The MIT Energy Initiative: Sustainable Energy and Terawatt-Scale Photovoltaics


Uploaded by GoogleTechTalks on 16.11.2009

Transcript:
Doug Spreng: We're excited to be here to talk to you about MIT's Energy Initiative and some
of the specifics within that. My name is Doug Spreng. I'm a MIT alumnus Class of '65. I'll
be the host for our two speakers here today, who I'll introduce in a second.
First I'd like to express my appreciation to Kevin Chen, who unfortunately couldn't
be with us today, but he helped set this up and also Alice Ryan, who's helped us with
the logistics, so thank you, Alice.
So as we all know, the world's energy challenges are enormous in scope and it would be hard
enough just to meet the growth in energy requirements, but when you start talking about replacing
petroleum fuel based ways of, or, or coal based ways of generating energy with renewable
and you consider the terawatt scale that you're talking about and everything, you realize
that to address this challenge you need some major, major breakthroughs in the cost and
the scalability of the renewable energy.
And we're seeing some successes here and there along the way, but we're a long, long way
from achieving that result.
One of neat things is though is trained engineers or scientists or people who have worked in
technology industries here at Google, for instance, and elsewhere in Silicon Valley,
we know that technology can be a major leverage factor in terms of addressing issues like
cost and scalability. And so that's the one of the reasons for being optimistic about
what we can do.
So we're here this afternoon to explore with, with you all some of the possibilities of
addressing this challenge particularly in the area of solar and we'll tell you a lot
more about it once we get into it.
We call this program MITei On The Road and it's the MIT Energy Initiative so M-I-T-e-i
Energy Initiative, "mighty" is how we pronounce it. The On The Road basically means that we
wanna bring MIT to you, and tell you more about the research that's going on at MIT
in the area of energy. Because we know many of you are extremely interested in this and,
and we wanna tell our story.
The Energy Initiative, and I need to advance this slide, okay, is this what I wanted to
do? Yeah.
The Energy Initiative at MIT was kicked off about three or four years ago by Susan Hockfield,
who was at that time the new President and since then at this time even over 20 percent
of the faculty is, is actively engaged in energy research.
And there's an MIT Energy Club that was established about the same time; it currently has 1700
members. It's a completely student run organization and, and, and the consequence of looking at
this is the, is that energy has become a huge program at MIT and has evoked a really passionate
response on the part of both students and faculty.
Daniel Enderton who is sitting down here will be our first speaker. He's a former President
of the MIT Energy Club. He's gonna give you a brief overview of MITei followed by a more
in depth presentation on the Sustainable Energy Revolutions Program or affectionately called
SERP, S-E-R-P. So Daniel's gonna tell us about that.
First a little history on, on, on why we're doing this MITei On The Road.
In April of this year, the MIT Alumni Association conducted a survey of its alumni and that
wanted to find out, they wanted to test basically the alumni's interest and, and ideas for further,
more, more significant engagement in the area of energy and environment.
And so we designed a survey for the alumni and sampled the base and, and the results
that we got were pretty amazing. And, and one of the things was surprising about the
results were that they were totally uniform. It didn't matter what kind of course you took,
what, what generation you graduated in, or anything like, whether, whether you donated
or not donated to the Institute; everybody had the same kind of feeling about energy
and the environment.
And in fact, 70 percent of the alumni surveyed described their personal interest as either
strong or passionate. And, and about 20 some percent of them actually said 'passionate',
which is kind of an off the charts kind of a result, I think.
I don't know how many of you have ever answered passionate to a survey I [laughs] it's the
first time I ever did, so, so I found this kind of amazing.
And so we felt like we really had something, but it gave us the, the reason for bringing
MIT more out basically on the road, if you will, to, to the people that are out in the
regions. 'Cause the regions, people in the regions said, "Hey we don't wanna have to
go back to Cambridge, Massachusetts to find out what's goin' on. We'd like to know more
about it where we live and work." And so that's one of the reasons why were doing this here.
Also in the survey alumni indicated their areas of interest. And the biggest one was
in the area of renewable energy particularly solar was the strongest; wind was number two;
and also in storage as a key enabler for being able to store these more intermittent sources
of energy to, to be used for later on.
And so those were all really oriented, indicated as very, very important subjects to, to go
over and the technology research going on in those areas what they want to know about.
So that's why we asked Professor Tonio Buonassisi to come here; talk specifically about solar.
It seemed to be one of the more interesting areas for just our MIT alums, but I think
out here in California, it's even more sig, significant.
Now why would we bring MIT to Google? Well, yeah we know there's quite a few MIT alumni
here, but a lot more that are not.
And two reasons I think: number one is your company is extremely well known as a very
environmentally oriented and sensitive company, and we understand that, that there's a lot
of personal interest in the employees here as well. And, and the second thing is a large
number of technical people and who can grasp the technical significance of the kind of
work that we're doing. So, so those are the two major reasons for this.
But, we have to admit it's kind of a test to see how it goes; see what the reaction
is and so on.
Now of course out here in California the solar revolution is in, in full force; you got dozens
of solar startups; you got millions of dollars of venture capital money that have gone into
this in the Bay area alone and the State of California with its Million Rooftops program
is promoting solar on the roof and so on and so forth. So we live in an environment out
here where this is a really big issue and we're blessed with having wonderful climate
too to help solar along.
So that's another reason why we asked Tonio Buonassisi to be our first speaker in what
we hope to be actually a series of events. So thinking maybe about once every three months
to have a new speaker in, in these topical areas that, that we mentioned.
Okay, so, so Professor Buonassisi is gonna follow Daniel Enderton in the program. We'll
talk about some of the technical challenges and some of the possibilities of how we can
scale up to the terawatt level with solar photovoltaics.
So here's the agenda. I just did your welcome. Daniel's gonna give the overview of the Energy
Initiative and then followed by Tonio to the terawatt-scale photovoltaics and then we'll
have time for Q & A at the end as long as you'd like actually.
So Daniel's gonna be about 15, 20 minutes; Tonio's gonna be about 30 or so and then Q
& A.
There we go. Daniel, take it over.
Daniel Enderton: Thanks, Doug.
So as Doug indicated I work for the Energy Initiative. I'm the Executive Director of
the Sustainable Energy Revolutions Program which is basically an extremely long winded
way of saying that my job is to help build renewable energy research programs.
And so actually what I wanna do real quick is just kind of give you a quick overview
of some of the things that are going on in energy at MIT. I would be a potentially good
sort of resource just to sort of help facilitate connections between yourselves as engineers
and engineers back at, at MIT, if that's of interest.
And I'll also get to talk a little bit more about my program, something sort of close
to my heart.
So actually I don't know if you can read this from back here, but we are fortunate to have
the, the President come by a couple weeks ago and he gave a, a talk on energy, but also
did a quick whirlwind tour of some labs beforehand. And some students talked their Professor Vladimir
Bulovic into getting the President to sign a deposition chamber for him and so it says,
"Good work. Barack Obama." And so now they're trying to figure out how to sort of encapsulate
that so that at, the rigors of lab use doesn't sort of wear it away over time.
