摘要:

ANALYST, JUNE 1993, VOL. 118 707 Simultaneous Kinetic Fluorimetric Determination of Amoxycillin and Clavulanic Acid by the Stopped-flow Mixing Technique* Pilar Izquierdo, Agustina Gomez-Hens and Dolores Perez-Benditot Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, E- 14004, Cordoba, Spain Kinetic methodology was applied t o the resolution of mixtures of amoxycillin and clavulanic acid. The method thereby developed is based on the formation of fluorescent derivatives in the presence of cerium(iv) in an acidic medium, and the simultaneous determination is made possible by the difference between the reaction rates of the two analytes. By using the stopped-flow mixing technique, kinetic data can be obtained within a few seconds of the reactants mixing, which allows the proposed method t o be readily applied t o routine analyses.The calibration graphs were linear over the range 0.06-15.0 pg ml-l for amoxycillin and 0.10- 17.5 pg ml-l for clavulanic acid. The relative standard deviation was less than 3%. Mixtures of amoxycillin and clavulanic acid in ratios between 6 : 1 and 1 : 4 were satisfactorily resolved. The proposed simultaneous method was applied t o the determination of these compounds in pharmaceutical preparations. Keywords : Am ox ycillin; cla vu Ian ic acid; stopped- flo w; flu0 rime tr y; p h a rm a ce u tica I p re pa ra ti0 ns Clavulanic acid exhibits only weak antibacterial activity, but is a powerful inhibitor of (3-lactamases, which hydrolyse the cyclic amide bond of (S-lactam antibiotics such as amoxycillin, which are converted into antibiotically inactive open-ring compounds.' Hence, clavulanic acid is generally commercial- ized in formulations containing amoxycillin, where the anti- bacterial activity of the latter is potentiated by the acid.There is a wide variety of methods available for the determination of these compounds. Almost all of them are equilibrium methods; in fact, kinetic methodology has only been used for the determination of clavulanic acid by measuring the time required for the enzymic degradation of penicillin G to penicilloic acid.2 The calibration graph for this fixed-absorbance method is non-linear, and the time required for each assay is 1-2 h. Kinetic methods based on the stopped-flow mixing approach3 are gaining interest in chemical analysis as a means of accomplishing automation whenever fast reactions are involved.Hence, this work represents the first attempt at assaying amoxycillin and clavulanic acid mixtures by use of kinetic methodology. The aim of this study was to develop a very simple and fast method without using separation techniques such as liquid chromatography.4 For this purpose the reaction between these compounds and cerium(1v) in an acidic medium was used, which yields very intensely fluor- escent derivatives. The reactions are so fast that their rate cannot be monitored by using the batch technique; hence it is necessary to use the stopped-flow mixing technique, which allows analytical data for each sample to be obtained within a few seconds. Experimental Instrumentation A Perkin-Elmer (Beaconsfield, Buckinghamshire, UK) LS-50 luminescence spectrometer, fitted with a stopped-flow module5 supplied by Quimi-Sur Instrumentation (Seville , Spain), was used for fluorescence measurements.The instru- ment was controlled via a Hewlett-Packard (Avondale, PA, USA) Vectra computer. Reaction-rate data were obtained by using the 'Kinetic Obey' application program. The excitation and emission slits were adjusted to provide a 5 nm bandpass, and the observation cell of the stopped-flow module had a * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992. t To whom correspondence should be addressed. pathlength of 1.0cm. The solutions in the stopped-flow module and the cell compartment were kept at a constant temperature (60 "C) by circulating water from a thermostatic- ally controlled tank.Reagents All the reagents used were of analytical-reagent grade. Amoxycillin was purchased from Sigma (St. Louis, MO, USA) and clavulanic acid (lithium salt) was kindly supplied by Beecham Pharmaceuticals. Stock solutions were freshly prepared in distilled water before use. A cerium(rv) solution was prepared from Ce(SOp)2-4H20 (Merck, Darmstadt , Germany) in 1 mol 1-1 sulfunc acid. Procedures Individual determination of amoxycillin and clavulanic acid One of the two 10ml reservoir syringes was filled with a solution containing 0.4 ml of 1.6 X mol 1-1 cerium(rv) and 1.0 ml of 1 mol I-' sulfuric acid in a final volume of 10 ml.The other syringe was filled with a solution containing 1.0 ml of 1 mol 1-1 sulfuric acid and amoxycillin or clavulanic acid standard or sample solution at a final concentration between 0.06 and 1 5 . 0 ~ g m l - ~ amoxycillin or from 0.10 to 17.5 pg ml-l clavulanic acid in a final volume of 10 ml. After the two 2 ml drive syringes had been filled, 0.15 ml of each solution was mixed in each run in the mixing chamber. Fluorescence increments during the reaction were monitored at he, = 256 nm and hem = 356 nm. Fluorescence values were obtained over 10 s and subsequently processed by the microcomputer, furnished with a linear-regression program for application of the initial-rate method ('Kinetic Obey'). The reaction rate was determined in about 3 s and each sample was assayed in triplicate.The blank signal was found to be negligible. All measurements were carried out at 60°C. Determination of amoxycillin and clavulanic acid in mixtures The solutions used to fill the two 10 ml reservoir syringes were prepared as described above; however, that containing the analyte was prepared from mixtures of arnoxycillin and clavulanic acid. Reaction-rate measurements were carried out in two different regions of the kinetic curve obtained at he, = 256 nm and hem = 356 nm. In the first region, the reaction rate was calculated during the first 3 s, and this corresponded to the sum of the contributions of amoxycillin and clavulanic acid. The reaction of amoxycillin was completed in 15 s, so the708 ANALYST, JUNE 1993, VOL. 118 reaction-rate values obtained between 15 and 30s only depended on the clavulanic acid concentration.The calibra- tion performed in this region of the kinetic curve (reaction rate versus clavulanic acid) allowed us to determine the concentra- tion of this analyte in the mixture. The difference between the total initial rate obtained for the first region of the kinetic curve and the initial rate ascribed to clavulanic acid from the corresponding calibration graph was thereby directly propor- tional to the amoxycillin concentration. The blank signal was found to be negligible. All measurements were carried out at 60 "C. Determination of amoxycillin and clavulanic acid in pharmaceuticals No sample pre-treatment was needed for these analyses other than appropriate dilution of the sample to obtain a concentra- tion level that was within the linear range of the calibration graphs.An accurately weighed amount of sample (powder for suspension) (approximately 0.125 g) was transferred into a 1OOOml calibrated flask and was diluted to volume with distilled water. The suspension was shaken for 5 min in an ultrasonic bath and then filtered. A volume of the filtrate (1 ml) was treated as described above. Results and Discussion Study of the Fluorescence Reactions A systematic study of the effect of various oxidants on the fluorescence behaviour of amoxycillin and clavulanic acid, which was carried out in order to find an appropriate system for the kinetic resolution of mixtures of these compounds, revealed that both analytes were highly fluorescent at hex = 256nm and he, = 356 nm (Fig.1) in an acidic medium containing cerium(1v); also, the intensity of clavulanic acid was higher than that of amoxycillin. Because these wave- lengths are the same as those for cerium(m), it was initially thought that the fluorescence signal corresponded exclusively to the cerium(rI1) formed in the redox reaction between amoxycillin (or clavulanic acid) and cerium(1v). However, replacing these compounds with arsenite in the presence of iodide resulted in a lower fluorescence intensity (Fig. 1, curves C,C') than for amoxycillin and clavulanic acid. This should be ascribed to the formation of a complex between the corres- 500 A' I 0 200 240 280 320 360 400 hln rn Fig. 1 Excitation (kern = 356 nm) and emission (Aex = 256 nm) spectra obtained for the following s stems: A and A', clavulanic acid- CeIV; B and B', amox cillin-Ce'?; and C and C', Aslll-CelV-I- H2S04] = 0.1 rnol 1-l; &avulanic acid] = [amoxycillin] = 1 pg ml-l; [CeIV] = 3.2 x rnol 1-I; [As"'] = 2.4 x 10-smoll-l; [I-] = 25 ng ml-l; temperature = 30 "C; and time = 10 min ponding oxidation product of amoxycillin and clavulanic acid and the ceriurn(rI1) formed in the redox reaction, which would have the same excitation and emission wavelengths, but a markedly different fluorescence intensity. Other oxidants such as hydrogen peroxide provided different results, so the fluorescence signals obtained with cerium(1v) could not be exclusively ascribed to the oxidation products of amoxycillin and clavulanic acid.Fig. 2 shows the kinetic curves obtained for amoxycillin, clavulanic acid and a mixture of both by using the stopped- flow mixing technique.As can be seen, the kinetic behaviour of the two compounds was different. Amoxycillin reached equilibrium faster than clavulanic acid, so after about 15 s the reaction rate obtained depended on the concentration of clavulanic acid alone. This allowed the development of a very simple, fast method for the simultaneous resolution of mixtures of these analytes by measuring the reaction rate in two different regions of the kinetic curve. In the first region, the initial rate corresponded to the joint contribution of the amoxycillin and clavulanic acid concentrations, whereas after 15 s, where a slight decrease in the slope of the kinetic curve was observed, the reaction rate obtained only depended on the clavulanic acid concentration. Effect of Reaction Variables The amoxycillin and clavulanic acid systems were optimized by altering each variable in turn while keeping all others constant.All stated concentrations are initial concentrations in the syringes (viz., twice the actual concentrations in the reaction mixture at time zero after mixing). Each kinetic result was the average of three determinations. Both amoxycillin and clavulanic acid systems required an acidic medium for the fluorescence signals to be obtained. Hence, various acids (H2S04, HC104 and HCI) were tested at final concentrations of 0.1 rnol 1-1 in each syringe; the best results were provided by sulfuric acid. The effect of the concentration of this acid was studied over the range 0.054.2 mol 1-1 [Fig.3(a)]. The reaction rate for the amoxycillin system was constant and independent of this variable above a concentration of 0.09 moll-'. However, the rate obtained for the clavulanic acid in the two regions of the kinetic curve was constant over the range 0.094.125 mol I-' and decreased slightly at higher sulfuric acid concentrations. Fig. 3(b) shows the effect of the cerium(iv) concentration on the two systems. Both reaction rates increased with increasing temperature over the range 2040°C; also, additivity of the reaction rates was better with the first region of the kinetic curve when a temperature of 60°C was used. The decrease in the relative permittivity of the solutions, resulting from the addition of r I a, .- - ;;J 100 a, CT B D 0 10 20 30 40 50 Tirnels Fig.2 Kinetic curves obtained for: A, amox cillin; B, clavulanic and H2S04 (0.1 mol I - I ) and D, blank signal. Temperature = 50 "C acid; C, a mixture of both in the presence of Ce' Y (3.2 X rnol I F 1 )ANALYST, JUNE 1993, VOL. 118 709 0 0.05 0.1 0.15 0.2 0 2.5 5.0 7.5 10 [H,SO,]/mol 1-1 CeIV/lO-5 mol 1-1 Fig. 3 Effect of the sulfuric acid (a) and CerV ( h ) concentration on the reaction rate obtained for clavulanic acid (A, Af), amoxycillin (B) and a mixture of both (C). (Af: reaction rate measured between 15 and 30 s on the kinetic curve) Table 1 Features of the individual determinations of amoxycillin and clavulanic acid Clavulanic acid Amoxycillin 1* 2+ Linear range/pg ml- 0.06-15.0 0.1-17.5 0.1-15.0 Pearson's correlation Detection limit/yg ml-1 0.02 0.04 0.04 Precision* (YO) At 2 pg mIk1 1.9 2.8 2.4 At 10 pg ml-l 0.7 2.9 2.5 * 1: From the first region of the kinetic curve.t 2: From the second region of the kinetic curve. * Relative standard deviation 0, = 0.05, n = 11). coefficient (F-) 0.999 0.999 0.993 Table 2 Effect of various species on the determination of 5 pg ml-I of amoxycillin and 5 pg ml-l of clavulanic acid Tolerated species-to-analyte ratio Amox ycillin 1 1 1 2 1 2 0.5 1 Clavulanic acid 2 1 1 5 0.5 1 2 1 Foreign species Ampicillin Cephalexin Cephaloridine Cephalothin Cephradine Cloxacillin Dicloxacillin Oxacillin ethanol (up to 30%), resulted in a concomitant decrease in the reaction rate. The slopes of the fluorescence-time graphs obtained under the optimum experimental conditions were indicative of a first-order reaction with respect to amoxycillin and clavulanic acid.Features of the Proposed Method Under the optimum experimental conditions described above, the fluorescence-time curves obtained at different concentra- tions of clavulanic acid were processed by calculating reaction- rate data in two different regions of the kinetic curve: at the beginning of the reaction ( i e . , during the first 3 s) and between 15 and 30 s after the reaction had been started. In this way, two calibration graphs were obtained for clavulanic acid. The second calibration graph was required to resolve mixtures of the two analytes. As the amoxycillin reaction was faster than that of clavulanic acid, and equilibrium was reached in about 15 s, the calibration graph (initial rate versus amoxycil- lin concentration) was only obtained in the first region of the ki.netic curve.The features of the determinations are sum- marized in Table 1. The detection limits were calculated Table 3 Resolution of mixtures of amoxycillin and clavulanic acid Amox ycillid Clavulanic pg ml-1 acidlyg ml- Amoxycillin : clavulanic acid ratio Added Found* Added Found* 6 : 1 5 : l 4: 1 3 : 1 2 : 1 1 : l 1: 1 3 :2 1 : 4 1.8 2.0 2.0 1.5 2.0 2.0 1 .o 1.5 0.5 1.71 2.04 2.03 1.50 1.90 1.90 0.92 1.50 0.52 0.3 0.4 0.5 0.5 1 .o 2.0 1 .o 3.0 2.0 0.29 0.39 0.48 0.48 0.96 2.07 0.96 2.92 2.07 * Average of three determinations. according to IUPAC recommendations.6 As shown in Table 1, the relative standard deviations obtained were less than 3% in all instances.The selectivity study carried out (Table 2) showed that none of the penicillins and cephalosporins tested interfered with the determination at the same concentration levels as both analytes except for dicloxacillin and cephradine, which interfered with the determination of amoxycillin and clavu- lanic acid, respectively. Resolution of Mixtures Amoxycillin and clavulanic acid can be determined simul- taneously, taking into account that the initial rate obtained during the first 3 s of reaction is additive for both analyte concentrations and that the reaction rate obtained after 15 s depends on the clavulanic acid concentration only. The proposed method can be applied to amoxycillin concentra- tions lower than 2.5 pg ml- *, above which amoxycillin yields a small slope for the second region of the kinetic curve.The method was applied to the analysis of various synthetic mixtures of the two compounds. The results obtained are summarized in Table 3; as can be seen, mixtures of amoxycil- lin and clavulanic acid in ratios between 6 : 1 and 1 : 4 can be resolved with errors equal to or less than 5%. Applications In order to check the applicability of the proposed stopped- flow method to the resolution of mixtures of amoxycillin and clavulanic acid, it was used to analyse various pharmaceutical preparations. The results obtained are summarized in Table 4. Recoveries were measured by adding two different amounts of amoxycillin and clavulanic acid standards to each sample suspension and subtracting the results obtained for the samples prepared similarly, but with no amoxycillin or clavulanic acid added. The recoveries obtained were between 85.0 and 112.570, with an average of 103.8% for amoxycillin and 92.0% for clavulanic acid.These results are also listed in Table 4. Although the results obtained range between rela- tively high and low values, particularly at the lowest concen- tration of each analyte, one should take into account that the amounts added were small. Conclusions The results obtained in this work show that the proposed stopped-flow method allows the simple, fast and precise determination of amoxycillin and of clavulanic acid in pharmaceutical preparations. Also, the method can be readily adapted to routine analyses because each determination takes only a few seconds.The need to use a final amoxycillin concentration below 2.5 pg ml-l can be readily met by appropriate dilution of the sample.710 ANALYST, JUNE 1993, VOL. 118 Table 4 Determination of amoxycillin and clavulanic acid in pharmaceutical preparations* Recovery Stated A/mg Clmg - Amoxyplus 500 125 Augmentine 500 125 Bigpen 500 125 Burmicin 500 125 Clavepen 500 125 Clavumox 500 125 (Nova@ (Beecham) (Fides) (CcPa) (Prodes) (Antibioticos) * A = Amoxycillin; C = clavulanic acid. + Average of three determinations. Found? Nmg Clmg 502.1 131.2 489.6 129.2 504.2 129.6 484.2 129.2 479.2 130.0 489.6 129.2 A C Added/ Found?/ Recovery Addcdl Found?/ Recovery pgml-1 pgml-1 (%) pgml-1 pgml-l (%) 4.0 3.95 98.7 1 .o 0.85 85.0 8.0 7.96 99.5 2.0 1.85 92.5 4.0 4.25 106.2 1.0 0.92 92.0 8.0 8.26 103.2 2.0 1.90 95.0 4.0 3.90 97.5 1.0 0.89 89.0 8.0 7.90 98.7 2.0 1.94 97.0 4.0 4.38 109.5 1 .o 0.90 90.0 8.0 8.40 105 .O 2.0 1.88 94.0 4.0 4.50 112.5 1 .0 0.88 88.0 8.0 8.52 106.5 2.0 1.93 96.5 4.0 4.24 106.0 1 .o 0.90 90.0 8.0 8.24 103.0 2.0 1.89 94.5 The authors are grateful to the CICYT (Comision Interminis- terial de Ciencia y Technologia, Project PB91-0840) for financial support and to Beecham Pharmaceuticals for kindly supplying a sample of lithium clavulanate. 3 Gomez-Hens, A., and Perez-Bendito, D., Anal. Chim. Acta, 1991, 242, 147. 4 Foulstone, M., and Reading, C., Antzrnicrob. Agents Che- mother., 1982, 22, 753. 5 Lorinuillo, A., Silva, M., and Perez-Bendito, D., Anal. Chirn. Acta, 1987, 199, 29. Long, G. L., and Winefordner, J. D., Anal. Chern., 1983, 55, 712A. References 6 Chemistry and Biology of (J-Lactam Antibiotics, eds. Morin, R. B., and Gorman, M., Academic Press, New York, 1982, vol. 3. Gutman, A. L., Ribon, V., and Leblanc, J . P., Anal. Chem., 1985. 57, 2344. Paper 21052950 Received October 2, 1992 Accepted January 4, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800707

