VLSI TECHNOLOGY

UNIT I
CRYSTAL GROWTH, WAFER
PREPARATION, EPITAXY AND
OXIDATION

VLSI TECHNOLOGY UNIT 1

Advantages of Siover Ge

Si has a larger bandgap(1.1 eVfor Siversus 0.66 eVfor Ge)

Si devices can operate at a higher temperature (150oC vs100oC)

Intrinsic resistivities higher (2.3 x 105-cm vs47 -cm)

SiO2is more stable than GeO2which is also water soluble

Siis less costly.

The processing characteristics and some material properties of silicon

wafers depend on its orientation.

The <111> planes have the highest density of atoms on the surface, so
crystals grow most easily on these planes and oxidation occurs at a higher

pace when compared to other crystal planes.

Traditionally, bipolar devices are fabricated in <111> oriented crystals
whereas <100> materials are preferred for MOS devices.

Defects
Any non-silicon atoms
incorporated into the
lattice at either a
substitutionalor
interstitial site are
considered point
defects

Point defects are important in the kinetics of diffusion and oxidation.

Moreover, to be electrically active, dopantsmust occupy substitutionalsites
in order to introduce an energy level in the bandgap.
3

Dislocations

are

line

defects.

Dislocations in a lattice are dynamic

defects. That is, they can diffuse under
applied stress, dissociate into two or
more dislocations, or combine with
other dislocations.
Dislocations in devices are generally
undesirable, because they act as sinks
for

metallic

impurities

and

alter

diffusion profiles.
4

Electronic Grade Silicon

Electronic-grade silicon (EGS), a polycrystalline material of high purity, is the

starting material for the preparation of single crystal silicon. EGS is made from
metallurgical-grade silicon (MGS) which in turn is made from quartzite, which
is a relatively pure form of sand. MGS is purified by the following reaction:

Si(solid) + 3HCl (gas) SiHCl3 (gas) + H2 (gas) + heat

The boiling point of trichlorosilane(SiHCl3) is 32oC and can be readily

purified

using

fractional

distillation.

EGS

is

formed

by

reacting

trichlorosilanewith hydrogen:

2SiHCl3 (gas) + 2H2 (gas) 2Si (solid) + 6HCl (gas)

5

CzochralskiCrystal Growth

The

Czochralski(CZ)

process,

which

accounts for 80% to 90% of worldwide

silicon consumption, consists of dipping a
small single-crystal seed into molten
silicon and slowly withdrawing the seed
while

rotating

it

simultaneously.The

crucible is usually made of quartz or

graphite with a fused silica lining. After
the seed is dipped into the EGS melt, the
crystal is pulled at a rate that minimizes
defects and yields a constant ingot
diameter.
6

Oxygen in Silicon

Oxygen forms a thermal donor in silicon

Oxygen increases the mechanical
strength of silicon
Oxygen precipitates provide
getteringsites for unintentional impurities
7

Thermal Donors

Thermal donors are formed by the

polymerization of Siand O into
complexes such as SiO4in interstitial
sites at 400oC to 500oC
Careful quenching of the crystal
annihilates these donors
8

Wafer Preparation

Gross crystalline imperfections are detected visually and defective crystals

are cut from the boule. More subtle defects such as dislocations can be
disclosed by preferential chemical etching

Chemical information can be acquired employing wet analytical techniques

or more sophisticated solid-state and surface analytical methods

Silicon, albeit brittle, is a hard material. The most suitable material for
shaping and cutting silicon is industrial-grade diamond. Conversion of
silicon ingots into polished wafers requires several machining, chemical,
and polishing operations

10

 After grinding to fix the diameter, one or

more flats are grounded along the length of the
ingot. The largest flat, called the "major" or
"primary" flat, is usually relative to a specific
crystal orientation. The flat is located by x-ray
diffraction techniques.
The primary flat serves as a mechanical
locator in automated processing equipment to
position the wafer, and also serves to orient the
IC device relative to the crystal. Other smaller
flats are called "secondary" flats that serve to
identify the orientation and conductivity type
of the wafer.
The drawback of these flats is the reduction
of the usable area on the wafer. For some
200 mm and 300 mm diameter wafers, only a
small notch is cut from the wafer to enable
lithographic alignment but no dopanttype or
crystal orientation information is conveyed.
11

Slicing determines four wafer

parameters:

Surface orientation (e.g., <111> or

<100>)
Thickness (e.g., 0.5 0.7 mm, depending
on wafer diameter)
Taper, which is the wafer thickness
variations from one end to another
Bow, which is the surface curvature of
the wafer measured from the centerof
the wafer to its edge
12

