Kiliani's classic reduction of cr-glucoheptonic (D-gluco-D-guloheptonic) lactone led to small, variable yields of n-heptanoic acid and varying amounts of the lactone of a hydroxy-11-heptar~oic acid. A repetition of this work showed that the combined yields of these two products reached a maximum near 70% of theory when the reduction with constant-boiling hydriodic acid (b.p. 127" C.) and phosphorus was limited to two hours. Under these conditions up to Y/lOths of the product consisted of the lactone. After purification through the hydrazide, m.p.8'3" C., the fragrant smelli~~g lactone boiled a t 61" to 62' C. a t 10 mm. pressure, had a density of 0.9948, and a refractive index of nD, 1.4405, both a t 20' C. The supposition that this lactone referred to 4-hydroxy-n-heptanoic acid was confirmed by preparing it from the known 4-keto-n-heptanoic acid, which in turn was synthesized by an improved method.The optimum time for reducing a n u~lcrystallized mixture of a-and 8-fructoheptonic acids under similar conditions was three to four hours, and the optimum yield of product was again near 70% of theory. This prodi~ct consisted almost exclusively of the lactone of 2-methyl-4-hydroxyhexanoic acid. After p~~rification through the crystalline hydrazide, 1n.p. 122' C., the lactone was recovered a s a fragrant oil boiling a t 48' to 4'3" C. a t 2 mnl. pressure, with density 0.0806 and refractive index n~ 1.4332 a t 25' C. The structure of this lactone was not in doubt.These lactones, prepared in good yield and with well defined hydrazides, were rnore readily characterized than the fully reduced heptanoic acids upon which ICiliani relied.
INTRODUCTIONWhile considering possible ways of locating carbonyl groups in oxystarches and oxycelluloses, it appeared that the method Kiliani used in 1885 to 1888 to solve the same problem for reducing sugars might be capable of development. In the case of fructose (structure I), for example, the addition of the elements of hydrogen cyanide gave the cyanohydrin which was hydrolyzed to the a,Pmixture of the corresponding fructoheptonic acids (11). Reduction of these acids with red phosphorus and hot hydriodic acid then yielded methyl-n-butyl acetic acid (III), and the result clearly showed that the ketone group occupied the second position in the fructose. Application of the same sequence to glucose first gave the lactone of glucoheptonic acid and then n-heptanoic acid. Mo~vry (29) reviewed the general applications of the cyanohydrin synthesis, and its value