Intermediate Organic

Chemistry











Michael S. Leonard

ii


iii








Copyright 2013 by Michael S. Leonard
All rights reserved. This book, or parts thereof, may not be
reproduced in any form without permission.


iv















For Laura Leigh

v

Table of Contents

Part I: Structure Solving............................................................................................................1

Chapter 1: Mass spectrometry ......................................................................................2

Chapter 2: Nuclear Magnetic Resonance spectroscopy ................................................27

Chapter 3: Infrared spectroscopy ..................................................................................67

Chapter 4: Structure elucidation ...................................................................................74

Part II: Aromatic Chemistry Expanded.....................................................................................108

Chapter 5: Aryl amines .................................................................................................110

Chapter 6: Aryl halides .................................................................................................127

Part III: Radical Reactions ........................................................................................................145

Chapter 7: Basic radical chemistry ...............................................................................146

Chapter 8: Advanced radical chemistry ........................................................................175

Part IV: Ionic Reactions ............................................................................................................192

Chapter 9: Synthesis of alkenes ....................................................................................193

Chapter 10: Synthesis and protection of alcohols .........................................................211

Chapter 11: Synthesis and protections of amines .........................................................228

Chapter 12: Synthesis and protection of aldehydes, ketones, and their derivatives .....246

Chapter 13: Synthesis and protection of carboxylic acids and their derivatives ..........263

Part V: Polymers .......................................................................................................................282

Chapter 14: Chain-growth polymers .............................................................................283

Chapter 15: Step-growth polymers ...............................................................................297

Solutions to End-of-Chapter Problems .....................................................................................312
1





Part I:
Structure Solving

2





Chapter 1:
Mass spectrometry

3

Ionization and analysis

Mass spectrometry is a technique for the identification of the molecular mass of an analyte.
However, the technique gives far more information as well. The name spectrometry conveys
both similarities to and differences from spectroscopic techniques, which use light to study
matter. In spectrometry, light is not used as the means to acquire information about the analyte;
however, the method generates a spectrum, analogous to those acquired through the
spectroscopic techniques such as NMR and IR.

Mass spectrometry can be achieved in multiple ways, and a complete discussion of the
instrumental methods is beyond the scope of this text. What follows is a brief overview of
classical methods and modern variations.

A classical method uses electron impact (EI) to generate the ions that are critical to the
technique. The sample is vaporized and then bombarded by a high energy beam of electrons.
Occasionally this bombardment will result in an electron being ejected from the analyte. What
results is called a radical cation because it has both an unpaired electron and a positive charge.
Assuming that, as is normal, all of the electrons were paired up in the original analyte, then the
loss of one electron leaves the molecule with a single unpaired electron somewhere in its
structure. A molecule with an unpaired electron is termed a radical. Additionally, if the original
molecule was neutral, then the loss of one electron leaves it with a positive charge.



The radical cation is a high-energy species, and as a result it is fairly unstable. Consequently,
this radical cation will frequently break down into smaller fragments, some of which will be
detected through mass spectrometry as well. There will often be multiple degradation pathways
possible, resulting in a wide variety of fragments. Later in the chapter, well see how the
analysis of these fragments can help us to determine molecular structure.



The charge of the radical cation serves as a means to manipulate it. The radical cation can be
accelerated using an electric field. Attraction toward an oppositely charge plate will draw the
radical cation through the mass spectrometer.
4




With large molecules, like proteins, the time that it takes the radical cation to traverse the
distance from one plate to another can be correlated to the mass of the radical cation. This
method of analysis is called time of flight (TOF). It is important to note that the mass of the
radical cation is essentially the same as the mass of the original analyte (M) since the two species
differ by only one electron, whose mass is negligible.

With smaller analytes, such as those molecules of interest to organic chemists, TOF analysis is
not as practical. Instead, a magnetic field is used to deflect the course of the radical cation as it
travels through the instrument. The extent of the deflection depends upon the ratio of the radical
cations mass to charge (m/z). For most small molecules, the charge will be +1, so the mass-to-
charge ratio essentially equates directly to mass.

If the magnetic field is constant, then a smaller radical cation will be deflected more than a larger
radical cation. In this example, the location at which the various radical cations strike the
detector reflects their masses.



5

Alternatively, if the magnetic field can be adjusted, then the instrument can scan for certain
mass-to-charge ratios at a particular point of impact.

A more compact method of sorting ions is called a quadrupole mass analyzer. This device
consists of two sets of parallel rods. The voltage applied across each set of parallel rods can be
adjusted so that only an analyte with the proper mass can travel between the rods. Other
molecules having different masses develop trajectories that lead them to collide with the rods. In
this fashion, it is possible to search for a single ion or scan for a range of ions.



It is worth noting that other methods of ionization exist and play a critical role in the analysis of
large molecules in particular. Until the 1980s, EI was the primary mode of ionization. However,
very large molecules, like biomolecules, are extremely susceptible to fragmentation under these
conditions. Electrospray ionization (ESI) was developed as a gentler method of ionization that
enabled the successful analysis of much larger molecules. In ESI, electrospray is used to
generate an aerosol of solvent containing the analyte. Small charged droplets are produced
during ionization, and as the solvent molecules are shed, charged analytes are released. The
development of ESI by Professor John B. Fenn of Virginia Commonwealth University led to his
2002 Nobel Prize in Chemistry.
1


Another widely used gentle, or soft, ionization technique is matrix-assisted laser
desorption/ionization (MALDI). In this method, the analyte is dispersed in a matrix of
molecules, such as cinnamic acid derivatives, that are typically acidic and absorb ultraviolet or
infrared radiation effectively. The exposure to a laser causes the matrix to absorb energy leading
to desorption and ionization through proton transfer. The charged analyte produced in this way
allows for mass spectral analysis.

Tandem techniques

Mass spectrometry is a very useful analytical tool, but the quality of the results is dependent
upon the quality of the sample. A mass spectrum of a mixture of analytes will understandably be
much more difficult to interpret than a mass spectrum of a single, pure substance. Consequently,
mass spectrometry (MS) is sometimes used in tandem with a separation technique, such as gas
chromatography (GC). In GC-MS, a mixture can be injected into the gas chromatograph and
separated into its constituent components. As each analyte elutes from the column, it passes into
a mass spectrometer for analysis. As a result, each peak in the GC has mass spectral data
associated with it. This allows for the rapid analysis of mixtures.

1
For additional information on the 2002 Nobel Prize in Chemistry, see:
http://www.nobelprize.org/nobel_prizes/chemistry/laureates/2002/
6


A simple mass spectrum

Lets consider a simple organic molecule, like methane (CH
4
). Its bombardment by a high
energy beam of electrons in electron impact ionization produces a radical cation.



Since the radical cation has only lost a single electron, its mass is essentially the same as that of
methane. That mass is 16 amu. When calculating mass for mass spectrometry, it is important to
remember that individual molecules are detected in this technique. Therefore, the masses of the
most commonly occurring isotopes are used to determine the most common molecular mass.
This stands in contrast to using average molecular mass, which we commonly do when consider
bulk samples, as in lab.



Right now, the difference appears to be insignificant because to the nearest whole number (and
even to the nearest tenth) both methods produce the same value. However, in subsequent
examples as the molecules become larger, those small differences will add up to create greater
disparities between the calculations.

On the mass spectrum of methane, we do indeed see a signal at m/z 16 corresponding to the
radical cation. This signal is often called the molecular ion peak since it corresponds simply to
the mass of the ionized molecule. This same signal also happens to be the most abundant one in
the mass spectrum and is therefore called the base peak. The relative abundance of the base peak
is arbitrarily set to 100%. It is coincidental in this instance that the molecular ion peak is also the
base peak. The two are not necessarily the same. In other words, the heaviest ion is not
necessarily the most abundant one, and we will see many cases where the molecular ion peak
and base peak are different signals.

7


In the mass spectrum of methane, peaks at m/z 15, 14, 13, and 12 are also evident. These
correspond to fragments of methane in which hydrogen atoms are successively lost.

Additionally, a very small peak is visible at m/z 17. At first glance, it would not seem possible
to have a peak with a mass greater than that of the molecular ion. This is possible due to the fact
that of 1.1% of carbon is
13
C, an isotope of
12
C with one additional neutron and therefore one
additional mass unit. The peak in the mass spectrum generated by
13
CH
4
is sometimes called the
M+1 peak, referring to the fact that it is one mass unit higher than the molecular ion peak (M).

The M+1 peak and the number of carbons in the molecule

The relative abundance of the M+1 signal can be used to predict the number of carbons in the
molecule, provided that carbon is the only element in the molecule with an isotope of significant
abundance that is one mass unit heavier. We can use a simple formula to determine the relative
abundance of the M+1 signal for a molecule containing n number of carbons. For each carbon,
there is a 1.1% chance of finding a
13
C in that location. Therefore, multiplying the number of
carbons by 0.011 will determine the chance of finding a
13
C somewhere in the molecule. If the
relative abundance of the molecular ion peak is multiplied by this factor, the abundance of the
M+1 peak is the result.

(Rcloti:c obunJoncc o H + 1) = n (u.u11) (Rcloti:c obunJoncc o H)

This equation can be rearranged if we are interesting in finding the number of carbon atoms in
the molecule.

n =
(Rcloti:c obunJoncc o H + 1)
(u.u11) (Rcloti:c obunJoncc o H)


8

As noted above, for this equation to hold true, it is important that carbon be the only element
with an isotope contributing to the M+1 peak. For instance, if nitrogen is present in the
molecule, the equation wont hold true due to the 0.4% abundance of
15
N, which is one mass unit
heavier than the predominant isotope
14
N.

Using the molecular ion peak to determine molecular formula possibilities

A simple application of mass spectrometry results is the prediction of reasonable molecular
formulas for an unknown substance. Consider an unknown substance whose mass spectrum
shows a molecular ion peak of m/z 180. The maximum number of carbons that such a substance
could contain is 15.

18u omu
12 omu pcr C
= 1S

However, the molecule probably doesnt contain 15 carbon atoms and 0 hydrogen atoms.
Instead, a much more reasonable molecular formula could be produced by subtracting one
carbon from the formula and replacing it with 12 hydrogen atoms to give C
14
H
12
. Such a
molecule would be highly unsaturated. Recall that degrees of unsaturation (DOU) are bonds or
rings. To calculate the degrees of unsaturation, the number of hydrogen atoms present in the
formula is subtracted from the number of hydrogen atoms that a given number of carbons could
hold (2n+2, where n = the number of carbons). This difference is then divided by two, since
hydrogens are removed in pairs to form bonds or rings.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2


In this instance, for C
14
H
12
there are nine degrees of unsaturation. While this is a highly
unsaturated molecule, such structures are possible. Two such examples follow.



Of course, other structures with the same formula could also be drawn, and it isnt possible to
determine which is the correct structure of the unknown substance without additional
information.

It is also important to note that C
14
H
12
is not the only viable molecular formula for this unknown.
Several other reasonable formulas can be derived from the original one. A carbon could be
replaced with 12 more hydrogens, giving C
13
H
24
, and reducing the amount of unsaturation to just
2 degrees. One of many possible structures with that formula follows.

9

C
13
H
24


Additionally, a CH
4
unit (mass 16 amu) can be removed from either formula and replaced with
an oxygen atom (mass 16 amu). This produces even more viable molecular formulas for the
unknown substance, as shown below. As you consider these examples, recall that the
introduction of oxygen atoms into the formulas has no impact on the calculation of degrees of
unsaturation (DOU).



The previous example of an unknown was a molecular ion peak at m/z 180 and its many viable
formulas suggests that a better method is needed to reduce the number of possible formulas.
This need is further highlighted by the current inability to distinguish between certain analytes.
A classic example is the comparison of carbon monoxide (CO), nitrogen (N
2
), and ethylene
(H
2
C=CH
2
). When measured to the nearest whole number, each of these three substances has a
mass of 28 amu. Low-resolution mass spectrometry (LRMS) provides mass spectral data to the
nearest whole number. However, high-resolution mass spectrometry (HRMS) provides several
more decimal places. The masses of the elements are measured relative to
12
C as the standard,
which has been assigned a mass of exactly 12.0000 amu. The HRMS results below show that,
while LRMS cannot distinguish these three analytes, HRMS can.



It is also the case that molecular ion masses measured using HRMS are likely to correspond to
only one (or at most a very few) possible formulas. For instance, a molecular ion at m/z
180.0939 would correspond to just one of the formulas we considered earlier as possible
identities of our unknown substance.

10



These calculations reveal that the unknown substance must have the molecular formula C
14
H
12
.

Chlorine, bromine, and their isotope patterns

Earlier, when we examined the mass spectrum of methane, we saw that isotopes can lead to the
present of peaks heavier than the molecular ion peak. For instance, we saw that the 1.1% of
methane molecules containing a
13
C would give rise to the M+1 peak. Carbon is not the only
element with an isotope having significant enough abundance to lead to an observable signal in
the mass spectrum. Two of the halogens also lead to distinctive isotope patterns.

Chlorine has two predominant isotopes,
35
Cl and
37
Cl, that are present in nearly a 3:1 ratio.
Consequently, a molecule containing chlorine will exhibit a molecular ion peak (M), as well as a
peak two mass units higher (M+2). The M+2 peak will be about one-third the height of the
molecular ion peak.

Bromine also has two isotopes,
79
Br and
81
Br, whose masses differ by two, leading to an M and
M+2 peak as well. However, for bromine these isotopes are present in nearly a 1:1 ratio,
meaning that the M and M+2 peak heights will be almost identical.

These distinctive isotope patterns are useful in identifying the present of a halogen. Prominent
M and M+2 peaks suggest the presence of a halogen, and the relative abundance of the two
signals reveals whether that halogen is chlorine or bromine.

Alkane fragmentation

With a very small molecule, like methane, the fragmentation options are quite limited as we saw
previously. However, with slightly larger molecules containing a larger number of bonds, the
options will be more varied, so it will help us to understand mass spectra if we make a systematic
study of the fragmentation possibilities.

Lets consider the ionization of pentane. As pentane is bombarded by a high energy beam of
electrons, an electron will be ejected from a few molecules, leading to their ionization. Carbon-
carbon bonds are typically weaker than carbon-hydrogen bonds. Bond dissociation energies
11

(BDE) highlight this fact. The typical bond dissociation energy of a carbon-carbon bond is 83
85 kcal/mole, while the typical BDE for a carbon-hydrogen bond is 96 99 kcal/mole. The
weaker carbon-carbon bond is more likely to experience the loss of an electron.

There are two types of carbon-carbon bonds in pentane, colored blue and red below, and an
electron could be ejected from either to provide one of two possible radical cations. These are
the molecular ions that will generate the molecular ion peak (M) at m/z 72.



One bond in each of these radical cations has been dramatically weakened by the loss of an
electron. Therefore, fragmentation is likely. The first radical cation, which lost a blue electron
from a terminal C-C bond, can fragment in one of two ways:

(1) the methyl group can retain the blue unpaired electron, leaving a butyl carbocation or
(2) the butyl group can retain the blue unpaired electron, leaving a methyl carbocation

Remember that the ions will be detected in mass spectrometry, while the neutral radicals will not
be directly observed.



The second radical cation, the one that lost an electron from the red interior C-C bond, can also
undergo fragmentation in an analogous fashion to yield a propyl or ethyl carbocation.



All of these signals at m/z 72, 57, 43, 29, and 15 can be seen in the mass spectrum, although the
signal at m/z 15 is miniscule and only fairly prominent signals are shown in the following
diagram. The stability of the fragments determines their relative abundance. In general, more
highly substituted carbocations and radicals will be more stable fragments.

12



Additionally, we see that some signals on the mass spectrum have not yet been explained. For
instance, there are prominent signals at m/z 42 and 41. These result from further fragmentation
of the propyl cation. Successive fragmentations are sometimes possible if they enhance the
stability of the molecule. A carbocation can be rendered more stable through resonance, and the
loss of two hydrogens from the propyl cation would introduce a bond, giving an allylic radical
that enjoys resonance stabilization.



Differentiation of isomers

Sometimes isomers will experience fragmentation patterns that reflect their structural
differences. 2-Methylbutane, an isomer of pentane, is an illustrative example. While the propyl
cation remains the dominant fragment (i.e. the base peak), the abundance of the butyl and ethyl
cation fragments have grown in intensity relative to it. This is also true of the methyl cation
signal at m/z 15, which is not shown in the spectrum below.

13



Examination of the structure and the relative stability of the possible fragments explains these
differences in spectra. 2-Methylbutane has three different types of carbon-carbon bonds, shown
in blue, red, and green below. Any of these bonds could conceivably fragment during mass
spectral analysis.



Notice that the fragmentation of a blue bond results in a secondary butyl carbocation (m/z 57).
When a butyl carbocation was produced from pentane, it was primary. The more stable
secondary butyl carbocation resulting from 2-methylbutane is therefore understandably greater in
relative abundance.

Fragmentation of the central red bond leads to a primary ethyl carbocation (m/z 29) and a
secondary radical. In contrast, when an ethyl carbocation was released from pentane, it was
accompanied by a primary propyl radical. The more stable radical formed in the fragmentation
of 2-methylbutane helps to explain the greater relative abundance of the m/z 29 signal.


14

Functional groups and their effect on fragmentation: alkyl halides

When we considered alkane fragmentation, the initial ionization resulted from the loss of an
electron from a sigma () bond. However, when heteroatom-containing functional groups are
present in a molecule, there are non-bonding electrons present (i.e. the lone pairs). These
electrons are less tightly held than -bonding electrons, and as a consequence, they are more
likely to be displaced during the ionization step. An alkyl halide provides an illustrative
example.



The molecular ion produced in this fashion can fragment in one of two ways. One option is
known as heterolytic cleavage. In this case, the adjacent sigma bond breaks with the electrons
flowing to the halogen as expected. The result is a neutral halogen radical and a carbocation.



Alternatively, -cleavage may occur. In this scenario, the halogens electron is used to form
of a bond. The other electron needed to complete the bond comes from the homolytic
cleavage of one of the bonds stemming from the -carbon. As this sigma bond fragments, a
radical is released. Since each of these events involves the movement of a single electron,
single-headed fish hook arrows are used to denote them. Also, note that the charged fragment
still contains the halogen and therefore still exhibits an isotope pattern.



-Cleavage is common for alkyl chlorides because the C-Cl bond has a bond dissociation energy
( 80 kcal/mole) close to that of C-C bonds ( 85 kcal/mole). For alkyl bromides, in which the
carbon to halogen bond is weaker ( 65 68 kcal/mole), heterolytic cleavage occurs more
frequently (through the dissociation of this weaker bond) than -cleavage, which would require
the cleavage of the significantly stronger C-C bond.

Functional groups and their effect on fragmentation: ethers

The behavior of ethers follows the same paradigms outlined above: heterolytic cleavage or -
cleavage. If we consider an unsymmetrical ether, it becomes clear that multiple fragmentations
will be possible using only these two mechanistic paradigms.
15


In the ionization step, we would again expect an electron to be lost from the less tightly held
non-bonding electrons to generate the molecular ion.



This radical cation could undergo two different heterolytic cleavage events.



We might expect the fragment containing a secondary carbocation to be more prominent than the
one containing a primary carbocation, but both fragments will likely be observed in the mass
spectrum.

By the same token, there are multiple -cleavage pathways. Two of these pathways (1 and 3,
below) release a methyl group and therefore result in fragments that, though they are structurally
distinct, have the same mass. The other pathway (2, below) releases an ethyl group and therefore
results in a fragment with a different mass.


16


Functional groups and their effect on fragmentation: alcohols

Alcohols present another instance of a heteroatom-containing functional group. They too lose an
electron from a lone pair during ionization. However, alcohols are unique in that they rarely
exhibit a prominent molecular ion peak in the mass spectrum. The electron-deficient oxygen is
unstable enough that fragmentation is very likely. While -cleavage is still possible, well see
that dehydration is another highly probable fragmentation.

Examining a particular alcohol, we see that ionization yields a radical cation as expected.



This radical cation rapidly fragments; however, heterolytic cleavage, which would result in a
hydroxy radical, is generally unfavorable. Instead, -cleavage may occur via one of two
pathways.



Alternatively, when the alcohol is sizable enough to have a hydrogen at a distance of five atoms
from the oxygen, dehydration may occur. This process begins with the abstraction of a hydrogen
atom from the -carbon. Another way of phrasing that is to say that the hydrogen that is 5 atoms
from the oxygen is abstracted. A note on terminology is in order here. We would not say that
the -carbon is deprotonated. Deprotonation refers to the removal of a proton (i.e. H
+
is
removed). In this case, a hydrogen atom (i.e. H) is removed, and so we say that a hydrogen
atom was abstracted.


17


This process has not yet resulted in any change in mass because the hydrogen atom was simply
transferred from one location to another within the same molecule. The significance of this step
is that the molecule now contains a good leaving group: water. The dissociation of water (or
dehydration) occurs at this stage.



The new radical cation is a fragment of the original molecular ion with a mass that is 18 amu
smaller because 18 amu is the mass of a water molecule. Consequently, this new radical cation
gives rise to what is sometimes called the M 18 peak. The entire two-step process is shown
together below. It is important to note that step 1 is homolytic (and therefore uses fish hook
arrows), while step 2 is heterolytic (and therefore uses regular mechanistic arrows).



In the analysis of alcohol mass spectra, it may not always be possible to see the molecular ion
peak (M), but it will usually be possible to see the dehydration fragment (M 18).

Functional groups and their effect on fragmentation: ketones

Ketones have unique fragmentation behavior, but there is some analogy to the behavior of
alcohols. For both functional groups -cleavage is possible. Also, both functional groups can
fragment in an alternative fashion that involves intramolecular hydrogen atom abstraction.

Ionization of ketones proceeds as expected, with the loss of an electron from a lone pair to
produce the molecular ion.



Fragmentation can occur via -cleavage in one of two pathways, given the asymmetry of the
molecule.

18



The CO species formed are acylium ions. These are resonance forms of carbonyl-containing
groups bearing an alkyl substituent (i.e. acyl groups), which happen to also be positive (hence the
ium suffix). You saw acylium ions during Introductory Organic Chemistry when you studied
Friedel-Crafts acylation (one of the Electrophilic Aromatic Substitution, or EAS, reactions).



An alternative fragmentation motif is known as the McLafferty rearrangement, and it involves
hydrogen abstraction from the -carbon. However, because of a difference in the use of Greek
lettering for alcohols and ketones, the -carbon is at a different distance from the oxygen atom.
With alcohols, the carbon bearing the functional group is the -carbon; whereas, with carbonyl-
containing compounds, it is the carbon adjacent to the one bearing oxygen that is labeled as the
-carbon. This means that for alcohols the -hydrogen is five atoms from the oxygen but for
ketones the -hydrogen is six atoms from the oxygen.



When alcohols dehydrate, the hydrogen five atoms from the oxygen is abstracted because atoms
that are separated by this distance are close in space to one another as the molecule rotates
through different conformations. When ketones undergo McLafferty rearrangement, the
hydrogen that is abstracted is slightly further away (six atoms) because of the larger bond angle
of the sp
2
-hybridized carbonyl carbon. This increase in bond angle from 109.5 to 120
separates the reactive oxygen from the closer hydrogen and renders its removal unlikely.

As the -hydrogen is abstracted (arrows labeled 1 in the figure below), the remaining electron
from the bond that is homolytically cleaved falls in between the - and -carbons (arrow labeled
2 in the figure below). The -carbon reciprocates through donation of an electron from the ,
19

bond (arrows labeled 3 in the figure below). This completes a bond between what used to be
the - and -carbons, and it also severs this olefin from the rest of the molecule



Using mass spectra data in problem solving

Now that we understand how molecules behave during mass spectrometry, it might be useful to
look at the information differently. In the research laboratory, you might not know the structure
of your analyte. Instead, you may acquire a mass spectrum of an unknown substance. Perhaps
you have some clues as to its structure, but you may not know all of the details. And, you may
be able to use mass spectrometry (mass spec) to clarify the analytes structure. Heres an
example.

Consider an analyte with the molecular formula C
7
H
14
O. It exhibits the following mass
spectrum.



Lets try to propose a structure that is consistent with this information.

Our first task should be to calculate degrees of unsaturation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2


20

cgrccs o unsoturotion =
|2(7) + 2] 14
2
= 1

The presence of an oxygen and a single degree of unsaturation suggests that a ketone is a
possible functional group. Of course, it is also possible that we have an alcohol or an ether with
a ring in the structure. Since the functional groups behave quite differently, the mass spectrum
should help us to identify which possibility is more probable.

Lets begin by considering a ketone. It would be challenging to look at the prominent peaks at
m/z 114, 85, 72, 57, etc. and devise structures for the fragments. Chemists usually prefer to look
at the difference in mass between the molecular ion peak and the other prominent peaks. These
mass differences will typically be smaller values that correspond to the loss of commonly
occurring fragments. Some examples are given below.

Mass Difference
(amu)
Group Lost
15 methyl radical (CH
3
)
29 ethyl radical (CH
3
CH
2
)
43 propyl radical (CH
3
CH
2
CH
2
) or the isomeric isopropyl radical [(CH
3
)
2
CH]
57 butyl radical (CH
3
CH
2
CH
2
CH
2
) or one of its isomers

Looking back at our mass spec through this lens, we see two mass differences (29 and 57 amu)
that correspond perfectly to the loss of ethyl and butyl radicals. The third mass difference does
not match up perfectly with the loss of a propyl radical, and there must be an explanation for this.



We know that ketones can undergo -cleavage, thereby releasing radical fragments.
Furthermore, we know that an unsymmetrical ketone will release two different radical fragments
from two unique -cleavage pathways. We can use this knowledge to reconstruct a possible
parent ketone from the fragments at hand. We know that butyl and ethyl radicals were released.
21

We also know that all ketones must contain a carbonyl. Lets simply combine those pieces to
obtain a possible ketone structure: 3-heptanone.



We still need to explain the loss of 42 amu via a third fragmentation. Ketones can undergo
McLafferty rearrangement. For 3-heptanone, McLafferty rearrangement would release
propylene (a loss of 42 amu), producing the fragment signal at m/z 72 for the oxonium ion.



Consequently, this is a structure that is consistent with the data at hand. Another structure would
also be consistent with the data. Remember that loss of 57 amu corresponds to the release of a
butyl radical or one of its isomers. Therefore, we could have constructed other ketones using
isobutyl, sec-butyl, or tert-butyl fragments. Only one of these would give fragments of the same
mass in McLafferty rearrangement though, and that is the ketone constructed using an isobutyl
fragment: 5-methyl-3-hexanone.



McLafferty rearrangement of 5-methyl-3-hexanone still releases propylene (for a loss of 42 amu)
and yields an oxonium ion with m/z 72.



Deciding between 3-heptanone and 5-methyl-3-hexanone is more subtle. For instance, we might
notice that 5-methyl-3-hexanone has more -hydrogens than 3-heptanone (6 vs. 2). Therefore, it
would seem as though the peak resulting from McLafferty rearrangement (m/z 72) would be
large if the analyte were 5-methyl-3-hexanone and small if it were 3-heptanone. Since the signal
at m/z 72 is fairly small, this suggests that the analyte is more likely 3-heptanone.
22

Chapter 1 Problems

1. Propyl guaiacol is a solvent that can be created by breaking the bonds in lignin, which is found
in great abundance in wood. This helps to address the issue of sustainability and has
implications for energy production as well (Chemical & Engineering News, June 11, 2012, p.
36). If the molecular ion peak for propyl guaiacol has a relative abundance of 18.6% and the
M+1 peak has a relative abundance of 2.05%, how many carbon atoms are present in propyl
guaiacol?

2. Propose two molecular formulas that are consistent with a molecular ion peak of m/z = 166.
One formula should include oxygen, and the other should not. For each formula, draw a
potential structure that corresponds to the formula.

3. Bosutinib is an anti-cancer agent that was at the center of some controversy (Chemical &
Engineering News, May 21, 2012, p. 34). There is concern that some investigators conducting
research many have used an isomer of the compound, thereby potentially calling the relevance of
their results into question.

In the mass spectrum, bosutinib exhibits the molecular ion peak (M), as well as prominent M+2
and M+4 peaks. Based on this clue, which of the following compounds could be bosutinib?

(a)


(b)
(c)


(d) None of these choices could be
bosutinib.





23

4. Describe a significant distinction between the mass spectra of these hydrocarbon isomers.



5. The mass spectra for the isomeric compounds 1-chlorobutane, 2-chlorobutane, and 2-chloro-2-
methylpropane are shown below. Match the compounds to their spectra based on differences in
their expected fragmentation patterns.



24





6. Predict the signals that will appear in the mass spectrum of the following ether.



7. For the following alcohol, show the mechanism for formation of the M-18 peak.


25


8. For the following ketone, provide the expected m/z ratio for the molecular ion peak, as well as
for the major fragments expected.




26

9. The following is the mass spectrum of a molecule with the formula C
7
H
14
O. Propose a
structure for this molecule that is consistent with the mass spectrum below.



10. A compound with the molecular formula C
6
H
11
Cl exhibits the following mass spectrum.
What is its probable structure?









27






Chapter 2:
Nuclear Magnetic
Resonance
Spectroscopy

28

A brief review of the fundamentals of NMR

Lets begin with a brief review of the fundamentals of Nuclear Magnetic Resonance (NMR) that
you learned about in Introductory Organic Chemistry. Certain nuclei (e.g.
1
H,
13
C) have a spin,
and this is sometimes designated by drawing an arrow through a circle. Ordinarily, these nuclear
spins are randomly oriented.



However, when placed in a magnetic field, the orientation of nuclear spins is no longer
haphazard. Instead, the spins align in one of two ways: with the external field or against the
external field. Many more nuclei will be aligned with the applied field, since this corresponds to
the lower energy state.



There is an energy difference (E) between these two states, and as you might expect, the energy
difference depends upon the strength of the applied field. If the applied field is miniscule, then
so is the difference in energy between the two spin states. However, as the field increases in
strength, the magnitude of the energy difference between states increases as well. This heightens
the probability of finding nuclei in the lower energy state (i.e. aligned with the applied field).



This phenomenon can be used to probe nuclei and to learn quite a bit about their environment.
The process begins with the excitation of nuclear spin. Radio waves (abbreviated as h since
29

that designates the energy carried by this electromagnetic radiation) are used to promote the
nuclei from the lower energy state (aligned with the applied field) to the higher energy state
(opposed to the applied field).



The nuclei are then allowed to relax back to their initial state. When they do so, radiofrequency
radiation is emitted. This radiation can be detected, giving us information about the sample.



At this point, there is one key piece of detail that is missing. If we are conducting proton NMR,
you might expect all protons (i.e.
1
H nuclei) to give off the exact same amount of energy, leading
to a single signal. This, however, is not the case. Protons in distinct chemical environments emit
unique amounts of energy that reflect the electron density around them. Electrons, being moving
charges, generate a magnetic field, and this magnetic field induced by the electrons opposes the
external field in most cases.

Lets consider two examples. In the first example, a nucleus is surrounded by a great deal of
electron density. This results in the induction of a prominent field opposed the applied field.
The difference between the applied and induced fields is the effective magnetic field, or the field
that the nucleus will actually feel. In this case, the effective field is small, so the energy
difference between spin states is small.



In a second scenario, a nucleus has little electron density in its immediate surroundings. As a
result, the induced magnetic field is small and diminishes the applied field to a much lesser
30

extent. In this case, the effective magnetic field is still quite strong, so the energy difference
between spin states is large.



The ultimate take home message from all of this discussion is that each nucleus in a unique
chemical environment will experience a unique effective magnetic field, resulting in a unique E
between the two spin states. Consequently, as each chemically distinct type of nucleus relaxes
during NMR, it gives off an amount of energy reflective of its surroundings. These emissions
translate into the signals observed in the NMR spectrum.

As we seek to plot those signals on an NMR spectrum, we would like to use an x-axis that is
directly comparable from one lab to another. As it stands, the E between spin states depends on
the effective magnetic field, which in turn depends on the strength of the applied magnetic field.
NMR spectrometers can have a wide range of magnetic field strengths. You may have heard
your instructor refer to the 60 MHz NMRs of the old days, with permanent magnets. Or perhaps,
your instructor has referred to a 300 or 400 MHz instrument in your department.

In order to report signals in such a way that we can make direct comparisons, regardless of the
field strength of our individual magnets, chemical shift () is used. Chemical shift measures the
ratio of the frequency difference between a signal of interest and that of an internal standard to
the operating frequency of the magnet.

o (ppm) =
Frcqucncy o o signol (Ez) Frcqucncy o intcrnol stonJorJ (Ez)
0pcroting rcqucncy o mognct (HEz)
1u
6


The internal standard is typically tetramethylsilane (TMS), which is the zero-point marker for
many NMR spectra. A small quantity of TMS is often introduced into NMR solvents, so that a
little bit is present in each NMR sample.



31

By calculating chemical shift in this way, the dependence upon field strength is removed, and we
can compare data even if you are using a 400 MHz instrument while I am using a 300 MHz
instrument.

As weve established, chemical shift tells us about the environment in which a nucleus resides.
For proton NMR, the correlation between chemical shift and proximity to functionality is shown
below. This list is not exhaustive by any means, but it covers a reasonable number of
possibilities.



Protons with chemical shift values closer to 0 ppm are referred to a shielded. This stems from
the fact that they experience a small effective magnetic field due to the large magnetic field
induced by the electrons around them. In other words, the high electron density they possess
shields them from the applied magnetic field. Protons with high chemical shift values are
referred to as deshielded since they lack this protection from the applied field and therefore
feel an effective field that is closer to the full magnitude of that which is applied.

32



Anisotropy

To this point, the shielding weve considered has been isotropic, which means the same in all
directions. However, clouds are not spherically symmetrical, and as a consequence, the field
induced by the electrons depends on the orientation of the protons relative to this cloud.
Benzene is probably the most commonly cited example of this phenomenon. For benzene, the
ring current caused by the electrons generates the magnetic field shown below. As you can see
from the diagram, whether the induced field opposes or reinforces the applied field depends on
the point in space under consideration.



However, it is the protons that interest us in proton NMR, and if we consider their location, the
induced field reinforces the applied field.


33


In benzene itself, all of the protons fall along the outer periphery of the ring, so they all
experience this same deshielding. But, if you look at the contour of the induced field, you can
see that above or inside the ring the induced field actually opposes the applied field. This seems
irrelevant because there are no protons in that location, but there are molecules where these
points in space could be occupied by protons. Consider a larger annulene. Annulenes are totally
conjugated rings. The name annulene is preceded by a number in brackets that tells the
number of carbons in the ring. So, [18]annulene is a totally conjugated, eighteen-carbon ring.



[18]Annulene illustrates how protons can be inside a ring, and cyclophanes illustrate how
protons can reside above a ring. Cyclophanes contain a benzene ring with a tether between the
meta or the para positions. In cyclophanes with smaller tethers, some protons will be held above
the ring, such as those stemming from the bond highlighted in purple in the following diagram.



These hydrogens could be expected to exhibit chemical shifts relatively close to 0 ppm.

Alkenes and alkynes also have pi clouds which behave in an analogous fashion. In alkenes,
deshielding occurs for protons connected to the double bond (i.e. vinyl protons); whereas,
protons held above the alkene could be shielded. On the other hand, the orientation of the
current for alkynes is such that protons directly connected to the triple bond are more shielded
than you would expect ( 2 3 ppm).

Coupling

Well begin this section with a review of the basics of coupling (formally spin-spin coupling),
which results in splitting. By this we simply mean that a signal can be split into more than one
spike, so signals can have complex shapes. Youll recall from your Introductory Organic
Chemistry course that the general idea of splitting is that the effective field experienced by a
proton depends to a small extent on the orientation of its neighboring protons spins.

To review this briefly, lets begin by considering a proton with only a single neighbor. The
squiggly lines in the drawing below simply mean that we are cutting out a fragment of a
molecule and ignoring the rest of its structure. The blue proton has one (red) neighbor. This red
neighbor could have its nuclear spin oriented either up or down (i.e. with or against the applied
field). These two possibilities place the blue hydrogen into one of two slightly different energy
states, so its signal is split into two spikes, which we call a doublet. The extent of the splitting
34

(i.e. the separation between the two spikes of the doublet) is described by a coupling constant, or
J value, which is typically expressed in Hz. J values usually range from 0 to 20 Hz.



Splitting is reciprocal, so if the red proton splits the blue proton into a doublet, the blue proton
will also split the red proton into a doublet with the same J value.



These diagrams are sometimes referred to as splitting trees.

It is important to note though that chemically equivalent protons do not split each other. For
example, the average methyl group (-CH
3
) consists of three chemically equivalent hydrogens.
While they may be split by neighbors, they do not split each other.

When a proton has two neighbors, there are two broad scenarios. These two neighbors may be
equivalent to one another or distinct. Lets consider both eventualities. First, lets examine a
fragment of a molecule (something we could also call a spin system) in which one proton (blue)
has two equivalent neighbors (red). Each of those neighboring protons could be in the spin up or
spin down state. Consequently, when all of the possible combinations are considered, there are
three states for the blue hydrogen, meaning that it will absorb energy at three unique frequencies,
all of which are in close proximity to one another.



This same concept can be expressed using a splitting tree. The top half of the splitting tree is
exactly the same as we saw previously. The blue protons signal is split into two peaks by a red
neighbor. Then, each of those new peaks is split into two by the second red neighbor. Since the
red neighboring hydrogens are equivalent, the J values for the two splitting events are equivalent.
This means that, during the second splitting, the central branches of the tree meet and overlap,
hence the doubling of intensity for that central peak.

35



The three peaks that result give the classic triplet, in which the intensity of peaks is 1 : 2 : 1.

If the two neighbors are inequivalent, the splitting tree looks a bit different. In the spin system
below, well assume that some asymmetry in the portion of the molecule not shown causes the
red and green neighbors to be chemically inequivalent. If that is the case, then each of these
neighbors splits the blue hydrogen with its own unique J value. As a result, the central branches
of the tree may well not meet, leading to four peaks of equal intensity. This is known as a
doublet of doublets.



As it was shown here, the smaller coupling constant was associated with the second splitting
event. The same doublet of doublets results regardless of the order in which we consider the
splitting events.



Of course, it is not uncommon for a proton to have three neighbors. Being adjacent to a methyl
group will provide three chemically equivalent neighbors. The upper two thirds of the splitting
tree matches what we drew for the triplet perfectly. We simply add one more splitting event for
the third red neighbor. Every time that branches meet, the intensities of the preceding signals are
additive. It follows then that the last splitting event yields the classic quartet shape with four
signals having intensities of 1 : 3 : 3 : 1.
36




Of course, if the three neighbors are inequivalent, a much more complex splitting pattern will
result. It is possible that two of the three neighbors could be equivalent.

H
H
H
H
J J
Each neighbor
may be spin up
or down
J
J J
J


In this case, the splitting pattern could be called a triplet of doublets or a double of triplets. The
name depends on your perspective.



Alternatively, it is conceivable that all three neighbors could be chemically inequivalent, in
which case the end result would be a quartet of doublets.


37



There are certainly more patterns that could be envisioned than we have covered here, but this
overview of splitting will enable us to predict those patterns in new scenarios.

The magnitude of coupling constants

In order for nuclear spins to couple, the nuclei typically need to be in relatively close proximity.
Most of the time, we expect nuclei to couple if they are separated by two or three bonds. The
typical neighbor that you will have considered in your Introductory Organic Chemistry course
is a vicinal hydrogen. Vicinal hydrogens are separated by three bonds.



In unconstrained systems, the vicinal J value will usually be in the range of 6 8 Hz. It is worth
noting that, in some systems, rotation about bonds can be constrained. Conformational
constraint can impact the observed vicinal coupling constant. The Karplus-Conroy equation
allows prediction of coupling constants in these situations. One formulation of the Karplus-
Conroy equation follows.

Iicinol [ = o cos
2
1 u.28 (o = 8.S or 1 9u; o = 9.S or 1 > 9u)

The angle is the dihedral angle between the vicinal hydrogens. Remember that the dihedral
angel is the angle between vicinal hydrogens when looking down the axis of the carbon-carbon
bond between them (green in the following diagram). A dihedral angle of 0 would mean that
the hydrogens overlap when viewed from this perspective; whereas, a dihedral angle of 180
would mean that they are on opposite sides of the central carbon-carbon axis.

38



Plotting the predicted vicinal coupling constants for every yields the following curve, showing
the sharp dependence of J on dihedral angle. In a molecule where the dihedral angle is
constrained near 90, you may not observe any vicinal coupling, but in a molecule where the
dihedral angle is constrained near 0 or 180, the J value could be sizable.



Usually, the focus is on the sort of vicinal coupling weve discussed thus far. However, there are
certainly other types of coupling observed. One small permutation is the coupling of cis- and
trans-vicinal hydrogens of an alkene.

-1
1
3
5
7
9
0 20 40 60 80 100 120 140 160 180
V
i
c
i
n
a
l

C
o
u
p
l
i
n
g

(
H
z
)

39



Geminal coupling refers to two-bond separation between the nuclei. This can occur in more than
one situation. One example would be in an olefin, where the geminal coupling is relatively
small. In order to observe this coupling, the two substituents on the alkene must differ in some
way.



A second example would be two hydrogens stemming from the same sp
3
-hybridized carbon.
These protons must be diastereotopic (i.e. different even in achiral environments) in order for
geminal coupling to take place. Diastereotopic is a term that we will define more fully later in
this chapter.



Aromatic rings also exhibit vicinal as well as long-range coupling. The magnitude of the
coupling diminishes with increasing distance between the protons. Ortho coupling is the largest,
while para may even be too small to observe.




40

Chemical equivalence vs. magnetic equivalence

Our analysis of a spin system can be further complicated by magnetic equivalence, or lack
thereof. To understand this issue of magnetic equivalence, lets first revisit chemical
equivalence. Chemical equivalence describes protons that exist in identical chemical
environments. As a result, they contribute to a single signal. A methyl group is a convenient
example of chemical equivalence. Typically, there is free rotation about the bond connecting
the methyl group to the rest of the molecule, so each of the three hydrogens can occupy any point
in space.



Magnetic equivalence is a more stringent comparison between protons. To be magnetically
equivalent, the protons must not only exhibit the same chemical shift (i.e. be chemically
equivalent) but they must also couple equally with other chemically equivalent protons. Another
way of expressing couple equally is to say that the J values must be the same for coupling with
other chemically equivalent protons.

Lets take a look at some examples. A tert-butyl group contains three methyl groups that are
likely to be both chemically and magnetically equivalent. Under normal circumstances, there
will be free rotation about the bond linking the tert-butyl group to the rest of the molecule.
The color coding here parallels that which was used for the hydrogens in the methyl group above
to illustrate that the methyls of the tert-butyl group will interconvert in the same way. This
illustrates that the methyls are chemically equivalent. Furthermore, these methyl groups must be
equidistant from any hydrogens in the rest of the molecule, meaning that they will couple
identically to those hydrogens.



For instance, the simplest substituent we can insert to complete a molecule with the tert-butyl
fragments is H. Doing so gives 2-methylpropane.



41

When we replace the squiggly line with an H, we see that the hydrogens of the methyl groups are
all the exact same distance from the black hydrogen. There are three bonds separating the black
hydrogen from any other hydrogen in the molecule. Therefore, we expect all of the J values to
be identical and to have a magnitude consistent with vicinal coupling ( 7 Hz).

To see an example with magnetically inequivalent hydrogens, lets consider some of the protons
of cyclopentene. The vinylic hydrogens (red) are chemically equivalent and cause a single
signal. Similarly, the neighboring allylic hydrogens (blue) are chemically equivalent and cause
one signal. Now, lets turn our attention to the coupling between the allylic hydrogens and one
of the vinylic hydrogens. Notice that one set of allylic hydrogens is close to the vinylic hydrogen
weve isolated. These neighboring allylic hydrogens will split the vinylic hydrogen with a
coupling constant somewhere between 3 11 Hz depending on the dihedral angle. However, the
other set of allylic hydrogens are further from the vinyl hydrogen under consideration. We
expect the four-bond separation to lead to little or no coupling (J 0 Hz).



We describe this situation by saying that that blue allylic hydrogens are chemically equivalent
but magnetically inequivalent. This is an instance where magnetic inequivalence simplifies the
NMR spectrum. Since the second J value is zero, we see no additional splitting of the vinylic
hydrogen due to the distal allylic hydrogens.

We can arrive at the same conclusion about the red vinylic hydrogens by considering their
interaction with a single set of allylic hydrogens. Here again, one relationship is three-bond
separation, while the other is four-bond separation. We expect the three-bond separation to
result in observable coupling, but we do not expect to see coupling from the four-bond
separation. We can therefore say that the red vinylic hydrogens are also chemically equivalent
but magnetically inequivalent.



In the previous example, magnetic inequivalence was a factor in simplifying the proton NMR
spectrum. However, that is not always the case. A molecule containing an aromatic ring can
provide an illustrative example of magnetic inequivalence that complicates the
1
H NMR
spectrum. Catechol is shown below. It has two sets of chemically equivalent aromatic protons
(red and blue). It quickly becomes evident that both sets are magnetically inequivalent when we
consider their coupling. If we examine the coupling between a single red hydrogen and both of
the blue hydrogens, we see that both ortho (J = 6 9 Hz) and meta (J = 1 3 Hz) splitting will
42

result. As in the previous example, one separation is by three bonds; whereas, the other is by
four bonds. The difference is that aromatic rings exhibit long-range splitting, so this time we see
the four-bond coupling as well as the three-bond coupling.



This time, the magnetic inequivalence complicates the NMR spectrum. The reason is that the
red hydrogen will be split by both its ortho and its meta neighbors, so its splitting pattern will be
more complex due to that long-range coupling with the meta neighbor

If we reexamine the molecule to consider the interaction between a single blue hydrogen and
both of the red hydrogens, we again see both ortho and meta splitting at play.



Defining spin systems

The concepts of chemical and magnetic equivalence allow us to assign simple abbreviations for
commonly occurring types of spin systems. There are only a few rules to follow for this process:

Spin systems are groups of protons that couple with each other.

A letter is used for each type of proton in the spin system.

If there is more than one proton of a certain type (i.e. chemically equivalent), a subscript
number tells how many there are. For example, a methyl group (-CH
3
) could be designated A
3
.
Two chemically (and magnetically) equivalent methyl groups could be designated A
6
.

If protons are chemically, but not magnetically, equivalent then a prime () is used to designate
the magnetic inequivalence. For instance, two chemically equivalent but magnetically
inequivalent methyl groups could be designated as A
3
and A
3
.

When selecting letters for different types of protons in the spin system, the separation between
the letters in the alphabet should roughly correspond to the chemical shift separation between the
signals. For instance, very different protons might be called types A and X. Protons that are a
43

little more similar might be termed A and M. Protons that are really similar in chemical shift
might be called A and B.

Lets reexamine some molecules that weve already discussed to demonstrate the method. We
saw 2-methylpropane earlier, but now well define its spin system. The methyl (CH
3
) and
methine (CH) hydrogens will certainly couple with one another due to their three-bond
separation. There is only one methine hydrogen in the molecule, so it could be assigned the
letter A. This choice is somewhat arbitrary because we certainly could have assigned it a letter
from the middle or end of the alphabet. There are nine methyl hydrogens in the molecule. These
hydrogens are all chemically and magnetically equivalent, as we determined previously, so they
will be assigned a single letter with the subscript 9. You do have some discretion in the choice
of the letter because the extent of the difference between chemical shifts will be interpreted
slightly differently by different individuals. You could conclude that methine and methyl
hydrogens will have somewhat distinct chemical shifts but that the difference wont be
tremendous. This is a fair conclusion and would lead you to choose a letter from the middle of
the alphabet for the methyl hydrogens, perhaps M.



Alternatively, you might have concluded that, given the absence of functionality in the molecule,
the methine and methyl hydrogens will have pretty similar chemical shifts, in which case you
might have chosen a letter from early in the alphabet for the methyl hydrogens.

Also, its worth noting that you shouldnt worry about which protons are assigned the A and
which get the M. You could have just as easily called this an A
9
M spin system. What matters is
not the specific letters or their order. Instead, the meaning comes from communicating that there
are two similar types of protons in the molecule, one having a single hydrogen and the other
having nine hydrogens.

We also considered part of cyclopentene earlier. Lets reexamine it to determine the spin
system. We previously concluded that the vinylic hydrogens (red) were chemically equivalent
but magnetically inequivalent. They are assigned one letter, A, due to their chemical
equivalence, but to express their magnetic inequivalence, well call them A and A. The allylic
hydrogens (blue) were also determined to be chemically, but not magnetically, equivalent.
Therefore, they too are assigned one letter, M, and well use M and M to denote the magnetic
inequivalence. Subscripts of 2 denote the two hydrogens in each allylic set. Finally, there is one
more type of hydrogen in the molecule (green) that we had not previously considered. The green
hydrogens are clearly another unique type of hydrogen since they are neither vinylic nor allylic,
so they need a letter from a different portion of the alphabet, X. There are two hydrogens in this
set of chemically and magnetically equivalent hydrogens, so theyll be denotes as X
2
.

44



Finally, we had previously considered catechol and determined that its two sets of aromatic
hydrogens are each chemical equivalent but magnetically inequivalent. The hydrogens ortho to
the hydroxyl groups are assigned one letter, A, and the prime is used to convey the magnetic
inequivalence. The hydrogens meta to the hydroxyl groups will be assigned a different letter, B,
and again the prime signifies their magnetic inequivalence.



This is another instance where judgment is involved in selecting appropriate letters. Would it be
unreasonable if you had called this spin system AAMM? Depending on the extent of your
experience with NMR, this may also be a totally reasonable conclusion. Youre still
communicating that there are two types of hydrogens in the molecule, and that they differ
(although not to a tremendous extent). It takes some experience to know that the aromatic
hydrogens in this molecule will have chemical shifts that are quite similar.

Methylene (CH
2
) protons and magnetic inequivalence

To this point, weve assumed that hydrogens stemming from the same carbon are identical. This
is often, but not always, true. For a methyl group, the three hydrogens will be identical unless its
rotation is somehow restricted. This is quite rare. However, the two hydrogens of a methylene
group (-CH
2
-) are commonly distinct from one another.

Methylene hydrogens can be placed into one of three categories, which will tell us about their
spectroscopic behavior. These categories are: homotopic, enantiotopic, and diastereotopic. In
order to place methylene hydrogens into one of these categories, we have to consider the
symmetry of the molecule, and you can do this in one of two ways. The first method is purely a
consideration of internal symmetry. The second method involves isotopic substitution of the
methylene hydrogens. Lets examine each method for the three categories of methylene protons.

Homotopic methylene hydrogens are completely identical to one another in all respects and in all
environments. They are totally indistinguishable, and this extends to spectroscopy. Homotopic
methylene protons contribute to the same signal in NMR under all conditions. Methylene
protons are homotopic when there is a rotational symmetry axis within the molecule. By this, we
45

mean that there is an axis about which you can rotate to interconvert the two hydrogens without
making any changes to the molecule.



In this generic example, the 180 rotation has switched the location of the red and blue protons.
When the R groups are identical, the structures before and after rotation are superimposable.



A specific example can be seen with the methylene protons of propane. Rotation about the
internal axis reverses the locations of the highlighted protons, but the two structures are
superimposable.



