Features of the Hydrogenation of Magnesium with a Ni-Graphene Coating

Article abstract of DOI:10.1134/S0036024420050222

Abstract: The kinetics of the hydrogenation of magnesium composites with graphene-like material (GLM) on which nickel particles are deposited (Ni/GLM) and dehydrogenation of MgH₂+Ni/GLM composites is studied. The mechanism of the reactions is revealed using the obtained results and the mathematical processing of curves with the Avrami–Erofeev equation. It is found that Ni/GLM considerably accelerates the hydrogenation of Mg and the decomposition of MgH₂ in the composite. The hydrogen-accumulating Mg+Ni/GLM composite is shown to have cyclic stability with rapid and virtually complete reversible hydrogenation.

Full text of DOI:10.1134/S0036024420050222

ISSN 0036-0244, Russian Journal of Physical Chemistry A, 2020, Vol. 94, No. 5, pp. 996–1001. © Pleiades Publishing, Ltd., 2020.
Russian Text © The Author(s), 2020, published in Zhurnal Fizicheskoi Khimii, 2020, Vol. 94, No. 5, pp. 772–777.
PROCEEDINGS OF THE CONFERENCE “PHYSICAL CHEMISTRY IN RUSSIA AND
BEYOND: FROM QUANTUM CHEMISTRY TO EXPERIMENT” (CHERNOGOLOVKA)
Features of the Hydrogenation of Magnesium
with a Ni-Graphene Coating
B. P. Tarasova,*, S. A. Mozhzhukhina, A. A. Arbuzova, A. A. Volodina, E. E. Fokinaa,
P. V. Fu r sik ov a, M. V. Lototskyyb, and V. A. Yartysc
a Institute of Problems of Chemical Physics, Russian Academy of Sciences, Chernogolovka, 142432 Russia
b University of the Western Cape, Bellville, 7535 South Africa
c Institute for Energy Technology, Kjeller, NO 2027 Norway
*e-mail: tarasov@icp.ac.ru
Received July 22, 2019; revised July 22, 2019; accepted September 17, 2019
Abstract—The kinetics of the hydrogenation of magnesium composites with graphene-like material (GLM)
on which nickel particles are deposited (Ni/GLM) and dehydrogenation of MgH2+Ni/GLM composites is
studied. The mechanism of the reactions is revealed using the obtained results and the mathematical process-
ing of curves with the Avrami–Erofeev equation. It is found that Ni/GLM considerably accelerates the
hydrogenation of Mg and the decomposition of MgH2 in the composite. The hydrogen-accumulating
Mg+Ni/GLM composite is shown to have cyclic stability with rapid and virtually complete reversible hydro-
genation.
Keywords: hydrogen, magnesium, carbon, magnesium hydride, hydrogen-accumulating composite, Avrami–
Erofeev equation
DOI: 10.1134/S0036024420050222
INTRODUCTION
graphene-like structure of the carbon material. Com-
posite particles also do not sinter during MgH2 dehy-
drogenation, due to the carbon coating.
Magnesium hydride is a promising material for
hydrogen storage, due to its high reversible weight
(7.6 wt % H2) and volumetric (110 g H/L) hydrogen
capacity [1–3]. We have patented ways of obtaining a
nickel–graphene catalyst for hydrogenation [4] and
hydrogen-accumulating materials based on magne-
sium [5]. The originality of our way of forming magne-
sium systems is the mechanochemical treatment of
magnesium with a Ni-graphene composite in hydro-
gen. This approach eliminates the main disadvantages
of magnesium as a hydrogen-accumulating material:
(1) the bad kinetics of hydrogenation because of the
formation of a continuous hydride layer through
which hydrogen atoms diffuse poorly; (2) the high
temperature of hydrogenation and dehydrogenation;
(3) the sintering of small magnesium particles during
the thermal decomposition of hydride; (4) and the
weak heat capacity of magnesium hydride, which hin-
ders the creation of a metal hydride accumulator [1, 2,
6]. A composite consisting of submicronic magnesium
particles with a Ni-graphene coating has good kinetics
of hydrogenation, due to the presence of catalytic Ni
particles and small Mg/MgH2 particles on the surface,
Studying the kinetics and mechanism of solid-
phase chemical reactions in metal–hydrogen systems
is of special interest, since it yields results allowing tar-
geted searches for materials with improved hydrogen-
sorption characteristics. The aim of this work was to
investigate the kinetics of the reversible hydrogenation
of magnesium particles with a Ni-graphene coating in
order to identify the limiting stages and the mecha-
nism behind the process.
