J. Am. Ceram. Soc.
, 1–10 (2014)
Journal
Ryan P. Haggerty,‡,§ Pankaj Sarin,‡,¶,* Zlatomir D. Apostolov,‡,k,* Patrick E. Driemeyer,‡,†† and
Waltraud M. Kriven‡,†,**
‡
Department of Materials Science and Engineering, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801
The thermal expansion of a low symmetry crystal can be much
more interesting than the lattice parameter expansion would
suggest. Here, the complete thermal expansion tensors for
monoclinic and tetragonal phases of ZrO2 and HfO2 have been
measured in air, by high-resolution, high-temperature X-ray
diffraction. These results reveal the highly anisotropic nature of
thermal expansion in the monoclinic phase as well as a cooper-
ative movement of ions and the existence of a zero thermal
expansion plane.
I. Introduction
(1) Motivation
T WO technical ceramics, zirconia (ZrO2) and hafnia
(HfO2), are known for their similar properties, and are
important for several advanced applications. ZrO2 is mostly
used for its structural properties. The toughening afforded by
its martensitic phase transformation has made it the material
of choice in many high-temperature structural applications.
Other common applications include fuel cell electrodes,1 ther-
mal barrier coatings,2,3 cutlery, and jewelry. HfO2 is often
thought of as a higher temperature substitute for ZrO2. This
is mostly due to the similarities in the chemical properties of
these ceramics. HfO2 has other uses such as in nuclear appli-
cations, due to its high neutron absorption. More recently, it
has found use as a high-k dielectric material and has replaced
SiO2 as the gate oxide in transistors.
With so many advanced applications, there has been a large
drive to understand the basic properties of these ceramics. Since
many of these applications are for use at high or extremely high
temperatures, and all of these applications require processing
at high temperatures, the thermal properties of these materials
are especially important. Unfortunately, most studies on these
materials have been done ex situ or in inert atmosphere. Many
of the applications for these ceramics are due to their stability
in air and it is desirable to measure the properties in the envi-
ronment in which the ceramic will be used.
One major goal of this study was to perform in situ studies
of ZrO2 and HfO2 at temperatures exceeding 1800°C in air,
to obtain the true thermal expansion of the material. This
would, hopefully, lead to a better understanding of the per-
formance of the material. Studies done in inert atmosphere
affect the properties measured. When heated in an inert
L. Gauckler—contributing editor
Manuscript No. 34505. Received February 3, 2014; approved April 4, 2014.
*Member, The American Ceramic Society.
**Fellow, The American Ceramic Society.
§
Present address: Sandia National Laboratory, Albuquerque, New Mexico
¶
Present address: School of Materials Science and Engineering, Oklahoma State
University, Tulsa, Oklahoma
k
Present address: Air Force Research Laboratory, Wright Patterson Air Force Base,
Dayton, Ohio
††
Present address: Alstom Power, Windsor, Connecticut
†
Author to whom correspondence should be addressed. e-mail: kriven@illinois.edu
1
DOI: 10.1111/jace.12975
© 2014 The American Ceramic Society
Thermal Expansion of HfO2 and ZrO2
atmosphere, increased concentration of oxygen vacancies
change the properties of the material.1,4–7
A common method for reporting the thermal expansion of
a material is to give the bulk value of thermal expansion.
Although this method is elegant in its simplicity, in crystals
with less than cubic symmetry, it does little to depict the
types of changes that the crystal incurs as the temperature
changes. A more thorough approach is to give the expansion
of the lattice parameters, but even this does not fully portray
the expansion of monoclinic and triclinic crystals. These low
symmetry crystals can have complex thermal expansions,
where the maximum and minimum values of expansion do
not necessarily lie along the lattice directions. ZrO2 and
HfO2 have monoclinic symmetries and are thus best
explained by more thorough methods. Furthermore, the ther-
mal expansion of these crystals can be related to their atomic
arrangements to understand these properties from a vantage
point not previously explored.
(2) Previous Work
There have been many studies on the thermal properties of
ZrO2 and HfO2 in the past. Curtis et al. carried out early in
situ X-ray diffraction studies on the high-temperature phases
of ZrO2 and HfO2.8 These experiments were performed using
an X-ray camera with CuKa wavelength, while the sample
was heated using tungsten elements.9 The thermal expansion
of ZrO2 and HfO2 was later measured by X-ray diffraction
in 1961 by Grain and Campbell.10 The first in situ, high-tem-
perature X-ray diffraction to be done in air on ZrO2 and
HfO2 was performed by Patil and Subbarao in 1969.11 This
work only reached 1400°C, far below the phase transforma-
tion from monoclinic to tetragonal in HfO2(~1700°C). A few
years later Stacy et al. conducted a thorough X-ray study of
the structure of HfO2 through the monoclinic to tetragonal
transformation, to get the thermal expansion of both phases
and the volume change at transformation. Unfortunately,
this work too was done in a vacuum furnace.12 The structure
of HfO2 was again studied through the transformation in
1987 by Coutures and Coutures, but these experiments were
done in air up to 1500°C, and then in a He-2%H atmosphere
to 2000°C.13
(A) Objectives: The major goal of this work was to
report the thermal expansion of ZrO2 and HfO2. These prop-
erties are obtained in situ, in air and are given as a function
of temperature and crystallographic direction. The properties
reported in this work are to a higher precision and a higher
temperature than have been obtained previously in air. In
addition, the importance of expressing thermal expansion in
its tensor form is addressed. The impact of the crystal struc-
ture of the low symmetry phases on the thermal expansion is
discussed along with the unique properties of these crystals.
(3) Crystal Structure of HfO2 and ZrO2 Phases
In 1959, Adam and Rogers14 were the first to suggest the simi-
larity between the crystal structures of monoclinic HfO2 and
2 Journal of the American Ceramic Society—Haggerty et al.
ZrO2.15 The cause for the parallelism between HfO2 and ZrO2
is the similarity in the Hf4+ and Zr4+ ions. Both elements are
in Group 4 on the Periodic Table, giving them similar valence
electrons: 4f145d26s2 for Hf and 4d25s2 for Zr. Due to the lan-
thanide contraction, the ionic size of Hf4+ and Zr4+ is very
similar. The Shannon–Prewitt ionic radii for sevenfold coordi-
nated Hf4+ and Zr4+ are 76 and 78 picometers, respectively,
and for eightfold coordinated ions are 83 and 84 picometers,
respectively.16 The similar electronic structure and atomic size,
in addition to similar electronegativities (1.23 and 1.22 for Hf
and Zr, respectively17) are indicative of the similar chemistries
and properties of these elements.
HfO2 and ZrO2 have the same polymorphs upon heating.
The room-temperature polymorph, known as the baddeleyite
structure, has a monoclinic space group, P21/c, and is stable
up to about 1170°C and 1750°C in ZrO2 and HfO2, respec-
tively. This structure is shown in Fig. 1, which depicts the
baddeleyite unit cell of HfO2. It is much easier to understand
this low symmetry structure by breaking it up into parts. In
this structure, the M4+ ions (Zr4+ or Hf4+) are in sevenfold
coordination with oxygen atoms, and the oxygen atoms are
either in three or fourfold coordination with M4+ ions.
Figure 1(d) shows the structure of the seven oxygens around
a M4+ ion. There are three oxygens labeled O1 bonded to
each M4+ ion, where these oxygens are each bonded to three
M4+ ions such that all four atoms are in a plane. The other
four oxygens labeled O2 in Fig. 1(d) are bonded to four
M4+ ions in a tetrahedral configuration. The sevenfold coor-
dinated M4+ ion polyhedra are oriented in such a way that
there are layers of O1 oxygens and O2 oxygens which are
perpendicular to the a-axis of the monoclinic unit cell. At
room temperature the HfO2 monoclinic unit cell is slightly
smaller than the ZrO2 radius.
At higher temperatures the oxides transform to take on the
tetragonally distorted fluorite structure with P42/nmc space
(a)
(c)
Fig. 1. The crystal structure of monoclinic baddeleyite for HfO2 at room temperature, given in (a), (b), and (c) as the unit cell projected along
the a-, b-, and c-axes, respectively. (d) A (Zr4+ or Hf4+) ion sevenfold coordinated with oxygen O1 and O2 sites.
group. The unit cell is shown in Fig. 2. This crystal structure
was first accurately described by Teufer in 1962 using high-
temperature X-ray diffraction.18 The highest temperature solid
phase of HfO2 and ZrO2 is the cubic fluorite structure with the
Fm3m space group. In the tetragonal and cubic phases, the
metal ions have eightfold oxygen atom coordination and oxy-
gen is fourfold coordinated with metal ions. There are few
studies conducted on pure ZrO2 or HfO2 due to the extremely
high temperatures, but due to the technological importance of
doped ZrO2 there is a plethora of information regarding the
properties of this phase when doped. In this work, the thermal
expansion of the monoclinic and tetragonal phases of pure
ZrO2 and pure HfO2 was investigated.
II. Theory
(1) Thermal Expansion of HfO2 and ZrO2
It is very common to describe the thermal expansion of a
material by calculating the expansion of the lattice parame-
ters of a crystal. For many systems this method is satisfac-
tory and describes the anisotropies in expansion that the
crystal exhibits. However, for systems of low symmetry (i.e.,
monoclinic or triclinic symmetry crystals) the thermal
expansion can be much more involved than just the expan-
sion of the lattice parameters can describe. The thermal
expansion tensor relates a change in temperature to the
strain of a crystal, Xi,j. In a three-dimensional system the
3 9 3 tensor has only six independent parameters because
of the inversion symmetry of thermal expansion, taking the
form:
0 1
a a1;2 a1;3
dXi;j @ 1;1
¼ a1;2 a2;2 a2;3 A (1)
dT a a2;3 a3;3
1;3
(b)
(d)
 Thermal Expansion of HfO2
(a)
Fig. 2. The unit cell of tetragonal HfO2, (a) a-axis projection, and (b) c-axis projection.
(A) Tetragonal and Monoclinic Phases: The Neumann
principle requires that “the symmetry of any physical prop-
erty of a crystal must include the symmetry elements of the
point group of that crystal”.19 Therefore, the thermal expan-
sion matrix possesses all the symmetry elements of the crys-
tal. For a monoclinic system (with a 6¼ b 6¼ c and b 6¼ 90°)
and for a tetragonal system (with a = b 6¼ c and b = 90°),
the respective thermal expansion tensors are be given by:
0 1 0 1
a1;1 0 a1;3 a1;1 0 0
amonoclinic ¼ @ 0 a2;2 0 A; atetragonal ¼ @ 0 a1;1 0 A
a1;3 0 a3;3 0 0 a3;3
In the tetragonal case, the a1,1 value will be equivalent to the
thermal expansion along the a and b lattice directions, so
that this value will be constant in any direction in the a–b
plane. Whereas the a3,3 value will be equal to the thermal
expansion along the c lattice direction, the larger of the a1,1
and a3,3 values will be equal to the largest thermal expansion
that the crystal exhibits in any direction. The smaller of the
values will equal the smallest thermal expansion that the
crystal exhibits.
It is not so straightforward in the monoclinic case. The
axis with the highest symmetry, the b lattice direction, will
have thermal expansion described by the a2,2 value. The a
and c lattice directions cannot be described by any single
value in the thermal expansion tensor. The largest and small-
est values of thermal expansion the crystal exhibits will be
equal to the largest and smallest eigenvalues of the thermal
expansion tensor. Although one of the eigenvectors of the
tensor must be along the b lattice direction, the other eigen-
vectors are only constrained to two mutually perpendicular
directions in the a–c plane.
(B) Temperature Dependence of Thermal Expansion: In
addition, it is important to recognize the temperature depen-
dence of thermal expansion when describing the properties of
a crystal over a very large temperature range. A varying-order
polynomial in powers of temperature can be used to fit thermal
expansion coefficients:20
1 ddhkl
ahkl ¼ ¼ A0;hkl þ A1;hkl T þ A2;hkl T2 þ . . .
dhkl dT
To extract the polynomial using high-temperature X-ray dif-
fraction data Eq. (3) can be rearranged and integrated with
respect to temperature over a range from T0 to T resulting in:
3
(b)
ln dhkl;T  ln dhkl;T0 ¼ A0;hkl ðT  T0 Þ þ A1;hkl ðT  T0 Þ2
þ A2;hkl ðT  T0 Þ3 þ . . .
(4)
In the above equation, dhkl,T is the experimentally measured
d-spacing of hkl plane at temperature T, and dhkl,T0 is the
experimentally measured d-spacing of hkl plane at T0, the
chosen reference temperature. The values for the coefficients
A0,hkl, A1,hkl, A2,hkl etc. are obtained by least-squares fitting
of the above equation. They can subsequently be used to cal-
culate ahkl using Eq. (3). The procedure is outlined in detail
by Jones et al. (2013).21
(2)
III. High-Temperature X-ray Diffraction
(1) Instrumentation
To obtain any crystallographic information for HfO2 in
the tetragonal phase, advanced instrumentation was needed
and some novel procedures were used. The major issues
were the extremely high temperatures (>1700°C), the high
X-ray and neutron absorption of Hf and sensitivity of
ZrO2 and HfO2 to oxygen partial pressure. To overcome
these challenges, significant amounts of planning and test-
ing of new experimental techniques were required. In addi-
tion, several advanced pieces of equipment were designed,
purchased, and developed. A quadrupole lamp furnace was
chosen for its ability to reach the melting point of alu-
mina, that is, 2050  4°C in air.22–27 A curved image plate
(CIP) detector was used to measure the intensity of the
diffracted X-rays. The CIP detector could obtain high-reso-
lution powder diffraction patterns in less than a minute.28
The Advanced Photon Source at Argonne National Labo-
ratory, beamline 33 BM was chosen as an X-ray source
for its particularly good optics. In addition, the bending
magnet beamline allowed for variable X-ray energy beams
to accommodate a variety of samples that had X-ray
absorption edges at common wavelengths. The CIP detec-
tor has been used at energies ranging from 14 to 22.5 keV.
The X-ray optics of 33-BM are described by Zschack.29
Secondary experiments were conducted at the National
Synchrotron Light Source at Brookhaven National Labora-
(3) tory, beamline X14A using a silicon strip detector. Com-
bining these instruments has allowed for unique
experimental instrumentation capable of observing the crys-
tal properties of ZrO2 and HfO2 at high resolution even at
temperatures above the transformation temperature of
HfO2 (~1750°C) to the tetragonal phase, in air.
4 Journal of the American Ceramic Society—Haggerty et al.

900°C (2) Diffraction Experiments

The starting chemicals for ZrO2 and HfO2 were acquired
from Sigma-Aldrich (Milwaukee, WI), 99.95% pure on a

Cooling
Intensity(Counts)

metals basis. Each powder sample was mixed with Pt powder

and placed into a sapphire capillary, produced by Saint-Gob-
ain (Valley Forge, PA). These capillaries were small enough
1822°C (inner diameter of 600  100 lm) to allow for transmission
of X-rays through lightly packed HfO2 and could withstand

8 9 10 11
Fig. 3. High-temperature X-ray diffraction of HfO2 mounted in a
sapphire capillary in air. From bottom to top, the patterns displayed
were collected from about 1000°C up to 1822°C and back down to
about 900°C, although the temperature step size is not constant.
Intensity(Counts)
Heating
temperatures of nearly 2000°C. Several experimental runs
were conducted by heating ZrO2 and HfO2 separately, to a
Monoclinic-HIO2 1000°C series of set temperatures and taking 1 to 3 X-ray diffraction
patterns at each temperature. The powders were packed into
HIO2-Tetragonal the capillaries by thin alumina tubes that were fixed in place
with yttria-stabilized zirconia paste around the outside. The
12 13 14 15 16 17 18 19 20 21
Two- Theta (deg)
hollow tubes allowed the sample access to air throughout the
experiment. The experiments were all conducted using X-ray
energies between 17 and 23 keV, which provided sufficient
transmission through the sample, while still giving excellent
d-spacing range and resolution. Diffraction from the sapphire
capillary was limited, but still required some small sections
of the diffraction pattern to be removed. The temperature of
the sample was determined from the diffraction of platinum,
which was mixed into the sample at ~10 wt%. Once the sam-
ple surpassed the melting temperature of platinum (only in

the case of HfO2), the peaks from the sapphire capillaries

25°C were used to determine the sample temperature. The rela-
tively large wall thickness of the sapphire capillaries caused
Cooling

the diffraction peaks from the sapphire capillary to cover a

1530°C
larger 2h range than that of the platinum peaks. This, com-
bined with the lower thermal expansion of sapphire com-
Heating

pared to platinum, made the error in temperature

25°C determination greater, once the platinum had melted or evap-
Tetragonal - ZrO2 orated.
Monoclinic - ZrO2

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
(3) Data Analysis
Two- Theta (deg) The X-ray data acquired were analyzed using structural
Fig. 4. High-temperature X-ray diffraction of ZrO2 mounted in a refinement software, Jade (Materials Data Incorporated, Liv-
sapphire capillary in air. From bottom to top, the patterns displayed ermore, CA). The crystallographic data were extracted using
were collected from room temperature up to about 1530°C and back the Rietveld30 refinement or Pawley’s methods,31 to fit a
down to room temperature, although the temperature step size was structure to the diffraction pattern while taking into account
not constant. the geometry and temperature effects.

(a) (b)

Fig. 5. Lattice parameters of the monoclinic phase of ZrO2 and HfO2 as a function of temperature in air (a) ZrO2 (b) HfO2 (error is smaller
than the symbols used).
 Thermal Expansion of HfO2 5

(a) 5.35 20
a,HfO2 a,ZrO2

a X 21/2

CTE - Crystallographic Axes (1 x 10-6 K-1)

b,HfO2 b,ZrO2
c
5.30 15 c,HfO2 c,ZrO2
Lattice Parameter (Å)

5.25
10

5.20

5.15

0
5.10
800 900 1000 1100 1200 1300 1400 1500 1600

Temperature (°C) -5
0 200 400 600 800 1000 1200 1400 1600 1800
(b) 5.35
Temperature (°C)
a X 21/2
c Fig. 8. Thermal expansion of the lattice directions in monoclinic
5.30
phases of ZrO2 and HfO2 as a function of temperature (error is
Lattice Parameter (Å)

smaller than the symbols used).

5.25

Table I. The Thermal Expansion Along Each Lattice

5.20 Direction for Monoclinic ZrO2 is Provided, as Well as the
Eigenvalues of Thermal Expansion

5.15 Thermal Expansion (9106 1/K)

Temp °C a b (EV3) c EV1 EV2

5.10
1350 1400 1450 1500 1550 1600 1650 1700 1750 1800 1850 50 2.89 (9) 5.57 (7) 9.06 (9) 10.7 (1) 4.71 (5)
Temperature (°C) 100 5.94 (6) 3.04 (7) 9.20 (7) 10.7 (1) 5.14 (4)
150 6.29 (5) 3.16 (6) 9.35 (6) 10.79 (9) 5.53 (3)
Fig. 6. Lattice parameters of the tetragonal phase of ZrO2 (a) and 200 6.62 (4) 3.25 (4) 9.51 (4) 10.91 (7) 5.87 (2)
HfO2 (b) as a function of temperature in air.
250 6.92 (3) 3.31 (3) 9.69 (3) 11.08 (5) 6.17 (2)
300 7.20 (2) 3.34 (2) 9.87 (2) 11.30 (4) 6.42 (1)
350 7.46 (1) 3.34 (2) 10.07 (2) 11.57 (3) 6.62 (1)
35
eig1,HfO
2
400 7.69 (2) 3.31 (2) 10.28 (2) 11.88 (3) 6.77 (1)
eig2,HfO 450 7.90 (2) 3.25 (2) 10.50 (2) 12.25 (4) 6.88 (1)
30
eig3,HfO
2
500 8.08 (2) 3.16 (3) 10.73 (3) 12.67 (4) 6.94 (1)
2
550 8.24 (2) 3.04 (3) 10.98 (3) 13.14 (5) 6.95 (1)
25 eig1,ZrO
CTE Eigenvalues (1 x 10-6 K-1)

2 600 8.37 (2) 2.89 (3) 11.24 (3) 13.66 (5) 6.91 (1)
eig2,ZrO 650 8.49 (2) 2.70 (3) 11.52 (3) 14.23 (5) 6.82 (1)
2
20
eig3,ZrO 700 8.58 (2) 2.49 (3) 11.81 (3) 14.86 (5) 6.69 (1)
2

750 8.64 (2) 2.25 (3) 12.11 (3) 15.53 (5) 6.51 (1)
15
800 8.69 (2) 1.98 (3) 12.42 (3) 16.26 (5) 6.28 (1)
850 8.71 (2) 1.67 (3) 12.76 (3) 17.03 (5) 6.01 (1)
10
900 8.70 (2) 1.34 (3) 13.10 (3) 17.86 (4) 5.69 (1)
950 8.68 (2) 0.97 (2) 13.46 (2) 18.75 (4) 5.32 (1)
5
1000 8.63 (2) 0.58 (3) 13.84 (2) 19.68 (4) 4.90 (1)
1050 8.57 (2) 0.15 (3) 14.24 (3) 20.68 (5) 4.44 (1)
0
1100 8.50 (4) 0.30 (5) 14.63 (5) 21.75 (8) 3.91 (2)
1150 8.47 (6) 0.75 (7) 14.91 (7) 22.8 (1) 3.33 (3)
-5
0 200 400 600 800 1000 1200 1400 1600 1800 1200 8.33 (7) 1.33 (9) 15.51 (9) 24.1 (1) 2.74 (4)
1250 8.3 (1) 1.8 (1) 15.8 (1) 25.5 (2) 1.94 (6)
Temperature (°C)

Fig. 7. Eigenvalues of the thermal expansion tensor calculated for

monoclinic phases of ZrO2 and HfO2 as a function of temperature a single temperature. Using a least-squares method and
(error is smaller than the symbols used). taking the symmetry of the crystal into account, the thermal
expansion tensor for that temperature is fitted. By repeating
this for a series of relevant temperatures a progression of the
An algorithm has been developed, called the Coefficient of thermal expansion tensor is produced.
Thermal Expansion Analysis Suite, to describe thermal
expansion as a function of crystallographic direction and of
IV. Results
temperature.21 This process involves first determining the
accurate temperature at which each refined pattern was (1) X-ray Diffraction
acquired. The d-spacing of each diffraction peak is associated A sample of the high-temperature diffraction data acquired
with the direction of that peak’s hkl pole. The temperature- for ZrO2 and HfO2 is shown in Figs. 3 and 4, respectively.
dependent thermal expansion is then determined along the The figures show a reduced range of 2h for a clear presenta-
direction of each hkl input, independently, by fitting the tion of the peaks, however, usable X-ray data were collected
expansion along that hkl to Eq. (3). Then, the thermal from 3° to 35° 2h for all patterns. As can be seen in Fig. 3 at
expansions of all directions determined are aggregated for around 14°, the sapphire peak was observed, due to the low
6 Journal of the American Ceramic Society—Haggerty et al.

Table II. The Thermal Expansion Along Each Lattice always an eigenvector of thermal expansion. In the case of
Direction for Monoclinic HfO2 is Provided, as Well as the ZrO2 and HfO2, the b-axis is always the direction of mini-
Eigenvalues of Thermal Expansion mum thermal expansion. The minimum thermal expansion of
ZrO2 is between 4 9 106 and 2 9 106 1/K. In the case
Thermal Expansion (9106 1/K) of HfO2 the minimum thermal expansion is lower than in
Temp °
C a b (EV3) c EV1 EV2
ZrO2, in the range of 2 9 106 to 4 9 106 1/K. The other
two eigenvectors of the thermal expansion tensor lie perpen-
50 3.7 (4) 0.8 (6) 7.5 (6) 9.0 (8) 3.0 (3) dicular to the b-axis in the a–c plane. These two largest
150 4.8 (4) 1.2 (4) 8.6 (4) 9.9 (7) 4.2 (3) eigenvectors, represent the minimum and maximum thermal
250 5.8 (3) 1.5 (3) 9.5 (3) 10.8 (5) 5.2 (2) expansion within the a–c plane. The largest thermal expan-
350 6.7 (2) 1.8 (3) 10.4 (3) 11.8 (4) 6.0 (2) sion in ZrO2 ranges from 11 9 106 to 25 9 106 1/K as it
450 7.5 (2) 1.9 (2) 11.3 (2) 12.8 (3) 6.7 (1) completes its transformation to the tetragonal phase. HfO2
550 8.2 (1) 2.0 (2) 12.1 (2) 13.9 (3) 7.25 (9) has an even larger anisotropy at transformation, its maxi-
650 8.8 (1) 2.0 (2) 12.8 (2) 15.1 (2) 7.57 (8) mum thermal expansion ranging from 8 9 106 1/K near
750 9.3 (1) 1.8 (2) 13.5 (2) 16.3 (2) 7.70 (8) room temperature to about 32 9 106 1/K at 1750°C during
850 9.7 (1) 1.6 (2) 14.1 (2) 17.6 (3) 7.63 (8) transformation. The minimum thermal expansion in the a–c
950 10.0 (1) 1.4 (2) 14.7 (2) 19.0 (3) 7.37 (7) plane for ZrO2 and HfO2 is 2 9 106 to 7 9 106 1/K
1050 10.2 (1) 1.0 (2) 15.3 (1) 20.5 (2) 6.92 (7) and 1 9 106 to 8 9 106 1/K, respectively. These eigen-
1150 10.1 (1) 0.1 (2) 16.4 (2) 22.3 (3) 6.46 (7) values do not match the thermal expansion along the a and c
1250 10.2 (1) 0.0 (1) 16.1 (1) 23.6 (2) 5.40 (5) directions of the unit cell. The deviation of the extreme ther-
1350 9.88 (9) 0.7 (1) 16.8 (1) 25.4 (2) 4.28 (5) mal expansions in the a–c plane from the expansion along
1450 10.09 (4) 1.35 (5) 16.95 (5) 27.15 (7) 3.25 (2) the lattice directions can be seen in Fig. 8. This difference
1550 9.4 (1) 2.1 (1) 17.2 (1) 28.4 (2) 1.90 (5) illustrates the importance of examining the properties of the
1650 9.4 (1) 3.0 (1) 17.6 (1) 30.9 (2) 0.27 (5) thermal expansion tensor rather than the changes in the lat-
1750 9.1 (2) 4.0 (2) 17.2 (2) 32.2 (3) 1.31 (9) tice parameters. A full list of the calculated thermal expan-
sion is given in Tables I and II, for ZrO2 and HfO2,
respectively.
A useful way to gain qualitative understanding of the
amount of scattering from HfO2. This portion of the pattern thermal expansion as a function of crystallographic direc-
was removed from the structural refinement for hafnia. tion is to plot the property ellipsoid in spherical coordi-
nates. Figure 9 shows the thermal expansion ellipsoid
surface for a few different temperatures of both ZrO2 and
(2) Lattice Parameters HfO2. By convention, the orthogonal x–y–z axis system is
(A) Monoclinic Phase: The lattice parameters for the chosen such that the z-axis is parallel to the c-axis of the
monoclinic phase of ZrO2 and HfO2 are given in Fig. 5 as a unit cell, the y-axis is parallel to the b-axis and x is perpen-
function of temperature. Even though the unit cell lengths of dicular to the b–c plane. The anisotropy in both oxides is
the monoclinic phase are distinct in each direction, the differ- clearly depicted by the shape of the ellipsoid. The b-axis is
 The gen-
ence is relatively small, ranging from 5.1 to 5.4 A. constrained to be an extreme of thermal expansion due to
eral trends in the lattice parameters of ZrO2 are similar to symmetry, but the ellipsoid is free to rotate around the
that of HfO2. As expected, the lattice parameters of ZrO2 are b-axis. For this reason, it is useful to view the a–c plane
larger than for HfO2 at room temperature. In both ceramics, (the x–z plane in orthogonal space). This plane is shown
the largest changes can be seen in the c-axis, followed by the for several temperatures in Fig. 10. The direction of the
a-axis. The b-axis changes the least with temperature. The eigenvectors in this plane can be related to the crystal struc-
b-axis of ZrO2 changes very little within a few hundred ture of the ceramic.
degrees of the transformation to tetragonal, while in HfO2 The direction of the eigenvectors relative to the crystal
the b-axis decreases in length above approximately 1000°C. structure of the ceramic can provide some interesting
While the b angles of the room-temperature ZrO2 and HfO2 insights. In particular, as the ceramic expands, all the bonds
are almost equivalent, the change in b from room to trans- along the eigenvectors remain in line with those eigenvectors.
formation temperature is much larger in HfO2. All other bonds turn away from the eigenvector correspond-
(B) Tetragonal Phase: The change in lattice parame- ing to the minimum direction of thermal expansion, toward
ters in the tetragonal phase is relatively simple when com- the eigenvector for maximum thermal expansion. For ZrO2,
pared to that for the monoclinic phase. The lattice the eigenvector corresponding to the maximum value of ther-
parameters for the tetragonal phase of ZrO2 and HfO2 are mal expansion is between 31° and 37° from the c-axis,
given in Fig. 6 as a function of temperature. The a-axis lat- toward the a-axis at temperatures between room temperature
tice parameter has been multiplied by √2 to make the tetrag- and 700°C. Above 700°C, the eigenvector is between 37° and
onal phase equivalent to the monoclinic phase in number of 40° from the c-axis. In crystallographic terms this is approxi-
formula units per unit cell, that is, Z = 4. Once adjusted, the mately, moving from the direction perpendicular to the (1 0
lattice parameters of the tetragonal phase are in a similar 2) plane pole at low temperatures, to just past the perpendic-
range to that of the monoclinic cell. ular to the (3 0 4) plane pole. In HfO2, the eigenvector corre-
sponding to the largest eigenvalue is in approximately at the
V. Discussion same location, tilted by ~29° from the c-axis at room temper-
ature. As the temperature rises, the angle between the c-axis
(1) Monoclinic Phase and the eigenvector increases, to about 42° near the start of
(A) Thermal Expansion Tensor: The thermal expan- the transformation to the tetragonal phase (~1750°C), again,
sion calculated for the monoclinic phase is highly anisotropic rotating from approximately perpendicular to the (1 0 2)
for both ZrO2 and HfO2, as can be seen by examining the plane pole to the (3 0 4) plane pole.
eigenvalues of the thermal expansion matrix in Fig. 7. The (B) Zero Thermal Expansion Plane: The eigenvalues
eigenvalues of the property tensor represent rate of expan- represent only the values of extreme thermal expansion, but
sion in the directions of the extreme values. In both ceramics, from their values much more interesting phenomena can be
the anisotropy of thermal expansion, i.e., the ratio of maxi- predicted. To have a direction of zero thermal expansion,
mum to minimum eigenvalue, increases with temperature. either one of the eigenvalues of thermal expansion must be
Due to the symmetry of the monoclinic crystal, the b-axis is zero, or there must be a mix of positive and negative values
 Thermal Expansion of HfO2 7

(a) (b)

(c) (d)

Fig. 9. Spherical diagram displaying the thermal expansion in all directions of the monoclinic phase crystal of ZrO2 in (a) at 250°C and in (b)
at 1000°C; and HfO2 in (c) at 100°C and in (d) at 1600°C. By convention the c-axis is parallel to the Z direction and the b-axis is parallel to the
Y direction, the X direction is perpendicular to the b–c plane.

of thermal expansion. A plane of zero thermal expansion
is rare. This requires that either two of the eigenvalues of
thermal expansion are zero, or if only one eigenvalue is zero,
the remaining two must be mixed, one positive and one
negative. The latter case occurs in HfO2, at ~1670°C. The
plane of zero thermal expansion contains both the eigenvec-
tor corresponding to an eigenvalue of zero and the direction
of zero thermal expansion in the plane of the other two ei-
genvectors. In HfO2 this is approximately the (3 1 4) plane.
However, with eigenvalues constantly changing with temper-
ature, the plane of zero thermal expansion is only shown to
exist for a single temperature. This does not take away from
the uniqueness of this property, as the thermal expansion of
this plane remains small in all directions over a range of tem-
peratures. Between 1500°C and the transformation tempera-
ture, the thermal expansion in the (3 1 4) plane has an
absolute value of less than 1 9 106 1/K.
A plane of zero thermal expansion has a wide range of
potential applications, particularly in thermal-shock-resistant
materials. The lack of materials with a plane of zero thermal
expansion has left the potential uses unexplored. The engi-
neering of a ceramic with such properties could lead to more
applications. Since the thermal expansion of ZrO2 near trans-
formation also has a low magnitude in the (3 1 4) plane, it
is likely that a solid solution of ZrO2 and HfO2 will also
have this property. It is likely that by adjusting the
composition of a ZrO2–HfO2 solid solution, the temperature
range over which a plane of near zero thermal expansion
exists can be tailored. It is not necessary to make a single
crystal of the ceramic to gain the benefits of a plane of zero
thermal expansion. A polycrystalline ceramic with grains ori-
ented such that the (3 1 4) plane is parallel to a substrate
can form a plane of low thermal expansion while still having
minimal thermal stress between grains.
(C) Relation to Crystal Structure: The thermal expan-
sion is a continuum property that relates the ways that the
different bonds in the crystal accommodate each other’s
expansions. The crystal structure of these ceramics is a
source of information about the cause of the large anisotropy
of thermal expansion. The crystal structure of monoclinic
ZrO2 and HfO2, viz., baddeleyite is introduced in Section I
(3), but warrants more examination.
The sevenfold coordinated M4+ ions (Zr4+ or Hf4+) have a
rather uncommon shape. They share edges with surrounding
8 Journal of the American Ceramic Society—Haggerty et al.

(a) a,HfO2
22

CTE - Crystallographic Axes (1 x 10-6 K-1)

c,HfO2

20
a,ZrO2

c,ZrO2
18

16

14

12

10

800 1000 1200 1400 1600 1800

Temperature (°C)
(b)
Fig. 12. Thermal expansion of the lattice directions in tetragonal
phases of ZrO2 and HfO2 as a function of temperature (error is
smaller than the symbols used).

Table III. The Thermal Expansion Along Each Lattice

Direction for Tetragonal ZrO2
Thermal Expansion (9106 1/K)

Temp (°C) a (EV1) c (EV2)

900 10.9 (1) 16.8 (1)

950 10.86 (6) 16.37 (6)
1000 10.87 (6) 15.97 (6)
1050 10.90 (4) 15.60 (4)
1100 10.96 (4) 15.27 (4)
Fig. 10. Polar diagram of thermal expansion in a–c plane of
1150 11.04 (7) 14.98 (7)
monoclinic (a) ZrO2 from 50°C (blue) to 1200°C (red) (b) HfO2 from
50°C (blue) to 1750°C (red). 1200 11.2 (1) 14.7 (1)
1250 11.3 (1) 14.5 (1)
1300 11.5 (1) 14.3 (1)
1350 11.7 (1) 14.2 (1)
(a) 1400 11.9 (1) 14.2 (1)
1450 12.1 (1) 14.2 (1)
1500 12.4 (1) 14.2 (1)

way to better understand the influence of the bonds on the

(b)
Fig. 11. The monoclinic crystal structure of the O1 polyhedra with
the O2 sites removed (a) a-axis projection, (b) normal to (3 0 4) plane.
polyhedra, which is not expected of highly anisotropic thermal
expansion materials that often have buckling of polyhedral
chains. The crystal structure needs to be viewed in another
thermal expansion. By simplifying the structure in terms of the
different type of oxygen sites, the interactions of the polyhedra
can be much more easily understood. First, consider the poly-
hedra of M4+ ions coordinated around O2 ions. For the O2
oxygen sites, these polyhedra consisting of one O2 site sur-
rounded by four M4+ ions, which form tetrahedra that are all
edge sharing with four other like tetrahedra. They are aligned
in an isolated layer perpendicular to the a-axis. The O1 sites
are threefold coordinated, with each bonded M4+ ion in a sin-
gle plane. These polyhedra are edge sharing with one other
similar polyhedron. Two edge sharing O1 polyhedra line up in
a single plane and can be thought of as a single polyhedron
containing 4M4+ ions arranged in a planar parallelogram sur-
rounding two O1 sites. This parallelogram polyhedron is cor-
ner shared with four other like polyhedra, interlocking to form
a continuous layer between the O2 polyhedral layers. Fig-
ure 11 depicts the room-temperature structure of the O1 poly-
hedron with the O2 atoms removed. The corner sharing of the
parallelogram polyhedra indicated that the O1 layers dictate
the highly anisotropic thermal expansion. The plane of each
polyhedron lines up almost perpendicular to its neighbors,
forming an accordion-like structure that allows the polyhedra
to rotate slightly and collapse in the b-axis direction. The ori-
entation of the parallelogram polyhedra supports the theory
 Thermal Expansion of HfO2 9

Table IV. The Thermal Expansion Along Each Lattice (2) Tetragonal Phase
Direction for Tetragonal HfO2 (A) Thermal Expansion Tensor: Due to increased
symmetry, the thermal expansion tensor that describes the
Thermal Expansion (9106 1/K) tetragonal phase is much less complex. The eigenvectors
Temp (°C) a (EV1) c (EV2) always lie along the a-, b-, and c-axes. The thermal expan-
sion in the a–b plane is constant in all directions. To show
1350 8.4 (2) 12.5 (2) the maximum and minimum values of thermal expansion,
1400 9.3 (1) 12.8 (1) the expansion along the a and c lattice directions is given as
1450 10.00 (7) 13.16 (7) a function of temperature in Fig. 12. The values are listed in
1500 10.55 (4) 13.51 (4) Tables III and IV for tetragonal ZrO2 and HfO2, respec-
1550 10.94 (5) 13.89 (5) tively. To describe the thermal expansion fully, it is sufficient
1600 11.17 (6) 14.31 (6) to provide the expansion in the a–c plane. All other planes
1650 11.24 (5) 14.76 (5) containing the c-axis will be identical to the a–c plane, shown
1700 11.16 (3) 15.24 (3) in Fig. 13.
1750 10.92 (5) 15.76 (5) The tetragonal phase thermal expansion has similar values
1800 10.5 (1) 16.3 (1) in both ZrO2 and HfO2, as was the case in the monoclinic
1850 10.0 (2) 16.9 (2) phase. The largest expansion is in the c-axis, between
13 9 106 1/K. Although the values are similar, the trend with
temperature is different. The thermal expansion along the c-
axis decreases with temperature for ZrO2, but it increases with
temperature for HfO2. The minimum thermal expansion, in
(a) the a-axis direction, is lower in HfO2 (8.5 9 106 1/K) than in
ZrO2 (11 9 106 to 12.5 9 106 1/K).

(b)
Fig. 13. Polar diagram of thermal expansion in a–c plane of
tetragonal (a) ZrO2 from 900°C (blue) to 1500°C (red) (b) HfO2
from 1350°C (blue) to 1850°C (red).
that these polyhedra are related to the odd thermal expansion
of ZrO2 and HfO2. The directions across the parallelogram,
from M4+ ions at opposite corners, approximately line up with
the eigenvectors of thermal expansion when projected into the
a–c plane. The shorter distance across the parallelogram lines
up with the direction of largest thermal expansion. It rotates
from about 30° angle with the c-axis in both ceramics and
rotates slightly toward the a-axis as it is heated. This is the
same behavior as is observed by the direction of the
eigenvector.
VI. Conclusions
Accurately determining the thermal expansion tensor of
ZrO2 and HfO2 allows the highly anisotropic nature of these
materials to be understood. Using high-resolution, high-tem-
perature X-ray diffraction the thermal expansion of the
monoclinic and tetragonal phases of these ceramics has been
determined. This is the first presentation of the thermal
expansion of tetragonal HfO2 in air.
In addition to the characterization of two technologically
important ceramics, this work has shown the significance of
describing the thermal expansion of low symmetry materials
using the tensor form. The thermal expansion of the crystals
in this study is largely dependent on the direction that it is
measured in. The direction of least thermal expansion was,
as expected, found to be in the b-axis. The maximum thermal
expansion of ZrO2 and HfO2 was found to be in the a–c
plane rotating from the (1 0 2) plane normal to the (3 0 4)
plane normal as temperature rises. The anisotropy in thermal
expansion reaches a maximum as the monoclinic phase
reaches the transformation temperature to the tetragonal
phase in both ceramics. By examining the directionally
dependent thermal expansion thoroughly, a plane of zero
thermal expansion has been found in HfO2 at high tempera-
tures. This plane is approximately normal to the (3 1 4)
plane of HfO2. The same plane of ZrO2 was found to have
very low thermal expansion, indicating that the solid solution
of ZrO2 and HfO2 can be exploited to engineer a ceramic
with a near zero thermal expansion plane in a desired tem-
perature range. The monoclinic crystal structure of the low-
temperature phase has been discussed and related to the
highly anisotropic thermal expansion.
Acknowledgments
This work was supported by the National Science Foundation Division of
Materials Research (NSF DMR) under grant no. 0706606. The work of P. E.
Driemeyer was supported by the AFOSR under grant number FA9550-09-1-
0322. Z.D. Apostolov was supported by an AFOSR SMART Scholarship by
the Department of Defense, USA. The Advanced Photon Source (APS), ANL
was supported by the US DOE, BES-Materials Sciences, under contract no:
W-31-109-ENG-38. The National Synchrotron Light Source (NSLS), BNL
was supported by the US DOE, Division of Materials Sciences and Division
of Chemical Sciences, under contract no: DE-AC02-98CH10886. The research
was carried out in part in the Frederick Seitz Materials Research Laboratory
Central Facilities, University of Illinois, which were partially supported by the
U.S. Department of Energy under grants DE-FG02-07ER46453 and DE-
FG02-07ER46471. The technical assistance of beam line scientists Dr. Paul
Zschack and Evgenia Karapetrova at the APS, ANL and Dr. Jian Ming Bai
at the NSLS, BNL are gratefully acknowledged.
10 Journal of the American Ceramic Society—Haggerty et al.
Supporting Information
Additional Supporting Information may be found in the
online version of this article:
Table S1. Lattice parameters of monoclinic HfO2 phase as
determined from in situ high-temperature XRD studies.
Table S2. Lattice parameters of tetragonal HfO2 phase as
determined from in situ high-temperature XRD studies.
Table S3. Lattice parameters of monoclinic ZrO2 phase as
determined from in situ high-temperature XRD studies.
Table S4. Lattice parameters of tetragonal ZrO2 phase as
determined from in situ high-temperature XRD studies.
References
1 Z. A. Jones, P. Sarin, R. P. Haggerty, and W. M. Kriven, “CTEAS: A
A. Weyl and D. Janke, “High-Temperature Ionic Conduction in Multicom-
ponent Solid Oxide Solutions Based on HfO2,” J. Am. Ceram. Soc., 79 [8]
2145–55 (1996).
2
H. Ibegazene, S. Alperine, and C. Diot, “Yttria-Stabilized Hafnia-Zirconia
Thermal Barrier Coatings: The Influence of Hafnia Addition on TBC Struc-
ture and High-Temperature Behaviour,” J. Mater. Sci., 30 [4] 938–51 (1995).
3
K. Matsumoto, I. Yoshiyasu, and Y. Ishiwata, “Thermal Conductivity and
Sintering Behavior of Hafnia-Based Thermal Barrier Coating Using EB-PVD”
in International Gas Turbine Conference Proceedings, 8th International Gas
Turbine Congress, Tokyo, Japan, 2003.
4
A. S. Foster, F. Lopez Gejo, A. L. Shluger, and R. M. Nieminen,
“Vacancy and Interstitial Defects in Hafnia,” Phys. Rev. B, 65 [17] 174117,
13pp (2002).
5
R. P. Haggerty and R. Seshadri, “Oxygen Stoichiometry, Crystal Structure,
and Magnetism of La0.5Sr0.5CoO3,” J. Phys.: Condens. Matter, 16 [36] 6477,
11pp (2004).
6
W. M. Kriven, J. W. Palko, S. Sinogeikin, J. D. Bass, A. Sayir, G. Bru-
nauer, H. Boysen, F. Frey, and J. Schneider, “High Temperature Single Crys-
tal Properties of Mullite,” J. Eur. Ceram. Soc., 19 [13–14] 2529–41 (1999).
7
C.-H. Lu, J. M. Raitano, S. Khalid, L. Zhang, and S.-W. Chan, “Cubic
Phase Stabilization in Nanoparticles of Hafnia-Zirconia Oxides: Particle-Size
and Annealing Environment Effects,” J. Appl. Phys., 103 [12] 124303 (2008).
8
C. E. Curtis, L. M. Doney, and J. R. Johnson, “Some Properties of
Hafnium Oxide, Hafnium Silicate, Calcium Hafnate, and Hafnium Carbide,”
J. Am. Ceram. Soc., 37 [10] 458–65 (1954).
9
J. R. Johnson, “A High-Temperature X-ray Camera,” J. Am. Ceram. Soc.,
37 [8] 360–2 (1954).
10
C. F. Grain and W. J. Campbell, “Thermal Expansion and Phase Inver-
sion of Six Refractory Oxides”; BM-RI-5982, U. S. Bureau of Mines Report
of Investigations, August 1961.
11
R. N. Patil and E. C. Subbarao, “Axial Thermal Expansion of ZrO2 and
HfO2 in the Range Room Temperature to 1400°C,” J. Appl. Crystallogr., 2 [6]
281–8 (1969).
12
D. W. Stacy, J. K. Johnson, and D. R. Wilder, “Axial Thermal Expansion
of HfO2,” J. Am. Ceram. Soc., 55 [9] 482–3 (1972).
13
J. P. Coutures and J. Coutures, “The System HfO2-TiO2,” J. Am. Ceram.
Soc., 70 [6] 383–7 (1987).
14
J. Adam and M. D. Rogers, “The Crystal Structure of ZrO2 and HfO2,”
Acta Crystallogr. A, 12 [11] 951 (1959).
15
R. Ruh and P. W. R. Corfield, “Crystal Structure of Monoclinic Hafnia
and Comparison with Monoclinic Zirconia,” J. Am. Ceram. Soc., 53 [3] 126–9
(1970).
16
R. D. Shannon, “Revised Effective Ionic Radii and Systematic Studies of
Interatomic Distances in Halides and Chalcogenides,” Acta Crystallogr. Sec.,
32 [5] 751–67 (1976).
17
E. J. Little and M. M. Jones, “A Complete Table of Electronegativities,”
J. Chem. Educ., 37 [5] 231–2 (1960).
18
G. Teufer, “The Crystal Structure of Tetragonal ZrO2,” Acta Crystallogr.,
15 [11] 1187 (1962).
19
R. E. Newnham, Properties of Materials: Anisotropy, Symmetry, Structure.
Oxford University Press, Inc., New York, NY 2003.
20
H. K€ uppers, “Thermal Expansion”; pp. 99–104 in Physical Properties of
Crystals, Vol. D. International Tables of Crystallography, Edited by A. Authier.
John Wiley & Sons, Ltd, Dordrecht, 2006. [Published Online]
21
Graphical-User-Interface-Based Program to Determine Thermal Expansion
from High-Temperature X-ray Diffraction,” J. Appl. Crystallogr., 46 [2] 550–3
(2013).
22
W. M. Kriven, P. Sarin, and L. F. Siah, “Phase Transformations in Rare
Earth Niobates,” Solid-Solid Phase Transform. Inorg. Mater., 2, 1015–22
(2005).
23
P. Sarin, P. E. Driemeyer, R. P. Haggerty, D.-K. Kim, J. L. Bell, Z. D.
Apostolov, and W. M. Kriven, “In Situ Studies of Oxidation of ZrB2 and
ZrB2-SiC Composites at High Temperatures,” J. Eur. Ceram. Soc., 30 [11]
2375–86 (2010).
24
P. Sarin, W. Yoon, R. P. Haggerty, C. Chiritescu, N. C. Bhorkar, and W.
M. Kriven, “Effect of Transition-Metal-Ion Doping on High Temperature
Thermal Expansion of 3:2 Mullite-An In Situ, High Temperature, Synchrotron
Diffraction Study,” J. Eur. Ceram. Soc., 28 [2] 353–65 (2008).
25
P. Sarin, W. Yoon, K. Jurkschat, P. Zschack, and W. M. Kriven, “Quad-
rupole Lamp Furnace for High Temperature (up to 2050K) Synchrotron Pow-
der X-ray Diffraction Studies in Air,” Rev. Sci. Instrum., 77 [9] 093906, 9pp
(2006).
26
L. F. Siah, W. M. Kriven, and J. Schneider, “In Situ, High-Temperature,
Synchrotron, Powder Diffraction Studies of Oxide Systems in Air, Using a
Thermal-Image Furnace,” Meas. Sci. Technol., 16, 1–8 (2005).
27
W. Yoon, P. Sarin, and W. M. Kriven, “Growth of Textured Mullite
Fibers Using a Quadrupole Lamp Furnace,” J. Eur. Ceram. Soc., 28 [2] 455–63
(2008).
28
P. Sarin, R. P. Haggerty, W. Yoon, M. Knapp, A. Berghaeuser, P. Zsc-
hack, E. Karapetrova, N. Yang, and W. M. Kriven, “A Curved Image-Plate
Detector System for High-Resolution Synchrotron X-ray Diffraction,” J. Syn-
chrotron Radiat., 16 [2] 273–82 (2009).
29
P. Zschack, “33 Bending Magnet Beamline Technical Information”; Avail-
able at http://www.aps.anl.gov/Sectors/33_34/admin/desc-33bm.html, 2009.
30
H. M. Rietveld, “A Profile Refinement Method for Nuclear and Magnetic
Structures,” J. Appl. Crystallogr., 2 [2] 65–71 (1969).
31
G. S. Pawley, “Unit-Cell Refinement from Powder Diffraction Scans,” J.
Appl. Crystallogr., 14, 357–61 (1981). h