Some people have a difficult time visualizing the rotational axis of symmetry. There is an
alternative method to identify homotopic methylene protons. Draw the original structure twice,
replacing each proton in turn with deuterium. Then, compare the two isotopically labeled
structures. If they are identical (i.e. if they can be superimposed), then the methylene protons are
homotopic.



With our specific example, propane, we can clearly see that the sequential isotopic substitution
results in identical compounds, meaning that propanes methylene hydrogens are homotopic.
46




Methylene protons may be enantiotopic instead. Enantiotopic protons contribute to the same
signal in NMR under normal circumstances. So, ordinarily they are spectroscopically
indistinguishable. However, in a chiral environment, enantiotopic protons will yield different
signals in the NMR spectrum. This chiral environment can be achieved through the addition of a
chiral shift agent to the NMR sample.

Enantiotopic protons are found when there is no rotational axis of symmetry in the molecule but
there is an internal plane of symmetry. If reflection through this plane interconverts the
methylene hydrogens and makes no changes in the structure of the molecule, then those protons
are enantiotopic.



In this generic example, the plane of symmetry cuts right through the R-C-R backbone and
passes between the blue and red hydrogens. Reflection through that plane interconverts the
position of the red and blue hydrogens but makes no change in the structure of the molecule.
When this is the case, the methylene protons are enantiotopic.



Ethanol provides a specific illustration of this principle. Reflection through the plane of
symmetry cutting between the red and blue protons and passing directly through the H
3
C-C-OH
backbone interconverts the colored hydrogens without altering the structure of the molecule.

47



Again, some people find symmetry planes to be challenging to visualize. If that is the case for
you, the isotopic-substitution method works equally well for the identification of enantiotopic
hydrogens. The structure is drawn twice and each hydrogen in turn is replaced with deuterium.
If the resulting compounds are enantiomers, then the methylene protons are enantiotopic.



Our specific example, ethanol, illustrates this method. Here, sequential isotopic substitution
produces compounds that are enantiomers, meaning that the methylene protons are enantiotopic.



The final possibility is that methylene protons could be diastereotopic. Diastereotopic protons
cause their own unique signals despite the fact that they stem from the same carbon. This is a
very important moment to take a step back and review the big picture. Very often we have
assumed that hydrogens bonded to a single carbon would give a single signal. We are now
seeing that this is not necessarily the case. In some molecules, hydrogens bonded to the same
carbon can cause different signals.

Diastereotopic protons can be identified in two ways as well. If there is no rotational axis of
symmetry and no internal plane of symmetry, then the protons are diastereotopic. In generic
structures we could see this by envisioning an R group containing chirality, which well call R*.
Rotation about an axis through the methylene carbon yields a structure that cannot be directly
superimposed with the first because R* and R do not line up.

48



Additionally, this generic structure does not contain an internal symmetry plane. If reflection
were attempted through a plane cutting along the R*-C-R backbone, the chirality in R* would be
inverted from (R) to (S), or vice versa.



A specific instance of such a molecule is (S)-2-butanol. Rotation about an axis through the
methylene carbon yields a structure that cannot be directly superimposed on the first because the
methyl and alcohol-containing substituents no longer line up.



Similarly, we are unable to find an internal symmetry plane in (S)-2-butanol. Reflection through
the plane dividing the red and blue hydrogens does interconvert them; however, the chirality of
the alcohol has been inverted in the process. Therefore, the resultant structure is not
superimposable on the original one.



The isotopic substitution method provides an alternative way to identify diastereotopic
hydrogens. Again, the structure is drawn twice and each methylene hydrogen is replaced in turn
with deuterium. When the structure contains a pre-existing chiral center, the resulting structures,
which now bear a second chiral center, are diastereomers. If the structures formed by isotopic
substitution at the methylene group are diastereomers, then the methylene protons are
diastereotopic.

49



Continuing with our specific example, (S)-2-butanol, sequential isotopic substitution yields
diastereomers, revealing that the methylene hydrogens are diastereotopic.

OH
H H
Replace each in turn with deuterium
These compounds are
diastereomers.
OH
D H
OH
H D


The ramifications of homotopic, enantiotopic, and diastereotopic hydrogens are observed in
NMR spectra rather commonly. For example, if we compare dimethyl succinate to dimethyl (S)-
malate, well see that their methylene hydrogens are of different types, and this will affect their
spectra in a pronounced way.



Dimethyl succinate possesses an internal plane of symmetry that can be used to interconvert the
red and blue hydrogens via reflection. This reveals that the methylene protons are enantiotopic.
Alternatively, if we were to conduct sequential isotopic substitution, the compounds generated in
that fashion would be enantiomers, leading us to the same conclusion. Consequently, we expect
the blue and red methylene protons to contribute to a single signal. We can see this in the
following
1
H NMR prediction.

50



In fact, we see that this compound only exhibits a total of two signals due to the fact that the two
methyl groups are equivalent and the two methylene groups are equivalent as well.

However, the methylene protons of dimethyl (S)-malate are another matter. Dimethyl (S)-malate
does not exhibit rotational symmetry about an axis through the methylene carbon, nor does it
possess an internal plane of symmetry. Consequently, its methylene protons are diastereotopic.
We could also have arrived at this answer by citing the fact that sequential isotopic substitution
of the methylene protons yields diastereomers. As a result, we expect the methylene protons to
cause their own unique signals. This stands to reason from a common-sense perspective. In the
conformation shown below, the blue hydrogen is closer to the hydroxyl group than the red
hydrogen, so they exist in different chemical environments. Granted, many conformations about
the central carbon-carbon bond are possible, but in any of them, the red and blue protons will
reside in different locations relative to the adjacent hydroxyl group. This serves to differentiate
them spectroscopically.



Furthermore, since the methylene protons are inequivalent, they will split one another. The red
proton will be split by its vicinal (methine, CH) neighbor and by its geminal blue neighbor into a
doublet of doublets. The blue proton will exhibit a similar coupling. The following
1
H NMR
prediction illustrates this. There are two doublets of doublets between 2.5 and 3.0 ppm. The
alcohol proton is a broad signal that happens to fall in between the two doublets of doublets in
this prediction.

51



Two other features of this spectrum deserve comment. First, the methine (CH) signal is an
apparent triplet. Formally, its two neighbors (the red and blue methylene protons) are different
and should split the methine hydrogen with different J values, leading to a doublet of doublets.
However, the red and blue neighbors are not that different from one another, so their J values are
not that different in magnitude. As a result, accidental overlap of the central peaks can result in
something that looks very much like a triplet. This is sometimes referred to as an apparent
triplet.



The second feature of the NMR prediction that warrants comment is the single peak for the two
methyl groups. The methyl groups are formally different from one another because one ester is
closer to the hydroxyl group than the other. As a result, the two methyl groups should produce
two singlets, and they do in the actual spectrum. This serves as a reminder that computer
predictions of spectra are quite good but not necessarily perfect.

Second-order coupling

Second-order coupling is a phenomenon that is also termed strong coupling or roofing. As the
difference in chemical shift between two coupled signals decreases, the inside peaks grow in
intensity while the outer peaks diminish in intensity. An example using two doublets is
52

presented graphically below. When the chemical shift difference is large relative to the coupling
constant (i.e.
6 (Hz)
] (Hz)
1u), the doublets appear exactly as you would expect. Each doublet
contains two peaks of equal height. As the chemical shift difference diminishes relative to the J
value, the interior peaks grow larger while the exterior peaks grow smaller. There are still two
doublets, but they no longer appear as we would expect since there is a height difference
between the two peaks of each doublet. These are sometimes called AB quartets, pulling from
our spin-system nomenclature.



This phenomenon can be useful. In a spectrum with multiple doublets, roofing can help to
illustrate which pairs reside within a single spin system. However, this phenomenon can also be
confusing since you may not know whether a signal is a true quartet or an AB quartet. This is
one of the reasons that strong magnets are helpful. As the magnetic field increases in strength,
the chemical shift difference between two signals will grow more pronounced; however, the
coupling constant is an intrinsic property and does not change. Therefore, the ratio
6 (Hz)
] (Hz)
will
increase as the field strength increases, and roofing will disappear.

Carbon-13 NMR

Carbon-13 is another NMR-active nucleus, but
13
C NMR faces some unique challenges. First,
there is a relatively low abundance of the NMR-active isotope. This was not the case with
1
H
NMR, in which the vast majority of hydrogen atoms are the NMR-active isotope. Carbon-13
comprises only about 1.1% of any sample of carbon. This, in combination with other factors,
typically lengthens
13
C NMR acquisition times.

Another challenge is the coupling of
13
C nuclei to their protons. This can give carbon-13 signals
with splitting patterns that, when they overlap one another, complicate the spectrum
dramatically. For this reason,
13
C NMR spectra are often proton decoupled. This simply means
that the protons are irradiated during the acquisition to keep them in the excited spin state. If the
protons exist in only one spin state during this time, then coupling, which depends on a neighbor
being spin up or spin down, does not occur.

Proton-decoupled
13
C spectra show all carbon signals as singlets, which greatly simplifies the
analysis. In this process, information was of course lost. For the sake of the obtaining readily
discernible signals, we have sacrificed the ability to determine how many protons exist in close
proximity to each carbon. If that information is needed,
13
C spectra can be acquired with
coupling; however, that may not necessarily be desirable since splitting may be observed due to
protons directly bonded to a given carbon, as well as those adjacent to that carbon (i.e. H-C-C).
53

A technique known as off-resonance decoupling will reveal only the splitting due to direct
carbon-to-hydrogen connectivity (i.e. H-C). Well see later in this chapter that there are other
NMR experiments that can reveal this same information in a more straightforward fashion.

The chemical shift range for
13
C NMR is much larger than what we observed in
1
H NMR.
13
C
spectra typically range from 0 to over 200 ppm. One convenient ramification of this is that
carbon signals rarely overlap. If two carbons contribute to a single signal, it is typically because
they are in chemically equivalent environments and are therefore indistinguishable. Accidental
overlap of similar but distinct carbons is rare. Carbons involved in a double bond appear above
100 ppm. They will be in the range of 100 150 ppm if they are alkene or aromatic carbons and
will appear over 160 ppm if they are carbonyl carbons. Most other types of carbons appear
below 100 ppm. This typically means sp
3
-hybridized carbons, although it is worth noting that
alkyne carbons (sp hybridized) appear from about 65 90 ppm. This list is not completely
inclusive, but provides enough categories for most situations.



When
13
C NMR spectra are proton decoupled, integration is rendered unreliable. Due to this
phenomenon and the fact that accidental overlap of carbon signals is rare, most carbon signals
are simply assumed to represent one carbon. When chemically equivalent carbons are present
within a molecule, you may sometimes see carbon signals that are much taller than the others.
This is a clue to the greater integration of those signals, but a specific integration value is still not
typically calculated. We can see an example of this in the predicted spectrum for tert-
butylbenzene. This molecule contains 10 carbons but only 6 different types of carbons.


54


Just above 30 ppm in the predicted spectrum, we see two signals for the quaternary carbon (b) of
the tert-butyl group as well as the three methyl groups (a). One signal is clearly larger than the
other, and this is a clue to the fact that the larger signal likely represents the three equivalent
carbons of the tert-butyl group. Similarly in the aromatic region, two signals are clearly larger
than the others, suggesting that they are due to the two sets of equivalent methine carbons
(labeled d and e). In the actual NMR spectrum, these height differences are even more
pronounced, reinforcing the fact that attempting to extract exact integrations from proton-
decoupled
13
C spectra is not feasible.



In general, the interpretation of carbon-13 NMR spectra is simpler than the interpretation of
proton NMR spectra because there are fewer pieces of information. In
13
C NMR, our primary
focus is the number of signals and their chemical shift. Splitting information has been erased
through proton decoupling, and this has also made integration unreliable. It is important to
realize that, while less information may initially seem appealing, it presents a challenging when
you are doing real research in the laboratory. If you are trying to determine the structure of a
newly isolated natural product or an unexpected reaction product, you benefit from having as
much information at your disposal as possible.

DEPT and APT spectra

DEPT and APT spectra are
13
C NMR experiments that help us to reclaim some of the
information lost through proton decoupling. It is desirable to use the
13
C spectrum to determine
the number of protons bonded to each carbon. One way of doing this is to run an off-resonance
decoupled spectrum, but the splitting can cause nearby peaks to overlap, resulting in complex
signals that are difficult to interpret. DEPT and APT spectra provide the same information in an
easier to interpret format.

DEPT (Distortionless Enhancement by Polarization Transfer) exploits the proton pulsing angle
to distinguish methine (CH), methylene (CH
2
), and methyl (CH
3
) carbons. The data is typically
presented as a series of stacked
13
C NMR plots. One plot will show methine and methyl signals
55

pointed up but methylene signals pointed down. A second plot will show only methine signals.
A final plot will show all protonated carbons.

An example of a predicted DEPT spectrum for sec-butylbenzene is given below. Plot I shows all
methine and methyl signals up, while methylenes are down. This immediately allows us to
identify the carbon labeled (c) as the only methylene in the molecule. Plot II shows only the
methine signals. The methylenes and methyls are suppressed. This allows us to identify carbon
(b) as the sole methine in the sec-butyl portion of the molecule. By process of elimination, we
can now return to Plot I and label the methyl groups (a) and (d). We can surmise that methyl (d)
would be more shielded since it is further from the electronegative sp
2
carbon of the benzene
ring.



Plot II also showed us the three types of methines of the aromatic ring. Based on relative
intensities, we can assign carbon (h) to the smallest signal. The remaining two aromatic
56

methines would have to be (f) and (g), although determining which is which would be difficult
due to their similarity. Notice that carbon (e) did not appear in any of the spectra since it is not
protonated. This is an idiosyncrasy of DEPT spectra. Quaternary carbons do not appear at all.
However, you can compare your DEPT spectrum to a normal proton-decoupled
13
C spectrum to
determine which carbons are quaternary by difference. The carbons that appear in your normal
13
C spectrum but not in the DEPT are quaternary.

An APT (Attached Proton Test) spectrum is similar. It provides a plot in which quaternary
carbon and methylene signals point down, while methine and methyl signals point up. An APT
spectrum for sec-butylbenzene is shown below above the full
13
C NMR spectrum. The
quaternary carbon (e) and the methylene carbon (c) point down, while the remaining carbons
(which are all methines and methyls) point up. This method has the advantage of including
quaternary carbons; however, it has the disadvantage of not differentiating CH from CH
3
or C
from CH
2
.



Two-dimensional NMR spectroscopy: homonuclear experiments

As informative as the previously discussed one-dimensional NMR experiments are, there are
times when additional information is needed to determine the structure of complex molecules.
Two-dimensional (2D) NMR spectra can provide that information. In 2D spectra, youll find a
one-dimensional NMR spectrum on each axis. It may be the same spectrum on both axes, in
57

which case the experiment is homonuclear, or the axes may contain different spectra, in which
case the experiment is heteronuclear. In this section, well discuss homonuclear experiments.

One example is COSY (correlation spectroscopy). In a COSY spectrum, youll find the proton
NMR on both axes. The two-dimensional field reveals vicinal coupling (H-C-C-H). You can
draw lines from the signals in the two-dimensional field to the peaks on each axis to determine
which COSY signals go with a given proton NMR signal. For instance, a predicted COSY
spectrum for ethyl propionate is shown below. This spectrum contains two triplets and two
quartets. Right away, we can surmise that there are two spin systems, both of the type A
2
X
3
.
But, we might want some help determining which quartet is in the same spin system with a given
triplet. COSY can provide that assistance.

If we follow the red line down from the
1
H NMR peak marked (a), we first encounter a signal on
the diagonal axis cutting across the spectrum. In COSY spectra, any signals falling on the
diagonal are meaningless. If you were to follow that signal on the diagonal over to the vertical
axis, youd see that it ends at the
1
H NMR peak (a). This doesnt give us any information.
However, if you continue following the red line down from the
1
H NMR signal (a), youll
encounter a second signal. Following this one over to the vertical axis shows a correlation to the
peak at about 2.3 ppm. Thats the proton marked (b). We have now identified that signals (a)
and (b) compose one spin system.

If you now follow the purple line down from the proton NMR peak labeled (d), you first
encounter the uninformative signal on the diagonal. However, continuing downward, we find a
second COSY signal. When we follow it across to the vertical axis, a correlation to the peak at
about 4.1 ppm is revealed. Thats the proton labeled (c), so now weve identified that the second
spin system does in fact consist of protons (d) and (c).

Its also worth noting that real experimentally acquired COSY spectra may contain some noise
that could be mistaken for meaningful peaks. Theres an easy way to determine if a peak is
real or if it is due to noise. Actual signals appear on both sides of the diagonal. Above, we
identified (a)-(b) and (d)-(c) correlations. It follows that (b)-(a) and (c)-(d) correlations should
also exist. You can find these by following the blue and green lines downward from the peaks
labeled (b) and (c), respectively. By contrast, if an apparent signal is only due to noise, it will
not appear on both sides of the diagonal, which is a sign that it should be ignored.

58

0 1 2 3 4 5
PPM
0
1
2
3
4
5
O
O
a
b
c
d
a d
b
c


Another very useful type of two-dimensional spectroscopy is known as NOESY (nuclear
Overhauser effect spectroscopy). While COSY reveals proximity through bonds, NOESY
illustrates proximity through space. It is based on the nuclear Overhauser effect (NOE or nOe),
which is the change in population of one protons spin states due to the proximity of another
NMR-active nucleus.

The format of a NOESY spectrum is very similar to that of a COSY, but the information that it
conveys is quite different. NOESY will show correlations for protons that might be quite distant
59

from one another when counting the number of bonds that separate them. However, if two sets
of protons happen to be held close in space to each other, the NOESY correlation appears. The
predicted NOESY spectrum of 2,5-dimethylanisole provides an illustrative example. There are
three CH
3
signals in the proton NMR spectrum of this compound, and we can use NOESY to
probe which benzylic methyl (a or c) is close to the methoxy group (b).



Following the red line down from the methyl signal labeled (a), we first come to a NOESY
signal on the diagonal. As in COSY, peaks falling on the black diagonal line bisecting the
60

spectrum are not informative. Continuing down the red line, there is a second NOESY signal
illustrating a correlation with the hydrogens labeled (b). We expect the methoxy hydrogens to be
more deshielded than the other methyl groups, so it is clear that signal (b) is the methoxy group.
We now know that the methoxy group (b) is close to the methyl group causing the signal (a).
Again, by close we now mean close in space. Despite the six-bond separation of the methyl
and methoxy protons, in certain conformations they will pass quite close by one another,
resulting in the nOe correlation.

If we continue even further down the red line, there is another NOESY signal showing a
correlation between methyl (a) and one of the aromatic protons (d). We have now established
their proximity in space as well, despite their four-bond separation. From the two correlations
described thus far [(a)-(b) and (a)-(d)], we have established connectivity within half of the
molecule.



If we follow the green line down from the methyl group labeled (c), there are two NOESY
signals illustrating correlations to the aromatic protons (e) and (f). This establishes the
connectivity of the other half of the molecule.



We can gain further confidence in our assignments and link the two halves of the structure using
an additional clue from the NOESY. There is a correlation between the aromatic hydrogen (f)
and the methoxy group (b).

CH
3
OCH
3
CH
3
a
b
c
d
e
f


Additionally, it is worth noting that three-bond (vicinal) coupling, which would normally be
revealed in COSY, should be suppressed in a NOESY spectrum; however, occasionally, that
three-bond coupling is still visible to some extent. So, an experimentally acquired NOESY
spectrum might also show a correlation between aromatic hydrogens (d) and (e).


61

Two-dimensional NMR spectroscopy: heteronuclear experiments

There are also many heteronuclear two-dimensional NMR experiments that can provide a wide
array of valuable information. As the name suggests, a heteronuclear experiment shows
correlations between different types of nuclei. Therefore, the axes will contain two different
NMR spectra. Proton-carbon coupling is especially useful since it can essentially allow us to
pair proton and carbon signals, showing which hydrogens stem from a particular carbon. There
are two experiments that give the same 2D NMR, and they are known as HSQC (heteronuclear
single-quantum correlation spectroscopy) and HMQC (heteronuclear multiple-quantum
correlation spectroscopy). While the instrumental methods differ, they produce the same
spectrum, which correlates protons with the carbons to which they are bonded.

Earlier, we used COSY to pair signals within the two spin systems of ethyl propionate. Now, we
can use the predicted HSQC spectrum below to pair the two methyl carbons with their own
hydrogens and the two methylene carbons with their own hydrogens. Since the axes differ in
heteronuclear experiments, each peak has meaning. Recall that the homonuclear experiments
had uninformative peaks that fell on the diagonal bisecting the spectrum. That is not the case
with heteronuclear experiments.

By following the red line down from the proton NMR signal labeled (a), we encounter one
HSQC signal. Tracing this to the vertical axis shows us that these particular protons are bonded
to the carbon appearing just under 10 ppm. In the same fashion, the other protons in this A
2
X
3

spin system can be assigned to their carbon. The blue line from the quartet for protons (b) can be
followed down to an HSQC signal correlating them with the carbon just under 30 ppm. Weve
now identified the entire A
2
X
3
spin system, including both the protons and the carbons.

The same process can be applied to assign the remainder of the other A
2
X
3
spin system. The
triplet proton signal (d) correlates with the carbon near 15 ppm, while the quartet proton signal
(c) couples with the carbon appearing just above 60 ppm.

Notice that the carbonyl carbon near 175 ppm does not correlate with any hydrogens. This of
course stands to reason since it is not directly bonded to any protons. Chemical shift would have
already revealed the identity of this carbon to us; however, the lack of HSQC signal is an
additional clue that this carbon must be the carbonyl carbon.

62




63

Chapter 2 Problems

1. The natural product hirsutellone C was synthesized by the research group of K.C. Nicolaou at
The Scripps Research Institute (Organic Letters, 2011, 13(20), 5708-5710).



The
13
C NMR of hirsutellone C displays 10 signals in the C=C region of the spectrum (100 150
ppm). This indicates that each alkene and aromatic carbon in the molecule is unique. Why does
each carbon in the aromatic ring of hirsutellone C give its own signal, while closely related 4-
methylanisole (shown below) gives only four aromatic
13
C signals?



2. Sawama, Sawama, and Krause (of the Dortmund University of Technology in Germany)
reported a highly selective ring-opening reaction of dihydrofurans of the type shown below
(Organic Letters, 2009, 11(21), 5034-5037):



How could you distinguish between the following two dihydrofurans using
1
H NMR
spectroscopy? To answer this question, name two notable differences in the number and type of
signals that you would expect to see in their spectra.



3. 3,4-Methylenedioxypyrovalerone (MDPV) is a so-called designer drug used as a stimulant. A
designer drug attempts to circumvent laws against stimulants such as methamphetamine by
creating a new but analogous chemical structure with similar effects. MDPV has sometimes
been sold in products labeled as bath salts. This compound has been outlawed due to the
disturbing behavior it causes in users.
64


Explain how
1
H NMR could be used to distinguish between MDPV and the isomeric compound
shown below. Focus on the attributes of the signal for H
a
in your answer.



4. A review article is a paper that summarizes the developments in a particular area of research.
The chemistry of the aporhoeadane alkaloids has been reviewed [Archive for Organic Chemistry
(ARKIVOC), 2013, i, 1-65]. One of the alkaloids that is a central focus of the review is
lennoxamine. Draw a splitting tree for the proton labeled H
a
. Assume that no long-range
coupling will be observed (i.e. consider only three-bond coupling).



5. Draw a splitting tree for the proton indicated by the arrow in the structure of morphine shown
below. Assume that long-range coupling is not observed (i.e. only three-bond coupling or less
will be apparent).



6. Mercks Fosamax (shown below) is a bisphosphonate drug used in the treatment of
osteoporosis. Some patients experience musculoskeletal pain when taking bisphosphonates. Are
the protons indicated by the arrow chemically equivalent? Are they magnetically equivalent?




65

7. Are the protons labeled H
a
in the following structure chemically equivalent? Are they
magnetically equivalent?



8. Melohenine B is a novel alkaloid that was isolated from the plant Melodinus henryi (Organic
Letters, 2009, 11(21), 4834-4837). Assuming that the hydroxyl proton is rapidly exchanging,
what are the spin systems for the isolated portions of the molecule indicated by the arrows.



9. Koszelewski et al. (of the University of Graz in Austria) published a paper in Organic Letters
[2009, 11(21), 4810-4812] detailing their work with mexiletine, an orally effective anti-
arrhythmic agent. Classify both sets of indicated protons in the structure of mexiletine below as
homotopic, enantiotopic, or diastereotopic.




66

10. In the review of the aporhoeadane alkaloids mentioned in question 4, one of the reactions
discussed produces a byproduct in two isomeric forms (shown below). What type of NMR
experiment could readily distinguish the two?









67





Chapter 3:
Infrared spectroscopy
68

A brief review of the fundamentals of IR

This chapter will be fairly brief because your Introductory Organic Chemistry course has given
you a good background in IR that is sufficient for many situations. However, there are a couple
of subtleties of IR that you may not have seen previously and which could help you as we soon
venture into structure elucidation.

Before we get to those subtleties, a brief review is in order. Remember that IR excites molecular
vibrations. Of particular interest are stretching and bending. Often, when discussing IR, atoms
are conceptualized as weights and the bond between them is visualized as a spring. The stronger
the bond, the tighter the spring and the higher the frequency of oscillation. Weaker bonds
constitute weaker springs, leading to lower frequencies of oscillation. This spring strength is
typically described as the springs force constant. Additionally, heavier atoms will lower the
frequency of oscillation, much as you would expect for a spring stretched between two heavy
weights.

stretch compress
Stretching
normal
state
normal
state
stretch compress
Bending


As infrared light passes through a sample, certain frequencies resonate with particular modes of
stretching and bending. These frequencies are absorbed, while others pass through the sample.
The IR spectrum itself reports frequency in wavenumbers (), which are simply the inverse of
wavelength. The vertical axis of the spectrum displays the percentage of IR light transmitted
through the sample (% Transmittance). More polar bonds tend to yield more intense signals (i.e.
signals with a lower % transmittance).

An IR spectrum can be divided into two major segments: the fingerprint region and the
functional group region. In general, signals in the fingerprint region have less predictive value.
As the name suggests, they are, by and large, a fingerprint of the molecule, and of course you
cannot extrapolate what a person looks like by his/her fingerprints. Nor can you extrapolate the
69

structure of a molecule from most peaks in the fingerprint region. Therefore, this information is
most useful when comparing a spectrum to entries in a spectral library. In contrast, the
functional group region contains peaks that do have general predictive value. They signify
certain types of bonds within a molecules structure.



Given that the fingerprint region wont be all that useful for us, we can focus more on the
important segments of the functional group region. The following list is not all inclusive, but it
is a quick reference that covers most situations. Carbonyl stretching occurs around 1700 cm
-1
.
Carbon-to-hydrogen stretching occurs near 3000 cm
-1
. Importantly, when the carbon is sp
3

hybridized, the stretching occurs just below 3000 cm
-1
, but when the carbon is sp
2
hybridized, the
stretching is just above 3000 cm
-1
. Finally, alcohol O-H stretching and N-H stretching occur
around 3400 cm
-1
.

70


We can elaborate a bit more on some of these segments of the functional group region.

The carbonyl stretching region

Simple aldehydes and ketones absorb near 1715 1720 cm
-1
. Esters typically absorb a bit
higher, around 1740 cm
-1
. On the other hand, amides typically absorb a bit lower, around 1650
cm
-1
. We can explain this by considering the effect of the heteroatom on the carbonyl in each
instance. While esters can most certainly experience resonance, the electronegative character of
the carboxyl (sp
3
) oxygen makes the dipole between the carbonyl carbon and the carboxyl
oxygen the predominant influence. The carboxyl oxygen is less inclined to release electron
density through resonance than it is to pull electron density toward itself inductively. This has
the effect of strengthening the spring, or enhancing the springs force constant. Thus, the
frequency of oscillation increases. On the other hand, the nitrogen of an amide is much more
electron releasing, so resonance becomes a significant effect. The resonance of electrons from
nitrogen into the carbonyl lends some single bond character to the carbonyl. Since single bonds
are weaker than double bonds, this weakens the spring, or reduces the springs force constant,
resulting in a lower frequency of oscillation.



The carbonyl stretch of a carboxylic acid is influenced by hydrogen bonding. A carboxylic acid
in isolation will show a carbonyl resonance around 1760 cm
-1
, not unlike an ester. However, if
the concentration of the sample is high enough to allow hydrogen bonding, the carbonyl
71

resonance will fall below 1700 cm
-1
. The hydrogen bonding withdraws some electron density
from the carbonyl, weakening the force constant and reducing the frequency of oscillation.




There are two other notable influences on carbonyl stretching: conjugation and ring size.
Conjugation lowers the frequency of a carbonyl by about 30 cm
-1
. Resonance due to conjugation
will introduce single-bond character to the carbonyl, reducing its force constant and frequency.



Ring size also plays a role in adjusting carbonyl frequency. A carbonyl in a six-membered ring
is effectively unstrained and will exhibit the expected frequency. However, as the ring size
decreases, the bond angle of the carbonyl is constrained, which results in a higher energy
absorption.



The heteroatom-to-hydrogen stretching region

The shape of the peaks in this region of the spectrum (around 3400 cm
-1
) can provide some
excellent structural clues. An alcohol O-H stretch gives a relatively broad signal in this region;
whereas, N-H stretching tends to be sharper. Additionally, while an R
2
N-H group gives a single
sharp peak, an RNH
2
group displays two peaks due to symmetric and asymmetric stretching.



72

Carboxylic acid O-H stretching is significantly different from alcohol O-H stretching. The
carboxylic acid O-H stretch is much broader, covering a range from 2500 3500 cm
-1
and often
obscuring the C-H stretching region.

Chapter 3 Problems

1. Which of the following ketones is more consistent with the portion of the IR spectrum shown?





2. Lets say that, in the course of your research, you conduct the [4+2]cycloaddition outlined
below, which is followed by spontaneous oxidation in air to provide a tetrasubstituted benzene
ring. From your study of related Diels-Alder reactions, you know that there are two possible
routes (A and B) that would produce the regioisomers shown. How could you use IR to
distinguish between the final products of Routes A and B?




73

3. How could IR spectroscopy be used to distinguish between ethyl formate and urethane (also
known as ethyl carbamate)?



4. Explain the fact that 2-pyrrolidinone has a carbonyl resonance at about 1700 cm
-1
.



5. Predict the carbonyl resonance in the IR spectrum of azaadamantane-2-one.










74







Chapter 4:
Structure elucidation
75

Weve spent some time discussing modern tools for the elucidation of chemical structures, but it
can still be very daunting to apply all of these tools in a real structure solving problem. So, lets
try a few together so that you can develop an approach to using all of the techniques in your
repertoire.

Structure Solving Example #1

Lets try to determine the structure of a molecule that has the molecular formula C
6
H
10
O
2
, as
well as the following
1
H and
13
C NMR spectra.

1
H NMR spectrum:



13
C NMR spectrum:



76

Sometimes the most challenging part of such a problem is simply knowing where to begin.
Anytime you are presented with a molecular formula, you can always calculate the degrees of
unsaturation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2


cgrccs o unsoturotion =
|2(6) + 2] 1u
2
= 2

We wont try to decide whether these are two rings, two bonds, or one of each just yet. Well
wait to see if the spectra can help us to make that decision.

The
13
C NMR spectrum is usually fairly easy to interpret because it contains fewer pieces of
information than the
1
H NMR spectrum, so it might be wise to extract what we can from the
carbon NMR next. First, the fact that there are six carbons in the molecule and six signals in the
spectrum shows that this molecule does not contain any symmetry. Each of the carbons is in a
unique chemical environment. Next, we can identify three sp
3
carbons with chemical shifts less
than 100 ppm, as well as two alkene carbons ( 100 150 ppm) and one carbonyl carbon ( >
160 ppm). Weve just figured out what our two degrees of unsaturation are. If an alkene and a
carbonyl are present, then there are two bonds.

As we start to examine the
1
H NMR spectrum, it may be easier to work from the higher chemical
shift signals (downfield) because they are associated with more distinctive groups. The peak at
11 ppm is probably a carboxylic acid proton (remember to refer back to the graphic of chemical
shift regions in the NMR chapter), so now we have a small fragment.



We had already determined from the
13
C NMR spectrum that an alkene was present. There is
also a proton signal near 7 ppm that could be associated with an alkene (it certainly isnt
aromatic because we dont have enough degrees of unsaturation for a benzene ring). However,
theres only one vinyl (alkene) proton, so the alkene must be trisubstituted.



As we continue to move upfield on the proton NMR spectrum, we encounter signals for 3, 2, and
3 hydrogens. These must be two methyls and a methylene. One methyl group is isolated. It has
no neighbors and therefore no splitting.



The remaining methyl and methylene form an A
2
X
3
(or A
2
M
3
) spin system.
77




Lets pause to tally up all of our fragments.



All of the atoms in the molecular formula are accounted for, so all that remains is to connect
these fragments in a reasonable fashion. There is one trivalent fragment, along with three
monovalent fragments. So, it makes a lot of sense to place the trivalent fragment in the center
and surround it with the ethyl, methyl, and carboxylic acid groups. This results in only two
possible structures, which are the E and Z isomers of 2-methyl-2-pentenoic acid.



A NOESY spectrum could help us to distinguish between the two. We would look for the
presence of a diagnostic correlation between the vinyl hydrogen and the methyl group to the
carbonyl. The following NOESY spectrum does not show that correlation, but it does show
correlation between the methylene and the -methyl, suggesting that the correct isomer is (E)-2-
methyl-2-pentenoic acid.



78



Structure Solving Example #2

A molecule has the formula C
12
H
16
O
3
. It displays a prominent peak near 1740 cm
-1
in the IR
spectrum and has the following
1
H and
13
C NMR.

1
H NMR spectrum:

79



13
C NMR spectrum:



Again, we should begin with the calculation of degrees of unsaturation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
=
|2(12) + 2] 16
2
= S

When there are a sizable number of degrees of unsaturation in a compound, it is useful to
remember that an aromatic ring constitutes a set of four degrees of unsaturation (three bonds
and one ring). Therefore, for four or more degrees of unsaturation, it is often expedient to
assume the presence of an aromatic ring. In this case, we would have one aromatic ring, as well
as one additional degree of unsaturation, which would be either a bond or a ring.

We can be a bit more specific about both the aromatic ring and the remaining degree of
unsaturation. A quick glance at the proton NMR reveals four aromatic hydrogens, so we know
80

that the aromatic ring is disubstituted. We know that the aromatic ring has no symmetry because
each of its six carbons appears as a unique signal in the
13
C NMR. Also, the shape of the
aromatic proton signals in the
1
H NMR is informative. Although two of the aromatic proton
signals overlap causing a complex multiplet, we can clearly see a singlet. The only way to
achieve that splitting is from a meta substitution pattern.



Furthermore, the remaining degree of unsaturation is the bond of a carbonyl, specifically an
ester. We see one carbonyl carbon resonance in the carbon-13 spectrum just above 170 ppm, and
the IR signal near 1740 cm
-1
is suggestive of an ester.



The proton NMR also clearly shows three methyl groups. Two of these methyl groups are
similarly deshielded, and their chemical shifts between 3.5 and 4.0 ppm suggest proximity to
oxygen atoms (oxygen atoms being the only heteroatoms in the compound). On the basis of this
clue, it seems reasonable to expand our ester fragment to a methyl ester and to add a methoxy
fragment as well.



The remaining methyl group is a shielded doublet, so it must have a single (methine) neighbor.



At this point, weve amassed as sizable number of fragments, so it might be useful to compare
the molecular formula to the fragments. Recall that the formula is C
12
H
16
O
3
. Weve accounted
for 11 of those carbons (6 in the aromatic ring, 2 in the methyl ester, 1 in the methoxy group, and
2 in the methyl with methine neighbor). All three oxygens are accounted for between the methyl
ester and the methoxy group. Finally, of the 16 hydrogens in the compound, 14 are present in
the fragments identified thus far (4 on the aromatic ring, 3 in the methyl ester, 3 in the methoxy
group, 4 in the methyl with methine neighbor). All that remains is one carbon and two
hydrogens, or a methylene group.

81

Notice though, that no methylene signal appears in the
1
H NMR. We would expect a signal with
an integration of 2H below 4.5 ppm. Instead, the two proton signals that we havent yet
addressed are a pair of peaks, each integrating for 1H, near 2.5 ppm. These could be the two
hydrogens of a methylene group if those hydrogens are diastereotopic. To have diastereotopic
methylene hydrogens, there would need to be chirality in the molecule. If the two R groups
attached to the methine are unique, then that methine would be a stereocenter, and its chirality
would differentiate the methylene protons if they are located nearby.



The methine hydrogen is fairly deshielded ( 3.5 ppm). Therefore, it must be next to some
deshielding portion of the molecule, like the aromatic ring. The methoxy group could also be
placed on the aromatic ring.



This would leave one place to position the methyl ester.



An additional clue that enabled us to assemble the fragments correctly was the fact that the ester
IR resonance was not consistent with conjugation, which would have lowered the carbonyl
vibration from 1740 to 1710 cm
-1
. Therefore, the methyl ester could not have been attached
directly to the aromatic ring.

Its always a good idea to double check your structural hypothesis by assigning all of its protons
and carbons to their corresponding signals. You may not always be able to make assignments
with perfect specificity. For instance, it would be quite challenging to know which
diastereotopic proton (b) corresponds to which signal (i.e. the one slightly above 2.5 ppm or the
one slightly below 2.5 ppm). Nevertheless, the exercise helps you to uncover any errors that
might have been made during the process.

82




Assigning each and every carbon signal perfectly can be quite challenging without additional
information (e.g. DEPT, APT, HSQC), but we can certainly group similar signals. For instance,
the six signals from 110 160 ppm can be identified as the aromatic carbons. The two signals
between 50 and 60 ppm can be assigned to the methyl groups connected to oxygen. The two
signals near 40 ppm can be attributed to the moderately deshielded methine and methylene
carbons. Finally, the methyl group remote from functionality (d) can be assigned to the signal
near 20 ppm, and the carbonyl appears just above 170 ppm.

0 20 40 60 80 100 120 140 160 180
PPM
aromatic carbons (e, f, g, h, i, j)
C=O
a and k c and b
d

83


Structure Solving Example #3

A compound has the molecular formula C
11
H
12
O
2
. It shows an IR signal at about 1715 cm
-1
. It
also exhibits the following predicted spectra.

1
H NMR spectrum:



13
C NMR spectrum:



A degrees of unsaturation calculation is always a reasonable place to begin.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
=
|2(11) + 2] 12
2
= 6

84

This suggests one aromatic ring and two additional degrees of unsaturation. A quick glance at
the carbon-13 NMR shows that the aromatic ring contains no symmetry, and the proton NMR
shows three aromatic hydrogens. Therefore, the ring must be trisubstituted, leaving us with two
choices for the generic substitution pattern.



If the first pattern (1,2,3-trisubstituted) were the case, all of the hydrogens would experience
ortho coupling of significant magnitude. However, we can see in the proton spectrum that one of
the hydrogens (just below 7 ppm) is split with coupling of a small magnitude. This suggests that
the substitution pattern is 1,2,4-trisubstituted instead.



The IR and
13
C NMR both show evidence of a carbonyl, likely a ketone (an aldehyde would
contain a proton with a chemical shift of 9 10 ppm, which is not present). This is one of our
remaining degrees of unsaturation. No additional carbonyl appears to be present, so the sixth
degree of unsaturation is a ring.



The four signals below 4 ppm in the proton NMR are readily interpreted and provide another set
of fragments. There is a fairly deshielded methyl group with no neighbors. This could well be a
methoxy group. There is a methylene with no neighbors, and there are two methylenes (triplets)
that are part of an A
2
M
2
spin system.

R
O
H
3
C
R
H
2
C
R
R
C
H
2
CH
2
R
methoxy group methylene
(no neighbors)
A
2
M
2
spin system


At this point, all of the atoms in the molecular formula (C
11
H
12
O
2
) are accounted for. We have
identified the 11 carbons (6 in the aromatic ring, 1 in the carbonyl, 1 in the methoxy group, 1 in
the methylene without neighbors, and 2 in the A
2
M
2
spin system), the 12 hydrogens (3 on the
aromatic ring, 3 on the methoxy group, 2 on the methylene without neighbors, and 4 in the A
2
M
2

spin system), and the 2 oxygens (1 in the carbonyl and 1 in the methoxy group). All that remains
is to assemble them in a reasonable fashion.

85

We know that the carbonyl cannot be directly bonded to the ring because that would introduce
conjugation, which would lower the IR frequency of 1715 cm
-1
by about 30 cm
-1
. However, we
could in theory bond any other fragment to the aromatic ring. Lets arbitrarily choose the
methoxy group. Knowing that we need to incorporate one more ring into the structure, we
should not place the methoxy group on either of the adjacent R groups since these would be the
most logical spot for our final ring.



We could choose to place the methylene with no neighbors next. It could be added to either of
the remaining positions.



In order for the methylene not to have neighbors, it must next connect to the carbonyl.



Finally, the two remaining valences can be joined by the A
2
M
2
spin system.



These structures could be drawn more concisely as:



To distinguish between these structures, we would probably need some additional information,
like a NOESY spectrum. The NOESY spectrum of the compound would show different
correlations depending on which structural hypothesis was correct. For instance, we could
choose to focus on the isolated aromatic hydrogen. It would correlate with a methylene of the
A
2
M
2
spin system in one structure or the methylene with no neighbors in the other structure.

86



The predicted NOESY spectrum is shown below.


87


The isolated aromatic hydrogen (f) correlates with one of the triplet methylenes of the A
2
M
2
spin
system. This shows us that the following structure is the correct one.

H
3
CO
O
observed
nOe
f b


Structure Solving Example #4

A compound has the molecular formula C
15
H
12
O
2
. In the IR spectrum, there is a peak at
approximately 1685 cm
-1
. It also has the following predicted spectra.

1
H NMR spectrum:




88

13
C NMR spectrum:



The degrees of unsaturation calculation shows a total of ten bonds and/or rings. Remembering
that an aromatic ring can readily account for four degrees of unsaturation, we can predict that
this compound likely contains two aromatic rings, along with two other degrees of unsaturation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
=
|2(1S) + 2] 12
2
= 1u

The proton NMR spectrum reveals a total of nine aromatic hydrogens, suggesting that one ring is
monosubstituted, while the other is disubstituted. The carbon-13 NMR supports this because
there are a total of ten aromatic carbon signals between 100 and 160 ppm, indicating some
symmetry in the molecule. The monosubstituted ring would exhibit only four different carbon
signals, and the disubstituted ring could have as many as six signals if the R groups differ.



Of course, there are three options for the substitution pattern (ortho, meta, or para) of the
disubstituted ring. Para substitution would give two doublets, assuming that the R groups differ.
Meta substitution would yield a singlet (or weakly coupled doublet), two doublets, and a doublet
of doublets (or apparent triplet). Since neither pattern appears in the proton spectrum, we might
tentatively assume ortho substitution. However, we should recognize that, with the significant
signal overlap that occurs in the proton NMR spectrum, it is possible that one of the other
splitting patterns occurred and was merely obscured.
89




The IR spectrum suggests a conjugated ketone since the frequency falls below 1700 cm
-1
.



We have thus far accounted for 13 carbons, 9 hydrogens, and 1 oxygen. Subtracting this from
the formula (C
15
H
12
O
2
) leaves us with only 2 carbons, 3 hydrogens, and 1 oxygen. Notice that
the
13
C spectrum shows two signals below 100 ppm, which must correspond with our two
residual carbon atoms. This is very important to recognize because the signal near 5.5 ppm in
the proton spectrum might lead you to incorrectly assume the presence of an alkene. However,
having an alkene in the molecule is not consistent with the carbon-13 NMR data.

The remaining hydrogens must be part of an ABX spin system. There are two doublets of
doublets between 3.0 and 3.5 ppm. Also, there is an apparent triplet at 5.5 ppm. We have only
two remaining carbons onto which to place these three hydrogens. This means that we have no
choice but to link a methine to a methylene. However, the methylene hydrogens must be
diastereotopic in order to explain the observed splitting.



One oxygen atom is still unaccounted for, but we can surmise that it should be bonded to the
methine. The methine proton is unusually deshielded (so much so that it falls in the alkene
proton region of the
1
H NMR). Consequently, it must be bonded to multiple deshielding
substituents, so the methylene-methine fragment can be expanded to contain the remaining
oxygen.



By taking stock of the fragments once more, it quickly becomes apparent that there arent many
ways to connect the pieces. We must also remember that one more ring must be inserted into the
structure in order to have the correct degrees of unsaturation (and therefore the correct formula).

90



In order for the carbonyl to be conjugated, it must be bonded to one of the aromatic rings.



Since the structure contains one more ring, and six-membered rings are quite commonly
occurring, it stands to reason to attach the methylene-methine fragment to the two open valences
above.



Then, the remaining phenyl group can be added to complete the structure, which happens to be
known as flavanone.



We can now see why the methine hydrogen is so deshielded. In addition to the electron
withdraw of the oxygen atom, it is also deshielded by the electronegative sp
2
carbon of the
phenyl group to which it is bonded. Furthermore, the methine is a chiral center, which would in
fact render the adjacent methylene hydrogens diastereotopic.

Structure Solving Example #5

A compound has the molecular formula C
10
H
9
NO
2
. In the IR spectrum, there is a sharp peak at
3400 cm
-1
, as well as a signal near 1700 cm
-1
. The molecule also has the following predicted
spectra.


91

1
H NMR spectrum:



13
C NMR Spectrum:



In previous calculations of degrees of unsaturation, we could simply ignore the heteroatom
because the presence of oxygen does not alter the calculation. However, this molecule contains
nitrogen, which does impact the calculation. There is more than one approach for adjusting the
calculation to account for nitrogen. One simple option is to count it as half of a carbon for the
purpose of the calculation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
=
|2(1u.S) + 2] 9
2
= 7

The degrees of unsaturation suggest that there is one benzene ring in the structure, as well as
three additional degrees of unsaturation. As we try to assess the level of substitution of the ring,
92

the aromatic region of the proton NMR spectrum becomes very important. At first glance, there
are five hydrogens in this region, suggesting a monosubstituted ring. However, one of the
signals is a singlet. How could we possibly have a monosubstituted benzene ring where one of
the hydrogens experiences no coupling? The answer is simple: we couldnt.



Therefore, the ring must be disubstituted, explaining the presence of four of the hydrogens in this
region. The fifth aromatic hydrogen must be part of some other fragment of the molecule, likely
a second aromatic ring of some type.



Introducing another aromatic ring without exceeding the allowable number of degrees of
unsaturation could be tricky, so well need to keep that challenge in mind.

Turning to the IR spectrum, it suggests an N-H stretch (sharp peak at 3400 cm
-1
) and a carbonyl.



The proton NMR allows us to add some specificity regarding these fragments. First, we can see
that a carboxylic acid proton is present ( 11 ppm). Since there is only one carbonyl signal in
both the IR and the
13
C NMR, that carbonyl must be part of the acid. Furthermore, the other
broad, exchangeable hydrogen in the proton NMR is located at about 10 ppm. This is quite
unusual for an amine N-H, which we would normally expect to see much further upfield for a
simple amine. There must be something very unusual about the N-H in this molecule.



The carbon-13 NMR shows eight aromatic carbons between 100 and 140 ppm. This presents
somewhat of a conundrum. If there is one benzene and a second aromatic ring of some type,
how can we add a second aromatic ring using only two more carbons. We could consider fusing
them directly to the benzene ring. This seems unreasonable on two fronts: the strain energy
would be high and the second ring has anti-aromatic character.

93



Our only other choice would be to enlarge the second ring by inserting another atom. There are
no more aromatic carbons, so we would have to use a heteroatom. The oxygens have been used
for the carboxylic acid, so nitrogen would be our only choice.



This fragment is much more reasonable. You might recognize this as an indole ring, which is
commonly occurring. It is also consistent with all of the data, notably explaining why the N-H
appears at such a special chemical shift. Between the indole ring and the carboxylic acid
fragment, we have accounted for most of the atoms in the formula. However, the proton NMR
reveals the presence of an additional methylene unit that has no neighbors.



There is only one way to put these fragments together. The methylene must unite the indole ring
and the acid.



The final question would of course be: which isomer is correct? Here, we could gain insight
from a NOESY spectrum. If we focus on C-H of the five-membered ring of the indole, there
would be one expected nOe interaction with the methylene group if the correct structure is 3-
indoleacetic acid. However, if the correct structure is 2-indoleacetic acid, there would be two
nOe interactions expected: one with the methylene and one with the proximal proton of the
benzene ring.

94




95


The NOESY spectrum reveals only one interaction for indole hydrogen (c). There is a nOe
correlation with methylene hydrogens (a), showing that the correct structure is 3-indoleacetic
acid.



Structure Solving Example #6

A compound has the molecular formula C
8
H
12
O
4
. It shows a signal at about 1710 cm
-1
in the IR
spectrum and has the following predicted spectra.

1
H NMR spectrum:




96

13
C NMR spectrum:



The degrees of unsaturation calculation shows that the molecule contains a total of three bonds
and/or rings.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
=
|2(8) + 2] 12
2
= S

There is clearly one carbonyl based on the IR (1710 cm
-1
) and
13
C NMR (just below 170 ppm).
There is also clearly an alkene based on the signals near 6 and 7 ppm in the
1
H NMR and the
signals near 120 and 150 ppm in the
13
C NMR. The alkene bears only two hydrogens, so it must
be disubstituted. Given the stated coupling constant (15 Hz), the alkene is trans-disubstituted.
There is no further evidence of bonds, so the final degree of unsaturation is a ring.



The
13
C NMR shows that, while the remainder of the carbons are sp
3
hybridized, some of them
are unusually deshielded. Two have chemical shifts near 100 ppm, which suggests adjacency to
multiple electron-withdrawing groups.

The proton NMR spectrum is complex but provides some tremendous clues nevertheless. For
example, one of the alkene protons is exceptionally deshielded and appears near 7 ppm, which is
usually associated with aromatic hydrogens. One factor that commonly results in this extensive
deshielding is conjugation to a carbonyl.

97



However, if the fragment were an ,-unsaturated ketone, the IR resonance would appear at
about 1685 cm
-1
, not 1710 cm
-1
. An ,-unsaturated ester would provide an IR resonance in the
correct location (1740 cm
-1
30 cm
-1
for conjugation).



As we begin to untangle the rest of the proton NMR spectrum, thinking about spin systems helps.
For instance, there is a distinctive methyl group that has two neighbors. There is also a
methylene group with three neighbors. Together, these two signals constitute an A
2
X
3
spin
system, or an ethyl group with no additional neighbors.



Subtracting the fragments that we have so far (5 carbons, 7 hydrogens, and 2 oxygens) from the
molecular formula (C
8
H
12
O
4
), we are left with 3 carbons, 5 hydrogens, and 2 oxygens. Each of
the remaining protons causes its own NMR signal, but there arent enough carbons to put each
hydrogen on a different carbon. Therefore, we must have some diastereotopic methylene
hydrogens in order to explain all of the unique proton NMR signals.



It is sometimes helpful to tally all of the fragments before attempting to assemble the molecule.
We know that the two remaining oxygen atoms must be involved in ether functionalities because
there are no additional carbonyls and there are no broad O-H signals in the proton NMR.


98


As we start to join fragments, we might note that the oxygen atoms must have at least one
intervening carbon. Youll recall that there are some especially deshielded carbon atoms with
chemical shifts near 100 ppm in the carbon NMR. A carbon bonded to two oxygen atoms would
be consistent with this extensive deshielding.



