Eur. Phys. J.

Plus (2019) 134: 286

DOI 10.1140/epjp/i2019-12652-8
THE EUROPEAN
PHYSICAL JOURNAL PLUS
Regular Article

Characterization and setting protocol for a simultaneous X-ray

Diffraction – X-ray Fluorescence system (XRD/XRF) for in situ
analysis

Valentina Aguilar1,a , José Luis Ruvalcaba-Sil1,b , Lauro Bucio1 , and Eric M. Rivera-Muñoz2
1
Instituto de Fı́sica, Universidad Nacional Autónoma de México, Universidad Av. 3000, Mexico City, 04510, Mexico
2
Centro de Fı́sica Aplicada y Tecnologı́a Avanzada, Universidad Nacional Autónoma de México, Universidad Nacional Au-
tonoma de México, A.P. 1-1010, Queretaro, 76000, Mexico

Received: 23 January 2019 / Revised: 5 March 2019

Published online: 25 June 2019

c Società Italiana di Fisica / Springer-Verlag GmbH Germany, part of Springer Nature, 2019

Abstract. A simultaneous XRD/XRF device was set to perform non-invasive and in situ analyses. It has a
curved position-sensitive detector (CPS) for X-ray Diffraction (XRD), with a 180 mm radius and a 120◦ 2θ
range with a reflection mode asymmetric geometry, and with an additional SDD detector for simultaneous
X-ray Fluorescence analyses (XRF). This device allows the use of two different configurations: one for
powder samples and objects smaller than 200 mm, and another for larger flat objects (e.g. paintings). This
work presents the design, optical features and characterization of the system, as well as some examples of
successful XRD identification in solid samples using XRF data with analysis by diffraction in situ (ADIS).
The reliability of this system for crystalline phase identification was tested with a 2θ position, profile shape
and intensity evaluation, using different reference materials. In addition, this device was compared with
portable XRD-XRF systems and two laboratory XRD devices.

1 Introduction

X-ray Diffraction (XRD) is the most reliable technique for the characterization of crystalline materials, with powder
diffraction also being a powerful method for the analysis of materials that are present in polycrystalline form.
Although modular XRD systems are the most suited equipment to analyze crystalline materials, there are cases in
which it is necessary to develop portable devices and to apply non-invasive methods, in particular for those materials
that cannot be transported due to their dimensions, their importance or fragility, such as cultural heritage objects
like paintings or other objects; and when some materials are in limited quantities to the point that sampling and
destructive techniques must be avoided like minerals in rocks.
Advances in technology have given rise to smaller devices and more powerful equipment. Miniaturization of XRD
systems is not a new issue. Since the 1960’s, one of the first attempts to develop portable devices for in situ XRD
analysis was part of a NASA program to analyze Moon soil and later to study Mars rocks and soil [1,2]. At first,
radioactive X-ray sources were used, until smaller and lighter X-ray tubes were developed by MOXTEK and AMPTEK
that could be used for in situ XRD analysis systems [3]. With a single source, X-ray Fluorescence (XRF) can be applied
simultaneously with XRD in portable devices, thus, elemental analysis provides complementary information that
facilitates phase identification and the detection of trace elements may provide information on the raw materials source.
The most common portable commercial XRD/XRF devices, DUETTO and XTRA, have a similar geometry and
detection system. The first is appropriate for non-invasive analysis, and the latter is suitable for invasive analysis
that require small amounts of sample [4]. ASSINGTM developed one of the lightest devices, characterized by Robin De
Wolf [5], for analyses on paintings [6,7]. Some modifications have been adapted from the Surface Monitor system by two
different groups to improve the diffraction peak shape [8] and to create a database of XRD patterns of 50 pigments of
pure powder pellets and dummy frescoes to evaluate the analysis of paintings [9]. Techno-X is another development of
a
e-mail: [email protected]
b
e-mail: [email protected]
Page 2 of 14 Eur. Phys. J. Plus (2019) 134: 286

Fig. 1. Optical path of X-rays in the XRD system.

