Effect of chemical and external pressure on the structure of intermetallic compound CeNi

Article abstract of DOI:10.1016/j.jallcom.2006.10.080

Neutron powder diffraction was employed to study the structural modifications of the intermediate-valence compound CeNi at room temperature induced by either chemical or external pressure. For the first time we were able to record the diffraction pattern resulting from the pressure-induced first-order phase transition occurring in CeNi at 300 K. At pressure P = 2 GPa we observe the coexistence of two phases while only single pressure-induced phase is visible at P = 5 GPa. The results obtained are indicative of a higher symmetry of the collapsed structure as compared to the CrB-type ambient pressure structure of CeNi. The critical pressure range around 2 GPa is found to be agreement with the previous estimation derived from the thermopower measurements.

Full text of DOI:10.1016/j.jallcom.2006.10.080

Journal of Alloys and Compounds 444–445 (2007) 281–284
Effect of chemical and external pressure on the structure of
intermetallic compound CeNi
a,∗
a
b
b
A. Mirmelstein , E. Clementyev , V. Voronin , Yu. Akshentsev ,
c
a
a
a
D. Kozlenko , A. Kutepov , A. Petrovtsev , Yu. Zuev
Russian Federal Nuclear Center, Institute of Technical Physics, Snezhinsk 456770, Russia
Institute for Metal Physics, Russian Academy of Sciences, 620041 Ekaterinburg GSP-170, Russia
a
b
c
Frank Laboratory of Neutron Physics, JINR, Dubna 141980, Russia
Received 29 June 2006; received in revised form 27 September 2006; accepted 17 October 2006
Available online 20 November 2006
Abstract
Neutron powder diffraction was employed to study the structural modifications of the intermediate-valence compound CeNi at room temperature
induced by either chemical or external pressure. For the first time we were able to record the diffraction pattern resulting from the pressure-
induced first-order phase transition occurring in CeNi at 300 K. At pressure P = 2 GPa we observe the coexistence of two phases while only single
pressure-induced phase is visible at P = 5 GPa. The results obtained are indicative of a higher symmetry of the collapsed structure as compared to
the CrB-type ambient pressure structure of CeNi. The critical pressure range around 2 GPa is found to be agreement with the previous estimation
derived from the thermopower measurements.
©
2006 Elsevier B.V. All rights reserved.
Keywords: Rare earth alloys and compounds; Phase transitions; High-pressure; Neutron diffraction
1
. Introduction
phase diagram is established only at relatively low pressures
<0.8 GPa) in the limited temperature range below 150 K and the
(
The study of mechanisms of the transitions between dif-
structure of the high-pressure phase remains unknown. Recent
thermopower measurements of CeNi as a function of external
pressure at 300 K indicate possible phase transitions to occur at
∼3 and 8 GPa [4].
ferent structural modifications for the materials with almost
empty or almost filled f5/2 electronic configurations may pro-
vide a new insight into the origin of their ground-state proper-
ties, especially in the case if the structural transitions involve
volume discontinuity driven by electron correlation effects.
The intermediate-valence compound CeNi exhibits anomalous
behavior of many physical properties including the first-order
structural phase transition with volume discontinuity ∼5% [1,2]
and well-defined ground-state coherent effects in the magnetic
response function observed by the inelastic neutron scattering
technique [3]. Thus, CeNi is an attractive model system to inves-
tigate the influence of the coherent effects on the mechanism of
the volume-collapse phase transitions. However, the CeNi P–T
In the present work the neutron powder diffraction technique
was used to study the pressure-induced phase transition in CeNi
as well as the structural modifications of CeNi resulting from the
chemical pressure due to either La (negative chemical pressure)
or Lu (positive chemical pressure) substitutions for Ce in the
CrB-type CeNi lattice.
2
. Experiment
2
.1. Chemical pressure
CeNi, Ce1−xLaxNi (0 ≤ x ≤ 1) and Ce1−xLuxNi (0 ≤ x ≤ 0.4) samples were
∗
Corresponding author at: Department of Experimental Physics, Russian Fed-
prepared by arc melting of high purity elements (∼99.98) in the argon atmo-
sphere. Obtained compositions were then recrystallized in vacuum and founded
in cylindrical ingots of a diameter ∼10 mm. Ingots were then annealed in
dynamical vacuum for 120 h. According to X-ray and neutron powder diffrac-
tion measurements (see Fig. 1 as an example), all samples were found to
eral Nuclear Center, E.I. Zababakhin Institute of Technical Physics, Vasil’ev
Street 13, Snezhinsk, 456770 Chelyabinsk Region, Russia.
Tel.: +7 35146 52070; fax: +7 35146 52070.
E-mail address: mirmelstein@mail.ru (A. Mirmelstein).
0
925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2006.10.080