The structure must contain a ring as its final degree of unsaturation. The monovalent ethyl group
cant be part of the ring, and the trans ,-unsaturated ester is not amenable to incorporation into
a ring of reasonable size. Therefore, it makes sense to close the ring using the remaining
methylene and methine groups.

C C
O
C
H
2
O
H H
H R


This leaves only one way to attach the remaining pieces of the structure.



As we double check our work, there are some very distinctive features of this molecule that
match the data. One such feature is the chiral center that causes the hydrogens of the two
proximal methylene groups to be diastereotopic. You might very reasonably wonder why the
methylene of the ethyl group doesnt exhibit diastereotopic hydrogens. Formally, those
hydrogens are diastereotopic; however, they are too distant from the chiral center for their
diastereotopic nature to manifest itself in different chemical shifts. Basically, they are different
protons but they arent different enough to resolve into separate signals since the chiral center is
so far away.

99



Another distinctive feature is the pair of highly deshielded carbon atoms. One is bonded to two
electron-withdrawing oxygen atoms. The other is bonded to an oxygen as well as an electron-
withdrawing sp
2
-hybridized carbon.



Summary

Even after six thoroughly worked examples, these problems may still seem overwhelming to
you. Its important to recognize that it takes a great deal of practice to become adept at
interpreting spectral data. You wont become an expert overnight, but every additional problem
that you solve will move you one step closer to that goal.

As you begin to solve problems on your own, try to follow an algorithm to stay organized, at
least at the start of every new problem. Remember to:

- calculate the degrees of unsaturation.

- check the IR spectrum for quick clues about functional groups.

- keep track of how many degrees of unsaturation appear to be bonds and how many appear to
be rings.

- scan the
13
C NMR spectrum for clues about carbonyls and aromatic / alkene carbons. Also,
look for uniquely deshielded signals (i.e. those near 100 ppm but still in the sp
3
-hybridized
range).
100


- scan the proton NMR for distinctive types of hydrogens. Remember that these are likely to be
the downfield protons.

- analyze spin systems as you interpret the
1
H NMR.

- tally your fragments and periodically compare them to the molecular formula.

- tally all of the atoms before assembling the molecule to ensure that you arent missing any
pieces and that you havent accidentally added any extra atoms to the formula.

Chapter 4 Problems

1. Recall from introductory Organic Chemistry that the aldol reaction is a transformation in
which an enolate of an aldehyde or ketone attacks the carbonyl of another aldehyde or ketone.
When the two ketones or aldehydes are different, the reaction is called a crossed aldol. Below is
an example of a crossed aldol reaction.

O
1. LDA, -78
o
C
2.
H
O
O OH
vs.
O
OH


2-Butanone is treated with a base known as LDA at -78 C. Then, benzaldehyde is added to the
mixture in step 2. You obtain a single product from this reaction, but you arent sure which of
the two potential structures is the correct one. Based on the
1
H NMR spectrum shown below,
what is the reaction product?

1
H NMR spectrum:


101

2. Cuelure (C
12
H
14
O
3
) is an attractant for the male melon fruit fly and is significant due to its
usage in the trapping of this pest to stave off infestation. The IR,
1
H NMR, and
13
C NMR spectra
of Cuelure are provided below. What is cuelures structure?

Segment of IR spectrum:


1
H NMR spectrum:


102

13
C NMR spectrum:


3. A compound with the molecular formula C
5
H
8
O
4
gives the following proton and carbon NMR
spectra. In the IR spectrum, this compound exhibits peaks just below 3000 cm
-1
and at about
1740 cm
-1
. Provide a structure of the unknown compound that is consistent with the data.

1
H NMR spectrum:


103

13
C NMR spectrum:


4. In 2009, Fang and Hou of the Shanghai Institute of Organic Chemistry published a paper in
Organic Letters describing an Asymmetric Copper-Catalyzed Propargylic Substitution Reaction
of Propargylic Acetates with Enamines.
2
Among the reactions they reported in the paper was
the following:



The product has the molecular formula C
18
H
16
O
2
. The IR spectrum has a notable resonance at
approximately 1680 cm
-1
, as well as a notable C-H stretch at about 3300 cm
-1
. The product gave
the following proton NMR spectrum. Provide the structure of the product that they generated.



2
Fang, P.; Hou, X.-L. Organic Letters, 2009, 11(20), 4612-4615.
104

1
H NMR spectrum:


5. In 2013, Reddy and co-workers published an article in Tetrahedron Letters entitled
Diastereoselective synthesis of 1-(tetrahydrofuran-3-yl)-1,3-dihydroisobenzofuran derivatives
via Prins bicyclization.
3
One of the reactions reported in that paper was the following.



The product has the molecular formula C
18
H
18
O
2
. The IR spectrum reveals sp
2
and sp
3
carbon-
hydrogen stretching; however, there are no notable signals around 1700 cm
-1
or 3400 cm
-1
. The
product also has the following predicted spectra. What is its structure?



3
Reddy, B. V. S.; Jalal, S.; Borkar, P.; Yadav, J. S.; Reddy, P. G.; Sarma, A. V. S. Tetrahedron
Letters, 2013, 54(12), 1519-1523.
105

1
H NMR spectrum:


13
C NMR spectrum:



106

HSQC spectrum:



107

COSY spectrum:
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5 6 7 8
PPM






108




Part II:
Aromatic
Chemistry Expanded

109

A brief review of what we know about aromatic chemistry

In your Introductory Organic Chemistry course, you learned about Electrophilic Aromatic
Substitution (EAS). In EAS reactions, the aromatic ring acts as a nucleophile and attacks an
electrophile (E
+
). The intermediate that results has an additional carbon-to-electrophile bond and
is known as the complex. The positive charge is stabilized through resonance delocalization.
Finally, loss of a proton from the sp
3
-hybridized carbon of the complex restores aromaticity.



There are five common EAS reactions:

(1) halogenation

(2) nitration

(3) sulfonation

(4) Friedel-Crafts alkylation

(5) Friedel Crafts acylation

All that differs between these reactions is the electrophile. These five reactions that youve
learned previously enable us to synthesize an impressive array of substitution patterns on the
benzene ring. However, there are some prominent substituents that we cannot currently add to
the ring. For example, we cannot yet add a hydroxyl group to the ring to make a phenol.



Additionally, there are certain substitution patterns that we cannot yet achieve. As an example,
try to use the EAS reactions above to prepare 1,3,5-tribromobenzene. Youll find it impossible
to accomplish.



As we expand our knowledge of aromatic chemistry, well be able to tackle these challenges and
others.
110





Chapter 5:
Aryl amines


111

Nitrous acid

Nitrous acid is formed from sodium nitrite (NaNO
2
) and hydrochloric acid.



In the presence of additional acid, the nitrosonium ion can be formed.



The nitrosonium ion is an electrophile that reacts with amines.

Reaction of the nitrosonium ion with primary amines

The nitrosonium ion reacts with primary amines to produce a diazonium ion.



Since the nitrosonium ion is generated in situ from NaNO
2
and HCl, there is chloride present,
which serves as the counterion for the diazonium ion.



When R is an alkyl group, nitrogen (N
2
) is expelled as a leaving group, and the resultant
carbocation undergoes a mixture of substitution and elimination. This is not particularly useful.



However, when the R group is aryl (i.e. benzene or a substituted benzene), the diazonium salt is
stable enough to persist until additional reagents are added to initiate another transformation.
We will use aryl diazonium salts in multiple transformations.

112



Reaction of the nitrosonium ion with secondary amines

The nitrosonium ion can also react with secondary amines. The products are N-nitrosamines.



N-nitrosamines are carcinogenic compounds. It is worth noting that they can sometimes appear
in food. Reaction of nitrites with secondary amines in proteins can lead to the production of N-
nitrosamines. This disturbing occurrence isnt widespread, but certain situations, like frying
meat at high temperatures, can be of concern.

Aryl diazonium salts: an overview

There are two principal reactions of aryl diazonium salts. These are substitution and coupling.

Substitution:

N N Cl
Nuc
Nuc + N
2
+ Cl


Coupling:



Well consider each of these processes in detail.

Substitution reactions of aryl diazonium salts

There are six types of substitution reactions. The Sandmeyer reaction involves the use of
copper(I) salts to form aryl chlorides or bromides.



The mechanism involves a series of single-electron transfers.
113



The utility of the Sandmeyer reaction may not be immediately apparent. It is true that we
already know how to achieve chlorination and bromination using the EAS paradigm, but we will
see shortly that it is advantageous to have a second approach for the installation of these
substituents.

The Schiemann (or Balz-Schiemann) reaction allows the installation of fluorine.



It is worth noting that HBF
4
is sometimes used as the acid in the generation of the nitrosonium
ion so as to arrive directly at the diazonium tetrafluoroborate, which is subsequently thermally
decomposed to the aryl fluoride.



Yet another halogen can be installed using potassium iodide. The installation of fluorine and
iodine on the aromatic ring is a new capability compared to what we could achieve with EAS
alone.



A reaction similar to the Sandmeyer process allows the formation of benzonitrile, or its aryl
analogues.



Heating an aryl diazonium salt with water yields a phenol.



Finally, hypophosphorous acid enables the conversion of the diazonium ion to a hydrogen
substituent.
114




The utility of this last transformation may also not be readily apparent, since it is rare to delete
functionality altogether. However, well soon see that it provides a very versatile synthetic tool.

Four of the six substitution reactions provide us with entirely new capabilities. We can now add
F, I, CN, and OH to an aromatic ring. Those are transformations that EAS would not allow.

Using substitution reactions of diazonium salts in synthesis

There are many ways that we could use diazonium salts in synthesis but the most interesting
involve creating substitution patterns that were previously inaccessible. For instance, consider
the synthesis of 1,3-dibromobenzene from benzene.



We can certainly install one of the bromine atoms using EAS bromination.



However, the next step proves to be an intractable problem. Halogens are deactivating but are
the only deactivating groups that direct ortho and para. As a result, the product is a mixture of
1,2- and 1,4-dibromobenzene, neither of which was our target.



The solution to this synthetic challenge lies with diazonium salts. Lets examine the procedure
for installing a diazonium ion on benzene in the hope that, at some stage of the process, the ring
will bear a meta director. Diazotization results from the reaction between an aryl amine and
nitrous acid. Therefore, we need to prepare an aryl amine, which in this case is aniline (i.e.
aminobenzene). While there is no EAS reaction that installs an amino group directly, there is an
EAS reaction that adds a nitrogen-containing group to the ring. Nitration should therefore be our
first step.



115

After each transformation, it is useful to pause and consider the directing effect of the
substituent. In order to rationalize directing effects (rather than simply memorizing them), youll
need to be able to see the entire Lewis structure of the substituent.



If the substituent can participate in resonance, the resonance structures are likely to be critical in
assessing the directing effect. The nitro group has a nitrogen-to-oxygen bond in conjugation
with the ring. Resonance withdraws electron density from the ring, placing a positive charge
ortho and para to the substituent. These positions are certainly deactivated toward further EAS
reaction, in which the aromatic ring would serve as a nucleophile. Further EAS could only occur
at the meta positions, which are the least deactivated (i.e. most nucleophilic) sites on the ring.



So, it is at this very stage that we have a meta director on the ring, and it would be wise to exploit
this to install a bromine in the meta position.



Now, weve achieved a 1,3-substitution pattern on the ring. The only problem, of course, is that
one of the substituents is the nitro group rather than bromine, which appears in our target. We
can remedy this situation through the use of a diazonium salt, but diazonium salts are made from
aryl amines. Consequently, we need to reduce the nitro group to an amino group. There are a
variety of ways to do this. You might recall H
2
and a metal catalyst (e.g. Pd/C) as reducing
conditions. Alternatively, an active metal reduction using tin or iron and hydrochloric acid
works well.



Now, diazotization can take place, and a Sandmeyer reaction will replace the diazonium ion with
the desired halogen.

116



The entire synthesis is shown below.



It is very important to note that the timing of the EAS bromination is critical. Had we reduced
the nitro group prior to EAS bromination, we would not have been able to produce the desired
substitution pattern. The reason is that the amino group is an ortho / para director due to the
donation of its lone pair into the ring via resonance. This places negative character on the ortho
and para positions, rendering them the most nucleophilic sites.


117


Another excellent example of the utility of diazonium salts in synthesis is the preparation of
1,3,5-tribromobenzene from benzene. As we saw earlier, placing halogens on the ring meta to
one another is not possible with conventional EAS reaction since the halogens are ortho / para
directors.



It is tempting to recycle the strategy used in the synthesis of 1,3-dibromobenzene above, but if
we try to do that, we encounter an intractable problem partway through the synthesis. After
placing the first bromine meta to the nitro group, the halogen becomes the dominant director on
the ring. Recall from your discussions of EAS in Introductory Organic Chemistry that the
strongest activating group on the ring controls the position of further EAS reaction. In this case,
the ring bears two deactivating groups. That, in and of itself, is a problem because powerfully
deactivated rings undergo sluggish EAS reaction (if they react at all). Even if we could force the
subsequent reaction through heating, it is the bromine, not the nitro group, that will direct further
substitution. While the halogen is not a strong activating group, it is stronger than the powerfully
deactivating nitro group. Consequently, the two isomers that result from a second bromination
have substitution patterns that are not useful to us in this particular synthesis problem.



The way around this problem comes from the substitution reaction of diazonium salts that
employs hypophosphorous acid to replace the diazonium ion with a hydrogen. This gives us the
ability to install a directing group to orchestrate the desired substitution pattern and then delete
that directing group. This transient directing group will be effectively invisible from the
standpoint of the starting material and ultimate product of the synthesis. It will appear only in
the synthetic intermediates.

118



The generic scheme above illustrates that the transient directing group (Y) should be an ortho /
para director in order to position the bromines in the correct locations. In order to be a transient
directing group, Y must be something that could be converted to a diazonium ion. Therefore, we
can begin the synthesis by adding substituents en route to the diazonium ion, pausing at each
intermediate to consider its directing effects. The synthesis begins with EAS nitration, but the
nitro group is a meta director and wont help us here. The next step is the reduction of the nitro
group to an amino group. The amino group is a strongly activating ortho / para director, and we
should exploit this property to install the halogens as desired before continuing to diazotization.
The bromination can simply be carried out using an excess of Br
2
. In other words, there is no
need to perform each of the brominations as a separate step.



Now that the desired substitution pattern has been attained, all that remains is the deletion of
the transient directing group. This is accomplished through diazotization follow by treatment
with hypophosphorous acid.



The take-home message is that seemingly impossible substitution patterns can be achieved by
creatively using the substitution reactions of aryl diazonium salts.

Coupling reactions of aryl diazonium salts

Earlier we saw the general motif of coupling reactions. In these reactions, the diazonium ion
serves as the electrophile in an EAS reaction with another aromatic ring.

119



The mechanism is no different from that of any other EAS reaction. You merely have to keep in
mind that the diazonium ion is a new electrophile (E
+
). Attack of the electron-rich aromatic ring
on the terminal nitrogen of the diazonium salt yields a complex, which loses a proton to restore
aromaticity.



The resultant compounds with the general formula (Ar-N=N-Ar) are known as azo compounds,
or sometimes as azo dyes since they have extended conjugation and therefore tend to be colored.

The requirement for the electron-donating group (EDG) on the nucleophilic ring stems from the
fact that the diazonium ion is a relatively weak electrophile. To compensate for its reduced
reactivity in this sense, the nucleophile must be stronger if the reaction is to proceed at a
reasonable rate. The electron-donating group enhances the electron-density of the ring, and
therefore its nucleophilicity. Electron-donating groups will direct ortho and para; however,
whenever possible the reaction will take place para to the electron-donating group because that
is the less sterically encumbered site.

Using coupling reactions of diazonium salts to make synthetic dyes

Synthetic dyes transformed chemistry. Prior to their development in the mid-1800s, chemistry
was not a profession. With the exception of academic chemists, there werent individuals who
practiced chemistry as a career. It had yet to be demonstrated that chemistry could be used for
some profitable endeavor. August Wilhelm von Hofmann was attempting to change this.
Professor Hofmann was convinced that chemistry could be used in a very practical sense to solve
the pressing problems of the day. One of these problems was malaria, which claimed the lives of
many, especially those in the military who traveled to tropical regions. It was known that
quinine was an effective treatment, but the isolation of sufficient quantities of quinine from the
bark of the cinchona tree was problematic. Hofmann believed that a laboratory synthesis of
quinine would address a major human health concern while simultaneously proving the practical
value of chemistry.

120



William Henry Perkin was a young student at the Royal College of Chemistry, who began his
studies with Professor Hofmann. Perkins research project was to prepare quinine from simple
starting materials. At the time, there was much less known about the structure of organic
molecules, so the state of the art was a sort of arithmetic approach to synthesis. Perkin believed
that he could make quinine (C
20
H
24
N
2
O
2
) from allyl toluidine (C
10
H
13
N) if two molecules of allyl
toluidine combined and then oxidized to incorporate two oxygen atoms and remove two
hydrogen atoms.



Today, any student of Introductory Organic Chemistry can readily see that this arithmetic
approach to synthesis is doomed to failure because the structures are too disparate, but at the
time, it seemed like a reasonable approach based on the information at hand.



Of course, Perkin did not make quinine via this method, but as he was cleaning up glassware
from his failed reaction, he noticed the purplish (or mauve) color of his actual product. Perkin
had prepared a synthetic dye that came to be known as mauveine. Mauveine is a mixture of
compounds one of which is shown below.



121

As you can see, Perkin didnt get anywhere close to the structure of quinine, but he did make a
compound with extended conjugation, which resulted in the mauve hue. Perkin was savvy
enough to recognize the potential value of this discovery. Purple colors were the shades of
royalty because the purple dyes available from natural sources were laborious to isolate and
prepare. Perkin knew that, if he could dye cotton with mauveine, it would be quite popular,
much as the latest fashion trend is always popular nowadays. This took some doing, but Perkin
eventually found a mordant that allowed mauveine to dye cotton, and he eventually produced the
dye on an industrial scale. One of the truly remarkable things about this story is that Perkins
work began when he was only 18 years old. He conducted the initial experiments that led to the
synthesis of mauveine during his Easter break from school.
4


You might wonder what all of this has to do with coupling reactions of diazonium ions. Perkins
discovery essentially founded science-based industry. Many people, including skeptical
Professor Hofmann, quickly realized the value (both scientific and financial) of his work. A
number of investigators decided to make other synthetic dyes that might be of commercial
interest. Quickly the palette of synthetic colors exploded, and this had far-reaching
consequences for science and humanity. It turns out that many of these synthetic dyes were
prepared through the coupling reactions of diazonium ions. One example is butter yellow.



As the name suggests, this compound was used as a coloring additive in food before its potential
carcinogenicity was discovered.

Using coupling reactions of diazonium salts to make sulfa drugs

Believe it or not, the development of chemotherapy was one of the consequences of the boom in
synthetic dyes. Scientists did many things with these newly available dyes, including treating
cells with them. Some individuals noticed that certain portions of the cell could be stained by
particular dyes, leading to the birth of molecular biology as organelles and biomolecules were
discovered. Other folks noticed that some dyes would actually kill microbes, leading Paul
Ehrlich to propose the magic bullet theory. Ehrlich believed that it should be possible to use
chemical agents to selectively target and kill disease-causing microbes. After 605 unsuccessful
attempts to apply his theory to practice, Ehrlich managed to prepare, on his 606
th
trial, a chemical
agent that could be used to treat the syphilis spirochete. This compound was called salvarsan.


4
For more details about this story and its consequences for history, see the book Mauve by
Simon Garfield (Norton: New York, 2000).
122



Others made important contributions in this area as well, and some relate to our discussion of
coupling reactions of diazonium ions. Gerhard Dogmak was investigating antibacterial agents
when his young daughter fell ill with a streptococcal infection. In desperation, he treated her
with a compound he called prontosil, which had puzzlingly shown promising activity in vivo but
not in vitro. With the benefit of this treatment, Dogmaks daughter recovered completely.



Prontosil is an early example of a prodrug. Prodrugs are administered in an inactive form but
undergo a reaction in the body to unveil the active drug molecule. In the body, prontosil is
metabolized to sulfanilamide, which is the active antimicrobial agent.



The mechanism of action centers around folic acid. Folic acid is an essential nutrient for
humans. We must acquire it through dietary sources. Conversely, some bacteria can synthesize
their own folic acid. Folic acid consists of a pteridine unit, a para-aminobenzoic acid (PABA)
unit, and a glutamic acid residue. Sulfanilamide is similar in size and shape to para-
aminobenzoic acid. As a result, the bacterial enzymes involved in the synthesis of folic acid
inadvertently include sulfanilamide in place of para-aminobenzoic acid. This results in an
inactive form of folic acid and eventually bacterial death.
123




Chapter 5 Problems

1. A joint publication from investigators at Purdue and the National Cancer Institute reported the
following conversion.
5
Provide reagents and intermediates for this transformation.



2. Show how 3-methoxybenzonitrile could be prepared from benzene.

3. Methyl orange is an azo dye that is often used as a pH indicator.


5
Conda-Sheridan, M.; Reddy, P. V. N.; Morrell, A.; Cobb, B. T.; Marchand, C.; Agama, K.;
Chergui, A.; Renaud, A.; Stephen, A. G.; Bindu, L. K.; Pommier, Y.; Cushman, M. Journal of
Medicinal Chemistry, 2013, 56(1), 182-200.
124



It is red-orange below pH 3.1 and yellow above pH 4.4. Show how you would prepare methyl
orange from the correct starting material (N,N-dimethyl-4-nitroaniline or 4-nitrobenzenesulfonic
acid) and any other reagents needed.



4. In 2012, Suh and co-workers published a paper in the Journal of Medicinal Chemistry entitled
Design, Synthesis, and Biological Evaluation of Novel Deguelin-Based Heat Shock Protein 90
(HSP90) Inhibitors Targeting Proliferation and Angiogenesis.
6
The following transformation
was reported in this paper. Show how it can be achieved.



5. Methyl red is another useful pH indicator with the following structure.



Show how you could prepare methyl red. Select the proper starting material (either o-
nitrobenzoic acid or N,N-dimethyl-4-nitroaniline) and use any other reagents needed.


6
Chang, D.-J.; An, H.; Kim, K.-s.; Kim, H. H.; Jung, J.; Lee, J. M.; Kim, N.-J.; Han, Y. T.; Yun,
H.; Lee, S.; Lee, G.; Lee, S.; Lee, J. S.; Cha, J.-H.; Park, J.-H.; Park, J. W.; Lee, S.-C.; Kim, S.
G.; Kim, J. H.; Lee, H.-Y.; Kim, K.-W.; Suh, Y.-G. Journal of Medicinal Chemistry, 2012,
55(24), 10863-10884.
125



6. The smoking cessation aid varenicline (shown below) is perhaps more commonly known as
Chantix.



In an effort to prepare Chantix analogues, Collum and co-workers at Cornell in collaboration
with Pfizer began with 1-chloro-3-fluorobenzene.
7


Show how you could prepare 1-chloro-3-fluorobenzene from benzene. We will then use this
product in questions at the end of Chapter 6.

7. Triclosan is a somewhat controversial antibacterial agent that has been added to many
products. One synthesis of triclosan uses a precursor bearing a nitro group.
8
Show how you
could complete this synthesis of triclosan.





7
Riggs, J. C.; Ramirez, A.; Cremeens, M. E.; Bashore, C. G.; Candler, J.; Wirtz, M. C.; Coe, J.
W.; Collum, D. B. Journal of the American Chemical Society, 2008, 130(11), 3406-3412
8
Lourens, J. G. Process for producing diphenylethers and esters, EP 1007501 A1
(WO1999010310A1), 2000.
126

8. Prepare the following alcohol from benzene and any other reagents needed.









127





Chapter 6:
Aryl halides





128

Synthesis of aryl halides

We now know about two mechanistically distinct approaches to the synthesis of aryl halides.
The first, EAS halogenation, affords access to aryl chlorides and bromides.



The second approach allows access the entire array of aryl halides using the reactions of aryl
diazonium ions.







Next, well turn our attention to two significant types of reactions of aryl halides. These can be
broadly described as addition-elimination and elimination-addition.

Addition-elimination reactions of aryl halides

The following reaction looks a bit unusual based on what weve learned about aromatic
chemistry thus far. The first striking thing is that the ring appears to be interacting with a
nucleophile (methoxide, CH
3
O). In EAS reactions, it is the ring itself which behaves as the
nucleophile. This is the first clue that there must be a different mechanistic paradigm at play
here.



Additionally, we see that the nucleophile is replacing a halogen on the ring. This too is
surprising because in EAS reactions only a hydrogen could be replaced. This reaction looks like
a substitution, but we know that S
N
1 and S
N
2 reactions do not occur on sp
2
-hybridized centers.
So, this is a new type of substitution reaction. Its called nucleophilic aromatic substitution, or
S
N
Ar.

129

There are some empirical observations that provide clues about the mechanism as well. For
instance, having a nitro group ortho or para to the halide results in a relatively fast reaction, but
the rate of reaction is reduced if the nitro group resides in the meta position. However, even
when the nitro group is meta to the halogen, the reaction is still faster than it would be for an
unsubstituted aryl halide. Also, the more nitro groups the ring has, the faster the reaction
proceeds.

NO
2
X
Rate of S
N
Ar reaction increases
X
NO
2
X
NO
2
X
NO
2
X
NO
2
NO
2
X
NO
2
O
2
N


These facts reinforce our initial impression that the polarity of the reactants is the reverse of what
we observed in EAS. The ring is now behaving as an electrophile, so the presence of electron-
withdrawing groups (e.g. NO
2
) enhances the rings electrophilicity and accelerates the reaction.
Additionally, the dependence of rate on the location of the electron-withdrawing group suggests
that resonance will play a role in the mechanism.

The kinetics of the reaction provide some insight into the rate-determining step of the
mechanism. S
N
Ar reaction follows second-order kinetics. The rate of the reaction depends on
the concentration of both the aryl halide and the nucleophile, which means that both must be
mechanistically involved in the rate-determining step of the reaction.

Rotc = k |oryl oliJc] |nuclcopilc]

Finally, another significant observation is that the order of aryl halide reactivity is the reverse of
the order of alkyl halide reactivity in S
N
1 and S
N
2 reactions. While alkyl fluorides give the most
sluggish substitution reactions, aryl fluorides proceed through substitution more rapidly than
another other aryl halide.

130



Taken together, all of these pieces of experimental evidence suggest a mechanism that involves
addition of the nucleophile to the carbon of the ring that bears the leaving group, followed by
subsequent loss of the leaving group in a second mechanistic step. The intermediate anion is
known as the Meisenheimer complex.



Initial attack of methoxide or another nucleophile on the carbon bearing the leaving group pushes
a -bonding pair of electrons to the adjacent carbon. This anion can resonate to the ring carbons
ortho and para to the leaving group. Finally, the unshared pair re-forms a bond, extruding the
halide, to yield the substitution product.

This mechanism is consistent with a rate acceleration that results from an electron-withdrawing
group on the ring. The electron-withdrawing group helps to stabilize the Meisenheimer complex
by removal of electron density through induction and/or resonance. An ortho or para electron-
withdrawing group that contains a bond provides a more pronounced rate acceleration because
of the additional resonance stabilization of the anion.



And, of course, additional electron-withdrawing groups would lead to cumulative stabilization of
the Meisenheimer complex. Its really important to note that electron-withdrawing groups,
which were deactivating for EAS, are activating for S
N
Ar. This is due to the reversal of the
aromatic rings role in S
N
Ar.
131


The second-order rate law shows that the first step of the reaction must be the rate-determining
step. Only the first step of the reaction mechanistically involves both the aryl halide and the
nucleophile. This conclusion also provides the explanation for the reversal of the rate of reaction
when compared to S
N
1 or S
N
2 reaction. In S
N
1 and S
N
2 reactions, the rate-determining step
involves the loss of leaving group (either before the attack of nucleophile or concurrently with
the attack of nucleophile). Therefore, the leaving group ability of the halide has a pronounced
influence on rate. However, in S
N
Ar reaction, the halogen is not lost during the rate-determining
step, so its leaving group ability is less significant. Instead, what matters most is its ability to
draw in a nucleophile. Fluorine, being the most electronegative halogen, creates the most
polarized carbon-to-halogen bond, and the carbon with the greatest
+
charge attracts the
nucleophile most intensely, thereby accelerating the reaction.

F Cl Br I
= dipole
= partial positive charge
Rate of S
N
Ar reaction increases


Examples of addition-elimination reactions (S
N
Ar)

There are two main variations that you may encounter in different S
N
Ar reactions. The
nucleophile can vary, and the electron-withdrawing groups on the ring can vary.

Earlier, we saw that methoxide could serve as the nucleophile in addition-elimination reactions
to yield a substituted anisole.



It stands to reason then that hydroxide could also perform this function, and the result is a
phenol. The hydroxide may come directly from sodium hydroxide, or it can be produced from
an aqueous solution of sodium carbonate (Na
2
CO
3
).

132

NO
2
NO
2
Cl
H
2
O,
NO
2
NO
2
OH
OH


Neutral nucleophiles, such as ammonia, can also participate in the reaction to provide aniline
derivatives.



With a sufficiently activated aromatic ring, it is even possible for a weak neutral nucleophile,
like water, to add. In the specific example shown below, the product is picric acid. Picric acid is
an explosive compound, as are many heavily nitrated molecules. It was used in munitions prior
to trinitrotoluene (or TNT), but it is difficult to detonate when wet and forms shock-sensitive
picrates, which cause shells to explode on contact rendering them ineffective for the penetration
of armor.



It is also possible to see variation in the nature of the electron-withdrawing group on the ring.
The nitro group appears commonly in this capacity because it is powerfully electron-
withdrawing, but it is not the only possible choice. A heavily halogenated aromatic ring can
experience S
N
Ar displacement of one of its halogens.

F
F F
F
F
F
CH
3
O
CH
3
OH,
F
F
OCH
3
F
F
F


In this case, the Meisenheimer complex is stabilized by the inductive electron withdraw of the
halogens. In each intermediate resonance form, the negative charge resides on a carbon bearing
an inductively electron-withdrawing fluorine which lends stability, as does the inductive electron
withdraw of the other fluorine atoms nearby.

133



Aromatic heterocyclic molecules can also participate in S
N
Ar reaction.



Here electron-withdrawing groups are not necessarily needed on the ring. The heteroatom in the
ring provides the stabilization by bearing the anion in one of the resonance forms. The diagram
below shows only that most prominent resonance contributor to the hybrid, but of course the
anion can also be delocalized onto two carbons of the ring to generate lesser resonance
contributors.



S
N
Ar and Agent Orange

During the Vietnam War, Agent Orange was used by the United States as a defoliant. The goal
was to make guerilla warfare more difficult by eliminating the cover in which guerilla fighters
would conceal themselves. The name derived from the orange stripe on the barrels used in
shipping. Agent Orange was a mixture of two compounds, known as 2,4-D and 2,4,5-T.



The herbicide 2,4,5-T is produced from 1,2,4,5-tetrachlorobenzene in a two-step sequence. The
first step entails displacement of a chloride with hydroxide via S
N
Ar reaction. The resultant
phenol loses its proton in basic media to become the phenoxide ion. In the second step, the
phenoxide ion displaces the chloride of chloroacetate via S
N
2 reaction to yield 2,4,5-T.

134



If the temperature is not carefully controlled and rises too high, a side reaction of the phenoxide
ion results. Two phenoxide ions can condense through two S
N
Ar reactions, the first being
intermolecular and the second being intramolecular. The product that results is often simply
called dioxin; however, this is an abbreviated form of its correct name, 2,3,7,8-
tetrachlorodibenzo-p-dioxin.



2,3,7,8-Tetrachlorodibenzo-p-dioxin is carcinogenic, and it is also quite persistent in the
environment once it has been introduced. Many serious health conditions of Vietnam War
veterans and of the Vietnamese people have been attributed to 2,3,7,8-tetrachlorodibenzo-p-
dioxin exposure due to extensive and prolonged spraying of forested areas with Agent Orange.

Elimination-addition reactions of aryl halides

Weve been discussing the addition-elimination reactions of aryl halides, but it turns out that aryl
halides can also undergo elimination-addition transformations. The discovery of this stems from
experimental observations that were perplexing and inconsistent with the S
N
Ar paradigm.

Sodamide or potassium amide will undergo reaction with chlorobenzene in liquid ammonia (-33
C) to yield aniline.



Additionally, phenol can be produced from the analogous reaction of chlorobenzene with
aqueous hydroxide under high temperature and pressure.



135

John D. Roberts is famous for the C-14 labeling studies which confirmed that not only the
carbon bearing the halogen but also the neighboring carbons are involved in the reaction.
9

Roberts experiment showed that the addition-elimination mechanism could not be at play in
these situations. The S
N
Ar mechanism only allows for substitution at the carbon bearing the
leaving group. However, the products that Roberts obtained in equal amounts contained the
amino group at the labeled carbon and adjacent to it.



This phenomenon can be observed regiochemically as well. The following reactions are
inconsistent with the addition-elimination motif for multiple reasons. First, the rings are
deactivated toward S
N
Ar. In other words, the rings bear electron-donating groups, rather than the
electron-withdrawing groups that would favor S
N
Ar reaction. Additionally, mixtures of isomers
result from each reaction, and this regiochemical outcome is not predicted by the addition-
elimination mechanism.

o-Bromoanisole yields a pair of regioisomers, o- and m-anisidine (i.e. 2- and 3-methoxyaniline).



Similarly, p-bromoanisole yields both m- and p-anisidine.



m-Bromoanisole presents an even greater conundrum, since it yields all possible anisidine
regioisomers.



9
For a fascinating description of this work and Roberts career, see A Perspective Distilled
from Seventy Years of Research, Journal of Organic Chemistry, 2009, 74(14), 4897 4917.
136


The only way to reconcile the empirical results with a mechanistic theory is to consider
elimination-addition. The first step of this process, elimination, results in the formation of an
unusual intermediate, known as benzyne. Depending on the aryl halide used, the elimination
may be drawn as a stepwise or concerted process. In the stepwise sequence, the amide ion
deprotonates the aromatic ring adjacent to the carbon bearing the leaving group. The
intermediate phenyl anion extrudes the halide leaving group to produce benzyne.



In the concerted approach, there is concomitant deprotonation and loss of leaving group to afford
benzyne directly.



Benzyne is a very surprising intermediate. The presence of an apparent triple bond in a six-
membered ring is not consistent with geometric constraints. After all, sp-hybridized carbon
atoms adopt 180 bond angles, which cannot be accommodated in six-membered rings.
However, the second bond is not a classical bond between adjacent p orbitals. Instead,
adjacent sp
2
hybrid orbitals each house one electron. There is an overlap between them but to a
lesser extent than we would find in a traditional bond.



The second step of the reaction is addition. Another amide anion attacks the weak bond of
benzyne, forming a bond to the ring and a carbanion on the adjacent center. This phenyl anion
then deprotonates the solvent, ammonia, to regenerate one amide anion.




137

Benzyne isomers and their involvement in other reactions

Benzyne, or more broadly arynes, need not necessarily have the two additional electrons
adjacent, or ortho, to one another. The other isomers, meta- and para-benzyne, are also possible.



o-Benzyne can be generated in elimination-additions reactions with reagents other than amide
ion, allowing the synthesis of other aromatic derivatives. For instance, treatment of an aryl
halide with phenyllithium can lead to the production of benzyne via either stepwise or concerted
elimination, depending on the halide chosen.

Stepwise pathway:



Concerted pathway:



A second phenyllithium can then add to the weak bond. In this case, there is no protic solvent
to protonate the phenyl anion. Consequently, it can be treated in a separate step with either acid
or an electrophile to quench the anion and form product.



o-Benzyne can also be generated during reactions other than dehydrohalogenation and can
therefore be used for transformations other than the additions seen above. One example is the
138

elimination of two halogens from an ortho-dihalobenzene using magnesium or lithium.
Oxidative insertion of the metal into one carbon-to-halogen bond produces the Grignard or
organolithium species, depending on the choice of metal. The electrons in the carbon-to-metal
bond can then form the additional bond of benzyne as the other halide is displaced from the
molecule.



Benzyne may then be used in a subsequent reaction, such as [4+2] cycloaddition (i.e. Diels-Alder
reaction). In the transformation shown below, benzyne acts as the dienophile and adds to
cyclopentadiene to form a bicyclic adduct.



To this point, weve focused only on ortho-benzyne, but there is also a very well-known instance
where para-benzyne plays a pivotal role. The Bergman cyclization of enediynes produces para-
benzyne.



In the presence of a hydrocarbon or other hydrogen atom source, the para-benzyne will abstract
two hydrogens to form benzene.



Isotopic-labeling studies provide some evidence for this mechanism. If the alkynes of the cis-
enediyne bear deuterium atoms, those deuterium atoms can also be observed in the vinylic
position through equilibration.



139

The reaction requires a good deal of heating (200 C), but ring strain can be used to elevate the
energy of the reactant thereby reducing the activation energy barrier to reaction. The enediyne of
a ten-membered ring undergoes Bergman cyclization at much more reasonable temperatures.



Dynemicin A is a natural product isolated from bacteria in a soil sample from India. It contains
an enediyne moiety. Upon intercalation into DNA and cleavage of the epoxide, the
conformational change that results allows Bergman cyclization to take place. The resultant para-
benzyne abstracts hydrogen atoms from DNA, ultimately leading to DNA cleavage and cell
death.
10




Chapter 6 Problems

1. Okuma and co-workers reported the reaction of benzyne with salicylaldehydes to prepare
xanthene derivatives.
11
Xanthene derivatives are utilized as dyes, and they are also found in
natural products and pharmaceuticals.

They generated benzyne from ortho-trimethylsilylphenyltriflate as shown below:




10
For more information about the isolation, mechanism of action, and total synthesis of
Dynemicin A, see Chapter 4 of Classics in Total Synthesis II by K. C. Nicolaou and Scott A.
Snyder (Wiley: Germany, 2003).
11
Okuma, K.; Nojima, A.; Matsunaga, N.; Shioji, K. Organic Letters, 2009, 11(1), 169-171.
140

Provide a mechanism for this transformation. As you do this, it will be helpful to know that
fluoride possesses a high affinity for silicon.

2. The benzyne prepared in the problem 1 was then treated with salicylaldehyde. Provide a
mechanism, using curved-arrow notation, to explain the generation of the xanthene product
shown.



3. Smith and Kim of the University of Pennsylvania reported a [4+2]cycloaddition (or Diels-
Alder reaction) between a substituted benzyne and a furan to provide the product shown below.
12

What reactants fitting this description would give rise to this product?



4. A more complete depiction of the reaction referenced in problem 3 is shown below. The
ketone substrate is treated with methyllithium. Then, a Brook rearrangement occurs. The Brook
rearrangement is an intramolecular transfer of a silicon-containing group from carbon to oxygen.
It is driven by the formation of a more favorable Si-O bond. The Brook rearrangement sets the
stage for benzyne formation and [4+2]cycloaddition. Finally, in step 2, tetrabutylammonium
fluoride is used as a fluoride source to cleave the trimethylsilyl group from the molecule (note
that there is some similarity here to the use of CsF in problem 1).

Based on the information provided above, show a complete mechanism for this transformation.



12
Smith, A. B., III; Kim, W.-S. Proceedings of the National Academy of Science, 2011, 108(17),
6787-6792.
141


5. In the Chapter 5 problems, we prepared 1-chloro-3-fluorobenzene, which used in a synthesis
of analogues of Chantix (shown below).
7




In this synthesis, 1-chloro-3-fluorobenzene is subjected to ortholithiation to prepare a substituted
benzyne. The halogens guide butyllithium in to the central proton, which is removed by this
very strong base to give the intermediate shown in brackets. Show mechanistic arrows to explain
how the intermediate shown in brackets becomes a substituted benzyne.



6. The substituted benzyne prepared in the previous problem is then treated with a
cyclopentadiene derivative as shown below. Provide the product that results from this Diels-
Alder cycloaddition.



7. Chloroquine has been used as an alternative to quinine in the treatment and prevention of
malaria. Chloroquine can be formed from the reaction between 4,7-dichloroquinoline (A) and
the diamine B shown below.

Provide a mechanism for this transformation.



142

8. In 2010, Hu and co-workers of the East China University of Science and Technology in
Shanghai reported the synthesis of rod-coil brush polymers in the journal Macromolecules.
13

The brush polymers can be represented using the following schematic.



The straight portions of the schematic are formed by linked aromatic rings; whereas, the squiggly
portions of the diagram denote ester-containing chains.



The aromatic portion is formed by Bergman cyclization of the following molecule. Show the
para-benzyne that results from this Bergman cyclization, as well as a mechanism for its
formation.



9. A team of investigators at the University of Pennsylvania and the University of the Sciences in
Philadelphia developed a chloroquine analogue (Lys05, shown below) that inhibits autophagy.
14

Autophagy is a process through which cells use the lysosome to repair damage, protect
themselves from reactive oxygen species, and recycle intermediates needed to sustain their
metabolism in response to stress.
15
Cancer cells rely heavily on this pathway, so it is hoped that
this compound may prove to be therapeutically useful.


13
Cheng, X.; Ma, J.; Zhi, J.; Yang, X.; Hu, A. Macromolecules, 2010, 43(2), 909-913.
14
McAfee, Q.; Zhang, Z.; Samanta, A.; Levi, S. M.; Ma, X.-H.; Piao, S.; Lynch, J. P.; Uehara,
T.; Sepulveda, A. R.; Davis, L. E.; Winkler, J. D.; Amaravadi, R. K. Proceedings of the National
Academy of Science, 2012, 109(21), 8253-8258.
15
Borman, S. Chemical & Engineering News, May 14, 2012, page 7.
143



Imagine that investigators attempting to prepare analogues discover that the 3-chloro isomer
needed as a reactant is significantly less reactive toward nucleophiles than the 4-chloro isomer
(as shown below).



Using a generic nucleophile, Nuc, draw structures of the reaction intermediates to explain why
this is the case.

10. Calicheamicin is a compound with potent DNA-cleaving properties. It was isolated from
bacteria found in chalky rocks near a Texas highway. Its extreme potency toward cancer cells
led to efforts to prepare this complex natural product.
16


Below is the core of calicheamicin as it looks at the critical DNA-cleaving stage. Bergman
cyclization follows. Draw the result of Bergman cyclization below. You can leave the product
at the para-benzyne stage (i.e. prior to H abstraction from DNA).





16
For an excellent discussion of this, see Chapter 30 of Classics in Total Synthesis by K. C.
Nicolaou and E. J. Sorensen (VCH: Weinheim, 1996).
144

11. Parker and co-workers at AstraZeneca published a paper in the journal Organic Process
Research & Development describing the preparation of compounds with therapeutic potential for
type II diabetes.
17
The final step in the synthesis of one of these compounds is shown below.
This step involves an addition-elimination mechanism. Use curved-arrow notation to illustrate
the mechanism of this process, and provide the structure of the product.







17
Parker, J. S.; Bower, J. F.; Murray, P. M.; Patel, B.; Talavera, P. Organic Process Research &
Development, 2008, 12(6), 1060-1077.
145




Part III.
Radical Reactions
146




Chapter 7:
Basic radical chemistry
147

Radicals: formation, stability, and behavior

Radicals (also known as free radicals) are formed when bonds undergo homolytic cleavage. The
majority of reactions youve studied in Introductory Organic Chemistry and thus far in our
current discussion involve heterolytic (or uneven) bond cleavage, in which both electrons of a
given bond flow to one atom as the bond breaks. In homolytic (or even) bond cleavage, each
atom acquires one of the bonding electrons as the bond breaks. Homolytic bond cleavage is the
reaction described by bond dissociation energies (BDE).



The homolytic cleavage of a bond results in species with unpaired electrons. These are termed
radicals or free radicals. Often, these radicals will not possess formal charges; nevertheless, they
are electron-deficient species because they lack a complete octet. This makes them reactive
intermediates.



The electron deficiency can also be illustrated by considering the similarity in bonding patterns
between a free radical and a carbocation. Both are sp
2
-hybridized centers and therefore have
trigonal planar geometry. While the carbocation has a completely empty unhybridized p orbital,
the free radical has a singly occupied p orbital. The carbocation has a sextet of electrons around
it, and the free radical is surrounded by a septet of electrons.



Incidentally, this trigonal planar geometry means that the stereochemical ramifications of radical
reactions are the same as those of reactions involving carbocation intermediates. In short, there
is no reason to expect a preference for reaction on one face of the molecule as opposed to the
other, so if a radical reacts to form a stereocenter, both configurations are expected.


148


Since radicals are electron-deficient species, their stability parallels that of carbocations, which
are also electron deficient. The more highly substituted the radical, the more stable it is.
Additionally, as with carbocations, resonance stabilizes radicals



These differences in stability are reflected in bond dissociation energies (BDE). For example,
there are two types of carbon-to-hydrogen bonds in propane: primary and secondary. The
secondary carbon-to-hydrogen bond has a lower BDE since the cleavage of this bond produces
the more stable secondary radical.



The same information can be communicated graphically as shown in the reaction energy profile
below. Both reactions begin with the same reactant, propane. While both bond cleavage events
are endothermic, it requires more energy to form the less stable primary radical than it does to
form the radical at the more highly substituted secondary center.

149

H +
H +
BDE = 98 kcal / mole BDE = 95 kcal / mole
Less stable
primary radical
More stable
secondary radical
Reaction Coordinate


One significant difference between radicals and carbocations is that radicals do not rearrange as
carbocations do. In the following example, a secondary carbocation rearranges to a tertiary
carbocation through 1,2-hydride shift.



The corresponding migration of a radical intermediate does not occur.



The rational for the difference is provided by molecular orbital theory. A carbocation
rearrangement is a three-center, two-electron process, in which the two involved electrons
occupy a bonding molecular orbital.

150



The analogous process for a free radical would be a three-center, three electron process. The
instability of the electron in the partially filled anti-bonding orbital raises the energy barrier to
rearrangement.



Continuing our comparisons between heterolytic and homolytic processes, there are differences
in the way that the two types of mechanisms are drawn. In heterolytic mechanisms, the
progression from reactant to intermediate(s) to product is conveyed as a series of steps, one
immediately after the other.



On the other hand, the steps of homolytic mechanisms are divided and drawn separately. This
allows the classification of each step as initiation, propagation, or termination.


151


As their name implies, initiation steps start the radical process. They are often steps that begin
without radicals but produce radicals. The spontaneous homolysis of a bond requires an input
of energy from light (h) or heat (). We will also see that sometimes a step that begins and
ends with radicals will be termed initiation if it produces the key radical that is active in the
subsequent propagation steps.



Propagation steps begin and end with radicals. In these steps, there are two primary options for
the free radical. It can incite the homolysis of a bond, or it can add to a bond. In either case,
the radical process continues (or propagates) because a new radical results.



In termination steps, any two radicals combine, resulting in a compound with no unpaired
electrons.



Radical substitution reactions

Lets consider the reaction of ethane with bromine. Overall, the process yields bromoethane and
HBr. The net result is that a hydrogen of an alkane has been substituted with a halogen.



152

This is quite important because it enables the functionalization of alkanes, which are otherwise
relatively unreactive. Once functionalized as alkyl halides, there are many transformations that
can be used to produce a wide variety of compounds.



The reaction begins with the homolysis of bromine upon exposure to light or heat. It is critically
important though to realize that only a small number of bromine molecules undergo the
homolysis. Most remain unreacted and will be encountered again in the second propagation step.



The propagation then begins as the electron-deficient bromine radical offers its electron and the
hydrogen of ethane reciprocates with one electron, leading to the formation of HBr and an ethyl
radical.



In the second propagation step, the ethyl radical incites the homolysis of bromine forming ethyl
bromide and a new bromine radical.



There are a few important comments regarding this step of the reaction. First, students are
sometimes surprised to see Br
2
again in this step because they feel as though it was consumed in
the initiation step. However, as we noted earlier, only a few molecules of Br
2
homolyze during
initiation. Most are unreacted and are consumed during the second propagation step.
Additionally, product is formed in this step, and in fact this is where the vast majority of product
results from. You might feel as though the product should be formed at the end of the reaction,
153

in the steps labeled termination, but in reality termination steps merely explain the fate of the few
residual radicals that remain after propagation concludes. Finally, to further illustrate this point,
bromine radical is regenerated during the second propagation step. Notice that bromine radical
is used in the first propagation step but reformed in the second propagation step. This
phenomenon makes the process a chain reaction. Bromine radical enters the propagation
sequence and ultimately produces product while replicating itself so that it can enter the cycle yet
again. This cycle repeats time and again until all of the ethane and bromine are consumed. You
could even illustrate this with a diagram that is more evocative of the cyclic nature of the
propagation sequence.

Br C H
3
C H
H
H
C H
3
C
H
H
H Br
Br Br
Product formed
Propagation cycle
Reactant consumed
Reactant consumed
Product formed
Bromine radical
continually regenerated
Step 1 Step 2
C H
3
C
H
H
Br


The termination steps are almost an afterthought in this case. They arent particularly important
if your goal is merely to explain mechanistically how the products are formed. The termination
steps arise from a very logical question. This propagation cycle cannot continue forever.
Eventually, both reactants are consumed, so you might wonder: what happens to the remaining
radicals at that point? These radicals combine in the terminations steps. Any pair of radicals can
bond, so there are a number of termination possibilities.

154



The reactivity-selectivity principle

The free-radical halogenation of propane differs from that of ethane in that there is a
regiochemical consideration. In principle, it is possible to halogenate at the terminal or internal
carbon to make either 1-halopropane or 2-halopropane.



The choices of halogen are typically limited to chlorine and bromine. Radical halogenation with
fluorine is severely exothermic, making it a potentially explosive reaction; whereas, the reaction
with iodine is not favorable enough to be synthetically useful.

There is a pronounced difference in the selectivity of the chlorination and bromination reactions.
Free-radical chlorination of propane yields roughly a 1 : 1 mixture of 1-chloropropane and 2-
chloropropane



On the other hand, the bromination reaction produces 2-bromopropane almost exclusively


155


To account for this dramatic difference in selectivity, we need to consider the energy profile for
each reaction. Lets examine chlorination first. Remember that Gibbs free energy changes
during a reaction depend upon changes in both enthalpy and entropy.

0

= E

IS

However, with many organic reactions, the entropy term is negligible. This is true in this
reaction where two reactant molecules (propane and chlorine) form two product molecules (alkyl
chloride and hydrochloric acid). Consequently, the change in Gibbs free energy can be
approximated as the change in enthalpy for the reaction.

0

The change in enthalpy is equal to the difference in energy between the bonds broken and
formed.

E

= |E

bonJs brokcn] |E

bonJs ormcJ]

Lets focus for the moment on the chlorination that proceeds through the more stable secondary
radical to yield 2-chloropropane. The relevant bond dissociation energy values are provided in
the table below.

Bond Bond Dissociation Energy (kcal/mole)
(CH
3
)
2
CH-H 95
Cl-Cl 58
(CH
3
)
2
CH-Cl 80
H-Cl 103

The reaction is overall exothermic and releases 30 kcal/mole of energy.

E

= |9S + S8] |8u +1uS] = Su kcolmolc

We can be quite specific about the change in energy at each stage of the reaction as well. The
propagation steps constitute the bulk of the process, as we saw earlier, so well only consider
those two steps. In the first propagation step, the carbon-hydrogen bond is broken and the H-Cl
bond is formed.



This step is exothermic and releases 8 kcal/mole.

E

= |9S] |1uS] = 8 kcolmolc

156


In the second propagation step, the Cl-Cl bond is broken and the carbon-to-chlorine bond is
formed.



This step is also exothermic, releasing 22 kcal/mole.

E

= |S8] |8u] = 22 kcolmolc

Both steps are plotted on the following energy diagram.



To address the selectivity (or lack thereof in this case), we need to consider the energetic
difference between forming the primary and secondary radicals in the first propagation step. We
already know that the formation of the secondary radical releases 8 kcal/mole. The following
table lists the bond dissociation energies needed to determine the change in enthalpy for the
formation of the primary free radical.

Bond Bond Dissociation Energy (kcal/mole)
CH
3
CH
2
CH
2
-H 98
H-Cl 103

The only value that differs is the bond dissociation energy of the carbon-to-hydrogen bond. This
primary carbon-to-hydrogen bond is slightly stronger than the equivalent bond at the secondary
center. Consequently, the first propagation step releases slightly less energy (5 kcal/mole) as the
primary radical is formed.

E

= |98] |1uS] = S kcolmolc

157


These two possibilities for the first propagation step are plotted together on the energy diagram
below. To sketch the transition states for these two reactions, it is important to remember the
guiding principle provided by the Hammond postulate, which predicts that the transition states of
exothermic reactions will more closely resemble the reactants. If that is the case, then the 3
kcal/mole difference in energy between the two free radical products has little relevance for the
transition states. The transitions states for these exothermic reactions are quite close in energy
because they resemble the common reactant from which both processes originate. Since the
transitions states are very close in energy, theres no selectivity for one pathway versus the other.



Lets examine free-radical bromination in an analogous fashion. The bond dissociation energies
needed to calculate the changes in energy for the formation of 2-bromopropane are found in the
following table.

Bond Bond Dissociation Energy (kcal/mole)
(CH
3
)
2
CH-H 95
Br-Br 46
(CH
3
)
2
CH-Br 68
H-Br 88

In the initial propagation step, the secondary carbon-to-hydrogen bond is broken, and the HBr
bond is made.



158

This results in an endothermic reaction, which requires an energetic input of 7 kcal/mole.

E

= |9S] |88] = +7 kcolmolc

The final propagation step involves cleavage of bromine and formation of the secondary carbon-
to-bromine bond.



This step is exothermic and releases 22 kcal/mole.

E

= |46] |68] = 22 kcolmolc

The energy profile for this reaction looks very different from that of chlorination, in which both
propagation steps were exothermic. Now, in bromination, the first step is endothermic instead.
This, it turns out, makes all the difference.



Lets now consider solely the first propagation step and how it differs depending on whether the
primary or secondary radical is formed. We already know that formation of the secondary
radical requires an input of 7 kcal/mole. The bond dissociation energies needed to calculate the
change in enthalpy associated with formation of the primary radical are found in the table below.
Again, all that differs is the carbon-to-hydrogen bond strength, since now we are considering
cleavage at the primary center.

Bond Bond Dissociation Energy (kcal/mole)
CH
3
CH
2
CH
2
-H 98
H-Br 88

159

The primary carbon-to-hydrogen is broken, while the HBr bond is formed, resulting in an
endothermic process requiring 10 kcal/mole.

E

= |98] |88] = +1u kcolmolc

These two propagation step 1 alternatives are plotted together in the energy diagram below. As
we sketch the transition state, it is still important to remember the Hammond postulate, which
predicts that the transition states of endothermic reactions will more closely resemble the
products. Theres a 3 kcal/mole difference in energy between the two radicals, just as there was
for the chlorination reaction. The difference is that, in bromination, that energy gap is expressed
more prominently at the transition states, which now are more similar to products than reactants.
Consequently, there is a much greater selectivity for the lower energy pathway, resulting almost
exclusively in the formation of the secondary radical and hence 2-bromopropane

HBr +
HBr +
E
n
e
r
g
y
Reaction Coordinate
+ Br
+
+ Br
2
Br
2
Propagation step 1,
formation of a
primary radical
H
o
= +10 kcal / mole
Propagation step 1,
formation of a
secondary radical
H
o
= +7 kcal / mole


The phenomenon that weve discussed in depth here has sometimes been termed the reactivity-
selectivity principle. This principle states that often when a species is more reactive the
selectivity of the reaction is reduced. The principle is certainly not without exception,
18
but it
does help us to attach a name to what has been observed with free-radical halogenation.

Radical substitution reactions at allylic or benzylic centers

Thus far, weve discussed radical stability solely from the perspective of alkyl group
substitution. The more highly substituted the radical, the more stable it is. It stands to reason
that resonance should also be a powerful stabilizing factor when it is present. For instance, in the

18
For more discussion of this, see The Reactivity-Selectivity Principle: An Imperishable Myth
in Organic Chemistry, Angewandte Chemie, International Edition, 2006, 45(12), 1844 1854.
160

allylic system, propene, homolysis of an sp
3
carbon-to-hydrogen bond results in a free radical
that can be resonance delocalized to the other terminus of the allylic system.



The benzylic system naturally affords even greater resonance delocalization. Both allylic and
benzylic radicals are even more stable than tertiary radicals.



This fact is reflected in their bond dissociation energies. Increasing radical stability results in
continually lower bond dissociation energies.

R
C R
R
H
C R
R
H
C H
R
H
C H
H
Decreasing Bond Dissociation Energy (BDE)
methyl primary secondary tertiary allylic benzylic
H H H H
H H
BDE =
103
kcal/mole
BDE =
98
kcal/mole
BDE =
95
kcal/mole
BDE =
93
kcal/mole
BDE =
88
kcal/mole
BDE =
85
kcal/mole


Ultimately, all of this means that, when an allylic or benzylic hydrogen is present, it should be
particularly reactive in radical substitution reactions. Since we know that alkenes can react with
both X
2
and HX via electrophilic addition reactions, which you learned about in Introductory
Organic Chemistry, it would be prudent to select alternative conditions for these reactions. A
reagent known as N-bromosuccinimide (NBS) supplies a constant low concentration of bromine
and consumes HBr as it is formed, making it particularly convenient for this transformation.



161

The sequence initiates with the homolysis of the weak nitrogen-to-bromine bond upon exposure
to light or a radical initiator. Again, it is important to remember that this initiation step occurs
with only a small number of N-bromosuccinimide molecules. Most NBS molecules remain
unreacted, and their fate is explained later in the mechanism.



The bromine radical thus formed enters propagation step 1, where it abstracts a hydrogen from
the benzylic carbon, leading to the formation of the most stable radical possible.



The HBr then reacts with NBS to form bromine. This is an ionic step. It is rather unusual to see
ionic steps amidst what is otherwise a radical mechanism. The HBr, being a strong acid,
protonates NBS that did not homolyze during initiation. The bromide that is released during this
step then attacks the weak nitrogen-to-bromine bond, cleaving it to form succinimide and
bromine.



This process ensures two things: (1) HBr is consumed as it is formed and (2) a constant, low
concentration of bromine is available throughout the reaction. Now that bromine has been
generated, it interacts with the benzylic radical in propagation step 2, leading to product
formation. Also in this step, the bromine radical needed for propagation step 1 is regenerated so
that the cycle can repeat.



162

The use of NBS is more critical when the substrates are alkenes, which readily undergo ionic
reaction with both HBr and Br
2
. If an allylic substrate were used and HBr were not consumed as
it was formed, hydrohalogenation (electrophilic addition of HX across the bond of an alkene)
would lead to an undesired side product.



Similarly, if a large concentration of Br
2
were present, halogenation (electrophilic addition of X
2

across the bond of an alkene) would also lead to an undesired side product.



To achieve solely radical substitution with an allylic substrate, it is important to use NBS as the
brominating agent.

H
NBS
h
Br


Radical addition to bonds

Early in this chapter, it was noted that radicals could cause the homolysis of bonds or they
could add to bonds. However, to this point, weve focused exclusively on the interaction of
radicals with bonds in radical substitution reactions. Now, lets turn our attention to the
addition reactions that radicals can undergo.

We saw a reminder above that, when alkenes undergo hydrohalogenation, Markovnikov
selectivity results.



You might recall from Introductory Organic Chemistry that, when peroxides are added to this
reaction, the regiochemistry is reversed, and anti-Markovnikov selectivity is observed.



163

Peroxide homolyzes in the presence of light or heat to begin the initiation sequence. The peroxy
radical thus formed abstracts hydrogen from HBr to yield an alcohol and the critical bromine
radical that will cycle through the propagation steps.



The bromine radical adds to the bond so as to form the more stable secondary radical
intermediate. Note that we arent breaking any rules of stability in order to achieve anti-
Markovnikov regiochemistry. What differentiates this reaction from its ionic counterpart is
merely what adds first to the alkene. In the ionic reaction, a proton is added to the alkene first;
whereas in the radical process, it is the bromine that adds first to the alkene. In both instances,
the addition occurs so as to give the more stable intermediate (carbocation for the ionic process
and radical here), but the differing initial bond formation leads to the reversal of
regiochemistry.



In the second propagation step, the carbon-centered radical incites the homolysis of a molecule
of HBr, leading to the formation of product and the regeneration of the critical bromine radical.



It is worth taking a step back at this moment to answer a question that might be on your mind.
During the first propagation step of radical substitution, bromine radical abstracts an allylic or
benzylic hydrogen atom. During the same step of radical addition, bromine radical adds to the
bond of an alkene instead. So, naturally you might wonder how you are supposed to know when
bromine radical interacts with a bond or a bond.

164



The answer comes from looking at the big picture. In radical substitution, the alkene is treated
with NBS, which is effectively a source of a steady and low concentration of Br
2
. On the other
hand, in radical addition the alkene is treated with HBr. The difference is reagent, Br
2
versus
HBr, must somehow explain the difference in reactivity.



Lets consider what would happen if bromine radical abstracted the allylic hydrogen during
radical addition. To help you put this what if? scenario in context, the actual, correct
mechanism for radical addition is summarized below. Take special note of what happens during
propagation step 1: bromine radical adds to the bond.

165

+ Br
Propagation step 2
Br
H Br +
Br
H
Product formed
Bromine radical regenerated
Initiation
RO OR 2 RO
H Br RO + RO H
Propagation step 1
+ Br
Br
h or
+ Br
Radical addition as it
actually occurs


Now, lets consider our what if? scenario. What if bromine radical abstracted an allylic
hydrogen during propagation step 1? The answer is shown below. This would result in the re-
formation of HBr and the production of an allylic radical. In propagation step 2, the allylic
radical would abstract a hydrogen from another molecule of HBr to re-form the initial organic
substrate, along with bromine radical.

+ Br
Propagation step 2
H Br +
Reactant
re-formed
Bromine radical regenerated
Initiation
RO OR 2 RO
H Br RO + RO H
Propagation step 1
+ Br
h or
+ Br
What if hydrogen
abstraction happened in
propagation step 1?
H
H Br +
H
Reactant
re-formed


166

So, while bromine radical could abstract a hydrogen in propagation step 1 of the radical addition
reaction, we would be unable to detect that this process had occurred because all that results is
the re-formation of our original reactants. The only reaction of bromine radical in the addition
process that leads to new products is its addition to the bond of the alkene.

Weve seen that radical substitution reactions involving chlorine or bromine are
thermodynamically favorable, although one is more selective than the other. You might also
wonder whether the radical addition of HBr to the alkene bond is favorable. The bond
dissociation energy values needed to ascertain this are provided in the table below.

Bond Bond Dissociation Energy (kcal/mole)
C-C 83
C=C 146
(CH
3
)
2
CH-H 95
CH
3
CH
2
CH
2
-Br 68
H-Br 88

As before, well be considering the difference in enthalpy between the bond broken and formed.
However, youll notice that only the bond of the alkene is broken, and the table above does not
contain a bond dissociation energy for the bond. This value is readily calculated though. It is
merely the difference in bond dissociation energy of the C=C and C-C bonds (146 83 = 63
kcal/mole).

E

= |E

bonJs brokcn] |E

bonJs ormcJ]

Therefore, we now have all of the values needed for the calculation.

E

= |6S + 88] |68 + 9S] = 12 kcolmolc

Furthermore, we can determine that both propagation steps of radical addition are exothermic. In
propagation step 1, the bond is broken and the carbon-to-bromine bond is formed.



This results in a step that is exothermic by 5 kcal/mole.

E

= |6S] |68] = S kcolmolc

In propagation step 2, the H-Br bond is broken and the secondary carbon-to-hydrogen bond is
formed.

167



This results in a step that is exothermic by 7 kcal/mole.

E

= |88] |9S] = 7 kcolmolc

Radicals and polymerization

Radical reactions play a very important role in the synthesis of useful polymers. This aspect of
radical chemistry will be discussed in depth in Part V of the text (Polymers).

General interest radical chemistry: oxidation of fats and oils, vitamin E, and preservatives

Fats and oils are biomolecules of the larger category known as lipids. Lipids are those naturally
occurring, biologically relevant molecules whose water solubility is low. Fats and oils have
quite similar structures. They are formed from the condensation of three fatty acids with
glycerol, a three-carbon triol. A carboxylic acid is termed a fatty acid if its R group consists of a
long hydrocarbon chain.



What distinguishes fats from oils is the structure of the fatty acid R groups. Fats, being solids at
room temperature, have R groups that allow relatively close packing. On the other hand, oils
possess R groups that disfavor this close packing, making them liquids at room temperature. The
presence of unsaturation in the R group will tend to disfavor packing because the naturally
occurring alkenes in fatty acids have the cis configuration. This introduces a bend or kink into
the hydrocarbon chain, which makes the close association of adjacent molecules awkward.

These unsaturated R groups naturally contain allylic hydrogens, and as a consequence, they can
undergo radical substitution reaction. Oxygen is the radical species with which they react.
Oxygen behaves more like a diradical than a molecule with an oxygen-to-oxygen bond. The
two structures are resonance forms, but the diradical explains much of oxygens behavior.


168


In the diagram below, most of the oils structure is omitted. Only a small portion of the
unsaturated R group is shown. An oxygen radical can abstract hydrogen from an allylic center



The coupling of the carbon-centered radical and another molecule of oxygen establishes a new
carbon-to-oxygen bond.



Finally, the abstraction of another allylic hydrogen by the peroxy radical yields a hydroperoxide.



This process has been implicated in the oxidation, and therefore spoilage, of unsaturated fats or
oils upon exposure to air. Saturated fats lack the bonds that make radical substitution viable
and as a result typically have longer shelf lives.

Preservatives provide one solution to the problem of oxidation of foodstuffs. Preservatives like
BHT and BHA are anti-oxidants and therefore share structural similarity and a common mode of
action with well-known anti-oxidants, like vitamin E.



Vitamin E contains a phenolic hydroxyl group. In the presence of a radical (perhaps the allylic
radical seen above in the oxidation of unsaturated fats and oils), this group can serve as a
hydrogen donor to quench the radical, returning it to its original form. The phenoxy radical that
results on vitamin E is much less reactive by virtue of its extensive resonance delocalization and
the steric shielding provided by the flanking methyl groups.

169



Preservatives like BHT and BHA mimic vitamin E. They too contain phenolic hydroxyl groups
that can quench radicals. Furthermore, the resultant phenoxy radicals are similarly resonance
stabilized and sterically shielded by nearby tert-butyl groups.



General interest radical chemistry: CFCs and ozone-layer depletion

The ozone layer protects life on Earths surface from high energy cosmic radiation that is
incident on our atmosphere. Ozone (O
3
) has multiple resonance forms, one of which is shown
below.



There is a natural equilibrium between ozone and its degradation products, atomic and molecular
oxygen. A method of envisioning the electron flow for this process follows.



The natural equilibrium between these processes can be pictured as a cycle. Ozone is degraded
into molecular oxygen (O
2
) and atomic oxygen (O) upon exposure to high energy solar radiation.
When molecular and atomic oxygen recombine to form ozone again, heat is evolved.
Consequently, this cycle essentially converts high energy light into heat.

170



Unfortunately, this natural balance can be disrupted by ozone-depleting substances, such as
chlorofluorocarbons (CFCs). In the 1920s, CFCs were developed as non-toxic and non-
flammable refrigerants by Thomas Midgley, Jr. and Albert Henne of General Motors. These
compounds are termed CFCs or Freons or R, followed by a number that is a code for their
structure. This code typically consists of two to three digits. The first digit is the number of
carbons minus 1. If the value is zero, it is not written. The second digit is the number of
hydrogens plus 1. The final digit is the number of fluorine atoms. The number of chlorine atoms
is then inferred (i.e. the remainder of carbons valences are filled with bonds to chlorine). When
bromine is present in the molecule, the initial set of digits is followed by B and one more
number to denote the bromine count in the molecule.



Midgley, who had a penchant for risky demonstrations, announced the development of the
successful refrigerant R-12 at the 1930 American Chemical Society meeting and illustrated its
safety by inhaling the substance and then blowing out a candle as he exhaled the CFC over the
flame.
19


The stability of CFCs was both a blessing and a curse. They did serve as valuable refrigerants.
However, once released into the environment, their stability enabled them to persist intact until
they reached the upper atmosphere. There, exposure to high energy light causes the homolysis
of a carbon-to-chlorine bond, which generates a chlorine radical.


19
See http://www.epa.gov/ozone/snap/refrigerants/safety.html (accessed February 11, 2013) for
more on Refrigerant safety and history.
171



This chlorine radical can facilitate the degradation of ozone in a propagation step. One way of
envisioning the electron flow during that step is illustrated below. The chlorine essentially
abstracts oxygen from ozone, thereby producing molecular oxygen.



In a second propagation step, the abstracted oxygen atom is donated to another atom of oxygen
to form a second molecule of molecular oxygen. The chlorine radical is also regenerated.



The regeneration of chlorine radical means that this is another chain reaction. Chlorine can
facilitate the degradation of many, many ozone molecules (about 100,000) before it is quenched.
Therefore, the release of a relatively small amount of CFC can lead to a great deal of ozone layer
depletion. Since the discovery of this phenomenon in the 1970s, ozone-depleting substances
have been continually phased out of usage.

Moses Gomberg and the discovery of free radicals

Moses Gomberg, who was a professor at the University of Michigan, is widely considered to be
the father of free radical chemistry. A few years prior to 1900, he prepared tetraphenylmethane,
which had proven to be quite the synthetic challenge.



His next endeavor was the synthesis of the homologous hexaphenylethane.

172

C C
hexaphenylethane


As he encountered unexpected experiment results, Gomberg hypothesized that he had observed a
relatively stable free radical, the triphenylmethyl radical.



This was quite the controversial idea at the time; however, Gombergs ideas about the existence
of radicals were eventually proven correct.
20


Chapter 7 Problems

1. Draw a mechanism for the free radical fluorination of ethane. Label the steps as initiation,
propagation, or termination as appropriate.





20
For a more thorough discussion of Moses Gombergs life and accomplishments, see Moses
Gomberg (1866 1947): A Biographical Memoir by John C. Blair, Jr. (National Academy of
Sciences: Washington, D.C., 1970) which can be accessed online via
http://www.nasonline.org/publications/biographical-memoirs/memoir-pdfs/gomberg-moses.pdf.
173

2. Using the Bond Dissociation Energies provided in the table below, explain why the
fluorination shown in question 1 is not typically used. Focus on the propagation steps, which are
the most significant steps in the process.

Bond Bond Dissociation Energy (kcal/mole)
CH
3
CH
2
H 98
HF 136
FF 38
CH
3
CH
2
F 107

3. Draw a mechanism for the free radical iodination of cyclopentane. Label the steps as
initiation, propagation, or termination as appropriate.



4. Using the Bond Dissociation Energies provided in the table below, explain why the iodination
shown in question 3 is not typically used. Focus on the propagation steps, which are the most
significant steps in the process.

Bond Bond Dissociation Energy (kcal/mole)
(CH
3
)
2
CHH 95
HI 71
II 36
(CH
3
)
2
CHI 53

5. Draw a mechanism for the free radical bromination shown below. Be sure to pay attention to
the selectivity that bromination exhibits.



6. Using the Bond Dissociation Energies provided below, show the change in enthalpy (H) for
propagation step 1, propagation step 2, and the overall reaction. Use these values to briefly
justify how bromination is both highly selective and favorable.

Bond Bond Dissociation Energy (kcal/mole)
Br-Br 46
1 C-H 98
2 C-H 95
3 C-H 91
H-Br 88
1 C-Br 68
2 C-Br 68
3 C-Br 65
174


7. In a classic synthesis of the thiopeptide antibiotic thiostrepton, K. C. Nicolaou and co-workers
employed a radical halogenation of the following epoxide using NBS and heat to initiate the
reaction.
21
The halogenation occurs at the benzylic methylene group. Provide the structure of
the product and a mechanism to explain its formation.



8. Compound A is a trisubstituted alkene that is treated with HBr and peroxides to yield
compound B. This substance is, in turn, treated with tert-butoxide to afford compound C, whose
ozonolysis yields cyclopentanecarboxaldehyde and formaldehyde. Provide the structures of
compounds A, B, and C.








21
For more information about this synthesis, see Chapter 5 of Classics in Total Synthesis III by
K. C. Nicolaou and Jason S. Chen (Wiley-VCH: Weinheim, 2011).
175




Chapter 8:
Advanced radical
chemistry
176

Additional radical initiators

Many introductory textbooks deal only with classical methods of radical initiation, which rely
upon homolysis of a reactants bond or use a peroxide as an added initiator. However, there are
other radical initiators, and youll see examples of these in the primary literature. One option is
AIBN (azobisisobutyronitrile). AIBN decomposes upon exposure to heat or light to yield two
isobutyronitrile radicals and nitrogen gas.



Another radical initiator option is hexabutylditin, which undergoes photolysis to yield two
tributyltin radicals.



Well be using these alternative initiators (particularly AIBN) in quite a few radical reactions.

Radicals containing a bond

When radicals contain a bond, the possibility for intramolecular reaction exists. What actually
transpires depends to a large extent on the concentration (abbreviated [ ] in the following
diagrams) of the hydrogen donor. When the concentration of the hydrogen donor is high, then
simple hydrogen abstraction may be rapid enough to compete with cyclization. When the
concentration of the hydrogen donor is low, then cyclization will almost certainly be faster than
the intermolecular abstraction of hydrogen to quench the radical.



It is key to remember that the rate laws for the inter- and intramolecular processes differ. The
rate of intermolecular hydrogen abstraction will, of course, depend upon both the concentration
of the radical and the hydrogen donor.

Rotc = k |roJicol]|yJrogcn Jonor]

On the other hand, the rate of intramolecular addition of the radical to a bond within the same
molecule will depend only on the concentration of the radical itself.
177


Rotc = k |roJciol]

This latter rate law does not provide much opportunity to exert any influence. However, the rate
law for intermolecular hydrogen abstract does give us an opportunity for control. Boosting the
hydrogen donor concentration enhances the rate, making this the dominant process. Greatly
reducing the hydrogen donor concentration can slow the rate dramatically enough to effectively
shut down that option, allowing the intramolecular cyclization to predominate.

When the cyclization dominates and the radical adds to an intramolecular bond, there are two
ring sizes that can result. If the radical adds to the proximal alkene center, the ring size will be
smaller. If it adds to the distal alkene center, a larger ring size will result. One example is
shown below. In this six-carbon substrate, if the radical at carbon (a) adds to the closer alkene
carbon (e), the product is a five-membered ring with a radical outside the ring.



Conversely, if the radical at center (a) adds to the further alkene carbon (f), the product contains
a six-membered ring, and the radical is on a ring carbon.



Baldwins rules are a set of empirical observations that help us to decide on the viability of a
particular cyclization. The rules are presented in the following table. To use them, we must first
be able to categorize a cyclization. The categorization contains three elements. First, the size of
the ring being formed is noted. Second, the orientation of the breaking bond is stated. If the
breaking bond is outside the incipient ring, the cyclization is exo. If the breaking bond is inside
the ring, the cyclization is endo. Finally, the geometry/hybridization of the carbon to which the
addition is occurring is noted. If the addition is to a tetrahedral (sp
3
) carbon, this component of
the classification is tet. If the carbon undergoing addition is trigonal planar (sp
2
), it is trig. If the
carbon is linear (sp), it is dig (for digonal referring to the 180 bond angle).

Type of ring
closure
Ring size Reactive
center 3 4 5 6 7
exo favored favored favored favored favored tet
endo disfavored disfavored disfavored disfavored disfavored
exo favored favored favored favored favored trig
endo disfavored disfavored disfavored favored favored
exo disfavored disfavored favored favored favored dig
endo favored favored favored favored favored
178


Looking back on our two simple cyclization examples, we can now classify them according to
this scheme. The first example led to the formation of a five-membered ring, so the first part of
the classification is 5. The bond that is breaking is positioned outside the forming ring,
making the cyclization exo, and the carbon undergoing addition is trigonal planar, hence
trig. Therefore, this is a 5-exo-trig cyclization.



In the second of our simple examples, a 6-membered ring is being formed. The breaking bond
is located within the ring (endo), and the carbon to which the addition occurs is trigonal planar
(trig). Thus, this is termed a 6-endo-trig cyclization.



Baldwins rules do not go so far as to state whether a particular cyclization is allowed or
forbidden. Instead, Baldwins rules provide an indication of whether a certain mode of
cyclization is favorable or unfavorable. Both 5-exo-trig and 6-endo-trig cyclizations are favored
(see Table above), so both are quite possible. When both possible cyclizations are favored, it is
often the case that the reaction will occur more rapidly with the closer center to form the smaller
ring. In general, five-membered rings form more rapidly than six-membered rings, which in turn
form more quickly than seven-membered rings. This, however, assumes that the steric
considerations around both reactive centers are comparable. Otherwise, steric encumbrance
could slow down one reaction relative to the other.

Well soon start to consider examples in which a radical is generated and then has the potential
to undergo cyclization. If the concentration of hydrogen donor is held sufficiently low, we can
be sure that cyclization will occur in good yield. The radical formed from the cyclization will
then persist until it finally encounters and is quenched by the hydrogen donor.

Making radicals from functional groups

In the chapter on basic radical chemistry, we saw a limited number of ways to produce a radical.
Radical substitution reactions would entail the homolysis of a carbon-to-hydrogen bond to yield
the most stable radical possible. We also saw that radical addition reactions would yield a
radical from a bond. Nevertheless, the options for generating radicals at various centers have
been limited to this point.

179

We are now going to embark on a discussion of how to convert commonly occurring functional
groups to radicals in a selective fashion. This will give us significantly more synthetic
capability, since well be able to design a substrate bearing a certain functional group knowing
that it can be transformed into a radical for some specific purpose.

Making radicals from halogens

Organohalides can be converted to radicals through exposure to tin radical. Tin radical will
abstract a halogen due to the fact that tin-to-halogen bonds are reasonably strong (Sn-F, 99
kcal/mole; Sn-Cl, 77 kcal/mole; Sn-Br, 65 kcal/mole; Sn-I, 49 kcal/mole). The abstraction is not
substantively different from other -bond homolysis weve seen. Comparing the two processes
side-by-side, we can clearly see the mechanistic similarity. All that differs is the identity of the
specific radical doing the atom abstraction and the identity of the atom that is abstracted.



Lets examine a specific example of such a reaction. 6-Bromo-1-hexene undergoes radical
cyclization in the presence of tributyltin hydride and AIBN when exposed to light or heat.



The reaction begins with two initiation steps. The first is the decomposition of AIBN that weve
seen previously, yielding two isobutyronitrile radicals and nitrogen gas.



In the second initiation step, an isobutyronitrile radical abstracts hydrogen from tributyltin
hydride to afford tributyltin radical. This is the radical that will cycle through the subsequent
propagation steps.

180



The propagation cycle beings with the abstraction of bromine by the tributyltin radical.
Knowing that radicals can cause the homolysis of bonds or add to bonds, you might wonder
why the tributyltin radical does not add to the bond in this instance. Bond strengths provide
the rationale. The Sn-Br bond has a bond dissociation energy of approximately 65 kcal/mole;
whereas, the Sn-C bond has a bond dissociation energy of only about 46 kcal/mole.
Consequently, it is more energetically favorable to form the tin-to-halogen bond, which explains
why halogen abstraction is favored over addition.



The carbon-centered radical thus formed can undergo the 5-exo-trig cyclization discussed earlier.
The low concentration of tributyltin hydride slows the rate of reduction (i.e. hydrogen atom
abstraction), allowing this rapid cyclization to occur preferentially.



Once the cyclization has occurred, the resultant radical persists until it encounters tributyltin
hydride. When it does, it abstracts a hydrogen atom providing methylcyclopentane as the
product. The tributyltin radical is also regenerated and can reenter propagation step 1.



181

If the concentration of tributyltin hydride were high, simple reduction would occur in preference
to cyclization. However, it is perhaps less likely that this would be the desired outcome since it
essentially amounts to the deletion of functionality.



Making radicals from carboxylic acids

Carboxylic acids can also be treated as radical precursors. However, unlike alkyl halides, the
carboxylic acid cannot be directly abstracted as a single unit by a radical. Instead, the acid must
first be converted to an ester that will then interact with tributyltin radical. The reactive ester is
formed by coupling of the carboxylic acid with N-hydroxypyridine-2-thione



The esterification is commonly performed as a coupling reaction using the coupling reagent
dicyclohexylcarbodiimide (DCC). DCC and its mechanism of action will be covered in Chapter
13 (The Synthesis and Protection of Carboxylic Acids and Their Derivatives). For the
moment, we need only know that DCC facilitates the formation of carboxylic acid derivatives
(e.g. esters and amides) through dehydration. Dimethylaminopyridine (DMAP) functions as an
acylation catalyst, which might be employed in this situation. It too will be covered in detail in
Chapter 13. This is an ionic process to prepare an ester that will subsequently be used in a
radical reaction



182

With this particular activated ester in hand, treatment with a low concentration of tributyltin
hydride and AIBN in the presence of light or heat yields methylcyclohexane, carbon dioxide, and
a pyridine derivative.



As we turn our attention to the mechanism, the initiation steps are no surprise since weve seen
them previously.



Once tributyltin radical is formed, it adds to the carbon-to-sulfur bond. The new bond forms
between tin and sulfur. This sets off a cascade in which carbon forms a new bond with the
adjacent nitrogen, the nitrogen-to-oxygen bond homolyzes, and a new carbon-to-oxygen bond
forms, effectively severing carbon dioxide from the substrate.



The carbon-centered radical could, in theory, undergo 6-exo-trig or 7-endo-trig cyclization. Both
options are favored according to Baldwins rules, so yet again proximity dictates the result.
Cyclization of the radical (g) onto the closer carbon of the alkene (b) via 6-exo-trig cyclization
yields a methylcyclohexyl radical.


183


Finally, in the last propagation step, the radical eventually encounters tributyltin hydride and
abstracts a proton from it, reducing the radical to methylcyclohexane and regenerating the
tributyltin radical needed for the first propagation step.



As we saw previously, it would certainly be possible to simply reduce the radical after its
formation by using a high concentration of tributyltin hydride, which would accelerate this
process relative to cyclization. This is typically a less common choice since it amounts to
nothing more than the deletion of functionality. This is occasionally necessary in total synthesis,
but it is not the norm.



Making radicals from alcohols

Alcohols are a third common functional group that can be converted to radicals. Much as with
carboxylic acids, ionic steps are needed to first prepare a substrate amenable to radical reaction.
In the case of alcohols, the hydroxyl group must be converted to a thiocarbonate via a three-step
sequence.



The first step of the sequence entails the use of sodium hydride to deprotonate the alcohol
quantitatively. In the second step, the alkoxide attacks the carbon of carbon sulfide. Finally, the
thiocarbonate ion is alkylated with methyl iodide to complete the sequence.

184



This thiocarbonate is now ready for radical reaction. The conditions weve used previously will
suffice here as well. This example will serve not only as an example of radical reaction of
alcohol derivatives but also as an illustration of tandem cyclizations, which are possible in any
category of radical reaction weve studied.



The reaction begins with the same two initiation steps weve seen previously. As a reminder,
these steps generate tributyltin radical. Another recurring theme is the affinity of tin for sulfur.
In the first propagation step, the tributyltin radical forms a bond to sulfur via the homolysis of the
carbon-to-sulfur bond. This causes a cascade resulting in the formation of a carbon-to-oxygen
bond and the cleavage of a carbon-to-oxygen bond. A carbon-centered radical is produced
from this cascade.



The radical at carbon (a) is five atoms from an alkene carbon (e). Cyclization onto this closer
alkene carbon in a 5-exo-trig fashion yields a cyclopentane derivative with a new radical at
carbon (f).



It just so happens that, in this substrate, the radical at carbon (f) is five atoms from another
alkene carbon (j). Consequently, a second cyclization is possible. This also proceeds in a 5-exo-
trig manner to create a second ring, leaving the final radical at carbon (k).
185




This radical has now exhausted its intramolecular reactivity options. Therefore, it persists until it
is reduced upon encountering a molecule of tributyltin hydride. This final propagation step also
re-forms the tributyltin radical, which can reenter the cycle in propagation step 1.



As weve noted before, a change in the conditions can alter the reactivity. If the concentration of
tributyltin hydride is sufficiently high, simple reduction of the radical is possible. The first
propagation step is identical.



In the second propagation step, the radical immediately encounters tributyltin hydride, given its
presence in high concentration, and hydrogen abstraction results.



When the radical is immediately reduced using a high concentration of hydrogen donor, such as
tributyltin hydride, the reaction is known as Barton deoxygenation. This is a convenient way to
delete functionality if a total synthesis requires such an action. Sometimes a functional group is
used to orchestrate a series of reactions, after which it is no longer needed. That is a situation in
which Barton deoxygenation may be used.

186



General advantages of radical reactions

Radicals often possess no charge. As a consequence, they are relatively non-polar, which
presents some attractive attributes for their use in synthesis. Given their lack of polarity, radicals
are not as broadly reactive with a range of functional groups as some ionic reagents are.
Therefore, it is often possible to conduct radical reactions without the use of protecting groups.
We know that protecting groups are frequently needed to ensure chemoselectivity (i.e. reaction at
only one functional group) in ionic reactions, but they are not desirable since they lengthen
syntheses and thereby reduce overall yields. Radical reactions can help to avoid the over-usage
of protecting groups.

Furthermore, radicals dont typically exhibit the sharp dependence on inductive effects and
solvent that we see with ionic reagents and reactions. This allows radical reactions to be
conducted in a broader array of analogues (whose inductive effects may differ) and in a wider
variety of solvents.

Radicals in synthesis

Weve now seen the potential of radicals to engage in tandem reactions, and weve discussed
some of the general advantages of radicals in synthesis. These topics can be tied together nicely
with an illustrative example from the primary literature. In a 2012 publication in the journal
Organic Letters, Gharpure, Niranjana, and Porwal reported the Stereoselective Synthesis of
Oxa- and Aza-Angular Triquinanes Using Tandem Radical Cyclization to Vinylogous
Carbonates and Carbamates.
22
There are three categories of triquinanes with differing patterns
of five-membered ring fusion: angular, linear, and propellane.




22
Gharpure, S. J.; Niranjana, P.; Porwal, S. K. Organic Letters, 2012, 14(21), 5476 5479.
187

The authors noted the relative paucity of synthetic methods for the preparation of triquinanes
containing both oxygen and nitrogen heteroatom substitution in the rings. They endeavored to
make such a target using tandem radical cyclization.



The overall reaction begins with an aryl iodide radical precursor. There are also two alkenes in
this substrate, one of which is part of a vinylogous urea functionality. Exposure to tributyltin
hydride with AIBN as the radical initiator leads to the formation of two new five-membered
rings.



The initiation sequence is no different than what weve encountered before. The tributyltin
radical formed during initiation then abstracts iodine in the first propagation step. This produces
an aryl radical.

O
I
N
N
O
O
Bz
Bu
3
Sn
O
N
N
O
O
Bz
+ Bu
3
Sn I
Propagation step 1


188

The aryl radical is five atoms away from the proximal alkene carbon. It adds to the alkene
bond in 5-exo-trig fashion, resulting in the formation of the first new five-membered ring, as
well as a new carbon-centered radical. As you are defining the type of cyclization as 5-exo-trig,
remember that, although the bond undergoing addition is within a ring, it is outside the ring
that is forming. Consequently, it is an exo process.



This newly formed radical also happens to be five atoms away from the closer carbon of yet
another bond. Addition to that bond (also in a 5-exo-trig manner), yields the second new
five-membered ring, as well as a radical to the carbonyl.



With no further intramolecular reaction possible, this radical is eventually quenched by
tributyltin hydride to form the azaoxatriquinane and tributyltin radical, which can continue the
propagation.



Notice that cis-fusion of the five-membered rings is favored. This is the lower energy
configuration and is generally favored over trans-fused five-membered rings.

189

The reaction proceeds in 55% yield. Although that is not an exceptionally high yield, this
reaction nevertheless has tremendous synthetic value since it leads to the formation of two new
rings and four stereocenters (with high diastereoselectivity) in a single transformation!

Chapter 8 Problems

1. In an article published in the Journal of Organic Chemistry, Orito and co-workers described
the synthesis of the isoindolobenzazepine lennoxamine.
23
The synthesis utilized radical
conditions in the final step to generate the last ring of the core structure. That transformation is
shown below. Provide the product of the reaction and a mechanism to explain its formation,
including the relevant initiation and propagation steps.



2. Investigators at the Australian National University published A Chemoenzymatic Total
Synthesis of (+)-Clividine.
24
The natural product clividine was isolated from the bush lilly
Clivia miniata Regel. The extracts of this plant have been used in traditional Zulu medicine.
The following reaction was employed in the synthesis. Show the intermediates A, B, and C, as
well as a mechanism for the reaction of C with tributyltin hydride and AIBN to yield D.



3. Yasuda, Ryu, and co-workers investigated the use of tetrahydropyran (THP) as a solvent for
radical reactions.
25
In their communication, one of the reactions studied was the following.
Provide a mechanism and product for this transformation. Assume that a relatively low
concentration of tributyltin hydride is used.




23
Onozaki, Y.; Kurono, N.; Senboku, H.; Tokuda, M.; Orito, K. Journal of Organic Chemistry,
2009, 74(15), 5486-5495.
24
White, L. V.; Schwartz, B. D.; Banwell, M. G.; Willis, A. C. Journal of Organic Chemistry,
2011, 76(15), 6250-6257.
25
Yasuda, H.; Uenoyama, Y.; Nobuta, O.; Kobayashi, S.; Ryu, I. Tetrahedron Letters, 2008,
49(2), 367-370.
190

4. Imagine that the diene below were subjected to similar conditions. What product would you
expect? Show a mechanism to explain its formation.



5. In the same paper references in question 3, radical additions using sulfur were explored.
Under the conditions shown below, a hydrogen is abstracted from hexanethiol (HexSH). Using
these clues, predict the product of the following reaction. Assume that a relatively low
concentration of hexanethiol is used.



6. Investigators at the Instituto de Qumica Orgnica General in Madrid reported on 6-endo
versus 5-exo radical cyclization: streamlined syntheses of carbahexopyranoses and derivatives by
6-endo-trig radical cyclization.
26
Show the structure of compound A. Given the description of
the reaction provided in their title, show a mechanism for the conversion of A to B. Assume a
relatively low concentration of tributyltin hydride.



7. In this same paper, the following reaction was reported. In this process, tributyltin radical
adds to the alkyne. Using this clue predict the product and show a mechanism for its formation.
Assume a relatively low concentration of tributyltin hydride.




26
Gmez, A. M.; Company, M. D.; Uriel, C.; Valverde, S.; Lpez, J. C. Tetrahedron Letters,
2007, 48(9), 1645-1649.
191

8. The Synthesis of the Core of Actinophyllic Acid Using a Transannular Acyl Radical
Cyclization was reported by investigators at the Institute of Medical, Pharmaceutical and Health
Sciences in Japan.
27
A key complexity-building step in this process is the radical cyclization
shown below. The initial substrate bears protecting groups known as Boc and Cbz. Well learn
about those in Chapter 11. For the moment, they can merely be treated as inert. The first
transformation involves hydrolysis of the ethyl ester and its conversion to a selenoester. The
selenoester is then transformed into an acyl radical using tributyltin hydride and ACN (a radical
initiator very similar to AIBN). Show how this acyl radical cyclizes to yield the product shown.







27
Zaimoku, H.; Taniguchi, T.; Ishibashi, H. Organic Letters, 2012, 14(6), 1656-1658.
192




Part IV:
Ionic Reactions

193




Chapter 9:
Synthesis of alkenes
194

Birch reduction

The Birch reduction is a nice segue from our discussions of aromatic chemistry and radicals to
this segment focusing on other functionalities and ionic mechanisms. The reason is that the
Birch reduction is a reaction of an aromatic ring, and it involves radical anions. Therefore, this
important named reaction blends concepts from the previous parts of this text.

The net result of the Birch reduction is the conversion of an aromatic ring to a 1,4-
cyclohexadiene. The reaction requires lithium or sodium metal, which is dissolved in liquid
ammonia. An alcohol is also included as a proton source.



The process begins with the donation of the metals lone valence electron to the aromatic ring.
The intermediate thus formed is a radical anion (i.e. it bears both an unpaired electron and a
negative charge).



An alternate way to envision the first step of the mechanism is to show the donation of the
metals electron to a particular carbon of the ring. That carbon becomes the anion as its bond
homolyzes, and the adjacent center develops radical character. In making this donation, the
metal becomes a cation and attains the preceding noble gas configuration.

NH
3(l)
ROH
Li
+ Li


This radical anion is part of a conjugated system and therefore benefits from resonance
delocalization.



The most important resonance contributor for what follows is the one that places the radical and
the anion on opposite ends of the ring. Once weve completed our discussion of the mechanism,
195

well return to this issue and examine why it is the case. The alcohol serves as a proton source,
and protonation of the anion yields a radical intermediate.



This radical accepts an electron from a second molecule of neutral metal. In the process, an
anion is formed.



That anion is again protonated by alcohol, yielding the 1,4-cyclohexadiene product.



At this point, it is logical to ask a question about the regiochemistry of the reaction. Given the
emphasis placed on the stabilizing effect of conjugation, you might wonder why the Birch
reduction results in a 1,4-cyclohexadiene rather than a 1,3-cyclohexadiene. After all, the two
bonds of the 1,4-cyclohexadiene are isolated (i.e. no conjugation), while the two bonds of 1,3-
cyclohexadiene are conjugated, which would make it the more stable isomer and therefore
seemingly more desirable as a product.



196

The answer to this question lies in the principle of least nuclear motion, which is often shortened
to the principle of least motion or PLNM.
28
This principle essentially invokes the idea that a
transformation requiring a significant rearrangement of nuclei will be less favorable than one
which requires less nuclear motion. Another way of conceptualizing this is that a reaction
involving a dramatic change in atomic position will have a higher energy transition state (hence a
higher activation energy barrier) and will therefore proceed more slowly than a reaction
requiring little change in position of the atoms involved.

Applying this to the current situation requires that we examine the resonance structures of each
intermediate in the reaction. The radical anion has three resonance forms, which weve seen
previously.



Lets assign a bond order (1 for a single bond and 2 for a double bond) to each site involved in
the resonance delocalization.



If we now average those bond orders, it becomes apparent that two of the bonds involved in
resonance have more double bond character than the others. This does assume though that each
resonance form is an equal contributor to the resonance hybrid.



Heres where the principle of least motion comes into play. If the protonation yields the alkenes
in a 1,4-arrangement, there is little nuclear motion. The bonds of the hybrid that were closer to
double bonds (bond order = 1
2
3
) become the double bonds of the radical. The bonds of the
hybrid that were closer to single bonds (bond order = 1
1
3
) become the single bonds of the radical.



28
For a full discussion of this phenomenon, see Hine, J. Advances in Physical Organic
Chemistry, 1978, 15, 1 61.
197


On the other hand, lets consider protonation to give the 1,3-arrangement of the alkenes. In this
case, greater nuclear motion is needed to transition from the radical anion to the radical. One of
the bonds with a bond order of 1
1
3
becomes a double bond, and one of the bonds with a bond
order of 1
2
3
becomes a single bond. This process is therefore less likely and is not generally
observed.



A similar analysis of the final anions protonation also justifies the formation of the 1,4-
cyclohexadiene. The anion has the following three resonance forms.



As before, a bond order can be assigned to each site involved in the resonance delocalization of
the charge.



Well assume that each resonance form is an equal contributor to the hybrid, which allows us to
simply average the bond orders as shown below.



The protonation that explains the formation of product involves minimal atomic motion. Each
bond with a bond order of 1
1
3
becomes a single bond (bond order = 1), and each bond with a
bond order of 1
2
3
becomes a double bond (bond order = 2).

198



On the other hand, protonation to afford the 1,3-cyclohexadiene requires a more significant
shifting of atoms since one bond with a bond order of 1
1
3
becomes a double bond and one bond
with a bond order of 1
2
3
becomes a single bond. This is undesirable, and consequently this
product is not formed.



Hofmann elimination

The Hofmann elimination is an alkene synthesis that utilizes an amine reactant. The reaction
involves three steps: (1) treatment with excess methyl iodide; (2) treatment with silver oxide; and
(3) heating. The final product is the less substituted, or Hofmann, alkene.



In step 1, the amine is exhaustively alkylated. Attack of nitrogens lone pair on the electrophilic
and unhindered carbon of methyl iodide (in S
N
2 fashion) results in an ammonium ion that
subsequently loses a proton.



This process repeats until the nitrogen can be alkylated no further. Note that the reactant does
not have to be a primary amine. However, regardless of the substitution of the amine reactant,
exhaustive alkylation is performed.



Treatment with silver oxide in the second step of the process results in a counterion switch.
Silver has a high affinity for halides, so silver iodide forms and precipitates. The hydroxide
199

formed from silver oxide in aqueous media then takes the place of iodide as the cations
counterion.



In the final step of the sequence, heating results in elimination of a proton and trimethylamine.
This occurs in a concerted E2 manner, which will become exceedingly important as we examine
the surprising regiochemistry of this reaction.



You might recall from Introductory Organic Chemistry that typically a small enough base will
remove a proton from the more substituted -carbon so as to form the more highly substituted,
and more stable, Zaitsev product. This, however, is not observed in this instance.



The rationale comes from an examination of transition state energies for the two possible modes
of elimination. Lets first analyze the elimination of a proton from the more highly substituted -
carbon. If we look down the -carbon to -carbon axis (the purple bond in the diagram below),
we can draw a Newman projection for the Zaitsev elimination. Remember that, in E2
elimination, the groups being removed must be anti-periplanar to one another. Notice that, when
the proton and the trimethylamino group are arranged anti-periplanar, the Newman projection
reveals an unavoidable gauche interaction between the trimethylamino group and the methyl
group on the -carbon.



On the other hand, a Newman view of the removal of a proton from the less substituted -carbon
shows no gauche interactions during the Hofmann elimination.

200



Plotting these observations on an energy diagram reveals an unusual situation. The more highly
substituted alkene product (Zaitsev) is known to be the more stable product. However, the
transition state leading to its formation is higher in energy because of the gauche interaction.
There is a higher activation energy barrier to the formation of the more stable product, which
slows the process. On the other hand, the Hofmann product, while less stable, can be reached
through a lower energy transition state. This makes formation of the Hofmann product the faster
process. Consequently, we would say that this reaction is under kinetic control, rather than
thermodynamic control. The more thermodynamically favorable product is not obtained.
Instead, the less thermodynamically favorable product is produced through a lower energy
pathway.

Reaction Coordinate
NMe
3
Me H
H
Me H
gauche interaction
NMe
3
Et H
H
H H
no gauche
interactions
Hofmann product
Zaitsev product
N
Me Me
Me
OH


Cope elimination

The Cope elimination is a second amine-based alkene synthesis, but it proceeds via a mechanism
very different from that of the Hofmann elimination. The amino group is initially oxidized to a
201

tertiary amine oxide using either hydrogen peroxide or a peroxy acid. Subsequent heating leads
to the elimination of a hydroxylamine from the molecule, yielding the alkene product.



The mechanism for the elimination involves a cyclic transition state. The oxygen anion removes
a proton from the position. As the -carbon to hydrogen bond is cleaved, the ,- bond forms
and the -carbon to nitrogen bond fragments.



If both the -carbon and -carbon are secondary or tertiary, then the geometry of the alkene
product must be considered. For instance, in the example below, the -carbon bears an
additional methyl group. This will ultimately impact the choice of the -proton that is
eliminated. Since there are two -protons in this substrate, either one could conceivably be
removed by the oxyanion. The removal of one of the -protons will place the -methyl group on
the same side of the cyclic transition state. This is a sterically encumbered transition state. If
that same transition state is viewed from the Newman perspective, the unfavorable gauche
interaction between the methyl groups is readily apparent. Consequently, this elimination
pathway is disfavored, and the cis-alkene is not obtained.



Removal of the other -proton, places the - and -methyl groups on opposite sides of the cyclic
transition state. The Newman diagram of this transition state shows that there are no gauche
interactions. Therefore, this is the preferred mode of elimination, and the trans-alkene is the
reaction product.
202




There is a piece of stereochemical nomenclature that could be applied to these two transition
states. When the backbone of a molecule is drawn in the eclipsed fashion and two similar groups
are on the same side of the chain, the form is called erythro. When drawn in the same fashion
and those groups are on opposite sides of the chain, the form is termed threo. In the transition
states we discussed, one form places the methyl groups on the same side (erythro), while the
other places them on opposite sides (threo).



Selenoxide elimination

The selenoxide elimination is an analogous alkene synthesis involving a cyclic transition state;
however, it utilizes a completely different reactant. The overall process installs an unsaturation
, to a carbonyl-containing functional group. The example below utilizes a ketone, but other
functionalities, such as esters, can be used as well.

203



There is more than one way to install the selenoxide group, which causes the actual elimination.
One representative method is to first prepare the enolate quantitatively using a strong base, such
as lithium diisopropylamide (LDA). Note that this can be a misleading piece of nomenclature.
The term amide refers to two things: (1) the carbonyl-containing functional group bearing a
nitrogen and (2) the conjugate base of ammonia or an amine. It is the latter meaning that applies
in this situation.



In the second step of the reaction, some source of electrophilic selenium, such as diphenyl
diselenide, is added. There are a number of other reagents that can serve in its place. The
enolate attacks the electrophilic selenium, installing a new carbon-selenium bond at the -
position.



Oxidation to the selenoxide is performed by exposure to hydrogen peroxide or a comparable
oxidant.



This selenoxide then undergoes elimination via a 5-membered cyclic transition state that is very
similar to that observed in the Cope elimination. The oxyanion removes a -proton, and the
carbon-hydrogen -bonding electrons collapse toward the -carbon, forming the bond and
expelling the selenium leaving group.

204



Peterson elimination

The Peterson elimination is a reaction that forms both bonds of an alkene. This is a significant
departure from the reactions weve studied thus far, in which only the bond of the alkene is
formed. The process begins with the addition of a silyl carbanion to a ketone or aldehyde.
Typically, this process yields a mixture of diastereomeric products, and the isomers are then
separated. For our purposes, well take a look at one mode of addition yielding just one
diastereomer, recognizing that the reaction would likely not be totally selective. A transition
state is shown (looking down the axis of the incipient carbon-carbon bond), which has relatively
low steric strain. The resultant oxyanion is protonated in a second step upon exposure to water.



