Jan.

20, 1952 Reaction of d-Glucose, d-Mannosb and d-Fructose in Sodium Hydroxide 499

acetic acid after refluxing one-half hour also gave the un- 5-Alkyl-5-( ß-isothioureidoethyl )-barbituric Acids.—Five
changed substance. The acid (1.0 g.) was now heated in grams (0.013 mole) of 5-isoamyl-5-(jS-isothiouronium-
a sealed tube with 7.0 cc. of hydrochloric acid (1.19) at bromidoethyl)-barbituric acid was dissolved by warming
160° for two hours. Some pressure developed and hydrogen in 65 cc. of water. The solution was filtered immediately
sulfide was identified among the gases evolved. No definite and 1 cc. (0.015 mole) of aqua ammonia (0.90) was added
crystalline product was isolated from this reaction mixture. gradually to the filtrate with stirring while it was cooled
5-AUtyl-5-( /3-isothiouroniumbromidoethyl )-barbituric in an ice-bath. The white crystalline product obtained by
Acids.—In a typical experiment 15 g. (0.049 mole) of 5- filtration required no further purification, yield 4.0 g.
isoamyl-5-(/3-bromoethyl)-barbituric acid and 3.8 g. (0.050 (80%).
mole) of thiourea were periodically swirled with 50 cc. of Mercapto Acid from the Isothiouronium Compound.—
alcohol as it heated to the temperature of complete solution One gram (0.0026 mole) of 5-»-amyl-5-( ß-isothiouronium-
(80°). In five minutes at this temperature the rapid bromidoethyl)-barbituric acid was dissolved in the minimum
precipitation of the product began. When this had subsided amount (5 cc.) of ice-cold 10% sodium hydroxide. The
the temperature of the bath was raised to 90° and the mix- solution was allowed to stand for an hour in an ice-bath.
ture was heated for five hours. After cooling in ice the solid The 5-«-amyl-5-( /3-mercaptoethyl )-barbituric acid was pre-
product, obtained by filtration and concentration of the cipitated by hydrochloric acid, yield 0.6 g. (95%), m.p.
filtrate, weighed 17.8 g. (95%). It is very soluble in hot 132-133.5°. It was identical with the mercapto add pre-
water but quite insoluble in cold water and was purified by pared from the corresponding xanthate.
crystallization from water. Newark, Delaware Received August 13, 1951
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

[Contribution from the Radiochbmistry Laboratory, Department of Chemistry, Washington University]

Downloaded via TU MUENCHEN on November 30, 2022 at 13:50:27 (UTC).

The Reaction of D-Glucose, D-Mannose and D-Fructose in 0.035 N Sodium Hydroxide

at 35°
By John C. Sowden and Robert Schaffer1
The reactions of individual molar solutions of D-glucose, D-mannose and D-fructose in 0.035 N sodium hydroxide at 35°
have been studied using radioisotopic dilution analysis for D-glucose and D-fructose in the reaction mixtures and a corrected
phenylhydrazone precipitation procedure for D-mannose. After reaction times of four to eight weeks, summation of the
analyses for these three isomeric sugars accounted in each instance for only 77-80% of the starting carbohydrate. The re-
mainder was shown, by fermentation experiments, to have been converted nearly quantitatively to a mixture of non-ferment-
able sugar products. Some exploratory experiments were performed concerning the nature of these non-fermentable sub-
stances.

The complex reaction sequence promoted by the Conclusive evidence for the presence of material
action of aqueous alkali on reducing sugars includes epimeric with D-glucose at carbon-3 has been ob-
isomerization, fragmentation and fragment recom- tained recently, however, by characterization of
bination. Two kinds of isomeric products form: the products obtained by electroreduction of the
carbohydrates8 and saccharinic acids.3’4,6 The mixture that results from the interaction of D-glu-
products of fragmentation are themselves at the cose and aqueous alkali.11
oxidation level of carbohydrate and they also may The generally accepted mechanism of the Lobry
isomerize6 or recombine to larger molecules.7 In de Bruyn-Alberda van Ekenstein isomerization re-
addition, there are formed colored products of high action postulates enediol intermediates, as illus-
molecular weight and undetermined structure. trated in the scheme
Lobry de Bruyn and Alberda van Ekenstein
demonstrated the interconvertibility of D-glucose,
D-mannose and D-fructose in aqueous alkali. They 0=0 0=0
also isolated fractions from the reaction which they
considered to be, respectively, the 3-epimer of d-
HCOH \X
x
rR
1
1 A
r
HOCH
|
fructose (“pseudo-fructose,” “psicose,” D-ribohexu- A' C—OH R'
lose) and a mixture of 3-ketohexoses. However,
11

