Synthesis, crystal structure and hydrogenation properties of the novel metal compound LaNi2Mn3

Article abstract of DOI:10.1016/S0925-8388(02)00838-1

LaNi₂Mn₃ has been synthesised from the elements and investigated with respect to crystal structure and hydrogenation properties. Single crystal X-ray diffraction reveals a YNi₂Al₃ type structure (space group P6/mmm, a = 9.2099(9), c = 4.1892(4) ?, c/a = 0.4548) that is distinctly different from the CaCu₅ type structure of LaNi₅. The compound absorbs reversibly up to six hydrogen atoms per formula unit near ambient conditions and shows a hydrogen plateau pressure of ～1 bar near 60 °C. The hydrogen atoms in the expanded metal atom structure were localised by neutron powder diffraction on deuterides of composition LaNi₂Mn₃D_4.6 (a = 9.7125(2), c = 4.3057(1) ?, c/a = 0.4433) and LaNi₂Mn₃D_5.6 (a = 9.8379(2), c = 4.3038(1), c/a = 0.4375). They occupy one octahedral coordinated site [La₂Mn₄] with full occupancy and three tetrahedral coordinated sites [La₂Mn₂], [La₂NiMn] and [LaMn₂Ni], and a deformed square pyramidal coordinated site [LaMn₂Ni₂] with partial occupancies. Elsevier Science B.V. All rights reserved.

Full text of DOI:10.1016/S0925-8388(02)00838-1

Journal of Alloys and Compounds 348 (2003) 176–183
L
www.elsevier.com/locate/jallcom
Synthesis, crystal structure and hydrogenation properties of the novel metal
compound LaNi2Mn3
*
´ ´
L. Guenee, K. Yvon
´
`
`
Laboratoire de Cristallographie, Universite de Geneve, 24 Quai E. Ansermet, CH-1211 Geneve 4, Switzerland
Received 24 May 2002; accepted 4 June 2002
Abstract
LaNi2Mn3 has been synthesised from the elements and investigated with respect to crystal structure and hydrogenation properties.
˚
Single crystal X-ray diffraction reveals a YNi2Al3 type structure (space group P6/mmm, a59.2099(9), c54.1892(4) A, c/a50.4548) that
is distinctly different from the CaCu5 type structure of LaNi5. The compound absorbs reversibly up to six hydrogen atoms per formula
unit near ambient conditions and shows a hydrogen plateau pressure of |1 bar near 60 8C. The hydrogen atoms in the expanded metal
˚
atom structure were localised by neutron powder diffraction on deuterides of composition LaNi2Mn3D4.6 (a59.7125(2), c54.3057(1) A,
c/a50.4433) and LaNi2Mn3D5.6 (a59.8379(2), c54.3038(1), c/a50.4375). They occupy one octahedral coordinated site [La2Mn4]
with full occupancy and three tetrahedral coordinated sites [La2Mn2], [La2NiMn] and [LaMn2Ni], and a deformed square pyramidal
coordinated site [LaMn2Ni2] with partial occupancies.
2002 Elsevier Science B.V. All rights reserved.
Keywords: Hydrogen absorbing materials; Metal hydrides; Crystal structure; Neutron diffraction
1. Introduction
relatively poor hydrogen absorbers. The purpose of the
present work was to synthesise a compound of com-
position LaNi2Mn3, to investigate its crystal structure and
to study its hydrogenation properties. It will be shown that
the compound crystallises with the expected YNi2Al3 type
structure and shows excellent hydrogen storage properties.
Intermetallic LaNi5 and its substitution derivatives are
of interest for applications in rechargeable batteries (for a
recent review see Ref. [1]). The manganese substituted
series LaNi52xMnx has been studied in detail in Ref. [2]
and found to crystallise with a partially disordered CaCu5
type structure for manganese contents 0#x#2.2 (800 8C).
