ISO 22309:2006

INTERNATIONAL ISO
STANDARD 22309
First edition
2006-04-15
Microbeam analysis — Quantitative
analysis using energy-dispersive
spectrometry (EDS)
Analyse par microfaisceaux — Analyse élémentaire quantitative par
spectrométrie à sélection d'énergie (EDS)

Reference number
©
ISO 2006
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.

©  ISO 2006
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 2006 – All rights reserved

Contents Page
Foreword. iv
Introduction . v
1 Scope . 1
2 Normative references . 2
3 Terms and definitions. 2
4 Specimen preparation . 6
5 Preliminary precautions. 7
6 Analysis procedure. 7
7 Data reduction. 9
7.1 General. 9
7.2 Identification of peaks. 9
7.3 Estimation of peak intensity . 9
7.4 Calculation of k-ratios . 10
7.5 Matrix effects. 10
7.6 Use of reference materials. 10
7.7 Standardless analysis . 11
7.8 Uncertainty of results. 11
7.9 Reporting of results. 12
Annex A (informative) The assignment of spectral peaks to their elements . 14
Annex B (informative) Peak identity/interferences . 16
Annex C (informative) Factors affecting the uncertainty of a result. 18
Annex D (informative) Analysis of elements with atomic number < 11 . 20
Annex E (informative) Example data from a reproducibility study within a laboratory and between
laboratories . 21
Bibliography . 23

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 22309 was prepared by Technical Committee ISO/TC 202, Microbeam analysis.
iv © ISO 2006 – All rights reserved

Introduction
X-rays generated when a high-energy electron beam interacts with a specimen have energies (wavelengths)
which are characteristic of the chemical elements (atom types) present in the specimen. The intensity of these
X-rays from each element is related to the concentration of that element in the specimen. If these intensities
are measured, compared with those from a suitable reference material or set of reference materials, and
corrected in an appropriate manner, the concentration of each element can be determined. “Standardless”
procedures also provide quantitative information, but involve a comparison with previously measured
reference intensities that are stored within the software package or are calculated theoretically; such
procedures may, depending on any assumptions made, be inherently less accurate than the method
employing reference materials (see References [1] to [8] in the Bibliography). There are two common methods
of detecting the characteristic X-rays that are produced, one which relies on wavelength dispersive
spectrometry (WDS) and the other which uses energy-dispersive spectrometry (EDS). This International
Standard relates to the latter, energy-dispersive spectrometry.
Using EDS, the quantitative analysis of light elements (i.e. atomic number Z < 11, below Na) is more complex
and some of the problems are discussed in this International Standard.

INTERNATIONAL STANDARD ISO 22309:2006(E)

