HETEROCYCLES

Review | Regular issue | Vol. 83, No. 3, 2011, pp. 447-489
Received, 3rd October, 2010, Accepted, 16th December, 2010, Published online, 11th January, 2011.
DOI: 10.3987/REV-10-682

■ Tandem Reactions Initiated by the Conjugate Addition of Chalcogen Compounds ― Utilization and Synthesis of Heterocycles ―

Tadashi Kataoka and Shin-ichi Watanabe*

College of Pharmacy, Kinjo Gakuin University, 2-1723 Omori, Moriyama-ku, Nagoya 463-8521, Japan

Abstract

Alkynyl- and alkenylselenonium salts reacted with nucleophiles at the α- or β-carbon depending upon the nucleophiles. The α-attack caused the addition-elimination reaction; i.e., the apparent substitution reaction and the β-attack (the conjugate addition) generated an ylide, which brought about the tandem reaction to form a variety of heterocyclic compounds. Some new reactions proceeded via the selenuranes formed by the attack of a nucleophile on the positively charged selenium atom.
An interesting tandem Michael-aldol reaction of enones (ynones) bearing a chalcogenide or a thioamide was developed. The reactions of the 1-[2-(methylchalcogeno)phenyl]propenones gave α-(α-hydroxyalkyl)enones (Morita-Baylis-Hillman adducts) after a work-up with Et3N. The reactions of the 3-cinnamoyl-1,3-oxazolidine-2-thiones with aldehydes gave tricyclic compounds with a bridgehead bound to four heteroatoms. The asymmetric reactions simultaneously induced four stereocenters, three of which are contiguous. Removal of the chiral auxiliary provided 1,3-diols bearing three consecutive stereocenters.

1. INTRODUCTION
The Michael reaction is the most significant C-C bond-forming reactions and a powerful synthetic tool in organic synthesis due to the various combinations between donors and acceptors.1 One of the properties of the Michael reaction is the formation of a stabilized carbanion at the α−position of the electron-withdrawing group in an acceptor after the 1,4-conjugate addition. The formation of this Michael adduct followed by the next C-C bond-forming reaction is applied to useful tandem reactions such as the Michael-aldol reaction known as Robinson annulation,2 Michael-Claisen reaction,3 and the double Michael reaction.4 Chalcogen compounds, such as thiolate and selenolate ions, are very useful nucleophiles for Michael reactions, and asymmetric Michael addition reactions have recently been studied intensively.5,6 This review describes the Michael reaction of alkenes or alkynes substituted by an electron-withdrawing chalcogen group and the intramolecular Michael addition-aldol reaction of chalcogenides or thiocarbamates in which novel heterocyclic compounds are produced and the heterocyclic systems play important roles.

2. STUDIES OF ALKYNYL, ALKENYL, AND ALLENYLSELENONIUM SALTS
The Michael acceptor possesses an electron-withdrawing and resonance-stabilizing activating group, such as the carbonyl, nitrile, sulfonyl and nitro groups, which stabilizes the anionic intermediate. However, the synthesis of alkynes and alkenes bearing onium cation groups as an electron-withdrawing group and their use for Michael acceptors have been rarely reported. The 1,4-conjugate addition for an alkenyl-onium salt produced the ylide as a conjugate adduct. For example, the conjugate addition of β-formyl alkoxide to a vinylphosphonium salt followed by the intramolecular Wittig reaction produced heterocyclic compounds.7 The utilization of alkynyl- and alkenyliodonium salts as the acceptor showed intriguing reactivity derived from the hypervalent iodan chemistry.8 The application of these acceptors bearing sulfonio and selenonio groups as the electron-withdrawing group was a somewhat unexplored field when we began our research.9 We were particularly interested in selenonium salts that have an unsaturated carbon-carbon bond.
Because of these factors, we report on the synthesis of novel alkynyl, alkenyl, and allenylselenonium salts and the investigation of their reactivities as the Michael acceptor with a variety of nucleophiles. Furthermore, the novel Michael-aldol reaction discovered in the research of alkenylselenonium salts is also described.

3. THE REACTION OF ALKYNYLSELENONIUM SALTS WITH ACTIVE METHYLENE CARBANIONS10
We first investigated the reactions of alkynylselenonium salts, which were prepared from trimethyl(arylethynyl)silane and diaryl selenoxide in the presence of trifluoromethanesulfonic anhydride, with active methylene carbanions. The reaction of diphenyl(phenylethynyl)selenonium triflate 1 with 2,4-pentanedione 2a and t-BuOK in THF gave the furan derivative 3a (40%) and diphenyl selenide 6 (62%) (entry 1 in Table 1). The yield of this reaction was much improved when the reaction mixture was heated to the refluxing temperature (entry 2). The reactions with other active methylene compounds (2b-f) under the refluxing conditions similarly gave the corresponding furan derivatives in high yields (entries 3-7). However, the reaction with 1,3-indandione 2g afforded the furan derivative 3g in poor yield (2%) with 6 (24%), 4b (30%), and 5 (8%) in entry 8. In sharp contrast, the reactions with benzoylacetonitrile 2f and 1,3-indandione 2g at room temperature only gave the products 4a and 4b in yields of 70% and 39%, respectively (entries 9 and 10).

Based on the result of the thermal reactions of compounds 4a and 4b as described below, it was anticipated that these products were the reaction intermediates that form the furan derivatives 3f and 3g or the ring-opened coupling product 5. Thus, we confirmed that compounds 4a and 4b underwent the ligand coupling reactions to form 3f, 3g and 6 (Table 2). When a solution of 4a in chloroform was allowed to stand at room temperature for 3 days, the furan derivative 3f (77%) and 6 (81%) were obtained (entry 1). The reaction was completed in 8 h under reflux in chloroform (entry 2). The reaction of 4b slowly proceeded to give the furan derivative 3g (13%), 6 (60%) and the ring-opened product 5 (29%) in entry 3.
The crystal structure of the intermediate 4a was more clearly established by an X-ray diffraction analysis (Figure 1). The apical-bond distance of Se-O (2.553Å) is shorter than the sum of the van der Waals radii for selenium and oxygen (3.40Å). The O-Se-C17 bond angle of 173.5° is approximately collinear and close to those of other selenuranes.11 The quadruple average angle of 4a (111.3°) coincided with the mean values (ca. 112°) in the literature.12 The configuration about the selenium atom is a slightly distorted trigonal bipyramid which consisted of two apical Se-O1 and Se-C17 bonds and two equatorial Se-C4 and Se-C24 bonds and a lone-pair at the third equatorial position. These structural features are consistent with a σ-selenurane structure. We isolated the first selenuranes with three carbon-selenium bonds and an oxygen-selenium bond [10-Se-4(C3O)] as the reaction intermediates.13

A plausible mechanism is shown in Scheme 1. The 1,4-conjugate addition of an active methylene carbanion to the alkynylselenonium salt 1 gives a selenonium ylide. Subsequent proton transfer from methine generates an enolate ion, of which the alkoxide ion attacks the selenium atom to form the selenurane intermediate 4. Finally, the ligand coupling reaction produced the furan 3 through path a cleavage or a ring-opened product 5 through path b cleavage.

When amides 7a, b were used instead of the active methylene compounds, oxazole derivatives 8a (50%) and 8b (27%) were obtained (Scheme 2). They would be formed by the ligand coupling reaction of the selenurane 10 through the ylide 9 in a similar manner to the formation of furans.

4. THE REACTION OF ALKENYLSELENONIUM SALTS WITH ACTIVE METHYLENE CARBANIONS14
The tandem Michael addition-cyclization route between the alkynylselenonium salts and active methylene carbanions afforded highly functionalized furan derivatives through vinylselenonium ylide intermediates that caused the intermolecular deprotonation of the active methine moiety. We were interested in the reactivity of the alkylselenonium ylide which would be formed by the Michael addition of an active methylene carbanion to an alkenylselenonium salt instead of an alkynyl one.
(E)-β-Styrylselenonium triflate 11 was prepared from (E)-trimethylstyrylsilane and diphenyl selenoxide in the presence of trifluoromethanesulfonic anhydride in good yield. The reactions of 11 with the active methylene carbanions were examined (Scheme 3). The reaction with a carbanion generated from benzoylacetone and sodium hydride in DMF at 70 °C for 3 hours gave the cyclopropane derivative 12a, the relative configuration of which was determined to be (1R*, 2S*, 3R*) by the NOE measurement, in 69% yield together with a (1R*, 2R*, 3S*) isomer (12%) and a (1S*, 2S*, 3S*) isomer (12%). Other cyclopropanes, 12b and c, with anti relationships between the phenyl group and the other functional groups, were prepared as the main products in moderate to good yields from 2,4-pentanedione or ethyl benzoylacetate. On the other hand, the reaction with diethyl or dibenzyl malonate afforded the diethyl or dibenzyl 2-phenylcyclopropane-1,1-dicarboxylate 13a (62%) or 13b (97%), respectively.

