Using the Zinin Reaction to Prepare Pure Aromatic Regioisomers

In an effort to choose from among the synthetic reaction schemes created by retrosynthetic analysis,  it is often an unconscious assumption that all of the same functional groups in an intermediate react with about the same ease. Thus, it is assumed that all instances of aldehyde, ketone, nitrile, nitro etc. display commensurate reactivity.  The only common exception to this is that sometimes a distinction is drawn between the same functionality attached to an aryl as opposed to an alkyl group. The combined electronic and steric effects in the vicinity of a particular functional group are considered inconsequential to rough first-order planning.  
In the case of a structure with two different functional groups of the same class,  ie two ketones, this simplification most often is a good one because to obtain a quantitatively selective reaction in which only one of two functional group of the same class reacts, the more reactive group must transform more than 100 times faster than the other. This is true because when the reaction is 99% complete, the final 1% of starting material must be more reactive than the 99% of product, which could react further at that second less responsive functional group. Thus, if one reactivity is not one hundredfold greater than the other, the reaction cannot be expected to be quantitative and for practical purposes selective (kinetic control assumed here).

This reasoning explains the widespread use of protecting groups when a substrate contains two functionalities of the same type and the reaction of one only is required. However, there are documented cases where the same functional groups, in two different environments, but in the same molecule, react at usefully different rates.  When the two nominally identical functional groups are in two separate molecules, this can be the basis for separating the mixture by reaction.  When the competing reactivity is between the same groups within the same substrate, the reaction becomes a practical process step that leads to even more complete distinguishing of the functional entities. The reaction is an intramolecular competition.

Suppose one is presented with a mixture of regioisomers: 4-methylethylbenzonitrile and 2-methylethylbenzonitrile.  Although each isomer contains a nominal nitrile, the nitrile groups are not equally reactive. For example, it is reported that ortho substituted aryl nitriles do not readily for imidates by reaction with ethanol and anhydrous hydrogen chloride. We can confidently predict that 2-methylethylbenzonitrile (2-isopropyl benzonitrile) will be unreactive. 

The low reactivity of hindered ortho-substituted aromatic compounds is fairly well known. A less known instance is the application of the sulfide reduction useful for the preparation of isomerically pure aromatic nitro-compounds and anilines.

Thomas R. Nickson wrote a research article[ J. Org. Chem. 1986, 51, 3903-3904] that taught. a specific case. When 3-trifluoromethyl toluene is nitrated the compound formed in the largest amount is the 2-nitro-3-trifluoromethyl-toluene.  It could be separated and purified from the other region-isomers because it was the only isomer than did not undergo the Zenin reduction to an aniline by treatment with sodium sulfide and sulfur. Dr. Nickson however also taught that 3-methyl benzaldehyde and 3-methyl benzoic acid both nitrated preferentially in the 2 position.  From my own experience,  I know that the compound 3,4-dichlorobenzaldehyde nitrates preferentially in the 2-position.  It is possible that all 1,3-substituted compounds with one electron-withdrawing group and one electron-donating group, nitrate preferentially in the 2 position. All of these may be separable by their failure to react in the Zenin reduction!  Nickson tells us that one electron-withdrawing group besides the nitro itself  is advantageous to achieve a fast reduction. Even so, he was able to make 2-nitro m-xylene and separate it cleanly, although in poor yield, by reducing the other isomers, but this reduction went slowly. Also, when the two substituents were both ortho-para directing the yield of the 2 isomer is much lower (10%) as in the case of m-xylene. It is not clear whether the reaction scheme would work with two deactivating groups meta to each other.

Thus, the same general approach could be used to make a variety of polysubstituted compounds with specific substitution patterns. The nitro group has the advantage that it can be reduced to primary amine and then removed completely or converted to a range of other functional groups by diazotization. 1,2,3-substituted aromatics would be available. Similarly, it could be used to prepare a 1,3 disubstituted isomer without adding any substituent.  Suppose one treats toluene with sufficient brominating agent to dibrominated predominantly and then separated out only the dibromides by distillation.  There could be five compounds: the 2,3; 2,4-; 2,5, 2,6-- and 3,4- dibromides. If one mono-nitrates this mixture, only three of these compounds could give a mono nitro compound in which the nitro would have two ortho substituents. If without attempting purification, one uses the Zinin reduction, all the nitro compounds with zero or one ortho substituent would be reduced to amines. Extracting an organic solution of the reaction products with aqueous acid can be expected to remove these substances from the organic phase. Four specific compounds: 3,4-dibromo-2-nitrotoluene; 2,4-dibromo-3-nitro-toluene; 3,6-dibromo-2-nitro-toluene; and 3,5-dibromo-4-nitrotoluene.  Of these only two are likely to be present in significant amounts, because of the directing influences of methyl and bromine on the second bromination.  These two are: 3,4-dibromo-2-nitro-toluene and  2,4-dibromo-3-nitro-toluene. After reduction to the corresponding anilines, there will be a predictable difference in the pKas of the two amines which will allow their separation. 2,4-dibromo-3-amino-toluene will have its amine flanked on each side by the sterically bulky bromines that will also be electron-withdrawing, making it a very weak base. The remaining compound:  3,4-dibromo-2-amino-toluene will be not as weak a base, because it will have the electron-donating methyl in one ortho position.  Consequently protonation will not be as severely inhibited.

The compound 3,4-dibromo-2-amino-toluene, more properly named 2,3-dibromo-6-methylaniline is in a position to be diazotized and reduced to give back 3,4-dibromotoluene.