So as, as Doug said the, the Energy Initiative's been around for three years. It was started
by our, our President who came in just before that and this is actually one of her major
initiatives. She's really working to sort of harness the enthusiasm and direct the Institute
towards energy. And out of which was formed the MIT Energy Initiative. It has four main
parts: research, education, campus energy and outreach.
Research sort of spans quite a gamut of research and it's really driven by what the faculty's
interests are. There's quite a bit, I mean 85 percent of our primary energy comes from
fossil sources, so a lot of it just thinks about how to consume, produce and consume
fossil energy more efficiently both supply side and demand side.
A lot of it and probably the majority of the research that's going on is sort of looking
towards renewable energy and storage options sort of transforming the energy systems of
tomorrow. A lot of also just sort of basic undercutting science that sort of cross cuts
a lot of technologies systems analysis, so on and so forth.
Education, there is a, there's definitely a dearth of talent to be able to go into the
energy workforce, develop technologies, implement new business plans and that sort of thing.
And so there's a major effort underway to sort of educate, sort of develop curriculum
and programs towards educating students to go off and, and work in the energy field.
Campus energy. If we're gonna get serious about doing work in energy we should look
at our own carbon footprint,our own energy use and see what we can do about it. And we
have a rather old campus that affords us many opportunities to think about how one retrofits
existing building stock.
And then finally outreach. There's, there's definitely, I mean there, there's an elevated
level of public discourse on energy at the national scene. A university can have a really
good role in terms of trying to be an honest broker about sort of the in, inner section
of technology, policy, economics, regulation and try and sort of help to sort of guide
policy and, and, and decision making and so forth. And so there's been a lot of outreach
work really focused at kind of a national level which I'll highlight.
And so what I'll do is I'll just kind of give a quick example of, of something that's going
on in each one of these areas. The one that I'll do the least is research simply 'cause
Tonio's gonna give you some really directed comments on research so, so all I just wanna
do here is say that solar energy is our, our most sort of robust program [ ]. There's over
40 faculty that are working in this area.
The only thing that I kinda wanna get across with this slide is just simply that we're
tryin' to sort of work with all, all sorts of different key players. We think this is
really important. Working with industry both big like Eni, but also small, a lot of startups.
Building philanthropically supported programs, building government supported programs, working
with other non-profits like Fraunhofer and Masdar; they all have, they all have really
important roles to play and we, and we think it's really important to sort of deal with
a team of all of them.
Education. So, so at a lot of universities you take a very deep dive into your program
of chemical engineering, double e, whatever it is that you're doing, but if you wanna
go on and work in energy oftentimes a lot of context is needed.
So for example, let's say a student from Tonio's lab is, is sort of studying some defect engineering,
but they're interested in maybe commercializing something that they're developing. Well they,
they really need to know some things about business, they need to know some things about
sort of economics, policy, basic science, engineering, so on so forth.
And so the idea is, is that undergraduates now have the opportunity to take this minor
which is the first kind of cross cutting institute minor that we actually have at MIT. What the
idea is that they complement their sort of core course of study with other areas like
energy science, social science and sort of technology and engineering. The idea is to
sort of prepare more well-rounded students when they, when they wanna then exit the Institute
and go off into the workforce.
And actually a really important part of this, that as well is that it includes sort of research
and very hands-on projects seen at the undergraduate level.
Campus energy. So, we have a lot of really old buildings that don't have sort of the
latest sensors and so forth and so here's a really good example of actually a, a student
project aimed at reducing electricity consumption for lighting purposes.
The, the building on the, the upper panel is the Frank Gehry Stata Center; it's, it's
the, the e, electrical engineering and computer science building. And all around it are these
older buildings and so they've set up cameras in the windows and this is the, the lower
left panel is the camera looking at one of these buildings and it's, there's an algorithm
such that it takes a picture every 20 minutes; it can, it's figured out sort of which lights
are on and which are off and then sends an automated email to the, the occupants of that
room if they've left their lights on. Just sort of give them a sort of a nudge, sort
of a behavioral prompt so to say, which is kind of a nice work around when you have an
old building stock.
Something, something really similar: chemistry fume hoods. If you leave like a modest sized
fume hood open, it's about the same power consumption as a mod, as a house. And so it's,
it's very easy to leave these up when you're done with your experiments or when you're
in the middle of it and so similar sort of thing; putting in sensors to automatically
sort of prompt the users to close the fume hoods when they're done. And the Department
of Chemistry expects to save $100,000 annually just, just from implementing these sensors
in the fume hoods.
So these are a couple things that we're tryin' to, to do on campus.
Actually something that may be of interest to the, to the, to you guys is there's a lot
of interesting research that's going on too in terms of things like trying to do remote
sensing to detect opportunities for energy efficiency improvement so, some thermal imaging
sort of a spectrum from satellite or land-based systems.
Actually it's a Professor Sanjay Sarma, someone who works with Tonio Buonassisi; so if anyone's
interested in going into that in more detail, I'm happy to chat about it afterwards.
And then finally, I'll, I'll just mention a little bit about outreach. So like I said
we've tried to take on these series of studies that are really focused on, on doing sort
of honest assessments of different parts of the energy sector.
Here's a few of our studies that we've done on the left; we've done it on coal, nuclear
power, a DOE commissioned one on geothermal, another one on sort of looking forward and
the trans, the light duty vehicle transportation fleet in the U.S.
All, actually the, the three, the three on the kind of upper left have actually all resulted
directly in legislation at the national level.
For coal in the case of policies to encourage demonstration scale facilities for carbon
sequestration because that's kind of one of the impedances to understanding if, if that
is gonna be a viable technology to deploy at scale.
The geothermal report was kind of one of a number of pieces that sort of helped prompt
the DOE to resurrect their, their geothermals program which had previously been zeroed out.
The three that we have ongoing right now are solar energy, natural gas and the future of
the electric grid. You can actually bother Tonio about it afterwards; he's one of the
authors on the solar study.
For natural gas it's about halfway through the study and they're already, they're coming
up with some really interesting results. They've, they've done a really in depth sort of supply
assessment of shale gas and, and they've, they've really helped to show just the, the
absolute abundance of shale gas that we have in this country. I mean sort of available
at the, the five to seven dollars in MMBtu level now, but, but that, that price is going
down and so that really chances sort of the energy dynamic in the U.S. And so these are
all available online; you can read the reports; you can see the studies. I can give you the
links to 'em. They might be of interest.
And so with that, what I'll do is I'll just describe a little bit about the program that
I'm working on. And so the, there's kind of two pieces to it.
The first is the Solar Revolution Project. And so here what it is is, is we find this,
this case where oftentimes you have a professor that has a really good idea and it doesn't
really fit in a good box. So, for example, it may be one of the specific programs that
the DOE or, or, or companies aren't sort of ready for, and so oftentimes what you need
to be able to move the, the technology development along is, is just a little bit of a kick start:
fellowships, a little bit of support.
[pause]
Go manual here.
