摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1135 Automated Determination of Sulphite by Gas-phase Molecular Absorption Spectrometry Toyin A. Arowolo and Malcolm S. Cresser" Department of Plant and Soil Science, University of Aberdeen, Meston Building, Aberdeen A69 ZUE, UK Conditions for the automated determination of sulphite in aqueous solution based on measuring the absorbance of sulphur dioxide evolved upon acidification of samples are described. The proposed method has a detection limit of 0.20 pg ml-1 and relative standard deviations of 2.3 and 1.8% for 20 and 10 pg ml-1 of sulphite, respectively. The calibration graph is linear up to 120 pg ml-1 and samples can be analysed at the rate of 20 per hour. The method has been applied to the determination of sulphur dioxide in synthetic samples and in white wines.Keywords: Gas-phase molecular absorption spectrometry; sulphite determination; automated analysis Sulphite and sulphur dioxide are widely used as preservatives in the food and pharmaceutical industries because of their ability to prevent oxidation and bacterial growth. Strict control over their concentrations in industrial products is mandatory because of their toxicity. Sulphur dioxide is also a major air pollutant that has been causing environmental concern over recent decades. It has been identified as one of the main causes of increasing acidification of the environment through the generation of acid rain. Sulphur dioxide is released into the atmosphere primarily from the combustion of coal and petroleum, the smelting of sulphur-containing ores, production of sulphuric acid, and the paper manufactur- ing industry. Its adverse effects on living organisms are well known.1 With the growing concern over sulphite as a food additive and sulphur dioxide as an important gaseous pollutant, it is not surprising that reliable methods for the determination of these compounds are continuously being sought. Both sulphite and sulphur dioxide have been determined by a variety of analytical techniques.2 These include titrimetry,3 ion chromat- ography,4 coulometry,5 conductimetry,6 flame photometry,7 molecular-emission cavity analy~is~8.9 chemiluminescencelo-11 and ultraviolethisible (UVNIS) spectrophotometry.12-16 The reference method adopted by the United States Environmen- tal Protection Agency'7 for the determination of atmospheric concentrations of sulphur dioxide is the modification by Scaringelli et a1.16 of the West-Gaeke method.15 The method, which is simple, sensitive and specific, has been automated.18 Cresser,1g720 Syty,21 and Winkler and Syty22 utilized the characteristic UV absorption bands of sulphur dioxide for its determination in solution by gas-phase molecular absorption spectrometry (GPMAS).Their methods were manual and involved considerable manipulation. The GPMAS technique is a sensitive, reproducible and versatile non-flame molecular absorption method in which the ionic species of interest is converted into a volatile product. The absorbance of the gas is measured as it is swept into a flow-through absorption cell positioned in the space normally provided for the burner in a flame atomic absorption spectrometer.We have recently described23 an automated method for the determination of sulphide by GPMAS. The purpose of this paper is to report the application of a similar procedure to the determination of sulphite in solution. Experimental Apparatus A Baird (Alpha 1) atomic absorption spectrometer was used with a deuterium hollow cathode lamp and a Bryans Southern * To whom correspondence should be addressed. Instruments 28000 chart recorder. The absorption cell was fastened tightly to the burner with a rubber band. This allowed the use of the burner controls to align the cell in the optical path. The procedure for introducing the carrier gas into the system and the gas-liquid separator were as described previously.23 Reagents All reagents were of analytical-reagent grade and de-aerated, de-ionized water was used throughout.Sulphite stock solution, 500 pg ml- *. This solution was prepared by dissolving 0.197 g of anhydrous sodium sulphite, Na2S03 [Merck (formerly BDH)], in 250ml of antioxidant solution. Working standards were freshly prepared each day in antioxidant solution by the least number of dilution steps possible. Stock sodium tetrachloromercurate(I1) ( TCM) solution , 0.04 rnol dm-3. Prepared by dissolving 10.86 g of mercury(I1) chloride, 4.68 g of sodium chloride, and 0.066 g of disodium dihydrogen ethylenediaminetetraacetate (Na2H2EDTA) in water, and diluting to the mark in a 11 calibrated flask. Stock sulphamic acid solution, 0.6%. Prepared by dissolving 0.6 g of sulphamic acid in 100 ml of water.Znterferent solutions, 1000 pg ml-1. Solutions of a range of cations and anions were prepared from analytical-reagent grade salts. Stability of Sulphite Solutions Aqueous sulphite solutions are unstablelg-22 in the presence of oxygen because they are very susceptible to oxidation. In the literature, various compounds have been suggested as stabil- izers for sulphite in aqueous solutions. Preliminary investiga- tion was carried out to find a stabilizer which is compatible with the proposed method by selecting three of the com- pounds or mixtures quoted in the literature, glycerol, 0.1 rnol dm-3 NaOH-O.OO1 rnol dm-3 EDTA and TCM, for initial use. Analysis of standard solutions of sulphite stabilized by each of the three compounds gave similar results. Stock solution of sodium sulphite was therefore prepared in 0.1 rnol dm-3 NaOH4.001 rnol dm-3 EDTA and also in 0.04 rnol dm-3 TCM.Procedure The operating parameters for the various parts of the instrument and manifold are shown in Table 1. The experimental procedure is the same as previously reported for sulphide .231136 ANALYST, NOVEMBER 1991, VOL. 116 Table 1 Operating conditions Spectrometer- Wavelength setting Mode Lamp current Slit-width Bandpass Recorder- Sensitivity Chart speed 198 nm Absorbance 10 mA 1.0 mm 3.0 nm 100 mV-2.5 V full scale 5 mm min-1 Proportioning pump and automatic sampler- Sampling cycle 45 s Wash cycle 45 s Air 3 mol dm-3 HCI flow rate Sample or wash uptake rate Carrier gas (air) flow rate Mixing coil 0.6 ml min-I 0.32 ml min-l 3.90 ml min-l 15.0 ml min-I 29 turns Results and Discussion Optimization of Experimental Conditions Effect of carrier gas flow rate Variation of the carrier gas flow rate which carried the evolved sulphur dioxide to the absorption cell was investigated by analysing a 20 pg ml-1 sulphite solution at different flow rates.The results obtained indicate a 15% difference in absorption signal intensity (absorbance changed from 0.035 to 0.041) when the carrier gas flow rate was varied from 10 to 30 ml min-1. Therefore, a convenient flow rate of 15 ml min-1 was selected for further work. Choice of wavelength Examination of the sulphur dioxide absorption spectrum obtained by Cresser and Isaacson24 in a similar procedure involving the addition of aliquots of 6 mol dm-3 hydrochloric acid to sodium suphite revealed wide bands with a maximum between 190 and 200 nm, and considerable fine structure (see Fig.1). Absorption intensity for various concentrations of sulphite solution (0-20 pg ml-1) at different wavelengths in the range 190-215 nm was measured and the results are shown in Fig. 2. A wavelength of 198 nm was employed throughout this study. Effect of temperature The effect of temperature on the rate of evolution of sulphur dioxide was investigated by passing the manifold tube carrying the reaction mixture (sample-HCI) through a heating-bath coil (Hampshire Lab Aid) which was connected to a tempera- ture controller. Analysis of standard sulphite solutions by the proposed method at three different temperatures was then carried out.The results obtained, shown in Fig. 3, indicate that an increase in temperature enhances the signal obtained. This is possibly due to a reduction in the solubility of sulphur dioxide with increasing temperature, and a reduction in the viscosity of the reaction mixture. However, condensation of water vapour at the top of the gashquid separator and in the connection tubing leading to the absorption cell was noticed when analysis was performed at elevated temperatures. This eventually resulted in poor precision, which was probably due to the re-absorption of evolved sulphur dioxide by the water droplets and the obstruction of the smooth flow of the gas into the absorption cell. In order to solve this problem, the gas-liquid separator and the absorption cell temperatures were raised above room temperature by wrapping them with a heating tape (HT3, Fisons) connected to a temperature controller.Although this eliminated the condensation prob- 200 250 300 Wavelength/nm Fig. 1 Absorption spectrum of sulphur dioxide 0.035 0.03 0.025 a, 2 0.02 e s: 9 0.015 0.01 0.005 0 5 10 15 20 25 Sulphite concentration/pg ml-' Fig. 2 Effect of choice of wavelength on sulphite calibration graphs: A, 190; B, 198; C. 205: D, 210; and E, 215 nm 0.014 I 0.01 2 0.01 $ 0.008 a, e s: 9 0.006 0.004 0.002 0 1 2 3 4 5 6 Sulphite concentration/pg ml-' Fig. 3 Effect of temperature of the mixing coil on sulphite calibration graphs: A. 22; B, 42; and C, 50 "C lem, it also resulted in increased background noise and baseline drift.Therefore, analysis at high temperatures will only be worthwhile if samples containing low concentrations of sulphite are to be determined. Except where otherwise stated, the data in this study were collected at 25 "C.ANALYST, NOVEMBER 1991, VOL. 116 Calibration graph, reproducibility and detection limit Fig. 4 shows a typical recording for a series of sulphite standards analysed by the proposed method. The linear portion of the calibration graph prepared by plotting absor- bance as a function of sulphite concentration under the optimum conditions extends to 120 pg ml-1. The equation of the calibration graph obtained by the method of least squares was A = 0.00168~ + 0.007, with r = 0.9980 (or A = 0.00176~ if the assumption Y = bx is used).The relative standard deviation (RSD) values for ten determinations of 20 and 10 pg ml-1 of sulphite were 2.3 and 1.8%, respectively. The limit of detection (signal-to-noise ratio = 2) of the proposed method was 0.2 pg ml-1 of sulphite. The slope of the straight line portion of the calibration graph is 0.00168 ml pg-1 which represents the sensitivity of the method. Under identical conditions, determination of sulphide (0-100 pg ml-1) by the same technique was more sensitive23 (A = 0.00547~ + 0.0081) than that of sulphite. Similar observations have been reported for molecular-emission cavity analysis.25 Interferences The effects of several anions were studied for potential interference upon the evolution of sulphur dioxide and its 0.27 8 0.18 5 B 9 0.09 K 0 100 , 10 min , Time - Fig.4 Typical recorder tracing for a series of sulphite standards. Numbers above peaks denote concentrations in pg ml-l 1137 absorbance in the gas phase at 198nm using the proposed method. A series of solutions containing 20pgml-1 of sulphite plus 500 pg ml-1 of potential interfering anion (as Na or K salts) and solutions containing the potential interfering anion (500 pg ml-1) in the absence of sulphite, were analysed by the procedure described above. The results obtained were compared with those from an uncontaminated sulphite solution (20 pg ml-1). The following anions had no effect on the determination of sulphite by the proposed method: C1-, Br-, I-, Po43-, S042-, C032- and NO3-. As expected, sulphide interfered by giving an enhanced absorption signal due to the evolution of hydrogen sulphide which absorbs at 198nm.An absorbance value of 0.045 was obtained when 5 pg ml-1 of sulphide were analysed alone. The potential effect of nitrogen dioxide, a common interfer- ent in the determination of atmospheric sulphur dioxide, was also studied. Nitrite at a concentration of 10 pg ml-1 inter- fered with the determination by causing a substantial depres- sion of the sulphite signal. Addition of 5 ml of 0.6% sulphamic acid to the solution containing 20 pg ml-l of sulphite and 10 pg ml-1 of nitrite prior to analysis restored the peak. It was observed that the volume of the 0.6% sulphamic acid solution required to eliminate the interference depended on the amount of nitrite present in the sample. For example, the addition of 5 ml of 0.6% sulphamic acid to a sample containing 20 pg ml-1 of both sulphite and nitrite resulted in about 70% elimination of the interference.It was also observed that the addition of a similar volume of 0.6% sulphamic acid solution to an uncontaminated sulphite solution (20 pg ml-1) did not affect the absorption signal of sulphur dioxide. This confirmed that sulphamic acid can be used to eliminate the effect of nitrite in the proposed method. The interference effects of various cations were investigated by a procedure similar to that used for anions. Solutions of sulphite and cation were made up in de-aerated, de-ionized water, 0.001, 0.01 or 0.1 mol dm-3 Na2H2EDTA, depending on the degree of interference. In some instances, a precipitate was formed after mixing of solutions in de-aerated, de-ionized water but preparation of similar solutions in EDTA resulted in the dissolution of this precipitate.The ions Na+, K+, Mg2+, ~ ~ ~- Table 2 Effect of metal ions in aqueous and EDTA solutions on 20 pg ml-1 of sulphite Observed signal intensity* relative to that of 20 pg ml-1 of sulphite (%) Concentration/ Ion pg ml-1 Ion + sulphite None - Ni2+ 500 -t Pb2+ 500 --t Ba2+ 500 45 Cr3+ 500 70 50 92 10 100 Fe2 + 500 0 50 0 10 92 Fe3+ 500 0 50 0 10 89 co2+ 500 0 50 0 10 0 Mn2+ 500 --t 50 0 10 8 cu*+ 500 15 50 0 10 0 100 * Mean of two values. t Precipitate formed after mixing of solutions, -$ Not determined. Ion + sulphite + 0.001 mol dm-3 EDTA 100 0 0 55 70 90 100 0 54 100 0 59 100 --$ 100 100 0 0 95 15 100 100 Ion + sulphite + 0.01 mol dm-3 EDTA 100 34 89 64 73 92 -$ 0 54 92 0 89 100 27 100 100 0 0 0 30 100 100 Ion + sulphite + 0.1 mol dm-3 EDTA 100 97 --$ 100 79 --$ --$ 0 38 73 0 86 78 39 -$ --$ 0 0 0 24 --$ --$1138 ANALYST, NOVEMBER 1991, VOL.116 Ca2+, and Cd*+ at concentrations of 500pgml-1 did not interfere with the determination of 20 pg ml-1 of sulphite in water or EDTA solution. A similar concentration of Zn2+ or A13+ caused a slight depressive interference effect on sulphite determination in water, but the presence of 0.001 rnol dm-3 EDTA restored the peak. Table 2 shows that EDTA at concentrations in the range 0.001-0.1 rnol dm-3 is effective in removing the interference effect of 50 pg ml-1 of Cr3+, Fe3+, Co2+ and Cu2+. However, Mn2+ and Fez+ were the most severe interferents of all the cations investigated and only 10pgml-1 of these two cations could be tolerated in the determination of 20 pg ml-1 of sulphite.These cationic interferences are due to the formation of stable complexes or insoluble compounds with sulphite. However, concentration of the EDTA, the pH of the solution and the interfering metal ion are important factors in the ability of the masking agent (EDTA) to suppress the effect of an interferent completely. For example, 0.1 rnol dm-3 EDTA gave comparatively lower results for some of the metal ions. This might be due to the premature evolution of sulphur dioxide during the preparation of such solutions. Introduction of EDTA into the manifold resulted in its precipitation in the mixing coil owing to the addition of acid.Applications Determination of sulphur dioxide The most widely applicable method for the collection and determination of sulphur dioxide in air is that of West and Gaeke and modifications thereof;15-1* In this method, sulphur dioxide in an air sample is absorbed into a solution of sodium or potassium TCM. The resulting complex, monochlorosul- phonatomercurate(ir), is treated with formaldehyde and specially purified, acid-bleached pararosaniline, to form a red-violet complex, pararosanilinemethylsulphonic acid. The absorbance at 548 or 575 nm is measured in order to determine the concentration of sulphur dioxide. Interference from nitrogen dioxide is overcome by the addition of sulphamic acid. The application of the proposed method to the determina- tion of sulphur dioxide in synthetic samples stabilized by TCM was carried‘out in an effort to establish whether it would be applicable to air pollution studies.Sulphite solutions (10 pg ml-1) were prepared in various concentrations of TCM (0.01-0.1 rnol dm-3) and the absorbance was measured by using the proposed method. No significant difference was obtained in the sulphur dioxide absorption intensity of these solutions and those prepared in both de-aerated, de-ionized water and in 0.1 rnol dm-3 NaOH-O.OO1 rnol dm-3 EDTA. It was therefore decided to prepare samples used for this study in 0.04 rnol dm-3 TCM. The proposed method was compared, for synthetic samples, with the West-Gaeke method by measuring the sulphur dioxide content of pure solutions covering a wide range of concentrations.The latter were prepared by absorbing sulphur dioxide generated from stock sulphite solution (after acidification) in 0.04 rnol dm-3 TCM. The results obtained with the two methods are in good agreement, as shown in Table 3. Recovery of sulphite from wines The proposed method was also applied to the determination of total sulphite in white wines. Analysis was performed after reacting a 50 ml aliquot of wine with 10 ml of 25% NaOH for 1Omin at room temperature (to release sulphite from com- pounds which bind it, e.g., aldehydes). A 2 ml volume of the treated wine sample was diluted to 100ml with de-aerated, de-ionized water and analysed by the proposed method. For recovery studies, a series of sulphite standards varying from 1.76 to 8.99 pg ml-1 of sulphur dioxide were added to 2 ml of the treated wine in a 100 ml calibrated flask and made up to Table 3 Determination of sulphur dioxide in synthetic samples Sulphur dioxide*/pg ml-1 Automated West-Gaeke Difference Sample? GPMAS method method i”/.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 3.04 2.35 1.73 2.81 4.30 5.51 0.38 0.67 0.79 2.87 6.33 4.49 5.24 4.20 5.33 6.12 3.13 2.39 1.75 2.88 4.29 5.50 0.39 0.69 0.78 2.97 6.53 4.48 5.22 4.30 5.20 6.06 -2.9 -1.7 -1.1 -2.4 +0.2 +0.2 -2.6 -2.9 +1.3 -3.4 -3.1 +0.2 +0.4 -2.3 +2.5 +1.0 Average: 1.8 * Mean of three values. 7 Samples 1-6 were prepared from stock sulphite solution. Samples 7-16 were prepared by absorbing sulphur dioxide generated from stock sulphite solution (after acidification) in 0.04 rnol dm-3 TCM.Table 4 Recovery of sulphite from wines Sulphur dioxide*/pg ml-1 Initially Recovery Sample? present Added Found (% 1 2.26 1.76 3.52 5.27 7.03 8.79 2 3.46 1.76 3.52 5.27 7.03 8.79 3 1.84 1.76 3.52 5.27 7.03 8.79 4.04 5.89 7.62 9.64 11.38 5.18 7.13 8.86 10.67 12.16 3.68 5.49 7.30 9.16 10.48 101.1 103.1 101.7 105.0 103.8 97.7 104.3 102.5 102.6 99.0 104.5 103.7 103.6 104.1 98.3 * Mean of three values. ? Samples 1 and 2 are sweet white wines and sample 3 is a dry white wine. the mark with de-aerated, de-ionized water. Analysis was performed by the proposed method. The results, which are shown in Table 4, indicate recovery in the range 97-105%. Unfortunately, reproducible results were not obtained with red wines. Conclusion The methodology described in this paper should be applicable to polluted air and to other samples that will readily evolve sulphur dioxide upon acidification.Collection of sulphur dioxide samples from air should follow an official procedure.26 The method is particularly attractive because of its simplicity, reproducibility and speed of analysis. Interferences in the proposed method are few and are readily overcome. The interference effect of nitrite is eliminated by adding sulphamic acid while that of heavy metals such as Mn2+, Cu2+ and Fe3+ is minimized by the addition of EDTA. The application of the proposed method to the determination of sulphur dioxide in synthetic samples and in wine has demonstrated the usefulnessANALYST, NOVEMBER 1991, VOL. 116 1139 of the method. Twenty samples can be analysed in 1 h and the results obtained by the proposed method correlate well with those obtained using an official method.The authors thank the Commonwealth Scholarship Commis- sion and the University of Agriculture, Abeokuta, Nigeria, for financial support and leave of absence, respectively, for T. A. A. 1 2 3 4 5 6 7 8 9 10 11 References Environmental Health Criteria &-Sulfur Oxides and Suspended Particulate Matter, WHO, Geneva, 1979. Williams, W. J., Handbook of Anion Determination, Butter- worth, London, 1st edn., 1979, p. 587. United States Environmental Protection Agency, Federal Regis- ter, 1981,46 (January 26). 8352. Frezier, C. D., Ion Chromatogr. Anal. Environ. Pollut., 1979, 2,211; Chem. Abstr., 1980.93, 30944~. Bruno, P., Casalli, M., Delle Monica, M., and Di Fano, A., Talanta, 1979, 26, 1011.Janak, J., and Vecera, Z., Mikrochim. Acta, Part III, 1990,29. Brody, S . S., and Chaney, J. E., J. Gas Chromatogr., 1966, 4, 42. Calokerinos, A. C., and Townshend, A., Fresenius 2. Anal. Chem., 1982,311, 214. Grekas, N., and Calokerinos, A. C., Analyst, 1985, 110,335. Al-Tamrah, S. A., Townshend, A., and Wheatley, A. R., Analyst, 1987, 112, 883. Koukli, 1. I., Sarantonis. E. G., and Calokerinos, A. C., Analyst, 1988, 113,603. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Chaube, A., Baveja, A. K., and Gupta, V. K., Analyst, 1984, 109,391. Balasubramanian, N., and Kumar, B. S. M., Analyst, 1991,116, 207. Bhatty, M. K., and Townshend, A., Anal. Chim. Acta, 1971, 55,263. West, P. W., and Gaeke, G. C., Anal. Chem., 1956,28, 1816. Scaringelli, F. P., Saltzman, B. E., and Frey, S. A., Anal. Chem., 1967,39, 1709. United States Environmental Protection Agency Federal Regis- ter, 1971,36 (April 30), 8187. Logsdon, 0. J., 11, and Carter, M. J., Environ. Sci. Technol., 1975,9, 1172. Cresser, M. S., Proc. Anal. Div. Chem. SOC., 1978, 15, 68. Cresser, M. S., Eur. Spectrosc. News, 1978, 19, 36. Syty, A., Anal. Chem., 1973,45, 1744. Winkler, H. E., and Syty, A., Environ. Sci. Technol., 1976,10, 913. Arowolo, T. A., and Cresser, M. S., Analyst, 1991, 116, 595. Cresser, M. S., and Isaacson, P. J., Talanta, 1976,23, 885. Grekas, N., and Calokerinos, A. C., Anal. Chim. Acta, 1985, 173,311. Methods of Air Sampling and Analysis, American Public Health Association, Intersociety Committee, Washington, DC, 2nd edn., 1977, p. 696. Paper 1102229F Received May 13th, 1991 Accepted July 19th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601135

