[Paper Review] Millisecond-long electron spin lifetime in CsPbI$_3$ perovskite nanocrystals revealed by optically detected magnetic resonance

V. V. Belykh, Mikhail M. Glazov+8 more|arXiv (Cornell University)|Mar 3, 2026

Chemical and Physical Properties of Materials0 citations

TL;DR

The authors measure long electron and hole spin lifetimes in CsPbI3 nanocrystals using resonant spin inertia ODMR, finding electron T1 up to 0.9 ms and modeling with a two-LO-phonon Raman process.

ABSTRACT

Perovskite nanocrystals are a convenient model system for optical spin orientation and manipulation. However, its real potential might be underestimated due to the incomplete knowledge on spin relaxation times, which are obscured by the limited sensitivity of measurement techniques as well as by the insufficient understanding of the spin relaxation mechanisms in perovskites. In this work, we study the spin relaxation of charge carriers in perovskite nanocrystals both experimentally and theoretically. We address the electron and hole spins in CsPbI$_3$ nanocrystals embedded in a glass matrix by the resonant spin inertia technique based on optically detected magnetic resonance. It allows us to determine the longitudinal spin relaxation time $T_1$ separately for electrons and holes, the $g$ factors, and the effective Overhauser field of the nuclear spin bath. At a temperature of 1.6 K, the $T_1$ time for electrons can be as long as 0.9 ms. We reveal the effect of the time-varying nuclear field fluctuations, which enhances the electron spin relaxation at low magnetic fields, and measure a rather long nuclear spin correlation time of about 60 $μ$s. We develop a model of the spin relaxation in nanocrystals based on a two-LO-phonon Raman process, which explains the observed temperature dependence of the time $T_1$.

Motivation & Objective

• Motivate spin-based applications of perovskite nanocrystals and determine exact longitudinal spin relaxation times T1 for electrons and holes.
• Identify g factors and the effective Overhauser nuclear field influencing spin dynamics in CsPbI3 NCs.
• Quantify how T1 depends on magnetic field, temperature, and optical power, and attribute relaxation mechanisms.
• Develop and validate a theoretical model for spin relaxation via two-LO-phonon Raman processes in nanocrystals.

Proposed method

• Use optically detected magnetic resonance with Faraday rotation to probe spin dynamics in CsPbI3 NCs embedded in glass.
• Employ resonant spin inertia by optically pumping and applying RF fields to measure T1 for specific spin resonances.
• Extract g factors from linear B-field dependencies of ODMR resonances and determine g-factor spreads and Overhauser field fluctuations.
• Measure T1 by fitting spin inertia curves with S = S0 / sqrt(1 + (2π T1 f_mod)^2) across modulation frequencies.
• Fit T1(B) with a nuclear-fluctuation–limited model T1 = τs / [1 + (ΔN/B)^2 (τs/τc)], obtaining τc and ΔN.
• Link temperature dependence to a two-LO-phonon Raman relaxation mechanism and extract an activation energy.

Experimental results

Research questions

• RQ1 What are the electron and hole g factors in CsPbI3 NCs and how do they vary with optical transition energy (NC size)?
• RQ2 What are the longitudinal spin relaxation times T1 for electrons and holes, and how do they depend on magnetic field, temperature, and excitation power?
• RQ3 How do nuclear field fluctuations influence electron spin relaxation, and what are their characteristic time scales?
• RQ4 Can a two-LO-phonon Raman process account for the temperature dependence of T1 in perovskite NCs?

Key findings

• Electron T1 reaches up to 0.9 ms at 1.6 K, high field, and low laser power.
• Electron T1 increases with magnetic field and saturates around B > ΔN, indicating suppression of nuclear-field fluctuations.
• Nuclear-field fluctuation width ΔN ≈ 12 mT and the corresponding Overhauser field spread Δg ≈ 0.15.
• Nuclear field variation correlation time τc ≈ 60 μs, longer than in GaAs QDs.
• Temperature dependence of T1 is activation-like, with Ea ≈ 3.2 meV, consistent with a two-LO-phonon Raman relaxation mechanism.
• Electron g factor varies with optical transition energy as g ≈ 1.68 at 1.824 eV, and decreases with energy; holes show a weaker, secondary resonance at |g| ≈ 0.8.
• Two-LO-phonon mechanism quantitatively captures the activation behavior and provides an order-of-magnitude estimate for the relaxation rate γa.