Stability of lithium hydride in argon and air

Article abstract of DOI:10.1021/jp060068m

The oxidation behaviors of LiH under a high purity argon atmosphere, an argon atmosphere with some O₂ and H₂O impurities, and ambient air at both room and high temperatures, are investigated using a variety of analytical instruments including X-ray diffractometry, thermogravimetry, mass spectrometry, scanning electron microscopy, and specific surface area analysis. The oxidation behaviors of the ball-milled LiH under different atmospheres are also studied and compared with those without ball milling. It is shown that no oxidation of LiH occurs under a high-purity argon atmosphere. However, oxidation of LiH takes place when the argon atmosphere contains some H₂O and O₂ impurities. At temperatures higher than ～55 °C, oxidation of LiH proceeds via the reaction of LiH + 1/4 O₂ = 1/2 Li₂O + 1/2 H₂, whereas at room temperature oxidation of LiH is likely caused by the simultaneous reactions of LiH + H₂O = LiOH + H₂ and LiH + 1/2 O₂ = LiOH. The oxidation behavior of LiH in ambient air with a 27% relative humidity can be well described by the Johnson-Mehl - Avrami equation. Furthermore, the ball-milled LiH oxidizes faster than the unmilled one, which is due to the finer particle size and larger surface area of the ball-milled powder.

Full text of DOI:10.1021/jp060068m

J. Phys. Chem. B 2006, 110, 10567-10575
10567
Stability of Lithium Hydride in Argon and Air
Ruiming Ren,†,‡ Angel L. Ortiz,§ Tippawan Markmaitree,† William Osborn,† and
Leon L. Shaw*,†
Department of Materials Science and Engineering, UniVersity of Connecticut, Storrs, CT 06269-3136, School of
Materials Science and Engineering, Dalian Jiaotong UniVersity, Dalian, China, and Departamento de
Electro´nica e Ingenier´ıa Electromeca´nica, UniVersidad de Extremadura, Badajoz, Spain.
ReceiVed: January 4, 2006; In Final Form: April 7, 2006
The oxidation behaviors of LiH under a high purity argon atmosphere, an argon atmosphere with some O2
and H2O impurities, and ambient air at both room and high temperatures, are investigated using a variety of
analytical instruments including X-ray diffractometry, thermogravimetry, mass spectrometry, scanning electron
microscopy, and specific surface area analysis. The oxidation behaviors of the ball-milled LiH under different
atmospheres are also studied and compared with those without ball milling. It is shown that no oxidation of
LiH occurs under a high-purity argon atmosphere. However, oxidation of LiH takes place when the argon
atmosphere contains some H2O and O2 impurities. At temperatures higher than ∼55 °C, oxidation of LiH
proceeds via the reaction of LiH + 1/4 O2 ) 1/2 Li2O + 1/2 H2, whereas at room temperature oxidation of LiH
is likely caused by the simultaneous reactions of LiH + H2O ) LiOH + H2 and LiH + 1/2 O2 ) LiOH. The
oxidation behavior of LiH in ambient air with a 27% relative humidity can be well described by the Johnson-
Mehl-Avrami equation. Furthermore, the ball-milled LiH oxidizes faster than the unmilled one, which is
due to the finer particle size and larger surface area of the ball-milled powder.
I. Introduction
of microseconds.4 As such, escaping of NH3 from the hydrogen
storage system is prevented. However, if LiH is oxidized, the
Li3N-based materials have recently been investigated for their
potential as hydrogen storage materials.1-10 It is generally agreed
that the hydrogen storage in Li3N takes place in two steps, as
shown in equations 1 and 2, with a total gravimetric density of
10.4 wt % H2 and volumetric density of ∼140 kg H2/m3.1,6,7,11
oxidation product(s) would no longer capture NH3, and escaping
of NH3 occurs.4
Because of the unique role played by LiH in the hydrogen
storage reactions of Li3N-based materials, the stability of LiH
under a high-purity argon atmosphere, an argon atmosphere
containing some O2 and H2O impurities, and ambient air, would
be of interest from both scientific and technological viewpoints.
The understanding of such stability would shed light on how
to handle and store LiH powder in order to prevent its oxidation.
It may also provide a base to understand the cyclic stability of
Li3N-based materials in hydrogen uptake and release cycles.
This becomes especially important if impurities enter the
hydrogen storage tank with each refueling.
Li3N + H2 ) Li2NH + LiH
(1)
Li2NH + H2 ) LiNH2 + LiH
(2)
Lithium hydride (LiH) is present in both steps, and plays a
critical role in preventing the formation of ammonia (NH3).4,5
It has been shown that the reverse process of eq 2, i.e.,
dehydriding of the lithium amide (LiNH2) and lithium hydride
mixture, proceeds with two elementary reactions, as defined by
equations 3 and 4.4,5
The reaction of LiH with water vapor has been investigated
quite extensively.12-15 It is established that when the amount
of water is only sufficient to react with one surface layer of
LiH, Li2O forms via the following reaction13
LiNH2 ) 1/2 Li2NH + 1/2 NH3
(3)
(4)
LiH + 1/2 H2O ) 1/2 Li2O + H2
(5)
1/2 NH3 + 1/2 LiH ) 1/2 LiNH2 + 1/2 H2
However, when the amount of water is sufficient to react with
several surface layers of LiH, LiOH forms via the following
reaction13
As shown in eqs 3 and 4, one complete cycle of these two
elementary successive reactions consumes 50% LiNH2 and 50%
LiH in the powder mixture. Such successive reactions would
continue until all LiNH2 and LiH completely transform to Li2-
NH and H2. Reaction 4 takes place extremely fast, on the order
LiH + H2O ) LiOH + H2
(6)
At even higher moisture levels, the formation of LiOH‚H2O is
observed.13-15
* Corresponding author. Phone: 1 (860) 486-2592. Fax: 1 (860) 486-
4745. E-mail: leon.shaw@uconn.edu.
† Department of Materials Science and Engineering, University of
Connecticut.
