Reversible, metal-free hydrogen activation by frustrated Lewis pairs

Article abstract of DOI:10.1039/c0dt01255f

The Lewis acid cyclohexylbis(pentafluorophenyl)boron 1, which exhibits about 15% lower Lewis acidity in comparison with B(C₆F ₅)₃, activates H₂ in the presence of the bulky Lewis bases 2,2,6,6-tetramethylpiperid

Full text of DOI:10.1039/c0dt01255f

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 1091
www.rsc.org/dalton
PAPER
Reversible, metal-free hydrogen activation by frustrated Lewis pairs†
Chunfang Jiang, Olivier Blacque, Thomas Fox and Heinz Berke*
Received 17th September 2010, Accepted 1st November 2010
DOI: 10.1039/c0dt01255f
The Lewis acid cyclohexylbis(pentafluorophenyl)boron 1, which exhibits about 15% lower Lewis acidity
in comparison with B(C6F5)3, activates H2 in the presence of the bulky Lewis bases 2,2,6,6-
tetramethylpiperidine (TMP), 1,2,2,6,6-pentamethylpiperidine (PMP), tri-tert-butylphosphine (t-Bu3P)
leading in facile reactions at room temperature to heterolytic splitting of dihydrogen and formation of
the salts [TMPH][CyBH(C6F5)2] 2, [PMPH][CyBH(C6F5)2] 3 and [t-Bu3PH][CyBH(C6F5)2] 4, which
could be dehydrogenated at higher temperatures. The related Lewis acid 1-phenyl-2-[bis-
(pentafluorophenyl)boryl]ethane 5 exhibiting about 10% lower Lewis acidity than B(C6F5)3 is also
capable of splitting H2 in a heterolytic fashion in the presence of TMP, PMP and t-Bu3P yielding
[TMPH][PhC2H4BH(C6F5)2] 6, [PMPH][PhC2H4BH(C6F5)2] 7 and [t-Bu3PH][PhC2H4BH(C6F5)2] 8.
Under comparable conditions as for 2–4, the dehydrogenations of 6–8 were much slower. 4b and 6 were
characterized by single crystal X-ray diffraction studies.
by the work of Stephan and his co-workers established a new
Introduction
so-called “metal-free” way of hydrogenation/dehydrogenation.
Dehydrogenation and hydrogenation processes are potential “fu-
Frustrated Lewis pairs (FLPs), which are main group element
elling” and “refuelling” reactions of a chemical storage system.1,2
“unquenched” Lewis acid–base pairs. The constituents stay re-
mote by steric hindrance forming encounter complexes.14–16 For
instance, the addition of H2 to the intramolecular Lewis pair
(C6H2Me3)2P(C6F4)B(C6F5)2 resulted in a zwitterionic phospho-
nium borate salt (C6H2Me3)2PH+(C6F4)B-H(C6F5)2, thus activat-
ing H2 in a heteropolar fashion at two reaction centers.17 Due to
their high molecular weights, such FLP compounds are certainly
not suited for H2 storage. But the given low energy pathways
for H2 activation may show new directions to develop efficient
storage materials. In particular we thought it necessary to acquire
more insights into the conditions for reversibility of the underlying
chemical processes.
Following the pioneering work of Stephan, an increasing
number of FLPs were studied exploring the chemical influence
of the Lewis base.18–38 It was found that the Lewis base function
is not limited to phosphine centers, but could be extended to
sterically demanding amines, imines, or N-heterocycylic carbenes.
In contrast, the influence of the Lewis acid was studied much
less. In the majority of FLPs B(C6F5)3 was used. To extend the
series of functional Lewis acids, our group has explored the
role of 1,8-bis(dipentafluorophenyl)naphthalene, ClB(C6F5)2 and
HB(C6F5)2 in the activation of H2 with bulky Lewis bases.39,40
The hydrogenation process was found to proceed in a facile
manner, but reversibility of the conversions could not be achieved.
Tuning the Lewis acidity it was found that B(p-C6F4H)3 with
about 5% less acidity compared to B(C6F5)3 (Gutmann–Beckett’s
method41–44) and presumably due to the changed electronics the
FLP with (o-C6H4Me)3P enabled reversible activation of H2 at
room temperature.28 Since in this case the change in the sterics
and electronics of the Lewis acid was not much, we thought that
These reactions could occur as 1,2-eliminations/additions
from/to a(n) saturated/unsaturated substrate or as binuclear
activation processes. Both reactions should be reversible and
facile, i.e. possessing low kinetic barriers. This is usually the
case when the involved H atoms bear opposite polarizations.3
Among the compounds with lighter main group elements, which
were considered suitable for the given purpose, ammonia borane
became a major research focus.4–9 In recent years respective studies
emphasized the dehydrogenation occuring as a 1,2-elimination
process. As yet the system is irreversible partly due to thermody-
namics, but also due to kinetic short-comings mainly caused by
too strong bonds involved. Apparently compounds with heavier
main group elements involving generally weaker E–H bonds allow
dehydrogenation/hydrogenation reactions with lower barriers.
Early model studies on homopolar hydrogenations were carried
out in vapor phase with the heavier group 13 elements by trapping
intermediates in frozen matrices.10–12 Similarly, Power and co-
workers reported that the highly unsaturated model compound
“digermyne” can directly react with H2 affording a mixture of
digermene, digermane and germane.13 One might speculate that
the process takes the course of a direct 1,2-addition processes
with simultaneous addition and splitting of the H2 molecule or
with bicentered radical type additions.3 The mentioned reversible
heteropolar binuclear activation pathway of H2 as introduced
Anorganisch-Chemisches Institut, Universita¨t Zu¨rich, Winterthurerstrasse
190, CH-8057, Zu¨rich, Switzerland. E-mail: hberke@aci.uzh.ch; Fax: +41
44 6356802; Tel: +41 44 6354681
† CCDC reference numbers 793471 and 793472. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/c0dt01255f
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 1091–1097 | 1091
©

View Article Online
related variations of the B(C6F5)3 master compound might not
only lead to reversibly activating systems, but also to a conceptual
conclusion of how to steer FLPs toward reversibility.
of a boron bound H atom. Interestingly, 2 was quite unstable in
solution indeed gradually releasing H2 at room temperature. The
H2 liberation became accelerated when the solution was heated
to 50 ◦C. The dehydrogenation reaction of 2 proceeded then so
quickly that within 30 min the yield of free CyB(C6F5)2 and TMP
amounted to more than 80% (Fig. 1, curve 1). This is indeed
one of the rare examples of a B,N FLP reversibly activating H2 in
solution, but apparently also in the solid state at room temperature,
where 2 showed also slow release of H2.
