The solar-powered streetlights that are being tested in Las Vegas get most of their power from high-efficiency, thin-film solar panels made by SunCore Corporation. I wondered how thin-film PV panels could achieve 22% efficiency when most rigid PV panels only reach 16% or so. SunCore vice president Michael Swan enlightened me on the matter: "The differences that I am willing to share with you are primarily in the silicon lattice.  We do several things with diffusion in our process that enhance our ability to convert photons more quickly. Additionally, our starting material is N-type silicon which also provides advantages if you know how to run it in a fab process."

Mr. Swan also told me that SunCore's PV cells are able to convert light from a wider spectrum of wavelengths than most other cells, allowing them to produce more electricity during cloudy days and near sunrise and sunset.

N-Type Silicon

Unlike the majority of photovoltaic panels, SunCore's PV cells start with N-type silicon as their substrate. I'll avoid the deep physics but in short, N-type silicon produces cells that are less susceptible to light-induced degradation (LID), and they're not as likely to be damaged by heat during the production process. N-type PV cells are also more tolerant of certain manufacturing defects. In general, silicon PV cells made with N-type substrates have higher junction efficiencies than their P-type counterparts.

So why are the majority of PV cells made with P-type substrates? Corporate inertia.

P-Type Silicon? That's So 1950s

When Bell Labs created the first photovoltaic cell in the early 1950s, the cells were made from N-type silicon substrates. Engineers quickly discovered that the prototypes were prone to degradation when exposed to cosmic rays, and since the space program was the first major application of PV technology, this was unacceptable. They began to experiment with P-type silicon substrates and found that the new technology was slightly less efficient but able to withstand the pummeling of "cosmic debris." Thus, P-type silicon was dubbed the standard, and its manufacturing process became refined.

Although the majority of PV panels now sit on the Earth's surface where they're safely protected from cosmic rays, companies have been reluctant to adopt the manufacturing process that creates N-type substrates. According to Mr. Swan, "It is a very difficult process to run and more expensive to use N-type and get a consistent yield. Only two companies that we know of can do it: SunPower and SunCore.  Though our wafers are ultimately different we are the only two that can mass produce them.  Others may have a lab sample but cannot get it into production. I think the industry would like to know how to move to N-type but given the price sensitivity in the solar panel industry I don’t see it happening anytime soon except from those interested in a premium product line."

Apples and Oranges

One reason that companies are sticking with the older technology is the simple "cost per watt" metric. Simply put, it's more expensive to produce N-type PV cells. That's not an inherent flaw; it's really due to economies of scale (i.e., fewer companies are doing it). Since designers, installers, and consumers look for the lowest cost per unit of power, the P-type panels look like they're the best ones for the money.

Mr. Swan suggests that cost per watt doesn't take everything into account. For example, PV panels are tested under various lighting conditions, but they're advertised according to peak output, which occurs under full sun. A traditional panel produces very little under clouds, but that's not factored into the peak output measure. On the other hand, SunCore's N-type panels deliver more power at full sun, and they significantly outperform other panels in diffuse light and under cloudy skies. Therefore, they're likely to deliver more energy over the course of a year than their competitors.

Most PV system designers use a simple formula (peak power x peak sun hours per day) to estimate daily production from an array. Although peak sun hours takes cloudiness into account, the formula does not account for all variables, such as a panel's efficiency at different wavelengths of light. (It's like education critics using standardized test scores to "measure the quality" of a school, but I digress.) Swan suggests that cost per annual megawatt-hour would be a more revealing measure of a PV panel's productivity.

From Small Scale to the Vegas Strip

Due to the higher manufacturing costs, the majority of SunCore's cells don't make it to rooftop solar arrays or solar farms. Instead, they end up in small products such as portable electronics, solar powered battery chargers, and even energy harvesting devices used by the military. In those applications, battery charging is the primary purpose of the solar cells. To address that need, SunCore developed advanced maximum power point transfer (MPPT) algorithms that optimize battery charging. Mr. Swan adds, "SunCore taught the major chip companies how to do dynamic MPPT years ago. TI, Microchip, and Linear are all using the old version of our algorithms. We have advanced our algorithm since then and have kept that code to ourselves, which we program into our circuits independently. Our sampling rate of battery/PV panel is fully programmable and is configured to match the product's intended use."

On the larger scale, EnGoPlanet's solar powered street lights use SunCore's technology because they needed a panel that follows the contours of the fixture and they wanted higher efficiency and greater annual energy production.

That's How We've Always Done It

The QWERTY keyboard, the internal combustion engine, and the P-type solar panel; they're all less efficient than their alternatives, yet they're still the standard technologies in their respective products, largely due to human inertia. If Isaac Newton were alive today, he might suggest that it's time for an unbalanced force to act upon us.

Images courtesy of SunCore