Effective hydrogen release from ammonia borane and sodium borohydride mixture through homopolar based dehydrocoupling driven by intermolecular interaction and restrained water supply

Article abstract of DOI:10.1039/d0ta04720a

Non-catalytic dehydrogenation of solid-state ammonia borane (AB) and sodium borohydride (SB) mixtures is achieved in this work by water vapor facilitated hydrothermolysis with a maximum release rate of 560 ml g^?1s^?1at 53.5 °C and dehydrogenation capacity between 7.8 wt% (measured) and 11.7 wt% (theoretical). A novel dehydrogenation pathway is identified, in which intermolecular interaction causes the formation of intermediates and further enables the prevalence of homopolar dehydrocoupling between -BH moieties in AB and SB. The low-temperature heat used in hydrothermolysis enhances the intermolecular interaction, and the oxygen and hydrogen supplied by water vapor promote the dehydrogenation by forming products. This new hydrogen liberation method can uniquely operate without any catalyst, liquid solvent, or high temperature, and provides a compelling solution for disposable hydrogen source application. The present work also unveils for the first time the unique hydride-hydride homopolar dehydrogenation mechanism, which may unlock the true potential of chemical hydride based hydrogen storage systems.

Full text of DOI:10.1039/d0ta04720a

Journal of
Materials Chemistry A
View Article Online
View Journal
PAPER
Effective hydrogen release from ammonia borane
and sodium borohydride mixture through
homopolar based dehydrocoupling driven by
intermolecular interaction and restrained water
supply†
Cite this: DOI: 10.1039/d0ta04720a
Pei Pei, Mark Cannon, Grace Quan and Erik Kjeang *a
a
b
b
Non-catalytic dehydrogenation of solid-state ammonia borane (AB) and sodium borohydride (SB) mixtures
is achieved in this work by water vapor facilitated hydrothermolysis with a maximum release rate of 560 ml
À1 À1
ꢀ
g
s
at 53.5 C and dehydrogenation capacity between 7.8 wt% (measured) and 11.7 wt% (theoretical). A
novel dehydrogenation pathway is identified, in which intermolecular interaction causes the formation of
intermediates and further enables the prevalence of homopolar dehydrocoupling between –BH moieties
in AB and SB. The low-temperature heat used in hydrothermolysis enhances the intermolecular
interaction, and the oxygen and hydrogen supplied by water vapor promote the dehydrogenation by
forming products. This new hydrogen liberation method can uniquely operate without any catalyst, liquid
solvent, or high temperature, and provides a compelling solution for disposable hydrogen source
application. The present work also unveils for the first time the unique hydride–hydride homopolar
dehydrogenation mechanism, which may unlock the true potential of chemical hydride based hydrogen
storage systems.
Received 6th May 2020
Accepted 4th September 2020
DOI: 10.1039/d0ta04720a
rsc.li/materials-a
Generally, certain catalytic additives are also needed to facilitate
the dehydrogenation process. Commonly used catalytic addi-
1
. Introduction
4
5
6
Hydrogen is a promising energy carrier that can be stored in tives include transition metals, metal salts, nanostructures
7
large quantities and used to produce power in fuel cells with no and protic/aprotic agents. Recent attempts of using different
harmful emissions. For mobile and stationary power applica- hydrides as catalytic additives were also reported, and several
tions, however, the energy density and storage duration of metal hydrides have been revealed to facilitate AB dehydroge-
8
existing hydrogen storage systems are generally inadequate. nation. Kang and co-workers showed that mechanical milling
Theoretically, solid state chemical hydrides could relieve these of AB with LiH results in 7 wt% hydrogen release at tempera-
ꢀ
constraints due to their high available hydrogen density and tures around 100 C through the yield of destabilized LiNH
BH
2
3
9
relatively benign dehydrogenation behaviour and long shelf life. material. Dovgaliuk et al. showed that the integration of AB
Ammonia borane (AB) and sodium borohydride (SB) as repre- with Al(BH forms a new coordination complex of Al(BH
sentatives of high capacity chemical hydrides possess 19.6 wt% $NH BH
4
)
3
4 3
) -
3
3
, in which AB can release two equivalent hydrogen
ꢀ
(
ref. 1) and 10.8 wt% (ref. 2) theoretical hydrogen gravimetric molecules at 70 C through producing Al(BH
4
)
3
$(BH–NH)
n
capacity, respectively. Strategies to effectively and efficiently intermediate. Similar efforts were also devoted on the combi-
release hydrogen from AB and SB at conditions that support nation of AB and SB, motivated by their high theoretical
practical applications are therefore desired.
hydrogen densities and benign dehydrogenation. However,
AB and SB can release hydrogen through thermolysis and intensive study has shown that mechanical milling and
10
hydrolysis. Hydrothermolysis is also used if heat, besides water, thermal treatment of an AB and SB mixture cannot release
3
is an essential driving force during dehydrogenation.
hydrogen below the decomposition temperature of pristine AB
at ꢁ104 ꢀC. The hydrolytic approach was also employed to
dehydrogenate the mixture of AB and SB, and showed that the
kinetic improvement of hydrogen evolution mainly derives from
the active Co catalyst reduced in situ by the mixture rather than
the interaction between AB and SB. These results indicate that
conventional thermolysis and hydrolysis processes are unable
to render facile hydrogen release from AB and SB mixtures.
a
Fuel Cell Research Laboratory (FCReL), Simon Fraser University, 250-13450 102
Avenue, Surrey, BC V3T0A3, Canada. E-mail: ekjeang@sfu.ca
b
Hydrogen in Motion (H2M), Unit 3B 2350 Beta Avenue, Burnaby, BC V5C 5M8,
11
Canada
†
Electronic supplementary information (ESI) available. See DOI:
0.1039/d0ta04720a
1
This journal is © The Royal Society of Chemistry 2020
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Here, we propose an original method to release hydrogen a condenser (trapping volatile gas), 0.1 N sulfuric acid solution
from AB and SB simultaneously through a novel dehydrogena- (removing alkaline gas), and silica gel (removing water vapor).
tion pathway which is dictated by intermolecular interaction The puried hydrogen gas was nally measured in volume by
and homopolar dehydrocoupling. No catalyst is required in this a burette. Prior to measurements, the system was purged by
process, and the mixing of AB and SB is achieved simply by high purity nitrogen for 10 min to eliminate residual air. The
blending. The method can be categorized as water vapor facil- supercial volume reading derived from thermal expansion of
itated hydrothermolysis, however, the functions of heat and gas in the system was corrected by subtracting the calibration
water to dehydrogenation are distinct from conventional test data from isothermal measurements. To start the dehy-
hydrolysis and thermolysis. Furthermore, only a trace amount drogenation experiment, the reactor ask containing ABSB and
ꢀ
of water vapor is utilized instead of liquid water to create water vapor source was transferred into an oil bath held at 55 C
a critical water rareed condition. The exemption from using while monitoring hydrogen evolution.
catalyst and large quantities of liquid water together with low-
temperature heat endues the potential of high effective
2.4 Characterization
capacity and facile operation.
The compositions of gases and volatile dehydrogenation prod-
ucts were analyzed by an ULT100 universal gas analyzer (UGA)
from Stanford Research Systems with maximum resolution of
2. Experimental
25 points per AMU. X-ray diffraction (XRD) spectra were
2.1 Chemicals
collected on a Bruker D8 Advance (Cu Ka) with 0.6 mm slit
Bragg–Brentano primary optics to identify phase structures of
solid residues. Solid state B magic angle spinning (MAS)
nuclear magnetic resonance (NMR) was measured on a Bruker
AVANCE III 400 system to examine chemical states of solid
residues. Raman spectra were collected on a Thermo Scientic
DXR2 Raman microscope system using 0.1 mW 532 nm exci-
tation laser and 25 mm slit aperture to probe dehydrogenation
pathway and mechanism.
Ammonia borane (97% purity), sodium borohydride (>99%
purity), deuterium oxide (99.9% purity), and silica gel (>99%
purity, 200–425 mesh) were purchased from Sigma-Aldrich.
Sulfuric acid (98% purity) was purchased from Alfa Aesar.
Nitrogen protection gas with purity of 99.999% was purchased
from Praxair. The resistivity of deionized water (DI water) used
11
À1
in this research was >18.1 MU cm . ND BH used in the
3
3
isotope experiment was prepared following the procedures in
12
literature.
3
. Results and discussion
2.2 Dynamic dehydrogenation
3
.1 Dehydrogenation performance and inuencing factors
Dynamic dehydrogenation was conducted on a custom-made
13
ꢀ
Sievert-type apparatus from 22 to 90 C with ramp rate of The proposed water vapor facilitated hydrothermolysis of AB
.5 C min . A dry glass vial containing 0.169 g ABSB of 3 : 2 and SB mixtures (ABSB) was rst demonstrated by dynamic
ꢀ
À1
0
mass ratio mixed by mild agitation was loaded into the sample dehydrogenation measurement with results shown in Fig. 1.
ꢀ
chamber with a water vapor source pre-loaded beneath it. The The release of gas started around 30 C and progressed with
ꢀ
sample chamber was sealed and leak tested, and then evacuated a slow release rate until 50 C. At this temperature, the gas
À4
to 2 Â 10 Pa for 10 min using a turbo molecular vacuum release rate increased drastically, and an eruption of gas was
ꢀ
À1 À1
pump. Heat was thereaer applied while the pressure and observed at 53.5 C with a peak release rate of 562.7 ml g
s .
À1
temperature were recorded with acquisition frequency of 1 s
.
The pressure and temperature data were converted to gas
volume using the ideal gas law, where the sample chamber
volume was obtained by helium gas calibration. The water vapor
concentration expressed in relative humidity (RH) was around
4
9% at the beginning of the dynamic experiment, which was
ꢀ
evaluated by directly measuring water vapor pressure at 25 C.
2.3 Isothermal dehydrogenation
Experimental apparatus for isothermal dehydrogenation
measurement is schematized in Fig. S1 (ESI†). The ABSB sample
was loaded into a dry beaker placed on the bottom of a dry
conical ask reactor also containing a water vapor source, in
this case is a small piece of wet cotton, separated from the ABSB
sample. A thermocouple was attached to the beaker containing
ABSB for temperature measurement, and a capillary tube was
inserted to extract gas samples for compositional analysis by
universal gas analyzer (UGA) prior to purication. The gas
Fig. 1 Dynamic gas release from the ammonia borane and sodium
ꢀ
borohydride mixture heated from room temperature to 90 C at
ꢀ
À1
0
.5 C min . The quantity of gas released was normalized by the initial
released from the reactor was puried by passing it through mass of the mixture.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
Journal of Materials Chemistry A
ꢀ
Further heating to 90 C did not signicantly improve the gas contained 86% (98.2%, if N
2
excluded) hydrogen and only 0.6%
À1
yield, and the total gas released was 1407 ml g
.
(0.7%, if N excluded) ammonia was identied as volatile by-
2
To understand the dehydrogenation behaviour and the product. The rest of the gas comprised of nitrogen as protec-
inuence of the water vapor concentration, isothermal tion gas and a trace amount of water vapor.
ꢀ
measurements were conducted at 55 C. Isothermal experi-
ments were performed with three different deionized water
sources to create low, medium, and high RH of 17% (ABSB-
3.2 Solid dehydrogenation residues
The solid residues from the three isothermal hydrothermolysis
experiments were analyzed by X-ray diffraction (XRD), shown in
Fig. 3. The XRD spectrum of the pristine ABSB mixture before
dehydrogenation was also measured (ESI, Fig. S3†) for refer-
ence. The residue from ABSB-LRH hydrothermolysis comprised
LRH), 48% (ABSB-MRH), and 97% (ABSB-HRH) in the reactor
ꢀ

ask at 55 C and 1 atm pressure. The RH values reect the
initial levels of water vapor concentration inside the reactor
measured by universal gas analyzer (UGA) in ABSB-free cali-
bration tests. The hydrogen release capacities of the obtained
ABSB hydrothermolysis isotherms, shown in Fig. 2, were
calculated aer purication and thermal expansion correction.
The hydrothermolysis of ABSB-LRH released 2.90 wt% of
hydrogen aer 64 min. Similarly, ABSB-HRH released 4.29 wt%
aer 134 min. Interestingly however, the hydrothermolysis of
ABSB-MRH gave a much larger hydrogen release of 7.78 wt%
aer 149 min. These experiments showed good reproducibility
within Æ0.5 wt%. There were no apparent induction periods for
any of the measurements, and immediate gas release was
observed upon activation with a predominantly consistent
release rate until the full capacity was reached in each case. The
measurements show that the RH is an important factor during
the hydrothermolysis of ABSB, and an appropriate water vapor
concentration is required in order to obtain a high degree of
dehydrogenation.
of NaB
unreacted AB and SB. The residue from ABSB-HRH hydro-
thermolysis contained a pentahydrated borate of Na B O -
3 5 2
O $2H O (JCPDS no. 20-1081) borate together with
2
4 7
$
5H O and similarly unreacted AB and SB. In contrast, the XRD
2
spectrum of ABSB-MRH shows a complex pattern identied as
a mixture of sodium borates, ammonium borate, and polymeric
borates. No unreacted AB or SB was detected in this case by
XRD, indicating a complete dehydrogenation of both AB and SB.
1
1
B magic angle spinning (MAS) nuclear magnetic resonance
NMR) was employed to further examine the chemical states of
(
the residues from dehydrogenation (ESI, Fig. S4†). In the ABSB-
LRH residue, resonances corresponding to unreacted solid AB
and SB together with tricoordinated and tetracoordinated
boron in borate environment were identied. The ABSB-HRH
spectrum resembles that of ABSB-LRH, showing AB, SB, and
borate resonances, but with evidence of higher degree of
dehydrogenation, especially for SB. In the case of dehydro-
genated ABSB-MRH, the resonance corresponding to SB is
undetectable, implying a complete dehydrogenation of SB. AB
was almost depleted as well, but a weak signal between À22 and
À30 ppm remained. This signal has similar chemical shis as
AB, but with apparent new resonance peaks indicating its new
Two control tests were conducted, where either mixing (CT 1)
or water vapor (CT 2) was omitted intentionally during hydro-
thermolysis (Fig. 2). Only marginal gas release was observed
from each test (#0.5 wt%). Thus, separated AB and SB failed to
produce a comparable amount of hydrogen even under optimal
(48%) RH, indicating the critical role of AB–SB interaction. Lack
of water vapor also largely prevented hydrogen release from
ABSB, showing that water vapor presence is required for the
desired reaction process to take place. The gas released from
ABSB-MRH was analyzed by UGA before being puried with
quantitative results appended (ESI, Fig. S2†). The gas detected
14,15
identity of AB mobile phase (AB*),
which features a rapidly
rotating B–N bond relaxing the extensive dihydrogen network in
AB and facilitating the formation of diammoniate of diborane
16
(
DADB). The AB* mobile phase found in ABSB-MRH was not
observed in the other two experiments where the majority of the
AB remained unreacted, implying a potentially important
Fig.
2
Isothermal hydrogen release from hydrothermolysis of
ꢀ
ammonia borane and sodium borohydride mixtures at 55 C with three
different water vapor concentrations of 17%, 48%, and 97% RH. Control
ꢀ
tests of isothermal hydrogen release at 55 C with either the separation
of AB and SB (CT 1) or absence of water vapor (CT 2) are also displayed Fig. 3 X-ray diffraction spectra of solid residues from 55 C dehy-
ꢀ
and compared. The capacity of H
2
release was measured in weight drogenation of (a) ABSB-LRH, (b) ABSB-MRH, and (c) ABSB-HRH. A
percent after purification.
phase identification list is provided in the inset.
This journal is © The Royal Society of Chemistry 2020
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
intermediate role of AB* for complete ABSB dehydrogenation. pristine AB, the intensity of the B–N band was signicantly
The chemical shi region of borates is also complex, supporting reduced aer the mixing of AB and SB, which can be interpreted
17
11
the existence of ammonium borate and polymeric borate in as weakening and even breaking of B–N bonds. Heating the
solid products. These results are consistent with the XRD mixture further attenuated the B–N band, as higher molecular
results, representing dehydrogenated reaction products. The kinetic energy caused by heating enhanced the interaction. AB*
collective information about products and isothermal H
2
mobile phase was possibly formed during this stage. Upon the
release capacities indicates that moderate RH values are the key introduction of water vapor and the commencement of dehy-
to deep dehydrogenation of ABSB, where both AB and SB are drogenation, the intensity of the B–N band was revived, indi-
sufficiently dehydrogenated. In contrast, low or high RH only cating the strengthening and recovery of B–N bonds. This
allows partial dehydrogenation of SB, which is also facilitated by occurred concurrently with the emergence of a shoulder peak in
À1
ABSB interaction though, forming regular SB hydrolysates and the B–H band at 2441 cm , as shown in Fig. S5b (ESI†), which is
leaving AB largely unreacted.