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1141 Determination of Cobalt in Two Glasses by Atomic Absorption Spectrometry After Flow Injection Ion-exchange Preconcentration M. C. Valdes-Hevia y Temprano, J. Perez Parajon, M. E. Diaz Garcia and A. Sanz-Medel Department of Physical and Analytical Chemistry, Faculty of Chemistry, University of Oviedo, 33006 Oviedo, Spain A flow injection ion-exchange preconcentration scheme, using a minicolumn loaded with Chelex-100 coupled on-line with the atomic detector, is described for the simple and quantitative determination of trace amounts of Co by atomic absorption spectrometry. The detection limit obtained for Co was 20 ng ml-1 with a precision of 21.5% a t the 0.5 pg ml-1 level, by preconcentrating 1 ml of the sample solution. The applicability of the optimized procedure to real problems is demonstrated by determining Co in two commercially available soda-lime-magnesia silica glasses produced in a local factory.The results were satisfactory as indicated by the results of the analysis of the same two samples carried out by inductively coupled plasma atomic emission spectrometry. Keywords: Flow injection on-line; ion-exchange preconcentration; flame atomic absorption spectrometry; cobalt determination; glass analysis Since 1983, when Olsen et al. 1 described the incorporation of a minicolumn into a flow injection atomic absorption spec- trometry (FI-AAS) system for on-line solid-liquid preconcen- tration and matrix removal, a wide variety of papers describ- ing new designs and employing different solid active materials have been published.2.3 The filling materials used in the columns can be classified into three broad groups: (i) organic and inorganic ion exchangers;4?5 (ii) chelating agents such as Chelex-100 (a chelating ion-exchange resin containing imino- diacetate groups) 1.6 and 8-hydroxyquinoline or its derivatives, covalently bound to solid supports, etc;73 and (iii) hydro- phobic materials able to retain metal chelates previously formed in the flow system (sorbent extra~tion).~ It is interesting to note that the success of FI-AAS is mainly related to its potential for routine analysis,lO and to date, the use of minicolumns for on-line sample pre-treatment has been mostly restricted to the determination of heavy metals in waters.In fact, very few applications to solving problems of industrial interest have been reported.1' The absorption of light by glass may be greatly affected by the presence of transition metal impurities present in the original raw materials ( e . g . , iron ions, which are powerful absorbers at long wavelengths). Iron, as Fe3+, imparts a yellow colour to glass, while Fez+ imparts a greenish-blue colour. Cobalt and selenium are the most widely used decolorizing agents for compensating for and neutralizing the green iron colour.12 Because of the high intensity of the cobalt blue colour, control of the amount of cobalt added to aid the decolorization is critical; concentrations of 0.01 ppm of cobalt or less can reduce glass transmittance substantially (about lO%).13 The determination of cobalt at such low concentra- tions in glass is not trouble free and requires sensitive and reliable analytical methods.The usual cobalt concentration in commercially available glass is close to or below the detection limit attainable by flame atomic absorption spectrometry (FAAS). Thus, this work was undertaken to overcome the lack of sensitivity in the determination of cobalt in glass by conventional AAS, by resorting to the use of an on-line preconcentration step. An on-line flow injection preconcentration method for the determination of cobalt in glass by FAAS is reported. This method makes use of a small column loaded with Chelex-100: the affinity of Chelex-100 for the main glass matrix elements (Na, Ca, and Mg) is low while its affinity for cobalt is high.14 Thus, the removal of the bulk glass matrix can be achieved simultaneously with the preconcentration of cobalt with this resin.The procedure could also be extended to the analysis of raw materials used in the production of glass, and other siliceous materials. Experimental Apparatus and Chemicals Atomic absorption measurements were made with a Perkin- Elmer Model 2280 spectrometer and an air-acetylene flame. When necessary, for comparison, cobalt was determined using a Perkin-Elmer inductively coupled plasma (ICP) atomic emission spectrometer, ICP-5000, with direct nebulization. Peak heights were recorded on a Perkin-Elmer Model 56 recorder connected to the spectrometers. Operating condi- tions for absorption and emission measurements are sum- marized in Table 1.A WTW pH meter equipped with a Radiometer GK-2401-C combined glass-calomel electrode was used for all pH measurements. Analytical reagent-grade chemicals were employed for the preparation of all the solutions: sodium citrate (10Y0 m/v, Merck), boric acid (6% d v , Merck), concentrated hydroflu- oric and hydrochloric acids (Merck) and concentrated per- chloric acid (Scharlau). All solutions were prepared with Milli-Q water (Millipore). Standard laboratory ware and glassware was repeatedly cleaned with 10% nitric acid (Merck) and rinsed with water. Standard cobalt solutions were prepared by appropriate dilution of a 1000 pg ml-1 cobalt stock solution (Merck) with a 0.1 mol dm-3 HCl-NH40Ac buffer. The following buffer solutions were used: 0.1 mol dm-3 HC1-NH40Ac (Merck) (pH 2.7) and 1 mol dm-3 NH3-NH4C1 (Merck) (pH 8.5).The resin used was Chelex- 100 (100-200 mesh, sodium form, Bio-Rad). Flow Injection Preconcentration Procedure The flow manifold used is shown in Fig. 1. As can be seen, the flow injection set-up consisted of two Gilson Minipuls 2 peristaltic pumps, a septum for sample injection constructed from Plexiglas, two mixing coils, a rotating valve (Omnifit), connecting poly(tetrafluoroethy1ene) (PTFE) tubes, connec- tions (all made from low-pressure tubing with 0.8 mm i.d.), fittings, a minicolumn to preconcentrate cobalt and the detector. The ion-exchange minicolumn was manufactured from a glass tube (3 mm i.d.). The length of the resin bed was 100 mm, and the bed was held in position by PTFE membranes. The procedure for the preconcentration of cobalt was as follows: the sample (adjusted to a pH of 7.0 k O.l), usually 11142 ANALYST, NOVEMBER 1991, VOL.116 Table 1 Operating conditions for the determination of cobalt FAAS- Observation height Bandwidth Aspiration rate Air flow rate Acetylene flow rate Gain Lamp intensity Absorption line Electrothermal AAS- Spectrometer Graphite furnace Absorption line Measurement Integration time Tube Gas Sample Chemical modifier Heating programme: Temperature/ Step "C 1 90 2 150 3 800 4 1300 5* 2600 6 2700 ICP-A ESS- Spectrometer Wavelength R.f. power Observation height Injector gas pressure Coolant gas flow rate Auxiliary gas flow rate 16.5 cm 0.2 nm 2.5 ml min-1 15.5 I min-1 2 1 min-1 75 v 10 mA 240.7 nm Perkin-Elmer Model 3030 240.7 nm Peak area (deuterium corrector) 4s Pyrolytic graphite coated graphite Ar N-48 0.005 mg of Pd HGA-500 tubes with L'vov platform 10 p1 Flow rate/ Ramp/s Holdls ml min-l 10 20 300 15 10 300 15 10 300 10 15 300 0.t 4 10 1 2 300 Perkin-Elmer ICP-5000 228.62 nm 1.25 kW 15 mm 20 psi 16 1 min-l 2.0 1 min-1 Pumping rate to the sample 1 ml min-l Nebulizer Cross-flow * Read.t Maximum power heating. $ ICP-AES = Inductively coupled plasma atomic emission spec- trometry. 1 ml of sample I FI-AAS AAS w Elution N H~OAC-HCI N H4CI-N H3 (0.1 rnol dm-3) (1 rnol dm-3) buffer buffer Fig. 1 Schematic diagram of the flow manifold ml, was injected through the septum into a channel containing 0.1 rnol dm-3 NH40Ac buffer carrier (pH = 2.7); the sample and the carrier were mixed by flowing through a 3 m mixing coil at an optimum flow rate of 0.5 ml min-1, after which they merged in a T-piece with a 1 rnol dm-3 NH3-NH4CI buffer carrier (pH = 8.5); the resulting solution was pumped at an optimum flow rate of 2.5 ml min-1 along a 5 m coil to achieve the appropriate pH of 7 for retaining cobalt in the mini- column.In order to ensure complete passage of the sample through the minicolumn, 5-6 min were allowed after injection for the preconcentration step (i.e., for a 1 ml sample). When the preconcentration was completed, a 200 pl plug of 5 rnol dm-3 nitric acid was injected through the injection valve in order to release cobalt directly into the nebulizer of the spectrometer. The resulting transient signal was recorded. Optimization of the whole analytical procedure was carried out using aqueous cobalt solutions. Procedure for the Determination of Cobalt in Two Commercially Available Glasses Glass samples, ground to 100 mesh, were provided by Cristalena Espaiiola. A 0.5 g sample was weighed into a clean, dry PTFE dish and in a fume hood, 1.5 ml of concentrated nitric acid and 1.5 ml of concentrated perchloric acid were added.The dish was swirled to wet the sample completely and prevent the formation of lumps. A 7 ml volume of hydroflu- oric acid (48%, Merck) was cautiously added, after which the dish was gently swirled, placed on a hot-plate and heated until fumes of perchloric acid were evolved. This treatment was repeated with 10 ml of hydrofluoric acid and the solution fumed to dryness. The hydrofluoric acid addition and fuming to dryness was repeated four times (the amount required for the decomposition depends on the silica contents of the glasses). After washing the dish carefully with a stream of water and evaporation of the solution to fumes, the contents of the dish were quantitatively transferred into a 25 ml calibrated flask.Five millilitres of hydrochloric acid (1 + l ) , 1 ml of boric acid (6% m/v), 1.5 g of sodium hydroxide, to neutralize most of the acidity, and 1.25 ml of sodium citrate (10% d v ) were added. The solution was diluted to the mark with the 0.1 rnol dm-3 NH40Ac-HCI buffer. Typically, 1 ml of the glass solution was injected into the flow system. The release/detection proceeded as described above. Cobalt determination was accomplished by peak height measurements using the average of three independent readings. Results and Discussion Separation of Cobalt and Optimization Studies The optimum pH range for the retention of cobalt was fairly large (4.5-10).A final pH value of 7 was selected for the uptake of cobalt by Chelex-100 in order to prevent some glass matrix interferences (see under Analytical Performance). Elution of cobalt from the Chelex-100 minicolumn was investigated by using sodium hydroxide , nitric acid, hydro- chloric acid and sulphuric acid solutions, at different concen- trations, as stripping agents. The results showed that cobalt was not eluted in basic solutions but was flushed from the resin by strong acids. A 200 1-11 volume of 5 rnol dm-3 nitric acid was finally selected and used throughout for releasing the cation.The recovery of cobalt was virtually unchanged at carrier flow rates of 2-4 ml min-1. When changing the flow rates, the aspiration rate on the spectrometer was altered accordingly. In further studies, the flow rate was kept constant at 2.5 ml min-1. The ionic strength of the sample solutions was also tested. The effect on the cobalt signal of adding increasing amounts of sodium chloride to the injected sample (1 ml of 0.5 pg ml-1 cobalt solution) was investigated. The presence of up to 0.5 rnol dm-3 sodium chloride had no effect on the retentiordreleasing characteristics of cobalt on Chelex-100 and, therefore, on the analytical signal. The effect of column length on the preconcentration of cobalt was also examined: two columns were prepared with lengths of 5 and 10 cm, both with the same internal diameter of 3 mm.The results were obtained by repeated injections of the same amounts of cobalt for each column (0.5 and 1 pg for both) and indicated very similar peak shapes and heights forANALYST, NOVEMBER 1991, VOL. 116 1143 Table 2 Influence of foreign species on the recovery of 0.3 yg of cobalt Species Na K Ca Mg Fe A1 F Si Species : Co mass ratio 10 10 000 10 1000 10 10 OOO 10 5Ooo 10 700* 10 2000 10 500 8 l0-t Recovery 100 98 99 98 98 95 100 97 99 99 100 95 100 100 100 83 (Yo 1 * In the presence of 0.5% m/v citrate. t During hydrofluoric acid treatment, Si must be removed to levels below 10 yg ml-1. both minicolumns. Dispersion appeared to take place mainly between the column exit and the nebulizer.By using a minicolumn of length 10 cm and 20 mm of connecting tubing downstream from the column to the nebulizer, the observed cobalt response was not dependent on the volume of sample injected at least in the interval tested (from 1 to 10 ml), provided that the total amount of cobalt was kept constant. These conditions give a favourable enrichment. In order to determine the efficiency of cobalt uptake, 0.0847 g (dry mass) of Chelex-100 resin was used for packing and 1 ml of sample containing 0.5 pg of cobalt, adjusted to the optimum pH value, was injected. After elution, the collected effluents were analysed for cobalt by electrothermal AAS (see Table 1 for operating conditions) in order to obtain the breakthrough retention efficiency of the minicolumn for the metal.The recovery of cobalt at a flow rate of 2.5 ml min-1 was 102 k 5% , which corresponds to 1.02 x 10-4 mmol per gram of dry resin. The relatively small retention capacity of the Chelex-100 minicolumn is not a problem here because of the very small amounts of cobalt present in the samples to be analysed. One drawback of this FI system was that the Chelex-100 resin swelled and contracted as the solution composition changed from the carrier pH to the strong acid stripping solution, which hampered compaction of the resin and led to difficulties in using the column for extended periods of time (even when changing flow directions for elution). In order to obtain reliable results, the resin in the minicolumn should be renewed after about 50 sample injections.Analytical Performance Calibration graphs for AAS prepared from the results of triplicate 1 ml injections of cobalt standard solutions were linear up to 1.2 pg ml-1 of cobalt (linear relation: A = 5.50 x 10-3 + 0.24[cobalt], r = 0.9998). The detection limit, calculated as three times the standard deviation of the blank signal, was 20 ng ml-1 and the relative standard deviation (n = 10) at the 0.5 pg ml-1 level was 1.5%. The detection limit observed was about 4-fold better than that obtained by direct aspiration without preconcentration. Chelex-100 has an unusually high preference for heavy metal cations over alkali and alkaline earth metal cations. Thus, the study of interferences was mainly focused on aluminium(iii) and iron(iii),which are present at significant concentration levels in some types of commercially available glasses.Previous studies showed that the optimum retention of aluminium(iii) on Chelex-100 takes place within the pH Table 3 Results of the determination of cobalt in two glasses Cobalt concentrationlyg g-' Flow injection Direct preconcentration nebulization Samples AAS ICP-AES Bronze glass 29.5 ? 1.1 30.0 k 1.4 TSA + 2 glass 19.1 k 0.6 18.9 k 0.3 range 4.5-6.5.6 Therefore, in order to prevent potential interferences by aluminium(i1i) a pH of 7.0 k 0.1 was selected.6 Having selected a working pH of 7.0 k 0.1, the influence of major glass constituents on the recovery of cobalt was investigated and the results are shown in Table 2. As can be seen, iron(iii) did not interfere with the determination of cobalt up to an Fe : Co ratio of 100 : 1.As iron could be present in the glass matrix at a concentration 500 times higher than cobalt, the possibility of removing this interference was investigated. It was found that a final concentration of 0.5% sodium citrate solution had no effect on the exchange of about 0.3 pg of cobalt on the resin, while completely masking the iron (even at a ratio of 700 : 1). After treatment of the sample with hydrofluoric acid, to remove silica, trace amounts of fluoride may be present in the solution. Therefore, this potential interference was studied in more detail. The results demonstrated that a 500-fold concen- tration of fluoride over cobalt had no influence on the uptake of cobalt by the resin. The addition of 0.24% m/v boric acid completely removed the deleterious effects of fluoride on glass laboratory ware.Interference from silica (the major constituent of glass) proved to be fairly serious: Co : Si ratios as low as 1 : 5 lowered the peak heights by more than 20% and hence complete removal of silica during the treatment with hydrofluoric acid must be ensured for reliable results. After the proposed exhaustive treatment of the sample with hydrofluoric acid had been carried out, the relative level of silicon in solution was Si : Co a2-3, which is well below the tolerance level. Analysis of Glass Samples The proposed AAS method was applied to the determination of cobalt in two samples of coloured soda-lime-magnesia silica glasses. The results, summarized in Table 3, were compared with those obtained by direct nebulization of the glass solution by ICP atomic emission spectrometry (AES).Each value is the mean of four independent sample dissolution analyses. The results confirm that the proposed method can be used routinely to determine trace amounts of cobalt with good precision and accuracy in the named commercially available glasses by AAS. In conclusion, the proposed FI separation-preconcentra- tion system coupled with FAAS provides sufficient sensitivity, accuracy, precision and speed to meet the routine quality control requirements of low levels of cobalt in the named commercially available glasses. We gratefully acknowledge the Fundacion para el Foment0 en Asturias de la Investigacion Cientifica Aplicada y la Tecnolo- gia (FICYT) and the Comision Interministerial de Ciencia y Tecnologia (CICYT) for financial support and grants to M.C. V.-H. y T. and J. P. P. Assistance from R. Pereiro Garcia with FI preconcentration is also acknowledged. References 1 Olsen, S., Pessenda, L. C. R., RGiiEka, J., and Hansen, E. H., Analyst, 1983, 108, 905.1144 ANALYST, NOVEMBER 1991, VOL. 116 2 3 4 5 6 7 8 9 10 Tyson, J. F., Anal. Chim. Acta, 1990, 234, 3. Fang, Z., Zhu, Z., Zhang, S., Xu, S., Guo, L., and Sun, L., Anal. Chim. Acta, 1988, 214, 411. Zhang, S., Xu, S., and Fang, Z., Quim. Anal., 1989, 8, 191. McLeod, C. W., J. Anal. At. Spectrom., 1987,2,549. Pereiro Garcia, R., Lopez Garcia, A., Diaz Garcia, M. E., and Sanz-Medel, A., J. Anal. At. Spectrom., 1990, 5 , 15. Marshall, M. A., and Mottola, H. A., Anal. Chem., 1985, 57, 729. Beauchemin, D., and Berman, S. S., Anal. Chem., 1989, 61, 1867. RGiiEka, J., and Arndal, A., Anal. Chim. Acta, 1989,216, 243. Valcarcel, M., and Luque de Castro, M. D., Flow Znjection Analysis-Principles and Applications, Ellis Horwood, Chiches- ter, 1987. 11 McLeod, C. W., Cook, I. G., Worsfold. P. J., Davies, J. E., and Queay, J., Spectrochim. Acta, Part B , 1985,40, 57. 12 Pye, L. D., Stevens, H. J., and La Course, W. C., introduction to Glass Science, Plenum Press, New York, 1972. 13 Campbell, D. E., and Adams, P. B., Glass Technol., 1969, 10, 29. 14 Kingston, M. A., Anal. Chem., 1978,50, 2064. Paper 1/00698C Received February 14th, 1991 Accepted June 25th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601141

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1145 Determination of Bismuth in Geological Materials by Flame Atomic Absorption Spectrometry Using a Selective Extraction Technique Partha Chattopadhyay" and S. S. Nathan Geological Survey of India, Chemical Division, Nongrim Hills, Shillong-793 003, India An investigation has been carried out to establish a rapid method for the determination of bismuth in geological samples (especially geochemical exploration samples) by flame atomic absorption spectrometry using a boosted-output hollow cathode lamp. Bismuth iodide is first extracted into isobutyl methyl ketone and then bismuth is stripped into an aqueous ethylenediaminetetraacetic acid solution for nebulization into an air-acetylene flame. In order to determine the accuracy of the method, geochemical exploration reference samples issued by the US Geological Survey and other international reference samples were analysed using the proposed method and the results obtained were compared with those obtained using other atomic absorption spectrometric procedures.The proposed method can be applied to a variety of geological samples for the determination of bismuth at levels as low as 0.4 ppm. Keywords: Bismuth determination; atomic absorption spectrometry; boosted-output hollow cathode lamp; geological reference samples; extraction with isobutyl methyl ketone The determination of bismuth at the pprn or sub-ppm level is often required in order to study various magmatic processes. The determination of the bismuth content not bound or occluded in insoluble silicates is also desirable for geochemical exploration purposes, where it is used as a pathfinder for the exploration of ores.' A variety of procedures have been described for the determination of bismuth in silicate and other rocks, which include a spectrophotometric method using dithizone after extraction of bismuth iodide into isopentyl acetate,2 a method using atomic absorption spectrometry (AAS) involving extraction of soluble bismuth into nitric acid and nebulization into an air-acetylene flame3 and a polarographic method applied after extraction with diethylammonium diethyldithio- carbamate.4 A method based on hydride generation AAS has been regarded as one of the most convenient and fairly accurate methods for this purpose.5-9 Chattopadhyaylo has reviewed the AAS procedures covering both flame and electrothermal techniques.Bismuth iodide can be extracted quantitatively into isobutyl methyl ketone (IBMK) or butyl acetate," but the direct introduction of an organic solvent extract into a flame or graphite furnace has been unpopular. Difficulties have been reported in accurately dispensing small volumes of organic solvent using electrothermal AAS (ETAAS)'* and for flame AAS (FAAS), direct aspiration of an organic solvent, particularly IBMK, can damage the aspirating assembly unit of some AA spectrometers. 13 Additionally, the volatility of some metal-organic compounds is reported to cause loss of the analyte during the drying and ashing steps of the furnace cycle. Direct introduction of the organic solvent can be avoided as bismuth iodide in IBMK can be stripped with ethylenediaminetetraacetic acid (EDTA) into an aqueous phase for use in the final nebulization step.The Reston Laboratory of the United States Geological Survey (USGS) less frequently uses this procedure for the determination of bismuth by ETAAS and it has reported the validity of this method using reference rocks containing 0.02-1.00 ppm of bismuth.14 The enhancing effects on the absorbance of bismuth caused by aluminium and iron(n) could be minimized by the addition of aluminium (-100 mg for a 0.5 g sample) to both samples and calibration standards8 or more effectively by a solvent extraction procedure. The proposed method mini- * Present address: Regional Research Laboratory, Bhubaneswar- 751 013, India.mizes the presence of concomitant elements in the final stage thereby giving more reliable values for bismuth. With a view to examining the applicability of the method using FAAS (with deuterium arc background correction and a boosted-output bismuth hollow cathode lamp), this paper describes the determination of bismuth in various inter- national geochemical reference samples with particular emphasis on the geochemical exploration samples (GXR- series) issued by the USGS. Recoveries of bismuth from various rock reference samples were also examined in order to study the extraction behaviour of bismuth from different geological matrices. A variety of digestion procedures were also investigated as some of them convert bismuth into an unextractable chemical form.Finally, commonly used sol- vents were evaluated in order to study the extraction behaviour . Experimental Apparatus The analyses were carried out using a Model 902 AA spectrometer equipped with a background correction device obtained from GBC Scientific Equipment Pty., Australia. A bismuth 'super lamp' was used which is a boosted-discharge hollow cathode lamp (HCL). This lamp uses the existing instrument lamp current supply of a standard HCL but adds a boosted-discharge to increase the excitation of atoms sput- tered by the existing lamp supply. A 'super lamp' ensures that all ground-state atoms sputtered from the cathode are excited, which completely removes self-absorption (atomic absorption in and above the cathode). A sharp non-broadened resonance line results.Instrumental parameters were as follows: wavelength, 223.1 nm; slit-width, 0.5 nm; lamp current, 10 mA; integration time, 2 s; and flame, air-acetylene (oxidiz- ing). Reagents All chemicals were of analytical-reagent grade unless other- wise indicated. A bismuth solution (lo00 pg ml-1) was pre- pared by dissolving 1.OOg of bismuth metal (99.99% purity, Merck) in 50 ml of concentrated HN03 and diluting to lo00 ml with distilled water. By successive 10-fold dilutions with 5% HN03 in water, standard solutions of 100, 10 and 1 pgml-1 resulted. All the solvents were distilled before use.1146 ANALYST, NOVEMBER 1991, VOL. 116 Reference Samples Different geochemical reference samples were analysed in order to check the validity of the procedure in different geological matrices.They were: GXR-1 (jasperoid), GXR-2: (soil), GXR-4 (copper mill head) and GXR-5 (soil, B-zone) provided by the USGS; ASK-3 (sulphide ore) provided by Analytisk Sporelement Komite (ASK), Norway; and JG-1 (granite rock), JR-1 (rheolite) and JR-2 (rheolite) provided by the Geological Survey of Japan (GSJ). Detailed sample descriptions are given elsewhere. 15-16 Procedure Weigh 0.2-1.Og of powdered sample accurately into a borosilicate beaker (100ml). Add 10ml of a concentrated HN03-HCI mixture (80% HN03 and 20% HCI) and mix. After allowing to stand for 15-20 min, evaporate the mixture almost to dryness on an electrically heated hot-plate (=2OO0C). Repeat this step. Add 10ml of 1 + 1 H2SO4 solution and heat to fume. Add =30 ml of 5% H2SO4 solution and digest for 30 min, cool and filter through a Whatman No.42 filter-paper. Wash the residue with 2% H2SO4 solution (5-6 times is sufficient). Transfer the filtrate into a 125 ml separating funnel, add 2 ml of aqueous NH2OHeHCl solution (3%), 5 ml of IBMK and 5 ml of aqueous KI solution (4%) in 1% NH20H.HCl. Shake for 2 min and leave until the layers separate clearly. Reject the lower aqueous phase and wash the organic layer twice with 2 ml of aqueous KI in NH2OH.HCI and then 5ml of water. Add 2ml of EDTA (disodium dihydrogen salt, 5 X 10-3 mol dm-3) solution, shake for 2 min, collect the lower aqueous layer in a calibrated tube and repeat this step first with 2 ml and finally with 1 ml of EDTA solution. The final volume should be exactly 5 ml. Aspirate this aqueous solution into an air-acetylene flame.Any dilution, if required, should be performed with the same EDTA solution. Calculate the bismuth concentration directly from a calibration graph prepared under identical conditions. This method was checked with a maximum of 5 g of sample using 20ml of the acid mixture for each step. If the samples contain high levels of organic carbon and sulphur, ignite the samples at about 500°C for 1 h before starting the acid treatment. Results and Discussion Results of Determinations In order to test the accuracy of the dissolution and extraction procedure, the reference samples were analysed. Table 1 shows the results obtained for five replicate analyses of the Table 1 Results for the determination of bismuth (ppm) in geological reference samples and some reported values Sample GXR-1 GXR-2 GXR-4 GXR-5 ASK-3 JG-1 JR-1 JR-2 This work* 1338 0.507 0.407 0.507 0.507 0.617 18.50 34 RSD 0.11 7.61 4.27 3.95 2.32 10.00 6.55 2.83 (Yo) Other AAS values Proposed Terashimat USGSS values§ 1643 1700 1300 0.315 NDll 0.44 19.50 21.20 19.00 0.31 ND 0.37 0.516 - 0.52 0.510 - 0.510 0.650 - 0.650 32 - - * Average of five determinations.t Ref. 9. $ Ref. 17. 0 Ref. 15. 7 A 5 g sample was taken for analysis. (1 ND = Not determined. samples and the relative standard deviation (RSD) and also other AAS val~es.9~17 An excellent agreement with the reported values indicates that the method is accurate for the various types of geological materials examined, except for GXR-1. Recovery Study Known amounts of a standard solution of bismuth were added to 0.5 or 1.0 g of reference rock samples from the GSJ [JG-3 (granodiorite), JGb-1 (gabbro), JP-1 (peridotite), JB-la (basalt) and JA-2 (andesite)].The standard deviations (SDs) for the determination of 5-1Opg of bismuth were less than 0.20 pg, and the recovery was 98-102%, as shown in Table 2. Extraction Behaviour With Different Solvents Although both IBMK and butyl acetate can extract bismuth iodide completely, it was found that stripping with EDTA into an aqueous solution took place most effectively when IBMK was used. Despite a clear and rapid extraction with butyl acetate, this solvent was not preferred because a correspond- ing reduction in sensitivity (in terms of absorbance value for bismuth) was observed. Standard solutions of bismuth (5, 10 and 20 pg) were taken and the general procedure was followed with both IBMK and butyl acetate.Table 3 shows the difference in absorbance values at 223.1 nm. Similar studies with ethyl acetate and diisopropyl ether indicated that these solvents are even less effective than butyl acetate. Decomposition With Different Acids Samples were decomposed using an HF-HC104-HN03 mix- ture [for a 1 g sample add 8 ml of HF (48%), 4 ml of HCI04 (70%) and 4 ml of HN03 (69%) in a poly(tetrafluoroethy1ene) beaker] and after fuming with 1 + 1 H2S04, bismuth was digested with 5% H2SO4 and the usual extraction procedure ~ ~~ Table 2 Precision and recovery in the determination of bismuth BiJvg Sample Takedg Added Found* JG-3 0.5 0.0 NDt 1.0 10.0 10.10 0.5 5.0 5.11 JGb-1 0.5 0.0 ND 1.0 10.0 10.01 0.5 5.0 5.08 JP-1 0.5 0.0 ND 1.0 10.0 9.96 0.5 5.0 5.04 1.0 10.0 9.95 0.5 5.0 4.96 1.0 10.0 9.88 0.5 5.0 4.88 JB-la 0.5 0.0 ND J A-2 0.5 0.0 ND * Average of five determinations.t ND = Not determined. SDhg - 0.16 0.14 0.14 0.19 0.09 0.18 0.14 0.14 0.10 0.10 - - - - RSD Recovery (%) (%) - - 1.58 101 2.74 102 1.39 100 3.74 102 0.90 99 3.50 101 1.41 99 2.82 99 1.01 99 2.05 98 - - - - - - - - Table 3 Comparison of bismuth absorbance at 223.1 nm using IBMK and butyl acetate for extraction Absorbance Absorbance using Bilygml-1 using IBMK butyl acetate 1 0.025 0.011 2 0.048 0.023 4 0.096 0.045ANALYST, NOVEMBER 1991, VOL. 116 1147 was followed. As some samples, particularly those of large sample mass, required repeated attack with acids, this mixture was only chosen for the determination of the total bismuth content in a sample. As for KC103-HCI digestion, the HN03-HCI mixture dissolves most metals not bound or occluded in insoluble silicates, which is desirable for most geochemical exploration applications.Decomposition of sam- ples with a mixture of HN03-HC104, or HN03 or H2S04 yielded incomplete recovery of bismuth. Conclusion The results of the analyses of reference samples and the recovery of bismuth from different rocks indicate that the proposed method can be applied to different geological and related samples of diverse matrices. As most national geo- logical survey organizations are equipped with flame AA spectrometers, this method has a wider application potential compared with the more sensitive ETAAS method (using a carbon tube or quartz cell atomizer).We thank the Deputy Director General, Geological Survey of India (GSI), North Eastern Region (NER) Shillong, for granting permission to carry out this work. We also thank S. Rajarao and the Director (Geochemistry), GSI, NER, for their time, help and encouragement. We are grateful to the USGS, GSJ and ASK, Norway, for providing the reference samples and associated documents. References 1 Moore, C. B., and Canepa, J. A., Anal. Chem.. 1985,57,88. 2 Mottola, H. A., and Sandell, E. B., Anal. Chim. Acta, 1961,25, 520. 3 Ward, F. N., and Nakagawa, H M., US Geol. Surv. Prof. Pap., 4 Russel, H., Fresenius 2. Anal. Chem., 1962, 189,256. 5 Chan, C. Y., Baig, M. W. A., and Pitts, A. E., Anal. Chim. Acta, 1979, 111, 169. 6 Smith, A. E., Analyst, 1975, 100, 300. 7 Thompson, K. C., and Thomerson, D. R., Analyst, 1974, 99, 595, 8 Terashima, S., Anal. Chim. A m , 1984, 156, 301. 9 Terashima, S., Geostand. Newsl., 1984, 8, 157. 10 Chattopadhyay, P., Indian Miner., 1986, 40, 62. 11 Kane, J. S., Anal. Chim. Acta, 1979, 106, 325. 12 Akatsuka, K., Nobuyama, N., and Atsuya, I., Anal. Sci., 1988, 4, 282. 13 Tewari, R. K., Tarasekar, V. K., and Lokhande, M. B., At. Spectrosc., 1990, 11, 125. 14 Wilson, S. A., Kane, J. S., Crock, J. G., and Hatfield, D. B., US Geol. Surv. Bull., 1987, D10, 1770. 15 Govindaraju, K., Geostand. Newsl., 1984, 8, 3 (appendix 111). 16 Ando, A., Mita, N., and Terashima, S., Geostand. Newsl., 1987, 11, 159. 17 O’Leary, R. M., and Meier, A. L., US Geol. Surv. Circ., 1984, 25, 948. 1967,575-D, 239. Paper IfO2552J Received May 30th, I991 Accepted June 25th, I991

