[Ifeffit] Fitting Commercial Molybdenum Disulfide

Michael A Groves dingoboy at u.washington.edu
Thu Oct 28 13:00:13 CDT 2004


Dan, Everyone else...

I have two suggestions for optimizing the fit.  First, from looking at the crystal structure of the MoS2, I would probably not use the same sigma-squared for all three paths.  The crystal structure I looked at online had the Mo connected to various Sulfur atoms, but not directly connected to the other Mo atoms, for example.  If this is the case, the motion of the Mo and its neighboring S atoms should be much more highly correlated than the motion of th Mo and the nearest Mo atoms-- So, the Mo-S path should have a smaller debye-waller factor than the Mo-Mo path.

Second, you're using alpha*reff for your delr term.  Are you sure that the only difference between your model and your actual structure is an isotropic expansion?  If bond angles change you won't be able to make that assumption, and from looking at the MoS2 structure, that seems likely.

So, in summary, I'd use a different sigma-squared term for each path, and try to make a more sophisticated guess about the form of the delr terms.

Hope this helps.
Mike Groves

On Thu, 28 Oct 2004 dmc at pdx.edu wrote:

>
> Hi Mailing list,
>
>        I'm back at fitting commercial Molybdenum Disulfide to a Hexagonal
> p 63/m m c space group.  Although my present fit is reasonable, thanks to the
> mailing list in August,  I was hoping someone might have more suggestions to
> optimize it.  Also, in order to better my fit, I still have seperate e0's, one
> for the first path, one for the remaining 2 paths.  I haven't gotten a
> satisfactory physical explanation for this and I thought someone might have a
> good one.
>
>        Also, the data on the same sample from two different beamlines seems to
> be somewhat different, namely the Reduced chi-square is very different
> (75.6 in one,  636.4 in the other).  In both cases I use the first 3 single
> scattering paths.  I may have included too much info, but better too much than
> too little.
>
>      Thank You, Dan Carter
>
>
>
>        Here are my parameters in both cases, values differ somewhat:
>
> Set:     Amp =  0.7
> guess:   eO     ~  1.57  (1.52)  (used for first path)
> guess:   eO_1   ~  -4.77  (2.43)  (used for remaining 2 paths)
> guess:   ss     ~  0.0027   (0.000427)
> guess:   alpha  ~ -0.004   (0.003255)     (positive for second fit)
> delr_1:  alpha*reff
>
>
>
>
> First Fit:
>
>
> Project title :  Fitting merge 10_15_04.chi
> Comment       :
> Prepared by   :
> Contact       :
> Started       :  11:49:36 on 22 October, 2004
> This fit at   :  11:10:57 on 28 October, 2004
> Environment   :  Artemis 0.7.010 using Windows 2000, perl 5.006001, Tk 800.024,
> and Ifeffit 1.2.6
>
> ============================================================
>
> Independent points          =      14.312500000
> Number of variables         =       4.000000000
> Chi-square                  =     779.303886008
> Reduced Chi-square          =      75.568861674
> R-factor                    =       0.041210572
> Measurement uncertainty (k) =       0.000827712
> Measurement uncertainty (R) =       0.022111003
> Number of data sets         =       1.000000000
>
>
> Guess parameters +/- uncertainties  (initial guess):
>  e0_1            =    -4.7737110   +/-      2.4268840    (guessed as -4.773687
> (2.426889))
>  ss              =     0.0027360   +/-      0.0004270    (guessed as 0.002736
> (0.000427))
>  e0              =     1.5684690   +/-      1.5209160    (guessed as 1.568507
> (1.520918))
>  alpha           =    -0.0040060   +/-      0.0032550    (guessed as -0.004006
> (0.003255))
>
> Def parameters:
>  delr_1          =    -0.0158980
>
> Set parameters:
>  amp             =  0.7
>
>
> Correlations between variables:
>        e0_1 and alpha      -->  0.8622
>          e0 and alpha      -->  0.7870
