Solar Cells: The Photovoltaic Effect From Photon to Grid

01 An Energy Problem Measured in Terawatts

Humanity consumes primary energy at an average rate near 19 terawatts, and electricity demand is climbing faster than at any point in decades as electrification spreads and data centers multiply. The only renewable resource sized for that scale is sunlight, which delivers about 173,000 terawatts continuously to the top of the atmosphere.

The obstacle was never the size of the resource but the economics of conversion: turning diffuse photons into dispatchable electrons at prices power grids will pay. For most of the twentieth century, photovoltaics lived aboard satellites, where cost per watt was irrelevant.

What changed was a manufacturing learning curve. Each doubling of cumulative production cut module costs by roughly a fifth, transforming a spacecraft technology into the cheapest new source of bulk electricity across most of the world.

02 The Photovoltaic Effect

The underlying physics was observed by Edmond Becquerel in 1839 and explained by Albert Einstein in 1905, work that later earned a Nobel Prize. A photon carrying more energy than the semiconductor's band gap excites an electron out of its valence position, leaving a mobile hole behind.

The solar cell's trick is architectural. A p-n junction carries a built-in electric field that sweeps electrons one way and holes the other, so the freed charges cannot simply recombine; instead they flow through an external circuit as usable current.

A solar cell is, in effect, a diode operated in reverse of its usual job: the same junction that rectifies signals in a microchip generates power when sunlight floods across it.

03 Silicon, Doping, and the Cell Stack

Almost all commercial photovoltaics are crystalline silicon, holding about 95 percent market share. Manufacturing starts with ultra-pure polysilicon, grows cylindrical ingots by the Czochralski process, saws wafers around 150 micrometres thick, dopes them with phosphorus and boron, and screen-prints metal contacts finished with an anti-reflective coating.

Cells are wired in series into modules of 60 to 72, laminated between glass and polymer, and warrantied for 25 to 30 years against degradation of roughly half a percent per year. Silicon's band gap of 1.12 electronvolts sits close to ideal for the solar spectrum: photons below it pass through unused, those above it convert with some energy lost to heat.

Around the cells themselves, the balance of system, meaning inverters, mounting, trackers and wiring, now matters as much to total cost as cell efficiency does.

04 Efficiency: Measured and Theoretical

The 1961 Shockley-Queisser analysis set the benchmark: a single silicon junction tops out at 33.7 percent conversion under standard sunlight. Commercial monocrystalline modules reach roughly 20 to 22 percent, laboratory silicon cells about 26 to 27, gallium arsenide single junctions 29.1, and perovskite cells have climbed past 26 in little more than a decade of development.

Laboratory records and factory products differ for structural reasons: manufacturing tolerances, module encapsulation, temperature coefficients and field soiling all subtract from headline numbers. The gap is real but narrow, which is itself remarkable.

Each absolute percentage point carries disproportionate value because land, racking and labour costs are fixed per square metre, so more watts from the same panel area flow almost entirely to the bottom line.

05 Terawatt-Scale Deployment

Installed capacity tells the deployment story in one line. The International Renewable Energy Agency counted roughly 40 gigawatts worldwide in 2010, 228 by 2015, 714 by 2020, and about 1,419 gigawatts by the end of 2023, with single-year additions now in the hundreds of gigawatts.

Module prices fell roughly 90 percent over the same period, and in most markets new solar now undercuts new fossil generation on levelized cost before any subsidy is counted. Capacity factors climbed in parallel as trackers and bifacial panels squeezed more output from each installed watt.

The supply chain is concentrated: China produces on the order of four-fifths of manufacturing capacity across the value chain, from polysilicon to modules, a fact that shapes energy security debates alongside the technology's economics.

06 Limits and Tradeoffs

Sunlight is intermittent, and the famous duck curve of solar-heavy grids captures the consequence: a midday glut of cheap power, an evening ramp that other resources must cover. Storage coupling, curtailment management and flexible demand are now the binding constraints on further growth.

Land use draws scrutiny, though estimates of a fully solar-powered United States require on the order of a tenth the area already under cultivation, and agrivoltaic designs share ground with crops and grazing. Recycling of end-of-life modules and, for thin-film products, cadmium handling remain open industrial problems.

Heat is a quiet tax on output: cell efficiency drops roughly 0.3 to 0.5 percent for every degree Celsius above 25, which is why mounting airflow and hot-climate derating figure so prominently in system design.

07 Legacy and Outlook

The frontier is tandem cells: stacking a perovskite layer on silicon to harvest the spectrum in two passes, with laboratory tandems already exceeding 34 percent, past the single-junction limit. The open question is whether perovskite stability survives decades outdoors as silicon does.

Meanwhile, photovoltaics behaves less like an energy technology and more like a manufacturing one, riding the same cost-decline logic as chips and displays. Forecasts that once extrapolated cautiously now look conservative year after year.

The endgame, if the curves hold, is a grid whose constraint shifts from the cost of generation to the cost of flexibility, with storage and transmission, not panels, setting the pace. The photovoltaic effect, demonstrated in a laboratory 187 years ago, would then underwrite the bulk of world electricity.