Tris(pyrazolyl)borate amidoborane complexes of the group 4 metals

Article abstract of DOI:10.1039/c1dt10709g

Treatment of the tris(pyrazolyl)borate metal triamides Tp′M(NMe ₂)₃, where Tp′ = (C₃H₃N ₂)₃BH (Tp) or (3,5-Me₂C₃HN ₂)₃BH (Tp*) and M = Ti, Zr and Hf, with the Bronsted acidic Lewis adduct (C₆F₅) ₃B·NH₃ in toluene solution leads to the formation of Tp′M(NMe₂)₂{NH₂B(C₆F ₅)₃} complexes. The exception to this was the attempted preparation of Tp*Ti(NMe₂)₂{NH₂B(C ₆F₅)₃} which was unsuccessful. Where Tp′ = Tp and M = Ti and Zr and where Tp′ = Tp* and M = Zr the complexes have been characterized by single crystal X-ray diffraction methods, revealing the first examples of octahedral amidoborane complexes of the group 4 metals. Attempts to drive the reactions to completion resulted in competing preferential hydrolysis of the amidoborane group, regenerating (C₆F ₅)₃B·NH₃.

Full text of DOI:10.1039/c1dt10709g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 7434
www.rsc.org/dalton
PAPER
Tris(pyrazolyl)borate amidoborane complexes of the group 4 metals†
Anna-Marie Fuller, David L. Hughes and Simon J. Lancaster*
Received 19th April 2011, Accepted 24th May 2011
DOI: 10.1039/c1dt10709g
Treatment of the tris(pyrazolyl)borate metal triamides Tp¢M(NMe2)3, where Tp¢ = (C3H3N2)3BH (Tp)
or (3,5-Me2C3HN2)3BH (Tp*) and M = Ti, Zr and Hf, with the Brønsted acidic Lewis adduct
(C6F5)3B·NH3 in toluene solution leads to the formation of Tp¢M(NMe2)2{NH2B(C6F5)3} complexes.
The exception to this was the attempted preparation of Tp*Ti(NMe2)2{NH2B(C6F5)3} which was
unsuccessful. Where Tp¢ = Tp and M = Ti and Zr and where Tp¢ = Tp* and M = Zr the complexes have
been characterized by single crystal X-ray diffraction methods, revealing the first examples of
octahedral amidoborane complexes of the group 4 metals. Attempts to drive the reactions to completion
resulted in competing preferential hydrolysis of the amidoborane group, regenerating (C6F5)3B·NH3.
Introduction
Interest in amineboranes as molecular hydrogen storage materials
has resulted in intense recent investigation of amidoborane
complexes of early transition metals as models for intermediates
in catalytic dehydrogenation.1 Furthermore, the isoelectronic
relationship between amidoborane and alkyl ligands has prompted
studies of their structure and bonding.2 Complexes of amidobo-
ranes are also believed to be intermediates in the technologically-
important production of metal nitride materials.3
We have shown that the elimination of amine in reactions
between metal amides and the Brønsted acidic ammonia adduct
of tris(pentafluorophenyl)borane, (C6F5)3B·NH3, (1) provides a
route to complexes of the amidoborane ligand [(C6F5)3BNH2]-.4–6
For example, treatment of Ti(NR2)4 (R = Me, Et) with 1 results
in replacement of one dialkylamide ligand with an amidoborane
ligand to give Ti(NR2)3{NH2B(C6F5)3} (I, Scheme 1), with the
Scheme 1
subsequent evolution of the corresponding dialkylamine; I is a
tetrahedral complex containing an NH2B(C6F5)3 ligand. However,
Zr(NMe2)4 reacts with 1 to produce the five-coordinate ami-
doborane complex Zr(NMe2)3(NHMe2){NH2B(C6F5)3} (II) with
dimethylamine being retained in the metal coordination sphere.4,5
Under certain circumstances, the remaining protic hydro-
gens on nitrogen are lost leading to the formation of an-
ionic trigonal bipyramidal coordination complexes, such as
III, of the formally trianionic nitridoborane ligand.7 The
tris(pentafluorophenyl)borane serves as a protecting group for the
terminal nitride functionality. Current investigations are exploring
both the mechanism of the formation of these nitridoborane
complexes and their subsequent derivitization.
Substitution of one of the amide ligands in complexes such as I
and II for a monoanionic facially coordinating tridentate ligand is
of interest since it should lead to the first examples of octahedral
group 4 metal amidoborane complexes. It may also provide
a favourable steric environment for a nitridoborane complex,
providing complementary shielding for the metal centre. The
utilization of a sterically-demanding polydentate ligand has been
demonstrated to be effective in restricting, but not eliminating,
oligomerization of the metal nitrogen multiply bonded species.8
The tris(pyrazolyl)borate ligands (Tp¢), are versatile facially
binding tridentate donors,9 in which the substituents on the
pyrazolyl rings allow for tuning of the steric and electronic
properties.10,11 The most common of these is the hydridotris(3,5-
dimethylpyrazolyl)borate anion (Tp*).10 Studies of group 4
Tp¢ complexes have encompassed chloro,12,13 alkyl,14 amido,15
carbonyl,16 hydrazido,17 and imido ligands,18 while derivatives have
been used in catalysis.19
Wolfson Materials and Catalysis Centre, School of Chemistry, University
of East Anglia, Norwich, NR4 7TJ, UK. E-mail: S.Lancaster@uea.ac.uk;
Fax: +44 1603 592003; Tel: +44 1603 592009
† Electronic supplementary information (ESI) available. CCDC reference
numbers 823223–823226. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c1dt10709g
7434 | Dalton Trans., 2011, 40, 7434–7441
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Scheme 2
Group 4 monocyclopentadienyl and metallocene amidoborane
complexes of the form CpnMR3 - n{NH2B(C6F5)3}, (n = 1, R =
NMe2, NEt2, M = Ti, Zr; n = 2, R = Me, M = Zr, Hf) have
been previously described.5 Group 4 Tp amidoborane complexes
are unprecedented in the literature.20 Herein the reactions of
Tp¢M(NMe2)3 with (C6F5)3B·NH3 are described and the structure
and reactivity of the resulting products reported.
case the 19F NMR spectrum displayed signals for both 1 and 2. On
the basis of these spectroscopic data, compound 2 was formulated
as TpTi(NMe2)2{NH2B(C6F5)3}.