related to the formation of DADB. Similar efforts to observe the
dehydrogenation of SB in ABSB were also made, with results
shown in Fig. S6 (ESI†), exhibiting the in situ Raman spectra
before ABSB being mixed and aer ABSB being mixed and sub-
jected to water vapor and heat. The mixing of SB with AB did not
cause any notable change in the Raman spectrum; neither
apparent band shiing nor new vibrational bands were seen.
3.3 Dehydrogenation pathway
Further efforts were also made to understand the process and
transformation during the ABSB-MRH hydrothermolysis by using
in situ Raman spectroscopy measurement with spectra evolution
of AB in ABSB shown in Fig. 4a for the 2000–3400 cm region
and in Fig. S5a (ESI†) for the 560–2000 cm region. Initially, all
À1
However, aer applying heat and water, the intensities of all BH4
À1
bands were attenuated and the combinational band n2+4 was
AB vibrational bands are weakened and broadened aer heat and
vapor are applied, due to the reduction of crystallinity and
interaction among components. With the progression of hydro-
thermolysis, DADB, which is an important intermediate for AB
thermolysis, emerges and increases in signicance over time.
Concurrently, evidence of borate formation is observed. Aer 3
hours of hydrothermolysis, all previously detected bands from AB
and DADB intermediates have vanished, and only weak peaks
from borates remain, indicating that the dehydrogenation
process is complete. The impact of ABSB interaction on AB can be
more clearly observed when comparing the Raman spectra of AB
before being mixed with SB (pristine-AB) and at sequential time
points in the early stage of dehydrogenation aer ABSB being
mixed, heated, and exposed to water vapor. The most signicant
changes were observed in the B–N vibration bands as shown in
Fig. 4b, where the intensities are normalized to the strongest
peaks of symmetric B–H in each spectrum to reect relative
intensity changes (IBN/IsyBH). The corresponding B–H band
evolution is shown in Fig. S5b (ESI†) as well. Compared to
suppressed and eliminated, which could imply the alteration of
+
quantity and chemical environment for BH
4
due to the occur-
rence of dehydrogenation interaction and reaction. With the
+
4
progression of dehydrogenation, the bands related to BH
van-
3 4
ished gradually, and concurrently, the B –O and B –O bands
from borates emerged.
In order to unravel the dehydrogenation pathway for ABSB,
two isotope modied ABSB-MRH isothermal dehydrogenation
experiments were conducted: one with D
oxide, Sigma-Aldrich) vapor replacing H
and another one with ND BH compound replacing AB in ABSB
together with D O vapor (AB(D)SB–D O). The volatile products
2
O (99.9 at% deuterium
2
O vapor (ABSB–D O);
2
3
3
2
2
collected aer 50–60 min of each experiment were examined by
UGA in the high resolution mode, and the sections correspond-
ing to hydrogen and isotopes are shown and compared to the
spectrum of regular ABSB-MRH (Fig. 5). The UGA spectrum for
2
ABSB–D O (Fig. 5a) shows the strong dominance of regular
2
hydrogen (H ) compared to hydrogen deuteride (HD) originating
À1
Fig. 4 In situ Raman spectra for dehydrogenation of AB in ABSB: (a) spectrum evolution with time in the 2000–3400 cm Raman shift region,
where ST indicates stretching mode; and (b) comparison of B–N bands before ABSB being mixed (pristine AB) and after ABSB being mixed (ABSB-
mixed), heated (ABSB-heated), and exposed to water vapor (ABSB-in situ-5–35 min) up to 35 min. The intensities in (b) are normalized by the
corresponding symmetric B–H peaks with arrows indicating the changes compared to the previous spectrum.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
Journal of Materials Chemistry A
from the above results and are schematically illustrated using
AB(D)SB as example to differentiate between homopolar and
heteropolar dehydrogenations, as shown in Fig. 6. Upon
blending of AB and SB, their interaction weakens and even
breaks B–N bonds and relaxes the dihydrogen bonding network
in AB. This process is greatly enhanced by mild heating of the
ꢀ
mixture (e.g. at 55 C) and eventually enables the formation of
DADB. Evidence also indicated a higher reactivity of SB in ABSB.
