HEWITTG . FLETCHER, JR.,
3682 [ CONTRIBUTION FROM
THE
AND
C. S. HUDSON
Vol. 71
LABORATORY OF CHEMISTRY AND CHEMOTHERAPY, EXPERIMENTAL BIOLOGY AND MEDICINE INSTITUTE, NATIONAL INSTITUTES OF HEALTH]
Relations between Rotatory Power and Structure in the Sugar Group. XXXVI.’ The 1 $-Anhydrides of the Glycitols and Related Sugar Derivatives BY HEWITTG. FLETCHER, JR.,
AND
C.S. HUDSON
Twelve years ago, when Werner Freudenberg tion to exclude an actual reversal of the sign of a and E. F. Rogers2 published their article on “the large epimeric difference. These two agreeing inchemistry of naturally occurring monoanhydro- dications that were noted by Freudenberg and hexitols” i t had already been determined that Rogers have been shown by the later conclusive polygalitol and styracitol are 1,5-anhydrides of structural and configurational chemical proofs6 to D-mannitol and D-glUCitOl but i t was not known have led them to the correct assignment of conwhich of these epimeric structures applies to each figurations to polygalitol and styracitol. Since anhydride. Styracitol, to be sure, had been syn- this example shows that the isorotation comparithesized through the catalytic hydrogenation of sons gave valuable early indications of configuratetraacetyl-2-hydroxy-~-glucal by Zervass but tions, we seek in the present article to extend such this synthesis did not distinguish between the comparisons to a rather large group of sugar deepimeric anhydrides since carbon atom two of tet- rivatives that are related through the common posraacetyl-2-hydroxy-~-g1ucal is symmetric. Freu- session of a tetrahydropyrane structure, with pardenberg and Rogers obtained two indications that ticular emphasis on the 1,5-anhydroglycitols and styracitol is of the D-mannitol series and polygali- their 2-desoxy derivatives, the hydroglycals. Examination of Table I will demonstrate that to1 of the D-glucitol series. As the first of these indications, the rate of oxidation of styracitol by the isorotation hypothesis holds well for the 1,5lead tetraacetate was found to be much faster anhydropentitols and for the four known 1,5-anthan that of polygalitol, from which they in- hydrohexitols as well as for the fully acetylated ferred for styracitol the &-relationship of hy- derivatives of all these substances. The large droxyl groups on carbon atoms two and three that epimeric differences of molecular rotation are of occurs in the 1,5-anhydro-~-mannitolconfigura- the correct sign throughout and they range in tion in contrast to the absence of such a pair of ad- value between 5,500 and 7,650 for the l,&anhyjacent hydroxyl groups in the 1,5-anhydro-~-glu- drides and between 8,480 and 13,400 for the acecitol pattern; their inference was based upon the tates of these anhydrides. I n the hexitol series previous generalization by Criegee4 that the cis the 1,5-anhydrides of allitol, altritol, gulitol and arrangement is attacked more rapidly than is the iditol are still unknown; they constitute two epitrans by this type of selective oxidation. The meric pairs and its seems reasonable to predict second indication came from their comparison of that the epimeric difference for each pair will have optical rotatory values by the isorotation method. the sign that is shown by the known pairs of TaThe pyranose (i, e., 1,5-ring) substances of the ble I. If a member of one of these epimeric pairs D-glucose series are much more dextrorotatory should be discovered through a synthetic method than their epimers of the D-mannose series; thus that does not distinguish between epimers, as for the molecular rotation of methyl a-D-glucopyrano- example through the catalytic reduction of an acylside is +30,860 and that of methyl a-D-mannopy- ated 2-hydroxyglycal, is i t possible t o assign the ranoside is $15,380, and the epimeric difference is full configuration through the use of isorotation thus 30,860 - 15,380 = +15,480, a large value. hypotheses? The answer to this question may be For polygalitol the molecular rotation is +6,950, obtained through the following considerations. The hydroglycal that is common to the D-glufor styracitol i t is -8340 and the epimeric difference is +15,290. The closeness of the numerical cose and D-mannose series (111) is closely related agreement may possibly be fortuitous, since a to polygalitol (I) and styracitol (TI); i t is a 2comparison of the molecular rotations of methyl desoxypolygalitol, which is synonymous with 20-D-glucopyranoside (- 6,640) and methyl P-D- desoxystyracitol. If the molecular rotation of A and that of stymannopyranoside (- 13,550) gives +6,910 for the polygalitol is expressed as B epimeric difference. I n either case, however, the racitol as B - A, where A represents the contribucomparison of rotations indicated in a qualitative tion from carbon atom two and B that from the way that polygalitol belongs in the D-glucose ser- remainder of the structure, the molecular rotation ies and styracitol in the D-mannose series, on the of hydroglucal, in the structure of which carbon assumption that the isorotation hypothesis ap- atom two is symmetric, may be written [(B (B - A)]/2 = B, if i t be assumed that isoplies to the substances with sufficient approxima- A)
+
+
(1) Number XXXV, by W. T. Haskins, R. M. Hann and C. S. Hudson, was published in THIS JOURNAL, 69, 1G68 (1947). (2) W.Freudenberg and E. F. Rogers, i b i d . , 69, 1602 (1937). (3) L.Zervas, BEY.,63, 1689 (1930). (4) R. Criegee, Ann., 491, 211 (1932): 107, 159 (1933).
+
(5) (a) L. Zervas and I. Papadimitriou, Bcr., 73, 174 (1940): (b) N. K.Ricbtmyer and C. S. Hudson, THIS JOURNAL, 61,64(1943); (c) N. K.Ricbtmyer, C. J. Carr and C . S. Hudson, ibid., 66, 1477 (1943); (d) R. C. Hockett and Maryalice Conley, i b i d . , 66, 464
(1944).
THE1,S-ANHYDRIDESOF THE GLYCITOLS AND RELATED SUGARS
Nov., 1949
3683
TABLE I PROPERTIES O F 1,5-ANHYDROGLYCITOLSDERIVEDFROM
MONOSACCHARIDES
CrOH (or OAc) Formula group
M. p.,
Mol.
OC.
Wt.
[a]"D
Solvent
[1bfI2OD
B
BAo
4-Desoxypentitol Series
1,5-Anhydro-~-threo-4-desoxypentitol" XV 1,5-Anhydro-~-erythro-4-desoxypentitol" XI Diacetate of X V Diacetate of XI 1,5-Anhydro-n-arabitol 1,5-Anhydroribitol Triacetate of XVII Triacetate of X I X 1,5-Anhydro-~-lyxitol" 1,5-Anhydroxylitol Triacetate of XVII Tiiacetate of XVI
XVII XIX
XVII XVI
-
67-6Sb
118 -44.9*
H20
+ +
Amorph.c Amorph.b Amorph.d
118 $48.2' 202 -38.Sb 202 $45.1d
HzO CzHsOH CHC13
-
+ + + +
Pentitol Series 96-97" 134 128-129' 134 58' 260 133-134' 260 96-97" 134 116-117" 134 58O 260 122' 260
-98.6E 0 (meso) -74.26 0 (meso) -98.6e 0 (meso) -74.2E 0 (meso)
HzO CHClj H20 CHCL
-
5,300
5500
+ 5,690 -
7,840
+ 9,110 -13,200 0 -19,300 0 -13,200 0 -19,300 0
8,480
6600 9,660
6600 9,660
Hexitol Series 1,5-Anhydro-~-mannitol I1 - 155h 164 -50.9* H20 - 8,340 7650 141-142< 164 $42.4' H20 6,950 1,5-Anhydro-~-glucitol I - 66-67, 58' 332 -42.0k CHCl, -13,900 Tetraacetate of I1 13,400 73-74' 332 $38.9' CHC1, +12,900 Tetraacetate of I - Amorph." 164 -11.4" HzO - 1,870 7240 1,5-Anhydro-~-talitol VI 1,5-Anhydro-~-galactitoi VI1 114* H20 +12,600 164 +76.6" - 106-107m 332 -16.2" CHC13 - 5,380 Tetraacetate of VI 10,800 Tetraacetate of VI1 332 $49.1" CHCla $16,300 75-76" a The names of XV, X I and XVII are systematic alternative designations of the known hydro-~-xylal, hydro-L-arabinal and 1,5-anhydro-~-arabitol. M. Gehrke and F. Obst, Ber., 64, 1724 (1931). G. E. Felton and W. Freudenberg, THISJOURNAL, 57, 1637 (1935). See text of present paper and also M. Gehrke and F. X. Aichner, Ber., 60, 918 (1927). e H. G. Fletcher, Jr., and C. S. Hudson, THISJOURNAL, 69, 1672 (1947). f R. Jeanloz, H. G. Fletcher, Jr., and C. S. R. C. Hockett and Hudson, ibid., 70, 4052 (1948). 0 H. G. Fletcher, Jr., and C. S. Hudson, ibid., 69, 921 (1947). Maryalice Conley, ibid., 66, 464 (1944). N. K. Richtmyer and C. S. Hudson, ibid.,65, 64 (1943). i Y. Asahina, Ber., This hitherto unpublished value was obtained in this Laboratory using a concentration of 2.634 g. per 45,2363 (1912). 100 ml. of solution. Asahina (Zoc. cit.) reported a rotation of -20.9'' in alcohol. N. K. Richtmyer, C. J. Cam, Jr., 65, 1477 (1943). D. A. Rosenfeld, N. K. Richtmyer and C. S. Hudson, ibid., 70, and C. S. Hudson, THISJOURNAL, H. G. Fletcher, Jr., and C. S. Hudson, ibid.,70,310 (1948). 2201 (1948).
