Low-valent titanium bis(phospholyl) chemistry: A configurationally stable chiral phosphatitanocene

Article abstract of DOI:10.1021/om020961h

A low-valent phosphatitanocene was prepared and structurally characterized. A chiral Ti(II) phosphametallocene was also prepared. It was found that the Ti(II) phosphametallocene has a higher barrier to isomerization than the Ti(IV) analogue. The difference between the molecular orbitals of the two complexes was also explained.

Full text of DOI:10.1021/om020961h

1432
Organometallics 2003, 22, 1432-1436
Low -Va len t Tita n iu m Bis(p h osp h olyl) Ch em istr y: A
Con figu r a tion a lly Sta ble Ch ir a l P h osp h a tita n ocen e
T. Keith Hollis,* Yi J oon Ahn, and Fook S. Tham
Department of Chemistry, University of California, Riverside, California 92521-0403
Received November 20, 2002
The first example of a low-valent phosphatitanocene has been prepared and structurally
characterized, and a chiral Ti(II) phosphametallocene was prepared and was found to have
a significantly higher barrier to isomerization than the Ti(IV) analogue. Descriptions of the
molecular orbitals of the two complexes offer an explanation of the difference.
Resu lts a n d Discu ssion
In tr od u ction
The achiral Ti(II) complex 1 was prepared in analogy
Heterocyclic π-ligands are of significant interest in the
continuing development of transition metal ligands, and
phospholyl anions (C4R4P-), analogues of the cyclopen-
tadienyl anion (C5R5-), have been receiving much at-
tention.1,2 Reports of the synthesis of phospholyl tran-
sition metal complexes have been summarized in several
reviews.3 The phosphatitanocenes and phosphazir-
conocenes in the +4 oxidation state have been known
for more than two decades.4 They have been structurally
characterized, providing unequivocal evidence for the
bonding modes of these complexes.5 While the low-
valent chemistry of bis(cyclopentadienyl) complexes of
the Ti triad is rich, plentiful, and well-documented,6,7
reports of the corresponding low-valent chemistry of the
bis(phospholyl) complexes have begun to appear only
recently.8,9 We report here the first examples of bis-
(phospholyl)Ti(II) complexes.
with established procedures for the reduction of met-
allocenes. Mathey and co-worker’s recently reported Mg
reduction conditions for phosphazirconocenes proved
most productive for Ti also.9 (η5-C4Me4P)2TiCl2 was
prepared in situ, according to literature procedures from
1-(trimethylstannyl)-2,3,4,5-tetramethylphosphole, and
was reduced using Mg powder under an atmosphere of
CO at 40 °C (eq 1). The desired complex 1 was obtained
in 39% yield. Titanium(II) complex 1 was characterized
by IR and NMR spectroscopy (1H, 31P, and 13C). The CO
stretches are readily apparent in the IR spectrum (1959,
1900 cm-1), and satisfactory elemental analysis was
obtained.
* Corresponding author. Fax: 909-787-4713. E-mail: keith.hollis@
ucr.edu.
(1) For recent examples of other heterocyclic π-ligands see: Bo-
ratabenzene: (a) Ashe, A. J .; Al-Ahmad, S.; Fang, X. D.; Kampf, J . W.
Organometallics 2001, 20, 468-473. (b) Ashe, A. J .; Al-Ahmad, S.;
Fang, X. G. J . Organomet. Chem. 1999, 581, 92-97. (c) Bazan, G. C.;
Cotter, W. D.; Komon, Z. J . A.; Lee, R. A.; Lachicotte, R. J . J . Am.
Chem. Soc. 2000, 122, 1371-1380. (d) Rogers, J . S.; Bu, X. H.; Bazan,
G. C. J . Am. Chem. Soc. 2000, 122, 730-731. (e) Lee, B.; Wang, S.;
Putzer, M.; Bartholomew, G.; Bu, X.; Bazan, G. J . Am. Chem. Soc.
2000, 122, 3969-3970. (f) Hoic, D. A.; Davis, W. M.; Fu, G. C. J . Am.
Chem. Soc. 1996, 118, 8176. Borabenzene: (g) Qiao, S.; Hoic, D. A.;
Fu, G. C. Organometallics 1997, 16, 1501-1502.
X-ray quality crystals of 1 were grown from a toluene
solution at -35 °C, and the red needles were used for
the structural analysis.10 An ORTEP plot of Ti(II)
complex 1 is presented in Figure 1 with pertinent
structural details. This structure is related to the Zr
analogue reported by Mathey,9 but it lacks crystal-
lographic symmetry. The dihedral angle between the P
atoms (P-Ctr-Ctr-P, Ctr ) centroid) is approximately
180°, which results in a staggered conformation of the
rings, unlike the corresponding Cp* analogue that has
the rings in an eclipsing conformation.11 The staggered
conformation produces an antiparallel arrangement of
the P-lone pair vectors, resulting in dipole minimization.
(2) For a recent review of chiral phospholyl chemistry see: Ganter,
C. J . Chem. Soc., Dalton Trans. 2001, 3541-3548.
(3) For reference to recent reviews see: (a) Ashe, A. J .; Al-Ahmad,
S. Adv. Organomet. Chem. 1996, 39, 325-353. (b) Mathey, F. Coord.
Chem. Rev. 1994, 137, 1-52. (c) Dillon, K. B.; Mathey, F.; Nixon, J . F.
Phosphorus: The Carbon Copy. From Organophosphorus to Phospha-
Organic Chemistry; J ohn Wiley & Sons: New York, 1998.
(4) The first spectroscopically characterized complexes were reported
by Meunier, P.; Gautheron, B. J . Organomet. Chem. 1980, 193, C13-
C16.
(5) For well-characterized examples see: (a) Nief, F.; Mathey, F. J .
Chem. Soc., Chem. Commun. 1988, 770-771. (b) Nief, F.; Mathey, F.;
Ricard, L. Organometallics 1988, 7, 921-926. (c) Nief, F.; Ricard, L.;
Mathey, F. Organometallics 1989, 8, 1473-1477.
(8) Hollis, T. K.; Wang, L.-S.; Ahn, Y. J .; Freeman, W. P. Division
of Inorganic Chemistry, Abstracts of Papers of the American Chemical
Society 2001, 222, 540.
(9) Buzin, F.-X.; Nief, F.; Ricard, L.; Mathey, F. Organometallics
2002, 21, 259-263.
(10) Crystal data for C18H24O2P2Ti, 1: M ) 382.21, monoclinic, C2/
c, a ) 16.018(3) Å, b ) 7.4328(13) Å, c ) 32.774(5) Å, R ) 90°, â )
102.498(4)°, γ ) 90°, Z ) 8, T ) 223(2) K, dcalcd ) 1.333 Mg/m3, F(000)
) 1600, µ(Mo KR) ) 0.623 mm-1, λ(Mo KR) ) 0.71073 Å, 3886
independent reflections measured, R1 ) 0.0479 (I > 2.00σ(I)), wR(F2)
) 0.1323 (all data).
(11) Sikora, D. J .; Rausch, M. D.; Rogers, R. D.; Atwood, J . L. J .
Am. Chem. Soc. 1981, 103, 1265-1267. See also the reviews listed in
ref 6 and references therein.
(6) For reviews see: (a) Pez, G. P.; Armor, J . N. Adv. Organomet.
Chem. 1981, 19, 1-50. (b) Erker, G.; Kru¨ger, C.; Mu¨ller, G. Adv.
Organomet. Chem. 1984, 24, 1-39. (c) Sikora, D. J .; Macomber, D. W.;
Rausch, M. D. Adv. Organomet. Chem. 1986, 25, 317-379. (d)
Buchwald, S. L.; Nielsen, R. B. Chem. Rev. 1988, 88, 1047-1058.
(7) For other recent low-valent chemistry of the Ti triad with
heterocyclic ligands see: (a) Boroles: Kiely, A. F.; Nelson, C. M.; Pastor,
A.; Henling, L. M.; Day, M. W.; Bercaw, J . E. Organometallics 1998,
17, 1324-1332. (b) Germoles: Dysard, J . M.; Tilley, T. D. J . Am. Chem.
Soc. 2000, 122, 3097-3105. (c) Boratabenzenes: Ashe, A.; Al-Ahmad,
S.; Kampf, J . Organometallics 1999, 18, 4234-4236.
