Thin film solar cells

Thin Film Solar cells
S. GOMATHY M.E.,M.B.A
AP(SrG)
KEC
S.Gomathy M.E.,M.B.A

Thin film solar cell
❖ Single crystals are expensive to produce and so there is a great deal of
interest in finding photovoltaic materials of less demanding material quality which
can be grown more cheaply.
❖A number of materials have been identified of which the best developed at
present are amorphous silicon (a-Si), polycrystalline cadmium telluride (CdTe)
and microcrystalline thin film silicon (P-Si).
❖These ‘thin film’ materials are usually produced by physical or chemical
deposition techniques which can be applied to large areas and fast throughput.
Note that the term ‘thin film’ refers more to solar cell technologies with mass-
production possibilities rather than the film thickness: GaAs p-n junction cells,
with an active layer a few µm thick, are thin, but do not belong to this class.
❖ Polycrystalline and amorphous semiconductors contain intrinsic defects
which increase the density of traps and recombination centres. For solar cells, this
has the consequence that:
❖ Diffusion lengths are shorter, so the material needs to be a strong optical
absorber.

Continued…
❖Alternatively, multiple junctions must be used to make the device optically
thick. In the case of very short diffusion lengths, it may be necessary to use
extended built-in electric fields to aid carrier collection. This is the case in
amorphous silicon, where p-i-n structures are preferred.
❖Losses in the layers close to the front surface are greater, so it is advantageous
to replace the emitter with a wider band gap window material.
❖ The presence of defect states in the band gap can make the materials
difficult to dope, and can limit the built-in bias available from a junction through
Fermi level pinning.
❖ The presence of grain boundaries and other intrinsic defects increases
the resistivity of the films particularly to low doping densities, and makes the
conductivity dependent on carrier density, so influencing the electrical
characteristics of devices.
❖ The presence of defects similarly means that minority carrier lifetime
and diffusion constant are carrier density dependent.

Thin film photovoltaic materials
Requirements for suitable materials
❖ Good thin film materials should be low cost, non-toxic, robust and stable.
They should absorb light more strongly than silicon. Higher absorption reduces the
cell thickness and so relaxes the requirement for long minority-carrier diffusion
lengths, allowing less pure polycrystalline or amorphous materials to be used.
❖ Notice how weakly crystalline silicon absorbs, in comparison with the
other materials. Suitable materials should transport charge efficiently, and should
be readily doped. Materials are particularly attractive if they can be deposited in
such a way that arrays of interconnected cells can be produced at one. This greatly
reduces the modules cost.
❖ Of the elemental semiconductors, only silicon has a suitable band gap for
photovoltaic energy conversion. Compound semiconductors greatly ex-tend the
range of available materials and of these a number of II-VI binary compounds and
I-III-VI ternary compounds have been used for thin film photovoltaic. Many of
these are direct band gap semiconductors with high optical absorption relative to
silicon. The I-III-VI compounds are analogous to II-VI’s where the group II
species. At present The leading compounds semiconductors for thin photovoltaics
are the II-VI semiconductors, CdTe, and the chalcogenide alloys, CuInGaSe 2 and
CuInSe2.

Amorphous silicon
Materials properties
❖ Amorphous silicon (a-si) is the best developed thin film materials and has
been in commercial production since 1980, initially for use in hand help
calculators.
❖ As a materials for photovoltaics, it has the advantages of relatively
cheap, low temperature (<300ºc) deposition and the possibility of growing on a
variety of substrates, including glass, metal and plastic, with diverse commercial
applications. The amorphous nature has several important consequences for
photovoltaics. Absorptions of visible light is better than for crystalline silicon, but
doping and charge transport are more difficult. The availability of alloys with
different band gap enable the design of heterostructure and tandem devices.

Defects in amorphous materials
❖ In amorphous materials, the lattice contains a range of bond lengths and
orientations, as well as unsatisfied ‘dangling’ bonds.
❖Although the nearest neighbors of any atom are co-ordinate almost exactly as in
the crystalline materials, the combined effects of small bond distortions means that
there is virtually no correlation between an atom and its more distant neighbors.
❖ The long range order of the crystal is gone. IN a- Si, Si atoms are
arranged in an approximated tetrahedral lattice but with a variation of up to 10º in
the bond angles.
❖ This increased absorption in one reason why amorphous silicon is of
interest for photovoltaics.
❖ The variations in SI-Si distance and orientation gives rise to a spreading
in the electron energy levels, relative to the perfect crystal.
❖This spreading appears as a tail in the density of states at the top of the valence
band and the bottom of the conduction band, known as a urbach tail.
❖ Dangling bond states are due to Si atoms which are co-ordinate only to three
neighboring Si atoms, leaving one valence orbital which is not involved in
bonding.

