| 查看: 1871 | 回复: 20 | |||
| 当前主题已经存档。 | |||
coolrainbow木虫 (著名写手)
未来国家冻凉
|
[交流]
【转帖】AIM理论将代表理论化学的发展方向？（原创+转载）
|
||
|
coolrainbow注：此帖转自CCL. Prof Bader是AIM理论的重要代表人物. 目前的分子电子结构（即不包括固体量化）的量子化学的计算方法，主流方法即 是从头算和密度泛函理论.这些计算方法应用极其广泛，并且从理论上有系统的 改进方法，如对于从头算，即是不断的增大基组的数量和扩大组态空间；对密度泛函， 即是改进交换泛函Exc的质量等等.但是，这些计算，往往越是精确，越缺乏物理意义. 试想，当初我们用一个两个基函数变分出来的VB函数解释H2分子，尽管解离能同实验 值相差甚远，但是那漂亮的曲线在定性上同实验完全一致，受此启发人们得到了与分子 电子结构有关的一系列结论.而更精确的波函数出现后，诸如Kolos-Wolniewicz 那个有一千多个变分函数的泛函，我们除了能得到一个比光谱实验还精确的解离能量 数值外，对其中如何由H2的电子结构得出那个36117.3cm-1的结果一无所知.到更大的 体系，例如一个大型的分子，虽然能用复合方法算出精确的热力学、光谱等数据，但这也 靠无数的组态堆积和经验矫正得到的.也就是说，到目前为止，“理论”化学在“计算” 上取得了巨大的进步，在“理论”上却远远没有达到相应水平（也就是文中所说， technique, not conceptual）. 相比之下，AIM理论具有非常明确的物理意义，并且 能够非常“物理”的解释很多化学问题.当然，同传统的QC方法相比，AIM在计算方法 和普及程度上都要差许多，究竟能有多少发展前途，还需经历时间的考验. CCL: Why Bader's theory represents the future of theoretical chemistry? Sent to CCL by: Sengen Sun Thanks to five people who responded to my last post (3 private and 2 public). Three of them agreed with me, and two don't. To me, Bader's theory is conceptually superior to Pople's and Kohn's. I am very disappointed that Dr. Nicholls judges a good theory based on its current popularity. (It is accidentally ironic!) As pointed out by Dirac, the major task in Quantum mechanics is to solve the Schrodinger equation, or is about mathematics. Pople and Kohn have just done that in different ways - techniques (in my best knowledge)! I have no doubt that Pople and Kohn well deserved the Nobel Prize. They created wonderful mathematical tools to solve some aspects (but not all) of chemical puzzles. But they just created tools! In DFT, the key sentence related to physics is that electron density (ED) decides everything. Unfortunately, ED is simply a mathematical tool and does not have an absolute meaning in DFT. Although some physical principles are referenced (used) quite often in order to solve, ED - the cause of chemical phenomena is not the major concern in DFT, but other observable properties such as geometries, energies, etc.. DFT abandons its original conceptual statement in the end of the mathematical processes. Bader's focus has been the cause of natural puzzles - ED. He and his group of people have proved that there is absolute ED in the nature that is computational accessible and experimentally provable by x-rays crystallography. If there is computationally accessible ED today, there will be computationally accessible ED tomorrow, and 100 years later, and 1000 years later, and 10000 years later, and.....! Until this world ends! Will DFT or LCAO still be there by then? I am not sure! This is an authentic example of theoretical revolution of chemistry about solving natural puzzles as Kuhn discussed. Who else did the same? Pople? Kohn? Hoffmann? Fukui? I don't know. > Dr. Anthony Nicholls: "The comments by Sengun Sun reflect one > of the main criticisms of Kuhn's work at the time it was first > published. It was easy to see the shift in Newton-Einstein or > Classical-Quantum as paradigm shifts but since these only happen > every fifty years or so, what use is it?" Are you asking "What use is it?"? Please forgive me for being straightforward. But do you think this a good opportunity to make some money (commercial opportunity?), or to make decoration symbolizing that the main stream community has reached a new level of civilization? > > Dr. Anthony Nicholls: "If Sun had attended any of the Kuhn >> Symposia he would have heard my introduction wherein I give many >> examples of such and in particular champion the "Kuhn-lite" >> description by the great British geneticist J. B. S. Haldane (J. >> Genetics #58, 1963, p 464): >> Four stages of acceptance of a