Alan Stanley
Wed, 11/03/2021 - 06:49
Edited Text
METHODS OF CLAY MINERAL ANALYSIS USING THE AGILENT XCALIBUR NOVA XRAY DIFFRACTOMETER

by
Silene DeCiucies

A thesis submitted in partial fulfillment of the requirements
for graduation with Honors in Geology

Whitman College
2014

Certificate of Approval
This is to certify that the accompanying thesis by Silene DeCiucies has been accepted
in partial fulfillment of the requirements for graduation with Honors in Geology.

________________________
Nick Bader

________________________

Kirsten Nicolaysen

Whitman College
May 12, 2014

i

TABLE OF CONTENTS

Introduction ................................................................................................................................................. 1
Background .................................................................................................................................................. 3
Methods........................................................................................................................................................12
Oriented Sample Preparation .............................................................................................................................. 12

Random Powder Sample Preparation ........................................................................................... 15
Loading and Running oriented samples on the Agilent Xcalibur Nova .......................... 16
Loading and running random powder samples ........................................................................ 17
Loading and running oriented samples on a Siemens D-500 .............................................. 18
Data reduction and analysis .............................................................................................................. 18
Results ..........................................................................................................................................................20
Discussion and Conclusions ................................................................................................................24
Acknowledgements.................................................................................................................................28
Works Cited ................................................................................................................................................29
Appendices .................................................................................................................................................30

ii

IMAGE DIRECTORY

Figure 1

Ilustration of Bragg’s Law

5

Figure 2

Clay mineral structure

7

Figure 3

TO and TOT clay structure

8

Figure 4

Comparison of Agilent Xcalibur Nova vs Siemens D-500 XRD

10

Figure 5

Schematics of θ-2θ and single crystal XRDs

11

Figure 6

Glass coverslip sample

16

Figure 7

Top-down schematic view of sample and detector movement

17

Figure 8

Kaolinite on glass coverslip; glass slide and Millipore methods

20

Figure 9

Kaolinite on glass coverslip

21

Figure 10

Montmorillonite on glass coverslip

22

Figure 11

Kaolinite on glass coverslip vs. glass slide on Siemens D-500

22

Figure 12

Random powder sample preparation comparison

23

Figure 13

Peak migration

26

Table 1

Procedure for oriented sample preparation

14

Table 2

Procedure for random powder sample preparation

15

Table 3

Commands in CrysAlis Pro Data Reduction

19

iii

ABSTRACT

The purpose of this project was to design a working method for analyzing
clay minerals on Whitman College’s Agilent Xcalibur Nova single-crystal
diffractometer.

We experimented with sample preparation, programming and

orientation of the diffractometer, and data analysis techniques. We were able to
achieve comparable results for oriented samples of clay mineral standards on both
the Xcalibur Nova and the Siemens D-500 diffractometer at the Geoanalytical Lab at
Washington State University. The most successful method for preparing oriented
samples was pipetting the sample directly onto small circular glass coverslips,
mounting the coverslips vertically in the sample holder, and collecting a sequence of
oriented scans. Additionally, we determined that random air-dried powder samples
have the strongest signal relative to background when loaded in 0.7 mm capillary
tubes. In the future, standard spikes added to the samples may be useful for
correcting peak positions in the final scan.

1

INTRODUCTION
Clay minerals are critical for understanding Earth’s surficial geologic
processes. These minerals provide information about the geologic past in that clays
represent the ultimate weathered states of aluminosilicate minerals and therefore
reflect parent rock and weathering processes over time. Unfortunately, clay
minerals are somewhat difficult to identify because they are generally poorly
crystallized, have a high concentration of defects, and have very similar chemical
compositions (Brindley, 1980). X-ray diffraction (XRD) is currently the only widelyused technique for identifying clay mineral species, but not all types of X-ray
diffractometers have established methods for analyzing clays.
Single crystal XRD diffractometers are widely used today, but differ from the
 -2θ diffractometers typically used for clay X-ray diffraction (Chipera and Bish,
2001). Single-crystal diffractometers such as the Agilent Xcalibur Nova are
successful in many applications such as in the analysis of protein crystals (e.g. Pan et
al., 2012), but are not used extensively for clay mineral analysis. There have been
many innovations in sample preparation, including the spray drying of clays to form
randomly oriented spheres (Hillier, 1999), preparation of filter cakes to maximize
sample thickness (Moore and Reynolds, 1989), varying designs of sample holders
for powder samples, and solution washes for stabilizing and organizing clay samples
(Whittig and Allardice, 1986). However, no published method describes clay
mineral analysis using a single crystal diffractometer.
The goal of this project was to devise a reproducible and accurate method for
XRD analysis of clay minerals using the Agilent Technologies Xcalibur Nova
2

diffractometer. We experimented with different sample preparation techniques for
both oriented and random powder samples, with programming and orientation of
the Agilent XRD detector, and with pattern analysis. Our immediate goal was to
qualitatively identify clay mineral standards with consistent accuracy. Our longterm goal using this method is to obtain quantitative analysis of the constituent clay
minerals in a soil or rock sample.
BACKGROUND
X-Ray Production
X-rays are a type of electromagnetic radiation with wavelengths between 0.01
and 10 nm. This wavelength is approximately the same length as inter-atomic
spacings in crystals, which makes X-rays useful for studying crystal structures. Xrays are generated in a vacuum-sealed cathode ray tube by bombarding a cathode
(usually copper or molybdenum) with electrons sourced from a tungsten filament.
When these electrons collide with the copper source, outer valence electrons
become excited to higher, unstable energy levels, and then return to lower valence
states. As electrons fall to a lower-energy state they release X-ray photons. A
collimator tube directs X-ray photons at a sample, where they are transmitted,
refracted, scattered, or absorbed (Whittig and Allardice, 1986).
History and Concepts of XRD
Moore and Reynolds (1989) summarized the history of the development of XRD.
XRD was developed in 1912 by Max von Laue and his team of researchers at the
University of Munich. X-rays had been discovered in 1885, and by the turn of the
century X-rays were being used in medicine and throughout the scientific
3

community. Von Laue was inspired by the notion that crystals might have internal
regularity, and set about proving this by sending small-wavelength beams through a
crystalline solid. Von Laue’s first successful X-ray image was that of a copper sulfate
crystal, and this image confirmed that the regular spacings between 3-dimensional
planes of atoms could diffract X-rays. Von Laue’s experiment proved that crystals
are orderly and systematically-repeating three-dimensional arrangements of atoms.
The distances between these planes of repeating atomic structures are roughly the
same as X-ray wavelengths and can be diagnostic of the particular crystalline solid.
This conclusion inspired Sir William H. Bragg and his son William Lawrence Bragg
to extend von Laue’s work, and the team defined the field of X-ray crystallography
with Bragg’s Law (Moore and Reynolds, 1989).
Bragg’s Law explains that when a collimated beam of monochromatic X-rays of a
wavelength  at an incident angle  strikes a crystal, these rays will penetrate the
crystal and scatter off of the successive parallel atomic planes within the crystal
structure (Fig.1). Because the spacings of these atomic planes are constant
throughout the crystal, for a given spacing (d) there will be a critical angle at which
the refracted photons will be in phase with one another as they leave the crystal.
This critical angle occurs only where a beam reflecting off a lower plane will travel
exactly n further wavelengths (n) than a beam reflecting from the plane above it,
where n is an integer (Fig. 1).

4

Figure 1: Illustration of Bragg's Law: X-rays at a given angle of incidence are “reflected” from
planes of a crystal structure (Moore and Reynolds, 1989).

The greatest degree of constructive interference occurs where the angle of
refraction is equal to the angle of the incident beam, and a succession of these
beams from equally spaced atomic planes within a crystal at the same angle (in
phase) will constructively interfere to create reflected beams at angles according to
Bragg’s Law (Whittig and Allardice, 1986):
n= 2d sin 
As long as the wavelength and angle of the incident beam is known, the internal dspacings and identity of a crystal can be determined. X-ray diffractometers consist
of a source cathode with a known characteristic wavelength (usually Cu or Mo) a
sample holder and a sample detector, and a precise goniometer that allows the
incident angle as well as the angle of the detector to be controlled. This allows d
spacings to be calculated using the Bragg equation.

5

Clay XRD
X-ray diffraction was first applied to clays in the 1920’s by Hadding and Rinne. In
the 1930’s, Hendricks and Fry, and Kelly et al. were the first to prove that the clay
fraction of soils contained minerals, i.e. crystalline materials that could yield X-ray
diffraction patterns (Whittig and Allardice, 1986). Following these discoveries came
an explosion of knowledge of clay mineral structures (Moore and Reynolds, 1989).
Clay is operationally defined in two ways. The word “clay” refers to any sediment
particles less than 2 μm in diameter (Brady and Weil, 2009). “Clay minerals “refer to
sheet-structured alumino-silicate minerals that primarily occur in the clay-sized (2
μm) fraction of soils, sediments, sedimentary rocks, and weathered or altered rocks
(Nesse, 2000). Clay minerals form as a result of the weathering of silicate rocks
under low temperature conditions (Nesse, 2000). They are part of the phyllosilicate
or layer silicate group, which is an abundant mineral group at the surface of the
Earth due to weathering of felsic crustal rocks.
Phyllosilicates consist of tetrahedral and octahedral phyllosilicate sheets,
primarily composed of silica, oxygen and aluminum that combine to create different
mineral structures (Fig. 2). SiO4 silica tetrahedrons linked together at basal oxygens
make up tetrahedral sheets, and these tetrahedral sheets (T) are always combined
with an octahedral sheet (O). Octahedral sheets are made up of a cation such as Al3+
or Al2+ and 6 oxygen or hydroxyl anions. These are linked together so that each
oxygen or hydroxyl is surrounded by 3 divalent cations (trioctahedral sheet
silicates) or by 2 trivalent cations (dioctahedral sheet silicates) (Fig. 2). Clay
minerals are made up of an arrangement of tetrahedral and octahedral sheets.
6

These arrangements consist of either one tetrahedral and one octahedral sheet (1:1
or TO clays) or an octahedral sheet between two tetrahedral sheets (2:1 or TOT
clays) (Figs. 2,3). TO clays are held together by weak electrostatic bonds, whereas
TOT clays require cations between TOT layers to maintain charge balance (Nesse,
2000).

Figure 2: Clay mineral structure at the atomic level. Tetrahedral and octahedral sheets are
made of oxygen and silicon atoms; inter-sheet spaces are occupied by water molecules and/or
cations. (Brady and Weil, 2009).

7

Figure 3: TO clay structure (e.g. kaolinite) and TOT clay structure (e.g. smectite); interlayer
spaces can be occupied by ions or water molecules (Moore and Reynolds, 1989).

The area between TOT layers is occupied by cations, water molecules, or may be
empty; cations and the positive side of water molecules adhere due to the net
negative charge characteristic of clay minerals. The sheets develop their negative
charge due to broken sheet edges that expose negatively-charged oxygen atoms, as
well as due to ion substitution within the structure of the mineral (Nesse, 2000).
Clay minerals’ high cation exchange capacity, as well as their small size and high
surface area-to-volume ratio make them a unique class of minerals. Clays are an
important part of soil in that they provide structure, water holding capacity,
catalytic properties, and enable the soil to hold nutrients (Brady and Weil, 2009).
Illitic materials, for example, provide the principal source of the essential nutrient
potassium for plants (Moore and Reynolds, 1989). Clays also have many industrial
uses: they are used in determining the quality of petroleum reservoirs based on
clays present in sandstones (Eslinger and Pevear, 1988), in remediation of
hazardous wastes, production of ceramics, construction planning, and widely in

8

agriculture (Sivalingham, 1995; Moore and Reynolds, 1989). Additionally, clay
minerals can be diagnostic of weathering intensity in soils; for example, kaolinite
tends to indicate a stronger degree of weathering than illite, and therefore is used to
reconstruct paleoenvironment (Wilson, 1999).
As XRD remains the only commonly available technique for identifying clay
minerals, improvements in clay XRD methodology are essential for expanding our
scientific knowledge of this unique class of minerals. In this study, we test methods
for analyzing clays on the Agilent Xcalibur Nova diffractometer. Single-crystal
diffractometers such as the Xcalibur Nova use a large detector that intercepts all
refracted beams (Fig. 4, top; Fig. 5, left). In contrast, traditional θ-2θ diffractometers
such as the Siemens D-500 use a small detector positioned at an angle of 2θ from the
sample (Fig. 4, bottom; Fig 5, right). The goal of this project is to create a repeatable
sample preparation and analysis method for both oriented and random powder
samples that yields accurate results in a single-crystal XRD.

9

Figure 4: Top: Sample, X-ray source, and detector geometry in Agilent Xcalibur Nova at
Whitman College; Bottom: Siemens D-500, a traditional XRD for clay analysis, located at
Washington State University. Both diffractometers have a fixed X-ray source. Note that on the
Agilent Xcalibur, the detector and sample can rotate in the horizontal plane, whereas on the
Siemens, the sample and detector rotate around the circular track.

10

Figure 5. Left: single crystal XRD schematic: Right: classic θ-2θ diffractometer schematic. Note that
the small detector moves depending on incident angle in the θ-2θ diffractometer, whereas the large
detector intercepts the entire array with the single crystal diffractometer. Left: International Union of
Crystallography, 2006. Right: Grebenkemper, Jason, UC Davis Department of Chemistry, 2014)

11

METHODS

Oriented Sample Preparation
We used clay mineral standards of kaolinite (KGa-2) and montmorillonite
(STx-1b) from the Clay Minerals Society at Purdue University to analyze our
diffractometer’s performance on known oriented clay mineral samples. Oriented
samples are meant to align clay minerals parallel to their c-axes by having them rest
on their basal cleavages and crystal faces. This is useful because the distances
between these basal planes are diagnostic (Fig. 2 and 3). Oriented samples intensify
the diffraction peaks resulting from these c-axis spacings, making clay mineral
identification more likely. In all oriented-sample preparation methods, orientation
is achieved by suspending the clay particles in a liquid and allowing them to slowly
settle in their preferred sheet orientation as the liquid is removed.
The glass slide method (Moore and Reynolds,1998, pp. 195-196) was used
to create oriented samples. Briefly, this method entails making a slurry of clay and
water, washing with a homo-ionic solution to occupy interlayer cation exchange
sites with Mg2+ cations and then water to remove excess salts; the slurry is vortexed
and centrifuged in between each wash, and then pipetted onto a glass substrate and
allowed to dry. For our experiments we used three washes with with 0.1 M MgCl2
solution, followed by three washes with DI H2O. Samples were vortexed and
centrifuged at 2000 rpm for 3 min each between washes. In order to create uniform
samples, the same amount of clay was used in each sample (0.30-0.32 g) and after
the last spin, the same amount of liquid was left within the tube to be pipetted onto
12

the slides (about 1 cm of liquid in tube). This solution was vortexed for 2 minutes
before being applied to either a glass slide or a glass coverslip. Proportionally, the
same amount of solution was placed on each slide (1 mL of sample for a circular
slide of 2.54 cm2, and 4.33 mL of sample for a 4.82 cm2 rectangle). For expanding
clays such as montmorillonite, the final wash used 50% ethanol and 50% water to
minimize curling of the sample.
Glass coverslips were also prepared using the Millipore method (Moore and
Reynolds, 1998, pp. 192-195). This method uses a specialized Millipore vacuum
filter device with glass fiber filter papers (EDM Millipore, Billerica, Massachusetts,
USA) to create a thick, clay filter-cake. Our Millipore method preparations included
sequential washes with 0.1 M Mg Cl2, H2O, and 1% poly-vinyl-alcohol (PVA) solution
in that order. The PVA was used to give the filter cake a gel-like quality to reduce
curling. One gram of slurry was mixed with 30 mL of water and set on a stir plate for
5 minutes before pouring through the Millipore system. The filter cakes were
manually transferred to glass cover slips. See Table 1 for specific instructions.

13

Table 1: Procedure for oriented sample preparation (steps 1-4 not tested in this paper
because clay standards were used).
Description
Reference
1
Preliminary disaggregation with ball mill Moore and Reynolds, 1989,
then blender.
pp.181-182
2
Chemically remove iron oxides, organic
Moore and Reynolds, 1989, pp.
matter, unconsolidated materials,
182-186
carbonates or sulfates if necessary.
3
Sonicate and centrifuge
Moore and Reynolds, 1989, p. 187
4
Isolate clay fraction using centrifuge
Jackson and Barak, 2005
5
Saturate interlayer sites with Mg2+ions:
Moore and Reynolds, 1989, p. 186
Add 0.1 M MgCl2 solution. Add solution
to clay, vortex for 30 seconds, centrifuge
for 3 minutes at 6000 rpm, pour off
supernatant. Repeat 3 times.
6
Rinse off excess salts with water. Add DI
Moore and Reynolds, 1989; this
water to clay, vortex for 30 seconds,
paper
centrifuge for 3 minutes at 6000 rpm,
pour off supernatant. Repeat 3 times,
After last wash, pour off water but leave
about 1 cm of supernatant above clay in
bottom of test tube. For expanding clays,
last rinse contains 50% water 50%
ethanol. Vortex this slurry for at least 2
minutes.
7
Pipet onto glass coverslip, let air dry.
This paper pp. 12-13
8
Place dry sample in modeling clay on
This paper p. 16
Hampton base oriented as vertically as
possible. Place in XRD so that clay side
faces away from user and sample is
oriented parallel to beam.
9
Center sample holder using goniometer
Oxford Diffraction, 2003
wrench; check coverslip angle using 5
This paper pp. 16-17, Table 3
second runs.
10 Complete powder run as detailed in p. 17
of this report and in Appendices 1 and 2.

14

Random Powder Sample Preparation
In order to test random powder samples, shale samples (sample 1124404)
were acquired from the James Hutton Institute in Scotland, UK. These samples were
prepared using a spray drying method (Hillier, 1999) which yields a powder
composed of microspheres roughly 0.1 to 0.2 mm in diameter. The samples were
loaded in several different mounts for XRD analysis including packing microspheres
into glass capillary tubes of various diameters (0.3, 0.5 and 0.7mm), mounting
microspheres on a static micro-mesh micro-mount (MiTeGen, Ithaca NY), mounting
microspheres in Loctite® Superglue on Hampton Cryo Loops, and dissolving
microspheres in water to create an oriented sample. The capillary tubes were
observed under a microscope to ensure tight packing. See Table 2 for specific
instructions.
Table 2: Procedure for random powder sample preparation in capillary tubes (steps 1
and 2 not tested as these samples were made for us at the James Hutton Institute in
Scotland, UK.
Description
Reference
1 Disaggregate and isolate clay fraction as in Table 1.
2 Grind slurry and load into airbrush. Spray dry in
Hillier, 1999
drying chamber and collect spheres.
3 Top-load spheres into capillary tubes: drag a set of This paper, p. 15
staples over the top of the tube to vibrate the
spheres so that they pack tightly at the bottom of
the tube. Fill up to about 1 cm.
4 Insert capillary tube vertically into modeling clay
on a Hampton base
5 Set sample in XRD and center using the video feed
Oxford Diffraction, 2003
and goniometer wrench.
6 Complete simple powder run as detailed on pages
17-18 of this report and in Appendix 3.

15

Loading and Running Oriented Samples on the Agilent Xcalibur Nova
The oriented samples on glass coverslips were loaded into the XRD using a Hampton
base with a small dome of modeling clay stuck to the top. The sample was inserted
into the clay vertically, ensuring that it was as straight as and perpendicular to
horizontal as possible (Fig 6). The goniometer/sample controller in the XRD must
be set at 0° when inserting sample.

Figure 6: Kaolinite on glass coverslip, mounted in clay on Hampton base. Coverslip has a
diameter of 1.8 cm.

The base was then set on the base holder in the diffractometer with the clay side
facing away from the user, and adjusted so that the disc was parallel with the beam
(Fig. 7, bottom). Proper orientation was tested by activating the X-ray source for 5
seconds and observing the pattern created on the detector. The goal was to observe
a pattern with the thinnest possible shadow- this indicated that the sample was
parallel to the beam.

16

Figure 7: Top-down schematic view of
sample and detector movement
throughout an oriented sample scan

Once the sample was set, the detector was set to -40 (Fig. 7, top and Table 3) for the
maximum 2 angle of refraction. The diffractometer was controlled using the
powder setup in CrysAlis Pro (Agilent Technologies, Santa Clara CA) with one run, a
dwell time of 10 seconds, a scan length from 320-360, a step size of 1, and 2 darks.
(Appendix 1).
Loading and Running Random Powder Samples
The net static mount and Superglue mounts were mounted on Hampton
bases and centered using the sample camera and goniometer wrench. The capillary
tubes stood vertically in a clay plug on a Hampton base and were centered in the
same fashion. These were analyzed using a simple powder run with max 2 set to
80, a dwell time of 10 seconds and 2 dark frames. The 2 angle for the powders was

17

the same as for oriented samples, but the run did not require a step size because the
sample is analyzed continuously as it rotates on its stand (Appendix 2).
Loading and Running Oriented Samples on a Siemens D-500
Identical chemical sample preparation methods were used for
montmorillonite and kaolinite samples run on a Siemens D-500 at the Geoanalytical
Lab at Washington State University. Samples were prepared on 2.2 x 5 cm cover
slips. A sample on a large glass slide was also used because large slides are standard
on this diffractometer. A 2-32 scan was completed with a step size of 0.05 degrees
and a count time of 10 seconds. Each run took roughly 2.5 to 3 hours. These
parameters were chosen based on the standard runs typically used at this lab for
clay samples, and the low 2 peaks we were looking for. These parameters were
chosen to maximize peak intensity in the diagnostic low -2 region.
Data Reduction and Analysis
Output from the XRD was uploaded to Jade (Minerals Data Incorporated,
Livermore CA) and compared with Jade’s RDB minerals database. Before uploading
into Jade, oriented sample data was first integrated in the data reduction version of
CrysAlis Pro. Each image was read in the command shell, and then integrated using
CrysAlis software powder radial command (1024 bins, 0 min, 80 max). In Jade,
these files were viewed separately by opening each file individually, and also
overlaid to analyze the data using the overlay tab in the open file window; this
allowed individual scans to be selected, and outlying data frames to be removed by
18

noting their image number and excluding that image from the summed scan.
Remaining frames could be summed using merge overlays  take summation
feature in Jade to create one scan. Once a single scan was created, the background
must be removed either by accepting the default background or manually fitting the
background to the curve. XRD peak locations were compared with the RDB minerals
database to preliminarily identify clay minerals. Two  and intensity data was also
exported into Excel and R (R Development Team, Auckland New Zealand) where dspacing was calculated. Charts plotting 2 against intensity were created in R.

Table 3: Commands in CrysAlis Pro Data Reduction
Purpose
Command in CrysAlis Pro
5 second run to test
“sm i 5”
initial position of
sample
Set detector to -40°
“gt a 0 -40 0 0”
Read image
“rd IMAGE TITLE” (this
can be achieved by typing
rd into command shell
and selecting desired
image from list)
Powder Radial
“POWDER RADIAL 1024
integration
0 80 image name_new
image name

19

Example

rd z:\XcaliburData\Silene\
powder _6349\40

POWDER RADIAL 1024 0 80
D:\XcaliburData\Silene\powder
_6349\int_6349_1

RESULTS
Oriented Samples
The Millipore filter cake method yielded a scan with more background noise,
and less defined peaks than the glass coverslip method (Fig. 8).

Figure 8: Kaolinite on glass coverslip. Top scan shows glass slide method and bottom shows
millipore filter cake method; the Agilent Xcalibur Nova measured both spectra. Numbers on
peaks are d-spacings associated with each plane of atoms causing the diffraction. Note that
intensities are relative and that these peaks have not yet been corrected to known standard
(see Discussion).

20

Figure 9: Kaolinite on glass coverslip. Top scan from Agilent Xcalibur Nova, and bottom from
Siemens D-500. Dashed line represents diagnostic peaks for kaolinite documented in Moore
and Reynolds, 1989. Peak corrections are not corrected to the standard.

The kaolinite coverslip scan comparison on the Agilent Xcalibur and Siemens D-500
shows that both diffractometers measure peaks within about 0.2 Å of the accepted
d-spacings for the standard mineral (Fig. 9). Both are close to the diagnostic peaks
with narrow peaks and low background, especially the Siemens D-500.
Montmorillonite yielded similar results (Fig. 10); however, the Xcalibur Nova peaks
were closer to diagnostic peaks, and had less background noise in the crucial range
below 40 2θ. For samples run on the Siemens D-500, samples on glass coverslips
have much higher signal relative to noise than samples on glass slides (Fig. 11).

21

Figure 10: Montmorillonite on glass coverslip. Top scan from Agilent X-calibur Nova, bottom
scan from Siemens D-500. Dashed peaks are diagnostic montmorillonite peaks. Not corrected
to the standard.

Figure 11: Kaolinite prepared using glass slide method. Glass coverslip substrate is the top
scan and glass slide is the bottom scan. Note that scans appear to be far from diagnostic peaks
because they have not yet been corrected to the standard.

22

Random Powder Samples
Many different methods of sample mounting were used with the spray-dried
shale microspheres from the James Hutton Institute. The net static mount and
oriented glass coverslip scans produced the most background noise (Fig. 13), seen
in the large mounds beneath the peaks. In both of these scans, the peaks are difficult
to distinguish from this background noise. The two capillary tube mounts yield very
similar scans, but the 0.7 mm capillary tube has more distinct, narrow peaks than
the 0.5 mm relative to the background (Fig. 12).

Figure 12: Analysis of Hillier shale sample 1124404 using different sample mounting methods
on the Xcalibur Nova. Right side, bottom to top scans represent: net static mount, 0.5 mm
capillary tube, 0.7 mm capillary tube, and oriented sample on glass coverslip. Note that for
this comparison the background was not removed from any of the scans.

23

DISCUSSION AND CONCLUSIONS

Based on our analysis of oriented samples, mounting clays on a glass
coverslip using the glass slide method is a reliable working method for analyzing
clays. Compared to the Millipore filter cake method, it yields a scan with less
background noise and narrower, more distinct peaks. Furthermore, when
compared with data from the Siemens D-500, the data from the Agilent Xcalibur
Nova contained the same diagnostic peaks, and in the case of montmorillonite,
yielded results closer to diagnostic peaks in the RDB database. This comparison is
important, because the Siemens D-500 is widely used for clay XRD analysis (Bish
and Chipera, 2001).
Many of our scans of clay mineral standards (e.g. Figs. 8-12) produced peaks
that are not exactly in the expected locations for diagnostic c-axis d-spacings. This
lateral shift is most likely a function of the slide or coverslip not being aligned with
the detector at the start of the experiment, despite our initial stage orientation using
the X-ray shadow cast on the Agilent XRD’s detector by the coverslip. Surprisingly,
this peak shift is most pronounced in data obtained from the Siemens D-500, which
has a dedicated rigid sample holder (Figs. 4, 11). In the future, the best course of
action may be to add a powder of a known highly crystalline standard (e.g. quartz)
with a narrow intense peak, and adjust the scan output so that the observed quartz
peak matches its diagnostic 2θ spacing. To further reduce variation, we recommend
building a dedicated sample holder that secures the coverslips vertically.

24

Correcting the peak positions in this way may improve our ability to automatically
identify minerals in Jade.
Our analysis of individual scans from the Agilent diffractometer revealed that
peaks migrate somewhat over the course of the analysis, and seemed least accurate
at a high angle of incidence (Fig. 13). We cannot satisfactorily explain this behavior.
However, the resulting integrated scan matches diagnostic peaks well enough to be
identified automatically by Jade. Adding a known crystalline standard, with a thin
and intense peak at a known location, may allow us to determine whether to restrict
scans to a low angle of incidence.

25

Figure 13: Peak migration in a kaolinite scan: the angle of incidence alters the shape and
location of the peak but when all separate peaks are summed, the desired peak location is
attained. The diagnostic peak at 24.9 2 is given by the vertical black line. The vertical red
line represents the peak achieved when these scans are summed together. Note that these
peaks have not been corrected to a known standard which explains why they are off from
diagnostic peak.

For analyzing random powder samples, the recommended method of sample
mounting is filling a 0.7 mm capillary tube with spray-dried spheres (Table 2). This
method yields a scan with relatively low background, and narrow peaks that are
more distinguishable from the background than other methods. This may be
because the beam passed through more of the clay spheres, but the same amount of
glass as in a smaller capillary tube. Because we lack a spray-drying apparatus
26

(Hillier, 1999), we were not able to experiment with clay mineral standards for this
method, and the samples contained many constituent minerals, so further research
in random powder sample preparation is recommended; random powders may
prove simpler to run than oriented samples on the Agilent XRD.
The apparently better performance of coverslip substrate samples on the
Siemens D-500 warrants further thought and experimentation. Glass slides are the
traditional substrate for this diffractometer, but our results show that the coverslip
sample had much higher intensity peaks- perhaps this should be investigated
further on this diffractometer to make improvements to the widely accepted
method.
Another issue that will always be a problem in clay analysis is the curling and
peeling of expanding clay samples. We ameliorated this to some extent with a 50/50
water-ethanol wash, but pure montmorillonite samples still cracked and curled; this
affects the scan of the sample. Experimentation with blends of expanding and nonexpanding clays is recommended for further research.
In the future it will be important to experiment with natural soil samples and
mixed standards. Continuing with the random powder method also warrants
further research as this method requires less sample preparation time and scan
time. Quartz coverslips or a substrate with known diffraction peaks such as a mica
might reduce background noise for the oriented sample method.

27

ACKNOWLEDGEMENTS
Many thanks to the Stephen Hillier at the James Hutton Institute for his donation of
shale samples, to Owen Niell at Washington State University for his assistance in
running their Siemens D-500, to Doug Juers for his XRD support on the Agilent
Xcalibur Nova and creative ideas, and lastly to my endlessly supportive thesis
adviser, Nick Bader.

28

WORKS CITED
Bish, D.L, Chipera, S.J. 2001. Baseline Studies of the Clay Minerals Society Source
Clays: Powder X-ray Diffraction Analyses. Clays and Clay Minerals 49:5 pp.
398-409.
Brady, N C and Weil, R R. 2009. Elements of the Nature and Properties of Soils 3e.
Prentice Hall
Brindley, G.W. 1980. Quantitative X-Ray mineral analysis of clays. Monograph 5,
Mineralogical Society of London
Eslinger, E and Pevear, D.R.1988. Clay Minerals for Petroleum Geologists and
Engineers. Society fpr Economic Paleontologists and Mineralogists. Tulsa, AZ.
Hillier, S. 1999. Use of an Air Brush to Spray Dry Samples for X-ray Powder
Diffraction. Clay Minerals 34:1. pp.127-35.
Jackson, M.R. 2005, Soil Chemical Analysis: Advanced Course. UW-Madison
Libraries, Parallel Press.
Moore, D.M, and Reynolds, R.C. 1989. X-ray Diffraction and the Identification and
Analysis of Clay Minerals. Oxford: Oxford UP
Nesse, W.D. 2000. Introduction to Mineralogy. New York: Oxford UP
Oxford Diffraction Limited. 2003. User Manual; Xcalibur series Single Crystal
Diffractometers. Version 1.4. Oxforshire, UK.
Pan, Q., J. Duan, and Q. Chang. 2012. Comparative Crystal Structure Determination of
Griseofulvin: Powder X-ray Diffraction versus Single-crystal X-ray Diffraction.
Chinese Science Bulletin 57:30
Sivalingham, S. 1995. Atomic Structure and Crystal Chemistry of Clays Used In
Remediation of Hazardous Waste. Great Plains/Rocky Mountain Hazardous
Substance Research Center. TapanAm Associates, Leawood, KS.
Whittig, L. D., and W. R. Allardice. 1986. "X-Ray Diffraction Techniques." Methods of
Soil Analysis, Part 1: Physical and Mineralogical Methods. Madison: American
Society of Agronomy
Wilson, M.J. 1986. The Origin and Formation of Clay Minerals in Soils: Past, Present
and Future Perspectives. Madison: American Society of Agronomy

29

APPENDICES

Appendix 1: Runs list for oriented sample as described on p. 17 of this report. To
get to this window, open the Powder power experiment dialog box and select the
Edit runs list tab. Note that one run has been selected and the file is routed to your
personal folder. Select the single run and press Edit to edit the run as in Appendix 2.
Dwell time is adjusted in the first powder experiment window.

30

Appendix 2: Editing run for oriented samples. Note that based on the start and end
(in degrees), the run is completing a 40° scan. Once these specifications have been
set, press OK button to return to the runs list press OK to return to powder
experiment window select start with analysis.

31

Appendix 3: Powder experiment window for random powder samples as described
on pp. 17-18 of this report. Once all of these specifications have been set, select Start
with analysis to begin experiment.

32