The stability of LiH in an environment with the co-presence
of O2 and H2O, on the other hand, is less investigated.14 This is
especially true for LiH under an argon atmosphere containing
trace amounts of O2 and H2O simultaneously. With this in mind,
this study investigates the oxidation behavior of LiH under a
‡ School of Materials Science and Engineering, Dalian Jiaotong Uni-
versity.
§ Departamento de Electro´nica e Ingenier´ıa Electromeca´nica, Universidad
de Extremadura.
10.1021/jp060068m CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/09/2006

10568 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Ren et al.
high purity argon atmosphere, an argon atmosphere with some
O2 and H2 impurities, and ambient air at both room and high
temperatures. Comparisons in the oxidation behavior among
these different environments are performed. Furthermore, since
ball milling is widely used to enhance the kinetics of hydrogen
uptake and release of hydrogen storage materials,16-20 the
oxidation behaviors of the ball-milled LiH under these different
atmospheres are also investigated and compared with those
without ball milling.
exposure to air (less than 30 s). The system was then flushed
immediately with Ar of 99.999% purity for 90 min before
heating from 20 to 600 °C with a heating rate of 5 °C/min. The
argon flow rate was maintained at 60 mL/min in the entire
holding and heating process. The outlet gas from TGA was
constantly analyzed using a quadrupole residual gas analyzer
(RGA) equipped with a mass spectrometer (model ppt-c300-
F2Y). The gases monitored included H2, N2, O2, H2O, and Ar.
To quantify the composition of the outlet gas, the RGA was
calibrated using two calibration gases, with one containing 21.71
vol % H2 and 78.29 vol % Ar and the other 301 ppm N2, 1210
ppm O2, 1990 ppm NH3, and balance Ar. Furthermore, the
detection limits of the RGA, determined using a flowing Ar of
99.999% purity, were 47, 49, 20, and 916 ppm for H2, N2, O2,
and H2O, respectively.
The second set of experiments to quantify the oxidation
behavior of LiH under an argon atmosphere was X-ray diffrac-
tion analysis of various types of LiH powders held inside a
sealed capillary tube which helped to prevent oxidation of LiH
powder during the XRD data collection. The loading of the
powder to the capillary tube was performed in a glovebox filled
with Ar of 99.999% purity. The capillary quartz tube had a wall
of 0.01 mm thick and thus was transparent to the X-ray. Four
types of LiH powders were studied in this set of experiments.
They included (i) the as-purchased LiH powder, (ii) the ball-
milled powder, (iii) the powder with ball milling and subse-
quently being stored in a glovebox filled with Ar of 99.999%
purity for 4 months (called the milled and exposed LiH
hereafter), and (iv) the powder without ball milling, but stored
in a glovebox filled with Ar of 99.999% for 4 months (termed
the purchased and exposed LiH hereafter).
II. Experimental Section
Lithium hydride with 95% purity was purchased from Sigma-
Aldrich. High-energy ball milling was performed using a
modified Szegvari attritor that has been shown to be effective
in preventing the formation of the dead zone and producing
uniform milling products within the powder charge.21 The
canister of the attritor and balls 6.4 mm in diameter were both
made of stainless steel. The loading of balls and the powder to
the canister was performed in a glovebox filled with ultrahigh
purity argon that contains Ar 99.999%, H2O < 1 ppm, O2 < 1
ppm, H2 < 3 ppm, N2 < 5 ppm, and THC < 0.5 ppm (to be
referred to as Ar of 99.999% purity hereafter). The ball-to-
powder weight ratio was 60:1, the milling speed was 600 rpm,
and the milling temperature was maintained at 20 °C, achieved
by water cooling at a flowing rate of 770 mL/min. The high-
energy ball milling for all the samples studied lasted for 3 h.
All the samples before and after high-energy ball milling were
subjected to various characterizations and handled in a glovebox
filled with Ar of 99.999% purity. The specific surface area
(SSA) of the LiH powders before and after high-energy ball
milling were determined through nitrogen adsorption at 77 K
based on the Brunauer-Emmett-Teller (BET) method22 using
a gas sorption analyzer (NOVA 1000). The loading of the LiH
sample (∼0.40 g) into a sample cell with a Teflon stem filler
was performed in a glovebox filled with Ar of 99.999% purity.
The measurement was performed immediately after the sample
was loaded in the instrument without a degassing treatment.
The relative pressure (P/Po) was 0.05-0.3 and the reported SSA
data were calculated based on 5 points BET method.
The phase composition of the oxidized LiH powders as a
function of the exposing time was calculated from the XRD
patterns by the normalized-intensity-ratio (NIR) method. The
NIR method uses the kinematics XRD theory to calculate the
volume fraction of the jth crystalline phase (X(jV)) in a combi-
nation of n crystalline phases through the equation:
IjgVj2e2B
jg
X(jV)
)
(7)
2
KmjgLPjgFjg
The oxidation behavior of LiH in ambient air was evaluated
by installing LiH powders with and without ball milling
individually in an X-ray diffractometer holder and exposing the
powders to ambient air directly at room temperature. X-ray
diffraction (XRD) patterns of the interested powder as a function
of the exposing time were collected from the same sample
repetitively in a pre-designed time interval. The operation
conditions for the XRD data collection were CuKR radiation,
40 kV, 40mA, 5°/min, and 0.02°/step using a D8 ADVANCE
diffractometer. Before being installed in the XRD holder, some
LiH samples were mixed with pure silicon powder. Since silicon
is stable in air at room-temperature, the position of Si peaks
can provide a baseline for a quick correction of the instrumental
aberrations in the XRD patterns, and is thus useful for a better
phase identification and lattice parameter determination in the
oxidized LiH powder. A dehumidifier was installed next to the
XRD instrument in order to maintain a constant relative
humidity of ambient air during the XRD data collection. The
relative humidity was constantly monitored and all the data were
collected with 25-30% relative humidity.
