XPS分峰

First some questions back to you and some fundamental comments.
- The above 2 curves from journal paper or they are your data?
- The fit on the right spectrum, there is a peak at ~289eV. That peak seems to have much bigger FWHM than other 2 peaks?
- For C1s peak fitting, usually you would expect "similar" FWHM amongst all fitted peaks, e.g. C-O, C=O, -COO >> BUT important is usually C-C/C-H peak is a little fatter (bigger FWHM) than other C1s-chemical state peaks. It is because this peak actually contain 2 forms (C-C and C-H) which cannot resolve by XPS, so due to this overlap this peak is a bit fatter always.


Next is answers to your questions:
- Using the above 2 spectrum as example, I think we can start with shifting the righter-most peak to 284.8eV as assigning it to be C-C/C-H peak.
- From there onward, you can start with for example the left spectrum and perform fitting. The fit shown above left spectrum looks quite reasonable. Make sure the FWHM are similar with C-C/C-H peak could be with a little bigger FWHM.
- Then for right spectrum, you can try using same 4 peaks to fit. From above fit with 3 peaks, the peak at 289eV seem fatter than C-C/C-H peak > which mentioned above base on fundamental it may indicate the fit is not correct. There should be a 4th peak inside this C1s peak.
- The fitted 4 peaks parameters should be same if you are assigning them to be same 4 chemical state. So expect them to be exact SAME peak position and do NOT allow even a few eV difference. (This is ideal case, but if smaller than 0.1~0.2eV difference can make better fit it can be tolerable but it also means your data is not good. It means if your data is good it should be no problem for fitted 4 peaks to fall at exactly same peak energy)
- You should NOT perform SMOOTH on the spectrum. If the spectrum is not smooth, it means your original data S/N is bad so you should go back and re-run your analysis to get good enough S/N spectrum. After SMOOTH, data can be a little change so always NOT suggested.

上面两个图是我自己的数据，都是分成4个峰。

右边那个图也是4个峰，在287.7ev处有一个峰，比较弱（见下图红色位置）；在288.8ev那个峰的FWHW是2.32ev(在1-3ev范围都正常吧)。

我遇到的问题是：左边有个峰在286.2ev,右边对应的位置却在286.9ev,怎么都拟合不到同一个位置上。

I understand your questions.

Let me ask some questions:

1. Are these 2 data using same analysis parameters (e.g. X-ray spot size, Pass Energy, step size.... all same)???

2. I hope you understand that your target should NOT just think you want to "fit 4 peaks". First, for example start from left spectrum, the fit result gave 4 peaks. Can you comment each of the 4 peaks Chemical state? You should be able to assign Chemical state of each of these 4 peaks if you believe this fitted result is correct.
> Then next is if exactly same 4 Chemical state of Carbon (you assigned from left spectrum sample) also confirm to appear on the other sample (right spectrum). If the answer is YES, then you should be able to use same fitting parameters to fit the curve.
***But from your data, the spectrum shape seems quite different between the 2 samples.
> After all, the objective for doing curve fit = separating overlapping peaks and in your case the overlapping peaks meant for different Chemical state of Carbon. So you need to assign them first and prove your fitting is making sense.

3. Can you attached your raw data files? What software do you use to fit the spectrum? I can possibly try to fit your data too.

两个样品是同一次测得，测试条件应该是一样的；
图1是氧化石墨烯，图2是石墨烯（还原），我分峰的目的是想说明氧化石墨烯碳的含氧基团在还原后含量大大减少；图1的4个峰分别为284.8ev（C=C）;286.9ev（C-O）;287.7ev（C=O）;288.8ev（COOH）图2的4个峰分别为284.8ev（C=C）;286.2ev（C-O）;287.7ev（C=O）;288.8ev（COOH）
XPS分峰软件：XPSPEAK41
原始数据见附件

寻求高手解答

I do not know why but I cannot download your data files (shown as 未验证). Anyway, base on your Chemical state assignment, I try to give more of my suggestion:

When assigning Chemical, my experience for C1s peak is:
C-C = 284.8eV Binding Energy (BE)
C-O > +1.5eV shifting to high BE
C=O > +2.75~3.2eV shifting to high BE
-COO > ~3.95~4.4eV shifting to high BE

Looking back to your spectrum.

(1) First on left spectrum above
It seems to have high C=O bonding peak (the highest peak), and it is about ~2.5V shifting from main C-C peak (so quite close to above C1s Chem state shift). I think there exist some C-O in this spectrum too. So I think you need to make curve fit of this with peaks at:
284.8 C-C
286.3 C-O
~287.5 C=O
~289 -COO (COOH)

(2) Then to the right spectrum
If you believe they should carry the same 4 Chemical state and both data run at same analysis condition, you should use same 4 peaks to fit the spectrum (fixed the peak position).

Honestly speaking I am not expert in your material but from your spectrum, your left spectrum is showing much higher C-O and C=O, and also a little more -COO also. While right spectrum seems more of the C1s is of C-C/C-H bonding.

非常感谢！我重新拟合了一下，左边那个图在286.3ev的那个峰还是拟合不出来（我拟合的286.8ev），你能把邮箱留给我一下吗？我把原始数据发给你，你帮我看看。

Okay.
Please email sactup@163.com

受益匪浅，学习了