Cationic Ti(IV) and neutral Ti(III) titanocene-phosphinoaryloxide frustrated Lewis pairs: Hydrogen activation and catalytic amine-borane dehydrogenation

Article abstract of DOI:10.1039/c2dt30168g

Titanium-phosphorus frustrated Lewis pairs (FLPs) based on titanocene-phosphinoaryloxide complexes have been synthesised. The cationic titanium(iv) complex [Cp₂TiOC₆H₄P( ^tBu)₂][B(C₆F_5<

Full text of DOI:10.1039/c2dt30168g

View Online / Journal Homepage / Table of Contents for this issue
This article is published as part of the Dalton Transactions themed
issue entitled:
Frustrated Lewis Pairs
Guest Editor: Douglas Stephan
Published in issue 30, 2012 of Dalton Transactions
Articles in this issue include:
Communication
Hydrogen activation by 2-boryl-N,N-dialkylanilines: a revision of Piers’ ansa-aminoborane
Konstantin Chernichenko, Martin Nieger, Markku Leskelä and Timo Repo
Paper
Frustrated Lewis pair addition to conjugated diynes: Formation of zwitterionic 1,2,3-butatriene
derivatives
Philipp Feldhaus, Birgitta Schirmer, Birgit Wibbeling, Constantin G. Daniliuc, Roland Fröhlich,
Stefan Grimme, Gerald Kehr and Gerhard Erker
Paper
Fixation of carbon dioxide and related small molecules by a bifunctional frustrated pyrazolylborane
Lewis pair
Eileen Theuergarten, Janin Schlösser, Danny Schlüns, Matthias Freytag, Constantin G. Daniliuc,
Peter G. Jones and Matthias Tamm
Visit the Dalton Transactions website for more cutting-edge inorganic chemistry research

Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 9067
www.rsc.org/dalton
PAPER
Cationic Ti(IV) and neutral Ti(III) titanocene–phosphinoaryloxide frustrated
Lewis pairs: hydrogen activation and catalytic amine-borane
dehydrogenation†
Andy M. Chapman and Duncan F. Wass*
Received 23rd January 2012, Accepted 16th March 2012
DOI: 10.1039/c2dt30168g
Titanium–phosphorus frustrated Lewis pairs (FLPs) based on titanocene–phosphinoaryloxide complexes
have been synthesised. The cationic titanium(IV) complex [Cp2TiOC6H4P(tBu)2][B(C6F5)4] 2 reacts with
hydrogen to yield the reduced titanium(III) complex [Cp2TiOC6H4PH(tBu)2][B(C6F5)4] 5. The
titanium(III)–phosphorus FLP [Cp2TiOC6H4P(tBu)2] 6 has been synthesised either by chemical reduction
of [Cp2Ti(Cl)OC6H4P(tBu)2] 1 with [CoCp*2] or by reaction of [Cp2Ti{N(SiMe3)2}] with
2-C6H4(OH){P(tBu)2}. Both 2 and 6 catalyse the dehydrogenation of Me2HN·BH3.
activation offers an exciting extension of FLP chemistry with
new possibilities for exploitation in activation pathways and reac-
Introduction
Solution phase combinations of sterically hindered Lewis acid–
Lewis base pairs, so-called Frustrated Lewis Pairs (FLPs), have
been the subject of recent interest, particularly because of the
high latent reactivity of such species in the activation of small
molecules. Initial studies have focused on the reversible hetero-
lytic cleavage of dihydrogen, which offers the promise of metal-
free catalytic hydrogenation.1–4 However, the diversity of the
reactions reported is now large and continues to grow.5 The
bulky phosphine and fluorinated borane systems (such as PtBu3/
B(C6F5)3) pioneered by Stephan have been modified so that the
specific reactivity of FLP systems can be controlled by subtle
steric and electronic alterations to either the Lewis acidic or
basic components.6,7 A great deal of work has also focused on
extending the range of main group FLPs to other main group
Lewis acids (e.g. simple alkyl boranes, alanes8,9 allenes10) or
bases (e.g. amines11 carbenes12 and sulfides13). Linking the two
components of the FLP into a single amphoteric molecule has
also led to interesting results.14,15
tivity patterns.
During our study of cationic Zr and Hf metallocene–phosphi-
noaryloxy complexes we have noted that both the nature of the
metal and ancillary ligands have a dramatic effect on the reactiv-
ity of such complexes.20 We were keen to complete the series by
examining the reactivity of the analogous Ti complex but were
limited by the synthetic challenge of obtaining pure samples of
such complexes. Adopting a modified synthesis based on chlo-
ride abstraction from the parent compound [Cp2Ti{OC6H4P-
(tBu)2}Cl] 1 with [Ph3Si][B(C6F5)4] has finally allowed us to
access this complex 2 (Scheme 1), as we reported recently.21
With this compound now in hand, we report here our initial
exploration of its FLP-type reactivity and, in particular, facile
reduction chemistry to Ti(III).
