Highly efficient colloidal cobalt- and rhodium-catalyzed hydrolysis of H3N·BH3 in air

Article abstract of DOI:10.1021/ic700806b

The compound H₃N·BH₃ (1) is currently attracting considerable attention as a potential hydrogen storage material. Group 9 catalysts which rapidly and conveniently hydrolyze aqueous 1 in air are described. When treated with 1 mol % [{Rh(μ-Cl)(1,5-cod)}₂] (cod = cyclooctadiene) in air, aqueous 1 undergoes rapid hydrolysis to afford the ionic species [NH₄][BO₂] in ～40 s. Higher catalyst loadings (3 mol %) result in a reduction in reaction time to 10 s. Quantification of the hydrogen evolved revealed that, on average, 2.8 of a maximum possible 3.0 equivalents (93%) were generated during the course of the reaction. Rh(0) species (e.g., Rh black, Rh stabilized on alumina, aqueous Rh colloids) were also found to be active hydrolysis catalysts, and evidence for a heterogeneous mechanism is provided. Significantly, although Ir(0) colloids are less active, aqueous Co(0) colloids are also effective catalysts for this process. This result is particularly important as Co, a first-row metal, is considerably more economical than the precious metal catalysts typically employed.

Full text of DOI:10.1021/ic700806b

Inorg. Chem. 2007, 46, 7522−7527
Highly Efficient Colloidal Cobalt- and Rhodium-Catalyzed Hydrolysis of
H N BH in Air
‚
3
3
Timothy J. Clark,† George R. Whittell,‡ and Ian Manners*,†,‡
Department of Chemistry, UniVersity of Toronto, 80 St. George Street, Toronto, Ontario,
Canada, M5S 3H6, and School of Chemistry, UniVersity of Bristol, Cantocks Close, Bristol,
United Kingdom, BS8 1TS
Received April 27, 2007
The compound H
Group 9 catalysts which rapidly and conveniently hydrolyze aqueous 1 in air are described. When treated with 1
mol % [ Rh( -Cl)(1,5-cod) ] (cod cyclooctadiene) in air, aqueous 1 undergoes rapid hydrolysis to afford the
ionic species [NH ][BO ] in 40 s. Higher catalyst loadings (3 mol %) result in a reduction in reaction time to 10
s. Quantification of the hydrogen evolved revealed that, on average, 2.8 of a maximum possible 3.0 equivalents
93%) were generated during the course of the reaction. Rh(0) species (e.g., Rh black, Rh stabilized on alumina,
3 3
N‚BH (1) is currently attracting considerable attention as a potential hydrogen storage material.
{
µ
}
2
)
4
2
∼
(
aqueous Rh colloids) were also found to be active hydrolysis catalysts, and evidence for a heterogeneous mechanism
is provided. Significantly, although Ir(0) colloids are less active, aqueous Co(0) colloids are also effective catalysts
for this process. This result is particularly important as Co, a first-row metal, is considerably more economical than
the precious metal catalysts typically employed.
Introduction
as low hydrogen storage densities and slow reaction kinetics.5
Ammonia-borane, H N‚BH (1), is now being studied
extensively, as it possesses one of the highest densities of
hydrogen available (19.6 wt % H ).
2
Our group has previously investigated the catalytic dehy-
3
3
Transition metal-catalyzed dehydrocoupling and dehydro-
genation reactions are of growing interest as a preparative
route to rings, chains, polymers, and materials based on main
group elements. Further impetus to this area has recently
been provided by the realization that the catalytic dehydro-
genation of hydrogen-rich inorganic compounds may play a
key role in their utility as hydrogen storage materials. Thus,
as environmental concerns over carbon-based emissions
grow, the search for an alternative vehicle fuel to petroleum
6
1
,2
drogenation of a variety of amine-borane adducts, including
1
(Scheme 1). Our results have shown that both heteroge-
7
,8
9
neous (e.g., Rh(0)) and homogeneous (e.g., Ti) catalysts
effectively promote amine-borane dehydrocoupling with
useful scope and activity. Other groups have recently reported
1
0
11
3
key advances in the area of thermal or catalytic elimina-
has become increasingly critical. To this end, hydrogen
storage materials have attracted significant attention, as they
(5) Gross, K. J.; Thomas, G. J.; Jensen, C. M. J. Alloys Compd. 2002,
330-332, 683.
4
can potentially store H
2
safely and efficiently. The most
(
6) H3N‚BH3 readily meets the targets for 2010 set by the U.S. Department
of Energy for an on-board chemical hydrogen storage material (6.0
wt% H2); see ref 3 for details.
studied materials, metal hydrides, suffer from drawbacks such
*
To whom correspondence should be addressed. E-mail: Ian.Manners@
(7) (a) Jaska, C. A.; Temple, K.; Lough, A. J.; Manners, I. Chem. Commun.
