M. L. WOLFROM, S . M. OLIN AND EVANF. EVANS
204
from gramine methiodide and potassium silver cyanide. The tertiary amine (gramine) reacts with malonic ester in the presence of a small amount of the sodium derivative of the ester to yield ethyl
Vol. 66
a-carbethoxy-~-(3-indole)-propionate.However, the yield of alkylation product obtained from the amine is inferior to that obtained from the amine methiodide. RECEIVED OCTOBER 2, 1943
U R ~ A N A , ILLINOIS
[CONTRIBUTION FROM THE CHEMICAL LABORATORY OF THEOHIQ STATE UNIVERSITY]
The Action of Diazomethane upon Acyclic Sugar Derivatives. VI.l
D-Sorbose2
BY M. L. WOLFROM, S. M. OLINAND EVANF. EVANS*
In continuation of our work upon the, action tion of L-sorbose. The acyclic nature of ketoof diazomethane upon acyclic sugar derivatives, L-sorbose pentaacetate was demonstrated by we have synthesized l-desoxy-l-diazo-keto-o-sor-Cramer and PacsulO and shortly later by Arrabose tetraacetate (IV) from D-xylonyl chloride goal1 We report herein a crystalline oxime of tetraacetate (111). D-Xylonic acid tetraacetate keto-L-sorbose pentaacetate. To our knowledge, (11) was prepared by the oxidation of aldehydo- this is the first nitrogen condensation product D-xylose tetraacetate (I) according to the pro- obtained for an acyclic or keto form of a ketohexose cedure reported by Major and Cook.' We pentaacetate, previous attempts1°*12having been describe an improved preparation of aldehydo- unsuccessful. We record also the D,L form of D-xylose tetraacetates from D-xylose diethyl keto-sorbose pentaacetate. mercaptal tetraacetate, utilizing the general Careful saponification of keto-D-sorbose pentatechniques of Wolfrom and RonigsbergSs The acetate led to the synthesis of D-sorbose. Preconstants of Major and Cook4 for I (m. p. 90- vious saponifications of the enantiomorphic de91', spec. rot. $22.5' in absolute chloroform, rivative have been recorded by Schlubach and for the form) have been verified. The rotation Vorwerk* and by Arragon." This synthesis of value differs appreciably from that previously D-sorbose from D-xylose compares with the synreporteds (m. p. 87-89', spec. rot. - 16' in chloro- thesis of L-sorbose reported by Gatzi and Reichform, for the D-form). D-Xylonyl chloride tetra- stein. l 3 These workers oxidized diethylideneacetate has apparently not been recorded, al- (levo)-sorbitol to diethylidene-L-xylonic acid. though Major and Cook7described the D,L form. From the crystalline acid chloride of the latter HC=O
COOH COCI I I HCOAC [O] HCOAC PC16 HCOACCHsN,
I
I --+I + AcOCH I I HCOAc HCOAc I I
AcOCH
CHzOAc
CHZOAC
I
CHNt I do HOAc
+ I - I - l HCOAc I I HCOAc AcOCH I I
AcOCH
CHzOAc
HCOAC
I
I1
I11
CHIOH
co
60
I
I
AcOCH 1
HCOAc
I
CHiOAc
v
IV
I
Ba(OH)*
HCOAc
CHiOAc
I
CHtOAc
HCOH
I
HOCH
I I
HCOH CHiOH
VI
CHa I CO
I I* AcOCH I
HCOAc
AcOCH
I I co I
HCOAc
CHa VI1
CHa
I CO
I
HOCH
I
HCOH
I co I
CHI
VI11
Treatment of the diazomethyl ketone IV with substance a s h p y diazomethyl ketone was obacetic acid yielded keto-D-sorbose pentaacetate, tained which on acid hydrolysis yielded crystalenantiomorphic with the substance prepared by line L-sorbose. D-Sorbose (earlier psuedo-tagaArragon* and, nearly simultaneously, by Schlu- tose, 1-sorbose) was first synthesized by Lobry bach and V0rwerk9 through the direct acetyla- de Bruyn and Alberda van Ekenstein14 by the (1) Previous publication in this 3aies: M. L. Wolfrom and Robert action of dilute aqueous alkali upon D-galactose. L. Brown, THISJOURNAL, 66, 1516 (1943). It was later formed from D-gulose and D-idose (2) Presented before the Division of Sugar Chemistry and Technology at the 106th meeting of the American Chemical Society, Pittsburgh, Pennsylvania, September 7, 1943. (3) Allied Chemical and Dye Corporation Fellow, 1942-1943. (4) R. T.Major and E. W. Cook, Tars JOURNAL, 68, 2474 (1936). (5) M. L. Wolfrom and Mildred R. Newlin, ibid., 68, 4379 (1931). (6) M. L. Wolfrorn and M. Konigsberg, ibid., 61, 574 (1939) (7) R. T.Major and E. W. Cook, ibid., 68, 2477 (1936). ( 8 ) 6.Arragon, Compl. rend.. lB6, 1733 (1933). (9) H H. Schlubach and J. Vorwerk, En.,66B,1251 (1933).
