COMMUNICATION
Precipitation of 2 is not observed in this reaction, and the proton
resonances do not match those of the MeAB dehydrogenation
product.
Evaluation of the potential viability of hydrogen storage
materials requires data for the thermodynamics of hydrogen
release from the materials. There have been several computa-
tional investigations6,19-21 of AB dehydrogenation, but experi-
mental verification under relevant conditions has remained
elusive. Wolf and co-workers have reported the enthalpy of AB
dehydrogenation in the solid state, but no solution data are
available.22 The ability of 1 to catalyze rapid H2 loss allowed
for the direct determination of the enthalpy of these reactions.
The reaction enthalpy (∆H) for AB dehydrogenation in THF
at 303 K (2.5 mol % 1) was found to be -6.7 ( 0.6 kcal mol-1
by differential scanning calorimetry (DSC). For the dehydro-
genation of MeAB and 1:1 AB/MeAB catalyzed by 1, ∆H
values of -6.8 ( 0.6 and -6.7 ( 0.1 kcal mol-1, respectively,
were determined. A value of ∆H ) -10.7 kcal mol-1 was
calculated by Matus and co-workers for AB dehydrogenation
to the cyclic trimer.21 Thus, the results reported here provide
experimental support for computational studies, suggesting that
reversible hydrogen absorption and release under mild condi-
tions using amine boranes may not be feasible. Alternate
methods of regenerating the parent amine borane compounds
from spent materials will be needed.
In the 11B NMR spectrum, broad resonances at -8.7 and
-11.4 ppm were observed, while the 13C{1H} NMR spectrum
showed a broad resonance at 36-37 ppm. As with the soluble
MeAB dehydrogenation product, the presence of broad reso-
nances in both the 13C{1H} and 11B NMR spectra is most
consistent with various inequivalent boron and carbon environ-
ments, suggesting that similar mixtures of oligomers were
formed from the 1:1 mixture of AB/MeAB.
Because the product of dehydrogenation of 1:1 mixtures of
AB/MeAB remained in solution, subsequent reactions were
carried out with increasing ratios of AB/MeAB in an attempt
to form products that would remain soluble but would contain
a higher weight percent of hydrogen. The addition of catalyst
1 to a 2:1 mixture of AB/MeAB in THF resulted in rapid
evolution of 1 equiv of H2, accompanied by precipitation of a
white solid. The addition of 1 to a 5:1 mixture of AB/MeAB
in THF proceeded in a similar fashion but produced significantly
more solid. In each case, the IR spectrum of the insoluble
product was consistent with the formation of 2.9
In summary, the iridium complex 1 is an efficient catalyst
for the rapid dehydrogenation of MeAB and AB/MeAB
mixtures. The rates of dehydrogenation are similar to those seen
for AB dehydrogenation. The dehydrogenation reactions of both
MeAB and mixtures of AB/MeAB with 1 result in soluble
products. Dehydrogenation of MeAB results in a mixture of
cyclic and noncyclic oligomers of varying chain lengths. The
dehydrogenation of mixtures of AB and MeAB also yield
multiple cyclic and noncyclic oligomeric products. The amount
of AB incorporation appears limited to 1:1 relative to MeAB
as evidenced by the observation of ca. 1:1 ratios of dehydro-
genated [NH2BH2] and [MeNHBH2] fragments and the ad-
ditional formation of insoluble pentamer 2 with reactions where
AB/MeAB ratios greater than 1:1 were employed. The mea-
sured exothermicity of the dehydrogenation reactions suggests
that direct regeneration under mild conditions is not feasible.
These AB/MeAB reactions were repeated in THF-d8 using
the same total substrate concentration but varying ratios of AB/
MeAB, and the reactions were monitored by 1H NMR
spectroscopy (Figure 1). Regardless of the initial ratio of AB
to MeAB, the signals observed in each spectrum are the same
as those observed in the 1:1 AB/MeAB reaction, with the
exception of what appears to be an additional methyl resonance
observed at 2.28 ppm in the 5:1 reaction. Although the NH
resonances have the same chemical shifts in each of the separate
reactions, the signal intensities vary depending on the initial
AB/MeAB ratio. Similarly, ESI-MS data show identical isotopic
distribution patterns for each reaction but with varying intensities
of the peaks.18 The results from the different mixed AB/MeAB
reactions suggest that the soluble product consists of oligomers
of varying chain lengths, with the initial ratio of AB/MeAB
determining which oligomers are formed preferentially.
1
For all three reactions, integration of the N-Me H NMR
Acknowledgment. This work was supported by the U.S.
Department of Energy (DOE) as part of the Center of Excel-
lence for Chemical Hydrogen Storage. PNNL is operated for
the DOE by Battelle. Thanks to A. St. John, L. Kruse and M.
Sadilek at UW for skilled technical assistance. Thanks to Abhi
Karkamkar of PNNL for assistance with the calorimetry.
resonances confirms that all MeAB is incorporated into the
soluble products. Notably, integration of the N-Me resonances
versus the NH protons of the products from both the 2:1 and
5:1 AB/MeAB reactions is consistent with a 1:1 ratio of
[NH2BH2] and [MeNHBH2] fragments incorporated into the
soluble product. This observation parallels the larger production
of insoluble 2 as the ratio of AB to MeAB was increased.
ESI-MS spectra of the soluble products show isotopic
distribution patterns consistent with cyclic compounds of the
general formula [NH2BH2]x[MeNHBH2]y and noncyclic oligo-
mers of the general formula [NH2BH2]x[MeNHBH2]y[H2]. All
observed distribution patterns can be described by formulas in
which x and y do not differ by more than 3, which is consistent
Supporting Information Available: Characterization of products,
quantification of H2 evolved, DSC experimental details, and ESI-MS
data. This material is available free of charge via the Internet at
IC801161G
(19) Gutowski, M.; Autrey, T. Prepr. Symp.sAm. Chem. Soc., DiV. Fuel
Chem. 2004, 49, 275.
(20) Li, J.; Kathmann, S. M.; Schenter, G. K.; Gutowski, M. J. Phys. Chem.
C 2007, 111, 3294.
(21) Matus, M. H.; Anderson, K. D.; Camaioni, D. M.; Autrey, T. S.; Dixon,
D. A. J. Phys. Chem. A 2007, 111, 4411.
(22) Wolf, G.; Baumann, J.; Baitlow, F.; Hoffman, F. P. Thermochim. Acta
2000, 343, 19.
1
with the 1:1 ratio of N-Me versus NH H NMR signals.
However, because isotopic distribution patterns in which the
disparity between x and y > 3 encompass mass ranges similar
to those signals already described, larger differences cannot be
ruled out.
Inorganic Chemistry, Vol. 47, No. 19, 2008 8585