Microbeam analysis — Quantitative analysis using energy-
dispersive spectrometry (EDS)
1 Scope
This International Standard gives guidance on the quantitative analysis at specific points or areas of a
specimen using energy-dispersive spectrometry (EDS) fitted to a scanning electron microscope (SEM) or
electron probe microanalyser (EPMA); any expression of amount, i.e. in terms of percent (mass fraction), as
large/small or major/minor amounts is deemed to be quantitative. The correct identification of all elements
present in the specimen is a necessary part of quantitative analysis and is therefore considered in this
International Standard. This International Standard provides guidance on the various approaches and is
applicable to routine quantitative analysis of mass fractions down to 1 %, utilising either reference materials or
“standardless” procedures. It can be used with confidence for elements with atomic number Z > 10.
Guidance on the analysis of light elements with Z < 11 is also given.
NOTE With care, mass fractions as low as 0,1 % are measurable when there is no peak overlap and the relevant
characteristic line is strongly excited. This International Standard applies principally to quantitative analyses on a flat
polished specimen surface. The basic procedures are also applicable to the analysis of specimens that do not have a
polished surface but additional uncertainty components will be introduced.
There is no accepted method for accurate quantitative EDS analysis of light elements. However, several EDS
methods do exist. These are the following.
a) Measuring peak areas and comparing intensities in the same way as for heavier elements. For the
reasons explained in Annex D, the uncertainty and inaccuracy associated with the results for light
elements will be greater than for the heavier elements.
b) Where the light element is known to be combined stoichiometrically with heavier elements (Z > 10) in the
specimen, its concentration can be determined by summing the appropriate proportions of concentrations
of the other elements. This is often used for the analysis of oxygen in silicate mineral specimens.
c) Calculation of concentration by difference where the light element percentage is 100 % minus the
percentage sum of the analysed elements. This method is only possible with good beam-current stability
and a separate measurement of at least one reference specimen and it requires very accurate analysis of
the other elements in the specimen.
Annex D summarises the problems of light element analysis, additional to those that exist for quantitative
analysis of the heavier elements. If both EDS and wavelength spectrometry (WDS) are available, then WDS
can be used to overcome the problems of peak overlap that occur with EDS at low energies. However, many
of the other issues are common to both techniques.
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 14594, Microbeam analysis — Electron probe microanalysis — Guidelines for the determination of
experimental parameters for wavelength dispersive spectroscopy
ISO 15632:2002, Microbeam analysis — Instrumental specification for energy dispersive X-ray spectrometers
with semiconductor detectors
ISO 16700:2004, Microbeam analysis — Scanning electron microscopy — Guidelines for calibrating image
magnification
ISO/IEC 17025:2005, General requirements for the competence of testing and calibration laboratories
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
absorption correction
matrix correction arising from the loss of X-ray intensity from an element due to photoelectric absorption by all
elements within the specimen while passing through it to the detector
3.2
accuracy
closeness of agreement between the “true” value and the measured value
3.3
accelerating voltage
potential difference applied between the filament and anode in order to accelerate the electrons emitted from
the source
NOTE Accelerating voltage is expressed in kilovolts.
3.4
atomic number correction
matrix correction which modifies intensity from each element in the specimen and standards to take account
of electron backscattering and stopping power, the magnitudes of which are influenced by all the elements in
the analysed volume
3.5
beam current
electron current contained within the beam
NOTE Beam current is expressed in nanoamperes.
3.6
beam stability
extent to which beam current varies during the course of an analysis
NOTE Beam stability is expressed in percent per hour.
3.7
bremsstrahlung
background continuum of X-rays generated by the deceleration of electrons within the specimen
2 © ISO 2006 – All rights reserved

3.8
certified reference material
CRM
reference material, one or more of whose property values are certified by a technically valid procedure,
accompanied by or traceable to a certificate or other documentation which is issued by a certifying body
3.9
characteristic X-ray
photon of electromagnetic radiation created by the relaxation of an excited atomic state caused by inner shell
ionisation following inelastic scattering of an energetic electron, or by absorption of an X-ray photon
3.10
dead time
the time that the system is unavailable to record a photon measurement because it is busy processing a
previous event
NOTE This is frequently expressed as a percentage of the total time (see also live time).
3.11
energy-dispersive spectrometry
EDS
form of X-ray spectrometry in which the energy of individual photons are measured and used to build up a
digital histogram representing the distribution of X-rays with energy
3.12
electron probe microanalysis
a technique of spatially resolved elemental analysis based on electron-excited X-ray spectrometry with a
focused electron probe and an interaction/excitation volume with micrometre to sub-micrometre dimensions
3.13
escape peaks
peaks that occur as a result of loss of incident photon energy by fluorescence of the material of the detector
NOTE 1 These occur at an energy equal to that of the incident characteristic peak minus the energy of the X-ray line(s)
emitted by the element(s) in the detector (1,734 keV for silicon).
NOTE 2 They cannot occur below the critical excitation potential of the detector material, e.g. Si K escape does not
occur for energies below 1,838 keV.
3.14
fluorescence
photoelectric absorption of any X-ray radiation (characteristic or bremsstrahlung) by an atom which results in
an excited atomic state which will de-excite with electron shell transitions and subsequent emission of an
Auger electron or the characteristic X-ray of the absorbing atom
3.15
fluorescence correction
matrix correction which modifies the intensity from each element in the specimen and standards to take
account of excess X-rays generated from element “A” due to the absorption of characteristic X-rays from
element “B” whose energy exceeds the critical (ionisation) energy of “A”
3.16
full width at half maximum
FWHM
measure of the width of an X-ray peak in which the background is first removed to reveal the complete peak
profile
NOTE FWHM is determined by measuring the width at half the maximum height.
3.17
incident beam energy
energy gained by the beam as a result of the potential applied between the filament and anode
3.18
k-ratio
net peak intensity (after background subtraction) for an element found in the specimen, divided by the
intensity, recorded or calculated, of the corresponding peak in the spectra of a reference material
3.19
live time (s)
time the pulse measurement circuitry is available for the detection of X-ray photons
See also dead time (3.10).
NOTE 1 Live time is expressed in second (s).
NOTE 2 Live time = real time for analysis − dead time. Real time is the time that would be measured with a
conventional clock. For an X-ray acquisition, the real time always exceeds the live time.
3.20
overvoltage ratio
ratio of the incident beam energy to the critical excitation energy for a particular shell and sub-shell (K, LI, LII,
etc.) from which the characteristic X-ray is emitted
3.21
peak intensity
total number of X-rays (counts) under the profile of a characteristic X-ray peak after background subtraction
NOTE This is sometimes referred to as the peak integral.
3.22
peak profile
detailed shape of a characteristic peak which depends on the relative intensities and energies of the individual
X-ray emissions that are unresolved by the energy-dispersive spectrometer
3.23
precision
closeness of agreement between the results obtained by applying the experimental procedure several times
under prescribed conditions
3.24
quantitative EDS
procedure leading to the assignment of numerical values or expressions to represent the concentrations of
elements measured within the analysis volume
3.25
reference material
RM
material or substance, one or more properties of which are sufficiently well established to be used for the
calibration of an apparatus, the assessment of a method, or for assigning values to materials
NOTE A reference material is said to be homogeneous with respect to a specified property if the property value, as
determined by tests on specimens of specified size, is found to lie within the specified uncertainty limits, the specimens
being taken either from a single or different supply unit.
4 © ISO 2006 – All rights reserved

