Base-promoted ammonia borane hydrogen-release

Article abstract of DOI:10.1021/ja905015x

The strong non-nucleophilic base bis(dimethylamino)naphthalene (Proton Sponge, PS) has been found to promote the rate and extent of H ₂-release from ammonia borane (AB) either in the solid state or in ionic-liquid and tetraglyme solutions. For

Full text of DOI:10.1021/ja905015x

Published on Web 09/11/2009
Base-Promoted Ammonia Borane Hydrogen-Release
Daniel W. Himmelberger, Chang Won Yoon, Martin E. Bluhm, Patrick J. Carroll, and
Larry G. Sneddon*
Department of Chemistry, UniVersity of PennsylVania, Philadelphia, PennsylVania 19104-6323
Received June 18, 2009; E-mail: lsneddon@sas.upenn.edu
Abstract: The strong non-nucleophilic base bis(dimethylamino)naphthalene (Proton Sponge, PS) has been
found to promote the rate and extent of H2-release from ammonia borane (AB) either in the solid state or
in ionic-liquid and tetraglyme solutions. For example, AB reactions in 1-butyl-3-methylimidazolium chloride
(bmimCl) containing 5.3 mol % PS released 2 equiv of H2 in 171 min at 85 °C and only 9 min at 110 °C,
whereas comparable reactions without PS required 316 min at 85 °C and 20 min at 110 °C. Ionic-liquid
solvents proved more favorable than tetraglyme since they reduced the formation of undesirable products
such as borazine. Solid-state and solution 11B NMR studies of PS-promoted reactions in progress support
a reaction pathway involving initial AB deprotonation to form the H3BNH2- anion. This anion can then initiate
AB dehydropolymerization to form branched-chain polyaminoborane polymers. Subsequent chain-branching
and dehydrogenation reactions lead ultimately to a cross-linked polyborazylene-type product. AB dehy-
-
drogenation by lithium and potassium triethylborohydride was found to produce the stabilized Et3BNH2BH3
anion, with the crystallographically determined structure of the [Et3BNH2BH3]-K+ ·18-crown-6 complex
showing that, following AB nitrogen-deprotonation by the triethylborohydride, the Lewis-acidic triethylborane
group coordinated at the nitrogen. Model studies of the reactions of [Et3BNH2BH3]-Li+ with AB show evidence
of chain-growth, providing additional support for a PS-promoted AB anionic dehydropolymerization H2-
release process.
85 °C.4,5 For AB to be useful for hydrogen storage, a fast and
Introduction
controlled H2-release rate, as well as a viable AB regeneration
The development of a safe and efficient storage medium for
hydrogen is vital to its use as an alternative energy carrier. One
important approach to hydrogen storage has focused on the use
of high-capacity chemical hydrides, including borohydrides,
alanates, amineboranes, and metalloamidoboranes, as storage
materials.1,2 Understanding and controlling the many possible
pathways by which dehydrogenation and regeneration of these
materials can occur is key to their ultimate utilization and has
therefore been a common goal of research efforts in chemical
hydrogen storage.
method, must be achieved. A number of approaches have now
been shown to induce AB H2-release by a range of mechanistic
pathways, including activation by transition metal catalysts,6-21
(4) Stowe, A. C.; Shaw, W. J.; Linehan, J. C.; Schmid, B.; Autrey, T.
Phys. Chem. Chem. Phys. 2007, 9, 1831–1836.
(5) Bluhm, M. E.; Bradley, M. G.; Butterick, R., III.; Kusari, U.; Sneddon,
L. G. J. Am. Chem. Soc. 2006, 128, 7748–7749.
(6) Jaska, C. A.; Temple, K.; Lough, A. J.; Manners, I. Chem. Commun.
2001, 962–963.
(7) Jaska, C. A.; Temple, K.; Lough, A. J.; Manners, I. J. Am. Chem.
Soc. 2003, 125, 9424–9434.
Owing to its high hydrogen content, ammonia borane (AB)
has been identified as one of the leading candidates for chemical
hydrogen storage that could potentially release 19.6 wt % H2
according to eq 1.3
(8) Jaska, C. A.; Manners, I. J. Am. Chem. Soc. 2004, 126, 2698–2699.
(9) Clark, T. J.; Lee, K.; Manners, I. Chem.sEur. J. 2006, 12, 8634–
8648.
(10) Clark, T. J.; Russell, C. A.; Manners, I. J. Am. Chem. Soc. 2006, 128,
9582–9583.
H3NBH3 f BN + 3H2
(1)
(11) Denney, M. C.; Pons, V.; Hebden, T. J.; Heinekey, D. M.; Goldberg,
K. I. J. Am. Chem. Soc. 2006, 128, 12048–12049.
(12) Fulton, J. L.; Linehan, J. C.; Autrey, T.; Balasubramanian, M.; Chen,
Y.; Szymczak, N. K. J. Am. Chem. Soc. 2007, 129, 11936–11949.
(13) Jiang, Y.; Berke, H. Chem. Commun. 2007, 3571–3573.
(14) Keaton, R. J.; Blacquiere, J. M.; Baker, R. T. J. Am. Chem. Soc. 2007,
129, 1844–1845.
Utilization of waste heat from a proton exchange membrane
fuel cell can provide for AB H2-release reaction temperatures
near 85 °C. However, at 85 °C in the absence of any additive,
H2-release from solid-state AB has been shown to exhibit an
induction period of up to 3 h. After hydrogen release begins,
only the release of ∼1 equiv of H2 can be achieved, rather than
the 3 equiv predicted by eq 1, even with prolonged heating at
(15) Paul, A.; Musgrave, C. B. Angew. Chem., Int. Ed. 2007, 46, 8153–
8156.
(16) Pun, D.; Lobkovsky, E.; Chirik, P. J. Chem. Commun. 2007, 3297–
3299.
(17) Blacquiere, N.; Diallo-Garcia, S.; Gorelsky, S. I.; Black, D. A.; Fagnou,
K. J. Am. Chem. Soc. 2008, 130, 14034–14035.
(18) Douglas, T. M.; Chaplin, A. B.; Weller, A. S. J. Am. Chem. Soc. 2008,
130, 14432–14433.
(1) Graetz, J. Chem. Soc. ReV. 2009, 38, 73–82.
(2) Hamilton, C. W.; Baker, R. T.; Staubitz, A.; Manners, I. Chem. Soc.
ReV. 2009, 38, 279–293.
(19) Staubitz, A.; Soto, A. P.; Manners, I. Angew. Chem., Int. Ed. 2008,
47, 6212–6215.
(3) Stephens, F. H.; Pons, V.; Baker, R. T. Dalton Trans. 2007, 25, 2613–
2626.
(20) Yang, X.; Hall, M. B. J. Am. Chem. Soc. 2008, 130, 1798–1799.
9
10.1021/ja905015x CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 14101–14110 14101

A R T I C L E S
Himmelberger et al.
acid catalysts,22 nano- and mesoporous scaffolds,23-25 and ionic
liquids.5 There has likewise been significant recent progress in
the development of efficient AB regeneration processes.26
We have previously shown27,28 that the addition of small
amounts of either LiH or LiNH2 to AB in the solid state
eliminated the induction period and increased both the rate and
extent of H2-release at 85 °C. As outlined in eqs 2 and 3, the
initial step in these reactions was proposed to be AB deproto-
dimethylimidazolium chloride (bdmimCl), and 1-ethyl-2,3-dime-
thylimidazolium ethylsulfate (edmimEtSO4) ionic liquids (Fluka)
were dried by toluene azeotropic distillation. Tetraethylene glycol
dimethyl ether (tetraglyme, Sigma, 99%) and ethylene glycol
dimethyl ether (glyme, Sigma, 99%) were vacuum distilled from
sodium with heating. Bis(dimethylamino)naphthalene (Proton Sponge,
PS, Aldrich) was sublimed and stored under inert atmosphere and
light-free conditions. Lithium triethylborohydride (1.0 M solution
in THF) and potassium triethylborohydride (1.0 M solution in THF)
(Aldrich) were used as received.
