Concentration-dependent dehydrogenation of ammonia-borane/triglyme mixtures

Article abstract of DOI:10.1002/ejic.201100795

The dehydrogenation reaction of the hydrogen-storage compound ammonia-borane (AB) dissolved in triglyme (TG) reveals an unexpected but favourable concentration dependence of its dehydrogenation behaviour. Formal reaction-order considerations and ¹¹B NMR spectroscopic analytics point towards a complex AB decomposition and hydrogen-release scheme that in highly concentrated solutions partly begins to resemble that of the solid-state case. The analysis of the gas-release characteristics identified two hydrogen desorption events that display an increasing peak overlap with increasing initial AB concentration. ¹¹B NMR spectroscopic studies indicate that the concentration-dependent stability of the intermediate diammoniate of diborane, the presence of which appears to influence the rates of several dehydrogenation steps by interacting with various other intermediates, is one of the key factors for the overall hydrogen release.

Full text of DOI:10.1002/ejic.201100795

FULL PAPER
DOI: 10.1002/ejic.201100795
Concentration-Dependent Dehydrogenation of Ammonia–Borane/Triglyme
Mixtures
Johannes F. Kostka,[a] René Schellenberg,[b] Felix Baitalow,[b] Tom Smolinka,[a] and
Florian Mertens*[b]
Keywords: Ammonia borane / Dehydrogenation / Hydrogen storage / Kinetics / Reaction mechanisms / Thermolysis
The dehydrogenation reaction of the hydrogen-storage com- of the gas-release characteristics identified two hydrogen de-
pound ammonia–borane (AB) dissolved in triglyme (TG) re-
veals an unexpected but favourable concentration depen-
dence of its dehydrogenation behaviour. Formal reaction-or-
der considerations and 11B NMR spectroscopic analytics
point towards a complex AB decomposition and hydrogen-
release scheme that in highly concentrated solutions partly
begins to resemble that of the solid-state case. The analysis
sorption events that display an increasing peak overlap with
increasing initial AB concentration. 11B NMR spectroscopic
studies indicate that the concentration-dependent stability of
the intermediate diammoniate of diborane, the presence of
which appears to influence the rates of several dehydrogena-
tion steps by interacting with various other intermediates, is
one of the key factors for the overall hydrogen release.
melting point of AB.[2] These findings suggest a kinetic en-
hancement for the dehydrogenation due to the presence of
mobile phases.
Introduction
Ammonia–borane (AB) is regarded as a promising can-
didate to fulfil the requirements for a high-density and easy-
to-handle hydrogen source material.[1] With its high hydro-
gen content of 19 wt.-% and a release of 13.5 wt.-% at tem-
peratures below 140 °C,[1,2] it possesses outstanding quali-
ties for this application. In this respect, one important ob-
stacle to using pure AB is given by its poor dehydrogenation
kinetics, which would limit its use to applications with a
high energy-demand-to-power ratio. To achieve the neces-
sary improvements in dehydrogenation kinetics, a detailed
understanding of structure, hydrogenation/dehydrogenation
mechanisms and their rate-limiting steps is imperative and
have been the subject of a significant number of studies
during the past decade.[1–5]
Consistent with earlier findings that the thermally in-
duced dehydrogenation reaction of solid AB has an induc-
tion period, recent investigations show that the dehydroge-
nation of solid AB may be characterized by a nucleation
and growth process.[3] After the induction period, the re-
lease seems to follow a two-step scheme.[2] The first release
step has its onset at about 120 °C, which is just above the
In a recent publication,[3] thermal decomposition of solid
AB was found to be initiated by a species AB*, which is
assumed to be mobile. Thereafter, two mobile AB* units
may form the intermediate called diammoniate of diborane
[BH2(NH3)2]+[BH4]– (DADB), which is regarded as a nu-
cleus for the dehydrogenation reaction. Once formed, or ex-
ternally added to pure AB, DADB seems to promote the
dehydrogenation kinetics significantly.[3] The kinetics of this
reaction were also substantially enhanced in ionic liquids
(e.g., bmimCl, in which bmim = 1-butyl-3-methylimidaz-
olium).[4] Here, too, DADB was described as an intermedi-
ate.[4] From their recent work on AB in diglyme, Autrey et
al.[6] reported on possible traces of DADB in a 2 m solution
at 50 °C. From these results they concluded that DADB in
solutions requires stabilization by a polar environment.[4,6]
Hence, controlling the DADB formation by the means of
the reaction environment could enhance the mobility of AB
and seems to be a promising approach to improve the dehy-
drogenation kinetics. For high gravimetric hydrogen densi-
ties in the storage medium, however, the weight fraction of
the reaction environment (e.g., the solvents or catalysts) and
the chemical performance has to be optimized.