3.26
repeatability
closeness of agreement between results of successive measurements of the same quantity carried out by the
same method, by the same observer, with the same measuring instruments, in the same laboratory, at quite
short intervals of time
3.27
reproducibility
closeness of agreement between the result of measurements of the same quantity, where the individual
measurements are made by different methods, with different measuring instruments, by different observers, in
different laboratories, after intervals of time that are quite long compared with the duration of a single
measurement, under different normal conditions of use of the instruments employed
3.28
resolution
〈energy〉 width of a peak measured by an energy-dispersive spectrometer and expressed as the peak width at
half the maximum peak intensity
NOTE This is usually expressed as the value for Mn Kα (5,894 keV), although peaks from other suitable elements
can be used.
3.29
resolution
〈spatial〉 spatial specificity of microanalysis
NOTE This is usually expressed in terms of a linear or volumetric measure of the region of the specimen that is
sampled by the measured characteristic radiation.
3.30
standardless analysis
procedure for quantitative X-ray microanalysis in which the reference peak intensity in the k-value expression,
(unknown/reference) is supplied from purely physical calculations or from stored data from a suite of reference
materials; adjustments being made to match analysis conditions and augment incomplete reference data
3.31
sum peaks
artefact peaks that occur as a result of pulse co-incidence effects that occur within the pulse pair resolution of
the pile-up inspection circuitry
NOTE These peaks appear at energies corresponding to the sum of those energies of the two photons that arrive
simultaneously at the detector.
3.32
traceability
ability to trace the history, application or location of an entity by means of recorded identifications
3.33
uncertainty
that part of the expression of the result of a measurement that states the range of values within which the
“true” value is estimated to lie for a stated probability
3.34
validation
confirmation by examination and provision of objective evidence that the particular requirements for a specific
intended use are fulfilled
3.35
X-ray absorption
attenuation of X-rays passing through matter, arising primarily from photoelectric absorption for X-ray energies
and ranges appropriate to EPMA/EDS and SEM/EDS
4 Specimen preparation
4.1 Material for analysis shall be stable under variable pressure conditions and the electron beam. As-
received specimens can be examined after simple cleaning, but surface inhomogeneity or topography will
adversely affect the quality of the quantitative analysis.
4.2 For reliable quantitative analysis, the specimen shall present a flat, smooth surface normal to the
electron beam. This requirement is usually met by the application of conventional metallographic or
petrographic techniques. The area for analysis should be homogeneous over a region, typically a few
micrometres in diameter, around the electron beam.
4.3 Solid specimens can be reduced to an appropriate size, making sure they undergo no transformation
during the process. Prior to examination in the as-received condition, all surface debris should be removed
using appropriate techniques, such as ultrasonic cleaning.
4.4 Specimens for sectioning should be embedded, where possible, in a conducting medium prior to
metallographic or petrographic polishing using standard procedures. The medium shall be chosen with care,
to avoid the possibility of the conducting component becoming smeared onto the specimen surface and
mistaken for a component of the specimen by effectively altering the composition of the analysed volume.
NOTE 1 Polishing may be carried through to 1/4 µm grade diamond, provided this can be done without introducing
relief effects. Complete removal of all scratches is not essential, provided areas for analysis are clean and relief-free.
Damage to the specimen during preparation should be avoided. Potential damage mechanisms include:
a) the effects of lubricant;
b) the removal of second phases (precipitates);
c) differential polishing of phases with different hardness, thus introducing relief to the surface;
d) strain introduced into the surface;
e) edge curvature.
With cross-sections, the specimen can be coated with a harder material to improve edge retention.
[9]
NOTE 2 See ASTM E3-01 for further guidance.
4.5 If optical examination is to be used for the location of areas for analysis, either prior to or while in the
instrument, etching of the specimen may be necessary. The depth of etching should be kept to a minimum,
being aware of the possibility of surface compositional changes or the development of undesirable
topographic effects. Polishing away the etching may be needed after locating and marking the regions for
analysis by reference to existing or added features, such as a scratch or hardness indent.
4.6 The specimen should have a good conductivity to avoid charge-up generated by electron beam
irradiation. The specimen shall be connected to the instrument ground either through a conducting mount, or
by a stripe of silver or carbon paint. Any exposed non-conducting mounting material may be covered with a
conducting medium to avoid disturbance of the electron beam during analysis.
Carbon to a thickness of about 20 nm can be used as a conductive coating, although a metallic (e.g.
aluminium) coating of lesser thickness may be used if carbon is unacceptable.
NOTE Coating with an element already present in the specimen changes the apparent composition of that element in
the specimen, the magnitude of the effect being dependent on accelerating voltage and the thickness of the coating.
4.7 The prepared specimens shall be positioned in the instrument stage in such a way that, for the majority
of work, the surface can be reliably held normal to the incident electron beam.
6 © ISO 2006 – All rights reserved

