WEBVTT

00:00:00.060 --> 00:00:02.500
The following content is
provided under a Creative

00:00:02.500 --> 00:00:04.010
Commons license.

00:00:04.010 --> 00:00:06.350
Your support will help
MIT OpenCourseWare

00:00:06.350 --> 00:00:10.720
continue to offer high quality
educational resources for free.

00:00:10.720 --> 00:00:13.340
To make a donation or
view additional materials

00:00:13.340 --> 00:00:17.226
from hundreds of MIT courses,
visit MIT OpenCourseWare

00:00:17.226 --> 00:00:17.851
at ocw.mit.edu.

00:00:25.930 --> 00:00:28.690
PROFESSOR: So folks,
we're going to get started

00:00:28.690 --> 00:00:31.960
into thin films for a moment,
but I saw two of you, at least,

00:00:31.960 --> 00:00:37.080
in the class at Eli Yablonovitch
talk on, gosh, what was it,

00:00:37.080 --> 00:00:37.810
Tuesday?

00:00:37.810 --> 00:00:39.110
Tuesday is was.

00:00:39.110 --> 00:00:41.821
How many attended the
talk-- show of hands?

00:00:41.821 --> 00:00:42.320
Three?

00:00:42.320 --> 00:00:43.710
OK, three, awesome.

00:00:43.710 --> 00:00:46.960
I must have missed one of
you-- very interesting talk.

00:00:46.960 --> 00:00:48.580
This was a talk
about solar cells

00:00:48.580 --> 00:00:50.430
given from the
perspective of somebody

00:00:50.430 --> 00:00:52.050
who does light management.

00:00:52.050 --> 00:00:55.760
And so I wanted to share with
you a book that is essentially

00:00:55.760 --> 00:00:59.260
from where he takes his
efficiency calculations, which

00:00:59.260 --> 00:01:01.570
are based largely
on thermal dynamics

00:01:01.570 --> 00:01:04.349
and less on the
continuity equations--

00:01:04.349 --> 00:01:07.150
Peter Wurfel's book
Physics of Solar Cells,

00:01:07.150 --> 00:01:08.660
a brilliant, brilliant book.

00:01:08.660 --> 00:01:10.110
I'm going to pass it around.

00:01:10.110 --> 00:01:14.740
On page 33, very easy number
to remember-- 2 times 3, 3.

00:01:14.740 --> 00:01:17.570
On page 33, he starts
delving into the derivation

00:01:17.570 --> 00:01:20.860
that Eli Yablonovitch
presented during his talk,

00:01:20.860 --> 00:01:24.010
so folks can follow along from
a thermodynamics point of view

00:01:24.010 --> 00:01:26.260
and maybe read up a
little more and understand

00:01:26.260 --> 00:01:27.620
that perspective.

00:01:27.620 --> 00:01:31.260
But he very, very
briefly touched

00:01:31.260 --> 00:01:35.017
upon essentially
the same physics

00:01:35.017 --> 00:01:36.600
but from the perspective
of what we've

00:01:36.600 --> 00:01:38.570
been talking about a
class in terms of carrier

00:01:38.570 --> 00:01:40.140
densities and current flows.

00:01:40.140 --> 00:01:41.840
He had it on the
bottom of a slide,

00:01:41.840 --> 00:01:43.537
perhaps halfway
through the talk,

00:01:43.537 --> 00:01:44.870
on four different bullet points.

00:01:44.870 --> 00:01:47.580
Does anybody remember
what those were?

00:01:47.580 --> 00:01:50.100
Why did he achieve such a
high efficiency conversion

00:01:50.100 --> 00:01:53.650
efficiency with the
gallium arsenide cell?

00:01:53.650 --> 00:01:54.825
Anybody remember that one?

00:01:54.825 --> 00:02:00.192
He had a thin device, so by
thinning the device down,

00:02:00.192 --> 00:02:01.900
if he's able to
concentrate the carriers,

00:02:01.900 --> 00:02:03.750
in other words, if he's able
to collect all of the charge

00:02:03.750 --> 00:02:05.760
carriers inside of
that very thin layer,

00:02:05.760 --> 00:02:08.889
he'll have a higher
charge carrier density.

00:02:08.889 --> 00:02:10.610
And the charge
carrier density is

00:02:10.610 --> 00:02:13.190
what influences the
separation of the quasi Fermi

00:02:13.190 --> 00:02:16.040
energies, which is what
influences the voltage output

00:02:16.040 --> 00:02:17.270
of the device.

00:02:17.270 --> 00:02:19.050
So he was able to
obtain a higher voltage

00:02:19.050 --> 00:02:21.380
output because he had
a thinner solar cell.

00:02:21.380 --> 00:02:24.490
He was able to concentrate the
carriers in that thinner region

00:02:24.490 --> 00:02:27.440
by light trapping,
by light management.

00:02:27.440 --> 00:02:30.104
And so as a result of having a
higher carrier concentration,

00:02:30.104 --> 00:02:32.270
he had a higher separation
of the quasi Fermi levels

00:02:32.270 --> 00:02:35.130
and hence a higher voltage
output of his device.

00:02:35.130 --> 00:02:37.740
So in reality, it was very
simple from the perspective

00:02:37.740 --> 00:02:39.490
of what we've been
learning in class here,

00:02:39.490 --> 00:02:41.656
how he was able to obtain
the very high efficiencies

00:02:41.656 --> 00:02:43.040
of gallium arsenide.

00:02:43.040 --> 00:02:45.100
The physics is well known;
it's not new physics.

00:02:45.100 --> 00:02:48.560
It's actually quite old physics,
and that that approach has

00:02:48.560 --> 00:02:51.530
been used within the crystalline
silicon solar cell community

00:02:51.530 --> 00:02:52.970
for some time as well.

00:02:52.970 --> 00:02:55.080
The back surface reflectors
off of the devices

00:02:55.080 --> 00:02:57.610
are highly optimized
and the texture,

00:02:57.610 --> 00:02:59.560
as well, to scatter the light.

00:02:59.560 --> 00:03:01.710
So I would invited
you to take a look,

00:03:01.710 --> 00:03:03.240
and this is another
example of how

00:03:03.240 --> 00:03:06.980
technologies can flow from
one photovoltaic system

00:03:06.980 --> 00:03:07.970
into another.

00:03:07.970 --> 00:03:10.530
So you can learn a lot
from material systems

00:03:10.530 --> 00:03:13.350
that you aren't working
on necessarily yourself.

00:03:13.350 --> 00:03:15.940
That's another take-home
message from the talk, at least

00:03:15.940 --> 00:03:16.940
what I walked away with.

00:03:16.940 --> 00:03:18.523
Any other impressions
that folks would

00:03:18.523 --> 00:03:20.952
like to share before we
dive into the lecture?

00:03:20.952 --> 00:03:21.452
Yeah.

00:03:21.452 --> 00:03:25.220
AUDIENCE: I just have question
about carrier collection.

00:03:25.220 --> 00:03:27.575
How is it possible
to extract any energy

00:03:27.575 --> 00:03:30.300
from carriers
which are generated

00:03:30.300 --> 00:03:31.934
in front of the junction?

00:03:31.934 --> 00:03:33.850
Because even if they
diffuse another junction,

00:03:33.850 --> 00:03:35.705
they have nothing to fall down?

00:03:35.705 --> 00:03:37.830
PROFESSOR: OK, so you have
to think about it always

00:03:37.830 --> 00:03:39.370
from the perspective of
the minority carrier.

00:03:39.370 --> 00:03:41.250
So if you generate an
electron-hole pair,

00:03:41.250 --> 00:03:44.039
your minority carrier is now
a hole, in the n plus region.

00:03:44.039 --> 00:03:45.830
And that hole diffuses
across the junction.

00:03:45.830 --> 00:03:46.990
The electron stays.

00:03:46.990 --> 00:03:47.823
AUDIENCE: I see, OK.

00:03:49.950 --> 00:03:51.550
PROFESSOR: So did
anybody else pick up

00:03:51.550 --> 00:03:54.100
on the point at the very
beginning of his presentation?

00:03:54.100 --> 00:03:57.980
He said a P-N junction isn't
necessary to separate charge.

00:03:57.980 --> 00:03:58.652
OK, that's fine.

00:03:58.652 --> 00:04:00.110
We've talked about
heterojunctions.

00:04:00.110 --> 00:04:02.840
We all agree there are other
ways to separate charge.

00:04:02.840 --> 00:04:05.220
And he said an
electric field is not

00:04:05.220 --> 00:04:07.580
necessary to separate
charge, but then

00:04:07.580 --> 00:04:11.140
he immediately went into
discussing how the chemical

00:04:11.140 --> 00:04:13.030
potential was slightly
lower in the contact

00:04:13.030 --> 00:04:14.750
than it was in the
semiconductor, which

00:04:14.750 --> 00:04:16.260
would result in a
charge imbalance, which

00:04:16.260 --> 00:04:17.260
would result in a field.

00:04:17.260 --> 00:04:19.339
And I think Gene Fitzgerald
from material science

00:04:19.339 --> 00:04:21.380
and engineering department--
Professor Fitzgerald

00:04:21.380 --> 00:04:23.766
called him out on it and
said, isn't there a field

00:04:23.766 --> 00:04:24.890
there at the metal contact.

00:04:24.890 --> 00:04:27.570
He said, quiet, wise guy,
we'll get back to you later.

00:04:27.570 --> 00:04:30.910
But essentially his point was
a very small electric field

00:04:30.910 --> 00:04:32.330
is necessary.

00:04:32.330 --> 00:04:35.650
So his point was a
matter of degrees,

00:04:35.650 --> 00:04:37.810
that you don't need a
massive electric field.

00:04:37.810 --> 00:04:39.510
A very slight
field is all that's

00:04:39.510 --> 00:04:44.232
necessary to start driving a
current through your system.

00:04:44.232 --> 00:04:46.790
I just wanted to
make sure we didn't

00:04:46.790 --> 00:04:50.120
leave that talk thoroughly
confused with our head

00:04:50.120 --> 00:04:52.570
on backwards.

00:04:52.570 --> 00:04:54.980
We're going to talk about
thin film materials today.

00:04:54.980 --> 00:04:56.510
Why thin film solar cells?

00:04:56.510 --> 00:04:59.260
Well, we've been talking about
crystalline silicon solar cells

00:04:59.260 --> 00:05:02.100
that have a lower optical
absorption coefficient, so you

00:05:02.100 --> 00:05:03.660
need a larger
amount of material,

00:05:03.660 --> 00:05:05.620
or a larger optical
path length, to absorb

00:05:05.620 --> 00:05:07.340
a significant
fraction of the light.

00:05:07.340 --> 00:05:09.380
Already, in lecture
number two, we

00:05:09.380 --> 00:05:11.230
saw how other
material systems that

00:05:11.230 --> 00:05:13.260
have higher optical
absorption coefficients

00:05:13.260 --> 00:05:15.690
are able to absorb this
equivalent amount of light

00:05:15.690 --> 00:05:19.060
in a thinner amount of
material and less material.

00:05:19.060 --> 00:05:20.770
So to put this in
perspective, what

00:05:20.770 --> 00:05:23.270
we're talking about on one hand
with the crystalline silicon

00:05:23.270 --> 00:05:25.330
devices is we might have
a device that's maybe

00:05:25.330 --> 00:05:27.680
three or four times the
thickness of your hair

00:05:27.680 --> 00:05:30.810
in crystalline silicon, and
for the other materials, so

00:05:30.810 --> 00:05:32.310
these thin film
materials, you might

00:05:32.310 --> 00:05:35.650
be talking about a material
absorber that has maybe

00:05:35.650 --> 00:05:37.410
100th the thickness,
so something

00:05:37.410 --> 00:05:40.900
under a micron or a 50th
of the width of your hair.

00:05:40.900 --> 00:05:42.857
So that's the
perspective of scale

00:05:42.857 --> 00:05:44.065
that we want to have in mind.

00:05:44.065 --> 00:05:45.564
When we're talking
about thin films,

00:05:45.564 --> 00:05:48.260
we're talking about thin
materials, really on the order

00:05:48.260 --> 00:05:49.480
of one micron or so.

00:05:49.480 --> 00:05:51.890
And even brittle materials,
at one micron thickness,

00:05:51.890 --> 00:05:57.710
if deposited on compliant
substrates, can be flexible.

00:05:57.710 --> 00:06:00.090
Another thing to keep in
mind in thin film technology

00:06:00.090 --> 00:06:02.520
is that the scale of
the thin films industry

00:06:02.520 --> 00:06:05.590
is about 1/10 that of
silicon industry right now.

00:06:05.590 --> 00:06:07.450
So the crystalline
silicon industry

00:06:07.450 --> 00:06:11.440
is going full force,
gangbusters right now,

00:06:11.440 --> 00:06:14.920
and the thin film is
a growing fraction,

00:06:14.920 --> 00:06:18.070
but it's on the order of 10%
of the total world market.

00:06:18.070 --> 00:06:21.740
And many, many, many startup
companies, which are young,

00:06:21.740 --> 00:06:25.090
dynamic, fun-- and that's why
today I'm not wearing a tie;

00:06:25.090 --> 00:06:26.230
I'm in startup mode.

00:06:26.230 --> 00:06:27.451
I'm a lot more relaxed.

00:06:27.451 --> 00:06:29.700
We're going to be talking
about thin film technologies

00:06:29.700 --> 00:06:32.420
and diving into some fun work.

00:06:32.420 --> 00:06:36.590
So we'll talk about these
specific technologies

00:06:36.590 --> 00:06:39.390
of thin film materials, and
before we get into those,

00:06:39.390 --> 00:06:42.020
I'm going to address some
general topics about deposition

00:06:42.020 --> 00:06:45.090
and, of course, general
parameters that affect

00:06:45.090 --> 00:06:48.010
all thin film material systems.

00:06:48.010 --> 00:06:50.750
We have to appreciate the sheer
diversity of technologies that

00:06:50.750 --> 00:06:51.916
are out there on the market.

00:06:51.916 --> 00:06:54.890
We have a variety of different
solar cell materials that

00:06:54.890 --> 00:06:57.140
are available, some of
which are thin films,

00:06:57.140 --> 00:07:00.120
other ones, wafer-based
crystalline silicon.

00:07:00.120 --> 00:07:03.220
And all of these
technologies have

00:07:03.220 --> 00:07:06.360
to consider cost resource
availability and, eventually,

00:07:06.360 --> 00:07:08.000
environmental impact as well.

00:07:08.000 --> 00:07:09.416
So these or some
of the things I'd

00:07:09.416 --> 00:07:12.364
like you to keep on the
forefront of your mind as we

00:07:12.364 --> 00:07:14.030
talk about these
different technologies.

00:07:14.030 --> 00:07:16.720
Think about the broader picture,
and ultimately, this cost,

00:07:16.720 --> 00:07:21.840
or the amount of money
per unit energy produced,

00:07:21.840 --> 00:07:24.900
is really paramount in
determining marketability

00:07:24.900 --> 00:07:28.370
and determining the
scales to which they'll

00:07:28.370 --> 00:07:30.150
penetrate the market.

00:07:30.150 --> 00:07:32.410
This is the one slide
that you have printed out.

00:07:32.410 --> 00:07:35.250
You have one per pair of
students, or you should.

00:07:35.250 --> 00:07:40.000
So if you don't have access
to that particular slide,

00:07:40.000 --> 00:07:43.370
feel free to share it with
the person next to you.

00:07:43.370 --> 00:07:45.700
This is representing
as a function of time,

00:07:45.700 --> 00:07:49.800
going back to the 1970s, the
record solar cell conversion

00:07:49.800 --> 00:07:51.210
efficiency.

00:07:51.210 --> 00:07:54.221
The chart is maintained by
a certain Larry Kazmerski

00:07:54.221 --> 00:07:54.720
at NREL.

00:07:54.720 --> 00:08:00.580
Actually, he used to be the
head of NREL's solar program.

00:08:00.580 --> 00:08:03.810
He stepped down a few years ago
after a very successful run--

00:08:03.810 --> 00:08:04.730
many years.

00:08:04.730 --> 00:08:08.190
And he's a bit of a father
of the US PV industry.

00:08:08.190 --> 00:08:10.310
He's been around for
a long time and has

00:08:10.310 --> 00:08:12.310
been tracking the growth
of the PV industry

00:08:12.310 --> 00:08:16.540
and, of course, the improvement
of performance over time.

