Origin of 1,4-regiochemistry in the dicouplings of ketones with 6,6-dimethylcyclohexadienyl complexes of titanium and zirconium: A mechanism arising from five-electron donation by alkoxide ligands

Article abstract of DOI:10.1021/om058018b

In an earlier study of the coupling reactions of Ti(C₅H ₅)(6,6-dmch)(PMe₃) (dmch = dimethylcyclohexadienyl) with ketones, the initially formed metal complex could not be isolated, but after hydrolysis an unprecedented 1,4-dicoupling product was isolated. To gain further insight into the origin of this regiochemistry, the reaction of a zirconium analogue, Zr(C₅H₅)(6,6-dmch)(PMe₃)₂, with acetone was investigated. A solid dimeric product could be isolated and structurally characterized, which revealed that both acetones had coupled, with the unusual 1,4-regiochemistry. This then necessitated the proposal of a new mechanism for these reactions, in which an alkoxide ligand, generated from a coupling of the initially incorporated ketone, serves as a formal five-electron donor, thereby altering the mode of coordination of the diene fragment. This proposal has been supported by the characterization of a mono(benzophenone) coupling product of Ti(C₅H₅)(6,6-dmch)(PMe₃). This coupling product possessed both a large Ti-O-C angle, indicative of at least a partial contribution of a resonance form in which the alkoxide ligand serves as a five-electron donor, and the expected η²-diene coordination.

Full text of DOI:10.1021/om058018b

3982
Organometallics 2005, 24, 3982-3986
Origin of 1,4-Regiochemistry in the Dicouplings of
Ketones with 6,6-Dimethylcyclohexadienyl Complexes of
Titanium and Zirconium: A Mechanism Arising from
Five-Electron Donation by Alkoxide Ligands
Rehan Basta, Atta M. Arif, and Richard D. Ernst*
Department of Chemistry, University of Utah, 315 S. 1400 E.,
Salt Lake City, Utah 84112-0850
Received March 31, 2005
In an earlier study of the coupling reactions of Ti(C5H5)(6,6-dmch)(PMe3) (dmch )
dimethylcyclohexadienyl) with ketones, the initially formed metal complex could not be
isolated, but after hydrolysis an unprecedented 1,4-dicoupling product was isolated. To gain
further insight into the origin of this regiochemistry, the reaction of a zirconium analogue,
Zr(C5H5)(6,6-dmch)(PMe3)2, with acetone was investigated. A solid dimeric product could be
isolated and structurally characterized, which revealed that both acetones had coupled, with
the unusual 1,4-regiochemistry. This then necessitated the proposal of a new mechanism
for these reactions, in which an alkoxide ligand, generated from a coupling of the initially
incorporated ketone, serves as a formal five-electron donor, thereby altering the mode of
coordination of the diene fragment. This proposal has been supported by the characterization
of a mono(benzophenone) coupling product of Ti(C5H5)(6,6-dmch)(PMe3). This coupling
product possessed both a large Ti-O-C angle, indicative of at least a partial contribution
of a resonance form in which the alkoxide ligand serves as a five-electron donor, and the
expected η2-diene coordination.
Introduction
Pentadienyl complexes of titanium and zirconium
have been found to undergo coupling reactions with a
wide variety of unsaturated molecules, including al-
kynes,1 ketones, aldehydes, imines, nitriles, and isoni-
triles. Quite often, multiple couplings can be achieved;
in fact, more often than not, these occur in preference
to single coupling processes. Until recently, such cou-
plings have inevitably involved the terminal dienyl
carbon atoms first. As a result, numerous 1,5-dicoupling
processes have been observed for ketones, imines,
nitriles, and even combinations of these substrates.2
However, it was subsequently found that Ti(C5H5)(6,6-
dmch)(PMe3), 1, underwent di(coupling) reactions with
ketones with unprecedented 1,4-regiochemistry.3 Al-
though the resulting metal complex could not readily
be isolated and characterized, the organic product could
be removed from the metal by either hydrolysis or a
combination of oxidation and hydrolysis and subse-
quently characterized.
