Electronic Effect of Substituents on Phenoxyacetic Acids 435 



data in Table 1. Aberg (1,2) in carefully controlled experiments 

 has also observed that the substitution of hydrogen by methyl results 

 in lowered activity in a variety of molecules. The deactivating effect 

 of methyl and methoxy groups on phenylacetic acids is also clear in 

 the data of Melnikov ct al. (24). The fact that the activating ability 

 of the halogens follows their increased inductive effect ( — I effect) 

 also points against attack by X-|-. Aberg (1) has observed the same 

 order of activity for the 4-substituted halophenoxyacetic acids which 

 we report in Table 1. He pointed out that the decrease in activity in 

 going from F to I might be due to the increasing atomic radius of 

 each substituent. This hypothesis would not explain the great dif- 

 ference in going from H to F in the 4 position, nor would this ex- 

 plain the difference between CI or Br and CH3, or CH;. and CF3. 

 3-Trifluoromethylphenoxyacetic acid is the most active biologically of 

 all the monosubstituted phenoxyacetic acids which we have tested. Ftn- 

 ther evidence against attack by X-|- has been fotnid by Fukui et al. (10) 

 for the benzoic acids. They have shown that there seems to be no cor- 

 relation between pi electron distribution on the benzene ring and 

 attack by an electrophilic reagent. 



It should be noted that both in the phenyl thioglycolic acids (17) 

 and in the phenylacetic acids (24) the order for activity on substitution 

 into the 4 position is Cl>Br>CH3>I, with the latter two substitu- 

 ents giving compounds of very low activity. The only groups which 

 consistently increase activity in rings other than the benzene ring in 

 the benzoic acids are F, Br, and CI. (It seems very likely that CF3 will 

 be added to this group when more such derivatives are studied.) The 

 nitro group sometimes gives weakly active compounds, but amino and 

 hydroxyl groups invariably give low activity. This has usually been at- 

 tributed to the fact that polar groups would greatly redtice the lipo- 

 philic character of the ring (37). 



Since attack by electrophilic reagents seems unlikely, we have ad- 

 vocated attack by an electron-rich group. Of the two possibilities X: 

 seems more likely than X-. In a recent critical review (3) of the na- 

 ture of free radical attack on aromatic nuclei, Augood and Williams 

 point out that almost all substitution on benzene results in an in- 

 creased rate of attack by radicals at points ortho and para to the sub- 

 stituent. The only groups which do not have total rate factors greater 

 than 1 are isopropyl, tert-butyl and trifluoromethyl. The latter has 

 a value of 0.99, which means essentially no change in over-all reaction 

 from benzene itself. Methyl and methoxyl groups are quite strongly 

 activating. Again this is at odds with the biological activating effects 

 of these groups in phenoxyacetic and phenylacetic acids. More inter- 

 esting is the fact that the order for activation by halogen on benzene 

 for radical attack is I>Br>Cl>F (3). This of course is just opposite 

 to the order found for the biological activating effect of the substi- 



