摘要:

ANALYST, FEBRUARY 1986, VOL. 111 171 Flow Injection - Hydride Generation System for the Determination of Arsenic by Molecular Emission Cavity Analysis M. Burguera and J. L. Burguera Departamento de Quimica, Facultad de Ciencias, Universidad de Los Andes, Apartado Postal 542, Merida 5101-A, Venezuela A system is described that permits the simple and rapid determination of nanogram amounts of arsenic (0.1-10 pg ml-1) in microlitre volumes of sample solutions by flow injection and hydride generation coupled with molecular emission cavity analysis. The effect of some interfering ions, including Ni, Ag, Zn, Cu, Te and Se, was eliminated or minimised by using a 0.01 M EDTA - 0.2 M sodium iodide - 1.5 M hydrochloric acid carrier solution in which the sample was injected. The method permits a sampling rate of ca.100 measurements per hour. The method is illustrated by the determination of arsenic in NBS orchard leaves reference material with good accuracy. Keywords: Arsenic determination; h ydride generation; flow injection analysis; molecular emission cavity analysis There is much published work on the determination of arsenic in different kinds of matrices.1-4 In addition to photometric methods,5-8 hydride generation in conjunction with atomic absorption spectrometry (AAS) ,941 inductively coupled plasma atomic emission spectrometry (ICP-AES) ,12J3 atomic fluorescence spectrometry14>15 and molecular emission spec- trometryl6 are widely used for the determination of arsenic. Conventional molecular emission cavity analysis (MECA)17 and hydride generation in conjunction with MECAls-21 have also been applied to the determination of nanogram amounts of arsenic.Flow injection analysis (FIA)22 has been applied together with amperometry and voltammetry for the determination of arsenic at pg ml-1 levels.2525 Astrom26 and Liversage and Van Loon27 introduced the hydride generation approach combined with FIA for the determination of bismuth and arsenic, respectively. By using their procedure, nanogram amounts of both elements were determined with good accuracy and up to 200 measurements per hour were possible.27 FIA has also recently been used in combination with MECA for the determination of nanogram amounts of sulphur28 and pho~phorus-containing29~3~ compounds. A combination of FIA and hydride generation with MECA detection has not been reported before.This paper describes the advantages of such a combination. The determination of arsenic is based on the generation of arsine in an FIA system followed by its separation from the liquid phase and transport with argon to the MECA cavity, where the As0 emission intensity is measured at 400 nm. The method has been applied success- fully to the determination of arsenic in NBS orchard leaves standard reference material. previously described.28129 A water-cooled steel MECA cav- ity, which was similar to that described earlier,28 was used. However, in addition a small hole was made in the side-wall of the cavity (Fig. 2) in order to introduce a flow of oxygen. The introduction of oxygen is necessary in order to restrict the emission within the oxy-cavity.17 The gas - liquid separator, which was essentially the same as that described by Vijan et aZ.,31 was miniaturised as far as possible using tubing of 2.0 mm i.d. Reagents All reagents used were of analytical-reagent grade, unless stated otherwise. De-ionised, doubly distilled water was used throughout. Arsenic stock solution, 1000 pg ml-1. Dissolve 0.6602 g of arsenic(II1) oxide in 15 ml of 20% sodium hydroxide solution. The solution was transferred into a-500-ml calibrated flask and HCI NaBH4 Gas to MECA cavity .iquid to waste Fig. 1. Flow injection manifold for the determination of arsenic with hydride generation and MECA detection. Coils a, b and c, 0.5 mm i.d.; S, point of sample injection; GLS, gas - liquid separator.[NaBH,], 0.2 M; [HCI], 1.5 M; other conditions as in Table 1. Experimental Apparatus The configuration of the experimental system used is shown in Fig. 1. The flow injection system consisted of a five-channel peristaltic pump (Sage, Orion Research) and a rotary injec- tion valve (Rheodyne Model 7125), to which a loop of given volume was attached. All tubing was made of Tygon (0.5 mm id.), except the side-tube, which connected the MECA cavity to the FIA system, made of stainless steel (0.5 mm i.d.) as previously described.28 The confluence T-joints and connec- tions were made of Perspex. The instrument for MECA measurements, the cavity holder support device and the circular emission burner were as - Detector and carrier gas Fig. 2. generation system (distances in mm).Conditions as in Table 1 Cross-section of the MECA cavity for FIA - hydride172 ANALYST, FEBRUARY 1986, VOL. 111 8 - > E z . c .- g 6 - .; 4 I C C .- v) w .- E 2 - Table 1. Experimental parameters used in the FIA - hydride generation - MECA system for the determination of arsenic FIA - hydride generation parameters MECA parameters - 0 - Sample volume/yl . . . . . . . . Coil b lengthkm . . . . . . . . Coil c lengthkm . . . . . . . . Solutions pumping rate/ml min- 1 . . Carrier gas flow-rate/ml min- 1 . . . . HCl concentration in carrier solutionh NaBH4 concentration in reductant solutionh . . . . . . . . . . EDTA concentration in carrier solutionh NaI concentration in carrier solution/M . . 120 Wavelengthlnm . . . . . . . . . . 400 Slit widthhm .. . . . . . . . . 1 . . 40 Cavity . . . . . . . . . . . . . . At flame centre . . 1.5 Cavity angle below horizontal . . . . 2.0" . . 1.5 Flame N2/1 min- 4.5 . . 20 0.5 Flame H2/l min-l . . . . . . . . . . 2.5 . . . . . . . . . . . . O2 to cavityhl min- 1 . . . . . . 90 . . 0.2 Watercoolingflow-rateikmin-1 . . . . 10 * . 0.01 . . 0.2 I 9.0 I 1 min , 4.5 1.5 M Time - Fig. 3. Emission - time responses for arsenic determination (num- bers on the peaks indicate the concentration of As in pg ml-I). Conditions as in Table 1 6 > E . .- 5 4 v) C al 4-l .- C .; 2 v) .- E w 0 P I 0.5 1 .o 1.5 [HCI]/M I I I I 0 0.1 0.2 0.3 [NaBH41/~ Fig. 4. Effect of hydrochloric acid and sodium tetrahydroborate(II1) concentrations on the sensitivity taking 5 pg ml-1 of As. Conditions as in Table 1 I 0 0.2 10-2 lo-' [ N a l ] / ~ L+ O l l I I I [ E DTA ]/M Fig.5. Effect of sodium iodide and EDTA concentrations on arsenic emission (5 pg ml-1 of As). The NaI effect was studied in the absence of EDTA and vice versa. Conditions as in Table 1 -' i d - 3 diluted to the mark with 2 M hydrochloric acid. Working solutions were prepared daily by appropriate dilution of the stock solution with 2 M hydrochloric acid. Sodium tetrahydroborate(III) solution, 1 M. Dissolve 3.800 g of powdered laboratory-reagent grade NaBH4 in 100 ml of 0.5 M sodium hydroxide solution. The strong base stabilises the NaBH4 solution.* Preliminary tests and optimisation of the experimental parameters were carried out with a 5 ng 1-11-1 As solution. Procedure The instrumental parameters were adjusted to the optimum values (Table 1).The MECA cavity was situated within the flame, and the pump was turned on for a few minutes to fill the U-tube of the gas - liquid separator. The argon used as the carrier gas was allowed to flow. The sample solution was injected via the rotary valve into the flowing stream of lo-* M EDTA - 0.2 M sodium iodide - 1.5 M hydrochloric acid, which merged with the 0.2 M NaBH4 flowing stream (Fig. 1). After gas - liquid separation, the generated arsine was carried by the argon carrier gas to the MECA cavity and the transient emission signal from the As0 species32 was measured at 400 nm against time and recorded. All emission values were based on peak-height measurements. Results and Discussion Optimisation of Operating Parameters A systematic investigation was necessary in order to establish the optimum conditions so that a high sensitivity could be achieved with good repeatability.Various chemical and instrumental operating parameters were varied individually while the others were kept constant, and were optimised with respect to the emission intensity. Regarding hydride genera- tion and MECA parameters, the highest peaks with least tailing (Fig. 3) were obtained under the conditions listed in Table 1. The effects of various chemical parameters are shown in Figs. 4 and 5. In this study, the dependence of analytical sensitivity on the concentration of HC1 was examined. The analytical sensitivity increased with increasing HC1 concentra- tion from 0.25 to 1.0 M, above which no further increase was observed.The concentration of HCl subsequently used was 1.5 M. It is reported that reduction of arsenate (As5+) or arsenite (As3+) to AsH3 with NaBH4 is affected by the pH of the solution.l6As3+ is reduced with maximum efficiency in the pH range 0-5, whereas Ass+ is reduced at pH 0-1. In the present system, the pH was about 0.1 after the merging of the sample and the NaBH4 flowing streams (Fig. 1). Therefore, under the present experimental conditions, total arsenic, including As3+ and Ass+, is measured. Sodium tetrahydroborate(II1) was chosen as a reductant, as it is powerful enough to reduce trivalent and pentavalent arsenic to AsH3.33 The emission intensity increased withANALYST, FEBRUARY 1986, VOL. 111 > > E . 4- .- z 4 - 4- C C 0 m .- .- .E 2 1 73 - 61 > 5: E 4 t.t .- q w E /” I I I I 0 40 80 120 Sample volumeipl Fig. 6. Effect of sample volume on emission intensity for 0.6 pg ml-1 of As each time. Conditions as in Table 1 2 h h 0 20 40 60 80 Coil lengthicm Fig. 7. for 5 pg ml-* of As. Conditions as in Table 1 Effect of the lengths of coils b and c on the emission intensity increasing NaBH4 concentration up to 0.2 M (Fig. 4), with no further improvement at higher concentration. EDTA and sodium iodide were added to the sample carrier solution in order to minimise interferences as described below. The addition of EDTA in the range 10-3-10-1 M did not have any effect on the emission intensity (Fig. 5), whereas the addition of sodium iodide increased the emission intensity at concentrations up to 0.1 M, above which the intensity was independent of concentration.This increase in the emission intensity could be due to the fact that the concentration of iodide effects a complete reduction of the arsenic ions.27 Concentrations of 10-2 M EDTA and 0.2 M sodium iodide are recommended. The influence of sample volume injected is shown in Fig. 6. The sensitivity increased with increasing sample volume, but peak broadening and tailing occurred above 120 pl. A sample volume of 120 pl was found to be reasonable to ensure smooth arsine generation in the FIA system, and therefore a suitable flow of this species into the MECA cavity. In this way the emission from the arsine was restricted within the oxy-cavity. The effect of different coil lengths was investigated (Fig. 1).In the screening experiments it was established that the length of coil a was unimportant and therefore can be disregarded in the discussion of coil lengths. As illustrated in Fig. 7, the lengths of coils b and c were of great importance. The sensitivity decreased when coil b was longer than 40 cm and increased with increasing length of coil c up to 30 cm. Lengths of coils b and c of 20 and 40 cm, respectively, are recommen- ded. We used equivalent sample carrier and NaBH4 solution flow-rates. It was found that the emission intensity increased with increasing pumping flow-rates of both solutions (Fig. 8). The emission intensity was found to be almost independent of the pumping flow-rate above 1.0 ml min-1. A pumping flow-rate of 1.5 ml min-1 in each channel was chosen as the optimum.The flow-rate of the argon carrier gas was found not to be critical, provided it was held in the range 0.4-0.6 ml min-1 (Fig. 9). However, higher flow-rates were found to decrease the emission intensity. This effect could be attributed to an insufficient gas - liquid separation at higher argon flow-rates, and to excessive amounts of gases swept into the cavity (under I I I I I 0.5 1 .o 1.5 2.0 Flow-rateiml min-1 Fig. 8. Effect of hydrochloric acid and sodium tetrahydroborate(II1) solution pumping rates on the sensitivity for 5 pg ml-1 of As. Conditions as in Table 1 t 0 m .- .- L P 1 I I I I 0.2 0.4 0.6 0.8 Argon flow-rateiml min-’ Fig. 9. As. Conditions as in Table 1 Effect of argon flow-rate on arsenic emission for 5 pg ml-1 of such conditions, the emission of arsenic occurred in the flame above the cavity, therefore decreasing the sensitivity).When a constant argon flow-rate of ca. 0.5 ml min-’ was maintained, good sensitivity and reproducibility were obtained. The MECA parameters, such as position of the cavity, water cooling and flame composition, were also optimised. The cavity was positioned at the centre of the flame, 15 mm above the burner top, and pitched 2” downwards, in line with the detector. The cavity was cooled with water, in order to protect it from incandescence and the support device from over-heating and deformation; the best sensitivity was achieved with a water-cooling rate of 10 ml min-1. It is well known that As0 species are only induced within the cavity by the use of an oxy-cavity.34 Therefore, an oxygen flow-rate to the cavity (Fig.2) of 90 ml min-1 restricted the emission within the cavity space and made the cavity environment hotter, so improving the sensitivity. The intensity of the emission of As0 increases with increasing temperature of the ~avity.3~ Therefore, the flame composition should be carefully optimised. The optimum temperature for promoting As0 emission was achieved with hydrogen and nitrogen flow-rates of 2.5 and 4.5 1 min-1, respectively. This flame composition was also found to give the best signal to noise ratio.35 Calibration Graph, Precision and Detection Limit The calibration graph obtained by injection of 120 yl of arsenic solutions was rectilinear from 0.1 to 10 pg ml-1 of As. The maximum emission intensity increased linearly with arsenic concentration as expressed by the equation EAs = -0.12 + 1.52XAS, r = 0.9997 (seven points), where EAs is the peak height (mV) and XA, the arsenic concentration (pg ml-1).The relative standard deviations for the determination of 0.8 and 8.0 pg ml-1 of As, obtained from eight replicate analyses, were 5.5 and 2.8%, respectively. The detection limit was evaluated by calculating the mean standard deviation of eight sets of replicate blank determinations. The detection limit , considered as twice the signal to blank ratio, was 0.08 pg ml-1 (9.6 ng) of As.174 ANALYST, FEBRUARY 1986, VOL. 111 ~~ ~ Table 2. Effect of foreign ions, EDTA and sodium iodide on arsenic emission intensity. Concentrations: arsenic, 1 pg ml-1; foreign ions, all at 100 pg ml-1; EDTA and NaI in the sample carrier solution, 0.01 and 0.2 M, respectively Change in emission intensity,* % Ion added As alone As + NaI As + EDTA As + NaI + EDTA Ni2+ .. . . -50 -43 -3 -2 Ag+t . . . . -25 -20 -1 -1 Bi3+ . . . . -30 -28 -2 -1 zn2+ . . . . -10 -8 0 0 Te4+ . . . . -6 -5 -5 -4 Se4+ . . . . -7 -1 -7 -1 cu2+ . . . . -15 -1 - 10 -1 Fe3+ . . . . -30 -1 0 0 cr3+ . . * . -2 0 0 0 A13+ . . * . -1 0 0 0 co2+ . . . . -34 -33 -1 0 * Compared with the emission in the absence of interfering ions. i AgCl suspension injected. Table 3. Analysis of NBS orchard leaves standard reference material Arsenic content/pg g-1 Sample No. Certified Measured* 1 1 1 f 2 9.2 k 0.3 2 14+2 12.0 k 0.6 3 10+2 9.5 + 0.4 * Four determinations; 1.0 g of dried orchard leaves sample was taken for each analysis Interferences Comprehensive studies of interferences in the determination of arsenic by hydride generation in several systems have been reported.3638 A variety of likely interfering ions were studied in our system under the recommended optimum conditions.An interference was defined as significant if a change of more than two standard deviations in the measurements was observed. Of the ions studied, alkali and alkaIine earth elements, Al3+ and Cr3+ did not interfere, whereas other ions suppressed the emission intensity. Most of these interferences were effectively eliminated or reduced by the addition of EDTA (Table 2), except for Se4+ and Cu2+. However, the interference effect of these ions was almost eliminated by also including NaI (0.2 M) in the sample carrier solution (Table 2).Determination of Arsenic in Standard Samples In order to evaluate the applicability of the present method to real samples, we determined arsenic in NBS orchard leaves standard reference material. The samples were digested as indicated by Liversage and Van Loon,27 although the final volume was reduced by evaporation to ca. 4-5 ml. The results are in good agreement with the certified values, as shown in Table 3. Conclusion The flow injection - hydride generation system used for the determination of arsenic by MECA is a very simple, sensitive and rapid technique. About 100 measurements per hour can be made. The sensitivity of this method is nearly an order of magnitude less than that reported for atomic absorption spectrometry; however, our method allows the determination of arsenic in real samples, such as plant material, with good accuracy.In addition, the system may also be adaptable for the determination of other hydride-forming species that would generate chemiluminescence emission in a MECA cavity, such as Se, Sb, Sn and Ge, in rocks, sediments and minerals. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. References Haywood, M. G., and Riley, J. P., Anal. Chim. Acta, 1976,85, 219. Fiorino, J. A., Jones, J. W., and Capar, S. G., Anal. Chem., 1976, 48, 120. Pierce; F. D., and Brown, H. R., Anal. Chem., 1976,48,693. Cladney, E. S . , and Owens, J.W., Anal. Chem., 1976, 48, 2220. Johnson, D. L., Environ. Sci. Technol., 1971, 51, 411. Portmann, J. E., and Riley, J. P., Anal. Chim. Acta, 1964,31, 509. Haywood, M. G., and Riley, J. P., Anal. Chim. Acta, 1976,85, 219. Reay, P. F., Anal. Chim. Acta, 1974, 72, 145. Thompson, K. C., and Thomerson, D. R., Analyst, 1974, 99, 595. Maruta, T., Anal. Chim. Acta, 1975, 77, 37. Van Loon, J. C., and Brooker, E. J., Anal. Lett., 1974,7,505. Thompson, M., Pahlavanpour, B., Walton, S. J., and Kirk- bright, G. F., Analyst, 1978, 103, 568. Pahlavanpour, B., Thompson, M., and Thorne, L., Analyst, 1981, 106, 467. Robbins, W. B., and Caruso, J. A., Anal. Chem., 1979, 51, 889A. Ebdon, L., Wilkinson, J. R., and Jackson, K. W., Anal. Chim. Acta, 1982, 136, 191. Matsumoto, K., and Fuwa, K., Anal.Chem., 1982,54,2012. Belcher, R., Bogdanski, S. L., Ghonaim, S . A., and Towns- hend, A., Anal. Chim. Acta, 1974, 72, 183. Belcher, R., Bogdanski, S. L., Ghonaim, S . A., and Towns- hend, A., Anal. Lett., 1974, 7, 133. Belcher, R., Bogdanski, S . L., Henden, E., and Townshend, A., Analyst, 1975, 100, 522. Henden, E., Analyst, 1982, 107, 872. Henden, E., Pourreza, N., and Townshend, A., Prog. Anal. At. Spectrosc., 1979, 2 , 355. Tyson, J . F., Analyst, 1985, 110, 419. Lown, J. A., and Johnson, D. C.,Anal. Chim. Acta, 1980,116, 41. Lown, J. A., Koile, R., and Johnson, D. C., Anal. Chim. Acta, 1980, 116, 33. Fogg, A. G., and Bsebsu, N. K . , Analyst, 1981, 106, 1288. Astrorn, O., Anal. Chem., 1982, 54, 190. Liversage, R. R., and Van Loon, J. C . , Anal. Chim. Acta, 1984, 161, 275. Burguera, J. L., and Burguera, M., Anal. Chim. Acta, 1984, 157, 177. Burguera, J. L., Burguera, M., and Flores, D., Anal. Chim. Acta, 1985, 170, 331. Burguera, M., and Burguera, J. L., Anal. Chim. Acta, 1986, in the press. Vijan, P. N., and Wood, G. R., At. Absorpt. Newsl., 1974,13, 33. Belcher, R., Bogdanski, S. L., Henden, E., and Townshend, A., Anal. Chim. Acta, 1977, 92, 33. Braman, R. S . , Justen, L. L., and Foreback, C. C., Anal. Chem., 1972,44,2195. Burguera, M., Bogdanski, S. L., and Townshed, A., CRC Crit. Rev. Anal. Chem., 1980, 10, 185. Burguera, J. L., Burguera, M., and Townshend, A., Acta Cient. Venez., 1984,35, 165. Haywood, M. G., and Riley, J. P., Anal. Chim. Acta, 1976, 85, 219. Nakahara, T., Anal. Chim. Acta, 1981, 131, 73. Peacock, C. J., and Singh, S. C., Analyst, 1981, 106, 931. Paper A51191 Received May 23rd, 1985 Accepted August 12th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100171

