Novel chemoselective and diastereoselective iron(III)-catalysed Michael reactions of 1,3-dicarbonyl compounds and enones

摘要

J. Chem. Soc. Perkin Trans. 1 1997 3141 Novel chemoselective and diastereoselective iron(III)-catalysed Michael reactions of 1,3-dicarbonyl compounds and enones Jens Christoffersdagger; Technische Universitauml;t Berlin Institut fuuml;r Organische Chemie Sekretariat C3 Straszlig;e des 17.Juni 135 D-10623 Berlin Germany Iron(III) chloride hexahydrate catalyses the Michael reaction of 1,3-dicarbonyl compounds with middot;,lsquor;-unsaturated ketones under mild and neutral conditions with extraordinary efficiency. The chemoselectivity of this FeIII-catalysed process is superior to that of the classic base-catalysed Michael reaction since the latter suffers from various side reactions namely drawbacks such as aldol cyclisations and ester solvolysis. Excellent yields and chemoselectivities together with the environmentally friendly nature of FeIII catalysis makes this an important alternative to classic base catalysis.Moreover the reaction procedure is reasonably easy FeIII catalysis does not require inert or anhydrous conditions and in most cases no solvent is needed. In terms of diastereoselectivity the FeIII-mediated reaction may also prove superior to a base-catalysed one. In at least one example FeIII catalysis forms a diastereoisomer as the major kinetic product which is disfavoured in the base-mediated Michael reaction where a thermodynamic mixture is obtained. The relative configuration of the diastereoisomeric Michael products has been determined for two examples by synthesis and structure elucidation of the cyclic aldol derivatives. Introduction The Michael reaction of the 1,3-dicarbonyl compounds 1 and the enones 2 is classically a high yielding base-mediated process 1 and can even be performed with high stereoselectivity.2 However there are some disadvantages with base catalysis including side reactions of the starting materials and subsequent reactions of the Michael product 3.Incompatibilities with base-sensitive groups ester solvolysis and aldol processes leading to cyclic products or retro-aldol type decompositions can significantly decrease yields of the base-mediated Michael reactions in some cases. Alternatively it has been reported that the Michael reaction can be catalysed by lanthanide 3 or transition metal 4 compounds e.g. group VIII 1,3-dionato complexes 5 although not always with satisfactory efficiency. Recently we reported that iron(III) chloride hexahydrate is an extraordinarily efficient catalyst for the Michael reaction of 1,3-dicarbonyl compounds and enones,6 a hitherto unknown fact,Dagger; despite the known tendency of FeIII to form 1,3-dionato complexes.8 This ability together with ecological and economical considerations make iron the transition metal of choice in such work.FeIII catalyses the Michael reaction under mild and neutral conditions and thus the chemoselectivity of the FeIII catalysis is superior to that of the classic base-mediated process since both side-reactions and subsequent reactions under basic conditions are avoided. Moreover the reaction conditions for FeIII catalysis are reasonably easy no inert or anhydrous conditions are required and in some cases even solvents are unnecessary. This high efficiency together with excellent yields make the FeIII catalysis of the Michael reaction an important alternative to the classic base catalysis.Results and discussion Iron(III) catalysis of the Michael reaction We have been investigating the catalytic activity of several tran- dagger; E-Mail jchr@wap0105.chem.tu-berlin.de Dagger; In one case the application of a combination of Ni(acac)2 and FeCl3 was reported but the role of FeIII was ascribed to its Lewis acid character (activation of the enone).7 sition metal compounds in the Michael reaction of 1a with 2a to give 3a9 (Scheme 1). Although a number of the investigated systems showed activity only with FeCl3?6 H2O was there fast clean and complete conversion at room temperature. With 5 mol the reaction was quantitative within 1 h and with 1 mol within 3 h.Results obtained with another FeIII catalyst 4c and some NiII compounds are listed in Table 1. Using Fe(acac)3 which is not active itself but needed further Lewis acid activation rapid consumption of starting materials was observed too but the conversion was not clean 4a (Table 2) was formed as a by-product via a hetero-Dielsndash;Alder dimerisation of 2a.10 Of the NiII compounds only Ni(acac)2 5 gave full conversion but a high temperature was required. All other transition-metal compounds investigated in our studies are less efficient than the NiII compounds reported in Table 1. FeIII catalysis is generally very efficient in the conversion of various Michael donors 1 (see Table 2) with methyl vinyl ketone 2a; a list of products 3 prepared is shown in Table 3. Best results were obtained with the cyclic keto esters to give 3a 3b,6 3d11 and 3e.6 With only 1 mol FeCl3?6 H2O full conversion is achieved within a few hours at room temperature even if bulkier ester functions are present (3b).Generally no ester solvolysis side-reactions occur with ethyl or higher alkyl esters. Methyl esters like 1e are partially solvolysed by the hydrate water of the catalyst,sect; so that the product yield drops to 72 if 5 mol FeCl3?6 H2O is used; with only 1 mol of catalyst this side reaction is negligible (91 isolated yield). Reactions of acyclic keto esters to give 3f,12 3g 13 and 3j 14 as well as of b-diketones to give 3c,15 3h,16 3i 17 and 3k18 proceed a Scheme 1 O OEt O O Me O CO2Et Me O 3a 2a 1a + cat. sect; Decomposition product cycloheptanone was identified in the reaction mixture by GCMS.3142 J. Chem. Soc. Perkin Trans. 1 1997 little slower but use of 5 mol of catalyst results in full conversion within a few hours at room temperature and gives satisfactory product yields. It should be emphasised that since the starting materials and products in Table 3 are liquid at room temperature no solvent is necessary for the transformations (except for 3i which is solid at room temperature). Moreover as long as reactions are quantitative with no by-products formation the work-up procedure is reasonably simple filtration using a small column of silica gel removes all iron-containing materials. In addition since water is tolerated no inert or anhydrous conditions are required and reactions are carried out simply by mixing starting materials and the catalyst.para; Fast and quantitative conversions together with a straightforward work-up procedure makes the iron(III) chloride hexahydrate-catalysed Michael reaction a very efficient alternative to the classic base-mediated methodology.Chemoselectivity In Table 4 products 3lndash;s resulting from Michael reactions of various keto esters 1 (Table 2) with substituted enones 2bndash;d (Table 2) are listed. Generally these transformations need solvent because the products are either solid or viscous oils at room temperature and higher reaction temperatures (up to Table 1 Comparison of NiII and FeIII catalysis of the Michael reaction Catalyst a Conversion () b FeCl3?6 H2O Fe(acac)3 1 BF3?OEt2 NiCl2?6 H2O Ni(OAc)2?4 H2O Ni(acac)2 1 h RTc 100 40 mdash; mdash; mdash; 3 h RT mdash; 90d 5 21 19 24 h RT mdash; mdash; 41 68 61 3 h 50 8C mdash; mdash; 81 52 100 a Conditions 1a (1 equiv.) 1 2a (1.1 equiv.) 1 catalyst.(0.05 equiv.) no solvent. b By 1H NMR. c Room temp. d By-product 4a was formed. Table 2 List of starting materials 1 and 2 and by-products 4 O X O X = OEt X = OBui X = Me 1a 1b 1c Me O 2a O OEt O 1d Me O Ph 2b O OMe O 1e Ph O Ph 2c R O OEt O R = Me R = Ph 1f 1g Me O Ph 2d R O R O R = Me R = Ph 1h 1i O Me Me O 4a Me O X O Me X = OEt X = Me 1j 1k O Me Ph Ph O Me 4b para; Scaling up (more than 20 mmol) requires cooling of the mixture to prevent 2a from being evolved since the reactions are slightly exothermic. 