The reactions gave different cyclopropanes depending upon the substituents on the active methylene group, i.e., bearing at least one ketone group or two ester groups (malonates). The interesting results can be explained by the plausible mechanism shown in Scheme 4. Selenonium ylides 14 are formed from the conjugate addition of the carbanion to the alkenylselenonium salt. When the active methylene compound possesses one or more ketone group(s), the ylide carbanion intramolecularly attacks the ketone carbonyl group to form a cyclobutane ring 15, which is followed by the 1,2-migration of the endo carbon-carbon bond accompanied with the elimination of diphenyl selenide to form a cyclopropane derivative bearing 1,2-dicarbonyl groups (path a). This ring contraction reaction is similar to the semibenzylic pathway for the Favorskii rearrangement.15 On the other hand, since the carbonyl groups of the ylide 14 with the malonate moiety are less reactive toward the ylide carbanion, deprotonation of an active methine proton preferentially occurs to form a betaine, followed by intramolecular nucleophilic substitution to give a cyclopropane derivative bearing 1,1-diester groups (path b). Thus, we have found the first example of a tandem Michael-Favorskii-type process in the reactions of alkenyl-substituted onium salts with active methylene carbanions bearing at least one ketone group to produce α,β-dicarbonylcyclopropane derivatives. It is noteworthy that the formation of the cyclopropane skeleton depends on the properties of the functional groups in active methylene compounds.

5. THE REACTION OF ALLENYLSELENONIUM SALTS WITH ACTIVE METHYLENE CARBANIONS16
The addition of nucleophiles to electron-deficient allene compounds is one of the most widely used construction methods in organic synthesis.17 In spite of their interesting features, allenylselenonium salts, which have still not been synthesized, should be targeted to determine their properties.
Allenyl methyl selenides were alkylated with methyl trifluoromethanesulfonate to afford the desired selenonium salts 16a and 16b. Table 3 shows the reactions of the allenylselenonium salt 16 with active methylene carbanions. We first conducted the reaction of 16a with 2,4-pentanedione 2a in DMF (entry 1) and t-BuOH (entry 2) at room temperature for 3 h to afford the dihydrofuran derivative 17a in 32% and 42%, respectively. The reaction with the ketoester 2h in a mixture of t-BuOH and DMF gave 17b in 41% yield (entry 3). On the other hand, the methylene cyclopropane derivative 19 was obtained from the reaction with dibenzyl malonate 2i (entry 4). The structures of 17 and 19 were determined on the basis of their spectral data. Interestingly, the reactions of 16b gave different products from those of 16a. The tetra-substituted furan derivative 18a was produced in 60% yield by the reaction of 16b with 2,4-pentanedione 2a in the presence of sodium hydride in DMF for 24 h (entry 5). The ketoester 2h also reacted with 16b to afford the corresponding furan 18b in 56% yield (entry 6). Furthermore, the reactions with the dibenzyl malonate 2i produced the furan derivative 18c (entry 7), differently from the reaction forming the methylenecyclopropane 19 from 16a.

The reactions of 16 with various active methylene compounds gave furan derivatives 17 and 18 except for the formation of the cyclopropane derivative 19 from 16a and dibenzyl malonate 2i. A plausible reaction mechanism is shown in Scheme 5. The selenonium ylides 20 are formed from the conjugate addition of an active methylene carbanion to the allenylselenonium salt 16. Deprotonation of an active methine proton generates betaine 21. In the case of the selenonium salt 16a, the intramolecular nucleophilic attack of the enolate occurs to afford the dihydrofuran derivatives 17 (path a) when the active methylene compound possesses not less than one acyl group. On the other hand, an active methylene carbanion of the diester derivative 21 (R3 = R4 = OBn), which has greater electron density on the carbon, attacks the primary α-carbon (R2 = H) of the selenonio group to give the methylenecyclopropane derivative 19 in spite of their ring strain (path b). Since carbanions attack the center carbon of the allenyl moiety in 16 from the backside of the R1 (phenyl) group, the geometry of the phenyl group and the nucleophiles has a trans configuration. Meanwhile, the reactions of 16b with active methylene carbanions proceed through path a regardless of the structure of the active methylene compounds. Due to the steric hindrance of the secondary α-carbon (R2 = Bn) to the selenonio group, the nucleophilic attack of the carbanion to the α-carbon (path b) is inhibited and the enolization predominantly proceeds to produce unstable dihydrofurans, which are easily transformed into the more stable furan derivatives 18 under the given reaction conditions.

6. THE REACTION OF AN ALKYNYLSELENONIUM SALT WITH A HYDROXY NUCLEOPHILE18
Although some stable selenonium ylides bearing two electron-withdrawing substituents on the carbanionic center were prepared by diverse methods,19 selenonium ylides bearing an electron-withdrawing group were generated only by the deprotonation of the corresponding selenonium salts.20 Only a β-ketodiarylselenonium ylide stabilized by a carbonyl group was prepared but its reactivities have not been studied.21 Based on our research on the reactivities of the diarylalkynylselenonium salts, we expected that the diarylalkynylselenonium salts would react with an hydroxide ion followed by the enol–keto tautomerization to afford a ketodiarylselenonium ylide (Scheme 6).22 We investigated the reactions of alkynylselenonium salts with lithium hydroxide and the capture of the resulting ylides with aldehydes.

The reaction of p-nitrobenzaldehyde with 2 equiv. of the alkynylselenonium salt 1a in the presence of 3 equiv. of lithium hydroxide at room temperature slowly proceeded, and trans-2-benzoyl-3- (4-nitrophenyl)oxirane and diphenyl selenoxide were obtained in 17% and 17% yields, respectively, as shown in Table 4 (entry 1). The optimization of the experimental conditions clarified that the addition of silver triflate and triethylamine improved the yield of the oxirane. The best yield of 22a was obtained when using 4 equiv. of the alkynylselenonium salt 1a and silver triflate and 6 equiv. of triethylamine and lithium hydroxide toward the aldehyde for 12 h (entry 2). Under the same conditions, several aromatic aldehydes underwent the reactions, and good yields of the desired oxiranylketones 22 were obtained. Electron-withdrawing groups on the aromatic ring accelerated the reaction rate (entries 3-7). The reactions with aliphatic aldehydes were carried out under the same reaction conditions as those used for the aromatic aldehydes. The desired compound 22g was obtained in 54% yield from the reaction with 3-phenylpropanal (entry 8). Additionally, isobutyraldehyde as a chain-branching aldehyde reacted with the ketoselenonium ylide to produce the oxiranyl ketone 22h in 37% yield, and the reaction with 3-methylbutanal showed a better result, producing a 63% yield of the oxirane 22i (entries 9 and 10). NMR analysis showed that these oxirane compounds obtained from the above reactions were only trans-isomers.

On the basis of these results, we propose a plausible mechanism for the reaction of an alkynylselenonium salt with aldehydes in the presence of hydroxide, silver salt and triethylamine (Scheme 7). The triple bond of the alkynylselenonium salt is activated by a silver cation, and a hydroxide ion attacks the β-carbon of the alkynylselenonium salt to form the vinyl ylide, which is transformed into the ketodiphenylselenonium ylide 23 by enol–keto tautomerization accelerated by triethylamine. The ylide reacts with aldehydes to give oxiranylketones together with diphenyl selenide (route A). On the other hand, diphenyl selenoxide is formed by the attack of the hydroxide on a selenonium cation without activation of a triple bond by the silver ion (route B).23 The reason that only trans-oxiranes are formed can be explained by the assumption that an active methine hydrogen in the betaine intermediate is easily deprotonated by excess bases in this system and the resulting thermodynamically stable conformer cyclizes to produce the trans-epoxides 22.

In anticipation of the production of oxiranyl imines, p-toluenesulfonamide was selected as a nucleophile (Scheme 8). The reaction of the alkynylselenonium salt 1 with p-chlorobenzaldehyde and sodium p-toluenesulfonamide in the presence of triethylamine and silver trifluoromethanesulfonate was undertaken under conditions similar to those in Table 4. Unexpectedly, 2-benzoyl-3-(p-chlorophenyl)-1- tosylaziridine 25a, not the oxiranyl imine was obtained in 44% yield as a single cis-isomer. Other aldehydes also reacted to give the tosylaziridine derivatives in moderate yields. The coupling constants of the methine protons on the aziridine ring of 25 (7–8 Hz) indicate that these isomers have a cis geometry. The initial step of the above reaction is presumed to be the formation of an N-sulfonylaldimine and a hydroxide ion from a p-toluenesulfonamide monosodium salt and an aromatic aldehyde. The subsequent reaction of the N-sulfonylaldimine with the β-ketoselenonium ylide 23, which is generated by an alkynylselenonium salt and a hydroxide ion, would lead to producing an aziridine derivative. However, the reaction of the alkynylselenonium salt 1 with N-tosyl-4-nitrobenzaldimine and lithium hydroxide in the presence of triethylamine and silver trifluoromethansulfonate at room temperature for 10 h gave the oxiranylketone 22b in 42% yield, and the desired aziridine derivative was not obtained. Recently, the preparation of aziridines starting from ylides and imines was reported.24 We have achieved a novel type of aziridine formation using the alkynylselenonium salt, aldehyde, and sodium p-toluenesulfonamide.

7. THE REACTION OF ALKYNYLSELENONIUM SALTS WITH SULFUR NUCLEOPHILES22d,e
Next, the reactivities of a sulfinic acid and a thiol as soft nucleophiles toward the selenonium salt were investigated. The reaction of the alkynylselenononium salt 1 with benzenesulfinic acid in i-PrOH gave the (Z)-β-sulfonylvinylselenonium salt 26 in good yield. The stereochemistry of 26 was determined as (Z) by the NOE measurement (8.0%) between the vinylic proton and ortho-protons of the Z-phenyl group. This result indicated that the alkynylselenonium salt underwent the anti-addition of the Michael-type reaction of benzenesulfinic acid in a way similar to the general nucleophilic addition to alkynes (Scheme 9).