a low-weight portable system used for paintings [10]. In the last few years, the prototype developed at C2RMF [11] has
been modified including a polycapillary optics and tested in fieldwork as a portable device to reduce acquisition time
significantly [12]. Most of these systems have a θ-θ geometry with mobile parts. Other systems such as DUETTO and
the portable XRD/XRF from C2RMF have a static X-ray source and a fixed 2D X-ray detector. Usually XRD/XRF
portable devices work in Angle Dispersive XRD (AD-XRD), but few systems operate with Energy Dispersive X-ray
Diffraction (ED-XRD) [13,14]. The latter systems have the advantage of lighter weight and since XRF elemental peaks
are detected at specific energies, the geometry can be modified to reduce the overlapping with diffraction signals. On
the other hand, a new promising XRD/XRF prototype has been developed [15] from a handheld XRF system but it
is still limited by the low resolution of diffraction peaks and interference by overlapping fluorescence peaks.
In comparison with bigger laboratory equipments, portable devices work with lower power and, therefore the
resulting diffractograms are of a lesser intensity and poorer resolution. Also, because of the mobility and the necessary
assembly and disassembly of the device it is important to design an easy way to set-up and align the system. To
evaluate the reliability of an XRD/XRF portable device, it is important to measure the accuracy in 2θ position as
well as the intensity and profile shape of the X-ray reflections. Also, it is necessary to characterize the XRF detection
system in order to determine which elements are detectable from its sensitivity and what are their detection limits.
This work presents the Analysis by Diffraction In Situ (ADIS) system, a recent development of a transportable
prototype for non-invasive XRD/XRF analyses. For the development of ADIS, a position detector was considered in
order to avoid moving parts that may shift in the process of assembly, disassembly and beam alignment. With Curve
Position Sensitive (CPS) detectors, all reflections within range are acquired at the same time, and the device has a
larger 2θ range than the usual XRD/XRF systems with position detectors. The features and reliability of ADIS are
discussed in this paper, and a methodology for evaluation of the assembly of a portable XRD/XRF system is proposed.
This prototype is compared with other XRD/XRF portable devices, and with two laboratory XRD devices with more
conventional configurations (Bruker D8 Discover and Rigaku Ultima IV). Some examples of the variety of materials
that ADIS can analyze are also presented. A diffractogram of a powdered malachite pigment sample is compared with
acquisitions from the laboratory devices. A mineral jadeite reference (both in powdered and solid forms) is compared
with an archaeological artefact, to stress the mean differences that can arise when a solid polycrystalline specimen
that cannot be sampled is analyzed.

2 Experimental set-up
ADIS is a system set for in situ and non-invasive XRD/XRF analysis. This system was developed using some compo-
nents from an INEL Equinox 100 diffractometer. It consist in an Curved Position-Sensitive X-ray Detector, model CPS
180 from INEL with a curvature radius of 180 mm. The X-ray source is a micro-focus Cu tube Petrick GmbH with a
water-cooling system, operational with a maximum power of 50 W, a 100 μm thick Be window, and a micro-focus size
of 50 μm. As a Cu Kβ filter the tube has a graphite (002) monochromator and a divergence slit with variable height,
fixed at 0.48 mm. A Matsusada Power Supplies series XR-605-60-02 is used for the X-ray tube. The XRD system has
an asymmetric Debye-Scherrer geometry for flat samples, in transmission mode [16]. Figure 1 shows the X-ray optics,
with the corresponding components that affect its path. The CPS has an angle step of 0.029◦ , and XRD FWHM is
0.2968◦ at 28.47◦ 2θ. The diffractometer’s angular divergence is 0.2233◦ , approximately. The CPS is a proportional
counter that works with a gas mixture with: 10% Xenon, 13% CO2 and Argon balance. For maximum performance of
the CPS, the valve’s pressure must be set between 5.5 and 5.7 bars.
Eur. Phys. J. Plus (2019) 134: 286 Page 3 of 14

Fig. 2. Design of the CPS detector and X-Ray tube support. a) Side view showing the plane of diffraction, parallel to the
supporting plate, b) frontal view of the plate with the X-ray tube and the removable powder sample support.

Fig. 3. Configurations allowed by the supporting plate: (a) for small objects, with a 2θ range: 0◦ –110◦ , and a 60 mm distance
from the X-ray tube to the center of the CPS; (b) for flat objects, with a 2θ range: 25◦ –140◦ , and a 113 mm distance from the
X-ray tube to the center of the CPS.

To perform simultaneously XRF analysis, an Amptek Silicon Drift Detector (SDD) detector has a 2.5 mm2 active
area, 12.5 μm thick Be window, and a FWHM of 130 eV at 5.9 keV. The detector is placed 45◦ over the horizontal
plane, with an 80◦ angle between the detector and the X-ray tube incidence.
Sensitivity and detection limits of the XRF detector have been measured with a series of standard references from
Micromatter Company: Mg 14686, Al 14687, SiO2 14688, GaP 14689, CuSx 14690, NaCl 24691, KCl 14692, CaF2
14693, Ti 14694, V 14695, Cr 14696, Mn 14697, Fe 14698, Co 14699, Ni 14700, Cu 14701, ZnTe 14702, GaAs 14703,
CsBr 14704, RbI 14705, SrF2 14706, MoO3 14707, Rh 14708, Ag 14709, Sn 14710, BaF2 14711, Au 14712, Pb 14713.
These references consist in metals and compounds deposited on 2 μm thick Kapton and Mylar films in concentrations
of micrograms per cm2 . Detection limits and sensitivity were calculated according to previous works [17,18]. For SDD
characterization, the detector was set at 3 cm from center of the sample surface, with a 180 s acquisition time under
typical XRD measurement conditions for the X-ray tube (35 kV, 0.8 mA). No filter was added.
To evaluate the XRD performance of ADIS, it was compared with two laboratory diffractometers, Bruker D8
Discover and Rigaku Ultima IV. For 2θ calibration and accuracy evaluation, a silicon standard reference material
(NIST SRM 640c) was used, and for intensity and peak resolution a ZnO standard reference material (NIST SRM
674b) was used. Bruker D8 Discover has a 2θ range: 4◦ –110◦ with a Cu X-ray tube operated at 40 kV and 35 mA for
645 s per XRD pattern, while Rigaku Ultima IV has a similar 2θ range (5–110◦ ) and the X-ray tube Cu was operated
at 40 kV and 30 mA for 690 s per XRD pattern.
ADIS works with 36 kV and 0.8 mA for the materials analyzed, while acquisition time changes depending on the
analyzed object. Pigments like malachite require 1800 s, while jadeite samples can be analyzed in 600 s. For crystalline
phase identifications the Match! 2.0 and the ICDD’s PDF-4 programs were used. For Rietveld refinement the GSAS II
program was used.

