One New Cation Additive Blocks Perovskite’s Efficiency-Killing Yellow Phase — and Lifts Output to 14.5%

The additive that saved this perovskite solar cell never actually joined its crystal structure.

Formamidinium lead iodide (FAPbI₃) is one of the most promising light absorbers in perovskite photovoltaics — its bandgap is close to ideal for single-junction solar cells, and the broader perovskite solar cell (PSC) field has already pushed past 25% power conversion efficiency (PCE) in record devices. But FAPbI₃ has a persistent problem: at room temperature, it doesn’t stay in its photoactive black (α) phase. It readily slips into δ-FAPbI₃, a non-photoactive yellow phase that simply doesn’t generate current — and once it forms, the device underneath it is compromised before a single J-V curve is even measured.

The Obstacle — Stabilizing the Right Phase Without Changing the Recipe

Researchers have long known that additives and alternative cations can push FAPbI₃ toward its stable black phase, but most of the working strategies come with a tradeoff: they either require pre-synthesized FAPbI₃ powder as a precursor, or they use additives that get incorporated directly into the perovskite’s A-site — meaning the fix also changes the material’s composition, and with it, unpredictable downstream effects on band position, absorption, and long-term behavior. What was missing was an additive that could suppress the yellow phase during crystallization without becoming part of the finished film at all.

The Innovation — An Additive That Works and Then Leaves

In a study published in Materials Advances (Raifuku, Chiang, Hou, Li, Lin, Lin, Shyue & Chen, 2021, Mater. Adv., 2, 2272–2277, DOI: 10.1039/d0ma00945h), researchers at National Cheng Kung University, Academia Sinica, and National Chi Nan University — working through NCKU’s Hierarchical Green-Energy Materials (Hi-GEM) Research Center — synthesized a new compound, formamide iodide (FoAI), by mixing formamide with hydroiodic acid. Structurally, FoAI closely resembles formamidinium iodide (FAI), the well-known cation source for FAPbI₃ itself, so the natural hypothesis was that FoAI would behave like another A-site cation.

It doesn’t. Using time-of-flight secondary ion mass spectrometry (ToF-SIMS), X-ray diffraction (XRD), and Fourier-transform infrared spectroscopy (FT-IR), the team confirmed FoAI is essentially absent from the finished perovskite films — it isn’t incorporated into the crystal structure at all. Instead, the evidence pointed to FoAI forming an adduct or complex with lead iodide (PbI₂) in solution, improving its solubility and altering how the film crystallizes as it dries — without changing what the film is made of once it’s done.

Without FoAI, FAPbI₃ crystallizes into the non-photoactive yellow δ-phase. With FoAI, an FoAI–PbI₂ adduct forms in solution and redirects crystallization toward the photoactive black α-phase — and FoAI itself is gone from the finished film.

Key Results — A Cleaner Crystal Without a Different Composition

The case builds in three layers: show the device numbers improved, confirm the structural cause, then confirm the mechanism behind it. Three image cards carry that argument; two supporting results that don’t have a figure of their own follow in prose.

5 mol% FoAI Lifts PCE From 12.3% to 14.5% — Every Metric Moves Together

Adding 5 mol% FoAI to the FAPbI₃ precursor raised power conversion efficiency from 12.29% to 14.49% (best devices, reverse scan; 11.67% ± 0.63% to 13.90% ± 0.54% averaged). Short-circuit current density, open-circuit voltage, and fill factor all improved together — a genuine across-the-board gain, not a single-metric tradeoff.

XRD Confirms the Yellow Phase Is Suppressed, Not Masked

Films without FoAI show a clear δ-FAPbI₃ diffraction peak — the non-photoactive yellow phase. Films with FoAI show none: direct structural evidence the additive prevents the phase from forming, not just compensates for it.

PL on ProTrusTech’s MRI System Reveals Fewer Defects

Photoluminescence measurements (532 nm diode laser, 127 mW) show FoAI-added triple-cation films emitting more strongly than the control at 0.625–1.25 mol% loading — a signature of fewer non-radiative defect sites, consistent with the improved open-circuit voltage seen in the same devices.

Two more results round out the picture without a figure of their own. Dose-response wasn’t simply “more is better”: in triple-cation devices, only the lowest tested loading, 0.625 mol% FoAI, improved PCE over the control (17.58% ± 0.10% vs. 17.36% ± 0.22%); PCE declined slightly at 1.25%, 2.5%, and 5 mol%, an honest result the authors report without over-claiming a monotonic trend. And under continuous 1-sun illumination without encapsulation, FoAI-added devices held onto more of their initial output than the unprotected control (49.3% vs. 42.7% after 90 minutes) — a modest but consistent stability edge.

Why ProTrusTech Is Featuring This Research

Every photoluminescence measurement behind this paper’s crystal-quality argument — the data in Fig. 4(b) showing FoAI-added films emitting more strongly than the control — was collected on a ProTrusTech MRI system. PL intensity is the paper’s evidence for reduced defect density, which is in turn the paper’s explanation for why open-circuit voltage improved. That’s a load-bearing measurement, not a supporting figure, and it’s exactly the kind of defect- and crystallinity-sensitive characterization our systems are built for.

Engineering the Solution

FoAI’s practical appeal is that it’s a drop-in additive, not a process overhaul. It’s mixed directly into an existing FAPbI₃ or triple-cation precursor solution at a low mol% — no new deposition step, no new solvent system, no changes to the spin-coating or annealing protocol already used to fabricate these devices. Since it also doesn’t alter the final film’s composition, band position, or absorbance, it’s a lower-risk lever than swapping in a genuinely new cation.

Featured Instruments:

Why This Matters

This paper is a clean illustration of a broader principle in perovskite processing: an additive doesn’t have to end up in the finished structure to change how that structure forms. FoAI’s entire effect happens in solution — it improves PbI₂ solubility and redirects crystallization — and by the time the film is finished, FoAI itself is gone. That decouples additive engineering from cation engineering: you can tune nucleation and crystal growth kinetics without touching the final film’s composition.

What’s Next

• Clarifying the exact mechanism. The authors describe the FoAI–PbI₂ adduct hypothesis as the likely explanation but say further investigation is needed to confirm it directly.
• Better dose-response optimization. Since 0.625 mol% outperformed higher loadings in triple-cation devices, there’s an open question of where the real optimum sits.
• Extending encapsulated, longer-duration stability testing. The 90-minute unencapsulated test shows a real but modest advantage; longer and encapsulated testing isn’t yet reported.

Working on a perovskite film-quality or defect-characterization question of your own? Get in touch with our team.

Key Takeaways

• FoAI, a new formamide-based additive, suppresses δ-FAPbI₃ (the non-photoactive yellow phase) without being incorporated into the finished perovskite film.
• FAPbI₃ solar cells with 5 mol% FoAI reached 14.49% PCE (best device, reverse scan), up from 12.29% for the additive-free control — with J_SC, V_OC, and fill factor all improving together.
• Photoluminescence measurements on a ProTrusTech MRI system showed FoAI-added films emitting more strongly, evidence of fewer defects and better crystal quality.
• In triple-cation devices, only the lowest tested FoAI loading (0.625 mol%) improved performance — higher loadings gave no further gain.
• FoAI-added devices retained more of their output (49.3% vs. 42.7%) after 90 minutes of unencapsulated 1-sun illumination.

Featured Publications: