Carbenoid insertion into alkenylzirconocenes - A convergent synthesis of functionalised allylmetallics

Article abstract of DOI:10.1016/S0040-4039(00)01024-8

Insertion of lithium alkylcarbenoids RC(A)LiX, A = CN, P(O)(OEt)₂, SO₂Ph, OMe into alkenylzirconocene chlorides derived by hydrozirconation of terminal alkynes affords functionalised allylzirconium reagents which may be further elaborated. (C) 2000 Elsevuer Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01024-8

Tetrahedron Letters 41 (2000) 6211±6216
Carbenoid insertion into alkenylzirconocenesÐa convergent
synthesis of functionalised allylmetallics
Alexander N. Kasatkin and Richard J. Whitby*
Department of Chemistry, Southampton University, Southampton, Hants SO17 1BJ, UK
Received 21 June 2000
Abstract
Insertion of lithium alkylcarbenoids RC(A)LiX, A=CN, P(O)(OEt)2, SO2Ph, OMe into alkenylzirconocene
chlorides derived by hydrozirconation of terminal alkynes a€ords functionalised allylzirconium reagents
which may be further elaborated. # 2000 Elsevier Science Ltd. All rights reserved.
An important aim of synthetic organic chemistry is to develop new methods for forming carbon±
carbon bonds. Ecient synthesis of complex molecules requires methods which form several
carbon±carbon bonds in one operation. Within organometallic chemistry a hardly investigated
reactionÐthe insertion of lithium carbenoids into metal±carbon bonds via a 1,2-metallate
rearrangement (Eq. (1))1Ðholds particular promise for the above since the product retains the
organometallic functionality of the starting material, i.e. the process is inherently iterative. We
now describe one application of this idea with the convergent formation of allylzirconium
compounds by insertion of an alkyl carbenoid into a alkenylzirconocene, and their further elaboration.
ꢀ1†
Allylmetallic reagents are generally highly reactive towards electrophiles and have an important
place in organic synthesis.2 The possibility of controlling regio- and stereochemistry when using
allyl organometallics has prompted the investigation of many metal systems, and those of the
group IV transition metals have proven extremely useful.2 The normal route to allyl organometallics
is from the allyl halide by oxidative addition reaction, or by transmetallation from a species so
formed. The use of low valent metal complexes Cp2Zr(1-butene) and (iPrO)2Ti(propene) allows
* Corresponding author. E-mail: rjw1@soton.ac.uk
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01024-8

6212
allyl alcohol derivatives to be used as precursors.3 Hydrozirconation of allenes is another route to
allylzirconocene reagents.4 A particularly attractive route to allylmetallic species, which provides
an easy access to a wide range of previously inaccessible reagents, is the insertion of a saturated
metal carbenoid into an alkenyl±metal bond. However, this `homologation' approach is almost
unexplored, the insertion of LiCH2Cl, LiCHCl2 and LiCH(Cl)SiMe3 into alkenyl boronates,5
LiCH(Cl)SiMe3 into (E)-1-octenylzirconocene chloride,1b LiCH(Cl)SiMe2Ph into (E)-1-
nonenyl(diisobutyl)aluminium,1b and ICH2ZnI into alkenylcopper species6 are the only examples
of which we are aware. Following our interest in the insertion of carbenoids into zirconacycles7
and organozirconocene chlorides,8 recently we communicated the high yielding formation of silyl-
substituted allylzirconium compounds by insertion of silylated lithium carbenoids into alkenylzirco-
nocene chlorides.9 We now report an extension of this work to form a range of functionalised
allylzirconium species.
Deprotonation of methyl chloromethyl ether (MOM-chloride) with lithium 2,2,6,6-tetra-
methylpiperidide (LiTMP) in THF at ^100ꢁC in the presence of (E)-1-octenylzirconocene chloride
(1) (readily formed by hydrozirconation of 1-octyne) followed by warming to ^60ꢁC and hydrolysis
gave the (E)-allyl ether 4a indicating intermediate formation of the methoxy-substituted allyl-
zirconocene 3a (Scheme 1, Table 1, entry 1). The carbenoid 2b derived by deprotonation of
2-methoxyethyl chloromethyl ether (MEM-chloride) a€orded a mixture of the allyl ether 4b (E-
isomer) and the alkenyl ether 5b (Z-isomer) in 59% total yield (entry 2), higher than in the
Scheme 1.
Table 1
Protonation products from allylzirconocences 3 (Scheme 1)a