These observations may be attributed to the alteration of elec-
tron structure and distribution in AB and SB through interac-
À
tions between –BH /BH
moieties, e.g. electrostatic
3
4
interaction, in the presence of water molecules. DADB is also
generally seen as an intermediate during thermolysis of AB at
temperatures above 80 C. However, the present ABSB dehy-
2
Fig. 5 High resolution UGA spectra of hydrogen (H ) and hydrogen
deuteride (HD) released by hydrothermolysis of (a) ABSB using D O
2
ꢀ
vapor, (b) ND
3
BH
3
+ SB mixture using D
2
O vapor, and (c) ABSB using drogenation process does not follow either of the regular
2
H O vapor (as reference case). All three experiments used medium RH
pathways of thermolysis or hydrolysis, as can be expected due to
the relatively low temperature and absence of catalyst and
liquid water in this case. Instead, a new dehydrogenation
pathway is observed at this stage as shown by the isotope and in
situ Raman experiments. The new pathway mostly relies on
(
MRH).
in part from D
possible dehydrogenation paths for ABSB–D
Fig. S7a (ESI†). As schematized, when D O vapor is involved in
the process, HD will be generated by hydrolysis with either AB or
SB; on the other hand, when the dehydrogenation happens
through the reaction/interaction between AB and SB or through
2
O, with an H
2
/HD integral area ratio of 25.9. The
2
O are illustrated in
homopolar dehydrogenation, specically through dehydrocou-
À
2
pling between –BH
AB, and NaBH
2
–/–BH
3
/BH
4
moieties contained in DADB,
4
. Homopolar dehydrogenation is counterintui-
tive given the repulsive force between the negatively charged
moieties. However, emerging evidence suggests an important,
common role of homopolar interaction during dehydrogena-
tion of chemical hydrides. According to the study by Wol-
the self-decomposition of AB, only H is generated. The high
2
integral area ratio thus excludes the possibility of any substantial
hydrolysis of AB or SB. Furthermore, the self-decomposition of
AB can also be excluded, given the absence of SB hydrolysis. This
implies that most of the hydrogen gas released originated from
AB and SB reaction/interaction. The product from hydro-
18
stenholme et al., even for the regular thermolysis of pure AB
homopolar B–H/H–B interaction besides the well-known
interaction of N–H/H–B contributes a substantial portion of
dehydrogenation when the distance between B–H/H–B has
been shortened by melting. In the present case, the quantity of
HH from homopolar dehydrogenation is about ꢁ6Â greater
than the HD amount from heteropolar dehydrogenation.
Furthermore, the derivation of D for heteropolar dehydroge-
nation is more likely from –ND moieties than from water (D O)
2
thermolysis of ABSB–D O was examined by Raman spectroscopy,
and the results shown in Fig. S8 (ESI†) indicate the existence of
N–D stretching vibration bands, which suggests that water vapor
promotes the formation of ammonium in products by providing
hydrogen atoms, besides supplying oxygen to borates. A deeper
insight into the dehydrogenation mechanism, especially into the
role of homopolar dehydrogenation, comes from the analysis of
the UGA results for AB(D)SB–D
possible dehydrogenation paths for AB(D)SB–D
tized in Fig. S7b (ESI†). Besides any HD released from hydrolysis
of AB(D)SB, heteropolar dehydrogenation between protonic D in
3
2
molecules. This is distinct from regular hydrolysis, in which
heteropolar dehydrogenation is the dominant process with
a signicant portion of hydrogen supplied by water. Neverthe-
less, water is still essential during the dehydrogenation. Besides
assisting the formation of DADB, water molecules also provide
oxygen to form borates and hydrogen to form ammonium,
which enables the formation of products and thus facilitates the
process of dehydrogenation. Importantly, the amount of water
vapor also determines the approach and extent of dehydroge-
nation, which is favoured by moderate RH levels. With insuffi-
cient RH, the available water vapor is unable to facilitate the
homopolar interaction and to sustain the formation of products
by providing adequate O and H atoms, whereas excessive water
vapor will rapidly deplete SB through regular hydrolysis into
sodium borates, and thus deprive the interaction between AB
and SB and prevent further dehydrogenation of ABSB.