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 71 1 Decomposition of 2,6=Dichlorophenolindophenol in Strongly Acidic Solutions: a Potential Kinetic Method for the Determination of pH” Constantina N. Konidari, Christos G. Nanos and Miltiades 1. Karayannis De pa rtm en t o f Chemistry, La bo ra to r y o f A na I ytica I Ch em is tr y, University o f loan n ina, loan n ina, Greece The problem of pH measurements at extreme pH values is well known. Higher pH values are obtained at pH values of below 1 because of the so-called ‘acid error‘. The acid error depends on various parameters and it is not always reproducible. In order t o determine accurate pH values at high acidities, a kinetic study of the decomposition of 2,6-dichlorophenolindophenol (DCPI) in strongly acidic solutions was investigated by applying the stopped-flow technique.The DCPl is unstable at lower pH values (below 2) and the reaction rate constant for its decomposition in the pH range 0-2 is dependent on the pH. A good correlation between pH and the observed reaction rate constant of the decomposition of DCPl ( k o b ) was found for low pH values. The reaction rate constant, k,, was calculated by three different approaches for the evaluation of the experimental data. The weighted mean value of kl was found t o be (77 & 1) x 10-3 I mol-1 s-1’ with confidence limits of 95%. A knowledge of kl is also important in analysis as DCPl is the main reagent for the determination of ascorbic acid in acidic solutions. Keywords: pH kinetic determination; 2,6-dichlorophenolindophenol; stopped flow; acid error; ascorbic acid In the analytical laboratory, the pH meter with a glass-calomel electrode system is, almost exclusively, used for pH measure- ments. The convenience and direct-reading features of these instruments have led to an increase in the number of methods utilizing pH control.However, this convenience has led to many erroneous results because the limitations of the method are not always taken into account. Some of the errors are inherent properties of the glass electrode. The typical glass electrode works reliably in the pH range 1-9 and it is unaffected by poisoning, species, strong oxidizing and reduc- ing agents, proteins and gases. In alkaline solutions (pH >9) the pH value recorded tends to be lower than the true value (‘alkaline error’).At the acid extreme the glass electrode exhibits an error in solutions of pH <1 (‘acid error’). As a consequence, pH readings tend to be too high in this region. The magnitude of the error is dependent on a variety of variables and is generally not very reproducible. 1-3 2,6-Dichlorophenolindophenol (DCPI) is a widely used reagent for the determination of ascorbic acid in a variety of samples. An important source of errors in this determination is the reaction of DCPI with other reducing substances present in the samples4>5 and also the decomposition of DCPI at high acidities. Because DCPI is unstable at low pH values it was assumed that the rate of its decomposition is related to the acidity of the medium. A kinetic study was undertaken in order to establish a relationship between the rate of DCPI decomposition and pH.Based on the kinetic data a simple equation is proposed for the calculation of accurate pH values. Also, the dissociation constant, KD1, of the dye was calculated from the experimental kinetic data. Experimental Apparatus The experimental apparatus consists of a Durrum D-110 single-beam stopped-flow spectrophotometer (Durrum Instruments, Palo Alto, CA, USA), equipped with a logarith- mic photometric unit, a Model 5516ST storage oscilloscope (Kikusui Electronics, Kawasaki, Japan) and an Omnigraphic 2000 recorder (Houston Instruments, Houston, TX, USA). Data acquisition was accomplished with an IBM data acquisi- tion and control adapter (IBM, Boca Raton, FL, USA) and an IBM-compatible personal computer.Data acquisition soft- * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992. ware was written in Microsoft QuickBASIC and Macroas- sembler (Microsoft, Redmond, WA, USA), and uses the computer’s interrupt system for rapid data acquisition.6 The software allows the collection of 2500 points from the reaction curve, the measurements of the slopes at any point on the curve and the calculation of the observed reaction rate constant of the decomposition of DCPI, kob, applying the infinite time method, e . g . , graphical presentation of ln(A, - A,) versus time, where A, and A , are the absorbance values at time t and infinity, respectively.7 The system is also connected to a printer. A pHM83 Autocal pH meter (Radiometer, Copenhagen, Denmark) was used for pH measurements.Reagents 2,6-Dichlorophenolindophenol. A stock standard solution 9 X 10-4mol1-1 in DCPI (BDH, Poole, Dorset, UK), containing 210 mg 1-1 of NaHC03, was prepared. This solution was standardized by titration with freshly prepared ascorbic acid solution and stored in a refrigerator. Working standard solutions with different concentrations of DCPI were prepared daily from this stock standard solution by diluting with 210 mg 1-1 NaHC03 solution. The concentration of DCPI solution was verified spectrophotometrically at 522 nm (isobestic point of DCPI), where the molar absorptivity is 860 m2 mol-1.8 Hydrochloric acid. A stock standard solution 2 moll-’ in HCl (Titrisol, Merck, Darmstadt, Germany) was prepared.Working standard solutions of different concentrations were prepared daily from this stock standard solution by diluting with distilled water. Procedure The stopped-flow spectrophotometer was calibrated for 0 and 100% transmittance with HC1 in the observation cell. The concentration of the reagents was adjusted to create pseudo- first-order conditions with respect to DCPI. The stopped-flow apparatus was flushed several times to remove bubbles from the system and the observation cell. The reaction was repeated until the trace observed on the oscilloscope was reproducible. Data from the next run were collected and evaluated with the computer. The number of points collected and the intervals between them were adjusted according to the total reaction time.712 ANALYST, JUNE 1993, VOL. 118 1 0.8 0.6 c 0 m & 0.4 .- CI 0.2 0 0.5 1 1.5 2 2.5 3 PH Fig.1 Distribution diagram of the species of DCPI at low pH values In all the experiments the concentrations given in the figures and tables are the initial concentrations of the reagents in the observation cell. All measurements were performed at the isobestic point of DCPI, i.e., 522 nm.8 All solutions and the cell compartment were kept at 25.0 f 0.1 "C via a circulating thermostat. The pH adjustment was effected with HC1. Treatment of the Experimental Data The experimental data obtained were treated in a manner similar to that described previou~ly.~79-~2 In the pH range 0-2, DCPI exists in two states,l3 DH2+ and DH, in equilibrium with ionization constant KD1 = 0.3.By using this value the distribution diagram of DCI was construc- ted (Fig. 1). The over-all reaction for the decomposition of DCPI is a summation of the decomposition of both DH2+ and DH KD 1 (DH2+ DH + H+): kl DH2+ + products (1) k2 DH -+ products (2) From the treatment of the experimental data it was found that k2 is very small compared with kl, and the second reaction was therefore ignored (see below, Effect of pH). The initial reaction rate of the reaction in eqn. (1) is R. - (9) = kl[DH2+] dt in in (3) The concentration of DH2+ can be calculated from the known function, which gives the fraction ao: (4) aH+ - w 2 + 1 ao= - - cd a H + + KD1 where cd is the initial analytical concentration of DCPI and CYH+ is the activity of H+. Substitution of eqn.(4) into eqn. (3) gives The observed reaction rate constant is given by By taking logarithms of eqn. (6), eqn. (7) is obtained, which can be converted into eqn. (8): (7) The aH+ was calculated by the equation aH+ =fH+cH+ (9) where fH+ is the activity coefficient and cH+ the analytical concentration of hydrogen ion. The fH+ was calculated from the following equation" QZH+21i -logfH+ = 1 + BaiP where Q and B are constants, equal to 0.509 and 0.329 x 108, respectively, for water at 25 "C, ZH+ is the charge of hydrogen ion, ai is the distance of the closest approach of the ions to each other, equal to 5 x 10-8 cm, and I is the ionic strength. Determination of kl In order to determine the pH value from kinetic data it is necessary first to calculate the reaction rate constant kl.Three different methods were used to determine the value of this constant .9 Method I . Based on eqn. (3) and Beer's law ( A = Ebc) , the initial rate of change of the absorbance in the reaction mixture is Combining eqns. (11) and (4), we obtain where E is the molar absorptivity of DCPI at its isobestic point and b is the pathlength, equal to 2.00 cm. Plotting RA against cd the value of kl can be calculated at a given pH from the slope S = klebao. Method 2. Dividing the kob values by ao, kl is calculated at a given pH value based on eqn. (6). Method 3. On the basis of eqn. (8) and plotting logkob against log[ 1 + (KDl/aH+)], kl is calculated from the intercept. Results and Discussion Effect of DCPI Concentration The kob and RA values were calculated for different pH values, varying the DCPI concentration in the range (0.5 x 10-5-5.5 x lO-5mol- 1-1.Based on methods 1 and 2, kl was calculated. The results are presented in Table 1. The dependence of the initial rate on the initial analytical concentration of DCPI for the different pH values is shown in Fig. 2. Effect of pH The kob values were calculated at different pH values. Based on these experimental data, kl was calculated by applying methods 2 and 3. The results are presented in Table 2. From the distribution diagram for DCPI it is evident that in the pH range 0-2 the protonated (DH2+) and the undisso- ciated (DH) species of the dye predominate (Fig. 1). After the treatment of the experimental data it was found that DH2+ is the species participating in the reaction leading to a colourless product.If both species were active, eqn. (6) would read kob = klao + k2al, where al is the fraction of DH. Applying the previous equation for two pH values, 2.18 and 1.85, and solving the system, the values kl = 0.074 1 mol-l s-1 and k2 = 7 X 10-5 1 mol-1 s-1 were calculated. It is evident that k2 is very small compared with kl and can be neglected for the calculation of kob. Also, as Gupta and Gupta14 suggest, in thisANALYST, JUNE 1993, VOL. 118 713 the dissociation constant KD1 and the reaction rate constant k1 can be calculated simultaneously. Converting eqn. (6) into r 0) 3 0.02 tc" 0.01 0 1 2 3 4 5 6 [DCPlldlO-5 mol I-' Fig. 2 Dependence of initial rate, RA, on the initial analytical concentration of DCPI. T = 25.0 "C. A, pH = 0.47, RA = 2 x 10-4 + 718.8cd, r = 0.9998.B, pH = 0.83, RA = -7 x + 440.6cd, r = 0.9996. C, pH = 1.28, RA = 8 x + 195.0cd, r = 0.9998. D, pH = 1.54, RA = 2 X 10-6 + 115.2cd, r = 0.9997 Table 1 Calculation of kl according to methods 1 and 2 at different pH values at 25.0 "C ( n = 4). k1/10-3 1 mol-1 s-1 PH Method 1 Method 2 78 f 2 0.20 - 0.47 79 f 5 77 f 2 0.83 78 k 7 78 f 4 1.28 77 _t 4 77 f 3 1.54 76 f 6 77 f 1 Mean (77 k 3) x 10-3 (77 k 1) x 10-3 Table 2 Dependence of the observed reaction rate constant, kob, on pH. Calculation of kl applying methods 2 and 3. [DCPIIO = 3.29 x moll-l; temperature, 25.0 "C kJ1O-3 Method pH k,ds-l an 1 mol-1 s-1 2 0.20 0.0525 0.680 77 0.47 0.0403 0.530 76 0.83 0.0250 0.329 76 1.28 0.0114 0.148 77 1.54 0.0068 0.088 77 1.85 0.0034 0.045 76 2.18 0.0017 0.022 77 Mean, El = 77 x 10-3 1 mol-1 s-1 k1 = (77 f 1) x 10-3 1 mol-1 s-1 n = 7 Skl = 7.1 x 10-4, Ski = 2.7 x 10-4 3 logkob = f [log(l + KD1/aH+)]: y = -1.116 - 0.99x, r = 0.9999 Intercept = -1.12 f 0.01 Slope = 0.99 k 0.01 kl = (76 ? 2) x 10-3 1 mol-1 s-1 type of reaction scheme only the protonated form is reactive.The reaction was found to follow first-order kinetics with respect to DCPI. The reaction rate constant (kl) of the decomposition of DH2+ was calculated from the experimental data by applying the above methods of evaluation. Considering the data in Tables 1 and 2, a weighted mean value of kl = (77 k 1) X 10-3 1 mol-1 s-1 was calculated with confidence limits of 95%. Calculation of KD1 Based on the study of the pH effect and the experimental data collected for the observed reaction rate constant (kob), both and plotting l/kob against l/aH+, kl and KD1 can be calculated from the intercept and the slope, respectively.The result of the treatment is the regression equation kob-l = (12.8 k 0.9) -I- (3.93 k 0.04)aH+-', r = 1.0000 from which k1 = (78 k 5 ) x 10-3 1 mol-1 s-1 and KD1 = 0.306 k 0.003 were calculated. The value KD1 = 0.306 is in very good agreement with the value of 0.3 given in the literature.l3 The ionic strength does not affect the value of KDl because of the equilibrium DH2+ DH + H+, which contains one univalent ion on each side. Finally, an equation was derived for the determination of the pH value from kinetic data. It gives the pH as a function of the dissociation constant of DCPI (KD1), the reaction rate constant of the decomposition of DH2+ (kl) and the observed reaction rate constant (kob) at each pH value: pH = PKD~ + log - - (14) (::b l) The pH can also be calculated using the initial reaction rate RA, i.e. , the initial slope of the reaction curve, -(dA/dt)in, and applying the equation Calculations of propagation errors based on eqns. (14) and (15) have shown that the over-all relative error for the kinetic determination of pH in the range 0-2 is 0.2%. The rate of decomposition of DCPI in acidic solutions determines the errors in the volumetric determination of vitamin C with DCPI, because samples of ascorbic acid are mainly prepared and preserved in acidic solutions. Although otherwise very fast,13J5 the rate of the reaction of ascorbic acid with DCPI near the equivalence point of the titration becomes small and is comparable to the rate of decomposition of DCPI, which may lead to positive errors. Titration at pH >3 eliminates this drawback.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 References Pease, B. F., Basic Instrumental Analysis, Van Nostrand, New York, 1980, p. 276. Brenziski, D. P., Analyst, 1983, 108, 425. Metcalf, R. C., Peck, D. V., and Arent, L. J., Analyst, 1990, 115, 899. Konidari, C. N., and Karayannis, M. I., Talanta, 1991,38,1019. Konidari, C. N., Tzouwara-Karayanni, S. M., Bowman, L. E., and Karayannis, M. I., Talanta, 1992, 39, 863. Bowman, L. E., Victor, M. A., and Crouch, S. R., TRAC Trends Anal. Chem., 1990, 9, 111. Fleck, M. G., Chemical Reaction Mechanisms, Holt, Rinehart and Winston, New York, 1971, p. 36. Armstrong, J. McD., Biochim. Biophys. Acta, 1964, 86, 194. Konidari, C. N., and Karayannis, M. I., Anal. Chim. Acta, 1989, 224, 199. Yeremin, E. N., The Foundations of Chemical Kinetics, Mir, Moscow, 1979, p. 21. Frost, A. A., and Pearson, A. A., Kinetics and Mechanism, Wiley, New York, 2nd edn., 1961, p. 150. Marc Loudon, G., J . Chem. Educ., 1991, 68, 973. Tonomura, B., Nakatani, H., Ohnishi, M., Yamaguchi-Ito, J., and Hiromi, K., Anal. Biochem., 1978, 84, 370. Gupta, K. S . , and Gupta, Y. K., J . Chem. Educ., 1984,61,972. Karayannis, M. I., Talanta, 1975, 23, 27. Paper 2105293H Received October 2, 1992 Accepted December 14, 1992

