« first day (2102 days earlier)      last day (2285 days later) » 

7:43 AM
@Secret Reminds me of old lab instructions like "the mixture of x and y is converted into z by grinding in a mortar until exhaustion of the laboratory assistant"
 
better to let the machines to do the hardwork, they do it better anyway
 
 
4 hours later…
11:58 AM
Guys, what software are available to calculate QTAIM that is as modern as possible and can run on a mac?
 
 
1 hour later…
1:28 PM
@Secret I think Multiwfn should do the trick
At least they list a Mac version: sobereva.com/multiwfn
If i recall correctly Goeff mentioned that Avogadro can do it, too.
 
I see, thanks!
 
1:51 PM
hmmm... the tutorial of QTAIM in avogadro is very brief and I cannot see any bond orders showing up, only the critical points
and it would be better for them to explain what each of the colors mean, for I don't want to make assumptions on where I expect the critical points will be when I apply it to those complicated molecules in my PhD
 
@Secret I have no idea what I see there. Try multiwfn, it is a fantastic program... makes nice analyses and gives you all the values you need (and then some).
 
ok
 
@Secret In case you have money to throw away... aim.tkgristmill.com/readme.html
 
O AIMALL is out because a) I have 60 atoms in the molecules and b) most importantly, they don't support PBE functionals which is what we used in the calculations
(Our complexes are organometallic molecules and it is well known in the literature that the platinum group does the best with PBE and other pure functionals)
so I am installing multiwfn now
 
2:11 PM
as far as i know a wfx file doesn't specify how the wave function was created and I don't see any limitations on pbe then. And I had no problems analysing molecules with more than 60 atoms. It just is very, very slow, and expensive...
 
yeah, that is not a very good idea since I have 5*144*5*5 molecules in total to analyse
 
with qtaim? Wow, what is it you're looking for? That seems like a load of work...
 
Basically, we are looking for structural electronic trends of some piano stool complexes in the platinum group. The first 5 in the multiplication refers to the 5 types of ligand 1, the 144=12*12 refers to the 12 types of ligands 2 and 3. After that the 2nd 5 is for repeating this whole batch by replacing the Cp containing group with an electron rich one to check electronics, and the final 5 are trigonal planar complexes (which I must have undercount since they have on average 4 minimas each)
upon discussion with my supervisior and our past experience, our complexes are highly delocalised systems that CMOs and NBOs are too messy to give useful info, and we then found out QTAIM and thinking it will help us to pin down how the electron density migrates as a function of the ligands
Luckily, I have wrote heaps of scripts that automated the 5*144 of that multiplication, which takes a couple of days to finish the batch on the supercomputer
so in theory, the analysis of the whole batch is really limited by the time taken for parallel computation of 200+ jobs (which has 5 tasks each) in the cluster, thus reasonably fast
I am not sure if I can automate multiwfn, but that might not be a good idea because analysis of electronic density is quite visual in nature
but will see...
 
 
2 hours later…
4:14 PM
@Secret you can automate multiwfn, at least the time consuming part, creating pdb files for the critical points etc. You can then match those with the output of the properties... I've written a bash script for helping with this. It has some problems, especially with the setup, but it can give you an idea of how it could work, find it here: github.com/polyluxus/runMultiwfn.bash
there are a couple of example files and I think i included a full qtaim in there...
so you can probably run those analyses quite easily and the difficult part will be making sense of it all ;) Good luck :D
 
 
3 hours later…
6:50 PM
@Loong o/
 
@Mithoron hi
 
Seems there was a bit more activity on site last few days? Exams?
 
Anyone got any recommended resources for implementing analytical gradients for molecular integral schemes?
 
7:07 PM
@obackhouse Why would someone have?
@obackhouse Well maybe someone will have, I don't even know what you're talking about :D
 
 
1 hour later…
8:16 PM
@Mithoron none of my texts etc. seem to have much explanation on it, and was trying to implement it into my hartree fock program
 
 
2 hours later…
9:53 PM
@obackhouse how are you getting your current integrals?
 
I'm using Obara-Saika for overlap/kinetic, and McMurchie-Davidson for Coulomb and 2-electron
But i can work with either scheme
 

« first day (2102 days earlier)      last day (2285 days later) »