[a] Fraunhofer Institute for Solar Energy Systems, Dept. Energy
Technology
In this paper, the findings of significant amounts of
DADB in highly concentrated solutions of AB in the
weakly polar solvent triglyme (TG) and its positive conse-
quences for the dehydrogenation kinetics of AB and the
possible underlying reaction pathways will be discussed. For
this purpose, we combined the method of gas volumetric
measurements with ex situ 11B NMR spectroscopy.
Heidenhofstrasse 2, 79110 Freiburg, Germany
Fax: +49-761-4588-9320
E-mail: johannes.kostka@ise.fraunhofer.de
[b] Technische Universität Bergakademie Freiberg, Fakultät für
Chemie und Physik
Leipziger Strasse 29, 09599 Freiberg, Germany
Fax: +49-3731-39-35-88
E-mail: Florian.Mertens@chemie.tu-freiberg.de
Eur. J. Inorg. Chem. 2012, 49–54
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
49

J. F. Kostka, R. Schellenberg, F. Baitalow, T. Smolinka, F. Mertens
FULL PAPER
t
x = 1/4(H2) on [AB]0–1 can be estimated, which would corre-
Results and Discussion
Dehydrogenation Kinetics
spond to a second-order kinetics (see Figure 2a). In con-
trast, the kinetics of the later dehydrogenation phase are
better characterized by a linear dependence of t1/2(H2) and
t3/4(H2) on [AB]0–2, which would correspond to a formal
third-order rate law for the formation of H2 as a function
of [AB]0 (see Figure 2b).
Dehydrogenation characteristics of the thermolysis reac-
tion (at T = 70 °C) were investigated over a wide range of
concentrations, from dilute solutions of [AB]0 = 0.1 m up to
suspensions of [AB]0 = 23.0 m (Figure 1a). The time needed
to release 1 mol of H2 per mol of AB [t1/2(H2)] as a function
of the initial AB concentration ([AB]0) decays asymptoti-
cally to (40Ϯ10) min at the saturation concentration
[AB]0 ≈ [AB]sat = 7.5 m and rises again for concentrations
[AB]0 Ն 9.0 m (suspensions). Experiments with concentra-
tions well above [AB]sat show nucleation and growth char-
acteristics that are typical of solid AB according to the lit-
erature[3] (see Figure 1). Hence, the best dehydrogenation
kinetics are achieved in highly concentrated AB solutions.
Therefore, the work presented will only focus on systems
with AB in solution.
Figure 2. Time elapsed until the release of 0.5, 1.0 and 1.5 mol-
equiv. of H2 [tx = 1/4, 1/2, 3/4(H2)] at T = 70 °C plotted versus (a)
–1
[AB]0 (a distinct deviation from a second-order behaviour is ob-
served for tx Ն 1/2) and (b) [AB]0–2 (as a test for third-order kinetics).
The simple assumption that one third-order reaction is
governing the dehydrogenation process appears to be very
unlikely if one takes into account that AB decomposes in
glymes according to a second-order rate law with respect to
the AB concentration.[6] In addition, calculations of the rate
constants according to a third-order law do not result in a
single value for all curves in Figure 2b, thus indicating that
this rate law also does not apply.
