i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Available online at www.sciencedirect.com

ScienceDirect
journal homepage: www.elsevier.com/locate/he

Mathematical analysis of the hydrogeneuranium

reaction using the shrinking core model for
hydrogen storage application
Rupsha Bhattacharyya*, Dibyendu Bandyopadhyay, Kalyan Bhanja,
Sadhana Mohan
Heavy Water Division, BARC, Mumbai 400 085, India

article info

abstract

Article history:

The reaction of metallic uranium particles with hydrogen at ambient or above ambient

Received 27 February 2015

temperature has been used as the basis for the solid state storage of hydrogen in the form

Received in revised form

of uranium hydride for various applications in the nuclear industry. This work models the

5 May 2015

reduction of a single particle of metallic uranium to uranium hydride using available ki-

Accepted 15 May 2015

netic data, the well known shrinking core model and the pseudo-steady state hypothesis.

Available online 6 June 2015

No single rate controlling regime was assumed apriori and the various interfacial gas
concentrations were calculated by an iterative procedure as function of the core radius at

Keywords:

any time and for given operating conditions. The volumetric expansion or increase in the

Hydrogen storage

outer radius of the uranium particle as it is progressively hydrided was also considered in

Uranium

this work. The time required for complete conversion of the pellet to the hydride was then

Shrinking core

calculated by numerical integration and the rate controlling regime was identified. A

Non-linear kinetics

spherical geometry was considered in this work for illustrating the technique, but the
method is applicable to any kinetic model and any geometry of the pellets after simple
modifications to the governing equations.
Copyright 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.

Introduction
The use of hydrogen as a clean energy carrier and a fuel of
the future have been extensively documented [1]. The
widespread use of hydrogen needs the development of
efficient storage technology for it. Reversible metal hydride
based solid state hydrogen storage systems have been given
particular importance in this context. Uranium is one of
the most extensively studied metallic species for this purpose, especially for storage of heavier hydrogen isotopes [2].

The reaction of metallic uranium takes place quite readily

depending on the gas pressure, temperature and the uranium particle size. The hydride that forms starts decomposing at temperatures of about 250 deg C to yield very
pure hydrogen gas. Owing to the reversibility of the reaction, the maximum hydriding reaction rates have been
observed to occur in a temperature range of 225e250 deg C
after which the dehydriding reaction assumes greater
importance [3]. Thus the importance of this reaction system
with respect to hydrogen storage technology cannot be
undermined.

* Corresponding author. Tel.: 91 22 2559 2962.

E-mail address: rupshabhattacharyya1986@gmail.com (R. Bhattacharyya).
http://dx.doi.org/10.1016/j.ijhydene.2015.05.099
0360-3199/Copyright 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

8918

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Table 1 e Summary of kinetic studies of the hydrogeneuranium reaction.

Serial
number
[Reference]
1 [4]

2 [5]

3 [6]

4 [7]

5 [8]

6 [9]

7 [10]

8 [11]

9 [12]

10 [13]

11 [14]

12 [15]

Area of focus

The kinetics for reaction of hydrogen with

uranium powder over the temperature range
of 50e250 deg C and at two hydrogen pressures
(13.3 kPa and 26.6 kPa) were investigated using
a gravimetric technique involving an
electro-microbalance.

