• From left to right, Associate Professor Michael Strano with graduate student Ardemis Boghossian and postdoctoral fellow Moon-Ho Ham, in one of the labs where they carried out their experiments.

    Photo: Patrick Gillooly

    Full Screen
  • This proof-of-concept version of the photoelectrochemical cell, which was used for laboratory tests, contains a photoactive solution made up of a mix of self-assembling molecules (in a glass cylinder held in place by metal clamp) with two electrodes protruding from the top, one made of platinum (the bare wire) and the other of silver (in a glass tube).

    Photo: Patrick Gillooly

    Full Screen

Solar cell, heal thyself

New self-assembling photovoltaic technology can keep repairing itself to avoid any loss in performance.

Plants are good at doing what scientists and engineers have been struggling to do for decades: converting sunlight into stored energy, and doing so reliably day after day, year after year. Now some MIT scientists have succeeded in mimicking a key aspect of that process.

One of the problems with harvesting sunlight is that the sun’s rays can be highly destructive to many materials. Sunlight leads to a gradual degradation of many systems developed to harness it. But plants have adopted an interesting strategy to address this issue: They constantly break down their light-capturing molecules and reassemble them from scratch, so the basic structures that capture the sun’s energy are, in effect, always brand new.

That process has now been imitated by Michael Strano, the Charles and Hilda Roddey Associate Professor of Chemical Engineering, and his team of graduate students and researchers. They have created a novel set of self-assembling molecules that can turn sunlight into electricity; the molecules can be repeatedly broken down and then reassembled quickly, just by adding or removing an additional solution. Their paper on the work was published on Sept. 5 in Nature Chemistry.

Strano says the idea first occurred to him when he was reading about plant biology. “I was really impressed by how plant cells have this extremely efficient repair mechanism,” he says. In full summer sunlight, “a leaf on a tree is recycling its proteins about every 45 minutes, even though you might think of it as a static photocell.”

One of Strano’s long-term research goals has been to find ways to imitate principles found in nature using nanocomponents. In the case of the molecules used for photosynthesis in plants, the reactive form of oxygen produced by sunlight causes the proteins to fail in a very precise way. As Strano describes it, the oxygen “unsnaps a tether that keeps the protein together,” but the same proteins are quickly reassembled to restart the process.

This action all takes place inside tiny capsules called chloroplasts that reside inside every plant cell — and which is where photosynthesis happens. The chloroplast is “an amazing machine,” Strano says. “They are remarkable engines that consume carbon dioxide and use light to produce glucose,” a chemical that provides energy for metabolism.

To imitate that process, Strano and his team, supported by grants from the MIT Energy Initiative and the Eni Solar Frontiers Center at MIT, produced synthetic molecules called phospholipids that form disks; these disks provide structural support for other molecules that actually respond to light, in structures called reaction centers, which release electrons when struck by particles of light. The disks, carrying the reaction centers, are in a solution where they attach themselves spontaneously to carbon nanotubes — wire-like hollow tubes of carbon atoms that are a few billionths of a meter thick yet stronger than steel and capable of conducting electricity a thousand times better than copper. The nanotubes hold the phospholipid disks in a uniform alignment so that the reaction centers can all be exposed to sunlight at once, and they also act as wires to collect and channel the flow of electrons knocked loose by the reactive molecules.

The system Strano’s team produced is made up of seven different compounds, including the carbon nanotubes, the phospholipids, and the proteins that make up the reaction centers, which under the right conditions spontaneously assemble themselves into a light-harvesting structure that produces an electric current. Strano says he believes this sets a record for the complexity of a self-assembling system. When a surfactant — similar in principle to the chemicals that BP has sprayed into the Gulf of Mexico to break apart oil — is added to the mix, the seven components all come apart and form a soupy solution. Then, when the researchers removed the surfactant by pushing the solution through a membrane, the compounds spontaneously assembled once again into a perfectly formed, rejuvenated photocell.

“We’re basically imitating tricks that nature has discovered over millions of years” — in particular, “reversibility, the ability to break apart and reassemble,” Strano says. The team, which included postdoctoral researcher Moon-Ho Ham and graduate student Ardemis Boghossian, came up with the system based on a theoretical analysis, but then decided to build a prototype cell to test it out. They ran the cell through repeated cycles of assembly and disassembly over a 14-hour period, with no loss of efficiency.