The -silyl alcohol can be eliminated in one of two ways. The choice of conditions impacts the
alkene geometry that results. In basic conditions (e.g. sodium hydride), one alkene
stereochemistry is formed, while the other alkene stereoisomer is formed in acidic conditions.



In a strong base, like hydride, the alcohol is completely deprotonated, forming an alkoxide and
hydrogen gas.



Rotation about the central carbon-carbon bond enables the oxyanion to attack silicon. There
are multiple interpretations of what happens next. One hypothesis is that a 4-membered ring is
formed, which subsequently eliminates silicon and oxygen to yield the alkene. A second
hypothesis is that, as the silicon is attacked by oxygen, an intermediate carbanion forms. In any
205

event, the silicon and oxygen are both eliminated before any additional rotation about the
carbon-carbon bond can occur. The diagram below shows the oxygen attacking silicon, the
carbon-silicon bond fragmenting to form the bond, and the carbon-oxygen bond cleaving as a
consequence. The product, a (Z)-alkene, preserves the erythro arrangement of the methyl and
ethyl groups seen in the intermediate alkoxide.



In acidic conditions, the alcohol is protonated, and the good leaving group thus formed is lost
simultaneously with the trimethylsilyl group. This is an E2-style reaction, so the trimethylsilyl
group and the oxonium ion must be anti-periplanar to one another during the elimination. As a
result, the cis arrangement of the methyl groups in the oxonium ion intermediate is preserved in
the product, which has the (E)-configuration.



Wittig reaction

The Wittig reaction has an important attribute in common with the Peterson elimination: it forms
both the and bonds of an alkene. The reaction begins with an alkyl halide, which is
converted to an intermediate known as an ylide in two steps. The ylide then reacts with an
aldehyde or ketone. The original Wittig reaction happens to be selective for the formation of cis
alkenes. There are a number of modifications of the Wittig reaction that have differing
selectivity. However, those modifications will not be discussed here.



The reaction sequence begins with treatment of an alkyl halide with triphenylphosphine. The
nucleophilic triphenylphosphine attacks the -carbon, displacing the halide in S
N
2 fashion to
yield a phosphonium salt.


206


In the second step, the phosphonium salt is treated with a strong base, such as butyllithium
(BuLi). Deprotonation affords a doubly charged species known as an ylide. More specifically,
we could say that this is a phosphorus ylide. This ylide has a neutral resonance form.



In the final step of the sequence, the aldehyde or ketone is added. A good deal happens in this
step, so lets examine each portion carefully. To understand the connectivity of the initial
intermediate formed, it is easiest to use the doubly charged ylide resonance form. This clearly
illustrates why the carbon with anionic character attacks the carbonyl carbon to yield the
intermediate known as a betaine.



Lets pause at this point for a note on nomenclature. Weve seen some unusual terms, ylide and
betaine. An ylide contains an adjacent cation and anion. A betaine also contains a cationic
group (with no hydrogen atoms) and an anionic group, but these groups need not be adjacent. A
betaine is a specific type of zwitterion, a term you may have encountered when studying amino
acids. Zwitterions contain both a cation and an anion at different sites within the molecule, but
there is no stipulation about whether or not the cation can bear hydrogen atoms, as there is with a
betaine.

To understand the betaines stereochemistry, we need to look a little deeper into the mechanism.
Here, it is convenient to use the ylide resonance form containing a bond. This helps to
illustrate that, in the transition state, the systems of the ylide and carbonyl should align so that
the incipient carbon-carbon bond (green in the Newman projection below) is formed from the
direct overlap of a p orbital from each reactant. That being said, there are two possible transition
states that abide by this alignment of p orbitals. In one, there are no gauche interactions, but the
other does suffer from this steric encumbrance. Consequently, the transition state that minimizes
steric repulsion leads to the observed cis-betaine.

207



This cis-betaine undergoes ring closure to yield a 4-membered ring known as an
oxaphosphetane.



Finally, the oxaphosphetane extrudes triphenylphosphine oxide to yield the cis-alkene. Since
both bonds are broken simultaneously during this step, the cis-arrangement of the methyl and
ethyl groups in the oxaphosphetane is maintained in the alkene product.



Chapter 9 Problems

1. The regiochemistry of the Birch reduction is sensitive to substituent effects. Electron-
withdrawing groups on the ring yield one outcome, while the presence of electron-donating
groups yields a different outcome. Use mechanism to explain the results shown below.


208




2. Boll and co-workers reported the Reversible Biological Birch Reduction at an Extremely
Low Redox Potential in which they alluded to the following traditional Birch reduction and then
compared it with transformations that are possible using enzymatic catalysis.
29
Predict the
product of the reaction shown below.



3. Investigators at the Tokyo Institute of Technology reported A fluorous-assisted synthesis of
oligosaccharides using a phenyl ether as a safety-catch linker.
30
In this article, the following
Birch reduction was described. Predict the product of this reaction.



4. Guo and Schultz reported the following transformation in the Journal of Organic Chemistry.
31

This involves a slight modification of the traditional Birch reduction that is sometimes called
Birch alkylation. In Birch alkylation, the final anion produced during the mechanism can attack

29
Kung, J. W.; Baumann, S.; von Bergen, M.; Mller, M.; Hagedoorn, P.-L.; Hagen, W. R.;
Boll, M. Journal of the American Chemical Society, 2010, 132(28), 9850-9856.
30
Tanaka, H.; Tanimoto, Y.; Kawai, T.; Takahashi, T. Tetrahedron, 2011, 67(51), 10011-10016.
31
Guo, Z.; Schultz, A. G. Journal of Organic Chemistry, 2001, 66(6), 2154-2157.
209

an electrophile to install an alkyl group on the ring. Given this information, predict the product
of the reaction shown below.



5. Kini and Ramana of the University of Mumbai reported A convenient synthesis of
phenanthrene alkaloids from 1-arylmethyl-1,2,3,4-tetrahydroisoquinolines.
32
In their paper, the
following reactions are reported. Provide structures for compounds A and B.



6. In their paper entitled Oxidation-Cope elimination: a REM-resin cleavage protocol for the
solid-phase synthesis of hydroxylamines, Sammelson and Kurth of UC Davis describe the
following sequence. Compound A is formed by conjugate addition of piperidine to the polymer-
bound acrylate reactant (the polymer is represented by a sphere). Then compound A is treated
with meta-chloroperoxybenzoic acid to yield an intermediate, compound B, that fragments to
yield the products, compounds C and D. Product structures for each intermediate, as well as a
mechanism to explain the formation of compounds C and D.



7. Investigators from the University of Melbourne reported the following conversion.
33
Show
reagents that could be used to achieve this transformation.




32
Kini, S. V.; Ramana, M. M. V. Tetrahedron Letters, 2004, 45(21), 4171-4173.
33
Macdougall, P. E.; Smith, N. A.; Schiesser, C. H. Tetrahedron, 2008, 64(12), 2824-2831.
210

8. In their synthesis of lancifolol (shown below), Galano, Audran, and Monti of the Laboratoire
de Ractivit Organique Slective at the Facult des Sciences et Techniques de St. Jrme
utilized at Peterson elimination.
34




The Peterson elimination is shown below. Provide the structure of the product and a mechanism
to explain its formation. Note that CHx is an abbreviation for cyclohexyl.



9. Investigators from Zhejiang University and the Hong Kong University of Science and
Technology reported the following sequence involving tandem Wittig reaction and
intramolecular Diels-Alder cycloaddition under microwave (MW) heating. Provide structures
for compounds A through C.







34
Galano, J.-M.; Audran, G.; Monti, H. Tetrahedron Letters, 2001, 42(35), 6125-6128.
211




Chapter 10:
Synthesis and
protection of alcohols
212

Synthesis of alcohols

In Introductory Organic Chemistry, you learned a number of ways to prepare alcohols. Some of
these are briefly summarized below. Alkenes are a common source of alcohols via hydration.
There are three general methods of hydration. Simple, acid-catalyzed hydration yields
Markovnikov regioselectivity; however, in some instances carbocation rearrangements may
occur. To obtain Markovnikov regioselectivity without the danger of carbocation rearrangement,
oxymercuration-demercuration is typically employed. Finally, H. C. Browns hydroboration-
oxidation protocol affords anti-Markovnikov regioselectivity.



The synthesis of alcohols via reduction is an alternative to hydration. Ketones or aldehydes may
be reduced with sodium borohydride (NaBH
4
) to the corresponding secondary or primary
alcohols, respectively. Grignard reagents (or organolithium reagents) can also be used to install
new carbon-carbon bonds on the incipient alcohol.



Esters can be similarly reduced to primary alcohols, but they require the use of the more potent
reducing agent lithium aluminum hydride (LiAlH
4
). Reaction with excess Grignard or
organolithium reagent converts the ester to a tertiary alcohol bearing two identical alkyl groups.

213



Protection of alcohols: rationale

Alcohols, like several other functional groups well discuss, present some unique challenges in
multi-step synthesis. They can undergo unintended side reactions as you attempt to manipulate
other functional groups. For instance, alcohols are nucleophilic, so they may attack an
electrophile intended for interaction at another site in the molecule. Additionally, alcohols may
be deprotonated under basic conditions, making them even more nucleophilic.



In acidic conditions, protonation of the alcohol transforms the hydroxyl group into a good
leaving group. This may result in unintended substitution or elimination.



Consequently, it is often necessary to temporarily render the alcohol inert through a process
known as protection. To protect a functional group is to derivatize it such that it is unreactive
under the conditions planned for a synthesis. The derivatization must also be readily reversible
so that the alcohol can be later unveiled when it is needed. Generically, protecting groups will
sometimes be denoted as PG.



It is easiest to see the need for protecting groups through examples. Lets consider the
Dieckmann condensation shown below. You may remember the Claisen condensation from your
Introductory Organic course. The Claisen condensation is the reaction of two esters with the loss
214

of alcohol to form a -keto ester. When this reaction is performed intramolecularly to form a
ring, it is called the Dieckmann condensation.



The mechanism of the Dieckmann condensation is shown in the following diagram. An alkoxide
base is used to deprotonate the position to the ester. It is important to note that the alkoxide R
group must match the ester R group in order to prevent transesterification. The enolate thus
generated attacks the carbonyl carbon of the other ester, leading to a cyclic tetrahedral
intermediate. Collapse of the tetrahedral intermediate expels an alkoxide and yields a -
ketoester. To this point, each step of the reaction has been freely reversible. However, once the
-ketoester is formed, its deprotonation (at the especially acidic position to two carbonyls)
drives the equilibrium. The highly resonance-stabilized enolate thus formed persists until it is
quenched by an acidic workup.



This mechanism was presented as you would have drawn it in Introductory Organic Chemistry.
There is, however, a convenient shortcut that you can begin to use now that you are more
comfortable with Organic mechanism. Drawing tetrahedral intermediates that subsequently
collapse can become tedious as mechanisms get more lengthy. Therefore, it is common to
condense the mechanism by bypassing the tetrahedral intermediate. It still forms, but we wont
explicitly draw it. Instead, well draw the mechanistic arrows to imply its transient existence.
The diagram below shows this shortcut. After deprotonation of the -carbon, the enolate attacks
the other carbonyl carbon. Rather than showing the -bonding pair of electrons coming to rest
on the carbonyl oxygen, well draw an arrow that loops around the carbonyl oxygen. This
reminds us that the electrons come to rest on the carbonyl oxygen and then collapse to re-form
the bond, but it is shorter since we wont have to take the time to draw the tetrahedral
intermediate. As the green arrow shows the re-formation of the bond, methoxide must then be
215

expelled. This leads us directly to the -ketoester, which is subsequently deprotonated and then
reprotonated upon workup.



Now that weve reviewed the basic Dieckmann condensation, lets imagine that you want to
perform a Dieckmann condensation on a substrate containing an alcohol. The methoxide will
likely deprotonate the hydroxyl group, which has a pK
a
( 15) lower than that of the carbon to
the ester ( 25). The alkoxide thus formed could attack the carbonyl carbon that is six atoms
away. Here again, we can use our shorthand notation showing that the electrons move onto the
carbonyl oxygen and then collapse to re-form the bond. Methoxide is lost, and a lactone (i.e. a
cyclic ester) is the product. The reaction did not proceed as expected to give a -ketoester due to
the presence of the alcohol functionality.



Protection of alcohols: silyl ethers

There are many, many ways to protect each functional group, and entire books have been
devoted to the possibilities.
35
Well take a look at just a couple of examples. Silyl ethers are
commonly used to protect alcohols. A silicon bearing three R groups is attached to the alcohol

35
For two excellent examples, see Protecting Groups by P. J. Kocieski (Thieme: New York,
1994) and Greenes Protective Groups in Organic Synthesis, Fourth Edition by P. G. M. Wuts
and T. W. Greene (Wiley: New Jersey, 2007). The Kocieski book is more of a textbook and is
a good general introduction to the wide array of protecting groups. Greenes book is much more
encyclopedic and is useful if you want to become a practitioner in the field.
216

oxygen, taking the place of its proton and rendering the oxygen relatively unreactive. There are
a wide range of silyl ethers that can be formed, and the R groups dictate their relative stabilities
under different sorts of conditions. The tert-butyldimethylsilyl ether is a common choice. This
group is abbreviated as TBDMS or simply TBS, and it is reasonably stable under a wide variety
of conditions between pH 4 and 12. There are also many particular sets of reagents that can be
used to install the TBS group. One such set is tert-butyldimethylsilyl chloride (TBSCl) and
imidazole.



The imidazole serves as both a base and a catalyst for the etherification. As a base, it can
hydrogen bond to the hydroxyl proton and remove it as the oxygen attacks silicon.



As a catalyst, it can attack the silicon directly, forming a more active silyl-transfer reagent. The
silicon now bears an even better leaving group than chloride.



The imidazole is displaced as the oxygen attacks silicon, and then a proton is lost from the
oxonium ion to this base.

HN N
Si
Cl
MeO
O
OMe
O O
H
MeO
O
OMe
O O
H Si
HN N +
MeO
O
OMe
O O TBS
HN N +
H
Cl


217

With the protected alcohol in hand, we can now perform the Dieckmann condensation
successfully. Since the alcohol has been rendered inert through protection, it does not participate
in the reaction at all. The end product is a cyclic -ketoester as expected.



The protecting group has served its function, and it can now be removed. TBS groups can be
removed in acid, but that might not be a good choice here since ester hydrolysis could also result.
Fluoride has a high affinity for silicon, so fluoride sources (like HF) are quite selective ways to
remove silyl ethers.



The oxygen can be protonated to make it a better leaving group, and the fluoride can then
displace the silicon to unveil the free alcohol.



Protection of alcohols: alkyl ethers

Lets consider another synthesis to showcase the utility of a different type of protecting group:
the alkyl ether. The Robinson annulation is especially useful for building up fused-ring systems,
such as the one shown below. The process consists of a Michael reaction followed by an aldol
condensation.

218



In the Michael reaction, an enolate is generated through removal of the most acidic proton in the
system. This enolate adds to the -carbon of an ,-unsaturated carbonyl, generating a new
enolate.



This enolate can be protonated in aqueous media, and another enolate can be formed at a
different center since the medium is basic. One such enolate will place a nucleophilic -carbon
six atoms from an electrophilic carbonyl.



The resultant attack yields a new six-membered ring bearing an alkoxide, which is protonated by
the medium. The -hydroxyketone intermediate could revert to starting material through a
reversal of each step examined thus far. Consequently, it is common to employ mild heating,
which facilitates irreversible dehydration. When heated in basic media, the -hydroxyketone is
deprotonated at its -position. The carbon-hydrogen electrons collapse in between and to
form a bond, and hydroxide is ejected from the -position. While hydroxide is not generally a
good leaving group, in this case one hydroxide molecule was merely traded for another, but in
the process three molecules were generated from two, thereby increasing the entropy of the
system. The ,-unsaturated ketone is the final product of the Robinson annulation.

219



Now, lets imagine that we need to perform an analogous reaction on a system bearing a
phenolic hydroxyl group.



The phenolic proton is the most acidic in the system. Its pK
a
( 10) is significantly lower than
that of a position to a ketone ( 20). So, we can expect the phenol to be deprotonated. The
phenoxide thus produced may well be resistant to a second deprotonation at the -carbon.
Therefore, we should consider protecting the phenol.



Alkyl ethers would not generally be employed to protect alcohols because the resultant ethers
would be difficult to cleave. However, since phenoxide is a much better leaving group than an
alkoxide, alkyl ethers can be used for the protection of phenols.



The ether can be generated by quantitatively deprotonating the phenol with a strong base, like
sodium hydride. The nucleophilic phenoxide can then attack an electrophilic methyl group, such
as that of methyl iodide.
220



With the substrate suitably protected as its methyl ether, we can now proceed with Robinson
annulation. The annulation will proceed exactly as expected via the mechanism outlined
previously. The methyl ether is stable even at very high pH, so it is totally inert under these
conditions.



Once the annulation has been achieved, the protecting group has served its purpose and can be
removed. There are a number of ways to do this. One method employs boron trichloride
followed by treatment with base.



The boron trichloride is a Lewis acid, and undergoes Lewis acid-base reaction with the aryl
methyl ether. The adduct formed can lose a chloride, and this chloride can attack the labile
methyl group of the oxonium ion, severing the oxygen-to-methyl group bond.



The intermediate boron-oxygen adduct is cleaved in aqueous base, as borate salts are formed.

221

O
O
B
Cl
Cl
Na
2
CO
3
H
2
O
O
O H


Protection of alcohols: alkoxyalkyl ethers

Another common way to protect alcohols is as their alkoxyalkyl ethers. The functional group
contained in the protected alcohol is an acetal, so the behavior is as you would expect for an
acetal: stable in base but less so in acid. One specific alkoxyalkyl ether is the methoxymethyl (or
MOM) ether. The MOM ether is installed using methoxymethyl chloride (MOMCl) and a non-
nucleophilic base, such as diisopropylethylamine (DIEA). This amounts to a straightforward
substitution reaction, and the DIEA serves to sequester the acid liberated during the
transformation.



As you would expect, this acetal can be cleaved in aqueous acid much like other acetals.



Lets examine a scenario in which this protecting group could be useful. Imagine that you want
to prepare the following diol from a hydroxyketone precursor. A Grignard or organolithium
reagent would clearly be the best way to install the new methyl group at the carbonyl carbon.



However, direct treatment with methylmagnesium bromide is not feasible because the basic
reagent will deprotonate the alcohol. This will destroy the Grignard reagent and fail to install the
new carbon-carbon bond as desired.

222



Protection of the hydroxyl group as the MOM ether prior to the Grignard reaction solves the
problem. The acetal is unreactive under the basic conditions of a Grignard reaction, so the
tertiary alcohol can be synthesized. Then, the MOM group can be removed through hydrolysis.
It may even be possible to condense the synthesis by combining the Grignard workup step with
the acidic hydrolysis, thereby protonating the alkoxide product of the Grignard reaction and
cleaving the MOM ether in one step.



Protection of alcohols: advantages and disadvantages

At first glance, protecting groups are extremely appealing because they allow us to ensure very
chemoselective reaction (i.e. reaction at only a single functional group). However, it is important
to remember that each protecting group used increases the length of a synthesis by two steps: one
for the protection and one for the deprotection. Although protecting groups can typically be
installed and cleaved in very high yields, any additional transformations reduce the overall
efficiency of a synthesis. Consequently, reactions that exhibit high chemoselectivity in the
absence of protecting groups are highly coveted. Such reactions allow for shorter, more direct
synthetic routes.


223

Chapter 10 Problems

1. In their synthesis of 5-epi-peacilomycin F, Jana and Nanda of the Indian Institute of
Technology formed an ester from the carboxylic acid and alcohol shown below.
36
The
esterification they used is known as the Mitsunobu reaction. Notice that it results in an inversion
of configuration at the alcohol center. Well learn more about the Mitsunobu reaction in Chapter
13.



Show how the alcohol used in the esterification could be prepared from the following substrate.
Note that TBDPS is a tert-butyldiphenylsilyl group, which is akin to the TBS group discussed in
this Chapter.





36
Jana, N.; Nanda, S. Tetrahedron: Asymmetry, 2012, 23(10), 802-808.
224

2. Ogura and Usuki reported the Total synthesis of acerogenins E, G and K, and centrolobol in
the journal Tetrahedron.
37
The acerogenins were isolated from the stem bark of Acer nikoense
MAXIM, which has been used in folk medicine to treat diseases of the eye and liver disorders.

Show how acerogenin G and centrolobol can be prepared from the ketone precursor shown
below.



3. Based on the reactions in question 2, suggest an appropriate precursor and synthesis for the
preparation of acerogenins E and K as described in the same publication.



4. Uang and co-workers of the National Tsing Hua University and the Taipei Medical University
reported the Synthesis of ()-pterosin A via Suzuki-Miyaura cross-coupling reaction.
38
Fill in
the missing reagents, intermediates, and product in the scheme below. As you work through this
problem, note the following:

- LAH is an abbreviation for lithium aluminum hydride.

- TESCl is triethylsilyl chloride, which is similar to TBSCl.

- TIPS stands for triisopropylsilyl.

37
Ogura, T.; Usuki, T. Tetrahedron, 2013, 69(13), 2807-2815.
38
Hsu, S.-C.; Narsingam, M.; Lin, Y.-F.; Hsu, F.-L.; Uang, B.-J. Tetrahedron, 2013, 69(12),
2572-2576.
225


- PDC is pyridinium dichromate, which is analogous to PCC (pyridinium chlorochromate).

- As noted in problem 1, TBAF is an alternate source of fluoride.



5. Barua and co-workers reported a synthesis of a particular oxylipin.
39
Oxylipins are oxygen-
containing natural products derived from fatty acids. Their target, a 6,9,10-trihydroxyoctadec-7-
enoic acid (shown below), was an compound known to be an immunostimulant.



The synthesis included the following sequence. Working backwards from the epoxide product,
provide the structures of compounds A D.





39
Saikia, B.; Devi, T. J.; Barua, N. C. Tetrahedron, 2013, 69(9), 2157-2166.
226

6. The racemic epoxide prepared in the previous question was subjected to hydrolytic kinetic
resolution (HKR).
40
In HKR, a racemic epoxide is treated with a chiral (salen)Co
III
catalyst,
which facilitates the opening of the epoxide with water. However, given the catalysts chirality,
one epoxide enantiomer opens more rapidly than the other. As shown below, this results in an
enantioenriched epoxide and a diol.

(R,R)-(salen)Co
III
OAc
H
2
O, 0
o
C
OTBS
O
OTBS HO
OH
OTBS
O
(R)
(S)
+


The (S)-diol was then selectively tosylated at the primary alcohol, as shown below. The tosylate
was subsequently treated with potassium carbonate. What is the structure of compound A?



7. Weve seen that methyl ethers are typically only used as protection for phenolic hydroxyl
groups. Danishefskys diene is an example in which a methyl ether can serve as a sort of proxy
for a hydroxyl group that is not phenolic. Honda and co-workers at Stony Brook University
reported the following microwave (MW)-assisted Diels-Alder reaction between Danishefskys
diene and a TBS-protected -(hydroxymethyl)acrylate.
41
Treatment of the Diels-Alder adduct
with camphorsulfonic acid yields an ,-unsaturated ketone. Camphorsulfonic acid is a source
of H
+
that is soluble in organic solvents, such as THF (which often contains some moisture).
Provide the structures of compounds A and B, as well as a mechanism to account for their
formation.



40
For more information on HKR, see Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga,
M.; Hansen, K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N. Journal of the American
Chemical Society, 2002, 124(7), 1307-1315.
41
Zheng, S.; Chowdhury, A.; Ojima, I.; Honda, T. Tetrahedron, 2013, 69(8), 2052-2055.
227


8. Mehta and Bera published An approach toward the synthesis of PPAP [Polycyclic
Polyprenylated Acyl Phloroglucin] natural product garsubellin A: construction of the tricyclic
core in which an interesting fragmentation / aldol cyclization cascade is employed.
42
This
cascade reaction is shown below. Provide a mechanism for the transformation. Recall from
introductory Organic Chemistry that DIBAL-H (diisobutylaluminum hydride) is a hydride-
transfer reagent.







42
Mehta, G.; Bera, M. Tetrahedron, 2013, 69(7), 1815-1821.
228




Chapter 11:
Synthesis and
protections of amines

229

Reductive synthesis of amines

During Introductory Organic Chemistry, you learned that amides can be reduced to amines using
lithium aluminum hydride.



Nitriles can be similarly reduced to amines. Since the installation of a nitrile is easily
accomplished by S
N
2 reaction between a substrate bearing a leaving group and cyanide, this
affords an opportunity to lengthen the backbone by a single carbon atom.



During our discussion of aromatic chemistry, we were also reminded of the active metal
reduction of aromatic nitro groups to aniline or its derivatives.



Alkylation of ammonia or amines

A classical amine synthesis is direct alkylation with an alkyl halide, or another comparable
electrophile. However, over-alkylation is a problem associated with this method. As an
example, consider the alkylation of ammonia with methyl bromide. The initial attack of
ammonia on electrophilic methyl group yields an ammonium ion, which can be deprotonated by
another molecule of ammonia. A primary amine is the result.



However, this primary amine is now more reactive than ammonia. It bears an electron-donating
methyl group that enhances the electron density on nitrogen. As a result, methylamine
outcompetes ammonia for the remaining methyl bromide.

230



A second round of alkylation and proton loss yields a secondary amine. Yet again, the
nucleophilicity of the amine has been enhanced, making dimethylamine the strongest nucleophile
in the system.



Consequently, another iteration of methylation follows to produce the even more nucleophilic
trimethylamine.



Finally, a fourth alkylation affords the quaternary ammonium ion, which can be alkylated no
further.



One solution to the problem of over-alkylation is the use of a large excess of ammonia. This
lowers the statistical probability that methylamine can actually encounter a second molecule of
the electrophile in the reaction mixture. As a result, the primary amine can be synthesized
without additional alkylation.



231

Gabriel synthesis

There are also mechanistically distinct approaches to amine synthesis that do not suffer from the
problem of over-alkylation. One such method is known as the Gabriel synthesis, and it uses
phthalimide as a reactant. Deprotonation yields potassium phthalimide, whose negative charge is
significantly stabilized through resonance. This nucleophile can then be used in alkylation to
afford a substituted phthalimide.



The primary amine can then be excised from the substituted phthalimide using hydrazine. The
mechanism involves two rounds of nucleophilic acyl substitution and displaces the free primary
amine as a leaving group during the formation of the phthalyl hydrazine byproduct.



Reductive amination

Reductive amination is a useful synthesis of primary, secondary, or tertiary amines. It uses a
different source of electrophilic carbon, namely a ketone or an aldehyde. Ammonia or a primary
amine can condense with the ketone or aldehyde to form an imine. This reaction will be
discussed further in Chapter 12 (Synthesis and protection of aldehydes, ketones, and their
derivatives).



The imine can then be reduced by sodium borohydride in a protic solvent, such as methanol.
The product in this case is a secondary amine.

232



The reduction involves the donation of hydride from sodium borohydride to the imine carbon. It
is unlikely that a free amide (i.e. amine anion) is formed since this would be rather high in
energy. The nitrogen likely acquires a proton from the solvent concurrently with the addition of
hydride, or coordination with boron may stabilize the nitrogen.



Tertiary amines are formed by in situ reduction of the iminium ion that results when a secondary
amine condenses with a ketone or aldehyde. This requires the usage of a milder reducing agent,
like sodium cyanoborohydride, which will not reduce the ketone or aldehyde reactant.



Staudinger reduction

The Staudinger reduction generates primary amines, like the Gabriel does; however, the method
is altogether different. The nitrogen is introduced in the form of azide and is then reduced
through the action of triphenylphosphine.



The organic azide is formed via the S
N
2 reaction of a substrate bearing a good leaving group,
such as an alkyl halide, with sodium azide.



233

It is easier to understand what happens next if we consider a resonance form of the azide that
places the anion on the interior nitrogen atom. Triphenylphosphine attacks the terminal nitrogen
of the azide, and the betaine closes to form an unusual four-membered ring. This intermediate
fragments so as to expel nitrogen and generate an iminophosphorane. In the presence of water,
this iminophosphorane hydrolyzes, much as an imine would, to liberate the primary amine along
with triphenylphosphine oxide.



Hofmann rearrangement

The Hofmann rearrangement is also a primary amine synthesis; however, it differs from the other
syntheses in a pronounced way: it is a degradation that results in the loss of one carbon atom.



The reaction begins with the bromination of the amide. The amide can be deprotonated in basic
media, and the enhanced nucleophilicity of the nitrogen results in its attack on bromine.

N
O
H
H
Br Br
N
O
Br
H
Na
OH
N
O
H


A second deprotonation incites the key rearrangement. As a lone pair of electrons from nitrogen
forms a new carbon-nitrogen bond, the adjacent carbon-carbon bond migrates, attacking
nitrogen and displacing its bromine. The intermediate that results contains the isocyanate
functional group.

234



The isocyanate carbon is then attacked by water. A proton is lost from the incipient oxonium ion
and gained by the incipient nitrogen anion, yielding a carbamic acid.



Additional proton transfers lead to the ionization of the carbamic acid. At this point, carbon
dioxide is extruded from the molecule, affording the truncated primary amine.



Protection of amines: amides

In synthetic endeavors, it is almost always necessary to protect amino groups. Amines are
amphoteric and are potent nucleophiles, making them problematic for many otherwise
straightforward transformations. For instance, the direct nitration of aniline would be expected
to yield mostly para-nitroaniline. However, in actuality a mixture of meta and para-substituted
products is obtained. The para product results from the expected EAS reaction. The meta
product can be explained by the protonation of aniline in this acidic medium. The resultant
ammonium ion is a deactivating group and a meta director.

235



The protection of anilines amino group prior to nitration can allow the expected para product to
predominate. There are a number of ways to protect amino groups, and well consider only a
few in this chapter. One simple approach is the formation of an amide. Acetylation with acetyl
chloride (AcCl) proceeds via nucleophilic acyl substitution, and a non-nucleophilic base, such as
triethylamine (TEA) or pyridine, can be used to trap the acid that is released during this reaction.

The acetamide thus formed is still an ortho / para director because donation of nitrogens lone
pair into the ring through resonance enhances the electron density on the ortho and para
positions. However, nitrogens lone pair is also resonance delocalized into the amide carbonyl.
This renders that lone pair of electrons much less basic than in the free amine. Consequently,
EAS nitration proceeds without complication giving the para product. The amide can now be
cleaved, and one way to accomplish this is via nucleophilic acyl substitution with ammonia.
This particular deprotection method could also be termed transamidation, since the amide is
transferred from the substrate to the acetamide byproduct.


In general, amide protecting groups will be used mostly with aryl amines. Aliphatic amines are
not sufficiently good leaving groups, which makes the transamidation deprotection challenging.
As with each of the protecting groups that well study, there are a number of structural variations
on this theme. Other amides can be used to modulate the protecting groups stability for the
needs of the synthesis at hand.


236

Protection of amines: benzylamines

Lets consider another possible synthesis. Imagine that you have a substrate bearing both an
alcohol and an amine, and you wish to acetylate the alcohol only.



In practice, obtaining the desired selectivity without protecting group chemistry would be a
challenging endeavor. An attempt to directly acylate the alcohol will likely exhibit the undesired
chemoselectivity. The amine is the more nucleophilic of the two functional groups, so it is more
likely to be acylated, resulting in the formation of an undesired product.



Conversion of the amine to its benzylamine derivative is one potential solution. Treatment of the
amine with benzyl bromide (BnBr) and a non-nucleophilic base will result in alkylation of the
nitrogen. The tertiary benzylamine derivative is now non-nucleophilic. At this point, acetylation
can be carried out on the hydroxyl group to form the desired ester.



Benzylamines are convenient amine protecting groups because there is a way to cleave the
carbon-nitrogen bond to unveil the original amine substrate. Hydrogenolysis cleaves the carbon-
237

nitrogen bond through the addition of hydrogen. The original amine is liberated and toluene is
released as a byproduct.



Protection of amines: carbamates

Peptide synthesis is a common procedure during which amino group protection is necessary.
Peptides are synthesized by linking together amino acids in a condensation reaction. Well learn
more about peptide bond formation and the reagents used for this transformation during Chapter
13 (Synthesis and protection of carboxylic acids and their derivatives). For the moment, lets
consider the condensation from the standpoint of selectivity. Alanine (Ala) and valine (Val) are
two amino acids that could be joined to form a dipeptide (i.e. a peptide consisting of two amino
acids). However, if the two free amino acids are subjected to condensation (i.e. amide bond
formation with the concomitant the loss of water), a mixture of four dipeptides will be formed.
Two of the dipeptides are homodimers (Ala-Ala and Val-Val), while the other two are
heterodimers (Ala-Val and Val-Ala). Without examining the structures, Ala-Val and Val-Ala
may sound like the same thing. However, the structures show that they are, in fact, constitutional
isomers. In one isomer (Ala-Val), the alanine is the amino-terminus of the dipeptide. In other
words, it is the side with the free amine group. In the other isomer (Val-Ala), the alanine is the
carboxy-terminus, meaning that it has the free acid.


238


In order to conduct the reaction selectively so as to obtain a single product, protecting groups
must be employed. One carboxylic acid must be protected, and one amino group must be
protected. This leaves only one free acid and amine available for condensation.



The acid may be protected as its methyl ester (more on carboxylic acid protecting groups in
Chapter 13). The amino group may be shielded from undesired reaction by converting it to its
tert-butoxycarbonyl (also called t-Boc or simply Boc) derivative. This is an example of a
carbamate protecting group.



The Boc group is installed using a reagent known as di-tert-butyl dicarbonate. This reagent is
also sometimes referred to as Boc anhydride (Boc
2
O). Nucleophilic acyl substitution results in
the acylation of the amino group. The byproduct degrades to carbon dioxide and tert-butanol.
This protected amino acid can be called Boc-Ala.



The resonance delocalization of nitrogens lone pair into the carbamate carbonyl is the salient
feature. This delocalization makes the nitrogen much less basic and nucleophilic than it was in
the free amino group. Consequently, the peptide bond formation can be carried out without
interference from this nitrogen. As we saw earlier, amides are also protecting groups that
delocalize nitrogens lone pair through resonance. What differentiates carbamates from amides
is the greater ease with which carbamates can be removed when they are no longer needed. The
239

cleavage of amide protecting groups requires rather strong conditions if the amine is aliphatic.
However, carbamates can be cleaved from aliphatic amines using gentler conditions that are
more compatible with a wide variety of substrates. The Boc carbamate is susceptible to cleavage
under acidic conditions. For instance, in this example, treatment with trifluoroacetic acid (TFA)
will release Ala-Val-OMe.



The mechanism of cleavage involves initial protonation of the carbamate carbonyl. The tert-
butyl to carboxyl oxygen bond cleaves rather readily because the relatively stable tert-butyl
carbocation is released. The substrate is thus converted to a carbamic acid.



The carbamic acid can ionize through proton transfer. Subsequent loss of carbon dioxide
completes the deprotection.



The tert-butyl carbocation that was released during this process can be captured by
trifluoroacetate. However, trifluoroacetate is not especially nucleophilic, and as a result, a
substrate containing nucleophilic substituents might be alkylated by the tert-butyl carbocation
before this cation is captured by trifluoroacetate. This alkylation would generally be undesirable.

240



To avoid any undesired alkylation by the tert-butyl carbocation, a carbocation scavenger is
sometimes added to the reaction mixture. Anisole is such a scavenger because it can undergo
EAS alkylation with the carbocation, thereby trapping it. Anisole possesses a reasonably
nucleophilic aromatic ring due to the resonance electron donation of the methoxy group. This
allows it to capture the carbocation, but by the same token, anisole is not so nucleophilic that it
would cause undesired side reactions of the substrate.



Carbamates are widely used amine protecting groups, and there are carbamates with a range of
stabilities and cleavage conditions. The Boc group is fairly resistant to base and moderate
nucleophiles because the tert-butyl group sterically shields the carbamate carbonyl. As we saw,
the Boc group is cleaved in acid. There are orthogonal protecting group options. This means
that there are other carbamates that are stable under conditions that would cleave the Boc group,
but cleave under conditions where Boc would be stable. In other words, they have stability and
cleavage conditions complementary to those of the Boc group. Two examples follow.

The 9-fluorenylmethoxycarbonyl (FMOC) group is one carbamate protecting group orthogonal
to Boc. It is installed via nucleophilic acyl substitution using FMOC-Cl and a non-nucleophilic
base.



Unlike Boc, the FMOC group is stable under mildly acidic conditions, and it cleaves upon
exposure to base. For instance, a tertiary amine base, like piperidine, will deprotonate the
FMOC group at the 9-position. The carbon-hydrogen -bonding electrons form a carbon-carbon
bond. In the process, carbon dioxide is lost, and the nitrogen is protonated by the conjugate
acid of piperidine.
241




Another carbamate protecting group orthogonal to Boc is the carboxylbenzyl group, which is
abbreviated as Cbz or sometimes simply as Z. It is also installed through nucleophilic acyl
substitution using a non-nucleophilic base.



The Cbz group is reasonably stable in mild acid and mild base, which would cleave the Boc and
FMOC groups, respectively. However, the Cbz group is cleaved by hydrogenolysis, which
would not affect the Boc or FMOC groups. As we saw with the protection of amines as their
benzylamine derivatives, hydrogenolysis adds hydrogen across the benzylic carbon to
heteroatom bond, thereby releasing toluene. The intermediate carbamic acid expels carbon
dioxide, as weve seen in previous examples, to liberate the free amine.



Now that we know a bit about orthogonal protecting groups it is possible to envision syntheses
involving polyfunctional substrates. If needed, each of these functional groups can be protected
with a group that can then be selectively removed to allow reaction at solely that center. Many
actual total syntheses reported in the primary literature utilize just such an approach.


242

Chapter 11 Problems

1. In their synthesis of the alkaloid serratezomine A, Johnston and co-workers of Vanderbilt
University and the Indiana University Molecular Structure Center prepared the imine shown
below from a haloalkyne starting material.
43
Show a method for this conversion.



2. In their synthesis of cleavamine-type indole alkaloids, Bennasar et al. employed the following
reaction sequence.
44
Provide the structures of compounds A and B. Note that sodium
triacetoxyborohydride is a weak reducing agent, much like sodium cyanoborohydride.



3. Later in the synthesis discussed in question 2, the following transformation was performed.
What is the expected product?



4. Matsuda and co-workers of Hokkaido University in Sapporo, Japan published the total
synthesis of the natural product kalkitoxin.
45
Kalkitoxin was isolated from a cyanobacterium and
is an ichthyotoxin (i.e. it is toxic to fish). Kalkitoxin is shown below, and the segment
highlighted in the box was prepared using a Staudinger reduction.


43
Pigza, J. A.; Han, J.-S.; Chandra, A.; Mutnick, D.; Pink, M.; Johnston, J. N. Journal of
Organic Chemistry, 2013, 78(3), 822-843.
44
Bennasar, M.-L.; Sol, D.; Zulaica, E.; Alonso, S. Tetrahedron, 2013, 69(12), 2534-2541.
45
Umezawa, T.; Sueda, M.; Kamura, T.; Kawahara, T.; Han, X.; Okino, T.; Matsuda, F. Journal
of Organic Chemistry, 2012, 77(1), 357-370.
243



The Staudinger reduction used in this synthesis is shown below. Provide a mechanism for this
transformation.



5. Silodosin is a medication for the treatment of symptoms associated with enlargement of the
prostate. Fill in the intermediates in the synthesis of silodosin shown below.
46




6. Reagents other than bromine (Br
2
) can be used to induce Hofmann rearrangement. In a report
by Hoffmann-La Roche, Hofmann rearrangement of the following amide is used as a step in the
synthesis of carmegliptin, a drug candidate with therapeutic potential for type 2 diabetes.
47
Show

46
Barve, I. J.; Chen, L.-H.; Wei, P. C. P.; Hung, J.-T.; Sun, C.-M. Tetrahedron, 2013, 69(13),
2834-2843.
47
Abrecht, S.; Adam, J.-M.; Bromberger, U.; Diodone, R.; Fettes, A.; Fischer, R.; Goeckel, V.;
Hildbrand, S.; Moine, G.; Weber, M. Organic Process Research & Development, 2011, 15(3),
503-514.
244

a mechanism and product for the Hofmann rearrangement of this substrate using NBS as the
source of electrophilic bromine.



Note that the authors actually used an even more modern alternative reagent for this
transformation that is known as iodosobenzene diacetate (PIDA).

7. Bunce, Squires, and Nammalwar of Oklahoma State University reported the reaction of
enones, like the one shown below, with amines, such as benzylamine.
48
If the product of this
reaction (compound A) is subjected to hydrogenolysis, compound B results. Compound B has
the molecular formula C
9
H
10
N
2
O. Shown the structures of compounds A and B, and provide a
mechanism for the formation of compound A.



8. Helicenes are interesting molecules because their aromatic rings bend out of planarity in order
to avoid repulsion of proximal groups, which would occur if they were flat structures.
Consequently, they can attain a helical chirality. In their synthesis of helicenes, such as the
[5]helicene shown below, Roy and Basak performed the following conversion of a primary
allylic alcohol to the corresponding primary allylic amine.
49
Provide reagents and intermediates
for this transformation.




48
Bunce, R. A.; Squires, S. T.; Nammalwar, B. Journal of Organic Chemistry, 2013, 78(5),
2144-2148.
49
Roy, S.; Basak, A. Tetrahedron, 2013, 69(9), 2184-2192.
245

9. A team of investigators from Arizona State University and the National Cancer Institute
reported the synthesis and biological evaluation of 3,4-methylenedioxy-5,4-dimethoxy-3-
amino-Z-stilbene.
50
Using their starting materials (shown below), propose a synthesis of this
target molecule.



10. Harada, Fujii, Odai, and Kato published a scaleable route to a key intermediate in the
synthesis of a compound dubbed AJ-9677, which has clinical potential for the treatment of
obesity, type 2 diabetes, and related conditions.
51
Provide the structures of compounds A C in
their route. Show mechanism for the formation of compounds A and C.

In the transformation that leads to the formation of compound A, it will be helpful to know that
the indole ring is nucleophilic at the 3 position. Additionally, there will be a tendency to restore
aromaticity following the initial nucleophilic attack.

O
N
H
MeMgBr
O
N
MgBr
1
2
3
O
Cl
H
N
Me
FMOC
Compound A
NaBH
4
cat. HCl
EtOH
Compound B
HN
CH
2
Cl
2
Compound C
1.
2. H
3
O
+









50
Pettit, G. R.; Anderson, C. R.; Herald, D. L.; Jung, M. K.; Lee, D. J.; Hamel, E.; Pettit, R. K.
Journal of Medicinal Chemistry, 2003, 46(4), 525-531.
51
Harada, H.; Fujii, A.; Odai, O.; Kato, S. Organic Process Research & Development, 2004,
8(2), 238-245.
246





Chapter 12:
Synthesis and
protection of aldehydes,
ketones, and their
derivatives
247

Oxidative synthesis of aldehydes

In Introductory Organic Chemistry, you learned that chromic acid (H
2
CrO
4
), which is formed
from sodium dichromate in sulfuric acid, is a convenient oxidant. However, it is a very strong
oxidant as well and, as a result, is not useful for the synthesis of aldehydes. When a primary
alcohol is treated with chromic acid, the aldehyde is produced, but only transiently. It quickly
forms a hydrate in this aqueous medium, and then a second round of oxidation yields the
carboxylic acid. You also learned that pyridinium chlorochromate (PCC) is an alternative
oxidant that does afford the aldehyde. PCC is soluble in organic solvents, like dichloromethane,
so once the aldehyde is formed, it is not converted to the hydrate since no water is present.
Consequently, the oxidation stops at this stage.



An alternative oxidation that also yields the aldehyde is the Swern oxidation. It employs
dimethyl sulfoxide (DMSO) and oxalyl chloride [(COCl)
2
] in its first step and a base, such as
triethylamine, in its second step.



Dimethyl sulfoxide has two resonance forms. The second resonance form shown below is an
important contributor to the resonance hybrid, which has significant partial charges on sulfur and
oxygen. One justification for the significance of this resonance form is that the 2p-3p overlap to
form the sulfur-oxygen bond is poor due to the size difference between these orbitals. In any
event, the electron-rich oxygen attacks a carbonyl carbon of oxalyl chloride in a nucleophilic
acyl substitution reaction. The sulfonium ion is subsequently attacked by chloride, which
initiates a fragmentation resulting in the release of carbon dioxide, carbon monoxide, and
chloride as byproducts. The dimethylchlorosulfonium ion is instrumental in the oxidation of the
alcohol.

248



The alcohol now attacks the sulfonium ion, thereby displacing chloride. Loss of a proton
generates a sulfonium ion that persists until the addition of base occurs.



A base of modest strength, like triethylamine (TEA), is sufficiently strong to deprotonate the
position adjacent to sulfur. The ylide that results benefits from the attraction of the adjacent
charges to one another. It is at this stage that the organic substrate is finally oxidized. The ylide
deprotonates the -carbon of the alcohol. The carbon-hydrogen -bonding electrons form the
carbonyl bond while the oxygen-sulfur bond cleaves and forms a lone pair of electrons on
sulfur. The products are an aldehyde if a primary alcohol was the substrate (or a ketone if a
secondary was used instead) and dimethyl sulfide.



Additionally, it is worthwhile to brief review from your Introductory course that ozonolysis is an
oxidative method to prepare aldehydes (or ketones) that involves scission of a carbon-carbon
double bond.



Protection of aldehydes: acetals

Aldehydes are reasonably electrophilic, and it is therefore often necessary to protect them during
syntheses that involve nucleophilic reagents. As an example, we might consider the following
transformation in which the goal is to add a Grignard or organolithium reagent to the ketone
while leaving the aldehyde untouched. This is not possible with direct addition of the Grignard
or organolithium reagent of course. Such a strong nucleophile would add indiscriminately to
both the aldehyde and ketone. Even if some selectivity could be achieved, it would undoubtedly
249

be for reaction at the more electrophilic aldehyde. Consequently, protection of the aldehyde will
be necessary.



An acetal is convenient for this purpose. Recall from your Introductory course that an aldehyde
treated with two equivalents of alcohol in the presence of catalytic acid will establish an
equilibrium with the acetal and water. This freely reversible process can be driven toward the
acetal through the addition of excess alcohol.



Furthermore, if the two equivalents of alcohol happen to be tethered to one another, a cyclic
acetal results. In general, acetals can be useful aldehyde protecting groups. They are stable in
base and stable to nucleophiles; however, they can be readily removed in aqueous acid. Cyclic
acetals are often favored because they are more robust and therefore less likely to undergo
premature hydrolysis.



In our example synthesis, the aldehyde can be converted to its cyclic acetal using ethylene glycol
and catalytic acid. While it is true that ketones can also form acetals, selectivity can be achieved
because the more electrophilic aldehyde will be the first to react with the nucleophilic alcohol.
After protection as the cyclic acetal, ethylmagnesium bromide can be added to the carbonyl.
Workup yields the tertiary alcohol, and acidic hydrolysis unveils the aldehyde.



This synthesis could also be shortened by one step. If the workup for the Grignard reaction
utilizes aqueous acid, instead of merely water, then the alkoxide can be protonated and the
protecting group removed in a single operation.

250



It is also worth noting that, in this specific example, intramolecular attack of the alcohol on the
aldehyde would likely occur spontaneously, providing a cyclic hemiacetal. Cyclic hemiacetals
are often referred to as lactols.



Oxidative synthesis of ketones

The oxidative methods that are useful for the synthesis of aldehydes are also employed in the
synthesis of ketones, but with less complication. Since ketones are less susceptible to over-
oxidation, chromic acid, PCC, or the Swern oxidation can be used with equal efficacy to oxidize
a secondary alcohol to the ketone. In a complex substrate, the selection of conditions may be
driven by concerns about undesired reaction with other functional groups in the molecule.



Additionally, ozonolysis of alkenes can yield ketones if the substitution of the alkene is
appropriate. Here, a tetrasubstituted alkene is ozonolyzed to two equivalents of 2-butanone.



Friedel-Crafts acylation, which you learned about in your Introductory Organic course, is another
oxidative method for the synthesis of ketones, specifically alkyl aryl ketones.

251



At first, this may not appear to be an oxidative method because we often approximate oxidation
state in Introductory courses by simply counting the bonds to heteroatoms. It is worth noting
that the determination of oxidation state can be slightly more subtle than that. Formally, each
bond to an element more electronegative than carbon yields a +1 contribution to the oxidation
state, while each bond to an element less electronegative than carbon offers a -1 contribution.
Consequently, the carbon of benzene undergoing acylation begins with a -1 oxidation state (3
bonds to carbon and 1 bond to hydrogen) and ends with an oxidation state of 0 (4 bonds to
carbon). Since the oxidation state increases for this carbon, it is oxidized during the reaction.



Synthons and synthetic equivalents

E. J. Corey developed and codified many important strategies that enabled chemists to approach
synthesis rationally and with great success. His contribution in developing the language of
retrosynthetic analysis was recognized by the 1990 Nobel Prize in Chemistry.
52
Corey coined
the term synthon to refer to a building block that you might like to have but cannot necessarily
produce and store in a bottle. This concept enables you to fully apply your imagination to
synthesis. Think of any building block that you wish you could have for your synthesis, no
matter how crazy it might appear. For instance, the diagram below shows some examples of
carbonyl synthons.

C
O
C
O
C
O
Carbonyl synthons


We can then turn our attention to identifying synthetic equivalents for these synthons. The
synthetic equivalents are the real molecules that exhibit reactivity like the synthon you require
for your synthesis. The following diagram shows synthetic equivalents corresponding to each of
the synthons listed above.

52
For information about this Nobel Prize, see
http://www.nobelprize.org/nobel_prizes/chemistry/laureates/1990/ (accessed February 18, 2013).
For E. J. Coreys full treatment of synthetic strategies, see The Logic of Chemical Synthesis by E.
J. Corey and X.-M. Cheng (Wiley: New York, 1995).
252




Phosgene (COCl
2
) can undergo two rounds of nucleophilic addition, so in that sense, it behaves
like the doubly positive carbonyl synthon, which would be receptive to two nucleophilic attacks.
Well examine this behavior in Part V, when we study polymers. It turns out that phosgene is an
important polymer building block.

Cyanide can behave as a carbonyl that can both perform a nucleophilic attack and accept a
nucleophile. For instance, cyanide can be readily alkylated via S
N
2 reaction with an alkyl halide.
Here, it performs a nucleophilic attack.



The nitrile thus formed can subsequently be attacked by a nucleophile. For example,
methylmagnesium bromide can add to the nitrile. Workup with aqueous acid yields the imine,
and the imine undergoes hydrolysis to the ketone via nucleophilic additionelimination, which
well review shortly in our discussion of imines and enamines. This ketone was synthesized by
adding one R group as an electrophile and one R group as a nucleophile, meaning that the
carbonyl behaved as a synthon with both a positive and negative charge.



It is also possible to employ a doubly negative carbonyl synthon. The synthetic equivalent for
this fragment is 1,3-dithiane, which can be deprotonated by a strong base like butyllithium. The
corresponding 1,3-dioxane (the oxygen-containing analogue of 1,3-dithiane) cannot be similarly
deprotonated. 1,3-Dithianes greater acidity is due to the ability of sulfur to stabilize the anion
through a resonance form in which the sulfur exhibits an expanded octet. Oxygen is unable to
provide this resonance stabilization, accounting for 1,3-dioxanes reduced acidity.
253




The anion thus formed can be alkylated with a substrate such as benzyl bromide.



