Production of H2 from combined endothermic and exothermic hydrogen carriers

Article abstract of DOI:10.1021/ja806721s

One of the major limitations to the use of fuel cell systems in vehicular transportation is the lack of hydrogen storage systems that have the required hydrogen storage density and moderate enthalpy of dehydrogenation. Organic liquid H₂ carriers that release H₂ endothermically are easier to handle with existing infrastructure because they are liquids, but they have low storage densities and their endothermicity consumes energy in the vehicle. On the other hand, inorganic solid H₂ carriers that release H₂ exothermically have greater storage densities but are unpumpable solids. This paper explores combinations of an endothermic carrier and an exothermic carrier, where the exothermic carrier provides some or all of the necessary heat required for dehydrogenation to the endothermic system, and the endothermic carrier serves as a solvent for the exothermic carrier. The two carriers can be either physically mixed or actually bonded to each other. To test the latter strategy, a number of chemically bound N-heterocycle:BH ₃ adducts were synthesized and in turn tested for their ability to release H₂ by tandem hydrolysis of the BH₃ moiety and dehydrogenation of the heterocycle. To test the strategy of physically mixing two carriers, the hydrolysis of a variety of amine-boranes (H ₃N:BH₃, Me₂HN:BH₃, Et ₃N:BH₃) and the catalytic dehydrogenation of indoline were carried out together.

Full text of DOI:10.1021/ja806721s

Published on Web 11/17/2008
Production of H from Combined Endothermic and
2
Exothermic Hydrogen Carriers
†
†
†
,‡
,†
Dominik Wechsler, Yi Cui, Darrell Dean, Boyd Davis,* and Philip G. Jessop*
Department of Chemistry, Queen’s UniVersity, Kingston, Ontario, Canada K7L 3N6, and
Queen’s-RMC Fuel Cell Research Centre, 945 Princess Street, Second Floor, Kingston,
Ontario, Canada K7L 5L9
Received September 2, 2008; E-mail: jessop@chem.queensu.ca; boyd.davis@mine.queensu.ca
Abstract: One of the major limitations to the use of fuel cell systems in vehicular transportation is the lack
of hydrogen storage systems that have the required hydrogen storage density and moderate enthalpy of
dehydrogenation. Organic liquid H
infrastructure because they are liquids, but they have low storage densities and their endothermicity
consumes energy in the vehicle. On the other hand, inorganic solid H carriers that release H exothermically
2 2
carriers that release H endothermically are easier to handle with existing
2
2
have greater storage densities but are unpumpable solids. This paper explores combinations of an
endothermic carrier and an exothermic carrier, where the exothermic carrier provides some or all of the
necessary heat required for dehydrogenation to the endothermic system, and the endothermic carrier serves
as a solvent for the exothermic carrier. The two carriers can be either physically mixed or actually bonded
to each other. To test the latter strategy, a number of chemically bound N-heterocycle:BH
synthesized and in turn tested for their ability to release H by tandem hydrolysis of the BH
dehydrogenation of the heterocycle. To test the strategy of physically mixing two carriers, the hydrolysis of
3
adducts were
2
3
moiety and
a variety of amine-boranes (H
were carried out together.
3 3 2 3 3 3
N:BH , Me HN:BH , Et N:BH ) and the catalytic dehydrogenation of indoline
Introduction
are organic liquids that can be dehydrogenated on-board (in the
vehicle or device) and rehydrogenated off-board (i.e., in a
factory). The use of liquids greatly facilitates the engineering
because they are pumpable and, if they are to be used in vehicles,
can be stored and delivered using existing infrastructure.
Endothermicity is also desirable because it makes the reaction
more controllable. However, one drawback for implementing
these endothermic systems in vehicles is the large amount of
energy that must be supplied within the vehicle to satisfy the
enthalpy of dehydrogenationsthis significantly detracts from
the effective energy storage density. Naturally, the reverse
hydrogenation reaction (off-board regeneration) releases large
Due to rising concerns about emissions of carbon dioxide
and other pollutants from mobile emission sources such as
vehicles, much attention is being paid to H as a clean fuel and
2
1
,2
2,3
energy carrier. While fuel cell technologies are advancing,
the development of hydrogen storage systems, beyond com-
pressed hydrogen and low storage density metal hydrides, is
being stymied by significant technological roadblocks. Currently,
several hydrogen storage systems are being developed for
vehicular transportation; these include physical systems involv-
ing compressed or adsorbed H
metal hydrides and organic liquids.
can be divided into those that release H
those that release H exothermically.
Endothermic carriers, such as cyclic hydrocarbons (e.g.,
2
and chemical systems such as
4
-6
The chemical systems
endothermically and
2
2
(
11) Klvana, D.; Touzani, A.; Chaouki, J.; Belanger, G. Int. J. Hydrogen
Energy 1991, 16, 55–60.
7
-14
15-21
cyclohexanes),
and N-heterocycles (e.g., piperidines),
(12) Touzani, A.; Klvana, D.; Belanger, G. Int. J. Hydrogen Energy 1984,
9
, 929–936.
(
13) Tsuji, T.; Shinya, Y.; Hiaki, T.; Itoh, N. Fluid Phase Equilib. 2005,
228-229, 499–503.
†
Queen’s University.
Queen’s-RMC Fuel Cell Research Centre.
‡
(14) Pez, G. P.; Scott, A. R.; Cooper, A. C.; Cheng, H.; Bagzis, L. D.;
Appleby, J. B. WO2005000457, Air Products and Chemicals, 2005.
(15) Clot, E.; Eisenstein, O.; Crabtree, R. H. Chem. Commun. 2007, 2231–
2233.
(16) Cui, Y.; Kwok, S.; Bucholtz, A.; Davis, B.; Whitney, R. A.; Jessop,
P. G. New J. Chem. 2008, 32, 1027–1037.
(17) Moores, A.; Poyatos, M.; Luo, Y.; Crabtree, R. H. New J. Chem. 2006,
30, 1675–1678.
(18) Pez, G. P.; Scott, A. R.; Cooper, A. C.; Cheng, H. EP 1475349, 2004.
(19) Pez, G. P.; Scott, A. R.; Cooper, A. C.; Cheng, H.; Wilhelm, F. C.;
Abdourazak, A. H. U.S. Patent 7351395, 2008.
(
(
1) Schlapbach, L.; Z u¨ ttle, A. Nature 2001, 414, 353–358.
2) Felderhoff, M.; Weidenthaler, C.; von Helmolt, R.; Eberle, U. Phys.
Chem. Chem. Phys. 2007, 9, 2643–2653.
(
(
(
(
(
3) Janssen, L. J. J. J. Appl. Electrochem. 2007, 37, 1383–1387.
4) Gray, E. M. AdV. Appl. Ceram. 2007, 106, 25–28.
5) Ni, M. Energ. Explor. Exploit. 2006, 24, 197–209.
6) Ross, D. K. Vacuum 2006, 80, 1084–1089.
7) Cacciola, G.; Giordano, N.; Restuccia, G. Int. J. Hydrogen Energy
1
984, 9, 411–419.
(
(
8) Davtyan, O. K.; Burshtein, I. I. Arm. Khim. Zh. 1971, 24, 1044–1049.
