Breakthrough allows inexpensive solar cells to be fabricated from any semiconductor


August 3, 2012

A new technique allows photovoltaic solar cells to be produced using any semiconductor (Photo: Shutterstock)

A new technique allows photovoltaic solar cells to be produced using any semiconductor (Photo: Shutterstock)

Image Gallery (3 images)

Despite their ability to generate clean, green electricity, solar panels aren't as commonplace as the could be. The main sticking point, of course, is price. Due to their need for relatively expensive semiconductor materials, conventional solar cells don't yet have a price-efficiency combination that can compete with other sources of electricity. Now Profs. Alex Zettl and Feng Wang of Lawrence Berkeley National Laboratory and the University of California at Berkeley have developed seriously unconventional solar cell technology that allows virtually any semiconductor material to be used to create photovoltaic cells.

A solar cell works according to these steps. First, sunlight hits the solar cell and is absorbed by the semiconductor of which the solar cell is made. In the absorption process, electrons are freed from their atoms, allowing them to flow through the semiconductor. The presence of a p-n junction acts as a diode, only allowing the electrons to move in a single direction. (Electrons and holes move in opposite directions, but the electrical current only moves in one.) Metal electrodes then transfer the light-generated electron flow into an electric circuit for use. A p-n junction is the interface between a region of the semiconductor where the dominant charge carriers are holes and a region where the carriers are electrons.

How a solar cell works (Image: Brian Dodson)

A conventional solar cell is made of a thin wafer of a semiconductor with a metallic electrode deposited on its rear side. The side facing the light source is polished more finely than any optical lens, cleaned to the atomic level, and then dopant atoms are deposited onto the front side, whereupon the entire wafer is placed in a high-temperature diffusion furnace.

The purpose of a dopant is to change the dominant charge carrier in the semiconductor from hole-rich to electron-rich, or v.v. In this process the p-n junction that converts incident light into a flow of electrons is formed. Following diffusion, the wafer is again cleaned, and a metallic electrode is grown on the front surface, using arcane rituals to ensure an ohmic contact with the active semiconductor material. (An ohmic contact is an electrical contact that obeys Ohm's law, having no rectifying or diode-like properties.)

The efficiency of conventional solar cells is also limited by the semiconducting materials which are suitable for the manufacture of solar cells by some approximation of the above process. It must be possible for the dominant charge carrier of the semiconductor to be changed between p (hole) and n (electron)-dominated conduction by introduction of chemical dopants, so that a well-behaved p-n junction is formed. It must also be possible to make a satisfactory electrical contact between the electrodes and the semiconductor.

There are many semiconductor materials with optical properties and electronic band-gaps well suited to conversion of light to electricity for which one or another of the rather rigid criteria for manufacture of conventional solar cells fails. These include many metal oxides, sulfides and phosphides, which are plentiful and inexpensive, but have been considered unsuitable because it is so difficult to alter their electronic structure chemically (e.g., through doping). For example, zinc oxide is a semiconductor well suited for capturing violet and near ultraviolet light, which is wasted by most conventional solar cells.

We now have Zettl and Wang's unconventional solar cells. Broken down, the expensive parts of making regular solar cells are the semiconductor wafer, forming a high-quality p-n junction under the surface of the wafer, and making ohmic electrical contact with the front and back of the wafer. Aside from providing a semiconductor wafer (usually of a cheaper material), the new solar cells require none of this.

The new technology is called "screening-engineered field-effect photovoltaics" (SFPV). An electrode is deposited on the front of the semiconductor wafer, which partially screens the semiconductor from an electric field generated between the front and rear electrodes. Assume the semiconductor is naturally p-type, so that it has an excess of holes. The applied electric field then penetrates the semiconductor surface slightly, attracting electrons toward the surface and repelling holes. As a result, the semiconductor near the surface changes from p-type to n-type (electron-rich), and a buried p-n junction has been generated – not by chemistry, but by use of carefully tailored electric fields. An extra bonus is that the front electrode automatically forms an ohmic contact with the semiconductor wafer.

“Our technology requires only electrode and gate deposition, without the need for high-temperature chemical doping, ion implantation, or other expensive or damaging processes,” said lead author of a paper describing the new technology, William Regan. “The key to our success is the minimal screening of the gate field which is achieved through geometric structuring of the top electrode. This makes it possible for electrical contact to and carrier modulation of the semiconductor to be performed simultaneously.”

