Structural and vibrational properties of Ca2FeH6 and Sr2RuH6

Article abstract of DOI:10.1016/j.jpcs.2011.01.001

The structural and vibrational properties of the isostructural compounds Ca₂FeH₆ and Sr₂RuH₆ are determined by periodic DFT calculations and compared with their previously published experimental crystal structures as well as new experimental vibrational data. The analysis of the vibrational data is extended to the whole series of alkaline-earth iron and ruthenium hydrides A₂TH₆ (A=Mg, Ca, Sr; T=Fe, Ru) in order to identify correlations between selected frequencies and the TH bond length. The bulk moduli of Ca₂FeH₆ and Sr₂RuH₆ have also been determined within DFT. Their calculated values prove to compare well with the experimental values reported for Mg₂FeH₆ and several other compounds of this structure.

Full text of DOI:10.1016/j.jpcs.2011.01.001

Journal of Physics and Chemistry of Solids 72 (2011) 286–289
Contents lists available at ScienceDirect
Journal of Physics and Chemistry of Solids
journal homepage: www.elsevier.com/locate/jpcs
Structural and vibrational properties of Ca2FeH6 and Sr2RuH6
H. Hagemann a,n, V. D’Anna a, L.M. Lawson Daku a, S. Gomes b,1, G. Renaudin b,1, K. Yvon b
a
ꢀ
De´partement de Chimie Physique, Universite´ de Geneve, 30, quai E. Ansermet, CH 1211 Geneva 4, Switzerland
b
ꢀ
Laboratoire de Cristallographie, Universite´ de Geneve, 24, quai E. Ansermet, CH 1211 Geneva 4, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
The structural and vibrational properties of the isostructural compounds Ca2FeH6 and Sr2RuH6 are
determined by periodic DFT calculations and compared with their previously published experimental
crystal structures as well as new experimental vibrational data. The analysis of the vibrational data is
extended to the whole series of alkaline-earth iron and ruthenium hydrides A2TH6 (A¼Mg, Ca, Sr;
T¼Fe, Ru) in order to identify correlations between selected frequencies and the T–H bond length. The
bulk moduli of Ca2FeH6 and Sr2RuH6 have also been determined within DFT. Their calculated values
prove to compare well with the experimental values reported for Mg2FeH6 and several other
compounds of this structure.
Received 30 September 2010
Received in revised form
26 November 2010
Accepted 13 January 2011
Available online 21 January 2011
Keywords:
A. Inorganic compounds
C. Ab initio calculations
D. Crystal structure
D. Phonons
& 2011 Elsevier Ltd. All rights reserved.
1. Introduction
2. Experimental
2.1. Sample preparation
The title compounds belong to the class of so-called complex
transition (T) metal hydrides that derive their name from the
presence of T metal hydrido complexes in the structure. For T¼Fe,
Ru, Os their general formula is A2TH6 (A¼Ca, Sr, Ba), and their
hydrido complexes have octahedral shape and the composition
[TH6]4ꢀ. Their structural and physical properties have been studied
abundantly in the past, in particular in view of their potential as
hydrogen storage materials (for a critical and comprehensive
review see [1]). Their prototype Mg2FeH6 was discovered in the
1980s and characterized with respect to a variety of properties [2],
including structural, vibrational and hydrogen sorption properties.
The compound turned out to have one of the highest hydrogen
density known among condensed materials (i.e. more than twice
that of liquid hydrogen), but was thermally too stable for reversible
hydrogen storage applications at room temperature. Later, how-
ever, it was shown to be of interest as a material in thermal heat
energy storage systems [3]. As to the vibrational properties of the
A2TH6 series several members have been studied in the past, such
as Mg2FeH6 or Ca2RuH6 [4–9]. A recent review [10] summarizes
the data available so far. In this paper we present the vibrational
spectra of two less well studied members, Ca2FeH6 and Sr2RuH6,
and compare the results of DFT calculations and calculated bulk
moduli to those of other members in this series.
