Titanocene-promoted intermolecular couplings of epoxides with nitriles. An easy access to β-hydroxyketones

Article abstract of DOI:10.1021/jo900479v

(Chemical Equation Presented) Radical couplings of epoxides and nitriles mediated by Cp₂TiCl provide a diastereoselective route to the synthesis of β-hydroxyketones. The conditions of this "aldol- like" reaction are mild enough to avoid the dehydration of the β-hydroxyketone. The scope of the coupling reaction with functionalized and tetrasubstituted epoxides has been studied. The radical character of the coupling reactions is demonstrated.

Full text of DOI:10.1021/jo900479v

Titanocene-Promoted Intermolecular Couplings of Epoxides with
Nitriles. An Easy Access to ꢀ-Hydroxyketones
A. Ferna´ndez-Mateos,* S. Encinas Madrazo, P. Herrero Teijo´n, and R. Rubio Gonza´lez
Departamento de Qu´ımica Orga´nica, Facultad de CC. Qu´ımicas, UniVersidad de Salamanca, Plaza de los
Ca´ıdos 1-5, 37008 Salamanca, Spain
afmateos@usal.es
ReceiVed March 3, 2009
Radical couplings of epoxides and nitriles mediated by Cp2TiCl provide a diastereoselective route to the
synthesis of ꢀ-hydroxyketones. The conditions of this “aldol-like” reaction are mild enough to avoid the
dehydration of the ꢀ-hydroxyketone. The scope of the coupling reaction with functionalized and
tetrasubstituted epoxides has been studied. The radical character of the coupling reactions is demonstrated.
Introduction
by titanocene chloride, onto acceptors with polar multiple bonds
such as aldehydes,6 ketones,6 esters,7 and nitriles.8 The products
from these cyclizations are diols, hydroxyacetals, or hydroxy-
ketones obtained in good yields from the corresponding epoxy
derivatives. Thus, titanocene chloride has been used to broaden
the range of useful radical traps; a recent mechanistic study9 of
the cyclization of epoxyderivatives revealed the triple role of
Ti(III): a radical initiator (homolytic cleavage of oxirane), a
Lewis acid accelerator (coordination to CN or CdO), and a
terminator (reduction of iminyl or alkoxyl radicals). In a related
cyclization of cyanoketones mediated by Cp2TiPh reported by
Itoh et al., the initial radical was generated from the carbonyl
group, and the acceptor was the nitrile group.10
The radical generation method based on titanocene-mediated
opening of epoxides through single electron transfer introduced
by Nugent and RajanBabu1 and the catalytic version developed
by Gansa¨uer and his group2 have been the object of many
interesting applications in synthesis.3 The main acceptors in all
of these radical reactions are alkenes.
Although radical additions to polar multiple bonds such as a
carbonyl group4 or a cyano group5 are in general unfavorable
processes due to reversibility of the former and the slowness of
the latter, our group has developed radical cyclizations, induced
(1) (a) Nugent, W. A.; RajanBabu, T. V. J. Am. Chem. Soc. 1988, 110, 8561.
(b) RajanBabu, T. V.; Nugent, W. A. J. Am. Chem. Soc. 1994, 116, 986.
(2) (a) Gansa¨uer, A.; Pierobon, M.; Bluhm, H. Angew. Chem., Int. Ed. 1998,
37, 101. (b) Gansa¨uer, A.; Bluhm, H.; Pierobon, M. J. Am. Chem. Soc. 1998,
120, 12849.
Recently, the group of Gansa¨uer has reported the catalytic
radical cyclization of some epoxy derivatives.11 They found that
epoxycarbonyl compounds react with standard titanocene chlo-
(3) (a) Gansa¨uer, A.; Bluhm, H. Chem. ReV. 2000, 100, 2771. (b) Gansa¨uer,
A.; Narayan, S. AdV. Synth. Catal. 2002, 344, 465. (c) Gansa¨uer, A.; Lauterbach,
T.; Narayan, S. Angew. Chem., Int. Ed. 2003, 42, 5556. (d) Cuerva, J. M.; Justicia,
J.; Oller-Lo´pez, J. L.; Oltra, J. E. Top. Curr. Chem. 2006, 264, 63. (e) Cuerva,
J. M.; Justicia, J.; Oller-Lo´pez, J. L.; Bazdi, B.; Oltra, J. E. Mini ReV. Org.
Chem. 2006, 3, 23. (f) Barrero, A. F.; Qu´ılez del Moral, J. F.; Sa´nchez, E. M.;
Arteaga, J. F. Eur. J. Org. Chem. 2006, 1627. (g) Gansa¨uer, A.; Justicia, J.;
Fan, C.-A.; Worgull, D.; Piestert, F. Top. Curr. Chem. 2007, 279, 25.