Attempts to isolate an analytically pure sample of 2 through
fractional crystallization were frustrated by the similarity in solu-
bility, across a range of solvents, between both starting materials
and the products and as a result elemental analysis data were not
obtained. A mixture of red and yellow crystals, which were not
readily separated, was obtained from a solution containing an
excess of TpTi(NMe2)3 in light petroleum. X-ray analysis revealed
the red crystals to be the starting material TpTi(NMe2)3 which
has previously been structurally characterized15 and the yellow
crystals to be 2 (Fig. 1).
Results and Discussion
The substituent-free tris(pyrazolyl)borate triamide TpTi(NMe2)3
was treated with one equivalent of 1 in toluene solution (Scheme 2).
Monitoring the reaction by 1H and 19F NMR suggested incomplete
conversion of the reactants to a new compound 2. Sparging the
reaction mixture with nitrogen did not improve conversion and
lead to suspected partial hydrolysis to {TpTi(NMe2)2}2{m-O} (3)
and this observation is discussed further in the supplementary
Compound 2 exhibits a distorted octahedral geometry. The
bond to the pyrazolyl ring of the Tp ligand trans to the
amidoborane ligand (Ti(1)–N(1)) is significantly shorter than the
other two Ti–N(Tp) bonds lengths (Table 1). This may be due to a
greater trans influence of the amide ligands versus the amidoborane
ligand. In the solid state TpTi(NMe2)3 is distorted such that one
of the Ti–N(Tp) bonds is longer than the other two.15 The two
longer Ti–N(Tp) bonds in 2 are comparable to the shorter of the
bond lengths observed in TpTi(NMe2)3. The amidoborane Ti–N
information.† If an excess of TpTi(NMe2)3 was used a simple 19
F
NMR spectrum was obtained, with one set of signals indicating
equivalence of all three C6F5 rings; the o-F signal was found at
higher frequency (d -132.7 ppm) from that observed for 1, while
the p-F and m-F signals appeared at lower frequencies than the
starting material (d -159.3 ppm and -164.3 ppm, respectively).
The 11B NMR displayed a broad resonance at -3.36 ppm
corresponding to the Tp boron atom and a singlet at -7.7 ppm for
the (C6F5)3B group. Whilst the Tp boron chemical shift displays
minimal change, the (C6F5)3B boron signal is indicative of a
reaction and appears at significantly lower frequency to that of
˚
bond (2.254(2) A) is slightly longer than that observed for similar
˚
titanium amidoborane complexes (2.125(2)–2.170(2) A) with the
Ti–N(dimethylamido) bond lengths lying within the expected
range.4,5 Whilst the B–N bond length is unremarkable the Ti–N–B
bond angle, at 139.44(12)◦, is significantly larger than previously
observed, even for the analogous CpTi(NMe2)2{NH2B(C6F5)3}
complex (133.76(10)◦).5 The opening of the bond angle and
the extension of the Ti–N bond are presumably due to steric
interactions with the other ligands. Compound 2 is the first
structurally characterized titanium amidoborane to exhibit a
definitively octahedral geometry.21 The hydrogen atoms of the
amidoborane ligand participate in intramolecular hydrogen bonds
1
1. As one would expect, the H NMR displayed a mixture of
signals corresponding to both the starting material and 2.
1
Simplification of the H NMR spectrum with respect to the
Tp and NMe2 signals was achieved through the addition of
an excess of 1 (the NH3 signal from 1 was still visible in the
1H NMR spectrum). The three pyrazolyl rings are no longer
equivalent in the Cs symmetric product and the ring trans to the
amidoborane has distinctly different spectroscopic characteristics.
A pair of doublets, in the ratio 2 : 1, corresponding to the protons
in the C5 positions are observed and a pair of triplet signals at
significantly lower frequency are assigned to the protons attached
to the C4 positions. There is also a doublet signal integrating
to 2H at d 7.52 ppm belonging to the proton on C3 for two
of the pyrazolyl rings; the third C3 hydrogen atom signal was
coincident with the solvent peak, (Fig. S1†). The NH2 group
gave a characteristic broad signal at d 3.18 ppm similar to
those previously observed.4,5 The resonance for the dimethylamide
ligands was found at d 3.07 ppm, shifted to very slightly lower
frequency relative to the starting material at d 3.13 ppm. In this
◦
˚
Table 1 Selected bond lengths (A) and angles ( ); estimated standard
deviations are in parentheses
2
4·CH2Cl2
6·pentane
M(1)–N(1)
2.163(2)
2.254(2)
2.276(2)
1.903(2)
1.900(2)
2.215(2)
1.596(3)
1.659
2.299(2)
2.396(2)
2.421(2)
2.037(2)
2.017(2)
2.347(2)
1.596(3)
1.656
2.308(7)
2.356(6)
2.362(7)
2.052(6)
2.004(7)
2.386(7)
1.596(12)
1.661
M(1)–N(11)
M(1)–N(21)
M(1)–N(31)
M(1)–N(41)
M(1)–N(51)
N(51)–B(52)
Mean B(52)–C
M(1)–N(51)–B(52)
139.44(12)
135.41(14)
145.1(5)
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7434–7441 | 7435
©

View Article Online
Fig. 1 ORTEP of 2 with displacement ellipsoids drawn at the 50% probability level. Tp and methyl hydrogen atoms have been omitted for clarity.
