NMR studies of the aluminum hydride phases and their stabilities

Article abstract of DOI:10.1016/j.jallcom.2007.01.115

Multinuclear and multidimensional solid state NMR techniques including magic-angle-spinning (MAS) and multiple-quantum (MQ) MAS experiments have been used to characterize various AlH₃ samples. At least three distinct polymorphic AlH₃ phases have been prepared by desolvating the alane etherate product from its organometallic synthesis. MAS-NMR spectra for the ¹H and ²⁷Al nuclei have been obtained on a variety of AlH₃ samples that include the β- and γ-phases as well as the α-phase. ²⁷Al MAS NMR was found to respond with high sensitivity for showing differences in spatial arrangements of AlH₆ octahedra in the three polymorphs studied. Based on the characteristic NMR signatures determined, phase transition of the γ-AlH₃ to the α-AlH₃ was studied at room and high temperatures. Direct decomposition of the γ-AlH₃ to aluminum metal at room temperature was also unambiguously confirmed by NMR studies.

Full text of DOI:10.1016/j.jallcom.2007.01.115

Journal of Alloys and Compounds 446–447 (2007) 290–295
NMR studies of the aluminum hydride phases and their stabilities
a,∗
b
c
Son-Jong Hwang , R.C. Bowman Jr. , Jason Graetz ,
c
d
d
J.J. Reilly , W. Langley , C.M. Jensen
a
The Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA 91125, USA
b
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA
c
Brookhaven National Laboratory, Department of Energy Sciences and Technology, Upton, NY 11973, USA
d
Department of Chemistry, University of Hawaii, Honolulu, HI 96822, USA
Received 27 October 2006; received in revised form 11 January 2007; accepted 11 January 2007
Available online 28 January 2007
Abstract
Multinuclear and multidimensional solid state NMR techniques including magic-angle-spinning (MAS) and multiple-quantum (MQ) MAS
experiments have been used to characterize various AlH samples. At least three distinct polymorphic AlH phases have been prepared by
desolvating the alane etherate product from its organometallic synthesis. MAS-NMR spectra for the H and Al nuclei have been obtained on
a variety of AlH
samples that include the ␤- and ␥-phases as well as the ␣-phase. 27Al MAS NMR was found to respond with high sensitivity
for showing differences in spatial arrangements of AlH octahedra in the three polymorphs studied. Based on the characteristic NMR signatures
determined, phase transition of the ␥-AlH to the ␣-AlH was studied at room and high temperatures. Direct decomposition of the ␥-AlH to
aluminum metal at room temperature was also unambiguously confirmed by NMR studies.
2007 Elsevier B.V. All rights reserved.
3
3
27
1
3
6
3
3
3
©
Keywords: Hydrogen storage materials; Gas–solid reactions; Nuclear resonances
1
. Introduction
by Yartys et al. [12]. Since the original organometallic synthesis
of ␣ and 6 other polymorphs by Brower et al. [13], the ␣-AlH3
phase was, until recently, the only phase with a known complete
crystalstructure[14], implicatingtheremarkableadvancesbeing
made in short period of time.
Aluminum hydride (AlH3, alane) is a metastable solid
with large hydrogen gravimetric (∼10 wt%) and volumetric
(
0.148 kg H2/L) capacities. Because of its high potential as a
hydrogen fuel carrier for automotive applications, there have
recently been considerable interests in understanding and devel-
opmentofthematerial. Recently, variousalanephases(e.g., ␣, ␤,
and ␥) were freshly prepared, characterized [1–4], and thermo-
dynamics were determined [2,5]. Enhanced kinetics in thermal
decomposition of AlH3 was observed by Sandrock et al. [6,7]
when dopants (e.g., LiH, Ti) were incorporated or particle sizes
were varied. More importantly a number of structural studies of
alane phases have appeared. There was a report on NMR char-
acterization of three different AlH3 phases by Hwang et al. [8]
identifying distinctive Al sites depending on polymorphs, and
In this work, we present a brief overview of high resolution
NMR studies on the structures and thermal stabilities of AlH3
polymorphs (␣, ␤, and ␥ phases). In the literature, solid state
NMR studies of pure AlH3 material were reported by Zogal et al.
