Handout 18 Phonons in 2D Crystals: Monoatomic Basis and Diatomic Basis

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Handout 18

Phonons in 2D Crystals: Monoatomic Basis and Diatomic Basis

In this lecture you will learn:

• Phonons in a 2D crystal with a monoatomic basis


• Phonons in a 2D crystal with a diatomic basis
• Dispersion of phonons
• LA and TA acoustic phonons
• LO and TO optical phonons

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Phonons in a 2D Crystal with a Monoatomic Basis


General lattice vector:
  
Rnm  n a1  ma2

y Nearest-neighbor vectors:
 
n1  axˆ n2  ayˆ
  
a2 n3   axˆ n4   ayˆ
x

a1 Next nearest-neighbor vectors:
 
p1  axˆ  ayˆ p2   axˆ  ayˆ
 
p3   axˆ  ayˆ p4  axˆ  ayˆ
Atomic displacement vectors:
Atoms, can move in 2D therefore atomic displacements are given by a vector:

 

u x Rnm , t 
 
 
u Rnm , t   
 

u y Rnm , t 

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

1
Vector Dynamical Equations
   
  
u R2 , t  u R1  m, t 

  
m  R2  R1 If the nearest-neighbor vectors are known then
 the dynamical equations can be written easily.
m

 
u R1, t  ˆ  
m
m

Vector dynamical equation:


 
M

d 2u R1, t     u R , t   u R , t  . mˆ  mˆ    u R  m , t   u R , t  . mˆ  mˆ
2 2 1 1 1
dt

Component dynamical equation:


To find the equations for the x and y-components of the atomic displacement, take
the dot-products of the above equation on both sides with x̂ and ŷ , respectively:

     u R

d 2u x R1, t
M 2 1

 
 
ˆ m
m , t  u R1, t . m  
ˆ . xˆ 
dt


d 2u y R1, t     u R 
  u R1, t  . mˆ  mˆ .yˆ 

M 2 1  m, t
dt
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Vector Dynamical Equations for a 2D Crystal


General lattice vector:
1   
Rnm  n a1  ma2
2 Nearest-neighbor vectors:
 
n1  axˆ n2  ayˆ
y  
n3   axˆ n4   ayˆ
 Next nearest-neighbor vectors:
a2  
x p1  axˆ  ayˆ p2   axˆ  ayˆ
  
a1 p3   axˆ  ayˆ p4  axˆ  ayˆ

M

 
d 2u Rnm , t    u Rnm  n j , t   u Rnm , t  . nˆ j  nˆ j

summation
2 1 over 4 nn
dt j 1,2,3,4

   u Rnm  p j , t   u Rnm , t  . pˆ j  pˆ j
     summation
 2
j 1,2,3,4 over 4 next
nn

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

2
Dynamical Equations

M

 
d 2u Rnm , t     u Rnm  n j , t   u Rnm , t  . nˆ j  nˆ j
2 1
dt j 1,2,3,4

   u Rnm  p j , t   u Rnm , t  . pˆ j  pˆ j
    
 2
j 1,2,3,4

If we take the dot-product of the above equation with x̂ we get:

    u R , t   u R  n , t    u R , t   u R  n , t 



d 2u x Rnm , t
M 2 1 x nm x nm 1 1 x nm x nm 3
dt

2
2
u x Rnm , t   u x Rnm  p1, t   22 uy Rnm , t   uy Rnm  p1, t 

2
2
u x Rnm , t   u x Rnm  p2 , t   22 uy Rnm , t   uy Rnm  p2 , t 
2
u x Rnm , t   u x Rnm  p3 , t   22 uy Rnm , t   uy Rnm  p3 , t 
   

2

2
2
u x Rnm , t   u x Rnm  p 4 , t   22 uy Rnm , t   uy Rnm  p4 , t 

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Dynamical Equations

M

 
d 2u Rnm , t     u Rnm  n j , t   u Rnm , t  . nˆ j  nˆ j
2 1
dt j 1,2,3,4

   u Rnm  p j , t   u Rnm , t  . pˆ j  pˆ j
    
 2
j 1,2,3,4

If we take the dot-product of the above equation with ŷ we get:

    u R , t   u R  n , t    u R , t   u R  n , t 



d 2u y Rnm , t
M 2 1 y nm y nm 2 1 y nm y nm 4
dt

2
2
uy Rnm , t   uy Rnm  p1, t   22 u x Rnm , t   u x Rnm  p1, t 

2
2
uy Rnm , t   uy Rnm  p2 , t   22 u x Rnm , t   u x Rnm  p2 , t 

2
2
uy Rnm , t   uy Rnm  p3 , t   22 u x Rnm , t   u x Rnm  p3 , t 

2
2
uy Rnm , t   uy Rnm  p4 , t   22 u x Rnm , t   u x Rnm  p 4 , t 

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

3
Solution of the Dynamical Equations

Assume a wave-like solution of the form:

 
 
u x Rnm , t  u x q  i q . R
 

u Rnm , t   
u

R , t


 
u q
 e
 
nm e  i  t

 y nm   y 
Then: 
 
 
u Rnm  n j , t   

 

u x Rnm  n j , t  u x q  i q . R nm  n j   i  t
       e e


u y Rnm  n j , t  u y q  

i q . n j u x q  i q . Rnm  i  t
  
e u q  e e
 y 
 
i q . nj  
e u Rnm , t  
We take the above solution form and plug it into the dynamical equations

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Dynamical Matrix and Phonon Bands


 2  qx a 
 
 41 sin  2   2 2 1  cosq x a  cos q y a  2 2 sinq x a  sin q y a  
 u x q  
u x q 
      2M u q 
u q 


 
2 2 sinq x a  sin q y a
 qy a 
41 sin2   
  2 2 1  cosq x a  cos q y a   y
 
  y 
 2 

Compare with the standard form:


 
 u x q  M 0  u x q 
D q      2  0 M u q 
u y q     y 
Solutions:
1  200 N/m
 2  100 N/m
M  2x10-26 kg
X M

2
a 

FBZ
2
a

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

4
Transverse (TA) and Longitudinal (LA) Acoustic Phonons

 2 

2
q a
 41 x    2 q x a   q y a
2
 
2 2 q x a  q y a   
u x q 

u x q 
  2   2
2 u q    M u q 

 
2 2 q x a  q y a   qy a 
41 
   2 q x a   q y a
2
 2 


 y   y 
  2  

Case I: q x  0 , q y  0

1   2 u x q x   1
LA q x   qx a u q   A   LA
M  y x  0 
Longitudinal acoustic phonons: atomic motion in TA
the direction of wave propagation

2 u x q x  0 
TA q x   qx a u q   A  
M  1 M
 y x 
Transverse acoustic phonons: atomic motion in the
direction perpendicular to wave propagation  X

FBZ
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Transverse (TA) and Longitudinal (LA) Acoustic Phonons



 2 

2
q a
 41 x    2 q x a   q y a
2
 
2 2 q x a  q y a   u q  
  2    x     2M u x q 
2  u y q  u q 

 
2 2 q x a  q y a   qy a 
41 
   2 q x a 2  q y a  2 

 y 
  2  

LA
Case II: q x  0 , q y  0 q x  qy  q

1  4 2 u x q  1 LA
LA q   qa u q   A   LA TA
M  y  1
TA
Longitudinal acoustic phonons: atomic motion in
the direction of wave propagation TA

1 u x q  1
TA q   qa u q   A  
M  y    1 M
Transverse acoustic phonons: atomic motion in the
direction perpendicular to wave propagation
 X

FBZ
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

5
Transverse (TA) and Longitudinal (LA) Acoustic Phonons

In general for longitudinal acoustic phonons


LA
near the zone center:

u x q  A q x  LA
u q    q 
 y  q  y LA TA
TA
And for transverse acoustic phonons near the
TA
zone center:

u x q  A   q y 
u q     
 y  q  qx 
In general, away from the zone center, the LA phonons are not entirely longitudinal
and neither the TA phonons are entirely transverse

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Transverse (TA) and Longitudinal (LA) Acoustic Phonons