We have been exploring the chemistry of cationic group 4
(Zr and Hf) metallocene–phosphinoaryloxide complexes as ana-
logues of linked main group FLPs where the Lewis acidic
borane component is replaced with an electrophilic transition
metal centre. Our initial results have established the analogy
with main group FLPs16,17 but also demonstrated additional reac-
tivity, for example the catalytic dehydrogenation of amine-
boranes,17 a reaction only achieved in a stoichiometric sense
with main group FLP systems.18,19 It is our view that combining
the ability of transition metal complexes in catalysis with the
capability of FLPs to activate substrate molecules via ditopic
Scheme 1 Synthesis of 2 (ref. 21).
Results and discussion
Hydrogen activation with [Cp2TiOC6H4P(tBu)2][B(C6F5)4] 2
School of Chemistry, University of Bristol, Cantock’s Close, Bristol,
BS8 1TS, UK. E-mail: duncan.wass@bristol.ac.uk
†CCDC 864986–864987. For crystallographic data in CIF or other elec-
tronic format see DOI: 10.1039/c2dt30168g
The archetypal reaction of FLPs is arguably the heterolytic clea-
vage of dihydrogen and we have already reported analogous
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 9067–9072 | 9067

Scheme 2 Attempted reaction of H2 with 2, 3 and 4, and suggested
mechanism for the reduction of 2 by H2 to give 5. Reagents and con-
ditions: PhCl, 1–3 bar H2, 25 °C, 20 h.
Fig. 1 Molecular structure of 5. The borate anion and all hydrogen
atoms (except P–H) have been omitted for clarity. Thermal ellipsoids are
drawn at the 50% probability level. Selected bond lengths (Å): Ti1–O1
1.941(2), C16–P1 1.773(2), P1–C17 1.846(3), C21–P1 1.850(3), Cp–
Ti1 2.039. Angles (°): Ti1–O1–C11 141.7(1), C16–P1–H25 107(1),
H25–P1–C17 104(1), H25–P1–C21 103(1), C16–P1–C17 110.0(1),
C21–P1–C16 111.8(1), C21–P1–C17 119.8(1), Cp–Ti1–Cp 138.54.
reactivity when our Zr and Hf complexes have at least one Cp*
ligand. By contrast, the Cp derivatives 3 and 4 do not appear to
react with H2 under the conditions tested to date (1–3 bar, −30
to 50 °C, weeks). Given this apparent inertness of 3 and 4, we
were surprised to find that 2 reacts smoothly with even low
pressures of H2 over the course of 20 hours (Scheme 2). During
the course of the reaction, the intense purple colour of 2 slowly
1
faded to a light brown colour and the H and 31P{1H} NMR
indicated complete conversion of 2 to a new species. However,
the complete loss of the Cp resonances and broadening of the
1
aromatic signals in the H NMR spectrum for this new species
suggest it is a paramagnetic Ti(III) complex, observations which
are also consistent with the genuine Ti(III) species 6 (vide infra)
and other related compounds.22,23 A weak signal in the 31P{1H}
NMR spectrum at δ 22.5 ppm is indicative of a P–H centre.
The product was crystallized and unambiguously determined by
X-ray crystallography to be 5 (Fig. 1).
Clearly, 2 is reduced under the reaction conditions to 5. It is
possible that the reduction occurs via an initial heterolytic activa-
tion of H2, followed by Ti–H bond homolysis to give a hydrogen
radical (and ultimately hydrogen after dimerization)24 and a cat-
ionic Ti(III) species. The formation of [Cp2TiH2]2 by the hydro-
genation of [Cp2TiMe2] provides a related example in which a
one-electron reduction of a Ti(IV) species is affected by
hydrogen.25
Scheme 3 Synthesis of 6. Reagents and conditions: (a)
1 eq.
[CoCp*2], benzene, 25 °C, 20 h. (b) 2 eq. KHMDS, THF, 25 °C, 20 h.
(c) 1 eq. t-Bu2P(C6H4)OH, hexanes, 25 °C, 30 min. (d) 0.98 eq.
[(C5H4Me)2Fe][B(C6F5)4], PhCl, 25 °C, 5 min.
