ISSN:0003-2654    

DOI:10.1039/AN98914FX001

出版商:RSC    

年代:1989

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98914BX003

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 15 Studies on Bis(crown ether)-based Ion-selective Electrodes for the Potentiometric Determination of Sodium and Potassium in Serum G. J. Moody, Bahruddin B. Saad and J. D. R. Thomas* School of Chemistry and Applied Chemistry, University of Wales College of Cardiff, PO Box 972, Cardiff CFI 3TB, UK Bis(crown ether)-based ion-selective electrodes for sodium and potassium are described, based on the bis[( 12-crown-4)-2-ylmethyI]-2-dodecyl-2-methyl malonate sensor(1) for sodium and the bis[( benzo-I 5- crown-5)-15-ylmethyl] pimelate sensor(l1) for potassium. The best results were obtained when the sensors were used in association with 2-nitrophenyl octyl ether as plasticising solvent mediator and potassium tetrakis(4-chlorophenyl)borate as anion excluder in poly(viny1 chloride) matrices.Electrode slopes were near-Nernstian, with detection limits of <lO-5 M. The electrode features are compared with those of a sodium glass membrane electrode, for sensor I, and with a valinomycin-based potassium electrode, for sensor II. The electrodes are also discussed in relation to others reported for sensors I and II and are shown to be superior. However, although the electrodes described offer promising alternatives to glass electrodes for sodium and valinomycin electrodes for potassium, data for sodium and potassium measurements in blood serum indicate a need for further research in order to improve the correlation with flame photometric measurements. Keywords : ion-selective electrodes; blood serum measurements of sodium and potassium; analate addition; flow injection analysis; bis(crown ether) ion-selective electrodes for sodium and potassium A recent review1 has indicated that an average sized district general hospital in the UK, serving a population of 250000, would carry out 150-200 analyses of sodium and potassium on each working day.Until recently, such analyses were carried out solely by flame photometry, but ion-selective electrodes (ISEs) are being used increasingly for the determination of sodium and potassium in body fluids.l-4 All the potassium ISEs in clinical analysers use valinomycin as the electroactive component, whereas sodium glass mem- branes are the most widely used for the determination of sodium. However, the use of glass membranes presents several difficulties, such as contamination of the glass mem- brane surface by proteins, high resistance, interferences from hydrogen ions and technical problems in the design of various configurations and shapes of cell assemblies.3 With regard to these drawbacks, solvent polymeric membranes are superior.The observation by Mallinson and Truters that potassium forms a 1 : 2 complex with benzo-15-crown-5 has resulted in the synthesis by Kimura and co-workers61" of bis(crown ethers) in which the two crown ether rings are linked together by a single bridge. Owing to the co-operative effect of the two crown ether rings, bis(crown ethers) exhibit remarkable selectivity for alkali metal ions, forming 1 : 1 cation : bis(crown ether) intramolecular sandwich complexes.10 These bis(crown ethers) and their derivatives have been used in neutral carrier type ISEs, for example, a bis(l2-crown-4) [(I), Fig.11 for sodium7-8 and a bis(benzo-15-crown-5) [(11), Fig. 13 for potassium.6.s Earlier work68 has been restricted mainly to poly(viny1 chloride) (PVC) membranes containing a sensor plasticised with 2-nitrophenyl octyl ether. This paper is concerned with studies of the effects of a wider range of plasticising solvent mediators with sensors I and 11, in the presence and absence of an anion excluder, in PVC for the potentiometric sensing of sodium and potassium ions, respectively. The best systems were evaluated for the determination of sodium and potas- sium in serum using analate addition and flow injection analysis (FIA) techniques.To whom correspondence should be addressed. Experimental Materials The reagents and materials used were obtained from the following sources: bis[ (12-crown-4)-2-ylmethyl]-2-dodecyl-2- methyl malonate (I), bis[(benzo-l5-crown-5)-15-ylmethyl] pimelate (11), valinomycin, bis(2-ethylhexyl) adipate (BEHA), glucose and urea from Aldrich (Gillingham, Dor- set, UK); dioctyl sebacate (DOS), a-, b- and y-globulins and I 0 I CH2 I 0 I CH2 W W It Fig. 1. Structures of bis(crown ether) compounds: (I) bis (12-crown- 4)-2-ylmethyl]-2-dodecyl-2-methyl malonate; and (11) bis[(henzo- 15- crown-5)-15-ylmethyl] pimelate16 ANALYST, JANUARY 1989, VOL. 114 human albumin from Sigma (Poole, Dorset, UK); 2-nitro- phenyl octyl ether (NPOE) and bis( 1-butylpentyl) adipate (BBPA) from Fluka (Buchs, Switzerland); and poly(viny1 chloride) (PVC) and Breon resin I11 EP from BP Chemicals (Barry, South Glamorgan, UK).AnalaR grade chlorides of calcium, lithium, magnesium, potassium and sodium were supplied by BDH (Poole, Dorset, UK). The Kent EIL Model 1048-2 sodium glass electrodes were obtained from Kent Industrial Measurements (Chertsey, Surrey, UK). PVC Membrane Electrode Fabrication and Measurements The PVC immobilised sensors were prepared as membranes and assembled into conventional ISEs using established procedures. 11 Unless stated otherwise, the master membranes contained 5 mg of sensor, 360 mg of solvent mediator and 170 mg of PVC. In some instances, a 50% mole ratio of potassium tetrakis(4-chloropheny1)borate (KTCIPB) anion excluder, relative to sensors I and 11.was also added to the membrane. Electrochemical measurements were made with respect to a Corning Model 003 11 602H 4M calomel reference electrode with a Beckman Model 4500 digital pH - millivoltmeter coupled to a Servoscribe 1s potentiometric recorder. Elec- trode calibrations were carried out by spiking with successive aliquots of metal chlorides of known concentration into doubly de-ionised water thermostated at 25 2 0.5 "C. The electrodes were conditioned before use by immersing them for 24 h in a 0.1 M chloride solution of the appropriate primary ion. Between calibrations, the electrodes were stored in the same 0.1 M chloride solutions. The pH - e.m.f. profiles of the best systems for sensors I and 11, and for an EIL sodium glass electrode, were obtained by adding ammonia solution (ca.1 M) to give a pH of about 10.5 and then adding appropriate amounts of concentrated hydrochloric acid to reduce the pH. The membrane resistances of the assembled electrodes were measured using a Bruker E130M potentio/galvanostat. These measurements were performed on electrodes that had been stored for 24 h in a 10-1 M chloride solution of the appropriate primary ion. The separate solution method, at a 10-2 M concentration of the cation, was used to calculate the selectivity coefficients. Determination of Sodium and Potassium in Serum This was performed by analate addition and flow injection analysis (FIA) on blood serum samples stored at 40°C and obtained from the University Hospital of Wales, Cardiff, who also supplied the flame photometric data.Immediately prior to carrying out measurements, the samples were allowed to equilibrate to room temperature over a period of about 1 h. The analate addition procedure is based on the following equation : where A E is the e.m.f. difference on addition of an analyte of volume V,, and concentration C,, to a larger volume of a standard V,, but of lower concentration, C,, and S is the calibration slope of the electrode for a graph of e.m.f. versus log[A], where [A] is the concentration of the primary ion. For FIA, a flow-through sandwich potentiometric detector, in which a PVC-based sensor cocktail, with components in the proportions mentioned above for master membrane construc- tion. was attached to a conductive epoxy support as described by Alegret et al.12 was used [Fig. 2(a)]. The manifold of the flow system is outlined in Fig. 2(b). Solutions were propelled by a multi-channel peristaltic pump (Ismatec Model 1P4) through PTFE tubing (0.8 mm i.d.). Samples were injected into an Omnifit injection valve and the distance between the injection valve and the flow cell was about 5 cm. (a) -5 U Fig. 2. (a) Details of the flow-through ISE detector (a gift from Professor A. A. S. C. Machado, Oporto); and ( b ) the FIA apparatus. A, Block with sensing chamber; B, cover; C, assembled sandwich detector; 1, conductive epoxy; 2, PVC-based ISE membrane; 3, connecting lead; 4, clamping screws; 5 , inlet channel; 6, Corning Model 003 11 602H calomel reference electrode; 7, effluent fluid pool; 8, outlet from effluent pool; 9, peristaltic pump; 10, pulse suppressor; 11, injection valve; 12, sandwich flow-through cell with reference electrode; 13, potentiometer; and 14, recorder For studies relating to serum measurements, all standard sodium and potassium solutions were prepared in mock serum A (consisting of 5 mM potassium chloride, 1 mM calcium chloride and 1 mM magnesium chloride in 0.05 M Trizma base buffer) and mock serum B (consisting of 140 mM sodium chloride, 1 mM calcium chloride and 1 mM magnesium chloride, also in 0.05 M Trizma base buffer), respectively, with the pH adjusted to 7.5 with hydrochloric acid.A direct potentiometric approach (i. e., without prior dilution) was used for the determination of sodium and potassium levels in serum in the analate addition method.Electrode 5, based on bis(l2-crown-4) and NPOE plus anion excluder, was considered to be the best because of its freedom from noise and its general quality. It was used in conjunction with a Corning ceramic junction saturated calomel reference electrode and was placed in mock serum electrolyte A (15 cm3) containing 1 mM sodium chloride, spiked with undiluted serum (0.15 cm3) and the e.m.f. readings were taken. For the determination of potassium in serum, the valinomycin or bis( benzo-15-crown-5) macro-ISE, as appro- priate, was immersed in 0.5 mM potassium in mock serum electrolyte B (15 cm3), spiked with serum (0.15 cm3) and the e.m.f. readings were recorded. For sodium measurements in serum with the FIA mode, an indirect potentiometric approach was used in which serum (0.15 cm3) was diluted to 5 cm3 with electrolyte A. Samples were not diluted for the potassium measurements.For the FIA measurements, the carrier stream consisted of mock serum A and B for the determination of sodium and potassium, respectively, using a flow-rate of 1.98 cm3 min-1 and sample volumes of 100 mm3.ANALYST, JANUARY 1989, VOL. 114 17 Interferences from Biochemicals The bis( 12-crown-4) and bis(benzo-15-crown-5) electrodes were tested for possible interferences from some biochemicals that are normally present13 in blood serum and plasma. Standard amounts of the biochemicals at the upper concentra- tion range (Table 1), e.g., 0.03% urea, were dissolved in 5 mM sodium or potassium chloride, as appropriate, in background mock serum electrolytes A and B and injected into the FIA apparatus while running the corresponding electrolyte (A or B) carrier stream.The mean peak heights for five replicate injections of the same sample containing the same concentra- tion of sodium or potassium ( 5 mM) with, and without, biochemicals were compared. Lifetime Studies The lifetimes of the bis( 12-crown-4) and bis(benzo-15-crown- 5 ) electrodes were studied using membranes plasticised with NPOE and with no anion excluder. For the assessment, both types of conventional electrodes, in conjunction with a Corning Model 003 11 6024 calomel reference electrode, were immersed in doubly de-ionised water (15 cm3) and spiked with non-diluted, pooled serum samples (150 mm3). After expo- sure of the electrodes to the solution for about 90 s, fresh charges of doubly de-ionised water and spikes of pooled serum samples were taken.This procedure was repeated until the supply of pooled serum samples was exhausted, i.e. , after 89 Table 1. Content of biochemicals in plasma and their effect on optimised bis(crown ether) electrodes Normal range in AE*lmV human Relative plasma13 Bis(benzo- molecular lg per Bis( 12-crown-4) 15-crown-5) Constituent mass 100 cm3 (electrode 5 ) (electrode 13) Glucose . . 180 0.065-0.09 O(0.005) O( 0.004) Albumin . . 69000 2.8-4.5 +9.0(0.7) -3.0(0.4) a-Globulin 41 000- 0.3-0.6 +10.2(0.76) 0( 0.006) P-Globulin 5-20 X 106 0.6-1.1 -0.5(0.01) y-Globulin 150 000 0.7-1.5 +8.0(1.2) +3.0(0.8) Mixture of the above six com- ponents - - +10.0(1.6) -2.2(0.05) * A E = difference in e.m.f.of 5 mM solutions of the primary ions with and without the biochemicals (standard deviation given in parentheses for n = 5 ) Urea . . 60 0.02-0.03 t-O.j(O.2) O(0.02) 54 000 r P 50 0 20 40 60 80 ” 0 10 20 30 No. of serum spikes Electrode ageid Fig. 3. during lifetime studies on (H) sodium and (a) potassium ISEs Electrode slope and membrane resistance data obtained spikes. After exposure to the serum, the ISEs were kept in a 0.1 M sodium or potassium chloride solution, as appropriate, and calibrations and resistance measurements were made at various intervals (Fig. 3). Results and Discussion In optimising the performances of the ISEs based on sensors I and 11, the effect of several plasticising solvent mediators was investigated (Tables 2 and 3).The use of NPOE as solvent mediator, together with a 50% mole ratio of KTClPB anion excluder to sensor, yields the best over-all electrode for both sensors (electrodes 5 and 13). These conditions u’ere therefore used in all subsequent studies, unless stated otherwise. The electrodes exhibited fast dynamic response times (<10 s) and a wide pH range (Fig. 4). The electrodes (5 and 13) are superior to those reported pre~iously6~7.14 in that they have improved calibration slopes and wider e.m.f. - log[A] linear ranges, e.g., slopes of 61.0 and 59.2 mV decade-1 (Tables 2 and 3) compared with 53 and 57 mV decade-1. For comparison purposes, a commercial sodium glass electrode (EIL Model 1048-2) was evaluated after condition- ing it in a 10-1 M sodium chloride solution for 2 d before use. This electrode showed poor discrimination for sodium over potassium, lithium and hydrogen ions (Table 2 and Fig.4) and was inferior to the bis(12-crown-4) (sensor I) electrode with respect to dynamic response times (1-2 min as opposed to <10 s). The discrimination of the electrode based on sensor I over other physiologically important cations is generally better than for the ionophores ETH 157 and ETH 227.15 The best bis(benzo-15-crown-5) (sensor 11) potassium electrode possessed properties comparable to those of the valinomycin electrode (Table 3). However, its lifetime was limited. Hence, its slope decreased (Fig. 3) and the membrane resistance increased, relative to the sodium bis( 12-crown-4) electrode, after 70 exposures to serum in a conventional ISE mode (Fig. 3).The sodium bis(l2-crown-4) electrode, on the other hand, showed no significant differences in slope or resistance over the full range of 89 contact periods in serum and continued to function well for more than 1 month (Fig. 3). The results from five replicate analyses on each sample by analate addition using the conventional type electrode and by the FIA approach are depicted in Figs. 5 and 6 together with the flame photometric data obtained at the University of Wales Hospital. Typical FIA peaks for the calibration of electrodes 5 and 13 (constructed from sensors I and 11, respectively, plus NPOE mediator and KTClPB anion excluder) are shown in Fig. 7 together with some data for the serum samples. For the FIA method, the electrodes were recalibrated after every 30 injections of serum.For analate addition, the electrodes were calibrated twice, both before and after each set of serum runs, and it was found that the over-all electrode characteristics (slopes of 58.0 and 59.0 mV decade-1 for sensors I and 11, respectively) did not change under such conditions. However, with respect to precision, it should be borne in mind that for ISE measurements4 with blood serum, precisions of k0.10 and k0.63 mV are required for the sodium and potassium readings, respectively, as the normal range in samples is 135-145 mM (1.9 mV e.m.f. spread) for sodium and 3.5-5.0 mM (9.5 mV e.m.f. spread) for potassium. The correlation between the analate addition and FIA techniques and flame photometry is low (Fig.5), particularly for the determination of sodium using the FIA technique. Similarly, poor correlation is evident on inspection of the Eilood serum data obtained by Tamura et al. ,8 particularly for sodium. However, these data8 were extended greatly by artificial controls, as a result of which correlations of 0.97 or better were obtained. Correlations for the blood serum samples alone were not calculated by these workers.8 For the determination of sodium using the FIA approach, and despite18 ANALYST, JANUARY 1989, VOL. 114 I Table 2. Some characteristics of sodium bis(l2-crown-4) (sensor I) electrodes compared with those of a commercial glass membrane sodium electrode 0 . 1 ( b ) .. r. Electrode to solvent relative to Slope/ Detection Resistance type according KTCIPB Log kEit, B t Electrode No.mediator sensor/mol-% mV decade-' limit/l0-6M /MR B = K B = Li B = Ca B = Mg 1 BBPA" 0 53.0 4.0 23 - 1.43 Not determined owing 2 BEHA* 0 52.0 1.8 44 -0.81 to poor slopes 3 DOS* 0 60.0 1.3 60 -1.38 -2.93 -4.06 -3.96 4 NPOE* 0 60.8 6.3 4 -1.74 -2.40 -3.88 -3.94 5 NPOE* 50 61 .0 6.0 0.08 -1.85 -1.80 -3.68 -3.15 6 EIL glass membrane 60.5 3.2 - -1.58 -1.41 -4.17 -4.21 electrode (Model 1048-2) i Separate solution method, [MI = 10-2 M. Sensor I . Table 3. Some characteristics of potassium bis(benzo-15-crown-5) (sensor 11) and valinomycin electrodes KTCIPB Electrode type relative Detection Log krtBT No. solvent mediator /mol-% mV decade 1 110-6 M /MR B = N a B = L i B = C a B = M g Electrode according to to sensor Slope/ limit Resistance 7 8 9 10 11 12 13 14 BBPA* 0 BEHA* 0 DOS* 0 NPOE* 0 BEHA* 50 DOS* 50 NPOE* 50 NPOE 0 (Valinomycin) Sensor 11.-t Separate solution method, [MI = 10 - ? M. 52.0 60.0 60.5 61 .O 45.5 57.5 59.2 59.6 7.6 7.5 2.5 3.2 5.5 3.5 7.5 8.0 34 43 110 5 3 7 1 8 -3.16 -2.72 -2.53 -2.58 -2.67 -3.05 -3.08 -3.02 Not determined owing to poor slopes -3.23 -4.21 -4.18 -3.25 -4.20 -4.08 -3.28 -4.00 -4.04 Not determined owing to poor slopes -3.16 -3.94 -4.09 -3.14 -3.88 -3.92 -2.88 -3.80 -3.96 -40 I - 80 2 4 6 8 10 PH Fig. 4. E.m.f. responses of various ISEs at various pH values. ( W ) Sodium ISE based on sensor I; (0) EIL Kent Model 1048-2 glass membrane sodium ISE; and (0) potassium ISE based on sensor I1 the convenience and speed of the technique (about 40-60 samples h-1 can be analysed), the relative standard deviation 130 1 .8% a I 2. 120 LL 140 150 120 130 140 150120 130 [ N a * ] / m ~ (flame photometry) Fig. 5. Correlation of mean sodium ISE (sensor I) potentiometric (n = 5 ) and flame photometric ( n = 2) data for (a) analate addition; and ( b ) FIA potentiometry. Equations: (a) y = 1 . 4 3 ~ - 63.4, r = 0.68; and ( h ) y = - 0 . 3 0 ~ + 181.0, r = 0.012 is greater (mean, 3.7%) than for analate addition (mean, For FIA the low correlation with the flame photometric data may be due to poisoning of the membrane with serum components. However, studies of the effect of interferences from various biochemicals on the electrode proved to be inconclusive (Table 1), as sodium (2-6%) is frequently present in the formulation used.16 In commercial ISE analysers, the usual method of eliminating interferences from biochemicals is to place an exclusion membrane over the TSE sensor surface. However, the use of such membranes sometimes fails to remove the interferences from unidentified low relative molecular mass serum components.17918 The optimised potassium electrode based on sensor I113 was found to suffer negligible interference from the biochemicals 2.6%).ANALYST, JANUARY 1989, VOL. 114 - 19 Serum sample Serum sample 0 t 0 a ’ a 0 i __L_ n t 0 L o 0 0 0 0 2 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 7 [ K + ] / ~ M (flame photometry) Fig. 6. Correlation of mean potentiometric ( n = S j and flame (n = 2) photometric data for ( a ) potassium ISE (sensor 11): ( b ) potassium ISE (valinomycin); and (c) FIA (ISE of sensor 11) potentiometry.Equations: ( a ) y = 0 . 8 6 ~ + 1.32, r = 0.89; ( b ) = 0 . 8 9 ~ + 1.42, r = 0.88; and (c) y = 0 . 