[18]
4.8 Reference materials should conform to ISO 14595 and shall be prepared in a similar way to the
unknown specimens, e.g. polished, carbon coated, and positioned in the same orientation relative to the
electron beam.
5 Preliminary precautions
5.1 With a suitable vacuum level and the electron beam on, preliminary checks to ensure beam stability
(better than 1 % per hour variation in counts from a standard specimen) and stable detector performance
should be made prior to commencing calibration and analysis procedures. Detector stability shall be
monitored by measuring the detector resolution and energy calibration (see ISO 15632).
5.2 The energy scale of the detector system shall be checked or re-calibrated at regular intervals (e.g.
daily) or whenever there is any doubt over the identity of spectral peaks. All calibration data and any departure
from calibration shall be recorded.
5.3 The energy scale shall be checked using the line positions for two peaks, one of low energy (for
example, Al Kα at 1,486 keV) and one of high energy (for example, Cu Kα at 8,040 keV). Suitable individual
reference materials for this purpose are readily available. Alternatively, Kα peaks from two elements, such as
Al and Cu, present in the same reference material can be collected into the same displayed spectrum.
NOTE 1 Where it is possible to monitor the zero of the energy scale, only a single spectral peak is needed for accurate
calibration.
NOTE 2 Use of the line positions for two peaks from the same element, e.g. Cu Lα and Cu Kα, along with a simple
fitting procedure may give reduced calibration accuracy because of the distortion of Cu Lα by Cu LI and Cu Lβ.
5.4 The full width at half maximum (FWHM) of a selected peak provides a measure of the detector
resolution. It shall be measured at regular intervals following the procedure in ISO 15632, and shall be less
than an acceptable limit (typically that corresponding to a deterioration of more than 10 % of the as-delivered
resolution) for given settings of the pulse processing electronics. If these limits are exceeded, the system
performance shall be investigated and revalidated if necessary. All such measurements shall be performed
under identical conditions and be recorded.
NOTE For accurate quantitative analysis using peaks with an energy close to 1 keV, it is not advisable to use a
detector for which the FWHM of Mn Kα exceeds 160 eV, and is preferably less than 135 eV.
5.5 The ratio of the intensities of the Cu Lα and Cu Kα peaks, or equivalent peaks from other elements, e.g.
nickel, provides a measure of the detector efficiency and shall be determined at regular intervals and recorded.
Where this decreases to an unacceptable level (typically not less than 2/3 of the as-delivered value), the
system shall be revalidated by the manufacturer, particularly where standardless analysis is being carried out.
The detector should conform to ISO 15632:2002, Annex B.
5.6 For standardless analysis, check the measured spectrum using the Duane-Hunt rule
NOTE Deviations of the high energy end of the bremsstrahlung background from the indicated high voltage means
either charging of the specimen or a wrong high voltage indicated by the SEM. No quantitative analysis can be done in
these cases.
6 Analysis procedure
6.1 The filament shall be saturated and sufficient time allowed for it to achieve adequate stability, for
example better than 1 % over the anticipated duration of the analysis run. An accelerating voltage, of typically
between 10 kV and 25 kV, shall be chosen to meet one or more of several criteria.
a) For efficient excitation and good peak intensity, an overvoltage of at least 1,8 is desirable. Thus, when
utilising high energy lines of the order of 8 keV to 10 keV, a minimum accelerating voltage of 20 kV is
recommended.
b) For analysis of low energy lines, e.g. 1 keV to 3 keV, it is desirable to minimise the magnitude of the
absorption correction and possible errors arising from it, by operation at low beam voltages, e.g. 10 kV
providing the excitation requirement a) is satisfied for the lines of interest.
NOTE At 10 kV, analysis of the elements from Z = 24 (Cr) to Z = 29 (Cu) will not be possible since the K lines are not
adequately excited.
c) For accurate analysis, it is essential that the analysed volume shall be wholly contained within the feature
under examination throughout the analysis period; such features may be surface layers, small inclusions,
areas near to an interface, etc. Methods for the estimation of the influence of accelerating voltage on
analysis area and analysis depth are given in ISO 14594, and may be used in the selection of an
appropriate accelerating voltage.
6.2 The beam current shall be set to a value that is sufficiently large to enable an adequate count rate for
the whole spectrum from the specimen, but not so large as to introduce electronic distortion or sum peaks into
the spectra from pure elemental materials.
NOTE 1 Typical conditions for a Si-Li detector are input count rates of up to 10 000 counts/s and dead times less than
35 %. Total spectrum counts of approximately 250 000 are usually appropriate, although this depends on the
concentration of the element of interest.
The count rate capabilities of the system should be checked by comparing a spectrum obtained at a low count
rate with a spectrum obtained at the highest count rate to be used, to look for peak shift and pile-up distortion.
A minimum of two checks on beam stability, using a Faraday cup on a known reference specimen, should be
made prior to and following the analysis.
NOTE 2 In the SEM, however, the beam current available may be limited by the requirement to operate with a small
spot size in the pursuit of good image resolution.
6.3 The specimen shall be correctly positioned under the electron beam:
a) Using the energy-dispersive spectrometer with an EPMA requires the specimen surface to be located in
the focal plane of the optical microscope, and at the instrument-defined angle to the beam (usually 90 °).
Positioning or accidental tilt errors may degrade the accuracy of the data reduction procedures that are
applied.
b) In the SEM, the specimen height will be determined by the working distance that has been selected or is
given by the manufacturers as appropriate for EDS. Specimen tilt shall be set to zero.
ISO 16700 provides details for setting the stage at 0° tilt.
6.4 Pulse-processing settings shall be chosen to give the optimum resolution possible, consistent with the
desired count rate and dead-time settings suggested in 6.2.
6.5 Positions for analysis on the specimen shall be selected by reference to either the optical or scanning
electron images. Homogeneity in the chosen regions shall be checked by carrying out preliminary line scans
of the X-ray intensity from one or more of the major elements in the specimen, or point analyses at several
random locations sufficiently far from phase boundaries.
Where an optical system is available for locating positions on the specimen, care should be taken that the
centres of the optical and electron optical systems are coincident.
6.6 When reference materials are used, they shall be positioned, in relation to the electron beam, in exactly
the same way as the specimen and coated with a similar thickness of conducting material. Spectra shall be
collected from the specimen and all reference materials using the same analytical conditions, with at least one
reference material being examined at the beginning and end of the run in order to confirm stability of the beam
and other system parameters, e.g. baseline shift, gain drift, resolution and beam current.
8 © ISO 2006 – All rights reserved