+ + +
+
+
tween those for 1,5-anhydro-~-arabitol (- 13,200) and 1,5-anhydroribitol (0), the molecular rotation of hydro-D-xylal (-5,300) lies between those for 1,5-anhydro-~-lyxitol(synonym, 1,5-anhydro-Darabitol) (- 13,200) and 1,5-anhydroxylitol (0) while the molecular rotation of hydro-D-galactal (+7,120) is between those for 1,5-anhydro-~I I 1 (- 1,870) and 1,5-anhydro-~-galactitol talito1 HCOH HCOH (+12,600). As is to be expected, the molecular +B +B I /+B HCI H L HC--J rotations of those acetates of these compounds I I I that are known are similarly related. Thus triCHzOH CHzOH CHzOH acetylhydro-D-glucal has a molecular rotation I I1 I11 (+9,760) which is between the values for tetraPolygalitol Styracitol Hydro-D-glucal acetylstyracitol ( - 13,900) and tetraacetylpolymathematical precision, i t may be that the ap- galitol (4 12,900); in the five-carbon series proximation will be such that the molecular rota- the molecular rotation of diacetylhydro-D-xylal tion of hydroglucal will lie well between those of (- 7,840) is seen to lie between those for triacetylpolygalitol and styracitol. Such is indeed the 1,5-anhydro-~-lyxitol(synonym, triacetyl-1,5-ancase, as is shown by the data in Tables I and 111; hydro-D-arabitol) (- 19,300) and triacetyl-1,5polygalitol (+6,950), hydro-D-glucal (+2,440) anhydroxylitol (0). and styracitol ( - 8,340). Likewise the molecular The literature reveals that an amorphous dirotation of hydro-D-arabinal (- 5,690) lies be- acetylhydroarabinal having a dextrorotation in rotation holds for the three substances. Even though i t cannot be expected that i t will hold with
I
HEWITTG. FLETCHER, JR., AND
3684
PROPERTIES O F THE
c. s. HUDSON
Vol. 71
TABLE I1 SUBSTITUTED 1,5-ANHYDROGLYCITOLSDERIVEDFROM THE DISACCHARIDES CrOH (or OAc) Formula group
Mol. wt.
M. p.
[a]aQD
1,5-Anhydro-4-(@-D-glUCOpyI anosy1)-D-glucitol ( = 1,5-anhydrocellobiitol) VI11 172" 326 29.3" 75-76" 621 4.0" Heptaacetate of VI11 1,5-Anhydro-6- (P-D-glucopyranosy1)-D-gluci to1 ( = 1,5-anhydrogentiobiitol) XXI 239-240" 326 $ 3.6" 153" 621 13.0" Heptaacetate of X X I 1,5-Anhydro-4-(P-D-galactopyranosy1)-D-glucitol ( = 1,5-anhydrolactitol) XI1 233-237 (dec.)b 326 49.4b 1,5-Anhydro-4-(a-~-glucopyranosyl) -D-glucitd ( = 1,banhydromaltitol) XI11 Amorph.b 326 +132b 133-134b 621 $ 82.0b Heptaacetate of XI11 H. G. Fletcher, Jr., and C. S. Hudson, THISJOURNAL, 70, 310 (1948). H. G. Fletcher, Jr., C. S. Hudson, ibid.,71, 3679 (1949).