Physical Measurements. The Toepler pump system used for
hydrogen measurements was similar to that described by Shriver31
and is diagrammed in the Supporting Information (Figure S1). The
gases released from the reaction vessel were first passed through a
liquid nitrogen trap before continuing on to the Toepler pump (700
mL). The released H2 was then pumped into a series of calibrated
volumes, with the final pressure of the collected H2 gas measured
((0.5 mm) with the aid of a U-tube manometer. After the H2-
measurement was completed, the in-line liquid nitrogen trap was
warmed to room temperature, and the amount of any volatiles that
had been trapped was then also measured using the Toepler pump.
The automated gas buret was based on the design reported by
Zheng et al.32 but employed all glass connections, with a cold trap
(-78 °C) inserted between the reaction flask and buret to allow
trapping of any volatiles that might have been produced during the
reaction.
-
nation to produce the H3BNH2 anion, with this anion then
inducing anionic dehydropolymerization of AB to produce a
growing polyaminoborane polymer (eq 4).
-
H3BNH3 + Li+NH2 f H3BNH2-Li+ + NH3
(2)
(3)
H3BNH3 + Li+H- f H3BNH2-Li+ + H2
H3BNH2-Li+ + H3BNH3 f H3BNH2BH2NH2-Li+ + H2
(4)
Unfortunately, these reactions stopped after the release of
∼1.5 equiv of H2 due at least in part to the formation of the
LiBH4 side product via the reactions in eqs 5 and 6. Since LiBH4
does not decompose until >350 °C, its formation reduces the
extent of AB H2-release at 85 °C.29
Differential scanning calorimetry (DSC) was carried out on a
Setaram C80 calorimeter. Samples containing 50 mg of AB and
50 mg of bmimCl, without and with 5 mol % (18 mg) of PS, were
loaded into the cells under a N2 atmosphere. The ramp rate was 1
°C/min, and samples were taken to either 85 or 110 °C.
H3BNH2-Li+ f (H2BNH2) + Li+H-
(5)
(6)
-
H3BNH3 + Li+H- f Li+BH4 + NH3
While bmimCl is a liquid at 85 °C, it is a solid at room
temperature; therefore, solid-state 11B NMR analyses (at Pacific
Northwest National Laboratories, 240 MHz machine spun at 10
kHz) were used to monitor the products of reactions carried out in
bmimCl. All solid-state 11B chemical shifts were measured relative
to external NaBH4 (-41 ppm). The solution 11B NMR (128.4 MHz
Bruker DMX-400 instrument) studies in the room-temperature ionic
liquid mmimMeSO4 were carried out by heating the reaction
mixtures in sealed NMR tubes at 85 °C for the indicated times,
with the spectra taken at 25 °C. The 11B NMR spectra of the
reactions in tetraglyme were collected with the NMR heated at 80
°C. All solid-state and solution 11B NMR chemical shifts are
referenced to external BF3 ·O(C2H5)2 (0.0 ppm), with a negative
sign indicating an upfield shift.
Procedures for AB H2-Release Reactions. For the experiments
where the released H2 was measured with the Toepler pump, the
AB (250 mg, 8.1 mmol) was loaded into the reaction flasks under
N2. The solid-state reactions of AB/PS mixtures were carried out
in evacuated 500 mL break-seal flasks that were heated in an oven
preheated to the desired temperature. The solids were initially only
crudely mixed, since upon heating the solid mixtures were found
to form a melt before the onset of H2-release. Reactions in solution
were loaded into ∼100 mL flasks with the ionic liquid and PS in
the amounts given in the tables. The flasks were then evacuated,
sealed, and placed in a hot oil bath preheated to the desired
temperature. The flasks were opened at the indicated times, and
the released hydrogen was quantified using the Toepler pump
system. After reaction, the flasks were evacuated for 30 min through
the cold trap to remove any volatile products from the reaction
residue. The product residues and volatiles in the cold trap were
extracted with dry glyme and analyzed by 11B NMR.
In order to avoid the formation of stable alkali-metal
borohydrides, we investigated the use of alternative nitrogen-
based deprotonating agents to induce AB polymerization. We
report here that the strong (pKa ∼12), non-nucleophilic base
bis(dimethyamino)naphthalene (Proton Sponge, PS)30 can also
induce AB H2-release via an anionic dehydropolymerization
mechanism with the advantage that the formation of a stable
-
BH4 salt is avoided.
Experimental Section
Materials. All manipulations were carried out using standard
high-vacuum or inert atmosphere techniques, as described by
Shriver.31 Ammonia borane (AB, Aviabor, 97% minimum purity)
was ground into a free-flowing powder using a commercial coffee
grinder. The 1-butyl-3-methylimidazolium chloride (bmimCl), 1,3-
dimethylimidazolium methylsulfate (mmimMeSO4), 1-butyl-2,3-
(21) Forster, T. D.; Tuononen, H. M.; Parvez, M.; Roesler, R. J. Am. Chem.
Soc. 2009, 131, 6689–6691.
(22) Stephens, F. H.; Baker, R. T.; Matus, M. H.; Grant, D. J.; Dixon,
D. A. Angew. Chem., Int. Ed. 2007, 46, 746–749.
(23) 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. J.; Gutowski,
M.; Autrey, T. Angew. Chem., Int. Ed. 2005, 44, 3578–3582.
(24) Paolone, A.; Palumbo, O.; Rispoli, P.; Cantelli, R.; Autrey, T.;
Karkamkar, A. J. Phys. Chem. C 2009, 113, 10319–10321.
(25) Sepehri, S.; Feaver, A.; Shaw, W. J.; Howard, C. J.; Zhang, Q.; Autrey,
T.; Cao, G. J. Phys. Chem. B 2007, 111, 14285–14289.
(26) Davis, B. L.; Dixon, D. A.; Garner, E. B.; Gordon, J. C.; Matus, M. H.;
Scott, B.; Stephens, F. H. Angew. Chem., Int. Ed. 2009, 48, 6812–
6816.
(27) Sneddon, L. G. DOE Hydrogen Program Review, 2007; http://
www.hydrogen.energy.gov/pdfs/review07/st_27_sneddon.pdf.
(28) Bluhm, M. E.; Bradley, M. G.; Sneddon, L. G. Prepr. Symp.sAm.
Chem. Soc., DiV. Fuel Chem. 2006, 51, 571–572.
For reactions using the automated gas buret, the AB (150 mg,
4.87 mmol) samples were loaded into ∼100 mL flasks with
(29) Fang, Z. Z.; Wang, P.; Rufford, T. E.; Kang, X. D.; Lu, G. Q.; Cheng,
H. M. Acta Mater. 2008, 56, 6257–6263.
(32) Zheng, F.; Rassat, S. D.; Helderandt, D. J.; Caldwell, D. D.; Aardahl,
C. L.; Autrey, T.; Linehan, J. C.; Rappe, K. G. ReV. Sci. Instrum.
2008, 79, 084103-1–084103-5.
