Crystal Structure / Imperfections

Almost all materials crystallize when they solidify; i.e., the atoms are arranged in an
ordered, repeating, 3-dimensional pattern. These structures are called crystals and are
discernible during microstructural characterization [by optical microscopy, x-ray diffraction,
transmission electron microscopy, etc.] of metals. Microstructure is of critical importance to the
properties of metals.

Aspects of Microstructure :

• grain size and shape

• grain orientation
• phase(s) present
• size and distribution of second phases
• nature of grain boundaries and precipitates
• crystal imperfections
Each grain is a single crystal with a specific orientation.

Imperfections
"FORTUNATELY crystals are seldom, if ever, perfect."

These defects or imperfections include :

• thermal vibrations
(0-D) • point ----- {vacancies, interstitials, substitutional, ... }
(1-D) • line ------- {dislocations}
(2-D) • surface --- {grain/phase boundaries, surfaces,... }
(3-D) • volume --- {amorphous, liquid state, voids, ... }

Without the imperfections, materials would have little ductility

[malleability] and thus cannot be formed into useful shapes.

• Increase strength by impurities / second phases

• Diffusion processes are possible because of vacancies / interstitials
• Grain boundaries affect mechanical properties [both at low and high
temperatures]

KL Murty page 1 NE409/509

 Crystal Structure

• Lattice - of points is an arrangment of points in 3-D such that each point has
identical surroundings

• Different Arrangements - 7 Systems (symmetry) & 14 Bravais Lattices

• Unit Cell - Building block describing the 3-D lattice by translation only

• Type of Arrangement ⁄ Atomic Bonding

• Metallic • Ionic • Covalent

• Lattice Parameter / Lattice Constant - repeat distance along a, b, c

(x, y, z) or (a1, a2, a3)

• Relations between directions (a, b, c) and angles (α, β, γ) define the 7 systems

Important Parameters
SC BCC FCC hcp
• Number of atoms per unit cell
Primitive vs Conventional 1 2 4

• Coordination Number (CN) :

number of equivalent nearest neighbors 6 8 12

• R (atomic radius) vs a (lattice constant) : a = 2R 3 a=4R 2 a=4R

• Packing Fraction (Atomic Packing Factor) :

(no.atoms/u.c.) (vol.atom)
APF (%) = vol. u.c. x100 52% 68% 74.1%

• CPDs <100> <111> <110>

• CPPs {100} {110} {111}

stacking of CPPs : fcc – ABCABCABC vs hcp – ABABAB (see figures)

KL Murty page 1 NE409/509

 Diamond structure (Fig. 3.4) – C, Si, Ge, etc :
FCC with 4 interior atoms (FCC with 2 atoms at each lattice point)
CN=4, 8 atoms/uc, cannot be reduced to a primitive unit cell

Polymorphism – more than one crystal structure

as T increases close packed structures transform to open structures

In general : hcp → fcc → bcc as Temp increases (reverse as P increases)

Notable exception – Fe : bcc at RT (α iron) → fcc (γ iron) → bcc (δ iron)

910C 1400C (see table on p.47 – note on the bottom of the table)

Zr : hcp (α) → bcc (β) @865C

----------

NaCl Structure : (Na+ cation and Cl- anion) Fig. 3.10 KCl, MgO, UC
a = 2(RNa + RCl) CN=6 4 ion pairs/uc (like fcc) – 2 fcc lattices

CsCl Structure : (Cs+ cation and Cl- anion) Fig. 3.11 AgMg, RbCl, CuZn (β -Brass)

a vs R CN=8 1 ion pair/uc (like sc) – 2 sc lattices

Fluorite (CaF2) Structure : (Ca++ cation and F- anion) Fig. 3.12 UO2
4
Ca (fcc with a) and F (sc with a/2) a = (R + + R − ) 4 CaF2 /uc (like fcc)
3
CN = 8 for Ca and 4 for F ----- advantage over UC for nuclear applications

KL Murty page 2 NE409/509

Planes and Directions – Miller Indices –
least integers with no common factors
Directions : [i j k] <i j k> - family
Planes : (h k l) {h k l} – family
reciprocals of intercepts on the x y z axes ↔ Directions are ┴ to Planes

Examples : cubic ----- group work

hcp : hkil so that h+k+i=0 or hk•l ---- examples
(planes and then directions as normal to the planes)

Cubic :
h1h2 + k1k 2 + l1l 2
Angle between 2 directions : Cosθ =
h12 + k12 + l12 h22 + k 22 + l 22

Intersection of 2 planes (direction) :

cross product of normals to the planes (h1 k1 l1) and (h2 k2 l2) :
[h k l] = [h1 k1 l1]x[h2 k2 l2]
Implies that if [h1 k1 l1] falls in a plane (h k l) → [h1 k1 l1] • [h k l]=0

Bragg’s Law : nλ = 2dhkl sinθ

1 h2 + k 2 + l 2
where dhkl is plane separation ----- Cubic : 2
=
d hkl a2

1 4 ⎛⎜ h 2 + hk + k 2 ⎞ l2
hcp : = ⎟+
2
d hkl 3 ⎜⎝ a2 ⎟ c2
⎠

KL Murty page 3 NE409/509

KL Murty NE409/509

Kinetics of Vacancy Formation

Why vacancies occur or exist ? ⇒ to disrupt the ordered structure or increase
randomness ⇒ increases entropy although work (energy) is required to form or
create a vacancy (i.e., to move a host atom from interior to surface).

