Hi again, this email concerns the spin.f program.
I do not understand, how you can get SPXAS at N, since the spin on N is zero. I would use Cr K or L3 edge instead.
Although there is no spin on the nitrogen, there are two distinct magnetic environments, arising from the spin orientation of the surrounding chromiums. Consequently, SPIN 2 generates a different N K-edge from SPIN -2, so i have two different xmu.dat files. I use spin.f to obtain an 'overall' K-edge.
> elseif (icase.eq.3) then > c factor=0.5 always for SPXAFS > t1 = (y1*ap - z1*am)/2.0/xnorm > t2 = (y2*ap - z2*am)/2.0/xnorm > t3 = (y3*ap - z3*am)/2.0/xnorm > > which are obviously subtractions, which will give negative values if, say, > z1>y1. > Does this explain the origin of the 'negative' peak, and if so, i assume > its perfectly reasonable to convert all negative numbers generated to > positive values? No, it is not reasonable. You have to adjust the spin.f to correspond to the measured signal. First, since you measure N K-edge, I do not understand how you can get spin-up and spin-down signals at all.
The "spin-up" N K-edge comes from xmu.dat with SPIN 2, the "spin-down" N K-edge comes from xmu.dat with SPIN -2. There are two unique magnetic sites for the N atom, although it itself does not have spin. This comes from the fact that an octahedrally coordinated N atom has 4 Cr-up spin neighbours and 2 Cr-down spin neighbours. The relative spin orientations specified by the SPIN card mean a different edge shape is obtained each time.
The spin.f assumes that one wants to know the DIFFERENCE between spin-up and spin-down signal. The difference obviously can have either sign.
I think we're talking about the same thing here, that spin.f could come up with a negative sign difference that would be have to be converted to positive for a sensible plot, i.e. the modulus of the difference is required.
I've attached the unmodified plot obtained from spin.f. Cheers again, Dave Eustace
Attachment:
Doc1.doc
Description: MS-Word document