Main results

Surface adsorption phenomenon of hydrogen on uranium on reaction

kinetics was offered as the explanation of the observed decrease in
reaction rate with increase of temperature, even when adsorption
was not the rate controlling step. Hydriding of uranium was found to
follow 2/3rd order kinetics for 13.3 kPa hydrogen pressure which rose
to 1st order kinetics at the higher pressure. Decomposition or
dehydriding was found to follow zero order kinetics.
The effect of the shape of uranium particles on
A diffusion based hydriding mechanism was considered and using
basic symmetrical particle shapes like cylinders, slabs and spheres for
the hydriding kinetics under isothermal and
isobaric was studied.
mathematical and experimental analysis, it was shown that the particle
shape had negligible influence on overall hydriding kinetics.
Development of a mathematical model considering Experimental data and the mathematical model developed were used to
four steps in series (nucleation, skin development, evaluate the kinetic parameters for the various steps in hydriding.
skin growth, and final saturation) for hydriding of
uranium was carried out.
A survey of the existing literature for various
Emphasis was given on identifying the gap areas from the materials
properties of the uraniumehydrogen system was
science perspective before uranium can be effectively used as a long
carried out to identify areas which require further term, reliable storage medium for hydrogen isotopes. The work
study for making extensive use of uranium for solid presents a fair amount of thermodynamic and kinetic information
state storage of hydrogen isotopes in the nuclear
about this system as also highlights issues like accidental air ingress
industry.
into a storage system and the release of helium gas from the solid
matrix when it is used for tritium storage.
Comparisons of the predicted rates of hydriding and dehydriding were
The various rate models for hydriding and
made between the most well known models and rate equations viz. the
dehydriding of uranium based spent nuclear fuel
sent for final disposal in geological repositories
Bloch-Mintz and Condon-Kirkpatrick models. It was seen that the
second model provides higher rates of hydriding as also dehydriding at
were reviewed critically for various conditions of
temperature and pressure, with a focus on
any given temperature and hydrogen overpressure.
ensuring the safety and reliability of the
multi-canister overpacks for spent fuel storage.
The hydriding kinetics of activated metallic
Kinetic data were obtained and explanations were provided as to the
uranium was studied at near equilibrium
origin of the parabolic pressure dependence of the linear hydriding rate
pressure over a temperature range of
constant by assuming a nucleation and growth model for the reaction
290e370 deg C to arrive at the reaction mechanism. for the low pressure ranges. At higher pressures there is a shift in the
pressure dependence, as noted by various other researchers as well.
An investigation of the fundamental mechanisms Sticking probabilities and binding energy of the gases to the metal
of reaction of hydrogen and moisture with clean
surface have been calculated. Desorption rates of hydrogen from the
and contaminated uranium surfaces was carried
surface have also been evaluated.
out using sophisticated experimental techniques.
Temperature programmed desorption and
Two modes of hydrogen adsorption on uranium were observed-a non
supersonic molecular beams have been used to
activated mode and an activated mode. Sticking probabilities were also
study the dynamics of hydrogeneuranium
calculated.
reaction on very clean metallic surfaces.
Reaction kinetics of high purity uranium rods with Linear rate laws as function of temperature and pressure were obtained
hydrogen was studied over the temperature range and reported. The reaction mechanisms corresponding to the different
of 96e400 deg C and hydrogen overpressure
conditions of reaction were not explicitly studied or discussed but it was
of 70e430 mm Hg.
mentioned that there could be at least two different rate controlling
regimes in the range of conditions examined in this work.
Reaction kinetics of uranium and hydrogen
Owing to the large pressure range examined, different pressure
over a significantly large range of pressure and
dependencies of the rate law were obtained. An extensive mathematical
temperature was determined and two different
analysis was also performed to arrive at a rate law that takes into account
reaction models were used to explain the results.
both the hydriding and dehydriding processes in a single equation.
The hydriding rate was found to have a first order dependence on the
Reaction kinetics of fine uranium powder and
hydrogen was studied and a new reaction
mass of uranium remaining unreacted and a hydrogen pressure
mechanism was proposed to explain the
dependence of 0.5. Dehydriding appeared to follow zero order kinetics.
observations.
Reaction kinetics of massive uranium with
Linear hydriding rates were shown to follow 0.5 order pressure
hydrogen was studied over a wide range of
dependence. Isotope effects were determined to be less significant
pressure and temperature. The effect of
for this reaction. The importance of previously reported studies in
conditions on the formation of the alpha or
explaining the rate data over different ranges of conditions was also
the beta hydride of uranium was also studied.
highlighted.
Isotope effects have also been considered.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

8919

Table 1 e (continued )
Serial
number
[Reference]
13 [16]

Area of focus

Detailed description of a complete system for

hydriding and dehydriding of uranium for the
purpose of hydrogen isotope storage has been
provided. Representative data showing the
kinetics of uptake and release of hydrogen
have been included.

