非包爾-歐本海默分子動力學理論與在光化學中的應用

Abstract

自從2005 年8 月我被聘為國立交通大學助理教授以來，我們所取得的一個重要的 成就是;(1) 推導出一個新的非絕熱的隧穿幾率的解析公式，結合半古典 non-Born-Oppenheimer trajectory surface hopping 方法，該公式可以用於 non-Born-Oppenheimerdon 動力學的模擬中，並對分子動力學的模擬的計算精度有很大 的提高作用，特別是對光化學中各種錐形交叉引起的非絕熱轉移貢獻可以更好的定量 描述。(2) 另一工作是使用非絕熱的幾率解析公式調節鐳射參數來設計和控制化學反 應。(3) 新的計劃涉及到半導體材料熱電效應，使用熱力狀態函數和第一原理電子結 構計算來研究。 對涉及三原子和四原子的非絕熱的反應系統，我們將研究非絕熱的隧穿效應，這個 效應可以解釋O(3P)+ CO 反應的振動轉動分佈的實驗結果，特別是對低振動態，能夠 更合理解釋理論和實驗產生的差異。這個效應也可以用來解釋實驗S(1D) + H2/ D2 反應 中發現的非尋常同位素效應，而至今沒有合理的理論計算結果及解釋。目前，我們使 用high-level 第一原理電子結構計算方法(CASSCF)來計算O+CO 非絕熱反應勢能表 面。 當涉及到更大的多原子系統，我們應該考慮一種QM/MM 方法，處理非絕熱躍遷問 題。對於反應中涉及到鍵的形成和斷裂的原子和分子，我們用量子力學或者我們開發 的精密半古典方法來處理；對於其他沒有鍵的形成和斷裂的原子和分子，我們可以用 經典系綜方法來描述。我們計畫用這個思路首先研究azobenzene 光致異構化的機制的 trans-to-cis 和 cis-t-trans 變換和pyrazine 吸收光譜。目前，我們使用Born-Oppenheimer 近似及Frank-Condon 重疊的方法研究了pyrazine 的第一和第二激發態吸收光譜。 對於非絕熱動力學涉及到能量和電子在不同相的介面傳遞（特別是超快過程）。核 的運動必須用量子化的概念來處理，而通常激發態勢能面很難精確得到，我們將運用 原分所林聖賢院士發展的一種簡諧微擾近似方法來處理該難題，一個改進的方向是包 含高階的非簡諧貢獻，沿著這個思路我們已經成功地推導出非簡諧效應的一階修正， 並正在用於解釋一些實驗結果。二階的簡諧效應的推導將是下一步工作。目前，我們 做了初步的計算關於ZnCAPEBPP 吸收光譜，電子轉移速率將是下一步工作。

I have been hired as an assistant professor in NCTU since August, 2005. A key accomplishment since then is that we have found an analytical formula for nonadiabatic tunneling probability in non-Born-Oppenheimer dynamics. This new formula can be implemented in semiclassical non-Born-Oppenheimer trajectory surface hopping method to treat nonadiabatic tunneling transition with great precision for MD simulation, especially in photochemistry in which various types of conic intersections play an important role. Another work is done involved controlling chemical reaction by using analytical switching probability formula to adjust laser parameters. The new project is carried out; thermoelectric effect of magnesium silicide is studied by using thermodynamical method with the presence of electric field. Those projects will be investigated further. For nonadiabatic reaction involved in triatomic and teratoimc systems, we propose to investigate nonadiabatic tunneling effects. This effect could explain experimental results of final rotation-vibrational distribution of O(3P)+ CO reaction with lower vibrational states in which the present theoretical studies are not in agreement with experiment. This effect could also explain unusual isotopic effect in S(1D) + H2/ D2 reaction which is observed in experiment but not in the present theoretical calculations. At the present, we have used high-level (CASSCF) a. b. initio method to compute very accurate potential energy surfaces for O+CO nonadiabatic reaction. For nonadiabatic transitions involved in large polyatomic systems, we consider kind of QM/MM method, in which the atoms and molecules involved in bond forming and breaking can be treated by quantum mechanics or by the our present sophisticated semiclassical method and those atoms an molecules not involved in bond forming and breaking can be treated as bath with using normal coordinates. We plan to apply this idea to investigate the mechanism of the trans-to-cis and cis-t-trans photoisomerization of azobenzene and the absorption spectrum of pyrazine. At the present, we have calculated absorption spectra of the first and the second excited states within the Born-Oppenheimer approximation plus Frank-Condon effect. For nonadiabatic dynamic simulation involved in interfacial electron transfer as well as energy transfer (especially for utrafast processes), the nuclear motions must be treated quantum mechanically. As in this case, accurate potential energy surfaces for excited states are not usual feasible, the perturbation method with harmonic approximation developed by Professor S. H. Lin (IAMS) are usually applied. The direction within the perturbation method will be studied in anharmonic effects, in which the first–order correction has been successfully derived and applications are on the way. The second-order correction should be derived in near future. At the present, we have done preliminary calculation about absorption spectra of ZnCAPEBPP in gas phase. We will study it in condense phase with connection to TiO2 semiconductor and simulate interfacial electron transfer rate.