Isolation and characterisation of transition and main group metal complexes supported by hydrogen-bonded zwitterionic polyphenolic ligands

Article abstract of DOI:10.1039/b303618a

Reaction of Zr(OⁱPr)₄ or Sn[N(SiMe₃)₂]₂ with the tris-phenol amine ligand H₃L^(Me/Me) results in the formation of zirconium or tin complexes containing the new C₃-symmetric zwitterionic ammonium-trisphenolate ligand HL^2-, while increasing the steric bulk of the ligand results in the isolation of a zirconium complex containing the known trianionic ligand L^3-.

Full text of DOI:10.1039/b303618a

Isolation and characterisation of transition and main group metal
complexes supported by hydrogen-bonded zwitterionic polyphenolic
ligands†
Matthew G. Davidson,a Cheryl L. Doherty,a Andrew L. Johnsona and Mary F. Mahonb
a Department of Chemistry, University of Bath, Claverton Down, Bath, UK BA2 7AY.
E-mail: m.g.davidson@bath.ac.uk; a.l.johnson@bath.ac.uk; Fax: +44 (0)1225 386231; Tel: +44 (0)1225
384467
b Bath Chemical Crystallography Unit, Department of Chemistry, University of Bath, Claverton Down, Bath,
UK BA2 7AY; Fax: +44 (0)1225 386231; Tel: +44 (0)1225 383752
Received (in Cambridge, UK) 2nd April 2003, Accepted 23rd May 2003
First published as an Advance Article on the web 19th June 2003
Reaction of Zr(OiPr)4 or Sn[N(SiMe3)2]2 with the tris-phenol
amine ligand H3L(Me/Me) results in the formation of zirco-
nium or tin complexes containing the new C3-symmetric
zwitterionic ammonium-trisphenolate ligand HL22, while
increasing the steric bulk of the ligand results in the isolation
of a zirconium complex containing the known trianionic
although small amounts of an insoluble white precipitate are
formed over a period of hours after exposure of solutions to
atmospheric moisture.
An X-ray crystallographic study‡ reveals 3a to be an
unprecedented C3-symmetric zwitterionic complex, in which
two formally trianionic [L(Me/Me)] ligands are bound to a central
6-coordinate, octahedral, zirconium centre, resulting in a formal
dianionic charge at the metal centre (Fig. 1). This charge is
ligand L32
.
Trianionic, chelating amido and/or alkoxo ligands, which are
capable of providing a range of stable steric and electronic
environments for catalytically active metal centres, have been
the subject of considerable research for some time, not least for
their ability to bind metal centres in a C3-symmetric fashion.1
We and others have recently been exploring the rich
chemistry of amine trisphenolate ligands, 1, and their use in the
stabilisation of highly reactive, Lewis acidic Ti(IV) metal
centres.2,3 For example, Kol et al.have shown that reaction of
ligands 1a and 1b with one equivalent of titanium tetra(iso-
propoxide) yields the corresponding monomeric [L]Ti(OiPr)
complexes 2a and 2b.3a This is in contrast to the oligomeric
systems formed using other tri-ol ligands.1f We now report on
complexes of 1 with both zirconium and tin which exhibit a new
coordination mode for ligands of this type, which potentially
increases their utility as scaffolds for reactive metal centres and
provides a highly unusual hydrogen bonding environment.
Treatment of ligand 1a with one equivalent of [Zr(OiPr)4-
(HOiPr)] in toluene under ambient conditions yields the
colourless complex (HL(Me/Me))2Zr, 3a, whereas, a similar
reaction with ligand 1b yields the pale yellow complex
t
t
(L( Bu/ Bu))Zr(OiPr), 3b (Scheme 1).† Both complexes are
readily soluble in a range of aromatic, and aliphatic solvents and
can be crystallised from toluene solutions. In the case of
complex 3a these crystals are air stable but contain toluene
molecules of crystallisation which are gradually lost on
standing. Complex 3b shows some stability to hydrolysis,
Fig. 1 ORTEP (50% probability ellipsoids) diagrams of 3a. Top: A view
down the N–Zr–N vector showing the C3-symmetry of the complex.
Bottom: A side view of 3a showing H-bonding interactions. Selected bond
lengths (Å): Zr(1)–O(1) 2.058(2), Zr(1)–O(2) 2.064(2), Zr(1)–O(3)
2.057(2), Zr(1)–N(1) 3.617(2), Zr(1)–H(1) 2.688(3), N(1)–H(1) 0.87(3),
N(1)–O(1) 2.868(2) (O(1)–H(1) 2.21(3)), N(1)–O(2) 2.920(2) (O(2)–H(1)
2.29(2)), N(1)–O(3) 2.834(2) (O(3)–H(1) 2.15(3)); Bond angles (°): Zr(1)–
O(1)–C(16) 157.23(15), Zr(1)–O(2)–C(26) 154.51(15), Zr(1)–O(3)–C(36)
160.04(15), N(1)–H(1)–O(1) 130.21(19), N(1)–H(1)–O(2) 129.51(19),
N(1)–H(1)–O(3) 133.39(19). Toluene and remaining hydrogen atoms have
been omitted for clarity.
Scheme 1 Synthesis of 2 and 3.
† Electronic supplementary information (ESI) available: full synthetic and
spectroscopic details. See http://www.rsc.org/suppdata/cc/b3/b303618a/
1832
CHEM. COMMUN., 2003, 1832–1833
This journal is © The Royal Society of Chemistry 2003