9 4 ~ - 0.41, r = 0.58 - Scan 100 mM -- 10 min - Scan Fig. 7. Typical FIA peaks obtained for (a) the sodium (sensor I) ISE (0.15-cmi serum samples were diluted to 5 cm’ with electrolyte A before taking a 100-mm3 aliquot for FIA); and ( h ) the potassium (sensor 11) ISE studied (Table 1). Also, there was better correlation between the direct (analate addition) ISE and the flame photometric methods (Fig. 6) than for the sodium measurements. Again, the analate addition technique yielded better correlation with flame photometry. Conclusion In the optimisation of ISEs based on the bis(l2-crown-4) (sensor I) and bis(benzo-15-crown-5) (sensor In) compounds, the best PVC matrix membrane electrodes are those based on NPOE as plasticising solvent mediator and which contain a 50% mole ratio of KTCIPB anion excluder relative to the sensor.Both the optimised bis( 12-crown-4) sodium ISE, which has a number of similar, desirable characteristics to the sodium glass electrode, and the bis(benzo-15-crown-5) potas- sium ISE, which has comparable characteristics to the valinomycin electrode, should, therefore, be suitable for monitoring sodium and potassium in body fluids. However, the FIA assessments are disappointing, although the analate addition approach seems promising, particularly for the determination of potassium. Nevertheless, the relative eco- nomical price of these sensors [the bis(12-crown-4) sensor (I) is only one-sixth of the cost of another neutral carrier type sodium sensor (ETH 227)] provides sufficient encouragement for continued research in this area, particularly to improve the ISE - flame photometric correlation by directing attention to the calibrant systems as in the various activities of the International Federation of Clinical Chemists,lq such as those of the European Working Group on Ion-selective Elec- t r odes.20 Financial support and leave of absence from the Universiti Sains Malaysia to one of us (B. B. S.) is gratefully acknow- ledged. Also, Dr. Keith Davies of the University Hospital of Wales is thanked for providing the serum samples and the flame photometric data, and Professor A. A. S. C. Machado of the University of Oporto is thanked for discussions on flow injection analysis, made possible by NATO Grant No.0069. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. References Worth, H. G. J . , Analyst, 1988, 113, 373. Savory, J., Bertholf, R. L., Boyd, C. J . , Bruns, D. E., Felder, R. A . , Lovell, M., Snipe, J . R . , Wills, M. R . , Czaban, J . D., Coffey, K . F., and O’Connell, K. M., Anal. Chrm. Acta, 1986, 180. 99. Oesch, U., Ammann, D., and Simon, W., Clin. Chem., 1986, 32, 1448. Byrne, T. P., Zon-Sef. Electrode Rev., 1988, 10, 107. Mallinson, P. R . , and Truter, M. R., J . Chem. Soc., Perkin Tram. ZZ, 1972, 1818. Kimura, K., Maeda, T., Tamura, H., and Shono, T.. J . Electroanal. Chem., 1979, 95, 91. Shono, T . , Okahara, M., Ikeda, I., Kimura. K., and Tamura, T . , J . Electroanaf. Chem., 1982, 132, 99. Tamura, H., Kumami, K., Kimura. K., and Shono, T., Mikrochim. Acta, 1983, 11, 287. Kimura, K . , Tamura, H., avd Shono, T . , J . Chem. Soc., Chem. Commun., 1983, 492. Kimura, K., and Shono, T., Anal. Chem. Symp. Ser., 1985,22, 155. Craggs, A., Moody, G. J . , and Thomas, J . D. R.. J. Chem. Educ., 1974, 51, 541. Alegret, S . , Alonso, J . , Bartroli, J . , Lima, J. L. F. C., Machado, A. A. S . C . , and Paulis, J. M., Anal. Leu.. 1985, 18, 2291, White, A . , Handler, P., and Smith, E . L., “Principles of Biochemistry,” Fourth Edition, McGraw-Hill, New York, 1972, p. 705.20 14. 15. 16. 17. 18. Tamura, H., Kimura, K., and Shono, T., Anal. Chem., 1982, 54, 1224. Anker, P., Jenny, H-B., Wuthier, U., Asper, R., Ammann, D., and Simon, W., Clin. Chem., 1983, 29, 1508. Davies, 0. G., Moody, G. J., and Thomas, J. D. R., Analyst, 1988, 113, 497. Gadzekpo, V. P. Y . , Moody, G. J., and Thomas, J. D. R., Analyst, 1986, 111, 567. Ladenson, J. H., and Koch, D. D., Anal. Chem., 1983, 55, 1807. ANALYST, JANUARY 1989, VOL. 114 19. 20. Maas, A. H. J., Siggaard-Andersen, O., Weinsberg, H. A . , and Ziglotra, W. G., lnt. Fed. Clin. Chem. News, 1982, 31, 5. Maas, A. H. J., Editor, “Proceedings of the International Federation of Clinical Chemistry, 1st-8th, European Working Group on ISEs,” Private Press, Copenhagen, 1979-1987. Paper 8103378A Received August 22nd, I988 Accepted October 4th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400015

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 21 Potentiometric Determination of Cephalothin Ryszard Durn kiewicz Department of Analytical Chemistry and Instrumental Analysis, Institute of Chemistry, UMCS, 2003 1 Lublin, Poland The composition of a pseudo-liquid potential-determining phase for the cephalothin-selective electrode has been determined and the following basic electrode parameters were examined: measurement range, slope, limit of detection, selectivity and lifetime. This paper discusses the effect of side-chain substituents on the electrode properties, the electrode having been used for cephalothin determination in the range 43-436.5 mg 1-1 (standard deviation, 0.64-4.5 mg 1-1). Keywords: Cephalothin-selective electrode; cephalothin determination; potentiometry Semi-synthetic cephalosporin antibiotics have been in use since the middle of the 1960s. Because the cephalosporin structure is similar to that of penicillins, both groups of compounds are characterised by similar properties and can be determined by the same methods.1 Cephalothin, the oldest known cephalosporin, has been determined by spectrophotometric,2-8 polarographic"l0 and high-performance liquid chromatographic (HPLC)11,*2 methods.In addition, for the general determination of cephalosporins, iodimetryl3 and titration in non-aqueous solutionsl4~~5 have been used. These methods are charac- terised by a low determination limit, 10-5 M, with an RSD of 2%. However they have the disadvantage of requiring complicated and laborious sample preparations prior to determination.Indeed, methods for the preparation of electrodes that are selective towards cephalosporins or for the determination of these antibiotics using ion-selective elec- trodes do not appear in the literature. Cephalosporins react with quaternary ammonium salts to form association complexes that are only slightly soluble in water and which may therefore be used as liquid exchangers in a membrane phase for the cephalothin-selective electrode. This paper discusses the method of electrode preparation and the resulting electrode properties. Experimental Reagents Analytical-reagent grade sodium chloride, sodium nitrate and sodium acetate were obtained from POCh (Gliwice, Poland). The antibiotics used were cephalothin sodium (Inst. Bioch. Pirri., Milan, Italy), cephradine (Sefril; Polfa Tarchomin, Poland, licensed by E.R. Squibb), cephazolin sodium (Laboratoires Allard, Paris, France) and cephapirin sodium (Laboratoires Bristol, Paris, France). The liquid exchanger Aliquat 336 was purchased from General Mills (Minneapolis, MN, USA). Cephalosporin stock solutions were prepared by dissolving a weighed sample in doubly distilled water; working solutions were prepared by appropriate dilution of the stock solutions. The stock solution of the sodium salt of cephradine was prepared by neutralisation of cephradinic acid with sodium hydroxide. Cephalosporin solutions were prepared fresh each day and were stored in a refrigerator between successive measurements. All solutions were prepared using doubly distilled water. Electrode Structure The major component of the electrode is a pseudo-liquid potential-determining phase based on plasticised PVC which is in direct contact with an Ag - AgCl electrode.This is placed in a cylindrical PTFE container screwed on to the electrode frame. The electrode resistance depends on both the organic phase composition (e.g., addition of tributyl phosphate leads to a significant decrease in the resistance) and on the distance of the Ag - AgCl electrode from the sensor surface (at a distance of about 1 mm the resistance is ca. 17-22 MQ). Fig. 1 shows a schematic diagram of the electrode structure. This type of electrode construction, having a pseudo-liquid potential-determining phase and no inner solution, has been used on many occasions by the author16 and recently it has been used as the basis of electrodes that are selective towards penicillin anions.These electrodes possess all the advantages of the "coated-wire" electrodes but their lifetime is longer owing to the large volume of the potential-determining phase, which acts as a reservoir of the active substance. Electrode Potential-determining Phase The Aliquat 336 - cephalosporin complex (the electrode potential-determining phase) was prepared by first placing 10 ml of the liquid exchanger (Aliquat 336) and 20 ml of a 10-1 M solution of cephalosporin in a separating funnel. The solutions were shaken for about 10 min and the liquid phase was separated. The extraction process was repeated until the reaction of the cephalosporin with C1- in the organic phase 1 2 Fig.1. Schematic diagram of the cephalothin electrode.(1) Body; (2) cable; ( 3 ) PTFE sensor; (4) Ag - AgCl electrode; and (5) pseudo-liquid potential-determining phase22 ANALYST, JANUARY 1989, VOL. 114 had ceased. After drying, the organic phase was used to prepare the potential-determining phase. To achieve this, the following de-aerated mixture was introduced into the elec- trode sensor into which an Ag - AgCl reference electrode had been inserted: cephalothin - Aliquat 336 complex (0.1 g), dibutyl phthalate (0.59 g), tributyl phosphate (0.01 g) and PVC (0.3 g). The mixture was then gelatinised at 8&85 "C for 30 min and, after cooling, the electrode was conditioned in a 10-1 M solution of cephalothin for 4 h. Measurement of E.m.f. Measurement of the e.m.f.of the cephalothin electrode - reference electrode (Orion 90-02) system was carried out in a thermostated vessel (25 Ifr 0.1 "C) using an Orion 901 Ionmeter microprocessor - ion analyser or a PHM-62 standard pH meter (Radiometer, Copenhagen). Potentiometric data were recorded using a BME 79812 recorder (Muszar-Meresz- technike, Hungary). Results and Discussion The potential of the cephalosporin electrode is given by the following equation: 2.303RT nF E = Eo - - log[cephalosporin] To study the analytical parameters of the cephalothin elec- trode, calibration graphs were obtained for solutions contain- ing cephalothin and solutions of possible interferent ions (Fig. 2). The calibration graph was linear in the concentration range 10-1-10-4 M cephalothin with a slope of 60.6 mV (pC)-l where C is the concentration of cephalothin.Response Time The response time of the cephalosporin electrode was established by determining the rate of the potential change following the rapid injection of a solution of higher concentra- tion, C,, and volume V,, into the stock solution (concentra- tion, C,; volume, V,), with C,: C, = 1 : 100 and Vp: V, = 20 : 1. The electrode potential changes were recorded and the results are shown in Fig. 3. The cephalothin electrode possesses a short response time, viz. , 3&60 s. 100 50 > E G O c Lu - 50 a) '9 - 5 4 3 2 PC / 5 I I 5 4 3 2 1 Fig. 2. Calibration graphs for the cephalothin electrode. (a) 1, Cephalothin; 2, cephazolin; 3, cephapirin; and 4, cephradine. (6) 1, Cephalothin; 5 , nitrate; 6, chloride; and 7, acetate Effect of pH The dependence of the electrode potential on the solution pH was investigated by adding hydrochloric acid or sodium hydroxide solution ( 5 X 10-2 M) dropwise to the stirred cephalothin sample (V = 20 ml, C = 10-3 M).After each addition of acid or base, the electrode potential and pH of the solution were measured. The results are shown in Fig. 4, from which it can be seen that the electrode can work effectively in the pH range 4.5-10.0. Selectivity The potentiometric selectivity coefficients of the cephalothin electrode were determined by two methods, viz., the separate solution and the mixed solution methods. To obtain the selectivity coefficients by the former method, the e.m.f. of 10-2 M solutions containing either the main ion or the interfering ion was measured.Selectivity coefficients were then determined using the following equation: E2 - El logkg;,,, = 2.303 R TI F Using the mixed solution method, the selectivity coefficients for interfering cephalosporins, cloxacillins and nitrates were determined from the following equation: where aM- is the activity of cephalothin, the activity of cephalothin after the addition of interferent and a',- the activity of the interferent. However, the following equation is used for interfering chlorides and acetates: J Dilution 1 + 1 Time - Fig. 3. C, = 10-1 M; Vp = 20 ml; V , = 1 ml; and dilution, 1 + f Response time of the cephalothin electrode. C = 10-3 M; t + E i Fig. 4 4 6 8 10 PH pH - e.m.f. response curve for the cephalothin electrodeANALYST, JANUARY 1989, VOL.114 23 This procedure results from the need to adjust the method to take account of the expected values of the selectivity coefficients. A5 Table 2 shows there is fairly good agreement between the selectivity coefficients determined by the two methods. The results indicate that the cephalothin electrode selectiv- ity for other cephalosporins is low. This is due to the structure of the compounds and the slight effect of the substituents on the density of the charge collected on the carboxyl group. Cephalothin has an acetoxymethyl group at position 3 of the dihydrothiazine system and a 2-thienylmethyl substituent on the @-lactam ring. These substituents, via an inductive effect, cause a decrease of the charge density on the carboxyl group, which, in turn, causes a weakening of the interaction between the cephalothin ion and water and hence an increasing interaction between the cephalothin ion and the liquid ion exchanger.Of those cephalosporins examined, cephalothin is characterised by having the strongest affinity for the ion exchanger. Cephapirin has a 4-pyridylthiomethyl substituent on the (3-lactam ring. This substituent has less effect on the charge density on the carboxyl group. Of the cephalosporins exam- ined, cephradine, which possesses inner salt characteristics, has the lowest affinity for the ion exchanger. From these results it can be seen that the cephalosporin affinity for the anion exchanger decreases in the order cephalothin > cephazolin > cephapirin > cephradine. Nitrate and penicillin (cloxacillin) ions have a greater affinity for the ion exchanger than the cephalosporin ions, Table 1.Analytical parameters of the cephalothin electrode Slope . . . . . . . . 58.4 mV (logCCeph) SD . . . . . . . . 3.7mV Correlation coefficient . . 0.9842 Intercept . . . . . . 151.5 mV Limitofdetection . . . . 10 5~ Measurementrange . . 10-1-5 x 10 5~ Response time . . . . 30s Lifetime . . . . . . . . 3 months pH range . . . . . . 4.5-10 Resistance . . . . . . 20MQ 4.2 pg ml-I 41.8-0.002 mg ml ~ ~~ Table 2. Comparison of selectivity coefficients, kf$,h,N, for two methods, using the cephalothin electrode. N represents the interfer- ent ion Method p t ceph,N Cephazolin . . . . Cephapirin . . . . Cephradine . . . . Cloxacillin . . , . Ampicillin .. . . Amoxicillin . . . . Nitrate . . . . . . Chloride . . . . Acetate . . . . . . * Solutions, M. Separate solution * 0.78 0.49 0.29 0.61 0.52 5.24 0.10 0.04 10.2 Mixed solution 0.65 0.38 0.25 9.5 0.74 0.41 2.58 0.05 0.001 having values of kg$h,N greater than unity. The unfavourable values of the selectivity coefficients for cloxacillin do not limit the applicability of the electrode as penicillins are very rarely used in mixtures. The low value of k&&C1 indicates the potential of the electrode for the determination of cephalothin in body fluids. Cephalothin Determination To check its analytical applicability, the electrode was used to determine cephalothin. Standard additions to the sample were performed using an Orion 901 microprocessor ion analyser.Table 3 gives the results. Cephalothin was determined in aqueous standard solution. Although slightly high results were observed in the concentra- tion range 4&400 mg 1-1 this had no effect on the accuracy of the method which was typical of analytical methods employing ion-selective electrodes. The results of the determination were compared with those given by the iodimetric method17 which is recommended in the British Pharmacopoeia. The latter method was regarded as being reliable and suitable for application. Table 3 shows that there is good agreement between the two methods; the results of the potentiometric determination are only about 2% higher than those obtained by the method recommended in the British Pharmacopoeia, demonstrating the applicability of the proposed electrode to the determination of cephalothin, The shorter analysis time of the potentiometric method is another advantage, especially for serial determinations.Conclusions An ion-selective electrode with a pseudo-liquid potential- determining phase has been prepared that allows the determi- nation of cephalothin in aqueous solutions by direct poten- tiometric methods in the range 40-400 mg 1-1. It has been shown that the affinity of cephalosporins for the ion exchanger used decreases in the order cephalothin > cephazolin > cephapirin > cephradine. This order may be attributed to the substituents present in the side-chain of 7-aminocephalosporanic acid. The results of this work demonstrate the analytical appli- cability of the electrode to a cephalothin function.The electrode may be used in manufacturing laboratories to control the cephalothin content of ready-made preparations. The potentiometric method has been shown to have an accuracy comparable to that of the iodimetric method recommended in the British Pharmacopoeia. However, its use in manufacturing laboratories may reduce the cost of the analysis significantly owing to the shorter determination times, an important factor in serial determinations. The low value of k&$cl suggests that the method can be applied to work carried out in clinical laboratories, e.g., to the determination of cephalothin in plasma. Potentiometric apparatus, which is not too costly, the use of cheap, simple reagents and the high accuracy of the determi- nation make the potentiometric method for the determination of cephalothin in the range 41.8-0.002 mg ml-1 competitive in comparison with other instrumental methods, e.g., spectro- photometry or HPLC.Table 3. Results of the determination of cephalothin Sample Method Takenimg 1-1 n Found/mg 1-1 SDimg 1 - 1 Recovery,% Cephalothin sodium . . Potentiometric 41.8 7 43.0 0.64 102.9 418.0 6 436.5 4.5 104.4 I ~ d i m e t r i c ' ~ 420 5 430.4 7.4 102.524 ANALYST, JANUARY 1989, VOL. 114 I 2 3. 4 5. 6. 7. 8. 9. 10. References Chan, C. Y., Chan, K., and French, G. L., J. Antirnicrob. Chernother.. 1986, 18, 537. Abdalla, M. A., Fogg, A. G., and Burgess, C., Analyst, 1982, 107, 213. Schroder, J . , Noschel, M., and Bonow, A., Pharrnazie, 1977, 35, 544. Mays, D. L., Bagert, F. K., Cantrell, W. C., and Evans, W. G., Anal. Chern., 1975, 47, 2229. Beltagy, Y. A., Zentralbl. Pharrn., 1977, 116, 925. Valenzuela, C. C., Bernalto, G., and Obrades, R. A., Ars Pharrn., 1986, 27, 95. Sengon, F. J., and Fedai, I., Talanta, 1986, 33, 366. Sengon, F. J., and Ulas, K., Talanta, 1986, 33, 363. Fogg, A. G., Fayed, N. M., Burgess, C., and McGlynn, A., Anal. Chim. Acta, 1979, 108, 205. Hall, D. A , , Berry, D. M., and Schneider, C. J., J. Electroanal. Chern., 1977, 80, 155. 11. Wold, J. S . , and Turmipsedd, S . A., Clin. Chirn. Acta, 1977, 78, 203. 12. Buhs, R. P., Maxim, T. E., Allon, N., Jacob, T. A., and Wolf, J . F., J. Chrornatogr., 1974, 99, 609. 13. Okada, S . , Mattori, K., and Takane, T., Bull. Chern. Soc. Jpn., 1965, 38, 2186. 14. Fogg, A. G., Abdalla, M. A., and Henriques, H. P., Analyst, 1982, 107, 449. 15. Kaminski, J. J., Dirnrnock, J. R., and Bodor, N . , Znt. J. Pharrn., 1979, 3, 151. 16. Sykut, K., Dumkiewicz, R., and Dumkiewicz, J., Ann. Univ. M . Curie-Skiodowska, Sect. A A , 1978, 13, 1. 17. “British Pharmacopoeia 1973,” HM Stationery Office, Lon- don, 1973. Paper 8100433A Received February 9th, 1988 Accepted August 5th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400021