For higher manganese contents (x.0.2, 600 8C) a new
ternary compound appeared in a phase diagram [3] the
composition of which was stated to be La3Mn10Ni7. A
ternary compound of composition LaNi2Mn3 and its
hydrogenation properties have so far not been reported. A
recent study of structure maps has shown [4] that this
composition is outside the stability range of the CaCu5
type structure but inside that of the hexagonal YNi2Al3
type structure (also called HoNi2.6Ga2.4 type). The latter
structure is ordered and closely related to, but significantly
different from, the CaCu5 type structure. So far, only
compounds based on aluminium, gallium and tin were
found to crystallise with that structure, and they were
2. Experimental
2.1. Synthesis and hydrogenation properties
Samples were obtained by placing compressed powder
mixtures of the elements (La rod (Johnson Mattey,
99.99%), Ni and Mn powders (Johnson Mattey, 5N) in an
induction furnace and by re-melting them several times in
a silver crucible under argon (6N). The ingots (|1 g) were
then wrapped into Ta foils, placed into quartz ampoules
that were sealed under 0.3 bar argon pressure, and an-
nealed at various temperatures. In spite of negligible
melting losses (,0.1 wt.%) initial attempts to obtain
LaNi2Mn3 as a single phase at the stoichiometric com-
position failed, even after long-time annealing of the
samples at 600 or 800 8C. The sample which was annealed
at 600 8C during 4 days, for example, contained a well-
*
Corresponding author. Tel.: 141-22-702-6231; fax: 141-22-702-
6864.
E-mail address: klaus.yvon@cryst.unige.ch (K. Yvon).
0925-8388/02/$ – see front matter
PII: S0925-8388(02)00838-1
2002 Elsevier Science B.V. All rights reserved.

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
177
crystallized majority phase that could be attributed to the
hexagonal YNi2Al3 type structure (space group P6/mmm,
˚
refined cell parameters a59.19089(4), c54.17861(4) A)
and two Ni(Mn)-rich secondary phases of likely com-
position La2(Ni,Mn)17 (rhombohedral Th2Zn17 type struc-
ture) and LaNi5Mn6 (tetragonal CeNi5Mn6 type structure).
On the other hand, the sample which was annealed at
800 8C during 24 h contained well crystallized LaNi5Mn6
as a majority phase and traces of orthorhombic LaNi (CrB
type structure) and possibly some La2O3, while the
LaNi2Mn3 and La2(Ni,Mn)17 phases had disappeared. In
order to avoid the formation of these unwanted phases a
lanthanum excess of up to 5 wt.% (nominal composition
La1.07Ni2Mn3) was used during the subsequent syntheses
and the samples were annealed between 600 and 630 8C
for up to 11 days (see below). X-ray powder diffraction
patterns showed these samples to be nearly single-phase,
containing mainly LaNi2Mn3 (YNi2Al3 type structure),
very little Mn(Ni) (|4 wt.%) and some traces of La2O3
(,1 wt.%).
For the hydrogenation (deuteration) experiments various
alloy samples were powdered in a glove box to an average
grain size of 63 mm. One sample (|0.5 g mass) was filled
into a Ni container and placed into an autoclave that was
evacuated. The temperature was increased up to 80 8C
while hydrogen was introduced to a pressure of 2 bar, and
the temperature was kept at that level for one night. X-ray
synchrotron patterns indicated that the hydride samples
were nearly single-phase but tended to desorb hydrogen
Fig. 2. Pressure-composition isotherm of La1.07Ni2Mn3 at T560 8C in
microbalance (composition corrected for 7 wt.% Mn, maximum H content
6.08 H/f.u.).
(see Section 2.2). Other samples (nominal composition
La1.07Ni2Mn3, |0.05 g mass) were placed into a microbal-
ance (Hiden Analytical, model IGA001) and hydrogenated
at various temperatures (50–80 8C) and pressures (0.01–3
bar). As can be seen from Fig. 1 their absorption at 60 8C
and 1.5 bar showed good kinetics and led to a maximum
hydrogen content of about 6 H atoms per formula unit
(H/f.u). The sample was not activated prior to the mea-
surement. A pressure-composition isotherm at 60 8C for
the same sample during absorption is presented in Fig. 2. It
Fig. 1. Mass increase (mg) versus time of La1.07Ni2Mn3 in microbalance, T560 8C, 1.5 bars hydrogen pressure, initial sample weight 0.06113 g, total
weight variation Dm5287.6 mg, maximum hydrogen content 6.03 H/f.u., values corrected for the presence of 7 wt.% Mn impurity.