Continue….
❖ The dangling bond may be positively charged (D+), neutral (Dº) or
negatively charged (D-). an excess of D- states gives rise to n type a-Si while
excess D+ gives rise to p-type materials.
❖More defects can be created by irradiating the material, by heating and sudden
cooling, or by extrinsic doping. The dangling bond states give rise to energy levels
deep in the band gap.
❖ The defects in amorphous materials differ from those in polycrystalline
materials in that they occur uniformly throughout the materials and not only at gain
boundaries.

Absorption
❖ The loss of crystal order means that the absorption coefficient is higher
than in crystalline silicon.
❖ In crystalline silicon that band gap is indirect for most visible wavelengths and
so absorption of a photon requires the simultaneous absorption of a phonon to
conserve crystal momentum.
❖ Absorption thus in limited by the availability of phonons.
❖In amorphous silicon there is no well defined E-K relationship and no
requirement to conserve crystal momentum.
❖ Absorption depends simply upon the availability of photons and the
density of states in valence and conduction bands, much as it does in direct gap
crystals.
❖The absorption coefficient is about an order of magnitude greater than in
crystalline silicon (c-Si) at visible wavelengths.
❖ Passivation of a Si with hydrogen (discussed below) increases both the
absorption coefficient and the band gap.
❖ Material used for solar cells typically has a band gap of 1.7 eV. This is higher
than the optimum for solar energy conversion, but material with lower
concentrations of hydrogen is unusable due to poor doping and transport properties.

Doping
❖ The extra carriers which would be introduced into the conduction or
valence band by donor or acceptor impurities are captured by dangling bond
defects.
❖ However, the background density of defect states may be reduced by saturating
the dangling bonds with atomic hydrogen.
❖ Passivation with 5-10% hydrogen reduces the density of dangling bonds
to around 10 cm%^-3, and produces materials from which workable p-n junction
can be made.
❖The material may be doped n or p type but the doping efficiency is low.
❖ In amorphous silicon, however, the addition of a an electron upsets the
equilibrium between neutral and negatively charged dangling bonds, and drives the
following reaction to the right.
❖ When the density of dangling bonds is very high, the Fermi level is
pinned amongst the defect states.
❖Low doping efficiency means that the majority carrier activation energy– which
is the difference between Fermi level and band edge—is high.
❖ In p-type a Si the activation energy is around 0.4eV.
❖These large activation energies limit the size of the built-in bias which can be
achieved at a p-n orp-i-n junction, increasing the difference betweem the built=-in
bias and the band gap. S.Gomathy M.E.,M.B.A

Transport
❖ The defect states which remain after hydrogen passivation act both as charge
traps and recombination centers and dominate charge transport in a-Si.
❖ The distribution of tail states below the conduction and valence band edges
act as traps for mobile carriers.
❖Charge carries in these states move by a sequences of thermal activation and
trapping events.
❖ This distribution in energies leads to a distribution in the time constants for
any transient process, and a strong dependence upon the occupation of the states, and
hence carrier density.
❖ Consequently the usual transport parameters of mobility, lifetime, and
constant are density dependent and hard to determine.
❖Charge transport in such a defective medium is sometimes called dispersive
transport.
❖ With doping the defect density increases and diffusion lengths are much
reduced.
❖This means that carrier collection in a p-n junction would be extermely poor, and
consequently p-i-n structures are used.

Stability
❖ Amorphous silicon suffers from light-induced degradation known as the
stealer Wronski effect.
❖The defect density in a Si:H increases with light exposure, over a time scale of
months, to cause an increase in the recombination current and reduction in efficiency.
❖ it is believed that light energy breaks some Si-H bonds to increase the
density of dangling bonds.
❖The systems excited into in a higher energy configuration with more active defects.
Annealing at a few hundred degree centigrade allows the structure to relax, and the
dangling bonds to be resaturated.
❖For this reason a-Si solar cells may perform rather better in high temperature
environments.
❖ In a typical a-Si solar cell the efficiency is reduced by up to 30% in the frist
six months as a result of the stables Wronki effect, and the fill factor falls from over
0.7 to about 0.6.
❖this light induced degradation is the major disadvantage of a-Si as a photovoltaic
material.

Amorphous silicon cell design
Amorphous silicon p-i-n structures
❖ The basic a-Si solar cell is a p-i-n junction. Since diffusion length are so
shorts in doped a-Si, the central undoped or intrinsic region is needed to extend the
thickness over which photons may be effectively absorbed.
❖ The built-in bias is dropped across the width of the i-region, creating an
electric filed which drives charge separation.
❖In the p-i-n structure photo carriers are collected primarily by drift rather than by
diffusion.
❖ The thickness of the I region should be optimized for maximum current
generation.
❖Although more light is absorbed in a thicker region, charged defects reduce the
electric field across the I region, and at some thickness the width of the i region will
exceed the space charge width as shown.
❖ The remaining, neutral part of the i layer is a ‘dead layer’ and does not
contribute to the photocurrent.

