Читать книгу Polar Organometallic Reagents - Группа авторов - Страница 15

An alternative approach to directed aromatic metalation has focused not upon replacing the alkyllithium base but on activating it. Two methods by which to achieve this were developed some time ago. The first was centered on TMEDA‐activation (TMEDA = N,N,N′,N′‐tetramethylethylenediamine) and the second involved the use of tert‐butoxide‐complexed alkyllithium reagents in the form of LICKOR superbases. The former route was employed to achieve site‐selective deprotonation with different selectivity to that achieved using the alkyllithium alone. Meanwhile, whereas unimetallic superbases are known [41], 1966–1967 saw the introduction by Lochmann [42] and Schlosser [43] of heterobimetallic (Li–Na/K) superbases. Subsequently extended to incorporate a range of alkyllithium adducts of potassium alkoxides [44], the most widely known example deploys traditional organolithium reagents in tandem with KOt‐Bu. Such heterobimetallic systems have shown enormous reactivity toward deprotonative metalation [45, 46]. As such, they enable the smooth deprotometalation of low acidity hydrocarbons [47] and weakly activated or nonactivated benzene derivatives [48] with, in some cases, unique regioselectivity [49] and also the facility for multideprotonation [50]. Whilst the synthetic importance of heterobimetallic superbases was quickly established, the characterization of such air‐sensitive materials lagged behind.

Structural information on superbases has been gathered from a number of areas. Concerning the activation of organolithium bases, early evidence for the existence of organolithium‐alkoxylithium aggregates of the type n‐Bu _x Li₄(OR)_4‐x (R = n‐Bu or t‐Bu; x = 1–4) in solution [51] was later substantiated by the crystallographic characterization of (n‐BuLi)₄(LiOt‐Bu)₄30 [52]. This sparked a search for similar complexes capable of acting as structurally well‐defined models for superbases. Hence, alkoxy‐ and/or amido(alkali metal) combinations, now demonstrating the at least partial replacement of lithium with a higher group 1 congener, were investigated. These revealed not only a number of heterobimetallic structures [53] but also ternary alkali metal aggregates such as the twelve‐vertex cage [{PhN(H)}₂(Ot‐Bu)LiNaK(TMEDA)₂]₂31 ₂ (Figure 1.4a) [54]. However, whilst providing fascinating insights into alkali metal structural chemistry, their study has yet to allow a relationship to synthetically useful Li–K superbases to be established. A significant advance towards this was made with the study of the tetralithium–tetrapotassium amide‐alkoxide [{t‐BuN(H)}₄(Ot‐Bu)₄Li₄K₄(C₆H₆)](C₆H₆) 32 (Figure 1.4b). Upon dissolution, crystals of this material achieved the smooth metalation of toluene [55]. Meanwhile, the combination of three alkali metal alkoxides yielded similar reactivity towards toluene through forming the fully characterizable ternary complex [Li₄Na₂K₂(Ot‐Bu)₈(μ‐L)] 33 (L = η⁶ : η⁶‐C₆H₆, TMEDA) at ambient temperature. In contrast, none of the individual alkali metal alkoxides MOt‐Bu (M = Li, Na, K) or their binary combinations detectably reacted with toluene under the same conditions (Scheme 1.9) [56].

Attempts to structurally investigate organo/alkoxy alkali metal aggregates, which might better represent the more frequently employed types of superbasic reagent already introduced [42, 43] have been hampered by the tendency of in situ preparations to precipitate microcrystalline (often organopotassium) products [57, 58]. Accordingly, a strategy was devised to overcome this problem through the in situ formation of a ligand providing both alkoxy and alkyl functionalities. To this end, sodium 2,4,6‐trimethylphenoxide was treated with n‐BuLi to result in lateral sodiation, and this enabled the crystallization of [{4,6‐Me₂C₆H₂(O)(CH₂)}LiNa(TMEDA)]₄34 ₄ (Figure 1.5), in which each phenoxide ligand had undergone a single benzylic deprotonation [59]. Consistent with previous reports that Li–O bonding takes precedent over Na–O interactions (logically a reflection of hard/soft interactions) [54, 55], X‐ray diffraction revealed a structure based upon a (LiO)₄ pseudocubic core, with the Na centres peripheral and coordinated perpendicular to the plane of the benzyl group.

Figure 1.4 Molecular structures of (a) [{PhN(H)}₂(t‐BuO)LiNaK(TMEDA)₂]₂31₂ and (b) [{t‐BuN(H)}₄(t‐BuO)₄Li₄K₄(C₆H₆)](C₆H₆) 32.

Sources: Adapted from Mackenzie et al. [54]; Kennedy et al. [55].

Scheme 1.9 Reactivity of metal alkoxides towards toluene in C₅D₅N. L = η⁶ : η⁶‐C₆H₆, TMEDA; M, M' = Li, Na, K.

Figure 1.5 Molecular structure of [{4,6‐Me₂C₆H₂(O)(CH₂)}LiNa(TMEDA)]₄34₄.

Source: Adapted from Harder and Streitwieser [59].

Moving forward a decade, the first organo/alkoxy Li–K species to be isolated from a LICKOR mixture was reported. To achieve this, n‐BuLi‐KOt‐Bu was reacted with C₆H₆ in THF, yielding crystalline (PhK)₄(PhLi)(t‐BuOLi)(THF)₆(C₆H₆)₂35, an aggregate containing all the expected constituents of a superbase. Furthermore, its place in this family of reagents was confirmed by its ability to metalate toluene. From a structural perspective, in spite of the substoichiometric Li content of 35, interactions with the lighter alkali metal dominate: bonds from tert‐butoxide and phenyl moieties are both shorter and more strongly directed towards Li⁺ compared to peripheral K⁺, again likely reflecting a competing preference for hard/soft stabilization (Figure 1.6) [60].

Whilst providing encouraging evidence for the reactivity of mixed Li–K species, the high potassium content of 35 leaves some doubt surrounding its validity as a model superbase since pure PhK can achieve the same reactivity towards toluene [61]. However, further evidence for Li–K cooperativity has been gathered by the judicious choice of neopentyllithium (NpLi, Np = CH₂C(CH₃)₃) as a stable and hydrocarbon‐soluble alkyl equivalent. When partnered with t‐BuOK, to form a LICKOR mixture, its use has allowed for the isolation of mixed‐metal aggregate Li₄K₄Np_2.75(t‐BuO)_5.2536 [62]. Though detailed analysis of the structure was complicated by the positional disorder of the ligands and metal, the fundamental motif – a central, square planar arrangement of K centres, bicapped by [RLi(Ot‐Bu)₂LiR]^2⁻ (R = Np/t‐BuO) – was clear (Figure 1.7a). Meanwhile, the NMR spectroscopic monitoring of experiments in which the NpLi : LiOt‐Bu : KOt‐Bu ratio was varied revealed evidence for equilibria involving Li‐rich mixed Np/Ot‐Bu species (represented by the series Li₄K₃Np _x (Ot‐Bu)_{7‐ x} ) in solution. In this work, the excess of Li could be rationalized by the tendency of poorly soluble NpK to precipitate. Consistent with this, cooling solutions rich in Np^⁻ led to crystallization of the Li‐rich mixed metal species, Li₄K₃Np_3.16(Ot‐Bu)_3.8437. X‐ray diffraction exposed in this the coordination of alkyl moieties to both K and Li centres (Figure 1.7b). This suggested hybrid Li/K–C bond polarity, which was in turn postulated to be responsible for superbasic activity.

Recently, this work has been extended to examine other Li–K compounds of the type described above, whose compositions are dominated by either metal. Thus, the structure of Li₄KNp₂(Ot‐Bu)₃38 has been elucidated in the solid‐state, revealing a square pyramidal metal architecture wherein interaction of apical K with the CH₃ component of an Np unit in an adjacent monomer results in a solid‐state dimer (Figure 1.8) [63]. Meanwhile, exploring K‐rich end members has lately provided the first crystallographic evidence for the existence of a mixed organo/alkoxypotassium species, K₄Np(Ot‐Am)₃39 (t‐Am = CH₂C(CH₃)₂CH₂CH₃). Based on its favourable solubility profile and donor solvent‐free constitution, reactivity comparable with or superior to organopotassium reagents was anticipated. In the event, this was realized in the polymetalation of ferrocene. Though isolation of metalated intermediates was not possible, the regioselectivity of metalation could be competently assessed through the study of carboxylated intermediates and the crystallization of selected ferrocene methyl ester derivatives (Scheme 1.10).

Figure 1.6 Molecular structure of the core of (PhK)₄(PhLi)(t‐BuOLi)(THF)₆(C₆H₆)₂35 (K = dark purple, Li = pink, O = red).

Source: Adapted from Unkelbach et al. [60].

Figure 1.7 Molecular structures of (a) Li₄K₄Np_2.75(t‐BuO)_5.2536 and (b) Li₄K₃Np_3.16(Ot‐Bu)_3.8437. Minor ligand disorder omitted (K = dark purple, Li = pink, O = red).

Source: Adapted from Benrath et al. [62].

Figure 1.8 Molecular structure of [Li₄KNp₂(Ot‐Bu)₃]₂38₂ (K = dark purple, Li = pink, O = red).

Source: Adapted from Jennewein et al. [63].

Scheme 1.10 The elaboration of ferrocene employing 39. Fc = ferrocenyl, n = 2–4.

The issues of reactivity that plagued the use of organolithium reagents have then, in many cases, been overcome by the advent of LICKOR superbases. However, the introduction of higher alkali metal reagents has led to new issues that have limited the applicability of heterometallic reagents except in the hands of specialists in air‐sensitive and nonstandard techniques. So though they offer significant, and often unique advantages, the limitations expressed here of the bases conventionally used for metalation have led to the search for new organometallic combinations capable of similar functionality, ideally under mild conditions. This has led researchers to utilize alkali metals in tandem with metals from elsewhere in the periodic table. This search has ultimately revealed new chemistry for the metalation of aromatic compounds and the advent of ate complexes as synthetic tools [64]. However, first it is worth briefly exploring the use of group 2 elements in conjunction with alkali metals–covered at more length in Chapter 3.