One Vapor Reaction Builds a Self-Sorting Perovskite Film — 3D on Top, Protective 2D Below
Two process knobs — a chemical ratio and a reaction temperature — are enough to turn a perovskite film’s color, and its internal structure, from yellow to dark brown.
Three-dimensional (3D) perovskite solar cells have already crossed 25% power conversion efficiency, but long-term stability against moisture and oxygen remains the main thing standing between the lab and commercialization. Two-dimensional (2D) layered perovskites — built by adding a large “spacer” cation that slices the 3D crystal into stacked layers — are far more resistant to degradation, but that stability trade-off usually comes at the cost of the charge-transport performance that makes 3D perovskites so efficient in the first place. The obvious answer is to combine both in one film: a stable 2D layer doing the protecting, a high-performance 3D layer doing the work. The hard part is controlling exactly where each phase ends up.
The Obstacle — Building a Graded Film Without a Multi-Step Process
Most attempts to combine 2D and 3D perovskites in one film rely on multiple deposition steps or careful interface engineering to place each phase where it’s needed. What was missing was a simpler route: a single process capable of producing a film that naturally grades itself, with the moisture-resistant 2D phase where it’s protective and the high-performance 3D phase where light absorption and charge extraction actually happen.
The Innovation — Letting the Reaction Do the Sorting
In a study published in The Journal of Physical Chemistry C (Chen, Shiu, Hsu, Mundt, Hung, Ohigashi, Li & Chen, 2021, DOI: 10.1021/acs.jpcc.1c07795), researchers at National Cheng Kung University and National Chi Nan University — working with Taiwan’s National Synchrotron Radiation Research Center, Germany’s Fraunhofer Institute for Solar Energy Systems (ISE), and Japan’s UVSOR Synchrotron at the Institute for Molecular Science — used a low-pressure vapor-assisted solution process (LP-VASP) to build 2D/3D mixed perovskite films with a built-in vertical phase gradient, from a single reaction step.
The process starts by spin-coating a film of lead iodide (PbI2) loaded with a high concentration of phenylethylammonium iodide (PEAI) — the large cation that forms 2D layered perovskite. That film is then heated under vacuum alongside methylammonium iodide (MAI) powder, which sublimates into methylamine gas that diffuses into the film and reacts with it. As the reaction proceeds, the PEA cations themselves progressively evaporate out of the film as gas, while methylamine gas diffuses in — so the region of the film that reacts first and most completely converts furthest toward the 3D phase, while the least-reacted region stays closer to the original 2D phase.
The reaction itself does the sorting: as PEA cations evaporate out and methylamine diffuses in, the film converts from the surface inward, leaving the 3D phase on top and the 2D phase at the bottom.
Key Results — Reading the Gradient in Light, Color, and Position
The team built the case in layers: first show the phases exist and can be controlled, then confirm they’re spatially organized exactly where the mechanism predicts.
Steady PL Fingerprint Reveals Four Coexisting 2D Phases
Measured on ProTrusTech’s MRI system (532 nm diode laser), the steady-state PL spectrum of the PEAI/PbI2 = 4 film at 125°C shows distinct emission peaks corresponding to the n = 1, 2, 3, and 4 layered perovskite phases — direct evidence that multiple 2D phases coexist in the same film. This is also the exact condition used for every mapping and surface measurement later in the paper.
Push the Temperature Higher and the Film Collapses to Pure 3D
At 135°C, the same PEAI/PbI2 = 4 film shows only a single PL peak near 780 nm — pure 3D MAPbI3. Between 120°C and 135°C, reaction temperature alone shifts the film from a multi-phase 2D mixture to a single 3D phase, as the PEA cations progressively evaporate out at higher heat.
Film Color Visibly Tracks the Same Conversion
Photographed across the same temperature range, the films shift from yellow to maple to dark brown — a simple visual proxy for the 2D-to-3D phase conversion the PL spectra measure quantitatively.
Confocal PL Mapping Shows the Phases Are Spatially Organized, Not Random
Mapping the PL emission specific to each n-value phase across the film surface shows the n = 1 phase concentrated in a branch-like pattern, with the higher-n phases distributed more uniformly around it — the phases aren’t just present together, they occupy distinct regions.
Surface Topography and Potential Confirm 3D Dominates the Top
SEM shows an uneven, flake-stacked surface; Kelvin probe force microscopy (KPFM) shows a surface work function matching 3D MAPbI3 — independent confirmation that the outermost layer of the film is the high-performance 3D phase.
An Independent Synchrotron Technique Confirms the Same Surface Composition
Scanning transmission X-ray microscopy (STXM), performed at Japan’s UVSOR-III synchrotron, chemically mapped the carbon signature across the film and reached the same conclusion by a completely different physical method: the surface composition matches 3D MAPbI3.
Why ProTrusTech Is Featuring This Research
The MRI 532 carried this paper’s phase-identification story end to end — from the ensemble PL spectra in Figure 1 to the spatial confocal mapping in Figure 4 — without switching instruments. That’s the kind of reliability we’ve seen matter across the perovskite and PL characterization work we’ve featured: researchers get consistent, trustworthy data whether they need a quick spectrum or a full spatial map, all on one system they already trust.
Engineering the Solution
What makes this approach practical is that it’s genuinely single-step: one spin-coating pass, one vapor-reaction step, no additional depositions or interface-engineering procedures. The compositional gradient emerges from tuning just two process parameters — the PEAI/PbI2 molar ratio and the reaction temperature — a small, controllable knob set rather than a complex multi-material engineering problem.
Featured Instruments:
Why This Matters
The broader idea here isn’t just that 2D and 3D perovskites can coexist in one film — it’s that a compositional gradient, not just a bulk mixture, can emerge from a single reaction step by exploiting a molecule’s own volatility. Instead of fighting to hold a target composition uniform throughout the film, this process lets the PEA cation’s evaporation do the differentiating work as the reaction proceeds. That reaction-driven self-grading principle isn’t inherently perovskite-specific — it’s a strategy worth considering anywhere a volatile spacer component naturally depletes over the course of a reaction.
What’s Next
- Device-level validation. The authors note this gradient structure has great potential for stable perovskite solar cells and efficient light-emitting diodes — the next step is building and testing actual devices.
- Finer control of the gradient profile. The current work shows the gradient exists and roughly where each phase sits; sharper depth-profiling and tighter control over the gradient’s steepness remain open.
- Scaling LP-VASP to larger areas. As with most vapor-phase perovskite processes, moving from small lab-scale films to production-relevant areas is an open manufacturing question.
Working on perovskite phase identification, thin-film crystallization monitoring, or PL-based characterization of your own? Get in touch with our team.
Key Takeaways
- A single low-pressure vapor-assisted reaction step builds 2D/3D mixed perovskite films with a built-in vertical gradient — 2D layered phases at the bottom, 3D MAPbI3 at the surface.
- Steady-state PL spectra, measured on a ProTrusTech MRI system, identified four coexisting 2D layered phases (n = 1–4) plus the 3D phase by their distinct emission wavelengths.
- Two process parameters — PEAI/PbI2 ratio and reaction temperature — control which phase dominates, visibly tracked by the film’s color shifting from yellow to dark brown.
- Confocal PL mapping, two-photon PL mapping, KPFM, and STXM independently confirmed the same top-3D/bottom-2D spatial gradient.
- The authors point to stable perovskite solar cells and efficient LEDs as the target applications for this gradient-structure approach.
