Mechanistic studies of the palladium-catalyzed ring opening of oxabicyclic alkenes with dialkylzinc

Article abstract of DOI:10.1021/ja010498k

The mechanism of the palladium-catalyzed ring opening of oxabicyclic alkenes with dialkylzinc has been studied. Experiments which rule out a π-allyl mechanism were carried out. Trapping carbometalated products and synthesis and successful reaction of alkyl palladium species provided strong evidence in favor of an enantioselective carbopalladation as the key step in the mechanism. The studies also suggest that a cationic palladium species is responsible for carbopalladation of the alkene. The combination of palladium and dialkylzinc is unique in that the dialkylzinc functions both in the transmetalation to palladium and as a Lewis acid in forming the reactive cationic palladium species.

Full text of DOI:10.1021/ja010498k

6834
J. Am. Chem. Soc. 2001, 123, 6834-6839
Mechanistic Studies of the Palladium-Catalyzed Ring Opening of
Oxabicyclic Alkenes with Dialkylzinc
Mark Lautens,* Sheldon Hiebert, and Jean-Luc Renaud
Contribution from the Department of Chemistry, UniVersity of Toronto,
Toronto, Ontario, Canada M5S 3H6
ReceiVed February 23, 2001
Abstract: The mechanism of the palladium-catalyzed ring opening of oxabicyclic alkenes with dialkylzinc
has been studied. Experiments which rule out a π-allyl mechanism were carried out. Trapping carbometalated
products and synthesis and successful reaction of alkyl palladium species provided strong evidence in favor of
an enantioselective carbopalladation as the key step in the mechanism. The studies also suggest that a cationic
palladium species is responsible for carbopalladation of the alkene. The combination of palladium and dialkylzinc
is unique in that the dialkylzinc functions both in the transmetalation to palladium and as a Lewis acid in
forming the reactive cationic palladium species.
Introduction
Scheme 1
The use of stereochemically rigid oxabicyclic templates to
achieve stereoselective functional group introduction in car-
bocyclic products has been investigated in our laboratory for
some years. The products from ring opening of these substrates
are potentially very useful in the synthesis of many natural
products.1 Recently we reported a highly enantioselective
alkylative ring opening with dialkylzinc reagents catalyzed by
palladium, which greatly extends the utility of the ring-opening
methodology (Scheme 1).2 We now report detailed studies on
the mechanism of this transformation, which indicate that an
enantioselective carbometalation is the predominant pathway
for the reaction.
Carbopalladation and ring opening or ionization and alkyla-
tion were considered to be the two most likely reaction
pathways. The enantiodetermining step would then either
involve an enantioselective carbopalladation of the alkene or a
selective ionization of the bridging carbon-oxygen bond. A
reversible ionization and enantioselective alkylation was con-
sidered unlikely due to the release of strain upon opening the
bridging carbon-oxygen bond.
Scheme 2
As shown in Scheme 2, a Pd(II) alkyl species could be
involved in a carbopalladation of the alkene. Fiaud3 and
Hayashi4 both propose a carbopalladation mechanism in their
Pd(0)-catalyzed addition of aryl and alkenyl halides to bicyclic
alkenes. Furthermore, Brookhart has also demonstrated that alkyl
insertion into ethylene from a cationic Pd(II) species occurs in
palladium-catalyzed polymerizations.5 In the substrates we
studied carbopalladation would be followed by elimination of
the â-oxygen.6
allylic leaving groups, palladium catalysts and chiral ligands
could also be occurring.7 We observe overall retention, so a
retention-retention or a double inversion pathway would be
required. The intermediate formed, following the ionization, is
closely related to the π-allyl species proposed by Trost in his
work on desymmetrization of meso-diesters.7b However, the
regioselectivity reported by Trost is the product from attack of
the nucleophile distal to the alkoxide. No trace of this isomer
was detected in our studies regardless of the substrate.
Alternatively, an allylic substitution via a π-allyl palladium
intermediate, in analogy with the well-documented results using
(1) (a) Lautens, M. Synlett 1993, 177-185. (b) Chiu, P.; Lautens, M.
Top. Curr. Chem. 1997, 190, 1-85.
(2) (a) Lautens, M.; Renaud, J.-L.; Hiebert, S. J. Am. Chem. Soc. 2000,
122, 1803-4. (b) Lautens, M.; Hiebert, S.; Renaud, J.-L. Org. Lett. 2000,
3, 1971-3.
(6) For other examples of â-oxygen elimination by palladium see: (a)
Larock, R.; Ding, S. J. Org. Chem. 1993, 58, 804. (b) Larock, R.; Ding, S.;
Tu, C. Synlett 1993, 145.
(7) (a) Matsushita, H.; Negishi, E. J. Am. Chem. Soc. 1981, 103, 2882.
(b) Trost, B. M.; Van Vranken, D. L. Chem. ReV. 1996, 96, 395. (c) Pfaltz,
A.; Lautens, M. In ComprehensiVe Asymmetric Catalysis; Jacoben, E. N.;
Pfaltz, A.; Yamamoto, H., Eds.; Springer-Verlag: Berlin, 1999; p 833.
(3) Moinet, C.; Fiaud, J.-C. Tetrahedron Lett. 1995, 36, 2051.
(4) Ozawa, F.; Kobatake, Y.; Kubo, A.; Hayashi, T. J. Chem. Soc., Chem.
Commun. 1994, 1323.
(5) (a) Rix, F.; Brookhart, M.; White, P. J. Am. Chem. Soc. 1996, 118,
4746. (b) Mecking, S.; Johnson, L.; Wang, L.; Brookhart, M. J. Am. Chem.
Soc. 1998, 120, 888.
