Enantiospecific synthesis of a planar chiral bidentate indenyl-alkoxide complex of zirconium using an axially chiral indene ligand

Article abstract of DOI:10.1039/a903740c

The reaction of axially chiral (M)-1-(2'-methyl-3'-indenyl)-naphthalene-2-methanol 3 with Zr(NEt₂)₄ enantiospecifically provides the planar chiral complex (P)-bis(diethyl-amido)-[ 1-(2'-methyl-1'-indenyl)naphthalene-2-methoxo]-zircon

Full text of DOI:10.1039/a903740c

Enantiospecific synthesis of a planar chiral bidentate indenyl–alkoxide complex
of zirconium using an axially chiral indene ligand
Robert W. Baker* and Brian J. Wallace
School of Chemistry, The University of Sydney, NSW 2006, Australia. E-mail: r.baker@chem.usyd.edu.au
Received (in Cambridge, UK) 10th May 1999, Accepted 22nd June 1999
The reaction of axially chiral (M)-1-(2A-methyl-3A-indenyl)-
naphthalene-2-methanol 3 with Zr(NEt2)4 enantiospeci-
CO2R*
H
Me
fically provides the planar chiral complex (P)-bis(diethyl-
amido)-[1-(2A-methyl-1A-indenyl)naphthalene-2-methoxo]-
zirconium 7.
Me
R
Planar chiral cyclopentadienylmetal complexes feature amongst
the more successful stereoselective catalysts. They are usually
obtained in enantiomerically pure form through resolution
procedures, there being few methodologies allowing asym-
metric synthesis. The most common approach1 designed to
avoid resolution utilises cyclopentadienyl ligands containing
other stereogenic elements, rendering the faces of the ligand
diastereotopic; however, stereoselectivity for the planar chiral
element generated during metallation reactions is not guaran-
teed and separation of diastereoisomers may still be required.
An approach extensively employed for 1,2-disubstituted ferro-
cenes involves the diastereoselective lithiation of complexes
containing a chiral ortho-directing group.2 An alternative to this
is the enantioselective metallation of ferrocenes with achiral
ortho-directing groups using a palladium(ii) salt with an N-
protected amino acid,3 or a chiral lithium amide or an
alkyllithium/chiral ligand complex.2 Two other recently re-
ported methods for the asymmetric synthesis of planar chiral
cyclopentadienylmetal complexes are the displacement of
chiral auxiliary ligands (Lch) from half-sandwich complexes
1 R* = (1R)-menthyl
2 R = CO2R*
3 R = CH2OH
overall yield. Oxidation of 5 with dimethyldioxirane, generated
in situ using OXONE®, provided the sulfoxide 6 as a 58+42
mixture of epimers at sulfur (1H NMR analysis) in 51% yield,
accompanied by 46% of recovered 5.† Reaction of 6 with
2-methylindenyllithium during 40 min at 0 °C furnished 1 and
the +ac(R) isomer of 1, in a ratio of 44+56, respectively (1H
NMR analysis), in 92% combined yield. This mixture was then
quantitatively isomerised to a mixture of 2 and the P-epimer of
2, in a ratio of 46+54, respectively,‡ on treatment with
triethylamine. Fractional crystallisation, then recrystallisation
from hexane solution furnished 2 in > 99% de‡ and 32% yield
(maximum theoretical yield 46%). A single-crystal X-ray
structure determination on 2§ establishes the M absolute
configuration at the chiral axis, which had been previously
proposed on the basis of comparisons of CD spectra.9 The rate
of epimerisation of 2 to the P-epimer (ultimately affording a 1+1
mixture) has been determined by HPLC analysis in xylenes
ch
Cp*RuL2 by planar prochiral cyclopentadienylmetals,4 and
the cyclisation of ferrocenyl diazoketones catalysed by chiral
rhodium complexes.5 We have proposed a ligand design
whereby a second metal coordination site is constrained to one
face of an asymmetrically substituted cyclopentadienyl ring
through hindered rotation about the bond between the cyclo-
pentadienyl ring and a naphthalene moiety; formation of
chelating complexes using these ligands was anticipated to lead
to the enantiospecific generation of a planar chiral cyclopenta-
dienylmetal complex. Towards this objective, we have de-
scribed approaches to the asymmetric synthesis of precursors to
axially chiral chelating fluorene6,7 and indene8,9 ligands,
employing ligand coupling reactions of sulfoxides. Here we
demonstrate the potential of our ligand design with the
enantiospecific synthesis of a planar chiral bidentate indenyl–
alkoxide complex of zirconium using an axially chiral indene
ligand.
solution at reflux,‡ providing a barrier to rotation DG‡
=
413
150.6 kJ mol21. Reduction of 2 of > 99% de‡ with LAH affords
3 quantitatively with > 99% ee.¶ While the barrier to rotation in
3 has not been determined, it can be anticipated to be higher
again than that of 2.8
Br
R
CO2H
CO2R*
i, ii, iii
5 R = SPh
4
iv
6 R = S(O)Ph
We have described previously9 that the reaction of (1R)-
menthyl
(R)-1-(p-tolylsulfinyl)naphthalene-2-carboxylate7
v, vi, vii
with 2-methylindenyllithium during 30 min at 0 °C provides
exclusively the 2ac-rotamer of (1R)-menthyl (S)-1-(2A-methyl-
1A-indenyl)naphthalene-2-carboxylate 1 in 50% yield and 70%
de. The C1–C1A bond of 1 was shown to be a stable stereogenic
element that was conserved during subsequent prototropic shifts
to 1-(2A-methyl-3A-indenyl)naphthalenes; thus, treatment of 1
with triethylamine provided 2 in 97% yield and 65% de, while
LAH reduction gave 3 in quantitative yield and 70% ee. In order
to obtain ligand 3 in enantiomerically pure form, a convenient
procedure amenable to operation on a large scale has been
developed (Scheme 1). Esterification of 1-bromo-2-naphthoic
acid 4 (readily prepared from 2-methylnaphthalene)10 with
(1R)-menthol followed by displacement of the bromide with
sodium thiophenoxide, furnished the phenylsulfide 5 in 94%
viii
3
Et2N
Et2N
Me
O
Zr
7
Scheme 1 Reagents and conditions: i, SOCl2 (15 equiv.), 25 °C, 20 h; ii,
(1R)-menthol (2 equiv.), pyridine (2 equiv.), CH2Cl2, 25 °C, 30 min; iii,
PhSNa (1.2 equiv.), DMF, 60 °C, 20 h; iv, OXONE® (3 equiv.), NaHCO3
(5 equiv.), acetone–MeCN–water (20+10+1), 5 °C, 20 h; v, 2-methyl-
indenyllithium (1.3 equiv.), THF, 0 °C, 40 min; vi, NEt3–toluene (1+1),
reflux, 24 h; vii, LAH (5 equiv.), diethyl ether, 25 °C, 1 h; viii, Zr(NEt2)4 (1
equiv.), toluene, reflux, 16 h.
Chem. Commun., 1999, 1405–1406
1405

§ Details of the single-crystal X-ray structure determination of 2 will appear
in the full account.
¶ Enantiomeric purity was determined by HPLC: a 4.6 3 250 mm column
(Pirkle Type 1A, Regis) was used with 5% propan-2-ol/hexane as eluent at
a flow rate of 1.5 ml min21, detection 254 nm, tr: 9.5 min for 3 and 10.2 min
for ent-3.
∑ Somewhat related to this work, Pregosin and coworkers12 have reported
that
reaction
of
Ru(OAc)2[(M)-6,6A-dimethoxybiphenyl-2,2A-diyl-
bis{di(3,5-di-tert-butylphenyl)phosphine}] with MeLi affords a planar
chiral RuMe2 complex where Ru is coordinated to one phosphine together
6
with an h -C6H3 moiety from one of the biaryl rings; however, the
enantiomeric purity of this complex is not discussed.
** Complex 7 was isolated as an orange viscous oil which, to date, has
resisted crystallisation. Consequently, the purity of 7 is dependant on the
precision of the stoichiometry of the reaction between 3 and Zr(NEt2)4, e.g.
the reaction product shown in Fig. 1 contains ca. 5 mol% unreacted
Zr(NEt2)4.
Fig. 1 1H NMR (400 MHz, C6D6) spectrum of complex 7; unobscured
signals of residual toluene and Zr(NEt2)4 are indicated by • and *,
respectively.
†† NMR data for 7: dH (400 MHz, C6D6, primed numbers refer to the
indenyl moiety) 0.94, 2.65 and 2.74 (10H, ABX3, J 14.0, 6.8, NEt2), 1.05,
3.17 and 3.38 (10H, ABX3, J 14.0, 6.8, NEt2), 1.96 (3H, s, 2A-Me), 5.43 and
5.56 (2H, AB, J 15.6, CH2O), 6.29 (1H, s, H-3A), 6.76 (1H, dd, J 7.6, 7.6, 6A-
H), 6.89–6.94 (2H, m, 5A- and 7A-H), 7.07 (1H, dd, J 8.4, 6.8, 7-H), 7.15 (1H,
d, J 8.4, 3-H), 7.22 (1H, dd, J 8.0, 6.8, 6-H), 7.41 (1H, d, J 8.4, 8-H), 7.53
(1H, d, J 8.4, 4A-H), 7.66 (1H, d, J 8.4, 4-H) and 7.71 (1H, d, J 8.0, 5-H); dC
(100 MHz, C6D6) 140.3, 134.8, 133.5, 130.9, 128.5 (each C), 128.4, 128.1
(each CH), 127.6 (C), 126.8, 126.6 (each CH), 125.9 (C), 125.6, 125.0,
123.7, 123.5, 123.0, 121.7 (each CH), 110.4 (C), 97.2 (CH), 74.0 (CH2),
45.02, 44.99 (each 2 3 CH2), 16.2, 15.7 (each 2 3 CH3) and 14.0 (CH3).
With multigram quantities of enantiomerically pure 3
available, we next sought to prepare a bidentate indenyl–
alkoxide metal complex11 in order to demonstrate that the axial
chirality of the ligand would indeed translate into planar
chirality in the complex without loss of enantiomeric purity.∑
Initial attempts at metallation of 3 through formation of the
dilithio species with BuLi, followed by reaction with ZrCl4 in a
variety of solvents, failed to produce any of the desired
complex. Metallation was readily achieved, however, employ-
ing the amine elimination method.13 Reaction of 3 with 1 equiv.
of Zr(NEt2)4 in toluene solution under reflux for 16 h, followed
by removal of the solvent, led to essentially quantitative
conversion to the complex 7,** as evident from inspection of
the 1H NMR spectrum of the crude product (Fig. 1).†† In
particular, the methylene signal of the indene moiety of 3 (AB
pattern centred at dH 3.65) is replaced by a singlet at dH 6.29 in
7; the methylene signal of the naphthalenemethanol moiety of 3
(AB pattern centred at dH 4.63) is significantly deshielded on
formation of 7 (AB pattern centred at dH 5.50); and the
diastereotopic NEt2 ligands of 7 appear as two distinct ABX3
patterns. Consistent with the mononuclear structure proposed
1 R. L. Halterman, Chem. Rev., 1992, 92, 965 and references cited therein;
R. L. Halterman, Synthesis of Chiral Titanocene and Zirconocene
Dichlorides, in Metallocenes, ed. A. Togni and R. L. Halterman, Wiley-
VCH, Weinheim, 1998, ch. 8, and references therein.
2 C. J. Richards and A. J. Locke, Tetrahedron: Asymmetry, 1998, 9, 2377
and references therein.
3 V. I. Sokolov, L. L. Troitskaya and O. A. Reutov, J. Organomet. Chem.,
1979, 182, 537.
4 U. Koelle, K. Bu¨cken and U. Englert, Organometallics, 1996, 15,
1376.
5 S. Siegel and H.-G. Schmalz, Angew. Chem., Int. Ed. Engl., 1997, 36,
2456.
6 R. W. Baker, T. W. Hambley and P. Turner, J. Chem. Soc., Chem.
Commun., 1995, 2509.
1
for 7, the high field H NMR spectrum of the crude product
obtained in the reaction of rac-3 with Zr(NEt2)4 was identical to
that obtained with enantiomerically pure 3, with additional
signals being completely absent. Hydrolysis of complex 7 with
dilute HCl solution quantitatively returned ligand 3 with > 99%
ee.¶ It follows that complex 7 is also enantiomerically pure and
can be assigned P absolute configuration (a single descriptor for
configuration is used as the atropisomerism of the precursor
ligand 3 is considered to be latent in the complex). We are
currently examining the expansion of this approach to a wide
range of ligands and metal complexes.
7 R. W. Baker, M. A. Foulkes and J. A. Taylor, J. Chem. Soc., Perkin
Trans. 1, 1998, 1047.
8 R. W. Baker, T. W. Hambley, P. Turner and B. J. Wallace, Chem.
Commun., 1996, 2571, (Corrigendum, 1997, 506).
9 R. W. Baker, J. N. H. Reek and B. J. Wallace, Tetrahedron Lett., 1998,
39, 6573.
10 S. L. Colletti and R. L. Halterman, Organometallics, 1991, 10, 3438.
11 For other examples of bidentate cyclopentadienyl–alkoxide complexes
of Group IV metals see: B. Reiger, J. Organomet. Chem., 1991, 420,
C17; Y.-X. Chen, P.-F. Fu, C. L. Stern and T. J. Marks, Organo-
metallics, 1997, 16, 5958; E. E. C. G. Gielens, J. Y. Tiesnitsch, B.
Hessen and J. H. Teuben, Organometallics, 1998, 17, 1652; S. D. R.
Christie, K. W. Man, R. J. Whitby and A. M. Z. Slawin, Organo-
metallics, 1999, 18, 348.
Notes and references
† A reaction sequence analogous to that from 4 to 6 was first developed in
this laboratory by M. A. Foulkes using the isopropyl esters.
‡ Diastereoisomeric purity was determined by HPLC: a 4.6 3 250 mm
column (Zorbax 5 m silica, Jones) was used with 0.4% ethyl acetate–hexane
as eluent at a flow rate of 1.5 ml min21, detection 254 nm, tR: 17.4 min for
2 and 20.0 min for the P-epimer of 2.
12 N. Feiken, P. S. Pregosin and G. Trabesinger, Organometallics, 1997,
16, 3735.
13 G. M. Diamond, R. F. Jordan and J. L. Petersen, J. Am. Chem. Soc.,
1996, 118, 8024 and references therein.
Communication 9/03740C
1406
Chem. Commun., 1999, 1405–1406