Fig. 1 e Schematic diagram of the shrinking core model of

uraniumehydrogen reaction.

The work presents an analysis of the hydriding reaction

of metallic uranium with hydrogen using the well known
shrinking core model under the pseudo-steady state hypothesis for gas transport through various transport resistances in series. Unlike most analyses of this nature, no
controlling regime has been identified prior to the analysis in
this work and all the intermediate gas concentrations have
been calculated by iterative calculations based on the
pseudo-steady state assumption. Further, the change in the
overall dimensions of the particle as it is hydrided has also
been taken into account. This is because of the significant
difference in the density of metallic uranium and uranium
hydride due to which a constant particle size cannot be
assumed as the reaction progresses. The uraniumehydrogen
reaction to form the hydride follows fractional order kinetics
(i.e. non-linear kinetics) according to a significant amount of
data reported in literature and this makes it necessary to use
numerical techniques for the solution of the equations. For a
given value of the bulk gas concentration of hydrogen (which
may be assumed to be present in pure form or in an inert gas
(e.g. helium) carrier gas stream) and a given initial pellet
radius, the gas concentrations at the interfaces have been
calculated as function of the unreacted core radius of the
pellet. The reaction at the hydride-metal interface occurs
provided the partial pressure of hydrogen at that location is
above the equilibrium pressure of hydrogen over uranium at
the reaction temperature which has been assumed to
remain spatially and temporally constant over the entire
course of the reaction in this study. It is then possible to use
the results of these calculations to determine the time for

Main results

The procedure for activation of uranium for hydrogen adsorption has

been stated clearly. Temperature and pressure conditions for rapid
uptake of hydrogen have been identified. Isotope effects have also
been considered in both hydriding and dehydriding steps.

complete reaction of the pellet to form the hydride and thus

gauge the loading or hydrogen storage time in an actual
storage vessel. The complete reaction time is one of the most
important parameters used to judge the suitability of a
particular material for hydrogen storage applications, which
is what the focus of the present study is. Based on the
calculated gas concentration gradient along each resistance,
the controlling regime (i.e. external mass transfer, ash/hydride diffusion or chemical reaction) for the overall
hydrogen storage process can therefore be identified and it
may even be possible to choose or modifying some of the
operating conditions in such a way that the limitations on
the rate of the hydriding process can be overcome to some
extent. For commercial application of metal based hydrogen
storage systems (e.g. in automobiles where a large amount of
hydrogen has to be stored on board the vehicle at a time), a
metal like uranium may not be suitable owing to the low
gravimetric storage density available from it. Lighter metal
based systems will be more useful and desirable in this
context. But for applications in the field of fusion energy and
others which require handling hydrogen isotopes like
tritium, the various kinetic and thermodynamic advantages
offered by uranium are rather desirable, thereby making it a
good choice for a getter material for hydrogen isotope
management [2]. Study of hydrogeneuranium reaction by
the shrinking core model may therefore be looked at as a
case study to establish the analysis method. This analysis
can then easily be extended to any other metal-hydrogen
isotope combination for which the required reaction and
transport data are available.

Kinetics of the reaction of hydrogen and uranium

There are quite a few experimental studies on the kinetics of
the reaction of uranium with hydrogen (or its isotopes in
certain cases) that have been reported in literature [4e16].
Their major findings are summarised in Table 1. These studies
cover different ranges of the operating parameters like
hydrogen pressure, temperature and the nature of solid used
for the experiments.
The present work illustrates a technique that allows the
shrinking core model be used to study non-catalytic gas solid
reactions with non-linear kinetics and with no apriori
assumption of the rate controlling regime. The non-linear
kinetic data reported by Balooch and Hamza [10], taken from
Powell et al. [17]. have been used for the subsequent calculations in this work.

