Thin film solar cells: amorphous silicon, CdTe, CIGS and organic cells

Thin film solar cells: amorphous silicon, CdTe, CIGS and organic cells


[MUSIC] All right, so what I want to focus on
today is these four, tenform technologies. And these are all made on, amorphous
material and we’ll use some of those concepts to
understand this technology. And most of these are located at this
bottom of this curve, so there’s a you know famous quote saying it’s love the
fortune is always at the bottom of the pyramid. So this, there’s lot potential over here,
the efficiency of now are all of these think film cells, they max
out at maximum of 20% for a six cell, most of the organic cells, they
max out at 10 and 11%. But there’s, there’s there’s still a lot
of potential, or Lacks some kind of options, which exist over there.
So, let’s go look at them one by one, and so here’s some.
Let’s look at amorphous silicon first. Most of these amorphous silicon cells,
which are shown in that green color, or that opal green there, line up over here.
They max out at efficiencies of around 11% and amorphous silicons. Let me give you some facts about amorphous
silicons. Amorphous silicon has a band gap as we saw
of around 1.7. It has a electron mobility of 20.
It has a whole mobility maximum of two. So, question, which, which of these
configurations assuming light is coming front the top is better for your amorphous silicon solar
cell? Is it this PIN which would be better or
would it be I mean this is better or this is
better? How many people? So this is, is much better because most of
your carriers are not most, but a large fraction of your carriers
would be generated at the very top, especially your blue wavelengths are
absorbed at the very top. And if you have an inservice over here
your electrons would be collected more but your holes would have to you
know, diffuse through over here. And since it’s amorphous material, the
recombination rates are much higher, mobility is much lower.o A lot of them will recombine in this configuration.e So this
configuration is preferred for for amorphous Typically most amorphous
silicon cells you will see that it’s the p material at the
top because when the carriers are absorbed at the top, you want to collect the wholes
very quickly and let the electrons use their mobility to travel
until here and then get connected. So, this is kind of intuitive of what we saw in, in crystal
and silicon, right? So I’m saying over here that amorphous silicon is, is mostly configured like
this. Last lecture we saw that crystal and
silicon is configured like this. So why is crystalline silicon like that
and why? Why is this ambiguity among? So the reason crystal and silicon has been
like that is because the P plus region has been
created typically by this backside feel, which comes from your
aluminum. So because of this aluminum layer, you get
this P plus layer for free, that’s why. And the difference between your, or the
ratio of your electron and holes mobility, in the case of silicon it’s
four, over here it’s like 10. And also there mobilities are much higher
over here. But even for this P plus on the top and and, and plus on the bottom would have
been better. And most of all, many people are starting to use this backside passivisation and you
are actually don’t use that field from
aluminum, you use aluminum oxide or, a dielectric layer
to create that field. So in that case in fact, even in this case
P on the top, N becomes better. So, [COUGH] so this is a typical, how a sin, single junction amorphous silicon
cell looks like. And again it’s, it’s a very simple process
layer of flow to make this. So you start with either a, a plastic or a
glass substrate and the first thing you put is your metal electrode and that’s a transparent metal
electrode. So, it’s a metal oxide, typically a zinc oxide or a tin oxide, right? And then you grow these different PIN
layers, and they’re all grown by a CVD process
typically, which grows all three of them together so you
grow, You put in a tool and you PIN layers. And then you cap it up with another oxide
or simply cap it up with a material. And that’s how you get the cell. And then you essentially the way it’s
shown now is then you invert it and then you get
light. But efficiency of these most of these
amorphous silicon crystalline cells, they saturate
out at 8% because of the reason that we just
discussed about all the bad things about amorphous
material. The mobility is low, recombination is
higher. So if you use a thicker film, it doesn’t really give you increase in increase in
your cell efficiency. So the thing that people have tried, and
not, with not with too much success, I’m showing this from you know, a two 237 course is that people have tried to go
tandem. Even in amorphous silicon cells, so they have tried to combine this amorphous
silicon which has a band gap of 1.7, with either
amorphous silicon germanium, or they have tried to combine it with a
multi crystalline cell.