ISSN:0003-2654    

DOI:10.1039/AN9911601145

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1149 Improved and Rapid Method for the Spectrofluorimetric Determination of Trace Amounts of Polyacrylamides in Waters Yateman Arryanto and L. S. Bark* University of Salford, Salford M6 5WT, UK The determination of trace amounts of non-ionic, cationic and anionic polyacrylamides in waters has been critically investigated. The polyacrylamides are reacted with hypochlorite from either sodium hypochlorite or Chloramine T to produce amines, which are then reacted with fluorescamine to give intensely fluorescent compounds. There are good linear correlations between the concentration and the fluorescence intensity for all three types of polymers studied. A procedure is proposed for the determination of the polyacrylamides in both fresh and saline waters.The over-all time of the determination is approximately 2 h, which compares favourably with previously reported methods which require 48 h. The relative standard deviation at the 5 mg I-’ level is approximately 5%. Keywords: Polyacrylamide determination; fluorescence; potable water; hypochlorite; fluorescamine Polyacrylamide flocculants play an ever increasing and im- portant role in many aspects of modern life. Whilst some applications, such as the use of the polyelectrolytes to disperse finely ground coal sufficiently for it to be used as a liquid fuel and the treatment of water-based drilling muds with poly- acrylamides to produce rapid changes in the rheology of the muds without a concurrent increase in the mass of the mud, use high concentrations, others in the oil and water industries use very low initial concentrations to cause flocculation of suspended solids.In the oil industry the polyacrylamides are added to the injection waters, used to increase oil and gas supplied from wells, which necessitates forcing sea-water down into the well to ensure that the gadoil reaches the surface. The injection of polyacrylamides ensures the precipitation and dispersion of crystalline products such as barium sulphate produced during the sea-water purge. The current costs of the polyacrylamides are such that it is commercially desirable to ensure that only the minimal excess of that required is introduced into the wells. The potential ecological hazards of the effects of the polymers on sea fauna are such that determination of the trace amounts remaining in the discharge water is potentially important.When polyacrylamides are used in the production of potable waters to remove turbidity, some colour is removed and both organic and inorganic salts can be precipitated. The use of polyacrylamides can increase the rate of settling of the solids,’ reduce over-all costs, and improve the resultant water quality.2 Although the toxicity of polyacrylamides has been reported to be very low at the levels expected in water treatment applications,3 the use of polyacrylamides in the treatment of potable waters is not without potential disadvan- tages because of some of the production processes of the polyacrylamides, which can retain residual acrylamide monomer or other materials. These will remain in the treated water.Results of the analysis of some industrial polyacrylamides show4 that the acrylamide content is of the order of 300 mg kg-1 which is less than the manufacturers’ general specification. McCollister et al.5 have reported a high chromic toxicity for acrylamide and hence there is a potential contaminant problem. In potable water applications, it is recommended6 that the polyacrylamide used should not contain more than 0.05% m/m of residual acrylamide and that it should not be applied at concentrations of greater than 1 mg 1 - 1 for water clarification. However, other poly- * To whom corrcspondcnce should be addressed. acrylamides containing 0.1-5% monomer can be used as industrial coagulants and the effluents from these processes can enter the water supply and pass unchanged through the treatment process of a watenvorks.7 Whilst the toxicity of the polyacrylamides is sufficiently low so as not to constitute a major health hazard for animals, the products formed from the impurities of the polyacrylamides, during the methods used for water purification, have considerable toxicity.4 Although the polyacrylamides react rapidly with ozone to produce compounds of lower relative molecular mass such as acrylic acid, the ozonolysis products of which, appear to be more mutagenic than those of the polyacrylamides, the over-all mutagenicity is not significant with the trace amounts of impurities normally encountered.Whereas the poly- acrylamides show little reactivity towards chlorine, the con- taminants show a great reactivity; however, the chlorinated products have not been reported to have any mutagenic effect. Reviews of methods available8.9 for the determination of polyacrylamides indicate that there are few methods that can be applied to concentrations of less than 20 mg 1-1; in addition, some methods are only applicable to certain types of polyacrylamides and several have limitations with respect to interferences. Some chemical methods involve reactions of :he amide group; conversion into the amine is the basis of a relatively simple and sensitive spectrofluorimetric method10 which, however, requires up to 48 h per determination.Conversion of the amide group into ammonia and subsequent determination of the latter is subject to matrix effects and the methods using a Nessler-type reagent ,I1 a gas-sensing electrode12 or gas chromatography13 have problems with reproducibility. Any laboratory method for possible field use must be robust, easy to operate, fairly rapid and be applicable to a range of polyacrylamide types and concentrations, without being affected by saline conditions such as those found in purge water discharges and estuarine waters which can be used as sources of potable water.It was therefore decided to investigate the possibility of using a rapid method involving conversion of the amide into an amine followed by derivatiza- tion to obtain a fluorescent compound and measurement of the fluorescence using a simple instrument. Scoggins and Miller14315 developed a method for the determination of a variety of amides in aqueous solution which was based on the oxidation of the amide functional group with bromine.They devised conditions to allow the application of this technique to polyacrylamides in the presence of high chloride concentrations. The technique used by Hendrickson and Neuman10 was based on this method and used sodium hypochlorite, but the conditions used by these workers did not1150 ANALYST, NOVEMBER 1991, VOL. 116 give sufficient reproducibility and the method was not sufficiently rapid to be used as a field method, as it required 48 h per determination. It was, therefore, decided to investigate the reactions involved in these methods in order to ascertain the standard conditions that could be used to obtain reproducible results in the minimum period of time.Experimental In preliminary experiments, the following oxidants were used to oxidize known amounts of three types of polyacrylamides: (a) sodium hypochlorite (NaOCl); (b) bleaching powder [CaCl(OCl)]; (c) Chloramine T (the sodium salt of N-chloro- p-toluenesulphonamide); (d) N-bromosuccinimide; and (e) sodium hypobromite. Aliquots of solutions [or suspensions for CaCl(OCl)] of these substances containing the same amounts of ‘available halogen’ (determined by the amount of iodine liberated when reacted with potassium iodide) were reacted with the polyacrylamides under controlled conditions of temperature and time to produce the amines. These were then reacted with a fluorogen and the fluorescence was measured. Several fluorogens were investigated.These included o- phthalaldehyde added in the presence of 2-mercaptoethanol under alkaline conditions,lO ninhydrin16 and fluorescamine,17 which was used in the manner reported by Wormerser and Zbinden. 18 Polyacrylamides As all types of polyacrylamides can be used in the treatment of waters, etc., samples of an anionic polyacrylamide (AA) with an intrinsic viscosity of 764, a cationic polyacrylamide (CA) with an intrinsic viscosity of 28.3 and a non-ionic poly- acrylamide (NA) with an intrinsic viscosity of 21 were obtained from Allied Colloids (Bradford, UK) and were used as supplied. Spectrofluorimeter A Schoeffel fluorescence spectrophotometer interfaced with a low-cost PC (a BBC Microcomputer Type B) was used to measure the fluorescence intensity and to display the plot of fluorescence intensity versus concentration.A simple program was used to give a direct readout of concentration. Hofmann Reaction Although the Hofmann reaction can be carried out under hot or cold conditions,19-21 temperatures above ambient favour the formation of carboxyl groups and do not favour amination; hence a range of temperatures between 0 and 30 “C was studied. Reactions were carried out in thermostated baths at various temperatures for various times with a sample of each type of polyacrylamide and after cooling the solutions for various fixed times, the fluorescence intensity was measured. An aliquot (5 ml containing 5 mg 1-1) of the polyacrylamide under test was placed in a stoppered borosilicate tube and allowed to come to temperature equilibrium.Sodium hydrox- ide (1 ml containing 1 mg ml-1) was added, followed by the halogen oxidant. The concentration of available halogen (as OC1-) was between 0.00015 and 0.20 mg ml-1. After ensuring complete mixing, the sample was allowed to equilibrate for a fixed time at the temperature of the bath and was then allowed to come to ambient temperature for 30 min. The ‘converted’ polyacrylamide was treated with buffer (2.5 ml of the appropriate phosphate buffer) and the fluorogen (0.5 ml containing 1 mg ml-1). The solutions were mixed for a fixed time (20 s) and after standing for 5 min the fluorescence intensity of the solution was measured at the selected wavelength using a 1 cm pathlength cell. A blank determina- tion was carried out for each system. For comparison of the stability and reproducibility of the procedures, experiments were carried out in triplicate.Results From Preliminary Experiments Choice of reagents and experimental parameters Fluorogen. From the standpoint of stability of the fluorogen solutions used, the reproducibility of the results and the intensity of the fluorescence produced, it was decided to use fluorescamine in all further investigations. Oxidant. In general there was very little difference in the results obtained using fresh solutions of any of the oxidants. The results obtained with some samples of bleaching powder were not as reproducible as were other results. The oxidants varied in stability and from this parameter and the fact that it is possible to have dilute solid solutions of Chloramine T prepared in sodium chloride, to help in the provision of suitable reagents for field work, either Chloramine T or sodium hypochlorite is recommended. In all laboratory work, sodium hypochlorite was used and when experimental condi- tions had been established, solid solutions of sodium chloride and Chloramine T were used.Temperature. At temperatures greater than 20°C the intensities of the samples and the blank were not significantly different. Optimum results were obtained at temperatures between 1 and 3°C. Cooling time. The results from experiments using sodium hypochlorite as the oxidant and fluorescamine as the flu- orogen (Table 1) indicate that cooling times of between 30 and 120 min do not have any significant effect on the intensity of Table 1 Effect of cooling time on the Hofmann reaction.Concer tra- tion of polyacrylamide CA used = 5 mg 1-l (intrinsic viscosity = 28.3) Cooling Average fluorescence intensity* time/min (arbitrary units) 0 15 30 f;n 90 120 24h 48h 3.51 3.75 3.80 3.77 3.92 3.64 3.68 3.77 * Each result is the average of three separate readings, each versus a separate blank result. I I I I I I 0 0.02 0.04 0.06 0.08 0.10 0.12 Concentration of OCI-/mg ml-’ Fig. 1 Variation of average fluorescence intensity with concentration of OCI-. 1, Polyacrylamide CA (5 mg I-l); 2, polyacrylamide AA (5 mg 1-1); and 3, polyacrylamide NA (5 mg 1-l). There was no OCI- ion in the blank. The fluorescence is caused by the reaction of the fluorogen with the residual active groups on the polymerANALYST, NOVEMBER 1991, VOL. 116 1151 the fluorescence.Hence a cooling time of 30 min was used in all further experiments. Concentration of oxidant. The concentration of the oxidant was varied between 0 and 0.15 mg ml-1 of available OCI- (from sodium hypochlorite or Chloramine T). The results for the three polyacrylamides are shown in Fig. 1. Proposed Method Reagents Sodium hydroxide solution. Dissolve sodium hydroxide (1 g, general purpose reagent) in water (100 ml). Dilute to a concentration of 1 mg ml-1 prior to use. Hypochlorite solution. Dissolve Chloramine T (2.15 g) in water (500 ml). Dilute to a concentration of 0.215 mg ml-1 prior to use. Fluorescamine. Dissolve 0.1 g of solid in 100 ml of 1,4-dioxane. Use without further dilution. Phosphate buffer. Adjust the pH of a solution (20 ml of 0.05 mol dm-3) of trisodium phosphate to between 8.5 and 9.0 by the addition of a solution of sodium dihydrogen phosphate (0.3 mol dm-3).Dilute to 100 ml with water. Stock solutions of polyacrylamides, 200 mg 1-1. To the polyacrylamide (10 mg) in a 50 ml flask, add absolute ethanol (0.5 ml) followed by de-ionized water to about 25 ml. Shake the mixture gently for 5-10 min, store overnight and dilute to the mark with water. Shake the mixture to obtain a uniformly viscous solution. Calibration solutions are prepared by appropriate dilution of the stock solution with de-ionized water. Procedure Cool an aliquot (5 ml) of the filtered sample in a stoppered tube to between 1 and 3°C. Add cooled sodium hydroxide solution (1 ml containing 1 mg ml-1 at 1-3°C and mix thoroughly. Add cooled (1-3 "C) hypochlorite solution (1 ml containing 0.025 mg ml-1 of OC1- (0.037 mg ml-1 of NaOCl or 0.215 mg ml-1 of Chloramine T).Mix thoroughly and allow to stand at 1-3°C for 30 min. Then allow to come to ambient temperature for 30 min. Add phosphate buffer (2.5 ml, pH 8.5-9.0) and fluorescamine (0.5 ml containing 1 mg ml-1). Mix thoroughly and allow to stand for 5 min. Transfer the solution into a suitable cell (1 cm pathlength) and then measure the fluorescence intensity of the solution using an excitation wavelength of 390 nm and an emission wavelength of 485 nm. Compare the fluorescence intensity with that obtained using a blank solution. Calculate the concentration of the polyacrylamide using a previously prepared calibration graph.Calibration graphs obtained, for the three types of poly- acrylamide used, are shown in Fig. 2. Comparison of the Hendrickson and Neuman Method With the Proposed Method The proposed method was compared with the Hendrickson and Neuman method which uses a mixture of o-phthaldehyde and 2-mercaptoethanol in ethanol as the fluorogenic reagent and a total waiting time of 48 h. It was then compared with s 'modified Hendrickson and Neuman method' using fluor- escamine as the fluorogenic reagent. The results are given in Table 2. Results and Discussion The preliminary results indicated that there is an optimum concentration required to effect the Hofmann rearrangement in the time allotted: below this value it is probable that there are residual unreacted amide groups; above this value it is probable that some of the amine groups are oxidized to nitrogen.Trace amounts of polyacrylamides will be affected by the prolonged presence of residual hypochlorite at the concentra- tions encountered in treated potable waters; however, the effect will be so small that it is virtually zero unless the contact time is of the order of days. It can be noted, from Fig. 1, that variations in the concentrations of hypochlorite ion used of between 0.025 and 0.05 mg ml-1 cause less than a 3% variation in the results. As the concentration of hypochlorite recommended for use is 0.025 mg ml-1 and the concentrations of residual hypo- chlorite generally found in treated water are less than this, the 0 4 8 12 16 20 24 Concentration of polyacrylamidel mg 1-1 Fig.2 Calibration graphs for the determination of polyacrylamides by spectrofluorimetry using fluorescamine. 1, Polyacrylamide CA; 2, polyacrylamide AA; and 3, polyacrylamide NA Table 2 Comparison of the proposed method with the Hendrickson and Neuman method (modified and original) Method Fluorescence intensity (arbitrary units) Wavelength used Time Fluorogen O* 1* 2* 4* 8* require& used L / n m h,,/nm Proposed (2.00)t 0.47 0.95 1.92 3.90 --2 Fluorescamine 390 Hendrickson and Neuman (modified) (2.32) 0.50 1.01 1.88 3.80 48 Fluorescamine 390 Hendrickson and Neuman (original) (1.88) 0.25 0.48 1.00 1.80 48 o-Phthalaldehyde 348 and 2-mercapto- ethanol in ethanol * Concentrations of polyacrylamide in mg 1-1. t Values in parentheses were measured against the blank.485 485 4581152 ANALYST, NOVEMBER 1991, VOL. 116 total concentration of hypochlorite will be less than 0.50 mg ml-1 and the results will be in the acceptable range of variation. Hence, as a result of the reaction of the polyacrylamide with residual hypochlorite from the water treatment, the blank values might be high in samples of water treated with hypochlorite and allowed to stand for some time before being assayed; however, the normalized values will reflect the amounts of residual (previously unreacted) polyacrylamide . When using the method for waters which are to be used for drinking, it is suggested that the assay method is used prior to the chlorination treatment. If the user of the water is supplied with treated water, and is then adding polyacrylamide , the assays for the polyacrylamide content should be carried out immediately before and after use of the solution in order to ensure that the differences in the amounts of polyacrylamide will be caused by the flocculation process and not by the removal of the polyacrylamide by the residual hypochlorite from the water treatment.When used as a routine analytical laboratory method, the proposed method has several advantages over previously reported methods with respect to speed, ease of application and reproducibility. The Hendrickson and Neuman method10 requires 48 h before a result can be obtained. This is of little use in many field operations. Usable results can be obtained with the proposed method within 1.5-2 h from sampling. The proposed method is also less time consuming in the prepara- tion of the fluorogenic reagents.It is suggested that the method is capable of being used under ‘field conditions’ for the determination of trace amounts of polyacrylamides in waters. Any method proposed for use in the field must be sufficiently robust to be able to be used in a routine manner by workers who are only semi-skilled with respect to analytical chemistry. Once the calibration graphs have been established by a skilled worker, the other operations should be made as undemanding as possible. Solutions should be easily prepared and not require frequent calibration; parameters such as temperature and times of reaction should be able to be attained and maintained with ease, with the minimum of attention and without the need for highly specialized equip- ment.When considered for use as a field method, the proposed method has additional advantages. It is sufficiently sensitive for routine field use having a linear range of 0-20 mg 1-1 for each of the three types of polyacrylamide polymers studied. Normally in field operations the actual commercial polymer being used is either known or samples of it are available, and whilst the identity of the polymer might not be, and need not be, known, it is possible to prepare calibration graphs for routine use. The reproducibility of the method at the various levels usually encountered in effluent waters is 5% or less, and in feed waters it is less than 2.5%. Such variations are generally commercially acceptable and hence the sensitivity and repro- ducibility of the proposed method are acceptable.After consideration of the temperatures concerned, it seemed feasible to propose a temperature range of between 1 and 3”C, as this is readily obtained, in practice, by having a water-bath which contains an ice-water mixture and which has been allowed to stand for some time. Hence a temperature of between 1 and 3°C is proposed. Although the spectrofluorimeter uses monochromators rather than optical filters it is still very stable and for field purposes the monochromators can be locked into position so that semi-skilled operators do not have to arrange the optical parameters. There is no doubt that a suitable filter instrument of relatively low cost could be constructed for dedicated instruments.Although either sodium hypochlorite or Chloramine T can be used as the source of the hypochlorite, the main reason for suggesting Chloramine T for use in the field is that it has a greater stability both in the solid form and in solution than does sodium hypochlorjte. Aqueous Chloramine T solutions are very stable. If stored in amber-coloured bottles, their titre does not change for several months. Variations in temperature have very little effect and it is reported that boiling for several hours does not alter the true titre.22 For use in the proposed method, it is recommended that the solution be changed at least every 3 months. Secondly, as Chloramine T is a solid, it is very easy to prepare a ‘solution’ of known concentration of Chloramine T in sodium chloride.The concentration can be arranged so that it can be dispensed in gelatine capsules so that the material from one capsule dissolved in a known amount of water (say, 0.5 or 1 1) provides sufficient solution for, say, assay purposes for 1-3 months. In an analogous way, if the assays are to be carried out on a ‘random one-off basis’, then the amount of solid solution per capsule can be arranged to be sufficient for, say, 250 ml and the solution can be easily and, as weighing is not necessary, rapidly prepared fresh for each batch of assays. The use of an easily transported solid rather than a corrosive solution is another reason for Chloramine T to be recommen- ded for use by non-specialist operators working under field conditions. The method is sufficiently robust to tolerate a slight variation in the amount of hypochlorite used.A sufficiently accurate amount of the hypochlorite, from Chloramine T, can readily and easily be dispensed if capsules containing a fixed amount of the oxyhalogen are used, these can then be dissolved in a known (within 1-2%) volume of water. Fluorescamine is a very stable reagent; previous work23 has shown that solutions of the reagent were stable when kept in amber-coloured bottles for up to 3 months without any noticeable alteration in the fluorogenic properties. The blank does not alter and hence it is assumed that any decomposition products are not fluorescent in the wavelength range used for the fluorescarnine. The fluorogenic reagent used in the Hendrickson and Neuman method has to be prepared immediately prior to use and the separate solutions must be stored in amber-coloured glass bottles under an atmosphere of oxygen-free nitrogen.It is also recommended10 that a calibration graph be obtained each day for the polymer on account of the potential oxidation of the constituents (2-mercaptoethanol and o-phthalaldehyde solutions) which make up the buffered fluorogenic reagent. If this is not carried out, errors can be encountered.10 Although it is feasible under laboratory conditions to have these solutions stored with a rubber septum stopper and dispensed via a hypodermic syringe, such an arrangement would not be desirable under field conditions. Linearity The statistics (Table 3) associated with the method indicate that it gives a good linear correlation (average 0.996) between fluorescence intensity and concentration over the range 0-20 mg 1-1.It is therefore easy to have a simple program giving a direct readout in concentration terms for any set of calibra- tions. The standard deviation (Tables 3 and 4) for any of the polyacrylamides investigated indicates an acceptable repro- ducibility at the commercially relevant levels of below Table 3 Statistical parameters of the calibration graphs for the three types of polyacrylamides Type of Standard Correlation polymer Slope Intercept deviation coefficient Anionic (AA) 0.376 2.125 0.213 0.996 Cationic (CA) 0.475 2.203 0.236 0.997 Non-ionic (NA) 0.295 2.044 0.204 0.995ANALYST, NOVEMBER 1991, VOL. 116 1153 Table 4 Fluorescence intensity obtained using the proposed method Fluorescence Polymer intensity concentration/ (arbitrary units) mg 1-1 Cationic polymer (CA)- 0 2.51 1 2.70 2 3.12 5 4.09 10 6.89 15 9.09 20 11.70 0 2.00 1 2.70 2 2.88 5 3.80 10 6.29 15 7.36 20 9.62 0 1.96 1 2.40 2 2.88 5 3.10 10 5.05 15 5.77 20 8.44 Anionic polymer (AA)- Non-ionic polymer (NA)- 2.10 2.92 3.22 4.70 7.02 9.47 11.89 2.08 2.62 3.20 3.75 6.16 7.56 9.74 1.86 2.54 2.69 3.41 4.83 6.23 8.32 2.42 2.60 3.30 4.63 7.15 9.60 12.06 1.92 2.48 3.16 3.98 6.06 7.78 9.83 2.03 2.51 2.66 3.45 4.93 6.30 7.90 2.32 2.85 3.28 4.12 6.96 9.25 12.12 2.07 2.36 2.74 4.07 6.22 7.67 9.92 2.10 2.42 2.78 3.50 5.16 6.51 7.86 2.41 2.69 3.01 4.53 6.82 9.37 11.92 2.12 2.51 2.98 4.18 6.12 7.42 10.06 2.06 2.30 2.76 3.61 5.10 6.44 8.10 Average fluorescence intensity (arbitrary units) 2.36 2.75 3.20 4.41 6.97 9.36 11.94 2.04 2.53 2.99 3.96 6.17 7.56 9.83 2.00 2.43 2.75 3.41 5.01 6.25 8.12 Relative standard deviation (%) 5.02 3.86 3.03 5.60 1.34 1.59 1.02 3.06 2.85 5.02 3.60 1.10 1.80 1.26 3.67 2.99 1.63 3.77 2.13 3.10 2.52 10 mg 1-1.By using the convention that the detection limit is defined as a signal-to-noise ratio of 2 : 1, the method has a detection limit of 0.5 mg 1-1. In order to ascertain the effect of common salts, solutions of the polyacrylamides were prepared in a synthetic sea-water mixture, the main constituents of which were: sodium chloride ( ~ 2 7 g 1-1); magnesium chloride (=11 g 1-1); calcium chloride ("1 g 1-1); and sodium sulphate (=4 g 1-1). Determinations of concentrations of the polyacrylamides in the range 1-5 mg 1-1 were not significantly different from those obtained using distilled water as the medium.The proposed method is recommended as a rapid, sensitive method for the determination of polyacrylamides in waters and is capable of being adapted for field use. One of us (Y. A.) thanks the Ministry of Technology, Indonesia, for financial support during the period of this work. References Yeh, H., and Ghosh, M. M., J. Am. Water Works Assoc., 1981, 73,211. Robinson, C. N.. Jr., J. Am. Water Works Assoc., 1979,71,226. McCollister, D. D., Hake, C. L., Sadek, S. E., and Rowe, V. K., Toxicol. Appl. Pharmacol., 1965, 7, 639. Mallevialle, J.. Bruchet, A., and Fiessinger, F., J. Am. Water Works Assoc., 1984, 76, 87. McCollister, D. D., Oyen, F.. and Rowe, V. K.. Toxicol. Appl. Pharmacol., 1964, 6, 172. Report of Committee on New Chemicals for Water Treatment, Ministry of Housing and Local Government (USA), Water Treat. Exam., 1969, 18,90. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Brown, L., Rhead, M. M., Bancrott, K. C. C., and Allen, N., Water Res., 1980, 14, 779. Shripchuk, V. G., and Kozubovskii, A. I., J. Anal. Chem. USSR, 1987,6, 299. Wickramanayake, G . B., Vigon. B. W., and Clark, R., in Flocculation, Biotechnology and Separation Systems, ed. Attia, Y. A., Elsevier, Amsterdam, 1987, p. 125. Hendrickson, E. R., and Neuman, R. D., Anal. Chem., 1984, 56, 354. Crummett, W. B., and Hummel, R. A., J. Am. Water Works Assoc., 1963,2, 209. Palma, R. J., Sr., Barad, J., and Palkowetz, J. M., Anal. Lett., Part A , 1984, 17,897. Hawn, G . G., and Talley, C. P., Anal. Chem., 1981, 53, 731. Scoggins, M. W., and Miller, J . W., Anal. Chem., 1975,47,152. Scoggins, M. W., and Miller, J. W., SOC. Pet. Eng. J . , 1979, 19, 151. Samejima, K., Dairman, W., and Udenfriend, S., Anal. Biochem., 1971,42, 222. Weigele, M., De Bernado, S. L., Tengi, J. P.. and Leimgruber, W., J. Am. Chem. Soc., 1972,94, 5927. Wormerser, U., and Zbinden, G., Clin. Chem., 1985,31,1079. Schiller, A. M.. and Suen, L. J., Ind. Eng. Chem., 1956, 48, 2132. Arcus, C. L., J. Polym. Sci., 1952, VIII, 365. Tanaka, H., J. Polym. Sci., Polym. Chem. Ed., 1979,17,1239. Berka, A., Vulterin, J., and Zyka, J., Newer Redox Titrants, Pergamon Press, Oxford, 1965, p. 38 and references cited therein. Bark, L. S., and Cross, L. P., unpublished results. Paper 1/01 994E Received April 29th, 1991 Accepted July 23rd, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601149