At this stage, there are two options. Hydrolysis of the thioacetal can be achieved in water with
the addition of a mercury(II) salt. Mercurys high affinity for sulfur facilitates the cleavage of
the thioacetal, and mercury behaves as a Lewis acid during the cleavage. An aldehyde would be
the product if we choose this option.



Alternatively, a second round of deprotonation and alkylation can be performed prior to the
hydrolysis. In this case a ketone is formed.



When two rounds of alkylation are conducted, the net result is the formation of a ketone product
through the addition of both R groups as electrophiles. Hence, the carbonyl behaves as though it
were doubly nucleophilic. Sometimes this strategy is referred to as using an umpolung.
Umpolung is the German word for reversed polarity. By this, we mean that carbonyls are
typically electrophilic. If they behave as nucleophiles, then they are exhibiting umpolung
behavior.

254



Protection of ketones: acetals

The protection of ketones is often accomplished through acetal formation, akin to what we
studied for aldehydes. Lets consider one synthetic example to showcase the occasional need for
ketone protection. Imagine that you need to convert to the following ketoester in the -ketoester
shown.



This synthesis would call for Claisen condensation with the enolate of methyl acetate. However,
direct addition of that enolate to the original substrate would result in a mixture of aldol and
Claisen products. If selectivity were achieved, it would be for addition to the more electrophilic
ketone. Since this is not the objective, it is necessary to protect the ketone as its cyclic acetal.
Then, the crossed Claisen condensation can be achieved. Workup with aqueous acid protonates
the enolate and hydrolyzes the acetal, unveiling the ketone.



Synthesis of imines and enamines: nucleophilic additionelimination

Imines and enamines are generally synthesized from the corresponding aldehydes and ketones.
You learned this reaction in Introductory Organic. It occurs through nucleophilic addition of the
amine to the carbonyl. Subsequent elimination of water yields the imine. Mildly acidic
255

conditions are usually employed. The mild acid is needed for the protonation of the hydroxyl
group to allow its departure as water; however, strong acid must be avoided because it will cause
the amine to exist mainly in the protonated form, which is not nucleophilic.



Enamine synthesis is largely similar, except that the elimination of water yields an iminium ion,
which must lose a proton from an adjacent carbon to neutralize nitrogen, forming the enamine.



Both reactions are freely reversible, so it is necessary to drive the equilibrium through the
removal of water. This is often accomplished using a Dean-Stark apparatus (or Dean-Stark trap).
The mixture is heated at reflux in a solvent such as toluene that is immiscible with water. A
Dean-Stark trap is inserted between the round-bottom flask and the condenser. As solvent
condenses, it fills the Dean-Stark trap before returning to the reaction vessel. Since water is
more dense than toluene, it settles at the bottom of the trap. The water removed can be measured
to monitor the reactions progress. Some Dean-Stark traps also allow for the water to be drained
off.




256

Synthesis of 1,5-dicarbonyl compounds using enamines: Stork enamine synthesis

Enamines can be used to prepare substituted ketones. Using a ketone in a Michael reaction
typically does not work. The Michael donor (i.e. the enolate) must be stabilized by two
carbonyls for Michael reaction to proceed smoothly. Gilbert Stork developed a solution to this
problem, which employs an enamine as a proxy for the ketone enolate. First, the enamine is
prepared from the ketone as described above. This enolate can serve as the nucleophile in a
Michael reaction, and it adds to the -carbon of an ,-unsaturated carbonyl compound. Workup
with aqueous acid protonates the enolate and hydrolyzes the iminium ion to yield a 1,5-
dicarbonyl.



The net result of the Stork enamine synthesis is that the -carbon of the ketone acted as an
enolate in a Michael reaction, even though its direct behavior in that fashion would have been
ineffective.




257

Chapter 12 Problems

1. During the synthesis of a stereoisomer of sanjoinine G1 (shown below), a member of the
cyclopeptide alkaloid class of natural products, Joulli and co-workers employed a Jones
oxidation.
53




The Jones oxidation used is portrayed below. It is an alternative way of generating chromic acid.
Provide the final product of this oxidation.

O
HN
BocHN
OH
O
ZHN
OH
CrO
3
aq. H
2
SO
4
acetone


2. Rutjes and co-workers of Radboud University reported the total synthesis of integric acid.
54

Integric acid is an HIV integrase inhibitor and therefore has interesting therapeutic value. The
penultimate step in the total synthesis is an ozonolysis reaction. The alkene highlighted in the
box is the one that is subject to ozonolysis. Provide the structure of the ozonolysis product.



53
East, S. P.; Shao, F.; Williams, L.; Joulli, M. M. Tetrahedron 1998, 54(44), 13371-13390.
54
Waalboer, D. C. J.; van Kalkeren, H. A.; Schaapman, M. C.; van Delft, F. L.; Rutjes, F. P. J. T.
Journal of Organic Chemistry, 2009, 74(22), 8878-8881.
258

3. In Chapter 10, we saw a question about the synthesis of ()-pterosin.
38
The indanone skeleton
used in that question was prepared via tandem Friedel-Crafts / Nazarov cyclization. The first
step in the sequence is Friedel-Crafts acylation. Show the mixture of products expected for this
reaction.



The second step of the sequence is a Nazarov cyclization. A generic Nazarov cyclic is shown
below.

O
O
H
O
H
O
H


Show how this reaction paradigm can be used to explain the conversion of compounds A and B
into indanones C and D.



4. In Chapter 10, we also saw a problem dealing with the preparation of acerogenins E, G and K,
and centrolobol.
37
The precursors shown in that problem were prepared as follows. Provide the
structures of compounds A H.



As you approach this portion of the question, it will help to know that PPTS is pyridinium para-
toluenesulfonate, which acts as a mild acid.

259



Finally, the end game of this synthesis employs the Claisen-Schmidt condensation, which is the
crossed aldol reaction between a ketone and an aldehyde with no -protons. Pyrrolidine first
forms an enamine with the ketone. The enamine then serves as the nucleophile, in a manner
reminiscent of the Stork enamine synthesis. Show a mechanism for the formation of compound
G.



5. In this Chapter, weve learned about the use of cyanide as a synthetic equivalent for the
carbonyl group. In their synthesis of amphidinolide B
1
, a team of investigators from Oregon
State University and the Beckman Research Institute utilized this methodology.
55




In Chapter 13, well learn how they introduced the nitrile, and well also learn about a catalyst
(DMAP) that they used in the preparation of compound C. For the moment, show the structures

55
Lu, L.; Zhang, W.; Nam, S.; Horne, D. A.; Jove, R.; Carter, R. G. Journal of Organic
Chemistry, 2013, 78(6), 2213-2247.
260

of compounds A, B, and C in their synthesis. The red coloration shows where this fragment is
incorporated into the natural products structure.



6. Lee and Zhao of Berlex Biosciences used dithiane chemistry en route to marine natural
products with novel bioactivity.
56
Fill in the structures of A D in their synthetic scheme below.
Note that THP is the abbreviation of a protecting group for alcohols (2-tetrahydropyranyl).



7. In their synthesis of (-)-alkaloid 205B, Smith and Kim used a dithiane as a synthetic
equivalent for the carbonyl group.
12
They call their approach anion relay chemistry (ARC). The
dithiane anion first attacks an electrophile. This is followed by a Brook rearrangement.
Remember from problems youve encountered earlier in this text that the Brook rearrangement
involves the transfer of a silyl group from carbon to oxygen. The carbanion generated from the
Brook rearrangement then attacks a second electrophile. Thus, the dithiane becomes a linchpin
in the synthesis. Using these clues, provide the structures for compounds A C below.


56
Lee, S.; Zhao, Z. Organic Letters, 1999, 1(4), 681-683.
261



8. Investigators at Pfizer reported a stereoselective synthesis of spiropiperidines, among which
was a lead compound that functions as an inhibitor of -site amyloid precursor protein cleaving
enzyme (BACE-1).
57
One route that they utilized to reach the target employed a Strecker
reaction, in which an iminium ion is formed and then attacked by cyanide. Using this clue, show
the product in their reaction sequence below.



9. Yehia, Polborn, and Mller of the Organisch-Chemisches Institut in Heidelberg, Germany
reported A novel four component one-pot access to pyridines and tetrahydroquinolines.
58
In
devising their one-pot protocol, they employed a union between compound A and the ,-
unsaturated ketone shown below to yield compound B. This process later becomes a part of their
one-pot procedure. Show the structures of compounds A and B.




57
Lee, C.-W.; Lira, R.; Dutra, J.; Ogilvie, K.; ONeill, B. T.; Brodney, M.; Helal, C.; Young, J.;
Lachapelle, E.; Sakya, S.; Murray, J. C. Journal of Organic Chemistry, 2013, 78(6), 2661-2669.
58
Yehia, N. A. M.; Polborn, K.; Mller, T. J. J. Tetrahedron Letters, 2002, 43(39), 6907-6910.
262

10. Investigators at Ghent University reported an interesting ring expansionoxidation protocol
for the synthesis of piperidin-4-ones.
59
The mechanism bears some analogy to the Swern
oxidation. Propose a mechanism for this transformation.














59
Mollet, K.; Dhooghe, M.; Broeckx, L.; Danneels, B.; Desmet, T.; De Kimpe, N. Tetrahedron,
2013, 69(12), 2603-2607.
263




Chapter 13:
Synthesis and
protection of carboxylic
acids and their
derivatives

264

Oxidative synthesis of carboxylic acids

Since carboxylic acids are highly oxidized, it comes as no surprise that they are frequently
prepared through oxidative methods. We saw earlier that chromic acid (produced in situ from
sodium dichromate in sulfuric acid) would oxidize a primary alcohol to the carboxylic acid, via
an ephemeral aldehyde intermediate.



You may also recall from your Introductory course that the benzylic position has unusual
reactivity in that it can be oxidized to a carboxylate by potassium permanganate even if this
entails the cleavage of carbon-carbon bonds. The only requirement is for at least one benzylic
hydrogen to be present. The carboxylate is then protonated upon workup. Alternatively,
chromic acid oxidation yields the carboxylic acid directly.



Ozonolysis can also yield carboxylic acids. Here, there are two choices. An alkene can be
ozonolyzed to carboxylic acid products (provided that the alkene has the correct substitution
pattern) if hydrogen peroxide is employed during the workup step. Alkynes can also be
ozonolyzed to acids.



Reductive synthesis of carboxylic acids

While it may seem initially surprising, there are also some reductive approaches to the synthesis
of carboxylic acids. For instance, cyanide can serve as a synthetic equivalent for the carboxylic
acid synthon that is nucleophilic. Cyanide can be alkylated with an alkyl halide, and the nitrile
can subsequently be hydrolyzed to the carboxylic acid. The substrate is reduced in this process.

265



The use of cyanide as a synthetic equivalent is an example of umpolung reactivity because it is a
reversal of the usual polarity of a carboxylic acids carbonyl carbon. An alternative reductive
approach uses carbon dioxide as a synthetic equivalent for the carboxylic acid synthon that is
electrophilic. Carbon dioxide can be attacked by a Grignard reagent to install the carboxylate,
which is protonated on workup.



Note that both of these methods extend the carbon skeleton by one.

Protection of carboxylic acids: esters

Carboxylic acids contain an acidic proton that is readily removed. This can be problematic if the
molecule is exposed to basic reagents. Additionally, once deprotonated the reactivity of
carboxylic acids toward electrophiles is enhanced. Consequently, it is often necessary to protect
this functional group. Carboxylic acids are most frequently protected as esters. In the next
section, well turn our attention to the synthesis of esters, so we wont consider how these esters
are formed at the moment. We will, however, study the methods of deprotection.

Saponification, or basic ester hydrolysis, is a straightforward method for the deprotection of
simple esters, like the methyl ester. Hydroxide attacks the carbonyl carbon, and nucleophilic
acyl substitution results in the displacement of methoxide. Methoxide then deprotonates the
newly formed carboxylic acid. This drives the equilibrium toward the saponification products,
and the carboxylate is protonated during workup.

266



Orthogonal deprotection can be achieved with tert-butyl esters, which are labile in acid. An acid
such as trifluoroacetic acid (TFA) can protonate the carbonyl oxygen. The scission of the
carboxyl oxygen to carbon bond unveils the free acid and releases the tert-butyl carbocation. As
we saw when considering Boc deprotection in Chapter 11, the carbocation may be captured by
trifluoroacetate, but it may also be desirable to employ a carbocation scavenger if there is
concern about alkylation of the substrate.



Since it is often convenient to utilize multiple protecting groups that can be removed under
different conditions, it is also worthwhile for us to consider -substituted ethyl esters. One such
example is the 2,2,2-trichloroethyl ester. This ester is cleaved reductively through the addition
of zinc in acetic acid. Zinc effectively serves as a two-electron donor to the halogen. As
chloride is lost from the ester, its -bonding electrons form an incipient bond and displace the
carboxylate, which is immediately protonated by acetic acid.



Another -substituted ethyl ester is the 2-(trimethylsilyl)ethyl ester. In this instance, the
deprotection exploits the affinity of fluoride for silicon. The flow of electrons is similar to that in
the previous example. It is simply initiated differently. One possible fluoride source for this
deprotection is tetrabutylammonium fluoride (TBAF).

267



Now that we have discussed a number of functional groups and their protection, it is important to
point out that some protecting groups may be used with multiple functionalities. A good
example is the benzyl group. Benzyl esters can be made from carboxylic acids by deprotonating
the acid with a base like DBU. The resultant carboxylate will attack benzyl bromide, resulting in
alkylation to give the ester. Benzyl ethers can be made by quantitatively deprotonating an
alcohol with a strong base like sodium hydride and alkylating the alkoxide with benzyl bromide.
As we saw when discussing amines, their protection as the benzylamine is readily achieved
through alkylation as well.



This can be particularly helpful if several functional groups need to be protected until the end of
a synthesis. If analogous protecting groups are used for each functionality, the synthesis can be
shortened through a final global deprotection using a single set of conditions. As a simple
example, every benzyl group in the following molecule can be simultaneously removed by
hydrogenolysis.



Synthesis of acid chlorides

A classical way to prepare an acid chloride involves the treatment of a carboxylic acid with
thionyl chloride (SOCl
2
). It is likely that you encountered this in your Introductory Organic
course, but it is less likely that you discussed the catalysis of this reaction by dimethylformamide
(DMF).

268



Dimethylformamide is reasonably nucleophilic on the carbonyl oxygen. Attack of this oxygen
on thionyl chloride yields an iminium ion intermediate that adds chloride. Subsequent
displacement of sulfur dioxide yields the Vilsmeier reagent, which is a more active chlorinating
agent than thionyl chloride.



The Vilsmeier reagent is then attacked by the carboxylic acid. Chloride is ejected from the
tetrahedral intermediate to yield another iminium ion. Nucleophilic attack of chloride at the
carbonyl carbon results in the displacement (and regeneration) of the catalyst, along with the
production of the acid chloride product.



Oxalyl chloride, which we saw previously in the Swern oxidation, is an alternative chlorinating
agent that may be used in place of thionyl chloride. Attack of the carboxylic acid on one of the
carbonyls of oxalyl chloride results in an anhydride intermediate. When attack of chloride and
nucleophilic acyl substitution occur, carbon dioxide, carbon monoxide, and chloride are released
and the acid chloride is formed.

269



Oxalyl chlorides lower boiling point can make it an attractive alternative to thionyl chloride.
Furthermore, the reaction with oxalyl chloride can also be catalyzed by DMF.

Synthesis of esters: making the carbonyl carbon-carboxyl oxygen bond

There are two broad categories of ester syntheses: (1) forming the bond between the carbonyl
carbon and the neighboring carboxyl oxygen and (2) forming the bond between the carboxyl
oxygen and its neighboring R group. In this section, well consider the former category. Most
examples in this class of ester syntheses are nucleophilic acyl substitutions. One classic example
is the Fischer esterification reaction, in which a carboxylic acid and alcohol condense with the
loss of water to form an ester. This reaction is acid catalyzed and freely reversible.
Consequently, Le Chteliers principle must be exploited to drive the equilibrium. It is common
to use a large excess of the alcohol to drive the reaction toward completion when that alcohol is
cheap and readily available.



Intramolecular Fischer esterifications have the advantage of being entropically favored and
therefore do not require excess reagent. In the example below, a -hydroxyl acid cyclizes to
form a -lactone. A single reactant molecule forms two product molecules. Since the entropy of
the system increases through this process, it favors products.



When both the acid and the alcohol employed in the esterification are expensive or more difficult
to prepare, using an excess of one reagent is generally undesirable. In such cases, the carboxylic
acid can first be converted to the corresponding acid chloride, using thionyl chloride and
catalytic dimethylformamide. The alcohol (or phenol, as in the case shown below) can then be
acylated by exposure to this acid chloride. A non-nucleophilic base, such as TEA, is often used
270

to trap the acid liberated during the nucleophilic acyl substitution. If the acylation of the alcohol
is carried out with an acid chloride or anhydride in aqueous base (e.g. NaOH in H
2
O), it is
known as the Schotten-Baumann reaction.



This type of reaction can be catalyzed by dimethylaminopyridine (DMAP), which is an acyl-
transfer catalyst. In the example below, the acylating agent is an acid anhydride instead of an
acid chloride, but the two reactions are extremely similar.



The acylating agent is attacked by the nucleophilic DMAP catalyst. This nucleophilic acyl
substitution results in the acylation of DMAP. The pyridinium ion thus formed is an even more
active acylating agent than the original anhydride.



Consequently, it reacts rapidly with the alcoholic substrate in a second nucleophilic acyl
substitution. The catalyst is regenerated in this step, and the ester is formed upon loss of a
proton.


We will see DMAP used as a catalyst in other acylation reactions as well.

271

Synthesis of esters: making the carboxyl oxygen-alkyl group bond

A mechanistically distinct approach to ester synthesis is the formation of the bond between the
carboxyl oxygen and the adjacent R group. A simple way to achieve this involves the
deprotonation of the carboxylic acid, followed by its alkylation. One drawback of this method
lies in the weak nucleophilicity of the carboxylate. Any steric hindrance at the electrophilic
center renders this process low yielding. Consequently, it is most often employed for simple
methylation.



An alternative methyl ester synthesis uses diazomethane.



The electron-rich carbon of diazomethane first deprotonates the carboxylic acid. Then, the
carboxylate attacks the methyl group, which bears an exceedingly good leaving group (N
2
). This
is a very clean methylation, since the only byproduct is nitrogen gas. However, extreme care
must be taken with this reaction. Diazomethane is toxic, and it is an explosion hazard. Even
contact with ground-glass joints can cause an explosion. Therefore, special glassware must be
used in its preparation, and adequate ventilation must be assured as well.



There are two additional prominent examples of ester syntheses that form the carboxyl oxygen-
to-R group bond: the Baeyer-Villiger oxidation and the Mitsunobu reaction.

Synthesis of esters: Baeyer-Villiger oxidation

The Baeyer-Villiger oxidation is unique in that a ketone is the starting material for the
preparation of an ester. To this point, weve only considered ester syntheses starting with
carboxylic acids or their derivatives (i.e. acid chlorides or anhydrides). Treatment of the ketone
272

with a peroxy acid, such as meta-chloroperoxybenzoic acid (mCPBA) results in the formation of
an ester due to the migration of one of the ketone R groups.



The reactant ketone is protonated by the peroxy acid, and the conjugate base of the peroxy acid
then attacks the electrophilic carbonyl carbon. This step amounts to nucleophilic addition to the
ketone bond. The resultant tetrahedral intermediate contains a weak oxygen-oxygen bond,
which cleaves upon migration of one of the R groups. The hydroxyl group concomitantly loses a
proton to form the carbonyl bond. The ester is formed as the final product, and meta-
chlorobenzoic acid is a byproduct.



The ketone bears two R groups either of which could in principle shift from what used to be the
carbonyl carbon to the neighboring oxygen. Therefore, there are two conceivable isomeric ester
products that could be formed. However, the ability of R groups to shift (i.e. their migratory
aptitude) determines which isomer is actually obtained.



The diagram above shows that tertiary groups, like the tert-butyl group in the example reaction,
migrate more rapidly than aryl groups, like the phenyl group in the example reaction.
Consequently, only the product resulting from tert-butyl migration is observed.



273

Synthesis of esters: Mitsunobu reaction

The Mitsunobu reaction also results in the formation of an ester through synthesis of the
carboxyl oxygen to R group bond. However, it does so through a mechanism that leads to
inversion of configuration. Stereochemically, this is of course a very important attribute of the
reaction.



Mitsunobu reaction employs a reagent named diethyl azodicarboxylate (DEAD), as well as
triphenylphosphine.



The initial interaction is between these reagents. The nucleophilic triphenylphosphine adds to
the nitrogen-nitrogen bond. The nitrogen anion is then quenched through deprotonation of the
reactant carboxylic acid. Subsequent attack of the substrate alcohol on phosphorus severs the
triphenylphosphine moiety from the residue of DEAD, and proton transfer follows.



After this proton transfer, the critical S
N
2 displacement occurs. The hydroxyl group of the
reactant alcohol has been converted to a good leaving group, and the carboxylate displaces it as
triphenylphosphine oxide. Inversion of configuration results.
274




In addition to offering an ester synthesis, the Mitsunobu reaction also affords a way to invert the
configuration of an alcohol. If the ester is saponified, an alcohol results that differs from the
original reactant alcohol only in its configuration. This process is sometimes used in natural
product synthesis. If the alcohol configuration present in the natural product (i.e. the correct
configuration) cannot be prepared directly, it can be converted to the desired configuration using
this two-step process.

OH
DEAD
PPh
3
O
HO
O
O
NaOH
OH
Alcohol product with
inverted configuration


Synthesis of amides

Much like esters, amides are also frequently synthesized through Schotten-Baumann-style
reaction with an acid chloride or anhydride. Since amines are more basic than alcohols, the acid
liberated during the reaction presents a greater problem here. Two equivalents of amine can be
used with the understanding that one equivalent will be incorporated into the amide product and
the other equivalent is sacrificial, as it merely traps the acid released as the reactants condense.
If the amine is valuable, an alternative is to use only a single equivalent along with an equivalent
of a non-nucleophilic base, such as TEA or pyridine, to trap the acid formed.

OH
O
SOCl
2
DMF (cat.)
Cl
O
H
2
N
N
H
O
2 equivalents
+ H
3
N Cl
H
2
N
TEA (or pyridine)
N
H
O
+
H Et
3
N Cl
N
H
Cl
or


275

Amines cannot be condensed directly with carboxylic acids unless very high temperatures are
used. The reason is that a carboxylic acid and an amine will merely undergo acid-base reaction
to yield a salt. However, coupling reagents can be employed to effect the union of these
substrates. One such coupling agent is known as DCC.



DCC stands for N,N-dicyclohexylcarbodiimide



The reaction begins with a proton transfer from the acid to DCC. The carboxylate subsequently
attacks the electrophilic carbodiimide carbon. The result is a rather unusual looking ester in
which the carboxyl oxygen has been activated as a good leaving group. This intermediate is
sometimes called the activated ester as a result.



After protonation on nitrogen of the DCC moiety, attack of the amine on the carbonyl carbon
displaces the good leaving group. The product amide is formed along with a byproduct called
dicyclohexylurea.



Looking at the overall change in the reagent, it becomes apparent that DCC is merely a
dehydrating agent. It adds water across the carbodiimide as the reaction progresses.



276

An acyl group is transferred in this reaction from the activated ester to the amine. As we saw
with previous acyl-transfer reactions, DMAP can serve as a catalyst for the process. Sometimes
a non-nucleophilic base, like TEA, is also added to the reaction mixture to ensure a basic
environment.



The formation of amide bonds is especially important when the amide bonds in question are
between two amino acids. These specific amide bonds are referred to as peptide bonds.
Therefore, DCC is often called a peptide coupling reagent when it is used to link amino acids. If
you refer back to the example used during our discussion of carbamate protecting groups for
amines, youll see that DCC could be used for the peptide bond formation mentioned in that
example.

Peptide synthesis is so important that a wide array of peptide coupling reagents have been
developed to facilitate the synthesis of particular peptides of interest. One alternative to DCC is
EDAC. DCC is a useful reagent, but it is not without its drawbacks. DCC is a sensitizer. It is
often challenging to scrape out of the reagent bottle since it tends not to exist as a powder, which
would be easier to dispense. Additionally, the dicyclohexylurea byproduct can be challenging to
remove from the product mixture. It is usually precipitated in cold ether, but several rounds of
cooling and filtration can be needed to remove it completely. EDAC [1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide, also frequently abbreviated as EDCI or EDC] is a similar
carbodiimide-based coupling reagent that circumvents the disadvantages of DCC. EDAC bears a
dimethylaminopropyl group. The amine can be protonated to form the HCl salt of EDAC, which
is a nice, fluffy powder that is easy to dispense.



When used in amide bond formation, it too adds water across the carbodiimide. The substituted
urea byproduct can be easily removed from the product mixture using an aqueous acid wash.
Protonation of the dimethylamino group yields a salt that is reasonably water soluble, and can
therefore be washed away from the desired amide.



These sorts of reagents can also be used to form esters. When we discussed the protection of
carboxylic acids as esters, we deferred any consideration of how the installation of these
277

protecting groups is achieved. One convenient method is called the Steglich esterification. In
the Steglich esterification, a carboxylic acid and alcohol are united in an ester by DCC. DMAP
catalyzes the esterification. This method can be used to protect a carboxylic acid as its 2,2,2-
trichloroethyl ester. It can also be utilized in similar reactions to install other protecting groups.



Chapter 13 Problems

1. In the Chapter 12 problems, we saw an oxidation reaction involving a precursor in the
synthesis of sanjoinine G1 and its C-11 epimer.
53
Below are two additional steps in this
synthesis.

In this Chapter, weve discussed the activation of carboxylates for the formation of amides. An
additional method for carboxylic acid activation is the formation of a pentafluorophenyl (Pfp)
ester. The Pfp ester is formed in the following sequence by coupling pentafluorophenol with a
free acid using EDAC and catalytic DMAP. Using this clue, show a mechanism for the
formation of compound A. Hydrogenolysis then cleaves a protecting group, leading to
spontaneous cyclization. Show a mechanism for this cyclization that explains the formation of
compound B.



2. In the previous Chapter, we examined a series of steps from the synthesis of amphidinolide
B
1
.
55
The nitrile substrate for that sequence was prepared as follows.



This is an example of a Mitsunobu reaction using a nucleophile other than a carboxylate. Here,
the nucleophile is cyanide, which is released from acetone cyanohydrin. Show the product of
this step and a mechanism for its formation.

278

3. In the final step of their synthesis of (-)-platensimycin, Ghosh and Xi of Purdue University
treated a -trimethylsilylethyl ester with TASF.
60
TASF [tris(dimethylamino)sulfonium
difluorotrimethylsilicate] is an anhydrous source of fluoride, and its structure is shown below as
well. Provide a mechanism to explain the formation of the novel antibacterial natural product
platensimycin.



4. Investigators at the Changchun Institute of Applied Chemistry reported a one-pot synthesis of
pyrimidin-4(3H)-ones from 3-aminopropenamides using the Vilsmeier reagent.
61
Provide a
plausible mechanism for this transformation.

O
N
H
O
Ar
NH
2
N
N
O
Ar
Cl
2 equivalents
Vilsmeier
reagent


5. In the Chapter 12 problems, we saw one approach to spiropiperidines. The same authors
presented a second approach, part of which is illustrated below.
57
This approach utilizes the
Corey-Link reaction, in which -trichloromethyl alcohols are converted into amino esters (and
then amino acids if desired). Provide the structures of compounds A D, as well as mechanisms
to explain their formation. Note that compound A is actually a mixture of epimers, which are
separated by crystallization. Simply choose one epimer to utilize in the sequence.


60
Ghosh, A. K.; Xi, K. Journal of Organic Chemistry, 2009, 74(3), 1163-1170.
61
Zhang, R.; Zhang, D.; Liang, Y.; Zhou, G.; Dong, D. Journal of Organic Chemistry, 2011,
76(8), 2880-2883.
279

N
O
Cbz
CHCl
3
LiHMDS
-78
o
C
Compound A
(DBU)
N
N
Compound B
NaN
3
Compound C
MeOH
Compound D
azide
reduction
N
Cbz
NH
2
MeO
O


6. Magauer, Mulzer, and Tiefenbacher of the University of Vienna prepared (+)-echinopine B, a
novel sesquiterpenoid with an exceedingly interesting structure.
62
The final step in the synthesis
involved treatment of the carboxylic acid precursor with diazomethane. Provide the structure of
the product and the mechanism for the process.



7. Romn-Leshkov and co-workers reported the use of nanosheets for the Baeyer-Villiger
oxidation of cyclic ketones.
63
These nanosheets consisted of stannosilicate zeolites, which
served as Lewis acid catalysts for the process. Provide the product of the following Baeyer-
Villiger oxidation reported in their paper.



8. In the Chapter 10 problems, we examined the synthesis of a fragment used in the preparation
of 5-epi-paecilomycin-F by Jana and Nanda.
36
Now, lets consider the Mitsunobu reaction that
allows union of two major fragments of the molecule near the end of the synthesis. Show the
product of this transformation. Note that DIAD is diisopropyl azodicarboxylate, which is akin to
the reagent DEAD.


62
Magauer, T.; Mulzer, J.; Tiefenbacher, K. Organic Letters, 2009, 11(22), 5306-5309.
63
Luo, H. Y.; Bui, L.; Gunter, W. R.; Min, E.; Romn-Leshkov, Y. ACS Catalysis, 2012, 2(12),
2695-2699.
280



9. In their synthesis of the marine natural product haliclamide, Altmann and co-workers
performed the following conversion.
64
What is the expected product?



10. The subsequent steps in the synthesis of haliclamide are shown below. HATU is a different
sort of coupling reagent, and its structure is shown below as well. Using the mechanistic
principles that apply to coupling reagents, provide the structures of compounds A and B, and
provide a reasonable mechanism for the conversion of A to B.



11. To study resonance energy transfer, Cavazzini and co-workers prepared compound D, which
contains coumarin units that act as energy donors, as well as a bis(alkylaminostyryl) unit that
acts as an energy acceptor.
65
From problem 10, we now know that HOBt can act as an acyl

64
Pfeiffer, B.; Speck-Gisler, S.; Barandun, L.; Senft, U.; de Groot, C.; Lehmann, I.; Ganci, W.;
Gertsch, J.; Altmann, K.-H. Journal of Organic Chemistry, 2013, 78(6), 2553-2563.
65
Cavazzini, M.; Quici, S.; Orlandi, S.; Sissa, C.; Terenziani, F.; Painelli, A. Tetrahedron, 2013,
69(13), 2827-2833.
281

transfer catalyst, much like DMAP. Using this knowledge, provide structures for compounds A
D in their synthesis below.








282





Part V:
Polymers

283






Chapter 14:
Chain-growth polymers


284

Polymer categories

Youve likely encountered the concept of polymers in previous coursework, so youll recall that
polymers are large molecules (sometimes called macromolecules) that consist of repeating
subunits, known as monomers. There are two broad categories of polymers that derive from the
way in which those polymers are prepared. Chain-growth polymers are made from reactions in
which the polymer chain is extended by one monomer at a time. Two monomers link to form a
dimer. The dimer adds another monomer to form a trimer. The trimer adds another monomer to
form a tetramer. This process continues through a series of oligomers (i.e. species built from a
few monomer units) until the chain ultimately becomes a polymer.



The other major way in which polymers are prepared is known as a step-growth process. In step-
growth polymerizations small oligomer units can link to one another. The chain is not
necessarily grown by one monomer at a time. For instance, two dimers could combine to form a
tetramer, bypassing the trimer entirely. The small fragments that join can be of any size.



In this chapter, well discuss chain-growth polymerizations, and the next chapter will focus on
step-growth processes.

Radical polymerizations

Through our earlier discussion of radicals, we became quite familiar with the fact that radical
processes require an initiator, such as AIBN or peroxides. When such an initiator homolyzes, the
resultant radical can add to the bond of an alkene. In the example below, ethylene is the
specific monomer used. The initial radical addition yields a new carbon-centered radical, which
can subsequently add to another ethylene monomer. The process continues to add one ethylene
285

unit at a time, making this a chain-growth process, like all of the polymerizations that well
discuss in this chapter.



The termination of the radical can occur in different ways. It is possible that two radicals can
simply combine to terminate the polymerization.



Alternatively, a process known as disproportionation can take place. In disproportionation,
analogous species are simultaneously oxidized and reduced to yield two different products. In
this case, one radical can abstract a hydrogen from the carbon to another radical. One radical is
reduced through the abstraction of hydrogen. The other radical is oxidized to a bond as the
carbon-hydrogen bond homolyzes.



Ethylene is not, of course, the only monomer that can be used in radical polymerization. Other
monomers, like styrene, can polymerize via an analogous pathway. The initial radical addition
to the bond of styrene occurs so as to form the more stable, resonance-stabilized radical
intermediate. Each subsequent addition happens in the same fashion, which is term head-to-tail
since structurally different ends of the monomer units bond to one another.

286



Polymer stereochemistry

With substituents pendent to the carbon backbone of the polymer, the issue of stereochemistry
becomes important. The relative stereochemistry of the chiral centers in a polymer is known as
its tacticity. There are three principal stereochemical outcomes. It is conceivable that all of the
substituents could reside in the same orientation relative to the backbone. This scenario is named
isotactic. Syndiotactic polymers contains substituents on alternating sides of the backbone, and
atactic polymers contain a completely random orientation of their substituents. In radical
polymerizations, there is little basis for stereochemical control, so the polymer is typically
atactic.



The tacticity of a polymer will impact its properties. For instance, isotactic and syndiotactic
polymers have more regular structures. As a consequence, adjacent polymer strands are likely to
pack more effectively, leading to stronger materials with greater rigidity than those formed from
atactic polymers whose disorder leads in ineffective packing.

During the 1950s, Karl Ziegler and Giulio Natta developed the so-called Ziegler-Natta catalysts,
which allow for greater control in polymerization. There are many types of Ziegler-Natta
catalysts, and well consider just one illustrative example. The reaction of an organoaluminum
species with titanium tetrachloride yields a Ziegler-Natta catalyst through transfer of an alkyl
group from aluminum to titanium.

287



The titanium of the Ziegler-Natta catalyst then coordinates with a monomers bond. This is
followed by insertion of the monomer into the growing carbon chain.



Titanium then coordinates with another monomers bond. The insertion repeats itself, and the
chain continues to grow.



As noted above, there are many Ziegler-Natta catalysts, and the choice of catalyst will provide
good selectivity for one particular stereochemical outcome of the polymerization.

Branching

Radical polymerizations can lead to significant branching. However, the mechanism we outlined
above does not seem to allow for that eventuality, so lets revisit it. Imagine that a growing
polymer chain encounters a separate polymer chain. The hydrocarbon nature of the polymer
provides ample opportunity for hydrogen abstraction. This quenches the reactivity of the
growing chain, but at the same time, the interior of a different chain now becomes reactive again.

288



This new radical on the chains interior can add to a monomer, which installs a branch point in
the carbon backbone. As subsequent radical additions occur, the branch grows in length until it
eventually terminates.



This branching significantly impacts the polymers properties. For instance, polyethylene
formed via a radical polymerization tends to contain a fair bit of branching. Consequently,
adjacent polymer strands do not pack well, and the material is more pliable as a result. This is
known as low-density polyethylene (LDPE). It has the resin ID code (i.e. the recycling code) 4.
It is used is applications like plastic bags.

One the other hand, polyethylene made using a Ziegler-Natta catalyst will not contain branching.
The adjacent polymer chains pack much more tightly, and the result is high-density polyethylene
(HDPE). HDPE has the reside ID code 2, and it is used to make plastic bottles like milk jugs.


289

Cationic polymerization

Polymerization can occur through more than one mechanism, and the reactions are not limited to
radical intermediates. In fact, there are both cationic and anionic mechanisms possible. Lets
consider cationic polymerization first. For these polymerizations, a source of acid is needed.
While multiple possibilities exist, it is common to combine boron trifluoride and water. The
Lewis acid-base reaction that follows generates a good proton donor.



Ethylene would not be particularly receptive to cationic polymerization since it would form an
especially unstable primary carbocation intermediate. However, monomers bearing an electron-
donating substituent do readily undergo cationic polymerization. For instance, vinyl acetate can
be protonated to form a resonance-stabilized carbocation. This carbocation can undergo addition
of another vinyl acetate monomer, creating a cationic dimer. The polymerization continues until
the cation is finally quenched by donation of hydroxide from the Lewis acid-base complex.



Anionic polymerization

If the monomer bond bears an electron-withdrawing substituent, a reversal of polarity is
possible during the polymerization, and anionic intermediates are involved. Notice how a subtle
change, like the inversion of the ester, can completely change the groups effect on the bond.
Addition of a strong nucleophile, such as an organolithium reagent, to the -position of the ,-
unsaturated ester yields a resonance-stabilized anion (specifically, in this case, an enolate). The
enolate acts as a donor in Michael reaction with another ,-unsaturated ester monomer, which
acts as the Michael acceptor. Polymerization continues, but this reaction differs from radical and
cationic polymerization in a significant way.

290

O O
R Li +
O O
R
O O
R
C
O
2
t
B
u
C
O
2
t
B
u
R
C
O
2
t
B
u
C
O
2
t
B
u
C
O
2
t
B
u
R
O O
n


There is no proton or electrophile present in the medium that can quench the enolate. Such a
proton source or electrophile would have been incompatible with the organolithium initiator.
Consequently, this process is called a living polymerization. Even after all of the monomer
has been consumed, the enolate persists, and the addition of more monomer at any point will
revive the polymerization process. The reactivity is not quenched until, in a separate second
step, a proton source (like water) or an electrophile (like carbon dioxide) is added.



Copolymers

To this point, we have considered only those polymers formed from identical monomer subunits.
Such polymers are termed homopolymers. However, it is also possible to form polymers using
more than one type of monomer subunit. Such polymers are called copolymers. The illustration
below shows alternating monomer units of two types (blue and green), but the two types of
monomers need not necessarily repeat in such an orderly fashion. There could be a long stretch
of one monomer followed by a long stretch of the other monomer, which would be called a block
copolymer. Alternatively, there could be a random or statistical distribution of the two monomer
subunits.


291


Styrene-butadiene rubber is one example of a copolymer. It was particularly important during
World War II when supplies of natural rubber were cut off. The preparation of styrene-butadiene
rubber began in earnest as a synthetic replacement for natural rubber. The two subunits used in
this polymerization are styrene and 1,3-butadiene. The diagram below shows the addition of the
initiator to styrene followed immediately by the addition of the styrene radical to the terminus of
butadiene.



Throughout this polymerization there is much variation in what can transpire at each step. The
butadiene radical can add next to a styrene monomer, or it could just as easily have added to
another butadiene monomer.



Another element of variation is whether butadiene is incorporated into the backbone in its
entirety or whether only two of its carbons become part of the chain.



If the resonance form that places the radical on the interior of the butadiene fragment adds to the
next monomer, the backbone will bear a vinyl substituent.



Polymers with heteroatoms in the backbone

Thus far, every polymer weve formed has had a backbone consisting solely of carbon.
Polymers can certainly be formed with heteroatoms in the backbone. Well see many examples
in the next chapter. For the moment, well simply consider the preparation of polyethers through
the polymerization of epoxides. A nucleophile like hydroxide can attack ethylene oxide, thereby
severing one of the carbon-to-oxygen bonds. The oxyanion thus formed can attack another
ethylene oxide monomer to lengthen the chain. This process continues until an oxygen anion is
protonated by water.

292



This polymer is called polyethylene glycol (PEG) because of the structural similarity to ethylene
glycol.

HO
O
O
O
n
H
Polyethylene glygol
Ethylene glycol
HO
OH


While the termini of the polymer may bear hydroxyl groups, the majority of the polymer
contains ether functionalities. These affect the properties in significant ways. For instance, PEG
has water solubility due to its ability to hydrogen bond with water. The same cannot be said of
polymers like polyethylene, which have no affinity for water. Consequently, PEG finds usage in
some personal care products, like lotions and creams.

The epoxide used in the polymerization need not be symmetrical. If propylene oxide is used in
place of ethylene oxide, the reaction proceeds as expected for nucleophilic attacks on epoxides.
That is to say that the reactions are S
N
2-like, so the attack occurs at the less hindered side of the
epoxide.



Rubber

Rubber is a polymer of isoprene, which is a very important biological building block not only for
polymers but also for other compounds as well. The polymerization of isoprene leads to two
isomeric polymers. One contains the Z configuration of the olefins and is natural rubber, while
the other contains the E configuration and is known as gutta-percha. Gutta-percha is a harder
polymer, which has been used as an electrical insulator and in golf balls.

293



Natural rubbers desirable properties are somewhat diminished by its tendency to become hard
and brittle in winter and soft and sticky in summer. It was the inventor Charles Goodyear who
discovered the process of vulcanization, in which sulfur is used to create some disulfide bridges
or linkages between polymer chains. These disulfide linkages act as staples to improve the
elasticity of rubber. As the chains are moved away from each other and then released, they
return to their original shape due to the presence of these covalent staples.



Unfortunately, Goodyear was a better inventor than he was a businessman. He was plagued by
patent infringement, and even had others steal his ideas. In fact, Goodyear spent quite a bit of
time in debtors prison as a result of his business woes.
66


Chapter 14 Problems

1. In the Chapter 6 problems, we saw the work of Hu and co-workers, who reported the synthesis
of rod-coil brush polymers.
13
The brush polymers can be represented using the following
schematic.




66
For an interesting historical perspective, see Chapter 8: Isoprene in Napoleons Buttons: 17
Molecules That Changed History by P. Le Couteur and J. Burreson (Penguin: New York, 2003).
294

Recall that the straight portions of the schematic are formed by linked aromatic rings; whereas,
the squiggly portions of the diagram denote the ester-containing chains.



The monomer is shown below. Show how Bergman cyclization can lead to the formation of the
polymer described.



2. Draw the mechanism for radical polymerization of vinyl acetate (CH
2
=CHOCOCH
3
).

3. As it is shown below, this particular __________________ chain has __________________
stereochemistry.



4. (a) Show how a Ziegler-Natta catalyst can add propylene units to a growing polymer chain.
This mechanism will not involve curved arrow notation.


(b) As it is shown above, this growing polymer chain has ______________________
stereochemistry.

5. Draw a mechanism to explain the branching that can occur within polystyrene. Use the
growing polymer and adjacent polymer chain shown below to begin your mechanism. Pay
special attention to the stability of the radical intermediates.

295



6. Draw a mechanism for the following polymerization, which yields polyisobutylene.



Polyisobutylene (aka butyl rubber) has diverse uses, including forming the airtight core of
basketballs and forming the gum base of chewing gum.

7. Methyl 2-cyanoacrylate (shown below) is a monomer found in some super glue products.
Show how the polymerization of this substrate can occur in the presence of adventitious
moisture.



8. Acrylonitrile butadiene styrene (ABS) is a copolymer used in car bumpers and crash helmets.
The monomers are shown below. Draw a possible structure for ABS.



9. What monomer would give rise to the following polymer when treated with NaOH to initiate
polymerization?



10. Draw a mechanism for the following polymerization, which begins with the addition of a
trace of butyllithium and terminates with the introduction of carbon dioxide.
296


CN
1. trace n-BuLi
2. CO
2


The product is polyacrylonitrile, which has been used in fabricating sails. Copolymers of
polyacrylonitrile (marketed as Orlon by DuPont) have been used in clothing labeled as
acrylics. Additionally, polyacrylonitrile can be used as a source of carbon fiber for aircraft,
although this requires further treatment of the polymer.




297





Chapter 15:
Step-growth polymers

298

Step-growth polymers

In the previous chapter, we noted that step-growth polymers differ from chain-growth polymers
in that they can add units to the growing chain that are larger than a single monomer. This stems
from the fact that the monomers for step-growth processes possess two reactive termini, and
either terminus can undergo reaction to add additional length to the chain. Since these reactions
often occur with the loss of a small molecule, like water, alcohol, or HCl, these polymers are
sometimes also called condensation polymers.



There are a number of functional groups that can serve as the linchpins for condensation
polymers. In this chapter, well examine these possibilities.

Polyamides

Condensation polymers that contain amide linkages between the monomers are called
polyamides. A classic example is called nylon 6-6. Nylon 6-6 can be formed by reaction
between adipoyl chloride, a six-carbon diacid chloride, and 1,6-diaminohexane, a six-carbon
diamine. The reaction itself is nothing more than simple nucleophilic acyl substitution. Once
the first amide bond is formed, the dimer still bears both an acid chloride and an amino terminus.
Either of these termini can react next with the appropriate partner to lengthen the chain. The
final product is called a nylon (a synonym for polyamide). It is nylon 6-6 because it consists of
two building blocks, both of which contain six carbon atoms.



299

Nylon 6-6 can also be made industrially through the condensation of adipic acid and 1,6-
diaminohexane. Initially, an acid-base reaction occurs to form a salt; however, heat can be used
to drive off water and push the equilibrium toward the covalent polymer.



In recent years, greater attention has been devoted to the sustainability of chemical processes.
Traditionally, adipic acid and 1,6-diaminohexane have been synthesized from benzene isolated
from petroleum. However, there is now technology that enables adipic acid to be prepared from
a different six-carbon substrate: glucose. Glucose could, in theory, be derived from the cellulose
in biomass. An efficient conversion of cellulose to glucose would ensure a renewable source of
many important chemical building blocks.



As with each category of condensation polymer well discuss, there are many examples. A
second illustrative nylon is nylon 6, which consists of only a single building block. -
Caprolactam is a seven-membered lactam (i.e. cyclic amide). The lactam ring can be opened by
a nucleophile via nucleophilic acyl substitution.



The amino acid thus formed can attack another molecule of -caprolactam using its amino
terminus. Again, this reaction follows the nucleophilic acyl substitution paradigm. Continued
incorporation of -caprolactam residues ultimately produces nylon 6.

300


Nylon 6 presents a unique opportunity for chemical recycling. Traditionally, we think of
recycling materials through the melting and remolding of plastics. However, certain polymers
can be degraded back into their original monomers. These monomers can then be used in a new
polymerization reaction. Such is the case with nylon 6. A nucleophile like ammonia can sever
one of the amide bonds of the backbone via nucleophilic acyl substitution. This unleashes a free
amino terminus, which can intramolecularly attack the nearest carbonyl to snip a -caprolactam
unit from the diminishing chain. This process can continue until the polymer reverts entirely to
monomer.



Polyesters

Polyesters contain ester linkages in the backbone, and a common example of this class is
poly(ethylene terephthalate), or PET. PET is widely distributed in products such as water bottles
and fleece. It has the resin ID code 1. PET can be formed from the condensation of terephthalic
acid and ethylene glycol via Fischer esterification.

301



Poly(ethylene terephthalate) is another polymer that can be chemically recycled. When treated
with methanol and catalytic acid, a transesterification takes place. This proceeds through
effectively the same mechanism as Fischer esterification, with the difference being that the
nucleophile is an alcohol in transesterification and water in Fisher esterification. The
transesterification cleaves the ester linkages of the backbone, releasing dimethyl terephthalate
and ethylene glycol monomers. These can be used in polymerization to re-form PET.



Polyurethanes

The name urethane is another name for the carbamate functional group. We encountered
carbamates during the discussion of amino group protection. When the carbamates are
incorporated into the backbone of polymers, these macromolecules are called polyurethanes.
Polyurethanes can be formed from isocyanates and alcohols. An alcohol, such as ethylene
glycol, will add to the central carbon of the isocyanate functional group. The carbon-to-oxygen
-bonding electrons are displaced onto the oxygen. They then collapse to re-form the carbonyl,
302

and the carbon-to-nitrogen -bonding electrons are used to acquire a proton. This process will
repeat with the second hydroxyl group of ethylene glycol to lengthen the chain.



Polycarbonates

Polycarbonates are formed by the reaction between phosgene (COCl
2
) and hydroxyl-group
containing monomers. One such polycarbonate building block is bisphenol A, also known as
BPA, which has received a great deal of attention due to a number of possible health risks
associated with exposure to this compound in its free form. The polycarbonate is formed as
phosgene undergoes two successive nucleophilic acyl substitution reactions. In this case, the
polymeric product is called Lexan. Lexan is used in many applications, including bulletproof
glass.



Epoxy resins

Epoxy resin (sometimes simply called epoxy) utilizes a clever twist on the themes weve
encountered so far. There are two liquids which are combined to form the final hardened epoxy
resin. One is a prepolymer fluid, and the other is a hardener fluid. The prepolymer fluid consists
of relatively small polymer chains bearing reactive termini. The hardener contains a molecule
with multiple pendent nucleophiles. These nucleophiles form covalent bonds with the reactive
303

termini of the prepolymer chains, thereby creating a cross-linked polymeric network in the final
resin.

The prepolymer fluid is form by a reaction like that of the conjugate base of bisphenol A with
epichlorohydrin. The phenoxide of BPA attacks the less hindered side of the epoxide of
epichlorohydrin, cleaving open the ring in the process. The resultant alkoxide attacks the
adjacent center bearing a chloride leaving group, and this introduces a new epoxide functionality.



This newly formed epoxide can be attacked by another BPA anion. The reaction is performed
with an excess of epichlorohydrin, which results in prepolymer chains that terminate with
epoxides.



These prepolymer chains are then treated with the hardener, a substance such as
diethylenetriamine (DETA). Each of the nucleophilic amino groups of DETA can attack the
304

terminal epoxide of a prepolymer chain. As this occurs with many DETA molecules and many,
many prepolymer chains, a long cross-linked covalent network is formed and the epoxy resin has
been cured.


Bakelite

Bakelite was first prepared by Leo Baekeland in the early 1900s. It was one of the first truly
synthetic plastics, and it was used in applications as diverse as billiard balls, telephone casings,
and electric insulators. Bakelite is an example of a thermosetting polymer. While thermoplastics
can be melted and remolded, thermosetting polymers cannot be because they contain extensive
cross-linked covalent networks, as we saw with epoxy resins above.

Bakelite is formed by the condensation of phenol and formaldehyde through electrophilic
aromatic substitution (EAS). In acidic media, formaldehyde can be protonated to enhance its
electrophilicity. The nucleophilic ortho (or para) position of phenol then attacks the
electrophilic carbonyl carbon of formaldehyde. The complex that results will lose a proton to
restore aromaticity, as would be expected with any other EAS reaction. This yields a benzylic
alcohol, which can be protonated to afford a good benzylic leaving group. The dissociation of
water produces a resonance-stabilized carbocation, which serves as the electrophile in a second
EAS reaction. In this way, two molecules of phenol are linked by one carbon that originated
from formaldehyde. Loss of a proton from this second complex yields a bisphenol. The
reaction continues with many formaldehyde and phenol molecules. Since each phenol can
undergo EAS at its ortho and para positions, the polymer that results is an intricate cross-linked
network.

305


Dendrimers

Dendrimers are a different kind of polymer in which many chains and branches radiate out from
a common, central core. They are sometimes also called cascade molecules. While dendrimers
are usually symmetrical around their core, dendrons bear highly branched chains emanating from
a focal point. Each time a new branch point is reached, the dendrimer or dendron adds another
generation. Dendrimers have been widely explored since their initial preparation in the late
1970s, and they are finding application in nanotechnology, catalysis, sensors, and drug delivery,
among other areas.