EXPERIMENTAL
To prepare our composites, we used magnesium
powder with particle sizes of 0.5–1 mm and a nickel-
graphene catalyst obtained by combined reduction of
Ni(II) and graphite oxide according to the procedure
described in [6, 7]. This was a graphene-like material
(GLM) with a specific surface area of 600–650 m2/g,
on which Ni nanoparticles (25 wt %) were deposited
uniformly. The MgH2 composites with 10 wt % of the
Ni-graphene catalyst (MgH2+Ni/GLM) were made
via high-energy grinding in a Pulverisette-6 ball mill at
an initial hydrogen pressure of 25 atm. Hydrogen
and the high heat capacity resulting from the (99.9999% purity) from a laboratory metal hydride
996

FEATURES OF THE HYDROGENATION
997
(a)
(b)
α-MgH2
γ-MgH2
1
2
20
30
40
50
60
70
80
1 μm
2θ, deg
Fig. 1. (Color online) (a) SEM micrograph and (b) XRD patterns of the composite obtained in the (1) first and (2) after ten cycles
of hydrogen desorption–sorption.
accumulator was used. The volume of the steel grind- of magnesium hydride: α-MgH2 with a rutile-type
ing bowl was 80 mL; the ratio of the weight of steel
balls to the weight of the material being ground was
40/1; the mill rotation speed was 500 rpm; and the
grinding time was 10 h.
The phase composition of the resulting composites
were analyzed via powder X-ray diffraction (XRD)
using DRON-UM2 X-ray diffractometer (CuKα radi-
ation). The microstructure of the samples was studied
using Zeiss LEO SUPRA 25 scanning electron micro-
scope.
Kinetic hydrogen desorption and sorption curves
were measured on a Siverts-type laboratory apparatus
in the temperature range of 300–350°C and the pres-
sure range of 1.2–5.5 atm. The hydrogen pressure was
1.2 and ≤1.3 atm at the beginning and end of dehydro-
genation, respectively, so the conditions of the process
were close to isobaric. The experimental data were
mathematically processed using the Avrami–Erofeev
kinetics equation in the form
structure and γ-MgH2 with an orthorhombic struc-
ture. Figure 1 presents micrograph and the XRD pat-
tern of the composite. The positions of diffraction
peaks (Fig. 1b) correspond to the Bragg reflections of
α-MgH2 and γ-MgH2 phases with unit cell parameters
а = 0.4515 nm, с = 0.3019 nm and а = 0.4526 nm, b =
0.5448 nm, с = 0.4936 nm, respectively [8]. The
γ-MgH2 phase is known to be metastable under the
conditions of hydrogen sorption–desorption cycles,
and as we showed in [8, 9], disappears after the first
step of dehydrogenation. It follows from the scanning
electron microscopy (SEM) data (Fig. 1a) that the
particle size of magnesium hydride in the synthesized
MgH2+GLM and MgH2+Ni/GLM composites
ranged from 500 nm to several μm and did not change
sappreciably during cycles of hydrogen desorption and
sorption [8, 9], while the size of coherent scattering
domains (CSDs) grew for the α-MgH2 phase. After
the tenth cycle of hydrogen sorption, the size of the
CSD in the 110 direction is thus 30–40 nm (initial
value, 7–10 nm; Fig. 1b). In the MgH2 reference sam-
ple without additives, the particle size grows abruptly
during cycling, which indicates their sintering. We
may assume that introducing the GLM into the com-
posites prevents the sintering of submicronic magne-
sium particles under the conditions of the thermal
decomposition of hydride.