LA
TA

TA and LA

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

6
Periodic Boundary Conditions in 2D
General lattice vector: General reciprocal lattice vector inside FBZ:
     
Rnm  n a1  ma2 q  1 b1   2 b2   1 2  1 ,  2  1 2
Our solution was:
 
 
u x Rnm , t  u x q  i q . R

 
u Rnm , t    
 
 e
 
nm e  i  t

u y Rnm , t  u y q 
Periodic boundary conditions for a lattice of N1xN2 primitive cells imply:
 
u x q  i q . R nm  N1a1   i  t   u x q  i q . R nm  i  t

  
u Rnm  N1a1, t     e e  u Rnm , t    e  e 
u y q  u y q 
 
 e i q . N1a1  1
 
 q . N1a1  m1 2  where m1 is an integer
 1 1 
 2 1 N1  m1 2  where -  1  b2 
 2 2
b1
m1  N1 N1 
 1   where - m
N1  2 2 FBZ
Similarly: m2  N2 N
2   where -  m2  2
N2  2 2
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Counting Degrees of Freedom


In the solution the values of the phonon wavevector are dictated by the periodic
boundary conditions:
  
q  1 b1   2 b2

m1  N1 N
1   where - m 1
N1  2 2
FBZ
m2  N2 N
2   where -  m2  2
N2  2 2
There are N1N2 allowed wavevectors in the FBZ
(There are also N1N2 primitive cells in the crystals)

There are N1N2 phonon modes per phonon band

Counting degrees of freedom:

• There are 2N1N2 degrees of freedom corresponding


to the motion in 2D of N1N2 atoms

• The total number of different phonon modes in the


two bands is also 2N1N2

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

7
Phonons in a 2D Crystal with a Diatomic Basis
  
Rnm  n a1  ma2

Atomic displacement vectors: 2


The two atoms in a primitive cell can move 1 y
in 2D therefore atomic displacements are
given by a four-component column vector: 
a2
 
  x
 u1x Rnm  d1, t
  
   

 


 u1 Rnm  d1, t   u1y Rnm  d1, t a1

 
 
 
   
u2 Rnm  d 2 , t  u2 x Rnm  d2 , t 2nd nearest-neighbor vectors (red to red):
 
   
u2 y Rnm  d2 , t n1  axˆ n2  ayˆ
 
n3   axˆ n4   ayˆ
1st nearest-neighbor vectors (red to blue):
 axˆ  ayˆ   axˆ  ayˆ 3rd nearest-neighbor vectors (red to red):
h1  h2   
2 2 p1  axˆ  ayˆ p2   axˆ  ayˆ
  axˆ  ayˆ  axˆ  ayˆ  
h3  h4  p3   axˆ  ayˆ p4  axˆ  ayˆ
2 2

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Diatomic Basis: Force Constants

1
y

a2
x

a1
The force constants between the 1st
2nd and 3rd nearest-neighbors need
plus
to be included (at least)
2
3


a2
x

a1
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

8
Diatomic Basis: Dynamical Equations
Dynamical equation for the red(1) atom:
 
   

d 2u1 Rnm  d1, t
M1 1   u2 Rnm  d1  h j , t   u1Rnm  d1, t  . hˆ j  hˆ j summation
over 4 1st nn
dt 2 j 1,2,3,4

   u1Rnm  d1  n j , t   u1Rnm  d1, t  . nˆ j  nˆ j


       summation
 2
j 1,2,3,4 over 4 2nd nn
   u1Rnm  d1  p j , t   u1Rnm  d1, t  . pˆ j  pˆ j
       summation
 3
j 1,2,3,4 over 4 3rd nn

Dynamical equation for the blue(2) atom:


 
  

d 2u2 Rnm  d 2 , t
M2 1   u2 Rnm  d2  h j , t   u2 Rnm  d2 , t  . hˆ j  hˆ j summation
over 4 1st nn
dt 2 j 1,2,3,4

   u2 Rnm  d 2  n j , t   u2 Rnm  d 2 , t  . nˆ j  nˆ j
       summation
 2
j 1,2,3,4 over 4 2nd nn
   u2 Rnm  d 2  p j , t   u2 Rnm  d 2 , t  . pˆ j  pˆ j
      
summation
 3
j 1,2,3,4 over 4 3rd nn

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Diatomic Basis: Dynamical Equations


Assume a solution of the form:
 
