Manufacturing Processes

(ME361)- Lecture 3
Shantanu Bhattacharya

Review of previous lecture

Manufacturing properties of materials.
Bonding in solids (Attractive and
repulsive forces).
Crystal structure (BCC, FCC and CPH)
Lattice parameters.
Crystal formation process (Epitaxial
silicon, polysilicon, single crystalline
silicon, Model studies for siliconCzochralshky and float zone method for
silicon wafers)

Crystal structures of some

common metals

Silicon for fabrication

Silicon is the most popular Microelectronic material.
Silicon and its compounds (Oxides, Nitrides, polysilicon) can
be categorized into three classes based on the amount and
range of order.
These are single crystalline, polycrystalline and amorphous
materials.
In single crystalline materials, almost all of the atoms in the
crystal occupy well defined and regular positions known as
lattice sites.
Materials like silicon dioxide are amorphous which means
that they do not posses any long range order.
The third class of material is polycrystalline and these are
single small crystals randomly oriented with respect to each
other.

Single Crystalline Silicon

They are characterized by crystalline
orientation of their surfaces.
The classification is based on Miller
indices as shown in the figure below.
A particular direction is indicated with
square bracket such as [100].
The set of equivalent directions is
described in angle brackets <100>.
If this direction is the normal vector of a
plane, it is denoted with parenthesis
(100).
The set of equivalent planes is described with braces, such as {100}.
Single crystalline silicon is mostly fabricated with Czocharalski growth method. A small
seed crystal with a given orientation is dipped into a highly purified silicon melt. The seed
is slowly pulled out of the melt while the crucible is rotated.
The other method is floating zone method where a polysilicon rod is used as a starting
material.
A seed crystal at the end of the rod defines the orientation . A radio frequency heater
locally melts the polysilicon rod. Crystal growth starts with the end from the seed.

Single Crystalline silicon formulation

(Czochralskis growth method)

Single crystal silicon is formulated

with Czochralski growth method.
A small seed crystal with a given
orientation is dipped into a highly
purified silicon melt.
The seed is slowly pulled out of
the melt while the crucible
containing the melt is rotated.
The material is polycrystalline
silicon and is 99.9999% pure.
The poly is loaded into a fused
silica crucible that is contained in
an evacuated chamber.
The chamber is back filled with
inert gas and the crucible is
heated to 1500 deg. C.

The seed crystal is a small chemically etched crystal lowered into contact with the
melt. This must be carefully oriented since it will serve as the template for growth of
the much larger crystal.

Czochralskys Growth
Method

Czochralskys Growth
Method

Czochralskys Growth
Method

Czochralskys Growth
Method

In reality, the maximum pull rate is not normally used.

The crystalline quality is a sensitive function of the pull rate.
The material near the melt has a very high density of point defects. So quick cooling
would help to prevent these defects to go into the formulating crystal.
However, too much gradient may create large thermal stresses and thus
dislocations, particularly in larger diameter wafer.

Single Crystalline silicon formulation (Float

zone method)
This method is used for extremely
high purity silicon growth.
A rod of high purity polycrystalline
material is held in a chuck while a
metal coil driven by a high power radio
frequency signal is slowly passed
along its length.
Alternatively, a focussed e-beam can
also be used for heating the rod.
The field setup by the RF power leads
to eddy currents and joule heating and
the material is melted.
To enhance the growth along the preferred crystal orientation a seed crystal is
injected into the top of the molten rod.
In this technique a thin neck of 3mm diameter and 10-20mm long is pulled and
the pull rate and the temperature lowered to shoulder the crystal out to a larger
diameter.

Epitaxial Silicon

Epitaxy is the single crystalline layer growth from another

single crystalline substrate.
The most important technique for epitaxy growth is CVD.
The table below lists the set of reactions for CVD using
silane or dichlorosilane at high temperature 1200 deg. C.
The epitaxial layer can be doped if
dopant gases like diborane for p-type or
phosphine for n-type are mixed during
CVD process.
Epitaxy can be also grown by MBE
(molecular beam epitaxy). The process
is similar to an evaporation process
using silicon melt in a crucible. MBE is
carried out under ultra high vacuum and
temperatures between 400 and 800
deg. C.