ISSN:0003-2654    

DOI:10.1039/AN9931800711

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 715 Use of the Triiodide-Hexadecylpyridinium Chloride Micellar System for the Kinetic Determination of Molybdenum(v1)" Maria Loreto Lunar, Soledad Rubio and Dolores Perez-Benditot Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, 14004 Cordoba, Spain The associate formed by triiodide ion and hexadecylpyridinium chloride (cetylpyridinium chloride; CPC) micelles was used t o enhance the kinetic spectrophotometric determination of MoV1 by its catalytic action on the oxidation of iodide with hydrogen peroxide in an acidic medium. The reaction rate was monitored at 500 nm, the maximum absorption wavelength of the I3--CPC associate. The micellar medium allowed the determination of MoV1 at concentrations between 2 and 150 ng ml-I, with a detection limit of I .3 ng ml-' (i.e., about 6-8 times lower than those of methods implemented in aqueous media).The relative precision achieved for 80 ng mi-' of MoVl was 2.6%. The selectivity towards interferents, yielding yellowish solutions, was significantly enhanced. The proposed method was applied to the determination of Mo in aqueous extracts of soil samples with no prior separation. Keywords: Cetylp yridinium chloride; micellar media; triiodide; molybdenum (w); soil Many methods for the determination of trace amounts of MeV' in plant materials, soils and waters rely on the MoV1 catalysed oxidation of iodide ion by hydrogen peroxide. Enthalpi- metric,' ion-selective electrode2 and biamperometric3 meth- ods have so far been used for measuring the rate of this reaction; however, the lowest detection limits for Mo"' have been achieved by spectrophotometric detection of triiodide,4,5 which is, therefore, the most commonly used detection method for this purpose in spite of the poor spectral features of the triiodide complex in aqueous media (viz., maximum absorbance at 350 nm and low absorptivity).Measurements at such a short wavelength pose serious problems in analysing soil samples because the extracts usually absorb significantly, so corrections must be made in order to determine accurately the Mo content in the soil concerned. In addition, some metal ions, such as Fell', Vv and Crvl, pose serious interference effects; hence, the determination of MoV1 in soils often entails using a separation technique (e.g., extraction).As Mo is an essential trace element for both plants and animals, which require very small amounts of this element, but is toxic at high concentrations,6 accurate, precise determinations of Mo in soils are crucial to understand the status and fate of Mo in the environment, and to detect Mo pollution in order to take timely control measures. The spectrophotometric methods most commonly used for the determination of MoV1 essentially involve measuring the absorbance of the Mo-thiocyanate complex, which is only formed if MeV' is completely reduced to Mo", and extracting the complex so as to achieve enhanced sensitivity and selectivity. As spectrophotometric detection of triiodide is highly sensitive, development of improved meth- ods for monitoring this species is essential.Micellar systems have been successfully used to enhance existing analytical methods.7 The selectivity of analytical kinetic methods can be improved by including micelles in the reaction medium. This can be accomplished by altering the reaction pH or its mechanism,s and by inducing spectral shifts9 and excluding interfering species from micelle surfaces ,lo among others. In this context, the interaction between triiodide ion and hexadecylpyridinium chloride (cetylpyri- dinium chloride; CPC) micellesll is of special interest. At surfactant concentrations slightly above its critical micelliza- tion concentration (c.m.c), triiodide ion undergoes a batho- * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992.t To whom correspondence should be addressed. chromic shift from 350 to 500-530 nm, depending on the experimental conditions used, in addition to a substantial increase in its stability constant [(5.4 k 0.2) x 1041 mol-l, i e . , approximately 50 times that in water] and absorptivity [E = (3.89 k 0.008) x lo3 m2 mol-I, i.e., approximately 1.6 times that in water]. These effects can be used to overcome completely, or at least minimize, selectivity and sensitivity problems that confront many of the original spectrophoto- metric procedures for monitoring iodine, involving aqueous media. The interaction between triiodide ion and CPC has been studied in depth." In this work, it was exploited to enhance the spectrophotometric determination of MeV' by its catalytic effect on the oxidation of iodide with hydrogen peroxide.The method thereby developed was successfully applied to the determination of MeV' in soils. The analytical features of the I3--CPC system excel in comparison with those of the 13--starch system. Hence, the sensitivity of the calibration graph for the determination of iodine and the detection limit obtained, in the presence of CPC, are both greater by one order of magnitude.11 Experimental Apparatus Kinetic measurements were performed on a Hitachi (Tokyo, Japan) U-2000 spectrophotometer fitted with a 1 cm path- length cell. The spectrophotometer cell compartment was thermostatically controlled by circulating water, from a Neslab (Newington, NH, USA) RTE bath, with a temper- ature stability of kO.1 "C throughout.Reagents All reagents used were of analytical-reagent grade and were used as received. An MeV' stock solution (5.21 x mol I-') was prepared by dissolving 0.1261 g of Na2MoO4-2H20 (Merck, Darmstadt, Germany) in 100 ml of doubly distilled water. More dilute solutions (2.08 x rnol I - l ) were prepared from this stock solution, before each set of experiments, by dilution with doubly distilled water. A hydrogen peroxide solution (2.44 x mol 1-l) was prepared daily. A 6.6 x rnol I-' aqueous iodide solution was prepared and stored in a dark bottle. Aqueous solutions of the surfactant CPC (1.4 X moll-l; Serva, Heidelberg, Germany) and hydrochloric acid (3.16 x mol I-l) were also prepared. All working standards and reagent solutions were kept in a water-bath at 20 k 0.1 "C when used.716 ANALYST, JUNE 1993, VOL.118 Recommended Procedure for the Determination of Mo In a 10 ml calibrated flask, place in sequence the volume of hydrochloric acid (3.16 X mol 1-I) required to obtain a final pH of 2.5 (approximately 1 ml for standard samples), appropriate volumes of the MoV1 stock solution (2.08 x 10-5 moll-l) to provide a final concentration of between 2 and 150 ng mIk1, 0.2 ml of hydrogen peroxide (2.44 x rnol I - l ) , 0.75 ml of CPC (1.4 X rnol 1-l iodide. Start the stopclock immediately after the iodide is added and then dilute the solution to the mark with doubly distilled water. Transfer an aliquot of the reaction mixture into a cell kept at 20 k 0.1 "C and measure the solution absorbance ( h = 500 nm) as a function of time. Measurements are to be started exactly 45 s after the iodide is added.The reaction rate is calculated by the tangent method from the absorbance-time curves obtained and the rate value yielded by a blank containing no MoV1 is subtracted. moll-I) and 0.7 ml of 6.6 x Determination of Mo in Soils The soil is air dried, crushed, and sieved through a screen with 0.6 mm holes for removing debris before use. Portions of about 15 g of sample are placed in Soxhlet extraction cellulose thimbles (Whatman, Maidstone, Kent, UK) and extracted with 100 ml of water for 4 h. The cycle time for the extractors is about 15 min. The solution is then dried and the residue is incinerated at 500 "C. The ash is suspended in approximately 15 ml of boiling water, then filtered off and washed with boiling water, and the filtrate is diluted to 25 ml with water.Aliquots of this solution are then used for the kinetic spectrophotometric determination of MoV1. Results and Discussion Both the bathochromic shift in the maximum absorbance of thc triiodide ion from 350 to 500 nm and the increased absorptivity provided by the CPC micellar medium can be used to enhance the selectivity and sensitivity of analytical methods involving 12-I- system. Hence, it can be exploited for the determination of catalysts that accelerate the conversion of excess iodide into 13- by hydrogen peroxide, as is the case with the determination of MoV1. Fig. 1 shows the absorbance- time kinetic curves obtained for the Movl-catalysed iodide- hydrogen peroxide system in three media, namely, water (curve 1), CPC micelles (curve 2) and starch (curve 3).As can be seen by comparing the kinetic curves obtained, the micellar system provides enhanced sensitivity in the determination of MoV1. The variation in the absorbance of the 13--CPC associate with time includes an induction period, the logar- ithm of the length of which, expressed in time units, is proportional to the MoV1 concentration. Greater reproducibil- ity was obtained by using measurements of rate calculated from the slope of the kinetic curves; hence, this was the measured parameter used for determining MoV1. 0 3 6 Time/mi n 9 Fig. 1 Kinetic curves obtained for the Movl-catalysed iodide- hydrogen peroxide system in 1, an aqueous medium; 2, a CPC micellar medium; and 3, in the presence of starch.Curves were recorded at 1, 350; 2, 500; and 3, 560 nm. [MoV1] = 60 ng ml-' Study of the Experimental Conditions As the signal intensity and temporal stability of the 13--CPC associate depend on parameters" such as the temperature, iodide concentration and pH, some parameter values affecting the catalytic determination of MoV1 were found to behave differently from previously performed studies.35 Deviant variables were changed individually in turn to study their effects. Because the uncatalysed oxidation of iodide by hydrogen peroxide also proceeds in an acidic medium, the ratio of the catalysed to the uncatalysed reaction also required a selection of the experimental conditions of reaction. Fig. 2(a) shows the dependence of the rate of the MoV1- catalysed iodide-hydrogen peroxide system on the CPC concentration.The rates yielded by the blanks (containing no Nfovl) were subtracted before constructing the graph. The surfactant concentration affected both the reaction rate and the final absorbance obtained. Concentrations of CPC below 7 X rnol 1-1 could not be used as the absorbance increments obtained as a function of time were very small; hence, they provided irreproducible rate measurements. On the other hand, surfactant concentrations above 1.1 x mol I-' caused the reaction rate to decrease gradually, probably owing to dilution of reagents on the micellar surface. The spectral feature of the 13--CPC associate did not change over the surfactant concentration range assayed.A concentra- tion of 1 X moll-' was, therefore, chosen for subsequent experiments. The c.m.c. of CPC in the reaction medium was calculated from surface tension measurements performed with a stalagmometer (Brand, Germany). A value of 4.5 x mol 1-1 CPC was found. The concentration is lower than the c.m.c. calculated for distilled water (1.2 x mol 1-I) and the analytical concentration used (1 x rnol 1-1), which suggests that micelles are indeed formed under the experimen- tal conditions used. The dependence of the rate of the Movl-catalysed reaction on the iodide concentration is illustrated in Fig. 2(6). As can be seen, the rate increases linearly with iodide concentrations up to 4.5 X mol l-l, above which it remains virtually constant. Because iodide ion yields an insoluble salt with CPC, excess of this ion should be avoided in order to prevent precipitation.A 4.62 x 10-3 rnol I-' iodide concentration was selected in order to ensure maximum sensitivity. The effect of hydrogen peroxide concentrations in the range 0-1 x rnol I-' on the uncatalysed and catalysed reaction was similarly studied. Up to about 3.7 x mol l-l, the rate of both reactions increased linearly with hydrogen peroxide concentration, but no significant increase was observed above 4.5 x mol 1-1 [Fig. 2(c); Movl-catalysed reaction]. VJ a m I 0 7 . .c. I C 0 m U .- .c. 1 0 2 [CPC1/10-4 rnol I-' I ?I .t 0 3 6 9 12t 0 3 6 9 2 4 PH [H2021/10-4 rnol 1-1 w Fig. 2 Influence of the concentration of ( a ) CPC, (b) iodide, hydrogen peroxide, and (d) hydrogen ion on the rate of the Mo catalysed reaction between iodide and hydrogen peroxide.[MoV1] = 10 ng ml-1ANALYST, JUNE 1993, VOL. 118 717 Fig. 2(d) shows the variation of the rate of the MeV'- catalysed reaction with the hydrogen ion concentration. The effect of this parameter was studied over the pH range 0.7-5, by adjusting it with hydrochloric acid. The catalytic effect of Mo"' was only observed at pH values between approximately 1 and 5. Because the reaction was rather too rapid below pH 1.5, rate measurements performed by using conventional techniques were highly irreproducible. Even though higher acidities resulted in slightly enhanced sensitivities [Fig. 2 ( d ) ] , the rate of the blank reaction was also higher; hence, no real advantage in terms of the detection limit was gained.A pH of 2.5 was selected as a compromise between adequate sensitivity and reproducibility in the measurements. On the other hand, the formation of the 13--CPC associate was independent of pH over the range 2-10.11 Increased temperatures had an adverse effect on the absorbance of the 13--CPC associate, at 500 nm,ll which, however, remained virtually constant between 10 and 20 "C. On the other hand, as the Movl-catalysed reaction between iodide and hydrogen peroxide had a relatively small temperat- ure coefficient,l2 and the blank signal increased as the temperature increased, 20 "C was chosen as the optimum. The kinetic curves obtained for the reaction in the CPC micellar medium at temperatures above 20 "C were irreproducible and showed an atypical dependence of the absorbance on time.The effect of the ionic strength was previously found to be dependent on the salt used;4 as a general rule, the effect is to depress the reaction rate. The effect of this parameter was tested by adjusting it with sodium chloride and potassium nitrate; the former salt was found to have a more marked effect than the latter at equivalent concentrations for both salts. The effect is ascribed to an interaction4 between chloride and iodine ions. With the proposed procedure, however, no significant effect was observed for up to 0.1 mol I-' potassium nitrate or 0.03 mol 1-1 sodium chloride. The order in which the reactants were mixed also influenced the reaction rate. Hence, the greater sensitivity was obtained when CPC was added before iodide, which was added last in order to control the start of the reaction.Analytical Features of the Proposed Method A calibration graph for the determination of MoV' in the CPC micellar medium was established under the optimum condi- tions described above. The determination of this species was feasible over the range 2-150 ng ml- I. The standard error of the estimate (4.6 x s-I) and the correlation coefficient (0.999; n = 8) both indicated good linearity. The sensitivity achieved was (1.98 k 0.03) x ml s-l ng-l. The detection limit (30) was 1.3 ng ml-l, i.e., approximately 6-8 times lowcr than those of typical spectrophotometric methods based on the detection of triiodide ion at 350 nm.4,12 The precision, expressed as relative standard deviation, was 2.6% (n = 11) for 80 ng ml-l of MeV'.Earlier comprehensive interference studies on the determi- nation of MeV' by its catalytic effect on the iodide-hydrogen peroxide system3.5 showed it to be inadequately selective; hence, most of the methods concerned entail use of masking agents and/or separation procedures in order to remove interfering species. The interference most commonly encoun- tered in soil samples arises from the yellowish colour of the soil extracts, resulting from organic matter or hydrated iron(rr1) oxide; other species such as Fe"', TiIV, Vv, Wv' and Crvl, which also catalyse oxidation of iodide by hydrogen peroxide, and phosphate, which reportedly reduces the catalytic effect of Mo, also pose serious interference problems. The proposed method is not disturbed by yellowish soil extracts. On the other hand, as micelles enhance the selectivity of some kinetic methods,g710 the effects of the ions most likely to cause interference in soil samples were investigated.A given ion was considered not to interfere with the determination if the interferent plus analyte mixture yielded a signal in the range -tS,o, where S, is the signal provided by the analyte in the absence of interferent and o is the standard deviation of the method. Iron(ni) did not interfere at concentrations where it is soluble in the aqueous soil extracts (viz., between 1 and 1.5 pg ml-l). The effects of other ions tested are shown in Table 1. No significant selectivity enhancement for ions that catalyse the oxidation of iodide by hydrogen peroxide was achieved in the presence of micelles. The enhanced selectivity provided by the CPC micellar medium relative to the aqueous medium essentially arose from the bathochromic shift in the maximum absorbance of the triiodide ion in the micellar system, Determination of Mo"' in Soil Extracts The applicability of the proposed method was tested by determining the MoV1 content of soils of different texture (sandy, loamy and clayey). The results obtained in the determination of this species in six soil samples are sum- marized in Table 2 and compared with those obtained by using the thiocyanate method.13 As the MoV1 content in most of the samples studied was somewhat low, it could not be quantified by the thiocyanate method (only soils with MeV' concentra- tions greater than 0.08 pg g-1 can be analysed by this method). On the other hand, the proposed method permits the analysis of soils with MoV1 contents greater than 4 ng g-l.Recoveries were calculated in order to obtain soils with Mo contents sufficiently high for its determination by the thiocyanate method so as to detect if the triiodide-CPC method was subject to any interference. Several spiked samples were therefore prepared by adding aliquots (a few microlitres) of an Mo solution to homogenized soil samples that were subse- quently extracted. As shown in Table 2, the recoveries obtained by applying the triiodide-CPC method to soil samples, to which were added different amounts of Mo"', were satisfactory. The authors gratefully acknowledge financial support from the CICyT (Project No.PB91-0840). Table 1 Tolerated foreign ion concentrations in the determination of 0.06 pg ml-' of MoV1 Tolerated concentration/ Foreign ion pg ml-I C1-, Br- 1000 Zn" 300 Mg" 200 Mn" 100 Ca" 10 POI3 ~ 0.8 VV 0.3 Cr"', TiIV 0.1 WV' 0.0s Table 2 Determination of Mo"' in soil samples MoV1 content/pg g-1 Sample Added Soil 1 - 0.28 Soil 2 - 0.40 Soil 3 - 1 .00 Soii 4 - 0.50 Soil 5 - 0.75 Found by proposed method 0.034 0.310 0.035 0.439 0.031 1 .002 0.123 0.654 0.069 0.801 Found by Recovery thiocyanate (70 1 method - Undetectable 98.6 0.311 Undetectable 101 .0 0.440 - Undetectable 97.1 1.040 - 0.12s 106.2 0.631 - Undetectable 97.6 0.830 -718 ANALYST, JUNE 1993, VOL. 118 References 1 2 3 Feys, R., Devynck, J., andTremillon, B., Talanta, 1975,22,17. Altinata, A., and Perkin, B., Anal. Lett., 1973, 6, 667. Trojanowicz, M., Hulanicki, A., Matuszewski, W., Palys, M., Fursiewicz, A., Hulanicka, T., Raszewski, S., Szyller, J., and Augustyniak, W., Anal. Chim. Acta, 1986, 188, 165. 4 Quin, B. F., and Woods, P. H., Analyst, 1979, 104, 552. 5 Fang, Z.-L., and Xu, S.K., Anal. Chim. Acta, 1983, 145, 143. 6 Molybdenum in the Environment, eds. Chappel, W. R., and Petersen, K. K., Marcel Dekker, New York, 1976. 7 McIntire, G. L., CRC Crit. Rev. Anal. Chem., 1990,21, 257. 8 Sicilia, D., Rubio, S . , and Perez-Bendito, D., Talanta, 1991,38, 1147. 9 Sicilia, D., Rubio, S., and Perez-Bendito, D., Fresenius’ J. Anal. Chem., 1992,342, 327. 10 Lunar, M. L., Rubio, S., and Perez-Bendito, D., Talanta, 1992, 39, 1163. 11 Lunar, M. L., Rubio, S., and P6rez-Bendito, D., Anal. Chim. Acta, 1992, 268, 145. 12 Hadjiioanou, T. P., Anal. Chim. Acta, 1966,35, 360. 13 Snell. F., Photometric and Fluorimetric Methods of Analysis. Metals. Part 2, Wiley, New York, 1978, p. 1326. Paper 2106065 E Received November 16, 1992 Accepted February 8, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800715