(10) P. B. Cramer and E. Pacsu. Tars JOURNAL, 69, 1467 (1937). (11) G.Arragon, Comgl. rend., PM,735 (1937). (12) M. L. Wolfrorn and A. Thompson, TITISJOURNAL. 66, 880
(1934).
(13) K.GLtzi and T. Reichstein, Bels Chim. A d o , 01, 186 (1938). (14) C . A. Lob- de B N and ~ W. Alberda van Ekenstein, RUG.lrov. chim., 16,262 (1897): 19, 1 (1900): J. U. Nef, A n n . , 40011, 342 (1914).
Feb., 1944
THE
ACTIONOF DIAZOMETHAIW ON D&RBOSB
by aqueous alkali" and by pyridine.'* Fischer and Baerl' have demonstrated the formation of Dsorbose (and D-fructose) by the action of dilute alkali upon =glyceraldehyde. We include herein the synthesis of a derivative of a dicarbonyl-octose. The diazomethyl ketone obtained18 from mucyl dichloride tetraacetate and designated 1,8-bisdiazomucyldimethane tetraacetate, was reduced with hydriodic acid according to the procedure of Wolfrom and Brown,lg yielding the crystalline substance VII, designated mucyldimethane tetraacetate. In the c6 series, a dicarbonyl hexose of a similar type has been isolated in crystalline form by Micheel,40 who employed an entirely different method of synthesis. In this excellent work, the longknown 1,6-dichloro-~-mannitol~~ was converted through a series of crystalline intermediates into D-(ho)-tartaryldimethane,VIII, designated 1,4dimethyl- 1,Cdiketo-tetrose by Micheel and
206
~-Xylanyl Chloride Tetraacetate (IU).CLDxylanic acid tetraacetate (3.0 g.), prepared by the oxidstion of &ehydo-D-xylose tetraacetste according to the procedure of Major and Cookt was shaken for one hour in ether (160 cc.) and phosphorus pentachloride (2.0 8.). Tht product (2.0 g) crysfrom the fltertd reaction muturr on standmg o w h t at icebox temperature and a further quantity (0.8 g.f was obtained from the mother liquor on addition of petroleum ether; total yield 2.6 g. (89%). Pure material was obtained on rectystallization from anhydrous ether; m. p. 72-73', spec. rot. 14' cb, abs. CHCS, D line). The substance a y s h l b d Pnsmatic crystals. A d . calcd. for CfiO&l(CH&O)~: C1, 10.05; saponification value (5 equiv.), 143 cc. 0.1 N NaOH per 100 mg. Found: C1.9.80; sapodkation value, 14.0 cc.