6.7 A collection time shall be chosen which is sufficient to give the required total count within peaks of
interest. This will depend on the precision that is required in the final results. However, where there is any
doubt about precision, the analysis shall be repeated so that repeatability of the result can be verified.
6.8 All the relevant measurement parameters (take-off angle, specimen tilt, accelerating voltage, etc.) shall
be recorded.
NOTE In EDS, since the spectrometer resolution is of the order of 100 eV, the complex series of characteristic X-ray
line emissions appear as a few broad peaks in the digitised energy spectrum. Electronic counting losses are usually
corrected automatically by slowing down the system clock and the spectrum is recorded for a preset “live time” in seconds,
where “live” seconds are slightly longer than real seconds. The area of a peak, in counts divided by the preset live time,
gives the intensity of X-rays striking the active area of the detector. On the assumption that the same incident electron
beam current and same recording live time are used, the peak area in counts is an appropriate measure to compare X-ray
intensities for quantitative analysis, and therefore the terms “peak intensity” and “peak area” are used in this International
Standard.
7 Data reduction
7.1 General
Information contained in the spectra from the specimen consists of peaks arising from the characteristic
X-rays of the elements present in the analysed volume of the specimen. Although the intensity of a peak is
related to the amount of the element present in the specimen, the relative intensity of the peaks from various
elements shall not be taken to represent the relative concentrations of the elements.
7.2 Identification of peaks
The identity of all peaks in the spectrum shall be determined, taking care to allow for the possibility of peak
overlap, see Annex B (see References [9] and [10] in the Bibliography).
Automatic peak identification software may be provided by the manufacturer of the equipment and some
software will deal with overlapping peaks. However, performance depends on the accuracy of modelling of
peak profile and, consequently, different combinations of software, electronics and detectors can produce
different results for the same specimen.
Alternatively, manual identification of peaks should be performed in a systematic manner beginning with the
most intense peaks, see Annex A. The operator shall confirm the identification of all elements by reference to
published peak intensity data, with particular reference to the energies and relative intensities of the lines
within each of the observed K, L or M series. Incorrect peak intensity ratios or irregularly shaped peaks may
indicate the presence of an interfering element and should be investigated whilst making allowance for the
loss of intensity which can occur for low energy lines near the lower limit of detector response, or when the
overvoltage is inadequate for the line in question. All the observed peaks in the spectrum shall be accounted
for, including the possible appearance of escape peaks or sum peaks that occur when high count rates are
used.
Suitable peaks, from which to analyse for the concentration of each element present, shall then be selected.
For example, for an accelerating voltage of 20 kV, the following lines are suitable for quantification: Z = 11-30
K lines, Z = 29-71 L lines, Z = 72-92 M lines.
7.3 Estimation of peak intensity
In order to calculate the net peak intensity of interest, it is necessary to subtract the underlying background
(normally done by the software). This can be removed by modelling and subtraction from the whole spectrum,
by digital filtering, or, by linear interpolation between points on either side of the peak. Background subtraction
is less likely to be accurate if estimated over a large energy range. The choice of method will be largely
dependent on the facilities available within the analysis software, and their accuracy in relation to the intended
overall accuracy of the analysis.
In the event of overlapping peaks, deconvolution routines may be available with the software, or manual
estimates, made using the relative peak intensities of the various peaks that appear in the spectrum of a pure
element, offer ways of correcting for this. Whatever approach is used, validation shall be undertaken using