Figure 1. Hydrogen-release characteristics of AB in TG at
T = 70 °C for mixtures with concentrations between 1.5 and 18 m.
(a) Plot of the half-time of the total hydrogen release (i.e., the
equivalent of 1 mol of H2 per mol of BH3NH3 versus the concen-
tration of the initial AB concentration in TG. (b) Amount of hydro-
gen released as a function of time given in mol H2 per mol
BH3NH3.
One possible explanation for the aforementioned dehy-
drogenation behaviour can be extracted from the time de-
rivative of the released hydrogen gas volume (see Figure 3).
In a first step to determine the kinetics of the dehydroge- It can be seen that at least two distinct desorption events
nation reaction, the emitted gas volume was characterized more strongly overlap with increasing [AB]0 concentrations.
by a plot of the time required to release 0.5, 1.0 and 1.5 mol The simultaneous occurrence of these two desorption
H2 per mol of AB (t1/4, t1/2 and t3/4, respectively) as a func- events leads to an enhanced hydrogen development at
–1
–2
tion of [AB]0 (Figure 2a) and [AB]0 (Figure 2b). For higher concentrations during the early reaction phase (see
the early dehydrogenation phase, a linear dependence of Figure 2a, left-hand side).
50
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 49–54

Dehydrogenation of Ammonia–Borane/Triglyme Mixtures
reaction of two AB molecules to DADB with a subsequent
cyclization to cyclodiborazane (CDB, c-B2N2H8) under the
release of 1 mol-equiv. H2 (see Scheme 1). Such a mecha-
nism was recently proposed for AB/glyme mixtures by Au-
trey et al. but could not yet be confirmed since the authors
found only traces of DADB at low reaction temperatures.
Scheme 1. Proposed dehydrogenation pathways of AB in TG at
70 °C based on NMR spectroscopy. Solid boxes denote unambigu-
ous observed signals; dashed boxes indicate proposed species based
on results of Shaw et al.;[6] dotted boxes frame BH2 species. DADB
was observed only at [AB]0 Ͼ 1.5 m but seems to promote the for-
mation of B-(cyclodiborazanyl)aminoborohydride (BCDB).
Figure 3. Time derivation of the gas volume of the hydrogen re-
leased at 70 °C from AB/TG solutions with different initial AB
concentrations.
Moreover, the first event displays its maximum hydrogen
flow significantly after t0. Such behaviour indicates that the
hydrogen release is not governed by a single elementary de-
composition reaction of any order but needs to proceed
through an intermediate compound.
To understand the underlying mechanistic steps responsi-
ble for such behaviour, we coupled the dehydrogenation ex-
periments with ex situ 11B NMR spectroscopy.
Our 11B NMR spectra of the early reaction of dilute
solutions ([AB]0 Ͻ 1.5 m) also show only two significant
resonances. Beside the prominent AB-related quintet at
δ = –21.5 ppm, a signal at δ = –10.5 ppm is observed, which
is assigned to CDB according to other publications (see
Table 1).[6,7]
For highly concentrated solutions ([AB]0 Ͼ 1.5 m),
we observe an additional quintet resonance at δ =
(–36.5Ϯ0.5) ppm (see Figure 4a), which can be assigned to
DADB.[3,4] This assignment is also supported by the corre-
11B NMR Spectroscopy of the Reaction Phase t < t1/4
As stated above, during the early reaction phase, the de- lation with the signal of the BH2 group at δ = –10.7 ppm.
hydrogenation kinetics can be fit to a second-order law, Because various components {CDB, cyclotriborazane
which may be satisfactorily explained by the bimolecular (c-B3N3H12, CTB), B-(cyclodiborazanyl)aminoborohydride
Table 1. Assignment of 11B NMR spectroscopic signals to the relevant species found during AB/TG dehydrogenation experiments.