Mixtures of the binary hydrides (deuteride) CaH2, CaD2 and
SrH2 and metallic Fe and Ru powders (Johnson Matthey, 99.99%)
with molar ratio 2:1 were pressed to pellets to prepare the three
ternary hydrides (deuteride) Ca2FeH6, Ca2FeD6 and Sr2RuH6,
respectively. The starting materials were prepared by hydrogena-
tion (deuteration) of metallic calcium powder (STREM 99.9%) and
strontium powder (STREM 99.9%) at 775 K under a hydrogen
(deuterium) pressure of 20 bar for two days. All materials were
handled in an argon-filled glove-box since they were sensitive to
air and moisture. The three pellets were placed in nickel crucibles
and sintered in an autoclave at 775 K under a hydrogen (deuter-
ium) pressure of 150 bar for four days. X-ray powder diffraction
showed that the samples were single-phase and contained the
expected ternary hydride (deuteride) having a cubic Fm-3m
symmetry with K2PtCl6-type structure [11,12].
2.2. Vibrational spectra
The Raman spectra of Ca2FeD6 were collected on a Kaiser
Holospec f/1.8 spectrograph with a 50-
m
m slit (ꢁ3–4 cmꢀ1
resolution) that was equipped with a liquid nitrogen cooled CCD
camera, using a 488 nm Ar laser excitation with a power of ca
15 mW. These spectra were subject to a very strong background.
In the case of Ca2FeH6, no Raman spectrum could be observed due
to the very strong background. The new Raman spectrum of
Sr2RuH6 was measured using a Renishaw InVia Raman micro-
scope using 785 nm excitation.
n Corresponding author. Tel.: +41 22 379 6539.
E-mail address: Hans-Rudolf.Hagemann@unige.ch (H. Hagemann).
1
Present address: Clermont Universite´, ENSCCF and Universite´ Blaise Pascal,
Laboratoire des Mate´riaux Inorganiques, UMR 6002, BP 10448, F-63000 Clermont-
Ferrand, France.
0022-3697/$ - see front matter & 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jpcs.2011.01.001

H. Hagemann et al. / Journal of Physics and Chemistry of Solids 72 (2011) 286–289
287
FTIR spectra were obtained using a Paragon 1000 (Perkin-
Elmer) instrument equipped with a Golden Gate Single Reflection
Diamond ATR set-up (P/N 10 500 Graseby-Specac Series).
The powder samples to be analyzed were pressed against the
ATR crystal with a calibrated strength in an Ar filled glove-box
and protected by an airtight protective polyethylene film during
measurement.
The IR active modes transform as T1u, the Raman active modes
as A1g, Eg, T2g
.
Fig. 1 presents the IR spectra of CaH2, Ca2FeH6 and Ca2FeD6. A
weak band at 1000 cmꢀ1 for Ca2FeH6 is probably due to a CaH2
impurity. 3 strong bands are observed for Ca2FeH6 at 924, 1442
and 1570 cmꢀ1, while only two IR active fundamentals (u3 and u4)
are expected. The band at 1442 cmꢀ1 can be assigned to the
combination of u6+libration (ca 900+540 cmꢀ1), which is
enhanced by Fermi resonance, in agreement with previous results
for Ca2RuH6 and Sr2RuH6 [8]. In the case of Mg2FeH6, this
combination mode is not observed. Indeed, the corresponding
frequency would be expected at 460+836¼1296 cmꢀ1, which
appears to be too different from the frequency of the allowed u3
band at 1746 cmꢀ1 to give rise to a Fermi resonance.
2.3. Computational methods
The computational study of Ca2FeH6 and Sr2RuH6 was done
within DFT [13,14], using the ABINIT [15–17] package which is
based on plane waves and pseudopotentials.
The weak features around 1860 and 1900 cmꢀ1 also arise
from combination tones (u5+u6 and u4+u5, respectively). It is
interesting to note that the relative intensity of the u4 band is
extremely weak for Mg2FeH6 [9], while it is stronger for Ca2FeH6
(see Fig. 1) and Ca2RuH6 [9,33]. The IR spectrum of Ca2FeD6 shows
the expected isotopic shifts. The Raman spectrum of Ca2FeD6
For both systems, the calculations were performed in the
Generalized Gradient Approximation (GGA) [18,19] using norm
conserving Troullier–Martins [20] and optimized Rappe–Rabe–
Kaxiras–Joannopoulos (RRKJ) [21] pseudopotentials in the separ-
able Kleinman–Bylander form [22], generated with the
FHI98PP [23] and OPIUM [24] programs.