8920

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Table 2 e Physical and transport properties of gas and

solid phases
Property

Value/Expression

PMH2/RgT kg m3 (ideal gas law)

PMHe/RgT kg m3 (ideal gas law)
8.76*106((293.85 72)/(T72))
*(T/293.85)1.5 Pa s [23]
Viscosity of helium
3.674*107*(T0.7) Pa s [24]
Density of uranium hydride
10950 kg m3 [25]
Density of uranium
18960 kg m3 [25]
Diffusivity of hydrogen in helium 1.032*104*(T)1.74 cm2 s1 [26]
Diffusivity of hydrogen in
2.11*104 exp(-2300/RgT) [27]
thorium hydride
Density of hydrogen
Density of helium
Viscosity of hydrogen

Model equations and solution procedure

The shrinking core model [18] used for the analysis in this
work is shown schematically in Fig. 1. The core radius at any
time is rc and the overall pellet radius has a value of R at any
time, starting with Ro at the instant the reaction is assumed
to begin. Since uranium is a pyrophoric material, even more
so when it is in the form of small particles or a powder, the
selection of the starting size of the metal particle will actually depend on the handling conditions e.g. the temperature
and whether it is to be handled in air or under inert gas
blanketing [19]. As the reaction proceeds, the reactive core
radius decreases from Ro to rc and finally to zero when the
pellet has completely reacted to form the hydride. Thus there
is a moving reaction interface or front which grows smaller

with time and a particle whose overall dimension grows with

time. The layer of solid product (uranium hydride in this
case) accumulates around the core region and its thickness
increases as reaction proceeds with consumption of the
uranium, in such a way that there is an increase in the
overall particle dimension with time as the reaction progresses. This takes into account the differing densities of
solid reactant and product. It is assumed that the ash or solid
product layer is porous enough through which reactant gas
(hydrogen) can diffuse while the core is impermeable to gas
and the reduction reaction takes place at the interface between the hydride shell (i.e. ash) and the metal core. The
relatively fast intrinsic surface reaction kinetics in this
particular system also enables intra-particle diffusion effects
to be neglected for the metallic core. Additionally isothermal
conditions have been assumed to be maintained and the gas
and pellet have also been assumed to be at the same temperature throughout the entire time of reduction, for the
purpose of establishing the technique and the trends in this
reacting system. The bulk or feed concentration Cb is taken to
remain unchanged with time (owing to unchanging bulk
conditions of hydrogen storage) while Ca and Cc change as
rc and R change on account of the reaction. In actual
hydriding of uranium it often happens that the outer hydride
layer flakes and falls off as it expands and thus allows easy
access of the metal to hydrogen gas and minimises diffusional resistance through the hydride shell [20]. In the present work the hydride shell has been assumed to remain
intact and adherent with the metal core, thus effectively
allowing the worst case scenario likely to arise (with respect

Fig. 2 e Effect of temperature on uraniumehydrogen reaction characteristics (a: uraniumeuranium hydride interfacial gas
concentration, b: Thiele Modulus, c: Core radius variation with time, d: Overall particle radius variation with time).

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

8921

Fig. 3 e Effect of feed gas concentration on uraniumehydrogen reaction characteristics (a: uraniumeuranium hydride
interfacial gas concentration, b: Thiele Modulus, c: Core radius variation with time, d: Overall particle radius variation with time).

to the time taken for hydrogen storage) to be mathematically

modelled.
The chemical reaction for hydrogen storage on uranium is
represented as [3].
1

U 1:5H2 UH3 ;DHrx 87:5kJmol H2

(1)

The reaction rate equation obtained from literature that is

considered in this work is [10] of the following form:
 rH2 kr Cnc

(2)

The following rate equations (Equation (3), (4) and (5)) can
be written for the external convective flux of hydrogen to the
particle surface, diffusive flux of hydrogen through the hydride layer and the reaction rate on the unreacted metal core
surface respectively at any instant of time [21]:
M1 4pRt2 kg bCb  Ca

(3)

4prc tbDe Ca  Cc



1  rc t=
Rt

(4)

M2

M3 4prc t2 bkr Cnc



rc kr Cnc 1  rc=R
r2c kr Cnc
Cb  2
Cc
De
R kg

(6)

From material balance, the relation between initial particle

diameter, particle diameter at any time t after the reaction
starts and the unreacted core radius at that time can be
expressed as follows:

13

 rU
Rt R3o  r3c
r3c
(7)
rUH3
The coefficient kg is calculated from the following correlation for convective mass transfer to a single spherical particle [22]:
Sh

kg 2*R
2 0:6 Sc0:333 Re0:5
p
DAB

(8)

The disappearance of uranium by reaction can be

described by the following equation:
(5)
drc bkr Cnc

dt
rU

where kr is obtained as [10].