the former material does not conform in its proper- R ¿—OH R

ties to synthetic D-ribohexulose8·9 and their “3- hAoh j//f Li- J ho¿h
ketohexose” mixture (“glutose”) is now considered
to be a complex mixture containing fructosans.10 0=0 0=0
(1) Abstracted from the thesis of Robert Schaffer presented in
partial fulfillment for the degree Doctor of Philosophy, Washington
University, October, 1950. (R=H, CH,OH, etc.; R'=H,
CHaOH, etc.)
(2) C. A. Lobry de Bruyn and W. Alberda van Ekenstein, Rec. trav. Observations, based on the measurement of deute-
chim., 14, 203 (1895); 15, 92 (1896); 16, 257, 262, 274, 282 (1897); rium exchange when the reaction was conducted in
18, 147 (1899); 19, 5 (1900).
(3) B. Peligot, Ann. chim. ci phyi., [2] 67, 154 (1839). heavy water, apparently in conflict with the ene-
(4) C. Schiebler, Bcr., IS, 2212 (1880). diol mechanism have been recorded.12·18 These
(5) H. Kiliani, ibid., 16, 701, 2953 (1882). objections to the mechanism, however, have not
(6) J. U. Nef, Ann., 857, 294(1907); 876,1 (1910); 408,204(1913). been sustained in more recent studies with heavy
(7) B. Fischer and J. Tafel, Bcr., 80, 2566 (1887).
(8) Marguerite Steiger and T. Reichstein, Bclv. Chim. Ada, 19, 187 (11) M. L. Wolfrom and co-workers, This Journal, 68, 122, 678,
(1936). 1443, 2342 (1946).
(9) M. L. Wolfroin and E. F. Evans, This Jouhnal, 67, 1793 (12) H. Fredenhagen and K. F. Bonhoeffer, Z. physik. Chtm,, 4181,
(1946). 392 (1938).
(10) L. Settler and F. W. Zerban, Ini. Bag. Chtm,, 87, 1188 (1945). (13) X» Goto, Chtm. Sac. Japan, 88, 817 (1943).
500 John C. Sowden and Robert Schaffer Vol. 74

water.14·15 It also has been reported16 that the and purified by paper chromatography, was kindly sup-
mechanism of isomerization shows cationic depend- plied by Drs. M. Gibbs and S. Udenfriend.19
Amberlite I.R.-I00-A.G.—This product of Rohm and
ence in that different types of enolic intermediates Haas Co., Philadelphia, Pa., was put through several
are involved when alkalies containing, respectively, sodium hydroxide-hydrochloric acid cycles and then washed
mono- or divalent cations are employed. In the free of chloride ions prior to use.
Duolite A-4.—This product of Chemical Process Co.,
present work we have confirmed qualitatively Redwood City, Cal., was put through several hydrochloric
Kusin’s observation that the initial course of the acid-sodium hydroxide cycles and then washed thoroughly.
reaction is different in mono- and divalent bases of Before use, the resin was treated with several volumes of
0.5 N concentration. However, at lower concen- 5% sodium chloride solution and then washed free of
chloride ions.
trations (0.035 N) of alkali, we could detect no dif- The Isomerization Reaction.—Samples (180.00 g.) of
ference in the catalysis of the isomerization by cal- each of the three sugars were dissolved in freshly distilled
cium hydroxide or sodium hydroxide. water in tared 1-1. volumetric flasks. Sufficient 1 N sodium
Previous quantitative studies of the course of the hydroxide solution (35.0 ml.) was added to each to give a
final concentration of 0.035 N and the flasks were filled
isomerization reaction have been handicapped by a to the mark with water. After weighing, the resulting
lack of specific analytical methods for individual sug- solutions were transferred, together with 0.1 ml. of xylene,
ars other than D-mannose in complex mixtures. to 2-1. round-bottomed flasks, previously blackened to avoid
Analyses for D-glucose and D-fructose in the isomeri- photochemical effects20 and designed to deliver samples
zation mixture previously were based on methods by application of nitrogen pressure. The flasks were ther-
mostated at 35° and their air content was replaced with
generally applicable to the aldose and ketose classes nitrogen by repeated alternate evacuation (water-aspirator)
of sugars.17 With the advent of isotopic labeling, and refilling with washed nitrogen. Periodically there-
it has been possible to apply the method of radioiso- after, samples of the reaction mixtures were removed under
nitrogen pressure and aliquots of each sample were used for
topic dilution analysis to the course of the isomeri- determinations of D-glucose, D-mannose and D-fructose,
zation reaction. In the present work, the isomeri- acid-base titration, pH determination and measurement of
zation of molar solutions of D-glucose, D-mannose optical activity. Amounts of solution below 5 ml. were
and D-fructose in 0.035 N sodium hydroxide at 35° weighed; the remainder pipetted.
has been studied. The growth or disappearance of Standardization of Radioactive Sugar Solutions. 1-C14-
o-Glucose.—To a series of weights (15-80 mg.) of d-
D-glucose and D-fructose was followed by radioiso- glucose were added 0.050-ml. aliquots of an aqueous stock
topic dilution analysis while variations in the d- solution of l-C14-o-glucose containing about 2.5 mg./ml.
mannose content were measured by a corrected of radioactive sugar. After solution in water and mixing,
the resulting diluted l-C14-D-glucose samples were crystal-
phenylhydrazone precipitation procedure. In this lized, washed and dried as described below. Accurately
manner, by summation of the progressive concen- weighed amounts (ca. 2.5 mg.) of each sugar sample were
trations of D-glucose, D-mannose and D-fructose, dissolved in water on individual stainless steel dishes (12.6
there was detected in the reaction mixtures the sq. cm. in area) and 2 drops of a 0.1% solution of the
detergent "Santomerse” (a product of Monsanto Chemical
rapid growth of a sugar fraction containing none of Co., St. Louis, Mo.) added to each. By manipulation
these three hexoses. Fermentation experiments under an infrared lamp, the solutions were evaporated to
then showed that this fraction consisted nearly produce evenly distributed, thin layers of the sugar. The _

