NaBH4 formation mechanism by reaction of sodium borate with Mg and H2

Article abstract of DOI:10.1016/j.jallcom.2006.07.119

It has been reported that sodium (potassium) borohydride can be formed by reaction of sodium (potassium) borate with Mg and hydrogen or magnesium hydride. However, few investigations on reaction mechanism have been reported. Here, we studied the NaBH₄ formation mechanism of Mg + NaBO₂ + 2H₂ = NaBH₄ + MgO through morphology observations, structure and micro composition analyses. It was found that when heating the reactor to 400 °C, NaBO₂ particles were agglomerated with Mg particles and a NaBO₂ network was formed due to the sintering effect. With further heating the reactor, a porous product layer (composed of NaBH₄ and MgO) was formed on Mg particles. It was found that no matter whether Mg was hydrogenated or not, Mg could react with sodium borate and hydrogen to form sodium borohydride. Elevating the reaction temperature was of benefit to NaBH₄ formation. Higher NaBH₄ yield can be obtained by using partially hydrogenated then dehydrogenated Mg.

Full text of DOI:10.1016/j.jallcom.2006.07.119

Journal of Alloys and Compounds 437 (2007) 311–316
NaBH formation mechanism by reaction of sodium borate with Mg and H2
4
a,∗
b
a
c
c
Z.P. Li , B.H. Liu , J.K. Zhu , N. Morigasaki , S. Suda
a
Department of Chemical and Biochemical Engineering, Zhejiang University, Hangzhou 310027, PR China
b
Department of Materials and Engineering, Zhejiang University, Hangzhou 310027, PR China
c
Materials & Energy Research Institute Tokyo Ltd., 1-1 sawarabi-daira tateshinachuo-kogen, Kitayama, Chino-shi, Nagano, Japan
Received 6 July 2006; received in revised form 26 July 2006; accepted 26 July 2006
Available online 1 September 2006
Abstract
It has been reported that sodium (potassium) borohydride can be formed by reaction of sodium (potassium) borate with Mg and hydrogen or
magnesium hydride. However, few investigations on reaction mechanism have been reported. Here, we studied the NaBH formation mechanism of
Mg + NaBO + 2H = NaBH + MgO through morphology observations, structure and micro composition analyses. It was found that when heating
the reactor to 400 C, NaBO particles were agglomerated with Mg particles and a NaBO network was formed due to the sintering effect. With
further heating the reactor, a porous product layer (composed of NaBH and MgO) was formed on Mg particles. It was found that no matter whether
Mg was hydrogenated or not, Mg could react with sodium borate and hydrogen to form sodium borohydride. Elevating the reaction temperature
was of benefit to NaBH formation. Higher NaBH yield can be obtained by using partially hydrogenated then dehydrogenated Mg.
2006 Elsevier B.V. All rights reserved.
4
2
2
◦
4
2
2
4
4
4
©
Keywords: Sodium borohydride formation mechanism; Magnesium; Magnesium hydride; Sodium borate; Reaction temperature
1
. Introduction
researches are needed on the development of new borohydride
synthesis technology to effectively lower down the cost.
Sodium borohydride is usually synthesized from NaH and
methyl borate, which was invented by Schlesinger et al. [36]:
Borohydrides are a group of compounds with high hydrogen
contents. Recently they have been attracted more attentions
as a hydrogen storage medium for hydrogen generation or
as a fuel for the direct borohydride fuel cell (DBFC) due to
their high hydrogen storage capacity. Many papers related
to hydrogen generation from borohydrides [1–14] and the
DBFC [15–26] have been published. Comparing with topics
of hydrogen generation from borohydrides and the DBFC, few
papers related to borohydride synthesis were published [27–32].
Recent progresses on hydrogen generation from borohydrides
and the DBFC have been summarized in review articles
4NaH + B(OCH3)3 = NaBH4 + 3NaOCH3
(1)
Beside sodium hydride, alkali-earth metal or their hydrides
have been used for reaction with NaBO2 to prepare NaBH4
[27–31,37]. However, the detailed reaction mechanism has not
been reported. In this paper, we studied the NaBH4 formation
mechanism through a reaction of dehydrated sodium borate with
◦
Mg and hydrogen at 550 C based on morphology observa-
[
33–35].
Currently the price of sodium borohydride is about 55 $/kg.
tions, structure and micro composition analyses. Based on them,
effects of magnesium hydrogenation and reaction temperature
on sodium borohydride formation were investigated.
In order to enlarge borohydride applications for hydrogen gen-
eration and the DBFC, the cost reduction of the borohydride
production is necessary. If the cost of borohydride production
can be reduced to 0.55 $/kg, borohydride will be a powerful com-
petitor as a hydrogen storage medium. Therefore, more detailed
2
. Experimental details
In this research, Mg powder (purity: 99.9%, <200 mesh), sodium borate
powder (purity: 98%, <50 mesh) and hydrogen (purity: 99.99%) were used as
the reactants. A diagram of test apparatus is illustrated in Fig. 1. Confirmation of
sodium borohydride formation was conducted by XRD. The borohydride yield
was determined by hydrogen consumption amount or iodimetric analysis [38].
∗
Corresponding author. Tel.: +86 571 87951977; fax: +86 571 87953149.
E-mail address: zhoupengli@zju.edu.cn (Z.P. Li).
0
925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2006.07.119

