1. What solution coating is, in three points

Solution coating means spreading a liquid in which the perovskite raw materials are dissolved (the precursor ink) onto a substrate, then driving off the solvent to leave a crystalline film. One reason this material attracted so much attention is that the absorber can be made without the high-temperature crystal growth or high-vacuum equipment that silicon solar cells require.

• What is happening: as the solvent leaves, the solute becomes supersaturated, nuclei form and crystals grow. The aim is to make that sequence happen uniformly across the whole film. The NREL review explains that uncontrolled drying makes crystals grow unevenly into a dendritic (branching) morphology that fails to cover the substrate and creates shunt paths in the cellSourced
• The lab standard: spin coating is the most common deposition method, but it is described as generally limited to substrates under about 4 inches, with most of the precursor ink (more than 90%) wastedSourced
• Methods for manufacturing: blade coating, slot-die coating, spray coating, inkjet printing and screen printing are all being studied, with roll-to-roll (R2R) production in viewSourced

2. Four ways to coat: spin, blade, slot-die and inkjet

All Sourced (NREL, Li et al. 2018 [Reference 1]). The same review points out that blade and slot-die coating both spread the wet film with a moving meniscus (the liquid edge).

In slot-die coating, thickness is set by how much you pump

Slot-die coating is a pre-metered method: all the liquid pumped to the head ends up in the film. The NREL review by Park and Zhu summarises that the final film thickness is proportional to the flow rate and the solids concentration, and inversely proportional to the substrate (web) speed, the coating width and the density of the dry filmSourced. The review also notes that slot-die coating was invented by Beguin of Eastman Kodak in 1951, and that it was first applied to perovskites in 2015Sourced.

A 2015 report (Hwang et al.) described cells in which every layer except the electrodes was printed by slot-die coating, reaching a power conversion efficiency of up to 11.96%, and showed that the process could be used for R2R productionSourced.

In blade coating, thickness responds to speed in opposite directions

Blade coating has a distinctive behaviour pointed out in the Park and Zhu review. At low speed, solvent evaporates at the meniscus and the solids gather near the contact line, so the faster you go the thinner the film (the evaporation regime). At high speed, viscous drag pulls out a wet film that dries afterwards, so the faster you go the thicker the film (the Landau–Levich regime)Sourced. On the same machine, the rule governing thickness flips depending on the speed range.

In inkjet printing, thickness is set by drop spacing

In 2018 Mathies and colleagues at KIT (Karlsruhe Institute of Technology) inkjet-printed a triple-cation perovskite containing 10% Cs, reporting that film thickness could be controlled between 175 and 780 nm simply by changing the drop spacing. They reported a power conversion efficiency of 12.9% for stabilised output at constant voltageSourced. Because the pattern is laid down digitally, the method suits products with free-form shapes or applications where appearance matters (our commentary).

3. Controlling crystallisation: when and where supersaturation happens

The NREL review divides the ways of avoiding the dendritic morphology that becomes a problem at large area into two groups: physical methods that promote supersaturation, and chemical additives that retard crystal growthSourced. The logic runs like this: remove the solvent fast → supersaturation rises sharply → many nuclei form, and no single nucleus has time to grow large → the film comes out flat and denseSourced.

(1) Antisolvent dripping: the turning point of 2014

Partway through spin coating, a solvent that does not dissolve perovskite (an antisolvent) is dripped on to make the film precipitate all at once. Two groups reported the method almost simultaneously in 2014.

• Jeon et al. (Nature Materials): coated from a mixed solvent of γ-butyrolactone (GBL) and dimethyl sulfoxide (DMSO) and dripped toluene on. Passing through a CH3NH3I–PbI2–DMSO intermediate phase, the process gave extremely uniform, dense films and a certified efficiency of 16.2% with no hysteresisSourced
• Xiao et al. (Angewandte Chemie): exposed a spin-coated DMF solution of CH3NH3PbI3 to chlorobenzene immediately after coating to induce crystallisation. The flat films consisted of large, micron-scale grains, and the authors reported an average of 13.9 ± 0.7% and a steady-state efficiency of 13%Sourced

The NREL review gives the requirement for an antisolvent as being miscible with the precursor solvents (DMF, DMSO, GBL and so on) while barely dissolving the perovskite itself, and lists toluene, chlorobenzene, ethyl ether, ethyl acetate, hexane and anisole as examplesSourced.

(2) Gas quenching: blowing dry gas across the film

Instead of an antisolvent, a stream of nitrogen or another gas strips the solvent off rapidly. The NREL review notes reports of gas flow speeding up crystallisation during spin coating, and examples where this was carried over to slot-die coating with an air knife to assist dryingSourced.