Thus, Energy ⇑ & Entropy ⇓ ⁄ minimize Gibbs Free Energy (G) ⁄ dG=0
(F = Free Energy, H = Enthalpy) G = F + PV = H - TS = E + PV -TS
[E] = eV, J per atom (1 eV = 1.6x10-19 J); Q = Cal per mole & 1eV=23kCal/mole

Equilibrium Vacancy Concentration :

G(o) = GFE of perfect lattice and G(n) = GFE of lattice with n vacancies
If N = Number of Atoms, the introduction of the vacancies increases the entropy
⇒ Sconf = k lnW configurational entropy with W being the number of
different ways of arranging the N atoms and nv vacancies on N+n lattice cites : W
(N+n)!
= (N+n)Cn = (N+n)CN = N! n! , where N! = N(N-1)(N-2)...1

Thus G(n) - G(o) = ∆G = n gv - kT lnW , where gv is the GFE to form a

vacancy. For thermal equilibrium,

∂∆G ∂ ln W ∂ ln W N+n
= 0 = g v − kT To show = ln n , use Stirling’s approximation
∂n ∂n ∂n
⇒ lnN! = NlnN - N for large N ⇒ True for n also since in general n is also
very large (millions)

n gv
Thus, ln(N+n ) = - kT ⇒
slope = E/k or Q/R
n n gv lnC
≈
N N+n = Cv = exp(- kT) , since
N>>n. Cv is the vacancy
concentration and
gv = Ev + PVv - TSv ≈Ev(orEf) high T low T
1/T
Arrhenius Equation

KL Murty page 4 NE409/509

 Divacancies
At very high temperatures, many vacant lattice sites find neighboring site vacant
(2)
too (Fig. 6.1 text) ⇒ divacancies ⁄ Εv = 2Ev - B , B is the binding energy of a
divancy (this is the energy required to separate a divacancy into two isolated
monovacancies).

(2) β 2Ev - B
Cv = 2 exp (- kT ) , β is the cordination number ⇒ see text Eq. 6.15-6.22

(2)
Cv β B
or 2 = 2 exp (kT)
Cv

Interstitials
ni Ei
(self) Similary, N = Ci = exp(-kT ) , ni = interstitial atoms and
i
Ni = No. of interstitial sites (≠ N, no. of atoms).

Ei >> Ev ⇒ thus the no. of interstitials (self) produced by thermal energy is

relatively small.

Example Calculation :
For Copper (fcc) Qv = 20 kCal/mole B = 7 kCal/mole Qi = 90 kCal/mole

T Cv (2) i
Cv Cv
˚C (K)
RT (293) 1.20x10-15 1.45x10-24 7.3x10-68
500 (773) 2.21x10-6 2.80x10-9 3.57x10-26
1073 (1346) 5.65x10-4 2.63x10-5 2.43x10-15
(TM = 1083) (1 vac per 1769 atoms) ***
***Self interstitials are produced more easily by radiation

KL Murty page 5 NE409/509

Ionics (MX and MX2 types)
⎧ gf
gs ⎪ vM
C = CiM = exp(− )
Schottky ⇒ CvX = CvM = exp(-2kT ) Frenkel ⇒ ⎨ 2kT
⎪ CvX = CiX = exp(− g f )
⎩ 2kT

Divalent Impurity ⇒ Extrinsic vs Intrinsic

e.g. Cation (+ive) vacancies created

due to addition of Ca++ in NaCl
⇒ charge neutrality.
exp(-Es/2kT)
lnCvM
n ppm of Ca++ means nx10-6 Cation
vacancies created
C++
(these are athermal and not
dependent on temperature but equal
to the divalent impurity
concentration) 1/T

⇒ Can get the relationship (CvM vs T) as before

1. qDNDM + qMNM = qXNx ⇒ charge neutrality, where qM = qx(MX type)

2. NDM + NM + NvM = Nx + Nvx ⇒ no. sites for M+++M+ = no. X- sites

+
• Can do the same calculation for Cv for MX2 type (such as CaF2 and/or UO2)
- see Text

Experimental Determination of Cv :
• Simmons-Balluffi Experiment • Quenched-in Resistivity
• FIM - imaging of atoms (for low Cv)

KL Murty page 6 NE409/509