View Article Online / Journal Homepage / Table of Contents for this issue

FULL PAPER

www.rsc.org/dalton
Studies of oxalate-bridged MM quadruple bonds and their radical

Dalton
cations (M = Mo or W): on the matter of linkage isomers†‡
Malcolm H. Chisholm,* Jason S. D’Acchioli, Christopher M. Hadad* and Nathan J. Patmore
Department of Chemistry, The Ohio State University, 100 W. 18th Avenue, Columbus, OH,
43210, USA
Received 7th February 2005, Accepted 23rd March 2005
First published as an Advance Article on the web 15th April 2005
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.

Electronic structure calculations employing density functional theory (DFT) and time-dependent density functional
theory (TD-DFT) have been carried out on the model complexes {[(HCO2 )3 M2 ]2 (l-O2 CCO2 )}0/+ (M = Mo or W) in
D2h symmetry, where the oxalate bridge forms either five- or six-membered rings with the M2 centres; the complexes
are hereafter referred to as l(5,5)0/+ and l(6,6)0/+ , respectively. The calculations predict that the neutral complexes
should exist as the l(5,5) linkage isomer, while the radical cations favour the l(6,6) isomer by ca. 4–6 kJ mol−1 . For
the l(5,5) isomers, the rotational barriers about the oxalate C–C bond have been calculated to be 15.9 and
27.2 kJ mol−1 for M = Mo and W, respectively. For the cationic l(5,5)+ isomers the barrier is higher, being 36.8 and
50.6 kJ mol−1 for M = Mo and W, respectively. The calculated Raman and visible near-IR spectra for the l(5,5)0/+
and l(6,6)0/+ are compared with experimental data obtained for the {[(t BuCO2 )3 M2 ]2 (l-O2 CCO2 )}0/+ complexes,
hereafter referred to as M4 OXA0/+ (M = Mo or W). The experimental data more closely correlate with that
calculated for the l(5,5)0/+ linkage isomers, and the 13 C-NMR spectrum of the mixed metal complex Mo2 W2 OXA
indicates the presence of the 5-membered oxalate-bridged species (J CC = 100 Hz).

Introduction We have previously proposed for the DOP-bridged complexes,

Since the initial studies of electron delocalization in the mixed
valence Creutz–Taube ion,1,2 {[(NH3 )5 Ru]2 pz}5+ (pz = pyrazine),
the extent and mechanism of charge delocalization in complexes
near the Class II (strongly coupled)/Class III (fully delocalized)
border continue to attract both experimental and theoretical
interest.3,4 Closely related pairs of metal complexes involving
second and third row transition metal ions afford the unique
opportunity of comparing ions of almost identical size as a result
of the lanthanide contraction. The ions differ very significantly,
however, with regard to the relative orbital energies and the
diffuseness of the valence 4d and 5d atomic orbitals. For a
given ligand bridge, the 5d metal ions favour greater electron
delocalization mediated by M dp to bridge p* donation.5
However, the ligand bridge may greatly attenuate this effect.6
When oxalate bridges two metal–metal (M2 ) quadruple bonds,
the extent of electronic coupling is much greater for tungsten rel-
ative to molybdenum. This is evidenced by the values of K C , the
comproportionation constant,7 deduced from electrochemical
measurements, where K C is ∼1012 versus ∼104 for tungsten as
opposed to molybdenum, respectively.8 However, for the 3,6-
dioxypyridazine (DOP) bridge, the K C values are much closer
to each other, being ∼107 for molybdenum versus ∼1010 for
tungsten.9,10 The oxalate and DOP ligand bridges are depicted
in Scheme 1.
[M2 (O2 Ct Bu)3 ]2 (l-H2 C4 N2 O2 ), that when M = Mo the p-
orbitals of the bridge facilitate electron delocalization in the
radical cation by a hole-hopping mechanism which is not as
favourable for M = W.11 In the case of the oxalate analogues
[M2 (O2 Ct Bu)3 ]2 (l-O2 CCO2 ) (hereafter referred to as M4 OXA),
the highest energy filled p-orbital of the bridge is too low in
energy to contribute to the hole-hopping mechanism. However,
this line of reasoning suggests that even at distances of greater
than 6 Å, electron hopping via the oxalate p* system is
remarkably effective for M = W, given that K C = ∼1012 .
Could there be another explanation for the relative stabil-
ities of the oxalate cations? Recently Cotton and co-workers
reported the preparation and characterization of linkage isomers
involving five- and six-membered oxamidate rings (Fig. 1).12
Furthermore, they found that electron delocalization was much
greater in the six-membered ring isomer as opposed to the
five-membered one. Could the oxalate-bridged complex of
tungsten also possess a six-membered ring isomer in solution?
We undertook further computational and experimental studies
on oxalate-bridged M4 OXA complexes and their single-electron
oxidized radical cations, in order to interrogate this possibility
and report the results of this work herein.

