Beruflich Dokumente
Kultur Dokumente
Structure of Solids 1. Introduction 2. Types of Solids 3 Atomic 3. At i Bonding B di 4. Growth of Semiconductor Materials Reference D.A. Neamen, Semiconductor Physics and Devices (McGraw-Hill, Devices (McGraw-Hill 2003), 2003) Chapter 1 1.
1. Introduction
An IC comprises millions of interconnected transistors transistors, resistors resistors, capacitors capacitors, etc etc. Metals used for interconnects, contacts, and in the transistor structure structure. Examples: aluminium, copper, tungsten, polysilicon, metal silicides. Dielectrics Di l t i for f electrical l ti li insulation, l ti mechanical support and environmental protection; part of device structure (transistor, capacitor dielectric). dielectric) Examples: silicon dioxide, silicon nitride, various polymers. Semiconductors used to make the active device (transistors, LEDs, lasers). Examples: silicon, germanium, GaAs.
Intel Core i7 Processor 45nm process technology 731 million transistors
Metal
Dielectric insulation
Semiconductor
* Group III (B, Ga, In) and Group V (P, As, Sb) elements used to dope Si.
2. Types of Solids
Classification according to the size of the ordered region: Amorphous Order in the range of a few atoms or on molecular dimensions Single-crystal One large crystal making up entire volume of material. Regularity of atomic arrangement present throughout entire solid. Polycrystalline y y High degree of order over many atomic / molecular dimensions. Ordered single-crystal regions are called grains. Grains vary in size and orientation. Grains are separated by grain boundaries.
Symbol
Schematic
Amorphous oxide
Single-crystal silicon
3. Atomic Bonding
3.1 Covalent Bonding in Silicon Silicon (atomic number = 14) has 14 electrons occupying 3 shells: shell n = 1: 2 electrons (complete shell) shell n = 2: 8 electrons (complete shell) shell n = 3: 4 electrons (partially filled shell) Th The outermost t t shell h ll (or ( the th valence l shell h ll) is i only l partially ti ll filled, fill d with ith 4 out of 8 possible states being occupied by electrons. Having g 4 electrons in the valence shell p places Silicon in Group p IV of the periodic table. Other Group IV elements include carbon (C) and germanium (Ge).
300 mm wafer
Wafer Polishing
Band Theory
Band Theory 1. Energy Concepts 2. Formation of Energy Bands 3 Electrical 3. El t i l C Conduction d ti 4. Metals, Insulators, and Semiconductors Reference D.A. Neamen, Semiconductor Physics and Devices (McGraw-Hill, Devices (McGraw-Hill 2003), 2003) Chapter 3 3.
1. Energy Concepts
1 1 Potential & Kinetic Energy of an Electron 1.1 Electron is a negatively-charged particle with a charge of ( e) or ( q) where e (or q) = 1.602 1 602 x 10-19 C is the fundamental unit of charge.
Consider a stationary electron (i.e., no kinetic energy, EKE = 0) in vacuum i b in between t 2 plates l t of f a bi biased d parallel plate capacitor (LH plate = 0V, RH plate = +5V). Between the plates, the voltage increases linearly from 0V to +5V.
Platinum ( (Pt) ) Gold (Au) Tungsten (W) Al i i Aluminium (Al) Cesium (Cs)
3. Electrical Conduction
The energy band model of a solid has important implications: 1. An empty band cannot support electrical conduction there are no electrons in such a band to carry current. 2 A band that is fully occupied cannot support conduction either! A full 2. band has plenty of electrons, but it has no available states within the band for the electrons to move into. 3. For conduction to take place in a particular band, the band must have BOTH electrons AND empty states, i.e., the band is only partially occupied.
Bandgaps of selected semiconductors at 300K (Note: Bandgap varies slightly with temperature)
The Semiconductor in Equilibrium 1. Charge Carriers in Semiconductors 2. Doping of Semiconductors 3 Carrier 3. C i C Concentration t ti Reference D.A. Neamen, Semiconductor Semiconductor Physics and Devices (McGraw-Hill, 2003), Chapter 4.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors 1.1 Bond Model Purpose is to relate the energy band model d l with ith th the more elementary l t bond model which is less abstract. Silicon (Z = 14) ) has 4 valence electrons. In the bond model of silicon, silicon is bonded to 4 other silicon atoms by covalent bonds each covalent bond involves the sharing of a pair of electrons (one from each atom) atom). At T = 0K, all the valence electrons are involved in covalent bonds. No free electrons for electrical conduction. (Inner core electrons are tightly bound to the nucleus and are thus not mobile.) bil )
Silicon at T = 0K
The Semiconductor in Equilibrium Charge Carriers in Semiconductors At temperatures above 0K, some of the electrons have sufficient thermal energy to break free from the covalent bond. The free Th f electron l t are now mobile bil can t take k part t in i electrical l t i l conduction. d ti They are thus known as conduction electrons.