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 7 19 Trial Measurements in Flow Analysis* Boaventura Freire dos Reis, Elias Ayres Guidetti Zagatto,+ Patricia Benedini Martelli and Sandra Maria Boscolo Brienza Centro de Energia Nuclear na Agricultura, Universidade de Siio Paulo, Caixa Postal 96, 134 16-000 Piracicaba, Sao Paulo, Brazil Trial measurements allow analytical results to be obtained by matching the sample volume with the dynamic range of the detector. In flow analysis, implementation of this approach can be accomplished by carrying out sequential injections or using concentration gradients. Trial measurements are performed using an increasing concentration sequence and the results are compared with a previously set threshold value. Once this value is surpassed, the remainder of the sample by-passes the detector and goes to waste.The potential and limitations in unsegmented and monosegmented flow systems are discussed. Improved systems for the automated determination of manganese in rocks by atomic absorption spectrometry are proposed. The typical measurement precision of flow injection was maintained and no baseline drift was observed. Results were in agreement with those obtained by inductively coupled plasma atomic emission spectrometry. Keywords: Trial measurements; flow analysis; atomic absorption spectrometry In analytical work, detection can be impaired after measure- ment of successive samples owing to the sample matrix being in contact with the detector; e.g., the continuous deposition of crystals on the burner of the atomic absorption spectrometer, which in extreme cases can cause clogging, which has often been reported for samples with a high salt content.1 This is also observed in flow-based methodologies, where the passage of several processed samples through the detector can sometimes cause a slow deterioration of measuring condi- tions, which is manifested as baseline drift, loss of sensitivity, increase in response time, etc.Efforts to minimize this drawback have been carried out by reducing the amount of sample in contact with the detector and/or the measuring time. Effects of matrix constituents on the performance of the detector are then reduced. In this context, trial measurements become an important approach. Trial measurements consist of selecting sample aliquots with different degrees of dilution, performing measurements on them according to an increasing concentration sequence, and comparing the results with a pre-set value.A reliable measurement is obtained when this value is surpassed. In flow analysis, trial measurements are accomplished efficiently and the remainder of the processed sample is discarded without flowing through the detection unit (Fig. 1). In this work, two different set-ups are proposed and critically examined, namely, sequential injection and gradient exploita- tion. In the former approach [Fig. l(a)], several sample plugs with different initial lengths are introduced into the same carrier stream. The sample zones established partially overlap with each other downstream, yielding a sample distribution with several concentration maxima, which reach the detector.Measurements then refer to peak maxima where concentra- tion gradients are less pronounced. After a reliable measure- ment has been achieved, a gate valve is switched and the sample portion still in the analytical path is directly discarded without flowing through the detection unit. For trial measurements based on gradient exploitation [Fig. l(b)], the sample is injected as a single plug, which undergoes dispersion, for successive trial measurements, before reaching the detector. When the analytical signal surpasses the magnitude of a threshold value, the gate valve is switched and the remainder of the sample is discarded directly. In order to * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992.+ To whom correspondence should be addressed. accomplish this approach, a high-speed data acquisition system is mandatory. In addition, the sample zone can be trapped and reversed, allowing aliquots with increasing concentrations to be delivered for detection. The sample portion, still trapped after a reliable measurement is achieved, is discarded directly. After evaluation of the potential and limitations of both strategies for trial measurements, the systems were applied to the determination of manganese in rocks. Experimental Solutions All solutions were prepared with analytical-reagent grade chemicals and freshly distilled, de-ionized water. A stock standard solution, 10.00 g I-' in manganese, was prepared by dissolving 7.690 g of manganese(I1) sulfate monohydrate in about 200 ml of water, adding 3.5 ml of concentrated nitric acid and diluting to 250 ml with water.Working standards between 5.00 and 5000 mg 1-l were also made up in 1% v/v HN03. - Scan T C 4 w -Scan S C V//A W Fig. 1 Trial measurements with ( a ) sequential injections and (b) gradient exploitation. S = sample, C = carrier stream, G = gate valve, D = detector, W = waste, T = pre-set threshold value. The shaded area represents the sample portion through the detector720 ANALYST, JUNE 1993, VOL. 118 Rock samples were decomposed in 40 ml PTFE [poly- (tetrafluoroethylene)] bombs following a procedure similar to that described by Langmyhr and Paus.2 Powdered rocks (about 500 mg) were accurately weighed, mixed with 1 ml of aqua regia (HCI-HN03, 3 + 1) and left at room temperature for at least 2 h.Then, 4 ml of hydrofluoric acid were added, the bombs were closed and heated in a digestion block at 130 "C for 2 h. After the bombs had been cooled and opened, 1 ml of perchloric acid was added, and the bombs placed again in the block, now at 210°C. After about 2 h, when evolution of white fumes was observed, the bombs were cooled. The residues were diluted to 20.0 ml with a 1.0% v/v nitric acid solution. Apparatus The peristaltic pump, automatic injector with electronic control, tubing and accessories of the flow system have been described previously.3 NResearch Model 161T031 3-way solenoid valves were employed. A laboratory-made micro- computer based on an 8085 microprocessor was used to control the valves and injector.4 Data acquisition was per- formed by using an analogue-to-digital (ND) converter pro- viding three measurements per second.Operating software was written in ASSEMBLER. A Perkin-Elmer Model 503 atomic absorption spectrometer was operated following the manufacturer's recommendations for maximum sensitivity with an air-acetylene flame.5 A damping factor of 1.3 s (TC2 setting) was set. Results and Discussion To study the potential of trial measurements in flow analysis, different set-ups were designed with the main reactor of the flow systems being connected to the aspiration tubing of the spectrometer, as already described.6 As the pneumatic sample aspiration rate was about 6 ml min-I, this rate was selected for the stream reaching the spectrometer.' Trial Measurements With Sequential Injections The system (Fig.2) was designed to use loop-based injec- tions.8 The sample is aspirated to fill different sampling loops, Li, and movement of the central sliding bar of the injector introduces the selected sample plugs into the carrier stream, C. The plugs, with increasing initial lengths and separated by coils B, and B2, partially overlap each other while travelling through coil B3, in such a way that a sample distribution with three concentration maxima reaches the spectrometer. Peak maxima are quantified following an increasing concentration sequence, and measurements are compared with a threshold value which was selected as an absorbance of 0.3. After this value is surpassed, the injector is switched, and the remainder of the processed sample is discarded without flowing through the detector.In view of the relatively narrow dynamic range of atomic absorption spectrometry, the system was set-up to provide three overlapping peaks with heights in a ratio of about 10 : 3 : 1. The distance between the injector and the spec- trometer (B4) was set to be as short as possible (15 cm) in order to minimize the amount of sample reaching the detector after cleavage of the sample zone. Coil B3 was defined as 50 cm in length. Preliminary tests indicated that with a shorter coil, a full reading of the maximum associated with loop L3 was not attained due to the damping factor of the detector plus accessories. With a 5 cm loop L3, the volumetric fraction x (sample relative to the volume contribution of the fluid element yielding the analytical readoutg) associated with the first peak was determined as 0.08.The length of coils B1 and B2 were selected as 100 and 50 cm to provide suitable separation between overlapping peaks, and L1 and L2 were experimentally fixed as 50 and 300 cm. Volumetric fractions related to the second and third peak maxima were calculated as 0.25 and 0.97. The system shown in Fig. 2 is remarkably stable. After a 4 h working period, only slight variations (4%) in the coeffi- cients of the calibration line equation were found and baseline drift was not observed. It should be stressed that manual operation of the system is possible. As the measurements refer to regions with less pronounced concentration gradients, the results are potentially less susceptible to variations in experi- mental conditions such as flow rates, timing, etc.Accuracy can be confirmed as can be seen in Table 1. For some samples two measurements were considered and an additional accuracy assessment obtained by comparing the results based on two calibration line equations. The sampling rate.depended on the number of trial measurements required for most of the samples assayed and the threshold value set. Considering the worst situation, in which three measurements per sample are required, about 100 samples can be run per hour. The sampling rate was improved by introducing air in tandem with the last plug to produce monosegmentation6 with a consequent reduction in washing time.Loop L1 was reduced to 30 cm without a pronounced increase in dispersion and the loop for air addition was set at 8 cm. As in earlier work,6 it was found that this parameter played no significant role on total dispersion. The air plug should not be increased: with a 50 cm plug, erratic results (relative standard deviation > 10%) were B 4 H20 W Fig. 2 Flow diagram of the system for manganese determination in rocks involving trial measurements without gradient exploitation. S, Sample aspirated at 4.0 ml min-*; L1, and L3, sampling loops; C, carrier stream. 1.0% v/v HN03, 6.0 ml min-'; B1, B2 and B3, coiled reactors; B4, transmission line; HzO, compensating stream; D. atomic absorption spectrometer; and W, waste. Alternative position of the injector specified by shaded area.For full details, see text Table 1 Comparative results. The data refer to the manganese contents in rocks (g kg-l) as determined with the proposed systems and inductively coupled plasma atomic emission spectrometry (ICP- AES). A and B refer to the flow systems in Figs. 2 and 3 , respectively. The numbers in parentheses are the estimated standard deviations (in %) based on four replicate measurements on the same digest. ICP- AES data are characterized by relative standard deviations of about 2%. Sample 1 2 3 4 5 6 7 A 4.25 (0.7) 0.55 (3.3) 3.48 (1.2) 14.90 (0.8) 87.30 (1.1) 4.35 (0.0) 4.79 (1.6) B 4.20 (1 .l) 0.57 (2.9) 4.30 (0.3) 3.43 (1.7) 14.30 (0.2) 86.93 (0.5) 4.72 (0.7) ICP-AES 4.17 0.58 4.33 3.50 14.52 83.36 4.70ANALYST, JUNE 1993, VOL.118 72 1 observed. It should be emphasized that monosegmentation becomes a more attractive proposition when saline effects are more pronounced. However, as the different degrees of sample dispersion are achieved with different loop lengths, the system lacks versatility. Trial Measurements With Gradient Exploitation Concentration gradients inherent to sample dispersion were also exploited to implement trial measurements. Initially, efforts were made to use a single-line system. The injected sample underwent dispersion and reached the detector at which point successive trial measurements were initiated. This configuration, although potentially beneficial for other appli- cations, was not examined further because preliminary tests involving spectrophotometry emphasized the poor measure- ment reproducibility inherent to regions at the front portion of the dispersed sample.'" The system was then modified so that the flow was reversed after sample dispersion and the trailing portion of the sample zone (associated with a longer travelled path and usually with better measurement precision) arrived at the detector earlier.However, data could not be collected properly because the sample zone passed through the detector too quickly. This drawback was circumvented by trapping the processed sample, re-sampling aliquots with increasing concentrations and directing them back towards the detector. In this system (Fig. 3), the injector V" is used for conventional loop-based injection and for zone trapping, and the three valves are used for directing the stream.In the situation specified, the sample is aspirated through loop L as the previous sample is being processed. When the injector is switched to the alternative resting position, the selected sample plug is introduced into stream C and trapping coil T is placed in the same path. After a time interval t l , when the sample zone is flowing through T, the injector is switched back to the position specified in Fig. 3, and the dispersed sample is trapped inside T. Valve V2 directs stream C towards the confluence points y and x. By successive switching of this valve, the slow flow rate of C' (1.0 ml min-I) delivers slices of the trapped zone with different concentrations to the detector. A total flow rate of 6.8 ml min-' reaching the spectrometer is C 1 w 1 C' A C" Fig. 3 Flow diagram of the system for manganese determination in rocks involving trial measurements with gradient exploitation.L, Sampling loop; T, trapping coil; C, C' and C", 1 .O% vlv HN03 carrier streams; Vo, injector; V1, V2 and V,, three-way valves; x, y , z and t , confluence points. Other symbols as indicated in Fig. 2 maintained by adding the confluent stream C" (5.8 ml min-1) either at confluence point y or t. Both valves Vz and V3 are switched simultaneously and timing is selected so that the aliquots delivered were not diluted at the confluence points. After attaining a reliable signal, the injector is switched again and the remainder of the sample zone is discarded simul- taneously on injection of the next sample.In order to minimize carryover effects, valve V1 is switched immediately before each movement of the injector. If the threshold value is not attained, the most concentrated portion of the trapped sample is delivered to the detector. The lengths of L, B1 and T were selected as 10, 20 and 200 cm to permit trapping of the entire sample zone. With a flow rate of C of 3.2 ml min-l, tl was experimentally set as 15 s. In this situation, the trapped sample was very dispersed (x < 0.2): for other applications requiring limited sample disper- sion, the initial volume injected should be increased and only the trailing edge trapped. The aliquoting process can be initiated any time after t,: here, a 0.5 s delay was chosen. Resting times of V2 and V3 in the alternative position relative to Fig.3 ( t ' ) defined the volume of the sample zone delivered to the detector. When this interval was too short (<1 s) measurement reproducibility was not acceptable and dispersion inside B2 and B3 was too pronounced. As the value of t' increased, the number of trial measurements per sample diminished and sample dispersion inside B2 and B3 became less pronounced (Fig. 4). The lessening in the dispersion is explained by the larger sample plug introduced into C'. It should be stressed that for lower t' values, the shape of the sample distribution inside T is better emphasized. For extreme values of t' > 50 s, the entire trapped zone was delivered without being sectioned. More- over, valve V2 became ineffective because portions of the long sample slice arrived simultaneously at confluence points x and y at the moment of valve switching.With t' values of 3, 7 and 10 s, three measurement conditions were defined and exploited for manganese deter- mination. In order to avoid carryover, the resting time in the position specified in Fig. 3, t", was kept as 12.8 s. It should be stressed that the distance between the trapping coil T and confluence point x can constitute an additional carryover source. Here, the effect was avoided by drilling a hole on the external plate of the injector for the confluence connection (Fig. 3 ) . With a length for B2 of 100 cm, good mixing conditions were attained and the simultaneous presence at x and y of portions of the dispersed slice was avoided. The length of B3 was as short as possible (20 cm).In this situation, the rate of the trial measurements was determincd as about 230 h-I. System stability, measurement reproducibility and accuracy for the system shown in Fig. 2 were also observed (see also Table 1). Monosegmentation is not compatible with the 4 min - a C (D e 0.15 - 2 0 - -Time Fig. 4 Influence oft' value. From right, set of peaks corresponding to t' values of 3,4, 5 and 7 s. Figure refers to flow system shown in Fig. 3 with injection of a 200.0 mg 1- * solution of Mn and a t" value of 12.8 sANALYST, JUNE 1993, VOL. 118 L 4 min - A -Time Fig. 5 Calibration plots obtained with the flow system shown in Fig. 3. From the right, peaks obtained after triplicate processing of: 10.0, 25.0, 50.0, 100.0, 200.0, 400.0 and 600.0 mg I-' of Mn standard solutions with t' values of 3 .7 and 10 s system shown in Fig. 3 because zone trapping is involved: the presence of an air phase inside coil T causes a loss of reproducibility of the volumes delivered due to compression effects. With the system shown in Fig. 3, versatility is improved in view of the easy modification of the timing involved. This potential enables the system to be used also for wide range spectrophotometry using gradient exploitation, as several calibration plots are easily definable (Fig. 5 ) . Sampling rate was improved relative to the system shown in Fig. 2 as cleavage occurred immediately after the reliable measurement was achieved and not after reaching the peak maximum. As far as gradient exploitation is concerned, the system shown in Fig.3 is potentially more susceptible to variations of experimental conditions. Hopefully, this drawback was not verified. As several measurements can be considered per sample, the final results for some samples are average values, thus they are intrinsically more reliable. Studies focusing on this point are presently in progress. Conclusions Flow systems are excellent set-ups for solution management and, therefore, attractive for trial measurements, which require on-line aliquoting/measuring processes. The system described here [Fig. l ( a ) ] exploits sequential loop-based injections and has been used for routine analysis without any major problems being noted. However, for particular sample lots with high variability in analyte concen- tration, the system lacks versatility.With gradient exploitation, the peaks are also quantified, so that no modifications to the detector are needed. Although potentially more susceptible to variations in experimental conditions, the system provides results as precise as those obtained with the sequential set-up. The capacity of the routine analytical laboratory underwent a pronounced expansion after implementation of the trial measurement approach because manual dilutions for trial assays of unusual samples are no longer required. In addition, maintenance of the equipment became less frequent as measurements are associated with reduced amounts of sample reaching the detector. Partial support from CNPq (Conselho Nacional de Desenvol- vimento Cientifico e Tecnologico) is greatly appreciated. The authors thank A. N. Araujo from Porto University for critical comments. 1 2 3 4 5 6 7 8 9 10 References Fang, Z., Xu, S., Wang, X., and Zhang, S . , Anal. Chim. Acta, 1986, 179, 325. Langmyhr, F. J., and Paus, P. E., Anal. Chim. Acta, 1968,43, 397. Krug, F. J., Bergamin Fo, H . , and Zagatto, E. A. G., Anal. Chim. Acta, 1986, 179, 103. Reis, B. F., GinC, M. F., and Kronka, E. A. M., Quim. Nova, 1992, 15, 231. Analytical Methods ,for Atomic Absorption Spec fro photometry, Perkin-Elmer, Norwalk, CT, 1973. Reis, B. F., Arruda, M. A. Z., Zagatto, E. A. G., andFerreira, J. R., Anal. Chim. Acta, 1988, 206, 253. Brown, M. W., and RiiiiEka, J., Analyst, 1984, 109, 1091. Zagatto, E. A. G., Gink, M. F., Fernandes, E. A. N. Reis,B. F., and Krug, F. J., A n d Chim. Acta, 1985, 173,289. Zagatto, E. A. G., Reis, B. F., and Bergamin Fo, H . , Anal. Chim. Acta, 1989, 226, 129, Tuon, R. L., M.Sc. Thesis, Universidade de S5o Paulo, Piracicaba, Brazil, 1989. Paper 210641 5 D Received December I , 1992 Accepted February 8, 1993