to o-F atoms, one of which is a bifurcated intramolecular hydrogen
bond to two o-F atoms, while the second hydrogen displays a
somewhat longer intermolecular contact (2.41(2) A) to an o-
F atom of the third C6F5 ring. Similar interactions have been
described for the solid state structure of 1 and other amine
adducts22,23 and intramolecular hydrogen bonding interactions are
invariably present in related amidoborane complexes.24
The zirconium analogue of 2, TpZr(NMe2)2{NH2B(C6F5)3} (4)
was prepared through the reaction of TpZr(NMe2)3 and 1 in
toluene solution, following a similar procedure to that employed
for 2. Here too following the reaction by 1H and 19F NMR
spectroscopy indicated incomplete conversion. Reaction of 1 with
a slight excess of TpZr(NMe2)3 was sufficient for the reaction to
reach completion with respect to 1, yielding a clean 19F NMR
spectrum. For zirconium the solubility differences were such that
pure 4 could be obtained as a yellow solid simply by washing with
light petroleum.
peaks against the Tp signals confirmed the substitution of only
one dimethylamide ligand. Treatment with a further equivalent of
1 at room temperature did not result in the substitution of a second
dimethylamide ligand. The 11B NMR spectrum consisted of a
broad signal assigned to the Tp boron atom, and a sharp singlet at
lower frequency corresponding to the amidoborane boron atom.
Compound 4 was successfully crystallized from both light
petroleum and dichloromethane/light petroleum solutions. X-ray
diffraction studies revealed that both structures were in the space
group P21/n and the complexes are essentially isostructural. The
minor differences in the unit cell parameters were a consequence
of the solvent of crystallization and small variations in the solid
state structure of 4, Fig. 2. The results described here are those
for the dichloromethane solvate the refinement of which was
marginally better than that for the hexane solvate. The Tp ligand
coordinates to one face of a distorted octahedral zirconium centre.
The pyrazolyl ring trans to the amidoborane ligand displays
a significantly shorter Zr–N bond length than the other two
pyrazolyl rings, as observed for 2. The Zr–N bond lengths to
the dimethylamide ligands are slightly shorter than observed in
the starting material15 but are consistent with those observed
for Zr(NMe2)3(HNMe2){NH2B(C6F5)3}.4 The bond lengths found
for 4·CH2Cl2 are very similar to those observed for the anal-
ogous complex CpZr(NMe2)2{NH2B(C6F5)3}; the Zr(1)–N(51)
˚
The 19F NMR spectrum of 4 gave one set of resonances
indicating that the C6F5 groups are equivalent, which is consistent
with unhindered rotation about the B–N bond. The chemical
shift difference between the para and meta signals (D(dp-F - dm-F))
for the product was observed to be smaller and the peaks at
lower frequency than in 1; this is consistent with 2 and related
amidoborane complexes and reflects greater anionic character.4,5
1
˚
The H NMR spectrum includes pairs of doublets at d 7.75 and
(amidoborate) bond length, at 2.347(2) A in 4·CH2Cl2, is only
˚
7.74 (C5) and 7.70 and 7.39 (C3), and a pair of triplet signals at
d 6.28 and 6.25 ppm (C4). For all three pairs the signals were
in a 2 : 1 ratio, consistent with Cs symmetry in solution. The
dimethylamide methyl signal was observed at d 2.70 ppm, with
the NH2 of the amidoborane at d 3.06 ppm. Integration of these
slightly longer than in the Cp analogue (2.3053(13) A). The M–
N–B bond angle is smaller in 4·CH2Cl2 than that observed for 2
which reflects the reduction in steric crowding at the larger metal
centre. However, the Zr–N–B bond angle remains larger than the
corresponding angle in the Cp complex.4,5
7436 | Dalton Trans., 2011, 40, 7434–7441
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 2 ORTEP of 4·CH2Cl2: the displacement ellipsoids are drawn at the 40% probability level. Hydrogen atoms not engaged in hydrogen bonding and
the dichloromethane molecule have been omitted for clarity.
Table 2 Selected torsion angles (◦), estimated standard deviations are in
parentheses
One of the NH2 hydrogen atoms is involved in a single
intramolecular hydrogen bond to an o-F atom whilst the other
hydrogen atom forms a bifurcated hydrogen bond to two o-F
2
4·CH2Cl2
6·pentane
atoms, one on each of the other two C6F5 rings (Table S1†).
The most obvious difference between the solid state structures
of 2 and 4 is the orientation of the (C6F5)3B moiety with respect to
the Tp ligand, as illustrated by the N(dimethylamide)–M–N(51)–
B(52) torsion angles (Table 2). This is more easily demonstrated by
representing the structures as Newman style projections, viewed
along the N(51)–M bond, (Fig. 3). It should be noted that while
the nitrogen is four-coordinate, the metal centre is six-coordinate,
with approximately 90◦ angles, thus there are seven groups in
the Newman projection of the metal complex that need to be
considered. The conformation in 4 appears to minimize the steric
interactions between the two bulky ligands but it does leave one of
the NH2 hydrogen atoms, H(51a), in an eclipsed configuration with
N(11). In 2, the Ti–NH2 bond has been rotated by approximately
90◦ with respect to the dimethylamide ligands versus 4 and as
a result H(51a) now eclipses N(21). This would suggest that the
N(31)–M(1)–N(51)–B(52)
N(41)–M(1)–N(51)–B(52)
141.6(2)
40.3(2)
54.0(2)
-51.1(2)
-89.2(9)
179.5(9)
orientation of the (C6F5)3B group with respect to the Tp ligand is
a result of the subtle influence of lattice packing forces rather than
solely intramolecular steric interactions between groups bound to
the metal and amido nitrogen.