balanced by the protonation of both nitrogen atoms to form
acidic (dH = 11.8 ppm) cationic ammonium centres, which are
involved in unusual trifurcated N–H…O3 hydrogen bonding.
Overall this arrangement represents a new binding mode for
these ligands. Compound 3a can be considered a ‘truly
zwitterionic’ complex, as the formal charge separation cannot
be modified by resonance of p-electrons.4
the tendency of Sn(II) complexes containing tris-alkoxide
ligands to form polymetallic clusters.6 As with complexes 3, the
three arms of the trisphenolate ligand in 4, are orientated in a C3-
symmetric propeller arrangement in the solid state, but unlike
1
3a, the H NMR spectrum of 4 shows only a single resonance
for the CH2 protons, indicating that rapid inversion of the
structure occurs at room temperature. As a consequence of
chelation by HL22, the pyramidal geometry at Sn(II) is
reinforced, and the metal centre lies out of the plane of the three
phenolic oxygen atoms by 1.259 Å. In spite of this apparent
nakedness, complex 4 exhibits considerable stability and
resistance to hydrolysis. For example, on addition of D2O to a
CDCl3 solution of 4 the NH unit (d = 10.7 ppm) disappears
rapidly but there is no further change to the NMR (1H, 13C and
119Sn) spectra even after several weeks.
In conclusion, the new zwitterionic coordination mode of
HL22 observed in complexes 3a and 4 appears to be a general
feature of the metal chemistry of H3L ligands which is dictated
by the size of the metal centre and/or ligand-ligand repul-
sions.
In contrast, the solid state molecular structure of 3b‡ (as
represented in Scheme 1) is very similar to analogous titanium
complexes.2,3a Comparison of 3a with 3b highlights a number
of structural differences between the two bonding modes. The
Zr metal centre in 3a is displaced out of the plane of the oxygen
atoms by 1.255 Å, compared to only 0.365 Å for the same
parameter in 3b. Furthermore, the Zr–O bonds in 3b, are
significantly shorter than the corresponding Zr–O bonds in 3a
(av. Zr–O 1.95 and 2.06Å, for 3b and 3a respectively). These
features are consistent with significant steric repulsion between
both the two ligands and the NH unit and the metal centre in
3a.
In solution at room temperature, 1H NMR spectra are
consistent with the C3-symmetry of 3a and 3b being main-
tained, as indicated in both cases by the presence of an AB spin
system for the CH2 protons.3a
We thank the EPSRC and the Royal Society (MGD) and the
University of Bath (ALJ).
These structural observations can be interpreted as follows.
In the case of tetravalent metals, the formation of HL22
complexes is dependant on two factors. First, the metal centre
must be large enough to accommodate the N–H…O3 structural
motif without imposing repulsive NH….M interactions.5 Sec-
ond, the steric demands of the ligand must not inhibit formation
of the a stable metal HL22 complex. Hence, we predict that Ti
is too small to allow HL22 coordination and L32 complexes
will always result.2,3 However, in the case of larger tetravalent
metals, (e.g., Zr and Hf), coordination of HL22 is possible, but
only provided that the ligand is sufficiently small (e.g., ortho-
methyl-substituted) to allow two ligands to approach the metal
centre.
These results inspired us to explore other metal complexes in
order to test the generality of the new HL22 coordination mode.
Thus, we speculated that a large, divalent metal such as Sn(II)
should favour the zwitterionic bonding mode observed in 3a.
Reaction of one equivalent of Sn[N(SiMe3)2]2 and 1a yields the
colourless complex 4. The solid state molecular structure of
HL(Me/Me)Sn, 4,‡ (Fig. 2) shows a monomeric complex with a
very similar metal–ligand bonding motif to that found in 3a.
Indeed, the similarity between 3a and 4 in terms of key