Scheme 1

† This paper is dedicated to Ian P. Rothwell, in memoriam.

‡ Electronic supplementary information (ESI) available: MO represen-
DOI: 10.1039/b501938a

tations along with symmetry labels used for oxalate can be found in Table
S1 and Fig. S1. Electronic absorption spectra for {Mo4 OXA}+ {PF6 }−
can be found in Fig. S2. Plots of the potential energy surface scans for all
model complexes can be found in Fig. S3, and Cartesian coordinates of
all geometry optimized structures can be found in Tables S2 through S9.
Vibrational frequencies for each model compound can be found in Tables Fig. 1 Oxamidate isomers prepared by Cotton and co-workers (see
S10 through S17. See http://www.rsc.org/suppdata/dt/b5/b501938a/ ref. 12).

1852 Dalton Trans., 2005, 1852–1857 This journal is © The Royal Society of Chemistry 2005
 View Article Online

Experimental Results and discussion

Physical techniques
UV/Vis/NIR absorption spectra were recorded using a Perkin-
Elmer Lambda 900 UV/Vis/NIR spectrometer, with nitrogen
purging. A 1.00 mm or 10.00 mm IR quartz cell was employed.
A spectrum of the neat solvent (THF) was subtracted. Spectra of
the radical cations {[M2 (O2 CBut )3 ]2 (l-O2 CCO2 )}+ {PF6 }− (M =
Mo, W) were obtained by adding 1 equivalent of AgPF6 to the
corresponding neutral compound dissolved in THF.
Raman spectra were acquired with a 514.5 nm laser on an
f/1.5 Kaiser spectrograph with a holographic grating. A video
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.
Computational results
Density functional theory (DFT) and time-dependent density
functional theory (TD-DFT) calculations were used in order to
determine the preferred ground state structures of the model
complexes {[(HCO2 )3 M2 ]2 (l-O2 CCO2 )}0/+ (M = Mo, W) for
both linkage isomers; they are hereafter referred to as l(5,5)0/+
and l(6,6)0/+ , respectively, and are shown in Scheme 2.

