[Ifeffit] Fwd: Re: why ss_2 is negative?

[Bruce Ravel]

On Friday 23 February 2007 21:41, hw26 at njit.edu wrote:
> Yes, I am modeling the uranyl ligand and two equatorial ligands, one of
> which is hypothesized as U-F (from BF4 anion), and the other as U-O (from
> hydration),because in the uranyl nitrate standard sloution(without
> MOEMIMBF4), there is only one peak that represents the U-O equatorial
> ligand, but with the MOEMIMBF4, the equatorial peak is split to two peaks.
> So I guess the other equatorial U-F formed.

Hao,

A few more comments:

1.  It is well advised for anyone starting to look at uranyl compounds
    with XAS to read Shelly's various publications on that topic.
    Geochem. Cosmo. Acta, 66(22) 3875-3891, (2002) is particularly
    useful.  Google scholar can help you locate more of her papers.

2.  I am the first to admit that Artemis and Ifeffit are difficult
    tools for the beginning exafs practitioner.  These tools try to
    provide a usable face to a difficult problem.  Ifeffit was written
    with enough flexibility and power to allow the so-called experts
    to do the analysis they want to do.  That means that Artemis has
    enough freedom to allow the unsuspecting user to do ill-advised
    things.

    You really need to think about what each of the parameters means.
    For example, just because Artemis allows you to float e0 freely
    for each path does not mean that doing so is a good idea.  There
    are situations where it is a very good idea to have more than one
    e0 parameter in your fit, but I suspect that yours is not one of
    them.  Consequently, you should consider constraining all your
    paths to have the same e0.

    What do I mean by "constrain"?  Well, on the Guess,Def,Set page,
    you can define a single e0 parameter as a guess parameter, then
    use that parameter on each of the path pages.  In that way, you
    will have one floating parameter in your fit that is used to
    describe the e0 of all your parameters.

    Similarly, you should think hard about the numbers of each kind of
    path.  A uranyl moiety is unlikely to have more or less than 2
    uranyl ligands.  The uranyl bond is very strong, consequently it
    is energetically unfavorable in a uranyl for those two bonds to be
    over- or under-filled.  Consequently, you can consider setting the
    number of yl oxygens to 2.  Furthermore, the equatorial plane is
    likely to have between 4 and 6 oxygens, so you should think of
    ways to constrain the total number of equatorial oxygens to be
    something reasonable.  Since the coordination number is highly
    correlated with sigma^2 in a statistical sense, thinking well
    about coordination number should help address your original
    question.

3.  Scott has made a very nice Artemis tutorial which can be found at
       http://cars9.uchicago.edu/iffwiki/HoraeSoftware#contrib
    It comes in the form of a series of Artemis projects.  Newcomers
    benefit by working through his tutorial before trying to analyse
    their own data with Artemis.

4.  The Ifeffit web site, my home page, and especialy http://xafs.org
    are all filled with useful information for the XAS novice.  If you
    haven't already looked through the resources on those sites, you
    really should budget some time to do so.


HTH,
B


-- 
 Bruce Ravel  ---------------------------------------------- bravel at anl.gov

 Molecular Environmental Science Group, Building 203, Room E-165
 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007

 Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
 Argonne IL 60439, USA                                fax: (1) 630 252 9793

 My homepage:    http://cars9.uchicago.edu/~ravel 
 EXAFS software: http://cars9.uchicago.edu/~ravel/software/