where Ijg, e2B , mjg, LPjg, and Fjg are the integrated intensity,
the Debye-Waller temperature factor, the multiplicity, the
Lorentz-polarization factor, and the structure factor of the gth
Bragg reflection of the jth phase, respectively. The term Vj is
the unit cell volume of the jth phase, whereas K is a constant
that depends on both diffractometer and sample features. In
practice, Ijg and Vj in eq 7 are obtained from the XRD data,
whereas the rest of factors in eq 7 are readily calculated from
the crystal structure of the jth phase, thus avoiding the use of
internal or external standards. Consequently, the volume fraction
of all phases in the specimen are calculated by formulating and
solving a square system constituted by a set of eq 7. However,
usually the constant K is not evaluated explicitly, and therefore,
the magnitudes calculated directly from the resolution of the
jg
square system are the product KX(jV). After knowing all KX(jV)
,
the corresponding X(jV) are then calculated by imposing the
normalization condition as follows:
KX(jV)
Two types of experiments were performed to quantify the
oxidation behavior of LiH under an argon atmosphere. The first
was the thermogravimetric analysis (TGA) which was conducted
using a TA instrument (TGA Q500). LiH samples of 25-55
mg were loaded into a Pt-microbalance pan with a short
X(jV)
)
(8)
n
KX(V)
∑
j
j)1

Stability of Lithium Hydride in Argon and Air
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10569
Figure 1. Experimental XRD patterns for the as-purchased LiH (solid
line) and the purchased and exposed LiH (dashed line). For comparison,
the two patterns have been normalized by imposing the same peak
intensity for the 111 Bragg reflection of the LiH phase.
Figure 2. Experimental XRD patterns for the ball-milled LiH (solid
line) and the milled and exposed LiH (dashed line). For comparison,
the two patterns have been normalized by imposing the same peak
intensity for the 111 Bragg reflection of the LiH phase.
In the present study, the terms Ijg and Vj for eq 7 were obtained,
respectively, from the area and position of the CuKR1 component
of the Bragg reflections of the phases present, which were
determined by analytical profile fitting assuming Voigt functions
with CuKR1 and CuKR2 components for the modeling of the
Bragg reflections and a first-order polynomial function for the
modeling of the background; the fits were conducted iteratively
by the Levenberg-Marquardt nonlineal least-squares algorithm.
These areas directly yielded the integrated intensities, whereas
these positions yielded Vj through the Bragg law coupled with
the plane-spacing equations for the different crystal systems.
0.5 vol %), it can be concluded that no oxidation takes place
when the ball-milled LiH powder is exposed to an argon
atmosphere of 99.999% purity for 4 months at room temperature.
Two additional conclusions can be made by comparing the
LiH with and without ball milling. First, there is no oxidation
during the ball milling process, indicating that the seal of the
canister of the attritor is airtight, and there is no leaking of argon
gas during ball milling. Second, the oxidation of LiH, if any,
under a high-purity argon atmosphere is extremely slow at room
temperature, regardless of the powder particle sizes. It is well-
known that ball milling can reduce powder particle sizes
dramatically, often from tens of micrometers to tenths of
micrometers.23,24 The measurement of the specific surface area
(SSA) of the LiH powder with and without ball milling confirms
this trend. Using the BET method, it is found that the
as-purchased LiH has a SSA of 4.614 m2/g, whereas the ball-
milled LiH has a SSA of 13.104 m2/g. Based on these SSA
data and assuming that the particles are spherical, the equivalent
particle radius can be calculated, and is found to be 0.79 µm
for the as-received LiH and 0.28 µm for the ball milled LiH.
Thus, the above data unambiguously show that both coarse and
fine LiH particles are stable at room temperature under an argon
atmosphere of 99.999% purity, as one would expect since argon
is an inert gas.
3.2. Stability of LiH in Ambient Air at Room Tempera-
ture. Shown in Figure 3a and 3b are selected experimental XRD
patterns of LiH with and without ball milling, collected in situ
during the exposure to ambient air at room temperature. It can
be seen that as the exposure time increases, the integrated
intensities of LiOH peaks increase, which is accompanied by a
continued decrease in the integrated intensities of LiH peaks.
This is true for both milled and unmilled LiH powders,
indicating that LiH is oxidized and the oxidation product is
LiOH for both milled and unmilled powders.
III. Results and Discussion
3.1. Stability of LiH Under a High-Purity Ar Atmosphere
at Room Temperature. The experimental XRD patterns of the
as-purchased LiH and the purchased LiH with a subsequent
exposure to an argon atmosphere of 99.999% purity at room
temperature for 4 months are shown in Figure 1. Note that the
ratios of LiH/LiOH integrated intensities are identical for the
LiH powder with and without 4-month exposure to the argon
atmosphere. The phase composition calculated from eqs 7 and
8 for the as-purchased LiH is 94.1 vol % LiH + 5.9 vol %
LiOH. The corresponding value for the purchased LiH with a
subsequent exposure to an argon atmosphere for 4 months is
93.9 vol % LiH + 6.1 vol % LiOH. The minute difference
between the two samples is within the experimental error (about
0.5 vol %), and thus it is concluded that there is no discernible
difference in the phase composition between the two samples.
This result indicates that there is no oxidation when the as-
purchased LiH is stored at room temperature under an argon
atmosphere of 99.999% purity with H2O < 1 ppm, O2 < 1 ppm,
H2 < 3 ppm, N2 < 5 ppm, and THC < 0.5 ppm. The result
also reveals that the as-purchased LiH contains 6.0 vol % LiOH.
Figure 2 shows the experimental XRD patterns of the as-
milled LiH and the milled LiH with a subsequent exposure to
an argon atmosphere of 99.999% purity at room temperature
for 4 months. Note that the same phenomenon as the LiH
without ball milling is found, that is, the ratio of LiH/LiOH
integrated intensities are identical for the ball-milled LiH with
and without 4-month exposure to an argon atmosphere. The
phase composition for the as-milled LiH is found to be 94.0
vol % LiH + 6.0 vol % LiOH, and the corresponding value for
the milled and exposed LiH is 93.7 vol % LiH + 6.3 vol %
LiOH. Since the minute difference in the phase composition
between the two samples is within the experimental error (about
The result from the quantitative phase composition analysis
based on the XRD data is shown in Figure 4. As expected, the
volume fraction of the oxidation product, LiOH, continues to
increase, while the volume fraction of LiH decreases with the
exposure time. Both the milled and unmilled LiH exhibit the
same trend. However, the oxidation rate of the milled LiH is
faster than that of the unmilled LiH, as shown in Figure 5.