charge coupled device (CCD) camera was used to focus the laser
at low power.
The 1 H and 13 C{1 H} NMR spectra were recorded on a
400 MHz Bruker DPX Avance spectrometer and referenced to
residual protio signals of d8 -THF at d = 3.58 and 1.72. Values
are quoted in ppm.
All compounds were manipulated by using standard dry
box and Schlenk line techniques, and THF was dried by
Scheme 2
refluxing over sodium metal, then degassed prior to use.
Mo2 (O2 CBut )4 ,13 W2 (O2 CBut )4 ,14 [Mo2 (O2 CBut )3 ]2 (l-O2 CCO2 )8
and [W2 (O2 CBut )3 ]2 (l-O2 CCO2 )8 were prepared according to Frontier molecular orbital interactions. The key molecular
previously published literature preparations. HO2 13 C13 CO2 H orbital (MO) interactions involved in the l(5,5)0/+ and l(6,6)0/+
was purchased from Cambridge Isotope Laboratories and used isomers are between the [M2 ]2 d and the oxalate p-orbitals (see
without further purification. ESI‡ for a description of the oxalate orbital symmetry labels,
as well as our convention for labelling); this is already well
Preparation of (But CO2 )3 Mo2 (l-O2 13 C13 CO2 )W2 (O2 CBut )3 established5 for the molybdenum and tungsten complexes of
(Mo2 W2 OXA*) l(5,5)0 . In D2h symmetry, the [M2 ]2 d orbitals form in-phase and
A Schlenk flask was charged with W2 (O2 CBut )4 (300 mg, out-of-phase combinations with the oxalate p-orbitals as shown
0.39 mmol), Mo(O2 CBut )4 (232 mg, 0.39 mmol) and in Fig. 2. The interactions which lead to the splitting of the d
HO2 13 C13 CO2 H (18 mg, 0.20 mmol) and ca. 10 ml of toluene. orbitals are shown qualitatively in Fig. 3. For both the l(5,5)0/+
The suspension was stirred for 7 days, during which time a and l(6,6)0/+ isomers, the [M2 ]2 d MO of b2u symmetry involves
blue precipitate formed. The dark blue solid was isolated by the same combinations of [M2 ]2 d and oxalate p5 character; the
centrifugation, and washed with 2 × 10 ml portions of toluene interaction is one of metal-to-ligand back donation. The [M2 ]2
and 1 × 10 ml of hexanes before drying in vacuo. The 1 H NMR d MO of b3g symmetry which is formed from the interaction
spectrum of the final product mixture indicated the formation of between the filled–filled [M2 ]2 d and oxalate p MO’s, however,
Mo4 OXA*, Mo2 W2 OXA* and W4 OXA* in a 1 : 2 : 1 ratio. d 1 H differs between the l(5,5)0/+ and l(6,6)0/+ isomers. For the
(400 MHz, d8 -THF): 1.43 (9H, s), 1.42 (18H, s), 1.41 (9H, s), l(5,5)0/+ isomers, the oxalate orbital contribution is CO2 p-
1.39 (18H, s). d 13 C{1 H} (101 MHz, d8 -THF): 161.02 (d, J CC = bonding and resembles the oxalate p2 MO, while for the l(6,6)0/+
100 Hz), 159.25 (d, J CC = 100 Hz) isomers, the oxalate contribution is oxygen p-nonbonding, and
resembles the oxalate p3 MO. The latter interaction contributes
Electronic structure calculations to the greater splitting of the [M2 ]2 d MO’s in the l(6,6)0/+
isomers, by ca. 0.3 eV (Fig. 4). The calculated total energies
Molecular and electronic structure determinations on the model of the two linkage isomers indicate the l(5,5) complexes are
l(5,5)0/+ and l(6,6)0/+ complexes (M = Mo, W) were performed lower in energy for the neutral molecules, but that for the radical
using density functional theory (DFT) as implemented in the cations, the l(6,6) isomers are more stable. For tungsten, the
Gaussian03 suite of programs.15 We have used the B3LYP16–18 l(6,6)+ isomer is predicted to be thermodynamically more stable
functional along with the 6-31G* basis set for H, C and O,19 and by ca. 4 kJ mol−1 (Table 1).
the SDD energy-consistent pseudopotentials for Mo and W.20
All geometries were fully optimized in D2h symmetry. It should
be noted that the symmetry labels of the MOs presented herein
differ slightly from those published previously.5,27 The difference
is due to a reorientation of the model complex along its Cartesian
axes.
Vibrational frequency analyses of the optimized structures
found them to be local minima on their respective potential
energy surfaces. The barriers for rotation of the l(5,5)0/+ model
complexes were determined by performing a relaxed potential
energy surface (PES) scan. The geometry of each complex was
optimized in C 1 symmetry in 15◦ increments from 0◦ (“D2h ”)
to 90◦ (“D2d ”). Orbital and vibrational frequency analyses were
completed with GaussView.21 Electronic absorption spectra were
predicted using the time-dependent DFT (TD-DFT) method
as implemented within Gaussian03.22–24 Visualizations of the
computed TD-DFT absorption spectra were facilitated through Fig. 2 a) Out-of-phase d orbital combination with oxalate p2 orbital.
the use of the SWizard program.25 The spectra produced were b) In-phase d orbital combination with oxalate p5 orbital.
useful for gauging the positions of calculated absorptions only;
no attempt was made to adjust for any Gaussian, Lorentzian, Spectroscopic properties for which experimental data were
or pseudo-Voigt character of the absorptions. readily available have also been calculated. The Raman