2001, 962. (b) Jaska, C. A.; Temple, K.; Lough, A. J.; Manners, I. J.
Am. Chem. Soc. 2003, 125, 9424. (c) Jaska, C. A.; Manners. I. J. Am.
Chem. Soc. 2004, 126, 9776.
(8) Autrey and co-workers have also studied the dehydrocoupling of Me2-
NH‚BH3 using Rh catalysts under conditions that differ from those
used in our group. On the basis of their observations, they propose
that the process may be homogeneous: Chen, Y.; Fulton, J. L.;
Linehan, J. C.; Autrey, T. J. Am. Chem. Soc. 2005, 127, 3254.
(9) Clark, T. J.; Russell, C. A.; Manners. I. J. Am. Chem. Soc. 2006, 128,
9582.
(10) (a) Gutowska, A.; Li, L.; Shin, Y.; Wang, C. M.; Li, X. S.; Linehan,
J. C.; Smith, R. S.; Kay, B. D.; Schmid, B.; Shaw, W.; Gutowski, M.;
Autrey, T. Angew. Chem., Int. Ed. 2005, 44, 3578. (b) Bluhm, M. E.;
Bradley, M. G.; Butterick, R., III; Kusari, U.; Sneddon, L. G. J. Am.
Chem. Soc. 2006, 128, 7748.
bristol.ac.uk.
†
University of Toronto.
‡
University of Bristol.
1) (a) Gauvin, F.; Harrod, J. F.; Woo, H. G. AdV. Organomet. Chem.
998, 42, 363. (b) Tilley, T. D. Acc. Chem. Res. 1993, 26, 22. (c)
Reichl, J. A.; Berry, D. H. AdV. Organomet. Chem. 1998, 43,
97.
(
1
1
(
2) For a recent review on metal-catalyzed dehydrocoupling reactions
see: Clark, T. J.; Lee, K.; Manners, I. Chem. Eur. J. 2006, 12, 8634.
3) (a) Dresselhaus, M.; Crabtree, G.; Buchanan M. Eds. Basic Research
Needs for the Hydrogen Economy; Basic Energy Sciences, Office of
Science, U.S. Department of Energy: Washington, DC, 2003. (b) The
Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs;
National Academies Press: Washington, DC, 2004.
(
(
4) Grochala, W.; Edwards, P. P. Chem. ReV. 2004, 104, 1283.
7522 Inorganic Chemistry, Vol. 46, No. 18, 2007
10.1021/ic700806b CCC: $37.00
© 2007 American Chemical Society
Published on Web 07/31/2007

Group 9 Metal-Catalyzed Aqueous Dehydrogenation of H
3
N‚BH
3
in Air
be prepared in four steps.11d,14 In this paper, we report our
Scheme 1
studies on the metal-promoted hydrolysis of 1, which can
be synthesized in one step or purchased, using relatively low
loadings of both bulk and colloidal Group 9 metals as
catalysts in water. Moreover, we show that these efficient
catalytic reactions proceed under very convenient conditions
in air.
Scheme 2
Experimental Section
General Procedures and Materials. All reactions and manipu-
lations were performed in air unless stated otherwise. As H is
2
released during dehydrogenation reactions and is a flammable gas,
these experiments should be conducted in a fumehood. Distilled
water and ACS grade tetrahydrofuran were used without any further
tion of H from 1 and other high-hydrogen-content amine-
borane adducts. Regeneration of the hydrogen storage
material is a key issue, and although multistep processes can
2
be considered, the recent observation of reversible H release
2
purification. H
diene), Rh black, and Rh/Al
received. [Co (CO) ] was sublimed immediately before use. Aque-
ous colloidal Rh(0) and Ir(0) solutions were prepared according
to literature procedures. Generally, a solution of the MCl ‚xH
M ) Rh, Ir) salt in H O was treated, with vigorous stirring, with
an aqueous mixture of the reductant Li[BH ] and surfactant
3
N‚BH
3
, [{Rh(µ-Cl)(1,5-cod)}
2
] (cod ) cycloocta-
2
O
3
were purchased and used as
2
8
15
16
from a phosphonium-borate species is particularly note-
worthy in this regard.12,13
3
2
O
(
2
Hydrolysis reactions provide an additional method for H
release from hydrogen-rich amine-borane adducts such as
(Scheme 2). For example, Xu and Chandra have described
2
4
N-hexadecyl-N-(2-hydroxyethyl)-N-dimethylammonium bromide,
giving a black suspension in both cases.
