Perovskites and Perovskite Solar Cells: An Introduction


The rapid improvement of perovskite solar cells has made them the rising star of the photovoltaics world and of huge interest to the academic community. Since their operational methods are still relatively new, there is great opportunity for further research into the basic physics and chemistry around perovskites. Furthermore, as has been shown over the past few years - the engineering improvements of perovskite formulations and fabrication routines has led to significant increases in power conversion efficiency, with recent devices reaching over 23%, as of June 2018.1



What are Perovskites?

The terms "perovskite" and "perovskite structure" are often used interchangeably. Technically, a perovskite is a type of mineral that was first found in the Ural Mountains and named after Lev Perovski (who was the founder of the Russian Geographical Society). A perovskite structure is any compound that has the same structure as the perovskite mineral.

True perovskite (the mineral) is composed of calcium, titanium and oxygen in the form CaTiO3. Meanwhile, a perovskite structure is anything that has the generic form ABX3 and the same crystallographic structure as perovskite (the mineral). However, since most people in the solar cell world aren’t involved with minerals and geology, perovskite and perovskite structure are used interchangeably.

The perovskite lattice arrangement is demonstrated below. As with many structures in crystallography, it can be represented in multiple ways. The simplest way to think about a perovskite is as a large atomic or molecular cation (positively-charged) of type A in the centre of a cube. The corners of the cube are then occupied by atoms B (also positively-charged cations) and the faces of the cube are occupied by a smaller atom X with negative charge (anion).


generic perovskite crystal structure in ABX form
A generic perovskite crystal structure of the form ABX3. Note that the two structures are equivalent – the left-hand structure is drawn so that atom B is at the <0,0,0> position while the right-hand structure is drawn so that atom (or molecule) A is at the <0,0,0> position. Also note that the lines are a guide to represent crystal orientation rather than bonding patterns.

Depending on which atoms/molecules are used in the structure, perovskites can have an impressive array of interesting properties, including superconductivity, giant magnetoresistance, spin-dependent transport (spintronics) and catalytic properties. Perovskites therefore represent an exciting playground for physicists, chemists and material scientists.

Perovskites were first successfully used in solid-state solar cells in 2012,2,3 and since then most cells have used the following combination of materials in the usual perovskite form ABX3:

  • A = An organic cation - methylammonium (CH3NH3+) or formamidinium (NH2CHNH2+)
  • B = A big inorganic cation - usually lead(II) (Pb2+)
  • X3= A slightly smaller halogen anion – usually chloride (Cl-) or iodide (I-)

Since this is a relatively general structure, these perovskite-based devices can also be given a number of different names, which can either refer to a more general class of materials or a specific combination. As an example of this, we’ve created the table below to highlight how many names can be formed from one basic structure.


A B X3
Organo Metal Trihalide (or trihalide)
Methylammonium Lead Iodide (or triiodide)
Plumbate Chloride (or trichloride)
The perovskite 'name-picking' table: Pick any one item from columns A, B or X3 to come up with a valid name. Examples include: Organo-lead-chlorides, Methylammonium-metal-trihalides, organo-plumbate-iodides etc.

The table demonstrates how vast the parameter space is for potential material/structure combinations, as there are many other atoms/molecules that could be substituted for each column. The choice of material combinations will be crucial for determining both the optical and electronic properties (e.g. bandgap and commensurate absorption spectra, mobility, diffusion lengths, etc). A simple brute-force optimisation by combinatorial screening in the lab is likely to be very inefficient at finding good perovskite structures. 

The majority of efficient perovskites are based on Group IV (specifically, lead) metal halides, and moving beyond this has proved challenging. It is likely more in-depth knowledge than currently available is required to fully explore the range of possible perovskite structures. Lead-based perovskite-based solar cells are particularly good because of a range of factors, including strong absorption in the visible regime, long charge-carrier diffusion lengths, a tunable band gap, and easy manufacture (due to the high defect tolerance and the ability to process at low temperatures).

 


Why are Perovskite Solar Cells So Significant?

There are two key graphs which demonstrate why perovskite solar cells have attracted such prominent attention in the short time since 2012. The first of these graphs (which uses data taken from the NREL solar cell efficiency chart)1 demonstrates the power conversion efficiencies of the perovskite-based devices over recent years, in comparison to emergent photovoltaic research technology, and also traditional thin-film photovoltaics.

