Hindawi Publishing Corporation

Advances in Materials Science and Engineering

Volume 2015, Article ID 290593, 7 pages
http://dx.doi.org/10.1155/2015/290593

Research Article
Element Analysis Based on Energy-Dispersive
X-Ray Fluorescence

Min Yao, Dongyue Wang, and Min Zhao

College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, 29 Yudao Street, Nanjing 210016, China

Correspondence should be addressed to Min Yao; ym nuaa@163.com

Received 30 August 2015; Revised 13 November 2015; Accepted 16 November 2015

Academic Editor: Santiago Garcia-Granda

Copyright © 2015 Min Yao et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Energy-dispersive X-ray fluorescence (EDXRF) spectrometry is a nondestructive, rapid, multielement, highly accurate, and
environment friendly analysis compared with other elemental detection methods. Thus, EDXRF spectrometry is applicable for
production quality control, ecological environment monitoring, geological surveying, food inspection, and heritage analysis,
among others. A hardware platform for the EDXRF spectrometer is designed in this study based on the theoretical analysis of
energy-dispersive X-ray. The platform includes a power supply subsystem, an optical subsystem, a control subsystem, and a personal
computer. A fluorescence spectrum analytical method is then developed to obtain the category and content of elements in a sample.
This method includes qualitative and quantitative analyses. Finally, a series of experiments is performed. Results show that the
precision of the proposed measurement method is below 8%, whereas its repeatability is below 2%.

1. Introduction 2. EDXRF Principle

Energy-dispersive X-ray fluorescence (EDXRF) spectrom- Direct excitation is a process by which atoms in a specimen
etry exhibits several advantages over other methods that are excited by primary photons from external sources, such
measure elemental content. This technique can measure as an X-ray tube, radioactive source, and synchrotron beam,
numerous sample forms and can satisfy different measure- to produce primary fluorescence. An alternative process is
ment requirements [1]. Moreover, EDXRF spectrometry can indirect excitation, in which the observed fluorescence is
detect a wide range of elements, even several elements produced as a secondary process by photons or particles
(electrons) originating from direct excitation or other sec-
simultaneously. This technique is nondestructive, fast, highly
ondary processes within the specimen. X-ray is an elec-
accurate, and environment friendly. EDXRF spectrometry
tromagnetic radiation generated by high-energy particles
can be used on different types of sample, such as bulk, liquid,
bombarding atoms. This radiation has wave-particle duality.
powder, and gas. It can also detect particles in the air. X-ray fluorescence (XRF) spectrometry uses primary X-ray
The European Union enforced a rule called “Restriction photons or other microscopic particles to excite the atoms
of Hazardous Substances” on July 1, 2006, to eliminate several in the test material to produce secondary XRF for material
harmful substances in electrical and electronic products. composition analysis and chemical state research. Qualitative
EDXRF spectrometry can rapidly measure these harmful analysis of X-ray spectroscopy is based on Moseley’s law, and
elements in products. the energy equation is as follows:
This study introduces the theoretical analysis of energy- 1 1
dispersive X-ray, the internal structure of the EDXRF spec- 𝐸X = 𝑅ℎ𝐶 (𝑍 − 𝜎)2 ( 2
− 2), (1)
𝑛1 𝑛2
trometer, and the algorithm for the rapid analysis of multiple
elements. This EDXRF equipment can measure minor ele- where 𝐸X is the characteristic X-ray energy, 𝑅 is Rydberg
ments in materials with contents below 0.03%. constant (𝑅∞ = 1.09737 × 107 m−1 ), ℎ is Planck’s constant
2 Advances in Materials Science and Engineering

High-voltage CCD Detector

power supply
Image
reading PC
Detector Data
communication Control communication
Be window circuit

Other
operations
Sample
Filter + collimator table

X-ray tube

Power supply

Figure 1: Structural diagram of the EDXRF spectrometer.

(ℎ = 6.6262 × 10−34 J⋅s), and 𝐶 is the speed of photons. 𝑍 is a power supply, a light path subsystem, a control circuit, and
the atomic number, 𝜎 is Shielding constant, and 𝑛1 and 𝑛2 are a personal computer (PC). High-voltage power is supplied to
the energy series. For the spectrum 𝐾𝛼1 , shielding constant the X-ray tube to emit a primary X-ray, which irradiates the
𝜎 = 1, 𝑛1 = 1 (𝐾-shell), and 𝑛2 = 2 (𝐿-shell). Thus, (1) can be sample. The sample is then stimulated to emit XRF, which
rewritten as follows: is received by an XRF detector. The detector classifies the
3𝑅ℎ𝐶 (𝑍 − 1)2 received photons according to energy and counts the number
𝐸𝐾𝛼1 = . (2) of photons that correspond to different energy levels. The
4
detector then sends the results to the PC, which completes
This law reveals the relationship between the X-ray energy the qualitative and quantitative analyses.
and atomic number. This law is the theoretical basis for the
qualitative analysis of material composition using XRF. A
positive relationship exists between the count rate of the 3.2. Light Path Subsystem. The light path subsystem, which
characteristic X-ray and the content of an element of the includes the X-ray tube, filter, collimator, detector, and a
tested sample, as follows: charge-coupled device (CCD) camera, is shown in Figure 2.
𝐾𝐼0 The light path subsystem is responsible for emitting,
𝐼𝐾 = × 𝑊𝐾 , (3) receiving, and counting the XRF photons. Its operation
𝜇0 + 𝜇𝐾
is as follows. A high-voltage power supply provides high-
where 𝐼𝐾 and 𝐼0 are the 𝐾 layer characteristics of the X-ray of voltage energy to the X-ray tube, which is stimulated to emit
the measured elements and the count rates of the incident X- primary X-ray. The primary X-ray passes through the Be
ray, respectively. Moreover, 𝜇0 and 𝜇𝐾 are the absorption coef- window, filter, and collimator, finally irradiating the sample.
ficients of the tested substance to the incident X-ray and the The sample is stimulated to emit XRF that can be recognized
tested element to the layer 𝐾 characteristic X-ray, respectively. by the detector. The received XRF is transformed into a
𝐾 is the constant related to the specific measurement device low-voltage pulse by the preamplifier. The pulse amplitude
and should be determined by the calibration instrument that is strictly proportional to the energy of the received
𝐾𝐼0 /(𝜇0 + 𝜇𝐾 ). 𝑊𝐾 is the measure of the content elements. XRF is further amplified by the main amplifier. The analog-
Quantitative analysis of the measured elements using XRF is to-digital converter then transforms the amplified voltage
theoretically based on (3). into a digital signal. The digital signal is further shaped,
sorted, and transformed into a pulse counter with amplitude
3. XRF Spectrometer Hardware information. This information is stored in a multichannel
analyzer according to its amplitude and finally formatted
3.1. Structure of the EDXRF Spectrometer. The EDXRF spec- to an XRF spectral line. The detector transmits spectral
trometer is designed according to Moseley’s law. The sys- information to the PC through a USB hub in the control
tem is illustrated in Figure 1. The spectrometer consists of circuit for qualitative and quantitative analyses.
Advances in Materials Science and Engineering 3

Sample Table 1: X-123 key parameters.

Energy resolution (Mn 𝐾𝛼 (5.9 eV) at half 145 eV

Sample table peak height)
Filter Collimator Energy 1.5–25 keV
MCA channel number 2048
Be window Detector
Detection area 5 mm2 –25 mm2
Counting rate 2 × 105 cps

CCD
X-ray tube

Figure 2: Structural diagram of the light path subsystem.

Figure 4: X-123 detector.