To explain the 1,4-regiochemistry, a mechanism was
suggested in which 2 was proposed to be the key
intermediate.4 In this species, only one of the ketones
(1) (a) Wilson, A. M.; Waldman, T. E.; Rheingold, A. L.; Ernst, R.
D. J. Am. Chem. Soc. 1992, 114, 6252. (b) Tomaszewski, R.; Hyla-
Kryspin, I.; Mayne, C. L.; Arif, A. M.; R. Gleiter, R.; Ernst, R. D. J.
Am. Chem. Soc. 1998, 120, 2959.
(2) (a) Waldman, T. E.; Wilson, A. M.; Rheingold, A. L.; Mele´ndez,
E. M.; Ernst, R. D. Organometallics 1992, 11, 3201. (b) Kralik, M. S.;
Rheingold, A. L.; Hutchinson, J. P.; Freeman, J. W.; Ernst, R. D.
Organometallics 1996, 15, 551. (c) Tomaszewski, R.; Wilson, A. M.;
Liable-Sands, L.; Rheingold, A. L.; Ernst, R. D. J. Organometal. Chem.
1999, 583, 42. (d) Tomaszewski, R.; Rheingold, A. L.; Ernst, R. D.
Organometallics 1999, 18, 4174. (e) See also: Paz-Sandoval, M. A.;
Coyotzi, R. S.; Villareal, N. Z.; Ernst, R. D.; Arif, A. M. Organometallics
1995, 14, 1044.
had undergone coupling, while the second simply coor-
dinated to the metal center. Such a possibility found
support from the fact that reactions with some diones
led to one carbonyl group coupling to the dienyl frag-
ment, while the other appeared to simply coordinate,
(3) Wilson, A. M.; West, F. G.; Arif, A. M.; Ernst, R. D. J. Am. Chem.
Soc. 1995, 117, 8490.
(4) Wilson, A. M.; West, F. G.; Rheingold, A. L.; Ernst, R. D. Inorg.
Chim. Acta 2000, 300-302, 65.
10.1021/om058018b CCC: $30.25 © 2005 American Chemical Society
Publication on Web 07/01/2005

1,4-Regiochemistry in the Dicouplings of Ketones
Organometallics, Vol. 24, No. 16, 2005 3983
as after hydrolysis the latter had been converted to a
secondary alcohol.5 In any event, it was presumed that
after hydrolysis of the other dienyl terminus of 2 an allyl
complex (3) would result, and coupling of the remaining
coordinated ketone with the 6,6-dmch ligand, leading
to the more usual 1,5-regiochemistry. Herein we report
the unexpected outcome of such an attempt and ad-
ditional studies designed to help elucidate the origin of
the 1,4-regiochemistry in some of these reactions. As
will be shown, it appears that the potential of alkoxide
ligands, resulting from the ketone couplings, to serve
as five-electron donors7 can be implicated as the source
of the unusual regiochemistry.
Experimental Section
All synthetic procedures and handling of compounds were
carried out under a nitrogen atmosphere. Solvents were dried
using activated alumina under a nitrogen atmosphere. Spec-
troscopic data were acquired as described previously.8 Elemen-
tal analyses were obtained from Complete Analysis Labora-
tories, Inc.
Structural data were obtained with a Nonius Kappa CCD
autodiffractometer. Single crystals were protected from the
atmosphere with Paratone oil. Initial structural solutions were
obtained using direct methods, with further atoms located via
difference Fourier maps.9 For the zirconium compound, all non-
hydrogen atoms were refined anisotropically, while all hydro-
gen atoms were located and successfully refined isotropically.
The refinement for the titanium compound was carried out
similarly, except that the PMe3 hydrogen atoms were placed
in idealized positions, as were three of the C5H5 ligand’s
hydrogen atoms. Half a molecule of ether was present in the
asymmetric unit. Its non-hydrogen atoms were refined aniso-
tropically, while its hydrogen atoms were placed in idealized
positions.