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1155 Co-fluorescence of Europium and Samarium in Time-resolved Fluorimetric lmmunoassays Yong-Yuan Xu Institute of Atomic Energy of China, Analytical Chemistry Laboratory, P.O. Box 275-88, Beijing, Republic of China llkka Hemmila," Veli-Matti Mukkala and Sirkku Holttinen Wallac Biochemical Laboratory, P.O. Box 10, 20701 Turku, Finland Tim0 Lovgren Department of Biochemistry, University of Turku, Finland People's In the presence of an excess of Y3+, the fluorescence intensities of Eu3+ and Sm3+, chelated with benzoyltrifluoroacetone (BTA) or thenoyltrifluoroacetone ( T A ) in an aqueous solution containing 1 ,lo-phenanthroline, were increased by factors ranging from 209- to 81 I-fold. This co-fluorescence phenomenon was used in a highly sensitive time-resolved fluorimetric detection of the lanthanides, Eu3+ and Sm3+. The detection limits of Eu3+ in the BTA- and TA-based solutions were 4 and 15 fmol dm-3, respectively.The detection limits of Sm3+ were 0.1 1 and 0.12 pmol dm-3, respectively. The co-fluorescence enhancement systems were also applied in the double-label time-resolved fluorimetric immunoassay of luteinizing hormone and follicle stimulating hormone using specific antibodies la belled either with Eu3+ or Sm3+. The co-fluorescence enhancement solution was superior as compared with the commercial 'direct' fluorescence enhancement solution based on the acidic solution of P-naphthoyltrifluoroacetone, trioctylphosphine oxide and Triton X-100, in respect to the signal level obtained and the sensitivity. It is suited to time-resolved fluorimetric immunoassays in which particularly high detection sensitivities are required, and it can also be used in double-label assays employing Eu3+ and Sm3+ chelate labels.Keywords: Time-resolved fluorimetry; europium and samarium; double-label; fluorimetric immunoassay; lu tein king hormone and follicle stimulating hormone Time-resolved fluorimetry, when combined with fluorescent lanthanide chelate labels, has proved to be a very promising alternative to radioisotopic techniques in the field of clinical immunoassays and deoxyribonucleic acid (DNA) probe assays.1.2 The technique relies on the unique fluorescent properties of the chelates, the long Stokes shift, the narrow- band emission lines, the high quantum yields and the long fluorescence lifetimes, ranging from microseconds to milli- seconds.The long emission duration enables the efficient usage of time-resolved fluorimetry in discriminating the specific signal from background noise .3 Amongst the assay principles published, a technique based on the use of reagents labelled with non-fluorescent Eu3+ chelates, with solid-phase separation and a separate dissocia- tive fluorescence enhancement step is the most widely and successfully applied.4.5 The technique combines a stable labelling reagent, optimized for labelling antibodies,6 with the high fluorescence of a (3-diketone-trioctylphosphine oxide (TOPO) chelates in aqueous detergent solution. Depending on the (3-diketones used, Eu3+-, Sm3+- or Tb3+-labelled antibodies or their combinations can be used.Under certain colloidal conditions in the presence of some other lanthanide ions or yttrium, it is possible to improve considerably the fluorescence intensities of the P-diketone chelates of Eu3+ and Sm3+. This phenomenon was first observed in 1967 with the Eu3+ (and Sm3+)-thenoyltriflu- oroacetone (TTA)-collidine chelate in the presence of Gd3+ or Tb3+.7 The interchelate energy transfer enhancement has recently been studied with the TTA and dibenzoylmethane (DBM) chelates of Eu3+ and Sm3+ and is known as 'co-fluorescence' .8-10 A co-fluorescence effect has been found with a number of P-diketone chelates. A co-fluorescence enhancement system comprising TTA or benzoyltrifluoroacetone (BTA) as the primary ligand, 1,lO-phenanthroline (phen) as the synergistic * To whom correspondence should be addressed.ligand, Triton X-100 as the detergent and Y3+ as the enhancing ion forming the energy donating chelates, was applied in the dual-label time-resolved fluorimetric im- munoassay of luteinizing hormone (LH) and follicle stimu- lating hormone (FSH). Experimental Apparatus Fluorescence spectra and decay times (z) were recorded using a Model LS 5 luminescence spectrometer (Perkin-Elmer, Beaconsfield, Buckinghamshire, UK), in the phosphor- escence mode. The time-resolved fluorimetric measurements of the ions and calculation of immunoassay results were performed on a Model 1234 Delfia research fluorimeter combined with a Model 1221 laboratory computer and the MultiCalc program (Wallac, Turku, Finland).The instrumen- tal parameters for the measurement of Eu3+ were: emission filter, 613 nm; cycling time, 2 ms; delay time, 0.5 ms; and counting time, 1.5 ms. The respective times for the Sm3+ measurement using a 643 nm filter were 1 ms, 0.05 ms and 0.15 ms. Reagents The purity grade of the lanthanide oxides (or their chlorides) was >99.9%. Yttrium oxide (99.9999%) was obtained from Alfa (Karlsruhe, Germany). The stock solutions were pre- pared by dissolving the oxides in hydrochloric acid (or chlorides in water). The BTA was obtained from ICN Pharmaceuticals (Plainview, NY), the TTA and phen from Aldrich (Steinheim, Germany), the DBM, tri(fluoroacety1)- D-camphor, 3-(heptafluorobutyryl)-~-camphor and TOPO from Fluka (Buchs, Switzerland), 2,2'-dipyridine (DP) from Merck-Schuchard (Hohenbrunn, Germany) and P-naphthoyl- trifluoroacetone (P-NTA) from Wallac.2-Furoyltrifluoroace- tone, p-fluorobenzoyltrifluoroacetone and 1,1,1,2,2-~entaflu-1156 ANALYST, NOVEMBER 1991, VOL. 116 oro-5-phenylpentane-3,5-dione were synthesized by the Claisen condensation reaction from the ethyl ester of triflu- oroacetic acid (or pentafluoropropionic acid) and the re- spective ketones using sodium hydride as the condensing agent - 1 1 Di-p-fluorobenzoylmethane, and dithenoylmethane (DTM) were synthesized from the respective acid chlorides by the method described by Siegliz and Horn.12 The stock solutions of BTA, TTA and phen were prepared by dissolving them in ethanol. The monoclonal antibodies against LH, clones 526 and 543, were obtained from Clonatec (Paris, France), and the monoclonal antibodies against a- and P-FSH, the LH and FSH standards, Delfia assay buffer, Eu-labelling reagent, Sm- labelling reagent, wash solution and the direct Delfia enhance- ment solution (containing (3-NTA, TOPO and Triton X-100) were obtained from Wallac.Polystyrene microtitration strips were from Eflab (Helsinki, Finland). Enhancement solution In the experiments on developing the co-fluorescence en- hancement solution, reagents were added to the wells of the microtitration strips in the following order: E d + (or Sm3+); Y3+; Triton X-100; BTA (or TTA); and phen, whereafter the volume was adjusted to 200 p1 with de-ionized water and the pH was adjusted with 20 pl of 10% m/v hexamine. The strips were shaken for 1 min (for the BTA-system) or 8-10 min (for the TT'A-system) before reading the fluorescence intensities on the time-resolved fluorimeter. The final solution is a suspension and not stable during long-term storage.On the other hand, owing to the neutral pH of the final solution, the ions are not dissociated from the chelates used in the labelling of the antibodies. These problems were avoided by storing the solutions in two parts: E,, which contained the P-diketones, Y3+ and Triton X-100 in the acidic buffer (pH 3.1); and Eb, which consisted of phen and the buffer required to raise the pH after dissociation. Labelling of the Antibodies The monoclonal anti-P-LH antibodies were labelled with a 50-fold molar excess of the Eu3+ labelling reagent. The labelling was performed at pH 9.5 by overnight incubation at room temperature. The labelled antibodies were separated from both the free chelates and protein aggregates by gel filtration on a combined column of Sephadex G-50 and Sepharose 6B (Pharmacia).The incorporation yield was 3.8 Eu3+ ions per immunoglobulin G (IgG) molecule. Mono- clonal anti-P-FSH antibodies were labelled at pH 8.6 with a 400-fold molar excess of the Sm3+ labelling reagent. The incorporation yield was 26 Sm3+ ions per IgG molecule. Immunoassa y The dual-label immunoassays were performed in microtitra- tion strip wells coated with monoclonal antibodies against the a-subunits of LH and FSH. The coating was performed with 2 pg of IgG in 200 p1 of 0.1 rnol dm-3 NaH2P04 solution (pH 4.5) by incubation overnight at room temperature.After coating, the strips were washed, saturated with 0.1% bovine serum albumin and stored wet. In the assay the strips were first incubated with 25 pl of LH and FSH standards in 200 p1 of the assay buffer for 1 h, with constant shaking. After washing (twice with the wash solution) the strips were further incubated with a mixture of Eu-anti-P-LH (25 ng) and Sm-anti-6-FSH (500 ng) in 200 p1 of assay buffer for 1 h, with constant shaking. After six washings the specifically bound labels were dissociated with 200 pl of the E, solution, mixed for 3 min and the fluorescence was enhanced by the addition of 20 pl of the Eb solution. After further shaking for 1 min (for the BTA-solution) or 10 min (for the TTA-solution) the fluorescence levels were measured on the time-resolved fluorimeter and the results calculated using the MultiCalc program.Results and Discussion Optimization of the Measurement Conditions In the co-fluorescence enhancement solution two types of (3-diketone chelates are formed, non-fluorescent chelates with Y3+ and fluorescent chelates with either Eu3+ or Sm3+. Under optimum conditions these chelates exist in aqueous solution as clusters, i.e., small aggregates, where the excitation energy is primarily absorbed by the Y3+ chelates, present in a high excess, and the energy is subsequently transferred to the fluorescent chelates, thus producing the highly enhanced Eu3+ or Sm3+ fluorescence levels. As aromatic (3-diketones are the best-suited ligands for the fluorimetric determination of Eu3+ and Sm3+, different aromatic and cyclic (3-diketones with different synergistic ligands and enhancing ions were screened with respect to co-fluorescence enhancement.A co-fluorescence effect was observed with a number of P-diketones, with phen, TOPO or DP as the synergistic ligand and Y3+ as the enhancing ion. With the exception of DTM, all of the tested P-diketones, described under Reagents, showed co-fluorescence to some extent. The strongest fluorescence enhancement was observed with the chelates of TTA and BTA, which were chosen for further studies. Fig. 1 shows the relation between the Y3+ concentration and the fluorescence intensities of Eu3+ and Sm3+ in the BTA-based solution. The optimum concentration of Y3+ was 7.5 pmol dm-3. The effect of varying the concentrations of 2000 I 0 2 4 6 8 10 12 14 16 [Y3+]/10-6 mol dm-3 Fig.1 Effect of Y3+ concentration on the co-fluorescence of: A, Ed+-; and B, Sm3+-BTA-phen chelates. Conditions: 5 x 10-1* rnol dm-3 Eu3+ or 3.5 x mol dm-3 S d + : 5 x 10-5 rnol dm-3 0 20 40 60 80 [BTA]/10-6 rnol dm-3 Fig. 2 Effect of BTA concentration on the fluorescence intensity of: A, Eu3+-; and B, Sm3+- BTA-phen chelates. Conditions: 7.5 x 10-6 rnol dm-3 Y3+; other components (except BTA) as described in Fig. 1ANALYST, NOVEMBER 1991, VOL. 116 I Eu3+ 1157 Sm3+ 2000 I I 0 20 40 60 80 1 00 120 [Phen]/lO-6 rnol dm-3 Fig. 3 Effect of phen concentration on the fluorescence intensity of: A, Ed+-; and B, Sm3+-BTA-phen chelates. Conditions: 7.5 X 10-6 rnol dm-3 Y3+; other components (except phen) as described in Fig.1 2500 m 3 w 8 2000 m s-- . w 0 > 'S 1500 Q) w .- $1000 P C 500 3 LL - t 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 [Triton X-1001 (%) Fig. 4 Effect of Triton X-100 concentration on the fluorescence intensity of: A, Eu3+-; and B, Sm3+-BTA-phen chelates. Con- ditions: 7.5 x 10-6 mol dm-3 Y3+; other components (except Triton X-100) as described in Fig. 1 BTA and phen are shown in Figs. 2 and 3, respectively. At a concentration range of about 50 pmol dm-3 both phen and BTA produce maximum enhancement to both Eu3+ and Sm3+. Similar results were obtained with TTA-based solutions (data not shown). The final solution forms a suspension and was noticed to be unstable for long-term storage. With detergents, however, the stability and fluorescence intensity could be improved. Fig.4 shows the effect of a non-ionic detergent, Triton X-100, on the fluorescence of Eu3+ and Sm3+ in the BTA-based co-fluor- escence solution. The maximum fluorescence intensity was found with 0.005% Triton X-100, but the fluorescence level was not stable (see Fig. 5). At higher Triton X-100 concentra- tions the fluorescence decreased but also stabilized. With 0.02% Triton X-100 the maximum fluorescence was reached after 1 min of shaking, and that concentration was chosen for subsequent experiments. The pH optimum of the system was around 6.8-7.3 (Fig. 6) and the co-fluorescence measurements of Eu3+ and Sm3+ ions were performed in a buffer (pH 7.0) containing 50 pmol dm-3 BTA, 50 pmol dm-3 phen, 7.5 pmol dm-3 Y3+ and 0.02% Triton X-100.The respective values with the TTA-based solution were, pH 7.3, 60 pmol dm-3 TTA, 175 pmol dm-3 phen, 7.5 pmol dm-3 Y3+ and 0.06% Triton X-100. In order to apply the co-fluorescence solutions in time- resolved fluorimetric immunoassays, the components were divided and stored in two parts: the acidic dissociative solutions, E,, containing either 60 pmol dm-3 BTA, 8.5 pmol dm-3 Y3+ and 0.02% Triton X-100 or 70 pmol dm-3 TTA, 7.5 pmol dm-3 Y3+ and 0.075% Triton X-100 in acetate buffer (pH 3.1); and the enhancing solutions, Eb, containing 2500 I m c, ; 2000 D m C 0 E B .c 1500 F . S 9) c. .- A Q) 1000 L! 500 I I I I I I 0 20 40 60 80 100 Time/min Fig. 5 Effect of standing time on the fluorescence intensities of the Eu3+-BTA-phen chelate at different Triton X-100 concentrations.A, 0; B, 0.005; C, 0.01; D, 0.02; and E, 0.04%. Conditions: 7.5 X mol dm-3 Y3+; other components (except Triton X-100) as described in Fig. 1 2000 I m C 3 4.4 8 ,, 1500 0 7 \ > c, ln C .- s 1000 .- 9) C Q) 500 E 3 LL - 4 4.5 5 5.5 6 6.5 f 7.5 8 8.5 9 PH Fig. 6 Effect of pH on the fluorescence intensities of: A, Eu3+--; and B, Sm3+-BTA-phen chelates. Conditions: 7.5 x 10-6 mol dm-3 Y3+; other components (except pH) as described in Fig. 1 either 500 pmol dm-3 phen in 0.2 rnol dm-3 aqueous solution of tris(hydroxymethy1)aminomethane (TRIS) (in the BTA- based system) or 1.75 mmol dm-3 phen in 0.21 rnol dm-3 TRIS solution (in the TTA-based system). Besides Y3+, some of the lanthanides, such as Gd3+, Lu3+, Tb3+ and La3+, could also be used as enhancing ions in the proposed system.In the BTA system Lu3+ and in the TTA system Gd3+ actually yielded more intense enhancement than Y3+ (data not shown), but their use was prevented by the high1158 ANALYST, NOVEMBER 1991, VOL. 116 Table 1 Fluorescent properties of Eu3+ and Sm3+ in BTA-based and TTA-based co-fluorescence enhancement solutions Fluorescence of Fluorescent & X I L n l T I 1 nmol dm-31 Enhancement Background1 Detection limit/ chelate nm nm P 103 counts s-1 factor* counts s-1 pmol dm-3 Eu3+-BTA-phen 333 612 764 41 940 209 1860 0.0043 Eu3+-l'TA-phen 365 612 1062 84 130 526 5700 0.015 Sm3+-BTA-phen 337 647 79 324 358 400 0.11 Sm3+-TTA-phen 358 648 96 602 81 1 750 0.12 * Enhancement factor is defined as the ratio of fluorescence in the presence of Y3+ to that in its absence.10 P A 9 5 4 4 2 3 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 IEu3+] or [Sm3+1/10~mol dm-3 Fig. 8 Calibration gra hs of TTA-based co-fluorescence enhance- ment system for: A, Euf+; and B, Sm3+; and C and D, the respective values with direct enhancement solution. Standards giving signals exceeding lo7 counts s-l for Eu3+ and lo6 for Sm3+ were re-measured with an insensitized parameter group using a delay time of 3.5 ms for Eu3+ and 1 ms for Sm3+ 7 I I J 0.1 1 10 100 1000 [LH] or [FSH]/U I-' Fig. 9 Dose-response curves for dual-label time-resolved fluori- metric immunoassays of: A, LH with Eu3+; and B, FSH with Sm3+. Measured with TTA-based (closed symbols) or BTA-based (open symbols) co-fluorescence enhancement solutions background, which was a result of the E u ~ + and Sm3+ trace contaminations with these ions.Fluorimetric Properties Fig. 7 shows the excitation and emission spectra of the BTA chelates of Eu3+ and Sm3+ in the proposed co-fluorescence solution. At the Eu3+ and Sm3+ concentrations used the emissions of the BTA chelates could not be recorded without Y3+. Similar spectra were recorded from the respective TTA chelates. Fig. 8 shows the calibration graphs for both metals in the co-fluorescence enhancement solutions and in the conven- tional commercial 'direct' enhancement solution. Table 1 summarizes the fluorescence properties of the chelates. The detection limits were calculated from the fluorescence values and the precisions of the zero points. The background present in the TTA solution (contamination of the Eu3+) resulted in a lower practical sensitivity with the TTA-based system.The theoretical sensitivity (signal compared with the background obtained from the plastics and instrument without contamina- tions) reached values below 1 fmol dm-3 (10-15 mol dm-3; <lO-19 mol per well). Double-label Time-resolved Fluorimetric Immunoassay of LH and FSH Fig. 9 shows the dose-response curves for assay of LH and FSH obtained with the two co-fluorescence enhancement solutions. The sensitivities, calculated as the concentration of the hormones in the samples giving a signal equal to the zero response plus three standard deviations, were 0.045 U 1-1 for LH and 0.043 U 1-1 for FSH with the BTA-based solution and 0.045 U 1-1 and 0.029 U 1-1 with the TTA-based solution, respectively.An identical assay with a direct enhancement solution (P-NTA-TOPO-Triton X-100) showed sensitivities of 0.14 U 1-1 and 0.49 U 1-1, respectively. Owing to the improved signal levels the co-fluorescence enhancement system is particularly advantageous for im- munoassays that require especially high sensitivities, provided that antibodies with low non-specific binding properties are available. The proposed co-fluorescence enhancement based on aromatic P-diketones is also suited for use in double-label assays with Eu3+ and Sm3+ as the labels, and further work is being carried out to apply a co-fluorescence enhancement solution, composed of aliphatic P-diketones, to triple- and quadruple-label assays with E d + , Sm3+, Tb3+ and Dy3+ as labels. 1 2 3 4 5 6 7 8 9 10 11 12 References Hemmili, I. A., ZSZAtlas Sci.: Zmmunol., 1988, 1, 231. Soini, E., Hemmila, I., and DahlCn, P., Ann. Biol. Clin. (Paris), 1990, 48, 567. Soini, E., and Hemmila, I., Clin. Chem. (Winston-Salem), 1979, 25, 353. Hemmila, I., Clin. Chem. (Winston-Salem), 1985. 31, 359. Lovgren, T., Hemmila, I., Petterson, K., and Halonen, P., in Alternative Zmmunoassays, ed. Collins, W. P.. Wiley, Chiches- ter, 1985. ch. 12, pp. 203-217. Mukkala, V.-M., Mikola. H., and Hemmila, I., Anal. Bio- chem., 1989, 176. 319. Melenteva, E. B., Poluektov, N. C., and Kononenko. L. I., Zh. Anal. Khim., 1967, 22, 187. Yang. J.-H., Zhu. G.-Y., and Wu, B . , Anal. Chim. Acta, 1987, 198, 287. Ci, Y.-X., and Lan, Z.-H., Anal. Lett., 1988, 21, 1499. Yang, J.-H., Ren, X.-Z., Zou, H.-B., and Shi, R.-P., Analyst, 1990, 115, 1505. Reid, J. C., Calvin, M., 1. Am. Chem. SOC.. 1950, 72, 2948. Sieglitz, A., and Horn, O., Chem. Ber., 1951, 84, 607. Paper 1102137K Received May 7th, 1991 Accepted July 16th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601155

