应《网络安全法》要求，自2017年10月1日起，未进行实名认证将不得使用互联网跟帖服务。为保障您的帐号能够正常使用，请尽快对帐号进行手机号验证，感谢您的理解与支持！

24小时热门版块排行榜    

小木虫论坛-学术科研互动平台 » 计算模拟区 » 量子化学 » Gaussian » 审稿人对我只用B3LYP有意见
52/1返回列表
查看: 3179  |  回复: 18
只看楼主 @他人 存档 新回复提醒 (忽略) 收藏    在APP中查看
当前只显示满足指定条件的回帖，点击这里查看本话题的所有回帖

hee_csu

金虫 (著名写手)

[求助] 审稿人对我只用B3LYP有意见

审稿人意见：B3LYP is a functional known to have very large exact exchange component (20%) favoring too loose electron densities and low transition energies. State-of-the-art is to at least use also a functional of the GGA class with zero exact exchange, which will give an impression of the error/spread obtainable from DFT, currently not described.
我的回答：B3LYP, as a hybrid functional parameterized to fit data primarily for main group molecules, has achieved outstanding accuracy and has been widely used in theoretical chemistry calculation. However, B3LYP is known to have very large exact exchange component (20%) favoring too loose electron densities and low transition energies. So, using the pure functional BP86 is necessary, which is a GGA class with zero exact exchange (0%). It can provide another side of the coin. So, we have also performed the geometry optimization, vibrational analysis and excited states calculation using pure functional BP86. The results have been added into the new version of manuscript.

本人纯属量化菜鸟，有个问题，请高手指点：
1、B3LYP中的20%交换相关项是什么?
2、为什么要用纯的泛函BP86来验证?
回复此楼

» 猜你喜欢
今天有点凉
1楼 2011-10-23 10:13:26
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

hee_csu

金虫 (著名写手)
送鲜花一朵
引用回帖:
13楼: Originally posted by yongleli at 2011-10-25 17:16:47:
后边几位量化专家说得很好。
我这里再赠送一个公式：

A*EXSlater+(1-A)*EXHF+B*ΔEXBecke+ECVWN+C*ΔECnon-local

这个式子是说，传统的纯泛函描写电子的行为是过分定域化了。
但是传统的HF计算又过分离域 ...

请高手指点：具体在哪篇文献中提到过相关内容，导师要我回答为什么考虑HF项，我实在是答不上来（导师非量化方向）。
赞一下 回复此楼
今天有点凉
18楼2011-10-26 16:07:57
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
查看全部 19 个回答

yongleli

木虫 (正式写手)

【答案】应助回帖

★ ★ ★
小红豆(金币+3): 感谢分享 2011-10-23 11:09:58
hee_csu(金币+10): 非常感谢，还望继续指教。 2011-10-23 14:43:20
引用回帖:
1楼: Originally posted by hee_csu at 2011-10-23 10:13:26:
审稿人意见：B3LYP is a functional known to have very large exact exchange component (20%) favoring too loose electron densities and low transition energies. State-of-the-art is to at least use also ...

审稿人是说，你用的是混合泛函，但是现在公认B3LYP计算反应能会偏低。
而且他们说，B3LYP计算得到的电子云过于离域化了，
所以让你用纯泛函验证一下，电子结构是不是你report的那么回事，
你计算得到的反应能是不是由于B3LYP误差相消，
导致你计算的值碰巧跟实验值对上了。
你必须证明你的计算原理正确、不是误差相消碰巧弄出来的。
赞一下(1人) 回复此楼
4楼2011-10-23 10:55:08
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

yongleli

木虫 (正式写手)

【答案】应助回帖

★ ★
小红豆(金币+2): 感谢分享 2011-10-23 11:10:13
hee_csu(金币+5): 放了电子轨道图，请继续指点，非常感谢。 2011-10-23 14:46:47
引用回帖:
1楼: Originally posted by hee_csu at 2011-10-23 10:13:26:
审稿人意见：B3LYP is a functional known to have very large exact exchange component (20%) favoring too loose electron densities and low transition energies. State-of-the-art is to at least use also ...

还有，你如何证明你计算得到的电子结构是正确的（你放了电子轨道图在文章里吗？）。
赞一下(1人) 回复此楼
5楼2011-10-23 10:56:06
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

lyy880530

木虫 (正式写手)
我的审稿意见也是这个意思， it is well known that the B3LYP tends to favor high spin states over low spin states. This is due to the HF exchange energy mixture that has this intrinsic feature and it should always be properly calibrated. One way to do this would be to compare the hybrid vs pure GGA functionals and check the energy difference between the quintet and the septet states. Or by smoothly change the quantity of HF exchange and look at the energy gap. Did the authors check this ? This can be quite important because along the reaction energy path you may get “earlier” or “later” spin-crossing.
看到你的帖子了，正好想请教你一下
赞一下 回复此楼
Life,withallofitsfragilewonders,shouldalwaysbepullingusin!
6楼2011-10-23 11:29:36
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
查看全部 19 个回答
普通表情 龙 兔 虎 猫
信息提示
关闭
请填处理意见
关闭