9) Hama, S.; Li, X.; Yukawa, K.; Saito, Y. Chem. Lett. 1992, 233, 2463–
(20) Toseland, B. A.; Pez, G. P.; Puri, P. S. U.S. Patent Application
20060143981, 2006.
(21) Pez, G. P.; Scott, A. R.; Cooper, A. C.; Cheng, H. U.S. Patent
Application 20050002857, 2004.
2
466.
10) Kariya, N.; Fukuoka, A.; Ichikawa, M. Appl. Catal. A 2002, 233, 91–
02.
(
1
10.1021/ja806721s CCC: $40.75

2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 17195–17203 9 17195

A R T I C L E S
Wechsler et al.
amounts of energy, requiring the reaction vessels to be cooled
to help dissipate the heat evolved.
Exothermic carriers, such as boron hydrides and metal
2
hydrides (when the H is released by hydrolysis), have the great
advantages of significantly higher gravimetric hydrogen storage
densities and not requiring heat to be supplied on-board.
However, these carriers require more complicated on-board
engineering because they are solids and because the exothermic
release of hydrogen results in heat management issues and the
risk of overheating or runaway reaction. The exothermic carriers
are also much more difficult to regenerate. This causes a
significant barrier to their use since it means that their cost of
production is high, and therefore their potential applications are
limited to niche high value products such as remote power and
military devices. However, chemical methods for the regenera-
Figure 1. ORTEP diagram for 2a, shown with 50% displacement ellipsoids.
this work, the endothermic carrier is a N-heterocycle and the
exothermic carrier contains B-H bonds which release H by
2
22,23
4
tion of hydrolyzed NaBH , for example, have been reported.
exothermic hydrolysis. Specifically, combinations of N-hetero-
An electrolytic method of directly reversing the hydrolysis with
cycles as the endothermic carrier with different amine-boranes
as the exothermic carrier were evaluated.
2
4
hydrogen gas has been proposed and may be followed by
others that would significantly reduce the overall cost of
regeneration.
Results and Discussion
2
The energy content (enthalpy of combustion) of H is 285.8
Chemically Bound Nitrogen-Boron Complexes: Synthesis
2
5
kJ/mol, which is only partially converted to electrical energy
by the polymer electrolyte membrane (PEM) fuel cell. PEM
fuel cells used for transportation purposes have an electrical
and Reactivity. Adducts of saturated N-heterocycles and BH
3
are advantageous because the two carriers are combined into
one molecule; accidental phase separation of the two carriers
is impossible, energy transfer from the exothermic to the
endothermic carrier is facile, and the ratio of exothermic to
endothermic carriers is fixed. Although there are many reports
efficiency of 53-58% (151.5-165.8 kJ/mol H
remaining 42-47% (120.0-134.4 kJ/mol H ) is released as heat
operating temperature 50-100 °C), which could possibly be
used to provide some of the heat for the dehydrogenation of
2
), while the
2
(
2
7-30
on using ammonia-borane and other amine-boranes
as
2
6
endothermic hydrogen carriers.
hydrogen carriers, none of the reported amine-boranes have an
endothermic hydrogen releasing portion. Several such com-
pounds have now been synthesized and evaluated to determine
whether these compounds can be hydrolyzed (exothermic
portion) and dehydrogenated (endothermic portion) at 100 °C.
A significant potential disadvantage of binding BH3 to
N-heterocycles is the resulting loss of electron density from the
heterocycle. Electron density strongly influences whether a
compound is easily dehydrogenated. We have shown previ-
A combination of an endothermic system and an exothermic
system would combine the advantages of both systems. The
energy released from the exothermic system would provide some
or all of the energy required for the endothermic process. The
liquid endothermic carrier would dissolve the solid exothermic
carrier, eliminating the engineering difficulties associated with
the use of a solid. The high hydrogen storage density of the
exothermic system would increase the overall storage capacity
16
of the combination relative to the low H
2
storage capacities of
ously that the substituents on N-heterocycles have a significant
influence on the ability of dehydrogenation. Electron-donating
substituents or conjugated substituents promote dehydrogenation,
while electron-withdrawing substituents have the opposite effect.
4-Aminopiperidine served as the endothermic carrier for the
first systems studied. Adducts of 4-aminopiperidine with two
BH molecules (1a) and with one BH molecule (2a) were
endothermic systems. In order for such a combined system to
operate adequately, the endothermic and exothermic carriers
must be chosen so that neither prevents the hydrogen release
from the other nor renders it unstable to decomposition or
premature H
if both are liquids) or the liquid carrier should dissolve the
solid, so that the mixture is liquid from -40 °C to well above
the temperature required for H release.
2
release. The two carriers should also be miscible
(
3
3
31
prepared (eq 1). An ORTEP diagram of 2a is shown in Figure
1. The nitrogen-boron bond distance (1.597(3) Å) agrees with
similar compounds in the literature
2
3
2-34
In this paper, we describe the feasibility of combining a carrier
that releases hydrogen by endothermic dehydrogenation with a
carrier that releases hydrogen by exothermic hydrolysis. Two
different approaches were considered for combining the two
carriers: (a) chemical binding of one carrier to the other and
exhibiting single bond
character. The connectivity of 1a was confirmed by standard
1
13
11
NMR experiments ( H, C, B, COSY, HSQC, and HMBC)
1
11
1
1
as well as H{ B} and H- H NOESY NMR experiments. In
the H{ B} spectra (Figure 2), each of the boron moieties
1
11
(
b) physical mixtures of the carriers. Ideally, the combined
system would release H at or below 100 °C to best match the
temperature of operation of a PEM fuel cell operation. In all of
2
(27) Marder, T. B. Angew. Chem., Int. Ed. 2007, 46, 8116–8118.
(28) Mohajeri, N.; T-Raissi, A.; Adeibiyi, O. J. Power Sources 2007, 167,
4
82–485.
(
29) Stephens, F. H.; Baker, R. T.; Matus, M. H.; Grant, D. J.; Dixon,
(
(
(
(
(
22) Chin, A. A.; Lipiecki, F. J.; Park, W. S. EP1645644, Rohm & Haas,
D. A. Angew. Chem., Int. Ed. 2007, 46, 746–749.
2
006.
(30) Stephens, F. H.; Pons, V.; Baker, R. T. Dalton Trans. 2007, 2613–
2626.
23) Liu, B. H.; Li, Z. P.; Zhu, J. K.; Morigasaki, N.; Suda, S. Energy
Fuels 2007, 21, 1707–1711.
(31) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(32) Amezcua, C. A.; Bell, K. E.; Kelly, H. C. Inorg. Chim. Acta 1999,
290, 80–85.
24) Calabretta, D. L.; Davis, B. R. J. Power Sources 2007, 164, 782–
7
91.
25) Lide, D. R. CRC Handbook of Chemistry and Physics; 80th ed.; CRC
Press: Boca Raton, FL, 1999.
(33) Chitsaz, S.; Breyhan, T.; Pauls, S.; Neumuller, B. Z. Anorg. Allg. Chem.
2002, 628, 956–964.
26) Types of Fuel Cells. http://www1.eere.energy.gov/hydrogenandfuel-
cells/fuelcells/fc_types.html (July 2008), U.S. Department of Energy.
(34) Khasnis, D. V.; Lattman, M.; Siriwardane, U. J. Chem. Soc., Chem.
Commun. 1989, 1538–1540.
