Innovation in solar cell technology has slowed as startups struggle to get a foothold in a tough market and solar panel manufacturers delay purchasing the equipment they need to manufacture more efficient cells. But First Solar, one of the world’s largest solar companies, continues to invest in boosting the efficiency of its solar cells.

The company, which is based in Tempe, Arizona, announced this week that it had set a new world record in efficiency for thin-film cadmium telluride solar panels. The equipment it uses to produce the record-setting panels will eventually be installed on all its production lines. It also announced the acquisition of Tetrasun, a startup with high-efficiency silicon technology that First Solar hopes to bring to market next year. First Solar’s stock jumped from $29 to over $40 on Tuesday and is still above $35 a share.

First Solar is able to make these investments because it is in a much better position than other major solar manufacturers, most of which are either declaring bankruptcy or on the brink of it (see “Why We Need More Solar Companies to Fail” and “Solar Downturn Casts a Shadow over Innovation”). It’s doing better for at least two reasons. Its solar panels are cheaper to make than conventional silicon solar panels, which has given it better profit margins. And it was one of the first companies to expand beyond making solar panels to become a project developer, designing and installing complete solar power plants. These projects create a steady market for First Solar’s panels and help it drive down costs in areas besides the panels themselves, which account for less than a quarter of the expense of solar power. The company’s good balance sheet and project experience also help lower the risk to investors, helping it secure better financing rates. And financing is now the biggest single contributor to the cost of solar power, accounting for 36 percent of the total for large installations and even more for smaller ones.

The industry is “getting so good on the technology that finance costs and project development costs are becoming dominant,” says Raffi Garabedian, First Solar’s chief technology officer. “We’re entering an era—the next five years, I think—where a lot of effort is going to be applied to reducing the cost of financing these systems.”

By next year solar power could cost as little as 10 cents per kilowatt-hour without subsidies and thus become cheap enough to compete with fossil fuels in many markets around the world, First Solar says. It expects to lower costs to 7.5 cents per kilowatt-hour by 2016, bringing solar power close in price to one of the cheapest sources of power in the United States—a new natural-gas power plant, which the U.S. Energy Information Administration expects to average 6.5 cents per kilowatt-hour over its lifetime.

Thin-film solar panels are less efficient than conventional silicon panels, and efficiency not only determines the number of panels needed for a project but affects costs for installation and financing. However, First Solar has been closing the efficiency gap. By the end of last year, the company’s cadmium telluride panels were converting about 13 percent of the energy in sunlight into electricity, compared with roughly 15.5 percent for silicon. As it installs more of the new equipment that allowed it to make its record-setting panel, that figure should increase. By 2016, the company expects its average solar panels to reach nearly 17 percent efficiency, with much of the improvement coming from advances that have already been demonstrated it in its labs. Silicon panels will also improve over that time, but First Solar expects to have comparable efficiency at that point. Garabedian thinks it could be possible to reach 19 percent efficiency within five years.

Even with these advances, First Solar doesn’t expect to be able to compete in certain markets—especially in places like Japan. The solar market has been booming there since the Fukushima disaster caused nuclear power plants to be shut down, but most of the demand is for rooftop solar. With that technology, space is limited and costs for installation are relatively high, putting a premium on efficiency.

That’s why First Solar is turning to Tetrasun, which has developed a single-crystal silicon panel with an efficiency rate over 20 percent. First Solar says Tetrasun’s version of this high-efficiency technology will be cheaper to manufacture than similarly efficient—but expensive—silicon solar panels from SunPower and Panasonic.

The Tetrasun investment could be risky, says Shyam Mehta, a senior analyst at GTM Research. “Tetrasun’s technology is far from being proven as commercially viable—right now, it is a company of 14 people with little more than a pilot cell manufacturing plant,” he says. New solar technologies have been difficult to scale up, so a goal of mass-producing its panels by next year may be overly ambitious.

