LabCanvas
Free account · save your work, full figure editorSign up free

XPS peak fitting online (C 1s, O 1s, N 1s deconvolution)

Paste a charge-corrected XPS region (binding energy and counts), list the chemical states you expect (for C 1s: C–C/C–H 284.8, C–O 286.3, C=O 287.8, O–C=O 289.0 eV) or let the tool find the peaks, and it fits them together with a linear background as pseudo-Voigt GL(30) peaks of one shared FWHM, the usual constraints for a C 1s region. You get each component’s binding energy, area, % of the total area and FWHM ± SE, drawn as filled components on a reversed binding-energy axis. Atomic ratios between elements also need the relative sensitivity factors.

What it does
Deconvolutes an XPS core-level region into chemical-state components (pseudo-Voigt, one shared FWHM, linear background), giving binding energies, areas, area percentages and FWHM ± SE.
Formula
I(E)=b0+b1E+∑kAk[ηLk(E)+(1−η)Gk(E)]η=0.3 (GL(30)), one FWHM%k=Ak/∑A
Example
The built-in C 1s region of an oxidised polymer (142 points) fitted with four GL(30) peaks of one FWHM gives C–C/C–H at 284.80 eV (59.2 %), C–O 286.29 eV (22.0 %), C=O 287.79 eV (8.6 %) and O–C=O 289.02 eV (10.2 %), FWHM 1.206 ± 0.003 eV, R² = 0.9997.

Last updated

FreeNo sign-upRuns in your browserData analysis

Loading XPS peak fitting…

At a glance

XPS peak fitting at a glance
InputBinding energy and counts of one region (CSV, TXT, XY)
ShapePseudo-Voigt GL(30) by default; Gaussian, Lorentzian, Voigt
ConstraintsOne FWHM for all components; positions from your list
BackgroundLinear, constant or quadratic, fitted with the peaks (no Shirley)
OutputBinding energy, area, % area, FWHM ± SE; 95% CIs; R², AIC
EngineThe LabCanvas Graph app’s peak fit (XPS preset), checked against lmfit
Runs inYour web browser; nothing is uploaded
PriceFree, no account needed

What it does

How to use the XPS peak fitting

  1. Paste or drop your XPS regionTwo columns, binding energy (eV) and intensity, for one core level, charge-corrected (for example adventitious C 1s at 284.8 eV).
  2. Choose the line shapePseudo-Voigt with η = 0.3 matches the GL(30) shape used in CasaXPS; leave η empty to fit it.
  3. List the chemical states or find peaksKeep the C 1s list or type your own positions with labels, one per line. The fit moves each centre; tick “one FWHM” off only with a reason.
  4. Check and reportLook at the residuals and the component positions, then copy the table of binding energies and area percentages.

How it works

I(E)=b0+b1E+∑kAk[ηL(E;ck,w)+(1−η)G(E;ck,w)]%k=Ak∑jAj

Components. Each chemical state of one core level is a peak at its binding energy. The usual constraints keep a fit physical: one FWHM for all components of a region (they share the instrument resolution and similar lifetimes), a fixed line shape, and positions that start from known chemical shifts. For C 1s in organic material, typical positions are C–C/C–H 284.8, C–O 286.3, C=O 287.8 and O–C=O 289.0 eV; the fit refines each centre.

Line shape. The pseudo-Voigt ηL+(1−η)G with one width approximates the Voigt profile of a core level; η = 0.3 corresponds to the GL(30) shape common in CasaXPS. Asymmetric metal and graphitic sp² peaks need an asymmetric shape this tool does not offer.

Background. The background here is a polynomial (usually a straight line) fitted with the peaks. The Shirley background (Shirley 1972), which rises in proportion to the integrated intensity at higher kinetic energy, is the standard for most regions and is not offered; over a narrow C 1s window with similar levels on both sides a linear background gives similar areas, but over wide regions or metal peaks use Shirley in your XPS software.