4. XRF Spectral Software

Qualitative analysis is performed to identify the elements
in the sample. By contrast, quantitative analysis is used not
only to identify the elements but also to determine elemental
content. According to Moseley’s law, the XRF photons of
a certain element possess a fixed energy; that is, the XRF
Figure 3: XY-501 X-ray tube.
photons of elements are unique. The abscissa of a channel
address reflects the photon energy in the spectral line. The
peak position of the abscissa corresponds to the energy of
the characteristic XRF of a certain element. Channel address
The X-ray tube in the light path subsystem (Model XY- range of a detector in this system is designed from 0 to 2048.
501; Dandong Oriental Electron Tube Factory, China) is Each XRF photon energy ranges from 103 eV to 104 eV. There-
shown in Figure 3. fore, the enlargement factor of the energy/channel address is
This X-ray tube can work at high voltage (50 kV) and low designed as 20 eV/channel. For example, the photon energy of
current (1 𝜇A). It can also work at low voltage (4 kV) and high 𝐾𝛼 for Ag is approximately 22.162 keV. The channel address
current (mA). This tube possesses good stability, keeping low should be near 1108 according to the energy calibration of
errors of 0.4% for 4 h. It has a small focal spot size, which is 20 eV/channel. The 𝐾𝛼 ray of Ag frequently appears in the
only 0.4 mm × 0.4 mm with a large X-ray flux. channel address of 1106.
The detector (Model X-123; Amptek Company, USA) is
the key part in the light path subsystem. This part mostly
decides the performance of the EDXRF spectrometer. The 4.1. Qualitative Analysis. Qualitative analysis is the basis of
detector, which is actually a sensor, transforms photons to quantitative analysis. Existing element types can be deter-
electrical pulse. The amplitude and number of electrical mined using the former analysis. Qualitative analysis is
pluses are related to the energy and intensity of X-ray. generally divided into three steps.
The high-performance detector is shown in Figure 4. The
key parameters are listed in Table 1. (1) Peak Location. The uncertainty of a large peak is approxi-
The entire measurement is observed using a CCD camera mately ±10 keV, whereas that of a small peak is up to ±50 eV.
and sent to the PC. The control circuit is responsible for Small peaks can be neglected when they overlap with a
reading or setting the parameters of the tube pressure, switch- large peak, particularly when its energy level is below 12 keV.
ing the filter and the collimator, controlling the interlock, Spectral overlap and interference frequently occur in the 𝐾
detecting motor position, receiving instructions from the PC, line (where the atomic number is between 22 and 35) and
and controlling USB communication. the 𝐿 line (where the atomic number is between 56 and 96).
4 Advances in Materials Science and Engineering

Table 2: Information of the samples.

Name Elemental concentration % Manufacturer

Cast iron (QD 2021A) C (3.57), S (0.053) East Institute of Standard Material, Jinan (2012)
High-Mn steel (QD 98-22) C (0.68), Si (0.254) East Institute of Standard Material, Jinan (2010)
Lead-based alloy (Pb60 Sn40) Pb (59.02), Sn (40.54) Shandong Research Institute of Metallurgical Science (1995)
Brass (GBW(E) 020012) Cu (63.34), Pb (0.028) Shanghai Institute of Electrical and Mechanical Engineering (2002)

42000 When the energy of an X-ray photon is higher than the

38000 Cu K𝛼 detection limit, some energy of the characteristic X-ray can
34000 escape because of its high transparency. The escaped energy
Photon number/200 s

30000 forms an escape peak in the low-energy position. The energy

26000 difference between the main and escape peaks is equal to that
22000 of the energy of the characteristic X-ray photons, which is
18000 recognized by the detector. The Si-PIN detector shows that
14000 the escape peak energy is 1.74 keV lower than the main peak,
10000 Zn K𝛼 Cu K𝛽 which is the Si 𝐾𝛼 energy. The height of the escape peak is
Sn K𝛼
6000 approximately 1/1000 to 2/100 of the main peak. However, the
Zn K𝛽 Pb L 𝛼 Sn K𝛽
2000 Pb L 𝛽 escape peak does not appear when the atomic number is over
0 200 400 600 800 1000 1200 1400 1600 1800 2000 30 [3].
Address channel X-ray scatters when it passes through objects, and a
scattering peak appears. Two types of scattering peaks are
Figure 5: Spectral line of Cu.
observed, namely, coherent scattering (or Rayleigh scatter-
ing) and incoherent scattering (or Compton scattering). The
scattering peak in the EDXRF spectrometer mainly results
All aforementioned factors hamper an accurate qualitative from the stimulation of the anode target of the X-ray tube.
analysis.
4.2. Quantitative Analysis. Quantitative analysis depends on
(2) Peak Recognition. The uncertainty of the peak position the standard curve established by standard samples. The
generally increases the difficulty of peak recognition. More measured intensity value of the unknown element in an
than one peak corresponds to the energy peak in the spec- actual measurement is fed into the standard curve to obtain
trum in several cases. Accumulated, escape, and scattering the elemental contents. In particular, the standard samples
peaks can also interfere with recognition. and unknown element should be measured under similar
conditions.
(3) Element Determination. Except for light elements, such Detailed steps are illustrated in Figure 6.
as Na, Mg, Al, Si, P, and S, identifying an element typically
requires more than two characteristic spectral peaks. When
the voltage of the X-ray tube is over 30 keV for atomic 5. Experiments
numbers 19 Z to 42 Z, the spectral peaks 𝐾𝛼 and 𝐾𝛽 appear
The EDXRF spectrometer is shown in Figure 7. The sample
simultaneously. Furthermore, relative intensities in different
table in Figure 7(c) holds the sample. Figure 7(d) shows the
spectral peaks should also be considered [2].
control circuit that controls the instrument and communi-
Figure 5 is the spectral line of Cu when the tube voltage is
cates with the PC.
40 keV, tube flow is 600 𝜇A, and the optical filter is 800 𝜇m.
The characteristic peak area of the relative element in Four types of samples are used in the tests. Sample
the qualitative analysis is calculated. This area corresponds characteristics are listed in Table 2.
to the photon number of the relative energy. The intensity The EDXRF spectrometer follows a set of operating steps,
of the characteristic peak can then be achieved. However, namely, preheating, initialization, mode selection, measure-
interference peaks should be considered in calculating the ment starting, qualitative analysis, and quantitative analysis.
peak area. The EDXRF spectrometer should be preheated for
The peak in the peak location, which is an interfer- approximately 40 min before the experiment. Preheating is
ence peak to the main peak, is called a pseudopeak. Such performed to increase tube pressure and tube flow gradually
peak includes accumulated, escape, and scattering peaks. to protect the light pipe and stabilize the instrument more.
The accumulated peak is also called the sum peak, which Initialization is necessary to calibrate channel magnifica-
is a phenomenon of peak hyperplasia resulting from the tion. A standard Ag sample is stimulated and magnification
accumulation of signal pulses while counting at a high rate. is adjusted during initialization. This step does not terminate
In a sum peak, the peak position does not correspond to until the main characteristic XRF stimulated from Ag is
the characteristic energy but to the sum of two independent located near the address of 1106. This step can ensure that
peaks [2]. system hardware remains in the best situation. Figure 8 shows
Advances in Materials Science and Engineering 5