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 25 Applicability of Gallium as Copper Scavenger in the Determination of Zinc in Samples of High Copper Content by Potentiometric Stripping Analysis Stavros V. Psaroudakis and Constantinos E. Efstathiou" Laboratory of Analytical Chemistry, University of Athens, 104 Solonos Street, Athens 106 80, Greece ~ ~~ The determination of zinc in the presence of copper by potentiometric stripping analysis is hindered by the formation of Cu - Zn intermetallic compounds on the working electrode during the deposition step. An excess of gallium is usually added to the sample to prevent the formation of Cu - Zn by forming much more stable Cu - Ga intermetallics. If the Cu : Zn mass ratio is higher than about 5, only part of the analytical signal of zinc is restored after addition of gallium and both the precision and accuracy of the determination are reduced.A simple extraction procedure is recommended to remove the bulk of copper from the sample. The pH of the working sample solution is adjusted to 4 (0.5 M acetate buffer) and is extracted with half of its volume of a 3% VNsolution of acetylacetone in benzene. Each extraction removes about 80% of the copper present without affecting the zinc. The determination of zinc is subsequently carried out after adding Hg" as an oxidant and gallium to mask the remaining copper. The method was applied to the determination of zinc in low-zinc, hig h-copper brass samples. Keywords: Potentiometric stripping analysis; zinc determination; intermetallic compounds; acetylacetone extraction of copper The determination of zinc and copper in the presence of each other by electrochemical stripping techniques such as anodic stripping voltammetry (ASV) and potentiometric stripping analysis (PSA) is hindered by the formation of intermetallic compounds between these two metals.These Cu - Zn intermetallics are formed during the deposition step and remain dissolved or as a separate crystalline phase on the mercury film of the working electrode. Kemula et al. 1 were the first to report on this type of interference in determinations of zinc by ASV. Shuman and Woodward,2 using data obtained by ASV and cyclic voltammetry, showed the formation of three compounds of the type CuZn, (n = 1-3), soljuble in the mercury phase, and the formation of an insoluble compound CuZn3, at high amalgam concentrations.More recently, Piccardi and Udisti3 examined the problems arising from the formation of Cu - Zn intermetallics in the determination of the apparent complexing capacity (versus copper) of sea water by ASV. Their findings indicated the formation of insoluble CuZn and soluble CuZn2 intermetallics. During the stripping step, the Cu - Zn intermetallics are oxidised at potentials close or equal to the stripping potential of copper. Therefore, depending on the acutal Cu : Zn ratio, the resulting zinc signal is decreased or completely eliminated, whereas the copper signal is increased. This is a particularly problematic situation as zinc and copper are ubiquitous metals found in a wide diversity of samples. In the determination of copper, the above error can be readily prevented by using a deposition potential less cathodic than that required for the deposition of zinc.Unfortunately, the deposition of zinc without the parallel deposition of copper is not feasible. Therefore, the analytical problem related to the determination of zinc cannot be alleviated as easily. Copeland et al.4 suggested the addition of an excess of gallium to the sample. Gallium forms more stable intermetal- lics with copper, allowing the oxidation of zinc at its normal stripping potential. This method is widely used in determina- tions of zinc by ASV, and it has been adopted by Jagner and co-workers5.6 in analogous determinations performed by PSA. Hoyer and Kryger7 applied the generalised standard additions method (GSAM)8 in PSA in order to resolve overlapping signals and to correct those signals affected by the formation of intermetallics.* To whom correspondence should be addressed. On trying to apply the gallium treatment in the determina- tion of zinc by PSA to samples containing copper in excess over zinc, we found that the analytical signal of zinc is only partially restored. Also, non-linear and/or heavily scattered zinc signal versus [Zn"] plots were obtained on trying to apply the multiple standard additions technique to the calculation of the concentration of zinc. Therefore, the sensitivity, precision and accuracy were reduced under these conditions. The limits of the scavenging action of gallium on copper in the determination of zinc in samples of high copper content by PSA are examined in this paper.Experimental Instrumentation The microcomputer-controlled PSA system, the cell and the electrodes used in this study have been described briefly elsewhere.9 The deposition and stripping steps took place under continuous delivery of oxygen-free nitrogen and con- stant mechanical stirring. A glass propeller rotated at 1300 rev min-1 was used for stirring instead of the magnetic bar used previously. Reagents Stock solutions (1000 mg 1-1) of zinc, copper and mercury are prepared by dissolving the appropriate amount of the corre- sponding nitrate (analytical-reagent grade) in distilled, de- ionised water. Spectroscopic quality (Johnson-Matthey) cop- per sulphate is used for the preparation of solutions with high Cu : Zn ratios.Gallium stock solution (10000 mg 1-1) is prepared by dissolving the appropriate amount of pure Ga203 in excess of 6 M HCl. Boiling is required to facilitate the dissolution. The excess of HC1 is expelled on a steam-bath before diluting to the appropriate volume. An acetate buffer (pH 4.0,0.50 M in total acetate) is used as an ionic medium. Procedure Measurements First step: determination of copper. The PSA system is programmed as follows (all potentials are vs. SCE) : deposi-26 ANALYST, JANUARY 1989, VOL. 114 tion potential, -0.80 V; deposition time, 60 s; final potential, 0.00 V; resting potential, 0.00 V; and stripping mode, "convective. " A 30.00-ml aliquot of the working sample solution contain- ing copper in the range 0.1-5 pg ml-1, spiked with t-Ig(N03)2 to a final HgII concentration of 10 pg ml-1, is transferred into the measurement cell and de-aerated for 10 min.Three plating - stripping cycles serve as an in situ plating procedure for the formation of the mercury film on the glassy carbon electrode. The potential is recorded during the stripping cycle and the copper E - t plateau is evaluated on the recordings graphic- allylo (Fig. 3 in reference 10, curve 111). The concentration of copper is calculated by the multiple standard additions method. Three additions of the appropriate volumes of the stock C U ( N O ~ ) ~ solution are usually made for better accuracy. Second step: determination of zinc. The working sample solution is spiked with Ga"' to a final concentration approxi- mately ten times that of Cut1 (in pg ml-1).The deposition potential is set to - 1.40 V and a measurement is obtained as before. The zinc E - t plateau, followed by the gallium E - t plateau, is measured graphically and the concentration of zinc is calculated by the multiple standard additions method. Three additions of the appropriate volumes of the stock Zn(N03)2 solution are made. For better accuracy, each standard addition should increase the zinc signal by a step within the range 25-50% of the initial signal. If the calculated concentra- tion of zinc is less than one fifth of the copper concentration, the result should be suspected to be inaccurate and this step is then repeated after extracting the working sample solution as described below. With high initial concentrations of copper, the first step should also be repeated to confirm the efficient removal of the bulk amount of copper.Extraction of copper from the working sample solution A 40-ml aliquot of the working sample solution is extracted with 20 ml of a 3% V/V solution of acetylacetone in benzene. For working solutions with higher copper concentrations, the extraction can be repeated with fresh amounts of organic phase. Each extraction removes about 80% of the copper. Five successive extractions remove almost all of the copper (>99.5%) without any noticeable decrease in the concentra- tion of zinc. Preparation of the working solution of the brass samples An accurately weighed amount of a low-zinc, high-copper brass sample (1-5% Zn, 80-95% Cu) in the range 0.8-1 g is dissolved in 20 ml of 8 M HN03.The solution is evaporated to about 5 ml and diluted to about 100 ml with water. The white hydrous SnOz precipitate is filtered off and washed thoroughly with warm 0.8 M HN03 solution. The filtrate is diluted with water to 200.0 ml and a 1.00-ml aliquot of this solution is diluted to 500.0 ml with the acetate buffer. This is the working sample solution containing zinc in the range 0.08-0.50 pg ml-1 and copper in the range &lo pg ml-1. Results and Discussion Zinc - Copper Mutual Interference Patterns The analytical signal ( E - t plateau) of zinc was measured in ZnII solutions after successive additions of Cull. Four different concentrations of Zn" were used and the resulting plots of zinc signal versus Cu : Zn molar ratio are shown in Fig.1 . All plots indicate a sharp decrease in the zinc signal which can be described by two intersecting lines. The average value of the Cu : Zn molar ratio, at the intersection points, is 0.33 k 0.04. The ratio of metal amalgam concentration to metal solution concentration, after a given deposition period (accumulation coefficient), is approximately the same for both metals.2 Therefore, the intersections at this particular ratio are further h 1 v) C N -. A .c 0 0.5 1 .o 1.5 [ C ~ ~ i I i l Z n ~ ~ l Fig. 1. Effect of increasing Cu: Zn molar ratio on the analytical signal of zinc at Zn" concentrations of (A) 0.33; (B) 0:67; (C) 1.?0; and (D) 1.33 pg ml-1. Deposition time, 60 s; deposition potential, -1.40 V 0 1 2 3 4 5 6 [ZnllI/[C~~~l Fig. 2. Effect of increasing the Zn : Cu molar ratio on the analytical signal of copper at Cull concentrations of (A) 0.13; (B) 0.33; (C) 0.67; and (D) 1.33 pg ml-1.Deposition time, 60 s; deposition potential, -1.40 V evidence that a CuZn3 intermetallic compound is formed when zinc is present in excess over copper. Crosmun et al." have reported an analogous experiment using differential-pulse ASV. An intersection point at a Cu : Zn molar ratio of 1 was noted, denoting the formation of intermetallic CuZn. These contradictory results may be attributed to the much thinner mercury film that we used (plating with an Hg" concentration of 1300 pg ml-1 was used by Crosmun e f ~1.11). Probably, the formation of CuZn and CuZn3 is favoured at low and high amalgam concentrations, respectively.This experiment was repeated using gallium instead of zinc. The resulting plots of gallium signal versus Cu : Ga molar ratio gave a single intersection point at a molar ratio of 1, denoting the formation of CuGa species. The effect of increasing Zn : Cu molar ratio on the copper signal was studied in a similar experiment. The resulting plots with four different initial concentrations of Cut1 are shown in Fig. 2. In all instances there are intersections of linear and non-linear graphs. An intersection point at a Zn : Cu molar ratio of about 0.9 appears consistently in all graphs, whereas a second intersection point, at a Zn : Cu molar ratio of about 1.9, is apparent in only two of them. These plots should be considered as evidence of the successive formation of CuZn and CuZn2 intermetallics. Obviously, each intermetallic is oxidised at a different rate by Hg", resulting in the patterns shown in Fig.2. A further increase in ZnII concentration increases the copper signal because of the enhanced deposi-ANALYST, JANUARY 1989, VOL. 114 27 0 2 4 6 8 PH Fig. 3. Effect of the pH on the analytical signal of gallium for a solution containing 0.67 pg ml-1 of Ga"'. Deposition time, 60 s; deposition potential, - 1.40 V; ionic medium, HCI - CH,COOH - NH4Cl. each 0.10 M; the pH was adjusted by adding small volumes of 5 M NaOH 0 ' ' I I I 0.1 1 10 100 [Gallllipg ml-' Fig. 4. Effect of the concentration o f gallium on the analytical signal of zinc (0.100 ug ml-1) in the presence of copper at concentrations of (A)O.lOOpgml-1; (B) l.OOpgml-';and(C)S.00p.gml-~. Deposition time, 60 s; deposition potential, -1.40 V tion of copper during the stripping of zinc (class ii interfer- enceg).The absence of purely linear relationships between the analytical signals of zinc and copper and their respective concentration ratios precludes computational corrections based on simple linear models of interferences.8 The computa- tional approach can be applied only over a relatively narrow concentration ratio range, and therefore it is of limited analytical importance. Selection of Ionic Medium The proposed ionic medium has the appropriate pH for the efficient separation of copper from the sample by extraction with acetylacetone without affecting the zinc present,l2 and it is also appropriate for the efficient deposition of zinc and gallium. Therefore, the over-all procedure is simplified, as further pH adjustments are not required.It also has the appropriate buffering capacity, as the gallium analytical signal is critically dependent on the pH of the solution (Fig. 3). In acidic solutions (e.g., 0.1 M HCl) the deposition of gallium is very poor. The relatively high concentrations of acetic acid and sodium hydroxide used for the preparation of the ionic medium make blank determinations for zinc necessary. Blanks for zinc were determined by PSA with each batch of the ionic medium. The average value of the blank ZnII concentrations was 0.010 k 0.002 pg ml-1. These blank values are much lower than the actual lower determination limit of ZnII (0.080 pg ml-1) in the present work, and therefore all analytical results can be simply corrected.If the determination limit is extended toward lower ZnII concentrations, e.g., by increasing the deposition time, the ionic medium should be purified electrolytically. Restoration of Zinc Signal With Gallium The percentage restoration of the zinc signal as a function of Gal11 concentration for three solutions containing 0.100 pg ml-1 of Zn" and different concentrations of Cu" (0.100, 1.00 and 5.00 pg ml-1) is shown in Fig. 4. In all instances the zinc signal reaches a steady value after adding about a 10-fold excess of Ga"' over Cur1 (concentrations in pg ml-1). These steady values correspond to a 100, 60 and 25% restoration of the original zinc signal (in the absence of CU" and GaII') for the three concentrations of copper examined.The observed partial restoration of the zinc signal may be attributed to saturation of the mercury film with Cu - Ga intermetallics and partial coverage of the mercury film with them, with a concomitant reduction in the hydrogen over- potential. This reduction in the hydrogen overpotential decreases the accumulation coefficient of zinc, also adversely affecting the reproducibility of the zinc signal. The decrease in the zinc signal appears to be the limiting factor preventing the application of the simple gallium treatment to samples of high copper content. Using ASV it has been reported that the strong overlap of the high gallium current peak with the small zinc current peak is the actual limiting factor.4 In PSA, the gallium E - t plateau (E+ = -0.88 V) not only does not obscure the development of the preceding zinc E - t plateau (Ei = -1.10 V), regardless of their relative widths, but also facilitates the evaluation of the latter on the recordings by the graphical procedure we have used.Removal of Bulk Amounts of Copper Large amounts of copper can be effectively removed from solutions containing minute amounts of zinc by controlled potential bulk electrolysis. This procedure is extremely time consuming and therefore it was not considered further for the present application. Removal of copper by precipitation with one of the many available copper precipitants always involves a risk of the coprecipitation of zinc. On the other hand, the excess of the precipitating reagent would also precipitate Hg", which is subsequently added to the working sample solution as an oxidant.Solvent extraction seems the most efficient method for removing large amounts of copper in the minimum time period. A wide variety of copper extractants can be found in the literature but only those which do not extract zinc can be considered for the present application. The extractability of various metals by weakly acidic organic extractants can be readily differentiated by controlling the pH of the aqueous phase. A literature search revealed that P-diketones show large differences between pHt,cu and pH,,,, values ( P H ~ , ~ corresponds to the pH of the aqueous phase where the extractability of the metal M is half of its maximum value). Acetylacetone (pentane-2,4-dione), the simplest P-diketone, is inexpensive, it can be used at high concentra- tions in a variety of organic solvents and it establishes extraction equilibria relatively rapidly.Copper is extracted by an acetylacetone solution in benzene (pH,,,, =: 3), whereas zinc is almost non-extractable when the pH of the aqueous phase is less than about 7 (pH:,,, = 8).12 The extraction of copper with acetylacetone is not as complete as might be required and more than one extraction is needed to reduce high Cu : Zn ratios effectively. Acetylacetone is an even weaker zinc extractant. Other more lipophilic P-diketones such as benzoylacetone and dibenzoylmethane can also be used instead of acetylacetone. These P-diketones extract copper almost completely without affecting zinc over a sufficiently wide pH range (pH;,,, = 2.5, PH;,~, = 6), but they are more expensive and relatively slow extractants.28 ANALYST, JANUARY 1989, VOL.114 Table 1. Determination of zinc in synthetic samples containing a fixed zinc concentration (0.100 pg ml-1) and various concentrations of copper using gallium as copper scavenger Concentration of zinc found k SD (n = 5)/ PSml ' Concentration of Without prior With prior copper/pg ml-1 extraction of copper extraction* of copper 0.100 0.100 L 0.008 0.109 k 0,008 (1) 2.50 0.142 i 0.031 0.113 t 0.01 1 (2) 10.0 t 0.107 k 0.025 (3) 50.0 i- 0.102 k 0.019 (4) 100 t 0.102 k 0.006 ( 5 ) 0. so0 0.106 k 0.012 0.099 i 0.005 (1) * The number of extractions required is shown in parentheses. t Not applicable owing to the high concentration of Ga"' needed.Table 2. Determination of zinc in low-zinc, high-copper brass samples Zn found k SD ( n = 5),% Certified content,% Without prior With prior extraction of extraction of 74 92.88 1.00 0.53 * 0.40 1 .00 k 0.08 78 81.04 4.69 4.51 t 1.05 4.79 k 0.30 89 85.50 1.50 1.29 2~ 0.13 1.39 k 0.09 Brasssample Cu Zn copper copper Results Zinc was determined in solutions containing a fixed ZnII concentration of 0.100 pg ml-1 and various CuII concentra- tions in the range 0.1-100 pg ml-1. The determination took place without (wherever possible) and with prior extraction of the working sample solution with acetylacetone. In all instances GaflI was added in a 10-fold excess over the remaining CU" prior to the PSA measurement. The results are given in Table 1.The accuracy of each result and the precisions between the results obtained with both methods were examined statistically at the 95% confidence level. Application of the Student's f-test revealed a significant error in the determination of Zn" concentration in the presence of a Cu*I concentration of 2.5 pg ml-1 without prior extraction of copper. At the same CU" concentration, application of the F-test revealed a significant difference between the precisions obtained, the determination of zinc with prior copper extrac- tion being more precise. The results for the determination of zinc in low-zinc, high-copper brass samples of known composition (Thorn Smith, Beulah, MI, USA) are shown in Table 2. Statistical examination of these results revealed a significant error in the determination of zinc in brass sample 89 without prior extraction of copper, and a significant difference in the precisions of the results obtained with samples 78 and 74, the determination of zinc with prior copper extraction always being more precise.Conclusions The determination by PSA of zinc in samples containing copper with a Cu : Zn mass ratio exceeding about 5 should not be conducted by simple addition of gallium to prevent the formation of Cu - Zn intermetallics. The high concentrations of gallium needed and the reduction in the over-all sensitivity and accuracy of the determination make it imperative to remove the bulk of the copper. The proposed scheme for the extraction of copper with acetylacetone is simple and rapid. Other separation schemes may also be applied.The Zn - Cu mutual interference patterns are indicative of the formation of CuZn, ( n = 1-3) intermetallic compounds during the deposition of these metals on the mercury film electrode. These findings obtained with PSA parallel previous results obtained with ASV. In a previous paper we reported on the strong interference of nickel and cobalt on the zinc signal due to the formation of stable Ni - Zn and Co - Zn intermetallic compounds.' Preliminary results have shown that gallium can also be used as a scavenger of nickel and cobalt. The optimisation of the determination of small amounts of zinc in the presence of large amounts of these two metals by PSA is under investigation. This work was supported in part (50%) by a research grant from the Ministry of Industry, Energy and Technology. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. References Kemula, W., Galus, Z . , and Kublik, Z., Nature (London), 1958, 182, 1228. Shuman, M. S . , and Woodward, G. P., Jr., Anal. Chem., 1976, 48, 1979. Piccardi, G . , and Udisti, R., Anal. Chim. Acta, 1987,202, 1.51. Copeland, T. R., Osteryoung, R. A . , and Skogerboe, R. K., Anal. Chem., 1974, 46, 2093. Danielson, L. G . , Jagner, D., Josefson, M., and Westerlung, S . , Anal. Chim. Acta, 1981, 127, 147. Jagner, D., Josefson, M., and Westerlund, S . , Anal. Chirn. Acta, 1981, 129, 153. Hoyer, B., and Kryger, L., Anal. Chim. Actu, 1985, 167, 11. Saxberg, B. E . H . , and Kowalski, B . R., Anal. Chem., 1979, 51, 1031. Psaroudakis, S. V., and Efstathiou, C. E., Analyst, 1987, 112, 1587. Jagner, D., Anal. Chem., 1978, 50, 1924. Crosmun, S . T., Dean, J. A . , and Stokely. J. R., Anal. Chim. Acfa, 1975, 75, 421. Stary, J., in Irving, H., Editor, "The Solvent Extraction of Metal Chelates." Pergamon Press, Oxford, 1964, p. 54. Paper 8102365 D Received June I4th, I988 Accepted September 12th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400025