Convergence studies performed on both systems led to the
choice of a 10 ꢂ 10 ꢂ 10 Monkhorst–Pack grid [25] and a kinetic
energy cut-off of 60 hartree for the two systems.
The optimizations were performed with the crystal symmetry
kept fixed to Fm-3m (number 225) space group; the Wyckoff
positions of the alkaline-earth (Ca, Sr), transition metal (Fe, Ru)
and H atoms are 8c, 4a and 24e, respectively. The structures of the
compounds were fully relaxed by letting simultaneously evolve
the cell parameter (a) and the reduced coordinate xH of the
hydrogen atom, which is the only adjustable internal coordinate.
In order to calculate the bulk moduli of Ca2FeH6 and Sr2RuH6,
their equations of state have been determined by optimizing their
structures for fixed values of the lattice parameter (see below).
For both systems at ambient pressure, their phonons were
calculated at the gamma point, within density-functional pertur-
bation theory [26,27]. The symmetry analysis of these frequencies
were performed with the ABINIT ‘‘get_irreps.py’’ utility, which
makes use of the symmetry information available from the Bilbao
Crystallographic Server [28,29].
1140
1060
Ca2FeH6
924
Ca2FeD6
1570
1442
CaH2
Phonon calculations were also performed for Ca2FeD6 and
Sr2RuD6.
600
800
1000
1200
Wavenumber [cm-1]
1400
1600
1800
Fig. 1. IR spectra of CaH2, Ca2FeH6 and Ca2FeD6 (blue) at room temperature. (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)
3. Results and discussion
3.1. Experimental vibrational spectra
1170
710
The members of the series A2TH6 (A¼Ca, Sr, Ba) based on
T¼Fe, Ru, Os have fully ordered structures. They crystallize in the
Fm-3m (no. 225) space group with Z¼4 and contain a fcc lattice
of [TH6]4ꢀ anions separated by A2+ cations in tetrahedral
voids [1]. Recently, some partially disordered members of this
series have been reported for manganese (T¼Mn) and some rare
earths (A¼Y, Er, Dy) [29–31]. Interestingly, while these hydrides
display octahedral [MnH6]4 complexes, their general composition
is AMn2H6. Given that the disorder occurs on the A site, the
compounds must be formulated as (A,Mn)MnH6. For the sake of
comparison, they will be included in the considerations below.
The symmetry analysis shows the following optical modes:
8000
Ca2FeD6
6000
851
4000
134
Sr2RuH6
2000
ꢃ
Internal modes of the octahedral [MH6]4ꢀ anions: A1g+Eg+
2T1u+T2g+T2u; A1g (u1), Eg (u2) and T1u (u3) are stretching
modes, T1u (u4), T2g (u5) and T2u (u6) are bending modes;
libration mode of [MH6]4ꢀ anions: T1g
200
400
600
800
1000
1200
Raman Shift [cm-1]
ꢃ
ꢃ
Fig. 2. Raman spectrum of Sr2RuH6 (as obtained with the Renishaw instrument)
and Ca2FeD6 (obtained with the Holospec Monochromator, a strong background
was substracted for clarity) at room temperature.
translational motions of the cation against the anion: T1u
and T2g
.

288
H. Hagemann et al. / Journal of Physics and Chemistry of Solids 72 (2011) 286–289
(see Fig. 2) shows two bands at 1170 (A1g (u1)) and 710 cmꢀ1
(T2g (u5)). For Mg2FeD6, the relative intensity of the u5 band was
also found to be very weak [9], in contrast to the spectrum
in Fig. 2.