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 175 Determination of Sulphur, Phosphorus, Magnesium, Silicon and Aluminium in Washing Powders by X-Ray Fluorescence - Spectrometry John Williams Interox Research and Development Department, P.O. Box 2, Moorfield Road, Widnes, Cheshire WA8 OJU, UK A method for preparing and analysing samples of washing powders by X-ray fluorescence spectrometry is described. The sample preparation involves the use of an automatic machine for the preparation of fused dilithium tetraborate beads and demonstrates the feasibility of using this technique for difficult samples containing a mixture of organic and inorganic components. The results of the X-ray fluorescence method are compared with those from wet chemical analysis. The determination of total sulphur, phosphorus, silicon, magnesium and aluminium by X-ray fluorescence represents a considerable time saving compared with wet chemical analysis.Keywords: Washing powders; X-ray fluorescence spectrometry; sulphu-; phosphorus, silicon, magnesium and aluminium determination; automatic sample preparation A wide range of compounds are used in the formulation of washing powders; in addition to organic surfactants, optical brighteners and softening agents, a washing powder may contain phosphates, sulphates, silicates, magnesium salts of bleaching agents, etc. The analysis of washing powders for elements such as sulphur, phosphorus, silicon, aluminium and magnesium can be a time-consuming operation and normally involves the use of several different analytical techniques.A widely used method of analysis involves initial fusion with a mixed nitrate - carbonate flux, to destroy organic matter, followed by dissolution of the fused melt and determination of individual elements by a combination of gravimetric, spectro- photometric and physical methods.' X-ray fluorescence spectrometry is a technique that can be used to determine a number of elements quickly on one sample, and usually the sample for analysis is prepared as a fused dilithium tetraborate bead. The initial fusion of the sample therefore is common to both the traditional and X-ray techniques of analysis, but the preparation of a bead followed by X-ray fluorescence analysis should save time and effort compared with traditional techniques. In the preparation of a sample bead for X-ray fluorescence analysis, the direct fusion of the sample with a flux such as dilithium tetraborate is perfectly satisfactory for inorganic materials in a fully oxidised state.However, problems are encountered if the sample contains organic matter or metals in a reduced form. In these instances the sample has to be fully oxidised before a fused bead can be made. Washing powders require such an oxidation step before fusion. This paper describes a method whereby the total concentra- tions of the above elements can be rapidly determined by an X-ray fluorescence technique using a sample in the form of a single fused dilithium tetraborate bead. The manual effort required for sample preparation is kept to a minimum by using an automatic bead maker. Experimental Apparatus and Reagents A Per1 'X-2 automatic bead machine, manufactured by Soled and supplied by Laborlux S.A., Luxembourg, was used for sequential oxidation and fusion of the sample into a dilithium tetraborate bead.This is a programmable radiofrequency furnace unit, which can also agitate the molten sample and pour the melt into a casting mould. The crucible for sample fusion and the casting mould were made from a platinum - gold alloy. The mould produced beads of 25 mm diameter. A Philips PW1400 X-ray fluorescence spectrometer with a Philips P851M minicomputer was used for the fluorescence measurements. The dilithium tetraborate used for the sample fusion was Spectroflux 100 from Johnson Matthey Chemicals Ltd. Other reagents used were of analytical-reagent grade.These operations produce a melt that is sufficiently fluid to give a flat area on top of the bead suitable for X-ray fluorescence measurement.2 The beads must be handled with clean gloves to avoid contamination and to prevent cuts from sharp edges. Calibration A series of standard calibration beads were prepared from mixtures of sodium sulphate, sodium orthophosphate, silicon dioxide, aluminium oxide and magnesium oxide. The ranges covered were equivalent to samples containing 0-250 g kg-1 of SO4, 0-400 g kg-1 of Si02 and PO4, 0-200 g kg-1 of A1203 and 0-100 g kg-1 of MgO. A reference bead was prepared, containing all five elements at about the sample equivalent of the 50 g kg-1 level, which was placed in the monitor position of the X-ray spectrometer.Table 1. X-ray spectrometer operating parameters Parameter S P Si Mg A1 Peak angle/"28 . . , , . . 110.710 141.030 144.710 45.165 145.125 Backgroundangle/"28 . . . , 111.710 140,030 143.710 44.165 146.625 Crystal . . . . . . . . Germanium Germanium Indium Thallium Pentaerythritol an tirnonide acid phthalate176 ANALYST, FEBRUARY 1986, VOL. 11 1 X-ray fluorescence measurements were made using the conditions given in Table 1 for each element in each calibration sample. The X-ray count for each measurement was corrected for background and the ratio of corrected count against the corrected monitor count, for the same element, was calculated. The use of a count ratioed to a monitor bead compensates for any long-term instrumental drifts. A regression analysis was made to calculate the inter- element interferences and to obtain a calibration equation for .each element.These calculations were made using a Philips 851 computer operating with Philips X14 An2-3 (UK) software. The operating parameters of the X-ray spectrometer are shown in Table 1. A rhodium tube operating at 50 kV and 50 mA was used for all measurements. All measurements were of 50 s duration, with a coarse collimator and a flow proportional counter. The peak angles given were for the Ka peak of the element in each instance. Sampling Washing powders are generally heterogeneous in nature, so care must be taken in sampling from packets. The procedure used in our laboratories was to rotary divide successively the contents of an entire packet down to a mass of approximately 30 g.This sample was then ground to a fine powder in a high-speed blade mill to provide a homogeneous, representa- tive sample from the packet , for subsequent analysis. Preparation of Samples for X-ray Fluorescence Analysis The sample fusion used was a two-stage procedure.s5 In the first, the oxidation stage, 0.25 g of the ground sample was mixed with 1 g of sodium carbonate and 1 g of a mixture containing 3 parts by mass of sodium nitrate and 1 part by mass each of potassium nitrate and strontium nitrate. This mixture was heated in a platinum - gold crucible at approximately 800 OC for 6 min without any agitation. In the second, the fusion stage, this mixture was agitated with 5 g of dilithium tetraborate at approximately 950 "C for 3 min. All weighings were carried out to +0.001 g.The fused melt was finally poured into a casting mould to give a glass bead suitable for X-ray fluorescence analysis. The oxidation and fusion stages were carried out using the automatic bead machine. The microprocessor in this instru- ment controls six parameters, namely time, heating power, crucible movement, angle through which the crucible is rotated, the speed of movement and the presence or absence of cooling air. The microprocessor also allows these para- meters to be grouped into up to six operations for the preparation of the bead. In the sequence for this particular fusion, the sample and the nitrate - carbonate mixture were placed on one side of the crucible and the dilithium tetraborate on the other. The first heating period, without any agitation, allows complete oxida- tion of any organics in the sample.The second heating period, with agitation, then mixes the oxidised sample with the tetraborate and produces a homogeneous melt. A similar fusion technique can be carried out manually. In this instance,, the oxidation stage is carried out in a furnace at Table 2. X-ray calibration data Concentration Element range/g kg-1 olg kg-1 K SasSO, . . . . . . 0-250 4.0 1.4 PasPO, . . . . . . 0-400 1.7 0.6 SiasSiO,? . . . . . . W O O 4.9 1.3 MgasMgO . . . . &lo0 3.0 1.6 AlasAI,O, . . , , 0-200 2.5 2.2 800 "C for 6-10 min. The fusion stage is then carried out either in a second furnace at 950 "C, or over the flame of a Meker burner. The fusion time for the second stage for manual operation is about 20 min, with manual swirling at 5-min intervals.The melt is poured into a heated casting mould to give a glass bead similar to that prepared on the automatic bead machine. Results Calibration The calibration equations for each element were calculated using the Philips X14 software, using the de Jongh model, with influence coefficients (a) calculated from the experimental data during the regression analysis. The values of the standard deviation, 0, and the propor- tionality factor, K , from the regression analyses are given in Table 2. Precision The precision of the sample preparation and measurement was demonstrated by fusing and measuring six sample beads from one ground sample. The results are shown in Table 3. With the exception of magnesium, the relative standard deviation of 1-3% for the sample preparation and measure- ment was of the same order as the over-all accuracy that could be expected from residual sample mass fluctuations on differing samples.The precision of the method for these four elements therefore satisfied our original aims. The poorer precision for magnesium , with a relative standard deviation of lo%, reflects the low concentration being measured and the fact that magnesium is the lightest of these elements. Accuracy The samples, which contain a mixture of organic and inorganic components, were prepared by oxidation and fusion with fixed masses of flux. The loss of volatile material during fusion is the limiting factor in the accuracy of the method, which is about k2Y0 relative.The accuracy of the method was demonstrated by analysing a known washing powder base and then adding known Table 3. Precision of sample preparation and X-ray measurement Concentratiodg kg-1 Fusion No. 1 2 3 4 5 6 Standarddeviation . . . . . . Relative standard deviation, YO Mg 27 28 26 27 30 35 3.3 11.4 so4 107 107 106 107 110 106 1.5 1.4 PO4 203 204 203 208 207 203 2.2 1.1 Si02 39 39 39 39 39 37 0.85 2.2 A1203 48 49 48 48 48 47 0.63 1.3 Table 4. Analysis of washing powders of known composition Concentratiodg kg-l Sample No. Mg so4 A . . . . . . . . Known 0 178 X-ray 2 182 B . . . . . . . . Known 14 142 X-ray 13 135 C . . . . . . . . Known 28 107 X-ray 29 107 D . . . . . . . . Known 42 71 X-ray 40 67 PO4 339 342 27 1 272 203 205 136 135 SiO, AI2O3 66 0 69 <1 53 25 49 23 40 50 39 48 26 75 26 78ANALYST, FEBRUARY 1986, VOL.111 177 Table 5. Comparison of wet chemical and X-ray results Concentration/g kg - 1 Sample No. Method 1 Chemical 2 Chemical 3 Chemical 4 Chemical X-ray X-ray X-ray X-ray Mg so4 188 <0.1 191 - 191 - 151 <0.1 149 - 153 <0.1 150 - <0.1 196 PO4 1 1 295 289 313 314 325 330 Si02 A1203 42 - 36 <0.1 53 - 51 <0.1 37 - 39 <0.1 63 - 59 <0.1 amounts of magnesium and aluminium to the base. The magnesium was added as magnesium monoperoxyphthalate and the aluminium as aluminium oxide. In each instance 30-g aliquots of material were prepared and ground in a high-speed grinder with stainless-steel blades to give a homogeneous sample. The results are given in Table 4. These results are in agreement within about k2%, which is the limit of the accuracy expected.As a further check on the accuracy, a number of samples of commercial washing powders were analysed by X-ray fluorescence and by wet chemical methods for sulphur, silicon and phosphorus. The wet chemical analysis was carried out on the solution resulting from dissolution of the melt from a sodium carbo- nate - sodium nitrate fusion. Sulphate was determined gravimetrically as barium sulphate, silicon by acid precipita- tion followed by volatilisation with hydrofluoric acid and phosphate spectrophotometrically with molybdovanadate. The results are given in Table 5 . These results again are within the expected k2%. Discussion The total analysis of washing powders is a complex procedure. The X-ray method described here gives a determination of total sulphur, phosphorus, silicon, aluminium and magne- sium.Sulphur particularly can exist in inorganic forms such as sodium sulphate and organic forms such as anionic surface- active agents. If necessary, these different forms of sulphur can be characterised by carrying out a preliminary separation of organic and inorganic fractions by solvent extraction, chromatography, etc. Alternatively, if the sulphur is present as an anionic surface-active agent then this component can be titrated with a cationic titrant.’ Similarly, silicon can be present as both sodium silicate and as zeolite. In this instance the aluminium and silicon figures would have to be used together to assess the formulation. Conclusions X-ray fluorescence spectrometry combined with an automated bead maker provides a rapid method for the determination of total sulphur, phosphorus, silicon, aluminium and magnesium in washing powders. The results obtained are within +2% relative of those obtained by wet chemical analysis. The oxidative fusion technique described demonstrates the feasibility of preparing glass beads for X-ray fluorescence spectrometry from difficult samples containing varying amounts of organic and inorganic constituents. References 1. 2. 3. Longman, G. F., “The Analysis of Detergents and Detergent Products,” Wiley, Chichester, 1975. Oliver, G. J., Eur. Spectrosc. News., 1976, No. 8, 4. Petin, J., Bentz, F., and Wagner, A., paper presented at the International Conference on Industrial Inorganic Elemental Analysis, Metz, June lst-5th, 1981. Wagner, A., Petin, J . , Hein, J., and Bentz, F., “Spectrometer- tagung,” Walter de Gruyter, Berlin, New York, 1981, pp. Petin, J., and Wagner, A., Labor Praxis, 1983, May, 416. 4. 71-81. 5. Paper A41416 Received November 27th, 1984 Accepted September 3rd, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100175