While improvements in efficiency may be the most powerful way to reduce the cost of solar power in the long term, most of the reductions in the next two years are expected to come in financing, which is far more expensive for solar than for many other capital projects, says Travis Bradford, a professor at Columbia University’s School of International and Public Affairs and president of the Prometheus Institute for Sustainable Development, a nonprofit research firm. He says First Solar is in a good position to lower these costs. “Suntech [the Chinese solar panel manufacturer] just went bankrupt, and most of the Chinese competitors have really bad balance sheets,” Bradford says. “But nobody thinks First Solar is going away.”

Innovation in solar cell technology has slowed as startups struggle to get a foothold in a tough market and solar panel manufacturers delay purchasing the equipment they need to manufacture more efficient cells. But First Solar, one of the world’s largest solar companies, continues to invest in boosting the efficiency of its solar cells.

The company, which is based in Tempe, Arizona, announced this week that it had set a new world record in efficiency for thin-film cadmium telluride solar panels. The equipment it uses to produce the record-setting panels will eventually be installed on all its production lines. It also announced the acquisition of Tetrasun, a startup with high-efficiency silicon technology that First Solar hopes to bring to market next year. First Solar’s stock jumped from $29 to over $40 on Tuesday and is still above $35 a share.

First Solar is able to make these investments because it is in a much better position than other major solar manufacturers, most of which are either declaring bankruptcy or on the brink of it (see “Why We Need More Solar Companies to Fail” and “Solar Downturn Casts a Shadow over Innovation”). It’s doing better for at least two reasons. Its solar panels are cheaper to make than conventional silicon solar panels, which has given it better profit margins. And it was one of the first companies to expand beyond making solar panels to become a project developer, designing and installing complete solar power plants. These projects create a steady market for First Solar’s panels and help it drive down costs in areas besides the panels themselves, which account for less than a quarter of the expense of solar power. The company’s good balance sheet and project experience also help lower the risk to investors, helping it secure better financing rates. And financing is now the biggest single contributor to the cost of solar power, accounting for 36 percent of the total for large installations and even more for smaller ones.

The industry is “getting so good on the technology that finance costs and project development costs are becoming dominant,” says Raffi Garabedian, First Solar’s chief technology officer. “We’re entering an era—the next five years, I think—where a lot of effort is going to be applied to reducing the cost of financing these systems.”

By next year solar power could cost as little as 10 cents per kilowatt-hour without subsidies and thus become cheap enough to compete with fossil fuels in many markets around the world, First Solar says. It expects to lower costs to 7.5 cents per kilowatt-hour by 2016, bringing solar power close in price to one of the cheapest sources of power in the United States—a new natural-gas power plant, which the U.S. Energy Information Administration expects to average 6.5 cents per kilowatt-hour over its lifetime.

Thin-film solar panels are less efficient than conventional silicon panels, and efficiency not only determines the number of panels needed for a project but affects costs for installation and financing. However, First Solar has been closing the efficiency gap. By the end of last year, the company’s cadmium telluride panels were converting about 13 percent of the energy in sunlight into electricity, compared with roughly 15.5 percent for silicon. As it installs more of the new equipment that allowed it to make its record-setting panel, that figure should increase. By 2016, the company expects its average solar panels to reach nearly 17 percent efficiency, with much of the improvement coming from advances that have already been demonstrated it in its labs. Silicon panels will also improve over that time, but First Solar expects to have comparable efficiency at that point. Garabedian thinks it could be possible to reach 19 percent efficiency within five years.

Even with these advances, First Solar doesn’t expect to be able to compete in certain markets—especially in places like Japan. The solar market has been booming there since the Fukushima disaster caused nuclear power plants to be shut down, but most of the demand is for rooftop solar. With that technology, space is limited and costs for installation are relatively high, putting a premium on efficiency.