Dalton Trans., 2005, 1852–1857 1853

 View Article Online

Table 1 Relative stability (kJ mol−1 ) of the l(5,5)0/+ isomers vs. the l(6,6)0/+ isomers

[M2 ]2 -l(5,5)0 vs. [M2 ]2 -l(6,6)0 [M2 ]2 -l(5,5)+ vs. [M2 ]2 -l(6,6)+

M = Mo M=W M = Mo M=W
Relative stability −18.8 −19.2 3.8 5.9

shown schematically in Fig. 5. TD-DFT calculations were also

used to determine the energies of the two most salient electronic
transitions, namely the fully allowed [M2 ]2 d → oxalate p5
transition for all species, as well as the fully allowed d(HOMO-
1) → d(HOMO) inter-valence charge transfer (IVCT) transition
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.

for the radical cations.

Fig. 3 Qualitative MO diagram showing the formation of Mo4 OXA

from interactions of the frontier orbitals of an [Mo2 ]2 fragment with
oxalate p orbitals. Splitting of the d orbitals via symmetry allowed
combinations is attributed to a filled–filled interaction (d orbital of b2u
symmetry with oxalate p2 ) and a filled–empty interaction (d orbital of
b3g symmetry with oxalate p5 ). For simplicity, only the HOMO is shown
with a pair of electrons.

Fig. 5 Vector representations of the calculated m1 , m2 , and m3 normal

modes of the oxalate moiety of the D2h l(5,5) isomer (adapted from
ref. 5).

The calculated values for the electronic transitions and

vibrational frequencies are shown in Tables 2 and 3, respectively.
There are three particularly noteworthy observations to be

Table 3 Calculated (D2h symmetry) and experimentally observed Ra-

man vibrational frequencies (experimental values are in parentheses)

m1 /cm−1 m2 /cm−1 m3 /cm−1

Fig. 4 Qualitative MO diagram comparing the relative energies of
the neutral l(5,5) oxalate-bridged complexes with the l(6,6) model [Mo2 ]2 -l(5,5)0 1422 (1384)a , b 945 (910)a , b 584 (565)a , b
complexes. The interaction of the oxalate p3 MO in the l(6,6) isomers [Mo2 ]2 -l(6,6)0 1375 971 473
with the [M2 ]2 d based MOs leads to a greater splitting of these orbitals [Mo2 ]2 -l(5,5)+ 1395 (1322)a , b 945 (922)a , b 592 (578)a , b
relative to the l(5,5) isomers. For simplicity, only the HOMO is shown [Mo2 ]2 -l(6,6)+ 1324 984 484
with a pair of electrons. [W2 ]2 -l(5,5)0 1390 (1324)a 941 (923)a 594 (589)a
[W2 ]2 -l(6,6)0 1281 1866 481
stretching frequencies for the l(5,5)0/+ and l(6,6)0/+ isomers [W2 ]2 -l(5,5)+ 1399 (1383)a 926 (917)a 612 (617)a
were calculated in order to gauge trends in m1 , m2 , and m3 for [W2 ]2 -l(6,6)+ 1273 993 503
the oxalate moiety. These are the totally symmetric vibrational a
Spectra obtained from a THF solution. b Taken from ref. 26.
modes involving the mixing of m(CO2 ), m(CC), and d(CO2 ),

Table 2 Comparison of experimental (in parentheses, recorded in THF) with calculated (D2h symmetry) electronic transition energies for l(5,5)0/+
and l(6,6)0/+ isomersa , a

[M2 ]2 d → oxalate p5 [M2 ]2 d → [M2 ]2 d (IVCT)

Excitation Excitation Oscillator Excitation Excitation Oscillator
wavelength/nm energy/eV strength (f ) wavelength/nm energy/eV strength (f )