The graph shows a meteoric rise compared to most other technologies over a relatively short period of time. Within 4 years of their breakthrough, perovskite solar cells had equalled efficiencies of Cadmium Telluride (CdTe), which has been around for over 40 year. Furthermore, as of June 2018 they have now exceeded all other thin-film, non-concentrator technologies - including CdTe and Copper Indium Gallium Selenide (CIGS). Although it could be argued that more resources and better infrastructure for solar cell research have been available in the last few years, the dramatic rise in perovskite solar cell efficiency is still incredibly significant and impressive.


Perovskite crystal structure in the form ABX3
Perovskite solar cells have increased in power conversion efficiency at a phenomenal rate compared to other types of photovoltaics. Although this figure only represents lab-based "hero cells", it heralds great promise.

The second key graph below is the open-circuit voltage compared to the band gap for a range of technologies that compete against perovskites. This graph demonstrates how much of a photon’s energy is lost in the conversion process from light to electricity. For standard excitonic-based, organic-based solar cells, this loss can be as high as 50% of the absorbed energy, whereas perovskite solar cells regularly exceed 70% photon energy utilisation, and have the potential to be increased even further.4

This is approaching the values of state-of-the-art technologies (such as GaAs), but at a significantly lower cost. Crystalline silicon solar cells, arguably the closest comparator to perovskites in terms of efficiency and cost, are already up to 1000 times cheaper than state-of-the-art GaAs.5 Perovskites have the potential to become even cheaper than this.



The maximum photon energy utilisation (defined as the open circuit voltage Voc divided by the optical bandgap Eg) for common single junction solar cells material systems. Calculated from state of the art cells detailed in NREL efficiency tables.6


 


What Issues do Perovskites Face?

The biggest issue in the field of perovskites currently is long-term instability. This has been shown be due to degradation pathways involving external factors, such as water, light, and oxygen,7 and also as a result of intrinsic instability, such as degradation upon heating, because of the properties of the material.4 For an overview of the causes of perovskite degradation, see Ossila’s guide.

Several strategies have been proposed in order to improve stability, most successfully by changing component choice. Using mixed-cation systems (for example by including inorganic cations such as rubidium or caesium) has been shown to improve both stability and efficiency.8 The first perovskite cells to exceed 20% efficiency used a mixed organic cation system,9 and many of the highest-efficiency systems published recently use inorganic components. Movement towards hydrophobic, UV-stable interfacial layers has also improved stability – for example by replacing TiO2, which is susceptible to UV degradation, with SnO2.10

Stability has also been improved through use of surface passivation11 and by combining 2D-layered (Ruddlesden-Popper) perovskites (which show better intrinsic stability, but poorer performance) with conventional 3D perovskites.12 These efforts (along with factors such as better encapsulation)13 have vastly improved the stability of perovskites since their initial introduction, and lifetimes are well on their way to meeting industrial standards – with recent work showing cells able to withstand a 1000-hour damp heat test.14 For a more in-depth discussion of methods to improve perovskite stability, see Ossila’s guide.

Comparison between 2D & 3D perovskite structure
Conventional 3D perovskite (left) compared to a generic 2D perovskite structure (right).

Another issue yet to be fully addressed is the use of lead in perovskite compounds. Though it is used in much smaller quantities than those currently present in either lead- or cadmium-based batteries, the presence of lead in products for commercial use is problematic.  Concerns still remain about exposure to toxic lead compounds (through leaching of the perovskite into the environment), and some studies have suggested large-scale implementation of perovskites would require complete containment of degradation products.15 In contrast, other life cycle assessments have found the toxicity impact of lead to be negligible in comparison to other materials in the cell (such as the cathode).5

There is also potential for a lead alternative to be used in perovskite solar cells (such as tin-based perovskites), but the power conversion efficiency of such devices is still significantly behind lead-based devices, with the record for a tin-based perovskite currently standing at 9.0%.16 Some studies have also concluded that tin may actually have a higher environmental toxicity than lead,17 and other less toxic alternatives are required.

Another major issue in terms of performance is the current-voltage hysteresis commonly seen in devices. The factors impacting hysteresis are still under debate, but it is most commonly attributed to mobile ion migration in combination with high levels of recombination.18 Methods to reduce hysteresis include varying cell architecture, surface passivation, and increasing lead iodide content,19 as well as general strategies to reduce recombination.


Graph showing current-voltage hysteresis in perovskite solar cells
An approximation of current-voltage hysteresis often encountered in perovskite solar cells.

To enable a truly low cost-per-watt, perovskite solar cells need to have achieved the much-heralded trio of high efficiency, long lifetimes, and low manufacturing costs. This has not yet been achieved for other thin-film technologies, but perovskite-based devices currently demonstrate enormous potential for achieving this.