+ + + + + + +
PROPERTIES O F THE
+ + + +
Solvent
[MID
HIO CHCl,
+ 9,550 + 2,500 + 1,200 + 8,070
H20
+lS,lOO
HzO CHCl,
Hz0 +43,000 CHCI, +50,900 L. H. Koehler and
TABLE I11 HYDROCLYCALS (1,5-ANHYDRO-2-DESOXYGLYCITOLS) Formula
M. p.,
OC.
Mol. wt.
[a]D
Solvent
[MID
Hydro-D-axabinal" XVIII Amorph." 118 -48.2" H20 -5690 Diacetate of XVIIIb Amorph? 202 -45." CHCls -9110 XV 67-68' 118 -44.9" HzO -5300 Hydro-D-xylal Amoiph.' 202 -38.8" C Z H ~ O H-7840 Diacetate of XV Hydro-D-glucal I11 86-87d 148 +16.5d Hz0 +2440 Amorph.# 274 +35, 6E C Z H ~ O H$9750 Triacetate of I11 H ydro-D-galactal IV 128d 148 $48.lC HzO +7120 4-(~-~-Glucopyranosy1)-hydro-~-g~ucal (hydrocellobial) IX 222' 310 4.5' H20 +1400 133-134' 563 +11.2' C2H4Clz $6310 Hexaacetate of IX 6-(~-~-Glu~opyranosyl)-hydro-~-g~ucal (hydrogentiobial) XX . . . . . . . 310 -10.2a H20 -3160 132-133' 563 $11.8' CsHiN $6640 Hexaacetate of X X 4-(~-~-Galactopyranosyl)-hydro-~-g~uca~ (hydrolactal) XIV 204-205A 310 +28 6h H 2 0 +8800 Hydro-D-arabinal has not been reported; its enantiomorph was made by G. E. Felton and W. Freudenberg, THIS JOURNAL, 57, 1637 (1935), and, with changed sign, we use here the rotational value reported by these authors. * See M. Gehrke and F. Obst, ibid., 64, text of present paper and also M. Gehrke and F. X. Aichner, Ber., 60, 918 (1927). E. Fischer 1724 (1931). H. Lohaus and 0. Widmaier, Ann., 520, 301 (1935). e E. Fischer, Ber., 47, 196 (1914). M. Bergmann and W. Freudenberg, ibid., 62,2783 (1929). * E. Fischer and and K. von Fodor, ibid., 47, 2057 (1914). G. 0. Curme, Jr., ibid., 47, 2047 (1914).
+
chloroform of +43.1° was reported by Gehrke and CH~OAC
CHtOH
H IV Hydro-D-galactal [MID$7,123
Aichner.6 These authors, however, failed to state to which series (D or L) their substance belonged and internal evidence was lacking since both Dand L-arabinose derivatives are reported elsewhere in the same paper; the present authors have therefore repeated this preparation. Diacetyl-D-
H
V Tetraacetyl-2-hydroxyD-galactal CH20H
CHzOH
I
I
I
CHzOH CHzOH CHzOH X VI11 IX 1,5-Anhydro-4-(P1,5-Anhydro-4-(P-~- Hydrocellobial 1400 D-glucopyranosy1)glucopyranosy1)-D- [.WID glucitol D-mannitol [ M I D =+9550 (unknown) (G1 = B-D-glucopyranosyli
+
H
H
VI 1,5-Anhydro-~-talito1 [ M I D -1,870
H
OH
VI1 1,5-Anhydro-~-galactitol [MID +12,600
(6) M. Gehrke and F. X. Aichaer, Bey., 60,918 (1927).