1
7196 J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008

H
2
from Combined Endothermic and Exothermic Hydrogen Carriers
A R T I C L E S
1
1
11
Figure 2. H NMR spectrum of 1a (top); H{ B} NMR spectra of 1a, boron decoupled at -14.7 ppm (middle), and boron decoupled at -21.2 ppm
bottom).
(
1
1
1
Figure 3. (A) The H NMR spectrum of 1a. (B) A cross-section slice of the H- H NOESY NMR spectrum for 1a was taken along the borane proton shift
1
1
(
1.3 ppm; NH-BH
3
and NH
2
-BH
3
) depicting any correlations to other protons within the molecule. As seen in the figure, the H- H NOESY correlations
]) as well as the other methylene protons within the molecule.
correspond to the amine protons (5.8 [NH] and 5.2 ppm [NH
2
1
1
[11B NMR:
(
attached to the NH [ B NMR: δ -14.7] and NH
2
of the resulting product by NMR spectroscopy showed a change
in the boron chemical shifts, while the C–H protons have the
same splitting pattern as the starting material with only a slight
δ -21.2]) is decoupled individually, resulting in two separate
singlets at 1.30 and 1.24 ppm, respectively. The structure was
1
1
11
further confirmed by through-space H- H NOESY correlations
change in chemical shift. The boron chemical shifts of 1a [ B
1
1
(
see Supporting Information Figures S9 and S10 for H- H
NMR: δ -14.5 (R
2
NH:BH
3
) and -20.8 (RNH
2 3
:BH )] shifted
NOESY NMR spectrum) between the methylene, amine, and
downfield to a very broad singlet at about 10-11 ppm
1
1
-
the borane protons. A cross-section slice of the H- H NOESY
NMR spectrum along the borane proton shift was extracted and
referenced to the H NMR spectrum. Clear correlations are
corresponding to a borate species, such as B(OH)
3 2
or BO ;
Xu and co-workers have reported that such borate species exist
1
35-37
in an equilibrium mixture.
For simplicity, only the boric
observed between the borane protons and the amine and
methylene protons (Figure 3); this provides additional evidence
besides the H{ B} NMR spectra for the presence of two
distinct borane moieties.
acid derivative of 1a (1b) is shown in eq 2. The change in the
boron chemical shifts suggests complete borane hydrolysis,
while the lack of proton chemical shifts in the olefinic and
1
11
1
aromatic regions of the H NMR spectrum confirms the
complete absence of ring dehydrogenation. In contrast, the
dehydrogenation of 4-aminopiperidine (without BH
3
groups
attached) with Pd/SiO occurs in 100% conversion over 0.5 h
2
1
6
at 170 °C and in 15% conversion over 3 h at 100 °C.
Reaction of 2a with water under the same reaction conditions
11
as 1a showed only borane hydrolysis [ B NMR: δ -14.5
(
R
2
NH:BH of 2a) shifted downfield to a very broad singlet at
3
(
(
35) Chandra, M.; Xu, Q. J. Power Sources 2006, 156, 190–194.
36) Chandra, M.; Xu, Q. J. Power Sources 2006, 159, 855–860.
When 1a is heated to 100 °C for 3 h in the presence of water
and catalyst (Pd/silica), gas evolution is observed. Examination
(37) Xu, Q.; Chandra, M. J. Alloys Compd. 2007, 446-447, 729–732.
J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008 17197

A R T I C L E S
Wechsler et al.
1
11
1
1
by standard NMR experiments as well as H{ B} and H- H
1
NOESY NMR experiments. In the H NMR spectrum of 3, the
borane protons appear as a broad apparent quartet ( H NMR: δ
3
), while in the H{ B} spectrum,
the quartet disappeared and was replaced by a singlet (Figure
). The structure is further confirmed by through-space H- H
NOESY correlations (see Supporting Information Figures S14
and S15 for the H- H NOESY NMR spectrum) between the
C7-H, amine, and the borane protons. A cross-section slice of
the H- H NOESY NMR spectrum (Figure 5) was extracted
along the C7 proton shift, showing any correlations toward other
protons within the molecule. The correlations that are observed
1
1
1
11
2
.48, JHB ) ∼90 Hz, 3H, BH
1
1
4
about 10.4 ppm], but again no dehydrogenation of the ring
system was observed. For simplicity, only the boric acid form
of 2a (2b) is shown in eq 3. The dehydrogenation of the ring
system seems to be unattainable at 100 °C when borane is
attached. The lack of dehydrogenation indicates that the
hydrolysis reaction is unable to provide enough energy to
facilitate the dehydrogenation at this temperature and/or that
the dehydrogenation is inhibited by electron density loss from
the ring to the boron moiety.
1
1
1
1
1
can be referenced to the H NMR spectrum, providing evidence
for the C7 proton being in close proximity to the amine and
borane protons, confirming the proposed indoline-borane struc-
ture 3.
The dehydrogenation and hydrolysis of 3 was performed using
palladium on carbon in the presence of water at 100 °C over
1
the course of 2 h. Analysis of the product mixture by H and
1
1
B NMR spectroscopy showed good indoline dehydrogenation
1
(
71%) and complete borane hydrolysis (eq 5). The H NMR
) shows an NH peak
spectrum of the major product (in DMSO-d
6
In order to determine whether the lack of dehydrogenation
is due to the temperature used or to the presence of the electron-
withdrawing borane moieties, the dehydrogenations of com-
plexes 1b and 2b were explored in the presence of palladium
on silica without water at elevated temperature (170 °C) in
deuterated DMSO (solvent was added because the melting points
of 1b and 2b are >220 °C). Even at 170 °C, no dehydrogenation
was observed for either 1b or 2b over the course of 3 h, which
is quite a contrast to the case of the parent molecule 4-ami-
nopiperidine which completely dehydrogenates over the course
of 0.5 h under similar conditions.
at 11 ppm that clearly correlates, in the COSY spectrum, to the
olefin hydrogens, demonstrating that the product retains an NH
proton and therefore only a single bond between N and B. This
reaction constitutes an example of H release from both an
2
endothermic carrier and an exothermic carrier, with the two
carriers actually being chemically bonded to each other. The
gravimetric hydrogen storage capacity of the system is 6.0%
(not including the water, which we assume will be obtained
from the exhaust of the fuel cell). The overall reaction is
3
exothermic, however, because the enthalpy of BH hydrolysis
3
6,37
(approximately -156 kJ/mol)
added to the enthalpy of
On the basis of the lack of dehydrogenation of 1b and 2b
indoline dehydrogenation (52 kJ/mol for unsubstituted indo-
line) gives a total of -104 or -26 kJ/mol H . Also, the
dehydrogenation of 3 was incomplete, even though the parent
3
8
(vide supra), we wanted to determine if the type of boron moiety
2
[
BH vs B(OH) ] has an effect on the dehydrogenation of the
3
3
ring. Therefore, the dehydrogenation of 1a was explored under
similar conditions but without the addition of water. The
dehydrogenation of compound 1a (melting point: >240 °C) was
performed in triglyme or DMSO at 170 °C over several hours
without any dehydrogenation of the ring (confirmed by the lack
of proton chemical shifts in the olefinic and aromatic regions).
compound indoline dehydrogenates completely in that time, in
keeping with the suggestion that the boron moiety acts as an
electron-withdrawing group, making the dehydrogenation more
difficult.
