ISO 20903:2011

INTERNATIONAL ISO
STANDARD 20903
Second edition
2011-11-01
Surface chemical analysis — Auger
electron spectroscopy and X‑ray
photoelectron spectroscopy — Methods
used to determine peak intensities and
information required when reporting
results
Analyse chimique des surfaces — Spectroscopie des électrons Auger
et spectroscopie de photoélectrons par rayons X — Méthodes utilisées
pour la détermination de l’intensité des pics et informations requises
pour l’expression des résultats
Reference number
ISO 20903:2011(E)
©
ISO 2011

---------------------- Page: 1 ----------------------
ISO 20903:2011(E)
COPYRIGHT PROTECTED DOCUMENT
© ISO 2011
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or ISO’s
member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2011 – All rights reserved

---------------------- Page: 2 ----------------------
ISO 20903:2011(E)
Contents Page
Foreword . iv
Introduction . v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Symbols and abbreviated terms . 1
5 Methods for peak‑intensity determination — Direct spectrum . 1
5.1 General . 1
5.2 Selection and subtraction of an inelastic background . 3
5.3 Measurement of peak intensity . 3
5.4 Measurement of a peak intensity with computer software . 4
5.5 Measurement of peak intensities for a spectrum with overlapping peaks . 5
5.6 Uncertainty in measurement of peak area . 5
6 Methods for peak intensity determination — Auger‑electron differential spectrum. 6
6.1 General . 6
6.2 Measurement of Auger‑electron differential intensity . 6
6.3 Uncertainties in measurement of Auger‑electron differential intensity . 7
7 Reporting of methods used to measure peak intensities . 8
7.1 General requirements . 8
7.2 Methods used to determine peak intensities in direct spectra . 8
7.3 Methods used to obtain and determine peak intensities in Auger‑electron differential spectra. 9
Annex A (informative) Instrumental effects on measured intensities .10
Annex B (informative) Useful integration limits for determination of peak intensities in XPS spectra . 11
Bibliography .13
© ISO 2011 – All rights reserved iii

---------------------- Page: 3 ----------------------
ISO 20903:2011(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the International
Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 20903 was prepared by Technical Committee ISO/TC 201, Surface chemical analysis, Subcommittee
SC 5, Auger electron spectroscopy.
This second edition cancels and replaces the first edition (ISO 20903:2006), which has been revised to include
an additional annex (Annex B) giving advice on the selection of the limits between which the peak intensity is
measured in X-ray photoelectron spectroscopy.
iv © ISO 2011 – All rights reserved

---------------------- Page: 4 ----------------------
ISO 20903:2011(E)
Introduction
An important feature of Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) is the
ability to obtain a quantitative analysis of the surface region of a solid sample. Such an analysis requires the
determination of the intensities of spectral components.
There are several methods of peak-intensity measurement that are applicable to AES and XPS. In practice, the
choice of method will depend upon the type of sample being analysed, the capabilities of the instrumentation
used, and the methods of data acquisition and treatment available.
This International Standard is expected to have two main areas of application. First, it provides a description of
methods that may be used in the determination of the intensity of a peak for an element in a given spectrum.
Information is given on the origin of uncertainties in the processes involved, and on how these uncertainties
may be reduced. Second, this International Standard specifies reporting requirements for the methods used for
peak-intensity measurements so that other analysts may use published results with confidence.
© ISO 2011 – All rights reserved v

---------------------- Page: 5 ----------------------
INTERNATIONAL STANDARD ISO 20903:2011(E)
Surface chemical analysis — Auger electron spectroscopy
and X‑ray photoelectron spectroscopy — Methods used to
determine peak intensities and information required when
reporting results
1 Scope
This International Standard specifies the necessary information required in a report of analytical results based
on measurements of the intensities of peaks in Auger electron and X-ray photoelectron spectra. Information on
methods for the measurement of peak intensities and on uncertainties of derived peak areas is also provided.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced document
(including any amendments) applies.
ISO 18115-1, Surface chemical analysis — Vocabulary — Part 1: General terms and terms used in spectroscopy
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 18115-1 apply.
4 Symbols and abbreviated terms
A peak area
AES Auger electron spectroscopy
b number of channels over which intensities are averaged to obtain a baseline
eV electron volts
n number of channels in a spectrum
XPS X-ray photoelectron spectroscopy
y number of counts in the ith channel of a spectrum
i
ΔE channel width (in electron volts)
Δt dwell time per channel (in seconds)
σ(A) standard deviation of calculated peak area
5 Methods for peak‑intensity determination — Direct spectrum
5.1 General
Figure 1 a) shows a portion of an X-ray photoelectron spectrum in which intensity is plotted as a function of
kinetic energy increasing to the right or of binding energy increasing to the left. The intensity is plotted usually
in units of counts or sometimes in units of counts per second. Intensities may also be plotted as a digitized
© ISO 2011 – All rights reserved 1