Results and discussion
With regard to reversible release of H2, we aimed at the
exploration of the Lewis acids cyclohexylbis(pentafluorophenyl)-
boron 1 (CyB(C6F5)2) and 1-phenyl-2-[bis(pentafluorophenyl)-
boryl]ethane 5 (PhC2H4B(C6F5)2) as FLP components modified
for reduced Lewis acidity with respect to B(C6F5)3. According to
Gutmann–Beckett’s method41–44 these Lewis acids showed reduced
acidities by about 15% and 10%. They were tested for reversible
H2 up-take in the presence of various bulky N,P-Lewis bases as
shown in Scheme 1.
Fig. 1 Dehydrogenation of 2, 3, 6 and 7 in C6D6 at various temperatures.
Curve 1: 2 at 50 ◦C, curve 2: 3 at 65 ◦C, curve 3: 3 at 80 ◦C, curve 4: 6 at
50 ◦C, curve 5: 6 at 65 ◦C, curve 6: 7 at 80 ◦C.
Modifying also the base component we then studied the H2
reaction of the FLP PMP ◊ ◊ ◊ BCy(C6F5)2 (PMP = 1,2,2,6,6-
pentamethyl piperidine) according to Scheme 1 by exposure of
a toluene solution of the stoichiometric mixture of PMP and
BCy(C6F5)2 to an atmosphere of H2 (1000 mbar) for 1 h at room
temperature. The reaction observed resulted in the formation
of the ionic product [PMPH][HBCy(C6F5)2] 3 isolated as an
Scheme 1
To prepare CyB(C6F5)2 (1), HB(C6F5)2 was treated in a hydrob-
oration reaction with 1 equiv. of cyclohexene in toluene, which
afforded 1 as a white solid in 92% yield. The 1H NMR spectrum
showed signals for the cyclohexyl group at 2.04 (m, 1H, CH), 1.67
(m, 4H, CH2), 1.20 ppm (m, 6H, CH2). The 19F NMR spectrum
[d -132.0 (o-), -149.9 (p-), -162.2 (m-C6F5) (C6D6)] was consistent
with the presence of a three-coordinate boron atom (Ddm,p = 12.3).
We reckoned that the steric congestion of 1 would still be close to
that of B(C6F5)3 and any difference in the FLP reactivity of 1 was
expected to originate from an electronic effect, i.e. its lower Lewis
acidity, which was thought to be translated into a considerably
weaker B–H bond of the tetracoordinate borate species with
concomitant higher propensity for dehydrogenation.
1
analytically pure, white solid in 72% yield (Scheme 1). In the H
NMR spectrum 3 featured broad NH and BH resonances at 5.42
and 2.70 ppm and in the 11B NMR spectrum a signal at -15.7 ppm
with a 1JBH coupling of 80 Hz. The reaction with H2 turned out to
be reversible, as well, occu◦rring, however, at the somewhat elevated
temperature of about 65 C. The maximum conversion rate in a
closed system was found to be about 80% within 1 h (Fig. 1. curve
2). At 80 ◦C, the maximum conversion was 85% within 30 min
(Fig. 1, curve 3).
When the FLP t-Bu3P ◊ ◊ ◊ Cy(C6F5)2 containing a phosphorus
Lewis base, was reacted with H2 at room temperature, a white
product was isolated in 74% yield after work-up. The NMR pursuit
of the reaction mixture revealed in the 11B NMR spectrum a broad
signal located at -22.2 ppm. Surprisingly, in the 1H NMR spectrum
(C6D6) two PH resonances appeared at 4.61 (4a) and 5.71 ppm
(4b) with 1JHP couplings of 438 and 453 Hz and in the 31P NMR
spectrum the corresponding resonances were found at 56.4 and
52.2 ppm (Fig. 2). The MAS 31P NMR of solid 4 also exhibited
two signals at 66.7 (broad) and 59.2 (sharp) ppm. Especially based
on the relatively large chemical shift difference these could be
attributed to the presence of a crystal mixture of 4a and 4b.
Exposure of a toluene solution of the stoichiometric mixture of
1 and 2,2,6,6-tetramethylpiperidine (TMP) to an atmosphere of
H2 (1000 mbar) for 30 min afforded the H2 cleaved ionic product
[TMPH][HBCy(C6F5)2] 2 (Scheme 1), which was isolated as a
1
white solid in 76% yield. In the H NMR spectrum 2 displayed
a broad BH resonance at 2.67 ppm and a NH resonance at
4.74 ppm. The 19F NMR spectrum featured a set of typical C6F5-
borate resonances at d -132.7 (o-), -164.6 (p-) and -167.1 (m-C6F5)
ppm. The 11B NMR signal was found to be split into a doublet at
-15.5 ppm with a 1JHB coupling of 88 Hz supporting the presence
1092 | Dalton Trans., 2011, 40, 1091–1097
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
1
Fig. 2 31P{ H} NMR (insert) and 1H NMR spectrum of 4 in C6D6 at r.t.
It is interesting to note that when the FLP t-Bu3P ◊ ◊ ◊ CyB(C6F5)2
was reacted with D2, the 1H NMR spectrum also gave two PH reso-
nances at 5.52 and 4.44 ppm, and the 31P NMR spectrum exhibited
four resonances at 56.8 (s, PH), 55.8 (t, PD), 52.9 (s, PH), 52.0
(t, PD) (Fig. 3). This observation suggested an exchange reaction
between D and H. One assumption was that the reaction was
initiated by the release of the tert-butyl cation from the generated
[t-Bu3PD]+ cation. The relatively stable carbocation [(CH3)3C]+
became then deprotonated by another molecule of t-Bu3P to
generate isobutylene CH2 C(CH3)2, which then inserted into
the intermediate [t-Bu2PD] compound to generate a D-labelled
t-Bu3P. The release of isobutene and t-Bu2PH from the cation
[t-Bu3PH]+ has also been put forward for the [t-Bu3PH][HB(p-
C6F4H)3] case published by Stephan’s group.45 Also in this case
evidence for the appearance of free CH2 C(CH3)2 could not
be provided, even in presence of a great excess of t-Bu3P. We
assumed the isobutylene insertion reaction was approximately as
fast as the isobutylene generation, a circumstance, which naturally
is expected to prevent detection of the free olefin. Once the
isobutylene generated, it quickly inserts (Scheme 2).