Two approaches to engineering the screening properties of SFPV electrodes - fingered electrodes on the left, graphene electrodes on the right

Two electrode configurations that exhibit the SFPV concept have been developed. In one, the electrode in contact with the semiconductor wafer is formed of a row of narrow fingers, while in the other the partial screening is accomplished by placing a layer of graphene atop the semiconductor wafer. In both cases, high-quality p-n junctions are formed in semiconductors for which this structure was previously impossible.

Low cost, high efficiency solar cells? Sounds like a winner here – let's see when and if it hits the market.

The team's research is published in the journal Nano Letters.

Source: Lawrence Berkeley National Laboratory

About the Author
Brian Dodson From an early age Brian wanted to become a scientist. He did, earning a Ph.D. in physics and embarking on an R&D career which has recently broken the 40th anniversary. What he didn't expect was that along the way he would become a patent agent, a rocket scientist, a gourmet cook, a biotech entrepreneur, an opera tenor and a science writer. All articles by Brian Dodson

Excellent explanation, Brian! Good article! The only thing I would crit is the conclusion: you left many things implied which I think should be explicitly stated, such as the notion that this new technique can be applied to the wide variety of potential semi-conductor materials successfully. A more "robust" wrap-up to your very detailed (and informative) explanation would have been more appropriate. Very nice, tho - and exciting news for this industry! Now, could we expedite commercialization of this tech this time? :-)


Oh no. Yet another reason to wait before buying solar panels


graphene and low cost are not exactly on the same page. And for what we know now, it will be this way for many years to come.


"Low cost, high efficiency solar cells? Sounds like a winner here – let's see when and if it hits the market."

I read the article - twice. Where is the "high-efficiency"? What I read states the achievement is in allowing the use of less costly materials to achieve what expensive materials currently do.

That may be less costly, but it doesn't make it "high-efficiency". To date, solar technology is horribly inefficient. On average, solar operates at between 10-25% efficiency (converting light to useable energy). The MOST efficient (and most expensive) make it to 40% efficiency.

Gasoline, Coal, Hydro-electric - even relatively inefficient natural gas - are all roughly 3X more efficient than solar.

Don't get me wrong, I love the idea of solar and I support R&D to get it where it needs to be - and this breakthrough IS a breakthrough - but, "High-efficiency"? No.

Joseph Boe

Any technology that permitted energy to be captured from the far blue and UV spectrum is very desirable and worth further development. Just obtaining efficiency close to existing PV would net much more POWER since this region of light has more energy per photon.

This approach may well improve space based PV also.

Manufacturing and testing means at best the material is five to ten years away from the market.


If the source is free, the sun, efficiency is not as important as with high priced oil.


"Despite their ability to generate clean, green electricity, solar panels aren't as commonplace as the could be. The main sticking point, of course, is price. "

Don't get me wrong, I love hearing about new technology which might just breakthrough. However, in the world of solar based electricity the issue is not the price of the solar cells. The issue is the price of acceptable storage. Generating power at noon which is really needed after dark is the issue I believe. It's certainly why I've never invested in solar power.


"Generating power at noon which is really needed after dark is the issue I believe."

Tell that to the people who are suffering through rolling blackouts during the Summer when it is 106 degrees in the afternoon. Everyone is blasting their air-conditioners on a day like that, and it is a major strain on the Grid. More power at noon is precisely, exactly what we need.

Timothy Neill

@ Joseph Boe "Gasoline, Coal, Hydro-electric - even relatively inefficient natural gas - are all roughly 3X more efficient than solar."

First you question the numbers in the article, and than to counter argue... you pull "3x" out of your ass showing you have no concept of the laws of thermodynamics.

Efficiency in photovoltaics refers to their proficiency at turning photon energy into electrical energy, NOT cost relative to something else!!!!!!!!!!!!!!!!!!!

A one meter square solar cell can only convert about 20% of the total energy output of a ONE SQUARE METER "cut" of the light hitting the earth. On the other hand to get hydrocarbon energy out of one square meter of light, first it has to shine on vegetation, which converts it to carbohydrates, which then goes trough the long process of sequestration over millions of years, before it can turn into hydrocarbons... so exactly how efficient is this process??????