The reactions of diphenylalkynylselenonium triflate 1 with various thiols in i-PrOH at room temperature were examined (Scheme 10). The reaction with 2 equiv. of thiophenol for 24 h afforded an inseparable mixture of a conjugate adduct 27a and unreacted starting material 1 in 10% and 20% yields, respectively. The reaction with 1.1 equiv. of thiophenol in the presence of 0.1 equiv. of triethylamine for only 30 min afforded 27a in 62% yield. To confirm the effect of an amine, the reaction with o-aminothiophenol without triethylamine was carried out to afford the desired compound 27b in 90% yield only for 10 min. The reaction with other arenethiol derivatives bearing neutral substituents, such as a chloro, hydroxy, or hydroxy alkyl group, also produced the vinylselenonium salts 27c-f in good yields. Based on these findings, the presence of a small amount of a weak base is very important for the preparation of the β-arylthioalkenylselenonium salts. The interaction between the base and a thiol group activates the nucleophilicity of the thiols. In contrast to arenethiols, the alkanethiols did not give good results. The vinylselenonium salt 27g was generated from the reactions with 2-mercaptoethanol in only 15% yield after neutralization of the reaction mixture with aqueous NaHCO3. The (Z) stereochemistry of 27 was determined by the NOE technique, and the results were attributable to the anti-addition of the thiols with the alkynyl moiety.

8. THE REACTION OF (Z)-β-SULFONYLVINYLSELENONIUM SALT WITH NUCLEOPHILES22b,e
Vinyl sulfones are some of the most important building blocks in organic synthesis because of their versatility. Various synthetic methods to achieve vinyl sulfones have been developed.25 However, there have been only a few reports on the introduction of a substituent at the β-position by manipulation of a simple vinyl sulfone.26 We attempted to prepare β-functionalized vinyl sulfones by using the β-sulfonylvinylselenonium salts because the selenonium group is an effective leaving group.
The reactions of the β-phenylsulfonylvinylselenonium salt 26 with alkoxides in MeCN were investigated (Scheme 11). The selenonium salt smoothly reacted with 1,3-dibromo-2-propanol as an acyclic secondary alcohol in the presence of NaH at -30 °C for 30 minutes to produce the (Z)-β-alkoxyvinyl sulfone 28a in 91% yield. Employment of a bulky secondary alcohol, i.e., diphenyl methanol, also reacted to afford the corresponding vinyl sulfone 28b in high yield under the same conditions. Application of this method to the chiral O-alkyl enol ether synthesis was achieved by the reaction with (+)-1-phenylethanol, and a chiral (Z)-β-alkoxyvinyl sulfone 28c was obtained in 91% yield. The β-alkoxyvinyl sulfones 28 shown in Scheme 10 were only single geometrical isomers. The stereochemistry of compound 28a was determined to be (Z) by NOE enhancement of the ortho-protons of the Z-phenyl group (7.3%) or the methine proton of the geminal 2-bromo-1-(bromomethyl)ethoxy group (18.6%) upon irradiation of the vinyl proton.

The reactivity of alkynylides as carbanions with the vinylselenonium salts was also examined (Scheme 12). The reaction of the diphenylvinylselenonium salt 26 with 1.2 mole equivalents of lithium phenylethynylide, prepared from the reaction of phenylethyne and n-butyllithium in THF at -78 °C, afforded the (Z)-β-alkynylvinyl sulfone 28d in 52% yield. The reactions with other alkynylides similarly produced the corresponding (Z)-enyne sulfone derivatives in good yields. The Z configuration was determined by the NOE experiment on 28f, showing the enhancement of the ortho-protons of the Z-phenyl group (6.1%) upon irradiation of the vinyl proton.

A plausible mechanism for the formation of the β-substituted (Z)-vinyl sulfones 28 from the reactions of the vinylselenonium salt 26 with nucleophiles is shown in Scheme 13. Route A proceeds via the pathway whereby the Michael-type addition of a nucleophile to the β-carbon in the vinyl sulfonyl moiety forms betaine 29 and the subsequent elimination of a selenide leads to the (Z)-vinyl sulfones with retention of its configuration.27 Another pathway, route B, involves the formation of the selenurane intermediate 30, via direct attack of the nucleophile on the selenium atom in the vinylselenonium salt, followed by ligand coupling between the Nu and the vinyl group of 30.28 Both pathways provide feasible explanations of the stereochemical outcome observed in these reactions. The attack of nucleophiles at the α position on the vinyl group in 26 was attributed to the steric hindrance around the β-carbon of 26.29

9. THE REACTION OF (Z)-β-THIOVINYLSELENONIUM SALT WITH NUCLEOPHILES22c,d
The (Z)-β-sulfonylalkenylselenonium salts with two kinds of electron-withdrawing groups on the double bond reacted with a variety of nuclephiles to produce (Z)-β-substituted vinyl sulfones (not

vinylselenonium salts) with retention of their configuration. Interestingly, if the sulfonyl group in the β-sulfonylalkenylselenonium salts is replaced by an alkylthio group, the reactivity of the β-alkylthioalkenylselenonium salts against nucleophiles will be changed, and it is anticipated that the Michael-type addition against the selenonio group would occur at only the β-carbon. On the basis of this background, we investigated the reactions of (Z)-β-thioalkenylselenonium salts with nucleophiles.
First, the reactivity of the alkoxide toward the β-thioalkenylselenonium salts was examined (Scheme 14). The reaction of 27c with isopropoxide in MeCN afforded the 2-alkoxy-1-arylthioethene 31a in 79% yield. A good result was also obtained from the reaction of 27a in the case of the phenoxide derivative. In addition, we tried to prepare a chiral β-alkoxylvinyl sulfide from the reaction with a chiral alkoxide. The reaction of 27a with sodium (+)-1-phenylethanolate, which was prepared from the corresponding alcohol and NaH, produced a chiral β-alkoxylvinyl sulfide 31c in 74% yield at room temperature for 90 min. The NOE experiment of 31a showed an enhancement of the ortho-protons of the cis-phenyl group and the methine proton of the geminal 2-propoxy group upon irradiation of the vinyl proton.

Due to the success of the reactions of the alkenylselenonium salts 27 with an alkoxide in MeCN, we decided to try the synthesis of medium-sized heterocyclic compounds containing sulfur and oxygen atoms via an intramolecular cyclization reaction (Table 5). The reaction of the alkenylselenonium salt 27d with NaH in MeCN at −10 °C for 24 h produced a complex mixture, and the ring closure product 32a was obtained in only 17% yield (entry 1). On the other hand, the reaction of the selenonium salt 27e smoothly proceeded to afford the desired compound, 2-phenyl-5H-4,1-benzoxathiepine 32b, in 67% yield (entry 2). 2-Phenyl-5,6-dihydro-4,1-benzoxathiocine 32c was also prepared from 27f in 54% yield (entry 3). In contrast to the compound bearing a hydroxyphenylthio group 27d, the compound with a hydroxyethyl side chain, 27g, produced the cyclic product 32d in 51% yield. The syntheses of six- and seven-membered heterocycles including sulfur and oxygen atoms have been reported;30 however, only a few studies on the preparation of 5H-4,1-benzoxathiepine derivatives have been published.31 Furthermore, there has been no report on the synthesis of 4,1-benzoxathiocine and related compounds. This new method of preparing medium-sized heterocycles containing sulfur and oxygen atoms is expected to have a wide number of applications.

Next, the reactions of the β-thioalkenylselenonium salt 27a with alkynylides were conducted (Scheme 15). The reaction of 27a with lithium phenylethynylide produced (Z)-4-phenyl-1-(phenylthio)-1-buten-3-yne 31d in 74% yield in THF at −78 °C for 5 h. The reactions with other alkynylides, such as 1-hexynylide and 3,3-dimethyl-1-butynylide, also afforded the desired compounds under the same conditions in good yields. The structures of 31d−f were identified by spectral data. In particular, the 1H NMR spectrum of 31e showed long-range coupling between a vinyl proton and protons of the 5-position (2 Hz). The (Z) stereochemistry of 31d−f was established by observation of the NOE enhancement (10%) between the vinylic proton and ortho-protons of the cis-phenyl group in compound 31d.

A plausible reaction mechanism to produce (Z)-vinyl sulfides consists of the formation of the selenurane intermediate followed by ligand coupling between the Nu group and the alkenylcarbon with retention of the configuration (Scheme 16).