3
12
Z.P. Li et al. / Journal of Alloys and Compounds 437 (2007) 311–316
Fig. 3. Hydrogen consumption behavior during reaction of hydrogen with Mg
or Mg + NaBO2.
Fig. 1. Diagram of test apparatus for NaBH4 formation.
The stoichiometric Mg powders and NaBO2 powders were well mixed
according to the following reaction:
3. Results and discussion
◦
NaBO2 + Mg + 2H2 = NaBH4 + MgO,
ꢀG = −342 kJ/mol-NaBH4
3.1. Sodium borohydride formation
(
2)
Fig. 3 shows the hydrogen consumption behaviors when
hydrogen reacted with Mg or a mixture of Mg and NaBO2. In
both of the H2 + Mg and the H2 + Mg + NaBO2 system, there was
then put into a stainless steel reactor in which a thermocouple was placed in the
reactant bed. The reactor was degassed during heating process before charge of
hydrogen into the reactor. The hydrogen pressure was set at 3.1 MPa. Heating
◦
rate was 10 C/min.
a small hydrogen pressure drop when heating the reactor from
The samples were subjected to SEM, EPMA and XRD analyses at different
stages of borohydride formation. The product (NaBH4) was extracted by anhy-
drous ethylenediamine with a purity of 99%. Hardened filter paper was used to
separate the extraction solution from formed oxides and remaining reactants.
The solid product (white powder) was obtained by evaporating the extraction
solution under 0.01 MPa at room temperature. A cold trap was applied to cap-
◦
4
00 to 426.8 C. It can be attributed that MgH2 formed according
to Van’t Hoff plot of Mg–H2 system [39]. Because the heating
◦
rate was 10 C/min, Mg only can absorb little hydrogen within
2.6 min. Then the hydrogen pressure of the system increased a
◦
little when heating reactor from 426.8 to 525 C due to dehy-
◦
ture the solvent vapor (ethylenediamine) at −10 C. The obtained powders were
drogenation of the formed MgH2. In following processes, the
H2 + Mg and the H2 + Mg + NaBO2 system began to show dif-
ferent hydrogen consumption behaviors.
subjected to XRD analysis for qualitative identification of the borohydride for-
mation. The amount of formed borohydride was quantitatively determined by
iodimetric analysis.
To investigate the effect of Mg hydrogenation on borohydride formation, we
did following experiments as shown in Fig. 2:
In the H2 + Mg system, while the reactor was heated
to 550 C, then held at 550 C for 185 min, no hydrogen
pressure drop was found. The reactor was then cooled from
◦
◦
•
•
•
Experiment 1: introduce hydrogen into the reactor after degassing the reactor
at 25 C, then heating the reactor to 550 C.
◦
◦
Experiment 2: introduce hydrogen into the reactor after degassing the reactor
◦
◦
at 400 C, and then heating the reactor to 550 C.
Experiment 3: introduce hydrogen into the reactor after degassing the reactor
◦
at 550 C.
Fig. 2. Hydrogen introduction into the reactor at different temperatures. Hydro-
gen pressure, 3.1 MPa.
Fig. 4. Confirmation of NaBH4 formation by XRD.

