@Jan I used it for Heterocycles.

Oooh yes, very hard to access journal D=

@hBy2Py Reminds me of Time Cube: timecube.2enp.com

@pentavalentcarbon Yeah, or the wall of text on bottles of Dr. Bronner's soap.

the thing is some of that stuff is probably fine if you ignore the nonsense text, that can make it hard to tell the difference between marketing and nonsense

@Jan :-)

Our Finnish safety trainer said the purpose is just to get drunk and naked.

What happened to @dhmo?

Hi! =)

@DSVA hi

@hBy2Py @pentavalentcarbon Here's another one:

chemistry.stackexchange.com/users/39913/…

Actually, I am about to destroy this user account, so here is his profile detail.

> I am a freelance Physicist/Chemical-Engineer who really enjoys the study of dark matter, wormholes & dare I say it, has a burning desire along with a deep, deep passion to someday work towards proving or at the very least, adding my positive contribution to the inevitable technology of actual human time travel.

> I am, at present time, working on experimental projects. Some of which require the documented notes and information to remind classified upon contract. These are for some high-profile clients.

Sometimes we could all use a little burning desire...

@pentavalentcarbon do you have experience calculating bond dissociation energy (or radicals in general)? I'm not quite sure what method to use:-D

Hahaha... yeah....

@DSVA Not really, other than the general knowledge. 2 electrons? CCSD. 3 electrons? CCSDT. 4 electrons? CCSDTQ.

I jest. It depends on the system.

Better ask Martin.

@DSVA it obviously depends on a lot of factors, but you can use unrestricted hartree fock

just can't use restricted

(that is if you want to use Ab initio.) I'm not so sure about DFT for open-shell systems

if they are mono- or di-radicals (as you'd expect with the organic species generated by homolytic cleavage for BDE calculations), you can also use ROHF

(it's just that ROHF isn't suitable for things like metals that might have unpaired electrons in 3 or more orbitals)

Actually, I guess those coupled cluster methods penta mentioned would be better.

What is a "freelance physicist?"

@Zhe It's a phrase that identifies you as a probable crackpot.

Well, if you've got a bunch of systems, try a functional and MP2 on all of them, then pick one that might be "tough" and use CCSD on that one too.

That way for one rxn you can compare all methods and not pay coupled cluster cost needlessly for everything.

well, that's a simple and good idea. :-D

Also don't mix RO and U results, so what you can do is dictated by programs in a sense. Dunno if NWChem can do ROMP2 or ROCCSD.

Gaussian can

@DSVA sorry to state the obvious I guess :P, I just recall that RO is sometimes good when you don't have to worry about many unpaired electrons (organic radicals), So U is not the only option.

Every time I've done this I used MP3 though

penta's suggestion to "benchmark" seems like a good idea

@pentavalentcarbon why CCSD instead of regular HF? (I'm curious, don't know anything about coupled cluster)

yeah have to look if NWChem can do that. Gaussian is fine, I'm just not sure if it's possible on our server to use more than one node with that (I've never tried it) while it's simple for NWChem. :-D

watcha trying to figure out with BDEs @DSVA?

@gannex Because when you break a bond, you need to account for electron correlation, both static and dynamic correlation, but you may assume the static correlation is small and only account for dynamic correlation. RHF has no dynamic electron correlation, UHF has almost none.

Oh OK, so CCSD takes care of that?

@gannex some colleagues from the macromolecular chemistry division want to know the dissociation energies for some of their, I guess, monomers or polymerization modifiers. In this case they want to know the difference between a C-Si bond and C-Ge bond.

CCSD will account for dynamic correlation between 2 electrons. CCSD on the Helium atom is equivalent to FCI for a given basis in that it is exact in that basis.

sometimes I don't get gaussian....I need to know the node names to correctly set up a LINDA job? that's ridiculous. :-D

@Jan Your bounty prompted me to re-think-through that answer; it needed some clarification/refinement (@DSVA).

@mhchem, something funny happened to some a \ce in this post of mine -- I think it swapped the closing curly bracket someplace it shouldn't have.

@pentavalentcarbon OK. Sounds like it merits account deletion then. Proceed.

@Zhe Then again, I could become a "freelance chemist" after I graduate.

Is that a euphemism for unemployed? ;)

freeland chemistry sounds like it could get you in trouble

frelance*

:P

@pentavalentcarbon what about MPn methods, correcting for correlation?

I guess CCSD is still the best