10.1021/om020961h CCC: $25.00 © 2003 American Chemical Society
Publication on Web 03/07/2003

Low-Valent Ti Bis(phospholyl) Chemistry
Organometallics, Vol. 22, No. 7, 2003 1433
Ta ble 1. Com p a r ison of ν(CO)a v a n d Str u ctu r a l
P a r a m eter s for L2M(CO)2 Ben t Meta llocen es (TMP
) 2,3,4,5-tetr a m eth yl p h osp h olyl)
a
L
ν(CO)av
OC-M-COb
Ctr-M-Ctrb
L2Ti(CO)2
Cpc
1938
1930
1899
87.9
88.7
83.3
138.6
144.0
147.9
TMPd
Cp*c
L2Zr(CO)2
89.2
Cpe
1930
1934
1900
143.4
143.9
147.4
TMPf
Cp*e
91.3
86.3
a
b
d
cm-1
.
deg. c See ref 6c and references therein. This work.
e Ref 11. f See ref 9.
between Cp and Cp* in electron-donating character (π-
basicity). This conclusion is different from that recently
reported for the Zr analogue where (η5-C4Me4P-) was
found to be a weaker donor than Cp.9
F igu r e 1. ORTEP plot of the monoclinic Ti(II) complex
bis(η5-2,3,4,5-tetramethylphospholyl)titanium dicarbonyl,
1, shown at 50% probability thermal ellipsoids. Selected
structural parameters: Ti-Ctr ) 2.066, 2.090 Å; Ti-C(O)
) 2.040, 2.044(3) Å; Ti-P ) 2.5611(10) Å, 2.5825(11) Å;
C-O ) 1.150, 1.147(4) Å; Ctr-Ti-Ctr ) 144°; (O)C-Ti-
C(O) ) 89°; P-Ctr-Ctr-P ) 179.1°.
The difference in apparent donor ability of the tet-
ramethylphospholyl (TMP) ligand relative to Cp and
Cp* in Ti versus Zr complexes presents an intriguing
conundrum. The two phosphametallocenes are isoelec-
tronic and, in gross features, isostructural. Yet, exami-
nation of the subtle structural differences allows an
adequate explanation to be offered, which is analogous
to that recently offered for the ansa-effect in the
monumental paper by Parkin, Bercaw, Green, Keister,
and co-workers.13 Presented in Table 1 is a comparison
of ν(CO)av for the Ti and Zr complexes of Cp, TMP, and
Cp* alongside the key structural parameters (O)C-M-
C(O) and Ctr-M-Ctr. If the change in (O)C-M-C(O)
angle was the dominate factor contributing to the
stretching frequency differences, then the angle for the
TMP complexes should be at different positions relative
to Cp and Cp* for the Ti versus Zr series. The observed
(O)C-M-C(O) angles follow the same order in Ti and
Zr (TMP > Cp > Cp*). In contrast to the OC-M-CO
structural parameters, the Ctr-M-Ctr angle follows a
different pattern in the Ti and Zr series, a pattern that
correlates with the stretching frequencies. The Ctr-M-
Ctr angle increases for the Ti complexes in the order
Cp < TMP < Cp*, which is inversely correlated with
the stretching frequency, whereas the Ctr-M-Ctr
angles for the Zr complexes are observed to vary such
that Cp ≈ TMP < Cp*, and the ν(CO)av frequencies are
The Ti-P distance is greater for the P atom bisecting
the carbonyl groups (2.5825(11) vs 2.5611(10) Å). The
Ti-Ctr distances for the Ti(II) complex are 2.066 and
2.090 Å. The phospholyl ring with the shorter Ti-Ctr
distance has the P atom at the narrow side of the
metallocene wedge, and the one with the P atom
bisecting the (O)C-Ti-C(O) angle has the longer Ti-
Ctr distance. The Ti-Ctr distance is expected to de-
crease for the Ti(IV) analogue since the metal center
is formally more electron deficient. But, the Ti-Ctr
distances in the closest Ti(IV) analogue for which a
crystal structure has been reported, (3,4-dimethyl-
phospholyl)2TiCl2, are on average 2.154 Å compared to
2.088 Å for the Ti(II) complex.5c Part of this increase in
the Ti-Ctr distance for the Ti(IV) analogue can be
attributed to increased steric repulsion between the
phospholyl ligand and Cl compared to the corresponding
interaction with CO (covalent radii: C(sp) ) 0.603 Å
vs Cl ) 0.99 Å).12 In addition, there is a bonding
interaction between the filled metal orbital and the
ligand π*-orbital Ti(II) complex (vide infra).
The CO stretching frequencies for the Ti(II) complexes
of Cp, Cp*, and (η5-C4Me4P) provide a means of compar-
ing the donor ability of each ligand. The CO stretching
frequencies in hexane found for (η5-C4Me4P)2Ti(CO)2
(1959, 1900 cm-1) are between those reported for Cp2-
Ti(CO)2 (1977, 1899 cm-1) and Cp*2Ti(CO)2 (1940, 1858
cm-1).11 The symmetric CO stretching frequency ob-
served for the phosphametallocene is halfway between
that of Cp and Cp* (1940 < 1959 < 1977 cm-1), while
the asymmetric stretch is almost identical to that of Cp
(1899 ≈ 1900 > 1858 cm-1). Considering the average of
the symmetric and asymmetric stretches for each of the
complexes, the order observed is Cp*, 1899 cm-1 < (η5-
C4Me4P), 1930 cm-1 < Cp, 1938 cm-1. These observa-
tions lead to the conclusion that the tetramethylphos-
pholyl ligand (TMP ) (η5-C4Me4P)) is intermediate
inversely correlated with the angle (Cp*, 1900 cm-1
<
TMP, 1934 cm-1 ≈ Cp, 1930 cm-1).
Bent metallocene dicarbonyl ν(CO)av stretching fre-
quencies have been correlated recently with the Ctr-
M-Ctr angle for a large number of zirconocene dicar-
bonyl complexes. The explanation offered was that the
overlap, and therefore the back-bonding, between the
metal orbitals and the “cyclopentadienyl ligand acceptor
orbital” improves as the Ctr-M-Ctr angle decreases.13
The increased back-bonding to the cyclopentadienyl
ligand decreases back-bonding to the CO ligand, result-
ing in higher frequency CO stretches for more bent
complexes. The data for the tetramethylphospholyl
complexes of Ti and Zr are consistent with this explana-
tion. The differences between the apparent donor ability
of TMP relative to Cp and Cp* in the Ti and Zr series
(12) Pauling, L. The Nature of the Chemical Bond and the Structure
of Molecules and Crystals; An Introduction to Modern Structural
Chemistry, 3d ed.; Cornell University Press: Ithaca, NY, 1960; p 224.
(13) Zachmanoglou, C. E.; Docrat, A.; Bridgewater, B. M.; Parkin,
G.; Brandow, C. G.; Bercaw, J . E.; J ardine, C. N.; Lyall, M.; Green, J .
C.; Keister, J . B. J . Am. Chem. Soc. 2002, 124, 9525-9546.

1434 Organometallics, Vol. 22, No. 7, 2003
Hollis et al.
Sch em e 1. Mech a n ism for th e Isom er iza tion of
Heter ocyclop en ta d ien yl Com p lexesa
correlate with the Ctr-M-Ctr angle. The apparent
electron-donating ability of the TMP ligand relative to
Cp and Cp* has been found to vary in the closely related
analogues (η5-C4Me4P)2Zr(CO)2 and (η5-C4Me4P)2Ti-
(CO)2. The source of the difference is attributable to
subtle geometric changes in the Ctr-M-Ctr angle that
change the energy of the a1 (HOMO) orbital. Finally,
the change in apparent ligand donor capacity is not due
to a change in the geometry of the TMP complexes, but
is due to a large Ctr-M-Ctr angle for Cp2Zr(CO)2
(Table 1). The large Ctr-M-Ctr angle for Cp2Zr(CO)2
produces decreased back-bonding to the Cp ligand and
increased back-bonding to the CO, which results in an
“anomalously” low ν(CO)av stretching frequency (1930
cm-1). This geometric flexibility of organometallic com-
plexes makes overgeneralization regarding ligand donor
ability and substituents effects in organometallic com-
plexes easy. The comparisons presented herein suggest
that generalization even between very closely related
isoelectronic complexes (Ti vs Zr) can fail.
a
See ref 15.
ally, heating of a THF solution of chiral dicarbonyl 2 to
50 °C did not result in broadening of the 31P NMR
signals, which clearly indicates a significantly higher
barrier to isomerization of the rac and meso isomers in
the Ti(II) complex. The mechanism of isomerization in
the M(IV) complexes is depicted in Scheme 1. Ring
slippage, inversion, and ring slippage change the abso-
lute configuration of the chiral plane.15 These structures
resemble known metallocene phosphido complexes18 and
the recently reported examples of η1 pyrrolyl com-
plexes.19 Metal oxidation state and d-electron count
appear to play an important role in the isomerization
of phosphatitanocenes.