Synthetic routes to neutral [Cp2TiOC6H4P(tBu)2] 6
The mixed alkoxy/chloride precursor 1 provides a convenient,
highly soluble precursor for the synthesis of the Ti(III) compound
6. Treatment of 1 with a stoichiometric quantity of [CoCp*2]
resulted in clean reduction, evidenced by the recovery of almost
exactly one equivalent of the expected by-product [CoCp*2][Cl]
(Scheme 3). The reaction is reversible, so that treatment of 6
with 1 eq [(C5H4Me)2Fe][B(C6F5)4] gave quantitative conver-
sion to the Ti(IV) species 2. This reduction procedure was free of
side reactions but is a rather uneconomical route to 6 because of
the expense of [CoCp*2]. An alternative, and potentially more
general, synthesis of 6 was realised by protonolysis of [Cp2Ti(N-
(TMS)2)] by t-Bu2P(C6H4)OH. This procedure provided a con-
venient method to prepare 6 in high yield and excellent purity.
Deep purple single crystals suitable for X-ray diffraction were
Despite being less electrophilic, the group 4 M(III) compounds
(of the type [Cp2MX]) may still be considered as Lewis acids.26
For example, intramolecular hydroaminations that are typical of
the highly Lewis acidic group 4 M(IV) complexes are still poss-
ible with Ti(III) compounds.27 Furthermore, it was anticipated the
presence of a d-electron would facilitate the binding of π-accep-
tor ligands such as CO, hydrogen, ethene, and even nitrogen,
opening new reaction pathways for FLP chemistry. Ti(III) com-
pounds are mildly reducing and it is conceivable that a reduction
may accompany an FLP-type activation of a small molecule with
these compounds. The unexpected reduction with H2 described
in the previous section led us to investigate reliable routes to
neutral Ti(III) titanocene–phosphinoaryloxy complexes.
9068 | Dalton Trans., 2012, 41, 9067–9072
This journal is © The Royal Society of Chemistry 2012

obtained from hexane at low temperature and the structure
confirmed (Fig. 2). The X-ray structure shows the monomeric
structure of 6 and the presence of a very long Ti–P interaction
(2.907(2) Å) exists in the solid state.28 The presence of the 5-
membered ring in 6 creates an unusually acute Ti–O–C bend
angle compared to the simple alkoxy-analogues A22 and B23
(Fig. 3).
In addition, 6 contains a rather long Ti–O bond in comparison
to these other two compounds. It has been suggested30,31 that the
extent of Ti ← O donation is proportional to the polarity of the
Ti–O bond, and hence the electron donor capability of the
O-atom. Importantly, these studies concluded that the Ti ← O
donation is largely unaffected by the Ti–O–C angle, although a
coincidental correlation is observed in the complexes presented
here. This may explain why the Ti–O bond is longer in 6
compared with A, but not why it is longer than in B, where the
electron donor capabilities are expected to be very similar. The
long Ti–O bond is best explained by the presence of the Ti–P
interaction in 6. The bonding picture in complexes of the type
[Cp2ML] and [Cp2ML2] has been studied in some detail,29 and
the MO pictures30,31 of these complexes help to explain these
observations. Accommodation of the extra pair of electrons from
the Ti–P bond in 6 is expected to result in population of the b2
orbital, reducing its availability to participate in a stabilizing Ti
← O π-bonding interaction. The net result is an elongated Ti–O
bond and a geometry that is very similar to the M(IV) analogues
(Table 1). The absence of paramagnetic broadening of the 31P
resonance in the 31P{1H} NMR spectra of 6 suggests that this
interaction is not present in solution. This was later confirmed by
an ESR experiment in which no 31P–47Ti or 31P–49Ti fine or
hyper-fine coupling was observed.32
Catalytic dehydrocoupling of Me2NHBH3 by 2 and 6
We have already reported that transition metal FLPs are active
catalysts for the dehydrocoupling of amine-boranes, in contrast
to main group systems which only mediate this reaction in a stoi-
chiometric sense.16 Our investigations into the mechanism of
this catalysis suggest that these systems are operating via an
FLP-type mechanism, in contrast to the superficially related
[Cp2M(II)] (M = Ti, Zr and Hf) systems reported by Manners
et al.33 and Chirik et al.34 In both these literature examples, Ti is
the most active metal, with little or no activity observed with the
analogous Zr(IV) and Hf(IV) systems. These trends have been
rationalized in terms of the relative stability of +2 and +4 oxi-
dation states of the group 4 metallocenes and thus the relative
accessibility of an M(II)/M(IV) redox manifold during cata-
lysis.33,35 Both 2 and 6 were found to be active catalysts (entries
3 and 4, Table 2), albeit with substantially decreased activity
Fig. 2 Molecular structure of 6. All hydrogen atoms have been omitted
for clarity. Thermal ellipsoids are drawn at the 50% probability level.
For selected bonds lengths and angles see Table 1.