scientific theory: > i) It is worthless nonsense. > ii) This is an interesting, but perverse, point of view. > iii) It is true, but quite unimportant. > iv) I always said so. > Every time I have given this introduction I have seen heads nod in > rueful acknowledgement because most of us who have had original ideas > have experienced just this process." Therefore, you think you are the problem-solver by recruiting "experts" and "judges". I would say that you should be awarded a nobel prize more than Bader. Best regards, Sent to CCL by: Sengen Sun [sengensun_-_yahoo.com] > Quote: > "Still Sun's > argument that DFT is just a computational tool does not > hold water. > Admittedly this is how the majority of people use it and > this is why they > awarded Kohn the Nobel prize with Pople (purely > computational tool > there), but there have been significant chemical insights > from conceptual > DFT due to Parr and coworkers which are on par (pun > intended) with those > of Bader." > > This is a part of the first of 4 messages I received on 10-08-2008. All 4 questioned about my remarks on DFT. I must address this issue after nearly a decade hesitation (I may hurt too many!). DFT starts with good physics, and ends with contradictory "conceptual". In the nature, there are electron density and electrostatic forces, but not Fukui functions derived from HOMO and LUMO. By the end of 2007 and early 2008, I was asked twice by 2 Editors to review 2 papers from 2 different groups, which were about use of conceptual DFT in prediction and understanding of regiochemistry of cycloadditions. In the end of my review, I said something like this: "I don' t mind publishing this paper as there have been many papers of this kind in the literature any way, and there will be more. But I must express my strong personal opinion that this is not what scientific theory is about." I asked the editors to release my identity to the authors as I welcome the authors to argue with me directly in order to be fair. They did not contact me afterward and I have not seen their papers out yet. Simply for the known 1,3-dipolar cycloadditions in the literature, prediction using conceptual DFT may fail as much as 30-40% (my very rough estimate)! A simple Diels-Alder reaction it cannot address is acrolein dimerization. I would like to challenge any one in this community, you come forward to rationalize its regioselectivity based on matching the nucleophilicity and electrophilicity indexes. If you claim you have a conceptual theory, you must be able, within your theoretical frame, to address these "exceptions" and draw a boundary to use your theory. Otherwise, prediction and understanding are meaningless! Who else understands this "conceptual"? Please look at the back-to-back debate (JPC A, 2001, 10943-10948). How do you understand the "conceptual DFT" there, my friends? ...... By the way, 2 people said that I was not accurate in describing Bader's theory in my last post. I have not be convinced. At least, I think I am close. They are just my own words to say it after I digested the literature. Somewhere in the literature, it says something like that the computed ED from AIM can be closely compared with that from X-rays crystallogrophy. I interpreted this to my own words in my last e-mail. Bader is my hero in science, and is in the right direction of //Good comment theoretical chemistry to me. If you use physics to explain chemistry, chemis try can be easily understood. Theoretical chemistry is about connections between physics and chemistry. If you use chemistry to explain chemistry, it is just like you ask me "why" and my answer is "because why". Further, no one is perfect, and our theory is far from being perfect. There are some particular things, on which I don't agree or others don't agree with Bader. There is no way that two people can think exactly the same. I encourage open and fair discussions. Regards to every one. Sengen [ Last edited by cadick on 2009-12-11 at 02:04 ] |
