XPS Detection Limits

In general, detection limits for XPS range from 0.1 to 1 atomic percent.  However there are cases where limits could be much better or much worse.  An article from Alexander Shard [1] gives an excellent look at detection limits in over 6000 binary systems for both Al and Mg X-ray sources.  In such systems a heavy element, such as gold, in a light element matrix, say carbon, would have detection limits closer to 0.01 atomic percent.  For the opposite situation, carbon in a gold matrix, detection limits for carbon would be around 3 atomic percent.  Spectral overlaps can also degrade detection limits significantly and are incorporated into the tables presented.  Overlaps with Auger peaks can be overcome by changing X-ray sources used (i.e. using Mg K instead of Al K).  One can also work with advanced curve-fitting techniques to overcome some of these overlap issues.

Download printable PDF of detection limits for Al X-ray spectra.

Download printable PDF of detection limits for Mg X-ray spectra.

XPS detection limits using Al K(alpha) radiation [1].

Reference:
[1] A.G. Shard, Surf. Interface Anal. 46 (2014) 175-185.

Carbon 1s for Organic Compounds

The seminal work of Graham Beamson and Dave Briggs in their "High Resolution XPS of Organic Polymers – The Scienta ESCA300 Database" has been utilized since 1992 as an invaluable resource for the XPS analysis of polymers and organic materials.  A summary of carbon 1s binding energies for organic functional groups from this work are presented here. The original work calibrates the binding energy scale to 285.0 eV for aliphatic carbon C 1s. The values presented here are now calibrated to 284.8 eV for aliphatic carbon, in line with recent results [2].  

Figure 1. Summary of the mean, maximum, and minimum carbon 1s binding energies for different organic functionalities according to the work of Beamson and Briggs [1].  Binding energy calibration presented here have been adjusted to the main aliphatic C 1s peak at 284.8 eV.

Table 1. Summary of the mean, maximum, and minimum carbon 1s binding energies for different organic functionalities according to the work of Beamson and Briggs [1].  Binding energy calibration presented here have been adjusted to the main aliphatic C 1s peak at 284.8 eV.

Table 2. Summary of beta shift effects from Beamson and Briggs (Appendix 2) [1].

The effects of various functional groups on beta carbon binding energies can be significant (Table 2).  Note that, in this context, the alpha carbon is the carbon directly attached to the functional group, and the beta carbon is attached to the alpha carbon. These effects have been included in the refinement of the binding energy value for the aliphatic carbon component in adventitious carbon [2]. 

References:
[1] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers - The Scienta ESCA300 Database, Wiley Interscience, 1992, Appendices 1 and 2.
[2] L.H. Grey, H.-Y. Nie, M.C. Biesinger, Appl. Surf. Sci. 653 (2024) 159319.

Oxygen 1s for Organic Compounds

Figure 1. Summary of the mean, maximum, and minimum oxygen 1s binding energies for different organic functionalities according to the work of Beamson and Briggs [1].  Binding energy calibration presented here have been adjusted to the main aliphatic C 1s peak at 284.8 eV [2].

Figure 2. Oxygen 1s binding energy means and ranges for various organic compound types. Plotted from data supplied in Beamson and Briggs [1]. Referenced to main aliphatic C 1s peak at 285.0 eV as in the original source data.

Additional Notes: 
C-OH (aliphatic) Ref to C 1s at 284.8 eV: Average 532.7 eV, Min. 532.5 eV, Max. 532.9 eV

C-OH (aliphatic) Ref to C 1s at 285.0 eV: Average 532.9 eV, Min. 532.7 eV, Max. 533.1 eV

C-OH (aromatic) Ref to C 1s at 284.8 eV: 533.4 eV 
C-OH (aromatic) Ref to C 1s at 285.0 eV: 533.6 eV 

Also note that Si 2p3/2 for PDMS (silicone) is at 101.79 eV (Si 2p = 102.0 eV) with the C 1s at 284.38 eV and O 1s at 532.00 eV (referenced to aliphatic C at 285.0 eV).  If we shift the C 1s to 285.0 eV then Si 2p3/2 is at 102.41 eV (Si 2p = 102.6 eV) and O 1s is at 532.62 eV for silicone. If we then shift the C 1s to 284.8 eV then Si 2p3/2 is at 102.21 eV (Si 2p = 102.4 eV) and O 1s is at 532.42 eV for silicone.