ISSN:0003-2654    

DOI:10.1039/AN9931800719

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 723 ’ Immobilization of Glutamate Oxidase on Non-porous Glass Beads. Automated Flow Injection System for the Assay of Glutamic Acid in Food Samples and Pharmaceuticals* 0.040 ml min-’ I Constantine D. Stalikas, Miltiades 1. Karayannis and Stella M. Tzouwara-Karayannit Department of Chemistry, University of loannina, loannina, 45 I 10 Greece An enzymic method is proposed for the determination of glutamic acid in food samples and pharmaceuticals. The method is useful in the range 0.01-0.5 mmol I-I, with a detection limit of 0.005 mmol I-1 for an injection volume of 52 1.11. L-Glutamate oxidase from Streptomyces sp. X-I 19-6 was immobilized on non-porous glass beads, and the H202 produced was caused t o react with Trinder’s reagent. Various parameters were studied for the establishment of the optimum operating conditions for an in-house flow injection manifold with a spectrophotometric detector. Many interfering species and several amino acids were tested in order t o verify the specificity of the enzyme reactor.Experimental data for interfering organic and inorganic substances are reported. The system works very selectively for glutamic acid and glutamates. The method is ideally suited t o the assay of glutamates in a large number of samples within a single working day, as the frequency is 40 samples h-I. The relative standard deviation is 1.8% for six assays. The immobilized reactor is stable for a long period under proper conditions of storage. The accuracy of the proposed method was tested by comparison of the results with those of the Association of Official Analytical Chemist‘s method and with the manufacturer’s specifications for the analysed samples, where available.Good agreement was obtained. Recovery studies yielded results between 97 and 105%. Keywords: L-Glutamic acid; glutamate oxidase; single-bead string reactor; flow injection; food and pharmaceutical samples Glutamic acid is an important ingredient of various foods, such as soups, seasonings and meat products, being used as a flavour enhancer, and is also present in many biological materials and pharmaceuticals. It plays an important role in neurological pathways of the central nervous system1 and in the pathophysiology of mental disorders.2 In the past 10 years, many methods have been developed for the reliable and accurate determination of low concentrations of glutamic acid in small samples.The more recently isolated specific flavin adenine dinucleotide (FAD)-glutamate oxi- dase3 (GLOD) has been used in packed-bed reactors4 and enzyme electrode@-7 with relative success. The novelty of our method is in the immobilization of the above-mentioned enzyme on non-porous glass beads for the construction of single-bead string reactors (SBSRs) with concomitant advantages such as limited dispersion, longer residence time, low consumption of the expensive enzyme and excellent conformation to flow injection (FI) systems.819 The Trinder reactionlo was used for the entrapment of H202, generated by the enzymic reaction, by means of a reagent consisting of 4-aminoantipyrine (AAP), 3,5-dichloro-2-hyd- roxybenzenesulfonic acid (DCPS) and peroxidase (PO), which has been extensively used in the past for the enzymic determination of glucose. The quinoneimine dye was then monitored at 510 nm in a flow-through cell.The soluble PO used in the system does not reduce the economical feature of the method, owing to its relatively low cost and consumption. Several parameters were studied and the optimum operating conditions for the system were carefully established. These were flow rate, pH, temperature, concentration of the Trinder reagent, composition of the buffer solutions and conditions for immobilization. * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992.+ To whom correspondence should be addressed. Experimental Apparatus Fig. 1 is a schematic diagram of the experimental set-up. The in-house FI system consisted of a four-way pneumatically actuated injection valve (Kheodyne Type 50 Teflon, Cotati, CA, USA), an eight-channel peristaltic pump (Ismatec, Glattbrugg, Zurich, Switzerland) and a filter spectropho- tometer11 equipped with a fibre optic for the transmission of the light from the source to the flow-through cell (2 pl) for measurement. The temperature of the reagents and of the reactor was regulated by a thermostatically controlled water-bath. Data processing and collection was performed with an IBM-com- patible PC by means of software written in Microsoft Q-Basic. The interface unit was an RTL 800/815 multifunction Input/ Output Board. Reagents All the chemicals used were of analytical-reagent grade, and the solutions were prepared in doubly distilled water (DDW).Glutamate oxidase (E.C. 1.4.3.11) from Streptomyces sp., 5 U mg-l solid (1 U = 16.67 nkat), was a gift from S ? P I I I - Waste Fig. 1 Schematic representation of FI manifold employed for glutamic acid determination. GLOD, glutamate oxidase; PR, plain reactor; C, carrier; R, reagent; P, pump; S, sample; D, detector; and PC, personal computer724 ANALYST, JUNE 1993, VOL. 118 Yamasa Shoyu (Chiba-Ken, Japan). Amino acids, glutar- aldehyde and 3-aminopropyltriethoxysilane were purchased from Sigma (St. Louis, MO, USA). The Trinder reagent was prepared by mixing the desired amount of horseradish peroxidase (2.5 mg ml-l) (Sigma) and volumes of 20 mmol I-' AAP (0.125 ml) (Sigma) and 20 mmol I-1 DCPS (0.470 ml) (Sigma) and then diluting with the appropriate working buffer solution.Glutamic acid standards were prepared from a 0.01 mol I-' stock solution by dilution with buffer. Preparation of Samples For soups and pharmaceuticals, a 0.2-0.4 g sample was accurately weighed and homogenized in a mortar. For meat products, a 5 g sample was mixed with sand and homogenized in a mortar. In both instances the sample was transferred into a 100 ml beaker with 35 ml of water at room temperature, and sonication was applied until all the water-soluble substances had dissolved. The sample was then vacuum-filtered, the residue on the filter was washed several times with DDW, and the filtrate plus washings were collected in a 50 ml calibrated flask.Procedure for Immobilization of the Enzyme The immobilization of the enzyme is described elsewhere.12.13 The immobilization procedure consists of the following steps. The glass beads were cleaned with ethanolic 5% m/v potassium hydroxide, then treated with a saturated solution of ammonium hydrogen fluoride in methanol and kept at 450 "C for 3 h to achieve roughness of their surface. They were then silanized with 1% v/v 3-aminopropyltriethoxysilane in acetone for 15 h at 70 "C. Finally, the silanized beads were equilibrated with a 2% v/v solution of glutaraldehyde in 0.05 mol I-' phosphate buffer at room temperature. This treatment helps the enzyme to be efficiently attached to the beads when they are placed in a solution of glutamate oxidase for 24 h at 4 "C.The data presented in Fig. 2 (curve B) were taken at pH 7, which is the optimum value for the immobilization of the enzyme. Procedure for Analysis The sample solution containing glutamic acid or glutamate is injected into the carrier stream (phosphate buffer, pH 7.8) and flows through the SBSR composed of treated non-porous glass beads, 0.6 mm diameter, with the immobilized GLOD in a 0.8 mm i.d. Teflon tube. Here, the H202 is produced according to the reaction: G LOD/m g 0.36 \ I 0.30 I I I I I 6.0 6.5 7.0 7.5 8.0 8.5 PH Fig. 2 Effect of A, pH and B, the amount of glutamate oxidase on the efficiency of immobilization. Glutamic acid standard solution, 0.2 mmol 1-1 GLOD L-glutamic acid - a-ketoglutarate + NH3 + H202 (1) H20 + 0 2 The new reaction mixture meets, at the output of the enzymic reactor, the Trinder reagent (R) in phosphate buffer solution and proceeds through a second reactor (PR), which contains untreated glass beads, with a length offering suffi- cient time for the thorough mixing of the new reacting solution and the complexation of the reaction: (2) PO H202 + AAP + DCPS + Reddye Finally, the reaction mixture reaches the cell of the spectrophotometer, where the maximum of the absorbance- time curve, A,,,, is measured automatically at 510 nm.The A,,, of the absorbance curve provides a measure of the concentration of glutamate in the sample. Results and Discussion To the best of our knowledge, glutamate oxidase was immobilized for the first time on non-porous glass beads during this work.The optimization of the immobilization was studied on the maximum amount of the enzyme immobilized on the glass beads and on the pH of the buffer solution used in immobilization. Fig. 2 (curve A) reveals that the optimum is around pH 7.0, with use of 0.1 rnol 1-1 phosphate buffer. The immobilization of the enzyme was performed by transferring 0.12 g of pre- treated non-porous glass beads into 1 ml of buffer solution bearing different amounts of the GLOD each time. Fig. 2 (curve B) shows that about 1.0 mg ml-I of the enzyme is adequate to saturate the surface of the beads under the above conditions. A 0.12 g amount of glass beads is sufficient for fabricating three 5 cm reactors. A calibration graph for the soluble enzyme was used for assessment on the efficiency of immobilization.The supernatant phase, containing the non- immobilized enzyme, was analysed to locate the absorbance on the calibration graph. It was found that 82% of the enzyme was attached to the beads. Conditions for the determination of glutamic acid were optimized by studying the effect of various parameters. For the optimization of the analytical system the univariate approach was applied, in which one parameter was varied while all the others were kept constant. The pH of the system was first investigated. Several types of buffers of different concentrations were used. Fig. 3 shows that phosphate buffer of pH 7.8 is well suited for the operation of the system. The above buffer with a concentration of 0.05 mol I-1 results in relatively high peaks, maintains the pH of the system and is easily handled.From Fig. -3, it is also apparent that the experimental system works effectively over a wide pH range. With Trizma buffer, 0.1 mol l-l, the peak is 0.20 , I I I .$ 0.10 x0.15 1 + + 0.05 1 0 1 5 6 7 8 9 10 11 PH Fig. 3 phosphate; A , Trizma; 0, carbonate; and 0, borate Effect of pH on the response ofethe system. W , Acetate; +,ANALYST, JUNE 1993, VOL. 118 725 slightly higher, but this buffer is not stable. Other buffers based on triethanolamine and imidazole, working in the same pH range as phosphate, yield lower peaks. An insignificant change in the peak height was observed by changing the concentration of phosphate from 0.05 to 0.5 mol I-'. A 5 cm reactor length yields acceptable absorbance values, A,,,, while minimizing the back-pressure in the flowing stream.Increase in the length of the reactor creates a broadening of the peak shape and a decrease in the peak height. The length of the reactor (PR), filled with plain beads, was 40 cm. A sample volume of 52 p1 was chosen for the measurements as no increase in the response was observed by further increasing the volume. The response of the system increases and approaches a plateau with peak shapes acquiring 'shoul- ders'. Several flow ratios were also studied, keeping the flow rate constant. A 16 : 3 ratio (buffer carrier: Trident reagent) was selected as both low consumption of the reagent and dilution of the sample was compensated for by satisfactory peak height. An over-all flow rate of 0.250 ml min-1 reconciles fairly high sensitivity, good reproducibility and satisfactory sampling throughput (40 samples h-l).Higher flow rates result in rapid decrease in peak heights, while lower rates minimize the frequency of sampling. Concentrations of AAP, DCPS and PO in the Trinder reagent were adopted from Stults et aZ.12 with a small modification in the concentration of PO, which was 0.4 instead of 0.8 mg ml-l. The final variable studied was the temperature of the SBSR. The bath temperature was adjusted and the sample was injected after thermal equilibrium had been reached in the system. The sensitivity increased sharply with temperature, reaching a plateau between 35 and 52 "C and then decreasing again rapidly, probably as a result of denaturation of the enzyme.The specificity of the enzymic reactor was also studied. Various amino acids were investigated at concentrations of 1 mmol 1-1 as was a mixture of them. L-Glutamate was oxidized almost exclusively. Only L-glutamine interfered slightly at concentrations above 20 mmoll-l. Many inhibitors were also tested for their influence on the measurements. It was found that the immobilization of the enzyme on the glass beads minimizes the effect of many inhibitors compared with that found for the soluble enzyme. The enzymic reaction is strongly and irreversibly inhibited by Ag+ and Hg2+. Glutathione and ascorbic acid, as reducing compounds, are sources of serious interference. 14,15 The results are presented in Table 1. Standard solutions of glutamic acid were prepared in 0.05 mol I-' phosphate buffer, pH 7.8.Under the optimum working conditions very good linearity was achieved between A,,, and glutamic acid in the concentration range 0.01-0.5 mmol 1-l. The correlation coefficient was 0.9999. The maximum response of the system was achieved in the temperature range 35-52 "C. If better sensitivity of the working curve is required, temperatures between 35 and 52 "C can be applied. The data presented in this work were obtained at 25 "C. The SBSR retains its activity after a 2 month working period and use on about 1300 runs, when stored at 4 "C. In the first 2 d, 20% of its original activity was lost, probably as a result of removal of physically adsorbed enzyme on the surface of the glass beads. The reactor needed more frequent calibration during the first 2 d of use.The method was applied to food samples and pharmaceuticals. The results for various real samples are summarized in Table 2. The method was compared with the standard Association of Official Analytical Chemist's (AOAC) method16 and good agreement between results was attained. All assays were performed, after appropriate dilution of the samples, by both direct and standard addition procedures. No matrix interference was observed, as the dilution of the samples was higher than 1 + Table 1 Interference effects of various compounds on the assay of glutamic acid in standard solution of 0.1 mmoll-'. The concentration of the interfering compound was 1 mmol 1-'; working temperature 25 "C Interfering Relative activity compound (Yo ) None EDTA* 2,2'-Bipyridyl a-Ketoglutaric acid Glutathione Ascorbic acid Starch KCl NaCl cuc12 BaClz ZnS04 FeS04 CaC12 NaN3 HgC12 &NO3 Strong alkaline solution * EDTA = Ethylenediaminetetraacetic acid.100.0 100.7 95.4 100.3 93.0 46.1 25 .0 90.0 104.8 101.3 100.0 95.9 97.1 102.1 93.5 0.0 0.0 20.0 Table 2 Determination of glutamic acid in various real samples Glutamic acid (YO d m ) Proposed AOAC Manufacturer's Product method* methodt specification Tomato cream Beef cube Beef cream Chicken cube Mushroom sauce Zwan (canned Your Life tablet* Brain tabletVmg per meat) tablet 5.46 6.65 1.55 8.54 3.23 0.04 8.6 77.0 5.43 - 6.82 - 1.48 - 8.68 - 3.36 - 0.04 - - 8.2 - 75 .0 * Average value of three runs. t Ref. 16. J; P. Leiner Nutritional Products, Torrance, CA, USA.Power Health Products, Pocklington, York, UK. Relative error (YO) +0.5 -2.5 +4.7 -1.6 -3.9 0.0 +4.8 +2.7 Table 3 Recovery of L-glutamic acid in some samples obtained by the proposed analytical procedure.* Working temperature 25 "C Added Taken Found Recovery Sample (Yo 1 (Yo 1 (Yo 1 (Yo) Tomato cream 5.73 11.19 11.14 99.5 3.27 8.73 8.85 101.4 Beef cream 1.15 2.70 2.75 101.8 0.72 2.27 2.39 105.3 Brain tablet 7.4t 84.4t 84. 0t 99.5 4.2+ 81.2t 79.0t 97.3 * The difference between the values in columns 3 and 2 is the t Values in mg per tablet. original concentration of L-glutamic acid in the analysed sample. 20. The relative standard deviation of the method was 1.8% for a standard glutamate solution of 2 x mol 1-l (n = 6). The accuracy was also checked with recovery studies by adding standard glutamic acid solutions t o appropriate real samples.Recoveries of 97-105% were achieved, as shown in Table 3. The authors thank Dr. H. Kusakabe (Yamasa Shoyu Co.), who kindly provided a gift sample of the enzyme glutamate oxidase (from Streptornyces sp.). C. D. S. thanks the 'G. Stavrou' Foundation and the European Environmental Research Institute for financial support.726 ANALYST, JUNE 1993, VOL. 118 References 1 Perry, T. L., Hansen, S., and Kennedy, J . , J. Neurochem., 1975, 24, 587. 2 Engelsen, B., Acta Neurol. Scand., 1986, 74, 337. 3 Kusakabe, H., Midorikawa, Y . , Fujishima, T., Kuninaka, A., and Yoshino, H . , Agric. Biol. Chem., 1983, 47, 1323. 4 Yao, T., Kobayashi, N., and Wasa, T., Anal. Chim. Acta, 1990, 231, 121. 5 Yamauchi, H., Kusakabe, H., Midorikawa, Y., Fujishima, T., and Kuninaka, A., paper presented at the Third European Congress on Biochemistry, 1984, vol. I, p. 705. 6 Villarta, R., Cunningham, D., and Guilbauit, G., Talanta, 1991,38. 49. 7 Vahjen, W., Bradley, J., Bilitewski, U., and Schmid, R., Anal. Lett., 1991, 24, 1445. 8 Valcarcel, M., and Luque de Castro, M. D., in Flow Injection Analysis, eds. Chalmers, R. A . , and Masson, M., Ellis Horwood, Chichester, UK, 1987, p. 84. 9 10 11 12 13 14 15 16 Reijn, J . , van der Lindern, W., and Poppe, H., Anal. Chim. Acta, 1981, 123, 229. Tnnder, P., Ann. Clin. Biochem., 1969, 6 , 24. Patton, C., and Crouch, S., Anal. Chim. Acta, 1986,179, 189. Stults, C., Wade, A., and Crouch, S., Anal. Chim. Acta, 1987, 192, 155. Kiranas, E., Karayanni-Tzouwara, S., and Karayannis, M. I . , Acta Chim. Hung., 1992, 129, 461. White-Stevens, R., Clin. Chem. (Winston-Salem, N. C . ) , 1982, 28, 578. White-Stevens, R., and Stover, L., Clin. Chem. (Winston- Salem, N.C.), 1982,28, 589. OfJicial Methods of Analysis of the Association of OfJicial Analytical Chemists, AOAC, Washington, DC, 12th edn., 1975, Sections 20.149-20.151. Paper 21052 90C Received October 2, 1992 Accepted November 30, 1992

