I694
WARREN C. VOSBURGH.
There may be some question of the accuracy of results obtained by this method, because of the possible presence of other reducing substances, and the fzct that the error is multiplied by such a large factor; but it seems fair, a t least to consider that they represent proximate maximum values. The difference between the values for “Total carbohydrates extracted by water and 1% HC1” and for “total carbohydrates by difference” would indicate the presence of hemicellulose-like substances. oreover, the structure oi the artichoke makes this seems the more probbab4e. It has been shown‘ that the human digestive tract secretes no enzymes capable of digesting either hemicellulose or inulin, and there is some evidences that inulin is not utilized by the animal body. There is some reason to doubt, therefore, whether the larger part of the carbohydrate of the glbbe artichoke is in a form that can be utilized by the human organism. BERHELEY, CALIF
[CONTRIBUTION FROM THE
DEPARTMARNT OF CHEMISTRY, COLUMBIA UNIVERSITY,
No. 336.1
THE SPECIFIC ROTATION OF FRUCTOSE. BY
WARREN
6.vOSBURGH.3
Received May 21, 1920
In the course of an investigation of the determination of the amount of sucrose and invert sugar in their mixtures it was found that considerable disagreement exists in the literature as to the specific rotation of fructose, For example the specific rotation for a Concentration of 5 g. per 1o(3 cc. (4.1337~by weight) according to several observers is as follows.
b12,O
bl?
89.64
86.84 Jungfleisch and Grimbert“ 89.40 86.04 Honig and Jesserb 92.45 89.65’ Ostd Nelson and Beeglef 91 .oe 88 .o a Jungfleisch and Grimbert, [ a ] ; = -(1o1.38-0~56 t 4- 0.108 (c- IO)), Comfit. rend., 107,390 (1888). Konig and Jesser, [ a ] ; = -(88.r3 -0.2385 p 4- 0.6714 (t-- zo)), J. Deut. Zuckeri 1028 (1888). Using the temperature coefficient of Jungfleisch and Grimbert. Ost, [a12 = -(g~.go 0 . 1 1 1 p)$ Ber., 24, 1636 (1891). e Interpolated between 15’ and 2 5 ’ . Nelson and Beegle, Tnrs JOURNAL, 41, 559 (1919).
’
+
1 Cf. M. D. Swartz, Trans. Conn. Acad. Sci., 16, 297 ( I ~ I I )for , a review of this literature. 2 R. Okey, J. Bioi. Chem., 39, 149 (1919). 8 Xationzd Research Fellow.
SPECIFIC ROTATION OF FRUCTOSE.
I697
There is disagreement also as to the temperature coefficient of the specific rotation. In order, therefore, to obtain accurate values of the ecific rotation for different concentrations a t 2 5 ' which were needed the above mentioned investigation, it was necessary to make several measurements for various concentrations. In this investigation it has been found that: (I) The specific rotation of fructose a t 2.5' can be expressed, in terms of the per cent. by weight, p j and the concentration in grams per IOO cc., c, respectively, by means of the equations, [ a ] g= -(88.50 4-0.145 p) (11 [a/: = -(88.50 0.150 c - 0.00086 cz). (II) These equations are based on experimental results for the range p = 2.6 t o p = 18.6 and c = 2.6 to c = 20. ( 2 ) The temperature coefficient of the specific rotation is a function of the concentration. (3) The relation between specific rotation and temperature for temeratures between 15 ' and 37 O can be expressed as [a]b = [CY]: f (0.566 f Q.0028 C) (i! - 2 5 ) . (110 (4) The values for the specific rotation found in this investigation are from 2' to 3' higher than those calculated from the formulas of Jungfleiseh and Grimbert, and Honig and Jesser. (5) The formula of Ost gives values higher a t concentrations below TO% and lower above 10% than the specific rotations found in this investigation. (6) The equation [cy]$ = -(g1.50 f 0.133 p ) expresses Ost's results better than the equation given by him. (7) Ost's experimental values for the specific rotation for concentrations 10% are in good agreement with the values calculated from EquaI and I11 of this investigation. wification of the Fructose.--Five different lots of fructose were purified by recrystallization from strong acetic acid solution as recommended by Hudson and Dale' for the purification of glucose. The purified fructose was dried for several hours a t 40' in vacuo over conc, sulfuric acid and preserved in glass-stoppered bottles in a desiccator O V ~ K cone. sulfuric acid. Preparations I and 2 resulted from recrystallization of some fructose previously purified in this laboratory by Nelson and Beegle.2 Preparation 3 was obtained by 2 recrystallizations of some commercial fructose. reparation 4 was the result of a third recrystallization of a portion of
+
1
Hudson and Dale, THISJOURNAL, 39, 320 (1917) LOC. E i t .