References:
[1] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers - The Scienta ESCA300 Database, Wiley Interscience, 1992, Appendices 3.1 and 3.2.

[2] J.D. Henderson, B.P. Payne, N.S. McIntyre, M.C. Biesinger, Surf. Interface Anal. 57 (2025) 214-220.

Calculating Oxygen Content from Adventitious Carbon 1s Spectra

Adventitious carbon (AdC) is commonly detected in X-ray photoelectron spectroscopy (XPS) analyses of most samples. While AdC can be beneficial in some cases, such as for charge correction purposes during the analysis of insulators, its associated C–O functionalities can complicate the interpretation of O 1s spectra. Accurately accounting for AdC’s contribution within the O 1s spectrum is essential but challenging due to significant spectral overlap and poorly resolved components in the high-resolution O 1s spectrum.

Rather than assigning multiple components without clear physical meaning—potentially leading to misinterpretations—incorporating stoichiometry offers a more reliable approach to improving data accuracy. However, applying stoichiometry can be tedious and challenging, particularly for novice users.

A recently published article [1] describes an approximation to enhance oxygen spectra interpretation by estimating oxygen linked to AdC. This publication provides background information, key assumptions, and an easy-to-use Excel calculator to assist XPS researchers in analyzing their own O 1s spectra.

This approach is particularly useful for accurately quantifying survey spectra when AdC influence must be minimized and for modelling high-binding-energy components in the oxygen 1s spectrum. The latter example is important to many transition metal oxides which have overlapping hydroxide and/or defect oxide components in the same binding energy window. Detailed examples of these applications are presented and discussed in reference [1]. These types of calculations were originally introduced in [2].

This Excel based calculator (also available at supplementary material in [1]) takes information from the survey and high resolution carbon 1s spectra and determines the amount of oxygen that is present from adventitious carbon species. This amount can then be deducted from the overall oxygen concentration.
(Note: you must download the file to Excel to use it - it is locked in Google Docs).

References:

[1] J.D. Henderson, B.P. Payne, N.S. McIntyre, M.C. Biesinger, Surf. Interface Anal. 57 (2025) 214-220.
[2] B.P. Payne, M.C. Biesinger, N.S. McIntyre, J. Electron Spectrosc. Relat. Phenom. 184 (2011) 29-37.

Caesium

 

Cs 3d5/2 binding energy values.

Note: Cs 3d5/2 - 3d3/2 splitting is 13.94 eV

Cs 3s: 1219 eV

Cs 3p3/2: 1002 eV

Cs 3p1/2: 1069 eV

Cs 4s: 234 eV

Cs 4p3/2: 161 eV

Cs 4p1/2: 173 eV

Cs 4d5/2: 77 eV

Cs 4d3/2: 80 eV

Cs 5s: 25 eV

Cs M5N45N45: 931 eV (Al)

Cs M4N45N45: 918 eV (Al)

Advanced Analysis of Indium

Analysing indium and indium-based compounds using X-ray Photoelectron Spectroscopy is challenging due to only slight shifts in the binding energies of the commonly used In 3d5/2 core line. A recent paper shares a comprehensive set of reference data for indium and its compounds, covering the In 3d, 3p, and 4d core lines, the In MNN Auger signal, as well as relevant counter ion signals [1]. Valuable tools, such as the modified Auger parameter and chemical state (Wagner) plots, which aid in differentiating indium species are also discussed.

Figure 1 and Table 1 present average literature values for the In 3d5/2 core line, highlighting both the average and standard deviations. These values illustrate the apparent challenge in distinguishing between various indium compounds. Factors like natural line widths, line shapes, and potential errors in charge correction add further complexity to accurate speciation.

Figure 1. Average In 3d5/2 literature values for indium compounds. 

