Dissociation Extraction and so called Dissociative Extractive Crystallization

In the opinion of KiloMentor, the most under-appreciated and underutilized method of separation that can be used in chemical process development is dissociation extraction. Dissociation extraction crystallization and dissociation leaching are two powerful variations on this same method. Fractional crystallization, which is hardly ever used, in practice is frequently mentioned while this method, which is more frequently applicable, is effectively unknown.

Dissociation Extraction

Dissociation extraction is a separation method that can be applied to a mixture of two (or more) chemical substances that are each ionizable in solution and have at least a small difference in pKas. Following the method, the mixture is partitioned between two liquid phases and then partially neutralized with a reagent so that the partition coefficients of the components of the mixture are further favorably enhanced. Most optimally the moles of the neutralizing agent are equal to the moles of the more reactive component of the binary mixture.  This is best made clear with an example.

A mixture of 3- and 4- picoline (3-methylpyridine and 4-methylpyridine) cannot be separated by distillation because the boiling points differ by only a fraction of a degree and they have similar solubilities in common solvents. Their dissociation constants, however, are 4.54 and 10.62 X 10^-9 respectively. Contacting it with a stoichiometric deficiency of an aqueous acid solution separates a mixture. The two picolines compete for the available acid, which preferentially reacts with the base with the higher dissociation constant to form a salt that prefers the aqueous phase. The weaker base remains unassociated and is preferentially taken into an organic solvent. What is not obvious at all is that the separation efficiency is a sensitive function of the organic solvent used with the aqueous acid.  It is quite remarkable how a small difference in pKas and an optimized selection of an organic solvent can produce surprisingly complete separations.

This technique is applicable to the concentration of a trace impurity in a substantially pure single substance.  The points of structural difference between the two molecules may be quite remote from the acidic or basic functional group upon which the dissociation depends so long as the effect of that difference can be transmitted electronically or sterically to the ionizing function. Since pharmaceutical substances are most often salt-forming species, this technique provides a powerful method to identify an impurity that has only minor differences from the drug substance which result in a close retention time by HPLC.

Selection of the Solvent for Dissociation Extraction

Organic solvents can be classified according to their ability to interact with carboxylic, phenolic, and basic substances. Aliphatic hydrocarbons, such as heptane, cyclohexane, etc. represent the inert solvents with negligible tendency to interact with CO2H, OH, and NH2 groups. These are followed by aromatic hydrocarbons, halogenated hydrocarbons, ethers and ketones, esters, and finally alcohols, in the order of increasing tendency to interact with the solutes. The knowledge of these particular interactions can help in the selection of the proper organic solvent with which to partner the dissociative extraction operation.

Gaikar and Sharma formulated guidelines on the basis of the thermodynamic considerations of solute-solvent interactions, solute-solute interactions, and steric hindrance to functional group solvation as follows:

(i)                Use an inert solvent if the more easily ionized component has a relatively free aquophilic ionizing group and

(ii)             Use a polar solvent if the less easily ionized component has a relatively free aquophilic group.

Examples of polar solvents that associate well with free ionizable groups are di-butyl ether and n-octanol.

Being chemical engineers, Gaikar and Sharma [V.G. Gaikar and M.M. Sharma, Separations through Reactions and Other Novel Strategies, Separation and Purification Methods, 18(2) 111 (1989).]  devote a considerable part of their article to means for reducing the cost of dissociative extraction. If this turns out to be important in your application refer to the original article.  For high-value items like pharmaceuticals and pharmaceutical intermediates, the cost of the chemicals required for dissociative extraction is very reasonable for the simplicity, power, and ruggedness that the method promises.

Dissociative Extractive Crystallization

What is called dissociative extractive crystallization in the literature is not extraction at all. What is happening is that a mixture of similar compounds is competing for an insufficiency of a salt-forming reagent and the winning substance is precipitating. In dissociative extraction, the winning substance is extracted into an aqueous phase.

The Effect and the Utility of the Hydrolysis of Weak Salts

On page 196-197 of an old chemistry laboratory text [Avery Adrian Morton, Laboratory Technique in Organic Chemistry, McGraw-Hill Book Company, Inc. New York and London 1938 ], the degree of the partitioning of thymol which has been exactly neutralized with aqueous sodium hydroxide was studied as a function of the organic solvent with which it is equilibrated.  The amounts of thymol extracted into the organic phase were reported to be: ether 88%; benzene 38%; carbon tetrachloride 25%; and petroleum ether 22%.  Author Morton explains that these results arise from the partial hydrolysis of the sodium thymolate back into thymol and sodium hydroxide and the extent to which extraction of the free thymol into the organic layer is favorable.

In another interesting table in the same section, Morton documents that as the structure of the sodium phenolate becomes more hydrophobic and more particularly as the steric hindrance to selective hydrogen bond solvation of the phenol becomes greater, the percentage extracted into ether increases. Thus for phenol, it is just 7.5% extracted, for p-cresol 13.3%, for o-cresol 20.8%,  for p-propylphenol 28.7%, for o-propylphenol 68.5%,  and for di o-allylphenol 91.7%.

he also notes that as the excess of base over the amount required for the neutralization increases, the tendency for solvolytic extraction is suppressed.

Taken with teaching concerning dissociative extraction, this understanding of partial solvolysis can help determine how much of the non-stoichiometric amount required for neutralization, we need to get an optimal result. For example, if we are using a strong solvent like amyl alcohol in a dissociative extraction there may be a need to use more than the equimolar reagent for the neutralization of the more reactive component if we want to hold it in the aqueous layer against the power of the hydrolysis effect.