1
However, the number of detectable amine protons in the H
NMR spectrum has gone from 3 (for compound 1a) to 1,
suggesting B-N dehydrocoupling. The product was not isolated
or characterized.
The lack of observable dehydrogenation for the cyclic
moieties of the boron complexes 1b and 2b at 170 °C suggests
that, even though the hydrolysis reaction is exothermic, the
coordinated boron moieties are sufficiently electron-withdrawing
to increase the enthalpy of dehydrogenation and thereby prevent
dehydrogenation.
In order to avoid this problem, the boron moiety must not be
bonded directly to the heterocycle. One method of separating
the two moieties would be to have each in separate independent
molecules, which is the strategy described in the next section.
Physical Combination of Endothermic/Exothermic Systems.
In order to separate the borane moiety from the cyclic moiety,
we explored physical combinations of boranes and indoline. This
should avoid the inhibition problem described above (only
partial dehydrogenation of indoline) as long as no borane transfer
to indoline occurs.
Due to the lack of clean dehydrogenation of the ring moiety
of 4-aminopiperidine in 1a,b and 2a,b even at elevated
temperatures, the use of indoline was explored as an alternative
endothermic carrier. Indoline has been shown to dehydrogenate
1
7
Having two independent systems gives one the opportunity
to adjust the ratio of the two systems; this is advantageous in
fully over half an hour at 100-110 °C (eq 4).
(
38) ∆Hrxn for the dehydrogenation of indoline was calculated with
38a,b
Gaussian03
M. J.; Gaussian 03, revision B.04; Gaussian Inc.: Wallingford, CT,
004. (b) Dennington, R., II; Keith, T.; Millam, J.; Eppinnett, K.;
using the B3LYP/6-311++G** basis set. (a) Frisch,
2
3
Compound 3 (the BH adduct of indoline) was synthesized
Hovell, W. L.; Gilliland, R. GaussView, version 3.08; Semichem, Inc.:
in a similar manner as 1a. The connectivity of 3 was confirmed
Shawnee Mission, KS, 2003.
1
7198 J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008

H
2
from Combined Endothermic and Exothermic Hydrogen Carriers
A R T I C L E S
1
1
11
Figure 4. H NMR spectrum of 3 (top) and H{ B} NMR spectrum (bottom).
1
1
1
Figure 5. (A) H NMR spectrum of 3 is shown as a reference spectrum. (B) Cross-section slice of the H- H NOESY NMR spectrum for 3 was taken along
1
1
the C7 proton shift (7.4 ppm) depicting any correlations to other protons within the molecule. As seen in the figure, the H- H NOESY correlations
correspond to the amine proton (4.4 ppm) and the broad “quartet” of the borane protons (2.9-2.2 ppm).
order to achieve an overall heat balance within the system.
Because the exothermic and endothermic carriers that will be
used in this system have different enthalpy values, a 1:1 ratio
might not always be ideal for a heat balanced system. The first
combination of carriers that was explored was the exothermic
In order to avoid carrying any extra weight in the vehicle,
we wanted to look at a solventless system. Fortunately, indoline
is liquid at room temperature and has a boiling point of 220
°C, much higher than the dehydrogenation temperature and
the PEM fuel cell operating temperature. On the other hand,
the amine-boranes explored are not always liquid under the same
conditions; therefore, the solubility of the amine-boranes in the
endothermic carrier is of importance. The solubility of each
amine-borane within indoline is part of the following discussion.
Utilization of Triethylamine-Borane. Triethylamine-borane
2
,39-42
hydrolysis reaction of amine-boranes
(eq 6, e.g., H
3
N:
3
6,37
3
BH : ∆Hrxn ) -156 kJ/mol BH
3
) and the endothermic
43
dehydrogenation of indoline (eq 4, ∆H ) 52 kJ/mol). In order
to achieve heat balance in this system, an indoline to amine-
borane ratio of 3:1 would be required.
(
3 3
Et N:BH ) was first examined because it is a liquid at room
temperature and is fully miscible with indoline.
11
1
The reactivity of Et
96 Hz)) was first explored in the absence of indoline.
Triethylamine-borane is stable over palladium on carbon at 100
C; no decomposition products are observed after 2 h (Table 1,
3 3
N:BH ( B NMR: δ ) -13.6 (q, JBH
)
°
(
39) Amendola, S. C.; Sharp-Goldman, S. L.; Janjua, M. S.; Spencer, N. C.;
entry 1). When water is added under the same conditions, gas
and heat evolution is observed and the only product seen by
NMR spectroscopy is the hydrolysis product Et N:B(OH) ( B
3 3
NMR: δ 18; eq 7; Table 1, entry 2). Triethylamine-borane will
also not hydrolyze at room temperature unless catalyst is added
(Table 1, entries 5 and 3, respectively).
Kelly, M. T.; Petillo, P. J.; Binder, M. Int. J. Hydrogen Energy 2000,
2
5, 969–975.
11
(
(
(
40) Davis, W. D.; Mason, L. S.; Stegeman, G. J. Am. Chem. Soc. 1949,
1, 2775–2781.
41) Fakio g˘ lu, E.; Y u¨ r u¨ m, Y.; Verirolu, T. N. Int. J. Hydrogen Energy
7
˘
2
004, 29, 1371–1376.
42) Kojima, Y.; Suzuki, K.; Fukumoto, K.; Sasaki, M.; Yamamoto, T.;
Kawai, Y.; Hayashi, H. Int. J. Hydrogen Energy 2002, 27, 1029–
1
034.
(
43) ∆Hrxn for the dehydrogenation of indoline was calculated with
38a,b
Gaussian03
using the B3LYP/6-311++G** basis set.
J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008 17199

A R T I C L E S
Wechsler et al.
, Water, and
a
Table 1. Combinations of Indoline, Et
3 3
NBH , Water, and Catalyst
Table 2. Combinations of Indoline, Me
2
HN:BH
3
a
Catalyst
reagent equivalent
reagent equivalent
entry indoline Et
3
N:BH
3
2
H O
catalyst % conv. of indoline hydrolysis of BH
3
entry indoline Me
2
H:NBH
3
H
2
O
catalyst % conv. of indoline
hydrolysis of BH
3
1
2
3
4
5
6
7
8
9
1
1
1
1
1
0
0
0
0
0
1
1
1
1
1
1
1
1
3
1
1
1
1
1
0
0
0
0
1
1
1
1
1
0
3
3
3
3
0
3
0
3
0
0
3
3
3
Pd/C
Pd/C
Pd/C
none
none
none
none
Pd/C
Pd/C
none
Pd/C
none
Pd/C
Pd/C
no change
complete
complete
complete
no change
1
2
3
4
5
6
7
8
9
0
0
0
0
1
1
1
1
3
3
1
1
1
1
1
1
1
1
1
1
3
3
3
3
0
0
3
3
3
3
none
none
Pd/C
Pd/C
partial (∼95%)
no change
complete
b
b
b
b
complete
0
0
>99
>99
0
none
Pd/C
none
Pd/C
Pd/C
Pd/C
0
0
0
shifted to 6 ppm
shifted to 6 ppm
partial (∼95%)
complete
complete
complete
91
>99
72
0
1
2
3
4
no change
no change
complete
complete
complete
c
>99
10
0
a
>99
>99
Catalyst: 10% Pd/C at 2 mol % loading, time ) 2 h, temp ) 100
b
°C, under N
2
flow to remove produced hydrogen. Reaction performed
c
at room temperature. Reaction performed without N
2
flow.
a
Catalyst: 10% Pd/C at 5 mol % loading, time ) 2 h, temp ) 100
4
4
b
carriers. Thorn and co-workers, on the other hand, have
reported the use of exothermic non-H -releasing reactions to
supply the heat for the release of H from endothermic carriers.
