Alloy Analysis with X-ray Spectrometry
Wavelength dispersive X-ray fluorescence instruments offer a solution for the
analysis of nickel-, cobalt- and iron-based alloys, super alloys and steels.
By Dr. Simon Milner



Alloy identification and analysis is an
essential aspect of numerous industrial
applications, ranging from scrap metal
recycling to the manufacture of special
steels for aircraft engines. X-ray fluores-
cence (XRF) spectrometry is one method
that can be used for alloy analysis. The
advantages of XRF have established the
technique as an industry standard.
The importance of alloy analysis can-
not be overestimated. Many alloys are
similar in appearance and, therefore,
can be easily confused after identifying
labels, such as mill test reports, have been
removed. Errors can have major cost
implications—product failure, machine
downtime and loss of consumer confi-
dence all being possible consequences.
The production of high-specification
steels requires process and quality control
with very tight tolerances. XRF spectrome-
try, capable of producing fast, accurate and
very precise chemical analysis, is one of the
principal methods for production control.
The challenges of analyzing a large
variety of alloying elements with wide
ranges of concentrations by XRF are well
known to the metals industry, in particu-
lar the matrix corrections needed and
the large number of reference materials
required for high-quality results.

WDXRF Analysis of Alloys
The data presented here demonstrate the
use of a wavelength dispersive X-ray fluo-
rescence (WDXRF) instrument, together
with a package for the analysis of alloying
elements with wide concentration ranges
(See Figures 1 and 2).
The package comprises specially pre-
pared calibration setup samples for nickel-,
iron- and cobalt-based alloys, and a dedi-
cated application setup using software that
employs a fundamental parameter (FP)
matrix correction algorithm that enables
the analysis of alloys, super alloys and
steels with a single measurement program.
The specially formulated setup samples
can be used to transfer calibrations for up
to 21 elements in nickel-, iron- and cobalt-
based alloys.
Preparation of Standards
and Samples
Calibration standards (CRMs), setup
samples and the CRMs used to verify accu-
racy were resurfaced using a fine-surface
milling machine with 150 grit Al
For the harder alloys, initial resurfacing
required 80 grit ZrO
Measurement Conditions
Total measurement time was 470 sec-
6 NDT Special Section | August 2007
Above: Figure 1. Accuracy overview from 0 to 100 weight percent comparison of certifed and measured values for
eight elements in the wide variety of CRMs listed in Table 4. Source: PANalytical B.V.
Below: Figure 2. Accuracy overview from 0 to 1.0 weight percent comparison of certifed and measured values for
eight elements in the wide variety of CRMs listed in Table 4. Source: PANalytical B.V. August 2007 | NDT Special Section 7
onds, with the spectrometer operated at
2.4 kilowatts. The program includes mea-
surements on both peak and background
positions for 21 elements and ensures
accuracy at low concentrations.
The elements from vanadium to
rhenium were measured using a LiF220
crystal to reduce spectral interferences.
For copper, tantalum and platinum, line
interferences were avoided by using Cu
K1, TaL1 and Pt L1 lines, respectively.
Master Calibration
The setup samples were calibrated against
master calibrations containing up to 126
CRMs. Master calibration accuracies are
shown in Table 1, where lower Root Mean
Square (RMS) and K-factor values indi-
cate more accurate calibrations (Figures
3-6 are calibration plots for Ni, Co, Si and
P). Typical compositions for the setup
samples are shown in Table 2.
The FP model automatically calculates
the matrix correction for the calibrations.
The wide range calibrations shown in
Figures 3-6 demonstrate the power of the
FP model. Corrections for inter-element
effects are optimized for each standard,
making it possible to calibrate concentra-
tion ranges from 0 to 70 weight percent
or more.
The accuracy of the WDXRF instrument
when used in conjunction with the setup
samples is illustrated in accuracy overview
plots for 11 elements, which compare certi-
fied and measured values for 12 CRMs of
widely varying compositions (See Table 3).
Precision and Stability
Precision was determined by repeated
measurement of a Nimonic alloy CRM
against the calibration. The CRM was
analyzed 20 times consecutively for a
short-term precision test and 15 times
over three weeks for long-term data (See
Table 4, Figure 7).
Comparison of the RMS value (1)
with the Counting Statistical Error (CSE)
shows both the precision of the method
and the stability of the spectrometer. The
Figures 3-6. Master calibration plots for Ni, Co, Si and P. Source: PANalytical B.V.


8 NDT Special Section | August 2007
similar magnitude of the RMS and CSE
values shows that errors originating from
the instrument are negligible.
Elemental analysis of alloys is essential
for identification, classification, quality
and process control. WDXRF spectrome-
ters are used routinely for these purposes.
The data shown here demonstrate
that a WDXRF instrument offers a solu-
tion for the analysis of Ni-, Co- and
Fe-based alloys, super alloys and steels in
a single, wide-range analytical program.
Analyses are accurate and precise, with
excellent system stability.
Dr. Simon Milner is senior product manager for
PANalytical B.V. (Almelo, The Netherlands). For more
information, call +31 (0) 546 534444, e-mail info@ or visit
Table 1: Calibration data for the master calibration.
Source: PANalytical B.V.
Figure 7: Short-term and long-term repeatability results for Al in a Nimonic alloy CRM (SS 310/1). The dashed lines
indicate the 3-σ levels. Source: PANalytical B.V.
Table 3: Analytical accuracy: comparison of certifed and measured values for 18 elements in 12 CRMs of various
types. Data for Y, Hf, Re and Pt are not included in this table because no reference values are available for these
CRMs. Source: PANalytical B.V.
BASICS August 2007 | NDT Special Section 9
Table 4: Analytical precision derived from repeated measurement of a Nimonic alloy (SS 310/1). Source: PANalytical B.V.
Table 2: Typical concentration values for the setup samples. Each set has a unique identifcation number (-xxx) because
the concentration values may vary from one set to another. Source: PANalytical B.V.
Reprinted with permission from Quality Magazine. Copyright August 2007.