Short intramolecular C–H ◊ ◊ ◊ F contacts are present in both 2
and 4 and must play a major role in the orientation of the (C6F5)3B
group with respect to the Tp ligand. In 4·CH2Cl2 there is one
˚
intramolecular short H ◊ ◊ ◊ F contact (2.22(3) A) from a pyrazolyl
C3 hydrogen atom H(15) to an o-F atom F(62). In 2 a similar
˚
short H ◊ ◊ ◊ F interaction, C(15)–H(15) ◊ ◊ ◊ F(72) at 2.36(2) A, is
observed.25
The adduct 1 reacts with TpHf(NMe2)3 and, as observed for 4,
complete conversion to TpHf(NMe2)2{NH2B(C6F5)3} (5) required
the addition of a slight excess of TpHf(NMe2)3. The reaction was
monitored by 1H and 19F NMR and, once complete, the volatiles
were removed under reduced pressure and the resulting solid was
washed with light petroleum to afford the pure product as a
colourless solid. The 1H, 19F and 11B NMR data of 5 are in close
agreement with those of 4 suggesting the solution state structure
of 5 is essentially the same as that of 4.
Following the method employed for the preparation of 2,
1 was treated with one equivalent of Tp*Ti(NMe2)3. In this
case the 1H, 11B and 19F NMR spectra of the crude reaction
Fig. 3 Newman style projections looking down the N(51)–M(1) bond.
The Tp and Tp* nitrogen atoms cis to the NH2 group are indicated in red
but have not been labelled for clarity.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7434–7441 | 7437
©

View Article Online
Scheme 3
mixture indicated a complex mixture of different products, which
proved intractable. In light of the selective, if incomplete, reaction
to form 2 and successful preparation of the Tp* zirconium
complex described below it seems probable that the generation
of Tp*Ti(NMe2)2{NH2B(C6F5)3} is disfavoured on steric grounds
leading to a number of possible competing reaction pathways.26
at slightly higher frequency (d 2.88 ppm) than observed for 4
(d 2.70 ppm). The 19F NMR spectrum of 6 displays one set of
three signals: the o-F signal is in close agreement with 4, while
the p-F and m-F signals are found at slightly lower frequencies.
The 11B NMR displays a singlet peak belonging to the (C6F5)3B
boron atom at d -7.9 while the BH peak is very broad, almost
disappearing into the baseline at d -9.2 ppm; the BH coupling
could not be resolved.
The solid state structure of 6 was elucidated by X-ray diffraction
methods (Fig. 4) and contains a disordered solvent molecule
(believed to be n-pentane) in the crystal lattice. As observed for
the Tp analogues, the zirconium centre in 6 exhibits a distorted
octahedral geometry with the expected facial Tp* coordination
geometry. The Zr–N bond lengths to the Tp* ligand are all shorter
than observed in the starting material15 and display the now
familiar pattern with the Zr–N bond length to the pyrazolyl ring
Tp*Zr(NMe2)3 and
1
react to form Tp*Zr(NMe2)2-
{NH2B(C6F5)3} (6) (Scheme 3) and, as found with the substituent-
free tris(pyrazolyl)borate systems 4 and 5, the use of a small
excess of Tp*Zr(NMe2)3 facilitates the isolation of analytically
1
pure 6 as yellow crystals. The H NMR spectra of the product
showed two singlets in the ratio 2 : 1 at d 5.68 and 5.75 ppm,
which corresponded to the CH group at C4 of the pyrazolyl ring.
Two pairs of singlets in the 2 : 1 ratio were observed at d 2.39
and 2.33 ppm and d 2.11 and 1.93 ppm corresponding to the
methyl signals from the pyrazolyl rings of the Cs symmetric Tp*
complex, the signal at higher frequency presumably being due
to the methyl group on C3. The signal for the methyl groups of
the dimethylamide ligand in 6 was observed as a sharp singlet
˚
trans to the amidoborate ligand (2.308(7) A) being significantly
shorter than the other two bonds to the Tp* ligand. The bonding
parameters for the dimethylamide ligands are consistent with those
Fig. 4 ORTEP of 6 drawn at the 40% probability level. The disordered solvent molecule and hydrogen atoms not bonded to nitrogen have been omitted
for clarity.
7438 | Dalton Trans., 2011, 40, 7434–7441
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
of the zirconium amidoboranes previously reported.4,5 The Zr(1)–
under vacuum and the product extracted with light petroleum.
Cooling the solution to -25 ◦C yielded a mixed crop of red crystals
of the titanium starting material and yellow crystals of the product
TpTi(NMe2)2{NH2B(C6F5)3}, 2. Due to the separation problems
discussed above satisfactory elemental analysis results could not
be obtained. dH/ppm (300.1 MHz, CDCl3) 7.74 (1H, d, JHH 1.8,
CH, C5), 7.70 (2H, d, JHH 2.1, CH, C5), 7.52 (2H, d, JHH 2.0, CH,
C3), 6.24 (1H, t, JHH 2.2, CH, C4), 6.19 (2H, t, JHH 2.2, CH, C4),
3.18 (2H, br, NH2), 3.07 (12H, s, NCH3). dC/ppm (75.5 MHz,
CDCl3) 141.4 (CH), 139.8 (CH), 135.4 (CH), 135.1(CH), 105.8
(CH), 105.5 (CH), 49.4 (CH3). dB/ppm (96.3 MHz, CDCl3) -3.36
(1B, br, BH), -7.69 (s, (C6F5)3B). dF/ppm (282.4 MHz, CDCl3)
-132.7 (6F, d, 3JFF 21, o-F), -159.3 (3F, t, JFF 21, p-F), -164.3 (6F,
m, m-F).
˚
N(51) bond (2.386(7) A) in 6 is slightly longer than observed for
4, presumably as a result of increased steric hindrance at the metal
centre. This also accounts for the Zr–N–B bond angle at 145.1(5)◦
being far larger than those observed previously. The N–B bond
length is consistent with those other complexes reported herein.