There are several possible reaction routes that can lead to
the formation of LiOH when LiH is exposed to air at room
temperature. Thermodynamic calculations25 indicate that all of
the reactions 5, 6, and 9-11, listed below along with their

10570 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Ren et al.
Figure 3. Experimentally measured XRD patterns of LiH exposing
to ambient air at room temperature as a function of the exposure time:
(a) without milling and (b) with milling for 3 h at 20 °C in Ar of
99.999% purity.
Figure 4. Volume fraction of the phases present in (a) the as-purchased
LiH and (b) the ball-milled LiH as a function of the exposure time to
ambient air at room temperature. The error bars are lower than the
point size, and the solid curves are guides for the eyes.
corresponding standard Gibbs free-energy changes, are ther-
modynamically possible.
LiH + 1/2 H2O ) 1/2 Li2O + H2
∆G ) - 93.98 kJ/mol
(5)
LiH + H2O ) LiOH + H2
∆G ) - 133.01 kJ/mol (6)
LiH + 1/2 O2 ) LiOH ∆G ) - 370.24 kJ/mol (9)
LiH + 1/4 O2 ) 1/2 Li2O + 1/2 H2 ∆G ) - 212.59 kJ/mol
(10)
Li2O + H2O ) 2 LiOH ∆G ) - 78.06 kJ/mol (11)
Both reactions 6 and 9 lead to the formation of LiOH directly,
while reactions 5 and 10 produce Li2O first which subsequently
reacts with H2O to form LiOH. Since in the entire exposure
period there is not any trace of Li2O, it can be concluded that
either reactions 5 and 10 do not occur at all, or reaction 11
proceeds extremely fast so that Li2O disappears as soon as it
forms from reactions 5 and 10. The latter is very unlikely
because a previous study26 has shown that the conversion
between Li2O and LiOH as defined by eq 11 can be easily
followed using elastic proton backscattering in a matter of
several hours, indicating that reaction 11 is sluggish. Thus, the
oxidation of LiH in ambient air with ∼27% relative humidity
must proceed with reaction 6, 9, or both.
Figure 5. Comparison in the oxidation rate under ambient air at room
temperature between the LiH powders with and without ball-milling.
The error bars are lower than the point size. The solid lines are generated
using the Johnson-Mehl-Avrami models (to be discussed in the text).
reacts with H2O as well as O2 even when LiH is exposed to an
argon atmosphere containing 515 ppm O2 (i.e., 0.0515 vol %)
and 5692 ppm H2O (0.569 vol %) at room temperature. For the
ambient air studied here, the concentrations of O2 and H2O are
20.9 and 0.81 vol %, respectively, both of which are higher
than those of O2 and H2O in the argon atmosphere to be
discussed in Section 3.3. Thus, it is reasonable to deduce that
LiH reacts with both O2 and H2O in ambient air with a 27%
relative humidity at 25 °C to form LiOH, as indicated by
reactions 6 and 9. In what follows, the reaction kinetics will be
Both reactions 6 and 9 are thermodynamically possible and
have indeed occurred. As will be shown in Section 3.3, LiH

Stability of Lithium Hydride in Argon and Air
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10571
where k4 is a rate constant. Finally, if the oxidation of LiH is
controlled by nucleation and subsequent growth of LiOH, the
volume fraction of LiOH as a function of time can then be
expressed by the well-known Johnson-Mehl-Avrami (JMA)
equation30
f ) 1 - exp(- k5tm)
(17)
where m is the mechanism constant which varies from 1 to 4,
depending on the detail of nucleation and growth of LiOH, and
k5 is the rate constant affected by nucleation and growth
mechanisms as well.
Figure 6. Schematic of the oxidation of LiH particles in ambient air.
discussed for both reactions 6 and 9. However, reaction 6 will
be analyzed first, and the result will be applied to reaction 9.
Based on the formula of reaction 6, it can be stated that the
rate control mechanism for oxidation of LiH in air can be (i)
evaporation of the H2 gas, (ii) concentration of the reactant,
LiH, (iii) diffusion of oxygen, hydrogen, or water through the
LiOH layer, (iv) movement of the LiH/LiOH phase boundary,
or (v) nucleation and growth of LiOH. It is well established
that if the oxidation is controlled by evaporation of the gaseous
product, the volume fraction of the oxidation product will follow
a linear-rate law27
To find out the rate-controlling mechanism, eqs 12, 13, 15,
16 and 17 should be fitted to the experimental data obtained
from the quantitative XRD analysis, and the degree of discrep-
ancy between the theoretical prediction and the experimental
data be quantified in terms of the reduced chi-squares (ør2).