In 2023 Geistert and colleagues at KIT aimed an inclined slit nozzle delivering nitrogen at slot-die-coated wet films and refined the conditions while watching the drying in situ with a camera. They showed how to adjust the nozzle to prevent gas backflow, and reported up to 18.6% for cells, 17.2% for a 50 × 50 mm² mini-module, and only very small losses when scaling up the areaSourced. Gas quenching, in other words, is a step in which gas flow strength, drying rate and the position where crystallisation begins decide the film morphology.

(3) Vacuum flash: a sudden drop in pressure

In 2016 Li and colleagues at EPFL (the Swiss Federal Institute of Technology in Lausanne) reported in Science a "vacuum flash" in which the freshly coated substrate is suddenly exposed to reduced pressure to pull the solvent out. They achieved a best of 20.5% and a certified 19.6% for cells with an aperture area above 1 cm², noting that the previous certified record for comparable areas had been 15.6%Sourced.

(4) Two-step deposition: PbI2 first, organic salt second

In 2013 Burschka et al. (Nature) used sequential deposition, first loading PbI2 into nanoporous TiO2 and then converting it by exposure to a CH3NH3I solution, improving morphological control and reproducibility and reaching an efficiency of about 15%Sourced. The NREL review cites as advantages of the two-step method that uniform PbI2 films are relatively easy to make over large areas, and the volume expansion on conversion to perovskite fills pinholes, and says the method is widely used with slot-die coating, electrodeposition and spray coatingSourced.

4. A materials engineer's view (1): the ink is not a solution but a piece of coordination chemistry

5. Choosing solvents: ones that coordinate, and ones that leave quickly

When you try to coat large areas fast, solvent choice sets the upper limit on coating speed. The framework laid out in 2019 by Deng and colleagues at the University of North Carolina (Science Advances) makes this easy to see.

• DMSO and DMF coordinate strongly to Pb²⁺, so they dissolve PbI2, but they dry slowly at room temperature. Even with high-speed blade coating the wet film stays wet, and it takes minutes to tens of minutes to dry and crystalliseSourced
• 2-methoxyethanol (2-ME) and acetonitrile (ACN) evaporate easily and barely coordinate to Pb²⁺. On their own, however, they give films with poor crystallinity and poor adhesion to the substrateSourced
• The answer was to use the volatile, non-coordinating solvent as the main solvent and add a small amount of the less volatile, coordinating solvent (DMSO). This achieved both fast drying and large grains at room temperatureSourced

With this formulation, Deng and colleagues blade-coated in ambient air at room temperature at 99 mm/s, obtaining a certified efficiency of 16.4% for a module with an aperture area of 63.7 cm²Sourced. The paper also records that drying was accelerated right after coating with a nitrogen knife (a stream of nitrogen applied at a set distance behind the blade), and that coating faster than 10 mm/s leaves a wet film a few micrometres thickSourced.

6. Scaling up: antisolvent baths, gas flow and vacuum

What gets in the way when a technique developed on spin coating is moved to large areas? Here are the key points from the NREL reviewSourced.

All Sourced (NREL, Li et al. 2018 [Reference 1]). The Park and Zhu review [Reference 2] also cites a report that, in slot-die coating, single-step coating onto a substrate preheated to about 65 °C followed by a cool air stream worked better than other conditions.

7. A materials engineer's view (2): a few tens of ppm of surfactant changes the coating

8. Our calculation: coating speed, output and the solvent to be evaporated

Citing Deng et al. 2018, the Park and Zhu review states that a coating speed of 180 m/h would give a production capacity of 236 MW a year on a line 1 m wideSourced. Let us check what assumptions lie behind that figure.

9. A materials engineer's view (3): solvents are also regulated substances

10. Open problems, and what this article could not confirm

(1) The best efficiencies still come from small spin-coated cells

The NREL review (2018) points out that mini-module efficiencies clearly lag behind the gains made in small-area cellsSourced. The Park and Zhu review (2020) likewise says that spin coating, used on small substrates (typically 2.5 × 2.5 cm), cannot be carried straight into commercialisation, and that materials and methods for large areas need to be developedSourced. The relationship between area and efficiency is covered in detail in our explainer on roll-to-roll processing and scale-up.

(2) Few devices are made with every layer coated

The NREL review says that most studies deposit only the absorber, or one or two layers, by scalable methods, and that reports of fully printed devices are few. It also notes that the back electrode is mainly made by vacuum evaporationSourced. Charge transport layer materials and their deposition are covered in our explainers on charge transport layers and cell architecture, and on self-assembled monolayers.

(3) Actual coating conditions on production lines

Which coating methods, solvents and drying methods are used on actual production or pilot lines could not be confirmed from companies' primary sources within the scope of this article, so they are not stated. Company developments are covered in our company profilesNot yet confirmed.

