MjS
B. H. ALEXANDER, K. J. DIMLERAND C. L. X~EHLTRETTER
[CONTRIBUTION FROM
THE
Vol. 73
STARCH AXD DEXTROSE DIVISION,SORTHERN REGIONAL RESEARCH LABORATORP~]
D-Galactosan had [ c Y ] ~ ~1-73.6' The D-galactosan < 1,4 > a < 1,6> was unusually difficult distillation was complete (one hour). The yield of brown, viscous pyrolysate was usually 19-22 g. [54 g. of pyrolysate to methylate with methyl iodide and silver oxide. Later experiments, however, showed that it was completely will give about 3 g. of pure D-galactosan< 1,4>a< 1,6>]. methylated after two treatments with sodium hydroxide and Isolation of the Crude D-Galactosan< 1,4>0( .Fifty-four grams of pyrolysate, collected from several runs, dimethyl sulfate. 2,3,5-Tr~ethyl-~-galactonolactone.-Onegram of triwas acetonated by the addition of 450 ml. of anhydrous acetone and 45 g. of anhydrous copper sulfate. The reac- methyl-D-galactosan was hydrolyzed by heating for six tion flask was shaken for 15 hours a t room temperature. hours a t 98' in 0.5 N hydrochloric acid. The specific roDuring this period the pyrolysate, initially incompletely tation of the resulting trimethyl galactose was calculated soluble in the acetone, went completely into solution. The to be -8" on the basis of the optical activity of the hydrolysis copper sulfate was filtered from the reaction mixture and solution. Luckett and Smith' reported an equilibrium [ a ]l s ~ -5' for the sirupy 2,3,5-trimethyl-~-galactose in the combined filtrate and washings were shaken with 40 g. of anhydrous copper sulfate for 35 hours. The reaction mix- water. The trimethyl-&galactose was converted t o the trimethylture was filtered and the filtrate concentrated in vacuo to a D-galactonolactone (0.53 9.) by bromine oxidation and subthin sirup. Chloroform, 300 ml., was added to the thin sirup to dis- limation. The crystalline 2,3,5-trimethyl-~-galactonolacsolve the undesired 3,4-isopropylidene-~-galactosan.The tone, after recrystallization from a mixture of ethanol and [cY]*~D mixture was cooled to 5" and the residue remaining after petroleum ether (b.p. 63-70'), melted a t 91-92'; -35' (after one-half hour; c 0.6, H?O). decantation of the chloroform was dissolved in cold water and thoroughly extracted with chloroform. The combined Anal. Calcd. for C9&06: OCHI, 42.3. Found: OCHI, water layers, after filtering, were concentrated in vacuo to a 42.1. light brown sirup which contained the unsubstituted DLuckett and Smith' reported m.p 90" and [CY]'~D -37" galactosan < 1,4>a< 1,6 > . (initial) for 2,3,5-trimethyl-~-ga~actonolactone. A mixture 2,3,5-Triacetyl-~-gdactosan< 1,4>a< 1,6> .-The sirup of a sample of their compound with our product melted a t was dried by the repeated concentration in vacuo of its solu- 90'. tion in pyridine. Acetic anhyride, 120 ml., was added to a 2,3,5-Trimethyl-~-galactonamide.-Treatment of the solution of the dry sirup in pyridine a t 0 " . After 24 hours crystalline lactone (0.12 g.) with methanolic ammonia for a t room temperature the acetylation mixture was concen- 24 hours a t 5' readily gave the 2,3,5-trimethyl-~-galactontrated in vacuo. The nearly black sirup (31.5 g.), was dis- amide, which crystallized on removal of the solvent; m.p. tilled a t or below 0.1 mm. pressure. The distillate was col- 156-157.5°. After recrystallization from absolute ethanol, lected in three fractions, 0.6 g., 2.3 g. and 14.7 g., over the the amide melted a t 162-163'; [a]*'D+5" ( c 0.5, HgO). temperature ranges 70-90', 90-132" and 133-140', reAnal. Calcd. for C ~ H I Q O ~ N C,: 45.6; H, 8.0; pi, 5.9; spectively. The latter two fractions, which crystallized on standing, were recrystallized from three parts isopropyl OCH,. 39.2. Found: C. 45.4; H. 8.0; X. 6.0; OCH3.39.1. alcohol. The product, 6.5 g., was 2,3,5-triacetyl-~-galacto- This melting point is higher than that (152") reported by Luckett and Smith' for the amide; however, the optical rosan< 1,4>0(< 1 , 6 > . Anal. Calcd. for CleHl~O~: acetyl, 44.8. Found: acetyl, tations are essentially the same, ca. $3" compared t o +5' for our amide. A mixture of their amide and our product 44.5. had an intermediate melting point of 156-158'. The pure triacetate melts a t 79-80'; [a]"D 144.9' Resistance of D-Gdactosan< 1,4>a< 1,6> to Periodic (c 1, CHC13). The melting point of the triacetate was un- Acid and Lead Tetraacetate Oxidation.-Samples of Dchanged by admixture with a sample of Hann and Hudson's galactosan< l , 4 > a < 1,6> were treated with periodic acid, triacetate (m.p. 79-80'). sodium periodate and lead tetraacetate essentially as deD-Gdactosan< 1,4>0(< 1,6> .-Catalytic deacetylation scribed by Dimler, et ~ 1 Analysis . ~ of aliquots of the oxidaof 4 g. of 2,3,5-triacetyl-~-galactosan< 1,4>a< 1,6> with tion mixtures showed that no reaction occurred over a period barium methylate gave 2 g. of D-galactosan (87% of seven days. of theory). Approximately 0.5 g. of the first crop of crystals, melting a t 183-184", was recrystallized from 10 parts absoAcknowledgment.-Grateful acknowledgment lute ethanol to obtain D-galactosana< 1,6>; m.p. is made t o Prof. C. S. Hudson for his encourage183-184', [a]"D +56' ( C 1, H@). ment and interest in the planning and progress .4naZ. Calcd. for C B H L O OC, ~ : 44.4; H , 6.2. Found: of this work, and to Dr. F. Smith for samples of C, 44.4; H , 6.0. acid. This D-galactosan preparation melts higher than the D- derivatives of 2,3,5-trimethyl-~-galactonic galactosan reported by Hann and Hudson, apparently be- Thanks are also expressed t o Mr. C. H. Van Etten cause of more complete purification when prepared via the and Mrs. Mary Wiele for the microanalyses. triacetate. The melting point of a mixture of this D-galacRECEIVED MAY19, 1951 tosan with a sample of D-galactosan (m.p. 174-177') fur- PEORIA,ILL.
+-
( 8 ) The pyrolysis was carried out according to the directions in an unpublished report "1,evoglucosan" by I. A. Wolff and D. W. Olds at this I ~ l ~ o r a t o r y .
(9) F. J. Bates and Associates, "Polarimetry, Saccharimetry and the Sugars,'' IT. S. Government Printing Office, Washington, 1942, pp. .r,o(i-~5ll7,