Wafers
The wafer as cut varies enough in thickness to warrant an
additional lapping operation that is performed under pressure using
a mixture of Al2O3and glycerine.
Subsequent chemical etching removes any remaining damaged and
contaminated regions.Polishing is the final step. Its purpose is to
provide a smooth, specularsurface on which device features can be
photoengraved.
13

Typical Specifications for Silicon Wafers

14

Epitaxial Growth
Deposition of a layer on
a substrate which
matches the crystalline
order of the substrate
Homoepitaxy
Growth of a layer of the
same material as the
substrate
Si on Si

Heteroepitaxy
Growth of a layer of a
different material than the
substrate
GaAs on Si

Ordered,
crystalline
growth; NOT
epitaxial

Epitaxial
growth:

General Epitaxial Deposition

Requirements
Surface preparation
Clean surface needed
Defects of surface duplicated in epitaxial layer
Hydrogen passivation of surface with water/HF

Surface mobility
High temperature required heated substrate
Epitaxial temperature exists, above which
deposition is ordered
Species need to be able to move into correct
crystallographic location
Relatively slow growth rates result
Ex. ~0.4 to 4 nm/min., SiGe on Si

General Scheme

Thermodynamics

Specific thermodynamics varies by process

Chemical potentials
Driving force

High temperature process is mass transport controlled, not

very sensitive to temperature changes
Steady state
Close enough to equilibrium that chemical forces that drive
growth are minimized to avoid creation of defects and allow
for correct ordering
Sufficient energy and time for adsorbed species to reach
their lowest energy state, duplicating the crystal lattice
structure
Thermodynamic calculations allow the determination of solid
composition based on growth temperature and source
composition

Kinetics
Growth rate controlled by kinetic
considerations
Mass transport of reactants to surface
Reactions in liquid or gas
Reactions at surface
Physical processes on surface
Nature and motion of step growth
Controlling factor in ordering

Specific reactions depend greatly on

method employed

Kinetics Example

Atoms can bond to flat surface,

steps, or kinks.
On surface requires some critical
radius
Easier at steps
Easiest at kinks

As-rich GaAs surface

As only forms two bonds to
underlying Ga
Very high energy

Reconstructs by forming As
dimers
Lowers energy
Causes kinks and steps on
surface

Results in motion of steps on

surface
If start with flat surface, create
step once first group has bonded
Growth continues in same way

Vapor Phase Epitaxy

Specific form of chemical vapor deposition (CVD)

Reactants introduced as gases
Material to be deposited bound to ligands
Ligands dissociate, allowing desired chemistry to
reach surface
Some desorption, but most adsorbed atoms find
proper crystallographic position
Example: Deposition of silicon

SiCl4 introduced with hydrogen

Forms silicon and HCl gas
Alternatively, SiHCl3, SiH2Cl2
SiH4 breaks via thermal decomposition

Precursors for VPE

Must be sufficiently volatile to allow
acceptable growth rates
Heating to desired T must result in
pyrolysis
Less hazardous chemicals preferable
Arsine highly toxic; use t-butyl arsine
instead

VPE techniques distinguished by

precursors used

Varieties of VPE
Chloride VPE
Chlorides of group III and V elements

Hydride VPE
Chlorides of group III element
Group III hydrides desirable, but too
unstable

Hydrides of group V element

Organometallic VPE
Organometallic group III compound
Hydride or organometallic of group V
element

Not quite that simple

Combinations of ligands in order to optimize
deposition or improve compound stability
Ex. trimethylaminealane gives less carbon
contamination than trimethylalluminum

Other Methods
Liquid Phase Epitaxy
Reactants are dissolved
in a molten solvent at
high temperature
Substrate dipped into
solution while the
temperature is held
constant
Example: SiGe on Si
Bismuth used as
solvent
Temperature held at
800C
High quality layer

Fast, inexpensive
Not ideal for large area
layers or abrupt interfaces
Thermodynamic driving
force relatively very low

Molecular Beam Epitaxy

Very promising technique
Elemental vapor phase
method
Beams created by
evaporating solid source
in UHV

Molecular Beam Epitaxy

25

Source: William R. Wiley Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA

Objective: To deposit single crystal thin films !

Molecular Beam Epitaxy: Idea !