>        e0_1 and e0         -->  0.7064
> All other correlations are below 0.25
>
>
>
> Project title :  Fitting merge 10_15_04.chi
> Comment       :
> Prepared by   :
> Contact       :
> Started       :  11:49:36 on 22 October, 2004
> This fit at   :  11:10:57 on 28 October, 2004
> Environment   :  Artemis 0.7.010 using Windows 2000, perl 5.006001, Tk 800.024,
> and Ifeffit 1.2.6
>
> ============================================================
>
> Independent points          =      14.312500000
> Number of variables         =       4.000000000
> Chi-square                  =     779.303886008
> Reduced Chi-square          =      75.568861674
> R-factor                    =       0.041210572
> Measurement uncertainty (k) =       0.000827712
> Measurement uncertainty (R) =       0.022111003
> Number of data sets         =       1.000000000
>
>
> Guess parameters +/- uncertainties  (initial guess):
>  e0_1            =    -4.7737110   +/-      2.4268840    (guessed as -4.773687
> (2.426889))
>  ss              =     0.0027360   +/-      0.0004270    (guessed as 0.002736
> (0.000427))
>  e0              =     1.5684690   +/-      1.5209160    (guessed as 1.568507
> (1.520918))
>  alpha           =    -0.0040060   +/-      0.0032550    (guessed as -0.004006
> (0.003255))
>
> Def parameters:
>  delr_1          =    -0.0158980
>
> Set parameters:
>  amp             =  0.7
>
>
> Correlations between variables:
>        e0_1 and alpha      -->  0.8622
>          e0 and alpha      -->  0.7870
>        e0_1 and e0         -->  0.7064
> All other correlations are below 0.25
>
>
>
>
> ===== Data set >>merge 6_17_04.chi<< ========================================
>
>  file: C:\Ifeffit\horae\stash\artemis.project.3\chi_data\merge 6_17_04.chi
>  title lines:
>    Athena data file -- Athena version 0.8.032
>    Saving  merge 6_17_04 (group=merge) as chi(k)
>    .  Element=Mo   Edge=K
>    Background parameters
>    .  E0=20007.492  Eshift=0.000  Rbkg=1.000
>    .  Standard=None
>    .  Kweight=2.0  Edge step=0.749
>    .  Fixed step=no    Flatten=yes
>    .  Pre-edge range: [ -150.000 : -75.000 ]
>    .  Normalization range: [ 150.000 : 895.199 ]
>    .  Spline range: [ 0.952 : 995.207 ]   Clamps: None/Strong
>    Foreward FT parameters
>    .  Kweight=2   Window=kaiser-bessel   Phase correction=no
>    .  k-range: [ 2.000 : 16.100 ]   dk=0.00
>    Backward FT parameters
>    .  R-range: [ 1.000 : 3.000 ]
>    .  dR=0.00   Window=kaiser-bessel
>    Plotting parameters
>    .  Multiplier=1   Y-offset=0.000
>    .
>    Merge in k space of:
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.101 (mos2_101)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.102 (mos2_102)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.103 (mos2_103)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.104 (mos2_104)
>
>  k-range             = 2.000 - 13.500
>  dk                  = 2.000
>  k-window            = kaiser-bessel
>  k-weight            = 2
>  R-range             = 1.000 - 3.000
>  dR                  = 0.100
>  R-window            = kaiser-bessel
>  fitting space       = R
>  background function = none
>  phase correction    = none
>
>
>  R-factor for this data set   = 0.74613
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> Second Fit:
>
>
> Project title :  Fitting merge 10_15_04.chi
> Comment       :
> Prepared by   :
> Contact       :
> Started       :  11:49:36 on 22 October, 2004
> This fit at   :  11:23:36 on 28 October, 2004
> Environment   :  Artemis 0.7.010 using Windows 2000, perl 5.006001, Tk 800.024,
> and Ifeffit 1.2.6
>
> ============================================================
>
> Independent points          =      14.312500000
> Number of variables         =       4.000000000
> Chi-square                  =    6562.747150466
> Reduced Chi-square          =     636.387602469