(33) (a) CrystalClear; Rigaku Corporation, 1999. (b) Crystal Structure
Analysis Package; Rigaku Corp. Rigaku/MSC, 2002.
(30) Alder, R. W. Chem. ReV. 1989, 89, 1215–1223.
(31) Shriver, D. F.; Drezdzon, M. A. Manipulation of Air SensitiVe
Compounds, 2nd ed.; Wiley: New York, 1986.
9
14102 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Base-Promoted Ammonia Borane H2-Release
A R T I C L E S
Table 1. Summary of Structure Determination of [Et3BNH2BH3]-
K·18-crown-6
calibrated volumes, along with the ionic-liquid (150 mg) or
tetraglyme (0.15 mL) solvents and PS. Under a flow of helium,
the flask was attached to the buret system. The system was
evacuated 30 min for reactions with the ionic-liquid solutions and
5 min for tetraglyme solutions. The system was then backfilled with
helium and allowed to equilibrate to atmospheric pressure for ∼30
min. Once the system pressure equalized, the data collection
program was started, and the flask was immersed in the preheated
oil bath. The data are reported from the point where the flask was
initially plunged into the oil bath, but H2-release was not observed
until the ionic-liquid/AB mixture melted. Data were recorded at
2-5 s intervals, depending on the speed of the reaction. (Control
experiments conducted without AB showed negligible pressure
change, corresponding to <0.01 gas equivalent, when a helium-
filled reaction flask connected to the gas buret was heated to 120
°C.) The product residues were extracted with dry glyme and
analyzed by 11B NMR.
Reactions of bmimCl and bmimCl/PS with partially dehydro-
genated AB followed the procedures for the automated gas buret.
Initially, two separate samples of neat AB (150 mg, 4.87 mmol)
were heated for 23 h at 85 °C to release ∼1 equiv of H2. The
reaction flasks were removed from the gas buret system under a
flow of helium and then taken into a glovebox, where bmimCl (150
mg, 50 wt %) was added to one sample and bmimCl (150 mg, 50
wt %) and PS (55 mg, 5 mol %) were added to the second. After
thorough mixing, the flasks were reattached to the gas buret system
and heated again at 85 °C. Data were recorded on the gas buret
system until H2-release stopped.
formula
C18B2H44NO6K
431.26
orthorhombic
Pbca (no. 61)
8
formula weight
crystal class
space group
Z
a, Å
b, Å
11.5950(7)
14.7764(9)
29.239(2)
5009.5(5)
2.42
c, Å
V, Å3
µ, cm-1
crystal size, mm
Dcalc, g/cm3
F(000)
0.46 × 0.30 × 0.25
1.144
1888
radiation
2θ range, deg
hkl collected
Mo KR (λ ) 0.71073 Å)
5.26-54.96
-12 e h e 15; -19 e k e 14;
-31 e l e 37
143
temperature, K
no. reflections measured
no. unique reflections
no. observed reflections
no. reflections used in
refinement
20472
5721 (Rint ) 0.0233)
4944 (F > 4σ)
5721
no. parameters
430
R indices (F > 4σ)a
R1 ) 0.0374
wR2 ) 0.0892
R1 ) 0.0459
wR2 ) 0.0943
1.031
R indices (all data)a
GOFb
Syntheses of [Et3BNH2BH3]-M+ (M ) Li and K). AB (50
mg, 1.6 mmol) was dissolved in THF or glyme (5.0 mL) and the
solution cooled at -70 °C. Lithium triethylborohydride (1.0 M
solution in THF, 1.6 mL) or potassium triethylborohydride (1.0 M
solution in THF, 1.6 mL) was then added dropwise. After the
solutions were warmed to room temperature and stirred for 1 h,
their 11B NMR spectra indicated the clean formation of
[Et3BNH2BH3]-M+ (M ) Li or K). 11B NMR (128.4 MHz, glyme,
ppm): [Et3BNH2BH3]-Li+ -7.5 (s, BEt3) and -23.8 (q, BH3);
[Et3BNH2BH3]-K+ -7.8 (s, Et3B) and -23.6 (q, BH3). DFT/GIAO
calcd 11B NMR chemical shift values for Et3BNH2BH3-: -10.6
(Et3B) and -25.3 (BH3). An electrospray mass spectrum of the
[Et3BNH2BH3]-Li+ solution confirmed the formation of the
final difference peaks, e/Å3
+0.240, -0.203
a R1 ) ∑|Fo| - |Fc|/∑|Fo|; wR2 ) {∑w(Fo - Fc2)2/∑w(Fo )2}1/2
.
2
2
b
GOF ) {∑w(Fo - Fc2)2/(n - p)}1/2, where n is the number of
reflections and p is the number of parameters refined.
2
Crystallographic Data for [Et3BNH2BH3]-K+ ·18-Crown-6.
Single crystals of the [Et3BNH2BH3]-K+ ·18-crown-6 (C12H24O6)
adduct (Penn 3310) were grown from 1:1 mixtures of
[Et3BNH2BH3]-K+ and 18-crown-6 in THF/hexanes at room
temperature.
Crystallographic data and structure refinement information are
summarized in Table 1. X-ray intensity data were collected on a
Rigaku Mercury CCD area detector employing graphite-monochro-
mated Mo KR radiation (λ ) 0.71073 Å). Preliminary indexing
was performed from a series of twelve 0.5° rotation images with
exposures of 30 s. Rotation images were processed using
CrystalClear,33a producing a listing of unaveraged F2 and σ(F2)
values which were then passed to the CrystalStructure33b program
package for further processing and structure solution on a Dell
Pentium III computer. The intensity data were corrected for Lorentz
and polarization effects and for absorption.
Et3BNH2BH3 anion (calcd m/e for 12C6 H2011B214N 128, found
128).
-
1
Reactions of [Et3BNH2BH3]-Li+ with AB. AB (50 mg, 1.6
mmol) was added to a glyme solution of [Et3BNH2BH3]-Li+ (1.6
mmol, 5.0 mL) and then stirred at room temperature. After 18 h,
the 11B NMR spectrum (128.4 MHz, glyme, ppm) showed, in
addition to the resonances from unreacted AB (-22.4) and
Et3BNH2BH3- (-7.5 and -23.8), new resonances at -6.0 (s, Et3B),
-10.9 (t, BH2), and -19.6 (q, BH3) that have chemical shift values
consistent with the DFT/GIAO calculated values for the
-
Et3BNH2BH2NH2BH3 anion: -8.2 (Et3B), -12.0 (BH2), and
The structure was solved by direct methods (SIR97).34 Refine-
ments were made by full-matrix least-squares based on F2 using
SHELXL-97.35 All reflections were used during refinement (F2
values that were experimentally negative were replaced by F2 )
0). Non-hydrogen atoms were refined anisotropically, and hydrogen
atoms were refined isotropically.
-23.5 (BH3). After 72 h, a Toepler pump measurement showed
that 0.6 equiv of H2 (relative to AB) had been released, and
electrospray mass spectrometry confirmed the presence of both
Et3BNH2BH3- (m/e 128) and Et3BNH2BH2NH2BH3- (calcd m/e for
12C6 H2411B314N2 157, found 157).