quantitatively of non-fermentable, non-acidic sugar specific radioactivities then were determined by timing
material. Similar non-fermentable substances have 10,000 counts from each sample in an R. C. L. nucleometer
(a product of Radiation Counter Laboratories, Chicago,
been obtained previously2 as products of consider- 111.). A single check determination was made similarly
ably more drastic treatment with alkali. Prelimi- with known D-glucose at the conclusion of the analyses of
nary experiments with this non-fermentable frac- the isomerization mixtures to demonstrate the stability of
tion have shown it to be a complex mixture, and a the stock l-C14-o-glucose solution. The results of the
standardization are shown in Table I.
detailed study of its nature is contemplated.
A quantitative study of the isomerization of d- Table I
fructose by aqueous alkali apparently has not been
Standardization of Stock 1-C14-d-Glucose Solution
recorded previously. Its inclusion here seems espe-
Specific Total
cially appropriate since, according to the enediol o-Glucose, radioactivity, radioactivity,
cts./min./mg. cts./min.
mechanism, it is the precursor of the 2,3-enediol Sample no. mg.
which in turn may give rise to non-fermentable 1 23.27 391.1 9100
aldo- and ketohexoses. 2 29.96 308.1 9230
3 40.20 230.6 9270
Experimental 4 51.06 179.6 9170
Materials. o-Glucose, o-Mannose and o-Fructose.— 5 64.64 142.0 9180
Commercial preparations of the sugars were recrystallized 6 78.26 115.5 9040
until they possessed physical properties in accord with the Check 44.12 210.2 9270
accepted values. 9180 ± 70
l-C14-D-Glucose.—The labeled sugar (ca. 0.05 µ ./mg.) Average
was prepared from D-arabinose via the nitromethane syn-
thesis with C14-nitromethane.11 It is apparent from Table I that the total radioactivity
Clt-D-Froctose.—Uniformly labeled C14-D-fructose (ca. found for each sample was essentially constant. Thus, the
1.2 ¿ic./mg.), prepared photosynthetically from C140, principle of isotopic dilution analysis is applicable and
quantitative analysis for the amount of the sugar in an un-
(14) Y. J. Topper and DeWitt Stetten, Jr,, J. Biol. Chem., 189, 191 known mixture to which the tracer has been added is given
<1951). by the quotient of the appropriate constant with the ob-
(15) J. C. Sowden and R. Schaffer, This Journal, 74, 505 served specific radioactivity of a sample of the sugar isolated
(1952). from the mixture.
(16) A. Kusin, Bet., 69, 1041 (1936).
(17) M. L. Wolfram and W. L. Lewis, This Journal, BO, 837 (1928); (19) S. Udenfriend and M. Gibbs, Science, 110, 708 (1949).
R. D. Greene and W L. Lewis, ibid., 60, 2813 (1928)· (20) A. L. Bernouli and R. Cantieni, Helv. Chim. Acta, 16, 119
(1.9) J. C, ftQWden, /, Biol. Chem., 180, 55 0940). (1982).
Jan. 20, 1952 Reaction of d-Glucose, d-Mannose and d-Fructose in Sodium Hydroxide 501