> R-factor                    =       0.032979787
> Measurement uncertainty (k) =       0.000181935
> Measurement uncertainty (R) =       0.004860095
> Number of data sets         =       1.000000000
>
>
> Guess parameters +/- uncertainties  (initial guess):
>  e0_1            =    -4.0720090   +/-      1.9194040    (guessed as -4.773711
> (2.426884))
>  ss              =     0.0046360   +/-      0.0004110    (guessed as 0.002736
> (0.000427))
>  e0              =     1.7659580   +/-      1.1543880    (guessed as 1.568469
> (1.520916))
>  alpha           =     0.0010900   +/-      0.0028430    (guessed as -0.004006
> (0.003255))
>
> Def parameters:
>  delr_1          =     0.0043250
>
> Set parameters:
>  amp             =  0.7
>
>
> Correlations between variables:
>        e0_1 and alpha      -->  0.8416
>        e0 and alpha      -->  0.8045
>        e0_1 and e0         -->  0.7141
> All other correlations are below 0.25
>
>
>
>
> ===== Data set >>merge 10_15_04.chi<< ========================================
>
>  file: C:\Ifeffit\horae\stash\artemis.project.3\chi_data\merge 10_15_04.chi
>  title lines:
>    Athena data file -- Athena version 0.8.032
>    Saving  merge 6_17_04 (group=merge) as chi(k)
>    .  Element=Mo   Edge=K
>    Background parameters
>    .  E0=20007.492  Eshift=0.000  Rbkg=1.000
>    .  Standard=None
>    .  Kweight=2.0  Edge step=0.749
>    .  Fixed step=no    Flatten=yes
>    .  Pre-edge range: [ -150.000 : -75.000 ]
>    .  Normalization range: [ 150.000 : 895.199 ]
>    .  Spline range: [ 0.952 : 995.207 ]   Clamps: None/Strong
>    Foreward FT parameters
>    .  Kweight=2   Window=kaiser-bessel   Phase correction=no
>    .  k-range: [ 2.000 : 16.100 ]   dk=0.00
>    Backward FT parameters
>    .  R-range: [ 1.000 : 3.000 ]
>    .  dR=0.00   Window=kaiser-bessel
>    Plotting parameters
>    .  Multiplier=1   Y-offset=0.000
>    .
>    Merge in k space of:
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.101 (mos2_101)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.102 (mos2_102)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.103 (mos2_103)
>    ** C:/USERS/Dan2/raw data MoS2 6_17_04/MoS2.104 (mos2_104)
>
>  k-range             = 2.000 - 13.500
>  dk                  = 2.000
>  k-window            = kaiser-bessel
>  k-weight            = 2
>  R-range             = 1.000 - 3.000
>  dR                  = 0.100
>  R-window            = kaiser-bessel
>  fitting space       = R
>  background function = none
>  phase correction    = none
>
>
>  R-factor for this data set   = 0.71585
>
>
> Paths used to fit Both Data Sets, degen, s02 the same.
>                               other values different:
>
>  F1 p 63/m m c: feff0001.dat ..
>    feff   = C:\Ifeffit\horae\stash\artemis.project.3\data0.feff2\feff0001.dat
>    id     = reff= 2.4135, nlegs= 2, path= Mo<->S
>    label  =
>    r      =    2.416130
>    degen  =    6.000000
>    s02    =    0.700000
>    e0     =    1.765958
>    dr     =    0.002630
>    ss2    =    0.004636
>
>  F1 p 63/m m c: feff0002.dat ..
>    feff   = C:\Ifeffit\horae\stash\artemis.project.3\data0.feff2\feff0002.dat
>    id     = reff= 3.1500, nlegs= 2, path= Mo<->Mo
>    label  =
>    r      =    3.153433
>    degen  =    6.000000
>    s02    =    0.700000
>    e0     =   -4.072009
>    dr     =    0.003433
>    ss2    =    0.004636
>
>  F1 p 63/m m c: feff0003.dat ..
>    feff   = C:\Ifeffit\horae\stash\artemis.project.3\data0.feff2\feff0003.dat
>    id     = reff= 3.9683, nlegs= 2, path= Mo<->S
>    label  =
>    r      =    3.972625
>    degen  =    6.000000
>    s02    =    0.700000
>    e0     =   -4.072009
>    dr     =    0.004325
>    ss2    =    0.004636
> _______________________________________________
> Ifeffit mailing list
> Ifeffit at millenia.cars.aps.anl.gov
> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>





More information about the Ifeffit mailing list