And so what we've tried to do is pair a, a faculty with really novel ideas with foundations
and donors so Doug being one of them. Chesonis family being another one to try and move an
idea to the point where you can get it where you have a proof of concept; you can have
a paper, a patent, something where you can take and you can really demonstrate you're,
you're on to something such that, such that you can apply for a relatively large research
program with the DOE or, or engage industry in sort of a fruitful collaboration; and really
build large, robust research programs.
And so an, an example of, so this isn't just some, some random picture off the Web of,
of some dapper looking gentleman. It's actually Don Sadoway who's a materials professor at
MIT and he does actually dress like that every day. He, he's kinda to the nines with his
suits.
So he, he had this really interesting idea where there's this kind of challenge in large
scale storage where, what a lot of people have been thinking about in terms of batteries
is tryin' to teach, or trying to get a device that you know how, that can storage hard and
try to teach it how to handle a high current which is you need some, something you need
for a grid scale application.
And it's been really hard to sort of scale up something like a lithium ion battery. And
so he's kind of turned it on his head where he had this idea, "Well, let's take a high
current device and try and see if we can't teach it how to store a charge." And so he
comes from a metallurgical background and so he's very familiar with aluminum smelters,
which is like, there's no better example of a high current device.
And so what he did was kind of turn an aluminum smelter into a battery by, by using a liquid
anode, a liquid electrolyte and a liquid cathode that are all density separated and through,
through just basically a fellowship supported by the Chesonis Foundation was able to, you,
you can't see this too well, but this is actually sort of maybe about five centimeters in the
vertical, was able to sort of demonstrate this liquid metal battery concept at the crucible
scale.
And so you can actually see the electrodes and the electrolyte here and, and this sort
of proof of concept just in terms of thinking about like the, the materials that were used
and their abundance in the earth, how much they cost, the numb, the cycle performance
variable to get kind of at this scale, so on so forth. They think they're on to something
and they think they can do it at a much larger scale.
And because they were able to kind of do some of these preliminary explorations with through
the help of the Foundation, they're now able to, over the course of basically a year and
a half, leverage that into an 11 million dollar research program, both with Totall for home
scale applications and ARPA-e which is a new DOE program for much larger grid scale applications.
So we really think there's something to this model. If, if we can kind of help these projects
along at just the earliest stages, that we can really get them ready to sort of really
launch off with government and industry.
So the program that I'm running is, the idea is to take this first to continue to grow
it within the solar and storage communities at MIT; that's, that's where we have kind
of some of our, our, our biggest areas of, of faculty interest and there's, there's so
many ideas there that, that still need sort of taking off, but then also really to leverage
it and to, to take this model and apply it to other renewables areas and as well as enabling
technologies. So things like grid, materials, so on so forth.
And so here, here's an example of a potential project which is perfect since Google has
a, a sort of a keen interest in engineered geothermal systems. So Paul Woskov on the
left, he's a, he's a, he does fusion power research. He works on a levitated dipole experiment
at MIT.
To do these, to do this work you need a millimeter wave length source at, at low power ratings
to diagnose the plasma and, and get some characteristics about it and a high wave length source to
actually inject, be able to inject energy into, into the plasma.
And so what his idea is, is to take this millimeter wave length source and use it for drilling.
So I don't know if any of you guys have, have thought about laser drilling for, for engineered
geothermal systems, being able to use it for, to drill hard rock.
They, they think that it had a lot of advantages over laser drilling because the millimeter
wave length sources are much more efficient; you have less, much less scattering. And so
they're kind of at the point now where they, they have the source and they just wanna build
a wave guide to be able to sort of vol, try and volatize some rocks.
And if that works, then they think they can sort of take that to the DOE. So crazy idea,
but also a really transformative one. And so this is the sort of thing where if we can
just bring it to the point where sort of if, if something is there we can sort of take
it to the next level.
So that's the program that I'm involved in and I'll turn it over to Tonio now. And he's,
so he's gonna give kind of launch into to, to some of his solar photovoltaics research
and gonna kind of take a much deeper dive. So, Tonio.
[pause]
Tonio Buonassisi: Thank you, Daniel. And thank you ladies and gentlemen for coming today.
It's an honor, but foremost a pleasure for me to be here today with you, especially given
Google's commitment to renewable energies and, and sustainability.
So today I'll be chatting a bit about our program at MIT focused on photovoltaics and
in particular the unique niche that MIT and universities can play in, in, in photovoltaics
and the development thereof; focused on the scaling and, and cost challenges of PV.
So I, I'll start with a basic slide that presents to you my motivation; my personal motivation
on a, on a month to month, year to year basis for working in PV. Obviously on a, on a day
to day, week to week I'm, I love being in the lab; I, I like working with students and
I like teaching and doing some fun research. So that's what, what keeps me going in the
short term.
But longer term when I look in the mirror and figure, "Where am I going? What am I doing
with my life? And, and what am I dedicating my, my time and energy to?" which is really
the investment, biggest investment any of us can make, it's about solar, precisely because
of the large resource bases available.
So on the left hand side here we have the resource, the solar resources provided by
the sun around 105 terawatts. And on the right hand side that tiny little blue cube represents
the human energy use, estimated in the mid to late century.
And even if we consider the losses that occur in atmospheric absorption, what's reaching
the earth's surface is still several orders of magnitude larger than our energy use. Hence,
if we can capture even a small percent of that total, we'll be in pretty good shape.
So if I look at solar energy today and ask, "What is the status quo of the industry?"
We have a few pictures here of different energy, solar energy installations.
On the upper right there's a, a winery in, in Napa Valley. On the lower right there's
a house in, in New York State. And on the lower left we have a picture of a large field
installation. The snaky white feature on the lower right part of that image is a road leading
up to a large field installation of solar panels in the middle of the image and the
little green specks are trees. So you can get a sense of the scale there. It's a, an
installation of several tens of megawatts.
So solar today is a professional industry. It's on the order of 50 billion dollar in
terms of, of, of total movement. If you go to any of the large trade shows, you'll see
it's a very professional environment akin to any of the other large even software or,
or hardware development in the, in the IC and computer industries.
It's competitive with bulk power in some markets and I stress the point some in this case,
due to the fact that still today the cost is, is high relative to conventional power
sources.
And as a consequence total grid penetration, meaning the total amount of solar that's on
the grid competing with fossil fuels comprises less than one percent of total electricity
generation. Meaning there's a huge upside for this technology, if we can get the cost
down and allow it to scale.
In terms of technologies this is, this chart here represents the total growth of the solar
industry broken down by technology. So in the vertical axis we have here starting from
1980 and going up to a couple years back, and in the horizontal axis abscissa we have
the total capacity, the total annual production of PV. And since we're on an exponential growth
curve here, this is growing quite fast and already out of date.
But as we see the breakdown of the technology this is still roughly representative of what
we have today.
So back in 1985 there was about a third, third, third split between monocrystal and silicon,
which is representative of the integrated circuit industry; these circular wafers grown
by the Sharkovsky growth method achieving solar cell efficiencies in the range of 15
to 23 percent.
We have about 50 percent of the market or, today back in 1985 about a third of the market
comprised of a large grained polycrystalline variety of silicon known as multicrystalline
silicon and that usually achieves efficiencies on the order of 14 to 16.5 percent due to
the higher defect density that inhibits solar cell performance.