Species
11B (20 °C)[a]
[ppm]
1JB,H
[Hz]
Referenced
literature
[3,6]
[6,7]
[7,9]
[6]
DABD
AB
μ-ADB
BCDB
CTB, CDB, BCDB
Boroxide (traces)
Polyborazylene
Borazine
(–36.5Ϯ0.5) (BH4)[b]
–21.5 (BH3)
–26.5 (H-N-BH2)
–4.5 (NH2BH3)
–10 to –13 (BH2)
1.5 (BO)
79 (quint)
78 (quint)
56 (d), 113 (t)
[c]
[c]
[3,4,6]
–
–
–
[d]
[4,6]
25.0 (BHNH)x(BH2)x[(dB] H3)x
[3,4,6,7]
30.5 (HNBH)3
113 (d)
[a] 11B chemical shifts are referenced to the IUPAC standard BF3·OEt2 (δ = 0 ppm) at 20 °C. [b] Species only found in samples
[AB] Ͼ 1.5 m. [c] No multiplets observable. [d] [AB] Ͻ 18 m.
Eur. J. Inorg. Chem. 2012, 49–54
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
51

J. F. Kostka, R. Schellenberg, F. Baitalow, T. Smolinka, F. Mertens
FULL PAPER
(cyc[NH2BH2NH2BH]-NH2BH3, BCDB)} also contribute
to this signal, it is not correlated with the DADB signal at
δ = –36.5 ppm and thus does not follow its decrease with
the progress of the dehydrogenation reaction. Our findings
in highly concentrated solutions strongly support Autrey
and co-workers’[6] assumption that the AB decomposition
proceeds by means of the formation of DADB and CDB to
CTB. Because our resolution of the δ = –10.7 ppm peak is
not sufficient, it is not evident whether or not CDB forms
after or simultaneously with DADB.
Figure 5. Excerpt of 11B NMR spectra representing only BH2 sig-
nals. All spectra are normalized by their integrals, representing the
total boron content of each sample and thus allowing the direct
comparison of the intensities. Grey: spectrum of a DADB-enriched
AB/TG solution before dehydrogenation. Black: spectra of AB/TG
solutions at different stages of dehydrogenation. The very shallow
maximum at δ = –10.7 ppm at the beginning of the recording of
the hydrogen evolution (rel. H2 = 0) is most likely due to a minute
hydrogen release that was not detected during the sample prepara-
tion.
Figure 4. (a) 11B{1H} NMR spectra depicting the dehydrogenation
evolution form an AB/TG solution ([AB]0 = 6.0 m) at 70 °C. (b)
The quintet signal at δ ≈ (–36.5Ϯ0.5) ppm in the coupled spectra
is identified as the BH4 resonance of DADB and can be traced
almost along the whole dehydrogenation process.
11B NMR Spectroscopy of the Reaction Phase t Ͼ t1/2
By the determination of the relative peak area of the sig-
nal at δ ≈ 30.5 ppm, we estimate that about (15Ϯ2)% of
Although it cannot be stated for diluted solutions the total boron content is bound in borazine in all samples
([AB]0 Յ 1.5 m) whether or not the absence of DADB-re- with [AB]0 Ͼ 1.5 m after the reaction has proceeded to
lated signals is due to DADB concentrations that are below t1/2(H2) (see Figure 4).
the detection limit, the observation, however, that the hy-
Samples with lower concentrations in which no DADB
drogen release rate for [AB]0 = 1.5 m reaches its maximum was observed show significantly less borazine [about
shortly after t0 (see Figure 3) supports the assumption of (5Ϯ1)%] at this point of reaction. These findings suggest
the presence of an early intermediate, be it DADB or CDB. that the hydrogen release from these mixtures is 15% higher
During the early reaction phase t Ͻ t1/4, the BH2 signal than would be expected from the reaction of AB through
at δ ≈ –10.7 ppm grows and slightly shifts to δ = –10.0 ppm DABD to CDB alone.