10. REACTIONS OF ELECTRON-DEFICIENT ALKENES WITH ALDEHYDES MEDIATED BY CHALCOGENIDE-TiCl432-35
We discussed the reactions of vinyl selenonium salts with nucleophiles in Section 8 and showed the addition-elimination mechanism as one of their reaction processes in Scheme 13. If the reverse reaction of this mechanism, the conjugate addition of a chalcogenide to an electron-deficient alkene proceeds in the presence of an aldehyde, the reaction shown in Scheme 17 might possibly occur.
A chalcogenide adds to an electron-deficient alkene and forms betaine 33. The betaine 33 reacts with an aldehyde to afford a zwitterionic intermediate 34, the alkoxide moiety of which intermolecularly abstracts the proton α to an electron-withdrawing group and brings about the β-elimination to afford an allyl

alcohol 35. This reaction is regarded as a chalcogenide version of the Morita-Baylis-Hillman （MBH） reaction (the chalcogeno-MBH reaction).
The chalcogeno-MBH reactions successfully proceeded using only a chalcogenide with a Lewis acid. The chalcogeno-MBH adducts were purified by preparative TLC on silica gel to give the desired products.33
Various Lewis acids such as AlCl3, BBr3, BCl3, and TiCl4, were useful, and, especially, TiCl4 exerted excellent effects.33b The chalcogenides shown in Scheme 18 were used. The 8-membered heterocycles 40 and 41 possessing two chalcogen atoms at the 1,5-positions effectively worked because the intermediary onium ion 44 was stabilized by the transannular interaction between the chalcogen atoms forming a hypervalent bond. Selenopyran-4-one, thiopyran-4-one and their 4-thione congeners 42 formed the stable 6π cations 43 by coordination of a Lewis acid at the 4-carbonyl or thiocarbonyl group and were also efficient catalysts.34
The reaction of but-3-en-2-one (37) with p-nitrobenzaldhyde (36a), shown in Scheme 19 as a typical example, afforded the MBH product 38a in good yield. When the product was purified by column chromatography on silica gel, the syn- and anti-chloromethyl aldols 45a were isolated.35
Although the results were different from the anticipated ones, in which a chalcogenide would nucleophilically attack the positively charged β-carbon of an enone, the reaction using a chalcogenide and a Lewis acid was the tandem Michael-aldol reaction from the viewpoint of the reaction mechanism (Scheme 20) and proceeded rapidly in comparison with the MBH reaction. Chalcogenide 46 coordinates with TiCl4 and assists it in releasing a chloride ion. The resulting TiCl4-chalcogenide complex 47, the exact structure of which has not been determined, reacts with enone 37 via the cyclic transition state 48. The stereoisomers of the chloride 45a were easily transformed into the α-methylene-aldol 38a upon treatment with a base or by preparative TLC on silica gel. Therefore, our newly developed reaction can be used as an alternative to the MBH reaction.36

Aromatic and aliphatic aldehydes33b and α-keto esters37 51 were reactive as carbonyl compounds for the reactions. Acyclic and alicyclic enones were very reactive as a Michael acceptor, but acrylonitrile and the acrylates were less reactive than the enones. Acrylic acid thioesters38 52 were quite active in the reactions and therefore, can be used for the synthesis of acrylic acid derivatives instead of the acrylates. Alkynyl ketones and acetylenic acid esters39 53 were applicable for this reaction and gave the β-halo-substituted MBH products 54 (Scheme 21).

The Michael-aldol reactions using TiCl4 have been reported one after another by the research groups of Li,40 Ohshima,41 and Shi.42 Various kinds of Lewis acids and Michael donors have been used for the tandem Michael-aldol reactions of electron-deficient alkenes with aldehydes.43 Recently, Verkade and his co-workers found that the bicyclic proazaphosphatrane sulfide acted as an efficient catalyst for the MBH reaction.44 This catalyst was used with TiCl4 and the MBH reaction was complete within 30 minutes. The reactions of the β-substituted enones, acrylates and acrylonitrile were enabled by the catalyst to give the MBH adducts in high yields. The reactions using this catalytic system did not afford the α-chloromethyl aldols, but only the MBH adducts. This is different from the results of Li40 and us.33,35,38
The MBH reaction that forms a chiral allyl alcohol and asymmetric synthesis using the MBH reaction has been thoroughly studied.36d,g Some reactions induced a high enantioselectivity.45 We conducted the catalytic asymmetric reactions using 0.1 equiv. of various bifunctional catalysts containing a chalcogenide and alcohol, ether or amine group at −20 oC, as shown in Scheme 22. When 0.1 equiv. of chalcogenide was used, the chemical yields were very high, but the optical yields were very low. Good enantioselectivity (71%ee) was achieved using 1 equiv. of 10-methylsulfanylisoborneol, but the chemical yield was decreased.

The asymmetric reaction of a chiral glyoxylate using Me2S-TiCl4 and that using a chiral sulfide-BF3·Et2O have been reported by Bauer46 and Goodman,47 respectively. Our reaction has the merits that it proceeds quickly and that the reactions of α-dicarbonyl compounds, alkynyl ketones and esters, which do not occur under the MBH reaction conditions, proceed smoothly.37,38,46,48

11. TANDEM MICHAEL-ALDOL REACTION OF CHALCOGENIDE-ENONES WITH CARBONYL COMPOUNDS32,50
The reactions of enones with aldehydes using the chalcogenide-TiCl4 were the tandem Michael-aldol reaction initiated by the nucleophilic attack of a chloride ion on the electron-deficient β-carbon of enones and gave the β-chloro-α-hydroxymethyl adducts. Therefore, we returned to the starting point of this research and reconsidered the plan again that the reaction began with the Michael addition of a chalcogenide to an enone.
Vinyl sulfide 55 was formed, although the yield was low, from the reaction of p-nitrobenzaldehyde (36a) with methyl propiolate in the presence of Me2S-TiBr4, as shown in Scheme 21.39b Nenajdenko et al. reported that 1-ethyl-4-oxo-2,3-dihydrothiopyranium perchlorate was formed upon the treatment of 1-(ethylsulfanyl)pentane-1,3-dien-3-one with perchloric acid.49
These results indicate that sulfides work as Lewis bases and undergo the Michael reaction. We made a new plan for the tandem Michael-aldol reaction in which a key step was the intramolecular Michael addition of the chalcogenide group to the enone moiety activated by a Lewis acid whose conjugate base had a very low nucleophilicity. We first confirmed the intramolecular Michael addition of 1-[2-(methylsulfanyl)phenyl]propenone (56) in the presence of BF3·Et2O to form the cyclic sulfonio-enolate 57 by comparison of the 1H-NMR spectrum of 57 with that of the TMS-enol ether 58 (Scheme 23).50

The reaction conditions for the chalcogeno-enones 56 and 59 with aldehydes 36 were closely examined, and BF3·Et2O and triethylamine were selected as the Lewis acid and quenching base, respectively. The reactions of the sulfide-enone 56 gave MBH products 60, but those of the seleno congener 59 afforded the selenochromanone 62 together with MBH products 61 (Table 6).
To elucidate the reaction mechanism, several experiments involving the intermediates, the sulfonium salts 64, were conducted. Four stereoisomers of the selenonium salts 64 were synthesized by methylation of the syn- and anti-aldols 63 and their stereostructures were determined from the NOEs between the methyl group and the 3-proton (Scheme 24).
Treatment of the syn-sulfonium salt 64a (syn-cis 64a:syn-trans 64a = 3:1) with saturated aqueous NaHCO3 gave the MBH adduct 60a, sulfonium salts 64a, p-nitrobenzaldehyde (36a), and sulfide-enone 56. On the other hand, treatment of the anti-sulfonium salts 64a (anti-cis 64a:anti-trans 64a = 1:1) with

saturated aqueous NaHCO3 gave the Morita-Baylis-Hillman adduct 60a, p-nitrobenzaldehyde (36a), and sulfide-enone 56, and the starting material 64a was not recovered. These findings indicate that the anti-64a more easily undergoes β-elimination than the syn-isomer and that the retro-aldol reaction of 60a also occurs in competition with the β-elimination upon treatment with a base. When the reaction mixture of 56 and 36a was worked up in different ways, the isolated sulfonium salts were different (Scheme 25).

The difference in reactivity between the syn- and anti-64 against a base is explained as shown in Scheme 26. In both isomers, the gauche conformation 66, in which the hydroxybenzyl group is in the equatorial position, is more stable than the antiperiplanar conformation 65 with the axial hydroxybenzyl group. β-Elimination occurs via the antiperiplanar conformation 65. Steric repulsion between the lone pair electrons and the benzylic hydrogen in the anti-65 is much less than that between the lone pair electrons and the group R in the syn-65. Therefore, the anti-65 causes the β-elimination more easily than the syn-65. This is the reason that the anti-64a was not obtained from the reaction of 56 and 36a after working up the reaction mixture with a base.

Based on these results, a possible mechanism for the tandem Michael-aldol reaction mediated by a Lewis acid is shown in Scheme 27. The activation of the enone by BF3∙Et2O allows a chalcogenide to add to the β-carbon of an enone moiety. The diastereoselectivity would be induced in the reaction step of the boron-enolate 57 with the aldehyde 36. The transition state 68 with an equatorial R group is favored over the other transition state 67 with an axial R group, and the anti-isomer anti-64 is favorably formed. When the aldehyde 36 approaches the enolate 57, the methyl group on the chalcogen atom takes a position opposite to the aldehyde, i.e., the trans-configuration. The anti-trans-64 undergoes β-elimination via the antiperiplanar anti-65 on the treatment with a base to give the MBH adduct 60a.

Carbonyl compounds other than aldehydes are used for the MBH reaction, but their use is limited. Ketones except for the trifluoromethyl derivatives reacted with electron-deficient alkenes only under high pressure.51 No reports on the reactions of the enolizable α-diketones have been published. The reaction of α-keto esters progressed only in the good match of an electron-deficient alkene with a Lewis acid.37a,52
Boron enolate is an intermediate of the BF3-mediated intramolecular Michael reaction from the findings reported above and generally reacts with carbonyl compounds under mild reaction conditions.53 Therefore, we next examined the reactions of 56 and 59 with various carbonyl compounds.
The reactions of the acetophenone derivatives 69a, b, and cyclohexanone (69c) at 0 ºC for 30 min gave products 70 and 71 in low to moderate yields (Table 7).
The α-diketones and α-keto esters reacted with 56 and 59 to afford the products 73-75 in low to high yields.50b These results were not satisfactory, but these are the first examples of the MBH-type reactions of the α-dicarbonyl compounds 72b and c which are enolizable under mild conditions (Table 8).