---------------------- Page: 6 ----------------------
ISO 20903:2011(E)
voltage; this procedure is often used when the intensity of an Auger differential spectrum is obtained from an
analogue detection system. Energies are commonly expressed in electron volts.
Key
X1 binding energy (eV)
X2 kinetic energy (eV)
Y intensity
Figure 1 — Illustration of procedure involved in the determination of the intensity of a single peak
in an X‑ray photoelectron spectrum (as described in 5.2 and 5.3)
The intensity of a single peak in an X-ray photoelectron spectrum can be measured by using the procedure
described in the following two subclauses (5.2 and 5.3) or by using computer software as described in 5.4. The
measurement of peak intensities for a spectrum containing overlapping peaks is described in 5.5. Information
on the uncertainty of a measured peak area for a single peak is given in 5.6.
The intensity of a single peak in a direct Auger-electron spectrum can be measured by following the procedure
[1][2]
described in 5.2 and 5.3, although it may be necessary first to subtract a secondary-electron background .
Alternatively, computer software can be used to measure the peak intensity as described in 5.4.
In some cases, the peak of interest may be superimposed on a sloping background. This background could
arise from multiple inelastic scattering of Auger electrons or photoelectrons of initially high energy, from
multiple inelastic scattering of primary electrons (in AES), or from photoemission by bremsstrahlung radiation
(for XPS with an unmonochromated X-ray source). It may be necessary (e.g. with use of the Tougaard inelastic
background described in 5.2) or desirable to subtract this background from the spectrum in the vicinity of the
peak before proceeding with the peak-intensity measurements described in 5.2 to 5.5. This subtraction can
usually be performed by fitting a straight line to the sloping background at energies between about 10 eV and
30 eV above the peak of interest, extrapolating this line to lower energies, and subtracting the spectral intensities
from this linear background. If a linear function is judged to be invalid for describing the sloping background
over the spectral range of interest (e.g. for modelling the background of scattered primary electrons in AES),
[3]
an exponential function can be utilized .
2 © ISO 2011 – All rights reserved