50 8C) are required in some cases (see Experimental section for details). The iron(III)-catalysed conversion of the keto ester 1a with the enone 2d to the Michael product 3n provides a typical example of the chemoselectivity achieved with this method (Scheme 2).Base catalysis namely 5 mol KOEt in absolute EtOH failed to give any of the desired product 3n the enone dimer 4b19 being obtained instead from a hetero Dielsndash;Alder reaction of two equivalents of 2d. A similar result was obtained in the base-mediated conversion of Michael donors 1f and 1g with the enone 2d. In these cases no Michael products 3q20 or 3s 20 were detected in the reaction mixtures only the dimer 4b being isolated. Thus the FeIII-catalysed reaction of the keto esters 1 with the enone 2d seems to be the only way to perform a Michael reaction and indeed products 3n 3q and 3s were isolated in good to moderate yields although the dimer 4b was also found in these reactions as a by-product (see Experimental section). Obviously with FeIII catalysis the Michael reaction of 2d becomes fast enough to compete seriously with the Dielsndash; Alder dimerisation side-reaction.The Michael adduct 3o20 was formed by FeIII-catalysed conversion of 1g with the acceptor 2b in good yield (Scheme 3). In contrast under conditions of base catalysis the primary product 3o cyclises in a subsequent aldol reaction to give 5a,21 a byproduct which is not detectable in an FeIII catalysed reaction. Thus the optimum yield achievable in the base-mediated Scheme 2 O Me Ph Ph O Me O Ph Me O EtO2C Fe(III) base 4b 91 3n 76 1a + 2d Table 3 Michael reactions of 1 with 2a to give 3andash;k Product O COX Me O O CO2Et Me O O Me O CO2Me R O CO2Et Me O R O COR Me O Me O Me O Me COX X = OEt X = OBui X = Me R = Me R = Ph R = Me R = Ph X = OEt X = Me 3a 3b 3c 3d 3e 3f 3g 3h 3i 3j 3k FeIII (mol) a 1 1 5 1 1 5 5 5 5 5 5 Yield () b 97 95 86 94 91 c 90 87 77 100 d 78 84 a Conditions 1 (1 equiv.) 1 2a (1.1 equiv.) 1 catalyst FeCl3?6 H2O no solvent room temp.b Isolated yields. c Use of 5 mol catalyst resulted in 72 yield. d Solvent (CHCl3) was applied. J. Chem. Soc. Perkin Trans. 1 1997 3143 process is 48 and much lower than in a FeIII-catalysed reaction in fact if the base-catalysed reaction were to run for infinite time the yield of 3o would drop to zero all the material by then having been consumed by the aldol process. A more extreme example is compound 3r 22 which whilst readily accessible by FeIII catalysis was not formed at all by base catalysis since subsequent aldol reactions led to isomeric mixtures of unseparated and uncharacterised cyclic products.In summary we have outlined how FeIII catalysis of the Michael reaction can provide access to compounds which cannot be prepared by classic base catalysis as a result of sidereactions or subsequent reactions taking place with the latter. Scheme 3 O Ph CO2Et HO Ph Ph O CO2Et Me O Ph 5a 3o 0 12 85 A / B = 84 16 48 A / B = 39 61 FeIII base + 1g + 2b Table 4 FeIII and base-catalysed Michael reactions of 1 with 2bndash;d to give 3lndash;s compared FeIII a Base b Product O CO2Et Me O Ph O CO2Et Ph O Ph O Ph Me O EtO2C Ph O CO2Et Me O Ph Ph O CO2Et Ph O Ph Ph O EtO2C Me O Ph Me O CO2Et Me O Ph Me O EtO2C Me O Ph 3l 3m 3n 3o 3p 3q 3r 3s yield () 90 93 76 d 84 85 74 d 76 46 d A/Bc 20:80 49:51 82:18 16:84 78:22 57:43g 57:43g 55:45g yield () 93 100 0 e 48 f 100 0 e 0 h 0 e A/Bc 31:69 46:54 mdash; 61:39 90:10 mdash; mdash; mdash; a Conditions 1 (1.0 equiv.) 1 2 (1.0 equiv.) 1 FeCl3?6 H2O (0.01ndash;0.05 equiv.) solvent CHCl3 stirring overnight.b Conditions 1 (1.0 equiv.) 1 2 (1.0 equiv.) 1 KOH (0.05 equiv.) solvent abs. EtOH stirring overnight. c In all cases the diastereoisomer with the ester-CH3 triplet in the 1H NMR spectrum at higher field was assigned as isomer A the one with the triplet at lower field is isomer B. d By-product 4b was formed. e Product 4b only. f By-product 5a was formed. g Isomers in equilibrium with each other. h Isomeric mixtures of cyclic aldol products were obtained. Consequently FeIII has to be considered as an alternative in the synthesis of novel Michael reaction products particularly if base catalysis fails to give satisfactory results or fails altogether.There are examples of course in which FeIII and base-mediated Michael reactions are both efficient e.g. to prepare 3l,23 3m or 3p.24 Diastereoselectivity If Michael donors 1 are allowed to react with the enones 2bndash;d which bear a substituted carbonndash;carbon double bond each product consists of two diastereoisomers A and B. In all the examples listed in Table 4 formation of these two diastereoisomers is observed and their ratio A/Bverbar;verbar; has been determined for FeIII as well as for base catalysis (the latter provided that a Michael product was formed at all). For the products in Table 4 the diastereoisomers were separated by chromatography and each completely characterised except for 3q 3r and 3s. In these cases products contain a b-keto ester moiety and the two diastereoisomers were in equilibrium via the corresponding enol form which was actually detectable in the 1H NMR spectra of 3q 3r and 3s (ca.0.5 mol; integral of the OH resonance at ca. d 13 ppm). Thus isomers were not separable. In contrast isomers of 3l 3m and 3n were separable because of a quaternary carbon centre; however with regard to diastereoselectivity the results of FeIII and base catalysis were not significantly different. Interestingly compounds 3o and 3p which also bear a b-keto ester moiety with a formal acidified CH bond failed to equilibrate and in both cases the two diastereoisomers obtained with FeIII catalysis could be separated by chromatography. They failed to interconvert both under neutral conditions even at temperatures up to 150 8C and also under the reaction conditions of the FeIII catalysis.Nevertheless an equilibrium mixture of 3o was obtained with 1 equiv. of NEt3 in CDCl3 within 2 h and with 5 equiv. of CF3CO2H in CDCl3 within 10 d. In both cases the thermodynamic equilibrium was found to be A/B = 60/40 a similar result to that obtained with the basecatalysed formation of 3o (A/B = 61/39). FeIII catalysis led to a kinetic mixture with A/B = 16/84; contrastingly FeIII and basemediated formation of 3p gave roughly the same product ratio. In summary FeIII catalysis of the Michael reaction introduces the opportunity to obtain kinetic product mixtures in cases in which base catalysis yields thermodynamic ratios. As shown for 3o these kinetic and thermodynamic mixtures can differ significantly. The inability of the isomers of 3o and 3p to interconvert seems to be closely linked to the 1,3-diphenyl constitution of the products thus if the phenyl groups are in a 1,4 relationship to each other lsquo;3qrsquo; or one is missing lsquo;3rrsquo; the corresponding species are in equilibrium again.Presumably the enol form which is the intermediate species in an equilibrium of the two diastereoisomers cannot be stabilised by H-bonding because of allylic 1,3-strain,25 and thus is not formed. Relative configuration The formation of 5a in the base-catalysed conversion of 1g and 2b (as a by-product at room temp. and main product at elevated temp.) allowed the assignment of the relative configuration of the diastereoisomers A and B of 3o. In independent experiments starting from pure isomers of 3o it was shown that base catalysis (either NEt3 or KOH in absolute EtOH) converts isomer B directly into the cyclic product 5a whilst isomer A reacted relatively slower andmdash;after equilibration with Bmdash;gives 5a as well (Scheme 4).Consequently the relative configuration of the stereocentres in isomer B should be the same as in 5a in verbar;verbar; In all cases the diastereoisomer with the ester-CH3 triplet in the 1H NMR spectrum at higher field is assigned to be isomer A the one with the triplet at lower field is isomer B. 3144 J. Chem. Soc. Perkin Trans. 