(4) (a) Kim, S. AdV. Synth. Catal. 2004, 346, 19–32, and references therein.
(b) Curran, D. P. In ComprehensiVe Organic Synthesis; Trost, B. M., Flemming,
I., Eds.; Pergamon Press: Oxford, U.K., 1991; Vol. IV, pp 815-817. (c) Giese,
B.; Kopping, B.; Go¨bel, T.; Dickhaut, J.; Thoma, G.; Kulicke, K. J.; Trach, F.
Org. React. 1996, 48, 311–312. (d) Beckwith, A. L. J.; Raner, K. D. J. Org.
Chem. 1992, 57, 4954–4962.
(5) (a) Fallis, A. G.; Brinza, I. M. Tetrahedron 1997, 53, 17543–17594, and
references therein. (b) Beckwith, A. L. J.; O’Shea, D. M.; Westwood, S. W.
J. Am. Chem. Soc. 1988, 110, 2565–2575. (c) Beckwith, A. L. J.; Hay, B. D.
J. Am. Chem. Soc. 1989, 111, 230. (d) Beckwith, A. L. J.; Hay, B. D. J. Am.
Chem. Soc. 1989, 111, 2674. (e) Curran, D. P.; Liu, W. Synlett 1999, 117–119.
(f) Bowman, W. R.; Bridge, C. F.; Brookes, P. Tetrahedron Lett. 2000, 41, 8989–
8994. (g) Rychnovsky, S. D.; Swenson, S. S. Tetrahedron 1997, 53, 16489–
16502. (h) Crich, D.; Bowers, A. A. J. Org. Chem. 2006, 71, 3452–3463.
(6) (a) Ferna´ndez-Mateos, A.; Martin de la Nava, E.; Pascual Coca, G.; Ramos
Silva, A.; Rubio Gonza´lez, R. Org. Lett. 1999, 1, 607. (b) Ferna´ndez-Mateos,
A.; Mateos Buro´n, L.; Rabanedo Clemente, R.; Mart´ın de la Nava, E. M.; Rubio
Gonza´lez, R.; Sanz Gonza´lez, F. Synlett 2004, 2553–2557. For applications of
this method to the synthesis of ꢀ-lactams, see: (c) Ruano, G.; Martian˜ez, J.;
Grande, M.; Anaya, J. J. Org. Chem. 2003, 68, 2024–2027. For applications of
this method to the synthesis of bioactive terpenoids see: (d) Bermejo, F. A.;
Ferna´ndez-Mateos, A.; Marcos-Escribano, A.; Mart´ın-Lago, R.; Mateos-Buro´n,
L.; Rodr´ıguez-Lo´pez, M.; Rubio-Gonza´lez, R. Tetrahedron 2006, 62, 8933–
8942. (e) Mart´ın-Rodr´ıguez, M.; Gala´n-Ferna´ndez, R.; Marcos-Escribano, A.;
Bermejo, F. A. J. Org. Chem. 2009, 74, 1798–1801.
(7) Ferna´ndez-Mateos, A.; Herrero Teijo´n, P.; Rabanedo Clemente, R.; Rubio
Gonza´lez, R. Tetrahedron Lett. 2006, 47, 7755–7758.
(8) (a) Ferna´ndez-Mateos, A.; Mateos Buro´n, L.; Rabanedo Clemente, R.;
Ramos Silva, A. I.; Rubio Gonza´lez, R. Synlett 2004, 1011. For a radical cascade
policyclization terminated on nitrile, see: (b) Ferna´ndez-Mateos, A.; Herrero
Teijo´n, P.; Rabanedo Clemente, R.; Rubio Gonza´lez, R.; Sanz Gonza´lez, F.
Synlett 2007, 2718. For applications of this method to the synthesis of ꢀ-lactams,
see: (c) Monleo´n, L. M.; Grande, M.; Anaya, J. Tetrahedron 2007, 63, 3017–
3025. (d) Monleo´n, L. M.; Grande, M.; Anaya, J. Synlett 2007, 1243–1246.
