Chapter 6: Diffusion in Solids

MSE 280: Engineering Materials

Professor Lane W. Martin

(lwmartin@illinois.edu)

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 1

 Overview

ISSUES TO ADDRESS...

• How does diffusion occur?

• Why is it an important part of processing?

• How can the rate of diffusion be predicted for

some simple cases?

• How does diffusion depend on structure and

temperature?

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 2

 Diffusion

Mechanisms
• Gases & Liquids – random (Brownian) motion
• Solids – vacancy diffusion or interstitial diffusion

Importance of diffusion?
• All materials processing
• Energy storage (e.g. fuel cells, batteries…)
• Filtration
• Chemical reactions
• Semiconductor devices
• And more…
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 3
 Processing Using Diffusion (1)

• Diffuse carbon atoms into the host

iron atoms at the surface

• Example of interstitial diffusion is a

case hardened gear

• Result
• The “case” is: Callister, Fig. 6.0
• Hard to deform  C atoms
“lock” planes from shearing
• Hard to crack  C atoms
put surface in compression

From Callister 6e resource CD.

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 4

 Processing Using Diffusion (2)
• Silicon with P for n-type semiconductors:
• Process:
1. Deposit P rich
layers on surface.

silicon

2. Heat it.
3. Result: Doped
semiconductor regions.

silicon
Fig. 18.0, Callister 6e.

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 5

 Diffusion: The Phenomenon
: Atoms of one material diffusing into another and
vice versa (e.g. In an alloy, atoms tend to migrate from regions of
large concentration)
Initially After some time

Adapted from
Figs. 5.1 and
5.2, Callister 6e.

100%

0
Concentration Profiles

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 6

 Diffusion: The Phenomenon

: Atoms within one material exchanging

positions. (i.e. In an elemental solid, atoms also migrate)

Label some atoms at t = 0 Observe at t > 0

C
A
D
B

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 7

 Diffusion Mechanisms

For diffusion to occur:

1. Adjacent site needs to be empty (vacancy or
interstitial)
2. Sufficient energy must be available to break
bonds and overcome lattice distortion

There are different possible mechanisms but let’s

consider the simple/more common cases:
1. Vacancy diffusion
2. Interstitial diffusion

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 8

 1) Vacancy Diffusion

First, bonds with the neighboring

atoms need to be broken

Lattice
distortion

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 9

 2) Interstitial Diffusion

Callister Fig. 6.3

Typically more rapid than vacancy diffusion 

SMALLER species diffuse FASTER

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 10

 Modeling Diffusion: Flux

Consider atoms (or mass, M) going through a plane

2 atoms passed from left to

right (+ direction)
t=0
1 atom passed from right
t = t’
to left (- direction)

1atom M
Net result: or
area  t ' At '

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 11

 Modeling Diffusion: Flux
• :

• Directional Quantity

From Callister 6e resource CD.

• Flux can be measured for:

--vacancies
--host (A) atoms
--impurity (B) atoms In general: diffusion flux may or may not be
the same over time
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 12
 Derivation of Fick’s First Law (1)
What causes net flow of atoms?
• Concentration Profile, C(x): [kg/m3]
Cu flux Ni flux

Concentration Concentration Adapted from

of Cu [kg/m3] of Ni [kg/m3]
Fig. 5.2(c),
Callister 6e.

Position, x
• Fick's First Law:

From Callister 6e resource CD.

• The steeper the concentration profile, the greater the flux!

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 13
 Derivation of Fick’s First Law (1)

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 14

 Derivation of Fick’s First Law (2)

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 15

 Steady State Diffusion
: the concentration profile doesn't
change with time

dC
• Apply Fick's First Law: J x  D
dx
dC  dC
• If Jx,left = Jx,right , then     
dx left dx right

• Result: the slope, dC/dx, must be constant

(i.e., slope doesn't vary with position)!