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 29 Effect of Pre-treatment of Platinum for Modified Platinum Wire Glucose Oxidase Amperometric Electrodes S. K. Beh, G. J. Moody and J. D. R. Thomas* School of Chemistry and Applied Chemistry, University of Wales College of Cardiff, PO Box 972, Cardiff CF7 3TB, UK A micro-amperometric enzyme electrode, suitable for flow injection analysis and based on platinised platinum wire (100 pm diameter), is described. The system was tested for glucose oxidase covalently attached to the activated electrode surface with glutaraldehyde. Comparisons were made with micro-enzyme electrodes based on anodised platinum and thermally oxidised platinum wire, each linked to glucose oxidase. The response and wash times of each system were <25 and ca. 30 s, respectively.The platinised platinum enzyme electrode exhibited enhanced signals and a wider range towards glucose (0.005-30 mM) compared with the anodised platinum and thermally oxidised platinum systems (0.05-30 mM). Also, the lifetimes of 15 h obtained for both the platinised platinum wire and the thermally oxidised platinum wire systems considerably exceeded the 9 h obtained for the anodised platinum electrode when these electrodes were subjected to a continuous flow of 10 mM glucose. The lifetimes for all four systems during normal flow injection analysis exceeded 10 d. Keywords: Amperometric glucose sensor; enzyme electrode; flow injection analysis; modified platinum electrode Enzyme electrodes combine the selectivity of an enzyme reaction with a suitable electrochemical detection system.Such an electroanalytical approach offers considerable poten- tial because of its relative simplicity and ease of interfacing to other equipment. The methods based on oxidase enzyme electrodes frequently involve either monitoring the consump- tion of oxygen using a Clark electrode? as demonstrated by Updike and Hicks,' or monitoring amperometrically the hydrogen peroxide formed using a platinum electrode.2.3 Usually, the enzyme electrode consists of a layer of enzyme immobilised on a suitable matrix held over the sensing electrode tip, but more recently, electrodes have been described in which the enzyme is either adsorbed on modified carbon4 or chemically immobilised directly on to silanised anodised platinum .s Such thin membrane electrodes offer the advantages of improved diffusion of substrate to, and the diffusion of product(s) from, the active sensor zones.In a previous report5 describing the immobilisation of glucose oxidase on platinum wire, the platinum surface was first activated by pulsing the electrode between anodic and cathodic potentialsh.7 so that the platinum surface was cleaned and roughened.' The anodic - cathodic treatment essentially results in metal from the electrode surface dissolving during the anodic sweep and a fraction of it being redeposited during the cathodic sweep. This anodisation process is equivalent to surface evaporation and selective condensation and it pro- duces a clean, fresh metal surface. Alternative approaches to cleaning and roughening the platinum surface include polishing it with alumina powder followed by soaking in hot concentrated nitric acid and thermal oxidation, or platinisation.Both approaches were studied here, each being followed by silanisation and, finally, immobilisation of the glucose oxidase. The performance of the resulting electrodes was compared with that obtained by the anodisation of platinum prior to enzyme immobilisation as described by Moody et a1.5 Experimental Reagents Glucose oxidase (E.C. 1.1.3.4., 100 U mg-1, purified from Aspergillus niger) , 3aminopropyltriethoxysilane, hydrogen * To whom correspondence should be addressed. Table 1. Summary of pre-treatment of platinum wires up to the silanisation step Platinum wire Pre-treatmentstep A B C D Aluminacleaning . . V v V V Hotnitricacid .. . . V' V' V' V Anodisation . . . . .\/ v - - v Platinisation . . . . Thermaloxidation . . - - V V Silanisation(20°/o V/V) V - V V Silanisation (10% VW) - V - - _ - - hexachloroplatinate( IV) hydrate, lead acetate, 25% glutaral- dehyde solution and P-D( +)-glucose were obtained from Sigma (Poole, Dorset, UK) and platinum wire from Goodfel- low Metals (Cambridge, UK). The enzyme was stored in a desiccator in a freezer ( - 5 "C). All other reagents used were of the best analytical-reagent grade available and were used without further purification. A pH 4 sodium dihydrogenorthophosphate buffer (100 mM) was prepared; it was adjusted to appropriate higher pH values by spiking with sodium hydroxide solution (4 M). Glucose standards were prepared from a fresh stock solution of P-D(+)-glucose (0.1 M) in 0.1 M sodium di- hydrogenorthophosphate buffer (pH 7).The flow injection analysis (FTA) carrier stream consisted of the same buffer. Pre-treatment of the Platinum Wire Prior to use, four platinum wires (100 pm diameter) (two were used for control experiments) were cleaned by rubbing them with fine alumina powder followed by soaking in hot concentrated nitric acid, rinsing with de-ionised water and drying at 120 "C. Each wire was then treated differently prior to the silanisation step (Table 1). For the two control experiments involving glucose oxidase immobilised on platinum,5 the platinum wires (A and B) were anodised at +2.50 V (relative to silver - silver chloride) in sulphuric acid for 5 min using a potentiostat.This produces a layer of oxide on the electrode surface. In the first approach used here for oxidising the electrode surface (electrode C) the platinum wire was oxidised in an electric furnace under atmospheric pressure at 900 "C for 5 h .30 ANALYST, JANUARY 1989, VOL. 114 For the second approach, i . e . , platinisation of the platinum wire (electrode D), hexachloroplatinate was reduced galvos- tatically at 450 nA for 2 h in the presence of lead acetate with an electrode system consisting of two platinum wires. The electrolyte solution contained 33 mg cm-3 of hexachloroplati- nate and 0.6 mg cm-3 of lead acetate. Prior to silanisation the platinised platinum electrode was also heated in an electric furnace for 5 h at 900°C. The following scheme illustrates the platinum electrode pre-treatment , silanisation and enzyme immobilisation steps.I NH2 Glutaraldehyde Glucose 1 C H=N- E n zy m e - oxidase Immobilisation of the Enzyme Three of the four pre-treated platinum wires, namely A, C and D, were refluxed for 1 h with a solution of anhydrous 3-aminopropyltriethoxysilane in toluene (20% VlV), whereas wire B was refluxed for 1 h with a 10% V/V solution of anhydrous 3-aminopropyltriethoxysilane in toluene in order to observe the effect of different concentrations of silanising agent. After silanisation, each platinum wire was placed in glutaraldehyde solution (5% V/V in 100 mM phosphate buffer) in a stoppered sample tube for 1 h. To attach the enzyme to the treated wire, the electrode was dipped overnight in a solution of glucose oxidase (30 mg) in phosphate buffer (1 cm3 of 100 mM pH 7 buffer) at 4°C.The same batch of enzyme was used throughout this work. Apparatus An FIA system (Fig. 1) for monitoring glucose was used to evaluate the four different electrodes. The electrode potential was controlled with, and the current monitored by, a Metrohm E 61 1 VA-detector potentiostat. A Servoscribe chart recorder was used to record the output signal. The carrier stream and sample were propelled by a four-channel Watson Marlow peristaltic pump, and an Omnifit sample injection valve was used. All connecting tubes were made from either silicone rubber or PTFE and had a nominal internal diameter of 1.27 mm. A pulse suppressor was fitted between the pump and the injection valve. The hydrogen peroxide produced by the enzymatic reaction was monitored with a three-electrode amperometric flow- through cell system designed for FIA (Fig.1) at 600 mV (relative to a silver - silver chloride electrode). The Perspex flow-through cell [Fig. l(b)] was laboratory-built and con- sisted of a platinum wire based enzyme electrode (I) as the working electrode, in addition to a platinum wire auxiliary ( a ) Servoscri be chart recorder Pump Waste Amperometric flow-through cell I I Ill Fig. 1. Schematic representation of (a) the FIA apparatus and ( b ) the amperometric flow-through cell (A). (I) Enzyme working electrode; (11) auxiliary platinum electrode; (111) silver - silver chloride reference electrode; (E) silicone rubber seals; (F) plastic electrode holders; (G) potassium chloride solution; and (H) sample inlet 2 400 22 a 4.3 5.3 6.3 7.3 PH Fig.2. pH optimisation rofiles for the four different enzyme electrodes. ( A ) A; (0) B; 0) C; and (A) D electrode (11) and a silver - silver chloride reference electrode (111). The cell was designed so that the auxiliary and reference electrodes were placed in a stationary solution of saturated potassium chloride and were in contact with a flowing buffer stream by means of a T junction [Fig. l ( b ) ] . Each electrode was held in place with silicone rubber seals. When not in use the enzyme electrode was stored at 4 "C in 100 mM pH 7 phosphate buffer. Results Optimisation of Conditions for Glucose Determination The sample volume and carrier stream flow-rate were optimised for each electrode using a modified Simplex optimisation algorithm for which the control parameters were the flow-rate and sample volume with respect to the response criteria of peak height and the time taken from sample injection to the attainment of the maximum signal.The bias of the optimisation is the maximisation of peak height but for the minimum time taken from injection to signal with a greater emphasis on peak height. Hence the carrier stream flow-rate was fixed at 2.0 cm3 min-1 and the corresponding sample size was optimised at 0.50 cm3 for all further work.ANALYST, JANUARY 1989, VOL. 114 31 The pH was optimised for each of the four enzyme electrodes by varying the pH between 5 and 9 by increments of 0.5 pH. The resulting peak height - pH plots showed plateaux at pH 6-7.5 (Fig.2). Therefore, all further work was carried out at pH 7 for the different enzyme electrodes. Other workers have reported a broad pH range of 4.0-7.0 with a maximum response around pH 5.5 for solubilised glucose oxidase.*.g The optimum pH range is a direct result of the micro-environment of the enzyme and is related to the immobilisation technique employed and to the nature of the supporting material. Electrode Calibration The electrodes were calibrated with glucose standards over the range 0.005-100 mM using the optimised conditions described above. Fig. 3 shows a typical chart recorder output obtained with electrode D for the triplicate injection of glucose standards and illustrates the utility of the electrode under FIA conditions.The lowest detectable concentration of glucose using the platinised electrode (D) was 0.005 mM, but was only 0.1 mM for the two control anodised platinum electrodes (A and B) and for the thermally oxidised platinum wire electrode (C). The calibration graph was linear up to 30 mM glucose [Fig. 4(a)]; the log - log plots are shown in Fig. 4(b). 12.5 pA I 100 r n M I\ 0 r n M 1 Omn 750 nA1 I 5 min H I Scan time --- Fig. 3. electrode D Typical recorder output towards glucose for glucose oxidasc 40 000 30 000 P a 20000 10 000 0 Discussion The various pre-treatments applied to the platinum wire prior to enzyme immobilisation lead to enzyme electrodes with different response characteristics, particularly for the platin- ised platinum system (D) compared with the other electrodes.Hence electrode D shows considerable enhancement of signal response over both electrode C and the control electrodes A and B, and permits improved sensitivity towards low levels of glucose (down to about 0.005 mM) (Fig. 4). No analytically useful signals were obtained for glucose standards below 0.05 mM with electrodes A, B and C. However, the speed of response, under the optimised conditions, is similar for all four electrodes, being ca. 24 s from base line to peak response in each instance. The wash times, i.e., the times required for the signal to revert to the base line, are only slightly longer, being ca. 30 s (see, for example, reference 5 and Fig. 3). For electrodes A, B and C the platinum wire was oxidised either by anodisation or by thermal means, whereas the platinised platinum wire D was oxidised by means of a final thermal stage aimed at preventing losses or redistribution of the newly formed platinised surface.Each approach was adequate for silanisation of the electrode surface as can be seen in systems A and C for the same concentration of silanising agent. The effect of doubling the concentration of the silanising agent for systems A and B produced little change in the signal. The platinisation of platinum wire increases both the surface area of the electrode and also the number of active sites available for silanisation. This in turn enhances the amount of enzyme that can be immobilised on the electrode surface, thus leading to significantly larger current outputs for electrode D, viz., about seven times greater than for the control electrodes A and B and the thermally oxidised electrode C (see Fig. 4).The advantages of electrode D are apparent from a one-way analysis of variance (ANOVA), the results of which are shown in Table 2. The analysis indicates that electrode D gives a highly significant improvement in response over the other electrodes and that this improvement is considerably greater than for any comparison between electrodes A, B and C. With regard to optimisation of the experimental conditions it should be noted that these were similar to the various electrode types. This, however, is to be expected from the A A A I I -2.5 I -1.5 I 0 ' --5.5 -4.5 -3.5 Log ([glucosel~M) Fig. 4. Glucose calibration plots. ( u ) A1 ( y ) versus [glucose] (x).( A ) Electrode A: y = 0.770(+0.066) + 10.5.510(k2001.9)~, r2 = 0.999: (0) electrode B: v = 0.720(+0.016) + 188970(5889.0)~, r2 = 1.000; (0) electrode C: y = 0.860(+0.066) + 117234(+222.5.8)~, r2 = 0.999; and (A) electrode D: y = .5.300(+0.066) + 732713(+13906.8)x, r2 = 1.000. Prediction intervals for single y values. (A) Electrode A: i2001.9; (0) electrode B: f889.0; (0) electrode C: 5222.5.8; and (A) electrode D: k13906.8. ( h ) Log AZ b) versus log [glucose] (x). (A) Electrode A: y = .5.0380(k0.066) + 1.0053(+0.02l)x, r2 = 0.999; (0) electrode R: y = .5.040.5(~0.016) + 1.0053(+0.00.5)x, r2 = 1.000; (0) electrode C: y = .5.0838(50.066) + 1.00.53(50.021)~, r* = 0.999; and (A electrode D: y = 5.8785(+_0.066) + 1.0049(+_0.021)~, rz = 1.000. Prediction intervals for single y values.( A ) Electrode A: k0.066; (01 electrode B: k0.016; (0) electrode C: 50.066; and (A) electrode D: k0.06632 ANALYST, JANUARY 1989, VOL. 114 Table 2. One-way analysis of variance for the four types of platinum wire electrode. Critical values are 4.41 atp = 0.05 and 8.29 atp = 0.01 with numerator = 1 and denominator = 18 Comparison of electrodes AandB , . AandC . . AandD . . BandC . . BandD . . CandD . . . . . . Difference at Difference at highly significant F-factor significant level level 347.75 v v 8.20 v X 1.3 x 106 \’ v 134.84 v v 1.6 x 106 v v 3.1 x 105 v v nature of the electrode under study, which simply consists of an enzyme layer immobilised directly on the electrode surface. Hence, diffusion conditions are more ideal than in instances where an enzyme membrane on a separate support is placed over the electrode tip.Electrode lifetimes and storage stability are significant factors when considering a wider practical role for biosensors. Hence, when a 1 mM glucose solution was pumped continu- ously through the electrode system until failure occurred, it was found that the platinised platinum system (D) and the thermally oxidised system (C) had lifetimes of 15 h, which was an improvement over the lifetimes of 10 h found for the control anodised systems (A and B). Also, when stored in a buffer overnight at <6”C, and with daily use, all the electrodes (A-D) had a lifetime in excess of 10 d. Therefore, there is a general improvement in the performance of electrodes C and D compared with those in which the enzyme is immobilised on anodised platinum as reported previously.5 The critical factor with regard to lifetime may relate to the stability of the Pt-0 bonds, i.e., cleavage of these bonds would lead to a loss of enzyme and hence deactivation of the electrode.However, the thermal pre-treatment stage employed for electrodes C and D seems to be beneficial. Interestingly, Masoom and Townshendl0 reported a glucose oxidase activity of up to 1 year for 3-aminopropyltriethoxy- silane - glutaraldehyde - glucose oxidase on controlled poros- ity glass (CPG). One factor that can reduce electrode lifetimes is the setting of the electrode to an anodic potential which may facilitate desorption of oxygen, resulting in premature loss of enzyme activity by cleavage of the Pt-0 bonds.Conclusion Platinisation of platinum wire significantly enhances the electrode surface features for the immobilisation of glucose oxidase and yields an improved enzyme electrode (type D) with respect to current response, linear range, detection limits, lifetime and storage stability. Electrodes (type C) prepared from thermally oxidised platinum compare favour- ably with the type D sensor except for the lower signal output and shorter linear response range. Hence both types of electrode offer advantages over the control electrodes (types A and B). The authors thank the Trustees of the Analytical Chemistry Trust Fund of the Royal Society of Chemistry for the award of an SAC Research Studentship (to S. K. B.). References 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. Updike, S. J . , and Hicks, G. P., Narure (London), 1967, 214, 986. Guilbault, G. G . , “Analytical Uses of Immobilized Enzymes,” Marcel Dekker, New York, 1984. Guilbauit, G. G., Ion-Sel. Electrode Rev., 1982, 4, 187. Marko-Varga, G., Appelquist, R., and Gorton. I . , Anal. Chim. Acta, 1986, 179, 371. Moody, G. J . , Sanghera, G. S . , andThomas, J. D. R . , Analyst, 1986, 111, 1235. Woods, R.. Electroanal. Chem., 1976, 9, 9. Gilman, S . , Electroanal. Chem., 1967, 2, 111. Bright, H. J . , and Appleby, M., J. Biol. Chem., 1969, 244, 3625. Weibel, M. K., and Bright, H. J . , J. Biol. Chem., 1971, 246, 2734. Masoom, M., and Townshend, A., Anal. Chim. Actu, 1984, 166, 111. Paper 8l031.501 Received August 2nd, 1988 Accepted September 26th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400029