Inventors: J.R. Arthur and Alfred Y. Chuo (Bell

Labs, 1960)
Very/Ultra high vacuum (10-8 Pa)
Important aspect:
micron/hour)
Slow deposition
better vacuum.
26

slow

rates

deposition

require

rate

(1

proportionally

Ultra-pure elements are heated in separate

Molecular Beam Epitaxy: Process
quasi-knudson effusion cells (e.g., Ga and As)
until they begin to slowly sublimate.
Gaseous elements then condense on the wafer,
where they may react with each other (e.g.,
GaAs).

27

The term beam means the evaporated atoms

do not interact with each other or with other
vacuum chamber gases until they reach the
wafer.

Molecular Beam

Oven
A collection of gas molecules moving in the same direction.

Orifice

Simplest way to generate: Effusion cell or Knudsen cell

Test Chamber

Pump
28

Knudson cell effusion beam system

Sample

Oven contains the material to make the beam.

Molecular beam
Oven is connected to a vacuum system
through a hole.
The substrate is located with a line-of-sight to
the oven aperture.
From kinetic theory, the flow through the
aperture
is
simply
the
molecular
impingement rate on the area of the orifice.
29

Impingement rate is:

Molecular BeamI 1 nv 1 p 8kT

4

4 kT

The total flux through the hole will thus be:

pr 2
Q IA
spatial distribution of molecules from the orifice
The
mkT
of a2knudsen
cell is normally a cosine distribution:

30

1 cos
I ' nv

The intensity drops off as the square of the distance

from the Beam
orifice.
Molecular

I sub

cos 1
IA
2
L

or ,

r

2mkT L

I sub

cos

High velocity, greater probability; the appropriate

distribution:
v3
v2
dnv

2 4 exp

2
n

31

where 2kT / m

dv

Integrating the equation gives:

Molecular Beam

Etr 2kT

as the mean translational energy of

Intensity is the
maximum
in the
molecules

rection normal to the orifice and

ecreases with increasing , which
auses problems.

Use collimator, a barrier with a

mall hole; it intercepts all of the
32
ow
except for that traveling towards the sample.

RHEED (Reflection High Energy Electron Diffraction) is used to

monitor the growth of the crystal layers.

MBE: In-situ process diagnostics

Computer controlled shutters of each furnace allows precise

control of the thickness of each layer, down to a single layer
of atoms.
Intricate structures of layers of different materials can be
fabricated this way e.g., semiconductor lasers, LEDs.
Systems requiring substrates to be cooled: Cryopumps and
Cryopanels are used using liquid nitrogen.

33

ATG Instability

Ataro-Tiller-Grinfeld (ATG) Instability: Often encountered during MBE.

If there is a lattice mismatch between the substrate and the growing film, elastic energy is accumulated in the growing film.
At some critical film thickness, the film may break/crack to lower the free energy of the film.
The critical film thickness depends on the Youngs moduli, mismatch size, and surface tensions.

34

OXIDATION
What is oxidation?
Formation of oxide layer on wafer
High temperature
O2 environment

Principle uses of Si dioxide (SiO2)

layer in Si wafer devices
Surface passivation
Doping barrier
Surface dielectric
Device dielectric

Doping barrier
In doping need to create holes in a
surface layer in which specific dopants are
introduced into the exposed wafer surface
through diffusion or ion implantation
SiO2 on Si wafer block the dopants from
reaching Si surface
All dopants have slower rate of movement in
SiO2 compared to Si
Relatively thin layer of SiO2 is required to
block the dopants from reaching SiO 2

Cont..
SiO2 possesses a similar thermal
expansion coefficient with Si
At high temperature oxidation process,
diffusion doping etc, wafer expands and
contracts when
it is heated and cooled
Dopants
close thermal expansion coefficient, wafer
does not warp
Si

SiO2 layer as dopant barrier

Surface dielectric
SiO2 is a dielectric does not conduct
electricity under normal circumstances
SiO2 layer prevents shorting of metal
layer to underlying metal
Oxide layer
MUST BE continuous; no holes or voids
Thick enough to prevent induction
If too thin SiO2 layer, electrical charge in metal
layer cause a build-up charge in the wafer
surface cause shorting!!
Thick enough oxide layer to avoid induction
called field oxide

Metal layer
Oxide layer

Wafer

Dielectric use of SiO2 layer

source

Drain

Field oxide

MOS gate

Types of oxidation
1. Thermal oxidation
2. High pressure oxidation
3. Anodic oxidation

