950 Analyst, October, 1990, Vol. 105, pp. 950-954 Spectrophotometric Determination of Hydrogen Peroxide Using Potassium Titanium( IV) Oxalate Robin M. Sellers Central Electricity Generating Board, Berkeley Nuclear Laboratoracs, Berkeley, Gloucestershtre, GL 13 9PB A simple and rapid method for the spectrophotometric determination of hydrogen peroxide using potassium titanium(1V) oxalate is described. The method can be used to measure peroxide concentrations down to about 10 p~hz (0.3mgkg-l) under the most favourable conditions. A variety of corn- plexing and reducing agents, and catalysts of peroxide decomposition, known to interfere with the alternative iodide method for peroxide determination, had no effect. Keywords: Hydrogen peroxide determination; spectrophotounetry; titunium(I V ) Fluoride was found to interfere.oxalate Many methods have been described in the literature for the spectrophotometric determination of hydrogen peroxide.l-lO One of the most sensitive and widely used is that based on the oxidation of I- to This test is not specific for hydrogen peroxide (organic peroxides and many other oxidising agents convert I- into 13-), although the analysis is usually performed in the presence of molybdate, a specific catalyst of the reactiom2 Other redox reactions, such as iron(I1) to iron(III), have also been employed.3 These too measure total peroxides, rather than hydrogen peroxide alone. A specific spectrophotometric test based on the formation of a complex, often written as Ti022 +, between hydrogen peroxide and the titanium(1V) ion has been de~cribed.~-S The published procedures involve lengthy preparations of titanium( IV) sulphate, and little seems to be known about either the optimum conditions for carrying out the measurements, or interference by compounds such as complexing or reducing agents.This paper describes the development of a method for the determination of hydrogen peroxide using potassium titanium(1V) oxalate, the only analytical-reagent grade salt of titanium readily available commercially, and the influence of various additives known to interfere with the iodide method. Experimental Reagents The reagents used in the determina- tion of the hydrogen peroxide, i.e., hydrogen peroxide, potassium titanium(1V) oxalate, [K,TiO(C,04),.2H,0] and sulphuric acid, were from BDH, AnalaR.grade, and were used as received. The organic compounds used in the tests for interference were of laboratory-reagent grade, and were obtained from BDH [ethylenediamine, iminodiacetic acid (IDA)], Aldrich [ethylene- diaminediacetic acid (EDDA), N-(2-hydroxyethyl)ethylenediaminetriacetic acid (HEDTA)] and Eastman Organic Chemicals [ethylenediaminetetraacetic acid (EDTA), nitrilotriacetic acid (NTA)]; glycine and 2,2’-bipyridyl were BDH AnalaR grade, and picolinic acid was Aldrich laboratory-reagent grade, re-crystallised once from water. All other inorganic compounds were BDH AnalaR grade. Hydrogen peroxide solutions were standardised by the iodide method of Allen et a1.,2 taking E for I,- at 350 nm as 25700 1 mol-l cm-I. All solutions were prepared from triply distilled water.Apparatus Absorbance measurements were made with a Cecil CE505 spectrophotometer. Results Development of the Method The titanium(1V) - peroxide complex is yellow - orange in colour and absorbs with a Amax. of about 400 nm. The intensity of this absorption was found to be dependent on theSELLERS 1200 c 1000 E 800 u .- - 0 600 E 8 400 - . 200 0 n I I 951 ~ 0 0.5 1 0 1 5 2 0 2.5 3.0 3 5 4 t log [titanium (IV) oxalatel Fig. 1. Dependence of the absorption of the titanium(1V) - peroxide complex on titanium(1V) oxalate concentration. Measure- ments were made in solutions containing 6 x lo-* M H,O, and 0.1( A), l.O(n) or 4.0(0) M H,SO,,. The arrow indicates the point a t which the concentrations of the titanium(1V) oxalate and hydrogen peroxide were equal. concentrations of titanium(1V) oxalate and sulphuric acid as shown in Figs. 1 and 2 .The measurements were made in solutions containing 6 x M hydrogen peroxide; for the measurements of dependence on the concentration of titanium(1V) oxalate, the sulphuric acid concentrations were 0.1, 1.0 or 4.0 M; and for the measurements of dependence on the concentration of sulphuric acid the titanium(1V) oxalate concentrations were 0.01, 0.02 or 0.04 M. Absorption by components other than Ti0,2+ was allowed for by subtracting the absorbances of solutions prepared in the same way but omitting hydrogen peroxide. -u 600 2.0 2.5 3.0 3.5 4.0 4.5 5.0 4 + log [H*S041 Fig. 2. Dependence of the absorption of the titanium(1V) - peroxide complex on sulphuric acid concentration.Measurements were in solutions containing 6 x M H,,O, and 0.01( 0). 