# Analysis Problems with calibrate shifts in energies

1. Feb 12, 2015

### cenicitas

Hello! Well this is a quite embarrassing problem but I think my maths has disappeared the last years. So I"m working on some calculations and i want to compare with the expermental data but there's a shift on energy that it has to be apply. So I thought I just need to to a simple calibration (experimental data vs calculated data and then a did a fit with origin where a get y= a+bx)
Now if I'm doing this..its not like i'm just applying a shif of the same X number to everything (let's says i'm not doing shift= experimental data + X), so i'm changing the espectra.
I guess I'm doing something wrong?

2. Feb 12, 2015

### Quantum Defect

If you are calibrating a spectrum, you should try to have a number of landmarks that are used to calibrate the data. If you have only a single landmark, you can use a simple offset, but the calibration will get worse, the farther you get away from the landmark.

It is important that you have complete confidence that your calbration peaks are what you say they are.

3. Feb 12, 2015

### cenicitas

sorry I'm not exactly sure what are you saying. So the calibaration that I'm doing is wrong?

4. Feb 12, 2015

### Quantum Defect

If you only have one data point to calibrate against, then what you are doing may be the best that you can do.

5. Feb 12, 2015

### cenicitas

that's what I though but, again if my "energy shift" is given by y= a+bx, then i'm not just shifting the function, i'm changing it since i'm not modifying the y data.. Well so I'm talking about k-ege for Fe and td dft calculations (it's reported that the shift should be close to + 180 ev)

6. Feb 12, 2015

### Quantum Defect

Could you maybe upload a picture of what you have, and what you are doing? It sounds like you aren't just correcting for an instrumental calibration error, but are rather trying to compare experimental and computational results.

My experience is in fairly high resolution gas-phase, molecular spectroscopy.

7. Feb 12, 2015

### cenicitas

So I'm just starting a course and trying to learn reproducing stuff already done, here for example the have some exp. data and did some tddft calculations, what given values for the energy that requiered to be shifted.

Fig. 5 Experimental (top) and calculated (bottom) Mo K-pre-edge of the various compounds investigated in this study. For clarity, the plots show the measured HERFD-XAS spectra subtracted from all the contributions except the pre-edge features (derived from the fits). The calculations used the BHLYP functional with DKH2 relativity correction. A constant shift of 60.9 eV and a broadening of 3.5 eV were applied to all calculated spectra. The calculations using other functionals produce similar results.

So this constant shift of 60.9 , I don't know how they determine that value :(

8. Feb 12, 2015

### Quantum Defect

Density functional theory gives some values that are very good, while others are known to have a consistent systematic bias. In my area of research, DFT gives vibrational frequencies of molecules that are all skewed in a single direction. When looking for a molecule's spectrum, you often apply a "fudge factor" to the DFT numbers to get a ball-park estimate of what the actual values should be. It sounds like the 60.9 eV shift to the DFT values above might be something along these lines. The particular value might be picked to simply line things up, or it may be based upon a known systematic problem with calculating the spectra in this region with this flavor of DFT.

9. Feb 12, 2015

### cenicitas

but, how can I calculate this value?

10. Feb 12, 2015

### Quantum Defect

I don't think that this is a calculated number. It is either a value that is obtained from comparing the experiment and theory and coming up with a shift to bring things into agreement (basically shifting the axis horizontally to get the broad peaks to match up). Or, it could be a number that is tabulated somwhere for practitioners to use when comparing Mo K-edge spectra with DFT theory. Read the section of the paper a bit more carefully where they discuss the results. If it is the latter, they will give a reference, if it is the former, they may say how they determined the best number for the shift and the width parameters (i.e. least squares fit to the experimental data, by eyeball, etc.)