- (q',
l-Desoxy-l-diazo-&eto-D9orboeeTetraacetate 0. ' L
The diazomethane used in this work was prepared by the method of Werner* as modified by Arndt." The ethereal solution of diazomethane was dried over potassium hydroxide pellets and was used at concentrations of 0.0160.025 g. of diazomethane per cc. of solution. The concentration was determined by titration against benzoic acid.= A solution of D-xylonyl chloride tetraacetate (8.3 g.) in sodium-dried ether (30 cc.) was poured slowly with Horn. stirring into 100 cc. of a dry, ether solution of diazomethane All compounds described herein were obtained containing 2.3 g. of diazomethane. There was a vigorous evolution of nitrogen gas during the addition, and the in crystalline form. crystallizationof the reaction product, which began almost Experimental immediately after the addition of the acid chloride was as allowed to complete itself at icebox temImproved Preparation of aldehyydo-D-Xylose Tetraacetate complete, w yield 7.7 g. (92%), m. p. 12!2-124", spec. rot. (I).*Z-To a solution of 40 g. of mercuric chloride in 100 cc. perature; +43' (25', c 5, abs. CHCb, D line). Pure material was of acetone in a 1-liter 3-necked 5ask was added 5 cc. of obtained on recrystalbahon from five parts of acetone by water and 25 g. of powdered cadmium carbonate and the the addition of like volumes each of ether and petroleum mixture was stirred vigorously for ten minutes. To this ether; m. p. 124.5-125.5', spec. rot. 4-44.6" (26'. c 5, abs. was added a solution of 15 g. of D-xylose diethyl mercaptal D line). tetraacetate' in 100 cc. of acetone and the resulting mixture CHCla, This diazomethyl ketone formed individual canarywas maintained a t room temperature and under vigorous mechanical stirring for twelve hours, whereupon it was yellow crystals that were insoluble in ether and petroleum filtered and the solvent removed from the filtrate under ether but were very soluble in acetone, chloroform and reduced pressure and in the presence of a little fresh cad- alcohol. Anal. Calcd. for CUHUGNZ: C, 48.93; H, 5.06; N, mium carbonate. The residual cake was extracted thrice with 75-cc. portions of warm chloroform and the combined 7.82. Found: C,46.94; H, 5.09; N.7.33. extracts washed with 10% aqueous potassium iodide (to keto-D-hbose Pentaacetate ( V ) . * L A solution of 1the disappearance of the red complex first formed) and desoxy-1-diazo-keto-D-sorbosepentaacetate (2.0 9.) in then with water (to a negligible halide test with silver glacial acetic acid (25 cc.) was relluxed until the cessation nitrate). The sirup obtained on solvent removal (reduced of nitrogen evolution (cu. twenty minutes). The crystalpressure) from the washed and dried extract was crystal- line solid obtained on pouring the cooled solution onto a lized from anhydrous ether; yield 6 g. (3 crops), m. P . * ~ mixture of ice and water and partially neutralizing the 87-89'. spec. rot. -22.5' (20°, t 2.5, abs. CHCS, D line). acetic acid with sodium bicarbonate, was removed by Further purification from anhydrous ether pelded pure filtration and washed with water; yield 1.6 g. (73%). m. aEdeydo-D-xylose tetraacetate; m. p. 90-91 , spec. rot. 91-94'. spec. rot. -1.9' (25". c 4, abs. CHCS, D he?: -23.3" (22', t 4, abs. CHCIa, D line). These constants Pure material (elongated prisms) was obtained on reare in agreemznt with those reported by Major and Cook4 crystallization from methanol (decolorizing carbon)(m. p. 90-91 , spec. rot. +22.5' in abs. CHCb for the water; m. p. 97.E-98.5', spec. rot. -2.5' (%", c 4, abs. enantiomorph) but differ in rotation from those reported' CHCL, D line). For the enantiomorph,the followingconpreviously from this Laboratory (m. p. 88-89", spec. rot. stants are raorted: m. D. 99'. wee. rot. +2.8' (20". -16' in CHCk). The melting point of 90-91' and the CHCS, 5780-AA.)aJ1;m. i. 96.6-9f.So, spec. rot. +2.90 specsc rotation of -23" (CHCls) may then be accepted (20'. . . c 2,. CHCL. D line).' for ddehydo-D-xylosetetraacetate. ' Anal. Calcd. for &yOs(CHaCO)r: C, 49.23; H, Anal. Calcd. for C&Os(CH&O)r: CHaCO, 12.6 cc. 5.68; CHsCO, 12.8 cc. 0.1 N NaOH per 100 mg. Found: 0.1 N NaOH per 100 mg. Found: CHaCO, 12.5 cc. C, 49.16; H, 5.66; CHaCO, 12.8CC. keto-D,L-Sorbose Pentaacetate.