Perovskite Solar Cells Beat New Records (In The Lab)

---

Support CleanTechnica's work through a Substack subscription or on Stripe.

---

By James Wimberley

News from the frontier of perovskite solar cell research:

• A team at China’s Huazhong University in Wuhan, working with pioneer Michael Grätzel at Lausanne, have fabricated cells without a hole transport layer, and done it by dripping a solution of the ingredients on to a carbon mesh. The cells have gone 1000 hours, a typical year’s exposure outdoors, without degradation.

• Meanwhile researchers at Sheffield University in England have made a working cell simply by spraying a solution of their patent gunk on to a surface – could be anything really, flat or curved.

Neither cell is efficient. At 12.8% and 11% respectively, they are well behind the perovskite solar cell record of 19.3%, let alone the 21% of production monocrystalline silicon cells. So why the fuss? Isn’t this another scheme which looks good in the lab but fails in the market, defeated by silicon through simply getting cheaper?

To get perspective, we need to backtrack. What is a perovskite? The first thing is that it’s not a material but a crystal structure similar to the original mineral found in Siberia in 1839. Take a cubic lattice A of positive ions of a metal like calcium, magnesium or lead. Fit another cubic lattice B of a different metal like titanium or a metalloid like silicon inside it, with each B atom (also positively ionised) at the centre of an A cube. The B ions have to be smaller than the A ones – this has nothing to do with mass. Finally, fit a third negatively charged element O, oxygen or a halogen like chlorine, in the face centres of the A lattice. That makes three more cubic lattices inside the same structure, if you are still with me. Since you have a range of elements to choose from in each lattice, there are dozens of possible perovskites.

Most don’t exist in nature, but those that do are very common. Magnesium-silicon perovskite minerals make up a substantial proportion of the Earth’s mantle: a few million trillion tons (the mantle is around 4 x 10^21 tonnes). Far from being exotic, they form naturally in many circumstances, as a stable low-energy configuration. So with skill it’s not too difficult to make them, without high energies or weird processes. There is enough variety using common elements for rare ones like tellurium to be unnecessary.

The perovskite family has an amazing range of properties, including superconductivity and “colossal magnetoresistance.” Suitably for this Swiss Army knife of a material, the grandfather of all the research on PV perovskite is Swiss: Michael Grätzel of EPFL Lausanne, the French-speaking younger sibling of Einstein’s ETH in Zurich.

He first worked on dye-sensitive cells, which have not made it commercially, and switched to the related perovskites less than a decade ago.

The progress since 2009 has been astonishing, even for a dynamic sector like solar: from 4% to 19% efficiency in five years. Chart from Burn & Meredith, Nature, January 2014:

The chart is out of date – the record has risen to 19.3%, announced by Yang Yang at UCLA.

We’ve seen many promising solar technologies crash and burn. Why should this one be different?

There is no guarantee of course, but perovskites have a lot going for them. The underlying materials are cheap. There’s no fundamental reason why efficiency can’t increase further. Most are inorganic, so less likely to degrade than organics. It’s inherently a thin-film technology. You build up layers, not saw thin slices off ingots. Panels are inherently thin and light. They are translucent, so you could put them on windows, or on silicon cells.

There are many challenges to meet before commercial success, and any one could be fatal. But there is progress to report on all of them.

• Toxicity. Some good candidates use lead, a nasty pollutant. Co-workers of Snaith’s and a team at Northwestern University have made working cells using tin instead. The efficiency is only 6%, but it’s the early days.

• Easy fabrication. The first cells used complex bespoke fabrication methods. Last year, Snaith at Oxford announced 15% efficiency cells made with vapour deposition, a standard method in the semiconductor industry. Mei’s lab has now gone further, using drop casting – dribbling the gunk from pipettes (the cheapest go for $10) into a prepared matrix. The Sheffield team has used simple spraying.

Mei in Wuhan has added another important twist. Up to now, everybody has thought it essential to transport the “holes” away as well as the electrons knocked loose by sunlight, to prevent them recombining. But it turns out this doesn’t happen fast enough to matter. Cutting out the “hole transport layer” means a great simplification of design and fabrication.

• Durability. Mei has just achieved 1000 hours exposure without noticeable degradation. This is the one that will wake up Yingli. Silicon PV is very long-lived. Early panels are running fine after 40 years, at slightly less output than when they started. Any competitor technology has at least to match the 25 years for which silicon panels are conservatively warrantied.

• Escaping the gravity well. Incumbency creates an economic gravity well that often traps angel investors. However good it looks in the lab, any rival scheme has to overcome the economies of scale and learning that silicon has already achieved. New silicon PV fabs are being built at 500 MW of annual capacity each – impossible to beat in a garage. All the published perovskite research so far has taken place in university or government research labs, or their tiny spinoffs. Grätzel’s collaboration with Chinese researchers is probably intended to speed up commercial adoption. But so far there have been no announcements of licensing by major solar firms. It will take deep pockets to fulfil perovskite’s promise. To spread the risks, major manufacturers could set up a consortium like Sematech.

We are eagerly waiting for someone in the lively and fast-growing research community to announce a working tandem cell, perovskite on silicon, with a combined efficiency of 30%. Fortunes and Nobel prizes beckon. The EPFL would surely like to notch a score against older brother ETH – they haven’t got a Nobel yet, in spite of top international rankings.

* * *

About the Author: James is one of our regular commenters. He blogs at samefacts.com, a site covering current affairs, drugs, health and arts policy, and sometimes energy.