Chapter 4 4. y Phonons I.

Crystal Vibrations

Crystal vibrations
The lattice will vibrate more as temperature is raised. The study of lattice vibrations is important for understanding gy how energy is absorbed in solids. At low temperatures, most of the energy that is absorbed is of two type
Lattice energy: the heat can go to vibrations of the positive charges in the lattice Electronic energy: the heat can go to the electrons

For a metal, C = T+T3

electronic term C/T lattice term

C/T +T C/T=+T2
Slop = Y-axis intercept = T2

Later (by the end of Chapter 6), we will show this can be measured with the heat capacity: C=dU/dT.

 Lattice is not rigid. Atoms can move from equilibrium.

basis atom

Vibrations of crystal with monatomic basis

r = ru1 ,u 2 ,u3 + rj + u(t)
lattice point vibration

Wh wave propagates along a certain direction, entire planes When t l t i di ti ti l of atoms move in phase with displacements us either parallel or p p perpendicular to the direction of wavevector k. k
us-1 us+1 us+2 us+3 us+4 us-2 s2 us-1 s1 us us+1 us+2

k
a s-3 s-2 s-1 s s+1 s+2 Two transverse modes

a s-1 s s+1 s+2 s+3 s+4 One longitudinal mode

Monatomic lattice vibrations

C M C M C M C M C us-2 us-1 F us F us+1 L R Consider only nearest-neighbor t i hb coupling

M a = Fs = FR + FL = C(u s +1 u s ) + C(u s 1 u s ) ( ( d2us M 2 = C(u s +1 + u s 1 2u s ) dt

Set of coupled, linear, second order differential equations. Hard to solve in general if Ms and Cs are different. Ms Cs different Cs will be different for longitudinal and transverse waves. Method : a trial solution (good guess)

u s (t) = A exp i ( kx s t ) where xs= sa (discrete)

Dispersion relation
M( 2 )A eiksa e i t = C ( A eik(s +1)a + A eik(s 1)a 2A eiksa ) e i t M2 = C ( eika + e ika 2 ) = C 2 ( cos(ka) 1) 2 = = 2C 2C ka (1 cos(ka) ) = 2sin 2 ( ) M M 2 4C ka sin M 2
1.0

sec ]
[(4C/ /M)

Dispersion relation
When k =

0.8 0.6 0.4 0.2 0.0 00 -1.5

a
: maximum

set the boundary of first BZ

4C M

1/2

-1

-1.0

-0.5

0.0

k [(2 /a) m ]

-1

0.5

1.0

1.5

Long wavelength limit

For a small k (ka<<1),
4C ka = M 2 C a k = M Ca k= M /a Ca

k = vsk

continuum elastic wave limit Dispersion : vk What is the wave velocity?

phase velocity vp group velocity vg

phase velocity v p = k

d group velocity v g = k (k) the velocity of energy propagation in the medium dk

Group velocity and phase velocity

time increas time increasing e e sing s (a) (d)

group velocity

(b)

(e)

d = k (k) ( ) vg dk
phase velocity

(c) time increasing

(f)

vp = k
(g)

(d)

phase velocity
4C/M sin(ka / 2) i (k vp = k k i (k a 4C sin(ka / 2) = 2 M ka / 2

group velocity
d sin(ka / 2) d vg = 4C/M dk dk a 4C ka = cos( ) 2 M 2

sin(ka / 2) vp = vs ka/2
When k=

vg = vs cos(ka/2)

, vg=0 group velocity vanishes ! vp=(2/)vs

(k lies on the boundary of first BZ)

r = nax 2 x Reciprocal lattice G = m a 1 BZ boundary G = x 2 a

Direct lattice Di t l tti

Bragg condition

Range of physically significant k

The motion of atoms The relative displacement between two adjacent atoms

u s (t) = A exp i ( ksa t )

same

u s +1 = eika For a fixed k, us

2 For the other fixed k' = k + n = k+G a ika in2 u s +1 i(k + n2 /a)a =e = e e = eika us
The displacement can always be described by a wavevector within the first BZ. BZ At k=q,

u s (t) = A exp i ( qsa t )

2 2 q,, u s (t) = A exp i ( q)sa t same as k= q At k= a a = A exp i ( 2 s qsa t ) = A exp i ( qsa t )

2 same as k= +q a

Equivalent wavevectors
=4a/3 =4a/7

1 = 4a,

4a 3 = , 5 4a 2 = , 3 4a 4 = , 7

2 4a 5 k3 = 2a 3 k2 = 2a 7 k4 = 2a k1 =

2a 2 = + a 2a 2 = a 2a 2 = 2 2a a

=4a

Just like adding a lattice constant to an atom has no information in real space, adding a wavevector 2/a has no new information in reciprocal space.

