Humidity-Induced Photoluminescence Shift and Structural Disorder in CH3NH3PbI3 Microwires

Abstract

We examined the long‑term effect of atmospheric humidity on self‑assembled CH₃NH₃PbI₃ (MAPbI₃) microwires (MWs) using photoluminescence (PL), Raman spectroscopy, and X‑ray diffraction (XRD). Over 11 weeks, the MWs displayed a 21 nm red‑shift of the PL peak (from 759 nm to ~780 nm) and an overall intensity increase, despite partial decomposition into PbI₂ and a monohydrate phase. Raman bands shifted and new peaks appeared, indicating defect formation and lattice distortion. XRD confirmed lattice expansion and contraction consistent with water incorporation. These results link moisture‑induced shallow trap states and structural disorder to the optical response of MAPbI₃ MWs, providing insight for stabilizing perovskite‑based optoelectronics.

Background

Hybrid halide perovskites, notably MAPbI₃, have attracted intense research interest due to their low‑cost, solution‑processable fabrication and exceptional optoelectronic properties [1–3]. While thin films dominate photovoltaic research, low‑dimensional morphologies such as microwires, nanowires, and microdisks offer advantages: high surface‑to‑volume ratio, reduced grain boundaries, and superior charge transport, enabling photodetectors, lasers, and waveguides [4–15]. However, environmental stability—particularly against humidity—is a major challenge. Water vapor induces the formation of hydrated intermediates and ultimately decomposes MAPbI₃ into PbI₂ and volatile by‑products [20–26]. Existing studies have focused on thin films, leaving the humidity response of MWs largely unexplored. Understanding this response is critical for the reliable operation of MW‑based devices.

Experimental

Synthesis of MAPbI₃ Microwires

MAI was synthesized by reacting hydroiodic acid with methylamine, followed by evaporation, washing, and drying to yield white powder. MAPbI₃ MWs were fabricated via a one‑step solution self‑assembly method, depositing a 20 µL precursor solution (MAI:PbI₂ in DMF) onto a glass slide, then exposing it to dichloromethane vapor at 65 °C for 3 h. XRD‑ready samples used a higher precursor concentration (24.7 mg MAI:72.3 mg PbI₂ in 3 mL DMF).

Humidity Exposure

MWs were stored in an airtight container at ~20 °C. Initial exposure (weeks 1–4) reflected ambient RH (45 ± 5 % for 3 weeks, 55 ± 5 % for week 4). From week 5 onward, a salt‑saturated solution maintained 80 ± 2 % RH. Samples were only removed for measurement to avoid additional light or thermal effects.

Characterization

PL & Raman: Renishaw InVia, 633 nm (5 µW) for PL and 532 nm (16 µW) for Raman, 10 s acquisition, 50× objective, backscattering geometry.
XRD: PANalytical X’Pert Pro, Cu‑Kα (λ = 1.5418 Å), 40 kV/40 mA, 0.026° step, 0.2 s per step, 5°–70° 2θ.
SEM & Optical: Hitachi SU8010 for SEM; Olympus BX51 for optical imaging.

Results and Discussion

Morphology and Initial Structure

SEM and optical microscopy revealed long, straight MWs (lengths up to cm, widths 2–5 µm). XRD of fresh MWs matched the tetragonal MAPbI₃ phase with peaks at 2θ = 14.11°, 28.45°, 31.90°, and 40.48°, corresponding to (110), (220), (310), and (224) planes. Lattice constants a = b = 8.8703 Å, c = 12.6646 Å confirm the expected crystal structure.

Photoluminescence Evolution

PL spectra remained single‑peaked throughout exposure. The emission peak shifted progressively from 759 nm (fresh) to ~780 nm after 11 weeks— a 21 nm red‑shift. The bandgap decreased from 1.63 eV to 1.59 eV. Despite partial decomposition to PbI₂ and a monohydrate phase (bandgaps > 2.5 eV), the shift originates from the intact MAPbI₃ lattice. PL intensity rose from week 4 to week 9, indicating reduced non‑radiative recombination, likely due to passivation of deep trap states by water. The concurrent red‑shift suggests activation of shallow trap‑mediated radiative recombination, consistent with defect‑induced sub‑gap states [35–40].

Raman Spectroscopy

Initial Raman peaks at 111 cm⁻¹ and a shoulder near 75 cm⁻¹ shifted to 108 cm⁻¹ after 3 weeks at 45 % RH. At 80 % RH, a new band emerged at 95 cm⁻¹, and the 111 cm⁻¹ peak oscillated slightly, reflecting lattice stress and MA vacancy formation due to water infiltration [21, 41]. The appearance of PbI₂‑like Raman features indicates partial decomposition, yet the PL still corresponds to MAPbI₃, underscoring incomplete degradation.

X‑ray Diffraction

After 5 days at 80 % RH, perovskite peaks weakened, while PbI₂ reflections strengthened and new peaks at 8.54° and 10.54° emerged, attributable to the MAPbI₃·H₂O monohydrate [21, 24, 43]. After 14 days, perovskite peaks diminished further, but PbI₂ and hydrate peaks remained weak, indicating limited decomposition. Notable shifts: (202) plane moved from 24.50° to 24.28°, implying increased d₂₀₂; (004) and (220) planes shifted to higher angles, indicating reduced d-spacing. These distortions are consistent with water molecules occupying the lattice, tilting PbI₆ octahedra, and altering bond lengths, which in turn lower the bandgap [44–46].

Mechanistic Insight

Humidity induces two concurrent effects: (1) passivation of deep non‑radiative traps, boosting PL intensity; (2) creation of shallow traps and lattice distortion, producing a red‑shifted, lower‑energy emission. The combined influence of water‑induced defects and structural disorder explains the observed optical evolution and aligns with Raman and XRD evidence.

Conclusions

Humidity exposure transforms MAPbI₃ microwires by partially decomposing them to PbI₂ and a monohydrate phase while simultaneously enhancing PL intensity and red‑shifting the emission peak. Structural analysis links these changes to water‑induced lattice distortion and the formation of shallow trap states. Controlling humidity‑induced defects and lattice deformation could preserve optical performance, improving the durability of perovskite MW‑based optoelectronic devices.