Fabrication of a-Si solar cells
❖ Amorphous silicon solar cells are normally deposited on glass substrates,
which are coated with a transparent conducting oxide (TCO) such as tin oxide or
indium tin oxide.
❖TCO coated plastics are also being developed.
❖ Cells are usually fabricated in a ‘superstrate’ design, where layers of
conducting oxide, p-type, undoped and n-type a-Si are deposited in sequence. The a-
Si is usually deposited by plasma decomposition of silence or ’glow discharge’, but a
number of other deposition methods such as sputtering and ‘hot wire’ are being
investigated.
❖For the rear contact, zinc oxide is deposited on to the n layer followed by a metal,
usually aluminum.
❖ Light trapping structures may be built-in by texting the front TCO layer,
and metallising or texturing the back surface in order to enhance light absorption.
❖ An alternative is the ‘substrate’ design where layers are deposited on a
metal substrate, such as steel, which forms the back contact.
❖ Here the p not need to be flat, and so higher efficiencies are possible.
❖ However, the substrate design is not so easy to process

Strategies to improve a-Si performance
Light induced degradation
❖ The staebler Wronski effect is the most important barrier to widespread use
of a-Si solar cells.
❖ Light-induced degradation is stronger in materials with a high hydrogen content,
because of the greater density of Si-H bonds, yet a high hydrogen content is needed for
suitable doping and transport properties.
❖ One possibility is the ‘hotwire’ technique which appear to produce good a-
Si when saturated with only 1 %
Improvement of voc
❖ The open circuit voltage in a-Si solar cells is substantially less than the
optical band gap (0.89 v compared to1.7eV) on account of the high activation energies
in amorphous material, and resulting low built-in bias.
❖ Voc can be increased by the use of either (i) a wider band gap emitter such as
a-Sic:H or(ii) a polycrystalline Si emitter, in which degenerate doping is possible.

Continue..
Improvement of Jsc
There are two problems:
i. Response to the blue light is poor in homogenous a-Si:H cells on account of
poor collection in the p layer. This can be resolved by replacing the a-Si p layer
with a wider band gap a-SiC:H window.
ii. Response to long wavelengths may be poor because of the limit to the i-
region thickness which arises from charged background doping in the i-region.
Improvement of limiting efficiency
❖ Multi-gap cell designs are possible using a-SiC:H as the material for a wider
gap cell and a-SiGe:H for a narrower gap cell.
❖Two-terminal cascade designs where the different p-i-n cells are collected in series
using tunnel junctions have been studied.
❖ The limiting efficiency for three cell devices is calculated at 335.
❖ the main problems have been the poorer quality of the alloy relative to a-Si
incorporating large area tunnel junctions.

Defects in polycrystalline thin film
materials
Grain boundaries
A polycrystalline materials is composed of microcrystallines or ‘grains’ of
the semiconductor arranged at random orientations to each other. The material is
crystalline over the width of a grain, which is typically the order of one µm.
The transport and recombination properties are strongly affected by the
presence of the interfaces or grain boundaries.
The different orientations of neighboring crystal grains give rise to
dislocations, misplaced atoms (‘interstitials’), vacancies, distorted bond angles and
bond distances at the interfaces.
Polycrystalline materials are also likely to contain extrinsic impurities
which contaminated the materials during growth. These extrinsic impurity atoms are
likely to concentrate at the grain boundaries.
The various types of defect introduce extra electronic states. These extra
states are spatially localized and because they do not need to obey the symmetry of
the crystal and they may have energies in the band gap, these intra band gap states
tend to trap carries and we will refer to them as ‘intra-band-gap states’ or simply
‘trap states.’

Continued…
❖ Because these intra band gap states are able to trap charge, they
influence the potential distribution close to the grain boundary.
❖ The electrostatic force sets up a potential barrier which opposes
further majority carrier migration. Minority carriers, however, see a
potential well at the grain boundary and are pulled towards it, where the
probability of recombining with a trapped majority carrier is high.

Effects of grain boundaries on transport
❖ The effect of grain boundaries on charge transport depends on
whether they lie normal to or parallel to the direction of current flow.
❖ In the first case, when current is flowing across a grain
boundary, the potential barriers slow down the transport of majority
carriers, limiting the majority carrier mobility, while the potential
wells drive minority carrier diffusion length and lifetime.
❖ The sizes of these effects depends upon the doping, the density of
interface states, and the photo generated carrier density.