Cl,-D-Fructose.—A similar standardization to that de- and the solution immediately de-ionized again by ion ex-
scribed above for l-C14-D-glucose was made for a stock solu- change. After observation of the optical rotation, the
tion of C14-D-fructose (ca. 0.033 mg./ml.) with known d- solution was concentrated at reduced pressure to a sirup.
fructose. The results are shown in Table II. The latter was dissolved in absolute ethanol (1 ml./100 mg.
of D-fructose indicated by optical rotation), the solution
Table II allowed to stand for two days, filtered if necessary, and then
seeded with a minute amount of D-fructose. The crystal-
Standardization of Stock C14-d-Fructose Solution lized sugar was washed by decantation with absolute
Specific Total
D-Fructose, radioactivity, radioactivity, ethanol, recrystallized from the same solvent, and dried
Sample no. mg. cts./min./mg. cts./min. for assay at 65° in high vacuum over phosphorus pentoxide.
1 15.44 409.5 6320
2 23.19 277.5 6440
3 41.92 153.8 6450
4 50.25 126.9 6380
5 64.86 100.1 6490
6 79.86 79.6 6360
Check 44.12 146.4 6460
Average 6410 ± 50

Isolation of o-Glucose for Isotopic Dilution Analysis.

From the n-Glucose Isomerization Reaction.—To each
aliquot of the reaction mixture, estimated to contain 100-400
mg. of D-glucose, was added 0.250 ml. of the stock radio-
active D-glucose solution. The solution was passed, with
washing, through Amberlite I.R.-10O-H (0.5 g./ml. of
reaction mixture) and the effluent was evaporated at reduced
pressure. The resulting sirup was dissolved in 10 ml. of
95% ethanol and, after standing overnight, the solution was
filtered, concentrated to 3-4 ml., and seeded with a few
minute crystals of D-glucose. After crystallization was
complete, the supernatant liquid was decanted and the
residual D-glucose was washed, ground and rewashed with
Fig. 1.—Yields of D-mannose phenylhydrazone from d-
95% ethanol. Recrystallization was performed from 3-4 mannose in the presence of D-glucose and D-fructose, total
ml. of 95% ethanol. Before assay, the sugar was dried at
105° in high vacuum over phosphorus pentoxide. sugar, 1.8 g. in 25 ml. of solution: A, D-mannose and d-
From the o-Fructose and o-Mannose Isomerization glucose; B, D-mannose and D-fructose.
Reactions.—Each alkaline reaction aliquot, estimated to
contain 100-400 mg. of D-glucose, was treated with 0.250
ml. of the stock radioactive D-glucose solution and then
neutralized with acetic acid. D-Mannose phenylhydrazone
was precipitated as described below and filtered off. The
filtrate, after addition of 0.5 ml. of benzaldehyde and 0.5 g.
of benzoic acid, was refluxed for 30 minutes, cooled, ex-
tracted three times with chloroform, and concentrated at
reduced pressure to a volume of 4-5 ml. Two grams of
sodium bisulfite was dissolved in the concentrate and 20 ml.
of absolute ethanol then added gradually. After several
hours, the precipitated bisulfite addition compound11 was
centrifuged off, dissolved in 25 ml. of water, and the solution
de-ionized by successive passage through 30 ml. of Amberlite
I.R.-100-H and 40 ml. of Duolite A-4. A total of 100 ml.
of effluent was collected. If the optical rotation of the
effluent was negative, bisulfite precipitation and de-ioniza-
tion were repeated. D-Glucose then was isolated and puri-
fied for assay as described above for the D-glucose isomeriza-
tion reaction.
Isolation of o-Fructose for Isotopic Dilution Analysis.—
To each aliquot of reaction mixture, estimated to contain
100-400 mg. of D-fructose, was added 0.250 ml. of the stock
radioactive D-fructose solution. For each 10 ml. of reaction
mixture, there then was added 65 ml. of water, 6 g. of
barium benzoate, 1 g. of benzoic acid and 0.6 ml. of bro-
mine.11 Oxidation was allowed to proceed at room tem-
perature in the dark for a minimum of 40 hours. Following
removal of barium by precipitation with sulfuric acid and
benzoic acid by extraction with chloroform, the solution
then was passed through 60 ml. of Duolite A-4. The
residual sugar acids and lactones in the effluent were
titrated at 0° with 1 N sodium hydroxide to the phenol-
phthalein end-point, acidity was just restored with 1 N
hydrochloric acid, and the solution then was concentrated
at reduced pressure to a volume of 25 ml. Following de-
ionization with Amberlite I.R.-100-H and Duolite A-4,
the titration with sodium hydroxide at 0° was repeated,
the solution was again made just acid with hydrochloric acid
and the concentration continued to a volume of 5 ml. Fig. 2.—Yields of D-mannose phenylhydrazone from d-
Alkalinity was just restored with sodium hydroxide at 0° mannose in the presence of D-glucose and D-fructose, total
sugar, 1.8 g. in 25 ml. of solution: A, D-mannose and d-
(21) P. A. Ashmarin and A. D. Braun, Byull. Rksptl. Biol. Med., 4,
374 (1937). glucose; B, D-mannose and equal amounts of D-glucose and
(22) C. S. Hudson and H. 8. Isbell, Tbxs Jowbkal, SI, 2226 (1929). D-fructose; C, D-mannose and D-fructose.
502 John C. Sowden and Robert Schaffer Vol. 74

