Optimizing CH₃NH₃PbI₃ Morphology for Enhanced Perovskite Solar Cell Performance

Abstract

We investigated how varying methylammonium iodide (MAI) concentrations in a two‑step wet process influences the morphology and photovoltaic performance of CH₃NH₃PbI₃ perovskite films. The device architecture—glass/FTO/TiO₂‑mesoporous/CH₃NH₃PbI₃/spiro‑OMeTAD/Ag—was fabricated on 1.5 × 1.5 cm² FTO substrates. By systematically adjusting spin‑coating speed and annealing temperature, we optimized film quality and achieved a record power conversion efficiency (PCE) of 17.42 %, with an open‑circuit voltage (Voc) of 0.97 V, short‑circuit current density (Jsc) of 24.06 mA cm⁻², and fill factor (FF) of 0.747.

Background

Organic‑inorganic perovskites have emerged as leading candidates for next‑generation thin‑film photovoltaics, achieving PCEs that have climbed from 3.8 % to over 22 % in the past decade. Two‑step fabrication—first depositing PbI₂, then converting it to CH₃NH₃PbI₃ via MAI exposure—is favored for its superior film uniformity and scalability.

While extensive work has optimized precursor chemistry and deposition parameters, the specific impact of MAI concentration on grain growth, surface coverage, and ultimately device performance remains underexplored. Here, we systematically vary MAI levels (10 mg mL⁻¹ vs. 40 mg mL⁻¹) and characterize the resulting morphology, crystallinity, and optoelectronic properties to elucidate their influence on PCE.

Methods

FTO glass (1.5 × 1.5 cm²) was cleaned sequentially with acetone, ethanol, and deionized water, then dried under nitrogen. A 50‑nm TiO₂ compact layer was deposited via spray pyrolysis at 500 °C, followed by a mesoporous TiO₂ layer (Dyesol 18NR‑T) spin‑coated and annealed at 450 °C. PbI₂ (1 M in DMF, 70 °C) was spin‑coated at 7000 rpm; the film was then dipped in 10 mg mL⁻¹ or 40 mg mL⁻¹ MAI/IPA for 30 s, dried, and annealed (100–120 °C) for 20 min.

The hole‑transport layer (spiro‑OMeTAD, 0.170 M) was doped with LiTFSI (60 mM) and TBP (200 mM) and spin‑coated at 4000 rpm. A 100‑nm Ag electrode was thermally evaporated through a shadow mask defining a 2 × 5 mm² active area. J‑V measurements were conducted under AM1.5G illumination (100 mW cm⁻²) with a Keithley 2400. SEM, XRD, UV‑vis, and PL (steady‑state and TRPL) analyses characterized morphology, crystallinity, optical absorption, and carrier dynamics.

Results and Discussion

SEM images (Fig. 2) reveal that low‑MAI films exhibit large, tetragonal grains with modest surface coverage, whereas high‑MAI films form dense, smooth, ~200–600 nm crystals (Fig. 3). XRD (Fig. 4) confirms that optimal annealing (>80 °C for high‑MAI) yields a single CH₃NH₃PbI₃ phase, while lower temperatures leave residual PbI₂, reducing efficiency.

Photoluminescence data (Fig. 5) show that low‑MAI films achieve peak emission at 768 nm (25 ns lifetime), whereas high‑MAI films shift to 773 nm with a shorter 14 ns lifetime—indicative of efficient carrier extraction at the TiO₂ interface.

Device performance (Fig. 8) demonstrates that high‑MAI cells deliver higher Jsc (~2 mA cm⁻² above low‑MAI) due to improved absorption and charge transport, while low‑MAI cells exhibit slightly higher Voc, likely from residual PbI₂. Across 50 devices, the high‑MAI process achieves an average PCE of 13.7 % (σ = 1.28 %) with >75 % exceeding 13 % under one‑sun illumination. The optimum cell attains 17.42 % efficiency, 0.97 V Voc, 24.06 mA cm⁻² Jsc, and 0.747 FF.

Conclusions

Adjusting MAI concentration in a two‑step perovskite deposition route effectively tailors grain size and surface morphology, directly influencing photovoltaic metrics. High‑MAI processing yields compact, defect‑free films that enable 17.42 % efficient, reproducible perovskite solar cells. This study underscores the critical role of precursor stoichiometry and thermal treatment in scaling high‑performance perovskite photovoltaics.

Abbreviations

FTO

Fluorine‑doped tin oxide

J‑V

Current density–voltage

MAI

CH₃NH₃I

MAPbI₃

CH₃NH₃PbI₃

PCE

Power conversion efficiency

PL

Photoluminescence

SEM

Scanning electron microscope

TRPL

Time‑resolved photoluminescence

XRD

X‑ray diffraction