The main result are that:
❖ Increasing the trap density increases the space charge stored at
the grain boundary, which increases the barrier height, reduces
conductivity and increases recombination.
❖ Increasing the doping first increases the barrier height, but at
higher doping levels the traps become saturated, the space charge region
begins to contract and the barrier is reduced.
❖ Increasing the density of free carriers by illumination reduces the
net charge stored at the grain boundary and hence the barrier height. At
high illumination levels the grain boundaries have the minimum effects:
conductivity reaches its maximum level, and recombination with trap
states saturates.

Continued…
❖ Grain boundaries which lie parallel to the direction of current
flow principally affect minority carriers. Majority carriers travelling
parallel to the grain boundary are not affected– they see no barrier– but
minority carriers are still likely to be trapped in the potential well and
recombine.
❖ When a grain boundary actually crosses the p-n junction, it
reduces the efficiency of charge separation by competing with the p-n
junction for minority carriers.
❖ A more serious problem arises when doping impurity atoms
from the emitter diffuse along the grain boundaries, through the nominal
p-n junction, into the base. This creates a shunt path trough the p-n
junction which reduces the rectification of the junction.

CdTe solar cell design
❖ The preferred design is a n-CdS p-CdTe hetero junction cell. As
in CuInSe2 based cells the CdS emitter acts as a window to improve
collection at short wavelengths.
❖ In this case, no barrier result in the conduction or valence band,
but lattice mismatch between the two materials leads to interface states.
❖ About 3-5 µm of CdTe are needed for sufficient optical depth.
This is generally weakly doped but a second doping treatment, e,g. with
copper or lithium is needed to improve the conductivity of the CdTe layer
at the rear contact.

Problems facing CdTe cell design are as follows:
❖ The CdS emitter still absorbs significantly in the green and very thin
(50—80nm) layers are needed for good response to green light. One objective
in CdTe cell development is to reduce the thickness of the CdS layer in mass
production.
❖ The high density of trap states at the grain boundaries give rise to a
high dark current (about 20 times that the otherwise similar material GaAs) and
low Voc (0.2 V less than GaAs). The low Voc is partly attributed to Fermi level
pinning at the trap states.
❖ Fermi level pinning by the trap states leads to difficulties doping
CdTe heavily p type and making ohmic contact to the substrate.
❖ Defects at the CdS-CdTe hetero junction, due to the formation of
other chemical compounds (such as CdTeO3) enhance junction recombination.
❖ The most efficient CdTe cells is a 16.4% efficient, 1cm² CdS/CdTe
device on glass, produced by the US National Renewable Energy Laboratories.
The challenge is now to improve CDtE solar cell production technology so as
to increase the efficiency of mass produced cells towards the values for lab
cells.

Thin film silicon solar cells
❖Materials properties
❖ Thin film microcrystalline silicon is characterized by grain sizes
of around 1µm. It has the optical properties of crystalline silicon, while its
electronic properties are dominated by the grain boundaries.
❖ Defects states at grain boundaries include dangling bonds
typical of amorphous silicon as well as extrinsic impurities introduced
during growth.
❖ The defects states appear to be distributed through the band gap
so that the gain boundaries are active in both n and p type material.
❖ Microcrystalline silicon (µ-Si) can be prepared by the same
techniques as multi-crystalline, normally by casting of molten silicon into
aggregates or sheets.
❖ A range of other techniques have been investigated for thin film
polycrystalline materials, including liquid phase epitaxial, chemical vapor
deposition and the crystallization of amorphous silicon.

Microcrystalline silicon solar cell design
❖ A 2µm thick µ- Si silicon solar cell on glass has been reported
with to have an efficiency of over 10%
Light trapping in thin film Si solar cells
❖ Relatively thin cells can be made from polycrystalline silicon if
light trapping techniques are used to increase the optical path length inside
the cells.
❖ Texturing front and back surfaces increases the optical depth by a
factor of around 20, which means that only a few tens of microns of µ-Si
are needed to absorb most of the incident light.
❖ This relaxes the need for a long diffusion length, and allows
higher base doping, which increases Voc.

Microcrystalline silicon solar cell design
Parallel multi junction thin film silicon solar cells
❖ An alternative concept in thin film silicon is the parallel multi
junction solar cell, where the cell is composed of consecutive micron-thick
layers of p and n type microcrystalline Si. Layers of similar polarity are
connected together to give a set of p- n junction connected in parallel.
❖ Because the layers are of similar thickness to the minority carrier
diffusion length, the probability is high that a photo generated minority
carrier diffusion laterally towards the contacts.
❖ The cells perform better than single junction µ-Si cells of the same
thickness because the fraction of minority carrier lost by recombination is
smaller.

Thin film solar cells

More Related Content

What's hot (20)

Similar to Thin film solar cells (20)

More from Gomathy Sengottaiyan (17)

Recently uploaded (20)

Thin film solar cells