摘要:

ANALYST, JULY 1993, VOL. 118 913 Experimental Method to Correct Fluorescence Intensities for the Inner Filter Effect Nanda K. Subbarao and Robert C. MacDonald Department of Biochemistr)/, Molecular Biology and Cell Biolog y, North western University, Evanston, I1 60208, USA An empirical procedure for the correction of measured fluorescence intensity for the inner filter effect is described. The procedure is more reliable than those described previously as fluorophores with the fluorescence properties of interest are sequestered from the external aqueous phase. The inner filter effect can therefore be assessed in the absence of chemical reaction or Forster energy transfer between the fluorophore and chromophore. A correction curve for the inner filter effect is obtained by measuring the fluorescence intensity of the sequestered fluorophore after adding a chromophore to increase the absorbance of the solution.The procedure was tested with liposomes containing calcein and two commercial polymer bead preparations containing embedded fluorophores. The former have the advantage of a wider choice of fluorophores and the latter of stability and convenience. The inner filter effect was varied using small aliquots of potassium chromate or pyridinium chloride solution. It was shown that the magnitude of the inner filter effect must be experimentally determined for each instrument and whenever the instrumental configuration is altered if an accurate correction for the inner filter effect is to be obtained. The magnitude of the correction depends on the wavelength range and the pathlength but not on slit-width or sample turbidity for the instrument employed for most of the experiments reported here. This empirical procedure makes possible studies of protein fluorescence quenching'by agents previously unsuitable because of high absorbances.Keywords: Fluorescence; fluorescence quenching; liposome; Fluoresbrite; calcein The inner filter effect (IFE) is the attenuation of fluorescence intensity due to absorbance by the fluorophore solution at its excitation and/or emission wavelengths. This attenuation has to be corrected for if one is to obtain the correct fluorescence intensity of the solution. Attempts are generally made to circumvent the problem of the IFE by reducing concentra- tions, by employing the purest available reagents and/or by adjusting the excitation wavelength.Considerable absorbance frequently persists in spite of taking such steps and a correction for the IFE may still be required. Earlier reports of empirically determined correction factors for the IFE have involved fitting parameters to the deviation from linearity of a fluorescence intensity (F) versus concentra- tion graph.1 Such an approach is correct only when the deviation is entirely due to the IFE and not to self-association of the fluorophore molecules. A second approach has been to use a protein containing a tryptophan that is not in direct contact with chromophores in the solution.*J This approach has the disadvantage that verification of the native state of the protein is inconvenient and sometimes impossible.Second, Forster energy transfer between the tryptophans and the chromophore cannot be ruled out as the distance of the tryptophans from the protein cavities that have access to the chromophore cannot be determined with certainty. More- over, it is limited to a single fluorophore. A more reliable method for experimentally determining the IFE is therefore desirable. A suitable system for establishing an experimental correc- tion for the IFE is a fluorophore-chromophore pair in which the chromophore absorbs at the fluorescence emission wave- length of the fluorophore of interest and the two moieties are physically separated so as to eliminate the possibility of chemical or physical reaction leading to altered spectral properties, collisional or static quenching and Forster energy transfer between the fluorophore and chromophore.The curve relating fluorescence attenuation (fluorescence with chromophore/fluorescence without chromophore) to the absorbance at the excitation wavelength (Aex) constitutes an experimental correction curve for the IFE. Liposomes containing calcein, a fluorescent dye,4 have been examined in conjunction with chromate added to the external buffer as the chromophore as a system that satisfies the above conditions for experimental correction for the IFE. Lipo- somes are vesicles with walls composed of a bimolecular layer of lipids. Virtually any polar molecule can be incorporated in their internal aqueous compartment. They may be prepared by a number of different procedures and from a variety of different lipids.Liposomes of a phosphatidylcholine-choleste- rol mixture prepared by the sonication or extrusion methods were suitable for our purpose. These are approximately 1000 nm (extruded) or 350 nm (sonicated) in diameter. The bilayer wall is impermeable to most molecules of high water solubility and low hydrophobicity. In addition, its thickness of 40 8, almost always means that potential Forster energy transfer between molecules on either side can be disregarded. The liposomes are colourless and do not interfere significantly with fluorescence measurements. Fluoresbrite particles (Polysciences, Warrington, PA, USA) were also studied as an easily available, stable alternative to liposomes for the determination of IFE correc- tion curves.These particles are beads with fluorophores embedded in an outer shell of the polymer matrix and are available with several fluorophores having spectral properties covering a wide range of wavelengths. As the light pathway is a major factor in the measured IFE, the correction curves were examined on two different spectrofluorimeters with different optical configurations. The effect of altering the slit-width, pathlength and wavelength range in one of the instruments to determine which of these changes warrants recalibration of the correction curve for the IFE was also studied. Experimental Phospholipids were obtained from Avanti Polar Lipids (Alabaster, AL, USA). Fluoresbrite YG and Fluoresbrite 273 particles were 1 and 0.45 pm in diameter, respectively, and were obtained from Polysciences.Potassium iodide (4 mol I - l ) , potassium chromate (50 mmol I - I ) and caesium chloride (1 mol I - l ) stock standard solutions were buffered with 50 mmol 1 -I tris( h y drox y me t h yl)me t h yl amine (Tris)-914 ANALYST, JULY 2993, VOL. 118 HCl (pH 7.4). The potassium iodide solution also contained 33 mmol 1-* sodium thiosulfate to eliminate elemental iodine. Pyridinium chloride (2 moll-l) was prepared by adjusting the pH of a spectroscopic-grade pyridine (Sigma) solution with HCl . Both sonicated and extruded vesicles were composed of dioleoylphosphatidylcholine (DOPC)-cholesterol (molar ratio 4.3 : 1). A chloroform solution containing 5 mg of DOPC and 0.5 mg of cholesterol was dried in a stream of argon and placed under vacuum for 2 h.Calcein [0.8 ml, 3.3 mmol 1-l in 5 mmol I-' Tris-HC1-100 mmol 1-1 NaCl (pH 7.4)] was added, the solution was vortex mixed and subjected to ten freeze-thaw cycles between a dry-ice-ethanol bath and a 40 "C heating block. One portion of this multi-lamellar preparation was sonicated in a bath sonicator until essentially clear. The remainder was extruded 19 times through a 0.1 pm diameter pore Nucleopore filter supported in a hand-held extrusion apparatus.5 Both unilamellar vesicle preparations were chro- matographed on a Sephadex G-75 column equilibrated with 5 mmol I-' Tris-HC1-100 mmol 1-l NaCl (pH 7.4). The lipo- somes were collected in the void volume. Most fluorescence measurements were made on a Photon Technology International Alphascan spectrofluorimeter equipped with a xenon lamp.The spectrum was scanned over a range extending 5 nm on either side of the wavelength of interest (step widths 0.5 nm; integration time 0.5 s). Repeated scanning of the samples under these conditions caused no measurable bleaching. Normally the bandwidths were 8 nm for both excitation and emission, the slit-width being 0.2 cm. Some fluorescence measurements were also made on a Farrand spectrofluorimeter with a 25 nm bandwidth and slit- widths of 0.5 cm. Correction curves for the IFE were obtained at vesicle concentrations that gave a fluorescence intensity of (1-1.7) x 105 counts s-l. Vesicles were suspended in 5 mmol 1-1 Tris- HCI-100 mmol I-' NaCl (pH 7.4) in a 0.5 cm pathlength cuvette. The excitation wavelength (hex) was 340 nm and the emission was monitored at 520 nm (hem). The vesicle suspen- sion in the cuvette was titrated with 5 mmol 1-1 potassium chromate and mixed by inverting the cuvette several times.The values of F and A,, were measured in the same cuvette. The A,, of the solution without chromate (blank, 0.04-0.06) was subtracted from the absorbances in the presence of chromate. The ratio FdF, where Fo is the fluorescence intensity of the solution before addition of chromate, was plotted against the corrected A,, . A second-order regression curve was fitted to the data set and the resulting expression was taken to be the IFE correction. The correction factor for the IFE was obtained by substituting A,, in the correction expression obtained at the same wavelength, slit-width and cuvette size used to measure the fluorescence intensity. The raw fluorescence intensity was multiplied by the correction factor to obtain the corrected fluorescence intensity.Similar procedures were followed for determination of the IFE correction curves with the fluorescent particles suspended in water. Potassium chromate (50 mmol l-I) and pyridinium chloride (2 mmoll-l) were used to generate the IFE for Fluoresbrite YG (hex = 440 nm, h,, = 520 nm) and Fluores- brite 273 (hex = 280 nm, he, = 350 nm) particles, respectively. An F versus concentration graph for quinine sulfate was obtained by measuring the F value (hex = 346 nm, hem = 450 nm) of 0.05 mol I-' sulfuric acid in a 1 cm pathlength cuvette as aliquots of a 0.1 g ml-1 solution of quinine sulfate in 0.05 mol I-' sulfuric acid were sequentially added.The value of A346 was measured in the same cuvette after each addition. Results Addition of potassium chromate to a suspension of calcein- containing liposomes reduces the fluorescence intensity and results in the IFE correction curve shown in Fig. 1 (closed 2.0 : correction 1 .o 0 0.1 0.2 0.3 0.4 0.5 A,, Fig. 1 Experimental correction curves obtained with a, sonicated vesicles and 0, extruded vesicles. The ratio of initial fluorescence intensity at 520 nm of each vesicle preparation with entrapped calcein to the measured fluorescence intensity in the presence of increasing concentrations of potassium chromate is plotted against A340 of the same sample measured in the same cuvette. For details see text. The data points reflect single measurements made in the same cuvette; the curve represents a quadratic expression fitted to the data using the least-squares procedure.Correction curves were reproducible to within 3% at any Aex. The PTI instrument was used in the determination of these curves. The pathlength was 0.5 cm and bandwidth 8 nm (slit-width 0.2 cm). The Brand and Witholt form of the theoretical correction for the inner filter effect6 is given for comparison. A slit-width of 0.2cm was used to calculate the theoretical correction circles). Direct contact between chromate and calcein due to leaky liposomes was ruled out by the absence of fluorescence quenching when CoCI2, which forms a non-fluorescent complex with calcein, was added to the suspension.In addition, the fluorescence intensity of the vesicles remained constant for more than 3 h when they were incubated in the presence of chromate at 2 mmol 1-1 (the highest concent- ration used in these experiments), indicating that quenching did not occur to any detectable extent by the passage of chromate into the liposomes. The light scattering at 600 nm also remained constant during this interval, showing that the chromate does not cause aggregation of the vesicles. Fluores- cence attenuation due to Forster energy transfer between calcein and chromate was also ruled out as the absorbance of potassium chromate at the highest concentration used is not detectable in the range of the emission band of calcein. Hence the chromate-attenuated fluorescence intensity curve (Fig.1) can serve as the correction curve for the IFE, as possibilities of other mechanisms of fluorescence attenuation have been eliminated. The theoretical curve for the IFE derived for a spectrofluorimeter with perfectly collimated beams6 is also plotted in Fig. 1 for comparison. The use of extruded vesicles, which are about three times larger than sonicated vesicles, gave an IFE correction curve (Fig. 1, open circles) that was closely similar to that obtained with sonicated vesicles. The attenuations of the fluorescence of Fluoresbrite YG and calcein-containing liposomes due to increasing A,, were compared and found to be very similar (data not shown). Effects other than TFE that could attenuate fluorescence are excluded on the basis of two considerations.First, direct contact between the aqueous phase and the fluorophore in the particles did not occur, as addition of KI (negatively charged quencher) or CsCl (positively charged quencher) did not reduce their fluorescence.7 Second, the overlap or the emission spectrum of Fluoresbrite YG with the absorbance band of chromate occurs over less than 5% of the fluorescence and absorbance wavelength ranges, suggesting that Forster energy transfer can make only a minor contribution to fluorescence attenuation. The measured IFE could be influ- enced by the difference in the excitaton wavelength (hex =ANALYST, JULY 1993, VOL. 118 915 340nm for calcein and 440nm for Fluoresbrite YG). The similarity of the two correction curves indicates that this factor is negligible.For convenience, Fluoresbrite YG was used for the remainder of the study. The IFE correction curve for a Farrand spectrofluorimeter with a divergent excitation beam was different from that determined o n the PTI instr~ment that has a convergent excitation beam. Introduction of the optical assembly system in the Farrand instrument that reduces the width of the light beam to accommodate narrow cuvettes and which also produces a convergent excitation beam resulted in a correc- tion curve similar to that obtained for the PTI instrument (Table 1). Changing the pathlength from 0.5 to 1 cm decreases the correction for the IFE (Table 1). The correction curve is not altered when the bandwidth is increased from 8 to 20 nm (Table 1). An important application of the IFE correction procedure is in protein tryptophan fluorescence-quenching studies.These are typically conducted at h,, and A,,, of 280-290 and 330- 360 nm, respectively, which are considerably different from the wavelength ranges of calcein of Fluoresbrite YG. Although no significant difference in the IFE correction curves obtained with calcein (hex = 340nm) and with Fluoresbrite YG particles (hex = 440nm) was observed, it seemed possible that a larger change in the wavelength might alter the IFE significantly. A correction curve was therefore determined using Fluoresbrite 273 microparticles (excitation maximum at 270 nm, emission maximum at 345 nm). Pyri- dinium chloride was used to increase the A,, of the solution. Direct contact between the aqueous phase and fluorophore and Forster energy transfer were ruled out by the procedure described earlier for Fluoresbrite YG.Comparison of the IFE correction curve obtained with Fluoresbrite 273 with that obtained with liposomes containing calcein reveals a modest effect of wavelength on the IFE (Table 1). Protein solutions often scatter light to a significant extent, particularly when they are membrane associated. A suitable system to examine the effect of altered light scattering was provided by Fluoresbrite YG particles, as they produce significant light scatter. It was found that the IFE correction curve was not significantly different for these microparticle suspensions at two different concentrations, even though the higher concentration produced a visibly turbid solution (Table 1).The IFE correction curve was applied to the F versus concentration graph for quinine sulfate (Fig. 2) to examine whether it would eliminate curvature, a requirement of any Table 1 Effects of different optical configurations on the TFE correction curve determined using the PTI and Farrand spectroflu- orimeters. The standard conditions for the PTI spectrofluorimeter are described in the caption to Fig. 1. Re-determination of thc correction curves resulted in less than a 3% variation in the corrcction factor at any Acx. Experimental details are given in the text IFE correction System IFE correction at A = 0.3 PTI, standardconditions 1.01 + 1.48Ae, + 1.64Acx2 1.607 Farrand, no optical accessory 1.01 + 2A,, + 0.82Ae,2 1.68 Farrand with optical accessory 1.01 + 1.48Ae, + 1.64Ae,* PTI, slit-width incrcascd to PTI, pathlength decreased to 0.5 cm 1 + 1.l5Ae, + 0.44Aex2 1.38 PTI, in 0.5 cm cuvette, turbid owing to high Fluoresbrite concentration 1 + 1 .lSAex + 0.44Acx2 0.5 cm cuvette 0.99 + 1.43Aex + 1.76Aex2 1.58 0.5 cm 1.01 + 1.48Aex + 1.64Aex2 PTI, hex = 270 nm in valid IFE correction curve. The correction curve used in this instance was for Fluoresbrite YG with A,, = 440 nm, whereas he, for quinine sulfate is 346 nm; however, this difference in wavelength has no significant effect on the correction curve. The fluorescence intensity curve of quinine sulfate was very close to linear up to a concentration of 0.2 mg ml-1 of quinine (A346 = 2). Beyond this point, the corrected curve showed a downward curvature, possibly indicating self-quenching of the fluorophore.Discussion It has been demonstrated that liposomes with entrapped fluorophores and fluorescent polymer particles are e€fective for determining correction curves for the IFE, even for relatively high absorbances. In these systems, any possibility of physical contact or Forster energy transfer between the fluorophore of interest and the chromophore that causes the IFE has been ruled out. The IFE correction curves determined with the two systems were very similar, suggesting that they can be used interchangeably. As would be expected, spectrofluorimeters with dissimilar optical paths gave different IFE correction curves.8 Further- more, use of the optical assembly system in the Farrand instrument, which changes the beam geometry, altered the TFE correction.Determining the IFE correction curve for each instrument and for each optical configuration is impor- tant especially in fluorescence quenching studies wherein small changes in the correction can drastically change the Stern-Volmer plot and the quenching constants derived from it. Factors such as slit-width, pathlength and beam configura- tion, which usually vary from instrument to instrument, can be expected to influence the IFE correction curve.8 However, our observation of a reduction in the IFE on increasing the pathlength was surprising in the light of the derivation of the theoretical expression for the IFE.6 The identical IFE curves that were obtained from a visibly turbid suspension of Fluoresbrite YG and a virtually clear suspension are explained on the basis that reduced light transmission due to light scattering and that due to true absorbances are equivalent with respect to their effect on the TFE.This situation may not be true on all spectrofluorimeters and may need to be verified on each instrument. In summary, the results presented here reveal that, to obtain accurate corrections for the IFE, it is necessary to 0 0.5 1 .o 1.5 2.0 Concentration of quinine sulfate/mg mi- Fig. 2 Plot of fluorescence intensity versus concentration of quinine sulfate in 0.05 mol I-' sulfuric acid (0) corrected for the IFE by the IFE correction curve obtained using Fluoresbrite YG microparticles (@). Fluorescencc intcnsities of quinine sulfate and thc microparticles and also their absorbances were measured in the same cuvette, with the same slit-width916 ANALYST, JULY 1993, VOL.118 determine experimental correction curves for each instrument and probably also when the configuration of a given instru- ment is changed. Liposomes can be employed fo,r this purpose over a wide wavelength range as they can entrap a wide variety of fluorophores. Sonicated liposomes are sufficient in those instances where Forster energy tFansfer is negligible owing to minimal overlap of the fluorophore emission band with the chromophore absorption band. When overlap is extensive, it may be necessary to use the larger extruded liposomes. When Fluoresbrite particles (or other suitable particles) of the appropriate wavelengths are available with a chromophore that does not absorb at the emission wavelength, these particles will be a more convenient alternative to liposomes because of their relatively long shelf-life. We thank Drs. R. I. MacDonald and S.-s. Feng for help and encouragement. This study was supported by NIH grant DK36634 and GM38244 and a Junior Research Fellowship from the American Heart Association of Metropolitan Chicago to N. K. S. References 1 Birdsall, B . , King, R. W., Wheeler, M. R., Lewis, C. A., Goodc, S. R., Dunlap, R. B., and Roberts, G. C. K., Anal. Biochem., 1983, 132, 353. 2 Wiechelman, K. J., Am. Lab., 1986, 18 (2), 49. 3 Kahana, E., Pinder, J. C., Smith, K. S., and Gratzer, W. B., Biochem. J., 1992,282, 75. 4 KendaIl, D., and MacDonald, R. C., Anal. Biochem., 1983, 134, 26. 5 Macdonald, R. C . , Macdonald, R. I., Menco, B. Ph. M., Takeshita, K., Subbarao, N. K., and Hu, L.-R., Biochim. Biophys. Acta, 1990, 1061, 297. 6 Brand, L., and Witholt, B., Methods Enzymol., 1967, 11,776. 7 Eftink, M. R., in Topics in Fluorescence Spectroscopy, ed. Lacowicz, J. R., Plenum, New York, 1991, vol. 2, pp. 53-120. 8 Standards in Fluorescence Spectroscopy, ed. Miller, J. N., Chapman and Hall, London, 1981, ch. 4. Paper 21051 83 D Received September 28, 1992 Accepted January 11, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800913