-Equal weights (1.0 g. (16) W. Alberda van Ekenstein and J. J. Blanksma, Rcc. trow of each) of keto-D-sorbose pentaacetate and kcfO-L-sorbose chim., 27, 1 (1908). were crystallized from methanol (16 cc.) (16) K. Ggtzi and T.Reichstein, A&. Chim. A d a , 91, 468 (1938). pentaacetate*-@ and water (25 cc.); yield 1.8 g. (90%), m. p. 83-84' (un(17) H. 0. L. Fischa and E. Baer, ibid., LO, 519 (1936). changed on further crystallization), spec. rot. 0' (25', c 2, (18) M. L. Wolfrom, S. W. Waisbrot and R. L. Brown, Tars abs. CHCS, D line). JOURNAL, W, 2329 (1942). (19) M. L. Wolfrom and Robert L. Brown, ibid., 66, 1516 (1943). Anal. Calcd. for C~I&(CHSCOL: C. 49.23: H. (20) F. Micheel, Ann., 496, 77 (1932); F. Micheel and IC. Horn, 5.68; CHaCO, 12.8 CC. 0.-1& NaOH p& 100 k g . F k n d : ibid., 616, 1 (1934). C, 49.16; H, 5.70; C a C O , 1 2 . 9 ~ ~ .
(21) G. Bouchudat, Am#. chim. phya., 151 6, 100 (1876). (22) Experimental work by Mr. Robert L. Brown of this Laboratory. (23) AU melting points herein reported were taken in Pyrex melting point tubes.
(24) Experimental work by bfr. S. M. O h . (25) E. A. Werner, J . C k n . Soc., 116, 1093 (1919). (26) F. Amdt, Org. Syulhercs, 16, 8, 48 (1935). (27) E. K. Manhall, Jr., and 5. F. Acree. Ber., 48, 2323
(1910).
206
ROBERTH.SAUNDERS AND M. j , M u m v
of &u-x.-Sorboes Peataecetate."-An amount of 5 g. of Rcle-csorbose pntaacctatesDQwas dissolved in 250 ca. of bailiiag watw. Hydroxylamine hydrochloride (2.5 g.) end potassium acetate (5 g.) were added to the hot solution which waa then allowed to cool slowly to room temperature. Upon cooling to 0' and scratching the sides of the &$, the product crystallized; yield 1.3 g., m. p. 113-114 , spec. rat. -42' (Zip,c 3, abs. CHCL, D line). An additionel 0.4 g. vas obtained by extraction of the mother liquor with ehlorofm. The constants were unaltered after recrystallization from solution in acetoneether by the addition of petroleum ether. Anal. Calcd. for C&LOsN(CHaC0)6: C, 47.41; H, 5.72: N.3.46: CHsCO. 12.3 cc. 0.1 N NaOH Der 100 mn. Calcd. for C&H~O~N(CHCO),:C, 46.28; 6, 5.83; G, 3.86; CHaCO, 1 1 . 0 ~ ~Found: . C, 47.40; H, 5.88; N, 3.68; CH&O (0-acetyl"), 12.1CC. D-Sorboee (VI) from keto-D-Sorbose Pentaacetate."-An amount of 2.43 g. cf keto-D-sorbose pentaacetate was added to 100 cc. of 0.6 N barium hydroxide solution, previously cooled to cu. 5'. With occasional shaking, the crystals dissolved in about thirty minutes. The solution was kept for one hour longer at cu.5" and then was saturated with carbon dioxide and the precipitated barium carbonate was removed by filtration. A slight excess of 2 N sulfuric acid was added slowly and the barium sulfate removed by filtration through a bed of Super-eel (Johns-Manville). The excess sulfuric acid was then removed exactly with barium hydrate and the filtered solution was concentrated under reduced pressure at 40"to a sirup which crystallized spontaneously. The crystalline mass was stirred with 10 cc. of ethanol for filtration; yield 0.90 g. (SO'%), m. p. 158160', spec. rot. $40.5' (so, c 3,HzO, .D line). Pure material was obtained on one recrystallization from water by the addition of absolute ethanol; m. p. 160-16Z0,spec. rot. +43 (ao, c 8, H20,D line, no significant mutarotation). For D-sorbose, Alberda van Ekenstein and Blank; sma16reported the constants: m. p. 166', spec. rot. +43 (c 1, HzO, D line). For the enantiomorphic L-sorbose, -.