00:08:16.540 --> 00:08:19.410
Many of these devices-- many
of these record efficiency

00:08:19.410 --> 00:08:23.490
devices-- are very small
area, and many of them

00:08:23.490 --> 00:08:26.430
were actually grown for the
intent, the explicit purpose,

00:08:26.430 --> 00:08:27.792
of getting onto this chart.

00:08:27.792 --> 00:08:30.000
And so when you're trying
to make a record efficiency

00:08:30.000 --> 00:08:32.464
device, you do things
a little differently.

00:08:32.464 --> 00:08:33.880
Let me give you
an example; you'll

00:08:33.880 --> 00:08:37.045
optimize your anti-reflection
coding for air not for glass,

00:08:37.045 --> 00:08:38.150
right?

00:08:38.150 --> 00:08:41.549
So if you want to
minimize the reflectance

00:08:41.549 --> 00:08:43.690
off the front surface,
you'll be optimizing it

00:08:43.690 --> 00:08:45.990
for air, which has a
refractive index of one, as

00:08:45.990 --> 00:08:49.250
opposed to glass, which has
a refractive index of 1.5.

00:08:49.250 --> 00:08:52.510
So there are some tricks
that one does and engages in,

00:08:52.510 --> 00:08:55.400
some are a little bit
under the table, too.

00:08:55.400 --> 00:08:57.410
It has been done in
the past that people

00:08:57.410 --> 00:09:00.350
would do an HF dip of their
silicon-based solar cells

00:09:00.350 --> 00:09:02.690
right before measuring
the efficiency.

00:09:02.690 --> 00:09:06.690
The hydrofluoric acid would
result in surface pacification,

00:09:06.690 --> 00:09:09.800
but, of course, it would
result in a very low surface

00:09:09.800 --> 00:09:13.020
adhesion of the metal, and so
the metal flake off afterwards.

00:09:13.020 --> 00:09:16.250
It wouldn't pass the tensile,
but you would nevertheless

00:09:16.250 --> 00:09:18.460
achieve instantaneously
higher efficiency.

00:09:18.460 --> 00:09:20.674
Those practices have
largely been weeded out.

00:09:20.674 --> 00:09:22.090
These were the
early days, when it

00:09:22.090 --> 00:09:24.460
was a wild west of solar cells.

00:09:24.460 --> 00:09:27.790
In more recent times, there
are some very strict standards,

00:09:27.790 --> 00:09:30.280
and there are only a few
laboratories around the world

00:09:30.280 --> 00:09:33.590
where you can take these
standard measurements.

00:09:33.590 --> 00:09:35.790
The one at NREL
is extremely well

00:09:35.790 --> 00:09:38.590
staffed in terms of the
quality of the people.

00:09:38.590 --> 00:09:40.450
They're notoriously
under resourced,

00:09:40.450 --> 00:09:41.940
but that's another issue.

00:09:41.940 --> 00:09:44.340
But in terms of the quality
of the people there,

00:09:44.340 --> 00:09:47.970
very, very good, very
thorough, very pedantic

00:09:47.970 --> 00:09:49.940
and careful about taking
their measurements.

00:09:49.940 --> 00:09:51.720
And if you ever have
a question about how

00:09:51.720 --> 00:09:53.740
to perform a solar cell
efficiency measurement,

00:09:53.740 --> 00:09:55.030
they're a very good resource.

00:09:55.030 --> 00:09:57.810
Their website would be
an excellent place to go.

00:09:57.810 --> 00:10:02.320
So these data points here versus
time represent the record cell

00:10:02.320 --> 00:10:03.390
efficiencies.

00:10:03.390 --> 00:10:06.020
They may be on very, very
small pieces of material.

00:10:06.020 --> 00:10:09.220
They may be on a centimeter
squared, perhaps even smaller,

00:10:09.220 --> 00:10:11.540
so they're not necessarily
representative of what is

00:10:11.540 --> 00:10:13.420
in commercial production today.

00:10:13.420 --> 00:10:16.950
Let me give you one example;
the record crystalline silicon

00:10:16.950 --> 00:10:22.060
cells, which are in blue here,
has been around 25% for about

00:10:22.060 --> 00:10:24.130
a decade-- actually,
a little more

00:10:24.130 --> 00:10:27.430
than a decade-- and
the record efficiency

00:10:27.430 --> 00:10:33.900
crystalline silicon device
has, in essence, not

00:10:33.900 --> 00:10:36.440
been so planted for
many, many years.

00:10:36.440 --> 00:10:38.740
There are a number
reasons for that.

00:10:38.740 --> 00:10:42.440
It's very much approaching its
theoretical efficiency limit.

00:10:42.440 --> 00:10:44.830
People haven't necessarily
tried specifically

00:10:44.830 --> 00:10:47.450
to get a record efficiency
crystalline silicon device.

00:10:47.450 --> 00:10:49.480
They're more intent on
making lower cost silicon

00:10:49.480 --> 00:10:51.560
devices than record
efficiency ones,

00:10:51.560 --> 00:10:54.720
and the average
module efficiencies

00:10:54.720 --> 00:10:56.310
are somewhere down
at around here--

00:10:56.310 --> 00:11:00.530
actually, somewhere
in the 13% to 15%

00:11:00.530 --> 00:11:02.960
range for module, average
module efficiency.

00:11:02.960 --> 00:11:06.960
You have some modules that
are in the 18%, 19% 20% range,

00:11:06.960 --> 00:11:08.900
but most of them are
significantly lower.

00:11:08.900 --> 00:11:10.941
And the record cell
efficiencies, as you can see,

00:11:10.941 --> 00:11:12.119
is 25%.

00:11:12.119 --> 00:11:13.910
So there's a significant
delta between what

00:11:13.910 --> 00:11:15.993
is commercially available
and what the record cell

00:11:15.993 --> 00:11:16.970
efficiency is.

00:11:16.970 --> 00:11:18.820
There are several
reasons for this.

00:11:18.820 --> 00:11:20.500
To make a record
efficiency cell,

00:11:20.500 --> 00:11:23.110
you have to throw everything
at Liebig's law of the minimum.

00:11:23.110 --> 00:11:25.693
You have to make sure that every
plank is really, really high.

00:11:25.693 --> 00:11:27.370
That costs a lot
of money typically,

00:11:27.370 --> 00:11:30.320
and so doing that cheaply
is a big challenge.

00:11:30.320 --> 00:11:32.090
Some companies,
like First Solar,

00:11:32.090 --> 00:11:34.110
for instance, has some
of the lowest cost

00:11:34.110 --> 00:11:35.290
models in the market.

00:11:35.290 --> 00:11:37.680
We'll describe how they're
made in a few slides.

00:11:37.680 --> 00:11:40.590
First Solar forwent the
anti-reflective coating

00:11:40.590 --> 00:11:43.220
on their glass for many
years, because it just

00:11:43.220 --> 00:11:45.080
didn't make cost sense.

00:11:45.080 --> 00:11:48.550
It didn't help optimize
this function right here.

00:11:48.550 --> 00:11:50.240
Although, you'd get
more energy out,

00:11:50.240 --> 00:11:52.690
the dollars that it took
to add that component just

00:11:52.690 --> 00:11:54.740
didn't make sense for them.

00:11:54.740 --> 00:11:57.900
So you have to think about a
few different perspectives.

00:11:57.900 --> 00:11:59.980
You have to think
both in terms of cost,

00:11:59.980 --> 00:12:01.520
and in terms of performance.

00:12:01.520 --> 00:12:04.130
The performance, what it
does or what it tells you

00:12:04.130 --> 00:12:06.850
is that this material
system has potential.

00:12:06.850 --> 00:12:10.490
It has been demonstrated we
can get the high performance.

00:12:10.490 --> 00:12:12.120
It's a proof of concept.

00:12:12.120 --> 00:12:15.390
The trick now is to
get there at low cost,

00:12:15.390 --> 00:12:17.570
and that's pretty much
what you should walk away

00:12:17.570 --> 00:12:19.114
from this chart having seen.

00:12:19.114 --> 00:12:21.530
Another thing to keep in mind
is that it takes a long time

00:12:21.530 --> 00:12:23.540
to improve the performance
of a new material.

00:12:23.540 --> 00:12:25.700
If you're starting out
somewhere down around here,

00:12:25.700 --> 00:12:28.580
it's going to take you a while
to reach higher efficiencies.

00:12:28.580 --> 00:12:31.800
Granted we can learn a lot from
the previous material systems.

00:12:31.800 --> 00:12:36.010
We could learn a lot by
reading those old NREL project

00:12:36.010 --> 00:12:38.480
reports that are available
online of all the people

00:12:38.480 --> 00:12:40.310
who were working towards these
record efficiencies-- what they

00:12:40.310 --> 00:12:41.910
did differently,
how they advanced,

00:12:41.910 --> 00:12:43.731
and how they improved
cell performance--

00:12:43.731 --> 00:12:45.230
and leverage that
information as you

00:12:45.230 --> 00:12:46.999
try to develop your material.

00:12:46.999 --> 00:12:49.540
But the fact of the matter is,
it'll still take a bit of time

00:12:49.540 --> 00:12:52.772
to develop new technologies,
and you can see that

00:12:52.772 --> 00:12:54.230
by some of the
newer materials that

00:12:54.230 --> 00:12:55.961
are coming along down
here, for example,

00:12:55.961 --> 00:12:57.210
the organic-based solar cells.

00:13:00.600 --> 00:13:03.700
So thin films,
general issues-- so we

00:13:03.700 --> 00:13:05.420
talked about the
advantages here,

00:13:05.420 --> 00:13:09.800
that we're squeezing the
cost of the absorber layer

00:13:09.800 --> 00:13:11.940
out of the module, which
is excellent from a cost

00:13:11.940 --> 00:13:13.170
point of view.

00:13:13.170 --> 00:13:15.650
But obviously, there
are trade-offs involved.

00:13:15.650 --> 00:13:17.470
If it was all a
walk in the park,

00:13:17.470 --> 00:13:21.057
we would be 100% thin films and
have abandoned silicon by now.

00:13:21.057 --> 00:13:22.890
There are both advantages
in this advantages

00:13:22.890 --> 00:13:24.440
with thin films.

00:13:24.440 --> 00:13:26.590
Instead of disadvantages,
perhaps a happier way

00:13:26.590 --> 00:13:28.800
of looking at this is
challenges and opportunities

00:13:28.800 --> 00:13:31.850
for getting PhDs and
other advanced degrees.

00:13:31.850 --> 00:13:33.920
So let's go up into advantages.

00:13:33.920 --> 00:13:35.720
The advantages of thin
films, quite simply,

00:13:35.720 --> 00:13:39.300
is that you're using a very thin
amount of material, so thin,

00:13:39.300 --> 00:13:42.170
in fact, that it's virtually
insignificant in terms

00:13:42.170 --> 00:13:44.510
of the total cost
structure of your module.

00:13:44.510 --> 00:13:47.930
One perspective is that if
you're depositing a thin film

00:13:47.930 --> 00:13:51.350
using a fairly
low-cost technique,

00:13:51.350 --> 00:13:55.940
like a c spaced
sublimation type process,

00:13:55.940 --> 00:13:59.760
you may be able to deposit
the material for as much cost

00:13:59.760 --> 00:14:01.550
as it takes the
cardboard that separates

00:14:01.550 --> 00:14:04.230
the modulus from each other in
the stack that's being loaded

00:14:04.230 --> 00:14:06.082
onto the 18-wheeler
out of the factory,

00:14:06.082 --> 00:14:07.290
to put things in perspective.

00:14:07.290 --> 00:14:10.000
It's very cheap to
deposit these thin layers.

00:14:10.000 --> 00:14:12.995
Now let's hop down to the
disadvantages real quick.

00:14:12.995 --> 00:14:14.620
If you're depositing
a very thin layer,

00:14:14.620 --> 00:14:17.300
and it's not high
efficiency, then you

00:14:17.300 --> 00:14:19.410
need more glass, more
encapsulants, more framing

00:14:19.410 --> 00:14:21.190
materials, more labor, and
everything for the same amount

00:14:21.190 --> 00:14:22.070
of power out.

00:14:22.070 --> 00:14:25.306
If your efficiency is low,
your costs will be higher.

00:14:25.306 --> 00:14:27.180
Even if you have a dirt
cheap absorber layer,

00:14:27.180 --> 00:14:29.180
you might as well get
the absorber for free.

00:14:29.180 --> 00:14:31.850
If your efficiency is too
low, all the other commodity

00:14:31.850 --> 00:14:34.860
materials are going to
outweigh that cost advantage

00:14:34.860 --> 00:14:37.220
because the commodity
materials scale with area.

00:14:37.220 --> 00:14:39.650
If you have low efficiency,
you need a larger area module

00:14:39.650 --> 00:14:42.080
to make the same
amount of power.

00:14:42.080 --> 00:14:44.290
So, at some point, if
you look at the cost

00:14:44.290 --> 00:14:46.080
of the material
versus efficiency,

00:14:46.080 --> 00:14:47.940
you start entering
negative territory.

00:14:47.940 --> 00:14:49.920
You actually have to be paid.

00:14:49.920 --> 00:14:52.410
If you're producing like
an 8% or a 7% module,

00:14:52.410 --> 00:14:55.620
typically, you would have to
pay your customer for them

00:14:55.620 --> 00:14:57.690
to accept your module.

00:14:57.690 --> 00:15:01.180
So you really have to
achieve a minimum efficiency

00:15:01.180 --> 00:15:04.230
target to be cost competitive,
and as a rule of thumb,

00:15:04.230 --> 00:15:08.390
that's typically 10% to 12%
for today's cost of glass

00:15:08.390 --> 00:15:12.690
encapsulance framing materials
and labor and installation

00:15:12.690 --> 00:15:14.650
and so forth.

00:15:14.650 --> 00:15:16.330
So back up to the
advantages-- there's

00:15:16.330 --> 00:15:19.030
a potential here for a
very low thermal budget.

00:15:19.030 --> 00:15:22.160
If we're able to print,
say, a micron-thick layer

00:15:22.160 --> 00:15:25.040
onto a substrate,
remember, we go back

00:15:25.040 --> 00:15:27.300
to that the high
speed printer analogy,

00:15:27.300 --> 00:15:29.480
there's a potential for a
low thermal budget, which

00:15:29.480 --> 00:15:32.260
means thermal budget
is the amount of heat

00:15:32.260 --> 00:15:35.440
that you're introducing
during the processing.

00:15:35.440 --> 00:15:37.680
As a result of a very
low thermal budget,

00:15:37.680 --> 00:15:39.530
you have a potential
cost decrease.

00:15:39.530 --> 00:15:41.540
Instead of heating things
up to 1,400 degrees

00:15:41.540 --> 00:15:45.789
C over several hours, having
all that massive amounts

00:15:45.789 --> 00:15:47.330
of electricity that
go into producing

00:15:47.330 --> 00:15:50.090
the crystalline silicon
wafers, here, potentially, we

00:15:50.090 --> 00:15:53.470
could be printing stuff
on flexible substrates.

00:15:53.470 --> 00:15:55.430
So that's the thermal
budget argument.

00:15:55.430 --> 00:15:59.540
In terms of conformal deposition
and flexible substrates,

00:15:59.540 --> 00:16:02.320
there's a potential here
for roll-to-roll deposition.

00:16:02.320 --> 00:16:05.080
Picture a newspaper plant,
where you have one roll of paper

00:16:05.080 --> 00:16:08.030
on one side being pulled on
to another spool in the other,

00:16:08.030 --> 00:16:10.520
with some deposition process
happening in between.

00:16:10.520 --> 00:16:14.140
If you can deposit on a flexible
substrate, this is the vision.

00:16:14.140 --> 00:16:16.855
And if you're not depositing
onto a flexible substrate

00:16:16.855 --> 00:16:19.390
but onto hard substrate
like this one right here--

00:16:19.390 --> 00:16:21.060
this is glass, a
thin film material

00:16:21.060 --> 00:16:23.180
deposited on glass right
here, a very small one.

00:16:23.180 --> 00:16:24.140
Oops, some tape on the front.

00:16:24.140 --> 00:16:25.431
Let me get rid of that for you.

00:16:27.768 --> 00:16:28.665
Here we go.

00:16:33.980 --> 00:16:38.320
It's in a nice little
protective coating here,

00:16:38.320 --> 00:16:41.370
so you can have a look at it
without worrying about getting

00:16:41.370 --> 00:16:43.790
your fingerprints all over it.

00:16:43.790 --> 00:16:48.610
And the company name is
fully removed-- check.