Conduction Electron
Silicon at T > 0K
The Semiconductor in Equilibrium Charge Carriers in Semiconductors The empty state left behind cannot in itself carry current, but it allows other electrons the opportunity of moving without any net input of energy into the system the energy to break a bond will be given given back back to the system when a new bond is formed. As another electron moves into the space, it leaves another behind it, etc.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Instead of seeing a successive movement of electrons, we can instead consider the motion of the space that is left behind. Th net The t charge h surrounding di th the space i is +e. The moving space may be treated as if it were a positively charged particle, called a hole.
The motion of the hole is independent of the electron that once occupied that state. Holes are thus independent carriers.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Comparing the bond model with the band model At T = 0K
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Comparing the bond model with the band model At T > 0K
Electron excited into CB is now mobile as there are unoccupied states available. il bl
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Comparing the bond model with the band model At T > 0K
Available state in VB now allows successive electron movement in VB. S Seen as the th motion ti of fah hole l i in th the VB VB.
The conduction band edge, EC, is the lowest energy level in the conduction band. Electrons in the CB normally reside at energy levels around EC as these are the lowest energy states in the CB. The valence band edge, EV, is the highest electron energy level in the valence band band. Holes in the VB normally reside at energy levels around EV as these are the lowest energy states for holes in the VB. The bandgap Eg = EC EV, is the minimum energy required to create an electron-hole pair (EHP). Note that thermal excitation across the bandgap always creates electrons and holes in pairs.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors 1.2 Carrier Properties 1.2.1 Charge Charge of electron = e = 1.602 x 10-19 C Charge of hole = + e = + 1.602 x 10-19 C
1.2.2 Effective Mass Consider electrons first. Electrons in CB are not entirely free as they are travelling in the background g p potential of the lattice atoms (nuclei, electron shells)
The Semiconductor in Equilibrium Charge Carriers in Semiconductors 1.2.2 Effective Mass Electrons experience internal forces Fint due d t to the th b background k d potential t ti l of f the lattice atoms. When an external force Fext ( (e.g. g an electric field) is applied to the conduction electrons, the equation of motion for the electrons can be written as:
Fext + Fint = m0 a
The Semiconductor in Equilibrium Charge Carriers in Semiconductors To treat the electrons as if they were truly free particles that respond to the external force (only), we wish to hide the internal force term in the equation of motion. We can re-write the equation of motion as:
* Fext = mn a
The effective mass now includes and takes care of the effects of the internal lattice forces on the electron. Electrons residing near the bottom of the CB behave like free electrons electrons, but * with an effective mass given by mn .
M Motion ti of f car is i governed db by th the f force you exert t (th (the external t lf force), ) and d th the true mass of the car.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors An analogy pushing a car (episode 2)
Th The internal i t lf forces (the (th applied li d b brakes) k ) opposes your exerted t df force and d makes the car feel heavier to you. It has a larger effective mass, as far as you are concerned. For the same external force, the car accelerates slowly.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors An analogy pushing a car (episode 3)
Th The internal i t lf forces (the (th engine) i ) assists i t your exerted t df force and d makes k th the car feel lighter to you. It has a smaller effective mass, as far as you are concerned. For the same external force, the car accelerates more.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors An analogy pushing a car (episode 3)
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Effective mass for holes Motion of holes at the top of the VB results from the collective effects produced by electrons moving in a nearly full band band. The motion of a hole is thus also affected by the internal lattice forces. The equation of motion for a hole is
Fext = m* pa
Material Si Ge GaAs
m* p m0
0.49 0.28 0.5
The Semiconductor in Equilibrium Charge Carriers in Semiconductors 1.2.3 Energy By convention, energy band diagrams are energy diagrams for negatively charged electrons electrons, with energy increasing from bottom to top top. For electrons in the CB, the lowest energy states are at the bottom of the CB, and hence electrons tend to stay in these low energy states near EC. For holes, which are positively charged, their energy increases from the top of a band to the bottom of a band. H Holes l i in the h VB are f found di in the h lower-energy l states for f holes, h l which hi h are at the top of the VB near EV.