Table III
Molar d-Glxjcose and 0.035 N NaOH at 35°
Sum:
n-Glucose,
D-fructose, Acid
Time, D-Glucose o-Fructose, D-Mannose, D-mannose, formed,
hr. % [al»D % (a]s0D % % [a]»D pH equiv.
4 95.0 0.2
11 88.8 52.7o1 8.5 -92.1° .3 97.6 40.2° 10.46 0.0010
20 82.3 13.0 .4 95.7 31.8 10.46 .0018
13.7
50 68.8 20.1 1.4 90.3 17.1 10.34 .0070
102 57.2“ 28.4 —

91.8 2.6 8.25 10.12 .0170

57.0 88.2
188 51.7 52.0 29.0 3.8 84.5 4.61 9.66 .0281
477 46.3 28.9 4.2 Opaque 8.40 .0350
46.6 52.1 28.1 78.9
646 45.4 27.7 4.3 77.2 7.50 .0360
1484 45.3 27.5 3.9 76.5 6.00 .0370
Sample calculations
Total sugar in sample (based on amount of reac- Total sugar in sample (based on amount of reac-
tion mixture taken), g. 0.187 tion mixture taken), g. 1.80
Specific radioactivity of D-glucose, cts./min./mg. 297.5 Specific radioactivity of D-fructose, cts./min./mg. 137.1
5 X 9180 X 1000 5 X 6410 X 1000
„
D-Glucose, g. =--—--
0.154 D-Fructose, g. = --
0.234

