SinoGreenTech Academic Portal
Official PDF TranslationSCIENCE CHINA Materials

Two-dimensional halide perovskite memristor arrays: linearly programmable for neuromorphic computing

Authors: Hudie Wei; Liang Chu

DOI: 10.1007/s40843-025-3257-yStatus: Verified Translated Edition
Sponsored AdvertisementAd Placement Area
reCAPTCHA Bot Shield Active

Preparing Secure Academic Download

Verifying human reader & generating high-resolution document...

Verifying Document Integrity15s remaining
← Back to Article
Protected by Google reCAPTCHA v3.PrivacyTerms
Sponsored ContentAdSense In-Feed Ad Slot

Key Findings in This Report

• • SCN additive incorporation into BDAMAn−1PbnI3n+1 (n = 3) films yields vertically aligned [PbI6]4− octahedral layers and superior crystallinity, enabling columnar grains with minimal grain boundaries; this microstructure directly reduces ion migration heterogeneity, a prerequisite for linear conductance updates in crossbar arrays. • • Moisture stability of SCN-modified DJ HP films reaches 7 months in ambient air, compared to 28 days for randomly oriented films—a 7.5-fold improvement. This threshold is critical for commercial neuromorphic hardware, where device lifetime must exceed 5 years under uncontrolled humidity. • • Potentiation and depression (P&D) characteristics are nearly perfectly linear and symmetric, with homogeneous migration of charged interfacial ions. This linearity minimizes weight update errors in analog neural network training, directly impacting inference accuracy and training convergence rates. • • First-principle DFT calculations identify the absence of van der Waals gaps between inorganic perovskite layers as the key structural feature governing resistive switching. This mechanistic insight provides a design rule for eliminating nonlinear conductance artifacts in 2D HP memristors.