Table 2 Comparison of the catalytic activities for the dehydrocoupling
of Me2NHBH3 to cyclo-[Me2NBH2]2
Entry
Compound
Loading/mol%
Solvent
TOFa/h−1
1
2
3
4
3
4
2
6
2
2
2
10
PhCl
PhCl
PhCl
C6D6
384
42
12
Fig. 3 Schematic representation of the three important angles in
[Cp2ML2] complexes (left) and examples of other [Cp2TiOR] com-
plexes, A and B.22,23 The two angles, α1 and α2, refer to the angles
made between the ligand, M centre and the imaginary normal that
bisects the two Cp rings in a staggered conformation. The 2a1 frontier
orbital is calculated to extend in this direction.29
0.8
a Calculated from monitoring of conversion by NMR spectroscopy at
23 °C.
Table 1 Comparison of some relevant structural parameters for 2, 3, 4, 6, A and B
Compound
d(M–O)/Å
d(Cp–M)/Åa
M–O–C/°
α1/°
α2/°
θ/°
φ/°
6
2.028(6)
1.810(3)
1.892(2)
1.901(5)
1.9972(1)
1.989(2)
2.07
124.3(5)
175.4(11)
142.3(2)
122.5(4)
124.51(1)
121.7(2)
41.7
0b
27.66
n/a
n/a
28.0
27.8
28.5
136.3
69.4(2)
c
c
c
A
B
2
3
4
c
2.36
2.06
2.20
2.19
0b
135.5
129.3
128.2
128.1
43.5
42.4
42.6
71.5(1)
70.18(4)
71.16(5)
a Mean of the two M–Cp centroid distances. b Estimated from the available X-ray structural data. c Data not available.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 9067–9072 | 9069

relative to both the Zr and Hf systems reported previously by us
(compounds 3 and 4, entries 1 and 2), and indeed those of
Manners33 and Chirik.34
compounds were recorded using sealable J-Youngs tap NMR
tubes. Microanalysis was carried out by the Microanalytical Lab-
oratory, University of Bristol using a Carlo Elba spectrometer.
X-ray diffraction experiments were carried out at 100 K on a
Bruker Apex II Kappa CCD diffractometer using Mo Kα radi-
ation (λ = 0.71073 Å). Spectral data for the [B(C6F5)4] anion are
reported separately:
This striking decrease in activity between 3 or 4 and 2 was
unexpected. If the Lewis acidity of the metal centre is key to the
catalytic activity, then it is not clear why the Ti compound 2
should be so much less active than the Zr and Hf analogues 3
and 4.36 Although bond dissociation energies of M–L bonds are
generally lowest for Ti compared to Zr and Hf, during the many
attempts at preparing 2 it was noted that the Ti–C bonds
appeared to undergo protonolysis far less readily than in either
the Zr of Hf compounds. Consequently, the slow rate may be a
factor of an increasing stable Ti–H bond that would be present in
the cationic intermediate [Cp2Ti(H)OC6H4PHtBu2]+ that is
postulated to form upon dehydrogenation of Me2NHBH3.16
Since the Ti–H bond is expected to be thermodynamically less
stable than either Zr–H or Hf–H bonds, it is presumed that the
stability is kinetic in origin and is a product of the smaller size
of Ti. When following the catalysis using in situ 11B{1H} NMR
spectroscopy, the resonance for free Me2NHBH3 was extremely
broad in the presence of 2.37 In addition, upon mixing the
reagents, an initial colour change from deep purple/black to light
brown was accompanied by a very brief period of vigorous effer-
vescence. One explanation is that the Me2NHBH3 is able to
reduce the Ti(IV) centre in 2 to a Ti(III) compound, akin to the
reduction of 2 to 5 by hydrogenation. Amounts of [Me2NBH2]2
consistent with the stoichiometric reduction of the catalyst were
observed upon acquiring the NMR immediately afterward.
Despite this evidence, the possible formation of 6 as the catalyti-
cally active species under the reaction conditions does not
account for the observation that 2 exhibits a TOF about 15 times
that of authentic 6. This Ti(III) compound shows low but repeata-
ble and catalytic activity (10 turnovers). It is interesting to note
that other Ti(III) sources such as [Cp2TiCl] and TiCl3 are com-
pletely inactive for this reaction.33 Once again, this highlights
and importance difference between these FLP-type compounds
and those that depend on metal-centered redox events.
1
13C {1H} (100 Hz, DCM-d2): δ 150.1 (dm, JCF = 244 Hz,
o-B(C6F5)4), 139.7 (dm, 1JCF = 246 Hz, p-B(C6F5)3), 137.8 (dm,
1JCF = 247 Hz, m-B(C6F5)4), 126.3 (m, ipso-B(C6F5)4). 11B{1H}
(96 Hz, DCM-d2): δ = 17.8 (s).
19F NMR (376 Hz, DCM-d2): δ = −129.6 (o-B(C6F5)4),
−162.0 (p-B(C6F5)4), −165.8 (m-B(C6F5)4).
