COMMUNICATION
Facile dihydrogen release from phosphino-borinate ester Lewis pairs†
Andy M. Chapman, Mairi F. Haddow, Jonathan P. H. Orton and Duncan F. Wass*
Received 2nd February 2010, Accepted 21st May 2010
First published as an Advance Article on the web 14th June 2010
DOI: 10.1039/c0dt00513d
New phosphino-borinate ester Lewis pairs of the type
tBu2PCH2C(R)2OB(C6F5)2 (R
=
Me or CF3) are syn-
thesised from the corresponding phosphino-alcohol and
HB(C6F5)2. Dihydrogen release from the zwitterionic
tBu2PHCH2C(R)2OBH(C6F5)2 is facile.
Solution phase combinations of sterically hindered (‘frustrated’)
Lewis acid–Lewis base pairs have been the subject of much recent
interest, not least because the high latent reactivity of such species
in, for example, the reversible heterolytic cleavage of hydrogen.
Stephan and co-workers have demonstrated that the precise bal-
ance of sterics and electronics between simple tertiary phosphine–
borane frustrated pairs plays a pivotal role in the reversibility of
this particular reaction. The same group showed that the P(Mes)3–
B(C6F5)3 pair reacts cleanly with hydrogen at room temperature
and the resulting ion-pair is stable at temperatures up to 150 ◦C.1
In a subsequent study,2 the combination of P(o-tol)3 with the
modified borane B(p-C6F4H)3 was also shown to cleave hydrogen
at room temperature but the resulting ion-pair decomposed slowly
under ambient conditions over nine days, releasing hydrogen and
reforming the starting materials. In the context of hydrogen storage
using such materials, facile release of dihydrogen is clearly crucial
and yet in all other examples,3–5 release is very slow and requires
heating to 60 ◦C or more. In most cases, a reduction in Lewis
acidity relative to B(C6F5)3 seems to be beneficial in creating a
reversible system.‡
In related work, Britovsek and co-workers6 synthesised
(C6F5)2B(OC6F5) and (C6F5)B(OC6F5)2 in order to compare the
effect that successive substitution of a -C6F5 group with -OC6F5
has on the Lewis acidity. Interestingly, they found that the
experimental Lewis acidity depended on which NMR method
was employed (Guttman7 or Childs8); B(OC6F5)3 appears to be a
stronger Lewis acid than B(C6F5)3 using Gutmann’s methodology
but a much weaker Lewis acid when using the method of Childs.
They concluded that substitution of a -C6F5 groups for -OC6F5
makes the boron centre a progressively harder Lewis acid. With
this in mind, we sought about designing and investigating the
reactivity of novel frustrated Lewis pairs featuring a borinate ester
as the Lewis acid component, reasoning that the harder Lewis acid
in such species would give rise to modified reactivity and a method
to tune dihydrogen activation and release.
Scheme 1 Synthesis of 1a–1d.
Compound 1c was prepared by reaction of the preformed lithium
phosphino-alkoxide 2a with ClB(C6F5)2 or by reaction of 1a with
HB(C6F5)2. The second route proved more practical since the only
by-product in an open system is hydrogen; removal of solvent
gave 1c as a highly moisture and oxygen sensitive, colourless oil in
analytical purity.
The 31P { H} NMR spectrum of 1c exhibits a sharp singlet at
1
24.5 ppm suggesting little or no P–B interaction. The absence of
this interaction is further supported by the broad singlet at 35 ppm
in the 11B { H} and characteristic pentafluorophenyl resonances
1
in 4 : 2 : 4 ratio, with Dd (m-F, p-F) separation of 11 ppm, typical
of trigonal planar boron and in good agreement with the related
borinate ester n-BuOB(C6F5)2.10 The VT NMR spectru◦m of 1c
reveals a shift in the 31P resonance upon cooling to -80 C from
24.5 to 48 ppm along with a simultaneous decrease in the Dd (m-F,
p-F) separation. This observation is consistent with the formation
of a P–B interaction at lower temperatures. Preliminary DFT
calculations reveal that this is best explained by an intermolecular
process rather than intramolecular; concentration studies are
underway to verify this.
We originally postulated that 1c may form via loss of dihydrogen
from the zwitterionic product 1d. However, no NMR signals
attributable to 1d were observed unless the reaction is performed in
a sealed system. In such instances, analytically pure 1d precipitates
from the reaction medium in varying quantities, but is not
observed in the supernatant. A further observed product is
[H2B(C6F5)2]-, suggesting dihydrogen activation by the Lewis pair
formed between HB(C6F5)2 and the phosphine moiety of 1a. The
isolable tBu3P–BH(C6F5)2 adduct and its subsequent reaction with
1 bar hydrogen at 80 ◦C has been previously reported.11
Phosphino-alcohol synthons (1a,b) may be synthesised via
the facile and highly selective ring opening of 2,2-disubstituted
oxarines using a modified literature procedure (Scheme 1).9
School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8
1TS, UK. E-mail: duncan.wass@bristol.ac.uk
† Electronic supplementary information (ESI) available: Experimental
details. CCDC reference numbers 764587 and 764588. For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/c0dt00513d
6184 | Dalton Trans., 2010, 39, 6184–6186
This journal is
The Royal Society of Chemistry 2010
©
Chapman, Andy M.
