ANALYST, JANUARY 1993. VOL. 118 23 Preconcentration of Aniline Derivatives From Aqueous Solutions Using Micellar-enhanced Ultrafiltration Edmondo Pramauro and Alessandra Bianco Prevot Dipartimento di Chimica Analitica, Universita di Torino, 10 125-Torino, Italy Piero Savarino and Guido Viscardi Dipartimento di Chimica Generale ed Organica Applicata, Universita di Torino, 10125-Torino, Italy Miguel de la Guardia and Empar Pewis Cardells Departamento de Quimica Analitica, Universidad de Valencia, 46I00-Burjassotl Valencia, Spain The preconcentration of aniline derivatives present in aqueous solutions containing ionic surfactants was performed using the micellar-enhanced ultrafiltration technique. The efficiency of the analyte recovery in the surfactant-rich retentate was significantly improved by exploiting the electrostatic, hydrophobic and specific interactions between the aggregates and the solute molecules.By working at low pH, in the presence of anionic micelles, the quantitative retention of amines having different su bstituent groups can be achieved. The enriched sample can be analysed directly using high-performance liquid chromatography. Keywords: Micellar-enhanced ultrafiltration; aromatic amine; preconcentration; aqueous sample; high-performance liquid chromatography analysis The unique properties of amphiphilic aggregates have been exploited in separation science to improve existing methods and to develop new procedures in which the use of organic solvents is avoided. 1-5 In particular, micellar-enhanced ultra- filtration (MEUF) is emerging as a promising technique for preconcentration and/or removal of organic and inorganic solutes of environmental concern from aqueous media.&" This technique is based on the association of the analytes to suitable surfact ant aggregates, which are successively blocked by membranes having an appropriate pore size.As the mean relative molecular mass of most micelles is generally higher than 10000 Da, membranes with a relative molecular mass cut-off in this range can be used, thus allowing operation at acceptable flow rates. The solute enrichment takes place in the retentate, whereas the permeate contains only small amounts of free surfactant and unbound solute. If biodegradable amphiphiles are used to treat polluted streams, their presence in the permeate does not present environmental problems.The extent of binding of substrates to the aggregates depends on hydrophobic, electrostatic and specific interac- tions operating between the micellized amphiphile molecules and solutes. Only a limited number of lipophilic solutes have been removed or concentrated from aqueous samples using MEUF (phenols, benzene, DDT, chlorinated hydrocarbons, aliphatic alcohols) and there is a lack of information about the basic relationships which control the performance of the process. In most instances, the efficiency of MEUF can be directly correlated to the partition coefficients of the analytes in micellar media, which are in turn dependent on the nature of the surfactant and on the conditions of the medium. For example, the effective removal of some chloroaromatic carboxylic herbicides from aqueous solutions can be accom- plished by regulating the pH and exploiting the electrostatic attraction between the ionized analytes and oppositely charged aggregates.10 Aromatic amines are an important class of anthropogenic compounds of environmental concern. Among their main applications, these harmful products are used in the synthesis of a variety of organic dyes. Hence, they may be present at trace levels in commercial products and in aqueous effluents discharged from dyestuff manufacturing and dyeing plants.]' Their determination in such samples often involves a time- consuming enrichment stage (typically a liquid-liquid extrac- tion from basic solutions), followed by analysis using gas chromatography (GC) or high-performance liquid chromato- graphy (HPLC).As GC usually requires the prior derivatiza- tion of the amines, HPLC is employed in a large number of analytical protocols. The sensitivity of this method varies with the detection mode, but it is sufficiently high to allow the determination of aromatic amines present at trace levels in effluents and environmental samples. 12-15 In the present work, a series of substituted anilines, including some hydrophobic dye intermediates, were separ- ated and concentrated from aqueous surfactant solutions using the MEUF approach. The use of organic solvents, which is required in other preconcentration techniques (e.g., solid- phase extraction) to dissolve the hydrophobic analytes, is avoided.The amphiphiles chosen were sodium dodecyl sulfate (SDS) and hexadecyltrimethylammonium bromide (HTAB) , which are commonly used and inexpensive products. Owing to their ionic character, these aggregates can exert electrostatic attraction or repulsion towards the amine cations, allowing the partition coefficients of the substrates investigated to be varied. The effects of pH, ionic strength, surfactant concentration and type of micelle on the solute recovery were examined. Experiments performed with some aniline derivatives indi- cated that the MEUF-based enrichment of these analytes can be coupled with micellar HPLC or conventional HPLC analysis without interference problems. Experimental Apparatus A Cary 219 (Varian) and a Uvikon 930 (Kontron) spectropho- tometer were used.Potentiometric measurements were per- formed with a Dosimat 655 (Metrohm) automatic titrimeter. A liquid chromatograph, consisting of a PM 330 pump (Violet) , a Merck-Hitachi LC-4200 ultraviolet/visible (UV/ VIS) detector and an ERC-7512 refractive index detector, was used for the HPLC measurements. Ultrafiltration cells (S-43- 70) supplied by Spectrum and Spectra/Por-C hydrophilic cellulose membranes were also employed. Reagents The aniline derivatives (Fluka and Merck) were used as received. Stock solutions of each compound were prepared in24 ANALYST. JANUARY 1993, VOL. 118 95% ethanol (Merck). The surfactants SDS and HTAB were of analytical reagent-grade (Merck) . Sodium chloride, HCI and NaOH were from Carlo Erba. Acetonitrile, methanol and propan- 1-01 of HPLC-grade (Merck) were used. Doubly distilled water was used throughout.The following heterocyclic intermediates [2-(4-amino-2- hydroxypheny1)benzoxazolesl , used in the synthesis of dis- perse azo dyes water) ,I6 were procedures. 17 (i.e., those dyes that are almost insoluble in prepared according to previously reported CiH I : X = O II: x = s 111: X = NH Procedure Measurement of the binding constants The micellar HPLC technique was mainly used to evaluate the extent of binding between the analytes and the anionic aggregates. The retention data were analysed as a function of the micellized surfactant concentration (c,,) in the mobile phase, according to the Armstrong-Nome partition model. 18 The following equation was used: where V,, V , and V, are the stationary phase, mobile phase and elution volume, respectively, PMw and Psw are the partition coefficients of the solutes between the micelles and the bulk aqueous phase, and between the stationary phase and the aqueous phase, respectively, and V is the paltial molar volume of the amphiphile.The term (PWM - l ) V gives the binding constants to the aggregates (KB).19 Mobile phases containing the micellized surfactants (with- out any organic modifier) were adjusted to the appropriate pH (about S), filtered through a 0.45 pm cellulose filter (Milli- pore) and carefully de-gassed. Each solute was dissolved in the surfactant solution before the runs and 20-50 pl of these samples were injected into the loop. The concentrations of the aniline derivatives were in the range from 1 x 10-3 to 3 X 10-3 mol 1-1.The elution was performed at constant flow rate 1 ml min-I), at room temperature (24 k 1 "C). The stationary phase of the column was LiChrospher 100-CN (10 pm) (Merck). The absorbances of the solutes were monitored at 240 nm. The absorbance variation method20 was also used to estimate the binding constants of some compounds to cationic micelles. The UV absorption of the undissociated analytes was measured at different wavelengths, varying the concentration of the micellized surfactant. The corresponding data were treated according to the following equation: where A , A , and A , are the measured absorbances in surfactant solution, in water and when the substrate is completely bound, respectively. For some polysubstitued hydrophobic compounds, the partition data were estimated by starting from the contribu- tion of each substituent to the free energy of transfer of the molecule to the micelle.The following equations21 are valid within each series of compounds: A p t = Ap*, - Apow = -RT ln(55.5 KB) (3) (4) where Apot is the free energy of transfer per mole of solute from water to micelles and ApoAn is the contribution of the aniline moiety to the total free energy of transfer. The term zApoSubst accounts for the contributions of the substituent groups. Ultrafiltration procedure The experiments were carried out in cylindrical ultrafiltration cells (capacity 70 ml) equipped with a magnetic stirrer rotating slightly above the UF membrane to reduce the concentration polarization effect.The Spectra/Por-C10 hydrophilic cellulose membranes have a relative molecular mass cut-off of about 10000 Da, which is sufficiently small to reject the micelles completely. The membranes were thoroughly washed with water before the runs in order to remove the incorporated wetting agents. A pressure of 300 kPa was maintained in the cell by nitrogen in order to obtain a regular flow of the solution through the membrane. The cells were filled with 30 ml of solution and each run was terminated after the collection of 25 ml of permeate. The stirrer bar was rotated at a constant speed in all the experiments. Each solute was determined both in the retentate and in the permeate by HPLC, according to the following standard procedure: aliquots of 20 pl of retentate solution were injected into the chromatograph, equipped with LiChrocart 125-4 columns (Merck) filled with LiChrosphcr 100 RP-18 (5 pm) stationary phase. Elution was performed using acetonitrile- water (50 + 50).The pH of the eluent was adjusted to about 8 using a borate buffer. The flow rate was kept constant (1 ml min-1) and the wavelength of the UV detector was selected in the range 250-320 nm. Calibration graphs were obtained from the corresponding standards. The analysis of retentate samples was performed in the same way after dilution with the hydro-organic eluent (usually 1 + 9). This step minimizes the interference effects arising from concentrated surfactants and allows the sample viscosity to be decreased. The efficiency of MEUF was estimated through the evaluation of the rejection factor, R , defined as follows: R = 1 - c,/c~ where cp and co are the analyte concentrations in the permeate and in the initial solution, respectively.Each reported R value represents the mean of three independent determinations. Preconcentration-determination experiments The preconcentration of a test mixture of aniline compounds from aqueous dilute solutions was performed using MEUF. Typically, aliquots of 70-100 ml of solutions containing the analytes at the ppb level and each surfactant at a concentration slightly above the critical micellization concentration (c.m.c.) were ultrafiltered until a volume ratio of retentate : permeate in the range 0.07-0.05 was obtained. Working with Spectra/ Por-C 10 membranes under a nitrogen pressure of 300 kPa, about 80 min are necessary for each ultrafiltration run.Samples were taken from the resulting viscous retentate solution with a syringe and immediately analysed by HPLC. As fairly concentrated amphiphile samples are obtained, the HPLC analysis using micellar eluents is recommended in order to eliminate the interference effects arising from concentrated amphiphiles. For example, SDS micellar solu- tions in the concentration range 0.1-0.3 mol 1-1 are able to elute the analytes fairly rapidly, preconcentrated using the same surfactant. The addition of 3% of propan-1-01 to the micellar eluent improves the chromatographic efficiency.2' A LiChrospher 100 RP-18 (5 pm) column was used and the flow rate was kept constant (1 ml min- 1). Ultraviolet detection was performed at 240 nrn for most anilines, whereas for com- pounds 18-20 detection at 340 nm is more sensitive.When HTAB is used to preconcentrate the analytes, the analytical step becomes more difficult because the surfactantANALYST, JANUARY 1993, VOL. 118 25 concentration in the chromatographic eluent is limited by phase separation phenomena (Krafft point). As the surfactant concentration in the eluent must be lower than about 0.04 mol 1-1 at room temperature, the elution becomes much too slow. In these instances, the determination step can be improved by using hydro-organic eluents [usually methanol-water (50 + 50 v/v)] containing an appropriate amount of dissolved cationic surfactant to reduce the interference effects. Higher organic solvent-to-water ratios are recommended to elute highly hydrophobic derivatives, such as compounds 18-20.Results and Discussion Partition and Ultrafiltration Data Most experiments were conducted in the pH range at which the aniline derivatives are present in the undissociated form. The binding constants of the analytes were measured or calculated under the same conditions. The precision of the measurements of K g , expressed as the relative standard deviation, is in the range 8-15%, depending on the hydro- phobicity of the solute.23 The precision of the determination of R is better than 6%. The rejection and partition data for all the compounds investigated are presented in Table 1. Binding constants higher than 2000 1 mol-1 were estimated for the heterocyclic derivatives 18-20 in HTAB micellar solution, from absorbance variation measurements.These lower limit values are in agreement with previous data concerning compounds 18 and 19 in strongly alkaline HTAB solutions, determined using a spectrofluorimetric approach .25 It can be seen that, irrespective of the surfactant used, partitioned compounds having binding constants lower than a threshold value of about 600-700 1 mol-1 are only partially recovered in the retentate. The data also indicate that cationic micelles are more efficient systems, because the positively charged head-groups strongly interact with the x-electrons of the aromatic ring. The results obtained here are in good agreement with previous findings concerning the removal of phenols,6 alcohols* and chloroaromatic pollutants10 from aqueous solutions by MEUF.Table 1 MEUF rejections and binding constants of aniline derivatives in the presence of SDS and HTAB micelles. Experimental conditions: pH 8.0-8.5 SDS HTAB Solute KB/l mol-I R KB/l mol-I R 1: Aniline 2: 4-Fluoroaniline 3: 4-Chloroanilinc 4: 4-Bromoaniline 5: 4-lodoaniline 6: 4-Methylaniline 7: 4-Ethylaniline 8: 4-Isopropylaniline 9: 4-tevt-Butylaniline 10: 4-Nitroaniline 11: 4-Cyanoaniline 12: 3-Fluoro-4-methylaniline 13: 5-Chloro-2-methoxyaniline 14: 2-Chloro-4-fluoroaniline 15: 2-Fluoro-4-bromoaniline 16: 2-Chloro-4-bromoaniline 17: 2-Nitro-4-methylanilinc 18: Compound 1 19: Compound I1 20: Compound I11 23 27 56 75 119 43 79 140 260t 33 27 50 133 707 90-1 220 58 800 lo00 1200 0.34 39* 0.40 0.35 SO* 0.50 0.61 175* 0.85 0.68 240" 0.91 0.87 430* 0.97 0.53 55* 0.61 0.75 110' 0.80 0.86 270* 0.92 0.85 47St 0.95 0.52 180 0.81 0.26 70" 0.55 0.62 94t 0.82 0.85 355 0.94 0.82 225-3- 0.85 0.88 350-t 0.96 0.97 12001- 1.00 0.70 350 0.93 1.00 - 1 .OO 1.00 - 1.00 1.00 - 1.00 * Data taken from ref.24. -3- Binding constants calculated from the hydrophobic contribution of each substituent. In order to improve the performance of MEUF it is necessary to increase the binding constants of the partitioned analytes to micelles by introducing additional interactions. In an attempt to achieve this, the effects of pH and ionic strength were investigated. Effect of pH The rejection factors of all the analytes examined in SDS can be significantly increased by lowering the pH, and this can be attributed to the electrostatic attraction between the anilinium ions and the negatively charged aggregates.The effect is, therefore, related to the mole fraction of the charged form present in the system. In order to evaluate this contribution, the apparent acid constant of some protonated aniline derivatives was determined spectrophotometrically in the presence of SDS and HTAB micelles. The surfactant concen- tration was the same as that of the initial working solutions (2 x 10-2 mol 1-1 for both surfactants). The uncertainty of these measurements is about 15%. The corresponding data are presented in Table 2. Fig. l(a) shows the variation of the rejection factor as a function of pH for some aniline derivatives in SDS. It appears that the effect is particularly important for the more hydro- philic compounds, whereas the electrostatic contribution is less relevant for analytes having larger binding constants. For example, ethylaniline can be quantitatively retained at pH 2-3 Table 2 Apparent acid constants of some anilinium ions in SDS and HTAB micellar media, at an ionic strength of 0.1 mol I-' (NaCI) Compound SDS HTAB Aniline 5.4 4.8 4-Ethylaniline 6.5 5.9 4-Isopropylaniline 6.5 5.8 4-Chloroaniline 5.2 4.0 1 .o 0.9 0.8 0.7 0.6 ~ 2 3 4 5 6 7 8 0.9 c 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 I I I I 1 2 3 4 5 6 7 8 PH Fig.1 Variation of the rejection coefficient with pH in the recovery of aniline (I), 4-chloroaniline (3) and 4-ethylaniline (7) by MEUF. ( a ) 2 x 10-2 moll-1 SDS and (b) 2 x 10-2 moll-' HTAB26 1.0 ANALYST, JANUARY 1993, VOL.118 -- ________ -+--------4 - 16 ---- *---- (at which the protonated form largely predominates) by SDS micelles, whereas at pH 8, about 25% of this solute was found in the permeate. For pH values in the pK,,,,,, range, the contribution of the electrostatic attraction becomes highly significant. Ultrafiltration experiments performed using SDS micelles at pH 2-4 gave quantitative recovery for all the compounds examined. As expected, the effect of pH on the retention yield is the opposite when cationic HTAB micelles are employed [see Fig. 1(6)]. Effect of ionic strength N o significant change in the value of R was observed in MEUF experiments conducted at higher pH with both anionic and cationic micelles, even after addition of 0.1 moll-' of an inert salt (NaCI).In contrast, marked effects were observed when the analytes were present in their protonated form, owing to the lowering of the electrostatic potential of the charged micelles. For the aniline-SDS system, where the electrostatic attrac- tion significantly increases the solute binding to the aggre- gates, the retention efficiency is lower at higher ionic strength. More hydrophobic compounds, such as alkylanilines, showed little effects because the main contribution to the binding is not electrostatic. These different situations are depicted in Fig. 2. The effect of the ionic strength on the rejection of protonated anilines in the presence of cationic HTAB micelles is the opposite, with a significant increase in binding at higher ionic strength for hydrophilic solutes.Effect of surfactant concentration Although the total surfactant concentration has to be mini- mized in MEUF experiments, in order to avoid the formation 1 .o 0.95 0.90 ry 0.85 0.80 0 0.02 0.04 0.06 0.08 0.10 cNaCl/mol I-' Fig. 2 Effect of ionic strength on rejection [SDS] = 2 x 10-2 moll-', pH = 3, added salt: NaCl. Analytes: aniline (l), 4-methylaniline (6) and 4-isopropylaniline (8) of viscous layers above the membrane, this experimental parameter can also be adjusted within certain limits. In particular, an increase in the number of micelles favours the retention of partially bound compounds, whereas the effect is less important for hydrophobic species. The MEUF experi- ments performed at pH 8.0 with compounds 1 and 16 in SDS micellar solutions showed the influence of this parameter (see Fig.3). Preconcentration Experiments The concentration factor in MEUF is limited by the viscosity of the retentate solution and/or by the Krafft point (when ionic surfactants are used). Hence, the preconcentration experi- ments were performed by starting from very dilute surfactant solutions, just above their corresponding c.m.c. The initial concentrations were 2 x 10-3 mol 1-1 (HTAB) and 1 X 10-2 mol 1-1 (SDS). Fig. 4(a) shows the chromatographic profile obtained by working with 0.15 rnol 1-1 SDS (3% v/v of propan-1-01 was added to this eluent), after direct injection of 20 pl of SDS-containing retentate. The separation of two aniline derivatives preconcentrated using HTAB is shown in Fig. 4(6), where methanol-water (50 + 50) (containing 3% v/v of 0.04 rnol 1-1 HTAB) was used as eluent.The inconvenience associated with the injection of concen- trated surfactant solutions into hydro-organic eluents, in particular the poor reproducibility of the peaks and the baseline disturbance, was significantly reduced by working a) 0.9 j- 0.4 f- 24 16 8 Retention time/min 0 0.1 1, I I I 0.01 0.02 0.03 0.04 csos/mol 1-1 Fig. 3 Effect of SDS concentration on rejection efficiency at pH 8.0. Aniline (1) and 2-chloro-4-bromoaniline (16) Fig. 4 ( a ) Separation of aniline derivatives preconccntrated using SDS and eluted with 0.15 rnol 1-1 SDS containing 3% v/v of propan-1-01. Analyte concentration: 7 X lo-' mol 1-'. Compound 5 : 1 x 10V mol I-'. 0.005 a.u.f.s. Detector wavelength: 240 nm. ( b ) Separation of anilines after preeoncentration with HTAB.Eluent: methanol-water (50 + 50) plus 3% v/v 0.04 mol I-' HTAB. Analyte concentration: 1 X 10-6 rnol 1-1. 0.005 a.u.f.s. Detector wavelength: 240 nmANALYST, JANUARY 1993, VOL. I18 27 Table 3 Sensitivity of the HPLC dctcrmination of some aniline derivatives after prcconcentration by MEUF Detection limit/ng Compound SDS HTAB CF* - 1 0.6 14 14 7 0.7 - 14 8 0.8 - 16 2.5 1.2 14/20 5 - 1.3 20 19 - 4.8 20 * cF = Concentration factor. with these surfactant-modified cluents. 'The added surfactant saturates the stationary phase, avoiding abrupt changes in the properties of the packing material when the sample is injected. Moreover, by using recrystallized surfactants, the interfer- ences arising from amphiphile impurities can also be mini- mized. The detection limits obtained by working with several aniline derivatives, after preconcentration of solutions containing 5 x 10-8 moll-' of each analyte, are presented in Table 3.The reported detection limits were determined under the same experimental conditions by injecting decreasing amounts of pure standards. The concentration factor attained is indicated by cF. Quantitative recovery of the analytes in the retentate phase was confirmed. Conclusions The results presented here indicate that effective preconcen- tration of several aniline derivatives from water (or aqueous wastes) can be performed by using MEUF. The percentage recovery of each analyte is related to the corresponding binding constants to the aggregates, which can in turn be increased by selecting the appropriate surfactant, the pH, and other experimental conditions.In particular, experiments conducted in acidic media with anionic micclles allow the quantitative recovery of both hydrophobic and polar derivatives. The retentate samples obtained by using these surfactants (e.g., SDS) can be readily injected into the chromatograph and analysed after elution with aqueous micellar solutions containing the same amphiphile. The use of hydro-organic HPLC eluents to analyse the retentate is also possible if surfactants are added to these solvents and the column is carefully conditioned. The final surfactant concentration in the retentate has to be limited in order to prevent the formation of very viscous layers (gel formation) and/or precipitation.The use of UF units equipped with stirrers rotating just above the membrane surface allows undesirable concentation polarization effects to be minimized. As high concentration factors are desirable when starting from very dilute aqueous samples, amphiphiles having low c.m.c. values (e.g., molecules possessing longer alkyl chains) are, in principle, the best candidates for these MEUF-based separations. Financial support from CNR (Progetto Finalizzato Chimica Fine) and from MURST (Rome) is gratefully acknowledged. References 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 Armstrong, D. W., Sep. Purif. Methods. 1985, 14, 213. Ordered Media in Chemical Separations, eds. Hinze, W. L.. and Armstrong, D. W.. ACS Symp. Ser.342, Amcrican Chemical Society. Washington, DC, 1987. Surfactants in ChemicallProcess Engineering. cds. Wasan, D. T., Ginn, M. E., and Shah, D. O., Marcel Dekker, New York, 1988. Pramauro, E., and Pelizzetti, E., TrAC, Trends Anal. Cizem., 1988, 7, 260. Surfuctant-Based Separation Processes, eds. Scamehorn, J. F., and Harwell, J . H., Marcel Dckker, New York, 1989. Dunn, R. O., Scamehorn, J . F., and Christian, S. D., Sep. Sci. Technol., 1985, 20, 257. Scamchorn, J . F., Ellington, R. T.. Christian, S. D., Penney, B. W., Dunn, R. O., and Bhat, S. N., AIChE Symp. Ser., 1986. 82, 48. Lane Gibbs, L., Scamchorn, J . F., and Christian, S. D.. Sasaki, K . J., Burnctt, S. L.. Christian. S. D., Tuckcr, E. E., and Scamehorn, J. F., Langmuir, 1989, 5, 363. Pramauro, E., Ann. C'him. (Rome), 1990, 80, 101. Clarke, E. A.. and Anliker, R., in The Handbook of Environ- mental Chemistry, cd. Hutzingcr, O., Springcr-Verlag, Berlin, Vantulder, P. J . M., Howard, C. C., and Riggin. R . M., HPLC Determination of Aromatic Amines in Body Fluids and Commer- cial Dyes, ACS Symy. Ser. 149, American Chemical Society, Washington, DC, 1981, pp. 413-427. Bjorkovist. B., J . Chromatogr., 1981, 204, 109. Lores. E. M., Bristol, D . W., and Moseman, R. F., J . Chromatogr. Sci., 1978, 16, 358. Narang, A. S . , Choudhury, D. R., and Richards, A., J. Chromatogr. Sci., 1982, 20, 235. Barni, E . , Savarino, P.. Carpignano, R., and Larovcrc, R., Dyes Pigments, 1985, 6, 83. Barni, E., Savarino, P., Marzona, M., and Piva, M., J . Heterocycl. Clzem., 1983, 20, 1517. Armstrong, D. W.. and Nome, F., Anal. Chem., 1981,53,1662. Bcrczin, 1. V., Martinek, K., and Yatsimirskii, A. K., Russ. Chem. Rev. (Engl. Trund.), 1973, 42, 787. Sepulveda, L.. J . Colloid Interface Sci., 1974, 46, 372. Bunton, C. A., and Sepulveda, L., J . Phys. Chem., 1979, 83, 680. Dorsey, J. G., De Echcgaray, M. T., and Landy, J. S . , Anal. Clzem., 1983, 55, 924. Pramauro, E.. Minero, C., Saini, G . , Graglia, R., and Pelizzetti, E . , Anal. Chirn. Acta, 1988, 212, 171. Graglia. R., Pramauro, E . , and Pelizzetti, E.. Ann. Chim. (Rome). 1984, 74, 41. de la Guardia, M., Peris-Cardells, E . , Sancenon, J . , Carrion, J . L., and Pramauro, E., Microchern. J., 1991, 44, 193. 1980, VOI. 3A, pp. 181-215. Puper 2/01285E Received March 10, 1992 Accepted August 28, I992