And we also have a small but growing percentage of thin films today that achieve efficiencies
usually somewhere in the range of six to ten percent. Now there are some newer technologies
that might be even pushing 13.
So if you look back to 1985 there was about a third, third, third split between these
three groups of technologies. Today the market is largely dominated by crystalline silicon
because this particular technology was able to scale faster and because of the higher
efficiencies there were less costs downstream.
For instance, if you have a higher efficiency cell, you need less encapsulant materials,
less labor to install the same amount of power, less racking materials and so forth.
So because of this efficiency lever, these particular types of technologies were cheaper
and able to scale faster. And we have the situation where we're at today: a large percentage
of the market dominated by crystalline silicon technologies; a small but growing percentage
of thin film technologies.
And in today's talk, we'll be talking a bit about both technologies. Both the crystalline
silicon and thin films. The incumbent and the emerging technologies.
So I told you a little bit about the status quo; where we're at today, but what I really
want to talk to you about is the vision for PV; where I think we can go from here.
And I think, I'm fairly well convinced based on, on the calculations that we've done and
also looking at other technologies that have gone through similar growth curves, that a
significant fraction of the, the world's energy portfolio can derive from PV within a few
decades. And the advantage obviously is that it's a clean renewable energy source; can
be produced and manufactured and used locally.
I emphasize here the word "portfolio," energy portfolio. I'm not exactly a nut who thinks
that all power will be coming from solar and we should turn off all of our other power
plants immediately. No, I think that solar will play a role; it will play a part in the
energy portfolio that will be comprised of other sources as well, but I hope for the
future of our planet that they, these will be renewable and, and, and clean in the sense
of non-CO2 emitting energy sources.
I also envision a future in which solar panels are so cheap to manufacture that they can
be produced locally. Say for example glass; glass is manufactured locally typically and
used locally because it's so heavy and, and cheap to manufacture, it would be impractical
to manufacture overseas and, and ship around.
So we can envision a future in which solar panels reach the same level if we can reduce
the cost of manufacturing.
There's another element to the vision here which relates to the map below and you can
see the solar distribution around the world and the sunlight heavily concentrated in the
so-called developing regions of the world.
And on the next slide here we have a, a cute little chart that I, I came up with some late
night thinking about solar, where we have the human development index, the HDI, which
is a United Nations index of human development versus insulation, the total amount of sunlight
that a particular region on average receives on a daily basis.
And you can see that those regions that are developing, that need energy the most, are
also the regions that fortuitously have the highest amount of sunlight available to them.
So I'm quite hopeful that PV can be developed an enabler and a conflict reduction vector
in, in the process of, of human evolution.
So this is the grand vision for solar and for PV in particular; the conversion of sunlight
into electricity.
Now regarding getting there; how do we actually move from where we are today to where we need
to get? How do we make this vision a reality?
Well we're here, we're at a relatively small fraction of the world's total energy mix and
if we maintain a steady clip, a steady growth rate, we should be reaching some meaningful
percentage of world electricity generation, and I emphasize electricity generation; it's
not fossil fuels that are dedicated to the transportation sector here as well. This is,
this is solely electricity. I, I do think that within a few decades if we keep it up
we can get there.
But there are two obvious challenges to, to scaling. The first which most people are familiar
with is the cost issue. We know that in, in most markets PV is not cost competitive. That
if you take power directly out of the wall, power being produced by a nearby natural gas
fired power plant, coal power plant, nuclear or hydroelectric, that PV will, will usually
not be competitive in most markets, with certain exceptions. California, tier 4 and 5; Hawaii
and so forth. But as the costs continue to come down, you'll hit grid parity in other
markets and the growth should accelerate.
The other element that most people are not necessarily aware of is the sheer scale of
manufacturing. Now since manufacturing went overseas in the United States, to, to East
Asia and, and other locations around the world where the labor costs and materials costs
are lower, we've somewhat forgotten this element, but it's very, very important that we keep
it in mind when we think about producing terawatts of power.
If we take the current manufacturing facilities for the dominant technologies out there in
the market today, crystalline silicon; right here what you see is, is a CAD rendering of
a one gigawatt fab for producing solar modules using crystalline silicon technology.
This is REC, it's a company based in Norway; it's REC's plant that is located in Singapore.
Now if we take that factory and scale it up a thousand fold to get to the terawatts level,
what we really need to be aiming for if we're going to be producing meaningful percentages
of the total world power, we're looking at land areas just for the factory alone, on
the order of the size of the State of Rhode Island. These are the left two bars in this
figure right here for crystalline silicon and thin films actual.
If we envision another reality where we have a much faster manufacturing process, for instance,
the Pilkington float glass process which is used to produce most of our glass; if we envision
that instead producing solar cells we would need much less land area because of the higher
throughput of that process.
And lastly on the far right portion of this plot, we have the high speed printer, the
LaserJet that's firing out 55 pages a minute; and if each of those were 15 percent solar
celled, we would need an area the equivalent to the size of five football fields; which
at that point everybody could do. Any country could have their own little football field
dedicated to solar cell manufacturing and off we go.
So it's not only a, a materials challenge, but also a manufacturing challenge. And I
wanna highlight these two points as we move forward.
Oh, one more thing.
If, even if we use a very, very thin layer of solar cell material, if we're going be
producing this on, on the terawatts level, we're producing tens of thousands of square
kilometers of this material per year just to replace the modules that are going out
of, out of, out of service and that equates to a very large mass of material; something
on the order of the Empire State Building per year in terms of materials processed to
keep these, these very thin layers of solar cell materials active in the field.
So that limits us in terms of what elements we can actually use on the periodic table.
There are many that are too rare to serve for large scale solar production.
Okay, so we've talked a little bit about the, the motivation for solar; the solar status
quo; the vision of what we can achieve in the future; some of the big challenges; and
now I'd like to talk a little bit about what we're doing at MIT to address these challenges.
And back in 2007 when I first landed there at MIT, started as a junior professor, the
vision was to produce a, or to create a laboratory that was 100 percent dedicated to solar. And
I think we've, we've done a, a decent job at growing up.
Now in September of '09 we have a group of around 15 people and a nice balance of, of
folks from around the world and different backgrounds and, and life experiences. But
100 percent dedicated to PV solar research R & D.
And we are use-inspired, which means, well this is a kind of funny term "use-inspired"
why, why do we, what, what does that mean? That's actually a, a code word for, for government
funding agencies.
You see if you go to the certain agencies within the Department of Energy and say the
word "basic research" you get kicked out; they only want applied research. And if you
turn it on its head and go to the National Science Foundation and, and utter the word
or the phrase "applied research" again they'll kick you out into the street and say, "We
only do basic research here."
So use-inspired is a way for us to say, "Look, we do, we need both. And we need basic research
because we need to understand some of the fundamentals and, and how to make breakthroughs
in areas that haven't been, haven't been made yet. But we also need the applied side, we
need to understand how we take these, these fundamental breakthroughs and introduce them
into real life systems to make a difference. If it just says in a thesis on a shelf, it's
not gonna do much good for the planet and it's not really gonna justify our existence."