If chalcogenide-ynones react with aldehydes in the presence of BF3·Et2O, the 3-(hydroxymethyl)-　chalcogenochromen-4-one derivatives can be synthesized. This reaction involves the 6-endo-dig cyclization and is interesting from the viewpoint of the Baldwin rule. The cyclization of the 1-(hydroxyaryl)-3-phenylpropynones proceeded in a 6-endo-dig or a 5-exo-dig manner depending upon the reaction conditions,54 whereas the selenium analogs were selectively cyclized in a 6-endo-dig manner under basic conditions.55 The reactions of the chalcogenide-ynone 76 or 77 with aldehydes occurred in a 6-endo-dig fashion to give the 3-(hydroxymethyl)chalcogenochromen-4-one 78 or 79, respectively (Table 9).56

No chalcogenonium salt 80 was obtained because coordination of a boron Lewis acid with an aldol moiety decreased the electron density of the chalcogenopyran ring and demethylation of the onium salt would have easily occurred (Scheme 28).

This method is convenient for the synthesis of the 2-unsubstituted 3-(hydroxymethyl)chalcogeno- chromen-4-one derivatives because the 2-unsubstituted chromenone was ring-opened, upon treatment with lithium diisopropylamide, before alkylation of the newly formed carbanion.57,58 Recently, Basavaiah and his co-workers synthesized the 2-unsubstituted 3-(hydroxymethyl)chromen-4-ones from chromenone using Et3N in methanol, but benzaldehyde and aliphatic aldehydes did not produce the products.59
Acetals function as electrophiles upon treatment with a Lewis acid.60 Noyori61 and other groups62-64 reported the α-alkoxyalkylation of α,β-unsaturated ketones with acetals or orthoesters in the presence of a silicon-Lewis acid. If acetals are used for the chalcogeno-MBH reaction instead of aldehydes, BF3·Et2O plays an important role in generating both the enolate-onium salts and α-alkoxy carbocations, and the α-alkoxyalkylation of the enones will be accomplished.
The reaction of benzaldehyde dimethyl acetal 81a at −40 ºC for 2 h gave the MBH-adduct 82 (78%) or 84 (79%) after work-up of the reaction mixture with Et3N and the onium salt 83 (40%) or 85 (51%) together with 82 or 84, respectively, after work-up with saturated NaHCO3 (Table 10).65 The structures of 83 and 85 were determined by X-ray crystallography and 1H- and 13C-NMR spectroscopies.

Cyclic acetal, 2-phenyl-1,3-dioxolane and trimethyl orthoformate were applicable for this reaction (Table 11).

12. TANDEM MICHAEL-ALDOL REACTION OF N-ENOYLCYCLIC THIOAMIDES WITH ALDEHYDES68,69
As reported above, we found that a chalcogenide group caused the intramolecular Michael addition to an enone moiety50,56,65 and that a thiocarbonyl compound, such as thiopyran-4-thione35 or tetramethylthiourea,39 catalyzed the tandem Michael–aldol reaction. When we started the study on the tandem Michael-aldol reaction of thioamides, the Michael addition of N-unsubstituted thioamides with α,β-unsaturated carbonyl compounds66 was already well known, but that of the N-substituted thiocarbamates was only slightly known.
Based on our findings on the intramolecular Michael cyclization of the chalcogenide group and the catalytic action of thioamides, we studied a new tandem Michael-aldol reaction of N-cinnamoyl cyclic thiocarbamates with aldehydes, as shown in Scheme 29. If this reaction goes well, we can synthesize the β-substituted MBH product 93, which is difficult to prepare by the MBH reaction.
When chiral 1,3-oxazolidine- or 1,3-thiazolidine-2-thione derivatives are used as a chiral auxiliary, an asymmetric tandem Michael–aldol reaction can be developed, and optically active products can be obtained. Palomo and co-workers separately reported the sulfur transfer reaction of a chiral N-enoyl–cyclic moiety with a Lewis acid followed by hydrolysis of the products.67 In their early report, SnCl4 was the most effective Lewis acid, but then the Lewis acid was changed into NbCl5 or BF3·Et2O.

The reaction conditions were examined for the reaction of the N-cinnamoyl-1,3-thiazolidine-2-thione 94a with p-chlorobenzaldehyde (36b) (Table 12), and the highest yield of 95 and 96 was afforded when using 3 equiv. of BF3·Et2O, 2 equiv. of 94a and 1 equiv. of 36b.68

The products were not the MBH adducts, but the tricyclic compounds with a bridgehead carbon bound with four heteroatoms. This reaction induced four chiral centers in the one-step reaction.
The structurally rare compounds 95-99 were formed via the reaction pathways shown in Scheme 30. Coordination of BF3·Et2O with the carbonyl oxygen of the N-cinnamoylthiazolidinethione 94 activates the enone moiety and the intramolecular conjugate addition of the thiocarbonyl group to the enone moiety forms the boron enolate-iminium salt 100. The aldol reaction of the boron enolate yields the aldol product 101, the alkoxide ion of which nucleophilically attacks the iminium carbon to afford the tricyclic products 95-99.

After investigation of the chiral auxiliary, 4S-methyl-5R-phenyloxazoline-2-thione showed the best diastereoselectivity. The reactions of N-cinnamoyl thioamide 102 with aromatic aldehydes afforded tricyclic compounds 103 in moderate to good optical yields, but the reactions with aliphatic aldehydes did not give satisfactory yields (Table 13).

The diastereomeric ratios were determined from the signal intensities of the 1H-NMR spectra. The structure of the m-nitro derivative 103g was determined to be a tricyclic adduct with 1R, 7R, 8R and 11R configurations based on the 3R and 4S-absolute configurations by X–ray crystallography (Figure 2). From this analysis, products contain three consecutive stereocenters and a chiral bridgehead carbon bound to four heteroatoms.

The asymmetric induction of four stereogenic centers can be explained as shown in Scheme 31. The boron enolate-iminium salt 106 consists of two diastereoisomers 106A and 106B, the anti- and syn-configurations between the phenyl group adjacent to the sulfur atom and the substituents of the oxazolidine ring, respectively. The approaches of an aldehyde from the Si- and the Re-faces to the boron enolate moiety in isomer 106A are prevented by the methyl and the phenyl groups of the oxazolidine ring and the phenyl group α to the sulfur, respectively. If the reaction were to proceed via 106A, the chiral carbon adjacent to the sulfur of the product should have an S-configuration, but they in fact had an R-configuration. Therefore, this pathway via intermediate 106A is inappropriate. On the other hand, isomer 106B with the S configuration has three substituents in the Si-face, and an aldehyde can easily attack the enolate carbon from the sterically relaxed Re-face.
Palomo and co-workers obtained β-sulfanylpropanoic acid derivatives 109 with the S-configuration, which is opposite to the 11R-configuration of 95–99 and 103-105, from the sulfur transfer reaction of the chiral N-enoyl-oxazolidinethiones.67 This reaction does not contain the subsequent aldol reaction with an aldehyde. Therefore, thiols 109 were afforded by hydrolysis of the diastereomer 106A, which is more stable than the other 106B.
The induction of three concecutive stereocenters in the aldol reaction of the cyclic enolate 106B is discussed by two possible processes via the cyclic transition state or the acyclic one. If the reaction of

106B with an aldehyde proceeds via the cyclic 6-membered ring transition state, two transition states 107A and 107B are considered. Transition state 107B bearing an equatorial R group is more stable than transition state 107A bearing an axial R group. The reaction progressed via 107B favorably to form the aldol product 108B, which leads to the tricyclic product 104. This is inconsistent with the finding that compound 104 is the minor isomer.
On the other hand, the acyclic transition states shown in Scheme 32 provide a reasonable explanation. In the acyclic transition models, stereoselection in the aldol reaction is governed by the transition state 110A or 110B. The transition state 110A is more favorable than 110B, which has a steric repulsion between the iminium moiety and the R group. The pathway via 110A produces the aldol adduct 108A, which cyclizes to product 103, which is the major product of the reaction. Based on the discussion above, we conclude that the tandem Michael-aldol reaction of 102 with aldehydes proceeds via the boron-enolate-iminium salt 106B and acyclic transition state 110A.
4R-Methyl-5S-phenyloxazoline-2-thione 111, an enantiomer of 102, reacted with p-nitrobenzaldehyde (36a) to give tricyclic compounds 112 and 113, the enantiomers of 103 and 104, respectively (Scheme 33).

