Dopant Diffusion

ESS4810 Lecture
Fall 2010

MOS Transistors

• Challenge: accurate control of the

placement of the active doping regions
 Structure of CMOS Inverter

High-Temperature Diffusion

• Introduce dopants into silicon to

control the majority-carrier type
and resistivity of layers formed in
the wafer
• Used in the formation of deep
layers exceeding a few tenths of a
micron in depth
 Diffusion Process

• The deposition of a shallow high-

concentration layer of the desired
dopants in the silicon surface
through windows etched in the
protective barrier layer
• At high temperatures, the dopant
atoms move from the surface into
the silicon crystal via the sub-
stitutional or interstitial diffusion
mechanisms

Pre-deposition:
Dose control

Drive-in:
Profile control
 Substitutional Diffusion

• Dopant atom hops from

one crystal lattice site
to another
• Vacancies must be
present in the silicon
lattice
• Statistically, a certain
number of vacancies
will always exist in the Large impurity atom

lattice

Interstitial Diffusion
• Interstitial sites exist
between atoms in the
silicon lattice
• Dopant atoms diffuse
through the crystal by
jumping from one
interstitial site to
another
• Vacancy not required
• Proceeds much more Small impurity atom
rapidly but difficult to
control
 Mathematical Model

• Fick’s first law of diffusion

∂C ( x, t )
J = −D
∂x
D: diffusion coefficient
[ D] = cm 2 / sec C(x,t)

C: concentration
J

Mathematical Model

• Continuity equation
∂C ( x, t ) ∂J ( x, t )
=−
∂t ∂x
• Fick’s second law of diffusion
C(x,t)
∂C ( x, t ) ∂ 2 C ( x, t )
= −D
∂t ∂x 2
D is independent of x J

x
Temperature Dependence of D

− EA
D = D0 exp( )
kT

EA = activation energy in eV
k = Boltzman constant
= 8.6 x 10-5 eV / kelvin

Diffusion Coefficient
Diffusion Profile – Pre-deposition

• Boundary conditions
C ( x = 0, t ) = C0
C ( x = ∞, t ) = 0
C0: solid solubility of dopant
• Initial conditions
C ( x , t = 0) = 0

Solid Solubility
Diffusion Profile – Pre-deposition
2 x
x
∫
− y2
C ( x, t ) = C0 ⋅ [1 − 2 Dt e dy ] = C0 ⋅ erfc( )
π 0
2 Dt
∞
Q = ∫ C ( x, t )dx = 2C0 Dt π ∝ t
0

∂C C0 − x2
=− e 4 Dt
∂x π Dt
C0 : surface concentration
(solid solubility limit)
Q : dose
2 Dt : characteristic distance for diffusion

N: concentration
 Diffusion Profile – Drive-in

• Boundary conditions
C ( x = ∞, t ) = 0
∂C ( x = 0, t )
=0
∂x

• Initial conditions
x
C ( x, t = 0) = C0 ⋅ erfc[ ]
2 ( Dt ) predep

(Dt) of Pre-deposition

Diffusion Profile – Drive-in

Q = 2C0 ( Dt ) predep π
2
Q −x
C ( x, t ) = e 4 ( Dt ) drive−in

π ( Dt ) drive −in

• Summary

2C0 ( Dt ) predep − x2
C ( x, t ) = e 4 ( Dt ) drive−in

π ( Dt ) drive −in
 N: concentration

Thermal Budget

• The sum of Dt products

• Dopants will redistribute when
subjected to various thermal cycles of
IC processing steps
• For small dimension IC devices, dopant
redistribution has to be minimized and
we need low thermal budget processes

( Dt) effective = ∑i Di ti
 Junction Depth

C(Xj)=NB (bulk concentration)

2
Q −x
C ( x, t ) = e 4 ( Dt )
= NB
πDt

X j = 2 Dt ln(C0 / N B )

Sheet Resistance

L L t
R =ρ = Rs
wt w I w L

• Rs=ρ/t
– sheet resistance of the layer of material
– ohm per square
– measured by four-point probe
 Sheet Resistance

• If C(x) >> NB and

μ(x) ~constant

Irvin’s Curves

• p-type and n-type erfc (2)

• p-type and n-type Gaussian (2)
• Establish the explicit relationship
between
• Co (surface concentration, hard to
measure)
• Xj (junction depth)
• NB (background concentration)
• RS (sheet resistance)
 Irvin’s Curves

• Once any three Co

NB

parameters are
known, the
fourth one can
be determined

RsXj

Diffusion Systems

• The open-furnace-tube system

using solid, liquid, and gaseous
sources is a common diffusion
technology used in IC fabrication
• Boron
• Phosphorus
• Arsenic
• Antimony
 Diffusion Systems

• Boron diffusion
– 2B2O3 + 3Si ↔ 4B + 3SiO2
• Phosphorus diffusion
– 2P2O5 + 5Si ↔ 4P + 5SiO2
• Arsenic diffusion
– 2As2O3 + 3Si ↔ 4As + 3SiO2
• Antimony diffusion
– 2Sb2O3 + 3Si ↔ 4Sb + 3SiO2
Ion Implantation

ESS4810 Lecture
Fall 2010

Ion Implantation
 Advantages

• Precise control of dose and depth

profile
• Low-temperature process
• Wide selection of masking materials
• Less sensitive to surface cleaning
procedures
• Excellent lateral dose uniformity

Ion Implanter
 Ion Implanter

• Ion source
– Produces a plasma containing the
desired dopant
• Mass spectrometer
– Selects the desired dopant ion
• High-voltage accelerator
• Scanning system
• Target chamber

Eaton High Energy Implanter

• Costs several millions (USD)

• 60 wafers/hour
 Mathematical Model

• Range and profile shape depend

on the ion energy (for a particular
ion/substrate combination)
• Concentration of profile depends
on the implantation dose
• Dose: Atoms/cm2
• Concentration: Atoms/cm3

Gaussian Approximation

Rp projected range

∆Rp Longitudinal straggle

 Gaussian Approximation
• Rp and ∆Rp values are given in tables or charts

Monte Carlo Simulation of 50keV

Boron Implanted into Silicon
Dose-Concentration Relationship

Gaussian approximation
Q = ∑0 C ( x)dx
∞

Q = 2π C p ΔR p
1 T
Q=
nqA ∫o
Idt

Q: dose
atoms/cm2

Selective Implantation
 A 2-D Formulation

Junction Depth

C(Xj)=NB (bulk concentration)

− ( X j − Rp )2
C p exp[ 2
] = NB
2ΔR p

X j = R p ± ΔR p 2 ln(C p / N B )
 Energy Loss Mechanisms

• Nuclear stopping
– Crystalline substrate
damaged by collision
– Heavier ions at lower
energy
• Electronic stopping
– Electronic excitation
creates heat
– Light ions at higher
energy

Implantation Damage
 Annealing

• Ion implantation is an ambient-

temperature process
• However, post-implant annealing
step (>900 ºC, ~30 mins) is required
to anneal out defects
– Restore silicon crystallinity
– Put dopants into silicon substitutional
sites for electrical activation

Annealing

as-implanted
profile, gaussian

after annealing for time t

profile becomes a broader
but lower “gaussian”
C p (t = 0) − ( x − Rp )2
C ( x, t ) = exp[ ]
4 Dt 2(ΔR p ) 2 + 4 Dt
1+
2(ΔR p ) 2
Ion Channeling

Comparison