And again the when you make a tandem cell, two important things to keep in mind
are, you need to choose a bandgap in the right way, so
the upper material should have a band higher bandgap, so it’s typically
amorphous silicon. And then on the lower material you have an
option of using either, a multi crystalline silicon
or an amorphous silicon germanium. And for all these tandem cells, since they
are connected in a series you need to match
their current. So, you need to choose the thickness of these layers such that their currents are
matched. Shown here is the configuration which is
also known as micromorph cell.
Are tandem of amorphous and micro crystalline.
So your amorphous silicon is a very thin layer which is on the top, and the below
is micro-crystalline the based cell. And they’re again grown in the same CVD
tool. It’s just that you grow this at a higher
temperature so it becomes micro crystalline. And the top one is grown at the lower
temperature so it’s still amorphous. The, when you take SCRTM, they don’t look
very pretty. These are amorphous materials so they look
something like this. One of the main problems with the
amorphous silicon based cells is that they degrade
with light. So, if you amorphous silicon cell and let
it lie in light for a large period of time, your efficiency or your
power will degrade as a function of time. And these degradation is, is even more
severe when you have these tandem cells, where you have a tandem of
amorphous silicon and micro-crystalline. So, it’s like, you know, you’re running in
a relay race and if one of your runners is slow, it
becomes, it brings everybody down. So, if you connect things in tandem and
one of them degrades, it degrades the whole
thing even more. So that’s why your your efficiency falls
or your delivered power falls as you soak it in sunlight.
The good thing is that it can be recovered back, so if you expose it to hot temperature again that efficiency
comes back. So the reason which occurs is is because
sorry the reason it occurs is because in this
amorphous silicon. You have these lot of hydrogen states, which are hydrogen molecules which are
bonded to your other dangling points, and when you
expose it to light that bond breaks, and you create more defect states. But, if you, again go to higher
temperature, it gets recovered back. So, it’s it’s a known problem, and this
light induced degradation is typical of all
these thin film technologies. So most of these amorphous materials, they
the bonds tend to break if you soak them to sunlight for a
long time. So they show much higher degradation as compared to crystalline solar cells. And then, the way people have been manufacturing, this is, you know, using
large tools, so this, this is the same tool that
I showed you for display. Most of these people try to modify, these
display-based tools, which are huge tools, could make panels of, 5.6
meters square size. But, at the same time, it adds to your
CapEx, and just like we did a bill of material
analysis, for the iPhone, people actually do bill of
material analysis for each of the process steps that you are
doing. How much material cost is coming. You know, it’s really counting the
pennies. You know, how much utilities are there,
how much rent is on the facility. Everything’s most of these amorphous silicon is cells, they
first of all max out at efficiency of 10%. If you need more than that, you need to go to this complex structure, but then it
degrades very fast with light so your light and if the degradation can reduce your efficiency
by 20% again. And then it requires these large and big
tools. There are a lot of action in these field
around 2007, 2008 but since then it has all cooled down because
of all of these limitations. So, amorphous silicon is is no longer a leading candidate for thin film
technology. So, because of all these reasons, people
thought you know, let’s look at another material and
that’s what I’ll, I’ll describe next. All right so the next thing, if you look
at again, this chart. And I love this chart because, you know,
it summarizes so much of information. The next thing beyond this amorphous
silicon that you have an efficiency is this cad
tel based technology. And it has been researched for a long time
and then it, it started to commercialize around 2000.
The company was started to commercialize this for the first solar.
And in terms of cost, it’s again, if you compare it’s much cheaper as
compared to crystalline silicon even, much cheaper
as compared to [UNKNOWN] and and, this is what you know, first solar typically shows.
