The cyclization of a 9,12-seco-11-nor keto acid in the hecogenin series in the presence of benzylamine and anhydrous ammonia is described. The benzylamine reaction follows an unusual course whereas the ammonia reaction proceeds normally t o provide a n en01 lactam. The sequence represents the first synthesis of an 11-aza derivative in the steroidal sapogenin series.In previous papers of this series (1, 2) we reported on the introduction of a nitrogen function into ring B of the steroid skeleton, and this work provided some 6-aza steroidal analogues. In extending these studies, we considered the synthesis of ring C aza steroids since these steroidal derivatives with an unexpanded C-ring system were not known (3 and references cited therein pertaining t o 12a-aza-C-homo steroids). However, a very recent article by Engel and Rakhit (4) on the first synthesis of 11-aza steroids possessing the normal skeleton prompts us to report some of our results in this area.I n considering a synthetic sequence, it was obvious that an extension of our work in -ring B should provide an entry into this series. I t was therefore necessary t o prepare the appropriate 9,12-seco steroidal intermediates which, in turn, could be cyclized in the presence of an amine t o provide the 11-aza steroid. The most direct approach to the desired 9,12-seco steroids involved ozonolysis of a 9,ll-unsaturated 12-keto steroid since this type of reaction had been successfully employed by the present authors (1, 2) and by Jacobs and Brownfield (5) in the synthesis of ring B seco derivatives and by Engel and co-workers (4, 6, 13) in the synthesis of sorne9,12-seco steroids. An attractive starting material for the ring-opening reaction was 9,11-dehydrohecogenin acetate, whieh was readily available from hecogenin acetate (7). When we subjected this substance t o ozonolysis we obtained in low yield, from the acidic fraction of the reaction product, a crystalline, high-melting compound which exhibited the expected properties of the desired 9,12-seco keto acid (I). The infrared spectrum indicated two bands in the carbonyl region (5.80 p, 5.93 p) and still retained the characteristic spiroketal bands (8, 9), indicating that the spiroketal side chain remained unaffected under these conditions. In spite of numerous experiments, we were not able t o improve the yield in this reaction and we turned to consider another approach to this problem. An attractive alternative was the use of the Lemieux reagent (10, 11) since, just a t this time, Edward (12) published his work which indicated that this reagent was superior t o ozone in the instance of cholest-4-en-3-one. Consequently, 9,ll-dehydrohecogenin acetate was treated with this reagent and a crystalline acidic product was obtained from this reaction. This substance was shown to be identical with the ozonolysis product (I). In addition, another crystalline compound was obtained from the acidic fraction of the reaction product which on the basis of its infrared spectrum (5.68 cl) is tentatively assigned an en...