To utilize the chiral products, removal of the chiral auxiliary from the tricyclic compounds was examined as shown in Scheme 34. Tricyclic compounds were stable to alkaline hydrolysis but were hydrolyzed with 2M HCl. The acid hydrolysis of 103a, b selectively cleaved the C-S bond of the six-membered ring to give N-(3-sulfanylpropanoyl)oxazolidines 114a (72%) and 114b (78%), respectively, which were converted to the S-methyl derivatives by methylation with MeI-Et3N. Removal of the oxazolidinone 117 was conducted by using EtSLi or LiBH4, but the retro-aldol reaction took place to give 116 and 36b or 118 and 119, respectively. These findings indicated that protection of the hydroxyl group is necessary to prevent compounds 115 from the retro-aldol reaction. Treatment of the trimethylsilyl ether 120 with LiBH4, EtSNa, or MeONa caused the oxazolidine ring to open to give the amide 121. Protection of the hydroxyl group inhibited the retro-aldol reaction, but the bulky trimethylsilyl group interfered with the attack of a nucleophile on the exo-carbonyl group. The nucleophile exclusively attacked the endo-carbonyl group. Therefore, returning to the original concept, various means were tried to remove the chiral auxiliary without protection of the hydroxy group of 115, and the reductive removal of the oxazolidinone moiety using sodium borohydride in aqueous THF successfully produced propanediols 122 and the oxazolidinone 117.

As reported above, the tandem Michael-aldol reaction of N-cinnamoyl-4S-methyl-5R-phenyl- oxazolidinethione with aldehydes afforded structurally rare tricyclic compounds with a bridgehead carbon bound to four heteroatoms. If acetals are used instead of aldehydes for this reaction, the cyclization of the aldol intermediate in the final step to form the tricyclic compounds is prevented, and it is anticipated that the 3-alkoxy-2-(α-sulfanylbenzyl)propionimides 124, which correspond to the hydrolysis-O-alkylation products of 114, will be formed. We have already achieved a simple procedure for the α-alkoxyalkylation of enones via the tandem Michael-aldol reaction of chalcogenide-enones with acetals using BF3∙Et2O, as shown in Table 10.65
SnCl4 as a Lewis acid promoted the reaction better than BF3∙Et2O for the acetals. This is probably because the chloride ion generated from SnCl4 would react with the iminium intermediate to give a more stable chloroamine intermediate. The reactions of N-cinnamoyl-4S-isopropyl-5,5-dimethyloxazolidine-thione (123) with aromatic aldehyde dimethyl acetals 81 are shown in Scheme 35.69
The configurations of the newly created stereogenic centers were assigned by an X-ray structural analysis of the crystalline dimer 125 (Figure 3). The absolute configuration of the benzylic position α to the sulfur

atom of 125 is different from that of the product obtained from the reaction with aldehydes using BF3∙Et2O.

Based on this finding, a plausible reaction mechanism is shown in Scheme 36. The intramolecular Michael addition of the thioamide group of 123 to the enone moiety activated by SnCl4 forms the iminium chloride 127 via the tin enolate 126. The iminium chloride 127 transforms into the more stable cyclic chloroamine 128. Since the Si face is sterically more crowded than the Re face and the SnCl3 moiety occupies the opposite side of the isopropyl, methyl, and phenyl groups, the chloride anion of 127 attacks the iminium carbon from the Re face to form the chloroamine 128. Steric repulsion among the phenyl, methyl, and isopropyl groups of 128B is quite strong on the Si face. The chlorine atom interferes with the Re face attack of the methoxycarbenium ion to the enolate ion. The other isomer 128A has a pseudoequatorial phenyl group, and this conformation relaxes the steric hindrance between the phenyl group and the isopropyl or methyl group and is more stable than 128B. The methoxycarbenium ions can attack from the Si face to form the Michael-aldol product 124 via the cyclic transition state 129. The phenyl group and the chloro substituent block the Re face approach of the carbenium ion.

We have developed the asymmetric tandem Michael−aldol reaction of N-cinnamoylthioamides with aldehydes and acetals. This reaction furnishes diastereomerically pure tricyclic compounds, 2-(α-hydroxybenzyl)- or 2-(α-methoxybenzyl)-3-phenyl-3-sulfanylpropionimides, which contain three contiguous chiral centers. The reductive removal of the chiral auxiliary from them provides 2-(α-methylsulfanylbenzyl)propane-1,3-diols and 2-alkoxybenzyl-3-sulfanylpropanols in good yields (Scheme 37).

ACKNOWLEDGEMENTS
This research was partly supported by a Grant-in-Aid (No. 20590021) from the Ministry of Education, Culture, Sports, Science and Technology (Japan).