α = A(1 – exp[–(kt)n]),
(1)
where α is the fraction of the reacted phase at moment
in time t; A is the asymptotic α value when t → ∞; k is
the rate constant (its reciprocal is the characteristic
reaction time), and n is the Avrami factor.
The samples of MgH2 and MgH2 + 10 wt % GLM
(MgH2+GLM)
prepared
similarly
to
the
The Avrami–Erofeev equation that we used to
study the mechanism of hydrogen sorption and
desorption by the composites has been actively used in
the kinetic modeling of the reactions of the hydroge-
nation and dehydrogenation of hydrogen-accumulat-
ing materials based on magnesium [10–16]. Phase
MgH2+Ni/GLM composites were used as reference
objects.
RESULTS AND DISCUSSION
According to the XRD data, the composites transformation reactions in metal–hydrogen systems
obtained in synthesis contained two crystalline phases are known [14] to be controlled by mechanisms of
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 94 No. 5 2020

998
TARASOV et al.
1.0
0.8
0.6
0.4
0.2
(a)
(b)
1.0
0.8
0.6
0.4
0.2
6
4
1
5
3
2
3
2
1
0
5
10
15
20
0
1.0
0.8
0.6
0.4
0.2
4
8
12
16
20
Time, min
Time, min
(c)
(d)
1.0
0.8
0.6
0.4
0.2
2
4
3
2
1
1
3
0
4
8
12
16
20
0
5
10
15
20
25
30
Time, min
Time, min
Fig. 2. (Color online) (a) Hydrogen desorption curves at 350°C and 1 atm for samples of (1) MgH , (2) MgH +GLM, and
2
2
(3) MgH +Ni/GLM; (b) hydrogen sorption curves at 5.5 atm and 300°C for samples of (1–3) Mg and (4–6) MgH +Ni/GLM
2
2
in the (1, 4) first, (2, 5) fifth, and (3, 6) tenth hydrogenation cycles; (c) (1) curve 5 from panel (b) and its approximation by the
Avrami–Erofeev equation with (2) one and (3) two summands in the right side; (d) dehydrogenation curves for the
MgH +Ni/GLM composite at temperatures of (1) 320, (2) 330, (3) 340, and (4) 350°C.
2
both nucleation and growth (N-G) and diffusion determined by geometry and nucleation occurring at
grain boundaries [15].
(DF). The contribution from these complementary
mechanisms is determined by numerical value n of the
Avrami factor: if 0.5 ≤ n < 1.0, the reaction is con-
trolled only by diffusion; if 2.5 ≤ n ≤ 4.0, the reaction
proceeds exclusively by the N-G mechanism; and if n
ranges from 1.0 to 2.5, both mechanisms work in par-
Figure 2 shows the kinetic hydrogen desorption
and sorption curves for composites with Ni/GLM
additives and the reference samples. It is seen that all
desorption curves and the sorption curves for the Mg
sample without additives is well approximated by
allel. For instance, when n = 1 there is both diffusion Eq. (1). The situation is more complicated for the
Table 1. Results from approximating the kinetic dehydrogenation curves of the composites (Fig. 2a)
k, min–1
R2
Curve
Sample
A
n
1
2
3
MgH2
0.8836(2)
0.8820(4)
0.8847(6)
5.817(2)
4.22(1)
2.507(9)
2.520(3)
1.241(3)
0.962(5)
0.99983
0.99887
0.99355
MgH2+10% GLM
MgH2+10% Ni/GLM
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 94
No. 5
2020

FEATURES OF THE HYDROGENATION
999
Table 2. Results from approximating the kinetic hydrogena-
tion curves of Mg without additives (Fig. 2b, curves 1–3)
Mg hydrogenation at lower temperatures when MgH2
is not subjected to ball milling [18]) after ten dehydro-
genation–hydrogenation cycles. The characteristic
reaction time and the Avrami factor grow from 9 to
>15 min and from 1.2 to 1.4, respectively.