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 179 Determination of a-Impurities in the P-Polymorph of lnosine Using Infrared Spectroscopy and X-ray Powder Diffraction David H. Doff Department of Geolog y, Trinity College, Dublin 2, Ireland Frank L. Brownen Newport Pharmaceuticals (Ireland), Baldo yle, Dublin 13, Ireland and Owen 1. Corrigan School of Pharmacy, Trinity College, Dublin 2, Ireland An X-ray powder diffraction method for the quantitative determination of the a-inosine content of mixtures of the crystalline polymorphic forms of a- and p-inosine is described. Oriented sample discs are prepared by pressing with a cellulose binder and the a-inosine content is calculated from measurements of the 8.25 A and 7.25 8, X-ray diffraction intensities of a- and (3-inosine, respectively. The method has a relative deviation of 7% and a detection limit of 0.4% a-inosine. This represents a 35-fold improvement in sensitivity over conventional infrared spectroscopy and a 14-fold improvement over Fourier transform infrared spectroscopy.Keywords: a-lnosine determination; polymorphs; X-ray diffraction; infrared spectroscopy; pharmaceuticals There is a growing interest in the pharmaceutical applications of polymorphism1 and hence in techniques for its identifica- tion and quantification. It is also required routinely by some national drug regulatory agencies that specific methods for detecting the absence of unwanted polymorphs be developed as routine quality control assays. Such unwanted polymorphs will not necessarily have any intrinsic toxicity but they must nevertheless be quantified.Inosine, C10H12N405, crystallises in two polymorphic anhydrous forms, one of which is orthorhombic and one mon~clinic.~J Inosine also forms a dihydrate that is mono- clinic. Suzuki and Nagashima2 designated the orthorhombic form of anhydrous inosine the a-form and the monoclinic form of anhydrous inosine the p-form. In our work it was important to detect and quantify, with a detection limit of less than 5% , the presence of a-inosine in samples of the P-polymorphic form. Preliminary investigations were made using differential- scanning calorimetry (DSC) but this technique proved unsuit- able because of the closeness of the melting-points of the two polymorphic forms. We therefore investigated the suitability of infrared spectroscopy and X-ray powder diffraction methods for this purpose.The application of infrared spectroscopy to the analysis of pharmaceutical compounds is well known. However, no methods for the determination of inosine in the solid state appear to have been reported. Quantitative X-ray diffrac- tometry is a technique that has considerable potential in the assay of pharmaceutical materials and was used by Christ et aL4 in their determination of crystalline sodium penicillin G. Shells pointed out that "published reports on quantitative diffraction applications to organic systems are almost non- existent ," and described several methods of determining crystalline components in drug systems. However X-ray powder diffraction has continued to find only occasional pharmaceutical application, such as in the analysis of intact tablets by Papariello et a1.6 and oral suspensions by Kuroda.7 More recently Imaizumi et aZ.8 reported a determination of the degree of crystallinity of indomethacin using lithium fluoride as an internal standard.A useful review is given by Zwell and Danko.9 An X-ray diffraction analysis of mixtures of inosine polymorphs was described briefly by Sumkilo and this formed the basis of our method. Experimental Instrumentation Conventional infrared spectra were obtained using a Pye Unicam SPS 300 instrument. Spectra were also recorded using a Nicolet 5MX Fourier transform infrared spectrometer using a continuous scan Michelson Interferometer, S/5 optics with spectral range 4600-400 cm-1, laser-assisted sample alignment with a resolution of 4 cm-1 and a wavelength accuracy of 0.01 cm-1 and automatic gain.A standard glower source was used and a two-pen digital plotter with variable abscissa and ordinate expansion. X-ray diffraction measurements were made using a Philips PW 1050 wide-range goniometer with a 1" dispersion slit, a 0.2-mm receiving slit and a 1" anti-scatter slit. The Cu anode X-ray tube was operated at 40 kV and 20 mA in combination with a Ni filter to give monochromatic Cu Ka X-rays of wavelength 1.5418 A. The angular calibration of the goniometer was based on the (020) reflection of cholesterol11 at 5.260 "28. Materials a-Inosine, crystalline. Ajinomoto Co. , Inc. , Japan. p-Inosine, crystalline. Ajinomoto Co., Inc. , Japan. Cellulose powder, Whatman CF 11.Sample Preparation For the infrared spectroscopic studies potassium bromide discs containing 1% of inosine were prepared by compression under vacuum at a pressure of 10 ton. To prepare samples for X-ray diffraction analysis the following method is used. Grind together by hand 1.00 g of inosine sample and 1.00 g of cellulose powder using an agate pestle and mortar. Transfer the mixture into a glass vial, add two 6 mm diameter polystyrene balls, stopper and further homogenise the mixture by shaking for 5 min in a Glen Creston ball mill. Transfer the resulting powder into a 32 mm diameter stainless-steel die and press into a disc at a pressure of 10 ton. The reasons for adopting this method of sample180 1.5 m 0 X Y c I, 1.0 In C S 0 0 w 0.5 c I I I I I 13 12 1 1 10 9 '20CuKa Fig.1. Portion of X-ray diffraction chart of a mixture of a-inosine (30%) and p-inosine (70%) showing the measurements used to derive the intensity values Z, and Zo. Pressed disc, cellulose binder. Z, = A - A'; ZB = B - B ' . Indexing of peaks from Suzuki and Nagashima2 I I I I I I I 1700 1600 1500 1400 Wavenumbedcm - l Fig. 2. 11, 25%; 111, 20%; IV, 15%; and V, 10% a-inosine Infrared transmittance spectra of inosine mixtures. I, 50%; al C m e In 2 ANALYST, FEBRUARY 1986, VOL. 111 1650 1630 1610 1590 1570 Wavenum bedcm - Fig. 3. Absorbance spectra of inosine mixtures acquired using the Fourier transform IR spectrometer. V, 10%; VI; 5% and VII, 2.5% a-inosine preparation as opposed to the conventional cavity mounts are discussed in detail below.Procedure for X-Ray Diffraction Analysis Insert the inosine - cellulose disc into the diffractometer and record the number of counts obtained in 60 s at 9.75 "28 (a-inosine background), 10.72 "28 (a-inosine peak), 11.74 "28 (p-inosine peak) and 12.50 228 (p-inosine background). Subtract the appropriate background counts from the peak counts to give the diffracted intensity values I, and Zp for a- and p-inosine, respectively (Fig. 1). Calculate the a-inosine concentration, C, from the following equation: 1001, c, = O/o (I, + 2.52 Ip) Determine also the blank using a sample of pure p-inosine reference standard. It is assumed that a preliminary diffractogram of the sample will have been recorded in order to check the peak positions and to verify that interfering peaks are absent.All our samples were known to contain only anhydrous inosine. Results and Discussion Infrared Methods Conventional infrared spectra provided a rapid means of identifying the form of inosine present. Visual examination of the spectra suggested that the a-band at 1593 cm-l and the P-band at 1577 cm-1 were the most suitable bands for quantification by the base-line method.12 Transmittance spectra illustrating these bands for inosine mixtures containing 10,15,20,25 and 50% of the a-polymorph are shown in Fig. 2. The peak at 1593 cm-1 is visible in spectra of samples containing 15% or greater of the a-form, but was absent in samples containing 10% or less of a-inosine. The detection limit for a-inosine was improved by using the Fourier transform instrument.Absorbance spectra (with subtraction) of inosine mixtures containing 10, 5 and 2.5% a-inosine are shown in Fig. 3 and indicate that a-inosine may be detected down to about 5% by this method. X-Ray Diffraction Method Peak height versus peak area On many diffractometers it is easier to measure peak heights rather than peak area and for the sake of simplicity we have adopted this practice here. In some instances peak-height measurement also has the added advantage of avoiding interference from adjacent peaks. On the other hand,ANALYST, FEBRUARY 1986, VOL. 111 181 peak-area measurement can result in improved precision and compensates for the loss of intensity owing to crystal defects and small particle size. Preferred orientation effects In our preliminary investigation of an X-ray diffraction assay for inosine polymorphs we used conventional cavity mounts.However, satisfactory precision could not be obtained and this was attributed to the crystal morphology of the two phases. a-Inosine forms long needle-like crystals while p-inosine crystals have a platy habit (shape), and it is very difficult to make a randomly oriented cavity mount of either of these materials because of the tendency for the crystals to adopt preferred orientations. Consequently, the reproducibility of X-ray diffraction intensities from cavity mounts is very poor even when careful packing procedures are employed. It is worth pointing out that preferred orientation affects (enhances or diminishes) all X-ray reflections from a crystal- line powder and not just those arising from lattice planes parallel to the orientation. The problem cannot, therefore, be approached by choosing to measure a different reflection.An attempt was made to overcome these orientation effects by grinding the sample with a small amount (2%) of carbon black, as recommended for this purpose by Christ et aZ.4 However, no improvement in precision could be obtained by this means and scanning-electron micrographs of the resulting powders showed the characteristic morphology of the inosine crystals to be unchanged. The use of cavity mounts was then abandoned and instead highly oriented mounts were made in the form of pressed discs with a cellulose binder. These discs are self supporting, easy to handle and can be inserted directly into the diffractometer.A comparison of the diffraction patterns of both a-inosine and p-inosine showed little differ- ence in relative peak intensities between the cavity mounts and the pressed discs, confirming that a high degree of preferred orientation was already present in the cavity mounts. The reproducibility of X-ray intensities from the pressed discs was, however, greatly improved and this method of sample preparation was therefore adopted. No deleterious effects on crystal structure owing to pressure were observed. It is possible that different crystallisation regimes may give rise to crystals of different habit and hence different orienta- tion behaviour. The grinding step is therefore of considerable importance as grinding alters the original crystal habit by producing cleavage flakes, thus ensuring that the morphology of the grains in the resulting powder is uniform for any given phase.It also ensures that the habits present in both standard and sample are essentially the same. Accuracy The accuracy of the X-ray method was assessed by analysing inosine mixtures containing 0,0.5,1,10,20,30,40 and 50% of the a-polymorph. The results are shown in Table 1. The Table 1. Results of X-ray diffraction analysis of mixtures containing known masses of inosine polymorphs Amount of winosine, % Taken 0.00 0.50 1 .oo 10.00 20.00 30.00 40.00 50.00 Found* 0.002 0.54 0.90 11.45 19.32 30.24 40.49 49.72 n on - 1 8 0.10 8 0.14 8 0.13 2 8 1.47 2 2 2 - - - - * Mean of n replicate analyses of the same disc. relationship between the a-inosine concentration taken and the a-inosine concentration found is described by a linear regression equation with slope equal to 0.997 and an intercept of 0.20% of a-inosine. This closely approximates to the ideal line of unit slope and zero intercept. Precision The precision of the method was determined from the analysis of eight replicate discs made from an approximately 1 + 3 mixture of a-inosine and p-inosine by mass.The mean a-inosine concentration found was 24.18% mlm with a standard deviation of 1.68% mlm. 'This represents a relative deviation of 6.95%. Counting statistics alone account for a relative deviation of about 2%. Detection limit In this work it was anticipated that most of the inosine samples that would be assayed would not contain any detectable amount of a-inosine. An accurate determination of the detection limit of the method was therefore of paramount importance.Accordingly, two approaches were adopted for the estimation of the detection limit. The first approach was based on the standard deviation of the blank. A blank disc was prepared using a pure P-inosine reference standard. This standard was considered to be pure because no a-inosine peak was present in the X-ray diffraction pattern and measurement of the parameter Z, gave a value identical to that of pure cellulose. Eight replicate analyses of this disc had a standard deviation of 0.095% mlm. The detection limit can therefore be taken to be three times this value, i.e., 0.29% a-inosine. However, we consider this to be an over-optimistic estimate of the detection limit because it assumes that the noise charac- teristics of the background are similar to the noise characteris- tics of the signal.In X-ray diffraction methods this is not a valid assumption as a significant component of the signal noise is likely to be represented by non-reproducibility in the sample preparation technique. The second approach is an extension of the first and is more realistic in that it incorporates information about the analytical precision into the definition of detection limit. Following the theoretical model of the relationship between precision and concentration developed by Thompson and Howarthl3 the detection limit is defined as the concentration at which the precision is 100%. For a 3u detection limit this value can be found from the following equation: where c d is the detection limit, o0 is the standard deviation at zero concentration and k is the precision of the method at high concentrations expressed as a relative deviation.Substituting the values uo = 0.095 and k = 6.95 leads to a detection limit of 0.4% a-inosine. Calculation of a-Znosine Concentration from I, and Ig Values The equation already given for the calculation of a-inosine concentrations can be derived from the fundamental equation for quantitative X-ray diffraction stated by Klug and Alex- ander.14 The diffracted X-ray intensity ( I ) arising from a component of a mixture is given by the equation where k is a constant depending on the nature of the component, x the mass fraction of the component, p the density, p the mass absorption coefficient and pm the mass absorption coefficient of the matrix.In the special case of182 ANALYST, FEBRUARY 1986, VOL. 111 mixtures of polymorphic forms of a substance, as is the case with the inosine system, we have and therefore equation (1) can be simplified to I= ( i ) x o r I = K x where the intensity is directly proportional to the concentra- tion and the constant K is a measure of the sensitivity. For a binary mixture of two components, A and B, we obtain the two equations: I A = K A XA . . . . * * (2) I B = K B x B . . . . . . * * (3) and As A and B are the only components (the cellulose binder can be neglected), XB = 1 - XA Substituting this value of XB in equation (3) and combining equations (2) and (3) leads to the equation The ratio KA/KB is the ratio of the sensitivities of the two components A and B.The value of this ratio was determined by measuring the diffraction intensities of the a- and P-diffraction peaks using a disc pressed from a mixture of equal masses of the two components and was found to be 2.52. It would be advisable for other workers to determine their own values for this ratio as it may vary slightly between different instruments and different laboratory procedures. Interestingly, Chung15 has shown that relationships similar to equation (4) are applicable to any binary system and not only those consisting of two polymorphs. This means that the method of calculation described above can be used in a wide range of pharmaceutical assays. Conclusion X-ray powder diffractometry, in combination with a pressed- disc sample-mounting technique, provides an accurate and sensitive method for determining the polymorphic composi- tion of crystalline inosine samples.The sensitivity is 35-fold greater than that obtainable using conventional infrared spectroscopy and 14-fold greater than that using Fourier transform infrared spectroscopy. X-ray diffraction methods such as this are capable of application to a wide variety of pharmaceutical assays. We are grateful to Mr. E. Pfadenhauer and Prof. J. M. D. Coey for their critical comments on an earlier version of this paper and to Dr. C. Brady and Ms. P. Rafferty for the use of the FTIR. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Haleblian, J., and McCrone, W., J . Pharm. Sci., 1969,58,911. Suzuki, Y., and Nagashima, N., Bull. Chem. SOC. Jpn., 1970, 43, 1600. Munns, A. R. I., and Tollin, P., Acta Crystallogr., 1970, B26, 1101. Christ, C. L., Barnes, R. B., and Williams, E. F.,Anal. Chem., 1948, 20, 789. Shell, J. W., J . Pharm. Sci., 1963, 52, 24. Papariello, G. J., Letterman, H., apd Huettemann, R. E., J. Pharm. Sci., 1964, 53, 663. Kuroda, K., J. Pharm. Sci., 1968, 57, 250. Imaizumi, H., Nambu, N., and Nagai, T., Chem. Pharm, Bull., 1980, 28, 2565. Zwell, L., and Danko, A. W., Appl. Spectrosc. Rev., 1975, 9, 167. Suzuki, Y., Bull. Chem. SOC., Jpn., 1974, 47, 2549. Kittrick, J. A., Proc. Soil Sci. SOC. Am., 1960, 24, 17. Williams, W. D., in Beckett, A. H., and Stenlake, J. B., Editors, “Practical Pharmaceutical Chemistry,” Part 2, Third Edition, Athlone Press, London, 1976, p. 331. Thompson, M., and Howarth, R. J . , Analyst, 1976, 101,690. Klug, H. P., and Alexander, L. E., “X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials,” Second Edition, Wiley-Interscience, New York, 1974. Chung, F. H., J . Appl. Crystallogr., 1974, 7, 519. Paper A5/150 Received April 25th, 1985 Accepted August 23rd, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100179

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 183 Continuous Spectrophotometric Monitoring of Chlorine in Air Aviva Shina and J. Gabbay Research Institute for Environmental Health, Israeli Ministry of Health, Sackler School of Medicine, Te I-A viv U n ive rsit y, Ram a t-A viv, Te I-A viv, Is ra e I Spectrophotometric monitoring of atmospheric free chlorine, based on the reaction with 4-nitroaniline, was assessed. A Beckman air quality Acralyzer Model K-1008 for nitrogen dioxide monitoring was modified and used for this purpose. Some reducing and oxidising compounds interfered with chlorine measurements, but these were removed efficiently using a scrubber consisting of chromic acid (prepared from phosphoric acid and chromium trioxide) absorbed in silica gel. This scrubber was found to be suitable for chlorine measurements in the presence of hydrogen chloride. Humid conditions resulted in some absorption of chlorine in the scrubber.At 69°C no such absorption was observed. Ozone was removed either by a glass-beads scrubber heated to 60-65 "C or by the addition of sodium nitrite to the absorbing solution. The sensitivity achieved using this monitor was 0.25 p.p.m. full-scale with solution and air flow-rates of 1 ml min-I and 2 I min-1, respectively. Keywords: Chlorine monitoring; spectrophotometric monitoring; air monitoring Various continuous chlorine monitors, based on different theoretical principles, have been developed and manufactured commercially by many companies. Some monitoring instru- ments have been in use within industrial areas for measuring high levels of chlorine, such as the personal monitors described by Langhorst.1 Others, used for ambient air monitoring, need to be more sensitive and selective owing to the greater interference of other pollutants.Some of the methods used for ambient chlorine monitoring include spectrophotometry, electrochemistry and indirect chemiluminescence. All of these suffer from interferences from other pollutants either oxidising or reducing. The extent of the interference differs from one to another depending on the method and on the way in which the monitor was manufactured. A spectrophotometric method, based on the reaction of chlorine with 4-nitroaniline in the presence of a barbitone sodium catalyst, was developed in our institute.2 This method exhibited some advantages over other methods using wet chemistry.2 The aim of this study was to assess this method for the continuous monitoring of chlorine in air.Two types of scrubbers are needed for the removal of interfering compounds: (1) those for reducing compounds and (2) those for oxidising compounds. 1. A large negative interference is caused by reducing sulphur compounds2 such as sulphur dioxide , hydrogen sulphide and thiols. In order to remove these interferences, it was necessary to use oxidising scrubbers, such as those used for monitoring oxidants. These scrubbers consisted of chromic acid absorbed on various inert carriers such as a fibre-glass filter,3 silica gel,374 pumice, boric acid4 and quartz chip^.^.^ Chromic acid was prepared by mixing chromium trioxide and sulphuric3 or phosphoric"5 acid.The scrubber, consisting of chromium trioxide, sulphuric acid and a fibre-glass filter, was originally prepared to remove sulphur dioxide.3 This scrubber was found to be efficient also in removing other reducing sulphur compounds, in addition to ammonia and hydrogen peroxide.2 No significant oxidation of HC12 or NaCP was observed using this scrubber. This scrubber was tried in the monitoring of chlorine at the Electrochemical Industries Frutarom factory, which is located along the seashore in the city of Acre, Israel, which has a very humid climate. The scrubber used became wet within a few days under the usual working conditions and it was therefore necessary to develop scrubbers having a higher capacity and higher resistance to humidity.It was thought that scrubbers made of silica gel and chromic acid would have higher oxidation capacities and for this reason the oxidation of chlorides by these scrubbers was studied. 2. Negative interference is caused also by ozone. Wartburg et aZ.6 eliminated this interference by passing the sampled air over heated scrubbers consisting of stainless-steel frits. Lindqvist4 used heated silver, aluminium or platinum scrub- bers for this purpose. In this work heated glass beads were tried. The effect on ozone removal of adding sodium nitrite to the reacting solution was also investigated. Experimental All chemicals were of analytical-reagent grade, unless indi- cated otherwise. Chemical Method The spectrophotometric method used was based on the chemical reaction of free chlorine with an alkaline solution of 4-nitroaniline in the presence of a barbitone sodium catalyst.2 The change in the absorbance of the solution at 485 nm is linearly proportional to the concentration of chlorine.Analy ser A Beckman air quality Acralyzer Model K-1008 for continu- ous monitoring of nitrogen dioxide was used. The instrument consists of three units: an analyser, a controller and a recorder.. The analyser is made up of a visible-range ratio photometer, a solution metering pump, an air pump, a glass flow system and a flow meter. The following modifications were made. 1. The original optical filter was replaced with a Corning No. 4060 CS-4-67 1.88-mm optical filter, suitable for reading the absorbance of the absorbing solution.2. The gain and the ratios of the Wheatstone bridge were changed. 3. A 0-5 1 min-1 Wisa diaphragm air pump was used instead of the original pump. The pump was connected to a variable voltage supplier. 4. A flow meter of range 0-2.5 1 min-1 was used. 5. The original flow system was modified and the final design is shown in Fig. 1. The sampled air is drawn into the glass flow system by the air pump (L) and enters through the bottom of the absorber184 ANALYST, FEBRUARY 1986, VOL. 111 b A +- D H 1 I I f & I h I I I _ _ J +-- _ _ I I I I .t I I I I -I Fig. 1. Modified flow system. Solid line, solution flow; and broken line, air flow. A, Solution reservoir; B, reference optical cell; C, solution metering pump, 0-2 ml min-1; D, 7-turn concurrent absor- ber; E and F, separators; G, sample optical cell; H, solution waste reservoir; I, trap; J, flow meter, 0-2.5 1 min-l; K, air flow control valve; and L, air pump, 0-5 1 min-1 (D), where chlorine reacts with the reagent.The direction of the air flow is the same as that of the reagent flow through the absorber. After the absorber, the air passes through separator E, where it is separated from the reagent and then drawn through a solution waste container (H), trap (I), flow meter (J), sample control valve (K) and out through the air pump. The solution is pumped from the reagent reservoir (A) through the reference optical cell (B) into the absorber, where the chemical reaction occurs. It is then drawn through separator E to the sample optical cell (G). Separator F provides a constant-pressure balance for smooth flow through the sample optical cell.From the optical cell it is drawn into the waste reservoir (H). Preparation of Scrubbers Coarse colourless silica gel crystals (BDH Chemicals) were initially saturated with doubly distilled water. The saturation was carried out slowly, to prevent breakage of the crystals. The system consisted of a midget impinger bubbler, which contained 20 ml of doubly distilled water, and was connected in series with two glass tubes. The first tube contained the colourless silica gel and the second contained dry blue silica gel. Air was drawn through the system at a flow-rate of about 200 ml min-1 until the blue silica gel crystals changed colour. The following scrubbers were prepared. A. A 30-g amount of the wet colourless silica gel crystals was immersed in 20 ml of solution containing 8.5 g of chromium trioxide (BDH Chemicals) and 3.5 g of concen- trated sulphuric acid (BDH Chemicals).The impregnated silica gel crystals were packed into a U-tube of 23 mm i.d. and 150 ml volume and dried overnight in an oven at 80 "C, by a stream of dry air. B. A 50-g amount of the wet colourless silica gel was immersed in a solution containing 10 g of chromium trioxide, 10 g (5.9 ml) of 85% orthophosphoric acid (Riedel-de Haen) and 10 ml of doubly distilled water. The silica gel crystals were then dried overnight at 70 "C, as described for scrubber A. Efficiency of Scrubber A in Removing Reducing Sulphur Compounds The efficiency of scrubber A at temperatures in the range 23-60 "C in removing SO2, CH,SH, C2H5SH and H2S was t out t G -11 Fig.2. PT calibration system. Solid line, air flow; and broken line, water flow. A, Air pump; B, silica gel air drier; C, activated charcoal filter: D. U-tube; E, glass beads; F, PT; G, thermometer; H, flow meter; I, thermoregulator, a heater and a water-pump (Haake Model E2); J , water-bath; K, polyethylene foam cover; and L, Haake DK12 cooling system +Line 3 Line 2 L-e-' Fig. 3. Experimental flow system used to study the efficiency of scrubber A in removing reducing sulphur compounds. A,A1 and A2, flow meters; B,B1 and B2, charcoal filters; C and C1, silica gel air drier; D and D1, ovens; E, chromic acid scrubber; F, Beckman Acralyzer; G, air pump; and a, b and c, glass tubes for PTs studied.Known concentrations of the reducing sulphur compounds and of C12 were obtained by using Metronics permeation tubes (PTs). Sulphur compound PTs were calibrated gravimetrically as follows. The calibrating system (Fig. 2) was constructed from a 9-1 water-bath (J) covered with polyethylene foam (K), which isolated it from the surroundings. A system (I) (Haake Model E2) consisting of a thermoregulator, a heater and a water pump and also a cooling system (L) (Haake Model DK 12) were used. A two-chamber glass U-tube (D) (20 cm in height, 2 cm in diameter) having an inlet and an outlet was housed in the water-bath. The two parts of the U-tube were separated by a glass disc, having a few holks that enabled the air to flow through. One chamber was filled with glass beads (E) and the other contained a PT (F) and a thermometer (G).Air was pumped through dried silica gel (B), activated charcoal (C), the U-tube and out through a flow meter (H). The temperature measured within the U-tube was accurate to within k0.05 "C. Calibration of H2S and thiol PTs was carried out just before and after the experiments, by measuring the descrease in mass during certain time periods, at the same temperature of the experiment. Fig. 3 illustrates the system used for the efficiency measure- ments. The air flows through lines 1 and 2 were combined at the entrance to the analyser (F). The two lines were connected to the instrument through a tee. All connections in the system were made of glass or short PTFE tubing. Glass tubes a, b and c in which the PTs were placed were thermostatically controlled throughout the range 2040 "C to with- in * 0.05 "C.The temperature of the scrubber (E) wasANALYST, FEBRUARY 1986, VOL. 11 1 185 thermostatically controlled between room temperature and 60 "C to within k1 "C. The chlorine concentration was regulated by three parameters: the emission rate of Clz from the PT (which was controlled by the thermostatic bath temperature), the rates of air flows through lines 1 and 2 and the rate of air flow bypassed through line 3. The concentra- tions of sulphur compounds were adjusted by the air flows through lines 1 and 2-and by the thermostatic bath tempera- ture. Each experiment included the following steps: (1) a, b and c without PTs; (2) a and b without PTs, c containing a C12 PT; (3) b containing a sulphur compound PT, a containing no PT and c containing a C12 PT; and (4) sulphur compound PT in a, C12 PT in c, b containing no PT.The instrument was zeroed by following step 1 and by changing the gain of the recorder. The span was adjusted according to step 2 and by changing the gain of the instrument. C12 concentrations were measured throughout step 2. The percentage of sulphur compound interference was calculated from steps 2 and 3. The scrubber efficiency was determined using steps 2 and 4. The experiments were carried out at a solution flow-rate of 2 ml min-1 and an air flow-rate of 2 1 min-1. The sensitivity was 0.5 p.p.m. full-scale. HCI Interference in the Presence of the Scrubbers and at Different Relative Humidities The flow system shown in Fig. 3 was also used to study HC1 interference in the presence of scrubbers A and B.A hygrometer was housed in a glass tube between the scrubber (E) and glass tube (b) (Fig. 3). The solution and air flow-rates were 1 ml min-1 and 2 1 min-1, respectively. The sensitivity was 0.25 p.p.m. V/V full-scale. Two groups of experiments were conducted as follows. 1. A few experiments were carried out with scrubbers A and B in the absence of C12 at a room temperature of 26 "C. Air was sampled through dry silica gel, and the dry scrubber was maintained at 57-69 "C. Under these conditions it was found that the relative humidity achieved was 10%. 2. Other experiments were carried out with scrubber B in the presence of C12. These were performed at a room temperature of 24-25 "C and scrubber temperatures in the range 25-53 "C.In these experiments the silica gel in line 1 was disconnected. Some of the experiments were carried out at room relative humidity (58%) and some at a relative humidity close to saturation at room temperature. Saturation was achieved by (i) replacing the silica gel tube with an impinger containing distilled water and (ii) bubbling air through the system overnight in order to achieve the proper conditions for the HCl PT and the scrubber. The experimental steps were the same as those described above. Effect of 0 3 on C12 Measurements In order to study O3 interference, certain changes were made to the system shown in Fig. 3. Line 1 was changed as follows: air was drawn to the analyser through an O3 generator and a scrubber of glass beads about 5 mm in diameter.The glass beads were packed in a U-tube of 23 mm i.d. and 150 ml volume. The scrubber was thermostatically controlled in the range 23.5-80 "C. The O3 concentration in these experiments was 0.70 p.p.m. at the mixing point of the two flows. The work was performed at [O,] to [C12] ratios of 1.6 and 2.4. Another means of removing 0 3 was to add 20 g 1-l of NaN02 to the absorbing solution. The system was the same as before, but without the glass beads scrubber. The 0 3 concentration at the mixing point of the flows was 0.10 p.p.m. and the [O,] to [C12] ratio was 2.3. Effect of Humid Scrubber B on C12 Measurements Line 1 in Fig. 3 was used in order to perform these experiments and line 2 was disconnected.The experiments were carried out under three humidity conditions and at different scrubber temperatures in the range 25-69 "C. Different periods of conditioning of the scrubber were used prior to each experi- ment (Table 3). The relative humidity was measured as described under HCI Interference in the Presence of Scrub- bers and at Different Relative Humidities. The three humidity conditions used were as follows. (a) The sampled air was saturated with water vapour at a room temperature of 25 "C. The saturation was achieved by bubbling the air through a midget impinger containing 20 ml of doubly distilled water. The impinger was located in place of the silica gel tube. (b) Air at a room relative humidity of 65% and a room temperature of 25 "C was used. The silica gel tube at line 1, Fig.2 was disconnected. (c) Air at a room temperature of 25 "C was sampled through dry silica gel, placed before the entrance to the scrubber. The relative humidity obtained at this temperature was 10%. The air flow-rate in all experiments was 2 1 min-1. Results and Discussion Performance of the Analyser The modifications that were carried out on the Beckman Acralyzer Model K-1008 enabled it to be used for chlorine monitoring. The sensitivity of the instrument was increased by changing its electronic system and by increasing the air flow. The new pump produced a more stable air flow than the original one. Originally it was connected to an exit from separator F, whereas in the modified system this exit was closed and the air pump was connected to the waste solution container through a trap.This reduced the solution level in separator E (Fig. l), lowering the lag time to about one third of the original (at a solution flow-rate of 2 ml min-1 it was about 10 min). The sensitivity achieved was 0.25 p.p.m. full-scale at a solution flow-rate of 1 ml min-1 and an air flow-rate of 2 1 min-1. Other desired sensitivities could be achieved by changing the solution and air flow-rates. It should be noted that rather than using the Beckman Acralyzer within the system, other instruments, manufactured by other companies, which are more compact and have smaller lag times, could be used. This may alleviate the necessity to modify the electronics and the flow systems. Efficiency of Scrubber A in Removing Reducing Sulphur Compounds Scrubbers consisting of pumice or silica gel as the absorbent, impregnated in a solution conraining chromium trioxide and sulphuric or phosphoric acid, have been tried.A preliminary investigation was carried out on scrubbers made of pumice. They were found to become humid within a short time, causing absorption of the chlorine. The chlorine concentration monitored rapidly decreased to zero, especially at low chlorine concentrations. Scrubbers A and B (see Experimen- tal) were the most efficient, and more resistant to humidity. In order to determine the efficiency of the scrubber, C12 was monitored first in the absence of sulphur compounds. In the next stage, measurements were carried out in the presence of these compounds, with and without the scrubber.If the interference of sulphur compounds in chlorine measurements is due only to the direct reaction between the two, the interference will be given by the percentage of chlorine that reacts. The amount of chlorine that reacts with sulphur compounds in a stationary state will be dependent on the stoicheiometry of the reaction, the concentration of each one of the two, the rate constant of the chemical reaction and also on the time from the moment they enter the solution until186 ANALYST, FEBRUARY 1986, VOL. 111 Table 1. Interference of sulphur compounds and removal efficiency of scrubber A Experi- ment No. 1 2 3i. 4 5 6 7 8 ClZ S Ratio of concentra- concentra- concentra- Sulphur tion, * tion,* tions, so2 0.458 0.110 0.24 so2 0.165 0.136 0.82 so2 0.145 0.127 0.88 so2 0.111 0.155 1.40 CH3SH 0.221 0.078 0.35 C2HSSH 0.536 0.188 0.35 H2S 0.186 0.813 4.40 compound (S) p.p.m.p.p.m. [S]:[C12] C2HSSH 0.159 0.188 1.18 Absorbance in the presence Scrubber of C1, temperature/ and in the "C absence of S 23 0.90 54 0.320 59 0.275 23 0.220 57 0.45 60 1.06 60 0.300 23 0.380 * The concentrations were calculated after mixing of the two flows (lines 1 and 2, Fig. 3) t In this experiment the air drawn through the scrubber was not dried over silica gel. Absorbance in the presence of of C1, and S Without With scrubber scrubber A A 0.57 0.86 0.182 0.335 0.105 0.275 0.035 0.220 0.000 0.45 0.000 1.06 0.000 0.300 0.000 0.380 Inter- ference, YO 36.7 43.1 61.8 84.0 100 100 100 100 Efficiency of scrubber A, O/o 96 100 100 100 100 100 100 100 they reach the optical cell.For certain conditions of reagent flow and geometry of the flow system, this time is constant. If the stoicheiometry of the reaction is 1 : 1 and the reaction is complete, the interference will be directly dependent on the relative concentration of the sulphur compound to that of chlorine. The results of experiments 1-4 in Table 1 show that an increase in the [SO,] to [Cl,] ratio causes a large decrease in the absorbance. With thiols it seems that more than one equivalent of chlorine reacts with each equivalent of thiol. No C12 could be detected in the presence of thiols and hydrogen sulphide at the concentrations indicated in Table 1. In all the experiments the efficiency of scrubber A was close to loo%, independent of the scrubber temperature. A qualitative experiment carried out on SO2 with scrubber B showed the same behaviour as that of scrubber A.Effect of Scrubbers on Oxidation of HCl During C12 Monitoring During the sampling of air containing 10.3 p.p.m. of HC1 and no Cl,, it was found qualitatively that some oxidation of HC1 occurred when using scrubber A. No such oxidation occurred when scrubber B was used. Monitoring of 0.17 p.p.m of C1, in the presence of 10.3 p.p.m. of HC1, as described under Experimental, was carried out using scrubber B. The results show no difference, or only a slight decrease, in the concentrations observed, compared with those obtained in the absence of HC1. This slight decrease was within the experimental error. However, no increase in concentration was obtained.Hence no oxida- tion of HCl occurred even at very high concentrations of HC1 when scrubber B was used. Effect of 0 3 on Clz Monitoring The interference of O3 in the presence of a glass-beads scrubber was studied in two separate experiments, as des- cribed under Experimental. The results show that while the glass-beads scrubber was at room temperature no C12 could be detected in the presence of 03. When the temperature was increased a rapid rise in C12 concentration was observed. At 50 "C about 80% of the value obtained in the absence of O3 was retained. Above 50 "C an additional but slower rise in C12 concentration as a function of temperature increase was observed. An optimum value of about 90% of the "original" reading was obtained at temperatures of 60-65 "C.At higher temperatures a slow decrease in those values was observed. Table 2. Efficiency of a solution containing NaNO, in removing O3 Concentration ratio, NaN02 [031: P 2 l present Absorbance 0 No 0.56 0 Yes 0.56 0.23 No 0.49 0.23 Yes 0.56 Table 3. Effect of humid scrubber on C1, monitoring Experiment No. 1 2 3 4 5 6 7 8 9 Relative humidity at 25 OC* a a a a a b b C C Scrubber tempera- ture/OC 33 44 45 69 69 40 69 25 69 Decrease in C12 concen- tration in Conditioning the presence period of the of scrubber scrubbert/h B,% 16 45 40 10 2 3 70 2 50 0 4 5 75 0 200 0 500 0 * The description of relative humidity a, b, and c is given under Effect of Humid Scrubber B on Cl, Measurements in the Experimental section. t Period during which the scrubber was in contact with the sampled air under the experimental conditions before the beginning of the experiment.As mentioned in a previous paper,2 the addition of NaN02 to the absorbing solution was also useful in preventing O3 interference. An experiment showing the efficiency of such a solution in a monitoring system was carried out. The results are given in Table 2. Effect of Humid Scrubber B on Clz Monitoring Table 3 shows that as the temperature of scrubber B increases, the interference of humid air is decreased. At high relative humidity and a low scrubber temperature (experiment l), the C12 concentration decreased by 45%. At such a low scrubber temperature, conditioning for 16 h was sufficient to increase the water content in the silica gel to a level that caused a significant absorption of C12. When the temperature of the scrubber was increased to 44 "C the interference was de-ANALYST, FEBRUARY 1986, VOL. 111 creased to only 5%, even though the conditioning time was 40 h (experiment 2). At 69 "C the interference found was close to zero, even when the conditioning time was as long as 70 h (experiments 3 and 4). At a lower relative humidity (experiments 6 and 7), the interference decreased. At a very low relative humidity the scrubber did not interfere even at low temperatures and with long conditioning periods. Conclusion The spectrophotometric method based on the reaction of chlorine with 4-nitroaniline can be used for the continuous monitoring of chlorine in air. Reducing sulphur compounds can be removed efficiently using a scrubber prepared from chromic acid absorbed in silica gel. A chromic acid scrubber prepared from chromium trioxide and phosphoric acid has a high capacity and a high 187 resistance to humidity when heated to 69 "C. O3 can be removed either by a glass-beads scrubber heated to 60-65 "C or by the addition of NaNOz to the absorbing solution. 1. 2. 3. 4. 5. 6. References Langhorst, M. L., Am. Znd. Hyg. Assoc. J., 1982,43, 347. Gabbay, J., Davidson, M., and Donagi, A. E., Analyst, 1976. 101, 128. Saltzman, B. E., and Wartburg, A. F., Anal. Chem., 1965,37, 779. Lindqvist, F., Analyst, 1972, 97, 549. Schulze, F., Anal. Chern., 1966,38, 748. Wartburg, A. F., Brewer, A. W., and Lodge, J . P., Int. J. Air Water Pollut., 1964, 8, 21. Paper A4184 Received February 28th, 1984 Accepted August 19th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100183

