We are attempting to synthesize ornithine δ-isonicotinate, in which the pyridinium nitrogen is connected to carbon-5 of a group derived from pentanoic acid. Under our conditions (1 eq. of trifluoroacetic acid and 20 equivalents of phenol), I expect our product to have two positive charges and for both carboxylic acids to be neutral. We would like to extract away the majority of the phenol before further purification by ion exchange chromatography. We have been searching through the literature for a good choice in extraction solvents. The solvent should remove the phenol but leave our product in the aqueous phase. We found one paper (Jabrou, Research Journal of Chemical Sciences, Vol. 2(4), 1-12, April (2012)) that recommended 2-butatone and isobutyl acetate as good solvents (http://www.isca.in/rjcs/Archives/vol2/i4/1.ISCA-RJCS-2012-001%20Done.pdf
). We also found a patent (US4160111 A) that mostly discussed fractional distillation but also discussed liquid-liquid extractions. The author (DF Strahorn) of this patent suggested mixtures of 2-butanone and benzene or alkylated benzenes were more effective than either solvent alone (http://www.google.com/patents/US4160111
"In a series of runs which were carried out under ambient conditions, that is, at a temperature of about 22° C. and a pressure of 1 atmosphere, the phenol partition coefficients for water and a series of ketone-modified hydrocarbon extracting phases were determined. The extracting phases used in these examples were benzene and a benzene solution containing a range of 2-butanone contents. The results from these runs are shown in FIG. 1. These results illustrate that while benzene or 2-butanone individually exhibit phenol partition coefficients of the relatively low order, for example, of less than about 3 or 5, mixtures of benzene and 2-butanone exhibit, in the partition of phenol between water and the organic extracting medium, partition coefficients which are exceptional. From FIG. 1 it is notable that the addition of but a minor amount of 2-butanone, for example, as little as 10 weight percent, to benzene about doubles the partition coefficient. The coefficient for a 50/50 mixture, relative to that for benzene alone, is larger by about a factor of 20! Similar and advantageous results obtained when lower alkyl substituted benzenes as described above are used in place of benzene and when other ketones as defined above are used in place of 2-butanone. These organic mixtures are also especially suitable for use as the organic mixture which is introduced to the fractionating column of the process herein."
Obviously we would avoid benzene. Right now I am leaning toward 2-butanone on the basis of simplicity, and maybe trying 40/60 2-butanone/toluene if a great deal of phenol is left in the aqueous phase when 2-butanone is used alone.