Peak intensity dependence

suneetkale

Hi,

I am having issues fitting peak intensities for primary phases in my fit.
My main peak intensity, coming from the primary phase at approximately 3.955 deg., changes by about 10-15% depending on the be of Sn in the main phase. If I refine the beq for Sn, I get beq as 0.840. However, the intensity calculated is 14% higher than the actual data. If I fix it to 0.3, the intensity matches up. If I refine the occupancy of Sn and Zn along with beq the intensity is 13% higher, and beq is around 0.82. I do not see any change in the lattice parameters or the phase % age. Is there a fix for this? I know that beq and occupancy are related, but this is too much of a difference.

Here is the str code:

str
local !CIF_PHASE_ID 1

Stephens_hexagonal(!eta, 0.45994_0.00801 min =0; max=1;, !s400, 1545.40850_44.37521 min =0;, !s004, 4.26120_0.14315 min =0;, !s202, 144.13499_7.37729 min =0;) 

r_bragg	2.47398707
phase_MAC	2.5038305
phase_name	"Nb2Sn0.9Zn0.1C"

MVW( 622.407, 124.180, 89.871`_0.082)
space_group 194
scale @  1.24039486e-005`_4.823e-008
 
Hexagonal(  3.214617_0.000038,   13.875997_0.000227)
site	Nb	num_posns	4	x	!xNb	=1/3;	y	!yNb	=2/3;	z	!zNb	0.58394_0.00003	occ	Nb	1	beq biso  0.34536`_0.00853 min 0
site	Sn	num_posns	2	x	!xSN	=1/3;	y	!ySn	=2/3;	z	!zSn	=1/4;	occ Sn !a  0.9 beq b  0.82488`_0.01521 min 0 occ Zn =1-a; :  0.10000 beq =b; :  0.82488`_0.01521
site	C	num_posns	2	x	!xC		0.00	y	!yC		0.00	z	!zC		0.00	occ	C	1	beq	1

johnsoevans

I'm not sure what the problem is in this specific case. However, as you say, atomic displacement parameters and scale factors are highly correlated, especially if you only have data over a narrow 2-theta range. They will also be correlated with any other parameters that have a smooth dependence on 2-theta (e.g. LP corrections, variable slit corrections, beam overspill corrections, absorption, surface roughness, etc). If you're worried about the intensity of a small number of reflections you might also have to consider preferred orientation/texture. Your peak is at very low 2-theta (though I don't know the wavelength); instrument/sample alignment issues can be important at low angle (particularly on lab instruments).
My first check is always to run a known material on the same instrument (e.g. Y2O3) to make sure I can fit those data correctly.