1
the metal-catalyzed hydrolysis of 1 under an inert atmosphere
of Ar with various late transition metals; Pt(0) gave the most
impressive results with complete reaction being observed in
ca. 2-10 min with 20 wt % Pt on C.1 In addition, these
workers investigated the addition of undissolved Rh catalysts
and observed slightly reduced reactivity and the elimination
Equipment. NMR spectra were recorded on a Jeol ECP 300
11
spectrometer operating at 96 MHz ( B). Chemical shifts are
reported relative to external BF ‚Et O.
Catalytic Hydrolysis of (1) by [{Rh(µ-Cl)(1,5-cod)}
solution of 1 (0.050 g, 1.62 mmol) in H O (ca. 5 mL), [{Rh(µ-
Cl)(1,5-cod)} ] (0.004 g, 0.0081 mmol, 1 mol %) in THF (ca. 1
3
2
1a
2
]. To a
2
2
mL) was added at 20 °C resulting in the immediate release of a
gas as evidenced by vigorous bubbling and a color change to black.
After ∼40 s, the bubbling ceased, an aliquot of the reaction mixture
of 2.5 of a possible 3.0 equiv of H
2
in ∼15 min. Of particular
relevance to this paper is their recent observation that
supported non-noble metals (e.g., Ni, Co) are also effective
catalysts for this process.1
was removed and the 11B NMR spectrum obtained showing a single
1e
- 11a,17
resonance at 9 ppm corresponding to [BO
2
] .
Sneddon and co-workers have recently described the acid-
Increasing the [{Rh(µ-Cl)(1,5-cod)}
2
] catalyst loading for the
and metal-mediated hydrolysis of ammonia-triborane,
above reaction to 3 mol % (0.012 g) resulted in complete conversion
(2).1 For example, 2 was reported to rapidly
1d
to [BO
] in 10 s as judged by the cessation of H
-
evolution using
NH B H
3 3 7
2
2
release H
2
in aqueous solution with moderately high loadings
a gas burette.
Hydrogen Gas Volumetric Measurements. A solution of 1
0.034 g, 1.10 mmol) in H O (ca. 5 mL) was prepared as above in
a two-necked round-bottom flask with one neck used for catalyst
solution injection through a rubber septum and the other neck
connected via plastic tubing to a 100 mL gas burette containing
of Rh-based catalysts (ca. 7 mol %) in 1.5 min. Although
the use of 2 can potentially yield more equivalents of H
per molecule compared to 1, the wt % H per molecule of 2
17.8 wt %) is actually less than that of 1. Furthermore,
unlike 1, it is not a commercially available material but must
(
2
2
2
(
2 2
H O. To the above solution, [{Rh(µ-Cl)(1,5-cod)} ] (0.009 g, 0.019
mmol, 3 mol %) in THF (ca. 1 mL) was added, causing 73 mL of
(
11) (a) Chandra, M.; Xu, Q. J. Power Sources 2006, 156, 190. (b) Chandra,
M.; Xu, Q. J. Power Sources 2006, 159, 855. (c) Denney, M. C.;
Pons, V.; Hebden, T. J.; Heinekey, D. M.; Goldberg, K. I. J. Am.
Chem. Soc. 2006, 128, 12048. (d) Yoon, C. W.; Sneddon, L. G. J.
Am. Chem. Soc. 2006, 128, 13992. (e) Xu, Q.; Chandra, M. J. Power
Sources 2006, 163, 364. (f) Stephens, F. H.; Baker, R. T.; Matus, M.
H.; Grant, D. J.; Dixon, D. A. Angew. Chem., Int. Ed. 2007, 46, 746.
H
H
2
O (of a maximum 79.5 mL based on the maximum of 3 mol of
2
that can be eliminated by the complete dehydrogenation of 1)
to be displaced in the gas burette. A repeat trial resulted in 75 mL
being displaced.
Testing Rh Catalyst Recyclability. To a solution of 1 (0.050
(g) Nguyen, M. T.; Nguyen, V. S.; Matus, M. H.; Gopakumar, G.;
2 2
g, 1.62 mmol) in H O (ca. 5 mL), [{Rh(µ-Cl)(1,5-cod)} ] (0.004
Dixon, D. A. J. Phys. Chem. A. 2007, 111, 679. (h) Cheng, F.; Ma,
H.; Li, Y.; Chen, J. Inorg. Chem. 2007, 46, 788. (i) Keaton, R. J.;
Blacquiere, J. M.; Baker, R. T. J. Am. Chem. Soc. 2007, 129, 1844.