CRMs/RMs as specimens (see Reference [12] in the Bibliography).
7.4 Calculation of k-ratios
The extracted peak intensities (after background subtraction) for the elements found in the specimen, divided
by the intensities of the corresponding peaks in the spectra of the pure elemental references, will give the
k-ratios. When pure elements are not available and compound reference materials are employed, the
observed reference peaks shall be corrected for matrix effects; such corrections are included in most software
packages.
As an alternative to the use of reference materials, the k-ratios may be obtained by comparing specimen peak
intensities to elemental intensities held in, or calculated by, one of the many “standardless” procedures that
are available (see 7.7).
7.5 Matrix effects
The set of k-ratios for the elements identified in the specimen are corrected for matrix effects by applying one
of many available correction routines. The corrections allow for atomic number (Z), X-ray absorption (A), and
fluorescence (F) effects leading to the frequently used generic title of ZAF. Various other procedures have
been evolved to optimise performance and may be preferred to the purely numerical ZAF routines. For
example, the “phi-rho-Z” model also allows the depth distribution of the generated X-rays to be displayed. This
is of particular value when assessing whether the analysis has been confined to the desired depth (or area)
within the specimen.
Attention shall be paid to the analytical totals after this correction, but before any normalisation is considered.
When using a procedure where a separate measurement is made on at least one reference material, un-
normalised analytical totals in the range 95 % to 105 % are considered acceptable. Values outside this range
shall be investigated to determine whether unidentified elements are present, including those of atomic
number 10 and below, or whether instrumental instabilities have occurred during the analysis.
Where the analysis total is < 100 % and a single element with Z < 11 is also known to be present, its
concentration can be inferred by difference, providing the effects of this element on the matrix corrections is
known. This is particularly true if the element is oxygen and it is combined in a stoichiometric manner.
Measurement of an element by difference can lead to large relative errors in the element's concentration,
particularly when concentrations are low.
7.6 Use of reference materials
For quantitative analysis by X-ray microanalysis, the reference materials should be certified wherever possible.
However, reference materials with a composition close to those of the specimens can be used in two ways:
a) a specimen from such a material can be included in every batch of analyses to verify that satisfactory
results are being obtained, and to provide information on the uncertainty associated with the analysis;
b) it may be acceptable to make a direct comparison between the peak intensities observed in the specimen
and in the reference material in order to obtain an estimate of the specimen composition. This may be the
optimum approach if suitable elemental reference materials are not available.
The operating conditions under which the specimen and reference material data are obtained shall be the
same.
NOTE Modern software can often compensate for any differences in operating conditions. Multi-element reference
materials are available which provide for a daily monitoring of calibration, i.e. checks on relative peak intensities and peak
positions.
10 © ISO 2006 – All rights reserved