Curve fitting through the least-squares method results in two
linear relationships from each of these attempts, one for the
unmilled LiH and the other for the milled LiH. Once the
parameters for all of the five possible rate-controlling mecha-
nisms are determined, comparisons between these possible rate-
controlling mechanisms and the experimental data can be
performed by plotting the predictions from various rate-
controlling mechanisms and the experimental data in a single
f-versus-t chart. Of course, the most likely rate-controlling
mechanism is the one yielding the lowest ør2. Shown in Figures
7a and b are such plots for the unmilled LiH and milled LiH,
respectively. It can be seen that the nucleation and growth
f ) k1t
(12)
where f stands for the volume fraction of the oxidation product,
k1 is a rate constant, and t is the time during an isothermal
reaction. If the oxidation rate is controlled by the concentration
of the reactant, then the functional dependency of f on t is given
by the following equation28
mechanism, i.e., eq 17, fits the experimental data best (ør2
)
0.00049 for the unmilled LiH and ør2 ) 0.0001 for the milled
LiH), followed by the movement of the LiH/LiOH phase
boundary mechanism (ør2 ) 0.0011 for the unmilled LiH and
ør2 ) 0.0042 for the milled LiH). Thus, it can be concluded
that the oxidation of LiH in air at room temperature is not
controlled by evaporation of H2 (ør2 ) 0.00711 for the unmilled
LiH and ør2 ) 0.02214 for the milled LiH), the higher order
1/n-1
n - 1
k2m0
f ) 1 -
(13)
n-1
[
]
t
where n is the order of reaction, m0 is the initial mass of the
reactant, and k2 is a rate constant. If the reaction is a first order
reaction, the volume fraction of the reaction product can then
be expressed in a relatively simple form28
reaction of the concentration of LiH in the powder (ør2
)
0.00952 for the unmilled LiH and ør2 ) 0.00815 for the milled
f ) 1 - exp(-k2t)
(14)
LiH), and diffusion of oxygen or hydrogen atoms through the
LiOH layer (ør2 ) 0.02518 for the unmilled LiH and ør2
)
If the oxidation product, LiOH, forms a dense layer on the
surface of LiH, the oxidation rate will be controlled by the
diffusion of water through the LiOH layer, or by the diffusion
of oxygen because the diffusion rate of hydrogen is typically
much faster than that of oxygen if water dissociates into oxygen
and hydrogen.29 A parabolic-rate law will be observed for such
a diffusion-controlled reaction.27 Given that the LiH in this study
is in a particle form, it is reasonable to assume that the oxidation
of LiH would occur via growth of a shell of LiOH into the
interior of a LiH particle if the oxidation of LiH is diffusion-
controlled (see Figure 6). It is easy to show that with such a
LiOH-growth mechanism, the volume fraction of LiOH as a
function of the reaction time is
0.01129 for the milled LiH). Instead, the oxidation rate of LiH
in air is controlled by the nucleation and growth of LiOH or by
the movement of the LiH/LiOH phase boundary. Note that the
difference between these two mechanisms is, in fact, very small,
as discussed below.
The JMA equation describing the mechanism of the nucle-
ation and growth of LiOH, obtained through the curve fitting
for the oxidation of the as-purchased LiH in air at room
temperature, is
f ) 1 - exp(- 0.018t1.01
)
(18)
and it becomes
f ) 1 - exp(- 0.028t1.01
)
(19)
1/2
3
k3
f ) 1 - 1 -
t1/2
(15)
[
]
for the ball-milled LiH. The equations describing the mechanism
of the LiH/LiOH phase boundary movement, obtained via curve
fitting, are
R
where R is the original average radius of LiH particles and k3
is a rate constant. However, if the oxidation of LiH is controlled
by the movement of the LiH/LiOH phase boundary, the volume
fraction of LiOH as a function of time would have the following
form:15
f ) 1 - (1-0.0056t)3
(20)
and
f ) 1 - (1-0.0080t)3
(21)
3
k
f ) 1 - 1 - 4 t
(16)
[
]
R
for the as-purchased and ball-milled LiH, respectively.

10572 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Ren et al.
However, the milled LiH powder oxidizes faster because it has
a larger k5 (see eq 17) or k4/R (see eq 16) than the unmilled
LiH powder, i.e., 0.028 vs 0.018 in eqs 19 and 18, and 0.0080
vs 0.0056 in eqs 21 and 20, respectively. The faster oxidation
rate of the milled LiH is due to its finer particle size and larger
surface area.
The conclusions obtained above are consistent with the SEM
observation. Shown in Figure 8 are SEM images of the typical
surfaces of (i) the as-purchased LiH particles with an exposure
to air at room temperature in less than 1 min before being coated
with a thin layer of Au and (ii) the purchased and exposed LiH
particles which have been intentionally exposed to air at room
temperature for 10 min before being coated with a thin layer of
Au. It is very clear that the density of microcracks on the surface
of the oxidized LiH particle increases from less than 1-min
exposure to 10-min exposure. The increase in the microcrack
density is consistent with the measurement of the specific surface
area, which has the experimental error of (0.35 m2/g and shows
that the SSA of LiH particles increases from 4.61 m2/g to 6.56
m2/g after exposure to air for 10 min at room temperature. This
increased density of microcracks eliminates the dependence of
the continued oxidation of LiH particles on the diffusion of
oxygen or hydrogen through the LiOH surface layer. As a result,
the oxidation of LiH is controlled by the formation of LiOH
nuclei and the subsequent movement of the LiH/LiOH phase
boundary.
Finally, it should be pointed out that the conclusions obtained
for reaction 6 are also applicable to reaction 9. This is true
because the formation of LiOH from reaction 9 will be on the
surface of LiH particles. As a result, the subsequent oxidation
will also rely on the movement of the LiH/LiOH phase
boundary.
Figure 7. Comparisons among all of the possible rate-controlling
mechanisms using the least-squares prediction equations and the volume
fraction of the experimentally determined LiOH as a function of the
oxidizing time for (a) the as-purchased LiH and (b) the ball-milled
LiH in air at room temperature. The error bars for experimental data
are lower than the point size. See the text for details.
3.3. Stability of LiH Under an Ar Atmosphere with O2
and H2O Impurities. The stability of LiH under an argon
atmosphere with some impurities of O2, H2O, N2, H2 and NH3
was evaluated using a TGA apparatus in conjunction with a
RGA for the outlet gas analysis. Shown in Figure 9 is the outlet
gas composition from the TGA apparatus with an empty pan
(i.e., a blank sample) and a flowing argon atmosphere of
99.999% purity. It is clear that immediately after the close of
the TGA chamber (the zero point of time), all of the gaseous
species (N2, O2, H2O, and H2) except Ar exhibit a general trend
of decreasing concentrations with time. Ar is the only exception
to this rule. It increases with time initially and in less than 10
min the argon concentration has reached its steady-state value.
In contrast, the concentrations of all other species decrease
continuously even after 180 min at which the experiment is
stopped. As expected, the onset of heating at 90 min and the
subsequent heating do not impose any interference to the
variation of the gaseous composition with time. All of these
phenomena can be explained based on the fact that ambient air
in the TGA chamber is gradually replaced by the flowing argon.
The quantitative analysis indicates that the concentrations of
H2O, O2, H2, and N2, in the outlet gas at 90 min are 5692, 515,
585, and 88 ppm, respectively, and the corresponding values
for these gaseous species at 180 min are 4541, 443, 530, and
72 ppm.