THE ANHYDRIDES
Nov., 1949
OF THE
GLYCITOLS AND RELATED SUGARS
3685
H
I I H OH Methyl cr-D-xylopyranoside [MID = +25,260
H
H XVI
H
OH OH XI (Table I) 1,5-Anhydro~-erytbro-4-descxypentitol ( = Hydro-L-arabinal) CHzOH
OH
1,5-Anhydro-xylitol meso, [MID = 0
I
I
HO
'
H/&
HO
CHzOH H OH X I1 (Table 11) 1,5-Anhydro-4- (@-D-galactopyranosyl)D-glucitol ( = 1,5-Anhydrolactitol) CHzOH CHpOH
I
\
H OH Methyl P-D-xylopyranoside [MID = -10,750
H
0, , / L O d h H
H
H
xv
H
H
Hydro-D-xyla1 ( = Hydro-D-lyxal) [MID = -5,300
I
H OH H OH XI11 (Table 11) 1,5-Anhydro-4-(cu-~-glucopyranosyl)-~glucitol ( = 1,5-Anhydrornaltitol) H OH CHpOH
H Methyl a-D-lyxopyranoside [MID = $9,750
XVII 1,5-Anhydro-~-lyxitol (= 1,5-AnhydroD-arabitol) [MID = -13,'200
H H CHiOH H H Methyl P-D-lyxopyranoside XIV (Table 111) 4-(B-~-Galactopyranosy1)-hydro-~glucal ( = Hydrolactal) [ h f ] D = -21,030 Fig. 1.-Some rotatory relations in the D-xylose and D-lyxose series.
arabinal, prepared from 2,3,4-triacetyl-~-arabino-tween those of the two corresponding I ,5-anhydrosyl bromide in 66% yield according to the method pentito1 acetates, triacety1-ll5-anhydro-~-arabitol of Karrer and co-workers' was hydrogenated, (- 19,300) and triacetyl-1,5-anhydroribitol(0). following the directions of Gehrke and AichnerI6 This relationship between the molecular rotayield diacetylhydro-D-arabinalas tions of a hydroglycal and its t w o corresponding to give in €%yo a colorless, viscid liquid boiling a t 0.2 mm. pres- 1,5-anhydroglycitols thus appears to be a general sure a t 90' and having n20D 1.4523. The specific one in a qualitative way and it may be said that rotation of this substance, [ a ]2 0 ~ ,in chloroform although the molecular rotation of a hydroglycal i s was found to be -45.1' (c, 3.02) and i t is there- not the exact mathematical average of those of the fore evident that the product having a rotation of corresponding epimeric 1,5-anhydroglycitols, i t lies +43.1° reported by Gehrke and Aichner was di- well between their greatly dijering values. acetylhydro-L-arabinal. In agreement with the If this generalization had been known a t an examples cited above the molecular rotation of earlier date, which was not possible because the diacetylhydro-D-arabinal (- 9,110) lies well be- pertinent data upon which i t is founded were not ( 7 ) P. Karrer. B. Becker, F. Benz, P. Frei, H. Salomon and P. then known, one could have inferred that the 1,5Schopp, Helu. Chim. Acta, 18, 1435 (1936). anhydrohexitol which Freudenberg and Rogers2
HEWITTG. FLETCHER, JR., AND C. S. HUDSON
3686
Vol. 71
H
1
/
\
\
XIX
oH
I
H H/fo\OCH,
I
I
I
OH H XVIII Hydro-D-ribal
[)MID =
I -5,690
CHzOH
I 1,5-Anhydro-~-glucit 01 ( = Polygalitol) [MID = f6,950
OH OH Methyl P-D-ribopyranoside [,WID = -18,650
( = Hydro-D-arabinal)
H OH Methyl a-D-glucopyranoside []WID +30,860 HO
1,5-Anhydro-ribitol meso, [MID = 0
H
'
I
OH OH Methyl a-D-ribopyranoside (not reported)
f
CH20H
H OH Methyl P-D-glucopyranoside [MID = -6,640
H
H I11 H ydro-D-glucal ( = Hydro-D-mannal) [ M I D $2,440
H H/LP0\H
FH1OH HI / ! T O \
J. H OH H Methyl a-D-arabinoDyranoside [MID -2,840 H I
1,5-Anhydro-~-arabitol (= 1,5-Anhydro-~lyxitol) [MID = -13,200
Fig. 2.-Some
H A P
1-1
OH H Methyl 8-D-arabinopyranoside [MID = -40,300
rotatory relations in the D-ribose and arabinose series.
D-
synthesized through the catalytic reduction of tetraacetyl-2-hydroxy-~-galactal (IV) was of the D-talitol rather than D-galaCtitOl series since the 1,5-anhydrogIycitol which these authors obtained had a molecular rotation of -1,870 and was thus more levorotatory than hydro-D-galactal (IV) (+7,120). Actual proof of the configuration of the substance in question as 1,5-anhydro-~-talito1 finally came through unequivocal independent syntheses of 1,5-anhydro-~-talito18and 1,sanhydro-~-galactitol.~ From the above i t is apparent that should the (8) D. A. Rosenfeld, N. K. Richtmyer and C. S. Hudson, THIS 2201 (1948). (9) H. G. Fletcher, Jr., and C.S. Hudson, ibid., 70, 310 (1948).
