April 5, 1956
DITHIOCARBONATE ESTERSOF ARABINOSE
from dextrin by an acetobacter strainlo gave thermogram C. The very sharp endotherm a t 220’ is its most prominent feature. The data in Figs. 1 to 5 demonstrate that the dextran thermograms show characteristic contours in the 100 to 310’ region. The 200” endotherms are common to the predominantly 1,6-linked dextrans while those near the 225” region occur with enzymatic preparations. Moreover, the thermal peaks near 245” appear to be typical of the most highly 1,3-like linked dextrans. Clinical size preparations, on the other hand, are recognized by a peak a t 185”. Significantly, almost all the thermograms show endotherms in the 310 f 5’ region. Aside from the dominant influence of the type and proportion of glucosidic linkages, further variations appear to be associated with marked changes in solubility, unique viscosity properties, and to a minor extent, purity of the sample. It must be emphasized here that other unknown factors may be involved, especially in view of the structural complexities of the dextrans and more particularly in view of the undeveloped state of this thermal technique. The differential thermal reactions responsible for the distinct endotherms appear to be similar to those encountered with starch.3 Analogous evidence given by infrared analysis of the pyrolysis
THE
residues, as well as examination of the gaseous pyrolyzates, suggest that transglucosidation”-13 and dehydration constitute the predoiriinant endothermic processes. Dehydration reactions are, moreover, generally recognized as being stereospecific, particularly when they are catalytic. Consequently, the interpretation of the thermal effects due to chain branching and molecular aggregation may be based predominantly on stereochemical considerations. The present investigation has shown that there may be some relationships between the thermal property of the bacterial dextrans and their molecular constitution which cannot be deduced readily by other analytical techniques. The results also indicate that differential thermal analysis may become an invaluable aid in elucidating the constitution of other polysaccharides and biologically important polymers. Acknowledgment.-The author thanks Dr. Allene Jeanes of the Northern Utilization Research Branch, United States Department of Agriculture, for her helpful advice and material assistance during this investigation. ( 1 1 ) B. Brimhall, Znd. Eng. Chem., 36, 72 (19.14). (12) F. G. Pautard, Chemistry & Znduslry, 1310 (1083) (13) I. A . Wolff, el a l . , I n d . Eng. Chem , 46, 7 . X (l!),:3)
OTTAWA, CANADA
(IO) E J Hehre, J Btol Chpm , 198, I G l (1951).
[CONTRIBUTION FROM
1399
DEPARTMENT O F CHEMISTRY OF THE OHIO STATE UNIVERSITY]
Dithiocarbonate Esters of Arabinose’ BY hl. L. WOLFROM AND A. B. FOSTER^ RECEIVEDSEPTEMBER 22, 1955 Certain 2-0-( S-alkyl dithiocarbonate) esters (“xanthates”) of methyl 3,4-f%sopropyfdene-@-~- and L-arabinopyranoside (I, Ia, 11) and of methyl 4,6-0-benzylidene-3-0-methyl-a-~-altropyranoside (IV) have been prepared and their behavior on pyrolysis studied. The derivatives did not undergo a Chugaev type elimination reaction, but the 2 - 0 4 S-methyl dithiocarbonate) of I11 rearranged to the isomeric 2-S-(S-methyl dithiocarbonate) (V) on distillation; V was reductively desul(“2-deoxy-~-ribose”)in low yield. The relation of the refurated and debenzylidenated t o 2-deoxy-~-erythro-aldopentose arrangement (of which other examples are known) to the Chugaev reaction is discussed.