{Zr(C5H5)[6,6-dmch-(C3H6O)2]}2. To a solution of CpZr-
(dmch)(PMe3)210 (0.78 g, 1.9 mmol) in 20 mL of hexane at room
temperature was added acetone (0.30 mL, 4.2 mmol). The
resulting yellow solution was stirred for 1 h. The solvent was
removed in vacuo, and the crude product was extracted using
ca. 40 mL of hexane. The solution was filtered through a Celite
pad on a coarse frit. The product was then crystallized by
concentration of the filtrate to ca. 4-5 mL and placement in
a freezer at -20 °C, which afforded yellow crystals (0.62 g,
52% yield). Anal. Calc for C19H28O2Zr: C, 60.11; H, 7.43.
Found: C, 60.10; H, 7.67.
ketone with an allyl terminus would lead to the ob-
served 1,4-regiochemistry and ultimately elimination of
an enediol such as 4a. As stated above, a sequential
combination of oxidation and hydrolysis could also be
employed, leading to an enetriol (e.g., 5a). Notably, the
stereochemistry of the product revealed that the oxida-
tion of the organic fragment occurred on the side
opposite to the metal, i.e., with inversion rather than
retention of stereochemistry. Also notable is that the
relative configurations of three chiral centers are set
selectively by this overall reaction, while five relative
configurations may be selectively set for some alde-
hydes.
Ti(C5H5)(η2-dmch-Ph2CO)PMe3. To an orange-brown so-
lution of Ti(C5H5)(6,6-dmch)(PMe3)3 (0.20 g, 0.68 mmol) in 10
mL of diethyl ether was added benzophenone (0.12 g, 0.68
mmol). The reaction mixture was left overnight without
stirring, resulting in a green solution. Placement of the
reaction flask in a -30 °C freezer resulted in the formation of
dark green crystals (0.13 g, 40% yield). 1H NMR (benzene-d6,
ambient): δ 0.53 (s, 3H, exo CH3), 0.58 (d, 9H, J ) 4.8 Hz,
PMe3), 1.53 (s, 3H, endo CH3), 1.84 (m, 1H4), 2.06 (d, 1H5, J )
9.6 Hz), 3.76 (d, 1H1, J ) 6.3 Hz), 5.48 (m, 1H2), 5.83 (s, 5H,
Cp), 6.38 (m, 1H3), 6.90 (m, 2H, Ph), 7.07 (m, 4H, Ph), 7.46
(m, 4H, Ph). 13C NMR (benzene-d6, ambient): δ 16.5 (broad
signal, PMe3), 31.6 (q, overlapping, 1C, J ) 126 Hz, exo CH3),
34.9 (q, overlapping, 1C, J ) 125 Hz, endo CH3), 35.6 (s, 1C6),
54.2 (d, 1C1, J ) 126 Hz), 62.9 (dd, 1C4, J ) 160, 10 Hz), 81.3
(d, 1C5, J ) 153 Hz), 102.7 (s, 1C9), 108.2 (d of quintets, 5C,
J ) 171, 7 Hz, Cp), 114.3 (dt, 1C2, J ) 160, 7 Hz), 124.0-
While the occurrence of the 1,4-regiochemistry was
much more interesting than simply finding another
example of the expected 1,5-regiochemistry, there was
reason to seek 1,5-regiochemistry in this case. It can
be noted that the 1,5-dicoupling of an appropriate dione
to a 6,6-dmch ligand would yield a fragment similar to
the AB ring system of Taxol.6 Hence, it was of interest
to coax a dmch metal complex to undergo the more
mundane 1,5-dicoupling process with simple ketones as
a precursor to a study of reactions involving more
complicated diones. As zirconium is more electropositive
and oxophilic than titanium, it would have to be
expected that in an analogue of 2 the presence of the
zirconium center would lead to a rapid coupling of the
(7) (a) Tomaszewski, R.; Arif, A. M.; Ernst, R. D. J. Chem. Soc.,
Dalton Trans. 1999, 1883. (b) Dobado, J. A.; Molina, J. M.; Uggla, R.;
Sundberg, M. R. Inorg. Chem. 2000, 39, 2831.