The Semiconductor in Equilibrium Charge Carriers in Semiconductors 1.2.3 Carrier motion in an electric field Compare an electron in vacuum to an electron in the CB of silicon:
The Semiconductor in Equilibrium Charge Carriers in Semiconductors Under the influence of an external electric field, holes move in the opposite direction to electrons. Currents are in the same direction (due to opposite charge direction of current is direction of flow of +ve charge). Note that the energy for holes increase from the top of the VB to the t e bottom botto o of t the e VB.
Electrons
Holes
Occupy states at the bottom of the Occupy states at the top of the conduction band (CB). valence band (VB). Charge g Effective mass Energy e( (negative) g ) Effective mass at the bottom of * the CB = mn Lowest energy states at the bottom of the CB. Energy increases upwards. + e (p (positive) ) Effective mass at the top of the VB * = mp Lowest energy states at the top of the VB. Energy increases downwards.
The Semiconductor in Equilibrium 1. Charge Carriers in Semiconductors 2. Doping of Semiconductors 3 Carrier 3. C i C Concentration t ti Reference D.A. Neamen, Semiconductor Semiconductor Physics and Devices (McGraw-Hill, 2003), Chapter 4.
2. Doping of Semiconductors 2 1 Undoped Semiconductors 2.1 2.2 Doped Semiconductors 2.3 Majority and Minority Carriers 2.4 Charge Neutrality 2.5 Semiconductor Doping Techniques 2.S Equation Summary
where ni is known as the intrinsic carrier concentration. The subscript 0 is used to denote thermal equilibrium conditions.
Thermal Equilibrium At thermal equilibrium, the energy in a system is everywhere equalized and evenly distributed, and is at equilibrium with its ambient temperature. There is no energy input (or output) from heat, voltage, or optical excitation. In this course, we specifically mean that there is no optical excitation (light shining on the semiconductor), and no net motion of charge (no net current t flow). fl )
At 300K, the average thermal energy of an electron is only about 3/2kT ~ 0 04 eV 0.04 eV, so statistically very few electrons have sufficient energy to jump across the bandgap. Actually the fraction of electrons is very small for silicon there are 2 x 1023 valence electrons / cm3 . Hence only around 1 out of 1013 valence electrons is excited across the bandgap!
B = Boron
P = Phosphorus
Ga = Gallium
As = Arsenic
In = Indium
Sb = Antimony
We can locate the position of the donor state in the energy band diagram (the energy of the loosely-bound loosely bound electron electron, before ionization) or the donor level. Since the freed electron is now a conduction electron, it travels at the bottom of the conduction band, at the conduction band edge EC. If the energy required to free the electron (the ionization energy) is 0.049eV (for As in Si), then the donor level must be 0.049eV below EC.
The As atom is known as a donor in the Si lattice since it donates a (conduction) electron when it is ionized ionized. Other Group V elements also act as donors in Group IV Si and Ge. An otherwise pure semiconductor that has impurities introduced (donors or acceptors) is said to be doped. The impurities are called dopants. The ionization energy EC ED required to ionize the donor depends on the dopant as well as the host semiconductor. Ionization energy (eV) Element Phosphorus (P) Arsenic (As) Antimony y( (Sb) ) Z 15 33 51 Type Donor Donor Donor Ge 0 0120 0.0120 0.0127 0.0096 Si 0 044 0.044 0.049 0.039
At T = 0K, the empty state resulting from the incomplete 4th bond is not free to move it is bound to the parent B atom atom, and hence it is not a hole hole. You need an ionization energy of ~ 0.05 eV in order to create a (mobile) hole. The reason is that a Si-Si bond is stronger than a B-Si bond. To break a SiSi bond, you need Eg ~1.1eV. When a B-Si bond is formed, you only get back (Eg - 0.05eV).