References

1. (a) M. E. Jung, ‘Comprehensive Organic Synthesis,’ Vol. 4, ed. by B. M. Trost and I. Fleming, Pergamon, Oxford, 1991, pp. 1-67; CrossRef (b) M. E. Jung, Tetrahedron, 1976, 32, 3; CrossRef (c) R. E. Gawley, Synthesis, 1976, 777; CrossRef (d) P. Perlmutter, ‘Conjugate Addition Reactions in Organic Synthesis,’ Pergamon, Oxford, 1992.
2. (a) W. S. Rapson and R. Robinson, J. Chem. Soc., 1935, 1285; CrossRef (b) J. R. Hawthorne and R. Robinson, J. Chem. Soc., 1936, 763. CrossRef
3. (a) J. J. Miller and P. L. De Benneville, J. Org. Chem., 1957, 22, 1268; CrossRef (b) R. L. Shriner and H. R. Todd, Org. Synth., 1943, 15, 14.
4. G. Stork and M. Tomasz, J. Am. Chem. Soc., 1964, 86, 471. CrossRef
5. (a) A. Kamimura, H. Mitsudera, S. Asano, and S. Kidera, J. Org. Chem., 1999, 64, 6353; CrossRef (b) A. Kamimura, R. Morita, K. Matsuura, H. Mitsudera, and M. Shirai, Tetrahedron, 2003, 59, 9931; CrossRef (c) K. Nishimura and K. Tomioka, Yakugaku Zasshi, 2003, 123, 9; CrossRef (d) D. Enders, K. Luttgen, and A. A. Narine, Synthesis, 2007, 959. CrossRef
6. (a) A. Kamimura, Y. Omata, H. Mitsudera, and A. Kakehi, J. Chem. Soc., Perkin Trans. 1, 2000, 4499; CrossRef (b) J. Jauch, J. Org. Chem., 2001, 66, 609. CrossRef
7. (a) P. K. Tarpey and M. Grayson, J. Org. Chem., 1964, 29, 631; CrossRef (b) R. W. Saalfrank, W. Hafner, J. Markmann, and H. J. Bestmann, Tetrahedron, 1988, 44, 5095; CrossRef (c) E. E. Schweizer, J. Am. Chem. Soc., 1964, 86, 2744; CrossRef (d) S. Catalán-Muñoz, C. A. Müller, and S. V. Ley, Eur. J. Org. Chem., 2010, 183.
8. M. Ochiai, Kikan Kagaku Sosetsu, 1998, 34, 181.
9. (a) D. L. Smith, Z. Wang, and L. S. Jimenez, J. Chem. Res. (S), 1995, 66; (b) J. Gosselck, F. Ahlbrecht, F. Dost, H. Schenk, and G. Schmidt, Tetrahedron Lett., 1968, 995; CrossRef (c) K. Takaki and T. Agawa, J. Org. Chem., 1977, 42, 3303; CrossRef (d) H. Yamanaka, J. Matsuo, A. Kawana, and T. Mukaiyama, ARKIVOC, 2004, 42; (e) Y. Watanabe, Y. Ueno, and T. Toru, Bull. Chem. Soc. Jpn., 1993, 66, 2042; CrossRef (f) S. Watanabe and T. Kataoka, Yuki Gosei Kagaku Kyokaishi, 2003, 61, 583.
10. T. Kataoka, S. Watanabe, K. Yamamoto, M. Yoshimatsu, G. Tanabe, and O. Muraoka, J. Org. Chem., 1998, 63, 6382. CrossRef
11. (a) T. Kataoka, H. Shimizu, K. Tomimatsu, K. Tanaka, M. Hori, and M. Kido, Chem. Pharm. Bull., 1990, 38, 874; (b) T. Takahashi, N. Kurose, S. Kawanami, Y. Arai, and T. Koizumi, J. Org. Chem., 1994, 59, 3262; CrossRef (c) T. Takahashi, N. Kurose, and T. Koizumi, Chem. Lett., 1995, 379; CrossRef (d) R. Paetzold and U. Lindner, Z. Anorg. Chem., 1967, 350, 295; (e) V. Horn and R. Paetzold, Z. Anorg. Chem., 1973, 398, 186; (f) H. J. Reich, J. Am. Chem. Soc., 1973, 95, 964; CrossRef (g) J. P. Marino and J. R. D. Larsen, J. Am. Chem. Soc., 1981, 103, 4642; CrossRef (h) W. Nakanishi, Y. Ikeda, and H. Iwamura, J. Org. Chem., 1982, 47, 2275; CrossRef (i) T. Kawashima, F. Ohno, and R. Okazaki, J. Am. Chem. Soc., 1993, 115, 10434. CrossRef
12. J. Bergman, L. Engman, and J. Sidén, ‘The Chemistry of Organic Selenium and Tellurium Compounds,’ Vol. 1, John Wiley & Sons Ltd., New York, 1986, pp. 517-558.
13. (a) S. Sato and N. Furukawa, Chem. Lett., 1994, 889; CrossRef (b) S. Sato, N. Kondo, and N. Furukawa, Organometallics, 1986, 13, 3393. CrossRef
14. S. Watanabe, I. Nakayama, and T. Kataoka, Eur. J. Org. Chem., 2005, 1493. CrossRef
15. J. Mann, ‘Comprehensive Organic Synthesis,’ Vol. 3, ed. by B. M. Trost and I. Fleming, Pergamon, Oxford, 1991, p. 839. CrossRef
16. S. Watanabe, Y. Miura, T. Iwamura, H. Nagasawa, and T. Kataoka, Tetrahedron Lett., 2007, 48, 813. CrossRef
17. H. F. Schuster and G. M. Coppola, ‘Allenes in Organic Synthesis,’ Wiley, New York, 1984.
18. S. Watanabe, S. Asaka, and T. Kataoka, Tetrahedron Lett., 2004, 45, 7459. CrossRef
19. (a) N. N. Magdesieva, R. A. Kyandzhetsian, and A. A. Ibragimov, J. Organomet. Chem., 1972, 42, 399; CrossRef (b) K. H. Wei, I. C. Paul, M. M. Y. Chang, and J. I. Musher, J. Am. Chem. Soc., 1974, 96, 4099; CrossRef (c) S. Tamagaki, R. Akatsuka, and S. Kozuka, Bull. Chem. Soc. Jpn., 1980, 53, 817; CrossRef (d) T. Kataoka, K. Tomimatsu, H. Shimizu, and M. Hori, Tetrahedron Lett., 1983, 24, 75; CrossRef (e) V. V. Semenov, L. G. Mel’nikova, S. A. Shevelev, and A. A. Fainzil’berg, Izv. Akad. Nauk SSSR, Ser. Khim., 1980, 138.
20. W. W. Lotz and J. Gosselck, Tetrahedron, 1973, 29, 917. CrossRef
21. T. Hashimoto, H. Kitano, and K. Fukui, Nippon Kagaku Zasshi, 1968, 89, 784.
22. (a) T. Kataoka, Y. Banno, S. Watanabe, T. Iwamura, and H. Shimizu, Tetrahedron Lett., 1997, 38, 1809; CrossRef (b) S. Watanabe, K. Yamamoto, Y. Itagaki, and T. Kataoka, J. Chem. Soc., Perkin Trans. 1, 1999, 2053; CrossRef (c) S. Watanabe, E. Mori, H. Nagai, and T. Kataoka, Synlett, 2000, 49; CrossRef (d) S. Watanabe, E. Mori, H. Nagai, T. Iwamura, T. Iwama, and T. Kataoka, J. Org. Chem., 2000, 65, 8893; CrossRef (e) S. Watanabe, K. Yamamoto, Y. Itagaki, T. Iwamura, T. Iwama, T. Kataoka, G. Tanabe, and O. Muraoka, J. Chem. Soc., Perkin Trans. 1, 2001, 239. CrossRef
23. (a) M. R. Detty, J. Org. Chem., 1980, 45, 274; CrossRef (b) B. Harirchian and P. D. Magnus, J. Chem. Soc., Chem. Commun., 1977, 522. CrossRef
24. (a) Y. Matano, M. Yoshimune, and H. Suzuki, J. Org. Chem., 1995, 60, 4663; CrossRef (b) D. Wang, L. Dai, and X. Hou, Chem. Commun., 1997, 1231; CrossRef (c) M. Ochiai and Y. Kitagawa, Tetrahedron Lett., 1998, 39, 5569; CrossRef (d) M. Ochiai and Y. Kitagawa, J. Org. Chem., 1999, 64, 3181; CrossRef (e) V. K. Aggarwal, E. Alonso, G. Fang, M. Ferrara, G. Hynd, and M. Porcelloni, Angew. Chem. Int. Ed., 2001, 40, 1433; CrossRef (f) V. K.Aggarwal, M. Ferrara, C. J. O’Brien, A. Thompson, and R. V. H. Jones, J. Chem. Soc., Perkin Trans. 1, 2001, 1635. CrossRef
25. N. S. Simpkins, Tetrahedron, 1990, 46, 6951. CrossRef
26. (a) G. E. Keck, J. H. Byers, and A. M. Tafesh, J. Org. Chem., 1988, 53, 1127; CrossRef (b) M. P. Gerard and N. S. Simpkins, J. Chem. Soc., Chem. Commun., 1987, 207.
27. (a) Z. Rappoport, Acc. Chem. Res., 1981, 14, 7; CrossRef b) M. Ochiai, K. Kawashima, Y. Masaki, Y. Kunishima, and S. Tani, Tetrahedron Lett., 1993, 34, 4820. CrossRef
28. S. Oae and Y. Uchida, Acc. Chem. Res., 1991, 24, 202. CrossRef
29. H. Hansen, S. R. Jensen, and J. Munch-Petersen, Acta Chem. Scand., 1972, 26, 1190. CrossRef
30. (a) D. S. Breslow and H. Skolnik, ‘Multi-Sulfur and Sulfur and Oxygen Five- and Six-Membered Heterocycles Part Two,’ Wiley, New York, 1966; (b) C. Song, F. Ailong, H. Ziming, and Z. Zheng, Synth. Commun., 1996, 26, 3979; CrossRef (c) S. Cabiddu, C. Floris, S. Melis, and F. Sotgiu, J. Heterocycl. Chem., 1986, 23, 1815; CrossRef (d) A. C. Gelebe and P. T. Kaye, Synth. Commun., 1996, 26, 4459; CrossRef (e) S. Cabiddu, S. Melis, and F. Sotgiu, Phosphorus Sulfur, 1983, 14, 151; CrossRef (f) H. Sugihara, H. Mabuchi, and Y. Kawamatsu, Chem. Pharm. Bull., 1987, 35, 1919; (g) H. Ishibashi, M. Okada, A. Akiyama, K. Nomura, and M. Ikeda, J. Heterocycl. Chem., 1986, 23, 1163. CrossRef
31. H. Meier and D. Gröschl, Tetrahedron Lett., 1995, 36, 6047. CrossRef
32. T. Kataoka and H. Kinoshita, Eur. J. Org. Chem., 2005, 45. CrossRef
33. (a) T. Kataoka, T. Iwama, and S. Tsujiyama, Chem. Commun., 1998, 197; CrossRef (b) T. Kataoka, T. Iwama, S. Tsujiyama, T. Iwamura, and S. Watanabe, Tetrahedron, 1998, 54, 11813. CrossRef
34. T. Iwama, H. Kinoshita, and T. Kataoka, Tetrahedron. Lett., 1999, 40, 3741. CrossRef
35. T. Kataoka, H. Kinoshita, T. Iwama, S. Tsujiyama, T. Iwamura, S. Watanabe, O. Muraoka, and G. Tanabe, Tetrahedron, 2000, 56, 4725. CrossRef