The di l Th displacement can always b d t l be described b ( / a) < k ( / a) ib d by

Reduced Brillouin zone scheme

Reduced zone scheme k1
1.0
k2 k3 k4

sec )] c
[(4C/M M

0.8 0.6 0.4 0.2 02 0.0 -2.0 -1.5 -1.0 -0.5 0.0

1/2

-1

k [(2 /a)

0.5

m ]

1.0

-1

1.5

2.0

2.5

3.0

First Brillouin zone

Consider that ks range over all reciprocal space, g p p , the expression for (k) is periodic. All th i f the information i i th fi t B ill i zone : th rest i repeated ti is in the first Brillouin the t is t d with periodicity 2/a, i.e., the frequencies are the same for (k) and (k+G) where G is any reciprocal lattice vector ( ) y p

2 G = n a

Periodicity in reciprocal space In fact, the motions of atoms with wavevector motion with wavevector k + G. All independent vibrations are described by

k is identical to the

k's inside BZ.

Standing wave at the BZ boundary

Group velocity of vibration wave

vg = vs cos(ka/2)
At the BZ boundary, k = and = 2a a vg = 0 zero group velocity
fundamentally different from elastic wave in a continuum any wave (vibration or others) is diffracted if k is on the zone boundary leads to standing wave with zero group velocity
1 .0 0 .8 0 .6 0 .4 0 .2 0 .0 -0 .5 -0 .4 -0 .3 -0 .2 -0 .1 0 .0 0 .1

[(4C/M) (

It is also a standing wave at k=0 (the wavelength is infinite)

0 .2 0 .3 0 .4 0 .5

1/2

sec ]

-1

k [(2 /a ) m

-1

Bragg condition k = (1/2) G

At the BZ boundary, vg=0

k= a
Alternate atoms oscillate Alt t t ill t in opposite phases.

No propagation (standing wave)

us=A exp(is)=(-1)s A
Transverse (shear) ( )
2a

2 Longitudinal (sound)

Contribution of other nearest planes

In metals the effective forces may be quite long range, carried from ion to ion through conduction electron sea sea. The generalization of the dispersion relation to p nearest planes,

2 =

2 Cp (1 cos(pka) ) M p>0
/a

Solving for the interplanar force constant Cp

/a

/a

dk dk cos(rka) = 2 C p (k )
2 p>0

/a

dk (1 cos(pka) ) cos(rka) ( k ) (k )

= 2

Cr a

Ma / M /a dk 2 cos(pka) Cp = 2 - /a

Two atoms per primitive basis

Diatomic lattice
M2 M1 C M2 C M1 C M2 C M1 C us-1 1 vs-1 1 us a vs us+1 1 M2 vs+1 1 M1

Fs = FR + FL

d2us M1 2 = C(vs u s u s + vs 1 ) = C(vs + vs 1 2u s ) dt d 2 vs M 2 2 = C(u s +1 v s v s + u s ) = C(u s +1 + u s 2v s ) dt A trial solution set t i l l ti t

u s (t) = u exp[i (kx s t )] () p v s (t) = v exp[i (kx s t )]

where xs=sa

M1 ( 2 )u eiksa e i t =

( C(v e
ika

iksa

+ v eik(s 1)a ) 2Cue iksa ) e i t

ika M12 u = Cv (1 + e ik ) 2Cu

( 2C M ) u C (1 + e ) v = 0 M ( )v e e = ( C(u e +u e M v = Cu ( e + 1) 2Cv C (1 + e ) u + ( 2C M ) v = 0
2 1 2 iksa i t ika ik(s +1)a 2 2 2 ika 2 2

iksa

) 2Cve iksa ) e i t

2C M 1 2 )( 2C M 2 2 ) C (1 + e ika ) C (1 + e ika ) = 0 ( M 1M 2 4 2C ( M 1 + M 2 ) 2 + C 2 ( 4 ( 2 + 2 cos(ka) ) ) = 0 2C ( M 1 + M 2 ) 2 2 sin 2 (ka/2) 4 + 2C 2 =0 M 1M 2 M 1M 2 1 1 2 4C 2 4 2C + sin 2 (ka/2) = 0 + M 1M 2 M1 M 2

Dispersion relation
1 1 2 = C + C M1 M 2
1 1 + = C 1 M1 M 2
2

1 1 4 ka sin 2 ( ) + M1 M 2 M1M 2 2
1

( M1 + M 2 )