Table 1. Average In 3d5/2 literature values for indium compounds.

Experimental data (Figure 2 and Table 2) presents a similar trend, underscoring that the In 3d5/2 core line alone is not enough to reliably distinguish between indium species. Notably, the In 3d5/2 core line of indium oxides shows variable asymmetry in line shape, which has led to differing interpretations in the literature. Some researchers attribute the high binding energy component to hydroxide or oxy-hydroxide species, while others suggest that it reflects electronic properties (screening effects). The current experiments [1] support the view that screening effects play an important role in this asymmetry. Excellent studies on these screening effects have been conducted by Körber [2] and Harvey [3].

Figure 2 In 3d spectra from [1].

Table 2. Experimental In 3d5/2 values from [1].

The In M4,5N4,5N4,5 transitions show a broader range of binding energy than the In 3d5/2 core level, making it better suited for accurate speciation, particularly by making use of the modified Auger parameter (Figure 3 and Table 3). In mixed-system analysis, i.e., a system containing multiple indium species, both the position and shape of the In M4,5N4,5N4,5 Auger electron signal can useful for speciation. Béchu and Fairley have provided an excellent discussion on the application of nonlinear and linear least-squares fitting methods to the In M4,5N4,5N4,5 signal, specifically for the oxidation of InSb [4]. Table 4 presents the fitting parameters needed to reproduce the M4,5N4,5N4,5 line shapes in order to fit complex experimental envelopes.

Figure 3. In MNN spectra for various indium compounds [1]. For reference, vertical lines indicating the kinetic energy (MNN) for metallic indium have been overlaid in each tile. Note that the additional signal present for InPO4 at 414.7 eV was due to Na contamination.

Table 3. In M4N4,5N4,5 and modified Auger parameter values [1]. 

Table 4. In MNN Auger peak fitting parameters [1].

Considering the information presented above, a comprehensive interpretation of XPS data involving indium and its compounds should involve a combination of the available data, including survey spectra (i.e., stoichiometry), the In 3d5/2 and In M4N4,5N4,5 Auger spectra, as well as the relevant counterion spectra (see [1]). For systems containing multiple indium compounds, the position and shape of the M4,5N4,5N4,5 transition can offer a more accurate approach than using the 3d5/2 core line alone.

References:

[1] J.D. Henderson, L.P. Pearson, H-Y. Nie, M.C. Biesinger, Surf. Interface Anal. 57 (2024) 81. https://doi.org/10.1002/sia.7356 

[2] C. Körber (et al.), Phys. Rev. B, 81 (2010) 165207

[3] S.P. Harvey (et al.), J. Phys. D Appl. Phys., 39 (2006) 3959.

[4] S. Béchu and N. Fairley, J. Vac. Sci. Technol. A, 42 (2024) 013202.

Iron

For the analysis of photoelectron spectra of relatively pure iron oxides, one can use peak shape and peak binding energy comparisons to standard compounds to derive oxide composition. McIntyre and Zetaruk’s [1] paper is widely cited and is still an excellent starting point for qualitative iron oxide determination. Pratt et al. [2] used a series of multiplet peaks to curve fit oxidized iron sulfide (pyrrhotite) surfaces.

Grosvenor et al.[3] fitted the various iron oxide, hydroxide and halide peak shapes with a close approximation of the Gupta and Sen[4] multiplet structure. Multiplet FWHM, splittings and weightings are presented. An analysis of satellite to main peak separation is also given. All Fe(II) (high spin only as low spin Fe(II) does not exhibit multiplet splitting) and Fe(III) species can be fitted with Gupta and Sen multiplet structure. Variation in peak spacing and intensity occur for different ligands. Broad satellite peaks of varying intensities at binding energies above the main Fe 2p3/2 structure are present in the spectra for all high spin compounds. However paper [3] only presents the main multiplet lines, excluding the details needed to fit the broader higher binding energy satellite structures.