The solid state structure of 6 displays a similar hydrogen
bonding pattern to that observed for 2, Table S1.† One of the
NH2 hydrogen atoms participates in a bifurcated intramolecular
hydrogen bond to two o-F atoms while the other makes a rather
longer contact to an o-F atom of the third C6F5 ring.27 The
(C6F5)3B group eclipses one of the 3,5-dimethylpyrazolyl rings
(Fig. 3), which is surprising considering the increased steric bulk
of the Tp* ligand and the apparent absence of any notable
intramolecular interactions.28
{HB(C3H3N2)3Ti(NMe2)2}2{l-O} (3)
Conclusion
During an attempt to repeat the preparation of 2, following the
general procedure outline above but using dichloromethane as the
solvent, adventitious water was introduced to yield the m-O species
and regenerate (C6F5)3B·NH3. Elemental analysis found: C 43.6,
H 6.1, N 31.3. Calculated for C26H44B2N16OTi2: C 43.7, H 6.2, N
31.4. dH/ppm (300.1 MHz, CDCl3) 7.74 (4H, m, CH), 7.57 (2H, d,
The triamides TpM(NMe2)3 (M = Ti, Zr and Hf) react with
(C6F5)3B·NH3 in toluene to form tris(pyrazolyl)borate amidobo-
rane complexes of the form TpM(NMe2)2{NH2B(C6F5)3}. Struc-
tural analysis of the titanium and zirconium complexes revealed
the expected distorted octahedral geometries.
The chemistry proceeds similarly for the Tp* complex of
zirconium and the solid state structure of 6 is similar to the Tp
complex, 4, with the differences being attributable to the greater
steric demands of the Tp* vs. Tp ligands. In contrast, the analogous
Tp*Ti complex proved inaccessible and attempted synthesis led
only to a complex product mixture.
J
HH 2.2, CH), 6.27 (1H, t, JHH 2.2, CH), 6.11 (2H, t, JHH 2.1, CH),
3.42 (12H, s, NCH3). dC/ppm (75.5 MHz, CDCl3) 142.3 (CH),
140.9 (CH), 134.6 (CH), 134.0 (CH), 105.0 (CH), 104.5 (CH),
50.33 (CH3). dB/ppm (96.3 MHz, CDCl3) -3.88 (d, JBH 112).
HB(C3H3N2)3Zr(NMe2)2{NH2B(C6F5)3} (4)
In summary, steric limitations notwithstanding, Tp¢M(NMe2)2-
{NH2B(C6F5)3} complexes are both accessible and relatively
stable. We have seen no evidence for further amine elimination
and the formation of imido or nitridoborane complexes.
A combination of TpZr(NMe2)3 (5.10 g, 11.7 mmol) and
(C6F5)3B·NH3 (5.69 g, 10.8 mmol) was dissolved in toluene (180
cm3) and nitrogen was bubbled through the solution for four hours.
The volatiles were removed under reduced pressure and the crude
product obtained as a foam which was washed with light petroleum
to yield a yellow solid (6.87 g, 6.9 mmol, 64%). X-ray quality
crystals were grown by cooling the light petroleum washings to
Experimental
◦
All reactions were carried out under a dry nitrogen atmosphere
using standard Schlenk techniques in pre-dried glassware. Solvents
were dried using an appropriate drying agent and distilled
under nitrogen prior to use: dichloromethane (CaH2), light
petroleum (Na/K alloy or sodium/dyglyme/benzophenone) and
toluene (sodium). Samples for NMR analysis were prepared
-25 C (4·hexane), or by recrystallization of the product from a
dichloromethane/light petroleum solution at -25 ◦C (4·CH2Cl2).
Elemental analysis found: C 40.3, H 2.7, N 13.6. Calculated for
C31H24B2F15N9Zr: C 40.5, H 2.6, N 13.7. dH/ppm (300.1 MHz,
CDCl3) 7.75 (2H, d, JHH 2.2 CH), 7.74 (1H, d, JHH 2.1, CH), 7.70
(2H, d, JHH 2.0, CH), 7.39 (1H, d, JHH 1.8, CH), 6.28 (2H, t, JHH
2.1, CH), 6.25 (1H, t, JHH 2.1, CH) 3.06 (2H, br, NH2) 2.70 (12H, s,
CH3). dC/ppm (75.5 MHz, CDCl3) 142.9 (CH), 141.2 (CH), 136.5
(CH), 136.2 CH), 106.0 (CH), 105.9 (CH), 43.00(CH3). dB/ppm
(96.3 MHz, CDCl3) -3.13 (br, BH), -8.24 ((C6F5)3B). dF/ppm
(282.4 MHz, CDCl3) -133.3 (6F, m, o-F), -159.0 (3F, t, JFF 20,
p-F), -164.1 (6F, m, m-F).