(9) Fernandez-Mateos, A.; Herrero Teijo´n, P.; Mateos Buro´n, L.; Rabanedo
Clemente, R.; Rubio Gonzalez, R. J. Org. Chem. 2007, 72, 9973.
10.1021/jo900479v CCC: $40.75
Published on Web 04/15/2009
2009 American Chemical Society
J. Org. Chem. 2009, 74, 3913–3918 3913

Ferna´ndez-Mateos et al.
SCHEME 1. Coupling of Epoxides with Nitriles Induced by
Titanocene
ride, but that a bulkier catalyst such as (c-C6H11C5H4)2TiCl was
needed for epoxynitriles.
Epoxynitrile cyclization mediated by titanocene chloride, in
both a stoichiometric and catalytic way, is a powerful route to
ꢀ-hydroxycycloalkanones. Taking this reaction as a starting
point, here we investigated the feasibility of the intermolecular
reaction between epoxides and nitriles mediated by titanocene
chloride, which would lead to the synthesis of aldols. To the
best of our knowledge, little attention has been devoted to the
topic apart from the paper published by Shono et al.12 on
intermolecular radical additions to nitriles. In that work the
radicals were generated electrochemically from ketones.
SCHEME 2. Reaction Pathways
Results and Discussion
Here we report our findings on the intermolecular reaction
between epoxides and nitriles mediated by titanocene chloride
(Scheme 1).
Reactions were carried out using 2 equiv of titanocene
dichloride and 5 or 10 equiv of the nitrile acceptor. In almost
all cases, the formation of the aldol products was observed. The
regiochemical outcome of the reaction is in agreement with what
has been previously observed with nonsymmetrical oxiranes.13
Epoxide opening is directed by nonbonding interactions during
electron transfer.
cis diastereoselectivity for the addition of acetonitrile in this
case was 86:14. The coupling reactions between the cyclopen-
tene oxide 9 and nitriles were completely diastereoselective;
only the trans-aldols were obtained in both examples (entries 7
and 8). Norbornene oxide 12 gave a 54:46 ratio of trans/cis
aldols with acetonitrile (entry 9). The norbornyl radical in this
example is attacked almost equally from both faces. The intrinsic
stereochemical outcome observed for norbornyl radicals is exo
selectivity, as reported in Giese’s seminal studies16 and subse-
quent work by Gansa¨uer,15a who for the reaction of norbornene
oxide and acrylate as an acceptor, mediated by titanocene
chloride, found an exo/endo ratio of 82:18. In the same work15a
Gansa¨uer reported that the bulky (t-BuCp)2TiCl catalyst changed
the exo/endo ratio to 53:47. The reaction of 12 with benzonitrile
gave a 50:50 ratio of trans/cis aldols (entry 10). The terminal
epoxides 15 and 18 gave the expected aldols in similar yields
as the previous epoxides. Finally, the reaction of the tetrasub-
stituted epoxide, terpinolene oxide, 21, gave only the reduction
product R-terpineol 22 regioselectively. The lack of coupling
products in the latter reaction means that it is sensitive to steric
factors.
The scope and limitations of the coupling reaction shown
above were studied with a new series of functionalized epoxides,
which could react with titanocene chloride by alternative
competing pathways.
The reaction of R,ꢀ-epoxyketones with Cp2TiCl in THF/
MeOH affords ꢀ-hydroxyketones in good yields.17 In the
postulated reaction mechanism, a single electron transfer from
Cp2TiCl to the oxirane generates the radical intermediate,
which upon reaction with Cp2TiCl produces the enolate.
The reaction takes place in two steps (Scheme 2). The initial
step is based on the well-documented titanocene-mediated
opening of epoxides, which affords the ꢀ-alkoxy radical
intermediate (A) that further reacts with Cp2ClTi-coordinated
nitrile8-10 to give B. The initial radical (A) could follow other
alternative pathways, such as reacting with Cp2TiCl, which
abstracts a hydrogen from the methyl group6a,9,14 when R )
Me to give C, or by coupling1 when R ) H to give D. This
1
coupling intermediate could evolve by elimination of (Cp2TiO)2
to afford E, or after hydrolysis by hydrogen interchange1 to
give F.
We carried out the work with a series of cyclic and acyclic
oxiranes and several nitriles. Our results are summarized in
Table 1.