Scheme 2
Fig. 4 Molecular structure of the ion pair 4b, 30% probability thermal
ellipsoids are shown. Hydrogen atoms except for the PH and BH
are omitted for clarity. Selected bond lengths [A]: B1–H1 1.205(12)
˚
B1–C1 1.651(2), B1–C7 1.653(2), B1–C13 1.630(2), P1–C19 1.8684(14),
P1–C23 1.8748(15), P1–C27 1.8663(15), P1–H2 1.298(13). Selected bond
angles [◦] C13–B1–C1 116.93(12), C13–B1–C7 111.01(12), C1–B1–C7
106.27(11), C13–B1–H1 108.2(6), C1–B1–H1 105.7(6), C7–B1–H1
108.3(6), C27–P1–C19 114.50(7), C27–P1–C23 114.37(7), C19–P1–C23
113.90(7), C27–P1–H2 103.1(6), C19–P1–H2 105.8(6), C23–P1–H2
103.4(6).
measurements did not provide any hint for correlation effects
between the cation and the anion. Another proposal for the
structure of 4a was the formation of [t-Bu2PH2][CyBH(C6F5)3] (or
[t-Bu2PHD][CyBH(C6F5)3]), based on the generation of t-Bu2PH
during the H-D (H) exchange reaction. But no J(HD) J(PD)
1
Fig. 3 31P{ H} NMR spectrum of the FLP t-Bu3P ◊ ◊ ◊ CyB(C6F5)2 to
react with D2 in C6D6 at r.t.
1
The products of the reactions of the FLP t-Bu3P ◊ ◊ ◊ CyB(C6F5)2
with H2 or D2 in benzene were two species coexisting in solution
and in solid state. One product should be the expected ionic
product [t-Bu3PH][CyBH(C6F5)3] 4b, which is consistent with the
coupled resonances were observed in the H NMR or the 31P
NMR spectrum and a 31P NMR resonance was also not detected
expected for the phosphonium cation [t-Bu2PH2]+ at 26 ppm.45 So
this proposal could not be verified by appropriate spectroscopic
compliance, as well.
Based on the fact that the isolated product mixture dissolved
in the non-polar solvent C6D6 giving two different species,
but in the polar solvent CDCl3 only one species, we first
of all reckoned that both compounds are structurally related.
We suspected the unusual hydrogen bonded species [t-Bu3P–
H ◊ ◊ ◊ PtBu3][CyBH(C6F5)2] to be 4a (Scheme 1). The structure
is assumed to possess a double minimum potential. A dynamic
process could lead to exchange of the binding modes of the two
P atoms. A dynamic process could not be “frozen” out even at
NMR spectra (1H NMR: 5.71 ppm; 31P{ H} NMR: 52.2 ppm) and
1
the X-ray diffraction analysis, in which the anion and the cation
are oriented “face-to-face” to each other with a non-bonding
46
˚
H1 ◊ ◊ ◊ H2 distance of 2.63 A (Fig. 4). For the other species,
several assumptions are given. One assumption was the Coulombic
ion pair of 4a, in which the cation and the anion were anticipated
to be held together additionally by multiple, but generally weak,
CH ◊ ◊ ◊ F, PH ◊ ◊ ◊ F or BH ◊ ◊ ◊ HP dihydrogen bonding interactions.
These weak interactions of the hydrogen or dihydrogen bonding
type were not detectable by NMR spectroscopy. T1(min) or NOE
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 1091–1097 | 1093
©

View Article Online
-80 ◦C. The ground state structures of a dynamic or a static
P-H ◊ ◊ ◊ P binding situation have only one strong covalent P–H
contact, which leads, in both cases, to detectable doublet coupling
in the 1H NMR spectrum. In order to prove this assumption, we
then varied the ratios of t-Bu3P to CyB(C6F5)2 in the reaction with
H2 and monitored the reaction course. The ratio of the products 4a
or 4b turned out to be sensitive with respect to the applied amount
of t-Bu3P, only if t-Bu3P was in great excess. For instance in the
cases of the 1 : 2 or 1 : 1 ratios with respect to CyB(C6F5)2, the
majority of the product is 4b [t-Bu3PH][CyBH(C6F5)3] and 4a is
formed in small amounts only. But when the t-Bu3P to CyB(C6F5)2
ratio was increased to 2 : 1, 4a became the predominant product
and 4b could not be detected at all when t-Bu3P was present in great
excess (t-Bu3P: CyB(C6F5)2 = 5 : 1). This observation suggested
that 4a is formed from 4b in an equilibrium reaction via the
addition of t-Bu3P (Scheme 1). 4a featured a PH resonance at
reversibility in the H2 splitting. In addition it was found earlier
that H2 can be heterolytically cleaved by the t-Bu3P ◊ ◊ ◊ B(C6F5)3,
TMP ◊ ◊ ◊ B(C6F5)3 or PMP ◊ ◊ ◊ B(C6F5)3 FLPs, but none of the
corresponding salts [t-Bu3PH][HB(C6F5)3], [TMPH][HB(C6F5)3]
and [PMPH][HB(C6F5)3]47 could release H2 even at higher tem-
peratures. This inability was prevailingly attributed to a too high
B–H bond strength in the [HB(C6F5)3]- anion related to the higher
Lewis acidity of B(C6F5)3.