[Mo2 ]2 -l(5,5)0 449 (460) 2.76 (2.70) 0.514 — — —

[Mo2 ]2 -l(6,6)0 480 2.59 0.414 — — —
[Mo2 ]2 -l(5,5)+ 485a (666) 2.56 (1.86) 0.000 1420 (2500) 0.87 (0.496) 0.231
[Mo2 ]2 -l(6,6)+ 537a 2.31 0.000 1196 1.04 0.194
[W2 ]2 -l(5,5)0 529 (704) 2.34 (1.76) 0.722 — — —
[W2 ]2 -l(6,6)0 600 2.07 0.536 — — —
[W2 ]2 -l(5,5)+ 523 (639) 2.37 (1.93) 0.601 1180 (1700) 1.05 (0.729) 0.178
[W2 ]2 -l(6,6)+ 592 2.09 0.526 1078 1.15 0.0597
a
The [M2 ]2 d MO in these transitions was calculated to be the HOMO-1. b Experimental values taken from ref. 8 and 26.

1854 Dalton Trans., 2005, 1852–1857


gleaned from these data. 1) The [M2 ]2 d → oxalate p5 electronic
transition will be lower in energy for the l(6,6) isomers relative
to the l(5,5) isomers (M = Mo, W). 2) The IVCT electronic
transition will be higher in energy for the l(6,6)+ isomers relative
to the l(5,5)+ isomers (M = Mo, W). 3) For M = W, the value
of m1 will be higher for the l(5,5)+ isomer relative to its neutral
analogue, while for l(6,6)+ , the value of m1 is lower in energy.
We have previously noted the thermochromic behaviour
of the M4 OXA compounds in 2-methyltetrahydrofuran, and
ascribed this phenomenon to a Boltzmann distribution of
rotamers arising from the relatively small barrier to rotation
about the oxalate C–C single bond;5 an alternative explanation,
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.
however, could be formulated based on a thermal equilibrium of
l(5,5) and l(6,6) linkage isomers. We consequently carried out a
relaxed potential energy surface (PES) scan by varying the O2 C–
CO2 dihedral angle in 15◦ increments for the l(5,5)0/+ isomers.
At each point, the geometry was optimized in C 1 symmetry and
the electronic excitations were calculated using TD-DFT. There
are two significant points that emerge from this study. 1) The
barriers to rotation in the l-(5,5) isomers about the central C–C
bond are notably greater in the oxidized radical cations, being
36.8 vs. 15.9 kJ mol−1 for M = Mo, and 50.6 vs. 27.2 kJ mol−1
for M = W. 2) There is a diminution in oscillator strength and a
blue shift in the energy of the [M2 ]2 d → oxalate p5 transition as
the O2 C–CO2 dihedral angle shifts from 0◦ (planar D2h ) to 90◦
(twisted D2d ); the effect on the calculated absorption spectra for
[W2 ]2 l(5,5)0 is shown pictorially in Fig. 6.
Fig. 6 The effect of rotation about the oxalate C–C bond in W4 -l(5,5)0
on the [M2 ]2 d → oxalate p5 transition. There is a noticeable blue-shift
in the energy of the transition as the dihedral angle increases from 0◦
to 90◦ , as well as a decrease in the calculated intensity. The presence of
more than one excitation band in the 45◦ , 60◦ , 75◦ , and 90◦ rotamers
results from the mixing of electronic states which occurs in C 1 symmetry.
Spectroscopic studies
NMR spectroscopy. The previous NMR spectroscopic data
for the neutral M4 OXA compounds provided no evidence for the
existence of two isomers in solution.8 However, the NMR data
did not preclude that two isomers could be undergoing a rapid
(on the NMR timescale) equilibration nor could it distinguish
between the two possible linkage isomers (a and b; see Scheme 1)
of M4 OXA. The appearance of cis and trans pivalates in the ratio
2 : 1, and a single oxalate 13 C signal could easily be accounted
for by the presence of the b-isomer. In an attempt to distinguish
between these possibilities, we carried out the reaction shown in
eqn. (1),
4 Mo2 (O2 CBu )4 + 4W2 (O2 CBu )4 + 4HO2 C CO2 H
t t 13 13
25◦ C,7days
−−−−−−→ Mo4 OXA∗ + W4 OXA ∗ +2Mo2 W2 OXA∗ (1)
toluene
where we anticipated the yield of a near statistical mixture of
products as indicated.
View Article Online
This product mixture was examined by 13 C{1 H} NMR spec-
troscopy, and as shown in Fig. 7, the two singlets are assigned to
the homometallic species, M4 OXA* (OXA* = O2 13 C13 CO2 , M =
Mo, W). In addition, there are four resonances which comprise
an AB spin system due to two non-equivalent 13 C nuclei with
J CC = 100 Hz. From this, we can conclude that only the oxalate
a-isomer is present for the mixed metal complex, Mo2 W2 OXA*,