Polysilicon
Polycrystalline silicon is refered to as polysilicon, which is
deposited with LPCVD process with silane.
The deposition temperatures range from 575 deg. C to 650 deg.
C.
At temperatures below 575 deg. C, the silicon layer is amorphous.
The grain size is .03-.3 microns.
Poly-silicon can be doped in situ with same gases as used for
amorphous silicon.
In surface micro machining poly-silicon is used directly as
mechanical material.
In microfluidics poly-silicon can be used for making channel walls
and also sealing etched channel structures.
Because of the high temperature annealing process the intrinsic
stresses are reduced.

Crystal Imperfections
As some of the properties of a crystalline solid heavily depend on the crystal
structure thus they should vary a lot based on imperfections in such structures.
These imperfections govern most of the mechanical properties of the crystalline
solid.
The study of crystal imperfections and their effects on the properties of a material
is a vast research area.
We would just like to study the various phenomena associated with different
manufacturing processes, e.g., plastic deformation and heat treatment.
The table below shows a classification of property dependent on basic crystal
structure and those dependent on crystal imperfection.

Classifications of Crystal
Point defect:
imperfections
If an imperfection is restricted to the neighborhood of a lattice point, the imperfection is
referred to as point defect .
There can be three different typed of point defect as shown in the figure below:
(1)Vacancy: This gets formulated as one lattice atom is missing. Since an atom
vibrates about its lattice position, the tendency of the atom to jump out of its regular
position creating a vacancy increases rapidly with its energy ,i.e., the temperature. For
example the usual order of vacancy at 500 deg. C is one in 1010 which increases to
one in 300 at 2000 deg. C. It is possible to increase the vacancy density at a given
temperature by rapid cooling or extensive plastic deformation.
(2) Interstitial impurity atom: This gets formulated when an atom occupies an
abnormal position. Such an atom is called an interstitial impurity atom. An interstitial
impurity can get realized when an atom possesses large enough thermal energy or
when its energy is increased by nuclear bombardment.
(3)Substitutional impurity atom: This happens when a regular lattice position is
occupied by an atom of a different material.

Crystal Imperfection
Line defect:
If an imperfection extending along a line has a length much larger than the lattice
spacing, the imperfection is called a line defect or, commonly a dislocation.
(1)Types of line defect:
When an extra half plane of atoms is accommodated by distorting the regular
lattice arrangement as done with plane marked XX in the figure, the resulting
defect is termed edge dislocation:
When the lattice
atoms move from
their regular ideal
position such a
defect is called
screw dislocation.
The dislocation density is
defined as the total length
of all dislocation lines per
unit volume.
In a single crystal, the minimum attainable dislocation density is of the order of 1001000 per cm2 .
In a polycrystalline solid this is as high as 107-108 per cm2

Crystal Imperfections
When an imperfection extends over a surface, the imperfection is known as surface
defect.
The figure below shows a common type of surface defect known as twin.
Twins are normally produced when a metal is stressed at low temperature.
The grain boundaries in a polycrystalline solid can also be considered as a surface
defect.

Slip and
Twinning

Figure 1.8 Permanent deformation (also

called plastic deformation) of a single
crystal subjected to a shear stress: (a)
structure before deformation; and (b)
permanent deformation by slip. The size
of the b/a ratio influences the magnitude of
the shear stress required to cause slip.

Figure 1.8 (a) Permanent deformation of a single crystal under a tensile

load. Note that the slip planes tend to align themselves in the direction
of the pulling force. This behavior can be simulated using a deck of
cards with a rubber band around them. (b) Twinning in a single crystal
in tension.

Slip Lines and Slip

Bands
Figure 1.8 Schematic illustration of slip lines
and slip bands in a single crystal (grain)
subjected to a shear stress. A slip band consists
of a number of slip planes. The crystal at the
center of the upper illustration is an individual
grain surrounded by other grains.

