Читать книгу Polar Organometallic Reagents - Группа авторов - Страница 18

Extending the earlier studies on forming aromatic and heteroaromatic zinc derivatives, the 1990s saw the development of halogen–zinc exchange reactions of aromatic halides using organozinc reagents to encompass the use of heterobimetallic ones. Lithium trialkylzincates (R₃ZnLi) have long been established as versatile reagents for the 1,4‐addition of alkyl groups to α,β‐unsaturated ketones [78, 79]. However, investigation of the potential of less reactive lithium aryldimethylzincates for smoothly effecting 1,4‐addition represents a more recent development. Hence, a novel preparation of lithium aryldimethylzincates using the halogen–zinc exchange reaction of a range of aromatic halides with lithium trimethylzincate 54, followed by reaction of the resulting intermediates with electrophiles has been reported [80]. In the first step, iodobenzene was treated with 54 in THF at −78 °C for 1 h to give assumed 55 (R = H). The introduction of benzaldehyde (R′ = Ph) gave 1,2‐adduct 56 in 65% yield. Aromatic iodides with a para substituent were examined for the same reaction, and various functional groups were found to tolerate the halogen–zinc exchange reaction – the corresponding 1,2‐adducts being obtained in good yields (Scheme 1.13). Especially noteworthy was the observed compatibility of a nitro group with this reaction sequence. Thus, the reaction of the arylzincate with propionaldehyde proceeded smoothly with various substituents; the arylzincate from p‐iodoanisole gave the 1,2‐adduct in 66% yield, with this level of effectiveness maintained using the corresponding nitro substrate (68%).

In seeking to extend the portfolio of this early class of ate chemical reactivity, and with the aim of preparing a new class of indolylzinc derivatives, the direct halogen–zinc exchange reaction of 2‐ and 3‐iodoindoles with 54 was next studied [81]. Preparation of indolylzincate by the treatment of the 3‐iodoindole with lithium trimethylzincate at −78 °C in THF was followed by the introduction of benzaldehyde to give the desired alcohol in 51% yield. A similar reaction with the indolylzincate and allyl bromide gave 3‐allylindole in 44% yield. The metalation gave the expected products in slightly higher yields when the reaction was instead conducted with a combination of lithium trimethylzincate and TMEDA. Importantly, although 3‐lithio‐1‐phenylsulfonylindole has long been known to isomerize easily to the thermodynamically stable 2‐lithioindole 57, the alternative use of lithium zincates avoided the formation of 2‐substituted isomers (Scheme 1.14). The heterobimetallic method, therefore, became considered to be advantageous for the selective metalation of the indole 3‐position, giving 58.

Scheme 1.13 Halogen–zinc exchange in a para‐substituted aromatic iodide.

The outer shell of the zinc centre in a lithium trialkylzincate is occupied by 16 electrons and, as will be explored more fully in Chapter 2, this leaves room for additional ligand coordination to form an 18 electron state [82]. With this in mind, reports on tetraalkylzincates have presented X‐ray studies that have disclosed a tetrahedral arrangement about the zinc [83–86]. However, the reactivities of tetraalkylzincates have not been well studied. Intramolecular carbometalation has been examined using the iodine–zinc exchange of allyl 2‐iodophenyl ether. When the iodo ether was treated with 54, halogen–metal exchange proceeded smoothly, but no intramolecular carbometalation was observed, yielding 59. In contrast, the reaction of the iodo ether with Me₄ZnLi₂60 was followed by hydrolysis to give 3‐methyldihydrobenzo[b]furan 61 in 42% yield, which behaviour was explained by invoking an intramolecular carbozincation (Scheme 1.15) [87, 88].

As they emerged as a practically applicable new subset of organozinc reagent, organozincates rapidly became regarded as attractive candidates for carbon–carbon bond formation. To evaluate the non‐transferability of alkyl groups, the migratory aptitude of various alkyls from a range of lithium aryldialkylzincates was investigated [89]. Tert‐butyl turned out to be the best non‐transferable group and lithium tri(tert‐butyl)zincate was therefore found to represent a highly effective reagent for the chemoselective halogen–zinc exchange of functionalized organic halides. As an exemplar, the migratory aptitude of lithium dialkylphenylzincates was investigated as illustrated in Scheme 1.16. Phenyllithium was reacted with a range of dialkylzinc substrates to form the corresponding ate complexes 62, and these were then treated with 0.5 equivalent benzaldehyde. Spectroscopic analysis of the resulting crude mixture of products then elucidated the ratio of the two alcoholic products.

The status of tert‐butyl as the most effective non‐transfer group led to the performance of lithium tri(tert‐butyl)zincate 65 also being examined in the halogen–zinc exchange of functionalized aryl halides. It was observed that the exchange reaction generally proceeded smoothly and that electrophilic quenching of the arylzincate intermediates also completed, giving 66–69, without transferring the tert‐butyl group (Scheme 1.17).

Scheme 1.14 Use of a lithium zincate to avoid intermediate rearrangement of the type undergone by the 3‐lithio‐1‐phenylsulfonylindole.

