计算化学公社

标题: DFT negative frequency. How small is small? [打印本页]

作者
Author: liuxiaogang0206 时间: 2016-5-6 16:40
标题: DFT negative frequency. How small is small?
很不幸，跑完了几个 DFT （B3LYP/6-31+G(d,p) ) S1 激发态的opt + freq.
发现有小小的 negative frequency。再跑一次成本太高。
想请教一下大家，这个可以容忍还是需要再次加精度优化？例外那个负频率对应的IR Int 对这个判定结果有多大影响了？

参考了帖子 http://www.gaussian.com/g_whitepap/vib.htm
里面有一段 “A note about negative frequency”

“You'll find that the frequencies for the translations are almost always extremely close to zero. The frequencies for rotations are quite a bit larger. So, how ``close to zero'' is close enough? For most methods (HF, MP2, etc.), you'd like the rotational frequencies to be around 10 wavenumbers or less. For methods which use numerical integration, like DFT, the frequencies should be less than a few tens of wavenumbers, say 50 or so.”

但这个解释有点模糊，另外只提到了频率，好像对于IR Inten 之类不很在意。想请大神指点详情。谢谢。

Case 1:
Full mass-weighted force constant matrix:
Low frequencies ---  -58.4689  -11.8250 -4.2095 -0.7156 0.0177 0.4100
Low frequencies --- 13.5391 45.2874 73.0230
****** 1 imaginary frequencies (negative Signs) ******
Diagonal vibrational polarizability:
   108.0064220    25.8468486    43.7482070
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
                  1                   2                   3
                  A                   A                   A
Frequencies -- -57.6117             45.1193             72.7695
Red. masses --    2.0793                7.0243                5.7788
Frc consts  --    0.0041                0.0084                0.0180
IR Inten --    0.0161                0.0340                1.6005

Case 2:
Full mass-weighted force constant matrix:
Low frequencies ---  -32.8770  -12.8801 -0.5486 -0.1589 0.2635 11.0292
Low frequencies --- 17.1687  114.6092  218.6386
****** 1 imaginary frequencies (negative Signs) ******
Diagonal vibrational polarizability:
   129.1962504    12.7364658    19.9316493
Harmonic frequencies (cm**-1), IR intensities (KM/Mole), Raman scattering
activities (A**4/AMU), depolarization ratios for plane and unpolarized
incident light, reduced masses (AMU), force constants (mDyne/A),
and normal coordinates:
                  1                   2                   3
                  A                   A                   A
Frequencies -- -32.6837             114.5989             218.6383
Red. masses --    10.2128                4.7052                4.7434
Frc consts  --    0.0064                0.0364                0.1336
IR Inten --    0.1134                1.0882                2.1035

作者
Author: sobereva 时间: 2016-5-14 19:26
还是int=ultrafine结合opt=tight再重新优化一次为宜。否则不仅是这一个模式是虚频，其它模式的红外强度、振动频率也不准。

为了节约时间，强烈建议去掉弥散，本身这对优化结构通常也没什么用。氢的极化也可以省。

作者
Author: liuxiaogang0206 时间: 2016-5-16 09:35
Thank you

欢迎光临计算化学公社 (http://bbs.keinsci.com/)