Condensed Fukui Function, Reactivity Index

Post here if you would like to suggest new features for Dalton
Post Reply
Rintintin
Posts: 9
Joined: 15 Dec 2013, 00:26
First name(s): Ariel
Last name(s): Gonzalez
Affiliation: none
Country: United States

Condensed Fukui Function, Reactivity Index

Post by Rintintin » 16 Jun 2016, 00:41

Hi...
I would like to see in Dalton the Condensed Fukui Function to calculate the reactivity index to study favorable sites for electrophilic, nucleophilic and radical attacks.

Thanks...

Bernd S.
Posts: 17
Joined: 19 Feb 2015, 13:31
First name(s): Bernd
Last name(s): Schimmelpfennig
Affiliation: KIT
Country: Germany

Re: Condensed Fukui Function, Reactivity Index

Post by Bernd S. » 16 Jun 2016, 06:29

Well, just implement it. Nobody will object ...

bast
Posts: 1190
Joined: 26 Aug 2013, 13:22
First name(s): Radovan
Last name(s): Bast
Affiliation: none
Country: Germany

Re: Condensed Fukui Function, Reactivity Index

Post by bast » 16 Jun 2016, 08:51

dear Ariel,
thank you for the suggestion. We value these suggestions and this is why we have this forum section.
best regards,
radovan

Rintintin
Posts: 9
Joined: 15 Dec 2013, 00:26
First name(s): Ariel
Last name(s): Gonzalez
Affiliation: none
Country: United States

Re: Condensed Fukui Function, Reactivity Index

Post by Rintintin » 16 Jun 2016, 15:24

Bernd S. wrote:Well, just implement it. Nobody will object ...
Hi...
The manual does not show anything on condensed fukui.
Could you show me how to get it?

Thanks...

taylor
Posts: 519
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Aarhus University
Country: Denmark

Re: Condensed Fukui Function, Reactivity Index

Post by taylor » 16 Jun 2016, 16:04

I think you misunderstood Bernd's posting --- he was light-heartedly saying that the best way to get new functionality into Dalton is to do it yourself! As Radovan said we always welcome suggestions, thoughts, and ideas, but implementing new functionality is something that somebody in the "Dalton mob" has to take on as a new task in addition to anything else they're doing. It is important to understand that Dalton is a purely voluntary effort and no-one is employed to take care of or to extend the code. This of course is closely tied to why the code is free: if no money is coming in, there's no mechanism to pay someone to maintain or extend it!

This is of course a simplification and indeed an oversimplification. Much of Dalton and LSDalton development is done in research groups where various individuals have research grants to develop certain methods and to implement them (linear-scaling coupled-cluster methods, say). However, with modern grant-driven research, as opposed to the curiosity-driven research that was now sadly a luxury of the past, one is usually required to attempt what one said in the grant, at least if you want another grant later! We do value suggestions but anything that is other than a very simple extension or modification of the code (and as Bernd says, someone could do this themselves) is likely to be put on a to-do list that will depend on when someone who is interested in doing it gets the financial and personnel resources to do it.

I am myself not really familiar with the Fukui function and reactivity index. But I will make a general point that I feel is particularly relevant to Dalton and LSDalton development. In very large part the Dalton mob focus on "observables". For example, the dipole moment, its response to an external electric field (the polarizability) which may be time-varying (dynamic polarizability), or nuclear spin-spin coupling constants. Yes, I am again oversimplifying, these things are not necessarily "observables" in the true quantum-mechanical sense. But they are well-established connections between theory and experiment. Contrast this with, for example, Mulliken population analysis (and you may have observed that Dalton is one of the few codes that does not print out such an analysis by default, although it does monitor some such quantities during the SCF iterations). While a population analysis may look like some convenient thing that makes chemists feel good, it is completely artefactual. Mulliken introduced it in the context of what we now call a minimal basis calculation, for which it can perhaps provide some helpful analysis. But we could (as I have posted before) conceive of a complete basis, which would give us the exact SCF result, and we could put that basis on the planet Venus, and our calculation would give us the exact SCF energy and properties, but the population analysis would be completely meaningless! All the population would be on Venus!

I do not want to discourage people from using simple, pictorial methods for trying to understand, for example, chemical reactivity. But one has to admit, from a theoretical perspective, it's all a bit, well, squishy. And most of the Dalton mob tend to have priorities that are less squishy...

Best regards
Pete

Rintintin
Posts: 9
Joined: 15 Dec 2013, 00:26
First name(s): Ariel
Last name(s): Gonzalez
Affiliation: none
Country: United States

Re: Condensed Fukui Function, Reactivity Index

Post by Rintintin » 16 Jun 2016, 23:43

I read in a publication that Hirshfeld charge and NPA charge are the ones most worth to be recommended for Fukui.

taylor
Posts: 519
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Aarhus University
Country: Denmark

Re: Condensed Fukui Function, Reactivity Index

Post by taylor » 17 Jun 2016, 10:38

I don't want to sound discouraging, but making further, detailed suggestions (if that's what your posting about NPA charges was intended for) is unlikely to move any particular project up the "Dalton priority ladder". At the risk of being tedious, let me repeat two key elements from my previous posting. To pick up a suggestion, a current Dalton developer needs to be funded to do it in some way. This might be a more senior researcher acquiring grant funds, or it might be a more junior researcher being paid as a student or postdoc on such grant funds. Other than the prospect that a developer will pick something up and do it in her or his spare time when not teaching, doing research they're funded for, or (most of all) attending lots of absolutely pointless meetings, it will take someone to get very interested in a topic for it to receive major attention.

The second element from my previous posting is observables. All population analyses are nonobservables and as a consequence one can cook up any answer one likes. For example, project the wave function from the actual basis onto a minimal basis and use Mulliken. Well, how do you deal with "hypervalent" (another nonobservable concept) compounds of second-row elements like P or S? I have said (for example in a lecture at a summer school last year) that it is essential that theoreticians master the squishy, fictional concepts that drive inorganic and organic chemists, whether it's "pushing arrows" for describing hypothetical "reaction mechanisms", or crystal-field theory for explaining the observed spectra of transition-metal complexes. But mastering them does not mean subscribing to them, just as understanding (some of) the things that Donald Trump says does not require agreeing with them! (They guarantee he's a horse's ass, but that's a different conclusion...) Much of the Dalton effort has focussed on properties of various sorts, electric and magnetic, for which there is a well-developed and formally argued connection between experiment (say, birefringence), and theory (in that case polarizability anisotropy). We can calculate the latter in various well-defined ways and improve our calculations both by using more elaborate methods and including effects such as vibration, or a solvent. "Populations" of any sort are a fiction, and which ones "work best" for particular purposes is just piling fiction on fiction. Most Dalton developers are focussed much more on properties for which there is a well-defined formal connection to experimental observations.

I was privileged to know Kenichi Fukui, as I am privileged to know his Nobel co-recipient, Roald Hoffmann. But I submit that brilliant success in explaining qualitative aspects of chemistry and chemical reactions is not necessarily a guarantee that this success can be made quantitative, in the sense of using it in high-level quantum-chemical calculations.

Best regards
Pete

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest