[Ifeffit] Artemis parameters

[Scott Calvin]

Hi Abhijeet,

The guess values are simply the starting point for ifeffit to work  
from, so they don't matter as long as they're "reasonable," which the  
default values are, at least if your normalization and background  
subtraction is OK.

But more importantly, the default parameters themselves are not  
appropriate for more than one shell of copper foil! If you're using  
more than one path, you need all the paths to expand by the same  
fraction, not the same number of angstroms. Therefore you want to  
guess a parameter (let's call it "alpha") and make the delr for each  
path alpha*reff. (ifeffit knows that reff means half the nominal  
length of that path.) Likewise, it's probably inappropriate to use the  
same sigma2 for all paths, but fortunately for metals the Debye model  
works fairly well. You can have a guessed variable for the Debye  
temperature (say, "omega"), and then have sigma2 for each path be  
Debye(temp, theta), where "temp" is the temperature you measured the  
data at.

As far as the ideal fitting range in R for copper foil, you can  
usually manage out to 5 angstroms or so with a metal foil without too  
much difficulty.

--Scott Calvin
Sarah Lawrence College

On Jan 29, 2009, at 5:11 AM, abhijeet gaur wrote:

> Hi
>      I am fitting a copper foil whose data is taken at room  
> temperature. So do I need to change the guess value of parameters  
> enot, ss,amp, delr to some other value or I should take them as the  
> default guess value. Also what should be the ideal fitting range in  
> R for the copper foil.
>
>                                                                                                                     With 
>  thanks
>                                                                                                                      Abhijeet