All energy comes from the sun and when measured side by side, square meter of light against square meter of light, hydrocarbons are least efficient of all. We have just been lucky over the last hundred years because those hydrocarbons have already been "pre-processed" by nature for us, but it took millions of years, and millions of hours, for plants to turn light into carbohydrates, get eaten by animals, bacteria, and then decay into hydrocarbons. Using solar directly bypasses this million year process and gives us energy directly from the sun for as long as the sun shines, without relying on evolution and its million year process. Plus hydrocarbons will run out sooner than later(since evolution takes millions of years to do its thing) and then we won't have such easy "pre-made" energy on demand as we had in our lifetimes. Please stop drinking the corporate cool-aid. How many Jonestowns do we need before people wake up to reality!!!

AnOld BlackMarble

@jocco, I disagree that efficiency is not as important as oil. It may not be important to you as you may be sitting on 20 acres of land in Arizona. But it's very important if you consider how much sun is achievable in the target area as well as its size in oirder to generate the energy that is needed for one's purpose. If I need 10KWh per day of electricity to run my house, I don't want to have to cover my roof AND my yard with solar panels to get it. Efficiency is ALWAYS important.

Cheers KT

Knowledge Thirsty

I actually agree with all the comments so far. But you could argue that these new style solar panels are efficient in the fact they will take less energy and effort to produce..... Here in Connecticut we have Net Metering. There is no storage of solar energy (unless you want to get completely off the grid). Most people are connected to the grid and send their excess solar energy out to the grid and receive credit. If the prices of solar panels came down enough I would throw a bunch of them up on my property. Putting them up on your roof is expensive and complicated your home. Overall I look forward to seeing this technology make solar panels cheap and common. This may be the tipping point we have been waiting for.

"... a metallic electrode is grown on the front surface, using arcane rituals to ensure an ohmic contact with the active semiconductor material."

At last, the proof I have been seeking that the fabrication of today's electronics requires witchcraft.

There is no other possible explanation for the fact that, for example, an average single E. coli bacterium dropped onto a contemporary DRAM chip would cover up several hundred memory cells.

Obviously no Earthly power could achieve that.


re; AnOld BlackMarble

Comparing using fossil fuel to the Jonestowns murders is over the top. Coal is just wood that has had a mountain dropped on it and a lot of nasty minerals washed into it. Oil and natural gas is just biomass that has been pressure cooked (It has been done in the laboratory but is not currently economically viable. Google synfuel. It does not need millions of years. ....................................................................................................

re; ralph.dratman arcane [ɑːˈkeɪn] adj requiring secret knowledge to be understood; mysterious; esoteric [from Latin arcānus secret, hidden, from arcēre to shut up, keep safe] arcanely adv arcaneness n rit·u·al 5. a. A detailed method of procedure faithfully or regularly followed: My household chores have become a morning ritual. b. A state or condition characterized by the presence of established procedure or routine: "Prison was a ritualreenacted daily, year in, year out. Prisoners came and went; generations came and went; and yet the ritual endured" (William H. Hallahan).

Therefor arcane rituals = secret techniques. However I don't under stand electrospeak either. .........................................................................................................

re; Timothy Neill

Yes there are times when more power is needed at noon and solar does that but sometimes more power is needed at night as well Solar does not do that without storage and if you are using solar as a primary generation source you either need a lot of storage or duplicated generating capacity.


please just let me know when i can ACTUALLY find an inexpensive solar panel. every other day with these articles. i want to install solar SO BAD but am SO POOR right now.


As another guy already said, graphene is not cheap to mass produce.

Fretting Freddy the Ferret pressing the Fret

@Knowledge Thirsty

I take it you have a coal mine and coal-fired power plant on your property then? No? I guess that's not very efficient either, then.

My roof is currently 0% solar efficient because solar panels are too expensive to install. But if this or some other technology makes solar panels cheap enough to put even 6 or 8% efficient panels on my roof to reduce my grid-draw by a third or a half, then I'm miles ahead.

And while I don't have 20 acres to install solar panels on, you can bet there's plenty of desert in the American southwest and virtually endless parking lots and highway all over that we could cover with PV if they got cheap enough.


