9.

Efficiency limits, losses and measurement

 10. Standard silicon solar cell technology

a. Describe the steps of solar module preparation

1. Sand to metallurgical grade silicon
1. sand to metallurgical-grad silicon
• Silicon is the second most a abundant element in the earths crust.
• The source material for the extraction of silicon is silicon dioxide the major
constituent of sand.
• In the present commercial extraction process, the crystalline form of silicon dioxide,
quartzite is used.
• Commercially the Si is extracted in large arc furnace.

• Quartz sand is reduced in large arc furnaces by carbon (in the form of a mixture of
wood chips, coke and coal) to produce silicon according to the following reaction
• SiO2(s)+C(s)-----------Si(l)+CO2 (g)

Solidification MG-Si (metallurgical grade silicon)

• Then silicon is periodically poured from the furnace and blown with O/chlorine
mixture to further purify it.
• Next poured into shallow troughs, where it solidifies and is subsequently broken
into chunks
• Electric arc produces melt
• Very energy consuming process: 14kWh/kg
Characteristics
• 1 million metric tons of this MG-si are produce globally each year (large
production).
• MG-Si is used in the steel and aluminum industries.
• The total processing energy requirement are acceptable. The result of the MG-Si is
inexpensive.
• purity of metallurgical grade Si is 98 to 99%
• the major impurities are iron and aluminum
• Typical concentrations of impurities in MG Si
2. MGS to SGS
2. MG-silicon to semiconductor-grade silicon
• For use in solar cells, silicon must be much purer than MG-Si.
• Powdered MG-Si is reacted with anhydrous HCl at 300°C in a fluidized bed reactor
to form Trichlorosilane, SiHCl3 in presence of cu catalyst
• MG-Si(s) +3HCl(g) --SiHCl3(g) +H2(g)
• During this reaction impurities such as Fe, Al, and B react to form their halides (e.g.
FeCl3, AlCl3, and BCl3) with different boiling points
• The SiHCl3 has a low boiling point of 31.8 °C and distillation is used to purify the
SiHCl3 from the impurity halides.
• The resulting SiHCl3 now has electrically active impurities (such as Al, P, B, Fe, Cu or
Au) of less than 1 Parts per billion atomic (ppba= 5 × 1013 𝑎𝑡𝑜𝑚𝑠⁄𝑐𝑚3 )

The gases emitted are passed through a condenser and the resulting liquid subjected to
multiple fractional distillation to produce SeG-SiHCl3 (trichlorosilane), the source material
for the silicon industry.
MG-Si(s) +3HCl(l) --SiHCl3(g) +H2(g)

Condenser and multiple SeG-SiHCl3 (l)

fractional distillations
Semiconductor-grade purity trichlorosilane

• Finally, the pure SiHCl3 is reacted

with hydrogen at 1100°C for ~200 –
300 hours to produce a very pure
form of silicon.
• SeG-SiHCl3(g) +H2SeG-Si(s) +
3HCl
• The reaction takes place inside large
vacuum chambers and the silicon is
deposited onto thin polysilicon rods
(small grain size silicon) to produce
high-purity polysilicon rods of
diameter 150-200mm.
• The process was first developed by
Siemens in the 60's and is often referred to as the Siemens process. About 80% of
the world’s polysilicon is produced using the Siemen’s process.
 • The resulting rods of semiconductor grade silicon are broken up to form the
feedstock for the crystallization process.
• Final polysilicon chunks ready for loading into a crystal growth furnace.

Characteristics
• The production of semiconductor grade silicon requires a lot of energy.
• Energy consumption 75-130kWh/kg silicon
• Low yield ~37%
• The high cost of this stage.
• Reach to purity 99.9999%

3. SGS to single crystal wafers

3. seG poly crystalline to single-crystal wafers
• Silicon must very pure and be in a single-
crystal form with essentially zero defect in
the crystal structure.
• The majority of silicon solar cells are
fabricated from silicon wafers, which may be
either single-crystalline or multi-crystalline.
• Single-crystalline wafers typically have
better material parameters but are also
more expensive.
• Crystalline silicon has an ordered crystal
structure, with each atom ideally lying in a
pre-determined position.
• Crystalline silicon exhibits predictable
and uniform behaviour but because of
the careful and slow manufacturing
processes required, it is also the most
expensive type of silicon.
• Single crystalline substrates are typically
differentiated by the process by which
they are made.
• Czochralski (Cz) wafers are the most
commonly used type of silicon wafer, and
are used by both the solar and integrated
circuit industry.
• The process of making a large single
crystalline silicon ingot by the
Czochralski process is shown in Figure

• For solar cells

• boron, a p-type dopant, is normally used, using a seed crystal
and
• with very close temperature control, it is possible to pull from
the melt a large cylindrical single crystal of silicon
• Crystals of diameter in excess of 12.5 cm and
• 1-2m in length are routinely grown in this manner.
• The use of quartz crucibles in the manufacture of Cz substrates
causes the incorporation of ppm (1018 cm-3) oxygen into the
silicon ingot.
• The oxygen itself is relatively benign but creates complexes
with boron doping that degrades the carrier lifetime.
 • N-type ingots fabricated with phosphorous dopants have similar oxygen
concentrations but do not show the degradation effect

Ingot to wafer
• Once an ingot has been grown it is then sliced up into wafers.
• The large single crystal is sliced up into wafers which are as thin as possible
• Silicon solar cells need only be 300μm or so thick to absorb most of the appropriate
wavelength in sunlight

characteristic
• The present wafering technology its difficult to cut wafers from the large crystals
which are any thinner than 300μm and still retain reasonable yields.
• More than half the silicon is wasted as kerfs or cutting loss in the process
• This step lowers the overall yields of single-crystal
• Cutting silicon into wafers leaves the surface covered with cutting slurry and the
surface is damaged due to the action of the saw.
• After cutting the silicon substrates needs to be etched off by several microns on both
sides. Usually aqueous alkali solution of NaOH or KOH at 80-90 °C are used for the
first removal of 10 µm of damaged silicon, for the as-cut wafer (as-cut wafers are
silicon wafers which are produced as slicing).
• The overall chemical reaction generally accepted are
• The etching rate of silicon depends on the product of hydroxide ions ([OH]-) and
free water concentration ([H2O]).

• Recently ingot cutting technology has been significantly improved which reduced
the importance of having an etching process.
• After the etching process, the surface of produced solar cells becomes shiny and the
average reflectance factor is more than 35 %
• To minimize reflection from the flat surface solar cell wafers are textured, this
means a creating a roughened surface, so that incident light will have a larger
probability of being absorbed into the solar cell.
• The absorption of incident light by the solar cell is increased by applying a
geometrical surface texture to the front surface which reduces average reflection to
less than 10 %
• For mono c-Si solar cells, an anisotropic texturization is applied using dilute alkaline
solution, KOH/NaOH, at 70-80°C and 1-2 wt % with isopropanol as a wetting agent.
• The surface morphology of the textured wafer is presented in Figs. 2a and 2b, where
the pyramid height is 3-5 μm.
 • The average reflection factor is decreased from 35 % to ~10 %.
4. Single crystal wafers to solar cells
4. single-crystal wafers to solar cells
• After etching the silicon wafers and cleaning them, additional impurities are
introduced into the cell in controlled manner by a high-temperature impurities
diffusion process.
• The emitter diffusion process is performed in a variety of ways.
• In this case of p type wafer a phosphorous containing coating is applied to the
surface. The wafers are then put in a belt furnace to diffuse a small amount of
phosphorous into the silicon surface.
 • The flow rate of carrier gas directly influences the production.
• Carrier gas must be dry (RH less than 40%), free of oxygen and chemically
inert. Helium is most commonly used because it is safer than, but comparable
to hydrogen in efficiency, has a larger range of flow rates and is compatible
with many detectors. Nitrogen, argon, and hydrogen are also used.
• A carrier gas is bubbled through phosphorus oxychloride (POCl 3), mixed with a
small a mount of oxygen, and passed down a heated furnace tube in which the
wafers are stacked, this grows an oxide layer on the surface of the wafers
• At the temp involved (800—1100) °C the phosphorus diffuses from the oxide into
the silicon
• After about 20min the P impurities override the B impurities in the region near the
surface of the wafers to give a thin, heavily doped n-type region