as the two oxalate carbons would be equivalent in the b-isomer.
This finding is consistent with expectations based on earlier
solid-state studies employing powder X-ray diffraction data.5
Fig. 7 Oxalate region of the 13 C{1 H} NMR spectrum of the mixture
of Mo4 OXA*, W4 OXA* and Mo2 W2 OXA* in d8 -THF.
Electronic absorption spectra
The spectra for the neutral compounds in 2-methyltetra-
hydrofuran (2-MeTHF) from 2 to 298 K, and tetrahydrofuran
(THF) at room temperature have been discussed previously.5
The focus of this work is on the radical cations and how these
observed transitions correlate with expectations for the model
compounds l(5,5)+ and l(6,6)+ .
The electronic absorption spectrum of Mo4 OXA+ is given in
the ESI (Fig. S2),‡ and shows a very broad band at low energy,
centred around 2500 nm, and a relatively sharp absorption at
666 nm. Unfortunately, the radical cation is not chemically per-
sistent, and bleaching occurs within ca. 90 minutes. Previously,
we have noted that the low-energy IVCT band is consistent
with expectations of Class III behaviour (H ab = 2000 cm−1 ).26
Based on these calculations, the higher energy absorption at
666 nm would arise from the [M2 ]2 d HOMO → oxalate p5
transition, a metal-to-ligand charge transfer, while the lower
energy band would be the d HOMO-1 → d HOMO transition.
Time-dependent density functional theory calculations predict
that the former transition occurs at lower energy than the related
[M2 ]2 dHOMO → oxalate p5 transition in the neutral compound,
but the energies of the calculated transitions are so different from
those observed as to make a distinction between the two linkage
isomers impossible.
The room temperature spectrum of W4 OXA+ in THF is
shown in Fig. 8, alongside spectra acquired at one day intervals.
Again, we see bleaching of the W4 OXA+ radical cation, but
this time over a time period of days. The spectral changes seen
in Fig. 8 are interesting, and we see the growth of new lower
energy absorptions corresponding to the formation of a new
species in solution. Based on the calculations, the higher energy
absorptions at 639 nm (15650 cm−1 ) would be assigned to the
[M2 ]2 d HOMO → oxalate p5 transition, while the lower energy
absorption at ca. 1700 nm (5880 cm−1 ) would be assigned to
the d HOMO-1 → d HOMO transition (the IVCT band). It is
interesting to note that the calculated [M2 ]2 d HOMO → oxalate
p5 bands for the l(5,5)+ and l(6,6)+ model compounds differ
significantly in energy. The calculations predict that the l(6,6)+
isomer will be at lower energy (longer wavelength) relative to
the l(5,5)+ isomer. Thus, the diminution of the absorption at
Dalton Trans., 2005, 1852–1857 1855
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.
Fig. 8 Electronic absorption spectra of W4 OXA+ PF6 − in THF recorded
at one day intervals. The first spectrum (t = 5 minutes) is red, and the
final spectrum (t = 5 days) is blue.
−1
639 nm (15650 cm ) and the growth of the absorption at
830 nm (12050 cm−1 ) could be taken as evidence for linkage
isomerization, and the formation of W4 OXA-l(6,6)+ . However,
this line of reasoning is not corroborated by the observed
changes in the low energy (near IR) region. The calculations
predict that the d HOMO-1 → d HOMO IVCT transition should
be at a higher energy for the W4 OXA-l(6,6)+ isomer. What
is observed is the growth of a broad electronic transition at
lower energy (longer wavelength), at ca. 2000 nm (5000 cm−1 ).
Based upon previous electronic structure calculations, we might
speculate that this arises from a ligand redistribution reaction
where a higher order oligomer, either a chain compound or a
square, is formed.27 What is clear is that the observed changes
are not in accordance with the calculated predictions for the
existence of the l(6,6)+ isomer for M = W.
Finally, we note that the electronic spectra of W4 OXA+ in 2-
MeTHF shows relatively little temperature dependence. At 4 K,
there is evidence for a vibronic progression at the higher energy
absorptions (ca. 700 nm). However, in contrast to the neutral
W4 OXA complex,5 the electronic transition is not significantly