Table 1 [5] presents full fitting parameters including the multiplet and satellite structure. FWHM values are reported for 10 eV pass energy only. To accommodate lower resolution settings slightly broader peaks would be necessary for best fit values. For these fits a Shirley background encompassing only the 2p3/2 portion of the spectrum is used. Also included in this Table are new spectral fitting parameters for FeCr2O4 and NiFe2O4, species that are important for the examination of oxide films on Fe-Cr-Ni alloys, as well as data for new analyses of α-Fe2O3 and γ-Fe2O3[5]. Fitting parameters for FeCO3, which has been noted in certain corrosion products, are also presented in Table 1. These analyses were collected from a mineral sample of siderite (cleaved in vacuum). Carbon 1s binding energy for FeCO3 is at 290.1 eV. The many spectra are best viewed in the original papers - see links in reference section.

New! Modified/updated fitting parameters for FeO, FeOOH and Fe3O4 are now included in Table 1 from work presented in reference [6]. In particular the analysis and fitting of FeO is improved substantially as the new FeO standard, sputter cleaned with a GCIS (not available during the original work in [3,5]), is free of low levels of surface Fe2O3/Fe3O4 contamination.

Table 1. Fe 2p3/2 spectral fitting parameters: binding energy (eV), percentage of total area, FWHM value (eV) and spectral component separation (eV) [5,6].

While these values [5] and reference spectra [1,3,5] will be useful for identification of pure oxide or oxy-hydroxide species, curve fitting of mixed systems quickly becomes complicated due to spectral overlaps. For example, it can be seen that various Fe(III) compounds have a similar range of Fe 2p binding energies and vary mostly in peak shape and satellite intensities. Any attempt at fitting two or more Fe(III) species to a spectrum will consequently contain an inherent degree of error. As well, overlap of the Fe(III) satellite structure with the Fe(0) and Fe(II) Fe 2p1/2 portion of the spectrum will result in setting the higher binding energy background endpoint placement at a point that will not cover the satellite structure of the Fe(III) species. This will require any fitting of mixed chemical state systems containing Fe(III) species to omit the higher binding energy Fe(III) satellite (e.g. Fe2O3, FeOOH) from the envelope of peaks. This will again increase the error associated with the curve fitting.  Finally, determination of the Fe species present, especially in a mix of Fe(III) species, should include corroborating evidence from O 1s analysis and even other analytical techniques such as Raman spectroscopy or, for thin crystalline films, grazing angle XRD. Some examples of fittings in mixed species samples are presented in [5].

Compared to the other transition metal species, the complex multiple species fitting of Fe is the most problematic. With so many possible species having overlapping binding energies erroneous interpretation can result. A sample with two distinct species can likely be fitted accurately, three species much less so, while four or more species must be looked at as indicative but unreliable. 

New! Recent work [6,7] has demonstrated the utility of these methodologies, with extremely good chemical state speciation achieved for oxide/hydroxide mixtures.  For metal/oxide/hydroxide mixtures, good success was found for low levels of metal content.  As the amount of metal grows (particularly above 25%) the amount of Fe(II) species tends to be underestimated. Results from the original curve-fittings from [5] were improved, particularly for FeO.  This work again emphasizes that the better the pure compound spectrum is, the better the final curve-fitting results!

Further work from Hughes et al. [8], published in the journal Corrosion Science in 2025, highlights the widespread erroneous analysis of the Fe 2p peak in XPS studies in corrosion science. It shows again the need for best practices when it come to fitting Fe 2p XPS spectra.

References:
[1] N.S. McIntyre, D.G. Zetaruk, Anal. Chem. 49 (1977) 1521.

[2] A.R. Pratt, I.J. Muir, H.W. Nesbitt, Geochim. Cosmochim. Acta 58 (1994) 827.
[3] A.P. Grosvenor, B.A. Kobe, M.C. Biesinger, N.S. McIntyre, Surf. Interface Anal. 36 (2004) 1564.
[4] R.P. Gupta, S.K. Sen, Phys Rev. B 12 (1975) 15.
[5] M.C. Biesinger, B.P. Payne, A.P. Grosvenor, L.W.M. Lau, A.R. Gerson, R.St.C. Smart, Appl. Surf. Sci. 257 (2011) 2717.