o-Glucose, % 82.3 o-Fructose, % 13.0

•
Where duplicate values are given, the second value was observed after recrystallization and re-assay of the sugar. 1 Spe-
cific optical rotation of the radio-assayed sample.
Table IV
Molar d-Fructose and 0.035 N NaOH at 35°
o-Fructose,
D-glucose, Acid
Time, D-Fructose, -Glucose, D-Mannose, D-mannose, formed,
hr. % [a]2»l> % [a]20D % [a]«D PH
4.5 0.7 -79.2° 10.47 0.0022
20 80.8“ 11.9 1.8 -59.8 10.36 .0086
80.8 -92.2o1 11.3 51.4° 93.9
50 3.4 -41.1 10.10 .0188
72 57.0 ..... 23.4 4.1 84.5 -33.3 9.93 .0237
23.4
100 52.0 ..... 4.6 -29.0 9.60 .0278
140 50.2 ..... 30.0 4.9 -25.4 9.35 .0307
49.9 28.5 51.2 83.3
260 44.7 ..... 29.5 5.3 Opaque 8.70 .0345
29.4 79.4
477 43.0 ..... 28.4 5.3 7.50 .0370
42.8 28.0 50.9 76.1
1437 43.3 ..... 29.8 5.1 5.85 .0374
28.9 48.8 77.2
“
Where duplicate values are given, the second value was after recrystallization and re-assay of the sugar. * Spe-
tifie optical rotation of the radio-assayed sample.
Determination of o-Mannose as the Phenylhydrazone.— samples of the D-glucose and D-fructose isomerization mix-
The determination of D-mannose as the phenylhydrazone23 is tures, 100 mg. of D-mannose. No additional sugar was
not applicable to low concentrations of the sugar because added to aliquots of the D-mannose isomerization mixture.
of a slight solubility of the derivative. A study of the A 5-ml. portion of a solution, prepared by combination
yields*of phenylhydrazone obtained, under specified condi- of 1.2 g. of freshly distilled phenylhydrazine, 6 drops of
tions, from known mixtures of D-mannose with D-glucose glacial acetic acid and 4.2 g. of water, then was added.
and D-fructose (Figs. 1 and 2) showed that amounts above After standing overnight at 0°, the hydrazone was filtered
106 mg. of D-mannose were readily determined. Accord- onto a tared, porous-bottomed crucible and washed with
ingly, smaller amounts were made measurable by the addi- 10 ml. of ice-water, 10 ml. of absolute ethanol at 0° and
tion of 100 mg. of D-mannose to the samples before hydra- 10 ml. of ether. Drying was performed at 110° for two
zone precipitation and the appropriate correction was hours. The observed weight of D-mannose phenylhydra-
made to the value for D-mannose thus obtained. Curve A, zone then was converted to weight of D-mannose by refer-
Fig. 1, was used for determinations of D-mannose in the d- ence to Figs, 1 and 2.
glucose isomerization mixture and curve B, Fig. 1, in the Determination of Total Acidic Products.—Periodically,
D-fructose isomerization mixture. Curve B, Fig. 2, was 10-ml. aliquots of the reaction mixtures were titrated to the
employed for the D-mannose isomerization mixture. phenolphthalein end-point with 0.1 N hydrochloric acid.
To each 10-ml. aliquot of reaction mixture was added, In the later stages of the reaction, when the reaction mix-
at 0°, 2 drops of glacial acetic acid, 10 ml. of water and, to tures were acidic to the indicator, titration was performed
with 0.1 N sodium hydroxide. The pH of the reactions
(18) B. Bourquelot and H. Heriwty, Com ft. US, 889 (1899). also was followed by use of a Beckman pH meter.
Jan. 20, 1952 Reaction of d-Glucose, d-Mannose and d-Fructose in Sodium Hydroxide 503

Table V
Molar d-Mannose and 0.035 N NaOH at 35°
Sum:
D-Mannose,
D-fructose, Acid
Time, D-Mannose, o-Fructose, D-Glucose, D-glucose, formed,
hr. % % [ap"D % [a]*°D % [o¡JtóD equiv.
5 95.8 12.1° 10.36 0.0007
19 92.1 5.9° 8.22 10.35 .0016
5.7 -91.8o11
52 82.2 12.1 ..... 2.67 10.32 .0035
73 73.7 14.2 ..... 6.4 1.03 10.23 .0057
14.8 5.9 48.0° 94.4
142 64.7 18.5 ..... 10.1 -1.36 10.15 .0169
9.9 49.1 93.1
260 55.3 20.1 ..... 12.7 -2.20 9.76 .0263
11.6 50.5 87.0
382 52.3 20.5 -91.1 -2.25 9.32 .0323
549 51.0 19.3 ..... 11.1 Opaque 8.45 .0350
19.7 10.5 50.0 81.2
765 51.0 6.90
“
Where duplicate values are given, the second value was observed after recrystallization and re-assay of the sugar. 1
Spe-
cific optical rotation of the radio-assayed sample.

