Optimizing Titanium Precursors for High‑Performance Compact TiO₂ Layers in Perovskite Solar Cells

Background

Since the first demonstration of methylammonium lead iodide (MAPbI₃) as a light absorber in 2009, perovskite solar cells have attracted significant interest due to their high performance and swift efficiency gains. Over the past five years, the PCE of PSCs has surged from 9 % to 22.1 % [3]. PSCs typically comprise a compact TiO₂ layer, an electron transport layer, the perovskite absorber, and a hole transport layer (HTL). While planar architectures and HTL‑free designs have been explored, the compact TiO₂ layer remains indispensable for high‑efficiency devices because it simultaneously facilitates electron transport and blocks hole‑FTO contact [7, 8]. Various deposition techniques—spray pyrolysis, spin‑coating, atomic layer deposition, sputtering, and electrochemical deposition—have been reported for creating compact TiO₂ films [9–13]. Spin‑coating, in particular, is favored for its simplicity, low cost, and scalability. Traditionally, titanium diisopropoxide bis(acetylacetonate) (c‑TTDB) and titanium isopropoxide (c‑TTIP) have been the most common precursors [14, 15]. Recent work introduced tetrabutyl titanate (c‑TBOT) as a promising alternative [16]. Despite these advances, a systematic comparison of precursor performance for spin‑coated compact layers in PSCs is still lacking. In this study, we synthesized compact TiO₂ layers from c‑TBOT, c‑TTIP, and c‑TTDB, and rigorously evaluated their structural, electrical, and optical properties as well as their impact on PSC performance. Our results demonstrate that the c‑TBOT precursor yields the best combination of conductivity, transparency, charge extraction, and recombination resistance, leading to superior device metrics.

Experimental

Preparation of compact TiO₂ layers

FTO glass substrates (≈15 Ω cm⁻², Japan) were etched with 2 M HCl and Zn powder, then sequentially cleaned in Hellmanex detergent, deionized water, acetone, 2‑propanol, and ethanol. Finally, the substrates were treated with UV‑O₃ for 15 min. The compact layer was deposited by spin‑coating at 3000 rpm for 30 s and annealed at 500 °C for 30 min. Three precursor solutions were prepared as follows:

• c‑TBOT: 0.25 mL tetrabutyl titanate (99 %, Aladdin) was dissolved in 5 mL ethanol, then 0.2 g polyethylene glycol, 1 mL nitric acid, and 0.5 mL deionized water were added. The mixture was stirred for 5 h, precipitated for 15 h, and filtered through a 0.45 µm PTFE filter [16].
• c‑TTDB: 0.15 M titanium diisopropoxide bis(acetylacetonate) (75 wt % in isopropanol, Sigma‑Aldrich) in 1‑butanol [14].
• c‑TTIP: 0.23 M titanium isopropoxide (99.999 %, Aladdin) plus 0.013 M HCl in isopropanol. 369 µL titanium isopropoxide and 35 µL 2 M HCl were diluted separately in 2.53 mL isopropanol; the HCl solution was added dropwise to the titanium precursor under vigorous stirring. The mixture was filtered through a 0.45 µm PTFE filter [19].

Device fabrication

A mesoporous TiO₂ (mp‑TiO₂) layer was spin‑coated over the compact layer using TiO₂ paste diluted 1:6 in ethanol at 4000 rpm for 30 s, then heated at 100 °C for 10 min and annealed at 500 °C for 30 min. The perovskite precursor—FAI (1 M), PbI₂ (1.1 M), MABr (0.2 M), and PbBr₂ (0.2 M) in DMF/DMSO (4:1)—was deposited by a two‑step spin‑coating (1000 rpm/10 s, 4000 rpm/30 s) with 200 µL chlorobenzene dropped 20 s before the end of the second step, followed by 1 h at 100 °C. After cooling to room temperature, a spiro‑OMeTAD layer (72.3 mg spiro‑MeOTAD, 28.8 µL TBP, 17.5 µL Li‑TFSI solution (520 mg Li‑TFSI in 1 mL acetonitrile), 1 mL chlorobenzene) was spin‑coated at 4000 rpm for 30 s. A 70 nm gold electrode was thermally evaporated on top of the HTL.

Characterization

The morphology and microstructure of the compact layers were examined by FESEM (JEM‑7001F, JEOL) and AFM (Bruker Multimode 8). X‑ray diffraction (XRD) patterns were recorded with a Bruker D8 Advance diffractometer (Cu‑Kα, λ = 0.1542 nm). Current‑voltage (J‑V) characteristics were measured under AM 1.5G (100 mW cm⁻²) using a Keithley 2440 source meter; the active area was 0.1 cm² defined by a shadow mask. Conductivity was evaluated with a Keithley 2400. Steady‑state and time‑resolved photoluminescence were measured using an Edinburgh FLS 980E fluorometer. UV‑vis absorption spectra were obtained with a Cary 5000 spectrophotometer. Electrochemical impedance spectroscopy (EIS) was performed with a CHI660e workstation at 0.8 V, 10 mV amplitude, 0.1 Hz–1 MHz. Incident photon‑to‑current conversion efficiency (IPCE) spectra were recorded with a Crowntech Qtest Station 500ADX.

Results and discussion

Figure 1a–d presents AFM images of the four substrates. The c‑TTIP film exhibits the smoothest surface, whereas the c‑TTDB and c‑TBOT films are comparatively rougher. The root‑mean‑square (RMS) roughness values on a 5 µm × 5 µm scan are summarized in Additional file 1: Table S1. The bare FTO RMS is 13.4 nm, which decreases to 11.4, 9.38, and 6.65 nm after coating with c‑TTDB, c‑TBOT, and c‑TTIP, respectively, indicating successful TiO₂ deposition.

AFM images of a bare FTO, b c‑TBOT, c c‑TTIP, and d c‑TTDB

X‑ray diffraction patterns of c‑TBOT, c‑TTIP, and c‑TTDB deposited on glass without FTO

Photovoltaic parameters of devices plotted as a function of different compact layers (J_SC, V_OC, FF, and PCE)

a Conductivity measurement results of various c‑TiO₂. The inset depicts the structure of the sample. b Plots of −dV/dJ vs (J_SC−J)⁻¹ derived from J‑V curves and the linear fitting curves

UV‑vis absorption spectra of perovskite films based on different compact layers

a PL and b TRPL of perovskite films based on different compact layers

a–d Current density‑voltage (J‑V) curves and IPCE for the best cells based on different compact layers

Nyquist plots of the solar cells based on different compact layers at 0.8 V under AM 1.5G. The inset is the equivalent circuit applied to fit the Nyquist plots

Conclusions

We have systematically compared three titanium precursor solutions—c‑TBOT, c‑TTIP, and c‑TTDB—for spin‑coated compact TiO₂ layers in perovskite solar cells. The c‑TBOT‑based compact layer delivers the highest conductivity, transparency, and charge‑transfer efficiency, which collectively reduce recombination and hysteresis while boosting J_SC, V_OC, and FF. Consequently, PSCs incorporating the c‑TBOT layer achieve an average PCE of 17.03 %. These findings provide clear guidance for selecting high‑quality titanium precursors in scalable, cost‑effective PSC fabrication.