CLIN. CHEM.

26/5, 658-660 (1980)

Method for Determination of Hydrogen Peroxide, with Its Application

Illustrated by Glucose Assay
Ernst Graf and John T. Penniston1

We describe a simple colorimetric method for determining in 0.5 molfL H2S04 [247.2mg of (NH4)6Mo7024.4H2O and 5.6
micromolar quantities of hydrogen peroxide, based on the mL of concentrated H2S04, diluted to 200 mL with waterj;
oxidation of iodide in the presence of ammonium molyb- starch solution (prepared daily) consisting of 5.0 g of soluble
date and photometry of the resulting blue starch-iodine starch (Lintner) made into a slurry with 10.0 mL of water and
complex. Color development is linearlydependent on then added, with stirring, to 90 mL of boiling water; hydrogen
analyte concentration, but only slightly time dependent, peroxide stock solution consisting of 1X mL of 300 g/L hy-
drogen peroxide diluted to 1 L with water (the exact concen-
and the color of the complex formed is stable for several
tration, determined iodometrically, was 8.347 mmolfL; the
hours. In the range of wavelengths that may be used (570
thiosulfate solution had been standardized against potassium
to 630 nm), lack of interferencefrom other biological
iodate); 2.21 mmolfL glucose stock solution in 0.1 mol/L
compounds makes thismethod seem suitable for routine
aqueous sodium acetate, pH 5.0; 10 g/L glucose oxidase stock
analyses.As one illustrative application of the method we
solution in 0.1 mol/L aqueous sodium acetate, pH 5.0. Glucose
quantitated glucose by measuring hydrogen peroxide oxidase (EC 1.1.3.4) type II, from AspergiUus niger, was
produced from it by glucose oxidase catalysis. This method purchased from Sigma Chemical Co., St. Louis, MO 63178; it
of quantitating glucose is more than five times as sensitive had a specific activity of 26 kU/g.
as the commonly used dianisidinemethod. With the ap-
propriate hydrogen peroxide-producing oxidases, this
method may be used to directly measure amino acids, Procedure
purines, uric acid, xanthine, and hypoxanthine. Standard hydrogen peroxide solutions were prepared by
diluting the above stock solution with various amounts of
AdditIonal Keyphrases: starch-iodine complex . spectro- water. These solutions were used to produce hydrogen per-
photometry- other potential applications (diseases and ana- oxide standard curves by the following two methods:
lytes) Method A: To 10 zL of standard hydrogen peroxide solu-
tion add, in order, 2.0 mL of HC1, 0.2 mL of KI, 0.2 mL of
Quantitative determination of hydrogen peroxide by use ammonium molybdate in H2S04, and 0.2 mL of starch solu-
of iodide and starch was first reported by Savage (1). We de- tion, in the concentrations specified above. Twenty minutes
scribe two modifications of this method and a more complete after adding the KI, measure the absorbance vs water in
investigation of the optimum wavelength, time dependence, 1.0-cm cuvets at 570 nm.
and the significance of the order in which reagents are Method B: The standard
curve was prepared as in method
added. A, but with an additional waiting period of 20 mm between
We also describe the application of the starch-iodine the addition of ammonium molybdate and the addition of
method to the analysis for glucose, using glucose oxidase to starch.
exemplify how various substrates that can be acted on by The absorption maximum of the iodine-starch complex was
specific oxidases to produce hydrogen peroxide may be as- determined by scanning a solution as prepared by method A
sayed. Most commonly, the hydrogen peroxide so generated from 850 to 400 nm with a spectrophotometer.
is analyzed by the o-dianisidine method (2-7), in which per- The color development and compliance with Beer’s law of
oxidase catalyzes the oxidation of o-dianisidine (or of some four different concentrations of hydrogen peroxide were de-
other suitable chromogenic oxygen acceptor) to form a colored termined over a 4-h period by measuring the absorbance (vs
product with an absorption maximum at 420 nm. Our method water) at 570 nm of standards prepared by method A.
presents several advantages over the dianisidine method, as Standard glucose solutions and standard glucose oxidase
we will discuss. solutions were prepared by diluting the stock solutions with
sodium acetate solution (0.1 mol/L, pH 5.0). The optimum
Materials and Methods glucose oxidase concentration was determined by adding 40
iL of glucose (2.2 mmol/L) to 20 L of glucose oxidase stan-
Materials dards and incubating the mixtures of 37 #{176}C
for 60 mm. The
All chemicals were reagent grade. hydrogen peroxide generated was then analyzed as described
We used the following solutions: 1.0 mol/L KI (prepared under method A. A glucose standard curve was established
daily); 50 mmolfL HC1; 1.0 mmol/L ammonium molybdate from data obtained by adding 40 tL of the above glucose
standards to 20 zL of glucose oxidase (250 mg/L), incubating
the mixtures at 37 #{176}C
for 60 mm, and measuring the hydrogen
Section of Biochemistry, Department of Cell Biology, Mayo
ClinicfFoundation, Rochester, MN 55901. peroxide produced. In all glucose determinations the ab-
‘To whom correspondence should be addressed. sorbance was measured vs water at 570 nm, in 1.0-cm cu-
Received Dec. 7, 1979; accepted Jan. 25, 1980. vets.