Fabrication and Measurement of Perovskite Solar Cells

Although perovskites come from a seemingly different world of crystallography, they can be incorporated very easily into a standard OPV (or other thin-film) architecture. The first perovskite solar cells were based on solid state dye-sensitised solar cells (DSSCs), and so used a mesoporous TiO2 scaffold. Many cells since have followed this template or used an Al2O3 scaffold in a ‘meso-superstructured’ architecture, but the high temperature steps required for manufacture, and UV instability of TiO2, led to introduction of a ‘planar’ architecture similar to other thin-film cells. After several years of lagging behind mesoporous cells in terms of efficiency, planar perovskites are now almost as efficient.20


Structures of perovskite cells (conventional, inverted planar, mesoporous)
Generic structures of conventional/inverted planar and mesoporous (conventional) perovskite cells.

The perovskite film itself is typically processed by either vacuum or solution methods. Film quality is very important. Initially, vacuum-deposited films gave the best devices, but this process requires the co-evaporation of the organic (methylammonium) component at the same time as the inorganic (lead halide) components, necessitating specialist evaporation chambers that are not available to many researchers. As a result, there have been significant efforts into improving solution-processed devices, as these are simpler and allow for low-temperature processing, and these now equal vacuum-deposited cells in terms of efficiency.21

Typically, the active layer of a perovskite solar cell is deposited via either a one or two-step process. In the one-step process, a precursor solution (such as a mix of CH3NH3I and PbI2) is coated that then converts to the perovskite film upon heating. A variation on this is the ‘antisolvent’ method, in which the precursor solution is coated in a polar solvent, and then quenched during the spin coating process by a non-polar solvent. Precise timings of the quench and volumes of the quenching solvents is required to give the optimal performance. To help with this, we built the Ossila Syringe Pump, which has allowed us to use this quenching process to push in-house power conversion efficiency values over 16%.

In the two-step process, the metal halide (such as PbI2) and organic components (such as CH3NH3I) are spin-coated in separate, subsequent films. Alternatively, metal halide films can be coated and annealed in a chamber filled with the organic component vapour, known as ‘vacuum-assisted solution process’ (VASP).22


Step-by-step approximation of anti-solvent quenching method for coating perovskite cells
An approximation of the anti-solvent quenching method often used to coat perovskites in a one-step process from a precursor solution.

Most state-of-the-art perovskites are based on a transparent conducting oxide/ETL/Perovskite/HTL/metal structure, where ETL and HTL refer to electron-transport and hole-transport layers respectively. Typical hole-transport layers include Spiro-OMeTAD or PEDOT:PSS, and typical electron-transport layers include TiO2 or SnO2. Understanding and optimising the energy levels and interactions of different materials at these interfaces offers a very exciting area of research that is still under discussion.

The main issues for practical device fabrication of perovskite solar cells are film quality and thickness. The light-harvesting (active) perovskite layer needs to be several hundred nanometres thick – several times more than for standard organic photovoltaics, and creating such thick layers with high uniformity can be difficult. Unless the deposition conditions and annealing temperature are optimised, rough surfaces with incomplete coverage will form. Even with good optimisation, there will still be a significant surface roughness remaining. Therefore, thicker interface layers than might normally be used are also required. Improvements to film quality have been achieved through a variety of methods. One such method is the addition of small amounts of acids, such as hydroiodic or hydrobromic acid, previously discussed in a post about the purity of MAI vs lead chloride solubility, or excess of lead iodide precursor.19

Through extensive research efforts, efficiencies of over 22% have been achieved using spin coating,23 and high efficiencies have also been achieved using other solution-processing techniques (such as slot-die coating).24 This suggests that large-scale solution processing of perovskites is very feasible.



The Future of Perovskites

Future research into perovskites is likely to focus on the reduction of recombination through strategies such as passivation and reduction of defects, as well as boosting efficiency through inclusion of 2D perovskites and better-optimised interface materials. Charge-extraction layers are likely to move away from organic materials to inorganic, to improve both efficiency and stability. Improving stability and reduction in the environmental impact of lead are likely to both continue to be significant areas of interest.