1
Computational Methods. DFT/GIAO/NMR calculations were
performed using the Gaussian 03 program.36 Geometries were fully
optimized at the B3LYP/6-31G(d) level without symmetry con-
straints. The 11B NMR chemical shifts were calculated at the
B3LYP/6-311G(d) level using the GIAO option within Gaussian
03. The 11B NMR GIAO chemical shifts are referenced to BF3 ·OEt2
using an absolute shielding constant of 101.58, which was obtained
The reaction rate increased when AB (0.1 g, 3.2 mmol) was
reacted with a glyme solution (5.0 mL) of [Et3BNH2BH3]-Li+ at
70 °C, and after only 2 h, electrospray mass spectrometry showed,
-
in addition to Et3BNH2BH3- (m/e 128) and Et3BNH2BH2NH2BH3
(m/e 157), the formation of the higher oligomers
Et3B(NH2BH2)2NH2BH3 (calcd m/e for 12C6 H2811B414N3 186,
-
1
-
found 186) and Et3B(NH2BH2)3NH2BH3 (calcd m/e for
12C6 H3211B514N4 215, found 215).
1
(35) Sheldrick, G. M. SHELXL-97: Program for the Refinement of Crystal
Structures; University of Gottingen: Gottingen, Germany, 1997.
(36) Frisch, M. J.; et al. Gaussian 03, Revision B.05; Gaussian, Inc.:
Pittsburgh, PA, 2003.
(34) Altomare, A.; Burla, M.; Camalli, M.; Cascarano, G.; Giacovazzo,
C.; Guagliardi, A.; Moliterni, A.; Polidori, G.; Spagna, R. J. Appl.
Crystallogr. 1999, 32, 115–119.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14103

A R T I C L E S
Himmelberger et al.
Table 3. H2-Release Data (Gas Buret) for AB/PS Solid-State
Reactions at 85 °C
time to H2-equivalents (min)
amount of PS
(mol %, mg, mmol)
total wt (mg)a
0.5 equiv
1 equiv
final (equiv)
150
205
295
191
1264
885
1333 (1.02)
1334 (1.08)
5.3, 55, 0.26
a 150 mg of NH3BH3 (4.87 mmol) was used for all reactions.
1
Figure 2. 11B{1H} NMR (128.4 MHz) spectra (insets show H coupled
spectra) recorded at 25 °C of the glyme extract of the reaction of (a) AB
(150 mg) and 5.3 mol % PS (55 mg) at 85 °C for 22 h and (b) AB (150
mg) at 85 °C for 22 h. The broad DADB resonance at -13 ppm is obscured
by the PAB resonance.
Figure 1. H2-release measurements.(a) Toepler pump measurement for
solid-state AB (250 mg) reactions with 0, 1.0, and 5.2 mol % PS (18 and
91 mg) at 85 °C. (b) Gas buret measurement for solid-state AB (150 mg)
reactions with 0 and 5.2 mol % PS (55 mg) at 85 °C.
just after the release of ∼1.1 equiv of H2. More detailed H2-
release data collected (Table 3 and Figure 1b) for the 5.2 mol
% PS reaction with the automated gas buret again clearly
demonstrated that the induction period was shortened and the rate
of H2-release was increased upon the addition of 5.2 mol % PS.
The NMR spectra in Figure 2 of the glyme-soluble residues
from the solid-state AB/PS reactions showed features similar
to those of the pure AB reactions. Thus, as the reaction
progressed, the AB resonance (-23.5 ppm) decreased and was
replaced by resonances arising from both the diammoniate of
Table 2. H2-Release Data (Toepler Pump) for AB/PS Solid-State
Reactions at 85 °C
H2 released
amount of PS
(mol %, mg, mmol)
time (h)
total wt (mg)a
equiv
mmol
1
3
21
250
250
250
0
0.34
0.84
0
2.74
6.79
-
diborane, [(NH3)2BH2]+BH4 , (DADB; -13.3 (overlapped) and
-37.6 ppm)37,38 and branched-chain polyaminoborane polymers
(PAB; -7, -13.3, and -25.1 ppm).5 Consistent with the faster
H2-release found for the PS/AB mixtures, the spectra show that
after 22 h the amount of unreacted AB was less in the PS/AB
reactions than in the pure AB reaction. The solid-state 11B NMR
spectra in Figure 3 of the final products of the reactions showed
an additional broad resonance centered near ∼23 ppm that is
characteristic of the sp2 boron framework of cross-linked
polyborazylene structures,39-41 thus indicating that AB dehy-
drogenation ultimately produces BdN unsaturated products.
Consistent with this conclusion, the 11B NMR spectra of the
volatile products of the PS/AB reactions collected in the cold
trap through which the released H2 had been passed also
indicated some borazine (B3N3H6) formation, with the amount
of the volatiles (ranging from 0.04 to 0.25 mmol, depending
upon the particular experiment) corresponding to <10% of the
AB converting to borazine.
1
3
21
1.0, 18, 0.084
1.0, 18, 0.084
1.0, 18, 0.084
268
268
268
0
0.65
0.95
0
5.23
7.73
1
3
21
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
341
341
341
0.35
0.92
1.12
2.86
7.52
9.04
a 250 mg of NH3BH3 (8.1 mmol) was used for all reactions.
from the GIAO NMR calculated shift of BF3 ·OEt2 at the B3LYP/
6-311G(d)//B3LYP/6-31G(d) level of theory.
Results and Discussion
H2-Release from AB/PS Solid-State Reactions. As shown in
Figure 1a and Table 2, initial H2-release measurements using
the Toepler pump of the solid-state reactions of AB in the
presence of 1.0 and 5.2 mol % PS at 85 °C clearly demonstrated
the activating effect of PS on AB H2-release. While the reaction
with 1.0 mol % PS still showed an induction period with no H2
released after 1 h, the extent of release was significantly
increased at 3 h (0.65 equiv) compared to that from pure AB at
the same time (0.34 equiv). The data for the 5.2 mol % PS
reaction indicated both a shortened induction period, with 0.35
equiv already released at 1 h, and significantly increased
amounts of H2-release at both 3 h (0.93 equiv) and 20 h (1.12
equiv) compared to those of either the pure AB or 1.0 mol %
PS reactions. However, the solid-state AB/PS reactions stopped
H2-Release from AB/PS Solution Reactions. We have previ-
ously shown that polar solvents, especially ionic liquids such
(37) Onak, T. P.; Shapiro, I. J. Chem. Phys. 1960, 32, 952.
(38) Shore, S. G.; Parry, R. W. J. Am. Chem. Soc. 1958, 80, 20–24, and
preceeding papers in the issue.
(39) Fazen, P. J.; Beck, J. S.; Lynch, A. T.; Remsen, E. E.; Sneddon, L. G.
Chem. Mater. 1990, 2, 96–97.
(40) Fazen, P. J.; Remsen, E. E.; Beck, J. S.; Carroll, P. J.; McGhie, A. R.;
Sneddon, L. G. Chem. Mater. 1995, 7, 1942–1956.
(41) Gervais, C.; Framery, E.; Duriez, C.; Maquet, J.; Vaultier, M.;
Babonneau, F. J. Eur. Ceram. Soc. 2005, 25, 129–135.
9
14104 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Base-Promoted Ammonia Borane H2-Release
A R T I C L E S
Figure 3. Solid-state 11B NMR (240 MHz) spectra recorded at 25 °C of
the reaction of (a) AB (150 mg) and 5.3 mol % PS (55 mg) at 85 °C until
1 equiv of H2 was released and (b) AB (150 mg) at 85 °C until 1 equiv of
H2 was released. The broad DADB resonance at -13 ppm is obscured by
the AB resonance. Borate resonances at 17 and 2 ppm result from exposure
of the sample to air.