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1159 Mathematical Models for the Fourier Transform Infrared Spectroscopic Determination of ortho-, meta- and para-Xylene in Xylol Salvador Garrigues and Miguel de la Guardia" Department of Analytical Chemistry, University of Valencia, 50 Or. Moliner St., 46100 Burjassot, Spain The analysis of infrared data of binary and ternary mixtures of 0-, m- and p-xylene has been carried out using a series of mathematical models, which permit the quantitative analysis of samples by Fourier transform infrared spectroscopy without the need for absorption cells with a known optical pathlength. The relative errors found in the analysis of binary mixtures are of the order of 1.4%, and those found for ternary mixtures are lower than 3.0% in most instances, for samples containing from 10 to 75% m/m of each component.The proposed method can be applied to the analysis of xylol, a paint solvent that contains the three compounds assayed. Keywords: Fourier transform infrared spectroscopy; quantitative analysis; determination of ortho-, meta- and para-xylene in xylol Fourier transform infrared spectroscopy (FTIR) is a rapid analytical technique that provides information about the qualitative composition of samples and chromatographic effluents. 172 However, quantitative analysis by FTIR requires the use of cells of known thickness and, in general, a previous dilution of the samples with a suitable solvent,3 limitations which often preclude the general application of the technique to quantitative determinations. In recent years, a series of simple models for carrying out quantitative determinations from infrared data, which do not require the use of a known absorption pathlength, have been described.These models, based on the use of the ratio of the absorbances at two well defined wavenumbers, have been applied in order to determine the average degree of condensa- tion of ethylene oxide condensate surfactants435 and the ratio of two compounds in a binary mixture, as applied to the analysis of mixtures of methylparathion and parathion6 and to a series of binary mixtures of oxazepam, medazepam and diazepam .7 In recent work two types of paint solvent have been characterized and analysed by FTIR.8 Problems related to band selection and to the determination of the baseline have been discussed in previous publications.However, the use of the ratio of the characteristic bands of two compounds in a mixture does not permit the determination of the concentration of each component, and only the proportion of the two compounds considered was obtained. In the present paper, the models mentioned above have been applied to the analysis of ternary mixtures and a series of different strategies have been compared in order to obtain accurate results on binary and ternary mixtures of 0-, rn- and p-xylene. Therefore, it is important to note that xylol, a solvent commonly used in the paint industry,g--" is a mixture of 0-, rn- and p-xylene, hence, the proposed treatment is of interest for the analysis of paint solvents. Experimental Apparatus and Reagents A Perkin-Elmer Model 1750 Fourier transform infrared spectrometer with a temperature stabilized coated detector (FR-DTGS) was employed to carry out the absorbance measurements. Potassium bromide cells were used to obtain the infrared spectra of thin films of the samples and 0, rn- and * To whom correspondence should be addressed. p-xylene (Panreac, analytical-reagent grade) were employed as standards.General Procedure Pure compounds, binary mixtures or ternary mixtures of 0-, rn- and p-xylene were weighed and homogenized, and a few drops were placed between two potassium bromide windows. The infrared spectra were recorded between 4800 and 400 cm-1 with a resolution of 4 cm-1 and the spectral range between 850 and 650 cm-1 was selected for the analysis of the compounds studied.A common baseline was established for all the IR absorb- ance spectra considered between 820 and 670 cm-1 and the peak heights or peak area absorbance values were established at the appropriate wavenumbers. The analysis of binary and ternary mixtures was carried out using the mathematical treatment indicated below. Mathematical Treatment Treatment of binary mixtures For a mixture of two compounds 0 and M, in the same spectrum, it follows readily from the Beer-Lambert law that where Al and A2 are the measured absorbance values at the two wavenumbers chosen such that the characteristic peaks for 0 and M do not overlap, ~1 is the molar absorptivity for 0 at wavenumber 1, E~ is the molar absorptivity for M at wavenumber 2, and [O] and [MI are the concentrations of 0 and M.For mixtures of compounds 0 and P, and M and P, respectively, the relevant equations are A3 E3 [PI \- I Experimentally the relationship between two compounds in a sample is established from the absorbance values at two wavenumbers characteristic of each of these compounds and from the ratio of their molar absorptivities. The ratio of the characteristic molar absorptivities of the compounds can be determined not only from a sample1160 ANALYST, NOVEMBER 1991, VOL. 116 containing a known proportion of both compounds6~7 but also from the regression between a series of samples in which the [O] : [PI or [MI : [PI ratios are known and Al : A3 or A2 : A3 was determined experimentally. In this sense two regression models were employed, a linear model, which is directly obtained from eqns.(2) and (3), and a double logarithmic model, which permits the linearization of the relationship between A l : A3 and [O] : [PI for samples containing very different proportions of both compounds. When the linear model is used the ratio E ~ : E ~ can be obtained from the slope of the regression line, and for the double logarithmic model : ~3 is obtained from the intercept. Treatment of ternary mixtures When the three compounds considered, 0, M and P, are present in the same sample a theoretical parameter, total absorbance, AT, can be established which is the sum of the experimental absorbance values obtained at each wavenum- ber considered AT = A1 + A2 + A3 (4) and taking into account the Beer-Lambert law for all the compounds in the same IR spectrum where b is the absorption pathlength.The ratio of the absorbance of one component of the mixture to the total absorbance does not depend on the thickness of the absorption cell and so a general expression, in which Ai, E~ and [i] correspond to any of the three compounds considered, can be written as follows AT = bE1 [0] + b ~ 2 [MI + b ~ 3 [PI ( 5 ) If = e2 = E~ = E ~ , then eqn. (6) can be modified to (7) which directly relates the concentration of one component of the mixture to the experimental results, independently of the optical pathlength of the cell. In order to obtain the expression given in eqn. (7) it is necessary to determine the coefficient which multiplied by ~2 provides the value of E~ and that which can transform ~3 to ~ 1 .From the rearrangement of eqns. (1)-(3) the following equations can be obtained (9) and from these the coefficients kl and k2 can be obtained as follows: By using the kl and k2 values the absorbance values, A2 and A3, at the wavenumber characteristic written as: A2 = bklE1 [MI A3 = bk2~1 [PI and so, of M and P can be and A3 - = 6 E~ [PI k2 Eqn. (5) can be written as: and the percentage of each of the components in a ternary mixture can be found. [O] in YO = [01 x 100 LO1 + [MI + [PI [MI in YO = x100 P I + [MI + [PI [P]in% = x 100 [OI + [MI + [PI - A31k2 x 100 (19) By using the values of A l , A2 and A3 obtained by experiment and from the values of kl and k2, previously established from eqns. (10) and ( l l ) , the concentration of each component can be determined.However, in order to establish the kl and k2 values different approaches can be used. Therefore, k2 is a constant equal to ~ 3 1 ~ 1 . This parameter can be obtained from the inverse of the slope of the regression line between Al : A3 and [O] : [PI or from the intercept of the regression of A1 + A2Ikl + A3Ik2 Log A1 - = log- 1 + log- [OI A3 E3IE1 [PI However, for pure solvents or solutions of high concentra- tion it was observed that E ~ , e2 and E~ did not remain constant for all the different levels of concentration of [O], [MI and [PI and so it is possible to obtain a function which relates kl and [O]:[M] or k2 and [O]:[P]. In order to establish more accurately the percentage of each component in a ternary mixture the optimum value of kl or k2 must be obtained from the kl and k2 functions.For this purpose several models, such as can be employed to obtain the kl value to be used in the analysis of a ternary mixture from the previous determination of the relationship between these two components in the mixture. In these expressions f indicates that k l , or log kl is a function of the ratio of [O] to [MI. Apart from these three basic models, kl and k2 can be obtained from the linear part of the simple logarithmic or the double logarithmic relationship between k and the ratio of the two compounds being studied, providing a series of numerical approaches based on the determination of an optimum estimation of the parameters kl and k2, for the sample considered. Resolution of Ternary Mixtures Using Equations Another way to determine [O], [MI and [PI in a ternary mixture, without the need to use a cell of known thickness is toANALYST, NOVEMBER 1991, VOL. 116 1161 use a compatible and determinate system of three equations which includes the three unknowns.From eqns. (l), (2) and (3) the ratio of the concentrations of two components in the mixture can be expressed as: hence the following equations can be written [O] - U [MI = 0 [O] - w [PI = 0 [MI - 2 [PI = 0 (27) (28) (29) taking into account that in a pure ternary system only three components are considered [O] + [MI + [PI = 1 (30) From eqns. (27)-(30) three systems can be written as follows: (Fl)(i -p -; )= (H) (32) On the other hand a series of equations can be considered which relates the proportion of each component in the mixture to the experimental data.From binary mixtures a series of equations similar to those found for ternary mixtures [eqns. (17)-(19)] can be obtained = R P I - A1 [O] + [MI -Al + A2/kl (34) = T (36) [MI - A2 [MI + [PI - A2 + A3/k3 and E3 k3 =- E2 (37) These equations can be written as a function of [O], [MI and [PI as follows: [O] = [O] R + [MI R [O] = [O] s + [PI s [MI = [MI T + [PI T (38) (39) (40) which are equivalent to: [O] (1 - R) - [MI R = 0 [O] (1 - S) - [PI s = 0 [MI (1 - T) - [PI T = 0 (41) (42) (43) For a sample containing only [O], [MI and [PI using eqn. (30) and eqns. (41)-(43) the following three systems can be proposed to determine [O], [MI and [PI. (F]!)(:t" -4 -!)=(!) (44) (1) (0" 1 3 -F)= (4) (45) 1 1 1 ( W ) ( ' i s 1 ' T - " ) = ( H ) 1 1 1 (46) Thus, [O], [MI and [PI are obtained from the experimental values of Al, A2, A3 and kl, k2 and k3 obtained from the IR spectra of binary mixtures using the different strategies described above.This treatment, and also the other proposed system, can be used for dilute samples if the total concentration of the compounds determined is known, taking into account that in this instance [O] + [MI + [PI = cT and not 1 as stated previously. Results and Discussion Infrared Spectra of ortho-, metu- and puru-Xylene The transmittance spectra of the three compounds considered are shown in Fig. l(a), while in Fig. l(b) the absorbance bands, in the range between 870 and 630 cm-1, corresponding to o-xylene, which has a characteristic band at 743 cm-1, rn-xylene, which has two bands at 770 and 693 cm-1, and p-xylene, which has a band at 796 cm-1, can be seen.As indicated above, a common baseline between 820 and 670 cm-1 can be defined for all the bands considered and, as can be seen in Fig. 2, the characteristic bands of the compounds do not overlap. For the determination of 0-, rn- and p-xylene, absorbance peak heights or absorbance areas can be employed. As indicated in Fig. 2(6), wavenumber intervals from 4 to 14 cm-1 were selected in order to determine the absorbance area. Analysis of Binary Mixtures For the analysis of binary mixtures of xylenes several options such as the selection of the absorbance band and the use of absorbance peak heights or peak areas are available. On the 1 I 1 I I 3142 2000 1428 857 Wavenumbedcm-' Y1 L I l l 837 771 705 Fig.1 PTIR spectra of pure I, o-xylene, 2, rn-xylene and 3,p-xylene. (a) Transmittance spectra; and (b) absorbance spectra in which the more interesting bands between 870 and 630 cm-1 of X, o-xylene-, Y1 and Y2; rn-xylene and Z, p-xylene are indicated1162 ANALYST, NOVEMBER 1991, VOL. 116 other hand, linear or double logarithmic adjustments can be used in order to determine the coefficients which relate the ratio of experimental absorbance data to the ratio of the compounds considered. Band selection For the analysis of binary mixtures of m-xylene with the other compounds considered, each of the two characteristic bands of m-xylene or the sum of the absorbance of both can be used. The kl values found for each of these three possibilities using both a linear regression model and a double logarithmic model are given in Table 1. It can be seen that in all instances good correlation coefficients were obtained, corresponding to the maximum sensitivity being achieved by the use of the M2 band (Y = 692 cm-1).However, good sensitivity and accuracy are obtained and a minimum of data treatment is required using only the band at 769 cm-1 (Table 2); therefore, this band was selected for the analysis of binary mixtures of m-xylene with the other xylenes. Use of peak heights versus peak areas in the analysis of binary mixtures As indicated in Fig. 2 it is possible to use both peak heights and peak areas to determine the ratio of two absorbance bands corresponding to each of the two components of a binary mixture.On the other hand it is possible to take different values of the peak area as a function of the wavenumber interval used for the calculations. 820 770 720 670 Wavenum berkm-’ Fig. 2 FTIR spectra of a ternary mixture of ortho-, meta- and para-xylene in which the common baseline of all the bands considered and the point at which peak height values were measured, in each instance, is indicated; and ( b ) absorbance band of rn-xylene in which the two different intervals employed to determine the peak area are indicated For binary mixtures of 0- andp-xylene the linear adjustment of the experimental data and the double logarithmic adjustment were employed for peak height and peak area values. The values obtained for the regression parameters and the k2 values are summarized in Table 3.The use of band ratios minimizes the sensitivity enhancement obtained from the use of peak area values instead of peak height values. The sensitivity, expressed as the slope of the analytical curve, provides a value of approximately 1.10 in all instances. However, when area values, with a 14 cm-1 wavenumber interval, were employed, a sensitivity enhancement of 15% was obtained. On the other hand it appears that the use of peak height values (see Table 4) provides the most accurate results in the analysis of synthetic samples, especially for samples with a low content of o-xylene, hence use of the ratio of peak height values is recommended. From the experiments carried out it can be concluded that a double logarithmic adjustment of the ratio of absorbance peak heights of the characteristic bands of 0-, m- and p-xylene and the ratio of two of these compounds in pure binary mixtures provide the best analytical performance.These relationships for the three binary mixtures considered are depicted in Fig. 3, which corresponds to the following calibration equations: A0 [ol Am [ml Log - = 0.863 log- + 0.41 Log - A0 = 0.882 log- [OI + 0.23 4 [PI where [o], [m] and [p] are the concentrations of o-xylene, m-xylene and p-xylene, respectively and A,, A, and A, their absorbances, respectively. Analysis of binary mixtures of xylenes By using peak height values of the bands at 744,769 and 796 cm-1 for 0-, m- and p-xylene, respectively, synthetic samples of binary mixtures were analysed by FTIR.The results were obtained using a previous double logarithmic adjustment of experimental data of a series of samples of known composition are given in Table 5 and it can be seen that good accuracy is achieved using the recommended procedure. The analytical figures of merit of the FTIR analysis of binary mixtures are summarized in Table 6, which indicates the improved performance of the double logarithmic data adjustment. Table 1 Adjustment of the experimental data to the [o] : [m] ratio as a function of the bands considered for m-xylene Bands considered Adjustment X (Y = 744 cm-1) Y (Y = 770cm-1) X (Y = 744 cm-1) Y (Y = 770 cm-1+ Y = 693 cm-l) X (Y = 744 cm-1) Y (Y = 693 cm-1) Linear A1 [OI A1 A2 [ml A2 [ml r = 0.998 - = 0.763 - + 0.17 r = 0.997 [OI -- - 1.19-+ 0.26 kl* = 0.840 f 0.014 kl = 1.31 k 0.02 A1 [OI A2 rml - = 2.13 - + 0.50 - _ r = 0.997 kl = 0.499 k 0.010 Double logarithmic A1 to1 A1 [OI A1 [OI A2 [ml A2 [ml A2 [ml Log- = 0.8541og- + 1.01 Log - = 0.863 log - + 0.41 Log - = 0.859 log - - 0.03 r = 0.9993 r = 0.9992 r = 0.9991 kl = 0.66 f 0.02 kl = 1.03 f 0.03 kl = 0.364 f 0.015 * k l = E ~ / E ~ f u.ANALYST, NOVEMBER 1991, VOL.116 1163 Table 2 Determination of the [o] : [m] ratio in binary mixtures as a function of the bands considered Found Linear adjustment Band Sample Added YI (Y = 770 cm-1)- [o] : [m] [o] : [m] Error (%) 1 0.20880 0.1408 -32.6 2 0.64324 0.6545 1.7 3 1.08701 1.1018 1.4 4 2.3948 2.6045 8.8 5 3.9859 3.7213- -6.6 E , = 10.2 YI (Y = 770 cm-l) + Y2 (Y = 693 cm-l)- 1 0.20880 0.1354 -35.2 2 0.64324 0.6471 0.6 3 1.08701 1.0911 0.4 4 2.3948 2.6013 8.6 5 3.9859 3.7076- -7.0 E , = 10.4 Y2 (Y = 693 cm-1)- 1 0.20880 0.1209 -42.1 2 0.64324 0.6285 -2.3 3 1.08701 1.0659 -1.9 4 2.3948 2.5865 8.0 5 3.9859 3.672% -7.9 E , = 12.4 E , = Average relative error.Double logarithmic adjustment [o] : [m] Error (%) 0.2322 11.2 0.6498 1 .o 1.0499 -3.4 2.5330 5.5 3.7254- -6.5 E , = 5.6 0.2213 6.0 0.6253 -2.8 1.0106 -7.0 2.5014 4.5 E , = 5.7 3.6641- -8.1 0.2288 9.6 0.6425 -0.1 1.0380 -4.5 2.5319 5.7 3.7179- -6.7 E , = 5.3 Table 3 Comparison between the use of peak height and peak area in the analysis of binary mixtures of 0- and p-xylene Adjustment Height Area (4)* Area (6) Area (8) Area (10) Area (12) Linear- Slope 1.022 1.030 1.056 1.088 1.122 1.155 Intercept 0.21 0.22 0.21 0.22 0.19 0.17 r 0.9988 0.9987 0.9989 0.9991 0.9993 0.9994 k2 k CJ 0.978 k 0.011 0.971 0.011 0.947 k 0.011 0.919 k 0.010 0.891 k 0.008 0.866 f 0.007 Slope 0.882 0.880 0.885 0.892 0.901 0.910 Intercept 0.23 0.25 0.26 0.28 0.29 0.30 r k2 k CJ 0.794 k 0.016 0.779 f 0.016 0.771 k 0.015 0.756 k 0.015 0.748 k 0.015 0.741 k 0.015 * Numbers in parentheses indicate the wavenumber interval, in cm-1, employed to determine the band area.Double logarithmic- 0.99970 0.99967 0.99967 0.99967 0.99968 0.99969 k2 = E~IEI * CJ. Area (14) 1.185 0.16 0.9994 0.844 k 0.006 0.919 0.31 0.99970 0.733 k 0.015 Table 4 Results of the determination of the o-xylene :p-xylene ratio using peak height and peak area values Found Peak Error Peak Error Sample Added height (YO) area (4)* (YO) 1 0.21601 0.2177 0.8 0.2143 -0.8 2 0.59955 0.6191 3.3 0.6193 3.3 3 1.14659 1.1180 -2.5 1.1108 -3.1 4 2.1534 2.0428 -5.1 2.0374 -5.4 5 3.9715 3.9500 -0.3 3.93_01 -1.0 E,? = 2.4 E , = 2.7 Peak Error area(6) (YO) 0.2131 -1.2 0.6254 4.3 1.1183 -2.5 2.0526 -4.7 3.94_01 -0.8 E , = 2.7 Peak Error Peak Error Peak Error Peak Error area(8) (Yo) area(l0) (YO) area(l2) (YO) area(l4) (YO) 0.6233 4.0 0.6265 4.5 0.6279 4.7 0.6243 4.1 0.2107 -2.5 0.2104 -2.6 0.2096 -3.0 0.2085 -3.5 1.1151 -2.7 1.1237 -2.0 1.1300 -1.4 1.1270 -1.7 2.0478 -4.9 2.0631 -4.2 2.0741 -3.7 2.0676 -4.0 3.91_09 -1.5 3.91_97 -1.3 3.92_39 -1.2 3.89-64 -1.9 E , = 3.1 E , = 2.9 E , = 2.8 E , = 3.0 *_Numbers in parentheses indicate the wavenumber interval, in cm-1, employed to determine the band area.TE, = Average relative error. Analysis of Ternary Mixtures of Xylenes Using k Constants and Functions As has been indicated, in the mathematical treatment of the results, the analysis of ternary mixtures can be carried out using a relationship between the absorbance at a wavenumber characteristic of each compound and the sum of the absorb- ance peak heights of the three bands considered, only if the molar absorptivities of all compounds are corrected in order to obtain equivalent data which permit the use of eqns.(17)- (19) * The kl and k2 values required to correct the absorbance values obtained by experiment can be calculated from the relationship between A,: A, and [o] : [m] and from that of A,: A, and [o] : [p] using both the linear adjustment and the double logarithmic adjustment (as can be seen in Tables 1 and 3).However, when experimental values of the E, : E, and cp : E, obtained from synthetic binary samples with a known proportion of xylenes are plotted versus the 0- : rn-xylene ratio and 0- :p-xylene ratio, respectively, a function and not a constant value is obtained. This may be due to the high1164 ANALYST, NOVEMBER 1991, VOL. 116 2.0 1 .o A d 9 0 0 J -1.0 -2.0 -1.0 0 1.0 2.0 Log [o-lm-xy lene ( mlm )I 2.0 1 .o 4 - a o 0 -I -1.0 I I I 1 I Loglo-lpxylene (m/m)l -2.0 -1.0 0 1.0 2.0 2.0 1 .o A qa s E O 0) 1 -1.0 I f -2.0 -1.0 0 1.0 2.0 Log [ m-/p-xylene (m/m 11 Fig. 3 Calibration equations for binary mixtures of (a) o-lm-xylene; (b) o-lp-xylene; and (c) m-lp-xylene. The graphs relate the logarithm of the ratio of the absorbance at each wavenumber characteristic of each compound and their relation in binary mixtures Table 5 Analysis of binary mixtures of o-, rn- and p-xylene by FTIR o-Xylene (YO) System Sample o-lm-Xylene 1 2 3 4 5 olp-Xylene 1 2 3 4 5 Added Found Error (YO) 17.23 18.84 9.3 39.14 39.39 0.6 52.08 51.22 -1.6 70.54 71.69 1.6 79.94 78.84- -1.4 E , = 2.9 17.76 17.87 0.6 37.48 37.86 1.0 53.41 52.77 -1.2 68.29 67.13 -1.7 79.88 79.84- 0.0 E , = 0.9 m-Xylene (Yo) Added Found mlp-Xylene 1 18.49 18.47 -0.1 2 36.21 36.26 0.1 3 50.31 50.28 0.0 4 68.66 68.08 -0.8 5 82.06 82.01 - -0.1 E , = 0.2 * E , = Average relative error.Table 6 Parameters for the analysis of binary mixtures of xylenes by FTIR. A linear and a double logarithmic calibration line between the ratio of absorbance peak height and the ratio of the two components considered in each mixture were employed Relative standard Sensitivity deviation (YO) Accuracy (Y) Double Double Double logarith- logarith- logarith- System Linear mic Linear mic Linear mic o-lm-xylene 1.19 1.51 ,0.3* 0.3* 6.8 2.9 o-lp-xylene 1.022 1.26 0.5t 0.5t 8.8 0.9 m-lp-xylene 0.786 0.914 0.4$ 0.4$ 4.2 0.2 * Calculated for a ratio [o] : [rn] of 0.64324.t Calculated for a ratio [o] : [p] of 2.1534. $ Calculated for a ratio [m] : [p] of 1.01238. 0.90 0.80 0.70 0.60 0.50 I 1 1 I 1 0 1.0 2.0 3.0 4.0 5.0 Ratio of o-lm-xylene (mlm) / 1 .oo O.’O r 0.60 I 1 I 1 1 I 0 1.0 2.0 3.0 4.0 5.0 Ratio of o-lpxylene (mlm) Fig. 4 Relationship between (a) kl and the o- : rn-xylene ratio and (b) k2 and the 0- :p-xylene ratio.For both, experimental values of k are indicated and A, linear; B, logarithmic; and C, double logarithmic adjustments can be made. A linearization of the first ten experimental values can also be carried out to give the curve D concentration of the samples which are pure xylene mixtures without any inert solvent, hence, under these conditions it is understandable that Beer’s law was not obeyed because of interactions between the different xylenes. The relationship between kl and the [o] : [m] ratio and k2 versus [o]:[p] is shown in Fig. 4. From these data three adjustment models can be employed. In all instances ternary mixtures can be processed from the previous determination of [o] : [m] and [o] : [p] ratios, as recommended earlier, and the appropriate kl and k2 values can be obtained from each of the models described, and from the linear portion of the logarith- mic relationships.ANALYST, NOVEMBER 1991, VOL.116 1165 Table 7 Coefficients employed for the analysis of ternary xylene mixtures kl From the calibration lines- Linear adjustment 0.840 Double logarithmic adjustment 0.664 From k = f(re1ationship)- [OI [ml Linear" k = 0.58 + 0.060 - Linear? [OI [ml k = 0.515 + 0.151- k2 0.978 0.794 [OI [PI k = 0.71 + 0.06 - [ol [PI k = 0.668 + 0.108- 101 [OI [ml [PI k = 0.800 + 0.098 log- Logarithmic k = 0.67 + 0.093 log - [OI [OI [ml [PI logk = 0.120L0g--O.23 Double logarithmic Log k = 0.137 log --0.41 * In the adjustment all the values were considered. t In the adjustment only the first ten values were considered.0 20 40 60 80 100 o-Xylene added (%) 0 20 40 60 80 100 rn-Xylene added (%) 0 20 40 60 80 100 p-Xylene added (%) Fig. 5 Regression between the results found and the amount of compound added in the determination of (a) o-xylene; (b) m-xylene; and (c), p-xylene in ternar mixtures using the logarithmic relation- ships between k , and log [or: [m] and k2 and log [o] : [p]. The solid line indicates the experimental values found and the broken line the theoretical values corresponding to a slope of 1 and an intercept of 0 The values of kl and k2 and the respective equations are summarized in Table 7. By using all these treatments, a series of synthetic samples of 0-, m- and p-xylene were analysed and the average relative error was used as a control parameter in an attempt to select the most appropriate analytical procedure.The regression between the results found, using the kl and k2 values obtained from the logarithmic relationship between kl and log([o] : [m]) and that between k2 and log([o] : [p]) and the real values of the spiked samples are shown in Fig. 5. It can be seen that relative errors lower than 7.0% were obtained in all instances, 3.4% being the average relative error of the total sample population assayed. Analysis of Ternary Mixtures Using Equations The use of the equations obtained from the relationships between [o] : [m], [o] : [p] and [m] : [p] and those from [o] : ([o] account the total concentration of xylenes in the samples considered, provides equations such as those indicated in eqns. (31)-(33) and (44)-(46) and from these the determina- tion of 0-, m- and p-xylene can be carried out.Results obtained using these six systems and also those found from the average of the concentrations obtained for both sets of systems were evaluated using the average relative error of the analysis of 12 synthetic samples as a control parameter. The most accurate results were obtained using the system described by eqn. (32) which provides relative errors of about 1.6%. The results found for each component in each of the samples as compared with the spiked values are given in Table 8. + [ml), [OI ([OI + [PI) and [PI : ([ml + [PI), taking into Conclusions The use of the ratio of absorbance peak heights of different compounds, in both binary and ternary mixtures, provides accurate results for the determination of each of the sample components without the need to use sample cells with a known optical pathlength.Results found in the analysis of xylene mixtures provide highly accurate values which can be used in the characteriza- tion of paint solvents.1166 ANALYST, NOVEMBER 1991, VOL. 116 Table 8 FTIR analysis of ternary mixtures of xylenes using an equation system o-Xylene m-Xylene p-Xy lene Sample 1 2a 2b 3a 3b 4a 4b 5 6a 6b 7a 7b 8 9a 9b 10 11 12a 12b Added 11.78 17.76 17.76 30.84 30.84 45.48 45.48 55.22 66.02 66.02 73.36 73.36 23.32 14.70 14.70 9.23 18.70 70.18 70.18 Found 11.98 17.24 17.97 30.41 30.85 45.26 45.58 54.39 66.64 65.72 73.07 72.56 23.22 14.32 15.12 9.52 18.34 69.57 70.72 Error (YO) 1.7 -2.9 1.2 -1.4 0.0 -0.5 0.2 -1.5 0.9 -0.5 -0.4 -1.1 -0.4 -2.6 2.8 3.2 -1.9 -0.9 0.8 Added 14.52 27.43 27.43 23.04 23.04 30.66 30.66 34.98 17.63 17.63 13.27 13.27 61.69 56.53 56.53 72.31 9.47 18.88 18.88 Found 14.11 27.62 27.38 23.85 22.78 30.53 30.18 35.38 17.07 17.52 13.33 13.60 61.62 56.08 55.61 70.97 9.07 19.09 18.40 Error (Yo) -2.8 0.7 -0.2 3.5 -1.1 -0.4 -1.6 1.1 -3.2 -0.6 0.5 2.5 -0.1 -0.8 -1.6 -1.8 -4.3 1.1 -2.5 Added 73.70 54.81 54.81 46.12 46.12 23.86 23.86 9.80 16.34 16.34 13.37 13.37 14.99 28.77 28.77 18.46 71.83 10.94 10.94 Found 73.91 55.13 54.65 45.74 46.37 24.21 24.24 10.22 16.29 16.76 13.60 13.84 15.15 29.59 29.27 19.50 72.60 11.33 10.87 Error (YO) 0.3 0.6 -0.3 -0.8 0.5 1.5 1.6 4.3 -0.3 2.5 1.7 3.5 1.1 2.8 1.7 5.7 1.1 3.6 -0.6 S.G. acknowledges the grant of the Conselleria d’Educacio i Cikncia de la Generalitat Valenciana to carry out PhD studies. References 1 Mackenzie, M. W., Advances in Applied FTIR Spectroscopy, Wiley, Chichester, 1988. 2 Nyquist, A. R., Leugers, M. A., McKelvy, M. L., Papenfuss, R. R., Putzig, C. L., and Yurga, L., Anal. Chem., 1990, 62, 223R. 3 McDonald, R. S., Anal. Chem., 1984,56, 349. 4 de la Guardia-Cirugeda, M., Carrion-Dominguez, J. L., and Medina-Esriche, J., Analyst, 1984, 109, 457. 5 Carrion-Dominguez, J. L., Sagiado, S., and de la Guardia- Cirugeda, M., Anal. Chim. Acta, 1986, 185, 101. 6 de Benzo, Z. A., Gomez, C., Menedez, S., de la Guardia- Cirugeda, M., and Salvador, A., Microchem. J., 1989,40,271. 7 de JuliBn-Ortiz, J. V., and de la Guardia-Cirugeda, M., Can. J. Spectrosc., 1990, 25, 44. 8 Garrigues, S., and de la Guardia-Cirugeda, M., Anal. Chim. Acta, 1991, 242, 123. 9 Durrans, T. H., Solvents, Chapman and Hall, London, 1971. London, 1982. Academic Press, New York, 1978. 10 Morgans, W. M., Outlines of Paint Technology, Griffin, 11 Lagowski, J. J., The Chemistry of Non-aqueous Solvents, Paper 1 I01 763 B Received April 16th, 1991 Accepted June 12th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601159