Whilst the commercialisation of standalone perovskite solar cells still faces obstacles in terms of fabrication and stability, their use in tandem c-Si/perovskite cells has progressed rapidly (with efficiencies above 25% achieved)25 and it is likely that perovskites will first see the PV market as part of this structure. Beyond solar, there remains significant potential for use of perovskites in other applications, such as light-emitting diodes26 and resistive memories.27



Perovskite Fabrication Video Guide

For those just beginning their perovskite research, we have produced a video guide demonstrating the entire process of fabricating and measuring perovskite photovoltaics. In our own labs, we have reached efficiencies in excess of 11% using this particular fabrication routine. The video below features an older, discontinued model of the Ossila Spin Coater - to see the current model, you can visit the product page here.

 


Ossila Products for Perovskite Solar Cells

Ossila's award-winning Solar Cell Prototyping Platform delivers exemplary scientific application and impact in solar cell research. It is a coherent collection of substrates, materials, and test equipment as part of a high-performance standard photovoltaic reference architecture. It enables researchers to produce high quality, fully-functional solar cells that can be used as a reliable baseline.

As researchers and scientists ourselves, we understand how time-consuming it is to gain expertise over all the materials, processes, and techniques required to produce a high-quality device – and how despite your best efforts, it can sometimes lead to inconsistent and non-reproducible results.

We developed this platform with the aim of enabling you to focus on your research (instead of designing/sourcing all your own components) and replicate a performance baseline. A significant benefit of this platform is the provision of pre-patterned ITO substrates and high-throughput processing equipment – resulting in a significant increase in your production rate for solar cell devices – thus helping you collect more data, much quicker. As such, more types of new materials or architecture variations can be tested and more statistical data can be collected - ensuring consistency and accuracy.

At the most basic level, most perovskite-based solar cells are based upon a transparent conductive oxide coated glass substrate with evaporated metal cathode and top encapsulation. As such, our existing substrate infrastructure and perovskite materials are already being used in high-performance solution-processed perovskite devices. Our standard encapsulation epoxy is also perfectly suited for laminating glass or other barrier layers – as used in Snaith's 2014 Nature paper.


The Ossila Spin Coater is routinely used for the deposition of our interface and active layers with high accuracy and simple operation.

A very useful companion to the Spin Coater (pictured above) is the Ossila Syringe Pump. It can be used for automatic dispensing and quenching of our perovskite layers to obtain high-quality films. Our academic colleagues have also made some exciting progress on solution-processed perovskite solar cells via spray coating onto our standard substrates. Furthermore, perovskite solar cells are being characterised using the Ossila Solar Cell I-V Test System, which automatically calculates device metrics and can perform stability measurements.


Perovskite Ink
I101 Perovskite ink available from Ossila. It is packaged as 10 individual vials containing 0.5 ml of solution. This is capable of coating up to 160 substrates. I101 can also be bought in bulk (30 ml), with a 25% discount compared to our standard order sizes.

Over recent months we have also worked with our academic collaborators to bring more perovskite-based products to market, including: High-purity Methylammonium Iodide, Methylammonium Bromide, Formamidinium Iodide, and Formamidinium Bromide. We have also released our first set of perovskite inks, the first of these is I101(MAI:PbCl2), is designed to be processed in air and has demonstrated efficiencies in our labs up to 11.7%. Our second ink, I201(MAI:PbCl2:PbI2) is formulated to be processed in a nitrogen atmosphere, and so far we have seen efficiencies up to 11.8%. Both inks are designed to help our customers reach high efficiencies incredibly quickly when first starting out with their perovskites research. We include optimised processing routines with both inks to maximise results.

We will continue to develop our range of perovskite products in order to simplify fabrication and measurement, while also developing better characterisation tools and techniques. As always, please contact us if you have any questions.