Edge isolation and PSG removal

• After the standard batch diffusion process, the phosphorus doped layer (n+) covers
the entire surface including edges.
• It is extremely important to disconnect the front emitter from the back contact
through etching the edge of the cell.
• Several inline and batch processes are successfully established; namely chemical
etching, plasma etching, laser cutting, mechanical sawing, grinding with sandpaper,
single side etching. Single side etching in PV industries is widely used.
• In the diffusion process, a glass, known as phosphosilicate glass, of phosphorus
pentaoxide (P2O5) and silicon dioxide (SiO2) are formed due to the uncontrolled
amount of dopant and irregular diffusion at the silicon surface.
• If the phosphorus concentration exceeds the solid solubility then it precipitates on
the surface and leads to PSG layer formation.
• The thickness of PSG layer is ~0.03 μm that can be removed by a wet chemical dip
by using diluted hydrofluoric acid (HF).
• In subsequent processing the oxide layer is removed as are the junction at the side
and back of the cell
Metal contact (Vacuum evaporation)
• The standard technology to metal contact are then
attached to both the n-type and the p-type region
• The metal to be deposited is heated in a vacuum to a
high enough temp to cause it to melt and vaporize, it
will then condense on any cooler parts of the vacuum
system in direct line of sight, including the solar cells
• The back contact is normally deposited over the entire
back surface, while the top contact is required in the
form of a grid.

Techniques for defining top grid

1. Use a metal shadow mask
2. The vacuum evaporation equipment is
expensive compared to its throughput.
3. Moreover, because of the nature of the
evaporation process, only a small fraction of
the material being evaporated ends up where
desired. This is quite extravagant when
dealing with material as expensive as Ag.
4. The metal can be deposited over the entire
front surface of the cell and subsequently
etched a way from unwanted region using a photographic technique known
photolithography

The contact made up three separate layer

1. Thin layer of titanium is used as the bottom layer for good adherence to silicon
2. Layer of silver in the top for low resistance and
solderability
3. The sandwiches layer is palladium which
prevents an undesirable reaction between the
Ti and Ag layers in the presence of moisture
After deposition the contacts are sintered at
500-600 °C to give good adherence and low
contact resistance,
Finally a thin antireflection (AR) coating is
deposited on the top of the cell by the same
vacuum evaporation process.
Characteristics
Yield of about 90% from starting wafers to completed terrestrial cells can be
obtained.
All processes above are done on a batch basis, where a batch of perhaps 40 to 100
wafers moves through the processing sequence at the same time
This make the processing very labor-intensive.

Anti Reflection Coating

• For non-encapsulated cells, the first optical loss is the
shading by the metal grid at the illuminated face, of the
order of 7% for industrial cells
• The next loss comes from the reflectance at the Si
interface,
more than 35% for bare Si in air due to its high refraction
index.
• An antireflective or anti-reflection (AR) coating is a
type of optical coating applied to the surface of lenses
and other optical elements to reduce reflection.
• A layer of non-absorbing material with a lower refractive index (nARC) on top of the
Si substrate decreases reflectance
• Many coatings consist of transparent thin film structures with alternating layers of
contrasting refractive index.
• Layer thicknesses are chosen to produce destructive interference in the beams
reflected from the interfaces, and constructive interference in the corresponding
transmitted beams.
 • For reflection to become zero at the minimum, the coating index should be the
geometric average of those of air and silicon, i.e. 2.4 at 600 nm for non-encapsulated
cells.

• The optimum refractive index for the ARC film to provide minimum reflection can
be calculated as,

• where nf is the refractive index of the film

• n1 is the refractive index of medium 1 (air) and
• n2 is the refractive index of medium 2 (silicon)

5. Solar cell to solar module

5. solar cell to solar module
After interconnecting between cells solar cells require encapsulation by glass to:
1. Mechanical protection
2. Electrical isolation
3. Chemical protection
4. Mechanical rigidity to support the brittle cells and their flexible interconnections
 • Most designs use a toughened or tempered glass with a low iron content for good
light transmission

• For the adhesive and pottant layers, silicones have been widely used.