structural parameters [e.g., M–O distances: 2.064(2) and
2.081(2) Å. N–H….M distances: 2.688(3) and 2.689(4) Å in 3a
and 4, respectively] suggests that a balance between attractive
N–H…O and repulsive N–H…M interactions has been reached
and that these impose an unusual rigidity on a normally flexible
ligand. The monomeric nature of 4 is noteworthy considering
Notes and references
‡
Crystal data were collected on a Nonius KappaCCD diffractometer
using Mo–Ka radiation (l = 0.71073 Å), and all structures were solved by
direct methods and refined on all F2 data using the SHELX-97 suite of
programs.7 Hydrogen atoms not involved in hydrogen bonding included in
idealised positions and refined using a riding model.
3a: C70H81N2O6Zr, M = 1137.59, yellow blocks, crystal size 0.25 3
¯
0.20 3 0.06 mm, Triclinic, space group P1, a = 11.338(3) Å, b = 11.768(3)
Å, c = 12.886(4) Å, a = 82.058(1)°, b = 83.645(1)°, g = 73.389(1)°, U =
1627.17(8) Å3, Z = 1, Dc = 1.161 g cm23, T = 150(2) K, 19089 reflections
measured, 7403 unique reflections (2qmax = 27.52°, Rint = 0.0290) against
384 parameters gave R1 = 0.0510 and wR2 = 0.1392 [I > 2s(I)] (R1
0.0583 and wR2 = 0.1462 for all data). CCDC 207513.
=
3b: C54H87N1O4Zr, M = 905.47, colourless blocks, crystal size 0.25 3
0.20 3 0.17 mm, Tetragonal, space group P43, a = b = 14.581(1) Å, c =
25.419(1) Å, U = 5404.27(7) Å3, Z = 4, Dc = 1.113 g cm23, T = 150(2)
K, 100213 reflections measured, 12336 unique reflections (2qmax = 27.48°,
Rint = 0.0486) against 768 parameters, with 55 restraints gave R1 = 0.0363
and wR2 = 0.0901 [I > 2s(I)] (R1 = 0.0450 and wR2 = 0.0956 for all data).
CCDC 207514.
4: C34H39N1O3Sn, M = 628.35, colourless blocks, crystal size 0.32 3
¯
0.20 3 0.16 mm, Trigonal, space group R3, a = b = 14.0560(4) Å, c =
25.9860(7) Å, g = 120°, U = 4446.2(2) Å3, Z = 6, Dc = 1.408 g cm23, T
= 150(2) K, 13702 reflections measured, 2246 unique reflections (2qmax
=
27.46°, Rint = 0.0759) against 130 parameters gave R1 = 0.0354 and wR2
= 0.0740 [I > 2s(I)] (R1 = 0.0594 and wR2 = 0.0812 for all data). CCDC
207515
See http://www.rsc.org/suppdata/cc/b3/b303618a/ for crystallographic
data in CIF format.
1 (a) R. R. Schrock, Acc. Chem. Res., 1997, 30, 9; (b) L. H. Gade, Chem.
Commun., 2000, 173; (c) J. G. Verkade, Acc. Chem. Res., 1993, 26, 483;
(d) C. Moberg, Angew. Chem. Int. Ed., 1998, 37, 248; (e) F. Di Furia, G.
Licini, G. Modena, R. Motterle and W. A. Neugent, J. Org. Chem., 1996,
61, 5177; (f) H. Lütjens, G. Wahl, F. Möller, P. Knochel and J.
Sundermeyer, Organometallics, 1997, 16, 5869.
2 S. D. Bull, M. G. Davidson, A. L. Johnson, D. E. E. Robinson and M. F.
Mahon, Chem. Commun., 2003.
3 (a) M. Kol, M. Shamis, I. Goldberg, Z. Goldshmit, S. Alfi and E. Hayut-
Salant, Inorg. Chem. Commun., 2001, 4, 177; (b) Y. Kim and J. G.
Verkade, Organometallics, 2002, 21, 2395; (c) Y. Kim, P. N. Kapoor and
J. G. Verkade, Inorg Chem., 2002, 41, 4834.
4 R. Chauvin, Eur. J. Inorg. Chem., 2000, 577.
5 (a) F. Cecconi, C. A. Ghilardi, P. Innocenti, C. Mealli, S. Midollini and
A. Orlandini, Inorg. Chem., 1984, 23, 922; (b) L. Brammer, J. C.
Mareque Rivas and C. D. Spilling, J. Organomet, Chem., 2000, 609,
36.
6 T. J. Boyle, J. M. Segall, T. M. Alam, M. A. Rodriguez and J. M. Santana,
J. Am. Chem. Soc., 2002, 124, 6904.
Fig. 2 ORTEP (50% probability ellipsoids) diagram of 4. Selected bond
lengths/Å: Sn(1)–O(1) 2.081(2), Sn(1)–N(1) 3.631(3), Sn(1)–H(1N)
2.689(4), N(1)–H(1N) 0.97(4), O(1)–N(1) 2.893(3), (O(1)–H(1N) 2.17(2));
Bond angles/°:Sn(1)–O(1)–C(11) 136.25(16), N(1)–H(1N)–O(1) 130.8(8).
Toluene and hydrogen atoms omitted for clarity.
7 G. M. Sheldrick, SHELXL-97, Program for refinement of crystal
structures, University of Göttingen, Germany, 1997.
CHEM. COMMUN., 2003, 1832–1833
1833