Chapter 2 Reciprocal Lattice

Phys 175A Dr. Ray Kwok SJSU

Crystal Lattice
Periodic f(r + T) = f(r) for any observable functions such as electronic density, electric potential.etc. which means they are all periodic functions because of the translational properties of the lattice vectors.

Fourier Transform
Fourier series (1D): f(r) = n An ei2nr/a where the period is a. such that f(r + ua) = f(r); u is an integer. Or write f(r) G AG eiGr where G = 2n/a (for 1D lattice) and AG = (1/a) cell f(r) eiGr dr

1D Translational Invariance
f(r + T) = G AG eiG(r+T) = G AG eiGr eiGT = f(r) = G AG eiGr i.e. eiGT = 1 For example, if T = ua (1D), then GT = 2un = 2(integer). and eiGT = 1.

3D Translational Invariance
For 3D lattice, eiGT = 1 and T = u1a1 + u2a2 + u3a3 is the translational vector or the lattice vector Define reciprocal primitive vectors b1 = (2/V) a2 x a3 b2 = (2/V) a3 x a1 b3 = (2/V) a1 x a2 where V is the volume of the primitive lattice cell V = | a1 a2 x a3|, such that bi aj = 2ij and ij is the kronecker delta function. Note: {a}s dont have to be orthogonal.

Reciprocal Lattice Vectors

Define G = v1b1 + v2b2 + v3b3 where the vs are integers so that GT = 2 ( u1v1 + u2v2 + u3v3) = 2(integer) [because bi aj = 2ij] and eiGT = 1 In other words, because of the translational invariance property of the crystal, there exist a set of vector G such that GT = 2(integer). This set of vector defines another set of lattice point in the Reciprocal Space.

Reciprocal Vector Space

G has a unit of 1/length. Similar to the wave-vector k in the plane wave expression eikr. G has a meaning in Fourier transform, kspace, or momentum space. It defines a set of lattice points in the kspace.

Direct and Reciprocal Lattice

For every crystal, there is a set of space lattice (crystal lattice) location of lattice points in real space where atoms and molecules are. There is also a set of reciprocal lattice in the momentum space (k-space) something we see in diffraction measurement. However, there is no physical object present at the reciprocal lattice sites.

Light Microscopy vs X-ray Crystallography

~ 500 nm

~ 0.1 nm

see real space

see reciprocal space

Tunneling Microscope vs X-Ray Diffraction

see Direct Lattice (surface only)

Reciprocal Lattice (volume)

Reciprocal nature of diffraction pattern

Braggs law: 2d sin = n d 1/sin

Diffraction maximum can be specified by the Miller indices [hkl]

What do we learn from x-ray?

Lattice parameters (Space Group) Symmetry (Point Group) Miller Index (h,k.l) for each point Intensity (square of structural factor) of each reflection

Properties of G reciprocal lattice

each G is normal to a lattice plane in real space GT = 2n. For a fixed G and n, there are many T vectors satisfied this equation. But they all lie on the plane perpendicular to G.
Crystal plane

Distance between planes = 2/G

If G has no common factor (prime #), then the distance between crystal plane perpendicular to the G is 2/G d = T cos = 2n/G d = Tcos = 2(n+1)/G separation of planes = d d = 2/G and G must be the smallest reciprocal lattice for a given direction in kspace (prime #).