36. Reviews for the Morita-Baylis-Hillman reaction: (a) S. E. Drewes and G. H. P. Roos, Tetrahedron, 1988, 44, 4653; CrossRef (b) D. Basavaiah, P. D. Rao, and R. S. Hyma, Tetrahedron, 1996, 52, 8001; CrossRef (c) E. Ciganek, Org. React., ed. by L. A. Paquette, Wiley, New York, 1997, 51, 201; (d) P. Langer, Angew. Chem. Int. Ed., 2000, 39, 3049; CrossRef (e) Y. Iwabuchi and S. Hatakeyama, J. Syn. Org. Chem. Jpn., 2002, 60, 2; (f) J. N. Kim and K. Y. Lee, Cur. Org. Chem., 2002, 6, 627; CrossRef (g) D. Basavaiah, A. J. Rao, and T. Satyanarayana, Chem. Rev., 2003, 103, 811. CrossRef
37. (a) D. Basavaiah, K. Muthukumaran, and B. Sreenivasulu, Synlett, 1999, 1249; CrossRef (b) D. Basavaiah, B. Sreenivasulu, R. Mallikarjuna, and K. Muthukumaran, Synth. Commun., 2001, 31, 2987; CrossRef (c) T. Bauer and J. Tarasiuk, Tetrahedron: Asymmetry, 2001, 12, 1741. CrossRef
38. (a) T. Kataoka, T. Iwama, H. Kinoshita, S. Tsujiyama, Y. Tsurukami, T. Iwamura, and S. Watanabe, Synlett, 1999, 197; CrossRef (b) T. Kataoka, T. Iwama, H. Kinoshita, Y. Tsurukami, S. Tsujiyama, M. Fujita, E. Honda, T. Iwamura, and S. Watanabe, J. Organomet. Chem., 2000, 611, 455. CrossRef
39. (a) T. Kataoka, H. Kinoshita, S. Kinoshita, T. Iwamura, and S. Watanabe, Angew. Chem. Int. Ed., 2000, 39, 2358; CrossRef (b) S. Kinoshita, H. Kinoshita, T. Iwamura, S. Watanabe, and T. Kataoka, Chem. Eur. J., 2003, 9, 1496. CrossRef
40. (a) G. Li, H.-X. Wei, and T. D. Caputo, Tetrahedron Lett., 2000, 41, 1; CrossRef (b) G. Li, J. Gao, H.-X. Wei, and M. Enright, Org. Lett., 2000, 2, 617; CrossRef (c) H.-X. Wei, S. H. Kim, T. D. Caputo, D. W. Purkiss, G. Li, Tetrahedron, 2000, 56, 2397; CrossRef (d) W. Pei, H.-X. Wei, and G. Li, Chem. Commun., 2002, 1856; CrossRef (e) W. Pei, H.-X. Wei, and G. Li, Chem. Commun., 2002, 2412; CrossRef (f) S. Karur, J. Hardin, A. Headley, and G. Li, Tetrahedron Lett., 2003, 44, 2991. CrossRef
41. (a) S. Uehira, Z. Han, H. Shinokubo, and K. Oshima, Org. Lett., 1999, 1, 1383; CrossRef (b) M. Shi and Y.-S. Feng, J. Org. Chem., 2001, 66, 406; CrossRef (c) Z. Han, S. Uehira, H. Shinokubo, and K. Oshima, J. Org. Chem., 2001, 66, 7854; CrossRef (d) K. Yagi, T. Turitani, H. Shinokubo, and K. Oshima, Org. Lett., 2002, 4, 3111. CrossRef
42. Shi and co–workers reported reactions conducted according to the methods of Li, Oshima and us, but their papers contain several mistakes, and the same data are repeatedly reported. Therefore, their papers and errata should be read carefully before references to them are made. (a) M. Shi and J.-K. Jiang, Tetrahedron, 2000, 56, 4793; CrossRef (b) M. Shi, J.-K. Jiang, and Y.-S. Feng, Org. Lett., 2000, 2, 2397; CrossRef (c) M. Shi, J.-K. Jiang, S.-C. Cui, and Y.-S. Feng, J. Chem. Soc., Perkin Trans. 1, 2001, 390; CrossRef (d) M. Shi, J.-K. Jiang, and S.-C. Cui, Tetrahedron, 2001, 57, 7343; CrossRef (e) M. Shi, J.-K. Jiang, and S.-C. Cui, Molecules, 2001, 6, 852; CrossRef (f) M. Shi and Y.-S. Feng, J. Org. Chem., 2001, 66, 406; CrossRef (g) M. Shi and S.-C. Cui, Chinese J. Chem., 2002, 20, 277. CrossRef
43. (a) T. Shono, Y. Matsumura, S. Kashimura, and K. Hatanaka, J. Am. Chem. Soc., 1979, 101, 4752; CrossRef (b) A. Itoh, S. Ozawa, K. Oshima, and H. Nozaki, Bull. Chem. Soc. Jpn., 1981, 54, 274; CrossRef (c) T. Yura, N. Iwasawa, and T. Mukaiyama, Chem. Lett., 1986, 187; CrossRef (d) C. Zhang and X. Lu, Synthesis, 1996, 586; CrossRef (e) G.-H. Deng, H. Hu, H.-X.Wei, and P. W. Pare, Helv. Chim. Acta, 2003, 86, 3510; CrossRef (f) X. Zheng, X. Xu, and Y. Zhang, Synlett, 2003, 2062. CrossRef
44. J. You, J. Xu, and J. G. Verkade, Angew. Chem. Int. Ed., 2003, 42, 5054. CrossRef
45. (a) Y. Iwabuchi, M. Nakatani, N. Yokoyama, and S. Hatakeyama, J. Am. Chem. Soc., 1999, 121, 10219; CrossRef (b) N. T. McDougal and S. E. Schaius, J. Am. Chem. Soc., 2003, 125, 12094; CrossRef (c) S. Kawahara, A. Nakano, T. Esumi, Y. Iwabuchi, and S. Hatakeyama, Org. Lett., 2003, 5, 3103. CrossRef
46. T. Bauer and J. Tarasiuk, Tetrahedron: Asymmetry, 2001, 12, 1741. CrossRef
47. L. M. Walsh, C. L. Winn, and J. M. Goodman, Tetrahedron Lett., 2002, 43, 8219. CrossRef
48. (a) R. Pathak, A. K. Shaw, and A. P. Bhaduri, Tetrahedron, 2002, 58, 3535; CrossRef (b) R. Pathak, C. S. Pant, A. K. Shaw, A. P. Bhaduri, A. N. Gaikwad, S. Sinha, A. Srivastava, K. K. Srivastava, V. Chaturvedi, R. Srivastava, and B. S. Srivastava, Bioorg. Med. Chem., 2002, 10, 3187. CrossRef
49. (a) V. G. Nenajdenko, M. V. Lebedev, and E. S. Balenkova, Synlett, 1995, 1133; CrossRef (b) V. G. Nenajdenko, M. V. Lebedev, and E. S. Balenkova, Tetrahedron Lett., 1995, 36, 6317; CrossRef (c) M. V. Lebedev, V. G. Nenajdenko, and E. S. Balenkova, Synthesis, 2001, 2124. CrossRef
50. (a) T. Kataoka, S. Kinoshita, H. Kinoshita, M. Fujita, T. Iwamura, and S. Watanabe, Chem. Commun., 2001, 1958; CrossRef (b) H. Kinoshita, S. Kinoshita, Y. Munechika, T. Iwamura, S. Watanabe, and T. Kataoka, Eur. J. Org. Chem., 2003, 4852. CrossRef
51. (a) J. S. Hill and N. S. Isaacs, Tetrahedron Lett., 1986, 27, 5007; CrossRef (b) J. S. Hill and N. S. Isaacs, J. Chem. Res. (S), 1988, 330; (c) J. S. Hill and N. S. Isaacs, J. Chem. Res. (M), 1988, 2641.
52. D. Basavaiah, T. Bharathi, and V. V. L. Gowriswari, Tetrahedron Lett., 1987, 28, 4351. CrossRef
53. B. M. Kim, S. F. Williams, and S. Masamune, ‘Comprehensive Organic Synthesis’, ed. by B. M. Trost, I. Fleming, and C. H. Heathcock, Pergamon Press, Oxford, 1991, Vol. 2, p. 240.
54. H. Garcia, S. Iborra, J. Primo, and M. A. Miranda, J. Org. Chem., 1986, 51, 4432. CrossRef
55. H. Sashida, Synthesis, 1998, 745. CrossRef
56. T. Kataoka, H. Kinoshita, S. Kinoshita, and T. Iwamura, Tetrahedron Lett., 2002, 43, 7039. CrossRef
57. A. M. B. S. R. C. S. Costa, F. M. Dean, M. A. Jones, and R. S. Varma, J. Chem. Soc., Perkin Trans. 1, 1985, 799. CrossRef
58. M. R. Detty and L. W. McGarry, J. Org. Chem., 1988, 53, 1203. CrossRef
59. D. Basavaiah and A. J. Rao, Tetrahedron Lett., 2003, 44, 436. CrossRef
60. T. Mukaiyama and M. Murakami, Synthesis, 1987, 1043. CrossRef
61. (a) M. Suzuki, T. Kawahigashi, and R. Noyori, Tetrahedron Lett., 1981, 22, 1809; CrossRef (b) R. Noyori, S. Murata, and M. Suzuki, Tetrahedron, 1981, 37, 3899. CrossRef
62. (a) S. Kim, Y. G. Kim, and J. H. Park, Tetrahedron Lett., 1991, 32, 2043; CrossRef (b) S. Kim, J. H. Park, Y. G. Kim, and J. H. Lee, Chem. Commun., 1993, 1188. CrossRef
63. F. Wang and R. Zibuck, Synlett, 1998, 245. CrossRef
64. M. Hojo, M. Nagayoshi, A. Fujii, T. Yanagi, N. Ishibashi, K. Miura, and A. Hosomi, Chem. Lett., 1994, 719. CrossRef
65. H. Kinoshita, T. Osamura, S. Kinoshita, T. Iwamura, S. Watanabe, T. Kataoka, O. Muraoka, and G. Tanabe, J. Org. Chem., 2003, 68, 7532. CrossRef
66. (a) T. H. Newby and E. C. Howlett, J. Am. Chem. Soc., 1951, 73, 4720; CrossRef (b) P. Ulrich and A. Cerami, J. Med. Chem., 1982, 25, 654; CrossRef (c) M. Dzurilla, P. Kutschy, and P. Kristan, Synthesis, 1985, 933; CrossRef (d) H. Takahata, Y. Banba, M. Mozumi, and T. Yamazaki, Heterocycles, 1986, 24, 947; CrossRef (e) K. Toda, K. Tanaka, and J. Sato, Tetrahedron: Asymmetry, 1993, 4, 1771; CrossRef (f) A. Hari and B. L. Miller, Org. Lett., 2000, 2, 3667. CrossRef
67. (a) C. Palomo, M. Oiarbide, F. Dias, A. Ortiz, and A. Linden, J. Am. Chem. Soc., 2001, 123, 5602; CrossRef (b) A. Ortiz, L. Quintero, H. Hernández, S. Maldonado, G. Mendoza, and S. Bernès, Tetrahedron Lett., 2003, 44, 1129; CrossRef (c) C. Palomo, M. Oiarbide, F. Dias, R. López, and A. Linden, Angew. Chem., 2004, 116, 3369; CrossRef (d) C. Palomo, M. Oiarbide, R. López, P. B. González, E. Gómez-Bengoa, J. M. Saá, and A. Linden, J. Am. Chem. Soc., 2006, 128, 15236. CrossRef
68. (a) T. Kataoka, H. Kinoshita, S. Kinoshita, T. Osamura, T. Iwamura, S. Watanabe, O. Muraoka, and G. Tanabe, Angew. Chem. Int. Ed., 2003, 42, 2889; CrossRef (b) H. Kinoshita, T. Osamura, K. Mizuno, S. Kinoshita, T. Iwamura, S. Watanabe, T. Kataoka, O. Muraoka, and G. Tanabe, Chem. Eur. J., 2006, 12, 3896. CrossRef
69. H. Kinoshita, N. Takahashi, T. Iwamura, S. Watanabe, T. Kataoka, O. Muraoka, and G. Tanabe, Tetrahedron Lett., 2005, 46, 7155. CrossRef