10.1021/ja010498k CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/22/2001

Pd-Catalyzed Ring Opening of Oxabicyclic Alkenes with Dialkylzinc
J. Am. Chem. Soc., Vol. 123, No. 28, 2001 6835
Scheme 3
Scheme 4
Results and Discussion
1. Investigations into the π-Allyl Pathway. (a) Reactions
of Oxabenzonorbornadiene with Palladium(0). If the π-allyl
pathway were occurring a Pd(0) species would initiate the
catalytic cycle. In reactions with Pd(CH3CN)2Cl2 it is conceiv-
able that displacement of a Pd-Cl with dialkylzinc could form
a Pd(0) species via an in situ reductive elimination. However,
when the reaction was attempted using Me2Zn and catalytic Pd2-
(dba)3/dppf less than 25% conversion was observed after 24 h,
whereas with Pd(dppf)2Cl2 the reaction is complete in less than
12 h. Also, when oxabenzonorbornadiene was reacted with
sodium dimethyl malonate (1.5 equiv) and either Pd2(dba)3/dppf
or Pd(dppf)Cl2/Et2Zn (5 mol % to reduce Pd(II) to Pd(0)) no
addition of the dimethyl malonate was detected.
Scheme 5
(b) Reactions with Unsymmetrical Substrates. Additional
evidence disfavoring an ionization mechanism is seen in the
reaction of unsymmetrical substrate, 1, substituted at the
bridgehead (Scheme 3). If a Lewis acid promoted reaction was
taking place ionization at the tertiary center would be expected
to predominate (pathway A); however, products exclusively
from pathway B were observed.
Scheme 6
These results are better explained if a carbopalladation
mechanism is occurring. The more sterically encumbered
palladium would migrate to the less hindered carbon, while the
smaller alkyl group would be transferred to the carbon next to
the substituted bridgehead.
2. Evidence for a Carbopalladation Pathway. (a) Isolation
and Reaction of Carbopalladated Products. Important evi-
dence supporting a carbopalladation mechanism arose while
studying the addition of Me2Zn to the [3.2.1] oxabicycles where
we found that addition of Zn(OTf)2 improved the yield of the
reaction. When the reaction of 2 was performed at ambient
temperature instead of refluxing dichloroethane, the major
product, 4a, was one arising from a palladium-catalyzed carbo-
zincation of the alkene followed by protonation (Scheme 4).
was trapped with iodine yielding the iodide 9 along with 7 and
8. Ring opening by lithiation of 9 with t-BuLi gave a mixture
of 7 and in similar enantiomeric excess to the direct conversion
of 6 to 7 strongly supporting an enantioselective carbopalladation
step.
To confirm that a carbozincated intermediate, 5, was indeed
being formed, trapping experiments were carried out. When the
reaction was quenched with dueterium oxide instead of water
4c was isolated with incorporation of dueterium on the carbon
After carbopalladation it is conceivable that the organopal-
ladium species directly undergoes a â-oxygen elimination.6
Alternatively a transmetalation to zinc followed by ring opening
could be occurring. To study this question substrate 2 was
reacted with Me2Zn, Pd(dppf)Cl2, and Zn(OTf)2 at room
temperature to give what we assume is the carbozincated product
in analogy to the results presented in Scheme 4 (via a
carbopalladation and transmetalation). Subsequent heating to
reflux for 20 h (Scheme 6) gave only 8% of the ring-opened
product, 3, whereas the reaction carried out entirely at reflux
gave 40% of the ring-opened product after 20 h. This suggests
that the ring opening via a palladium â-oxygen elimination is
faster than ring opening from the zinc species. However, with
certain substrates and conditions the transmetalation to zinc is
competing with the ring opening.
1
adjacent to the methyl group. H NMR showed less than 10%
of 4a present, which indicates that the organometallic intermedi-
ate is the major product prior to quenching.
When the reaction was quenched with iodine, 4b was isolated
in 62% yield. An X-ray structure was obtained which confirmed
the cis relationship between the methyl, iodide, and bridging
oxygen. The cis iodide rules out the possibility of forming
these products on workup by proto- or iodoetherification of
3, and provides additional support for the formation of inter-
mediate 5.
Evidence that a carbopalladation was part of the catalytic
cycle and not just a side reaction was obtained when similar
results were observed using the chiral catalyst Pd(t-Bu-DIPOF)-
Cl2 with substrate 6 (Scheme 5). The carbometalated product
(b) Synthesis and Reaction of Palladium Alkyl Species.
To better understand the nature of the palladium intermediate

6836 J. Am. Chem. Soc., Vol. 123, No. 28, 2001
Lautens et al.
Scheme 7
Scheme 9
Scheme 8
Scheme 10
tetraarylborate, 17 (NaB(ArF)4), a reagent Brookhart has shown
forms a cationic palladium species in situ with concomitant
formation of NaCl.9 We found the resulting cationic palladium
species underwent very rapid reaction with 14 to give complete
conversion in less than an hour (Scheme 10). This suggests that
the dialkylzinc or zinc triflate may form at least a small amount
of a highly reactive cationic palladium species.
undergoing the carbopalladation we undertook the synthesis
of palladium alkyl species 11-13 using the methods described
by Vrieze and co-workers and van Koten and co-workers
(Scheme 7).8
Treatment of 14 with a stoichiometric amount of 12 or 13
failed to induce any reaction; however, when Zn(OTf)2 was
added, rapid conversion to the ring-opened product was observed
for both the methylchloropalladium species, 12, and the dim-
ethylpalladium species, 13 (Scheme 8).