I 698
WARREN C. VOSBURGH.
Preparation 3. The good agreement of the specific rotation of this p r e p aration with that of the other preparations is good evidence of the purity of the material used in this investigation. Preparation 5 was the result of 3 recrystallizations 01 fructose similar to that used in Preparation 3. Procedure.--The required amount of fructose was weighed, dissolve in distilled water and made up to IBO cc. at the temperature of measurement in a calibrated volumetric flask. The solution was then weighe in order that the per cent. by weight could be calculated in addition t the volume concentration. In some of the experiments of Tables 116 and I V the solution was made up a t one temperature and the rotation measured a t another. In that case the density was determined a t the temperature in question as well as at the temperature a t which the solution was made up to volume and the required concentration calculated. Ala weights were reduced to a vacuum in the calculations. A11 solutions were allowed to stand until mutarotation was comp1ete.I The angular rotation of a t least 2 samples of each solution was then measured and the mean value taken as the angular rotation of the solution. For the measurement of the angular rotation a Schmidt and HaendB polarimeter sensitive to 0.01' was used. The solution being measured was maintained accurately a t the desired temperature by the use of the thermostat described by Nelson and Beegle.2 The polariscope tubes were measured and found to be 199.94 mm. and 200.39 mm. in length respectively and all observed angular rotations given were corrected to apply to B a o o mm. tube. At least 4 polariscope settings were made for each sample in the manner recommended by Browne.a Two different light sources were used in the polariscopic measurements. Most of the measurements were made with the use of a sodiu lamp, the light from which was not purified. The other source of light which was used in the experiments oi Table V, was a mercury-vapor lam the light from which was purified by passing through a Wratten No. 74 filter, giving light,of wave length X = 546ppl. The Specific Rotation at 25".---Determinations were made as given !,a Table I using sodium light. The first column gives the weight of fructose in IOQ cc. of solution. The second gives the density as determined by hing the IOO cc. of solution. The third column gives the composition in per cent. by weight calculated by dividing the weight per 100 cc. by 1 A calculation from the results of Nelson and Beegle, Zoc. cib., shows that the mutarotation of pure @-fructosewould be 99.9% complete after 84 minutes a t xg a atter 35 minutes a t 2 5 a t the hydrogen ion concentration giving minimum velocity. a LOC. C i t . 3 Browne, "A Handbook of Sugar Analysis," IQIP, p, 106.
tlic d m d t y of the solution The figures in die iourth column refer to the particular preparation of inictose used. The fifth colrimn gives ~ i i e aiiguiar rotation, and the sixth the 5pecif~crotations calculated from the oSserved. rotations and the volume concenlnations. TABtlG 1, Specific Rotntiau of Fructose at 25'. Cotle. in g" :!oocc.
2 632 2.632 I
Fructose.
Gbserved rotatian.
2.61
I
- 4.68
2.61
2
-- 4.67
t
.,
-- 4.69
I
-- 8.94 -.--8.91
Density.
% by wt.
I 0073 I .QQ71
I
2.632
I ,0062
2.62
-5 ,003 5.003 j ,003
1.0164 I .a160 I .or60
4.92 4.92 4~92
2
j ,265
s
5.18
1 2
,0170
3
5.255 7.900 7 . 900 IC.006
I .017g
5.18
..
.._
I
3
XO.006
1 .Q349
' I O .006
1 .oggo
15.wjog
1.0539
7.69 3.67 9.67 9.63 14.24
'is ,009
..
20.012
Y ,0729
I I
,0268 ,0348
.
I
.