Device oxide thicknesses

Most applications of semiconductor are
dependent on their oxide thicknesses

Silicon dioxide
thickness,

Applications

60-100

Tunneling gates

150-500

Gates oxides, capacitor

dielectrics

200-500

LOCOS pad oxide

2000-5000

Masking oxides, surface

passivation

3000-10000

Field oxides

Thermal oxidation mechanisms

Chemical reaction of thermal oxide
growth
Si (solid) + O2 (gas) SiO2 (solid)

Oxidation temperature 900-1200C

Oxidation: Si wafer placed in a
heated chamber exposed to
oxygen gas

SiO2 growth stages

Initial

In a furnace with O2 gas environment

Si wafer

Linear

Parabolic

Oxygen atoms combine readily with Si atoms

Linear- oxide grows in equal amounts for each time
Around 500 thick
Si wafer

Si wafer

Above 500, in order for oxide layer to keep growing, oxygen

and Si atoms must be in contact
SiO2 layer separate the oxygen in the chamber from the wafer
surface
Si must migrate through the grown oxide layer to the
oxygen in the vapor
oxygen must migrate to the wafer surface

Three dimension view of SiO2 growth by

thermal oxidation
Original SiO2 surface

SiO2 surface

SiO2

Si substrate

Linear oxidation
X

B
t
A

Parabolic oxidation of silicon

X Bt

where X = oxide thickness, B = parabolic rate constant, B/A

= linear rate constant, t = oxidation time

Parabolic relationship of SiO2 growth parameters

X
R
t

where R = SiO2 growth rate, X = oxide thickness, t =

oxidation time

Cont..
Implication of parabolic relationship:
Thicker oxides need longer time to grow than
thinner oxides
2000, 1200C in dry O2 = 6 minutes
4000, 1200C in dry O2 = 220 minutes (36 times
longer)

Long oxidation time required:

Dry O2
Low temperature

Dependence of silicon oxidation rate constants on

temperature

Kinetics of growth

Si is oxidised by oxygen or steam at high

temperature according to the following chemical
reactions:
Si (solid) + O2 (gas) SiO2 (solid) (dry oxidation)
Or
Si (solid) + 2H2O (gas) SiO2 (solid) + 2H2(gas) (wet
oxidation)
Also called steam oxidation, wet oxidation, pyrogenic steam
Faster oxidation OH- hydroxyl ions diffuses faster in oxide
layer than dry oxygen
2H2 on the right side of the equation shows H 2 are trapped in
SiO2 layer

Layer less dense than oxide layer in dry

oxidation
Can be eliminated by heat treatment in an inert
atmosphere e.g. N2

2 mechanisms influence the growth rate of the oxide

1. Actual chemical reaction rate between Si and O 2
2. Diffusion rate of the oxidising species through an already
grown oxide layer
No or little oxide on Si the oxidising agent easily reach the Si
surface
Factor that determine the growth rate is kinetics of the
silicon-oxide chemical reaction
Si is already covered by a sufficiently thick layer of oxide
Oxidation process is mass-transport limited
Diffusion rate of O2 and H2O through the oxide limit the growth
rate is diffusion of O2 and H2O through the oxide

A steam ambient is preferred for the growth of thick oxides:H 2O

molecules smaller than O2 thus, easier diffuse through SiO 2
that cause high oxidation rates

Si oxidation

Mass transport of O2 molecules from gas ambient

towards the Si through a layer of already grown oxide
Flux of O2 molecules is proportional to the differential in
O2 concentration between the ambient (C *) and oxide
surface (C0)
*

F1 h (C C0 )..............5.1

Where h is the mass transport coefficient for O 2 in the gas

phase

Diffusion of O2 through the oxide is proportional to the

difference of oxygen concentration between the oxide
surface and the Si/SiO2 interface. The flux of oxygen
through the oxide, F2 becomes

Where,

C0 Ci
F2 D
...................5.2
tox

Ci = oxygen concentration at theSi/SiO 2 interface

D = diffusion coefficient of O2 or H2O in oxide
tox = oxide thickness

Kinetics of the chemical reaction

between silicon and oxygen is
F3 k sCi .................5.3
characterised by reaction constant, k:
In steady state, all flux terms are
C
equal:
F
F
=
C
5.4 F3. Eliminating C0
1 =..........
2.........
k kt
1
from the
h
D flux equations, we can
obtain:
*

s ox

If N0x is a constant representing

the number of oxidising gas
molecules necessary to grow a
unit thickness
of
oxide,
one
can
dt
N kC
N FN kC
.......5.5
k kt
write:dt
1
*

ox

ox

ox s

ox s
s

s ox

k s k s tox
t
1

The solution
to
this
differential
h
D dt dt..........
5.6
ox
*

N ox kis:
equation
sC
0
0
tox

If tox=0 when t=0, th eintegration yields:

Or

tox2 D D

tox N oxC *dt 0........5.7
2 ks h

1 1
tox 2 DN oxC *t............5.8 *
Defining new constant A and
k s Bh in
terms of D, k s, Nox and C :
tox2 2 D

1 1
A 2 D
............5.9
We can obtain:
ks h
From which we findand
tox :
B 2 DC * N ox ................5.10

t 2 Atox Bt.....................5.11
tox

(t )
.................5.12
2
1
A / 4B

 is introduced to take into account the possible presence of an

oxide layer on the Si before thermal oxide growth being carry
out
Oxide layer can be a native oxide layer due to oxidation of bare Si
by ambient air or thermally grown oxide produced during a prior
oxidation step
=0 if the thickness of the initial oxide is equal to zero

When thin oxides are formed the growth rate is limited by the
kinetics of chemical reaction between Si and O2.
Eq. 5.12 becomes:
B
tox t ...........5.13
A
B

WhichA is linear with time.

The
ratio is called linear growth coefficient, and is
dependent on crystal orientation of Si

When thick oxides are formed, the growth rate is limited by the
diffusion rate of oxygen through the oxide. Eq 5.12 becomes:

tox B (t ) Bt ..............5.14
The coefficient B is called parabolic growth coefficient and is
independent on crystal orientation of Si.
The parabolic growth coefficient can be increased:
Increase the pressure of the ambient oxygen up to 10-20 atm (high
pressure oxidation)

The linear growth coefficient can be increased:

Si consists of high concentration of impurities e.g. phosphorous:
increase point defects in the crystal which increase the oxidation
reaction rate at the Si/SiO2 interface
Oxidation process also generate point defects in Si which enhance
diffusion of dopants. Some dopants diffuse faster when annealed in
oxidising ambient than in neutral gas such at N2

Oxidation rate
Controlled by:
1.
2.
3.
4.

Wafer orientation
Wafer dopant
Impurities
Oxidation of polysilicon layers

1. Wafer orientation
Large no of atoms allows faster oxide growth
<111> plane have more Si atoms than
<100> plane

Faster oxide growth in <111> Si

More obvious in linear growth stage and at low
temperature

2. Wafer dopant(s) distribution

Oxidised Si surface always has dopants;
N-type or P-type
Dopant may also present on the Si surface
from diffusion or ion implantation
Oxidation growth rate is influenced by dopant
element used and their concentration e.g.

Phosphorus-doped oxide: less dense and etch

faster
Higher doped region oxidise faster than lesser
doped region e.g. high P doping can oxidise 2-5
times the undoped oxidation region
Doping induced oxidation effects are more
obvious in the linear stage oxidation

Schematic illustration of dopant distribution as a function of position is the SiO 2/Si structure indicating
the redistribution and segregation of dopants during silicon thermal oxidation

Distribution of dopant atoms in Si after

oxidation is completed
During thermal oxidation, oxide layer
grows down into Si wafer- behavior
depends on conductivity type of dopant
N-type: higher solubility in Si than SiO 2, move
down to wafer. Interface consists of high
concentration N-type doping
P-type: opposite effect occurs e.g Boron doping
in Si move to SiO2 surface causes B pile up in
SiO2 layer and depletion in Si wafer change
electrical properties

3. Oxide impurities
Certain impurities may influence
oxidation rate
e.g. chlorine from HCl from
oxidation atmosphere increase
growth rate 1-5%

4. Oxidation of polysilicon
Oxidation of polysilicon is essential
for polysilicon conductors and gates
in MOS devices and circuits
Oxidation of polysilicon is dependent
on
Polisilicon deposition method
Deposition temperature
Deposition pressure
The type and concentration of doping
Grain structure of polysilicon

Thermal oxidation method

Thermal oxidation energy is supplied by
heating a wafer
SiO2 layer are grown:
Atmospheric pressure oxidation oxidation
without intentional pressure control (autogenerated pressure); also called atmospheric
technique
High pressure oxidation high pressure is
applied during oxidation

2 atmospheric techniques
1.Tube furnace
2.Rapid thermal system

Oxidation methods
Thermal oxidation
Atmospheric
pressure

High pressure

Tube furnace

Dry oxygen
Wet oxygen

Rapid thermal

Dry oxygen

Tube furnace

Dry or wet
oxygen

Chemical oxidation
Anodic
oxidation

Electrolytic cell

Chemical