In order to further substantiate the given assumption
of the Lewis acidity influence, we selected the Lewis acid
PhCH2CH2B(C6F5)2 5 with a Lewis acidity between those of
BCy(C6F5)2 1 and B(C6F5)3. 5 was then applied in FLPs in combi-
nation with the same Lewis bases as for 1. As expected, the FLPs
TMP ◊ ◊ ◊ PhCH2CH2B(C6F5)2 and PMP ◊ ◊ ◊ PhCH2CH2B(C6F5)2
induced heterolytic splitting of H2 at room temperature to form
white solids of the ionic products [TMPH][PhCH2CH2BH(C6F5)2]
6 and [PMPH][PhCH2CH2BH(C6F5)2] 7 in 85% and 83% yield.
1
1
4.61 ppm in H NMR spectrum with a JHP coupling of 438 Hz.
The lower J(HP) coupling of the PH signal of 4a compared to 4b of
453 Hz could indeed be related to [P–H ◊ ◊ ◊ P] hydrogen bonding,
since the hydrogen bonding of P(tBu)3 is expected to withdraw
electron densities from the P–H bond and decrease its bond order.
Additional experiments showed that both 4a and 4b were not stable
in solution and gradually underwent dehydrogenation reactions.
1
Both of these ionic compounds featured in the H NMR signals
of the NH resonances with chemical shifts at 4.52 and 5.49 ppm
in C6D6 similar to those of 2 and 3 indicating the presence of the
same cationic species. In addition, the 19F NMR resonances of
6 (-133.7 (o-), -164.3 (p-), -166.9 (m-C6F5) ppm) and 7 (-133.6
(o-), -164.1 (p-), -167.0 (m-C6F5) ppm) are comparable to those
of 2 and 3. As anticipated, both compounds 6 and 7 released H2
under mild conditions, but the rates were not quite as high as
1
According to H NMR and 31P NMR pursuits, the PH signals
of 4a disappeared first at room temperature accompanied by the
appearance of signals for free H2 located at 4.45 ppm and of free
t-Bu3P. 4b seemed somewhat more stable in this process, but it
disappeared also at elevated temperatures.
Further investigations demonstrated that the dehydrogenation
reaction of the mixture of 4 is quite temperature dependent.
At room temperature and based on NMR spectroscopy, the
conversion was about 30% within 3 h in C6D6 solution (Fig. 5,
curve 1). While at 50 ◦C, the conversion reached 60% (Fig. 5,
curve 2), and at the higher temperature of 80 ◦C more than 80%
of free t-Bu3P and CyB(C6F5)3 were regenerated within 1 h (Fig. 5,
curve 3).
◦
in the BCy(C6F5)2 cases. After 1 h at 50 C, the conversion of 6
was only 15% (Fig. 1, curve 4), while for 2 it had reached more
than 80% under the same conditions. In order to accomplish the
same yields in the dehydro◦ genation process as for 2, 6 required
temperatures of about 65 C. Due to the higher Lewis acidity of
PhCH2CH2B(C6F5)2, 5 was supposed to generate a stronger B–H
bond in the hydrido borate anion than BCy(C6F5)2 consequently
requiring more severe conditions in dehydrogenation batches.
Single-crystals of 6 were grown from a toluene–hexane solution
at 243 K and analyzed by X-ray diffraction. After refinement of
the NH2 and BH hydrogens a close B–H ◊ ◊ ◊ H–N bonding contact
48
˚
of 1.88 A could be extracted, consistent with strong dihydrogen
bonding (Fig. 6). Similar dihydrogen bonding was observed in
the X-ray crystal structure of [t-BuNH2-CH2Ph][HB(C6F5)3], in
which the distance between one of the NH2 protons and the B–H
49
˚
hydride was 1.87(3) A. Despite this close contact, no evidence
was found for spontaneous dehydrogenation or dehydrogenation
taking place at elevated temperatures. 6, however, released H2 at
65 ◦C was quite fast. This again provided evidence that reformation
of H2 from a [LBH][LAH] pair (LB: Lewis base; LA: Lewis acid)
is mainly correlated with the strength of the B–H bond rather than
with a kinetically feasible short contact between the protonic and
hydridic hydrogen atoms.
The FLP t-Bu3P ◊ ◊ ◊ BPhCH2CH2(C6F5)2 also enabled hetero-
lytic activation of H2 in the same way as the t-Bu3P ◊ ◊ ◊ CyB-
(C6F5)2 system. The isolated product exhibited again two
different PH resonances at 5.34 (8b, [t-Bu3P–H ◊ ◊ ◊ PtBu3]-
[PhCH2CH2BH(C6F5)2]) and at 4.13 ppm (8a, [t-Bu3P–H]-
[PhCH2CH2BH(C6F5)2]) in the 1H NMR spectrum in C6D6
solution with 1JHP coupling constants of 447 and 432 Hz and in the
31P NMR spectrum also two signals were located at 53.5 (8b) and
57.4 ppm (8a). The solid state 31P NMR also gave two different
resonances of 59.3 (sharp, 8b) and 64.0 (broad, 8a) ppm indicating
that both compounds are co-crystallizing. A similar observation
Fig. 5 Dehydrogenations of 4 and 8 in C6D6 at various temperatures
pursued by 1H NMR. Curve 1: 4 at r.t., curve 2: 4 at 50 ◦C, curve 3: 4 at
80 ◦C, curve 4: 8 at r.t., curve 5: 8 at 50 ◦C, curve 6: 8 at 80 ◦C.
The series of FLP combinations applied up to now, showed that
the Lewis acidity of the Lewis acid is a crucial factor to achieve
1094 | Dalton Trans., 2011, 40, 1091–1097
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
and were degassed by freeze–thaw cycles prior to use. All
organic reagents were purchased from Aldrich and used without
further purification. HB(C6F5)2 was synthesized according to the
1
literature,50,51 cyclohexene and styrene were dried before use. H
1
1
NMR, 19F NMR and 11B{ H} NMR 31P{ H} NMR data were
recorded on a Varian Gemini-300 spectrometer. Chemical shifts
are expressed in parts per million (ppm) referenced to deuterated
1
1
solvent used. 19F NMR, 11B{ H} NMR and 31P{ H} NMR were
referenced to CFCl3, BF3·OEt2 and 85% H3PO4, respectively.
Microanalyses were carried out at the Anorganisch-Chemisches
Institut of the University of Zu¨rich.