Figure 4. H2-release measurements (Toepler pump) of the reaction of AB
(250 mg) in bmimCl (250 mg) with 0, 1.0, and 5.2 mol % PS (18 and 91
mg) at 85 °C.
Table 4. H2-Release Data (Toepler Pump) for AB/bmimCl/PS
Reactions at 85 °C
H2 released
amount of PS
(mol %, mg, mmol)
time (h)
totala wt (mg)
equiv
mmol
1
3
6
23
500
500
500
500
0.69
1.00
1.28
1.97
5.62
8.10
10.36
15.94
1
3
6
22
0.5, 9, 0.04
0.5, 9, 0.04
0.5, 9, 0.04
0.5, 9, 0.04
509
509
509
509
0.83
1.65
2.10
2.23
6.70
13.37
17.04
18.04
1
3
6
21
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
591
591
591
591
1.40
2.10
2.18
2.23
11.33
17.02
17.67
18.04
1
3
6
22
24.9, 434, 2.03
24.9, 434, 2.03
24.9, 434, 2.03
24.9, 434, 2.03
934
934
934
934
0.79
1.75
2.01
2.13
6.37
14.14
16.28
17.23
a 250 mg of NH3BH3 (8.1 mmol) and 250 mg of bmimCl were used
for all reactions.
as 1-butyl-3-methylimidazolium chloride (bmimCl), can activate
H2-release from AB at 85 °C,5 with over 2 equiv being produced
in ∼5 h from a 50/50 wt % AB/bmimCl mixture. Comparisons
of the H2-release rates using Toepler pump measurements of
AB dissolved in bmimCl with different amounts of added PS
at 85 °C are summarized in Table 4 and Figure 4. In the absence
of PS, only 0.69, 1.00, 1.28, and 1.97 equiv of H2-equiv were
released at 1, 3, 6, and 23 h, respectively. In contrast, a similar
reaction of AB in bmimCl containing 5.2 mol % PS showed a
significantly increased H2-release rate, with 1.40 equiv of H2-
equiv released by 1 h and 2.10 equiv of H2 released after only
3 h. Even when the amount of PS was decreased to only 0.5
mol % PS, there were still significant increases in the H2-release
found at 1 h (0.83 equiv), 3 h (1.65 equiv), and 6 h (2.10 equiv)
compared to those of the reaction without PS. The plots in
Figure 4 show that the biggest differences in the H2-release rates
of the bmimCl reactions with and without PS occurred following
the release of the first equivalent of H2. For the reaction without
PS, the first equivalent was released in 3 h, and even after 23 h
only 1.97 equiv was released. On the other hand, by 3 h the
reactions with 5.2 and 1.0 mol % PS had already released 2.10
Figure 5. Differential scanning calorimetry analyses of the reactions of
AB (150 mg) in bmimCl (150 mg) with 5.3 mol % PS (55 mg) at (a) 85
and (b) 110 °C. The initial exotherm is an apparatus artifact.
and 1.75 equiv, respectively, and at 6 h, 2.18 and 2.01 equiv.
Thus, PS appears to have significantly enhanced the release rate
of the second H2 equivalent from AB. DSC measurements also
support this conclusion. The black curve in Figure 5a shows a
single, well-defined exotherm for the release of the first H2
equivalent from the reaction of an AB/bmimCl mixture. The
blue curve, which is from the reaction of a similar mixture
containing 5.3 mol % of added PS, clearly exhibits a second
exotherm, indicating the release of additional H2 by a different
process. Consistent with the H2-release measurements that
showed significant rate enhancements for the second H2
equivalent as the temperature was raised, when the upper
temperature of the DSC analysis was raised from 85 to 110 °C
(Figure 5b), the second exotherm grew and shifted to shorter
times such that it overlapped the first isotherm.
In order to better quantify the effects of both PS-loading and
temperature, more detailed H2-release data on a series of 50/50
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14105

A R T I C L E S
Himmelberger et al.
Table 5. H2-Release Data (Gas Buret) for AB/bmimCl/PS
Reactions
Table 6. H2-Release (Toepler Pump) Data for Partially
Dehydrogenated AB
time to H2 equivalents (min)
H2 released
amount of PS
total
temp (°C) (mol %, mg, mmol) wt (mg)a 0.75 equiv 1 equiv 1.5 equiv final (equiv)
time (h)
total wt (mg)a
equiv
mmol
75
300
228
205
161
140
52
50
45
35
21
20
17
16.3
5.8
4.9
3.8
4.0
459
320
215
208
84
74
61
41
37
27
21
20.9
7.4
5.9
4.3
4.3
1041 (1.43)
665 1040 (2)
0
250
500
500
500
500
500
0.99
1.05
1.11
1.27
1.78
1.79
8.04
8.51
9.02
10.30
14.42
14.51
1b
0.96, 10, 0.047 310
2b
5.3, 55, 0.26
355
342 723 (1.99)
378 782 (1.98)
199 316 (2)
118 179 (2)
95 171 (2)
55 77 (2)
87 126 (2)
44 72 (2)
31 52 (2)
3b
24.9, 260, 1.21 560
300
0.96, 10, 0.047 310
21b
24b
85
5.3, 55, 0.26
355
24.9, 260, 1.21 560
300
0.96, 10, 0.047 310
0
250
591
591
591
591
0.98
1.21
1.45
1.51
1.70
7.96
9.82
11.75
12.20
13.77
1c
2c
3c
24c
95
5.3, 55, 0.26
355
24.9, 260, 1.21 560
300
0.96, 10, 0.047 310
32 59 (1.91)
12 20 (2)
110
a 250 mg of NH3BH3 (8.1 mmol) for all experiments. b 250 mg of
bmimCl added after neat NH3BH3 reaction. c 250 mg of bmimCl and 91
mg of PS added after neat NH3BH3 reaction.
8
6
5
13 (2)
9 (2)
7 (2)
5.3, 55, 0.26
355
24.9, 260, 1.21 560
a 150 mg of NH3BH3 (4.87 mmol) and 150 mg of bmimCl were used
for all reactions.
Figure 7. H2-release measurements (gas buret) of partially dehydrogenated
AB (150 mg) where 1 equiv of H2 was initially released at 85 °C, and then
bmimCl (150 mg) and bmimCl (150 mg)/PS (55 mg/5.2 mol %) were added
to separate samples and heating resumed at 85 °C. the The dashed line
shows the observed AB H2-release when IL and/or PS are not added.
4.3, second 9 min), and 24.9 (first 4.3, second 7 min) mol %
PS were all again significantly faster. It is also noteworthy that
at 95 and 110 °C, unlike at 85 °C, the rate of H2-release for the
5.3 and 24.9 mol % PS reactions were similar, indicating that
less PS is required to induce H2-release at higher temperatures.
The above studies all indicate that both bmimCl and PS
promote the loss of more than 1 equiv of H2, with the
combination of PS in bmimCl being the most effective. To
further test this conclusion, two samples of neat AB were first
heated at 85 °C for 23 h to produce a partially dehydrogenated
material where only ∼1.0 equiv of H2 had been released. To
the first sample was added bmimCl, and to the second sample
were added both PS and bmimCl. The flasks were then reheated
at 85 °C to produce liquid suspensions, and any additional H2-
release was measured. While the heating of AB in the solid
state at 85 °C for more extended periods (∼48 h) gave no further
H2-release, both Toepler-pump (Table 6) and gas-buret mea-
surements (Figure 7) showed that the addition of either bmimCl
or bmimCl/PS to the partially dehydrogenated AB caused H2-
release to resume, ultimately yielding an additional ∼0.7 equiv
of H2 from both samples, with the bmimCl/PS reaction
exhibiting the fastest rate.