It is well established that, for solid-state transformations with
one-directional growth, the m constant in the JMA equation will
be between 1 and 2.30 Furthermore, when m ) 1, the reaction
product nucleates at the beginning of the reaction, and the
subsequent transformation is controlled by the one-directional
growth of the reaction product.30 Based on this established
mechanism and the m constant in eqs 18 and 19, it can be stated
that the oxidation of the as-purchased and ball-milled LiH in
air at room temperature proceeds by the formation of all the
LiOH nuclei on the surface of the LiH particles at the beginning
of the reaction, followed by the subsequent growth of these
nuclei into the interiors of the LiH particles, as shown
schematically in Figure 6.
The difference between eqs 18 and 20 is that eq 18 takes
into the consideration of nucleation events, whereas eq 20
assumes that nucleation takes place instantaneously and the
entire transformation process is controlled by the growth of
LiOH into the LiH particle only. Since eq 18 fits the experi-
mental data better than eq 20 as shown in Figure 7a, it is
reasonable to conclude that the nucleation event accounts for a
noticeable period of time during the oxidation of LiH in air at
room temperature. The same conclusion can be made for the
milled LiH powder by comparing eqs 19 and 21, as supported
by the observation of Figure 7b.
Figure 10 shows the weight change of the as-purchased LiH
as a function of time after the heating starts at 90-min of holding.
There are several interesting features in this figure. First, LiH
is oxidized during 90-min of holding at room temperature. This
oxidation has resulted in a 1% increase in weight and is due to
the presence of O2 and H2O impurities in the TGA chamber
(see Figure 9 and additional discussion below). The 1% increase
Note that since the m constants in eqs 18 and 19 are the same
for the uilled and milled LiH powders, it can be concluded that
the oxidation mechanisms for these two powders are identical.

Stability of Lithium Hydride in Argon and Air
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10573
Figure 8. SEM images of the as-purchased LiH with a subsequent exposure to air at room temperature for (a) less than 1 min and (b) about 11
min.
the weight gain appears to increase at ∼390 °C, suggesting an
increase in the oxidation rate above 390 °C.
Shown in Figure 11 is the outlet gas composition as a function
of temperature for the as-purchased LiH in the TGA chamber
with an identical experimental condition as that in Figure 10.
The important features to be noted from this figure are as
follows. (i) The hydrogen concentration exhibits obvious
increases at ∼40 °C, suggesting the increased oxidation rate
right after the increase in temperature. (ii) Two hydrogen
concentration peaks are observed, one at 130 °C and the other
at 600 °C after which the TGA chamber starts to cool. (iii) The
oxygen concentration exhibits a precipitous decrease at ∼55
°C, and continues to decrease as temperature increases. How-
ever, the oxygen concentration starts to increase when the
cooling is initiated after reaching 600 °C. (iv) Heating does not
affect the gradual decrease in the N2 concentration. The latter
Figure 9. The composition profile of the outlet gas for a blank sample
in the TGA apparatus as a function of the holding time. A flowing
is caused by the gradual replacement of ambient air by the
flowing argon. (v) The concentration of H2O exhibits a general
argon atmosphere of 99.999% purity was provided immediately after
trend of decrease as temperature and time increase. However,
the close of the TGA chamber at 25 °C. The time at which the TGA
two peaks are superimposed on this general trend with one
chamber is closed is taken as the zero point, and at 90 min the TGA
appearing at ∼150 °C and the other at 600 °C.
chamber is heated at a heating rate of 5 °C/min.
The composition change as a function of temperature and
time observed in Figure 11 provides direct evidence for deducing
the oxidation mechanism of LiH in a flowing argon with less
than 5692 ppm H2O, 515 ppm O2, 585 ppm H2, and 88 ppm N2
impurities. Based on the phenomena observed in Figure 11 and
the comparison with eqs 5, 6, 9, and 10, the following
conclusions can be made. First, the weight gain observed in
Figure 10 is not induced by N2 because N2 exhibits a continuous
and gradual decrease with time for the entire holding and heating
period. Second, the weight gain from ∼55 to 600 °C is caused
by the O2 impurity. Furthermore, the oxidation of LiH proceeds
with reaction 10 because this reaction results in a decrease in
O2 and an increase in H2 simultaneously. This is exactly what
is observed in Figure 11. Third, reactions 5 and 6, which also
produce H2, play a minimal role in the oxidation of LiH from
∼55 to 600 °C under the condition investigated. This is
Figure 10. TGA curve of the as-purchased LiH in a flowing argon
atmosphere with a constant heating rate of 5 °C/min. The sample has
been held in the TGA chamber with a flowing argon of 99.999% purity
for 90 min at room temperature before the heating starts, and the cooling
is initiated when temperature reaches 600 °C.
supported by the fact that the H2O concentration in the outlet
gas continues to decrease at the cooling stage after 600 °C,
whereas the O2 concentration increases, instead, at the cooling
stage. The increased O2 concentration at the cooling stage
[Figure 11c] indicates a decrease in the oxidation rate of LiH
and thus less consumption of O2 in the flowing argon as the
temperature decreases. In contrast, the continued decrease in
the H2O concentration after cooling from 600 °C [Figure 11b]
can only be explained by the continued replacement of the
in weight would represent oxidation of 0.41 vol % LiH in the
system if the oxidation occurs via either reaction 6 or 9. Second,
the weight gain (i.e., another 3% increase) is observed in the
entire heating period (from 25 to 600 °C), indicating the
continued oxidation of LiH during heating. Last, the slope of

10574 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Ren et al.
Figure 12. The variation of H2O and O2 concentrations in the outlet
gas as a function of the holding time near 90-min of holding for the
as-purchased LiH sample in the TGA chamber. The experimental
condition is identical to that shown in Figures 10 and 11.
subsequent cracking of this Li2O shell results in a substantial
increase in the oxidation rate at temperatures higher than 400
°C. This proposition needs to be verified in a future study.