After hydrolysis, the reaction of cyclohexene oxide 1 with
acetonitrile mediated by titanocene chloride affords a diaster-
eomeric mixture of the aldols 2a (trans) and 2b (cis) at a 77:23
ratio. This result is in agreement with previous intermolecular
additions of cyclohexyl radicals to activated alkenes1,15 or
nitriles12 in which the reaction took place preferentially at the
less hindered side to give the trans product. The addition of
benzonitrile occurs in a completely stereoselective manner to
give only the trans-aldol 3. The reaction with other nitriles such
as methyl cyanoacetate or 1-cyclopentenylacetonitrile also gives
the coupling products 5a/b and 4a/b, respectively. The best
couplings yields were obtained for acetonitrile. The reaction of
methylcyclohexene oxide 6 with acetonitrile and benzonitrile
afforded results similar to those described previously. The trans/
(14) (a) Barrero, A. F.; Oltra, J. E.; Cuerva, J. M.; Rosales, A. J. Org. Chem.
2002, 67, 2566–2571. (b) Bermejo, F.; Sandoval, C. J. Org. Chem. 2004, 69,
5275–5280. (c) Cuerva, J. M.; Campan˜a, A. G.; Justicia, J.; Rosales, A.; Oller-
Lopez, J. L.; Robles, R.; Ca´rdenas, D. J.; Bun˜uel, E.; Enrique Oltra, J. Angew.
Chem., Int. Ed. 2006, 45, 5522–5526.
(15) (a) Gansa¨uer, A.; Bluhm, H.; Rinker, B.; Narayan, S.; Schick, M.;
Lauterbach, T.; Pierobon, M. Chem.-Eur. J. 2003, 9, 531–542. (b) Curran, D. P.
Porter, N. A.; Giese, B. Stereochemistry of Radical Reactions; VCH: Weinheim,
1996; pp 128-135.
(16) Giese, B.; Heuck, K.; Lenhardt, H.; Lu¨ning, U. Chem. Ber. 1984, 117,
2132.
(10) (a) Yamamoto, Y.; Matsumi, D.; Hattori, R.; Itoh, K. J. Org. Chem.
1999, 64, 3224–3229. See the Ti(III)-nitrile complex in (b) de Boer, E. J. M.;
Teuben, J. H. J. Organomet. Chem. 1977, 140, 41–45. (c) de Boer, E. J. M.;
Teuben, J. H. J. Organomet. Chem. 1978, 153, 53–57.
(11) Gansa¨uer, A.; Piestert, F.; Huth, I.; Lauterbach, T. Synthesis 2008, 3509.
(12) Shono, T.; Kise, N.; Fujimoto, T.; Tominaga, N.; Morita, H. J. Org.
Chem. 1992, 57, 7175.
(13) (a) Daasbjerg, K.; Svith, H.; Grimme, S.; Gerenkamp, M.; Mu¨ck-
Lichtenfeld, C.; Gansa¨uer, A.; Barchuk, A.; Keller, F. Angew. Chem., Int. Ed.
2006, 45, 2041. (b) Gansa¨uer, A.; Barchuk, A.; Keller, F.; Schmitt, M.; Grimme,
S.; Gerenkampf, M.; Mu¨ck-Lichtenfeld, C.; Daasbjerg, K.; Svith, H. J. Am. Chem.
Soc. 2007, 129, 1359–1371.
(17) Hardouin, C.; Chevallier, F.; Rousseau, B.; Doris, E. J. Org. Chem.
2001, 66, 1046–1048.
3914 J. Org. Chem. Vol. 74, No. 10, 2009

Easy Access to ꢀ-Hydroxyketones
TABLE 1. Cp2TiCl-Mediated Coupling of Epoxides and Nitriles
Subsequent protonation by methanol affords the ꢀ-hydroxy-
ketone17 (Scheme 4).
of the reagent followed by epoxide, and with differents
equivalents of the nitrile. Coupling with other acceptors such
as acrylonitrile or ethyl acrylate was also attempted. The product
was always isophorone 24, in good yield. This result indicates
that the initial stabilized radical generated by the opening of
the oxyrane is trapped faster by the Cp2TiCl than the alternative
acceptors. In the absence of a proton donor, deoxygenation1
The isophorone oxide 23 with acetonitrile (10 equiv) and
titanocene chloride (2 equiv) gave only the deoxygenation
product isophorone 24 (Table 2). The reaction was carried out
under several conditions: reagent was added to a solution of
the substrate and vice versa; acetonitrile was added to a solution
J. Org. Chem. Vol. 74, No. 10, 2009 3915

Ferna´ndez-Mateos et al.