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 189 Morpholine as an Absorbing Reagent for the Determination Of Sulphur Dioxide V. Raman, J. Rai, M. Singh and D. C. Parashar Chemistry Division, National Physical Laboratory, New Delhi I10 012, India Several methods have been developed for the determination of sulphur dioxide in the environment, although some involve the use of hazardous mercury salts. A method is described of trapping sulphur dioxide by forming an adduct with morpholine. Sulphur dioxide is determined in the adduct using p-rosaniline hydrochloride, which forms a coloured complex with maximum absorption at 560 nm. The method is sensitive and a 0.025 pg ml-1 concentration of sulphur dioxide can be determined in solution. Interferences from Cu(ll), Pb(ll), Mn(ll), Fe(lll), Cr(lll), V(V), nitrite and hydrogen sulphide have been studied.In addition concentrations of morpholine thrown into the atmosphere by this method have been investigated to assess any possible hazards due to the use of morpholine as a trapping agent during the sampling process. Keywords: Sulphur dioxide determination; rnorpholine; spectrophotometry; p-rosaniline hydrochloride; formaldehyde A literature survey reveals that a considerable amount of work has been carried out on the development of various physico-chemical techniques for the determination of trace amounts of sulphur dioxide-a compound which may be harmful to animals and plants.132 The spectrophotometric determination of sulphur dioxide is a useful technique owing to its simplicity and sensitivity and many research papers are available on the subject.>g Most of the methods use mercury(I1) salts either for fixation or for chemical reactions in which excess of mercury(I1) salts are utilised for the determination of sulphur dioxide.3-5>8 Only a few methods that avoid the use of poisonous mercury(I1) salts are available.6.7.+11 Sulphur dioxide is known to be absorbed by many organic substances.sl6 Sulphur dioxide forms an adduct with mor- pholine and its derivatives.13JS This aspect is employed in this work in which morpholine is used as a trapping solution for sulphur dioxide.Sulphite or sulphurous acid is used as a source of sulphur dioxide and is added to morpholine. The product obtained is treated with acid bleached p-rosaniline hydrochloride and formaldehyde to yield a violet colour with an absorption maximum at 560 nm.The intensity of the colour is proportional to the concentration of sulphur dioxide and obeys Beer’s law. Sulphur dioxide in solution at a concentration as low as 0.025 pg ml-1 can be determined by this method. Experimental Apparatus A Hilger and Watts UV H 700 spectrophotometer with 1 cm cells was used for the absorbance measurements. Reagents All chemicals used were of analytical-reagent grade. Doubly distilled water was used for preparing all the solutions. Sodium sulphite solution. A 0.1 g mass of anhydrous sodium sulphite was dissolved in distilled water and the volume was made up to 250 ml. This solution was standardised iodimetric- ally and diluted further to obtain a 4 pg ml-1 solution of sodium sulphite. Fresh solutions were prepared before each experiment as sodium sulphite is unstable in solution.Sulphurous acid. Sulphur dioxide was prepared by the action of concentrated hydrochloric acid on sodium sulphite and passed through distilled water. The sulphurous acid thus formed was determined conductimetrically after oxidation to sulphuric acid using hydrogen peroxide and was further diluted to obtain a 2 pg ml-1 solution. Morpholine solution. Morpholine (4 cm3) was diluted to 1000 ml with distilled water. p-Rosaniline hydrochloride solution. A 100-mg mass of the p-rosaniline hydrochloride was dissolved in distilled water, de-colourised with 15 ml of concentrated hydrochloric acid and diluted to 250 ml. Formaldehyde solution. A 2.5-ml volume of 40% formal- dehyde was made up to 500 ml with distilled water.Sulphamic acid, 0.6%. The solution was prepared freshly. Procedure A 10-ml volume of morpholine solution was pipetted into a series of 50-ml calibrated flasks. Volumes of 2,4, 6, 8 and 10 ml of sodium sulphite or sulphurous acid were then pipetted into these flasks. The flasks were tightly stoppered and thoroughly shaken. This was followed by the addition of 10 ml each of p-rosaniline hydrochloride and formaldehyde solu- tions. The flasks were well shaken and the volume was made up to 50 ml. The pH of the solution was noted and the absorbances were measured at 560 nm against a blank after 20 min. The absorbances were plotted against the concentration of sulphite - sulphur dioxide to obtain the calibration graph.Results and Discussion It is well known that sulphur dioxide forms clathrates and charge-transfer complexes with organic substances.gJ2-16 In this work sulphur dioxide - sulphite was trapped in basic morpholine solution to form the adduct and the sulphur dioxide in the adduct was determined with acidified p-rosani- line hydrochloride. Effect of pH The effect of pH on the reaction of the morpholine - sulphur dioxide adduct with p-rosaniline hydrochloride and formal- dehyde was studied. It was found that the intensity of the violet colour obtained by the reaction was enhanced with an increase in pH. The colour intensity of the dye was also pH dependent. The effect of pH on the intensity of the colour of the dye was studied with the reagent blank and it was found that the change in absorbances due to dye was at a minimum when the pH of the final solution was between 0.8 and 1.0190 ANALYST, FEBRUARY 1986, VOL.111 Table 1. Comparison of the morpholine adduct method with the West - Gaeke method Absorbance at 560 nm 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 PH Fig. 1. Effect of pH on the reagent blank (Fig. 1). The acid present in p-rosaniline hydrochloride not only helps in releasing sulphur dioxide but also in achieving a pH of 0.8-1.0. Beer’s Law and Reproducibility Beer’s law was obeyed in the concentration range 1.25-20 yg of sulphur dioxide in 50 ml of solution. The method was found to be reproducible and the standard deviation was found to be 0.006. Comparison of Morpholine with Sodium Tetrachloromercu- rate(I1) for Trapping Sulphur Dioxide The validity of the proposed method was checked by using sodium sulphite and sulphurous acid, whose sulphur dioxide content had been determined by the West - Gaeke method. Care was taken to maintain the same pH in both the methods to avoid the variation in the absorbance of the solution owing to the effect of pH on the dye.It was observed that the absorbance was independent of the nature of the trapping solution. This is illustrated in Table 1. The stability of sulphur dioxide or sulphite in morpholine was also investigated. Anhydrous sodium sulphite (50 mg) was added to 20 ml of morpholine solution and the volume was made to up 100 ml with doubly distilled water. Aliquots were taken and analysed for the sulphur dioxide content once a day for 15 days.It was found that the sulphur dioxide - sulphite content remained unchanged. Absorption Efficiency Sulphur dioxide, prepared by the action of hydrochloric acid on sodium sulphite, was collected in a Corning glass sampler (500 ml) provided with a septum arrangement to draw known amounts of sulphur dioxide. Using this gas, a dilute sample solution with a sulphur dioxide concentration of 15 pg ml-1 was prepared. A 1-ml volume of the gas (15 pg ml-l) was directly injected into 10 ml of 0.4% morpholine using a gas-tight syringe. In another experiment 1 ml of the gas (15 pg ml-1) was injected into a gas sampler of 500 ml capacity, which was initially flushed with pure dry nitrogen. The sulphur dioxide - nitrogen mixture in the gas sampler was passed through two midget inpingers (25 ml capacity) containing 10 ml of 0.4% morpholine connected in series using pure nitrogen as a carrier gas at a flow-rate of 0.5 1 min-1 for 30 min.These impingers are similar to the midget impingers of Type 9700 of Ace Glass Company.18 The sulphur dioxide deter- Amount of sulphur dioxide West - Gaeke Morpholine addedlpg method adduct method 3.24 0.022 0.023 6.47 0.051 0.044 9.71 0.076 0.076 12.94 0.099 0.099 16.18 0.130 0.130 mined in both the impingers indicated that the first impinger contained 100% sulphur dioxide, which was comparable to the concentration of sulphur dioxide obtained by directly injecting 1 ml into 10 ml of 0.4% morpholine. No sulphur dioxide was found in the second impinger. Application of the Method To extend the method to the determination of sulphur dioxide in air, laboratory air was passed through 10 ml of 0.4% morpholine and through sodium tetrachloromercurate(I1) using two separate flow meters.The sulphur dioxide concen- tration determined by both the methods was found to be 0.08 p.p.m. Effect of Foreign Ions The interference of nitrite was investigated by adding 10-50 pg of nitrite ions to 10 ml of absorbing solution containing 30 pg of sulphite and the absorbances were measured as explained under Procedure. It was observed that the intensity of the coloured complex was diminished in the presence of nitrate. To eliminate this error, 2 ml of 0.6% sulphamic acid were added to the absorbing solution before the addition of nitrite and sulphite. The colour was developed and measured and the absorbance was found to be slightly lower than the expected value.The error was reduced by introducing a separate trap containing 0.6% sulphamic acid instead of adding sulphamic acid to the absorbing solution as suggested by Paul and Gupta. 17 The sulphide ion, which is also present in the atmosphere mainly as hydrogen sulphide, interferes in the proposed method and to study its effect,l6 yg of sulphide ion were added to 10 ml of 0.4% morpholine solution containing 18 pg of sulphur dioxide. These amounts, if present in 40 1 of air, correspond to 0.16 p.p.m. of sulphur dioxide and 0.28 p.p.m. of sulphide. The colour was developed as described under Procedure. It was found that the sulphide ion enhanced the absorbance; the content of sulphur dioxide was found to be 0.24 p.p.m.compared with the actual value of 0.16 p.p.m. The actual concentration of hydrogen sulphide in the environment is generally too low to cause any significant variation in the sulphur dioxide value, The influence of metallic ions on this method was studied by adding 10-50 yg of Cu(II), Pb(II), Mn(II), Fe(III), Cr(II1) and V(V) to 10 ml of absorbing solution containing sulphite before the colour development. It was observed that 20 pg of Cu(II), 50 pg of Pb(II), 50 pg of Fe(II1) and 50 pg of V(V) did not interfere. Mn(I1) interfered even when only 10 pg were added. Although it is not expected that the metallic impurity in the atmosphere is as high as the concentrations tested in this work, we have studied the effects of higher concentrations of metallic ions to simulate possible industrial situations.The addition of the sodium salt of ethylenediaminetetraacetic acid (2 ml of 0.066%) masks 100 pg of metallic ions without affecting the determination of sulphur dioxide.ANALYST, FEBRUARY 1986, VOL. 111 Morpholine Concentration in the Atmosphere The threshold limit of morpholine in air is 20 p.p.m.19 When air is passed through morpholine during sampling it is expected that the morpholine vapour will contaminate the atmosphere. Hence, it is essential to study the loss of morpholine from 10 ml of 0.4% morpholine, being used as the absorbing reagent in the determination of atmospheric sulphur dioxide. A 40-1 volume of air was passed through 10 ml of 0.4% morpholine at a rate of 0.5 1 min-1.The morpholine content in the solution was determined by adding an excess of 0.1 N hydrochloric acid ( 5 ml) and titrating the excess of acid with 0.02 M sodium hydroxide solution using methyl red as an indicator. The difference in the morpholine content between the two samples, one of them through which 40 1 of air was passed and the other which was not subjected to bubbling of air, was 1 mg, which corresponds to 3.7 p.p.m. in air. When a similar experiment was conducted by passing 375 1 of air, the loss of morpholine was found to be 2 mg, which corresponds to 1.4 p.p.m. of morpholine in air. These results indicate clearly that the loss of morpholine during the sampling process is very much below the threshold limit and poses no hazardous effects. 191 References Conclusion Sulphur dioxide - sulphite is trapped in morpholine solution to give an adduct that, when treated with p-rosaniline hydro- chloride and formaldehyde solutions, yields a violet colour with an absorption maximum at 560 nm.This investigation avoids the use of poisonous mercury(I1) salts and is as sensitive as the West - Gaeke method. A sulphur dioxide content as low as 0.025 pg ml-1 in solution can be determined by this method. Morpholine is very effective as an absorbent for sulphur dioxide as the adduct formed is stable. The morpholine concentration passing into the atmosphere during the sampling process is much below the threshold limit. The authors are grateful to Dr. A. P. Mitra, Director, National Physical Laboratory, for his encouragement throughout this investigation. 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Marr, I. L., and Cresser, M. S., “Environmental Chemical Analvsis,” Thomson Litho. East Kilbride. 1983, p. 250. Environ. Health, Sharma,’H. C., and Rao,’D. N., Indian’J. 1983, 25, 169. Hize, W. K., Kippenberger, D. J . , and € Microchem. J., 1973, 20, 43. Humphrey, R. E., Ingram, G. S., and Microchem. J., 1982, 27, 351. umphrey, R. E., Crump, D. K., Okutani, T., and Utsumi, S., Bull. Chem. SOC. Jpn., 1967,40, 1386. Ramakrishna, T. V., and Balasubramaniyan, N. B., Indian J. Chem., 1982,21,217. Raman, V., Singh, M., and Parashar, D. C., Microchem. J . , in the press. West, P. W., and Gaeke, G. C., Anal. Chem., 1956,28,1816. Bhatt, A., and Gupta, V. K., Analyst, 1983, 108, 374. Christopher, R. F., and Barry, D., Environ. Sci. Technol., 1982, 16, 62. DasGupta, P. K., Decesare, K., and Ullrey, J. C., Anal. Chem., 1980, 52, 1912. Grundnes, G., and Christian, S. D., Acta Chem. Scand., 1969, 23, 3583. Harris, P. K., and Spragg, R. A., Org. Magn. Reson., 1969, 1, 329. Kanamueller, J. M., J . Znorg. Nucl. Chem., 1971, 33, 4051. Sweeney, N. P., and Thom, K. F., U.S. Pat., 3943 146; Chem. Abstr., 1976, 85, 5637c. Yamaguchi, T. , Kawasaki, T., and Ernori, S., Kobunshi Kagaku, 1971, 28, 336; Chem. Abstr., 1971, 75, 152190~. Paul, K. R., and Gupta, V. K., Atmos. Environ., 1983, 17, 1773. Pagnotto, L. D., “Gas Vapor Sample Collectors,” in “Air Sampling Instruments For Evaluation of Atmospheric Contam- inants,” Fifth Edition, American Conference of Governmental Industrial Hygienists, Cincinnati, OH, 1978, R7. Hawley, G. G., “Condensed Chemical Dictionary,” Ninth Edition, Van Nostrand Reinhold, New York, 1977, p. 591. Paper A51168 Received May 8th, 1985 Accevted Sevtember 16th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100189