The amount of energy required to create a hole is of the order of 10s of meV in Silicon Silicon. As the temperature increases, more and more of the B atoms become ionized. At T > ~150K, practically all the B atoms in Si are ionized. The hole created can now conduct electricity. The B atom is now ionized and becomes an B ion, but this ion is covalently bonded to the Si S lattice. It is not a charge carrier, but a fixed f ve charge.
-
We can locate the position of the acceptor state in the energy band diagram (the energy of the empty state state, before ionization) or the acceptor level. The created hole travels at the top of the valence band, at the valence band edge EV. If the energy required to create the hole (the ionization energy) is 0.045eV (for B in Si), then the acceptor level must be 0.045eV above EV.
If there are NA (cm-3) acceptors in the lattice, they will contribute NA holes to the valence band when they are fully ionized. Note that an equal number of negatively charged acceptor ions are left behind, so the material remains electrically neutral. As with donors, at low temperatures, not all the acceptors are fully ionized in the freeze-out range.
2.2.3 Extrinsic semiconductors Carrier concentrations ( (electrons or holes) ) introduced by y dopants p are usually y much higher than the intrinsic carrier concentration of the semiconductor. e.g., n = 1015 cm-3 for Si doped with 1015 cm-3 donors, compared with 0 cm-3 3 at T = 300K. ni ~ 1010 300K The electron & hole concentrations of the Si are then no longer controlled by properties p of Si, , but by y the dopants. p the intrinsic p The semiconductor is then said to be an extrinsic semiconductor.
Donors Valence electrons For Group IV host Carrier created Ionized dopant Energy level of d donor/acceptor / t state t t More valence electrons than host semiconductor. semiconductor Group V elements Electrons Positive, fixed charge Slightly g y below the CB
Acceptors Fewer valence electrons than host semiconductor. semiconductor Group III elements Holes Negative, fixed charge Slightly g y above the VB
Consider donor doping. When you dope a semiconductor with donors, and if the temperature is sufficiently high for complete ionization, then each donor will contribute 1 electron to the conduction band band. Usually the doping levels are such that ND >> ni, so the electron concentration is n ND. Hence if ND = 1015 cm-3, then n = 1015 cm-3 >> ni. Q: What happens to the hole concentration?
n0 = ni p0 = ni
ni ~ 1010 cm 3
Electrons from ED fill up the low energy states (holes) in VB. Remaining electrons from ED thermally excited to CB
Most of the holes in the VB annihilated. Large concentration of electrons (mainly from the donors, n ~ ND) in the CB.
p0 n0 = ni2
Obvious for an intrinsic semiconductor where p0 = n0 = ni Also holds for extrinsic semiconductors. If you increase the electron concentration then the hole concentration will decrease and vice versa concentration, versa. Example: As (gp V, donor) doped Si
+ If we e dope S Si with t ND = 10 016 c cm-3, then t e (to be s shown o later): ate ) n0 N D = 10 016 c cm 3 + where N D = conc. of ionized donors (assumed fully ionized at room temp.).
1016
Only ~22,000 holes per cm3, or ~12 orders of magnitude fewer holes compared to electrons!
mobile holes
p0 n0 = ni2
ni2 + n0 = + ND n0
2 + n0 ND n0 ni2 = 0
n0 =
+ ND
+2 ND
+ 4ni2
We only take the +ve solution as negative values for electron concentration make no physical sense.
+ If N D >> ni (true for most practical cases of doping) then we can neglect the 4ni2 term in the square root, and write: + n0 N D
and
p0 = ni2 n0
p0 n0 = ni2
ni2 + NA = p0 p0
2 p0 NA p0 ni2 = 0
p0 =
NA
2 NA
+ 4ni2
We only take the +ve solution as negative values for hole concentration make no physical sense.
If N A >> ni (true for most practical cases of doping) then we can neglect the 4ni2 term in the square root, and write: p0 N A
and n0 = ni2 p0
and n0 = ni2 p0
and
p0 = ni2 n0
+ ND > If N A /> / ni then the exact values for the carrier concentrations need to
To see why only net concentration of ionized dopants matters (thought expt expt.) )
Undoped silicon @ TE
Electrons from CB occupy low-energy states at EA. Electrons from VB excited into remaining state at EA
Holes become maj. carrier, p0~NA. Electrons become min min. carrier, n0 << ni.
Now add ND (>NA) dopants. NA electrons from ED occupy low-energy states (holes) in VB.