00:16:48.610 --> 00:16:52.340
This is an example of a
thin film material deposited

00:16:52.340 --> 00:16:55.310
on glass without any
anti-reflective coating, just

00:16:55.310 --> 00:16:57.520
the absorber material,
so you can get a sense.

00:16:57.520 --> 00:16:58.630
It looks great.

00:16:58.630 --> 00:17:00.270
It's about a micron thick.

00:17:00.270 --> 00:17:04.270
It's about 170 times
thinner than those wafers

00:17:04.270 --> 00:17:06.555
that you saw on Tuesday.

00:17:06.555 --> 00:17:08.430
So that's an example of
a thin film material.

00:17:08.430 --> 00:17:10.025
It will be making its rounds.

00:17:10.025 --> 00:17:11.900
There's a large amount
of technology transfer

00:17:11.900 --> 00:17:14.849
with a thin film display, the
flat panel display industry,

00:17:14.849 --> 00:17:18.359
with deposition on glass
like that one right there.

00:17:18.359 --> 00:17:20.950
And there's a potential it'll
be very nice for building

00:17:20.950 --> 00:17:22.477
integrated PV applications.

00:17:22.477 --> 00:17:24.060
If you're able to
get rid of the glass

00:17:24.060 --> 00:17:25.670
and deposit on a
conformal substrate,

00:17:25.670 --> 00:17:30.540
you could envision roof shingles
or other flexible substrates

00:17:30.540 --> 00:17:33.120
that would allow you conformal
coverage on undulating roof

00:17:33.120 --> 00:17:35.950
tops and so forth.

00:17:35.950 --> 00:17:38.550
Radiation hardness-- this
is just a small aside,

00:17:38.550 --> 00:17:41.270
but there are some materials
that have better radiation

00:17:41.270 --> 00:17:42.420
hardness than silicon.

00:17:42.420 --> 00:17:43.860
What does radiation
hardness mean?

00:17:43.860 --> 00:17:46.151
It means that if I send
something to outer space, where

00:17:46.151 --> 00:17:48.190
we don't benefit
from the radiation

00:17:48.190 --> 00:17:51.170
shield of our own atmosphere in
the Van Allen belts on earth,

00:17:51.170 --> 00:17:56.740
and we have proton bombardment
and other forms of radiation

00:17:56.740 --> 00:17:59.600
striking are module and creating
damage within the absorber

00:17:59.600 --> 00:18:03.410
layer, some compounds are
naturally better at resisting

00:18:03.410 --> 00:18:05.860
degradation of
performance than others,

00:18:05.860 --> 00:18:07.789
and that's what
radiation hardness means.

00:18:07.789 --> 00:18:09.330
So there are some
thin film materials

00:18:09.330 --> 00:18:13.200
that are exceptional
for space applications.

00:18:13.200 --> 00:18:15.240
The challenges and--
oh, go ahead, Ashley.

00:18:15.240 --> 00:18:16.990
ASHLEY: Is gallium
arsenide one of them?

00:18:16.990 --> 00:18:19.114
PROFESSOR: We're going to
show you in a few slides.

00:18:19.114 --> 00:18:22.260
We'll compare them all as a
function of radiation exposure

00:18:22.260 --> 00:18:23.160
time.

00:18:23.160 --> 00:18:25.210
The disadvantages, or
shall we say challenges

00:18:25.210 --> 00:18:28.072
and opportunities for PhD
and master's students,

00:18:28.072 --> 00:18:30.280
lower efficiencies in
crystalline silicon potentially

00:18:30.280 --> 00:18:31.591
larger module costs.

00:18:31.591 --> 00:18:34.090
If you're able to improve the
performance of these thin film

00:18:34.090 --> 00:18:38.530
materials, wow, you have
now equivalent performance

00:18:38.530 --> 00:18:40.950
of crystalline silicon
but at much lower cost.

00:18:40.950 --> 00:18:43.460
Good for you-- you have
a marketable product.

00:18:43.460 --> 00:18:47.110
Potential for capital intensive
production equipment-- not all

00:18:47.110 --> 00:18:50.080
of the production equipment
is as low cost and as low

00:18:50.080 --> 00:18:52.900
thermal budget as simply
printing on a piece of paper.

00:18:52.900 --> 00:18:56.510
As a matter of fact, that's one
of the more avant garde and R&D

00:18:56.510 --> 00:18:58.730
type of deposition processes.

00:18:58.730 --> 00:19:01.280
Most deposition processes
and the vast majority

00:19:01.280 --> 00:19:04.590
of companies used are actually
quite capital intensive,

00:19:04.590 --> 00:19:07.300
and the cost of the
equipment can add up.

00:19:07.300 --> 00:19:09.930
Sometimes, not always,
but sometimes scarce

00:19:09.930 --> 00:19:11.760
elements are used.

00:19:11.760 --> 00:19:14.790
We're going to have a debate
about that on next class,

00:19:14.790 --> 00:19:16.822
on Tuesday.

00:19:16.822 --> 00:19:18.030
Put an asterisk next to that.

00:19:18.030 --> 00:19:20.238
I'll get back to those as
soon as this slide is over.

00:19:20.238 --> 00:19:23.580
And spatial uniformity is a
challenge during deposition.

00:19:23.580 --> 00:19:28.210
Imagine trying to deposit a
film one-micron thick over glass

00:19:28.210 --> 00:19:31.120
that is one meter in size.

00:19:31.120 --> 00:19:34.570
You're talking about a six order
of magnitude aspect ratio here.

00:19:34.570 --> 00:19:38.730
So we have to somehow deposit
a film a micron thick in layers

00:19:38.730 --> 00:19:40.170
that are even
thinner, that might

00:19:40.170 --> 00:19:43.320
be only a few tens or hundreds
of nanometers on top of that

00:19:43.320 --> 00:19:46.290
and below that absorber
layer to separate charge,

00:19:46.290 --> 00:19:48.910
for instance, and that's really
challenging to do on a very

00:19:48.910 --> 00:19:50.869
large scale, and that
is an engineering

00:19:50.869 --> 00:19:52.910
challenge or a process
engineering challenge that

00:19:52.910 --> 00:19:56.870
had many startup companies
flailing for a long time.

00:19:56.870 --> 00:19:59.232
Think of spatial homogeneity
in the following manner;

00:19:59.232 --> 00:20:00.940
if you have one region
of your solar cell

00:20:00.940 --> 00:20:03.356
that's producing a lot of
power, and the region next to it

00:20:03.356 --> 00:20:07.550
is not, and they're connected in
parallel through the contacts,

00:20:07.550 --> 00:20:10.160
power will flow from the good
region into the bad region.

00:20:10.160 --> 00:20:14.420
So you have internal current
loops inside of your module.

00:20:14.420 --> 00:20:16.610
That is essentially
decreasing the power output

00:20:16.610 --> 00:20:18.420
of your module itself.

00:20:18.420 --> 00:20:21.950
So that's why
homogeneity is important.

00:20:21.950 --> 00:20:24.109
This is just to
represent the vision

00:20:24.109 --> 00:20:26.650
of a roll-to-roll process in
the upper right-hand side there.

00:20:26.650 --> 00:20:32.500
Kind of a visionary cartoon
that is being enacted

00:20:32.500 --> 00:20:34.630
by one company, in
particular, Uni-Solar,

00:20:34.630 --> 00:20:35.660
based out of Michigan.

00:20:35.660 --> 00:20:37.750
They do have a
roll-to-roll process

00:20:37.750 --> 00:20:39.720
and PCBD-- we'll
describe what that

00:20:39.720 --> 00:20:42.830
is in a second-- deposition
of this material,

00:20:42.830 --> 00:20:44.381
so-called amorphous silicon.

00:20:44.381 --> 00:20:46.380
And here are some building
integrated solutions,

00:20:46.380 --> 00:20:47.950
just showing you what
you can accomplish

00:20:47.950 --> 00:20:49.190
or what the vision would be.

00:20:49.190 --> 00:20:53.170
If had have this really flexible
substrate that you could

00:20:53.170 --> 00:20:57.020
literally take it as a
roll from Home Depot,

00:20:57.020 --> 00:20:59.490
bring up to your rooftop,
splay it out on your roof,

00:20:59.490 --> 00:21:02.400
much like you'd lay down a
piece of tarp or plastic,

00:21:02.400 --> 00:21:07.030
and take a staple gun or a nail
gun and drill it into location,

00:21:07.030 --> 00:21:11.072
that would be an example of a
much reduced installation cost.

00:21:11.072 --> 00:21:13.530
So you have the potential here
of reducing the installation

00:21:13.530 --> 00:21:18.484
cost of solar as a result of
the form factor of your module.

00:21:18.484 --> 00:21:20.400
And this here is another
example of a building

00:21:20.400 --> 00:21:22.820
integrated photovoltaic
solution within films.

00:21:22.820 --> 00:21:25.040
The fact that it looks
really nice, is really sleek,

00:21:25.040 --> 00:21:27.430
you'd never guess that those
are solar panels there,

00:21:27.430 --> 00:21:29.860
and that's, of course, from
an aesthetic point of view,

00:21:29.860 --> 00:21:32.300
a huge benefit.

00:21:32.300 --> 00:21:34.040
Common growth
methods-- how do we

00:21:34.040 --> 00:21:37.980
make that sample of
copper indium gallium

00:21:37.980 --> 00:21:40.210
diselenide, that thin
film material that happens

00:21:40.210 --> 00:21:42.460
to be making its way around
the classroom right now,

00:21:42.460 --> 00:21:44.220
how do we actually make it?

00:21:44.220 --> 00:21:46.750
Well, not only the
material I just described,

00:21:46.750 --> 00:21:48.190
there are other
materials as well.

00:21:48.190 --> 00:21:51.460
We'll talk about the general
classes of growth method.

00:21:51.460 --> 00:21:55.840
So this is the material science
processing class condensed

00:21:55.840 --> 00:21:57.010
into a few slides.

00:21:57.010 --> 00:21:59.280
Bear with me; this
very high level,

00:21:59.280 --> 00:22:02.150
but it aims to highlight
the techniques that are

00:22:02.150 --> 00:22:04.500
most commonly used in PV today.

00:22:04.500 --> 00:22:05.950
We're going to
start with what are

00:22:05.950 --> 00:22:08.910
called vacuum-based thin-film
deposition technologies.

00:22:08.910 --> 00:22:11.960
And the reason I'm separating
vacuum from non-vacuum

00:22:11.960 --> 00:22:13.894
is because if you
have a system that

00:22:13.894 --> 00:22:16.060
is comprised of these large
stainless steel chambers

00:22:16.060 --> 00:22:18.435
that you typically see when
you go walking in the physics

00:22:18.435 --> 00:22:21.410
building, if you have
a vacuum chambers,

00:22:21.410 --> 00:22:22.930
those are typically
quite costly,

00:22:22.930 --> 00:22:24.355
at least the large
scale ones that

00:22:24.355 --> 00:22:26.470
are in commercial production.

00:22:26.470 --> 00:22:28.010
As the name would
suggest, you need

00:22:28.010 --> 00:22:32.970
to have pumps to suck out the
air inside of the chamber,

00:22:32.970 --> 00:22:34.760
and that's how you
create the vacuum.

00:22:34.760 --> 00:22:37.370
The vacuum is necessary
because typically you're

00:22:37.370 --> 00:22:40.230
transporting atoms from
some sort of source,

00:22:40.230 --> 00:22:43.785
either gas or a solid
target, onto the substrate.

00:22:43.785 --> 00:22:45.910
So you're transferring
individual atoms or clusters

00:22:45.910 --> 00:22:49.700
of atoms from some
source onto the substrate

00:22:49.700 --> 00:22:52.380
that will ultimately hold
your thin film device.

00:22:52.380 --> 00:22:55.640
And that process
requires a limited number

00:22:55.640 --> 00:22:58.970
of interactions of those
atoms or clusters of atoms,

00:22:58.970 --> 00:23:01.300
in other words, a
large mean-free path,

00:23:01.300 --> 00:23:04.140
as these make their
way to your substrate.

00:23:04.140 --> 00:23:06.250
And that's why the vacuum
is typically required

00:23:06.250 --> 00:23:07.730
in these deposition systems.

00:23:07.730 --> 00:23:11.090
There are a variety of ways
to accomplish this goal.

00:23:11.090 --> 00:23:13.890
One class of
techniques is called

00:23:13.890 --> 00:23:17.960
Chemical Vapor Deposition,
often referred to as CVD.

00:23:17.960 --> 00:23:21.250
This typically involves
flowing in some form of gas

00:23:21.250 --> 00:23:25.350
into your chamber and
then allowing that gas

00:23:25.350 --> 00:23:29.070
to react on the
surface of your sample

00:23:29.070 --> 00:23:31.070
or above the surface
of your sample

00:23:31.070 --> 00:23:33.870
and ultimately depositing
on the surface.

00:23:33.870 --> 00:23:38.240
The chemistries involved in CVD
processes can be quite complex,

00:23:38.240 --> 00:23:40.610
and the reaction
process itself can

00:23:40.610 --> 00:23:42.550
be very difficult to master.

00:23:42.550 --> 00:23:44.100
So you might have
some friends who

00:23:44.100 --> 00:23:46.852
are involved in spectroscopy
shining lasers at their system

00:23:46.852 --> 00:23:48.310
and looking at the
absorption lines

00:23:48.310 --> 00:23:50.880
and trying to figure out how
these molecules are evolving

00:23:50.880 --> 00:23:52.530
between when they're
inserted into the chamber

00:23:52.530 --> 00:23:54.390
and when they actually
wind up as your film,

00:23:54.390 --> 00:23:55.810
because understanding
the reaction,

00:23:55.810 --> 00:23:58.018
the chemical reactions, that
take place is essential,

00:23:58.018 --> 00:24:01.104
is key, to really
controlling the CVD process.

00:24:01.104 --> 00:24:02.770
The other class of
technologies involved

00:24:02.770 --> 00:24:05.440
is called PVD, or
Physical Vapor Deposition,

00:24:05.440 --> 00:24:07.740
and this tends to be a
bit more straightforward.

00:24:07.740 --> 00:24:11.750
We tend to have atoms
of a specific type.

00:24:11.750 --> 00:24:13.880
They may be ionized, or
they may be charge neutral,

00:24:13.880 --> 00:24:15.838
and they're making their
way to your substrate.

00:24:15.838 --> 00:24:17.860
And the chemistry tends
to be much more simple,

00:24:17.860 --> 00:24:21.850
but the apparatus around
it to give the incentive

00:24:21.850 --> 00:24:23.460
for the atoms to
leave the target

00:24:23.460 --> 00:24:26.980
and deposit on your substrate,
that tends to be more complex.

00:24:26.980 --> 00:24:29.960
And so some of these
tools, especially

00:24:29.960 --> 00:24:33.460
molecular-beam epitaxy can
be very expensive, very slow,

00:24:33.460 --> 00:24:36.640
but very high quality, but
very expensive as a result.

00:24:36.640 --> 00:24:39.290
And so a very
simple way to think

00:24:39.290 --> 00:24:42.900
about the vacuum-based
deposition technologies

00:24:42.900 --> 00:24:47.052
is a compromise-- this is an
oversimplification indeed,

00:24:47.052 --> 00:24:48.510
but it's an easy
way to get started

00:24:48.510 --> 00:24:51.130
about thinking of the parameter
space of all these techniques.

00:24:51.130 --> 00:24:54.860
It's a compromise between
speed and quality.

00:24:54.860 --> 00:24:56.980
Some of the techniques
that are fastest

00:24:56.980 --> 00:24:58.924
also tend to be the
lowest quality materials,

00:24:58.924 --> 00:25:00.840
and the other ones that
tend to be the slowest

00:25:00.840 --> 00:25:02.756
tend to produce the
highest quality materials.

00:25:02.756 --> 00:25:06.020
How do you optimize somewhere
in between, somewhere

00:25:06.020 --> 00:25:08.810
in that parameter space, to
get reasonably high material,

00:25:08.810 --> 00:25:11.430
just enough that you can produce
a high efficiency device--

00:25:11.430 --> 00:25:13.760
remember that saturation
of device performance

00:25:13.760 --> 00:25:14.980
versus diffusion length.

00:25:14.980 --> 00:25:16.400
At some point, it just
doesn't make sense

00:25:16.400 --> 00:25:17.710
to keep optimizing
your material.

00:25:17.710 --> 00:25:18.820
You've got it good enough.

00:25:18.820 --> 00:25:21.070
You're good to go.

00:25:21.070 --> 00:25:25.460
So that's one of the things to
consider when you're choosing

00:25:25.460 --> 00:25:26.970
your deposition system.