S The production of olefins by the controlled thert R ’I mal decomposition of the 0-(S-alkyl dithiocarbon- R-OH +R-ONa --+R-OC-SNn + ate) or “xanthate” esters of certain alcohols (ChuS gaev reaction)3 is now well known. Numerous ext amples of the Chugaev reaction have been described ROC-SR’ (1) in the literature, but the thermal decomposition of dithiocarbonate) .5 0-Acetyl-1-thioaldose 1-S(02-O-(S-alkyl dithiocarbonate) esters of simple caralkyl dithiocarbonate) esters have been prepared bohydrates has been little studied. In fact, few such esters are known. Freudenberg and Wolf4 by the action of potassium 0-ethyl dithiocarbonate on 0-acetylglycosyl halides.6 described , 1,2 :3,4-di-O-isopropylidene-a-~-galactoPyrolysis of 1,2 :5,6-di-0-isopropylidene-a-~-glupyranose 6-O-(S-methyldithiocarbonate), 2,3 :5,6-di0-isopropylidene-D-mannofuranose 1-0-(S-methyl cofuranose 3-O-(S-methyl dithiocarbonate) under dithiocarbonate) and 1,2 :5,6-di-O-isopropylidene-a- the conditions originally employed by Chugaev3 D-glucofuranose 3-0-(S-methyl dithiocarbonate), did not give the desired 3,4-unsaturated compound. The purpose of this investigation was the preparaof which the latter two derivatives were crystalline. These esters were prepared by the general route (1) tion of other dithiocarbonate (“xanthate”) esters of as was methyl a-D-glucopyranoside 2-0-(S-methyl simple carbohydrates, which might reasonably be expected to undergo a Chugaev type reaction ( 5 ) (1) A preliminary report of this work appears in Abslracls Papcrs and a study of their behavior on pyrolysis. A m . Chcm. Soc., 116, 23D (1954). (2) Rockefeller Foundation Fellow,1953-1954. (3) L. Chugaev, Bcr.. 32, 3332 (1899). (4) K. Freudenberg and A. Wolf, ibid., 60, 232 (1927).
(5) T. Lieser and E. Leckzyck, Ann., 619, 279 (1935); M. L. Wolfrom and M. A. El-Taraboulsi, THISJOURNAL, 76, 5360 (1953). ( 0 ) W. Schneider, R. Gille and K. Eisfeld, Bcr., 61, 1241 (1928).
1400
M. L. WOLFROM AND A. B. FOSTER
Vol. 78
TABLE I
DATAO N THE PREPARATION OF THE S-ALKYL XANTHATES [alZsD
Compounda
Yield,
70
M.p., O C .
(c 2, benzene)
Formula
Carbon, % Calcd. Found
Hydrogen, % Calcd. Found
Methyl 3,4-0-isopropylidene-@-~-arabinopyranoside 20-(S-methyl dithiocarbonate)b ( I ) 70-75 119-120 +217O CllH1805Ss 44.91 44.55 6 . 1 2 6.24 2Methyl 3,4-0-isopropylidene-P-~-arabinopyranoside 0-(S-methyl dithiocarbonate)* (Ia) 72 120-121 $210 CllH1805S2 44.91 44.74 6 . 1 2 6.35 Methyl 3,4-0-isopropylidene-~-~-arabinopyranoside 20-(S-triphenylrnethyl dithiocarbonate) (11) 54 169-171 54 Cz9H3105Sz 67.01 66.66 5.87 5 . 7 5 Methyl 4,6-O-benzylidene-3-O-methyl-a-~-altropyranoside 2-0-( S-methyl dithiocarbonate) ( I V ) 26 113-1 15 Ci7H2206S2 53.16 52.85 6 . 0 8 5.70 a Attempts to prepare methyl 3,4-0-isopropylidene-@-~-arabinopyranoside 2-0-(S-p-nitrobenzyl dithiocarbonate) and methyl 4,6-0-benzylidene-a-~-glucopyranoside 2,3-bis-O-(S-methyl dithiocarbonate) were unsuccessful. Mixed melting point of the D- and L-forms was 97-99'. Recrystallization of equal amounts of the enantiomorphs gave a product of m.p. 97-99",
+
*
The Chugaev reaction involves an elimination of action of the presence of methyl 3,4-O-isopropylithe 0-(S-alkyl dithiocarbonate) group together with dene-P-~-2-deoxy-~-erythro-aldopentoside (resulting an adjacent cis-hydrogen atom (5). The reaction is from cleavage a t b) or methyl 3,4-O-isopropylidene known to proceed readily in cyclohexane deriva- 2-O-methyl-p-~-arabinopyranoside (resulting from t i v e ~and ~ for this reason analogous pyran deriva- cleavage a t e). These findings are in accord with tives were selected, namely, methyl 3,4-0-isopro- those of Wolfrom and El-Taraboulsi in the D-glupylidene-P-D- and L-arabinopyranoside (111) in cose structure. l 