Переработка солнечных панелей: химические методы извлечения кремния и экономический потенциал до 2050 года

Snap the frame off a retired module and the glass comes away as one sheet. Underneath sit rows of dark blue squares, each thinner than cardboard, laced with hair fine silver lines. Those squares are why the thing existed. They are also why it will not come apart, because everything worth having is under laminated glass and behind a clear plastic layer engineered never to let go. The whole problem reduces to what you soak it in. Why the silicon will not come loose A module is a sandwich, not an assembly. Glass on the front, cells in the middle, a backsheet behind, and the whole stack fused by a clear polymer that melts during manufacture and locks everything in place. There are no screws in the middle of a panel. Each cell also carries metal printed onto its face, fine silver lines to collect current and an aluminum layer across the back, and that metal has to come off before the silicon underneath is worth anything to anybody. Strip it too gently and the metal stays. Strip it too hard and you eat the wafer, which is why the standard routes reach for hydrofluoric acid or a nitric mixture, or simply burn the plastic off at high temperature. Both approaches work. Both create a second problem where the first one was. What the object is actually made of By weight, a panel is mostly window. Glass is the bulk of it, the aluminum frame is the next largest piece, and the silicon everybody argues about is a sliver of mass. The silver is smaller still. It goes on in lines finer than a hair across the front of every cell, a fraction of an ounce per module, and it is most of the reason a scrapyard will accept the things at all. The object is upside down. Its heaviest part is the cheapest, its thinnest part is the most valuable, and the layer holding the two together is worth nothing whatsoever. The two soaks A group in Singapore published a route in 2023 that is almost aggressively boring. Soak the spent cells in hot diluted phosphoric acid for thirty minutes. Pour it off, add fresh acid, soak for thirty more. That is the entire method. Phosphoric acid is a food industry chemical, the sour note in cola, and it took the aluminum and the silver off the wafers while leaving the silicon standing at 99.2 percent purity and 98.9 percent recovery. Then they built batteries with it. The recovered material went into lithium ion anodes and performed comparably to commercial silicon bought new, which was the actual point, because a clean wafer nobody wants is an expensive paperweight. Other groups have pushed further since. A team in India published a recovery route in 2025 aimed at both anodes and capacitors, and another has printed electrodes directly from recovered cells. The slice of the panel that carries the money Here is where the chemistry meets the accounting. The two soak route is aimed at silicon, and silicon is not where the value in a dead module sits. A new assessment of end of life panels puts the recoverable material value by 2050 above twenty billion dollars a year, and breaks out where that value lives. Aluminum is nearly half of it. Silver is another third of the value, copper takes 7 percent, and silicon, the fraction all that careful acid work is designed to rescue, comes in at 11 percent. The frame is worth more than the cell. That does not make the recovery pointless, but it explains why plants running at scale take the easy fractions first, and why the clear plastic layer remains nobody’s solved problem. How much of this even exists yet The feedstock is the other constraint, and it is far smaller than the headlines suggest. Cumulative panel waste worldwide sits at roughly a million tons, with something like 440,000 tons added each year. That is not a supply chain. That is a pilot. The curve steepens later, past 13 million tons cumulative by 2035 and beyond 220 million tons by 2050, at which point recovered material could cover 70 percent of the solar industry’s own silver demand. Whether the material is even cheap to handle is a separate fight. A retired module currently spends 18 hours tumbling in weak acid to decide which bin it belongs in, and two panels off the same production line in the same year can land on opposite sides of that result. The method is ready. The material is not here yet. That gap is most of the story of solar panels at end of life, where every clever route shows up a decade ahead of the volume it was built for. Read the whole thing? Get the week's signal, not the noise Our sharpest reporting on energy, climate and nature — free, once a week.