Characterizing Infrared Spectra of OH–·(H2O)2 and OH–·(H2O)3 with Constrained Nuclear-Electronic Orbital Molecular Dynamics

The vibrational spectra of OH–·(H2O)n clusters for small n have been well established experimentally, with fundamental modes largely assigned. However, clear assignment of highly anharmonic modes and combination bands associated with strong hydrogen bonds, which often manifest as broad spectral features, remains challenging. In this work, we employ constrained nuclear-electronic orbital molecular dynamics (CNEO-MD) to provide detailed peak assignments and plausible physical interpretations for the vibrational spectra of OH–·(H2O)n clusters with n = 2 and 3. The CNEO framework incorporates nuclear quantum effects, particularly nuclear quantum delocalization, through the underlying effective potential energy surfaces. When combined with classical molecular dynamics, CNEO-MD further captures coupling effects between vibrational modes. Leveraging machine-learned potentials, we perform a series of temperature-dependent CNEO-MD simulations and use the resulting spectra to facilitate peak assignment. Our results largely confirm the experimental assignments reported by Johnson and coworkers [J. Chem. Phys. 2016, 145, 134304], while also providing direct, physically grounded interpretations of previously unassigned features.对于小尺寸n的OH⁻·(H₂O)ₙ团簇的振动光谱，已通过实验得到充分表征，其基频振动模式基本完成归属。然而，针对与强氢键相关、通常呈现为宽谱特征的高度非简谐振动模式及组合带的明确归属，仍颇具挑战。本研究采用约束性核电子轨道分子动力学（constrained nuclear-electronic orbital molecular dynamics, CNEO-MD）方法，针对n=2和n=3的OH⁻·(H₂O)ₙ团簇的振动光谱，实现了详细的峰归属并给出合理的物理解释。CNEO框架通过其底层有效势能面纳入了核量子效应，尤其是核量子离域效应。与经典分子动力学结合后，CNEO-MD还可捕捉振动模式间的耦合效应。本研究借助机器学习势函数，开展了一系列温度相关的CNEO-MD模拟，并利用所得光谱结果辅助峰归属。我们的结果在很大程度上验证了Johnson及其团队于[J. Chem. Phys. 2016, 145, 134304]报道的实验归属，同时还为此前未被归属的光谱特征提供了直接且基于物理依据的解释。