(see Figure 5). From DADB-enriched AB/TG solutions
The DADB-related signal peaks around t1/4 with a rela-
compared to pure AB/TG solutions, one has to assign tive peak area of (2Ϯ1)% for [AB]0 = 3.0 m and of (4Ϯ1)%
the signal at δ = –10.7 ppm to the BH2 group of DADB. for [AB]0 = 6.0 m. The correlation between the high DADB
Based on the experiments with the diluted solutions and the high borazine concentrations at t1/2 suggests that
([AB]0 Յ 1.5 m), we assign the δ = –10.7 ppm signal to CDB DADB is linked to an accelerated decomposition to boraz-
as well. Together with the observation described in the lit- ine and the accompanied hydrogen release.
erature[6] that the CTB-BH2 signal is slightly shifted to
Shortly after the rise of the DADB-related signal, a reso-
lower absolute ppm values in respect to the CDB-BH2 sig- nance at δ = –4.5 ppm occurs, the relative area of which
nal, we interpret the evolving peak at δ = –10.0 ppm as a also increases with increasing [AB]0 concentrations (see Fig-
result of the formation of CTB out of DADB and/or CDB. ure 6). We attribute this signal to the BNH3 fragment of the
Since the occurrence of the third-order decomposition recently identified BCDB.[6] The relative peak area of this
law could not be simply derived from the early reaction signal at δ = –4.5 ppm follows a two-step pattern. Here,
mechanism, an analysis of the formation of intermediates each step is followed by a short plateau (see Figure 6). The
by 11B NMR spectroscopy was also performed for the reac- levels of these plateaus rise with increasing [AB]0 concentra-
tion phase t Ͼ t1/4(H2).
tions. Prior to the onset of the second plateau, a sharp de-
52
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 49–54

Dehydrogenation of Ammonia–Borane/Triglyme Mixtures
cline of the signal’s relative area at δ ≈ –10.5 ppm is observ-
able, which we attribute to BH2 resonances and interpret as
a transition from CDB[6,7] to BCDB.
Figure 8. Coupled 11B{1H} NMR spectra[8] of samples of AB
(black line) and CTB (grey line) in glyme, which were dehydroge-
nated at 90 and 130 °C for 4 h, respectively. The spectrum of CTB
does not show any signal of BCDB or CDB, whereas highly con-
centrated AB/TG solutions react to form CDB, BCDB and signifi-
cant amounts of μ-ADB.
Figure 6. Analysis of the relative area of the BCDB-related peak
at δ = –4.5 ppm for different [AB]0 values. The BCDB formation
characteristics display two plateaus during the reaction. The rela-
tive peak areas rise with increasing [AB]0.
Hence, we interpret the noticed acceleration of the for-
mation of borazine from CDB through BCDB as being in-
duced or promoted by the increased presence of DADB.
Thus, in our measurements, higher [AB]0 values seem to
correlate with both higher [BCDB] and higher [DADB]
ones. An apparent correlation between increased [DADB]
and increased [BCDB] levels is supported by experiments
we conducted with separately prepared mixtures of AB and
DADB, which were prepared according to the procedure
described by Shore et al.[8] During the significantly en-
hanced dehydrogenation process, a rise of the relative peak
area at δ = –4.5 ppm (BCDB) was observed (see Figure 7).
11B NMR Spectroscopy of the Reaction Phase t Ͼ t1/2
For the reaction phase t Ͼ t1/2, the building of μ-amino-
diborane (μ-ADB, μ-[NH2BH5]) seems to be another vital
intermediate step to form borazine. The sextet signal at
δ = –26.5 ppm, which we assign to μ-ADB, shows up
shortly after t1/2(H2) and remains present until all hydrogen
is released (see Figure 4). In the separately prepared CTB
samples, μ-ADB is the only compound found beside boraz-
ine (see Figure 8).
The detected dehydrogenation of pure CTB/TG samples
is significantly slower than in CDB- or BCDB-rich AB/TG
samples with [AB]0 Ͼ 1.5 m. Thus, we assume two alterna-
tive pathways to borazine: an early, fast one that occurs
through BCDB and a late, slow one through μ-ADB. For
diluted solutions, however, with [AB]0 Ͻ 1.0 m both of these
pathways seem, at least partially, to be blocked at 70 °C
since neither DADB nor μ-ADB were detected.