ISSN:0003-2654    

DOI:10.1039/AN9931800723

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 727 Substitution of Peroxidase in Trinder's Reagent with Iron(ii) for the Determination of Hydrogen Peroxide in Enzymic Reactions by Applying Flow Injection" Efstratios R. Kiranas, Stella M. Tzouwara-Karayanni and Miltiades 1. Karayannist Department of Chemistry, University of loannina, loannina 451 10, Greece The kinetics and mechanism of the reactions between 4-aminoantipyrine and/or 3,5-dichloro-2-hydroxyben- zenesulfonic acid with Fez+ in the presence of H202 have been investigated. A differential spectrophotometric flow injection (FI) method for the determination of H202 has been developed, by substituting horseradish peroxidase (PO) in Trinder's reagent with Fez+. This method has been applied to the determination of glucose in human serum, with good results, by using an enzymic reactor with immobilized glucose oxidase and an automated FI system, after optimizing its parameters.The agreement between the results of this method with those of the routine method used by a hospital clinical laboratory for the determination of glucose in human serum was very good. The correlation line of the two methods shows a slope of 1.03 and a correlation coefficient ( r ) of 0.9997. The method was also applied to the determination of H202. The working curve, with Fe2+ as promoter, was linear for 10-400 pmol I-' H202 compared with 5-170 pmol I-' H202 when PO was used. The slopes of the curves for Fe2+ and PO are 1.69 and 4.35 A I mmol-l, respectively, showing a reduced sensitivity for the proposed method compared with the PO method.Keywords: Flow injection; glucose; serum; replacement of peroxidase About 100 years ago, Fenton reported that Fez+ strongly promotes the oxidation of malic acid by H202.1 He proposed for such oxidations the so-called Fenton's reagent (FR) , consisting of a mixture of H202 and Fez+ in aqueous solution. Forty years later, Haber and Weiss2 suggested that the hydroxyl radical is the actual oxidant in such systems. According to Walling's model,3 oxidations with FR involve many steps with a variety of products, depending on the experimental conditions. Barham and Trinder have shown,4 that, when H202 reacts with a mixture consisting of 4-aminoantipyrine (AAP), 3,5- dichloro-2-hydroxybenzenesulfonic acid (DCPS) and horseradish peroxidase (PO), a coloured dimer is produced, which absorbs at 510 nm with a molar absorptivity of 2200 m2 mol-I. This reaction has been used extensively by many workers5.6 for the determination of H202 produced during various enzymic reactions.The soluble peroxidase used in Trinder's reagent (TR) is its critical feature because, in routine use in batch experiments, PO increases the cost of a single run. In these studies, Fez+ is used to replace PO in TR as in both instances the action of the species is catalytic. The reactions of FR with AAP and/or DCPS were studied and absorption spectra of the reaction products were recorded 30 s after mixing of the reagents. Experimental Reagents All chemicals were of analytical-reagent grade. TriseHCl buffersolution, 0.05 mol 1-1 (pH7.0). Prepared by mixing in a 100 ml calibrated flask 50 ml of 0.1 mol 1-l tris(hydroxymethy1)methylamine (Tris) and 46.6 ml of 0.1 mol 1-1 HC1 and diluting to volume with water.Potassium hydrogen phthalate buffer solution (KHP) , 0.05 mol 1-l (pH 2-6). Prepared by mixing in a 100 ml calibrated flask 50 ml of 0.1 mol 1-1 KHP solution and the appropriate volume of 0.1 mol 1-1 HC1 and diluting to volume with water.7 Reagent solution. In a 10 ml calibrated flask, dissolve 0.8 mg * Presented at the 4th International Symposium on Kinetics in Analytical Chemistry, Erlangen, Germany, September 27-30, 1992. + To whom correspondence should be addressed. of PO (EC 1.11.1.7; Sigma, St. Louis, MO, USA), in water add 1 ml of 0.01 moll-' AAP and 1 ml of 0.01 mol 1-1 DCPS and dilute to volume with Tris-HC1 buffer (pH 7.0).Reagent 1 ( R l ) . In a 100 ml calibrated flask dissolve 0.477 g of AAP (Sigma) in 0.1 moll-' HCl (stable for 7 d at 4°C) and dilute to volume. Reagent 2 (R2). In a 100 ml calibrated flask, dissolve 2.491 g of DCPS (Sigma) and 1.568 g of Fe(NH4)2(S04)2-6H20 (Sigma) in 0.01 mol I-' HCl (stable for 7 d at 4°C) and dilute to volume. Standard glucose solution. Weigh an appropriate amount of P-D( +)-glucose (Sigma) into a 10 ml calibrated flask and dilute to volume with Tris-HCl buffer (pH 7.0). Human serum. The pre-analysed blood plasma samples from patients were obtained fresh from the University of Ioannina Hospital. Deproteinization of the samples was essential because the low diffusion velocities and the Tyndall effect of protein macromolecules resulted in high ,blanks and double-peak shapes during the flow injection (FI) measure- ments. Single-bead string reactor (SBSR).The non-porous glass beads were 0.5 mm in diameter. After immobilization of glucose oxidase (GO) from Aspergillus niger [EC 1.1.3.4; 15 000-25 000 U g-l solid (1 U = 16.67 nkat); Sigma], the glass beads were inserted into the Teflon reactor tubes (0.8 mm i.d.; Gilson, Worthington, OH, USA) by suction with a glass syringe.8~9 The reactors were filled with Tris-HC1 buffer (pH 7.0) and stored in a refrigerator. Plain SBSR (PR). This reactor, containing untreated plain glass beads and fabricated in a similar manner to the enzyme reactor, controls the completion of the reaction and improves the value of the dispersion coefficient.Equipment Kinetic measurements were performed in a filter spectropho- tometer [Metrohm (Herisau, Switzerland) Model 6621 equipped with fibre optics and an immersion-type optical cell. Data acquisition was performed with an IBM data-acquisition and control adapter and an IBM-compatible personal com- puter. The data-collection and acquisition software allows: ( i ) the automated collection of 2500 data points on the reaction curve, (ii) the plotting of the reaction curve A, = f(t), (iii) the728 ANALYST, JUNE 1993, VOL. 118 R1 R2 4 0.600 Waste 0.480 a, 0 0.360 9 E < 0.240 Fig. 1 Schematic diagram of the experimental set-up of the FI system for glucose determination. PP, Peristaltic pump; IV, injection valve; D, detector (spectrophotometer); PC, personal computer; GO, single bead string reactor with immobilized GO, 10 cm; PR, single bead string plain reactor, 25 cm; and C, carrier, Tris-HCl buffer, pH = 7.00.R1, AAP in 0.1 mol 1-l HCl; R2, DCPS + Fez+ in 0.1 moll-' HCl calculation of kobs, and (iv) the measurement of the slope dA/dt at any point on the reaction curve. The FT measurements were performed with an automated unit, constructed in-house and consisting of an eight-channel peristaltic pump, a four-way pneumatic rotary-injection valve and a filter spectrophotometer equipped with fibre optics and a 2.0 p1 microcell.10 The unit also contains the SBSRs mentioned above: one 10 cm bearing immobilized GO and another 25 cm PR, connected in series in order to retard the reaction mixture. The reaction was optimized for the ratio of the reagent concentrations, pH and temperature, in order to apply it to the kinetic determination of H202 and to FI.The interface unit RTI-800/815 multi-function input-output board was used for the collection of 120 data points on the FI reaction curves during a period of 60 s. The FT experimental set-up is shown in Fig. 1. Procedures Deproteinization of Serum Samples Two methods for the deproteinization of the sample were applied, with comparable results. Deproteinization with ZnS04 and Ba(OW2 A (1 + 4) mixture of solutions of ZnS04 (15% d v ) and Ba(OH)2 (5% d v ) yields a pH of 7.46, which is the optimum pH for deproteinization.11 The serum sample (1 ml) is mixed with 150 pl of the ZnS04 solution and then 600 1.11 of the Ba(OH)2 solution are added.After centrifugation the sample still retains an absorbance blank of 0.023, which interferes with the assay. Additional dilution (1 + 8) of the sample with Tris-HCl buffer (pH 7.0) reduces the absorbance of the blank to 0.003 and allows FI measurements for the determination of glucose in the sample. Deproteinization by gel filtration A Sephadex G-25 superfine column (Pharmacia, Uppsala, Sweden) was used for the deproteinization of the serum samples. The dimensions of the column were an inner diameter of 1.5 cm and length 4.8 cm. The fractionation range for peptides and globular proteins was 1000-5000 Da. The serum sample (400 1.11) was applied to the column and elution was effected with 8 ml of Tris-HCI buffer (pH 7.0). Determinations of glucose in the eluate fractions showed that 100% of the analyte appeared between 2.5 and 7.5 ml of the elution volume.This procedure guarantees an almost unde- tectable blank during the FI measurements. Determination of Glucose of H202 Each run starts by activation of the injection valve. The sample (glucose in Tris.HC1 buffer, pH 7.0) is injected into the carrier stream (Tris-HC1 buffer pH 7.0), which flows through 0.120 0 400 460 520 580 640 700 h/nm Fig. 2 Absorption spectrum of the final solution, 30 s after mixing of the reagents. T = 25"C, pH = 1.87. Fe2+ = 5.1 mmol 1-I; H202 = 0.07 mmol I-'. A, 3 mmol 1-I AAP; B, 21 mmol 1-' DCPS; and C, 3 mmol I-' AAP + 21 mmol I-' DCPS the GO reactor, where H202 is produced. The new reaction mixture meets at the output of the reagent solution R1 + R2, with which the pH is adjusted to the final value of 1.87.The new mixture proceeds through the PR and finally reaches the cell of the spectrophotometer, where the absorbance is measured at 510 nm. The reactions taking place during each measurement are: GO Glucose A G l u c o n i c acid + H202 (1) HzO + 0 2 H202 + AAP + DCPS + Fe2+ - Red product (2) The reaction curve and the maximum absorbance value A,,, are automatically presented on the screen. The calibration graph is established from the analysis of three to four glucose standards and then the unknown sample is analysed, following the same procedure. The concentration of glucose in the unknown sample is calculated by reference to the calibration graph. For the determination of H202 in samples containing this analyte, the enzyme reactor is omitted.In this instance the sample is injected into the carrier stream, and the concent- ration of H202 is calculated by use of a working curve. Results and Discussion Fig. 2 shows absorption spectra for the reaction products of FR with AAP, DCPS and their mixture. Values for kobs, A,,, and reaction rate R of fhe above reactions are presented in Table 1. The table also shows regression lines for the above parameters in terms of their dependence on the H202 concentration. From these results it can be concluded that a synergetic effect is operating for reaction (2). The optimum ratio of concentrations for the participating reagents is: AAP-DCPS-FeZ+ (1 + 4 + 1.7). Optimum pH values are pH 1.87 for the reaction of AAP + DCPS with FR, as shown in Fig.3, and pH 7.0 for the enzymic reaction. The optimum FI parameters are as follows. ( i ) The length of the PR should be 25 cm. (ii) The total flow rate should be 0.47 ml min-l; this ensures the arrival of the reacted zone in the observation cell 14 s after the addition of the reagents R1 and R2, a time which is sufficient for completion of the oxidation reaction. (iii) A loop of 30 1.11 ensures good values of A,,, with economy in the use of the reagents and in the time for a single run. Fig. 4 shows the calibration graphs for the determination of H202 with use of PO or Fe2+ in TR. Table 2 shows comparative results for the analytical characteristics of the twoANALYST, JUNE 1993, VOL. 118 729 Table 1 Experimental data from the study of the reaction of AAP, DCPS and their mixture with FR [AAP] = 2 mmol 1-l; [Fez+] = 5 mmol I-' [DCPS] = 2mmo11k1; [Fez+] = 5 mmoll-I [AAP] = 2 mmoll-l; [DCPS] = 2 mmol I-'; [Fez+] = 5 mmol I-' [ H 2 0 2 11 R = (dA/ R = (dA1 R = (dA1 10-6 dt)l10-3 kobsl dt)/10-3 kobsi dt)/10-3 kobs/ As-' 10-3 s- 1 A s s l 10-3 s- 1 Amax mol I-' A m a x A s-l 10-3 s-l A m a x 10 0.028 0.32 13.6 0.036 2.73 108.3 0.066 7.57 167.3 50 0.100 2.11 25.0 0.166 13.32 120.9 0.312 36.42 177.7 100 0.161 4.68 38.1 0.283 24.33 132.6 0.580 70.83 189.8 150 0.222 7.49 49.7 0.368 32.34 143.3 0.787 99.58 200.6 213.9 200 0.269 10.48 60.0 0.469 44.56 155.8 1.034 134.46 Regression equations- A,,, = (28 t 10) x 10-3 + A,,, = (38 & 18) x 10-3 + (1250 f.79) [H202] (54 -t 2) [H2021 kobs = (12 & 1) x 10k3 + (224 k 8) [H202] (2215 k 151) [H202] (213 -t 9) [H202] (244 k 9) [H202] R = ( - 4 f 2 ) X 10P4+ R = (17 k 11) X lop3 + k,& = (107 f.1) X + A,,, = (45 f. 23) x R = (25 k 16) x + (5011 k 191) [H202] (660 f. 13) [H202] (241 -I- 7) [H202] + kobs = (165 -t 1) x + 0.120 $ 0.080 Q I) 0.040 L 0 1 1 \ \ pH = 1.87 I I I I 2 3 4 5 6 PH Fig. 3 FR. T = 25°C. A, HCl + Tris-HC1 buffer; and B, KHP buffer pH profiles for the reaction of a mixture of AAP + DCPS with methods. The use of Fe2+ provides a larger linear range, while PO ensures better sensitivity. The standard additions method was applied to the determi- nation of glucose in the following way. To seven different portions of the serum were added known amounts of glucose before deproteinization.Each deproteinized serum sample was diluted about 14 times during the procedure to avoid double peaks and high blank values. The regression equations for the calibration graphs for standard glucose solution and serum are: A,,, = (4 ? 2) x [GLU] and A,,, = (174 k 2) x [GLU], respectively. The intercept of the regression equations corresponds to the absorbance of the original concentration of glucose in the sample tested. It is interesting to note that the standard additions method for the serum samples and the calibration graph obtained with standard solutions of glucose exhibit almost the same slopes, within experimental error. This allows the application of the method as described above. One hundred serum samples obtained from hospitalized patients were analysed for glucose by the proposed method with use of the manifold shown in Fig.1. The values obtained by the proposed method were compared with those of the official method applied in the clinical laboratory of the University Hospital [enzymic glucose determination with a Technicon (Tarrytown, NY, USA) RA-1000 System]. Very good agreement was achieved as the calculated slope of the correlation line was found to be 1.03. + (218 k 2) x + (244 k 2) x This work was financially supported partly from the STRIDE HELLAS 33 project of the University of Ioannina and partly from the Research Committee of the University of Ioannina, Project Number 170. o.800 I 1 ; 0.600 E Q ai 5 0.400 - e z a 0.200 - - II I I 1 0 0.080 0.160 0.240 0.320 0.400 HpOp/mmol I-' Fig. 4 Calibration graphs for H202 determination using the PO and Fez+ methods.T = 25 "C, pH = 1.87. A, Peroxidase: A,,,, = (-24 k 7) x + (4348 f. 78) x x H202], r = 0.9998; B , Fe2+: A,,, = (-12 2 4) x + (1692 f. 191 x x [H202], r = 0.9999 Table 2 Comparative analytical characteristics of the proposed Fe2+ and PO methods Analytical Proposed characteristics Fez+ method PO method Lincaritylymol I-' H202 10-400 5-1 70 SlopeIA 1 mmol- 1.69 f. 0.02 4.3.5 k 0.08 InterceptIA -0.012 f. 0.004 -0.023 k 0.006 Correlation coefficient 0.9993 0.9990 1 2 3 4 5 6 7 8 9 10 11 References Fenton, H. J. H . . J . Chem. Soc.. 1894,65, 899. Haber, F., and Weiss, J. J., Proc. R. Soc. London, Ser. A , 1934, 147, 332. Walling, C., Acc. Chem. Res., 1975, 8 , 125. Barham, D., and Trinder, P.. Analyst, 1972, 97, 142. Worsfold, P. J., Anal. Chim. Acta, 1983, 145, 117. Mottola, H. A., Anal. Chim. Acta, 1983, 145, 27. CRC Handbook of Chemistry and Physics, cd. Weast, R. C., CRC Press, Boca Katon, FL, 62nd edn., 1981-1982. Kiranas, R. E., Tzouwara-Karayanni, M. S . , and Karayannis, Stults, L. M. C., Ph.D. Dissertation, Michigan State University, 1987. Patton, J. C., and Crouch, R. C., Anal. Chim. Acta, 1986, 179, 189. Henry, J. R., Cannon, C. D., and Winkelman, W. J . , Clinical Chemistry. Principles and Techniques, Harper and Row, Hagerstown, MD, 2nd edn., 1974, pp. 389404. Paper 21061 37F Received November 18, 1992 Accepted December 23, 1992 I . M . , Acta Chim. Hung., 1992, 129, 461.

