It should be clear to anyone paying attention that we need to wean ourselves off fossil fuels as quickly as possible. The pollution they generate harms health, contributes to global warming, and causes long term damage to the economy. We are also experiencing a great example of how we will never truly be energy independent as long as our gas prices are determined by global markets over which we have little control, and which a single dictator can throw into chaos.

There is a lot of debate about what is the optimal path from where we are now to where we want to be in terms of our energy infrastructure, which I have discussed before and won’t repeat here. What is fairly clear, however, is that solar power is likely to play a critical and increasing role in our energy infrastructure going forward. Solar power is now responsible for about 4% of total US power generation (combined with wind, these renewables just hit 20% power production in March 2022). That is a 36% increase in solar power over last year. Really the main debate is about where renewable power will level off, and how to best integrate them into the overall power grid (which depends a lot on things like grid storage and updates). By 2050 a much greater portion of our energy will likely come from solar, so advances in solar technology will similarly have a huge impact on the cost and effectiveness of our solar infrastructure.

Right now silicon based solar photovoltaic (PV) panels are the industry standard. They have become incredibly cheaper and more energy efficient over the last 20 years, but there are concerns that they may be getting close to the limits of this technology. One huge limiting factor for silicon is that PV panels need to be manufactured at 3,000 degrees F, which itself requires a lot of energy, reducing the energy and carbon efficiency of silicon PV technology. Silicon is also stiff and opaque, which limit its applications.

For these reasons researchers have been studying perovskite for years. Perovskites are a class of crystaline chemical structures – “perovskite compounds have a chemical formula ABX3, where ‘A’ and ‘B’ represent cations and X is an anion that bonds to both.” They can be made into thin films, which are flexible and can be transparent. Importantly, they can be manufactured at room temperature. This would make their commercial production massively more efficient that silicon, and this is the primary reason that researchers have been focusing on perovskite PV cells as the likely successor to silicon. In 2012 the first perovskite PV cell with energy efficiency of >10% was produced. Currently silicon PV cells are 18-20% efficient.

Over the last decade huge strides have been made with perovskite – with a recent study showing a perovskite PV cell of 25% efficiency and after >1,500 hours of use maintaining 98% of this efficiency. That 25% is great, but this is not yet up to commercial standards because of the stability issue. That has been the Achilles heel of perovskite, it’s not stable. The first perovskite PV crystals lasted for seconds to minutes before breaking down. Over the years researchers extended that lifespan to hours, then months, and finally years. But the industry standard is that PV cells need to have a usable lifespan of >20 years, and no perovskite PV cell could get close.

However, Princeton Engineering researchers recently published a report of a new perovskite PV design that they claim has a conversion efficiency of 14.9% to 17.4% (within the commercial range) and a lifespan of 51,000 hours, > 5 years. That lifespan, however, is under continuous bright light exposure and high temperature. They estimate this would translate into a real world lifespan of >30 years, way past the 20 year industry standard. If true, this is a huge milestone for perovskite PV cells.

The primary innovation, which is not unique to this group but which they were able to optimize, is the use of a 2-dimensional cap layer (a thin film only several molecule thick) that protects the perovskite crystal and keeps it from breaking down. The researchers studied many different permutations until they found one that worked. The researchers also had to develop a method for testing the longevity of their PV cells. The longer they last, the harder that is to test. We can’t literally test them under 20 years of usage, for example, because that study would take 20 years. So they developed a method of accelerated aging, using continuous light exposure and high temperatures, and then extrapolate out to real-world use. Of course, this introduces an element of uncertainty, and we won’t know how well this method translates until we do the real world testing, but of course that will take 20-30 years. In the meantime this method may work as a reasonable proxy.

As always, it’s difficult to extrapolate from the lab to industry, and often that translation is where new technologies go to die. But there is reason for optimism here, in that the primary advantage of perovskite is that it is easier to manufacture. It terms of efficiency it’s essentially a lateral move from silicon (and may even start at a slight disadvantage), but eliminating the need for high temperatures during construction will mean much lower manufacturing costs and greater energy and carbon efficiency (in terms of the total carbon released per unit of energy obtained). Perovskites may be the technology to get us past the limits of silicon, and push the industry forward over the next few decades.

The other main competition comes from organic solar cells, which are cheap and flexible, but suffer from relatively low conversion efficiency. However, in February 2021 a National Renewable Energy Lab (NREL) researcher developed an organic solar cell with 18% conversion efficiency. Organic solar cells can be cheaply manufactured because they are basically plastics. They can be sprayed onto a substrate, which can be rigid or flexible. So maybe the future of solar belongs to organic PV technology. Or perhaps silicon, perovskite, and organic will all exist together as complementary technologies. It’s good to have options, and it will be interesting to see which technology ultimately dominates.