E348 Journal of The Electrochemical Society, 164 (12) E348-E351 (2017)

Rapid, Non-Invasive Method for Quantifying Particle Orientation

Distributions in Graphite Anodes
Paul Baade, Martin Ebner,∗ and Vanessa Woodz
Department of Information Technology and Electrical Engineering, ETH Zurich, Switzerland

For materials with 1D or 2D lithiation pathways and non-spherical particle shapes, knowledge of the crystallographic grain orientation
in the particle and the active particle orientation in the porous electrode is important for quantifying battery performance. Here we
study graphite anodes and show how X-ray diffraction based texture measurements can be used to quantify both the particle
orientation, which develops during the coating and calendaring process, and the grain orientation within a specific type of graphite.
This lab-based, non-invasive approach to study electrode structure and active particles can assist in engineering improved lithium
ion batteries.
© The Author(s) 2017. Published by ECS. This is an open access article distributed under the terms of the Creative Commons
Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any
medium, provided the original work is properly cited. [DOI: 10.1149/2.1291712jes] All rights reserved.

Manuscript submitted June 23, 2017; revised manuscript received August 14, 2017. Published September 8, 2017.

In order to optimize a lithium ion battery, it is important to under-
stand the lithiation pathways within the active material1–5 as well
as the lithium transport through the pore space of the electrode,
which is governed by the shape and orientation distribution of par-
ticles in the electrode.6,7 Here, we present X-ray diffraction (XRD)-
based texture measurements as a quantitative, non-invasive method
to characterize the crystallographic grain orientation of particles and
the orientation of these particles within lithium ion battery porous
electrodes.
The importance of this structural information is highlighted in the
case of graphite, which is the most widely used active material for
negative lithium ion battery electrodes. The two-dimensional struc-
ture of graphite is responsible for the anisotropy of its electronic,
ionic, and thermal conductivity as well as its mechanical properties.8
(De)lithiation occurs along the (002) planes, causing directional vol-
umetric expansion in the [001] direction, which can also translate to
directional changes in electrode porosity as the graphite particles swell
into the pore space during lithiation.9 The two-dimensional nature of
graphite is also reflected in the shape of graphite particles: they are
naturally platelet-shaped.10 Earlier work has shown that non-spherical
particles align during porous electrode fabrication, leading to large tor-
tuosity of the pore-phase, reducing effective transport of lithium in the
electrolyte.6 Tortuosity can be somewhat reduced by using spherical
graphites, which often exhibit complex grain distributions depending
on the manufacturing process.10
For graphite anodes, the electronic and ionic conductivity of the
graphite particles does not limit cell performance,11,12 so most efforts
have focused on quantifying and reducing tortuosity, which limits
the fast charging of high energy dense cells.13,14 To determine tor-
tuosity of an electrode, numerical diffusion simulations can be run
on a three-dimensional representation of the electrode,15 or, because
porous electrodes obey the Bruggeman relationship,16 an estimation
of tortuosity can be obtained from particle shape and orientation dis-
tributions within the electrode.6,17
Current techniques used to quantify orientation distributions of
particles in porous electrodes have drawbacks. Scanning electron mi-
croscopy (SEM) cross-sectional images are destructive to the sam-
ple under investigation and can present ambiguities due to their 2D
nature.14 Focused-ion-beam (FIB)-SEM tomography provides quan-
titative 3D information but it can require elaborate sample preparation
and is a time consuming measurement that can access only a small
electrode volume.14,18 X-ray tomography is data intensive and the in-
struments and computational resources may not be readily available
to a large number of users.19
In contrast, our XRD-based texture measurement approach pre-
sented here is performed with a laboratory XRD system on centimeter-
∗ Electrochemical Society Member.
z
E-mail: vwood@ethz.ch
size electrodes with no special sample preparation required. Further-
more, the experiments can be fast (<30 min) and the data generated
is small (<200 kB), enabling rapid assessment of a large number of
samples.
While texture measurements have been used to quantify orientation
of grains in metal sheets such as cold-rolled copper20 or thin films,21,22
here we show that this technique can be extended to porous electrodes.
For active materials such as graphite that exhibit a preferred orienta-
tion at the single particle level, this technique can be used to extract
the particle orientation distribution, the grain orientation distribution
within the particles, and information about particle deformation and
fracture during manufacturing.
Experimental
Measurement setup.—The concept of XRD texture measurements
is to obtain the orientation distribution of crystal planes in a sample.