---------------------- Page: 7 ----------------------
ISO 20903:2011(E)
5.2 Selection and subtraction of an inelastic background
It is necessary to select an appropriate inelastic background and then to subtract this background from the
measured spectrum. Three types of inelastic background are in common use:
a) linear background;
[4]
b) integral or Shirley background ;
[5][6][7] [8][9]
c) Tougaard background and Werner background , based on physical models describing inelastic
electron scattering in solids.
Information on procedures and software for determining the Shirley, Tougaard and Werner backgrounds is
[4]-[13] [14]
given in the scientific literature and ISO/TR 18392 .
From a practical viewpoint, the selection of a particular background will depend on (a) whether the relevant
software is conveniently available and (b) the type of sample analysed. For insulators, the linear background
is often satisfactory, while the Shirley background is often employed for metals. While these two backgrounds
are simple and convenient to apply, the limits of these two backgrounds (the starting and ending points on the
energy scale) should be chosen carefully so that the background is as nearly continuous as possible with the
spectrum in the region of overlap.
[5][6][7]
Tougaard’s approach, in particular, for background determination and subtraction has found favour over the
[15] [16]
Shirley background because it describes the physics of the inelastic-scattering process more accurately .
The Tougaard and the Werner approaches have a further advantage in that they are insensitive to the precise
positions of the starting and ending energy points providing they are clearly in the spectral region well away
from the main peak of interest (typically starting at an energy at least 10 eV higher than that of the peak of
interest and ending at an energy at least 50 eV lower). This requirement is a disadvantage in that spectra have
to be recorded over a larger energy range than if the linear or Shirley background is used.
As an example, Figure 1 a) shows an XPS peak whose intensity is to be measured. Vertical lines have been
drawn to indicate suitable limits for use of the Shirley background. The spectrum after subtraction of this
background is shown on an expanded energy scale in Figure 1 b). For clarity of display, the zero of the intensity
scale in Figure 1 b) has been placed at 2 % of the ordinate axis. The end points in Figure 1 b) are at the same
positions as those in Figure 1 a).
Averaging over neighbouring channels may be helpful in defining the signal level at the selected end points,
thus improving the precision of peak-height or peak-area measurement. The sets of points to be averaged may
be located inside or outside of the chosen end points or may be symmetrically placed about the end points. It is
important that the end points are chosen to be sufficiently far from the peak so that the averaging process does
[17]
not include significant peak intensity. Harrison and Hazell have derived an expression for the estimated
uncertainty in a peak-area measurement (see 5.6) and have shown that a large contribution to this uncertainty
comes from uncertainties arising from the choice of end points and the intensities at these end points.
[18]
Smoothing of a spectrum, using a Savitzky-Golay convolution with a width less than 50 % of the full width
at half-maximum intensity of the peak, may improve the precision of a peak-height determination. However,
smoothing should be avoided for peak-area determination since it cannot improve the precision and, if over-
done, will distort the spectrum.
Annex B gives information on the choice of suitable energy limits for the determination of peak intensities or
areas in XPS spectra.
5.3 Measurement of peak intensity
5.3.1 Measurement of peak height
A peak height is determined (i) by direct measurement from a chart output using a ruler, (ii) by using computer
software to obtain the intensity difference from the baseline to the peak maximum or (iii) by using computer
[10][11][12]
software to fit an appropriate analytical peak shape (Gaussian, Lorentzian or a mixture of the two )
to the experimental spectrum (that is, the group of data points defining the peak of interest). The length of the
© ISO 2011 – All rights reserved 3

---------------------- Page: 8 ----------------------
ISO 20903:2011(E)
vertical line with arrows in Figure 1 b) is a measure of the peak height in units defined by the intensity scale
(either counts or counts per second).
The use of peak heights in subsequent data processing has advantages arising from the speed of processing
and the ease with which this method can be applied with many instruments. However, using peak height as a
measurement of intensity has several disadvantages: (i) it is insensitive to peak-shape changes arising from
the complex chemistry of an element, (ii) it ignores spectral intensity from secondary features in the spectrum
(such as satellite peaks) and (iii) the measured height is very dependent on the choice of inelastic background.
NOTE Instruments should be operated with settings chosen to avoid significant nonlinearities in the intensity
[19]
scales ; alternatively, corrections should be made for counting losses due to the finite dead time of the counting
[19]
electronics . Spectra should be corrected for the intensity-energy response function of the instrument before peak
[20]
heights are measured . Further information is provided in Annex A.
5.3.2 Measurement of peak area
Methods used in the past to determine the peak area have included: (i) tracing the peak shape onto paper,
cutting the paper into the shape of the peak and weighing the paper, (ii) plotting the peak onto paper printed
with a square grid pattern and counting the grid squares and (iii) using a planimeter, a mechanical device for
making area measurements. However, most modern AES and XPS instruments have computer software that
can be used to determine the peak area (e.g. by summing the counts above the inelastic background or by
numerical integration). Alternatively, the peak area can be calculated from the parameters obtained after fitting
[10][11][12]
the peak with an appropriate analytical function . The shaded area in Figure 1 b) illustrates the peak
area obtained from integration of the peak defined by the end points and subtraction of the inelastic background
in Figure 1 a).
The measured intensity in each channel of an AES or XPS spectrum depends on a number of instrumental
[20]
parameters and settings . For specified instrumental conditions, the measured intensity for each channel can
be simply expressed as a number of counts (or counts/second) per eV; Annex A provides further information. A
peak area (or peak intensity) is then expressed as the total number of counts (or counts/second) for a specified
energy region of summation or integration.
NOTE Instruments should be operated with settings chosen to avoid significant nonlinearities in the intensity
[19]
scales ; alternatively, corrections should be made for counting losses due to the finite dead time of the counting
[19]
electronics . Spectra should be corrected for the intensity-energy response function of the instrument before peak
[20]
areas are measured . Further information is provided in Annex A.
In practical AES and XPS, an analyst generally wishes to compare intensities of peaks that were measured with
identical instrumental settings [e.g. analyser mode, pass energy (for the constant-analyser-energy mode) and
retarding ratio (for the constant-retarding-ratio mode)] but differences in certain other settings (e.g. different
energy channel widths or different dwell times). The analyst often will not know certain parameters that affect
the absolute intensities of measured peaks (see Annex A) since only relative intensities are needed for practical
analyses. In such cases, peak intensities can be determined from simple summations or integrations of
measured spectra for the particular conditions, and these intensities are often expressed in units of counts⋅eV
or counts⋅eV/second. Corrections of peak areas can then be made as needed for different channel widths and
dwell times. Annex A provides further information.
The use of peak intensities derived from measurements of peak areas has some clear advantages over the
use of measurements of peak heights. First, account can be taken in the measurement of peak areas of any
chemical changes that result in reduced peak height and increased peak width (compared to the corresponding
values for the elemental solid). Second, any satellite intensity can be easily included in the measurement
of peak area. However, the uncertainty of a peak-area measurement may increase for complex specimen
materials with many elemental components that could have overlapping spectral features (as described in 5.5).
In such cases, the value of the derived peak area may depend on the choice and placement of the inelastic
background function in 5.2.
5.4 Measurement of a peak intensity with computer software
Computer software can be used to fit a selected analytical function describing the shape of a peak and another
[10][11][12]
function describing the inelastic background to a measured spectrum . This process essentially
4 © ISO 2011 – All rights reserved