k, min–1
R2
Curve
Cycle
A
n
1
2
3
1
5
0.805(1) 9.40(2) 1.161(2) 0.99964
0.777(2) 13.74(5) 1.384(3) 0.99969
0.741(2) 15.98(5) 1.428(2) 0.99984
As we showed in [6], GLM layers cover the magne-
sium hydride particles during mechanochemical treat-
ment, preventing their agglomeration and sintering
during dehydrogenation. As a result of this and the
catalytic effect of Ni nanoparticles, the rate of hydro-
gen absorption by composites with Ni/GLM additives
is high in both the first and subsequent cycles of
repeated hydrogenation (Fig. 2b, curves 4–6). The
limiting α value grows to 0.92–0.95 (equaling the sum
А1 + А2 in Table 3), and 80% Mg hydrogenation is
reached in just 2–4 min. The considerable difference
between hydrogenation reaction rate constants k1 @ k2
for the two summands in the right side of the Avrami–
Erofeev equation, which describes the time depen-
dence of α, indicates that hydrogenation consists of
two stages: rapid and slow.
Note that the observed character of changes in the
kinetic parameters of repeated hydrogenation for the
investigated samples is similar to those for MgH2–
TiH2 nanocomposites obtained as a result of the ball
milling of Mg + Ti in hydrogen [12]. Performing ten
desorption–sorption cycles gradually reduces the con-
tribution from the slow stage A2/(A1 + A2) with a simul-
taneous drop in the characteristic reaction time for the
rapid stage (from 36 to 24 s) and an increase in the char-
acteristic reaction time (from 5 to 15 min) for the slow
one. Avrami factors for the rapid stage do not change
appreciably during the cycling (n = 1.3), while the slow
stage exhibits an increase in n, from 0.5 (diffusion-con-
trolled process) to 1.2 (mixed type: N-G and DF).
10
curves of hydrogen sorption by Mg+GLM and
Mg+Ni/GLM composites, and good approximation
was achieved with Eq. (1) whose right side consists of
at least two summands with different sets of parame-
ters: A1, k1, n1 and A2, k2, n2. Results from that approx-
imation are listed in Tables 1–3.
The curves in Fig. 2a and the data in Table 1 show
that introducing such additives as GLM or Ni/GLM
alter the dehydrogenation reaction mechanism in
comparison to MgH2 without additives. The Avrami
parameters are remarkably different. In addition,
composites with additives (Fig. 2a, curves 2 and 3)
exhibit substantially higher rates of transformation at
the beginning of dehydrogenation. When n = 2.5,
dehydrogenation of the MgH2 sample without addi-
tives proceeds according to the N-G mechanism. For
the composite with GLM additives, Avrami factor n =
1.25 means that the nucleation reaction is of zero
order; for the composite with Ni/GLM additives
(n = 1) there is nucleation at grain boundaries. For the
last two processes, it is very likely the contribution
from diffusion is the factor limiting the total rate of the
processes, as was assumed in [17]. The characteristic
reaction time becomes very short (less than 6 min for
MgH2). Introducing GLM additives slows it by
approximately 30%, and introducing Ni/GLM slows
it by approximately another 170%.
Figure 2d shows the kinetic dehydrogenation
curves for the MgH2 composite with Ni/GLM addi-
tives at four different temperatures in the range 320–
340°C, and rate constants k are given in Table 4. The
linear approximation of the lnk dependence on 1/T
yields the energy of activation for dehydrogenation:
Ea = 76 3 kJ (mol H2)−1, which is notably less than
In the dehydrogenation of MgH2 without additives,
the particles of the forming magnesium phase sinter
during the formation of agglomerates, deteriorating
the kinetics of hydrogen absorption in the subsequent
hydrogenation cycles appreciably (Fig. 2b, curves 1–
3). As is seen from Table 2, limiting α value for the the reported energies of activation for the dehydroge-
transformation of Mg into MgH2 falls from >0.8 to nation of magnesium-based materials. However, it is
worth noting that to calculate Ea correctly, the k values
0.75 as a result of the cycling (this is a typical value of
Table 3. Results from approximating the kinetic hydrogenation curves of Mg+Ni/GLM composites (Fig. 2b, curves 4–6)
Rapid stage
Slow stage
k2, min−1
5.4(1)
A2
A + A2
R2
A1 + A2
Curve
Cycle
k1, min−1
A1
n1
A2
n2
1
4
5
6
1
5
0.647(7) 0.566(1) 1.215(8) 0.270(8)
0.68(2)
0.50(2)
1.24(3)
0.92
0.95
0.93
0.29
0.20
0.16
0.99968
0.99972
0.99937
0.764(3) 0.474(1) 1.33(1) 0.188(7) 14.3(9)
0.7860(8) 0.422(1) 1.261(3) 0.148(4) 14.8(4)
10
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 94 No. 5 2020

1000
TARASOV et al.