°
2
C, under N flow to remove produced hydrogen. Reaction performed
at room temperature.
2
2
In certain cases, additional boron-containing species were
However, the text of the patent briefly mentions the dehydro-
genation of amine-borane using palladium or platinum (sup-
ported on carbon, etc.) combined with the dehydrogenation of
decalin at 250-280 °C, without providing experimental data.
The advantages of our system are that (a) each hydrogen
storage carrier can operate independently in a common chamber
without inhibiting the ability of the other to produce hydrogen,
(b) the reaction temperature is a moderate 100 °C, which is in
the operating temperature range of a PEM fuel cell, and (c) the
use of the endothermic carrier as a solvent to dissolve the
exothermic carrier facilitates material handling.
11
observed after the hydrolysis reaction was complete ( B NMR:
3
5-37
δ 19, 8, and 2). Xu and co-workers
have observed that
-
there is an equilibrium process between B(OH)
3
2
, BO , and
other borate species which is dependent on small changes in
the reaction conditions. The difference in chemical shift of the
borate species can potentially be attributed to a change in
available water equivalents and/or pH.
Indoline as the sole reagent in the absence of catalyst is stable
at 100 °C even in the presence of water (Table 1, entries 6 and
7
, respectively). However, in the presence of catalyst, indoline
completely dehydrogenates, even in the presence of water (Table
, entries 8 and 9, respectively).
Indoline and Et N:BH do not react with each other at 100
C (Table 1, entry 10). The hydrolysis of Et N:BH at 100 °C
The overall gravimetric hydrogen storage density of the
1
current system at a 3:1 ratio of indoline to Et
counting water because it could be supplied from the fuel cell)
is 2.5% (Et N:BH is 5.2%; indoline is 1.7%). The estimated
3 3
N:BH (not
3
3
°
3
3
3
3
overall enthalpy is 0 kJ/mol at a 3:1 ratio, based upon the
is not inhibited by the presence of indoline (compare Table 1,
entries 4 and 12). The catalytic dehydrogenation of indoline is
4
6
dehydrogenation enthalphy of indoline (52 kJ/mol H
the hydrolysis enthalpy of H N:BH (-156 kJ/mol BH
At a 1:1 mol ratio, the H
.4% and -104 kJ/mol, respectively. The feasibility of increas-
2
)
and
3
6,37
3
3
3
).
not inhibited by the presence of Et
and 11). This contrasts to the marked inhibition that was
observed when BH was bound directly to the endothermic
3
carrier (vide supra). When all reagents (indoline, Et N:BH ,
3 3
N:BH (Table 1, entries 8
2
capacity and overall enthalpy are
3
3
ing the hydrogen storage density by replacing the triethylamine-
borane with lower molecular weight boranes such as dimeth-
3
catalyst, and water) are reacted together, both indoline dehy-
drogenation and borane hydrolysis are observed, showing that
2 3 3 3
ylamine-borane (Me HN:BH ) and H N:BH will be described
next.
neither carrier system prevents the other from releasing its H
Table 1, entries 13 and 14). This confirms that endothermic
2
Utilization of Dimethylamine-Borane. Dimethylamine-borane
is a solid at room temperature but is completely soluble in
indoline at room temperature at a 1:1 ratio. The reactivity of
(
and exothermic hydrogen producing reactions can be operated
together in the same system without one carrier significantly
inhibiting or destabilizing the other.
Me
2 3
HN:BH in the presence of catalyst or acid and absence of
3
0
water has been summarized by Baker and co-workers.
In the current system, the BH
3
moiety remains bound to
Dimethylamine-borane does not undergo dehydrocoupling in
the presence of a Lewis acid, but if a precious metal catalyst is
implemented, dehydrocoupling occurs over several hours with
Et
3
N, and no evidence of the borane transferring to the
indoline is observed by NMR spectroscopy. No transfer was
expected because triethylamine is a better Lewis base than
indoline.
3
0
gentle heating (26-45 °C). We explored the reactivity of
Me HN:BH toward water in the absence of catalyst (Table 2,
2
3
To the best of our knowledge, there have been only two
4
4,45
entries 1 and 2); at 100 °C, only partial hydrolysis (∼20%) is
observed in 2 h, while at room temperature, no hydrolysis
occurred. In the presence of catalyst (eq 8; Table 2, entries 3
and 4), complete hydrolysis is achieved at both 100 °C and room
temperature.
reports
and exothermic dehydrogenation are described. Gelsey and co-
workers have first described the use of exothermic hydrolysis
with H release (NaBH + H O) and an endothermic dehydro-
genation (metal hydride), although the reactions were performed
in separate chambers to avoid direct contact between the two
where the simultaneous occurrence of endothermic
4
5
2
4
2
When Me
2 3
HN:BH is mixed with indoline and heated to 100
°
C in the absence or presence of catalyst, a new boron-
(
(
44) Thorn, D. L.; Tumas, W.; Ott, K. C.; Burrell, A. K. US20070183967,
007.
45) Gelsey, J. U. S. Patent 7,108,933, Intel Corporation, 2006.
2
(46) ∆Hrxn for the dehydrogenation of indoline was calculated with
Gaussian03 using the B3LYP/6-311++G** basis set.
1
7200 J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008

H
2
from Combined Endothermic and Exothermic Hydrogen Carriers
A R T I C L E S
a
Table 3. Combinations of Indoline, H
3 3
N:BH , Water, and Catalyst
reagent equivalent
1
1
containing species is produced ( B NMR: δ 6 ppm; Table 2,
entries 5 and 6, respectively). No dehydrogenation of the
indoline portion is observed even in the presence of catalyst
entry indoline
H
3
N:BH
3
H
2
O
catalyst % conv. of indoline
hydrolysis of BH
3
1
2
3
4
5
6
7
8
9
1
0
0
0
0
1
1
1
1
3
1
1
1
1
1
1
1
1
1
1
0
3
3
3
3
0
0
3
3
3
0
Pd/C
Pd/C
none
none
complete
complete
b
b
partial (∼75%)
(
Table 2, entry 6), suggesting a new amine borane complex
no change
similar to compound 3 is formed. The identity of this boron-
containing species is currently unknown and will be part of
future examinations.
none
Pd/C
none
Pd/C
Pd/C
Pd/C
0
0
0
shift to 25 ppm
shift to 25 ppm
complete
complete
complete
c
c
66
>99
>99
2 3
Combining all three reagents (Me HN:BH , indoline, and
water) in the absence of catalyst results in partial borane
hydrolysis (Table 2, entry 7 compared to entry 1). In the
presence of catalyst, complete hydrolysis of the borane and
good conversion of indoline are observed in 2 h (Table 2,
entry 8, eq 9). When the indoline to borane ratio is increased
to 3:1 (Table 2, entry 9), complete hydrolysis and dehydro-
genation are observed. All of these experiments, thus far,
0
a
Catalyst: 10% Pd/C at 2 mol % loading, time ) 2 h, temp ) 100
b
°C, under N flow to remove produced hydrogen. Reaction performed
2
c
at room temperature. Water was added to reaction after ∼30 min of
heating.