[15,16] who used ␣-AlH3 material originally produced at Dow
Chemical to measure 27Al quadrupolar coupling constant and
observe molecular hydrogen (H2) that they attributed to being
adsorbed on the hydride particle surfaces from a sharp line of
1
H static NMR seen at temperature as low as 10 K. Tarasov et
al. [4] more recently used 27Al static NMR to explore formation
of Al metal [Al(M)] during thermal decomposition of AlH3 and
AlD3 reporting the kinetics for temperatures between 360 K and
400 K, and giving an activation energy. Although there is a lack
of high resolution studies of pure AlH3 except our recent work
ꢀ
very recently crystal structures of ␣ , ␤, and ␥ phases in deu-
teride forms (AlD3) were reported by Brinks et al. [9–11] and
[8], use of high resolution NMR methods has been successful in
∗
addressing structures and chemical transformation of complex
metal hydride systems [17–19].
Corresponding author. Tel.: +1 626 395 2323; fax: +1 626 568 8743.
E-mail address: sonjong@cheme.caltech.edu (S.-J. Hwang).
0
925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2007.01.115

S.-J. Hwang et al. / Journal of Alloys and Compounds 446–447 (2007) 290–295
291
Table 1
general synthesis methods and procedures described by Brower et al. [13] were
followed to make all alane materials as reported previously [1,2]. Powder X-ray
diffraction was used to identify the initial phase compositions of the samples
characterized by NMR. The Dow material was primarily the ␣-phase alane with
8.3 wt% hydrogen content that was virtually unchanged after ∼30 years storage
in air at ambient conditions. All initial NMR sample handling was performed in
a glove box under an argon atmosphere. Later samples were exposed to air on
purpose and examined for the build-up of oxide layers [22].
AlH3 samples used for the NMR experiments
Sample description
ID code
␣-AlH3
␣-AlH3
␣-AlH3
BNL-H2963
Dow
UTRC
BNL-H2990, BNL-H3013
UH-3
␥
␥
-AlH3
-AlD3
Solid state MAS NMR spectra were recorded using a Bruker Avance
␣
, ␤, ␥-mixed
BNL-H2961
500 MHz spectrometer with a wide bore 11.7 T magnet and employing a Bruker
◦
4
mm CPMAS probe with extended variable temperature capability up to 300 C.
1
The spectral frequencies were 500.23 MHz for H nucleus and 130.35 MHz for
2
7
The magic angle spinning (MAS) NMR of AlH3 for 27Al
Al nucleus. NMR shifts were reported in parts per million (ppm) when exter-
nally referenced to tetramethylsilane (TMS) and 1.0 M of Al(NO3)3 aqueous
1
nucleus (I = 5/2) with high power H decoupling in general
yields well resolved 27Al signal of which isotropic chemical
shift (δiso) and quadrupolar coupling parameters (CQ, η, where
CQ is the quadrupolar coupling constant and η is the asym-
metry parameter) become distinctive signatures for different
polymorphs of AlH3 [8]. The fact that Al is positioned in
the center of a symmetric AlH octahedron (for all three ␣,
␤
quadrupolar coupling and consequently less broadened favor-
able NMR lineshapes. When more broadening is introduced
due to higher quadrupolar effect, which is the case for the
1
27
solution at 0 ppm for H and Al nuclei, respectively. The AlH3 powder sam-
ples were packed into a 4 mm ZrO2 rotor after minimal additional grinding and
were sealed with a tight fitting kel-F cap [8]. Sample spinning was performed
2
7
using dry nitrogen gas. A typical Al MAS NMR spectrum was obtained at a
sample spinning rate of 14.5 kHz and after a 0.3 ␮s single pulse (<π/18) with
1
application of strong H decoupling pulse of the two-pulse phase modulation
(
TPPM) scheme [23]. In situ variable temperature (VT) MAS NMR experi-
6
ments on AlH3 were performed at 8 kHz spinning rate using a ZrO2 cap that
allowed leaking of hydrogen gas desorbed at high temperatures. The MQ MAS
experiment was performed using the z-filter method [24] at a spinning speed
of 14 kHz and a two-dimensional (2D) spectrum was presented after a shearing
transformation [21].