[(C5H4Me)2Fe][B(C6F5)4]
A Schlenk flask was charged with [CPh3][B(C6F5)4] (461 mg,
0.5 mmol), which was dissolved in fluorobenzene. To this, a
fluorobenzene solution of [(C5H4Me)2Fe] (109 mg, 0.51 mmol)
was added dropwise with stirring, causing a colour change to
deep emerald green. Once the addition was complete, the sol-
ution was left to stir for 30 min then layered with hexane
(10 mL). After standing overnight, the light green supernatant
was removed via cannula and the green crystals washed with
several portions of hexane before drying in vacuo for 2 hours.
Yield: 402 mg, 0.45 mmol, 90%. Elemental analysis: calc. C,
48.41; H, 1.58; found C, 48.53; H, 1.70.
[Cp2TiCl]2
This compound was prepared by a modification to the published
procedure:41 Cp2TiCl2 (3 g, 12.05 mmol) and Al powder (0.8 g,
30 mmol) were weighed out into a Schlenk flask fitted with a
condenser and filter stick. The apparatus was connected to a
Schlenk line and THF (ca. 10 mL) was added by vacuum trans-
fer from Na/benzophenone at −78 °C. The solution was allowed
to warm to room temperature, then refluxed for 1 hour, during
which time the red/silver suspension became a dark green sol-
ution. Stirring was continued for a further 20 hours before the
solution was filtered and the filtrate concentrated to dryness.
The resulting dark green powder was suspended in ether
(20 mL), collected on a frit and washed with several portions of
ether (3 × 10 mL) until the colour of the washings had faded to a
light olive green. The olive green powder was dried in vacuo for
1 hour. Yield: 2.01 g, 9.4 mmol, 78%. Satisfactory elemental
analysis could not be obtained and the crude material was used
in the next step.
Experimental section
Unless otherwise stated, all manipulations were carried out under
an inert atmosphere of argon using standard Schlenk-line and
glovebox (M-Braun, O2 < 0.1 ppm, H2O < 0.1 ppm) techniques,
and all glassware was oven dried (200 °C) overnight and allowed
t
to cool under vacuum prior to use. The reagents Bu2P(C6H4)
OH38 and [CPh3][B(C6F5)4]39 were prepared according to the lit-
erature. Solvents were purified and pre-dried using an Anhydrous
Engineering column purification system40 then vacuum trans-
ferred from the appropriate drying agent (K/benzophenone aro-
matics, ethers, CaH2 for hydrocarbons and chlorinated solvents)
prior to use. NMR spectra were recorded using a JEOL ECP 300
spectrometer at 300 MHz, and Varian 400 and 500 spectrometers
at 400 and 500 MHz, respectively, (using the appropriate deuter-
ated solvent, purchased from Cambridge Isotope Labs or Sigma-
Aldrich and purified by vacuum transfer from the appropriate
desiccant) and referenced to an internal standard (residual
solvent signal for 1H, BF3·OEt2 for 11B, 85% H3PO4 for 31P, and
FCCl3 for 19F. NMR. Spectra of air and moisture sensitive
[Cp2Ti(N(TMS)2)]
[Cp2TiCl]2 (2.13 g, 5.0 mmol) and KN(TMS)2 (1.99 g,
10.0 mmol) were weighed out into a swivel frit assembly. The
apparatus was connected to a Schlenk line and THF (ca. 10 mL)
was added by vacuum transfer from Na/benzophenone at
−78 °C. The solution was allowed to warm to room temperature,
during which time the green suspension turned black/brown with
a grey precipitate. The suspension was stirred for a further
24 hours then concentrated to dryness. Pentane (ca. 15 mL) was
9070 | Dalton Trans., 2012, 41, 9067–9072
This journal is © The Royal Society of Chemistry 2012

distilled onto the solids from Na/benzophenone/tetraglyme at
0 °C. The resulting brown suspension was stirred for 30 min at
ambient temperature then filtered and the filtrate concentrated to
dryness leaving a dark brown oil. The residue was sublimed at
90 °C, 10−2 Torr as a deep brown/black microcrystalline solid.
Yield (crude): 1.51 g, 4.45 mmol, 89%. 1H NMR (300 Hz,
toluene-d8): δ 0.10 (br s, C(CH3)3). Satisfactory elemental analy-
sis could not be obtained even after repeated sublimation. The
material obtained after one sublimation was of sufficient purity
for use in the next stage.
the walls of the flask. The suspension was filtered and the solids
washed with benzene. The yellow solids that remained were
identified as essentially pure [CoCp*2][Cl] by 1H NMR
(MeCN-d3, δ 2.02 ppm, s, C5(CH3)5) and amounted to 84% of
the expected mass. Solvent was removed from the filtrate under
reduced pressure and the resulting purple microcrystalline solid 6
was washed with several small portions of hexane (1 mL) then
dried in vacuo. Yield: 85 mg, 0.205 mmol, 41%.