TABLE 2. Cp2TiCl-Mediated Coupling of Functionalized Epoxides and Nitriles
SCHEME 3. Reaction of Terpineol with Cp2TiCl and
CH3CN
SCHEME 4. Deoxygenation of r,ꢀ-Epoxyketones
was the predominant reaction (Scheme 4). Similar results were
found with the epoxide 25 (Table 2, entry 2).
It is known that in the absence of H-atom donors 2,3-
epoxyalcohols react with titanocene chloride to give allylic
alcohols regioselectively18 (Scheme 5). This dehydroxylation
has been explained as occurring through a radical intermediate,
which could be trapped with different kinds of acceptors, such
as hydrogen1b,19a,b (from 1,4-cyclohexadiene), acrylates, or
acrylonitrile.19c In these cases it is clear that acceptor trapping
is faster than the dehydroxylation step.
The epoxyalcohols 27 and 29 afforded opposite results (Table
2, entries 3 and 4), whereas the 2,3-epoxypropanol 27 gave only
3916 J. Org. Chem. Vol. 74, No. 10, 2009

Easy Access to ꢀ-Hydroxyketones
TABLE 3. Influence of Nitrile/Epoxide Ratio
SCHEME 5. Reaction Pathways of 2,3-Epoxyalcohols
entry
epoxide
MeCN equiv
aldol products
1
2
3
4
1
1
1
1
10
5
2.5
1
2a 57%; 2b 17%
2a 60%; 2b 11%
2a 20%; 2b 1.9%
2a 8.6%; 2b 2.4%
SCHEME 7. Competition between Hydrogen Transfer and
Coupling
SCHEME 6. Homocoupling of Vinyl Epoxides
SCHEME 8. Reaction of Pinene Oxide and Acetonitrile
with Titanocene
the aldol coupling product 28 in good yield. The related
epoxyalcohol 29 gave the dehydroxylation product 30. These
results can be explained by the different reaction rates of the
radicals generated from 27 and 29 with the nitrile acceptor.
The secondary radical arising from 27 is trapped quickly by
the nitrile, but the tertiary radical from 29 is not. This means
that dehydroxylation is faster than nitrile trapping when the
radical arising from 2,3-epoxyalcohols is tertiary. We have
previously shown20 that the dehydroxylation rate constant for
slowly23 than the allyl radical to give three kinds of products,
among which is the aldol resulting from the addition of the
benzyl radical to acetonitrile. A similar result was found when
benzonitrile was used as an acceptor. The methyl homologue
38 did not afford addition products with acetonitrile, but two
kinds of homocoupling products, 39 and 41, and a deoxygen-
ation product, 40. This result must be due to steric factors.
Finally, the dimethyl styrene oxide 42 was assayed to observe
the regioselectivity on the oxirane cleavage (benzyl versus
tertiary radical). This time, no addition products were obtained;
instead homocoupling, reduction, and deoxygenation products
were observed. The benzyl radical seems to be the only one
generated. The addition of this radical to nitrile failed presum-
ably as a result of steric requirements.
A series of reactions with different amounts of acetonitrile
were performed to check the influence of the radical acceptor
concentration on product yield (Table 3). As can be seen, the
yield of aldol products was similar for 10 or 5 equiv of
acetonitrile; below these amounts, the reaction yield decreased.
Diastereoselectivity was better for 2.5 equiv (entry 3).
It turned out that all above-reported reactions have a radical
character. The aldol product stereochemistry shown above is a
proof of this character. Indeed, the following three experiments
corroborate this hypothesis. The reaction of cyclohexene oxide
with titanocene chloride and a mixture of 5 equiv of acetonitrile
and 5 equiv of water gave a mixture of cyclohexanol 46 and
the aldol 2 at a ratio of 85:15, respectively. This means that the
addition of the radical to acetonitrile is nearly five times slower
than hydrogen transfer from the TiIII-H2O complex.24
Another fact in favor of the radical mechanism is the reaction
of pinene oxide 47 with Cp2TiCl and acetonitrile. The products
obtained were the trans-carveol 48 and the hydroxyketone 49,
at a ratio of 40:60. This result could be explained in terms of
the homolytic cleavage of oxirane, further cyclobutane open-
these compounds is higher than 1.1 × 107 s-1
.
