Predicting the substitution reactions of carboxylic acid derivatives

Amides can be formed from esters by an attacking amine, but esters cannot be form from amides by attack with alcohol.

Both cases would form a tetrahedral intermediate, but the RO^- would always be preferentially kicked off:

There are other factors than leaving group ability which decide whether a substitution takes place, such as how well the intermediate forms. The amide is so bad an electrophile that the immediate is very unlikely to form.

So for a sucessful attack we require:
1. A good enough electrophile
2. A good enough nucleophile
3. A leaving group on the defending molecule which is better at leaving than the attacking nucleophile.

Leaving group ability is a good guide to how good a nucleophile something is - good nucleophiles are bad leaving groups. Hence a good nucleophile will have a high pK_aH. Another way to think about it is that something which wants to form a bond to hydrogen will be likely to want to form a bond to carbon.

A reaction particularly suited to substitution will have more relaxed reaction conditions. Eg. Will react faster, at a lower temperature, and without needing a catalyst.

For example, below are four carboxylic acid derivatives with different leaving groups, each attacked by H₂O. Try to relate the reaction conditions to the table above:

The difficulty of attacking an amide is due to n -> pi* overlap. This can be depicted as resonance:

Or spacial overlap of the filled n and empty π* orbital:

Or as an MO diagram:

Now we can see why an amide is such a bad nucleophile - because resonance raises the LUMO of the molecule.

This MO approach also gives another explanation for why an amide is a worst base than an amine - a lower lone pair energy makes it less favorable to donate them into a proton's empty 1s orbital