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 193 Kinetic = Catalytic Determination of Cobalt by Oxidation of Pyrogallol Red by Hydrogen Peroxide M. Llobat-Estelles, A. Sevillano-Cabeza and J. Medina-Escriche Department of Analytical Chemistry, Faculty of Chemical Sciences, Valencia University, Bur-asot (Valencia), Spain A kinetic - spectrophotometric method for the determination of trace amounts of cobalt(l1) based on its catalytic effect on the hydrogen peroxide oxidation of Pyrogallol Red (PGR) is proposed. Under optimum conditions, cobalt can be determined in the range 0.20-13.00 ng ml-1. The limit of detection is 0.16 ng ml-1 of Co and the limit of quantitation is 0.54 ng ml-1 of Co. The precision and accuracy of the method using a fixed time of 5 min were also determined. The influence of 18 foreign species is discussed.Keywords: Cobalt determination; catalytic oxidation; Pyrogallol Red; hydrogen peroxide; kinetic - spectrophotometric method The catalytic action of cobalt may be observed in the oxidation of various substituted phenols with hydrogen peroxide. This effect is especially great if the phenols contain hydroxy groups in the ortho position,' and some cobalt-catalysed hydrogen peroxide reactions, such as those with alizarin,2 9-phenyl- 3,4,7-trihydroxyfluorone,3 Tiron,475 haematoxylin,6 cate- chol7-9 and 4-(4-nitrophenylazo)catechol,~0 have been des- cribed. There are some papers, however, that describe reactions with meta-substituted phenols such as 4-(4- nitropheny1azo)resorcinolll and with para-substituted phe- nols such as Alizarin Red S.12J3 The catalytic effect of cobalt on the oxidation of Pyrocate- chol Violet14 and Bromopyrogallol Red,lS dyes of the tri- phenylmethane series, have been studied.In relation to Bromopyrogallol Red, it is possible to carry out the determina- tion of the catalyst in the range 7.76 x 10-10-38.80 x 10-10 g cm-3 of Co in alkaline media. We earlier carried out a kinetic study of the uncatalysed reaction between Pyrogallol Red (PGR) and hydrogen peroxide and the cobalt(I1)-catalysed hydrogen peroxide oxidation of PGR.16 This paper reports the analytical applica- tion of the reaction. Experimental Reagents All reagents used were of analytical-reagent grade unless specified otherwise. CobaZt(1l) stock solution, 500 mg 1-1. A 1.247-g amount of cobalt(I1) nitrate [ C O ( N O ~ ) ~ .~ H ~ O , Merck] was dissolved in distilled water and diluted to 1 1. The solution was standar- dised with EDTA. A 50 ng ml-1 working solution was prepared by appropriate dilution of the stock solution. Hydrogen peroxide stock solution, 30% mlV. This was standardised by permanganate titration. Buffer solution, 0.1 M. Sodium tetraborate - 0.02 M hydro- chloric acid, pH 8. Pyrogallol Red solution, 8 x 10-4 M. Dissolve 0.160 g of PGR (Merck) in 500 ml of methanol. This solution is stable for at least 1 month. I Apparatus All spectrophotometric measurements were made on a Shimadzu UV-240 spectrophotometer with an OPI-2 coupled optional program unit, A Crison-501 pH meter, equipped with a Metrohm EA-121 Ag - AgCl electrode system, was used for measuring the pH of the solutions.The pH should be measured to an accuracy of ItO.01 pH unit. A Frigedor Selecta-396 refrigerator unit for use in a water-bath and a Thermotronic Selecta-389 immersion thermostat capable of maintaining the temperature to within k0.05 "C were also used. All the solutions were previously heated to the working temperature (25 It 0.05OC) in a thermostat and this temperature was maintained in the reaction cell during the experiment. Results and Discussion Optimum Conditions for the Cobalt-catalysed Reaction The kinetic study of the Co(I1)-catalysed hydrogen peroxide oxidation of PGR in alkaline medium (pH 8.3) resulted in the following rate equation because the uncatalysed reaction does not take place and the catalysed reaction is zero order with respect to PGR16: The optimum conditions should be selected such that the maximum sensitivity and the largest linear range are obtained together with the maximum correlation coefficient and precision.In order to find these optimum working conditions, the influence of reagent concentration was initially studied. The procedure and the order of addition of reagents have already been described.16 The kinetic data obtained for the dependence of the hydrogen peroxide concentration on the initial reaction rate show that the analytical sensitivity of the catalysed reaction increases with increasing hydrogen peroxide concentration up to 14 X 10-2 M. The reproducibility, however, is not good, possibly owing to the decomposition of hydrogen peroxide, which takes place to a measurable extent. Therefore, 3.5 X 10-2 M H202 was chosen as the recommended concentration as a compromise between reproducibility and sensitivity.Borate has an inhibitory effect on the initial reaction rate. This effect may be caused by the formation of a complex between the dye and borate and also by the cobalt-complexing ability of borate. There is also a linear relationship between absorbance (measured at a fixed time of 5 min, and at the 515-nm maximum wavelength of PGR at pH 8.3) against borate concentration in the range 2 X 10-2-8 X 10-2 M (y = 0.531 + 1.86cBdO72-; r = 0.998). The optimum pH is 8.3, and to achieve this the reaction system must be buffered as effectively as possible. It was found that a borate concentra- tion of 4 x 10-2 M is the best compromise between this purpose and the inhibitory effect.194 ANALYST, FEBRUARY 1986, VOL.111 Table 1. as follows: PGR, 3.3 x 10-2 M; pH, 8.3; I , 0.1 M; h = 515 nm; and temperature, 25 "C Calibration graphs for the fixed-time method. Conditions M; H202, 3.5 x 10-2 M; Na2B407, 4 x Time/ Calibration graphs min (0-13.0 ng ml-1 of Co) r 1 .o A = 0.7306 - 0.0119~ -0.992 3.0 A = 0.7277 - 0.0221~ -0.9989 5.0 A = 0.7274 - 0.0302~ -0.9998 10.0 A = 0.7256 - 0.0452~ -0.9988 15.0 A = 0.7167 - 0.0539~ -0.994 Table 2. Precision data for the method. Equation of calibration graph: y = 0.7274 - 0.0302 c ( r = -0.9998). The following conditions were used: PGR, 3.3 x M; H202, 3.5 X 10-2 M Na2B407, 4 X 10-2 M; pH, 8.3; I , 0.1 M; h = 515 nm; and temperature, 25 "C True concentration Mean concentration Standard of cobalt(II)/ of cobalt(I1) found/ deviation/ ng ml-1 ng ml-1 ng ml-1 0.40 1.00 2.00 4.00 8.00 12.00 0.44 0.07 0.99 0.08 2.05 0.09 4.1 0.14 8.0 0.14 11.99 0.09 No change in the initial rate of reaction was detected with variation in PGR concentration in the range 1.6 x 10-5-4.8 x 10-5 M.From the analytical point of view, a PGR concentra- tion should be used that provides an absorbance in the range of minimum photometric error, and for this purpose 3 x 10-5 M PGR was chosen as the most suitable concentration. The analytical sensitivity of the catalysed reaction (i. e. , the slope of the calibration graph) was found to increase with increasing temperature; this would suggest that the tempera- ture should be raised to 40 "C or even higher in order to optimise the method. With increasing temperature, however, the standard deviation increases. In addition , no dependence on temperature between 15 and 25 "C was noticeable for the uncatalysed reaction, but the initial reaction rate increased with increasing temperature in the range 30-40 "C.Therefore, 25 "C was chosen, as a compromise between precision and sensitivity (i.e., the slope of the calibration graph and the highest ratio of the catalysed and uncatalysed reaction rates). To summarise, the best experimental conditions for the catalytic determination of cobalt are H202 3.5 x 10-2 M, Na2B407 4 x 10-2 M, PGR 3 x 10-5 M, pH 8.3 and temperature 25 "C. Calibration Graphs To obtain the calibration graphs, four methods were used: fixed-time, fixed-concentration, initial-rate and rate-constant methods.The best method for our catalysed reaction was chosen on criteria of sensitivity [i.e., the slope ( S ) of the calibration graph] , linear range and correlation coefficient ( r ) . Fixed-time method The calibration graphs of absorbance versus cobalt concentra- tion at times of 1, 3, 5 , 10 and 15 min are shown in Table 1. From these results, it can be seen that the slope increases with time. The best correlation coefficient was obtained for a fixed time of 5 min and this was chosen as the most suitable measuring time. Fixed-concentration method The reciprocal of the time needed for the absorbance to decrease to 0.5500 (about 75% of the initial absorbance of the uncatalysed reaction) was plotted against cobalt concentration in the range 2.0-10.0 ng ml-1. This calibration, graph has the following equation: llt = -0.0014 + 0.0008~ with r = 0.997, where the time was measured in seconds.Initial- rate method A plot of the reaction rate (first 5 min) versus cobalt concentration yields the equation v = 0.0008 + 0.0055~ with r = 0.9996. Rate-constant method The plots of log A versus time (first 5 min) for cobalt concentrations in the range 0.2-13.0 ng ml-1 were straight lines. From the slope ( K ) plotted against cobalt concentration the equation K = -0.0015 + 0.0047~ with r = 0.995 was obtained. From the results obtained above, it can be inferred that the fixed-time method is the best as a better sensitivity, linear range and correlation coefficient were obtained.Limit of Detection and Limit of Quantitation The theoretical limit of detection (cL* = K&,/s)'7'18 for a numerical factor Kd = 3 (confidence level) is 0.16 ng ml-1 of Co at a fixed time of 5 min. The standard deviation (sb) for 10 independent measurements with 3 X 10-5 M PGR, 3.5 x 10-2 M H202 and 4 x 10-2 M Na2B407 (uncatalysed reaction) is 1.62 X 10-3 absorbance unit. The slope of the calibration graph is 0.0302 absorbance unit (ml ng-1 of Co). The experimental limit of detection is 0.20 ng ml-l of Co. The limit of quantitation ( C ~ Q = Kqsb/S)18 for a numerical factor Kq = 10 (confidence level) is 0.54 ng ml-1 of Co. The regions of analyte measurement are as follows: analyte not detected, <0.16 ng ml-1 of Co; region of detection, 0.16-0.54 ng ml-1 of Co; and region of quantitation, > 0.54 ng ml-1 of c o .Precision and Accuracy A study of the precision was performed by carrying out 10 independent measurements on solutions of various concentra- tions of cobalt(I1) and fixed concentrations of PGR, H202 and Na2B407. The cobalt concentration was calculated by substi- tuting the absorbance values, at a fixed time of 5 min, into the corresponding equation for the calibration graph. The results are shown in Table 2, the relative standard deviation (s,) being in the range 15.9-0.7% for a cobalt concentration of 0.40- 12.00 ng ml-1. The accuracyl9-22 was studied in the range 0.40-12.00 ng ml-1 of Co. The homogeneity of the variances of the analysed samples was confirmed by application of Bartlett's test and Hartley's test.The linear regression of the values obtained for-each analysis of each sample and the correspond- ing real values were obtained. The statistical t-test was applied to the study of the slope and the intercept of the straight line was obtained (Table 3). From this study we conclude that the proposed method does not present a constant-type error (a slope equal to unity) and it does not need a blank correction (an intercept equal to zero). Selectivity The effect of metal ions on the cobalt-catalysed reaction is shown in Figs. 1 and 2, where AA is the difference in absorbance between the cobalt-catalysed reaction and the cobalt-catalysed - possible interferent species reaction, at a fixed time of 5 min. From the results, the interfering processes may be classified as follows: (a) inhibition of the catalytic activity of cobalt in the presence of Cr(III), Ce(IV), Fe(III), Cu(II), Zn(II), Cd(I1) and Cr(V1); (b) acceleration of the catalytic effect, inANALYST, FEBRUARY 1986, VOL.111 195 I I I 8 4 I I Table 3. Statistical t-test applied to the study of the intercept and the slope t-Test (intercept) [-Test (slope) Equation of r tl-or/2 = t0.975 straight line S2Y ,x s a texp sb texp v = N - 2 y = 0.04 + 0.997~ 0.9997 0.0488 0.042 1.074 6.84 X -0.438 58 40 2.021 60 2.000 9 -0.02 - -0.04 - I I I 1 I 0.2 0.6 1.0 1.4 1.8 2.2 Metal ion concentrationipg mi-’ Fig. 1. Influence of metal ions on the kinetic - catalytic determina- tion of cobalt. 1, Pb(I1) 550; 2, Cr(1II) 300; 3, Ce(1V) 300; 4, Mn(I1) 150; and 5, Fe(II1) 100 (the numbers given after each element refer to maximum tolerable ratios of foreign metal ion to analyte concentra- tion).Conditions: PGR, 3.0 x 10-5 M; H202, 3.5 x 1 0 - 2 ~ ; Na2B407, 4 x 10-2~; Co, 2 ng ml-l; pH, 8.3; h = 515 nm; and temperature, 25 “C the presence of Pb(II), Mn(II), Mn(VII), V(V) and Ni(I1); and (c) lO4ng ml-1 of Po43-, Br03-, of C104-, C103-, 104- and 103- do not interfere. A species is considered to cause interference when a change of more than twice the standard deviation of the absorbance value of the catalysed reaction (s = 4.3 X 10-3 absorbance units, calculated from 10 indepen- dent measurements with a cobalt concentration of 2 ng ml-1) is obtained. Recommended Procedure Samples containing between 5 and 325 ng of cobalt(I1) were placed in 25-ml calibrated flasks, 10 ml of 0.1 M sodium tetraborate - 0.2 M hydrochloric acid buffer solution and 0.1 ml of 30% mlV hydrogen peroxide were added and the contents diluted to 20 ml with distilled water.This solution was shaken gently while 1 ml of 8 x 10-4 M PGR was added. The stop-watch was turned on when the last drop had fallen and the solution was diluted to the mark with distilled water. The absorbance was measured at 515 nm at 25 “C against water as a reference blank, for a fixed time of 5 min. The cobalt concentration was calculated from the corresponding equation for the calibration graph. References Otto, M., Muller, H., and Werner, G., Talanta, 1978,25,123. Chistyakov, N., Sb. Nauchn, Tr. Ivanov. Gos. Med. Inst., 1965, 31, 65. Popa, Gr., and Costache, D., Rev.Roum. Chim., 1967, 12, 963. Kucharkowski, R., and Doge, H. G., Fresenius 2. Anal. Chem., 1968,238,241. I I 0.08 1 P Metal ion concentration/pg ml-1 Fig. 2. Influence of metal ions on the kinetic - catalytic determi- nation of cobalt. 1, Cu(1I) 50; 2,Zn(II) 40; 3, Mn(VI1) 28; 4, V(V) 18; 5, Cd(I1) 12; 6, Cr(V1) 5; and 7, Ni(I1) 4 (the numbers given after each element reter to maximum tolerable ratios of foreign metal ion to analyte concentration). Conditions: PGR, 3.0 X 10-5 M; Hz02, 3.5 X 10-2 M Na2B407, 4 x 10-2 M; Co, 2 ng ml-1; pH, 8.3; h = 515 nm; and temperature, 25 “C 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Otto, M., Rentsch, J., and Werner, G., Anal. Chim. Acta, 1983, 147,267. Costache, D., Rev. Roum.Chim., 1970, 15, 1181. Prik, G. A,, and Orlova, M. N., Uch. Zap. Vladirnir. Gos. Pedagog. Inst., 1970,22,48. Kreingol’d, S. U., Sosenkova, L. I., and Vzorova, I. F., Metody Anal. Kontrol. Proizvod. Khim. Prom., 1976, No. 2, 38; Anal. Abstr., 1977, 22, 382. Alexiev, A. A., and Angelova, M. G., Mikrochim. Acta, 1980, 11, 187. Blazys, I., Paeda, R., and Jurevicius, R., Nauchn. Tr. Vyssh. Ucheb. Zaved. Lit. SSR, Khim. Khim. Tekhnol., 1971. 13,35. Costache, D., and Junie, P., Rev. Roum. Chirn., 1971,16,597. Vershinin, V. I., Chuiko, V. T., and Reznik, B. E., Zh. Anal. Khim., 1971, 26, 1710. Reznik, B. E., Chuiko, V. T., and Vershinin, V. I., Zh. Anal. Khim., 1972, 27, 395. Popa, Gr., and Costache, D., An. Univ. Bucuresti., Ser. Stiint. nut. Chim., 1968, 17, 21. Costache, D., and Popa, Gr., Rev. Rourn Chirn., 1970, 15, 1349. Sevillano-Cabeza, A., Llobat-Estelles, M., and Medina- Escriche, J., Analyst, 1985, 110, 1333. IUPAC, “Compendium of Analytical Nomenclature,” Per- gamon Press, Oxford, 1978. ACS Committee on Environmental Improvement, Anal. Chern., 1980,52, 2242. Mandel, J., and Linnig, F. J., Anal. Chem., 1957, 29, 743. Commissariat 1’Energie Atomique, “Statistique AppliquCe 1’Exploitation de Mesures,” Masson, Paris, 1978. de la Guardia, M., Salvador, A., and Berenguer, V., An. Quirn., 1981, 77, 129. de la Guardia, M., Salvador, A., and Berenguer, V., paper presented at the 5” Encontro Anual da Sociedade Portuguesa da Quimica, 1982. Paper A51180 Received May 17th, 1985 Accepted September 2nd, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100193