(4) Efficiencies in papers are not measured on the same basis

The efficiencies quoted in this article differ from paper to paper in whether they are cells or modules, in area, in whether they are certified, and in whether they are steady-state or scanned. For example, 16.2% (Jeon et al.), 19.6% (Li et al.) and the 16.4% module (Deng et al. 2019) are certified values, while 18.3% and 19.7% (Ahn et al.) and 15.3% (Deng et al. 2018) were measured within the papersSourced. Lining the numbers up does not tell you which method is better. Differences in measurement conditions are covered in our explainer on conversion efficiency and performance metrics.

All Sourced (abstracts of each paper [References 3, 4, 5, 7, 8, 9 and 11]). Lining them up in a table is this article's own arrangement; it is not a comparison under identical conditions.

11. Glossary

12. References (primary sources)

1. Li Z, Klein TR, Kim DH, Yang M, Berry JJ, van Hest MFAM, Zhu K (NREL) “Scalable fabrication of perovskite solar cells”, Nature Reviews Materials 3, 18017 (2018) (DOE Public Access accepted manuscript) https://www.osti.gov/servlets/purl/1430821
2. Park NG, Zhu K “Scalable fabrication and coating methods for perovskite solar cells and solar modules”, Nature Reviews Materials 5, 333–350 (2020) (DOE Public Access accepted manuscript) https://www.osti.gov/servlets/purl/1605079
3. Jeon NJ et al. “Solvent engineering for high-performance inorganic–organic hybrid perovskite solar cells”, Nature Materials 13, 897–903 (2014) https://doi.org/10.1038/nmat4014
4. Xiao M et al. “A Fast Deposition-Crystallization Procedure for Highly Efficient Lead Iodide Perovskite Thin-Film Solar Cells”, Angewandte Chemie International Edition 53, 9898–9903 (2014) https://doi.org/10.1002/anie.201405334
5. Ahn N et al. “Highly Reproducible Perovskite Solar Cells with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct of Lead(II) Iodide”, Journal of the American Chemical Society 137, 8696–8699 (2015) https://doi.org/10.1021/jacs.5b04930
6. Burschka J et al. “Sequential deposition as a route to high-performance perovskite-sensitized solar cells”, Nature 499, 316–319 (2013) https://doi.org/10.1038/nature12340
7. Li X et al. “A vacuum flash–assisted solution process for high-efficiency large-area perovskite solar cells”, Science 353, 58–62 (2016) https://doi.org/10.1126/science.aaf8060
8. Deng Y et al. “Surfactant-controlled ink drying enables high-speed deposition of perovskite films for efficient photovoltaic modules”, Nature Energy 3, 560–566 (2018) https://doi.org/10.1038/s41560-018-0153-9
9. Deng Y et al. “Tailoring solvent coordination for high-speed, room-temperature blading of perovskite photovoltaic films”, Science Advances 5, eaax7537 (2019) (full text on PMC) https://pmc.ncbi.nlm.nih.gov/articles/PMC6897546/
10. Hwang K et al. “Toward Large Scale Roll-to-Roll Production of Fully Printed Perovskite Solar Cells”, Advanced Materials 27, 1241–1247 (2015) https://doi.org/10.1002/adma.201404598
11. Geistert K et al. (KIT) “Controlling Thin Film Morphology Formation during Gas Quenching of Slot-Die Coated Perovskite Solar Modules”, ACS Applied Materials & Interfaces (2023) https://doi.org/10.1021/acsami.3c11923
12. Mathies F et al. (KIT) “Inkjet-Printed Triple Cation Perovskite Solar Cells”, ACS Applied Energy Materials 1, 1834–1839 (2018) https://doi.org/10.1021/acsaem.8b00222
13. European Commission “Commission Regulation (EU) 2021/2030 amending Annex XVII to Regulation (EC) No 1907/2006 (REACH) as regards N,N-dimethylformamide”, EUR-Lex https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32021R2030

13. Claim-to-source audit

Last updated 25 September 2026. Sources are limited to primary material (peer-reviewed papers, NREL accepted manuscripts published through DOE Public Access, and EU regulation). Because the article includes structural readings and materials and process interpretations, those are marked as Commentary and kept separate from sourced fact. The coating conditions actually used on production lines, and the regulatory status of alternative solvents, are not stated because this article did not confirm or investigate them. The efficiencies quoted here differ from paper to paper in area, certification and measurement method, and cannot be compared directly. All figures are explanatory concept graphics. Figs. 1, 2, 3 and 5 are vector drawings; the hero image and Fig. 4 are AI-generated images, and none of them shows real equipment, a micrograph or a physical product.