658 CLINICALCHEMISTRY,Vol. 26, No. 5, 1980

 E
C
0
I’,

0 10
a,
U
C 0.5
0
0

1.0 20 0
0 1.2 2.4
jig H202/2.6l ml pg H202/2.61 ml
Fig. 1. Standard curves for H202 prepared b method A (0) and
Fig. 3. Compliance with Beer’s law after incubation for 1.5 mm
by method B (#{149}) (X), 3 mm (A), 60 mm (S), or 240 mm (0)
Method A was used
Results
A standard curve for hydrogen peroxide determined by which removes both product and enzyme (8). The rea#{235}tion by
method A is shown in Figure 1. The non-zero intercepts in which the enzyme is destroyed is rather fast because the
Figure 1 are ascribable to the slight turbidity of the starch maximum amount of hydrogen peroxide is observed at the
solutions (the absorbance was measured vs water). The ab-
same enzyme concentration, even when K! and ammonium
sorbance was linearly proportional to hydrogen peroxide up molybdate are present during the glucose oxidase reaction.
to concentrations exceeding 0.1 mg/L. From the slope of the The optimum glucose oxidase concentration was used to set
curve at high concentrations, we calculated a molar absorp- up the standard curve for glucose assay (Figure 5). The curve
tivity of E = 39.45 mmolLcm1.L at 570 nm. Most of the color is linear down to 0.6mg of glucose per liter. From the slope we
develops within the first 8 mm; the absorbance then increases
slowly over the next 4 h (Figure 2). The rate of color devel-
opment is concentration independent, and the linear rela-
tionship between color and concentration is maintained at
various incubation times (Figure 3). Thus, the time-dependent E
C
0
increase in absorbance will not be a problem in routine anal-
yses as long as the interval between addition of K! and pho-
tometric measurement of standards and unknown samples 0
does not vary by more than ito 2 mm. Sensitivity may be in- a,
a
creased when hydrogen peroxide is determined by method B C
0
(Figure i); the absorptivity calculated from this slope was .0
58.38 mmol’.cm”L. 0
In
.0
The optimum amount of glucose oxidase for the determi- 4
nation of 15.88 xg of glucose in 2.66 mL of assay medium was
5.0 zg (Figure 4). At higher enzyme concentrations the amount
of assayable hydrogen peroxide decreases to less than 10% of jig Glucose Oxidase added
the maximum value, probably because of the oxidation of
Fig. 4. Effects of glucose oxidase concentration on colorde-
methionine residues of glucose oxidase by hydrogen peroxide,
velopment of 15.9 g of glucose in a total volume of 2.66 mL
H,O, assayed by method A

1.0 I.
E
C
0
E
C
0 0
I-
a,
a
I’)

C 0
0
.0 a,
U
0 C
In 0
.0
4 .0
0
U)
.0
60 ‘ 120 ‘ ISO ‘ ‘ 240 4

Time (mm) 0
0 5
Fig. 2. Rate of color developmentatfourdifferent H202 con- pg Glucose/2.66 ml
centrations: 0.57 tg (X), 1.14g (A), 1.70 zg (#{149}),
and 2.27 ig
(0) in a total assay medium of 2.61 mL Fig. 5. Standardcurveforglucoseinthepresenceof 5 g of
Method A was used glucose oxidase, assayed by method A