00:25:26.970 --> 00:25:30.530
So let's go into a few
examples of these vacuum-based

00:25:30.530 --> 00:25:31.860
deposition systems.

00:25:31.860 --> 00:25:34.150
Within the PVD techniques,
within the Physical Vapor

00:25:34.150 --> 00:25:37.120
Deposition techniques, one
of the most commonly used

00:25:37.120 --> 00:25:39.820
in manufacturing, at
least in some startups--

00:25:39.820 --> 00:25:44.160
you have examples like
MiaSole-- is sputtering.

00:25:44.160 --> 00:25:46.686
And this sputtering
process is essentially

00:25:46.686 --> 00:25:47.810
very, very straightforward.

00:25:47.810 --> 00:25:49.490
You have a plasma.

00:25:49.490 --> 00:25:53.730
The plasma consists of
atoms that are charged.

00:25:53.730 --> 00:25:57.340
These are accelerated
toward your target, which

00:25:57.340 --> 00:25:59.170
is comprised of the
elements that you want

00:25:59.170 --> 00:26:01.610
to deposit onto your substrate.

00:26:01.610 --> 00:26:03.190
Your substrate is
sitting up top.

00:26:03.190 --> 00:26:07.310
And this target material is
sputtered off and eventually

00:26:07.310 --> 00:26:10.350
makes its way up and sticks
to and eventually grows

00:26:10.350 --> 00:26:13.620
the film on that orange
platen up here at the top.

00:26:13.620 --> 00:26:15.370
That is your substrate.

00:26:15.370 --> 00:26:17.300
The substrate is facing down.

00:26:17.300 --> 00:26:19.000
Why is the substrate
looking down?

00:26:19.000 --> 00:26:21.230
Why wouldn't you invert
this and put the target

00:26:21.230 --> 00:26:23.270
on top and in the
substrate in the bottom?

00:26:23.270 --> 00:26:25.730
What could happened
then in terms of purity

00:26:25.730 --> 00:26:26.920
of the deposition process?

00:26:26.920 --> 00:26:28.070
Let's go to Kristy.

00:26:28.070 --> 00:26:30.084
AUDIENCE: Things
could fall onto it.

00:26:30.084 --> 00:26:32.500
PROFESSOR: So stuff, gunk,
could fall onto your substrate.

00:26:32.500 --> 00:26:34.870
You're trying to grow a
thin film a micron thick,

00:26:34.870 --> 00:26:36.770
and you're trying to
avoid any imperfection,

00:26:36.770 --> 00:26:39.080
and now gravity is working
against you in that case.

00:26:39.080 --> 00:26:40.910
Because, if you
were to invert this,

00:26:40.910 --> 00:26:42.220
your target would be on top.

00:26:42.220 --> 00:26:44.980
You could have stuff raining
down onto your substrate.

00:26:44.980 --> 00:26:46.930
There are a few people
who sputter down.

00:26:46.930 --> 00:26:47.722
It's very tricky.

00:26:47.722 --> 00:26:49.930
You have to be able to
control your process very well

00:26:49.930 --> 00:26:51.560
and avoid flakes
from coming off.

00:26:51.560 --> 00:26:54.050
There are folks who
sputter sideways,

00:26:54.050 --> 00:26:56.100
saves some ground
space in their factory.

00:26:56.100 --> 00:26:59.480
They might load things
vertically, put them in.

00:26:59.480 --> 00:27:03.240
And many people, at
least in R&D, sputter up.

00:27:03.240 --> 00:27:06.790
So again, you're
creating this plasma.

00:27:06.790 --> 00:27:10.040
The charged species are
accelerated toward the target.

00:27:10.040 --> 00:27:13.540
They sputter off
atoms, which are then

00:27:13.540 --> 00:27:16.330
deposited on to your substrate,
which is there at the top.

00:27:16.330 --> 00:27:19.380
And the film that was
just being passed around

00:27:19.380 --> 00:27:21.600
is an example of
a sputtered film.

00:27:21.600 --> 00:27:23.910
The spatial uniformity
of sputtering

00:27:23.910 --> 00:27:25.860
over large area
depositions can be

00:27:25.860 --> 00:27:27.810
in the order of a few percent.

00:27:27.810 --> 00:27:29.970
So the ability to
control this process

00:27:29.970 --> 00:27:34.970
in terms of spatial
uniformity is fairly good.

00:27:34.970 --> 00:27:37.060
You could also employ
radio frequency modulations

00:27:37.060 --> 00:27:38.080
to the bias voltage.

00:27:38.080 --> 00:27:41.060
That's called RF sputtering
for Radio Frequency.

00:27:41.060 --> 00:27:45.530
Industrial applications usually
involve large rotating targets.

00:27:45.530 --> 00:27:47.320
So for those of
you-- how many people

00:27:47.320 --> 00:27:48.770
actually work with some
sputtering materials

00:27:48.770 --> 00:27:49.936
or have done it in the past?

00:27:49.936 --> 00:27:52.716
One, two, three,
four, five, six, OK.

00:27:52.716 --> 00:27:54.590
So you know that, at
least in the laboratory,

00:27:54.590 --> 00:27:57.131
if you have a fixed target, you
wind up with that race track,

00:27:57.131 --> 00:27:57.750
right?

00:27:57.750 --> 00:27:59.210
So if you have a fixed
target in the lab,

00:27:59.210 --> 00:28:00.876
and you're trying to
deposit your films,

00:28:00.876 --> 00:28:03.490
if you wear it
down several hours,

00:28:03.490 --> 00:28:05.880
eventually the metal that
you're trying to deposit,

00:28:05.880 --> 00:28:07.980
or the ceramic that
you're trying to deposit,

00:28:07.980 --> 00:28:10.710
will usually wind up having
a bit of shape to it.

00:28:10.710 --> 00:28:12.620
Instead of being
flat on the surface,

00:28:12.620 --> 00:28:14.280
you'll have what's
called a race track;

00:28:14.280 --> 00:28:16.160
it'll dipped down
near the edges,

00:28:16.160 --> 00:28:19.440
and that can result in a
change of the deposition

00:28:19.440 --> 00:28:23.359
rate of the species that
you're trying to deposit.

00:28:23.359 --> 00:28:24.900
And from a homogeneity
point of view,

00:28:24.900 --> 00:28:26.960
that might be disastrous
in the company,

00:28:26.960 --> 00:28:30.440
and so there are methods
to move your target

00:28:30.440 --> 00:28:33.650
to avoid that sort of
effect from happening.

00:28:33.650 --> 00:28:36.040
And when we talk
about large targets,

00:28:36.040 --> 00:28:38.140
we're really talking about
large targets, right?

00:28:38.140 --> 00:28:40.900
These aren't your lab scale
two-inch or three-inch,

00:28:40.900 --> 00:28:45.240
these are much, much bigger
in commercial production.

00:28:45.240 --> 00:28:50.210
So in terms of comparing
sputtering against other growth

00:28:50.210 --> 00:28:54.750
technologies, there
are technologies

00:28:54.750 --> 00:28:56.330
that are more conformal.

00:28:56.330 --> 00:28:58.900
Because this is more of a
line-of-sight deposition

00:28:58.900 --> 00:29:02.440
technique, the atoms are
moving toward your substrates.

00:29:02.440 --> 00:29:07.370
But if you have some
shape to your substrate,

00:29:07.370 --> 00:29:10.380
maybe you have a ledge
or a ridge, in that case,

00:29:10.380 --> 00:29:13.090
you won't necessarily
coat that uniformly.

00:29:13.090 --> 00:29:16.780
You might have less being
deposited on that edge rather

00:29:16.780 --> 00:29:18.140
than the flat sections.

00:29:18.140 --> 00:29:21.727
And so conformality of coverage,
or conformal surface coverage,

00:29:21.727 --> 00:29:23.060
can be an issue with sputtering.

00:29:25.710 --> 00:29:28.190
Let's talk about
the next technique

00:29:28.190 --> 00:29:34.310
that is commonly used in
inorganic thin-film deposition.

00:29:34.310 --> 00:29:35.830
Excuse me.

00:29:35.830 --> 00:29:39.780
This is called metalorganic
chemical vapor deposition.

00:29:39.780 --> 00:29:41.780
So again we notice the
CVD appearing at the end.

00:29:41.780 --> 00:29:44.030
We know it's a Chemical
Vapor Deposition process.

00:29:44.030 --> 00:29:47.260
MO in this case, standing
for Metalorganic.

00:29:47.260 --> 00:29:50.070
The reason metalorganic
is because we typically

00:29:50.070 --> 00:29:53.170
have a metal, like this
representing the indium right

00:29:53.170 --> 00:29:57.700
here, and then little organic
compounds on the outside.

00:29:57.700 --> 00:29:58.700
Those are methyl groups.

00:29:58.700 --> 00:30:01.510
The little gray and the
two white dots, those

00:30:01.510 --> 00:30:05.060
represent three methyl
groups around the indium, so

00:30:05.060 --> 00:30:06.060
trimethylindium.

00:30:06.060 --> 00:30:10.710
And what we do is we flow these
molecules into our reaction

00:30:10.710 --> 00:30:13.835
chamber and control the
temperature gradients inside

00:30:13.835 --> 00:30:15.970
in such a way to
have those molecules

00:30:15.970 --> 00:30:19.610
deposit on the surface, leaving
the indium behind, or the metal

00:30:19.610 --> 00:30:22.130
behind, and the
reaction products flow

00:30:22.130 --> 00:30:27.057
away out the back, and that
is represented chemically

00:30:27.057 --> 00:30:27.890
here on the surface.

00:30:27.890 --> 00:30:31.090
This is zooming in right at
the surface of our sample

00:30:31.090 --> 00:30:33.690
so right where the gas interacts
with the thin film material

00:30:33.690 --> 00:30:35.090
that you're depositing.

00:30:35.090 --> 00:30:39.110
This is representing the
incoming metalorganic molecule

00:30:39.110 --> 00:30:40.450
reaching the surface.

00:30:40.450 --> 00:30:44.800
This represents, right
here, the separation where

00:30:44.800 --> 00:30:49.160
we have the indium shown
in black right here,

00:30:49.160 --> 00:30:52.420
and then the methyl
groups are moving off,

00:30:52.420 --> 00:30:54.290
and essentially, those
will be sucked out

00:30:54.290 --> 00:30:57.340
of the chamber, leaving behind,
in this particular case,

00:30:57.340 --> 00:31:00.560
you have a layer of indium
forming, probably another layer

00:31:00.560 --> 00:31:01.940
of material underneath.

00:31:01.940 --> 00:31:05.270
Say, for example, your
other species comprising

00:31:05.270 --> 00:31:08.740
the thin film may be phosphorus,
so it would be indium phosphide

00:31:08.740 --> 00:31:10.230
growth.

00:31:10.230 --> 00:31:13.500
This metalorganic
chemical vapor deposition

00:31:13.500 --> 00:31:15.120
is very nice from
the point of view

00:31:15.120 --> 00:31:18.590
that you tend to form
homogeneous films-- very

00:31:18.590 --> 00:31:20.690
good surface coverage.

00:31:20.690 --> 00:31:24.860
The disadvantages would be that
many of the inputs and outputs

00:31:24.860 --> 00:31:27.540
are toxix-- not always,
but many of them are.

00:31:27.540 --> 00:31:29.130
They have to be
volatile and reactive

00:31:29.130 --> 00:31:31.850
so that you can crack
the metal on your surface

00:31:31.850 --> 00:31:34.509
and create the thin film.

00:31:34.509 --> 00:31:36.050
If it wasn't reactive,
you would just

00:31:36.050 --> 00:31:38.380
have it flowing through
and leaving, not having

00:31:38.380 --> 00:31:40.060
a reactant with your substrate.

00:31:40.060 --> 00:31:42.180
But because of the reactivity
involved, oftentimes

00:31:42.180 --> 00:31:44.640
these are not very
friendly for human beings

00:31:44.640 --> 00:31:46.540
or for other organisms.

00:31:46.540 --> 00:31:48.997
It was not uncommon in the
early days of MOCVD reactor

00:31:48.997 --> 00:31:51.330
development where they'd have
this little stack going up

00:31:51.330 --> 00:31:52.630
to the roof, and then
when they'd do maintenance

00:31:52.630 --> 00:31:55.088
on the roof, they'd find all
these dead birds lying around.

00:31:55.088 --> 00:31:58.070
That obviously has improved
since people have put up

00:31:58.070 --> 00:32:02.030
the appropriate filtration
on the output of their growth

00:32:02.030 --> 00:32:05.950
system, so-called scrubbers, to
prevent toxic gases from being

00:32:05.950 --> 00:32:07.420
released into the atmosphere.

00:32:07.420 --> 00:32:11.450
But you do have
some old stories.

00:32:11.450 --> 00:32:14.300
So the proper design of
metalorganic precursors

00:32:14.300 --> 00:32:14.970
is essential.

00:32:14.970 --> 00:32:18.280
You can easily see how if you
change the molecule that you're

00:32:18.280 --> 00:32:20.790
bringing in, all of a sudden
now, your reaction temperatures

00:32:20.790 --> 00:32:21.331
are changing.

00:32:21.331 --> 00:32:23.870
The rate of deposition
is changing,

00:32:23.870 --> 00:32:26.850
and you have to optimize your
growth process all over again.

00:32:26.850 --> 00:32:29.010
So part of the trick
of doing good MOCVD

00:32:29.010 --> 00:32:32.360
is knowing your chemistry, being
able to design or synthesize

00:32:32.360 --> 00:32:35.410
these metalorganic precursors.

00:32:35.410 --> 00:32:38.720
And the deposition process is
very sensitive to temperature,

00:32:38.720 --> 00:32:42.830
pressure, the precise surface
orientation, and preparation,

00:32:42.830 --> 00:32:46.010
what carrier gases, as well, are
mixed in with the metalorganic

00:32:46.010 --> 00:32:47.790
precursor that
you're putting in,

00:32:47.790 --> 00:32:50.382
and the byproducts obviously
need to be managed.

00:32:50.382 --> 00:32:51.920
So that's MOCVD in a nutshell.

00:32:51.920 --> 00:32:52.420
Yes?

00:32:52.420 --> 00:32:55.724
AUDIENCE: And pure quality
is much better with MOCVD

00:32:55.724 --> 00:32:57.620
than it is for
sputtering, right?

00:32:57.620 --> 00:32:59.530
PROFESSOR: It depends
on a lot of factors.

00:32:59.530 --> 00:33:01.110
So the reason the
purity of MOCVD

00:33:01.110 --> 00:33:02.750
is generally better
than sputtering

00:33:02.750 --> 00:33:04.620
is because the mass
flow controllers

00:33:04.620 --> 00:33:08.500
necessary to control
the gas flow specific

00:33:08.500 --> 00:33:10.594
for particular types of gases.

00:33:10.594 --> 00:33:12.510
Now, in sputtering,
because of the versatility

00:33:12.510 --> 00:33:15.700
of the sputtering chamber,
you could take this target out

00:33:15.700 --> 00:33:18.020
and put-- maybe
Ashley comes along

00:33:18.020 --> 00:33:19.630
into your sputtering
chamber, and she

00:33:19.630 --> 00:33:21.470
puts in another target
of another metal.

00:33:21.470 --> 00:33:22.870
And now you're depositing
two different metals

00:33:22.870 --> 00:33:23.840
in the same sputtering chamber.

00:33:23.840 --> 00:33:25.507
You're going to get
cross contamination.

00:33:25.507 --> 00:33:28.006
There are things you can do to
minimize cross contamination.

00:33:28.006 --> 00:33:30.577
You can have a chimney around
your target to prevent flakes

00:33:30.577 --> 00:33:31.510
from coming down.

00:33:31.510 --> 00:33:34.940
You could sandblast
the sidewall coating

00:33:34.940 --> 00:33:37.246
and so forth to prevent
stuff, gunk, from building up

00:33:37.246 --> 00:33:38.620
around the side,
but you're still

00:33:38.620 --> 00:33:40.578
going to get a lot of
cross contamination here.

00:33:40.578 --> 00:33:42.920
And furthermore, the
purity of your film

00:33:42.920 --> 00:33:45.020
is dictated by the
purity of your re-target.