12) Welch, G. C.; San Juan, R. R.; Masuda, J. D.; Stephan, D. W. Science
g, 0.0081 mmol, 1 mol %) in THF (ca. 1 mL) was added at 20 °C
resulting in the immediate release of a gas as evidenced by vigorous
bubbling and a color change to black. After ∼40 s, the bubbling
ceased, an aliquot of the reaction mixture was removed, and the
(
(
2006, 314, 1124.
13) The catalytic dehydrocoupling of phosphine-borane adducts has also
been reported: (a) Dorn, H.; Singh, R. A.; Massey, J. A.; Lough, A.
J.; Manners, I. Angew. Chem., Int. Ed. 1999, 38, 3321. (b) Dorn, H.;
Singh, R. A.; Massey, J. A.; Nelson, J. M.; Jaska, C. A.; Lough, A.
J.; Manners, I. J. Am. Chem. Soc. 2000, 122, 6669. (c) Dorn, H.;
Rodezno, J. M.; Brunnh o¨ fer, B.; Rivard, E.; Massey, J. A.; Manners,
I. Macromolecules 2003, 36, 291. (d) Clark, T. J.; Rodezno, J. M.;
Clendenning, S. B.; Aouba, S.; Brodersen, P. M.; Lough, A. J.; Ruda,
H. E.; Manners, I. Chem. Eur. J. 2005, 11, 4526.
11
B NMR spectrum obtained showing a single resonance at 9 ppm
(14) Nainan, K. C.; Ryschkewitsch, G. E. Inorg. Nucl. Chem. Lett. 1970,
6, 765.
(15) Schulz, J.; Roucoux, A.; Patin, H. Chem. Eur. J. 2000, 6, 618.
(16) M e´ vellec, V.; Roucoux, A.; Ramirez, E.; Philippot, K.; Chaudret, B.
AdV. Synth. Catal. 2004, 346, 72.
(17) Smith, H. D., Jr.; Wiersema, R. J. Inorg. Chem. 1972, 11, 1152.
Inorganic Chemistry, Vol. 46, No. 18, 2007 7523

Clark et al.
Table 1. Reaction Times Required for the Complete Catalytic
Hydrolysis of Additional Equivalents of 1 by the Rh Catalyst Derived
from [{Rh(µ-Cl)(1,5-cod)}2]
Table 2. Reaction Times Required for the Complete Catalytic
Hydrolysis of Additional Equivalents of 1 by Aqueous Colloidal Rh(0)
no. of equiv of 1
reaction time (min)
no. of equiv of 1
reaction time
1
2
3
4
5
25
25
25
35
40
1
2
3
4
40 s
6 min
15 min
35 min
-
11a,17
corresponding to [BO
2
] .
Subsequent equivalents of 1 (0.050
of 1 (0.050 g, 1.62 mmol) were added and the reactions monitored
by B NMR spectroscopy in order to determine the amount of
time required for complete consumption of 1 (results summarized
in Table 2).
11
11
g, 1.62 mmol) were added and the reactions monitored by B NMR
spectroscopy in order to determine the amount of time required
for complete consumption of 1 (results summarized in Table 1).
Catalytic Hydrolysis of (1) by Rh(0) Species. To a solution of
Catalytic Hydrolysis of (1) by Aqueous Colloidal Ir(0). To a
1
0
(0.050 g, 1.62 mmol) in H
.032 mmol, 1 mol %) was added at 20 °C resulting in the
2
O (ca. 5 mL), Rh/Al
2
O
3
(0.033 g,
solution of 1 (0.049 g, 1.59 mmol) in H O (ca. 4 mL), a 9.33 ×
2
-
4
10 M solution of colloidal Ir(0) (17 mL, 0.0159 mmol, 1 mol %)
immediate release of a gas as evidenced by vigorous bubbling. After
1 min, the bubbling ceased and an aliquot of the reaction mixture
in H O was added at 20 °C resulting in the immediate release of a
gas as evidenced by bubbling. The reaction was monitored by B
NMR spectroscopy in order to determine the amount of time
required for complete consumption of 1 (105 min). A repeat trial
took 110 min.
2
1
1
∼
11
was removed. The B NMR spectrum obtained showed a single
-
11a,17
2
resonance at 9 ppm corresponding to [BO ] .
Similar results were obtained for Rh black.
Catalytic Hydrolysis of (1) by Aqueous Colloidal Rh(0). To
Treatment of Me NH‚BH (3) with a Catalytic Amount of
2
3
a solution of 1 (0.050 g, 1.62 mmol) in H
2
O (ca. 18 mL), a 4.74 ×
M solution of colloidal Rh(0) (3.40 mL, 0.0161 mmol, 1 mol
) in H O was added at 20 °C resulting in the immediate release
Co(0). This reaction was carried out in an argon-filled glovebox.