Method B. Crude [Cp2Ti(N(TMS)2)] (169.1 mg, 0.5 mmol)
t
and Bu2P(C6H4)OH (119.2 mg, 0.5 mmol) were mixed in
hexane and stirred for 10 min, resulting in the precipitation of
purple microcrystals of 6. The solution was allowed to stand at
−20 °C overnight to induce further precipitation. The super-
natant was removed via cannula and the solids dried in vacuo.
Yield: 127.8 mg, 0.305 mmol, 61%. 1H NMR (300 Hz,
benzene-d6): No aromatic (including C5H5) signals could be
unambiguously assigned due to paramagnetic line broadening.
δ 1.48 (br. S, C(CH3)3). 31P{1H} NMR (161 Hz, benzene-d6):
δ 12.8 (s). ESI-MS: 415.17 [M]. Elemental analysis: calc. C,
69.90; H, 7.77; found. C, 69.91; H, 7.78.
[Cp2Ti{OC6H4P(tBu)2}][B(C6F5)4] 2
Equimolar quantities of 6 (0.02 mmol) and [(C5H4Me)2Fe]-
[B(C6F5)4] were weighed into vials and each dissolved in PhCl
(ca. 1 mL). The two solutions were mixed, resulting in a dark
brown/black solution. The solution was layered with hexanes
and allowed to stand overnight, precipitating large black/purple
needles of 2. Yield: 0.0198 mmol, 99%. 1H NMR (500 Hz,
PhCl–benzene-d6, 5 : 1): HC3 HC4 and HC6 aromatic signals
are obscured by PhCl signals and could be unambiguously
Crystallographic details for 5 and 6 can be found in Table 3.
3
4
identified. δ 6.27 (dd, 1H, JHH = 8.2 Hz, JHH = 4.4, HC5),
3
5.87 (s, 5H, C5H5), 0.99 (d, JHP = 13.6 Hz, 18H, C(CH3)3).13C
{1H} (125 Hz, PhCl–benzene-d6, 5 : 1): δ 170.6 (d, JCP = 15.6
2
Representative procedure for the catalytic dehydrocoupling of
Me2NHBH3
2
3
Hz, C1), 133.2 (d, JCP = 2.0 Hz, C5), 132.6 (d, JCP = 2.5 Hz,
1
3
C4), 124.2 (d, JCP = 28.4 Hz, C2), 123.0 (d, JCP = 4.4 Hz,
C6), 116.6 (s, C5H5), 115.6 (d, 2JCP = 4.9 Hz, C3), 39.4 (d, 1JCP
Compound 2 or 6 and Me2NHBH3 (10–50 eq.) were each
weighed into an NMR tube as finely ground powders. The NMR
tube was connected to a Schlenk line and evacuated. The tube
was immersed in liquid nitrogen to a depth of 5 cm and 0.7 mL
PhF (2) or benzene-d6 (6) was added from a calibrated bulb via
vacuum transfer from CaH2. The tube was sealed and immedi-
ately placed in a dry ice/acetone bath and stored until required.
Immediately prior to loading into an NMR spectrometer
(300 MHz), the sample was inverted and thawed so that the
solids were not in contact with the solvent. The solvent was
allowed to warm to ambient then the tube inverted and shaken
2
= 3.4 Hz, C(CH3)3), 30.2 (d, JCP = 3.9 Hz, C(CH3)3). 31P{1H}
(161 Hz, PhCl–benzene-d6, 5 : 1): δ 70.2 (s). ESI-MS: 415.17
[M]. Elemental analysis: calc. C, 52.68; H, 2.95; found. C,
52.52; H, 3.27.
[Cp2Ti{OC6H4PH(tBu)2}][B(C6F5)4] 5
2 (139.0 mg, 0.2 mmol) was loaded into an NMR tube fitted
with a Teflon needle valve and dissolved in PhCl and benzene-d6
(ca. 0.5 and 0.2 mL, respectively) giving a deep purple solution.
The tube was connected to a Schlenk line and subjected to three
freeze–pump–thaw degassing cycles then backfilled with 2 bar
H2 at room temperature via liquid nitrogen trap. No conversion
of 2 was detected after ca. 3 hours. After standing overnight, the
solution had changed colour to orange/brown. The solution was
layered with hexanes (ca. 6 mL), precipitating brown/purple
plates which were collected and dried in vacuo. Yield: 130 mg,
94%. 1H NMR (500 Hz PhCl–toluene-d8, 5 : 1): No aromatic
signals (including C5H5) could be unambiguously identified due
to paramagnetic broadening; δ 0.9 (br. s, C(CH3)3). 31P{1H}
NMR (161 Hz, PhCl–toluene-d8, 5 : 1): δ 26.8 (s). Satisfactory
elemental analysis could not be obtained.