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 197 Effect of Anion-exchange Resin on the Formation of Iron(ll1) - Tiron Corn plexes Mohamed M. A. Shriadah" and Kunio Ohzekit Department of Chemistry, Faculty of Science, Hokkaido University, Sapporo, 060, Japan A comparison between iron(ll1) - Tiron complexes in solution and in the presence of a finely divided anion-exchange resin suspension (ARS) was studied. The effects of the ARS on the acid dissociation constants of Tiron (&, and Kaz) and on the stepwise stability constants (K,, K2 and K3) of iron(ll1) - Tiron complexes (FeL, FeL2 and FeL,) were investigated. Instead of the three complexes formed in solution at different pH values, only two complexes (FeLz and FeL3) were found to be fixed on the ARS. Although the ARS did not affect the acid dissociation constants of Tiron, the stability constants of FeL, and FeL3 were considerably improved.The stepwise stability constants were 20.0, 18.5 and 15.0, respectively, compared with reported values in solution of 20.4, 15.1 and 10.8, respectively. Keywords ; Iron( Ill) - Tiron complex formation; anion-exchange resin; stability constants The reactions of 1,2-dihydroxybenzene-3,5-disulphonate (Tiron) with metal ions are well documented and Tiron is well known as a colorimetric reagent for the determination of some metal ions. 1 Recently, molybdenum(V1) and vanadium(V) have been determined by densitometry after enrichment as the Tiron complex on a thin layer of anion-exchange resin.2.3 This paper presents an experimental comparison between the metal - ligand system in solution and the metal - ligand system in the presence of a finely divided anion-exchange resin suspension (ARS).As the reaction between iron(II1) and Tiron has been extensively studied, this system was selected in order to obtain a clearer insight into the effect of the anion-exchange resin on the complex formation between iron(II1) and Tiron. This study was undertaken with the objective of determining the three stepwise stability constants of iron(II1) - Tiron complexes in the presence of an ARS. Experimental Reagents Iron(ZII) standard solution, 10-3 M. Prepared by dissolving 0.4822 g of ammonium iron(II1) sulphate, FeNH4S04)2. 12H20, in 100 ml of water containing 5 ml of sulphuric acid and diluting to 1 1. Zron(ZIZ) standard solution, 1000 p.p.m. (1 mg ml-l).Prepared by dissolving 8.6340 g of ammonium iron(II1) sulphate in 100 rnl of water containing 5 ml of sulphuric acid and diluting to 1 1. A working solution, containing 10 pg ml-l, was prepared by appropriate dilution. Potassium hydrogen phthalate standard solution, 0.1 M. Prepared by dissolving 5.1056 g of potassium hydrogen phthalate, after drying for 2 h at 120 "C, in boiled water. The solution was diluted to 250 ml. Sodium hydroxide standard solution, 0.1 M. The solution was standardised against standard potassium hydrogen phtha- late solution. Potassium chloride solution, 1 M. Tiron solution, 0.1 M. The reagent was obtained from Dojindo (Japan) and used without further purification. A 0.1 M aqueous solution was prepared and standardised against standard sodium hydroxide solution.Working solutions of 10-2 and 10-3 M were prepared by dilution. Buffer solutions. The buffer solutions used for examining the effect of pH were 1 M sodium monochloroacetate - monochloroacetic acid, 1 M sodium acetate - acetic acid and 1 M ammonia - ammonium chloride solutions. * To whom correspondence should be addressed. t Present address: Department of Chemistry, Hirosaki University, Hirosaki, 036, Japan. Anion-exchange Resin This was macroreticular Amberlyst A-27 (Rhom and Haas) in the chloride form. The anion-exchange resin suspension (ARS) with particles smaller than 30 yrn was prepared according to the reported method.4 The exchange capacity of the suspension, as determined by conductimetric titration, was 7.12 pequiv.ml-1. A working suspension (3.56 pequiv. ml-l) was prepared by dilution. Apparatus A Shimadzu CS-920 Chromatoscanner was used for the measurement of the reflecting absorbance of the coloured complex in the thin layer of anion-exchange resin. A densitometer was used to linearise the convex calibration graph based on the Kubelka - Munk theory.5Jj A Shimadzu Type UV-240 UV - visible recording spectrophotometer was used for measuring the absorption spectra of iron(II1) - Tiron complexes in solution and in the resin phase. A Hiranuma Type RAT-11 recording autotitrator was used for carrying out the potentiometric titration of Tiron in the presence and absence of the ARS. A Hitachi - Horiba Type F7Lc pH meter was applied for the adjustment of pH. A Toyo KG-25 filter holder was used for the preparation of the thin layer by filtration under suction.A TM-1 filter-paper (0.65 pm) (Toyo) was used. Determination of Stepwise Stability Constants For the determination of K1 and K2, a 40-p1 portion of 10-3 M iron(II1) is placed in a dry 100-ml beaker followed by various amounts of 10-3 M Tiron solution. A 1.0-ml portion of acetate buffer solution and 2.0 ml of 1 M KCl solution are added. The sample volume is adjusted to 17.0 ml with water and then a 3.0-ml portion of ARS is added. The final pH of the solution is 3.67. The mixture is stirred for 10 min by means of a magnetic stirrer, then the resin is collected on a membrane filter under suction. A disc of coloured thin layer is formed that is 17 mm in diameter and about 0.2 mm in thickness.It is then wetted in a dipping solution containing 5 x 10-2 M acetate buffer solution (pH 3.7). The wet membrane filter holding the resin is placed on a white plastic plate in the densitometer, then the integrated absorbance is measured using the lineariser at 600 nm by scanning the thin layer over an area 24 mm wide and 30 mm long. The blank value is obtained by carrrying out the procedure without the addition of iron(II1). The third stability constant, K3, is determined in a similar manner except at pH 5.43 instead of pH 3.67. As the violet thin layer has been found to change to red during storage in the dipping solution, it is necessary to198 ANALYST, FEBRUARY 1986, VOL. 111 measure the absorbance immediately after the preparation of the thin layer.All the experiments were carried out at room temperature (20 k 1 "C). Results and Discussion Absorption Spectra The absorption spectra of iron(II1) - Tiron complexes were measured using the spectrophotometer for both the solution and the resin phase. The absorption spectra in solution were measured using a 1-cm cell against a reagent blank, whereas in the resin phase the thin layer of the sample and blank together were fixed on a glass plate fitted directly to the cell holder and the absorbance was measured with a slit width of 5 mm. Fig. 1 shows the absorption spectra of iron(II1) - Tiron complexes in the solution. As expected,' the reagent produced three complexes with iron(II1) ions depending on the pH of the solution; red (FeL3) at pH higher than 7, violet (FeL2) in slightly acidic solutions and blue (FeL) in more acidic solutions.The absorption maxima of each complex were 480, 560 and 670 nm, respectively. The absorption spectra of iron(II1) - Tiron complexes in the resin phase are shown in Fig. 2. Only two complexes are found to be fixed on the resin phase: red (FeL,) in basic and slightly acidic solutions (pH > 4.8) and violet (FeL2) in more acidic solutions. Effect of Stirring Time The influence of stirring time on the formation of the 1 : 2 complex was examined. The thin layers were prepared from solutions of pH 3.65, then wetted in a dipping solution containing 5 x 10-2 M acetate buffer solution (pH 3.65). The absorbance measurements were carried out at 600 nm immediately after filtration. The absorbance was found to be constant up to 20 min, which indicates that the reaction between iron(II1) and Tiron in the presence of the ARS is very fast and the reaction equilibrium can be attained quickly after the addition of the reagents.Effect of ARS on the Acid Dissociation Constants of Tiron The influence of the ARS on the acid dissociation constants of Tiron was studied by potentiometric titration. A 50-ml portion of the test solution containing 5.0 ml of 0.1 M Tiron was titrated with 0.1 M standard sodium hydroxide solution in the absence and presence of 40.0 ml of ARS (3.56 pequiv. ml-1). The ionic strength was kept constant (0.1). It was found that the titration curves in the presence and absence of the ARS coincide, which shows that the ARS has no effect on the acid dissociation constants of Tiron.The reported values of the acid dissociation constants of Tironl were used in the calculation of the three stepwise stability constants of iron(II1) - Tiron complexes. o.*51 ,,!, \ / '. \ \ I 400 500 600 700 800 Wavelengthhrn Fig. 1. Absorption spectra of iron(II1) - Tiron complexes in solution. Iron(III), 4.0 x 10-5 M; Tiron, 5.0 x 10-3 M; KCI, 0.1 M. pH: I, 2.01; 11, 4.95; and 111, 7.70 Calculation of K1 and Kz The total concentration of Tiron is given by CL = [L] + [HL] + [H2L] + [FeL] + 2[FeL2] + 3[FeL3] (1) Introducing the acid dissociation constant of Tiron, the free ligand concentration is expressed by CL - ([FeL] + 2[FeL2] + 3[FeL3]) * * (2) [HI + [HI2 1+- Ka2 KaiKa2 where Ka1 and Ka2 are 10-7.66 and 10-12.6, respectively.1 Although it has been found that almost all of the free Tiron is fixed on the resin phase, for simple calculation it is considered as if it is in the solution.At pH 3.67, as the formation of FeL3 is assumed to be negligible so the molar fraction of FeL2 is given by where CFe is the total concentration of iron(III), K1 and K2 are the stability constants of FeL and FeL2 and Kh is the hydrolysis constant of iron(III), reported as7 Kh = [FeOH][H]/[Fe] = 10-2.63 (4) As it was found that FeL could not be fixed on the resin phase, the colour of the thin layer is assumed to be due only to FeL2. The absorbance of the thin layer may be proportional to the concentration of FeL2 in the resin phase, that is, A = kFeL2[FeL21r (5) where ~ F ~ L ~ is the proportionality constant and the subscript r denotes the resin phase.As the distribution coefficient of F e k is given by the absorbance of the thin layer is proportional to the concentration of FeL2 in the solution: (7) At the maximum colour formation the above equation can be written as As shown in Fig. 3, the absorbance of the thin layer prepared from the solution containing 40 nmol of iron(II1) increased with increasing amount of Tiron and reached a constant and 0.5 Q, C m f! z n a 0.25 400 500 600 Wavelengthlnrn 700 Fig. 2. Absorption spectra of iron(II1) - Tiron complexes in the presence of the ARS. Iron(III), 2.0 pg; Tiron, 5.0 x M; buffer solution, 5 X 10-2 M; KCI, 0.1 M; ARS (3.56 pequiv. ml-l), 3.0 ml; sample volume, 20.0 ml. pH: I, 2.90; 11, 3.67; and 111, 5.43ANALYST, FEBRUARY 1986, VOL.111 199 maximum value. When the thin layer of maximum colour was wetted with the a’cetate buffer solution of pH 5.4, the colour changed as a result of the formation of FeL3. The absorbance of the resulting thin layer was compared with that of the thin layer that had the full colour produced by 40 nmol of iron(II1) as FeL3. The value was found to be 0.97, which indicates that . . . . * - (9) $2 = O.97A/Amax, (10) [FeLzlrnax. = o.97 CFe From equations (3), (7), (8) and (9), we obtain At pH 3.67, as the concentration of FeL3 is assumed to be trivial, the concentration of free Tiron is given from equation [L] = 10-12.92 {C, - ([FeL] + 2,[FeL2])} (11) By introducing the numerical values of Kh and pH into equation (3) and rearranging we obtain (2) as If [L] is properly calculated from equation (11) and the left-hand side of equation (12) is plotted against [L], a straight line is obtained.The results are shown in Fig. 4. From the slope of the line and the intercept, K1 and K2 were obtained. The logarithmic values of the stability constants K1 and K2 were found to be 20.0 and 18.5, respectively. Determination of K3 The experiments were carried out at pH 5.43, where the predominant species of iron(II1) can be assumed to be FeL2 and FeL3. It can be predicted that with increasing amount of Tiron, first the concentration of FeL2 will increase and the I I 0.25 0.50 0.75 1 .o +- P O 0 Added amount of Tiron/mI Y T C - Fig. 3. Effect of amount of Tiron on the absorbance of the iron(II1) - Tiron complex at pH 3.67.Iron(III), 2.0 x 10-6 M; buffer solution, 5 X 10-2 M; KCl, 0.1 M; ARS (3.56 pequiv. ml-l), 3.0 ml; sample volume, 20.0 ml. The distribution percentage was calculated accord- ing to equation (10) 3 40 i .- Oj 20 - w I , I 0 10 20 i L / i I O - ’ g Plots according to equation (12) for the determination of K , Fig. 4. and K2 absorbance also increases, then with further addition of Tiron FeL3 begins to be formed and the absorbance begins to decrease because the absorption coefficient of FeL3 is lower than that of FeL2 at 600 nm, as shown in Fig. 2. This prediction was supported by the results shown in Fig. 5. The absorbance was found to increase with increasing amount of Tiron; to reach a maximum and then to decrease to a constant value. The absorbance can be written as A = ~ ’ F ~ L & ’ F ~ L ~ [ F ~ L ~ I k‘FeL3d’FeL3[FeL3] * * (13) where k’ and d’ denote the proportionality constant and the distribution coefficient at pH 5.43, respectively. If it is assumed that at the maximum absorbance total iron is present as FeL2 and at the minimum absorbance it is present as FeL3, then Amax = k’FeL2d‘FeLzCFe (14) Amin = k’FeL3d’FeL3CFe (15) A = (Amax + Amin)/2 (16) and When [FeL2] = [FeL3] = cFe/2, the absorbance is given by and the corresponding concentration of free Tiron at pH 5.43 is given by equation (2) as (17) As K3 = [FeL3]/[FeL2][L], when the equation [FeL2] = [FeL3] holds, the formation constant K3 is given by K3 = l/[L] (18) From the results shown in Fig.5 , one obtains [L] = 10-15.0 and so the logarithmic value of the constant K3 is found to be 15.0.Effect of pH on the Distribution of Iron(II1) - Tiron Complexes at a Constant Tiron Concentration The stability constants of iron(II1) - Tiron complexes thus obtained in the presence of the ARS were K1 = 20.0, K2 = 18.5 and K3 = 15.0 at an ionic strength of ca. 0.1, the reported values1 in solution being 20.4, 15.1 and 10.8, respectively. To obtain a clearer picture of the effect of the ARS on the iron(II1) - Tiron complexes, the distribution of iron(II1) between various complexes as a function of pH, at a constant Tiron concentration , was calculated. The distribution curves of iron(II1) - Tiron complexes in solution are illustrated in Fig. 6(a) and in the presence of the ARS in Fig. 6(b). On comparing these curves, it is clear that the ARS has a marked effect on the distribution of the iron(II1) - Tiron complexes.The range of pH at which FeL, FeL2 and FeL3 occur in the solution was shifted to lower values in the presence of the ARS. Fig. 6 also shows that the distribution percentage of FeL, which was found not to be fixed on the anion-exchange resin, decreased in the presence of the ARS. The influence of I E I I UJ w Added amount of Tironiml - Fig. 5. Effect of amount of Tiron on the absorbance of the iron(II1) - Tiron complex at pH 5.43. Iron(III), 2.0 x 10-6 M; buffer solution, 5 X lo-* M; KCl, 0.1 M; ARS (3.56 pequiv. ml-I), 3.0 ml; sample volume, 20.0 ml200 8 6 ANALYST, FEBRUARY 1986, VOL. 111 / \ / \ ‘ II / \ 1‘. / v// / I 1 - 1 - /< 0 \. -_ I -- I r A S / 0 1 2 3 4 5 6 7 8 1 /---4-L ,’ 11 !i .\\-,, I \, I 4 I 1 I PH Fig.6. Effect of pH on the distribution of iron 111) - Tiron complexes calculated at a constant Tiron concentration:\L] + [HL] + [H,L] = 5.0 X M. ( a ) In solution; and ( b ) in the presence of the ARS. I, Fe3+; 11, FeOH; 111, FeL; IV, FeL2; and V, FeL3 L O 1 2 3 4 5 PH Fig. 7. Effect of pH on the absorbance of iron(II1) - Tiron complexes. Iron(III), 2.0 pg; Tiron, 5.0 X M; buffer solution, 5 X 10-2 M; KCI, 0.1 M; ARS (3.56 pequiv. ml-I), 3.0 ml; sample volume, 20.0 ml pH on the fixation of iron(II1) - Tiron complexes was examined as shown in Fig. 7. The fixation of iron(II1) as the Tiron complex on the ARS started at ca. pH 2 and the maximum fixation was observed at ca. pH 4, as predicted from the distribution curve of iron(II1) - Tiron complexes (Fig.6). Effect of Concentration of Tiron on theDistribution of Iron(II1) - Tiron Complexes at a Constant pH Fig. 8 shows the distribution of iron(II1) between various complexes in the presence and absence of the ARS as a function of the concentration of free Tiron at pH 3.67. Fig. 8(a) shows the distribution of iron(II1) - Tiron complexes in solution and Fig. 8(b) that in the presence of the ARS. Comparison of the two indicates that the presence of the ARS affects the distribution of the iron(II1) - Tiron complexes. In the presence of the ARS a smaller amount of Tiron than that in solution was found to be sufficient for the formation of the complexes. Calibration Graph The calibration graphs for the iron(II1) - Tiron complex in the presence of the ARS were constructed at pH 4.8 and 9.0.On using the lineariser, good linearity was observed up to 3.0 pg of iron(II1). It was also found that both lines had the same slope, which indicates that in the presence of the ARS the 1 : 3 complex is formed over a wide pH range. Effect of Potassium Chloride The effect of potassium chloride on the fixation of the 1 : 2 and 1 : 3 iron(II1) - Tiron complexes on the ARS was examined, as shown in Fig. 9. After fixation of iron(II1) as FeLz at pH 3.67 on the ARS from the solution containing various amounts of -Log[Ll Fig. 8. Effect of Tiron concentration on the distribution of iron(II1) - Tiron complexes calculated at a constant pH of 3.67. (a) In solution; and ( b ) in the presence of the ARS.I, Fe’+; 11, FeOH; 111, FeL; IV, FeL,; and V, FeL3 L - I I I 6 1 I I 0.1 0.2 0.3 0.4 0.5 0.6 0 Concentration of KCVM Fig. 9. Effect of potassium chloride on the fixation of FeL, and FeL3 complexes. Iron(III), 2.0 pg; Tiron, 5.0 x M; buffer solution, 5 X lo-, M; ARS (3.56 pequiv. ml-’), 3.0 ml; sample volume, 20.0 ml. I, FeL,; and 11, FeL, potassium chloride, the resulting thin layers of resin were wetted in the acetate buffer solution of pH 5.4 to convert FeL2 in the resin phase into FeL3, and then the absorbance was measured at 500 nm. The results were compared with those observed for FeL3 at pH 5.43. It was found that the fixation of iron(II1) as FeL2 was constant up to 0.13 M of potassium chloride, then it started to decrease sharply with increasing concentration of the salt. On the other hand, the fixation of the 1 : 3 complex was constant up to 0.3 M of potassium chloride, then it started to decrease gradually with increasing concentration of the salt. It is clear that the 1 : 3 complex, having nine negative charges,l is more strongly fixed on the resin than the 1 : 2 complex, which has only five negative charges. 1. 2. 3. 4. 5. 6. 7. References Cheng, K. L., Ueno, K . , and Imamura, T., “Handbook of Organic Analytical Reagents,” CRC Press, Boca Raton, FL, 1982, pp. 63-69. Shriadah, M. M. A., Kataoka, M., and Ohzeki, K., Analyst, 1985, 110, 125. Shriadah, M. M. A., and Ohzeki, K., Analyst, 1985, 110,677. Abe, M., Ohzeki, K., and Kambara, T., Bull. Chem. SOC. Jpn., 1978, 51, 1090. Yamamoto, H., Kurita, T., Suzuki, J., Hira, R., Nakano, K., and Makabe, H., Bunseki Kagaku, 1974, 23, 1016. Yamamoto, H., Kurita, T., Suzuki, J., Hira, R., Nakano, K., and Makabe, H., J . Chromatogr., 1976, 116, 29. Sillen, L. G., and Martell, E. E., “Stability Constants of Metal Ion Complexes,” Chemical Society, London, 1964, p. 54. Paper A511 10 Received March 22nd, 1985 Accepted August 28th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100197