2 which a substituted 0-(S-alkyl dithiocarbonate) The 0-(S-alkyl dithiocarbonates) I, Ia, I1 and 1V group, a t position 2, would be adjacent to a config- were subjected to 2 variety of thermal treatments urationally cis-hydrogen. The esters were pre- and typical experiments are summarized in Table pared by the general route (1) in which the use of 11. Some difficulty was experienced with the desodium hydrideg in the first step was found to be rivative of methyl 3,4-0-isopropylidene-/3-~-arabinomuch more expedient than the inore conventional pyranoside (I) which sublimed when heated a t 120use of sodium. The remainder of the sequence is 204' in an open tube a t atmospheric pressure. The well established. A 70-75% over-all yield of the use of a sealed and evacuated tube under the thercrystalline 2-0- (S-methyl dithiocarbonate) (I, Ia) mal conditions B and C, did not lead t o the eliminaof methyl 3,4-0-isopropylidene-~-~or L-arabino- tion of carbon oxysulfide and methyl mercaptan pyranoside and a 54y0over-all yield of the 2-0-(S- which would indicate the occurrence of a Chugaevtriphenylmethyl dithiocarbonate) (11) ester of the type reaction. As the period or temperature of L-isomer were thus obtained. The latter deriva- heating was increased, the amount of unchanged tive was prepared for study because of the known starting material recovered decreased. This is in factge10 that the more electrophilic the S-alkyl agreement with the subsequently described finding group in the xanthate esters, the greater the insta- of the rearrangement of I on distillation. It is of bility of the derivative on pyrolysis. I n this re- interest to compare the thermal conditions B and spect the S-triphenylmethyl group is the most C with those used by O'Connor and Nates who effective.1° A much lower over-all yield (25gib) of found that 0-cholestanyl 0-(S-methyl dithiocarbonmethyl 4,6-0-benzylidene-3-O-methyl-cr-~-altropyate) was converted in 94% yield to a mixture of A2ranoside" 2-O-(S-methyl dithiocarbonate) (IV) and A3-cholestene when heating a t 230' for 2.5 was obtained (see Table I). In this compound (IV) hours a t 20 mm. pressure. the dithiocarbonate group is doubly flanked by cisThe use of the more drastic thermal conditions D on I led to a 36% loss in weight (theoretical loss for hydrogen atoms. Reductive desulfuration of methyl 3,4-O-isopro- elimination of COS and CH,SH, 30.6%), no startpylidene-P-D-arabinopyranoside 2-0-(S-methyl di- ing materia1 was recovered and most of the residue thiocarbonate) (I) with Raney nickel gave the par- was insoluble in benzene. These results would indient alcohol I11 which was characterized as the cate extensive decomposition of the dithiocarbonate known crystalline 2-p-toluenesulfonate. The forma- ester. tion of I11 resulted from cleavage of the C-0 linkage Under thermal conditions (Table 11) similar t o a. There was no indication in the products of re- those used by O'Connor and Nace and mentioned above, I1 and IV did not undergo any elimination S 1 1 ; .--! ---, reaction or rearrangement and could be recovered CH-~-O--/-C-~-S-CH~ largely unchanged. It is clear from these results that the S-alkyl dithiocarbonate esters studied do I b a : not undergo elimination reactions under normal (7) E. R. Alexander, "Principles of Ionic Organic Reactions," Chugaev reaction conditions. John Wiley and Sons, Inc., New York, N. Y.,1950, p. 120. When compound I was subjected to distillation (8) G. L. O'Connor and H. R. Nace, THISJ O U R N A L , 74, 5454 (1952). a t atmospheric pressure (b.p. 270-280') the distil(9) Irene M. McAlpine, J . Chcm. Soc., 1114 (1931). late consisted largely of unchanged starting materid (10) G. L. O'Connor and H. L. Nace, THISJOURNAL, 7 6 , 2118 (1953).