Crystallographic data were collected at 183(2) K on an Oxford
Xcalibur diffractometer (4-circle kappa platform, Ruby CCD
detector and a single wavelength Enhance X-ray source with
Fig. 6 Molecular structure of 6, 30% probability thermal ellipsoids are
shown. Hydrogen atoms except for the NH and BH are omitted for
clarity. Selected bond lengths [A]: B1–H1 1.283(18), B1–C1 1.643(2),
B1–C7 1.639(2), B1–C13 1.634(2), N1–C21 1.542(2), N1–C25 1.528(2),
N1–H2 0.98(2), N1–H3 0.89(2). Selected bond angles [◦]: C25–N1–C21
120.63(12), C25–N1–H2 110.7(12), C21–N1–H2 104.9(12), C25–N1–H3
104.6(13), C21–N1–H3 109.1(13), H2–N1–H3 106.1(18), C13–B1–C7
109.27(12), C13–B1–C1 116.23(13), C7–B1–C1 108.43(12), C13–B1–H1
108.0(8), C7–B1–H1 110.6(8), C1–B1–H1 104.3(8).
52
˚
Mo-Ka radiation, l = 0.71073 A). The selected suitable single
˚
crystals were mounted using polybutene oil on the top of a glass
fiber fixed on a goniometer head and immediately transferred to
the diffractometer. Pre-experiment, data collection, absorption
correction and data reduction were performed with the Oxford
program suite CrysAlisPro.53 The structures were solved with
direct methods (SHELXS-97) and were refined by full-matrix
least-squares methods on F2 (SHELXL-97).54 All programs used
during the crystal structure determination processes are included
in the WINGX software.55 The program PLATON56 was used to
check the result of the X-ray analyses.
was made as for 4: when 8 was dissolved in the more polar solvent
CDCl3, only one species could be observed which was attributed
to the structure of the solvated ions of 8b.
By analogy to 4 we assume that there are two coexisting ion
pairs 8a and 8b in benzene solution, and in the solid state we
suppose in analogy to 4 the presence of a crystalline mixture of
8a and 8b (Scheme 1). Moreover, the ion pair 8a seemed to be
quite unstable in solution accompanied by fast H2 loss, while 8b in
CDCl3 was found to be more stable. At the elevated temperature
of 80 ◦C, the conversion of the dehydrogenation of the mixture of
8 also increased to more than 80% (Fig. 4, curve 6).
Synthesis of CyB(C6F5)2 1 and PhC2H4B(C6F5)2 5. These two
compounds were prepared in a similar fashion and thus for only
one is the preparation detailed. HB(C6F5)2 (0.346 g, 1 mmol) and
cyclohexene (0.1 g, 1.2 mmol) were dissolved in 5 mL of toluene.
The slurry turned clear after 5 min, the solution was stirred for
an additional 30 min. Then the solvent was removed in vacuo to
obtain 1 as a white solid which was washed by hexane and dried
in vacuo.
CyB(C6F5)2 1. Yield: 92%. 1H NMR (C6D6, 300 MHz, 298 K):
d 2.04 (m, 1H, CH), 1.67 (m, 4H, CH2), 1.20 ppm (m, 6H, CH2).
19F NMR (C6D6, 282 MHz, 298 K): d -132.0 (d, 4F, 3JFF = 23 Hz,
Conclusions
In a tuning effort the two Lewis acids CyB(C6F5)2 1 and
PhC2H4B(C6F5)2 5 were applied as FLPs in combination with
the bulky Lewis bases TMP, PMP and t-Bu3P to split dihydrogen
heterolytically. In comparison with the known chemistry of the
“parent” Lewis acid B(C6F5)3, reversibility of the H2 uptake
was achieved for CyB(C6F5)2 and PhC2H4B(C6F5)2. Based on
Gutmann’s method, the relative Lewis acidities of the fluorinated
o-C6F5), -149.9 (t, 2F, 3JFF = 20 Hz, p-C6F5), -162.2 (t, 4F, 3JFF
23 Hz, m-C6F5).
=
1
PhC2H4B(C6F5)2 5. Yield: 95%. H NMR (C6D6, 300 MHz,
3
298 K): d 7.12 (d, 2H, JHH = 6 Hz, Ph-H), 7.02 (m, 3H, Ph-H),
2.70 (t, 2H, 3JHH = 6 Hz, CH2), 2.22 ppm (t, 2H, 3JHH = 6 Hz, CH2).
19F NMR (C6D6, 282 MHz, 298 K): d -131.6 (d, 4F, 3JFF = 20 Hz,
boron Lewis acids are as follows: B(C6F5)3 > B(p-C6F4H)3
>
o-C6F5), -148.4 (t, 2F, 3JFF = 20 Hz, p-C6F5), -162.3 (t, 4F, 3JFF
23 Hz, m-C6F5).
=
PhC2H4B(C6F5)2 > CyB(C6F5)2. Changes in the Lewis bases
turned out to be of secondary importance for hydrogena-
tion/dehydrogenation reversibility. The main factor for reversible
activation of H2 was found to be the diminished Lewis acidity with
regard to B(C6F5)3 accomplished in this work by variation of one
of the boron substituents.
Lewis acidity tests according to the Gutmann–Beckett method
1
in CDCl3. 31P{ H} NMR: Ph3P O reference: d = 29.5.
(Ph3P O)B(C6F5)3 reference addcut: d = 45.5. Reference shift:
Dd = 16.0. (Ph3P O)B(Cy)(C6F5)2 reference adduct: d = 42.0. Ref-
erence shift: Dd = 12.5. Lewis-acidity relative to B(C6F5)3: 78.1%.
(Ph3P O)B(PhCH2CH2)(C6F5)2 reference adduct: d = 42.3. Ref-
erence shift: Dd = 12.8. Lewis-acidity relative to B(C6F5)3: 80.0%.
Et3P O reference: d = 52.8. (Et3P O)B(C6F5)3 reference adduct:
d = 76.3. Reference shift: Dd = 23.5. (Et3P O)B(Cy)(C6F5)2
reference adduct:d = 72.8. Reference shift: Dd = 20.0. Lewis-acidity
relative to B(C6F5)3: 85.1%. (Et3P O)B(PhCH2CH2)(C6F5)2 ref-
erence adduct: d = 74.4. Reference shift: Dd = 21.6. Lewis-
acidity relative to B(C6F5)3: 91.9%. Gutmann–Beckett in C6D6.