Z.P. Li et al. / Journal of Alloys and Compounds 437 (2007) 311–316
313
Fig. 5. SEM images in the process of NaBH4 formation.
◦
5
50 to 426.8 C, the system hydrogen pressure was back
drying. The XRD peaks clearly show the existence of NaBH4.
These results strongly proved that NaBO2 reacted with Mg and
H2 to form NaBH4 and MgO.
to original pressure. In the following cooling process, the
system showed a big drop in hydrogen pressure. It can be
attributed that Mg began to absorb hydrogen because magne-
◦
sium hydride formed below 426.8 C at 3.1 MPa of hydrogen
3.2. Morphology during NaBH formation
4
pressure according to Van’t Hoff plot of Mg–H2 system
[
39].
In the H2 + Mg + NaBO2 system, a big drop in hydrogen pres-
Fig. 5 shows the morphology changes of Mg and NaBO2
particles at each stage of the NaBH formation. Particles were
4
◦
sure occurred at 525 C and hydrogen pressure kept decreasing
in the following process of heating the reactor to 550 C and
holding the reactor at 550 C for 185 min. It can be concluded
that there were some hydrides formed at 525 C, which were
identified by EPMA through surface composition analyses (Na
distribution and Mg distribution) as shown in Fig. 6. It can be
seen that particles with smooth surfaces were Mg particles, and
◦
◦
◦
those with rough surfaces were NaBO . When heating the reac-
2
◦
different from magnesium hydride.
tor to 400 C, NaBO particles were agglomerated with Mg
2
The obtained samples from the H2 + Mg and the
H2 + Mg + NaBO2 system were subjected to XRD analysis. The
X-ray diffraction diagram verified the existence of MgH2 in the
H2 + Mg system and the existence of NaBH4 and MgO in the
H2 + Mg + NaBO2 system as shown in Fig. 4. It implied that
NaBO2 was converted to NaBH4, and Mg was oxidized into
MgO in the H2 + Mg + NaBO2 system. Fig. 4 gives the XRD
result of the sample obtained from H2 + Mg + NaBO2 system
after extraction with anhydrous ethylenediamine, filtration and
particles and a NaBO network was formed due to the sinter-
ing effect. While further heating the reactor, a porous product
2
layer (composed of NaBH and MgO) was formed on Mg par-
4
ticles. It was found that MgO and NaBH were homogeneously
4
distributed in the product layer as shown in Fig. 6(b-2) and (b-
3). This porous layer was of benefit to the NaBH formation
4
because it would not be a barrier against hydrogen molecules
to access to boundaries of Mg and NaBO . From our previous
2
results as shown in Fig. 6 from (c-1) to (c-3) [30], there was

3
14
Z.P. Li et al. / Journal of Alloys and Compounds 437 (2007) 311–316
Fig. 6. Mg and Na distribution before and after NaBH4 formation.
a clear boundary between the metallic Mg and the mixture of
MgO and NaBH4.
3
.3. Effect of reaction temperature on NaBH4 formation
Fig. 7showstherelationofNaBH4 yieldvs. reactiontempera-
ture. Thoughreaction (2)isanexothermalreaction, NaBH4 yield
increased with elevating the reaction temperature. It implied that
NaBH4 formation was controlled by a mass diffusion process.
Through morphology observations mentioned above, it can be
speculated that the process of sodium borohydride formation
would take place according to following steps:
(
a) NaBO2 particles were agglomerated with Mg particles at
◦
2−
4
00 C, then form a NaBO2 network, through which O
could diffuse to surfaces of Mg particles.
Fig. 7. NaBH4 formation at different reaction temperatures.