Finally, it is worthwhile to examine the oxidation mechanism
of LiH at room temperature under the flowing argon atmosphere
with H2O and O2 impurities. This can be done by studying how
the concentrations of H2O and O2 in the outlet gas vary with
the holding time for a LiH sample before heating starts at 90-
min of holding. Shown in Figure 12 are details of H2O and O2
concentrations as a function of the holding time near 90-min
of holding for a LiH sample in the TGA chamber. A comparison
with the blank sample (Figure 9) reveals that at 90-min of
holding the concentration of H2O in the outlet gas with a LiH
sample in the TGA chamber is lower than that of H2O without
a LiH sample by a factor of 0.1, indicating that some H2O in
the TGA chamber has reacted with the LiH sample. Similarly,
the concentration of O2 in the outlet gas with a LiH sample in
the TGA chamber is lower than that of O2 without a LiH sample
by a factor of 0.2, indicating that some O2 in the TGA chamber
has also reacted with the LiH sample.
The fact that the outlet gas has lower concentrations of O2
and H2O when the TGA chamber contains a LiH sample
provides clear evidence for the reaction of LiH with the O2 and
H2O impurities in the flowing argon at room temperature. Such
reactions are most likely due to reactions 6 and 9. This
conclusion is reached based on the following argument. First,
reaction 5 is unlikely because the H2O impurity in the present
study (5692 ppm) is much higher than what is needed to react
with one surface layer of LiH. Reaction 5 only occurs when
the amount of water is only sufficient to react with one surface
layer of LiH.13 Second, reaction 10 has been identified to occur
at temperatures higher than ∼55 °C. Below 55 °C different
oxidation mechanisms are present because a sudden increase
in the oxygen consumption takes place at ∼55 °C (Figure 11c).
Thus, reactions 6 and 9 are the remaining reactions that can
account for the decrease in the O2 and H2O concentrations
observed at room temperature. Finally, the conclusion that
reactions 6 and 9 are operational at room temperature is
consistent with the formation of only LiOH observed when LiH
is exposed to ambient air, as discussed in Section 3.2. It is also
in good agreement with a previous study31 which shows that at
a given oxidation environment, LiOH is preferentially formed
at lower temperatures, while Li2O is formed predominately at
higher temperatures.
Figure 11. The variation of the outlet gas composition as a function
of temperature for the as-purchased LiH sample in the TGA chamber
with the experimental condition identical to that in Figure 10: (a) H2,
(b) H2O and N2, and (c) O2. Note that the cooling is initiated at 600
°C. Thus, the composition changes after the onset of cooling should
be taken as the variation as a function of time.
residual H2O from ambient air by the flowing argon. Therefore,
it can be concluded that H2O plays a minimum role in the
oxidation of LiH in the heating stage from ∼55 to 600 °C.
Fourth, the two H2O peaks are generated via the following
reaction
H2 + 1/2 O2 ) H2O
(22)
This conclusion is supported by the fact that the two H2O peaks
appear at the same time as the two H2 peaks, that is, the
increased H2 concentration drives reaction 22 toward the
formation of H2O. Fifth, the two H2 peaks are believed to be
related to the change in the oxidation rate of LiH with
temperature and possibly with the formation of a Li2O shell on
the surface of the LiH particle. The exact mechanism(s) for this
phenomenon are not investigated in this study. However, one
possible explanation would be the formation of the Li2O shell
which leads to a reduction in the oxidation rate of LiH at
temperatures higher than 135 °C but below 400 °C. The
IV. Summary and Conclusions
The stability of LiH in high purity argon, argon with some
O2 and H2O impurities, and ambient air has been investigated.

Stability of Lithium Hydride in Argon and Air
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10575
(3) Hu, Y. H.; Ruckenstein, E. H2 storage in Li3N. Temperature-
programmed hydrogenation and dehydrogenation. Ind. Eng. Chem. Res.
2003, 42, 5135.
Effects of ball milling on the LiH stability have also been
examined. Based on this study, the following conclusions can
be offered.
1. No oxidation of LiH occurs at room temperature under an
argon atmosphere of 99.999% purity with H2O < 1 ppm, O2 <
1 ppm, H2 < 3 ppm, N2 < 5 ppm, and THC < 0.5 ppm.
2. This conclusion remains true even after LiH particles have
been subjected to ball milling with an increase in the specific
surface area to 13.1 m2/g and a reduction in the equivalent
particle size to 0.56 µm in diameter.
3. The oxidation product of LiH in ambient air with a relative
humidity of 27% at room temperature is LiOH.
4. The oxidation kinetics of LiH in ambient air at room
temperature is controlled by nucleation and growth processes.
LiOH nucleates on the surface of the LiH particle at the
beginning of the reaction, and the subsequent reaction is
controlled by the one-directional growth of LiOH into the LiH
core.
(4) Hu, Y. H.; Ruckenstein, E. Ultrafast reaction between LiH and NH3
during H2 storage in Li3N. J. Phys. Chem. A 2003, 107, 9737.
(5) Ichikawa, T.; Hanada, N.; Isobe, S.; Leng, H.; Fujii, H. Mechanism
of novel reaction from LiNH2 and LiH to Li2NH and H2 as a promising
hydrogen storage system. J. Phys. Chem. B 2004, 108, 7887.
(6) Ichikawa, T.; Isobe, S.; Hanada, N.; Fujii, H. Lithium nitride for
reversible hydrogen storage. J. Alloys Compd. 2004, 365, 271.
(7) Orimo, S.; Nakamori, Y.; Kitahara, G.; Miwa, K.; Ohba, N.;
Noritake, T.; Towata, S. Destablization and enhanced dehydriding reaction
of LiNH2: an electronic structure viewpoint. Appl. Phys. A 2004, 79, 1765.
(8) Ichikawa, T.; Hanada, N.; Isobe, S.; Leng, H.; Fujii, H. Composite
materials based on light elements for hydrogen storage. Mater. Trans. 2005,
46, 1.
(9) Kojima, Y.; Kawai, Y. IR characterization of lithium imide and
amide. J. Alloys Compd. 2005, 395, 236.
(10) Hu, Y. H.; Ruckenstein, E. Highly effective Li2O/Li3N with ultrafast
kinetics for H2 storage. Ind. Eng. Chem. Res. 2004, 43, 2464.