So we start from a cost model in general; if we're looking at a cost reduction. And
efficiency here is a huge lever on your cost. There are also other things that are completely
outside of our control like the interest rate on the loan you get to build your solar facility.
That has about the same influence as efficiency. So there's both business and, and technology
related challenges here.
And we also emphasize medium term real world impact. Now it would be so much easier for
myself as a professor just to say, "Oh, I'm gonna be working on some really far out project
and in, in 40 years when I retire the world can judge me." [laughs] Instead I put my reputation
on the line on a regular basis and say, "Okay within five years we hope to have this accomplished;
within two years we hope to have this accomplished." The reason I do that is, is because there
is an urgency to this problem of, of climate change and we have to get moving quickly.
There's another element to this as well. We wanna help the students who will be graduating
in two to five years. If we set them loose on a project that's of fundamental importance
today and that will be bearing fruit within two to five years, then when they go out into
industry they're, they're sought after commodity and they're much more likely to get a position.
So let me highlight for you a few of the examples of, of the instances in my own career that
have led me along this path of, of medium term impact.
As a graduate student in, in Berkeley we, we developed some technologies that spun off
into a company called Calisolar which is located just up the street here, using dirtier forms
of feed stock material to make high efficiency solar cells. And we also have developed some
newer technologies more recently. This is a map of current collection efficiency that
are improving bad regions of solar cell materials that are now implemented in commercial production.
So it was kind of fun to visit Calisolar's factory early today and see some of the technologies
under development being actually used out there in industry.
So it was thanks to these advances that were, I think one of the rarer research groups that's
actually carbon positive right now. Our efficiency increases in industry have offset the insane
amount of traveling I do running around giving presentations and the like. So I'm, I'm motivated
to keep it that way and hopefully we'll, we'll continue making advances that, that continue
to improve solar.
So let me give you a few case studies. We'll, we'll get into the science and I would like
to make this as interesting as possible for you so I'll, if I happen to breeze over some
of the subjects and you really wanna talk about them we can go back to some of the,
the details in the Q & A.
Number one, we're gonna talk about multicrystalline silicon. Again, the technology that comprises
about 50 percent of the world's market today.
And the reason we're focusing on this is because many of the established companies are very
risk adverse. This is a commodity industry where you're producing a product that will
last for 20 years. It's like the building construction industry. So there's a very high
risk if you do something new and it doesn't work out and you have a massive module recall
it's gonna cost a company quite a bit.
So at, at the same time breakthroughs are sorely needed. There is a strong driving force
for innovation to drive the cost down and improve the scalability of these technologies.
And it's, it's motivating for the students I think to have this shorter time to market
impact in this particular field. So let me give you an example here of multicrystalline
silicon.
This on the right hand side is an image of a, of a, of a silicon nitrite coded multicrystalline
silicon wafer on the order of about five by five inches. So you can get a sense of the
beautiful crystal pattern.
This is a picture of the crystal growth area within the production facility for these multicrystalline
silicon materials. And you can see here on the, on the lower right portion of that image
a small keyboard and monitor for size comparison. These are rather large furnaces.
Each furnace produces ingots between one and four ingots somewhere in the range of 250
to 400 kilograms; today they're even going up to 600 kilograms. And one ton ingots are
in R & D stage. So these are massive blocks of silicon.
And out of those blocks, you can see the size here of the crucibles with one of our students
Sarah Bernardis as a comparison. The number after the Gen refers to the number of wafers
along the linear dimension. So Gen 5 would be 5 wafers along the linear dimension, 25
wafers total along each thin delta. A delta Z of your, of your crucible.
And out of these crucibles come material that looks something like this. These are columnar
growth structures starting from the bottom all the way to the top. And then you slice
your wafers out horizontally and your green boundaries are running perpendicular to their
surfaces and don't interfere much with charge transport.
So we have this sort of material. What are, how can we do better? Where are we at today,
and where are the opportunities for innovation?
Well, if we were to look inside of the material with a little bit of x-ray vision here and
to detect the, the defect clusters within the material, we would see these defective
zones going from the bottom of the ingot all the way through the top.
And these are zones that contain dislocations which are essentially one dimensional line
defects within your material on order of 104 and 108 dislocations per square centimeter.
Anybody who's worked with integrated circuits or, or, or, or lead emitting diodes for instance
knows these are very, very high values; very detrimental for electronic properties of a
semi-conductor material.
Now to illustrate that in a more sciency form, we have a performance in, in the y axis and
dislocation density in the x axis here as you increase your dislocation density your
performance comes down. We have some magic number here where we have better solar cells
with lower dislocation densities and worse solar cells above.
If you get down to this number 104 per square centimeter, this magic number, then solar
cell performance is not really influenced much because the probability that some electron
inside the material will encounter a defect is very, very low.
Now let's take a wafer out of a vertical slice shown here by this little box. And if we scan
that region using an electrical performance mapper, we realize that there are some really
bad regions caused by regions of high dislocation density. Likewise if we take a wafer in the
cross section as we're apt to do when we actually produce a solar cell, we find that the, the
wafer itself has a check board like pattern of these bad regions together with the good
ones.
And what that means from a practical point of view, is that if you have the good regions
and bad regions all connected together through a parallel circuit through the front surface
emitter and the back surface metallization, the bad regions will drain current from the
good ones and the overall performance of your device will be reduced.
So these dislocations are public enemy number one in terms of solar cell performance. And
a lot of work has gone into it in the past to try to figure out how to reduce their impact
on solar cell performance.
Usually after the material is already made, and while you're doing the cell processing
steps, the phosphorus diffusion, the anti-reflection coating deposition, trying all sorts of things.
Well as a starting professor at MIT, had a completely blank slate, we said, "Well why
don't we try something very, very different here? We know that dislocations are manipulated
in metallurgy in the steel and aluminum industries. So why don't we go ahead and try to apply
some of that learning to multicrystalline silicon and solar cell material?"
We know that at low temperatures dislocations really don't move anywhere. Silicon is a brittle
material like glass; if you push it, it'll deform elastically, meaning just bonds will
stretch, but won't, won't slip. And eventually if you keep pushing, the thing will just shatter.
And the same is true for, for silicon. So at room temperature.
As you start increasing the temperature a bit you can actually have the, the material
deform just like a paperclip would; a piece of metal; a ductile material. You've, you've
gone above the brittle to ductile transition temperature. This for silicon occurs around
500 degrees Celsius, but because of the unique crystal structure you're still limited to
how mobile these dislocations are.
You really need to go above a thousand degree C before you start getting dislocations moving
freely throughout the lattice and note that silicon melts at around 1400 degrees Celsius.
So these are very high temperatures and this little dislocation core structure denoted
here with this, this defect is free to move around the lattice.
So with this knowledge in our minds, we sought out some ceramic annealing furnaces and developed
a process at MIT where we can eliminate the dislocations inside of these multicrystalline
silicon materials.
So on the bottom part of this figure we have a control wafer with these little pits that
correspond to dislocations after the sample has been defect etched, and on the top we
have the annealed sample which is essentially taken from an adjacent part of material where
we demonstrate that the dislocations have been effectively removed.