The influence of metal oxidation state on the rac/meso
isomerization of chiral phosphametallocenes is of fun-
damental interest. Following up our initial report of the
facile thermal isomerization of chiral bis(phospholyl)
complexes of Zr,14 we have recently reported the activa-
tion parameters for the isomerization process in a Ti-
(IV) analogue based on variable-temperature NMR.15,16
The activation barrier for the isomerization of rac/meso-
(η5-C4Me2(Ph)HP)2TiCl2 was found to be ∆Gq ) 11.5
kcal/mol. We undertook the preparation of chiral ana-
logues of 1, as outlined in eq 2, to prepare the chiral
An explanation for the difference between the rates
of isomerization in different oxidation states depends
on the d-orbital in which the d-electrons are found for
the Ti(II) complex (Figures 2 and 3). Employing the C2v
Cp2ML2 orbitals as a first-order model for the phospholyl
analogue, it is possible to picture the interactions of
importance in the ground state and at the transition
state for the isomerization. The a1-orbital pictured in
Figure 2c is the LUMO of the d0 complex (Figure 2a),
and it is the HOMO of a d2 complex (Figure 2b).20 In
the ground state of a Ti(II) complex there is a bonding
interaction between the filled metal a1-like orbital and
a π*-orbital of the phospholyl ligand, as depicted in
Figure 3a for (C4H4P)2Ti(CO)2. This interaction results
in additional ground state stabilization of the chiral Ti-
(II) complex relative to the Ti(IV) analogue. The ad-
ditional bonding interaction contributes to the decreased
M-Ctr distance in the Ti(II) molecules (vide supra). A
shorter M-Ctr distance in the low-valent state is also
observed when comparing Zr(II) and Zr(IV) analogues,
(η5-C4Me4P)2ZrCl2, Zr-Ctr(av) ) 2.284 Å21 and (η5-C4-
Me4P)2Zr(CO)2, Zr-Ctr(av) ) 2.201 Å.9
bis(phospholyl)Ti(II) dicarbonyl complex 2. This Ti(II)
complex has two distinct resonances in the 31P NMR
spectrum (4.13, -13.6 ppm, ratio 24:76 in C6D6) con-
sistent with rac and meso isomers. The 1H NMR
spectrum also revealed the expected duplications due
to rac and meso isomers. The IR spectrum shows two
CO stretches at 1971 and 1914 cm-1 for an Et2O cast
film. In hexane solution, the IR spectrum at 0.5 cm-1
resolution contains five CO stretches (1992, 1981, 1975,
1931, and 1918 cm-1) with a shoulder on the 1981
signal.17 The observation of two sharp signals in the 31
P
NMR spectrum of complex 2 is distinct from the Ti(IV)
dichloro analogue, which has a broadened resonance at
room temperature and does not have discrete signals
for the rac and meso isomers down to 0 °C.15 Addition-
But, additional ground state stabilization of the
Ti(II) complex is not the only contributor to the con-
(14) Hollis, T. K.; Wang, L. S.; Tham, F. J . Am. Chem. Soc. 2000,
122, 11737-11738.
(15) Hollis, T. K.; Ahn, Y. J .; Tham, F. S. Chem. Commun. 2002,
2996-2997.
(18) Baker, R. T.; Whitney, J . F.; Wreford, S. S. Organometallics
1983, 2, 1049-51. Hey-Hawkins, E. Chem. Rev. 1994, 94, 1661-1717.
For a recent reference see: Urnezius, E.; Klippenstein, S. J .; Prota-
siewicz, J . D. Inorg. Chim. Acta 2000, 297, 181-190.
(16) For the first report of an intramolecular isomerization of a
phospinoferrocene see: Curnow, O. J .; Fern, G. M. Organometallics
2002, 21, 2827-2829.
(17) At present we propose that multiple conformations of the rac/
meso isomers are being observed consistent with the report for
(indenyl)2Ti(CO)2. See: ref 6c, p 347. Demerseman, B.; Bouquet, G.;
Bigorgne, M. J . Organomet. Chem. 1975, 93, 199-204. Rausch, M. D.;
Moriarty, K. J .; Atwood, J . L.; Hunter, W. E.; Samuel, E. J . Organomet.