6213
reaction with 2a. We were delighted to ®nd that the easily formed and relatively stable carbenoid
(MeO)CH(PhSO2)Li (2c)10 inserted into 1 to form the allylzirconocene 3a in excellent yield
judging from the hydrolysis products (entry 3). In contrast to 2c, lithiated phenyl methyl
sulphone failed to insert, suggesting that a sulphone only acts as a leaving group when the carbon±
sulphone bond is weakened by the anomeric e€ect of an adjacent alkoxy-fragment.11 Unfortunately,
the substituted a-alkoxysulphones 6 and 7, and the a-acyloxysulphones 8 did not react with 1 in
the presence of LiTMP under conditions described above. Previously some alkoxy-substituted
allylzirconocenes have been formed by the reaction of acetals derived from a,b-unsaturated
aldehydes with Cp2ZrBu2.12
Successful insertion of the carbenoid 2c prompted us to examine other lithium carbenoids stabilised
by electron withdrawing groups. We were delighted to ®nd that the PhSO2-, (EtO)2P(O)- and
CN-substituted lithium carbenoids 2d±g13 reacted with 1 to a€ord the corresponding functionalised
allylzirconium compounds 3d±g in good yield. In all cases hydrolysis of the zirconium reagents
gave only the b,g-unsaturated products 4 as exclusively (Table 1, entries 4, 6, 7) or predominately
(entry 5) (E)-isomers. The `one pot' procedure provides a very convenient route to b,g-unsaturated
sulphones, phosphonates and nitriles widely used in organic synthesis.
Scheme 2.
Table 2
Reaction of oxygen-substituted allylzirconocenes 3a,b with aldehydes (Scheme 2)

6214
To make best use of the synthetic method we needed to elaborate the intermediate allylzirconocenes
with electrophiles other than H+. Addition of benzaldehyde to 3b followed by hydrolysis gave the
homoallylic alcohol 9a in reasonable yield and good anti-diastereoselectivity (Scheme 2, Table 2,
entry 1). The stereochemistry of 9a was proven by its conversion to the trans-g-lactone 10a.14
Reaction with an aliphatic aldehyde gave similar stereochemical results but the expected alcohol
9b was obtained in lower yield (entry 3). Remarkably, the addition of benzaldehyde to 3b in the
.
presence of BF3 Et2O a€orded predominantly the syn-(E)-stereoisomer of 9a, absent in the
uncatalysed reaction (entry 2).15 Subsequent treatment with acidic MeOH followed by oxidation
led to predominantly the cis-g-lactone.14 The methoxy-substituted allylzirconocene 3a formed by
the insertion of 2c reacted with both benzaldehyde and propanal to give the corresponding
alcohols 9c and 9d in excellent yield (entries 5, 6). The stereochemical results were similar to those
in the elaboration of 3b.
Although the reaction of the PhSO2- and (EtO)2P(O)-substituted allylzirconium reagents 3d,e with
benzaldehyde gave a complex mixture of products, the addition of PhCHO to the cyano-substituted
allylzirconocenes 3f and 3g led to the b,g-unsaturated b0-hydroxynitriles 11 (a mixture of
diastereoisomers in both cases) in good yield (Scheme 3). The lack of diastereocontrol, and the
fact that the regiochemistry does not depend on the presence of an a-substituent in 3 (A=CN,
R2=H or Me) is interesting and might be explained by localisation of the zirconium on the
nitrogen. We have previously noted that in similar silicon-substituted systems (3, A=SiMe3) the
presence of an a-substituent (R2ˆH) reversed the regiochemistry of the reaction with aldehydes.9
Elimination of water from 11 (R2=H) gave the 2-cyano-1,3-diene 12 as a mixture of (Z,E)- and
(E,E)-isomers.
Scheme 3.
Overall we have shown that the insertion of readily available functionalised lithium carbenoids
into alkenylzirconocene chlorides represents an ecient and practical route to a range of allyl-
zirconium compounds; the RO- and CN-substituted allylzirconocenes can be further elaborated
by the reaction with aldehydes.16
Acknowledgements
We thank the Engineering and Physical Sciences Research Council (EPSRC) for funding this
work (GR/MO3283). R.J.W. thanks P®zer for generous uncommitted support.