That’s why First Solar is turning to Tetrasun, which has developed a single-crystal silicon panel with an efficiency rate over 20 percent. First Solar says Tetrasun’s version of this high-efficiency technology will be cheaper to manufacture than similarly efficient—but expensive—silicon solar panels from SunPower and Panasonic.

The Tetrasun investment could be risky, says Shyam Mehta, a senior analyst at GTM Research. “Tetrasun’s technology is far from being proven as commercially viable—right now, it is a company of 14 people with little more than a pilot cell manufacturing plant,” he says. New solar technologies have been difficult to scale up, so a goal of mass-producing its panels by next year may be overly ambitious.

While improvements in efficiency may be the most powerful way to reduce the cost of solar power in the long term, most of the reductions in the next two years are expected to come in financing, which is far more expensive for solar than for many other capital projects, says Travis Bradford, a professor at Columbia University’s School of International and Public Affairs and president of the Prometheus Institute for Sustainable Development, a nonprofit research firm. He says First Solar is in a good position to lower these costs. “Suntech [the Chinese solar panel manufacturer] just went bankrupt, and most of the Chinese competitors have really bad balance sheets,” Bradford says. “But nobody thinks First Solar is going away.”

The value of a Bitcoin has grown in the four years since the digital currency was invented, but there’s been little interest from mainstream business or technology investors in using it.

News today from OpenCoin, a startup that today launched its own digital currency called Ripple and tools for making transactions in other currencies, including Bitcoins, suggests that may change. The company says it has attracted early investments, of an undisclosed size, from established venture capital firms including Andreessen Horowitz, which reportedly made over $100 million when Microsoft bought Skype, and Lightspeed Ventures, one of the first investors in networking company Nicira, which sold to VMWare last year for $1.6 billion. OpenCoin has also received investment from the early-stage wing the Founder’s Fund, a venture firm owned by PayPal cofounder Peter Thiel.

Ripple, the currency developed by OpenCoin, is similar to Bitcoin in that it uses math to prevent counterfeiting and fraud. However, Ripple has features intended to make it more practical to use, and that enable Ripple to be used to make transactions with existing currencies. One major difference is that transfers made with Ripple can be confirmed in seconds; Bitcoin transfers take, on average, 10 minutes to be confirmed, and many sites that accept Bitcoins make users wait an hour for confirmation of their transaction.

“Bitcoin is focused more on the currency and less on the payments system,” says Chris Larsen, a cofounder and CEO of OpenCoin. “We’re trying to offer both.” Larsen previously founded the online lending and savings company e-loan, and founded OpenCoin with chief technology officer Jed McCaleb, who created the once-popular file sharing client eDonkey as well as the largest Bitcoin exchange, Mt Gox. OpenCoin currently has 14 employees.

The company’s website for Ripple is more polished and easy-to-use than most sites built for Bitcoin users. As well as sending Ripples to other people, users can also send and exchange U.S. dollars, Euros, Bitcoins, and other currencies using the site. Ripple’s design has those transactions automatically routed through exchange companies that are working with OpenCoin. Tools are also available to allow others to offer software or websites that make use of Ripple.

Transferring Ripples is free, while transactions that involve converting between currencies involve small transaction fees—typically 0.02 percent. That’s significantly less than the fees levied by existing financial companies, such as PayPal, credit card issuers, or banks, says Larsen, an advantage that he says will lead to online merchants experimenting with Ripple, and one that attracted OpenCoin’s investors.

“You can send e-mails for free, but not payments,” says Larsen. “Finally we might get finance to the place where e-mail or social networking has taken communication.”

“OpenCoin’s Ripple protocol, which enables free instant global payments in any currency, including Bitcoin, supports many of the advantages of a math-based currency like Bitcoin while also addressing some of the drawbacks,” says Jeremy Liew, managing director of Lightspeed Venture Partners. If math-backed currencies are made useful, as OpenCoin is trying to do, they will last, he says.