4M1M 2

ka sin ( ) 2
2

1 1 + 2C M M2 1

+ optical branch
2C M2
2C M1

M1 > M2

Two branches correspond to signs of dispersion relation

- acoustic branch

- /a

/a

Zone boundary
At BZ boundary k=/a : sin (ka/2) ~ 1
1 1 + = C 1 M1 M 2
2

( M1 + M 2 )

4M1M 2

M1 M 2 M1 + M 2 = C 1 M1M 2 M1 + M 2 1 1 1 1 = C + M1 M 2 M1 M 2

M1>M2 acoustic branch optical ti l branch

1 1 1 1 2C = C M + M M M = M 2 1 1 1 2
1 1 1 1 2C + = + = C + M M M M1 M2 2 1 2
gap

Long wavelength limit (>>a)

For a small k (ka<<1), sin (ka/2) ~ ka/2
1 1 = C M + M 2 1
1/2

4M1M 2 ka k 1 1 2 ( M1 + M 2 ) 2
2

1/ 2

M1 + M 2 = C M1M 2 M1 + M 2 C M1M 2

M1M 2 1 1 2 ( M1 + M 2 )

( ka )
2

1/ 2

1/ 2

M1M 2 2 1 1 ka ) + ... 2 ( 2 ( M1 + M 2 )
C (ka ) k 2(M1 + M 2 )

1/ 2

acoustic branch optical branch

1 1 + 2C M + M 1 2

as k0

is finite

How about M1=M2

= 2C 1 M
1.4
1/2

1 sin 2 (

ka 2C ka )= 1 cos( ) ; back to monatomic chain 2 M 2

optical branch

(2C/ /M)

1.2 1.0 10 0.8 0.6 06

optical branch is higher k values folded back to diatomic BZ acoustic branch

0.4 0.2 0.0 -1.00 -0.75 -0.50 -0.25 0 00 0 25 1 00 0 75 0 50 0 25 0.00 0.25 ka/2 (2/a)

0.50 0 50

0.75 0 75

1.00 1 00

Acoustic and optical branches at k=0

2C M2

M1 > M2

2C M1

- /a
-acoustic branch k=0, =0; u=v -optical branch k=0, =
2C , M1 + M 2

/a k
M2 vs M1 us+1 M2 vs+1

M2

M1 us-1

M2 vs-1

M1 us

maximum
M2

M2

M2

M2

M u = 2 v M1
M1 M1 M1

Transverse acoustic and optical branches

-acoustic branch acoustic k=
a
M1 M2

, =

2C M1

, v=0

-optical branch
k= k , a

2C ,u 0 u=0 M2

-acoustic branch (k ) acoustic (k,

-optical branch (k ) optical (k,

Difference between acoustical and optical modes d ti l d

http://www.chembio.uoguelph.ca/educmat/chm729/Phonons/optmovie.htm

Transverse acoustical mode for diatomic chain

Transverse optical mode for diatomic chain

Oscillation of atoms in optical branch polarizes the matter electrically and this type of oscillation can be excited by infra-red optical radiation. This is the reason why this branch was called "optical".

Longitudinal optical mode for diatomic chain

k =0

k =

N atoms per primitive basis

If the lattice has a basis, there will be more degrees of freedom. The vibration of the atoms within a basis causes optical branches.

In three dimensions N atoms per primitive cell

3N vibration branches

Optic 3(N-1) : LO (longitudinal) TO (transverse) Acoustic (3) : LA (longitudinal) TA1 (transverse) TA2 (transverse)

Ge: 2 atoms per primitive basis

In three dimensions 2 atoms per primitive cell
2-fold degeneracy

3x2 vibration branches

Optical

(3) : LO (longitudinal) TO (transverse)

Ge T=80K

Acoustic (3) : LA (longitudinal) TA (transverse) The two phonon branches are p horizontal at the zone boundary position, Kmax=(2/a)(1/2,1/2,1/2). The LO and TO branches coincide at K=0.

2-fold degeneracy

So f S far, and k are continuous except that, d ti t th t

Quantization of elastic waves

k is restricted in the first BZ (discrete of lattice spacing) is forbidden in some gaps (splitting of acoustic and optical modes)
Additional quantization effects:

k is quantized by finite crystal size size.

Energy of vibration mode is quantized by quantum effect.

discrete k values di l

Finite crystal size N+1 atoms, atoms a= /N spacing N bounds

Eg. 1D monatomic chain w/. length g g

u s (k t) = A exp[i (k s t )] (k, ) kx
u1(k t)= uN 1(k t) (k,t) N+1(k,t) kNa=n2

where xs=sa

exp[ik(Na)] 1 exp[ik(Na)]=1
2 2 n = n discrete ! k= Na

Quantized energy of vibration mode

N=20, k=n(2/20a)=(n/10)(/a) only 20 modes are allowed. l d ll d Each k has a corresponding , what is the energy associated w/ w/. this mode?
Some quantum systems: photons: k=2/ =ck A particle in a box: = 2 n photon energy E= mode energy Ek,= nk ( ) Number of photon at (k,)

n 2 = n kn = n En

( =

2 2 kn ) 2 =n 2m 2m 2 2

Phonon: particle-like properties

number n i e gy Energy i wavevector k i crsytal momentum Pi = k i
Not a real momentum !