So it shows that your, it has been able to reduce it even further, and now they sell
it for around $0.60 per watt. And a lot if has come through efficiency
improvement, some of through, some of it has come through,
you know, scale, scaling up their plant, improving
the throughput and so on. [COUGH] So Cad tel, if you look at the material
itself, it’s a direct bandgap material and it has a bind
gap of 1.5, which falls within that optimum peak.
For your optimum efficiency of a single crystalline solar cell is between 0.6 to
1.5 [UNKNOWN] band gap. So the bind gap is nearly ideal and it has
a very high absorption coefficient as i see, so you don’t you
need to take off for material. It is a strong direct band gap absorber. So people have been mucking around with it
since a long time you know. The the first thing that you need to make
this cad tel cell is that . Again, you can’t dope it too high so you need that hetero-structure contact to
extract the electron. So the way it works is, you have cad-tel
layer and then you have a cad-sulfide layer on the top and they
form this type two kind of lineup. So your electrons, which are generated in
your in your cad-tel could be easily by this
cad-sulfide layer. And so again it falls back to that initial
description that, you know that PN junction is a very limited
definition of a solar cell. a more complete definition is that you all
you need is a photovoltaic action, so you need a
density of state button link, so you need a band gap,
and you need this selective contact to
separate your electrons and wholes. So, the way this selective contact over
here is made by, is by having this cad-sulfide
cad-tel hetero junction. So, this, too, has been, you know, played
around till the 1990’s people were, they already knew all this concepts
and they were playing around with it. But their efficiency never used to exceed
more than 10 or 11%. What First Solar did was they figured out,
a very good way to pacivate the cell and you know that
reduced some of these defects that you get in in this amorphous silicon material and
they went to, they now have a cell efficiency of 16.5
with this cell. And what they did was they, they figured
out this treatment that’s a cad chloride based
treatment and that they first deposit the whole stack, and then anneal
it in this gas and it improves your lifetime and your properties of this
cad-tel material very significantly. And they didn’t, you know, publish, not,
they still aren’t publishing, they really
didn’t reveal it for a long time, so for a long time it was a
well-kept secret how does their cell work. And they’re the only supplier which sells
this in, in volume. [COUGH] So that, that’s about cad tel. So but cad tel has these two very big
problems. That is the two letters in cad tel.
Each of them is a problem, trillium is a rare earth metal, and it’s in fact, this
is your plot showing how much is the quantity available and you see
that Trillium is right at the bottom, it’s one of the rarest metals.
So, one of the concern always with their technology is that you
know if trillium becomes short in supply. The costs will go up but that so far
hasn’t happened they’ve been able to find new new resources in a whole state in china band which has huge amount of
trillium. So it was said to be one of the borderlinks, but so far it hasn’t
been so. Another thing which is a border link is
cadmium is a poisonous materials. EU has this specification called
restriction of hazardous substance, so it comes under
that. In fact, these cells are not allowed to be sold in Japan because they need special
recycling ones. So, you know they have these cells
typically sell like with a lifetime of 20 years. So after 20 years, you need to recycle
them. In fact First Solar includes $0.05 charge
in their panel costs for their recycling. But still some countries don’t allow these
cells at all. So, their efficiency maxes out at about
16%, that’s the maximum people have been able
to go. But So, First Solar, including First
Solar, they have been looking into alternates to [UNKNOWN] and they in fact have a big group in
Oakland, around 80 people who were trying to look at this alternate technology,
which is the CIGS based solar cell. So the next, if you look up again, this,
in this efficiency later, the next thing after Cartel is
these CIGS based solar cell. So by CIGS, what I mean is copper, indium, gallium, selenium.