In recent years, technology investors have shown a lot of interest in payments companies such as Square and Dwolla, eyeing the lucrative transaction fees levied by credit-card companies. However, OpenCoin has a very different business model, since the company does not control Ripple even though it created it. OpenCoin plans to hand out some 50 billion Ripples in coming months, and more in the future, in an attempt to get the currency to function independently. Larsen says his company aims to turn a profit by retaining a chunk, likely 25 percent, of the total 100 billion Ripples that will ever exist, in the expectation that the currency gains value.

There are already signs that Ripple is finding favor with some devotees of Bitcoin, yet convincing people and companies outside that community to start using currency not issued by any government will be a tough sell.

Liew, of Lightspeed, notes that Bitcoin has already made significant progress. “While consumer usage is still largely confined to hobbyists, large Internet companies are starting to accept payments or donations in Bitcoin, including Expensify, WordPress, and Reddit.”

The ink is based on two advances from Lund University in Sweden. Professor Lars Samuelson demonstrated that nanowires can improve the efficiency of solar cells, and he developed a new way to manufacture nanowires that could make them practical to use.

Increasing solar cell efficiency is one of the most effective ways to reduce the cost of solar power, since it can lower the cost per watt of solar panels as well as reduce installation costs, because fewer solar panels would be needed (see “Alta Devices: Finding a Solar Solution”).

Lund, Sweden-based Sol Voltaics plans to develop equipment to produce nanowire ink, and then sell it to existing manufacturers.  The ink is expected to boost efficiency by helping solar cells absorb more sunlight.

Research in making nanowires for solar photovoltaics has been going on for years, but fabricating nanowires has never been done in an economical way (see “Nanowires Suck Up Light from Around Them” and “How to Double the Power of Solar Panels”). Nanowires are usually grown on a substrate in a batch process that is too expensive for large-scale production.

The Lund University team lead by Samuelson has developed an alternative method that does away with the substrate. It starts by vaporizing gold to produce aerosol nanoparticles, which flow into a tube-shaped furnace along with two other gases. The gold serves as a seed that catalyzes a reaction with the gases to form gallium arsenide nanowires. In a paper published in Nature last December, the Lund researchers said that the process, called aerotaxy, can grow gallium arsenide nanowires 20 to 1,000 faster than batch deposition methods. By controlling temperature and reaction time, they can control the dimensions of the nanowires, which is key to optimizing their performance for solar cells.

Brian Korgel, a professor of chemical engineering at the University of Texas, says aerotaxy “has the potential to be scaled to a continuous process.” Making large volumes would overcome one of the biggest technical challenges in scaling up gas-phase nanowire production, he says.

In a separate paper in Science earlier this year, the same researchers at Lund University showed that arrays of these nanowires made with the aerotaxy method improved the efficiency of indium phosphide solar cells by 13.8 percent by trapping more light.

The next step for Sol Voltaics is to demonstrate that the effect also works on silicon solar cells, the most common type, says CEO Dave Epstein. After it does that, it intends to develop equipment to manufacture the nanowires. He estimates that the ink would add one or two cents per watt to production costs—it currently costs less than 75 cents per watt to make solar panels. “It only takes one gram of nanowires to cover a square meter of a silicon solar panel, so they only need a very small amount of material,” he says.

 “Every indication is that even if there will be an additional cost, the increased efficiency will far outweigh the cost,” says Alf Bjørseth, the founder of REC Solar and an investor in Sol Voltaics.

If the nanowire-ink-on-silicon approach is effective, the company plans to begin small-scale production in 2015. It expects to need $50 million—much less than a full-scale solar factory—to produce at commercial scale, since it’s only selling an add-on product.

The Brain Research through Advancing Innovative Neurotechnologies, or BRAIN, as the project is now called, aims to reconstruct the activity of every single neuron as they fire simultaneously in different brain circuits, or perhaps even whole brains.

The “next great American project,” as Obama called it, could help neuroscientists understand the origins of cognition, perception, and other enigmatic brain activities, which may lead to new, more effective treatments for conditions like autism or mood disorders and could help veterans suffering from brain injuries.