JOURNAL,70,
1,5-Anhydro-~-mannitol ( = Styracitol) [MID = -8,340
H
H
Methyl 8-D-mannopyranoside [MID = -13,550 Fig. 3.-Some
rotatory relations in the D-glucose and Dmannose series.
catalytic reduction of an acetylated Z-hydroxyglycal ever give rise to both of the expected 1,5anhydroglycitolslo each of the two substances may a t once be assigned its proper configuration, solely on the basis of its rotation and without recourse t o comparison with the rotation of the corresponding hydroglycal. The foregoing discussion has been limited to derivatives of the monosaccharides; those of some (10) The palladium-catalyzed reduction of tetraacetyl-2-hydroxyD-glucal has, for instance, been shown (ref. 6c) to produce a Small quantity of 1,5-anhydro-~-glucitolalthough 1,5-anhydro-o-mannito~ is the predominating product. Hockett and Conley (ref. 5d) have indicated that the nature of the catalyst employed may influence the proportions of products formed.
THE1,S-ANHYDRIDES
Nov., 1949
OF THE
3687
GLYCITOLS AND RELATED SUGARS CH20H
CHzOH
CHgOH
CHzOH
I
I
H OH Methyl a-D-galactopyranoside [MID = +38,050 H
CH"/.
H
-I
\
RO
H OH Methyl p-D-galactopyranoside [MID = 0
CH2OH
CHZOH
VI11
VI1 1,5-Anhydro-~-galactitol [MID +12,600
t
H OH Methyl 1-cellobidpyranoside [MID = +34,490
1,5-Anhydro-4-(&~glucopyranosy1)-Dglucitol [MID = +9,550
t
I I H OH Methyl p-cellobiopyranoside [MID = -6,810
CHzOH
H
H
IV
H
Hydro-D-galactal (= Hydro-D-talal) [MID f 7,120
.1
CHgOH
H
IX Hydrocellobial ( = Hydroepicellobial) (R = fl-D-glUCOpyranosyl ) [MID = $1,400
HOL /I -".\
CHzOH
I I H H Methyl a-D-talopyranoside (not reported) CHzOH
p'
H/H
I.
H
VI
Methyl a-epicellobiopyranosid? [MID = +16,400 CH20H
1,5-Anhydro-~-talito1 [MID = -1,870
OH \!-
I
. -0.
HO/A
H
I'
H H Methyl 8-D-talopyranoside (not reported) Fig. 4.--Some
CH20H
rotatory relations in the D-galactose and talose series.
D-
disaccharides will now be considered. In no case are data now known for an epimeric pair of the pertinent 1,S-anhydro derivatives. The hydroglycals from cellobiose, gentiobiose and lactose were synthesized many years ago ; the data concerning them are shown in Table 11. Maurer and PIotnerll synthesized 1,j-anhydro derivatives in the cellobiose and gentiobiose series through the palladium-catalyzed reduction of the respective heptaacetyl-2-hydroxyglycals, and though they designated these anhydrides as substituted styracitols (i. e., as derivatives of D-mannitol by present knowledge) i t is apparent that there was no just (11) K. Maurer and K. Plotner, Ber., 64, 281 (1931).
X 1,5-Anhydro-4-(p-Dglucopyranosy1)-Dmannitol (not reported)
Fig. 5.-Some
I
H H Methyl P-epicellobiopyranoside [-WID = -17.700 rotatory relations in the cellobiose and epicellobiose series.