2
82
A. Mirmelstein et al. / Journal of Alloys and Compounds 444–445 (2007) 281–284
Fig. 3. Structural spectra for CeNi at ambient pressure, at 2 and 5 GPa measured
at room temperature using DN-12 time-of-flight high-pressure spectrometer
Fig. 1. Neutron powder diffraction pattern of CeNi at ambient pressure and
room temperature recorded using D7A diffractometer (Ekaterinburg) with
λ = 1.5324 A˚ . h k l Miller indexes are indicated below the correspondent Bragg
(
Dubna). (h k l) indexes correspond to the Cmcm space group.
peaks.
2
.2. High-pressure neutron measurements
Table 1
The crystal lattice parameters of CeNi as a function of La and Lu substitutions
for Ce
Using the time-of-flight high-pressure neutron spectrometer DN-12 (IBR-
pulsed reactor, Frank Laboratory of Neutron Physics, Dubna) and a sapphire
2
a ( A˚ )
b ( A˚ )
c ( A˚ )
V ( A˚ 3)
anvil cell, we were able to record the pressure-driven evolution of elastic neutron
spectra of CeNi at 300 K up to P = 5 GPa (Fig. 3). Ambient pressure spectrum
shows the pure CrB-type diffraction pattern, the same as in Fig. 1. A shift and/or
splitting of some spectral lines as well as appearance of additional maxima
at P = 2 GPa indicate clearly the development of the pressure-induced phase
transition. At 5 GPa the diffraction pattern changes significantly. The number of
relatively intensive peaks is reduced as compared to ambient pressure diffraction
pattern. In particular, the (1 1 1) and (0 2 0) peaks are absent at 5 GPa. Thus, we
can assume a single-phase character of the 5 GPa diffraction pattern associated
with a new phase of CeNi resulting from a first-order pressure-induced phase
transition.
Ce0.6Lu0.4Ni
Ce0.8Lu0.2Ni
Ce0.9Lu0.1Ni
Ce0.95Lu0.005Ni
CeNi
Ce0.9La0.1Ni
Ce0.8La0.2Ni
LaNi
3.693(2)
3.746(2)
3.759(1)
3.7699(5)
3.7858(6)
3.8014(6)
3.819(1)
3.9048(5)
10.286(3)
10.407(3)
10.461(3)
10.495(1)
10.548(1)
10.571(1)
10.587(2)
10.792(1)
4.310(2)
4.342(2)
4.352(1)
4.3568(5)
4.3631(6)
4.3683(7)
4.370(1)
4.3900(5)
163.7
169.2
171.15
172.4
174.2
175.6
176.7
185.0
be of single phase with the CrB-type of crystal structure (the Cmcm space
group).
3. Discussion
In order to study the effect of chemical pressure, neutron powder diffrac-
tion patterns for the samples under study were recorded at room temperature
using D7A diffractometer (Institute for Metal Physics, Ekaterinburg) with
La substitutions for Ce expand the crystal lattice in all the
main crystallographic directions, while substitutions with Lu
lead to the lattice compression. The main structural elements
of the CrB structure are the zig-zag chains of Ce and Ni ions
propagating along the c crystallographic direction. Respectively,
the lattice parameter c varies much less due to chemical pressure
than the a and b parameters.
λ = 1.5324 A˚ . The results are summarized in Table 1 and Fig. 2.
CeNi primitive cell contains two Ce and two Ni ions. Ce
and Ni are both located at (4c) (0, y, 1/4) positions. Fig. 2
demonstrates the variation of the atomic coordinate y for the
rare earth ions Ce/La/Lu and Ni as a function of the unit cell
volume derived from the Rietveld refinement of the neutron
diffraction patterns. While yCe/La/Lu increases slightly and
almost linearly with decrease in the unit cell volume, yNi shows
nonmonotonous behavior. It decreases also almost linearly
when going from LaNi to CeNi, reaches its minimum for the
slightly compressed (as compared to the stoichiometric CeNi)
composition Ce0.95Lu0.05Ni and then grows again. It should be
noted, that in CeNi both yCe and yNi position parameters exhibit
significant changes within the temperature range between 80
and 120 K, whereas for LaNi the values of these parameters
remain almost temperature independent [5]. Although at ambi-
Fig. 2. Position parameter y for Ce/La/Lu and Ni ions as a function of
the unit cell volume at room temperature and ambient pressure for the
Ce1−xLaxNi–CeNi–Ce1−xLuxNi series.

