Suppr超能文献

基于容错光子量子计算机的二芳基乙烯激发态计算的资源估计

Resource Estimates for Excited-State Calculations of Diarylethenes on Fault-Tolerant Photonic Quantum Computers.

作者信息

Kanno Shu, Kobayashi Takao, Keithley Kimberlee, Gao Qi

机构信息

Science & Innovation Center, Mitsubishi Chemical Corporation, 1000, Kamoshida-cho, Aoba-ku, Yokohama 227-8502, Japan.

出版信息

ACS Omega. 2025 Apr 30;10(18):18332-18337. doi: 10.1021/acsomega.4c09568. eCollection 2025 May 13.

Abstract

We estimate computational resources for computing excited-state energies of benchmark photochromic molecules of diarylethenes (DAEs) by using quantum phase estimation on photonic devices. The number of T gates, which determines the calculation time, and logical qubits for a simulation are estimated, considering the overhead of fault-tolerant computers. Three DAE molecules of increasing size are examined with active space sizes specified. Quantum resource estimation is conducted via Hamiltonian truncation within an active space of all valence π electrons in all valence bonding π and their antibonding π* orbitals. For small and medium molecules, complete active space configuration interaction generates reference energies and trial initial states, while for the large molecule, a trial state with perfect overlap with the exact excited state is assumed due to computational constraints. Notably, with 1.02 × 10 resource state generators, computation for the largest molecule with an active space of 22 electrons and 22 orbitals takes 7 h and 54 min. These results provide insights into the computational resources necessary for computing excited states on quantum hardware.

摘要

我们通过在光子器件上使用量子相位估计来估算计算二芳基乙烯(DAE)基准光致变色分子激发态能量所需的计算资源。考虑到容错计算机的开销,估算了决定计算时间的T门数量以及模拟所需的逻辑量子比特数。研究了三种尺寸不断增大的DAE分子,并指定了活性空间大小。通过在所有价层π键及其反键π*轨道中所有价层π电子的活性空间内进行哈密顿截断来进行量子资源估计。对于小分子和中等分子,完全活性空间组态相互作用产生参考能量和试探初始态,而对于大分子,由于计算限制,假设一个与精确激发态具有完美重叠的试探态。值得注意的是,使用1.02×10个资源态生成器,对于具有22个电子和22个轨道活性空间的最大分子的计算需要7小时54分钟。这些结果为在量子硬件上计算激发态所需的计算资源提供了见解。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/23f4/12079192/8af3e1d1c9b3/ao4c09568_0001.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验