Effect of Toluene and Similar Compounds
the Carbohydrate-Anthrone
on
Reaction
Arthur W, Devor, William L. Baker, and Kenneth A. Devor
Toluene interferes with the anthrone reaction for hexoses causing a decrease in the
absorption maximum at 625 m and producing a second absorption band with a
maximum at 475 mjz. Xylene, benzene, and cyclohexene also interfere, whereas
cyclohexane, o-chlorobenzene, and benzoic acid do not.
I
T WAS NOTED that, when
dialyzable
constituents
of
appeared
with a maximum
the test was applied to the
with sugar standards.
This
there was some vai’iable in
the anthrone
test was ap)lied
to tile 11011human
urine, an absorption
band always
at about 475 mn; this band was absent when
dialyzable
fraction
alid was usually al)sent
experience
with the sugars led us to 1)elieve
OUF procedure.
Methods
To Lumetron
tubes containing
2 ml. of carbohydrate
solution
5#{188}
ml. of anthrone
reagent
(0.2 gm. of anthrone
in 10 ml. of 1120 amId 100
ml. of H2SO4) was rapidly added, while mixing.
Tile tubes were heated
in a boiling-water
bath for 31/2 miii., after whicil they were wiped dry
and allowed to cool. The percentage
transmittance
was measured
in
1-cm. cells with a T)K-2 Beckman
recording
spectm.ophotometer*
using
a blank containing
HO in place of tile carbohydrate
solution
as a reference.
Some of the carbohydrate
samples
contained
reagent-grade
toluene or another
possible interfering
substance.
From the Department
of Physiological
Clmemnistry, Ohio State ITmiiversity,
Columbus,
Ohio.
This imlvestigation
was supported
by Grant A-3037 from the National
Institute
of Artlmri.
tis amid Metabolic
Diseases,
IT. S. Public
Ilealtlm
Seivice,
Bethmesda,
Md.
Received
for publication
Feb. 3, 1963.
*Beekmami Instrimmnemmts Co., Spimuco I)ivisioim, Palo Alto, Calif.
597
DEVOR fT AL.
598
Clinical
Chemistry
Results and Discussion
Figure
1 silows the effect of toluene on tile anthrone
reaction
with
galactose.
The usual ai)sorption
halid at 625 mm appears,
but it is (Iccrease(i
(by nearly half when saturated
with toluene)
when tOluelle is
present.
lii addition,
an absorption
maximum
appears
(greater
than
Fig.
1. Effeet
reactuon
H20.
with
Curve
of toluene
laO
sg.
on anthrone
galactose
1, no toluene;
in
Curve
tion saturated
with toluene;
Curve
drops
of toluene
added
to 2 ml.
I
tose
l)efore
test
was
2m1.
2, solu.
3, two
game.
applied.
Wavelengthin m,u
at 625 m when saturated
with toluene)
at 475 mp.. The effect increases
w’ith an increase
in toluene concentration
and is similar for all hexoses
tried (glucose,
galactose,
mannose,
and fructose).
In fact, tile al)sol’bance at 475 TUL is proportional
to the amount of toluene present.
The
results
were so consistent
that tile solubility
of toluene in a standard
galactose
solution was estimated
to be 38 mg./100 ml. of solution.
The amltilrone test for 6-deoxyhexoses
(rhamnose
and fucose)
is
only slightly affected by toluene and the second al)sorption
band does
not appear.
The toluene is very probably
incorporated
into the green
chromogen
resulting
(1) from the reaction
between anthrone
and hy(1roxymethyl
furfural,
producing
a new chromogen
with an absorption
maximum
at 475 m; this type of complex probably
does not occur with
methyl furfural.
The toluene effect is eliminated
by passing
nitrogen
gas through
the carboilydl’ate
solution for 10 to 15 mm.
An throne was recrystallized
according
to the method of Morris (2);
Vol. 10, Wo. 7, 1964
CARBOHYDRATE-ANTHRONE
REACTION
599
the crystalline
and tile filtrate
fractions
(after
evaporation
of tile
solveiit)
gave tile same results,
which l)roves that impurities
in the
anthrone
are not tile cause of the cllalige iii the absorption
spectra.
A 90-mI. sample of the toluene was fractionally
distilled
into 9 separate 10-mi. fractions.
Each of the fractions,
imicluding the residual
fraction,
gave the same effect, which strongly
indicates
that the change
in the absorption
is not (Iue to an impurity
in the toluene.
Xylene,
producillg
a second al)sorl)tion
han(1 at 485 m1., interferes
with the anthrone
reaction
to a lesser (legree than toluene.
Benzene
produces
a second maximum
at 455 m and interferes
even less than
xylene.
Cyclohexene
destroys
the color, while cyclohexane,
o-chlorobellzelle,
2,2,4-trimethyl
pentaime, and benzoic acid (0.2% w/v) have
no effect.
It has been known for some time that tryptophane
(.?) interferes
with the anthrone
reaction,
and this reaction
is not highly recommended
for determining
the starch content
of plants
(4) because
all
interfering
substances
cannot be removed.
When tile anthrone
reaction is used for the estimation
of carbohydrates
in the presence
of
other substances,
an absorption
curve should be run to check for possiide interferences.
References
1.
2.
3.
4.
Momose,
T., Ueda,
Y., and
Morris,
D. L., Science
107,
Tulier,
E. F., and Keiding,
Hoffpauir,
C. L., J. Assoc.
Sugi, A., Pjmarnm. Bull. 3, 323 (1955).
254 (1948).
N. R., Anal.
Gheni. 26, 875 (1954).
Offic.
Agr.
Chemists
37, 716 (1954).