A. Mirmelstein et al. / Journal of Alloys and Compounds 444–445 (2007) 281–284
283
by two different techniques are in reasonable agreement. At
the same time, our structural data are not consistent with the
results derived from the de Haas-van Alphen effect in CeNi [7].
One cannot exclude, however, a temperature-driven phase tran-
sitions in CeNi within the pressure range 2 GPa < P < 5 GPa, so
that the CeNi structures can be different at room and at low
temperatures.
The neutron diffraction data for Ce1−xLaxNi–CeNi–Ce1−x
LuxNi series listed in Table 1 (lattice parameters a, b and c)
and shown in Fig. 2 (position parameters) were extrapolated
to their values correspondent to the volume compression down
3
˚
to ꢀV/V0 = −25% (where V0 = 174.2 A is the unit cell vol-
ume of CeNi) and used for the electronic structure calculations
in terms of the spin-polarized relativistic density functional
theory in generalized gradient approximation (full potential
linearized augmented plane wave version). For a few values
of ꢀV/V0 the minimum of the full energy E was determined
as a function of yNi (this parameter demonstrates the most
Fig. 4. Thermopower S as a function of pressure for CeNi (from Ref. [4]). The
phase transitions are assumed to occur around ∼3 and ∼8 GPa (arrows), i.e. in
the regions of broad singularities of S vs. P curve.
pronounced variation in Fig. 2). At ꢀV/V0 = −25% E
was
min
found at ꢀy = 0.002 with respect to its stoichiometric value
Ni
yNi = 0.4254. Although ꢀyNi = 0.002 turns out to be much less
than the extrapolated value ꢀyNi ∼ 0.015, no signature of the
phase instability was detected since the steepness of the E ver-
sus yNi dependence remains unchanged at any |ꢀV/V0| ≤ 25%.
It cannot be excluded that a single-electron theory fails to pre-
dict a crystal lattice instability in CeNi driven by electronic
correlations.
ent pressure CeNi exhibits clear signatures of the crystal lattice
instability upon cooling [5,6], a structural phase transition does
not occur down to very low temperature. The question whether
the compressed Lu-doped compositions undergo a phase trans-
formation upon cooling at ambient pressure or not remains to
be open.
In 1985 Gignoux and Voiron demonstrated the existence in
CeNi of a first-order phase transition under pressure associated
with a large volume discontinuity [1,2]. The P–T phase diagram
was determined up to 0.8 GPa and 150 K. Since that time how-
ever to the best of our knowledge neither the phase diagram
out of this limited P–T domain nor the crystal structure of the
pressure-induced phase were published. The only indirect evi-
dence of unchanged crystal symmetry across the transition was
obtained from the de Haas-van Alphen effect under pressure [7],
which reveals that the topology of the Fermi surface of CeNi is
not change after transition, but the cyclotron effective mass is
strongly reduced.
The main result of the present study is the direct obser-
vation of the pressure-induced phase transition in CeNi at
room temperature. Although the structural identification of the
pressure-induced phase is by no means trivial because of a
rather limited number of well-reserved Bragg peaks, there are
no doubts that the 5 GPa diffraction pattern cannot be described
in terms of the ambient pressure Cmcm space group. There-
fore, the structure of pressure-induced phase does not belong to
the orthorhombic CrB-type and has, most probably, a higher
symmetry. From Fig. 3 it follows that at 300 K the phase
transition starts at or just below P = 2 GPa, while the 5 GPa
pressure range lies above the transition region. According to
our previous thermopower measurements the room tempera-
ture phase transitions in CeNi were assumed to occur at ∼3
and ∼8 GPa [4]. Since the singularities of thermopower ver-
sus pressure dependence around 3 and 8 GPa are rather broad
4. Conclusion
Neutron powder diffraction study of the structural varia-
tion of intermediate-valence compound CeNi due to chemi-
cal pressure-induced by either La or Lu substitutions for Ce
revealed nonmonotonous variation of the Ni position param-
eters yNi which has a minimum near the stoichiometric CeNi
composition. For the first time we were able to detect a pressure-
induced first-order structural phase transformation in CeNi by
direct experimental technique. The results obtained suggest that
the symmetry of the pressure-induced phase is higher that the
orthorhombic CrB-type structure of ambient pressure phase
of CeNi.
Acknowledgements
This work was performed under auspices of Russian Federal
Agency for Atomic Energy (State Contract #6.06.19.19.06.988).
Financial support by RFBR (grant #05-08-33456-a) is gratefully
acknowledged.
References
[1] D. Gignoux, J. Voiron, Phys. Rev. B 32 (1985) 4822.
[
[
2] D. Gignoux, C. Vettier, J. Voiron, JMMM 70 (1987) 388.
3] E.S. Clementyev, J.-M. Mignot, P.A. Alekseev, V.N. Lazukov, E.V. Nefe-
dova, I.P. Sadikov, M. Braden, R. Kahn, G. Lapertot, Phys. Rev. B 61 (2000)
6189.
[
4] A. Mirmelstein, E. Clementyev, A. Kutepov, V. Shchennikov, A. Karkin, A.
Petrovtsev, Yu. Zuev, B.S Goshchitskii, Proceeding of the Fifth International
(Fig. 4), the critical pressure values ∼2 and ∼3 GPa derived

2
84
A. Mirmelstein et al. / Journal of Alloys and Compounds 444–445 (2007) 281–284
Workshop, “Fundamental Plutonium Properties”, Snezhinsk, Chelyabinsk
Region, Russia, September 12–16, 2005, pp. 30–32.
5] V.N. Lazukov, E.V. Nefedova, V.V. Sikolenko, U. Staub, P.A. Alekseev, M.
Braden, K.S. Nemkovskii, C. Pradervant, I.P. Sadikov, L. Soderholm, N.N.
Tiden, Appl. Phys. A 74 (2002) S559 [Suppl.].
[6] E.S. Clementyev, P.A. Alekseev, M. Braden, J.-M. Mignot, G. Lapertot, V.N.
Lazukov, I.P. Sadikov, Phys. Rev. B 57 (2000) R8099.
[7] S. Araki, R. Settai, Y. Inada, Y. Onuki, Physica B 281–282 (2000)
736.
[