When the reaction (Table 3, entry 6) was repeated with the
addition of water in the absence of catalyst (Table 3, entry 7),
complete hydrolysis of the borane but no borane-indoline adduct
formation was observed.
were performed with N
remove H as it forms. Performing the reaction without N
gas flushing through the system, to allow the partial pressure
of H gas to build up, reduced the dehydrogenation of indoline
2
gas flushing through the system to
2
2
2
Both the hydrolysis of H
indoline are observed when a mixture of all three reagents
indoline, water, and H N:BH ) is exposed to palladium on
3 3
N:BH and the dehydrogenation of
to 72% conversion (entry 10 vs entry 9).
(
3
3
carbon (Table 3, entry 9). This suggests that the hydrolysis
reaction prevents or reverses adduct formation.
In order to determine the stability of the unknown indoline-
borane adduct, we combined indoline and ammonia-borane with
catalyst at 100 °C (as in Table 3, entry 6) for 1 h followed by
the addition of water and continued heating at 100 °C for 2 h.
The H and B NMR spectra were taken and showed complete
borane hydrolysis as well as partial indoline dehydrogenation
(Table 3, entry 8). This result suggests that once the BH moiety
3
is hydrolyzed the indoline dehydrogenation is no longer inhibited
to the same extent.
The overall hydrogen storage density of the current system
at a 3:1 ratio of indoline to Me
because it would be supplied from the fuel cell) is 2.9%
Me HN:BH is 11.4%; indoline is 1.7%). The estimated overall
enthalpy is the same as for Et N:BH . At a 1:1 mol ratio, the
capacity and overall enthalpy are 4.5% and -104 kJ/mol,
respectively.
Utilization of Ammonia-Borane. Use of H
Me HN:BH would again result in increased hydrogen capacity.
In this case, we were curious to see if the borane moiety of
2 3
HN:BH (not counting water
1
11
(
2
3
3
3
H
2
3 3
N:BH instead of
The overall hydrogen storage density of the current system
2
3
at a 3:1 ratio of indoline to H
because it would be supplied from the fuel cell) is 3.1% (H
BH is 19.4%; indoline is 1.7%). At a 1:1 mol ratio, the H
capacity is 5.3%. The estimated overall enthalpy is the same as
for Et N:BH and Me HN:BH
3
N:BH
3
(not counting water
1
1
1
3
N:
ammonia-borane ( B NMR: δ -23.4 (q, JBH ) 90 Hz)) would
transfer to indoline (as in the case of Me HN:BH ; Table 2,
2
3
3
2
entries 5 and 6) and in turn inhibit its dehydrogenation. The
stability and reactivity of ammonia-borane have been widely
3
3
2
3
.
3
0
Increasing the overall hydrogen storage density by employing
lower molecular weight amine-boranes only makes small
improvements to the overall H capacity because the amine-
2
borane is only one-quarter of the weight of the system. Larger
increases would be obtained by using a different endothermic
system that has a higher hydrogen storage density; this will be
part of future explorations.
studied. Unfortunately, ammonia-borane is a solid at room
temperature and is not soluble in indoline at a 1:1 or 3:1 indoline
to borane ratio until it is heated or in the presence of catalyst
and undergoes hydrogen loss.
3 3
We explored the reactivity of H N:BH toward water in the
absence of catalyst (Table 3, entries 3 and 4); at 100 °C, only
partial hydrolysis (∼75%) is observed in 2 h, while at room
temperature, no hydrolysis occurred. In the presence of catalyst
Conclusion
(
Table 3, entries 1 and 2), complete hydrolysis is achieved at
both 100 °C and room temperature.
In this paper, we explored the combination of two
independent hydrogen storage systems to determine if any
Whenever indoline and ammonia-borane are combined in the
absence of water, with or without catalyst, a shift to 25 ppm in
interactions occur between the two that might prevent H
2
11
the B NMR is observed, suggesting the formation of a new
boron-containing complex (Table 3, entries 5 and 6, respec-
release. Depending on the type of borane used, some
interaction was observed in the absence of water, while in
the presence of water, complete hydrolysis of the exothermic
carrier (triethylamine-, dimethylamine-, and ammonia-borane)
and complete dehydrogenation of the endothermic carrier
(indoline) are observed. This shows that, under the right
conditions, two hydrogen carriers can operate simultaneously
in the same system without inhibiting each other. The ratio
of the endothermic and exothermic systems can also be
altered to achieve overall heat balance.
1
tively). The H NMR spectra contained peaks matching those
3
of indoline plus additional peaks corresponding to a BH species.
11
The B NMR spectrum of this unknown borane-indoline adduct
does not match that of compound 3, suggesting that an
ammonia-borane derivative is formed which in turn complexes
with indoline. Although this adduct could not be identified, it
is clear that its formation inhibits the dehydrogenation of
indoline (compare Table 3, entry 6 with entry 10).
J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008 17201

A R T I C L E S
Wechsler et al.
The system is not, however, without its disadvantages.
Although the kinetics of the system have not been evaluated, it
is clear that hydrogen release from the exothermic system is
much faster than the hydrogen release from the endothermic
system; ideally, the reactions would have comparable rates. Also,
the borate hydrolysis products are not completely soluble in
indoline, so that precipitation of the product onto the catalyst
could, after some time, result in catalyst deactivation.
This work has also shown that a chemically bound
exothermic carrier tends to inhibit the dehydrogenation of
the endothermic carrier. In the case of compound 3, only
partial indoline dehydrogenation was observed in 2 h versus
complete dehydrogenation for indoline in the absence of any
borane. Therefore, the two moieties have to be separated to
prevent this inhibition.
NH), 5.26 (br s, 2H, NH
2
), 2.96 (d, 3JHH ) 12.1 Hz, 2H, C2-H
3
and C6-H), 2.44 (m, 1H, C4-H), 2.30 (q, JHH ) 11.1 Hz, 2H,
3
C2-H and C6-H), 1.95 (d, JHH ) 12.5 Hz, 2H, C3-H and
3
C5-H), 1.39 (q of d, JHH ) 13.0 Hz, JHH ) 3.5 Hz, 2H, C3-H
1
3
1
and C5-H), 1.26 (br m, 6H, BH
3
’s); C{ H} NMR (DMSO-d
6
) δ
11
5
2.4 (C4), 51.0 (C2 and C6), 29.4 (C3 and C5); B NMR (DMSO-
) δ -14.8 (br s, NH-BH ), -21.4 (br s, NH -BH ).
Synthesis of 1b. To a round-bottom flask equipped with a stir
d
6
3
2
3
bar and condenser were added compound 1a (0.039 g, 0.31 mmol)
and distilled water (∼1 mL). The mixture was heated to 100 °C,
and 5% palladium on silica (5 mol%) was added through the
condenser. After gas evolution ceased, more water (3 mL) was
added and the reaction mixture was filtered through diatomaceous
earth to remove the heterogeneous catalyst and other insolubles.
