| Findings |
Reason |
Line
of action |
Example |
Correspondence
between experimental spectra and calculated sum curve
unsatisfying even after several iterations
|
System
sticks in a local minimum |
Disturb
the system by varying start parameters sufficiently
strong |
Doubling
of intensity |
| Convergence
incomplete |
Min/Max
values to close to the fit parameters
|
Select
larger Min/Max range |
Increase
Max value for Lorentzian width from 2 to 2.5 |
| Mismatch
between flanks of experimental spectra and calculated
sum curve |
Selected
peak shape doesn't describe the experimental function
accurately |
Select
[Fit Procedure - Convolution] |
3d-
and 4f-signals of transition metals with high resolution
(comparable contribution of Gaussian and Lorentzian
function)
|
| Mismatch
between experimental spectra and calculated sum curve
in the region of peak tails |
Background
subtraction before starting peak fit |
Fit
baseline together with the peak |
Signals
with strong Lorentzian contribution
|
|
|
|
|