When Et2Zn was added to the reaction of 13 and 14 a 5:1
ratio of ethyl addition to methyl addition was observed. This is
best explained if the rate of addition of the Et-Pd complex is
much faster than that of the Me-Pd complex. This proposal is
consistent with competition experiments in which we have
observed that ethyl addition is >100 times faster than methyl
addition. It is evident that diethylzinc both activates 13 leading
to methyl addition and then participates in a rapid catalytic
process.
Palladium BINAP complex 11 reacted analogously to 12 and
13, and the enantiomeric excesses of the methyl and ethyl
addition products were similar to that observed in the catalytic
reaction (Scheme 9).
These experiments provide strong evidence that [RPdL2]
species are present in the catalytic cycle and undergo an
enantioselective carbopalladation. It appears that both the
alkylchloropalladium and dialkylpalladium species can react but
both require the zinc as an activating agent. We speculated that
perhaps the zinc species was acting as a Lewis acid to remove
an alkyl or chloro group from the palladium creating a cationic
intermediate that could then bind the substrate alkene. To test
this hypothesis we performed a reaction with 13 and sodium
The proposal of a cationic palladium is in accordance with
trends observed while studying additive effects on the yield of
reaction with the [3.2.1] oxabicycles. During these reactions a
Pd(0) black precipitate was frequently observed. The addition
of Lewis bases such as excess phosphine, tetrabutylammonium
chloride, or even THF to stabilize the palladium prevented
formation of the Pd(0) precipitate but also inhibited the reaction.
The addition of the Lewis acid, Zn(OTf)2, which is more Lewis
acidic than a dialkylzinc and is more apt to form the reactive
cationic catalyst, resulted in higher yields and reduced amounts
of precipitate.
1
(c) H NMR Studies on the Effect of Zn(OTf)2 on Pd-
(dppf)MeCl. To better understand the effect of the Lewis acidic
1
species on the catalyst, H NMR studies were undertaken and
the results shown in Figure 1. Pd(dppf)MeCl, 13, was used for
the studies and trace A shows the spectrum of 13 in CDCl3.
There is a signal at 0.79 ppm from the methyl group and four
cyclopentadienyl ring signals, two from the ring trans to the
methyl and two from the ring trans to the chloride. Addition of
Zn(OTf)2 to 13 has an influence on the cyclopentadienyl
resonances where a coalescence occurs but little change on the
methyl signal (trace B). The result is similar to the spectrum
obtained from addition of AgOTf to 13 (trace C) with small
differences in the chemical shifts observed. Trace D shows the
affect of the addition of NaB(ArF)4 to 13. Here again a
coalescence in the cyclopentadienyl signals is observed, however
in a slightly different pattern. The coalescence observed seems
to suggest that in each of these cases there is partial or complete
removal of the chloride to give a cationic palladium species.
The differences in the spectra could be attributed to differences
in the nature of the counterion, with the one obtained from
Zn(OTf)2 being more similar to that from AgOTf compared to
NaB(ArF)4.
(8) (a) Rulke, R.; Ernsting, J.; Spek, A.; Elsevier, C.; van Leeuwen, P.;
Vrieze, K. Inorg. Chem. 1993, 32, 5769. (b) de Graaf, W.; Boersma, J.;
Smeets, W. J. J.; Spek, A. L.; van Koten, G. Organometallics 1989, 8,
2907.
(9) Brookhart, M.; Wagner, M. J. Am. Chem. Soc. 1994, 116, 3641.

Pd-Catalyzed Ring Opening of Oxabicyclic Alkenes with Dialkylzinc
Scheme 12
J. Am. Chem. Soc., Vol. 123, No. 28, 2001 6837
Figure 1. 1H NMR of Pd(dppf)MeCl (A), Pd(dppf)MeCl + Zn(OTf)2
(B), Pd(dppf)MeCl + AgOTf (C), and Pd(dppf)MeCl + NaB(ArF)4
(D).
Scheme 13
Scheme 11
(d) Attempts To Observe a Carbopalladated Intermediate.
In hopes of observing a cationic palladium intermediate with
substrate 2, which did not undergo ring opening at room
temperature, reaction with 13 and Zn(OTf)2 or NaB(ArF)4 was
attempted (Scheme 11). To our surprise the ring-opened product,
which normally could only be obtained at elevated temperature,
was isolated. Thus, without the dialkylzinc species present for
transmetalation, the carbopalladated intermediate undergoes
rapid â-oxygen elimination to the ring-opened product.
3. Investigations into the Role of the Substrate Olefin. To
investigate the importance of the substrate olefin on the reaction,
different substrates were submitted to the reaction conditions.
Without substrate present, addition of Et2Zn to any of the
palladium species in dichloromethane causes rapid formation
of a black Pd(0) precipitate. The addition of Me2Zn to Pd(dppf)-
Cl2 in the absence of substrate forms Pd(dppf)Me2 which is
stable at room temperature but upon heating also forms a black
precipitate. Reaction with styrene gave no reaction, and
substrates 19 and 20 seemed to react via Pd(0) mechanisms
(Scheme 12).
species, i, via transmetalation with the dialkylzinc which then
binds the alkene substrate with loss of X-, assisted by a Lewis
acidic zinc agent. An enantioselective carbopalladation takes
place to give intermediate iii, and â-oxygen elimination can then
give the ring-opened product, iv, or transmetalation to zinc may
occur which competes with ring opening and gives v. The zinc
intermediate could potentially ring open, however it appears to
do so at a slower rate.
In conclusion, these studies have helped provide insight into
the mechanism of this new enantioselective reaction, and have
been useful in the extension of this methodology. They also
provide a better understanding of the interactions between alkyl
palladium species and zinc. Reactions involving catalysis by
cationic palladium are becoming more prevalent in the literature.
Our studies show the important qualities of the dialkylzinc which
functions both in the transmetalation to palladium and as a Lewis
acid in forming the reactive cationic palladium species.