• Pottant is the material, usually some form of silicone, like ethylene vinyl acetate
copolymer that holds crystalline solar cells in place inside their modules

• They have good UV stability, low light absorption and are appropriately elastic to
alleviate thermal stresses in the module, but they are expensive. Polyvinyl butyral
(PVB) and ethylene/vinyl acetate (EVA) have been used by several manufacturers
for the corresponding layers in the laminated approach.

we will connect many cells to make module and connect more than two modules to
make panel and the many panel to make array show below:

b. What is etching, texturing, edge isolation, PSG removal?

c. What is anti-reflection coating? Why is it necessary in solar cell?
d. What design parameters should be considered in designing and selecting an ARC
material?
e. What is lithography and Photolithography? Positive and negative
f. Describe the procedures of Photolithography.
g. How surface states are created at solar cell surfaces?
h. How the surface states reduce the output current from a solar cell?
i. What is Surface Recombination Velocity (SRV)?
j. How to reduce the surface states and effect of this in solar cells
k. Sketch and describe the operation of a standard solar cell composed of p type, n type,
ARC layer, surface passivation, metal connections, BSF, etc. with band diagram
 11. Crystal growth

a. Procedure, advantage and disadvantages of Czochralski (CZ) method

The Czochralski process is a method of crystal growth used to obtain single crystals of
semiconductors (e.g. silicon, germanium and gallium arsenide), metals (e.g. palladium,
platinum, silver, gold), salts and synthetic gemstones. The process is named after Polish
scientist Jan Czochralski, who invented the method in 1916 while investigating the
crystallization rates of metals.

Czochralski Growth (Cz)

❖ The equipment and the basic principle
for Cz pulling is shown in Figure
❖ The Cz equipment consists of a vacuum
chamber in which feedstock material,
that is, polycrystalline silicon pieces or
residues from single crystals, is melted
in a crucible and a seed crystal is first
dipped into the melt.
❖ Then the seed is slowly withdrawn
vertically to the melt surface whereby
the liquid crystallises at the seed.
❖ High vacuum conditions can be used as
long as the melt weight is small (<1–2
kg), but
❖ with larger melts (often more than 30
kg) only pulling under argon inert gas
stream is practicable.
❖ Owing to the reduced argon
consumption, the argon pressure is set
in the 5 to 50 mbar regime in the PV
industry
❖ whereas in the microelectronic industry atmospheric pressure is also used.
❖ Shortly before the desired diameter is reached, the pulling velocity is raised to the
specific value at which the crystal grows with the required diameter.
❖ Owing to the seed rotation, the crystal cross section is mostly circular.
❖ In general, the pulling velocity during the growth of the cylindrical part is not kept
constant, but is reduced towards the bottom end of the crystal. This is mainly caused
by the increasing heat radiation from the crucible wall as the melt level sinks. The
heat removal of the crystallisation thus becomes more difficult and more time is
needed to grow a certain length of the crystal.
❖ Standard pull speeds in the body range from 0.5 to 1.2 mm/min.
❖ The diameter of the crystal in PV is often chosen between 100 and 150 mm. This is
due to the short-circuit current of big solar cells where values of 6 A per cell are
exceeded. It is difficult to provide a proper contacting scheme in screen print
technology that can handle such high currents in the front contacts without high
series-resistance losses. With even larger cell sizes, this effect becomes more
problematic.
❖ To complete the crystal growth free of dislocations, the crystal diameter has to be
reduced gradually to a small size, whereby an end cone develops. For this purpose,
the pulling speed is raised and the crystal diameter is decreased. If the diameter is
small enough, the crystal can be separated from the melt without a dislocation
forming in the cylindrical part of the crystal.