Fig. 2 presents the Raman spectrum below 1000 cmꢀ1 of
Sr2RuH6. Besides the previously reported T2g bending mode (u5)
at 851 cmꢀ1 [5,8], the translational T2g mode at 134 cmꢀ1 is
observed for the first time.
and theoretical results from the literature and from this study, we
have plotted IR and Raman active stretching modes against the
M–H bond length in Fig. 3. The frequency of the IR active u3 mode
appears to decrease systematically with increasing M–H bond
length (as highlighted by the dashed lines in Fig. 3), however, the
data point for Sr2RuH6 does not appear to follow this trend. A
possible explanation may be that the reported experimental Ru–H
˚
bond length [12] (1.69(1) A) is in fact too short. A bond length of
˚
The experimental data are collected in Table 1.
ca 1.715 A would bring this value in line with the other results
(see Fig. 3). The DFT calculations on Sr2RuH6 and Ca2RuH6 support
this suggestion, as the calculated Ru–H bond length is larger in
the Sr compound than in the Ca compound, while the vibrational
frequencies for both stretching modes (u1 and u3) are smaller for
the Sr compound (see the open symbols in Fig. 3).
Finally, we have studied theoretically the pressure dependence
of the crystal structure of Ca2FeH6 and Sr2RuH6. The calculated
pressure–volume curves could be fitted using a Vinet equation of
state [35]. The results are summarized in Table 3 and compared
with experimental values of similar compounds [32,36]. The
calculated bulk moduli of Ca2FeH6 and Ca2RuH6 are very close. In
the case of Ca2RuH6 we had seen that under pressure the [RuH6]4ꢀ
octahedron is not compressed very much compared to the entire
lattice [7]. The bulk modulus reflects thus mainly the interaction
between the Ca2+ ions and the relatively rigid [MH6]4ꢀ anions.
3.2. Theoretical calculations
Table 2 compares the calculated structures of Ca2FeH6 and
Sr2RuH6 with the experimental structures reported by [4,11,34,35].
The agreement for the lattice parameter a is excellent. The Fe(Ru)–H
bond length is calculated to be about 0.04–0.05 A larger than the
experimentally reported value, which reflects the usual GGA over-
estimate of the structural parameters.
˚
Table 1 includes the theoretically calculated vibrational frequen-
cies for Ca2FeH6 and Sr2RuH6 and their deuterated analogs. The
agreement between theory and experiment is quite good; the slight
underestimation of the experimental values can be ascribed to the
tendency of the GGA calculations towards slightly smaller force
constants (also related to the overestimated structural parameters).
As recently shown in our study of Ca2RuH6 [7], the frequency
of the librational mode is particularly sensitive to the surround-
ings of the [MH6]4ꢀ complex and to pressure, thus making its
theoretical determination quite challenging. Thus noticeably
different values were predicted at the LDA and GGA levels for
the librational frequency of Ca2RuH6 [7]. This explains the
discrepancy observed between experiment and theory for the
librational frequency of Sr2RuH6, which is calculated at 478 cmꢀ1
and was observed at 561 cmꢀ1 by Barsan et al. [8].
Os
1800
Ru
1700
Fe
1600
1500
1400
It is interesting to note that the splitting between u1 and u3 is
much smaller for the Fe compounds than for the Ru compounds.
Kritikos and Nore´us [4] have shown a correlation between IR
active stretching frequencies and unit cell parameter. We have
previously [5] attempted to correlate Ru–H stretching frequencies
with Ru–H bond length using Badger’s rule, however the correla-
tion was not very satisfactory. Using both experimental [4–10]
1.55
1.60
1.65
1.70
1.75
M-H Bond Length [Å]
Fig. 3. M–H stretching frequencies vs. M–H bond lengths.
Table 1
Observed and calculated vibrational frequencies (cmꢀ1). The frequency of the librational band at 540 cmꢀ1 was estimated from the combination band at 1442 cmꢀ1
(see text). The experimental data for Sr2RuH6 were taken from [8].
M–H (D) stretching modes
M–H (D) bending modes
Librations
T1g (IA)
Translations
T2g (R)
A1g (R)
Eg (R)
T1u (IR)
T2g (R)
T1u (IR)
T2u (IA)
u1
u2
u3
u5
u4
u6
Ca2FeH6 (Ca2FeD6)
Experimental
Calculated
(1170)
1570 (1142)
1550 (1113)
(710)
924 (681)
877 (633)
ca 540
1615 (1143)
1584 (1121)
934 (661)
839 (539)
505 (357)
202 (203)
Sr2RuH6 (Sr2RuD6)
Experimental [8]
Calculated
1782 (1265)
1733 (1226)
1731 (1235)
1706 (1207)
1487 (1075)
1503 (1073)
850 (602)
793 (561)
879 (634)
839 (600)
856
561
134 (this work)
131 (131)
799 (565)
478 (338)
Table 2
Crystal structures of Ca2FeH6 and Sr2RuH6: comparison of theory with experiment.