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
178
shows a rather well-defined plateau at |1 bar that extends
to |6 H atoms per formula unit (H/f.u). The hydride was
found to be stable at ambient temperature under a hydro-
gen pressure of ,1 bar, but desorbed rapidly if exposed to
air. Complete desorption without further oxidation and/or
segregation occurred within 5 days.
remain of the YNi2Al3 type but showed a considerable
lattice expansion (a59.8951(3), c54.3061(1) A, DV/V5
˚
18.7%) compared to the alloy. Furthermore, the expansion
was anisotropic as shown by the cell parameter ratios
c/a50.4351 (hydride) and 0.4544 (alloy). Unfortunately,
the partial hydrogen desorption and the resulting sample
inhomogeneity prevented the metal atom substructure from
being refined. Subsequently, a deuterated sample (|7 g
mass) was prepared by deuteration of three alloy batches of
nominal composition La1.04Ni2Mn3 in an autoclave (4.5
bar, T565 8C, 12 h). Prior to the experiment the alloys
were annealed at 630 8C for 11 days, finely ground and
sieved (grain size ,64 mm). A quantitative X-ray phase
analysis showed that they contained mainly the LaNi2Mn3
phase (96 wt.%) and some Mn(Ni) (4 wt.%). After the
deuteration reaction was completed, the sample (called I)
was filled into a vanadium container under protective
atmosphere (1 bar argon) and measured on the neutron
2.2. Structure analysis
In order to study the structure of the alloy a single
crystal was isolated from the bulk and measured on an
X-ray diffractometer (STOE IPDS) equipped by an image
plate (Mo Ka radiation). The absence of systematic extinc-
tions confirmed space group P6/mmm and yielded the
˚
refined cell parameters a59.2099(9) and c54.1892(4) A.
A structure refinement (programme Xtal3.7 [5]) was
performed (on F2) by placing La, Ni and Mn on the Y, Ni
and Al sites, respectively, of the YNi2Al3 type structure
[6] and by varying one scale factor and 15 atomic
parameters. Results are summarised in Table 1. A refine-
ment of occupancy factors gave no indication for a
possible disorder of Ni and Mn on the three atom sites.
Finally, a phase analysis and structure refinement by using
synchrotron X-ray data was made on a polycrystalline
sample that was subsequently subjected to hydrogenation
studies. The results showed that LaNi2Mn3 was the
majority phase (89 wt.%) and Mn(Ni) (6.7 wt.%), LaNi
(1.5 wt.%) and La2O3 (2.7 wt.%) minority phases. Com-
pared to the single crystal results no significant structural
differences were observed for the LaNi2Mn3 phase except
for slightly smaller cell parameters (a59.20179(7), c5
˚
powder diffractometer HRPT (l51.494 A) at SINQ (PSI,
Villigen). As expected, the main phase had a strongly
˚
expanded hexagonal cell (a59.7125(2), c54.3057(1) A)
and the concentration of the secondary Mn(Ni) phase was
nearly the same as in the starting material. Five deuterium
sites in the metal atom substructure of the main phase were
located ab initio by the program FOX [7] and their
coordinates refined by the programme Fullprof [8]. The
following 32 parameters were allowed to vary: two scale
factors, five profile and 25 atomic parameters. The patterns
are presented in Fig. 3 and the refinement results are
summarised in Table 2. After completing the experiment,
sample I was again placed in an autoclave, deuterated at a
slightly higher temperature and pressure (T575 8C, 6 bar,
12 h), and re-measured on the neutron powder diffractome-
˚
4.18078(4) A).