(28) Experimental work by Mr. Evan F. Evans. (29) M. L. Wolfmm, Svr. Kongsborg and S. Soltzberg, THIS JOURNAL, 68, 490 (1936).
[CONTRIBUTION FROM THE CHEMICAL
Yol. 66
Schlubach and, Vorwerk9 have recorded the constants: m, p. 159-161 , spec. rot. -43' ( Z O O , c 2, HtO, D line). Pigman and Isbell" record a small, complex mutarotation for L-sorbose. Mucyldimethane Tetraacetate (VII).*LTo a chloroform (40 cc.) solution of 1.8-bisdiazomucvldimethane tetraacetate18 (2.00 g . ) , in a separatory funnel, was added 8 cc. of 47% hydriodic acid. The mixture was shaken until cessation of nitrogen evolution, diluted with water (40 cc.) and the chloroform layer removed and washed successively with water, aqueous sodium thiosulfate and water. A white crystalline solid was obtained on solvent removal from the dried chloroform solution; yield 1.60g., m. p. 200204 '. Pure material was obtained on recrystallization from acetone by the addition of absolute ethanol; yield 1.35 g. (78%), m. p. 204-206". The substance gave a positive iodoform test (iodoform identified by m. p. and mixed m. p.). Anal. Calcd. for C I B H ? ~C, ~ ~ 51.21; ~: H, 5.91. Found: C, 51.25;H, 5.85.
Summary 1. An improved demercaptalation procedure
for aldehyde-D-xylose tetraacetate (I) is described. 2. The tetraacetates of D-xylonyl chloride (111) and of 1-desoxy-1-diazo-keto-D-sorbose (IV) have been synthesized. Reaction of the latter with acetic acid led to the synthesis of keto-Dsorbose pentaacetate (V), from which D-sorbose (VI) was obtained on saponification. 3. An oxime of keto-L-sorbose pentaacetate has been obtained. 4. keto-D,L-Sorbose pentaacetate is described. 5. Hydriodic acid reduction of 1,s-bisdiazomucyldimethane tetraacetate yielded mucyldimethane tetraacetate (VII). (30) W.W.Pigman and H. S. Isbell. J . Rcscerch Null. Bur. Slotadnrds, 19, 443 (1937).
COLUMBUS, OHIO
RECEIVED SEPTEMBER 29,1943
LABORATORY OF ILLINOIS INSTITUTE OF
TECHNOLOGY]
The Aldol Condensation. 111. Aldol-aldehyde Addition Products and their Derivatives BY ROBERTH. SAUNDERS AND M. J. MURRAY H Spiith. Lorenz and Freund' and Hanschke* have recently independently reported the isolation of an addition product of acetaldehyde to OH aldol, tQgether with the acetate and benzoate of this compound. While this work was still unknown to us, we reported an analogous compound formed by the reaction of isobutyraldehyde with its aldol.8 Under the circumstances we are reporting the results of our research to date and are I. concluding our work on this phase of the problem. The following equation shows the addition of By acetylating the crude reaction mixtures of aldehyde to its aldol to form a 1,3-&oxane deriv- aldolized aldehydes related to acetaldehyde, ative propionaldehyde, n- and isobutyraldehydes we (I) Spith, Lorenz and Freund, Bn.,TBB, 57-68 (1943): C A , obtained the 6-acetoxy-1,3-dioxanes in about 37, 4695 (1943). 700/, yields. I n one instance, namely, that from (2) Hanschke, Bet., 768,180 (1943); C. A . , $7, 5374 (1943). isobutyraldehyde, we also prepared the 686, 1714 (3) Saunders, Murray and Cleveland THISJOURNAL, I1943).
methoxy-1,.7-dioxaIie.
Physical properties for