7.7 Standardless analysis
The standardless analysis routines mentioned in 7.4 can be employed when elemental reference materials
are not available at the time the analysis is undertaken. These routines provide estimates of elemental
concentration which are considerably more accurate than those derived from uncorrected relative peaks. The
standardless routines provide correction for specified excitation conditions and geometries. The k-ratios in
these procedures are obtained by reference to elemental peak intensities derived by calculation, or extracted
from a library of elemental or compound spectra (or profiles) provided by the manufacturer and augmented, or
replaced, by spectra obtained by the user on previous occasions. A total relative uncertainty better than
± 10 % relative may be achieved. However, greater errors can be expected when the analytical conditions
used are different from those specified for the standardless procedures and where low concentrations exist.
In its simplest form, the standardless analysis provides estimates of the relative elemental concentrations and
forces the sum to be 100 %, yielding a plausible result even if elements are omitted from the analysis, the
wrong elements are specified, or there are large errors in the determination of peak intensities. If some
elements cannot be analysed because there is no suitable peak available, then this shall be noted on the
printout of results, because the relative concentrations calculated by standardless analysis will not include the
effect of these missing elements on the X-ray intensity corrections for other elements (see Reference [1] in the
Bibliography).
A more reliable procedure may include measurement on one or more reference materials to enable estimates
of absolute concentrations to be obtained. Unnormalised totals may then be used as a tool to diagnose the
possibility of the presence of undetected elements in the specimen.
Validation is an essential part in the derivation of the uncertainty associated with the standardless approach
and shall be performed on known materials with similar characteristics to those of the specimens to be
analysed; particular attention should be paid to matching the experimental conditions to those required by the
standardless procedure (see ISO/IEC 17025, 2005, Subclauses 5.4.5 and 5.9).
Validation of the method (software and procedure), shall be performed prior to any analyses being done. This
can be accomplished using certified reference materials.
7.8 Uncertainty of results
7.8.1 General
Subclauses 7.8.2 and 7.8.3 offer guidance for uncertainty estimation where in-house development and
validation is performed (see References [13] to [16] in the Bibliography).
7.8.2 Routine analyses on repeated or similar specimens
The analyst shall establish the in-house reproducibility and accuracy of measurement for typical specimens
analysed in the laboratory, and should also validate the method to ensure that it is fit for the purpose.
The reproducibility of a measurement shall be established from repeat testing of the same specimen under
nominally the same conditions at intervals of time which are long compared with the time of analysis. This can
incorporate different operators and different analysis points within the same phase. This component of
uncertainty will include a number of the factors listed in Annex C.
Participation in proficiency testing schemes, round robin specimen analyses will provide a useful measure of
th
...