For the study of the hydride structure a hydrogenated
sample was first measured by synchrotron radiation. The
metal atom substructure of the majority phase was found to
˚
ter D1A (l51.9114 A) at ILL (Grenoble). As expected,
this new sample (called II) contained a more deuterium-
Table 1
Structure refinement results on a single crystal of LaNi2Mn3; e.s.d.s in parentheses
2
˚
Atoms
Site
x
y
z
Ueq. (A )
Occupancy
La(1)
La(2)
Ni(1)
Mn(1)
Mn(2)
1a
2d
6l
6k
3f
0
1/3
0.18159(5)
0.2985(1)
1/2
0
0
1/2
0
1/2
0
0.0177(3)
0.0172(2)
0.0175(3)
0.0178(4)
0.0176(5)
1.0
1.0
1.0
1.0
1.0
2/3
2x
0
0
3
˚
R
int. 512.9%
Electron density residuals61.5el./A
R2F 55.3%
wRF 52.3%
Nmeasa5246, Nuniqueb5146
GoF51.06
Anisotropic displacement amplitudes
U11
U22
U33
U12
U13
U23
La(1)
La(2)
Ni(1)
Mn(1)
Mn(2)
2U12
2U12
0.0177(4)
0.0209(4)
0.0170(5)
2U12
2U12
2U12
2U12
2U12
0.0188(6)
0.0182(5)
0.0156(6)
0.0142(7)
0.0182(9)
0.0086(2)
0.0084(1)
0.0100(2)
0.0086(2)
0.0091(3)
0
0
0
0
0
0
0
0
0
0
˚
Space group P6/mmm, a59.2099(9), c54.1892(4) A.
a Number of measured reflections.
b Number of unique reflections.

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
179
Fig. 3. Observed (points), calculated (line) and difference (bottom line) neutron diffraction patterns of La1.04Ni2Mn3D4.6 (sample I) measured on HRPT
˚
(PSI, l51.494 A). Vertical lines indicate Bragg peak position of the contributing phases LaNi2Mn3D4.6 (97%) and Mn (3%).
rich main phase, some Mn and traces of La2O3 (0.3 wt.%).
The refinements were performed along similar lines as for
sample I. A total of 30 parameters were refined (three
scale, seven profile and 20 atomic) while the displacement
parameters of the five deuterium sites were constrained to
be equal. The patterns are represented in Fig. 4 and the
refinement results are summarised in Table 2. Interestingly,
despite the anisotropic lattice expansion of the main phase
(deuterides: c/a50.4433 (I), 0.4375 (II); alloy
La1.04Ni2Mn3: c/a50.4543) no significant anisotropic
diffraction line broadening appeared. This contrasts with
the hydrides of LaNi5 and most of their derivatives that
show strongly broadened (hk0) reflections [9].
VYNi2Al3 53VCaCu5), but differ with respect to their metal
atom substructures, in particular that of lanthanum. In fact,
the La positions in LaNi2Mn3 (YNi2Al3 type) can be
derived from those in LaNi3Mn2 (CaCu5 type) by displac-
ing one out of three La atoms along 1/2c (for a com-
parison between a CaCu5 type structure and a YCo3Ga2
type structure which is a substitution variant of the
YNi2Al3 type structure, see Ref. [10]). As a consequence,
the Ni and Mn atoms in the vicinity of the displaced La
sites are able to relax and create atomic environments that
differ significantly between both structure types. In
LaNi3Mn2 (CaCu5 type) the partially disordered Ni/Mn
substructure consists of two mixed Ni/Mn sites that have
coordination number CN512 and shortest bond distances
that are relatively similar for both sites (Ni–Mn(Ni)52.56
˚
˚
A, Mn(Ni)–Mn(Ni)52.59 A [2]). In LaNi2Mn3 (YNi2Al3
type) the ordered Ni/Mn substructure consists of two Mn
sites having CN512 and one Ni site having CN511, and
the shortest bond distances differ between these sites
3. Results and discussion
3.1. Intermetallic compound
˚
˚
(Ni–Mn52.55 A, Mn–Mn52.75 A) as expected from
atomic size considerations (RNi 51.246, RMn 51.304 A).