1 1997 which the ethoxycarbonyl and the phenyl group have been reported to be both equatorial and trans to each other; in fact the 3J16 H,H-coupling constant in 5a was found to be 2.8 Hz. Thus isomer B of 3o the major product with FeIII is the (R*,S*)-isomer and isomer A which is the thermodynamic product is assigned to be (R*,R*).Analogously the same relative configuration applies for the two isomers of 3p since their 1H NMR spectra show ABMX patterns which are nearly identical with those of 3o. In the same manner isomer B of 3l cyclises under basic reaction conditions to give the bicyclic derivative 5b (Scheme 5) the assignment of structure to which was made on the basis of H,H-COSY and NOE experiments. In a similar way to 5a the ethoxycarbonyl and phenyl groups are both equatorial (for the chair-conformation of the six-membered ring as shown in Scheme 5) and trans to each other. Also the phenyl and methyl groups are trans and the hydroxy function is in an equatorial position. Isomer A of 3l failed to give any aldol product under either the same or even more drastic reaction conditions.A cyclisation product analogous to 5b would bear the phenyl group in an axial position and cis to either an hydroxy or a methyl group a situation which is impossible because of 1,3- strain. Consequently the relative configuration of isomer B is (R*,R*) and that of isomer A must be (R*,S*). By comparison of the 1H NMR data relative configurations of 3m can be assigned analogously. Conclusion Iron(III) catalysis is proposed as a highly efficient alternative to base catalysis of the Michael reaction for 1,3-dicarbonyl compounds and enones. Conversion is generally fast and clean performance and work-up easy and in some cases solvents are unneeded; further FeCl3?6 H2O is both an ecologically friendly compound and relatively cheap. In terms of chemoselectivity iron(III) catalysis can yield Michael reaction products even in cases where base catalysis gives poor results or fails altogether.Scheme 4 O Ph Me O CO2Et Ph O Ph Me O CO2Et Ph O EtO2C Ph HO Ph O EtO2C Ph Ph HO 5a Isomer A of 3o (R*,R*) Isomer B of 3o (R*,S*) EtOK / EtOH Scheme 5 Me O Ph O CO2Et Me O Ph O CO2Et EtO2C O OH Me H Ph EtO2C O OH Me Ph H 5b Isomer A of 3l (R*,S*) Isomer B of 3l (R*,R*) EtOK / EtOH As to diastereoselectivity iron(III) catalysis can produce kinetic mixtures of diastereoisomers even in cases where base catalysis leads to a thermodynamic equilibrium mixture. Experimental Column chromatography was accomplished with Merck silica gel (Type 60 0.063ndash;0.200 mm) using methyl tert-butyl ether (MTB). 1H NMR Bruker AM 400 (400 MHz) 25 8C TMS; structure elucidation was made using H,H-COSY and NOE experiments.13C NMR Bruker AC 200 (50 MHz) 25 8C TMS assignments were made using DEPT experiments. J values given in Hz. MS Varian MAT 711 and MAT 955Q (high resolution). IR Nicolet Magna IR 750. GC and GCMS HP 5890 II with FID resp. HP MSD 5971A. Elemental analyses Analytik Jena Vario EL. All starting materials were either commercially available or were prepared according to literature procedures (1b,26 1j,27 1k,28 2d 29). Ethyl 2-oxo-1-(3-oxobutyl)cyclopentanecarboxylate 3a A mixture of the oxo ester 1a (875 mg 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (15 mg 0.055 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.41) to afford 3a as a colourless oil (1.23 g 5.44 mmol 97); nmax/cm21 (ATR) 2976m 1748vs 1717vs 1448m 1406m 1367m 1318m 1260s 1232s 1165s 1116m 1029m and 861m; dH(400 MHz CDCl3) 1.23 (t J 7.2 3 H CH3) 1.82ndash;2.03 (m 4 H CH2) 2.03ndash;2.13 (m 1 H CHH) 2.12 (s 3 H CH3) 2.24ndash;2.49 (m 4 H CH2) 2.69 (ddd J 18 9.6 6.0 1 H CHH) and 4.14 (q J 7.1 2 H OCH2) 13C{1H} NMR (50 MHz CDCl3) d 13.29 (CH3) 18.84 (CH2) 26.24 (CH2) 29.00 (CH3) 33.22 (CH2) 37.07 (CH2) 38.01 (CH2) 58.23 (C) 60.23 (OCH2) 170.47 (C O) 206.61 (C O) and 213.75 (C O); m/z (EI 70 eV) 226 (11) M1 208 (10) M1 2 H2O 198 (86) M1 2 CO 169 (17) M1 2 Me(CO)- CH2 156 (49) M1 2 Me(CO)CH CH2 152 (19) M1 2 EtOCHO 141 (18) M1 2 Me(CO)CH CH2 2 Me 137 (50) M1 2 EtOCHO 2 Me 125 (100) M1 2 EtOCHO 2 CH2 CH 110 (54) M1 2 EtOCO 2 MeCO 55 (29) CH2COMe1 and 43 (86) COMe1 Found C 63.68; H 8.09; M (HRMS) 226.1207.Calc. for C12H18O4 C 63.70; H 8.02 M 226.1205. Isobutyl 2-oxo-1-(3-oxobutyl)cyclopentanecarboxylate 3b A mixture of the oxo ester 1b (1.03 g 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (15 mg 0.055 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.45) to afford 3b as a colourless oil (1.35 g 5.32 mmol 95); nmax/cm21 (ATR) 2964s 2892m 2876m 1749vs 1718vs 1471m 1370s 1356m 1259s 1230s 1165s 1117m and 998m; dH(400 MHz CDCl3) 0.91 (d J 6.6 6 H 2 CH3) 1.84ndash;2.15 (m 6 H CH CH2) 2.13 (s 3 H CH3) 2.25ndash;2.52 (m 4 H CH2) 2.70 (ddd J 18 9.1 6.2 1 H CHH) and 3.84ndash;3.92 (m 2 H OCH2); 13C{1H} NMR (50 MHz CDCl3) d 18.72 (2 CH3) 19.37 (CH2) 26.79 (CH2) 27.47 (CH) 29.62 (CH3) 34.08 (CH2) 37.72 (CH2) 38.62 (CH2) 58.73 (C) 71.08 (OCH2) 171.13 (C O) 207.38 (C O) and 214.42 (C O); m/z (EI 70 eV) 254 (1) M1 236 (5) M1 2 H2O 226 (41) M1 2 CO and 125 (100) M1 2 Bui 2 CO 2 Me2CO Found C 66.46; H 8.78; M (HRMS) 254.1522.Calc. for C14H22O4 C 66.12; H 8.72; M 254.1518. 2-Acetyl-2-(3-oxobutyl)cyclopentanone 3c A mixture of the diketone 1c (706 mg 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (76 mg 0.28 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.42) to afford 3c as a colourless oil (946 mg 4.82 mmol 86); nmax/cm21 2966m 2890w 1735s 1702vs 1421m 1407m 1358s 1317m 1276m 1248m 1163s 1147s 1116m 929m and 815m; dH(400 MHz CDCl3) 1.62ndash;1.69 (m 1 H) 1.85ndash;1.97 (m 4 H) 2.11 (s 3 H, J.Chem. Soc. Perkin Trans. 1 1997 3145 CH3) 2.16 (s 3 H CH3) 2.28ndash;2.38 (m 4 H) and 2.05ndash;2.57 (m 1 H); 13C{1H} NMR (50 MHz CDCl3) d 19.09 (CH2) 25.84 (CH3) 27.30 (CH2) 29.65 (CH3) 31.21 (CH2) 38.08 (CH2) 38.31 (CH2) 67.11 (C) 204.33 (C O) 206.86 (C O) and 215.72 (C O); m/z (EI 70 eV) 196 (1) M1 154 (100) M1 2 CH2CO 97 (84) M1 2 CH2CO 2 MeCOCH2 and 43 (94) MeCO1 Found C 66.92; H 8.24; M (HRMS) 196.1103. Calc. for C11H16O3 C 67.32; H 8.22; M 196.1099. Ethyl 2-oxo-1-(3-oxobutyl)cyclohexanecarboxylate 3d A mixture of the oxo ester 1d (953 mg 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (15 mg 0.055 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.46) to afford 3d as a colourless oil (1.26 g 5.26 mmol 94); nmax/cm21 (ATR) 2940s 2867m 1711vs 1445s 1367s 1244s 1212s 1188s 1168s 1137m 1096m and 1020m; dH(400 MHz CDCl3) 1.24 (t J 7.2 3 H CH3) 1.38ndash;1.48 (m 1 H CHH) 1.56ndash;1.65 (m 2 H CH2) 1.68ndash;1.76 (m 1 H) 1.81 (ddd J 14 10 5.4 1 H CHH) 1.92ndash;2.01 (m 1 H CHH) 2.05 (ddd J 14 10 5.0 1 H CHH) 2.09 (s 3 H CH3) 2.33 (ddd J 18 10 5.4 1 H CHH) 2.39ndash;2.50 (m 3 H CH2) 2.55 (ddd J 18 10 5.2 1 H CHH) and 4.11ndash;4.23 (m 2 H OCH2); 13C{1H} NMR (50 MHz CDCl3) d 13.47 (CH3) 21.90 (CH2) 26.85 (CH2) 27.72 (CH2) 29.14 (CH3) 35.86 (CH2) 37.99 (CH2) 40.30 (CH2) 59.25 (C) 60.64 (OCH2) 171.22 (C O) 206.67 (C O) and 206.87 (C O); m/z (EI 70 eV) 240 (3) M1 212 (17) M1 2 CO 194 (22) M1 2 CO 2 H2O 170 (100) M1 2 CH2CH(CO)Me 151 (54) M1 2 H2O 2 CH2CH2(CO)Me and 124 (62) M1 2 H2O 2 CO 2 CH2CH(CO)Me Found C 64.64; H 8.37; M (HRMS) 240.1362.Calc. for C13H20O4 C 64.98 H 8.39; M 240.1362. Methyl 2-oxo-1-(3-oxobutyl)cycloheptanecarboxylate 3e A mixture of the oxo ester 1e (929 mg 5.46 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (15 mg 0.055 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.45) to give 3e as a colourless oil (1.20 g 4.99 mmol 91); nmax/cm21 (ATR) 2935s 2861m 1734sh 1714vs 1444s 1437s 1356m 1295m 1257m 1229s 1198s 1165s 992s 941s; dH(400 MHz CDCl3) 1.38ndash;1.55 (m 2 H CH2) 1.58ndash;1.75 (m 5 H CH2) 1.79ndash;1.92 (m 1 H CHH) 1.88 (ddd J 14 10 J 5.6 1 H CHH) 2.10 (s 3 H CH3) 2.15 (ddd J 14 10 5.6 1 H CHH) 2.39 (ddd J 18 10 5.6 1 H CHH) 2.45ndash;2.65 (m 3 H CH2) and 3.70 (s 3 H OCH3); 13C{1H} NMR (50 MHz CDCl3) d 24.49 (CH2) 24.99 (CH2) 28.82 (CH2) 29.34 (CH2) 29.38 (CH3) 33.48 (CH2) 38.57 (CH2) 41.68 (CH2) 51.69 (OCH3) 61.32 (C) 172.39 (C O) 206.99 (C O) and 208.87 (C O); m/z (EI 70 eV) 240 (3) M1 222 (4) M1 2 H2O 212 (7) M1 2 CO 150 (70) M1 2 MeOCO 2 MeO 98 (57) COCH2CH2CH2CH2CH2 1 95 (85) MeCOCH2CH2CH2CH2 1 and 43 (100) MeCO1 Found C 65.03; H 8.39; M (HRMS) 240.1367.Calc. for C13H20O4 C 64.98; H 8.39; M 240.1362. Ethyl 2-acetyl-5-oxohexanoate 3f A mixture of the oxo ester 1f (728 mg 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (76 mg 0.28 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.38) to afford 3f as a colourless oil (1.01 g 5.04 mmol 90); nmax/cm21 (ATR) 2983m 2940m 1738vs 1713vs 1445m 1419m 1359s 1244s 1217s 1151s 1097m 1023m 957m and 858m; dH(400 MHz CDCl3) 1.22 (t J 7.2 3 H CH3) 1.97ndash;2.11 (m 2 H 3-CH2) 2.09 (s 3 H CH3) 2.19 (s 3 H CH3) 2.43ndash;2.48 (m 2 H 4-CH2) 3.45 (t J 7.2 1 H 2-CH) and 4.10ndash;4.18 (m 2 H OCH2); 13C{1H} NMR (CDCl3) d 13.55 (CH3) 21.20 (CH2) 28.45 (CH3) 29.32 (CH3) 39.92 (CH2) 57.70 (CH) 60.86 (OCH2) 168.98 (C O) 202.26 (C O) and 206.88 (C O); m/z (EI 70 eV) 200 (0.5) M1 158 (24) M1 2 CH2CO 112 (17) M1 2 EtOCO 2 Me 101 (23) M1 2 Et 2 MeCOCH2CH2 84 (35) M1 2 EtOH 2 MeCOCH2CH2 and 43 (100) MeCO1 Found C 59.85; H 8.05; M (HRMS) 200.1053.Calc. for C10H16O4 C 59.98; H 8.05; M 200.1049. Ethyl 2-benzoyl-5-oxohexanoate 3g A mixture of the oxo ester 1g (1.08 g 5.60 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (76 mg 0.28 mmol) was stirred overnight at room temp.after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.46) to afford 3g as a colourless oil (1.28 g 4.87 mmol 87); nmax/cm21 (ATR) 3063w 2982m 2940m 2905w 1735vs 1715vs 1685vs 1597m 1448m 1369m 1286m 1246s 1180s 1157s 1025m and 691m; dH(400 MHz CDCl3) 1.16 (t J 7.1 3 H CH3) 2.13 (s 3 H 6-CH3) 2.15ndash;2.30 (m 2 H 3-CH2) 2.52ndash;2.67 (m 2 H 4-CH2) 4.08ndash;4.20 (m 2 H OCH2) 4.44 (dd J 6.3 7.9 1 H 2-CH) 7.47ndash;7.50 (m 2 H ArH) 7.57ndash;7.61 (m 1 H ArH) and 8.01ndash; 8.03 (m 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.62 (CH3) 22.42 (CH2) 29.56 (CH3) 40.13 (CH2) 52.30 (CH) 61.00 (OCH2) 128.32 (2 CH) 128.44 (2 CH) 133.30 (CH) 135.67 (C) 169.42 (C O) 194.89 (C O) and 207.36 (C O); m/z (EI 70 eV) 262 (0.1) M1 105 (100) PhCO1 77 (56) Ph1 and 43 (86) MeCO Found C 68.36; H 6.90; M (HRMS) 262.1149.Calc. for C15H18O4 C 68.69; H 6.92; M 262.1205. 3-Acetylhepta-2,6-dione 3h The enone 2a (2.0 ml 24.0 mmol) was added to a mixture of the diketone 1h (2.40 g 24.0 mmol) and FeCl3?6 H2O (324 mg 1.20 mmol) at 0 8C. The resulting mixture was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.34) to afford 3h as a colourless oil (3.13 g 18.4 mmol 77) (mixture of tautomers NMR data are given for the major isomer). An analytically pure sample of 3h was obtained by Kugelrohr distillation (60 8C at 0.5 mm); nmax/ cm21 (ATR) 2970w 2941w 1699vs 1569w 1420m 1358s and 1154s; dH(400 MHz CDCl3) 2.07 (q J 7.0 2 H 4-CH2) 2.12 (s 3 H 7-CH3) 2.19 (s 6 H 1,19-CH3) 2.44 (t J 7.0 2 H 5-CH2) and 3.67 (t J 7.0 1 H 3-CH); 13C{1H} NMR (50 MHz CDCl3) d 20.98 (CH2) 28.81 (2 CH3) 29.32 (CH3) 39.99 (CH2) 66.13 (CH) 203.55 (2 C O) and 206.79 (C O); m/z (EI 70 eV) 170 (1) M1 152 (9) M1 2 H2O 137 (100) M1 2 H2O 2 Me 109 (26) M1 2 H2O 2 COMe Found C 63.72; H 8.11; M (HRMS) 170.0942.Calc. for C9H14O3 C 63.51; H 8.29; M 170.0943. 2-Benzoyl-1-phenylhexane-1,5-dione 3i A solution of the dione 1i (1.22 g 5.46 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (74 mg 0.274 mmol) in CHCl3 (1.5 ml) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.50) to afford 3i as colourless crystals (1.60 g 5.44 mmol 100) mp 60 8C (recryst. from MTB); nmax/cm21 (ATR) 3062m 3003w 2937m 1711vs 1694vs 1670vs 1596s 1448s 1430m 1409m 1368m 1349s 1324m 1306m 1285m 1257s 1233m 1209m 1199s 1181s 1160s 1075w 1064w 1000m 936m 921m 790m 758m 728m and 693vs; dH(400 MHz CDCl3) 2.13 (s 3 H 6-CH3) 2.32 (q J 6.5 2 H 3-CH2) 2.71 (t J 6.4 2 H 4-CH2) 5.49 (t J 6.7 1 H 2-CH) 7.44ndash;7.48 (m 4 H ArH) 7.55ndash;7.60 (m 2 H ArH) and 8.03ndash;8.05 (m 4 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 22.94 (3-CH2) 29.73 (6-CH3) 40.50 (4-CH2) 54.46 (2-CH) 128.38 (4 CH) 128.68 (4 CH) 133.39 (2 CH) 135.57 (2 C) 196.04 (2 C O) and 208.29 (C O); m/z (EI 70 eV) 294 (0.1) M1 276 (0.5) M1 2 H2O 251 (1) M1 2 MeCO 236 (2) M1 2 Me2CO 224 (4) M1 2 MeCOCH CH2 189 (10) M1 2 PhCO 172 (20) M1 2 PhCOOH 105 (100) PhCO1 and 77 (28) Ph1 Found C 77.53; H 6.29; M (HRMS) 294.1243.Calc. for C19H18O3 C 77.53; H 6.16; M 294.1256.3146 J. Chem. Soc. Perkin Trans. 1 1997 Ethyl 2-acetyl-2-methyl-5-oxohexanoate 3j A mixture of the oxo ester 1j (785 mg 5.44 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (74 mg 0.27 mmol) was stirred overnight at room temp. after which it was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.48) to afford 3j as a colourless oil (910 mg 4.25 mmol 78); nmax/cm21 (ATR) 2984m 2941m 1711vs 1447m 1423m 1357s 1296m 1255s 1227s 1183s 1167s 1118s 1103s and 1019m; dH(400 MHz CDCl3) 1.25 (t J 7.2 3 H CH3) 1.32 (s 3 H CH3) 1.97ndash;2.05 (m 2 H CH2) 2.13 (s 3 H CH3) 2.14 (s 3 H CH3) 2.37ndash;2.43 (m 2 H CH2) and 4.14ndash;4.20 (m 2 H OCH2); 13C{1H} NMR (50 MHz CDCl3) 13.74 (CH3) 18.95 (CH3) 25.85 (CH3) 28.10 (CH2) 29.59 (CH3) 38.26 (CH2) 58.35 (C) 61.11 (OCH2) 172.32 (C O) 204.99 (C O) and 206.94 (C O); m/z (EI 70 eV) 214 (0.1) M1 172 (77) M 1 H1 2 COMe 144 (14) M1 2 CH2CHCOMe 126 (38) M1 2 CH2CHCMe 2 H2O 115 (72) MeCOC(Me)COO1 98 (100) MeCOC(Me)CO1 and 87 (57) CH2COOCH2Me1 Found C 61.66; H 8.46; M (HRMS) 214.1194.Calc. for C11H18O4 C 61.66; H 8.47; M 214.1205. 3-Acetyl-3-methylhepta-2,6-dione 3k A mixture of the dione 1k (623 mg of a 85 containing mixture; 28 i.e. 530 mg 4.64 mmol) the enone 2a (0.500 ml 6.00 mmol) and FeCl3?6 H2O (74 mg 0.27 mmol) was stirred overnight at room temp. after which all the volatile materials were removed in vacuo (2 h; to remove 3,3-dimethylpentane-2,4- dione which was the contaminant of the starting material 1k28) and the residue was chromatographed on silica gel (hexanendash;MTB 1 5; Rf 0.33) to yield 3k as a colourless oil (721 mg 3.91 mmol 84); nmax/cm21 (ATR) 2977m 2938m 1715vs 1697vs 1423m 1358s 1293m 1209m 1167m 1099m 967m and 920m; dH(400 MHz CDCl3) 1.31 (s 3 H 3-CH3) 2.04ndash;2.08 (m 2 H 4-CH2) 2.10 (s 6 H 1,19-CH3) 2.11 (s 3 H 7-CH3) and 2.30ndash;2.34 (m 2 H 5-CH2); 13C{1H} NMR (50 MHz CDCl3) d 18.33 (CH3) 26.30 (2 CH3) 27.36 (CH2) 29.68 (CH3) 38.11 (CH2) 65.05 (C) and 207.10 (3 C O); m/z (EI 70 eV) 184 (0.1) M1 142 (5) M1 2CH2CO 114 (2) M1 2MeCOCH CH2 99 (8) MeCOCH2COCH2 1 and 85 (100) MeCOCH2- CO1 Found M (HRMS) 184.1105.Calc. for C10H16O3 M 184.1099. Ethyl 2-oxo-1-(1-phenyl-3-oxobutyl)cyclopentanecarboxylate 3l A mixture of CHCl3 (0.5 ml) the oxo ester 1a (500 mg 3.20 mmol) the enone 2b (468 mg 3.20 mmol) and FeCl3?6 H2O (43 mg 0.16 mmol) was stirred overnight at room temp.after which all the volatile materials were removed in vacuo and the residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford two fractions of 3l containing two diastereoisomers isomer B (174 mg 0.58 mmol 18; Rf 0.39) and isomer A (697 mg 2.31 mmol 72; Rf 0.35). Isomer A colourless oil nmax/cm21 (ATR) 2979m 2893w 1748s 1718vs 1495m 1453m 1405m 1365m 1357m 1315w 1292m 1224s 1159s 1143s 1109m 1022m and 704s; dH(400 MHz CDCl3) 1.19 (t J 7.1 3 H CH3) 1.45ndash;1.54 (m 1 H CHH) 1.66ndash;1.84 (m 2 H CH2) 1.92ndash;2.05 (m 1 H CHH) 1.98 (s 3 H 49-CH3) 2.31ndash;2.42 (m 2 H CH2) 2.73 (dd J 16 2.6 1 H 29-CHH) 3.29 (dd J 16 11 1 H 29-CHH) 3.96 (dd J 11 2.6 1 H 19-CH) 4.04ndash;4.19 (m 2 H OCH2) and 7.16ndash;7.27 (m 5 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.94 (CH3) 19.31 (CH2) 30.18 (CH3) 32.20 (CH2) 39.42 (CH2) 44.42 (CH) 45.68 (CH2) 61.62 (CH2) 64.02 (C) 127.26 (CH) 128.41 (2 CH) 129.08 (2 CH) 139.49 (C) 170.65 (C O) 206.12 (C O) and 215.18 (C O); m/z (EI 70 eV) 302 (7) M1 284 (37) M1 2 H2O 256 (23) M1 2 CO 2 H2O 229 (66) M1 2 COOEt 211 (100) M1 2 C7H7 187 (29) M1 2 COOEt 2 COCH 186 (28) M1 2 COOEt 2 COCH2 185 (52) M1 2 COOEt 2 COMe 171 (46) M1 2 PhCHCHCO 156 (82) M1 2 PhCHCHCOMe 147 (42) PhCHCHCOMe 1 H1 and 43 (43) COMe1 Found C 71.51; H 7.33; M (HRMS) 302.1509.Calc. for C18H22O4 C 71.50; H 