Crystal planes
T T G d

Small G dominates
The larger the G, the closer the crystal plane, and less atoms on the plane. For example, in sc, separation of (100) is a, of (110) is a/2 = 0.71a (110)
(100)

a1
(100)

0.71a

a2
(110)

(210)

Volume of the reciprocal cell

If the volume of a unit cell of the direct lattice is V, then the volume of a unit cell of the reciprocal lattice is (2)D/V where D is the dimension of the lattice, usually 3. 3D: Vg = | b1 b2 x b3| = (2/V)3 {(a2 x a3) (a3 x a1) x (a1 x a2)} = (2/V)3 {(a2 x a3) [a3(a1 x a2)] a1 - [a1(a1 x a2)] a3} = (2/V)3 {(a2 x a3) [a3(a1 x a2)] a1 } = (2/V)3 {(a2 x a3) V a1 } = (2)3/V2 {(a2 x a3) a1 } = (2)3/V

Reciprocal of Reciprocal
The direct lattice is the reciprocal of its own reciprocal lattice b1 = (2/V) a2 x a3 b2 = (2/V) a3 x a1 b3 = (2/V) a1 x a2 reciprocal c1 = (2/Vg) b2 x b3 = V/(2)2 (2/V)2 (a3 x a1) x (a1 x a2) = (1/V) { [a3(a1 x a2)] a1 - [a1(a1 x a2)] a3} = (1/V) { [a3(a1 x a2)] a1 } = a1

Primitive Reciprocal Lattice Cell

The Wigner-Seitz cell of the reciprocal lattice is called the Brillouin Zone

Reciprocal Lattice

Cubic reciprocal lattice

Direct Lattice

SC

r a 1 = ax r a 2 = ay r a 3 = az

Reciprocal Lattice r b1 = (2 / a )x r SC b 2 = (2 / a )y r b = (2 / a )z 3

Volume (k-space)

(2 / a )3

FCC

r ( ) a 1 = 1 2a x+y r 1 +z ) a 2 = 2 a (y r a 1 +x ) 3 = 2 a(z
r 1 +y z ) a 1 = 2 a(x r 1 +y +z ) a 2 = 2 a( x r a 1 y +z ) 3 = 2 a (x

r b1 = 2a (x +y z ) r 2 b 2 = a ( x + y + z ) BCC r b = 2 (x 3 a y + z)

4(2 / a )

BCC

r b1 = 2a (x +y ) r FCC 2 ( ) b = y + z r 2 a b = 2 (z 3 a + x)

2(2 / a )

bcc

Direct Lattice Primitive Cell

Wigner Cell

Reciprocal Lattice (fcc)

Brillouin Zone

fcc

Direct Lattice Primitive Cell

Wigner Cell

Reciprocal Lattice (bcc)

Brillouin Zone

Hexagonal - homework
Derive the reciprocal vectors for a 2D hexagonal lattice. Real space a Reciprocal space

a 4 a 3

a 4 a 3 Real and reciprocal lattices appear to be rotated from one another !

1D - Reciprocal Lattice
Direct Lattice
a

biaj = 2ij means ba = 2 so b = 2/a along the same direction

b=2/a

Reciprocal Lattice
2/a First BZ

2D Reciprocal Lattice
Direct Lattice

r a1 = a1x r a2 = a2y

a1

a2

biaj = 2ij means b1a1 = 2; b1a2 = 0 b2a1 = 0; b2a2 = 2

1st BZ

Reciprocal Lattice
b2=2/a2 b1=2/a1

r b1 = r b2 =

2 a1 2 a2

x y

Diffraction in quantum mechanic

Diffraction in quantum mechanic description: (Born approx) The transitional amplitude to scatter from k-state to kstate = Mkk = <k|V(r)|k> = *(r) V(r) (r) dr (r) = eikr is the plane wave (think of the Bragg scattering plane incident & diffracted waves) In Fourier series, V(r) = VG eiGr (with the translational property of the lattice) Mkk= VG ei(G + k - k)r dr = VG (G k) = VG if G = k , Mkk= 0 otherwise. So the condition for Bragg scattering becomes G = k = k - k

Bragg Scattering
For elastic scattering, frequency = ck is unchanged. Speed of light is the same before & after scattering off the target, so |k| = |k|. write k + G = k and square both sides gives 2 kG + G2 = 0,
k=G

k k

2kG = 2kG cos = -G2 -2k cos = -G 2k sin = G 2(2/) sin = 2n/d 2d sin = n which is the Bragg condition.