The filtrate was then dried in vacuo to remove any solvent and
other volatiles, resulting in a white solid (0.047 g, 0.21-0.25 mmol,
-
1
6
8-81%; depending on B(OH)
3
or BO
2
, respectively): H NMR
Future explorations will include a survey of other endothermic
carriers to increase the hydrogen storage density of the overall
system. Other exothermic carriers that would be more easily
regenerated will also be part of future explorations. Finally,
combinations of exothermic and endothermic carriers that have
3
(
1
D O) δ 3.31 (d, JHH ) 13.2 Hz, 2H, C2-H and C6-H), 3.07 (m,
2
3
H, C4-H), 2.86 (t of d, JHH ) 13.0 Hz, JHH ) 2.9 Hz, 2H, C2-H
3
and C6-H), 2.01 (d, JHH ) 13.4 Hz, 2H, C3-H and C5-H), 1.52
3
(
q of d, JHH ) 14.0 Hz, JHH ) 4.1 Hz, 2H, C3-H and C5-H);
13
1
C{ H} NMR (D
2
O) δ 45.8 (C4), 42.7 (C2 and C6), 29.8 (C3 and
O): δ 10.1 (br s).
comparable rates of H
2
release will be sought.
11
C5); B NMR (D
2
Synthesis of 2a. To a round-bottom flask equipped with a
Experimental Section
condenser containing a magnetically stirred solution of 1a (0.12 g,
0
.94 mmol) in THF (3 mL) was added 4-aminopiperidine (0.10
General Considerations. Manipulations of air-sensitive com-
pounds were conducted in the absence of oxygen and water under
an atmosphere of N , by use of standard Schlenk methods, utilizing
2
glassware that was oven-dried (130 °C) and evacuated while hot
prior to use. Indoline (Aldrich) was dried in vacuo at 80 °C for 2 h
mL, 0.94 mmol). The mixture was heated at 60 °C for 18 h. The
solvent and other volatiles were removed in vacuo. Methanol (4
mL) was added, and the solution was filtered through diatomaceous
earth. The filtrate was dried in vacuo, resulting in a white solid
1
3
(
0.16 g, 1.4 mmol, 72%): H NMR (CD
3
OD) δ 3.14 (d of d, JHH
prior to use. Triethylamine-borane (Aldrich), 1 M BH -THF
3
)
13.6 Hz, JHH ) 1.9 Hz, 2H, C2-H and C6-H), 2.78 (m, 1H,
solution in THF (Aldrich or Alfa Aesar), 10% palladium on
activated carbon (dry powder) (Strem Chemicals), hexanes (Fisher
Scientific), and dimethylamine-borane (Alfa Aesar) were used as
3
C4-H), 2.47 (apparent t, JHH ) 13.4 Hz, 2H, C2-H and C6-H),
3
1
.88 (d of d, J HH) 12.4 Hz, JHH ) 1.4 Hz, 2H, C3-H and C5-H),
3
47
1.39 (q of d, JHH ) 11.4 Hz, JHH ) 4.0 Hz, 2H, C3-H and C5-H);
received. Ammonia-borane was synthesized via literature methods.
1
3
1
C{ H} NMR (CD
3
OD) δ 53.2 (C2 and C6), 48.5 (C4), 35.3 (C3
Tetrahydrofuran and diethyl ether were obtained from Fisher
Scientific and passed through a double-column solvent purification
system purchased from Innovative Technologies, Inc. prior to use.
1
1
and C5); B NMR (CD
3
OD) δ -15.6 (apparent d, JBH ) 9.6 Hz);
OD) δ 35 (NH), 38 (NH ).
Synthesis of 2b. Ten percent palladium on carbon (0.054 g, 5
1
5
N NMR (CD
3
2
C
6
D
6
, CD
3 3 3 2
OD, CDCl , and (CD ) SO were purchased from
1
13
mol %, 0.051 mmol) was added to a round-bottom flask charged
with a stir bar and 2a (0.12 g, 1.0 mmol). Magnetic stirring was
initiated, and distilled water (0.5 mL) was added, resulting in
gas and heat evolution. The mixture was allowed to stir for 30
min followed by the addition of 2 mL of water and removal of
the catalyst by filtration through diatomaceous earth. The filtrate
was dried in vacuo, resulting in a pure white solid (0.12 g,
0.74-0.83 mmol, 73-83%; depending on B(OH)
respectively): H NMR (CD
Cambridge Isotope Laboratories Inc. and used as received. H, C,
1
1
and B NMR spectra were collected at 300 K on a Bruker AV-
00 spectrometer operating at 400.3, 100.7, and 128.4 MHz
respectively) with chemical shifts reported in parts per million
downfield of SiMe
and C NMR chemical shift assignments were made on the basis
4
(
1
13
11
1
4
(for H and C) or BF
3
2
-Et O (for B). H
1
3
1
3
1
1
1
13
of data obtained from C, H- H COSY, H- C HSQC, and
1
13
1
11
1
1
-
H- C HMBC NMR experiments. H{ B} and H- H NOESY
3
or BO
2
,
1
3
NMR experiments for compounds 1a and 3 were performed at 300
3
OD) δ 3.19 (apparent d, JHH
)
K on a Bruker AV-600 spectrometer at 600.2 MHz. Where reported,
13.0 Hz, 2H, C2-H and C6-H), 2.97 (m, 1H, C4-H), 2.75 (t
1
5
1
15
3
the N chemical shifts were assigned on the basis of a H- N
HMBC NMR experiment run at 300 K on a Bruker AV-600
spectrometer operating at 60.8 MHz with chemical shifts reported
of d, JHH ) 12.5 Hz, JHH ) 2.6 Hz, 2H, C2-H and C6-H),
3
1.97 (apparent d, JHH ) 12.6 Hz, 2H, C3-H and C5-H), 1.46
3
(q of d, JHH ) 12.1 Hz, JHH ) 4.0 Hz, 2H, C3-H and C5-H);
1
3
1
in parts per million with respect to CH
3
NO
2
at 381 ppm. Elemental
3
C{ H} NMR (CD OD) δ 49.0 (C4), 45.2 (C2 and C6), 33.8
15
1
1
analyses were attempted for compounds 1a,b, 2a,b, and 3, but each
sample returned lower than expected experimental values. This
could be due to the formation of thermally stable borocarbides in
(C3 and C5); B NMR (CD
(CD OD) δ 32 (NH), 36 (NH
Synthesis of 3. To a Schlenk flask charged with a stir bar were
added indoline (0.80 mL, 7.1 mmol) and 1 M BH -THF solution in
THF (14.3 mL, 14.3 mmol) and stirred overnight (16 h). Solvent was
removed in vacuo, leaving behind a yellow solid to which hexanes
was added (3 mL). The hexanes dissolved the yellow impurity, leaving
behind a white precipitate. The hexanes were decanted, and the
remaining solid was washed several times with fresh hexanes (3 × 5
mL) followed by drying in vacuo, leaving behind a white solid (0.80
g, 6.0 mmol, 85%). Storage of compound 3 at low temperature is
preferred to inhibit decomposition. It appears to be mildly pyrophoric
3
OD) δ 10.4 (br s); N NMR
3
2
).
4
8
boron-containing molecules during elemental analysis. Mass
spectroscopic techniques that could ionize these molecules without
3
3
loss of BH were not found.