(11) G, J. Robertson and C. F. Griffith, J . Chcm. Soc., 1193 (1935).
(12) M. L. Wolfrom and M. A. ECTarabould, Abslracls Pagers Am.
Chcm. Soc., 111, 8P
(1952).
1401
DITHIOCARBONATE ESTERSOF ARABINOSE
April 5, 1956
TABLEI1 THERMAL TREATMENT OF THE O-(S-ALKYLDITHIOCARBONATE) ESTERS Compound
Methyl 3,4-0-isopropylidene @-D-(Or L)- (A) Open tube, atm. press. arabinopyranoside 2-O-(S-methyl di- ( B ) Sealed tube in vacuo thiocarbonate)” (I, I a ) (C) Sealed tube in vucuo (D)
Sealed tube in vacuo
Methyl 3,4-0-isopropylidene 8-L-arabinopyranoside 2 - 0 4 S-triphenylmethyl diOpen tube a t 10-15 mm. thiocarbonate) (11) Methyl 4,6-0-benzylidene-3-O-methyl-aD-altropyranoside” 2-0-( S-methyl dithiocarbonate) (IV)b Open tube a t 10-15 mm. 5
Sample wt., 300 mg.
Temp., OC.
Time, hr.
120-205 200-210
1 1
210-218
4
253-254
5
Reaction conditions
b
200-230
203-207
Observation
Starting mater. subl. unchanged No evol. of CHaSH f COS; 230 mg. recovd. No evol. of CHsSH -I- COS; 153 mg. recovd. Loss in wt. 107 mg.; no starting mater. recovd. ; residue largely insol. in benzene
1-3 No loss in wt.; 50-80% recov.
2
No loss in wt.; 60 mg. recovd.
Sample wt., 80 mg.
H H which crystallized and was easily separated. It is \I/ probable that the non-crystalline fraction of the dis\s tillate contained a rearranged product, presumably (5) -3 CII C-SCHI I C N C-SCHa the substance V. This assumption was based on / I \o/ /I the fact that reductive desulfuration of the noncrystalline fraction with Raney nickel gave known molecular and that the transition state complex is methyl 3,4-0-isopropylidene-@-2-deoxy-~-erythro-a1dopentoside (VI). Hydrolysis of VI gave 2-deoxy- highly ordered (negative entropy of activation). D-erythro-aldopentose (“2-deoxy-~-ribose”), char- Thus the geometry of the transition state would apacterized by ionophoresis and by the formation of pear to be of importance. It is possible that the the crystalline anilide. The inferred rearrange- failure of the dithiocarbonate esters described in ment I += V is not without precedent. Thus this paper to undergo a Chugaev-type elimination Freudenberg and Wolf4obtained crystalline 1,2:5,6- reaction may have been due in part to interference di-0-isopropylidene cy-D-1-thioglucofuranose 3-s- with the probably critical geometry of the transi(S-methyl dithiocarbonate) in 30% yield on distilla- tion state by the heavy substitution of the pyran ring. Alternatively, the occurrence of an elimination of 1,2 :5,6-di-O-isopropylidene-a-~-glucofuranose 3-O-(S-methyl dithiocarbonate) a t atmospheric tion reaction in the 0-(S-alkyl dithiocarbonate) pressure (2). Similar rearrangements of such es- derivatives I, 11 and IV would lead to an enol ters have been noted by Laakso‘3 and in the closely ether and not an olefin and it is possible that the hydrogen @-cis to the 0-(5’-alkyl dithiocarbonate) S 0 group is inactivated by the oxygen functions att II R-0-C-S-R’ +R-S-C-SR’ (2) tached to the same carbon atom. It has been suggestedl7that the Chugaev reaction S 0 may involve a rearrangement to the dithiol carbont I1 R-S-C-0-R’ +R-S-C-0-R’ (3) ate (2) prior to elimination. This would seem imS 0 probable since the dithiol carbonates are known to t It be more stable than the