Table 4. Dehydrogenation rate constants for MgH2+Ni/GLM
ACKNOWLEDGMENTS
composites at different temperatures
Our physicochemical studies of the materials were per-
formed using the facilities at the Institute of Problems of
Chemical Physics and scientific organizations of the South
Africa and Norway.
kcor
,
p(T)/p(T =
350°С)
−1 P0, atm
Curve T, °C
k, min
min−1
1
2
3
4
320 0.3359 3.10
330 0.4196 3.98
340 0.5342 5.07
350 0.7054 6.41
0.62
0.76
0.89
1.00
0.2089
0.3196
0.4739
0.7054
FUNDING
M.V. Lototskyy is grateful to the Department of Science
and Technology (DST, project KP6–S01/HySA) and the
National Research Foundation of South Africa (NRF,
grant 109092). V.A. Yartys would like to thank the Research
Council of Norway (RCN; project 285146).
This work is supported by BRICS STI Framework Pro-
gramme (project 064-RICS-MH). Russian co-authors are
grateful to the Ministry of Science and Higher Education
for the financial support (agreement no. 14.613.21.0087,
k values were obtained by approximating the kinetic curves in
Fig. 2d; k , after multiplying by the correction factor. P is the
cor
0
equilibrium pressure; p(T)/p(T = 350°С) denotes coefficients of
correction. The real pressure for all curves is P = 1.25 atm.
obtained from a series of isobaric experiments must be unique identifier RFMEFI61318X0087).
multiplied by a temperature-independent correction
coefficient in the form of the ratio p(T)/p(T = 350°С),
REFERENCES
which accounts for thermodynamic characteristics of
1. B. P. Tarasov, M. V. Lototskii, and V. A. Yartys’, Russ.
the considered composite–hydrogen system: the tem-
perature-dependent value of the equilibrium hydrogen
pressure P0 in the system. For diffusion-controlled
dehydrogenation with the investigated material
(Fig. 2a, curve 3), it was proposed in [17] that the fol-
lowing expression be used for p(T): p = T(1 –
(P/P0)1/2). Introducing this correction for our k values
J. Gen. Chem. 77, 694 (2006).
https://doi.org/10.1134/S1070363207040429
2. R. V. Lukashev, S. N. Klyamkin, and B. P. Tarasov, In-
org. Mater. 42, 726 (2006).
https://doi.org/10.1134/S0020168506070077
3. B. P. Tarasov, Int. J. Hydrogen Energy 36, 1196 (2011).
4. A. A. Arbuzov, S. A. Mozhzhukhin, A. A. Volodin,
P. V. Fursikov, and B. P. Tarasov, RF Patent
No. 2660232, Byull. Izobret., No. 19 (2018).
5. A. A. Arbuzov, S. A. Mozhzhukhin, A. A. Volodin,
P. V. Fursikov, and B. P. Tarasov, RF Patent
No. 2675882, Byull. Izobret., No. 36 (2018).
yields Ea = 124 1 kJ (mol H2)–1, which is consistent
with the data of other works (e.g., [19–22]) where sim-
ilar systems were examined: the energy of dehydroge-
nation activation is 121.3 kJ (mol H2)−1 for MgH2 +
10% NiCl2 (compared to 160 kJ (mol H2)−1 for the
dehydrogenation of pure MgH2 [18]).