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1167 Simultaneous Stopped-flow Kinetic Determination of Gallium and Indium by a Ligand Substitution Reaction Ornella Abollino, Edoardo Mentasti," Corrado Sarzanini and Valerio Porta Department of Analytical Chemistry, University of Torino, Via Giuria 5, 10125 Torino, Italy Louis J. Kirschenbaum Department of Chemistry, University of Rhode Island, Kingston, RI 02881, USA A kinetic method for the analytical determination of binary mixtures of gallium and indium in liquid samples is described. The method is based on the spectrophotometric determination of the variation in absorbance connected with the displacement of a metallochromic ligand from the metal ions to be determined. The different reaction rates of gallium and indium enable the single metals to be distinguished and to be determined down to concentrations of 0.2 pg ml-1, in binary mixtures.Determinations in water samples from the River Po have been performed with satisfactory accuracy and precision. Standard deviations were around 4-5%, and mean relative errors were around 57%. Keywords: Trace metal; kinetic method; indium; gallium; spectrophotometry The determination of trace metal ions in real samples can be achieved with several instrumental techniques now available in almost any analytical chemistry laboratory. Among these, spectrometric methods (atomic absorption, atomic emission and spectrophotometry) and electrochemical methods (polar- ography and voltammetry) are the most widely employed. * Much less frequently utilized are the kinetic methods, which sometimes offer a series of advantages that would make their adoption very useful in certain applications.2J In common with spectrophotometric methods, kinetic methods offer the advantages of instrumental simplicity and low costs and, in common with the more sophisticated instrumental methods, offer the desired selectivities and, in some instances, specifici- ties which are lacking in the spectrophotometric methods.Kinetic methods of analysis for metal species can be divided into two broad c l a ~ s e s : ~ (i) catalytic methods, which indirectly measure a species following its effect on the rate of an indicator reaction; and (ii) true kinetic methods, which determine the amount of a species by measuring the rate of a reaction that directly involves the species under investigation as a reactant.Examples of the first class of methods are redox reactions involving organic substrates, which are catalysed by trace metals,3,4,5.6 and enzymic reactions.2-4.7 In the second class of methods are a wide range of reactions, which can be used for determining reactants in a non-catalytic kinetic method.8-'2 They usually exhibit higher selectivity than the first class of methods, with the exception of the enzymic reactions. This feature is particularly useful in the analysis of mixtures of metals with similar chemical behaviour, such as gallium and indium. In this paper a rapid and convenient method for the determination of trace amounts of gallium and indium in mixtures is described. A kinetic procedure for the determina- tion of trace amounts of indium and gallium has been presented previ0us1y.l~ The method was based on the accelerating effect on the Cu-catalysed oxidation of an organic bi-electron redox couple by hydrogen peroxide.The proposed method is based on a ligand substitution reaction in the complexes of gallium and indium with 4-(2-pyridylazo)resorcinol (PAR), and pertains to the second class of methods reported above. Substitution reactions involving the displacement of a ligand in a metal complex are widely used in inorganic differential kinetic analysis. In these reactions chromogenic ligands, suitable for spectrophoto- * To whom correspondence should be addressed. metrically monitoring the reaction rates are used as the primary ligands and are displaced during the analytical runs by substituting an aminopolycarboxylic ligand [e.g.ethylene- diaminetetraacetic acid (EDTA)]. Alternatively, the chro- mogenic ligand can be displaced by acid dissociation of the primary complex. For each metal of the binary mixture, these two mechanisms can be indicated as follows: M-PAR + EDTA M-EDTA + PAR (1) (2) M-PAR + H+ F M + PAR If a single metal is to be determined, conventional equilibrium absorbance measurements allow the spectrophotometric determination of the analyte; the method is more sensitive and accurate the higher the molar absorptivity shown by the complex, and the more stable the M-L species formed by the chromogenic ligand (the species is thus formed at a higher yield with a lower excess of ligand). Unfortunately, the latter property makes the ligand (usually a metallochromic indica- tor) a less selective reagent.Therefore, for mixtures of metals, a single species cannot be determined independently from the others. However, fortunately, when using kinetic methods this problem can be overcome. In fact, if the metals present in a mixture undergo ligand displacement according to either eqn. (1) or (2), the single contribution to the total absorbance of the mixture of complexes can be temporally separated owing to the particular lability or inertness of each metal.8 Metal species of similar chemical behaviour, such as, for example, niobium and tantalum,14 exhibit remarkably differ- ent kinetic behaviour and their rates of either formation or dissociation of complexes with the same ligand can differ by several orders of magnitude.Therefore, a variation in absorbance can be separated temporally for each metal ion and used as a parameter for the quantitative analysis of their mixtures. The displacement of PAR from gallium and indium, according to either eqn. (1) or (2), occurs with rates that are related to the ease with which the substitution reactions occur at the coordination sphere of these two metals. The mechanisms and rates of formation and dissociation of PAR complexes for both gallium and indium have previously been investigated in detail, and the results have indicated the feasibility of the present method.15 This paper also reports all of the relevant parameters of the system investigated. These include conditional stability constants of the complexes, specific rate constants and other equilibrium and kinetic parameters.1168 ANALYST, NOVEMBER 1991, VOL.116 Experimental Reagents Gallium and indium stock solutions were prepared from analytical-reagent grade nitrates (Merck) and standardized by EDTA compleximetric titrations. The PAR ligand solutions were prepared from analytical-reagent grade products (Fluka). All of the solutions contained 10% v/v methanol, in order to avoid possible separation of insoluble species (ligand or complex), and KN03 or NaC104 up to a total ionic strength ( I ) of 0.80 mol dm-3. All other chemicals such as acids, reagents and EDTA were obtained from Merck. Metal-PAR complex solutions The complexes were prepared by adding an excess of PAR to the solution or the sample containing the metals to be determined and the solutions were adjusted to the desired pH.Water Samples From the River Po Water samples from the River Po were collected in the city of Torino. The water samples were adjusted to pH 3 by the addition of HN03, filtered through 0.45 pm cellulose mem- branes (Millipore) and kept frozen until required. Instrumentation Kinetic measurements were carried out at 25.0 k 0.1 "C using a Durrum-Gibson stopped-flow spectrophotometer (path- length = 2.00 cm) equipped with a Tektronix Model 564 storage oscilloscope .16 Data acquisition, treatment and elab- oration were performed using an Apple 2E personal computer interfaced to the stopped-flow spectrophotometer. Graphical treatment of the experimental data, described below, can be accomplished automatically with simple algorithms within the data acquisition program of the com- puter.Procedure A solution containing gallium, or indium, or both (total concentration S4.0 x 10-5 mol dm-3) is added to a PAR solrsiion of up to [PAR],,, = 8.0 x 10-5 mol dm-3. The solution is adjusted to an ionic strength of 0.80 mol dm-3, and methanol is added up to 10% v/v. The pH of the solution is adjusted to within the range 3-4. After equilibration at temperature (7') = 25.0 "C, this solution is mixed in the stopped-flow spectrophotometer, with a solution of 0.10 mol dm-3 HC104 containing 10% v/v methanol and again adjusted to I = 0.80 rnol dm-3. After mixing, obviously, the final solution has half the initial metal, ligand and proton concentrations.The variation of absorbance, at h = 515 nm, as a function of time is then monitored. Results and Discussion Theoretical Foundation of the Method At a pH of 3 4 , gallium and indium react with PAR to form 1 : 1 M-L complexes. In the presence of an excess of PAR almost complete complex formation is achieved and no superior 1 : 2 complexes are formed. Previous findings on the spectral behaviour of these complexes, and of PAR as a function of added PAR, showed that in this pH range and under these metal and ligand concentration conditions, the distribution of both metals is shifted towards full complex- ation. 15 When a solution containing gallium, or indium, and an excess of PAR (adjusted to pH 3-4, I = 0.8 mol dm-3, and at 25.5"C) is mixed with a solution of HC104 (adjusted to the same ionic strength and temperature) so that the final concentration of H+ in the mixture is 0.050 mol dm-3, a rapid dissociation of the M-PAR complex occurs and a decrease of absorbance at 515 nm can be measured as a function of time.The relative rapidity of the reaction (the time taken for 50% completion of the reaction is about 0.0440 s, depending on acidity and type of metal) requires the use of a stopped-flow spectrophotometer. Under these conditions the rate of reaction is of pseudo-first order with respect to the displaced complex, all of the other parameters being constant. d [M-PAR] Rate of disappearance of M-PAR = - d t = kobs [M-PAR] (3) Integration of eqn. 3 yields: [M-PAR10 - [M-PAR], [M-PAR], - [M-PAR], In = kobs x t (4) or A0 -Am ( 5 ) In ~ = kobs X t At- Am where, A is the absorbance, kobs is the pseudo-first order rate constant, and subscript indexes 0, t and refer to zero time, time t , and time to attained equilibrium.A plot of ln(A, - Am) as a function of time should be linear, with a slope equal to -kobs and an intercept equal to ln(A0 - Am). As an example, the results obtained for a typical run are shown in Fig. 1. From the linear plots, such as that in Fig. 1, it is possible to evaluate from the intercept, the absorbance variation connected with the metal concentration initially present in the solution, and from it, the concentration of the metal: (A0 - Am) = AAtot = (EM-PAR - &PAR) x b x [Mltot (6) (7) where E is the molar absorptivity of the species and b the light pathlength. Kinetic Determinations Single metals According to the procedure described, it is possible to evaluate AA,,, for the dissociation of the gallium or indium complexes from semilogarithmic plots according to eqn.(5). The determination of the metal concentration is then effected through eqn. (7) using A& = ( E ~ - ~ ~ R - &PAR) obtained with the same equation from standard solutions of known concen- tration. To give an example, Fig. 2 shows the plot of AA,,, as a function of [MI,,, for a series of solutions of known composi- 4 3 4 I - 6 2 C 1 I I I I I 1 10 20 30 40 50 0 Time/s Fig. 1 Variation of In (A, - A, as a function of time according to eqn. (5), for a solution of Ga: [Gal. 1.0 x rnol dm-3; [H+], 0.050 mol dm-3 and [PAR], 4.0 x 10-5 mol dm-3ANALYST, NOVEMBER 1991, VOL.116 1169 tion of indium and gallium. The values of A& x b for gallium and indium were 6.8 x 103 and 1.0 X 104 1 mol-1 cm-1, respectively, at h = 515 nm. Mixtures Gallium- and indium-PAR complexes dissociate in acidic medium at different rates. This is the basis for the determina- tion of both metals in their mixtures with the present kinetic method. In Fig. 3, the variation of transmittance at 515 nm for a typical run obtained for a mixture of gallium and indium is shown. From the slower reaction (i.e. for gallium) it is possible to obtain by the graphical treatment described above, the total variation in absorbance, AAtot. This quantity makes it possible to determine gallium, by using eqn. (7). In addition, the absorbance value corresponding to completion of the faster reaction A’,, i.e.for indium, can be evaluated. A second plot of ln(A, - A’m), according to eqn. (4) finally enables the calculation of (A’o - A’,) = (AAtot)ln. From this quantity, the determination of indium becomes straightfor- ward using eqn. (7). Table 1 gives the results of a series of determinations in single and binary mixtures of the elements. Table 2 gives the results obtained for the determination of indium and gallium in spiked water samples from the River Po. The mean relative standard deviation estimate is around 4.5 and 5.4% for 0.2 8 lp o.l 0 .p I I 1 .o 2.0 [MIA05 rnol dm-3 Fig. 2 Calibration graphs for Ga and In, according to eqn. (6), obtained using standard solutions of Ga and In gallium and indium, respectively, and the mean error is 7.2 and 4.8% respectively.As can be seen, the accuracy and precision appear satisfactory, thus, showing the feasibility of the determination of these binary mixtures in real samples, at concentrations ranging between 0.2 and 2 yg ml-1. In addition, the particularly high stability of indium and gallium complexes with PAR, even at pH 3.5 (the present starting conditions before ligand replacement) makes the present method very selective and almost insensitive to the presence of other metal species. It must be noted that the remarkably high difference in rate for indium and gallium (3 orders of magnitude) makes the single reaction rates almost independent of each other, thus, the errors in the determination of the single species, in their mixtures, does not depend upon the relative ratios of analyte concentrations, but rather on the sensitivity of the spectro- photometric detection of AA.As a consequence, mean relative errors below +7% are expected down to concentra- tions of 0.2 yg ml-1. Conclusions The proposed method is based on the spectrophotometric determination of a mixture of two components; evaluation of the single analyte is performed using the temporal difference Time/s 0 I I I 0 50 100 150 Time/ms Fig. 3 Variation of transmittance, as a function of time, at h = 515 nm (cell lightpath, b, = 2.00 cm) for a mixture of In and Ga, both 1.0 x 10-5 mol dm-3; [H+], 0.050 mol dm-3; and [PAR], 4.0 X mol dm-3 Table 1 Determination of gallium, indium and gallium + indium in synthetic mixtures with the proposed substitution reaction [Ga3+]/105 rnol dm-3 [In3+]/105 mol dm-3 Error Error Added Found (Yo 1 Added Found (Yo) A, AAtot A ‘ , 0.30 0.60 0.90 1.20 1.50 1.80 2.10 - - - - - - - 1.60 1.40 1.20 1 .oo 0.80 0.60 0.40 0.33 0.64 0.90 1.21 1.51 1.78 2.11 - - - - - - - 1.64 1.38 1.11 0.95 0.76 0.62 0.37 + 10.0 +6.6 0 +0.8 +0.7 -1.1 +0.5 - - - - - - - +2.5 -1.4 -7.5 -5.0 -2.0 +3.3 -7.5 - - - - - - - 0.30 0.60 0.90 1.20 1.50 1.80 2.10 0.40 0.60 0.80 1.00 1.20 1.40 1.60 - - - - - - - 0.33 0.59 0.74 1.18 1.63 1.81 2.08 0.39 0.56 0.87 0.95 1.25 1.42 1.57 - - - - - - - +10.0 -1.7 -17.7 -1.7 +8.7 +0.6 +1.0 -2.5 -6.6 -5.0 +4.2 +1.4 +8.8 -1.9 0.4685 0.4685 0.4685 0.4660 0.4622 0.4687 0.4672 - - - - - - - 0.4609 0.4672 0.4647 0.4647 0.4622 0.4660 0.4672 0.0226 0.0438 0.061 1 0.0819 0.1029 0.1213 0.1434 - - - - - - - 0.1115 0.0938 0.0755 0.0646 0.0518 0.0421 0.0253 - - - - - - - 0.4672 0.4672 0.4660 0.4647 0.4685 0.4634 0.4660 0.5714 0.5610 0.5402 0.5293 0.5140 0.5081 0.4925 AA’m - - - - - - - 0.0331 0.0590 0.0738 0.1178 0.1627 0.1809 0.2081 0.0390 0.0565 0.0870 0.0948 0.1249 0.1420 0.15701170 ANALYST, NOVEMBER 1991, VOL.116 Table 2 Determination of gallium and indium in spiked water samples from the River Po, using the proposed substitution reaction. Each determination represents the mean of five totally independent individual measurements (ND = not detectable) [Ga3+]/105 mol dm-3 Added Found 0 ND 0.50 0.46 * 0.03 1.00 1.08 f 0.05 1.50 1.41 k 0.06 0.50 0.44 k 0.02 1.50 1.53 -C 0.04 [In3+]/105 mol dm-3 Error Error (%) Added Found (%) 0 ND -8.0 0.50 0.53k0.04 +6.0 +8.0 1.00 0.97-CO.06 -3.0 -6.0 1.50 1.55 k0.08 +3.3 -12.0 1.50 1.56-CO.07 +4.0 +2.0 0.50 0.54-CO.02 +8.0 of each reaction completion.The graphical treatment depic- ted by eqns. (3)-(5) enables the derivation of AA,,, for each metal from the linear regression analysis of several absor- bance-time data couples and this procedure allows a more accurate evaluation of AA even for very dilute solutions. Table 1 shows the results of the determination of variations in absorbance as low as 0.02. On one hand, this shows the sensitivity of the method and on the other, points out that extrapolation from a set of short time-sampled data allows the reduction of the inaccuracy from bias, signal noise and drift. Financial support from the National Research Council of Italy (CNR, Rome) is kindly acknowledged. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 References Hunt, D. T. E., and Wilson, A. L., The Chemical Analysis of Water, The Royal Society of Chemistry, Cambridge, 2nd edn., 1986. Kopanica, M., and Stara, V., Kinetic Methods in Chemical Analysis, Elsevier, Amsterdam, 1983. Perez-Bendito, D., and Silva, M., Kinetic Methods in Analytical Chemistry, Ellis Horwood, Chichester, 1988. Perez-Bendito, D., Analyst, 1984, 109, 891. Mottola, H. A., and Heath, G. L., Anal. Chem., 1972,33,2322. Rubio, S . , Gomez-Hens, A., and Valcarcel, M., Anal. Chem., 1984, 56, 1417. Perez-Bendito, D., Silva, M., and Gomez-Hens, A., TrAC, Trends Anal. Chem. (Pers. Ed.), 1989, 8, 302. Ridder, G. R., and Margerum, D. W., Anal. Chem., 1977,49, 2090. Lagrange, J., Lagrange, G., and Zarem, Z., Bull. SOC. Chim. Fr., 1978, I, 7. Mentasti, E., Anal. Chim. Acta, 1979, 111, 177. Mentasti, E., Dlask, V., and Coe, J. S., Analyst, 1985, 110, 1451. Funahashi, S., Ito, Y., Kakito, H., Inamo, M., Hamada, Y., andTanaka, M., Mikrochim. Acta, 1986, 1, 33. Marin, A., Silva, M., and Perez-Bendito, D., Anal. Chim. Acta, 1987, 197,77. Yamada, S., Anma, H., and Murata, A., Anal. Sci., 1988,449. Mentasti. E., Baiocchi, C., and Kirschenbaum, L. J., J. Chem. SOC. Dalton Trans., 1985, 2615. Mentasti, E., Znorg. Chem., 1979, 18, 1512. Paper 1l0181.51 Received April 18th, 1991 Accepted June 25th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601167