2
O dehydrogenation (Fig. 5b). The
O are schema-
2
+
4
–
ND and hydridic H in –BH /BH also produces HD. On the
3
3
contrary, all H
2
generated must come from homopolar dehy-
À
drogenation among –BH
UGA results of the AB(D)SB–D
ously distinguish the contributions from heteropolar and
homopolar dehydrogenations. Here, the reduced H /HD integral
area ratio for AB(D)SB–D O (6.2) compared to that of ABSB–D O
2
–/–BH
3
/BH
4
moieties. Therefore, the
2
O isotope experiment unambigu-
2
2
2
(25.9) shows the additional contribution from heteropolar
dehydrogenation, where the D derives from –ND moieties in AB,
as conrmed by the supplementary isotope experiment of AB(D)
SB–H O in Fig. S9 (ESI†). However, the preponderance of H
2 2
affirms the major role of homopolar dehydrogenation in the
overall hydrothermolysis of ABSB.
3
The regeneration of spent fuel and impurities in products
are thorny issues for AB and SB hydrogen storage materials.
Developing effective methods to solve these problems is a hot
The mechanism and pathway of ABSB dehydrogenation by
vapor facilitated hydrothermolysis can be collectively inferred
19
topic, and promising progress has been made. Theoretically,
the techniques developed for hydrolytic dehydrogenation
This journal is © The Royal Society of Chemistry 2020
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 6 Proposed pathway for dehydrogenation of AB(D)SB by hydrothermolysis. The methods used to identify the intermediates, products, and
processes are indicated in brackets.
should be also applicable for the regeneration and impurity high quality with less than 1% of ammonia as impurity. Water
suppression of ABSB. On the other hand, the new homopolar vapor concentration was a decisive factor for the approach and
based dehydrogenation mechanism of ABSB renders possibili- degree of dehydrogenation, and the highest extent of dehydro-
ties of new strategies to simplify spent fuel regeneration and genation (7.78 wt%) was achieved with moderate relative
impurity elimination. As aforementioned, the water in humidity. In situ Raman, isotope experiments, and thorough
conventional hydrolytic dehydrogenation commits to supply H/ analysis of the solid products unraveled the novel homopolar
O atoms for dehydrocoupling and also for solid product based dehydrogenation mechanism, which is driven by inter-
formation. Under the new mechanism, the main role of water is molecular interaction with AB* and DADB as intermediates.
more simple, which is to promote the formation of products. Distinct from conventional hydrolytic dehydrogenation, the role
Therefore, it is rational to hypothesize that other gaseous of water vapor is to promote the formation of products and the
reagents could be utilized to reduce the thermodynamic interaction rather than participating in dehydrocoupling. The
stability of the products without compromising dehydrogena- ratio between homopolar and heteropolar dehydrogenation was
tion. Although the ammonia emission from ABSB dehydroge- about 6 : 1. Based on this, a theoretical maximum capacity of
nation is low (0.7 vol%), further reduction and even elimination 11.7 wt% was predicted. Overall, this work validated the possi-
of ammonia are feasible through maximizing homopolar bility of dehydrogenating chemical hydride mixtures through
dehydrogenation and avoiding side reaction. These could be intermolecular interaction. The low-temperature heat, minimal
achieved by optimizing the parameters of hydrothermolysis. water use, and absence of catalyst for hydrogen liberation are key
The gap between theoretical capacity and measured capacity advantages for ABSB as an excellent disposable solid state
could also be narrowed by parameter optimization. Further hydrogen source with simple operation, high energy density, and
research is needed to systematically investigate the optimal AB rapid kinetics. Moreover, the homopolar based dehydrogenation
to SB ratio and optimal vapor concentration range under mechanism could render the possibilities of better spent fuel
a certain AB/SB ratio.
regeneration and impurity elimination strategies. Further study
based on this is suggested.