❖ SeG polycrystalline Si is molten in a silica (quartz) crucible, at T=1410 °C

❖ Owing to the reaction between the liquid Si and the quartz crucible, the crucible is of
considerable importance to the growth.
 ❖ The silica of the crucible supplies considerable amounts of oxygen to the melt and,
owing to the high purity of the silica, only small amounts of other impurities.
However, the crucible tends to dissolve after a long-standing time so that the risk for
particles in the melt from the crucible is increased with increased pulling time.
❖ The oxygen of the melt adds up to 1018 oxygen atoms/cm3 to the growing crystal,
whereas carbon is usually <1017/cm3 and has only little impact on the solar cell
performance.
❖ Oxygen impurities can have beneficial effects. Carefully chosen annealing conditions
can allow the formation of oxygen precipitates.
❖ These have the effect of trapping unwanted transition metal impurities in a process
known as gettering.
❖ Additionally, oxygen impurities can improve the mechanical strength of silicon
wafers by immobilising any dislocations which may be introduced during device
processing.
❖ dopants (B, P) can be added to the melt
❖ Oxygen impurities can react with boron in an illuminated environment, such as that
experienced by solar cells.
❖ This results in the formation of an electrically active boron–oxygen complex that
detracts from cell performance (Open circuit voltage and leakage current).
❖ Module output drops by approximately 3% during the first few hours of light
exposure.
❖ Ld=200 mm t=15 msec
❖ Cz technology is mature and cost-effective.
❖ Equipment and processes for semiautomated growing of crystals are commercially
available so that several Cz pullers can be run by a single operator.
❖ Owing to the robust making of the machines, many Cz growers more than 20 years
age are still in production.
❖ In CZ Si PV technology, approximately 30% of the costs are in the crystal ingot, with
20% in wafering, 20% in cell fabrication and 30% in module fabrication.

Disadvantages
❖ One of the main disadvantages of Cz crystallisation of silicon is the fact that square
cells are best suited to built a highly efficient solar module, whereas Cz ingots have a
round cross section.
❖ In order to use both the crystal and the module area in the best manner, the ingots
are usually cut into a pseudosquare cross section before they are cut into wafers.
❖ Additionally, the tops and tails of the ingots cannot be used for wafer production.
❖ The cropped and slapped materials, that is, tops and tails and so on, are then fed
back into the growth process again.

b. Role of oxygen in CZ method

❖ To complete the crystal growth free of dislocations, the crystal diameter has to be
reduced gradually to a small size, whereby an end cone develops. For this purpose,
 the pulling speed is raised and the crystal diameter is decreased. If the diameter is
small enough, the crystal can be separated from the melt without a dislocation
forming in the cylindrical part of the crystal.
❖ SeG polycrystalline Si is molten in a silica (quartz) crucible, at T=1410 °C
❖ Owing to the reaction between the liquid Si and the quartz crucible, the crucible is of
considerable importance to the growth.
❖ The silica of the crucible supplies considerable amounts of oxygen to the melt and,
owing to the high purity of the silica, only small amounts of other impurities.
However, the crucible tends to dissolve after a long-standing time so that the risk for
particles in the melt from the crucible is increased with increased pulling time.
❖ The oxygen of the melt adds up to 1018 oxygen atoms/cm3 to the growing crystal,
whereas carbon is usually <1017/cm3 and has only little impact on the solar cell
performance.
❖ Oxygen impurities can have beneficial effects. Carefully chosen annealing conditions
can allow the formation of oxygen precipitates.
❖ These have the effect of trapping unwanted transition metal impurities in a process
known as gettering.
❖ Additionally, oxygen impurities can improve the mechanical strength of silicon
wafers by immobilising any dislocations which may be introduced during device
processing.
❖ dopants (B, P) can be added to the melt
❖ Oxygen impurities can react with boron in an illuminated environment, such as that
experienced by solar cells.
This results in the formation of an electrically active boron–oxygen complex that detracts
from cell performance (Open circuit voltage and leakage current).

c. Material purification process through float zone method

d. Procedure, advantage and disadvantages of Float Zone method
Float Zone Method
❖ An alternative crystal growth method is the
float-zone method, which produces a silicon
moncrystal ingot with significantly lower
oxygen content.
❖ The float zone method was developed in the
1950s and produces the purest bulk silicon
crystals known to date.
❖ Video:Zone Refining.mp4
❖ A cylindrical polysilicon rod is placed in a
chamber
❖ A seed crystal placed at the lower end
❖ An encircling inductive heating coil melts the
silicon material
❖ The coil heater starts from the bottom and is raised pulling up the molten zone
❖ A solidified single crystal ingot forms below
❖ Impurities prefer to remain in the molten silicon so very few defects and impurities
remain in the forming crystal
❖ High purity regrown Si crystal: no crucible, liquid region prevents impurity from
entering the growing crystal
❖ Ld>500 mm t>100 msec
❖ uses the advantage of the small segregation coefficient of many impurities in silicon
❖ is a purification method - the impurities prefer to stay in the melt and hence are
swept to the end of the recrystallized crystal
❖ the melt remains in vacuum or inert gas, no incorporation of impurities
❖ produced crystals have a high quality, especially low oxygen concentrations
❖ more expensive than Czochralski growth