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 33 Selective Electrochemical Biosensors From State-switching of Bilayer and Monolayer Lipid Membranes by Lectin - Polysaccharide Complexes Ulrich J. Krull, R. Stephen Brown, Robinson N. Koilpillai, Roberto Nespolo, Ali Safarzadeh-Amiri and Elaine T. Vandenberg Chemical Sensors Group, Department of Chemistry, Erindale College, University of Toronto, 3359 Mississauga Road North, Mississauga, Ontario L5L I C6, Canada Interaction of the lectin concanavalin A with the polysaccharide glycogen can provide rapid spontaneous transients of the surface potential at bilayer and monolayer lipid membranes. The selective binding process can cause large, rapid potassium ion current fluctuations across bilayer membranes in a manner that is periodic and reproducible. The frequency of these transient ion current signals was shown to be related to sub-nanomolar concentrations of the reactive agents in aqueous solution.The physical mechanism responsible for ion current modulation was investigated by fluorescence methods using lipid vesicles, by the thermal dependence of the potassium ion current across planar bilayers and by pressure - area and dipolar potential measurements of lipid monolayers at an air - water interface. The mechanism is primarily associated with physical perturbations of lipid membranes by lectin - polysaccharide aggregates, resulting in the formation of localised domains of variable electrostatic potential and conductivity. Keywords: Lipid membrane; lectin; Langmuir - Blodgett trough; dipolar potential; fluorescence Most biosensors operate on the principle of the detection of an analyte that is a product of the interaction of a selective receptor with some targeted species.Example of such products are ions or small molecules produced during an enzyme - substrate reaction or by the displacement of a marker compound from the binding site of an antibody due to antibody - antigen complexation. Amplification of the ana- lytical signal is usually accomplished by conjugating an enzymatic reaction with the receptor binding event. The amplification process is based on the ultimate production of a high steady-state analyte concentration, and the response time and final signal is limited by the kinetics of the enzymatic reaction and by the continuous availability of the substrate for analyte evolution.Biosensors based on the detection of changes in the properties of bilayer lipid membranes (BLMs) caused by selective binding events do not require the production of secondary products for analysis. An example of a selective transduction process has been encountered in natural cell membranes where the control of ion-channel conductivity can occur as a switching between at least two conduction states.' Molecular switching systems may provide the basis for a new class of chemical sensing devices, which can be envisaged most simply as any chemically selective molecular arrangement that can exist in two or more discrete physical states. Switching between states could be represented as a change in molecular receptor conformation on binding a stimulant. However, the absolute analytical signal (e.g., ion current) associated with single ion channel events encountered in nature is usually very small, limiting the practical utility of such an analytical strategy. A primary binding event, which locally alters the structure or electrostatic field of a lipid membrane, provides a physical change that can readily be detected by electrochemical or optical methods.2 Amplification of the analytical signal can be achieved by inducing a greater spatially distributed structural change caused by the selective binding event. Ideally, this concept would converge in a phase transition of the lipid matrix caused by a very limited number of receptor binding events. Such macroscopic state-switching provides opportuni- ties for the sensing of very low concentrations of target species by monitoring the discontinuities in a relatively invariant analytical signal.The lectin concanavalin A (Con A) is known to bind selectively and polymerise with certain polysaccharides when Mn2+ and Ca2+ ions are complexed to two sites in the protein. We have previously reported a series of electrochemical experiments which involved the addition of Mn2+ - Ca2+, Con A and saccharide to one side of a BLM."4 The most outstanding feature of the time-dependent ion current profiles is their similarity to those of the lytic immunochemical system, which is an accepted ion channel conduction system.5 With Con A , step-current increases are not due to the opening and closing of pores; another mechanism, suitable for rapid current change related to a critical phenomenon, must be involved.Accordingly, these conductivity effects have been attributed to lectin - polysaccharide polymerisation and to precipitation on the membrane surface in a non-diffusion controlled process. The biochemical literature indicates that Con A has an effect on the ability of cell receptors to form patches 0; caps on the cell surface. The formation of these caps has been linked to various cellular events such as mitogenesis. One theory proposed to explain this phenomenon states that poly- saccharides located on the cell receptors bind to a relatively small number of Con A molecules. This leads to a phase transition in the lipids of the membrane, which decreases the fluidity and increases the phase transition temperature of the membrane.It has been noted that cap formation and its various effects are dependent on binding a critical number of molecules of tetravalent Con A.6 The nucleation - aggregation process proposed to explain the electrochemical action of Con A during the BLM experiments is that of a crtical event. Nucleation will only occur when a critical concentration has been achieved at which point an aggregate will form and grow. This implies that a concentration threshold-dependent mechanism of signal evol- ution should exist, which would provide exciting new possibili- ties for analytical strategies based on molecular switching phenomena. This paper describes an investigation of concentration- dependent state-switching responses derived from the interac- tion of Con A with a polysaccharide at both monolayer and bilayer lipid membranes.Electrochemical ion current and fluorescence studies of BLMs, in addition to dipolar potential and pressure - area data for lipid monolayers at an air - water interface, provided structural and electrostatic information about the mechanism responsible for step-current evolution.34 ANALYST, JANUARY 1989, VOL. 114 Radiolabelled Con A was used to determine the actual concentration of the lectin in solution as the macromolecules adhered tenaciously to all container surfaces. These results provided an indication of the care that must be exercised when quantitatively evaluating any form of biosensor at trace analyte concentrations. Experimental Chemicals Bilayer lipid membranes were formed from dry, purified decane mixtures of lyophilised egg phosphatidylcholine (PC), phosphatidylserine (PS) (Avanti Biochemicals, Birmingham, AL, USA) and cholesterol (C) (Sigma, St.Louis, MO, USA), where the composition of the mixture was 20 mg of phospho- lipid (10% PS) and 20 mg of cholesterol, in 1 ml of decane. Lipid solutions were stored under a nitrogen atmosphere at -20 "C and were discarded when sterol oxidation became evident from TLC analyses. Lipid monolayers were prepared from solutions containing 3 mg of phospholipid (1070 PS) and 3 mg of cholesterol in 5 ml of hexane. The optical fluorescence experiments were performed using pyrene-labelled Con A (Molecular Probes, Eugene, OR, USA), 4-heptadecyl-7-hydroxycoumarin (HDHC) (Mol- ecular Probes) and tranL~-4-dimethylamino-4'-( l-oxobuty1)stil- bene (DOS).' Concanavalin A and glycogen (relative molecular mass 106) (Sigma) were introduced as complementary selective reac- tants into aqueous electrolytic solutions containing analytical- reagent grade KCI, CaCI2 and MnC12.Water was purified with a Milli-Q cartridge filtration system (Millipore, Mississauga, ON, Canada). Radioactive Con A in the tritium-labelled form (Amersham Canada, ON, or DuPont Canada, PQ) was combined with unlabelled Con A in known ratios for the calibration experiments on solution concentration. The scintil- lation fluid used for detection contained 2,5-diphenyloxazole and 1,4-bis[2-(5-phenyloxazolyl)]benzene (Sigma). Apparatus The electrochemical cell used for the planar membrane studies consisted of two identical machined Perspex blocks separated by a PTFE sheet (0.1 mm thick) which contained a circular aperture (1 mm diameter) used for supporting the BLM.An external direct potential was applied across the membrane between two Ag - AgCl single-junction reference electrodes (Orion Research, Cambridge, MA, USA). The external circuitry consisted of a d.c. power supply and a digital electrometer (Model 616B, Keithley Instruments, Cleveland, OH, USA). The solution cell and sensitive equipment were isolated in a Faraday cage. An infrared lamp was used to provide variable temperatures in the range 21-27 "C. High speed acquisition of the time-dependent ion current signals from the BLMs was achieved with a circuit consisting of a Texas Instruments TL082CP current to voltage operational amplifier with a field-effect transistor input.This device had a low leakage current and a fast response and was connected to a virtual ground shared with the low output of the power supply for the electrochemical experiments. The output of the current to voltage device was amplified further and was then displayed on a 20-MHz Model HM 203-5 oscilloscope (Hameg, Frankfurt, FRG). The data were captured and later displayed with a Model DS102 fast sampling and storage instrument (Polar Instruments, Guernsey, UK). Lipid mono- layers were prepared on a Lauda Model 1974 Langmuir - Blodgett film balance (Sybron-Brinkman, Toronto, ON, Canada). The subphase volume of the trough was reduced from 1 to 0.3 1 by partitioning the unused surface area with PTFE barriers and by placing PTFE blocks into the base of the subphase compartment to displace the solution volume.Reagents were introduced into the trough subphase while a monolayer was in place at the air - water interface. An infusion assembly consisting of a peristaltic pump withdrew the subphase solution through a perforated tube located near one wall of the trough and returned the solution through a similar tube placed near the opposite wall. A Luer-lock connection in the flow line allowed the introduction of the reagents into the flowing stream with a syringe. Surface potential measure- ments were made with an electrostatic voltmeter (Isoprobe 162, Monroe Electronics, Lyndonville, NY, USA) using a non-contacting probe (Model 1015A, Monroe Electronics).The probe sampled a surface area of approximately 3 mm in diameter when placed within 1 mm of the surface of the monolayer and provided a time resolution of 0.2 s. The air gap beneath the probe was purged with a continuous stream of filtered air to reduce the instrumental noise. The fluorescence experiments were performed with an SLM 4800 spectrofluorimeter for all the emissive probes studied. Lipid vesicles for the optical studies were prepared with a Model VC250 sonicator (Sonics and Materials, Dunbury, CT, USA). Surface analysis by X-ray photoelectron spectroscopy (XPS), which was used to determine the occurrence of non-selective protein adsorption, was carried out with a McPherson Model ESCA 36 instrument using A1 Ka radia- tion.Quantitative measurements of the radioactivity of tritiated Con A were made with a Model LS 7000 scintillation counter (Beckman Instruments, Toronto, ON, Canada). Procedures BLM electrochemistry A lipid - cholesterol mixture was introduced into the PTFE aperture between the two compartments containing 0.1 M KCI electrolyte, by the brush technique and membranes were formed under an applied potential of +25 mV. The initial ion current through the membrane was measured after membrane stabilisation, but before the addition of various reagents to the solution compartment. The use of the brush technique to form the BLM prevented contamination of the solution compart- ment with excess of lipid. Concanavalin A and glycogen were each prepared as separate aqueous solutions and were then tested individually for their capability to initiate an electrochemical response. The experiments consisted of adding various concentrations of lectin, 10-4 M Ca2+, 10-4 M Mn*+ and 10-8 M polysaccharide solutions, both as individual constituents and in all possible combinations, to the stirred electrolyte which supported the BLM.Individual and sequential additions were accomplished in less than 1 min in all instances. The pH for all experiments was fixed at 6.5 with a phosphate buffer. The temperature of the solution was adjusted to a fixed value within the range 21-27 ? 1 "C before the lectin or polysaccharide solutions were added. Monolayer studies The non-contacting electrostatic probe was calibrated against an electrically grounded metal plate and was then positioned over the trough, which was filled with a 0.1 M KCl solution adjusted to pH 6.8 with Trizma base and re-adjusted to give a zero reading.The hexane solution containing the lipid was added slowly to the surface of the trough using a syringe. Compression was started after a period of 15 min during which the lipid was allowed to remain in a fully expanded state. The monolayer was compressed to a surface pressure of 32 mN m-l, after which the surface area was not changed. The peristaltic pump was activated and Ca2+ and Mn2? ions were infused to give a final concentration of each of 10-4 M. After a period of 5 min, sufficient Con A solution was added to achieve an apparent concentration of 10-9 M. A further 30 min elapsed before the glycogen solution was added to obtain concentrations in the range 5 x 10-10-5 x 10-7 M.After each experiment, the tubing used for subphase circulation and theANALYST, JANUAKY 1989. VOL. 114 surfaces of the trough, were washed extensively with a solution of sodium dodecyl sulphate for a period of at least 12 h. The system was then rinsed with water and wiped carefully with acetone followed by ethanol. Optical experiments The synthesis and purification of DOS has been described previously.7 Phosphatidylcholine and cholesterol, dissolved in chloroform, were placed in a small vial, and, after the addition of DOS from a concentrated stock solution in chloroform (the final concentration of the probe was typically 6 x 10-6-8 x 10-6 M), the solvent was evaporated under a stream of nitrogen.Then, 5-8 ml of Tris - HCl buffer solution (15 mM) (pH 7.4) were added and the samples were sonicated with a Model VC250 sonicator (Sonics and Materials, Dunbury, CT, USA) set at 40 W and equipped with a small direction insertion probe. Sonication was continued for 30 min at 5-10 "C in a nitrogen atmosphere. All sonicated solutions were centrifuged at 27 000 g for 30 min to remove insoluble material. The PC - C (50150 mol-%) vesicles for the HDI1C experiments were prepared by sonication for ca. 30 min at about 5 "C in a buffer solution (15 mM Tris - 0.1 M KCI) containing approximately 10-5 M HDHC. They were then centrifuged at 27 000 g and passed through a Sephadex (350 column. The fluorescence intensity of DOS and HDHC in the PC - C vesicles was measured as a function of pH in both the presence and absence of Con A , glycogen (= 10-8 M) and al\o Con A and glycogen. The degree of ioniwtion (a) was measured by dividing the fluorescence intensity (If) at a given pH by that at the highest pH, where If was constant (pH =r 11.5-12), and the pK, of I-IDHC was thus obtained.The experiments designed to evaluate the proximity of the pyrene-labels on Con A during Con A - glycogen interaction employed vesicles prepared as described above. Pyrene- labelled Con A was added stoicheiometrically with glycogen within a 10-6-10-8 M concentration range. Surface analysis Samples of thin (2 x I X 0.2 cm) Perspex sheets were cleaned rigorously, first with a 1.0 M NaOH solution, then with concentrated chromic acid and finally they were washed extensively with water.Handling only with forceps, these samples were allowed to stand for 1 h in stirred 0.1 M KCl solutions (buffered between pH 5 and 8) containing 10-6 M Con A. The sheets were withdrawn from the solution5 and rinsed gently with a small amount of water after which they were vacuum dried at room temperature and stored in a vacuum desiccator. Surface analyses were performed to determine the degree of nitrogen contamination caused by the non-selective adsorption of protein from solutions of pH 5-8. Con t r o 1 s a m p 1 e s were prepared si m u 1 tan e ou s 1 y . Rndiolabelled Con A Tritiated Con A (?H-Con A , 1 0 ~ 1 ~ mol) was added to one compartment of the cell used for the BLM electrochemical \tudies, which had been modified 50 that there was no aperture in the dividing PTFE sheet. After a 15-20 min incubation period (with stirring), an aliquot of the solution was taken.placed in a scintillation vial and allowed to evaporate. A 20-ml volume of scintillation fluid containing 3.0 g 1-' of 2,5-diphenyloxazole and 0.3 g 1 - 1 of 1,4-bis[2-(5- phenyloxazolyl)]benzene in toluene - dioxane - ethanol (66.6 + 23.3 + 10) was added and the solution wa\ counted in order to measure the radioactivity and calculate the concentration of Con A in solution. Two method5 of sub5equent Con A addition were used: in the first, the volume of the aliquot removed was replaced by a KCl buffer containing sufficient labelled and unlabelled Con A to give a ten-fold increase in the total amount of Con A in the solution.The cycle was repeated and the new value for the non-adsorbed 'H-Con A was determined. The amount of ?II-Con A increased a5 the total Con A concentration rose; however, the ratio of 7H-Con A to non-radioactive Con A decreased. In the Tecond method, the contents of the cell were disposed of after sampling and a new cell containing the next iH-Con A concentration wa\ prepared. For all Con A concentration\, the amount of ?H-Con A present was kept constant while the amount of unlabelled Con A was increased. The range of Con A concentrations tested in both methods was 10-'3-10-7 w . Results and Discussion Concanavalin A is a globular protein containing saccharide. ion and hydrophobic binding sites.8.: The maximum diameter of the Con A molecule is 40 X 39 A.Each monomer has an optimum affinity for one Calf and one Mn'f ion at specific sites located approximately 4.6 A apart.'(' It has been found that Mn'+ and Ca2+ cofactors are needed for optimum binding activity to occur. Con A contains approximately 67% (3-sheet but no a-helix (owing to the high percentage of [Mieet forming amino acid sequences). 1 0 H,'gh resolution structural studies of Con A (resolution 2.0 A) have shown that the molecular surface is relatively smooth and uninterrupted, except for one large depression or cavity extending deep into each protomer. It is likely that these cavities contain the hydrophobic saccharide binding sites. The pH of the solution used in these experiments induces the formation of a quaternary protein structure which takes the form of Con A tetramers.The fact that tetrameric Con A has multiple saccharide binding sites suggests that cross-linking with a Q r I 0 7 .. +j 1.0 3 C 0, (0 4- .- E 0.5 P 3 8 0 - 0 15 Timel rn in 30 Fig. 1. Time scale of the various ion current - time profiles observed during electrochemical expcrirnents using PC - C, BLMs in thc presence o f Con A and glycogen. 'Time: 0-15 min, Con A and glycogen, I 0 " M ; 15-20 min. Con A and glycogen, 10 '' M; 20-30 min, Con A and glycogen. 10 - M a, -0 3 +- .- El 2.0 m E i- C 1.0 L 3 z 2 .- a, .- 2 1 - (1: 0 1.0 2.0 3.0 4.0 5.0 Timeis Fig. 2. Fast sampled ion currcnt - time profiles observed during electrochemical experiments using PC - C BLMs in the presence o f Con A and glycogen. The rise tinic o f the ion current switching event is significantly longcr than t h a t associated with classical ion channel activity36 ANALYST, .JANUARY 1989, VOL.114 polysaccharide can occur. This provides the capability for polymerisation and aggregate growth when binding of poly- saccharides ensues. No significant electrochemical ion current variations are observed during the binding of Con A to mono- or disaccharides. However, binding of large polysaccharides such as dextran or glycogen provides electrochemical K+ ion current “steps,” supporting the hypothesis of aggregate formation. Irregular ion current steps are always preceded by an induction period lasting from 3 to 30 min, and they appear most frequently in situations where the lipid component of the membranes had undergone only minor chemical oxidation .4 A composite of various ion current profiles obtained for Con A - glycogen interactions at BLMs is shown i n Fig. 1. The different electrochemical characteristics result from variations in the concentrations of the reacting species and from the chemical composition of the lipid membrane. One of the observations that can be made from an examination of Fig. 1 is that ion current switching between two or more states can occur, thus providing a current - time profile analogous to that of the ion channel events observed for lytic antibody - antigen complement interactions.5 Phe- nomenologically similar ion current switching of much higher speed has been observed for ion channels associated with neural receptors. Generally, it appears that the rise and fall times of the switching events for true “ion channels” (which are polar pathways through proteins) are of the order of tens of microseconds.1 The results obtained with a rapid current - time sampling system used to study the electrochemistry of BLMs are shown in Fig.2. The two profiles in Fig. 2 point to exponential rise times of hundreds of milliseconds, giving results consistent with the concept of a switching phenomenon associated with nucleation and aggregate growth. The time resolution, based on the digitised output of the Keithley electrometer used to measure the ion currents of the BLhils, was not sufficient to provide this result, but either skewed the ion current steps or gave rise times which appeared to be instantaneous. The two profiles shown in Fig.2 indicate that both short-period (seconds) and long-period (minutes) ion current transients are observed experimentally. Recent reports have indicated that spontaneous electrical voltage fluctuations are observed in platinum electrode-based electrochemical cells which employ a thick oil membrane barrier and hydrophobic ions. II.12 Oil membranes doped with hydrophobic ions produced regular spontaneous voltage fluctuations only when the aqueous solutions were unstirred and when non-equilibrium conditions were maintained. Tem- porally, a non-equilibrium situation also exists in the Con A - BLM experiments as the reaction necessary for the formation and growth of the Con A - polysaccharide aggregates is relatively slow and may take many minutes.8 The rapid ion current transitions seen for the BLMs indicate that both Con A and polysaccharide accumulate in the unstirred layer on the membrane surface.A critical concentration of reactants or reactive precursor particles is achieved by diffusion into the unstirred layer, resulting in a rapid large-scale aggregative event (nucleation) which suddenly changes the BL,M ion current. Dramatic results for the selective binding of Con A to polysaccharides at lipid monolayers were reported recently from epifluorescence work.13 The introduction of the Con A - polysaccharide system caused a very rapid change in the domain distribution, resulting in a redistribution of the microcrystallites in the mixed phase monolayer. The growth of Con A - polysaccharide polymers could be seen and was consistent with two-dimensional growth on the surface of the monolayer.Although these results do not provide a complete description of the dynamic processes that occur at monolayer surfaces in the presence of Con A - polysaccharide systems, they do support the hypothesis that these binding events lead to aggregate growth on lipid membrane surfaces and can also lead to major changes in the structure of lipid membranes. Optical Results The mechanism of ion current alteration through BLMs is not apparent from epifluorescence studies and so it has been investigated using other fluorescence techniques. One pos- sible effect of the interaction of Con A with glycogen would be a condensation of the monomers of each of the two species to form aggregates of relatively high density.A rapid lateral migration of the large protein and polysaccharide species on the surface of the membrane could cause substantial structural perturbation of the BLM, resulting in increased ion leakage through zones of reduced lipid density. This hypothesis was tested by spectroscopic means, using BLM vesicles to study the interaction of pyrene-labelled Con A with glycogen. It was expected that a glycogen - Con A complexation event, associated with a condensation effect, would cause the number of pyrene moieties in close proximity to each other to increase. This i n turn would provide an increased emission of fluorescence from the dimer species compared with monomer emission. However, these experiments did not produce any evidence for a condensation effect that could be indicative o f intermolecular interactions of pyrene.1-t The water-insoluble fluorescent probe DOS has been used to measure the dielectric constant, microviscosity and phase behaviour of various bilayer lipid membranes. 15 It absorbs strongly in the near ultraviolet region and has a moderate emission in the blue to red region, depending on the environmental polarity and viscosity. This is a result of the stabilisation of the excited intramolecular charge-transfer state, which shifts the emission maximum to longer wavelengths with increasing environmental polarity. The fluorescence quantum yield (@*) of DOS increases with increasing viscosity as the value of @+ is governed by the freedom of rotation around the central ethylenic double bond.The polarity indicated by DOS in lipid vesicles is an average value of a significant cross-section of the membrane, as is evident from the heterogeneity of the decay time. The DOS is located in the hydrocarbon region of the bilayers, possibly near the deepest carbonyl groups, extending into the hydro- carbon interior. The spectral shift of the emission maximum of DOS to shorter wavelengths indicates that the addition of cholesterol to PC reduces the polarity of the probe environ- ment. Cholesterol is known to reduce the penetration depth of water within the bilayers to about the level of the glycerol backbone.’” Fluorescence intensity measurements of DOS indicate an exponential increase in viscosity with increasing cholesterol content in PC - C vesicles and are consistent with the ion conductivity results obtained for planar BLMs.17 The spectral properties of DOS in vesicles treated with Con A - glycogen do not indicate that any major average structural changes are induced by the selective interaction.It would seem that re-orientation of the domain structure observed in monolayers does not greatly affect the macroscopic structural properties of the hydrocarbon zone of BLMs. The mechanism responsible for ion current effects induced by Con A - glycogen interactions may therefore be associated with edge effects surrounding the domain structures and/or with elec- trostatic effects located at the surface of the membrane. Macromolecules such as Con A and glycogen are larger than the Debye length at the membrane - solution interface.Displacement of the ions in a double layer at the surface of a BLM in a d.c. voltage clamp would result in a significant change in the ion current measured during the electrochemical experiments. An experiment was therefore designed to study whether the complexation of Con A with glycogen signifi- cantly altered the surface ion concentration. 4-Heptadecyl-7- hydroxycoumarin is a pM indicator dye and can be used to measure the surface potential of BLMs with reference to a given environment. I K I ~ The reference environment can be either the PC or the PC - C vesicles. The surface potential (Y) is calculated from the equation Y = -(pK, - pK,”)2.3RT/FANALYST, JANUARY 154x9. VOL. 114 37 -100, -300 t I C --400 I C A - 500 L--- I 0 25 50 75 100 125 150 Ti me/m in Fig.3. Electrostatic potential - time profiles of monolaycrs at an air - water interface in thc presence o f Con A and glycogen indicating both rapid and long-lived voltage transients. (a) A. addition of C - PC - PS; B, subphase injection of lo---‘ M Ca’+, Mn” ; C. compression to 32 mN m 1 ; D. injection o f 10 ‘I M Con A: and E. injection of5 X 10 lob^ glycogen. ( b ) As for (a) except E. 5 x I O - ~ M glycogen where pK,” (for PC - C) and pK,, are the pH at which the dye is half ionised and R , T and Fare the gas constant, the absolute temperature and the Faraday constant, respectively. The pK,, values of the HDHC dye in PC and PC - C (molar ratio 50 : 5 0 ) are 10.5 and 9, respectively. This would give a surface potential of -89 mV for PC compared with the PC - C vesicles.This is consistent with the observation that the residual ion conductivity through PC - C BLMs decreases as the mole fraction of cholesterol increases. No significant change in the pK,, was observed when Con A or glycogen was added individually to solutions containing PC - C vesicles. The addition of both reagents to the same solution gave variable results, which indicated a change in the pK,, of between 0 and 0.3 units. The surface potential of the vesicles in the presence of Con A and glycogen changed by as much as -20 mV compared with the PC - C vesicles. This indicates that the surface cation concentration may increase in the presence of a reactive species. resulting in a higher concentration of ions available for transport and a correspondingly higher ion current.The BLMs used for the ion current investigations in this work had an intrinsic resistance of approximately l o 7 S2 cm-2. Application of Ohm’s law to the area of the BLMs and to the -20 mV maximum potential indicates that ion current modulation should occur with peak current magni- tudes of about 10-11 A, as is observed experimentally. It should be noted that this experiment again derives only the average properties of BLMs that are not time resolved. The variability of the results obtained from this experiment are probably the result of transient effects, which have been shown by electrochemical and epifluorescence studies to be caused by Con A - glycogen interactions. Lipid Monolayers The anisotropic electric field structure of a monolayer lipid membrane at an air - water interface permits the investigation of surface potential by means of a non-contacting electrostatic probe.”’ The surface potential of a lipid monolayer is usually large.even for uncharged or zwitterionic lipids, owing to the dipolar potential,21 which is commonly of the order of hundreds of millivolts with the electrically positive portion of the field directed towards the acyl chain region. In these experiments, the monolayers were compressed to a pressure of approximately 30 mN m-1 in order to represent the surface tension expected for a BLM.22 Simultaneous measurements of surface pressure, average molecular area and electrostatic potential allowed the resolution of effects that were purely structural from those that were electrostatic.Control experi- ments involving the independent addition of Con A, glycogen and Ca2+ - Mn2+ ions confirmed that these species alone produced no major changes in the lipid monolayers. Fig. 3 illustrates some representative electrostatic profiles for lipid monolayers subjected to a mixture of Con A and glycogen in a 0.1 M KCl electrolyte solution. Short-period (seconds) and long-period (minutes) transient electrostatic potential excur- sions were observed after allowing time for mixing and incubation to occur. No significant surface pressure or molecular area alterations were observed during these experi- ments, indicating that the effect was primarily electrostatic in nature. The voltage transients were not regularly periodic and ranged in magnitude from 15 to 110 mV.All the voltage spikes were of a negative voltage with respect to the PC - PS - C monolayer. indicating that the effect could be responsible for the increased K+ ion current permeability observed during the BLM experiments. A model of the dipolar potential and the associated voltage transients induced by Con A - glycogen interactions was recently proposed by Thompson et a1.2-3 The actual mechanism responsible for the voltage transients is still not known, but our results support the view that the electrostatic switching process is not an artifact of the experiment. The physical processes responsible for dipolar action in lipid membranes can be summarised as follows. (1) Re-orientation of dipoles associated with lipid - steroid molecules in localised domains; (2) alteration of dipole orientations that originate initially with membrane-embedded receptors and aggregates; (3) introduction of dipoles intrinsic to stimulants on complexation with a membrane-embedded receptor; and (4) a combination of the mechanisms described in (2) and ( 3 ) , resulting in an aggregate or complex with either a substantially different dipole magnitude than the initial receptor and/or a substantially different dipole orientation to the plane of the membrane.Concentration Threshold of Response Concentration - response calibration for the action of Con A - glycogen aggregations on BLMs and lipid monolayers required a determination of the loss of Con A from solution on to both hydrophilic and hydrophobic surfaces.An XPS study of the amount of protein lost on to the plastic housing material used for the electrochemical investigations of the BLMs indicated that there was a significant loss (Fig. 4) which was not pH dependent (pH range 5-8). A mixture of 3H-Con A with unlabelled Con A was used to trace both the location and amount of protein deposited on to the Perspex surface. All experiments used the same volume of aqueous solution and it was assumed that -3H-Con A was identical with unlabelled Con A in terms of solution equilibration (30 min) and non-selective surface adsorption. Fig. 5 shows the loss of Con A on to the Perspex walls of the solution compared with the amount remaining in solution. A loss of approximately 95% of the Con A originally in the solution cell used for the BLM studies was observed and this was linearly related to the concentration of the aqueous solution over the entire range of Con A concentrations employed (no surface saturation was obser- ved).A similar experiment to measure the loss of Con A on to a PTFE surface indicated that relatively little protein was adsorbed on to this type of plastic and that it represented only 4% of the total amount of Con A originally available. The calibration of the concentration of Con A in solutionANALYST, JANUARY 1989, VOL. 114 41 0 405 400 Binding energyieV Fig. 4. XPS spectra indicating the presence of a nitrogen signal on the surface o f Perspex after incubation in A , 0.1 M KCI; B, 0.1 M KCI containing 1 0 - (' M valinomycin; and C . 0 . 1 M KCI containing 10 31 Con /I.The large signal for the Con A treatment is indicative of thc high affinity of the protein for non-selective deposition on to the plastic surface 8 - 7 0 m 3 a, Lc L 8.0 5 2 - 9 0 C 0 U a a, - 1 0 0 L 0 In 2 -11.0 - Q 5 -12.0 0 0, 5 .- 13.0 -12.0 -11.0 -10.0 -9.0 -8.0 -7.0 LogICon A] in solution Fig. 5. Correlation o f the amount of 'H-Con A in 0.1 R.I KCI solution to that lvhich was non-selectively adsorbed on to the cxperimental cell used for the electrochemical investigations of BLMs. The results indicate that over 90°% of the Con A in solution is adsorbed on to thc plastic surfaces over the entire concentration range investigated. Thesc results were used to calibrate the concentrations o f Con A in solution that arc givcn in Figs. 6 and 7 permitted an investigation of the existence of a threshold concentration effect for the development of Con A - glycogen ion current transients in BLMs.At the low concentrations of Con A used in solution (maximum 10-9 M), the ion current response of BLMs to Con A - glycogen interactions always appeared as "steps." The frequency of the occurrence of these steps was monitored over periods of 30 min and the results are displajred in Fig. 6 for a Con A concentration range of 1 x 10-9 -1 x 10-1-7 M. A threshold concentration of 10-12 M Con A in solution was required before the ion current steps were ohscrved. and the modulation effect reached a maximum frequency at a Con A concentration of approximately 10-10 M. A maximum response in the ion current step frequency may be related to the precipitin reaction of antibody - antigen binding.In the latter process, quantitative binding increases until stoicheiometric reactant ratios are reached, followed by binding inhibition (or possibly solubilisation in this instance). At present it is impossible to model the exact process g 1.0 c v) C c 0.8 3 0 c .' 0.6 Lc 2. 6 0.4 ?! a, 0.2 U n 3 0- w- > m ._ Y - 10-9 10-10 10-11 10-12 U " 10-8 Con A in solutionimol I Fig. 6. Distribution of the frequency of ion current transients observed for PC - C BLMs exposed to various concentrations o f Con A (glycogen concentration. 10 M) associated with the evolution of a maximum response in the concentration analysis, because the concentration of both Con A and glycogen at the lipid membrane surface is not known ( i . ~ ., their partition coefficients are not known). This problem is currently being investigated because the loss of ion current modulation activity may also provide an insight into the mechanism responsible for the reversal of ion current tran- sients back to lower ion current levels (previously attributed to spontaneous loss of aggregate from the surface or to re- orientation of lipid dipoles in the domain structures). An analysis of the total number of voltage transients observed during monolayer experiments over a period of 1 h also supported the occurrence of a maximum in the frequency of transients as the concentration of glycogen was varied. These results are useful for correlation o f trends with the BLM experiments, but cannot be used quantitatively as the loss of Con A on the trough surface and in the tubing used for the circulation of the subphase was large and variable.One of the interesting analytical properties of the current - time profiles of the Con A - glycogen interactions at BLMs is that the voltage transients may become periodic and reprodu- cible. This effect is usually observed when the concentration of Con A is such that it provides the maximum frequency of transient response. A representative current - time profile for this process is shown in Fig. 7; it occurs only for a very narrow concentration range of Con A (within approximately 5% of the concentration associated with the maximum response frequency). An analysis of the time delays between the successive transients shown in Fig.7 is given in Fig. 8. This figure indicates that the data again suggest the presence of aggregative events, which decrease with time as the concentra- tion of the Con A and glycogen monomers becomes depleted. Mechanism of Signal Generation The ion current that passes through a planar BLM when a low d.c. voltage is applied provides an indirect indication of the ion energy barriers and kinetics associated with the ion translation mechanism. Two physical features of the BLM have been identified as the major parameters influencing the permion current'j: (1) the dipolar potential, which originates from the anisotropy of the polar headgroup alignment of lipids, can be observed during monolayer experiments. It controls both the electrostatic energy barriers and the local surface ion concentration5 at the membrane surface; (2) acyl chain and polar headgroup interactions result in average intermolecular packing - fluidity properties of bilayer mem- branes that correlate well with average molecular area data obtained from monolayer compression experiments.The temperature dependence of ion currents through BLMs provides a measure of the Arrhenius energy barrier to ionANALYST. JANUARY 1089, VOL. 114 39 v , I I .s 0 10 20 30 40 50 60 60 70 80 90 100 110 120 Timeimi n Fig. 7. the frequency of ion current transients Actual ion current - time profile for a PC - C BLM exposed to Con A and glycogen concentrations that induced a maximum response in 0 0 0 uo 0 00 0 0 0 I- 1 ' I 0 20 40 60 80 100 120 Ti me/m i n Fig. 8. Analysis of the timc delays between the successive ion current transients shown in Fig.7, indicating a reduction in the frequency o f the switching events which was consistent with the consumption of the monomeric forms o f Con A and glycogen Table 1. Correlation of bilayer lipid membrane Arrhenius energy barrier with monolayer lipid membrane structural data (glycogen concentration, 10-8 M) Approximate Arrhenius Avcragc concentration energy molecular of Con AIM barrierlmev areainm' lo-" Ik 20% 750 0.42 10 '0 690 0.42 10-9 710 0.42 0 860 5 SO 0.42 5 0.01 permeability. Such energy barrier results correlate well with average molecular area measurements derived from mono- layer experiments. The energy barrier is controlled primarily by the packing - fluidity of lipid molecules within the membranes, but the latter is not exclusively responsible for establishing the magnitude of the transmembrane ion current. The additional factor is the dipolar potential, which appears to be related to the Arrhenius prefactor and apparently controls the availability of ions within the headgroup zone of the lipid membranes.The temperature dependence of the magnitude of the ion currents through BLMs in both the absence and presence of Con A - glycogen complexes was used to calculate the Arrhenius energy barriers shown in Table 1 (ion currents measured during transients). Phase transitions for these membranes did not occur over the temperature range investi- gated and it was therefore assumed that the 6°C temperature range did not cause any major structural changes within the lipid monolayers.The corresponding data show that no macroscopic changes in the surface pressure or average molecular area of the lipids within the monolayers were observed during exposure to similar Con A - glycogen concentrations. It is apparent that surface potential changes occur during monolayer experiments and lead to a reduction in the dipolar potential. This in turn could lead to a reduction in the electrostatic barrier of BLMs to ion permeability. The Arrhenius barrier data samples average the effects within the BLMs and indicate that a reduction in the energy barrier occurs during ion current transients. As the monolayer results confirm that no macroscopic structural perturbations are induccd by the Con A - glycogen aggregates, it must be concluded that the reduction in the Arrhenius barrier is based on microscopic events that are probably associated with the domain structures within the lipid membrane recently repor- ted by McConnell and M o p A redistribution of the domain structures with respect to edge contact would cause little perturbation of the macroscopic structural data obtained from monolayers, but could cause major changes in the ion conductivity by the incorporation of defects between domains.Such defects would probably include areas of low lipid density associated with reduced dipolar electrostatic field strengths and would, therefore, yield high ion conductivity and local surface potential changes. Conclusions A bilayer lipid membrane provides the basis for the develop- ment of a biosensor based on chemoreceptive processes.The Con A - glycogen interaction is one example of a selective binding process that induces large analytical signals in the form of transmembrane ion current changes. These signals are derived primarily from the perturbation of electrostatic potentials and may be related to microscopic changes in the phase structure, although they are not related to significant changes in the macroscopic structural properties of lipid membranes. At low concentrations of Con A, the ion current response is usually in the form of transient ion current steps with a variable lifetime. The frequency of these steps has been shown to be directly related to the concentration of the binding protein in the aqueous solution. The development of reproducible periodic ion current oscillations demonstrates that a digital switching process caused by a selective chemical40 ANALYST, JANUARY 1989, VOL.114 reaction can be activated over a narrow range of reagent concentrations at sub-nanomolar levels. A response frequency analysis can therefore provide information about the analyte concentration; it also provides an alternative to the measure- ment of the conventional analogue signal magnitude. We thank the Natural Sciences and Engineering Research Council of Canada and Allied-Signal Canada for financial support of this work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Hille, B., “Ionic Channels of Excitable Membranes,” Sinauer Associates, Sunderland, Massachusetts, 1984. Krull. U. J . , and Thompson, M., IEEE Trans. Electron Devices, 1985, 32, 1180. Thompson, M., Krull, U. J., and Bendell-Young, L. I., Bioelectrochem. Bioenerg., 1984, 13, 255. Krull, U. J . , and Thompson, M., Biochem. Biophys. Res. Commun., 1986. 141,912. Michaels, D. W., Abramoritz, A . S . , Hammer, C. H . , and Mayer, M. M., Proc. Natl. Acad. Sci. U.S.A., 1976,73, 2852. Edelman, G. M., Yahara, I., and Wang, J . I,.. Proc. Natl. Acad. Sci. U.S.A., 1973, 70, 1442. Safarzadeh-Amiri, A , , J. Photochem. Photohiol., 1988,43,43. Hardman, K. D., and Goldstein, I. J . , in Atazzi, M. Z., Editor, “Immunochemistry of Proteins,” Plenum Press, New York, 1977, Volume 2, p. 373. Edelman, G. M., and Wang, J. L., J. Biol. Chem., 1978, 253, 3016. Becker, J . W., Reeke, G . N., and Wang, J . L., J. Biol. Chem., 1975, 250, 1513. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Yoshikawa, K., and Matsubara, Y., Biophys. Chem., 1983,17, 183. Yoshikawa, K., and Matsubara, Y., J. Am. Chem. Soc.. 1984, 106, 4423. Heckl, W. M., PhD Thesis, Technische Universitat Munchen, 1988, p. 126. Birks, J . B., “Photophysics of Aromatic Molecules,” Wiley- Interscience, New York, 1970, p, 301. Safarzadeh-Amiri, A . , Thompson, M., and Krull, U . J . , unpublished work. Simons, S . A., McIntosh, T. J . , and Latorrc, R., Science, 1982, 216, 65. Thompson, M., and Krull, U. J . , Anal. Chim. Actu, 1982, 141, 33. Fromherz, M. S . , and Fromherz, P . , J. Phys. Chem., 1077,81, 1755, Pal, R . , Petri, W. A.. Jr., Barenholz, Y., and Wagner, R. R . , Biochim. Biophys. Acta, 1983, 729, 185. Gaines, G. L., “Insoluble Monolayers at Liquid - Gas Interfaces,” Interscience. New York, 1966. Krull, U. J . , J. Electrochem. Soc., 1987, 134, 1432. Georgallas, A., Hunter, D. L., Lookman, T., Zuckerman, M. J . , and Pink, D. A . , Eur. Biophys. J., 1984, 11, 79. Thompson, M., Wong, H. E . , and Dorn, W. H . , Anal. Chim. Acta, 1987, 200, 319. Krull, U. J., Thompson, M., and Wong, H. E.. Bioelectro- chem. Bioenerg., 1986, 15, 371. McConnell, H. M., and Moy, V. T., J. Phys. Chem., 1988,92. 4520. Paper 81022 73 I Received June 7th, 1988 Accepted September 2nd, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400033