18.65
8.91 9.42 9.39
-i
--14.. 16 -14. I6 --r7.99 --IS.g8 ---I 7 .9') -27. I8
3 4
---27. z I --36.48
I 2
4
3 s t t and Honig and Jeaser' have expressed ihe specific rotation of fructose as a limar Gunction of the per cent. fructose by weight, while Jun$%xsch and Grimbertl express it as a linear function of the volume concentration However if the per centa by weight is plotted against the volume concentration a curve 4s obtained which is not quite a straight line sl:owing that there is not a direct proportionality between the ~ K V O . specific rulat;on therefore camot be a linear hinctioim of both. When the specific rotation was plotted against both per cent. and concentration 1. was found that in the fornzer case the observed poirits lay more nearly on L% straight line than in the latter. ?"ne relation between specific rotation and per cent. Imctose, p , is shown graphically i~Fig. I , the values for the rpecifx rotation being the inem valurs given in Table 11. The e c y a t h ~of the straight line in1 Fig. I is [a]: --- --(88.50 -f- 0.145 3). (1) Xz r t is often more cenvenient i o deal mitli conceutratioris by volume an q u a t i o ~expressing the specific rotation a5 a function of the concentration, F, in g. per POO ce. was derived by the method of least squares, from t91e vaL2es in Table 11. I5 - - ( 8 8 . ~ i s 4- 0.150 L - 0.00086 c2). Ia]r> (11)
' xuc
a:t
T h e s e equations
obviously have not been experimentally verified outside the limits 9 = -3' 2.6 to p = 18.6 in I, k P and c = 2.f) to c = 20 in 6 -so k 11. 'Cable I I gives a si1111mary of the mean valiiei 2 -89 oi the ohservcd specific 2 rotations as given i i i '3 -88 Table I, together with v a 1u e s calculated by means of Equations I 2 5 a \I I4 17 and I1 for the same coli PL-RCENT fRUCTOSE centrations. The equaFIG. 1.-Variation of the specific rotatioil of fructose tions are shown to -92
g
8
with conceutration. t losely
values which agree very
with the experimental values.
TABLB11. Coiiiparison of the Observed Specific Rotation with that Calculated from Equations I and 11. Conc. 6.
ill
/IO0 cc.
- - ~ _ _Specific rotation % by
wt.
15.009
2.61 4.92 3.18 7.69 9.67 14.24
20.012
18.65
2.632
5.003 5.265 7.900 10.006
Observed.
-88.91 -89.15 -89.32 -89.62 -89.89 -90. Go --91.14
Calc. I.
-88.88
-89.2 I -89.25 -89.62 -89.90 -m. 56 --g1 . 2 0
ut 2S0
_______ Culc. 11.
--88.89 -89.23 -89.27 -89.63 -89.91 -90.56
31. 16
Temperature Coefficient.-There is considerable disagreement also as to the temperature coefficient of the specific rotation of fructose. Jungfleisch and Grimbertl give 0.56 as the temperature coefficient while I-Ionig and Jesser' found 0.671 for a 9 % solution and 0.692 for a 23.5((solution. Nelson and Reegle' found a difference of (3.0' in the specific rotations a t 15' and 25' and a difference of 7.0' between 25' and 37'. Assuming a constant coeficient over this range their temperature coefficient is 9.59' per degree centigrade, determined for a concentration of 5 g. per 100 cc. 'fhe temperature coefficient was therefore determined by means of the experiments given in Table 111. The temperature was maintained a t the figures given within *o.o~'. 'l'he solutions were made up at t'lw 1
Lac. Cil.
~ ~ ~ ~ e r a taut which r e s the measurement was to be made except in case o€ the measurements at oQ, 15" and 20' in which case the 2 temperatures were determined and the concentration at the temperature si measurement calculated. In ail cases care was taken to allow the solution to s t m d at the temperature of measurement long enough for ~ ~ ~to be i completed ~ a r~ e a ~s ~ ~~~ were ~ before e m~ emade. ~ ~~ s
~
TABLG 111. Specific Ratation of Fructose a t Various Temperatures.
..J emperaturr. "G
5'. per. 100 cc.
Observed sotatioo.
0.2
5.038
15.3
37.1
5.039 5.003 5.003 5.003
--15.34 - 9.55
0.2
ao.o8o
25
.o
30.0
- 8.92
Specific rot.a?ion. ,.-. _ L Observed. Ca1culated.a
---1O2.63
- 94.76 .-
89.15 86.15
---103.6s 94.78
-
. I
UiSierence, --T.02 ---0.02
."
- 86.25 - 8a.x~
---o,Io
- 8.24
- 82.35
-20.89
-103.63
-1oq.75 - 9.5.84
-1.12
_ I
8,6a
15.0
10.080
20.5
10.028
25.0
ro.006
30.0
10.506
-19.32 --.r8.64 -17.99 --17.39
37.1
1o.oo6
-16.57
[a]i= [a]: + (0.566 -2- 0.0028
_95.83
-- 92.94
-- 89.90
- 86.90
- 82.80
--
92.87
..
- 86.93 -'
82-73
$0.20
+O.OX
+o.o7 1 "
1-43.03 +O,O'b
c](t -25).