0:5

3033
0:51 Rg T exp 
T

Under the pseudo steady state assumption, the rates M1,

M2 and M3 are all equal at any given time. This allows equation
(6) to be written by eliminating Ca to get an implicit equation
relating Cc with rc in terms of Cb and other known quantities.

(5a)

(9)

From the stoichiometry of the reaction under consideration, coefficient b has a value of 0.667. The total time required

8922

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Fig. 4 e Effect of particle size on uraniumehydrogen reaction characteristics (a: uraniumeuranium hydride interfacial gas
concentration, b: Thiele Modulus, c: Core radius variation with time, d: Overall particle radius variation with time).

for a uranium pellet to react completely (i.e. t actual) is then

obtained by numerically integrating Equation (9) between the
limits of rc R at t 0 to rc 0 to t t actual.
The various physical and transport properties of the reactant and product gases and solids used for the analysis are
presented in Table 2 below.
Cb is known from the conditions maintained during experiments or operational runs, as also the various kinetic and
transport coefficients from empirical equations. Thus equation (6) can be solved iteratively to obtain Cc as a function of
rc for the given set of conditions. The time of complete
conversion of a uranium pellet to the hydride can be obtained by numerically integrating equation (9) while accounting for the dependence of Cc on the instantaneous
value of rc from the calculations described in the previous
paragraphs. The effect of different reaction temperatures,
gas velocities and initial particle sizes has been subsequently
investigated.

Results and analysis

Parametric studies
The results of the calculations for some representative cases
are presented in this section. The conditions considered for the
evaluation of the interfacial concentrations were as follows:
Temperature: 20e50 deg C.

Pressure: 1 bar (a)

Hydrogen concentration: 0.1e100 % by volume in helium
(the actual value of the bulk concentration of hydrogen then
depends on the prevailing temperature and total pressure)
Uranium particle size (initial): 50 to 5000 micron.
Gas flow velocity: 0.5 m s1
The calculated values of Cc for the above conditions as
function of rc are presented in Figs. 2a, 3a and 4a on semilogarithmic axes.
It is observed that at every temperature considered, the
values of Cc are distributed almost symmetrically about the
core radius equal to half the pellet radius. At low values of rc
(i.e. when the reaction is almost complete and only a small
amount of uranium remains unreacted) the available surface
area for reaction is quite less and the ash layer thickness has
also increased, thus a higher concentration of reacting gas
should be available at the surface to ensure equality of
external and internal mass transfer rates. At high values of rc
(i.e. when the pellet has not reacted to any appreciable extent)
diffusional flux through hydride layer is quite high as it is thin,
so a relatively high concentration of gas is available at the core
surface too for reaction.
The above calculations depend strongly on the value of
effective diffusivity of hydrogen through the porous hydride
layer. This value again depends strongly on the reaction
conditions, especially temperature because the state of aggregation of the hydride layer is significantly affected by it.
Accurate experimental determination of the effective gas

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Table 3 e Complete hydriding reaction time under

various operating conditions.
T
R
% Assumed controlling
(C) (micron) H2
regime
20

50

10

100

20

500

10

100

50

50

10

100

50

500

10

100

Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction
Film diffusion
Ash diffusion
Chemical reaction

t calc
(sec)

t actual
(sec)

1.835
2.979*105
1.156*104
0.1835
2.979*104
3.656*103
0.0183
2.979*103
1.156*103
134.811
2.979*107
1.1561*105
13.476
2.979*106
3.656*104
1.337
2.979*105
1.156*104
1.729
2.275*105
4.420*103
0.173
2.275*104
1.398*103
0.0172
2.275*103
442.034
129.38
2.275*107
4.420*104
12.933
2.275*106
1.398*104
1.284
2.275*105
4.420*103