Chem. 1987, 327, 39-54.
(19) Tanski, J . M.; Parkin, G. Organometallics 2002, 21, 587-589.
(20) Adapted from: Albright, T. A.; Burdett, J . K.; Whangbo, M.-H.
Orbital Interactions in Chemistry; Wiley: New York, 1985; p 394-
396. Lauher, J . W.; Hoffmann, R. J . Am. Chem. Soc. 1976, 98, 1729-
42. Green, J . C. Chem. Soc. Rev. 1998, 27, 263-271.
(21) Mathey, F.; Ricard, L. Organometallics 1988, 7, 921-926.

Low-Valent Ti Bis(phospholyl) Chemistry
Organometallics, Vol. 22, No. 7, 2003 1435
To generalize, heterocyclic π-ligands that have a filled
nonbonding orbital (e.g., a lone pair as in the phospha-
metallocenes) will have a stabilizing interaction with a
vacant metal orbital in the planar transition state that
lowers the barrier to isomerization. A d0 metal is not
necessarily required. In contrast, if the metal orbital
that interacts with the lone pair is occupied, a desta-
bilizing interaction occurs that results in a higher
barrier to isomerization. These complexes should not
isomerize as easily. The Ti(II) results reported here and
previous observations are consistent with this conclu-
sion. These occupancy rules are reversed for heterocyclic
π-ligands with vacant nonbonding orbitals. For such
complexes isomerization is predicted to be more facile
if the metal has filled orbitals that can interact with
the vacant orbital of the ligand, producing a two-
electron-two-orbital stabilization at the transition state.
Therefore, d0 complexes are predicted to have higher
barriers to isomerization, whereas metals with high
d-orbital occupancy may have lower barriers to isomer-
ization. But, a low barrier to isomerization is not
guaranteed. An orbital of the proper symmetry, energy,
and orientation must be available to interact with the
vacant ligand orbital.
F igu r e 2. Frontier molecular orbitals for a C2v Cp2ML2
fragment as a model for phospholyl analogues. (a) Orbital
occupation for Ti(IV). (b) Orbital occupation for Ti(II). (c)
The metal-centered symmetry adapted combination for the
a1-orbital is pictured in the ML2 plane with L2 omitted for
clarity. (d) The P-lone pair in a p-orbital interacting with
the a1-orbital at the planar transition state of the inversion
with L2 omitted for clarity.
In summary, we report the first examples of low-
valent phosphatitanocenes and the much slower rates
of isomerization of a chiral Ti(II) d2 complex versus its
Ti(IV) d0 counterpart. An explanation for the difference
in isomerization rate has been advanced. Additional
work is underway to determine the structures of the
chiral Ti(II) complexes and to explore the Ti(II) phos-
phametallocene reactivity.
Exp er im en ta l Section 22
F igu r e 3. Orbital depictions of phosphatitanocenes (P is
orange) with simplified models for the phospholyl ligands
calculated with PCSpartan at the RHF/3-21G* level of (a)
the HOMO of (C4H4P)2Ti(CO)2 (side view) with the crystal
structure conformation, and (b) the geometry optimized
LUMO of (C4H4P)2TiCl2 (viewed from behind the narrow
side of metallocene wedge).
Bis(η5-2,3,4,5-tetr a m eth ylp h osp h olyl)tita n iu m d ica r -
bon yl, 1. 2,3,4,5-Tetramethyl-1-trimethylstannyl phosphole
(1.16 g, 3.83 mmol)5 and pentane (10 mL) were rapidly added
to a solution of TiCl4‚2THF (0.59 g, 1.76 mmol) and CH2Cl2
(20 mL). After stirring 30 min, all volatiles were removed. The
dark brown oil was extracted with pentane (40 mL). The
extract was concentrated, and THF (15 mL) and Mg powder
(86 mg, 3.54 mmol) were added. CO was bubbled through the
dark green solution for 5 min and stirred overnight at 40 °C.
