一個更普遍用於半經典理論研究化學反應動態學的絕熱勢能曲線

Abstract

摘要 此研究利用了現今發展完善的反應途徑(Reaction coordinate)以及正交於反應座標系的3N-7震動模(Normal modes)所建構的J = 0絕熱曲線(adiabatic curves)來進行OH + H2 □ H2O + H化學反應動態學的研究。發現由此絕熱曲線在只考慮穿隧效應下得到的J=0累積反應機率(Cumulative reaction probability)比Miller的J=0累積反應機率(Cumulative reaction probability)小很多。可能原因是反應端跟生成端轉動相關的震動模的數目不相符。吾人也利用了絕熱模型(adiabatic model)，改裝過的絕熱模型(modified adiabatic model)和透熱模型(diabatic model)來計算累積反應機率，並得到三個可能與Miller結果差異的原因。第一是能量曲線並不是拋物線的模型，第二是非絕熱偶合(non-adiabatic coupling)的貢獻和最後一個是延反應途徑的頻率分析的準確度。為了確定哪個是最主要的原因，吾人利用上述的三個模型到比較大的體系， Cl +CH4 □ CH3 + HCl。可是對於大於五個原子以上的體系，並沒有一個精確的解。所以只有降維度的資料可以比較，也就是說，吾人只能用降維度的絕熱模型來做比較。在與Nyman三個降維度；RLA，RLU和RBU,的結果比較，吾人發現絕熱模型的結果與Nynam的結果的趨勢很近，並且發現考慮的維度越小，共振的現象越大。從比較的結果，吾人發現以上三個原因中，以第三個原因為決定累積反應機率準確度的最主要原因。

Abstract We constructed more general adiabatic energy curves of J =0 for OH + H2 □ H2O + H by using the reaction coordinate and 3N-7 normal modes which is orthogonal to the reaction coordinate, but found the cumulative reaction probability of these adiabatic energy curves (without the effect of non-adiabatic coupling, only the effect of tunneling) is lower than the result of Miller. The possible reason would be the linkage between reactants and products. Since three rotational related vibration modes in reactant side, but two in product side. And we also applied the adiabatic model, modified adiabatic model and diabatic model to the cumulative reaction probability and then got three main possible reasons for the deviation from Miller’s result. First is that the energy curves are not parabolic model, the second one is that the contribution of non-adiabatic coupling and the last one is the accuracies of the frequencies along the reaction path. In order to make sure which one is the main reason, we applied these three models to a larger but prototypical system, Cl +CH4 □ HCl + CH3. But for more than 5-atoms system, there’s no exact solution no matter hyper-spherical (projected one is the same) or Jocabi coordinate. So only reduced dimensions information is available, that is, only the adiabatic model with reduced dimension could be applied to do comparisons. After comparing with the three models; RLA, RLU, RBU, by Nyman, we found that the trends of our adiabatic models goes well Nyman’s results and the effect of resonance becomes larger as the degree of freedom becomes smaller. From the result, we conclude that the last reason; the accuracies of the frequencies along the reaction coordinate dominate the accuracy of the result for the cumulative reaction probability.