(8) Newbound, T. D.; Stahl, L.; Ziegler, M. L.; Ernst, R. D. Orga-
nometallics 1990, 9, 2962.
(9) (a)Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.;
Giacovazzo, C.; Guagliardi, A.; Moliteni, A. G. G.; Polidori, G.; Spagna,
R. SIR97 (Release 1.02). (b) Sheldrick, G. M. SHELX97 (Release 97-
2); University of Go¨ttingen, Germany.
(10) Kulsomphob, V.; Harvey, B. G.; Arif, A. M.; Ernst, R. D. Inorg.
Chim. Acta 2002, 334, 17.
(5) Wilson, A. M. Ph.D. Thesis, University of Utah, 1994.
(6) Herz, W., Falk, H., Kirby, G. W., Eds. Progress in the Chemistry
of Organic Natural Products; Springer: Berlin, 2002; Vol. 84.

3984 Organometallics, Vol. 24, No. 16, 2005
Basta et al.
Table 1. Crystallographic Data for
{Zr(C5H5)(6,6-dmch-(C3H6O)2]}2 and
Ti(C5H5)(η2-dmch-Ph2CO)PMe3
formula
fw
C19H28O2Zr
379.63
triclinic
P1h
yellow
C31H40O1.5PTi
515.50
monoclinic
C2/c
green
34.3999(10)
10.2152(3)
18.2981(5)
90
cryst syst
space group
color
a (Å)
9.4042(2)
9.4956(2)
11.7123(3)
89.4653(9)
70.9506(10)
63.1912(15)
200(1)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
temp (K)
Z
120.8700(13)
90
150(1)
2
8
R (2σ)
wR2 (2σ)
GOF
0.0243
0.0429
0.0553
0.0973
1.063
1.037
127.6 (10C, Ph), 131.6 (dt, 1C, J ) 149, 5 Hz, C3), 151.2 (t, 1C,
J ) 7 Hz, Ph), 152.5 (t, 1C, J ) 7 Hz, Ph). Anal. Calc for C29-
H35OPTi: C, 72.79; H, 7.26. Found: C, 72.83; H, 7.58.
Figure 1. Perspective view of {Zr(C5H5)[6,6-dmch-(C3H6-
O)2]}2.
Results and Discussion
In contrast to the reaction of 1 with ketones, which
led to oily products, the reaction of Zr(C5H5)(6,6-dmch)-
(PMe3)2 with acetone led to a crystalline product.
Analytical, spectroscopic, and structural characteriza-
tion revealed that, as expected, both acetones had
coupled to each dmch ligand on a given zirconium
center, but the unusual 1,4-regiochemistry observed
earlier was, unexpectedly, once again present (6, Figure
1). It therefore became apparent that the 1,4-regiochem-
istry for titanium was not the result of a delayed second
coupling step, but instead represented an actually
preferred regiochemistry for a dicoupling process of
ketones with either a titanium or zirconium 6,6-dmch
complex.