---------------------- Page: 9 ----------------------
ISO 20903:2011(E)
combines the steps described in 5.2 and 5.3 into a single procedure. Prior removal of X-ray satellites from XPS
spectra recorded using unmonochromated radiation may be necessary if they contribute intensity in the region
of the spectrum defined by the integration limits (see 5.2).
Peak shapes in AES may be more complex than those in XPS, and analytical functions used to fit XPS spectra
may then be unsatisfactory for similar fits of AES spectra. In such cases, peak intensities can be derived using
spectral addition/subtraction, least-squares analysis with suitable reference spectra, or principal-component
[21]
analysis .
5.5 Measurement of peak intensities for a spectrum with overlapping peaks
In many practical cases, a spectrum in the region of interest may consist of two or more overlapping peaks
because of the presence of chemically shifted peaks from the same element, the presence of peaks from
multiple elements or the presence of peaks arising from X-ray satellites. As an example, Figure 2 shows an
X-ray photoelectron spectrum for an oxidized vanadium foil that was measured with an unmonochromated
Al Kα X-ray source. In this spectrum, the more intense peaks arise from vanadium 2p and oxygen 1s
photoelectrons; there is also a weaker peak due to oxygen 1s photoelectrons excited by the Al Kα satellite
3,4
line that overlaps the vanadium 2p peaks. Correct identification of chemical state requires calibration of the
[22]
instrumental binding-energy scale and, for non-conductive specimens, use of charge-control or charge-
[23]
correction procedures .
For a spectrum with overlapping peaks, it is necessary to measure intensities from fits of analytical functions
[10][11][12]
to a selected spectral region . Peak heights and peak areas can be determined from values of the
parameters found for each peak.
Key
X binding energy (eV)
Y intensity
1 O(1s) X-ray satellites
Figure 2 — X‑ray photoelectron spectrum measured with unmonochromated Al Kα X‑rays for an
oxidized vanadium foil
5.6 Uncertainty in measurement of peak area
The uncertainty in the result of a measurement (such as peak area) generally consists of several components
[24]
that may be grouped into two categories according to the method used to estimate their numerical values :
type A: those which are evaluated by statistical methods;
type B: those which are evaluated by other means.
© ISO 2011 – All rights reserved 5

---------------------- Page: 10 ----------------------
ISO 20903:2011(E)
The type A uncertainty indicates the component of uncertainty arising from a random effect while the type B
[24]
uncertainty indicates the component of uncertainty arising from a systematic effect . For measurements of
peak areas, type A uncertainties can arise from counting statistics in the measurement of a spectrum and the
fitting of an inelastic background to the spectrum (5.2). Type B uncertainties typically arise from the choice of
the inelastic background function (5.2), the selection of end points (5.2), the choice of a function to describe a
[25]
measured peak shape (5.3 and 5.4), the choice of computer software used for peak fitting and the choice of
initial parameter values in a non-linear least-squares fitting algorithm. The total uncertainty of a measurement
can be obtained from the standard deviation (for the type A uncertainties) and an evaluation of the type B
[24]
uncertainties .
The standard deviation, σ(A), of the measured peak area, A, of a single peak measured in counts⋅eV/second is
[17]
given by the following expression :
05,
n−1 2
 
