ANALYTICAL BIOCHEMISTRY 68, 87-94 (1975)

Percolation of Starch and Soluble Carbohydrates from

Plant Tissue for Quantitative Determination
with Anthrone

J ~ R G E N HANSEN AND IB M O L L E R

Department of, Plant Physiology and Anatomy, Royal Veterinary and

Agricultural University, Thorvaldsensvej 40,
DK-1871 Copenhagen V, Denmark
Received January 20, 1975; accepted April 4, 1975

The technique of percolation was applied as an improvement over the usual

technique of maceration for extraction of carbohydrates from small amounts of
plant material. Soluble carbohydrates were extracted by percolation with 80%
ethanol and starch was extracted by percolation with 35% perchloric acid. The
course of percolation was studied and the technique was demonstrated to give
reproducible results. The anthrone method was applied for starch determination.
The influenceby cellulose on the determinationof starch, as well as, the influence
of perchloric acid on the anthronereaction was investigated.An analyticalproce-
dure based on the obtained results is given.

The reaction of carbohydrates with anthrone in sulfuric acid, which

was first described by Dreywood (1) as a qualitative test for carbohy-
drates, has been improved for quantitative determination of different
carbohydrates (2-4). McCready et aL (5) improved further the method
for the determination of starch in pea plants and found that purified
starches reacted in accordance with their glucose equivalents. A system-
atic investigation of the anthrone reaction with the different hexoses and
pentoses which may be encountered in biological material was per-
formed by Yemm and Willis (6). Quantitative determination is only pos-
sible when the identity of the sugar component to be assayed is known,
because the colour development varies with the different sugars (6).
Thus, the method is not valid for determination of carbohydrates in mix-
tures (7). In some special cases noncarbohydrate compounds have been
reported to interfere with the carbohydrate-anthrone reaction (1,8-1 1).
Nevertheless, the anthrone method is widely used, especially for the
determination of starch in plant material.
Starch is usually extracted with perchloric acid. Nielsen (12) and
Nielsen and Gleason (13) first described perchloric acid as a solvent for
starch. At room temperature starch is easily dissolved in perchloric acid
at a concentration between 4.0 and 4.8 molal. Concentrations higher
than 4.8 molal resulted in hydrolysis of starch (I2).
87
Copyright© 1975by Academic Press, Inc.
All rightsof reproductionin any formreserved.
88 HANSEN AND MOLLER

Generally starch is extracted by maceration of the plant material. The

time for extraction of starch with perchloric acid is about 10-15 min
followed by centrifugation and decantation. With some modifications
this procedure was adopted by several investigators (5,14-17). The ex-
traction of starch by maceration, however, has several disadvantages.
Interfering carbohydrates have to be removed quantitatively. For accu-
racy the method of extraction by maceration, thus includes several repeat-
ing steps, the number of which will depend on different factors, e.g., the
amount and type of plant material and the carbohydrate content. For
these reasons the method is rather time consuming and not suitable for
routine measurements. Furthermore, a relatively large amount of plant
material is required per analysis.
The present paper describes the adaptation of the technique of perco-
lation for the extraction of carbohydrates from pea plants. With this
technique it is possible to measure the starch content of small amounts
of plant material (50-100 mg dry wt), which is often all that is available
from plants grown under controlled conditions. The course of percola-
tion of ethanol soluble carbohydrates and starch was investigated, as
well as, the effect of perchloric acid on another glucose polymer, cellu-
lose. Furthermore, we examined the interference by perchloric acid on
the determination of starch with anthrone.

MATERIALS AND METHODS

Reagents. Analytical reagent grade HC104, sp gr 1.53 was used to
prepare all perchloric acid solutions. The anthrone reagent consisted of
1.0 g of anthrone (Merck, for synthesis) dissolved in 500 ml 72% sul-
furic acid. This reagent has to be prepared daily. A stock starch solution
contained 1 mg maize starch/ml distilled water preserved with 0.1%
benzoic acid. The hydrolysis of the starch took place with 1 ml of 35%
perchloric acid before dilution. Standard starch solutions containing
0-10 mg starch] 100 ml were prepared daily. The perchloric acid concen-
tration of the standard starch solutions was adjusted to the concentration
in the test solutions. Cellulose (microcrystalline) for column chroma-
tography was from Merck.
Equipment. Spectrophotometric readings were made at 630 nm with a
Zeiss spectrophotometer PMQ II using the tungsten lamp. All readings
were made using matched l-cm pathlength glass cuvettes.
Glass percolators, having an inner diameter of 0.5 cm (Fig. 1) were
specially made for analysing small quantities of plant material. The glass
wool micro mineral fiber was from Bilsom International AB, Sweden.
Preparation of plant material. Pea plants were dried for 24 hr at 60°C
and ground in a ball mill (Retsch Miihle) for 5 min at high speed. A very
uniformly ground plant material was obtained, which was stored at room
temperature until extraction.
 PERCOLATION OF CARBOHYDRATES 89

Fie. l. Device for percolation of carbohydrates. A: Supply of solvent from a Mariotte

flask. B: Solvent. C: Glass wool stopper. D: Air space. E: Liquid surface. F: Plant mate-
rial. G: Glass filter (diam 0.5 cm). H: Stopcock.

Percolation of carbohydrates. A weighed sample of 50-100 mg of the

plant material was quantitatively transferred to a percolator and placed
on the glass filter (Fig. I). The plant material was wetted with 1 ml of
80% ethanol, and air bubbles were removed by stirring in order to avoid
the formation of channels. To prevent diffusion of carbohydrates into the
solvent, an air space was formed by inserting a stopper of glass wool as
shown in Fig. 1.
Soluble carbohydrates were percolated with ethanol. A continuous
flow of cold 80% ethanol was provided from a Mariotte flask. The rate
of percolation was adjusted to approximately 1.5 ml/hr. After the eth-
anol percolation, the residue containing the starch was percolated with
perchloric acid. The plant material was first thoroughly mixed in the
percolator with 1 ml of 35% perchloric acid resulting in a swelling of the
plant material, and starch was then percolated from the plant material by
a continuous flow of 35% perchloric acid, provided from a Mariotte
flask at approximately 1.5 ml[hr.
Determination of starch. From the test solutions, containing the
starch or an equivalent amount of glucose, 2.0 ml were pipetted into a
Pyrex glass tube and kept at 0°C. Ten milliliters of the anthrone-reagent,
which was cooled to 0°C before use, were added to the 2.0 ml test solu-
tion. The reaction mixture was shaken thoroughly and heated for exactly
11 min at 100°C (water bath). After this treatment the tube was rapidly
90 HANSEN AND MOLLER

cooled to 0°C and the absorbance at 630 nm was measured against dis-
tilled water within 1 hr.

RES U LTS
Effect of perchloric acid on starch and cellulose. Glass tubes, con-
taining 10.0 mg of maize starch or 10.0 mg of cellulose powder, were
shaken for 20 hr with 10 ml perchloric acid of different concentrations.
The content of each glass tube was filtered through glass wool and the
filtrate was diluted with distilled water to give a total volume of 100 ml.
The content of the carbohydrates in the diluted filtrates was determined
in a 2.0 ml sample by the anthrone method.
Starch was dissolved completely at a perchloric acid concentration
higher than 20% (Fig. 2). The perchloric acid solutions had a minimal ef-
fect on the solubility or hydrolysis of cellulose only at a concentration of
35% perchloric acid.
Interference of perchloric acid on the determination of starch by the
anthrone method. The influence of perchloric acid on the carbohydrate-
anthrone reaction has previously not been investigated. We therefore
examined to which extent the concentration of perchloric acid would af-
fect the carbohydrate-anthrone reaction. Standard starch solutions were
made containing different amounts of perchloric acid. The starch con-

100

8O

LU"
£D
o 60

g 40

~ 2o

t
5 10 15 20 25 30 35
PERCHLORIC ACID,%

FIG. 2. Solubility and/or hydrolysis of starch and cellulose in relation to the concentra-
tion of perchloric acid. The ordinate represents the content of starch (©) and cellulose (Q)
in percentage of the initial amount. Fixed amounts of starch or cellulose were shaken for
20 hr with the respective concentrations of perchloric acid. The solutions were filtered and
the filtrates were tested for the content of starch or cellulose.
 PERCOLATION OF CARBOHYDRATES 91