Experimental Section
In contrast, reaction of 21 provided 1,2-dimethylnaphthalene
which presumably proceeds via carbopalladation followed by
â-carbon elimination. This result is similar to the ring-opening
of cyclopropanes and cyclobutanes bearing a palladium at the
â position.10 Ring opening of these bicyclic substrates releases
significant ring strain providing a driving force for the reaction.
General. All 1H and 13C NMR spectra were recorded using a Varian
XL 400 spectrometer. IR spectra were obtained using a Nicolet DX
FT-IR spectrometer. High-resolution mass spectra were obtained on a
VG 70-250S (double focusing) mass spectrometer at 70 eV.
Reactions of Oxabenzonorbornadiene (14) with Palladium(0).
A flame-dried round-bottom flask was charged, under argon, with
Pd2dba3 (6.3 mg, 0.007 mmol), dppf (7.6 mg, 0.14 mmol), and
oxabenzonorbornadiene 14 (40 mg, 0.28 mmol). Distilled dichlo-
romethane (5 mL) was added and stirred 1 h at room temperature. To
this solution was added dimethylzinc (2.0 M in toluene, 0.21 mL, 0.42
mmol) which was stirred at room temperature for 24 h. The reaction
was quenched with a few drops of water and the solution was stirred
for 20 min, filtered through Celite, and evaporated to give the crude
product. The crude product was purified by flash chromatography on
silica gel (15% ethyl acetate in hexanes) to give (1R,2R)-2-methyl-
Conclusion
The mechanism we propose from the above evidence is
summarized in Scheme 13. Pd(dppf)Cl2 forms a palladium alkyl
(10) (a) Larock, R.; Varaprath, S. J. Org. Chem. 1984, 49, 3432. (b)
Donaldson, W.; Brodt, C. J. Organomet. Chem. 1987, 330, C33. (c) Brase,
S.; De Meijere, A. Angew. Chem., Int. Ed. Engl. 1995, 34, 2545. (d) Lautens,
M.; Meyer, C.; Lorenz, A. J. Am. Chem. Soc. 1996, 118, 10676.

6838 J. Am. Chem. Soc., Vol. 123, No. 28, 2001
Lautens et al.
1,2-dihydronaphth-1-ol (15, 11 mg, 25% yield) as a white solid. Rf )
0.33 on silica gel (hexanes:ethyl acetate 7:3); mp 64-65 °C; IR (CHCl3)
3456 (s), 3030 (m), 2966 (s), 1588 (w), 1485 (m), 1454 (m), 1364
(m), 1280 (m), 831 (w), 788 (s), 760 (s), 679 (w) cm-1; 1H NMR (400
MHz, CDCl3) δ 7.36 (1H, d, J ) 7.2 Hz), 7.30-7.22 (2H, m), 7.12
(1H, d, J ) 7.2 Hz), 6.52 (1H, dd, J ) 9.5, 2.3 Hz), 5.80 (1H, dd, J )
9.5, 2.5 Hz), 4.57 (1H, dd, J ) 7.8, 4.8 Hz), 2.67-2.63 (1H, m), 1.57
(1H, d, J ) 7.8 Hz), 1.25 (3H, d, J ) 7.4 Hz); 13C NMR (100 MHz,
CDCl3) δ 136.7, 132.6, 132.4, 128.6, 127.7, 127.4, 126.7, 126.6, 71.8,
35.4, 14.2; HRMS calcd for C11H12O (M)+ 160.0888, found 160.0888.
Attempts to add dimethyl malonate were carried out as above except
with the addition of a solution of sodium dimethyl malonate (formed
by addition of NaH (8 mg, 0.33 mmol) to dimethyl malonate (38 µL,
0.33 mmol) in THF at room temperature) to the substrate and catalyst
in THF at room temperature.
Si (M)+ 326.263281, found 326.264109; M(m/z) 309, 283, 265, 135,
123, 103, 95, 75, 61.
When the reaction was carried out at room temperature product 4a
was obtained in 88% yield as a white solid. Rf ) 0.5 on silica gel
(hexanes:ethyl ether 8:2); IR (CHCl3) 2951, 1458, 1057, 1015, 909,
1
881, 733 cm-1; H NMR (400 MHz, CDCl3) δ 4.02 (t, J ) 4.0 Hz,
1H), 3.98 (dd, J ) 7.6, 3.2 Hz, 1H), 3.47 (d, J ) 2.8 Hz, 1H), 2.70-
2.61 (m, 1H), 2.47 (dd, J ) 11.7, 8.6 Hz, 1H), 2.00-1.91 (m, 1H),
1.16 (m, 1H), 1.13(br s, 21H), 1.00 (d, J ) 7.2 Hz, 3H), 0.96 (d, J )
7.2 Hz, 3H), 0.91 (d, J ) 7.6 Hz, 3H); 13C NMR (100 MHz, CDCl3)
δ 86.2, 80.0, 74.4, 40.4, 40.3, 34.8, 32.1, 23.6, 18.9, 14.1, 14.0, 13.9;
HRMS calcd for C16H31O2Si (M - C3H7)+ 283.2093, found 283.2098.
When the reaction was carried out at room temperature for 20 h
followed by addition of iodine (2 equiv) and stirring an additional 2 h
product 4b was obtained in 62% yield as a white solid. Rf ) 0.6 on
silica gel (hexanes:ethyl ether 8:2); 1H NMR (400 MHz, CDCl3) δ 5.20
(d, J ) 8.0 Hz, 1H), 4.43 (d, J ) 3.2 Hz, 1H), 4.04 (t, J ) 3.8 Hz,
1H), 3.57 (d, J ) 3.2 Hz, 1H), 2.64 (quint., J ) 7.4 Hz, 1H), 1.91 (m,
2H), 1.22 (d, J ) 7.6 Hz, 3H), 1.13(br s, 21H), 1.08 (d, J ) 7.2 Hz,
3H), 0.97 (d, J ) 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ 92.1,
85.6, 73.8, 41.6, 39.5, 35.8, 35.1, 29.4, 18.6, 13.7, 13.6, 13.5; HRMS
calcd for C16H30O2SiI (M - C3H7)+ 409.1060, found 409.1069.