In accordance with this interpretation, the measured hy-
drogen yield strongly deviates from the theoretical expected
value of 2 mol H2 per mol of AB. For example, for the
initial concentrations of [AB]0 = 0.8, 0.5 and 0.15 m, only
1.8, 1.5 and 0.3 mol H2 per mol of AB were released,
respectively.
Figure 7. Evolution analysis of the relative areas of the BCDB-
related 11B NMR spectroscopic peak at δ = –4.5 ppm for different
[AB]0 values during the dehydrogenation reaction of (a) 6.0 m and
(b) 1.0 m solutions of AB/TG and AB/DADB/TG mixtures. The
addition of DADB significantly promotes the formation of BCDB.
Conclusion
We have found a concentration-dependent overlap of at
least two distinct dehydrogenation events by gas kinetic
analysis in AB/TG solutions. 11B NMR spectroscopy dem-
On the basis of experiments with separately prepared and onstrates that the relative amounts both of DADB as well
heated CTB/TG samples, we ruled out the formation as of BCDB formed during thermal decomposition are in-
of BCDB from CTB, since we could not detect the creased in highly concentrated solutions. Moving from de-
δ = –4.5 ppm signal, even at elevated temperatures (see Fig- composition experiments with highly concentrated AB/TG
ure 8).
solutions to the ones enriched with DADB, even higher rel-
Eur. J. Inorg. Chem. 2012, 49–54
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
53

J. F. Kostka, R. Schellenberg, F. Baitalow, T. Smolinka, F. Mertens
FULL PAPER
2.75 g (6.0 m), were mixed in a two-necked Schlenk flask. For sus-
pensions ([AB]0 from 16 to 23 m) the amount of AB was fixed at
2.0 g. The flask was connected to a reflux condenser and a silicon
oil trap. The hydrogen release of the samples was measured volu-
metrically. About 15–20 samples were dynamically taken during
each dehydrogenation experiment. Samples were taken by using a
metal-free syringe with a polytetrafluoroethylene (PTFE) needle to
prevent catalytic effects. The samples were filled into NMR spec-
troscopy tubes and diluted with TG from the 1:2 to the 1:5 ratio
(according to their original concentration) to reduce signal broad-
ative amounts of BCDB are found. In addition, the pres-
ence of the latter compound appears to accelerate the for-
mation of borazine (i.e., its formation is observed earlier
and in higher amounts with respect to the relative amount
of hydrogen released). If we recall that the borazine forma-
tion is accompanied by hydrogen emission, we can conclude
that through the formation of BCDB, DADB is indirectly
responsible for the observed accelerated hydrogen release by
the earlier occurrence of the second desorption event. In
this context, the observed experimental third-order law of ening.
the decomposition kinetics appears to be a consequence of
Pressure Measurements: The above-described AB/TG mixtures
a shift in the downstream reaction pathways.
(1–23 m, 0.3–3.0 g) were isothermally decomposed in a Schlenk
flask at temperatures that ranged from 70 to 110 °C. The vessel
was connected to a reflux condenser and a 2 L gas tank equipped
with pressure and temperature sensors. The pressure and tempera-
ture of the gas tank were constantly monitored during the experi-
ments.
As a consequence of above deduced relationship between
the DADB and BCDB concentrations, the absence of de-
tectable DADB in samples with [AB]0 concentrations below
1.5 m indicates that the BCDB pathway for the conversion
of CDB to borazine becomes less important. This could
also explain the incomplete dehydrogenation in such sam-
ples at 70 °C.
The presence of DADB in highly concentrated solutions
in a weakly polar organic solvent gives reason to reconsider
the hypothesis that such an environment destabilizes the ion
11B NMR Spectroscopic Measurements: All samples were measured
with
a Bruker DPX 400 spectrometer (7.06 T) operating at
1
400 MHz for H and 128 MHz for 11B. 11B shifts were referenced
externally to BF3·Et2O (δ = 0 ppm).