And we believe that this process, if done correctly, could be done for about a cent
per wafer, a penny for wafer, which is really, really promising. If we can implement this
and, and, and manage to scale it up, we have a few more tests to run in the laboratory
- I'm really telling you some of the cutting edge research that we're doing in our lab
- but if we're successful in ramping it up, we could for a mere penny per wafer be increasing
the efficiencies of these solar cell devices anywhere between 10 and 40 percent relative
to where they are today; which is a big deal, and we're pretty excited about it.
So let me zoom through these slides since we'll be discussing more about that in the
Q & A.
But the take-aways from this first part here is that there's plenty, plenty, plenty of
room for innovation in existing commercial PV technologies. And anybody out there who
says, "Well, you know it makes no sense to dedicate any time to commercial technologies
because there are big industries thinking about it," I would say actually the big industries
are, are, are looking in the areas where they think they can have the biggest impact and
the lowest risk, the lowest risk opportunities.
If you come along with a great idea that the industry thinks, "That's too risky, maybe
somebody tried it back in the 1970's; it didn't work; don't waste your time." I would encourage
you to go ahead because maybe, just maybe you'll strike upon something really interesting.
You really have to take advantage of the years of insights gained in, in parallel fields
especially those fields that are going out of vogue, like metallurgy. You, it's not often
that you hear a new department from a new university forming the metallurgy department.
No, actually if you speed back 50 years when steel was the, the big material that everybody
was researching, you'll find that many of the materials science and engineering departments
were called Department of Mining and Metallurgy, or something equivalent thereof.
Nowadays it's, it's the Department of Nanotechnology, which is great. We should definitely be looking
into new materials, but we also shouldn't forget the wealth of knowledge that there
are in these fields that are now going out of vogue.
Just because the light is shining in that corner of the room when you've lost your keys
doesn't mean that your keys are over there; they might be somewhere else in the room and
you might have to create your own light to go looking for them.
The third part is that I just wanted to, to, to mention this point that the model for funding
research within universities is changing a bit. As Daniel alluded to before, the involvement
of, of individuals within university research, philanthropic individuals, is very important.
And it was actually Doug and Barbara Spring that allowed this project to get off the ground
when it was still an idea in our minds; me as a starting professor without many resources
to get started with. They were the ones who actually oiled the machine and allowed things
to get started and some first proofs of principal. Eventually now it's funded by a DOE project.
So that's an interesting part.
So we're gonna change gears and move from the existing crystalline silicon technologies
into some of the thin film technologies.
If I might remind you of that previous slide at the beginning of the talk, thin film technologies
today comprise a very small percentage of the total solar produced. And, however, there's
a large potential for thin film technologies for the following reasons: there's a scalability
argument to be made; these, these thin film materials use a hundred times less materials
than standard crystalline silicon technology.
To give you some concrete numbers, the absorber layers, the active layer materials, are typically
on the order of 300 nanometers to 3 microns, as opposed to a standard silicon wafer which
is on the order of 150 to 180 microns thick, which is about four times the width of your
human hair. So you can imagine how thin these, these thin film layers really are.
But because they are so-called direct band gap semiconductors, or take advantage of very
good light trapping, they're able to get away with much less material.
There's also the potential for very high-throughput manufacturing. With a thin film technology,
one could, in principle, produce solar cells are shown here on the right. This is a picture
from UNI-SOLAR a, a, an engineer here processing solar cell materials onto a roll of stainless
steel foil. And that's of course a, a, a big step forward in terms of manufacturability,
where we're producing solar cell materials like newspapers.
The challenges, however, are two-fold. To date, thin film materials haven't been able
to achieve very high efficiencies and thus haven't been competitive in the bulk market
when you have more efficient solar panels that require less labor and less materials
to install per, per unit power produced. And secondly, the conventional thin film materials
will ultimately be limited in how fast they, or how large, how much they can scale because
of the material's availability. Now we'll, we'll get to that in, in a few slides. I'll
come back to that point.
So the research goal within our laboratory is to design scalable earth-abundant thin
film materials, meaning materials that are, that have a large enough natural abundance
in the earth to scale to the terawatts level.
So the functional requirements if you think about this from a design engineering perspective,
we have to have an efficiency of about 15 percent or more; it has to be able to absorb
sunlight very efficiently; generate a high volume of charged carriers; and minimize the
carrier collection loss within the material.
And so there are some design parameters that relate to the material properties that we
have to design for and the constraints are essentially the ability to scale up to the
terawatts level and low cost, and of course durability.
So let's focus on this scaling up to the ten terawatts level. I'll show you some numbers.
If we look at the Periodic Table and we assess which elements out there are abundant enough
to be deployed in the tens of terawatts scale, we're looking at about 50 of them. And out
of those elements there's a subset of the elements that are produced in a high enough
volume of manufacturing today, there's about 30, so these are materials today that are
manufactured at scale.
And if we take those 30 elements and begin combining them to form semiconductor compounds,
we have a number of, of possibilities to choose from. And by screening both experimentally
and theoretically using some very sophisticated computer modeling by a colleague of mine,
Gerb Ceder, in the Material Science and Engineering Department, were able to hone in on around
ten compounds that really shows some good promise for making thin film earth-abundant
solar cell materials.
However, if we start comparing the real efficiencies, the record efficiencies of these materials
versus the theoretical maximum, shown here on this plot is the theoretical maximum efficiency
as a function of band gap, and we compare the actual record solar cell efficiencies
for several compounds including iron sulfide, tin sulfide, copper zinc tin sulfide, cuprous
oxide and, and tungsten sulfide, we see that the record efficiencies of these earth-abundant
compounds are much, much, much lower than the theoretical maximum efficiency. And this
is after several years of work especially during the 1980's on, on these particular
compounds.
Compare that to crystalline silicon, the blue are the record cell and module efficiencies,
the red are the typical commercial cell and module efficiencies, head and shoulders above
the earth-abundant compounds.
So the question is since these materials underperform are they intrinsically limited by some fundamental
electronic property of the material or are they limited by defects the same way that
crystalline silicon is limited?
And of course when the word 'defect' comes along my ears perk up. I'm of course fascinated
by these defects. They're not only bad, but in many cases quite good. You can think of
examples in, in the real world around you where defects are, are actually helpful.
For example, anything from strengthening steel or aluminum, you add impurities to, to strengthen;
steel in particular. Stainless steel you add chromium to form a passivating layer of chromium
oxide that prevents oxygen from going in and corroding the steel.
Now you can, you can have examples of colored glass or, or gemstones that gain their color
by the presence of certain point defect impurities. So defects here are, are very valuable in
certain cases, but of course detrimental in others. And for solar cell materials as we
just saw in the example we have some detrimental effects.
So our first step is grain boundary engineering. We want to make sure that these materials
have large enough grains such that the grain boundaries do not inhibit carrier transport.
And let me just show you here the grain boundaries highlighted in the cross sectional scanning
electron microscope image of a thin film material. If this is our little electron at the bottom,
it has to make, be able to make its way out of the material without interacting with the
grain boundaries. So the grain size has to be on the order of five times the thickness
of your, of your material. If you have about a one micron thick solar cell material, you
need grains on the order of about five microns thick, and this is just a quick little simulation
to show you that point.