 
()ny−+2 ()y
 ΔE 
1 n
 
σ()A =  y + (1)
  ∑ i
 
Δt 4
 
 
i=2 
 
where ΔE is the channel width (or energy step) in the spectrum (in eV), Δt is the dwell time per channel (in
seconds), n is the number of channels in the spectrum and y is the number of counts in the ith channel. If the
i
peak area is measured in counts eV, σ(A) from Equation (1) should be multiplied by Δt.
Equation (1) includes contributions from statistical noise in the experimental spectrum and from the uncertainty
arising from placement of the baseline [the final term in Equation (1)], the latter being particularly significant for
noisy spectra. The latter uncertainty can be reduced as explained in 5.2, in which case σ(A) becomes:
0,55
n−1 2
 
 
  ()ny−+2 ()y
ΔE
1 n
 
σ()A =  y + (2)
 
∑ i
 
Δt 4b
 
 
i=2 
 
where b is the number of channels adjacent to the end points for which intensities are averaged to obtain the
baseline. The final term in Equation (2) will usually be larger than the previous term, in which case σ(A) will be
0,5 [17]
reduced by approximately b over the value obtained from Equation (1) .
The determination of uncertainties in peak areas for a spectrum with overlapping peaks is more complex than
for the case of a single peak. The type B uncertainties may be larger than the type A uncertainties, particularly if
there is a high degree of peak overlap and if assumptions have to be made concerning the inelastic background
function and the function describing the peak shape. Information on type A and type B uncertainties can be
[17][26]-[30]
found in the scientific literature .
6 Methods for peak intensity determination — Auger‑electron differential spectrum
6.1 General
In Auger electron spectroscopy, spectra are often displayed in the differential mode in which the first derivative
of the direct spectrum is plotted as a function of kinetic energy. If the direct spectrum is recorded as intensity
in counts versus energy in electron volts, the intensity in the differential spectrum will be in units of counts per
electron volt.
6.2 Measurement of Auger‑electron differential intensity
Two methods, illustrated in Figure 3, are commonly used to measure peak intensity of an Auger-electron
differential spectrum:
a) peak-to-peak intensity (measured from the positive to the negative excursions in a differential spectrum);
b) peak-to-background intensity (measured from the negative excursion to the background on the high-
kinetic-energy side of the spectral feature).
Measurements of these intensities were generally made electronically in early commercial AES instruments
[18]
but numerical methods for differentiation are now also employed . The measurements of peak-to-peak or
6 © ISO 2011 – All rights reserved

---------------------- Page: 11 ----------------------
ISO 20903:2011(E)
peak-to-background intensities can be made rapidly, and are therefore useful in, for example, depth-profiling
studies.
Key
X kinetic energy (eV)
Y intensity
a
Peak-to-peak.
b
Peak-to-background.
Figure 3 — Schematic diagram of an Auger‑electron differential spectrum indicating measurements
of peak‑to‑peak and peak‑to‑background intensities
6.3 Uncertainties in measurement of Auger‑electron differential intensity
Auger-electron lineshapes in direct spectra are generally more complex than XPS lineshapes, and these
[31]
lineshapes may change appreciably with change of chemical state . As a result, peak-to-peak and peak-to-
background intensities in AES differential spectra can change significantly with chemical state. Compositions
derived from AES differential spectra may therefore have larger uncertainties than the compositions obtained
from direct spectra in XPS. Additional complications can arise from energy shifts of Auger peaks (e.g. in
the vicinity of an interface due to different chemical states or to different amounts of surface charge) and
with overlapping peaks from different elements. Figure 4 is an example of an AES differential spectrum with
overlapping features due to chromium and oxygen. In such cases, elemental intensities may be derived from
spectral addition/subtraction, least-squares analysis with elemental or compound spectra, factor analysis or
[21]
principal-component analysis . Seah et al. have reviewed problems involved in the analysis of AES differential
spectra and, for quantitative analysis using the differential spectrum, recommend numerical broadening to
[32]
reduce the effects of peak-shape changes .
© ISO 2011 – All rights reserved 7

---------------------- Page: 12 ----------------------
...