Ring Opening and Carbometalation of endo-3-(tert-Butyldiphe-
nylsilyl)oxy-8-oxabicyclo[3.2.1]oct-6-ene (6). A flame-dried, round-
bottom flask was charged, under argon, with Pd(CH3CN)2Cl2 (5 mol
%) and (S)-t-Bu-(S)-DIPOF (5 mol %). Distilled dichloroethane (1 mL)
was added, the solution was stirred at room temperature for 1 h, and
then substrate 6 (50 mg, 0.137 mmol) in dichloroethane (6 mL) was
added via cannula and heated to reflux. A solution of dimethylzinc
(2.0 M in toluene, 0.14 mL, 0.28 mmol) was then added and stirred at
reflux for 20 h. The reaction was cooled to room temperature and a
solution of iodine (2 equiv) in dichloroethane was added and stirred
60 min. The reaction was quenched with a few drops of water and the
solution was stirred for 15 min, filtered through a short plug of Celite,
and concentrated. The crude mixture was purified by flash chroma-
tography on silica gel to afford products 7, 8,11 and 9.
(1R,2R)-1,2-Dimethyl-1,2-dihydronaphth-1-ol. A flame-dried, round-
bottom flask was charged, under argon, with Pd(dppf)Cl2 (3 mg, 0.004
mmol) and 1 (73 mg, 0.46 mmol). Distilled dichloromethane (13 mL)
was added prior to addition of dimethylzinc (2.0 M in toluene, 0.28
mL, 0.56 mmol) and the mixture was stirred at room temperature for
16 h. The reaction was quenched with a few drops of water and the
solution was stirred 20 min, filtered through Celite, and evaporated to
give the crude product, which was purified by flash chromatography
on silica gel (10% ether in hexanes) to give the product (50 mg, 62%
yield) as an oil. Rf ) 0.36 on silica gel (hexanes:ether 7:3); IR (neat)
1
3450, 3031, 2966, 1484, 1452, 1364, 788, 760 cm-1; H NMR (400
MHz, CDCl3) δ 7.57 (d, J ) 7.1 Hz, 1H), 7.23 (m, 2H), 7.03 (d, J )
6.8 Hz, 1H), 6.36 (d, J ) 9.6 Hz, 1H), 5.97 (dd, J ) 9.6, 5.2 Hz, 1H),
2.42 (quint., J ) 5.2 Hz, 1H), 1.82 (1H, OH), 1.51 (s, 3H), 1.01 (d, J
) 5.2 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ 141.0, 133.8, 132.1,
127.9, 127.4, 126.3, 125.6, 123.7, 73.9, 41.0, 28.1, 13.2; HRMS calcd
for C12H14O (M)+ 174.1045 found 174.1043.
(1R,2R)-1-Methyl-2-ethyl-1,2-dihydronaphth-1-ol. A flame-dried,
round-bottom flask was charged, under argon, with Pd(dppf)Cl2 (15
mg, 0.02 mmol) and 1 (73 mg, 0.46 mmol). Distilled dichloromethane
(8 mL) was added prior to addition of diethylzinc (1.0 M in hexanes,
0. 64 mL, 0.64 mmol) and the mixture was stirred at room temperature
for 4 h. The reaction was quenched with a few drops of water and the
solution was stirred 20 min, filtered through Celite, and evaporated to
give the crude product, which was purified by flash chromatography
on silica gel (10% ether in hexanes) to give the product (68 mg, 78%
yield) as an oil. Rf ) 0.4 on silica gel (hexanes:ether 7:3); IR (neat)
Product 7 was obtained in 19% yield as an oil, and the ee was
determined to be 81% using HPLC analysis on a CHIRACEL OD
column, λ ) 254 nm. Retention times in 0.5% iPrOH in hexanes were
9.4 min (major) and 15.6 min (minor). Rf ) 0.27 on silica gel (hexanes:
ethyl ether 8:2); IR (neat) cm-1 3443, 2954, 2923, 2867, 1649, 1451,
1109, 1067; 1H NMR (400 MHz, CDCl3) δ 7.68-7.65 (m, 4H), 7.44-
7.34 (m, 6H), 5.62-5.56 (m, 1H), 5.30-5.25 (m, 1H), 3.86-3.81 (m,
1H), 3.73 (br s, 1H), 2.58-2.48 (m, 1H), 2.40-2.22 (m, 3H), 1.86 (td,
J ) 10.4, 2.8 Hz, 1H); 13C NMR(100 MHz, CDCl3) δ 135.8, 135.1,
134.4, 134.2, 129.6, 129.5, 127.9, 127.5, 72.4, 65.8, 47.6, 38.2, 38.0,
26.9, 19.1, 18.9; HRMS calcd for C20H23O2Si (M - C4H9)+ 323.1467,
found 323.1482.
1
3447, 3033, 2963, 1588, 1461, 1384, 790, 758 cm-1; H NMR (400
MHz, CDCl3) δ 7.55 (d, J ) 8.0 Hz, 1H), 7.23 (m, 2H), 7.03 (d, J )
8.0 Hz, 1H), 6.41 (d, J ) 10.0 Hz, 1H), 6.11 (dd, J ) 9.8, 5.8 Hz,
1H), 2.22 (m, 1H), 1.77 (1H, OH), 1.75 (m, 1H), 1.50 (s, 3H), 1.26
(m, 1H), 0.91 (t, J ) 7.4 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ
141.9, 132.4, 132.1, 127.9, 127.2, 126.2, 126.1, 123.3, 74.4, 47.6, 29.3,
21.5, 11.7; HRMS calcd for C13H16O (M)+ 188.1201, found 188.1203.
Ring Opening and Carbometalation of endo-2,4-Dimethyl-3-
(triisopropylsilyl)oxy-8-oxabicyclo[3.2.1]oct-6-ene (2). A flame-dried,
round-bottom flask was charged, under argon, with Pd(dppf)Cl2 (6.5
mg, 0.008 mmol), Zn(OTf)2 (58 mg, 0.16 mmol), and substrate 2 (50
mg, 0.161 mmol). Distilled dichloroethane (6 mL) was added and the
reaction solution brought to the appropriate temperature. A solution of
dimethylzinc (2.0 M in toluene, 0.12 mL, 0.24 mmol) was then added
dropwise. After the reaction was complete it was cooled to room
temperature and quenched with a few drops of water and the solution
was stirred 15 min, filtered through a short plug of Celite, and
concentrated. The crude mixture was purified by flash chromatography
on silica gel (10% Et2O in hexanes) to give the product.
When the reaction was carried out at reflux product 3 was obtained
in 62% yield as an oil. Rf ) 0.33 on silica gel (hexanes:ethyl ether
8:2); IR (neat) 3403, 2945, 2865, 1653, 1461, 1382, 1212, 1089, 887,
854 cm-1; 1H NMR (400 MHz, CDCl3) δ 5.65 (ddd, J ) 11.5, 7.5, 1.6
Hz, 1H), 5.34 (dd, J ) 11.5, 5.3 Hz, 1H), 4.19 (dd, J ) 5.0, 3.6 Hz,
1H), 3.65 (td, J ) 7.0, 2.0 Hz, 1H), 2.73 (m, 1H), 2.51 (quint. d, J )
7.3, 3.5 Hz, 1H), 2.26 (quint., J ) 7.0 Hz, 1H), 1.60 (dd, J ) 7.0, 6.8
Hz, 1H), 1.17 (d, J ) 7.5 Hz, 3H), 1.14 (d, J ) 7.3 Hz, 6H), 1.07 (br
s, 18 H); 13C NMR (100 MHz, CDCl3) δ 133.1, 132.0, 77.2, 71.2,
45.6, 42.6, 36.6, 18.4, 17.6, 16.1, 14.4, 12.8; HRMS calcd for C26H36O2-
Product 8 was obtained in 38% yield as an oil. Rf ) 0.51 on silica
gel (hexanes:ethyl ether 8:2); IR (neat) 3155, 2972, 1486, 1381, 1103,
1
913, 737 cm-1; H NMR (400 MHz, CDCl3) δ 7.64-7.62 (m, 4H),
7.46-7.40 (m,2H), 7.39-7.35 (m, 4H), 4.34 (dd, J ) 7.6, 3.5 Hz, 1H),
4.08 (t, J ) 4.6 Hz, 1H), 3.80 (d, J ) 3.6 Hz, 1H), 2.95-2.85 (m,
1H), 2.72 (dd, J ) 11.6, 8.8 Hz, 1H), 1.89-1.81 (m, 2H), 1.65 (d, J )
14.4 Hz, 1H), 1.46 (ddd, J ) 11.4, 8.0, 3.8 Hz, 1H), 1.56 (dd, J )
14.4, 1.0 Hz, 1H), 1.09 (s, 9H), 1.07(d, J ) 6.8 Hz, 3H); 13C NMR
(100 MHz, CDCl3) δ 136.1 (4C), 134.3 (2C), 129.9 (2C), 127.8 (4C),
81.3, 75.1, 66.0, 38.8, 38.7, 38.6, 36.3, 27.2 (3C), 23.4, 19.3; HRMS
calcd for C20H23O2Si (M - C4H9)+ 323.1467, found 323.1466.
Product 9 was obtained in 29% yield as an oil. Rf ) 0.62 on silica
gel (hexanes:ethyl ether 8:2); IR (neat) 3148, 2930, 1817, 1795, 1638,
1
1483, 1468, 1382, 1102, 990, 941, 783 cm-1; H NMR (400 MHz,
CDCl3) δ 7.72-7.67 (m, 4H), 7.45-7.35 (m, 6H), 4.21 (t, J ) 5.9 Hz,
1H), 4.08 (t, J ) 5.7 Hz, 1H), 3.77 (br s, 1H), 3.70 (t, J ) 6.6 Hz,
1H), 3.24 (quint., J ) 6.6 Hz, 1H), 2.46 (AB, JAB ) 15.4 Hz, 1H),
2.11 (td, J ) 14.8, 6.0 Hz, 1H), 1.81 (td, J ) 14.8, 4.8 Hz, 1H), 1.49
(d, JAB ) 15.4 Hz, 1H), 1.15 (d, J ) 7.1 Hz, 3H), 1.12 (s, 9H); 13C
(11) This product was obtained by protonation of the carbometalated
product by water, presumably because the iodination was slow and
incomplete. Quenching the reaction with D2O instead of iodine gave only
7 with a deuterium next to the methyl along with ring-opened product.

Pd-Catalyzed Ring Opening of Oxabicyclic Alkenes with Dialkylzinc
J. Am. Chem. Soc., Vol. 123, No. 28, 2001 6839
NMR (100 MHz, CDCl3) δ 136.3 (2C), 136.2 (2C), 134.3, 134.0, 129.9,
129.8, 127.8 (2C), 127.7 (2C), 80.2, 78.6, 64.4, 44.3, 39.0, 34.9, 29.9,
27.5 (3C), 19.3, 18.9; HRMS calcd for C20H22O2SiI (M - C4H9)+
449.0434, found 449.0427.