From areas to composition. Area percentages within one region are fractions of that element’s atoms in each state. Atomic ratios between elements need each area divided by its relative sensitivity factor (photoionisation cross-section, Scofield 1976, with the analyser transmission).

Same numbers as the Graph app. The page runs the LabCanvas Graph app's peak-fit analysis (its XPS preset), which is checked against lmfit; the hand-off opens the same table and fit there.

Sources: Shirley DA. Phys Rev B 1972;5:4709, doi:10.1103/PhysRevB.5.4709. Scofield JH. J Electron Spectrosc Relat Phenom 1976;8:129, doi:10.1016/0368-2048(76)80015-1. Newville M et al. LMFIT: non-linear least-squares minimization and curve-fitting for Python. Zenodo, doi:10.5281/zenodo.11813 (line shapes; area = amplitude). Olivero JJ, Longbothum RL. Empirical fits to the Voigt line width. J Quant Spectrosc Radiat Transf 1977;17:233–236, doi:10.1016/0022-4073(77)90161-3.

Worked example: C 1s of an oxidised polymer surface

The built-in example is a simulated C 1s region with noise, not a measurement: four components (C–C/C–H 284.8, C–O 286.3, C=O 287.8, O–C=O 289.0 eV) as GL(30) peaks of one FWHM on a linear background, 142 points.

Pseudo-Voigt GL(30) fit, one FWHM, linear background (value ± SE)
ComponentBinding energy (eV)Area% of area
C–C/C–H284.798 ± 0.0017367 ± 1759.2
C–O286.289 ± 0.0042738 ± 1322.0
C=O287.794 ± 0.0121066 ± 148.6
O–C=O289.017 ± 0.0091263 ± 1510.2
  1. Shared width. All four share FWHM = 1.206 ± 0.003 eV, so the fit has 4 centres, 4 areas, one width and two background terms: 11 parameters for 142 points.
  2. Chemical shifts. Relative to C–C at 284.80 eV, C–O is +1.49 eV, C=O +2.99 eV and O–C=O +4.22 eV.
  3. Composition. 59.2 % of the carbon is unoxidised; 22.0 + 8.6 + 10.2 = 40.8 % is bound to oxygen.
  4. Fit quality. R² = 0.9997 with residuals scattered around zero.

Frequently asked questions

How do I fit an XPS C 1s spectrum?

Charge-correct the energy scale, subtract a background, and fit components at the expected chemical states (C–C/C–H 284.8, C–O about 286.3, C=O about 287.8, O–C=O about 289.0 eV) with one shared FWHM and a GL(30) line shape. Check that the residuals are flat and the positions are chemically sensible.

What is GL(30)?

A product or sum of Gaussian and Lorentzian line shapes with 30 % Lorentzian character, the default in CasaXPS. Here it is the pseudo-Voigt sum with η = 0.3.

Should all components have the same FWHM?

Usually yes for the chemical states of one core level measured in one region: they share the instrument resolution. Free widths let the fit trade width for area and give unphysical results; relax the constraint only with a reason, such as a differently charged phase.

Which background should I use for XPS?

Shirley for most core levels, Tougaard for quantitative work over wide ranges, linear for narrow regions with flat levels on both sides. This tool fits a linear (or quadratic) background with the peaks; for a Shirley background use your XPS software.

How do I get atomic percentages from XPS peak areas?

Within one region, the area percentages are the fractions of that element in each chemical state. Between elements, divide each region’s area by its relative sensitivity factor, then normalise the results to 100 %.

How do I charge-correct XPS data?

Shift the whole binding-energy scale so a reference peak sits at its accepted value, most often adventitious carbon C–C/C–H at 284.8 eV, and apply the same shift to every region of that sample.

Make the XPS fit publication-ready

Your data from this page opens in the LabCanvas Graph app with the same analysis, as an editable graph you can restyle, size to a journal column and place in a figure. The Graph app is free to use with a free LabCanvas account.

Open the Graph app