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 917 Analysis of Mixtures of Polycyclic Aromatic Hydrocarbons in Sea-water by Synchronous Fluorescence Spectrometry in Organized Media J. J. Santana Rodriguez, J. Hernandez Garcia, M. M. Bernal Suarez and A. Bermejo Martin-Lazaro Department of Chemistry, Faculty of Marine Sciences, University of Las Palmas de Gran Canaria, 35017 Las Palmas, Spain The spectrofluorimetric characteristics of chrysene, perylene, dibenz[a,c]anthracene and coronene were studied in aqueous medium and in the presence of cationic, anionic and non-ionic surfactants. The great enhancement of the fluorescence of these polycyclic aromatic hydrocarbons (PAHs) in hexadecyltrimethyl- ammonium bromide and synchronous fluorescence spectrometry were used for the establishment of two methods for the analysis of binary and ternary mixtures of these compounds. The former allows the simultaneous determination of dibenz[a,c]anthracene and coronene using Ah = 89 nm, with detection limits of 0.20 and 0.22 ng ml-l, respectively.In the latter, three different Ah values are used for the simultaneous determination of chrysene, perylene and coronene, viz., 41,3 and 140 nm, respectively. Limits of detection of 0.17 ng ml-I for chrysene, 0.13 ng ml-I for perylene and 0.14 ng ml-I for coronene were obtained. Good recoveries were obtained when both methods were applied to the analysis of binary and ternary mixtures of these PAHs added to sea-water. Keywords: Synchronous fluorescence spectrometry; polycyclic aromatic hydrocarbon; micellar media; chrysene, perylene, dibenz[a,c]anthracene and coronene Many potentially hazardous chemicals are commonly released into the environment through industrial and domestic pro- cesses.Among them, polycyclic aromatic hydrocarbons (PAHs) are widespread environmental pollutants, owing to their presence as natural and synthetic compounds in all areas of human activity (industry, agriculture, domestic use, etc.).1 These PAHs have been shown to display a wide range of cancer-inducing activity .2J The need to identify and deter- mine these potentially toxic and/or carcinogenic compounds has led to the establishment of methods that allow for the simultaneous determination of as many PAHs as possible. For several years, the use of organized media (aqueous and reverse micelles , bilayers , microemulsions , vesicles and lipo- somes) has been common practice in many areas of analytical chemistry for the modification of the reactivity between analytes and reagents and to improve the analytical proce- dures.4-8 Specifically.micellar media have been used to improve the sensitivity and selectivity of many luminescence determinations .9-12 Synchronous fluorescence spectrometry, a method first used by Lloyd13 to identify a number of PAHs, involves simulta- neous scanning of the excitation and emission mono- chromators, which are synchronized in such a fashion that a well-defined relationship is maintained between the two wavelengths: either a constant wavelength (Ah)13 or a constant energy (Av) 14 difference. The frequency range over which a given component of the fluorescent mixture can emit is made narrower to an extent dependent on the magnitude of Ah or Av.This technique has been used for the analysis of crude oils,15-19 pharmaceuticals20 and PAHs.21-24 This paper describes the use of synchronous fluorescence spectrometry and micellar media for the analysis of binary mixtures of dibenz[a, clanthracene and coronene and ternary mixtures of chrysene, perylene and coronene. The results show that the proposed methods can be used for the simple and sensitive simultaneous determination of these hydrocar- bons in sea-water. Experimental Reagents Stock standard solutions of 0.023 mg ml-l chrysene (Sigma, 95% purity), 0.025 mg ml-1 perylene (Sigma, 100% purity), 0.028 mg m1-I dibenz[a,c]anthracene (Aldrich, 97% purity) and 0.015 mg ml-l coronene (Aldrich, 97% purity) were prepared in ethanol.Working standard solutions were pre- pared by appropriate dilution with ethanol. The surfactants polyoxyethylene 10-oleyl ether (Brij-96), isooctylphenoxypolyethoxyethanol (Triton X-loo), hexa- decyltrimethylammonium bromide (HDTAB), benzyl- dimethylhexadecylammonium bromide (BDHAB), sodium lauryl sulfate (NaLS) and sodium dodecylbenzenesulfonate (NaDDBS) were obtained from Aldrich. Apparatus All fluorescence measurements were made with a Perkin- Elmer MPF-44A recording spectrofluorime ter equipped with a 150 W Osram XBO xenon arc lamp, a Scort EDM 2116 digital readout, a Selecta Frigitherm S-382 ultrathermostat- cryostat and 1 cm quartz cells. The emission intensity measuring system of the spectrofluorimeter was calibrated daily by using the Perkin-Elmer set of fluorescent polymer blocks. General Procedure for the Analysis of Mixtures To an aliquot containing 2.7 ng-13.9 pg of dibenz[a,c]an- thracene and 3.2 ng-8.3 pg of coronene for the binary mixtures and 2.5 ng-11.4 pg of chrysene, 2.5 ng-12.6 pg of perylene and 3.2 ng-8.3 pg of coronene for the ternary mixtures, and in the allowed ratios in a 25 ml calibrated flask, 4 ml of 2.5 x mol I-' HDTAB solution are added, followed by the necessary volume of ethanol to obtain a content of 2% v/v of organic solvent in the final solutions and de-ionized water up to the mark.For the simultaneous determination of dibenz[a, clan- thracene and coronene in a mixture, the synchronous fluores- cence spectra were recorded by scanning both monochroma- tors simultaneously (scan rate 120 nm min-l) at an 89 nm constant difference. The measurements of peak height were carried out at h", = 289 nm for dibenz[a,c]anthracene and 13, = 342 nm for coronene.For the ternary mixture of chrysene, perylene and coronene, the fluorescence intensities were determined by measuring peak heights to a synchronous value for each PAH: for chrysene hO, = 322 nm (Ah = 41 nm), for perylene LO, = 439 nm (Ah = 3 nm) and for coronene ho, = 304918 ANALYST, JULY 1993, VQL. 118 nm (Ah = 140 nm). Calibration graphs were obtained using standard solutions prepared under the same conditions. Analysis of Mixtures in Sea-water Prior to analysis, the sea-water samples were passed succes- sively through filters of different# porosity (50, 30 and 5 pm) and treated with ultraviolet radiation to avoid possible interferences or marine micro-organisms.After verifying that they showed no fluorescence, they were spiked with suitable amounts of the hydrocarbons present in the mixture and analysed according to the above method. Results and Discussion Conventional Spectral Characteristics in Different Media The fluorescence characteristics of chrysene, perylene , dibenz[a,c]anthracene and coronene were studied in 2% v/v ethanol-water medium (which is subsequently referred to as aqueous solution) and in the presence of different cationic, anionic and non-ionic surfactants (Table 1). The micellar solutions contain sufficient surfactant to ensure a higher concentration than the critical micellization concentration. The hydrocarbons were initially dissolved in ethanol to avoid modification of the spectra by any solubilizing effect possibly produced by micellar media.Aqueous medium Chrysene presents an excitation spectrum with two peaks of maximum intensity at 278 and 318 nm and another two peaks of lower intensity at 303 and 332 nm. These four peaks form a wide band with a peak half-width of 84 nm. Another peak of average intensity appears at 366 nm. The emission spectrum shows an intense maximum at 446 nm with a peak half-width of 24 nm, another peak of high intensity at 412 nm and a third peak of average intensity at 476 nm. Perylene shows an excitation spectrum with two clearly defined peaks at 410 and 439 nm, with peak half-widths of 25 and 18 nm, respectively.The emission spectrum shows a highly intense maximum at 442 nm with a peak half-width of 24 nm and another peak of average intensity at 469 nm. Dibenz[a, clanthracene shows an excitation spectrum with a broad peak at 293 nm with a peak half-width of 64 nm and another peak of average intensity at 348 nm. The emission spectrum shows a maximum at 388 nm and another peak at 407 nm. Coronene presents an excitation spectrum with a band at 300 nm with a peak half-width of 65 nm and a peak of average intensity at 358 nm. The emission spectrum shows a band at 500 nm with a peak half-width of 70 nm. Micellar media Except for perylene, it was observed that the surfactants studied modify the form of the spectra, making them narrower than those produced in the aqueous medium.For chrysene and dibenz[a, clanthracene, the surfactants used produced hypsochromic shifts of the excitation and emission wavelengths compared with those observed in the aqueous medium. For coronene the same shift was observed except for the excitation maximum, with a small shift towards longer wavelengths. For perylene, none of the surfactants produced shifts in the excitation and emission wavelengths compared with those observed in the aqueous medium. In Table 1 it can be seen that HDTAB, BDHAB, Brij-96 and Triton X-100 produce substantial enhancements of the fluorescence intensity emitted by the four hydrocarbons compared with when the aqueous medium is used. The HDTAB, which produces the maximum enhancement, was used for the study of these hydrocarbons.The excitation spectrum (hem = 381 nm) of chrysene in HDTAB shows a clearly defined peak at 272 nm with a peak half-width of 17 nm. The emission spectrum (Aex = 272 nm) presents two peaks of similar intensity at 381 and 402 nm within a band with a peak half-width of 38 nm. For dibenz[a,c]anthracene in HDTAB solutions, the excita- tion spectrum (hem = 378 nm) presents a clearly defined peak at 289 nm with a half-width of 20 nm. The emission spectrum (hem = 289 nm) presents a band with a peak half-width of 38 nm with the most intense peak at 378 nm and another peak at 398 nm of lower intensity. In the presence of HDTAB, the excitation spectrum of coronene presents a maximum at 304 nm with a peak of lower intensity at 341 nm (hem = 444 nm).The emission spectrum (hex = 304 nm) shows a highly intense peak at 444 nm, another of average intensity at 428 nm and a third peak of low intensity Table 1 Spectrofluorimetric characteristics of chrysene, perylene, dibenz[a,c]anthracene and coronene in different media. Values in italics are maximum excitation and emission wavelengths Chrysene* Medium Aqueous Brij-96 Triton X-100 HDTAB BDHAB NaLS NaDDBS h,,/nmT 278,318, 332,366 272,310, 32 1 310,322 272,310, 32 1 272,310, 323 272,307, 320 ** - h,,/nmT 141 446,412. 7.0 360,381, 80.0 360,381, 83.0 362,381. 85.2 360,381? 75.2 362,381, 25.5 476 400 400 402 402 402 - - Perylenet h,,/nmT h,,/nmT 410,439 442,469 410,439 442,469 41 1,440 444,470 412,439 442,469 412,439 442,469 414,439 441,471 414,439 442,471 ~- Zfll 0 89 70 93 87 15 35 Dibenz[a, c]anthraceneT h,,/nmT h,,/nmT 141 293,348 388,407 6.0 288 376,396 90.0 - - ** - 289 378,398 85.6 289 378,398 78.4 288 378,396 20.1 288 376.396 80.4 Coronenes h,,/nmr h,,/nmv 300,358 480,500 304,341 428,444, 304,341 428,444, 304,341 428,444, 304,341 428,444, 304,341 428,444, 304,341 428,444.472 472 472 472 472 472 14 3.3 75.3 85 .o 96.5 83.0 5.2 15.8 * Slits 7 nm. t Slits 3.5 nm. * Slits 7 nm. 5 Slits 6 nm. fl Wavelengths are all 2 2 nm. 11 Relative fluorescence intensity (%). ** Great interference of surfactant is observed.ANALYST, JULY 1993, VOL. 118 919 100 80 A 8 Y > m C a, 4- .- c.l .- c 60 a, C a, m ?? - 2 40 Y- a, m a, U .- .I- - 20 0 B A C -1 J 220 300 380 460 240 280 320 hlnm 220 300 380 460270 310 350 iilnrn Fig.1 Excitation (A), emission (B) and synchronous (C) (Ah = 110 nm) spectra of chrysene in micellar medium. Cchrysene = 2.0 x 10V mol I-' and cHDTA = 4.0 x moll-' Fig. 3 Excitation (A), emission (B) and synchronous (C) (AL = 89 nm) spectra of dibenz[a,c]anthracene in micellar medium. C&benz. [lu,,.]anlhracene = 2.0 x 10-6 mol 1-1 and cHDTAB = 4.0 x 0-3 moll-' F -1 100 80 8 - > m C a, 4- .- .- 60 a, C a, m 2 - 2 40 Y- a, m a, U .- 4- - 20 0 B 1 C 'r 1: 340 380 420 460 500 540420 460 hln m 260 340 420 500 270 310 350 l./nrn Fig. 2 nm) spectra of perylene in micellar medium. cperylene = 2.0 x moll-' and cHDTAB = 4.0 x Excitation (A), emission (B) and synchronous (C) (Ah = 3 mol 1-1 Fig. 4 Excitation (A), emission (B) and synchronous (C) (Ah = 140 nm) spectra of coronene in micellar medium.ccoronene = 1.0 x lop6 moll-' and cHDTAB = 4.0 x mol 1-l at 472 nm. The most intense peaks at 304 and 444 nm show peak half-widths of 15 and 18 nm, respectively. When HDTAB is used as the surfactant the spectra for all the hydrocarbons are similar to those obtained when they are dissolved in cyclohexane.25 This seems to confirm that the inner core of spherical rnicelles is hydrocarbon-like26 and suggests that in the presence of HDTAB micelles, chrysene, perylene , dibenz[a, clanthracene and coronene are mainly solubilized in the hydrophobic region of these micelles. Such a microenvironment allows for stabilization of the excited singlet states similar to that obtained in cyclohexane. Different studies carried out for the simultaneous determi- nation of these PAHs, using their conventional spectral characteristics, show great mutual interferences, which would lead to large errors in the analytical method.Synchronous Spectral Characteristics In aqueous medium, the synchronous spectra of the hydrocar- bons reveal wide bands of different intensities. However, in the presence of HDTAB, synchronous spectra with clearly defined peaks and small peak half-widths are obtained with the choice of a suitable Ak. For solutions of chrysene in HDTAB, the difference between the emission maximum (381 nm) and the excitation maximum (272 nm) is 109 nm. By using this, the synchronous spectrum shows a clearly defined peak of greatest intensity at A", = 272 nm (Fig. 1). When A1 = 5-30 nm the synchronous spectrum reveals bands of low intensity and large width.For A1 > 30 nm a peak appears at 330 nm which reaches its maximum intensity at A1 = 41 nm. From then on, the peak920 ANALYST, JULY 1993, VOL. 118 J I I 240 320 400 h/n m Fig. 5 Synchronous fluorescence spectrum of a binary mixthre of dibenz[a,c]anthracene (A) and coronene B in micellar medium. Ah - 1 ) 10-7 moll-1 an4 CHDTAB = 4.0 x 10-3 moll-1 - 89 nm, Cdibenz n,c]anthracene = 1 .o lo- mol I-'? Ccoronene = 5.0 decreases in intensity. When A1 > 50 nm, the peak splits into two peaks. One of them, which appears at 324 nm, increases each time Al is increased and reaches maximum intensity at A1 = 60 nm. For Al > 100 nm a peak appears that shows maximum intensity at Al = 109 nm. For perylene in HDTAB, the difference between the most intense excitation (439 nm) and emission (442 nm) peaks is 3 nm, which corresponds to the most frequent value of the Stokes shift.By using this value of AI, the synchronous spectrum (Fig. 2) reveals only one intense and well defined peak at LO, = 439 nm. If Al is increased, this peak decreases in intensity up to A1 = 20 nm. From then on, it splits into two clearly defined new peaks, the intensities of which decrease each time A1 is increased, and when A1 > 45 nm bands of low intensity and large peak half-width are produced. For dibenz[a, clanthracene in HDTAB solution, the differ- ence between the emission and excitation maxima (378 and 289 nm) is 89 nm. By using this value, the synchronous spectrum shows a peak of greatest intensity at LO, = 289 nm (Fig.3). This peak appears when A1 > 30 nm, and the intensity increases up to Al = 89 nm. When A1 > 89 nm, the peak intensity decreases and later increases up to Al = 109 nm, which is the difference between an emission peak at 398 nm and the maximum excitation peak at 289 nm. For values of A1 > 110 nm the peak decreases in intensity and it splits into two peaks when A1 > 120 nm. For coronene in the presence of HDTAB, the difference between the emission maximum (444 nm) and the excitation maximum (304 nm) is 140 nm. By using this, the synchronous spectrum reveals a peak of greatest intensity at hO, = 304 nm (Fig. 4). When A1 > 80 nm the synchronous spectrum reveals bands of low intensity and great width. For A1 from 85 nm upwards, a clearly defined peak appears at 342 nm that reaches its maximum intensity at Al = 140 nm.For values of Table 2 Simultaneous determination of dibenz[a, clanthracene (D) and coronene (C) in synthetic binary mixtures Found*/ng ml-l Ratio D added C added/ D:C ng ml-l ng ml-l D C 1 : l 3.0 3.0 3.2 3.1 1 : 3 3.0 9.0 2.9 9.9 3: 1 9.0 3.0 9.4 2.7 1 : s 3.0 15.0 3.3 15.5 5 : 1 15.0 3.0 15.8 2.5 1 : 8 3.0 24.0 3.2 24.6 1 : l O 3.0 30.0 3.3 30.8 1:30 3.0 90.0 3.1 87.3 * Mean of three determinations. Table 3 Simultaneous determination of chrysene (Ch), perylene (P) and coronene (C) in synthetic ternary mixtures Ch Ratio of added/ Ch:P:C ngml-I 1 : l : l 3.0 5:1:5 15.0 1 : 5 : 5 3.0 5 : 5 : 1 15.0 3:7:1 9.0 1 : 8 : 5 3.0 1:9:11 3.0 8 : 1 : 9 24.0 1:30:5 3.0 26:1 : S 78.0 1:30:33 3.0 P added/ ng ml-I 3.0 3.0 15.0 15.0 21 .o 24.0 27.0 3.0 90.0 3.0 90.0 C added - ng ml-l 3.0 15.0 15 .0 3.0 3.0 15.0 33 .0 27.0 15.0 15.0 99.0 * Mean of three determinations.Found*/ng ml-1 Ch P C 3.2 3.2 3.3 14.8 3.2 15.9 3.1 15.9 15.9 14.2 15.8 3.4 9.7 21.8 3.4 3.2 25.1 16.4 2.9 29.0 34.1 25.9 3.2 28.2 3.2 93.0 15.6 81.4 2.9 17.0 2.8 103.9 93.6 AI .> 140 nm the peak decreases in intensity. From A1 = 160 nm the peak splits into two new peaks. An exhaustive evaluation of the characteristics of the sj7n Zhronous spectra reveals that simultaneous determinations of dibenz[a, clanthracene and coronene in binary mixtures and chry sene, perylene and coronene in ternary mixtures are possible. Ana1ytic.d Considerations In order to develop a method for the simultaneous determi- nation of th: r,omponents of the binary mixture of dibenz[a,- clanthracene and coronene, a value of Al = 89 nm is the most appropriate for the simple and rapid determination of both hydrocarbons (Fig. 5).However, it is only possible to resolve simultaneously the ternary mixture of chrysene, perylene and coronene if three different A1 values are used (chrysene 41 nm, perylene 3 nm and coronene 140 nm). This allows for the maximum sensitivity and selectivity of the method. In the presence of HDTAB, linear calibration graphs for each PAH are obtained by plotting the peak height measured on the synchronous spectra against the hydrocarbon concen- tration (0.1-500 ng ml-l for chrysene, perylene and dibenz- [a, clanthracene and 0.1-330 ng ml- for coronene).Pearson's correlation coefficients for all the peak-height calibration graphs are higher than 0.998. In the binary mixture, the detection limits27 are 0.20 and 0.22 ng ml-I for dibenz[a,c]anthracene and coronene, respec- tively. When the method developed was applied to three series of 11 samples containing 3.0, 30.0 and 300.0 ng mI-l of dibenz[a,c]anthracene and 3.0, 30.0 and 200.0 ng ml-1 of coronene, the relative standard deviations were 1.8, 1.4 and 1.1% for dibenz[a,c]anthracene and 2.0, 2.3 and 4.0% for coronene , respectively. For the ternary mixture, the detection limits were 0.17 ng m1-l for chrysene, 0.13 ng ml-I for perylene and 0.14ANALYST, JULY 1993, VOL. 118 92 1 Table 4 Simultaneous analysis of binary mixtures of dibenz[a, clan- thracene (D) and coronene (C) in sea-water from different areas of the Canary Islands Found*/ng ml- Sea-wat'er 1 Sea-water 2 Sea-water 3 Ratio D added C added/ D : C ngml-' ngml-l D C D C D C 1 : l 3.0 3.0 3.2 2.6 3.1 2.6 3.3 2.6 1 : 9 3.0 27.0 2.8 26.7 3.1 26.9 3.3 27.2 1:15 3.0 45.0 2.9 45.3 2.9 45.3 3.2 45.0 3: 1 9.0 3.0 9.4 2.2 9.2 2.3 9.9 2.4 * Mean of three determinations.Table 5 Simultaneous analysis of ternary mixtures of chrysene (Ch), perylene (P) and coronene (C) in sea-water Ch P C Found*/ng ml-I Ratio added added/ added/ Ch : P : C ng ml-1 ng ml-l ng ml-I Ch P C 1 : 1 : 1 3.0 3.0 3.0 2.6 3.1 3.1 5:1:5 15.0 3.0 15.0 15.0 3.3 14.9 1:5:5 3.0 15.0 15.0 2.7 15.5 15.0 5:5:1 15.0 15.0 3.0 15.0 15.5 3.2 3 : 7 : 1 9.0 21.0 3.0 9.5 21.4 3.3 1 : 9 : 4 3.0 27.0 12.0 3.1 28.2 12.4 9 : l : l l 27.0 3.0 33.0 27.4 3.0 29.4 1 : 9 : 9 3.0 27.0 27.0 3.2 28.3 26.2 1:33:22 3.0 99.0 66.0 2.7 90.0 65.2 * Mean of three determinations. ng ml-I for coronene.When the method was applied to three series of 11 samples containing 3.0,30.0 and 300.0 ng ml-l for chrysene and perylene and 3.0, 30.0 and 200.0 ng ml-1 for coronene, the relative standard deviations were 2.8, 1.7 and 2.4% for chrysene, 1.9,1.7 and 1.6% for perylene and 3.3,2.0 and 1.4% for coronene, respectively. Analysis of binary and ternary mixtures of PAHs The developed procedures do not only result in very sensitive methods for the determination of dibenz[a, clanthracene and coronene in binary mixtures and chrysene, perylene and coronene in ternary mixtures, but also allow the analysis of these PAH mixtures in different matrices.To study the selectivity of the method, simultaneous analyses of samples of synthetic mixtures were carried out. Tables 2 and 3 show the excellent results obtained for the two mixtures studied when different ratios of the hydrocarbons were used. The proposed methods were applied to the simultaneous determination of the hydrocarbons in sea-water samples from different areas of the Canary Islands, previously spiked with suitable amounts of the hydrocarbons. The results given in Tables 4 and 5 for the binary and ternary mixtures, respecti- vely, indicate good recoveries. This work was supported by funds provided by the Gobierno Autonomo de Canarias, Spain (Research Project No. 20/3 1.07.89). 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 26 27 References Bjorseth, A., and Ramdahl, T., Handbook of Polycyclic Aromatic Hydrocarbons, Marcel Dekker, New York, 1985, vol. 2.Lowe, J. P., and Silverman, B. D., Acc. Chem. Res., 1984, 17, 332. Grimmer, G., Toxicol. Environ. Chem., 1985, 10, 171. Hinze, W. L., Singh, H. N., Baban, Y., and Harvey, N. G., TrAC, Trends Anal. Chem., 1984, 3, 193. Pelizzetti, E., and Pramauro, E . , Anal. Chim. Acta, 1985, 169, 1. Armstrong, D. W., Sep. Purif. Methods, 1985, 14,212. Hinze, W. L., in Ordered Media in Chemical Separations (ACS Symposium Series, vol. 342), ed. Hinze, W. L., and Armstrong, D. W., American Chemical Society, Washington, DC, 1987, Mcintire, G. L., Crit. Rev. Anal. Chem., 1990, 21, 257. Howard, J. H., and Fazio, T., J . Assoc. Off. Anal. Chem., 1980, 63, 1077. Kalyanasundaram, K., and Thomas, J. K., J . Am. Chem. SOC., 1977, 99, 2039. Femia, R. A., and Cline Love, L. J., Spectrochim. Acta, Part B , 1986, 169, 1. Santana, J . J., Sosa, Z . , Afonso, A., and Gonzalez, V., Fresenius'J. Anal. Chem., 1990, 337, 96. Lloyd, J. B. F., Nature (London), 1971, 231, 64. Inman, E. L., Jr., and Winefordner, J . D., Anal. Chim. Acta, 1982, 141, 241. Vo-Dinh, T., and Martinez, P. R., Anal. Chim. Acta, 1981,125, 13. Eastwood, D., Fortier, S. H., and Hendrick, M. S . , Am. Lab., 1978, 10, 45. John, P., and Soutar, I . , Anal. Chem., 1976, 48, 520. Lloyd, J. B. F., Analyst, 1980, 105, 97. Wakeham, S . G., Environ. Sci. Technol., 1977, 11,272. Andre, J . C., Bandot, Ph., and Niclaus, M., Clin. Chim. Acta, 1977, 76, 55. Vo-Dinh, T., Gammage, R. B., Hawthorne, A. R., and Thorngate, J . H., Environ. Sci. Technol., 1978, 12, 1297. Kerkhoff, M. J., Lee, T. M., Allen, E. R., Lundgren, D. A., and Winefordner, J. D., Environ. Sci. Technol., 1985,19,695. Files, L. A., Jones, B. T., Hanamura, S . , and Winefordner, J. D., Anal. Chem., 1986,58, 1440. Santana Rodriguez, J . J., Sosa Ferrera, Z . , Afonso Perera, A., and Gonzalez Diaz, V., Anal. Chim. Acta, 1991, 225, 107. Lloyd, J . B. F., and Evett, I . W., Anal. Chem., 1977,49, 1710. Fendler, J. H., and Fendler, E . J., Catalysis in Micellar and Macromolecular Systems, Academic Press, New York, 1975. Long, G. L., and Winefordner, J . D., Anal. Chem., 1983, 55, 712A. pp. 2-82. Paper 210561 1 I Received October 21, 1992 Accepted January 14, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800917