Radicals derived from the reaction of vinylepoxides with
Cp2TiCl undergo homocoupling,21a deoxygenation,21b,c or
reduction.21b,c Homocoupling has been observed for catalytic
conditions21a and seems to be general (Scheme 6); the deoxy-
genation occurs with 2 equiv of Cp2TiCl and has been observed
for terminal alkenes,21b,c while reduction has been reported for
internal alkenes.21b,c
The reaction of vinyl epoxide 31 with Cp2TiCl (2.2 equiv)
and acetonitrile (10 equiv) afforded quantitatively the homo-
coupling product 32 rather than the addition on the nitrile.
Attempts to trap the allyl radical intermediate with methyl
acrylate were fruitless. The same happened with acrylonitrile,
as has been reported earlier.21a
These results mean that in this case radical homocoupling is
a very fast reaction, with a rate constant higher than 107 M-1s-1,
since it is known that intermolecular addition of radicals to
acrylates or acrylonitrile22 are in the order of k ) 106 M-1 s-1
.
Styrene oxide 33 undergoes homocoupling, addition, and
deoxygenation processes in the reaction with Cp2TiCl (2.2 equiv)
and acetonitrile (10 equiv). The initial benzyl radical reacts more
(18) (a) RajanBabu, T. V.; Nugent, W. A,; Beattic, M. S. J. Am. Chem. Soc.
1990, 112, 6408–6409. (b) Yadav, J. S.; Shekharam, T.; Gadgil, V. R. J. Chem.
Soc., Chem. Commun. 1990, 843.
(19) (a) Chakraborty, T. K.; Das, S. Tetrahedron Lett. 2002, 43, 2313–2315.
(b) Chakraborty, T. K.; Dutta, S. J. Chem. Soc., Perkin Trans. 1 1997, 1257–
1259. (c) Chakraborty, T. K.; Samanta, R.; Das, S. J. Org. Chem. 2006, 71,
3321–3324, and references therein.
(20) Ferna´ndez-Mateos, A.; Herrero Teijo´n, P.; Rabanedo Clemente, R.;
Rubio Gonza´lez, R. Synlett 2008, 3208–3212.
(21) (a) Barrero, A. F.; Qu´ılez del Moral, J. F.; Sa´nchez, E. M.; Arteaga,
J. F. Org. Lett. 2006, 8, 669–672. (b) Yadav, J. S.; Shekharam, T.; Srinivas, D.
Tetrahedron Lett. 1992, 33, 7973–7976. (c) Yadav, J. S.; Shekharam, T.; Gadgil,
V. R. J. Chem. Soc., Chem. Commun. 1990, 11, 843–844. (d) The dimerization
of phenyl-substituted radicals is unique for epoxide opening but has been observed
for oxetane opening: Gansa¨uer, A.; Ndene, N.; Lauterbach, T.; Justicia, J.;
Winkler, I.; Mu¨ck-Lichtenfeld, C.; Grimme, S. Tetrahedron 2008, 64, 11839–
11845.
(23) Newcomb, M. Tetrahedron 1993, 49, 1167.
(24) Jin, J.; Newcomb, M. J. Org. Chem. 2008, 73, 7901–7905. The rate
constant for hydrogen atom transfer reaction of the water complex Cp2TiIIICl-
(22) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340–1371.
H2O to alkyl radicals using indirect kinetic methods is k25 ) 1.0 × 105 M-1 s-1
.
J. Org. Chem. Vol. 74, No. 10, 2009 3917

Ferna´ndez-Mateos et al.
ing,20 and trapping of the tertiary radical by acetonitrile. The
carveol product is formed by hydrogen elimination from the
tertiary radical. This means that the radical addition to aceto-
nitrile is slower than cyclobutane radical cleavage20,25 and faster
than hydrogen elimination. The rate constant of the cyclobutane
radical cleavage in pinene derivatives20 by the radical clock
CDCl3) δ 1.7-2.2 (6H, m), 3.69 (1H, m), 4.61 (1H, q, J ) 5.2
Hz), 7.4-7.6 (3H, m), 8.0-8.1 (2H, m) ppm; 13C NMR (100 MHz,
CDCl3) δ 22.8 (CH2), 28.7 (CH2), 34.9 (CH2), 55.2 (CH), 75.8 (CH),
128.5 (3CH), 133.1 (2CH), 136.6 (C), 201.9 (C) ppm; MS EI, m/z
(relative intensity) 172 (M+ - 18, 31), 148 (1), 144 (3), 133 (13),
105 (100), 77 (91), 51 (36); HRMS (ESI) 213.0884 (M+ + Na,
C12H14O2Na), calcd 213.0886. Anal. Calcd for C12H14O2: C, 75.76;
H, 7.42. Found: C, 75.87; H, 7.44.