The reaction was concentrated, and the resulting dark violet
residue was extracted with pentane (50 mL). The extract was
concentrated, yielding a red powder (270 mg, 39%). X-ray
quality crystals were grown from a saturated toluene solution
at -35 °C. 1H NMR (C6D6, 300.053 MHz): δ 1.81 (s, 12H,
figurational stability of the chiral Ti(II) complexes. The
a1-like orbital is the LUMO of the Ti(IV) complex
(Figures 2a,c and 3b). In the transition state for inver-
sion the Ti atom is coplanar with the phospholyl ring
and there will be overlap between the metal d-orbital
and the p-π-orbital of P (Figure 2d). In the Ti(IV)
complex the a1-orbital is unoccupied (Figure 2a), and a
two-electron-two-orbital stabilizing interaction occurs,
which lowers the energy of the transition state for
inversion. In contrast, in the Ti(II) complex the bonding
a1-orbital is occupied. Therefore, to undergo inversion
the energy of the additional bonding interaction must
be overcome. The major contributor to configurational
stability is the interaction between the filled P-lone pair
and the filled a1-orbital at the transition state that
results in a four-electron-two-orbital destabilizing in-
teraction, which increases the activation barrier for
inversion. Thus for the Ti(II) complex additional ground
state stabilization and transition state destabilization
both contribute to the increased configurational stabil-
ity. This first-order explanation illustrates the impor-
tance of metal d-orbital occupancy in isomerization of
heterocyclopentadienyl and other heterocyclic π-ligands.
3
â-CH3), 1.59 (d, J PH ) 9.75 Hz, 12H, R-CH3). 13C{1H} NMR
2
(C6D6, 75.475 MHz): δ 248. 4(TiCO), 123.0 (d, J PC ) 4.9 Hz,
1
2
â-C), 117.3 (d, J PC ) 52.5 Hz, R-C), 15.5 (d, J PC ) 23.2 Hz,
R-CH3) 15.3 (m, â-CH3). 31P{1H} NMR (C6D6, 121.474 MHz):
δ 28.1 (s). IR (hexane, solution): 1959 (CO), 1900 (CO) cm-1
;
(hexane, cast film) 1953 (CO), 1889 (CO) cm-1. Anal. Calcd
for C18H24O2P2Ti: C, 56.57; H, 6.33; P, 16.21. Found: C, 56.45;
H, 6.28; P, 15.96.
Bis(3,4-d im eth yl-2-p h en ylp h osp h olyl)tita n iu m d ica r -
bon yl, 2, was prepared in analogy with 1 from the isolated
dichloride precursor.15 A brown powder was obtained (75 mg,
35%). Isomer 1: 1H NMR (C6D6, 300.053 MHz): δ 7.50-6.94
(m, 10H, C6H5), 4.76 (d, 2J PH )38.0 Hz, 2H, PCH), 1.98 (s, 6H,
CH3), 1.58 (s, 6H, CH3). 13C{1H} NMR (C6D6, 75.475 MHz): δ
245.7 (m, TiCO). 31P{1H} NMR (C6D6, 121.474 MHz): δ -13.63
(s). Isomer 2: 1H NMR (C6D6, 300.053 MHz): δ 7.50-6.94 (m,
(22) For general experimental details see ref 14. Elemental analyses
were performed by Desert Analytics.

1436 Organometallics, Vol. 22, No. 7, 2003
Hollis et al.
2
10H, C6H5), 4.43 (d, J PH ) 38.0 Hz, 2H, PCH), 1.85 (s, 6H,
G.A.A.N.N. Fellowship for Y.J .A. We thank Professor
Chris Reed and Dr. Kee-Chan Kim for helpful sugges-
tions and assistance collecting IR data.
CH3), 1.61 (s, 6H, CH3). 13C{1H} NMR (C6D6, 75.475 MHz): δ
243.8 (m, TiCO). 31P{1H} NMR (C6D6): δ 4.13 (s). IR (Et2O,
cast film): 1971(CO), 1914(CO) cm-1; (hexane) 1992, 1981,
1975, 1931, 1918(CO) cm-1. Anal. Calcd for C26H24O2P2Ti: C,
65.29; H, 5.06; P, 12.95. Found: C, 65.34; H, 5.35; P, 12.81.
Su p p or tin g In for m a tion Ava ila ble: Full details of the
crystal structure determination including tables of structural
data and refinement, atomic coordinates, bond distances, bond
angles, torsion angles, and anisotropic displacement param-
eters for 1. This material is available free of charge via the
Internet at http://pubs.acs.org.
Ack n ow led gm en t is made to the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society for partial support of this re-
search. We thank the Department of Education for a
OM020961H