Experimental
General
All manipulations were carried out under an atmosphere of dry
nitrogen using standard Schlenk techniques or in a glovebox
(M. Braun 150B-G-II) filled with dry nitrogen. Solvents were
freshly distilled under N2 by employing standard procedures
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 1091–1097 | 1095
©

View Article Online
1
1
31P{ H} NMR: Ph3P O reference: d = 25.3. (Ph3P O)B(C6F5)3
4a). 31P{ H} NMR (C6D6, 121 MHz, 298 K): d 56.4 (4a), 52.2 (4b)
1
reference adduct:
d
=
45.6. Reference shift: Dd
=
20.3.
ppm. 11B{ H} NMR (C6D6, 96 MHz, 298 K): d -22.2 ppm (br).
(Ph3P O)B(Cy)(C6F5)2 reference adduct: d = 42.1. Reference
shift: Dd = 16.8. Lewis-acidity relative to B(C6F5)3: 82.8%.
(Ph3P O)B(PhCH2CH2)(C6F5)2 reference adduct: d = 42.3. Ref-
erence shift: Dd = 17.0. Lewis-acidity relative to B(C6F5)3: 83.7%.
Et3P O reference: d = 46.0. (Et3P O)B(C6F5)3 reference adduct:
d = 75.8. Reference shift: Dd = 29.8. (Et3P O)B(Cy)(C6F5)2
reference adduct:d = 72.5. Reference shift: Dd = 26.5. Lewis-acidity
relative to B(C6F5)3: 88.9%. (Et3P O)B(PhCH2CH2)(C6F5)2 ref-
erence adduct: d = 74.2. Reference shift: Dd = 28.2. Lewis-acidity
relative to B(C6F5)3: 94.6%.
Compound 6. PhCH2CH2B(C6F5)2 (0.09 g, 0.2 mmol) and
TMP (0.0283 g, 0.2 mmol). 6 was obtained as a white solid, yield
85%. Anal. Calcd. for C29H30BF10N: C, 58.70; H, 5.10; N, 2.36.
1
Found: C, 58.42; H, 5.02; N, 2.23. H NMR (C6D6, 300 MHz,
3
3
293 K): d 7.38 (d, 2H, JHH = 6 Hz, Ph–H), 7.18 (t, 2H, JHH
=
6 Hz, Ph–H), 7.03 (t, 1H, 3JHH = 6 Hz, Ph–H), 4.52 (br, 2H, NH),
2.85 (t, 2H, 3JHH = 6 Hz, CH2), 1.75 (br, 2H, CH2), 0.89 (m, 2H,
CH2), 0.78 (m, 2H, CH2), 0.69 (s, 12H, CH3). 19F NMR (C6D6,
282 MHz, 293 K): d -133.7 (d, 4F, 3JFF = 23 Hz, o-C6F5), -164.3
(t, 2F, 3JFF = 23 Hz, p-C6F5), -166.9 (t, 4F, 3JFF = 23 Hz, m-C6F5).
11B{ H} NMR (C6D6, 96 MHz, 293 K): d -18.8 ppm (br).
1
Preparation of [TMPH][CyBH(C6F5)2] 2, [PMPH][CyBH-
(C6F5)2] 3, [t-Bu3PH][CyBH(C6F5)2] 4, [TMPH][PhC2H4BH-
Compound 7. PhCH2CH2B(C6F5)2 (0.09 g, 0.2 mmol) and
PMP (0.034 g, 0.2 mmol). 7 was obtained as a white solid, yield
83%. Anal. Calcd. for C30H32BF10N: C, 59.32; H, 5.31; N, 2.31.
(C6F5)2]
6,
[PMPH][PhC2H4BH(C6F5)2]
7,
[t-Bu3PH]-
[PhC2H4BH(C6F5)2] 8. 2–4 and 6–8 were all prepared in
the same way. Their preparations are described in general form
(LA = Lewis acid. LB = Lewis base). Stoichiometric amounts
of LA (0.2 mmol) and LB (0.2 mmol) were dissolved in 5 mL
of toluene, giving a colorless solution. The reaction vessel
was filled with 1000 mbar of H2 and the solution was stirred
at room temperature for 1 h. The reaction mixture was then
concentrated to half of its volume, and hexane was added to
induce precipitation. The mixture was filtered, washed with
hexane and dried in vacuo.
1
Found: C, 59.53; H, 5.41; N, 2.33. H NMR (C6D6, 300 MHz,
3
3
293 K): d 7.42 (d, 2H, JHH = 6 Hz, Ph-H), 7.26 (t, 2H, JHH
=
6 Hz, Ph–H), 7.13 (t, 1H, 3JHH = 6 Hz, Ph–H), 5.49 (br, 1H, NH),
3
2.80 (t, 2H, JHH = 6 Hz, CH2), 1.77 (m, 2H, CH2), 1.61 (d, 3H,
3JHH = 6 Hz, N-CH3), 1.23 (t, 2H, 3JHH = 6 Hz, CH2), 0.79 (m, 4H,
CH2), 0.65 (s, 6H, CH3), 0.18 ppm (s, 6H, CH3). 19F NMR (C6D6,
282 MHz, 293 K): d -133.6 (d, 4F, 3JFF = 23 Hz, o-C6F5), -164.1
(t, 3F, 3JFF = 23 Hz, p-C6F5), -167.0 (t, 6F, 3JFF = 23 Hz, m-C6F5).
11B{ H} NMR (C6D6, 96 MHz, 293 K): d -18.6 ppm (br).
1
Compound 2. CyB(C6F5)2 (0.0856 g, 0.2 mmol) and TMP
(0.0283 g, 0.2 mmol). 2 was obtained as a white solid, yield 76%.