˚
In contrast to LaNi3Mn2 that crystallises with a partially
disordered CaCu5 type structure [2] LaNi2Mn3 crystallises
with an ordered YNi2Al3 type structure. Both structures
are related as far as their hexagonal cell parameters are
Altogether, these structural differences lead to rather
different atomic environments in both hydride structures
and thus to different hydrogenation properties (see Section
3.3).
concerned
(aYNi2Al3 5œ33aCaCu5
,
cYNi2Al3 5cCaCu5,

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
180
Table 2
Structure refinement results on polycrystalline La1.04Ni2Mn3D4.6 (sample I, first line) and La1.04Ni2Mn3D5. 6 (sample II,second line); e.s.d.s in parentheses
Phase 1:
LaNi2Mn3D4.6: space group P6/mmm, refined phase content |97 wt.%,
a59.7125(2), c54.3057(1) A, DV/V 514.7%
a
˚
LaNi2Mn3D5.6: space group P6/mmm, refined phase content |93 wt.%,
a
˚
a59.8379(2), c54.3038(1) A, DV/V 517.7%
2
˚
Atom
La1
Site
x
y
z
B (A )
Occupancy
1.0
1a
0
0
0
1.15(9)
0.81(6)
1.08(5)
BLa1
La2
Ni1
Mn1
Mn2
D1
2d
6l
1/3
2/3
2x
0
1/2
0
1.0
1.0
1.0
1.0
0.1841(2)
0.1847(3)
0.2899(5)
0.2856(7)
1/2
1.78(3)
1.95 (5)
1.95(8)
2.0 (1)
1.7 (1)
1.2 (1)
2.4 (2)
3.04 (8)
2.38 (9)
BD1
2.9 (1)
BD1
2.1 (2)
BD1
2.00 (8)
BD1
6k
1/2
0
3f
0
6m
3g
6j
0.072(1)
0.068(1)
0.5
2x
0
0.5
0.5
0
0.227(5)
0.288 (7)
0.987(7)
1.00 (1)
0.538(7)
0.68 (1)
0.164(3)
0.246 (6)
0.351(3)
0.413 (6)
D2
D3
0.7440(6)
0.7458(6)
0.8003(6)
0.7981(5)
0.5415(5)
0.4596(6)
0
D4
12o
12p
2x
0.630(2)
0.643(2)
0.0
0.0
32 variables
31
D5
0.3346(6)
0.6675(7)
Nrefb5205,
109,
R
Bragg 58.49%
5.65%
]
˚
Phase 2:
Mn (a-Mn type structure), space group I43m, a58.9474(2) A
(refined phase content |3 (7) wt.%)
R
Bragg 524.2%
Nbref 5110
57
wp 514.4%
12.3
18.7%
Rp 516.8%
R
Chi257.91
13.8
5.71
a Relative volume increase during deuteration.
b Number of reflections.
3.2. Deuterides
D–D distances (Table 4) and D site occupancies are
consistent with repulsive D–D interaction (D–D.2.1 A).
