Molecule, acetone left/right

Fig.3.1.9 (a) The adsorption-desorption isotherm (circles, right axis) and the self-diffusion coefficients D (triangles, left axis) for cyclohexane in porous silicon with 3.6-nm pore diameter as a function of the relative vapor pressure z = P/PS1 where Ps is the saturated vapor pressure, (b) The self-diffusion coefficients D for acetone (squares) and cyclohexane (triangles) as a function of the concentration 0 of molecules in pores measured on the adsorption (open symbols) and the desorption (filled symbols) branches. 

Typical carbonyl compounds behave as weak C-H acids. If, for example, one generates the enolate anion of acetone using even a relatively strong conventional base, such as sodium ethoxide, the resultant equilibrium will be shifted far to the left (Scheme 2.21). Carbonyl compounds themselves, as we soon will see, are active electrophiles due to the presence of the partially positive carbonyl carbon. Hence, the nucleophilic enolate generated in the above system can react with non-ionized acetone molecules abundantly present in this equilibrium. This reaction is the well-known aldol condensation (Scheme 2.21). Although useful in its own right, its ease of occurrence creates serious obstacles for the use of an in situ generated enolate anion as a nucleophile in reactions with other electrophiles. 

We have two options from which to choose Michael addition of the enolate of acetone to 2-phenylpropenal (above, left), and Michael addition of the enolate of 2-phenylacetaldehyde to 3-buten-2-one (above, right). Which is better Michael additions are normally carried out under basic conditions, and we know (Section 18-5) that aldehydes undergo aldol condensation with base much more readily than do sin5)le ketones. Thus, the second option above risks interference from the aldol reaction between two molecules of 2-phenylacetaldehyde. The first option is better because the enolate of acetone will undergo Michael addition to the aldehyde much more easily than aldol condensation with itself. Thus, we have all the pieces we need to put the synthesis together in the forward direction—it is a Robinson annulation ... 

---