0,95

0.90

E
c:

co

z
<
¢n
g
~
<
0.80

0.75
0 3.5 7.0 10.5 14.O 17.5
FINAL PERCHLORIC ACiD CONCENTRATION,°/°
FIG. 3. Relation between the absorbance of the colour developed by the carbohydrate-
anthrone reaction and the concentration of perchloric acid in the test solutions. The con-
centration of starch was 10 rag/100 ml test solution.

centration ranged from 0 to 10 mg starch/100 ml solution and the

amount of perchloric acid varied from 10 to 50 ml of 35% perchloric
acid/100 ml solution, equal to a final concentration ranging from 3.5 to
17.5% perchloric acid. The anthrone method was applied for the deter-
mination of the carbohydrate content.
It became evident from the standard curves that perchloric acid did
influence the carbohydrate-anthrone reaction. The slope of the starch
standard curve increased proportionally to the concentration of per-
chtoric acid. Therefore, it is important to adjust the concentration of
perchloric acid in the standard starch solutions according to that of the
test solutions. The results for one concentration of starch (10 rag/100
ml) at the different concentrations of perchloric acid are shown (Fig. 3).
Percolation of carbohydrates from plant material. The course of
percolation was studied to examine the effectivity of extraction of carbo-
hydrates by percolation. An amount of 0. t g of the ground plant mate-
rial, containing approximately 9% of the dry weight as starch, was trans-
ferred to a percolator and treated as described.
With a fraction collector samples of 2.5 ml of the ethanol percolate
were collected and tested for soluble carbohydrates by the anthrone
method. To make these determinations more sensitive to anthrone-de-
tectable carbohydrates, the fractions were tested without previous dilu-
92 H A N S E N A N D MOLLER

2.0

1.8

1.6

E
c 1.4
~D

~ 1.2
<
Ill

~. 1.0
<
m
~ 0.8
nn

< 0.6

0.4

0.2

0
i i i L i i i L

1 3 5 7 9 11 13 15 17
FRACTION NUMBER
FiG. 4. The course of percolation of soluble carbohydrates (C)) and starch (O). Fractions
of 2.5 ml were collected from the ethanol percolate. Fractions of 2.0 ml were collected
from the perchloric acid percolate. The carbohydrate content in 2.0 ml of each fraction
was determined by the anthrone method without previous dilution.

tion. A volume of 25 ml of 80% ethanol is sufficient to remove the solu-

ble carbohydrates (Fig. 4).
Starch was percolated with 35% perchloric acid as described and frac-
tions of 2.0 ml were collected and tested by the anthrone method
without dilution. Figure 4 shows that 30 ml of 35% perchloric acid is a
suitable volume to percolate starch.

ANALYTICAL METHOD
Procedure. Based on the results of these investigations, the following
procedure was formulated and demonstrated to give reproducible results
in the determination of starch in pea plants: Prepare solutions of 80%
ethanol, 35% perchloric acid, and 72% sulfuric acid. Dissolve 1 g of
anthrone in 500 ml of the sulfuric acid solution. Prepare this reagent
daily. Standard starch solutions containing from 0 to 10 mg starch/100
ml solution are prepared daily. The concentration of perchloric acid in
the standard solutions must be identical to the concentration in the test
solutions.
 PERCOLATION OF CARBOHYDRATES 93