˚
As expected, the main phase in sample
I
As can be seen from the radii of the various metal atom
(LaNi2Mn3D4.6) contains less deuterium than in sample II
(LaNi2Mn3D5.6), and the refined deuterium contents
(4.58(6) and 5.57(9) D/f.u., respectively) are slightly
inferior than, but consistent with those measured for the
hydride samples on the microbalance (|6 H/f.u). As
shown in Fig. 5 deuterium occupies five different inter-
stices, one octahedral coordinated site [La2Mn4] with full
occupancy, and three tetrahedral coordinated sites
[La2Mn2], [La2NiMn] and [LaMn2Ni], and a deformed
square pyramidal coordinated site [LaMn2Ni2] with partial
occupancies. The shapes of these interstices differ sig-
nificantly from those occupied in CaCu5 type hydrides
AB5Hx that have tetrahedral ([A2B2], [AB3], [B4]) and
octahedral ([A2B4]) but no square-pyramidal metal coordi-
nations. The metal–deuterium bond distances (Table 3) are
in the usual range for intermetallic hydrides [11], and the
interstices (see Rsite in Table 4) they are all bigger than 0.4
A, in agreement with Westlake’s model [12]. Thus, from a
purely geometric point of view the maximum deuterium
content possible in this series of compounds is 8 D/f.u.
˚
3.3. Hydrogenation properties
LaNi2Mn3 absorbs reversibly up to 5.6 H/f.u. near
ambient conditions (60 8C). Thus its maximum hydrogen
capacity exceeds that of the end member of the
LaNi52xMnx (CaCu5 type) series (LaNi3Mn2: 4.8 H/f.u. at
60 8C [2]) and those of other systems showing a structural
change from CaCu5 type to YNi2Al3 type structure such as
neodymium-based NdNi52xAlx (CaCu5 type: x50.5,1,1.5;
YNi2Al3 type: x51.89–2.97; maximum H content 1.4
H/f.u. at x51.89 under 50 bar at 20 8C, activated at

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
181
Fig. 4. Observed (points), calculated (line) and difference (bottom line) neutron diffraction pattern of La1.04Ni2Mn3D5.6 (sample II) measured on D1A
˚
(ILL, l51.9114 A). Vertical lines indicate Bragg peak position of the contributing phases LaNi2Mn3D5.6 (92.5%), Mn (3%) and La2O3 (0.5%).
300 8C [13]), or gadolinium-based GdNi3Al2 (1.2H/f.u, 50
bar at 20 8C, activated at 300 8C) [14]) and GdNi52xGax
(x52.5: no significant hydrogen absorption up to 427 8C
and 200 bar [15], x52–3: inert up to 100 bar and 400 8C
[16]). As to the hydrogen plateaus, that of LaNi2Mn3 (|1
bar at 60 8C) is relatively well developed and in a more
useful pressure range than that of LaNi3Mn2 (|1025 bar at
60 8C [2]). On the other hand, in the aluminium substituted
series NdNi52xAlx compositions within the CaCu5 type
structure domain (i.e., at low Al contents 0#x#1.5) show
relatively wide plateaus that decrease with aluminium
content, while compositions within the YNi2Al3 type
structure domain (i.e., at high Al contents 1.89#x#2.97)
show no plateaus and small capacities (NdNi2.6Al2.4: 1.1
H/f.u. [13]). Finally, the kinetics of LaNi2Mn3 is rela-
tively slow at the beginning of the hydrogenation reaction
but becomes rather fast once the reaction has started (see
insert of Fig. 1), which indicates that it can be improved
by activation. The calculated hydrogen densities of
LaNi2Mn3H6 are 1.42 wt.% and 82.5 g H2 l21
.
4. Conclusions
LaNi2Mn3 is a new intermetallic compound the struc-
ture and hydrogenation properties of which are distinctly
different from those of LaNi3Mn2 having similar com-
position. The relatively abrupt changes of hydrogenation
properties as a function of composition are attributed to the
structural differences between the alloys that lead to
different hydrogen environments and thus different metal–
hydrogen interactions in the hydrides. The compound has
Fig. 5. Metal coordinations of deuterium sites in hexagonal LaNi2Mn3Dx
(x54.6, 5.6). For clarity, only one representative of each site is drawn.
Site symmetries: D1, mm2; D2, mmm; D3, m2m; D4, .m.; D5, m.; large
atoms, La; small dark grey atoms, Ni; medium light grey atoms, Mn1;
medium dark grey atoms, Mn2.

