Solar energy is an essential part of the global move toward clean, renewable energy, and it is critical that the growing solar photovoltaic industry is itself truly safe and sustainable. Little attention is currently being paid to the potential risks and consequences of scaling up solar PV cell production. The solar PV industry must address these issues immediately, or risk repeating the mistakes made by the microelectronics industry.
Silicon-based solar PV production involves many of the same materials as the microelectronics industry and, therefore, presents many of the same hazards. Here is an overview of some of the hazards posed by crystalline silicon (c-Si) PV production technologies - the most common technology found in the solar sector.
Start with silicon
As with the production of silicon chips, production of c-Si wafers begins with the mining of silica, found in the environment as sand or quartz. Silica is refined at high temperatures to remove the oxygen and produce metallurgical grade silicon, which is approximately 99.6% pure. However, silicon for semiconductor use must be much purer.
Higher purities are achieved through a chemical process that exposes metallurgical grade silicon to hydrochloric acid and copper to produce trichlorosilane gas. The trichlorosilane is then distilled to remove remaining impurities, which typically include chlorinated metals of aluminum, iron and carbon. It is finally heated or “reduced” with hydrogen to produce silane gas. The silane gas is heated again to make molten silicon, used to grow monocrystalline silicon crystals or used as an input for amorphous silicon.
The next step is to produce crystals of either monocrystalline or policrystalline silicon. Monocrystalline silicon rods are pulled from molten silicon, cooled and suspended in a reactor at high temperature and high pressure. Silane gas is then introduced into the reactor to deposit additional silicon onto the rods until they “grow” to a specified diameter.
To produce multicrystalline silicon, molten silicon is poured into crucibles and cooled into blocks or ingots. Both processes produce silicon crystals that are extremely pure (from 99.99999% to 99.9999999%), which is ideal for microchips, but far more than required by the PV industry. The high temperatures required for c-Si production make it an extremely energy-intensive and expensive process, and also produces large amounts of waste. As much as 80% of the initial metallurgical grade silicon is lost in the process.
Sawing c-Si wafers creates a significant amount of waste silicon dust called kerf, and up to 50% of the material is lost in air and water used to rinse wafers. This process may generate silicon particulate matter that will pose inhalation problems for production workers and those who clean and maintain equipment. The U.S. Occupational Safety and Health Administration (OSHA) has set exposure limits to keep ambient dust levels low and recommends the use of respiratory masks. But it has been suggested that, despite the use of respiratory masks, workers remain overexposed to silicon dust.
The use of silane gas is the most significant hazard in the production of c-Si because it is extremely explosive and presents a potential danger to workers and communities. Accidental releases of silane have been known to spontaneously explode, and the semiconductor industry reports several silane incidents every year.