˚
using degassed deuterated solvents dried over activated 4 A
molecular sieves. NMR spectra were obtained using a Bruker
Avance DPX300 spectrometer at 22 ◦C, J values are given
in Hz. Chemical shifts are reported in ppm and referenced
1
to residual solvent resonances (1H, 13C{ H}); 19F is relative to
CFCl3; 11B is relative to Et2O·BF3. Elemental analyses were
carried out at Medac Ltd and at the Department of Health
and Human Sciences, London Metropolitan University. The syn-
theses of (C6F5)3B, (C6F5)3B·NH3, TpTi(NMe2)3, TpZr(NMe2)3,
TpHf(NMe2)3, Tp*Ti(NMe2)3 and Tp*Zr(NMe2) were conducted
according to the literature procedures.10,15,22,29–31
HB(C3H3N2)3Hf(NMe2)2{NH2B(C6F5)3} (5)
Samples of (C6F5)3B·NH3 (0.53 g, 1 mmol) and TpHf(NMe2)3
(0.52 g, 1 mmol) were combined and dissolved in toluene (20 cm3),
the solution was stirred at room temperature and the reaction
1
monitored by H and 19F NMR spectroscopy. The volatiles were
HB(C3H3N2)3Ti(NMe2)2{NH2B(C6F5)3} (2)
removed by vacuum distillation to yield the crude product as a
colourless foam. This was then washed with light petroleum to
yield the pure product as a colourless solid (0.61 g, 0.6 mmol,
61%). Elemental analysis found: C 36.9, H 2.3, N 12.5. Calculated
TpTi(NMe2)3 (0.39 g, 1 mmol) and (C6F5)3B·NH3 (0.53 g, 1 mmol)
were dissolved in toluene (80 cm3) and nitrogen was bubbled
through the solution for a period of 4 h. The volatiles were removed
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7434–7441 | 7439
©

View Article Online
Table 3 Crystallographic data
2
4·hexane
4·CH2Cl2
6·pentane
Chemical formula
M
Crystal system
C31H24B2F15N9Ti
877.1
Monoclinic
C31H24B2F15N9Zr, C6H14
1006.6
Monoclinic
C31H24B2F15N9Zr, CH2Cl2
1005.4
Monoclinic
C37H36B2F15N9Zr, C5H12
1076.7
Triclinic
¯
P1
Space group
I2/a
P21/n
P21/n
(equiv. to no. 15)
27.3497(5)
12.7238(2)
21.1149(4)
90
95.083(2)
90
7318.9(2)
140(1)
8
27.5
52275
8377, 0.054
4887
˚
a/A
12.6960(4)
24.1645(8)
14.1775(5)
90
105.824(3)
90
4184.7(2)
140(1)
4
25
44382
7352, 0.073
5407
12.9240(2)
23.2383(3)
13.4537(3)
90
103.201(2)
90
3933.80(12)
140(1)
4
27.5
53224
9024, 0.049
6101
11.2726(5)
11.8733(6)
17.9169(8)
80.540(4)
89.161(4)
81.252(4)
2337.8(2)
140(1)
2
22.5
18358
6062, 0.159
3533
˚
b/A
˚
c/A
a (◦)
b (◦)
g (◦)
3
˚
V/A
T/K
Z
q max/◦
Reflections measured
Unique reflections, Rint
Observed data
wR2, R1 (observed data)
wR2, R1 (all data)
Largest difference peak
0.072, 0.033
0.083, 0.068
0.31
0.088, 0.052
0.099, 0.082
0.57
0.070, 0.032
0.078, 0.059
0.75
0.118, 0.086
0.146, 0.171
0.47
for C31H24B2F15HfN9: C 36.95, H 2.4, N 12.5. dH/ppm (300.1
MHz, CDCl3) 7.77 (2H, d, JHH 2.1, CH), 7.75 (1H, d, JHH 2.2,
CH), 7.68 (2H, d, JHH 2.1, CH), 7.46 (1H, d, JHH 2.0, CH), 6.28
(1H, t, JHH 2.3, CH), 6.27 (2H, t, JHH 2.2, CH), 3.17 (2H, br, NH2),
2.77 (12H, s, CH3). dC/ppm (75.5 MHz, CDCl3) 143.3 (CH), 141.5
(CH), 136.6 (CH), 106.1 (CH), 43.00 (CH3). dB/ppm (96.3 MHz,
CDCl3) -3.12 (br, BH), -8.09 (s, (C6F5)3B). dF/ppm (282.4 MHz,
CDCl3) -133.3 (6F, d, JFF 23, o-F), -158.9 (3F, t, JFF 21, p-F),
-164.1 (6F, m, m-F).
CCD and RED programs.32 In each case, the structure was deter-
mined by direct method routines in the SHELXS program and
refined by full-matrix least-squares methods on F2 in SHELXL33
within the WinGX program suite.34 The results are collated in
Table 3. Scattering factors for neutral atoms were taken from
literature values.35
Full resolution of the structure of 4·hexane was hindered by a
poorly defined solvent molecule, presumed to be hexane, within
the crystal lattice. A full solution for the solvent molecule could not
be obtained as four carbon atoms were identified but disordered
over two positions and kept isotropic while two further carbon
atoms were located and refined anisotropically but displayed very
large displacement parameters. The six identified carbon atoms did
not link together in any sensible fashion to form a recognizable
molecule; furthermore location of these additional part atoms was
restricted due to low quality data. Hydrogen atoms for the solvent
molecule were therefore not included in the model but have been
included in the molecular formula.
Recrystallization of 4 from a dichloromethane/light petroleum
solution at -25 ◦C provided a crop of colourless crystals
(4·CH2Cl2). Analysis by X-ray diffraction techniques illustrated
both 4·hexane and 4·CH2Cl2 crystallizing in the same space group,
the unit cell dimensions being slightly different due to the presence
of a well resolved dichloromethane molecule in the crystal lattice
of 4·CH2Cl2. The structures of the zirconium complexes in both
4·hexane and 4·CH2Cl2 are very similar but are not perfectly
isostructural. The most noteworthy differences occur in the angles
about zirconium, with the N(41)–Zr(1)–N(1) angle differing by
over 2◦. The bond lengths about the zirconium atom and the Zr–
N–B bond angle in 4·CH2Cl2 agree with those in 4·hexane within
experimental error and in 4·CH2Cl2 there are no interactions with
the dichloromethane solvent molecule.