❖ The single-crystal growth methods, float-zoning (FZ) 20% and Czochralski growth
(CZ), 80%
 ❖ There are two principal technological advantages of the FZ method for PV Si growth.
• The first is that large lifetime values are obtained as a result of higher purity
and better microdefect control, resulting in 10% to 20% higher solar cell
efficiencies.
• The second is that faster growth rates and heat-up/cool-down times, along
with absences of a crucible and consumable hot-zone parts, provide a
substantial economic advantage.
❖ The main technological disadvantage of the FZ method is the requirement for a
uniform, crack-free cylindrical feed rod. A cost premium (100% or more) is
associated with such poly rods.
❖ At the present time, FZ Si is used for premium high-efficiency cell applications and
CZ Si is used for higher-volume, lower-cost applications.
❖ Electrical power requirements for these two methods 35-40kWh/kg for CZ growth
and on the order of 18 kWh/kg by Float Zone FZ
❖ Not only were energy requirements reduced, but also argon consumption was
reduced from 3 m3/kg of Si to 1 m3/kg of Si.
❖ Also, oxygen content in the crystals was reduced by 20%, the crystal growth rate
was increased from 1.28 kg/hr to 1.56 kg/hr, and relative solar cell efficiency
increased by 5%.
Bulk multi-crystalline silicon
❖ Multi-crystalline silicon offers advantages over monocrystalline silicon with respect
to manufacturing cost and feedstock tolerance, however slightly reduced
efficiencies.
❖ Another inherent advantage of multi-crystalline silicon is the rectangular or square
wafer shape
❖ Two different fabrication technologies for multi-crystalline silicon,
• the Bridgman and
• The block-casting process are employed.
❖ In both processes the solidification of high-quality multi-crystalline silicon ingots
with weights of 250 to 300 kg, dimensions of up to 70 × 70 cm2 and heights of more
than 30 cm have been successfully realized.
❖ The Bridgman technology is a quite commonly used technique,
❖ The only two companies mainly employing the casting technology are Kyocera
(Japan) and Deutsche Solar GmbH (Germany)
❖ The main difference between both the techniques is that
❖ for the melting and crystallization process only one crucible (Bridgman) is used,
❖ whereas for the crystallization process a second crucible (block casting) is used.

e. FZ method vs. CZ method.

f. Procedure, advantage and disadvantages of Bridgman and the block-casting process for
poly crystalline silicon production
Bridgman process
❖ In the case of the Bridgman
process, a silicon nitride (Si3N4)-
coated quartz crucible is usually
employed for melting of the silicon
raw material and subsequent
solidification of the multi-
crystalline ingot.
❖ The Si3N4 coating thereby serves as
an anti-sticking layer preventing
the adhesion of the silicon ingot to
the quartz crucible walls
❖ Usually, crystallization starts at the
bottom of the crucible by lowering
the temperature below the melting
temperature (1410◦C) of silicon.
❖ Within the Bridgman process the
temperature reduction is achieved by simply
descending the liquid silicon-containing
crucible out of the hot area of the
crystallization furnace.
❖ This process is also known as Gradient freeze
technique

Block-casting process
❖ In the block-casting process, the melting is performed in a quartz crucible without a
coating,
❖ Usually, crystallisation starts at the bottom of the crucible by lowering the
temperature below the melting temperature (1410◦C) of silicon.
❖ After pouring the molten silicon into
a second crucible – for the
crystallisation also a Si3N4-coated
one is used.
❖ The temperature control during the
block-casting process is achieved by
a corresponding adjustment of the
heaters, whereas the crucible itself
is not moved during solidification.
❖ After solidification starts in the
bottom region, the crystallisation
front, that is, the liquid–solid
interphase, moves in a vertical
direction upwards through the
 crystallization crucible.
❖ The directional solidification results in a columnar crystal growth and consequently
adjacent wafers fabricated out of the ingots show nearly identical defect structures
(grain boundaries and dislocations).
❖ Common crystallisation speeds used for the Bridgman technology are in a range of
about 1 cm/h (corresponding to a weight of approximately 10 kg/h for large ingots).
❖ Too high process speeds cause large thermal gradients within the solidified silicon
that may result in cracks or even destruction of the ingot.
❖ For the block-casting technology, however, owing to the more versatile and
sophisticated heater system, considerably higher crystallisation speeds can be
achieved.
Block casting: industrial furnaces