The semiconductor is then said to be compensated. If the concentrations of the ionized dopants are not equal, the material is said to be partially compensated. An important consequence of this compensation is that it is possible to change, for example, a p-type semiconductor into an n-type semiconductor by y doping p g the original g p p-type yp semiconductor with more donors than there were original acceptors. This trick can be repeated several times, so that the n-type semiconductor can be changed back into a p-type semiconductor by adding even more acceptors. However, the total concentration of dopants increases each time the material is changed from f p-type to n-type and vice versa. This can adversely affect other properties of the semiconductor.
y dope p a volume with ND = 5x1015 cm-3 donors. We now selectively As there are more donors than acceptors, the selected volume is now ntype with an electron concentration of n = ND NA1 = 3x1015 cm-3.
p0 =
ni2 n0
(1.5 10 )
3 1016
10 2
= 7.5 103 cm 3
Depending on the ion energy, the ions will come to a stop after a certain range of travel in the material material, be it the mask or the silicon silicon.
Simulated dopant distribution after implantation through oxide mask and metallization
Moreover, the implanted atoms are not electrically active unless they are in substitutional sites in the silicon lattice lattice. Hence an annealing process is needed after ion implantation to allow atomic diffusion to take place to repair the damage and activate the dopants.
Anneal
20/07/2010
J Thong
107
Source: www.ele.uva.es/~simulacion/MD.htm
where
(c) Majority carrier concentration (i) For n-type material, where N D > N A
+ n0 N D NA
+ NA >> ni provided N D
The Semiconductor in Equilibrium 1. Charge Carriers in Semiconductors 2. Doping of Semiconductors 3 Carrier 3. C i C Concentration t ti Reference D.A. Neamen, Semiconductor Semiconductor Physics and Devices (McGraw-Hill, 2003), Sect. 3.5, Chapter 4.
3. Carrier Concentration 3.1 Density y of States 3.2 Fermi-Dirac Distribution 3 3 Carrier Concentration in a Band 3.3 3.4 Electron and Hole Concentrations 3 5 Law of Mass Action for Semiconductors 3.5 3.6 Position of Fermi Level 37C 3.7 Carrier i C Concentrations t ti at t Hi High hT Temperature t 3.8 Fermi Level Considerations 3.S S Equation Summary S
g(E).dE is the number of states (per unit volume) of the material at energy level E in an energy interval of dE.
The total number of states in the band (per unit volume) is g ( E ).dE integrated from the bottom to the top of the band.
fF (E) =
1 E EF 1 + exp kT
where f F ( E ) is the probability of a state at energy E being occupied by an electron at thermal equilibrium. k is the Boltzmann constant (1 381x10-23 J/K), (1.381x10 J/K) and T is the absolute temperature temperature.
EF is the Fermi level and is defined as the energy level where the probability of a state being g occupied p by y an electron is 0.5.
1 E EF 1 + exp kT
We see that W th t all ll th the states t t below b l the th Fermi F i Level L l are occupied i d (fF(E) = 1), 1) while those above are empty (fF(E) = 0). Note that there are NO states in the forbidden band, , so there are no electrons there (although the probability of occupancy of a state here, if it had existed, would be 1).
Some electrons S l t near th the F Fermi il level lh have sufficient ffi i t th thermal l energy t to occupy states higher than the Fermi level, while leaving behind now unoccupied states below the Fermi level. Hence there is now a gradual transition of fF(E) from full occupancy (1) to zero occupancy (0) as the energy goes from below to above EF. Java applet
The product of g(E) and fF(E) then gives us the number of electrons per unit energy (per unit volume) at energy E,
N (E ) = g (E ) f F (E )
Hence in the energy interval between E and (E+dE), there are N(E).dE electrons (per unit volume). The total number of electrons per unit volume (or electron concentration) is thus:
n = N ( E )dE = g (E ) f F (E )dE
EB EB ET ET
(a) the probability of occupancy of the states at 0.05eV and 0.1eV relative to EF, and (b) the corresponding number of electrons at these energy levels.