I rarely post on this website, but two things stuck me, or I should say two post. OldBlackmarble. His post is not logical. Preprocessed hydrocarbons ie oil and gas has served us well, and there is plenty still around, would I like to get around the expensive stuff, yes, but solar is not there yet and is way to costly. JACCO is correct in what he is saying. During the normal times I need 10kw to run my home, I don't need it during the day, when there is sun light. Now I might need more if it is extra hot, but for the most part I need light at night. STORAGE is the problem of solar power. If I have an efficient storage capacity then I can store energy to cool me if it is hot or heat my home if it is too cold. So it is a three prong problem with solar energy. 1. Cost of material in by money and space. 2. Efficiency. You need a lot of panels to provide what you need. 3. Storage. Battery tech. is not sufficiently mature to store the energy I need at a reasonable cost. One other thing to: Timothy Neil... even all the people suffering from the 106 degree had solar power, THEY are still on the grid... grid fail... air conditioning stops.

S Michael

I'm in the process of building some conventional solar panels. Every time I see an article like this, I contemplate that perhaps I should postpone my project because it appears that newer, more efficient solar technologies are on the horizon. Then I realize that I have been reading Gizmag stories on the "newest discovery that will lead to more efficient solar panels" for YEARS, and none of these discoveries have yet to come to mass market as a significantly more efficient, yet affordable solution. Sigh....

Back to soldering together my solar panels.

Laura Ward

Even if the materials became cheap enough solar would unlikely ever become the ONLY source of grid electricity, things like hydro and nuclear electricity would still be produced at night time when solar doesn't work.

As for having to fill your rooftop and yard just to get 10kwh for your home in a day, your way off, a 10kw system would probably fill up half your roof, or likely just the half thats facing south. And that would produce 10kwh in 1 hour of direct sunlight and possible up to 160 kwh in a day depending on your location in the world.

Generally homeowners get 1-4kw systems on their roof which isn't all that big.

As for battery storage, yes this is a problem and battery tech isn't that great that its terrible cheap but I'd say its not all that expensive when compared to the price of the panels and i disagree that you couldn't buy the batteries at a "reasonable" cost.

Anyways as for the article, this is great news and its good more breakthroughs are happening in this field of technology, i have no doubt that solar will be one of the main sources of power in the future, however like a few other people have mentioned, there have been major breakthroughs in solar tech now going on for like the last 5 years and none of that is on the market yet either, the stuff we're using is at least 10-15 years old even the thin-film stuff they layer over windows. So we will not be seeing this on the market for awhile but it WILL eventually get to us.

Just my 2 cents.


Don't know where these guys were searching, but PVs (solar panels) are dirt cheap nowadays. Even here in Oz they are approaching $1 per watt and in the US they are well below that. Check out mobs like and you will see what I mean.

There seems to be considerable ignorance here about how grid-connected PVs work. With grid connect systems there is no battery, the grid is your battery. You generate excess during the day and get paid for it at whatever tariff your energy company will give you (some are good, some are crap) and then draw from the grid at other times. No battery, just your PV array and inverter.

As an example, you can buy a 3kW array and inverter here in Oz for around $5k, not including installation (see as an example). A system this size will average 10-15kWh a day, depending on where you are. At current elec prices here (approaching 30c per kWh) a system like this will pay itself off in a few years.

Given that US prices are similar, then you can install a similar system for similar cost.

However, the payback equation is different, as US power is massively subsidised and so your costs per kWh are much lower than they should be. Indeed, if you were all paying the real price for coal fired elec, most of you would not be able to afford the energy bills of your inefficient homes. In short, the biggest probs is that you all use too much energy. But, that aside, solar is still cheap, and on a level playing field, larger scale PV generators are already producing energy cheaper than fossil fuel generators.

The biggest issue off-grid systems face is battery costs. Lead-acid batteries are still the mainstay and prices have not fallen in recent years, in fact they have mostly increased. As EVs come onto the market and many more companies start making lithium batts en-masse, lithium prices will fall to the point where lead-acid won't be economical, so their prices will also drop, but that might take a few years.

Mr T

Laura, You are doing what a sensible person would do. There are always announcements about how some one is revolutionizing an industry but even when the idea is truly revolutionary it can take many years to become commercially available. Much of what is posted on gizmag is someone trying to get more money for further study, think of Solyndra or the many other solar outfits that have tried to make better solar cells. It will improve but over decades like most new technology. I do like the Gizmag news but I wish they would be a little less of a tease. They leave out information that we would like to know, such as the improvement in price/watt that we might get with this cheaper cell should it ever make it to market.


I sure would like to see more talk about Liquid Fluoride Thorium Reactors (LFTRs).

Solar should always be a part of the solution even though it is not as scalable as we might like it to be.