Anal. Calcd for C27H32BF10N: C, 56.76; H, 5.65; N, 2.45. Found: C,
56.51; H, 5.60; N, 2.50. 1H NMR (C6D6, 300 MHz, 298 K): d 4.74
(br, 2H, NH), 2.67 (br, 1H, BH), 2.03 (m, 4H, CH2), 1.70 (m, 4H,
CH2), 1.48 (m, 1H, CH), 1.18 (m, 2H, CH2), 0.72 (overlap, 18H,
CH2, CH3). 19F NMR (C6D6, 282 MHz, 298 K): d -132.7 (d, 4F,
3JFF = 21 Hz, o-C6F5), -164.6 (t, 2F, 3JFF = 20 Hz, p-C6F5), -167.1
Compounds 8a and 8b. PhCH2CH2B(C6F5)2 (0.09 g, 0.2 mmol)
and t-Bu3P (0.04 g, 0.2 mmol). 8a was obtained as a white solid,
yield 78%. Anal. Calcd. for C32H38BF10P: C, 58.73; H, 5.85. Found:
C, 58.59; H, 5.72. 1H NMR (C6D6, 300 MHz, 298 K): d 7.49, 7.25,
1
1
7.10 (Ph–H), 5.34 (d, JHP = 447 Hz, PH, 8b), 4.13 (d, JHp
=
432 Hz, PH, 8a), 3.00, 2.81, 1.88, 1.58, 1.22 (CH2), 0.74 (d, 27H,
3JH–P = 15 Hz P{(C(CH3)}3). 19F NMR (C6D6, 282 MHz, 298 K):
3
d -133.1 (overlap, o-C6F5), -165.6 (t, JF–F = 20 Hz, p-C6F5, 8b),
1
(t, 4F, 3JFF = 21 Hz, m-C6F5). 11B{ H} NMR (C6D6, 96 MHz, 298
3
3
-166.7 (t, JF–F = 20 Hz, p-C6F5, 8a), -167.7 (t, JF–F = 20 Hz,
m-C6F5, 8b), -168.3 (t, 3JF–F = 20 Hz, m-C6F5, 8a). 31P { H}NMR
1
K): d -15.5 ppm (br).
(C6D6, 121 MHz, 298 K): d 57.4 (8a), 53.5 (8b).11B{ H} NMR
1
Compound 3. CyB(C6F5)2 (0.0856 g, 0.2 mmol) and PMP
(0.034 g, 0.2 mmol). 3 was obtained as a white solid, yield 72%.
Anal. Calcd for C28H34BF10N: C, 57.45; H, 5.85; N, 2.39. Found:
C, 57.07; H, 5.47; N, 2.04. 1H NMR (C6D6, 300 MHz, 293 K): d
5.42 (br, 1H, NH), 2.70 (br, 1H, BH), 2.12 (m, 3H, Cy-H), 1.83 (d,
3H, 3JHH = 6 Hz, N-CH3), 1.45 (m, 2H, Cy-H), 1.26 (m, 6H, Cy-
H), 1.04 (m, 2H, CH2), 0.90 (m, 4H, CH2), 0.76 (s, 6H, CH3), 0.34
(s, 6H, CH3). 19F NMR (C6D6, 282 MHz, 293 K): d -132.5 (d, 4F,
3JFF = 23 Hz, o-C6F5), -164.9 (t, 2F, 3JFF = 23 Hz, p-C6F5), -167.3
(C6D6, 96 MHz, 298 K): d -18.5 ppm (br).
Notes and references
1 D. Sperling and M. A. Deluchi, Annu. Rev. Energy, 1989, 14, 375.
2 L. Schlapbach and A. Zu¨ttel, Nature, 2001, 414, 353.
3 H. Berke, ChemPhysChem, 2010, 11, 1837.
4 F. Cheng, H. Ma, Y. Li and J. Chen, Inorg. Chem., 2007, 46, 788.
5 R. J. Keaton, J. M. Blacquiere and R. T. Baker, J. Am. Chem. Soc.,
2007, 129, 1844.
6 M. C. Denney, V. Pons, T. J. Hebden, D. M. Heinekey and K. I.
Goldberg, J. Am. Chem. Soc., 2006, 128, 12048.
7 T. B. Marder, Angew. Chem., Int. Ed., 2007, 46, 8116.
8 F. H. Stephens, R. T. Baker, M. H. Matus, D. J. Grant and D. A. Dixon,
Angew. Chem., Int. Ed., 2007, 46, 746.
9 S. B. Kalidindi, M. Indirani and B. R. Jagirdar, Inorg. Chem., 2008, 47,
7424.
10 S. Aldridge and A. J. Downs, Chem. Rev., 2001, 101, 3305.
11 Z. L. Xiao, R. H. Hauge and J. L. Margrave, Inorg. Chem., 1993, 32,
642.
12 H. J. Himmel and J. Vollet, Organometallics, 2002, 21, 5972.
13 G. H. Spikes, J. C. Fettinger and P. P. Power, J. Am. Chem. Soc., 2005,
127, 12232.
1
(t, 4F, 3JFF = 23 Hz, m-C6F5). 11B{ H} NMR (C6D6, 96 MHz, 293
K): d -15.7 ppm (br).
Compounds 4a and 4b. CyB(C6F5)2 (0.0856 g, 0.2 mmol) and
t-Bu3P (0.04 g, 0.2 mmol). 4a and 4b were obtained as a white
solid, yield 74%. Anal. found: C, 56.77; H, 6.51. 1H NMR (C6D6,
300 MHz, 298 K): d 5.71 (d, 1JHP = 453 Hz, PH, 4b), 4.61 (d, 1JHP
=
438 Hz, PH, 4a), 3.16 (br, 1H, BH), 2.23 (m, 1H, CH), 1.97 (m,
4H, CH2), 1.76 (m, 2H, CH2), 1.52 (m, 2H, CH2), 1.34 (m, 2H,
CH2), 0.85 (d, 27H, 3JH–P = 15 Hz, P{(C(CH3)}3). 19F NMR (C6D6,
3
282 MHz, 298 K): d -132.1 (overlap, o-C6F5), -165.9 (t, JF–F
=
14 D. W. Stephan, Org. Biomol. Chem., 2008, 6, 1535.
15 D. W. Stephan, Dalton Trans., 2009, 3129.
16 D. W. Stephan and G. Erker, Angew. Chem. Int. Ed., 2010, 49, 46.
20 Hz, p-C6F5, 4b), -166.8 (t, 3JFF = 20 Hz, p-C6F5, 4a), -167.9 (t,
3JFF = 20 Hz, m-C6F5, 4b), -168.4 ppm (t, 3JFF = 20 Hz, m-C6F5,
1096 | Dalton Trans., 2011, 40, 1091–1097
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
17 G. C. Welch, R. R. S. Juan, J. D. Masuda and D. W. Stephan, Science,
2006, 314, 1124.