red shifted upon cooling. Also, the low energy band at 1700 nm
(5880 cm−1 ) does not shift significantly and is consistent with a
higher barrier to rotation about the central oxalate C–C bond.
Raman spectra. The Raman and resonance Raman spectra
for the M4 OXA compounds have been previously reported
and noted to have pronounced resonance enhancement of the
totally symmetric oxalate modes m1 , m2 and m3 (Fig. 5) when the
wavelength of Raman excitation is in the envelope of the [M2 ]2 d
HOMO → oxalate p5 electronic transition.5,26
The Raman spectrum of Mo4 OXA and Mo4 OXA+ in THF
have been previously reported.26 The Raman spectra of W4 OXA
and W4 OXA+ were obtained in THF with a laser excitation of
514 nm, and are compared in Fig. 9. Clearly different bands show
different resonance enhancements and, in general, the bands
associated with the radical cation W4 OXA+ are less enhanced
as perhaps expected, since the excitation wavelength at 514 nm
does not fall so clearly within the envelope of [M2 ]2 d HOMO →
oxalate p5 transition. Nevertheless, we see significant changes
with oxidation, as m1 shifts from 1324 to 1383 cm−1 ; m2 from 923
to 917 cm−1 and m3 from 589 to 617 cm−1 .
As shown in Fig. 5, the simple interpretation of m1 as m(CO2 )
and m2 as m(CC) and m3 as d(CO2 ) is far from accurate, as there is
extensive mixing of these totally symmetric vibrations. However,
the change in m1 that is observed upon oxidation is consistent
with less p-back bonding ([M2 ]2 d HOMO → oxalate p5 ) that
weakens the C–O bonds. The changes observed in the Raman
spectra upon oxidation of M4 OXA to M4 OXA+ are consistent
with the calculated trends for the l(5,5)0/+ isomers (see Table 3).
1856 Dalton Trans., 2005, 1852–1857
View Article Online
Fig. 9 Raman spectra of W4 OXA (blue) and W4 OXA+ PF6 − (black) in
THF solutions recorded using 514 nm laser excitation.
Conclusions
A broad range of theoretical and experimental techniques have
been used to probe the phenomenon of linkage isomerization in
oxalate bridged dimetal units. Based on the predictions for the
model compounds l(5,5)0/+ and l(6,6)0/+ and the experimental
data presented, we find no evidence for the existence of a linkage
isomer in the compounds M4 OXA0/+ having six-membered rings
being present in solution. Indeed, the 13 C NMR spectrum of the
labelled oxalate mixed metal complex requires the presence of
the l(5,5) isomer. The calculations do, however, suggest that
the isomer having six-membered rings would be significantly
stabilized for the radical cation relative to the isomer with
five-membered rings. We might expect to find these studies
significant and useful in the study of mixed-metal complexes of
the type shown below in Scheme 3, where the five-membered ring
involving the mononuclear metal centre contributes significantly
to the overall stabilization of the complex, as well as to electronic
communication between the M2 and M’ centres.
Scheme 3
Acknowledgements
The authors thank Ms Belinda Hurley and Prof. Richard
McCreery for their assistance in obtaining the Raman spectra
and the National Science Foundation for financial support.
We also thank the Ohio Supercomputing Centre for providing
resource units in order to carry out the computational studies
on the model compounds.
References
1 C. Creutz and J. Taube, J. Am. Chem. Soc., 1969, 91, 3988.
2 C. Creutz, Prog. Inorg. Chem., 1983, 30, 1.
3 S. F. Nelson, Chem. Eur. J., 2000, 6, 581.
4 B. S. Brunschwig, C. Creutz and N. Sutin, Chem. Soc. Rev., 2002, 31,
168.
5 B. E. Bursten, M. H. Chisholm, R. J. H. Clark, S. Firth, C. M. Hadad,
A. M. MacIntosh, P. J. Wilson, P. M. Woodward and J. M. Zaleski,
J. Am. Chem. Soc., 2002, 124, 3050.
6 F. A. Cotton, J. P. Donahue and C. A. Murillo, J. Am. Chem. Soc.,
2003, 125, 5436.
7 D. E. Richardson and H. Taube, Inorg. Chem., 1981, 20, 1278.
8 R. H. Cayton, M. H. Chisholm, J. C. Huffman and E. B. Lobkovsky,
J. Am. Chem. Soc., 1991, 113, 8709.
9 M. H. Chisholm, R. J. H. Clark, J. Gallucci, C. M. Hadad and N. J.
Patmore, J. Am. Chem. Soc., 2004, 126, 8303.
10 R. H. Cayton, M. H. Chisholm, E. E. Putilina and K. Folting,
Polyhedron, 1993, 12, 2627.