X-ray data

Bragg Conditions
r r 2k G + G 2 = 0 r |= G / 2 | k G
Can use different k to probe the same G. Different circles, lane P g different k and angles. Brag
k=G
k k

k
o

G/2

G/2

Or similarly, all k end on the perpendicular bisecting plane of G would give Bragg diffraction. (BZ !!)

BZ

Laue Equations
G = k = v1b1 + v2b2 + v3b3 bi aj = 2ij aj k = 2vj define sets of cones around {a}
a3 families of cones a2

a1

Allowed diffraction when the cones intersect. Note: Cone height is discrete, not a continue spectrum. Zero height is allowed) different integers vs.

Ewald Sphere
A sphere with radius k. Any point lands on the surface of the sphere is an allowed G value, i.e. diffraction occurs at that particular angle & k. In principle, it can map out the whole reciprocal lattice using different k.

G k 2 k

Example
RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

Note: a1b2 = 0 a2b1 = 0 here {a} are the direct lattice vectors & {b} are the reciprocal lattice vectors

REAL LATTICE a1 a2

(0,0)

RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

length=1/d0,1 (0,1) planes REAL LATTICE a1 a2 (0,0)

RECIPROCAL LATTICE b1
(3,1)

Note:

b2
(0,1)

(2,1) (1,1)

length=1/d1,1

length is longer than (0,1) since spacing between (1,1) planes is smaller.

REAL LATTICE a1 a2

(0,0)

(1,1) planes

RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

length=1/d2,1

REAL LATTICE a1 a2

(0,0) (2,1) planes

RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

length=1/d3,1

REAL LATTICE a1 a2

(0,0)

(3,1) planes

RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

(0,1) planes REAL LATTICE a1 a2 (0,0) (1,1) planes (2,1) planes (3,1) planes

RECIPROCAL LATTICE b1 b2
(0,1) (0,2) (2,1) (1,1) (3,1)

length=1/d0,2 (0,2) planes REAL LATTICE a1 a2 (0,0)

with basis, for example

RECIPROCAL LATTICE b1 b2
(0,1) (1,2) (0,2) (2,1) (1,1) (3,1)

length=1/d1,2

REAL LATTICE a1 a2

(0,0)

(1,2) planes

RECIPROCAL LATTICE b1 b2
(0,1) (2,2) (1,2) (0,2) (2,1) (1,1) (3,1)

length=1/d2,2

REAL LATTICE a1 a2

(0,0)

(2,2) planes

RECIPROCAL LATTICE b1 b2
(0,1) (2,2) (1,2) (0,2) (2,1) (1,1)

(3,2)

(3,1)

length=1/d3,2

REAL LATTICE a1 a2

(0,0)

(3,2) planes

RECIPROCAL LATTICE b1 b2
(0,1) (2,2) (1,2) (0,2) (2,1) (1,1)

(3,2)

(3,1)

REAL LATTICE a1 a2

(0,0)

(0,1) planes (0,2) planes (1,1) planes (1,2) planes (2,1) planes (2,2) planes (3,1) planes (3,2) planes

RECIPROCAL LATTICE b1
(3,1)

b2
(0,1)

(2,1) (1,1)

length=1/d0,1

How do we orient the crystal to observe diffraction from the (0,1) reflection?
(0,1) planes

REAL LATTICE a1 a2

(0,0)

(0,0)

(0,1) planes

n=2dsin Bragg condition-- upper beam has to be an integral number of wavelengths from the lower beam for constructive interference.

(3,1) (2, 1) (1,1) (0, 1)

(0,0)

(0,0)

(0,1) planes

n=2dsin

(3 , (2 , 1) (1 , 1) (0 , 1)

1)

3D
0) (0 ,

) 0,0

(1,1) planes

(0 ,0 )

(0 , (0 ,

1) (1 , 0)

1) (2, 1)

(3 ,

1)

(2,1) planes

Interpretation of the observed diffraction pattern

The reciprocal lattice can be constructed.

