最近看了篇文献,里面涉及到自旋-轨道耦合(SOC)和不同反应势能面跃迁几率的计算.
       一般来说,我们都是用Landau-Zener跃迁几率公式进行计算,可是文献中M.S.Gordon等人却直接从图11中读出了几率的大小,很是疑惑,请大家帮忙解答下!谢谢!

J. Chem. Phys., Vol. 112, No. 23, 15 June 2000.pdf (312.63 KB)

J. Chem. Phys., Vol. 112, No. 23, 15 June 2000.pdf (312.63 KB)
下载次数: 16

2008-6-18 16:17

搭车问啊！！

还有什么软件可以算这个SOC？看过的文章都是用Gamess的，Molpro可以吗？

F以前的原子(体系)可以用CASSCF进行计算
况且直接可以算出跃迁几率公式中的所有未知参数,如: 瞬时速率v, 不同势能面在交叉点附近的梯度差等.
但是对于过渡金属,尤其是重金属,相关很详细的报道还没见过
Molpro好象也能算SOC,但是大多好象是GAMESS,毕竟GAMESS-US是免费的,呵呵~
Dice,好长时间不见啊

I am very new about excited state calculations. My wording do not absolutely right. Sorry.

Based on Landau-Zener formula,

P(LZ) = exp[(-2pi*A2)/(hv*delta F)].

  The case Gordon studied is a crossing between doublet and quartet. The spin are different! Calculations may give a 0 value for derivative coupling between two different spins?

  Sometimes, I saw some people used SOC to estimate magnitude of the crossing.  I guess he used SOC to replace derivative coupling for Landau-Zener formula??? What's F paper in 2000? How they get velocity of the motion? Amazing!?!?

Do you have experience to calculate Landau-Zener formula?? I have to consider that in my study.

SOC can be calculated by Gaussian, Molpro too, but I did not use before.

个人见解,仅供参考.

Calculations may give a 0 value for derivative coupling between two different spins?
         不同自旋态之间的耦合系数应该有等于0的情况,但是我们主要计算关键电子和轨道之间的作用,所以耦合系数一般不为0,具体你可以参考这篇文献:
http://g.zhubajie.com/urllink.php?id=2580504uybdea2e5drxrohr

      I guess he used SOC to replace derivative coupling for Landau-Zener formula???
      应该不是吧,文献中已经明确提到了室温下的跃迁几率为13.4%.况且直接以SOC的大小衡量Landau-Zener的几率大小也是不准确的,见参考文献:
http://g.zhubajie.com/urllink.php?id=258050709l1ag5qz1cb12op
http://g.zhubajie.com/urllink.php?id=2580534wyz21qux5pz1f5fe

What's F paper in 2000?
F, fluoride atom.

How they get velocity of the motion? Amazing!?!?
CASSCF的CONICAL关键词好象可以计算得到,我没算过

Do you have experience to calculate Landau-Zener formula?? I have to consider that in my study.
和你一样,第一次接触,欢迎多交流

个人见解 too,仅供参考.
[quote]原帖由 ccidnetcom 于 2008-6-19 16:24 发表
个人见解,仅供参考.

Calculations may give a 0 value for derivative coupling between two different spins?
         不同自旋态之间的耦合系数应该有等于0的情况,但是我们主要计算关键电子和轨道之间的作用,所以耦合系数一般不为0,具体你可以参考这篇文献:
http://g.zhubajie.com/urllink.php?id=2580504uybdea2e5drxrohr


      I guess he used SOC to replace derivative coupling for Landau-Zener formula???
      应该不是吧,文献中已经明确提到了室温下的跃迁几率为13.4%.况且直接以SOC的大小衡量Landau-Zener的几率大小也是不准确的,见参考文献:
http://g.zhubajie.com/urllink.php?id=258050709l1ag5qz1cb12op
http://g.zhubajie.com/urllink.php?id=2580534wyz21qux5pz1f5fe


From the molpro 2006 manual, when we locate MECP between differentspin, we don't need to calculate derivative coupling (DC). "skip the DCevaluation if the conical intersection involves states with differentspin (e.g., a Singlet/Triplet crossing) because the coupling is thenzero." Interestingly, Shaik exactly used SOC to appromixate DC in LZ equation.


What's F paper in 2000?
F, fluoride atom.

D. G. Fedorov and M. S. Gordon, J. Chem. Phys. 112, 5611 ~2000!.

How they get velocity of the motion? Amazing!?!?
CASSCF的CONICAL关键词好象可以计算得到,我没算过

It should be wrong.  CONICAL关键词 calculates DC and 2 Gradients (to get gradient difference) only. People need velocity of trajectory when it crosses. Shaik guessed velocity by "the sum of translational energies of the two reactants at 300 K". Interestingly.

CONICAL是控制CASSCF计算中的指定态优化计算变成寻找这个态的圆锥交叉或者避免交叉的计算. 这个是我理解错了

Shaik guessed velocity by "the sum of translational energies of the two reactants at 300 K".
   S.Shaik 很巧妙的回避了这样定性猜测速率的不合理性,但倘若我们要仔细地讨论两不同势能面交叉点附近的自旋翻转行为的话则不能这样说明.
   回归到这个帖子的原始问题上,Gordon等人是用什么方法计算做出图11并得到跃迁几率的?用GAMESS计算速率和交叉点附近势能面梯度差?还是用G03的CASSCF方法?
CASSCF我也是刚入手,好多问题还不懂,请大家指教!
谢谢

You can send email to Gordon to ask your questions.
势能面梯度差 can be done by many QM softwares. You can also use other methods such as DFT for 势能面梯度差, when you get correct SCF solutions for these two states. CASSCF sometimes is expensive.
However, velocity can't be obtained easily. For my case, I have to run MD to get velocity.
No other idea!

引用:

原帖由 ccidnetcom 于 2008-6-19 21:21 发表
CONICAL是控制CASSCF计算中的指定态优化计算变成寻找这个态的圆锥交叉或者避免交叉的计算. 这个是我理解错了

Shaik guessed velocity by "the sum of translational energies of the two reactants at 300 K".
...

给Gordon写信已有10天了吧,还没见回复.
DFT方法计算势能面梯度差我没试过,麻烦你能再深入介绍下吗?
十分感谢你的帮助!