[6] A.E. Hughes, C.D. Easton, T.R. Gengenbach, M.C. Biesinger, M. Laleh, (2024) JVST A, 42 (2024) 053205.

[7] A.E. Hughes, C.D. Easton, T.R. Gengenbach, M.C. Biesinger, M. Laleh, (2024) JVST A, 42 (2024) 053206.

[8] A.E. Hughes, C.D. Easton, P.R. Anusuyadevi, T.J. Raeber, N.C. Wilson, A. Mol, Corr. Sci. 257 (2025) 113357.

Workshop Exercises: Advanced Chemical State Analysis

1) Carbon 1s - Set up a standard file for the fitting of adventitious carbon. Download the C 1s spectrum here.

2) Use it to charge correct a series of spectra. Download test file.

3) Oxygen 1s - General fitting of the O 1s spectra for metals.  Use file from number 2. Fit the O 1s spectrum, understand and explain all sources of oxygen.

4) Ti 2p - Set up curve-fitting parameters.  Download Ti 2p test file. See titanium literature fitting values.

5) Cr 2p - Set up curve-fitting parameters. Download Cr 2p test file. See chromium literature fitting values.

Notes:
•Cr(VI) Species – mix of oxide and hydrated species
   -One narrow peak FWHM 1.5 eV
  -Range from 579.0 to 580.0 eV
•Cr(III) Species – mix of oxide and hydrated species
  -Cr(OH)3 - One broad peak FWHM of ~2.5 eV, set to 577.5 eV
  -Cr2O3 - Five multiplet peaks of equal FWHM (~0.9 eV) with set areas and separations based on standard sample
•Cr(0) – Metal
  -One asymmetric peak with a FWHM of 0.9 eV
  -Range from 573.9 to 574.5 eV

6) Ni 2p - For the brave, download and give it a try. See Ni 2p literature fitting values and general instructions here.

What is Adventitious Carbon?

A thin layer of carbonaceous material is usually found on the surface of most air exposed samples, this layer is generally known as adventitious carbon. Even small exposures to atmosphere can produce these films. Adventitious carbon is generally comprised of a variety of (relatively short chain [1]) hydrocarbons species with small amounts of both singly and doubly bound oxygen functionality. The source of this carbon has been debated over the years. It does not appear to be graphitic in nature and in most modern high vacuum systems vacuum oils are not readily present (as they have been in the past) [1,2,3,4]. There may be some evidence that CO or CO2 species may play a role in the gradual appearance of carbon on pristine surfaces within the vacuum of the XPS chamber [3].

It’s presence on insulating surfaces provides for a convenient charge reference by setting the main line of the C 1s spectrum to 284.8 eV (although values ranging from 285.0 eV to 284.5 eV have been used in some cases, remember to check for this value when looking for binding energy references in the literature). The error in this value (284.8 eV) is, for most systems, on the order of +/-0.2 eV to 0.3 eV.  An in-depth look at the effectiveness of using AdC for charge correction purposes, including standardized fitting procedures, is presented in [5].

  

Work by Grey et al. [6] has explored the nature of adventitious carbon by XPS and time-of-flight secondary ion mass spectrometry (ToF-SIMS).  XPS D-parameter and ToF-SIMS analyses confirms that AdC is not graphitic in nature. An average C 1s spectrum for AdC (Figure 1, Table 1) was derived and shows that, on average, ~ 25 % of the carbon species in AdC is directly associated with oxygen functionality.  Similarly, ToF-SIMS analyses show that AdC is comprised of mainly short chain hydrocarbons with some oxygen functionality.

An advanced method for curve-fitting of the C 1s envelope for AdC (Table 2) was developed that included the effects of beta carbons (in this context, the alpha carbon is the carbon directly attached to the oxygen, and the beta carbon is attached to the alpha carbon) and were informed by the configurations of possible volatile organic compounds (VOC) that are the source of most AdC [6]. Using this method in combination with the dataset from [5], the average C–C/C–H AdC aliphatic peak position was shown to be 284.81 eV (+/- 0.25 eV) via verification with a secondary internal reference.