, ␥ polymorphs) [9,10,13,14] results in weak contribution of
␥
-AlH3, two-dimensional (2D) multiple quantum (MQ) MAS
technique [20,21] can be used to obtain spectra with improved
resolution. Overall, NMR is sensitive enough to distinguish
subtle differences originated from variations in arrangement of
3. Results and discussion
3.1. 27Al NMR of AlH phases
3
AlH octahedra in space: i.e., corner shared (␣, ␤) versus edge
shared (␥). In addition, NMR was proven to provide separate sig-
6
The characteristic 27Al NMR features are shown in Fig. 1
nals for corner shared octahedra packed in different space groups
(
for three different AlH phases. Pure ␣-AlH phases show rel-
atively well resolved static powder pattern arising from both
3
3
␣ versus ␤ and one of the ␥ sites) [8]. Once these 27Al spectral
fingerprints are determined, NMR becomes a straightforward
tool in exploring phase changes during hydrogen desorption pro-
cesses of AlH3 both at room and high temperatures. It is the aim
of the current work to demonstrate how effectively NMR can
be employed to study structure changes and kinetics during the
thermal decomposition of both the old Dow and freshly made
AlH3 samples with differing phase compositions.
satellite (m = ± 3/2 ↔ m ± l/2 and m = ± 5/2 ↔ m = ± 3/2) and
27
central transitions (m = −l/2 ↔ m = +l/2) of Al (I = 5/2) nuclei
at around ∼0 ppm, representing six-coordinated Al in AlH
6
octahedron units which are all connected by corner-sharing in
the 3D chain [14]. Its MAS spectrum under 14.5 kHz spin-
ning significantly improves the resolution to yield a single line
∼850 Hz wide at 5.5 ppm, which is ∼5 times sharper than he
static linewidth (4.5 kHz). In addition to the centerband, several
spinning sidebands are also present with their intensity distribu-
tion expressing the shape of the static powder pattern. Analysis
of both spectra using the QUASAR software [25] allowed us to
extract NMR parameters (see Table 2). As expected, Al sitting in
a symmetric octahedron experiences very little electric field gra-
dient (efg) as indicated by a small CQ value (∼250 kHz). Three
2
. Experimental description
The alane (i.e., AlH3) materials are from four sources: (1) originally manu-
factured by the Dow Chemical Company in the 1970s, (2) recently prepared at
Brookhaven National Laboratory (BNL), (3) freshly prepared at University of
Hawaii (UH), and passivated ␣-AlH3 from an unspecified Russian source pro-
vided by United Technologies Research Center (UTRC) as listed in Table 1. The
Table 2
The quadrupole coupling parameters for 27Al spectra in AlH3 phases
Sample
Code
δiso (ppm)
CQ (MHz)a
η
Relative conc.
1.0
␣-AlH3
BNL-H2963
BNL-H2990
5.8
␥-I: 10.9
0.25
2.9
4.0
0.38c
0.1
0.58
0.02
b
␥-AlH3
1.0
0.48
b
␥
-II: 36.0
c
c
␣
+ ␤ + ␥-AlH3
BNL-H2961
21.5
0.49 (0.14)
␣:␤:␥-I:␥-II = 0.22:0.36:0.31:0.11
a
The contribution of chemical shift anisotropy (ꢀδCSA) was not taken into account in the current work. The estimated quadrupolar parameters need further
refinements to provide more accurate values, which is currently in progress. For reference, a rough estimation [17] indicated that 2 Al NMR could show ꢀδCSA as
7
large as 25 (± 10) ppm for Li3AlH6 octahedron.
b
The intensity ratio was reported accidentally the opposite way in our previous report [8].