Figure 6. H2-release measurements (gas buret) of the reaction of AB (150
mg) in bmimCl (150 mg) with 0, 0.96, 5.3, and 24.9 mol % PS (10, 55,
and 260 mg) at (a) 85 and (b) 110 °C.
wt % AB/bmimCl systems containing from 0 to 24.9 mol %
PS at 75, 85, 95, and 110 °C were collected on the automated
gas buret (Table 5 and Figure 6). The rate of H2-release
increased as the mol % of PS in the mixture was increased. For
example, as shown in Figure 6a, while the reaction at 85 °C
without PS required 84 and 316 min to liberate the first and
second equivalents of H2, the reactions with 0.96 (first 74,
second 179 min), 5.3 (first 61, second 171 min), and 24.9 (first
41, second 77 min) mol % PS were all significantly faster, with
the fastest rate found for the highest loading. Substantial rate
increases were observed for all reactions when the reaction
temperature was increased. Thus, as illustrated in the plots in
Figure 6b, at 110 °C the reaction without PS required only 7.4
and 20 min to liberate the first and second equivalents of H2,
but the reactions with 0.96 (first 5.9, second 13 min), 5.3 (first
While bmimCl is a liquid at 85 °C, it is a solid at room
temperature; therefore, solid-state NMR was used to monitor
the bmimCl/AB/PS reactions at different stages by heating the
liquid mixtures at 85 °C and then quenching the reactions at
the indicated times by cooling to room temperature, with
subsequent analyses of the resulting solid materials by solid-
9
14106 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Base-Promoted Ammonia Borane H2-Release
A R T I C L E S
Figure 9. Solid-state 11B NMR (240 MHz) spectrum recorded at 25 °C of
the reaction of AB (250 mg) and 5.2 mol % PS (91 mg) in bmimCl (150
mg) at 85 °C for 23 h.
Figure 8. Solid-state 11B NMR (240 MHz) spectra recorded at 25 °C of
the reaction of AB (150 mg) and 5.3 mol % PS (55 mg) in bmimCl (150
mg) at 85 °C after the H2-release of (a) 0.5 equiv (30 min), (b) 1 equiv (61
min), (c) 1.5 equiv (95 min), and (d) 2 equiv (171 min). The broad DADB
resonance at -13 ppm is obscured by the AB and PAB resonances.
state 11B NMR. Consistent with both the H2-release measure-
ments and the 11B NMR studies of the solid-state PS/AB
reactions, the solid-state 11B spectrum (Figure 8) of the reaction
of a 50/50 wt % bmimCl/AB mixture containing 5.2 mol % of
added PS showed after 0.5 equiv of H2-release the resonances
of both DADB37,38 and branched-chain PAB.5 The small signal
(-21.2 ppm) that is apparent just downfield of the AB (-23.2
ppm) resonance has a chemical shift value consistent with that
previously reported42-45 for the H3BNH2- anion (-21.49 ppm)
and with its DFT/GIAO calculated value (-20.6 ppm) (Figure 2S,
Supporting Information). As the reaction continued, the resonances
broadened and diminished and a new resonance centered at 25.9
ppm grew in that is characteristic of unsaturated sp2-hybridized
boron. As shown in the spectrum in Figure 9, after prolonged
reaction (23 h) only the 25.9 ppm resonance remained, indicating
that all final products had unsaturated sp2-hybridized structures.
NMR studies of the dehydrogenated products of AB H2-release
promoted by solid-state thermal reactions4 have likewise shown
the formation of similar types of BdN unsaturated final products
after more than 2 equiv of H2-release.
Figure 10. Solution 11B NMR (128 MHz) spectra recorded at 25 °C of
the reaction of AB (50 mg) in mmimMeSO4 (450 mg) at 85 °C. (Right)
With 5.2 mol % PS (18 mg) after the release of (a) 0.03, (b) 0.08, (c) 0.32,
(d) 1.14, and (e) 1.63 equiv. (Left) Without PS after the release of (a*)
0.03, (b*) 0.08, (c*) 0.29, (d*) 1.07, and (e*) 1.62 equiv. The broad DADB
resonance at -13 ppm is obscured by the AB and PAB resonances.
In situ solution 11B NMR studies of the PS-promoted H2-
release from AB in the room-temperature ionic-liquid mmim-
MeSO4 at 85 °C also exhibited features similar to those observed
in the solid-state NMR spectra of the bmimCl reactions. It is
also significant that the 11B NMR spectra of the reactions with
and without PS in mmimMeSO4 (Figure 10) showed the
formation of similar types of products. In both reactions, DADB
and PAB polymers were formed initially; at the longer times where
more H2 was released, the DADB and PAB resonances decreased
and insoluble materials formed. Only a small intensity resonance
was observed near 25 ppm, indicating that if unsaturated BdN
products such as borazine or polyborazylene were formed, they
must have reacted rapidly to form the final insoluble product.
One of the barriers to the utilization of AB H2-release for
hydrogen storage is that AB dehydrogenation either in the solid
state or in solution normally produces considerable foaming.46
However, as shown in the photographs in Figure 11, Proton
Sponge, besides increasing both the rate and extent of H2-release,
was found to have the unanticipated but highly beneficial effect of
significantly reducing foaming during AB H2-release in the
bmimCl/PS reactions.
PS was also found to increase AB H2-release in both other
ionic liquids and tetraglyme, with both the extent and rate of
H2-release comparable to those of the PS/bmimCl reactions
(Tables 7 and 8 and Figure 12). Gas-buret data at 85 °C for the
tetraglyme reactions showed that fast rates could be achieved
with only 1 mol % of PS. The 11B NMR spectra of the
tetraglyme/PS reactions also showed evidence for the initial
(42) Myers, A. G.; Yang, B. H.; David, K. J. Tetrahedron Lett. 1996, 37,
3623–3626.
(43) Kang, X.; Fang, Z.; Kong, L.; Cheng, H.; Yao, X.; Lu, G.; Wang, P.
AdV. Mater. 2008, 20, 2756–2759.
-
formation of the H3BNH2 anion, followed by the appearance
(44) Xiong, Z.; Chua, Y. S.; Wu, G.; Xu, W.; Chen, P.; Shaw, W.;
Karkamkar, A.; Linehan, J.; Smurthwaite, T.; Autrey, T. Chem.
Commun. 2008, 5595–5597.
of the resonances for PAB polymers. However, at longer times,
unlike in the ionic liquids, the major product was borazine (30.1
(45) Xiong, Z.; Yong, C. K.; Wu, G.; Chen, P.; Shaw, W.; Karkamkar,
A.; Autrey, T.; Jones, M. O.; Johnson, S. R.; Edwards, P. P.; David,
W. I. F. Nat. Mater. 2008, 7, 138–141.
(46) Aardahl, C. L. DOE Hydrogen Program Review, 2008; http://
www.hydrogen.energy.gov/pdfs/review08/st_5_aardahl.pdf.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14107

A R T I C L E S
Himmelberger et al.