Manufacturing Processes
(ME361)- Lecture 4
Shantanu Bhattacharya

Elastic and plastic

deformation
Most conventional manufacturing
processes involve the deformation of the work
material.

Atomistically, in the absence of any external force, the distance between a pair of
atoms is de .
The net inter-atomic force varies with the
atomic spacing in a manner shown in the figure
on the left.
Under the application of an external tensile
force, the inter-atomic distance increases
beyond de to maintain the equilibrium.
If the external tensile load is of the magnitude
P , then the inter-atomic distance should be da
so that the inter-atomic force is an attractive
force of the same magnitude.
If da is not very much different than de (of the order of 5%), then upon removal of the
external force the atoms attain their original position.
With an external compressive force (when da<de) the interatomic force should be
repulsive in nature. This behavior is known as elastic behavior of solids.
The phenomenon we have described for a pair of atoms is true also for normal
solids even on a macroscopic scale.

Elastic and Plastic

deformation
Now, let us consider a crystal lattice with regularly spaced atoms, as shown in
Figure below.

Under the externally applied shear force, the upper layers of the atoms move to
the right and the lower layers move to the left.
When the applied force reaches a sufficiently high value, the crystal lattice looks
as in figure 1.20 (b).
Here all atoms are in equilibrium and will remain thus if external force is removed
and thus a permanent deformation is produced in the crystal lattice.
This permanent set is termed as plastic deformation and cannot be recovered
after withdrawal of the external load.

Elastic and plastic

deformation

Elastic and Plastic

Deformation

Elastic and Plastic

Deformation

The Carpet Analogy

Tensile Testing

Stress Strain Relationship for engineering materials

Tensile Testing
The strength of a material in the elastic and the plastic regions is represented
by the yield stress (Y ) and the ultimate strength (u ), respectively.
The capability of withstanding plastic deformation is another important
mechanical property of an engineering material.
This reflects the property of the material to distribute the localized stresses, thus
lowering the tendency of crack formation . This property is commonly referred to
as the ductility of the material.
Ductility is expressed by the percentage elongation, i.e., the percentage strain
at fracture point.
Thus the larger % elongation , i.e., the percentage strain at fracture point. Thus
larger % elongation means higher ductility.
Moreover, the strain at the point D (where necking starts) represents the
amount of plastic strain a material can withstand without localized deformation
In the linear region the constant of proportionality between the engineering
stress and strain is known as the Youngs modulus or the modulus of elasticity.

Tensile Testing
Unlike Mild Steel most other materials do not
have a sharply defined proportional limit or yeild
point.
In such instances, the yield stress is defined by
drawing a parallel lineto the tangent of the stress
strain curve at the origin from a point on the xaxis representing 0.2% strain.
The line intersects the stress strain curve at
the point Y. The stress level at Y is the yield
stress in this case.
The typical stress-strain curve for a brittle (nonductile)
material is shown in the figure on the left.
As can be seen, the material fractures with very little or no
plastic strain. (If the % elongation is less than 5 the material
is brittle)

Toughness
Yet another important material property is its ability to absorb energy in the
plastic range. (Toughness)
The index commonly used to describe this is the total area under the stressstrain curve up to the fracture point.
It represents the work done on a material per unit volume.
The figure on the bottom compares two different materials, of which one is
stronger and the other is more ductile.
It is obvious that toughness is representative of the effect of both strength and
ductility.
Referring to the stressstrain curve, it is obvious that
when the deformation of a
specimen is no longer
negligible, the actual stress
should be defined as the ratio
of the load (P) and the
instantaneous area (Ai) rather
than the original area.
This new ratio (P/Ai) is
known as the true stress of
the material

True Stress and True Strain

True Stress = P/ Ai

The true strain versus true

stress curve is indicated by
the dashed line.
This curve shows a
continuous strain hardening
of the material up to fracture
point, a phenomena not
revealed by a stress strain
curve.