Z.P. Li et al. / Journal of Alloys and Compounds 437 (2007) 311–316
315
Fig. 8. NaBH4 formation when introducing hydrogen into the reactor at different temperatures.
◦
−
hydrogenated), and then were dehydrogenated from 426.8 C.
In experiment 2, few magnesium hydrides were formed during
(
b) O2 from NaBO2, invaded into metallic Mg lattice to leave
vacancies in the NaBO2 lattice.
c) Hydrogen was dissociated to absorb on Mg surfaces.
d) Mg was combined with O to release two electrons to
◦
heating from 400 to 426.8 C within 2.6 min (about 2.5 wt.%
of Mg was hydrogenated), and then were dehydrogenated from
26.8 C. In experiment 3, no Mg was hydrogenated because
the reactor temperature was higher than the magnesium hydride
dissociation temperature.
(
(
2−
◦
4
absorbed H on Mg surfaces.
−
(
e) Absorbed H received electron to form H , and then move
to the vacancies for combining with boron to form borohy-
dride.
Fig. 8 shows the NaBH4 formation when introducing hydro-
gen into the reactor at different temperatures (experiments 1–3).
Elevating reaction temperature would be of benefit to O2
−
The NaBH yields were calculated from the hydrogen pressure
drop. It was found that no matter whether Mg was hydrogenated
4
diffusion in step (b), so that O2 ions could deeply invade into
−
or not, Mg could react with sodium borate and hydrogen to form
sodium borohydride. Higher NaBH4 yield can be obtained by
using partially hydrogenated then dehydrogenated Mg. It might
be attributed to that hydrogenation and dehydrogenation of Mg
metallic Mg lattice. Therefore, NaBH4 yield increased with ele-
vating the reaction temperature. Experimental results showed
◦
there was no NaBH4 formed at 360 C, but 4.8% of NaBH4
◦
yield was obtained at reaction temperature of 400 C. It can
generated some defects in Mg particles so that O2 ions could
deeply invade into metallic Mg lattice through these defects.
From Fig. 8, it was found that if the reactor temperature reached
−
be concluded that NaBH4 formation temperature was located
◦
between 360 and 400 C. Here, borohydride yield was deter-
minedbyiodimetricanalysis becausemagnesiumhydridewould
be formed in this temperature range.
◦
to reaction temperature of 550 C more quickly, the NaBH4 for-
mation was faster. This result was coincidence with our previous
conclusion that the rate of NaBH4 formation was proportional to
heating rate [31] and reconfirmed that the reaction temperature
played an important role in NaBH4 formation rate.
3
.4. Effect of Mg hydrogenation on NaBH4 formation
The start temperature of sodium borohydride formation was
located at the temperature range of magnesium hydride forma-
tion. It implied that some magnesium hydrides would be formed
during borohydride formation if reaction temperature was lower
than MgH2 dissociation temperature. Kojima et al. reported that
higher NaBH4 yield can be obtained by using MgH2 as the reac-
tant comparing with using Mg as the reactant [28]. However,
from our previous results, it was found there was no Mg hydride
existed during formation of NaBH4 [30].
4. Conclusions
◦
When heating the reactor to 400 C, NaBO2 particles were
agglomerated with Mg particles to form a NaBO2 network due to
the sintering effect. NaBH4 formation temperature was located
◦
between 360 and 400 C. It was found that no matter whether Mg
was hydrogenated or not, Mg could react with sodium borate and
hydrogen to form sodium borohydride. NaBH4 and MgO were
formed on surfaces of Mg particles. NaBH4 yield increased with
elevating the reaction temperature. Higher NaBH4 yield can be
obtained by using partially hydrogenated then dehydrogenated
Mg.
To find the role that Mg hydrogenation played during boro-
hydride formation, we did following experiments as described
in Section 2. In experiment 1, some magnesium hydrides were
formed during heating the mixture of Mg and NaBO2 from
◦
2
5 to 426.8 C under hydrogen (about 20 wt.% of Mg was

