Vinylketene complexes

By a photochemically induced elimination of CO, a chromium carbene complex with a free coordination site is generated. That species can coordinate to an alkyne, to give the alkyne-chromium carbonyl complex 4. The next step is likely to be a cycloaddition reaction leading to a four-membered ring compound 5. A subsequent electrocyclic ring opening and the insertion of CO leads to the vinylketene complex 6 ... 

A greatly enhanced chemoselective formation of phenol is observed for alkoxy(alkenyl)carbene complexes compared to alkoxy(aryl)carbene complexes. This behaviour reflects the ease of formation of the rf-vinylketene complex intermediate E starting from alkenylcarbene complexes for aryl complexes this transformation would require dearomatisation. 

The cyclobutenone 70 is transformed to the r/4-vinylketene complex 72 with (t/5-indenyl)Co(PPh3)2 71. The vinylketene complex 72 undergoes cyclization with alkynes to produce the corresponding phenols 73. FeCl3 oxidation of the (2-phenylvinyl)ketene complex, however, leads to the naphthol 74. A catalytic synthesis of phenols via the vinylketene intermediates 72 is achieved by the use of Ni(COD)2 as a catalyst [36]. (Scheme 26)... 

Cycloaddition of the carbene chromium complexes 97 with CO incorporation provides a versatile method for naphthol synthesis, in which the metallacy-clic intermediates 99 are involved [47]. An alternative entry to 101 is achieved by metal carbonyl-catalyzed rearrangement of the cyclopropenes 98 via the same metalla-cyclobutenes 99 and vinylketene complexes 100 [52], Mo(CO)6 shows a higher activity than Cr(CO)6 and W(CO)6. The vinylketene complex 103 is formed by the regioselective ring cleavage of 1,3,3-trimethylcyelopropene 102 with an excess of Fe2(CO)9 [53]. (Scheme 35 and 36)... 

In addition to the ring opening of cyclopropenes noted above, vinylketene complexes 103 have been prepared by (1) ligand initiated carbonyl insertion of vinyl carbene complexes 104 and (2) benzoylation of ,/3-unsaturalcd acyl ferrates 105 (Scheme 20)114. X-ray diffraction analysis of these vinylketene complexes indicates that the structure may be best represented as a hybrid between an /j4-dicnc type complex (103) and an jj3-allyl r/1 acyl complex (106). The Fe-Cl distance (ca 1.92 A) is shorter than the Fe-C2, Fe-C3, or Fe-C4 distances (ca 2.1-2.2 A)113a-C. In addition, the C—C—O ketene array is not linear (bend angle ca 135°). 

These structural considerations have provided a convenient way to delineate the scope of this review. We shall only be considering vinylketene complexes that fall within the bonding parameters laid out above. Complexes such as 5,46,5 and 76 are all -complexes and as such will not be covered here, unless they have direct relevance to related -complexes. Similarly, free vinylketenes7 will not be discussed, except where their chemistry has been related to their complexed analogues. The review includes, to the best of our... 

The most common citation of rj4-vinylketene complexes in the literature is to be found in mechanistic discussions of the Dotz annulation.8 In the years since the reaction was discovered,2 there has been an enormous amount of research conducted to determine a definitive mechanism. Nevertheless, certain transformations in the mechanism are still open to speculation, the most recent reassessment having been postulated by Sola.911 is not within the scope of this review to comprehensively discuss a subject which has been extensively covered elsewhere.8,10,11 However, the reactions of i74-vinylketene complexes of chromium, and indeed of all the other metals in this review, are so intrinsically linked with the Dotz annulation that we must acquaint ourselves with its intricacies. We shall see the same structures occurring in mechanisms time after time, and the same classes of organic molecules being isolated as final products. 

The vast majority of the circumstantial evidence concerning this reaction points toward the presence of i74-vinylketene complexes such as E-18 at some point along the reaction coordinate. Moreover, the chromacyclohexa-diene pathway cannot fully explain all of the data concerning the benzannu-lation reaction.13 The accumulation of evidence for the intermediacy of rf-vinylketenes has been recorded elsewhere,8,10,11,13,27 but will be considered here briefly to give an impression of the general methods that have been used to elucidate the reactivity of vinylketene complexes. 

Analysis of product distributions has often been used to evince the presence of 774-vinylketenes in the Dotz reaction. Wulff has studied the reactions of a wide variety of substituted aryl chromiumcarbenes and drawn conclusions from the data that point directly toward i74-vinylketene complexes as key intermediates.13 He has also isolated40 bicyclic lactone products (e.g., 23) characteristic of a van Halban-White cyclization41-44 from the reaction of vinylcarbenes and ketoalkynes. These bicyclic lactones are not generally... 