7.33; M 302.1518. Isomer B colourless oil nmax/ cm21 (ATR) 2978m 1747s 1715vs 1495m 1453m 1365m 1356m 1281m 1222s 1141s 1109m 1021m and 704s; dH(400 MHz CDCl3) 1.24 (t J 7.2 3 H CH3) 1.68ndash;1.86 (m 3 H CH2) 1.97ndash;2.05 (m 1 H CHH) 2.02 (s 3 H 49-CH3) 2.16ndash; 2.27 (m 1 H CHH) 2.39ndash;2.47 (m 1 H CHH) 2.95ndash;3.07 (m 2 H 29-CH2) 4.03 (dd J 9.1 5.5 1 H 19-CH) 4.10ndash;4.22 (m 2 H OCH2) and 7.16ndash;7.28 (m 5 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.71 (CH3) 19.06 (CH2) 29.59 (CH3) 29.68 (CH2) 38.04 (CH2) 43.03 (CH) 44.77 (CH2) 61.32 (CH2) 63.90 (C) 126.94 (CH) 127.99 (2 CH) 129.43 (2 CH) 138.54 (C) 169.63 (C O) 206.03 (C O) and 212.84 (C O); m/z (EI 70 eV) 302 (4) M1 284 (28) M1 2 H2O 256 (22) M1 2 CO 2 H2O 229 (50) M1 2 COOEt 211 (98) M1 2 C7H7 187 (31) M1 2 COOEt 2 COCH 186 (30) M1 2 COOEt 2 COCH2 185 (56) M1 2 COOEt 2 COMe 171 (52) M1 2 PhCHCHCO 156 (78) M1 2 PhCHCHCOMe 147 (49) PhCHCHCOMe 1 H1 91 (28) C7H7 1 and 43 (100) COMe1 Found C 71.70; H 7.33; M (HRMS) 302.1516.Calc.for C18H22O4 C 71.50; H 7.33; M 302.1518. Ethyl 2-oxo-1-(1,3-diphenyl-3-oxopropyl)cyclopentanecarboxylate 3m A mixture of CHCl3 (0.5 ml) the oxo ester 1a (500 mg 3.20 mmol) the enone 2c (660 mg 3.20 mmol) and FeCl3?6 H2O (8.7 mg 0.032 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo. The residue was chromatographed on silica gel (hexanendash;MTB 2 1; Rf 0.25) to afford 3m as an oil (1.13 g 3.09 mmol 96) comprising two diastereoisomers (ratio A/B = 53 47 by 1H NMR) which were not separated; nmax/cm21 (ATR) 3061w 3030w 2979m 2891w 1748vs 1723vs 1687vs 1597m 1580w 1496w 1448m 1404w 1366w 1293w 1224s 1178w 1144m 1109w 1033w 1017w 1002w 921w 860w 749m 691m and 702m; dH(400 MHz CDCl3) 1.19 (t J 7.1 3 H CH3) 1.25 (t J 7.1 3 H CH3) 1.45ndash; 1.53 (m 1 H) 1.73ndash;1.90 (m 4 H) 2.01 (dd J 18.5 9.0 1 H) 2.07ndash;2.14 (m 2 H) 2.21ndash;2.30 (m 1 H) 2.37ndash;2.44 (m 2 H) 2.47ndash;2.53 (m 1 H) 3.31 (dd J 17.1 J 2.7 1 H 29-CHH) 3.57ndash; 3.66 (m 2 H 29-CH2) 3.94 (dd J 17.1 J 10.8 1 H 29-CHH) 4.05ndash;4.22 (m 6 H 19-CH and OCH2) 7.15ndash;7.29 (m 10 H ArH) 7.39ndash;7.43 (m 4 H ArH) 7.49ndash;7.53 (m 2 H ArH) and 7.89ndash;7.91 (m 4 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.74 (CH3) 13.77 (CH3) 19.14 (CH2) 19.18 (CH2) 30.12 (CH2) 32.40 (CH2) 38.08 (CH2) 39.34 (CH2) 39.90 (CH2) 40.51 (CH2) 43.34 (CH) 44.48 (CH) 61.39 (2 OCH2) 63.85 (C) 64.04 (C) 126.85 (CH) 126.95 (CH) 127.76 (2 CH) 127.80 (2 CH) 127.93 (2 CH) 128.11 (2 CH) 128.21 (2 CH) 128.24 (2 CH) 128.99 (2 CH) 129.52 (2 CH) 132.65 (CH) 132.67 (CH) 136.62 (C) 136.64 (C) 138.80 (C) 139.45 (C) 169.90 (C O) 170.70 (C O) 197.35 (C O) 197.64 (C O) 212.99 (C O) and 215.21 (C O); m/z (EI 70 eV) 364 (2) M1 336 (6) M1 2 CO 319 (7) M1 2 OEt 291 (8) M1 2 COOEt 273 (54) M1 2 C7H7 245 (12) M1 2 PhCOCH2 209 (57) PhCHCH2COPh1 208 (55) PhCHCHCOPh1 207 (68) PhCHCCOPh1 155 (6) M1 2 PhCHCH2- COPh 143 (9) PhCCCOCH2 1 115 (14) PhCCCH2 1 105 (100) PhCO1 103 (16) PhCHCH1 91 (7) C7H7 1 and 77 (58) C6H5 1 Found C 75.74; H 6.75; M (HRMS) 364.1675.Calc. for C23H24O4 C 75.80; H 6.64; M 364.1675. Ethyl 2-oxo-1-(2-phenyl-3-oxobutyl)cyclopentanecarboxylate 3n A mixture of CHCl3 (0.5 ml) the oxo ester 1a (468 mg 3.00 mmol) the enone 2d (439 mg 3.00 mmol) and FeCl3?6 H2O (41 mg 0.15 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo. The residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford one fraction of 4b as a colourless oil (88 mg 0.30 mmol 20; Rf 0.58) and two fractions of 3n containing two diastereoisomers; isomer B (163 mg 0.54 mmol 18; Rf 0.42) and isomer A (490 mg 1.62 mmol 54; Rf 0.35).Isomer A a col- J. Chem. Soc. Perkin Trans. 1 1997 3147 ourless oil nmax/cm21 (ATR) 2978m 2937w 1749vs 1714vs 1454m 1405w 1355m 1276m 1231s 1190s 1155s 1111s 1029m 763m and 701s; dH(400 MHz CDCl3) 1.13 (t J 7.1 3 H CH3) 1.72 (dt J 13.0 J 7.7 1 H) 1.82ndash;1.92 (m 2 H) 2.02 (s 3 H CH3) 2.13ndash;2.23 (m 2 H) 2.30 (dt J 19.1 7.3 1 H) 2.45 (dd J 12.5 6.1 1 H) 2.52 (dd J 14.5 7.1 1 H) 3.81ndash;3.91 (m 2 H) 3.92ndash;4.01 (m 1 H OCHH) and 7.12ndash;7.30 (m 5 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.72 (CH3) 19.09 (CH2) 28.82 (CH3) 32.79 (CH2) 34.97 (CH2) 37.10 (CH2) 55.43 (CH) 59.79 (C) 61.19 (OCH2) 127.19 (CH) 128.14 (2 CH) 128.71 (2 CH) 138.83 (C) 170.17 (C O) 206.93 (C O) and 213.55 (C O); m/z (EI 70 eV) 302 (3) M1 198 (10) M1 2 CH2CHPh 185 (34) M1 2 CH2CHPhCH 156 (100) M1 2 CH2CHPhCOMe and 91 (11) C7H7 1 Found C 71.53; H 7.37; M (HRMS) 302.1515.Calc. for C18H22O4 C 71.50; H 7.33; M 302.1518. Isomer B colourless oil nmax/cm21 (ATR) 2978m 2938m 1774vs 1716vs 1276m 1229m 1182m 1161s 1113m 1029m 1005m 759m and 702s; dH(400 MHz CDCl3) 1.23 (t J 7.2 3 H CH3) 1.76ndash;1.83 (m 1 H) 1.91ndash;2.01 (m 2 H) 2.02 (s 3 H CH3) 2.07ndash;2.17 (m 2 H) 2.29ndash;2.33 (m 2 H) 2.76 (dd J 14.2 8.8 1 H) 4.09ndash;4.15 (m 3 H CH OCH2) and 7.16ndash;7.32 (m 5 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.93 (CH3) 19.55 (CH2) 29.11 (CH3) 35.87 (CH2) 35.99 (CH2) 38.02 (CH2) 55.05 (CH) 59.22 (C) 61.23 (OCH2) 127.16 (CH) 128.06 (2 CH) 128.88 (2 CH) 139.49 (C) 171.85 (C O) 207.56 (C O) and 215.83 (C O); m/z (EI 70 eV) 302 (2) M1 185 (18) M1 2 CH2CHPhCH 157 (100) M1 2 CHCHPhCOMe and 128 (65) 111 (87) Found C 71.21; H 7.32; M (HRMS) 302.1515.Calc. for C18H22O4 C 71.50; H 7.33; M 302.1518. Ethyl 2-benzoyl-5-oxo-3-phenylhexanoate 3o A mixture of CHCl3 (0.5 ml) the oxo ester 1g (577 mg 3.00 mmol) the enone 2b (439 mg 3.00 mmol) and FeCl3?6 H2O (8.1 mg 0.030 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo and the residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford two fractions of 3o containing two diastereoisomers isomer A (136 mg 0.40 mmol 13; Rf 0.36) and isomer B (716 mg 2.12 mmol 71; Rf 0.27).Analytically pure samples were obtained by crystallisation from hexanendash;CH2Cl2 (5:1) at 220 8C. Isomer A colourless crystals mp 71ndash;72 8C; nmax/cm21 (ATR) 2981m 1730vs 1720vs 1685vs 1597m 1448m 1366m 1279s 1258s 1210s 1185m 1150s 1027m 986m 759m 739m 701s and 690s; dH(400 MHz CDCl3) 0.88 (t J 7.1 3 H CH3) 2.00 (s 3 H 6-CH3) 2.81ndash;2.84 (m 2 H 4-CH2) 3.78ndash;3.85 (m 2 H OCH2) 4.18ndash;4.24 (m 1 H 3-CH) 4.80 (d J 10.1 1 H 2-CH) 7.15ndash;7.22 (m 1 H ArH) 7.24ndash;7.33 (m 4 H ArH) 7.46ndash;7.51 (m 2 H ArH) 7.58ndash;7.61 (m 1 H ArH) and 8.04ndash;8.07 (m 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.61 (CH3) 30.24 (CH3) 40.96 (CH) 47.68 (CH2) 59.67 (CH) 61.40 (CH2) 126.23 (CH) 128.44 (4 CH) 128.76 (4 CH) 133.79 (CH) 136.44 (C) 140.27 (C) 167.77 (C O) 193.54 (C O) and 206.47 (C O); m/z (EI 70 eV) 338 (4) M1 292 (42) M1 2 H 2 EtO 247 (78) M1 2 C7H7 233 (22) M1 2 PhCO 187 (49) M1 2 PhCO 2 EtO 2 H 147 (21) M1 1 H1 2 PhCO 2 CO2 2 COMe 105 (100) PhCO1 and 77 (70) Ph1 Found C 74.41; H 6.36; M (HRMS) 338.1517.Calc. for C21H22O4 C 74.54; H 6.55; M 338.1518. Isomer B colourless crystals mp 163 8C; nmax/cm21 (ATR) 2980m 2931m 1734vs 1719vs 1685s 1597m 1581m 1495m 1448m 1366m 1358m 1278m 1254s 1208m 1175m 1154s 1096m 1022m 701s and 690m; dH(400 MHz CDCl3) 1.16 (t J 7.2 3 H CH3) 2.03 (s 3 H 6-CH3) 2.97ndash;3.00 (m 2 H 4-CH2) 4.13 (q J 7.1 2 H OCH2) 4.15ndash;4.22 (m 1 H 3-CH) 4.79 (d J 9.6 1 H 2-CH) 7.08ndash;7.10 (m 1 H ArH) 7.15ndash;7.19 (m 2 H ArH) 7.21ndash;7.26 (m 2 H ArH) 7.36ndash;7.40 (m 2 H ArH) 7.48ndash; 7.52 (m 1 H ArH) and 7.80ndash;7.83 (m 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.91 (CH3) 30.29 (CH3) 40.70 (CH) 47.26 (CH2) 59.08 (CH) 61.66 (CH2) 126.96 (CH) 128.12 (2 CH) 128.37 (2 CH) 128.48 (2 CH) 128.53 (2 CH) 133.32 (CH) 136.57 (C) 140.89 (C) 168.53 (C O) 193.64 (C O) and 206.38 (C O); m/z (EI 70 eV) 338 (1) M1 292 (16) M1 2 H 2 EtO 247 (36) M1 2 C7H7 233 (20) M1 2 PhCO 192 (22) M 1 H1 2 PhCO 2 CH2CO 187 (55) M1 2 PhCO 2 EtO 2 H 147 (14) M 1 H1 2 PhCO 2 CO2 2 COMe 105 (100) PhCO1 and 77 (20) Ph1 Found C 74.21; H 6.51; M (HRMS) 338.1522.Calc. for C21H22O4 C 74.54; H 6.55; M 338.1518. Ethyl 2-benzoyl-5-oxo-3,5-diphenylpentanoate 3p A mixture of CHCl3 (0.5 ml) the oxo ester 1g (577 