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 923 Automated Kinetic-Spectrofluorimetric Method for the Determination of Morphine in Urine* Juana Cepas, Manuel Silva and Dolores Perez-Benditot Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, E- 14004 Cordoba, Spain The continuous addition of reagent technique was used with spectrofluorimetric detection for the kinetic determination of morphine in urine samples. The method thus developed is based on the oxidative dimerization of morphine to fluorescent pseudomorphine by potassium hexacyanoferrate(iit) in a basic medium. The optimum pH, oxidant concentration and instrumental variables were determined. The method permits the sensitive determination of morphine over a wide concentration range (15-925 ng ml-l) with high precision (relative standard deviation = 2%), selectivity and sample throughput (48 samples h-l).A novel sorption-desorption procedure was used to isolate morphine from whole urine, which provided recoveries of between 85 and 100%. The results showed the usefulness of the proposed method for controlling a wide variety of medicinal problems (e.g., therapeutic, overdose and doping analysis). Keywords: Kinetic-spectro fluorimetric determination; continuous addition of reagent technique; morphine; urine sample Opioid analgesics comprise a large group of substances that control pain by depressing the central nervous system. Of these, morphine, a narcotic analgesic, is the most important alkaloid of opium. This substance is widely used for treating severe pain in cancer patients and for relieving pain following surgical procedures.However, administration of this drug must be strictly controlled because excessive doses may have major side effects, whereas inadequate doses fail to alleviate pain. Hence, monitoring morphine levels in biological fluids (e.g., plasma and urine) is of great interest.' The determina- tion of morphine in these samples is also of interest in order to confirm overdosing in addicts1 or doping in sporting events2 (morphine can be present in urine as a metabolite of codeine). Morphine has been determined by using a wide variety of analytical techniques, particularly chromatographic tech- niques. However, analyses for morphine in biological fluids are still of great interest judging by the large number of papers published on this subject3 (e.g., about 35 references to the chromatographic determination of morphine appeared in 1990-1991).Specifically, liquid chromatography with electro- chemical detection is one of the most frequent choices in this respect because of the high sensitivity acheived due to the ease with which morphine can be oxidized.4-10 However, most of these methods pose problems arising from the composition of the mobile phase, because a high pH must be used in order to be able to apply low potentials in the detector. The combination of gas chromatography-mass spectrometry is currently another major choice for the determination of this drug"-16 as it provides high sensitivity and selectivity; however, it requires expensive equipment and skilled opera- tors. Immunoassay techniques [Abbott TDx,17-19 enzyme mediated immunoassay technique (EMIT) ,20-22 radioimmu- noassay23-2s and enzyme linked immunosorbent assay (ELISA)2"27] are also often used for the determination of morphine as they provide the required sensitivity (ng ml-I- pg ml-l level); however, they can only be applied when the drug has been identified, and positive results must be confirmed by alternative assays owing to potential cross- reaction by metabolites. Many of the above-mentioned methods are highly sensitive and selective, and sufficiently rapid for most purposes; however, fairly high concentrations of morphine in specific * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992.+ To whom correspondence should be addressed. samples (e.g., urine) could be determined by using a simpler, less time-consuming method of choice provided it were sufficiently selective. For this purpose, a kinetic method was developed, based on the classical oxidative dimerization of morphine by alkaline potassium hexacyanoferrate(n1) to form a highly fluorescent dimer.28 This reaction has attracted the attention of analysts in the clinica129~30 and forensic fields,31 yet has never been used from a kinetic point of view. Hence the continuous addition of reagent (CAR) technique was used in this work for the automated kinetic determination of morphine. The basis of this technique32 and its validation as an alternative methodology for accomplishing direct rate measurements on reactions on the millisecond time scale have been described elsewhere.33-3s In addition to a reliable means of automating kinetic methods for routine analyses by thorough mixing of sample and reagents, this technique overcomes some problems arising from the instability of the reaction products and undesirable interactions between excess of reagent and the reaction products [e.g., quenching due to excess of hexacyanoferrate(u1) in the oxidation of morphine], in contrast to conventional or even flow techniques.Experimental Reagents All chemicals used were of analytical-reagent grade and solutions were prepared in distilled water. Morphine hydro- chloride was a certified pharmaceutical sample, 299.5% pure, and was used as received. Stock solutions (100 pg m1-I) were prepared from the salt in distilled water.A potassium hexacyanoferrate(Ir1) solution (1.3 x mol I - l ) was also prepared by dissolving 42.7mg of the chemical (Merck) in 100 ml of distilled water. A 0.5 mol I-' Tris buffer solution was made by dissolving 60.5 g of tris(hydroxymethy1) methyl- amine (Tris) (Merck) and 66.4 g of sodium chloride (Merck) in water and adding sufficient sodium hydroxide (Merck) to adjust the pH to 10 in a final volume of 11. Apparatus The instrumental set-up used to implement the CAR tech- nique consisted essentially of an addition unit composed of an autoburette (Mettler DL40) and a fan stirrer which was fitted to a spectrofluorimeter (Perkin-Elmer 650-10s) furnished with a special rectangular reaction vessel (100 x 40 ~ 4 0 mm, volume = 100 ml) made of Herasil quartz the sides of which924 ANALYST, JULY 1993, VOL.118 were optically polished, and a data acquisition and processing unit consisting of a Mitac PC-AT 12-MHz compatible computer equipped with a PC-Multilab PCL-711 analogue-to- digital converter running a program developed by the authors for application to reaction-rate methods. Procedures Kinetic-spectrofluorimetric determination of morphine To the reaction vessel of the CAR system were added a known volume of the standard morphine solution containing between 1 and 55 pg of the drug, and 30 ml of 0.5 moll-1 Tris buffer solution of pH 10, and the mixture was accurately diluted to 60 ml with distilled water. Then, 1.3 X moll-l potassium hexacyanoferrate(II1) solution was added at 20 ml min-l, with stirring at 125 rev min- l .The reaction was monitored spec- trofluorimetrically at A,, = 315 nm and k,, = 435 nm (excitation and emission slit-widths, 10 and 15 nm, respec- tively) and the kinetic data (relative fluorescence intensity versus time) were acquired at a rate of 75 ms per point, the maximum reaction rate being measured in about 5 s. Determination of morphine in human urine In preparing the urine samples, pre-packed ion-exchange columns (Cat. No. 11113) were used as supplied by Bio- systems Laboratories. An appropriate volume of urine (2.0 ml) was diluted in 15 ml of 0.1% d m ethylenediamine- tetraacetic acid (EDTA) and adjusted to pH 6.5 with 0.5 mol I-' phosphate buffer (pH 6.5), and the mixture was applied to the column, which was then washed with 5.0 ml of 0.02 mol I-' phosphate buffer (pH 6.5) also containing 0.2% m/m EDTA, and with 5.0 ml of distilled water.Free morphine was eluted with 10 ml of 1.0 mol 1-I hydrochloric acid. The eluate was collected in the reaction vessel, to which 1 .O moll- sodium hydroxide was added in order to neutralize the hydrochloric acid, after which the above-described pro- cedure was carried out. + Fe(CNI63- NCH3 Tris buffer pH 10 _____) 1 2 NCH3 Results and Discussion Morphine (1) undergoes oxidative dimerization in an alkaline solution, by thc action of the potassium hexacyanoferrate(II1) added from the autoburette, to form highly fluorescent pseudomorphine (2). This reaction is fairly rapid, but can be followed using the CAR technique by monitoring the reaction rate spectrofluorimetrically at the excitation and emission wavelengths (315 and 435 nm, respectively) of the pseudo- morphine formed.Fig. l(a) shows the CAR kinetic curve provided by the spectrofluorimeter and processed by the data acquisition system in a single experiment. In the kinetic curve, the relative fluorescence intensity increases as potassium hexacyanofer- rate(1Ir) is added from the autoburette as a result of the formation of pseudomorphine by oxidation of the phenoxide ion at position 3. When the morphine concentration becomes negligible , excess of hexacyanoferrate(rI1) quenches the flu- orescence, probably through further oxidation of the pseudo- morphine formed (the structure possesses two phenol groups), thereby giving rise to the decaying portion of the kinetic curve following the maximum.As can be seen in Fig. l(b), when the reaction is monitored in an equilibrium mode [all ingredients are mixed instantaneously and the potassium hexacyanoferrate(n1) concentration corresponds to the actual concentration in the reaction vessel relative to the maximum in the CAR profile], the quenching effect of hexacyanofer- rate(u1) poses a severe problem, so measurements must be made at a controlled time after the reaction is started in order to obtain reproducible results. This requirement calls for a kinetic methodology based on reaction-rate measurements, which will provide increased selectivity, precision and throughput36 in routine analyses for morphine. 0.4 0.3 0.2 > C a, C a, C a, w 'jj 0.1 4- .- s o 5 10 15 20 a, m [I: .- c.- 0.3 0.2 0.1 I 1 0 20 40 60 80 Time/s Fig. 1 Relative fluorescence intensity versus time curves for the determination of SOOngml-' of morphine using: (a) the CAR technique and (6) equilibrium measurements at a fixed potassium hexacyanoferrate(n1) concentration of 9.75 X moll-l. (For other experimental conditions, see under Procedures)ANALYST, JULY 1993, VOL. 118 925 Effect of Reaction Variables From the curve in Fig. l ( a ) , three measurement parameters, viz., the maximum reaction rate (measured on the linear rising portion of the curve), the maximum fluorescence intensity and the quenching rate, can be used for the routine spectrc- fluorimetric determination of morphine. However, previous studies37 showed that the read-out method based on the maximum reaction rate is the most suitable in terms of sensitivity, precision and sample throughput.Therefore, the effects of a number of major factors, such as the pH that gave the maximum yield of the dimeric product and the concentra- tion of potassium hexacyanoferrate(II1) and its rate of addition, were studied from maximum reaction rate- concentration plots. The rate of addition of the potassium hexacyanoferrate(rI1) was studied over a wider range, viz., 0.25-25 mlmin-l. Tncreasing rates of addition resulted in increasing maximum reaction rates up to about 14 ml min-l, above which they remained virtually constant or even decreased slightly at very high addition rates. Fig. 2(a) shows a typical plot for the dependence of this variable on the measurement parameter (a square root relationship between the maximum reaction rate and the addition rate) based on the theoretical basis of the CAR technique,32 which conforms to the equation: Maximum reaction rate = - (:Irn k[hexacyanoferrate( I I ~ ) ] ( ~ U vo = K { r'; [morphinelo (1) where K is a proportionality constant, k is the second-order rate constant, u is the addition rate, Vo is the initial sample 15 I I 0 1 2 3 4 5 U .- c [Addition rate/ml min '1; $ 25 2 ( ( b ) I I I 1 2 3 4 5 6 0 ([K3Fe(CN)&mol I-'); X Fig.2 Effect of (a) addition rate and (b) potassium hexacyanofer- rate(m) concentration on the kinetic-spectrofluorimetric determina- tion of 500 ng ml-' of morphine at pH 8.45 (0.05 moll-I Tris buffer). Other experimental conditions: (a) [K3Fe(CN)6] = 3 X moll-'; (b) addition rate = 20 ml min-l volume and I F is the fluorescence intensity.As can be seen in Fig. 2(a), the maximum reaction rate was linearly dependent on the addition rate over the range 0.25-12.5 ml min-I; the deviations observed at higher values can be ascribed to the quenching by hexacyanoferrate(rr1) [increased addition rates also increase the actual concentration of hexacyanoferrate(n1) in the reaction vessel per unit time and hence its quenching effect]. From these results, an addition rate of 20 ml min-l was chosen, which lay in a region of minimal variations and no appreciable error in the measured rate; in addition, the kinetic curves included a sufficiently wide portion for the maximum reaction rate to be readily determined. The concentration of potassium hexacyanoferrate(Ir1) added from the autoburette was varied between 3.0 x and 3.0 X moll-l.A similar dependence to that on the addition rate was observed, as shown in Fig. 2(b), where the linear portion at low hexacyanoferrate(rI1) concentrations is consistent with the predictions of eqn. (1). A 1.3 x mol 1-1 potassium hexacyanoferrate(r1r) concentration was therefore selected for further experiments on the same grounds as above. The influence of pH on the maximum reaction rate was studied over the range 7.5-12.0 (Fig. 3). The plot of the maximum reaction rate versus pH shows a virtually zero-order dependence at pH 9.5-10.5, where morphine is dissociated via its phenol group according to the reported pK.8 The formation of pseudomorphine was favoured above pH 10; however, even though the oxidation of morphine is pH-independent in this range,8 the decrease in the maximum reaction rate reflected in the final portion of Fig.3 can be ascribed to faster oxidation of pseudomorphine. Based on this kinetic dependence, a pH of 10 was selected for use in the kinetic determination of this drug; the pH was adjusted in the reaction vessel by adding 30 ml of 0.5 mol 1-l Tris buffer (pH 10). Finally, other instrumental variables were also investigated. The stirring speed (65-145 rev min-I) did not significantly affect the measured rate; hence 125 rev min-l was selected. The influence of the initial sample volume, V,, was studied between SO and 80ml at a constant amount of added morphine.The relationships found were consistent with the dilution or concentration involved. An initial sample volume of 60 ml was therefore selected. Kinetic Determination of Trace Amounts of Morphine The maximum reaction rates calculated from the relative fluorescence intensity versus time curves for solutions contain- ing various amounts of morphine were analysed. Under the selected conditions, they were found to be linearly related to the morphine concentration through the following regression equation: Maximum reaction rate = -1.07 X + 2.5 x 10-4[morphine]o (Y = 0.999, rz = 15) 40 I v) m I 0 7 3 30 J+ 2 C 0 .- g 20 5 P 2 E 10 .- X 0 8 9 10 11 12 PH Fig. 3 Influence of pH on the maximum reaction rate. [Morphine] = 500 ng ml-l. (For other experimental conditions, see under Pro- cedures)926 ANALYST, JULY 1993, VOL.118 Table 1 Analytical figures of merit for the determination of morphine by the CAR technique Dynamic linear range 15-925 ng ml-l 4.5 X 10V-2.5 x mol I-' Sensitivity Detection limit Sampling rate 48 samples h- 2.5 x 10-4s-1 ml ng-' 2.76 ng ml-l 9.5 x 104 s-l I mol-' 7.4 x 10 -9 moll- 1 Precision (relative standard deviation) 1.77%* 2.79% 1 * Eleven determinations of 500 ng ml-' of morphine. t Eleven determinations of 50 ng ml-1 of morphine. Table 2 Influence of common narcotic drugs on the determination of SO ng ml-' of morphine Drug added Tolerated ratio (m/m) Nicotine, cocaine, codeine, methadone, papaverine Perphenazine 25 Chlorpromazine, promethazine, quinine 10 Ace to promazi ne 5 100 Table 3 Recovery of morphine added to urine samples Sample 1 2 3 4 5 6 7 8 Added/ pg 55 45 35 25 15 4 3 2 Concentration in urine/pg ml- 27.5 22.5 17.5 12.5 7.5 2.0 1.5 1 .0 Found/pg 55 .O 45.0 33.6 22.6 13.2 3.5 2.6 1.7 Recovery 100.0 100.0 96.0 90.4 88.0 87.5 86.0 85 .O (Yo 1 where [morphinelo is expressed in ngml-l.Table 1 sum- marizes the analytical features of the proposed CAR method. The detection limit was calculated by the IUPAC recommen- ded procedure38 and the sample throughput was determined from the time taken to perform three replicate analyses including sample changeover in the CAR system (about 30 s). As can be seen, morphine can be determined at the ng ml-l level over a wide concentration range with good precision and a high sample throughput. A study of potential interferents was carried out in order to determine the tolerated limits for various compounds related to morphine ( e .g . , pharmacological agents structurally similar to morphine and narcotic drugs). The maximum concentra- tion of foreign compound studied was 5.0 pg ml-l for a solution containing 50 ng ml-1 of morphine (tolerated limit, 100: 1 ndm ratio). The tolerated limits for the foreign compound were taken as the largest amounts yielding errors of less than +5% in the maximum reaction-rate measurements. As can be seen from Table 2, the proposed method is highly selective: of all the compounds tested, the only serious interference was caused by acetopromazine, although at a fairly high concentration (five times that of morphine). Determination of Morphine in Urine Samples As stated earlier, monitoring morphine in urine is of great interest for various purposes.Therapeutic concentrations39 of morphine in urine range from 0.5 to 10 pg m1-*, although concentrations can be as high as 14-81 pg ml-l in overdosed addicts.' The determination of this drug in urine samples can also be used in doping analysis2 to ascertain consumption of analgesic preparations containing codeine, for which mor- phine is a metabolite; in such cases, the morphine concentra- tion in urine is typically about 1.5 pg ml-I. Based on these levels, the sensitivity and selectivity of the proposed method provide a useful, straightforward alternative for the determi- nation of this drug in this type of sample. Reported methods for the determination of morphine in urine samples include a sample preparation step and isolation of the drug from the sample matrix prior to analysis, even if a chromatographic technique is used, in order to avoid interfer- ences arising from overlapping peaks.Liquid-liquid extrac- tion31 has been the traditional choice for the removal of such interferences; however, any emulsions formed may result in losses of analyte; solid-liquid extraction ,40 usually implemen- ted with Sep-Pak CIS cartridges, is a good alternative in this respect. Both methodologies were examined in this work for the isolation of morphine from biological material in the urine samples; however, the results were unsatisfactory: fluores- cence signals were too high before any hexacyanoferrate(ii1) was added from the autoburette and, rather than the typical kinetic curve, addition of the oxidant gave rise to fluorescence quenching.In this work, an alternative procedure for the isolation of morphine from the sample matrix was examined, based on the use of a commercially available cation-exchange column and the procedure described under Experimental. Urine samples were spiked with different amounts of morphine (it was impossible to obtain samples from patients being administered morphine) and recoveries determined in order to investigate potential matrix effects by comparing the results obtained before and after addition of the morphine solutions. As can be seen in Table 3, the recoveries (between 85 and 100%) are similar or even higher than others previously reported for a variety of sample preparations; hence the proposed method is very suitable for the analysis of these samples.The results refer to a 2.0ml urine sample volume; however, the volume can be increased up to 10 ml in order to preconcentrate the samples with no appreciable affect on the recoveries. The precision of this determination was assessed on 11 samples, each containing 10 pg m1-I of morphine; the relative standard deviation was found to be 8.3%. Based on the morphine concentrations in urine, as listed in Table 3, the proposed method is useful for the rapid routine control of morphine in the above-mentioned applications (therapeutic, overdose and doping analysis). As with previous alternatives, the proposed method can be extended to the determination of morphine metabolites. Morphine is excreted in urine mainly as morphine-3-glucuro- nide, whereas less than 10% of the drug is excreted as unchanged morphine.However, by using acidic or enzymic hydrolysis, 1 the former can be converted into the latter; hence morphine-3-glucuronide can be determined by difference in two separate analyses. Because the proposed method allows the determination of free morphine in urine, there should be no difficulties, as the metabolite occurs at a much higher concentration. Conclusions The CAR technique is a useful means of developing kinetic determinations for species involved in fast reactions by using spectrofluorimetric detection and inexpensive instrumenta- tion. As regards the determination of morphine, the proposed kinetic method allows the analyte to be assayed in the presence of other commonly used drugs with a high sensitivity and sample throughput and none of the disadvantages inherent in equilibrium spectrofluorimetric analyses.Also, the proposed method allows the determination of morphine inANALYST, JULY 1993, VOL. 118 927 urine samples, which is of interest in controlling a wide variety of medicinal problems (e.g., therapeutic, overdose and doping analysis). Even though high-performance liquid chromatography (HPLC) with electrochemical detection is currently one of the most frequently used techniques for the determination of morphine, the analytical feature; of the proposed method make it an advantageous choice for the determination of this drug in urine. On the other hand, the higher sensitivity of the chromatographic technique and the typically low volumes used make it particularly useful in those cases where the morphine concentration is outside the range of the proposed method.As regards the determination of morphine in urine, the throughput of the proposed CAR method, which requires only a few seconds per sample, is clearly superior to that obtainable with a chromatograph. The throughput times are obviously those expended in the determination step alone as both the proposed and HPLC methods require free morphine to be previously isolated from the sample. In addition, analytical costs are also lower as a result of using less expensive instrumentation, reagents and solvents. The authors gratefully acknowledge financial support from the Direccion General Interministerial de Ciencia y Tecnologia (DIGICyT) (Project No.PB91-0840). 1 2 3 4 5 6 7 8 9 10 11 12 13 References Clarke’s Isolation and Identification of Drugs, ed. Moffat, A. C., Pharmaceutical Press, London, 1986. Le Monde Cycliste, Report on the UCI Medical Comission meeting of 28 January 1990, April 1990. Analytical Abstracts, Computer Files (1980-June 1992), Royal Society of Chemistry, Cambridge. Konishi, M., and Hashimoto, H., J. Pharm. Sci., 1990,79,379. Tagliaro, F., Carli, G., Dorizzi, R., and Marigo, M., J. Chro- matogr., 1990,507,253. McLean, C. F., Mather, L. E., Odontiadis, J., and Sloan, P. A., J. Pharm. Pharmacol., 1990,42, 669. Staub, C., and Zwahlen, A. L., Lab. Med., 1990, 13, 392. Jordan, P. H., and Hart, J.P., Analyst, 1991, 116, 991. Mason, J. L., Ashmore, S. P., and Aitkenhead, A. R., J. Chromatogr. Biomed. Appl., 1991, 108, 191. Verwey-van-Wissen, C. P. W. G. M., Koopman-Kimenai, P. M., and Vree, T. B., J. Chromatogr. Biomed. Appl., 1991, 108, 309. Uhrich, M., and Tillmanns, U., GIT Fachz. Lab., 1990, 34, 1265. Solans, A., de la Torre, R., and Segura, J., J . Pharm. Biomed. Anal., 1990, 8, 905. Cone, E. J., Welch, P., Mitchell, J. M., and Paul, B. D., J. Anal. Toxicol., 1991, 15, 1. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Lillsunde, P., and Korte, T., J. Anal. Toxicol., 1991, 15, 71. Goldberger, B. A., Caplan, Y. H., Maguire, T., and Cone, E. J., J. Anal. Toxicol., 1991, 15. 226. Grinstead, G. F., J. Anal. Toxicol., 1991, 15, 293.McCord, C. E., and McCutcheon, J. R., J. Anal. Toxicol., 1988, 12, 295. Gerchow, J., Schmidt, K., and Raudonat, H. W., Fresenius’ Z . Anal. Chem., 1988, 330. 461. Tagliaro, F., Marigo, M., Dorizzi, R., and Rigolin, F.. Clin. Chem. (Winston Salem, N. C . ) , 1988,34, 1365. Standefer, J. C., and Backer, R. C., Clin. Chem. (Winston Salem, N.C.), 1991,37,733. Gjerde, H., Christophersen, A. S . , Skuterud, B., Klementsen, K., and Moerland, J., Forensic Sci. Znt., 1990, 44, 179. Blum, L. M., Klinger, R. A., and Rieders, F., J . Anal. Toxicol., 1989, 13, 285. Cody, J. T., and Schwarzhoff, R. H., J. Anal. Toxicol., 1989, 13, 277. Offidani, C., Carnevale, A., and Chiarotti, M., Forensic Sci. Int., 1989,41, 35. Ostrea, E. M., Jr., Parks, P. M., and Brady, M. J., Clin. Chem. (Winston Salem, N.C.), 1988, 34, 2372. Stanley, S . , Jeganathan, A., Wood, T., Henry, P., Turner, S . , Woods, W. E., Green, M., Tai, H. H., Watt, D., Blake, J., and Tobin, T., J. Anal. Toxicol., 1991, 15, 305. Laurie, D., Manson, A. J., Rowell, F. J., and Seviour, J . , Clin. Chim. Acta, 1989, 183, 183. Bentley, K. W., and Dyke, S. F., J. Chem. SOC., 1959, 2574. Nelson, P. E., Nolan, S. L., and Bedford, K. R., J. Chromat- ogr., 1982, 234, 407. Brodzinska, D., Chem. Anal. (Warsaw), 1988,33, 383. Jane, I . , and Taylor, J. F., J. Chromatogr., 1975, 109, 37. Velasco, A., Silva, M., and Perez-Bendito, D., Anal. Chem., 1992, 64,2359. PCrez-Bendito, D., Silva, M., and G6mez-Hens, A., TrAC, Trends Anal. Chem., 1989, 8, 302. Marquez, M., Silva, M., and PCrez-Bendito. D., Anal. Chim. Acta, 1990, 237, 353. MBrquez, M., Silva, M., and Perez-Bendito, D., Anal. Chim. Acta, 1990, 239,221. Perez-Bendito, D., and Silva, M., Kinetic Methods in Analytical chemistry, Ellis Horwood, Chichester, 1988. MBrquez, M., Silva, M., and Perez-Bendito, D., Anal. Lett., 1989, 22, 2485. Long, G. L., and Winefordner, J. D., Anal. Chem., 1983, 55, 712A. Fundamentals of Clinical Chemistry, ed. Norbert, W. T., Saunders, Philadelphia, PA, 1987. Svensson, J. O., Rane, A., Sawc, J., and Sjoqvist, F., J. Chromatogr. Biomed. Appl., 1982, 230, 427. Paper 31009983 Received February 19, 1993 Accepted April 2, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800923

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 929 ln Situ Determination of High Hydrogen Ion Concentrations in Aqueous Solutions Using a-Nafion Membrane Containing Rhodamine 6G Hari Mohan and R. M. lyer Chemistry Division, Bhabha Atomic Research Centre, Tromba y, Bomba y-400 085, India ~ ~~~ Based on the different optical absorption spectrum and fluorescence emission intensity of Rhodamine 6G in neutral and acidic solutions, a Nafion membrane containing Rhodamine 6G was successfully employed to determine indirectly high H+ concentrations (1-5 rnol dm-3) in aqueous solutions in the temperature range Keywords: Hydrogen ion concentration; Nafion membrane; Rhodamine 6G; spectrophotometry; spectro fluorimetr y 25-1 00 "C. The hydrogen ion concentration (pH) of a solution is usually determined using a pH meter or pH-sensitive papers.1 However, it is difficult to determine the H+ concentration if it is more than 1.0 rnol dm-3 because pH meters are normally calibrated for the pH range 0-14.High H+ concentrations can bc determined after suitable dilutions using standard meth- ods. The diluted solution may not represent the true H+ concentration owing to the difference in the dissociation constant of the acid in the concentrated conditions. A polymeric membrane, Nafion, has emerged as a potential medium in which high concentrations of organic dyes could be incorporated.2.3 The Nafion membrane consists of a tetraf- luoroethylene backbone with a perfluorinated ether side-chain and terminal sulfonic acid groups.4 This structure provides the membrane with thermal, mechanical and chemical stability and low optical absorption at wavelengths >250 nm.4 During our studies on the optical absorption characteristics of organic dyes in Nafion membranes, it was observed that Rhodamine 6G has different optical absorption spectra in dry and wet conditions of the membrane .5 This difference was explained as being due to the different H+ concentrations in dry and wet conditions of the membrane.5 Among the many uses of organic dyes in analytical chemistry, such as indicators in acid-base titrations, Hirsch- feld"8 proposed their use as fluorescent probes for the determination of antigens in biological fluids and examining weakly fluorescent dyes.Organic dyes are also used as an optical sensor for determination of pH, and metal and halide ions.9 Here we report the use of the organic dye Rhodamine 6G incorporated in an ion-exchange membrane for the determination of high H+ concentrations in aqueous solu- tions.The determination is based on the changes in the optical absorption spectrum of Rhodamine 6G and in the fluores- cence intensity at high acid concentrations. The results for the application of a Nafion membrane containing Rhodamine 6G as an optical and fluorescent probe for the in situ determina- tion of high acid concentrations are reported in this paper. Experimental Apparatus A Hitachi (Tokyo, Japan) Model 330 spectrophotometer was used to monitor the optical absorption spectrum of Rhoda- mine 6G in a Nafion membrane and in aqueous solutions. Steady-state fluorescence emission measurements were car- ried out with a Hitachi Model F 4010 spectrofluorimeter.Reagents De-ionized Nanopure water was used to prepare all solutions. Perchloric acid (60%) was obtained from Merck (Darmstadt, Germany), Nafion-117 (H+) from DuPont (Wilmington, DE, USA) and Rhodamine 6G (Basic Red 1, CI 45160, CAS Registry Number 989-38-8, laser grade, A,,, = 525 nm in water) from Lambda Physik (Gottingen, Germany) and used without any further purification. All other chemicals used were of analytical-reagent grade. Procedure The Nafion membrane was first treated with 2.0 mol dm-3 HCl to ensure its complete conversion into the H+ form. It was then stored in distilled water for at least 24 h to remove any free excess acid. The density and thickness were determined to be 1.675 g ~ m - ~ and 0.0196 cm, respectively.10 A small piece of Nafion membrane (1 X 3 cm) of known mass was equilibrated with a known volume of standard solution of Rhodamine 6G (1.0 x mol dm-3).The solution was continuously stirred with a magnetic stirrer. The concentra- tion of the dye incorporated into the membrane was deter- mined by measuring the decrease in the concentration of the dye in the aqueous solution. The equilibrated membrane was washed with distilled water and used for the determination of the H+ concentration, for which the membrane containing Rhodamine 6G was kept in a 1 X 1 cm spectrophotometric cell to which 3.5 cm3 of the aqueous solution whose H+ concentration was to be determined were added. The membrane was allowed to stand in the cell for 10 min.For optical absorption studies, the membrane was held at an angle of 90" to the incident light beam whereas for fluorescence emission studies it was held at 45" with respect to both the incident and emitted light beams. Results and Discussion Rhodamine 6G as an Optical Absorption Probe Fig. 1(A) shows the optical absorption spectrum of the Nafion membrane containing Rhodamine 6G (7.9 X mol dm-3), immersed in neutral water, which exhibits an absorption band with h,,, = 525 nm. This absorption spectrum is similar to that of Rhodamine 6G in neutral aqueous solution. When this membrane is immersed in HC104, the absorption at 525 nm decreases with a simultaneous increase at 470 nm [Fig. l(B)]. This new absorption spectrum corresponds to Rhodamine 6G in acidic solution.5 A Nafion membrane not containing Rhodamine 6G shows very little absorption in this region [Fig.l(C)]. Earlier studies5 have shown that Rhodamine 6G is present in the cationic form (A) in neutral aqueous solutions and in the protonated form (B) in acidic solutions with a pK value of -0.38 (at an H+ concentration of 2.4 mol dm-3). The absorption spectrum of Rhodamine 6G in a Nafion membrane changes from the cationic form, A [Fig. l(A)], to the930 ANALYST, JULY 1993, VOL. 118 1 .o Q, C m 0.5 s a n 0 400 500 600 Wavelengthlnm Fig. 1 ing Rhodamine 6G (7.9 x and C, in the absence of Rhodamine 6G Optical absorption spectrum of a Nafion membrane contain- mol dm-"): A, in water; B, in HClO,; protonated form, B [Fig. l(B)], as the concentration of acid is increased.The absorption spectrum of a Nafion membrane containing Rhodamine 6G at any given concentration of H+ attains constant absorption within 10 min and it remains constant as long at it is immersed in the solution, and the absorption at 525 nm decreases whereas that at 470 nm increases with increase in the concentration of HClO,. The ratio of the intensities at 525 and 470 nm is thus found to be a measure of acid concentration. A B Fig. 2 shows the variation of the 525 : 470 nm intensity ratio as a function of HC104 concentration. This ratio is found to be independent of Rhodamine 6G concentration in the mem- brane. Hence from this calibration graph, the acid concentra- tion of any unknown aqueous solution could be determined on equilibrating a Rhodamine 6G-containing Nafion membrane with the solution and determining the ratio of the absorbances at 525 and 470 nm.Fig. 2 could not be used, however, to determine the true H+ concentration of the aqueous solution as the Hammett acidity function (Ho)*l values are different for the same molar concentration of different acids. Therefore, a plot of absor- bance ratio versus Ho would essentially give the true H+ concentration in the aqueous solutions, irrespective of the nature of the acid. 8 1 400 0 5 10 HClO,/mol dm-3 Fig. 2 Variation of 0, the ratio of absorbance at 525 and 470 nm and a, fluorescence intensity (h = 550 nm) as a function of HC104 concentration Fig. 3 shows a plot of the ratio of absorbances at 525 and 470 nm versus Hammett acidity function for different acids.The values for the different acids lie close to the same line and show the suitability of this method for the determination of high H+ concentrations in aqueous solutions of strong acids. Rhodamine 6 6 as a Fluorescent Probe We have reported previously5 that the cationic form (A) of Rhodamine 6G is highly fluorescent whereas the protonated form (B) is non-fluorescent. As the relative fraction of each form is dependent on the concentration of H+ ions in the aqueous solution, the fluorescence emission from Rhodamine 6G will vary with H+ concentration and could be used as a measure of the H+ concentration in the aqueous solution. The fluorescence emission intensity at he, = 550 nm (hex = 345 nm) of Rhodamine 6G in a Nafion membrane when kept in a cell containing a solution of HC104 was found to decrease with increasing acid concentration (Fig.2). The variation in the fluorescence intensity followed a similar trend to that ob- served for optical absorption changes (Fig. 2) with HC104 concentration. Therefore, a Nafion membrane containing Rhodamine 6G could also be used as a fluorescent probe for the determination of high H+ concentrations in aqueous solutions. Leaching of Rhodamine 6 6 From the Nafion Membrane An important factor that should be considered before the use of this approach for determining H+ concentration is the stability of the membrane system. Leaching of Rhodamine 6G from the Nafion membrane in aqueous solution was studied by monitoring the absorbance (525 nm) of a membrane contain- ing Rhodamine 6G (1.8 x 10-3 rnol dm-3) that was kept immersed in water for various periods of time [Fig.4(A)]. The absorption spectrum remained constant even after immersion of the membrane for 20 d in water [Fig. 4(A)], showing it is stable in water. However, some leaching was observed when it was immersed in HC104 (9.8 rnol dm-3). Fig. 4(B) shows the variation in absorbance (470 nm) of Rhodamine 6G (1.9 X rnol dm-3) as a function of equilibration time in HC104 (9.8 rnol dm-3). As 10 min are sufficient for the membrane to attain constant absorption in HC104 solution, during this period leaching of Rhodamine 6G is negligible. Therefore, a Nafion membrane containing Rhodamine 6G can be used to determine the H+ concentration of aqueous solutions. The Nafion membrane as such is stable in HClO,.Therefore, this membrane containing Rhodamine 6G can be used for an indefinite time. From the calibration graph (Fig. 2), it is clear that appreciable changes in the ratio of absorbances at 525 and 470 nm and in the fluorescence emission intensity are observed when the acid concentration is within the range 1-5 mol dm-3.ANALYST, JULY 1993, VOL. 118 93 1 For higher acid concentrations, the changes in the ratio of the absorbances at 525 and 470 nm and the fluorescence intensity are very small. Therefore, it appears that the membrane would be more useful for the determination of acid concentra- tions in the range 1-5 rnol dm-3. Interferences From Other Solutes Rhodamine 6G is a cationic dye and the presence of other cations in the solution may affect the displacement of Rhodamine 6G from the membrane.In order to check this, a Nafion membrane containing Rhodamine 6G was kept in a 1 .O mol dm-3 aqueous solution of LiCl and the concentration of Rhodamine 6G in the aqueous solution was monitored as a function of time. Even after 5 d of equilibration of the membrane in the solution, leaching of Rhodamine 6G was not observed. A 1.0 mol dm-3 NaCl solution also failed to displace Rhodamine 6G from the membrane. Anions are not expected to displace Rhodamine 6G from the membrane. As the method depends on the optical absorption and emission in the range 400-600 nm, it implies that the test solution should not absorb or emit in this region. Determination of H+ Ions at High Temperature In order to assess the applicability of this membrane for the determination of H+ concentrations at high temperatures, the stability of the membrane containing Rhodamine 6G was tested in boiling water.It was found that the optical absorption spectrum remained unchanged even after heating for 10 h at 100°C. However, on heating the membrane at 150°C in water in an autoclave for 6 h, the optical absorption spectrum was decreased to approximately half of its original size. The aqueous solution did not show the presence of Rhodamine 6G. This suggests that Rhodamine 6G was not leached from the membrane, but might have decomposed in the membrane at 150°C. The membrane as such is stable at 150"C.4 In order to confirm this, an aqueous solution of Rhodamine 6G on heating at 150 "C for 6 h also showed similar behaviour. The aqueous solution of Rhodamine 6G remained unaffected on heating at 100°C for 10 h.Therefore, the membrane containing Rhodamine 6G could be safely used for the determination of H+ in aqueous solutions at temperatures as high as 100°C. A Nafion membrane containing Rhodamine 6G was found to be stable in water up to 100°C whereas leaching was observed in 5.0 mol dm-3 acid. Therefore, it is important to establish the behaviour of a Nafion membrane containing Rhodamine 6G in 5.0 rnol dm-3 acid at 100°C before it could be recommended for use at this temperature. Leaching of Rhodamine 6G from the Nafion membrane kept in 5.0 10 0 PI d < $ 5 0 - 1 - 2 -3 -4 -5 HO Fig. 3 Variation of the ratio of absorbance at 525 and 470 nm as a function of the Hammett acidity function (Ho) for: A, HClO,; 0, H,SO,; a, H,PO,; A, HCl; and x , HN03 mol dm-3 HC104 at 100°C was monitored as a function of time.After 10 h of heating at lOO"C, it was observed that the concentration of Rhodamine 6G had decreased from 2.83 x to 2.31 x rnol dm-3, i.e., a decrease of 18%, whereas at room temperature the decrease in the concentra- tion of Rhodamine 6G was only about 3% for the same period. Therefore, it appears that leaching of Rhodamine 6G by acid is accelerated at 100°C. However, as only 10 min are required for determination, leaching is expected to be negligible during this period. The ions Li+ and Na+ are not able to displace Rhodamine 6G from a Nafion membrane when kept in 1.0 mol dm-3 solutions of LiCl and NaCl at room temperature.Displacement was also not observed when the solution was heated to 100°C. Even in the presence of 5.0 rnol dm-3 acid at lOO"C, displacement of Rhodamine 6G by Li+ and Na+ ions was comparable to that observed in the absence of these ions (18%). Hence the presence of Li+ and Na+ ions does not enhance the displacement of Rhodamine 6G in 5.0 mol dm-3 acid at 100°C. Therefore, it is concluded that the membrane can be used to determine high H+ concentrations at tempera- tures as high as 100°C. Precision and Accuracy The optical absorption spectrum of Rhodamine 6G in a Nafion membrane was fairly reproducible, showing that Rhodamine 6G had been loaded uniformly into the Nafion membrane. The absorbance at 525 nm varied between 0.90 and 0.95 when the experiment was repeated eight times.For the same piece of membrane, the absorbance at 470 nm varied between 0.14 and 0.17. The variation in these results gave precisions (relative mean deviations)l2 of 1.4 and 5.2% at 525 and 470 nm, respectively. The standard deviation at 525 and 470 nm was 0.018 and 0.01 A, respectively. The accuracy of the method can be determined by comparison with the exact or most probable value of the acid concentration in the aqueous solution. Titration of the acid solution with standard alkali solution may not yield correct results as dilution would change the dissociation constant and hence the true H+ concentration. The accuracy of this method could, however, be calculated from Fig. 3, which shows the variation of the absorbance as a function of Ho for different acids.Between 230 values of -0.2 and -1.0 (acid concentra- tion approximately 1.0-2.5 rnol dm-3), the accuracy is within 0.3 rnol dm-3 concentration of the acid. The accuracy for Ho values between - 1.0 and -2.0 (acid concentration approxi- mately 2.5-5.0 rnol dm-3) is 0.2 rnol dm-3. For higher concentrations, the accuracy is within 0.1 rnol dm-3. Conclusions It is concluded that a Nafion membrane containing Rhoda- mine 6G can be used to determine high H+ concentrations (1- 5 mol dm-3) in aqueous solutions. The presence of other 1.0 ' I I 0 10 20 Tim e/d Fig. 4 Leaching of Rhodamine 6G from a Nafion membrane at ambient temperature on equilibrating in A, water (525 nm) and B, 9.8 mol dm-3 HC10, (470 nm) as a function of time932 ANALYST, JULY 1993, VOL.118 cations in the test solutions, such as Li+ and Na+, does not affect the determination. The membrane has many potential applications for the determination of high acid concentrations even at 100°C. 5 6 7 Mohan, H., and Iyer, R. M., J. Chem. SOC., Faraday Trans., 1992, 88, 41. Hirschfeld, T., Block Engineering, US Pat. 4 166 105, 1979; Chem. Ahstr., 1979, 91, 189324b. Hirschfeld, T., Block Engineering, Can. Par., 1059 785, 1978; Chem. Ahstr., 1979, 91, f06915q.- Hirschfeld, T., Block Engineering, Br. Pat., 1566427, 1980; Chem. Abstr., 1980, 93, 177071~. Sincere thanks are due to Dr. D. Nandan, Chemistry Division, for helpful discussions and the loin of the Nafion membrane. 8 References Vogel, A. I., A Text Book of Quantitative Inorganic Analysis, Longman, London, 1961, p. 50. Lee, P. C., and Meisel, D., Photochem. Photobiol., 1985, 41, 21. Mohan, H., Moorthy, P. N., and Iyer, R. M., Photochem. Photobiol., 1989, 49, 395. Sondheimer, S. J., Bunce, N. J., and Fyfe, C. A , , Macromol. Chem. Rev., 1986, C26, 353. 9 Seitz, W. R., in Biosensors: Fundamentals and Application, eds. Turner, A. P. F., Karube, I., and Wilson, G. S., Oxford University Press, New York, 1987, p. 599. 10 Nandan, D., Mohan, H., and Iyer, R. M., Proceedings of International Conference on Ion Exchange (ICIE), 1991, Tokyo, Kodansha, Tokyo, 1991, p. 491. Paul, M. A., and Long, F. A., Chem. Rev., 1957, 57, 1. Vogel, A. I., A Text Book of Quantitative Inorganic Analysis, Longman, London, 1961, p. 1122. 11 12 Paper 2105355A Received October 6, 1992 Accepted January 4, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800929

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 933 Spectrophotometric Determination of Lead in Tap Water With 5,10,15,20-Tetra(4-N-sulfoethylpyridinium)porphyrin Using Merging Zones Flow Injection Jeffery A. Schneider" and James F. Hornig Department of Chemistry, Dartmouth College, Hanover, NH 03755, USA Determination of lead at concentrations of 31 5 pg I-1 in aqueous samples was achieved by selective reaction with 5,10,15,20-tetra(4-N-sulfoethylpyridinium)porphyrin at pH 9.4 and 30°C in a merging zones flow injection manifold followed by detection at 480 nm after an elapsed time of 45 s. Interference by aluminium (125-fold excess), cadmium (25-fold excess), copper (25-fold excess), manganese (200-fold excess) and zinc (200-fold excess) was eliminated by the use of 1.0 mol I-1 NH3-NH4CI as a buffedmasking agent; interference by iron (10-fold excess) was eliminated by the addition of 1 % v/v of acetylacetone to the reagent mixture and use of the method of standard additions for sample analysis.Recoveries from tap water samples, to which various amounts of lead had been added, ranged from 98 to 109% with a detection limit of 10 pg I-' when iron was present in the sample matrix and 4.2 pg I-' when it was not. Results of analyses of tap water samples using this method were in good agreement with those obtained by electrothermal atomic absorption spectrometry. Keywords: Lead determination; flow injection; spectrophotometry; interference masking; tap water analysis The determination of lead in drinking water is becoming increasingly important. The maximum contaminant level (MCL) allowed by the US Environmental Protection Agency (EPA) was 50 pg 1-I at the tap, effective until the end of 1992.Beginning in 1993, the EPA replaced the concept of MCL with the goal of 'zero' lead, and the guidelines state that no more than 90% of the taps that are tested in a given location may show in excess of 15 pg 1-l of lead without corrective action being taken.' This means that for a given location being tested, if 90% or more of the taps are found to have 215 pg 1-1 of lead, procedures to decrease the amount of lead present will have to be implemented such that the location once again meets the EPA guidelines. Electrothermal atomic absorption spectrometry (ETAAS) is generally used to determine lead at these low concentrations but it is complex to use.Since the first reports of flow injection (FI)2,3 in 1974, there have been about 70 papers published relating to the determi- nation of lead in samples ranging from water4-9 to whole blood'(kl2 and gasoline .13 Limits of detection have ranged from ~0.514 to 20 pg l-I,15 with sample throughputs ranging from 69 to 120 h-l.16 Of these papers, only ~1~7.9117-20 reported the use of spectrophotometric detection and three of those included some type of preconcentration or extraction step. The method presented here bridges the sensitivity gap between flame and electrothermal (furnace) AAS and is faster than commercially available batch methods. Flame AAS is reliable down to about 0.2 mg 1-1 whereas ETAAS has an upper limit of about 0.2 mg 1-1 (unless the operator takes additional time to dilute the samples).Our method is based on the kinetically selective reaction at pH 9.4 of lead with the 'amphoteric ion type' water-soluble porphyrin 5,10,15,20- tetra(4-N-sulfoethylpyridinium)porphyrin21 [T(4-SP)P] (Fig. 1). In addition to its selective reaction with lead, T(4-SP)P does not adhere to surfaces over the pH range 2-12 and it does not precipitate from solution as readily as do similar water- soluble cationic porphyrins .21 Experimental General Method A 75 pl aliquot of 8 x 10-6 mol 1-1 T(4-SP)P in 1 mol I-' NH3- NH4C1 buffer containing 1% of acetylacetone is injected * Present address: Department of Chemistry, University of Cincin- nati, Cincinnati, OH 45221-0172, USA. simultaneously with a 150 pl aliquot of samplektandard into a de-ionized water carrier stream in a merging zones-FI manifold (Fig.2). The injected zones are combined at a mixing tee and the PbT(4-SP)P complex is detected spectrophoto- metrically approximately 45 s after injection. The carrier stream is propelled through the system by gas displacement from a reservoir using helium to provide the overpressure. A molar absorptivity for the PbT(4-SP)P complex of E = 25.6 x 104 1 mol-l cm-l at A,,, = 480 nm eliminates the requirement for preconcentration. The use of a three-dimensional (3D) knitted coil reactor provides improved mixing and at the same time reduces dispersion, thus improving detection limits. Apparatus A merging zones FI manifold was constructed from poly- (tetrafluoroethylene) (PTFE) tubing (Supelco, Bellefonte, PA, USA) of i.d.0.8 mm, two manually operated, six-port rotary injection valves (Type 5020, Rheodyne, Cotati, CA, USA) and a pressurized bottle for delivery of carrier. The manifold was connected to a variable-wavelength ultraviolet/ visible (UV/VIS) spectrophotometer (Model LC65T, Perkin- Elmer, Nonvalk, CT, USA). Peaks were recorded with a strip-chart recorder. The 3D knitted coil reactor was construc- ted as a series of intertwined knots using 3 m of 0.8 mm i.d. PTFE tubing. System flow rates were maintained at 2.0 ml min-I to provide a mean residence time of 45 s for all analyses. Reagents All water used was distilled and de-ionized and obtained from a Milli-Q water purification system (Millipore, Waltham, MA, USA). All metal standard solutions were prepared daily by dilution from atomic absorption-grade stock solutions (1000 pg m1-I) and stored in acid-washed low-density polyethylene bottles.Porphyrin reagent, T(4-SP)P, obtained from Por- phyrin Products (Logan, UT, USA), was 8.0 pmol 1-1 in tris(hydroxymethy1)methylamine (Tris) or NH3-NH4C1 buf- fer. The Tris buffer (0.1 mol l-l, pH 9.4) was prepared by dissolving 12.2 g of tris(hydroxymethy1)methylamine in water and adjusting to pH 9.4 with HCl. The NH3-NH4CI buffer (1.0 mol I-', pH 9.4) was prepared by addition of 80 ml of concentrated HC1 to approximately 500 ml of water followed by addition of 130 ml of concentrated aqueous NH3 and dilution to 1 1 with water.934 CH2CH2S03 L I ANALYST, JULY 1993, VOL. 118 Fig. 1 Structure of 5,10,15,20-tetra(4-N-sulfoethylpyridinium)porphyrin Helium Detector h = 480 nm Waste R = 75 pI Fig.2 Merging zones FI manifold Interference Studies Thirteen solutions each containing 100 pg 1-1 of lead were prepared. To 12 of them were added 2.5 or 20 mg 1-* of either aluminium, cadmium, copper, iron, manganese or zinc. The solutions were analysed using porphyrin reagent that had been buffered in either 0.1 mol I-' Tris or 1.0 moll-' NH3-NH4Cl buffer (pH 9.4) with calibration against non-acidified lead standards. For the iron interference studies, 15 solutions each contain- ing 100 pg 1-' of lead were prepared. To seven of them was added 1.0 mg 1-1 of iron. The solutions were analysed using porphyrin reagent, buffered in 1.0 mol 1-1 NH3-NH4Cl buffer (pH 9.4) to which either acetylacetone, citrate, glycerol, oxalate, salicylate, sulfamate or thiourea had been added as an iron-specific masking agent with calibration against non-acidi- fied lead standards.Recovery Studies After the suitability of each masking agent had been determined, a reagent mixture consisting of 8.0 pmol 1-1 T(4- SP)P in NH3-NH4Cl buffer (pH 9.4) containing 1% v/v of acetylacetone was used for the determination of lead in four spiked tap water samples that were prepared from the tap water in our laboratory without acidification. The method of standard additions was used for the analyses. Tap Water Analyses Thirty tap water samples were obtained from different locations in the Fairchild Science Building at Dartmouth College and analysed by ETAAS, and a sub-set of six samples and a blank were analysed spectrophotometrically by the method of standard additions using T(4-SP)P.Samples were not acidified. 0.01 A I H 3 min 100 Time - Fig. 3 Typical recorder output for a blank and a set 50-500 pg 1-1 Pb Results and Discussion )f standards, Whereas other FI methods for the determination of lead at concentrations below 100 pg 1-1 include a preconcentration step in order to increase the sensitivity of the method (see, e.g., refs. 6 and 8), the proposed method does not. From the absorbance versm lead concentration curve, optimum operat- ing variables were determined to be wavelength 480 nm, temperature 30 "C, pH 9.4, sample-to-reagent injection volume ratio 2 : 1, mean residence time 45 s, reactor type 3D knitted coil and buffedmasking agent NH3-NH4Cl containing acetylacetone.The optimum value for the variable of interest was taken to be at the point at which maximum sensitivity (i.e., the greatest slope) for the calibration graph was obtained, or for temperature, where an increase in sensitivity was offset by the increase in cost of maintaining a higher temperature. The optimum wavelength of 480 nm is the peak maximum of the Soret band for the Pb-T(4-SP)P complex. Although the molar absorptivity of the free porphyrin at this wavelength is at a minimum, it is not zero. Because of this, the blank generates a signal. Fig. 3 shows a typical recorder output for a series of lead standards including the blank. As reported by Igarashi and Yotsuyanagi,21 the reaction of lead and cadmium with T(4-SP)P is very pH dependent.The optimum pH of 9.4 was chosen because at higher pH theANALYST, JULY 1993, VOL. 118 1 - 5 0.8 v) 0.6 C m al .- .- w 0.4 U 935 - . . cadmium pseudo-first-order reaction rate increases rapidly, whereas the lead pseudo-first-order reaction rate begins to level out. At pH 9.4, the pseudo-first-order half-lives for the lead-porphyrin and cadmium-porphyrin reactions are 8 and 100 s, respectively. At the mean residence time of 45 s with our manifold, the lead reaction has gone to near completion (5.5 z), whereas the Cd reaction has nbt (0.45 z). Karlberg and Pacey22 stated that in order to obtain 'effective and reproducible' results, the reagent volume should be slightly larger than the sample volume in a merging zones manifold. Johnson and Petty23 reported that they could increase the generated analyte signal by reversing the roles of reagent and sample in a normal FI manifold, such that sample was continuously pumped and reagent was introduced as a discrete zone.Therefore, it was hypothesized that by making our merging zones manifold appear like a reverse FI manifold (z.e., making the injected sample volume larger than the injected reagent volume), the sensitivity of our method could be increased. From a series of studies in which the sample I I I I I I I Pb Pb + Al Pb + Cu Pb + Mn Pb + Cd Pb + Fe Pb + Zn Fig. 4 Relative signal for 100 yg 1-' Pb samples in the presence of: 20 mg I-' interferents, in A, Tris and B, NH,; and 2.5 mg 1-' interferents in C, NH3. All points are referenced to a Pb only sample, with relative signal = 1.Values of points for Cd and Fe are reduced as follows for clarity: a, values divided by 10; b, value divided by 5 1 injection volume was increased relative to the reagent injection volume, it was determined that the optimum ratio of sample-to-reagent volume was 2 : 1, which yields a porphyrin to lead ratio of 3 : 1 for a 250 pg 1-I lead standard. The sensitivity decreases at larger ratios as a result of deviation from pseudo-first-order reaction conditions. Fig. 4 shows the results of interference studies as relative signal for a 100 pg 1-1 lead sample as a function of added interferent in the absence and presence of masking agent. The Tris buffer provides no masking for interfering ions, whereas NH3-NH4CI buffer masks most of the interfering ions studied.It was observed that Al, Cd, Cu, Mn and Zn are all masked very effectively by NH3 when they are present in S125-, 2 5 , 2 5 , 200- and 200-fold excess, respectively. The only interfer- ing ion that ammonia does not mask is iron. Eight different masking agents were evaluated for their effectiveness at masking iron: acetylacetone, ascorbate, citrate, glycerol, oxalate, salicylate, sulfamate and thiourea. A solution of 1% ascorbate in pH 9.4 ammonia buffer is yellow and was not tested further as the signal from the blank was found to double. The remaining masking agents were tested by observing the increase in blank signal and by comparing the relative signals generated by a 100 pg I-' lead standard after reaction with T(4-SP)P, in the absence and presence of 1.0 mg 1-1 of iron.Fig. 5 shows the results of these tests. Salicylate and citrate appear to react with lead, as evidenced by the huge decrease in relative signal compared with when no masking agent was present. Glycerol, oxalate, thiourea, sulfamate and acetylacetone showed promise as masking agents because they did not react significantly with lead, an important consideration. It was found, however, that use of glycerol, thiourea or oxalate increased the size of the blank due to heavy metal impurities present in these compounds. Of those potential masking reagents tested, acetylacetone at a concentration of 1% v/v in ammonia buffer (pH 9.4) is the best choice for masking iron in the presence of lead, As iron forms a slightly coloured complex with acetylacetone with a molar absorptivity E = 1.01 x 102 1 mol-1 cm-l at 480 nm, the value of the blank is increased considerably and it is necessary to use the method of standard additions for the analyses in order to eliminate completely the effects of interference by iron.It has been suggested that use of a substituted form of acetylacetone may provide less background interference and thus eliminate the need for standard additions, although we have not yet pursued this suggestion. The recoveries for the spiked tap water samples are given in Table 1. Interferent levels in the tap water samples, deter- mined by ETAAS, ranged from 0 (Cd) to 0.3 mg 1-' (Fe). The results of the T(4-SP)P analyses for the tap water samples are shown as a regression line against the results for the same samples analysed by ETAAS (Fig.6). The data from the recovery studies are included on the same plot. It was observed that, on average, the precision of the proposed method is worse than that of ETAAS by about a factor of two. A concentration of 1.8 pg 1-1 was generated for the blank from three replicate determinations using the proposed method, with a 95% confidence interval of 7 pg I-'. When the regression line of Fig. 6 was calculated to compare ETAAS with the proposed T(4-SP)P merging zones FI method, a slope Table 1 Recoveries for spiked tap water determined using T(4-SP)P and the method of standard additions Nominal Mean Mean concentration Confidence concentration/ concentration found - unspiked interval Sample ygl-' found/pg 1- tap/pg 1- (95%)/pg 1-' Recovery (%) Unspiked tap water 23.7* 1 26.0 2 42.0 3 65.0 4 89.0 * Determined by ETAAS.25.9 53.3 68.7 91.6 112.8 - 27.4 42.8 65.7 86.9 3.0 7.0 6.2 6.0 2.8 109.3 105.4 101.9 101.1 97.6936 ANALYST, JULY 1993, VOL. 118 / n r I I -1 50 100 150 Pb concentration by ETAAS/pg I-' Fig. 6 Regression line of the results of ETAAS and T(4-SP)P merging zones FI. A, Unspiked samples; B, spiked samples. The error bars represent 95% confidence levels for each method. Slope = 1.04 -t 0.05; intercept = 2.6 k 4.5; and r2 = 0.997 of 1.04, an intercept of 2.6 and r2 = 0.997 were obtained with 95% confidence intervals of 0.10 and 9 for slope and intercept , respectively; the slope is not significantly different from 1.0 and the intercept is not significantly different from zero.The method detection limit (MDL)24 for the proposed method was determined in the absence and presence of 0.5 mg 1-l of iron. When iron is not present in the sample matrix, lead determinations can be made from calibration graphs yielding an MDL of 4.2 pg l - l , which for a 150 p1 injection gives an absolute detection limit of 0.6 ng. When iron is present in the sample matrix, the addition of acetylacetone to the reagent-buffer mixture and use of the method of standard additions is required for an accurate lead determination. The MDL in this instance is 10 pg I-'. Conclusions It has been demonstrated that use of the porphyrin T(4-SP)P as a spectrophotometric reagent in a merging zones FI manifold is suitable for the determination of lead in tap water in the presence of interfering heavy metal ions.It has also been shown that the use of the method of standard additions for the analyses is appropriate when high concentrations of iron are present. The method bridges the gap between ETAAS and flame AAS, being useful for concentrations down to 10 pg 1-l. Sampling rates for the method are approximately 20 h-l compared with approximately 6 h-1 for ETAAS, although flame AAS sampling rates are much higher. The major advantages of the method are its ease of use and the lack of a need for a preconcentration step. We believe this method has the potential to be a cost-effective substitute for AAS in laboratories in which an atomic absorption instrument is not available. The authors thank the Department of Chemistry at Dart- mouth College, NH, for financial support of this research project.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 References US Fed. Regist., 1991, 56, 26460. RfiiiCka, J.. and Hansen, E. H., Anal. Chim. Acta, 1975, 78, 145. Stewart, K. K., Beecher, G. R., and Hare, P. E., Fed. Proc. Fed. Am. SOC. Exp. Biol., 1974, 33. 1439. Bysouth, S. R., Tyson, J. F., and Stockwell, P. B., Analyst, 1990, 115, 571. Fang, Z.. Xu, S . , and Zhang, S . , Anal. Chim. Acta, 1984, 164, 41. Hu, A. , Dessy, R. E., and Graneli, A., Anal. Chem., 1983,55, 320. Kuban, V., and Bulawa, R., Collect. Czech. Chem. Commun., 1989, 54, 2674. Zhang, Y., Riby, P., Cox, A. G., McLeod, C. W., Date, A. R., and Cheung, Y. Y., Analyst, 1988, 113, 125. Zolotov, Y. A., Shpigun, L. K . , Kolotyrkina, I. Y., Novikov, E. A . , and Bazanova, 0. V., Anal. Chirn. Acta, 1987,200,21. Almestrand, L., Betti, M., Hua, C., Jagner, D., and Renman, L., Anal. Chim. Acta, 1988, 209, 339. Dean, J. R., Ebdon, L., Crews, H. M., and Massey, R. C., J . Anal. At. Spectrom., 1988. 3, 349. Nygren, O., Nilsson, C.-A., and Gustavsson, A., Analyst, 1988, 113, 591. Taylor, C. G., and Trevaskis, J. M., Anal. Chim. Acta, 1986, 179, 491. Bengtsson, M., and Johansson, G., Anal. Chim. A d a , 1984, 158, 147. Tougas, T. P., and Yuan, C. Y., Anal. Chim. Acta, 1987, 192, 327. Zhou, N., Frech, W., and Lundberg, E., Anal. Chim. Acta, 1983, 153, 23. Klinghoffer, O., RfiiiEka, J., and Hansen, E. H., Talanta, 1980, 27, 169. Calatayud, J. M., Mico, A. R., and Camplco, P., Anal. Lett., 1987, 20, 1379. Mortatti, J., Krug, I;. J., and Bergamin, H., Energ. Nucl. Agric., 1982, 4, 82. Novikov, E. A., Shpigun, L. K., and Zolotov, Yu. A., Anal. Chim. Acta, 1990, 230, 157. Igarashi, S . , and Yotsuyanagi, T., Chem. Lett., 1984, 1871. Karlberg, B., and Paeey, G. E . , Flow Injection Analysis-A Practical Guide, Elsevier, Amsterdam, 1989. Johnson, K. S . , and Petty, R. L., Anal. Chem., 1982,54,1185. Standard Methods for the Examination of Water and Wastewater, eds. Clesceri, L. S., Greenberg, A. E . , and Trussell, R. R., American Public Health Association, Washington, DC, 17th edn., 1989. Paper 2105274A Received October 1, 1992 Accepted December 14, 1992

ISSN:0003-2654    

DOI:10.1039/AN9931800933

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 937 Enzymic Micro-determination of Bronopol and Its Application to Pharmaceuticals Asis K. Sanyal, Bhabadeb Chowdhury and Arun B. Banerjee University College of Scienck, Department of Biochemistry, Microbiology Laboratory, 35 Ballygunge Circular Road, Calcutta-700 0 19, India A procedure is described for the rapid micro-determination of bronopol, based on the measurement of inhibition of a thiol protease, papain. The inhibition follows a linear relationship with concentration in the range 0.8-4.0 pg ml-l. The method utilizes the property of bronopol that is primarily responsible for its antimicrobial activity (ability to form disulfide bonds) and hence is also indicative of its antimicrobial potency. The method showed sensitivity towards the degradation of bronopol and can, therefore, be used in stability analysis.Results for the determination of bronopol in the presence of various pharmaceutical ingredients and also in the presence of much higher amounts of some antibiotics and anti-histamine and anti-inflammatory drugs confirmed the validity of the method in the determination of bronopol in complex pharmaceuticals. Keywords: Micro-determination of bronopol; inhibition of papain; liquid oral formulations; spectrophotometry Bronopol (2-bromo-2-nitropropane- 1,3-diol) is a broad-spec- trum antibacterial active against many gram positive and gram negative bacteria. It is important for its high activity against Pseudomonas aeruginosa, which is resistant to many antibac- terial agents. For the wide spectrum of activity and low toxicity on oral administration, bronopol was recommended as suitable for use in the preservation of oral medicaments.1 The compound had been officially accepted as a preservative for pharmaceutical use.2 Increasing use of bronopol in liquid oral formulations was probably because of its high activity over a wide pH range (5- 8).1 This property is not shared by other common preserva- tives such as benzoic acid and methyl p-hydroxybenzoate, which are less efficient antibacterial agents and are only effective at acidic pH. As a preservative, bronopol is present in liquid oral preparations in the presence of much higher amounts of active drugs and various pharmaceutical additives. A sensitive and selective method for the determination of this preservative in complex formulations is, therefore, highly desirable.The official analytical method2 described for the raw material is not suitable for the determination of bronopol present in low concentrations in complex liquid formulations. The mode of action of bronopol was previously investigated by Stretton and Manson.3 The inhibitory property of the compound against bacteria is primarily due to the inactivation of vital thiol proteins of bacterial cells caused by the formation of disulfide bridges. We observed that papain, a proteolytic enzyme with an active thiol group,4 is highly susceptible, and was inhibited at a very low concentration of bronopol. In this paper, a sensitive method for the determination of bronopol, based on the measurement of papain inhibition under standardized conditions, is described.The results of recovery experiments on various pharmaceutical ingredients demon- strated the applicability of the method to complex phar- maceuticals. Experimental Reagents and Samples All chemicals were of analytical-reagent grade and redistilled water was used throughout. Tris-HCl buffer (0.2 mol I-*; pH 8). Tris(hydroxymethy1)- methylamine (Tris) (4.85 g) was dissolved in 100 ml of water, the solution was adjusted to pH 8 with HC1, and the volume was made up to 200 ml with water. Substrate solution. An aqueous solution of azoalbumin (Sigma, St. Louis, MO, USA) of concentration 59.4 mg ml-l was prepared in Tris-HC1 buffer. Trichloroacetic acid (TCA) solution. A 5% TCA solution was used. Formate-EDTA solution.To 10 ml of formic acid (85%), 50 ml of 4 moll-' NaOH were added slowly, with stirring, under ice-cold conditions. Ethylenediaminetetraacetic acid (EDTA) (745 mg) was then added and dissolved by gentle heating. The pH of the solution was adjusted to 5.5, and the volume was made up to 100 ml with water. Enzyme solution. A solution of papain (Fluka, Buchs, Switzerland) of concentration 1.5 mg ml-* was prepared in formate-EDTA solution. The enzyme activity in the solution was so adjusted that the hydrolysis product generated (from azoalbumin) had an absorbance of 0.7 at 390 nm under the defined conditions of enzyme assay described below. This enzyme solution was suitable for use up to 48 h if stored at -20°C. Assay of proteolytic activity of the enzyme solution The incubation mixture contained substrate solution, 0.3 ml; water, 0.1 ml; and enzyme solution, 0.2 ml.The reaction was carried out for 15 min at 37 "C and stopped by the addition of 2 ml of TCA solution. The mixture was shaken on a vortex mixer, and the precipitate was removed by centrifugation at 3000g for 15 min. The absorbance of the hydrolysis product generated in the supernatant phase was measured at 390 nm (absorption maximum of azoalbumin in water) against a reaction blank prepared as above except that TCA solution was added before the addition of the enzyme. Preparation of standard bronopol solution About 30 mg of bronopol (Aldrich, Milwaukee, WI, USA) was accurately weighed and dissolved in 500 ml of water. This solution served as the stock standard and was stable for at least 1 month when stored in a refrigerator ( 5 "C).Preparation of Samples for Recovery Experiments on Pharmaceutical Ingredients The desired amount of pharmaceutical ingredient(s) to be analysed was added to 10 ml of stock standard solution (60 pg ml-l), and the mixture was shaken for 2 h on a rotary shaker at 40 "C to allow for maximum dissolution of the added ingredient(s). The volume of the solution was made up to 50 ml with water and mixed well. Undissolved material (if any) was removed by centrifugation at 3000g for 15 min. The clear supernatant phase was used in the assay.938 1 u 5 ; ANALYST, JULY 1993, VOL. 118 I I I Preparation of Calibration Graph and Assay of Bronopol in Samples A stock standard solution of bronopol (10 ml) was diluted to 50 ml with water.Various volumes (0.02,0.04,0.06,0.08 and 0.10 ml) of the diluted solution were placed in five test-tubes, and each was diluted to 0.1 ml *with water. The tubes were cooled in an ice-bath, 0.2 ml of ice-cold enzyme solution was then placed in each tube, and the contents were mixed well. Inhibition was carried out by incubating each mixture at 40 "C in a water-bath for 15 min. The tubes were then cooled immediately in an ice-bath to stop the reaction. In order to measure the residual enzyme activity, 0.3 ml of ice-cold substrate solution was added to the solution in each tube, the contents were mixed on a vortex mixer, and each reaction mixture was incubated at 37 "C for 15 min in a water-bath. The reaction was terminated by cooling the tubes immediately in an ice-bath, and the residual material, of high relative molecular mass, was precipitated by the addition of 2 ml of TCA solution.The precipitate was removed by centrifugation at 3000g for 15 min. In order to deal with the samples, 0.08 ml of sample solution was mixed with 0.02 ml of water (in duplicate tubes), cooled in an ice-bath and then treated with 0.2 ml of enzyme solution at 40 "C for 15 min. Subsequent processing was exactly the same as described for the standards. An uninhibited control was run in parallel, in the same manner, by using 0.1 ml of water instead of standard or sample solutions. Measurement of Absorbance and Determination of Bronopol From the Inhibition of Papain The absorbance of the supernatant phase obtained after TCA precipitation was measured for standards, samples and uninhibited control at 390 nm against a reaction blank (a clear supernatant phase obtained by centrifugation of a well-shaken mixture of 0.3 ml of water, 0.3 ml of substrate and 2 ml of TCA solution).A separate reaction blank must be run for each analysis (with 0.08 ml of sample solution) for coloured samples. A Hitachi Model U-3210 spectrophotometer (Hitachi, Tokyo, Japan) was used for the measurement of absorbance. The difference between the absorbance (at 390 nm) reading for the experiment, i . e . , with inhibitor (standard or sample) and that for the control (without inhibitor) was the measure of inhibition. The inhibition obtained for different standard dilutions was plotted against the respective bronopol concen- trations during the inhibition reaction.Bronopol in the sample was determined by interpolation from the linear portion of the calibration graph. The use of low-grade enzyme in the assay might cause changes in the assay sensitivity and in the nature of the inhibition-concentration curve. Results and Discussion Bronopol, a potent antibacterial compound, interacts with essential thiol proteins of bacterial cells by the formation of disulfide bridges.3 This property can be utilized with advant- age to detect and determine micro-amounts of bronopol selectively in complex systems such as pharmaceutical formu- lations, where it is nowadays used widely as a preservative. Inhibition of papain, which contains an active thiol group, showed promising potentialities in our attempt to determine bronopol by measuring its effect on thiol proteins.Cysteine, which is normally used as an activator of this enzyme, could not be used during the study of bronopol inhibition owing to the incompatibility of this amino acid with bronopol.1.3 While searching for a suitable activator, we observed that a strong reducing agent (formate) could serve as a papain activator, allowing bronopol to exert its inhibitory effect on the enzyme satisfactorily at a very low concentration. The inhibition was pronounced under slightly acidic conditions (pH 5.5), which were also reported to be most favourable for the antimicrobial 0.40 1 .c E I I 0.20 20 40 60 TemperaturePC Fig. 1 Effect of temperature on the inhibition of papain by bronopol. The inhibition was carried out at different temperatures and measured as described in the text ;, 0.40 I I 0 1.0 2.0 3.0 4.0 B ro no pol concent rat ion/pg m I - 1 Fig.3 Calibration graph for bronopol. Each point represents the mean of three determinations activity and stability of bronopol in aqueous solution. 1 In the presence of high concentrations of strong ultraviolet (UV)- absorbing compounds (present in pharmaceutical formulae), use of casein as the substrate (widely used in the assay of papain) is not possible as this necessitates measuring the absorbance of the hydrolysis product in the short-wave UV region (280 nm).4 Azoalbumin was found suitable for this purpose as, with this substrate, the hydrolysis product can be determined conveniently at 390 nm in the presence of high concentrations of UV-absorbing compounds.The hydrolysis of azoalbumin was measured by the method described earlier,s after certain modifications. Temperature had little effect on the enzyme inhibition between 30 and 50°C (Fig. 1). However, the inhibition was affected at higher temperature (60°C). The time course of inhibition (Fig. 2) indicated attainment of a maximum within 10 min. Under the standardized conditions of pH ( 5 . 3 , temperature (40°C) and time of incubation (15 min), the inhibition of papain showed a significant linear relationship with concentrations of bronopol in the range 0.84.0 pg ml-1 (Fig. 3; r = 0.9986; p t0.05). Recoveries of bronopol from authentic samples demonstrated satisfactory accuracy and precision for the proposed method (Table 1).ANALYST, JULY 1993, VOL.118 939 Table 1 Recovery of bronopol from authentic samples Experiment Addedmg per 50 ml* Recovery+ (YO) 0.4 97.8 0.5 97.2 0.6 ' 101.0 0.7 102.4 0.8 99.1 Average f standard deviation 99.5 f 2.2 * Solutions containing the stated amount of bronopol (pure bronopol used as the reference standard) represent authentic samples for the recovery studies. Processing of the samples for the determina- tion of bronopol is described under Experimental in the text. t Each recovery is the average of four determinations. Table 2 Recovery of bronopol from pharmaceuticals Added per 50 ml water* Ingredient mixture+/ BronopoV Bronopol drug mg recovered/mg* Recovery (YO)* MU37.9 mg M2/30.5 mg M3/0.02 ml Ampicillin/5 mg Cephalexin/S mg Erythromycin estolateh mg Nalidixic acid/5 mg Metronidazole/S mg Chlorpheniramine maleate/5 mg Paracetamol/S mg Ibuprofen/S mg 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.61 k 0.02 0.60 -C 0.02 0.59 f 0.02 0.58 -I- 0.02 0.61 k 0.02 0.60 f 0.02 0.55 f 0.02 0.60 f 0.02 0.60 f 0.02 0.60 k 0.02 0.59 f 0.02 102.0 f 3.2 100.5 f 3.5 98.2 f 3.0 96.9 f 2.8 101.1 f 3.4 100.7 f 3.0 92.2 f 3.2 100.5 f 3.3 100.8 f 3.6 99.6 f 3.1 98.8 f 3.3 * The processing of the mixtures and subsequent determination of bronopol is described under Experimental in the text.f Composition of the ingredient mixture-M1 : carboxymethylcellu- lose (5 mg); gelatin (5 mg); povidone ( 5 mg) bentonite ( 5 mg); sorbitol ( 5 mg); alginic acid ( 5 mg); silicon dioxide ( 5 mg); methyl paraben (0.63 mg); propyl paraben (0.63 mg); benzoic acid (0.63 mg); Sunset Yellow (0.13 mg); tartrazine (0.13 mg); carmoisine (0.13 mg); pineapple (0.31 mg); vanillin (0.31 mg).M2: sucrose ( 5 mg); agar ( 5 mg); acacia ( 5 mg); D-glucose (5 mg); aspartame (0.5 mg); mannitol(5 mg); fructose ( 5 mg). M3: Peppermint oil (0.02 ml); ice cream soda (concentrate, 0.02 ml); strawberry (concentrated juice, 0.02 ml); raspberry (concentrated juice, 0.02 ml); ethyl alcohol (95%, 0.1 ml). * Each result is the mean of five independent determinations f standard deviation. r 99.80 - $ 79.84 I - 0 Q 59.88 2 I) - ; 39.92 2 e, 19.96 0 24 48 72 Time/h Fig. 4 Response of the proposed enz mic method towards degrada- tion of bronopol. Bronopol solution 62 yg ml-*) prepared in Tris- HCI buffer (0.05 mol I-I; pH 8) was heated at 60°C for different time intervals.Bronopol in the solution was determined as described in the text syrups and suspensions. Hence, the proposed method might find important applications in the determination of bronopol in complex pharmaceuticals. The proposed enzymic method could be used to detect bronopol degradation in solutions buffered at pH >5, as reported previously by Croshaw et aZ.1 A marked lowering in bronopol concentration was observed when a solution buf- fered at pH 8 was heated at 60°C (Fig. 4). The assay results obtained by the proposed method are likely to reflect the antibacterial potency of bronopol as the determination is based on the ability of the compound to form disulfide bridges, which is known to be the primary mode of action of bronopol against bacteria. This enzymic assay is, therefore , particularly useful for evaluating bronopol potency in antibiotic preparations, where the assessment of antimicro- bial activity of this preservative is difficult by microbiological techniques owing to the presence of broad spectrum anti- biotics.A. K. S. is grateful to Professor A. K. Bhattacharya, Head of the Department of Biochemistry, Calcutta University, for his encouragement during the course of this work. A. K. S. also thanks Smith Stanistreet Pharmaceuticals, Calcutta, for co- operating in this research programme. Recovery experiments were performed on various exci- pients, preservatives and other additives used in liquid oral formulations in order to show the applicability of this enzymic method to complex pharmaceuticals. As the concentrations of different excipients and other additives might be different in different commercial formulations, their amounts in the mixtures prepared for recovery experiments were selected on the basis of the type of material taken. The results show satisfactory recovery of bronopol from various preservatives and suspending, colouring, sweetening and flavouring agents commonly present in liquid oral formulations (Table 2). Good recovery of bronopol was also observed in the presence of much higher amounts of some important antibiotics and anti- histamine and anti-inflammatory drugs widely marketed as References Croshaw, B., Groves, M. J., and Lessel, B., J . Pharm. Pharmacol., 1964, 16, 127T. British Pharmacopoeia, HM Stationery Office, London, 1988, Stretton, R. J . , and Manson, T. W., J . Appl. Bacteriol., 1973, 36, 61. Arnon, R., in Methods in Enzymology, eds. Perlmann, G. E., and Lorand, L., Academic Press, New York and London, 1970, vol. XIX, p. 226. Sanyal, A. K., Das, S. K . , and Banerjee, A. B., Sabouraudia, 1985,23, 165. vof. I, p. 80. Paper 2106321 B Received November 26, 1992 Accepted February 2, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800937