00:33:45.020 --> 00:33:47.830
And if you go online and look
at [INAUDIBLE] or CERAC or some

00:33:47.830 --> 00:33:51.770
of the big metal selling
firms, which are essentially

00:33:51.770 --> 00:33:55.340
from where the target
manufacturers are purchasing

00:33:55.340 --> 00:33:57.230
their precursors and
they compact them

00:33:57.230 --> 00:34:01.520
and make their targets, the
target purity, or the metal

00:34:01.520 --> 00:34:05.880
purity, is only on the order
of maybe 2/9 to 6/9 pure,

00:34:05.880 --> 00:34:07.660
typically within that range.

00:34:07.660 --> 00:34:10.370
So from an MOCVD
point of view, you

00:34:10.370 --> 00:34:12.210
could do a distillation
process and increase

00:34:12.210 --> 00:34:15.810
the purity of your precursor
gas and avoid that.

00:34:15.810 --> 00:34:20.820
So I think two big reasons why
MOCVD can produce higher purity

00:34:20.820 --> 00:34:23.199
films in the
sputtered system, one

00:34:23.199 --> 00:34:26.230
is the quality of the
target, and the other,

00:34:26.230 --> 00:34:29.110
I think, bigger parameter, at
least in our growth system,

00:34:29.110 --> 00:34:31.199
is cross contamination.

00:34:31.199 --> 00:34:32.964
And whenever you
deposit, say, an EML--

00:34:32.964 --> 00:34:35.172
they have a sputtering system
there from AJA, or over

00:34:35.172 --> 00:34:37.904
at Harvard CNS, there's another
AJA sputtering system there--

00:34:37.904 --> 00:34:39.570
you're going to get
cross contamination.

00:34:39.570 --> 00:34:41.080
Just look to the log
book and see what

00:34:41.080 --> 00:34:42.288
people have tried to deposit.

00:34:42.288 --> 00:34:44.260
It gets kind of scary.

00:34:44.260 --> 00:34:47.370
PECVD, Plasma Enhanced
Chemical Vapor Deposition--

00:34:47.370 --> 00:34:51.020
so similar to the previous
variety right here,

00:34:51.020 --> 00:34:53.370
but instead of saying,
OK, we're going

00:34:53.370 --> 00:34:56.889
to put the burden of the
design, the scientific design,

00:34:56.889 --> 00:34:59.620
onto this interface right here
and on to the chemist, who

00:34:59.620 --> 00:35:02.320
has to design this molecule
that reaches a surface

00:35:02.320 --> 00:35:05.740
and breaks up in just the right
way in an orderly fashion,

00:35:05.740 --> 00:35:08.740
leaving behind the metal and
letting the other gases go

00:35:08.740 --> 00:35:10.290
away, what we're
going to do here

00:35:10.290 --> 00:35:14.360
instead is to shift the burden
of separation onto the plasma.

00:35:14.360 --> 00:35:16.370
So the centers around
the physicists.

00:35:16.370 --> 00:35:17.580
We can flow in gases.

00:35:17.580 --> 00:35:19.570
We can break them up
inside of a plasma,

00:35:19.570 --> 00:35:22.460
atomize them or, at least,
create radicalized versions

00:35:22.460 --> 00:35:24.940
of them and then
allow them to it

00:35:24.940 --> 00:35:27.710
on to the substrate--
very simple in theory.

00:35:27.710 --> 00:35:29.700
In practice, what happens
inside that plasma,

00:35:29.700 --> 00:35:31.325
depending on the
temperature, depending

00:35:31.325 --> 00:35:33.255
on the frequency
and other factors,

00:35:33.255 --> 00:35:35.630
you'll get different types--
and the pressure, especially

00:35:35.630 --> 00:35:39.834
the pressure-- you'll get
different types of molecules

00:35:39.834 --> 00:35:40.750
forming in the plasma.

00:35:40.750 --> 00:35:42.790
They may be charged, and
they'll be accelerated

00:35:42.790 --> 00:35:44.290
toward your substrate
and eventually

00:35:44.290 --> 00:35:46.740
grow and form a thin film.

00:35:46.740 --> 00:35:49.910
But depending on what
species you have up there

00:35:49.910 --> 00:35:51.960
that is being deposited
on your surface,

00:35:51.960 --> 00:35:54.780
you'll get different
types of thin films

00:35:54.780 --> 00:35:56.920
growing-- different
quality material.

00:35:56.920 --> 00:35:59.840
And so, again, this
shifts the burden back

00:35:59.840 --> 00:36:03.270
to the spectroscopist to
measure what is exactly

00:36:03.270 --> 00:36:04.930
the composition of that plasma.

00:36:04.930 --> 00:36:08.122
What is the active molecule
that's being accelerated

00:36:08.122 --> 00:36:09.330
and deposited on the surface?

00:36:09.330 --> 00:36:12.080
And usually it's some
probability distribution

00:36:12.080 --> 00:36:15.330
function of varied species.

00:36:15.330 --> 00:36:18.355
The plasma is created
by this radio frequency.

00:36:21.460 --> 00:36:23.056
Let's put it this
way; usually you

00:36:23.056 --> 00:36:27.530
have a plasma frequency
of around 13.56 megahertz.

00:36:27.530 --> 00:36:30.162
Does anybody know why
this 13.56 keeps on coming

00:36:30.162 --> 00:36:31.120
up over and over again?

00:36:31.120 --> 00:36:31.619
Yeah?

00:36:31.619 --> 00:36:34.270
AUDIENCE: [INAUDIBLE] energy
to the ionized hydrogen, right?

00:36:34.270 --> 00:36:39.486
PROFESSOR: Well, if
we're thinking about eV,

00:36:39.486 --> 00:36:42.300
that would certainly be the
energy necessary to remove

00:36:42.300 --> 00:36:45.990
the electron from
the hydrogen atom,

00:36:45.990 --> 00:36:47.686
but this is another reason.

00:36:47.686 --> 00:36:48.186
Yeah?

00:36:48.186 --> 00:36:49.910
AUDIENCE: It's a
special bend that's

00:36:49.910 --> 00:36:54.037
dedicated for these crazy
noise-emitting medical and

00:36:54.037 --> 00:36:54.870
industrial purposes.

00:36:54.870 --> 00:36:57.100
PROFESSOR: Exactly,
so this is falling

00:36:57.100 --> 00:36:59.570
within the radio
frequency regime, which

00:36:59.570 --> 00:37:01.080
would affect communications.

00:37:01.080 --> 00:37:06.200
And if everybody was allowed
to run rough shod around,

00:37:06.200 --> 00:37:08.920
creating these very
high intensity emission

00:37:08.920 --> 00:37:11.020
sources of radio
frequency waves,

00:37:11.020 --> 00:37:15.320
we would very likely have
interruptions to our police

00:37:15.320 --> 00:37:19.370
communications or maybe even
our radios or cell phones.

00:37:19.370 --> 00:37:22.340
And so, at some point,
they had to say, look,

00:37:22.340 --> 00:37:25.860
we have to assign definite
bands within the radio frequency

00:37:25.860 --> 00:37:29.490
space and allocate them
to specific purposes.

00:37:29.490 --> 00:37:31.650
In one band, they allocated
to all the scientists

00:37:31.650 --> 00:37:33.200
and medical personnel
and said, you

00:37:33.200 --> 00:37:37.025
have to operate your equipment
in these specific bands,

00:37:37.025 --> 00:37:38.400
and we'll give
you a few of them,

00:37:38.400 --> 00:37:40.769
because we know that one
frequency doesn't work for all

00:37:40.769 --> 00:37:42.060
the things you're trying to do.

00:37:42.060 --> 00:37:45.820
But for medical equipment,
for scientific equipment,

00:37:45.820 --> 00:37:49.380
and I believe even some home
electronics, like microwaves,

00:37:49.380 --> 00:37:52.320
there are specific
bands dedicated to them.

00:37:52.320 --> 00:37:56.060
And that's why we have
this 13.56 number popping

00:37:56.060 --> 00:37:57.640
up over and over again.

00:37:57.640 --> 00:38:00.100
The reality is that if
you change the frequency,

00:38:00.100 --> 00:38:01.800
you'll change the
nature of your plasma.

00:38:01.800 --> 00:38:04.950
You may change the deposition
rates and the quality

00:38:04.950 --> 00:38:06.220
of your film as well.

00:38:06.220 --> 00:38:08.640
And so there are people
who get special permits

00:38:08.640 --> 00:38:12.040
and have these radio
frequency shielded

00:38:12.040 --> 00:38:14.880
rooms, where they do experiments
outside-- or excursions

00:38:14.880 --> 00:38:17.880
outside-- of the
13.56 megahertz range.

00:38:17.880 --> 00:38:21.580
So this is PCBD-- excellent
conformal surface coverage

00:38:21.580 --> 00:38:22.230
again.

00:38:22.230 --> 00:38:24.480
Because you're biasing
your substrate,

00:38:24.480 --> 00:38:26.321
you're able to conform.

00:38:26.321 --> 00:38:27.820
The electric field
is usually always

00:38:27.820 --> 00:38:30.820
perpendicular to the
surface, and so the angle

00:38:30.820 --> 00:38:35.660
of entry of those atoms or
molecules, the ionized species,

00:38:35.660 --> 00:38:38.150
entering the surface is going
to be normal to that surface.

00:38:38.150 --> 00:38:44.160
And you can get good coverage
around rough textured surfaces.

00:38:44.160 --> 00:38:47.060
The deposition is very sensitive
to temperature, pressure,

00:38:47.060 --> 00:38:48.570
power, carrier gases.

00:38:48.570 --> 00:38:51.470
Power of the-- here,
as well, shown.

00:38:51.470 --> 00:38:53.014
And the byproducts,
as well, need

00:38:53.014 --> 00:38:55.180
to be managed because
sometimes you're sucking out--

00:38:55.180 --> 00:38:58.936
in this particular case, you
could be pulling out silane,

00:38:58.936 --> 00:39:00.810
as shown right there,
and we talked about all

00:39:00.810 --> 00:39:05.580
of the risks involved a
silane in our last class.

00:39:05.580 --> 00:39:08.440
So, as you could guess, each
of those different deposition

00:39:08.440 --> 00:39:13.140
techniques is used or is favored
for specific material systems.

00:39:13.140 --> 00:39:16.269
And we shouldn't
forget, as we talk

00:39:16.269 --> 00:39:18.060
about all these fancy
vacuum equipment that

00:39:18.060 --> 00:39:20.140
look nice and cool as you
walk through the labs,

00:39:20.140 --> 00:39:22.490
and you see these big
stainless steel chambers,

00:39:22.490 --> 00:39:25.830
we shouldn't forget about the
simpler, lower cost, lower

00:39:25.830 --> 00:39:28.170
thermal budget, lower
capital equipment cost

00:39:28.170 --> 00:39:31.170
techniques-- the solution-based
deposition methods.

00:39:31.170 --> 00:39:33.480
And these involve printing.

00:39:33.480 --> 00:39:35.560
They involve a
electrodeposition,

00:39:35.560 --> 00:39:38.080
spin casting,
colloidal synthesis,

00:39:38.080 --> 00:39:40.530
layer-by-layer deposition--
developed here at MIT--

00:39:40.530 --> 00:39:44.080
and other technologies as well.

00:39:44.080 --> 00:39:46.600
I want to point out
two technologies,

00:39:46.600 --> 00:39:50.310
in general, the first of which
is still under some development

00:39:50.310 --> 00:39:50.890
printing.

00:39:50.890 --> 00:39:53.150
Obviously, we have
inkjet printers.

00:39:53.150 --> 00:39:54.400
That's pretty straightforward.

00:39:54.400 --> 00:39:56.690
But printing
fractional solar cells

00:39:56.690 --> 00:39:58.410
is something being
commercialized

00:39:58.410 --> 00:40:01.910
by only a handful of companies,
Nanosolar being one of them.

00:40:01.910 --> 00:40:10.260
And I would say there aren't
any authoritative textbooks that

00:40:10.260 --> 00:40:12.130
will describe for
you their technology,

00:40:12.130 --> 00:40:13.540
because it's
largely under wraps.

00:40:13.540 --> 00:40:16.562
They're a startup company, and
it's not publicly available.

00:40:16.562 --> 00:40:19.020
Electrodeposition, on the other
hand, is fairly well known.

00:40:19.020 --> 00:40:21.790
You're, again, applying
a voltage difference

00:40:21.790 --> 00:40:24.820
between two electrodes, one of
which will be your substrate,

00:40:24.820 --> 00:40:27.070
and depositing a
species contained

00:40:27.070 --> 00:40:29.850
within your electrolytic
solution onto that substrate,

00:40:29.850 --> 00:40:30.940
growing your layers.

00:40:30.940 --> 00:40:33.020
Because you're growing
it at room temperature,

00:40:33.020 --> 00:40:38.754
these films, I would say,
tend to have rough surfaces.

00:40:38.754 --> 00:40:40.670
That could be a downside
of electrodeposition.

00:40:40.670 --> 00:40:42.809
They might have some
pinholes as well.

00:40:42.809 --> 00:40:44.725
But you do get fairly
large grained materials.

00:40:44.725 --> 00:40:47.445
It can be a very gentle growth
process, and, of course,

00:40:47.445 --> 00:40:49.940
the advantage is
lower temperature.

00:40:49.940 --> 00:40:52.240
So you have a variety of
different growth techniques.

00:40:52.240 --> 00:40:55.770
Let's talk about the general
issues involved with thin films

00:40:55.770 --> 00:40:58.800
in general, and then we'll dive
into the specific materials.

00:40:58.800 --> 00:41:01.400
So taking the same tact as
we've taken the full class,

00:41:01.400 --> 00:41:03.530
going from fundamentals
toward the technologies.

00:41:03.530 --> 00:41:04.066
Yes?

00:41:04.066 --> 00:41:05.482
AUDIENCE: I'm just
curious; do you

00:41:05.482 --> 00:41:08.145
know of any companies that
actually use electrodeposition?

00:41:08.145 --> 00:41:09.645
PROFESSOR: I know
of some companies.

00:41:09.645 --> 00:41:11.730
Let me think which I
can talk publicly about.

00:41:15.270 --> 00:41:18.800
So IBM, they presented at the
Electrochemical Society meeting

00:41:18.800 --> 00:41:21.230
last Monday here in Boston.

00:41:21.230 --> 00:41:22.920
They're an example
of a company that

00:41:22.920 --> 00:41:25.180
is developing
electrochemical deposition

00:41:25.180 --> 00:41:28.800
processes for material systems,
including copper zinc tin

00:41:28.800 --> 00:41:31.260
sulfide and copper indium
gallium diselenide.

00:41:31.260 --> 00:41:33.580
We'll talk about the
latter in a few slides,

00:41:33.580 --> 00:41:35.200
but that's one
example of a company.

00:41:38.240 --> 00:41:42.250
So general issues in thin
films-- thin film compounds

00:41:42.250 --> 00:41:44.020
are typically, not
always, but typically,

00:41:44.020 --> 00:41:47.600
binary, ternary, quaternary,
or multinary semiconductors.

00:41:47.600 --> 00:41:50.450
Meaning you don't have
just one element comprising

00:41:50.450 --> 00:41:51.770
the semiconductor species.

00:41:51.770 --> 00:41:55.010
You might have several, and
they form a crystal structure

00:41:55.010 --> 00:41:58.920
with repeating structure but
alternating atoms typically.

00:41:58.920 --> 00:42:03.420
And so, if you have multiple
atoms in one compound,

00:42:03.420 --> 00:42:05.730
a couple of issues
could arise and need

00:42:05.730 --> 00:42:08.360
to be controlled
to grow good films.

00:42:08.360 --> 00:42:10.990
The first involves
phase stability.

00:42:10.990 --> 00:42:14.067
What is shown right here in
mulitnary parameter space,

00:42:14.067 --> 00:42:15.900
this is the chemical
potential zinc, copper,

00:42:15.900 --> 00:42:19.720
and tin in a so-called zinc
copper tin sulfide material

00:42:19.720 --> 00:42:20.600
system.

00:42:20.600 --> 00:42:24.650
This red fin right here is
showing you the parameter space

00:42:24.650 --> 00:42:27.500
within which this
compound is stable.

00:42:27.500 --> 00:42:31.630
If your stoichiometry takes an
excursion from that red fin,

00:42:31.630 --> 00:42:35.180
you could wind up in
a bi-phase regime.

00:42:35.180 --> 00:42:37.800
Meaning you have CZTS
and something else,

00:42:37.800 --> 00:42:40.110
a copper tin sulphide,
a zinc sulfide,

00:42:40.110 --> 00:42:41.710
or some other
species that happens

00:42:41.710 --> 00:42:44.190
to be nearby in phase space.