So we've been doing a lot of work on trying to increase the grain sizes of these thin
film materials. For example, this one a cuprous oxide earth-abundant solar cell material.
We have a layer somewhere in the order of a 100 nanometers thick and we're trying to
achieve grain sizes in the order of 500 nanometers or more.
You can see that we've managed to increase it quite a bit and just last week the student
on this project showed me a figure that I wasn't able to incorporate into this, which
basically had grains that were larger than the little box that I'm showing right here;
so really, really large grain material. And the resistivities of these materials were
such that they could be incorporated into devices; that's our next step that we're working
on.
And hopefully, hopefully with any luck, knock on wood, we'll be able to beat the efficiency
limit and have a record efficiency cell, if we can get above 2% efficiency. It's, it's
not a high bar to cross. So we're, we're working on this avidly at MIT and, and hopefully we'll
make some, some nice progress over the next year or so on this topic.
So take-aways. In this particular subset we have novel earth-abundant thin film materials
which we believe offer great potential for substituting out the commercial thin film
technologies on the order of five to ten years. And now this, this basically says, "Guys don't,
don't go out and sell all your stock in First Solar or give up on all the six startups in
the Bay area." No, no, no, no. There's, there's still several years that cadmium telluride
and copper uranium gallium celunite can continue producing before you start running up against
resource walls. But many people think by about 2020 we'll start hitting these, these walls
of, of material availability and we'll need new technologies to overcome that; the niche
that we're attempting to develop right now.
The second take-away is that current earth-abundant solar cell materials are underperforming.
We believe that it's in part due to defects; we some evidence for this already. And by
defect engineering the same types or general classes of techniques we've developed for
crystalline silicon applied to thin films, we believe we can overcome these efficiency
limits. So the next few years we'll be able to tell whether we're successful or not.
And in conclusion, so we can open it up for Q & A. The potential of the solar resource
is very vast. Terawatt scaling is possible if certain challenges due to scale and cost
can be overcome and most importantly at this very early stage of the solar market development
a very small innovation can go a long, long way. If you make an innovation today and the
market continues to grow up, you can envision a situation in which your small innovations
impact over tens of percents of total market share, if you develop a very successful product.
So with that I'd, I'd like to really conclude by acknowledging the team; it's not myself
doing this work, it's, it's an entire team of people comprised of, of students, undergraduate
and graduate, post-doctoral fellows and research scientists. And we have here the team up here
and our collaborators internationally and nationally, and our funding agencies of course.
With that I'd like to open it up for Q & A and perhaps turn it back to our MC or open
up for - we'll just open it up straight for Q & A; make it efficient that way.
So, question in the back. Yeah.
[pause]
[sound of typing on keyboard in background]
Tonio Buonassisi: So the question was are most materials, most earth-abundant materials
sulfides?
Yes, there are several earth-abundant materials that are sulfides. Each has some very interesting
material properties. They're very typically prevalently bonded materials or near covalently
bonded materials that allow for [inaudible] transport throughout the layers. You don't
have as [inaudible] localization as you might have in zinc oxide. So yes, they are, they
are quite popular.
[pause]
[sound of typing on keyboard in background]
unidentified male: Excuse me, sir on GVR, remote GVR can you hear us or?
male voice: Yes, sir.
[unintelligible voices in audience]
Would you like to stop the recording?
No, don't stop the recording we just have somebody typing on a keyboard and it's making
a little bit of –
[unintelligible talking in background]
Got it?
I think they – yep. Okay.
[unintelligible talking in background]
Tonio Buonassisi: Yeah, so the question is can, can sulfides react with, with the environment?
Most certainly. So this one of the disadvantages of sulfide materials. They are rather reactive
and as a result the encapsulation becomes very, very important. There is one type of
sulfide material in near commercial production today. They're just at the cusp of commercial
production. It's called CIGS. It means copper indium gallium disulfide. And this particular
type of material needs to be very well encapsulated to avoid any sort of degradation due to exposure
to the elements or ingress of, of, of gases and moisture through the encapsulants to the
active layer materials themselves. Because they're very, very thin layers, they're extremely
sensitive to, to this, to the ingress problems. Yeah.
[unintelligible speaking]
Question in the front. Yes.
male voice in audience: I can't remember the name of the elements but I've read about a
couple of elements that, that you adjust the ratios of to the band gap and you can make
multi-layer crystals that are very efficient. Do you have any comments about [inaudible]?
Tonio Buonassisi: Certainly. So the, the essence of, of, of tuning the band gap, so the band
gap basically means the, the, the energy at which the semiconductor begins to absorb light.
And anything with the higher energy than that semiconductor can absorb light and transform
that into electrical current. So the band gap is a very important material parameter.
You match that up with the solar spectrum and try to absorb as much, as large of a solar
spectrum as you can.
And the material will absorb light most efficiently at the band gap energy. So you can stack different
materials one top of one another that absorb most efficiently at different regions of the
solar spectrum. Thus achieving a very high efficiency solar cell.
So to make a long story short, yes, you can tune the band gap of materials by alloying
them; by say for example taking three elements and changing the ratio of the first and second
element so that you go from a low band gap material to a high band gap material and absorb
very efficiently the reds and the, and the blues of your solar spectrum.
While certainly there is, there's efforts in these directions, one usually does that
once one already has an established solar cell material. So the first step is to get
a solar cell material that works with some band gap. And then from there you can innovate
an, an, alloy and change the, the band gap property. So, so we're one step behind that
still in terms of the earth-abundant compounds. In terms of the non-earth-abundant compounds
they're already quite there. With the three five compounds certainly indium nitride, indium
gallium nitride and other compounds you could alloy in this way.
Yeah.
male in audience: So up there as one of the goals you had a dollar a watt module used
to be a dollar a watt[inaudible] the bar [inaudible]. First Solar is already there but they're not
economic. [inaudible]
How do you overcome that?
Tonio Buonassisi: Sir. So the question was the dollar a watt in the module level the,
the premise here is that First Solar produces solar cell modules that are a dollar a watt
in the module level, but because of the high installation costs these are low efficiency
panels. Because of the high installation costs required you, you wind up with a installed
system cost much higher than one dollar per watt and it's still not cost effective.
So as one of the requirements that I had earlier in the slides a, a base line efficiency of
15 percent was what we're targeting. And with those sorts of efficiencies if you have a
dollar per watt the installed costs become lower as a total percent. So there, there
is a, a, an element there embedded perhaps I didn't express it clearly enough. It, it,
a more precise way of saying that would be exactly as you recommend; the price at the
installed system level which does capture the efficiency trickledown effect.
Personally I think that a large fraction of the cost that can be squeezed out of solar
cell production is not only on the materials and, and the manufacturing side of things,
but also on the balance of system on the module level and on the installation side.
And we were touring some, some companies this morning that are doing innovations in this
area here in the Bay area in, in the installation and better framing materials and, and better
ways of installing them. Certainly there is a lot of, of leverage to be gained there.