The parameters are for the ␤-AlH3 phase.
c

2
92
S.-J. Hwang et al. / Journal of Alloys and Compounds 446–447 (2007) 290–295
Fig. 1. 27Al NMR spectra and simulations: (a) static and under MAS at 14.5 kHz of ␣-AlH3, (b) ␥-AlH3 under MAS at 14 kHz for a wide manifold of spinning
sidebands and the centerband, (c) characterization of ␤-AlH3, and (d) A 2D MQMAS spectrum of mixed phase (BNL-H2961).
different ␣-AlH3 materials (see Table 1: Dow, BNL, UTRC)
were examined with NMR in the present work. Although the
centerband of 2 Al MAS signal are nearly discernable among
three samples, the 27Al static NMR powder patterns (not shown)
and the corresponding spinning sideband pattern in MAS spectra
showed noticeable differences that are associated with variations
in crystallinity, the amount of surface oxide layers (Al2O3), and
crystal sizes of materials.
with the significantly larger quadrupolar coupling constants (see
Table 2) compared to that of AlH octahedra in the ␣-phase. The
6
7
presence of two crystallographically different AlH sites with
6
2:1 ratio was also revealed by recent crystal structure determina-
tions of the ␥-AlH3 (or ␥-AlD3) by Brinks et al. [10] and Yartys
et al. [12]. The fact that two AlH octahedral sites are distorted
6
in bond angles (Al–H–Al) and distances due to the nature of spa-
tial arrangement of corner-sharing (Al(I)H ) and edge-sharing
6
The NMR characterization of ␥-AlH3 phase was performed
using the same approach on a BNL material with relatively high
purity for the ␥ phase. Later, NMR analysis indicated that the ␥-
AlH3 material also contained ∼9 wt% of the ␣-AlH3 phase and
(Al(II)H ) octahedra in the ␥ phase is completely consistent with
6
higher values in the quadrupolar interactions of NMR spectra.
These associations between our NMR parameters with the crys-
tal structures have allowed us to assign the two sites based on
the site occupancy, and the assignment is listed in Table 2. The
∼
4 wt% of LiAlH4 with negligible amount of Al(M). Unlike the
phase, Fig. 1(b) shows two sets of wide manifold of spinning
␣
corner-sharing Al(I)H octahedra, ␥-II site, show well-defined
6
sidebands distributed over 6000 ppm. This is a typical NMR
signature of quadrupole nucleus under moderate quadrupolar
coupling and with high crystallinity. Spectral fitting using the
QUASAR (see simulated manifold of spinning sidebands for
powder pattern of the central transition (m = +1/2 ↔ m = −l/2)
in the centerband (see Fig. 1(b)) due to the so called second
order quadrupolar broadening. It is surprising since all corner-
shared AlH octahedra in the ␣ and ␤ phases show narrow and
6
␥
-I and ␥-II) disclosed the presence of two different sites (see
symmetric lineshapes. This is better understood now thanks to
crystal structure studies showing that the corner-sharing Al(I)H6
the centerband in Fig. 1(b)) with intensity ratio ␥-I:␥-II = 2:1 and

S.-J. Hwang et al. / Journal of Alloys and Compounds 446–447 (2007) 290–295
293
octahedra in the ␥ phase is highly distorted because of bonding
MQMAS spectrum of the mixture phase with improved resolu-
tion, supporting our deconvolution of the one-dimensional (1D)
spectrum. Note that the powder pattern seen for 1D MAS spec-
trum of the highly distorted ␥-II site appears to be free from the
second order broadening in the isotropic axis of 2D MQ MAS
spectrum.
to edge-sharing Al(II)H in two different ways both in the plane
6
and axial positions of octahedra and containing a large Al–H–Al
bond angle [10,12].
For the ␤-AlH3 phase, 27Al NMR determination was per-
formed using a material (i.e., BNL-H2961) crystallized with
mixture of three different phases, which were identified by their
27
X-ray diffraction peaks. Fitting of the experimental Al MAS
spectrum employing previously determined NMR parameters
for both ␣ and ␥ phases allowed us to separate out a spectral
component that is quite similar to that of the ␣ phase with an
exception of a downfield shift in the isotropic chemical shift
4
. Decomposition studies of AlH using NMR
3
2
1
7
Al NMR characterization in the previous section as well
2
as H and H MAS NMR of AlH3/AlD3 phases plays a valu-
able role in studying chemical transformation of various AlH3
2
7
(
see the deconvolution in Fig. 1(c) and Table 2). According to
phases: AlH3 ↔ Al(M) + 3/2H2. Note that unlike the Al NMR,
1
2
Brinks et al. [9], the ␤-AlH3 phase consists of all corner-shared
H or H MAS NMR in the solid AlH3/AlD3 phase did not pro-
vide distinctive signatures from different crystallographic sites.