00:42:44.190 --> 00:42:46.280
One way to think
about this is it's

00:42:46.280 --> 00:42:48.660
just you have a
homogeneous material.

00:42:48.660 --> 00:42:52.250
If you exceed a solubility limit
in one direction or another,

00:42:52.250 --> 00:42:55.270
you'll have precipitation
of a secondary phase.

00:42:55.270 --> 00:42:58.760
So you have to make sure
that in a gross perspective,

00:42:58.760 --> 00:43:02.060
on a percents basis,
you're in the right regime

00:43:02.060 --> 00:43:02.840
of stoichiometry.

00:43:02.840 --> 00:43:05.210
Stoichiometry being the
ratio of different elements

00:43:05.210 --> 00:43:06.280
in your system.

00:43:06.280 --> 00:43:08.780
So it's like cooking; you need
the right set of ingredients

00:43:08.780 --> 00:43:11.540
to make the right material.

00:43:11.540 --> 00:43:15.965
Now, that has to do with-- large
excursions from stoichiometry

00:43:15.965 --> 00:43:17.690
can result in phase
decomposition.

00:43:17.690 --> 00:43:19.990
Small excursions from
stoichiometry, a much more

00:43:19.990 --> 00:43:22.130
subtle effect can occur.

00:43:22.130 --> 00:43:25.890
Let's imagine for a moment that
we have two species comprising

00:43:25.890 --> 00:43:28.150
are binary material.

00:43:28.150 --> 00:43:30.870
One species has three
valence electrons.

00:43:30.870 --> 00:43:34.500
The other species has
five valence electrons.

00:43:34.500 --> 00:43:37.330
Now, because of a small
error in stoichiometry,

00:43:37.330 --> 00:43:39.700
maybe something in the
order of a few tens

00:43:39.700 --> 00:43:42.620
or hundreds of parts per
million in stoichiometry,

00:43:42.620 --> 00:43:44.170
we didn't get the
ratio just right.

00:43:44.170 --> 00:43:45.670
We were off by a little bit.

00:43:45.670 --> 00:43:48.540
Now we have one of our
compounds in excess

00:43:48.540 --> 00:43:51.390
and the other one in deficiency.

00:43:51.390 --> 00:43:55.540
If we have a different number
of electrons surrounding

00:43:55.540 --> 00:44:00.020
the atoms, we could wind up with
an excess free carrier density.

00:44:00.020 --> 00:44:02.250
In other words, you could
self-dope your material

00:44:02.250 --> 00:44:03.666
if you're unlucky,
in other words,

00:44:03.666 --> 00:44:06.170
if the material system
has a propensity for this.

00:44:06.170 --> 00:44:08.254
And you can change the
free carrier concentration,

00:44:08.254 --> 00:44:10.545
and because the free carrier
concentration is changing,

00:44:10.545 --> 00:44:12.460
you might even
change your mobility.

00:44:12.460 --> 00:44:14.460
So there are some
effects that can

00:44:14.460 --> 00:44:20.810
occur as a result of small
excursions from stoichiometry.

00:44:20.810 --> 00:44:22.470
As a result of the
self-doping, you're

00:44:22.470 --> 00:44:26.020
shifting the Fermi energy
inside your semiconductor.

00:44:26.020 --> 00:44:28.140
And as a result of
shifting the Fermi energy,

00:44:28.140 --> 00:44:32.325
it might lead to a
cascade series of events.

00:44:32.325 --> 00:44:33.700
There could be
other defects that

00:44:33.700 --> 00:44:36.950
form as a result of the
Fermi energy change.

00:44:36.950 --> 00:44:40.270
You could have other
so-called antisite defects.

00:44:40.270 --> 00:44:42.790
Atoms could switch positions
inside of your lattice,

00:44:42.790 --> 00:44:46.920
and as a result of that,
have very low minority

00:44:46.920 --> 00:44:49.570
carrier lifetime in
certain materials.

00:44:49.570 --> 00:44:53.490
So nailing the stoichiometry
both from a very large sense,

00:44:53.490 --> 00:44:55.141
to avoid phase
decomposition, instead

00:44:55.141 --> 00:44:56.890
of having a dalmatian
film, you have phase

00:44:56.890 --> 00:45:03.010
pure film, and from
a local perspective,

00:45:03.010 --> 00:45:05.340
once you get on to
this phase space

00:45:05.340 --> 00:45:07.850
where you can grow
your film well,

00:45:07.850 --> 00:45:10.510
you want to make sure that your
stoichiometry is controlled

00:45:10.510 --> 00:45:12.610
to avoid self-doping
and to prevent

00:45:12.610 --> 00:45:14.790
certain types of
intrinsic point defects

00:45:14.790 --> 00:45:17.550
from forming that might
lower minority carrier

00:45:17.550 --> 00:45:20.450
lifetime or change
carrier concentration,

00:45:20.450 --> 00:45:22.280
change other properties
of your film.

00:45:22.280 --> 00:45:24.530
For those who are working
on these sorts of materials,

00:45:24.530 --> 00:45:28.060
I'm happy to talk ad
nauseam about these topics,

00:45:28.060 --> 00:45:32.590
maybe after class, since this
is a more detailed subject.

00:45:32.590 --> 00:45:37.120
Another topic of interest
in thin films is grain size.

00:45:37.120 --> 00:45:39.800
At some point, grains
don't matter anymore.

00:45:39.800 --> 00:45:42.620
The grain size,
typically if you exceed

00:45:42.620 --> 00:45:45.240
the thickness of your film
by about a factor five.

00:45:45.240 --> 00:45:46.740
In other words, the
grain diameter's

00:45:46.740 --> 00:45:50.040
about five times wider than
the thickness of your film,

00:45:50.040 --> 00:45:52.050
grain size is not
as much of an issue.

00:45:52.050 --> 00:45:54.290
But if you do have
very small grains,

00:45:54.290 --> 00:45:57.120
they can impact performance,
because carriers will interact

00:45:57.120 --> 00:45:58.360
with those grain boundaries.

00:45:58.360 --> 00:46:00.580
And depending how
recombination active they are

00:46:00.580 --> 00:46:03.110
or where the grain boundary
is pinning the Fermi energy,

00:46:03.110 --> 00:46:05.310
the density of state at
that, at the grain boundary,

00:46:05.310 --> 00:46:08.130
will dictate the effect
on device performance.

00:46:08.130 --> 00:46:11.060
So these are some
very rough plots

00:46:11.060 --> 00:46:15.010
in crystalline silicon
for thin film devices

00:46:15.010 --> 00:46:16.640
and for some thicker
ones as well.

00:46:16.640 --> 00:46:19.046
So performance is a
function the grain size.

00:46:19.046 --> 00:46:21.420
And I show crystalline silicon
because the data is really

00:46:21.420 --> 00:46:23.750
well developed for it,
but you see similar types

00:46:23.750 --> 00:46:28.080
of plots for organic materials,
for some inorganic thin film

00:46:28.080 --> 00:46:30.450
materials, like
CIGS and so forth.

00:46:30.450 --> 00:46:33.760
And this convolutes a
few different parameters.

00:46:33.760 --> 00:46:35.970
You have to take into account
that the recombination

00:46:35.970 --> 00:46:37.969
activity of the grain
boundary is also a factor.

00:46:40.840 --> 00:46:43.790
The next topic, general
topic of interest,

00:46:43.790 --> 00:46:46.630
another tool that we'll want
to have an our material science

00:46:46.630 --> 00:46:48.820
toolkit as we start
designing these materials,

00:46:48.820 --> 00:46:50.470
we have to think
about the interfaces

00:46:50.470 --> 00:46:52.440
between the different materials.

00:46:52.440 --> 00:46:54.530
Especially in thin
films, interfaces

00:46:54.530 --> 00:46:57.940
are so important because
we don't much bulk anymore.

00:46:57.940 --> 00:46:59.860
So the device could
really be affected

00:46:59.860 --> 00:47:02.090
or device performance
really reduced

00:47:02.090 --> 00:47:05.410
if we don't pay proper
attention to our interfaces.

00:47:05.410 --> 00:47:07.490
What are these plots over here?

00:47:07.490 --> 00:47:11.040
These plots are used to grow
some very high efficiency

00:47:11.040 --> 00:47:15.100
materials, for example, by
MOCVD or molecular-beam epitaxy.

00:47:15.100 --> 00:47:18.080
And what is represented
on the horizontal axis

00:47:18.080 --> 00:47:19.780
is lattice constant.

00:47:19.780 --> 00:47:22.750
Lattice constant refers to the
equilibrium spacing of atoms

00:47:22.750 --> 00:47:24.070
inside of your material.

00:47:24.070 --> 00:47:27.940
So this regular repeating unit
cell that defines a crystal

00:47:27.940 --> 00:47:30.450
has a certain lattice
constant, a certain distance--

00:47:30.450 --> 00:47:33.810
physical distance--
shown here in angstroms.

00:47:33.810 --> 00:47:37.350
The energy of the gap is
shown on the vertical axis.

00:47:37.350 --> 00:47:39.920
And if we want to select two
or three of these materials

00:47:39.920 --> 00:47:42.600
to stack on top of one
another to absorb well

00:47:42.600 --> 00:47:44.820
at different portions
of the solar spectrum,

00:47:44.820 --> 00:47:48.610
we'll be choosing, for example,
one band gap at around 1.9 eV,

00:47:48.610 --> 00:47:50.386
another band gap
of 1 eV, or maybe

00:47:50.386 --> 00:47:51.760
if we want three
materials, we'll

00:47:51.760 --> 00:47:54.524
go even higher at the top
end and lower at the low end.

00:47:54.524 --> 00:47:55.940
So we'll stack
different materials

00:47:55.940 --> 00:47:58.050
on top of each other to
absorb preferentially

00:47:58.050 --> 00:48:00.420
in different regions
of the solar spectrum

00:48:00.420 --> 00:48:02.600
and hence exceed the
Shockley-Queisser efficiency

00:48:02.600 --> 00:48:05.600
limit, because now we're
absorbing well in two or three

00:48:05.600 --> 00:48:08.310
different colors as
opposed to just one.

00:48:08.310 --> 00:48:10.500
And the energy gap
here is important

00:48:10.500 --> 00:48:12.810
because you want
maybe one material

00:48:12.810 --> 00:48:14.914
at 1 eV, one material
at about 1.9 eV.

00:48:14.914 --> 00:48:16.580
But you also want to
make sure that they

00:48:16.580 --> 00:48:18.630
can grow on top of each
other, that you're not

00:48:18.630 --> 00:48:20.380
going to get a mismatch
of that interface,

00:48:20.380 --> 00:48:22.560
that the lattice constants
aren't so different that you

00:48:22.560 --> 00:48:24.680
wind up with these dangling
bonds at the interface, where

00:48:24.680 --> 00:48:26.380
you have an atom
coming down and nothing

00:48:26.380 --> 00:48:28.450
on the other side
for it to bond to.

00:48:28.450 --> 00:48:31.270
And so you need to make sure
that the materials that you

00:48:31.270 --> 00:48:34.770
grow are matched in lattice
constant but varying

00:48:34.770 --> 00:48:38.542
in band gap, if you're trying
to grow a multi-junction device,

00:48:38.542 --> 00:48:40.500
if you're trying to grow
a very high efficiency

00:48:40.500 --> 00:48:41.820
solar cell device.

00:48:41.820 --> 00:48:46.490
And so the growth or matching of
materials one on top of another

00:48:46.490 --> 00:48:49.300
is important, especially
for the multi-junctions,

00:48:49.300 --> 00:48:51.770
also for some of the
single junction materials

00:48:51.770 --> 00:48:54.730
if you really want to minimize
the interface recombination.

00:48:54.730 --> 00:48:56.460
So let's look at
this growth system

00:48:56.460 --> 00:48:59.810
up here, the one
that is typically

00:48:59.810 --> 00:49:02.100
used in high efficiency
solar cell materials.

00:49:02.100 --> 00:49:07.440
We have germanium right here,
gallium arsenide, and indium

00:49:07.440 --> 00:49:10.430
gallium phosphide,
which is essentially

00:49:10.430 --> 00:49:14.029
a mixture between
gallium phosphide up here

00:49:14.029 --> 00:49:15.320
and indium phosphide down here.

00:49:15.320 --> 00:49:17.940
You can alloy the two together
and get an indium gallium

00:49:17.940 --> 00:49:21.024
phosphide mixture and stack
these three materials on top

00:49:21.024 --> 00:49:22.940
of one another-- germanium,
gallium, arsenide,

00:49:22.940 --> 00:49:24.240
and indium gallium phosphide.

00:49:24.240 --> 00:49:25.190
They have three
different band gaps.

00:49:25.190 --> 00:49:27.550
The absorb in three different
regions of the solar spectrum.

00:49:27.550 --> 00:49:29.440
But they have a very
similar lattice constant,

00:49:29.440 --> 00:49:31.564
and so the interfaces will
be very well maintained.

00:49:31.564 --> 00:49:35.210
That's an example of using
a chart like this to design

00:49:35.210 --> 00:49:36.580
your solar cell materials.

00:49:39.760 --> 00:49:42.860
Next topic is
material abundances.

00:49:42.860 --> 00:49:46.917
If we're trying to engineer
all of these other parameters

00:49:46.917 --> 00:49:49.500
that we've been talking about--
the lattice constant, the band

00:49:49.500 --> 00:49:53.000
gap, the grain size
that also is a function

00:49:53.000 --> 00:49:55.510
of how the material grows,
the ability to self-dope.

00:49:55.510 --> 00:49:58.190
We have all of these material
issues that we have first

00:49:58.190 --> 00:49:59.800
and foremost in our minds.

00:49:59.800 --> 00:50:00.990
We go to the periodic table.

00:50:00.990 --> 00:50:02.364
We find some
compounds that work.

00:50:02.364 --> 00:50:03.890
We're really happy about it.

00:50:03.890 --> 00:50:06.250
But then, all of a sudden,
life comes along and slaps us

00:50:06.250 --> 00:50:07.860
the face and says,
well, we don't

00:50:07.860 --> 00:50:09.934
have enough of this
material to really scale

00:50:09.934 --> 00:50:12.100
to get all the way to the
terawatt cell [INAUDIBLE].

00:50:12.100 --> 00:50:14.530
Oh, I wish I had known about
this before when I first

00:50:14.530 --> 00:50:16.450
got started.

00:50:16.450 --> 00:50:20.180
So we're presenting to you
upfront the state-of-the-art

00:50:20.180 --> 00:50:22.870
of what is known about
material abundances.

00:50:22.870 --> 00:50:27.220
And these last two studies
right here, APS Energy Critical

00:50:27.220 --> 00:50:30.670
Elements and the DOE Critical
Material Strategy, both of them

00:50:30.670 --> 00:50:33.200
represent a synthesis of
the information, essentially

00:50:33.200 --> 00:50:35.440
the equivalent to the
IPCC reports in climate

00:50:35.440 --> 00:50:38.170
change, but the best
synthesis that we

00:50:38.170 --> 00:50:40.730
have right now
about the abundances

00:50:40.730 --> 00:50:42.065
of different elements out there.

00:50:42.065 --> 00:50:44.275
There are as well a
variety of different papers

00:50:44.275 --> 00:50:46.775
that have been published in the
subject over the last couple

00:50:46.775 --> 00:50:50.340
of decades or even earlier.

00:50:50.340 --> 00:50:54.020
So what we have to keep in mind
is that our stardust out there

00:50:54.020 --> 00:50:56.060
is not in infinite supply.

00:50:56.060 --> 00:50:58.530
Every element we have on
the planet that we know of

00:50:58.530 --> 00:51:00.699
came from fusion
reactions in stars,

00:51:00.699 --> 00:51:02.740
and there was a probability
distribution function

00:51:02.740 --> 00:51:06.280
of the appearance of those
elements as a function of z

00:51:06.280 --> 00:51:09.651
on the planet as a result biased
toward the lighter elements.

00:51:09.651 --> 00:51:11.150
And some of the
heavier elements are

00:51:11.150 --> 00:51:15.860
in lesser supply, that we
know of, on the Earth's crust.

00:51:15.860 --> 00:51:18.250
Not to say that the
deposits don't exist.

00:51:18.250 --> 00:51:20.470
Not to say that these
studies right here

00:51:20.470 --> 00:51:23.330
are the authoritative
end-all and be-all.

00:51:23.330 --> 00:51:25.370
We might discover next
year or next month

00:51:25.370 --> 00:51:28.680
for tomorrow huge deposit
of a particular element

00:51:28.680 --> 00:51:31.780
at a specific spot, let's
say, under the Arctic.

