Difunctional

The reactions described so far can be considered as alkylation, alkenylation, or alkynylation reactions. In principle all polar reactions in syntheses, which produce monofunctional carbon compounds, proceed in the same way a carbanion reacts with an electropositive carbon atom, and the activating groups (e.g. metals, boron, phosphorus) of the carbanion are lost in the work-up procedures. We now turn to reactions, in which the hetero atoms of both the acceptor and donor synthons are kept in a difunctional reaction produa. 

Before we start with a systematic discussion of the syntheses of difunctional molecules, we have to point out a formal difficulty. A carbonmultiple bond is, of course, considered as one functional group. With these groups, however, it is not clear, which of the two carbon atoms has to be named as the functional one. A 1,3-diene, for example, could be considered as a 1,2-, 1,3-, or 1,4-difunctional compound. An a, -unsaturated ketone has a 1.2- as well as a 1,3-difunctional structure. We adhere to useful, although arbitrary conventions. Dienes and polyenes are separated out as a special case. a, -Unsaturated alcohols, ketones, etc. are considered as 1,3-difunctional. We call a carbon compound 1,2-difunctional only, if two neighbouring carbon atoms bear hetero atoms. 

The most general methods for the syntheses of 1,2-difunctional molecules are based on the oxidation of carbon-carbon multiple bonds (p. 117) and the opening of oxiranes by hetero atoms (p. 123fl.). There exist, however, also a few useful reactions in which an a - and a d -synthon or two r -synthons are combined. The classical polar reaction is the addition of cyanide anion to carbonyl groups, which leads to a-hydroxynitriles (cyanohydrins). It is used, for example, in Strecker s synthesis of amino acids and in the homologization of monosaccharides. The ff-hydroxy group of a nitrile can be easily substituted by various nucleophiles, the nitrile can be solvolyzed or reduced. Therefore a large variety of terminal difunctional molecules with one additional carbon atom can be made. Equally versatile are a-methylsulfinyl ketones (H.G. Hauthal, 1971 T. Durst, 1979 O. DeLucchi, 1991), which are available from acid chlorides or esters and the dimsyl anion. Carbanions of these compounds can also be used for the synthesis of 1,4-dicarbonyl compounds (p. 65f.). 

A classical reaction leading to 1,4-difunctional compounds is the nucleophilic substitution of the bromine of cf-bromo carbonyl compounds (a -synthons) with enolate type anions (d -synthons). Regio- and stereoselectivities, which can be achieved by an appropiate choice of the enol component, are similar to those described in the previous section. Just one example of a highly functionalized product (W.L. Meyer, 1963) is given. 

The growing importance of cyclopropane derivatives (A. de Meijere, 1979), as synthetic intermediates originates in the unique, olefin-like properties of this carbocycle. Cyclopropane derivatives with one or two activating groups are easily opened (see. p. 69f.). Some of these reactions are highly regio- and stereoselective (E. Wenkert, 1970 A, B E. J. Corey, 1956 A, B, 1975 see p. 70). Many appropriately substituted cyclopropane derivatives yield 1,4-difunctional compounds under mild nucleophilic or reductive reaction conditions. Such compounds are especially useful in syntheses of cyclopentenone derivatives and of heterocycles (see also sections 1.13.3 and 4.6.4). 

Conventional synthetic schemes to produce 1,6-disubstituted products, e.g. reaction of a - with d -synthons, are largely unsuccessful. An exception is the following reaction, which provides a useful alternative when Michael type additions fail, e. g., at angular or other tertiary carbon atoms. In such cases the addition of allylsilanes catalyzed by titanium tetrachloride, the Sakurai reaction, is most appropriate (A. Hosomi, 1977). Isomerization of the double bond with bis(benzonitrile-N)dichloropalladium gives the y-double bond in excellent yield. Subsequent ozonolysis provides a pathway to 1,4-dicarbonyl compounds. Thus 1,6-, 1,5- and 1,4-difunctional compounds are accessible by this reaction. 

Heterocyde syntheses are often possible from difunctional open-chain precursors, including olefins as 1,2-difunctional reagents, and an appropiate nucleophile or electrophile containing one or more hetero atoms. The choice of the open-chain precursor is usually dictated by the longest carbon chain within the heterocyde to be synthesized. 

Oxidation of olefins and dienes provides the classic means for syntheses of 1,2- and 1,4-difunctional carbon compounds. The related cleavage of cyclohexene rings to produce 1,6-dioxo compounds has already been discussed in section 1.14. Many regio- and stereoselective oxidations have been developed within the enormously productive field of steroid syntheses. Our examples for regio- and stereoselective C C double bond oxidations as well as the examples for C C double bond cleavages (see p. 87f.) are largely selected from this area. 

Most syntheses of nitrogen heterocycles involve substitution and/or condensation reactions of nitrogen nucleophiles with difunctional halides or carbonyl compounds. Common nitrogen reagents are ... 

The longest carbon chain within a heterocycle indicates possible open-chain precursors. We use this chain as a basis to classify heterocycles as 1,2- to 1,6-difunctional systems. 

Since 1,2- to 1,6-difunctional opengeneral procedures (see chapter 1), it is useful to consider them as possible starting materials for syntheses of three- to seven-membered heterocycies 1,2-heterocycles can be made from 1,2-difunctional compounds, e.g. olefins or dibromides 1,3-difunctional compounds, e.g. 1,3-dibromides or 1,3-dioxo compounds, can be converted into 1,3-heterocycles etc. 

Regioselectivity becomes important, if unsymmetric difunctional nitrogen components are used. In such cases two different reactions of the nitrogen nucleophile with the open-chain educt may be possible, one of which must be faster than the other. Hydrazone formation, for example, occurs more readily than hydrazinoLysis of an ester. In the second example, on the other hand, the amide is formed very rapidly from the acyl chloride, and only one cyclization product is observed. 

Many successful regioselective syntheses of heterocydes, however, are more complex than the examples given so far. They employ condensation of two different carbonyl or halide compounds with one nitrogen base or the condensation of an amino ketone with a second difunctional compound. Such reactions cannot be rationalized in a simple way, and the literature must be consulted. 

The usefulness of the Knorr synthesis arises from the fact that 1,3-dioxo compounds and a-aminoketones are much more easily accessible in large quantities than rational 1,4-difunctional precursors. Such practical syntheses are known for several important hetero-cycles. They are usually limited to certain substitution patterns of the target molecules. 

In antithetical analyses of carbon skeletons the synthon approach described in chapter I is used in the reverse order, e.g. 1,3-difunctional target molecules are "transformed" by imaginary retro-aldol type reactions, cyclohexene derivatives by imaginary relro-Diels-Alder reactions. 

---