American Mineralogist, Volume85, pages 283-287, 2000
New diamond anvil cells for gigahertz ultrasonic interferometry and X-ray diffraction
W.A. BASSETT,"* H.-J. REICHMANN2, R.J. ANGEL2, H. SPETZLER,' AND J.R. SMYTH'
'Department of Geological Sciences, Comell University, Ithaca, New York 14853, U.S.A.
2
Bayerisches Geoinstitut, Universitat Bayreuth, D-95440 Bayreuth, Germany
3
Department of Geological Sciences and Cooperative Institute for Research in Environmental Sciences,
University of Colorado, Boulder, Colorado 80309, U.S.A.
ABSTRACT
Two new diamond anvil cells have been designed for ultrasonic and X-ray diffraction measurements on
a single crystal sample up to 6 GPa and 250 ºC. Advances in the generation and transmission of coherent
GHz ultrasonic signals with wavelengths of the order of micrometers now make it practical to measure
elastic properties of samples small enough to be subjected to pressure and temperature in diamond anvil
cells. The signal is carried from a thin transducer through a sapphire buffer rod coupled to one of the
diamond anvils by means of force. The signal traverses the diamond anvil and enters the single crystal
sample, which is coupled to the anvil face by cement, adhesion, or by a normal force. Interference of
superimposed waves reflected from the near and far faces of the single crystal is used to measure travel
time of the sound waves in the sample. One of the diamond anvil cells employs the conventional geometry
in which access for the X-rays is through the diamond anvils. The other provides access for X-rays at high
angles to the load axis so that they do not need to pass through the diamond anvils and can therefore have
access to the sample while the buffer rod is in place. Both diamond anvil cells make it possible to measure
d-spacings at several different orientations using a four-circle goniometer. This capability is used for
detecting and correcting displacement of the sample from the center of the goniometer. Measurement of
travel times and lattice parameters at the same pressure-temperature conditions allows conversion of travel
times to velocities and can provide simultaneous equations of state, which EOS can then be used to make
an independent determination of pressure vs. lattice parameter. This provides a primary pressure scale.
This paper describes instrumentation, which allows us
measurement of ultrasonic travel time as well as lattice
parameter measurements on a single crystal sample at high
pressure and temperature. We describe how combining these
two techniques can provide accurate sound velocities as a
function of pressure and temperature as well as two
simultaneously determined equation of state measurements on
the same sample. The former is important for the interpretation
of seismic observations whereas the latter can be used to
establish a primary pressure scale.
INTRODUCTION
Ultrasonic interferometry has, for many years, been
an important source of data on elastic properties of minerals at
high pressures and temperatures (e.g., Jackson and Niesler
1982), The recent development of GHz ultrasonic
interferometry (Spetzler et al. 1993) has significantly reduced
the wavelength and, therefore, the sample size required for
making such measurements. This, in turn, has increased the
number of pertinent phases that can be studied with ultrasonic
interferometry by including small samples available from high
pressure-temperature synthesis. In addition, small samples
greatly increase the perfection with which single crystals can be
obtained.
Techniques for making accurate lattice parameter
measurements on single crystal samples under pressure by Xray diffraction in a diamond anvil cell have been developed
(King and Finger 1979; Hazen and Finger 1982; Angel et al.
1997). These techniques have proved to be very successful for
making direct volume vs. pressure measurements.
Establishing a reliable pressure scale is of fundamental
importance in all areas of high pressure research. Yoneda et al.
(1994) proposed a method for establishing a new primary high
pressure scale by using simultaneous measurements of
redundant equation of state (EOS) parameters. This approach
was also the basis for the very accurate determination of the
freezing point of mercury two decades earlier by Ruoff et al.
(1973).
THE NEW DIAMOND ANVIL CELLS
The diamond anvil cells described in this paper are based
on the hydrothermal diamond anvil cell (Fig. 1) described by
Bassett et al. (1993). The modifications that allow ultrasonic
interferometry are similar to those found in the diamond anvil
cell described by Spetzler et al. (1996). Further modifications
described in this paper make it possible to collect single crystal
X-ray diffraction data in addition to ultrasonic data on the same
single crystal sample at high pressure and high temperature.
The sample is placed in a gasket between two brilliant-cut
gem-quality, 1/8 carat diamonds mounted on tungsten carbide
seats with axial holes that permit access for light, sound and
other forms of radiation. Molybdenum wire 0.25mm in
diameter is wound around the tungsten carbide seats to serve as
heaters. The tungsten carbide seats are supported on annular
ceramic plates that are accommodated in recesses in the two
283
American Mineralogist, Volume85, pages 283-287, 2000
steel platens that form the cell body. These serve to thermally
isolate the heaters, anvils and sample. Pressures up to
approximately 10 GPa can be produced in the sample when
screws pull the stainless steel platens and drive the diamond
anvils together. Temperatures up to approximately 1000 ºC can
be produced by less than 10 V and less than 10 A. At
temperature over 300ºC, the diamonds and heaters must e
surrounded by a slightly reducing gas to protect these parts
from oxidation. We continually flow a gas mixture consisting
of one part hydrogen to 100 parts argon. Thermocouples
attached to the outside of the diamond anvils provide
temperatures that never differ by more than 8º from the sample
location and, once corrected, yield temperatures accurate to ±2
'C. Several different methods including X-ray diffraction
(Decker 197 1), ruby fluorescence (Piermarini et al. 1975), and
H20 equation of state (Bassett et al. 1993) have been used to
measure pressure.
culet and table faces of the diamonds parallel to within a few
milliradians of arc. In addition, the mounting device that holds
the buffer rod assembly to the diamond anvil cell allows it to be
translated and tilted to bring its end into exact parallelism with
the diamond table face. Visible-light interference fringes
provide a sensitive method for accomplishing this. The
assembly also provides sufficient normal force to the buffer
rod/diamond interface to ensure that acoustic coupling is
maintained without use of a bond.
Compressional-wave ZnO transducers on sapphire are
commercially available. We have achieved a usable bandwidth
from 300 MHz to 1.4 GHz with commercial transducers and up
to 2.4GHz using transducers that we sputtered directly onto the
sapphire buffer rods. We have failed to date, however, to
produce shear transducers by sputtering and have reverted to
the procedure of bonding and subsequently thinning LiNbO3
shear transducers on the buffer rod. Although not consistently
successful some of these transducers produced acceptable shear
wave signals at a few hundred megahertz and once as high as
600 MHz. To accomplish this, we started with commercial 25
MHz LiNbO, shear transducers (-90 Rm thick) and thinned
them by a factor of -25 to a thickness of -4 to 5 @m. Because
of the lower velocity of shear waves, wavelengths comparable
to those of compressional waves can be achieved at lower
frequencies. We chose, therefore, to thin the transducers until
the frequency corresponded to about 600 MHz.
The
wavelengths are thus approximately the same for the P- and Swaves. A program to develop procedures that can more
consistently produce successful shear wave transducers capable
of high frequencies is under way. One approach being
developed is the conversion of p-waves to s-waves by means of
a carefully controlled path within the buffer rod.
SOUND VELOCITY MEASUREMENT
A buffer rod consisting of a single crystal of sapphire
oriented with its c-axis coincident with the long axis of the rod
is pressed against the table face of one of the diamond anvils
(Fig. 1). The ultrasonic signals are transmitted from a thin
transducer attached to the end of the buffer rod. They travel
down the length of the buffer rod, pass through the diamond
anvil, and reflect from the surfaces of the sample back through
the diamond and the buffer rod to the transducer. it is important
that all interfaces be as perpendicular as possible to the
direction of propagation of the signal and therefore as parallel
to each other as possible. Thus, not only the faces of the buffer
rod but also the faces of the diamond anvil must be cut and
polished with a high precision. To this end we have developed
a procedure using reflected light for cutting and polishing the
A single pulse of ultrasonic signal traveling from the
transducer, down the buffer rod, through the diamond anvil,
and through the sample produces a series of reflections created
at each of the interfaces that the signal passes through. These
travel back to the transducer and are recorded as a function of
elapsed time. Two closely spaced pulses can be carefully timed
so that the reflection of the first pulse from the far surface of
the sample is superimposed on the reflection of the second
pulse from the near surface of the sample (Fig. 2). The
interference of these two signals as they return to the transducer
provides a sensitive means for determining the travel time of
the signal through the sample. The correct fringe number, i.e.,
the number of waves within the sample, is the one that shows
the least change in travel time as a function of frequency. Any
slope that it has is a measure of dispersion, However, no
significant sample dispersion has been observed to date.
The initial sample thickness is calculated from the
measured travel time at ambient conditions and the known
velocity for MgO. Velocity at pressure is calculated from
travel time and the thickness of the sample based on the initial
thickness and the measured compression.
284
American Mineralogist, Volume85, pages 283-287, 2000
LATTICE PARAMETER MEASUREMENT
We have used two designs of diamond anvil cells. The
measurement of ultrasonic travel times is virtually the same in
both while the method of obtaining X-ray diffraction data is
different. One uses the conventional geometry in which access
for the X-rays is through the diamond anvils (Fig. 3a). The
other provides access for X-rays at high angles to the load axis
so that they do not need to pass through the diamond anvils and
can therefore be used while the buffer rod is in place (Fig. 3b).
We call this the transverse cell.
Once the ultrasonic measurements as a function of pressure
and temperature have been made on a sample in the
conventional cell (Fig. 3a), the buffer rod and transducer can be
removed and the diamond anvil cell mounted on a four circle
goniometer. Single crystal X-ray diffraction data as a function
of pressure and temperature can then be collected on the same
sample by passing the X-rays in through one diamond and out
through the other. The diffraction data in turn can be used to
obtain accurate lattice parameters. Techniques for making
direct volume vs. pressure measurements on single crystal
samples by X-ray diffraction have been developed and
successfully applied to a wide range of samples at high
pressures in diamond anvil cells (Merrill and Bassett 1974;
Angel et al. 1997). The ruby method (Pierinarini et al. 1975)
was employed by us for measuring pressure. Recently, we have
adopted an internal diffraction standard because it is better able
to yield precise pressures.
285
American Mineralogist, Volume85, pages 283-287, 2000
thus preventing the elimination of all of the goniometer
aberrations and crystal offset errors from the measurements of
the Bragg diffraction angles. Second, even for thicker Crystals,
not all of the crystal is illuminated by the X-ray beam in all
eight of the equivalent positions. Thus, the effective diffraction
center of the crystal changes from one equivalent to the next in
a set of eight, and the data reduction of the eight-position
centering algorithm of King and Finger (1979) is not strictly
valid. Both effects together reduce the precision of lattice
parameter measurements to around I part in 2000.
The transverse diamond anvil cell (Fig. 3b) can be used to
collect X-ray diffraction data at the same time that the
ultrasonic data are being collected especially if the cable
carrying the ultrasonic signals to and from the diamond anvil
cell is sufficiently flexible to permit changes in orientation of
the cell while mounted on the 4 circle goniometer.
Both designs described above permit d-spacing
measurements of several reflections from each {hkl) set of
lattice planes. The method of eight-position diffracted-beam
centering (Finger and King 1978; King and Finger 1979;
Hazen and Finger 1982) allows the diffraction setting angles to
be determined free of the effects of diffractometer aberrations,
circle zero errors, and crystal offsets. Measurements made by
this method have an accuracy of better than 1 in 10 000 (Angel
et al. 1997). Clearly measurements by both ultrasonic and Xray diffraction simultaneously, as in the transverse method (Fig.
3b), have the advantage of eliminating pressure and
temperature measurements as a source of error when combining
ultrasonic and X-ray diffraction data. Experience has shown,
however, that with steel gaskets that are opaque to X-rays, the
transverse geometry results in larger uncertainties in the lattice
parameter measurements. These larger uncertainties arise from
two effects, both in turn related to the shadowing by the gasket
of the sample from the incident X-ray beam. First, for thinner
crystals the X-ray path to the crystal is completely obscured for
four of the eight equivalent goniometer settings of a reflection,
The more conventional design (Fig. 3a) does not suffer
from this source of error but depends on how well both
ultrasonic and X-ray diffraction measurements can be made at
the same pressure-temperature conditions but at different times.
286
American Mineralogist, Volume85, pages 283-287, 2000
MgO
We have applied the techniques described above to the
measurement of the elastic properties of MgO (Reichmann et
al. 1998). In that study, cylindrical samples of single crystal
MgO 60 to 150 µm long and 200 to 250 µm in diameter were
cut from a polished section of MgO. The cylinder axis of the
sample was oriented parallel to the crystallographic axis of the
crystal within 0.5º. The flat face of a cylindrical sample was
attached to one of the diamond anvil faces by super glue. It
was then immersed in either pure ethanol or a 4:1
methanol:ethanol mixture along with a chip of ruby in a
stainless steel gasket hole 300 to 400 µm in diameter and 250
µm thick.
We plotted c11 calculated from the travel time
measurements vs. pressure as measured by the ruby method
(Fig. 4). This plot is in excellent agreement with the data of
Jackson and Niesler (1982) and Yoneda (1990) collected using
other techniques. The disagreement with the data of Shen et al.
(1998) can be attributed to their use of solid KBr as pressure
medium, whereas e our measurements were made with either
pure ethanol or a 4:1 e methanol-ethanol mixture and were
therefore completely hydrostatic.
We also plotted the
normalized travel time, t(P)/t(P = 0), vs. lattice parameter
calculated from X-ray data collected over a range of pressure
and temperature (Fig. 5). This plot in which pressure is
ignored shows little scatter and illustrates the capability of
making ultrasonic velocities and lattice parameter
measurements on the same sample. Measurement of travel
times and lattice parameters at the same pressure-temperature
conditions can provide simultaneous equations of state which
have the potential to be used to make an independent
determination of pressure vs. lattice parameter thus providing a
primary pressure scale.
The diamond anvil cell we used to collect data on MgO
was the transverse style shown in Figure 3b. However, the Xray data were collected at a different time from the ultrasonic
data. Thus, these measurements suffered from the sources of
error inherent in both methods illustrated in Figure 3. That we
obtained such consistent data and in such good agreement with
those of Jackson and Niesler (1982) and Yoneda (1990) attests
to the success of both designs.
ACKNOWLEDGMENTS
We acknowledge the support of the National Science
Foundation (NSF), North Atlantic Treaty Organization
(NATO), the Alexander-von-HumboltStiftung, and the
Geoinstitut of the University of Bayreuth, Bayreuth, Germany.
We also thank the technical staff at the University of Bayreuth
and at the University of Colorado, Boulder, Colorado.
Decker, D.L. (197 1) High-pressure equation of state for NaCl, KCI
and CsCl. Journal of Applied Physics, 42, 3239-3244.
Finger, L.W. and King, H.E, (1978) A revised method of operation of
the singlecrystal diamond cell and the refinement of the structure
of NaCl at 32 kbar. American Mineralogist, 63, 337-342.
Hazen, R.M. and Finger, L.W. (1982) Comparative Crystal Chemistry,
231 P. Wiley, New York.
Jackson, 1. and Niesler, H. (1982) The elasticity of periclase to 3 GPa
and some geophysical implications. In S. Aidmoto and M.H.
Manghnani, Eds., High Pressure Research in Geophisics.
Advances in Earth and Planetary Sciences 12, 93112. Centre for
Academic Publications, D. Reidel, Dordrecht.
King, H.E. and Finger, L.W. (1979) Diffracted beam crystal centering
and its application to high-pressure crystallography, Journal of
Applied Crystallography, 12,374-378.
Merrill, L. and Bassett, W.A. (I 974) Miniature diamond anvil pressure
cell for single crystal X-ray diffraction studies. Review of
Scientific Instruments, 45, 290294.
Piermarini, G.J., Block, S., Barnett, J.D., and Fonnan, R.A. (1975)
Calibration of the pressure dependence of the R1 ruby
fluorescence line to 195 kbar. Journal Applied Physics, 46, 27742780.
Reichmann, H.J., Angel, R.J., Spetzler, H., and Bassett, W.A. (1998)
Ultrasonic interferometry and X-ray measurements on MgO in a
new diamond anvil cell. American Mineralogist 83, 1357-1360.
Ruoff, A.L., Lincoln, R.C., and Chen, Y.C. (1973) High-pressure
calibration with a new absolute pressure gauge, Applied Physics
Letters, 22, 310-311.
Shen, A.H., Reichmann, H.J., Chen, G., Angel, R.J., Bassett, W.A.,
and Spetzler, H. (1998) GHz ultrasonic interferometiy in a
diamond anvil cell: P-wave velocities in periclase to 4.4 GPa and
207'C. in Murli H. Manghanani and Takehiko Yagi, Eds.,
Properties of earth and Planetary materials at high pressure and
temperature, P. 71-77. Geophysical Monograph Series, 101.
Spetzler, H., Chen, G., Whitehead, S., and Getting, I.C. (1993) A new
ultrasonic interferometer for the determination of equation of state
of sub-millimeter single crystals. Pure and Applied Geophysics,
141, 341-373.
Spetzler, H., Shen, A., and Chen, G., Hermannsdoerfer, G., Schulze,
H.; and Weigel, R. (I 996) Ultrasonic measurements in a diamond
anvil cell. Physics of the Earth and Planetary Interiors, 98, 93-99.
Yoneda, A. (1990) Pressure derivatives of elastic constants of single
crystal MgO and MgAI,O,. Journal Physics of the Earth, 38, 1955
Yoneda, A., Spetzler, H., and Getting, 1. (I 994) Implication of the
complete travel time equation of state for a new pressure scale. In
J.W. Schmidt, G.A. Shaner, Samara, and M. Ross, Eds., HighPressure Science and Technology-1993, Proceedings of the
International Association for Research and Advancement of High
Pressure Science and Technology and the American Physical
Society, P. 1609-1612. American Institute of Physics, New York.
MANUSCRIPT RECEIVED MAY 12,1999
MANUSCRIPT ACCEPTED SEPTEMBER 30, 1999
PAPER HANDLED By ROBERT C. LIEBERMANN
REFERENCES CITED
Angel, R.J., Allan, DR, Miletich, R., and Finger, L.W. (1997) The use
of quartz as an internal pressure standard in high pressure
crystallography. Journal of Applied Crystallography, 30, 461-466.
Bassett, W.A. Shen, A.H., and Bucknum, M. (I 993) A new diamond
anvil cell for hydrothermal studies to 2.5 GPa and from - 190 to
1200'C. Review of Scientific Instruments, 64, 2340-2345.
287