Computation of Relative Peak Areas: Relative 11B NMR spectro-
pair [NH3BH2NH3]+[BH4]–.[4,6] Either this hypothesis is un- scopic peak areas were determined by fitting the corresponding
peaks with the software PeakFit, which determined the value of
their integrals and divided them by the ones of the complete spec-
trum (the latter representing the total boron content). The given
error bars (when stated) reflect the influence of the integration bor-
ders of the signals in the case of overlapping peaks.
tenable or strongly concentrated solutions of AB in glymes
lead to an increasingly polar environment that stabilizes
DADB.
From these investigations it can be stated that the decom-
position behaviour in highly concentrated solution, and
naturally also in suspensions, resembles the decomposition
mechanism in solid AB with respect to the importance of
DADB.
Acknowledgments
We thank the German Federal Ministry for the Environment, Na-
ture Conservation and Nuclear Safety (BMU) for partly funding
this work and the Bundesministerium für Bildung und Forschung
for financial support of J. K. and F. B.
These results have important implications for the devel-
opment of possible high-capacity hydrogen-generation de-
vices based on AB in organic solvents, since the dehydroge-
nation kinetics is one critical factor in the reactor design
and layout in all conceivable applications.
[1] J. Baumann, F. Baitalow, F. P. Hoffmann, Thermochim. Acta
2000, 343, 19.
[2] F. Baitalow, J. Baumann, G. Wolf, K. Jaenicke-Rossler, G. Le-
itner, Thermochim. Acta 2002, 391, 159.
Experimental Section
[3] A. C. Stowe, W. J. Shaw, J. C. Linehan, B. Schmid, T. Autrey,
Phys. Chem. Chem. Phys. 2007, 9, 1831.
[4] M. E. Bluhm, M. G. Bradley, R. Butterick III, U. Kusari, L. G.
Sneddon, J. Am. Chem. Soc. 2006, 128, 7748.
[5] R. Schellenberg, J. Kriehme, G. Wolf, Thermochim. Acta 2007,
457, 103.
[6] W. J. Shaw, J. C. Linehan, N. K. Szymczak, D. J. Heldebrant, C.
Yonker, D. M. Camaioni, R. T. Baker, T. Autrey, Angew. Chem.
2008, 120, 7603; Angew. Chem. Int. Ed. 2008, 47, 7493–7496.
[7] J. S. Wang, R. A. Geanagel, Inorg. Chim. Acta 1988, 148, 185.
[8] S. G. Shore, K. W. Böddeker, Inorg. Chem. 1964, 3, 914.
[9] D. F. Gaines, R. Schaeffer, J. Am. Chem. Soc. 1964, 86, 1505.
[10] G. H. Dahl, R. Schaeffer, J. Am. Chem. Soc. 1961, 83, 3032–
3034.
Materials: All experiments and measurements were performed un-
der argon 5.0. AB was purchased from SAFC and purified by
means of Soxhlet extraction with dry diethyl ether. Triglyme
(2,5,8,11-tetraoxadodecane) was obtained from Aldrich (purity
98%) and distilled from sodium at 90 °C and 0.2 bar. DADB was
synthesized according to literature methods[8] and characterized by
1
XRD, and 11B and H NMR spectroscopy. The obtained product
was Ͼ75% pure with AB as the only byproduct. CTB was synthe-
sized as described in the literature[10] and was further purified by
vacuum sublimation to result in a purity of 98% as demonstrated
1
by 11B and H NMR spectra.
NMR Spectroscopy Sample Preparation: In a typical NMR-spec-
troscopy-monitored dehydrogenation experiment, triglyme (12–
16 g) and an amount of AB, which varied from 0.07 g (0.1 m) to
Received: July 29, 2011
Published Online: November 16, 2011
54
www.eurjic.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2012, 49–54