Researchers at Sejong University in South Korea have developed a self-powered perovskite photodiode for blue-light detection using a new anisole-assisted precursor engineering technique, reaching an external quantum efficiency (EQE) of 78.11% at 380 nm and a detectivity of 6.77 × 10¹² Jones at zero applied bias. The devices also retained over 97% of their initial EQE after 1,000 hours of unencapsulated storage.
Blue-light detection has applications spanning underwater sensing and communication, where blue wavelengths penetrate water more effectively than other visible light; medical diagnostics, for fluorescence-based assays and phototherapy; and wearable optoelectronics that monitor exposure to potentially harmful blue light. Conventional blue photodetectors rely on III-V semiconductors such as GaN, InGaN, and AlGaN, which perform well but require costly epitaxial growth on rigid substrates, limiting their use in flexible and wearable devices. Halide perovskites are solution-processable and offer a tunable bandgap, but most perovskite optoelectronics research has focused on narrow-to-moderate bandgap compositions for photovoltaics and LEDs; wide-bandgap perovskites for blue detection have lagged, in part because widening the bandgap with chlorine tends to produce poor film morphology, phase instability, and higher defect density.
The research team addressed this by adding anisole, a solvent with antisolvent-like behavior, directly into the chlorine- and bromine-based wide-bandgap perovskite precursor solution (a DMSO-based FA0.4MA0.6Pb(Br0.55Cl0.45)3 formulation with a 2.58 eV bandgap), then applying a brief anisole bathing step after deposition to strip residual coordinating solvent. Used as an internal antisolvent, anisole slowed and evened out crystallization, yielding pinhole-free films with grains exceeding 300 nm, versus predominantly sub-200 nm grains without it.
The best performance came at 40 vol% anisole; going higher, to 50 vol%, oversaturated the precursor and caused visible precipitation and poor film coverage. The finished photodiode uses an ITO/perovskite/PC60BM/C60/ZnO/Ag stack that deliberately omits a hole-transport layer, a design choice the team says reduces dark current and improves self-powered (0 V) performance.
That film-quality improvement translated directly into device performance: dark current density at 0 V fell to 4.89 nA/cm² for the 40 vol% anisole devices, down from 16.06 nA/cm² for the anisole-free control, and stayed low (5.24 nA/cm²) even under a small reverse bias, indicating better voltage tolerance. The on/off ratio reached 8.35 × 10⁵, about 90 times higher than the control, and a rejection ratio (comparing EQE at the 470 nm target wavelength against 530 nm) of 117.4 pointed to improved wavelength selectivity. Responsivity reached 0.27 A/W. Across 50 fabricated devices, the anisole-treated batch also showed markedly tighter device-to-device uniformity, with average dark current density dropping roughly three orders of magnitude compared to the control group.
On stability, unencapsulated devices stored in an argon-filled glovebox held onto more than 97% of their initial EQE after 1,000 hours, while control devices without anisole fell below 50% of their initial EQE after just 400 hours. The authors frame anisole-assisted precursor engineering as a broadly applicable, scalable route to wide-bandgap perovskite photodiodes for self-powered UV and blue-light sensing.