What is the real momentum? Physical momentum eg. k=0, k 0 corresponding to translation of the whole crystal by u(t) di t t l ti f th h l t l b (t) u(t) ()

k=0

du(t) but Pcrystal = k = 0 dt eg. k0, eg k0 corresponding to relative motions of atoms around the equilibrium Preal = Nm

k0
Preal = 0 but Pcrystal = k 0

Real and crystal momenta

Preal du d i (t) = M dt i Preal = M du(t) dt u s (t) = u(t)exp[iksa]

exp[iksa]
s

du(t) 1 exp[ikNa] =M dt 1 exp[ika] discrete k = 2 n Na Preal = 0 exp[ikNa] = exp[ i2 n] = 1

Only k=0 carries the real momentum. y

Lattice vibrations Specifying the vibrational states of the crystal by specifying number of phonon in each state ki (ki, ) is determined by structure and binding ni is determined by excitation (thermal, acoustic, etc..)

Phonons: quantized lattice vibrations

Simple harmonic oscillator:

1 2 U(x) = cx 2 1 c E n = n + where = 2 m
Zero point energy Lattice vibrations: wavevector k

frequency = f k

()

Dispersion relation SHO

1 energy E n = n + 2

Phonons: can be thought as particles interact with other electrons, phonons, etc

Determination of phonon structure

Neutron scattering neutrons only scatter off atoms, not electrons N t tt i t l tt ff t t l t Primary tool for obtaining phonon dispersion relation Elastic scattering

k = G

Bragg condition k ' = k + G

2 2 k'2 k = 2 M n 2M n 2

k'

k =k
'

Inelastic scattering phonon emission

p phonon absorption p

k q

k'

k q

k'

2 2 k'2 k k ' + q = k + G and + q = 2M n 2M n

2 2 k'2 k k q = k + G and q = 2M n 2M n '

Experimental measurements of dispersion curves di i

Dispersion curves as a function of wave vector q are measured by inelastic diffraction If the atoms are vibrating then diffraction can occur with energy loss or gain by scattering particles In principle, can use any particle neutrons from a reactor, X-rays from a synchrotron He atoms which scatter from surfaces synchrotron, surfaces, Neutrons are most useful for vibrations For temperature~ vibration energies F ~ atomic size, energies at room t t i i i t t ib ti i X-ray only recently has it been possible to have enough resolution meV resolution with K V X V l ti ith KeV X-rays

Experimental setup
detector Triple axis : rotation of sample sample Selected energy out gy Single crystal monochromater Single crystal monochromater Neutrons or X-rays with broad range of energies

Selected energy in

A triple axis neutron spectrometer at Brookhaven

(1012Hz)
Measure

[100] Na T=90K
q(2/a)

k, k', and E q and q d

for h f phonons [110]

q(2/a)

[111]

q(2/a)

Review
Fundamental concepts needed to understand the vibratory motion of atoms: Normal mode of vibration : all atoms oscillate with the same frequency. Only atomic vibrations with certain frequencies, determined by interatomic forces, occur in any given solid. --- periodic arrangement of atoms Normal mode displacements for these materials have an especially simple form and are relatively easy to discuss. If the displacements of atoms from their equilibrium sites are small, the forces they exert on each other are proportional to their displacements, as if the atoms were connected by ideal spring spring. Atomic motions are simple harmonic. substitute expressions for the forces into Newtons 2nd law generate a set of differential equations one for each atom equations, seek solutions for which all atomic displacements have the same freq.

a constant v group

k <<

v phase

M u 2 0 v M1 Center of mass is stationary

Dispersion relation
At BZ boundary

1 1 2C + M M2 1

+ optical branch
2C M2 2C M1

= constant

v phase

2C a = M2

M1 > M2

v group = 0
u=0 = constant

- acoustic branch

- /a

/a k
k <<

v phase = v group = 0

Long wavelength limit :

a u 1 v

2C a M1

= vs k

C a vs = 2 ( M1 + M 2 )

v=0

More complicated lattices

1st and 2nd nearest neighboring couplings in 1D monatomic chain

Two dimensional monatomic rectangular l tti T di i l t i t l lattice

How about two atoms per primitive basis in one dimension?