To a solution of 3 (0.250 g, 4.24 mmol) in toluene (ca. 3 mL), a
catalytic amount of bulk Co(0) powder (0.005 g, 0.085 mmol, 2
mol %) was added and stirred in an open vessel for 16 h. After
this time, an aliquot of the reaction mixture was removed and an
-
3
1
%
0
2
of a gas as evidenced by vigorous bubbling. The reaction was
monitored by 11B NMR spectroscopy in order to determine the
amount of time required for complete consumption of 1 (25 min).
Treatment of Aqueous Colloidal Rh(0) with Hg Prior to
11B NMR spectrum obtained showing only unreacted 3 (
B
δ -14
ppm).
-
3
Addition of (1). To a 4.74 × 10 M solution of colloidal Rh(0)
No reaction was observed when the same procedure was carried
out with [Co (CO) ].
Treatment of (1) with a Catalytic Amount of Co(0). To a
solution of 1 (0.200 g, 6.48 mmol) in H O (ca. 5 mL), bulk Co(0)
powder (0.004 g, 0.068 mmol, 1 mol %) was added and stirred for
h. After this time, an aliquot of the reaction mixture was removed
(
3.40 mL, 0.0161 mmol, 1 mol %) in H O, Hg (4.50 g, 22.43 mmol)
2
2
8
was added and stirred rapidly for 45 min. IMPORTANT: Good
stirring is known to be needed to ensure contact of the catalyst
with Hg(0), a condition required to aVoid false negatiVes in this
2
1
8
experiment. To this mixture, 1 (0.050 g, 1.62 mmol) in H
2
O (ca.
5
1
8 mL) was added at 20 °C and the reaction monitored by 11
B
11
and an B NMR spectrum obtained showing the presence of 10%
]- and 90% unreacted 1.
BO
NMR spectroscopy. After 25 min, an aliquot of the reaction mixture
[
2
was removed and the 11B NMR spectrum obtained showing only
Preparation of Aqueous Colloidal Co(0). Under an atmosphere
of dry N , lithium borohydride (0.032 g, 1.47 mmol) was added to
a deoxygenated aqueous solution (ca. 120 mL) of N-hexadecyl-N-
2-hydroxyethyl)-N-dimethylammonium bromide (0.360 g, 0.91
mmol). This solution was added all at once to a vigorously stirred,
deoxygenated aqueous solution (ca. 6 mL) of CoCl ‚6H O (0.077
-
3
5% conversion of 1 to [BO
2
] .
2
Filtration Experiment: Catalytic Hydrolysis of (1) by Aque-
ous Colloidal Rh(0). To a solution of 1 (0.050 g, 1.62 mmol) in
(
-
3
H
0
2
O (ca. 4 mL), a 4.74 × 10 M solution of Rh(0) (3.40 mL,
.0161 mmol, 1 mol %) in H O was added at 20 °C resulting in
2
2
2
the immediate release of a gas as evidenced by vigorous bubbling.
After 25 min, the reaction was judged to be complete by 11B NMR.
The reaction mixture was then filtered through a 200 nm porous
filter giving a black solution. To this solution, a second equivalent
of 1 (0.050 g, 1.62 mmol) was added. After 25 min, an aliquot of
g, 0.32 mmol) to give a black Co(0) colloidal suspension (126 mL).
Catalytic Hydrolysis of (1) by Aqueous Colloidal Co(0). To a
solution of 1 (0.050 g, 1.62 mmol) in H
2
O (ca. 18 mL), a 4.71 ×
M solution of colloidal Co(0) (3.40 mL, 0.0160 mmol, 1 mol
) in H O was added at 20 °C resulting in the immediate release
-
3
10
%
2
11
the reaction mixture was taken and the B NMR spectrum was
of a gas as evidenced by bubbling. The reaction was monitored by
B NMR spectroscopy in order to determine the amount of time
required for complete consumption of 1 (60 min). A repeat trial
also took 60 min.
-
obtained showing the presence of 58% [BO
2
] and 42% unreacted
11
1.