00:51:31.780 --> 00:51:34.790
But from what we know right
now, that's the stardust

00:51:34.790 --> 00:51:36.320
that we have to work with.

00:51:36.320 --> 00:51:38.070
These are our abundances.

00:51:38.070 --> 00:51:40.000
So if you'd like to
design around it,

00:51:40.000 --> 00:51:43.445
I'd advise looking into
those reports as well.

00:51:43.445 --> 00:51:45.320
And finally, radiation
hardness, getting back

00:51:45.320 --> 00:51:47.570
to Ashley's question, gee,
what are the most radiation

00:51:47.570 --> 00:51:49.040
hard species?

00:51:49.040 --> 00:51:53.950
This is the efficiency of solar
cell performance normalized

00:51:53.950 --> 00:51:56.130
at the very start
of a test, and this

00:51:56.130 --> 00:52:00.430
is the equivalent
radiation damage.

00:52:03.220 --> 00:52:06.460
You could also think about
this as the amount of momentum

00:52:06.460 --> 00:52:09.456
or energy depending
transferred to the atoms

00:52:09.456 --> 00:52:10.830
inside of your
semiconductor that

00:52:10.830 --> 00:52:13.300
would result in lattice
damage that would result

00:52:13.300 --> 00:52:16.604
in a decrease of minority
carrier lifetime or mobility,

00:52:16.604 --> 00:52:18.520
which ultimately would
impact cell performance

00:52:18.520 --> 00:52:19.544
and efficiency.

00:52:19.544 --> 00:52:20.960
We can see that
different material

00:52:20.960 --> 00:52:24.120
systems have different
degrees of radiation hardness.

00:52:24.120 --> 00:52:26.000
Some maintain their
high efficiency

00:52:26.000 --> 00:52:28.970
until very high radiation
dose, and others

00:52:28.970 --> 00:52:30.820
degrade much quicker.

00:52:30.820 --> 00:52:32.590
And look at this.

00:52:32.590 --> 00:52:35.490
This is a dose in orbit per
year, right around there.

00:52:35.490 --> 00:52:38.070
And you can already begin
to see that some of our most

00:52:38.070 --> 00:52:42.500
common compounds are not doing
too well out there-- not doing

00:52:42.500 --> 00:52:45.840
too well in outer space.

00:52:45.840 --> 00:52:48.910
So we have the
radiation hardness

00:52:48.910 --> 00:52:51.540
to take into account if we're
putting these solar panels

00:52:51.540 --> 00:52:53.550
out there into outer space.

00:52:53.550 --> 00:52:55.924
This is one older study
I would definitely

00:52:55.924 --> 00:52:56.840
encourage you to look.

00:52:56.840 --> 00:52:58.840
There may be some newer studies.

00:52:58.840 --> 00:53:01.340
As the solar cell
efficiency improves,

00:53:01.340 --> 00:53:03.540
they become more
sensitive because you

00:53:03.540 --> 00:53:06.580
begin decreasing your
efficiency with smaller

00:53:06.580 --> 00:53:09.550
variations in the minority
carrier diffusion length.

00:53:09.550 --> 00:53:12.070
So those charts make look a
different as time goes by.

00:53:12.070 --> 00:53:15.130
AUDIENCE: Is the effect of
the radiation cumulative?

00:53:15.130 --> 00:53:18.111
So for example, gallium
arsenide or any of these

00:53:18.111 --> 00:53:21.780
would just continue to degrade
as they're out in space?

00:53:21.780 --> 00:53:25.130
PROFESSOR: So is effective
radiation dose cumulative?

00:53:25.130 --> 00:53:28.090
I am not the expert on
this particular topic.

00:53:28.090 --> 00:53:30.940
But from what I know about
radiation exposure of detectors

00:53:30.940 --> 00:53:35.460
at synchrotrons, which
is a little similar,

00:53:35.460 --> 00:53:40.490
not quite the same, the
mechanisms involved with this

00:53:40.490 --> 00:53:42.790
essentially involve
atomic displacements

00:53:42.790 --> 00:53:43.630
within the lattice.

00:53:43.630 --> 00:53:45.710
You have atoms physically
being displaced

00:53:45.710 --> 00:53:47.760
from their equilibrium
positions as they interact

00:53:47.760 --> 00:53:49.220
with this incoming radiation.

00:53:49.220 --> 00:53:52.000
And the probability that it
occurs is a function of time,

00:53:52.000 --> 00:53:55.330
will increase per unit
volume, and hence it

00:53:55.330 --> 00:53:57.860
can be thought of as
accumulative exposure effect.

00:53:57.860 --> 00:54:00.610
The first order impact would be
on minority carrier diffusion

00:54:00.610 --> 00:54:04.440
length, impacting both
lifetime and mobility.

00:54:04.440 --> 00:54:08.230
And to the effect that you have
a relationship between exposure

00:54:08.230 --> 00:54:11.007
time-lattice displacement,
lattice displacement-minority

00:54:11.007 --> 00:54:13.173
carrier diffusion length,
minority carrier diffusion

00:54:13.173 --> 00:54:15.900
length and efficiency, you might
be able to model this effect.

00:54:15.900 --> 00:54:18.600
That would be my
uninformed answer.

00:54:18.600 --> 00:54:20.800
Again, you might want to
look this up yourself.

00:54:20.800 --> 00:54:21.345
Question?

00:54:21.345 --> 00:54:23.720
AUDIENCE: Do you know why the
cadmium telluride improves?

00:54:23.720 --> 00:54:26.620
PROFESSOR: Oh, why does it
improve with efficiency?

00:54:26.620 --> 00:54:28.680
I don't know
specifically about why

00:54:28.680 --> 00:54:30.214
that is for this
particular case,

00:54:30.214 --> 00:54:31.880
but do know that some
materials are what

00:54:31.880 --> 00:54:33.780
are called defect tolerant.

00:54:33.780 --> 00:54:37.840
Some are more naturally able
to withstand antisite defects

00:54:37.840 --> 00:54:41.360
or a certain concentration
of damage, internal surfaces,

00:54:41.360 --> 00:54:43.670
voids, grain boundaries.

00:54:43.670 --> 00:54:48.940
Cad-tel, cadmium telluride,
is fairly defect tolerant.

00:54:48.940 --> 00:54:54.190
It's one of nature's gifts
to humanity in that regard.

00:54:54.190 --> 00:54:57.250
The degree to which a material
can be defect tolerant

00:54:57.250 --> 00:54:58.927
depends partly on
the carrier density.

00:54:58.927 --> 00:55:00.510
If you have very
high carrier density,

00:55:00.510 --> 00:55:03.950
you tend to screen defects.

00:55:03.950 --> 00:55:07.430
Another reason why they could
be defect tolerant-- all

00:55:07.430 --> 00:55:10.305
of these compounds are
somewhere between an ionic

00:55:10.305 --> 00:55:12.350
and a covalent semiconductor.

00:55:12.350 --> 00:55:15.370
In a covalent semiconductor,
those materials

00:55:15.370 --> 00:55:23.060
tend to be very defect
intolerant because there's

00:55:23.060 --> 00:55:25.930
the conduction band and valence
band as a function of position,

00:55:25.930 --> 00:55:26.981
tend to be very flat.

00:55:26.981 --> 00:55:29.230
The material tends to be
fairly homogeneous throughout

00:55:29.230 --> 00:55:31.370
the electron densities,
fairly well distributed

00:55:31.370 --> 00:55:33.000
in a covalently bonded material.

00:55:33.000 --> 00:55:35.650
In an ionic material, you tend
to have charge localization.

00:55:35.650 --> 00:55:38.090
So energy as a
function of position

00:55:38.090 --> 00:55:40.205
might look like this
on an atomic scale

00:55:40.205 --> 00:55:42.580
as you go from one atom to
the next atom on your lattice,

00:55:42.580 --> 00:55:44.170
to the next one, to the
next one, to the next one,

00:55:44.170 --> 00:55:46.070
and that reflects the
localization of charge

00:55:46.070 --> 00:55:47.050
in your material.

00:55:47.050 --> 00:55:50.590
Those materials can be
more defect intolerant,

00:55:50.590 --> 00:55:53.470
because conduction can happen
more from a hopping mechanism

00:55:53.470 --> 00:55:57.300
than from a band
conduction mechanism,

00:55:57.300 --> 00:55:59.010
and this is really
a gradient between,

00:55:59.010 --> 00:56:01.850
and most materials, they tend
to be partially ionic partially

00:56:01.850 --> 00:56:04.550
covalent going
down the list here.

00:56:04.550 --> 00:56:07.170
And there's a bit of a shift
between cadmium sulfide

00:56:07.170 --> 00:56:10.680
in zinc sulfide in terms
of the ionicity covalence.

00:56:10.680 --> 00:56:14.510
So cad-tel would be
with the chalcogen

00:56:14.510 --> 00:56:16.530
two lower than sulfur
on the periodic table.

00:56:16.530 --> 00:56:18.821
I would imagine it would be
at this transition as well,

00:56:18.821 --> 00:56:21.020
but I'd have to look that up.

00:56:21.020 --> 00:56:24.030
Gives a place to get started
and read more about it.

00:56:24.030 --> 00:56:25.910
And then reliability
and degradation--

00:56:25.910 --> 00:56:27.310
this is important.

00:56:27.310 --> 00:56:28.810
This is a crystalline
silicon module

00:56:28.810 --> 00:56:30.880
being loaded into
environmental testing chamber.

00:56:30.880 --> 00:56:32.296
Inside of that
chamber, the module

00:56:32.296 --> 00:56:34.300
is going to be put
through hell and back.

00:56:34.300 --> 00:56:36.117
The temperature is
going to be raised.

00:56:36.117 --> 00:56:37.700
The humidity is going
to be pumped in.

00:56:37.700 --> 00:56:39.140
Sometimes ultraviolet
light is even

00:56:39.140 --> 00:56:41.400
put in there in some of the
more modern advanced ones.

00:56:41.400 --> 00:56:44.430
And then the temperature
can drop down

00:56:44.430 --> 00:56:47.564
to temperatures as low
as minus 40 degrees C,

00:56:47.564 --> 00:56:49.480
depending on what the
environmental chamber is

00:56:49.480 --> 00:56:51.890
designed to do, exactly
how it's designed

00:56:51.890 --> 00:56:53.820
to stress or test your module.

00:56:53.820 --> 00:56:56.770
And the idea is to promote
an accelerated degradation

00:56:56.770 --> 00:57:00.215
of the module on purpose to test
what its failure modes will be,

00:57:00.215 --> 00:57:02.340
and we'll see these we go
take a tour of Fraunhofer

00:57:02.340 --> 00:57:04.310
CSE in a couple weeks.

00:57:04.310 --> 00:57:06.540
This is a crystalline silicon
module being loaded in.

00:57:06.540 --> 00:57:09.665
If you were to put a
thin film material--

00:57:09.665 --> 00:57:12.440
and crystalline silicon are
materials very, very thick,

00:57:12.440 --> 00:57:16.030
and we said at the native
oxide was very tenacious.

00:57:16.030 --> 00:57:19.430
It was only a few tens or maybe
hundreds of angstroms thick,

00:57:19.430 --> 00:57:23.604
and the junction depth was
about on the order of a micron.

00:57:23.604 --> 00:57:25.520
If you have water attacking
the surface of you

00:57:25.520 --> 00:57:27.960
silicon wafer, water
vapor, really not too

00:57:27.960 --> 00:57:30.500
much of an issue, and silicon's
is fairly inert anyway.

00:57:30.500 --> 00:57:34.150
But now if you take a fairly
reactive material, a thin film

00:57:34.150 --> 00:57:37.110
material that might react with
air or might react with water,

00:57:37.110 --> 00:57:43.770
and it's so thin that the these
rusting modes, or reaction

00:57:43.770 --> 00:57:45.630
modes, the weathering
modes, can really

00:57:45.630 --> 00:57:49.437
impact a large fraction of
the thickness of your device.

00:57:49.437 --> 00:57:51.020
Now you've become a
lot more sensitive

00:57:51.020 --> 00:57:52.730
to accelerated degradation.

00:57:52.730 --> 00:57:55.290
Now you've become a lot more
sensitive to the elements,

00:57:55.290 --> 00:57:57.700
and this includes
both oxygen and water.

00:57:57.700 --> 00:58:03.400
So if the ambient
is able to penetrate

00:58:03.400 --> 00:58:05.770
through the encapsulate and
get to the active absorber

00:58:05.770 --> 00:58:08.220
material, you may have
accelerated degradation

00:58:08.220 --> 00:58:10.200
of module performance
as a result.

00:58:10.200 --> 00:58:12.229
And so there also
some unique failure

00:58:12.229 --> 00:58:13.520
modes within thin film modules.

00:58:13.520 --> 00:58:17.040
If you have two different
species comprising

00:58:17.040 --> 00:58:18.990
your compound, one
of them might be

00:58:18.990 --> 00:58:20.980
prone to move in electric field.

00:58:20.980 --> 00:58:24.250
For example, copper is
notorious for zipping along

00:58:24.250 --> 00:58:26.566
in electric field,
in electromigrating.

00:58:26.566 --> 00:58:27.940
And so that's a
failure mode that

00:58:27.940 --> 00:58:31.920
doesn't exist in large thick
crystalline silicon modules

00:58:31.920 --> 00:58:33.430
but could existent
in thin films.

00:58:33.430 --> 00:58:35.960
And so because of all of this,
and because of the growing

00:58:35.960 --> 00:58:38.860
realization that the way we
test crystalline silicon modules

00:58:38.860 --> 00:58:42.370
and drive them failure
is not the same

00:58:42.370 --> 00:58:45.610
that we might be able to achieve
failure in a thin film module.

00:58:45.610 --> 00:58:47.350
There are newer
testing protocols,

00:58:47.350 --> 00:58:50.750
such as this IEC 61853,
that have been introduced

00:58:50.750 --> 00:58:55.070
in an attempt to do test
appropriately thin film

00:58:55.070 --> 00:58:57.340
modules for their
respective failure modes.

00:58:57.340 --> 00:59:00.660
And this is, I would say,
still work in progress.

00:59:00.660 --> 00:59:02.520
So much so, that we
have a group project

00:59:02.520 --> 00:59:04.780
focused in part on this.

00:59:04.780 --> 00:59:07.499
It's still a work
in process to try

00:59:07.499 --> 00:59:09.040
to figure out how
do we appropriately

00:59:09.040 --> 00:59:11.930
test these thin film
modules toward the point

00:59:11.930 --> 00:59:14.180
where they can fail.

00:59:14.180 --> 00:59:16.670
Any questions so
far on these topics?

00:59:16.670 --> 00:59:21.100
Because these are general issues
that will affect all thin film

00:59:21.100 --> 00:59:23.020
materials, I wanted
to make sure that we

00:59:23.020 --> 00:59:24.920
were comfortable with
these general topics

00:59:24.920 --> 00:59:27.850
before we dove in any
detail into the technologies

00:59:27.850 --> 00:59:28.430
themselves?

00:59:28.430 --> 00:59:28.929
Yes?

00:59:28.929 --> 00:59:30.982
AUDIENCE: A question
about lattice matching--

00:59:30.982 --> 00:59:33.065
is it important to lattice
match the semiconductor

00:59:33.065 --> 00:59:34.520
to the contact as well?

00:59:34.520 --> 00:59:36.950
Or is that not as important

00:59:36.950 --> 00:59:38.640
PROFESSOR: So the
question was is it

00:59:38.640 --> 00:59:42.180
important to lattice match the
semiconductor to the contact?

00:59:42.180 --> 00:59:46.900
So let me emphasize that in
many semiconductor contact

00:59:46.900 --> 00:59:50.260
combinations you would have
a highly doped semiconductor

00:59:50.260 --> 00:59:52.840
right before the contact, a
very localized region of highly

00:59:52.840 --> 00:59:55.410
doped semiconductor that would
create a tunneling injunction.

00:59:55.410 --> 00:59:58.432
In that case, the density
of states at the interface

00:59:58.432 --> 01:00:00.640
doesn't matter because you
have a tunneling junction.

01:00:00.640 --> 01:00:02.973
You're be able to tunnel
straight from the semiconductor

01:00:02.973 --> 01:00:03.790
into the contact.

01:00:03.790 --> 01:00:05.564
The lattice matching
would matter, though,

01:00:05.564 --> 01:00:07.230
if you didn't have a
tunneling junction.