3
xH
˚
˚
˚
a (A)
M–H (A)
V (A )
Exp.
Calc.
Exp.
Calc.
Exp.
Calc.
Exp.
Calc.
Ca2FeH6
Sr2RuH6
7.036(1)
7.608[4]
7.032
7.627
0.230[11]
0.223[12]
0.234
0.228
1.618(5)
1.69[12]
1.6451
1.7359
348.319
440.54
347.726
443.67
Note: In the case of Ca2FeD6 the experimental values are in fact the values of Ca1.98Mg0.02FeD6 [11].

H. Hagemann et al. / Journal of Physics and Chemistry of Solids 72 (2011) 286–289
289
Table 3
Wiley, Chichester, 2005, pp. 1814–1846, ISBN 0-470-86078-2; DOI: 10.1002/
0470862106.ia087 (15 March 2006).
Experimental and theoretical bulk moduli of A2MH6 compounds.
[2] J.J. Didisheim, P. Zolliker, K. Yvon, P. Fischer, J. Schefer, M. Gubelmann,
A.F. Williams, Inorg. Chem. 23 (1984) 1953–1957.
[3] B. Bogdanovic, A. Reiser, K. Schlichte, B. Spliethoff, B. Tesche, J. Alloys Compd.
345 (2002) 77–89.
[4] M. Kritikos, D. Nore´us, J. Solid State Chem. 93 (1991) 256.
[5] H. Hagemann, R.O. Moyer, J. Alloys Compd. 330–332 (2001) 296–300.
[6] M.M. Barsan, R.O. Moyer Jr., I.S. Butler, D.F.R. Gilson, J. Alloys Compd. 424
(2006) 73–77.
[7] L.M. Lawson Daku, H. Hagemann, Phys. Rev. B 76 (2007) 014118.
[8] M.M. Barsan, I.S. Butler, D.F.R. Gilson, R.O. Moyer Jr., W. Zhou, H. Wu,
T.J. Udovic, J. Phys. Chem. A 112 (2008) 6936.
0
Compound
B0 (GPa)
B0
˚
a (A)
Mg2FeH6 [36]
75.4(4)
76.3
68.05
83.8
6.446
YMn2D6 [32]
6.28
4.59
5.85
5.20
4.28
4.0
6.708
6.720
6.677
6.6797
7.032
7.253
7.627
DyMn2D6 [32]
HoMn2D6 [32]
ErMn2D6 [32]
Ca2FeH6 (GGA)
Ca2RuH6 (GGA) [7]
Sr2RuH6 (GGA)
73.2
76.1
59.7
58.5
49.7
[9] S.F. Parker, K.P.J. Williams, M. Bortz, K. Yvon, Inorg. Chem. 36 (1997)
5218–5221.
4.29
[10] S.F. Parker, Coord. Chem. Rev. 254 (2010) 215.
[11] B. Huang, F. Bonhomme, P. Selvam, K. Yvon, P. Fischer, Journal Less-Common
Met. 171 (1991) 301–311.
[12] R.O. Moyer jr, C. Stanitski, J. Tanaka, M.I. Kay, R. Kleinberg, Journal Solid State
Chem. 3 (1971) 541–549.
[13] P. Hohenberg, W. Kohn., Phys. Rev. 136 (1964) B864.
[14] W. Kohn, L.J. Sham., Phys. Rev. 140 (1965) A1133.
[15] X. Gonze, et al., Z. Kristallogr. 220 (2005) 558.
Comparing the experimental and theoretical values of the bulk
moduli vs. unit cell parameters, it can be noted that the value of the
bulk modulus is larger for smaller unit cell dimensions (or molar
volumes), in agreement with trends reported for other systems [37].
