Detangling energy and charge transfer in Photosystem I

普渡大学的Lyudmila Slipchenko获得了化学系化学理论，模型和计算方法项目的奖励，以开发用于光合复合物中能量和电荷转移预测建模的计算方法和软件工具，并提供光系统I中能量转移和电荷分离主要步骤的计算分析。光系统I（PS I）是地球上所有用于太阳能转换的产氧光合生物所共有的主要复合体。PS I的结构和功能通过进化得到优化，并保存了数十亿年。然而，虽然解开这种高效天然太阳能电池的结构-功能关系对于设计太阳能转换的最佳途径至关重要，但对PS I中能量转移和电荷分离的分子水平理解缺乏关键细节。Slipchenko将开发高效和强大的计算方法，其中光合蛋白的混合量子经典描述与机器学习算法增强。使用这些模型，Slipchenko和她的研究小组将致力于阐明PS I中的结构-功能关系，这些关系可以在人工光电池中模仿和模板化。所开发的方法将为计算化学家提供新的算法和工具，用于复杂分子系统的有效和严格的建模。Slipchenko的研究计划将为参与的学生和博士后提供理论和计算电子结构理论和数据科学领域的培训机会。Lyudmila Slipchenko和她的团队将开发计算方法和软件工具，用于光合复合物中能量和电荷转移的预测建模。该方法的发展是基于（i）混合机器学习，量子和经典力学模型，大大加快了生物系统中的基态和激发态计算，以及（ii）电子激发态的能量分解模型，阐明了耦合系统中的结构-功能关系。在Fenna-Matthews-Olson（FMO）家族光合作用蛋白质建模的新方法发展和专业知识的支持下，Slipchenko小组正在寻求建立与光系统I中能量转移和电荷分离的光学特征一致的系统哈密顿量，并阐明初级电荷分离事件的动力学和PS I反应中心中辅助色素的作用。理解PS I中的结构-功能关系具有根本和实际的重要性，因为PS I包含了一种进化优化的光伏机器，可以在人工光电子器件中模仿和模板化。所开发的算法、计算代码和机器学习潜力为计算化学家提供了有效和严格建模复杂分子系统的工具。所有开发的代码将通过开源库和软件包libefp，GAMESS，GROMACS以及Q-Chem量子化学软件中的开源模块提供给社区。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Lyudmila Slipchenko of Purdue University is supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to develop computational methodologies and software tools for predictive modeling of energy and charge transfer in photosynthetic complexes, and to provide computational analysis of the main steps of energy transfer and charge separation in Photosystem I. Photosystems I (PS I) is a major complex shared by all oxygenic photosynthetic organisms utilized for sunlight energy conversion on Earth. The structure and function of PS I is optimized by evolution and preserved for billions of years. However, while disentangling the structure-function relations in this highly efficient natural solar cell is essential for designing optimal routes of solar energy conversion, molecular-level understanding of energy transfer and charge separation in PS I lacks critical details. Slipchenko will develop efficient and robust computational methodology in which hybrid quantum-classical description of the photosynthetic protein is augmented with machine learning algorithms. Using these models, Slipchenko and her research group will pursue to elucidate the structure-function relations in PS I that can be mimicked and templated in artificial photocells. The developed methodology will provide the community of computational chemists with new algorithms and tools for efficient and rigorous modeling of complex molecular systems. Slipchenko’s research program will provide training opportunity for participating students and postdocs in the field of theoretical and computational electronic structure theory and data science. Lyudmila Slipchenko and her group will develop computational methodologies and software tools for predictive modeling of energy and charge transfer in photosynthetic complexes. The method developments are based on (i) hybrid machine learning, quantum and classical mechanics models that dramatically speed up ground and excited state calculations in biological systems, and (ii) energy decomposition models for electronic excited states that elucidate the structure-function relations in a coupled system. Empowered by the new methodological developments and expertise in modeling photosynthetic proteins of the Fenna-Matthews-Olson (FMO) family, Slipchenko group is pursuing to build the system Hamiltonian consistent with optical signatures of energy transfer and charge separation in Photosystem I, and to elucidate kinetics of the primary charge-separation events and the role of accessory pigments in the PS I reaction center. Understanding of the structure-function relations in PS I is of both fundamental and practical importance as PS I encompasses an evolutionary optimized photovoltaic machinery that can be mimicked and templated in artificial photoelectronic devices. The developed algorithms, computational codes and machine learning potentials provide the community of computational chemists with tools for efficient and rigorous modeling of complex molecular systems. All developed codes will be available to the community via open-source libraries and software packages libefp, GAMESS, GROMACS, and open-source modules in the Q-Chem quantum chemistry software.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.