ISSN:0003-2654    

DOI:10.1039/AN9931800727

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993. VOL. 118 COM M U N ICATIO N 73 1 Material for publication as a Communication must be on an urgent matter and be of obvious scientific importance. Rapidity of publication is enhanced if diagrams are omitted, but tables and formulae can be included. Communications receive priority and are usually published within 5-8 weeks of receipt. They are intended for brief descriptions of work that has progressed to a stage at which it is likely to be valuable to workers faced with similar problems. A fuller paper may be offered subsequently, if justified by later work. Manuscripts are usually examined by one referee and inclusion of a Communication is at the Editor's discretion. Use of Poly(~-lysine) and Ascorbic Acid for Surface Enhanced Resonance Raman Scattering Analysis of Acidic Monoazo Dyes C.H. Munro, W. E. Smith and P. C. White* Department of Pure and Applied Chemistry, University of Strathclyde, Glasgow, UK GI IXL A surface enhanced resonance Raman scattering (SERRS) procedure is described for the analysis of acidic monoazo dyes. The SERRS spectra of these compounds can be obtained if a 0.01 % solution of poly(L-lysine) is added to a citrate-reduced silver sol. Further enhancement of the SERRS intensities can be obtained by the addition of ascorbic acid. The discrimination between dyes offered by this procedure is demonstrated and a mechanism is proposed for the interaction between the poly(L-lysine), ascorbic acid and a dye. Keywords: Surface enhanced resonance Raman scattering; poly(L-lysine); ascorbic acid; acidic monoazo dye Acidic dyes are used to colour a large number of materials including food, drink, water-based paints, cosmetics, inks, leather and a range of fibres (wool, nylon, silk and modified acrylic).It is often necessary within forensic science to examine these materials, and the ability to analyse and characterize the small amounts of dyes (typically 10 ng or less) used to colour them would be highly advantageous. A chromatographic technique combined with multivariate analysis of the data has been successfully employed for the study of these dyes, however, some difficulties are still experienced in their characterization. 1 Surface enhanced Raman scattering (SERS) has emerged as a potentially useful analytical technique for the identification of trace amounts of Raman-active compounds adsorbed onto silver surfaces, with enhancement of the scattering intensity being typically in the range lO3-lO6.2,3 Systems examined under electronic resonant conditions can show a multiplica- tion of the resonant and surface enhancements, the resulting surface enhanced resonance Raman scattering (SERRS) is extremely efficient, allowing detection levels in the region of 10-y-lO-ll mol I-' to be achieved.4-6 In addition, fluores- cence from adsorbed species is generally quenched by radiationless energy transfer to the metal surface, enabling Raman spectra to be obtained for compounds where this would normally be hindered by the presence of a strong fluorescent background.7 Surface enhanced Raman scattering has been observed for many compounds (including dyes) on electrode surfaces, colloidal solutions, gratings, paper substrates, silver island films and silver films deposited on quartz, Teflon particles or posts.637 Colloidal silver has been the most widely employed form of enhancing agent, giving strong SERS in aqueous solution without the requirement of specialized apparatus.The enhancement is greatest when controlled aggregation is induced in the sol. This is believed to produce greater field * To whom correspondence should be addressed. strength in the interstices of the aggregate. Citrate-reduced silver colloids show a more uniform particle size distribution and greater stability than the tetrahydroborate-reduced equi- valent indicating that SERRS on citrate-reduced silver sol particles might be applicable to the analysis and characteriza- tion of acidic azo dyes.6 In order to investigate this hypothesis SERRS analyses were performed on several acidic monoazo dyes and the results of this preliminary study are presented and discussed.Experimental All chemicals were of analytical-reagent grade and were obtained from several sources; sodium citrate, silver nitrate (Johnson Matthey), ascorbic acid (Aldrich) and poly( L-lysine) hydrobromide, relative molecular mass 23000 (Sigma). The following dyes were used in this study: Food Red 17 (CI 16035), Acid Red 13 (CI 16045), Acid Orange 10 (CI 16230), Acid Red 44 (CI 16250), Acid Orange 12 (CI 15970) and Acid Red 88 (CI 15620). Prior to their analysis pure dye samples were obtained by preparative thin-layer chromatography.A Spectra-Physics 2020 argon ion laser (100 mW) was used as the excitation source (457.9 nm), with conventional 90" geometry. The spectrum was dispersed by an Anaspec-modi- fied Cary 81 scanning monochromator. The spectral resolu- tion was 4 cm-1. A cooled Thorn EM1 9658R photomultiplier tube was used for detection, with photon-counting electronics for data acquisition. Silver sols were prepared according to the method of Lee and Meisel.8 [It is essential that all glassware is rigorously cleaned before use by treatment with aqua regia (HCL- HN03, 3 + 1) followed by gentle scrubbing in a soap solution.] A sample of silver nitrate (90 mg) was suspended in distilled water (500 ml) and heated rapidly to 100 "C. A 1% solution of sodium citrate (10 ml) was added under vigorous stirring and the solution was kept boiling for 65 min with continuous stirring.In order to promote adsorption of the dye and aggregation732 ANALYST, JUNE 1993, VOL. 118 of the silver colloid a 0.01% aqueous solution of poly(L-lysine) (150 pl) was added to an aliquot of the silver sol (2 ml), followed by an aqueous solution of the dye mol 1-l: 100 PI) and ascorbic acid (1 mol 1-I: 100 PI). Results and Discussion The ultraviolet visible absorption spectrum of the silver sols showed maxima, at approximately 404 nm, which can be I I 1 I I I I I 1600 1200 800 Wavenu m be r/cm - 400 Fig. 1 SERRS spectra of poly(L-lysine)-dye complexes in the presence of ascorbic acid. Dyes: ( a ) Acid Orange 12; ( b ) Food Red 17; (c) Acid Red 44; and ( d ) Acid Red 88.(Intensities are measured in counts s-1) associated with the dipolar surface plasmon for silver spheres that have small radii (approximately S20 nm) compared with the illumination wavelength. Addition of an aqueous solution of each dye, used in this study, to the silver sol produced no aggregation or SERRS spectra and this was attributed to poor adsorption of the anionic dye molecules to the sol particles due to the negatively charged citrate layer on the surface of the silver.9 However, on addition of 150 p1 of 0.01% poly(~- lysine) solution to the silver sol immediately prior to the addition of the dyes, aggregation and intense SERRS spectra of the poly(L-1ysine)-dye complexes were observed. Subse- quent addition of 150 pl of 1 mol I-' ascorbic acid to the colloidal suspension resulted in an overall increase in the observed SERRS intensity.Examples of the spectra recorded are illustrated in Fig. 1. At present, no assignment of the bands observed has been carried out, nor has the sensitivity of the technique been fully assessed, however, the intensity of scattering and the clear differences in the observed spectra indicate that SERRS in the presence of poly(L-lysine) and ascorbic acid provides a method for the discrimination at low concentrations of the acidic monoazo dyes studied. The bonding of the dyes to the poly(L-lysine) is analogous to the proposed action of acidic azo dyes in the colouring of wool fibres. The lysine residues of the wool a-keratin are the primary sites of attachment for the anionic dye molecules.Under the correct conditions for dyeing, the amino groups of the residues will become protonated, attracting the negatively charged dye molecules. In a similar fashion, the ionic interactions between the poly(L-lysine) and the dye molecules lead to the formation of poly(L-1ysine)-dye complexes. The presence of the ascorbic acid promotes the protonation of the amino groups, increasing the overall efficiency of the process. The protonated amino groups of the unbound lysine residues are attracted to, and readily absorbed onto, the citrate layer of the sol particles. This leads to aggregation of the particles and an intense SERRS effect. Poly(L-lysine) is commonly used for the determination of secondary structure content of proteins by Raman spectro- sc~py.