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 16

 Steady State Diffusion
Rate of diffusion independent of time
dC
Flux proportional to concentration gradient =
dx

C1 C1 Fick’s first law of diffusion

dC
C2 C2 J  D
dx
x1 x2
x
D  diffusion coefficient
dC C C2  C1
if linear  
dx x x2  x1

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 17

 Steady State Diffusion: Example
• Steel plate at 700°C
with geometry shown:

• Q: How much carbon

transfers from the rich to
the deficient side?
= 2.4x10-10 g/cm2s

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 18

Example: Personal Protective Equipment

• Methylene chloride is a common ingredient of paint

removers. Besides being an irritant, it also may be
absorbed through the skin. When using this paint
remover, protective gloves should be worn.
• If butyl rubber gloves (0.04 cm thick) are used, what is
the diffusive flux of methylene chloride through the
glove?
• Data:
– diffusion coefficient in butyl rubber:
D = 110 x10-8 cm2/s
– surface concentrations: C1 = 0.44 g/cm3
C2 = 0.02 g/cm3

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 19

 Example: Personal Protective Equipment
• Solution – assuming linear conc. Gradient (i.e. steady-state)
glove dC C2  C1
C1 J -D  D
tb 
2
6D
dx x2  x1
paint skin
remover Data: D = 110 x 10-8 cm2/s
C2 C1 = 0.44 g/cm3
x1 x2
C2 = 0.02 g/cm3
x2 – x1 = 0.04 cm

-8 2 (0.02 g/cm3  0.44 g/cm3 ) g

J   (110 x 10 cm /s)  1.16 x 10-5
(0.04 cm) cm2s
 Compare this result to J = 2.4x10-10 g/cm2s calculated in the previous
example of C diffusion in Fe (at higher T and smaller diffusing species).
FASTER diffusion through materials with weak (secondary) bonds

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 20

 Example – Fick’s First Law
We can purify hydrogen gas by diffusing it through a palladium
sheet. Compute the number of kilograms of hydrogen that
pass / hour through a 6 mm thick sheet of palladium with an
area of 0.25 m2 at 600°C.

Assume:
• D = 1.7x10-8 m2 / s
• CH = 2.0 kg / m3
• CL = 0.4 kg / m3
• That steady state conditions have been achieved.

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 21

Temperature Dependence of Diffusion

 Qd 
D  Do exp  
 R T 
D = diffusion coefficient [m2/s]
Do = pre-exponential [m2/s]
Qd = activation energy [J/mol or eV/atom]
R = gas constant [8.314 J/mol-K]
T = absolute temperature [K]

Note: Same type of T-dependence as

vacancies and interstitials
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 22
 Temperature Dependence of Diffusion

1500

1000

600

300
T(C)
10-8

D (m2/s) Dinterstitial >> Dsubstitutional

C in α-Fe Al in Al
10-14 C in γ-Fe Fe in α-Fe
Fe in γ-Fe

10-20
0.5 1.0 1.5 1000 K/T
Adapted from Fig. 6.7, Callister & Rethwisch 3e. (Date for Fig. 6.7
taken from E.A. Brandes and G.B. Brook (Ed.) Smithells Metals
Again, SMALLER
Reference Book, 7th ed., Butterworth-Heinemann, Oxford, 1992.) species diffuse FASTER

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 23

 Temperature Dependence of Diffusion
Example: At 300ºC the diffusion coefficient and activation energy for
Cu in Si are
D(300ºC) = 7.8 x 10-11 m2/s
Qd = 41.5 kJ/mol
What is the diffusion coefficient at 350ºC?

D transform data ln D

Temp = T 1/T
Qd1 Q  1
lnD2  lnD0    and lnD1  lnD0  d  
R  T2  R  T1 
D2 Qd  1 1 
 lnD2  lnD1  ln    
D1 R  T2 T1 
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 24
 Temperature Dependence of Diffusion

 Qd  1 1 
D2  D1 exp    
 R  T2 T1 

T1 = 273 + 300 = 573 K

T2 = 273 + 350 = 623 K

11 2   41,500 J/mol  1 1 

D2  (7.8 x 10 m /s) exp    
 8.314 J/mol - K  623 K 573 K 

D2 = 15.7 x 10-11 m2/s

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 25

 Crystal Structure and Diffusion

1500

1000

600

300
T(C)
10-8 C diffusion in γ-Fe
(FCC) is slower than
D (m2/s)
in α-Fe (BCC) even at
higher T!
10-14

10-20
0.5 1.0 1.5 1000 K/T

More OPEN STRUCTURES allow FASTER diffusion.

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 26

 And now for the complication…

Note that we’d have to remove carbon from the right

side and add to the left side to keep a constant flux.

Possible in the previous example of processing using

gas source that can be added and removed but in many
cases this may not be possible….