Table 3 Crystallographic details for 5 and 6
Compound
5
6
Chemical formula
Formula mass
Crystal system
a/Å
b/Å
c/Å
α/°
C48H33BF20OPTi C24H32OPTi
1095.42
Orthorhombic
18.5031(8)
17.8620(8)
27.4342(11)
90.00
415.37
Monoclinic
9.599(2)
10.528(3)
10.456(3)
90.00
β/°
γ/°
90.00
90.00
9067.1(7)
100(2)
Pbca
96.260(6)
90.00
1050.4(4)
100(2)
P21
Unit cell volume/Å3
Temperature/K
Space group
[Cp2Ti{OC6H4P(tBu)2}] 6
No. of formula units per unit cell, Z
No. of reflections measured
No. of independent reflections
Rint
Final R1 values (I > 2σ(I))
Final wR(F2) values (I > 2σ(I))
Final R1 values (all data)
Final wR(F2) values (all data)
Flack parameter
8
2
153 693
10 740
0.1055
0.0436
0.0933
0.0766
0.1092
—
12 230
3587
Method A. 1 (633.0 mg, 1.41 mmol) and [CoCp*2]
(463.1 mg, 1.41 mmol) were combined in a Schlenk flask.
Benzene (8 mL) was vacuum transferred onto the solids at 0 °C.
Upon thawing of the solvent, a black suspension was formed,
which was stirred for 24 hours. During this time, a yellow pre-
cipitate formed along with a small quantity of metallic mirror on
0.0432
0.0377
0.1000
0.0380
0.1004
0.005(5)
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 9067–9072 | 9071

12 P. A. Chase, A. L. Gille, T. M. Gilbert and D. W. Stephan, Dalton Trans.,
2009, 7179.
thoroughly to insure dissolution of all of the solids. The tube
was quickly loaded into the NMR spectrometer and an appro-
priate series of 11B{1H}NMR spectra acquired until complete
conversion was achieved.
13 C. A. Tanur and D. W. Stephan, Organometallics, 2011, 30, 3652.
14 V. Sumerin, F. Schulz, M. Atsumi, C. Wang, M. Nieger, M. Leskela,
T. Repo, P. Pyykkö and B. Rieger, J. Am. Chem. Soc., 2008, 130, 14117.
15 P. Spies, S. Schwendemann, S. Lange, G. Kehr, R. Fröhlich and G. Erker,
Angew. Chem., Int. Ed., 2008, 47, 7543.
16 A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am. Chem. Soc.,
2011, 133, 8826.
Conclusions
17 A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am. Chem. Soc.,
2011, 133, 18463.
We have synthesised titanium-phosphorus frustrated Lewis pairs
based on titanocene–phosphinoaryloxide complexes. The facile
reduction of the cation titanium(IV) complex [Cp2TiOC6H4P-
(tBu)2][B(C6F5)4] with hydrogen led us to synthesise the tita-
nium(III)-phosphorus FLP [Cp2TiOC6H4P(tBu)2]. Both of these
complexes catalyse the dehydrogenation of Me2HN·BH3, albeit
with lower activity than their zirconium(IV) or hafnium(IV)
counterparts. The facile reduction of these titanium(IV) species
by hydrogen suggests possibilities in replacing more expensive
stoichiometric reducing agents in catalytic cycles where such
one electron processes depend on this oxidation state manifold.42
We also believe the isolation of reduced d1 metal-containing
frustrated Lewis pairs may open yet new possibilities for small
molecule activation in cases where metal backbonding is impor-
tant for substrate binding and activation.
18 A. J. M. Miller and J. E. Bercaw, Chem. Commun., 2010, 46, 1709.
19 G. R. Whittell, E. I. Balmond, A. P. M. Robertson, S. K. Patra,
M. F. Haddow and I. Manners, Eur. J. Inorg. Chem., 2010, 3967.
20 In this context, it is worth noting that related complexes of the zircono-
alkenylphosphine family have been reported for ethene polymnerisation
catalysis: P. Hao, S. Zhang, J. Yi and W.-H. Sun, J. Mol. Catal. A:
Chem., 2009, 302, 1.
21 A. M. Chapman, M. F. Haddow and D. F. Wass, Eur. J. Inorg. Chem.,
2012, 1546.
22 K. Matsubara, S. Niibayashi and H. Nagashima, Organometallics, 2003,
22, 1376.
23 B. Cetinkaya, P. B. Hitchcock, M. F. Lappert, S. Torroni, J. L. Atwood,
W. E. Hunter and M. J. Zaworotko, J. Organomet. Chem., 1980, 188, 31.
24 Similar radical elimination processes are known to occur when [Cp2MR2]
complexes (M = Zr and Ti) are oxidised: R. F. Jordan, R. E. LaPointe,
C. S. Bajgur, S. F. Echols and R. Willett, J. Am. Chem. Soc., 1987, 109,
4111.