6215
References
1. (a) Negishi, E.; Akiyoshi, K. J. Am. Chem. Soc. 1988, 110, 646; (b) Negishi, E.; Akiyoshi, K.; O'Connor, B.;
Takagi, K.; Wu, G. J. Am. Chem. Soc. 1989, 111, 3089; (c) Kocienski, P.; Barber, C. Pure Appl. Chem. 1990, 62,
1933.
2. (a) Rousch, W. R. In Comprehensive Organic Synthesis; Trost, B. M.; Fleming, I., Eds.; Pergamon Press: Oxford,
1992; Vol. 5, pp. 1±54; (b) Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207; (c) Chan, T. H.; Wang, D. Chem.
Rev. 1995, 95, 1279.
3. (a) Kasatkin, A.; Nakagawa, T.; Okamoto, S.; Sato, F. J. Am. Chem. Soc. 1995, 117, 3881; (b) Ito, H.; Nakamura,
T.; Taguchi, T.; Hanzawa, Y. Tetrahedron 1995, 51, 4507.
4. (a) Maeta, H.; Hasegawa, T.; Suzuki, K. Synlett 1993, 341; (b) Chino, M.; Matsumoto, T.; Suzuki, K. Synlett
1994, 359; (c) Chino, M.; Liang, G. H.; Matsumoto, T.; Suzuki, K. Chem. Lett. 1996, 231.
5. (a) Brown, H. C.; Phadke, A. S.; Bhat, N. G. Tetrahedron Lett. 1993, 34, 7845; (b) Matteson, D. S.; Majumdar, D.
Organometallics 1983, 2, 1529; (c) Tsai, D. J. S.; Matteson, D. S. Organometallics 1983, 2, 236; (d) Matteson, D. S.;
Majumdar, D. J. Organomet. Chem. 1980, 184, C41; (e) Ho€mann, R. W.; Dresely, S.; Lanz, J. W. Chem. Ber.
1988, 121, 1501.
6. Sidduri, A.; Rozema, M. J.; Knochel, P. J. Org. Chem. 1993, 58, 2694.
7. (a) Fillery, S. M.; Gordon, G. J.; Luker, T.; Whitby, R. J. Pure Appl. Chem. 1997, 69, 633, and references cited
therein; (b) Gordon, G. J.; Whitby, R. J. J. Chem. Soc., Chem. Commun. 1997, 1045 and 1321; (c) Gordon, G. J.;
Luker, T.; Tuckett, M. W.; Whitby, R. J. Tetrahedron 2000, 56, 2113.
8. (a) Kasatkin, A.; Whitby, R. J. Tetrahedron Lett. 1997, 38, 4857; (b) Kasatkin, A.; Whitby, R. J. J. Am. Chem.
Soc. 1999, 121, 10208.
9. Kasatkin, A. N.; Whitby, R. J. Tetrahedron Lett. 1999, 40, 9353.
10. (a) Mandai, T.; Hara, K.; Nakajima, T.; Kawada, M.; Otera, J. Tetrahedron Lett. 1983, 24, 4993; (b) Habermann,
A.-K.; Julia, M.; Verpeaux, J.-N.; Zhang, D. Bull. Soc. Chim. Fr. 1994, 131, 965; a-substituted analogs of 2c were
also reported, see: (c) Ley, S. V.; Lygo, B.; Sternfeld, S.; Wonnacott, A. Tetrahedron 1986, 42, 4333.
11. Electrophilic behaviour of lithiated 1-alkoxyalkyl sulphones was previously demonstrated by Julia, see Ref. 10b.
12. (a) Ito, H.; Taguchi, T.; Hanzawa, Y. Tetrahedron Lett. 1992, 33, 7873. Alkoxy-substituted allylzirconium reagents
prepared by transmetallation of allyllithiums with Cp2ZrCl2 have also been reported: (b) Yamamoto, Y.;
Komatsu, T.; Maruyama, K. J. Organomet. Chem. 1985, 285, 31; (c) Yamamoto, Y.; Saito, Y.; Maruyama, K.
J. Organomet. Chem. 1985, 292, 311.
13. Lithiated (EtO)2P(O)CH2Cl: (a) Waschbusch, R.; Carran, J.; Marinetti, A.; Savignas, P. Chem. Rev. 1997, 97,
3401, and references cited therein. Lithiated PhSO2CH2Cl: (b) Durst, T.; Tin, K.-C.; Reinach-Hirtzbach, F.;
Decesare, J. M.; Ryan, M. D. Can. J. Chem. 1979, 57, 258; (c) Makosza, M.; Podraza, R.; Bialecki, M. Gazz.
Chim. Ital. 1995, 125, 601. Lithiated a-chloronitriles are well known intermediates in the Darzens reaction (see, for
instance: Mongelli, N.; Animati, F.; D'Alessio, R.; Zuliani, L.; Gandol®, C. Synthesis 1988, 310); however, to the
best of our knowledge there is no information in the literature on their generation in the absence of an
electrophile.
14. The stereochemistry of the cis- and trans-g-lactones 11 was determined by comparison of their 1H NMR data with
those of cis- and trans-4-propyl-5-phenyltetrahydro-2-furanones (for 11a, see: Sekine, M.; Nakajima, M.; Kume,
A.; Hashizume, A.; Hata, T. Bull. Chem. Soc. Jpn. 1982, 55, 224), and cis- and trans-4-pentyl-5-propyltetrahydro-
2-furanones (for 11b, see: Kosugi, H.; Tagami, K.; Takahashi, A.; Kanna, H.; Uda, H. J. Chem. Soc., Perkin
Trans. 1 1989, 935).
15. A similar reversal of diastereoselectivity in the reaction of alkoxy-substituted allylzirconocene species with
aldehydes has been reported, see Ref. 12.
16. The following procedure for the preparation of 9c (entry 4 in Table 2) is representative: A solution of 1 was
prepared by addition of 1-octyne (94 mg, 0.85 mmol) to a suspension of Cp2Zr(H)Cl (253 mg, 0.98 mmol) in THF
(3.5 mL) followed by stirring at 20ꢁC for 1 h. The reaction mixture was cooled to ^100ꢁC, and a solution of
methoxymethyl phenyl sulphone (206 mg, 1.11 mmol) in THF (4.0 mL) was added followed by slow addition of a
solution of LiTMP [preformed from TMP (157 mg, 1.11 mmol) and BuLi (0.44 mL, 2.5 M in hexanes, 1.11 mmol)
in THF (2.5 mL) at 0ꢁC]. After warming to ^60ꢁC over 1 h benzaldehyde (180 mg, 1.70 mmol) was added, the
mixture was allowed to warm to room temperature, stirred for 3 h and hydrolysed with sat. NaHCO3 aq. (10 mL).
The products were extracted into ether (3Â6 mL), the organic layer was washed with brine (2Â10 mL), dried over
MgSO4 and concentrated under reduced pressure. The residue was puri®ed by column chromatography (silica gel,