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 41 Supercritical Fluid Chromatography of Coal-derived Polycyclic Aromatic Hydrocarbons on Packed Columns Ian K. Barker, Jacob P. Kithinji, Keith D. Bartle," Anthony A. Clifford, Mark W. Raynor, Gavin F. Shilstone and Peter A. Halford-Maw Department of Physical Chemistry, University of Leeds, Leeds LS2 9JT, UK The supercritical fluid chromatographic behaviour of coal-derived polycyclic aromatic hydrocarbons using modified HPLC equipment with carbon dioxide ( C 0 2 ) and methanol - CO2 as the mobile phases is described. Factors affecting the separation and retention of these compounds are considered. Keywords: Supercritical fluid chromatography; packed columns; polycyclic aromatic hydrocarbons The tumorigenicity of coal-derived oils has been shown to be due to the presence of polycyclic aromatic hydrocarbons (PAHs)' and there is much current interest in devising rapid analytical methods for PAHs that can be applied to coal processing and combustion products.Coal oils can be screened rapidly for the presence of groups of PAHs by 1 electrochemical methods,2 but complementary separation procedures, which allow more detailed identification, are necessary, as the tumorigenic activities of PAH isomers vary widely. 1 Although high-performance liquid chromatography (HPLC) is commonly used for the analysis of PAH mix- tures,'-s supercritical fluid chromatography (SFC) has a number of advantages over HPLC when the same columns are employed. Above its critical point, a substance such as carbon dioxide (CO,) has properties that make its use as a chromato- graphic mobile phase very favourable, particularly in terms of faster analysis times.Solute diffusion coefficients in super- critical fluids are considerably greater than in liquids; the resulting mass transfer coefficients lead to the generation of considerably larger numbers of theoretical plates per unit time in SFC compared with HPLC. Further, the reduced viscosities of supercritical fluids allow greater mobile phase velocities.6 The separation of PAH standards using packed columns has been described by Gere et al.7 and the possibility of separating coal-derived PAHs by SFC on amino and octadecylsilane (ODS) columns has been reported.8 This paper describes the construction of a packed column SFC system. The retention characteristics of PAHs were investigated using C 0 2 and modified C 0 2 and compared with those obtained by HPLC and GC.The rapid determination of PAHs in a coal tar is demonstrated. Experimental Apparatus A diagram of the supercritical fluid chromatograph con- structed mainly from standard HPLC components is shown in Fig. 1. Liquid C02 (British Oxygen) or C 0 2 - methanol (95 + 5 ) (Electrochem) is supplied from the cylinder dip tube to a Varian 8500 syringe pump through A in stainless-steel tubing fitted with a 5-pm in-line filter. The pump is cooled by circulating ethanol at -2 "C through a 10-m long copper coil wound round the cylinder head. The pump outlet is connected via stainless-steel tubing to a Rheodyne 7010 or 7125 injection valve with 10- or 20-pl sample loops and mounted in a - Rheodyne (7125) .-- ODS 2 column (25 x 4.5 mm i.d.1 Cecil UV detector High-pressure Capillary restrictor cell (>300 bar) in water-bath I Brookes flowmeter ( ~ 8 .8 I m-1) Galaxy computer Fig. 1. Schematic diagram of the supercritical fluid chromatograph * To whom correspondence should be addressed.42 ANALYST, JANUARY 1989. VOL. 114 chromatographic oven (Dupont) maintained above the critical temperature of C 0 2 (31 "C). A 1-m length of the connecting tube is located in the oven to ensure pre-heating. Standard HPLC columns (10 or 25 cm x 4.5 mm i.d.) are joined to the injector and to a high-pressure cell mounted in a Cecil ultraviolet (UV) spectrophotometer located immediately below the oven. The cell (Fig. 2) has a path length of 10 mm and a volume of 8 p1 and is similar to that described by Hewlett-Packardg; polished quartz windows, 5 mm in diameter and 3 mm thick, were held in place by metal end pieces and 0.007 in PTFE ring seals.The outlet of the UV cell was connected to a restrictor to maintain supercritical (>75 bar) conditions. The restrictor was made from either 0.010 or 0.006 in i.d., in o.d. stainless-steel tubing held at 40 "C in a water-bath or a Tescom back-pressure regulator fitted with Buna N O-rings. All the connecting tubing had an internal diameter of 0.010 in and an outer diameter of in. Operation of the Chromatograph Complete filling of the pump cylinder is readily achieved if the pump head is cooled to 0 "C. The system can be operated under the flow control of the pump or under pressure control by means of a pressure transducer and microcomputer (Gemini) interfaced to the pump as described by van Leuten and Rothman.1 0 Pressure programming is achieved by varying the inlet pressure by means of a programme written in BASIC and ramp rates of up to 50 bar min-1 can be achieved. Samples were injected as solutions containing 3-5 g 1 - 1 of PAHs in either dichloromethane or methanol. Carbon dioxide is virtually transparent in the UV region above 190 nm and so (a) Inlet 3.5 crn + , \ Outlet Quartz windows (5 x 3 r n r n ) A / \ / \ \ / / / \ \ \ \ \ 10 rnrn \ \ \ \ / / \ \ \ / \ \ ' 4 PTFE seals Fig. 2. and ( b ) cross-section of the detector cell Diagrams of the high-pressure UV flow cell. ( u ) Detector cell there are few restrictions on the choice of wavelength when using either UV or fluorescence detection.Capillary tubing restrictors were less easily blocked than the back-pressure regulator. Samples Coke-oven tar distillation fractions (anthracene oils) were supplied by British Coal and gasifier tars and coal tar pitches by British Gas. Samples were stirred with HPLC grade dichloromethane or methanol and the solutions filtered before injection. Results and Discussion Injection The best results were achieved by the injection of 10-pl sample volumes. A 20-pl loop gave rise to peak broadening and splitting, particularly at low pressures. The effect was most marked for dichloromethane and toluene solutions and was attributed to an effect well known in HPLC": if a sample is injected in a large volume of solution in a solvent that is stronger than the mobile phase, then the resulting peaks may be distorted.Reducing the sample size generally led to symmetrical peaks. Column Packing The retention of PAHs was investigated for a range of different (standard) HPLC column packings with C 0 2 as the mobile phase (Table 1). Very rapid elution was observed from silica, generally with poor peak shapes which were attributed to adsorption on the active sites (residual silanol groups) on the silica. The retention times were longer on silica modified with aminopropyl groups (normal phase), but poor peak shapes were observed for late eluting compounds; this effect was attributed to the precipitation of PAHs caused by their poor solubility in the mobile phase, particularly at the lower pressures at the end of the column.The best chromatograms were obtained on octadecylsilane (ODS) modified silica (reversed-phase) columns, although the retention of PAHs was similar to that found on diol columns. A typical chromatogram of a number of standard compounds run on an ODS column is shown in Fig. 3. Effect of Modifier, Temperature and Pressure on Retention The addition of methanol to the C 0 2 mobile phase substan- tially reduced the retention times of the PAEls on both the diol and ODS columns (Table 1). For polar solutes this effect is commonly attributed to competition between the modifier and the solute for active sites on the stationary phase.]? However, other effects that can intluence retention are the inter m o 1 e c u 1 a r attraction between the met h a n o 1 and so 1 u t e molecules and the increased solvating power due to the presence of methanol.13 For PAHs, however, the reduced retention is probably caused principally by the last effect.As has been observed previously in SFC, increasing the column temperature at constant pressure results in an increase in retention because of the increase in the free volume of the mobile phase which leads to a reduction in the solubility and a shift in partition in favour of the stationary phase. A 20 "C rise in temperature from 34 "C results i n a 100% increase in the capacity factor, k ' , for chrysene. A further increase in temperature causes an increase in the vapour pressure of the PAHs and, consequently, there is an increase in the concen- tration in the mobile phase, which reduces the value of k ' .A graph of In k' versus the reciprocal of the temperature (1173 shows the typical turnover (Fig. 4). These graphs do not intersect for the PAHs so no selectivity can be induced by changing the operating temperature. 14 This dependence of retention on temperature in SFC has recently been explainedANALYST. JANUARY 1989, VOL. 114 43 Table 1. Retention times of PAHs at 45 "C on various 25-cm long packed columns with supercritical COz and modified C02 mobile phases Retention timeimin Mobile Pressure/ Column phase bar packing Toluene CO, . . . . 148 Silica 0.53 Aminopropyl 0.65 Diol 0.65 Octadecylsilane 0.50 CO2 - MeOH (95 + 5 ) . . 148 Silica - C 0 2 . . . . 215 Aminopropyl - Octadecylsilane - Diol - D i d - Naph- thalene 0.70 1.30 1.10 0.90 0.45 0.73 0.80 0.62 0.55 Fluorene 0.97 2.30 1.70 1.55 0.56 0.84 1.25 0.83 0.80 Phen- anthrene 1.22 4.70 2.65 2.30 0.59 1.06 2.40 1 .00 1.10 Pyrene 1.32 10.70 4.90 4.90 0.68 1.45 5.20 1.98 2.05 Chrysene 1.82 20.00 8.10 8.10 0.83 1.80 9.00 2.98 3.05 ! 5 7 3, 0 2.0 4.0 6.0 Ti me/m in Fig.3. Chromatogram of a standard mixture of 1) naphthalene; (2) fluorene; (3) phenanthrene; (4) fluoranthene; (5(, pyrene; (6) benz- [alanthracene; and (7) chrysene. Column: ODS, 25 cm X 4.6 mm i.d. packed with 5-pm particles. Inlet pressure, 212 bar; temperature, 40 "C: CO, flow-rate. 2.5 I min-1 at STP. Restrictor: 2 m x 0.25 mm i.d. Detector wavelength. 254 nm 1.6 0.8 t -J 0 2.6 2.8 3.0 3.2 IO~T-~IK-I Fig. 4. Graphs of In k' versus 1/T.(0) Pyrene; (A) phenanthrene; (0) fluorene; and (0) naphthalene. ODS column; inlet pressure, 148 bar. Other conditions as in Fig. 3 quantitativelyls: the retention is dependent on the fugacity coefficient (a) of the solute in the supercritical mobile phase 7.04 c 5.42 E .- 1 .- E" 4- 3.80 0 C (u (u .- 4- 4- a 2.18 n u ~ r L 0.55 155 174 194 215 Column inlet pressure/bar Fig. 5. Plots of the retention time of (0) pyrene; ( A ) phenanthrene; (0) fluorene; and (0) naphthalene as a function of the CO, inlet pressure on an ODS column at 40 "C. Other conditions as in Fig. 3 and on the density (p) of the mobile phase. Hence a linear relationship is observed when the total contribution of these terms, i.e., [ln(k') + In(@) - ln(p)], is plotted against the reciprocal of the temperature.The effect of pressure on retention is illustrated in Fig. 5 , which shows that increasing the pressure at constant tempera- ture results in a decrease in retention because of the increased density of the mobile phase, which in turn leads to increased solubility and a shift in partition in favour of the mobile phase. 16 The thermodynamics of this phenomenon have recently been studied by Yonker et al. 17 Comparison of SFC With HPLC and GC for the Separation of PAHs The retention behaviour of PAHs in SFC with C02 on reverged-phase columns is compared in Table 2 with that observed on similar columns in the HPLC mode18 and with retention on a non-polar stationary phase using GC.19 The retention of PAHs in SFC resembles that found in GC and gives smooth graphs when the logarithm of the capacity factor is plotted against either the relative molecular mass or the boiling point of the PAHs (Fig.6). The retention indices given in Table 2 show how the isomeric pairs phenanthrene - anthracene, benzo(b]fluoranthene - benzo[k]fluoranthene, benzo[a]pyrene - benzo[e]pyrene and dibenz[a,h]anthracene - picene either co-elute in SFC or have fairly similar retention, as in GC. There is little sign of the effect of molecular shape on retention in SFC, whereas in reversed-phase HPLC the length to breadth ratio exerts a 5trong influence and the above pairs of isomers are well separated.*() Alkyl derivatives are also well separated from their parent PAHs using SFC.44 0 6 6 . ANALYST, JANUARY 1989, VOL.114 Table 2. Comparison of the retention indices of PAHs in SFC and HPLC with octadecylsilane packings and in GC. The retention index, I , of a compound x is defined by the equation log R, + log R,l Z, = log R, + log R,, + 1 - log R,l where n and n + 1 are the bracketing standards naphthalene, phenanthrene, chrysene and picene SFC HPLC" GCt Mobile phase CH,CN CH,CN (SO+ (90+ -H20 -H20 co2 20) 10) H2 Stationary phase ODs-2 column Spheri- LiChro- 1 Naphthalene . . 2.00 1 -Methylnaph- thalene . . 2.14 Fluorene . . 2.43 Phenanthrene . . 3.00 Anthracene . . 3.00 2-Methylphen- anthrene . . 3.12 Fluoranthene . . 3.28 Benzo[b]- fluorene . . 3.40 Pyrene . . . . 3.48 Benz[ a] - anthracene . . 3.88 Chrysene . . 4.00 4-Methyl- chrysene . . 4.11 Benzo[b]- fluoranthene 4.36 Benzo[k]- fluoranthene Benzo[e]pyrene Benzo[a]pyrene 4.67 Perylene . . . . 4.75 Dibenz[a,h]- anthracene . . 5.00 Picene . . . . 5.00 * Reference 18. t Reference 19. 2 2.00 2.51 3.00 3.00 3.38 3.58 3.92 4.00 4.38 4.43 4.56 4.59 5 .OO sorb ODS 2.00 2.56 3.00 3.00 3.41 3.63 3.92 4.00 4.43 4.67 5.00 sorb VYDAC RP- 18 2.00 2.73 3.00 3.14 3.42 3.62 4.00 4.00 4.40 4.48 4.40 4.63 4.78 5.00 201 TP 2.00 2.73 3.00 3.24 3.49 3.56 4.00 4.00 4.30 4.43 4.25 4.52 4.74 5.00 SE-52 2.00 2.20 2.70 3.00 3.01 3.20 3.45 3.69 3.52 3.99 4.00 4.20 4.43 4.44 4.52 4.54 4.57 4.96 5.00 0.37 i e 5 e 4 * 3 e2 0 1 0 1 0 148 245 343 440 537 Boiling pointPC Fig. 6. Plots of log R, against (0) the relative molecular mass and 0) the boilin point of 1) fluorene; (2) phenanthrene; (3) pyrene; [4) chrysene; 8) benzo[e\pyrene; and (6) picene on an ODS column.Other conditions as in Fig. 3 J I I I I I I I I I 1 I 0 1 2 3 4 5 6 7 8 9 Time/m in Fig. 7. Chromatograms of heavy anthracene oil. (a) SFC on a 25-cm ODS column; CO, inlet pressure, 180 bar. ( b ) GC on a 14-m capillary column. (1) Benzofluoranthenes; (2) benzopyrenes; and (3) pcrylene 0 2 4 6 n5 6 1 I 8 10 12 14 16 ' Ti me/mi n 8 Fig. 8. Pressure-programmed chromatogram (148-292 bar) of coal tar pitch on a 25-cm ODS column at 40 "C. UV detection wavelength, 254 nm. (1) Phenanthrene; (2) fluoranthene; (3) pyrene; (4) benz[a]- anthracene; ( 5 ) chrysene; (6) benzofluoranthenes; (7) benzopyrenes; (8) dibenzoanthracenes and phenanthrenes; and (9) coronene Rapid Analysis of Coal-derived Oils by SFC The rapid analysis of PAH mixtures on packed columns is made possible by the high mobile phase flow-rates, which are a consequence of the low viscosity of supercritical fluids and the high diffusivities of solutes, which in turn result in higher efficiencies per unit time than are possible in HPLC.Moreover, COz is particularly suitable because of its low critical temperature and excellent solvating power for non- polar solutes. These advantages, coupled with the separation of PAH isomers such as chrysene - benz[a]anthracene and the ben- zopyrenes, and of alkyl derivatives from the parent PAHs discussed above, make chromatography with supercritical C02 on reversed-phase columns an ideal method for the analysis of coal-derived oils. Fig. 7 shows the rapid separation of such a mixture, the resolution of the benzofluoranthene and benzopyrene isomers occurring within 9 min and being almostANALYST, JANUARY 1989.VOL. 114 45 1 5 0 1 2 3 4 5 6 7 8 9 1 0 Ti rneirnin Fig. 9. Chromatogram of coal tar pitch on an ODS column (10 cm X 4.6 rnm i.d.) at 40 “C. Inlet pressure, 148 bar; UV detection wavelength, 254 nm. Peak identification as in Fig. 8 as good as that obtained by capillary GC. Programming the column inlet pressure from 148 to 292 bar allows both optimisation of the separation of the lower PAHs (Fig. 8) and elution of the PAHs (Fig. 9) up to coronene (relative molecular mass 300) as the density and hence the solubility in CO? are increased. On short (10 cm) columns, isoconfertic (constant density) analyses extending up to coronene are possible in less than 10 min.For mixtures containing predomi- nantly alkylated derivatives, all the methyl and dimethyl compounds are well separated from the parent PAHs. The support of this work through grants from the Science and Engineering Research Council (SERC) and the British Gas Corporation, and through studentships awarded by the Royal Society of Chemistry, the British Council (J. P. K.) and the SERC (G. F. S.) is gratefully acknowledged. 1 . 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 1.5. 16. 17. 18. 19. 20. References Lee, M. L., Novotny, M. V., and Bartle, K. D.. “Analytical Chemistry of Polycyclic Aromatic Compounds,” Academic Press, New York. 1981. Bartle, K. D., Taylor, N., Pappin, A . , Wallace, S . , and Mills, D. G., Fuel, 1987, 66, 10.50.Bjorseth, A . , Editor, “Handbook of Polycyclic Aromatic Hydrocarbons,” Marcel Dekker, New York, 1983. Wise, S. A., in Bjorseth, A . , and Ramdahl, T . , Editors, “Handbook of Polycyclic Aromatic Compounds ,” Volume 2, Marcel Dekker, New York, 1985, p. 183. Wise. S. A., Benner, A. B., Liu, H., and Byrd, D. G., Anal. Chem., 1988, 60, 630. Bartle, K. D . , Barker, I. K., Clifford, A. A , , Kithinji, J . P., Raynor, M. W., and Shilstone, G. F., Anal. Proc., 1987, 24, 299. Gere, D . R . , Board, R., and McManigill, D., Anal. Chem., 1982, 54, 736. Christensen, R. G., J . High Resolut. Chrornatogr. Chrornatogr. Comrnun., 1985, 8, 824. McManigill, D., Board, R . , and Gere. D . K., Publication No. 43-59.53-1647. Hewlett-Packard, Avondale, PA, 1982. van Leuten, F. T., and Rothman, L. D., Anal. Chem.. 1976. 48, 1430. Snyder, L. R., and Kirkiand, J . J . , Editors, “Introduction to Modern Liquid Chromatography,” Second Edition, Wiley, New York, 1979, p. 805. Levy, J . M., and Ritchey, W. M., J. Chromutogr. Sci., 1986, 24, 242. Raynor, M. W., Kithinji, J . P . , Barker, I. K., Bartle. K. D., and Wilson, I . , J. Chromatogr.. 1988, 436, 497. Chester, T. L., and Innnis, D. P., J . Nigh Resolut. Chromatogr. Chrornutogr. Cornmun., 1985, 8, 561. Bartle, K. D., Clifford, A. A . , Kithinji, J . P . , and Shilstone, G. F., J. Chern. SOC., Faraday Trans. I , in the press. Klesper, E., and Leydendecker, D . , Znt. Lab., 1986, 17. Yonker, C. R . , Gale, R . W.. and Smith, R . D., J . Phys. Chern., 1987. 91, 3333. Wise. S. A., Bennet, W. J . , and May, W. E.. in Bjorseth, A , , and Dennis, A. J . , Editors. “Polynuclear Aromatic Hydrocar- bons: Chemistry and Biological Effects,” Battelle Press, Columbus, OH, 1980, p. 791. Lee, M. L., Vassilaros. D . L., White, C . M., and Novotny, M., Anal. Chern., 1979, 51, 768. Wise, S. A.. Bennet, B. A , , Guenther, F. R . , and May, W. E., J . Chromatogr. Sci., 1981, 19, 457. Paper 8102823 K Received July 13th, 1988 Accepted August 5th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400041

