Analyst, September, 1973, Vol. 98, $9. 663-672 663 Assay of Micro-scale Amounts of Hydroperoxide and of Iodine in Aqueous Non-ionic Surfactant Solutions by a Spectrophotometric Method BY E. AZAZ, M. DONBROW AND R. HAMBURGER (Pharmacy Department, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem, P.O.B. 12065, Israel) An iodimetric - ultraviolet spectrophotometric method has been developed for the quantitative determination of trace concentrations of hydroperoxide (down to It involves a simple extrapolation procedure based on the kinetics of iodine fading in these systems. This method was also applicable to macro-scale amounts when the classical titrimetric method could not be used. Hydrogen peroxide and hydroperoxides of some polyethylene glycols, in the presence of Cetomacrogol 1000, gave reproducible results that were in good agreement with the results obtained by use of the titrimetric method.M) in the presence of etheric non-ionic surface-active agents. THE most universally used method for the determination of hydroperoxide is that developed by Lea.lS2 It is based on the titration of iodine liberated from potassium iodide by oxidising species. This method was found not to be applicable in the presence of non-ionic surfactants as up to 40 per cent. of the iodine was unavailable for titrati0n.~-6 Another disadvantage of Lea’s method is that it is sensitive only to concentrations above 5 x M ; lower concen- trations give errors of up to 50 per cent.7 Nevertheless, the necessity of carrying out such determinations was encountered in studies of autoxidisable substances, such as benzaldehyde,* linoleic acidg and vitamin A,l0 which were rendered soluble with the aid of non-ionic surface- active agents.The development of micro-scale methods for the quantitative determination of hydroperoxides in the presence of non-ionic surf ace-active agents would enable the initial stages of autoxidation to be investigated. Micro-assays of iodine by ultraviolet spectroscopy in solvents other than aqueous non- ionic surfactants have been de~eloped.~9ll According to Heaton and Uri,7 calibration graphs of absorbance against concentration for molecular iodine added to potassium iodide solutions were identical with those obtained for iodine liberated from potassium iodide by hydro- peroxides. They also found that up to concentrations of 5 x 10-4 M Beer’s law was obeyed.The object of the present work was to find conditions under which the ultraviolet absorbance of iodine would be applicable to the determination of hydroperoxides present in non-ionic surfactants. The ultraviolet absorbance of iodine at 360 nm in non-ionic surfactants was used by several workers for the determination of the critical micellar concentration.12-14 However, there seem to be discrepancies as to the position of the iodine absorption peaks. Earlier papers5?l5 reported the peak to be at 370 nm. Ross arid Olivier14 state that the critical micellar concentration has to be determined within 1 hour from the preparation of the solution as otherwise the ultraviolet spectrum changes. Spectral changes were also observed by other workers, occurring with change in concentration ratio, 1 2 9 1 3 9 1 6 storage timelsJ7 and electrolyte composition.13 A maximum absorbance in the region of 360 to 370 nm for I,- was also reported for non-aqueous solvents.18-20 After standardisation of the concentration of surfactant, the temperature, the storage time and the electrolyte composition, this absorption region (360 to 370 nm) provided reproducible assays of iodine liberated by peroxides present in the surfactant system, as can be seen from the results given below.A variety of other methods for the determination of micro-scale amounts of hydroperoxide, none of which was checked for systems containing surfactants, have been described. However, the iodimetric method is the most generally applicable.21 The method described in the present paper is used in routine testing in our laboratory for monitoring oxidisable drugs, for which it has proved to be indispensable and results obtained will be published elsewhere.@ SAC and the authors.664 MATERIALS AND REAGENTS- Low Mills, Leeds. AZAZ et al. : ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, Vol. 98 METHODS Cetomacrogol 1000 B.P.C. (Texofor A IP)-Supplied by Glovers Chemicals Ltd. , Wortley Iodine. Potassium iodide (iodate-free)-Analytical-reagent grade material was used. Sodium thiosulphate solution-BDH Chemicals Ltd. Concentrated Volumetric Solution Ammonium molybdate. Nitrogen gas, 99-999 per cent. Sodium dihydrogen orthophosphate. Disodium hydrogen orthophosphate.PoZyethylene glycol 400, 1500, 4000 and 6000. Hydrogen peroxide solution-E. Merck's 30 per cent. analytical-reagent grade material was used. Buflered solution of 9otassium iodide in 0.1 M phosphate (pH 6), 5 per cent.-This solution was prepared by mixing 87.7 ml of 0-2 M sodium dihydrogen orthophosphate solution with 12.3 ml of 0.2 M disodium hydrogen phosphate solution,22 adding 10 g of potassium iodide and diluting to 200 ml. Iodine, sodium thiosulphate and hydrogen peroxide solutions were prepared and standard- ised according to Kolthoff and Sande1LZ3 Cetomacrogol 1000 solutions were freshly prepared from approximately 20 per cent. stock solutions and standardised by making refractive index measurements on an Abbk 60 refractometer (Bellingham & Stanley Ltd., London) (see Fig.1). For each batch a separate calibration graph was drawn by using accurately weighed and diluted samples. The differences were not great (n: for 20 per cent. solutions ranged be- tween 1-3580 and 1.3602). was used. 1.330 -o 15 0 5 Concentration, per cent. Fig. 1. Refractive index of Cetomacrogol a t 25 "C uemws con- centration (batch A) APPARATUS- A Hilger Ultrascan H999 recording spectrophotometer was used for qualitative work, and a Hilger Uvispek H700 spectrophotometer with temperature-regulating water-jacket for quantitative, time-dependent readings. Readings were made in the concentration range of iodine from 1 x PROCEDURES- 1. Spectrophotometric determination of iodine in the presence of Cetomacrogo LIodine solutions in concentrations of about 0.1 M were diluted with 5 per cent.potassium iodide solution to about 1 x M. Final dilutions to concentrations of 1 x lou4 to to 1 x M by using 0.5 to 4-cm stoppered cells. to 1 xSeptember, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS 665 1 x M were made by adding appropriate amounts of these iodine solutions to calibrated flasks containing appropriate amounts of concentrated, freshly prepared Cetomacrogol solution (5 to 10 per cent.) in buffered (at pH 6) 5 per cent. potassium iodide solution. The contents of the flasks were diluted to volume with more buffered 5 per cent. potassium iodide solution to make a final sample of buffered solution (pH 6) containing 1 per cent. of Cetomacrogol and 5 per cent. of potassium iodide with a concentration of 1 x All of the dilution steps were performed in flasks protected from the light and under purified nitrogen and the iodide was free from iodate.The time of bringing the iodine solution into contact with Cetomacrogol solution was taken as zero time. Absorbances were measured a t a wave- length of 370 nm a t 5 minute intervals and at controlled temperatures for the first half hour. Results were extrapolated to zero time as in Fig. 5 (a) and the concentration of iodine was calculated by using 20 000 as the molar extinction coefficient. This determination and the following determination were performed in triplicate. The results were identical throughout. 2. Spectrophotometric determination of hydrogen pevoxide in the presence of Cetomacrogol- Standardised 0.05 M hydrogen peroxide solutions were diluted with buffered Cetomacrogol and potassium iodide solutions as described above for iodine, except that before final dilution to volume 0-05 ml of 3 per cent.ammonium molybdate solution was added to catalyse the liberation of iodine. The time of mixing hydrogen peroxide solution with Cetomacrogol and potassium iodide solution was taken as zero storage time of liberated iodine. For both iodine and hydrogen peroxide determinations, the reference cell was filled with a freshly prepared solution containing buffered 1 per cent. Cetomacrogol containing 5 per cent. of potassium iodide (iodate-free). 3. Spectrophotometric determination of hydroperoxides of etheric substances in the Presence of Cetomacrogol-Aged samples of polyethylene glycol 400 and of 10 per cent.solutions of polyethylene glycol 1500, 4000 and 600Q were diluted as described above for iodine and hydrogen peroxide and the hydroperoxides determined by the above spectrophotometric method. Spectrophotometric method f o r the determinatioa of unknown amounts of hydroperoxides in micellar solutions-The micellar solution containing the hydroperoxides was added to freshly prepared solutions of Cetomacrogol buffered with potassium iodide solution (pH 6) in amounts such that the final concentration was 1 per cent. of Cetomacrogol in 0-1 M phosphate buffer at pH 6 containing 5 per cent. of potassium iodide in flasks protected from the light and swept with purified nitrogen. The time of addition of the hydroperoxide to the potassium iodide solution was taken as zero time and the absorbances were read in stoppered, well filled cells every 5 minutes for half an hour at 370 nm, being extrapolated to zero time as in Fig.5 (a). The reference solution contained all components except the hydroperoxides. When the ab- sorbance was unsuitable, preliminary dilutions were performed to being the final concentrations of iodine within the range lo-* to Each hydroperoxide group is equivalent to 1 mol of iodine liberated and the concentration is calculated by using 20 000 as the molecular extinction coefficient. 5. Titrimetric determination of hydroperoxides in etheric materials-The hydroperoxide content of aged samples of polyethylene glycol 400 and 10 per cent. aqueous solutions of polyethylene glycol 1500, 4000 or 6000 were determined as follows.The sample was placed in a 250-ml stoppered Erlenmeyer flask, 20 ml of a mixture of acetic acid - chloroform (3 + 2) were added, the solution was purged with nitrogen and 2 ml of a freshly prepared saturated aqueous solution of iodate-free potassium iodide (nitrogen-saturated) were added as quickly as possible. The stoppered flask was then shaken for about 10 minutes. Water and starch solution were added and the sample was titrated against standard 0.01 M sodium thiosulphate solution. All determinations were carried out in triplicate. RESULTS AND DISCUSSION Fig. 2 (a) shows the ultraviolet spectrum of iodine, recorded at 25 "C, 5 and 20 minutes after the iodine was added to a buffered 1 per cent. Cetomacrogol solution in the presence of excess (5 per cent.) of potassium iodide.(See Factors that affect the rate of iodine fading in the presence of Cetomacrogol.) The spectrum shows peaks a t 294 and at 370 nm, the intensi- ties of which decrease with time. to 1 x 10-6 M of iodine. 4. M.666 AZAZ et d.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, VOl. 98 , 7 0.8 - fa) fb) 0.6 - - 0-4 - I 1 1 1 1 294 300 350 370 400 294 300 350 370 400 Wavelength] nm Fig. 2. Absorption spectra of iodine in a buffered 1 per cent. Cetomacrogol solution contain- ing 5 per cent. of potassium iodide at 25 "C, measured 5 minutes (solid line) and 20 minutes (broken line) after: (a), direct addition of iodine to the system; and (b), liberation of iodine by hydrogen peroxide added to the same system Table I summarises the data reported in the literature on aqueous and micellar iodine The differences may be accounted for in part by variations in the equilibrium spectra.situations encountered. For iodine in pure water, equilibrium in the reaction I, + I- + I, lies predominantly to the left and the low-intensity bands of hydrated molecular iodine at 270 and 460 nm are observed,lg whereas in the presence of potassium iodide, equilibrium is to the right and the high-intensity bands in the 290 and 350-nm regions are due to I,. The bands are sensitive to I- concentration,ll the reaction being incomplete even at high I- ion concentration^^^ (Table I, A and B). The I, bands are not detected at the concentrations used for measurement of the 1, absorption.TABLE I ULTRAVIOLET ABSORPTION DATA FOR IODINE SPECIES IN AQUEOUS AND I N MICELLAR SYSTEMS Medium Iodine species Amax./nm cmax. Water . . .. .. . . I, 270 121 Water with excess of KI . . I, 460 704, 746 See B - See A - 13- 1,- 350 26 040 352 352 25 120 353 26 400 285 - 287-5 40 000 289 41 300 25 417 to 26 968* - Aqueous Cetomacrogol (and I, micellar 390 other non-ionic surfactants) I, + I, micellar % I,- 388 * s $ 3 § , + * § *, 1, 9 ** $ 8 § * + 360 to 390 360 293 to 294 * + 370 D + I 9 Aqueous Cetomacrogol in I, micellar 4 I,- 362 to 363 + presence of excess of KI - 370 20 6oot 293 to 294 294 36 6OOt * Absorption intensities rise with addition of potassium iodide. t Data from present work. E values by extrapolation method. 3 Amax and intensity depend on surfactant concentration.Amax. and intensity depend on storage time. Reference 16, 19, 24 - - - 16, 19, 24 11 t 24 19 24 t 16 17 12, 13, 16, t 14 4, 15, t 16, t 16 t 16 tSeptember, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS 667 In the surfactant systems (Table I, C and D), peaks are observed between 360 and 390 nm, which are very sensitive to surfactant concentration in the absence of iodide, and at 293 to 294 nm.12-16 These are probably overlapping peaks of equilibrium mixtures of hydrated and micellar iodine and triiodide, together with a molecular complex of iodine and surfactant (Amax. 390 nm), the last peak being distinctive in the vicinity of the critical micellar con- centration.16 On addition of iodide to the iodine - surfactant system the intensities rise significantly.The 390-nm peak is replaced by the peak in the 360-nm region and the spectrum is much less sensitive to surfactant concentration,l6 presumably because of the formation of I;. The degree of conversion of I, into 1, is less than in non-micellar systems because, with the surfact- ant, species other than 1, may be formed. This may account for the lower extinction co- efficient obtained in the presence of Cetomacrogol, as well as for the discrepancies in the litera- ture (reporting peaks from 360 to 370 nm), which may be due to differences in composition, such as pH and salt concentrations, which alter the equilibrium. 0.50 I I I I I I I I 0 2 4 6 Ti me / hours Fig. 3. Dependence of iodine absorbance at 370 nm on time a t 25 "C. Iodine added to buffered 1 per cent.Cetomacrogol solution containing 5 per cent. of potassium iodide A systematic study of the absorption of iodine at 370 nm, covering 8 hours in controlled conditions (1 per cent. of Cetomacrogol and 5 per cent. of potassium iodide in phosphate buffer at pH 6 and at a temperature of 25 "C) is shown in Fig. 3. After a period of about 1 hour the intensity drops drastically and might reach such a low value as to be below the optimum range of the instrument. Moreover, the intensity does not reach a constant value within 24 hours. Therefore, no single time selected during this period would be applicable to quantitative analyses ; calculations reported in the literature are therefore incorrect. On the other hand, a graph of the logarithm of absorbance at either of the observed peaks (370 and 294 nm) against storage time gives a straight line within the first half hour after addition of iodine to the system (Fig.4). Extrapolation of the line to zero storage time gives an intercept that proves to be the absorption of the amount of iodine actually added to the system. At controlled pH and temperature, the slopes of such lines differ with the Cetomacrogol batch used and with the initial iodine concentration [Fig. 5 ( a ) ] . However, the values extrapolated to zero storage time depend solely on the amount of iodine added. A log - log graph of these extrapolated absorption values against initial I, concentration is a straight line and gives an intercept representing log E , where E is the molar extinction coefficient of iodine in Cetoma- crogol [Fig.5 (b)]. It is evident from Fig. 5 (a) and (b) that Beer's law is observed within the iodine concentration range from 1 x to 1 x 10-6 M . Similar behaviour was observed at668 the 244-nm peak. 36 000 at 294 nm. stant at 1:1.18 for the first 30 minutes of storage time. subsequent studies so as to minimise possible interferences. AZAZ et al.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, Vol. 98 The molar extinction coefficient was found to be 20 000 at 370 nm and As can be seen from Fig. 4, the ratio of the intensities at 370 nm and 294 nm remains con- The 370-nm peak was chosen for 0.2 I I I I 0 10 20 Time/ minutes Fig. 4. Semi-logarithmic graph of absorbance of iodine against time: A, a t 294 nm; and B, a t 370 nm (2.7 x M iodine in buffered 1 per cent.Cetomacrogol - 5 per cent. potassium iodide solution a t 25 "C) The validity of the above extinction coefficient for iodine liberated from potassium iodide by hydroperoxides present in the system is evident from Figs. 2 ( b ) and 6, and from Table 11. The spectrum of the iodine liberated from potassium iodide by hydrogen peroxide is identical with that of molecular iodine [Fig. 2 (a) and (b)]. Further, amounts of molecular iodine identical with those liberated by hydrogen peroxide from potassium iodide give coincident curves when the logarithm of absorbance is plotted against the storage time in minutes (Fig. 6). Table I1 shows that hydrogen peroxide concentrations determined by classical titrimetry are in good agreement with those obtained spectrophotometrically by extrapolating the readings at 370 nm to zero storage time at 25 "C.It is important to emphasise that the titrations were carried out on macro-scale amounts of peroxide in non-micellar stock solutions, hence the microdetermination provides an alterna- tive method of determining macro-scale amounts of peroxide in micellar solutions. Although Cetomacrogol was chosen for systematic investigation, similar linear relation- ships were obtained with other non-ionic surfactants such as Tweens. FACTORS THAT AFFECT THE RATE OF IODINE FADING IN THE PRESENCE OF CETOMACROGOL- Efect of temPerature-As the graph of the logarithm of absorbance of iodine against stor- age time is linear [Figs. 4 and 5 (a)], the slope can be used to evaluate an apparent first-order rate constant, which is convenient for evaluating the influence of factors on the rate of the reaction.Fig. 7 (a) shows that the rate constant of fading, for a given amount of iodine in the presence of Cetomacrogol and potassium iodide, rises with temperature. However, the intercept on the log-concentration axis is independent of temperature so that there is no need for temperature control in the analytical extrapolation procedure proposed, provided that solution temperatures do not change during the period of measurement, graphs obtained being linear only under such conditions. Rate constants given in Fig. 7 (a) are nevertheless specific for a given Cetomacrogol batch and for a given initial iodine concentration. Within these limitations, the apparent rate constant changes with temperature in accordance with the Arrhenius equation [Fig.7 ( b ) ] , the energy of activation for this particular system being 15.7 kcal mol-1. Batch efects-With controlled temperature, pH and initial iodine concentration, the slope of the graph of the logarithm of absorbance against time changes with the batch of Cetomacrogol used (Fig. 8). The extinction coefficient obtained by extrapolation to zero storage time was, however, independent of the batch used [Fig. 5 (a) and (b)].September, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS lo Oo0 t I t a, C m e z a n 0 10 20 30 40 6 4 u 2 669 Time/ minutes -Llog concentration Fig. 5. (u) Semi-logarithmic graph of absorbance of iodine a t 370 nm per l-cm path length against time.Intercepts show initial absorbances. Initial iodine concentrations as determined by titrimetry : 1, 1.05 x 2, 5-40 x 3, 4.50 x 4, 4.00 x 5, 2.00 x 6, 5.00 x and 7, 2.00 x M (all in buffered 1 per cent. Ceto- macrogol - 5 per cent. potassium iodide solution at 25 "C). Cetomacrogol batches: a, batch A ; 0, batch B; and A, batch C. (b) Extrapolated initial iodine absorbances vemxs initial conceri- trations corresponding to (a). The intercept of this graph (log absorbance = log E + log concentration, for a 1-cm path length) gives E = 20 000 Efect of initial iodine concentration-Fig. 5 (a) and (b) shows that a t controlled pH, temperature and with a given Cetomacrogol batch, the slope changes with initial iodine concentration without affecting the extinction coefficient obtained by extrapolation to zero time.TABLE I1 COMPARISON OF HYDROGEN PEROXIDE CONTENT AS DETERMINED IN THE PRESENCE AND ABSENCE OF CETOMACROGOL BY TWO METHODS Ultraviolet Experimental Theoretical absorption a t zero concentration from concentration (from Sample Cetomacrogol storage time ultraviolet standardised stock number batch used (l-cm cell) * absorptionliu x lo5* solution)/M x lO5t A A A A D D D D 0.501 0.794 0.318 1.047 0.603 0.50 1 1.189 0.479 2-51 3-97 1-59 5.24 3.02 2.51 5.95 2-40 2.43 3.90 1.67 5-25 3.00 2.44 6.00 2.40 * Proposed method in the presence of surfactant. t Absence of surfactant (Kolthoff and Sande1P3).670 AZAZ et d.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [AW4lySt, VOl. 98 Efect of CetomacrogoZ concentration-A minimal concentration of 0.5 per cent.of Ceto- macrogol was needed for a linear relationship of the type shown in Figs. 4 and 5 (a) to occur. Although throughout this work 1 per cent. of Cetomacrogol was used,-a range of 0.5 to 5 per cent. was also found to be satisfactory (Fig. 9). Efect ofpH-The pattern of iodine fading changes significantly from that described above when the pH extends outside the range 5.8 to 7.0. Only slight changes in the slopes of the graphs described in Fig. 5 ( a ) were observed within this range, which were without effect on the extrapolated zero storage time reading. Because of possible variations in the acidity or alkalinity of surfactant samples, all systems were buffered to pH 6.0 by using phosphate buffer.1.01 1 I I I I I 0 10 20 30 Time /minutes Fig. 6. Comparison of absorption a t 370 nm of iodine added to system (0) with that of iodine liberated from potassium iodide solution by equivalent amounts of hydrogen peroxide (a). Initial iodine concentration in buffered 1 per cent. Cetomacrogol - 5 per cent. potassium iodide solution a t 25 OC: A, 2.5 x M ; and B, 3.2 x 10-5 M APPLICABILITY OF THE METHOD- Clearly, changes in the slopes observed will occur as a result of differences in the batch of surf actant and the initial iodine concentration, whereas temperature, pH and surfactant concentration effects can be fully controlled. However, under the conditions described, slope variations have been shown not to affect the accuracy of the procedure, which invariably gives the correct initial iodine concentration on extrapolation of the absorbances measured over a half-hour period back to zero time.Table I11 summarises results obtained in the quantitative determination of hydro- peroxides present in aged samples of polyethylene glycols25 by two methods, titration in an acetic acid - chloroform medium and the spectrophotometric method in a micellar medium {see Methods). The applicability to hydroperoxides present in all the aged samples of polyethylene glycol 400 to 6000, together with the results obtained on standard peroxide solutions, establishes the technique as a general method for the analysis of macro-scale or micro-scale amounts of these hydroperoxides.September, 19731 AND IODINE I N AQUEOUS NON-IONIC SURFACTANTS 1.0.67 1 - I I I I 1 0 5 10 15 20 25 30 3-2 3-3 3.4 3-5 Time/minutes 1 xi03 T Fig. 7. (a) Rate of fading of iodine a t A, 18 "C; B, 22.5 "C; and C, 30 "C. First- order rate constants ( k ) obtained from slopes of semi-logarithmic graphs for Ceto- macrogol (batch A) and for an initial iodine concentration of 3.98 x M : 1.76 h-1 a t 18 "C; 2.63 h-1 at 22.5 "C; and 5.12 h-1 at 30 "C. (b) Arrhenius plot of results from (a). k = first-order rate constant (apparent) for fading of iodine; and T = ab- solute temperature. Energy of activation calculated from slope = 15.7 kcal mol-1 CONCLUSIONS The kinetics of iodine fading in the presence of an etheric non-ionic surfactant have been followed spectrophotometrically and factors influencing the kinetics have been examined.A procedure is given for the determination of micro-scale or macro-scale amounts of hydroperoxides in the presence of etheric non-ionic micelles based on the liberation of iodine ahd i& spectrophotometric determination o.2 ~ by the extrapolation method described. 0 10 20 3 Time/ minutes Fig. 9. Semi-logarithmic graph of iodine absorbance at 370 nm against time at three different Cetomacrogol concentrations : A, 5 per cent.; 0, 1 per cent.; and 0, 0.5 per cent. Cetomacrogol (batch A) buffered in 5 per cent. potassium iodide solution at 25 "C672 AZAZ, DONBROW AND HAMBERGER TABLE I11 COMPARISON OF HYDROPEROXIDE CONTENT OF AGED SAMPLES OF POLYETHYLENE GLYCOLS AS DETERMINED IN THE PRESENCE AND ABSENCE OF CETOMACROGOL BY TWO METHODS Hydroperoxide content/M x lo3 r -I A Determined by titration in Determined spectrophotometrically Material acetic acid - chloroform medium” in micellar medium PEG 400 2.30 2.30 PEG 400 4-45 4.50 PEG l500t 5.35 5.20 PEG 4000t 5.15 5.15 PEG 6000t 5-95 6-15 * In the absence of Cetomacrogol (Procedure 5).t 10 per cent. solutions. This method has been substantiated by comparison with titrimetric procedures on macro- scale amounts by use of both hydrogen peroxide and hydroperoxides contained in aged samples of polyethylene glycols. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. REFERENCES Lea, C. H., Proc. R. Soc., 1931, 108B, 175. -, J . Soc. Chem. Ind., Trans., 1946, 65, 286. Henderson, G., and Newton, J. M., Pharm. Acta Helv., 1966, 41, 228. , Ibid., 1969, 44, 129. Allawala, N. A., and Riegelman, S., J . Amer. Pharm. Ass., Sci. Edn, 1953, 42, 396. Hugo, W. B., and Newton, J. M., J . Pharm. Pharmac., 1963, 15, 731. Heaton, F. W., and Uri, N., J . Sci. Fd Agric., 1958, 9, 781. Mitchell, A. G., and Wan, L. S. C., J . Pharm. Sci., 1965, 54, 699. Rhodes, C. T., Can. J . Pharm. Sci., 1967, 2, 16. Kern, C. J., and Antoshkiw, T., Ind. Engng Chem., 1950, 42, 709. Custer, J. J., and Natelson, S., Analyt. Chem., 1949, 21, 1005. Carless, J. E., Challis, R. A., and Mulley, B. A., J . Colloid Sci., 1964, 19, 201. Elworthy, P. H., J . Pharm. Pharmac., 1960, 12, 293. Ross, S., and Olivier, J. P., J . Phys. Chem., 1959, 63, 1671. Osol, A., and Pines, C. C., J . Amer. Pharm. Ass., Sci. Edn, 1952, 41, 634. Woodward, R. J., Ph.D. Thesis, University of London, 1962. Ross, S., and Baldwin, V. H., jun., J . Colloid Interface Sci., 1966, 21, 284. Benesi, H. A., and Hildebrand, J. H., J . Amer. Chew. Sac., 1950, 72, 2273. Katzin, L. I., J . Chem. Phys., 1953, 21, 490. Klaeboe, P., Acta Chem. Scand., 1964, 18, 27. Johnson, R. M., and Siddiqui, J. W., “The Determination of Organic Peroxides,” Pergamon Press, Colowick, S. P., and Kaplan, N. O., Editors, “Methods in Enzymology,” Volume I, Academic Kolthoff, J. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Awtrey, A. D., and Connick, R. E., J . Amer. Chew. Soc., 1951, 73, 1842. Lloyd, W. G., J . Polym. Sci., Part A , 1963, 1, 3551. J -- Oxford, 1970, pp. 43 and 113. Press Inc., New York, 1955, p. 143. The Macmillan Company, New York, 1962. Received January 19th, 1973 Accepted April 3rd, 1973