Re: High Z difference densities question.

In article <[EMAIL PROTECTED]>,
Doug <[EMAIL PROTECTED]> wrote:
>For inorganic structures containing high atomic number elements such
>as Ta, W or Pb, I sometimes encounter large difference densities (with
>respect to spherical independent atom model) of the order 10 to 20 or
>even 30 eA^-3^, localised on or very close to the heavy atom sites.

>of order 5 electrons accompanied by charge concentration and/or
>polarisation of core densities could  give rise to such large
 
  Doesn't sound reasonable to me.


>Are  there any guidelines or rules of thumb which apply,  or any
>recommended reading material about such things? For delta rho as a
>function of Z, how large is too large for publication purposes?
>
 Depends on whether anybody even looks at your delta rho.

 From the IUCr CIF checking algorithms 
http://journals.iucr.org/services/cif/datavalidation.html

PROC-NAME: DIFMX01 Type_2

PURPOSE: To check that _refine_diff_density_max is within expected limits.

PROCEDURE:
DMAX = _refine_diff_density_max
DTEST = 0.1 * ZMAX

TEST

IF DMAX > DTEST*2 issue ALERT A
   "Alert A The maximum difference density is > 0.1*ZMAX*2.00"
> DTEST issue ALERT B
   "Alert B The maximum difference density is > 0.1*ZMAX*1.00"
> DTEST*.75 issue ALERT C
   "Alert C The maximum difference density is > 0.1*ZMAX*0.75"


   The maximum residual electron density is larger than normally expected, 
even after making an allowance for the heaviest element in the structure. 
This is often an indication that 
(a) the absorption corrections are inadequate; 
(b) the overall quality of the data may be poor, leading to 
spurious peaks and holes of residual electron density; 
(c) there is twinning which has not been allowed for where overlap from the
   second twin domain (which may have been ignored in the data collection) 
causes errors in the intensities of some reflections; 
(d) the model is incorrect or incomplete in terms of incorrect element 
assignment, missing atoms or unmodelled or inadequately modelled disorder 
or solvent atoms. 

   If you believe you have eliminated all potential causes of this alert, 
but the maximum residual electron
   density still remains larger than normally expected, you must specify 
the name of the nearest atom to this
   minimum and its distance from the minimum.  

-----------------------

 Or you could use weaker datasets or throw out some high angle data :-)

larry henling