Scheme 1.15 The contrasting reactivity of an allyl 2‐iodophenyl ether with Me₃ZnLi 54 and Me₄ZnLi₂60.

Scheme 1.16 Investigation of the migratory aptitude of lithium dialkylphenylzincates using phenyllithium, a dialkylzinc and benzaldehyde.

Scheme 1.17 Tert‐butyl as a nontransfer group; 65 incurs the halogen–zinc exchange of functionalized aryl halides.

To demonstrate the compatibility of this reaction with diverse electrophilic substituents, methyl 4‐iodobenzoate was treated with 65 and the resulting assumed arylzincate 70 was then treated with a range of electrophiles. Attempts at 1,2‐addition with benzaldehyde, alkylation with methyl iodide and allylation with allyl iodide all gave the expected products (e.g. 71, Scheme 1.18), and acylation with benzoyl chloride proceeded smoothly without a palladium catalyst to give the corresponding ketone 72. This avoidance of the use of a noble metal catalyst was advantageous not only on cost grounds but also on account of the strict guidelines that regulate trace amounts of transition metals in pharmaceuticals. In contrast to the reaction of 2‐iodoanisole with lithium trimethylzincate 54, which gave the corresponding alcohol in a low yield of 29% upon introduction of benzaldehyde, the use of tri(tert‐butyl)zincate 65 enabled the corresponding reaction in the much higher yield of 83%.

To further investigate the synthetic potential of this halogen–zinc exchange reaction of 65, transmetalation of the putative lithium di(tert‐butyl)phenylzincate 73 with thienylcyanocuprate was surveyed (Scheme 1.19, Th = thienyl) [89]. The phenyl group was found to smoothly undergo 1,4‐addition to cyclohex‐2‐enone in the presence of lithium thienylcyanocuprate 74 to give 75, while the reaction in the absence of 74 was very sluggish.

From a structural perspective, the first fully characterized lithium alkylzincate appeared in the 1960s [83]. The structure of 60 revealed tetravalent zinc, the completion of a full outer electron shell in which has more recently characterized the reactivity of alkali metal zincates in deprotometalation, of which more later. Moving on several decades, the imperative to complete zinc’s outer shell was similarly revealed by external solvation in simple amine adducts in spite of the deployment of sterically demanding ligands in tri[di(trimethylsilyl)methyl]zincates [90]. Competition between tri‐ and tetra(organyl)zincates was further elucidated at about this time through the use of the monoanionic, potentially C,N‐chelating C₆H₄CH₂NMe₂‐2 (DMBA) ligand. This allowed the observation of 18 electron Zn centres in both (DMBA)₃ZnLi 76 (following 4:1 (DMBA)₂Zn:DMBALi reaction; Figure 1.9a) and spirocyclic [91] (DMBA)₄ZnLi₂77 (2 : 1 reaction; Figure 1.9b) [92]. Interestingly though, reactions of these zincates with cyclohexanone gave a very early indication of disproportionation, a phenomenon seen more recently in alkyl, amido, and mixed alkyl–amido zincates [86]. This issue has been revisited in ate chemistry discussed elsewhere in this book that finds applications in directed aromatic deprotometalation [93, 94]. Further advances towards the isolation and full characterization of simple, trivalent, 16 electron Zn were made by the expansion of steric demands in silylmethylzincates [90, 95, 96] in combination with the introduction of Lewis bases capable of abstracting the alkali metal. Accordingly, with TMEDA present it proved possible to observe ion‐separated [Me_3‐n Zn{CH(SiMe₃)Ph} _n ][Li(TMEDA)₂] 78 (n = 1–3) [97]. Recently, remarkable insights into competing solvent‐separated ion‐pair (SIP, obtained in the presence of diglyme) and contact ion‐pair (CIP, obtained in the presence of PMDETA) formation saw the subject go full‐cycle, returning to simple lithium methylzincates but now probing 16 electron metal centres in the trimethylate (Scheme 1.20). The same work used the relatively new technique of DOSY to shed light on the solution behaviour of these species. Hence, though ¹H NMR spectroscopy revealed single methyl resonances for both PMDETA 79 and diglyme 80 systems in solution, DOSY was able to establish that this was due to exchange in the former case and not SIP formation. The power of DOSY to advance our understanding of the potentially elaborate solution chemistry of polar organometallics is the subject of detailed discussion elsewhere in this book.

Scheme 1.18 Treatment of methyl 4‐iodobenzoate with 65 preceded allylation (with allyl iodide) and acylation (with catalyst‐free benzoyl chloride).

Scheme 1.19 Addition reaction involving the transmetalation of putative lithium di(tert‐butyl)phenylzincate 73 with thienylcyanocuprate.

Figure 1.9 Molecular structures of (a) solvated (DMBA)₃ZnLi 76 and (b) (DMBA)₄ZnLi₂77, which can be selectively targeted by modulating the (DMBA)₂Zn:DMBALi ratio in reaction.

Sources: Adapted from Wyrwa et al. [91]; Rijnberg et al. [92].

Scheme 1.20 Competing SIP and CIP formation in Me₃ZnLi chemistry.