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1171 Kinetic Enthalpimetric Determination of Vanadium(v) by Its Catalytic Effect on the Oxidation Rate of Pyrogallol Red by Potassium Bromate Rafael Forteza, Jose Manuel Estela and Victor Cerda* Department of Chemistry, Faculty of Sciences, University of the Balearic islands, E-07071 Palma de Mallorca, Spain The Vv-catalysed oxidation of Pyrogallol Red by Br03- was monitored enthalpimetrically. The rate of the reaction was determined graphically from the initial linear slope of the temperaturetime curve recorded and was found to be proportional to the V V concentration over the range 10-150 ng ml-1. The relative standard deviation obtained for the determination of a 100 ng ml-1 concentration of the analyte was 6%. Foreign species did not interfere at concentrations as high as 20pg mi-1.The method was applied to the determination of V V in steel with good results. Keywords: Kinetic enthalpimetric method; vanadium(v) determination; Pyrogallol Red; catalysis There are a number of sensitive catalytic spectrophotometric methods for the determination of Vv.l-ll However, most of these methods are not very selective and, hence, are subject to interference from ions such as Fell1, Ti1', MeV' and Wvl, which sometimes makes them inapplicable to real samples or, more often, requires prior removal of the interferents. The catalytic spectrophotometric method" based on the oxidation of Pyrogallol Red (PGR) by Br03- catalysed by trace amounts of Vv is a representative example. In spite of its good sensitivity (detection limit, 0.61 pg I-]), broad linear range (between 0.61 ng ml-1 and 1.83 pg ml-1) and satisfactory precision and reproducibility, it is not very tolerant to the presence of other ions and therefore lacks selectivity.Hence, for a Vv concentration of 10 ng ml-1 and a tolerated interfer- ent level of 2%, the tolerated Vv to interferent ratios for several ions are as follows: 1 : 2 for TiIV, 1 : 5 for SnIv and MoV1, 1 : 10 for Cr"' and Wvl, 1 : 30 for All1' and 1 : 60 for Fell'. In previous work,12-15 kinetic catalytic thermometric methods for the determination of inorganic ions were developed that were more selective than their spectrophotometric counterparts in most instances. This, and the lack of reports on the kinetic catalytic oxidation of PGR and its potential application to the determination of different species, prompted us to carry out a kinetic thermometric study on the PGR-Br03--VV system.As a result, a kinetic catalytic direct injection enthalpimetric method for the determination of Vv was developed. The method was applied to the determination of Vv in steel samples with good results. Experimental Reagents A stock standard solution containing 1000 pg ml-1 of Vv was prepared by dissolving 2.297 g of analytical-reagent grade NH4V03 in 1 + 99 HN03. The solution was transferred into a 1OOOml calibrated flask and diluted to the mark with 1 + 99 HN03. From this stock solution, working-strength solutions were prepared daily by dilution with de-ionized water. Other solutions used included: a 2.2 mol dm-3 H3P04-NaOH buffer of pH 3; a 1.25 x mol dm-3 PGR solution prepared by dissolving 1 g of the reagent (Merck) in de-ionized water together with two NaOH pellets and then making up to 200ml; and a 1.4moldm-3 Br03- solution prepared by dissolving 10.56 g of NaBr03 (Merck) in 50 ml of de-ionized water.All the above solutions were stored in a thermostated room (22 k 1 "C), where the experiments were performed. * To whom correspondence should be addressed. Apparatus The temperature monitoring system consisted of a highly responsive thermometer-type thermistor with a nominal resistance of 100 kQ at 25 "C, a Wheatstone bridge supplied with 8.93 V from a stabilized source, and a chart recorder. The sensitivity was 0.0055 "C mV-1 with a full-scale deflection on the recorder of 20 mV (25 cm).The thermometric system has been described in detail elsewhere.12 Bromate was introduced into the thermometric cell by means of a syringe. A propeller stirrer operated by a synchronous motor was used to ensure thorough mixing of the reactants. The reaction rates were measured using the initial-rate method, i.e., by measuring the tangent dT/dt (where T denotes the temperature and t the time) over the initial linear portion of the temperature rise and given as tan a in Figs. 1-3, where a is the angle of slope of the linear portion. Procedure Into a 75ml polystyrene cell were placed 5ml of the PGR solution, 20 ml of the pH 3 buffer, and the required volume of the Vv solution to obtain a final Vv concentration of between 10 and 150 ng ml-1; the mixture was then made up to a volume of 49ml with de-ionized water.As soon as the temperature had stabilized, 1 ml of the NaBr03 solution was added rapidly with the aid of a syringe and the changes in the solution temperature were measured as a function of time. Procedure for the Determination of Vv in Steel A 0.5g amount of steel was accurately weighed and transferred into a 600 ml beaker to which 20 ml of 1 + 1 H2S04 were then added. The solution was heated until the effervescence had almost ceased after which 5ml of concentrated HN03 were added cautiously. The solution was boiled until complete dissolution was obtained and the nitrogen oxide fumes had ceased. Finally, the solution was filtered if required and diluted to a final volume of 100 ml. The working solution was prepared by dilution of the above solution with de-ionized water to give a Vv concentration in the range 10-150 ng ml- 1 .Results and Discussion Preliminary assays showed that the kinetics of the oxidation of PGR by Br03 in the absence of Vv gave no temperature changes, whereas trace amounts of Vv produced marked temperature changes that could be measured. Such changes were manifested by the decoloration of PGR.1172 ANALYST, NOVEMBER 1991, VOL. 116 1.00, I 0.80 tf 0.60 C I-" 0.40 0.20 0: 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 PH Fig. 1 [PGR] = 2.5 X 10-3 rnol dm-3; and [Vvf = 100 ng ml-1 Effect of pH on the initial rate: Br03-] = 0.028 mol dm-3; 0.20 I I I I I -1 0 0.5 1 .o 1.5 2.0 2.5 [PGR]/mmol dm-3 Fig. 2 0.028 mol dm-3; and [Vv] = 100 ng ml-* Effect of the PGR concentration on the initial rate: [Br03-] = 1.20 1 .oo 8 0.80 I-" 0.60 0.40 0.20 0 5.0 10 15 20 25 30 [Br03-]/mmol dm-3 Fig.3 1.25 x 10-3moldm-3 and [Vv] = 100ngml-1 Effect of the Br03- concentration on the initial rate: [PGR] = The influence of the acidity of the medium and the concentrations of PGR, Br03- and Vv on the reaction kinetics was studied in order to establish the optimum Vv concentra- tion to be used, namely that resulting in the maximum initial rate (i.e. , in pseudo-zero order conditions). The initial rate was not affected by small variations in the reactant concentra- tions. Fig. 1 shows the results obtained by studying the influence of the acidity on the reaction rate. The experimental conditions used are given in the figure caption, and also in all subsequent figures.As can be seen, the maximum slope was obtained at a pH of about 3. Higher or lower pH values resulted in lower reaction rates. Table 1 Tolerated interferent to Vv ratio: [PGR] = 1.25 X 10-3 rnol dm-3; [NaBr03] = 0.028 rnol dm-3; [Vv] = 100 ng m1-I; and pH = 3. Maximum tolerance = +2 SD Ion [Ion] : [Vv] Ni2+, Mg2+, Ag+, AP+, Cd2+, Ba2+, Cr3+, Ti", Ca2+, Zn2+, Pb2+, Cu2+, As02-, NO3-. S042-, C1-, Na+, K+ , Co2+ 200: 1* 150: 1 W042-, Fe3+, Mn2+, Sn2+ * Maximum ratio tested. M0042- 100: 1 Fig. 2 reflects the influence of the PGR concentration on the reaction kinetics. As can be seen, the optimization criterion established above was met over a broad PGR concentration range; hence a concentration of 1.25 x 10-3 rnol dm-3 was chosen for subsequent experiments.Fig. 3 shows the effect of the Br03- concentration on the process. As can be seen, the reaction rate increases virtually linearly with the Br03- concentration. Because of the solubility of NaBr03 and the advisability of using injected volumes no larger than 1 ml, it was decided to use the maximum possible Br03- concentration, viz. , 0.028moldm-3, in order to achieve the highest possible scnsitivi ty . Calibration Graph A calibration graph was constructed under the experimental conditions described above. The graph obtained was linear (five data points) over the concentration range 10- 150 ng ml-1, with a negligible intercept. The corresponding equation was: tan a = 0.0255 + 1 2 . 1 1 ~ (correlation coefficient = 0.9997) (c is the concentration of Vv in ngml-1) and the relative standard deviation for 100 ng ml-1 of Vv (n = 10) was 6%.Study of Interferences The tolerance of the proposed method to potential interfer- ents, taken as the concentration of foreign ion resulting in tan a values differing by +twice the standard deviation (SD) with respect to the values obtained in their absence, is illustrated in Table 1. As can be seen, the proposed kinetic thermometric method is very tolerant to most ions. The low tolerance of the spectrophotometric method to ions such as FeIr1, TirV, MoV1 and Wvl arises from the formation of coloured complexes with PGR or the reaction products involved in the process. The tolerance of the thermometric method is much higher as it is not subject to interference from coloured substances.Determination of Vv in Steel The applicability of the proposed method was tested on a steel sample (High-speed Steel 64b, Bureau of Analysed Samples) with the following composition: C, 0.90; Cr, 4.55; V, 1.99; Mo, 4.95; and W, 7.05%. The Vv content obtained (2.05 k 0.1%; average of two replicate determinations on two different samples) is consistent with the certified value. The authors gratefully acknowledge financial support from the Direccion General Interministerial de Ciencia y Techno- logia (DGICyT) (Grant PS89-0146). References 1 Fishman. M. J.. and Slougstad, M. W., Anal. Chem., 1964,36, 1643.ANALYST, NOVEMBER 1991, VOL. 116 1173 2 3 4 5 6 7 8 9 10 Costache, D., An. Univ. Bucuresti, Chim., 1972, 21, 145. Fuller, C. W., and Ottaway, J. M., Analyst, 1970, 95, 41. Kreingol’d, S. U., Panteleimonova, A. A., and Poponava, R. V., Zh. Anal. Khim., 1973, 28, 2179. Il’cheva, I . A., Degtereva, I. F., Dolmanova, I. F., and Petrukhina, L. A., Metody Anal. Kontrolya Kach. Prod. Khim. Zh. (Russ. Ed)., 1978,4, 65. Lisetskaya, G. S., and Bakal, G. F., Ukr. Khim. Zh. (Russ. Ed.), 1970, 36, 709. Motoharu, T., and Norio, A., Anal. Chim. Acta, 1967,39,485. Yamane, T., and Fukasawa, T., Bunseki Kagaku, 1976,25,454. Jarabin, Z . , and Szarvas, P., Acta Univ. Debrecen. Ludovico Kossuth Nominatae, 1961, 7 , 131. Weiguo, Q., Anal. Chem., 1983, 55,2043. 11 Sevillano-Cabeza, S . , Medina-Escriche, J., and Bosch-Reig, F., Analyst, 1984, 109, 1559. 12 Lumbiarres, J . , Mongay, C., and Cerda, V., J. Thermal Anal., 1981, 22,275. 13 Forteza, R., and CerdA, V., Anal. Chem., 1986, 58,453. 14 Forteza, R., Estela, J. M., and Cerda, V., Analyst, 1990, 115, 749. 15 Gomez, E., Estela, J. M., and CerdA, V., Thermochim. Acta, 1990, 165,255. Paper I I021 84 B Received May 9th, 1991 Accepted June 27th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601171