Diffraction by a monatomic lattice

+ k k +

monatomic lattice

Bragg condition: In phase to get constructive interference (+ + + +) Zero reflection otherwise (+ - + - + -)

primitive cell

From here we got: G = k // d G=2/d Bragg diffraction Laude Equations Construction of reciprocal lattice from EACH diffraction observed i.e. each reciprocal lattice point gives a spot on the screen

Diffraction by a lattice with a Basis

+ k k + + + d1 diatomic lattice d

conventional cell

For diatomic lattice (assume identical atoms) ALL in phase to get max amplitude (++ ++) Zero amplitude if the reflection from the basis cancel (+ +). Likewise, if the basis reflections are in phase, but the lattice reflections are out-of-phase, amplitude still cancels (++ ) i.e. to get max Bragg scattering, both lattice reflections and basis reflections should be in phase. Amplitude is less if they are partially in phase or out-of-phase.

With multiple atoms, or non-identical atoms, scattering amplitude would not be uniform. There are partial cancellations. The resulting reciprocal lattice of the direct lattice would be modified.

Fourier Analysis of a Basis

Scattering amplitude Mkk = <k|V(r)|k> = *(r) V(r) (r) dr where (r) = eikr V(r) = j V(r-rj) = G,j VG eiG(r-rj) rj is the position of the jth atom in the unit cell. Mkk= VG ei(G + k - k)(r-rj) dr = G j e-iGrj [VG ei(G + k - k)r dr] Define SG = j e-iGrj fj the Structure Factor and fj is the Atomic Form Factor such that Mkk SG

Now if we specify G and rj as G = v1b1 + v2b 2 + v3b3 and r j = x j a1 + y j a 2 + z j a3 we get: s 2 i (x j v1 + y j v2 + z j v3 ) SG = f j e j =1 Note that SG can be complex, because the scattering intensity involves the magnitude squared of SG. The evaluation of atomic form factor is complicated but for spherically-symmetric electron density, it can be written as (Kittel) sin (Gr ) f j = 4 n j ( r ) r 2 dr Gr 0

Example bcc
primitive cell bcc V=a3/2 reciprocal cell fcc V =(4/a)3/4
* see primitive vectors

lattice cell sc w/ 2-pts basis V=a3

reciprocal lattice sc V= (2/a)3 [8x smaller than the cube (4/a)3 ] extra lattice points extra diffraction spots?

bcc structure factor

For a BCC lattice, rj = (000) and ( ). The structure factor is then:

SG = 1 + e

i (v1 + v2 + v3 )

]f

The structure factor is maximum SG = 2f when the sum of the indices is even, i.e. v1+v2+v3=2n . The structure factor is SG = 0 when the sum of the indices is odd, i.e. v1+v2+v3=2n+1 .
The only G show up in diffraction pattern are the ones fit the fcc lattice points.
(0,0,0)

Same physical measurement regardless how the cell was defined.

Example fcc
primitive cell fcc V=a3/4 reciprocal cell bcc V =(4/a)3/2
* see primitive vectors

lattice cell sc w/ 4-pts basis V=a3

(0,0,0)

reciprocal lattice sc V= (2/a)3 [8x smaller than the cube (4/a)3 ] extra lattice points extra diffraction spots?

fcc structure factor

For a FCC lattice, we have four atoms per init cell located at (000), (0 ), ( 0 ) and ( 0). The structure factor is then: S = 1 + e i (v1 + v2 ) + e i (v1 + v3 ) + e i (v2 + v3 ) f
G

When all indices are even or odd, then SG = 4f. When the indices are partially even and partially odd, then SG = 0.
The only G show up in diffraction pattern are the ones fit the bcc lattice points.
(0,0,0)

Same physical measurement regardless how the cell was defined.

Homework
Problem # 1, 2 Problem # 5, 6, 7