+ All Categories
Home > Documents > EE-606: Solid State Devices Lecture 9: Fermi-Dirac - nanoHUB

EE-606: Solid State Devices Lecture 9: Fermi-Dirac - nanoHUB

Date post: 18-Mar-2022
Category:
Upload: others
View: 4 times
Download: 0 times
Share this document with a friend
24
www.nanohub.org NCN EE606: Solid State Devices EE606: Solid State Devices Lecture 9: FermiDirac Statistics Muhammad Ashraful Alam [email protected] Alam ECE606 S09 1
Transcript

www.nanohub.orgNCN

EE‐606: Solid State DevicesEE‐606: Solid State DevicesLecture 9: Fermi‐Dirac Statistics

Muhammad Ashraful [email protected]

Alam  ECE‐606 S09 1

Carrier Density

Carrier number = Number of states x filling factor

Chapters 2‐3 Chapter 4

Alam  ECE‐606 S09 2

Outline

1) Rules of filling electronic states1) Rules of filling electronic states 

2) Derivation of Fermi‐Dirac Statistics: three techniques

)3) Intrinsic carrier concentration

4) Conclusion

Reference: Vol 6 Ch 4 (pages 96‐105)

Alam  ECE‐606 S09 3

Reference: Vol. 6, Ch. 4 (pages 96 105) 

E‐k diagram and Electronic States

Energy‐Band Density of States

EE

2 3

22

−*

cm m* E E2 32π

E

E3

E1

E2

kaπ

− g(E)

Alam  ECE‐606 S09 4

Rules for filling up the States

E3E2

3

E11

Pauli Principle:  Only one electron per state

Total number of electrons is conserved T iiN N=∑

Total energy of the system is conserved

T ii∑T i ii

E E N=∑

Alam  ECE‐606 S09 5

Outline

1) R l f l i l t i t t1) Rules of placing electronic states 

2) Derivation of Fermi‐Dirac Statistics: three techniques

3) Intrinsic carrier concentration

4) Conclusion

In 1926, Fowler studied collapse of a star to white dwarf by F‐D statistics, before Sommerfeld used the F‐D statistics to develop a theory of 

Alam  ECE‐606 S09 6

electrons in metals in 1927. Wikipedia has a nice article on this topic. 

Illustrative Example: 3 Energy Levels

T iiN N=∑ T i ii

E E N=∑NT=5 and ET=12

E=2

E=4

E=0

E 2

2035!

0!5!2!

1!27!

3!5

4!3

!W = • •

=

1225!

2!3!72!

1!1!

5!2!!420

W = • •

=

0415!

4!1!2!

0!27!

6!5

1!3

!W = • •

=

Alam  ECE‐606 S09 7

53= 420= 53=

Occupation Statistics

E=4

E=0

E=2

E=4

E=0

122 420W = 041 35W =203 35W =

(E) Choose the most 

probable configuration

W ( probable configuration.

Alam  ECE‐606 S09 8

2,0,3          1,2,2           0,4,1

Occupation Statistics

E=4 * 2f = E

E=0

E=2

E=4 3

*2

725

f

f

=

=

E

E=0

122 420W =*

1 21f =

f(E)

(E)

W (

Alam  ECE‐606 S09 9

2,0,3 1,2,2 0,4,1

For N‐states

Ni

lnWSi

!iSW =∏

2,0,3  1,2,2   0,4,1

5!2! 7!WR ll( )! !ci i i i

WS N N

=−∏ 203 0!51 ! 3!4!!2!

= • •WRecall. 

[ ]ln( ) ln ! ln( )! ln != − − −∑W S S N N

[ ]ln ( ) ln( ) ( ) ln− − − − + − − +∑ i i i i i i i ii i i ii

NS S S N N NS NS N SStirling approx.

[ ]ln( ) ln ! ln( )! ln !=∑ i i i ii

W S S N N

10

[ ]ln ( ) ln( ) ln= − − − −∑ i i i i i i i ii

S S S N S N N N

Optimization with Lagrange‐Multiplierh hChoose the most 

probable configuration.

lnWlnln( ) ii i

WW dNN

δ ∂=

∂∑configurations

i

ln 1ii i i i

S dN dN E dNN

α β⎡ ⎤⎛ ⎞

− − −⎢ ⎥⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦

∑ ∑ ∑

l 1iS E dNβ⎡ ⎤⎛ ⎞⎢ ⎥⎜ ⎟∑

T i iiE E N=∑

i i iiN⎢ ⎥⎝ ⎠⎣ ⎦∑ ∑ ∑

ln 1ii i

i i

S E dNN

α β⎛ ⎞

= − − −⎢ ⎥⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦

∑T ii

N N=∑

i∑

Alam  ECE‐606 S09 11

i

Final steps …

ln 1 0ii

S EN

α β⎡ ⎤⎛ ⎞

− − − =⎢ ⎥⎜ ⎟⎢ ⎥⎝ ⎠ E

1iN

iiN

β⎢ ⎥⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦

E

EF1( )1 α β+≡ =+

iE

i

Nf ES e

1f(E)

fmax(E)=1EF

1At , ( ) 02FF FE f EE Eα β= ≡ ⇒ + =

1( ) iNf E1

( )1( )