AlH octahedra with a single crystallographic Al site and with
6
1
2
similar structural parameters to those of the ␣-phase, consistent
with our NMR result. The change in the isotropic shift can be
attributed to the difference in the connectivity of octahedra. The
mixture phase (␣ + ␤ + ␥) showed in addition a relatively strong
initial Al(M) peak (not shown) even for freshly prepared sam-
ples compared to the others, implying that degradation occurs
at room temperature much more rapidly. Fig. 1(d) shows a 2D
However, growth of sharp peaks at4.4 ppm in H or HNMR can
provide an easy check of desorption of hydrogen or deuterium
molecules from the solid phase [8]. The resulting signature in
27
Al NMR is a decrease of intensity for peaks from AlH3 phases
with simultaneous growth of the Al(M) peak at 1640 ppm. The
measure of peak intensities in the 27Al MAS NMR spectra in
turn gives rise to a handy in situ kinetics study. Here in the
Fig. 2. (a) 27Al MAS NMR spectra of ␥-AlH3 phase undergoing self-decomposition at room temperature. (b) Plot of fraction (φ) of ␥-AlH3, Al(M), and ␣-AlH3
2
7
during the decomposition of the ␥-AlH3 sample, and plot of [−ln(l − α)] vs. t, where α is the fractional decomposition of ␥-AlH3. (c) Al MAS NMR spectra
showing the 2nd and 3rd spinning sidebands at t = l h and 192 h, those of the ␣-AlH3, and spectrum after subtraction of the ␣-AlH3 from the spectrum at t = 192 h.
2
7
(
d) Al MAS NMR spectra obtained for decomposition of the ␥-AlH3 phase at high temperatures.

2
94
S.-J. Hwang et al. / Journal of Alloys and Compounds 446–447 (2007) 290–295
present work, we focus on showing a decomposition reaction
of the ␥-AlH3 phase at room temperature. The most important
issue in this study is to elucidate the decomposition pathway in
addition to obtaining the rate, i.e., to determine possible phase
transition from the ␥ phase to the ␣ phase at room temperature,
and possible direct decomposition of the ␥ phase.
position of the ␥-AlH3 phase to form Al metal also takes place at
room temperature. Finally, note that detail analyses of spectra in
Fig. 2(a) revealed that the decomposition of the ␥-AlH3 led the
formation of both ␣-AlH3 and Al metal at about equal amount
as plotted in Fig. 2(b).
Fig. 2(a) shows a series of 27Al NMR spectra recorded
at room temperature over the decomposition time period for
the centerband (−20 to 50 ppm) and for Al(M) at 1640 ppm.
From the integrated area of the ␥-AlH3 (␥-I and ␥-II sites,
vide supra) and forming Al(M) and ␣-AlH3, mole fraction (φ)
of each species was calculated and plotted as a function of
time in Fig. 2(b). The decomposition kinetics of the ␥-AlH3
was analyzed by using the Avrami–Erofeev equation consid-
ering the reaction as an irreversible isothermal decomposition,
as employed by Graetz et al. [1,2] who recently reported high
temperature decomposition of ␣, ␤, ␥-AlH3 phases. A plot of
ln[−ln(l – α)] versus ln(t), with α being the fractional decom-
position, in the current study yielded a value of n ∼ l, and
5. Conclusions
Static and MAS NMR characterization was performed on
27
various AlH samples. Al MAS NMR demonstrated high
3
sensitivity in distinguishing local environments of aluminum
nucleus depending on the type of connectivity of AlH octahe-
6
dra and their spatial arrangement. For example, NMR responded
with well-resolved lines for corner-sharing and edge sharing
AlH octahedra in the ␥-AlH polymorph. Once characterized,
6
3
NMR spectra were useful in the study of the decomposition
kinetics and phase transitions taking place in these materials.
The reaction rate at room temperature was obtained for unstable
␥ phase undergoing decomposition. The rate-limiting mecha-
nism appears to be the nucleation and growth model with the
dimensionality of one at room temperature. A phase transition
to the ␣ phase was also confirmed by analyzing spinning side-
band patterns. NMR evidence for direct decomposition of the ␥
phase at room temperature was also obtained.