0.02( A) or 0.04( 0) M titanium(1V) oxalate. When the titanium(1V) concentration was greater than the hydrogen peroxide concentra- tion (i.e., under conditions where all the hydrogen peroxide was complexed) only a small dependence on the concentration of titanium(1V) oxalate was found at titanium(1V) oxalate concentrations less than 0.1 M. At higher concentrations the intensity of the absorbance of the titanium(1V) - peroxide complex was much reduced. Varying the sulphuric acid con- centration a t a constant titanium(1V) oxalate concentration had only a small effect. The absorption was a t its most intense at a titanium(1V) oxalate concentration of about 0.02 M and 0.1-1.0 hi sulphuric acid.The A,,,. for the absorption increased with decreasing titanium(1V) oxalate concentration or increasing sulphuric acid concentration. Some experiments were also carried out in which the hydrogen peroxide concentration was varied at constant titanium(1V) oxalate and sulphuric acid concentrations. In all instances the absorbance a t a particular wavelength varied linearly with hydrogen peroxide concentra- tion, a t peroxide concentrations up to about 2 x M, the highest employed.952 SELLERS : SPECTROPHOTOMETRIC DETERMINATION OF HYDROGEN Analyst, Vol. 105 Recommended Procedure Based on the conditions under which the absorption of the titanium(1V) - peroxide complex was at its most intense (i.e., was most sensitive to hydrogen peroxide) the procedure described below appears most suitable for the determination of hydrogen peroxide using titanium(1V) oxalate.Titanium reagent Mix 272 ml of concentrated sulphuric acid (BDH AnalaR) with about 300 ml of distilled water (care should be taken and cooling is required). Dissolve in this mixture 35.4 g of potassium titanium(1V) oxalate, K,TiO(C204),.2H,0, (BDH AnalaR), and make up to 1 1 with distilled water. Procedure Pipette 5 ml of the titanium reagent and 5 ml (or as appropriate) of the sample into a 25-ml calibrated flask and make up to the mark. Measure the absorbance of the solution at 400 nm. A blank, consisting of 5 ml of the titanium reagent and 5 ml (or as appropriate) of sample without the hydrogen peroxide present made up to 25 ml, should also be measured.(The hydrogen peroxide may be destroyed by the addition of platinum black followed by filtration, or by boiling.) Calculation of hydrogen peroxide concentration the molar absorptivity of the titanium(1V) - peroxide complex. of solution the concentration of hydrogen peroxide (in moles per litre) is given by: The concentration of hydrogen peroxide is calculated taking elOO = 935 1 mol-l cm-1 as For x ml of sample per 25 ml where A and A , are the absorbances of the test and blank solutions, respectively, and 1 is the path length of the spectrophotometer cell in centimetres. Effect of Other Solutes The effect of other solutes on the method was investigated by measuring the apparent molar absorptivity of the titanium(1V) - peroxide complex at 400 nm in solutions made up to contain 0.02 M in titanium(1V) oxalate plus 1.0 M in sulphuric acid, 6 x M in hydrogen peroxide and about 0.02 M in solute.The results are summarisedl in Table I. The solutes included a variety of complexing and reducing agents and catalysts of peroxide decomposition. TABLE I EFFECT OF SOME ADDITIVES ON THE TITANIUM(IV) OXAL.ATE METHOD FOR THE DETERMINATION OF HYDROGEN PEROXIDE: All solutions also contained 0.020 M titanium(1V) oxalate, 1.0 IN sulphuric acid and 6.0 x lo-, M hydrogen peroxide. Additive concentration/M :2400/1. mol-1 cm -l Additive None - 935 NaF 0.020 805 coso, 0.020 956 cuso, 0.020 935 NiSO, 0.020 936 Hydrazinium sulphate 0.020 935 Ethylenediamine 0.018 949 Glycine 0.020 936 EDDA 0.002 935 EDTA 0.020 957 HEDTA 0.020 969 IDA 0.020 927 NTA 0.020 950 Picolinic acid 0.024 940 2,2’-Bipyridyl 0.012 949October, 1980 PEROXIDE USING POTASSIUM TITANIUM(IV) OXALATE 953 Only for fluoride was any appreciable interference found.A similar series of tests was per- formed in which the hydrogen peroxide was determined by the iodide method of Allen et aL2 The results can be summarised as follows: (i) Ethylenediamine, EDTA, HEDTA, IDA and NTA, no I,- was formed, as these solutes reduce I,- back to I-. (ii) 2,2'-Bipyridyl and picolinic acid, 1,- formed slowly, and with 2,2'-bipyridyl a purple - black precipitate was obtained. This behaviour probably results from complexation of the molybdate catalyst by these solutes. Hydrazinium sulphate is a reducing agent and not only converts I,- into I-, but also molybdate into a heteropoly blue.