[16] X. Gonze, et al., Comput. Phys. Commun. 180 (2009) 2582.
´
[17] ABINIT code, a common project of the Universite Catholique de Louvain,
4. Conclusions
ꢀ
ꢀ
Corning Incorporated, the Universite´ de Liege, the Commisariat a l’Energie
Atomique, Mitsubushi Chemical Corp., the Ecole Polytechinique Palaiseau,
and other contributors (URL /http://www.abinit.orgS).
The structural vibrational and elastic/mechanical properties of
Ca2FeH6 and Sr2RuH6 have been studied and compared with the
properties of other compounds of this family. The agreement
between experimental and theoretical values for structural para-
meters and vibrational frequencies is generally good. The calcu-
lated bulk moduli appear to be very reasonable compared to the
available experimental data of related systems. Periodic DFT
calculations appear therefore to be quite reliable to calculate
these properties for other systems in this family of compounds
where experimental data are lacking.
[18] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865.
[19] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 78 (1997) 1396E.
[20] N. Troullier, J.L. Martins, Phys. Rev. B 43 (1991) 1993.
[21] A.M. Rappe, K.M. Rabe, E. Kaxiras, J.D. Joannopoulos, Phys. Rev. B 41 (1990)
1227.
[22] L. Kleinman, D.M. Bylander, Phys. Rev. Lett. 48 (1982) 1425.
[23] M. Fuchs, M. Scheffler, Comput. Phys. Commun. 119 (1999) 67.
[24] /http://opium.sourceforge.netS (last visited: 2010-02-12).
[25] H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13 (1976) 5188.
[26] X. Gonze, C. Lee, Phys. Rev. B 55 (1997) 10355.
[27] X. Gonze, Phys. Rev. B 55 (1997) 10337.
[28] M.I. Aroyo, J.M. Perez-Mato, C. Capillas, E. Kroumova, S. Ivantchev,
G. Madariaga, A. Kirov, H. Wondratschek, Z. Kristallogr. 221 (2006) 15.
[29] M.I. Aroyo, A. Kirov, C. Capillas, J.M. Perez-Mato, H. Wondratschek, Acta
Crystallogr. A 62 (2006) 115.
Acknowledgements
[30] V. Paul-Boncour, S.M. Filipek, M. Dorogova, F. Boure´e, G. Andre´, I. Marchuk,
´
A. Percheron-Guegan, R.S. Liu, J. Solid State Chem. 179 (2005) 356.
[31] V. Paul-Boncour, S.M. Filipek, G. Andre´, F. Boure´e, M. Guillot, R. Wierzbicki,
This work was supported by the Swiss National Science
Foundation and by a Grant from the Swiss National Supercomput-
ing Centre (CSCS) under Project ID 103. The authors thank
´
I. Marchuk, R.S. Liu, B. Villeroy, A. Percheron-Guegan, H.D. Yang, S.C. Pin,
J. Phys.: Condens. Matter 18 (2006) 6409.
[32] S.M. Filipek, H. Sugiura, V. Paul-Boncour, R. Wierzbicki, R.S. Liu, N. Bagkar,
J. Phys. Conf. Ser. 121 (2008) 022001.
[33] R.O. Moyer Jr., J.R. Wilkins, P. Ryan, J. Alloys Compd. 290 (1999) 103.
[34] R.O. Moyer Jr., C. Stanitski, J. Tanaka, M.I. Kay, B. Kleinberg, J. Solid State
Chem. 3 (1971) 541.
ꢀ
´
Mr. C. Schnyder (Musee d’histoire naturelle de Geneve) for help
with the Raman spectrum of Sr2RuH6.
[35] P. Vinet, J.H. Rose, J. Ferrante, J.R. Smith, J. Phys.: Condens. Matter 1 (1989)
1941.
References
[36] L. George, V. Drozd, A. Durygin, J. Chen, S.K. Saxena, Int. J. Hydrogen Energy
34 (2009) 3410.
[37] L. Glasser, Inorg. Chem. 49 (2010) 3424.
[1] K. Yvon, G. Renaudin, Hydrides: solid state transition metal complexes, in:
Encyclopedia of Inorganic Chemistry, second ed., H. Crabtree (Ed.), John