[−ln(l − α)] versus t produced an acceptable liner plot as shown
in Fig. 2(b). The result indicates that the rate determining step
might be considered in the nucleation and growth model [26],
and the growth dimension in this case could be one dimension
unlike the previous cases found at high temperatures [1,2]. It
is quite interesting because this finding could support that the
gamma phase, which has a clear one-dimensional morphology
in SEM images [27], actually decompose and nucleate Al in
one dimension at low temperatures. The rate constant k obtained
Acknowledgements
This research was partially performed at the Jet Propul-
sion Laboratory, which is operated by the California Institute
of Technology under contract with the NASA. This work was
partially supported by DOE through Award Number DE-AI-
−
◦
7
−1
from [−ln(l − α)] versus t plot was 5.8 (± 0.2) × l0
s , which
is an order of magnitude smaller than that at 60 C [1,2]. The
␥
-AlH3 material (BNL-3013) used in this study was found
01-05EE11105. The NMR facility at Caltech was supported
to contain a similar level of contaminants to our previous ␥-
AlH3 material (H2990, see above), including ␣-AlH3 (∼7%).
by the National Science Foundation (NSF) under Grant Num-
ber 9724240 and partially supported by the MRSEC Program
of the NSF under Award Number DMR-0080065. This work
was supported at BNL by the DOE under contract DEA-
AC0298CH100886. Financial support received from the Office
of Hydrogen, Fuel Cells and Infrastructure Technologies of US
Department of Energy is gratefully acknowledged. We thank
J.G. Kulleck (JPL) for his X-ray diffraction measurements and
Dr. X. Tang from the United Technologies Research Center for
providing the Russian made ␣-AlH3 material labeled UTRC.
Finally, the authors are very grateful to Drs. H. Brinks and V.A.
Yartys for providing their detailed crystal structure results to us
prior to publication.
Recording of 27Al NMR spectra was terminated when hydro-
gen pressure build-up led opening of the tight-sealing kel-F cap
from the rotor.
In order to observe a phase change to the ␣-AlH3 phase from
the starting ␥-AlH3 phase during room temperature decompo-
sition, spinning sidebands were carefully examined because the
centerband of 27Al resonance cannot distinguish ␣-AlH3 from
the ␥-I site as shown in Fig. 2(a). The 2nd and 3rd spinning side-
bands (200–400 ppm) of two spectra in Fig. 2(a) are separately
displayed in Fig. 2(c) to show the growth of peaks marked by
arrows after 192 h at room temperature. Their positions are pre-
cisely matched with sideband positions of the ␣-AlH3 spectrum
shown in Fig. 2(c) (broken line). This result clearly indicates that
the phase changes from the ␥-AlH3 to the ␣-AlH3 takes places
even at room temperature. The bottom spectrum of Fig. 2(c)
was generated by subtracting the ␣-AlH3 component from the
spectrum of 192 h. The subtracted spectrum shows almost iden-
tical lineshapes to the top spectrum of 1 h as anticipated with
an exception of the reduced intensity. Such phase transition can
be easily visualized by NMR at high temperatures as shown in
Fig. 2(d), where spinning sidebands manifold of the ␥-AlH3 is
greatly reduced to form sidebands manifold of the ␣-AlH3 phase
at 373 K. Considering the stable nature of the ␣ phase, results
presented in Fig. 2 strongly support the view that direct decom-
References
[
[
1] J. Graetz, J.J. Reilly, J. Phys. Chem. B 109 (2005) 22181.
2] J. Graetz, J. Reilly, G. Sandrock, J. Johnson, W.M. Zhou, J. Wegrzyn, in: D.
Chandra, J.J. Petrovic, R. Bautista, A. Imam (Eds.), Advanced Materials
for Energy Conversion III, The Minerals, Metals, and Materials Society,
Warrendale, PA, 2006, pp. 57–63.
[
3] J.P. Maehlen, V.A. Yartys, in: D. Chandra, J.J. Petrovic, R. Bautista, A.
Imam (Eds.), Advanced Materials for Energy Conversion III, The Minerals,
Metals, and Materials Society, Warrendale, PA, 2006, pp. 77–85.