So [INAUDIBLE] CIGS is You can have any combination of
these materials in your group 11, 13, and 16 so you could pick copper then
indium, galium, so CIG. And then either of these sulfur selenium,
telium too, and all of them will give you a direct uh,bang
up material. The good thing about CIGS is that you can
also tune your Indium and gillium ratio. So it’s typically copper in them and them
two atoms of selenium and you could use that to tune your bang up from from
one to our one point to. And this again made in a way similar way. You start with a glass, you deposit the
layer there and you have the stack at the top and then
another DC on the top. so, you could make this thing either on
glass, or you could make them on sheets of metal. You could even make them on. So starting material could be glass, it
could be a flexible material. And it could be in sheet of metal, so
there are all the startups which are in the six
pales they have. Each of them has a unique process and each
of them has the unique starting material to
make these sigs material. The bottleneck for this is that it
requires four elements.
So Let me see, CIGS has least four elements.
And you have to be part of a fixed ratio across that big pan and that’s
one of the main bottlenecks, how you deposit on a big
panel that same symmetry of these four elements which all have to be (no period). So,it’s actually a good experiment to do
in a University lab where you have four different elements and you have a
filter which is a multi-target filter and you’ll get all different combinations
on one wafer and you can measure each of those and publish a nice chart,
which has all kinds of efficiencies, but If you manufacture this in an industry
you’ll need to get the same prospects all the
time. So each of these companies, you know a lot
of them are based here in the Bay Area. They are essentially either a processor or a company, a telecom company but each of
them have a different process to deposit this
(no period) And some of them are still surviving, but, you know a lot of them
are, are no longer this slide is from 2010, so a few of them are actually
just on the brink of collapse. This one actually went bankrupt. This one was.
Acquired by another Chinese company for, I think, one tenth of the
amount that was raised in terms of [UNKNOWN] A few of them other are also on the brink
of collapse. So, it’s again choosing the right material, choosing the right
processes is very important in the [UNKNOWN] material, because it’s all about can you
make them at a low cost and do you have a road map to increasing the
efficiency for them at the same cost? So that’s Those are very important,uh,
things, to answer for, So the final thing I want to describe, and I’m actually like this alot.
Its at the very bottom of this pyramid and it has the lowest efficiency in organic
based for all that it says they really suck in terms of efficiency.
The maximum you get is around 10%. But there’s a lot of potential in the
thing. So the first thing we need to understand
is, is how it works. So you know.
this is the value proposition for organic [INAUDIBLE] that you know you can get these organic
molecules which are you know, a way routinely used in, in
factories and laboratories. And there’s a lot of abundance of this. No scarcity, you can actually manufacture
them as you want. Very low cost. Efficiency, maximum 8% but, if you can
make them very cheaply, and if you actually achieve a modular efficiency of
30% you could achieve grid parity. So, it’s very important to understand how
it works. So for a lot of time people try to generate gate Electricity out of these
organic molecules. And, they applied a lot of current, they
applied a very high electric field, but you know, you can
never get efficiency of maximum of, you know, 0.1%, no matter what you
did. And The reason is that it when these,
these electron and holes in organic molecule, they’re not
generated like electron and hole separate. They’re generated as a exciton pair. So, what people discovered then was, you
know, instead of having that one layer Again, this hetero
junction. If you use this hetero junction like this,
where you have a donor material, which has a higher elec, lower electron
affinity and exceptor layer, which has a higher electron affinity, you could immediately get,
people started to get efficiency of 0, 1, 2, 3,
4, 5 percent. By just applying these two layers instead of one. So let me describe you in a little more
detail how does it work. So this, having these two layers in this
heterojunction is critical for having, making organic based so you need to have
this Donor and Acceptor list. Typically made of two different organic
molecules. And they need to be aligned in this space. The Donor has a low electron affinity as
compared to Acceptor. And the way it works is that when when an
incoming photon comes, it generates a the first thing it does is it generates
this electron in whole pairs, but this is generated as a x
ion, and it has a [INAUDIBLE] efficiency of, let’s say 15 percent.
