Mercuric acetate reacts with betulin (Ia) in chlorofonn-acetic acid to form the cyclic ether, 13P,28-epoxylup-20(3O)-en-3~-ol (11; R = H). Under the same conditions betulin diacetate (Id) forms the non-conjugated diene, 3@,28-diacetoxylupa-12,20(30)-diene (XIV; R = Ac). Comparable reactions are undergone by acetylbetulinic acid (Ib) and methyl acetylbetulinate (Ic) which form respectively the y-lactone 3P-acetoxylup-20(30)-en-28,13@-olide (IX), and the non-conjugated diene, methyl-3Pacetoxylupa-12,20 (30) -dien-28-oate (X) . MERCURIC ACETATE has frequently been employed as a dehydrogenating agent in the steroid field,l and in the references cited the reaction leads to the introduction of an ethylenic bond in conjugation with an existing double bond. Biedebach2 extended the reaction to triterpenes and in general found that whereas a-and p-amyrin derivatives did not react with mercuric acetate in chloroform-acetic acid, derivatives of a-lupene, viz. , betulin and lupeol, formed dehydro-compounds of unknown constitution, and it is the purpose of the present work to describe the nature of these compounds.Biedebach observed that the reactions of betulin and lupeol with mercuric acetate were not dependent on the presence of the alcohol group since their esters were also dehydrogenated. He established that the dehydrogenation was associated with the presence of the olefinic bond in these compounds, since the dihydro-compounds did not react with mercuric acetate.recently described the formation of a cyclic ether by reaction of the 78-unsaturated alcohol, bicyclo[2,2,l]hept-5-en-2aylmethanol, with mercuric acetate, and Brook, Rodgman, and Wright describe the similar formation of a cyclic ether from the $-unsaturated alcohol, 2,6-dimethylhept-5en-2-01.We have repeated Biedebach's experiments and find that betulin (Ia) with mercuric acetate forms a cyclic ether (11; R = H) which contains a vinylidene group (infrared bands at 1630 and 896 cm.-l). However, since betulin may isomerise5 in presence of acid with enlargement of ring E, there is a possibility that under the conditions of the experiment, in which acetic acid is present, ring enlargement may occur to give a compound (111; R" = CH,), which also contains a vinylidene group. That the cyclic ether is correctly represented by ( 11) is shown in the sequel.The nuclear magnetic resonance spectrum of the ether (11; R = Ac) in carbon tetrachloride shows the following features: weak bands at +63 and +89 cycles/sec. due to C*CH2*0 protons, and peaks at +140 and +150 cycles/sec. due to M e G protons. There is no evidence of a 7 cycles/sec. doublet peak in the Me*C< region which would be expected from the Me*CH< methyl group in the alternative structure (111; R" = CH,).Oxidation of the acetoxy-ether (11; R = Ac) with chromic acid or ozone gave the methyl ketone (IV; R = Ac, R' = H,) which was hydrolysed and converted into the benzoate (IV; R = Bz, R' = H,). The benzoate showed nuclear magnetic resonance bands at -102 and -85 cycles/sec. due to aromatic ring protons, and a peak at...
