Cycloaddition, 1,3-anionic other

Two major side reactions compete with the coupling reaction protonation of the radical anion followed by further reduction and protonation leading to the saturated dihydro product, and polymerization induced by the basic dianion formed by coupling of two radical anions. Other, less typical reaction pathways include reaction between a radical anion and a molecule of substrate. Scheme 2, dimerization of two radicals formed by protonation of the initial radical anion. Scheme 3, or, infrequently, cleavage of the radical anion followed by coupling. However, for radical anions derived from monoactivated alkenes, the pathway in Scheme 2 has only been unequivocally established as a major pathway in a few cases in which the final zero-electron product is a cyclobutane, that is, a cycloaddition product. 

Grigg and coworkers developed bimetallic domino reactions such as the electro-chemically driven Pd/Cr Nozaki-Hiyama-Kishi reaction [69], the Pd/In Barbier-type allylation [70], Heck/Tsuji-Trost reaction/1,3 dipolar cycloaddition [71], the Heck reaction/metathesis [72], and several other processes [73-75]. A first example for an anion capture approach, which was performed on solid phase, is the reaction of 6/1-134 and 6/1-135 in the presence of CO and piperidine to give 6/1-136. Liberation from solid phase was achieved with HF, leading to 6/1-137 (Scheme 6/1.30) [76]. 

The first steps involve coordination and cycloaddition to the metal. Insertion of a third molecule of ethene leads to a more instable intermediate, a seven-membered ring, that eliminates the product, 1-hexene. This last reaction can be a (3-hydrogen elimination giving chromium hydride and alkene, followed by a reductive elimination. Alternatively, one alkyl anion can abstract a (3-hydrogen from the other alkyl-chromium bond, giving 1-hexene in one step. We prefer the latter pathway as this offers no possibilities to initiate a classic chain growth mechanism, as was also proposed for titanium [8]. The byproduct observed is a mixture of decenes ( ) and not octenes. The latter would be expected if one more molecule of ethene would insert into the metallocycloheptane intermediate. Decene is formed via insertion of the product hexene into the metallo-cyclopentane intermediate followed by elimination. 

N-Unsubstituted azomethine ylides may be generated thermally (79), and the N-metalated, 2-azaallyl anion versions may be generated by action of nonmetalhc bases such as l,8-diazabicyclo[5.4.0]undec-7-ene (DBU) on certain imines (80). Although they are assumed to show similar chemical properties, these two species usually show different reaction patterns, as shown in Scheme 11.7, where the regio-and stereoselectivities of the cycloadditions are quite different (24,78-80). Metala-tion of (alkylideneamino)acetonitriles can be performed with metallic bases other than LDA. Thus, butyllithium, ethylmagnesium bromide, and magnesium bromide-diisopropylamide are also effective (78). The N-magnesioazomethine... 

Heating /V-lithioaziridines provides 2-azaallyl anions, which undergo concerted cycloaddition reactions with certain alkenes and other anionophiles (Scheme 39) (74AG(E)627, B-88MI101-02). 

All the other cycloadditions, such as the [4+2] cycloadditions of allyl cations and anions, and the [8+2] and [6+4] cycloadditions of longer conjugated systems, have also been found to be suprafacial on both components, wherever it has been possible to test them. Thus the trans phenyl groups on the cyclopentene 2.65 show that the two new bonds were formed suprafacially on the rrans-stilbene. The tricyclic adducts 2.61, 2.77, 2.79, and 2.83, and the tetracyclic adduct 2.82, show that both components in each case have reacted suprafacially, although only suprafacial reactions are possible in cases like these, since the products from antarafacial attack on either component would have been prohibitively strained. Nevertheless, the fact that they have undergone cycloaddition is important, for it is the failure of thermal [2+2], [4+4] and [6+6], and photochemical [4+2], [8+2] and [6+4] pericyclic cycloadditions to take place, even when all-suprafacial options are open to them, that is significant. 

A secondary orbital interaction has been used to explain other puzzling features of selectivity, but, like frontier orbital theory itself, it has not stood the test of higher levels of theoretical investigation. Although still much cited, it does not appear to be the whole story, yet it remains the only simple explanation. It works for several other cycloadditions too, with the cyclopentadiene+tropone reaction favouring the extended transition structure 2.106 because the frontier orbitals have a repulsive interaction (wavy lines) between C-3, C-4, C-5 and C-6 on the tropone and C-2 and C-3 on the diene in the compressed transition structure 3.55. Similarly, the allyl anion+alkene interaction 3.56 is a model for a 1,3-dipolar cycloaddition, which has no secondary orbital interaction between the HOMO of the anion, with a node on C-2, and the LUMO of the dipolarophile, and only has a favourable interaction between the LUMO of the anion and the HOMO of the dipolarophile 3.57, which might explain the low level or absence of endo selectivity that dipolar cycloadditions show. 

The essential features of the Diels-Alder reaction are a four-electron n system and a two-electron it system which interact by a HOMO-LUMO interaction. The Diels-Alder reaction uses a conjugated diene as the four-electron n system and a it bond between two elements as the two-electron component. However, other four-electron it systems could potentially interact widi olefins in a similar fashion to give cycloaddition products. For example, an allyl anion is a four-electron it system whose orbital diagram is shown below. The symmetry of the allyl anion nonbonding HOMO matches that of the olefin LUMO (as does the olefin HOMO and the allyl anion LUMO) thus effective overlap is possible and cycloaddition is allowed. The HOMO-LUMO energy gap determines the rate of reaction, which happens to be relatively slow in this case. 

This result has been confirmed with other dipolarophiles. Conservation of the original geometry is observed as exemplified by reactions with methyl dimaleate and fumarate. In the presence of a base such as the methoxide anion, the ri.v-isomer issued from fumarate was transformed into the more stable traw.v-cycloadduct form from maleate. Cycloaddition of benzaldehyde leads to TV-benzyl-5-phenyloxazolidine.65 443... 

In this account, we will focus on the transient analysis of these systems, which has strongly contributed to a deeper understanding of the diverse reaction modes (Patemo-Buchi, proton abstraction, cycloaddition). In general, aromatic ketones were selected as electron acceptors for reasons of suitable excitation and long wavelength absorption of the radical anion intermediates. Among them, fluorenone 3 is particularly well suited since the concentration, solvent, temperature, and cation radius dependence of the absorption spectra of pairs formed with metal cations are already known [29]. Hogen-Esch and Smid [30, 10] pointed out that a differentiation between CIP and SSIP is possible for fluorenone systems. On the other hand, FRI s and SSIP s cannot be differentiated simply by their UV/Vis absorption spectra, whereas for instance conductance measurements may be successful. However, the portion of free radical ions in fluorenyl salt solutions was shown to be less important [9, 31]... 

The reaction of the acid chloride phenylhydrazone (11) with base gives the nitrile-imine 1,3-dipolar compound (12) which reacts with potassium thiocyanate to give the A2-thiadiazo-line (13 Scheme 1). Thus the cycloaddition occurs at the C=S and not the C=N bond. This regioselectivity can be explained in terms of the frontier orbital treatment. Due to the electron rich nature of the thiocyanate anion, its reaction with (12) is expected to be controlled by the LUMO and HOMO of (12) and the thiocyanate respectively. As the HOMO of the thiocyanate anion has the larger orbital coefficient on the sulfur atom, it can be concluded that the larger orbital coefficient in the LUMO of (12) is on the carbon atom. This is also in agreement with other dipolar cycloadditions (82H( 19)57). 

---