CLINICAL CHEMISTRY,Vol. 26, No. 5, 1980 859

calculated an absorptivity of 29.14 mmol.cm.L, which
corresponds to 73.9% oxidation of the glucose.
Discussion
Our method for determining hydrogen peroxide
concen-
trations is sensitive, fast, inexpensive, and easily
reproducible,
adaptable to routine analysis for hydrogen peroxide. A simple
assay for hydrogen peroxide and peroxidase may be useful in
the study of certain dermatological disorders such as chronic
granulomatous disease, myeloperoxidase deficiency, and
Ch#{233}diak-Higashi syndrome. These diseases may be caused
by an inability to generate hydrogen peroxide (9), a lack of
eosinophil peroxidase (10), and a lack of myeloperoxidase (11),
respectively.
In our application of this method, determinations of glucose
concentrations with glucose oxidase produced hydrogen
peroxide in a yield of 73.9% of theoretical. This oxidation yield
is large enough for good reproducibility, and the known in-
hibitory effects of hydrogen peroxide and of gluconolactone
on glucose oxidase (8, 12) thus do not in any way hinder the
usefulness of this assay. This method of quantitating glucose
is more than fivefold as sensitive as the commonly used o-
dianisidine method, in which peroxidase (EC 1.11.1.7) cata-
lyzes the oxidation of o-dianisidine to form a colored product
having an absorption maximum at 420 nm. The peroxidase
reaction is inhibited by bilirubin (13), uric acid, ascorbic acid,
catechols, glutathione, and other hydrogen donors. The ab-
sence of peroxidase in our method obviates these problems
of interference.
Our assay for hydrogen peroxide could be particularly useful
for measuring certain metabolites by coupling it to specific
enzymic reactions that yield hydrogen peroxide. To compare
the respective sensitivities in the quantitation of some of these
metabolites, we calculated what concentrations of each would
be required to give an absorbance of 0.5 A at 570 nm in 1.0-cm
cuvets, based on the absorptivity of 58.38 mmol1.cm’.L.
Assuming that 1,5 mL of a sample was analyzed in a total assay
medium of 2.6 mL by method B, and assuming a 75% yield of
hydrogen peroxide, the following concentrations, in milligrams
of metabolite per liter of original sample, are needed to obtain
an absorbance of 0.5: 3.56mg of glucose, 3.13mg of uric acid,
3.01 mg of xanthine, 1.35 mg of hypoxanthine, and 2.71 mg of
amino acids (mean molecular mass of amino acids = 136.75
daltons). The increased sensitivity and favorable wavelength
880 CLINICAL CHEMISTRY, Vol. 26, No. 5, 1980
of colorimetric analysis of this coupled enzyme and hydrogen
peroxide assay may make possible an improved quantitative
analysis for xanthine, hypoxanthine, and uric acid. Assays
currently used in the clinical diagnosis of pathological states
such as hyperxanthinuria are cumbersome and not very reli-
able. At this time, the development of such an assay is being
investigated at the Mayo Clinic.
This work was supported by the Mayo Foundation.
References
1. Savage, D. J., The determination of hydrogen peroxide in radiation
experiments on aqueous solutions, Analyst 76, 224-226 (1951).
2. Teller, J. D., Direct, quantitative, colorimetric determination of
serum or plasma glucose. Abstr. 130th Meeting Am. Chem. Soc., 69C
(1956).
3. Huggett, A. St. G., and Nixon, D. A., Use of glucose oxidase, per-
oxidase, and o-dianisidine in determination of blood and urinary
glucose. Lancet ii, 368-370 (1957).
4. Kaplan, N. 0., Enzymatic determination of free sUgars. In Methods
in Enzymology 3, S. P. Colowick and N. 0. Kaplan, Eds., Academic
Press, New York, NY 1957, pp 107-110.
5. Saifer, A., and Gerstenfeld, S., Laboratory Methods. The photo-
metric microdetermination of blood glucose with glucose oxidase. J.
Lab. Clin. Med. 51, 448-460 (1958).
6. Washko, M. E.,and Rice,E. W., Determinationofglucoseby an
improved enzymatic procedure. Clin. Chem. 7, 542-545 (1961).
7. Bergmeyer, H. U., Gawehn, K., and Crassl, M., Glucose Oxydase.
In Methoden der Enzymatischen Analyse, H. U. Bergmeyer, Ed.,
VerlagChemie, Weinheim, F.R. G.,1970,p 416.
8. Kleppe, K., The effect of hydrogen peroxide on glucose oxidase
from Aspergillus niger. Biochemistry 5, 139-143 (1966).
9. Johnston, R. B., and Baehner, R. L., Chronic granulomatous dis-
ease: Correlation between pathogenesis and clinical findings. Pedi-
atrics 48, 730-739 (1971).
10. Salmon, S. E., Cline, M. J., Schultz, J., and Lehrer, R. I., Myelo-
peroxidase deficiency. Immunologic study of a genetic leukocyte
defect. N. EngI. J. Med. 282, 250-253 (1970).
11. Wolff, S. M., Dale, D. C., Clark, R. A., et aL, The Chbdiak-Higashi
syndrome: Studies of host defenses. Ann. Intern. Med. 76, 293-306
(1972).
12. Gibson, Q. H., Swoboda, B. E. P., and Massey, V., Kinetics and
mechanism of action of glucose oxidase. J. Biol. Chem. 239,3927-3934
(1964).
13. Witte, D. L., Brown, L. F., and Feld, R. D., Effects of bilirubin on
detection of hydrogen peroxide by use of peroxidase. Clin. Chem. 24,
1778-1782 (1978).