What about holes? The probability of a hole occupying a state in the valence band at energy E is simply (1 f F ( E ) ), since a hole is an unoccupied electron state in the VB VB. So the distribution of holes in the VB is
N p ( E ) = (1 f F ( E ) ) g p ( E )
We see that most of the electrons and holes occupy states at energy levels close to their respective band edges.
where NC is known as the effective density of states for electrons in the conduction band. At room temperature (300K), NC is about 2.8 x 1019 cm3 for silicon. Note that NC has a temperature dependence of
NC T 3 / 2
E EV p0 = NV exp F kT
where NV is known as the effective density of states for holes in the valence band. At room temperature (300K), NV is about 1.04 x 1019 cm3 for silicon. Note that NV also has a temperature dependence:
NV T 3 / 2
E EV p0 = NV exp F kT
2 points to note: The equations are only for carrier concentrations at thermal equilibrium. These derived equations are only accurate provided the Fermi level should not be too close to the conduction band (for n0 expression) and valence band (for p0 expression). Usually o.k. unless we are dealing with very heavy doping (> ~1018 /cm3).
we see that the electron concentration n0 is large g when ( EC EF ) is small (i.e., the Fermi level is close to the conduction band edge). At the same time, the hole concentration diminishes. C Conversely l f from
E EV p0 = NV exp F kT
the hole concentration p0 is large when ( EF EV ) is small (i.e., the Fermi level is close to the valence band edge). At the same time, the electron concentration diminishes.
The electron and hole concentrations are comparable when the Fermi level is in the middle of the bandgap (the difference being due to the slightly different values for NC and NV.
ermi i level l l in i the th middle iddl of f the th bandgap b d (b) Fermi F i level l l closer l to t CB edge. d Java applet
The right-hand-side is dependent only on: o the temperature, and o the h ki kind d of f semiconductor i d (b (bandgap d Eg, effective ff i DOS NC & NV). )
It does not depend on doping / Fermi level position, and hence applies to both extrinsic and intrinsic semiconductors. At a given temperature this becomes the Law of Mass Action,
p0 n0 = ni2 = constant
E + EV kT N C kT n0 ln + ln E Fi = C 2 2 NV 2 p0
where EFi is the intrinsic Fermi level. The term involving the effective DOS is quite small about -0.013eV in Si at 300K so the intrinsic Fermi level is almost in the middle of the bandgap (just below).
For n-type semiconductor n0 >> p0, & EF goes above EFi towards the CB. For p-type p type semiconductor p0 >> n0, & EF goes below EFi towards the VB VB. If the material is doped more heavily, EF goes up / down further.
E EFi ni = NC exp C kT
Dividing the two equations, we get
E EF NC exp C n0 kT = exp EF EFi n = n exp EF EFi = i 0 ni kT kT EC EFi NC exp kT
In n++ material, I t i l th the conduction d ti b band di is partially ti ll occupied, i d while hil i in p++ material, t i l th the top of the VB is empty, also giving you a partially occupied band. Hence degenerately-doped semiconductors behave like metals in electrical conduction.
If the net dopant concentration (in this case ND) is >> ni,
+ then n0 N D = N D
E EF n0 = NC exp C kT NC EC EF = kT ln n 0 NC EC EF = kT ln N D
Q: Calculate the Fermi level position for ND = 5x1016 cm-3 for Si at 300K. (NC = 2.8x1019cm-3)
We note that the effective DOS DOS, NC and NV, have a T3/2 dependence. dependence So the intrinsic carrier concentration is related to temperature via
ni T
3/ 2
Eg exp 2kT
The bandgap Eg is weakly dependent on temperature as the material expands d or contracts, t t but b t this thi i is small ll relative l ti t to th the T in i the th denominator. d i t
Room Temp.
At high temperature, ni can exceed the dopant concentration of a doped semiconductors, and the approximation of N A N D >> ni may no longer hold.
em = Evac EF
es = Evac EF
Clearly, the workfunction will depend on the position of the Fermi level If the semiconductor is n level. ntype, then its workfunction will be smaller than if it were p-type
e s = Evac EC
The workfunction can thus be determine if the position of the Fermi level relative to the conduction band edge is known:
es = e s + (EC EF )
Element Silicon, Si Germanium, Ge Gallium Arsenide, GaAs Electron affinity (eV) 4.01 4.13 4.07
fF (E) =
1 E EF 1 + exp kT
(b) Electron and hole concentrations (at thermal equilibrium) as function of Fermi level position:
E EF E EFi n0 = NC exp C or n0 = ni exp F kT kT E EV p0 = NV exp F or kT E EF p0 = ni exp Fi kT