Big brain science is also on the minds of Europeans; the European Union recently announced a nearly 1.2 billion Euro, 10-year proposal to computationally simulate the human brain from the level of molecules and neurons up through neuronal circuits.

Various tools—from genetics and molecular biology—have helped researchers understand how neurons behave as individuals. But neuroscientists are now able to study only the activity of a handful of these brain cells at a time using voltage-sensing electrode probes.

Other efforts to map the physical connections in the brain are already under way, but these projects either look at dead brains or provide only a rough, low-resolution view of how regions of the brain communicate. For example, the Allen Institute for Brain Science has developed several so-called Brain Atlases that map the physical connections between neurons in different species’ brains as well as the patterns of unique genetics in each neuron. While these static maps are great to learn about the architecture of the brain, they don’t provide information about how neuron activity leads to brain function.

It’s possible to get a rough view of whole-neural-circuit activity using tools like MRI and EEG, but only at a low resolution. And the behavior of the brain in between these two scales—how thousands or millions of neurons interact to control the behavior of discrete circuits in the brain—has been inaccessible. Scientists don’t yet understand how complex interactions among many neurons at once give rise to neural circuit function.

The BRAIN initiative proposes to develop new technologies that can record the activity from thousands, if not millions or billions, of neurons simultaneously at timescales matching behavior and mental activities. The initiative will likely tackle discrete brain circuits within different species of animals to understand how neurons work together to give rise to behaviors, moods, and other mental phenomena.

Developing novel technologies will be necessary to achieve the goals of BRAIN, and these will likely take advantage of recent advances in nanotechnology. Existing sensors can record the electrical activity of neurons, but can typically monitor fewer than 100 neurons at a time.

Emerging micro- and nanofabrication techniques could be used to create smaller chips bearing smaller electrical and even chemical probes that would be less invasive. Nanoprobes bearing several dozen electrodes, for instance, could be stacked to probe hundreds of thousands of recording sites and transmit data wirelessly.

Alternatively, nanoparticles carrying molecules that bring them to specific cell types could lodge in cell membranes so surgical placement wouldn’t be necessary. The nanoparticles could also carry molecules that can sense electrical activity, pressure, or even certain chemicals revealing brain activity.

Novel optical techniques could also aid the mapping project. When a neuron fires, the amount of calcium inside cells increases, so many research groups use calcium-sensitive fluorescent dyes to study neuron activity. But this measurement is once removed from the actual electrical activity of the neuron. A voltage-sensitive fluorescent molecule or other imaging agent could provide a more accurate view of activity.

Synthetic biology could be another useful tool. Enzymes that build strands of DNA are sensitive to ion concentration and will introduce more errors into their DNA production in the presence of calcium. As such, these enzymes could be used as sensors for neuron activity. A predetermined DNA sequence could be implanted into neurons and, as it is copied, the resulting strand of DNA would provide a record of the patterns of errors corresponding to patterns of neuron activity. Error-spotted strands from different neurons could later be sequenced.

Researchers sketched out a rough roadmap for the project in a 2012 proposal. The initiative will most likely start with the development of improved calcium-imaging methods for recording neuron firing, followed by voltage-imaging of neuron activity. Since these two methods would look only at surface structures (because light cannot travel far into brain tissue), the third step could be the development of large arrays of nanoprobes.

In the first five years, the initiative may start with small circuits, such as the whole nervous system of the nematode C. elegans (which has only 302 neurons and 7,000 connections) and discrete circuits of the fruit fly brain. Individual circuits in a mouse nervous system, such as that in the retina or olfaction center, could be tackled within 10 years, and within 15 years, scientists may be able to reconstruct the neuronal activity of the entire neocortex of a mouse.

Even without directly exploring the human brain, the resulting insights could have a profound impact on neuroscience and medicine—that is, if everything in this next great American project goes according to plan.