NMR (400 MHz, CDCl3) δ 7.31 (1H, d, J ) 6.9 Hz), 7.28-7.19 (2H,
m), 7.10 (1H, d, J ) 7.1 Hz), 6.52 (1H, dd, J ) 9.7, 2.5 Hz), 5.81 (1H,
dd, J ) 9.6, 1.8 Hz), 4.62 (1H, dd, J ) 7.3, 4.5 Hz), 2.39-2.32 (1H,
m), 1.84 (1H, dquin, J ) 13.6, 7.5 Hz), 1.63 (1H, dquin, J ) 13.6, 7.5
Hz), 1.52 (1H, d, J ) 7.3 Hz), 1.11 (3H, t, J ) 7.4 Hz). The
enantiomeric excess of products 15 and 16 was determined by HPLC
analysis on a CHIRACEL OD column, λ ) 254 nm. Retention times
for 15 in 1.25% iPrOH in hexanes were 33.6 and 35.3 min (major) and
Ring opening of 9 was achieved by addition of t-BuLi (1.7M in
pentane, 46 µL, 0.08 mmol) to substrate 9 (20 mg, 0.04 mmol) in
distilled THF at -78 degrees. This solution was allowed to slowly
warm to room-temperature overnight. Work up was done with water
and ether, and the ether layer was separated, dried (Na2SO4), and
evaporated to give crude 7 and 8 in a 1:1.5 ratio. The ee of 7 was
determined to be 86% by HPLC analysis as described above.
[(S)-BINAP]Pd(CH3)Cl (11).12 To a flame-dried, round-bottom
flask, under argon, was added Pd(CH3CN)2Cl2 (33 mg, 0.13 mmol)
and (S)-BINAP (80 mg, 0.13 mmol) and 4 mL of distilled dichlo-
romethane. The mixture was stirred 1 h at room temperature and then
SnMe4 (36 µL, 0.26 mmol) was added and stirring was continued for
65 h at room temperature. The reaction mixture was filtered through
Celite, evaporated, washed with ether (3 × 5 mL), and dried in vacuo
to give product 11 (71 mg, 70% yield) as a pale yellow solid. 1H NMR
(400 MHz, CDCl3) δ 7.88 (m, 2H), 7.75 (m, 2H), 7.54-7.21 (m, 18H),
i
for 16 in 1.25% PrOH in hexanes were 14.5 and 16.3 min (major).
1H NMR Studies on Pd(dppf)MeCl. A flame-dried, round-bottom
flask, under argon, was charged with Pd(dppf)MeCl (7.1 mg, 0.01
mmol) and Zn(OTf)2 (3.6 mg, 0.01 mmol), AgOTf (2.5 mg, 0.01 mmol),
or NaB(ArF)4 (8.8 mg, 0.01 mmol). CDCl3 (2 mL, stored over molecular
sieves) was added with stirring at room temperature. After 10 to 30
1
min the salts were allowed to settle and an aliquot was taken for H
1
NMR analysis. The H NMR of Pd(dppf)MeCl is given above. For
Pd(dppf)MeCl + Zn(OTf)2 signals are seen at δ 4.39 (4H), 4.24 (2H),
3.81 (2H), 0.79 (3H). For Pd(dppf)MeCl + AgOTf signals are seen at
δ 4.38 (4H), 4.26 (2H), 3.82 (2H), 0.82 (3H). For Pd(dppf)MeCl +
NaB(ArF)4 signals are seen at δ 4.35 (2H), 4.25 (4H), 3.81 (2H), 0.79
(3H).
7.03 (m, 2H), 6.79-6.53 (m, 8H), 0.77 (dd, J ) 7.8, 3.4 Hz, 3H); 13
C
NMR (100 MHz, CDCl3) δ 135.6-125.1 (aromatic C), 16.9 (d, J )
97.3 Hz); ESMS calcd for C45H35P2Pd (M - Cl)+ 743, found 743.13
[1,1′-Bis(diphenylphosphino)ferrocene]Pd(CH3)2 (12). To a flame-
dried, round-bottom flask, under argon, was added (TMEDA)Pd(CH3)214
(200 mg, 0.8 mmol) and dppf (444 mg, 0.8 mmol) and 7 mL of distilled
benzene. This was stirred at room temperature for 30 min and then the
benzene was removed in vacuo. The crude was washed with pentane
and dried in vacuo to give product 12 (550 mg, 99% yield) as an orange
General Procedure for Reactions with Olefins 18-21. A flame-
dried, round-bottom flask, under argon, was charged with Pd(dppf)Cl2
and Zn(OTf)2 (105 mg, 0.29 mmol). Distilled dichloromethane or
dichloroethane (2 mL) was added and after 10 min at room temperature
the substrate (0.29 mmol) in 4 mL of the reaction solvent was added
via cannula. The reaction was heated to the appropriate temperature
and dimethylzinc (2.0 M in toluene, 0.22 mL, 0.44 mmol) was added.
After 20 h the reaction was cooled to room temperature and quenched
by addition of a few drops of water. This was stirred 15 min and then
filtered through Celite and evaporated to give crude product.