basis for excluding the possibility that they were substituted polygalitols. 'Indeed, a consideration of the pertinent rotatory relations as developed previously in this article leads clearly to the inference that these two l15-anhydroderivatives] one from cellobiose and one from gentiobiose] are substituted polygalitols. I n each case the substituted 1,j-anhydride is more dextrorotatory than the corresponding hydroglycal (see Tables I1 and III), just as polygalitol is more dextrorotatory than hydro-D-glucal. All doubt concerning the allocations has been removed lately through the syn-
HEWITTG. FLETCHER, JR., AND
3688
CHzOR
H OH Methyl a-gentiobiopyranoside [MID = +23,340
CHzOR
I
I H
OH
XXI 1,5-Anhydro-6-(8-~glucopyranosyl)D-glUCitOl [MID +1,200
t
,
H
OH Methyl B-gentiobiopyranoside [MID = -12,800
CHzOR
H
xx
c. s. HUDSON
Vol. 71
syl) -D-glUCito112 affords an analogous example ; as will be seen from Tables I1 and I11 the anhydride (+16,loo), a substituted polygalitol, is more dextrorotatory than the long known hydrolactal (+ 8,800). With the data which are available in Tables I, I1 and 111, and assuming isorotation, i t is possible to predict the sign and order of magnitude of the rotations of some 1,5-anhydrides which have not as yet been reported. Thus, using the molecular rotation of the known 1,5-anhydro-4-(@-~-glucopyranosy1)-D-glucitol(+ 9,550) and themolecular rotation of hydrocellobial (+1,400), i t is simple to calculate the molecular rotation of the unknown epimeric anhydride, 1,5-anhydro-4-(@D-glucopyranosyl)-D-mannitol, as 2 X 1,4009,550 = - 6,750, corresponding to a specific rotation of -21'. Alternatively, the difference between the epimers polygalitol and styracitol (+6,950 - (-8,340) = +15,290) may be used to calculate a value of 9,550 - 15,290 = 5,740, corresponding to a specific rotation of -17' for the same anhydride. Attention is next directed to the Figs. 1-6.13 I n the left hand column of each of these figures one will observe that the arrangements portray the relationship that has been shown between the molecular rotation of a hydroglycal and those of the members of the related epimeric pair of 1,s-anhydroglycitols. An inspection of the two combined columns of these figures leads t o a new generalization which may be stated as follows: the molecular rotation of a 1,5-anhydroglycitol lies well between the molecular rotations of the anomeric methyl glycopyranosides of the corresfionding aldose. Each methyl a-D-glycopyranoside is much more dextrorotatory, and each methyl P-D-glycopyranoside is much more levorotatory, than the corresponding 1,5-anhydro-~-glycitol. This generalization is not a purely empirical one; i t has the same type of basis upon the isorotation hypothesis as does the relationship between the rotation of a hydroglycal and the corresponding 1,5-anhydroglycitol. From Figs. 1-6 it will be seen that the molecular rotation of a 1,5-anhydroglycitol is not the exact mathematical mean of the molecular rotations of the corresponding methyl glycopyranosides but that it does lie so well between them that the generalization can be applied without uncertainty.
-
H
Hydrogen tiobial ( = Hydroepigentiobial)
CHzOR
(R = 8-D-glUCOpyranOSyl) [MID = -3,160
CHzOR I
H H Methyl a-epigentiobiopyranoside (not reported) CHzOR H 1,5-Anhydro-6- (B-D-glucopyranosy1)-D-mannitol (not reported) H H Methyl b-epigentiobiopyranoside (not reported) Fig. 6.---Sorne rotatory relations in the gentiobiose and epigentiobiose series.
Summary thesis of authentic 1,s-anhydro-4-(0-D-glucopySome relationships between molecular rotation ranosyl) -D-glucitol by the reductive desulfurization of phenyl 1-thio-&cellobioside heptaacetate and configuration among the hydroglycals, the 1,5-anhydroglycitols and the methyl glycopyranoand of 1,5-anhydro-6-(~-~-g~ucopyranosy~)-~-g~ucitol from phenyl 1-thio-p-gentiobioside heptaace- sides have been pointed out. MARYLAND RECEIVED MAY5, 1949 tate in like mannerag The substances proved to BETHESDA, be identical with those discovered by Maurer and (12) H. G . Fletcher, Jr., L. H. Koehler and C. S. Hudson, THIS Plotner; they are substituted polygalitols rather JOURNAL, 71, 3679 (1949). (13) The molecular rotations of the methyl glycopyranosides in than substituted styracitols. 1 to 6 have been tsken from the book by F. J. Bates and AssociThe recent preparation of an anhydride in the Figs. ates, "Polarimetry, Saccharimetry and the Sugars," U. S. Govt. lactose series, 1,5-anhydr0-4-(P-~-galactopyrano-Printing Office, Washington, D. C., 1942.