Figure 1. Average of 80 adventitious carbon C 1s XPS spectra.

Table 1. Average adventitious carbon C 1s fitting parameters from an average of 80 AdC spectra.

Table 2. Curve-fitting parameters for AdC C 1s including shifted beta peaks (*) (peaks E, F and G). Areas for peaks A, B, C, and D should be left unconstrained. # If peak-shape for peak D is well-defined the FWHM constraint can be removed.

References:

[1] T.L. Barr, S. Seal, J. Vac. Sci. Technol. A 13(3) (1995) 1239.

[2] P. Swift, Surf. Interface Anal. 4 (1982) 47.
[3] D.J. Miller, M.C. Biesinger, N.S. McIntyre, Surf. Interface Anal. 33 (2002) 299.
[4] H. Piao, N.S. McIntyre, Surf. Interface Anal. 33 (2002) 591.

[5] M.C. Biesinger, Appl. Surf. Sci, 597 (2022) 153681.

[6] L.H. Grey, H.-Y. Nie, M.C. Biesinger, Appl. Surf. Sci. 653 (2024) 159319.

Using Adventitious Carbon for Charge Correcting

The C 1s spectrum for adventitious carbon can be fit as follows.  A single peak, ascribed to alkyl type carbon (C-C, C-H), is fit to the main peak of the C 1s spectrum.  A second peak is added that is constrained to be 1.5 eV above the main peak, of equal FWHM to the main peak (C-C, C-H). This higher binding energy peak is ascribed to alcohol and/or ester functionality (C-OH, C-O-C). Further high binding energy components can be added if required. For example: C=O at approximately 3 eV above the main peak and O-C=O at 3.8 to 4.3 eV above the main peak. One or both of these peaks may also have to be constrained to the FWHM of the main peak if they are poorly resolved.  Reference [1] and the table below outline standard starting fitting parameters for adventitious carbon. 

Adventitious carbon C 1s curve-fitting parameters [1].

Spectra from insulating samples can then be charge corrected by shifting all peaks to the adventitious C 1s spectral component (C-C, C-H) binding energy set to 284.8 eV. There is certainly error associated with this assignment. Swift [2] lists a number of studies showing errors ranging from ±0.1eV to ±0.4 eV.  “Newer” studies (late 1970's) range from ±0.1 to ±0.3 eV. “Older” studies (late 1960's to early 1970's) were in the ±0.4eV range - however, reproducibility and resolution of the spectrometers of the time may have played a role.  Barr's [3] work from 1995 states that error in using adventitious carbon is ±0.2 eV.  Our work [4] in 2002 also suggests error in the ±0.2eV to  ±0.3eV range.  Experience with numerous conducting samples (1995 to present) and a routinely calibrated instrument have shown that the C 1s signal generally ranges from 284.7 eV to as high as 285.2 eV [5].  Reference [1] presents a detailed assessment of the analysis of insulating samples from a multi-user facility from over a 5-year period that showed an adventitious C 1s (C-C, C-H) binding of 284.91 eV ±0.25eV.  A similar study confirming the utility of the adventitious carbon technique with a similar multi-user facility analysis has been published by Morgan [6].

For organic systems, especially polymers, it is convenient to charge correct to the C-C, C-H signal set to 285.0 eV. This makes for easier comparison to the polymer handbook [7] which uses this number for charge correction.

References:
[1] M.C. Biesinger, Appl. Surf. Sci, 597 (2022) 153681.

[2] T.L. Barr, S. Seal, J. Vac. Sci. Technol. A 13(3) (1995) 1239.
[3] P. Swift, Surf. Interface Anal. 4 (1982) 47.
[4] D.J. Miller, M.C. Biesinger, N.S. McIntyre, Surf. Interface Anal. 33 (2002) 299.
[5] M.C. Biesinger, unpublished results

[6] D.J. Morgan, Surf. Interface Anal. 57 (2025) 28.
[7] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers - The Scienta ESCA300 Database Wiley Interscience, 1992.