analysis of channel vs intensity diffraction data. pk_xo?

rowlesmr

I have some data coming off a diffractometer in channel vs intensity. Channel is directly equivalent to 2theta, with the proviso that channel 1 is ~160 deg2Th and channel 512 is ~145 deg2Th.

Is there a way to use pk_xo to define a calibration equation to change X (channel) into 2Th, so that I can use all the other bits and bobs (emission spectra, cs_L...) as per normal? I don't want to have to prechange the data, or have to redefine all the emission spectra and other macros.

My current attempt ends with the error "No hkls for det0.xye". The error is due to either the pk_xo line. Placing it before the XDD, before the LAMs, after the LAMs, before the STR, or after the STR, make no difference.

Thanks

Matthew

My entire input file is:

XYE("det0.xye")
	x_calculation_step 1
	
	prm !ch_width -0.0356183
	prm ch_offset 160 min 150 max 170
	pk_xo = ((2 Th)-ch_offset)/ch_width; 'this is the bit that I need to figure out.
	
   lam ymin_on_ymax 0.0001 
   Lam_recs
   {
      0.378  2.2896977	0.6160974  
      0.132  2.2900254	0.7444350 
      0.084  2.2905983	1.3279580 
      0.073  2.2915136	2.1807273 
      0.009  2.2882482	0.8395685 
      0.271  2.2936467	0.9436726 
      0.054  2.2943110	2.0124101 
   }

	bkg @ 1 0  ' add/remove coefficients as required
	
	Radius(35)

	str 
		CS_L(!csyo, 50)
		phase_name Y2O3
		MVW( 1, 1, 1)
		space_group I_A_3
		scale @ 9.92059e-008
		site Y1 num_posns 0 x  0.25 y  0.25 z  0.25 occ Y+3  1 beq  0.4545
		site Y2 num_posns 0 x  0.46756 y  0 z  0.25 occ Y+3  1 beq  0.394
		site O num_posns 0 x  0.10889 y  0.3489 z  0.11966 occ O-2  1 beq  0.7929
		Cubic(!ayo 10.60303 min =10.55; max =10.65;)

alancoelho

You could use pk_xo but that would mean rewriting the fwhm etc... in terms of channel. Easier in this case to load the raw data into excel and change the x-axis; ie.

New x-axis = ((ch - ch1) / (ch2 - ch1)) (x2 - x1) + x1

where ch = old x-axis (ie. channels)

ch1 = 1
ch2 = 512
x1 = 60
x2 = 14

You would then have a 2Th x-axis as before where zero error etc... can be applied

alancoelho

You could also use the program instead of Excel; ie.

XYE("det0.xye")

prm !ch1 1
prm !ch2 512
prm !x1 60
prm !x2 145

xdd_out new-file.xye load out_record out_fmt out_eqn
{
" %11.6f " = ((X - ch1) / (ch2 - ch1)) (x2 - x1) + x1;
" %11.6f " = Yobs;
" %11.6f\n" = SigmaYobs;
}

rowlesmr

Thanks for that Alan.

I'll have to go through and see what is more painful, changing the macros or pre-editing the files.

The main reason that I wanted to be able to do the channel conversion on the fly in Topas is that the detector position can be moved, and so I need to use a standard to get absolute 2Th values, and it would be nice to be able to refine this directly as just another xdd.

In that vein, what is the cause of the "no hkls" error?

Would I have to do some funkiness to the emission profile to get it to work in channel space?

alancoelho

>Would I have to do some funkiness to the emission profile
>to get it to work in channel space?

Editing the files is trivial and by far the easiest. Once edited you can refine the standard as normal. You could even set up a batch file with a INP file template for the file conversion.

rowlesmr

Roger that.

An idea of a work flow would be just to peak fit the standard peak in channel space to get its position, and then use that value to get Topas to do the channel-to-2T conversion with the known 2T position of the standard.

Once I'm in 2T space, I can refine the instrument parameters on the standard, and then apply them to the unknowns.

Now, to find some time to do it!

Thanks

Matthew