References

(please note that Ossila has no formal connection to any of the authors or institutions in these references):
  1. NREL’s ‘Best Research-Cell Efficiencies’. Available at: https://www.nrel.gov/pv/assets/pdfs/pv-efficiencies-07-17-2018.pdf. (Accessed: 14th December 2018)
  2. Efficient Hybrid Solar Cells Based on Meso-Superstructured Organometal Halide Perovskites., M. M. Lee et al., Science, Vol. 338, p643–647 (2012).
  3. Lead iodide perovskite sensitized all-solid-state submicron thin film mesoscopic solar cell with efficiency exceeding 9%., H.-S. Kim et al., Scientific Reports, Vol 2 (2012).
  4. Present status and future prospects of perovskite photovoltaics. H. J. Snaith, Nature Materials, Vol. 17, p372–376 (2018).
  5. Perovskite solar cells: An integrated hybrid lifecycle assessment and review in comparison with other photovoltaic technologies., T. Ibn-Mohammed et al., Renewable & Sustainable Energy Reviews, Vol. 80, p1321–1344 (2017).
  6. Solar cell efficiency tables (version 52), M. A. Green et al., Progress in Photovoltaics: Research and Applications, Vol 26, p427–436 (2018).
  7. Organometal halide perovskite solar cells: Degradation and stability., T. A. Berhe et al. Energy & Environmental Science., Vol. 9, p323–356 (2016).
  8. Cesium-containing triple cation perovskite solar cells: Improved stability, reproducibility and high efficiency., M. Saliba et al., Energy & Environmental Science, Vol. 9, p1989–1997 (2016).
  9. High-performance photovoltaic perovskite layers fabricated through intramolecular exchange., W. S. Yang et al., Science, Vol. 348, p2013–2017 (2015).
  10. Tailored interfaces of unencapsulated perovskite solar cells for >1,000 hour operational stability., J. A. Christians et al., Nature Energy, Vol. 3, p68–74 (2018).
  11. Potassium-and rubidium-passivated alloyed perovskite films: Optoelectronic properties and moisture stability., M. Abdi-Jalebi et al., ACS Energy Letters, Vol. 3, p2671–2678 (2018).
  12. Efficient ambient-air-stable solar cells with 2D-3D heterostructured butylammonium-caesium-formamidinium lead halide perovskites., Z. Wang et al., Nature Energy, Vol. 2, p1–10 (2017).
  13. High-Performance Multilayer Encapsulation for Perovskite Photovoltaics, M. Wong-Stringer et al., Advanced Energy Materials, Vol. 8, p1–11 (2018).
  14. 23.6%-Efficient Monolithic Perovskite/Silicon Tandem Solar Cells With Improved Stability., K. A. Bush et al., Nature Energy, Vol. 2, p1–7 (2017).
  15. Toxicity of organometal halide perovskite solar cells., A. Babyigit et al., Nature Materials, Vol. 15, p247–251 (2016).
  16. Highly Reproducible Sn-Based Hybrid Perovskite Solar Cells with 9% Efficiency., S. Shao et al., Advanced Energy Materials., Vol. 8 (2018).
  17. Assessing the toxicity of Pb-and Sn-based perovskite solar cells in model organism Danio rerio., A. Babayigit et al, Scientific Reports, Vol. 6, p1–11 (2016).
  18. Hysteresis Index: A Figure without Merit for Quantifying Hysteresis in Perovskite Solar Cells., S. N. Habisreutinger, ACS Energy Letters, p2472–2476 (2018).
  19. Understanding how excess lead iodide precursor improves halide perovskite solar cell performance, B. Park, et al., Nature Communications., Vol. 9, p1–8 (2018).
  20. Enhanced charge carrier mobility and lifetime suppress hysteresis and improve efficiency in planar perovskite solar cells., S. H. Turren-Cruz et al., Energy & Environmental Science, Vol. 11, p78–86 (2018).
  21. How to Make over 20% Efficient Perovskite Solar Cells in Regular (n-i-p) and Inverted (p-i-n) Architectures., M. Saliba et al., Chemistry of Materials, Vol. 30, p4193–4201 (2018).
  22. Perovskites photovoltaic solar cells: An overview of current status., P. Tonui et al., Renewable & Sustainable Energy Reviews, Vol. 91, p1025–1044 (2018).
  23. Iodide management in formamidinium-lead-halide-based perovskite layers for efficient solar cells., W. S. Yang et al. Science, Vol. 356, p1376–1379 (2017).
  24. Roll-to-Roll Printing of Perovskite Solar Cells., B. Dou et al., ACS Energy Letters, Vol. 3, p2558–2565 (2018).
  25. Minimizing Current and Voltage Losses to Reach 25% Efficient Monolithic Two-Terminal Perovskite-Silicon Tandem Solar Cells., K. A. Bush et al., ACS Energy Letters., Vol. 3, p2173–2180 (2018).
  26. Solution-processed perovskite light emitting diodes with efficiency exceeding 15% through additive-controlled nanostructure tailoring, M. Ban et al., Nature Communications, Vol. 9, p1–10 (2018).
  27. Resistive Switching Behavior in Organic-Inorganic Hybrid CH3NH3PbI3-xClxPerovskite for Resistive Random Access Memory Devices., E. J. Yoo et al., Advanced Materials, Vol. 27, p6170–6175 (2015).

Further Reading:

Recent Perovskite Solar Cell Reviews:

A general perovskite review by Henry Snaith:

Hot Topic Perovskite Papers: