Journal of Crystal Growth 49(1980)97—101 © North-Holland Publishing Company
GROWTH AND CHARACTERIZATION OF CePd3-SINGLE CRYSTALS R. TAKKE and W. ASSMUS Physikalisches Inst itut der Universität Frankfurt, Robert-Mayer-Str. 2—4, D-6000 Frankfurt am Main, Germany Received 2 Octobei 1979
CePd3 crystals have been grown from a levitated melt using the Czochralski method. Crystal characterization has been made using a Laue X-ray camera, SEM and EPMA. The temperature dependence of the elastic constants is measured and found to be normal. Anomalies in the elastic constants which might be caused by valence fluctuations are not seen.
prereacting was less than 0.001 wt%. Pre-reacted pellets were transferred to a cold crucible which was built into a commercial ADL crystal puller. The growth chamber was evacuated to less than lO-~mbar and then filled with argon (purity better than 5N5) up to a pressure of 5 bar. The pellets were subsequently melted and levitated for at least one hour before pulling, to ensure a completely homogenous melt. CePd3 melts congruently with a melt temperature of 1350°C. The crystals were grown from the levitated melt using the Czochralski method. The main problem was to produce a stable and vibration free melt. Growth parameters such as melt weight, RF power, crucible geometry and atmosphere must be carefully chosen as material properties such as melting temperature, electrical conductivity, vapor pressure, viscosity, surface tension and density all affect the levitation. A stabilized melt was obtained with the crucible geometry used with a melt weight of nearly 80 g and an RF power output of 12 kW at 200 kHz. Fig. 1 shows a growing crystal from a levitated melt and fig. 2 a crucible showing the effects of contact with molten CePd3 when it is not perfectly levitated. Crystals were pulled from the melt using a seed attached to a water cooled pulling rod. Homogeneous CePd3 polycrystalline material, cut with a spark cutter from a levitated melt, were first used as seeds. Single crystals were later used as seeds as they became available. The pulling speed was approximately 3 mm/h with no rotation so as to eliminate mechani-
Recent investigations on CePd3, a Cu3Au type intermetallic compound, have indicated that it is one of the best materials for the study of intermediate valence systems [1—3]. All investigations have been performed on polycrystalline samples as single crystals were not available, The difficulty in growing CePd3 crystals lies in its high affinity for other materials at its melting ternperature. Consequently there is a lack of suitable crucible materials and crucible free methods must be used. Growth experiments using the floating zone or the tn-arc furnace with the melt on a water cooled copper plate remain unsuccessful .
2. Czochralski growth from a levitated melt CePd3 crystals have been grown from a levitated melt using a cold crucible supplied by Metals Research Ltd., as developed by Hukin . The starting materials cerium (Rare Earth Prod. 4N) and palladium (Demetron 3N5) were cut, surface cleaned and weighed together stoichiometrically in a glove box under argon. The humidity of the argon gas was less than 30 ppm. Due to the highly exothermic reaction of Ce and Pd it was not possible to react the components together in the cold crucible. An argon arc furnace using a water cooled copper-plate was used to prereact. Argon purity 5N5 was used. Weight loss in 97
R. Takke and W. A ssmus / Growth and characterization of CePd
Fig. 1. CePd3 crystal, growing from a levitated melt, radiation shield removed.
cal vibrations. The effects of levitation on the mixing of the melt is far greater than the effects of crystal rotation. Temperature is controlled by adjustment of the RF output power. The melt becomes unstable and begins to vibrate when the RF output power is increased to appreciably more than is needed to melt the material. Therefore a good thermal insulation is required. A radiation shield made from molybdenum sheet lowers the heat losses due to radiation and convection considerably. Crystals up to 100 mm in length and 5—lU mm diameter have been grown.
3. Etching and characterization In contrast to pure cerium, CePd3 has no affinity for water vapor or oxygen. However surfaces of the as grown crystals were covered with a cerium layer. To clean the surfaces of the crystals the electropolishing methods often used for intermetallic rare earth corn-
Fig. 2. Crucible, damaged during a melt contact.
pounds was attempted with little success. Etching the crystals in concentrated HNO3 with the gradual addition of concentrated HC1 during the etch process produced the desired results. After etching most of the crystals showed longitudinal etch grooves fig. 3. Laue photographs of the crystals showed that large crystallites of different orientations were bounded by these etch grooves. In some cases the crystals were single and showed no etch grooves. Single crystals grown are typically 20 mm long and 4—5 mm in diameter. Crystals were cut with a spark cutter and orientated using a Laue camera. A typical Laue pattern is shown in fig. 4. CePd3 crystals were examined in a scanning electron microscope. X-ray element distribution analysis
R. Takke and W. Assrnus / Growth and characterization of CePd
Fig. 3. Etch grooves on the crystal, different crystallites are bounded by these grooves,
made using an energy dispersive solid state X-ray detector showed no deviation from stoichiometry within the accuracy of the system. The electron probe microanalysis of the as grown crystals showed a
Fig. 5. Region A: surface of the crystal. Region B: the inner crystal.
Number of ~ Counts 3 8x10
1Tp11T 1..~H TT~,,,
1,,,,1,,11,1...1.1, X-Ray- energy
6 5 Ce 3.
Fig. 4. Laue photograph of a CcPd3 single crystal.
Fig. 6. Microprobe analysis of regions A and B.
R. Takke and W. Assmus
/ Growth and characterization of CePd~-singlecrystals
Fig. 7. Pd-rich hillocks on a etched surface.
cerium layer of approximately 2 pm on the surface (penetration depth of electrons at 25 keV in cerium is approximately 2 pm). A part of an as-grown crystal surface was broken away to reveal the inner crystal. Fig. 5 shows clearly the crystal surface (region A) and the inner crystal (region B). A corresponding analysis made with the microprobe can be seen in fig. 6. The as grown surface consists solely of cerium whereas the inner crystal region is CePd3. An analysis of polished CePd3 surfaces show no changes in stoichiometry over a period of three months at room temperature. This would indicate that the cerium coating takes place only at high temperature. Therefore care should be taken in repeated annealing as first estimates indicate that for a 5 mm diameter crystal a 2 pm layer of Ce results in a change of stoichiometry in the crystal in the 1000 ppm range.
The etching of the crystals can also produce changes in the stoichiometry of the surface when cerium is preferably dissolved. Fig. 7 shows a non-
Table 1 Pertinent parameters of CePd3 Melting point cDensity c11 (RT) c12 (RT) 44 (RT)
1350°C 3 10.81 g cm 3 l6.88xlO~ ergcm3 6.82 X 1011 erg cm3 6.04 x 1011 erg cm
Fig. 8. Fcho patterns of the measured elastic modes: (a) c 1~. (b)c44,(c)~(c11_c12),cL~(c1I+cla+2c44).
R. Takke and W. .4ssmus / Growth and characterization of CePd
1 part in iO~and an absolute accuracy of 1%  was Fig. 8 shows the echo patterns of the measured modes ~(c11 c12), c44 and CL ~(c11 + + 2c44) at room temperature at a frequency of 10 MHz. It can been seen that the crystals are of good quality for ultrasonic attenuation experiments. Fig. 9 shows the temperature dependence of the elastic constants.
Above helium temperature there is no elastic anomaly which might be caused by valence fluctuations. Table 1 shows the pertinent parameters of this system.
This work was 65supported by the Sonderforschungsbereich “Festkorperspektroskopie” Darmstadt/Frankfurt, financed by special funds of
6.0 ~4. 52 0
the Deutsche Forschungsgemeinschaft. —
T(K) Fig. 9. Temperature dependence of the elastic constants.
stoichiometric crystal surface with palladium rich
hillocks. A further characterization of CePd3 crystals was made by measuring the elastic constants of the material. A phase comparison method with an accuracy of
[11 W.E. Gardner, J. Penfold, T.F. Smith and I.R. Harris, J. Phys. F2 (1972) 133.  A. Furrer and T.G. Purwins, J. Phys. C9 (1976) 1491. [31E. Loewenhaupt, W. Schmatz K. Holland-Moritz, Wohileben, Phys.M. Rev. Letters 38 (1977) 983. and D.  E. Bucher, E. Holland-Moritz, private communication.  D. Hukin, Brit. Patent No. 1, 269, 762 (1972).  T.J. Moran and B. LUthi, Phys. Rev. 187 (1969) 710.