A blank reaction, where the reaction mixture was not filtered
prior to addition of the second equivalent of 1, showed a single
11
resonance in the B NMR spectrum at 9 ppm corresponding to
Results and Discussion
-
after 25 min.11a,17
[
BO
2
]
Testing Aqueous Colloidal Rh(0) Catalyst Recyclability. To
a solution of 1 (0.050 g, 1.62 mmol) in H
O (ca. 5 mL), a 4.74 ×
M solution of Rh(0) (3.40 mL, 0.0161 mmol, 1 mol %) in
O was added at 20 °C resulting in the immediate release of a
gas as evidenced by vigorous bubbling. After 25 min, the reaction
was judged to be complete by 11B NMR. Subsequent equivalents
i. Rhodium-Catalyzed Hydrolysis of 1. We have previ-
ously shown that 1 undergoes catalytic dehydrocoupling in
diglyme with Rh (pre)catalysts such as [{Rh(µ-Cl)(1,5-
2
-
3
1
H
0
2
cod)}
2
] at 45 °C under N
2
to afford borazine [HN-BH]
3
and insoluble oligomeric/polymeric products over 48-84 h
7
a,b
(
Scheme 1). Notably, when an aqueous solution of 1 is
treated with a 1 mol % THF solution of the same Rh (pre)-
catalyst in air, an instantaneous color change to black
(
18) Widegren, J. A.; Finke, R. G. J. Mol. Catal. A.: Chem. 2003, 198,
317.
7524 Inorganic Chemistry, Vol. 46, No. 18, 2007

Group 9 Metal-Catalyzed Aqueous Dehydrogenation of H
3 3
N‚BH in Air
Figure 1. Photographs depicting an aqueous solution of 1 before 1 mol % Rh catalyst addition (t ) 0 s) and at various stages of reaction after Rh addition
until completion at t ) 40 s.
Scheme 3
consistent with the formation of colloidal Rh(0) was observed
along with vigorous gas evolution (Figure 1). After 40 s,
11
the bubbling subsided and B NMR spectroscopy confirmed
that the starting material had been completely consumed by
the absence of a resonance at -24.5 ppm and revealed a
new resonance at 9 ppm. We assign this chemical shift to
the borate species [NH
4
][BO
2
] (Scheme 3).1
1a,17
This remark-
Figure 2. Graph of % conversion (as determined by 11B NMR) versus
time for the catalytic hydrolysis of 1 employing 1 mol % of colloidal Co
ably short reaction time was further reduced to 10 s upon
increasing the Rh catalyst loading to 3 mol %. As noted
above, Chandra and Xu have recently studied the late metal-
catalyzed hydrolysis of 1 under an inert atmosphere. By
comparison, they reported that 1.8 mol % [{Rh(µ-Cl)(1,5-
(green triangles), Rh (blue squares), and Ir (red circles) aqueous solutions.
1
1a
prepared by the reduction of transition metal salts by borane
or borohydride species, exhibit reduced activity toward the
cod)}
in 15 min under an Ar atmosphere. Here, our results indicate
that over repeat trials, the addition of [{Rh(µ-Cl)(1,5-cod)}
as a solution in air significantly decreases the reaction time.
In order to quantify the amount of H gas eliminated during
2
], added in the solid state, catalytically hydrolyzed 1
19
dehydrocoupling of amine-borane adducts. This was
attributed to a passivating layer of the Group 13 element on
the metal surface blocking access of the substrate to catalyst
active sites.
The observed color change to black along with the
formation of visible dark precipitate is presumably due to
2
]
2
the course of the reaction, we measured the displacement of
water in a gas burette. Over two runs, an average 74 mL of
2
reduction of the Rh(I) (pre)catalyst [{Rh(µ-Cl)(1,5-cod)} ]
a maximum 79.5 mL of H
.8 out of a possible 3.0 equivalents of H
the stoichiometry shown in Scheme 3. The extent of H
2
were released, which equates to
to catalytically active Rh(0) metal. Although it is likely that
the catalyst is heterogeneous, a homogeneous metal species
cannot be completely ruled out at this stage.18 This is
supported by our previous studies on late metal-catalyzed
dehydrocoupling of amine-borane adducts under an inert
atmosphere, where a variable-time induction period was
observed as the catalytically active heterogeneous species
is formed prior to dehydrogenation of the adduct.7
2
2
(93%), confirming
2
release from 1 here, particularly in such a short period of
time at room temperature, makes this catalytic system, which
operates in air, highly promising.
The recyclability of the Rh catalyst was also explored.
Immediately following a hydrolysis reaction with 1, another
equivalent of 1 was added leading to further hydrogen
evolution from the reaction mixture. However, the time
needed to completely consume 1 increased to 6 min, which
can still be regarded as a short reaction time although
significantly lengthened relative to the 40 s required for the
initial equivalent. Subsequent treatment with a third and
fourth equivalent of 1 resulted in further increases in reaction
time to 15 and 35 min, respectively (Table 1). This apparent
decrease in catalytic activity may be attributed to aggregation
of the Rh(0) particles over time resulting in a decrease in
catalyst surface area. Another potential contributing factor
could be that species with B-H functionalities inhibit
hydrolytic activity by slowly poisoning the catalyst surface.