And then it you know, then it and the absorption lens would be this, typically, around 100 Millimeter, for
these, organic, molecules. And then, immediately, for this generation
happens very fast, it happens on a time scale of ten to power minus ten,
minus 16 second. And then it relaxes to this intermediate
state, which is, which is called a excitation,
exiton relaxation. And this, it forms this also very fast in ten to power minus, 30
second. And then this is essentially a stable
state. And it has a certain binding energy. And to break the state you, what you need
to do is take this exciton and diffuse it to
this tetra junction. And that tetra junction, so this is the
main numerating step that you need to, you have generated
that exciton. That’s okay. But now you need to take this to this
hetero junction, and so it has to diffuse through this third diffusion length, and
that’s typically, you know, 15, five, or between five to 20
nanometer. And only if it’s within that diffusion
length, it will diffuse to this this hetero
interface. And when it does then it’s actually breaks
up where it separates this exciton state, separates very easily
. it gives away this electron to this other
guy and the hole very nicely goes to the other
contant. And this happens at a very high efficiency
as well. So this is also very fast process and this
is also very efficient. So the bottleneck are you know the overall
efficiency of this organic [INAUDIBLE] is is limited by.
Generating this exciton in a region of, you know, 10 to 20
nanometers from this interface. So, over at the generator, within this 10
to 20 nanometers of this interface, is absorbed with, you know, almost a way higher quantum efficiency, but Everything
always essentially just, just not connected. So, but this, you know, this means, you
know, there’s something which is nice. And, you know, we could engineer it to,
you know, some solutions around it. So, if you look at the overall efficiency,
it’s essentially a multiple of these all these different efficiencies
you know for absorption then exitation diffusion and they you know separation of these you know exitation in then their
collection. And all of these otheres are you know
either 100% or at least half but the limiting step is this exitation
difusion and that happens only in the region of. twice the diffusion length from this
interface. And this is what limits the efficiency of
these cells. So, if you measure if you measure like you, what the absorption,
what is the efficiency of these cells. You see, you know, that most of these are,
these are not limited by, you know, the hole easily you collect
their income line on anything. They’re limited by How much you are able
to absorb within that diffusion [INAUDIBLE] absorption for different wavelengths.
Is very, what limits your efficiency of [INAUDIBLE]. So, there have been you know, a lot of ideas
Which people are working on to improve this organic
photovoltaic efficiency.
So remember, most of that useful electron and holes that we collect are
generated at that interface between our [INAUDIBLE] except a layer in this organic photo [INAUDIBLE] . So what people have been trying to do, at
least in the research community, is, is trying to
maximize that interface, again, this one of the devices that tries to leverage
that interface, that tries to maximize that So you can collect
as many of these [INAUDIBLE] and break them into electron and hole, and
collect them. So, one way to maximize that interface is that, have this highly four-layered
hetero-junctions. Essentially, these are randomly located
acceptors and donors, but you want to connect all the acceptors on one side, and
all the donors on. By the other side.
So this is one scheme of doing that. And lot of dye synthesis base mattered
trial, which would rate this kind of profile between your profile and acceptor
molecule. The other way to do it, is have a
controlled junction again. You are maximize that interface, so you
can collect that interface is would Breaks this
exciton into electron and [INAUDIBLE] . So again you want to maximize that
interface. So one way to do this is have this regular
controlled you know, hetero junction between your acceptors and
donors. Well these are all active areas of
research but I find the technology is pretty neat And, it’s, it’s cheap to start
with, and then you can, you have further, room for optimization.
And that’s why you see a lot of, there’s quite a few faculty at Sanford who work on
the, on the organic office These materials you could
deposit them either by DVD kind of process that
gives you good film quality but it it wastes a lot of these
materials. So a lot of that material would be you
know, deposited on your mask. Another way to do it is, you know, have
this roll to roll coater. So these are.
big, you know, the material, the [UNKNOWN] material is like a roll, and it’s [UNKNOWN] on this big film. And there are other techniques like dye,
like solution based processing. So, all of these are are ways you can make deposit these organic layers and
make organic [UNKNOWN]. [MUSIC]

5 COMMENTS

Leave a Reply

Your email address will not be published. Required fields are marked *