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
182
Table 3
Metal coordinations, coordination numbers (CN), and interatomic distancesa in LaNi2Mn3D4.6, LaNi2Mn3D5.6 and LaNi2Mn3; e.s.d.s in parentheses
Central
atoms
CN
Coordinated
LaNi2Mn3D4.6
LaNi2Mn3D5.6
LaNi2Mn3
a59.2099(9) A
c54.1892(4) A
˚
˚
˚
˚
a59.7125(2) A
a59.8379(2) A
˚
˚
c54.3057(1) A
c54.3038(1) A
distances
(A)
distances
(A)
distances
(A)
˚
˚
˚
La1 (1a)
La2 (2d)
18
18
12Mn1
6Ni1
3.544(4)
3.097(2)
3.540(5)
3.147(3)
3.456(1)
2.897(1)
6Mn1
6Ni1
6Mn2
3.467(3)
3.307(1)
3.5347(1)
3.538(6)
3.323(2)
3.5629(1)
3.242(1)
3.201(1)
3.3844(2)
Mn1 (6k)
Ni1 (6l)
Mn2 (3f )
12
11
12
2 Mn1
4 Ni1
2 Mn2
2 La2
2 La1
2.816(5)
2.655(1)
2.966(3)
3.467(3)
3.544(4)
2.810(5)
2.667(2)
3.013(5)
3.537(4)
3.539(6)
2.749(2)
2.558(1)
2.798(1)
3.242(1)
3.456(1)
4 Mn1
2 Mn2
1 La1
2 La2
2 Ni1
2.655(1)
2.669(2)
3.097(2)
3.307(1)
3.097(2)
2.667(4)
2.700(3)
3.147(3)
3.324(2)
3.147(4)
2.558(1)
2.548(1)
2.897(1)
3.201(1)
2.897(1)
4 Mn1
4 Ni1
4 La2
2.966(3)
2.669(2)
3.5347(1)
3.013(5)
2.700(3)
3.5629(1)
2.798(1)
2.548(1)
3.384(1)
D1 (6m)
D2 (3g)
4
6
2La1
2Mn1
2.470(4)
1.871(9)
2.442(6)
1.90(2)
2La2
2Mn1
2Mn2
2.803(3)
2.037(4)
2.153(1)
2.841(1)
2.109(6)
2.152(1)
D3 (6j)
5
4
4
1La1
2Mn1
2Ni1
2.486(6)
2.178(1)
1.561(2)
2.501(6)
2.174(1)
1.590(3)
D4 (12o)
1La2
2Mn1
1Ni1
2.316(5)
1.772(6)
1.616(9)
2.322(5)
1.835(5)
1.563(8)
D5 (12p)
2La2
1Mn2
1Ni1
2.476(2)
1.841(5)
1.584(5)
2.483(3)
1.878(6)
1.581(6)
a
˚
D–Ni(Mn) cut-off distance ,2.7(2.4) A.
Table 4
Shortest distances between D atom sites and interstitial hole size (Rsite) in LaNi2Mn3D4.6 and LaNi2Mn3D5.6; e.s.d.s in parentheses
Atom (site)
LaNi2Mn3D4.6
Neighbours
LaNi2Mn3D5.6
Neighbours
a
a
˚
˚
Distances
(A)
Rsite (A)
Distances
(A)
Rsite (A)
˚
˚
D1 (6m)
2 D1
2 D1
8 D4
4 D4
1 D4
1.21(1)
2.10(1)
2.632(5)
2.353(7)
1.12(1)
0.58
2 D1
2 D1
8 D4
4 D4
1 D4
2 D5
1 D5
1 D5
1 D4
1.15(2)
2.00(2)
2.665(5)
2.355(7)
1.23(1)
2.02(1)
1.23(1)
1.25(1)
2.02(1)
0.58
D2 (3g)
D3 (6j)
D4 (12o)
0.74
0.36
0.44
0.81
0.39
0.46
D5 (12p)
1 D5
1 D5
1.20(2)
1.24(2)
0.54
0.55
a Calculated from atomic radii RLa 51.877 A, RMn 51.304 A, RNi 51.246 A [17] within the hard sphere approximation.