回复此楼» 猜你喜欢
重庆交大26年硕士生招生拟调剂通知已出！欢迎加入光子学微结构与器件课题组。
已经有0人回复
-
UJN物理学专业调剂
已经有12人回复
-
物理学I论文润色/翻译怎么收费?
已经有198人回复
-
0702一志愿吉大B区求调剂 本科期间发表一篇Sci
已经有0人回复
-
北京纳米能源与系统研究所王中林院士/曹南颖研究员课题组招收2026硕士研究生1名
已经有0人回复
-
山西大同大学物理学硕士研究生招收调剂生
已经有0人回复
-
[调剂信息]211智能人工感知方向国家青年特聘专家课题组招收调剂研究生
已经有0人回复
-
［调剂信息］211智能人工感知方向国家青年特聘专家课题组招收调剂研究生
已经有0人回复
-
[调剂信息]211智能人工感知方向国家青年特聘专家课题组招收调剂研究生
已经有0人回复
-
一志愿物理所凝聚态物理352求调剂
已经有0人回复
» 本主题相关商家推荐: (我也要在这里推广)
snoopyzhao
至尊木虫 (职业作家)
- 计算强帖: 1
- 应助: 157 (高中生)
- 贵宾: 0.02
- 金币: 18844.7
- 红花: 29
- 帖子: 3803
- 在线: 1422.4小时
- 虫号: 183750
- 注册: 2006-02-13
- 专业: 污染生态化学
2楼2009-06-28 10:38:08
coolrainbow
木虫 (著名写手)
未来国家冻凉
- 应助: 1 (幼儿园)
- 贵宾: 0.1
- 金币: 3979.5
- 散金: 10
- 红花: 50
- 帖子: 1137
- 在线: 272.3小时
- 虫号: 239560
- 注册: 2006-04-07
- 性别: GG
- 专业: 理论和计算化学
★ ★ ★
zeoliters(金币+3,VIP+0):感谢介绍 6-28 13:27
zeoliters(金币+3,VIP+0):感谢介绍 6-28 13:27
|
最经典的几个例子：F在几乎所有分子里都有强吸电子效应，苯环在无论什么样的分子里都有相似的性质，烷烃的生成焓和C原子数几乎完美成正比，等等，而technique的理论只能算出无数的数据，至于怎么从他们之中解释出这些现象似乎一无所知（当然有定域轨道理论等，不过只对HF有用），我们能够算出精确的能量，却对这些熟知了一百年的普通化学现象说不清楚，这不能不说是传统理论的一大软肋，一个遗憾。 这急需理论化学家们发展更加conceptual的理论，多在基础理论上，在物理和化学本质上下功夫，而不要一昧的陶醉于越来越多的变分函数和越来越可怕的数学技巧。因此，我觉得AIM理论在这方面，即使不能大有作为，也一定能为以后出现的更加完备、更加美丽的理论提供有益的启发。 |
3楼2009-06-28 12:05:06
yalefield
金虫 (文坛精英)
老汉一枚
- 计算强帖: 2
- 应助: 129 (高中生)
- 贵宾: 0.17
- 金币: 21238.9
- 散金: 3440
- 红花: 66
- 帖子: 12101
- 在线: 759.1小时
- 虫号: 96063
- 注册: 2005-10-07
- 专业: 高等教育学
- 管辖: 计算模拟
4楼2009-06-28 13:00:42
★ ★ ★ ★
小木虫(金币+0.5):给个红包，谢谢回帖交流
wuchenwf(金币+1,VIP+0):谢谢 7-2 22:56
zeoliters(金币+2,VIP+0):谢谢回帖交流！ 7-5 10:53
小木虫(金币+0.5):给个红包，谢谢回帖交流
wuchenwf(金币+1,VIP+0):谢谢 7-2 22:56
zeoliters(金币+2,VIP+0):谢谢回帖交流！ 7-5 10:53
|
理论化学的方向将是Density-Matrix Renormalization Group (DMRG) 理论的研究表明：DMRG是局域，低标度，大小一致，多参考的凝聚态物理和量子化学方法，完美的符合John Pople所期待的终极量子化学方法。 DFT的时代即将终结，所谓Conceptual理论，说到底，就是玩Conceptual，没有任何物理甚至化学上的新东西。所谓AIM，只是一个posttool而已！ PS（个人意见，欢迎讨论） [ Last edited by 天空空 on 2009-6-29 at 02:25 ] |
5楼2009-06-29 02:15:58
★ ★ ★ ★
小木虫(金币+0.5):给个红包，谢谢回帖交流
wuchenwf(金币+1,VIP+0):谢谢 7-2 22:56
zeoliters(金币+2,VIP+0):谢谢回帖交流！ 7-5 10:54
小木虫(金币+0.5):给个红包，谢谢回帖交流
wuchenwf(金币+1,VIP+0):谢谢 7-2 22:56
zeoliters(金币+2,VIP+0):谢谢回帖交流！ 7-5 10:54
|
数据不是无用的，只是大多数理论化学家太缺乏物理素养（不乏像莱纳斯.卡尔·鲍林这样的天才，才会有''Resonating Valence Bond theory'' RVB这样同时具有深刻物理内涵又具有清晰的化学图像的理论），其实化学家不要太执着于“表象”，因为几乎所有的理论化学问题的本质上在物理上是清晰的（量子电动力学），只有精确的求解QED（DIRAC）方程，才能反馈的定义清晰的化学图像（当然大部分化学图像s是基于近似），而不能执着于（conceptual）概念。就像我们看到一个球，我们应该去测量它的直径，周长，再去精确的算出它的体积（这就是科学的力量，尽管它还远不是真理）。而不是我们这样定义一个直径，又那样定义一个周长，甚至自欺欺人的定义一个体积......这样我们永远不知道它的真实而有用的体积。玩弄conceptual是要建立在深厚而深刻的理论和计算的基础上。 [ Last edited by 天空空 on 2009-6-29 at 02:51 ] |
6楼2009-06-29 02:47:18
coolrainbow
木虫 (著名写手)
未来国家冻凉
- 应助: 1 (幼儿园)
- 贵宾: 0.1
- 金币: 3979.5
- 散金: 10
- 红花: 50
- 帖子: 1137
- 在线: 272.3小时
- 虫号: 239560
- 注册: 2006-04-07
- 性别: GG
- 专业: 理论和计算化学
7楼2009-07-01 22:13:46
yalefield
金虫 (文坛精英)
老汉一枚
- 计算强帖: 2
- 应助: 129 (高中生)
- 贵宾: 0.17
- 金币: 21238.9
- 散金: 3440
- 红花: 66
- 帖子: 12101
- 在线: 759.1小时
- 虫号: 96063
- 注册: 2005-10-07
- 专业: 高等教育学
- 管辖: 计算模拟
波恩-奥本海默近似在氯加氢高能反应中有效
★ ★ ★
小木虫(金币+0.5):给个红包，谢谢回帖交流
fegg7502(金币+2,VIP+0):thank you very much! 7-4 18:23
小木虫(金币+0.5):给个红包，谢谢回帖交流
fegg7502(金币+2,VIP+0):thank you very much! 7-4 18:23
8楼2009-07-02 22:23:00
9楼2009-07-03 01:07:27
coolrainbow
木虫 (著名写手)
未来国家冻凉
- 应助: 1 (幼儿园)
- 贵宾: 0.1
- 金币: 3979.5
- 散金: 10
- 红花: 50
- 帖子: 1137
- 在线: 272.3小时
- 虫号: 239560
- 注册: 2006-04-07
- 性别: GG
- 专业: 理论和计算化学
，弄得学生们都觉得BO近似是很正常的 10楼2009-07-03 20:21:11