Rnm  d1, t  u1x q  e i q . d 1 
 
 u1x
 

 u1 Rnm

 
 d1, t   u1y

Rnm  d1, t    u1y q e i q . d1  e i q . Rnm e  i  t
 

u2 Rnm

 
 d 2 , t  u2 x

Rnm  d2 , t  u2 x q e i q . d2 
u2 y Rnm  d2 , t  u2y q e i q . d 2 

To get a matrix equation of the form:


 
 u1x q   M1 0 0 0   u1x q  
 u q   0 M 0 0   u q  

D q      2    1y  
1y 1
u2 x q  0 0 M2 0  u2 x q 
u q    u q 
 2y  0 0 0 M2   2y 

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

9
The Dynamical Matrix
 
 u1x q   M1 0 0 0   u1x q  
   0 M 
 u q  
 u1y q   0 0 
D q      1y  
1
  2
u2 x q  0 0 M2 0  u2 x q 
u q    u q 
 2y  0 0 0 M2   2y 

The matrix D q  is:

q a
21  4 2 sin2  x  
 2  
2 3 sinq x a  sin q y a  q a  qy a 
 21 cos x  cos 
q a  qy a 
21 sin x  sin 
 2   2 
 
2 3 1  cosq x a  cos q y a   2   2 

 qy a 
21  4 2 sin2   

2 3 sinq x a  sin q y a   2 
q a  qy a 
21 sin x  sin 
q a  qy a 
 21 cos x  cos 
 
2 3 1  cosq x a  cos q y a   2   2   2   2 

q a
21  4 2 sin2  x  
q a  qy a 
 21 cos x  cos 
q a  qy a 
21 sin x  sin   2  
2 3 sinq x a  sin q y a 
 2   2   2   2  
2 3 1  cosq x a  cos q y a  
 qy a 
21  4 2 sin2 
q a  qy a 
21 sin x  sin  q a  qy a 
 21 cos x  cos  2 3 sinq x a  sin q y a   2 
 
 2   2   2   2   
2 3 1  cosq x a  cos q y a 

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Diatomic Basis: Solution and Phonon Bands


For calculations: Optical
LO
 26 bands
2M1  M 2  4  10 kg
LO LO
 1  300 N/m TO
TO
 2  200 N/m
TO
 3  100 N/m LA
LA
TA
LA TA

One obtains:
TA Acoustic
- 2 optical phonon bands (that have a bands
non-zero frequency at the zone
center)
X M
- 2 acoustic phonon bands (that have
zero frequency at the zone center)
2
a  X
2
FBZ a

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

10
Longitudinal (LO) and Transverse (TO) Optical Phonons

Case I: q x  0 , q y  0
1 1 1
 u1x q x   1   
 u q  Mr M1 M 2
21  0 
LO q x  0    1y x A 
Mr u q
 2x x     M1 M 2
u q   0 
 2y x 
Longitudinal optical phonons: atomic motion in the
direction of wave propagation and basis atoms
move out of phase
 u1x q x   0 
 u q   
21 1
TO q x  0    1y x A 
Mr u2 x q x   0 
u q    M M 
 2y x   1 2
Transverse optical phonons: atomic motion in the M
direction perpendicular to wave propagation and
basis atoms move out of phase
 X

FBZ
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

Longitudinal (LA) and Transverse (TA) Acoustic Phonons

Case I: q x  0 , q y  0
 u1x q x   1
 u q   
LA q x  0   ?  1y x   A 0 
u2 x q x   1
u q  0
 2y x 
Longitudinal acoustic phonons: atomic motion in
the direction of wave propagation and basis atoms
move in phase
 u1x q x  0 
 u q   
TO q x  0   ?  1y x   A  1
u2 x q x  0 
u q   1
 2y x 
Transverse acoustic phonons: atomic motion in the M
direction perpendicular to wave propagation and
basis atoms move in phase
 X

FBZ
ECE 407 – Spring 2009 – Farhan Rana – Cornell University

11
Counting Degrees of Freedom and the Number of Phonon Bands

Periodic boundary conditions for a lattice of N1xN2 primitive cells imply:


  
q  1 b1   2 b2

m1  N1 N
1   where - m 1
N1  2 2
m2  N2 N
2   where -  m2  2
N2  2 2
There are N1N2 allowed wavevectors in the FBZ
There are N1N2 phonon modes per phonon band

Counting degrees of freedom:

• There are 4N1N2 degrees of freedom corresponding to the motion


in 2D of 2N1N2 atoms (2 atoms in each primitive cell)

• The total number of different phonon modes in the four bands is


also 4N1N2

ECE 407 – Spring 2009 – Farhan Rana – Cornell University

12

You might also like