2.1 ADIS implementation

An aluminum plate was designed for the setting of the X-ray tube and the CPS detector as can be shown in fig. 2.
The electronic components were placed inside a wheeled table to facilitate the transportation of the diffractometer.
The plate was designed to accommodate an SDD X-ray detector for XRF analysis. This detector is located outside the
diffraction plane, fixed with a support that allows the change of the detection angle and distance to the sample surface.
The plate was designed to have two different configurations (see fig. 3): one for small pieces with a 6◦ incidence
angle and a 2θ range from 0◦ to 110◦ with a 60 mm distance from the area of analysis to the X-ray tube. The second
configuration is designed for larger, flat objects, has a 12.5◦ incidence angle, a 2θ range from 25◦ to 135◦ and a
113.8 mm distance from the area of analysis to the X-ray tube. The larger separation from the area of analysis to the
Page 4 of 14 Eur. Phys. J. Plus (2019) 134: 286

Fig. 4. (a) Complete ADIS system with water-cooling system inside the table. (b) Positioning of the incident X-rays on the
sample with beam spot on the fluorescent screen. (c) Intersection of the two cross line lasers in the center of the sample holder.
(d) Sample positioning for surface analysis of objects. Images (b) and (c) were acquired with the USB digital camera added to
the system.

X-ray tube in the later configuration is due to the fact that some parts of the X-ray tube impede the positioning of
the sample. The dimensions of detection head are 515 mm high, 238 mm wide and 12 mm thick.
To complete the adaptation of ADIS into an XRD/XRF portable system, the electronics, tank and hoses were set
inside the wheeled table. The complete system is presented in fig. 4(a). For transport, the assembled detection head
is carefully set inside a rigid suitcase and the table with the electronics is transported separately.

2.2 Calibration and sample preparation

Sample position is established with the Sample Spinner for Reflection and Transmission (SSRT), a sample holder
suited for powder analysis. The sample holder gives the position of the plane of analysis, the center of the sample
holder corresponds to the center of CPS detector. A fluorescent screen is used to align the X-rays with the desired area
of analysis (fig. 4(b)). Then, the plane of analysis is determined by the intersection of two cross line lasers in the center
of the sample holder (fig. 4(c)). The analyzed region of the objects is observed and registered with a high-resolution
USB digital camera that was added to the system. A circular collimator was added, allowing to change the spot size
to 2 mm diameter or less.
After setting the X-rays incidence area, the zero is calibrated and the linearization of the detector is calculated with
analysis of a silicon NIST SRM 640c standard reference. The SRM XRD pattern is compared with the corresponding
crystallographic sheet, to evaluate the accuracy and 2θ deviation of the calibration.
For solid samples, a lab jack is used to adjust the sample’s height until the position of analysis is reached, as can
be seen in fig. 4(d). An X-Y positioning stage (±35 mm in each direction) allows the movement of the object in order
to irradiate the required regions. This mechanical setup allows the examination of objects of different sizes and shapes
on their flattest surfaces.
Eur. Phys. J. Plus (2019) 134: 286 Page 5 of 14

Fig. 5. |Δd/d| curves of the different XRD devices for the NIST SRM 640c silicon standard.

3 Results and discussion

3.1 ADIS evaluation. XRD comparison with modular devices

Portable devices require a protocol and criteria to ensure the reliability of the measurements every time they are
reassembled. Accuracy gives certainty about phase identification, and resolution provides the capability to differentiate
one reflection from another that belongs to a different crystalline phase. Intensity hints about the required acquisition’s
time for a suitable analysis while relative intensity offers information about texture, detector efficiency at different
angles, and, when possible, weight proportion of crystalline phases.
In comparison with portable devices, modular devices do not need a constant evaluation for XRD accuracy. Due
the differences in configuration with modular XRD devices (D8 Discover and Ultima IV work with higher power and
have larger goniometers) it is expected that ADIS should have lower accuracy and intensity. To evaluate ADIS 2θ
calibration and accuracy, a silicon NIST SRM 640c was used. The same standard reference material was analyzed with
D8 Discover and Ultima IV diffractometers. The detector’s accuracy is given by the following equation [19]:

Δd
= − cot(θ) · Δθ, (1)
d
where d is the interplanar distance.
The absolute value of Δd/d vs. 2θ can be observed in fig. 5, where the difference between the modular devices is
negligible in comparison with ADIS. ADIS |Δd/d| curve is about 1.46 higher than the other two curves, that is, an
error in 2θ with ADIS will give an error in interplanar distances 1.46 times higher compared with the other devices.
Keeping in mind the capability of transportation, and the adaptability to work with different materials in less
controlled environments, the given error for ADIS still allows a crystalline phase identification for different materials
that belong to the same mineralogical group, as will be seen in the applications presented.
The quality of calibration affects the position of the reflected peaks, therefore the identification of crystalline phases.
Comparing the acquisition of NIST SRM 640c silicon standard, with its specific crystallographic card, it is possible
to evaluate the quality of acquisitions in relation with Δ2θ dispersion vs. 2θ. A comparison of Δ2θ dispersion for the
three devices is showed in fig. 6.
ADIS has a slightly lower accuracy in calibration, compared with the other devices. While dispersion in D8 Discover
is within −0.01◦ to 0.01◦ Δ2θ, Ultima IV and ADIS dispersions are within −0.02◦ to 0.02◦ Δ2θ. The differences in
goniometer size and beam focusing are factors that affect the lower accuracy for ADIS, as well as the lack of collimation
for diffracted peaks. The difference in ADIS’s 2θ position is still reasonable and permits a reliable crystalline phase
identification.
With respect to intensity, relative intensity, and profile shape evaluation a ZnO standard reference (NIST SRM
674b) was used. Profile shape is affected by the geometry and optics of the system, so it is necessary to model
the instrumental aspects of ADIS with a Rietveld refinement. For intensity comparison, a diffractogram with a 600 s
Page 6 of 14 Eur. Phys. J. Plus (2019) 134: 286

Fig. 6. Δ2θ dispersion for the three XRD devices, using NIST SRM 640c silicon standard.

Table 1. Relative intensities for the ZnO standard reference (NIST SRM 674b) acquired with all three XRD devices normalized
by the reflection at 2θ of 36.252◦ , compared with the corresponding PDF number for ZnO: 00-036-1451.

2θ [◦ ] Card ADIS ADIS D8 D8 Ultima IV Ultima IV

intensity intensity difference discover difference intensity difference
% intensity (%) (%)
31.769 57 63 9.5 62 8 59 3.4
34.421 44 43 2.3 50 12 46 4.3
36.252 100 100 0 100 0 100 0
47.538 23 25 8 24 4.2 21 9.5
56.602 32 32 0 32 0 30 6.7
62.862 29 30 3.3 27 7.4 27 7.4
66.378 4 7 42.8 8 50 4 0
67.961 23 28 17.8 23 0 22 4.5
69.098 11 15 26.7 14 21.4 11 0
72.560 2 3 33.3 5 60 2 0
76.953 4 6 33.3 7 42.8 4 0
81.368 1 4 75 5 80 2 50
89.6045 7 12 41.7 9 22.2 7 0
92.7808 3 6 50 6 50 3 0

acquisition was acquired with ADIS, and a 3600 s acquisition for a Rietveld refinement. The GSASII software in Debye-
Scherrer mode was used, considering that there are no appreciable differences between reflection and transmission
asymmetric Debye-Scherrer geometry [16].
To compare the diffraction pattern intensities, and evaluate the relative intensities, the NIST SRM 674b ZnO
standard reference was used. This standard reference is used for quantitative X-ray Diffraction analysis, has sharp
reflection peaks and high peak to background ratios, and because zinc generates a low fluorescence background when
using a Cu source, this component was selected for the evaluation.
The capability of a device to obtain diffractograms comparable with a reference data base is also evaluated by the
relative intensity of the diffractogram. Table 1 presents the relative intensities obtained with the different devices. The
measured differences for ADIS are certainly comparable to those obtained for the Bruker D8 Advance and Rigaku
Ultima IV diffractometers.
Eur. Phys. J. Plus (2019) 134: 286 Page 7 of 14

Table 2. Comparative chart of ADIS with two modular XRD devices.

ADIS Ultima IV D8 Discover

Goniometer size [mm] 180 285 305
Voltage [kV] 36 40 40
Current [mA] 0.8 30 35
Acquisition time (s) 600 645 690
◦ ◦
FWHM (at 36.252 ) [ ] 0.52 0.06 0.05
Intensity (at 36.252◦ ) [◦ ] 2458 30453 19380

Fig. 7. Optics and geometry of an asymmetric system.

For the ADIS system, it was necessary to evaluate the accuracy of the system and quality of calibration, as well
as any changes in resolution and profile shape and intensity (counts and relative intensity) in diffraction patterns.
Considering the difference in goniometer size and power of operation, the differences in accuracy and resolution are
considerable between a benchtop and a laboratory devices. The primary beam intensity is given by the following
equation [20]:
I = kZV P, (2)
where, k is a constant, Z is the atomic number, V the voltage and P the power. As D8 Discover and Ultima IV operate
with power 500 times higher than ADIS and considering the similar k for Cu and Mo anodes, according to eq. (2)
their primary beam intensity is 54 times higher.
If the acquisition intensities are compared with ADIS, for ZnO at 36.252◦ 2θ the intensity is about 11.5 times
higher with Ultima IV and about 6.8 times with D8 Discover. An overall comparison of ADIS with modular XRD
devices is shown in table 2.
ADIS device has the advantage that with short acquisition times it is possible to visualize all the diffraction peaks
at the same time, and it is possible to identify some crystalline phases. In the case of complex materials or Rietveld
analyses, acquisition times usually require more than 30 min.
To understand the effects of the prototype’s geometry and optics on the broadening and relative intensity of
reflection peaks it is important to evaluate the geometry of the system (see fig. 7). This device is equipped with a
0.48 × 9 mm2 divergence slit, according to a pseudo-parallel beam optics, the broadening of diffracted beams depends
on the angle of diffraction [21]. With an angle of 12◦ between the sample and the tube, the spot size has a 2.3 mm
width, approximately. The spot size is approx. 2 mm wide, as can be seen in fig. 4(b).
There is no collimation for diffracted beams, so the width of diffraction peaks is determined by eq. (3), that
according with the optics of a pseudo-parallel beam is given by
b sin(2θ − α) 180
w= , (3)
sin α R π
where w is the width of diffracted beams, b is the slit aperture, R is the radius of the detector and α is the angle
between the incident beam and the sample.
As can be observed in fig. 8, the reflection peaks are considerably wider in ADIS acquisitions in comparison with
the other modular devices.
The standard reference’s certificate reports a crystallite size = 0.2014 μm. With a residual wR = 8.798%, the
resulting TCHZ instrumental parameters are the following: U = 161.678, W = 177.558, V = −2.0, x = 4.886, y = 0.0,
and z = 0.158. The background was adjusted by Chebyschev polynomial of 25 coefficients.
Page 8 of 14 Eur. Phys. J. Plus (2019) 134: 286

Fig. 8. Diffractograms of a ZnO standard reference material (NIST SRM 674b) acquired with ADIS, D8 Discover and Ultima
IV. The different XRD devices worked under typical acquisition conditions for each equipment.

Fig. 9. FWHM of reflection peaks for ZnO standard (NIST SRM 674b) obtained with the different XRD devices.

With the same program the FWHM for ADIS, D8 Discover and Ultima IV were calculated. Figure 9 shows the
FWHM measurements for the three devices, ADIS FWHM is 10 times higher than D8 Discover and Ultima IV FWHM
at 31.723◦ 2θ, and this difference slightly increases with θ. This difference is due to the optics of a pseudo-parallel
beam of ADIS, the geometry of the system, and the absence of diffraction slits in the detector for reflected X-rays.

3.2 XRF detector characterization

The SDD detector employed for XRF analysis was characterized using the references mentioned above. Figures 10
and 11 show the sensitivity curve and the detection limits measurements for the ADIS system.
Eur. Phys. J. Plus (2019) 134: 286 Page 9 of 14

Fig. 10. XRF sensitivity curve of an Amptek silicon drift detector (SDD) with the X-ray Cu tube.

Fig. 11. Detection limits for the silicon drift detector with a Cu anode of ADIS.

The best sensitivity corresponds to Fe and Co —as can be expected for an X-ray Cu tube—, and it decreases
significantly for lighter elements such as Si, Al and Mg. The sensitivity is also very low for elements heavier than Cu
due to the poor excitation produced by the X-ray Cu tube. Nevertheless, for long acquisition times (more than 15 min)
it is possible to observe some heavier elements.
The minimum detection limit is around Z = 27 that corresponds to Mn, Fe and Co with less than 0.02 μg/cm2 .
For light elements such as Al, as well as heavier elements a minimum amount of 1 μg/cm2 is required in order to
get a detectable characteristic X-ray peak intensity. For elements in the Si to Fe range the detection limits less than
0.2 μg/cm2 .
Acquisition time in XRF is notoriously shorter than XRD because diffraction has a lower interaction probability
than fluorescence, typically by an order of magnitude or more [15]. A typical XRF measurement with suitable X-ray
intensity peaks (statistical error lower than 5%) generally requires 5 minutes for elements in the 1 to 8 keV range and
10 minutes for heavier elements. In comparison with other XRF portable devices with Mo or Rh X-ray tubes of similar
power [18], the acquisition time for ADIS is longer. The Kβ filter for Cu X-ray lines diminishes the photon counting
and there is lower contribution to fluorescence by the X-ray continuum background of the X-ray tube, so the generated
fluorescence is less intense.
Page 10 of 14 Eur. Phys. J. Plus (2019) 134: 286

Table 3. Main features and aspects of outstanding XRD-XRF devices for in situ analysis developed by other groups. ADIS
data are included in the bottom row.
Detector Detector Typical
X-ray Power XRD XRF XRF range 2θ range acquisition Weight
Name Reference
source [W] (FWHM) (FWHM) [keV] [◦ ] time [kg]
[◦ 2θ] [eV] [s]
◦ ◦
CheMin Co 10 CCD (0.35 ) CCD (250–300) 1–15 5–55 1800–9000 – [22]
◦
XRDF Cr, Cu, W 35 Si-PIN Si-PIN (150) 2–20 0–120 200 27 [23]
ASX-DUXT Fe 10 Si-PIN (0.12◦ ) Si-PIN (190) 1–12 24–134◦ 10800 30 [9]
◦
– Cu 30 Imaging plate SDD (150) 2–28 10–60 900 34 [11]
◦ ◦
DUETTO Co, Cu 10 CCD (0.3 ) CCD (250–300) 3–15 20–55 300 20 [4]
◦ ◦ ◦
XRD/XRF Cu 30 Si-PIN (0.12 ) Si-PIN (165) 2–12 10–140 3 s/ step 25 [8]
PT-XRD Cu 9 Si-PIN (0.2–0.3◦ ) Si-PIN (200) 2–10 0–90◦ 1800 11.5 [10]
Prototype 2 Ag 10 SDD (141–170 eV) SDD (136) 0–50 0–73◦ 300–600 7(a) [14]
◦
5–110 or
ADIS Cu 50 CPS 180 (0.3◦ ) SDD (130) 1–20 600–3600 12(a) This paper
25–140◦
(a)
Only detection head.

3.3 Comparison with other XRD/XRF portable devices

As was mentioned before, several strategies have been used in the development of portable XRD/XRF devices. Table 3
shows some examples of portable devices with different sources, geometries and detectors and reported data. ADIS
features area included in this table. DUETTO and Surface Monitor are the only commercially available devices on the
list.
Many of these equipments work with Bragg-Brentano geometry. The lightest one with a short acquisition time
is the PT-XRD system, developed in Japan [10]. Some of these devices have a better 2θ resolution than ADIS (e.g.
ASX-DUXT), but others have a lower resolution (CheMin). The resolution of the ADIS XRD detector is comparable
with the commercial device DUETTO. The devices that use position-sensitive detectors (PSD) have faster acquisitions,
but shorter angle ranges; as an XRD/XRF device based in position detectors, ADIS has an advantage in this feature.
Concerning XRF analysis, ADIS’ energy range is the same as most of the devices on the list and has the highest
resolution at 5.9 keV. The lightest device on the list is Prototype 2, that works with ED-XRD technique, nearly followed
by PT-XRD. Besides its dimensions and weight, ADIS is portable, and has the versatility to adapt accordingly with
the shape and size of the object: for objects smaller than 200 mm is possible to identify phases that have their principal
reflections from 5◦ to 20◦ 2θ.

4 Applications
A wide 2θ range allows the detection of reflection peaks from several materials: clays, pigments, metals, pyroxenes,
plasters, among others. Some examples of XRD/XRF analyses for powder and solid samples are presented in this
section to prove the versatility of ADIS. The ADIS prototype uses the same experimental conditions in the X-ray
tube for all materials (35 kV, 0.8 mA), as a consequence, acquisition time varies for different materials: metals and
lithics (e.g. greenstones pieces), plasters and limestone require short acquisition times (between 20 and 30 minutes);
while ceramics, sediments and pigments require longer acquisition times (about 1 hour). Background increments due
to iron fluorescence, induced by the copper X-ray tube of the system may increase the acquisition time significantly
in iron-rich materials (up to 2 hrs).

4.1 Powder analysis by XRD/XRF

With powdered samples, it is possible to reduce texture and detect all possible orientations within the 2θ range; this is
the reason why powder diffraction is the most reliable technique for crystalline phase characterization. Some materials
like pigments, soil samples or clays, are available in powdered form, in these cases the SSRT sample holder is used.
Eur. Phys. J. Plus (2019) 134: 286 Page 11 of 14

Fig. 12. Diffractogram of a malachite powder sample with ADIS, D8 Discover and Ultima IV.

Malachite is a mineral that can be found in many archaeological and historical contexts, as jewelry or pigments or
in minerals related to copper production. In archaeological sites through Mesoamerica, it can be found as a pigment
in murals and ceramic vases as well as green stones artifacts in offerings and burials. This pigment was also often used
in European paintings [24]. As its main element is copper, elemental characterization by XRF cannot differentiate
malachite from other copper-based green pigments or a mixture of azurite and a yellow pigment. With XRD is
possible to differentiate those materials and perform a suitable identification. This is also the case of raw materials of
copper minerals to produce copper artifacts and ores linked to the production of the metal [25].
Figure 12 shows the diffractograms of a powdered malachite pigment comparing acquisitions with ADIS, D8
Discover and Ultima IV. The main differences between ADIS and the other two devices are the peak width, and the
intensities (absolute and relative), as was expected from the features of these devices mentioned in previous sections.
In the corresponding ADIS XRF spectrum of the malachite powdered pigment, several elements were detected: Cu,
Fe, Mn, Ca, K, V, P and Si. With Match! 2.0 and PDF 4 database was possible to identify the main crystalline phase,
corresponding to PDF card 00-010-0399 malachite, but it was difficult to detect a minor phase rich in Si. Perhaps this
element is part of an amorphous component.

4.2 Application in solid samples

Due to the natural crystallization of some minerals, the diffraction patterns present preferred orientations and extinc-
tion of some reflection peaks in solid samples. Jadeite is a mineral of interest for archaeologists and geologists [26,27].
The effect of preferred orientations can be observed in fig. 13, where we present the comparison between a powdered
reference that corresponds to jadeite from a source of Motagua, Guatemala, a solid sample of the same source and
an analyzed archaeological artifact (3 cm diameter) made of jadeite discovered in the pre-Hispanic site of La Joya, in
the Gulf of Mexico. In this case, the XRD identification corresponds to jadeite with card number 04-011-6672. In the
corresponding XRF spectrum, several elements were detected: Fe, Ca, Si, Ti, K, Mn, Al, Cl, Cr and S.
Crystalline phase identification and preferred orientation analyses were made with ICCD’s PDF 4 data base, that
includes a preferred orientation simulation for some phases. In fig. 14, the preferred orientation of the solid sample
corresponds to [002].
The heterogeneity of this kind of minerals in the rocks could make difficult for techniques like non-invasive FTIR
or portable Raman to identify some minor phases or to differentiate phases that correspond to the same mineralogical
group. For example, diopside is a mineral that can be found, in most of cases, associated with jadeite. The composi-
tions of both minerals are different (CaMgSi2 O6 , for diopside, and NaAlSi2 O6 , for jadeite) but the crystal structure
(monoclinic system) and the highest reflection of both are close together. Figure 14 shows the diffractogram of another
archaeological piece (2 cm diameter), with a more complex mineralogical composition, that contains different crys-
talline phases (jadeite, diopside, quartz and kosmochlor) determined using the elemental information and the ICCD’s
PDF 4 data base considering the preferred orientation.
Page 12 of 14 Eur. Phys. J. Plus (2019) 134: 286

Fig. 13. Comparison between a powder sample of jadeite, a solid sample of jadeite and an archaeological object (earring).
Jadeite data sheet from ICDD’s PDF-4: 04-011-6672.

Fig. 14. Results of an XRD/XRF analysis for an archaeological object made of jadeite and other associated minerals.

4.3 Crystallite size

With the modeling of the ADIS geometry, Caglioti parameters were calculated, and with this data we calculated the
crystallite size of a piece of electrolytic copper. With a residual wR = 14.182%, and a preferred orientation on [111],
the calculated crystallite size is 0.039 μm. It has a small amount of cuprite in the surface: Cu wt = 89.8%, Cu2 O
wt = 10.2% (see fig. 15). The texture was calculated with a 20th-order spherical-harmonic polynomial.
Eur. Phys. J. Plus (2019) 134: 286 Page 13 of 14

Fig. 15. Image of Rietveld refinement of a piece of electrolytic copper, wR = 14.182%.

These results demonstrate that TCHZ parameters in Caglioti equation are not enough to get a good approximation
to the FWHM value of the reflection peaks generated by ADIS. Considering ADIS geometry, it could be useful to
consider the proposal of Rowles and Madsen [21] to include eq. (3) in the expression for FWHM value for the Gaussian
approximation.

5 Conclusions
ADIS is a versatile portable, XRD/XRF system that allows the analysis of either solid or powder samples. ADIS is
adaptable to the size of the samples and has a large 2θ range in comparison with other systems with position detectors.
In comparison with other XRD/XRF portable devices, ADIS is heavier, nevertheless, transportation is possible with
the wheeled table and the installation and calibration of the equipment is fast.
A general setting methodology was established for ADIS, ensuring an appropriate accuracy in 2θ position for
diffraction measurements. The Δ2θ dispersion in the reflection peaks position that are found below |0.03|◦ supports
the reliability of ADIS in 2θ calibration and crystalline phase identification. Considering the FWHM of reflection
peaks at 28.44 2θ (FWHM = 0.2986◦ ), it is possible to differentiate between phases, despite the overlapping peaks.
It was possible to identify minor phases in materials with two or more phases. The SDD sensitivity curve for the Cu
X-ray tube was also measured for XRF analysis. As for XRF, ADIS has a higher resolution in comparison to the other
mentioned XRD/XRF systems, and the energy range is the same for most of the devices.
ADIS’s acquisition times are between 10 minutes and 1 hour (depending on the complexity of the analyzed material),
while for XRF analyses, acquisition times range from 5 to 15 minutes.
ADIS has been developed mainly for cultural-heritage applications such as analysis of pigments in paintings,
metallic objects for manufacturing studies, and minerals and stone artifacts, that are difficult to analyze using other
non-invasive spectroscopic techniques. Nevertheless, its versatility allows performing analyses on other kinds of objects.