LFTRs, on the other hand, are entirely scalable and in a direction away from the norm.

I for one would love to have my community provide its own power, rather than moving it across hundreds of miles and multiple states.

If anyone has the information for the downside to LFTRs please post it. I have looked and found nothing significant to stop our moving on this technology other than politics.

Dr. Veritas


"First you question the numbers in the article, and than to counter argue... you pull "3x" out of your ass showing you have no concept of the laws of thermodynamics." 1) I did not question the numbers in the article, I questioned the statement "high-efficiency" 2) I don't need thermodynamics to fully appreciate the laws of The WALLET. Efficiency ultimately has a real, monetary cost. Solar is high, petrochemicals are low.

"A one meter square solar cell can only convert about 20% of the total energy output of a ONE SQUARE METER "cut" of the light hitting the earth. On the other hand to get hydrocarbon energy out of one square meter of light, first it has to shine on vegetation, which converts it to carbohydrates, which then goes trough the long process of sequestration over millions of years, before it can turn into hydrocarbons... so exactly how efficient is this process?????? " 1) What you meant to say is that for a soalr cell to convert any percentage of the light hitting the earth...first, the universe has to come into creation, then all the matter created has to coalesce into proto galaxies, then stars have to form, then planets have to form then our solar system has to form, then human life has to evolve and mature to a level where solar panels can be contemplated, developed and installed - a process taking at least 14 Billion years (assuming the Big Bang Theory were correct - and it isn't as most credible scientists today have already determined).

I can play that stupid "smart" game too.

Here is the point. Solar has a LONG way to go to rival the efficiency of petrochemical, nuclear, etc. that are all readily available, reliable and cheap.

This article's use of the term "high-efficiency" is nonsense. It should state "more efficient than existing".

The only thing less efficient than solar is Government's farcical attempts to promote its use. That will change when the incremental improvements - achieved thru the collective improvements in all the materials used to capture, store and deliver it - reach a critical mass that triggers genuine interest from the business community to exploit it.

And, while we're at it, stop looking at my ass.

Joseph Boe

Well dr veritas "lifters" as those thorium doohickeys are called make some nasty poisonous waste that contains really bad stuff like beryllium, flourine and various radionuclides. Its particularly nasty because it's water-soluble and therefore tricky to contain. No one has really done the hard yards on the practical yet and the green movement will react vigorously to prevent any.


What about the guys at Caltech? Harry Atwater and his team apparently produced remarkable results with flexible, cost-effective cells in 2010:


I think Boe needs to check his facts! Over one hundred years after the invention of the modern automotive combustion engine, efficiency clusters around an average 25%. If we check back on PV cells in another 50-60 years, I have no doubt we will see much greater improvement in efficiency. That is automotive efficiency, using gasoline to produce electricity (a fair comparison) doesn't even get to that level. That said, obviously the article refers to economic efficiency, cost per kWh in this case.

Dave Beachler

@ Dave Beacher -

You are correct. I was blending my terms and that was inaccurate. I was coming from the perspective of practical cost (as you point out) and because the useable energy produced by solar is still so much more expensive to obtain than its combustible equivalent, I simply called that "efficiency".

When solar (or any other non-petroleum) can power a car, rocket, airplane, etc. to the same level of performance (speed distance load capacity) at a similar cost, I'll consider it efficient.

Joseph Boe

I wanted to thank you for this great read!! I definitely loved every little bit of it. I have bookmarked your site to check out the new stuff you post.


I suspect that at this point, the fastest and easiest way to increase the "efficiency" of solar would be to find a way to drastically reduce the installation cost. It seems like the cost of the panels is $1 per Watt or less plus an inverter for a grid-tied system...let's say that is $500. So, if I want 2kW of installed power, I've got $2k + $500 or $2500 for raw materials, raw electric materials that is. We'll say $500 more for mounting material or $3000 total....totally cheap at that point...Now if it costs $3k - $6k to pay someone to install it (permits, mounting, wiring, inspection, etc)...Now it is not so cheap. Streamline this process and reduce the cost of installation for 2kW to something like $1000 (plus the aforementioned $500 in mounting hardware) and suddenly you have the benefit of an equivalent 10%-15% improvement in efficiency in solar...the benefit is almost instantaneous compared to the 5 to 10 years to market for lab efficiency breakthroughs.

Post a Comment

Login with your Gizmag account:

Related Articles
Looking for something? Search our articles