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1175 Differential-pulse Polarographic Investigation of L- and DL-5-Hydroxytryptophan in the Presence of Copper(!!) Ions V. Kapetanovic and Lj. Milovanovic Department of Analytical Chemistry, Faculty of Pharmacy, University of Belgrade, Yugoslavia The complexes of L- and DL-5-hydroxytryptophan have been investigated by differential-pulse polarography, in Britton-Robinson buffer in the pH range 5.0-10.3, ionic strength I = 0.2 mol dm-3, at room temperature. It has been established that the formation of the ML+ and ML2 complexes occurs. By applying Lingane's method, the over-all stability constant logp2 was 15.91 at pH 9.30. The stability constant of the ML+ complex, logp,, was determined to be 8.30 at pH 5.0 by Leden's method. No differences were found in stability constant values between L- and DL-5-hydroxytryptophan-Cull complexes a t pH values above 9.0.Some differences were obtained in the qualitative results at pH values below 9.0. Keywords: Polarograp h y; L- and DL -forms of 5-h ydroxytryptop han-copperl: 11) complexes; stability constant The aromatic amino acid L-tryptophan can be metabolized via the so-called 5-hydroxyindole route.' By this route, trypto- phan is first hydroxylated to 5-hydroxytryptophan (5-HTP) by the enzyme tryptophan hydroxylase. Coordinative interac- tions have been shown to play a role in serotonin metabolism,2 especially in the decarboxylation of 5-HTP. There is some evidence for the influence of metal ions on the action of monoamino oxidase .3 The significance of the decarboxylation of 5-HTP is the formation of the very important neurotrans- mitter, serotonin, i .e . , 5-hydroxytryptamine. Therefore, an investigation of the coordinative ability of the substances taking part in these metabolic pathways has been undertaken. The stability constants of the bis-L-tryptophanato complexes of Cd", Cu", Nil1 and Zn" have been reported by Albert.4 The experimental technique used enabled only the over-all stabil- ity constant p2 for the 2 : 1 complex to be measured. Weber and Simeons reported the equilibrium constants for the stepwise liberation of protons from D-, L- and DL-tryptophan, and the stepwise interactions of their ligand anions with Cd", Cu", Nil1, Pb" and Zn", in aqueous solution, at 20 "C and ionic strength I = 0.37 mol dm-3 (maintained using NaN03).No differences have been found either between the proton dissociation constants of the stereomeric forms of the ligands or between the stability constants of the respective complex species (MD and ML, or MD2, ML2 and MLD). The stability constants for the mono and bis complexes of Cu" and tryptophan were found to be 8.3O(logp1) and 15.51 (logpz), respectively. The interactions of 5-HTP with some divalent metal ions, Nil1, Cull, Cd" and Pb", have been studied potentiometrically by Weber and Simeon.6 The proton dissociation constants of the ligands and the stepwise stability constants of the simple and protonated complexes have been determined at 20 "C and I = 0.37 mol dm-3. These workers evaluated the complexes M(HL) and M(HL)2, with stability constants, logpl = 8.61 and logp2 = 15.76, respectively, showing the same stabilities as Cu"-tryptophan complexes.In a previous paper,' the Ni11-5-HTP complex was investigated by differential-pulse polarography (DPP). No literature data have been found on the polarographic investigation of 5-HTP 'complexes. In this paper the affinity data for the interactions of L- and DL-S-HTP with the CU" ion based on a polarographic method are reported, with the primary aim of establishing whether there is any difference between these forms. Experimental Apparatus A PO-2 polarographic analyser (Laboratorni Pristoje, Prague) and PHM-62 standard pH meter (Radiometer, Copenhagen) equipped with a combined glass-calomel electrode (Radio- meter GK2401B) were used. Polarographic measurements were performed with a three-electrode system [dropping mercury electrode (DME), standard calomel electrode and Pt electrode] by direct current (d.c.) polarography and DPP.Differential-pulse polarography was performed under the following conditions: forced drop time, 1 s; modulation amplitude, 25 mV; scan rate, 5 mV s-1 and mercury column height, 60 cm. Reagents The reagents used were Cu(N03)2, NaN03 and L-~-HTP (Merck) , and DL-S-HTP (Sigma). The pH was adjusted with NaOH or HN03 when the experiments were performed in an aqueous medium. A constant value of I was maintained with NaN03. For experiments in a buffered medium, Britton-Robinson buffer was used in the pH range 5-10. The 5-HTP solutions were prepared by exact weighing of the standard substance and were stable for 8 d.Procedure Appropriate amounts of 5-HTP and Cull stock solutions were transferred into 10 ml calibrated flasks. The ionic strength was adjusted to a final concentration of 0.2moldm-3. The solutions were de-aerated by passing a stream of nitrogen before recording the polarograms; polarograms were re- corded for 10min in a nitrogen atmosphere. The pH of the solutions was controlled throughout all the experiments. Results and Discussion Polarographic Investigation of L-~HTP in the Presence of Cu" Ion The waves of L-~-HTP (anodic), Cu" and the 5-HTP-Cu" mixture, obtained under the same experimental conditions of pH 9.4, I = 0.2 mol dm-3, are shown in Fig. 1. It can be seen that no cathodic activity is observed for either 5-HTP or for CuII, the latter being precipitated at the pH used.The appearance of a cathodic peak A, or wave B, at a potential of -0.25V is due to the reduction of the polaro- graphically active complex formed between 5-HTP and Cu". The wave height at E = -0.25 V is a linear function of the square root of the mercury column height, with a temperature coefficient (0) of 2% per "C, indicating that the limiting current can be diffusion controlled. By logarithmic analysis of the d.c. polarograms at different pH values, an average value for the electrochemical transfer coefficient (a,) of 1.8 was1176 ANALYST, NOVEMBER 1991, VOL. 116 t 4- E 3 0 A 0.4 pA (DPP 0.04 pA(d.c. I D I 1 1 1 1 1 1 I +0.1 0 -0.1 -0.2 -0.3-0.4 -0.5-0.6 PotentiaiN Fig. 1 Polarograms of: A and B, a mixture of 1 x 10-3mol1-1 5-HTP and 2 X molI-l5-HTP; and D, 2 X moll-' Cu" all under the same experimental conditions, pH 9.4, I = 0.2 rnol dm-3 (NaN03) moll-' Cu"; C, 1 X obtained, indicating that the electrode process is slightly irreversible.pH Dependence of the L-S-HTP-CUII Complex The dependence of the diffusion current of the 5-HTP-Cu1[ complex on pH and the corresponding distribution diagram of 5-HTP ionic species as a function of pH are shown in Fig. 2. The Cu" ion concentration was constant (1 x 10-4 moll-1) and the 5-HTP concentrations used were 1 x 10-4 moll-' [curve 1, Fig. 2(a)] and 1 X 10-3 moll-' [curve 2, Fig. 2(a)]. Curve 3 represents the dependence of the peak height on pH for the free Cu" ion. Up to pH 8, the peaks of the complexed and free Cu" ions are not well resolved.At higher pH, >8.5, the diffusion current of the complex increases, owing to increasing 5-HTP-anion concentration; according to the distribution diagram, Fig. 2(b), the diffusion current is parti- cularly pronounced when the concentration of 5-HTP is ten times higher than that of Cu". The shape of the graph of the wave height versus pH can be explained as being the result of a decrease in Cu" ion concentration due to increasing formation of the complex and the effect of the diffusion coefficient, which is lower for the complex than for the free Cu" ion. The dependence of the peak potential (Ep) on pH gives two parallel lines with almost the same slope (60 and 62mV), indicating the same type of electrode process involving two hydrogen ions.Composition and Stability Constants The metal to ligand ratio in the complex was determined by amperometric titration of a 1 X 10-3 moll-' 5-HTP solution with Cu" ions in Britton-Robinson buffer at pH 9.0 (curve 1, Fig. 3) and pH 9.4 (curve 2, Fig. 3). It can be seen that the diffusion current of the complex increases up to a CuIL concentration of 5 x moll-1 and then decreases sharply. Cross section points, obtained by extrapolating the initial and final points, indicate a 2 : 1 ligand to metal ratio. 1 40 120 100 80 .h- 60 40 20 4 6 8 1 0 8 E ; 80 2 60 ! E 40 8 5 100 '0 20 0 2 4 6 8 1012 PH Fig. 2 (a) Dependence of the diffusion current peak of the complex (curves 1 and 2) and Cu" free ion peak (curve 3) on pH; and (b) corresponding concentration distribution diagram of 5-HTP 240 I 160 E E 3 80 2 4 6 8 1 0 CuWO-4 mol 1-1 Fig.3 Amperometric titration curve of 1 x rnol I-' 5-HTP with Cu" in Britton-Robinson buffer: 1, pH = 9.0; and 2, pH = 9.4, 1 = 0.2 mol dm-3 (NaN03) The stoichiometric ratio of 5-HTP to Cu" in the complex and the stability constant were determined by Lingane's method,8 which according to Tanaka and Tamamushis can also be used for slightly irreversible reduction. Polarograms were recorded for solutions with constant Cu" concentration (2 X 10-4 moll-1) and various concentrations of 5-HTP (5 x 10-5-5 x lO-3mol1-1), at a constant pH of 9.0 and I = 0.2 mol dm-3 (NaN03). When the concentration of 5-HTP was higher than 5 x 10-4 moll-1, a negative shift of the peak due to the complex was observed. By plotting the dependence of the peak potential difference between the complexed and free Cu" ions, AEp, against the logarithm of the 5-HTP concentration, a coordination number of 2 and a relative stability constant, logp2' , of 15.10 were calculated from the slope and intercept of the straight line (Y = 0.9970).ANALYST, NOVEMBER 1991, VOL.116 1177 As the 5-HTP anion, formed by deprotonation of the zwitterion is participating in complex formation, a stability constant (logp2) of 15.88 for the complex was calculated from the relative stability constant value (logo2'), according to the following equation: 10gp2 = log p2' - log kd2 -k log CH+ (1) where kd2 is the dissociation constant of the 5-HTP amino group (kd2 = 2.3 X lo-'*) and CH+ is the hydrogen ion concentration, calculated from the activity coefficient at I = 0.2 rnol dm-3.Titration of Cu" ion (1 x 10-4moll-1) with 5-HTP, in the concentration range 5 X 10-5-3 x 10-3 mol I-*, at pH9.30 was also performed and the results obtained are presented in Table 1. Leden's methodlo was also applied in order to determine the stability constant of the 5-HTP-Cu" complex, Ampero- metric titrations of Cu" ion (2 X 10-4molI-1) with 5-HTP, ranging from 5 X 10-5 to 8 X moll-1, were performed at pH = 5, I = 0.2 rnol dm-3 (NaN03). A lower pH value was used with the aim of following the free Cu" peak height, i.e., of calculating the free metal ion concentration. Equilibrium concentrations of the metal, M, and the complex, MXi, were determined from the height of the free Cu" peak, while equilibrium concentrations of the free ligand, [5-HTP], were calculated from its analytical concentration and the concentra- tion of the complex.According to Leden's equation: F[S-HTP] = p1 + P2[5-HTP] (2) By applying the method of least squares or by plotting the dependence of Leden's function F versus [5-HTP], a straight line was obtained, the slope of which is p2' = 5.58 x 106 and the intercept PI' = 3.14 x 103. These findings indicate the formation of two different complexes where the ratios of 5-HTP to CuL1 are 1 : 1 and 2 : 1, respectively. These values were also corrected according to eqn. (l), and the results obtained by Leden's method are presented in Table 1. An attempt was made to use the dependence of AEp of the complex on pH for the evaluation of the stability constants. The effect of pH (ranging from 5.5 to 9.45) on the peak complex potential was investigated by keeping the concentra- tions of Cull and 5-HTP constant at 1 x 10-4 and 1 x 10-3 moll-], respectively.By changing the pH of the solution, the concentration of the 5-HTP complexing anion was changed. The shift of the peak potential caused by the corresponding pH change was found to be from -0.115 to -0.33V. The concentration of the 5-HTP- anion was calculated from its analytical concentration for each pH value as follows: log C5-HTP- = pH -k log C5-HTP - pkd, (3) Table 1 Stability constants of L- and DL-~-HTP-CU" complexes obtained by different methods ( I = 0.2 mol dm-', Britton-Robinson buffer, room temperature): Method Leden's Lingane's From pH PH 5.00 5.57 6.07 9.00 9.30 5.50-9.46 data - - - - 3.51 - 8.30 - - - - - LogP1'* Log@ 1 * hzB2'* 6.77 9.13 9.63 15.10 15.47 - LogB2 * 11.56 13.35 13.36 15.88 15.95 15.50 - - - 4.04 - - - - - 8.22 - - LogP1't - 7.30 - - 15.38 - LogP I t LogP2 ' t LogPat - 11.48 - - 15.92 - * Values for L-~-HTP-CU" complexes.t Values for DL-~-HTP-CU" complexes. From the pH values the corresponding hydrogen ion concen- trations were calculated, taking the activity coefficients into consideration. By plotting logc5-HTP- versus AEp, a straight line with a slope of 64mV was obtained, from which the number of ligands in the complex was found to be 1.93. From the intercept of the straight line the stability constant log p2 = 15.5 was calculated. As the calculations were performed for a wide pH interval (5.5-9.45), the degree of the reversibility was also checked for each pH value.The average a, value was found to be 1.8. Taking this into consideration, the calculation was performed by applying the equation of Matsuda and Ayabe.11 The values for stability constants obtained at lower pH values are given in Table 1. From the distribution diagram of the complexes formed in the 5-HTP-Cu" system as a function of the logarithm of the free anion concentration (Fig. 4), it follows that the predomi- nant complex species is C u b , when the ligand concentration is higher than 1 x 10-5 mol 1-1. Taking into account the distribution diagram of the 5-HTP itself [Fig. 2(b)], it can be concluded that the concentration of the 5-HTP anion, responsible for the complex formation, is lower at lower pH.For this reason the values obtained for stability constants under these conditions are as expected. Thus, at a pH of 6, the limiting values for the ligand concentration range from 7.41 x 10-4 to 2 x 10-3 moll-1, and the corresponding anion concentrations range form 1.97 x 10-7 to 5.26 x 10-7 moi 1-1, i.e., 7040% of the C u b complex and 10-20% of the CuL+ complex are formed according to the distribution shown in Fig. 4. At pH 9 and 9.30 the formation of a complex of the type CuL2 is quantitative (100%). The impossibility of determining the stability constant of the CuL+ complex by Lingane's method can easily be explained as being owing to the concentration distribution of the complexes formed in the 5-HTP-Cu11 system. Therefore, the stability constant of the CuL+ complex could be determined only by Leden's method.Lingane's method made possible the determination of the stability constant of the C u b complex with sufficient accuracy at pH values above 9. At the same time, on the basis of the data from the pH dependence, reasonable values were obtained for the over-all stability constant, taking into account the wide pH range and the changeable degree of reversibility. This is in good agreement with data reported for the Cull-glycinate system,l2 where only the C u b complex was polarographically detected, although the potentiometric and spectrophotometric measurements also indicated the existence of the CuL+ complex. I 9 8 7 6 5 -Log([5HTP]/mol I-') Fig. 4 Concentration distribution of the complexes formed in the 5-HTP-Cu" system as a function of the logarithm of the 5-HTP concentration for: 1, Cu"; 2, Cu(5-HTP); and 3, CU(S-HTP)~1178 ANALYST, NOVEMBER 1991, VOL.116 0.08 pLA 1.1 v 4 12 A 1 +O.l JkNdIJIL v +0.1 v +0.1 v ov 0 v +0.1 v Potential - Fig. 5 DPP curves obtained in the L-~-HTP-CU" system (1-6) and DL-~-HTP-CU" (7-12). Concentration of Cu" = 1 x moll-' and L- and DL-~-HTP = 3 x mol I-l, Britton-Robinson buffer, I = 0.2 mol dm-3 (NaN03). pH: 1,5; 2,6; 3,7; 4,8; 5,9; 6,9.75; 7,5; 8,6; 9, 6.7; 10, 7; 11, 8; and 12, 9 Polarographic Investigation of DLJ-HTP in the Presence of Cu" Ion An extensive study of the DL-~-HTP-CU'I complexes was carried out in a similar way as for the L-~-HTP-CU" complexes. The results obtained show that there is some evidence for the qualitative difference in the polarographic reduction of L- and DL-~-HTP in the presence of Cull ion, at a pH below 9, when an excess of ligand is present.The polarographic reduction of L-~-HTP proceeds with the appearance of only one peak, in a wide pH range (5.0-9.73, while the reduction of DL-~-HTP is followed by the appear- ance of two peaks in the same pH range. The concentration of the ligands was 3 X 10-3 moll-' and the concentration of the Cull ion was 1 x moll-' (Fig. 5). It has been shown above that the appearance of two peaks cannot be explained as a consequence of the successive formation of 1 : 1 and 1 : 2 complexes and their reduction at a DME. The appearance of two peaks at a pH below 9 for the DL-~-HTP-CU" complex required explanation; therefore, an investigation of their origin was undertaken.The polarograms of L-~-HTP-CUII and DL-~-HTP-CU" were recorded separ- ately, at pH values in the range 5.05-9.60 [Fig. 6(a) and (6)]. The concentration of the Cu" ion was 2 x 10-4 moll-1 and L- and DL-~-HTP concentrations were 1 X 10-3mo11-1. A mixture of 2 x 10-4 moll-1 Cu'I and equal concentrations of L- and DL-~-HTP was also prepared and the corresponding polarograms are presented in Fig.G(c). It can be seen that the peak potential of the first peak in the DL-~-HTP-CU'' system corresponds to the peak potential responsible for the L-5- HTP-Cu" complex. The heights of these peaks can also be compared; at a pH of 6.45, the peak height of DL-~-HTP-CUII is equal to half the peak height of the ~-5-HTP-cu'l complex.The appearance of three peaks can also be observed in the pH range 7-8.5 [Fig. 6 ( b ) ] . This can be explained as a consequence of the redistribution of Cull between different complex species: ML2, MD2 and MDL. This hypothesis is supported by the fact that the sum of these three peaks is almost the same as the peak height of the ~-5-HTP-cu'l complex at pH7, where the equilibrium concentration of these complexes is attained. la hLl kO.1 v ov -0.1 v "1 5*06 9.60 c, 15.05 Potential - Fig. 6 DPP curves obtained in: (a) L-~-HTP-DL-~-HTP-CU" mix- ture; (b) DL-~-HTP-CU"; and (c) L-~HTP-CU" system. Concentra- tion of Cu1I = 2 X 10-4mo11-1 and L- and DL-~-HTP = 1 X mol 1-l, Britton-Robinson buffer, I = 0.2 mol dm-3 (NaN03). The numbers on the peaks are pH values Further, the first peak height of the mixture [Fig.6(a)] is greater than the peak height of ~-5-HTP-cu~l alone, indicating the probable contribution of the corresponding L form, from DL-~-HTP. This contribution is limited by the maximum complex concentration, which has been achieved previously under the appropriate experimental conditions described above (large excess of ligand). Amperometric titrations of 2 X 10-4 moll-' Cull with L- and DL-~-HTP were also performed at pH 5.70 in order to obtain quantitative information. When DL-~-HTP was added to the system containing Cul' ions, the presence of two peaks was evident over the entire concentration range (1 x 10-4-2.5 x 10-3 moll-1). For L-~-HTP, only one peak appeared in the same concentration range. From the data of previous ampero- metric titrations, the stability constants were calculated, resulting in a greater stability of the L-~-HTP-CUII complex for 0.43 log units (logfi2' for L-~-HTP-CU" = 9.20 and 10gp2' for The results obtained can be considered from the viewpoint of stereoselectivity in the formation of complexes of optically active amino acids.13J4 According to Barnes and Pettit,14 the racemic complex (MDL) differs in stability compared with the optically pure species (ML and MD).These workers found significant enthalpy changes between optically pure and racemic species. The enthalpy changes accompanying the formation of Cull-histidine complexes resulted in a greater stability of the optically pure species (ML2 and MD2) compared with the racemic species. Marked stereoselectivity was also obtained by Brookes and Pettit,13 when the N atom in a side chain was protonated in mixed complexes of amino acids and dipeptides.Considering the results obtained in the pH range studied, where, according to the distribution diagram, the protonated species is present, the observed effects can be ascribed to the different behaviour of optically pure and racemic species observed in the presence of the Cu'I ion. DL-5-HTP-Cu" = 8.77).ANALYST, NOVEMBER 1991, VOL. 116 1179 Attempts by Morris and Martin15 to detect stereoselectivity in Cull-histidine complexes by potentiometric titration was unsuccessful. Weber and Simeons also failed to observe any stereoselectivity effect by applying the same technique. The results presented show the significant difference in the polarographic reduction of L- and DL-~-HTP in the presence of Cull ion, leading to the conclusion that the technique described in this paper enabled these effects to be observed. References 1 Udenfriend, S., Titus, E., Weissbach, H., and Peterson, R. E., J. Biol. Chem., 1956,219, 335. 2 Garattini, S., and Valzelli, L., Serotonin, Elsevier, Amsterdam, 1965, p. 31. 3 Garattini, S., and Valzelli, L., Serotonin, Elsevier, Amsterdam, 1965, p. 40. 4 Albert, A., Biochem. J . , 1950,47,531. 5 Weber, 0. A., and Simeon, V., Biochim. Biophys. Acta, 1971, 244,94. 6 7 8 9 10 11 12 13 14 15 Weber, 0. A., and Simeon, V., J. Inorg. Nucl. Chem., 1971,33, 2097. KapetanoviC, V., and MilovanoviC, Lj., Electroanalysis, 1989, 1,461. Lingane, J. J., Chem. Rev., 1941,29, 1. Tanaka, R., and Tamamushi, R., Bull. Chem. SOC. Jpn., 1949, 22, 227. Leden, I., 2. Phys. Chem., 1941, lWA, 160. Matsuda, H., and Ayabe, Y., 2. Elektrochem., 1959,63,1164. Keefer, R . M., J. Am. Chem. SOC., 1946,68, 1946. Brookes, G., and Pettit, L. D., Proceedings of the Sixteenth International Conference on Coordination Chemistry, Dublin, 1974, 1.32. Barnes, D. S., and Pettit, L. D., J . Inorg. Nucl. Chem., 1971, 33,2177. Morris, P. J., and Martin, R . B., J . Inorg. Nucl. Chem., 1970, 32, 2891. Paper 1 I01 599 K Received April 5th, 1991 Accepted July 19th, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601175

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYST, NOVEMBER 1991, VOL. 116 1181 Titration of Proteins in Dimethyl Sulphoxide-Water Mixtures Oswaldo E. S. Godinho, Ivo M. Raimundo, Jr., Luis M. Aleixo and Graciliano Oliveira Net0 lnstituto de Quimica, Universidade Estadual de Campinas, CP 6154, CEP 13081, Campinas, Sao Paulo, Brazil The acid-base behaviour of arginine, lysozyme and ovalbumin has been studied by potentiometric and catalytic thermometric titrimetry in a mixture of dimethyl sulphoxide-water, with acrylonitrile as the end-point indicator in the latter technique. It was observed that, with the exception of the SH groups, all the protonated groups, including the guanidine groups of lysozyme and ovalbumin, were titrated by catalytic thermometric titrimetry. By using potentiometric titrimetry, all the titratable groups of ovalbumin were determined, whereas the guanidine groups of lysozyme were not determined by this technique.Keywords: Protein; catalytic titrimetry; thermometric titrimetry; potentiometric titration; dimeth yl sulp hoxide An approximately equimolar mixture (80% m/m) of dimethyl sulphoxide (DMSO) and water has frequently been employed in potentiometric titrimetry (PT). 1 This mixture is convenient because it is not very hygroscopic, dissolves many organic and inorganic compounds and has a relative permittivity (78.5) close to that of water, but has very different acid-base properties from water. The use of a DMSO-water-acrylonitrile (AN) mixture for the titration of a number of strong and weak acids, including some amino acids, by catalytic thermometric titrimetry (CTT) has been reported.2 In this instance the rise in temperature caused by a cyanoethylation reaction and by the poly- merization of AN catalysed by the excess of base is used to indicate the end-point of the titration. The influence of the solvent composition on the base-catalysed reactions of AN in the titration of benzoic acid in mixtures of some dipolar aprotic solvents, water and AN has also been reported.3 In preliminary experiments involving the titration of acid groups of proteins in a DMSO-water-AN mixture, the influence of solvent composition has also been investigated.4 By consider- ing the shape of the titration curve and the solubility of proteins, it was decided to adopt a solvent composition of DMSO-water-AN of 7.0 + 0.8 + 4.0 ml for these titrations.It was observed that in addition to the carboxyl and a-amino groups of amino acids, the second carboxyl groups of glutamic and aspartic acids, the &-amino group of lysine, the phenolic group of tyrosine and the imidazole group of histidine were titrated under the above conditions. However, the SH group of cysteine cannot be determined by this technique. This has been attributed to the cyanoethylation of this group, in agreement with Greenhow and LOO,^ who reported the cyanoethylation of SH groups of thiols during an investigation of the determination of these compounds. In this work, the behaviour of the amino acid arginine and the proteins lysozyme and ovalbumin has been investigated by PT in a DMSO-water mixture (7.0 + 0.8 ml) and by CTT in the solvent mixture DMSO-water-AN (7.0 + 0.8 + 4.0 ml).Experimental Reagents Dimethyl sulphoxide was of laboratory-reagent grade and potassium hydroxide and propan-2-01 were of analytical- reagent grade. The solvents were dried over a molecular sieve 3A before use. The proteins employed were lysozyme, egg-white, from Aldrich (Cat. No. 85387-9, Lot No. 121567) and ovalbumin, from Sigma (Cat. No. A-5503, Lot No. 81F-8235). Potassium hydroxide solution, 0.1 mol dm-3 in propan-2-01, was prepared and standardized against benzoic acid in ethanolic solution by employing phenolphthalein as indicator. Apparatus The apparatus employed to introduce the titrant at a constant delivery rate6 and to detect the temperature change has been described elsewhere .7 In potentiometric titrations a glass electrode and a nichrome wire sealed in the syringe and immersed in the titrant were employed as indicator and reference electrode, respectively.The glass electrode was conditioned in the DMSO-water solvent mixture (7.0 + 0.8 vlv) for about 48 h before use. A derivative circuit constructed in this laboratory was used to obtain the first-derivative titration curves. Procedure For potentiometric titrations of protonated groups originally present in proteins and arginine, 20.00 mg of protein or 10.00 mg of arginine were weighed into a 50 ml beaker, dissolved in 0.8 ml of water and then 7.0 ml of DMSO were added. In the potentiometric titrations of the mixture of fully protonated proteins or arginine plus a strong acid, the same amounts of protein and arginine were dissolved in 0.8 ml of 0.1 mol dm-3 HCI instead of water.For thermometric titrations the same amount of protein or arginine was weighed into a 25 ml unsilvered Dewar flask, dissolved in 0.8 ml of water and then 7.0 ml of DMSO and 4.0 ml of AN were added. In both the potentiometric and thermometric techniques the titrant was added at a constant delivery rate of 0.40 ml min-1. During the titration the solutions were stirred with a magnetic stirrer. The end-point of the thermometric titrations was taken as the point where the titration curve left the horizontal. Table 1 Results of the titration of protonated groups originally present in arginine, lysozyme and ovalbumin. Relative molecular masses of lysozyme and ovalbumin used in the calculations were 14307 and 42 699, respectively No.of No. of groups groups found* found by No. of Compound PT CTT technique? groups another guanidine Arginine 2.01 k 0.01 1.98 k 0.03 1 1 Lysozyme 11.4 k0.5 23.6 kO.7 10.3 lli: Ovalbumin 52.3 k 1.5 47.6 k0.6 39.1 1% * f Standard deviation. t Titration of denatured protein in 8.0 mol dm-3 aqueous urea $ Reference 9. D Reference 10. solution (reference 8).1182 ANALYST, NOVEMBER 1991, VOL. 116 Results and Discussion Results from the titration of arginine (Table 1) indicate that although the guanidine group is a weak acid (pK, = 12.5 in water) it can be titrated by both CTT and PT in the medium employed. This observation is expected as DMSO is a protophilic dipolar aprotic solvent and hence it favours the titration of very weak acids."' Thermometric titration curves of arginine, lysozyme and ovalbumin are shown in Fig.1. The results of the titration of protonated groups originally present in lysozyme and oval- bumin by both PT and CTT are shown in Table 1. The results 0 0.5 1 .o Titrant volume/ml Fig. 1 Catalytic thermometric titration of protonated groups origi- nally present in arginine hydrochloride and in some proteins in a solvent mixture of DMSO-water-AN (7.0 + 0.8 + 4.0 ml): A, blank sample; B, 20.70 mg of ovalbumin; C, 21.11 mg of lysozyme; and D, 12.08 mg of arginine hydrochloride. Titrant, 0.0995 mol dm-3 KOH in propan-2-01 1 I 1 I 0 0.5 1 .o 1.5 2.0 Titrant volume/ml Fig. 2 First derivative of the potentiometric titration curve of arginine plus excess of a strong acid: 9.11 mg of arginine hydro- chloride in 0.809 ml of 0.1018 mol dm-' HCI, followed by the addition of 7.0 ml of DMSO.Titrant, 0.0995 mol dm-3 KOH in propan-2-01 are compared with those obtained by the potentiometric titration of these proteins denatured in urea8 when the guanidine groups are not determined. For lysozymeg it can be seen that although the guanidine groups are determined by CTT they are not determined by PT. The observation that the guanidine groups of lysozyme are not determined by PT might be explained by considering that the protein is not completely denatured in this solvent mixture and hence these groups are not accessible for titration. The cysteine groups of lysozyme are not determined because they form interlinking disulphide bridges.For ovalbumin the guanidine groups are determined by both techniques. However, the number of protonated groups obtained by CTT is about four less than that obtained by PT. This discrepancy is ascribed to the cyanoethylation of the SH groups of cysteine residues during the catalytic titration by analogy with the results obtained for the amino acid ~ysteine.~ This explanation is consistent with the fact that four cysteine residues are present per mole of ovalbumin. Potentiometric Titration of Fully Protonated Arginine and Proteins The first derivative of the potentiometric titration curve of arginine plus a strong acid is shown in Fig. 2 and the results are presented in Table 2. Four peaks corresponding to the titration of hydrochloric acid, carboxylic, amino, and guan- idine groups are observed.Similarly, first derivatives of the potentiometric titration curves of total protonated groups of proteins in the presence of a strong acid are shown in Fig. 3 and the results are presented in Table 2. By analogy with the results obtained for arginine and other amino acids,4 the first peak is related to the titration of the strong acid. The second and third peaks are related to the titration of acid and basic groups of proteins, respectively. Groups named as 'acid groups' can be identified as carboxylic groups by comparison with the potentiometric titration curves of arginine and other amino acids.4 On the other hand, groups named as 'basic groups' might include protonated imidazole, amino, phenolic, SH and guanidine groups. In fact the potentiometric titration curves of proteins do not exhibit discrete inflections corresponding to each of these basic groups.The number of acid groups, which is related to the number of carboxylic groups, and the total number of groups are shown in Table 2. For lysozyme, 10-11 carboxylic groups are expected but only 8.7 were determined, although ovalbumin has 47 carboxylic groups only 42.2 were determined. It is assumed that because of electrostatic effects12 the first carboxylic group being titrated has a pK, sufficiently low for it to be titrated as a strong acid. By considering the total number of groups titrated for lysozyme it was deduced that the guanidine groups were not determined in this protein. In fact 22.7 groups were deter- mined and the total number of groups present is 21, if the guanidine groups are excluded.This conclusion is in agree- Table 2 Results obtained by PT for total protonated groups of arginine, lysozyme and ovalbumin No. of Total No. of No. of groups Compound Found* Present Found* Present? present carboxylic groups groups Guanidine Arginine 1.01 f 0.01 1 2.96 f 0.03 2 1 Ly sozy me 8.7 k0.8 10-1 1 22.7 k0.8 21-22$ 11 Ovalbumin 42.2 k2.2 47 97.0 k2.1 88§ 15 * k Standard deviation. t Not including guanidine groups. $ 10-11 Carboxylic, 1 imidazole, 1 or-amino, 6 E-amino and 3 phenolic groups (reference 9). 0 47 Carboxylic, 7 imidazole, 20 E-amino, 10 phenolic and 4 SH groups (reference 10).ANALYST, NOVEMBER 1991, VOL. 116 1183 I I I I I 0 0.5 1 .o 1.5 Titrant volume/ml Fig. 3 First derivative of the potentiometric titration curves of lysozyme and ovalbumin plus excess of a strong acid.Protein dissolved in 0.809 ml of 0.1018 mol dm-3 HCl, followed by the addition of 7.0 ml of DMSO: A, 25.69 mg of lysozyme; and B, 19.09 mg of ovalbumin. Titrant, 0.0995 mol dm-3 KOH in propan-2-01 ment with the results obtained for the titration of the protonated groups originally present in lysozyme (Table 1). On the other hand, for ovalbumin a total of 103 protonated groups are present, if the 15 guanidine and 4 SH groups are included. Ninety-seven groups were found in the titration. This fact might be explained by considering, as mentioned above, that 4-5 carboxylic groups are titrated as if they were a strong acid and hence their effect is added to that of the hydrochloric acid.In this way it can be assumed that for ovalbumin the guanidine and SH groups are titrated. This assumption is in agreement with that based on the potentio- metric titration of the protonated groups originally present in the protein, as shown in Table 1. The results presented here offer some interesting possi- bilities: for example, the determination of cysteine in the presence of other amino acids, or even of cysteine residues in proteins, by comparing the results obtained by PT in DMSO-water with those of CTT in DMSO-water-AN. Similarly, it might be possible to determine the amino acid arginine or arginine residues in proteins by comparing the results obtained by PT in DMSO-water or CTT in DMSO- water-AN with the results of a titration in water where the guanidine groups are not determined. However, these methods needs to be applied to other proteins or to a mixture of amino acids resulting from the hydrolysis of proteins in order to establish their applicability. 1 2 3 4 5 6 7 8 9 10 11 12 References Georgieva, M., Velinov, G., and Budevsky, O., Anal. Chim. Acta, 1977,90, 83. Greenhow, E. J., and Shafi, A. A., Talanta, 1976, 23,73. Godinho, 0. E. S., and Greenhow, E. J., Anal. Chem., 1985, 57,1725. Godinho, 0. E. S., and Greenhow, E. J., unpublished results. Greenhow, E. J., and Loo, L. H., Analyst, 1974,99,360. Greenhow, E. J., and Spencer, L. E., Analyst, 1973, 98, 98. Chagas, A. P., Godinho, 0. E. S., and Costa, J. L. M., Talanta, 1977,24,593. Godinho, 0. E. S., and Silva, M. C., unpublished results. Canfield, R. E., J. Biol. Chem., 1963,238,2698. Nisbet, A. D., Saundry, R. H., Moir, A. J. G., Fothergill, L. A., and Fothergill, J. E., Eur. J. Biochem., 1981, 115, 335. Kolthoff, I. M., Anal. Chem., 1974,46, 1992. Linderstrom-Lang, K., C. R. Trav. Lab. Carlsberg, 1924,15 (7), 1. Paper 110071 9J Received February 15th, 1991 Accepted July 3rd, 1991

ISSN:0003-2654    

DOI:10.1039/AN9911601181

出版商:RSC    

年代:1991

数据来源: RSC