Tabulation of Results.—The specific radioactivities of the finally with addition of absolute ethanol, the residual sirups
isolated D-glucose and D-fructose samples were determined were dried at room temperature in high vacuum over phos-
as described above for the standardization of the stock radio- phorus pentoxide. The resulting material, on solution in
active sugar solutions. Representative data for the in- water and addition of fresh yeast, did not ferment further.
dividual isomerization reactions are shown in Tables III, Attempted pheriylosazone formation gave no precipitate.
IV and V. Aldose sugars were removed from aliquots of the non-
Isolation and Partial Characterization of Acidic Products. fermen table sirups by bromine oxidation followed by ion-
—A 100-ml. sample taken 43 days after initiation of the exchange as described above. The resulting unoxidized,
D-glucose reaction, containing 3.6 meq. of acid by titration, non-fermen table material gave a positive Fehling test with-
was passed successively through 10 ml. of Amberlite I.R.- out heating. When treated with acetone, anhydrous cupric
100-H and 15 ml. of Duolite A-4. The resulting effluent sulfate and sulfuric acid according to the directions of
was colorless. After washing free of reducing substances, Steiger and Reichstein,8 this material failed to yield 1,2;3,4-
the Duolite A-4 column was connected to deliver into a diisopropylidene-D-ribohexulose although indications of its
column containing 50 ml. of Amberlite I.R.-100-H, and 15 production in small amount were obtained by fractionation
ml. of 1 Nsodium hydroxide was washed through. Effluent and polarimetric observations.
from the Duolite A-4 column showed the original dark color Table VI lists observations of fermentation and oxidation
of the reaction mixture while the effluent from the Amberlite reaction yields.
I.R.-100 was yellow.
Volatile acid, amounting to 1.46 meq. by titration, was Table VI
distilled by concentration of the effluent at reduced pressure -acidic, Non-fermbntable Products from d-Glu-
with repeated additions of water. Following titration with COSB AND D-FRUCTOSE
sodium hydroxide, the distillate was concentrated to dry-
Vol. of Product
ness. The equivalent weight of the residual salt, deter- Reac- soln. unoxi-
mined by its conversion to sodium sulfate, coincided with tion fer- Prod- dized by
that of sodium acetate. The presence also of traces of time, mented, [of ]d in bromine, [<x]d in
Sugar ml. g.’ water g. water
formic acid in the volatile acid fraction was indicated by the
incipient ability of the latter to decolorize dilute acidified D-Glucose 43 100 3.90 7.7° 2.52 8.1°
potassium permanganate solution. (In preliminary experi- D-Fructose 60 36 1.48 8.5 1.05 7.0
ments, when air was not rigidly excluded from the alkaline
isomerization reaction mixtures, the equivalent weight of The summation of the above respective amounts of non-
the volatile acid produced was intermediate between that acidic, non-fermen table materials and the corresponding
of formic and acetic acids.) analytically observed amounts of D-glucose, D-fructose and
The volume of the non-volatile acid solution was reduced D-mannose accounts for 100 ± 1% of the initial amounts
to 0.5 ml. and lactic acid then was separated by continuous of the sugars employed in the isomerization reactions.
ether extraction followed by steam distillation of the ether Cationic Dependence of the Isomerization in 0.5 N
extract.24 Titration indicated 0.40 meq. of lactic acid. Alkali.—Kusin18 reported a difference in the initial course
Titration, with sodium hydroxide, of the acids not ex- of the isomerizations of D-glucose and D-fructose, respec-
tracted by ether produced considerable color in the solution. tively, in 0.5 N sodium and calcium hydroxides. A similar
Accordingly, the titration was followed with a pH-meter. effect is demonstrated readily with D-mannose by observa-
About one-half of these acids, 0.65 meq., was rapidly tion of its mutarotation and changes in its concentration
neutralized to pH 7. The remaining acids, 0.7 meq., in the two alkaline systems. The early mutarotations in
required stepwise additions of alkali over a long period of sodium and calcium hydroxides, recorded in Table VII,
time before a constant pH in the alkaline range was at- are in different directions. The effect of barium hydroxide
tained. appears to be intermediate between those of sodium and
Isolation and Partial Characterization of Non-acid, Non- calcium hydroxides. Experiments conducted at lower
fermentable Products.—Samples of the reaction mixtures alkali concentrations («.£., D-glucose with 0.035 N sodium
were de-ionized by ion-exchange and then slurried with fresh and calcium hydroxides at 35°) showed no detectable dif-
baker’s yeast equal in weight to the initial amounts of sugar ference in the course of the isomerization.
in each sample. Evolution of carbon dioxide was complete
after seven hours and the yeast was filtered off after 20 Discussion
hours. The filtrates were titrated to the phenolphthalein
end-point with dilute alkali and then de-ionized by ion- Accuracy of Analytical Results—Standardiza-
exchange. Following evaporation at reduced pressure, tion of the stock radioactive sugar solutions by
(34) W. L. Bran· end . B. Hew, Tan Jooemal, 43, 2703 (1826). isotopic dilution with pure sugars gave results
504 John C. Sowden and Robert Schaffer Vol. 74