We have previously reported that heterogeneous catalysts
treated with Group 13 hydrides, as well as colloidal metals
b,c
Consequently, we turned our attention to Rh(0) catalysts
in an effort to investigate and compare their catalytic activity
toward 1. Treatment of an aqueous solution of 1 in air with
1
2 3
-2 mol % Rh black or Rh/Al O again resulted in
immediate gas evolution and complete conversion to
1
1
[
4 2
NH ][BO ] by B NMR spectroscopy in ca. 1-4 min. As
the reaction appeared to be heterogeneous and is performed
in water, an aqueous colloidal Rh solution was also pre-
1
5
pared and its catalytic activity examined. Upon treatment
of 1 with a 1 mol % aqueous solution of colloidal Rh(0) in
air, bubbling was again observed and complete reaction was
detected after 25 min (Figure 2, blue squares). This increase
(
19) Jaska, C. A.; Clark, T. J.; Clendenning, S. B.; Grozea, D.; Turak, A.;
Lu, Z.-H.; Manners, I. J. Am. Chem. Soc. 2005, 127, 5116.
Inorganic Chemistry, Vol. 46, No. 18, 2007 7525

Clark et al.
in reaction time relative to the previously mentioned Rh
systems may be due to the dilute solutions required in order
to prevent aggregation of the colloids to give insoluble bulk
Rh metal. An alternative possibility is that the surfactant,
while stabilizing the colloidal Rh(0) against aggregation, may
also be hindering the access of 1 to catalyst active sites.
Significantly, rapid stirring of the Rh(0) solution with a large
excess of Hg followed by similar treatment with 1 resulted
Scheme 4
2
0
analogous to our previous observations on the catalytic
dehydrocoupling of Me NH‚BH (3) in organic solvents
under an inert atmosphere which yields the cyclic dimer
Me N-BH . In particular, we reported that the room-
temperature reaction with [{Rh(µ-Cl)(1,5-cod)} ] took 8 h
2
3
4 2
in only 35% conversion to [NH ][BO ] after 25 min. This
decrease in rate is likely caused by poisoning of the Rh(0)
[
2
2 2
]
2
7b
catalyst by Hg; a well-known heterogeneous catalyst poi-
2
while that of [{Ir(µ-Cl)(1,5-cod)} ] took 136 h. Although
in the case described here the difference in reaction time is
son.1
8,21
The heterogeneity of the process was further
supported by filtration experiments, which indicate an
insoluble catalyst is the active species if the activity is
lowered upon filtering through small (ca. 200 nm) pore
membrane filters.18 Two hydrolysis reactions of 1 were
carried out simultaneously with the aqueous colloidal Rh(0)
not as dramatic, the same trend is observed.
iii. Cobalt-Catalyzed Hydrolysis of 1. As noted by Baker
and co-workers in a recent report detailing their Ni-catalyzed
11i
dehydrocoupling of 1 (at 60 °C in diglyme over 4 h), there
is a general desire not just for greater catalyst activity and
stability but also for catalysts comprising more abundant,
economical first-row metals. In this context, our studies of
the Group 9 metal-catalyzed hydrolysis of 1 were extended
to Co. We found that treatment of 3 with 2 mol % of the
1
1
solution; after 25 min both reactions were complete by
B
NMR. A second equivalent of 1 was added to one of the
reactions and stirred, while the other reaction mixture was
filtered through a 200 nm porous filter giving a black solution
prior to addition of a second equivalent of 1. After a further
2 8
Co(0) source [Co (CO) ] or bulk Co(0) powder did not result
2
5 min, 11B NMR revealed complete reaction had occurred
in the case of the former, while in the latter case, only 58%
of the second equivalent had formed [NH ][BO ]. This
in dehydrocoupling and formation of the cyclic dimer
11
[Me
2 2 2
N-BH ] after 16 h as determined by B NMR
4
2
spectroscopy (Scheme 4). Furthermore, treatment of an
strongly supports the assertion that the key active catalytic
species in this process is heterogeneous.
The recyclability of the aqueous colloidal Rh(0) catalyst
was also tested in a similar manner to that used for the above
Rh catalyst. In contrast to the Rh catalyst derived from [{Rh-
aqueous solution of 1 with 2 mol % bulk Co(0) powder
4 2
resulted in only 10% conversion to [NH ][BO ] after 5 h as
11
judged by B NMR spectroscopy. These results contrast with
those of Chandra and Xu who reported that 1 was completely
consumed by various insoluble, supported Co(0) catalysts
1
1e
(µ-Cl)(1,5-cod)}
2
], the activity of the aqueous colloidal Rh-
2 3 2
including Co/Al O , Co/SiO , and Co/C.