The remote Alaskan village of Igiugig—home to about 50 people—will be the first to demonstrate a new approach to wind power that could boost power output and, its inventors say, just might make it more affordable.

For decades, the trend across the wind industry has been to make wind turbines larger and larger—because it has improved efficiency and helped lower costs.

John Dabiri, a professor of aeronautical and bioengineering at Caltech, has a heretical idea. He thinks the way to lower the cost of wind power is to use small vertical-axis wind turbines, while using computer models to optimize their arrangement in a wind farm so that each turbine boosts the power output of its neighbors.

Dabiri has demonstrated the basic idea at a 24-turbine test plot in southern California. Grants totaling $6 million from the Gordon and Betty Moore Foundation and the U.S. Department of Defense will allow him to see if the approach can lower wind power costs in Igiugig. The first 10 turbines will be installed this year, and the goal is to eventually install 50 to 70 turbines, which would produce roughly as much power as the diesel generators the village uses now. Dabiri is also installing turbines at an existing wind farm in Palm Springs, California, using his models to generate power by putting up new turbines between existing ones.

Ordinarily, as wind passes around and through a wind turbine, it produces turbulence that buffets downstream turbines, reducing their power output and increasing wear and tear. Dabiri says that vertical-axis turbines produce a wake that can be beneficial to other turbines, if they’re positioned correctly.

The blades of this type of wind turbine are arranged vertically—like poles on a carousel rather than spokes on a wheel, as with conventional wind turbines. Wind moving around the vertical-axis turbines speeds up, and the vertical arrangement of the blades on downstream wind turbines allows them to effectively catch that wind, speed up, and generate more power. (The spinning blades of a conventional wind turbine would only catch some of this faster wind as they pass through it—this actually hurts the turbine’s performance because it increases stress on the blades.) The arrangement makes it possible to pack more turbines onto a piece of land.

Dabiri’s wind turbines are 10 meters tall and generate three to five kilowatts, unlike the 100-meter-tall, multi-megawatt machines in conventional wind farms. He says the smaller ones are easier to manufacture and could cost less than conventional ones if produced on a large scale. He also says maintenance costs could be less because the generator sits on the ground, rather than at the top of a 100-meter tower, and thus is easier to access. The performance of the wind farm at Igiugig will help determine whether his estimates of maintenance costs are correct.

Dabiri says small, vertical wind turbines have other advantages. While the noise of conventional wind turbines has led some communities to campaign to tear them down, his turbines are “almost inaudible,” he says. They’re also less likely to kill birds. And their short profile has attracted a $1 million grant from the Department of Defense to study their use on military bases. Because they’re shorter, they interfere less with helicopter operations and with radar than conventional wind turbines.

The approach, however, faces some challenges. Vertical-axis wind turbines aren’t as efficient as conventional ones—half of the time the blades are actually moving against the wind, rather than generating the lift needed spin a generator. As the blades alternatively catch the wind and then move against it, they create wear and tear on the structure, says Fork Felker, director of the National Wind Technology Center at the National Renewable Energy Laboratory. Dabiri, and researchers such as Alexander Smits at Princeton University, say they are working on improved turbine designs to address some of these issues.

Felker notes that Dabiri’s approach will also require installing a thousand times more wind turbines, requiring potentially millions of wind turbines rather than thousands to generate significant fractions of U.S. power supply. And he notes that, over the last several decades, the wind industry has demonstrated that making ever larger wind turbines lowers costs (“Novel Designs are Taking Wind Power to the Next Level,” “Supersized Wind Turbines Head Out to Sea,” and “The Quest for the Monster Wind Turbine Blade.” “Going in the other direction, I believe, will not be successful,” he says. “I don’t think the math works out.”

Felker thinks that Dabiri’s approach might prove fitting for small, isolated places like Igiugig, where simpler construction and maintenance might be important. “But if you’re trying to transform the overall energy economy,” he says, “you’ve got to go big.”