Table VII which lead to acidic products, results in increased

Reaction of 0.555 Molar d-Mannose in 0.5 N Alkalies duration and extent of the isomerization reaction.
at 25° The comparative rate of reaction as well as the
Sodium hydroxide Calcium hydroxide Barium hydroxide preferred order of formation of the three sugars is:
D-fructose > D-glucose > D-mannose. Thus, the
Time,
hr.
‘ Time,
hr.
"; Time,
hr.
nose, reactivity of D-fructose in alkaline solution is much
greater than observations of its rate of disappear-
[a]25D % [a]25n [a]Mn
1.0 14.5° 1.2 —
6.2° 1.0 5.8° ance from its reaction mixture would indicate.
2.0 13.7 2.0 -4.3 2. U 4.75
2.7 12.5 3.3 -
1 85 3.0 3.9 The initial rate of formation of acidic products is
4.0 89.1 4.5 2.4 e 2 0 4.0 3.05 greatest in the D-fructose reaction and decreases
5.3
9.0
11.6 6.5 6 45 4.5 2.55 regularly in this system while, in the D-glucose and
9.6 9.0 7.95 5.5 1.8
22.5 -1.85
D-mannose systems, the rates at first increase and
11.0 6.8 11.0 8.45 26.5 -2,3 then decrease. D-Fructose, therefore, appears to
24.0 1.95 71.9 24.0 4.95 20 9 27.0 60.0 act as precursor for initial acid formation. From
Initial pH: 12.45 Initial pH: 11.85 Initial pH: 12.35 analysis of the acids formed, the major source of
Acid formed, 24 hr.: Acid formed, 24 hr.: Acid formed, 24 hr.: this acid, which eventually stops the isomerization
0.025 meq./ml. 0.125 meq./ml. 0.070 meq./ml. reaction by neutralization of alkali, is the isomeriza-
tion of fragmentation products. In this connec-
with average deviations of less than 1%. Isotopic
dilution analyses of samples of the reaction mix- tion, it may be noted that acetic and lactic acids
are, respectively, the saccharinic acids related to
tures, however, subject to greater error due to
were
glycolaldehyde and glyceraldehyde.
impurity—an that leads to reduced counting
error
The amounts of non-acidic, non-fermentable
rates and, therefore, increased calculated concen-
trations. Recrystallizations of sugar samples ob- products isolated correspond to the amounts of
tained from the reaction mixtures and redetermina- sugar unaccounted for by direct analyses. Again,
the D-fructose system provides the most rapid growth
tions of their radioactivities indicated unchanged as well as the greatest amounts of these products.
specific radioactivities for all D-fructose samples The nature of this material remains unknown.
and for D-glucose samples from the D-glucose and Conclusive evidence for the production of sugars
alkali reaction. The optical rotations of these
epimeric with the reagent carbohydrates at carbon-
products were in good agreement with accepted val- 3 was not obtained, although the enediol mecha-
ues. D-Glucose samples from the other two reac- nism predicts the formation of such substances.
tion mixtures could not be brought to constant
Indeed, the isomerization of D-galactose by alkali is
specific radioactivities even after three crystalliza- known to produce not only D-tagatose but also the
tions and the observed specific optical rotations
were low by as much as 8% in these instances.
epimeric D-sorbose,2 while D-manno-D-gaZu-heptose
produces D-glucoheptulose as well as D-mannohep-
Accordingly, these latter values for D-glucose are tulose.26 A significant portion of the non-ferment-
the least accurate of the analyses reported. able substances was oxidized by bromine under
Determinations of D-mannose as the phenylhy- buffered acid conditions and the oxidation products
drazone are affected by the nature and amounts of are a promising source of further information con-
other materials present, as shown in Figs. 1 and 2.
The substances other than D-glucose and D-fructose cerning the nature of the complex mixture. Reac-
tion of the unoxidized fraction with cold Fehling
formed in the reactions may add to the variance in solution suggests the presence of ketotetroses among
yields of D-mannose phenylhydrazone observed with the products.
the known sugar mixtures. The melting points of
the D-mannose phenylhydrazone analytical samples Acknowledgment.—The authors are pleased to
from the reaction mixtures were consistently in the acknowledge the generous support of Anheuser-
Busch, Inc., Saint Louis, throughout the course of
range 193-196°, which is quite satisfactory for this this research.
derivative on initial precipitation.
The Course of the Reaction.—The rigorous St. Louis, Missouri Received July 13, 1951
exclusion of conditions favoring side reactions, e.g., (25) Edna M. Montgomery and C. S. Hudson, This Journal, 61,
air oxidation and photochemical decomposition 1654 (1939).