(0) catalyst does not decrease appreciably until the addition
In contrast to our studies employing molecular Co(0)
species and bulk Co metal, an aqueous solution of colloidal
Co(0) (1 mol %) was found to be an active hydrolysis catalyst
for 1, as the reaction was found to be complete in only 60
min (Figure 2, green triangles). The distinct difference in
catalytic activity apparent in these results and our observa-
tions involving bulk and colloidal Co(0) is of importance
and underscores the change in behavior that can be observed
upon shifting to smaller particle sizes. By comparison,
Chandra and Xu reported that 1 undergoes complete hy-
drolysis when treated with 1.8 mol % of Co/C over the same
period of time, reflecting reasonably consistent reaction times
for the Co(0) catalyzed reaction. They also detailed that Co/C
catalysts gave the best results as other Co(0) systems such
as Co/Al O took longer to catalyze the hydrolysis of 1 (70
of a fourth equivalent of 1, as the initial three equivalents
all took 25 min to be completely converted to [NH ][BO ],
4
2
the fourth taking 35 min (Table 2). By comparison, a fourth
equivalent of 1 also required 35 min for complete conversion
to [NH
Cl)(1,5-cod)}
4
][BO
2
] with the Rh catalyst derived from [{Rh(µ-
]; however, the first equivalent only took 40
2
s. We attribute this significant difference in catalyst recy-
clability to the surfactant stabilizing the aqueous colloids,
which likely hinders aggregation resulting in reasonably
constant reaction times. Another possibility is that the high
surface area of the colloidal Rh(0) catalyst may also prevent
appreciable surface passivation by species containing B-H
1
9
bonds.
ii. Iridium-Catalyzed Hydrolysis of 1. Other aqueous
Group 9 metal colloid solutions were examined in order to
determine whether they exhibit catalytic activity toward the
2
3
1
1e
min) under the same conditions.
This difference in
reactivity was attributed to particle size effects as the Co/C
particles were found to be smaller than that of Co/Al O as
1
6
hydrolysis of 1. An aqueous solution of colloidal Ir(0) in
air was found to be effective, although not as efficient as
Rh(0), as complete conversion was found to have taken place
after 105 min (Figure 2, red circles). These findings are
2
3
determined by powder X-ray diffraction. Significantly, both
our results on colloids and those of Chandra and Xu on
supported systems demonstrate that the cheaper and more
abundant first-row metal Co exhibits greater reactivity toward
(
20) The various reaction times for the bulk or colloidal metal catalysts
are uncorrected for the different surface areas and number of active
sites of these different heterogeneous catalysts.
1
than the precious metal Ir.
Summary
(
21) (a) Anton, D. R.; Crabtree, R. H. Organometallics 1983, 2, 855. (b)
Whitesides, G. M.; Hackett, M.; Brainard, R. L.; Lavalleye, J. P. P.
M.; Sowinski, A. F.; Izumi, A. N.; Moore, S. S.; Brown, D. W.; Staudt,
E. M. Organometallics 1985, 4, 1819.
As part of our continuing research on the catalytic
dehydrocoupling and dehydrogenation of amine-borane
7526 Inorganic Chemistry, Vol. 46, No. 18, 2007

Group 9 Metal-Catalyzed Aqueous Dehydrogenation of H
3
N‚BH
3
in Air
efficient multistep regeneration of 1 from the hydrolysis
adducts, we have explored the bulk or colloidal Group 9
metal-catalyzed hydrolysis of the promising hydrogen storage
material 1. We have shown that the aqueous Rh-catalyzed
hydrolysis of 1 in air can occur at 25 °C at rates that
significantly exceed those of other reported catalysts, with
products and analogous studies of other amine-borane
adducts. The role of air in the promotion of the heterogeneous
catalytic dehydrogenation chemistry relative to corresponding
reactions under an inert atmosphere is also under investiga-
tion.
2
2.8 of a maximum possible 3.0 equivalents H eliminated in
seconds in some cases. Furthermore, we have studied the
extension of this method to aqueous solutions of other
colloidal Group 9 metals such as Co, a cheap and abundant
first-row transition metal. Thus, Co(0) colloids also ef-
ficiently and conveniently catalyze the hydrolysis of 1 in
air at 25 °C. Future work is directed toward exploring the
Acknowledgment. T.J.C. thanks the Ontario Government
for a Graduate Scholarship (OGS). I.M. thanks the European
Union for a Marie Curie Chair and the Royal Society for a
Wolfson Research Merit Award.
IC700806B
Inorganic Chemistry, Vol. 46, No. 18, 2007 7527