How do solar cells work?

 

 There are really only two possible endpoints for human energy production, and they’re both fusion. Either we find a way to create tiny, controlled fusion reactions here on Earth (fusion power) or we find a way to usefully collect a good portion of the energy already being released form the enormous fusion reactor our solar system has built right in (solar power). The nice thing about the solar option is that it can come about incrementally, giving us partial utility while inching ever-closer to the tipping point, when it could provide for the majority of our electrical needs. But what is a solar cell, the centrally important component of solar power, and how does it work?
A solar cell, also called a photovoltaic cell, is defined as any device that can capture some of the energy of a photon of light, and pass that energy on to a device or storage medium in the form of electricity. Not all solar power is photovoltaic in nature, as some solar technologies collect the heat of absorbed photons, rather than their energy, directly. Still, with such a general definition, the term photovoltaics encompasses a wide variety of different technologies.

People in bunny-suits making solar panels.

All of them have one thing in common, however: they use the energy of a photon to excite electrons in the cell’s semi-conducting material from a non-conductive energy level to a conductive one. What makes this complex is that not all photons are created equal. Light arrives as an unhelpful amalgamation of wavelengths and energy levels, and no one semi-conducting material is capable of properly absorbing all of them. This means that to increase the efficiency of capture of solar radiation, we have to make hybrid (“multi-junction”) cells that use more than one absorbing material.
Each semi-conducting material has a characteristic “band gap” or a spectrum of electron energies which the material simply cannot abide. This gap lies between the electron’s excited and unexcited states. An electron in its rest state cannot be excited into usefulness unless it receives enough excess energy to jump right over this band gap. Silicon has a nice, achievable band gap, one that can be bridged by a single photon’s-worth of extra energy. This allows silicon to be nicely either on (conducting) or off (not), as defined by the position of its potentially conductive electrons.
A material like graphene could, in one sense, be a far better basis for a photovoltaic cell than silicon due to its incredible electrical efficiency and the potential to be packed far more densely on the panels themselves — the big problem comes back to the band gap, and graphene’s inability to be properly excited by the power of an incoming photon. Some complex graphene devices like dual gate bilayer graphene transistors — but the problems with actually manufacturing such devices offsets the potential gains, at least for now.

Solar power is a lot easier to collect in space — but then you’ve got to actually get it down to the surface.

Real progress will have to wait for a suitably affordable super-material is found that can provide a useful band gap while also beating silicon’s mechanical and electronic properties by a fair margin. Until then, interim solutions have managed to greatly increase the functional abilities of silicon-based panels.
Anti-reflective coatings increase the amount of light absorbed overall, while chemical “doping” of the transistors themselves can improve silicon’s optical abilities. Some solar setups use fields of mirrors to concentrate as much solar radiation as possible on just a few high-capacity cells at the center. Many are now even designed as light-capture devices, so light that enters gets bounced around internally, forever, until it’s all eventually absorbed. Last fall, researchers at the University of Michigan even developed a fully transparent solar cell.
Heat may also be an increasingly important part of solar power rigs, since any radiation not electronically absorbed will at least be partially absorbed as raw heat. Using this heat to boil water, or even heat homes directly, could help civilian solar power improve overall efficiency even while electrical super-materials continue to play catch-up.
Even more out-there concepts, like space-based solar power, offer some potential by capturing light before it’s filtered through the Earth’s atmosphere; Japan wants to generate a gigawatt of solar power in space, for instance. The problem is getting the power down to the surface, where it could be useful to human beings. The Japanese initiative looks to use lasers for that purpose, but there’s no telling whether bypassing the atmosphere will prove to be a winning strategy, overall.
Solar cells have been hamstrung by several decades of premature headlines announcing such a winning overall strategy and the oncoming dominance of solar power. The reality is that there will almost certainly never be any such eureka moment in engineering. Solar cell technology will be amended and upgraded until it passes some abstract threshold based on affordability, the state of power storage and transmission technology, and the local annual level of sunlight.
All types of solar power will be important to any real attempt to roll out green power on a national scale. Unless fusion makes huge leaps forward, or classical nuclear power becomes a whole lot more popular, you can bet that solar will be a big part of our energy future.