25 Recent Advances in Hydride Chemistry, ed. M. Peruzzini, R. Poli,
M. Peruzzini and R. Poli, Elsevier Science, 2002.
26 R. S. P. Coutts, P. C. Wailes and R. L. Martin, J. Organomet. Chem.,
1973, 47, 375.
Acknowledgements
27 T. Janssen, R. Severin, M. Diekmann, M. Friedemann, D. Haase,
W. Saak, S. Doye and R. Beckhaus, Organometallics, 2010, 29, 1806.
28 Compare to 2.624(3) Å in a related titanium cyclopentadienyl phosphino-
aryloxide compound: C. A. Willoughby, R. R. Duff, Jr., W. M. Davis and
S. L. Buchwald, Organometallics, 1996, 15, 472.
29 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98, 1729.
30 R. Gyepes, V. Varga, M. Horacek, J. Kubistas, J. Pinkass and K. Mach,
Organometallics, 2009, 28, 1748.
The University of Bristol is thanked for funding.
References
1 G. C. Welch, R. R. San Juan, J. D. Masuda and D. W. Stephan, Science,
2006, 314, 1124.
2 D. W. Stephan and M. Ullrich, J. Am. Chem. Soc., 2009, 131, 52.
3 V. Sumerin, F. Schulz, M. Atsumi, C. Wang, M. Nieger, M. Leskel,
T. Repo, P. Pyykkö and B. Rieger, J. Am. Chem. Soc., 2008, 130, 14117.
4 J. M. Farrell, Z. M. Heiden and D. W. Stephan, Organometallics, 2011,
30, 4497.
5 D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2010, 49, 46;
A. J. P. Cardenas, B. J. Culotta, T. H. Warren, S. Grimme, A. Stute,
R. Fröhlich, G. Kehr and G. Erker, Angew. Chem., Int. Ed., 2011, DOI:
10.1002/anie.201101622; Z. M. Heiden and D. W. Stephan, Chem.
Commun., 2011, 47, 5729; X. Zhao and D. W. Stephan, J. Am. Chem.
Soc., 2011, 133, 12448.
31 R. Gyepes, V. Varga, M. Horacek, J. Kubistas, J. Pinkass and K. Mach,
Organometallics, 2010, 29, 3780.
32 E. Samuel, J. F. Harrod, D. Gourier, Y. Dromzee, F. Robert and
Y. Jeannin, Inorg. Chem., 1992, 31, 3252.
33 M. E. Sloan, A. Staubitz, T. J. Clark, C. A. Russell, G. C. Lloyd-Jones
and I. Manners, J. Am. Chem. Soc., 2010, 132, 3831.
34 D. Pun, E. Lobkovsky and P. J. Chirik, Chem. Commun., 2007, 3297.
35 Y. Luo and K. Ohno, Organometallics, 2007, 26, 3597.
36 Although not definitive, the Gutmann Lewis Acceptor Numbers (AN) of
2 (86.3) was found to be very similar to its Zr (78.9) and Hf (93.7)
homologues.
6 A. J. M. Miller, J. A. Labinger and J. E. Bercaw, J. Am. Chem. Soc.,
2010, 132, 3301.
7 G. Ménard and D. W. Stephan, J. Am. Chem. Soc., 2010, 132, 1796.
8 Y. Zhang, G. M. Miyake and E. Y. X. Chen, Angew. Chem., Int. Ed.,
2010, 49, 10158.
9 C. Appelt, H. Westenberg, F. Bertini, A. W. Ehlers, J. C. Slootweg,
K. Lammertsma and W. Uhl, Angew. Chem., Int. Ed., 2011, 50, 3925.
10 B. Inés, S. Holle, R. Goddard and M. Alcarazo, Angew. Chem., Int. Ed.,
2010, 49, 8389.
37 The observed line broadening in the free Me2NHBH3 resonance could be
a facet of fast exchange of free and coordinated Me2NHBH3 to the Ti
centre on the NMR timescale, especially if the Ti centre is paramagnetic.
38 M. Bornand, S. Torker and P. Chen, Organometallics, 2007, 26, 3585.
39 C. Wang, G. Erker, G. Kehr, K. Wedeking and R. Frohlich, Organometal-
lics, 2005, 24, 4760.
40 A. B. Pangborn and M. A. Giardello, Organometallics, 1996, 15, 1518.
41 R. S. P. Coutts, P. C. Wailes and R. L. Martin, J. Organomet. Chem.,
1973, 47, 375.
11 S. J. Geier and D. W. Stephan, J. Am. Chem. Soc., 2009, 131, 3476.
42 J. Streuff, Chem.–Eur. J., 2011, 17, 5507.
9072 | Dalton Trans., 2012, 41, 9067–9072
This journal is © The Royal Society of Chemistry 2012