Silicon ribbon and foil production

❖ Conventional methods for producing the silicon material from which wafers are cut,
the Czochralski and floatingzone techniques, "grow" single-crystal* silicon in ingots
from molten silicon starting with single-crystal seeds; then the ingots are sawn into
wafers.
❖ These are complex, expensive processes.
❖ New fabrication processes try to reduce expense by changing raw silicon into single-
crystal silicon more cheaply, and by forming single-crystal silicon directly into
usable wafers while avoiding steps such as sawing.
❖ A new generation of ribbon technologies has moved past the R&D stage into large-
scale manufacturing and is in competition with these conventional approaches.
❖ Ribbon technologies, have now reached maturity with the start up of manufacturing
on a megawatt (MW) scale, include
• Edge-defined Film-fed Growth (EFG),
• String Ribbon (STR) and
• Silicon Film (SF).
❖ Moving to pilot demonstration phases
• Dendrite web (WEB)
 • Ribbon growth on substrate

❖ The ribbon technologies may be grouped into two basic approaches: “vertical” and
“horizontal” growth (pulling) methods.
❖ The EFG, WEB and STR methods are examples of the vertical method category,
❖ while both the RGS foil and the SF methods grow crystals in a horizontal-pulling
configuration with the aid of a substrate.
g. Procedure, advantage and disadvantages of edge defined film fed growth (EFG) method
Edge-Defined Film-Fed Growth (EFG)
❖ In this technique, the geometry of the ribbon is controlled by a slotted graphite die
through which silicon is fed via capillary action
❖ A seed crystal is lowered until it contacts the liquid in the capillary.
❖ The liquid spreads out over the top of the die to
the edges where it is pinned by surface tension.
❖ The seed is withdrawn, pulling the liquid up
while more liquid flows upward through the
capillary.
❖ As the ribbon is withdrawn, the liquid freezes on
the solid crystal.
❖ The die and the crucible are integral, that is,
made of the same piece of graphite.
❖ Temperature control of a few degrees along the
interface is sufficient to prevent ribbon pull-out
or freezing of a growing ribbon to the die top.
❖ Ribbons with thicknesses from 400 μm to as
little as 100 μm have been grown.
❖ The growth rate is controlled by how fast heat
can be conducted away from the interface and
lost by radiation or convection from the solid
crystal.
❖ Very high production rates of silicon can be obtained by pulling several ribbons
simultaneously from the same melt
❖ There are several problems associated with edge defined film-fed growth.
❖ The die must be of a highly refined material that has the proper capillary
characteristics and does not contaminate the silicon.
❖ So far, carbon has been the best die material, but it still introduces some impurities.
This problem is exacerbated because there is very little volume (between the die
and the single-crystal product) in which contaminants can dissipate
❖ To avoid contamination, a highly pure silicon melt must be provided as starting
material.
❖ Additionally, repeated use tends to erode the die.
 ❖ EFG does not produce single-crystal silicon because of impurities and stress
resulting from process geometry, the ribbon has crystal lattice defects.
❖ For this reason, PV performance of cells made from EFG ribbon has not been as high
as, for instance, with Czochralski-grown wafers.
❖ Cells from laboratory grown ribbons have attained 12% efficiency.
❖ Projections (Backus, 1976, p. 395) are that EFG can be reduced in cost 10-100 fold,
which is the amount needed to support overall goals for making PV generally cost-
competitive.

EFG – Properties
❖ locally high dislocation densities
❖ twin boundaries
❖ a twin boundaries is a symmetric grain boundaries
❖ the red line (the twin grain boundary) is like a mirror

h. Procedure, advantage and disadvantages of Dendrite web method

Dendritic Web (WEB) growth process for ribbon
❖ Dendritic web growth has yielded cells with efficiencies (about 15%) as high as any
among the single-crystal silicon processes, notably Czochralski silicon; however, the
cost of web growth has yet to drop to competitive levels.