3
16
Z.P. Li et al. / Journal of Alloys and Compounds 437 (2007) 311–316
[18] Z.P. Li, B.H. Liu, K. Arai, S. Suda, J. Electrochem. Soc. 150 (7) (2003)
References
A868.
[
[
[
[
[
[
[
[
[
1] V.C.Y. Kong, F.R. Foulkes, D.W. Kirk, J.T. Hinatsu, Int. J. HydrogenEnergy
4 (1999) 665.
2] R. Aiello, J.H. Sharp, M.A. Matthew, Int. J. Hydrogen Energy 24 (1999)
123.
3] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, M.T. Kelly, P.J. Petillo,
M. Binder, J. Power Sources 85 (2000) 186.
4] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, N.C. Spencer, M.T.
Kelly, P.J. Petillo, M. Binder, Int. J. Hydrogen Energy 25 (2000) 969.
5] S. Suda, Y.M. Sun, B.H. Liu, Y. Zhou, S. Morimitsu, K. Arai, Y. Zhou, N.
Tsukamaoto, Y. Candra, Z.P. Li, Met. Mater. Korea 7 (2001) 73.
6] S. Suda, Y.M. Sun, B.H. Liu, Y. Zhou, S. Morimitsu, K. Arai, N. Tsukamoto,
M. Uchida, Y. Candra, Z.P. Li, J. Appl. Phys. A72 (2001) 209.
7] Y. Kojima, K. Suzuki, K. Fukumoto, M. Sasaki, T. Yamamoto, Y. Kawai,
H. Hayashi, Int. J. Hydrogen Energy 27 (2002) 1029.
[19] B.H. Liu, Z.P. Li, S. Suda, J. Electrochem. Soc. 150 (3) (2003) A398.
[20] M.V. Mirkin, H. Yang, A.J. Bard, J. Electrochem. Soc. 139 (8) (1992) 2212.
[21] E. Gyenge, Electrochim. Acta 49 (6) (2004) 965.
[22] B.H. Liu, Z.P. Li, S. Suda, Electrochim. Acta 49 (2004) 3097.
[23] H. Dong, R. Feng, X. Ai, Y. Cao, H. Yang, C. Cha, J. Phys. Chem. B 109
(21) (2005) 10896.
[24] J.-H. Kim, H.-S. Kim, Y.-M. Kang, M.-S. Song, S. Rajendran, S.-C. Han,
D.-H. Jung, J.-Y. Lee, J. Electrochem. Soc. 151 (7) (2004) A1039.
[25] Z.P. Li, B.H. Liu, K. Arai, K. Asaba, S. Suda, J. Power Sources 126 (1–2)
(2004) 28.
[26] Z.P. Li, B.H. Liu, K. Arai, S. Suda, J. Alloys Compd. 404–406 (2005) 648.
[27] Z.P. Li, N. Morigasaki, B.H. Liu, S. Suda, J. Alloys Compd. 349 (1) (2003)
232.
[28] Y. Kojima, T. Haga, Int. J. Hydrogen Energy 28 (2003) 989.
[29] Z.P. Li, B.H. Liu, N. Morigasaki, S. Suda, J. Alloys Compd. 354 (1–2)
(2003) 243.
[30] S. Suda, Y. Iwase, N. Morigasaki, Z.-P. Li, Advanced Materials for Energy
Conversion II, TMS, 2004, p. 123.
[31] S. Suda, N. Morigasaki, Y. Iwase, Z.P. Li, J. Alloys Compd. 404–406 (2005)
643.
[32] E. Jeon, Y. Cho, Mechanochemical synthesis and thermal decomposition
of zinc borohydride, J. Alloys Compd. 422 (2006) 273–275.
[33] Z.P. Li, B.H. Liu, K. Arai, N. Morigasaki, S. Suda, J. Alloys Compd.
356–357 (2003) 469.
[34] C. Ponce de Leon, F.C. Walsh, D. Pletcher, D.J. Browning, J.B. Lakeman,
J. Power Sources 155 (2) (2006) 172.
2
1
8] D. Hua, Y. Hanxi, A. Xinping, C. Chuansin, Int. J. Hydrogen Energy 28
(
10) (2003) 1095.
9] C. Wu, H. Zhang, B. Yi, Catal. Today 93–95 (2004) 477.
[
10] J.-H. Kim, K.-T. Kim, Y.-M. Kang, H.-S. Kim, M.-S. Song, Y.-J. Lee, P.S.
Lee, J.-Y. Lee, J. Alloys Compd. 379 (1–2) (2004) 222.
11] A. Zaluska, L. Zaluski, J. Alloys Compd. 404–406 (2005) 706.
12] J.-H. Kim, H. Lee, S.-C. Han, H.-S. Kim, M.-S. Song, J.-Y. Lee, Int. J.
Hydrogen Energy 29 (3) (2004) 263.
13] B.H. Liu, Z.P. Li, S. Suda, J. Alloys Compd. 415 (1–2) (2006) 288.
14] M. Aya, A. Midillib, I. Dincerc, J. Power Sources 157 (1) (2006) 104.
15] S. Amendola, P. Onnerud, M. Kelly, P. Petillo, M. Binder, Proc. Elec-
trochem. Soc. 47 (1999), 98-15 (selected Battery Topics);
[
[
[
[
[
S. Amendola, P. Onnerud, M. Kelly, P. Petillo, M. Binder, Proceedings of
the 194th Meeting of the Electrochemical Society, Boston, MA, November
[35] J.-H. Wee, J. Power Sources 155 (2) (2006) 329.
[36] H.I. Schlesinger, H.C. Brown, A.E. Finholt, J.R. Gilbreath, H.R. Hoekstra,
E.K. Hyde, J. Am. Chem. Soc. 75 (1953) 215.
1
–6, 1998.
[
[
16] S. Amendola, P. Onnerud, M. Kelly, P. Petillo, S. Sharp-Goldman, M.
Binder, J. Power Sources 84 (1) (1999) 130.
17] S. Amendola, P. Petillo, M. Kelly, S. Sharp-Goldman, M. Binder, Proceed-
ings of the 195th Meeting of the Electrochemical Society, Seattle, WA,
May 2–6, 1999.
[37] D. Goerrig, W. Schabacher, F. Schubert, DBP 1036222 (1956).
[38] D.A. Lyttle, E.H. Jensen, W.A. Struck, Anal. Chem. 24 (1952) 1843.
[39] P. Selvam, B. Viswanathan, C.S. Swamy, V. Srinivasan, Int. J. Hydrogen
Energy 11 (3) (1986) 169.