[Wien] convergence of Bader (aim) calculations

[Jorissen Kevin]

Dear Peter,
thank you very much for your swift help.  Here are some comments to your reply :
 
** About the K RMT : 
This is core leakage from outputst :
 TOTAL CORE-CHARGE:                  10.0000000000118                                                       <  K
 TOTAL CORE-CHARGE INSIDE SPHERE:    9.99941207977341
 TOTAL CORE-CHARGE:                  2.00000000000043                                                        < C
 TOTAL CORE-CHARGE INSIDE SPHERE:    1.99977147149244

This is not perfect, but I thought it would do.  I cannot increase RMT(C) any further, and tried to keep the two close together.
 
** About the 'total charge'  : 
As mentioned in my e-mail, I also calculated this 'total charge', and for the best calculations (rkmax 6.5 to 7) I have 0.35 electrons too many (roughly 0.006 per atom), so it seems the Bader surfaces are slightly overlapping.
 
** The accuracy required : 
The system is KC60, which means that charge from one K atom will be transferred to a molecule consisting of 60 C atoms.  So the charge transfer per atom is expected to be very small.  My current values  per C atom are comparable to the precision you mention.  An accuracy of 0.005 would be nice.
 
** The system
is indeed very open and is bound to have low density regions.  I had anticipated problems, and was surprised to see that extending the precision of the density calculation (to gmax 20 and charge components up to l=10 for all atoms) didn't matter at all (differences of at most 0.001 in the charge!).  Let me add that I converged all calculations using run -cc 0.00001.
 
So, it is obvious that if I want to get reliable numbers, I will need to increase accuracy a little further.  I was hoping for suggestions to how I could do this - by increasing the angular mesh for aim?  By increasing Rkmax?  (By enlarging RMT(K)?).  How critical is the exchange/correlation potential?
All options are rather time consuming, and I was hoping to get away with doing only one :-)
 
Kevin.
 

	-----Oorspronkelijk bericht----- 
	Van: Peter Blaha [mailto:pblaha at zeus.theochem.tuwien.ac.at] 
	Verzonden: wo 6/30/2004 5:03 
	Aan: wien at zeus.theochem.tuwien.ac.at 
	CC: 
	Onderwerp: Re: [Wien] convergence of Bader (aim) calculations
	
	

	Also my experience with Bader is limited, but:
	
	I usually trust charges to about 1-3 hundreds of an electron (X.yz+/-0.03)
	
	
	
	>Rkmax is really important - at rkmax 7 (for RMT(C)=1.3 and RMT(K)=1.45, approximately) I am starting to see (I hope :)) signs of convergence.  Unfortunately I doubt I'll be able to go significantly beyond 7.
	
	
	K seems a little bit small. Anyway, what means "convergence" ? To how many
	digits in the charge ? I'm surprised that you need to go that high with
	RKmax.
	
	> 1/ which values of the named parameters would you expect to be reasonable?  I'm especially interested in hearing about nth,nphi and npr - perhaps an increase of merely 10 points is ridiculously small?  (I'm sampling the full interval, by the way - I'm not using symmetry there).
	
	More or less use the defaults. Within my accuracy (above) it should be ok.
	
	I always check the integration by forming the sum of all charges*mult !! and
	this must be close to zero. (with charges I mean the last number in the
	outputaim file). This worked also (within 0.05) for 88 atom cells.
	
	If it is not, the only explanation could be, that your structure is so open
	(and with a very low density) that aim has problems,
	or it would also be explained when you have a non-nuclear maximum in the
	density (like an F-center).
	
	                                     P.Blaha
	--------------------------------------------------------------------------
	Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
	Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
	Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
	--------------------------------------------------------------------------
	
	_______________________________________________
	Wien mailing list
	Wien at zeus.theochem.tuwien.ac.at
	http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
	
	

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 8818 bytes
Desc: not available
Url : http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20040630/9bb75c43/attachment.bin