When the observed specific rotations are plotted against temperature between 15' and 39" for both concentrations, as .in Fig. 2, two straight lines are obtained which are not parallel, showing that the temperature coefficient does not vary with teaperattire over this range but that it does vary with concentration. The slope of the specific rotation-temperature line for a concentration of 5 g. per IOO cc., Curve IT, i s d [ a ~ ] b / d=t 0.580 and for a concentration of T O g. per IOO cc., Curve I, ~ ~ [ =; 0.594. ~ ~ Assuming ~ ~ the~difference / d in slopes ~ to be ~ r o p o ~ t ~ o n ~ ~ to the diflerence irr concentrations' the difference for one g. difference in ~ o ~ s c e n t ~ ais t ~ '/eo n (0.594--0@0) or 0 . 0 0 ~ 8and the slope for zero con~ ~ ~ ~ ris a t i ~ n (5 X 0.0028) I 0.566. d[a0] h / d t -- 0.580 ~ h e r e ~ for o ~ any e concentration d[or]k/di -. (0.566 0 . 0 0 ~ 8c). ~ ~ t ~ ~ r a t ~ ~ ~ , [a]EL)= (0.566 0.0028 c) t -1- C. When I. 25', [a]L = [ a ] g #and C = [ a ] E - 25 (0.566 3-0.0028 c). Therefore [a&, = [a]$ (0.566 0.0028 c> (t - 2 5 ) . (W The values for the specific rotation under the conditions of the experi-
+
+
-i
-+
+-
1 This assumption is sufficient for the purpose if only roughly true since the correction t o t h e temperature coefficient is small and a large error in the correction w o d d makc little elifhence in the result.
~
merits of Table JII were cd- lculated by meaos of this eyuation and are given in Table Iir. The maximum difference hctween observed and caladated, except at 0.2' is 0.2' or 2.4 parts pes thousand, corresponding to an error in the observed rotation of 0.02'. The equation evidently does not hold at 0.2' and C Q ~ W quently cannot be depended UPOII much bdQW.'$K eserlts of Other ~b~~~~~~~ --Comparison of the specific rotation for a concentration sf 5 g. p r loo cc. as given in -85 Table IP with the values given -114 by the formulas of Ju -83 and Grimbert, and Ik Jesser, as given above, shows -82 35 37 that the latter values are 2' to 15 20 25 3' too low, as Qst has pointed TEMP€RATUR€ FIG.z.--~ariation oi the specificrotation OE out.' fructose with temperature. Curve I, concenThe specific rotation as given tration of fructose, ro g. per roo cc. Carve bj7the formula of ostis toohigh IT, concentration of fructose, 5 g. per 100 cc. for concentrations up to about 10% and too low above this point as compared with the results of this in vestigatioxn , as the following figures show. % Fructose [a]z,oOst. . . ,
.. .. . . . . . . . ~. [a]? I and 111.. .. . ~., , . . .
5.
92.45 92.13
10
93.01 92.93
15.
93.57 93.73
20.
94-12 94 * 53
The specific rotation increases too slowly with concentration in the equation proposed by Ost and the value for zero concentration is too high if the results of this investigation are correct.. An examination of Ost's experimental results shows that they can he expressed as well or better by the equation [ c x ] ~= - (91.50 0.133 p). Moreover values calculated from above Equations 1 and 111 agree with Ost's experimental values up to the 10% solution, and the values for 5 arid I Q % given by Eqsnations I and 111 are closer to the experimental values than those
+
calculated from either of the other equations. Table IV demonstrates these facts. The first 3 columns are taken from Ost's paper and tlae last 3 were calculated by the author. 1
Lcl-. cit.
1703
SP$CEiW ROTATIOX OF FRUCITOSlZ. TABLB
iv. Vakes
Ost's Observed Values far the Specific Rotation of Fructose Compared with
Calciilated by Various Enipiricsl Formulas. Observed rotation.
-
3.67
- 3.74 -. 7 . 3 7 - 7.49 - 18.55
-- 18.59 - 18.64 - 29.86 ~
31.50
-- 3 7 . 8 8
- 4.0. 7 2 -- 7 4 6 1 - 82.47
-------
Specific rotatiou
--
Calc. C (3).
Calc. A (1).
Caic. B (2)
-90.64
- 9 2 . 01
--91.63
-91.50
-90.36 -91.79
-92. ox -92.12 - 9 2 . I3 -92.45
--gx .64
-91
-91 77 -91.77
-91.63 -91.65 ---92.12 -92.12 -92. I 2 --v.57
Observed.
-91.84
-92.30 -92. T I -92.10 -92
. go
-92.96
-s)z. 92 -93.00
-92
. 4j
-92 ' 45 -92 I77 -92.90 -92 ' 99 -93.07 -93.95 -94.15 -95.20 --gj .24.
