Pentachlorobenzoic acid

Submitted by D. E. Pearson and Dorotha Cowan 1 Checked by Virgil Boekelheide and Fred G. H. Lee 

Magnesium, turnings (39 g-, 1.6 g. atoms) and hexachloroben-zene (142.4 g., 0.5 mole, m.p. 228-229°) in 1 1. of dry ether are brought to gentle reflux in a 3-1. three-necked flask heated by a Glascol mantle at 20 volts (Note 1). Ethylene bromide (188 g., 1.0 mole) in 200 ml. of dry benzene is added through a Hersh- 

The atomic proportions of magnesium are not related to the mole quantity of hexachlorobenzene in this or any other entrainment reaction. The excess magnesium (1.1 g. atoms in this case) is used to react with ethylene bromide and leave 0.5 g. atom of dean-surfaced magnesium. Ordinarily 1 mole of entrainment reagent is used per mole of inert halide, but for this preparation 2 moles of entrainment reagent per mole of halide gives a better yield. 

Little attention is needed provided that the capillary tube is fitted properly. The capillary tube of the Hershberg dropping 

A T-tube in the carbon dioxide stream serves to bypass the gas if its rate of addition is too rapid. Also, the T-tube is large enough to permit the insertion of a plunger to dislodge particles within the mouth of the tube. 

---

Pentachlorobenzoic acid has been prepared by oxidation of pentachlorotoluene with nitric acid and mercury, by oxidation of pentachlorobenzaldehyde by potassium permanganate,6 and by chlorination of tetrachlorophthalyl chloride 6 and of dichloro-benzoic acids.7 Pentachlorobenzoic acid recently has been prepared by the exhaustive chlorination of benzoic acid in sulfuric acid containing iodine.8 The present procedure has been adapted from that of Pearson, Cowan, and Beckler.9... 

Although pentacblorophenylmagnesium chloride can be made in tetrahydrofuran without the use of the entrainment method, the Grignard reagent in this solvent does not react with carbon dioxide to give pentachlorobenzoic acid in good yield.10... 

When, instead of sulphuric acid, oleum is used, path 1 seems to be the only one available, since the product is the chloroketone [98], along with some pentachlorobenzoic acid (Ballester et al., 1986b). The latter cannot result from the acid [99] since under the reaction conditions, the latter gives pentachlorobenzaldehyde almost quantitatively. The chloroketone [98] does give pentachlorobenzoic acid in an excellent yield. Possibly, path 2 is not available here since the bulkiness of the HSOj species precludes its addition to the a-carbon, which is significantly shielded by its neighbouring o-chlorines (100). 

---