6. B. P. Tarasov, A. A. Arbuzov, S. A. Mozhzhukhin,
A. A. Volodin, and P. V. Fursikov, J. Struct. Chem. 59,
830 (2018).
https://doi.org/10.26902/JSC20180411
CONCLUSIONS
7. M. V. Klyuev, A. A. Arbuzov, N. A. Magdalinova,
P. A. Kalmykov, and B. P. Tarasov, Russ. J. Phys.
Chem. A 90, 1749 (2016).
Adding GLM, and especially Ni/GLM, greatly
increases the rates of hydrogen sorption and desorp-
tion in magnesium composites. The dehydrogenation
reaction rate constant grows by 1.4 and 2.3 times for
Mg + GLM and Mg + Ni/GLM, respectively, com-
pared to that of Mg without additives. The hydrogena-
tion rate constant for composites with Ni/GLM addi-
tives is 15–37 times higher than the one for the sample
without additives. For all processes, the effect of addi-
tives is also expressed as a substantial reduction in the
Avrami factor, testifying to the geometric constraint
on the regions of nucleation (with the rate of this pro-
cess growing) and the contribution from H atom diffu-
sion becoming the limiting step. Adding Ni/GLM
improves the cyclic stability as well. It provides rapid
and almost complete repeated hydrogenation for at
least ten cycles. During cycling, the contribution from
the slow process to the kinetics of hydrogen sorption
diminishes without reducing the hydrogen capacity of
the composite.
8. P. V. Fursikov and B. P. Tarasov, Russ. Chem. Bull. 67,
193 (2018).
https://doi.org/10.1007/s11172-018-2058-y
9. A. A. Arbuzov, S. A. Mozhzhukhin, A. A. Volodin,
P. V. Fursikov, and B. P. Tarasov, Russ. Chem. Bull.
65, 1893 (2016).
https://doi.org/10.1007/s11172-016-1530-9
10. R. V. Denys, A. B. Riabov, J. P. Maehlen, et al., Acta
Mater. 57, 3989 (2009).
11. L. Z. Ouyang, X. S. Yang, M. Zhu, et al., J. Phys.
Chem. C 118, 7808 (2014).
12. M. Lotoskyy, R. Denys, V. A. Yartys, et al., J. Mater.
Chem. A 6, 10740 (2018).
13. L. Z. Ouyang, Z. J. Cao, H. Wang, et al., J. Alloys
Compd. 586, 113 (2014).
14. L. Ouyang, Z. Cao, H. Wang, et al., J. Alloys Compd.
691, 422 (2017).
15. T. Forde, J. P. Maehlen, V. A. Yartys, et al., Int. J. Hy-
drogen Energy 32, 1041 (2007).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 94
No. 5
2020

FEATURES OF THE HYDROGENATION
1001
16. M. Lototskyy, J. Goh, M. W. Davids, et al., Int. J. Hy- 20. J. F. Fernández and C. R. Sánchez, J. Alloys Compd.
340, 189 (2002).
drogen Energy 44, 6687 (2019).
https://doi.org/10.1016/S0925-8388(02)00120-2
https://doi.org/10.1016/j.ijhydene.2019.01.135
21. V. A. Yartys, M. V. Lototskyy, E. Akiba, et al., Int. J.
17. P. S. Rudman, J. Appl. Phys. 50, 7195 (1979).
Hydrogen Energy 44, 7809 (2019).
https://doi.org/10.1016/j.ijhydene.2018.12.212
18. C. M. Stander, Z. Phys. Chem. Neue Folge 104, 229
22. B. P. Tarasov, A. A. Arbuzov, S. A. Mozhzhuhin, et al.,
(1977).
Int. J. Hydrogen Energy 44, 29212 (2019).
https://doi.org/10.1016/j.ijhydene.2019.02.033
19. J. Mao, Z. Guo, X. Yu, et al., Int. J. Hydrogen Energy
35, 4569 (2010).
Translated by L. Chernikova
https://doi.org/10.1016/j.ijhydene.2010.02.107
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 94 No. 5 2020