If you look to Germany, the price of installation is about half of what it is in the United
States in large part due to more streamlined processes for installation including the bureaucracy
involved. So there, there are a number of different pieces that can be, you can, you
can squeeze costs out of. And ultimately those efficiencies will need to come to bear if
we are to produce solar at the terawatt scale.
So, yes, that would be my answer.
[pause]
Yeah.
male voice: [inaudible]
Tonio Buonassisi: Yeah. So what are the materials innovations in concentrating solar power,
that was the, in the CPV.
So CPV is a technology in which you utilize optics to concentrate sunlight into a very,
very small cell. And you can spend more money to produce that cell because you're taking
light from a larger area and concen, concentrating it down so in terms of total cost of production
the balance of system components, the optics and, and the structure around it ends up being
comparable, if not much greater than the actual cost of the cell itself.
So it's fascinating that, that, that type of technology because you can envision a scenario
in which maybe a better cell comes along in three to five years and you just swap it out
and you haven't changed the total price of, of your system much.
A lot of innovation goes around managing the heat load in these devices. And because you're
concentrating 500 suns, up to 500 suns, it can be less, but up to hundreds of suns onto
a very, very small area - when I say a sun that's equivalent of going out and standing
in midday sun - imagine 500 times that amount of heat; that's a lot of, of power to dissipate.
Any inefficiency in the solar cell device will usually generate heat and you have to
extract it somehow. So there's been a lot of work going into the materials used for
the packaging and heat extraction from such devices. And as well into the materials that
comprise the active layer solar cell itself.
So in those cases, goodness if you concentrate enough maybe, maybe it becomes impractical
at, at very high concentrations to the heat load, but if you concentrate on it enough
it might reduce some of your material constraints as well. You might be able to instead of 30
elements; you might have about 34 or 35 to deal with on the Periodic Table which could
open up some possibilities as well.
Yeah. Another question.
male voice: You had had a slide that you talked about [inaudible] 2020
[inaudible]
Tonio Buonassisi: Yeah.
male voice: What about the area from the solar cell itself?
Tonio Buonassisi: Yeah.
male voice: [inaudible]
Tonio Buonassisi: Yeah.
male voice: [inaudible]
Tonio Buonassisi: Certainly. So today the, some of the most cost effective installations
are large field installations largely because you, you don't have to worry about some of
the safety regulations of installing it on the rooftop of somebody's house; you can just
put up a big fence around the area; install a bunch of modules all at once and then only
pay for one architect and one design engineer's fees for a very, very large field installation
and negotiate down the price per module with the supplier because you're buying it bulk.
So these field installations are, are price competitive today, but I, I really do hope
and, and secondly expect that the residential market will comprise ultimately a large portion
of the, the demand because we pay a higher price for our electricity than commercial
suppliers do.
And secondly if we produce the energy locally we're using it locally as well and we don't
have to pay for the transmission distribution of that electricity. It's not a commonly known
fact but if you're buying coal power, the total price that you pay or the total cost
to get it to your home, half of that is the transmission and distribution cost; the other
half is the production.
So if you produce your power locally you're producing it on your roof and consuming it
locally, I think that's ultimately where the market has to head. Now the question is, is
there enough roof space in the United States to actually make that happen. Barely.
And I would add furthermore that we shouldn't bank on it. I think rooftops definitely should
all be covered or at least the practical ones should be covered with good solar panels.
But there are also other opportunities of dead spaces around the United States that
we can consider covering.
Let's take for a second the total road coverage in the United States. If you look at the total
network of highways and roads around the U.S. – you can download these number from the
government – and you look at an average width of a road, you multiply the two together
to get the total land area; you're looking at a fraction between one and two percent
of all of the United States covered in asphalt. It is an amazingly large number.
You look at other types of things like the, the percentage of Virginia covered in tobacco
farms. It's also on the order of percent [laughs].
So there examples of human-made structures, perhaps not the tobacco farms, but roads and,
and parking lots that could be considered as well for, for installation of, of solar
panels. And the, the parking lots being a great example. If you install them on, on
covered lots.
So I think we have enough dead spaces already; spaces that aren't being utilized that we
don't have to take new pristine land and, and produce solar panels there. That might
be a good way to get the industry jump-started near term, but longer term there are certainly
a lot of options for us with, the spaces we've already created.
[pause]
Yeah. Question.
male voice: So your notion of [inaudible]
Tonio Buonassisi: Yeah.
male voice: [inaudible]
Tonio Buonassisi: Absolutely. So the, the earth-abundant materials; the question was
it, it's not such a new idea; there have been people investigating this for several years.
Absolutely true.
So the history of earth-abundant solar cell materials I believe started way back in, in
the 1970's or perhaps early 1980's. CIGS – Copper, Indium, Gallium, Diselenide -- actually started
from investigations of copper sulfide and the realization that this material wasn't
stable. The copper tended to electromigrate so let's add some heavier elements in there
to, to stabilize this and get a, a firmer structure; and then from there they added
indium.
But any rate, the, the research that's been going on, let's see some of the big names
in the field [inaudible] in, in Texas has been investigating this [inaudible] for quite
a bit. Other groups Cyrus Wadia from Berkeley produced a, a great paper where he looked
at the economic aspects together with Dan Kammen and there's been a growing enthusiasm
for this idea in the field.
Now the, the ultimate Achilles heel for all of this is that we haven't been able to, to
really make one of these earth-abundant materials work on a large scale, precisely because of
the efficiency issue. There are different types of problems that affect each type of,
of material and we're just now beginning to get a sense or appreciation of the limiting
factors for each of these.
Let's take sulfides for instance. A lot of work sent into sulfide research in, in here
in California several decades ago and then the Hahn Meitner Institute which is now the
Helmholtz-Zentrum Berlin, in Berlin, Germany during the 1990's working on compounds like
copper sulfide and iron sulfide. And we, we just didn't get critical mass is my belief
and I, I, I do think that the, the key efficiency limiting defects for each of these different
materials are on the cusp of being identified. And if we can identify them and if we can
overcome them then these materials have a chance. But if we just go by brute force alone,
trying to deposit them and make solar cells out of them, we won't get anywhere.
[pause]
Good. Well, I'll turn it back to our MC. I, I do thank you for your questions. It's been
fun being here and thank you.
Doug Spreng: Thank you, Tonio.
[applause]
And thank you Daniel, as well.
Okay, so just about done. Just to let you know that we do have some handout material
over here which give you a lot more information on some of the other programs that are involved
in MITei and the Sustainable Energy Revolution Program, and some actual, interesting background
articles about some projects that are possible in the area of wind, geothermal and storage.
So you're welcome to have those.
Also, if you'd like to sign up to receive the MIT Energy Initiative Newsletter on a
regular basis to see what's going on at MIT, some of you have already signed up over here,
but others that haven't yet, you're, please feel free to do it.
And in the bold hope that any of you out there may be actually interested in funding some
of this research, just come and talk to either Richard over there or me after this is over.
And we'll be happy to follow up with you.
So that's it. Thanks very much for coming and thank you out there viewing audience.
I hope y'all enjoy seein' it on YouTube.
Thank you very much. Bye-Bye.
[applause]
[techno music]