zirconium centers would attain an 18-electron configu-
ration through π (three electron) alkoxide coordination
from both O1 and O2. The Zr1-O2-C12 angle of
117.64(9)° is in fact fairly close to the ideal value of 120°
expected for the formal sp2 hybridization on O2.11
However, the significantly larger Zr1-O1-C9 angle,
144.93(11)°, and the accompanying shorter Zr-O1
distance of 1.923(1) Å suggest a stronger interaction for
O1. Given earlier results establishing the ability of
alkoxide ligands to serve as five-electron donors,7 it can
be assumed that a second π interaction between O1 and
Zr1 is occurring in competition with a (primarily) single
π interaction between O2 and Zr1, and likely also with
π Zr-Cp interactions.2c,12
The large Zr1-O1-C9 angle, associated with the
ketone that would have been involved in the first
coupling step, appeared to provide some insight into the
origin of the 1,4-regiochemistry. Since 1,5-regiochem-
istry has been observed for couplings of imines with 6,6-
dmch ligands,13 it was clear that it was the initially
incorporated, and coupled, ketone that had to be re-
sponsible. In examining the potential implications of a
large Zr1-O1-C9 angle, brought about by an attempt
for a species such as 7 (n probably ) 1) to achieve an
Before continuing any mechanistic discussion, we can
first consider the structural result for 6 and its implica-
tions for the regiochemical issues. One of the interesting
ramifications of the 1,4-regiochemistry is the formation
of a localized Zr-C(5) bond, noticeably shorter than the
Zr-C(Cp) bonds (2.353(2) vs ca. 2.59 Å), leading also to
a localized C2dC3 bond (1.330(3) Å). More telling,
however, are the parameters associated with the Zr-O
coordination. As a starting point, it can be observed that
the shortest Zr-O distances involve those oxygen atoms
that would be expected to be coordinated to Zr in the
initial monomeric units (e.g., the labeled O1 and O2 for
Zr1). Considering these as forming single bonds to Zr,
with lone pair coordination from the other attached O2
center (O2′), would lead to a 14-electron configuration
for the complex. It might then be expected that the
18-electron configuration, it appeared that the large Zr-
O-C angle would likely result in the Zr center being
(11) (a) A Ti-O-C value of ca. 133° has been proposed as the ideal
for three-electron donation.11b (b) Huffman, J. C.; Moloy, K. G.;
Marsella, J. A.; Caulton, K. G. J. Am. Chem. Soc. 1980, 102, 3009.

1,4-Regiochemistry in the Dicouplings of Ketones
Table 2. Pertinent Bonding Parameters for {Zr(C5H5)[6,6-dmch-(C3H6O)2]}2
Bond Distances (Å)
Organometallics, Vol. 24, No. 16, 2005 3985
Zr-C15
Zr-C16
Zr-C17
Zr-C18
Zr-C19
2.554(2)
2.607(2)
2.638(2)
2.612(2)
2.550(2)
C15-C16
C15-C19
C16-C17
C17-C18
C18-C19
1.398(3)
Zr-O1
Zr-O2
Zr-O2′
Zr-C5
C1-C2
1.9230(12)
2.1421(11)
2.2471(11)
2.3534(17)
1.512(2)
C1-C6
C2-C3
C3-C4
C4-C5
C5-C6
1.548(3)
1.330(3)
1.501(3)
1.531(2)
1.556(2)
1.402(3)
1.404(3)
1.403(3)
1.399(3)
Bond Angles (deg)
Zr-O1-C9 144.93(11)
Zr-O2-Zr′
O1-Zr-O2
O1-Zr-O2′
O1-Zr-C5
O2-Zr-O2′
O2-Zr-C5
O2′-Zr-C5
110.48(5)
C9-C1-C6
116.80(15)
124.85(17)
124.49(17)
111.01(15)
111.43(14)
109.08(14)
107.64(14)
118.21(5)
93.38(5)
83.60(6)
69.52(5)
75.98(5)
138.91(5)
Zr-O2-C12
Zr-O2′-C12′
Zr-C5-C4
Zr-C5-C6
O1-C9-C1
C9-C1-C2
117.64(9)
131.61(9)
107.74(11)
115.24(11)
109.57(14)
112.54(14)
C1-C2-C3
C2-C3-C4
C3-C4-C5
C4-C5-C6
C5-C6-C1
C6-C1-C2
Table 3. Pertinent Bonding Parameters for Ti(C5H5)[η2-dmch-Ph2CO]PMe3
Bond Distances (Å)
Ti-C22
Ti-C23
Ti-C24
Ti-C25
Ti-C26
2.334(3)
2.317(2)
2.369(2)
2.438(2)
2.424(3)
Ti-P
2.6150(7)
1.8219(13)
2.235(2)
2.158(2)
1.509(3)
C1-C6
C2-C3
C3-C4
C4-C5
C5-C6
1.562(3)
1.342(3)
1.445(3)
1.430(3)
1.530(3)
C22-C23
C22-C26
C23-C24
C24-C25
C25-C26
1.392(5)
1.361(6)
1.379(4)
1.370(4)
1.338(5)
Ti-O
Ti-C4
Ti-C5
C1-C2
Bond Angles (deg)
P-Ti-O
Ti-O-C9
O-C9-C1
96.02(4)
140.68(11)
107.37(14)
C9-C1-C2
C9-C1-C6
C2-C1-C6
C1-C2-C3
109.52(15)
C2-C3-C4
123.5(2)
116.9(2)
119.0(2)
111.2(2)
115.61(15)
108.61(16)
122.33(19)
C3-C4-C5
C4-C5-C6
C5-C6-C1
directed away from a location in which it could interact
with the entire diene fragment, particularly C2 and C3,
thus leading to attachment of Zr1 to only the C4dC5
bond of the C2-C5 diene fragment. In such a situation,
there is no longer an electronic difference between C4
and C5 that would favor coupling of C5, as there would
be for a diene. Coupling at C4 could then be favored
due to steric hindrance to coupling at C5 that the
C(CH3)2 bridge may provide, or due to electronic influ-
ences (vide infra).