mg 3.00 mmol) the enone 2c (625 mg 3.00 mmol) and FeCl3?6 H2O (8.1 mg 0.030 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo.The residue was chromatographed on silica gel (hexanendash;MTB 1 1; Rf 0.37) to afford 3p as a colourless solid (1.02 g 2.55 mmol 85) as a mixture of two diastereoisomers (A/B = 78 22 by 1H NMR). Only isomer A was obtained purely by crystallisation from hexanendash;CH2Cl2 (5 1) at 220 8C. Isomer A colourless crystals mp 138 8C; nmax/cm21 (ATR) 3062m 3029m 2981m 2906m 1734vs 1684vs 1597s 1579m 1495m 1448s 1411m 1367m 1335s 1261vs 1215vs 1181s 1152s 1096m 1078m 1034m 1017s 1001s 981m 749vs 700s and 689vs; dH(400 MHz CDCl3) 0.88 (t J 7.1 3 H CH3) 3.30 (dd J 16.1 9.8 1 H 4-CHH) 3.49 (dd J 16.0 4.0 1 H 4-CHH) 3.78ndash;3.86 (m 2 H OCH2) 4.37ndash;4.44 (m 1 H 3-CH) 4.92 (d J 9.9 1 H 2-CH) and 7.03ndash;8.10 (m 15 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.56 (CH3) 41.37 (CH) 42.94 (CH2) 59.88 (CH) 61.34 (CH2) 127.08 (CH) 128.09 (2 CH) 128.27 (2 CH) 128.44 (4 CH) 128.52 (2 CH) 128.76 (2 CH) 132.88 (CH) 133.74 (CH) 136.46 (C) 136.83 (C) 140.16 (C) 167.83 (C O) 193.64 (C O) and 197.92 (C O); m/z (EI 70 eV) 401 (0.5) M 1 H1 309 (78) M1 2 C7H7 295 (22) M1 2 PhCO 249 (82) M1 2 PhCO 2 EtO 2 H 209 (58) PhCOCH2CHPh1 105 (100) PhCO1 and 77 (67) Ph1 Found C 77.76; H 6.05; M (HRMS) 400.1677.Calc. for C26H24O4 C 77.97; H 6.04; M 400.1675. Isomer B (identified in a mixture with isomer A) dH(400 MHz CDCl3) 1.16 (t J 7.1 3 H CH3) 3.54ndash;3.57 (m 2 H 4-CH2) 4.16ndash;4.19 (m 2 H OCH2) 4.37ndash;4.44 (m 1 H 3-CH) 4.93 (d J 9.6 1 H 2-CH) and 7.03ndash;8.10 (m 15 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.89 (CH3) 41.05 (CH) 42.39 (CH2) 59.13 (CH) 61.63 (CH2) 126.82 (CH) 128.03 (2 CH) 128.17 (2 CH) 128.34 (4 CH) 128.37 (2 CH) 128.91 (2 CH) 132.72 (CH) 133.28 (CH) 134.84 (C) 136.58 (C) 140.89 (C) 168.63 (C O) 193.80 (C O) and 197.79 (C O).Ethyl 2-benzoyl-5-oxo-4-phenylhexanoate 3q A mixture of CHCl3 (0.5 ml) the oxo ester 1g (577 mg 3.00 mmol) the enone 2d (439 mg 3.00 mmol) and FeCl3?6 H2O (41 mg 0.15 mmol) was stirred overnight at room temp. after which all the volatile materials were removed in vacuo. The residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford one fraction (Rf 0.58) 4b as a colourless oil (75 mg 0.26 mmol 17) and a second fraction (Rf 0.38) containing 3q also as a colourless oil (751 mg 2.22 mmol 74). Product 3q consisted of two diastereoisomers (A/B = 57 :43 by 1H NMR) which were equilibrating and thus could not be separated; nmax/ cm21 (ATR) 2980m 1735vs 1712vs 1684vs 1597m 1493m 1448m 1355m 1274m 1228s 1193s 1183s 1173s 1156s 1096m 1028m 766m 752m 701s 690m and 668m; dH(400 MHz CDCl3) 1.12 (t J 6.6 3 H CH3) 1.16 (t J 7.1 3 H CH3) 1.99 (s 3 H CH3) 2.05 (s 3 H CH3) 2.25ndash;2.35 (m 2 H 3-CH2) 2.60ndash;2.72 (m 2 H 3-CH2) 3.77ndash;3.82 (m 2 H 2 4-CH) 4.04ndash;4.15 (m 4 H 2 OCH2) 4.14 (t J 7.1 1 H 2-CH) 4.22 (t J 7.3 1 H 2-CH) 7.08ndash;7.11 (m 2 H ArH) 7.17ndash;7.20 (m 2 H ArH) 7.23ndash;7.40 (m 6 H ArH) 7.41ndash;7.46 (m 3 H ArH) 7.50ndash; 7.57 (m 3 H ArH) 7.69ndash;7.71 (m 2 H ArH) and 7.92ndash;7.95 (m, 3148 J.Chem. Soc. Perkin Trans. 1 1997 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.80 (CH3) 13.86 (CH3) 28.92 (CH3) 29.09 (CH3) 30.84 (CH2) 30.91 (CH2) 51.20 (CH) 51.55 (CH) 56.53 (CH) 56.64 (CH) 61.28 (2 OCH2) 127.57 (CH) 127.75 (CH) 128.06 (2 CH) 128.16 (2 CH) 128.35 (2 CH) 128.40 (2 CH) 128.55 (2 CH) 128.79 (2 CH) 129.05 (2 CH) 129.13 (2 CH) 133.44 (2 CH) 135.54 (C) 135.88 (C) 137.60 (C) 137.77 (C) 169.54 (C O) 169.65 (C O) 194.94 (C O) 195.24 (C O) 206.97 (C O) and 207.35 (C O); m/z (EI 70 eV) 338 (2) M1 295 (40) M1 2 MeCO 278 (6) M 1 H1 2 MeCO 2 H2O 192 (100) M1 2 MeCOCPhCH2 146 (8) MeCOCPhCH2 1 105 (86) PhCO1 and 77 (16) C6H5 1 Found C 74.14; H 6.28; M (HRMS) 338.1515.Calc. for C21H22O4 C 74.54; H 6.55; M 338.1518. Ethyl 2-acetyl-5-oxo-3-phenylhexanoate 3r A mixture of CHCl3 (0.5 ml) the oxo ester 1f (390 mg 3.00 mmol) the enone 2b (439 mg 3.00 mmol) and FeCl3?6 H2O (41 mg 0.15 mmol) was stirred overnight at room temp. after which all the volatile materials were removed in vacuo.The residue was chromatographed on silica gel (hexanendash;MTB 1 1; Rf 0.25) to afford 3r as a colourless oil (630 mg 2.28 mmol 76) which solidified in the refrigerator and consisted of two equilibrating diastereoisomers (A/B = 57 43 by 1H NMR); nmax/cm21 (ATR) 2983m 1738vs 1714s 1495m 1454m 1419m 1357s 1292sh 1272m 1243s 1178m 1152s 1096m 1022s 758m and 701s; dH(400 MHz CDCl3) 0.95 (t J 7.2 3 H CH3) 1.25 (t J 7.1 3 H CH3) 1.96 (s 3 H CH3) 1.99 (s 3 H CH3) 2.00 (s 3 H CH3) 2.25 (s 3 H CH3) 2.77ndash;2.79 (m 2 H 4-CH2) 2.86ndash; 2.88 (m 2 H 4-CH2) 3.88 (q J 7.1 2 H OCH2) 3.85ndash;3.98 (m 4 H 2- and 3-CH) 4.17 (q J 7.2 2 H OCH2) and 7.15ndash;2.28 (m 10 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.27 (CH3) 13.61 (CH3) 29.03 (CH3) 29.58 (CH3) 29.85 (CH3) 29.96 (CH3) 39.77 (3-CH) 39.99 (3-CH) 47.06 (2 4-CH2) 60.83 (OCH2) 61.14 (OCH2) 64.13 (2-CH) 64.88 (2-CH) 126.70 (CH) 126.80 (CH) 127.70 (2 CH) 127.76 (2 CH) 128.03 (2 CH) 128.27 (2 CH) 140.14 (C) 140.33 (C) 167.51 (C O) 167.90 (C O) 201.45 (C O) 201.80 (C O) 205.64 (C O) and 205.97 (C O); m/z (EI 70 eV) 276 (1) M1 233 (17) M1 2 MeCO 230 (13) M1 2 EtOH 191 (12) M1 2 MeCOCH2 2 CO 187 (100) M1 2 Me 2 HCOOEt 185 (68) M1 2 C7H7 177 (16) M1 2 MeCOCH2 2 CH2CO 173 (26) M1 2 MeCOCH2 2 CO 2 H2O 147 (32) PhCHCHCOMe 1 H1 145 (54) PhCHCHCOMe1 2 H 131 (67) PhCHCHCO1 103 (24) PhCHCH1 Found C 69.72; H 7.27; M (HRMS) 276.1377.Calc. for C16H20O4 C 69.55; H 7.30; M 276.1362. Ethyl 2-acetyl-5-oxo-4-phenylhexanoate 3s A mixture of CHCl3 (0.5 ml) the oxo ester 1f (390 mg 3.00 mmol) the enone 2d (439 mg 3.00 mmol) and FeCl3?6 H2O (41 mg 0.15 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo.The residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford one fraction (Rf 0.58) containing 4b as a colourless oil (101 mg 0.35 mmol 23) and a second fraction (Rf 0.21) containing 3s as a colourless oil (381 mg 1.38 mmol 46). Product 3s consisted of two diastereoisomers (A/B = 55 :45 by 1H NMR) which were equilibrating and thus could not be separated; nmax/ cm21 (ATR) 2982m 2938m 1739vs 1712vs 1494m 1454m 1357s 1299m 1244s 1219s 1152s 1097m 1029s 958m 759s and 702s; dH(400 MHz CDCl3) 1.20 (t J 7.2 3 H CH3) 1.23 (t J 7.3 3 H CH3) 1.99 (s 6 H 2 CH3) 2.09 (s 3 H CH3) 2.12ndash; 2.22 (m 2 H 3-CH2) 2.15 (s 3 H CH3) 2.44ndash;2.55 (m 2 H 3-CH2) 3.18 (dd J 8.9 5.1 1 H 2-CH) 3.29 (t J 7.3 1 H 2-CH) 3.63ndash;3.69 (m 2 H 2 4-CH) 4.06ndash;4.19 (m 4 H 2 OCH2) 7.11ndash;7.14 (m 4 H ArH) and 7.26ndash;7.32 (m 6 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.88 (CH3) 13.93 (CH3) 28.58 (CH3) 28.85 (CH3) 28.96 (CH3) 29.09 (CH3) 29.60 (2 3-CH2) 56.44 (2 CH) 56.88 (2 CH) 61.27 (OCH2) 61.31 (OCH2) 127.61 (CH) 127.67 (CH) 128.14 (2 CH) 128.21 (2 CH) 129.08 (4 CH) 137.42 (C) 137.66 (C) 169.28 (C O) 169.31 (C O) 202.55 (C O) 202.65 (C O) 206.88 (C O) and 206.97 (C O); m/z (EI 70 eV) 276 (10) M1 161 (18) M1 2 2 MeCO 2 Et 159 (22) M1 2 HCOOEt 2 MeCO 145 (14) MeCO 2 CHPhCH2 1 143 (16) M1 2 MeCOCHPh 134 (22) PhCH2COMe1 130 (13) MeCOCH2- COOEt1 115 (14) MeCOCH2COOCH2 1 104 (100) PhCHCH2 1 and 91 (13) C7H7 1 Found C 69.55; H 6.94; M (HRMS) 276.1355.Calc. for C16H20O4 C 69.55; H 7.30; M 276.1361. 