stacking fault parameter of a_add/b_add

hanyifeng

Hi Alen,

I am studing stacking fault example which named fit-2.inp(list below) in Topas-academic v6.
I want to know how to determing the numerical of a_add and b_add.

Sincerely

Yifeng Han


      prm !pa 0.7
      seed
      generate_stack_sequences {
			save_sequences seq.txt
			'load_sequences seq.txt
         number_of_sequences = Nseq;
         number_of_stacks_per_sequence = Nv;
			prm !h5 5
			prm !h6 6
			Transition(1, h5) 
            to 1 = pa;     a_add = 2/3;   b_add = 1/3;  
            to 2 = 1-pa;   a_add = 0;     b_add = 0;    
         Transition(2, h6)                           
            to 2 = pa;     a_add = -1/3;  b_add = -1/3; 
            to 1 = 1-pa;   a_add = 2/3;   b_add = 0;    
      }

alancoelho

Hi Yifeng

The a_add and b_add keywords determine the offset of the layers in the a and b directions. Knowing what these offsets should be is aim of stacking fault analysis. The user need to guess these or drive the program in a batch mode (or using the #list command) to determine the stacking vectors that best fits the data.

cheers
alan

hanyifeng

Dear Alan,

Recently, I read the manual file of DIFFaX. But I still confused of how to get the probability of Layer A to Layer B.
1.Can you give us a tutorial of how to determing the probability and give a more complex example (like LiNbO3 it contain 12 layers).
2.We need do the Rietveld refinment but there are stacking fault in our pattern, what order should I take to deal with my pattern? (Rietveld first or stacking fault first? If I take stacking fault first, will this affect the accuracy of my Rietveld result?)

Sincerely

Yifeng

alancoelho

>I still confused of how to get the probability of Layer A to Layer B.

Do a series of refinements with the probability of A ranging from 0 to 1. Use #list to set up the INP file to make it easy.

The test_examples\stacking_faults\fit-1.inp has the following:

prm !pa 0.7
generate_stack_sequences {
number_of_sequences Nseqs 200
number_of_stacks_per_sequence Nv 200
num_unique_vx_vy 6
Transition(1, lpc)
to 1 = pa; a_add = 2/3; b_add = 1/3; n !n1 0
to 2 = 1-pa; a_add = 0; b_add = 0; n !n2 0
Transition(2, lpc)
to 1 = 1-pa; a_add = 0; b_add = 0; n !n3 0
to 2 = pa; a_add = -2/3; b_add = -1/3; n !n4 0
}

Here we see one stacking parameter pa and it cant be refined. Setup a similar transition matrix for your problem (this is for the scientist to do) and then do a series of refinement with pa ranging from 0.01 to 0.99 and then plot something like Rwp versus pa. Where to go from there is the science of the problem.

>what order should I take to deal with my pattern

Try Rietveld refinement first. If the fit is perfect then don't go further in regards to stacking faults.

johnsoevans

There are also some tutorials which relate directly to the diffax examples which might help you:

http://community.dur.ac.uk/john.evans/topas_workshop/tutorial_stacking_faults.htm