~".ll As such, the Raman spectra of the a-helical, p- sheet and radom coil conformations are well characterized.Under the conditions used in this current study the literature would support a preference for the random coil conforma- tion.12 For random coiled poly(L-lysine) in solution, the amide I and 111 frequencies are reported to be at 1665 cm-l and 1243-1248 cm-l , respectively. However, owing to the absence of these bands, examination of the SERRS spectra of the poly(L-1ysine)-dye complexes indicates that poly( L-1 ysine) makes no contribution to the observed spectra. Comparison of the SERRS spectra with the resonance Raman spectra of the individual dyes confirmed that the SERRS spectra observed can be attributed in full to the dye molecule. Conclusion The results of this preliminary study indicate that in the presence of poly( L-lysine) and ascorbic acid, SERRS analysis can be used for the fingerprint identification at low concentra- tions of the six acidic monoazo dyes examined.The results also support a continuation of the examination of SERRS as a technique for the characterization of acidic dyes. References 1 White, P. C. , and Catterick, T., Analyst, 1993,118, in the press. 2 Fleischmann, M., Hendra, P. J., and McQuillan, A. J., Chem. Phys. Lett., 1974, 26, 163. 3 Creighton, J . A., Blatchford, C. G., and Albrecht, M. G., J . Chern. SOC., Faraday Trans. 2 , 1979, 75, 790. 4 Hildebrandt, P., and Stockburger, M., J . Phys. Chem., 1984, 88, 5935.ANALYST, JUNE 1993, VOL. 118 733 5 Xu, Y., and Zheng, Y., Anal. Chim. Acta, 1989, 225, 227. 6 Sheng, R.-S., Zhu, L., and Morris, M. D., Anal. Chem., 1986, 58, 1116. 7 Montes, R., and Laserna, J. J., Analyst, 1990, 115, 1601. 8 Lee, P. C., and Meisel, D., J . Phys. Chem., 1982, 86, 3391. 9 Jolivet, J . P., Gzara, M., Mazieres, J . , and Lefebvre, J., J. Coil. Interface Sci., 1985, 107, 429. Paper 3/02155 F 10 Chen, M. C., and Lord, R. C., J. Am. Chem. Soc., 1974, 96, Received April 15, 1993 4750. Accepted May 5, 1993 11 12 Lippert, J. L., Tyminski, D., and Desmeules, P. J . , J. Am. Chem. Soc., 1976, 98, 7075. Yu, T.-J., Lippert, J. L., and Peticolas, W. L., Biopolymers, 1973, 12, 2161.

ISSN:0003-2654    

DOI:10.1039/AN9931800731

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

ANALYST, JUNE 1993, VOL. 118 735 Aboal-Somoza, Manuel, 665 Adams, Michael J., 229 Alder, J. F., 395 Allag, Houssein, 401 Alwarthan, Abdulrahman A. , Anderson, David R., 449 Andrew, B. E., 153 Andrews, William J., 425 Arrigan, Damien W. M., 355 Ashok Kumar, T., 293 Avidad, Ramiro, 303 Bae, Yea-Ling, 297, 301 Bangar Raju, G., 101 Bannon, Thomas, 361 Barclay, David A., 245 Barjat, Herve, 73 Barker, Philip G., 347 Barnard Howie, Judith A., 35 Bartlett, Philip N., 371 Banvick, Ian M., 489 Baxter, Douglas C., 495 Bayo, Javier, 171 Bell, Jimmy D., 241 Belton, Peter S., 73 Benmakroha, Farida, 401 Bermejo-Barrera, Pilar, 665 Bhaskar, Nilam, 1 Biondi, Cinzia, 183 Birmingham, John. J . , 1 Blaih, Salah M., 577 Blair, Neil, 371 Bos, Albert, 323 Bos, Martinus, 323 Boudjerda, Tarik, 401 Boufenar, Rabah, 401 Bradbury, Michael W.B., 533 Breen, William, 415 Brienza, Sandra Maria Boscolo, Brinkman. Udo A. Th., 11 Brown, Marc B . , 407 Bruns, Roy E., 213 Bui, Liin N., 463 Cai, Xiaohua, 53 Calokerinos, Antony C., 627, Campiglio, Antonio, 545 Canela, Ramon, 171 Capitan-Vallve y , Luis Fermin, Carlson, Robert G., 257 Casey, Vincent, 389 Cassidy, John F., 415 Cermak, Josef, 79 Chadima, Radko, 79 Chaisuksant, Rasamee, 179 Chartier. A, 157 Chen, Liang. 277 Chen, Tianyu, 541 Chokshi, Hitesh P., 257 Clark, Alastair, 601 Clarke, Colin G., 229 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 Crubellaji, Ricardo O., 529 Csizkr, Eva, 609 Cummins, Diane, 1 Cummins, Phillip G., 1 Cunningham, K., 341 639 719 633 303 CUMULATIVE AUTHOR INDEX JANUARY-JUNE 1993 Daenens, P., 137 Davis, Willard E., 249 de Andrade, Joiio Carlos, 213 de la Guardia, Miguel, 23 de la Rosa, Francisco F., 643 de Paula Eiras, Sebastiiio, 213 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-GonzBlez, Raquel, dos Reis, Boaventura Freire, 719 Dowle, Chris J., 17 Dreassi, Elena, 183 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 Faure, Uta, 475, 481 Fearn, Tom, 235 FernBndez Laespada, Ma. Esther, 209 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 Gaind, Virindar S . , 149 Garcia G6mez de Barreda, Daniel, 175 Garcia-Lopez, Trinidad, 303 Gardner, Julian W., 371 Georges, J., 157 Ghijsen, Rudy T., 11 Gibney, Patrick M., 425 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, Zhi-cheng, 105 Guiraum, Alfonso, 643 Haegel , Franz-Hubert ,703 Halbig, Peter, 689 Halvatzis, Stergios A., 633 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 Hidalgo de Cisneros, Jose L. 665 175 Hidalgo, 175 Hiraide, Masataka, 537 Hirose, Tsuyoshi, 517 Hokari, Norihisa, 219 Hollman, Peter C. H., 475, 481 Howard, Vyvyan C., 1 Huang, Ka-Lin, 205 Hunt, Terence P., 17 Idriss, Kamal A., 223 Iizuka, Ryuji, 165 Ishida, Junichi, 165 Iwachido, Tadashi, 273 Iwata, Tetsuharu, 517 Izquierdo, Pilar, 707 Izumi, Sigeru, 553 Janata, Jiii, 335 Jedrzejczak, Kazik, 149 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 , 13 1 Katz, Stanley E., 281 Kawaguchi, Hiroshi, 537 Kessler, Margalith, 235 Keyes, Emmetine T., 385 King, Bernard, 587 Kinoshita, Toshio, 161 Kiranas, Efstratios R., 727 Kiss, Attila, 661 Kobayashi, Atsushi, 273 Kobayashi, Shouichi, 131 Kogak, Ali, 657 Koh, Tomozo, 669 Konidari, Constantina N., 71 1 Konig, Monika, 703 Kotrly, Stanislav, 79 Kovanic, Pavcl, 145 Krishan Puri, Bal, 85 Kubal, Gina, 241 Kumar, Manjeet, 193 Lan, Chi-Ren, 189 Lang, Mark J., 425 Lauko, Anna, 609 Ledesma, Ariel G., 529 Ledingham, Kenneth W.D., 601 Lev, Ovadia, 557 Li, Ronghua, 563 Li, Xiang-Ming, 289 Liang, Wei-An, 97 Lin, Qingxiong, 643 Lin, Yuehe, 277 Littlejohn, David, 541 Lopez Ruiz, B . , 59 Lowe, Roger D., 613 Lunar, Maria Loreto, 715 Lunte, Susan M.. 257 Luque de Castro, Maria 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, 51 1 MacCraith, Brian D., 385 McDonagh, Colettc M., 385 McEvoy, John D. G., 447 McGilp, John F. , 385 McKeown, Neil B., 463 McKervey, M. A., 341 MacLaurin, Paul, 617 McLeod, Cameron W., 449 Magee, Robert J., 53 Malone, Michael A., 649 Marshall, Archibald. 601 723, 727 Dolores, 593 Martelli, Patricia Benedini, 719 Martin, J. P., 59 Martinez-Lozano, C., 567 Mathieu, Jacques, 11 Matsubara, Chiyo, 553 Mellidis.Antonios S., 179 Mertens, Bart, 235 Midgley, Derek, 41 Miller, James N., 407, 455 Miller, Richard M., 1 Miura, Yasuyuki, 669 Monks, Cheryl D., 623 Moreno, Miguel A.. 171 Mori, Yuichi, 553 Moriyama, Youichi, 29 Moss, Martin C., 1 Mottola, Horacio A., 675 Mufioz Leyva, Juan A., 175 Munro, C. H., 731 Nabekura, Tomiko. 273 Nacapricha, Duangjai, 623 Nagahiro, Tohru, 85 Nakagawa, Genkichi, 219 Nakamura. Kayoko, 29 Nakamura, Masaru, 517 Nanos, Christos G., 711 Narayanaswamy, Ramaier, 317 Navalon, Albcrto, 303 Neuhold, Christian, 53 Nicholson, Brenton C., 265 Nickel, Ulrich, 689 Nimura. Noriyuki, 161 Norman, Philip, 617 O’Beirn, Brcndan, 389 O’Donoghue, Eilish, 415 Ohkubo, Hiromi, 549 Okabe, Katsuaki, 669 O’Kane, Edward, 511 O’Keeffe, Gerard, 385 O’Kelly. Brendan, 38.5 O’Kennedy, Richard, 201, 411 O’Neill, Robert D., 433 O’Sullivan, Ciara, 411 Palaniappan, K., 293 Papageorgiou, Vassilios P., 179 Paukert, TomaS, 145 Paynter, J., 379 Pearce, Timothy C., 371 Perez Pavon, Jose Luis, 209 Perez-Bendito, Dolores, 707, Perez-Ruiz, T., 567 Peris Cardells, Empar, 23 Persaud, Krishna C., 419 Petelenz, Danuta, 335 Pitre, Krishna S . , 65 Pramauro, Edmondo, 23 Preston, Gaynor, 245 Prevot, Alessandra Bianco, 23 Prieta, Javier, 171 Proietti, Daniela, 183 Pyo, Dongjin, 253 Quencer, Brett M., 695 Radulovic, Stojan, 241 Radunovik, Aleksandar, 533 Ramachandran, Venkataraman Raurich, Josep Garcia, 197 Reckhow, David A., 71 Reid, Helen J., 443 Repasi, Janos, 661 Riley, David P., 407 Roe, Merrion P., 425 Romaschin.Alex D., 463 Ruan, Fu-Chang, 289 RubeSka, Ivan, 145 Rubio Leal, Amparo, 89 Rubio, Soledad, 715 Sadler, Peter J., 241 Saleh, Magda M. S., 223 Salinas, Francisco, 89 715 N., 511736 ANALYST, JUNE 1993, VOL. 118 Salvatore, Michael J., 281 Sanchis, Vicente, 171 Sander, Joseph, 601 Sanz, A., 567 Sanz Pedrero, P., 59 Satake, Masatada, 85 Savarino, Piero, 23 Sawai, Kaori, 549 Schneider, Siegfried, 689 Schwuger, Milan Johann, 703 Scare, Nichola J., 407 Shallow, A., 429 Sheppard, Robert C., 1 Shimoishi, Yasuaki, 273 Shortt, Desmond H., 447 Silva, Manucl, 681 Simpson, Michael, 449 Singhal, Ravi P., 601 Singleton, Scott, 1 Slangen, Jean H., 475, 481 Slater. Jonathan M., 379 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 Sramkova, Jitka, 79 Srivastava, P. K., 193 Stalikas, Constantine D., 723 Su, Hongbo, 309 Suarez, Guillermo, 171 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 Taylor, Colin G., 623 Teasdale, P. R. , 329 Thompson, Michael, 235,309, Timotheou-Potamia, Meropi Tomas, V., 567 Torrades, Francesc, 197 Tbyo’oka, Toshimasa, 257 Tsai, Suh-Jen Jane, 297,301,521 Tsionsky , Michael, 557 463 M., 633 Tsuzuki, Wakako, 131 Tucker, Alan, 241 Tuiion Blanco, Paulino, 649 Tzouwara-Karayanni, Stella M., 723, 727 Uchida, Takaaki, 537 Valcarccl, Miguel, 593 Van Boven, M., 137 van der Linden, Willem E., 323 Veiro, Jeffrey A., 1 Verma, Neerja, 65 Vijaya Raju, K., 101 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, P. C., 731 179 Williams, David M., 249 Williams, Kathleen E., 245 Wong, Kwok-Yin, 289 Worsfold, Paul J., 617 Worswick, Richard, 583 Wuchner, Klaus, 11 Xie, Yuefeng, 71 Xu, Hongda, 269 Yamaguchi, Masatoshi, 165, 517 Yamauchi, Shuji, 161 Y an, Hsiao-Tzu, 52 1 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-Iiang, 105 Zou, Shi-Fu, 97 719

ISSN:0003-2654    

DOI:10.1039/AN9931800735

出版商:RSC    

年代:1993

数据来源: RSC