-93.73 -94.03 -128.23 -95.51 --129.60 -95.39 ( r ) [a]? = - (9r.go -I- 0.111 p ) ( 2 ) [el',"= - (91.50 -I- 0.133 P) (3) [a];: = - (88.50 o 14s f (0.566
~
-92.16 -q2. r6 -92.16 -92.54 -92.69 -92.81 -92.90 -93 " 96 -94.20 -95.46 -95.50
*
50
--ga.y5 -92.89 --93. OCI
-94.28 -94.56 .*. a * .
0.0028 C)(20-25)
~ i n g Eynon , and 1,aneI find [ a ] g 5= -193.83 when c = EO for some fructose which they describe as of high purity. Calculation by means of the abovc Equations rT and III gives [a]g5= -93.77, agreeing with tlii.; experimental value vithin one part in a thousand. easiired with the ~~~~~w~~~~~~ Line of Inercury Vapor.-After most of the above determinations 'CVPTIScompleted a meicury v a p r :amp was installed in this laboratory for use in polariscopic nieasinemeuts, a n d a determination was made of the I? atiq of 4 he specific rotations for the unpurified S Q ~ ~ U I X light ac; used in the measurements above and the line X = 5 4 6 , ~ obtained by passing the light from the mercury vapor lamp though 5 Wratten No. 74 filter. For this purpose the rotation of the solutions used for tbe determination of $he ten?,peratwre coeficient was determined by the use of mercury light a s avell a? sodium l!ght, the light S O I V C ~ being changed during the measurements ~ i t h ~ ckrengirg ut the .;olutiiins in the polariscope tubes. The ratio of the rotation usjag mercury light to that using sodium light was calculated and is the same as the ratio of the corresponding specific rotations. The specific rotation of frucl;ose with the line X = 546pp as the iight source can therefore be ealcu!ated lrom the above results by mut-ciplying by rbc proper factor. The experiment.1 results and factors to be used BW given in Table k', ~~~~~~~
~~~~~~~~
' Ling, ?3:ymoa
and bane, J SOC.C h i n P n d , 28, 730 (rgcq)
ARTHUR J. HILL AND ERWIN Be XELSGY* TABLE
Determination of the Ratio
--
, _ A -
5.038 5.039 5 .003 5.003
for Fructosee.
Iff1 ;
Observed rotation.
Temperature Conc. in. g. /lo0 cc.
v.
t l a [ X = 546pp
D line.
A
-10.34 - 9.55 8.92 8.62 8.24
-
546 p p .
-19.32 -17 99 -17.39 ---*6.j7
15 .o 20.0
546
546pp.
I . 1809
-11.27
0.8474 0.8455 0.8451 0.8443 0.8482 0.8492 0.8458 0.8479
E * 1801
- 9.96 -24.63 -22 " 75 -21.27
pp.
Variation from mean.
[a]~"
cp .a468
-10.20
-20.89
-
0
-12.21
---x0.55
5.003 10.080 10.080 10. 006 10. 006 10.006
1
7 t
I . 1828
I. 1833
1.1845 I . 1790
1.1775 I. 1823
-20.51 -19.55
0.84.76
1. I799
-27.09
-32.00
0 . 8466
I , 1812
-36.34
-42.87
0.8477
1.1797
*
1794
The ratios of the specific rotations using mercury and sodium lights respec~,iv~ly seem to vary with temperature at a concentration of 5 g. per roo cc. becoming slightly larger as the temperature increases, but in the case of IO g. per IOO cc. the ratio seems to be constant over the temperature range considered. If the values a t 25' for the different concentrations are considered a regular variation with concentration amounting to less than 3 parts per thousand is noticed. Hotvevvcr these variations are hardly larger than the experiniental errors and for most purposes the mean value of the ratio, namely 1.1809,or its reciprocal, 0.8q67, can be used. The average variation from the mean in the above figures i s about one part in a thousand while the extreme variation is 3 parts per NEW
YOWL,
N.
v. _-- -
_ I
[C0N*lRIBVTION S'RROM THE DEPARTMENT OF CHEMISTRY, YALE UNWI3RSXTY. 1
VIII.' THE PREPARATION GETANILIDE. BY
ARTHUR J,
HILI, AND ERWIN Received M a y 24, 1920.
B. KELSSP.
In order to develop certain phases of an extended research on thiocyanates and isothiocyanates, now in progress in the Sheffield Laboratory, it became necessary to obtain representatives of a series of amines of fhe general formula RNHCOCP-12NX-12, of which amino-acetanilide, ~ ~ ~ ~ ~ ~ ~ OisCtheX prototype. - I ~ N ~OurI atteulion ~ , was first turned to this latter compound, since it was evident that the principles underlying a successful method for its preparation would apply directly to the I Paper VTI, J . I d . Eng. Chcm., l a , 636 (IWO).
