|
本帖最后由 不想飞的猫头鹰 于 2024-3-19 17:11 编辑
嗯嗯,感谢您提供的思路,我看了下orca5.0.4的手册,在60/1330页说这个功能不太成熟(THIS IS VERY RUDIMENTARY!),和88/1330页的“ On request, with the UseSym keyword or a %symmetry input block (for which the abbreviation %sym is allowed), the program detects the point group, cleans up the coordinates, orients the molecule and produces symmetry-adapted orbitals in SCF/CASSCF calculations. Note however that the calculation time will not be reduced. While for geometry cleanup the full point group is taken into account (as long as the number of group operations does not exceed 120), only D2h and subgroups are currently supported for electronic-structure calculations. The only correlation module that makes use of this information so far is the MRCI module. Here and in CASSCF calculations, the use of symmetry helps to control the calculation and the interpretation of the results. More symmetry is likely to be implemented in the future, although it is unlikely that the program will ever take advantage of symmetry in a very big way.”说目前不支持太高的对称性。
这种特殊情况也并不好判断,什么时候启用对称性或者用更高质量的格点也有点人为性。如果不按照gview中那么严格的判断对称性的标准的话,就按照卢老师多次调试过的生成的对于orca 5.0以上的关键词,在! wB97X-D3 TZVP def2/J RIJCOSX opt freq tightSCF verytightopt noautostart miniprint nopop关键词下,生成的18个键角为159.86327 159.87491 159.96979 160.07819 160.11089 160.08013 160.08016 160.01877 159.94186 159.86871 159.85146 159.98282 160.01989 160.13435 160.12767 160.07948 160.00316 159.91447 度,而加了defgrid3和verytightopt关键词的18个键角为160.01266 160.01485 160.00878 160.00682 159.99060 159.97376 159.97637 159.98734 160.01363 160.03075 160.02324 160.01787 159.98388 159.97666 159.97801 159.98562 160.00905 160.01013,也只是整体只靠近了均匀的160度一点而已,但是耗时翻了很多倍,之前收敛失败的refgrid重新读取初猜后又收敛失败了。这样的话,之后如果想研究类似的已经证明有对称性的分子,类似18碳环这种,用ORCA进行优化和振动分析是否合适,如果把键长键角列出来后说这已经十分接近D9h了的话是否足够有说服力呢?当前体系在wB97X-D3 TZVP的结果可以用来做分析吗?
|
|