1-Phenylbutene. The reaction was carried out as above with
substrate 19 in refluxing dichloroethane with 25 mol % Pd(dppf)Cl2
and no Zn(OTf)2. The crude product was purified by flash chroma-
tography on silica gel (hexanes) to give 18% yield of product. Rf )
1
solid. H NMR (400 MHz, CDCl3) δ 7.68 (m, 8H), 7.35 (m, 12H),
4.22 (t, J ) 1.6 Hz, 4H), 4.09 (t, J ) 1.6 Hz, 4H), 0.13 (d, J ) 2.8 Hz,
6H); 13C NMR (100 MHz, CDCl3) δ 135.0 (d, J ) 28.2 Hz), 134.6 (t,
J ) 6.8 Hz), 129.6, 127.9 (t, J ) 4.5 Hz), 79.5 (d, J ) 38.8 Hz), 74.7
(t, J ) 5.3 Hz), 71.6, 8.5 (dd, J ) 14.4, 107.9 Hz); ESMS calcd for
C36H34P2FePd (M)+ 690, found 690.13
1
[1,1′-Bis(diphenylphosphino)ferrocene]Pd(CH3)Cl (13). To a flame-
dried, round-bottom flask, under argon, was added 12 (250 mg, 0.36
mmol) and 5 mL of distilled benzene. Acetyl chloride (0.1 mL, 1.4
mmol) was added at room temperature and the reaction mixture was
stirred for 1 h. The reaction mixture was then evaporated in vacuo to
0.65 (5% ethyl acetate in hexanes); H NMR (400 MHz, CDCl3) δ
7.35-7.16 (m, 5H), 6.38 (d, J ) 16.0 Hz, 1H), 6.27 (d, J ) 16.0, 6.4
Hz, 1H), 2.23 (quint., J ) 7.2 Hz, 2H), 1.09 (t, J ) 7.2 Hz, 3H).
2-Methyl-4-phenylbut-3-enol and 4-Methyl-4-phenylbut-2-enol.
The reaction was carried out as above with substrate 20 at room
temperature in dichloromethane with 5 mol % Pd(dppf)Cl2 and no Zn-
(OTf)2. The crude product was purified by flash chromatography on
silica gel (40% ethyl acetate in hexanes) to give a mixture of isomers
in 79% yield. Rf ) 0.3 (40% ethyl acetate in hexanes)
1
give product 13 (250 mg, 97%). H NMR (400 MHz, CDCl3) δ 7.89
(m, 4H), 7.65 (m, 4H), 7.43-7.29 (m, 12H), 4.48 (s, 2H), 4.38 (s,
2H), 4.20 (s, 2H), 3.74 (s, 2H), 0.79 (dd, J ) 7.8, 4.6 Hz); 13C NMR
(100 MHz, CDCl3) δ 134.9 (d, J ) 12.2 Hz), 134.3 (d, J ) 12.2 Hz),
133.6 (d, J ) 31.9 Hz), 132.8 (d, J ) 53.2 Hz), 130.7 (d, J ) 3.0 Hz),
129.9 (d, J ) 2.3 Hz), 128.0 (d, J ) 10.7 Hz), 127.9 (d, J ) 9.1 Hz),
77.6 (d, J ) 7.1 Hz), 75.7 (d, J ) 11.4 Hz), 74.7 (d, J ) 9.1 Hz), 72.9
(d, J ) 6.9 Hz), 72.3 (d, J ) 4.5 Hz), 16.8 (d, J ) 98.8 Hz); ESMS
calcd for C35H31P2FePd (M - Cl)+ 675, found 675.13
1,2-Dimethylnaphthalene. The reaction was carried out as above
with substrate 21 in refluxing dichloroethane with 25 mol % Pd(dppf)-
Cl2. The crude product was purified by flash chromatography on silica
1
gel (hexanes) to give product in 30% yield. Rf ) 0.4 (hexanes); H
NMR (400 MHz, CDCl3) δ 8.03 (d, J ) 8.4 Hz, 1H), 7.79 (d, J ) 8.0
Hz, 1H), 7.62 (d, J ) 8.4 Hz, 1H), 7.49 (t, J ) 7.6 Hz, 1H), 7.41 (d,
J ) 7.5 Hz, 1H), 7.29 (d, J ) 8.0 Hz, 1H), 2.61 (s, 3H), 2.49 (s, 3H).
General Procedure for Reactions of Palladium Alkyl Species.
The reactions in Schemes 8-11 were carried out as follows: A flame-
dried, round-bottom flask, under argon, was charged with substrate 14
or 2 (1 equiv), palladium species 11, 12, or 13 (1 equiv), and CDCl3
(stored over molecular sieves). A small aliquot was taken after 1 h
Acknowledgment. We thank the AstraZeneca Research
Centre in Montreal, the Ontario Research and Development
Challenge Fund (ORDCF), NSERC (Canada), and the Univer-
sity of Toronto for financial support of this work. S.H. thanks
NSERC (Canada) for financial support in the form of a
postgraduate fellowship. J.L.R. thanks the Ministe`re Francais
des Affaires Etrange`res for a Bourse LaVoisier.
1
and diluted in CDCl3 for H NMR analysis. Then Zn(OTf)2 or Et2Zn
or NaB(ArF)4 (1 equiv) was added and the mixture was stirred at room
temperature. Aliquots were taken after 1-4 h for 1H NMR analysis to
determine percent conversion of substrate to product.
The 1H NMR of products 15 and 3 is given above. Product 16: 1H
(12) Recrystallization from dichloromethane/hexanes gave crystals suit-
able for obtaining an X-ray structure.
Supporting Information Available: Crystallographic data
and ORTEP diagram for the X-ray structure of compound 4b
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
(13) Electrospray mass spectrometry was the only method mild enough
to give molecular ion signals without loss of ligand. The pattern and intensity
of the isotope peaks for M+ were calculated and matched the experimental
data giving good evidence for the molecular formula.
(14) (TMEDA)Pd(CH3)2 was prepared according to the procedure in ref
8b.
JA010498K