Sangiogo Gil Eduarda, Oppel Markus, Kottmann Jakob S, González Leticia
Faculty of Chemistry, Institute of Theoretical Chemistry, Universität Wien A-1090 Vienna Austria
Institute for Computer Science, Center for Advanced Analytics and Predictive Sciences, Universität Augsburg Augsburg Germany.
Chem Sci. 2024 Nov 28;16(2):596-609. doi: 10.1039/d4sc04987j. eCollection 2025 Jan 2.
Recent developments in quantum computing are highly promising, particularly in the realm of quantum chemistry. Due to the noisy nature of currently available quantum hardware, hybrid quantum-classical algorithms have emerged as a reliable option for near-term simulations. Mixed quantum-classical dynamics methods effectively capture nonadiabatic effects by integrating classical nuclear dynamics with quantum chemical computations of the electronic properties. However, these methods face challenges due to the high computational cost of the quantum chemistry part. To mitigate the computational demand, we propose a method where the required electronic properties are computed through a hybrid quantum-classical approach that combines classical and quantum hardware. This framework employs the variational quantum eigensolver and variational quantum deflation algorithms to obtain ground and excited state energies, gradients, nonadiabatic coupling vectors, and transition dipole moments. These quantities are used to propagate the nonadiabatic molecular dynamics using the Tully's fewest switches surface hopping method, although the implementation is also compatible with other molecular dynamics approaches. The approach, implemented by integrating the molecular dynamics program package SHARC with the TEQUILA quantum computing framework, is validated by studying the - photoisomerization of methanimine and the electronic relaxation of ethylene. The results show qualitatively accurate molecular dynamics that align with experimental findings and other computational studies. This work is expected to mark a significant step towards achieving a "quantum advantage" for realistic chemical simulations.
量子计算领域的最新进展前景十分广阔,尤其是在量子化学领域。由于当前可用量子硬件存在噪声特性,混合量子 - 经典算法已成为近期模拟的可靠选择。混合量子 - 经典动力学方法通过将经典核动力学与电子性质的量子化学计算相结合,有效地捕捉非绝热效应。然而,由于量子化学部分的计算成本高昂,这些方法面临挑战。为了减轻计算需求,我们提出了一种方法,其中所需的电子性质通过结合经典和量子硬件的混合量子 - 经典方法来计算。该框架采用变分量子本征求解器和变分量子消去算法来获得基态和激发态能量、梯度、非绝热耦合矢量和跃迁偶极矩。这些量用于使用塔利最少开关表面跳跃方法来传播非绝热分子动力学,尽管该实现也与其他分子动力学方法兼容。通过将分子动力学程序包SHARC与TEQUILA量子计算框架集成来实现的该方法,通过研究亚甲亚胺的光异构化和乙烯的电子弛豫得到了验证。结果显示了与实验结果和其他计算研究相一致的定性准确的分子动力学。这项工作有望标志着在实现实际化学模拟的“量子优势”方面迈出重要一步。