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 Multi-component Quantitative Analysis of Fluorescent Obeying Beer's Law A. T. Rhys Williams and R. A. Spragg Perkin-Elmer Limited, Post Office Lane, Beaconsfield, Buckinghamshire HP9 I QA, UK 201 Mixtures Not The determination of mixtures by a linear least-squares regression analysis is a well established technique. However, this mathematical approach depends on the individual components obeying the Beer - Lambert law. Although this assumption is generally true in UV absorption, the analysis of mixtures by fluorescence spectrometry presents a number of problems. Distortion of the sample spectra can occur through a variety of reasons, including energy transfer. By using standards that are mixtures of the components, rather than pure components, compensation can be made for non-linear deviations.Keywords; Least-squares multi-component analysis; fluorimetry; non-linear deviation Multi-component analysis using regression techniques has received much attention during recent years. The growth in the use of the technique has been stimulated by new instrument designs using microprocessors and the advent of powerful desk-top computers using advanced software rou- tines. For the quantitative analysis of multi-component mixtures the most popular mathematical technique is that of least squares using standard matrix algebra.l-4 However, an assumption is made that the absorbance (or fluorescence) response is a function of the known components and that they obey the Beer - Lambert law. Several attempts have been made to accommodate these difficulties.Leggetts and Warner et aZ.6 proposed the use of non-negative least squares to overcome the problem of negative molar absorptivities. Haaland and Easterling7 and Brown et a/.* selected only the spectral regions that gave the best agreement to the calibration spectra. Over-determination of the regression by using more standards than there are components was found by Maris et aZ.9 to improve greatly the accuracy of most analyses. Osten and Kowalskilo used a modification of self-modeiling curve resolution as a means of testing and compensating for a background interferent. Least-squares curve fitting involves finding that combi- nation of a set of standard spectra which gives the best fit to a sample spectrum according to a least-squares criterion.These standard spectra have usually been those of the pure com- ponents, or spectra of the individual components derived from known mixtures. The calculated concentrations in the samples are then proportional to the contributions of these standard spectra to the best fit combination. An alternative is to use as the standards spectra from mixtures of known composition. The calculated concentration for each component is found by adding the contributions for that component from each standard. The approach has been successfully applied to the determi- nation of salicylic acid in aspirin.11 The low levels of salicylic acid present as a contaminant can easily be calculated even though very small differences occur between the fluorescence emission of acetylsalicylic acid and salicylic acid.An advantage of this method is that it allows a new approach to systems showing deviations from the Beer- Lambert law. The spectrum of such a system can be represented as the sum of the spectra of the pure components, with appropriate weightings, plus an additional spectral component. We can define the additional spectral component as the difference between the observed spectrum of the mixture and the spectrum calculated from the spectra of the pure components on the basis of the Beer - Lambert law. Previous workers12J3 have chosen as the additional spectral component the difference between the observed spectrum and the least-squares fit spectrum obtained from pure component spectra. Neither choice has any special physical significance.By using in the curve fitting procedure a number of standard spectra exceeding the number of known constituents in the sample, it is possible to take account of such additional spectral components. Accurate quantitative measurements for the constituents can be obtained in this way without any knowledge of the form or concentration dependence of the additional spectral components. 14 The only requirements are that the additional spectral components should not simply be linear combinations of the spectra of the known constituents and that the spectra of the mixtures used as standards should include these additional components. In this paper we shall show that this technique can be applied to samples whose emission spectra are distorted through energy transfer.Energy may be transferred from one component A (the donor), to another B (the acceptor), when the emission of A overlaps the absorption band of B. The reduction in the emission of A will generally not be the same for all wavelengths, but will depend on the absorption spectrum of B. The emission spectra will therefore not be a linear combination of the emission spectra of A and B with weightings proportional to their concentrations. It is possible to consider the deviations from linear behaviour as arising from an additional spectral component that is a combination of the emission of A and the absorption and emission of B, assuming that a constant proportion of the transferred energy is re-emitted by B. The example chosen is the energy transfer shown to occur between Tb3+ chelated to dipicolinic acid, Tb(DPA)3, and Rhodamine B.l5 Both the absorbance and fluorescence emission were directly proportional to their concentrations in the absence of each other.Quantitative analysis of mixtures by curve fitting using pure standards results in large errors. However, by using an additional standard, which is a mixture, there is an enormous improve- ment in accuracy. Experimental Instrumentation and Software A Perkin-Elmer Model LS-5 luminescence spectrometer fitted with a red-sensitive R 928 photomultiplier and equipped with an RS 232C serial interface was used for time-resolved measurements and all luminescence excitation and emission studies. Data were recorded using a Perkin-Elmer Model 7500202 ANALYST, FEBRUARY 1986, VOL.111 Professional computer using the PECLS-I11 applications software providing instrumental control, data manipulation and storage. Quantitative analyses were performed using the QUANT-I11 software, which can use up to 15 standard spectra to match the spectrum of the sample. Reagents All experiments were performed with analytical-reagent grade chemicals , except where stated otherwise. Water was distilled and passed through a 0.45-ym Millipore filter. Terbium(II1) chloride (TbC13.6H20), 99.9% , was obtained from Aldrich, Gillingham, UK, dipicolinic acid from Fluka, Buchs, Switzer- land, Rhodamine B from BDH Chemicals, Poole, UK, and Tris buffer (pH 8.0) from Sigma, Poole, UK. Procedure All reagents and experimental solutions were prepared as described by Thomas et al.15 Time-resolved fluorescence measurements were made using the Model LS-5 as described previously. 16 Fluorescence spectra were recorded with a delay time, fd, of 0.01 ms and a gate time, tg, of 8.0 ms. Emission spectra were recorded with a 5-nm spectral band pass and were corrected for instrumental response from 250 to 700 nm. I I 450 525 600 67 5 750 Wavelengthhm Fig. 1. (A) The corrected excitation and (B) the emission of Rhodamine B and (C) the corrected emission of Tb(DPA)3, all measured with a 5-nm spectral band pass at an excitation of 260 nm Results and Discussion The excited-state lifetime of Tb(DPA)3 in the absence of Rhodamine B was found to be 2.08 ms, with its emission spectrum overlapping the absorption spectrum of Rhodamine B (Fig.1). A series of mixtures were prepared containing terbium varying in concentration between 1.25 and 10.0 VM. The limit of 0.5 PM for the Rhodamine B concentration was chosen so as to keep the absorbance at 554 nm below 0.05 and to minimise the “inner filter” effect. A gate time of 8.0 ms was chosen so as to integrate the total emission from the terbium excited state. The effect of energy transfer on the emission spectrum of the sample mixture is shown in Fig. 2. Spectrum B is the actual emission of the mixture and spectrum A is the result of adding the two components mathematically. There is an increase in the Rhodamine B emission while at the same time there is a decrease in the terbium emission. The excited-state lifetime of the Tb(DPA)3 also decreased from 2.08 to 0.43 ms in the presence of 2 VM Rhodamine B.The latter also exhibited an emission lifetime virtually comparable to that of the Tb(DPA)3, indicating “triplet” to singlet energy transfer. The emission spectra were measured and quantitatively analysed using the QUANT-I11 software using either two pure stan- dards or with an additional standard that was a known mixture. The results are given in Table 1. 100 80 60 > v) Q) c 4- .- c - 40 20 450 525 600 675 7! 0 Wavelength/nrn Fig. 2. (A) Synthetic spectrum of the two components and (B) the corrected emission of a mixture containing 0.5 p~ Rhodamine B and 5.0 p . ~ Tb3+ Table 1. Analysis of Tb(DPA)3 - Rhodamine B solutions Caiculated/pM Initial concentrations/pM A* Tb(DPA)3 Rhodamine B 10.0 0.5 5.0 0.5 2.5 0.5 1.25 0.5 2.5 0.25 1.25 0.125 Tb(DPA)3 Rhodamine B 5.25 2.20 2.66 1.25 1.25 0.91 0.64 0.78 1.77 0.53 1.02 0.20 Tb(DPA)3 Rhodamine B 10.26 0.48 4.86 0.47 2.50 0.49 1.35 0.50 2.51 0.27 1.24 0.131 * A, Two standards, 2.5 p~ Tb(DPA)3 and 1.0 p~ Rhodamine B.t B, Three standards, as A plus 5.0 p . ~ Tb(DPA)3, 0.5 p~ Rhodamine B.ANALYST, FEBRUARY 1986, VOL. 111 203 Quantitative analysis using the spectra of the separate components results in gross errors in the relative concentra- tions. With a single additional standard that is a mixture there is a considerable improvement in accuracy, the errors being reduced on average by a factor of 20. By specifying that the mixture contains three components and setting the nominal concentrations of the three standards used in analysis B to 1.0, an estimate can be obtained of the contribution due to energy transfer.Normalising these values to the terbium concentration and plotting the results against the corresponding Rhodamine B concentration yields a linear relationship, indicating that the transferred energy is propor- tional to the Rhodamine B concentration up to 0.5 PM. Peak-area calculations on spectra for actual and synthetic mixtures indicate that there is an approximate 10% loss in the amount of energy transferred. Conclusion Accurate quantitative analysis of fluorescent mixtures using least-squares curve fitting can be achieved even when there are gross deviations from the Beer - Lambert law. By includ- ing a standard that is a known mixture, the errors that occur when separate components are used can be considerably reduced.References 1. Sternberg, J. C., Stillo, H. S . , and Schwendmeman, R. H., Anal. Chem., 1960,32, 84. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Weber, G., Nature (London), 1961, 190, 27. Ainsworth, S . , J. Phys. Chern., 1963, 67, 1613. Meites, L., Anal. Chirn. Acta, 1975, 74, 177. Leggett, D. J., Anal. Chem., 1977, 49, 276. Warner, I. M., Davidson, E. R., and Christian, G. D., Anal. Chem., 1977,49, 2155. Haaland, D. M., and Easterling, R. G.,Appl. Spectrosc., 1982, 36, 665. Brown, C. B., Lynch, P. F., Obremski, R. J . , and Lavery, D. S . , Anal. Chem., 1982, 54, 1472. Maris, M. A., Brown, C. W., and Lavery, D. S., Anal. Chern., 1983, 55, 1694. Osten, D. W., and Kowalski, B. P., Anal. Chem., 1985, 57, 908. Winfield, S. A., and Rhys Williams, A. T., J. Pharrn. Biorned. Anal., 1984, 2 , 561. Antoon, M. K., Koenig, J. H., and Koenig, J. L., Appl. Spectrosc., 1977, 31, 518. Ramana Rao, G., and Zerbi, G . , Appl. Spectrosc., 1984, 38, 795. Ford, M. A., and Spragg, R. A., in “Proceedings of the 1985 International Conference on Fourier and Computerised Infrared Spectroscopy, Ottawa,” SPIE, USA, to be published. 15. Thomas, D. D., Carlsen, W. F., and Stryer, L., Proc. Natl. Acad. Sci. USA, 1978, 75, 5746. 16. Rhys Williams, A. T., and Fuller, M. J., Computer Enhanced Spectrosc., 1983, 1, 145. Paper A51278 Received July 26th, 1985 Accepted August 22nd, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100201

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 205 Minimisation of Bilirubin Interference in the Determination Fluorescein Using First-derivative Synchronous Excitation FI uorescence Spectroscopy Frank V. Bright* and Linda B. McGownt Department of Chemistry, Oklahoma State University, Stillwater, OK 74078, USA The use of first-derivative synchronous excitation fluorescence is described for the minimisation of bilirubin interference in the determination of fluorescein. The limits of detection and the standard errorsof estimates are better using the synchronous first-derivative approach than with conventional subtraction of the bilirubin blank. Results are shown for the quantitation of fluorescein in the nanomole concentration range. Keywords: Fluorescein determination; first-derivative synchronous excitation fluorescence; interferent elimination; bilirubin interference The use of fluorescein-conjugated reagents has become widespread, especially for use in fluorescence immunoassay techniques.Many of the immunoassays are applied to the analysis of biological fluids, especially serum. However, the fluorimetric determination of fluorescein- labelled species in serum is subject to interference from high concentrations of serum bilirubin, such as in icteric sera of jaundiced patients.l.2 The interference is due to the broad fluorescence emission and excitation bands of bilirubin that entirely overlap those of fluorescein. Free bilirubin is very unstable in aqueous media and has a negligible quantum yield (lO-5).3 Bilirubin will, however, strongly bind to serum albumin, resulting in an increased quantum yield of 0.001.4 The extent of interference is determined by the relative fluorescence intensity contributions of fluorescein (F) and bilirubin (B) and, assuming a quantum yield of 0.90 for fluorescein,5 the relative bilirubin contribution will be 1% or more of the total fluorescence when CB 2 30CF (where Cis the concentration).Bilirubin interference is generally treated by simple subtrac- tion of the serum sample blank fluorescence from the total fluorescence signal (fluorescein + blank). The limits of detection (LOD) for the determination of fluorescein will then be limited by the relative magnitudes of the blank and the fluorescein contributions. At very high values of CB/CF this difference will be small relative to the signals causing an increase in the LOD.The coupling of synchronous excitation luminescence with derivative spectroscopy has been discussed previously.6 For example, second-derivative synchronous excitation lumines- cence was used for the analysis of mixtures of acenaphthene, biphenyl, chrysene, dibenzothiophene and phenol.7 The technique has also been used for the determination of cadmium using benzyl2-pyridyl ketone 2-quinolylhydrazone .8 The use of first-derivative synchronous excitation fluores- cence spectroscopy is described for the selective determina- tion of fluorescein in the presence of bilirubin. This approach offers a lower LOD and a wider range of linearity than is achieved using blank subtraction. Theory The theory and applications of synchronous excitation fluores- cence spectroscopy have been described in detail else- where,s13 as have those for derivative spectroscopy.14,15 Briefly, in synchronous excitation fluorescence the excitation * Present address: Department of Chemistry, University of t To whom correspondence should be directed. Indiana, Bloomington, IN 47405, USA. and emission monochromators are scanned simultaneously, maintaining a constant wavelength difference (Ah) between the two monochromators. The Ah is usually chosen to equal the difference between the 0-0 transitions for fluorescence excitation and emission. The result of synchronous excitation fluorescence scanning is a narrowing of the spectral bands owing to the synchronous multiplication of the simultaneously increasing and then decreasing fluorescence spectra as func- tions of excitation and emission wavelengths.The synchro- nous excitation fluorescence intensity ( I F ) takes the form where c is the analytical concentration of the analyte; k is a constant that contains various parameters (molar absorptivity, cell path length and instrumental parameters); Aex. and kern, are the excitation and emission wavelengths, respectively; and Ex and Em refer to the emission and excitation spectra, respec- tively. The coupling of synchronous luminescence with derivative techniques simply involves taking the derivative of the synchronous spectrum. This can be carried out in real time or after spectral acquisition.16 zF(hex.,hem.) = kc[Ex(hex.) Em(hex. -k Ah) . * (1) Experimental Materials Bilirubin was purchased from United States Biochemical Co.(Cat. No. 12110) and used without further purification. Human serum albumin was purchased from Sigma (HSA, Cat. No. A1887). A stock solution of 106 PM HSA was prepared by dissolving 0.7000 g of HSA in pH 7.00, 0.10 M phosphate buffer and diluting to 1.000 dl followed by sonication for 30 min. This solution was subsequently diluted 10-fold with the buffer to yield a 10.6 VM HSA working solution. A stock solution of 1.00 mM bilirubin was prepared in 1.00 mM HSA by dissolving 0.0059 g of bilirubin and 0.6700 g of HSA in the buffer and diluting to 0.100 dl followed by sonication, in the dark, for 30 min. Laboratory standards (Set 1) of bilirubin were prepared in 1.00 mM HSA by adding the appropriate volume of stock solution followed by dilution to 1.000 ml.A second series of bilirubin standards (Set 2) were purchased from Sigma (Cat. No. 550-11) and contained 2.2,9.9 and 15.0 mg dl-1 of bilirubin. Fluorescein was purchased as the sodium salt from Sigma (Cat. No. F-6377) and a 1.00 mM stock solution was prepared by dissolving 0.0367 g of fluorescein and diluting to 1.000 dl with buffer solution followed by sonication for 30 min. This stock solution was subsequently diluted 10 000-fold with 10.6 VM HSA. All bilirubin solutions were kept in the dark to minimise potential sample degrada- tion.206 All synchronous fluorescence excitation measurements were made using disposable polyethylene cuvettes (NSG Precision Cells, Inc.). ANALYST, FEBRUARY 1986, VOL. 111 Data Collection Synchronous fluorescence spectra were collected on an SLM 4800s (SLM-Aminco) spectrofluorimeter with a 450-W Xe arc source and PMT detection.All spectra were collected in the “10 average” mode in which an average value is obtained via integration and division over approximately 3 s. The spectro- fluorimeter was interfaced to an Apple 11+ microcomputer for spectral acquisition and subsequent calculation of the first- derivative spectra. Measurements were taken in a ratiometric mode to minimise effects from source output fluctuations. Fifteen synchronous spectra were recorded for each sample and blank solution and the average of the 15 spectra was used for the calculation of the derivative spectrum. The sample-chamber temperature was maintained at 20.0 k 0.1 “C using a Haake A81 temperature control unit. Monochromators were synchronously scanned from Aex, = 440 nm to 540 nm at 0.5-nm intervals (200 points) maintaining a Ah of 20 nm between the excitation and emission monochro- mators and requiring approximately 10 min per spectrum. All monochromator entrance and exit slits and the modulation tank chamber exit slit were set at 16 nm.Data Analysis The synchronous fluorescence excitation spectra were analy- sed by both blank subtraction and the first-derivative synchro- nous fluorescence excitation approaches. For the blank subtraction approach the synchronous spectra for fluorescein, bilirubin and mixtures of bilirubin and fluorescein were collected and the appropriate spectrum of bilirubin was subtracted from the spectrum of a mixture containing the same bilirubin concentration. The difference spectrum, correspond- ing to the fluorescein contribution, was used to determine the concentration of fluorescein from a calibration graph gener- ated from fluorescein standards.In the first-derivative synchronous excitation fluorescence approach, the first- derivative synchronous excitation spectra were numerically generated by the Apple 11+ and the differential intensities (dF/dh) for the mixtures were measured at the point at which the bilirubin first-derivative synchronous spectrum gave zero contribution (see below). Fluorescein concentrations were determined using a calibration graph generated with measure- ments of standard first-derivative synchronous spectra at the same zero bilirubin point.Results and Discussion The synchronous fluorescence spectra acquired at Ah = 20 nm for bilirubin, fluorescein and a mixture of the two species are shown in Fig. 1. Other AA values were also used, but the best results were achieved using 20 nm, which is very close to the 0-0 transition for fluorescein (30 nm). The bilirubin fluorescence spectrum began to broaden significantly above the 20-nm value. It is evident that one cannot determine fluorescein devoid of bilirubin contribution at any of the wavelengths shown in Fig. 1. The first derivatives of the synchronous spectra are shown in Fig. 2. The effects of the first-derivative transformation are dramatic and one feature is of special interest. The zero crossing point for bilirubin (481/501 nm) is very nearly at the maximum for the first-derivative spectrum of fluorescein, which is the optimum situation for fluorescein determination.Table 1 shows the composition of the fluorescein and bilirubin standards used in this study. Fifty-four mixtures were I I I I I I I I I 440 450 460 470 480 490 500 510 520 530 540 Excitation wavelength (Aem. = Aex. + 20 nm)/nrn Fig. 1. Synchronous fluorescence spectra (AA = 20 nm) for: A, 2.00 WM of bilirubin; B, 10 nM of fluorescein; and C, a mixture of 2.00 p~ of bilirubin and 10 nM of fluorescein, all in 10 VM HSA 100 I \-’ -100 I I I 1 I I I I I 440 450 460 470 480 490 500 510 520 530 540 Excitation wavelength (Aem, = Aex. + 20 nm)/nm Fig. 2. First derivatives of the synchronous spectra shownin Fig. 1: A, bilirubin; B, fluorescein; and C, mixture Table 1.Concentrations of standard solutions used in this study* Bilirubin/pM Fluorescein/nM Set I t Set 2$ 0.100 0.0977 3.67 0.200 0.244 1.65 0.398 0.489 2.51 0.990 0.977 1.96 1.47 3.23 1.95 * Analytical concentration in cuvettes containing 10.6 WM HSA. t Standards prepared from solid bilirubin (USB). $ Standards from Sigma. used for the fluorescein determinations, prepared using each fluorescein standard with every bilirubin standard listed in Table 1. The LOD, which is the concentration of fluorescein (C,) required to give a signal equal to the following: where Xave.,blank is the average value of the bilirubin blank signal and &,lank is the standard deviation of &e.,blank, determined as a function of bilirubin concentration at three wavelength pairs ( 4 8 ~ 0 1 , 495/515 and 500/520 nm).The CF = xave.,blank -k 3Sblank * * * (2)ANALYST, FEBRUARY 1986, VOL. 111 207 Table 2. LOD results for fluorescein using 15 synchronous excitation spectra LOD/nM Blank subtraction (non-derivative) Bilirubin/moll-l 481/501 nm* 4951515 nm* 5001520 nm * 9.77 x 10-*t 0.644 0.461 0.439 2.44 x 10-7t 0.896 0.531 0.533 4.89 x 10-7t 1.22 0.781 0.804 9.77 x 10-7t 2.07 1.42 1.56 1.47 x 10-6-t 2.54 1.97 2.00 1.95 x 10-9 2.99 2.21 2.54 3.67 X 10-6$ 1.02 0.694 0.661 1.65 X 10-6$ 2.65 2.04 2.15 2.51 x 10-6$ 3.25 2.57 2.90 r§ 0.9987 0.9981 0.9977 SEED 0.059 0.058 0.072 * Exci ta tion/emission wavelength pairs. t Standards prepared in our laboratory (USB). $ Standards from Sigma. 0 Correlation coefficients for LOD versus bilirubin plots (second-order polynomial fit).7 Standard error of estimate (nM). First derivative, 481/501 nm* 0.250 0.360 0.520 1.04 1.42 1.66 0.436 1.53 1.80 0.9973 0.052 standard deviation for the bilirubin blank increases as the bilirubin concentration increases, resulting in an increase in the LOD for the determination of fluorescein using blank subtraction. For the first-derivative work the dF/dh at the 481/501 nm wavelength pair was used and the fluorescein concentration determined from a calibration graph determined under the same set of conditions using the standards described in Table 1. Blank subtraction was not necessary as bilirubin does not contribute at the 481/501 nm wavelength pair under the first-derivative conditions. The limit of detection was also determined for this approach as a function of bilirubin concentration using the standard deviation for the blanks (bilirubin) at the various bilirubin concentrations (Table 1).Table 2 shows the LOD results obtained by the blank subtraction (non-derivative) method at the 481/501, 495/515 and 500620 nm wavelength pairs and those obtained by the first-derivative method at the 481/501 nm wavelength pair. The first-derivative approach gives significantly lower detec- tion limits for the determination of fluorescein in the presence of bilirubin, owing to the minimisation of bilirubin contribu- tion and therefore the imprecision associated with its contribu- tion under the first-derivative conditions. Table 2 also lists the standard errors of estimates (SEE) under the various condi- tions, which are a measure of the precision for a given determination method.The first-derivative synchronous exci- tation approach is again shown to be the better method, and would be worth the additional time required for the acquisi- tion of the synchronous spectrum in samples for which the minimisation of detection limits is of critical importance. The spectral acquisition time could be reduced to a negligible amount if a multi-channel detector was used. The authors acknowledge support of this work by the National Science Foundation (Grant No. CHE-8403759). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Chiamori, N., Henry, R. J., and Golub, 0. J., Clin. Chim. Acta, 1961, 8, 1. Bright, F. V., and McGown, L. B., Anal. Chim. Acta, 1984, 162, 275. Smith, D. S . , Hassan, M., and Nargessi, R. D., in Wehry, E. L., Editor, “Modern Fluorescence Spectroscopy,” Volume 3, Plenum Press, New York, 1981, Chapter 4. Chen, R. F., Biochem. Biophys., 1974, 160, 106. Weber, G., and Teale, F. W. J., Trans. Faraday SOC., 1957,53, 646. John, P., and Soutar, I., Anal. Chem., 1976,48,520. Sanchez, F. G., Navas, A., and Santiago, M., Anal. Chim. Acta, 1985, 167, 217. Vo-Dinh, T., in Wehry, E. L. Editor, “Modern Fluorescence Spectroscopy,” Volume 4, Plenum Press, New York, 1981, Chapter 5. Vo-Dinh, T., Gammage, R. B., Hawthorne, A. R., and Thorgate, J. H., Environ. Sci. Technol., 1978, 12, 1297. Lloyd, J. B. F., and Evett, I. W.,Anal. Chem., 1977,49,1711. Lloyd, J. B. F., Analyst, 1975, 100, 82. Vo-Dinh, T., Anal. Chem., 1978, 50, 396. Vo-Dinh, T., and Gammage, R. B., Anal. Chem., 1978, 50, 2054. Green, G. L., and O’Haver, T. C., Anal. Chem., 1974, 46, 2191. O’Haver, T. C., Anal. Chem., 1979, 51,91A. O’Haver, T. C., and Green, G. L., Anal. Chem., 1976,48,312. Paper A5/156 Received April 26th, 1985 Accepted August 27th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100205