corresponding S-alkyl diR-C-0-R‘ +R-C-SR’ (4) thiocarbonate esters.l 3 Further, from an infrared analogous thionocarbonates (3)’4 and thiono esters study, O’Connor and Nace8 concluded that no re(4). arranged product was present in quantity as the The isolation of 2-deoxy-~-erythro-?ldopentoseChugaev reaction proceeded. Thus i t would ap(“2-deoxy-~-ribose”)as described herein consti- pear that the rearrangement (3) encountered by tutes a new synthesis of this biologically important Freudenberg and Wolf4 and in this paper are not sugar. Because of the low over-all yield the syn- partially complete Chugaev-type reactions but althesis is of’little value as a preparative method. ternative reactions. It is significant that the therThe mechanism of the Chugaev reaction has been mal conditions necessary to effect the rearrangestudied extensively16 and i t is currently thought ments of the types 2, 3 and 44-13-15are generally that the /3-cis elimination which occurs invdves a more drastic than those necessary to bring about cyclic transition state complex in which the thiono the Chugaev reaction in suitable derivatives.s sulfur bonds to the @-hydrogen. O’Connor and Experimental Nace8 have shown that the reaction is strictly uni-
YJ
(13) P.V. Laakso, Suomcn Kcmisfilchli, 188, 19 (1943); C. A , 40, 4687 (1946). (14) A. SchBnberg a n d L. v. Vargha, Bcr., 63, 178 (1930); A. Sch6nberg, L. v. Vargha and W. Paul, Ann., 488, 107 (1930). (15) S. A. Karjala and S. M McElvain, TEISJOURNAL, 55, 2966 (1933). (16) See reference 8 and other references cited therein.
q \ \
\c.
Methyl 3,4-0-Isopropylidene-g-o-arabinopyranoside 2-0(S-Methyl Dithiocarbonate) (I).-A solution of 10.6 g. of methyl 3,4-O-isopropylidene-~-~-arabinopyranoside~~ in 300 ml. of dry benzene was added to a suspension of 6 g. (5 moles) of sodium hydride in 50 ml. of dry benzene.a Vigor(17) D. S. Tarbell and D. P. Harnish, Chem. Revs., 49, 60 (1951). (18) J. Honeyman, J . Chcm. Sac., 990 (1946).
1402
1-d.7s
CH?--C-
II
0
CIIa V, K
CII
-i.
~
0
ClIl =
CH3-
\. I
ous evolution of hydrogen occurred aiid tlie mixture \v;i. \%-ellstirred and boiled under reflux for 18 hr. T h e deep r c t l solution was cooled, decanted from excess sodium iiydritlc and treated with 20 ml. of carboii di-ulfide. 1nitnctli:ite separation of t h e red gelatinous sodium salt of the sugar dithiocarbonate occurred. T h e misture was boiled under reflux for 1 hr. and was then treated with excess methyl iodide. After a further 5 hr. of refluxing, the reaction mixture was filtered, partially decolorized with carbon and concentrated to small volume. Addition of escess hesarie gave colorless prisms; yield 11.1 g. (72%). Pure material was obtaiiied on recrystallization from benzene-hexane; m.p. 119-120°,L9 [ a ] * ' D -217' (c 2.48, benzene). By essentially the szme procedure, the derivatives listed in Table I were prepared. Reductive Desulfuration of Methyl 3,4-O-Isopropylidene8-L-arabinopyranoside Z-O-(S-Methyl Dithiocarbonate) (Ia) . -A solution of 1.31 g. of I a in 200 ml. of an ethanol-benzene mixture (equal parts by volume) was boiled under reflux with freshly prepared Raney froni 30 g. of :illoy. T h e mixture was filtered aiid the nickel w:is wLishccl times with hot methanol. Evaporatioii crf the COI filtrcite and washings gave :L sirup, yield 0.48 g. (52.8yo). b.p. 8&85' [bath temperature) rO.O5 mm.), n31 1.4600, la]31D +192' (c 1,-I cliloroform). , Honeyniaii'8 record> b.p. 82" (bath temperature) (0.1 inin.), [a]D +199.1" iii chloroform for methyl 3,4-O-isopropylidene-/3-x,-arabiiioygranoside. T h e product from the reductive desulfuration gave a crystalline p-tc~lueiiesiilfonate in good yield. 