3.482*105

3.529*104

3.872*103

3.475*107

3.476*106

3.482*105

2.656*105

2.667*104

2.748*103

2.654*107

2.654*106

2.656*105

diffusivity would be useful for such calculations albeit

difficult to accomplish. Inaccurate values or estimates can
lead to wrong interpretation of the controlling regime of the
reaction. Since exact data for the diffusivity of hydrogen
gas through uranium hydride layer under such conditions
could not be obtained from literature, values of effective
diffusivity characteristic of similar systems (e.g. diffusivity
of hydrogen through thorium hydride [28]) have been used
in this work.
The results are presented in Figs. 2e4 for the cases
considered in this section. The reaction profiles indicate that
ash layer diffusion is likely to be the controlling resistance for
this system. The reaction kinetics is quite slow at the low
temperatures considered in this study but the ash layer
diffusion still controls the rate of the process.

Controlling regime identification

The calculated times for complete conversion of the uranium
pellet to uranium hydride under the assumption of a single
controlling regime at isothermal conditions with constant

8923

particle size for some representative operating conditions

have been shown in Table 3 below. The values are obtained
from equations in literature [18]. It shows that in each case,
the actual time of complete reaction of the uranium pellet
matches very closely with the time calculated assuming only
the diffusion controlled regime, thus indicating the dominant
resistance to be originating from the ash layer. The calculations indicate that the diffusional resistance is indeed controlling in the process studied here even when temperature,
reactant concentration and size of uranium pellet are
changed, though for low temperatures and high concentrations of hydrogen in the feed gas, chemical reaction control
becomes almost equally important.
The values of Thiele modulus, defined here as [7].
s
n 1 kr Cn1
c
MT 2R
2
De

(10)

were calculated for some typical operating conditions and

results are reported in Figs. 3b, 4b and 5b. The high values of
the modulus, even for particles as small as 0.1 mm in diameter
clearly indicate the dominance of ash diffusional resistance
over the reaction kinetics. The value rises to a maximum as
the reaction proceeds and then drops when the reactive core
has completely disappeared, showing the dynamic behaviour
of the reaction and diffusional resistances in the case of a noncatalytic gas solid reaction.

Assumption of unchanging overall particle size:

consequences
For most gas solid non catalytic reactions analysed using the
shrinking core model, the overall particle is assumed to
remain constant during the course of the reaction. In case of
the uraniumehydrogen reaction, this assumption was relaxed
in this present study to account for the significantly different
densities of the metal and its hydride. This section looks at the
variation in the reaction profiles that would occur if this difference were to be overlooked. Fig. 5 presents the variation of
the core size with time (i.e. conversion versus time data) for a
typical set of operating conditions, both with and without the
assumption of unchanging particle size. The time for
completion of the reaction is expectedly less for the case of
unchanging particle size, given that the overall diffusional
resistance is smaller in this situation (so the values of Thiele
modulus are also lower at each point of time). The difference
is about 15.7% for the case considered in Fig. 5 and this value
remains almost constant for any set of operating conditions
considered. This gives an estimate of the error in hydrogen
storage time when calculated by neglecting the growth of the
hydride particle.

Summary and conclusion

A simplified mathematical analysis has been performed for
estimating the time of complete reaction of a metallic uranium particle or pellet by hydrogen gas, present either in pure
form or alongside an inert carrier gas like helium, using the
well known shrinking core model. Reaction kinetic data

8924

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

Fig. 5 e Reaction profiles with and without the assumption of changing particle size during uraniumehydrogen reaction (a:
uraniumeuranium hydride interfacial gas concentration, b: Thiele Modulus, c: Core radius variation with time, d: Overall
particle radius variation with time).