38 C. M. Mo¨mming, S. Fro¨mel, G. Kehr, R. Fro¨hlich, S. Grimme and G.
Erker, J. Am. Chem. Soc., 2009, 131, 12280.
39 C. Jiang, O. Blacque and H. Berke, Chem. Commun., 2009, 5518.
40 C. Jiang, O. Blacque and H. Berke,, Organometallics, 2009, 28, 5233.
41 U. Mayer, V. Gutmann and W. Gerger, Monatsh. Chem., 1975, 106,
1235.
18 G. C. Welch and D. W. Stephan,, J. Am. Chem. Soc., 2007, 129,
1880.
19 P. Spies, G. Erker, G. Kehr, K. Bergander, R. Fro¨hlich, S. Grimme and
D. W. Stephan, Chem. Commun., 2007, 5072.
20 J. S. J. McCahill, G. C. Welch and D. W. Stephan, Angew. Chem., Int.
Ed., 2007, 46, 4968.
21 P. A. Chase and D. W. Stephan, Angew. Chem., Int. Ed., 2008, 47, 7433.
22 D. Holschumacher, T. Bannenberg, C. G. Hrib, P. G. Jones and M.
Tamm, Angew. Chem., Int. Ed., 2008, 47, 7428.
23 V. Sumerin, F. Schulz, M. Nieger, M. Leskela¨, T. Repo and B. Rieger,
Angew. Chem., Int. Ed., 2008, 47, 5861.
24 D. P. Huber, G. Kehr, K. Bergander, R. Fro¨hlich, G. Erker, S. Tanino,
Y. Ohki and K. Tatsumi, Organometallics, 2008, 27, 5279.
25 S. J. Geier, T. M. Gilbert and D. W. Stephan, J. Am. Chem. Soc., 2008,
130, 12632.
26 V. Sumerin, F. Schulz, M. Atsumi, C. Wang, M. Nieger, M. Leskela¨, T.
Repo, P. Pyykko¨ and B. Rieger, J. Am. Chem. Soc., 2008, 130, 14117.
27 H. Wang, R. Fro¨hlich, G. Kehr and G. Erker, Chem. Commun., 2008,
5966.
42 V. Gutmann, Coord. Chem. Rev., 1976, 18, 225.
43 M. A. Beckett, D. S. Brassington, M. E. Light and M. B. Hursthouse,
J. Chem. Soc., Dalton Trans., 2001, 1768.
44 M. A. Beckett, D. S. Brassington, S. J. Coles and M. B. Hursthouse,
Inorg. Chem. Commun., 2000, 3, 530.
45 M. Ullrich, A. J. Lough and D. W. Stephan, Organometallics, 2010, 29,
3647.
46 Crystal data for 4b: Crystal system orthorhombic, space group Pbca,
3
˚
˚
a = 18.3337(3), b = 16.1172(3), c = 20.7593(3) A,V = 6134.12(18) A ,
Z = 8, m = 0.170 mm-1, 43845 reflections collected, 9356 independent
(Rint = 0.0455), R1 = 0.0427, wR2 = 0.0918 (for 5223 observed reflections
with I ≥ 2s (I) and 396 refined parameters). CCDC 793471.
47 C. Jiang, O. Blacque, T. Fox, H. Berke, unpublished results.
48 Crystal data for 6: Crystal system Triclinic, space group P¯ı, a =
˚
9.4346(2), b = 11.4068(3), c = 14.4546(3) A, a = 74.034(2), b = 87.133(2),
3
-1
˚
28 M. Ullrich, A. J. Lough and D. W. Stephan, J. Am. Chem. Soc., 2009,
g = 68.809(2), V = 1392.33(6) A , Z = 2, m = 0.128 mm , 28610
reflections collected, 7497 independent (Rint = 0.0257), R1 = 0.0562,
wR2 = 0.1720 (for 5114 observed reflections with I ≥ 2s(I) and 385
refined parameters). CCDC 793472.
131, 52.
29 P. Spies, G. Kehr, K. Bergander, B. Wibbeling, R. Fro¨hlich and G.
Erker, Dalton Trans., 2009, 1534.
30 A. Ramos, A. J. Lough and D. W. Stephan, Chem. Commun., 2009,
1118.
31 S. J. Geier and D. W. Stephan, J. Am. Chem. Soc., 2009, 131(10), 3476.
32 M. A. Dureen, G. C. Welch, T. M. Gilbert and D. W. Stephan, Inorg.
Chem., 2009, 48, 9910.
49 P. A. Chase, T. Jurca and D. W. Stephan, Chem. Commun., 2008, 1701.
50 D. J. Parks, W. E. Piers and G. P. A. Yap, Organometallics, 1998, 17,
5492.
51 D. J. Parks, R. E. v H. Spence and W. E. Piers, Angew. Chem., Int. Ed.
Engl., 1995, 34, 809.
33 M. A. Dureen and D. W. Stephan, J. Am. Chem. Soc., 2009, 131, 8396.
34 E. Otten, R. C. Neu and D. W. Stephan, J. Am. Chem. Soc., 2009, 131,
9918.
52 Oxford Diffraction (2007). Xcalibur CCD system. Oxford Diffraction
Ltd, Abingdon, Oxfordshire, England.
53 CrysAlisPro (Versions 1.171.32/33), Oxford Diffraction Ltd,
35 A. Ramos, A. J. Lough and D. W. Stephan, Chem. Commun., 2009,
Abingdon, Oxfordshire, England.
1118.
54 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
A64, 112–122.
55 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
56 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7–13.
36 M. Ullrich,K. S.-H. Seto, A. J. Lough and D. W. Stephan, Chem.
Commun., 2009, 2335.
37 S. J. Geier and D. W. Stephan, J. Am. Chem. Soc., 2009, 131, 3476.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 1091–1097 | 1097
©