This research has been supported by grants from CONACYT LANCIC LN 279749, LN 293904 and CB 239609, as well as PAPIIT
UNAM IN110416 and IN112018. Authors acknowledge the technical support of IFUNAM staff Hesiquio Vargas, Juan Gabriel
Morales, Francisco Jaimes, Guillermo Picco Ruiz Dr. Alejandro Mitrani and Dr. Edgar Casanova for the development of the
ADIS system, Dr. Beatriz Millán Malo for technical support in XRD analyses from CFATA-UNAM; and Antonio Morales-Espino
from LAREC-IFUNAM for the XRD measurements performed by the D8 Discover equipment.

Publisher’s Note The EPJ Publishers remain neutral with regard to jurisdictional claims in published maps and institutional
affiliations.

References
1. P. Sarrazin, D. Blake, D. Bish, D. Vaniman, S. Collins, J. Phys. IV 8, 465 (1998).
2. D. Blake, Adv. X-Ray Anal. 43, 487 (2000).
3. S. Cornaby, A. Reyes-Mena, P. Moody, A. Stradling, T. Grow, L. Knight, Adv. X-Ray Anal. 45, 34 (2002).
Page 14 of 14 Eur. Phys. J. Plus (2019) 134: 286

4. M. Sarrazin, P. Chiari, G. Gailhanou, in Denver X-Ray Conference on Applications of X-Ray Analysis, 2009 (2009) pp. 175–
186.
5. R. De Wolf, Master of Science Thesis, Ghent University, 2012.
6. L. Van De Voorde, B. Vekemans, E. Verhaeven, P. Tack, R. De Wolf, J. Garrevoet, P. Vandenabeele, L. Vincze, Spectrochim.
Acta - Part B: At. Spectrosc. 110, 14 (2015).
7. L. Van De Voorde, J. Van Pevenage, K. De Langhe, R. De Wolf, B. Vekemans, L. Vincze, P. Vandenabeele, M.P.J. Martens,
Spectrochim. Acta - Part B: At. Spectrosc. 97, 1 (2014).
8. A. Pifferi, G. Campi, C. Giacovazzo, E. Gobbi, Croat. Chem. Acta 82, 449 (2009).
9. G. Gatto Rotondo, F.P. Romano, G. Pappalardo, L. Pappalardo, F. Rizzo, Microchem. J. 96, 252 (2010).
10. I. Nakai, Y. Abe, Appl. Phys. A Mater. Sci. Process. 106, 279 (2012).
11. A. Gianoncelli, J. Castaing, L. Ortega, E. Dooryhée, J. Salomon, P. Walter, X-Ray Spectrom. 37, 418 (2008).
12. M. Eveno, A. Duran, J. Castaing, Appl. Phys. A: Mater. Sci. Process. 100, 577 (2010).
13. M. Uda, Nucl. Instrum. Methods Phys. Res. B 226, 75 (2004).
14. A. Mendoza Cuevas, F. Bernardini, A. Gianoncelli, C. Tuniz, X-Ray Spectrom. 44, 105 (2015).
15. G. Hansford, J. Appl. Crystallogr. 51, 1571 (2018).
16. O. Masson, R. Guinebretière, A. Dauger, J. Appl. Crystallogr. 29, 540 (1996).
17. P.J. Potts, M. West, Portable X-Ray Fluorescence Spectrometry: Capabilities for in Situ Analysis (RSC Pub., Cambridge,
UK, 2008).
18. J.L. Ruvalcaba Sil, D. Ramı́rez Miranda, V. Aguilar Melo, F. Picazo, X-Ray Spectrom. 39, 338 (2010).
19. R. Guinebretiere, X-Ray Diffraction by Polycrystalline Materials (ISTE Ltd, Chippenham, Wiltshire, 2007).
20. R.E. Van Grieken, A.A. Markowicz, Handbook of X-Ray Spectrometry, Second (Marcel Dekker, Inc., New York, 2001).
21. M.R. Rowles, I.C. Madsen, J. Appl. Crystallogr. 43, 632 (2010).
22. D. Blake, D. Vaniman, C. Achilles, R. Anderson, D. Bish, T. Bristow, C. Chen, S. Chipera, J. Crisp, D. Des Marais, R.T.
Downs, J. Farmer, S. Feldman, M. Fonda, M. Gailhanou, H. Ma, D.W. Ming, R.V. Morris, P. Sarrazin, E. Stolper, A.
Treiman, A. Yen, Space Sci. Rev. 170, 341 (2012).
23. M. Uda, in X-Rays for Archaeology (Springer, 2005) pp. 3–26.
24. A. Roy, Artists’ Pigments, Vol. 2 (National Gallery of Art, Washington, 1993).
25. G. Rapp, Archaeomineralogy, 2nd edition (Springer, Dulluth, Minnesota, 2009).
26. F. Gendron, D.C. Smith, A. Gendron-Badou, J. Archaeol. Sci. 29, 837 (2002).
27. H.S. Cheng, Z.Q. Zhang, B. Zhang, F.J. Yang, Nucl. Instrum. Methods Phys. Res. B 219-220, 30 (2004).