Reaction of 15 with Cp2TiCl/CH3CN. According to GP1,
reaction of 15 (100 mg, 0.78 mmol) and CH3CN (0.40 mL, 7.80
mmol) with Cp2TiCl followed by flash chromatography (hexane
6:4 diethyl ether) furnished 16 (101 mg, 76%): IR, ν 3408, 2941,
1701, 1462, 1361, 1191, 1046 cm-1; 1H NMR (200 MHz, CDCl3)
δ 0.86 (3H, t, J ) 6.8 Hz), 1.2-1.6 (10H, m), 2.18 (3H, s), 2.67
(1H, m), 3.71 (2H, m) ppm; 13C NMR (50 MHz, CDCl3) δ 14.2
(CH3), 22.7 (CH2), 27.4 (CH2), 28.2 (CH2), 29.5 (CH2), 29.9 (CH3),
31.8 (CH2), 54.7 (CH), 62.9 (CH2), 213.3 (C) ppm; MS EI, m/z
(relative intensity) 139 (M+ - 33, 1), 125 (1), 111 (2), 97 (2), 88
(91), 70 (88), 55 (100); HRMS (ESI) 195.1343 (M+ + Na,
C10H20O2Na), calcd 195.1356. Anal. Calcd for C10H20O2: C, 69.72;
H, 11.70. Found: C, 69.84; H, 11.73.
method was 1.1 × 107 s-1
.
Finally, we investigated the catalytic version of the reaction
using the conditions reported by Gansa¨uer.2 We found that the
reaction of cyclohexene oxide with 5 equiv of acetonitrile and
0.5 equiv of Cp2TiCl gave a mixture of aldols 2a and 2b at a
ratio of 76:24 and a 21% yield. For 0.2 equiv of Cp2TiCl the
ratio of 2a and 2b was 75:25 and the yield was 4%. For 0.1
equiv of Cp2TiCl no aldol products were observed. This result
also demonstrates the radical character of the reaction, because
in the catalytic conditions the double role of titanocene
(coordination to CN and reduction of iminyl radicals) seems
less likely due to the relatively low catalyst loading.
Reaction of 27 with Cp2TiCl/CH3CN. According to GP1,
reaction of 27 (99 mg, 1.35 mmol) and CH3CN (0.71 mL, 13.51
mmol) with Cp2TiCl followed by flash chromatography (hexane
4:6 diethyl ether) furnished 28 (94 mg, 59%): IR, ν 3382, 2929,
1695, 1443, 1096 cm-1; 1H NMR (200 MHz, CDCl3) δ 2.27 (3H,
s), 2.75 (1H, t, J ) 4.6 Hz), 3.97 (4H, dq, J1 ) 4.6 Hz, J2 ) 7.4
Hz) ppm; 13C NMR (50 MHz, CDCl3) δ 29.3 (CH3), 55.5 (CH),
61.9 (2CH2), 210.6 (C) ppm; HRMS (ESI) 141.0533 (M+ + Na,
C5H10O3Na), calcd 141.0528. Anal. Calcd for C5H10O3: C, 50.84;
H, 8.53. Found: C, 50.91; H, 8.52.
Experimental Section
General Procedure 1 (GP1). Reaction of epoxides and
nitriles with Cp2TiCl. A mixture of Cp2TiCl2 (2.20 mmol) and
Zn (6.60 equiv) in strictly deoxygenated THF (10 mL) was stirred
at room temperature until the red solution turned green. In a separate
flask, the epoxy compound (1 mmol) and the nitrile were dissolved
in strictly deoxygenated THF (10 mL). The green Ti(III) solution
was slowly added via cannula to the epoxide and nitrile solution.
After 30 min, an excess of saturated NaH2PO4 was added, and the
mixture was stirred for 20 min. The mixture was filtered to remove
insoluble titanium salts. The product was extracted into ether, and
the combined organic layers were washed with saturated NaHCO3
and brine, dried (Na2SO4), and filtered. After removal of the solvent,
the crude product was purified by flash chromatography.