Grind a composite sample of dried pea plants and transfer 50-100 mg

of the plant material to a percolator. Wet the plant material with 1 ml of
80% ethanol, remove air bubbles, and percolate with 25 ml of 80%
ethanol at a rate of 1.5 ml/hr. This percolate may be used for determina-
tion of one or more of the soluble carbohydrates by some convenient
method.
Mix the residue thoroughly with 1 ml of 35% perchloric acid and
percolate with 30 ml of the perchloric acid solution at a rate of 1.5 ml/hr.
Collect the percolate and dilute to a volume of 100 ml. Pipet 2.0 ml of
the test solution into a Pyrex glass tube and cool to 0°C. Add 10 ml of
ice cold anthrone reagent and mix thoroughly. Heat the mixture for ex-
actly 11 min at 100°C and cool rapidly to 0°C. Measure the absorbance
at 630 nm against distilled water. Carry also standard starch solutions
including a blank through the anthrone reaction procedure.
Reproducibility of starch determination. Two different batches of
plant material originating from pea plants grown at different irradiances
and thus containing different amounts of carbohydrates, were used to
study the reproducibility of extraction of starch by percolation. The
plant material grown under the low irradiance contained approximately
2% starch on dry weight basis, whereas the other batch contained
approximately 8% starch. Eight, respectively, nine samples of the plant
material were carried through the procedure as described and starch was
determined. Statistical evaluation was based on calculation of the coeffi-
cient of variation, V. For the plant material containing about 2% starch
V was 4% whereas V was 2% when the starch content was about 8%.

DISCUSSION
The extraction of ethanol soluble carbohydrates from plant material
by percolation with 80% ethanol is very effective (Fig. 4). Some varia-
tion probably exists in the course of percolation for different batches of
plant material containing different amounts of starch. However, by per-
colation with 25 ml of 80% ethanol the soluble carbohydrates should be
quantitatively removed from the plant material. The ethanol percolate is
suitable for the determination of one or more of the soluble carbohy-
drates present.
The percolation of starch from the plant material is more slow. As it
can be seen (Fig. 4) fraction number 16 still contains some anthrone-de-
tectable material. However, these measurements were made without
dilution. The absorbance of the last fractions would not have been signif-
icantly different from the blank if they were diluted as described in the
standard procedure. The actual content of anthrone-detectable material
in fraction number 16 was calculated to be 32 t~g of starch equal to 0.3%
of the total amount percolated in the first 15 fractions (30 ml).
94 HANSEN AND M ~ L L E R

A coefficient of variation ranging from 2 to 4% demonstrates that the

technique of percolation gives reproducible results and may be success-
fully employed for extraction of starch from plant material.
The present method offers several advantages over the maceration
technique. The percolation technique is easy to handle and the risk of
loss of plant material during the percolation procedure is minimal. Thus,
one important source of error which is connected with the extraction by
maceration is eliminated. A large number of determinations can be made
simultaneously as the number of percolators available is the only limiting
factor. The application of the technique of percolation for determination
of starch thus provides a suitable method for routine analysis.
REFERENCES
1. Dreywood, R. (1946) Ind. Eng. Chem. Anal. Ed. 18, 499.
2. Morris, D. L. (1948) Science 107, 254-255.
3. Morse, E. E. (1947) Ind. Eng. Chem. Anal. Ed. 19, 1012-1013.
4. Viles, F. J. and Silverman, L. (1949) Anal. Chem. 21, 950-953.
5. McCready, R. M., Guggolz, J., Silviera, V., and Owens, H. S. (1950) Anal. Chem. 22,
1156-1158.
6. Yemm, E. W. and Willis, A. J. (1954) Biochem. J. 57, 508-514.
7. Johanson, R. (1953) Nature (London) 171, 176-177.
8. Devor, A. W., Baker, W. L., and Devor, K. A. (1964) Clin. Chem. 10, 597-599.
9. Scott, T. A. and Melvin, E. H. (1953)Anal. Chem. 25, 1656-1661.
10. Tuller, E. F. and Keiding, N. R. (1954) Anal. Chem. 26, 875-878.
11. Tong, H. K., Lee, K. H., and Wong, H. A. (1973) Anal. Biochem. 51, 390-398.
12. Nielsen, J. P. (1943) Ind. Eng. Chem. Anal. Ed. 15, 176-179.
13. Nielsen, J. P. and Gleason, P. C. (1945) Ind. Eng. Chem. Anal. Ed. 17, 131-134.
14. Pucher, G. W., Leavenworth, C. S., and Vickery, H. B. (1948) Anal. Chem. 20,
850-853.
15. Clegg, K. M. (1956)J. Sci. FoodAgric. 7, 40-44.
16. Madsen, E. (1968) Physiol. Plant. 21, 168-175.
17. Gaines, T. P. (1973)J.A.O.A.C. 56, 1419-1424.