crystalline form and characterized (see Experimental
Section). In addition to analytical and spectroscopic
To gain support for the above mechanistic proposals,
it became of interest to isolate a mono(ketone) coupling
product that should show η2-coordination for the diene
fragment, as opposed to the η4-coordination observed for
mono(imine) products.14 Unfortunately, in most cases
with ketones, it appeared that the second coupling took
place more rapidly than the first, making it seem
unlikely that such a species could be isolated. Neverthe-
less, for Ti(C5H5)(6,6-dmch)(PMe3) it was observed that
reaction with benzaldehyde yielded a mono(coupling)
product,3,4 and attempts were therefore made to isolate
this species. However, that complex was found to be very
difficult to handle, being quite susceptible to protona-
tion, perhaps even from hydroxyl groups on the silica
surface of the glassware. After a number of attempts
with a variety of ketones and aldehydes, it was found
that benzophenone yielded a (deep green) product (8,
Figure 2) that, while also being quite susceptible to
oxidation and hydrolysis, could at least be isolated in
data, an X-ray structural determination clearly revealed
that a single benzophenone had coupled to the 6,6-dmch
fragment, leading to alkoxide coordination characterized
by a short Ti-O bond (1.822(1) Å) and a wide Ti-O-C
(12) It seems likely that many of the complexes previously reported
to have M-O-C angles exceeding 120° actually will have partial π
donations from both of the oxygen lone pairs, with other bonding
interactions suffering as a result.2c This should be observable through
appropriate theoretical studies. An analogous situation may be con-
sidered to exist for C5H5 compounds such as Ni(C5H5)2 or Zr(C5H5)3-
(η1-C5H5).
(13) Basta, R.; Ernst, R. D.; Arif, A. M. J. Organomet. Chem. 2003,
683, 64.
(14) (a) Tomaszewski, R.; Rheingold, A. L.; Ernst, R. D. Organome-
tallics 1999, 18, 4174. (b) Pillet, S.; Wu, G.; Kulsomphob, V.; Harvey,
B. G.; Ernst, R. D.; Coppens, P. J. Am. Chem. Soc. 2003, 125, 1937.
Figure 2. Perspective view of Ti(C5H5)(η2-dmch-Ph2CO)-
PMe3.