2-Acetyl-6-methyl-2,5-diphenyl-3,4-dihydro-2H-pyran 4b A mixture of absolute ethanol (0.3 ml) KOH (5 mg 0.09 mmol) and the enone 2d (292 mg 2.00 mmol) was stirred for 2 d at room temp. after which all the volatile materials were removed in vacuo. The residue was chromatographed on silica gel (hexanendash;MTB 1 1; Rf 0.58) to afford 4b as a colourless oil (277 mg 0.95 mmol 95); nmax/cm21 (ATR) 2919m 1718vs 1670s 1599m 1493s 1447m 1383m 1351m 1254m 1211s 1175s 1131m 1117s 1073m 1064m 1031m 989m 757s and 699s; dH(400 MHz CDCl3) 1.99 (s 3 H CH3) 2.09ndash;2.12 (m 1 H) 2.13ndash;2.18 (m 1 H) 2.18 (s 3 H CH3) 2.28ndash;2.37 (m 1 H) 2.61ndash;2.67 (m 1 H) 7.11ndash;7.14 (m 2 H ArH) 7.18ndash;7.22 (m 1 H ArH) 7.26ndash;7.34 (m 3 H ArH) 7.37ndash;7.41 (m 2 H ArH) and 7.52ndash;7.55 (m 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 18.06 (CH3) 24.35 (CH3) 24.72 (CH2) 29.93 (CH2) 85.69 (C) 111.34 ( C) 124.97 (2 CH) 126.11 (CH) 127.76 (CH) 128.00 (2 CH) 128.47 (2 CH) 128.65 (2 CH) 139.16 (C) 141.25 (C) 145.77 ( C) and 208.91 (C O); m/z (EI 70 eV) 292 (8) M1 249 (100) M1 2 MeCO 204 (52) M1 2 2 MeCO 129 (10) PhCCH2CH2C1 and 105 (50) PhCO1 Found C 81.98; H 6.68; M (HRMS) 292.1455.Calc. for C20H20O2 C 82.16; H 6.89; M 292.1462. Ethyl 2-hydroxy-4-oxo-2,6-diphenylcyclohexanecarboxylate 5a A mixture of absolute ethanol (1 ml) KOH (5 mg 0.09 mmol) the oxo ester 1g (577 mg 3.00 mmol) and the enone 2b (439 mg 3.00 mmol) was stirred overnight at 50 8C after which all the volatile materials were removed in vacuo to leave 5a as a colourless solid (1.01 g 3.00 mmol 100) which was purified by crystallisation from hexanendash;CH2Cl2 (1:1) at 220 8C; this afforded colourless crystals mp 211 8C; nmax/cm21 (ATR) 3354s 1718vs 1495m 1455m 1447m 1374m 1346m 1302m 1256m 1222m 1179m 1143m 1067m 1030m 750s 699s and 668m; dH(400 MHz CDCl3) 0.53 (t J 7.14 3 H CH3) 1.64 (s br 1 H OH) 2.70ndash;2.82 (m 4 H 3-CH2 and 5-CH2) 3.52ndash;3.55 (m 2 H OCH2) 3.72ndash;3.87 (m 1 H 6-CH) 4.44 (d J 2.78 1 H 1-CH) 7.24ndash;7.37 (m 8 H ArH) and 7.49ndash;7.52 (m 2 H ArH); 13C{1H} NMR (50 MHz CDCl3) d 13.23 (CH3) 43.28 (6-CH) 47.38 (CH2) 53.98 (CH2) 56.64 (1-CH) 60.60 (OCH2) 77.28 (2-C) 124.57 (2 CH) 127.51 (2 CH) 127.57 (2 CH) 128.43 (2 CH) 128.74 (2 CH) 140.28 (C) 144.13 (C) 174.17 (C O) and 205.88 (C O); m/z (EI 70 eV) 338 (4) M1 247 (25) M1 2 C7H7 177 (39) EtOCCH2CHPh1 162 (92) PhCHCH2CO2CH2 1 146 (19) CH2(CO)CH2CHPh1 131 (38) COCHCHPh1 105 (199) PhCO1 and 77 (24) C6H5 1 Found C 74.49; H 6.51; M (HRMS) 338.1526.Calc. for C21H22O4 C 74.54; H 6.55; M 338.1518. Ethyl 4-hydroxy-4-methyl-8-oxo-2-phenylbicyclo3.2.1octane-1- carboxylate 5b A mixture of absolute ethanol (1 ml) the oxo ester 1a (469 mg 3.00 mmol) the enone 2b (439 mg 3.00 mmol) and KOH (17 mg 0.30 mmol) was stirred for 12 h at 50 8C after which all the volatile materials were removed in vacuo. The residue was chromatographed on silica gel (hexanendash;MTB 1 1) to afford four fractions the first containing a mixture of starting materials 1a and 2b (63 mg; Rf ca. 0.5) the second isomer B of 3l (168 mg 0.556 mmol 19; Rf 0.39) the third isomer A of 3l (19 mg 0.064 mmol 2; Rf 0.35) and finally the cyclic product J.Chem. Soc. Perkin Trans. 1 1997 3149 5b (446 mg 1.48 mmol 49; Rf 0.20) a colourless solid mp 142ndash;143 8C (decomp.); nmax/cm21 (ATR) 3488s 2970s 2935s 1756vs 1722vs 1497m 1453m 1368m 1299m 1271s 1203m 1130m 1072m and 1022m; dH(400 MHz CDCl3) 1.08 (t J 7.16 3 H CH3) 1.36 (s 3 H CH3) 1.73 (ddd J 13.9 10.9 4.9 1 H 7-H) 1.85 (ddd J 14.5 5.01 0.8 1 H 6-H) 2.03ndash;2.15 (m 3 H OH 6-H 3-H) 2.35ndash;2.43 (m 2 H 2-H 7-H) 2.55ndash;2.64 (m 1 H 3-H) 3.99 (dd J 13.1 4.8 1 H 5-H) 4.01ndash;4.07 (m 2 H OCH2) and 7.19ndash;7.27 (m 5 H ArH); 13C{1H} NMR (CDCl3 50 MHz) d 13.93 (CH3) 20.09 (CH2) 20.67 (CH2) 27.75 (CH3) 39.48 (CH2) 46.66 (CH) 56.61 (CH) 61.01 (CH2) 61.23 (C) 77.64 (C) 127.09 (CH) 128.18 (2 CH) 128.29 (2 CH) 139.98 (C) 169.39 (C O) and 209.44 (C O); m/z (EI 70 eV) 302 (8) M1 284 (12) M1 2 H2O 274 (7) M1 2 CO 256 (11) M1 2 CH 2 H2O 244 (9) M1 2 Me2CO 229 (13) M1 2 COOEt 211 (66) M1 2 C7H7 156 (100) M1 2 Ph- CHCHCOMe 104 (47) PhCH CH1 91 (22) C7H7 1 and 55 (15) CH2CHCO1 Found M (HRMS) 302.1523.Calc. for C18H22O4 M 302.1518. Acknowledgements Support from Prof. S. Blechert and the Institut fuuml;r Organische Chemie der TU Berlin is gratefully acknowledged. The author also thanks the Fonds der Chemischen Industrie for financial support and a fellowship. References 1 D. A. Oare and C. H. Heathcock in Topics in Stereochemistry ed. E. L. Eliel and S. H. Wilen Wiley-Interscience New York 1989 vol. 19 p. 227. 2 M. Shibasaki H. Sasai and T. Arai Angew. Chem. 1997 109 1290; Angew. Chem. Int. Ed. Engl.1997 36 1236. 3 F. Bonadies A. Lattanzi L. R. Orelli S. Pesci and A. Scettri Tetrahedron Lett. 1993 34 7649; E. Keller and B. L. Feringa Tetrahedron Lett. 1996 37 1879; A. Soriente A. Spinella M. De Rosa M. Giordano and A. Scettri Tetrahedron Lett. 1997 38 289. 4 (a) C. P. Fei and T. H. Chan Synthesis 1982 467; (b) H. Brunner and B. Hammer Angew. Chem. 1984 96 305; Angew. Chem. Int. Ed. Engl. 1984 32 312; (c) P. Kocovsky and D. Dvorak Tetrahedron Lett. 1986 27 5015. 5 J. H. Nelson P. N. Howells G. C. DeLullo G. L. Landen and R. A. Henry J. Org. Chem. 1980 45 1246. 6 J. Christoffers J. Chem. Soc. Chem. Commun. 1997 943. 7 P. Laszlo M.-T. Montaufier and S. L. Randraimahefa Tetrahedron Lett. 1990 34 4867. 8 Literature cited in Gmelins Handbuch der Anorganischen Chemie Eisen Verlag Chemie Berlin 1932 vol.59 B pp. 554. 9 W. G. Dauben and J. W. McFarland J. Am. Chem. Soc. 1960 82 4245. 10 K. Alder H. Offermanns and E. Ruuml;den Chem. Ber. 1941 74 905. 11 A. S. Dreiding and A. J. Tomasewski J. Am. Chem. Soc. 1955 77 411. 12 H. Henecka Chem. Ber. 1948 81 179. 13 N. C. Ross and R. Levine J. Org. Chem. 1964 29 2346. 14 H. Plieninger and T. Suehiro Chem. Ber. 1956 89 2789. 15 H. Wynberg and R. Helder Tetrahedron Lett. 1975 4057. 16 R. N. Lacey J. Chem. Soc. 1960 1625. 17 Y. L. Chow S.-S. Wang and X.-E. Cheng Can. J. Chem. 1993 71 846. 18 S. Terashima S. Sato and K. Koga Tetrahedron Lett. 1979 3469. 19 W. Wilson and Z.-Y. Kyi J. Chem. Soc. 1952 1321; A. F. Renaldo and H. Ito Synth. Commun. 1987 17 1823. 20 W. Dieckmann and K. von Fischer Chem. Ber. 1911 44 966. 21 E. D. Bergmann S.Blumberg P. Bracha and S. Epstein Tetrahedron 1964 20 195; G. N. Walker J. Am. Chem. Soc. 1955 77 3664. 22 A. Garcia-Raso J. Garcia-Raso B. Campaner R. Mestres and J. V. Sinisterra Synthesis 1982 1037. 23 W. S. Rapson J. Chem. Soc. 1936 1626. 24 R. Connor and D. B. Andrews J. Am. Chem. Soc. 1934 56 2713. 25 R. W. Hoffmann Angew. Chem. 1992 104 1147; Angew. Chem. Int. Ed. Engl. 1992 31 1124; R. W. Hoffmann Chem. Rev. 1989 89 1841. 26 C. A. M. Fraga and E. J. Barreiro Synth. Commun. 1995 25 1133. 27 D. H. Grayson and M. R. J. Tuite J. Chem. Soc. Perkin Trans. 1 1986 2137. 28 A. W. Johnson E. Markham and R. Price Org. Synth. 1973 Coll. Vol. 5 785. 29 T. Sakai S. Matsumoto S. Hidaka N. Imajo S. Tsuboi and M. Utaka Bull. Chem. Soc. Jpn. 1991 64 3473. Paper 7/04873D Received 8th July 1997 Accepted 21st July 1997 J.Chem. Soc. Perkin Trans. 1 1997 3141 Novel chemoselective and diastereoselective iron(III)-catalysed Michael reactions of 1,3-dicarbonyl compounds and enones Jens Christoffersdagger; Technische Universitauml;t Berlin Institut fuuml;r Organische Chemie Sekretariat C3 Straszlig;e des 17.Juni 135 D-10623 Berlin Germany Iron(III) chloride hexahydrate catalyses the Michael reaction of 1,3-dicarbonyl compounds with middot;,lsquor;-unsaturated ketones under mild and neutral conditions with extraordinary efficiency. The chemoselectivity of this FeIII-catalysed process is superior to that of the classic base-catalysed Michael reaction since the latter suffers from various side reactions namely drawbacks such as aldol cyclisations and ester solvolysis. Excellent yields and chemoselectivities together with the environmentally friendly nature of FeIII catalysis makes this an important alternative to classic base catalysis.Moreover the reaction procedure is reasonably easy FeIII catalysis does not require inert or anhydrous conditions and in most cases no solvent is needed. In terms of diastereoselectivity the FeIII-mediated reaction may also prove superior to a base-catalysed one. In at least one example FeIII catalysis forms a diastereoisomer as the major kinetic product which is disfavoured in the base-mediated Michael reaction where a thermodynamic mixture is obtained. The relative configuration of the diastereoisomeric Michael products has been determined for two examples by synthesis and structure elucidation of the cyclic aldol derivatives. Introduction The Michael