1n.p. 131-132' alone or i i i admixture with autheiitic methyl :i , 4 - O - i ~ o p r o p y l i t i ~ i i e - ~ - ~ - ~ - t ~ ~ l ~ i e 1 i0 ~ -s Lu -l f:ir:ibinvpyu1iylr,iiioside. l a Isomerization of Methyl 3,4-O-Isopropylidene-B-~-arabinopyranoside 2 - 0 - (S-Methyl Dithiocarbonatej (I).-Isomcriz'ition of I wits effected by distillatioii a t atmcrsplieric ure.' Tlie tli,tillate \viis :illowed tu stand for 2 1 hr. rmit as complete crystallization aq poscihle o f unchanged starting material. Estraction of th'e distillate with hexane nhich contnined 10% of benzene or ethanol removed sirupy material. Concentration of the extracts normally afforded a small estra amount of starting material. T h e results of a series of distillatioiis are recorded in Table 111. Reductive Desulfuration of the Non-crystalline Distillate from I.--A solution of 15.77 g . of the non-crystalline distilliitc froni I (see Table 111) in 500 ml. of an etlianul-hexxrie mixture ( e q u d part< by volume) w;is boiled uiider reflux fur 5 hr. with Raney nickel freshly prepared from 100 g . of alIoy.*O T h e mixture was filtered and t h e nickel was exhaustively extracted with methanol using a Soxhlet apparatus. T h e extract was combined with tlie earlier filtrate, concentrated, and the residue distilled, ii-iiig :i LVidrner-
__~~___
(10) All melting points w e rincorrected. (20) H . A d k i n s o n d A . A P;t\lic, 'Tirrs J O I ~ R N A I . ,6 9 , 3039 (1947).
1.43 2.07 4.11 4.20 4.17 3.71 19 45 13 82
0.52 0.70 1.72 1.89 1.77 1.53 lo 00
7.50
0 1 1 0 1
93 03 12
40d
26 21
80
16
06 0 64 4 5.5 4 12
20 15 19 25
0 91 2 27 2.39 2.31 2.40 2.18 9.45 6.32
38 A7 44 46 41 50 39 3'5
April 5 , 19.3;
TDICNTIFIC.ITION
01: I ~ L D O S E S z
aldopentose in 5 ml. of ethnnol containing 0.1 g. of freshly distilled aniline was hoilcd uiider rcflux for :3 hr. Therecrystallized slowly; yield 20 mg. ‘Tlie crude miterial was recrystallized from ethanol; ni .I). 10:i l M 0 unchanged on
l k
TIIEIR
P I I E N ~ I . I I \ ~ D R . \ Z O N I ~ S 1403
adrnisture with an authentic specimen of 2-deosy-~-frylhroaldopentose a n i l i d ~ * 4 of , * ~m.p. 1BS-107°. “Ju”
‘‘JTG’’
COLUMBUS 10, OHIO
[ C O N T R I B t J T I O N FROM THE INSTITUTE O F O R G A N I C C H E M I S T R Y ,
TECIINICAL UNIVERSITY,
BUDAPEST]
Identification of Aldoses in the Form of Formazans via their Phenylhydrazones BY L. MESTERAND A. MAJOR RECEIVED JUNE16, 1955 Aldoses have been identified by the use of the formazan reaction. B y coupling with diazonium solutions, crystalline sugar formazans with Characteristic physicnl properties have been prepared from the reaction of the corresponding aldose and phenylhydrazine without isolation of the internmediate phenylhydrazone. T h e optimal conditions for formazan formation hnvr been established.