Mori has reported that in the reaction of chromium carbene 25 with an alkyne containing a tethered 4-amidobutyne unit (26), a postulated vinylketene complex (27) is intercepted by nucleophilic amide attack, yielding a mixture of lactams (28 and 29).15 The expected naphthol 30 was also isolated in low yield. 

Many other reactions designed to trap intermediate vinylketene complexes are known. Dotz has used alkynes with a pendant alcohol to produce the butyrolactones E-31 and Z-31 in a 70 30 ratio and 34% yield.16 These are formed by the nucleophilic attack of the pendant alcohol on the ketene... 

Alkenes may be used to trap intermediate vinylketene complexes (e.g.,... 

Perhaps the most compelling evidence for the intermediacy of rf-vinylketene complexes in the Dotz annulation has been provided by the isolation of both free and complexed vinylketenes from the reaction mixture. Dotz himself has reported52,53 the isolation of both arene-complexed (38) and... 

The single best piece of evidence for the intermediacy of vinylketene complexes is however the isolation and characterization by Wulff22 of amine-stabilized T74-vinylketenechromium(0) complexes (42) from the thermolysis of the chromium carbene complexes 43, containing a tethered alkyne functionality. This was the first time that a d6-rf-vinylketene complex of any group 6 metal had been isolated. 

Crucially, it was found22 that upon reaction of the vinylketene complex 42.a with 1-pentyne, the same product distribution was seen as for the direct thermolysis of 1-pentyne with the precursor carbene 43.a. The analysis was simplified by reduction of the crude reaction mixture with McMurry s reagent to produce a mixture of the isomeric indanols 44-46. 

The indanols 44 and 45 can only be the products of a formal [4 + 2] cycloaddition23 of the vinylketene complex 42.a with 1-pentyne. Note that upon reaction of 42.b with diethylpropynylamine a formal [2 + 2] cycloaddition65 is seen to take place, yielding the cyclobutenone 47 along with a tricarbonylchromium complex, tentatively identified as 48.66,67 As one would expect, the vinylketene complex 42.b underwent 1,2-additions with pyrrolidine and sodium methoxide in methanol, yielding 49 and 50, respectively. The CO-insertion step leading to vinylketene formation is reversible in some systems,51,68,69 but there is no evidence of this for complex 42.a. Heating a benzene solution of complex 42.a at 80°C under an atmosphere... 

Complex exo-60 is then protonated to give the 773-vinylcarbene complex exo-64, which subsequently inserts carbon monoxide in the well-established manner (see Sections II,B, V,B, VI,B, VI,C, VI,E, VI,J, and VII), affording the 16-electron species endo-65. Anion trapping of the unsaturated species finally yields the vinylketene complex endo-62. 

Among the metals investigated, manganese and vanadium have been the least explored in the synthesis of -(vinylketene) complexes, with only... 

The reactions between cyclopropenes and carbon monoxide in the presence of transition metals have been of some use in synthesis,93 and in 1978 Binger initiated a study of the reactions between metal carbonyls and cyclopropenes in order to elucidate the generality of these reactions.75 It was found that dicarbonyl 775-cyclopentadienyl(tetrahydrofuran)manganese(I) reacted with 3,3-dime thy Icy clopropene at 0°C to produce -(vinylketene) complex 81 in fair yield. The only other transition metal in Binger s study that was found to react with 3,3-dimethylcyclopropene in this manner was iron (see Section VI,B). 

Although the preparative chemistry of (vinylketene)cobalt(I) complexes is relatively limited in the literature, the methods used include all the major procedures that have been more widely exploited in the analogous chromium and iron systems. There are many similarities between the intermediates involved in the synthesis of vinylketene complexes of iron, chromium, and cobalt, but as the metal is varied the complexes containing analogous ligands often exhibit significant differences in stability and reactivity (see Sections II and VI). Comparison of such species has often been an important aim of the research in this area. The (vinylketene)cobalt(I) complexes have also been shown to be synthetically useful precursors to a variety of naphthols, 2-furanones, ce-pyrones, phenols,6,22,95 >8, y-unsaturated esters,51 and furans.51,96a... 

In 1986, Wulff reported the first synthesis of an -vinylketene complex from the reaction between a carbene and an alkyne51 the former had of course been postulated as key intermediates in analogous reactions of chromium8 and iron97 carbenes (see Sections II.B and VI.C, respectively). On treatment of the tricarbonylcobalt carbene (84) with 3-hexyne under mild conditions, the yf-vinylketene complex (85) and the y-keto unsaturated... 