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, FEBRUARY 1986, VOL. 111 209 Novel Aryl Oxalate Esters for Peroxyoxalate Chemiluminescence Reactions Kazuhiro Imai* Faculty of Pharmaceutical Sciences, University of Tokyo, 7-3- 1 Hongo, Bunkyo-ku, Tokyo 1 13, Japan and Hiroyoshi Nawa, Motoaki Tanaka and Hiroshi Ogata Wako Pure Chemical Industries, Ltd., 3- I0 Dosho-machi, Higashi-ku, Osaka 54 I , Japan Six oxalate esters of 2-nitro-4-alkoxycarbonylphenol and 2-alkoxycarbonyl-4-nitrophenol were synthesised for use in the peroxyoxalate chemiluminescence reactions that take place via several steps with 4-hydroxy-3-nitrobenzoic acid and 5-nitrosalicylic acid as starting materials, respectively. The addition of ethylene glycol to the alkoxy moiety enhanced the solubility in solvents such as acetone, acetonitrile and ethyl acetate.When reacted with hydrogen peroxide in the presence of a fluorescent compound, perylene, they gave similar chemiluminescence reaction curves to those of DNPO [bis(2,4-dinitrophenyl) oxalate] and TCPO [ bis( 2,4,6-t r ic h lo ro p hen y I ) oxa late]. Among them, bis[ 4-n it ro-2-( 3,6,9-t rioxadecyl oxyca r bo n y I ) p h en y I ] oxa- late was found to be the best for the high-sensitivity detection of both hydrogen peroxide and fluorescent compounds using the peroxyoxalate chemiluminescence reaction because of its extreme solubility in solvents such as acetone (767 mM) and acetonitrile (1010 mM). Keywords : A r yl oxala tes; bis[4-n itro-2- (3,6,9- trioxadec ylox yca rbon yl)p h en yl] oxa la te; ch em ilum inescence; fluorescent compounds; hydrogen peroxide Aryl oxalate esters and hydrogen peroxide generate chemi- luminescencel-3 in the presence of fluorescent compounds.This reaction (shown in Scheme 1) can be used for the detec- Aryl oxalate (I or Ill L J Dioxetane-3,4-dione 1: 81 C-C + Fluorophor - Fluorophor" + 2C02, 10-0 ] Fluorophor* -Light + Fluorophor Scheme 1. Proposed mechanism for the peroxyoxalate chemilu- minescence reaction tion of hydrogen peroxide produced in enzymic reaction^^.^ or enzyme immunoreactions.6 It has also been applied success- fully to the detection of fluorescent compounds on a thin-layer plate7 and in high-performance liquid chromatography (HPLC).g-15 In HPLC, its application to post-column reac- tions produced a sensitivity for the detection of the fluorescent compounds one to two orders of magnitude higher than did conventional light-induced fluorimetry; dansylated amino acids,g19 dansylated steroids,lo fluorescamine-labelled cate- cholamines,ll o-phthalaldehyde (0PA)-derivatised and nitro- benzofurazan (NBD)-labelled amines12 and polycyclic aro- matic hydrocarbonsl3-15 were detected at the femtomole level.For more sensitive detection using this reaction system, esters should be present in the reaction medium in larger amounts as the proposed active intermediate, dioxetane-l,2- dione, which excites the fluorescent compounds, can best be produced in the presence of large amounts of esters. However, bis(2,4,6-trichlorophenyl) oxalate (TCPO) and bis(2,4- dinitrophenyl) oxalate (DNPO), which are commonly used in the detection system for HPLC,S-16 dissolve in acetone and acetonitrile to give about 10 mM solutions, which might be suitable solvents for the post-column reaction in HPLC.s Therefore, we studied various oxalate esters that are more soluble than TCPO and DNPO in hydrophilic solvents.We selected as parent compounds 4-hydroxy(or 2-hydroxy)-3- nitro(or 5-nitro)benzoic acid (I11 or IV) (Fig. l), as the nitro group in the ortho- or para-position to the hydroxy group greatly withdraws the electrons to yield a higher reactivity. The other reason is that the nitro group does not produce quenching like the chloro moiety of 2,4,6-trichlorophenol, a hydrolysis product of TCP0.16 Therefore, bis(4-methoxy- carbonyl-2-nitrophenyl) oxalate (Ia) was first synthesised. Next, to produce the affinity of the oxalates to the hydrophilic solvents, the ethylene glycol moiety was attached to the carboxy group of the parent compounds to give Ib, Ic, IIb, IIc and IId (Fig.1). Synthesis To 4-hydroxy-3-nitrobenzoic acid(II1) (11.0 g, 60.1 mmol) in dry benzene (50 ml) containing dry pyridine (5 drops) was added dropwise thionyl chloride (9.20 g, 77.3 mmol) at 50- 60 "C. (Caution-Benzene is highly toxic and appropriate precautions should be taken.) After stirring for 4 h, the reaction mixture was condensed under vacuum to give an oily residue of 4-hydroxy-3-nitrobenzoyl chloride (IIIa),l7 12.0 g, yield 99.2%. IR (neat): 1750 cm-1 (-COC1) and 3250 cm-l ( Ar-OH) . The residue (IIIa, 12.0 g, 59.5 mmol) was dissolved in ethylene glycol monomethyl ether (22.8 g, 230 mmol) and heated at 70-75 "C for 2 h. The reaction mixture was evaporated under vacuum to give methoxyethyl4-hydroxy-3- nitrobenzoate (Vb) as yellow crystals, 14.1 g, yield 98.2%, m.p.49-51 "C. IR (KBr): 1725 cm-1 (CO-0) and 3250 cm-1 (Ar-OH). NMR (CDC13): 6 3.5 (3H, s, 0-CH3), 4.5 (2H, t, -COOCH2-), 10.8 (lH, s, -OH), 3.8 (2H, t, -CH20-) and 7.2-8.9 p.p.m. (3H, m, C6H3). uv (acetonitrile): h,,,, 240 nm (E = 3.90 x 104 1 mol-1 cm-1). Vb (4.80 g, 19.9 mmol) and triethylamine (2.20 g, 21.7 mmol) were dissolved in dry benzene (50 ml) and to this solution was added dropwise oxalyl chloride (1.13 g, 8.90 mmol) at 5-8 "C under a stream of nitrogen. After reaction for 3 h, the resulting precipitates were collected and dissolved in dry benzene. The filtrate was evaporated under vacuum to half the volume to yield pale yellow prisms, Ib (3.60 g, yield210 ANALYST, FEBRUARY 1986, VOL.111 I l l COOH IV COOR COCl COOR IVa VI Fig. 1. Synthetic route for the new oxalates Table 1. Physico-chemical properties of the new oxalates. 1N NMR spectra were recorded on a JEOL Model PMX-60-SI spectrometer at 60 MHz using tetramethylsilane as an internal standard. Abbreviations: s, singlet; t, triplet; and m, multiplet. IR spectra were recorded using KBr discs with a Jasco Model IRA-2 spectrometer. UV spectra were measured with a Hitachi-557 spectrophotometer Absorption spectra Analysis, % (in CH3CN) Com- Appear- pound ance M.p./'C Formula C H N E X 104/ IR ~HNMR Solubility (at 25 "C)/mM Ia Ib Ic IIb IIC IId Colourless 167-169 Cl8Hl2Ol2N2 powder Calcd.Found Pale yellow 114-116 C22H20014N2 prisms Calcd . Found Colourless 126128 C22H20014N2 needles Calcd. Found Almost 89-90 C28H32016N2 colourless Calcd . crystals Found DNPO Colourless 192-194 CI4H6o12N4 powder powder TCPO Colourless 196198 C14H404C16 48.23 2.70 47.69 2.64 49.26 3.76 49.34 3.61 51.54 4.94 51.61 5.32 49.26 3.76 48.92 3.62 51.54 4.94 51.02 5.43 50.56 5.09 50.94 4.81 6.25 6.03 5.22 5.19 4.29 4.26 5.22 5.00 4.29 4.09 3.93 3.78 k,,ax.' 1 mol-1 (KBr)/ (CDC13), nm cm-1 cm-1 p.p.m. CH3CN CH3COCH3 C2H50Ac 236 4.26 1790 4.0,s,6H 16 18 6 340 0.54 1725 7.2-9.0,m,6H 1530 1350 233 3.90 1780 3.4,s,6H 58 71 35 340 0.41 1725 3.7,t,4H 1535 4.5, t, 4H 1360 7.3-8.9, m, 6H 232 3.88 1780 1.2,t,6H 273 192 83 340 0.39 1725 3.6,m,16H 1535 4.5, t, 4H 1350 7.3-8.9, m, 6H 222 3.77 1770 3.4,s,6H 52 58 17 296 1.44 1720 3.7,t,4H 1530 4.5, t, 4H 1360 7.4-9.0, m, 6H 221 3.83 1770 1.2,t,6H 329 213 115 296 1.46 1720 3.6,m,16H 1530 4.5, t, 4H 1350 7.3-9.0, m, 6H 1710 3.7, m, 20H 1530 4.5, t, 4H 1350 7.5-9.0, m, 6H 220 3.90 1780 3.4,s,6H 1009 767 359 20 36 8 3 14 13ANALYST, FEBRUARY 1986, VOL.111 I 1 I I I 1 I 0 1 2 3 4 5 6 7 Ti meimi n Fig. 2. Chemiluminescence reaction curves obtained from the new oxalates, DNPO and TCPO. A 1-ml volume of 10 p~ perylene in acetone and 0.1 ml of 10 mM H202 in 50 mM imidazole - nitrate buffer (pH 6.50) were mixed in a quartz cuvette (1 X 1 X 5 cm). To this solution were added 1 ml of a 1 mM solution of each oxalate in acetonitrile. The generated chemiluminescence was measured with a fluorophotometer (UM-2S, Kotaki Seisakusho, Tokyo, Japan) with the li ht source off.Graphs: 0, Ia; V, Ib; 0, Ic; V, IIb; H, IIc; 0, IId; 1, TCPO; and 0, DNPO 67.1%, m.p. 114-116 "C). IR (KBr): 1780 cm-1 (CO-CO). 4.5 (4H, t, -COOCH2-) and 7.2-8.9 p.p.m. (6H, m, C6H3). UV (acetonitrile): A,,,, 233 nm (E = 3.90 x 1041 mol-1 cm-1). Analysis: calculated for C22H20014N2: C 49.26, H 3.76, N 5.22; found, C 49.34, H 3.61, N 5.19%. The other oxalate esters Ia, Ic, IIb, IIc and IId were synthesised via the intermediates Va, Vc, VIb, VIc and VId in a similar manner (Fig. 1). NMR (CDC13): 6 3.4 (6H, S , O-CH3), 3.7 (4H, t, -CH20-), Results and Discussion The physico-chemical properties of the oxalate esters are given in Table 1. It is clear that the hydrophilicity was enhanced as the number of ethylene glycol moieties in the ester increased.The synthesis of an extremely soluble oxalate in the hydrophilic solvent was attained by the introduction of the triethylene glycol moiety to the skeleton of IV to give IId. The introduction of the tetraethylene glycol moiety would be more preferable for enhancing the solubility further, but pure tetraethylene glycol was not commercially available. The chemiluminescence reaction curves for Ia, Ib, Ic, IIb, IIc and IId and also DNPO and TCPO are shown in Fig. 2. Within 2 min all the new oxalates gave the maximum intensity and decreased their chemiluminescence intensities at a similar rate. Under the experimental conditions, the new oxalates gave higher chemiluminescence intensities than those given by DNPO and TCPO.21 1 In preliminary experiments, the stabilities of IIb, IIc and IId in the presence of hydrogen peroxide18 were superior to those of Ia, Ib and Ic, which means that the type of oxalates having the structure of VI might be well suited for use in the post-column reaction in HPLC.s-16 When the concentration of the oxalate IId in the medium was increased in proportion to that of hydrogen peroxide, the relative chemiluminescence intensities increased proportion- ally to the concentration (the relative maximum intensities are 1, 4.27, 10.8 and 16 for 0.5, 1.0, 2.0 and 3.0 mM IId, respectively), meaning that the enhanced solubility of the oxalates would be helpful for more sensitive detection in the peroxyoxalate reaction. The application of IId to the detection system for HPLC is currently being studied and the details will be published elsewhere.In conclusion, a novel series of oxalate esters having hydrophilic properties have been developed, and especially bis[4-nitro-2-(3,6,9-trioxadecyloxycarbonyl)phenyl] oxalate, IId, which is extremely soluble in acetone, acetonitrile and ethyl acetate, is expected to be useful for the high-sensitivity detection of both hydrogen peroxide and fluorescent com- pounds. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Rauhut, M. M., Bollyky, L. J., Roberts, B. G., Loy, M., Whitman, R. H., Iannotta, A. V., Semsel, A. M., andclarke, R. A., J . Am. Chem. SOC., 1967, 89, 6515. Rauhut, M. M., Acc. Chem. Res., 1962, 2, 80. Mohan, A. G., and Turro, N. J.,J. Chem. Educ., 1974,51,528. Williams, D. C., Huff, G. F., and Seitz, W. R., Anal. Chem., 1976, 48, 1003. Williams, D. C., and Seitz, W. R., Anal. Chem., 1986, 48, 1478. Arakawa, H., Maeda, M., andTsuji, A., Chem. Pharm. Bull., 1982, 30, 3036. Curtis, T. G., and Seitz, W. R., J. Chromatogr., 1977,134,343. Kobayashi, S., and Imai, K., Anal. Chem., 1980, 52, 424. Mellbin, G., J. Liq. Chromatogr., 1983, 6, 1603. de Jong, G. J., Lammers, N., Spriit, F. J., Brinkman, U. A. Th., and Frei, R. W., Chromatography, 1984, 18, 129. Kobayashi, S., Sekino, J., Honda, K., and Imai, K., Anal. Biochem., 1981, 112, 99. Mellbin, G., and Smith, B. E. F., J. Chromatogr., 1984, 312, 203. Sigvardson, K. W., and Birks, J. W., Anal. Chem., 1983, 55, 432. Weinberger, R., Mannan, C. A., Cerchio, M., and Grayeski, M. L., J. Chromatogr., 1984, 288, 445. Sigvardson, K. W., Kennish, J. M., and Birks, J. W., Anal. Chem., 1984, 56, 1096. Honda, K., Sekino, J., and Imai, K., Anal. Chem., 1983, 53, 940. Wagner, G., and Singer, D., Pharm. Zentralbl., 1964,103,794; Chem. Abstr., 1965, 63, 9937d. Honda, K., Miyaguchi, K., and Imai, K., Anal. Chzm. Acta, in the press. Paper A51250 Received July Ilth, 1985 Accepted September 18th, 1985

ISSN:0003-2654    

DOI:10.1039/AN9861100209

出版商:RSC    

年代:1986

数据来源: RSC