Start

Calculating the relative concentration of the

sample
m measured concentration (Ii,u )
Ri =
elemental concentration (Ii )

The relative concentration is normalized to

the initial concentration
Ci,u = Rim /∑ Rim

Calculate and update the theoretical relative

concentrations Rit with Ci,u

Calculate the new concentration Ci is normalized to the

Rim Ci,u (1 − Rit ) initial concentration
Ci = m
Ri (Ci,u − Rit) + Rit (1 − Ci,u )

No
|Ci − Ci,u | < Ci,u /(121.2Ci,u + 28.8)

Yes

Ci is the elemental concentration

End

Figure 6: Steps of quantitative analysis, where 𝑚 is the measured value and 𝑡 is the calculated value.

(a) Instrument appearance (b) Light path subsystem

(c) Sample table (d) Control circuit

Figure 7: EDXRF spectrometer.

6 Advances in Materials Science and Engineering

1585 Main peak

Channel address = 1106
Photon number

1268
Content
Element Intensity
(PPM)
951 Mn 0.114403 12335

634

317

0 Results of the quantitative analysis

0
150
300
450
600
750
900
1050
1200
1350
1500
1650
1800
1950
Channel address Figure 10: Results of the quantitative analysis.

Figure 8: System initialization.

Table 3: Mn content of cast iron.

Mn
Number
Measured value Reference value Relative error
Element Mn Yes
1 1.32% 4.76%
Element Fe Yes 2 1.29% 2.38%
Element Ni Yes 3 1.20% −4.76%
4 1.24% −1.58%
5 1.23% −2.38%
6 1.27% 0.79%
7 1.28% 1.58%
Results of the qualitative analysis 8 1.22% 1.26% −3.17%
9 1.25% −0.79%
Figure 9: Results of the qualitative analysis. 10 1.20% −4.76%
11 1.21% −3.97%
12 1.22% −3.17%
13 1.32% 4.76%
the spectral line of Ag when the system has been completely 14 1.23% −2.38%
initialized. 15 1.22% −3.17%
Measurement is started after the suitable working mode
has been established. An experiment is performed to measure
the Mn content of cast iron sample. Figure 9 presents the
Table 4: Mn content of high-Mn steel.
results of the qualitative analysis, which show elements Mn,
Fe, and Ni in the cast iron. Meanwhile, Figure 10 presents Mn
Number
the results of the quantitative analysis, which show that Mn Measured value Reference value Relative error
content is 12,335 ppm or 1.23%. 1 12.10% −0.33%
The experiments are conducted 15 times to calculate the
2 12.12% −0.16%
repeatability of the results. Experimental results are presented
3 12.14% 0%
in Table 3.
Another experiment is conducted to measure the Mn 4 11.92% −1.81%
content of high-Mn steel. The results are presented in Table 4. 5 12.49% 2.88%
The relative error for the Mn content of cast iron is within 6 12.41% 2.22%
5%, whereas that for Mn content of high-Mn steel is within 7 12.21% 0.57%
3%. Repeatability of the results is calculated, and the results 8 12.13% 12.14% −0.08%
are presented in Table 5. 9 12.14% 0%
Table 3 shows that the repeatability of the measurement 10 11.91% −1.89%
is within 2%. 11 12.07% −0.57%
Two more experiments are conducted to measure minor 12 12.34% 1.65%
elements. The results are shown in Tables 6 and 7.
13 12.05% −0.74%
The results from the two experiments show that the
14 12.03% −0.91%
proposed EDXRF method can easily measure minor elements
in materials. 15 12.35% 1.73%
Advances in Materials Science and Engineering 7