obtained from literature were used for the analysis of the

process. A controlling regime for the overall storage process
was not identified beforehand and the necessary interfacial or
intermediate reactant concentrations were calculated as a
function of the instantaneous core radius while keeping all
other parameters unchanged by an iterative process based on
the pseudo-steady state hypothesis. Using the value of the
surface concentration of hydrogen actually available during
the reaction, the time for complete conversion of the pellet
was then calculated. The change in the overall particle
dimension owing to the significant difference in the densities
of the solid reactant and product was also accounted for in
this work. The calculated concentration gradients and reaction profiles enable one to determine which regime actually
controls the rate of hydrogen uptake by uranium and how, if
at all, the controlling regime changes with time. Diffusional
resistance to gas transport offered by the ash (i.e. uranium
hydride) layer was found to govern the kinetics of the process
overall. For heavier isotopes of hydrogen at similar temperature conditions, this resistance will be still larger, thus it may
be expected that the rate controlling regime would still remain
same. In reality though, the ash layer may not actually be
present during the hydriding or it may only have a partial
impact on the overall process kinetics since it tends to spall off
the surface and thus continually allows fresh metal surface to
be exposed to the reactant gas, thus totally eliminating the
ash diffusion aspect. The method illustrated here can be used

for any gasesolid reaction system with non-linear kinetics

and for any other particle shape, when the pseudo-steady
state assumption is taken as valid. Once the ash diffusion
controlling regime is conclusively established for this and
similar systems, further analysis can be significantly simplified. Isothermal conditions were assumed for this study but
the assumption may be relaxed to include temperature effects
in a more involved analysis in future. The analysis developed
for a single particle may also be extended to study the
behaviour of a packed bed of uranium particles which would
represent a typical hydrogen storage system on a larger scale.
Such an analysis is necessary and may be carried out for
various other alloy systems that have been studied in the
context of reversible hydrogen storage, as long as the kinetic
data and transport properties are even approximately known.

Nomenclature
b
Ca
Cb
Cc
DAB

stoichiometric coefficient for U in the reaction,

dimensionless
concentration of hydrogen gas on the ash layer,
mol m3
inlet or bulk concentration of hydrogen gas, mol m3
concentration of hydrogen gas, mol m3
concentration of hydrogen gas on the shrinking core,
mol m3
diffusivity of hydrogen through helium, m2 sec1

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 8 9 1 7 e8 9 2 5

De

effective diffusivity of hydrogen through the porous

metal hydride layer, m2 sec1
Dp
uranium pellet diameter, m
mass transfer coefficient for the spherical particle,
kg
m sec1
kr
intrinsic reaction rate constant, sec1
M1
rate of film diffusion, mol s1
M2
rate of ash diffusion, mol s1
M3
rate of surface reaction on unreacted core, mol s1
MH2
molecular weight of hydrogen, gm mol1
MHe
molecular weight of helium, gm mol1
MT
Thiele modulus, dimensionless
n
order of reaction (value 0.5 in this case),
dimensionless
core radius at any time, m
rc(t)
rate of reduction reaction, mol m3 s1
rH2
Ro
initial metal pellet radius, m
R(t)
pellet radius at any time t, m
universal gas constant, 8.314 J mol1 K1
Rg
Rep
particle Reynolds number, dimensionless
Sc
Schmidt number, dimensionless
Sh
sherwood number, dimensionless
t
time, sec
T
absolute temperature, K
rU
uranium molar density, kg m3
rUH3
uranium hydride molar density, kg m3
t actual actual time for complete reaction of U pellet, sec
calculated time for complete reaction of U pellet
t calc
based on one controlling regime, sec

references

[9]

[10]

[11]

[12]
[13]
[14]
[15]

[16]

[17]

[18]
[19]

[20]

[21]
[22]