HB(C5H7N2)3Zr(NMe2)2{NH2B(C6F5)3} (6)
Samples of (C6F5)3B·NH3 (0.79 g, 1.5 mmol) and Tp*Zr(NMe2)3
(0.77 g, 1.5 mmol) were combined and dissolved in toluene (180
cm3). The reaction was stirred at room temperature until the 19F
NMR indicated that the major component in solution was the
desired product (ca. 13 h). The volatiles were removed under
reduced pressure and the product extracted with light petroleum
and cooled to -25 ◦C whereupon an impure yellow precipitate
separated. The pure product was obtained in very low yield by
recrystallization from dichloromethane/light petroleum (0.03 g,
0.03 mmol, 2%). Elemental analysis found: C 44.1, H 3.5, N
12.6. Calculated for C37H36B2F15N9Zr: C 44.2, H 3.6, N 12.6.
dH/ppm (300.1 MHz, CDCl3): 5.75 (1H, s, CH), 5.68 (2H, s,
CH), 3.05 (2H, br, NH2), 2.88 (12H, NCH3), 2.39 (6H, s, CH3),
2.33 (3H, s, CH3), 2.11 (3H, s, CH3), 1.93 (6H, s, CH3). dB/ppm
(96.3 MHz, CDCl3): -7.90 (s, (C6F5)3B), -9.20 (br, BH). dF/ppm
(282.4 MHz, CDCl3): -133.2 (6F, d, JFF 23, o-F), -159.6 (3F, t,
JFF 21, p-F), -164.9 (6F, m, m-F). Collection of 13C NMR data
was precluded by the extremely low yield and priority was given
to providing sufficient material for crystallography and elemental
analysis.
All the hydrogen atoms within the zirconium complex in both
structures 4·hexane and 4·CH2Cl2 were located in difference maps
and refined freely. Both structures exhibit the same hydrogen
bonding arrangement, with the hydrogen bond lengths and angles
in good agreement.
Crystal Structure Analyses
Diffraction data were collected at UEA on an Oxford Diffraction
Xcalibur-3 CCD diffractometer, and processed using the CrysAlis-
7440 | Dalton Trans., 2011, 40, 7434–7441
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Full structure resolution of the solvent molecule in 6·pentane,
presumed to be a pentane fraction from light petroleum, proved
impossible due to its highly disordered nature. Five carbon atoms
were located in difference maps, two of which have been modelled
over two positions, to form a straight alkyl chain. All carbon atoms
were modelled anisotropically. The hydrogen atoms have not been
modelled but have been included in the formula. A low cut-off
value for theta was used due to there being no reliable/significant
intensity data beyond this value.
13 J. K. Kouba and S. S. Wreford, Inorg. Chem., 1976, 15, 2313–
2314.
14 (a) D. L. Reger, M. E. Tarquini and L. Lebioda, Organometallics,
1983, 2, 1763–1769; (b) H. Lee and R. F. Jordan, J. Am. Chem. Soc.,
2005, 127, 9384–9385; (c) H. Lee, K. Nienkemper and R. F. Jordan,
Organometallics, 2008, 27, 5075–5081.
15 H. Cai, W. H. Lam, X. H. Yu, X. Z. Liu, Z. Z. Wu, T. N. Chen, Z.
Y. Lin, X. T. Chen, X. Z. You and Z. L. Xue, Inorg. Chem., 2003, 42,
3008–3015.
16 (a) K. M. Chi, S. R. Frerichs, B. K. Stein, D. W. Blackburn and J.
E. Ellis, J. Am. Chem. Soc., 1988, 110, 163–171; (b) M. D. Curtis, K.
Shiu, W. M. Butler and J. C. Huffman, J. Am. Chem. Soc., 1986, 108,
3335–3343.
17 (a) D. L. Hughes, G. J. Leigh and D. G. Walker, J. Chem. Soc., Dalton
Trans., 1988, 1153–1157; (b) D. L. Hughes, G. J. Leigh and D. G. Walker,
J. Chem. Soc., Dalton Trans., 1989, 1413–1416.
18 (a) S. C. Dunn, A. S. Batsanov and P Mountford, J. Chem. Soc., Chem.
Commun., 1994, 2007–2008; (b) S. C. Dunn, P. Mountford and O. V.
Shishkin, Inorg. Chem., 1996, 35, 1006–1012.
19 (a) S. Murtuza, O. L. Casagrande Jr and R. F. Jordan, Organometallics,
2002, 21, 1882–1890; (b) M. P. Gil and O. L. Casagrande Jr, J.
Organomet. Chem., 2004, 689, 286–292.
20 The only Tp amidoborane complexes of which we are aware are
pentafluorophenyl-free osmium compounds: (a) T. J. Crevier and J. M.
Mayer, Angew. Chem., Int. Ed., 1998, 37, 1891–1893; (b) T. J. Crevier,
B. K. Bennett, J. D. Soper, J. A. Bowman, A. Dehestani, D. A. Hrovat,
S. Lovell, W. Kaminski and J. M. Mayer, J. Am. Chem. Soc., 2001, 123,
1059–1071; (c) A. Dehestani, W. Kaminski and J. M. Mayer, Inorg.
Chem., 2003, 42, 605–611.
Acknowledgements
The authors would like to thank the Engineering and Physical
Sciences Research Council and the University of East Anglia for
financial support.
References
1 (a) C. W. Hamilton, R. T. Baker, A. Staubitz and I. Manners, Chem.
Soc. Rev., 2009, 38, 279–293; (b) Y. Kawano, M. Uruichi, M. Shimoni,
S. Taki, T. Kawaguchi, T. Kakizawa and H. Ogino, J. Am. Chem. Soc.,
2009, 131, 14946–14957; (c) M. E. Sloan, A. Staubitz, T. J. Clark, C. A.
Russell, G. C. Lloyd-Jones and I. Manners, J. Am. Chem. Soc., 2010,
132, 3831–3841.
2 (a) T. D. Forster, H. M. Tuononen, M. Parvez and R. Roesler, J.
Am. Chem. Soc., 2009, 131, 6689–6691; (b) D. J. Wolstenholme, K.
T. Traboulsee, A. Decken and G. S. McGrady, Organometallics, 2010,
29, 5769–5772.
3 (a) L. E. Toth, Transition Metal Carbides and Nitrides, Academic Press,
New York, 1971; (b) D. M. Hoffman, Polyhedron, 1994, 13, 1169–
1179 and references therein; (c) L. H. Dubois, Polyhedron, 1994, 13,
1329–1336; (d) B. H. Weiller, J. Am. Chem. Soc., 1996, 118, 4975–
4983.