Table 5: Repeatability of the measurement results. References

Element Content Repeatability [1] T. Wakisaka, N. Morita, M. Wakasa, S. Terada, K. Nishihagi,
1.26% 0.08% and K. Taniguchi, “Development of energy dispersive X-ray
Mn fluorescence spectrometer with monochromatic excitation for
12.14% 1.56%
the direct determination of trace elements in organic matrices,”
Bunseki Kagaku, vol. 45, no. 10, pp. 933–939, 1996.
Table 6: Cu content of lead-based alloy. [2] A. Gürol, “Measurements of the K X-ray intensity ratios
by using energy-dispersive X-ray fluorescence spectrometry,”
Cu
Number Applied Radiation and Isotopes, vol. 66, no. 3, pp. 372–376, 2008.
Measured value Reference value Relative error
[3] R. Sitko, B. Zawisza, and E. Malicka, “Energy-dispersive X-
1 0.0268% −0.74% ray fluorescence spectrometer for analysis of conventional and
2 0.0267% −1.11% micro-samples: preliminary assessment,” Spectrochimica Acta
3 0.0271% 0.37% Part B: Atomic Spectroscopy, vol. 64, no. 5, pp. 436–441, 2009.
4 0.0268% −0.74%
5 0.0271% 0.37%
0.0270%
6 0.0272% 0.74%
7 0.0271% 0.37%
8 0.0270% 0
9 0.0273% 1.11%
10 0.0271% 0.37%

Table 7: Cu content of brass.

Cu
Number
Measured value Reference value Relative error
1 0.0850% 2.18%
2 0.0756% −3.07%
3 0.0831% 6.54%
4 0.0841% 7.82%
5 0.0743% −4.74%
0.0780%
6 0.0729% −6.54%
7 0.0727% −6.79%
8 0.0833% 6.79%
9 0.0718% −7.95%
10 0.0719% −7.82%

6. Conclusion
XRF analytical method is discussed according to Moseley’s
law. An EDXRF spectrometer, which consists of a power
supply, a light path subsystem, and a control circuit, is
designed. A series of experiments is conducted using this
instrument. The results show that the precision of this
equipment for measuring elemental content is below 8%, and
its repeatability is below 2%.

Conflict of Interests
The authors declare that there is no conflict of interests
regarding the publication of this paper.

Acknowledgment
This research has been funded by Nanjing University of
Aeronautics and Astronautics Research Funding (Grant no.
NS2015030).
Journal of International Journal of International Journal of Smart Materials Journal of
Nanotechnology
Hindawi Publishing Corporation
Corrosion
Hindawi Publishing Corporation
Polymer Science
Hindawi Publishing Corporation
Research
Hindawi Publishing Corporation
Composites
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of
Metallurgy

BioMed
Research International
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Nanomaterials
Submit your manuscripts at
http://www.hindawi.com

Journal of Journal of
Materials
Hindawi Publishing Corporation
Nanoparticles
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Nanomaterials
Journal of

Advances in The Scientific International Journal of

Materials Science and Engineering
Hindawi Publishing Corporation
Scientifica
Hindawi Publishing Corporation Hindawi Publishing Corporation
World Journal
Hindawi Publishing Corporation
Biomaterials
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Journal of Journal of Journal of Journal of

Nanoscience
Hindawi Publishing Corporation
Coatings
Hindawi Publishing Corporation
Crystallography
Hindawi Publishing Corporation
Ceramics
Hindawi Publishing Corporation
Textiles
Hindawi Publishing Corporation
http://www.hindawi.com
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Volume 2014