4
. Conclusion
Conflicts of interest
The present work revealed for the rst time the phenomenon and
mechanism of simultaneous and interactional dehydrogenation
of ammonia borane and sodium borohydride at near ambient
temperature without using any catalyst and liquid solvent. Mix-
ing, moderate humidity, and mild heat are the only requirements
for liberation of hydrogen. Major hydrogen release was demon-
A patent application has been led based on this work (US
62783620; PCT/CA2019/051841).
Acknowledgements
ꢀ
À1 À1
strated at 53.5 C with peak evolution rate of 562.7 ml g
s
via This research was supported by the Natural Sciences and
dynamic dehydrogenation experiment. The hydrogen gas was of Engineering Research Council of Canada (NSERC), Mitacs,
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
Journal of Materials Chemistry A
Hydrogen in Motion Inc. (H2M), Western Economic Diversi-
cation Canada (WD), Canada Foundation for Innovation (CFI),
9 I. Dovgaliuk, C. S. L. Duff, K. Robeyns, M. Devillers and
Y. Filinchuk, Chem. Mater., 2015, 27, 768.
British Columbia Knowledge Development Fund (BCKDF), and 10 L. H. Jepsen, J. Skibsted and T. R. Jensen, J. Alloys Compd.,
Simon Fraser University (SFU). The work made use of the 4D 2013, 580, S287.
LABS shared facilities at SFU supported by the CFI, BCKDF, WD, 11 J. Hannauer, U. B. Demirci, C. Geantet, J. M. Herrmann and
and SFU. This research was undertaken, in part, thanks to P. Miele, Phys. Chem. Chem. Phys., 2011, 13, 3809.
funding from the Canada Research Chairs program. We 12 G. H. Penner, Y. C. P. Chang and J. Hutzal, Inorg. Chem.,
acknowledge Marisa Tellis and Angus Hui for technical support.
1999, 38, 2868.
13 P. Pei, M. B. Whitwick, W. L. Sun, G. Quan, M. Cannon and
E. Kjeang, Nanoscale, 2017, 9, 4143.
References
1
4 A. S. Nathanson, A. R. Ploszajski, M. Billing, J. P. Cook,
D. W. K. Jenkins, T. F. Headen, Z. Kurban, A. Lovell and
S. M. Bennington, J. Mater. Chem. A, 2015, 3, 3683.
1
2
U. B. Demirci, Int. J. Hydrogen Energy, 2017, 42, 9978.
P. Brack, S. E. Dann and K. G. Upul Wijayantha, Energy Sci.
Eng., 2015, 3, 174.
15 W. J. Shaw, M. Bowden, A. Karkamkar, C. J. Howard,
D. J. Heldebrant, N. J. Hess, J. C. Linehan and T. Autrey,
Energy Environ. Sci., 2010, 3, 796.
3
4
5
6
7
J. Zhang, Y. Zhao, D. L. Akins and J. W. Lee, Ind. Eng. Chem.
Res., 2011, 50, 10407.
X. Zhang, Z. Wei, Q. Guo and H. Tian, J. Power Sources, 2013, 16 A. C. Stowe, W. J. Shaw, J. C. Linehan, B. Schmid and
31, 190.
T. Autrey, Phys. Chem. Chem. Phys., 2007, 9, 1831.
S. B. Kalidindi, M. Indirani and B. R. Jagirdar, Inorg. Chem., 17 C. Chou, D. Lee and B. Chen, Int. J. Hydrogen Energy, 2012,
008, 47, 7424.
37, 15681.
L. Chong, X. Zeng, W. Ding, D. J. Liu and J. Zou, Adv. Mater., 18 D. J. Wolstenholme, K. T. Traboulsee, Y. Hua, L. A. Calhoun
015, 27, 5070.
and G. S. McGrady, Chem. Commun., 2012, 48, 2597.
D. J. Heldebrant, A. Karkamkar, N. J. Hess, M. Bowden, 19 U. Sanyal, U. B. Demirci, B. R. Jagirdar and P. Miele,
S. Rassat, F. Zheng, K. Rappe and T. Autrey, Chem. Mater.,
ChemSusChem, 2011, 4, 1731.
008, 20, 5332.
2
2
2
2
8
X. Kang, Z. Fang, L. Kong, H. Cheng, X. Yao, G. Lu and
P. Wang, Adv. Mater., 2008, 20, 2756.
This journal is © The Royal Society of Chemistry 2020
J. Mater. Chem. A