3986 Organometallics, Vol. 24, No. 16, 2005
Basta et al.
angle, 140.7(1)°. Particularly important, however, is that
the metal center is coordinated only by the C4dC5 bond
of the diene unit, as had been surmised. An additional
indication of this can be gained from the comparison of
the C2-C3 and C4-C5 distances, 1.342(3) vs 1.430(3)
Å, respectively. It is of further note that there is also
coordination by phosphine, which impacts on at least
two issues. First, relative to an electron count, the
presence of the phosphine would allow the titanium
center to achieve a 14-electron configuration were the
alkoxide to serve just as a one-electron donor. Quite
clearly, the large Ti-O-C angle is consistent with π
donation by oxygen, and from a comparison with data
for other complexes, it can be expected that at least
partial donations are occurring from both oxygen lone
pairs. In any event, the phosphine ligand may be
providing one other important bit of information. Should
the second coupling step involve replacement of the
coordinated phosphine by the second equivalent of
ketone, the ketone would probably coordinate like the
phosphine, in a position closer to C4 than to C5, hence
favoring the observed 1,4-regiochemistry. It is notewor-
thy that the Ti-C4 distance of 2.235(2) Å is significantly
longer than that for Ti-C5, 2.158(2) Å, which may also
reflect a problem for the titanium center interacting not
only with C2 and C3 but C4 as well, thereby again
making it more favorable for coupling to occur at C4
rather than C5.
With the origin of the unusual 1,4-regiochemistry
seeming to be resolved, some attention may be given to
the impact of related regiochemical considerations in
other situations. First, while the dicoupling reactions
of either imines or ketones with titanium or zirconium
6,6-dmch complexes lead to different regiochemistries,
it is interesting that mixed imine/ketone couplings
should be able to lead to either regiochemistry, depend-
ing on which substrate is introduced first. Thus, an
imine/ketone sequence should lead to the usual 1,5-
regiochemistry, while the opposite sequence should
result in 1,4-regiochemistry. Second, given the 1,4-
regiochemistry in the dicoupling reactions of acetone
with 1 or Zr(C5H5)(6,6-dmch)(PMe3)2, it might be ques-
tioned why the dicoupling of ketones to the Ti(C5H5)-
(2,4-C7H11)(PEt3) complex (9, C7H11 ) dimethylpenta-
dienyl) was found to lead to the usual 1,5-regiochemistry.
In this case, even though an η2-diene intermediate may
again be proposed, the steric properties of 2,4-C7H11 and
6,6-dmch differ dramatically. As the 2,4-C7H11 ligand
does not possess the steric hindrance of the edge-bridge
in 6,6-dmch, but instead has a methyl substituent on
the C4 position, one could easily expect the C5 position
to engage more readily in the second coupling step.
Conclusions
The unusual 1,4-regiochemistry observed in the di-
coupling reactions of ketones or aldehydes with 6,6-
dmch complexes of titanium or zirconium demonstrates
that the previously established ability of an alkoxide
ligand to serve as a five-electron donor can play a
determining role in the outcome of a chemical reaction.
Concerning the original goal of achieving 1,5-regiochem-
istry for the dicoupling of a dione with a 6,6-dmch
ligand, it seems ironic that this could still be achievable
through a straightforward reaction of either a titanium
or zirconium 6,6-dmch complex with a dione, as the
tether between the two carbonyl groups could prevent
the second carbonyl group from approaching the C4 end
of the coordinated olefin. However, there is a regio-
chemical issue that would need to be addressed, in that
two orientations could be adopted by the incoming dione,
one with the tether directed toward the other dienyl
terminus, as desired, but the other having the opposite
orientation. Regiochemical results for imine couplings
have demonstrated that the less bulky substituent tends
to be directed toward the other terminus, so that one
could perhaps employ a particularly large substituent
such as SiMe3 or t-C4H9 on the ends of the dione. As
intramolecular dicoupling reactions between 6,6-dmch
ligands and diimines tethered at least through the
nitrogen atoms have been realized,15 there is every
reason to expect that diones should also be capable of
intramolecular dicoupling reactions.
Acknowledgment is made to the Donors of the
American Chemical Society Petroleum Research Fund
and the University of Utah for partial support of this
work. We thank Dr. Ben Harvey for experimental
assistance.
Supporting Information Available: A listing of posi-
tional coordinates, anisotropic thermal parameters, and ad-
ditional bonding parameters. This material is available free
of charge via the Internet at http://pubs.acs.org.
OM058018B
(15) Basta, R. Ph.D. Thesis, University of Utah, 2004.