˚
˚
˚

´ ´
L. Guenee, K. Yvon / Journal of Alloys and Compounds 348 (2003) 176–183
183
´ ´
[4] L. Guenee, PhD Thesis No. 3341, University of Geneva (2002).
[5] S.R. Hall, D.J. du Boulay, R. Olthof-Hazekamp, Eds. Xtal3.7
System. University of Western Australia (2000).
relatively good hydrogen storage efficiencies and fast
kinetics and thus is useful for hydrogen storage applica-
tions. Its discovery underlines the usefulness of modelling
the structural stability of intermetallic compounds for the
search of new hydrogen storage materials.
[6] J.L.C. Daams, P. Villars, J.H.N. Van Vucht, in: Atlas of Crystal
Structure Types For Intermetallic Phases, ASM International, Ma-
terials Park, OH 44073, 1991.
[7] V. Favre-Nicolin, R. Cerny, J. Appl. Cryst. (2002) submitted, see
also http://objcryst.sourceforge.net
´
[8] J. Rodrıguez-Carvajal, in: Abstract of the Satellite Meeting on
Acknowledgements
Powder Diffraction, Congress of the International Union of Crys-
tallography, Toulouse, France, 1990, p. 127. Fullprof Program,
version 3.5d Oct 98-LLB-JRC, 1998.
The authors thank the staff of the Swiss Norwegian
beamline at ESRF (Grenoble), C. Ritter (D1A, ILL,
Grenoble) and D. Sheptyakov (HRPT, PSI, Villigen) for
help with collecting the diffraction patterns. Thanks are
also due to R. Cerny (Geneva) and V. Favre-Nicolin
(ESRF, Grenoble) for discussions and help with the newly
developed computer program FOX. This work was sup-
ported by the Swiss National Science Foundation and the
Swiss Federal Office of Energy.
´
[9] R. Cerny, J.-M. Joubert, M. Latroche, A. Percheron-Guegan, K.
Yvon, J. Appl. Crystallogr. 33 (2000) 997.
[10] M.A. Fremy, D. Gignoux, J. Less-Common Met. 106 (1985) 251–
255.
[11] K. Yvon, P. Fischer, Topics in applied physics, in: L. Schlapbach
(Ed.), Hydrogen in Intermetallic Compounds I, Vol. 63, Springer,
New York, 1988, p. 87.
[12] D.G. Westlake, J. Less-Common Met. 90 (1983) 251.
[13] J.-L. Bobet, S. Pechev, B. Chevalier, B. Darriet, J. Alloys Comp.
267 (1998) 136–141.
[14] S. Pechev, J.-L. Bobet, B. Chevalier, B. Darriet, F. Weill, J. Solid
State Chem. 150 (2000) 62–71.
[15] Z. Blazina, B. Sorgic, A. Drasner, J. Phys.: Condens. Matter 11
(1999) 3105–3114.
References
[16] G.I. Miletic, A. Drasner, Z. Blazina, J. Alloys Comp. 296 (2000)
170–174.
[17] E. Teatum, K. Gschneider, J. Waber (1960), LA-2345, US Depart-
ment of Commerce, Washington, DC. (cf. book: W.B. Pearson, The
Crystal Chemistry and Physics of Metals and Alloys, p. 151).
´
[1] F. Cuevas, J.-M. Joubert, M. Latroche, A. Percheron-Guegan, Appl.
Phys. A 72 (2001) 225–238.
[2] C. Lartigue, A. Percheron-Guegan, J.C. Achard, F. Tasset, Rare
earth in modern science and technology, Rare Earths Res. Conf. 2
(1980) 585–591.
[3] Ya.M. Kal’chak, O.I. Bodak, E.I. Gladyshevskiy, Russian Metal-
lurgy, Translated from Izvestiya Akademii Nauk SSSR, Metaly (5)
(1980) 217–221.