considerable diminution of their capacity for COUpling, a phenomenon connected with optically observable structural changes. Obviously, two processes are taking place in the reaction of the sugar dnd phenylhydrazine, which from the point of view of formazan formation have opposing effects : (i) the formation of phen~l1iyclr:izoiie; (ii) the transformation of this into a form unsuitable for COLIpling. To decide on the usefulness of the simplified eaction in aldose identification, foriiiazan formati(111 frorqthe corresponding altloscs has been studied as a function of time. For each sugar the reaction with phenylhydrazine has been examined in three difpyridine 1 ferent solvents, z k , (i) water, (ii) XYc alcoholI CHz-OH CH2-OH water, (iii) pyridine. After identical periods of Aldose phenylhytlrnzone Aldose diphenylformazan time, all three solutions were coupled to yield the formazan in a pyridine medium by the atldition of TABLE I diazotized anili e. The data in the tables reveal M.P., that on the who e the yield was highest i n pyridine Crystalline form solution, between the 24th and 4Sth hour. U’ith D-Glucose diphenylformazan 177-178 Red ndls., freq. rosettes D-Galactose dil,henylforinazan 1G7-1G8 Bronze-red tablets mannose the nearly insoluble phenylhydrazone sepD-Mannose diphenylformazan 17-1-175 Bunches of microscopic rusarates very soon; therefore in this case the formset ndls. azan reaction was done with the separated phenvlL-Arabinose diphenylforrnaznn 173-174 Bunches of fine br.-red hydrazone and, of course, showed no deviation :is a ndls. L-Rhamnose diphenylformazan 17,7-176 Brill. red ndl. cryst. function of time. D-Xylose diphenylformazan 183-1 24 1.anceolate red ndls. Experiments on the separation of mixtures of Because they contain no active nicthine groups3f4 sugar forrnazans by means of chromatography arc the ketose phenylhydrazones do not yield formaz- in progress. Acknowledgment.-The authors wish to thank ans; this reaction is therefore suitable for clistinguishing aldoses from ketoses. It is more specific Professor G. ZemplCn for his interest in this w r k than the osazone test, frequently used in the identi- and for valuable advice. Thanks are also tlue to I . fication of sugars, for i t yields a different formazan GyBngy, who performed some of the experiments. for each aldose, while the latter does not differenExperimental tiate between epimers. Reagents .-D-Glucose puriss. ‘ ‘Gyogyert ; D-gnl,ictose For the purposes of identifying aldoses, we have puriss. “Riedel”; D-mannose “ ~ ~ o f f m a n n - ~ a ~ o c t i eI.-” ’ simplified this reaction. The sugar phenylhydra- arabinose p. u. sci. “Riedel”; L-rhnmncise ‘‘n.D.fT.”; zones derived from the reaction mixtures of the D-xylose puriss. “Fluka”; L-sorbose p u r . “Llerck”; Dsugar and phenylhydrazine are not isolated but, fructose p. u . sci. “Schuchardt.” Melting Points.-All m.p.’s were determined in capillary after addition of pyridine, are immediately coupled tubes and are uncorrected. As the 1 r . p . chunges ccinsiderto give the formazan. ably with the manner of heating,s our h t l i \vas heatcd rapIn a previous paper2i t has been reported t h a t on idly t o within about 10’ of the m.p. Thereafter its trmstanding sugar phenylhydrazone solutions suffer a perature was raised slowly, so t h a t the mercury culumn \IUS
In earlier communications1f2we have reported that when allowed to react in pyridine solution with cold solutions of diazoniu~ncompounds, aldose phenylhydrazones yielded brilliant red, readily crystallized sugar formazans. Owing to their melting poict and characteristic crystalline forms, these new nitrogenous sugar tlerivativcs can be used to advantage in identifying aldoses. +
O C .
4;
”
(1) G . ZemplCn and L. Mester, Acto C h i f n . A c a d . Sci. H n n g . , 2, 9 (1952). ( 2 ) L. Mester and A . Major, TEISJ O U R N A I . . 77, 4297 (1955). (3) H . v. Pechmann, Rer., 25, 3181 (1802). (4) S . IIiinig and 0. noes, Aiijr , 579, 2s (1953).
unchanging a t the moment of melting. Preparation of Diazonium Solution.--A solution of 9.3 g. of aniline in a mixture of 25 ml. of concd. hytlrocliloric acitl and an equal amount of water was diazotized a t ( J 5’ \r.itli --__ ( 5 ) H. v . I’erhmnnn. B w ,27, 1682 (1894).