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 94 1 Simple and Rapid Titrimetric Method for the Determination of Metoxuron in Commercial Formulations Balbir C. Verma, Devender K. Sharma, Amita Pandey and Hari K. Thakur Department of Chemistry, Wimachal Pradesh University, Summer Hill, Shimla- 171 005, India A simple and rapid titrimetric method for the determination of metoxuron, based on its oxidation with cerium(iv) in sulfuric acid medium, is described. The method permits detection of the end-point visually, potentiometrically or spectrophotometrically. The proposed method is accurate to +.0.6% with a relative standard deviation of 0.8% and is recommended for the routine determination of this herbicide in its co m mercia I form u I at ions. Keywords: Metoxuron determination; cerium(iv); visual and potentiometric titrations; spectrophotometric titrations; formulation analysis Metoxuron, N'-(3-chloro-4-methoxyphenyl)-N,N-dimethyl- urea (l), is a selective herbicide for post-emergence use in winter-sown varieties of wheat and barley, and for pre- and post-emergence use in carrots.It is also used for the control of annual grasses, blackgrass, loose-silky bentgrass, wild oats, rye-grass, meadow-grass and the main broad-leaved weeds.l-3 In view of the wide herbicidal use of this compound, a simple, rapid and reliable method for its determination in commercial formulations is highly desirable. The method4 commonly used for this purpose consists in separating the compound from the formulation with dichloromethane and extracting this solution with hydrochloric acid to remove interfering amines, followed by hydrolysis with potassium hydroxide-propane-1,3-diol in a specially designed distilling apparatus to liberate dimethyl- amine.The amine is adsorbed into boric acid solution and titrated with standard hydrochloric acid. The method, in addition to requiring special apparatus, is tedious and time consuming. cC (1) A simple and rapid method has been developed for the determination of metoxuron in its formulations. The method is based on the direct oxidimetric titration of metoxuron with cerium(rv) in sulfuric acid medium. The end-point can be detected visually (with ferroin as indicator), potentiometric- ally (with platinum-saturated calomel electrodes) or spectro- photometrically at 340 nm [cerium(Iv) solutions are orange]. Experimental Reagents Dirnethylforrnamide.Dime th ylformamide (BDH, Poole , Dorset, UK) was purified by storing over analytical-reagent grade anhydrous sodium carbonate for 2 d. The solvent was decanted off and distilled, and the fraction distilling at 148.5- 149 "C was collected in brown glass bottles to exclude light Ammonium hexanitratocerate(rv) solution (0.04 mol 1- l) in 0.5 mol 1-1 sulfuric acid. Prepared by dissolving slightly more than the calculated amount of the dried analytical-reagent grade compound in 0.5 mol I-' sulfuric acid and standardized by titration with ammonium iron(n) sulfate in sulfuric acid medium, with ferroin as indicator. Ferroin solution in water, 0.01 mol 1-l. Prepared by dissolving 0.594 g of 1,lO-phenanthroline and 0.278 g of analytical-reagent grade iron(I1) sulfate heptahydrate in 100 ml of water.Metoxuron. The standard material was supplied by courtesy of the US Environmental Protection Agency, Research Triangle Park, NC, USA. Apparatus Potentiometric titrations were performed with a Toshniwal (Delhi, India) potentiometer equipped with a platinum and saturated calomel electrode assembly. Spectrophotometric measurements were performed with a Bausch & Lomb (Rochester, NY, USA) spectrophotometer. Borosilicate glass microburettes graduated to 0.01 ml were used. Procedures Visual and potentiometric titrations Aliquots (0.5-2.5 ml) of dimethylformamide solutions of the pure compound were placed in titration vessels and diluted to 5 ml with dimethylformamide. Each solution was mixed with 20 ml of 0.5 mol 1-l sulfuric acid, 1 or 2 drops of ferroin indicator (for visual titrations) and titrated at room tempera- ture (approximately 24 "C) with 0.04 mol 1-1 ammonium hexanitratocerate(1v) visually or potentiometrically.In the Table 1 Determination of metoxuron with ammonium hexanitratocerate(1v): visual and potentiometric titrations Metoxuron found*/mg Visual method Potentiometric method Metoxuron Mean titration taken/mg value/ml Found 2.0 0.442 k 0.001 2.02 f 0.01 4.0 0.869 f 0.003 3.97 f 0.03 6.0 1.298 f 0.002 5.94 f 0.02 8.0 1.759 4 0.008 8.04 f 0.04 10.0 2.170 f 0.009 9.92 f 0.06 Mean titration 0.439 f 0.001 0.878 f 0.002 1.300 f 0.009 1.755 k 0.005 2.175 f 0.010 value/ml Found 2.01 f 0.01 4.02 f 0.02 5.95 f 0.03 8.02 f 0.03 9.94 f 0.05 * Mean of ten determinations f standard deviation (SD).942 ANALYST, JULY 1993, VOL.118 visual titrations, the end-point was marked by a sharp colour change from red to light blue. In potentiometric titrations, a sharp break in potential was observed at the equivalence point. The results are presented in Table 1. Spectrophotometric titrations Aliquots (0.2-1 .O ml) of dimethylformamide solutions of the pure compound were diluted to 2 ml with dimethylformamide, mixed with 5 ml of 0.5 mol 1-1 sulfuric acid and titrated at room temperature (approximately 24 “C), at 340 nm, with the standard oxidant. For plotting the titration curves, the absorbance values were corrected to initial volume by multiplying the absorbance reading by a factor V + v / V , where Vis the initial volume, and v is the volume of cerium(rv) added for a particular reading.A plot of absorbance versus volume of titrant (ml) was established and the best straight lines were drawn between points taken well before and after the equivalence point. The intersection of the linear segments was taken as the end-point. The results are presented in Table 2. Formulation analysis A wettable powder formulation ‘Dosanex’ , containing 80% of the active ingredient, was used. A single large sample of the formulation (600 mg for visual and potentiometric titrations Table 2 Determination of metoxuron with ammonium hexanitrato- cerate(1v): spectrophotometric titrations Mean titration Metoxuron Metoxuron taken/pg value/ml found*/pg 20.0 0.87 k 0.005 19.9 rt 0.1 40.0 1.74 rt 0.010 39.7 f 0.2 60.0 2.65 k 0.018 60.4 f 0.4 80.0 3.54 rt 0.023 80.6 rt 0.5 100.0 4.34 rt 0.022 99.0 f 0.6 * Mean of five detcrminations k SD.and 30 mg for spectrophotometric titrations) was weighed, shaken with 50 ml of dimethylformamide, and the solution was filtered. The residue (if any) was washed two or three times with dimethylformamide. The filtrate, plus washings, was diluted to a known volume with the same solvent. In order to check the accuracy of the method, known amounts of pure metoxuron were also added to the weighed formulation before extraction. Aliquots were then taken for titrations. The visual, potentiometric and spectrophotometric titrations were per- formed as described above. The results are presented in Tables 3 and 4. Results and Discussion Ammonium hexanitratocerate(1v) is a versatile oxidimetric reagent.The orange colour of the reagent, coupled with its properties of exhibiting a high oxidation potential and excellent solution stability, prompted us to use this reagent for the determination of metoxuron. In potentiometric titrations a sharp jump in potential of the order of 350-400 mV is observed at the equivalence point with the addition of 0.05 ml of 0.04 moll-’ oxidant solution. The results recorded in Table 1 show that metoxuron in the range 2-10 mg can be determined by visual and potentiometric titrations with a maximum relative standard deviation (RSD) of 0.8%. The spectrophotometric titrations were monitored at 340 nm, the A,,, of the reagent. The absorbance remains virtually zero until the equivalence point is reached.Thereafter, it increases linearly as the excess of reagent is added. With spectropho- tometric titrations, the compound could be determined in the range 20-100 pg with a maximum RSD of 0.6% (Table 2). For the commercial formulation, visual and potentiometric titra- tion procedures yielded values corresponding to 98.2-99.3% of the nominal content, with RSD values in the range 0.3- 0.5% (Table 3). The spectrophotometric titration procedure yielded 98.1-99.3% of the nominal content with RSD values in the range 0.3-0.4% (Table 4). The results were, however, compared with those of an independent method.4 Table 3 Determination of metoxuron in a commercial formulation nominally containing 80% active ingredient Foundt/mg Nominal amount Metoxuron Visual Recovery Potentiometric Recovery Comparison Recovery 3.00 - 2.97 99.0 -t- 0.4 2.98 99.3 2 0.3 2.96 98.7 k 0.9 5 .oo - 4.93 98.6 -t 0.4 4.94 98.8 k 0.4 4.93 98.6 k 0.6 3.00 2.0 4.94 98.8 k 0.3 4.95 99.0 +.0.3 4.94 98.8 k 0.6 3.00 4.0 6.88 98.3 k 0.5 6.90 98.6 k 0.4 6.90 98.6 k 0.8 98.7 k 0.3 8.84 98.2 k 0.7 3.00 6.0 8.84 98.2 k 0.3 8.88 ta ken*/mg added/ml method (Yo) method (Yo 1 met hod$ (”/) * Based on 80% active ingredient. t Mean k SD of ten determinations. * Ref. 4. Table 4 Determination of metoxuron in a commercial formulation nominally containing 80% active ingredient Fo undt/pg Spectrophoto- metric Nominal amount Metoxuron titration Recovery Comparison Recovery 25.0 - 24.8 99.3 & 0.4 24.8 99.0 f 0.6 75.0 - 73.7 98.3 k 0.4 73.6 98.1 f 0.5 25.0 25.0 49.5 99.0 k 0.3 49.3 98.6 f 0.7 25.0 50.0 74.0 98.7 f 0.4 73.6 98.1 f 0.5 25.0 75 .0 98.2 98.2 ? 0.3 98.1 98.1 ? 0.4 taken*/pg added/pg method (Yo) method* ( Y o ) * Based on 80% active ingredient.t Mean k SD of five determinations. t Ref. 4.ANALYST, JULY 1993, VOL. 118 943 The results indicate that each molecule of metoxuron consumes two equivalents of the reagent. The most plausible mechanism of oxidation is: V C H 3 0 0 N H - C - - N ( C H , ) 2 II 2Ce4+ + 2H20 - / CI 0 I1 2ce3+ + o 0 + (CH3)ZN-C-NHT + CHSOH + 2Hf CI That substituted p-alkoxyanilines, i. e., yield quinones on oxidation is well known.5 Owing to its simplicity and rapidity, the proposed method is recommended for the routine determination of metoxuron in its formulations. Amines, which cause interference in the commonly used method4 and have to be removed by extraction into hydrochloric acid, do not interfere in the proposed method. References Berg, W., 2. Pflanzenkrankh. (Pflazenpathol.) Pflanzenschutz, Sonderh., 1968,4,233. Glenister, M . , and Griffiths, G. P., Proc. of 9th British Weed Control Conf., London, 1968, vol. 1 , p. 46. Griffiths, G. P., and Ummel, E., Proc. of ZUth British Weed Control Conf., London, 1970, vol. 1, pp. 7 and 186; vol. 2, p. 849. Wisson, M., Van Hoek, C., and Saucer, H. H., in Analytical Methods for Pesticides and Plant Growth Regulators, ed. Zweig, G., Academic Press, New York, 1976, voi. VIII, p. 417. Gainelli, G., and Cardillo, G., Chromium Oxidations in Organic Chemistry, Springer-Verlag, Berlin and Heidelberg, 1984, p. 96. Paper 2104741A Received September 3, 1992 Accepted December 14, 1992

ISSN:0003-2654    

DOI:10.1039/AN9931800941

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 945 Catalytic Method for the Determination of Trace Amounts of Mercury in Environmental Samples A. Ramesh Department of Pesticide Chemistry, Fredric Institute of Plant Protection and Toxicolog y, Padappai, Chingelput District, Madras 601 301, India A catalytic method based on the abstraction of CN- from hexacyanoferrate(1i) by mercury(ii) in the presence of 4-(2'-th iazolylazo)resacetop henone oxi me was studied spectrop hotometrical I y. The met hod allows the determination of mercury(ii) in the range 0.014.9 ppm at a wavelength of 420 nm. The method was applied t o the determination of methyl- and ethylmercury chloride and the analysis of environmental water samples. Keywords: Mercury determination; catalysis; 4-(2 '-thiazolylazo)resacetophenone oxime; water analysis Chemical speciation and the accurate determination of mercury and its compounds have been studied by many workersl-5 owing to the highly toxic effects of this element in the environment.The rate of chemical reaction is often used as an analytical tool as it is dependent on the concentration of the reactants or in some instances on the concentration of a catalyst. The measurement of the initial rate has found wide application for reactions catalysed by enzymes. However, when the reaction to be followed is purely inorganic or involves complexes with organic reagents these are allowed to proceed for a fixed time and then quenched. The slow ligand-exchange reaction between hexacyanofer- rate(r1) and different types of organic ligands is catalysed by even small amounts of Hg" ions.6-11 Even though nitroso compounds and amino compounds are extensively used as organic ligands in this context, no report on the use of thiazolylazo compounds could be found.The analytical potential of 4-(2'-thiazoly1azo)resacetophenone oxime for trace metal analysis,12,13 was evaluated in this work and the ligand-exchange reaction was studied in detail for the possible determination of Hg" ions. Experimental 4-(2'-Thiazo1ylazo)resacetophenone oxime (TARPO) was prepared by diazotization of 2-aminothiazole and subsequent coupling of the resulting diazonium salt with 2,4-dihydroxy- acetophenone oxime. The compound was purified by recrys- tallization from ethanol and was characterized by spectral studies. Stock solutions (0.1 moll-I) of TARPO, potassium hexa- cyanoferrate(I1) [K4Fe(CN)6.3H20] (BDH) and mercury(i1) chloride (HgC12) (Merck) were prepared by dissolving the required amounts of the analytical-reagent grade substances in distilled water or distilled water containing 40% v/v dimethylformamide (DMF) .Buffer solutions containing appropriate volumes of 0.1 mol 1-1 sodium acetate and 0.1 mol 1-1 hydrochloric acid (pH 1-3) or 0.1 rnol I-' sodium acetate and 0.1 rnol I-' acetic acid (pH 4-5) were prepared. A Model 106 spectrophotometer (Systronics Instruments, Ahmadabad, India) was employed for absorbance measure- ments. Results and Discussion Addition of TARPO solution to a dilute solution of hexa- cyanoferrate(11) gives a green colour on standing of the test mixture for about 48 h.The reaction is very slow and a long time is required for the development of observable intensity. However, the addition of a small amount of mercury(n) catalyses the reaction strongly and a measurable absorbance was observed after a reaction time of only 15 min. The reaction was therefore studied in detail to develop a suitable procedure for the determination of small amounts of mercury. The green, soluble product formed in the reaction at pH 3.0 absorbs most strongly at 420 nm. The effects of pH, TARPO concentration, K4Fe(CN)6 concentration and reaction time on the absorbance of the experimental solution were therefore studied in order to establish the optimum conditions. The results in Tables 1-3 suggest that the absorbance is maximum at pH 3.0 and increases with increasing concentra- tions of TARPO and K4Fe(CN)6 up to 1 x and 2 x mol I-l, respectively, but subsequently remains un- changed.From these results, the optimum concentrations of TARPO and K4Fe(CN)6 were fixed at the above levels and the effect of mercury(r1) concentration was studied. Determination of Mercury(Vii) Known aliquots of mercury(r1) solution were placed in a series of 10 ml calibrated flasks, 4 ml of buffer solution (pH 3.0), 1 ml of DMF, 1 ml of 1 X mol 1-1 TARPO solution and 1 ml of 2 x mol 1-1 K4Fe(CN)6 solution were added and the contents were diluted to volume with distilled water. The Table 1 Effect of pH on absorbance. Conditions: [&Fe(CN),] = 2 x l0W4 mol l - l , [TARPO] = 1 x rnol 1-l, [mercury(ir)] = 1 x lo-" rnol l-l, [DMF] = 30% v/v, h,,, = 420 nm Absorbance PH 30 min 60 min 1 .0 0.15 0.18 2.0 0.23 0.34 3.0 0.42 0.51 4.0 0.31 0.42 5 .o 0.12 0.28 Table 2 Effect of TAKPO concentration on absorbance.Conditions: [K4Fe(CN),] = 2 x [DMF] = 30% v/v, pH = 3.0, h,,, = 420 nm rnol 1 - I , [mercury(ii)] = 1 X lo-" rnol [TARPOI/ lO-~rnoll-~ 0.2 0.4 0.6 0.8 1 .0 1.2 1.4 1.6 Absorbance 30 min 0.07 0.23 0.36 0.40 0.42 0.42 0.42 0.42 60 min 0.14 0.28 0.39 0.48 0.51 0.51 0.51 0.51946 ANALYST, JULY 1993, VOL. 118 Table 3 Effect of K4Fe(CN)6 concentration on absorbance. Condi- tions: [TARPO] = 1 X 10-5mo11-1, [mercury(i~)] = 1 X mol I F 1 , [DMF] = 30% v/v, pH = 3.0, A,,, = 420 nm [K4Fe(CN)61/ - moll-' 0.4 0.8 1.2 1.6 2.0 2.4 2.8 Absorbance 30 mi? 0.08 0.17 0.26 0.34 0.42 0.42 0.42 60 min 0.21 0.28 0.33 0.42 0.51 0.51 0.51 Table 4 Determination of mercury(1r).Conditions: [TARPO] = 1 X mol l-l, [DMF] = 30% v/v, pH = 3.0, A,,, = 420 nm mol l-l, [bFe(CN),] = 2 X Amount of mercury(r1) (ppm) Added 0.050 0.100 0.200 0.300 0.500 0.700 0.900 Found* 0.048 0.097 0.195 0.296 0.500 0.702 0.905 * Average of six determinations. Relative error (%) -4.00 -3.00 -2.50 -1.30 0 +0.28 +0.55 Relative standard deviation (%) 2.68 2.13 1.68 1.47 0.93 1.21 1.57 Table 5 Effect of foreign ions Tolerance Ion added limit (ppm) Pb", W"' 120 Pt'" 80 TI"', Ball 70 BilIl 60 Zn" 50 Mn" 40 Tolerance Ion added limit (ppm) Cd", Ni" 20 S2- 60 ~ s 0 ~ 3 - 30 Br- 20 I- 15 so42- 10 time for addition of half the K4Fe(CN)6 solution was taken as the zero time. The absorbance of the solution was measured at 30 and 60 min against a reagent blank.A calibration graph was constructed by plotting the measured absorbance versus the amount of metal ion. The calibration graph was linear over the range 0.01-0.50 pprn, suggesting that mercury(I1) can be determined in this range. The amount of Hg" ions in unknown solutions was determined using this calibration graph. The results are given in Table 4. Effect of Foreign Ions The tolerance limits of foreign ions in the determination of mercury(I1) were established by studying the effect of these ions on the absorbance of the experimental solution. The tolerance limit was taken as the maximum amount of the foreign ion that causes an error of less than f2% in the absorbance of the experimental solution containing 0.2 pprn Hg" ions.The results are presented in Table 5. Determination of Mercury in Spiked Water Samples Spiked water samples containing different mercury com- pounds [mercury( 11) chloride , methylmercury chloride and ethylmercury chloride] were analysed after adjusting the pH of the solutions as described earlier. The results are presented in Table 6. The validity of the values for mercury(I1) chloride was checked using the dithizonate method.14 Analysis of Environmental Water Samples Different water samples from sites with potential mercury pollution were collected in polyethylene containers, concen- Table 6 Determination of mercury in spiked water samples Amount Compound added/ng Amount found/ng* Mercury(i1) chloride 200 100 50 Methylmercury chloride 180 100 50 Ethylmercury chloride 150 100 50 198.8 k 0.34 (198.6) 99.33 t 0.41 (99.48) 49.86 t 0.26 178.94 t 0.28 99.64 k 0.33 49.21 t 0.41 149.37 k 0.47 99.18 k 0.28 49.36 t 0.26 (49.55) * Means k standard deviations for five determinations.Values in parentheses were determined by the dithizonate method. Table 7 Analysis of environmental water samples Location Mercury found/ng* Nellore 1.6 f 0.23 Tada 1.2 f 0.31 Renigunta 1.9 k 0.28 Guntakal 2.3 k 0.35 * Means t standard deviations for five determinations. trated, filtered and adjusted to the required pH. Mercury was determined as described earlier and the results are presented in Table 7. Conclusions The results show that the catalysed reaction is highly sensitive, accurate and allows the determination of mercury in polluted water at very low levels.The author is grateful to the University Grants Commission, New Delhi, India, for providing a Junior Research Fellowship and J.N.T. University, Hyderabad, India, for the necessary facilities. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 References Goto, M., Shibakawa, T., Arita, T., and Ishii, D., Anal. Chim. Acta, 1982, 140, 179. Ahmed, R., May, K., and Stoeppler, M., Fresenius' 2. Anal. Chem., 1987, 326, 510. Horvat. M., May, K . , Stoeppler, M., and Byrne, A. R., Appl. Organomet. Chem., 1988, 2, 515. Horvat, M., Byrne, A. R . , and May, K., Talanta, 1990,37,207. Ramadevi, P., Gangaiah, T., and Naidu, G. R. K., Anal. Chim. Acta, 1991, 249, 533. Asperger, S . , and Pavlovic, D., J. Chem. Soc., 1955, 1449. Raman, S., J. Inorg. Anal. Chem., 1981, 43, 1855. Phull, M., and Nigam, P. C., Talanta, 1981, 28, 591. Phull, M., Bajaj, H. C., and Nigam, P. C., Talanta, 1981, 28, 610. Venkateswarlu, T., and Raman, S . , Indian J. Chem. Sect. A , 1983, 22, 553. Mallikarjuna Rao, K . , Sreenivasulu Reddy, T., and Brahmaji Rao, S., Analyst, 1988, 113, 983. Ramesh, A., Krishnamacharyulu, J . , Ravindranath, L. K., and Brahmaji Rao, S., J . Radioanal. Nucl. Chem., 1991, 154, 357. Ramesh, A., Krishnamacharyulu, J . , Ravindranath, L. K., and Brahmaji Rao, S., Analyst, 1992, 117, 1037. Marczenko, Z., Spectrophotometric Determination of Elementmy, Ellis Horwood, Chichester, 1976, p. 350. Paper 2105615A Received October 21, 1992 Accepted December 21, 1992

ISSN:0003-2654    

DOI:10.1039/AN9931800945

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 947 Use of Elemental and Functional Group Analysis for Monitoring Compositional Changes Occurring on Air Blowing and Accelerated Weathering of a Natural Asphalt Lutchminarine Chatergoon, Robin Whiting and Clayton Smith* Asphalt Research Group, University of the West lndies, St. Augustine, Trinidad, West lndies The composition of a natural asphalt was examined using elemental analysis, functional group analysis and infrared spectrophotometry. Changes in composition resulting from air blowing and accelerated weathering were studied. The natural asphalt was found to have a composition that more closely resembles blown than straight-run petroleum asphalts. The heteroatom content, degree of unsaturation and functional group composition were found to be related to asphalt durability.Keywords: Natural asphalt; air blowing; accelerated weathering; elemental analysis; functional group analysis Asphalts are complex mixtures of organic molecules ranging from non-polar saturated hydrocarbons to highly polar, highly condensed aromatic ring systems. Although the molecules are composed predominantly of carbon and hydrogen, most molecules contain one or more of the heteroatoms nitrogen, sulfur and oxygen. These heteroatoms, particularly nitrogen and oxygen, in combination with the aromatic ring system contribute considerable polarity and polarizability to the molecules, and this leads to strong association forces that in turn affect the physical properties.' Moschopedis and Speight2 have shown that the elemental composition of both natural and petroleum bitumens may be used as a guide to the changes in composition that occur during air blowing of the materials.It is reasonable to expect that changes due to oxidative ageing may also be examined in this way. Compositional changes usually give rise to changes in the functional group chemistry and hence the physical properties of the material. Infrared (IR) spectrophotometry has long been used in the study of major functional groups in organic chemistry. These groups can be determined via differential IR analysis.3" In this paper, we report on the chemical composition of Trinidad lake asphalt (TLA), a natural asphalt that has found widespread use in both road and roofing applications owing to its stability, durability and adhesiveness.It is also used to upgrade the performance of petroleum asphalt binders.7 In an effort to standardize the material, research was undertaken to obtain an American Society for Testing and Materials (ASTM) Type IV material based on TLA. The results from previous unpublished work indicated that the latter can be achieved by blending TLA with petroleum bitumen and flux oil, the mixture being subsequently air blown so as to obtain the final product. The mechanism of the air- blowing process is still in the theoretical stage, but this work gives some insight to two possible pathways. The changes that occurred during accelerated weathering of the material were also examined, and compared with those which occurred in a Venezuelan petroleum roofing bitumen. The possibility of improving the resistance of the natural asphalt composition to weathering by decreasing the level of acidic functional groups was also studied.Experimental Dried and filtered TLA was obtained from Lake Asphalt of Trinidad and Tobago (1978) Ltd. The material consists of 55.4% bitumen, 37.0% mineral matter and 7.6% organic * To whom correspondence should be addressed. components that are not bituminous in nature.8 It has a penetration of 4 (xl/lO mm) and softening point (ring and ball) of 96°C. Samples for air blowing were prepared by blending TLA with a 150 penetration petroleum bitumen from Trinidad and Tobago Oil Co. Ltd. and a suitable flux oil. This feedstock was blown to an ASTM D 312-84 Type IV specification9 and subsequently weathered according to ASTM D 4799-88.10 The Venezuelan bitumen used for comparison was a 180 penetration roofing feedstock from Lagoven (Caracas, Venezuela).The crude asphalt, feedstock, air-blown and weathered samples were fractionated according to modified ASTM D 2172-81 and D 4124-84 methods as decribed elsewhere.8 The former method entails the removal of the inorganic material from TLA and samples that contain TLA. The effect of this inorganic material on the properties of the asphalt has yet to be evaluated, but current work indicates that the inorganics play an important role in the colloidal behaviour of the TLA samples. The ASTM D4124-84 method separates the pure bitumen into four fractions: saturates, naphthene aromatics, polar aromatics and asphaltenes. Elemental analysis (C, H, N, S and 0) was carried out on both the crude samples (Table 1) and the four fractions of each sample (Table 2), using a Perkin-Elmer Model 240C elemental analyser, and infrared spectra were determined using a Perkin-Elmer Model 983 infrared spectrophotometer. Functional group analyses were performed using base hydrolysis and silylation reactions that shifted the principal absorption bands to facilitate easy identification, as described by Petersen and co-workers .3 4 Quantification of the levels of each group was then achieved using differential IR spectro- photometry . 3 4 An investigation of the effects of lowering the free carboxylic acid and acid anhydride levels via lime treatment of the samples was also carried out. This was achieved by blending 0.1% of lime into some of the samples, which were then blown and weathered accordingly.The following is a list of sample codes used in the text: TLA = Trinidad lake asphalt (37% inorganics); TPB = refinery bitumen; TAB = blended TLA (14.8% inorganics); T.IV = blown TAB Type IV specification (ASTM); VPB = Venezuelan petroleum bitu- men; V.IV = blown VPB Type IV specification (ASTM); .W = weathered sample; and .L = lime-treated sample. Results Elemental composition data presented are the means of three determinations, which were normalized to give oxygen values by difference. Table 1 shows that TLA has a similar elemental composition to the refinery bitumens except for a higher sulfur948 ANALYST, JULY 1993, VOL. 118 content (6.2%). The H : C ratio is lower than that of the refinery materials.On air blowing the blend of TLA and TPB, oxygen uptake is observed, but weathering subsequently reduces the oxygen content. For the Venezuelan asphalt, however, there was a decrease in the oxygen content on air blowing. Table 2 shows that the oxygen uptake in the TLA blend occurs mainly in the asphaltene and polar aromatic fractions, whereas weathering leads to a decrease in oxygen mainly in the polar and naphthene aromatic fractions. Trinidad lake asphalt shows an average sulfur content of 3.1% for the different fractions. However, the crude material shows a much higher value of 6.2%, which is due to the existence of elemental sulfur in the inorganic fractions of the material. Although the Venezuelan roofing flux contains an average of 3.6% of total sulfur, only a negligible amount ~ ~~~~ Table 1 Elemental composition of natural asphalt and petroleum bitumen samples before fractionation Elemental composition (%) H : C Sample C H N S O* ratio TLA 82.3 8.8 TPB 83.9 10.6 TAB 83.3 9.9 T.IV 82.7 10.3 T.1V.W 83.8 11.1 VPB 84.1 9.7 V.IV 86.1 8.5 V.1V.W 86.1 8.7 * Oxygen by diffcrcnce.0.6 6.2 0.6 2.7 0.5 4.5 0.4 3.0 0.4 2.4 0.7 3.6 0.6 3.4 0.5 3.3 2.1 1.28 2.2 1.52 1.8 1.43 3.6 1.50 2.3 1.59 1.9 1.38 1.4 1.19 1.4 1.21 Table 2 Elemental composition of fractionated asphalt and bitumen samples Elemental composition (%) - Asphalt Sample fraction C H TLA Asphaltene 82.7 8.7 Polar 82.3 9.0 Naphthene 85.5 8.8 Saturates 90.0 10.0 TPB Asphaltene 81.2 10.8 Polar 83.7 11.1 Naphthene 84.0 9.0 Saturates 86.7 11.5 TAB Asphaltenc 82.0 9.8 Polar 82.1 10.5 Napthene 84.8 8.9 Saturates 88.0 10.8 T.IV Asphaltene 79.6 8.6 Polar 80.4 9.7 Naphthene 84.5 9.9 Saturates 86.0 12.9 T.1V.W Asphaltene 81.3 10.2 Polar 82.7 10.5 Naphthene 85.5 10.1 Saturates 86.0 13.6 VPB Asphaltene 84.4 7.1 Polar 82.8 9.9 Naphthene 82.8 10.1 Saturates 86.5 11.5 V.IV Asphaltene 85.0 7.2 Polar 85.1 8.2 Naphthene 84.2 9.3 Saturates 90.1 9.5 V.1V.W Asphaltene 85.2 7.3 Polar 85.3 8.4 Naphthene 84.0 9.5 Saturates 90.1 9.5 * Oxygen by difference.N 1.5 0.9 0.2 0 1.3 1.1 0 0 1.4 1.0 0.1 0 1.5 0.9 0.3 0 1.4 0.9 0.3 0 1.2 0.7 0.7 0 1 .o 0.6 0.7 0 0.9 0.6 0.7 0 - S 6.1 3.5 2.9 0 5.4 2.1 2.4 0.7 5.6 2.4 2.5 0.5 8.1 2.3 1.4 0.3 5.9 2.0 1.2 0.3 3.8 5.4 4.9 0.4 3.5 4.9 4.7 0.3 3.4 4.8 4.7 0.3 H:C O* 1 .o 4.3 2.6 0 1.3 2.0 4.6 1.1 1.2 4.0 3.7 0.7 2.2 6.7 3.9 0.8 1.2 3.9 2.9 0.1 3.5 1.2 1.5 1.6 3.3 1.2 1.1 0.1 3.2 0.9 1.1 0.1 ratio 1.26 1.31 1.23 1.33 1.59 1.60 1.29 1.60 1.40 1.53 1.25 1.47 1.30 1.45 1.41 1.81 1.25 1.48 1.47 1.83 1.01 1.44 1.47 1.60 1.01 1.16 1.32 1.26 1.02 1.18 1.36 1.27 occurs in an elemental form.The refinery bitumen shows the lowest sulfur content of 2.7%. For TLA the sulfur content varies from fraction to fraction, ranging from zero in the saturates to 6.1% in the asphaltenes. This variation is also noted for the oxygen content, which varies from zero in the saturates to 4.3% in the polar aromatics. These trends are also observed for the Trinidad refinery bitumen but not in the Venezuelan bitumen, which shows a more uniform distribu- tion of heteroatoms throughout the different fractions.From Table 2, it can be seen that TLA has only a narrow range of H : C ratios, from 1.23 in the naphthene aromatic fraction to 1.33 in the saturate fraction. This is in contrast to both the Venezuelan and refinery bitumens, which show relatively large ranges for the H : C ratio of their fractions. The Trinidad refinery bitumen has an average sulfur content of 2.7%, which increases on blending with TLA; however, note the introduction of sulfur-containing compounds into the saturate fraction. The refinery bitumen also contains 4.6% oxygen in the naphthene aromatic fraction, which is compar- able to the 4.3% in the TLA polar aromatic fraction. The blend produces a material that contains 4.0 and 3.7% oxygen in the polar and naphthene aromatic fractions, respectively.When TLA is blended with the 150 penetration refinery bitumen, an increase in the H : C ratios for all the fractions occurs. When this blend is air blown to the ASTM D 312 Type IV specification, the H : C ratio for the asphaltene and polar aromatic fractions is reduced, whereas for the saturate and naphthene aromatic fraction it increases. It is worth noting also that the oxygen content of all these fractions shows a steady increase on air blowing. When the refinery bitumen was air blown, a change in composition that closely resembled that of the TLA samples was observed. The former showed an average H : C ratio of 1.31 after blowing whereas the latter showed a ratio of 1.28. With the exception of the asphaltene fraction, the blown TLA material shows a reduction in oxygen content on weathering.This phenomenon is accompanied by an over-all decrease in sulfur-containing compounds. The IR spectra of the unmodified and blown asphalts are shown in Fig. 1. The natural asphalt and the Trinidad petroleum bitumen exhibited very weak bands at around 3500-3400 cm- l , which is attributed to stretching of N-H or 0-H bonds. These groups become more prominent after air blowing, but a similar change is not observed for the 4000 3000 2000 1600 1200 W avenu m ber/cm - 1 Fig. 1 Infrared spectra of selected asphalt samplesANALYST, JULY 1993, VOL. 118 949 Venezuelan material. Aromatic C-H stretching and meth- ylene vibrations are observed at around 3025 and 2890 cm-l, respectively. The small peaks at approximately 2728 cm-l are due to C-H stretching for aldehyde functional groups. The presence of these groups is confirmed by the peaks at around 1710 cm-l which are caused by C=O stretching for aldehyde and ketone groups.For the natdral asphalt samples these groups increase in concentration on blowing, as shown by the increased intensity of the peak. Conjugated systems show C=C stretching at around 1600 cm-l. The peaks at this frequency are more prominent for TLA than the petroleum bitumen, and become more intense on air blowing and weathering. Previous IR investigations of the bituminous materials have shown evidence of a few major functional groups affecting their rheological properties. 1 Functional group analysis was achieved via selective complexation of these groups, thus causing a disappearance or shift in the characteristic peaks associated with that functional group.Table 3 shows the results obtained from this type of analysis. A significant difference was found between the levels of free acid and acid anhydride groups occurring in TLA and the Venezuelan material. The quinolone levels decreased both on blowing and on weathering. These processes, however, cause an increase in the free acid and anhydride levels of the materials. Ketone levels seem to decrease on blowing but increase on weather- ing. The sulfoxide levels for both materials show the same trend, i . e . , an increase on blowing and weathering. Unblown TLA material, however, shows a decrease in sulfoxide level on weathering. The results for the lime-treated samples both before and after weathering are given in Table 4.It was observed that lime-treated TLA samples showed a reduction in the extent of surface cracking, which could be linked to inhibition of acidic functional groups in the materials. Discussion The type and number of functional groups present in an asphalt influence both the physical properties and the rate of ageing of the material.1 Elemental and IR analyses can help to identify the types present, whilst functional group analysis permits the major groups to be determined. These methods can also be used to follow the changes that occur during processes such as air blowing and weathering of the asphalt.2 Table 3 Functional group concentrations for bituminous materials (moll-' x 100) Quinol- Free Anhy- Sulf- Material 2-ones acids Ketones drides oxides TLA TPB TAB T.IV T.1V.W VPA V.IV v.1v.w 1.21 0.93 0.80 0.77 0.74 1.09 0.97 0.95 10.40 10.57 8.40 9.00 9.31 0.29 5.43 6.11 42.20 36.40 18.80 16.20 16.65 46.40 40.20 42.30 9.00 12.60 8.40 4.00 15.70 1.10 2.00 2.56 30.70 31.40 33.10 32.60 36.60 23.50 30.70 32.40 Table 4 Functional group concentrations for bituminous materials after lime treatment (mol I-' x 100) Quinol- Free Anhy- Sulf- Material 2-oncs acids Ketones drides oxides TLA .L 0.72 0.42 41.80 5.50 27.50 TLA.L.W 0.68 0.45 36.00 5.48 27.00 TAB. L 0.65 0.33 18.81 6.40 19.60 TAB.L. W 0.55 0.33 21.96 6.60 19.30 T.1V.L 0.66 0.42 16.20 6.20 34.30 T. IV . L . W 0.62 0.50 16.60 6.50 32.20 The heteroatom content of the material also gives an indication of the type of functional groups present. Generally, asphaltic materials contain carboxylic acids, dicarboxylic anhydrides, ketones, sulfoxides and quinol-2-ones.3 The existence of sulfur-containing compounds provides a channel for oxygen attack.During blowing and ageing, sulfur atoms are readily oxidized to sulfoxide groups. Sulfoxides are formed in preference to ketones, and are usually precursors to ketone and hence acid formation. The extent to which these groups are formed on ageing influences the physical properties of the asphalt and, consequently, its durability. Functional group composition is of particular importance in the asphal- tene and resin fractions'] as these fractions show the greatest variation in percentage composition and hence physical properties of the material.Comparison of the data from the elemental and functional group analyses for the TLA sample shows that there is an increase in the S : 0 ratio from 2 to 2.5 when the material is blended with refinery bitumen. Consequently, there is also an increase in the sulfoxide level. However, as the material is air blown, a decrease in the S : 0 ratio to 0.8 is observed, but there is little change in the sulfoxide level. This indicates that the oxygen introduced on air blowing is being incorporated into the material but not only as sulfoxide groups. This contrasts with the steady increase in both the S : 0 ratio and the sulfoxide level obtained as the Venezuelan samples are air blown. These observations suggest that the oxygen introduced during air blowing of TLA does not become incorporated into stable functional groups, but exists merely as transient groups that may serve a catalytic function in the polymeriza- tion process.This has been observed previously for some asphalts.1 The Venezuelan material, however, seems to show the formation of stable molecular species on air blowing. An increase in the percentage aromatic character occurs on air blowing but there is a slight decrease during the weathering process. Another factor that affects the stability of the materials is the heteroatom content. The electronegativity of atoms such as sulfur, nitrogen and oxygen causes considerable polarization of large molecules and thus polymerization due to ion-dipole interactions. Hence the delicate balance of hetero- atoms that exists is needed in order to maintain the stability of the colloidal system.The fractions with the higher percentages of heteroatorn species seem to undergo more changes on air blowing, indicating that the polarity of the components may play an important role in the stability of the material. The H : C ratio gives an indication of the degree of unsaturation of the material. For both Trinidad materials, the H:C ratio was the smallest in the naphthene aromatic fraction, possibly pointing to the polycyclic aromatic nature of these fractions. The large H : C ratio for the same fraction in the Venezuelan sample may indicate that the polycyclic aromatics here contain more aliphatic side-chains. As this is the fraction that is most affected by air blowing, the degree of unsaturation may be a limiting factor in this process. The H : C ratios and the IR analysis show that TLA has a greater aromatic character than both the local petroleum bitumen and the Venezuelan asphalt.However, the asphal- tene fraction of the latter does show extensive unsaturation. As the aromaticity and the degree of aliphatic side-chain interaction affect the stability of the material, TLA would be expected to be less stable than other petroleum bitumens of the same elemental composition. This suggests that the well- known durability of TLA arises from some other factor than its bitumen composition. Air blowing of the blended TLA caused a significant increase in oxygen content for all the different fractions, indicating that reactions occurred that caused the retention of oxygen in the molecules.This does not occur, however, for the Venezuelan sample in which the percentage of oxygen for the individual fractions decreases. The results indicate that two separate processes are occurring in the asphalts studied. For the TLA material the950 ANALYST, JULY 1993, VOL. 118 predominant trend was oxygen incorporation. This is substan- tiated by the observed increase in the 0-H stretching frequency for acids and acid anhydrides on air blowing. However, for the Venezuelan sample the molecules may have undergone condensation t o higher relative molecular mass compounds, with no net uptake of oxygen. This suggests that the oxygen in this instance is aiting solely as a catalyst for polymerization of polycyclic aromatic compounds via hydro- gen bonding, for example.Petersen et aZ.12 have suggested that in addition to oxygen uptake some elimination of nitrogen and sulfur may occur and it is possible that the eliminated sulfur may act as a condensing agent. Investigations were also carried out by Moschopedis and Speight,2 who postulated that should this type of reaction occur during asphalt ageing, it could well have an adverse effect on asphalt stability due to depeptization. The IR results show that the asphalts differ not so much in oxygen content, but rather in the type of oxygen-containing compounds. Unoxidized asphalts, which contain mainly ketone carbonyl compounds, contrast sharply with oxidized asphalts where the principal carbonyl is acidic (acids and acid anhydrides) in nature.However, TLA before modification shows properties similar to those of an oxidized asphalt, in that it contains a large percentage of acidic compounds, and also shows H: C ratios similar to those in blown bitumen samples. The microscopic examination of T.IV weathered material when compared with the V.IV weathered material showed more surface cracking. It was observed that this cracking can be reduced by neutralizing the acidic compounds that are present initially and produced on weathering by incorporating lime in the original blend. This indicates that it is primarily the excess acids or acid anhydrides that lead to changes in intermicellar interaction and hence cause premature harden- ing and eventual cracking. The lower acid levels originally and the small increases on weathering that are seen as the Venezuelan bitumen ages do not appear to upset the functional group equilibrium as much as that seen for TLA and as a result cracking is less apparent.Conclusions Heteroatom species may exist in asphalt in several forms, which, from their chemistry, impart characteristic physical properties to the material. The natural asphalt has the lowest H : C ratio and an average heteroatom content that falls between those of the Venezuelan and the local petroleum bitumens, which gives an indication that TLA is more stable than the local bitumen but not as stable as the Venezuelan material. The factors that seem to affect the durability of TLA are the relatively high degree of unsaturation, heteroatom content and the types of functional groups associated with the components.A high degree of unsaturation was shown to be the factor allowing the incorporation of oxygen atoms into the material. Heteroatom and functional group content also affect the physical properties of the material via chemical interac- tions, thus causing changes in the colloidal properties of the asphalts. With a simple understanding of the chemical changes that occur when air blowing and weathering take place, it was possible to reduce the effect that the acidic compounds had on the over-all physical behaviour of TLA. The authors are grateful to the following funding agencies for financial support for this research: National Institute for Higher Education, Research Science and Technology (NIHERST) , Port of Spain, Trinidad; Campus Research Fund Committee of the University of the West Indies, St. Augustine , Trinidad; and Board of Post-Graduate Studies, University of the West Indies. 1 2 3 4 5 6 7 8 9 10 11 12 References Petersen, J. C., Transp. Res. Rec., 1984, No. 999, 13. Moschopedis, S. E., and Speight, J. G., J . Mater. Sci., 1977,12, 990. Petersen, J. C., Anal. Chem., 1975,47, 112. Petersen, J. C., and Plancher, H., Anal. Chem., 1981,53, 786. Dorrence, S. M., Babour, F. A., and Petersen, J. C., Anal. Chem., 1974, 46, 2242. Petersen, J. C., Babour, F. A., and Dorrence, S. M., Anal. Chem., 1975,47, 107. Barth, E. J., Asphalt Science And Technology, Gordon and Breach, New York, 1962. Chatergoon, L., Whiting, R., and Smith, C., Analyst, 1992,117, 1869. Annual Book of ASTM Standards, Section 4, vol. 4.04, ASTM D 312-84, American Society for Testing and Materials, Phi- ladelphia, 1984. Annual Book of ASTM Standards, Section 4, vol. 4.04, ASTM D 4799-88, American Society for Testing and Materials, Phi- ladelphia, 1988. Goncharov, I. V., Babicheva, T. A., Bodak, A. N., Nemirov- skaya, G. B., and Mashigorov, A. A., Pet. Chem. USSR, 1985, 25,95. Petersen, J. C., Dorrence, S. M., Nazir, M., Plancher, H., and Babour, F. A., Proceedings of a Symposium on Characterization of Heavy Ends in Petroleum, vol. 26, American Chemical Society, New York, 1981, pp. 898-906. Paper 3100258F Received January 15, 1993 Accepted March 26, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800947

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JULY 1993, VOL. 118 95 1 CUMULATIVE AUTHOR INDEX JANUARY-JULY 1993 Aboal-Somoza, Manuel, 665 Abramovid, Biljana F., 899 Adams, Michael J., 229 Akasaka, Kazuaki, 765 Alder, J. F., 39.5 Allag, Houssein, 401 Al-Masri, M. S., 873 Alwarthan, Abdulrahman A., Anderson, David R., 449 Andrew, B. E., 153 Andrews, William J., 425 Aoki, Nobumi, 909 Arrigan, Damien W. M., 355 Ashok Kumar, T., 293 Avidad, Ramiro, 303 Bae, Yea-Ling, 297, 301 Banerjee, Arun B., 937 Bangar Raju, G., 101 Bannon, Thomas, 361 Barclay, David A., 245 Barjat, HervC. 73 Barker, Philip G., 347 Barnard Howie, Judith A., 35 Bartle, Keith D., 737 Bartlett , Philip N., 371 Barwick, Ian M., 489 Baxter, Douglas C., 495 Bayo, Javier, 171 Beauchemin, Diane, 815 Bell, Jimmy D., 241 Belton, Peter S . , 73 Benmakroha, Farida, 401 Bermejo Martin-Lazaro, A., 917 Bermejo-Barrera, Pilar, 665 Bernal Suarez, M.M., 917 Bhaskar, Nilam, 1 Biondi, Cinzia, 183 Birmingham, John. J., 1 Blackburn, R., 873 Blaih, Salah M., 577 Blair, Neil, 371 Boriitsky, Juri A., 859 Bos, Albert, 323 Bos, Martinus, 323 Boudjerda, Tarik, 401 Boufenar, Rabah, 401 Bovara, Roberto, 849 Bradbury, Michael W. B . , 533 Breen, William, 415 Brereton, Richard G., 779 Brienza, Sandra Maria Boscolo, Brinkman, Udo A. Th., 11 Brough, Paul A., 753 Brown, Marc B., 407 Bruns, Roy E., 213 Bui, LAn N., 463 Cai, Xiaohua, 53 Calokerinos, Antony C., 627, Campiglio, Antonio, 545 Canela, Ramon, 171 Capitin-Vallvey , Luis Fermin, Carlson, Robert G., 257 Carrea, Giacomo, 849 Casey, Vincent, 389 Cassidy, John F., 415 Catterick, Timothy, 791 Celeste, M., 895 Cepas, Juana, 923 Cerda, V., 895 Cermhk, Josef, 79 Chadima, Radko, 79 Chaisuksant, Rasamee, 179 Chan, Wing-Hong, 863,869 Chang, Qing, 839 Chartier, A., 157 Chatergoon, Lutchminarine, 947 Chen, Liang, 277 639 719 633 303 Chkn, Tianyu, 541 Chokshi, Hitesh P., 257 Chowdhury, Bhabadeb, 937 Cladera, A., 895 Clark, Alastair, 601 Clarke, Colin G., 229 Clifford, Anthony A..737 Clinton, Cathriona, 415 Comber, Sean, 505 Cooke, Michael, 449 Corbini, Gianfranco, 183 Cordero, Bernard0 Moreno, 209 Corti, Piero, 183 Costa Garcia, Agustin, 649 Cotsaris, Evangelo, 265 Crane, Michael, 617 Crean, G. M., 429 Crooks, Steven R. H., 447 Crosby, Neil T., 489 Crouch, Stanley R., 695 Crubellati, Ricardo O., 529 Cruz Or$iz, M., 801 CsizCr, Eva, 609 Cummins, Diane, 1 Cummins, Phillip G., 1 Cunningham, K., 341 Daenens, P., 137 Danielsson, Bengt, 845 Davis, Willard E., 249 dc Andrade, Jo20 Carlos, 213 de la Guardia, Miguel, 23 de la Rosa, Francisco F., 643 de Paula Eiras, SebastiFio, 213 Dean, John R., 747 Deasy, Brian, 355 Debrabandere, Lode, 137 Deftereos, Nikolaos T., 627 Delves, H.Trevor, 533 Dempsey, Eithne, 411 Diamond, Dermot, 347 Diaz, Susana, 171 Diewald, Wolfgang, 53 Djerboua, Ferhat, 401 Domansky, Karel, 335 Dominguez, Lucas, 171 Dominguez-Gonzalez, Raquel, dos Reis, Boaventura Freire, 719 Dowle, Chris J., 17 Dreassi. Elena, 183 Duran-Meras, Isabel, 807 Dvinin, Alexei, 859 Economou, Anastasios, 47 Edmonds, Tony E., 407,443 Efstathiou, Constantinos E., 627 Egan, Denise A., 201, 411 Eggins, Brian R., 439 Elliott, Christopher T., 447 El-Yazbi, Fawzy A., 577 Escobar, Rosario, 643 Espinosa-Mansilla, Anunciacion, 89, 807 Estela, J.M.. 89.5 Faure, Uta, 475, 481 Fearn, Tom, 235 Fernandez Laespada, Ma. Esther, 209 Ferri, Elida, 849 Feygin, Ilya, 281 Fielden, Peter R., 47 Finglas, Paul M., 475, 481 Fitzgerald, Catherine, 361 Flaherty, T., 429 Foster, Robert, 415 Fox, C. G., 157 Fraidias Becerra, Antonio J., Frech, Wolfgang, 495 Friel, Sharon, 371 Frutos, G., 59 Fu. Chengguang, 269 665 175 GaBl, Ferenc F., 899 Gaind, Virindar S., 149 Gallagher, Timothy, 753 Gangemi, Gianni, 849 Garcia G6mez de Barreda, Garcia-Lopez, Trinidad, 303 Garcia-Mesa, J. A., 891 Gardner, Julian W., 371 Georges, J., 157 Ghijsen, Rudy T., 11 Ghini, Severino, 849 Gibney, Patrick M., 425 Giosuk, Maria Antonietta, 849 Girotti, Stefano, 849 Givens, Richard S ., 257 Glennon, Jeremy D., 355 Gomez-Hens, Agustina, 707 Goodfellow, Brian J., 73 Gorog, Sandor, 609 Grau, Harald, 689 Greenfield, Stanley, 443 Greenway, Gillian, 17 Gregory, Donald P., 1 Grob, Robert, 11 Gu, Xiao-Hong, 863 Gu, Zhi-cheng, 105 Guiraum, Alfonso, 643 Haegel, Franz-Hubert, 703 Halbig, Peter, 689 Halls, David J., 821 Halvatzis, Stergios A., 633 Harnano, Takashi, 909 Hanai, Toshihiko, 769, 773 Hara, Hirokazu, 549 Harris, S. J., 341 Hartnett, Margaret, 347 Haswell, Stephen J., 245 Hawkesworth, K., 395 Hawkins, Peter, 35 Hembree, Jr., Doyle M., 249 Hernandez Garcia, J., 917 Hidalgo de Cisneros, Jose L. Hiraide, Masataka, 537 Hirose, Tsuyoshi, 517 Hokari, Norihisa, 219 Hollman, Peter C. H., 475,481 Hornig, James F ., 933 Horvath, KornClia S . , 899 Howard, Vyvyan C., 1 Huang, Ka-Lin, 205 Hunt, Terence P., 17 Idriss, Kamal A., 223 Tizuka, Ryuji, 16.5 Tmai, Kazuhiro, 759 Ishibashi, Mumio, 759 Ishida, Junichi, 165 Ito, Yoshio, 909 Ivaska, Ari, 885 Twachido, Tadashi, 273 Iwata, Tetsuharu, 517 Iyer, R. M., 929 Izquierdo, Pilar, 707 Izumi, Sigeru, 553 Janata, Jifi, 335 Jedrzejczak, Kazik, 149 Jefferies, Terry M., 753 Johnston, Brian, 355 Jones, Carol L., 1 Josowicz, Mira, 335 Ju, Doweon, 253 Kalcher, Kurt, 53 Kallury, Krishna M. R., 309 Kalman, Peter G., 463 Karayannis, Miltiades I., 711, Kasumimoto, Hanae, 131 Katz, Stanlcy E . , 281 Kawaguchi , Hiroshi, 537 Kessler, Margalith, 235 Daniel, 175 Hidalgo, 175 723, 727 Kcyes, Emmetine T., 385 Khayyami, Masoud, 845 King, Bernard, 587 Kinoshita, Toshio, 161, 769, 773 Kiranas, Efstratios R., 727 Kiss, Attila, 661 Kobayashi, Atsushi, 273 Kobayashi, Shouichi, 131 Koqak, Ali, 657 Koh, Tomozo, 669 Kojima, Nobuaki, 909 Konidari, Constantina N., 711 Konig, Monika, 703 Koshino, Yukihiro, 827 Kotrlg, Stanislav, 79 Kovanic, Pavel, 145 Krishan Puri, Bal, 85 Kubal, Gina, 241 Kumar, Manject, 193 Kundu, Dipali, 905 Kvalhein, Olav M., 779 Lan, Chi-Ren, 189 Lang, Mark J., 425 Larsson, Per-Olof, 845 Lauko, Anna, 609 Ledesma, Ariel G., 529 Ledingham, Kenneth W.D., 601 Lee, Albert Wai-Ming, 869 Lev, Ovadia, 557 Li, Ronghua, 563 Li, Xiang-Ming, 289 Liang, Wei-An, 97 Liang, Yi-Zeng, 779 Lin, Qingxiong, 643 Lin, Yuehe, 277 Littlejohn, David, 541, 821 Lopez Palacios, Jesus, 801 Lopez Ruiz, B., 59 Lowdon, John, 747 Lowe, Roger D., 613 Lunar, Maria Loreto, 715 Lunte, Susan M., 257 Luque de Castro, Maria Dolores, 593, 891 Lyons, Cormac H., 361 Lyons, Michael E.G., 361 Mc Monagle, James B., 389 McArdle, Fiona A., 419 McCallum, John J. , 401 McCaughey, William J., 447 McClean, Stephen, 511 MacCraith, Brian D., 385 McDonagh, Colette M., 385 MacDonald, Robert C., 913 McEvoy, John D. G., 447 McGilp, John F., 385 MacKay, Graham A., 741 McKeown, Neil B., 463 McKervey, M. A., 341 MacLaurin, Paul, 617 McLeod, Cameron W., 449 Magee, Robert J., 53 Malone, Michael A., 649 Maquieira, Angel, 855 Marshall, Archibald, 601 Martelli, Patricia Benedini, 719 Martin, J. P., 59 Martinez-Lazano. C., 567 Masotti, Piero, 849 Mathieu, Jacques, 11 Matsubara, Chiyo, 553 Meguro, Hiroshi, 765 Mellidis, Antonios S ., 179 Mertens, Bart, 235 Midgley, Derek, 41 Miller, James N. , 407, 455 Miller, Richard M., 1 Mills, Andrew, 839 Mitsuhashi, Yukimasa, 909 Miura, Y asuyuki, 669 Mohan, Hari, 929952 ANALYST, JULY 1993, VOL. 118 Monks, Cheryl D., 623 Moreno, Miguel A., 171 Mori, Yuichi, 553 Moriyama, Youichi. 29 Moss, Martin C., 1 Mottola, Horacio A., 675 Moulder, Robert, 737 Muiioz de la Pefia, Arsenio, 807 Muiioz Leyva, Juan A., 175 Munro, C. H., 731 Nabekura, Tomiko, 273 Nacapricha, Duangjai, 623 Nagahiro, Tohru, 85 Nakagawa. Genkichi, 219 Nakai, Chie, 769, 773 Nakamura, Kayoko, 29 Nakamura, Masaru, 517 Nanos, Christos G., 711 Narayanaswamy, Ramaier, 317 Narukawa, Akira, 827 Navalon, Alberto, 303 Neuhold, Christian, 53 Niazi, Shahida B., 821 Nicholson, Brenton C., 265 Nickel, Ulrich, 689 Nimura, Noriyuki, 161, 769, 773 Norman, Philip, 617 Nwosu, Titus, 845 O’Beirn, Brendan, 389 O’Donoghue, Eilish, 415 Ohkubo, Hiromi, 549 Ohrui, Hiroshi, 765 Oji, Yoshikiyo, 909 Okabe, Katsuaki, 669 O’Kane, Edward, 511 O’Keeffe, Gerard, 385 O’Kelly, Brendan, 385 O’Kennedy, Richard, 201,411 O’Neill, Robert D., 433 O’Sullivan.Ciara, 411 Palaniappan, R., 293 Pandey, Amita, 941 Papageorgiou, Vassilios P., 179 Pasha, Akmal, 777 Paukert, TomaS, 145 Paynter, J., 379 Pearce, Timothy C.. 371 Perez Pavon, Jose Luis, 209 Perez-Bendito, Dolores, 707, PCrez-Ruiz, T., 567 Peris Cardells, Empar, 23 Persaud, Krishna C., 419 Petelenz, Danuta, 335 Petrukhin, Oleg M., 859 715, 923 Pinatel, Henri, 831 Pitre, Krishna S . , 65 Pramauro, Edmondo, 23 Preston, Gaynor, 245 Prevot , Alessandra Bianco, 23 Prieta, Javier, 171 Proietti, Daniela, 183 Puchades, Rosa, 855 €‘yo, Dongjin, 253 Quencer, Brett M., 695 Radulovic, Stojan, 241 Radunovie, Aleksandar, 533 Rahmani, Ali, 779 Ramachandran, Venkataraman Ramesh, A., 945 Rauch, Pavel, 849 Raurich, Josep Garcia, 197 Reckhow, David A., 71 Reed, Peter I., 877 Reid, Helen J., 443 RCpGsi, Janos, 661 Riley, David P., 407 Roda, Aldo, 849 Roe, Merrion P., 425 Romaschin, Alex D., 463 Roy, S.K., 905 Ruan. Fu-Chang, 289 RubeSka, Ivan, 145 Rubio Leal, Amparo, 89 Rubio, Soledad, 715 RGiiEka, Jaromir, 885 Sabot, Jean-Franqois, 831 Sadler, Peter J., 241 Saez, Josk A., 801 Saleh, Magda M. S . , 223 Salinas, Francisco, 89, 807 Salvatore, Michael J., 281 Sanchis, Vicente, 171 Sander, Joseph, 601 Santana Rodriguez, J.J . , 917 Sanyal, Asis K., 937 Sanz, A., 567 Sanz Pedrero, P., 59 Satake, Masatada, 85 Savarino, Piero, 23 Sawai, Kaori, 549 Schneider, Jeffery A., 933 Schneider, Siegfried, 689 Schwuger, Milan Johann, 703 Seare, Nichola J., 407 Shallow, A., 429 Sharma, Devender K., 941 Sheppard, Robert C., 1 Shibata, Masaru, 909 N., 511 Shimoishi, Yasuaki, 273 Shiu, Kwok-Keung, 863, 869 Shortt, Desmond H., 447 Silva, Manuel, 681, 923 Simpson, Michael, 449 Singhal, Ravi P., 601 Singleton, Scott, 1 Slangen, Jean H., 475, 481 Slater, Jonathan M., 379 Smith, Clayton, 947 Smith, Roger M., 741 Smith, W. E., 731 Smyrl, Norman R., 249 Smyth, Malcolm R., 411, 649 Smyth, W. Franklin, 511 Snook, Richard D., 613 Somer, Guler, 657 Soto-Ferreiro, Rosa M., 665 Southgate, David A.T., 475,481 SrBmkovB, Jitka, 79 Srivastava, P. K., 193 Stalikas, Constantine D., 723 Su, Hongbo, 309 SuBrez, Guillermo, 171 Subbarao, Nanda K., 913 Suliman, Fakhr Eldin O., 573 Sultan, Salah M., 573 Svehla, Gyula, 341, 355 Svendsen, C. N., 123 Takamura, Kiyoko, 553 Tang, Gui-Na, 205 Taniguchi, Hirokazu, 29 Tanweer, Ahmad, 835 Taylor, Colin G., 623 Teasdale, P. R. , 329 Terao, Tadao, 759 Thakur, Hari K., 941 Thompson, Michael, 235, 309, Timotheou-Potamia, Meropi Tomas, C., 895 TomBs, V., 567 Torrades, Francesc, 197 Torro, Luis, 855 Toyo’oka, Toshimasa, 257, 759 Tsai, Suh-Jen Jane, 297,301,521 Tsionsky, Michael, 557 Tsuzuki, Wakako, 131 Tucker, Alan, 241 Tufi6n Blanco, Paulino, 649 Tzouwara-Karayanni, Stella M., Uchida, Takaaki, 537 Urusov, Yuri I., 859 463 M., 633 723, 727 ValcBrcel, Miguel, 593, 891 Van Boven, M., 137 van der Linden, Willem E., 323 Veiro, Jeffrey A., 1 Verma, Balbir C., 941 Verma, Neerja, 65 Vijaya Raju, K., 101 Vijayashankar, Yadathora N., Vilchez, Jose Luis, 303 Viscardi, Guido, 23 Vos, Johannes G., 385 Voulgaropoulos, Anastasios, Wada, Hiroko, 219 Wagstaffe, Peter J., 475, 481 Wallace, G. G., 329 Walton, Philip W., 425 Wang, Bao-Ning, 205 Wang, Joseph, 277, 411 Warwick, Peter, 489 Watt, E. J., 379 White, Peter C., 731, 791 Whiting, Robin, 947 Williams, David M., 249 Williams, Kathleen E., 245 Wong, Kwok-Yin, 289 Wong, Wai-Cheong, 869 Worsfold, Paul J., 617 Worswick, Richard, 583 Wuchner, Klaus, 11 Xie, Bin, 845 Xie, Yuefeng, 71 Xu, Guoping, 877 Xu, Hongda, 269 Yamaguchi, Masatoshi, 165,517 Yamauchi, Shuji, 161, 769, 773 Yan, Hsiao-Tzu, 521 Yang, Mengsu, 309 Yoshida, Tomohiko, 29 Yoshioka, Hiroshi, 553 Yuchi, Akio, 219 Zagatto, Elias Ayres Guidetti, Zenki, Michio, 273 Zhang, D., 429 Zheng, Minghui, 269 Zhou, Jie, 97 Zhou, Zhauro, 563 Zhu, Zhong-liang, 105 Zotou, Anastasia, 753 Zou, Shi-Fu, 97 777 179 719

ISSN:0003-2654    

DOI:10.1039/AN9931800951

出版商:RSC    

年代:1993

数据来源: RSC