4 A. J. Mountford, W. Clegg, R. W. Harrington, S. M. Humphrey and S.
J. Lancaster, Chem. Commun., 2005, 2044–2046.
5 A. J. Mountford, W. Clegg, S. J. Coles, R. W. Harrington, P. N. Horton,
S. M. Humphrey, M. B. Hursthouse, J. A. Wright and S. J. Lancaster,
Chem.–Eur. J., 2007, 13, 4535–4547.
6 For reviews of the chemistry of tris(pentafluorophenyl)borane see:
(a) W. E. Piers and T. Chivers, Chem. Soc. Rev., 1997, 26, 345–354;
(b) E. Y.-X Chen and T. J. Marks, Chem. Rev., 2000, 100, 1391–1434;
(c) G. Erker, Dalton Trans., 2005, 1883–1890; (d) W. E. Piers, Adv.
Organomet. Chem., 2005, 52, 1–76.
7 A. Fuller, W. Clegg, R. W. Harrington, D. L. Hughes and S. J. Lancaster,
Chem. Commun., 2008, 5776–5778.
21 All previous examples having had tetrahedral geometry, see references
4 and 5, with the exception of the rather remarkable ion pair reported
here: A. J. Mountford, S. J. Lancaster and S. J. Coles, Acta Crystallogr.,
Sect. C: Cryst. Struct. Commun., 2007, 63, m401–m404.
22 A. J. Mountford, S. J. Lancaster, S. J. Coles, P. N. Horton, D. L. Hughes,
M. B. Hursthouse and M. E. Light, Inorg. Chem., 2005, 44, 5921–5933.
23 S. J. Lancaster, A. J. Mountford, D. L. Hughes, M. Schormann and M.
Bochmann, J. Organomet. Chem., 2003, 680, 193–205.
24 For a review of nitrogen adducts of tris(pentafluorophenyl)borane see:
F. Focante, P. Mercandelli, A. Sironi and L. Resconi, Coord. Chem.
Rev., 2006, 250, 170–188.
25 Here there is also a short C–H ◊ ◊ ◊ F interaction from a dimethylamide
methyl group to the opposite o-F atom of the same C6F5 ring, C(33)–
˚
H(33c) ◊ ◊ ◊ F(76) (2.34(2) A).
26 A. Kayal, J. Kuncheria and S. C. Lee, Chem. Commun., 2001, 2482–
2483.
27 There are no close intramolecular C–H ◊ ◊ ◊ F interactions as observed
˚
in 4.CH2Cl2 but there is a short intermolecular interaction (2.37 A)
between one of the pyrazolyl methyl group hydrogen atoms and the
m-F of a C6F5 ring on a symmetry related molecule. The two molecules
are related by the symmetry operation x - 1, y, z.
8 For examples see: (a) H. W. Roesky, Y. Bai and M. Noltemeyer, Angew.
Chem., Int. Ed. Engl., 1989, 28, 754–755; (b) Z. Duan and J. G. Verkade,
Inorg. Chem., 1996, 35, 5325–5327; (c) A. Abarca, P. Go´mez-Sal, A.
Mart´ın, M. Mena, J. M. Poblet and C. Ye´lamos, Inorg. Chem., 2000,
39, 642–651; (d) M. Garc´ıa-Castro, A. Mart´ın, M. Mena, J.-M. Poblet
and C. Ye´lamos, Eur. J. Inorg. Chem., 2006, 1155–1160.
9 S. Trofimenko, J. Am. Chem. Soc., 1966, 88, 1842–1844.
10 (a) S. Trofimenko, J. Am. Chem. Soc., 1967, 89, 6288–6294; (b) S.
Trofimenko, Chem. Rev., 1993, 93, 943–980.
11 (a) M. J. Saly, M. J. Heeg and C. H. Winter, Inorg. Chem., 2009, 48,
5303–5312; (b) P. Biagini, F. Calderazzo, F. Marchetti, A. M Romano
and S. Spera, J. Organomet. Chem., 2006, 691, 4172–4180; (c) M. H.
Chisholm, N. W. Eilerts, J. C. Huffman, S. S. Iyer, M. Pacold and K.
Phomphari, J. Am. Chem. Soc., 2000, 122, 11845–11854; (d) M. C.
Keyes, B. M. Chamberlain, S. A. Caltagirone, J. A. Halfen and W. B.
Tolman, Organometallics, 1998, 17, 1984–1992.
12 A. Antin˜olo, F. Carrillo-Hermosilla, A. E. Corrochano, J. Ferna´ndez-
Baeza, M. Lanfranchi, A. Otero and M. A. Pellinghelli, J. Organomet.
Chem., 1999, 577, 174–180.
28 In one instance, adventitious water was introduced during the at-
tempted synthesis of 6, resulting in the precipitation of a yellow
crystalline solid. Poor solubility precluded spectroscopic analysis but
elemental analysis was consistent with the formulation as the hydrolysis
product {Tp*Zr(NMe2)2}2{m-O}. Elemental analysis found: C 47.0, H
6.9, N 23.0. Calculated for C38H68B2N16OZr2: C 47.1, H 7.1, N, 23.1.
29 (a) S. J. Lancaster, http://www.syntheticpages.org/pages/215; (b) J. L.
W. Pohlmann and F. E. Z. Brinckman, Naturforsch., 1965, 20b, 5.
30 G. M. Diamond and R. F. Jordan, Organometallics, 1996, 15, 4030–
4037.
31 S. Trofimenko, J. Am. Chem. Soc., 1967, 89, 3170–3177.
32 CrysAlis-CCD and -RED Oxford Diffraction Ltd, Abingdon, UK,
2008.
33 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
64, 112–122.
34 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
35 International Tables for X-Ray Crystallography, ed. A. J. C. Wilson,
Kluwer Academic Publishers, Dordrecht, The Netherlands, 1992, vol.
C, pp. 193, 219 and 500.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 7434–7441 | 7441
©