Synthesis of 1a. A Schlenk flask was charged with a stir bar
and 4-aminopiperidine (1.0 mL, 9.4 mmol) followed by the addition
of 2-3 equiv of 1 M BH -THF in THF (25 mL, 25 mmol) and
3
magnetically stirred for 16 h. The solvent was removed in vacuo,
leaving behind a white solid. The solid was washed several times
with distilled water and dried in vacuo, yielding 1a as a white solid
1
1
(
1.04 g, 8.1 mmol, 86%): H NMR (DMSO-d
6
) δ 5.78 (br s, 1H,
in air. In solution under air, it degrades over hours: H NMR (C D )
6
6
3
3
δ 7.42 (d, JHH ) 7.8 Hz, 1H, C7-H), 6.94 (t, JHH ) 7.3 Hz, 1H,
C6-H), 6.88 (t, JHH ) 7.2 Hz, 1H, C5-H), 6.75 (d, JHH ) 7.3,
C4-H), 5.01 (br s, 1H, NH), 3.07 (m, 1H, C2-H ), 2.87 (m, 1H,
3
3
(
(
47) Ramachandran, P. V.; Gagare, P. D. WO2007106459, 2007.
48) Borda, P. P.; Legzdins, P. Anal. Chem. 1980, 52, 1777–1778.
2
1
7202 J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008

H
2
from Combined Endothermic and Exothermic Hydrogen Carriers
A R T I C L E S
C2-H), 2.50 (m, 1H, C3-H
2
), 2.16 (m, 1H, C3-H
; partially underneath CH
) δ 146.5 (C3a), 134.9 (C7a), 128.0 (C6), 127.7 (C5),
2
), 2.49 (apparent
Data were processed on a PC using the Bruker AXS Crystal
Structure Analysis Package: Data collection: APEX2 (Bruker,
1
13
1
49
q, JHB ) ∼90 Hz, 3H, BH
NMR (C
6 6
25.2 (C4), 120.3 (C7), 55.3 (C2), 29.0 (C3); B NMR (C D ) δ -13.8
3
2
peaks); C{ H}
6 6
D
2006); cell refinement: APEX2; data reduction: SAINT (Bruker,
2005) and XPREP (Bruker, 2005); program(s) used to solve
structure: SHELXTL (Bruker, 2000); program(s) used to refine
structure: SHELXTL; molecular graphics: SHELXTL; software
used to prepare material for publication: SHELXTL. Neutral atom
11
1
(
1
q, JBH ) 91 Hz, BH
Dehydrogenation and Hydrolysis of Indoline/R
H or Et) Mixtures. The 1:1:3 reaction mixture of indoline, Et
BH , and water is described here as the general protocol. For other
reaction ratios, the same conditions were followed.
3
).
3
N:BH
3
(R
N:
)
3
50
3
scattering factors were taken from Cromer and Waber. The crystal
is orthorhombic space group Pna2 , based on the systematic
1
A round-bottom flask was charged with a stir bar, indoline (0.18
mL, 1.6 mmol), triethylamine-borane (0.24 mL, 1.6 mmol), distilled
water (0.087 mL, 4.8 mmol), and 10% palladium on carbon (0.085
g, 0.08 mmol) and equipped with a condenser. The reaction mixture
was immersed into a preheated oil bath (set at 100 °C) and stirred
absences, E statistics, and successful refinement of the structure.
The structure was solved by direct methods. Full-matrix least-
2
2 2
squares refinements minimizing the function ∑w (F
applied to the compound. All non-hydrogen atoms were refined
anisotropically. The hydrogen atoms -CH and dNH were
calculated (C-H ) 0.99 Å and N-H ) 0.93 Å) and refined using
o
c
- F ) were
2
2
magnetically for 2 h under a N flow for the duration of the
experiment. In order to get efficient borane hydrolysis, vigorous
stirring is required due to the insolubility of indoline and triethy-
lamine-borane in water to produce consistent dispersion of the water
throughout the reaction mixture. The reaction flask was then allowed
to cool to room temperature, and the reaction mixture was extract
a riding model, with Uiso(H) ) 1.2Ueq(N,C). The -NH
hydrogen atoms were located from difference Fourier maps and
refined isotropically.
2
and -BH
3
There are two molecules in the asymmetrical unit, and each
molecule is hydrogen-bonded to its two neighboring molecules
3
with CDCl followed by filtration through a small plug of
(
2 2
N-H· · ·NH and H N· · ·H-N), presenting a 1-D chain along the
diatomaceous earth to remove the remaining catalyst. The filtrate
was analyzed by NMR spectroscopy, showing indole and hydro-
lyzed triethylamine-borane.
b axis. Additional crystallographic information is provided in the
accompanying CIF.
Dehydrogenation and Hydrolysis of Indoline/Me
2
HN:BH
3
Acknowledgment. The authors gratefully acknowledge financial
support from the Natural Sciences and Engineering Research
Council of Canada, the Ontario Centres of Excellence (Energy),
Defence Research and Development Canada, AUTO21, Chrysler
Canada, and the Canada Research Chairs program. We also thank
Dr. Francoise Sauriol for her assistance in the acquisition of NMR
data.
Mixtures. The 3:1:3 reaction mixture of indoline, Me
2
HN:BH ,
3
and water is described here as the general protocol. For other
reaction ratios, the same reaction conditions were followed.
A 25 mL Schlenk tube was charged with a stir bar, indoline
(
0.34 mL, 3.0 mmol), dimethylamine-borane (0.059 g, 1.0 mmol),
distilled water (0.054 mL, 3.0 mmol), and 10% palladium on carbon
0.064 g, 0.06 mmol). The tube was equipped with a condenser,
immersed into a preheated oil bath, and stirred magnetically for
h at 100 °C under N flow for the duration of the experiment.
The reaction flask was then allowed to cool to room temperature,
and the reaction mixture was extracted with CDCl or CD OD and
filtered through a small plug of Al powder to remove any
(
Supporting Information Available: Single-crystal X-ray dif-
2
2
13
fraction data in CIF format for 2a. C NMR spectra for 1a,b,
1
1
11
3
3
2
a,b, and 3. H, H{ B}, and NOESY NMR spectra for 1a
2
O
3
and 3. This material is available free of charge via the Internet
at http://pubs.acs.org.
remaining catalyst. The filtrate was analyzed by use of NMR
spectroscopy.
Crystallographic Characterization of 2a. A crystal of the
compound (colorless, plate-shaped, size 0.39 × 0.28 × 0.11 mm) was
mounted on a glass fiber with grease and cooled to -93 °C in a stream
of nitrogen gas controlled with Cryostream Controller 700. Data
collection was performed on a Bruker SMART APEX II X-ray
diffractometer with graphite-monochromated Mo KR radiation
JA806721S
(49) Bruker AXS Crystal Structure Analysis Package; Bruker, 2000;
SHELXTL, version 6.14; Bruker AXS Inc.: Madison, WI, 2005;
XPREP, version 2005/2; Bruker AXS Inc.: Madison, WI, 2005; SAINT,
version 7.23A; Bruker AXS Inc.: Madison,WI, 2006; APEX2, version
(
λ ) 0.71073 Å), operating at 50 kV and 30 mA over 2θ ranges of
.82-52.00°. No significant decay was observed during the data
collection.
2
.0-2; Bruker AXS Inc.: Madison, WI.
4
(
50) Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystal-
lography; Kynoch Press: Birmingham, U.K., 1974; Vol. 4, Table 2.2A.
J. AM. CHEM. SOC. 9 VOL. 130, NO. 50, 2008 17203