(11) Ordaz, G.; Petrovic, J.; Read, C.; Satyapal, S. Hydrogen storage.
In CD of the 2005 Annual DOE Hydrogen Program ReView Meeting,
Washington, DC, May 23-26, 2005; U.S. Department of Energy: Wash-
ington, DC, 2005.
(12) Holcombe, C. H. Retardation of the reaction of lithium hydride
with water vapor. In USAEC Report, Union Carbide Corp., Y-12 Plant,
Y-1835/XAB; USAEC: Aberdeen Proving Ground, MD, 1972.
(13) Machin, W. D.; Tompkins, F. C. Kinetics of the reaction of water
vapor with crystalline lithium hydride. Trans. Faraday Soc. 1966, 62, 2205.
(14) Holcombe, C. E.; Powell, G. L. Some observations on the reaction
layer of LiOH on LiH. J. Nucl. Mater. 1973, 47, 121.
5. The oxidation kinetics of LiH in ambient air at room
temperature can be described very well by the Johnson-Mehl-
Avrami equation:
f ) 1 - exp(- 0.018t1.01) for the unmilled LiH
f ) 1 - exp(- 0.028t1.01) for the ball-milled LiH
(15) Dinh, L. N.; Cecala, C. M.; Leckey, J. H.; Balooch, M. The effects
of moisture on LiD single crystals studied by temperature-programmed
decomposition. J. Nucl. Mater. 2001, 295, 193.
(16) Gross, K. J.; Chartouni, D.; Leroy, E.; Zu¨ttel, A.; Schlapbach, L.
Mechanically milled Mg composites for hydrogen storage: the relationship
between morphology and kinetics. J. Alloys Compd. 1998, 269, 259.
(17) Liang, G.; Boily, S.; Huot, J.; Van Neste, A.; Schulz, R. Hydrogen
absorption properties of a mechanically milled Mg-50 wt % LaNi5
composite. J. Alloys Compd. 1998, 268, 302.
(18) Zaluski, L.; Zaluska, A.; Strom-Olsen, J. O. Hydrogen absorption
in nanocrystalline Mg2Ni formed by mechanical alloying. J. Alloys Compd.
1995, 217, 245.
(19) Orimo, S.; Fujii, H.; Ikeda, K. Notable hydriding properties of a
nanostructured composite material of the Mg2Ni-H system synthesized by
reactive mechanical grinding. Acta Mater. 1997, 45, 331.
(20) Reule, H.; Hirscher, M.; Weiâhardt, A.; Kronmuller, H. Hydrogen
desorption properties of mechanically alloyed MgH2 composite materials.
J. Alloys Compd. 2000, 305, 246.
The ball-milled LiH has a faster oxidation rate than the unmilled
LiH because the ball-milled LiH has a larger specific surface
area and finer particle size.
Oxidation of LiH occurs under an argon atmosphere contain-
ing 5692 ppm H2O, 515 ppm O2, 585 ppm H2, 88 ppm N2, and
1.3 ppm NH3. At temperatures ranging from ∼55 to 600 °C,
oxidation of LiH proceeds via the following equation:
LiH + 1/4 O2 ) 1/2 Li2O + 1/2 H2
7. At room temperature, oxidation of LiH in ambient air and
in an argon atmosphere containing 5692 ppm H2O, 515 ppm
O2, 585 ppm H2, 88 ppm N2, and 1.3 ppm NH3 is most likely
due to the following two reactions:
(21) Yang, Z.-G.; Shaw, L. Synthesis of nanocrystalline SiC at ambient
temperature through high energy reaction milling. Nanostruct. Mater. 1996,
7, 873.
(22) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of gases in
multimolecular layers. J. Am. Chem. Soc. 1938, 60, 309.
(23) Yang, Z.-G.; Ren, R.-M.; Shaw, L. Evolution of microstructures
and nitrogen sorption during high energy milling of Si in ammonia. J. Am.
Ceram. Soc. 2000, 83, 1897.
LiH + H2O ) LiOH + H2
LiH + 1/2 O2 ) LiOH
(24) Ren, R.-M.; Yang, Z.-G.; Shaw, L. Polymorphic transformation
and powder characteristics of TiO2 during high energy milling. J. Mater.
Sci. 2000, 35, 6015.
(25) Barin, I.; Knacke, O.; Kubaschewski, O. Thermochemical properties
of inorganic substances. Spinger-Verlag: Berlin, Germany, 1977.
(26) Myers, S. M. Ion-backscattering study of LiOH-to-Li2O conversion
on a LiH substrate. J. Appl. Phys. 1974, 45, 4320.
(27) Oxidation and corrosion of intermetallic alloys; Welsch, G., Desai,
P. D., Eds.; Purdue Research Foundation: West Lafayette, Indiana, 1996.
(28) Speyer, R. F. Thermal analysis of materials. Marcel Dekker: New
York, 1994.
Acknowledgment. This work was supported under the U.S.
Department of Energy (DOE) contract no. DE-FC36-05GO15008.
The vision and support of Dr. Carole Read, DOE Technology
Manager, is greatly appreciated.
(29) Diffusion in solid metals and alloys; Mehrer, H., Ed.; Springer-
Verlag: New York, 1990; Vol 26.
(30) Dinh, L. N.; McLean, W. II; Schildbach, M. A.; LeMay, J. D.;
Siekhaus, W. J.; Balooch, M. The nature and effects of the thermal stability
of lithium hydroxide. J. Nucl. Mater. 2003, 317, 175.
(31) Christian, J. W. The theory of transformations in metals and
alloys: Part I; Pergamon: Oxford, UK, 2002.
References and Notes
(1) Chen, P.; Xiong, Z.; Luo, J. Z.; Lin, J. Y.; Tan, K. L. Interaction
of hydrogen with metal nitrides and imides. Nature 2002, 420, 302.
(2) Chen, P.; Xiong, Z.; Luo, J. Z.; Lin, J. Y.; Tan, K. L. Interaction
between lithium amide and lithium hydride. J. Phys. Chem. B 2003, 107,
10967.