Conclusion
The titanocene-promoted intermolecular coupling of epoxides
with nitriles was successful in providing easy access to
ꢀ-hydroxyketones diastereoselectively. The coordination of
Cp2TiCl to the cyano group plays a key role in the reaction
addressed here. As a result, the LUMO of the cyano group is
lowered, and radical coupling proceeds irreversibly without the
formation of unstable iminyl radical intermediates. In this
situation, a low concentration of the Ti(III) reagent is unfavor-
able. An excess of the nitrile is required for better yields of the
coupling products. Functionalized epoxides such as epoxyke-
tones, epoxyalcohols, vinylepoxides, arylepoxides, and tetra-
substituted epoxides do not generally give addition products with
nitriles because of alternative faster radical reactions or steric
factors. Several proofs of the reactions’ radical mechanism are
given.
Reaction of 1 with Cp2TiCl/CH3CN. According to GP1,
reaction of 1 (100 mg, 1.02 mmol) and CH3CN (0.27 mL, 5.09
mmol) with Cp2TiCl followed by flash chromatography (hexane
7:3 diethyl ether) furnished 2a (81 mg, 56%) and 2b (30 mg, 21%).
Data for trans isomer 2a: IR, ν 3395, 2975, 1701, 1430, 1059
cm-1; 1H NMR (400 MHz, CDCl3) δ 1.20 (4H, m), 1.72 (2H, m),
1.93 (2H, m), 2.15 (3H, s), 2.34 (1H, m), 2.83 (1H, bs), 3.75 (1H,
m) ppm; 13C NMR (100 MHz, CDCl3) δ 24.3 (CH2), 25.2 (CH2),
27.9 (CH2), 29.0 (CH3), 33.7 (CH2), 58.8 (CH), 70.6 (CH), 212.8
(C) ppm; MS EI, m/z (relative intensity) 124 (M+ - 18, 3), 81
(45), 71 (100), 55 (38); HRMS (ESI) 165.0894 (M+ + Na,
C8H14O2Na), calcd 165.0886. Anal. Calcd for C8H14O2: C, 67.57;
H, 9.92. Found: C, 67.67; H, 9.94.
Data for cis isomer 2b: IR, ν 3427, 2970, 1698, 1434, 1062 cm-1
;
1H NMR (400 MHz, CDCl3) δ 1.1-2.0 (8H, m), 2.18 (3H, s), 2.46
(1H, m), 3.18 (1H, bs), 4.20 (1H, m) ppm; 13C NMR (100 MHz,
CDCl3) δ 19.7 (CH2), 23.3 (CH2), 25.3 (CH2), 28.7 (CH3), 31.8
(CH2), 53.9 (CH), 66.2 (CH), 213.9 (C) ppm; MS EI, m/z (relative
intensity) 124 (M+ - 18, 2), 81 (42), 71 (100), 55 (22); HRMS
(ESI) 165.0891 (M+ + Na, C8H14O2Na), calcd 165.0886. Anal.
Calcd for C8H14O2: C, 67.57; H, 9.92. Found: C, 67.64; H, 9.95.
Reaction of 9 with Cp2TiCl/PhCN. According to GP1, reaction
of 9 (100 mg, 1.19 mmol) and PhCN (1.21 mL, 11.90 mmol) with
Cp2TiCl followed by flash chromatography (hexane 7:3 diethyl
ether) furnished 11 (131 mg, 58%) as a colorless oil: IR, ν 3401,
Acknowledgment. Financial support for this work from the
Ministerio de Ciencia y Tecnolog´ıa of Spain (CTQ2005-05026/
BQU) and the Junta de Castilla y Leo´n (SA079A06) is gratefully
acknowledged. We also thank the Universidad de Salamanca
for the fellowship to P.H.T.
Supporting Information Available: . Experimental proce-
1
dures and copies of H and 13C NMR spectra for all new
2948, 1669, 1361, 1229, 1084, 995 cm-1; H NMR (400 MHz,
1
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(25) (a) Kang, Y. H.; Kice, J. L. J. Org. Chem. 1984, 49, 1507–1511. (b)
Kenney, R. L.; Fisher, G. S. J. Org. Chem. 1974, 39, 682–686.
JO900479V
3918 J. Org. Chem. Vol. 74, No. 10, 2009