6216
20% ethyl acetate in 40±60 petrol ether) to a€ord 1-phenyl-2-hexyl-4-methoxy-3-buten-1-ol (9c) (156 mg, 70%
yield) as a mixture of three stereoisomers. 1H NMR (300 MHz, CDCl3): ꢀ=0.82 (m, 3H), 1.05±1.50 (m, 10H), 2.15
(m, 1H), 2.37 (br. s, OH), 3.58 (s, 3H), 4.33 (m, 1H), 4.52 (dd, J=15.5 and 9.8 Hz, 1H), 6.40 (d, J=15.5 Hz, 1H),
7.36 (m, 5H) (anti,E-isomer); 2.48 (br. s, OH), 2.78±2.95 (m, 1H), 3.55 (s, 3H), 4.24 (dd, J=9.9 and 6.3 Hz, 1H),
6.20 (d, J=6.3 Hz, 1H) (anti,Z-isomer); 2.90 (m, 1H), 3.46 (s, 3H), 3.95 (dd, J=9.9 and 6.3 Hz, 1H), 4.58 (m, 1H),
5.90 (d, J=6.3 Hz, 1H) (syn,Z-isomer). 13C NMR (75 MHz, CDCl3): ꢀ=14.23, 22.76, 27.38, 29.25, 31.53, 31.90,
47.66, 56.31, 77.51, 102.97, 127.20, 127.69, 128.33, 142.82, 150.22 (anti,E-isomer); 27.29, 29.31, 43.22, 59.82, 78.00,
107.36, 127.53, 128.22, 143.26, 149.49 (anti,Z-isomer).