01:00:07.230 --> 01:00:09.954
If you had a regular
Schottky ohmic contact,

01:00:09.954 --> 01:00:11.870
then you might have to
worry about the density

01:00:11.870 --> 01:00:13.745
of interface states,
which would be regulated

01:00:13.745 --> 01:00:15.820
by the number of dangling
bonds, and then you

01:00:15.820 --> 01:00:17.830
might want every
single atom pairing up

01:00:17.830 --> 01:00:19.690
with a neighbor
on the other side.

01:00:19.690 --> 01:00:21.260
So lattice matching would be
important for the contacts

01:00:21.260 --> 01:00:21.759
there.

01:00:24.180 --> 01:00:26.640
All right, fun
stuff-- wow, we've

01:00:26.640 --> 01:00:29.770
had a good dose of material
science of the day.

01:00:29.770 --> 01:00:31.510
Thin film cost
structure-- I just wanted

01:00:31.510 --> 01:00:33.010
to highlight one quick thing.

01:00:33.010 --> 01:00:35.827
This material right here, that's
not the absorber material.

01:00:35.827 --> 01:00:37.785
The absorber material is
a really tiny fraction

01:00:37.785 --> 01:00:39.120
of the material.

01:00:39.120 --> 01:00:42.300
This comprises the other
materials within the module

01:00:42.300 --> 01:00:43.380
as well.

01:00:43.380 --> 01:00:46.530
So the encapsulates,
the glass, and so fort,

01:00:46.530 --> 01:00:49.470
just keep that in mind
as a kind of asterisks.

01:00:49.470 --> 01:00:51.295
So it's typical thin
film cost structure.

01:00:51.295 --> 01:00:52.680
It evolves with time.

01:00:52.680 --> 01:00:54.380
This is a few years
old, this slide,

01:00:54.380 --> 01:00:57.690
but it gives you a
sense, a feeling.

01:00:57.690 --> 01:01:02.740
In terms of global production,
this is a year-old data now

01:01:02.740 --> 01:01:05.360
from 2010.

01:01:05.360 --> 01:01:10.080
During this past year-- so this
was 2009 data shown in 2010.

01:01:10.080 --> 01:01:17.000
The 2010 market was very harsh
for the thin film producers,

01:01:17.000 --> 01:01:20.550
many of which tend to be in the
United States and in Europe.

01:01:20.550 --> 01:01:23.780
In 2010, prices
dropped quite a bit,

01:01:23.780 --> 01:01:25.785
and that favored the
low-cost Chinese solar cell

01:01:25.785 --> 01:01:27.160
manufacturers,
many of which were

01:01:27.160 --> 01:01:29.170
invested in crystalline
silicon technology.

01:01:29.170 --> 01:01:34.370
So by no detriment
to the technology

01:01:34.370 --> 01:01:37.590
itself, market
dynamics worldwide,

01:01:37.590 --> 01:01:40.320
due to other factors, tended
to favor crystalline silicon

01:01:40.320 --> 01:01:41.104
in the past year.

01:01:41.104 --> 01:01:42.520
And the market
share of thin films

01:01:42.520 --> 01:01:45.550
contracted a bit so it's now
about 90% crystalline silicon,

01:01:45.550 --> 01:01:49.410
10% then films
worldwide in 2010.

01:01:49.410 --> 01:01:52.290
But the break down between
the different thin film

01:01:52.290 --> 01:01:55.210
technologies, we had the
so-called cadmium telluride,

01:01:55.210 --> 01:01:57.660
CIGS, so that's Copper
Indium Gallium Diselenide,

01:01:57.660 --> 01:01:59.200
and amorphous silicon.

01:01:59.200 --> 01:02:02.280
And the dynamic
between 2009-2010

01:02:02.280 --> 01:02:04.800
was that the amorphous
silicon shrank a bit.

01:02:04.800 --> 01:02:08.160
CIGS grew, and cad-tel continued
growing, but more marginally

01:02:08.160 --> 01:02:10.240
because it was already
big to begin with.

01:02:10.240 --> 01:02:13.124
So you could think of
this red portion growing

01:02:13.124 --> 01:02:15.540
at the expense of the green,
if you want to translate this

01:02:15.540 --> 01:02:18.970
into 2010 numbers.

01:02:18.970 --> 01:02:22.230
So what is CIGS, cad-tel,
amorphous silicon--

01:02:22.230 --> 01:02:25.130
what are those materials?

01:02:25.130 --> 01:02:26.540
Well, we'll get into that.

01:02:26.540 --> 01:02:30.330
I think the best thing to do is
to leave this for next class.

01:02:30.330 --> 01:02:32.930
I'll briefly go
over cad-tel just

01:02:32.930 --> 01:02:34.710
because it is so important.

01:02:34.710 --> 01:02:38.470
It is the biggest-- the
single biggest US solar cell

01:02:38.470 --> 01:02:41.420
manufacturer is producing
cadmium telluride solar cells,

01:02:41.420 --> 01:02:42.860
or cad-tel for short.

01:02:42.860 --> 01:02:46.980
And to make just a description
of what you're solar panel

01:02:46.980 --> 01:02:48.620
would look like
in cross section,

01:02:48.620 --> 01:02:51.520
this is your glass
on the backside here.

01:02:51.520 --> 01:02:54.000
This ITO Indium Tin Oxide.

01:02:54.000 --> 01:02:55.620
It is a what?

01:02:55.620 --> 01:02:57.930
A transparent conducting
oxide, very good.

01:02:57.930 --> 01:03:00.082
So the ITO is a transparent
connecting oxide.

01:03:00.082 --> 01:03:02.290
Your light, your sunlight,
is coming in through here.

01:03:02.290 --> 01:03:04.200
So this is electrically
conductive layer,

01:03:04.200 --> 01:03:08.160
but it's transparent, so it's
allowing the sunlight through.

01:03:08.160 --> 01:03:10.330
Tin oxide, we'll get
to that in a second.

01:03:10.330 --> 01:03:12.700
Cadmium sulfide and
cadmium telluride--

01:03:12.700 --> 01:03:15.030
so the cadmium
telluride layer is

01:03:15.030 --> 01:03:17.020
a layer that's absorbing
most of the sunlight

01:03:17.020 --> 01:03:19.070
and producing
electron hole pairs.

01:03:19.070 --> 01:03:22.610
The cadmium sulfide is
providing the header junction

01:03:22.610 --> 01:03:25.520
separating charge at the header
junction between the cad-tel

01:03:25.520 --> 01:03:27.110
and the cad-sulfide.

01:03:27.110 --> 01:03:30.260
This tin oxide is generally
an intrinsic layer.

01:03:30.260 --> 01:03:35.237
It's assisting here with the
ITO on the front contact,

01:03:35.237 --> 01:03:37.570
and then you have your back
contact that extracts charge

01:03:37.570 --> 01:03:38.910
from the back.

01:03:38.910 --> 01:03:40.960
There are a couple
more tricks to creating

01:03:40.960 --> 01:03:44.710
a good cad-tel device that
involve chlorine treatment

01:03:44.710 --> 01:03:46.590
and passivation of defects.

01:03:46.590 --> 01:03:49.420
That's where some of
the activation comes in.

01:03:49.420 --> 01:03:51.450
This is another view
of the cad-tel device

01:03:51.450 --> 01:03:54.100
in cross section,
another example.

01:03:54.100 --> 01:03:56.100
This transparent conducting
oxide, in this case,

01:03:56.100 --> 01:04:02.130
is fluorine doped tin
oxide, another TCO material.

01:04:02.130 --> 01:04:04.480
But very similar structure
here, the cad-tel

01:04:04.480 --> 01:04:06.390
being the p-type
material, and cad-sulfide

01:04:06.390 --> 01:04:07.901
the n-type material.

01:04:07.901 --> 01:04:10.150
The thicknesses of these
different layers you can see.

01:04:10.150 --> 01:04:12.740
The cad-tel is only
a few microns thick,

01:04:12.740 --> 01:04:15.960
and the cad-sulfide
this is even thinner.

01:04:15.960 --> 01:04:18.710
It's a very thin layer just
serving to separate the charge.

01:04:18.710 --> 01:04:25.260
The band diagram of a cad-tel
solar cell is shown right here.

01:04:25.260 --> 01:04:28.530
We have the cad-tel here and
the cad-sulphide right here.

01:04:28.530 --> 01:04:29.930
So you can see the junction.

01:04:29.930 --> 01:04:34.890
Notice, because of the
thickness of this layer,

01:04:34.890 --> 01:04:36.650
the band bending
at these interfaces

01:04:36.650 --> 01:04:38.950
extends quite an
appreciable percentage

01:04:38.950 --> 01:04:40.550
of the total thickness
of your device.

01:04:40.550 --> 01:04:41.966
Whereas in crystalline
silicon, we

01:04:41.966 --> 01:04:44.750
have the band bending right near
the junction region, so right

01:04:44.750 --> 01:04:47.580
within a few hundreds of
nanometers, maybe a micron away

01:04:47.580 --> 01:04:50.710
from the junction, and the
device is 100 microns thick.

01:04:50.710 --> 01:04:54.200
So we had 100 microns
approximately of this flat band

01:04:54.200 --> 01:04:56.360
condition, at least in the dark.

01:04:56.360 --> 01:04:59.889
Here we have bending extending
an appreciable percentage

01:04:59.889 --> 01:05:01.430
of the total thickness
of our device,

01:05:01.430 --> 01:05:03.805
just by virtue of the fact
that we have such a thin film.

01:05:06.460 --> 01:05:12.920
And the characteristics,
the deposition technology

01:05:12.920 --> 01:05:16.180
of cad-tel, as I said, it's
nature's gift to humankind.

01:05:16.180 --> 01:05:18.160
If you put cadmium and
tellurium in together

01:05:18.160 --> 01:05:22.530
in a pot and start heating it
up, what evaporates is cad-tel.

01:05:22.530 --> 01:05:24.950
It congruently evaporates.

01:05:24.950 --> 01:05:28.320
So you could use a process
called close space sublimation,

01:05:28.320 --> 01:05:31.016
where you essentially
sublime your cad-tel,

01:05:31.016 --> 01:05:32.640
and you deposit it
onto your substrate.

01:05:32.640 --> 01:05:34.410
If you try to do this
with most other compounds

01:05:34.410 --> 01:05:35.784
in the periodic
table, you'll get

01:05:35.784 --> 01:05:38.679
either one element or the other
element evaporating first.

01:05:38.679 --> 01:05:39.970
They'll create an overpressure.

01:05:39.970 --> 01:05:40.960
They'll evaporate
off, and you won't

01:05:40.960 --> 01:05:42.626
get your compound
depositing, but you'll

01:05:42.626 --> 01:05:45.200
get one element depositing on
your substrate preferentially.

01:05:45.200 --> 01:05:48.120
cad-tel, again, nature's
gift to humankind.

01:05:48.120 --> 01:05:51.260
The two come off together and
form a cad-tel layer and so

01:05:51.260 --> 01:05:54.130
that congruent evaporation makes
it very nice from a deposition

01:05:54.130 --> 01:05:57.050
point of view-- very low cost,
high throughput deposition

01:05:57.050 --> 01:06:00.240
process for solar benefits from.

01:06:00.240 --> 01:06:03.290
Environmental
concerns-- cadmium has

01:06:03.290 --> 01:06:08.120
raised quite a bunch of
concerns amongst folks

01:06:08.120 --> 01:06:12.650
in environmental groups because
it's a known carcinogen.

01:06:12.650 --> 01:06:17.740
It is responsible
in Japan for, I

01:06:17.740 --> 01:06:20.300
believe it was called
the itai-itai ban, which

01:06:20.300 --> 01:06:22.320
means "the ouch-ouch disease."

01:06:22.320 --> 01:06:25.610
It was a disease
that was acquired

01:06:25.610 --> 01:06:28.950
by folks exposed to cadmium
during manufacturing.

01:06:28.950 --> 01:06:31.510
And as a result,
cad-tel solar panels

01:06:31.510 --> 01:06:34.070
are not allowed to be
installed in Japan.

01:06:34.070 --> 01:06:37.830
So First Solar cannot sell
its cad-tel products in Japan.

01:06:37.830 --> 01:06:41.650
There are very tightly
regulated emissions laws

01:06:41.650 --> 01:06:44.460
in the EU and the United States,
especially in the EU, where

01:06:44.460 --> 01:06:48.560
cradle-to-grave recycling
of cad-tel solar panels

01:06:48.560 --> 01:06:49.479
are necessary.

01:06:49.479 --> 01:06:51.270
So you'll see a lot of
cad-tel solar panels

01:06:51.270 --> 01:06:54.360
in large field installations or
in commercial buildings where

01:06:54.360 --> 01:06:56.955
it's very easy to collect
them all after their 20-

01:06:56.955 --> 01:06:59.600
or 25-year life span and bring
them back to the factory,

01:06:59.600 --> 01:07:01.141
as opposed to having
them distributed

01:07:01.141 --> 01:07:03.160
amongst hundreds of
thousands of smaller systems

01:07:03.160 --> 01:07:05.330
deposited on rooftops.

01:07:05.330 --> 01:07:09.780
The arguments in favor of having
cadmium inside of solar panels

01:07:09.780 --> 01:07:10.960
is the following.

01:07:10.960 --> 01:07:12.810
It's better to tie
up cadmium inside

01:07:12.810 --> 01:07:16.580
of a relatively inert compound,
cad-tel, then to have it go off

01:07:16.580 --> 01:07:19.240
and cause problems on its own.

01:07:19.240 --> 01:07:23.880
If you heat it up, the cad-tel
would evaporate congruently.

01:07:23.880 --> 01:07:27.010
Typically cadmium is
so-called "negligible" amounts

01:07:27.010 --> 01:07:29.510
are released during fires, and
they put it in between quotes

01:07:29.510 --> 01:07:32.290
because these are studies,
very good studies,

01:07:32.290 --> 01:07:37.570
and I trust the work coming
out of the [INAUDIBLE] group

01:07:37.570 --> 01:07:38.410
very much.

01:07:38.410 --> 01:07:41.470
His critics would argue
that, well, the studies

01:07:41.470 --> 01:07:43.820
were paid for, in
part, by First Solar,

01:07:43.820 --> 01:07:46.140
so how do you trust
studies like that?

01:07:46.140 --> 01:07:48.570
I would counter and say,
this is a pretty good group.

01:07:48.570 --> 01:07:50.730
Out of all the people
to life cycle analysis,

01:07:50.730 --> 01:07:54.410
he's within the top tier.

01:07:54.410 --> 01:07:57.430
So it's some
question as to that.

01:07:57.430 --> 01:07:59.554
People do question were
the temperatures used

01:07:59.554 --> 01:08:01.970
in these studies representative
of what you would actually

01:08:01.970 --> 01:08:04.030
get in the hot zone of
a fire and so forth.

01:08:04.030 --> 01:08:05.830
There's a public fear
and perception issue.

01:08:05.830 --> 01:08:06.810
It's a big deal.

01:08:06.810 --> 01:08:14.950
And the folks would argue that
much less cadmium is released

01:08:14.950 --> 01:08:16.955
per kilowatt hour than,
say, in a battery, where

01:08:16.955 --> 01:08:20.850
we would use a nickel cadmium
battery, for instance.

01:08:20.850 --> 01:08:24.080
And safe production methods
now-- fully automated,

01:08:24.080 --> 01:08:25.670
and recycling is
guaranteed by law.

01:08:25.670 --> 01:08:28.699
So have arguments in
favor and against.

01:08:28.699 --> 01:08:30.240
I'm going to stop
right here, and I'm

01:08:30.240 --> 01:08:33.020
going to pull aside the
teams during the last 15

01:08:33.020 --> 01:08:34.170
minutes of class.

01:08:34.170 --> 01:08:35.470
I emailed to you.

01:08:35.470 --> 01:08:38.810
If those of you had checked
your email before last night

01:08:38.810 --> 01:08:41.260
at 5 o'clock, should have
received an email saying you're

01:08:41.260 --> 01:08:43.260
on a particular project team.

01:08:43.260 --> 01:08:45.344
Find your partners,
cluster together.

01:08:45.344 --> 01:08:46.760
Joe and I are going
to come around

01:08:46.760 --> 01:08:48.399
to spend a few minutes
with each of you

01:08:48.399 --> 01:08:51.520
just to make sure that
our first steps are clear

01:08:51.520 --> 01:08:54.620
and that we have a forward
path and we gain some momentum.

01:08:54.620 --> 01:08:57.609
So self-assemble and don't
leave the room before the chance

01:08:57.609 --> 01:08:59.690
to come talk to you.