Table 8. H2-Release Data (Gas Buret) for AB/Tetraglyme/PS
Reactions at 85 °C
time to H2 equivalents (min)
amount of PS
(mol %, mg, mmol)
total wt
(mg)a
1 equiv
1.5 equiv
final (equiv)
300
310
355
560
248
166
164
193
587
234
263
458
1126 (2)
461 (2)
886 (2)
0.96, 10, 0.047
5.3, 55, 0.26
24.9, 260, 1.21
1238 (1.91)
300
310
355
560
40.7
36.1
27.8
34.9
80.4
51.4
51.6
107
211 (2)
125 (2)
253 (1.87)
325 (1.76)
0.96, 10, 0.047
5.3, 55, 0.26
24.9, 260, 1.21
Figure 11. Foaming resulting from the reaction of 250 mg of AB in 250
mg of bmimCl after 1 h at 100 °C: (a) without PS and (b) with 5.2 mol %
PS (91 mg).
300
310
355
560
17.2
14.4
12.1
18.5
39.6
22.9
25.2
87.2
122 (2)
77.1 (2)
169 (2)
0.96, 10, 0.047
5.3, 55, 0.26
24.9, 260, 1.21
Table 7. H2-Release Data (Toepler Pump) for AB/Ionic-Liquid/PS
Reactions at 85 °C
369 (1.88)
a 150 mg of NH3BH3 (4.87 mmol) and 150 mg of tetraglyme were
used for all reactions.
H2 released
amount of PS
(mol %, mg, mmol)
total wt
(mg)a
solvent
time (h)
equiv
mmol
bdmimCl
1
3
6
500
500
500
500
0.68
1.06
1.31
2.29
5.52
8.58
10.61
18.59
22
bdmimCl
1
3
6
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
591
591
591
591
1.26
1.82
2.03
2.13
10.19
14.74
16.47
17.23
22
edmimEtSO4
edmimEtSO4
tetraglyme
tetraglyme
1
3
6
500
500
500
500
1.00
1.32
1.64
2.20
8.09
10.7
11.3
17.8
22
1
3
6
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
591
591
591
591
1.52
1.91
2.04
2.16
12.33
15.50
16.51
17.47
Figure 12. H2-release measurements (Toepler pump) of the reaction of
AB (250 mg) in ionic liquids or tetraglyme (250 mg) with 5.2 mol % PS
(91 mg) at 85 °C.
21.5
proposed as a key step in thermally induced AB H2-release
reactions in the solid state and in organic solvents.4,48 While a
DADB pathway may contribute to the H2-release observed in
the bmimCl/PS solutions, their observed rate enhancements and
DSC properties suggest that there is also another mechanistic
pathway for H2-release in these systems.
Proton Sponge is a strong base and because of its low
nucleophilicity is frequently used as a deprotonating agent in
organic and inorganic syntheses that can eliminate the undesir-
able side reactions often found with more coordinating bases.30
The observation of a signal near -21.0 ppm in the 11B NMR
spectra at the beginning stages of the reactions of AB with PS
in both ionic liquids and tetraglyme strongly supports a PS-
promoted AB H2-release reaction pathway involving initial AB
deprotonation to form the H3BNH2-PSH+ ion.
1
3
6
500
500
500
500
1.19
1.80
1.97
2.12
9.65
14.57
15.97
17.20
22
1
3
6
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
5.2, 91, 0.43
591
591
591
591
1.32
1.66
1.86
2.13
10.73
13.45
15.10
17.22
22
a 250 mg of NH3BH3 (8.1 mmol) and 250 mg of ionic liquid or
tetraglyme were used for all reactions.
ppm)47 along with smaller amounts of BH4 (-36.8 ppm)47
-
and µ-aminodiborane (-27.5 ppm)47 (Figure 13). Thus, ionic-
liquid solvents are favored for AB H2-release since they suppress
or reduce the formation of these products.
We have previously proposed5 that the AB activation for H2-
release that is observed in ionic liquids may occur by a
mechanistic pathway involving (1) ionic-liquid-promoted con-
version of AB into its more reactive ionic DADB form, (2)
further intermolecular dehydrocoupling reactions between hy-
dridic B-H hydrogens and protonic N-H hydrogens on DADB
and/or AB to form neutral PAB polymers, and (3) PAB
dehydrogenation to unsaturated cross-linked polyborazylene
materials. The initial formation of DADB has also been
Deprotonation of AB by bases is well-known. For example,
-
the H3BNH2 Li+ salt has previously been prepared in situ from
AB and n-BuLi at 0 °C.42 More recently, ball milling43,45,49
and solution50,51 reactions of AB with metal hydrides have been
used to generate H3BNH2-M+ (M ) Li and Na) and
(H3BNH2-)2Ca2+. We found that AB deprotonation could also
be easily achieved by its reaction with either lithium or
potassium triethylborohydride (eq 7), but in these cases a Et3B-
(48) (a) Shaw, W. J.; Linehan, J. C.; Szymczak, N. K.; Heldebrant, D. J.;
Yonker, C.; Camaioni, D. M.; Baker, R. T.; Autrey, T. Angew. Chem.,
Int. Ed. 2008, 47, 7493–7496. (b) Heldebrant, D. J.; Karkamkar, A.;
Hess, N. J.; Bowden, M.; Rassat, S.; Zheng, F.; Rappe, K.; Autrey,
T. Chem. Mater. 2008, 20, 5332–5336.
(47) No¨th, H.; Wrackmeyer, B. Nuclear Magnetic Resonance Spectroscopy
of Boron Compounds; Springer-Verlag: New York, 1978; pp 188, 265,
394-395.
9
14108 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Base-Promoted Ammonia Borane H2-Release
A R T I C L E S
Figure 14. Selected bond distances (Å) and angles (°) for [Et3BNH2-
BH3]-K+ ·18-crown-6: B3-N2, 1.594(2); B1-N2, 1.630(2); B1-C4,
1.622(2); B1-C6, 1.637(2); B1-C8, 1.629(2); B3-H3a, 1.13(2); B3-H3b,
1.14(2); B3-H3c, 1.13(2); N2-H2a, 0.94(2); N2-H2b, 0.91(2); K1-H3a,
2.85(2);K1-H3c,2.76(2);B1-N2-B3,122.05(10);N2-B1-C4,108.75(11);
N2-B1-C6, 105.17(10); N2-B1-C8, 108.43(11); B1-N2-H2a, 107.0(11);
B1-N2-H2b, 107.6(10); H2a-N2-H2b, 106(2).
Figure 13. Solution 11B{1H} NMR (128 MHz) spectra (insets show
1H-coupled spectra) recorded at 80 °C of the reaction of AB (50 mg) and 5.2
mol % PS (18 mg) in tetraglyme (450 mg) at 85 °C after (a) 1, (b) 3, and (c) 6 h.
stabilized anion was formed, with the observed resonances in the
11B NMR spectra of the Li+ (-7.5 (s) and -23.8 (q) ppm) and
K+ (-7.8 (s) and -23.6 (q) ppm) compounds being assigned on
the basis of the DFT/GIAO calculations to the BEt3- (calcd -10.6)
and terminal -BH3 (calcd -25.3) units in the anion.