To perform a texture measurement, the intensity of X-rays diffracted
from a sample at a fixed scattering angle 2θ is recorded for all possible
orientations. The 2θ angle is chosen to satisfy the Bragg condition of a
crystal direction of interest. Here we measure the orientation of (002)
planes in a graphite electrode, so 2θ = 26.53◦ is chosen. Orientation
is described by the elevation angle α and an azimuth angle β of the
scattering vector k, which is normal to the crystal planes that contribute
to the recorded intensity.
Texture measurements can be obtained by either (1) keeping the
source and detector fixed to probe the desired 2θ angle and tilting the
sample using an Eulerian cradle, or (2) keeping the sample fixed and
moving both the source and detector along coupled trajectories. We
use the latter approach, which is shown in Figure 1a.22 The electrode
is positioned with the current collector parallel to the xy-plane. This
requires an XRD system with an in-plane axis, which allows the
detector to rotate out of the xz-plane.
Discrete steps for α and β are chosen. To vary α, the diffraction
plane sweeps from being orthogonal to the xy-plane at α = 0◦ to
parallel to the xy-plane at α= 90◦ . Figure 1 shows the in-plane mea-
surement position at α= 90◦ , where the diffraction plane is parallel to
the sample plane. At each α step, the sample stage is rotated to move
the electrode around its surface normal (which is collinear with the
z-axis of the goniometer) to allow scanning of β from 0◦ to 360◦ .
A pole figure is produced by attributing the recorded X-ray inten-
sity to points on a hemisphere defined by a unit vector with α and β and
projecting the result stereographically onto the plane (Figure 1b). As
described in the SI (part 2), the pole figure is corrected for defocusing,
absorption, and background effects. The result is normalized and the
intensity is represented in multiples of random distribution (m.r.d.).23
An m.r.d. value of 1 corresponds to a random orientation while values
higher than 1 indicate a preferred orientation.

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Journal of The Electrochemical Society, 164 (12) E348-E351 (2017)
Figure 1. a) Experimental setup for in-plane pole figure texture measure-
ments. b) Example result on a graphite electrode of the intensity projection
onto the plane is depicted in perspective below the setup, with the green to
yellow color scale representing increasing multiples of random distribution
values. The profile of the overall measured distribution by the pole figure
(solid line) is compared with that of uniform distribution (dashed line).
Rotationally symmetric samples lead to pole figures with no β
dependence. This is the case for the graphite electrodes studied here.
In this case, it is convenient to compute the average over β and display
the result as a function of the single parameter α (Figure 1b). The m.r.d.
value of 11 at α = 0◦ indicates that the [002] direction is 11-fold more
likely to be in the z-direction than for a uniform distribution. This
indicates that the (002) planes in graphite are preferentially aligned
parallel to the current collector, which is expected from previous
investigations using SEM and X-ray tomographic microscopy and
can be explained by the alignment of graphite particles with their
major axes parallel to the current collector during fabrication.6,10
Deconvoluting particle orientation and grain orientation within
particles.—In porous electrodes containing polycrystalline particles,
the measured orientation distribution f (α) is a convolution of the
particle orientation distribution p(α) and the grain orientation distri-
bution within a single particle g(α p ). Here, α p is the elevation angle
relative to the particle plane (Figure 2a). This is important for battery
grade graphite, where the basal planes exhibit a preferred orientation
around the particle pole.24,25 Indeed, both a grain orientation distri-
bution within particles as well as a particle orientation distribution
within the electrode can be seen from cross-sectional SEM images
(Figures 2b, 2c). Such images, however, are not suited to quantify
both of them unambiguously.
From a single texture measurement, the individual distributions
that contribute to the pole figure cannot be separated without prior
knowledge of either one of the distributions. This would require ei-
ther (1) tomographic measurements, particle identification, fitting, and
identification of major axes, or (2) measurement of the grain distri-
bution on a statistically significant number of particles through single
particle XRD.
Here, we show that it is possible to deconvolute both orientation
distributions by compressing electrodes from their initial thickness
after coating to various thicknesses. By obtaining pole figure mea-
surements on a series of electrodes of different but known thicknesses
and assuming that the grain orientation stays constant for each, we are
able to uniquely identify both distributions.
Here, the March model26 which was developed to describe fiber
textures in rock formations is used to describe the two contributions.
Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
E349
Table I. Thickness, porosity, orientation parameter, and degree
of preferred orientation for the six electrodes studied in order to
deconvolute the intra-particle grain orientation from the particle
orientation in the electrode.
Sample 1 2 3 4 5 6
Thickness d [μm] 135 129 121 115 109 101
Porosity ε 0.586 0.568 0.539 0.497 0.470 0.430
Orientation parameter r p 0.124 0.119 0.111 0.106 0.100 0.093
Degree of preferred 76.52 77.25 78.44 79.20 80.12 81.20
orientation [%]
For the grain orientation within the particle, the March-Dollase func-
tion is chosen, which was developed for volume conserving compres-
sion and is implemented in most Rietveld refinement programs as a
texture parameter to account for deviating peak heights:27
    −3/2
g α p = r g2 cos2 α p + 1 r g sin2 α p ,
where rg is the particle orientation parameter, which can have values
between 0 (in the case of single crystals) and 1 (for the case of a
uniform grain distribution).
For the particle orientation distribution, we derive a novel, modified
March function that approximates uniaxial compression of porous
electrodes
p(α) = r p (r p2 cos2 α + sin2 α)−3/2 ,
where rp is the orientation parameter of the particle distribution. This
allows for a change in the sample volume, which occurs in the battery
electrodes during calendaring. We demonstrate that this is a valid
expression by studying porous electrodes made with particles of boron
nitride single crystals (SI, part 3).
The full spectrum of possible distributions from uniform (r p =
1, r g = 1) to perfectly aligned or single crystalline (r p = 0, r g = 0)
can be described through a single orientation parameter r p/g . Further-
more, the orientation parameter r p of the particle orientation distribu-
tion can be linked to the thickness d of the electrode r p = d/d0 , where
d0 is a virtual thickness at which the particles in the electrode would
have a uniform orientation distribution. Since the grain orientation
parameter r g is constant and the particle orientation parameter r p is
linked to the measured thickness, there are only 2 fitting parameters
(r g and d0 ) for a series of multiple electrodes.
The orientation parameter can be expressed as the degree of pre-
ferred orientation (i.e., the fraction (in percent) of aligned crystallites
or particles) calculated from the respective distribution profile. A
definition for the degree of preferred orientation for March-Dollase
function was proposed by Zolotoyabko,28 and we adapt it here for
the modified March function describing the particle orientation in a
porous electrode (SI, part 3).
Figure 2a presents an approach to deconvolute the particle orien-
tation and intra-particle grain orientations for electrodes made from
commercially available graphite particles. A slurry is spread by a
doctor-blade onto Cu-foil (SI, part 1) such that the electrode has a
thickness of 135 μm after drying (58.6% porosity and 0.94 g/ml
density). This electrode is compressed to different thicknesses, and,
for each thickness, pole figures are measured and analyzed with the
MTEX texture package in MATLAB29 (SI, part 2). Figure 2d plots
the averaged intensity of the measured orientation distribution, f (α)
for electrodes with the thicknesses and porosities tabulated in Table I.
Each sample has its peak at α = 0, which means that most of the
graphite (002) planes are parallel to the current collector.
As described in the SI (part 2), a MATLAB script convolutes
particle and grain distributions and carries out a least squares fit to
determine the best distributions for all electrodes. This procedure pro-
vides uniquely defined distributions for grain and particle orientations
for each electrode (Figure 2e). These convolutions are shown in Fig-
ure 2d (green solid lines), which show good agreement with the mea-
sured data (symbols). The distributions of the 6 samples can be well
 E350 Journal of The Electrochemical Society, 164 (12) E348-E351 (2017)
described with r g = 0.285 (the corresponding degree of preferred ori-
entation is 61.17%) and d0 = 1086 μm. This solution converges and
the error is estimated in SI (part 2). Inserting the measured thicknesses
leads to the particle orientation parameters, which are summarized in
Table I together with the degree of preferred orientation.
Results and Discussion
Grain orientation within graphite particles and their orientation
in electrodes.—In Figures 2d and 2e the peaks in the m.r.d. at α = 0◦
of all distributions indicate that graphite particles are aligned parallel
to the current collector and that the grain distributions within the
particles show a preferred orientation around the particle poles (see
inset in Figure 2a). The lower m.r.d. value for the grain orientations in
the particle indicate that the intra-particle grains are less well oriented
than the particles themselves. With decreasing porosity, the degree of
alignment of the particles increases because of the compression forces
during the calendaring process.
Identifying particle deformation.—To date, it has been difficult to
observe the onset and extent of particle deformation or cracking that
can occur during electrode calendaring;6 however, texture analysis
presents an opportunity. The modified March model predicts that with
decreasing electrode thickness (i.e. decreasing porosity), the particles
will increasingly align parallel to the current collector, resulting in
a larger m.r.d. value at α = 0◦ . Figure 3 shows that for electrodes
compressed to below 40% porosity, the m.r.d. maximum decreases.
We explain this by noting that, at low porosities, the graphite particles
are not free to reorient during compression. Rather particle-particle
interactions can influence the positioning of particles. Indeed, an SEM
cross-sectional image inset of the 28% porosity electrode shows a
particle bent around two others. The modified March model proposed
above assumes a constant grain orientation distribution and an affine
transformation (i.e., one in which planes are preserved) and cannot
account for this behavior. Therefore, a deviation from linearity of the
maximum m.r.d. values of f (α) as a function of electrode compression
can be a robust method to determine at what point in the calendaring
Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
Figure 2. a) The measured orientation dis-
tribution (green symbols) and the deconvo-
luted individual contributions coming from
the particles in the electrode (blue line) and the
grains in the particle (red line). b,c) SEM im-
ages show these individual contributions. d)
Orientation distributions for electrodes with
porosities ranging from 0.43 to 0.586. The
solid green lines are the result of fitting a con-
volution of the individual orientation distribu-
tions. The corresponding particle orientation
distributions (blue) and grain orientation dis-
tribution with a single particle (red) are shown
in e).
process the deformation is dominated by inter-particle interactions. An
extreme deviation from the linear model is an indicator that individual
particles are deformed or fractured.
Summary
We show that texture measurements can be used to rapidly and
non-invasively quantify the particle orientation distribution in porous
electrodes, the grain distribution within single particles, and the poros-
ity at which particle interactions begin influencing particle packing.
This lets us monitor the onset at which particle deformation is likely to
occur. We have extended the March model of texture to porous elec-
trodes so that the individual distributions can be reliably described
with a single parameter. While the range of applicability of the model
will depend on particle shape, material properties, and processing
conditions, in general, texture measurements can contribute to under-
standing morphological anisotropies of various particle systems.
Figure 3. The maximum of the overall measured pole figure is plotted versus
the electrode porosity. The solid green line indicates the predicted maximum.
The shaded area marks a 10% error for the grain orientation distribution. At
low porosities, the m.r.d. maximum deviates (red circles) from prediction,
indicating inter-particle interactions as shown in the SEM.
 Journal of The Electrochemical Society, 164 (12) E348-E351 (2017)
Acknowledgments
The authors thank the Swiss Commission for Technology and
Innovation and the European Research Council for financial support.
Cross-sections of electrodes were prepared at ScopeM.
References
1. M. Winter, J. O. Besenhard, M. E. Spahr, and P. Novák, “Insertion Electrode Materials
for Rechargeable Lithium Batteries,” Adv. Mater., 10(10), 725 (1998).
2. J. R. Dahn, “Phase diagram of Lix C6 ,” Phys. Rev. B, 44(17), 9170 (1991).
3. R. Malik, A. Abdellahi, and G. Ceder, “A Critical Review of the Li Insertion Mech-
anisms in LiFePO4 Electrodes,” J. Electrochem. Soc., 160(5), A3179 (2013).
4. M. Wagemaker and F. M. Mulder, “Properties and promises of nanosized insertion
materials for li-ion batteries,” Acc. Chem. Res., 46(5), 1206 (2013).
5. J. Zheng et al., “Li- and Mn-Rich Cathode Materials: Challenges to Commercializa-
tion,” Adv. Energy Mater., 7(6), (2017).
6. M. Ebner, D. W. Chung, R. E. Garcı́a, and V. Wood, “Tortuosity anisotropy in lithium-
ion battery electrodes,” Adv. Energy Mater., 4(5), (2014).
7. D. Kehrwald, P. R. Shearing, N. P. Brandon, P. K. Sinha, and S. J. Harris, “Local Tor-
tuosity Inhomogeneities in a Lithium Battery Composite Electrode,” J. Electrochem.
Soc., 158(12), A1393 (2011).
8. M. S. Dresselhaus and G. Dresselhaus, “Intercalation compounds of graphite,” Adv.
Phys., 30(2), 139 (1981).
9. Y. Qi, S. J. Harris, and “,In Situ Observation of Strains during Lithiation of a Graphite
Electrode,” J. Electrochem. Soc., 157(6), A741 (2010).
10. M. Mundszinger, S. Farsi, M. Rapp, U. Golla-Schindler, U. Kaiser, and M. Wachtler,
“Morphology and texture of spheroidized natural and synthetic graphites,” Carbon,
(111), 764 (2017).
11. H. Buqa, D. Goers, M. Holzapfel, M. E. Spahr, and P. Novák, “High Rate Capability
of Graphite Negative Electrodes for Lithium-Ion Batteries,” J. Electrochem. Soc.,
152(2), A474 (2005).
12. M. Heß and P. Novák, “Shrinking annuli mechanism and stage-dependent rate capa-
bility of thin-layer graphite electrodes for lithium-ion batteries,” Electrochim. Acta,
106, 149 (2013).
13. I. V. Thorat, D. E. Stephenson, N. A. Zacharias, K. Zaghib, J. N. Harb, and
D. R. Wheeler, “Quantifying tortuosity in porous Li-ion battery materials,” J. Power
Sources, 188(2), 592 (2009).
Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
E351
14. M. Ebner and V. Wood, “Tool for Tortuosity Estimation in Lithium Ion Battery Porous
Electrodes,” J. Electrochem. Soc., 162(2), 3064 (2015).
15. S. J. Cooper, A. Bertei, P. R. Shearing, J. A. Kilner, and N. P. Brandon, “TauFactor:
An open-source application for calculating tortuosity factors from tomographic data,”
SoftwareX, 5, 203 (2016).
16. D. -W. Chung, M. Ebner, D. R. Ely, V. Wood, and R. Edwin Garcı́a, “Validity of the
Bruggeman relation for porous electrodes,” Model. Simul. Mater. Sci. Eng., 21(7),
74009 (2013).
17. D. A. G. Bruggeman, “Berechnung verschiedener physikalischer Konstanten von het-
erogenen Substanzen. III. Die elastischen Konstanten der quasiisotropen Mischkörper
aus isotropen Substanzen,” Ann. Phys., 421(2), 160 (1937).
18. M. F. Lagadec, M. Ebner, R. Zahn, and V. Wood, “Communication—Technique for
Visualization and Quantification of Lithium-Ion Battery Separator Microstructure,”
J. Electrochem. Soc., 163(6), A992 (2016).
19. P. Pietsch and V. Wood, “X-Ray Tomography for Lithium Ion Battery Research: A
Practical Guide,” Annu. Rev. Mater. Res., 471229(1), 1 (2017).
20. H. J. Bunge and F. Haessner, “Three-dimensional orientation distribution function of
crystals in cold-rolled copper,” J. Appl. Phys., 39(12), 5503 (1968).
21. M. D. Vaudin, M. W. Rupich, M. Jowett, G. N. Riley, J. F. Bingert, and I. Introduction,
“A method for crystallographic texture investigations using standard X-ray equip-
ment,” J. Mater. Res., 13(10), 2910 (1998).
22. K. E. Nagao Keigo, “X-ray thin film measurement techniques VII. Pole figure mea-
surement,” Rigaku J., 27(2), (2011).
23. U. F. Kocks, C. N. Tomé, and H. R. Wenk, Texture and Anisotropy: Preferred Orienta-
tions in Polycrystals and Their Effect on Materials Properties, Cambridge University
Press, 2000.
24. M. Wissler, “Graphite and carbon powders for electrochemical applications,” J. Power
Sources, 156(2), 142 (2006).
25. D. Aurbach, E. Zinigrad, Y. Cohen, and H. Teller, “A short review of failure mecha-
nisms of lithium metal and lithiated graphite anodes in liquid electrolyte solutions,”
Solid State Ionics, 148, 405 (2002).
26. A. March, “Mathematische Theorie der Regelung nach der Korngestalt bei affiner
Deformation,” Zeitschrift fuer Kristallographie, 81, 285 (1932).
27. W. A. Dollase, “Correction of Intensities for Preferred Orientation in Powder Diffrac-
tometry: Application of the March Model,” J. Appl. Cryst, 19, 267 (1986).
28. E. Zolotoyabko, “Determination of the degree of preferred orientation within the
March–Dollase approach,” J. Appl. Cryst, 42, 513 (2009).
29. F. Bachmann, R. Hielscher, and H. Schaeben, “Texture Analysis with MTEX
– Free and Open Source Software Toolbox,” Solid State Phenom., 160, 63
(2010).