1 F

iE

iE

Nf ES eβ −

= =+

/A ( ) E k Tf A

( ) /1 BFE kE Te −=

+

Alam  ECE‐606 S09 12

Bk Tβ⇒ =/At , ( ) BE k T

BoltzmanE f E Ae−→ ∞ =

Derivation by Detailed Balance

E 0

E=2E=4

0 1 0 2 0 3 0 4( ) ( )[1 ( )][1 ( )]0

f E f E f E f EA⎧ ⎫− − −⎪ ⎪⎨ ⎬

E=0

0 1 0 2 0 3 0 4

0 3 0 4 0 1 0 2

( ) ( )[ ( )][ ( )]0

( ) ( )[1 ( )][1 ( )]f f f f

Af E f E f E f E

⎪ ⎪ =⎨ ⎬− −⎪ ⎪⎩ ⎭

1 E1 + E2 = E3 + E4 Only solution is …. 0 ( )

1( )1 β −=+ FE Ef E

e

Alam  ECE‐606 S09 13

Pauli Principle, energy, and number conservation all satisfied

Derivation by Partition Function

β β− − − −F Fiiii( E ) ( E )NE E Ne eP β− −= ≡∑ i i Fi ( E N E )

i

e ePe Z

E=2E=4

E=0 1β = Bk T

( )0000 0 β− − × F

i i i

E

state E N P

e Z( )1

00

11

0

1 β− − ×i FE E

e Z

e ZEi

Alam  ECE‐606 S09 14

Derivation by Partition Function1

( )0000 0 β− − × F

i i i

E

state E N P

e Z

1

( )111 1 β− − ×i FE Ee Z

( ) 1

0 1

=+Pf E

P PProbability that state is filled ….

E 0 1

1

− −

− −=i F B

i F B

( E E ) / k T

( E E ) / k T

e / Z/ Z / Z1

f(E)

EF

E

1 + i F B( E E ) / k T/ Z e / Z1

1 −= ( E E ) / k T

11/2f(E)

F

Alam  ECE‐606 S09 15

1 −+ i F B( E E ) / k TeEFE

1 )1/2

Few comments on Fermi‐Dirac Statistics

Applies to all spin‐1/2 particles

Information about spin is not explicit; multiply DOS by 2Information about spin is not explicit; multiply DOS by 2. May be more complicated for magnetic semiconductors. 

C l b i t ti ti l i l t dCoulomb‐interaction among particles is neglected,Therefore it applies to extended solids, not to small molecules 

Alam  ECE‐606 S09 16Lx

Outline

1) Rules of placing electronic states1) Rules of placing electronic states 

2) Derivation of Fermi‐Dirac Statistics: three techniques

3) I i i i i3) Intrinsic carrier concentration

4) Conclusion

Alam  ECE‐606 S09 17

Carrier Distribution 

( )1 f E ( ) ( )tE

∫DOS F‐D concentration

E

E

( )1 f E− ( ) ( )top

c

E

cEn g E f E dE= ∫E E

( )cg EFE

( )g E

cE( ) ( )cg E f E

( ) ( )1g E f E⎡ ⎤−⎣ ⎦Eυ( )g Eυ

( )f E

( ) ( )1g E f Eυ ⎡ ⎤⎣ ⎦

1

( ) ( )1bot

E

Ep g E f E dEυ

υ ⎡ ⎤= −⎣ ⎦∫

18

Electron Concentration in 3D solids

( ) ( )= ∫top

c

E

cEn g E f E dE

( )( )2 3

2 122 1 βπ −

−= ×

+∫top

F

* *E n n C

E EE

m m E EdE( )2 1 βπ +∫ FcE e

( )2 1∞ −∫

*Cn

*nm m

dEE E

( ) ( )2 3 1 β βπ − −+∫ c Fc cE E EE E dEe e

( ) ( )1 22 η η β= ≡ −FC CF EN E

3 22 * dmπ β ξ ξ∞⎛ ⎞

( ) ( )1 2 η η βπ c c FC CF EN E

Alam  ECE‐606 S09 19

( )2 1 20 1

22 nC

dF

emNh ξ η

π β ξ ξη −=

⎛ ⎞≡ ⎜ ⎟

⎝ ⎠ +∫

Boltzmann vs. Fermi‐Dirac Statistics

( ) ( )1 22 3ηη η βπ

= → − ≡ − >cC c C c C Fn N F N e if E E

ηce( )ηF

( ) ( )f

( )1 2 ηcF

( ) ( )cg E f E

( ) ( )1g E f Eυ ⎡ ⎤−⎣ ⎦FE

Alam  ECE‐606 S09 20

Effective Density of States

( )2 E Eβ( ) ( )1 2

2 3c FE EC c C c Fn N F N e if E Eβη β

π− −= → − >

( ) ( )fCN

( ) ( )cg E f E

( ) ( )1g E f Eυ ⎡ ⎤−⎣ ⎦ VNFE

FE

V

As if all states are at a single level E

Alam  ECE‐606 S09 21

As if all states are at a single level EC

Law of Mass‐Action

( )β− −= c FE En N e

( )β+

= C

E E

n N e

( )β+ −= v FE EVp N e

FE

( )β

β

− −× = c vE EC V

E

n p N N eβ−= gE

C VN N e

Alam  ECE‐606 S09 22

Fermi‐Level for Intrinsic Semiconductors

2 β

= = i

E

p nn

2βEFE

2 β−= gV

Ei C Nn eN

2β−=

gEi

F

V

i

Cn eE

NE

N≡F iE E

( ) ( )β β− − + −= ⇒ =c i v iE E E En p e eNN E

12 2β

= ⇒ =

= +Gi

V

VCn p e eEE ln

NN

N

N

3

Alam  ECE‐606 S09 23

2 2β CNk

2

Conclusions

We discussed how electrons are distributed in electronic states defined by the solution of Schrodinger equationstates defined by the solution of Schrodinger equation. 

Since electrons are distributed according to their energy,irrespective of their momentum states, the previously developed concepts of constant energy surfaces, density of states etc. turn out to be very useful.y

We still do not know where EF is for general semiconductors … If we did we could calculate electron concentrationIf we did, we could calculate electron concentration. 

Alam  ECE‐606 S09 24


Recommended