G0
G1
G2
G3
Dendrimer
(G = generation)
Dendron
G1
G2
G3


306

While there are many, many example of dendrimers and dendrons that we could consider, a
relatively straightforward example from the primary literature follows.
67
A diamino acid
building block can be acylated with trifluoroacetic anhydride (TFAA) via nucleophilic acyl
substitution. This leaves a single reactive site: the carboxylic acid. That acid can be converted
to the corresponding acid chloride by treatment with thionyl chloride (SOCl
2
).



Two equivalents of this acid chloride are then combined with one equivalent of the original
diamino acid building block. Amide linkages are formed at both accessible amino groups,
yielding a dendron.



67
This example was reported in Endo, K; Higashihara, T.; Ueda, M. Synthesis of a novel water-
soluble polyamide dendrimer based on a facile convergent method European Polymer Journal,
2009, 45(7), 1994-2001.
307


The iterative nature of the dendron synthesis is illustrated by the following steps in which the
accessible carboxylic acid is again converted to its acid chloride. Reaction of two equivalents of
this acid chloride with the original diamino acid incorporates a third generation of branching to
the dendron.

CO
2
H
O
O
N
N
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
SOCl
2
CO
2
H
O
O
H
2
N
H
2
N
(stoichiometric ratio of 2 : 1)
CO
2
H
O
O
H
N
N
H
O
O
N
N
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
O
O
O
N
N
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
O
O
N
N
H
H
F
3
C
O
F
3
C
O
O
H
O


308

This third-generation dendron is then activated as its acid chloride and linked to the core of the
incipient dendrimer. The core contains four reactive amino groups, so four equivalents of the
activated dendron are attached to it.



The synthesis is completed with the use of hydrazine to cleave the trifluoroacetyl protecting
groups from the peripheral amines. The ultimate product of this global deprotection is a
dendrimer with three generations of branching.

309

Chapter 15 Problems

1. What monomers would be required to synthesize the following condensation polymer?



2. Kevlar is a polymer well known for its use in bulletproof vests. A short strand of Kevlar is
shown below. Show the monomers needed for its synthesis, and propose reasons for the rigidity
of this polymer.



3. Quiana has been used to make wrinkle-resistant fabrics. It is formed from the reaction shown
below. Give the structure of Quiana.



4. Kodel is a polyester fiber with the following structure. It is commonly used in films for
packaging. What monomers would be required for the synthesis of this condensation polymer?



5. Compare polystyrene to Kodel (in the question above). Which polymer would you expect to
be more stable? Briefly explain your rationale.

6. In the Chapter 13 problems, we learned about the role of the pentafluorophenyl ester as an
activated carboxylate for the synthesis of amides. Engler, Hendrick, and co-workers reported
Accessing New Materials through Polymerization and Modification of a Polycarbonate with a
310

Pendant Activated Ester in the journal Macromolecules.
68
In this publication, ring opening
polymerization (ROP) was used to access aliphatic polycarbonates. A trace of a small alcohol,
such as benzyl alcohol, was used to incite ROP.

The activated polymer could then be converted to a polymer bearing different functionalities by
exploiting the reactivity of the Pfp ester. Show the structures of the activated polymer and
polymer below, and provide mechanisms to explain their formation.



7. Give the structures of the prepolymer and the epoxy resin that are formed from the following
sequence?



8. Melamine is an interesting compound that has found its way into a number of applications,
such as magic erasers. It has also been at the center of controversy. For instance, in 2007
there was a recall of pet foods to which melamine had been added to artificially elevate the
apparent protein content of the food.

Melamine has also been used as a monomer in the synthesis of Melmac, which is used for dishes
and countertops. Melamine condenses with formaldehyde to give a highly cross-linked network
much like we saw with Bakelite. However, this does not occur through EAS reaction as it did
with Bakelite. Instead, the process is akin to reductive amination. Remember that, in reductive
amination, an amine and an aldehyde or ketone condense to give an imine. This imine then
undergoes the addition of hydride as a nucleophile.

In the synthesis of Melmac, an amine from melamine condenses with formaldehyde to give an
iminium ion that then undergoes the addition of a different nucleophile (an amino group from a
new melamine molecule). Using this description as a guide, show a small segment of Melmac in
which you have linked at least four melamine monomers and at least four formaldehyde
monomers. Make sure that this segment also shows some of the cross-linking of chains that is so
important to the properties of thermosetting polymers.


68
Engler, A. C.; Chan, J. M. W.; Coady, D. J.; OBrien, J. M.; Sardon, H.; Nelson, A.; Sanders,
D. P.; Yang, Y. Y.; Hendrick, J. L. Macromolecules, 2013, 46(4), 1283-1290.
311



9. TMXDI (m-tetramethylxylylene diisocyanate) is a common isocyanate used in the preparation
of polyurethanes.
69
Show the polyurethane that would result from the following reaction.



10. Investigators at the University of Maryland reported the synthesis of fluorocarbon dendrons
as part of an effort toward the preparation of defect-free macromolecules.
70
Show the structures
of compounds A F. Note that compound F is a model G2 dendron.



Show the structure of compound G and compound H, a model G3 dendron.







69
Delebecq, E.; Pascault, J.-P.; Boutevin, B.; Ganachaud, F. Chemical Reviews, 2013, 113(1),
80-118.
70
Yue, X.; Taraban, M. B.; Hyland, L. L.; Yu, Y. B. Journal of Organic Chemistry, 2012,
77(20), 8879-8887.
312





Solutions to End-of-
Chapter Problems

313

Solutions to Chapter 1 Problems

1. If the number of carbons in a molecule is known, we can predict the relative abundance of the
M+1 peak for a given intensity of the molecular ion peak (M). This equation is based on the fact
that 1.1% of naturally occurring carbon will be
13
C, which will contribute an extra mass unit to
the molecular mass.

(Rcloti:c obunJoncc o H + 1) = n (u.u11) (Rcloti:c obunJoncc o H)

This equation can also be rearranged to allow determination of the number of carbons if we
know the relative intensities of the M and M+1 peaks.

n =
(Rcloti:c obunJoncc o H + 1)
(u.u11) (Rcloti:c obunJoncc o H)


In this case, the number of carbon atoms in the structure is 10.

n =
(2.uS%)
(u.u11) (18.6%)
1u

The structure of propyl guaiacol is shown below, although we cannot determine this from the
information given.



2. First, it is important to determine how many carbons the molecule is likely to contain.
Dividing the molecular mass by the mass of carbon shows that there cannot be more than 13
carbons in the structure.

Hoximum # Cs =
166
12
= 1S.8

A molecular formula of C
13
H
10
is therefore an acceptable possibility. This is a highly
unsaturated molecule with 9 degrees of unsaturation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns prcscnt
2
; (n = # corbons)

314

cgrccs o unsoturotion =
|2(1S) + 2] 1u
2
= 9

One possible structure with this formula follows.

C
13
H
10


If you wanted to draw possible structures with fewer degrees of unsaturation, one of the carbon
atoms could be traded for 12 hydrogen atoms.

C
13
E
10
C + 12 E = C
12
E
22


Now, there molecule contains only 2 degrees of unsaturation, so there are many possible
structures, including the following.



To derive possible formulas containing oxygen, a CH
4
unit (mass of 16) is replaced by oxygen
(also mass of 16).

C
13
E
10
CE
4
+ 0 = C
12
E
6
0

or

C
12
E
22
CE
4
+ 0 = C
11
E
18
0

The first formula (C
12
H
6
O) has so many degrees of unsaturation that it becomes challenging to
draw reasonable structures for it, so you might prefer the second choice, which has only 3
degrees of unsaturation and could encompass structures like the following.



315

3. There has been no discussion in the text of an isotope that could lead to an M+4 peak.
However, we do know that the presence of chlorine or bromine in a molecule results in M+2
peaks. Notice that the choices contain differing numbers of chlorine atoms. Choice (a) has 1
chlorine. Choice (b) contains 2 chlorines, and choice (c) contains no chlorine atoms. Choice (c)
would therefore have no isotope pattern. Choice (a) would have an isotope pattern, but it would
contain significant M and M+2 peaks only. Choice (b), however, could show the observed
isotope pattern.



Its worth noting that the isotope pattern for bosutinib will be slightly complicated by the
presence of nitrogen, which has an isotope (
15
N) that contributes an additional mass unit.
However, the abundance of
15
N (0.4%) is quite low relative to that of
37
Cl (24.2%).

4. For 3,3-dimethyloctane, cleavage around the quaternary center is expected to account for the
most prominent fragments because these cleavage events afford the most stable carbocations
(and radicals). Notably, M-15, M-29, and M-71 peaks are expected to be significant since they
all result from tertiary carbocation fragments.

316


With 2,3-dimethyloctane, cleavage about the bond between the two tertiary carbons is expected
to be among the most important fragmentation pathways because it will yield the most highly
substituted carbocation and radical. As a result, and M-43 peak is expected to be significant, as
well as a peak at m/z 43.



5. To answer this question, it would be useful to examine the heterolytic and -cleavage for each
analyte upon ionization. 1-Chlorobutane yields the fewest and lowest mass fragments.



317

Therefore, it pairs most logically with the first mass spectrum. Heterolytic cleavage appears to
be the predominant fragmentation pathway, suggesting that the chloronium ion derived from -
cleavage is not well stabilized. Additional cleavage of C-C bonds is observable in the m/z 43
peak corresponding to the loss of a propyl cation and its subsequent formation of the allylic
carbocation (m/z 41) through the loss of two hydrogen atoms.


2-Chloro-2-methylpropanes fragmentation can explain the formation of a m/z 57 peak (the tert-
butyl carbocation), as well as a m/z 77 or 79 peak corresponding to the chloronium ion
containing either
35
Cl or
37
Cl.

Cl
high E beam
of e
-
Cl heterolytic
cleavage
+ Cl
m/z 57
Cl
+
-cleavage
Cl
H
3
C
m/z 77
or 79


Therefore, it pairs with the second spectrum.

318


Finally, 2-chlorobutane exhibits not only heterolytic cleavage but also two -cleavage
fragmentations.



These explain the signals apparent in the third mass spectrum. Notice that the prominent m/z 57
signal appears to fragment further, undergoing loss of two additional hydrogen atoms to produce
a signal at m/z 55.

319


6. Since this ether is unsymmetrical, there are multiple heterolytic and homolytic fragmentation
pathways. The two heterolytic cleavage events predict a peak at m/z 43. Further fragmentation
of the propyl carbocation, through the sequential loss of two hydrogen atoms, could also give
rise to signals at m/z 42 and 41. The -cleavage pathways predict peaks at m/z 73 and 83.



7. The M-18 peak results from dehydration of the molecular ion. The oxygen radical of the
molecular ion initially abstracts the -hydrogen, which is five atoms away from this reactive
center. This creates a new, more stable carbon-centered radical. Additionally, the molecule now
bears water as a good leaving group. Its heterolytic dissociation from the carbon backbone
results in a radical cation that is 18 mass units lighter than the molecular ion.

320

HO
high E beam
of e
-
HO H O
H H
M
O
H H
+
M-18


8. Ionization of the ketone through loss of an electron from an unshared pair forms the molecular
ion with a m/z ratio of 224.



The major fragmentation pathways for ketones are -cleavage and McLafferty rearrangement.
When the ketone is unsymmetrical, two -cleavage fragmentations are possible. The -cleavage
process entails donation of oxygens unpaired electron to the adjacent carbon to form one half of
a bond. As the carbonyl carbon reciprocates to complete the bond, it donates an electron
from one of its bonds, thereby cleaving that bond. In this case, the acylium ions that result
yield peaks at m/z 147 and 105.





McLafferty rearrangement occurs through abstraction of a hydrogen atom from the -carbon.
This hydrogen is six atoms removed from the reactive center. The abstraction of the -hydrogen
results in a cascade that forms a bond between the - and -carbons. In the process, the ,-
bond is severed, releasing styrene. The resultant radical cation appears at m/z 120.

321



9. Whenever a problem provides a molecular formula, it is wise to begin with a calculation of
degrees of unsaturation. In this instance, the molecule contains a single degree of unsaturation.

cgrccs o unsoturotion =
|2(7) + 2] 14
2
= 1

This degree of unsaturation could be a ring or a bond. When the molecule is small and there
are multiple fragments, a ring seems less probable. The reason is that possible structures
including rings of a reasonable size will yield few fragments, since breaking a single ring bond
does not cleave the molecule into smaller pieces. While we certainly cannot rule out a ring at
this stage, it might be most productive to first consider the possibility of a bond.

The molecule could contain an ether or alcohol in addition to an alkene. Alternatively, it could
contain a ketone. While two functionalities are certainly possible, a logical first consideration
would be the ketone, which combines the oxygen and bond into a single functional group.

We know that ketones engage in -cleavage and McLafferty rearrangement. When a ketone is
unsymmetrical, the mass spectrum typically reveals two -cleavage fragments and up to two
McLafferty rearrangement fragments (depending on whether the R groups are large enough to
contain -hydrogens). The mass spectrum in this case reveals only two fragments, so well need
to consider an explanation for that observation.


322



The mass differences between the molecular ion peak and the fragments reveal the loss of ethyl
radical and a butyl radical (or one of its isomers). These fragments likely result from -cleavage,
so we are left with four possible structures in which the carbonyl carbon is bonded to an ethyl
group as well as a butyl, isobutyl, sec-butyl, or tert-butyl group.



Remember that McLafferty rearrangement results in cleavage of the ,-bond and formation of a
bond between the - and -positions. This allows us to eliminate the first two ketones bearing
butyl and isobutyl groups, both of which would release propylene during McLafferty
rearrangement to yield a peak at M-42, which we do not observe in the spectrum.

The ketone bearing a sec-butyl group would release ethylene during McLafferty rearrangement.
This would result in an M-28 peak, which is also not observed in the mass spectrum. So, this
ketone can also be eliminated, leaving 2,2-dimethylpentanone as the only reasonable choice.



2,2-Dimethylpentanone does not possess a -carbon, so McLafferty rearrangement is not
possible. The only major fragmentations result from the two -cleavage events, which release
323

ethyl and tert-butyl radical. The release of the more highly substituted tert-butyl radical would
certainly be more favorable, explaining the pronounced intensity of this base peak relative to the
other signals.

10. Since this problem also provides us with a molecular formula, it is helpful to begin with a
degrees of unsaturation calculation. Remember that a halogen, being monovalent, essentially
takes the place of a hydrogen atom and must therefore be counted as a hydrogen equivalent in
the calculation.

cgrccs o unsoturotion =
|2n + 2] EyJrogcns (or E cqui:olcnts)prcscnt
2


In this case, there is one degree of unsaturation present, so the structure contains a single ring or
bond.

cgrccs o unsoturotion =
|2(6) + 2] 12
2
= 1

An initial analysis of the mass spectrum provides some important clues. The molecular ion peak
reveals the expected isotope pattern for a chlorine-containing compound. There are M and M+2
peaks in an intensity ratio of approximately 3:1.

The signal at m/z 83 is 35 amu lighter than the molecular ion peak at m/z 118. This suggests the
loss of chlorine, an assumption which is supported by the absence of an isotope pattern for this
signal. Therefore, this fragment likely results from simple heterolytic dissociation of the halogen
radical.



The fragments at m/z 49 and 51 appear to still contain the chlorine, since they are separated by 2
mass units and have the expected 3:1 abundance ratio. If chlorine is part of this fragment, then
324

only 14 mass units remain for the other atoms in the fragment. Since 14 amu corresponds to a
methylene (CH
2
) group, this fragment is likely the chloronium ion resulting from -cleavage.



All of the other atoms in the molecule must be bound to the methylene carbon in a single R
group that was broken off during -cleavage. There are 5 remaining carbon atoms and 9
hydrogens, as well as a single degree of unsaturation. Since there are no other major
fragmentations (as we would expect if an alkene were present in the structure), it seems logical
for the degree of unsaturation to be a ring. While we could draw more than one feasible ring
size, five-membered rings are the least strained of our options, so it seems likely that a
cyclopentyl radical was released during -cleavage.



Putting these fragments back together, a structure consistent with the data is
(chloromethyl)cyclopentane.



Solutions to Chapter 2 Problems

1. Hirsutellone C contains a paracyclophane (i.e. a benzene ring bearing a tether between its para
positions). The tether is a relatively small, nine-atom chain, making this a [9]paracyclophane.



Given the constraints of this small tether, it is unlikely that the benzene ring can freely rotate. If
free rotation of the benzene ring is prevented, then carbons (a) and (b), which would have
ordinarily been equivalent, are actually held in separate, distinct chemical environments. As a
result, they will each give rise to their own
13
C NMR signal. Similarly, carbons (c) and (d) are
rendered unique by this restricted rotation and each cause their own signals as well.

325



On the other hand, 4-methylanisole experiences no such conformation constraint. As a result,
carbons (a) and (b) give rise to only one signal. Similarly, carbons (c) and (d) are also
chemically equivalent and cause just one signal. The remaining two aromatic
13
C signals result
from the carbons bearing the methoxy and methyl groups.



2. One pronounced difference between the structures is the presence of a vinyl proton ( 4.5
6.5) in the first structure but not the second.



Additionally, the two compounds have differing numbers of methylene, methine, and quaternary
carbons. A DEPT or APT spectrum would readily reveal these distinctions.



3. The isomer of MDPV bears only one neighbor to H
a
. This would result in H
a
being split into a
doublet.



326

On the other hand, MDPV itself contains two hydrogens neighboring H
a
. Since the carbon
bearing H
a
is a chiral center, these two neighboring hydrogens (H
c
and H
d
) will each split H
a
with
a unique coupling constant. This will result in the signal for H
a
appearing as a doublet of
doublets if the J values are reasonably different in magnitude.



If the J values are similar in magnitude, then the H
a
signal may be an apparent triplet.



4. It is important to notice that H
a
of lennoxamine originates from a stereocenter. Therefore, we
cant merely say that it has two neighbors, resulting in a triplet signal. Instead, we have to
acknowledge that H
b
and H
c
are diastereotopic and interact with H
a
via unique coupling
constants. This should lead to a doublet of doublets.



However, if the J values are not significantly different, then an apparent triplet may result.

327



5. The proton indicated in the question has been labeled as H
a
below. It has three neighbors, but
we wont expect to see a quartet here because all three neighbors are unique. H
b
and H
c
are
diastereotopic. Their adjacency to a chiral center places them in different chemical
environments, so they will interact differently with H
a
. Splitting of H
a
by H
b
and then by H
c

results in a doublet of doublets. However, we must still consider the interaction with the third
neighbor, H
d
. One more round of splitting results in a quartet of doublets. This is the
multiplicity that we would expect for H
a
.



6. Since the indicated protons stem from the same carbon and there are no stereogenic carbons in
Fosamax, these hydrogens are chemically equivalent. Furthermore, there is no rigidity or other
conformational constraint that prevents free rotation, so these hydrogens will interact identically
with all other chemically equivalent protons in the molecule, making them magnetically
equivalent as well.

7. The protons of 1,2-diacetylbenzene that are labeled H
a
are in identical chemical environments
(i.e. they are both ortho to the acetyl groups) and are therefore chemically equivalent. However,
they are magnetically inequivalent. This is due to the fact that H
a
and H
a
are located at different
distances relative to H
b
. H
a
splits H
b
with an ortho coupling constant of about 6 9 Hz, while
H
a
splits H
b
with a meta coupling constant of about 1 3 Hz.

328

O
O
H
a
H
a
'
H
b
J
Ha-Hb
J
Ha'-Hb


8. The stereochemistry of this molecule plays an important role in the conclusions that we draw.
Both spin systems exist in proximity to stereocenters. Therefore, the methylene hydrogens will
not formally be identical. They are diastereotopic and each give rise to their own unique signals.
However, we would not expect the signals of diastereotopic methylene hydrogens to be vastly
different in chemical shift under ordinary circumstances, so they have all been assigned adjacent
letters in alphabetical order.



For the spin system between the carbonyl and the amine, one methylene grouping is AB and the
other has been designated XY since we would expect the chemical shifts of hydrogens to a
ketone to be fairly different from those of hydrogens next to an amine. It is worth noting the
possibility that A and B may be far enough removed from the chirality to cause them to
accidentally overlap in an actual NMR spectrum.

For the spin system near the alcohol (really a hemiaminal in this case), the methylene hydrogens
have been assigned AB and the methine hydrogen has been designated as X since we would
expect it to have a fairly different chemical shift as a result of its proximity to the functional
group.

9. The methyl hydrogens are homotopic. Since there is no reason to expect the rotation of the
methyl group to be restricted, these hydrogens rapidly interconvert positions and are completely
identical to one another.

329



The methylene protons are diastereotopic. The amine-bearing carbon is a chirality center (even
if that chirality is not specified). Therefore, sequential isotopic substitution of the methylene
hydrogens will result in diastereomers, leading us to the conclusion that these hydrogens exist in
different chemical environments and are diastereotopic as a result.

10. Whenever the isomers in question present a clear difference in spatial orientation, NOESY
spectroscopy should come to mind as a way to differentiate between them based on through
space interactions of protons that are separated by several bonds.

O
O
HN
O
MeO
MeO
OMe
O
O
NH
O
MeO
OMe
MeO
Z isomer E isomer
H
H
expected
nOe
expected nOe
between
methoxy and
aromatic
hydrogens


Here the Z isomer places the vinyl and aromatic hydrogens close in space, so we would expect a
nOe correlation between them. The E isomer, on the other hand, places the methoxy group of
the dioxolane system close the aromatic hydrogens of the isoindolinone system, and we would
expect a corresponding nOe.

Solutions to Chapter 3 Problems

1. The carbonyl resonance for a ketone is expected to appear at about 1715 cm
-1
. The six-
membered ring has no impact on the expected resonance. In order for the resonance to fall
below 1700 cm
-1
as it does in this spectrum, the ketone must be conjugated (i.e. ,-unsaturated).



2. Both esters are conjugated to the ring; however, the ester resulting from route B is also in
conjugation with the ortho methoxy group. This additional donation of electron density into the
330

carbonyl should increase the single-bond character and therefore lower the resonance even
further.



3. Ethyl formate is expected to exhibit a typical ester carbonyl resonance in the vicinity of 1740
cm
-1
.



On the other hand, urethane demonstrates two competing influences on the carbonyl. The
electron-releasing nature of the nitrogen via resonance is counterbalanced by the electron-
withdrawing behavior of the carboxyl oxygen. These two factors very nearly cancel each other,
resulting in a carbonyl resonance around 1710 cm
-1
.



4. There are two competing factors in this problem as well. The inclusion of a carbonyl in a five-
membered ring is expected to raise the frequency to approximately 1750 cm
-1
. Conversely, the
electronic contribution of nitrogen to an amide carbonyl is expected to lower the frequency to
about 1650 cm
-1
.



331

Both factors are relevant in 2-pyrrolidinone, which is a five-membered lactam (i.e. cyclic amide).
In this case, their opposing effects more or less cancel, resulting in the observed resonance at
nearly 1700 cm
-1
.

5. Typically, we expect amide carbonyl resonances to appear at approximately 1650 cm
-1
. This,
however, is dependent upon the electron-releasing nature of the nitrogen through resonance.



In this case, that conjugation is not possible. The perspective in which this unusual lactam is
drawn may not be the easiest one to use to visualize this. We can rotate azaadamantane-2-one
90 toward the viewer to arrive at an orientation from which the justification will be more readily
apparent.



In highly constrained systems, it is useful to consider the positioning of the orbitals and any
impact that may have on their behavior. This kind of relationship between geometry and
electronics is sometimes termed the stereoelectronic effect. In order for resonance to occur
between the nitrogen lone pair and the carbonyl, the nitrogen must be able to place its lone pair
of electrons into a p orbital that is parallel to the system of the carbonyl (as shown above for a
generic amide).



The lone pair in azaadamantane-2-one is constrained in a position nearly 90 relative to the
carbonyl system. In other words, the lone pair and the system are essentially orthogonal and
are therefore totally unconjugated. For this reason, azaadamantane-2-one has been used as a
reference compound with 0% amidity.
71



71
Musci, Z.; Tsai, A.; Szori, M.; Chass, G. A.; Viskolcz, B.; Csizmadia, I. G. Journal of
Physical Chemistry A, 2007, 111(50), 13245-13254.
332



Consequently, we would expect the carbonyl resonance to appear closer to the normal ketone
frequency of 1715 cm
-1
. It is even possible that the frequency may be slightly elevated by the
carbonyl carbon to nitrogen dipole, which withdraws some electron density from the carbonyl.



Solutions to Chapter 4 Problems

1. The nature of this problem makes it a bit easier to tackle since we need not start from scratch
and try to build up a structure de novo. Instead, we are given two choices to decide between. A
straightforward strategy is to predict the expected integrations and multiplicities and see which
structure then matches the given spectrum more closely. The assignments listed below use the
abbreviations s (singlet), d (doublet), t (triplet), q (quartet), br s (broad singlet), dd (doublet of
doublets), and m (multiplet).



Even a cursory glance at the proton NMR shows that the first -hydroxyketone matches the
spectrum more closely. Lets take a closer look to make sure that we can explain everything.
The ethyl group is an easy place to begin. As expected, this forms an A
2
M
3
spin system,
meaning that a triplet and quartet are observed with integrations of 3H and 2H, respectively.
Formally, the methylene hydrogens are diastereotopic due to the presence of a chiral center in the
molecule; however, given their distance from the chiral center, this difference may well not
manifest itself in distinct signals.


333


When considering the methylene-methine fragment, the chirality center is a more significant
consideration. It renders the two adjacent methylene protons diastereotopic, and they can be
expected to cause separate signals. We do indeed see two doublets of doublets near 3 ppm. The
methine hydrogen (the M of the ABM spin system) should also formally be a doublet of
doublets. However, the coupling constants for interaction with the diastereotopic methylene
hydrogens might not be all that different in magnitude. Hence, it is not surprising that this signal
(near 4.75 ppm) is an apparent triplet.



Finally, we also observe a broad singlet for the hydroxyl proton near 3.6 ppm and a multiplet for
5H near 7.4 ppm accounting for the aromatic hydrogens.

2. Since we were provided with a molecular formula, it is wise to begin our analysis with a
degrees of unsaturation calculation.

cgrccs o unsoturotion =
|2(12) + 2] 14
2
= 6

When large numbers of degrees of unsaturation are present, it is helpful to recall that a benzene
ring will account for four degrees of unsaturation (three bonds and one ring). We might
therefore suspect the presence of a benzene ring in this unknown, and a quick glance at the
proton NMR confirms this due to the presence of signals in the vicinity of 7 ppm.

The IR is also relatively straightforward to interpret, so we should examine any clues that it may
provide early in the structure-solving process. There appear to be two carbonyls, one of which is
likely an ester and the other of which is probably an aldehyde or ketone. There may be other
explanations for these signals, but we'll consider the simplest possibilities first.

334



We can narrow down the choices even further by a quick glance at the NMR. Aldehydes show a
distinctive proton NMR signal around 9 10 ppm, which is absent, suggesting that the carbonyls
in cuelure are likely an ester and a ketone.



With these two fragments in hand, we can turn our attention to the proton NMR. This spectrum
shows two methyl groups with no neighbors.



We can also ascertain that the two methylene signals (i.e. those that integrate for 2H) must be
adjacent to one another, since both peaks appear as triplets signifying two neighbors. Youll also
notice some roofing (or second-order splitting) in the shapes of their signals, which further
supports our assertion that they couple to one another. Furthermore, these methylene groups
have no additional neighbors, which would have complicated their splitting pattern.



Finally, the aromatic hydrogen splitting pattern suggests a para-disubstituted benzene ring.


335


This is the only way to account for an aromatic ring bearing four hydrogens that appear as a pair
of doublets.

As we start to assemble the fragments, it is important to keep in mind that the carbonyls cannot
be conjugated to the ring, as this would alter their IR signals. Bonding either of the monovalent
methyl groups to the ring will make it difficult to fit all of the pieces together. This leaves us
with two divalent groups to bond to the ring: the ester (through the carboxyl oxygen) and the two
methylenes.



There is then only one way to place the remaining fragments into a final structure if we bear in
mind that the methylenes can have no additional neighbors.



This structure is consistent with the carbon-13 NMR. There are four carbon atoms appearing at
< 100 ppm. These correspond with the two methyls and two methylenes. There are four types of
aromatic carbons, which is consistent with the symmetry of the ring. And, there are two
carbonyls: the ester appearing around 170 ppm and the ketone at just under 210 ppm.

3. The degrees of unsaturation calculation is always a good place to begin these problems, and
here we see that the structure will have two bonds, two rings, or one of each.

336

cgrccs o unsoturotion =
|2(S) + 2] 8
2
= 2

The IR reveals sp
3
carbon-hydrogen stretching, as well as what is probably an ester (which will
account for one degree of unsaturation).

The
1
H NMR reveals a methyl group with no neighbors and a series of five inequivalent
hydrogen atoms. Given that two carbons are already accounted for (one as the ester carbonyl
and one as the methyl group). There are only three carbons remaining to house all of these
hydrogens. Consequently, there are probably two sets of diastereotopic methylene hydrogens
and one methine.



We can condense these fragments a bit by noting that two of the proton NMR signals are
doublets of doublets appearing between 4 and 4.5 ppm. To achieve this splitting pattern, one
pair of diastereotopic methylene hydrogens must be adjacent to the methine, which is a chiral
center. The methine itself is an apparent triplet at about 4.75 ppm, which is consistent with the
fact that J values for diastereotopic methylene hydrogens are not always significantly different in
magnitude.



At this point, weve accounted for most of the atoms in the formula, with the exception of the
two oxygen atoms.



Note that the remaining set of diastereotopic methylene hydrogens accounts for the most
deshielded signals in the spectrum (near 5 ppm). They are both doublets, meaning that they have
no additional neighbors. One way to achieve this powerful deshielding and exclude neighbors is
to place the remaining methylene group between the two oxygen atoms.

337



Recall that we still have one degree of unsaturation to account for. It must be a ring since we
saw no evidence for an additional bond. In order to form a ring of reasonable size and prevent
the methyl group from having neighbors, it is reasonable to form a methyl ester and attach it as a
substituent to the five-membered ring that can be made from the remaining two fragments.



This structure has some unusual features that should be apparent in the carbon-13 NMR if weve
assigned the structure properly. There are two highly deshielded carbon atoms: the one
appearing between two oxygen atoms and the one appearing between the ester and an oxygen
atom.



There are two moderately deshielded carbons: the methyl ester and the methylene adjacent to a
single oxygen atom. Finally, the ester carbonyl carbon appears as expected around 170 ppm.

4. As always, a degrees of unsaturation calculation is a logical place to begin. In this instance,
there is a particularly large extent of unsaturation in the molecule.

cgrccs o unsoturotion =
|2(18) + 2] 16
2
= 11

With large numbers of degrees of unsaturation, we often think in terms of benzene rings, each of
which will account for four degrees of unsaturation. Therefore, it is possible that the product
contains two benzene rings and three other degrees of unsaturation. This seems reasonable since
the reactants each contain an aromatic ring. If the reactants are united in some way to form the
product, it would indeed contain two aromatic rings.

The IR spectrum gives some insight into the remaining degrees of unsaturation. The signal at
1680 cm
-1
suggests a conjugated ketone or aldehyde. Since there is no aldehyde signal in the
338

proton NMR ( 9 10 ppm), a conjugated ketone seems probable. This will account for one of
the three remaining degrees of unsaturation. Additionally, the unusual C-H stretch around 3300
cm
-1
suggests something other than ordinary sp
3
C-H or sp
2
C-H stretching. If the alkyne is
retained in the product, then this would afford a distinct type of C-H stretching involving sp-
hybridized carbon. That would also account for the two remaining degrees of unsaturation.

We have already amassed a few fragments.



The
1
H NMR contains a fair number of signals. We can pare these down most easily by
considering the aromatic rings first. There are a total of nine protons in the aromatic region of
the spectrum. This is consistent with a monosubstituted and a disubstituted ring. Since we had
each of these structural features in the reactants, we might first check to see whether their
continued, unaltered presence in the product would explain the observed signals.



Indeed, it does. The disubstituted ring would yield a pair of doublets each integrating for two
hydrogens. These are observed near 7 ppm. Additionally, there is a singlet for three hydrogens
near 3.75 ppm, which is consistent with the continued presence of the methoxy group. The
remaining five aromatic hydrogens are conveniently explained by a phenyl substituent.

The fragments that we have now account for 16 carbons, 13 hydrogens, both oxygens, and all 11
degrees of unsaturation.



We have only two more carbon atoms and three hydrogens to complete the structure.
Considering the four signals in the proton NMR that we havent yet assigned (in colored boxes
below), the one with the least splitting is probably the terminal alkyne proton, which has no
immediate neighbors. The narrowly split doublet in the purple box would therefore correspond
to the alkyne proton, which experiences some long-range splitting.

339



This leaves three hydrogens (signals in the red, blue, and green boxes) to place on our remaining
two carbons. We therefore are forced to consider a methylene and methine fragment, in which
the methylene hydrogens are diastereotopic.



This provides a splitting pattern consistent with the spectrum. The diastereotopic methylene
hydrogens will each appear as a doublet of doublets. It may be difficult to know exactly which
one corresponds to which doublet of doublets, so they are highlighted together in the graphic
above. The methine hydrogen will have a fairly different chemical shift. It is common to see
such signals as apparent triplets if the J values for interaction with the methylene hydrogens are
not all that different in magnitude. Here, you may be able to detect some additional splitting in
the signal, which suggests some long range splitting may be occurring.

Our fragments now account for all of the atoms in the formula and all of the degrees of
unsaturation, so we can turn our attention to assembling them.



Since we do have reactant structures in this case, they can serve as a useful guide. For instance,
it would make sense for the alkyne and para-methoxyphenyl groups to be connected to the same
carbon, as they were in one of the reactants.

340



There is then only one way to piece together the remaining fragments.



5. A degrees of unsaturation calculation reveals a total of ten bonds and/or rings in the
products structure.

cgrccs o unsoturotion =
|2(18) + 2] 18
2
= 1u

Given that the reactants each contained a benzene ring, it is probable that eight degrees of
unsaturation can be attributed to those aromatic rings in the product. There are two additional
bonds and/or rings.

The IR data suggests that the oxygen atoms are not part of carbonyls or alcohols due to the
absence of signals in the 1700 and 3400 cm
-1
ranges. This implies that ethers may be important
functional groups as we move forward.



The proton NMR may seem daunting at first, so lets deal with the simplest portion. The
aromatic region of the spectrum reveals a total integration of nine hydrogens. This is consistent
with a disubstituted and monosubstituted ring. Since we had one of each in the reactants, it is
logical to assume that they are unchanged in the product.



This leaves us with six carbons and nine hydrogens. While all of the hydrogens are inequivalent,
there are not enough carbon atoms for nine methines. Consequently, there would need to be
three methylenes and three methines. Each of the methylenes must contain diastereotopic
hydrogens in order to explain the number of signals.

341



At this point, we have accounted for all of the atoms in the formula. Given that we didnt turn up
any more bonds, the remaining two degrees of unsaturation must be rings.

R
R
R
R CH
2
R
R CH
2
R
R CH
2
R
R
C R
R
H
R
C R
R
H
R
C R
R
H
diastereotopic methylenes
R
O
R R
O
R
two more rings


Due to the complexity of many of the proton NMR signals, it will be helpful to use some two-
dimensional NMR data to gain deeper insight into the connectivity. The HSQC spectrum is
exceedingly useful because it shows us which pairs of protons are bonded to the same carbon.
The similar proton NMR peak shapes could also have helped us toward this conclusion, but the
HSQC spectrum may be more clear. The proton signals that stem from the same carbon are
given the same color in the HSQC spectrum and are further highlighted in the spectrum below.

342



Now lets look at the adjacencies of these diastereotopic methylenes using the COSY spectrum.
For instance, the red diastereotopic hydrogens show a correlation not only to each other but also
to the green diastereotopic methylene hydrogens and to the blue methine. Piecing together this
fragment condenses the building blocks available for the structure.

343



If we examine the blue methine further, we see that it correlates with the two other methines in
the molecule (orange and light blue). This allows us to condense the fragments even more.

R
R
R
R CH
2
R
R
O
R R
O
R
two more rings
R CH
2
CH
2
C C
C
H
H
R
R
H
R
R


Both the light blue methine and the green methylene are deshielded. The green methylene falls
in a chemical shift range that would be consistent with close proximity to a heteroatom, like
oxygen. The light blue methine is even further deshielded. We can introduce one of the needed
rings and explain these chemical shifts by making a cyclic ether and tethering the
monosubstituted benzene ring to the light blue methylene. The oxygen deshields both the green
methylene and the light blue methine. The phenyl group further deshields the light blue methine
because its sp
2
carbon is electronegative and therefore withdraws additional electron density.



The orange methine is deshielded to a similar extent as the green methylene, so it seems
reasonable that it, too, is adjacent to an oxygen atom.



The remaining methylene (purple) experiences powerful deshielding, not unlike the light blue
methine which is bound to an oxygen and an electron-withdrawing sp
2
-hybridized carbon. We
344

can provide the purple methine with a similar chemical environment and complete the last ring
as follows.



Solutions to Chapter 5 Problems

1. In this problem, the nitro group must ultimately be replaced by an iodine. We know that
iodine can be installed via a substitution reaction of an aryl diazonium salt, so the principal
challenge is to convert the nitro group to a diazonium salt. This is a two-step process. First, the
nitro group is reduced to the amino group. Active metal reduction using tin or iron in acid will
accomplish the transformation. Alternatively, reduction using hydrogen and Pd/C is a fine
choice as well. The amino group is then diazotized, and the diazonium ion is appropriately
substituted by treatment with potassium iodide.



2. It is always a good idea to begin such synthesis problems with an analysis of the activating or
deactivating nature of the substituents, as well as their directing capabilities. The methoxy group
is electron donating via resonance, making it an activating group and an ortho / para director. In
contrast, the nitrile is electron withdrawing via resonance, so it deactivates the ring and directs
meta. Therefore, it seems that the nitrile could be useful in attaining the correct substitution
pattern. However, another important observation is that neither of these groups can be directly
installed on the ring through EAS reaction.

345



Consequently, we suspect that a diazonium ion will be an important intermediate in the addition
of each group to the ring. In order to obtain a diazonium ion, we must first nitrate the ring,
reduce the amino group, and diazotize. With the diazonium ion in hand, substitution to yield
benzonitrile affords a ring with the directing capability to give us the desired substitution pattern.



A second EAS nitration installs a nitro group meta to the nitrile. Its reduction, diazotization, and
displacement with water provide a phenol.



In order to convert the phenol to a methoxy group, we need only deprotonate and methylate.
Any of several bases could suffice to remove the phenolic proton, but given its acidity, we can
use the relatively gentle base potassium carbonate. Methyl iodide serves as the source of
electrophilic carbon.



346

3. It is important to remember that coupling reactions of diazonium ions require a fairly
nucleophilic aromatic ring, due to the low electrophilicity of the diazonium ion. This
necessitates an electron-donating group on the ring that attacks the diazonium ion.



This requirement factors into our selection of an appropriate starting material. If N,N-dimethyl-
4-nitroaniline is chosen as the starting material, then it will be converted to the diazonium ion.
This diazonium ion would be treated with benzenesulfonic acid; however, the sulfonic acid
group is electron withdrawing by resonance. Therefore, it is not a sufficiently nucleophilic ring
for this transformation. Additionally, even if we were to somehow force the coupling to occur,
the electron-withdrawing sulfonic acid group would direct the coupling to occur at the meta
position, rather than the desired para position.



Instead, 4-nitrobenzenesulfonic acid should be chosen as the starting material. It can be reduced
and diazotized. Subsequent treatment with N,N-dimethylaniline yields methyl orange because
the dimethylamino group is electron donating by resonance, making the ring of N,N-
dimethylaniline sufficiently nucleophilic to attack the diazonium ion.



4. In this synthesis, a nitro group is converted to a halide, which suggests the intermediacy of a
diazonium ion. Reduction and diazotization yield the diazonium ion. A subsequent Sandmeyer
reaction provides the desired aryl bromide.

347

O
O
2
N
MeO
O
Br
MeO
Sn
HCl
O
H
2
N
MeO
NaNO
2
HCl
O
N
2
MeO
CuBr


5. The important considerations here mimic those that we encountered in question 3. The
nucleophilic ring of the coupling reaction must be electron rich. Had we selected N,N-dimethyl-
4-nitroaniline as the starting material, the coupling partner would be the electron-poor benzoic
acid. Benzoic acid will not be a good coupling partner because the ring is too weakly
nucleophilic. Furthermore, even if the coupling reaction were somehow forced, the meta-
directing carboxylic acid would not allow for formation of the desired isomer.



Instead, o-nitrobenzoic acid is a better choice. Its reduction and diazotization can be followed by
coupling with the nucleophilic ring of N,N-dimethylaniline to yield methyl red.



6. Recall that halogens are the most unusual substituents on aromatic rings in that they deactivate
the ring inductively yet direct ortho / para due to their ability to stabilize the complex through
resonance. Having a pair of halogens meta to one another therefore appears to be impossible to
achieve. This suggests that at least one halogen should be installed using substitution of a
diazonium ion.

348



We know only one way to install fluorine, and it requires a diazonium ion. On the other hand,
chlorine can be installed through EAS chlorination or Sandmeyer reaction. Since theres only
one way to put the fluorine on the ring, we should begin that process, which will give us the most
flexibility with regard to placing the chlorine.

Nitration of benzene provides us with a meta director. Since we want the halogens to be situated
meta to each other, it makes sense to conduct EAS chlorination at this stage. Then, reduction of
the nitro group and diazotization of the amino group set the stage for fluorination.



7. This part of the triclosan synthesis amounts to a straightforward phenol preparation via a
diazonium ion. Reduction and diazotization are followed by treatment with water to obtain the
phenolic hydroxyl group.

O
Cl OH
Cl Cl
O
Cl NO
2
Cl Cl
triclosan
Sn
HCl
O
Cl NH
2
Cl Cl
NaNO
2
HCl
O
Cl N
2
Cl Cl
H
2
O


8. This synthesis problem is a bit less intuitive than some of the previous ones. Therefore, it
would be a good idea to do some retrosynthetic analysis to provide guidance on how best to
proceed. Recall from your introductory Organic Chemistry course that retrosynthetic analysis is
basically working backwards from the target to its precursors. With each retrosynthetic arrow,
we are asking ourselves what the compound could be made from in a single transformation.

349

The alcohol-bearing substituent is certainly the harder one to install on the ring. We know no
reaction that will introduce such a group directly. Furthermore, Friedel-Crafts acylation
followed by reduction can be used to make benzylic alcohols but wont work here because the
acid chloride reagent can possess only one R group attached to the carbonyl. The acid chloride
could not therefore bring with it both the cyclopentyl and methyl groups needed at the benzylic
carbon.



Lets step back and think more broadly about the problem. In other words, lets think outside the
context of EAS and diazonium ion reactions. We know that organometallic (e.g. Grignard or
organolithium) reagents can be used to make new carbon-carbon bonds at a carbonyl center.
This provides us with a new idea. What if the product were made from an arylmagnesium
bromide and a ketone? The Grignard reagent could, in turn, be made from 3-bromoanisole,
which can be readily prepared from 3-bromophenol.



Now that we have a more manageable immediate target (3-bromophenol), we can focus on its
preparation from benzene. Nitration of the ring affords a meta director and a prime opportunity
to add the bromine at the desired location through EAS bromination. Then, reduction and
diazotization set the stage for installation of the phenol through treatment of the diazonium ion
with water.

350



3-Bromophenol can now be methylated with methyl iodide (or a similar electrophile) using a
base, like potassium carbonate, to deprotonate the phenol. The Grignard reagent can then be
made through oxidative insertion of magnesium. It will add to cyclopentyl methyl ketone to
afford the desired product after workup.



Solutions to Chapter 6 Problems

1. The clue presented at the end of the question is that fluoride has a high affinity for silicon.
The formation of a strong Si-F bond is what drives the attack of fluoride on the trimethylsilyl
group. It may look unappealing to have an S
N
2-style attack on a highly substituted center;
however, since silicon is a larger atom than carbon, it makes longer bonds, which removes the
hindrance from the reactive center.

During this attack, the trimethylsilyl group is cleaved from the ring. The C-Si -bonding
electrons form the additional bond of benzyne as the triflate (i.e. trifluoromethanesulfonyl)
group is displaced.



351

2. In this reaction, the nucleophilic hydroxyl group can attack benzyne, while benzyne in turn
attacks the electrophilic carbonyl carbon of the aldehyde. This results in a doubly charged
intermediate, and a proton transfer yields the xanthene product.



3. We are told that the product is the result of [4+2]cycloaddition. Therefore, it makes sense to
draw mechanistic arrows for a retro-Diels-Alder process in order to unveil the necessary
reactants. Recall that a cyclohexene core is formed during Diels-Alder.



In retro-Diels-Alder, we can envision the undoing of the events that led to the formation of this
ring. The green bond is shifted over, as the red bond breaks. As the red bond is re-formed,
the blue bond breaks to re-form the blue bond.



Applying this strategy to the product of interest reveals the nature of the reactants, a substituted
benzyne and a dimethylfuran.



4. To begin this problem, we know from introductory Organic Chemistry that Grignard reagents
and organolithium species will add to ketones. This nucleophilic addition adds the second
methyl group to this center and forms an alkoxide intermediate.

352



At this point, Brook rearrangement becomes significant. In the text of the question, we are told
that the process involves intramolecular transfer of a silicon from carbon to oxygen. The
nucleophilic alkoxide oxygen will attack the silicon, cleaving it from the sp
2
-hybridized carbon.



The phenyl anion that results collapses to form a substituted benzyne, as the triflate group is
ejected from the substrate.



This benzyne undergoes [4+2]cycloaddition, analogous to the reaction discussed in question 3.



Finally, in a second step, a source of fluoride is added. This is different than the fluoride source
employed in question 1 (CsF); however, all that differs is the counterion. Fluoride cleaves the
trimethylsilyl group from the Diels-Alder adduct, and the resultant alkoxide will be protonated
upon workup to yield the product.

353



5. In the lithiated intermediate, the C-Li bond is polarized toward the more electronegative
carbon atom. Therefore, this carbon atom behaves much like a carbanion. Consequently, the red
electron pair displaces the better adjacent leaving group (Cl) and forms a fluorobenzyne.



6. The fluorobenzyne formed in question 5 now undergoes [4+2]cycloaddition analogous to the
reactions in questions 3 and 4 above. The adduct is bicyclic because the diene reactant was
already cyclic. As an additional ring is formed during the Diels-Alder reaction, the molecule
therefore becomes bicyclic.



7. Nucleophilic attach of the unhindered primary amine on the carbon at position 4 yields a
Meisenheimer complex that places the electron density on the electronegative heteroatom. There
are two chlorine-bearing carbons in the molecule that are electrophilic and could attract this
nucleophile. However, this particular one allows the stabilized Meisenheimer complex to be
formed without disrupting the aromaticity of both of the rings.

After attack, the nucleophile sheds a proton to the medium. The Meisenheimer complex then
restores aromaticity through the displacement of chloride, yielding chloroquine.

354



8. This Bergman cyclization follows the basic mechanistic paradigm for the process. Although
the substrate and product both bear an unusual long-chain substituent, this does not alter the
mechanism. One electron from each of the alkynes is contributed to form the bond of the new
ring. The remaining electron of each broken bond resides on the carbon that did not acquire a
new bond, giving the diradical known as a para-benzyne.



9. Reaction with the 4-chloro isomer very closely parallels question 7. A particularly stabilized
Meisenheimer complex forms, which has a resonance contributor that places the negative charge
on the most electronegative element in the ring.



The mechanism for reaction with the 3-chloro isomer illustrates why this compound is less
reactive. In the Meisenheimer complex, the negative charge cannot be placed directly on the
nitrogen. While this complex still benefits from resonance stabilization, it lacks the optimal
resonance contributor that is present in the other pathway, so the reaction is therefore slower.


355


10. This reaction also follows the basic Bergman cyclization mechanism. The challenge here is
clearly the drawing of the polycyclic product. In these sorts of situations, it is a good idea to
number or letter the atoms so that you can keep track of them and their substituents. One way to
do this is to label one of the alkyne carbons that will participate in the new bond as a.
Continue to label the carbons around the horizontal plane until you reach the other alkyne carbon
that obtains a new bond (f). This allows you to make sure that the correct ring size is drawn in
the product, which contains a para-benzyne.



11. One of the principal challenges of this question is that the nucleophile could conceivably
attack the ring carbon bearing chlorine or the ring carbon bearing fluorine. Remember that the
order of reactivity is determined by the halogens electronegativity. Fluorine yields the most
reactive center since it withdraws a great deal of electron density from the carbon to which it is
bonded, making that center especially electrophilic.

Another challenge is deciding which nitrogen atom should act as the nucleophile. The secondary
amine will be able to shed a proton after attack so as to become neutral. The tertiary amine will
not be able to do so, meaning that it would remain an excellent leaving group and would likely
be displaced from the Meisenheimer complex to re-form the reactants. Therefore, it makes the
most sense for the secondary amine to serve as the nucleophile in this reaction.

Now that the reactive sites have been determined, the S
N
Ar process itself is fairly
straightforward. The secondary amine attacks the carbon bearing fluorine. The resultant
Meisenheimer complex is stabilized not only by the usual resonance forms that place the
negative charge on multiple ring atoms but also by resonance into the sulfone. The nucleophile
sheds a proton, and fluoride is displaced as aromaticity is restored.

356



Solutions to Chapter 7 Problems

1. This very closely parallels the bromination of ethane discussed in the text. Initiation occurs
when a fluorine-fluorine bond homolyzes upon exposure to light or heat.



The first propagation step consists of hydrogen abstraction by the fluorine radical, forming HF as
a byproduct.



The carbon-centered radical thus formed then abstracts a fluorine atom from an unreacted
molecule of F
2
. This results in the formation of product and the regeneration of the fluorine
radical, making this a chain process.


357


The termination steps include the combination of any two radicals formed during propagation
and yield tiny amounts of fluorine, fluoroethane, and butane.



2. Recall that the change in free energy for a radical reaction is often about equal to the change in
enthalpy, provided that the change in entropy for the process is negligible. The change in
enthalpy is the enthalpy of the bonds broken minus the enthalpy of the bonds formed.

E

= |E

bonJs brokcn] |E

bonJs ormcJ]

In propagation step 1, the C-H bond is broken (98 kcal/mole) and the H-F bond is formed (136
kcal/mole). Therefore, the change in enthalpy for this step is -38 kcal/mole.

E

= |98] |1S6] = S8 kcolmolc

In propagation step 2, the F-F bond is broken (38 kcal/mole) while the C-F bond is made (107
kcal/mole). This results in a change in enthalpy of -69 kcal/mole.

E

= |S8] |1u7] = 69

These calculations illustrate that both propagation steps of radical fluorination are highly
exothermic. The overall enthalpy change for the reaction is -107 kcal/mole. As a consequence,
not only is free radical fluorination unselective (which is relevant in more complex substrates)
but it is also hazardous.

3. This mechanism also closely parallels the bromination shown in the text as well as the
fluorination shown in question 1. Initial homolysis forms iodine radicals.