Strano says that in devising novel systems for generating electricity from light, researchers don’t often study how the systems change over time. For conventional silicon-based photovoltaic cells, there is little degradation, but with many new systems being developed — either for lower cost, higher efficiency, flexibility or other improved characteristics — the degradation can be very significant. “Often people see, over 60 hours, the efficiency falling to 10 percent of what you initially saw,” he says.

The individual reactions of these new molecular structures in converting sunlight are about 40 percent efficient. Theoretically, the efficiency of the structures could be close to 100 percent, he says. But in the initial work, the concentration of the structures in the solution was low, so the overall efficiency of the device — the amount of electricity produced for a given surface area — was very low. They are working now to find ways to greatly increase the concentration.

Philip Collins ’90, associate professor of experimental and condensed-matter physics at the University of California, Irvine, who was not involved in this work, says, “One of the remaining differences between man-made devices and biological systems is the ability to regenerate and self-repair. Closing this gap is one promise of nanotechnology, a promise that has been hyped for many years. Strano's work is the first sign of progress in this area, and it suggests that ‘nanotechnology’ is finally preparing to advance beyond simple nanomaterials and composites into this new realm.”

Topics: Biology, Chemistry and chemical engineering, Energy, Solar, Eni, Graduate, postdoctoral


This is a wonderful thing for the future, but we should never forget what we could do today. So support CSP and unleash the Solar Millennium
If we could integrate the regeneration and self-repairing process into conjugated polymer based electrochemical solar cells, It would be definitely a solutions for future energy crisis. Dr Joseph Makkolil,CUSAT
The most common comment I'm seeing about this, is this is the last we hear of it. I don't blame them. We often see stories of such breakthroughs, only to never hear about them again. The view seems to be that breakthroughs where energy are concerned will be swallowed up by big business and government and mankind will never see the benefits. Money talks where modern science is involved, or should I say money and power (politics) talks. Do you suppose this is the last we will hear of this and will it be buried with all the others?
it is seriously a fantastic and awesome invention by the MITians ... hats off to you guys.
This is great news. Nature nurtures. This research is ground breaking as it gives a new tools and methodologies to understand Natures's secrets and harness them for our sustainability on this beautiful planet - EARTH...
There is a serious problem with the paragraph comparing the solar conversion efficiency of this system with photovoltaics. The 40% given by the authors is a quantum yield for a single wavelength of light (785nm). The best solar cells have a quantum yield of better than 95% at all wavelengths across the entire visible and near IR spectra, more than twice as efficient as what is reported here. The 20% efficiency given for solar cells is the efficiency for converting the entire solar spectra into electricity. The solar conversion efficiency of this system is several orders of magnitude lower than that. That paragraph needs to be deleted. <strong><em>Editor’s note: The article has been corrected to remove the comparison with the efficiencies of conventional PV devices (“…or about double the efficiency of today’s best solar cells.”). The paper does not make this comparison.</em><strong>
I agree that we see lots of breakthroughs that seem to have a real benefit for us all. The rub comes in when somebody tries to commercialize those breakthroughs. Many things may work well in the lab, but may be impractical or unsustainably costly when implemented on any scale that may provide some benefit. It may be true that in some cases, nobody even tries to commercialize it, or it is attempted and is jettisoned by people who have little creativity or interest in anything except the tried-and-true. But a vast conspiracy -- doubtful. There are many other more reasonable explanations as to why seemingly fantastic discoveries vanish. For example in this case: 1) This article does not state how long the phospholipids maintain their structure. Is it seconds, minutes, or weeks? So how often does the cell need to be reconstituted? 2) How much time does it take to reconstitute a photocell? And how much time is this a fraction of its operating time? A long reconstitution time and a short operating time means that it is not commercially viable. 3) Can the reconstitution be automated, with fallible equipment? How much equipment is needed for the reconstitution, and how intelligent must the equipment be? 4) And of all this, the big question is ... how expensive would be any *retail* system that used this process? Something to think about.....
then how does the 100% theoretical efficiency compare ?
This is great news. I understand that the effency is reduced slightly, but I am sure that you will develope new ideas. Keep up the good work.
Back to the top