4] V.P. Tarasov, Y.B. Muravlev, S.I. Bakum, A.V. Novikov, Doklady Phys.
Chem. 393 (2003) 353–356.
[
[5] J. Graetz, J.J. Reilly, J. Alloys Compd. 424 (2006) 262.

S.-J. Hwang et al. / Journal of Alloys and Compounds 446–447 (2007) 290–295
295
[
[
[
6] G. Sandrock, J. Reilly, J. Graetz, W. Zhou, J. Johnson, J. Wegrzyn, Appl.
Phys. A 80 (2005) 687.
7] G. Sandrock, J. Reilly, J. Graetz, W. Zhou, J. Johnson, J. Wegrzyn, J. Alloys
Compd. 142 (2006) 185–189.
8] S.-J. Hwang, R.C. Bowman, J. Graetz, J.J. Reilly, in: J.C.F. Wang, W.
Tumas, A. Rougier, M.J. Heben, E. Akiba (Eds.), Hydrogen Storage Mate-
rials, Mater. Res. Soc. Symp. Proc., vol. 927E, Warrendale, PA, 2006, pp.
[17] J.W. Wiench, V.P. Balema, V.K. Pecharsky, M. Pruski, J. Solid State Chem.
177 (2004) 648.
[18] B. Bogdanovi c´ , M. Felderhoff, M. Germann, M. H a¨ rtel, A. Pommerin,
F. Schuth, C. Weidenthaler, B. Zibrowius, J. Alloys Compd. 350 (2003)
246.
[19] J.L. Herberg, R.S. Maxwell, E.H. Majzoub, J. Alloys Compd. 417 (2006)
39.
0
927-EE03-03.
[20] L. Frydman, J.S. Harwood, J. Am. Chem. Soc. 117 (1995) 5367.
[21] J.P. Amoureus, M. Pruski, in: D.M. Grant, R.K. Harris (Eds.), Encyclopedia
of NMR, 9, John Wiley & Sons, Chichester, 2002, pp. 226–251.
[22] For all the ␣-AlH3 phase studied in the present work, the formation of
surface oxide layer was too slow to be detected by NMR.
[23] A.E. Bennett, C.M. Rienstra, M. Auger, K.V. Lakshmi, R.G. Griffin, J.
Chem. Phys. 103 (1995) 6951.
[
9] H.W. Brinks, W. Langley, C.M. Jensen, J. Graetz, J.J. Reily, B.C. Hauback,
J. Alloys Compd. 433 (2007) 180.
10] H.W. Brinks, W. Langley, C.M. Jensen, J. Graetz, J.J. Reily, B.C. Hauback,
unpublished.
[
[
[
[
11] H.W. Brinks, A. Istad-Lem, B.C. Hauback, J. Phys. Chem. 110 (2006)
2
5833.
12] V.A. Yartys, R.V. Denys, J.P. Maehlen, C.h. Frommen, M. Fichtner, B.M.
Bulychev, H. Emerich, Inorg. Chem. 46 (2007) 1051.
13] F.M. Brower, N.E. Matzek, P.F. Reigler, H.W. Rinn, C.B. Roberts, D.L.
Schmidt, J.A. Snover, K. Terada, J. Am. Chem. Soc. 98 (1976) 2450.
14] J.W. Turley, H.W. Rinn, Inorg. Chem. 8 (1968) 18.
[24] J.P. Amoureux, C. Fernandez, Solid State NMR 10 (1998) 211.
[25] J.P. Amoureux, C. Fernandez, L. Carpentier, E. Cochon, Phys. Stat. Sol.
(a) 132 (1992) 461.
[26] C.H. Bamford, C.F.H. Tipper (Eds.), Comprehensive Chemical Kinetics,
vol. 22, Elsevier, New York, 1980 (Chapter 3).
[
[
[
15] O.J. Zogal, B. Stalinski, S. Idziak, Z. Physk. Chem. N.F. 145 (1985) 167.
16] O.J. Zogal, M. Punkkinen, E.E. Ylinen, B. Stalinski, J. Phys.: Condens.
Matter 2 (1990) (1941).
[27] G.K. Lund, M.W. Lesley, S.R. Glaitli, Proceedings of the JANNF-32nd
PEDCS Meeting, 2004, p. 1.