❖ Dendritic web growth is similar to EFG, except the mold is eliminated.

❖ The dendritic web process is used for growing long thin ribbon crystals of silicon
and other semiconductors
❖ Growth is initiated from a thin wire-like dendrite seed which is brought into contact
with the melt surface.
❖ Initially, the seed grows laterally to form a button at the melt surface; when the seed
is withdrawn, needlelike dendrites propagate from each end of the button into the
melt, and the web portion of the crystal is formed by the solidification of the liquid
film supported by the button and the bounding dendrites.
 ❖ The performance of solar cells
fabricated from dendritic web material
is indistinguishable from the
performance of cells fabricated from
Czochralski grown material.
❖ Accurate temperature control, within a
few tenths of one degree, is necessary
to ensure a uniform ribbon width and
thickness.
❖ The temperature of the melt surface
must be constant over the width of the
growing web to prevent the dendrites
from growing in or out.
❖ Typical WEB thicknesses range from 100 to 150 μm and widths up to 8 cm have
been grown. Growth lengths between seedings in pilot production extend to many
tens of meters.
❖ The dominant impurity in the WEB in production today is oxygen since a quartz
crucible is used.
❖ The growth velocity is determined by the rate of removal of the latent heat into the
ribbon and of the heat conducted through the melt.
❖ Typical growth rates are from 1 to 3 cm/min. The relatively low production rate of
silicon crystal material by this process has been its major disadvantage

i. Procedure, advantage and disadvantages of STR

String Ribbon (STR) growth system
❖ In this technique, ribbon growth takes place
directly from a pool of melted silicon without a
die in a process mirroring the WEB
geometrically.
❖ Rather than dendrites, as with WEB, the
position of the ribbon edges in STR is
maintained by two strings fed through holes in
the bottom of the crucible.
❖ The strings are drawn upward out of the melt to
support the meniscus and the ribbon, and their
pull rate determines the growth speed of the
ribbon.
❖ The thickness of the ribbon is controlled by
surface tension, heat loss from the sheet and
pull rate.
❖ An important difference of the STR process
from WEB growth is that the constraints of
maintaining propagating dendrites and a
 supercooled melt are eliminated, and this relaxes the high degree of temperature
control required in the WEB furnace.
❖ Careful adjustment of the growth parameters can allow very thin ribbon, down to 5
μm, to be grown

❖ In the String Ribbon technique, two high temperature strings are pulled vertically
through a shallow silicon melt, and the molten silicon spans and freezes between the
strings.
❖ The process is continuous: long strings are unwound from spools; the melt is
replenished; and the silicon ribbon is cut to length for further processing, without
interrupting growth.
❖ This advantage in material efficiency means String Ribbon yields over twice as many
solar cells per pound of silicon as conventional methods.
❖ Additionally, the resulting distinctive shape of the solar cell allows for a high
packing density.
j. Procedure, advantage and disadvantages of RGS
Ribbon Growth on Substrate (RGS)
❖ In this growth technique, the silicon melt reservoir and die are placed in close
proximity to the top surface of a substrate, on which the ribbon/foil grows.
❖ The substrate may be graphite or ceramic
❖ The principle is to have a large wedge-shaped crystallisation front.
❖ The die contains the melt and acts to fix the width of the foil.
❖ The thickness of the foil is controlled by the heat-removal capacity of the substrate,
pull rate and surface tension.
❖ The direction of crystallisation and growth are nearly perpendicular.
❖ The area of the growth interface now can be very large compared to the foil
thickness.
❖ The latent heat is extracted by conduction into the substrate.
❖ Growth rates from 4 to 9 m/min have been demonstrated.

❖ Kerf losses due to ingot and wafer sawing can be avoided by solidifying the silicon
wafers directly from the melt by the Ribbon Growth on Substrate (RGS) process,
thus significantly reducing the wafer cost.
❖ However, up to now solar cells made from standard RGS material suffered from
shunting problems due to current collecting structures.
❖ This resulted in lower fill factor values and hence in lower efficiencies compared to
solar cells made from block-cast multicrystalline silicon (mc-Si) materials.