Tetramethylene zwitterion intermediates

Electron-rich bifunctional vinyl ethers (e.g. ethylene glycol divinyl ether) react with electron-poor alkenes (e.g. TCNE) to produce cyclobutanes in good yields via tetramethylene zwitterion intermediates. In some cases, cyclobutanes reacted with the solvent (MeCN) to yield tetrahydropyridines.9 Trifluoromethanesulfonimide is an... 

This cautionary attitude toward the stand d enumeration of mechanistic criteria seems fully justified when one reflects on the sorts of evidence advanced in support of 1,4-tetramethylene zwitterionic intermediates in thermal [2 + 2] cycloadditions. 

The tetramethylene is a resonance hybrid of 1,4-diradical (, = y in Eq. (22)) and zwitterionic (, = +,—) limiting structures. The character of the tetramethylene is determined by the nature of the terminal substituents. A very strong donor substituent at one of the terminal carbons and a very strong acceptor substituent at the other leads to zwitterionic intermediates. Otherwise, for instance, phenyl or vinyl group at the donor terminal and diester, cyano-ester or anhydride at the acceptor terminal, will favor the diradical form. 

Hall has introduced an empirical test to estimate the relative importance of diradical and zwitterionic forms in tetramethylene intermediates rrans-1,4-tetramethylene diradical intermediates may initiate alternating radical copolymerizations if they add to another alkene faster than they undergo conformational isomerization to the gauche form and give a cyclobutane product through carbon-carbon bond formation, while zwitterionic 1,4-tetramethylene intermediates may initiate ionic homopolymerizations. 

The case for 1,4-zwitterionic intermediates based on chemical trapping experiments depends on obtaining such products using reagents reactive toward the tetramethylene and unreactive toward both addends and the [2 + 2] cycloadduct. 

From organic chemistry it is known that cycloaddition reactions leading to cyclobutanes are required to be stepwise reactions, according to the Woodward-Hoffmann rules [131]. A bond is formed between the two olefins, leading to a tetramethylene intermediate (T). In a subsequent step, the second bond is formed, yielding the cycloadduct. Depending on the reactants, either zwitterionic or diradical tetramethylenes can be proposed as intermediates [132, 133]. 

A unifying hypothesis for the observed organic chemistry was advanced by Huisgen [132], who suggested that all tetramethylenes lie on a continuous scale between zwitterionic and diradical structures and may be regarded as resonance hybrids of the two extreme forms. The predominant nature of the tetramethylene intermediate is determined by the terminal substituents, and the termini can interact with each other by through-bond interaction [132, 134]. 

Lewis examined two systems, tra/w-a-phenylcinnamonitrile with a 2,4-hexadiene [109] and 9-cyanoanthracene with a butene [39], in an attempt to find evidence to support a proposed 1,4-zwitterionic tetramethylene intermediate. His results, including the lack of a pronounced solvent effect when the polarity was varied, failed to support such an intermediate. Curiously, he did not propose a 1,4-diradical intermediate which would not be susceptible to solvent stabilization. 

Other theoretical studies have been concerned with diradical or zwitterionic tetramethylenes approached from [2 + 2] cycloreversions. For l,l-dicyano-2-methoxycyclobutane, cleavage of C-1—C-2 gives a gauche intermediate (98) which may isomerize to the trans intermediate (99). They are nearly isoenergetic, and both have predominant dir ical character. ... 

This empirical yardstick draws a boundary between [2 + 2] cycloadditions that occur through essentially diradical or essentially zwitterionic tetramethylenes the resonance hybrid view which encompasses a continuous range of more or less dipolar intermediates is neglected in favor of a more decisive either/or discrimination. 

Numerous [2 + 2] cycloadditions which may proceed through dipolar intermediates (tetramethylenes which may be represented as resonance hybrids of singlet diradicaJ and zwitterionic forms) react with partial or even complete loss of the stereochemical relationships present in starting materials. Reactions between the cis and trans isomers of l,2-bis(trifluoromethyl)-l,2-dicyanoethylene and cis- and trans-propenyl n-propyl ether are not completely stereoselective. The trans electrophilic alkene (107) and the cis isomer of the enol ether (108), for instance, give rise to products (109), (110) and (111), the last having trifluoromethyl groups in cis relationship one to the other. ... 

In summary, then, the diradical versus zwitterionic issue seems to have shifted in recent years, away from an either/or dichotomy and toward a more integrated view, one seeing tetramethylene intermediates as more or less dipolar diradicals. A small amount of zwitterionic character does not obviate the essentially diradical nature of these intermediates, any more than monoradicals forego radical character if they are more or less electrophilic or nucleophilic. 

A little bit of zwitterionic character goes a long way energetically, especially in transition states for cyclobutane formation. 1,4-Zwitterionic tetramethylene intermediates in [2 -i- 2] cycloadditions may well be important in some reactions, where donor and acceptor substituents are so strong that diradical character is overshadowed and yet one-electron transfer does not take over, but they do not at this point seem common. 

The question whether the intermediates 6.24 and 6.26 are zwitterions or biradicals has not yet been answered. As known from the tri- and tetramethylene species (see Hoffmann, 1968 Hoffmann et al., 1970 Hiberty, 1983 Harcourt and Little, 1984 and Ejiri et al., 1992), zwitterions and biradicals are probably extremes of a structure-dependent degree of charge separation. ... 

The experimental evidence cited above indicates that this does not occur. As suggested at the right of Fig. 6.6, the in-plane glide (6i) begins well before the intended HOMO-LUMO crossing, which is avoided because all four MOs have the same irrep (o ) in The HOMO and LUMO of the extended tran-soid zwitterion are qualitatively similar to those in the biradical illustrated in Fig. 6.4, but are more widely separated in energy As a result, the orbital approximation - and the symmetry analysis based upon it - is no less reliable than for the biradical mechanism. The zwitterionic mechanism can be accomodated by Fig. 6.5, with self-evident modifications arising from the polar nature of the tetramethylene intermediate. ... 

First, the most important step in the analysis of the above scheme requires us to characterize the structure of the intermediate since it is only when its structure is known with sufficient certainty that the predictions based on the value of the overlap determinants can be reliable. In general, the question of the structure of the intermediate can, of course, be quite complicated, but in the case of pericyclic reactions, which are of concern here, the situation is slightly more simple. This is due to the fact that the set of structures which could play the role of the eventual intermediates is restricted only to species of a biradical and/or zwitterionic nature [60,61], so that the proposal of the structure of the eventual intermediate need not be so complicated. Thus, e.g., in the case of 2j + 2g ethene dimerization, the corresponding intermediate can be naturally identifi with the tetramethylene biradical. In such a case, the whole two step reaction scheme can be desribed as follows ... 

The zwitterionic tetramethylene intermediate has been trapped with methanol as a linear 1-methoxybutane, indicating that the addition of tetracyanoethylene to NVK occurs in a stepwise manner. Once the zwitterions are formed, the course of the reaction is determined largely by experimental conditions. With excess NVK, cationic polymerization occurs. If the concentrations of NVK and tetracyanoethylene are comparable, collapse of the intermediate to the corresponding cyclobutane is favored. The cyclobutane formation is reversible, and the isolated cyclobutane alone is capable of initiating the cationic polymerization of NVK. The polymerization of NVK by the stable radical cation salts phenothiazin hexafluoroantimonate [464] and tris-p-bromophenylamminium hexachloro-antimonate [465] were shown to occur via charge transfer initiation too. 

Returning to the initiation process H K Hall and his collaborators have been investigating the spontaneous copolymerization of vinyl monomers containing donor and acceptor groups They have presented strong evidence for the existence of a tetramethylene intermediate 30 (Eq 7) which is a resonance hybrid of a zwitterion and a spin paired biradical, the terminal carbon atoms of which sense each other by through bond coupling. ... 

This intermediate has been identified in many small molecules reactions such as cycloaddition The tetramethylene 30, depending upon conditions, may cyclize or react with more monom to form either a copolymer or a homopolymer. It has been proposed that a largely zwitterionic produces cationic homopolymers whereas a largely biradical tetramethylene produces an alternating copolymer ... 

Our earlier studies led to conceive a mechanistic hypothesis based on the bondforming initiation theory and on the polar character of the exocycUc double bond of BFl. Thus, we assumed that the interaction of the strongly polarized exocycUc double bonds of two BFl molecules could produce the formation of an activated dimer featuring a zwitterionic tetramethylene moiety (Scheme 16). This zwitterionic dimer intermediate was supposed to show a negative charge stabilized as an aromatic ind-enyl anion and an unstabilized which could initiate the cationic polymerizadon (initiation step) process leading to the corresponding polymer [20,21],... 

---