[1] Rand DAJ, Dell RM. Hydrogen energy-challenges and

prospects. 1st ed. Cambridge: RSC Publishing; 2008.
[2] Shmayda WT, Heics AG, Kherani NP. Comparison of uranium
and zirconium cobalt for tritium storage. J Less-Common
Metals 1990;162:117e27.
[3] Katz JJ, Rabinowitch E. The chemistry of uranium. 1st ed.
United States of America: Mc-Graw Hill Book Company Inc;
1951.
[4] Stakebake JL. Kinetics for the reaction of hydrogen with
uranium powder. J Electrochem Soc 1979;126(9). 1956e1601.
[5] Chernov I, Bloch J, Voit A, Gabis I. Influence of metal powder
particle's shape on the kinetics of hydriding. Int J Hydrog
Energy 2010;35:253e8.
[6] Chernov I, Bloch J, Gabis I. Mathematical modeling of UH3
formation. Int J Hydrog Energy 2008;33:5589e95.
[7] Kolasinki RD, Shugard AD, Tewell CR, Cowgill DF. Uranium
for hydrogen storage applications: a materials science
perspective. In: Sandia national laboratories report No.
SAND2010e5195; 2010.
[8] Kirkpatrick JR, Mackay NS, Wheatley PD. Review of uranium
hydriding and dehydriding rate models in GOTH_SNF for

[23]

[24]

[25]

[26]

[27]

8925

spent fuel MCO calculations. In: DOE report No. DOE/SNF/

REP-0890; 2003.
Bloch J. The hydriding kinetics of activated uranium powder
under low (near equilibrium) hydrogen pressure. J Alloys
Compd 2003;361:130e7.
Balooch M, Hamza AV. Hydrogen and water vapor
adsorption on and reaction with uranium. J Nucl Mater
1996;230:259e70.
Bazley SG, Nunney TS, Mormiche C, Hayden BE. The
dynamics of hydrogen adsorption on polycrystalline
uranium. Appl Surf Sci 2008;254:6376e9.
Albrecht WM, Mallett MW. Reaction of hydrogen with
uranium. J Electrochem Soc 1956;103(7):404e9.
Bloch J, Mintz MH. Kinetics and mechanism of the U-H
reaction. J Less-Common Metals 1981;81:301e20.
Condon JB, Larson EA. Kinetics of the uranium-hydrogen
system. J Chem Phys 1973;59(2):855e65.
Powell GL, Harper WL, Kirkpatrick JR. The kinetics of the
hydriding of uranium metal. J Less-Common Metals
1991;172e174:116e23.
Paek S, Ahn D, Kim K, Chung H. Characteristics of reaction
between hydrogen isotopes and Depleted uranium. J
Industrial Eng Chem 2002;8(1):12e6.
Powell GL, Ceo RN, Harper WL, Kirkpatrick JR. The kinetics of
the hydriding of uranium metal II. Z fur Phys Chem
1993;181(1e2):275e82.
Levenspiel O. Chemical reaction engineering. 3rd ed. New
Delhi: Wiley India Pvt. Ltd; 2004.
Peacock HB. Pyrophoricty of uranium. Available from. In:
WSRC-TR-92e106 (DOE technical report); 1992 [Last accessed
19.09.13], http://www.osti.gov/scitech/servlets/purl/
10155050.
Maienschien JL. Assessment of a chemical getter for
scavenging tritium from an inert gas. In: Preprint UCRL77728, lawrence livermore laboratory; 1976.
Smith JM. Chemical engineering kinetics. 3rd revised ed. US:
Mc-Graw Hill Inc; 1981.
Dutta BK. Principles of mass transfer and separation
processes. 1st ed. New Delhi: Prentice Hall of India Pvt. Ltd.;
2007.
Wilcock DL. Properties of gases in CRC handbook of
lubrication. In: Bruce RW, editor. Theory and practice of
tribologyII. Boca Raton, Florida: CRC Press LLC; 1983.
Petersen H. The properties of helium: density, specific heats,
viscosity, and thermal conductivity at pressures from 1 to
100 bar and from room temperature to about 1800 K. In:
Danish atomic energy commission research establishment
report No-224; 1970. Kopenhagen, Denmark.
Riso
Penzhorn R-D. Tritium storage. In: Mannone F, editor. Safety
in tritium handling technology. Dordrecht: Kluwer Academic
Publishers Group; 1993.
Chung HK, Dalgarno A. Diffusion of hydrogen atoms in
helium gas and helium atoms in hydrogen gas. Phys Rev A
2002;66. 012712-1-3.
Libowitz GG. The actinide hydrides. In: Mueller WM,
Blackledge JP, Libowitz GG, editors. Metal hydrides. New
York: Academic Press; 1968.