358



In propagation step 1, hydrogen abstraction leads to the formation of a carbon-centered radical
and HI.

I + + H I
Propagation step 1
H


In propagation step 2, this cyclopentyl radical abstracts an iodine atom forming product and
regenerating the iodine radical for this chain reaction.



Termination entails the combination of any two radicals previously formed.



4. In propagation step 1, the C-H bond is cleaved (95 kcal/mole) and the H-I bond is formed (71
kcal/mole). The result is a step that is endothermic by 24 kcal/mole.

E

= |9S] |71] = +24 kcolmolc

In propagation step 2, the I-I bond is cleaved (36 kcal/mole) and the C-I bond is formed (53
kcal/mole). This step releases 17 kcal/mole worth of energy.
359


E

= |S6] |SS] = 17 kcolmolc

However, the net result is that the sum total of the propagation steps is endothermic. Since the
propagation steps as a whole require an input of 7 kcal/mole, the process is not favorable.
Unlike free radical bromination, the exothermicity of the second step is not sufficient to
overcome the endothermicity of the first step. Free radical iodination is therefore not often used
due to the fact that it is unfavorable.

5. Weve seen through this Chapter and the four previous problems, that free radical bromination
is the sweet spot for reactivity. Unlike fluorination, bromination is not hazardous. Unlike
iodination, bromination is favorable. Unlike chlorination, bromination is selective. These
features combine to make it the most attractive option for free radical halogenation.

The initial homolysis is identical to what weve seen previously, so it is not shown here. In
propagation step 1, the bromine radical preferentially abstracts hydrogen from the most highly
substituted carbon to make the most stable radical possible, which in this case is tertiary.



In propagation step 2, that tertiary carbon-centered radical abstracts a bromine to yield the
product and a new bromine radical.



As seen previously, termination steps involve the combination of any two radicals formed during
propagation. Since these steps are not especially significant in explaining product formation,
they arent explicitly illustrated here.

6. In propagation step 1, the 3 C-H bond is broken (91 kcal/mole) and the H-Br bond is formed
(88 kcal/mole). This step is endothermic, requiring an input of 3 kcal/mole.

E

= |91] |88] = +S

In propagation step 2, the Br-Br bond is broken (46 kcal/mole) and the 3 C-Br bond is made (65
kcal/mole). This step is exothermic and releases 19 kcal/mole.

E

= |46] |6S] = 19

360

Therefore, the overall process is exothermic with a change in enthalpy of -16 kcal/mole.

E

= |91 + 46] |88 + 6S] = 16

The selectivity of free radical bromination stems from the fact that propagation step 1 is
endothermic. Therefore, by the Hammond postulate, we expect the differences in energy
between possible radical intermediates to be reflected in nearly comparable transition state
energy differences. Since the energies of the transition states leading to 1, 2, and 3 radicals
differ significantly, the reaction is highly selective for the formation of only the most stable 3
radical.

The favorability of the process is the result of the overall change in enthalpy. The
endothermicity of propagation step 1 is more than compensated for by the exothermicity of
propagation step 2. Since the overall process is exothermic and the change in entropy is
negligible, the free energy change is expected to be negative, and the reaction is therefore
favorable.

7. This reaction is a free radical bromination at the benzylic position. Although the substrate is
more complex, the reaction does not substantively differ from the benzylic bromination of
toluene described in the text. Initiation occurs through the homolysis of the N-Br bond upon
heating.



The bromine radical thus formed abstracts a hydrogen from the benzylic methylene as mentioned
in the question. This provides a resonance-stabilized radical intermediate.



The HBr formed during propagation step 1 protonates another molecule of NBS. This enables
the nucleophilic attack of bromide on the electrophilic bromine of the conjugate acid of NBS.
Succinimide is displaced as a leaving group, and Br
2
is formed. As you draw this portion of the
mechanism be careful to note that this particular step is a heterolytic process, so it is important to
use the regular mechanistic arrows (as opposed to fish-hook arrows) in this step.

361



A bromine atom is abstracted from the newly formed Br
2
during propagation step 2, leading to
product formation. Bromine radical is also regenerated and can reenter propagation step 1.



Notice that a squiggly line is drawn for the new C-Br bond. This indicates that a mixture of
epimers is formed. Epimers are diastereomers that differ specifically in their configuration at
only one stereocenter. This was not problematic in the actual synthesis because the next step is
the elimination of HBr, which results in an alkene and therefore deletes this stereocenter.

8. The best approach for a problem like this is to work backwards one step at a time from the
given product structure. You know from introductory Organic Chemistry that ozonolysis cleaves
alkenes into aldehydes (and/or ketones) upon reductive workup with dimethylsulfide (DMS).
This analysis reveals that vinylcyclopentane is compound C.



The next step entails treatment of compound B with tert-butoxide to yield vinylcyclopentane.
Since tert-butoxide is a bulky base, it seems likely that an elimination occurs in this step.
Compound B therefore needs a leaving group on one of the two carbons pendent to the ring.
That leaving group should be bromine, based on the conditions of the step that forms compound
B. In theory the bromine could be at either the internal or terminal carbon. However, it is only
by positioning the bromine at the interior location that well be able to work backwards to a
compound A that is a trisubstituted alkene (as stated in the question).

362



Compound A is therefore the trisubstituted alkene to which anti-Markovnikov addition of HBr
will yield (1-bromoethyl)cyclopentane.



The overall reaction sequence is shown below.



Solutions to Chapter 8 Problems

1. Heat incites the fragmentation of AIBN, yielding two isobutyronitrile radicals in initiation step
1.



In the second initiation step, the tributyltin radical is formed as it undergoes hydrogen
abstraction.


363


Given the affinity of tin for halogens, we can expect it to abstract bromine from the substrate in
the first propagation step, resulting in the formation of a phenyl radical.



This radical happens to be situated five atoms from an alkene, so subsequent 5-exo-trig
cyclization yields a new benzylic radical.

N
O
O
O
OMe
OMe
Propagation step 2
N
O
O
O
OMe
OMe


In the final propagation step, this benzylic radical is reduced through hydrogen abstraction,
which also regenerates the tributyltin radical needed for another round of propagation steps.



2. The overall sequence is a Barton deoxygenation, which necessitates the conversion of the
alcohol to its corresponding thiocarbonate, also known as a xanthate. This is a three-step
process. Initial deprotonation of the alcohol yields the alkoxide. The alkoxide attacks carbon
disulfide to form the xanthate salt, which is then alkylated with methyl iodide.

364

Br
NC
OH
O
O
NaH CS
2 MeI
Br
NC
O
O
O
Br
NC
O
O
O
S
S
Br
NC
O
O
O
S
S


This substrate is now amenable to radical reaction due to the affinity of tin for sulfur. Tributyltin
radical (generated from AIBN and tributyltin hydride as seen in question 1) adds to the C-S
bond. This results in the formation of a C-O bond, as well as the homolysis of a O-C bond.



The resultant carbon-centered radical abstracts a hydrogen atom from tributyltin hydride to
afford the deoxygenated product.



3. As in question 1, we expect the tin radical to abstract bromine, leaving a phenyl radical
behind.



This phenyl radical happens to be five atoms from an alkene, so cyclization occurs to yield a new
five-membered ring. Labeling of the atoms (as shown below) will help you to accurately draw
the cyclization product.

365



Finally, hydrogen abstraction from tributyltin hydride reduces the radical.



4. Again, halogen abstraction would be expected as the first propagation step.



The phenyl radical thus produced is six atoms from one alkene and four atoms from another
alkene. Reaction with the closer alkene is not feasible for two reasons:

(a) The formation of a four-membered ring fused to a benzene ring would be unfavorable due to
high strain energy.

(b) Reaction with the same alkene in 5-endo-trig fashion is disfavored according to Baldwins
rules (refer to the Baldwins rules table in the text of this Chapter).

Consequently, reaction occurs with the distal alkene to yield a 6,6-fused system. Labeling of the
atoms (as shown below) continues to be an effective tool for accurate rendering of the product
structure.



The newly formed radical now happens to be six atoms from the second alkene, so another
cyclization results to yield yet another fused six-membered ring.

366



In the final propagation step, the radical is reduced and tributyltin radical is regenerated.



5. In the initiation, AIBN fragments as expected.



The isobutyronitrile radical then abstracts a hydrogen atom from hexanethiol as mentioned in the
text of the question.



The question also said that the reaction involved radical addition of sulfur. This suggests that the
thioradical should now add to a bond in the first propagation step.



The newly formed radical on carbon is five atoms from the other alkene bond, so cyclization is
expected to occur quickly.

367



Finally, the radical is reduced by hydrogen abstraction from another molecule of hexanethiol,
which regenerates the sulfur radical needed for propagation step 1.



6. The first three steps of the sequence are the preparatory steps of Barton deoxygenation, which
ultimately yield the methyl xanthate shown below (compound A).



Compound A then fragments in the presence of tributyltin radical. The addition of tin to the C-S
bond leads to the formation of a new C-O bond and the cleavage of a C-O bond.



The resultant carbon-centered radical is in fairly close proximity to the alkene. In theory, 5-exo-
trig and 6-endo-trig cyclization are both possible. The title of the paper provides a clue about
which process is favored in this instance when it refers to syntheses of carbahexopyranoses and
derivatives by 6-endo-trig radical cyclization. Therefore, following the 6-endo-trig paradigm, a
new six-membered ring is formed and the radical is then reduced by hydrogen abstraction.

368



7. In this question, we are told that tributyltin radical adds to the alkyne. Doing so results in a
vinyl radical. This vinyl radical can undergo a 6-endo-trig cyclization akin to the cyclization that
occurred in the previous question.



The final carbon-centered radical is reduced to afford an organostannane.



8. Although ACN contains rings that AIBN did not, its core structure is the same, so during
initiation it forms analogous radicals, which then abstract hydrogen from tributyltin hydride as
expected. The tributyltin radical then abstracts selenium to afford the acyl radical whose
cyclization we are asked to show. Remember that an acyl group refers to a carbonyl bearing a
single R group.

For the moment, it is best to ignore the product structure. As drawn, it looks quite complex and
may obscure what is an otherwise relatively straightforward process. We would expect the acyl
radical could cyclize onto the alkene of the eight-membered ring. This would be the transannular
(i.e. across the ring) acyl radical cyclization referred to in the title of the paper. Cyclization onto
either site would seem reasonable; however, the cyclization shown below affords a more highly
stabilized benzylic radical, which is desirable. The reduction of that radical yields the
transannular cyclization product.

369



To see if we have accurately identified the product, labeling of the atoms of the eight-membered
ring is useful. Doing so shows that we can label identical positions in both drawings. Also, we
can see that both structures contain a methylene-to-carbonyl bridge from C
b
to C
f
. This labeling
has illustrated that we did, in fact, correctly predict the product.

Solutions to Chapter 9 Problems

1. When the ring bears an electron-donating group, the regiochemistry is driven by avoiding the
inherent instability of an intermediate with a negative charge on the substituted carbon.
Consequently, the initial electron transfer yields a radical anion where the negative charge cannot
be delocalized onto the substituted carbon.



The following resonance structures further illustrate this point.



Protonation quenches the first anion, leaving a radical.



370

A second single-electron transfer to this radical yields the second anion.



This anion is quenched by proton transfer from the alcohol. The final product therefore has the
substituent on a vinylic carbon.



On the other hand, when the ring bears an electron-withdrawing group a different regiochemical
outcome is observed. The first single-electron transfer again yields a radical anion as expected.



This radical anion has the following resonance forms.



Protonation quenches the first anion, leaving behind the following radical.



The second electron transfer now generates an anion that is stabilized by the adjacent electron-
withdrawing group.

371



The final proton transfer affords the product.



2. This question reinforces what we saw with electron-withdrawing substituents in question 1.
The carboxylic acid will be deprotonated under the reaction conditions, so it appears as the
carboxylate throughout. The initial electron transfer yields the radical anion.



The negative charge is delocalized onto the carbon para to the substituted center.



This anion is then protonated.

RO H
H H
+ RO
O O O O


A second electron transfer yields an anion to the carbonyl. This anion is stabilized by its
conjugation to the carbonyl, which provides it with an especially stable resonance form that
places the negative charge on the carbonyl oxygen. This clearly illustrates that the carboxylate
serves as an electron-withdrawing group.

372



Protonation of this anion yields the product as a carboxylate.


This carboxylate would be protonated upon workup to give the corresponding carboxylic acid.



3. Both substituents on this aromatic ring are alkyl ethers, which are electron-donating by
resonance. Consequently, following the paradigm explained in question 1, we expect the product
to have both electron-donating groups on vinylic carbons.



As an aside, this Birch reduction also proceeds to cleave all of the benzyl (Bn) ethers from the
molecule. These groups served as alcohol protecting groups. Well learn more about benzyl
groups as protecting groups in Chapters 11 and 13.

373

Additionally, the product of this reaction sets the stage for the cleavage of the tagged
hydroxyquinone from the sugar. This is accomplished using acid. You might be interested in
trying your hand at a mechanism to explain this cleavage.

4. This ring bears two substituents, one electron-withdrawing and one electron-donating.
Therefore, we expect the regiochemistry to place the electron-donating group on a vinylic carbon
and the electron-withdrawing group on an allylic carbon.

However, there is an additional twist to this transformation. As stated in the question, the last
anion formed undergoes reaction with an electrophilic alkyl halide. This S
N
2 process yields the
substituted isoindolone shown below.



5. This is a Hofmann elimination that does not use silver oxide as a reagent. Compound A is the
result of methylation of the tetrahydroisoquinoline.



Subsequent treatment with base results in elimination. There are two -positions, and loss of a
proton anti-periplanar to the nitrogen from either site is feasible. However, the protons adjacent
to the electron-rich aromatic ring (purple -position) are not as acidic as those next to the phenyl
group (blue -position). Loss of a proton from the blue -position therefore cleaves open the
ring, giving an (E)-stilbene derivative.



6. The question states that compound A is the result of conjugate addition of piperidine to the
polymer-bound acrylate. This occurs through attack of piperidine at the -carbon. The enolate
374

that results undergoes a proton transfer. The ammonium ion loses a proton and the enolate gains
one to give compound A.



Compound A is then oxidized to the amine oxide. In the text of this Chapter, it was noted that
hydrogen peroxide or a peroxy acid could be used for this purpose. Here, the specific peroxy
acid used is mCPBA. The amine oxide is compound B.



Compound B then undergoes Cope elimination by removal of a proton to the carbonyl. This
re-forms the polymer-bound acrylate (compound C) and releases a hydroxylamine, which in this
case is piperidinol (compound D).



7. A selenoxide elimination is a convenient tool to install an ,-unsaturation. The process
would begin with deprotonation of the -carbon using a strong base, such as LDA. Then a
source of electrophilic selenium, like diphenyl diselenide, is added. Oxidation to the selenoxide
is achieved with an oxidant such as hydrogen peroxide. This is followed immediately by
elimination, which affords the desired product. In this case, that product is a cholestenone.

O O
1. LDA
2. Ph
2
Se
2
3. H
2
O
2


375

8. Lithium dicyclohexylamide is a strong base (much like LDA), which deprotonates the alcohol.



The alkoxide thus formed attacks the adjacent trimethylsilyl group to initiate a Peterson
elimination. The C-Si -bonding electrons form the new alkene bond, and the C-O bond is
cleaved to release the silyloxy anion.



9. The initial step of this sequence is formation of the phosphonium salt (compound A) through
displacement of bromide by triphenylphosphine.



Compound A is then deprotonated by 2,6-lutidine. Here, a weaker base than normal is used
because the proton to be removed happens to be to a carbonyl, which enhances its acidity. The
resultant ylide condenses with formaldehyde in a Wittig reaction to form an ,-unsaturated ester
as the product (compound B).



Compound B contains both an electron-rich diene and an electron-poor dienophile. These can
undergo intramolecular Diels-Alder reaction. It is very helpful to label the atoms, which will
376

assist you in accurately drawing the product of this [4+2]cycloaddition. It is also beneficial to
redraw compound B so as to emphasize the two six-membered rings that will form as a result of
this reaction.



Solutions to Chapter 10 Problems

1. An orthogonal protecting group is needed for the free secondary alcohol. A methyl ether
would be a poor choice here since it is difficult to cleave unless it is an aryl methyl ether.
Consequently, the MOM group is a good choice. It can be installed using MOMCl and N,N-
diisopropylethylamine. Stating the solvent is not necessary; however, solvents have been shown
below for your edification. Then, the TBDPS group can be removed through treatment with a
fluoride source. Using HF would be fine. The authors of this paper happened to use an
alternative fluoride source, known as TBAF (tetrabutylammonium fluoride). Well see the use
of TBAF in Chapter 13 as well.



2. Acerogenin G can be prepared simply by cleaving the aryl methyl ether protecting groups.
This is accomplished using boron trichloride or boron tribromide at low temperature. The free
phenols are liberated by subsequent treatment with aqueous base. It would be fine to show
sodium carbonate, as used in the text of this Chapter. The authors of the paper chose to use
saturated aqueous sodium bicarbonate instead.

The ketone can then be reduced to the secondary alcohol using sodium borohydride.
Alternatively, if you were concerned about possible complication resulting from the phenolic
hydroxyl groups, you could have reversed the steps for the synthesis of centrolobol (i.e.
reduction followed by deprotection).

377



3. Based on the synthesis above, a corresponding ketone seems to be a viable candidate for the
preparation of acerogenins E and K.



4. Treatment of the -ketoester with lithium aluminum hydride yields the diol (compound A).



The second step is a protection. The primary alcohol is less sterically hindered than the
secondary alcohol. Consequently, it reacts more rapidly with the TESCl and is selectively
protected.



The next transformation is an anti-Markovnikov hydration of the alkene. This could be
accomplished using conditions that you are familiar with from introductory Organic Chemistry
378

(e.g. BH
3
THF followed by hydrogen peroxide and sodium hydroxide as Reagents C). In the
article, dicyclohexylborane was used, which behaves in an analogous fashion.



The next step calls for the installation of the TIPS protecting group. This can be accomplished
with TIPSCl and imidazole (Reagents D), much like the TES group was installed. The size of
this group facilitates the selective protection of only the primary hydroxyl group.



Since PDC behaves like PCC, we expect that Compound E will be the ketone resulting from
oxidation of the secondary alcohol.



Finally, treatment with fluoride from tetrabutylammonium fluoride accomplishes global
deprotection, cleaving both of the silyl ethers to afford pterosin A.



5. Recall that mCPBA is a peroxy acid (specifically meta-chloroperoxybenzoic acid) and is
commonly used to prepare epoxides from alkenes, such as compound D.



Given that this alkene is the result of Wittig reaction of compound C with a single-carbon ylide,
we can surmise that compound C must be the aldehyde shown below.



379

Aldehydes result from the PCC oxidation of primary alcohols, such as compound B.



Monoprotection of compound A, a diol, would afford the silyl ether.



The synthesis of the epoxide can, therefore, be summarized as follows.



6. Deprotonation of the alcohol (either by potassium carbonate directly or by methoxide formed
from potassium carbonate in methanol) yields an alkoxide that is in close proximity to a carbon
bearing a tosylate leaving group. Intramolecular S
N
2 reaction affords the epoxide. The
stereochemistry is unaltered since the S
N
2 inversion occurs at the primary center.



In this way enantioenriched samples of both the (R) and the (S)-epoxide are obtained. The (R)-
epoxide is obtained directly from the HKR. The (S)-epoxide is obtained through this
intramolecular S
N
2 reaction of the activated diol.

7. The regiochemistry of the Diels-Alder reaction is one of the first decisions that we face.
Resonance structures of the reactants reveal a high electron density on the carbon of
Danishefskys diene furthest from the methoxy group. The acrylate, being an ,-unsaturated
ester, is naturally electron poor at the -position. These termini with complementary polarity
align during the [4+2]cycloaddition, which yields compound A.

380



Compound A is then treated with camphorsulfonic acid in THF. Protonation of the oxygen of
the trimethylsilyl ether renders it a better leaving group. The TMS group can therefore be
cleaved from the molecule by the attack of water (recall that the problem mentioned that THF
often contains moisture).



The resultant enol can be expected to rapidly tautomerize. Acid-catalyzed tautomerization
occurs through protonation of the bond followed by loss of a proton from the hydroxyl group.



The ketone loses methanol to install an ,-unsaturation. This occurs through the protonation of
the methoxy group, and the subsequent displacement of methanol as a proton is removed from
the -carbon. A concise way of representing this essentially pushes the light blue electrons
directly out to a proton as the methoxy group is lost. This reduces the drawing a bit (since the
intermediate with a protonated methoxy group is not explicitly drawn), but it also avoids the
implication that methoxide is lost, which would be highly unlikely in an acidic medium.

381



8. The cascade begins with the transfer of hydride from DIBAL-H to the lactone (i.e. cyclic
ester) carbonyl. The -bonding electrons will be displaced on the carbonyl oxygen. Drawing
that tetrahedral intermediate would certainly be fine. To condense the drawing a bit, the
tetrahedral intermediate is merely implied below by looping the -bonding electrons around the
carbonyl to suggest both the formation of the tetrahedral intermediate and its eventual collapse to
re-form the carbonyl, thereby displacing the carboxyl oxygen. For the sake of clarity, the
aldehyde proton (bonded through the red electron pair to the carbonyl carbon) is not explicitly
drawn in the following structures.



The carboxyl oxygen, once displaced, is actually part of an enolate. Resonance more clearly
illustrates the nucleophilic character of the carbon to the ketone.




382

This enolate can then attack the aldehyde in an intramolecular crossed aldol reaction to afford the
bicyclic product. The alkoxide would merely be protonated upon workup.



Solutions to Chapter 11 Problems

1. To prepare the necessary imine, well first need an amine corresponding to the segment of the
structure highlighted in the red box below.



This amine bears great homology with the chloroalkyne reactant. To convert the chloroalkyne to
the amine, you could have performed alkylation with a large excess of ammonia. Alternatively,
you could have displaced chloride with azide and used the Staudinger reduction. Yet one more
approach (and the approach used by the authors of the paper) is to employ the Gabriel synthesis.
Deprotonation of phthalimide is followed by treatment of potassium phthalimide with the
chloroalkyne. S
N
2 displacement of chloride yields the critical carbon-nitrogen bond.



Treatment of the alkylated phthalimide with hydrazine incites double nucleophilic acyl
substitution, which liberates the desired amine. This amine can then simply be condensed with
the appropriate ketone to afford the imine target.



383

2. The aldehyde will be expected to condense with allylamine to yield an iminium ion
intermediate. This intermediate can be reduced in situ to yield the secondary allylic amine.

N
SO
2
Ph
CHO
H
2
N
NaBH(OAc)
3
N
SO
2
Ph
N
H
N
SO
2
Ph
NH
Compound A


Treatment with Boc anhydride and a moderate base, like triethylamine, results in the protection
of the secondary allylic amine as its tert-butyl carbamate.



3. The Boc group is labile in acid, so step 1 results in its cleavage to unveil the secondary amine.
With the nucleophilicity of this nitrogen now restored, it can serve as a reactant in an S
N
2
process. Treatment with the dihalide in step 2 results in displacement of bromide. Remember
that vinyl halides are unreactive in both S
N
1 and S
N
2 reactions. Diisopropylethylamine (DIEA)
acts as a non-nucleophilic base to deprotonate the ammonium ion that results from the S
N
2
reaction. The reaction is dilute enough and the tertiary amine product is hindered enough that no
further alkylation results.



4. The azide has two resonance forms, and latter form shown below more clearly illustrates why
the attack of triphenylphosphine on the terminal nitrogen is feasible.

384

N N N
N N N
PPh
3
N N
N
Ph
3
P
TBDPSO
TBDPSO
TBDPSO


The resultant betaine undergoes ring closure through attack of the nitrogen anion on the
phosphonium ion.



Fragmentation of the four-membered ring extrudes nitrogen and yields an iminophosphorane.



The iminophosphorane is analogous to an imine and is therefore similarly susceptible to
hydrolysis. Hydrolysis of an imine would yield a free amine and a ketone or aldehyde.
Hydrolysis of the iminophosphorane unveils the free primary amine and releases
triphenylphosphine oxide as a byproduct.

385



5. The first step of the reaction is an alkylation of the primary amine. This reaction was sluggish,
so the authors utilized a Finkelstein reaction to accelerate the process and improve the yield. In
the Finkelstein reaction, a leaving group (such as mesylate) is first displaced by a halide (such as
iodide). Iodide is a fairly good nucleophile, so the displacement of the leaving group is facile.
Iodide is also a very good leaving group, so with this excellent leaving group on the electrophile,
S
N
2 union with the nucleophilic amine is accelerated.

N
O
NC NH
2
CH
3
+
O F
3
C
O
OMs
K
2
CO
3
, KI
AcN,
N
O
NC
H
N
CH
3
O
O CF
3
O F
3
C
O
I
Compound A


We are then told that the enantiomers are resolved (i.e. separated) and that a particular one is
utilized for the remainder of the synthesis. Boc-protection of the secondary amine yields
compound B.



Heating with potassium carbonate and ethanol leads to the cleavage of the acetyl group from
compound B. This is a case where the amide was serving as a protecting group. Ethoxide is
generated under the reaction conditions, and it can serve as the nucleophile in a nucleophilic acyl
substitution to unveil the secondary amine. In the text of this Chapter, we saw ammonia fulfill
that role, but it is not the only nucleophile that can suffice. Notice that the amine leaving group
is an aryl amine. As we noted, protection as the amide is typically used for aryl amines because,
in general, aliphatic amines are not sufficiently good leaving groups when cleavage of the
protecting group is attempted.

386



The secondary amine of compound C is then deprotonated by the strong base sodium hydride.
The nitrogen anion displaces bromide from the electrophile, yielding compound D via alkylation.



We are then shown that hydrogenolysis cleaves the benzyl ether. This is similar to cleavage of
the benzylic carbon-oxygen bond during the removal of a Cbz group. Well see this fact further
highlighted in Chapter 13. Subsequent treatment with trifluoroacetic acid (TFA) cleaves the Boc
group, liberating another secondary amine (compound E).



Compound E is then converted to silodosin by hydration of the nitrile.

6. We expect Hofmann elimination to begin with the deprotonation of the amide. Attack of the
conjugate base on the electrophilic bromine of N-bromosuccinimide halogenates the nitrogen.



A second deprotonation of the amide generates the anion that is primed for rearrangement. A
nitrogen lone pair (blue) collapses toward the carbonyl, forming a nitrogen-carbon bond. This
causes the shifting of the carbon-carbonyl bond (green), which results in the displacement of
bromide.

387



The isocyanate thus formed is receptive to the attack of water, which yields a carbamic acid.



Proton transfer allows for subsequent decarboxylation, yielding the primary amine with a carbon
skeleton that is reduced in size by one carbon atom.



7. The ,-unsaturated ketone is electrophilic at the -position, and more moderate nucleophiles,
like benzylamine, will attack at that center. The -bonding electrons are then displaced into the
carbonyl. The enolate is subsequently protonated, while the ammonium ion is deprotonated.
This affords the neutral product of conjugate addition to the enone. However, this is not
compound A.



388

The conjugate addition product can readily undergo intramolecular S
N
Ar reaction. As nitrogen
attacks the carbon of the pyridine ring bearing chlorine, the electrons can be displaced onto the
pyridine nitrogen atom or into the carbonyl. The initial intermediate sheds a proton, and then
aromaticity is restored as chloride is displaced to yield compound A.



This sequence may be called a tandem Michael-S
N
Ar annulation. Although formally the
Michael reaction refers to the conjugate addition of carbon-based nucleophiles, the name is often
used more broadly to refer to conjugate addition of other nucleophiles as well. The term
annulation refers to ring formation.

Compound B results from hydrogenolysis, which adds H
2
across the carbon-nitrogen bond of
the benzylamine. This releases the benzyl group as toluene and provides compound B as the free
amine.



8. The alcohol must first be converted to a good leaving group. This can be accomplished
through formation of a sulfonate. If you made the tosylate, using TsCl and a non-nucleophilic
base (pyridine, triethylamine, etc.), that would be fine. The authors of this paper selected the
mesylate instead, but the principle remains the same. The mesyl group is commonly abbreviated
as Ms.


389


The mesylate must then be displaced by a nucleophile. We know that simple alkylation of
ammonia is plagued by overalkylation unless a large excess of ammonia is used, so we might
consider a more elegant solution. Had you used potassium phthalimide for a Gabriel synthesis
that would certainly be reasonable. Alternatively, the mesylate can be displaced by sodium azide
to set the stage for Staudinger reduction.



Finally, the reduction of the azide to the primary amine is accomplished using
triphenylphosphine in the presence of water.



9. This synthesis requires the union of a benzylic halide with an aldehyde via formation of an
alkene, which suggests that a Wittig reaction would be a useful approach. The Z olefin in the
product is further evidence for the Wittig route, since the Wittig reaction leads to cis alkenes.

To begin the Wittig process, the benzylic halide must be converted to the corresponding
phosphonium salt by treatment with triphenylphosphine. Subsequent deprotonation using a
strong base, such as butyllithium, provides the ylide.



The ylide and the aldehyde can then be joined to yield the Z-stilbene.



The nitro group should then be reduced using hydrogen and a metal catalyst or using active metal
reduction. Tin and hydrochloric acid would be acceptable conditions for this transformation.
The authors happened to use zinc in acetic acid instead.

390



10. As stated in the question, the indole ring is nucleophilic at the 3 position. Electrons from the
nitrogen-magnesium bond flow into the ring and the -bonding electrons then attack the acid
chloride carbonyl. This results in acylation at C3. During the workup with aqueous acid, the
nitrogen will be protonated, which facilitates tautomerization to the restore the aromatic indole
ring system, yielding compound A.

O
N
MgBr
O
Cl
H
N
Me
FMOC
Compound A
O
N
O
HN
FMOC
Me
H
O
N
O
HN
FMOC
Me
H
O
N
O
HN
FMOC
Me
H
H
H


Sodium borohydride reduction of the ketone yields a mixture of epimeric alcohols (compound
B).

O
N
H
O
HN
FMOC
Me
Compound A
NaBH
4
cat. HCl
EtOH
Compound B
O
N
H
HO
HN
FMOC
Me

391


Finally, treatment with piperidine cleaves the FMOC protecting group. This occurs through
deprotonation of the fluorenyl moiety at C9. These -bonding electrons (blue) displace the
carboxylate, which concurrently decarboxylates by losing carbon dioxide. The nitrogen is
protonated by the conjugate acid of piperidine.



Solutions to Chapter 12 Problems

1. There are two alcohols that are susceptible to oxidation in this reaction. They are primary and
secondary. Primary alcohols are oxidized first to aldehydes and then further to carboxylic acids
under conditions utilizing chromic acid. The secondary alcohol is oxidized to the ketone, and no
further oxidation is observed.

O
HN
BocHN
OH
O
ZHN
OH
CrO
3
aq. H
2
SO
4
acetone
O
HN
BocHN
O
O
ZHN
OH
O


The oxidation of the secondary alcohol is undesired in this synthesis, so the ketone is reduced
back to the secondary benzylic alcohol in the next step of the sequence.

2. Ozonolysis with a reductive workup yields ketones and/or aldehydes, depending on the level
of substitution of the alkene substrate. In this case, acetone is cleaved from the molecule,
leaving the substrate with a new aldehyde functionality.

392



3. Friedel-Crafts acylation can occur at either site on the ring, yielding a mixture of compounds
A and B.



If we consider just one isomer for the moment, compound A can tautomerize in acid through
protonation of the carbonyl and subsequent loss of an -proton. The resultant enol can eject the
-chlorine to introduce an ,-unsaturation.



This ,-unsaturated oxonium ion can now undergo Nazarov cyclization. The aromatic ring
attacks the electrophilic -carbon, pushing electrons into the carbonyl. The resultant complex
(since in this case the reaction is also a sort of Electrophilic Aromatic Substitution) sheds a
proton to restore aromaticity. Tautomerization of the enol then yields indanone C.

393



In a similar fashion, compound B will yield indanone D.



4. The first step is protection of the ketone as an acetal (compound A). Careful DIBAL-H
reduction at low temperature then yields the aldehyde (compound B).



The phosphonium salt is deprotonated with butyllithium to afford ylide C, which then undergoes
Wittig reaction with the aldehyde previously prepared. The result is a mixture of alkene
geometric isomers (compound D). Reduction of the bond causes both isomers to converge on
a single product, acetal E. Finally, the ketone is unveiled via hydrolysis of the acetal.

394



This ketone (compound F) is condensed with pyrrolidine to yield the enamime. The enamine
then acts as a nucleophile and attacks the aldehyde of para-methoxybenzaldehyde. The resultant
-hydroxyiminium ion can undergo protonation on oxygen. Water is lost as an unsaturation is
introduced, and workup hydrolyzes the iminium ion to afford the ,-unsaturated ketone
(compound G).



The alkene of compound G is then reduced in the final step of this sequence.
395


5. Much like an ester can be reduced to an aldehyde using DIBAL-H, so too can a nitrile. A
single hydride addition to the nitrile occurs. Upon workup, the nitrogen anion is protonated, and
the imine thus formed is hydrolyzed to release the aldehyde (compound A). Sodium borohydride
reduction then yields the primary alcohol (compound B), which is subsequently acetylated with
acetic anhydride to produce compound C as the acetate ester.



6. In order for 1,3-dithiane to become nucleophilic, it must first be deprotonated using a strong
base, such as butyllithium. The carbanion thus produced will displace chloride from the THP-
protected alcohol to produce the substituted dithiane shown. A second deprotonation with strong
base yields a new carbanion, which is similarly alkylated to afford the disubstituted dithiane
product.



7. As in the previous problem, a strong base (this time tert-butyllithium) is used to generate a
carbanion (compound A).



396

This carbanion attacks the less hindered side of the epoxide. We are told that a Brook
rearrangement follows. The oxyanion attacks the silicon of the TBS group. This transfers the
silyl group to oxygen and generates a new carbanion (compound B).



This carbanion then attacks the less hindered side of the aziridine ring to install a second new
carbon-carbon bond. Upon workup, the charge on nitrogen is quenched to afford compound C.



8. Under acidic conditions the ketone and aniline derivative can certainly condense to form an
iminium ion, as the hint in the text of the question suggested. Cyanide then attacks the carbonyl-
like center of the iminium ion, displacing the electrons onto nitrogen to form the product.



9. Initial condensation of cyclopentanone and morpholine yields an enamine (compound A).
This enamine can then attack the -carbon of the ,-unsaturated ketone in a Stork enamine
synthesis. Upon workup, the enolate is protonated and the iminium ion is hydrolyzed, unveiling
the expected 1,5-dicarbonyl product.
397




10. Silver, having an affinity for halides, aids the dissociation of bromide as it is displaced by
intramolecular attack of the nitrogen. This results in an unusual bicyclic intermediate, which
naturally fragments quite readily. Attack of DMSO cleaves the [2.2.0]bicyclic system into a
single six-membered ring. Then, removal of a proton from the carbon bearing DMSO results in
oxidation to the ketone. The base could be the carbonate ion.



This mechanism has been presented as it was by the authors of the paper. However, had you
chosen to first deprotonate a methyl group of DMSO and then use that anion in an intramolecular
deprotonation (as the Swern mechanism was shown in the text of this Chapter), that too would be
reasonable.

Solutions to Chapter 13 Problems

1. Overall, the coupling yields the pentafluorophenyl (Pfp) ester shown below.

398



The Pfp ester is the result of a stepwise process in which the acid is first activated by EDAC.
Proton transfer enhances the nucleophilicity of the carboxylate while also increasing the
electrophilicity of the carbodiimide of EDAC. DMAP subsequently attacks the carbonyl of the
active ester, thereby displacing the urea byproduct. Finally, DMAP is regenerated when the
pentafluorophenoxide ion attacks the carbonyl to afford the Pfp ester.

O
O
O
BocHN
HN
O
ZHN
OH
HO
F F
F
F F
EDAC
DMAP (cat.)
C N N
N
H
O
O
O
BocHN
HN
O
ZHN
OH
N N
N
H
N
N
H
O
O
BocHN
HN
O
ZHN
OH
N N
N
H H
O
+
N
N
O
F F
F
F
F
O
OPfp
O
BocHN
HN
O
ZHN
OH


The Z (or Cbz) group is the only protecting group in the molecule that is susceptible to
hydrogenolysis. Once the free amine is liberated, it can perform intramolecular nucleophilic acyl
substitution. This attack on the active ester carbonyl displaces pentafluorophenoxide and yields
the macrocycle.

399



2. Triphenylphosphine initially attacks DEAD (diethyl azodicarboxylate). The nitrogen anion
that results deprotonates acetone cyanohydrin, which facilitates the release of cyanide from this
molecule. Subsequent attack of the alcohol on the electrophilic phosphonium ion cleaves it from
the DEAD residue. A final proton transfer yields an intermediate that is primed for S
N
2
displacement because the hydroxyl group has been converted to a good leaving group.

400



The last step is the second-order nucleophilic substitution that yields the nitrile product.



3. Although TASF is a different fluoride source than those weve seen previously, all that
ultimately matters is that it is a fluoride source. Fluoride is transferred from TASF to the -
trimethylsilyl group. This results in the cleavage of the Si-C bond and the formation of a carbon-
carbon bond in its place. The carboxylate is displaced as a resonance-stabilized leaving group.
Upon workup, protonation affords (-)-platensimycin.

401

O
O
Me
N
H
O
OH
OH
O
O
Si
TASF
Si
F
F
O
O
Me
N
H
O
OH
OH
O
O
workup
O
O
Me
N
H
O
OH
OH
O
O
H
(-)-platensimycin


4. There are two principle changes occurring in this transformation: the conversion of the ketone
to a vinyl halide and the formation of the pyrimidinone ring. The two tasks could be addressed
in either order. Lets examine the formation of the vinyl chloride first. Attack of the ketone on
the Vilsmeier reagent initiates the process. This ketone is rendered more nucleophilic by virtue
of the fact that it is actually a vinylogous amide. In other words, the ketone and the amine are
merely separated by an intervening vinyl group. Therefore, this functional group experiences the
same type of resonance donation into the carbonyl as an amide does.

Chloride is then displaced from the Vilsmeier reagent, and a chloride also attacks the carbonyl
carbon. Finally, loss of an -proton expels DMF and yields the vinyl chloride.



The second major task in this transformation is the formation of the pyrimidinone ring. This
process can begin with the attack of the amine on a second equivalent of the Vilsmeier reagent.
402

The Vilsmeier reagent then ejects chloride to re-form the carbon-nitrogen bond. The substrate
amine loses a proton, and tautomerization extends the conjugation in the system. At this point,
cyclization occurs, and the carbon of the Vilsmeier reagent is incorporated into the ring. Finally,
proton transfer activates the dimethylamino group as a much better leaving group, and its
displacement affords the pyrimidinone ring.

Cl
NH
2
N
H
O
Ar
Cl
N H
Cl
Cl
N
N
H
O
Ar
H
H
Cl N
Cl
NH
N
H
O
Ar
N
H
Cl
N
N
H
OH
Ar
N
- H
Cl
N
N
OH
Ar
N
proton
transfer
Cl
N
N
O
Ar
N
H
Cl
N
N
O
Ar
- H


5. Lithium hexamethyldisilazide (LiHMDS) is a strong base, which deprotonates chloroform.
The resultant trichloromethyl anion attacks the ketone to afford an -trichloromethyl alcohol
after protonation during workup.



Treatment of the -trichloromethyl alcohol with DBU (1,8-diazabicyclo[5.4.0]undec-7-ene),
another fairly strong base, incites attack of the alcohol on the adjacent trichloromethyl group. A
dichloroepoxide results, and it is then opened by nucleophilic attack of azide. If attack occurs at
the spirocyclic center (i.e. where two rings meet at a single point), then an acid chloride is
formed as chloride is expelled from the molecule. Finally, methanolysis of the acid chloride
affords the methyl ester. Subsequent reduction of the azide (as shown in the text of the
question) yields the amino ester product.

403



6. Diazomethane has two prominent resonance forms, one of which places high electron density
on the carbon atom. This resonance form clearly illustrates why the carbon atom can behave as a
base and deprotonate the carboxylic acid. The carboxylate thus formed attacks the methyl group
of diazomethane, which now bears nitrogen as an excellent leaving group. This results in
methylation and formation of (+)-echinopine B.



7. In this instance, hydrogen peroxide is used as the oxidant instead of a peroxy acid; however,
this makes little difference in the overall transformation. The two R groups attached to the
carbonyl carbon are identical, so only a single lactone (i.e. cyclic ester) product is possible.



8. The transformation proceeds exactly as expected for a Mitsunobu reaction. There is almost no
deviation from the mechanism shown in the text of this Chapter. Consequently, the ester is the
predicted product. However, it is important not to overlook the inversion of configuration that
occurs during the Mitsunobu reaction.

404



9. Near the end of this Chapter, we saw that DCC and DMAP can be used not only in the
synthesis of amides but also in the preparation of esters. That is precisely what happens in this
transformation, which unites the secondary alcohol of the reactant with the carboxylic acid of
Boc-N-Me-L-phenylalanine.



10. Initial treatment with trifluoroacetic acid (TFA) simply cleaves the acid-labile Boc group
from the substrate, unveiling the free secondary amine (compound A).



In the second step, a carboxylic acid is added, along with HATU and DIEA
(diisopropylethylamine, a base). DIEA will deprotonate the carboxylic acid. The carboxylate
then attacks the electrophilic carbon of HATU. The electrons flow onto nitrogen and then
collapse to expel the conjugate base of hydroxybenzotriazole (HOBt). This activate ester is
attacked by the HOBt anion, much like the active ester generated when using DCC can be
attacked by DMAP. In this way, a second active ester is formed and tetramethylurea is released
as a byproduct.

405



This second active ester is attacked by the free secondary amine of compound A. The
nucleophilic acyl substitution that results releases HOBt and yields compound B as the amide.

Compound A
O
N
H
OBn
O
O
N
H
Me
+
O
Me
O
N
N
N
N
N
N
O
O
N
H
OBn
O
O
N
Me
H
HOBt
O
Me
Compound B


11. Recall from the discussion of the Gabriel synthesis (Chapter 11) that hydrazine will cleave
the phthalimide through double nucleophilic acyl substitution to release the free primary amine
(compound A).



In the next step of the synthesis, two equivalents of this primary amine are joined with the
linchpin diacid using DCC as the coupling reagent and HOBt as the acyl transfer catalyst. This
yields compound B as a diamide.

406

CO
2
Me
CO
2
H HO
2
C
HOBt, DCC
Compound B
Compound A
O
H
2
N
O O
2
CO
2
Me
H
N
O
H
N
O
O O O O O O


The ester of the linchpin is then hydrolyzed via saponification to afford compound C as the free
carboxylic acid.



Finally, this free acid and the terminal amine of the bis(alkylaminostyryl) unit are coupled using
DCC and HOBt once again. The triamide (compound D) is the molecule required by the authors
for their study of resonance energy transfer.
407


408


Solutions to Chapter 14 Problems

1. Recall that Bergman cyclization will yield a para-benzyne.



These para-benzyne units can then combine to form the rigid polymer backbone, which bears
more flexible ester side chains.



2. An initiator, such as peroxides, would of course be needed for this process. However, the
specific initiator selected is immaterial for the general mechanism. Once radicals are formed
from the initiator, addition to the vinyl acetate bond can occur. This addition occurs so as to
form the more highly stabilized radical, which then adds to another monomers bond. The
process continues until the chain becomes quite long.

409



At this point, termination can occur by the joining of two radicals.



Alternatively, disproportionation can be the method by which chain growth is terminated.

410



3. As it is shown below, this particular polystyrene chain has atactic stereochemistry.



This polymer is one that we will recognize from the text of the Chapter as polystyrene. It would
be formed by the polymerization of styrene monomers.



Since the relative stereochemical orientation of the side chain phenyl groups is random, the
stereochemistry would be designated as atactic.

4. (a) The Ziegler-Natta catalysts operate through iterations of: (1) coordination between the
metal center and the monomer bond followed by (2) insertion of the new monomer unit into
the growing chain.



411

(b) As it is shown above, this growing polymer chain has isotactic stereochemistry. Since all of
the stereocenters have the same relative configuration, the polymer is described as isotactic.

5. The growing polymer chain can abstract a hydrogen atom from an adjacent polymer chain.
This will occur so as to place the new radical on a benzylic center, thereby making the most
stable radical possible.



This benzylic radical can then add to the bond of a styrene monomer. This introduces the
branch point.



The branch can continue to elongate through the addition of more styrene monomers.



6. A Lewis acid-base reaction between boron trifluoride and water generates the proton source
that initiates cationic polymerization. Isobutylene is protonated so as to make the more stable
tertiary carbocation intermediate, which then attracts another isobutylene monomer. The
polymerization continues in this fashion until the process is terminated by the transfer of
hydroxide from the Lewis acid-base complex to the cation.

412



7. The bond of methyl 2-cyanoacrylate is particularly susceptible to nucleophilic addition
because the anion that results is stabilized by two groups that are electron withdrawing by
resonance. This highly stabilized carbanion can then add to the -position of another methyl 2-
cyanoacrylate monomer. This may be termed a Michael addition since it is the conjugate
addition of an enolate. The process continues until eventual protonation terminates the sequence.



8. The order and frequency of occurrence of each monomer can of course vary, so many different
answers are feasible. One example is shown below. In this instance, all four carbons of the 1,3-
butadiene monomer have been incorporated into the polymer backbone.



It is also possible, however, for only two of the 1,3-butadiene carbons to become part of the
polymer backbone, in which case the backbone could be the following.

413



9. The monomer needed for the synthesis of this polyglycol is the following epoxide.



Opening of the epoxide by hydroxide at the less hindered side yields an alkoxide. This oxyanion
attacks another epoxide monomer at the less hindered side, lengthening the chain. The process
continues until protonation terminates the polymerization.



10. Butyllithium adds to the -position of acrylonitrile. The stabilized carbanion thus formed
will undergo conjugate additions to another monomer. This process repeats until all of the
monomer is consumed.



This process is a living polymerization, meaning that there is no way to terminate the
polymerization until another reagent is added to the system. If carbon dioxide is added, it serves
as an electrophile, and the polymerization is terminated by addition of the carbanion to CO
2
. The
result is a carboxylate.

414



Solutions to Chapter 15 Problems

1. The condensation polymer shown is a polycarbonate. Recall that, in the synthesis of
polycarbonates, it is the carbonyl carbon-to-carboxyl oxygen bonds that are formed. Therefore,
these bonds should be retrosynthetically disconnected to unveil the required reactants, 2,2-
dimethyl-1,3-cyclohexanediol and phosgene.



2. Kevlar is a polyamide, and when polyamides are prepared, it is the carbonyl carbon-to-
nitrogen bond that is formed. Therefore, this is the place for retrosynthetic disconnections,
which reveal the needed reactants to be 1,4-diaminobenzene and terephthalic acid (note that the
diacid chloride would also have been a reasonable selection).


415


Kevlars backbone contains many planar benzene rings, which help to account for the polymers
rigidity. Additionally, by aligning two adjacent strands, it is apparent that extensive interstrand
hydrogen bonding is also possible. This lends additional structural rigidity to the polymer.



3. The diamine and diacid can be condensed under high heat with the concomitant loss of water.
The result is the polyamide known as Quiana.



4. Kodel is a polyester, so it is the carbonyl carbon-to-carboxyl oxygen bond that should be
formed during its synthesis. Disconnecting at each of these junctures suggests terephthalic acid
and 1,4-cyclohexanedimethanol as reasonable starting materials. Had you chosen terephthaloyl
dichloride (the diacid chloride) as a reactant, that too would have been reasonable.

416



In actuality, Kodel can also be prepared from dimethyl terephthalate alone. Its reduction yields
1,4-cyclohexanedimethanol, and subsequent transesterification affords the polyester.



5. In general, polymers containing hydrocarbon backbones (like polystyrene, which is shown
below) will be more stable than those that contain other functionalities in the polymer chain.



The ester linkages of Kodel are likely to be more labile than the hydrocarbon core of
polystyrene.

6. The small alcohol initiator attacks the cyclic carbonate carbonyl and displaces an alkoxide.
The acid catalyst can protonate the carbonyl oxygen to enhance the electrophilicity of the
417

carbonyl carbon. Additionally, this acid catalyst can enable the leaving group to be the more
stable neutral alcohol, as opposed to the alkoxide. For simplicity, the acid catalyst has been
omitted from the scheme below.

Attack of the alkoxide on another monomer lengthens the chain. As this process repeats many
times, the activated polymer is formed.



The activated polymer can now be treated with a range of nucleophiles to install various side
chains. In this instance, 1-hexylamine is the nucleophile shown. Its attack on the carbonyl of the
Pfp ester results in nucleophilic acyl substitution, which displaces pentafluorophenoxide and
yields an amide side chain.

NH
2
polymer
activated polymer
OPfp O
O O
O
O O
O
O
OPfp O
H
n
NH O
O O
O
O O
O
O
NH O
H
n
NH
2
- H


7. This question is quite similar to the example epoxy resin found in the text of this Chapter. The
principal difference is that bisphenol F (made from phenol and formaldehyde) is used as the
linker instead of bisphenol A (made from phenol and acetone). Remember that the prepolymer is
prepared using a slight excess of epichlorohydrin so that the small chains terminate in epoxides.

418



In this instance, a diamine was chosen as the linker, so these ethylene diamine units open the
terminal epoxides and link the prepolymer chains to form the epoxy resin.



8. As explained in the text of the question, a melamine unit first condenses with formaldehyde to
yield an iminium ion. This iminium ion is then attacked by another melamine unit. As a result,
two melamine units are tethered by a methylene (i.e. CH
2
) from formaldehyde. This process
continues to afford the highly cross-linked structure of Melmac.

419



9. Attack of 1,4-butanediol on an isocyanate of TMXDI affords a urethane linkage. The alcohol
is drawn to the carbonyl carbon. The C-O bonding electrons are displaced onto oxygen but
then collapse to re-form the carbonyl. The C-N bonding electrons are subsequently converted
to a N-H bond. The net result is that a proton is transferred from the alcohol to the isocyanate
nitrogen as the urethane functional group is produced.



10. Treatment of the tetraol with base and tert-butyl acrylate results in conjugate addition to yield
compound A.

420



Mitsunobu reaction with three equivalents of perfluoro-tert-butanol (the G1 dendron) yields
compound B.



Reduction of the tert-butyl ester affords the alcohol (compound C).



This alcohol is then converted to the corresponding mesylate (compound D).

421



Displacement of the mesylate with thioacetate yields compound E as the thioester.



Reduction of this thioester affords the thiol (compound F), a model G2 dendron.



Monoprotected ethylene glycol is deprotonated by sodium hydride, and the alkoxide thus formed
displaces one of the halides of the tetrabromide reactant to generate compound G.

422



Under basic conditions, three equivalents of thiol F displace the remaining three bromide leaving
groups. The result (compound H) is a model G3 dendron.


423






About The Author



Michael S. Leonard earned his B.A. in Chemistry from Goucher College in 1998 under the
direction of Professor David E. Horn. He then transitioned to the University of Pennsylvania for
his doctoral studies in the laboratory of Professor Madeleine M. Joulli. After obtaining his
Ph.D. in 2003, he joined the faculty of Washington & Jefferson College, where he is currently
Associate Professor and Chair of Chemistry. The author gratefully acknowledges his mentors,
Professors Horn and Joulli, for their tireless dedication to education.