reaction of the 1,3-dicarbonyl compounds 1 and the enones 2 is classically a high yielding base-mediated process 1 and can even be performed with high stereoselectivity.2 However there are some disadvantages with base catalysis including side reactions of the starting materials and subsequent reactions of the Michael product 3.Incompatibilities with base-sensitive groups ester solvolysis and aldol processes leading to cyclic products or retro-aldol type decompositions can significantly decrease yields of the base-mediated Michael reactions in some cases. Alternatively it has been reported that the Michael reaction can be catalysed by lanthanide 3 or transition metal 4 compounds e.g. group VIII 1,3-dionato complexes 5 although not always with satisfactory efficiency.Recently we reported that iron(III) chloride hexahydrate is an extraordinarily efficient catalyst for the Michael reaction of 1,3-dicarbonyl compounds and enones,6 a hitherto unknown fact,Dagger; despite the known tendency of FeIII to form 1,3-dionato complexes.8 This ability together with ecological and economical considerations make iron the transition metal of choice in such work. FeIII catalyses the Michael reaction under mild and neutral conditions and thus the chemoselectivity of the FeIII catalysis is superior to that of the classic base-mediated process since both side-reactions and subsequent reactions under basic conditions are avoided. Moreover the reaction conditions for FeIII catalysis are reasonably easy no inert or anhydrous conditions are required and in some cases even solvents are unnecessary.This high efficiency together with excellent yields make the FeIII catalysis of the Michael reaction an important alternative to the classic base catalysis. Results and discussion Iron(III) catalysis of the Michael reaction We have been investigating the catalytic activity of several tran- dagger; E-Mail jchr@wap0105.chem.tu-berlin.de Dagger; In one case the application of a combination of Ni(acac)2 and FeCl3 was reported but the role of FeIII was ascribed to its Lewis acid character (activation of the enone).7 sition metal compounds in the Michael reaction of 1a with 2a to give 3a9 (Scheme 1). Although a number of the investigated systems showed activity only with FeCl3?6 H2O was there fast clean and complete conversion at room temperature. With 5 mol the reaction was quantitative within 1 h and with 1 mol within 3 h.Results obtained with another FeIII catalyst 4c and some NiII compounds are listed in Table 1. Using Fe(acac)3 which is not active itself but needed further Lewis acid activation rapid consumption of starting materials was observed too but the conversion was not clean 4a (Table 2) was formed as a by-product via a hetero-Dielsndash;Alder dimerisation of 2a.10 Of the NiII compounds only Ni(acac)2 5 gave full conversion but a high temperature was required. All other transition-metal compounds investigated in our studies are less efficient than the NiII compounds reported in Table 1. FeIII catalysis is generally very efficient in the conversion of various Michael donors 1 (see Table 2) with methyl vinyl ketone 2a; a list of products 3 prepared is shown in Table 3.Best results were obtained with the cyclic keto esters to give 3a 3b,6 3d11 and 3e.6 With only 1 mol FeCl3?6 H2O full conversion is achieved within a few hours at room temperature even if bulkier ester functions are present (3b). Generally no ester solvolysis side-reactions occur with ethyl or higher alkyl esters. Methyl esters like 1e are partially solvolysed by the hydrate water of the catalyst,sect; so that the product yield drops to 72 if 5 mol FeCl3?6 H2O is used; with only 1 mol of catalyst this side reaction is negligible (91 isolated yield). Reactions of acyclic keto esters to give 3f,12 3g 13 and 3j 14 as well as of b-diketones to give 3c,15 3h,16 3i 17 and 3k18 proceed a Scheme 1 O OEt O O Me O CO2Et Me O 3a 2a 1a + cat. sect; Decomposition product cycloheptanone was identified in the reaction mixture by GCMS.3142 J. Chem. Soc. Perkin Trans. 1 1997 little slower but use of 5 mol of catalyst results in full conversion within a few hours at room temperature and gives satisfactory product yields. It should be emphasised that since the starting materials and products in Table 3 are liquid at room temperature no solvent is necessary for the transformations (except for 3i which is solid at room temperature). Moreover as long as reactions are quantitative with no by-products formation the work-up procedure is reasonably simple filtration using a small column of silica gel removes all iron-containing materials. In addition since water is tolerated no inert or anhydrous conditions are required and reactions are carried out simply by mixing starting materials and the catalyst.para; Fast and quantitative conversions together with a straightforward work-up procedure makes the iron(III) chloride hexahydrate-catalysed Michael reaction a very efficient alternative to the classic base-mediated methodology.Chemoselectivity In Table 4 products 3lndash;s resulting from Michael reactions of various keto esters 1 (Table 2) with substituted enones 2bndash;d (Table 2) are listed. Generally these transformations need solvent because the products are either solid or viscous oils at room temperature and higher reaction temperatures (up to Table 1 Comparison of NiII and FeIII catalysis of the Michael reaction Catalyst a Conversion () b FeCl3?6 H2O Fe(acac)3 1 BF3?OEt2 NiCl2?6 H2O Ni(OAc)2?4 H2O Ni(acac)2 1 h RTc 100 40 mdash; mdash; mdash; 3 h RT mdash; 90d 5 21 19 24 h RT mdash; mdash; 41 68 61 3 h 50 8C mdash; mdash; 81 52 100 a Conditions 1a (1 equiv.) 1 2a (1.1 equiv.) 1 catalyst.(0.05 equiv.) no solvent. b By 1H NMR. c Room temp. d By-product 4a was formed. Table 2 List of starting materials 1 and 2 and by-products 4 O X O X = OEt X = OBui X = Me 1a 1b 1c Me O 2a O OEt O 1d Me O Ph 2b O OMe O 1e Ph O Ph 2c R O OEt O R = Me R = Ph 1f 1g Me O Ph 2d R O R O R = Me R = Ph 1h 1i O Me Me O 4a Me O X O Me X = OEt X = Me 1j 1k O Me Ph Ph O Me 4b para; Scaling up (more than 20 mmol) requires cooling of the mixture to prevent 2a from being evolved since the reactions are slightly exothermic. 50 8C) are required in some cases (see Experimental section for details). The iron(III)-catalysed conversion of the keto ester 1a with the enone 2d to the Michael product 3n provides a typical example of the chemoselectivity achieved with this method (Scheme 2).Base catalysis namely 5 mol KOEt in absolute EtOH failed to give any of the desired product 3n the enone dimer 4b19 being obtained instead from a hetero Dielsndash;Alder reaction of two equivalents of 2d. A similar result was obtained in the base-mediated conversion of Michael donors 1f and 1g with the enone 2d. In these cases no Michael products 3q20 or 3s 20 were detected in the reaction mixtures only the dimer 4b being isolated. Thus the FeIII-catalysed reaction of the keto esters 1 with the enone 2d seems to be the only way to perform a Michael reaction and indeed products 3n 3q and 3s were isolated in good to moderate yields although the dimer 4b was also found in these reactions as a by-product (see Experimental section).Obviously with FeIII catalysis the Michael reaction of 2d becomes fast enough to compete seriously with the Dielsndash; Alder dimerisation side-reaction. The Michael adduct 3o20 was formed by FeIII-catalysed conversion of 1g with the acceptor 2b in good yield (Scheme 3). In contrast under conditions of base catalysis the primary product 3o cyclises in a subsequent aldol reaction to give 5a,21 a byproduct which is not detectable in an FeIII catalysed reaction. Thus the optimum yield achievable in the base-mediated Scheme 2 O Me Ph Ph O Me O Ph Me O EtO2C Fe(III) base 4b 91 3n 76 1a + 2d Table 3 Michael reactions of 1 with 2a to give 3andash;k Product O COX Me O O CO2Et Me O O Me O CO2Me R O CO2Et Me O R O COR Me O Me O Me O Me COX X = OEt X = OBui X = Me R = Me R =