11 M. H. Chisholm, R. J. H. Clark, C. M. Hadad and N. J. Patmore,
Chem. Commun., 2004, 80.
12 F. A. Cotton, C. Y. Liu, C. A. Murillo, D. Villagran and X. Wang, 16
J. Am. Chem. Soc., 2003, 125, 13564. 17
13 A. G. Brignole and F. A. Cotton, Inorg. Synth., 1972, 13, 81. 18
14 D. J. Santure, J. C. Huffman and A. P. Sattelberger, Inorg. Chem., 19
1985, 24, 371.
15 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. 20
Robb, J. R. Cheeseman, J. Montgomery, J. A. T. Vreven, K. N. Kudin,
J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. 21
Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Pettersson, M. 22
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, 23
Published on 15 April 2005. Downloaded by UNIVERSIDAD SAO PAULO on 5/3/2020 9:52:50 PM.
X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, J. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratman, O. Yazyev, A. J. Austin, R. Cammi, C. 24
Pomelli, J. W. Ochterski, P. Y. Ayala, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuk, 25
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, 26
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- 27
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
View Article Online
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople,
Gaussian03, Revision B.04, Gaussian Inc., Pittsburgh, PA, 2003.
A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
A. D. Becke, Phys. Rev. A, 1988, 38, 3098.
C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
W. J. Hehre, L. Radom, P. v. R. Schleyer and J. A. Pople, Ab initio
Molecular Orbital Theory, John Wiley & Sons, New York, 1986.
D. Andrae, U. Hauessermann, M. Dolg and H. Preuss, Theor. Chim.
Acta, 1990, 77, 123.
Gaussview 3.0, Gaussian Inc., Pittsburgh, PA, 2003.
R. Bauernschmitt and R. Ahlrichs, Chem. Phys. Lett., 1996, 256,
454.
M. E. Casida, C. Jamorski, K. C. Casida and D. R. Salahub, J. Chem.
Phys., 1998, 108, 4439.
R. E. Stratman, G. E. Scuseria and M. J. Frisch, J. Chem. Phys., 1998,
109, 8218.
S. I. Gorelski, Swizard Program, 2003, http://www.sg-chem.net/.
M. H. Chisholm, B. D. Pate, P. J. Wilson and J. M. Zaleski, Chem.
Commun., 2002, 1084.
B. E. Bursten, M. H. Chisholm, C. M. Hadad, J. Li and P. J. Wilson,
Isr. J. Chem., 2001, 41, 187.
Dalton Trans., 2005, 1852–1857 1857