H3BNH3 + M+BEt3H- f [Et3BNH2BH3]-M+ + H2
(M ) Li or K) (7)
The structures of H3BNH2-M+ (M ) Li and Na) and
(H3BNH2-)2Ca2+ have recently been established from the
analyses of their high-resolution powder diffraction data,49,51
and the structure of (H3BNH2-)2Ca2+ ·2THF has been estab-
lished by a single-crystal X-ray determination.50 Our crystal-
lographic determination of the [Et3BNH2BH3]-K+ ·18-crown-
6 complex in Figure 14 confirmed that, following AB nitrogen-
deprotonation by the triethylborohydride, the Lewis-acidic
triethylborane group coordinated at the nitrogen. Selected bond
lengths and angles are listed in the figure caption. The angles
around B1 indicate sp3 hybridization resulting from the nitrogen
lone-pair donation to the vacant boron orbital on the triethylbo-
rane fragment. The observed N2-B3 length of 1.594(2) Å is
similar to that found in AB (1.58(2) Å)52 but longer than those
determined in the powder diffraction studies of H3BNH2-Li+
(1.561(7) Å) and (H3BNH2-)2Ca2+ (1.575(4) Å).51 The longer
distance of the B1-N2 1.631 (2) Å bond compared to that of
N2-B3 is consistent with the higher Lewis acidity of the H3B
(49) Ramzan, M.; Silvearv, F.; Blomqvist, A.; Scheicher, R. H.; Lebegue,
S.; Ahuja, R. Phys. ReV. B: Condens. Matter Mater. Phys. 2009, 79,
132102-4.
Figure 15. Possible anionic polymerization pathway for PS-promoted H2-
release from AB.
(50) Diyabalanage, H. V. K.; Shrestha, R. P.; Semelsberger, T. A.; Scott,
B. L.; Bowden, M. E.; Davis, B. L.; Burrell, A. K. Angew. Chem.,
Int. Ed. 2007, 46, 8995–8997.
fragment (and hence stronger NfB dative bonding donation)
compared to the Et3B unit. The K+ cation is coordinated to the
six oxygen atoms of the 18-crown-6 ring with distances ranging
from 2.79 to 2.92 Å. Consistent with the hydridic nature of the
BH hydrogens of Et3BNH2BH3-, the closest interactions of the
(51) Wu, H.; Zhou, W.; Yildirim, T. J. Am. Chem. Soc. 2008, 130, 14834–
14839.
(52) Klooster, W. T.; Koetzle, T. F.; Siegbahn, P. E. M.; Richardson, T. B.;
Crabtree, R. H. J. Am. Chem. Soc. 1999, 121, 6337–5343.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14109

A R T I C L E S
Himmelberger et al.
Figure 17. DFT (B3LYP/6-31G(d))-optimized geometry and GIAO
(B3LYP/6-311G(d)) calculated 11B NMR shifts for [Et3BNH2BH2NH2BH3]-.
indicated formation of Et3BNH2BH2NH2BH3- (eq 8). As shown
in the 11B NMR spectra in Figure 16, new resonances grew in
at -6.0 (s, Et3B-), -10.9 (t, -NH2BH2-), and -19.6 (q,
-BH3) that are in good agreement with the calculated 11B NMR
shifts for the DFT-optimized geometry of the chain growth
-
product Et3BNH2BH2NH2BH3 given in Figure 17.
[Et3BNH2BH3]-Li+ + H3BNH3 f
[Et3BNH2BH2NH2BH3]-Li+ + H2 (8)
When the reaction was carried out at higher temperatures, the
growth of the BNHx oligomers Et3B(NH2BH2)2NH2BH3- and
Et3B(NH2BH2)3NH2BH3- was observed in the electrospray mass
spectra. These results provide additional strong support for a PS-
induced H2-release mechanism involving an intermolecular anionic
dehydropolymerization pathway initiated by the NH2BH3- anion.
In summary, base-activation by Proton Sponge of AB H2-release
has been demonstrated in solid-state and in ionic-liquid and
tetraglyme solution reactions, with the experimental observations
and model studies supporting an anionic dehydropolymerization
mechanism initiated by the H3BNH2- anion. For the bmimCl/PS
reactions, significantly increased rates of AB H2-release, yielding
over 2 equiv of H2, were achieved with as little as 1 mol % Proton
Sponge. These studies also further demonstrate that H2-release from
chemical hydrides can occur by a number of different mechanistic
pathways and strongly suggest that optimal chemical-hydride-based
H2-release systems may require the use of synergistic dehydroge-
nation methods to induce H2-loss from chemically different
intermediates formed during release reactions.
Figure 16. Reaction progression monitored by 11B{1H} NMR of the
reaction of [Et3BNH2BH3]-Li+ and AB at room temperature, showing the
growth of the [Et3BNH2BH2NH2BH3]- resonances: (a) 0, (b) 18, (c) 44,
(d) 67, and (e) 144 h.
K+ to the Et3BNH2BH3 anion are with two of the BH
-
hydrogens, K1-H3a 2.85(2) Å and K1-H3c 2.76(2) Å.
Electronic structure calculations on [H2NBH3]-M+ have
predicted49,51 that the B-H hydrogens in these complexes will
be much more negatively charged and, as a result, have a higher
basicity than in AB. This increases the ability of these anions
to release H2 by the intermolecular reaction of their hydridic
B-H hydrogens with an N-H proton on an adjacent complex.
Thus, the most likely second step in PS-promoted AB H2-release
is the intermolecular dehydrocoupling of one of the hydridic
B-H hydrogens of [H2NBH3]-PSH+ with a H3NBH3 to
produce, as shown in Figure 15a, a growing anionic PAB
polymer. DFT/GIAO calculations of model anionic polymers
showed that their 11B NMR chemical shifts (Figure 2S, Supporting
Information) would be indistinguishable from those of the corre-
sponding neutral PAB polymers,5 thus explaining the similarity
of the spectra for the reactions with and without PS in Figure 10.
The expected decreased acidity of the internal NH2 hydrogens
of a PAB polymer relative to those on terminal NH3 units should
result in the NH2 protons being less reactive for dehydrocoupling
reactions that could liberate more than 1 equiv of H2. Thus, the
increased release rate for the second equivalent of H2 observed
for the PS-induced reactions of AB may result from the greater
ability of the more basic B-H hydrogens of an anionic PAB to
induce H2-elimination reactions with the protons at these NH2 sites.
As a model for the anionic polymerization step, we explored
the reactions of [Et3BNH2BH3]-Li+ with AB. The higher Lewis
acidity of the BH3 group compared to the Et3B unit should
significantly increase the nucleophilic character of the boron-
hydrogens and enhance the ability of this anion to undergo
dehydropolymerization with AB. In agreement with this ex-
pectation, Toepler-pump measurements confirmed the loss of
H2 during the reaction of equal equivalents of NH3BH3 with
[Et3BNH2BH3]-Li+ at 23 °C, and an electrospray mass spectrum
Acknowledgment. We thank the U.S. Department of Energy
for grants from the Center of Excellence for Chemical Hydrogen
Storage and the Division of Basic Energy Sciences for the support
of this research. A portion of the research was performed in the
Environmental Molecular Sciences Laboratory, a national scientific
user facility sponsored by the Department of Energy’s Office of
Biological and Environmental Research, located at Pacific North-
west National Laboratory. We especially thank Dr. Tom Autrey for
his generous invitation to visit PNNL, Dr. Wendy Shaw and Dr. Sarah
Burton for their assistance with the solid-state NMR, and Dr. Ahbi
Karkamkar for assistance with the DSC used in this study.
Supporting Information Available: Diagram of the Toepler
pump apparatus used for some H2-release measurements; tables
listing the Cartesian coordinates for all DFT-optimized geometries;
X-ray crystallographic data for the structure determination of
[Et3BNH2BH3]-K+ ·18-crown-6 (CIF); complete ref 36. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JA905015X
9
14110 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009