Non steady-state diffusion

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 27

 Non-Steady State Diffusion

• The concentration of diffusing species is a function of

both time and position C = C(x,t)
• In this case is used

C C 2
D 2
t x

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 28

 Non-Steady State Diffusion
From Fick’s 1st Law: dc
J  D
dx
Jl Jr
Take the first derivative w.r.t. x:
dJ d  dc 
  D 
dx dx  dx 
dx c = conc.
inside box
Conservation of mass: dc J r  J l dJ
  
dt dx dx
i.e. fluxes to left and to right have to correspond to concentration change.
Sub into the first derivative:
dc d  dc 
 D  Fick’s 2nd law
dt dx  dx 
Partial differential equation. We’ll need boundary conditions to solve…
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 29
 Example Non-Steady State Diffusion

• Copper diffuses into a bar of aluminum (semi infinite solid).

C( x, t )
Cs At to, C = Co inside the Al bar
At t > 0, C(x=0) = Cs and C(x=∞) = Co

t2 t3
t t1
Co o
position, x

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 30

 Solution

C x ,t   Co  x 
 1  erf  
Cs  Co  2 Dt 

C(x,t) = Conc. at point x at CS

time t
erf (z) = error function
2 z y 2 C(x,t)

  0
e dy
Co
erf(z) values are given in
Table 6.1 in text

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 31

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 32
MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 33
 Non-Steady State Diffusion Cont.
Suppose it is desired to achieve a specific concentration
(C1) in an alloy, therefore:
C ( x, t )  Co C1  Co
  constant
Cs  Co Cs  Co

which leads to:

Specified with C1
x
 constant
2 Dt

Known for given system

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 34

 Example Processing Question
• Copper diffuses into a bar of aluminum
• 10 hours at 600°C gives desired C(x)
• How many hours would it take to get the same C(x) if we
processed at 500C?
Key point 1: C(x,t500C) = C(x,t600C).
Key point 2: Both cases have the same Co and Cs.
• Result: Dt should be held constant.

2 Dt

Note: values
• Answer: of D are
provided here

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 35

 Diffusion Design Example

• During a steel carburization process at 1000oC, there is

a drop in carbon concentration from 0.5 at% to 0.4 at%
between 1 mm and 2 mm from the surface (γ-Fe at
1000oC).
• Estimate the flux of carbon atoms at the surface given:
Do = 2.3 x 10-5 m2/s for C diffusion in γ-Fe.
Qd = 148 kJ/mol
ργ-Fe = 7.63 g/cm3
AFe = 55.85 g/mol

If we start with Co = 0.2 wt% and Cs = 1.0 wt% how long

does it take to reach 0.6 wt% at 0.75 mm from the surface
for different processing temperatures?

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 36

 Diffusion Design Example

11
10

10
9 T (oC) t (s) t (h)
7 300 8.5 x 1011 2.4x108 27,782 yrs!
10
time (hours)

10
5
900 106,400 29.6
950 57,200 15.9
1000
1000 32,300 9.0
10
1050 19,000 5.3
0.1
200 400 600 800 1000 1200
T (K)

Need to consider factors such as cost of maintaining furnace at

different T for corresponding times

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 37

 Fick’s 2nd Law Example
Determine the carburizing time necessary to achieve a carbon
concentration of 0.30 wt% at a position 4 mm into an iron-
carbon alloy that initially contains 0.10 wt% C. The surface
concentration is to be maintained at 0.90 wt% C and the
treatment is to be conducted at 1100°C.

Assume: D = 5.3x10-11 m2 / s

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 38

 Diffusion in Ionic Solids

Need to consider coulomb interactions between ions…

- + - + - + - + - + -

+ - + - + - + - + - +

- + - + - + - + - + -

+ - + - + - - + - +

- + - + - + - + - + -

+ - + - + - + - + - +

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 39

 Diffusion in Ionic Solids

Which do you expect to diffuse faster, cations

or anions?

• Smaller cations will usually diffuse faster

• But analogous to charge neutrality required for
defect formation, each ion will need counter
charge to move with it (e.g. vacancy, impurity or
free electrons or holes).

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 40

 Structure and Diffusion

• Open crystal structures • Close-packed structures

• Lower melting T materials • Higher melting T materials

• Materials w/secondary • Materials w/covalent

bonding bonding

• Smaller diffusing atoms • Larger diffusing atoms

• Lower density materials • Higher density materials

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 41

 Concepts to Remember

• Diffusion mechanisms and phenomena

– Vacancy diffusion
– Interstitial diffusion
• Importance/usefulness of understanding diffusion
(especially in processing)
• Steady-state diffusion
• Temperature dependence – similarity to vacancies and
interstitials
• Non steady-state diffusion
• Structural dependence (e.g. size of the diffusing
atoms, bonding type, openness of medium)

MSE 280: Engineering Materials ©Lane W. Martin Spring 2011 42