(iv) Copper sulphate, large amounts of I, were formed, due to oxidation of I- by Cu2+. ( v ) Glycine, cobalt sulphate, nickel sulphate and sodium fluoride, no interference. (iii) Hydrazinium sulphate, no I,- was formed and the solution turned blue. Discussion The recommended procedure gives concentrations of 0.02 M titanium(1V) oxalate and 1.0 M sulphuric acid in the analysis solution. The high acid concentration prevents precipitation of titanium(1V) hydroxide and ensures that relatively large amounts of base can be present in the peroxide containing solution without interfering unduly with the method through the consumption of protons. An upper limit of about 0.1 M of base in the peroxide sample is probably advisable. The precision of the method is around 1% for peroxide concentrations of 0.5-2.0 x lo-, M (an absorbance of 0.5-2.0 in l-cm cells).Thus, the 13 measurements in Table I that relate to solutes that do not interfere have a relative mean deviation of 1.1%. The relative mean error of the values is also 1.1%. At lower peroxide concentrations the method becomes less precise, the size of the blank reading becoming important. This averaged about 0.001 absorbance units, as measured in l-cm cells, in the absence of any additional solutes, and about double this in the presence of solutes, such as the complexing agents, or higher still in the presence of coloured substances such as Cu2r. The upper limit for detection of peroxide is about 2 mM using l-cm spectrophotometer cells (with an absorbance of 2).This is well below the titanium(1V) oxalate concentration in the mixture, ensuring that all peroxide is complexed. The calculation of the hydrogen peroxide present is based on the measured molar absorp- tivity of the complex a t 400 nm. The intensity of the absorption follows the Beer - Lambert law, so that little or no additional precision is obtained by constructing a calibration graph, although some improvement should be possible by adaptation of the method described by Nea1.O It is clearly important to check for possible interference by other solutes that may be present. The experiments summarised in Table I suggest that the method is relatively free from interference, and appears to be particularly suitable for the determination of hydrogen peroxide in the presence of complexing and reducing agents.Interference by fluoride is not unexpected in view of the ease of formation and stability of the fluorotitanate ion, TiFe2-. I t may also be noted that the method involves considerable savings in time over the earlier procedures, which required lengthy preparations of titanium(1V) by digestion of titanium(1V) oxide in sulphuric acid. It is reportedlo that the sensitivity of the method can be improved by addition of xylenol orange, and experiments have shownll that potassium titanium(1V) oxalate is suitable for use as a source of titanium(1V) under these conditions. The pH of the solution seems to be critical, however, and the effect of complexing and reducing agents is unknown. The lower limit with 1-cm cells is 10 PM under favourable circumstances. I am indebted to Mr. B. Daniel for assistance with some of the measurements. This paper is published by permission of the Central Electricity Generating Board.954 SELLERS References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Schumb, W. C., Satterfield, C. N., and Wentworth, R. L., “Hyclrogm Peroxide,” Reinhold, New Allen, A. 0.. Hochanadel, C. J., Ghormley, J. A., and Davies, R. W., J . Phys. Chem., 1952, 56, 575. Michaels, H. B., and Hunt, J . W., Anal. Biochem., 1978, 87, 135. Jackson, E., Chem. News, 1883, 47, 157. Richardson, A,, J . Chem. Soc., 1883, 63, 1109. Allsopp, C. B., Analyst, 1941, 66, 371. Eisenberg, G. M., Ind. Eng. Chem., Anal. Ed., 1943, 15, 327. Humpoletz, J. E., Aust. J . Sci., 1949, 12, 111. Neal, W. T. L., Analyst, 1954, 79, 403. Gupta, B. L., Microchem. I., 1973, 18, 363. Sellers, R. M., unpublished data. York, 1955, p. 561. Received April 9th, 1980 Accepted May 29th, 1980