The 774-vinylketene complex (85) could be oxidatively decomplexed with Ce(IV) to afford the lactone (87). Although no reaction was observed with methanol (unlike a postulated chromium analogue16,18 26), treatment with sodium methoxide produced the expected /3, y-unsaturated ester (88). Thermolysis of complex 85 afforded no trace of the naphthol that one would expect33 from a proposed chromium vinylketene complex with the same syn relationship between the phenyl group and the ketene moiety. Instead, only the furan (89.a) was seen. Indeed, upon exhaustive reaction of tricarbon-ylcobalt carbenes (84 and 90) with different alkynes, the furans (89.a-d) were isolated as the exclusive products in moderate to excellent yields. 

In a later paper by Weiss,68 the methodology was extended to a more complex cyclopropene, and an intermediate cobaltacyclobutene (103) was proposed. In an analogous insertion reaction with nonacarbonyldiiron, a vinylcarbene complex was isolated along with the expected vinylketene complex (see Section VI,B). However, no such vinylcarbene cobalt complex was isolated, even when cyclopentadienyl bis(ethene) cobalt was used in place of dicarbonylcyclopentadienyl cobalt, and the only product isolated was the vinylketene complex 104, represented here in the rf -allylacyl structure. 

O Connor proposed a mechanism involving deinsertion of carbon monoxide from the vinylketene complex 106 to form the new cobaltacyclobutene 109. The cobalt may then undergo a 1,3-shift to the carbonyl of the ester group to create the oxycobaltacycle 110, before deinsertion of the cobalt moiety forms the furan 108. Alternatively, 109 may rearrange to the vinyl-carbene 111, which then undergoes ester-carbonyl attack on the carbene carbon to form the zwitterionic species 112, which finally aromatizes to yield the furan 108. Notice that this latter postulate is identical to the final steps of the mechanism formulated by Wulff (see Section V,B) for the reaction between a cobalt carbene and an alkyne, in which a cobaltacyclobutene is a key intermediate.51... 

It has already been seen that cobalt may insert into cyclopropenes to yield, after further insertion of a carbonyl ligand, -vinylketene complexes... 

It should be noted that upon reaction with an electron-rich cyclobutenone (R1 = R3 = H, R2 = OEt), the major product formed was a cobaltacyclopen-tenone, which may also be considered to be an 772-vinylketene complex. A similar restructure was isolated after heating 114.a with a large excess of triphenylphosphine, which replaces the ligand site vacated by the central C2 unit. Interestingly, such 772-vinylketene complexes are the expected products from the analogous insertion of rhodium into cyclobutenones (e.g., 7). 

This chemistry was developed with the specific intention of application to the synthesis of phenols. There had only been one previous example of the addition of alkynes to 774-vinylketene complexes yielding phenols,22 despite a cornucopia of other organic fragments having been isolated from such reactions (see Sections II, V, and VI). The best results were obtained on reaction of the 3-phenylvinylketene complex 114.a with several alkynes. 

Vollhardt s investigations100 into electrocyclic transformations on a CpCo template produced the following unusual result. After photolysis of CpCo(CO)2 in the presence of the tosyl hydrazone of trans-4-pheny -3-buten-2-one, the only isolated product was the vinylketene complex 123. Note that the tricarbonyliron analogue of this complex has also been isolated.3,87 The mechanism of formation was not discussed, but it seems likely that the ketene carbonyl originated as a carbonyl ligand that replaced the hydrazone moiety, perhaps via a vinylcarbene intermediate. 

In a complementary study, Weiss reacted equimolar amounts of diiron nonacarbonyl and various cyclopropenes, yielding the products 130.a-130.f in fair to good yield.103 Notice that (on crystallographic evidence), he prefers to represent the products in the rf r -allylacyl structure (see Section 1). Once isolated, the phenyl-substituted vinylketene complexes were reacted... 

Weiss studied68a the reactivity of both new complexes, and found that a variety of phosphines and phosphites would also convert the vinylcarbene complex 139 into the corresponding vinylketene complex (140), capturing one of the carbonyl ligands from the coordination sphere of the metal to become the ketene carbonyl. Only in the case of triphenylphosphine was the dicarbonyl(phosphine)vinylcarbene complex (141) isolated, which then required addition of carbon monoxide to convert it to the dicarbonyl(triphe-nylphosphine)vinylketene complex 140.a. This interconversion was reversible and proceeded quantitatively. 

The vinylketene complex 140.g was shown to react toward electrophiles at the ketene oxygen. Although iodomethane had no effect, reaction with trimethyloxonium tetrafluoroborate caused methylation to produce the cationic complex 143 in fair yield. 

---