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 47 Determination of Volatile Aromatic Hydrocarbons in Estuarine and Coastal Sediments Using Gas Syringe Injection of Headspace Vapours and Gas Chromatography With Flame-ionisation Detection Alexander Bianchi Environmental Laboratory, Exxon Chemical Compan y, Cadland Road, H ythe, Southampton, Hampshire SO4 4WH, UK Mark S. Varney Department of Oceanography, Building 3, University of Southampton, Southampton, Hampshire, UK A simple, low cost flame-ionisation detection method for the determination of volatile aromatic compounds in estuarine and coastal sediments is described. Headspace vapours are drawn from a modified sample vessel at 80 "C and injected by means of a gas-tight syringe with a valved needle. The method eliminates the difficulties norma I ly encou ntered with solvent extraction and dynamic " non-equ i I i bri u m " heads pace methods.The effect of varying the sample preparation parameters is discussed and results giving the optimised values are presented. Relative standard deviations of less than 2% were achieved for a variety of sub-marine and tidal sediments and the results were found to be superior to those given by an existing solvent extraction method. The limits of detection of the method are below 0.5 pg kg-1 (dry mass) for ten key volatile organic aromatic compounds and the response is linear up to at least 200 pg kg-1. Keywords: Volatile aromatics; estuarine sediments; static headspace; gas syringe; flame-ionisation detection gas chromatography The study of the temporal and spatial variation of volatile organic carbon (VOC) in coastal and estuarine sediments has received comparatively little attention in the past.Geochem- ical studies carried out on the coastal marine environment have tended to restrict themselves largely to the study of metals,' total dissolved carbon' and the concentrations and distribution of higher relative molecular mass hydrocarbons.' Numerous studies in the last area have been carried out by the Ministry of Agriculture, Fisheries and Foods (MAFF), principally to widen the understanding of environmental pollution caused by oil from both shipping and industry.4 Consequently, studies of VOC in sub-marine sediments have been conducted almost exclusively as part of large scale programmes associated with petroleum exploration and pet- rogenesis studies.5 This is perhaps not surprising considering the extensive practical difficulties involved in sampling sedi- ments at depths in excess of 10 m from smaller marine vessels.Further, until comparatively recently, few analytical metho- dologies had been developed sufficiently to enable analyses other than simple "oil content" and alkane fingerprint profiles to be performed on recovered sediments. Examples of newer techniques are provided by Readman et a1.6 who developed a complete method for the analysis of sewage, oil and polycyclic aromatic hydrocarbon (PAH) pollution using a single sedi- ment sample. The term VOC, however, covers a group of compounds ranging approximately from methane to dodecane, which originate from both anthropogenic and biogenic sources and contain a variety of organic groups including alkanes, alde- hydes, furans, alcohols, organochlorines and alkylbenzenes.Benzene and substituted benzene compounds, although thought to be derived principally from man-made sources,7 e.g., fuel oils and gasolines, can also be generated selectively as by-products of biological processes via plant metabolic pathways.8 The major VOCs are benzene, toluene, ethylben- zene, m-xylene. o-xylene, cumene, propylbenzene, 13-5- trimethylbenzene, 1,2,4-trimethylbenzene and 1,2,3- trimethylbenzene. Many studies, including those of Laskin and Goldstein," have pointed to the carcinogenicity of benzene in animals. In addition, the United States Environ- mental Protection Agency (EPA) has defined benzene, toluene and ethylbenzene among its published criterion list of 65 priority pollutants, based on factors such as the frequency of occurrence in water, chemical stability and the structure and mass of the pollutant produced.10 The EPA policy states that there is no scientific basis for calculating safe levels of carcinogens, nevertheless they set certain risk levels for various toxic substances, e.g., the risk that 1 in 100000 people will contract cancer is at the 1 .0 pg 1 - 1 level. Levels exceeding 120 pg 1-1 have been reported in the water column of the Southampton estuary. 11 Therefore, in order to screen sediments for VOC and in particular those compounds classified as priority pollutants, it is desirable to have an analytical method capable of detection at the 1.0 pg kg-1 level (dry mass).The majority of methods developed so far for the determi- nation of hydrocarbons in sediments involve liquid - liquid extraction into a purified colvent, using either sonication or Soxhlet extraction processes.1*,13 However, these approaches are hindered by inadequate and time consuming solvent purification steps,'4 by the need for clean-up stages to remove interfering compounds (i. e., oxygenates) and by Soxhlet extraction times of up to 72 h.15 Once extraction is complete, the analysis is normally conducted using gas - liquid chromato- graphy and high-performance fused silica capillary columns. Accordingly, unless strict quality control procedures are applied throughout the preparation of the sample, contamina- tive effects and volatilisation losses will make the analysis both complicated and difficult.Among the various headspace alternatives, purge and trap (P and T) methods, used largely for aqueous samples,Ih have been applied only rarely to sediment samples, and have been reported to be potentially complex. 17 In view of the difficulties encountered in solvent extraction and P and T analysis, we have investigated a simple manual "static" equilibrium head- space approach using a gas syringe and a re-designed sediment headspace sampling vessel. The developed technique avoids the necessity for solvents and clean-up steps, retains the integrity of the sample, facilitates handling strategies and is an improvement over an existing method for static headspace48 ANALYST, JANUARY 1989, VOL.114 sampling of sediments. 18 Manual headspace sampling by means of a gas syringe has been reported to give poor reproducibility, particularly in its application to the determi- nation of organohalogens.19 However, Croll et af.2" have developed an improved version of this method and demon- strated that relative standard deviations (RSDs) of less than 2% could be attained by strict control of the key method parameters. Equilibrium headspace methods for the analysis of marine sediments have been reported by Burke et af.,21 but a more comprehensive method for the specific determination of VOCs in sediments was subsequently developed by Hunt and Whelan.22 The sampling protocol involves placing 100 g of frozen sediment in a 600-ml capacity metal can fitted with two silicone rubber septa.The can is filled with degassed water, sealed and a headspace created by removing 150 ml of water, which are replaced by an equal volume of helium. The can is allowed to thaw overnight, shaken vigorously for 3 min and heated in boiling water for 30 min. Headspace gas aliquots (1.CL50.0 ml) are then injected either directly with a gas syringe or via gas sampling valves into a gas chromatograph. This method was duplicated in the Department of Oceano- graphy, but was found to present practical difficulties. These included leakage of gas from the can, leakage at the septum - can joints and the adherence of sediment to the roof of the can after shaking, causing subsequent fouling of the sampling syringe needle. The repeatability was found to be inadequate with a typical RSD in excess of 8% for all components (possibly owing to fugitive leaks).Accordingly, the headspace vessel was re-designed in an all-glass construction and a more detailed procedure adopted in order to improve the performance of the method. Good precision can be obtained by using modern gas-tight syringes and by maintaining strict control of the experimental paramet- ers, i.e., temperature and equilibration time. This paper describes a simple and rapid method for the routine determi- nation of volatile aromatic compounds in coastal, estuarine and beach sediments. The method involves the collection of sediment cores, scrapings, etc.. into septum-sealed glass vessels, equilibration at 80 "C in a water-bath and headspace gas sampling with a gas syringe.The test data generated using the proposed method are presented. In addition, the tech- nique has been used successfully in an industrial environmen- tal laboratory for the routine analysis of waste sediments and sludges. Experimental Apparatus A Perkin-Elmer Sigma 3B gas chromatograph with a flame- ionisation detector and an LCT-100 computing integrator - plotter was used together with a fused silica capillary column, SO m X 0.22 mm i.d., of WCOT (BP-1) (0.S-vm film thickness) (SGE, Milton Keynes, UK) under the following conditions: injector temperature, 300 "C; detector temperature, 300 "C; initial column temperature, 60 "C for 5 min then increased at 10 "C min-1 to give a final column temperature of 200 "C held for 1 min; carrier gas, helium; and column head pressure, 25.3 lb in-2 (pressure control).Further apparatus used included the modified sampling vessels (Hampshire Glassware, Southampton, UK) , nominal capacity 850 ml; aluminium foil coated PTFE septa (Perkin- Elmer HS-6 septa modified for use with the sampling vessels); a gas-tight syringe with a valved needle, Pressure-Lok Type A-2, 2.0-ml capacity (Precision Sampling, Baton Rouge, LA, USA): and plunger-in-needle liquid syringes (SGE), 5- and 25-pI capacity. A commercially available "picnic" insulation box (Geeco Coolbox) was used for storage and transportation of the vessels and was packed with dry-ice (Cardice) for sub-ambient cooling. Plastic safety containers (BDH, Poole, UK) were used to stand the sampling vessels upright in the Alu rn inium-faced PTFE septa (16 mm) ,Schott screw caps \ Glass seal spring I oca t i n g " h o r n s " Nominal 150 ml 700 ml Liquid - solid capacity Fig. 1.Bulk sample headspace equilibration vessel box. A thermostated water-bath (Grant Instruments, Cam- bridge, UK) was also required for high-temperature equilibra- tion of the samples. The modified headspace sampling vessel is shown in Fig. 1. Reagents De-ionised water, containing less than 10 ng 1-1 of total aromatics, was used. It was purged with filtered ultrapure nitrogen 24 h prior to use. Undecane. Redistilled, containing less than 10 ng 1-1 of total aromatics. Aromatic standards. Benzene, toluene, ethylbenzene, m-xylene, o-xylene, cumene, propylbenzene, 1,3,5-trimethyl- benzene, 1,2,4-trimethylbenzene and 1,2,3-trimethylbenzene were of chromatographic grade.Sodium azide. AnalaR grade. Stock standard solution. A combined solution containing 10 mg 1-1 of each of the ten aromatic hydrocarbons (each added with its own syringe) in undecane was prepared. The standard preparation method was based on a published CONCAWE method.23 The solution remained stable for at least 1 week if stored at <4 "C in a glass-stoppered flask. Dilute standard solutions. Prepared using further syringes to dilute the solutions serially to give concentrations of 1,10,20,50,70 and 100 pg 1-1 in undecane. Standard .sediment. The optimum matrix to use has been stated to be that of the matrix itself.17 Sediment was taken from a relatively unpolluted estuarial site on the Southampton estuary.The sediment was subjected to rigorous clean-up procedures in order to remove all the volatile organic compounds. These included solvent extraction methods nor- mally used for sediment clean-up,24 in addition to boiling, stripping with nitrogen and washing with water prior to drying in an oven at 105 "C for 72 h. It was recognised that these procedures could destroy or modify the adsorptive sites within the grain - particle matrix; however, a suitable compromise was sought between the necessity for a representative "blank" sediment and the physico-chemical integrity of the original sediment itself. Method Batches of sediment (100-150 g) were cooled in a desiccator and added gravimetrically to the sample vessels, which were then sealed and purged with ultrapure nitrogen for 3 min.ANALYST, JANUARY 1989, VOL.114 49 Known aliquots of the standard solutions were spiked into the vessels (via the septum inlets) and the sediment was shaken gently to incorporate the spiked material. Blanks were prepared by spiking the sediment with undecane only. Standards were analysed immediately after preparation. The exact concentration (in pg kg-1) of the aromatic hydrocarbons was calculated from the initial concentration of the liquid standards and the exact mass of sediment taken (both known accurately to four significant figures). Sample Collection and Storage The sample vessels were cleaned with detergent, washed with acid and water and stored overnight at 150 "C. New septa were fitted to the vessels for each sample. Sample cores and scoops were taken (with Van-Veen sediment grabs for shipborne sampling) and approximately 100 g were placed in each vessel.The vessels were capped and stored on a dry-ice bed inside the collection box. Sodium azide was added (approximately 0.5 g) Table 1. Analytical procedure Step Experimental procedure Notes 1 Conduct steps 2-14 for duplicate calibration standards. Plot mean integrator counts (or peak heights) against correctcd con- centrations on calibration graphs for each standard. Inte- grator counts should not differ by more than 2-3% of each other Begin with low concentration standards. A linear plot should be obtained. Intermediate stan- dards may be deleted later if linearity is reproducible. A computing integrator can be programmed to identify and calibrate all peaks 2 Samples are removed from the insulation box.Standards and blanks are removed from ref- rigerated storage. Ensure the securing springs are attached to the glass mountings Samples and blanks should be analysed immediately after removal from the refrigerator or from coolbox storage 3 Remove one septum cap and decant 500 ml of de-ionised water into the vessel This step should be executed promptly. A pre-cleaned glass funnel can be used 4 Purge ultrapure helium into the A flow-rate of approximately 200 ml min-1 should be suffi cient to achieve a helium atmosphere headspace. Use only pre-cleaned stainless- headspace above the water level through a stainless-steel tube for 1 min 2 5 s. Refit the septum cap and switch off the helium steel tubing 5 Allow frozen sediment samples This step was unnecessary with to thaw at room temperature for 1 h k 2 min standards and blanks, unless they were re-frozen 6 Agitate the headspace vessel for approximately 5 rnin to achieve dispersion of the solid phase into the water phase Vigorous shaking is not neces- sary 7 Immerse the headspace vessel containing the blank, standard or sample in a hot water-bath pre-set at 80 "C.The vessels can be clamped to avoid instability in the bath. An aluminium lid with apertures cut to support up to eight vessels is a recommen- ded accessory The water line should extend to at least 75% of the height of the vessel. The bath temperature will drop by about 5 "Con entry of the vessel depending on the geometry of the bath and the mass of water 8 Allow the headspace vessel(s) to equilibrate in the bath for 45 k 1 min Agitate the vessel(s) every 10 rnin for approximately 1 min to poison any biological processes occurring in the sediment and principally to minimise the degradation of organic compounds by bacteria.On returning to the laboratory, the box can be replenished with dry-ice if short-term storage (i. e., 2-3 h) is intended prior to analysis. Overnight storage in the deep-freeze compartment of a refrigerator is possible, although same-day analysis is recommended whenever prac- tical. Analytical Procedure The analytical procedure is given in Table 1. Results and Discussion A headspace gas sample volume of 2.0 ml was found to provide a satisfactory chromatographic peak area response (>lo00 pV s) at the 1.0 pg kg-1 level for a range of sub-marine, estuarial and beach sediments.Calibration graphs were linear from 0.1 to 200 pg kg-1. Step Experimental procedure 9(a) During equilibration of the headspace check the syringe by injecting a 2.0-ml aliquot of nitrogen into the gas chromato- graph. Draw 2.0 ml of nitrogen slowly into the syringe, close the syringe valve and slide the needle into the injection port of the gas chromatograph ( h ) Compress the gas against the valve to the 1.0-ml mark, open the valve and inject the gas to the 2.0-ml limit. Switch on the integrator - plotter. Use this technique for all injections. Repeat step 9(a) if contami- nant peaks are found 10 Conduct steps 3-14 for succes- sive blanks. Sample the head- space gas directly from the vessel immersed in the bath.It should not be necessary to remove the bath 11 Equilibrate and analyse each sample (steps 2-14). Include a quality control standard every two analyses. Calculate the absolute concentration of each aromatic hydrocarbon 12 Allow the vessel to cool after removal from the bath. Remove the headspace vessel cap and filter the sediment sample into a pre-weighed Pyrex glass drying dish (300-ml capacity) 13 Dry the sediment in an oven at 105 "C for 8-12 h. Place in a desiccator and allow to cool for 1-2 h, then re-weigh 14 Calculate the mass of aromatic hydrocarbon relative to the mass of sediment (dry mass) recovered for each sample. Express the concentration as pg kg (dry mass) Notes This ensures that the syringe is free from organic contami- nants which would generate "ghost" peaks on the chro- matogram. This procedure should be accomplished in one smooth step Set plotter - chart recorder at 10 mm min-1.An integrator delay can be used to ignore both pressure and air peaks which may appear at the start of the chromatogram. The integrator can be set up to start the GC run automat- ically A satisfactory blank should contain <1 pg kg of each aromatic hydrocarbon. If levels in excess of this are recovered, check the syringe and, if necessary, check with a third blank sample Plot - integrate as described in step 1 Weigh the drying vessel to within 0.001 g. Spread the sediment evenly across the vessel surface. Agitate and stir during drying Continue to agitate the sedi- ment to assist the removal of water This calculation should be programmed into the comput- ing integrator if possible50 ANALYST, JANUARY 1989, VOL.114 Table 2. Results of replicate analyses of standard sediment(s) expressed as relative standard deviation (YO) Component Benzene . . . . . . Toluene . . . . . . Ethylbenzene . . . . m-Xylene . . . . . . o-Xylene . . . . . . Cumene . . . . . . Propylbenzene . . . . 1.3.5-Trimethylbenzene 1.2,4-Trimethylbenzene 1.2,3-TrimethyIbenzene No. of samples: Concentration of standard in sediment(s)/pg kg-1 (dry mass) 1 10 20 50 75 100 . . . . 1.8 1.8 1.9 2.0 1.9 1.8 . . . . 1.8 2.0 2.0 2.1 2.1 2.0 . . . . 2.0 2.2 2.2 2.2 2.3 1.9 . . . . 1.5 1.8 2.2 1.8 1.7 1.7 . . . . 1.5 1.7 2.3 1.8 1.6 1.7 . . . . 2.8 2.5 2.5 2.7 2.0 1.9 . . . . 1.9 2.2 2.2 2.3 2.2 2.2 .. . . 1.5 1.9 1.9 1.4 1.5 1.4 . . . . 1.6 1.8 1.8 2.0 2.0 2.1 . . . . 1.7 1.8 1.8 1.9 2.2 2.0 9 9 5 8 5 9 c a, - I- A C ID II 0 5 10 Timeimin 15 Fig. 2. Headspace analysis chromatogram for 10 pg kg-1 of each aromatic compound. (A) Benzene; (B) toluene; (C) ethylbenzene: (D) m-xylene; (E) o-xylcnc; (F) cumene; (G) propylbenzene; (14) 1.3.5-trimethylbenzene; (I) 1,2,4-trimcthylbenzene; and (J) 1.2,3- trime t hylbenzenc The precision of the method was evaluated and the data obtained are presented in Table 2. A specimen chromatogram of the 10 pg kg-1 calibration standard is shown in Fig. 2. The total GC run-time is just under 15 min per sample; however, the analysis is run up to 200 "C in order to remove higher boiling compounds from the column. Note that undecane elutes at approximately 17.0 min under the standard condi- tions described under Apparatus.Standards were spiked with known aliquots of higher relative molecular mass organic compounds, e.g., naph- thalene, to determine if there were any quantitative interfer- ence effects on the calibration values. Subsequent analyses yielded data falling within the RSD values given in Table 3. Comparison of the headspace method with an existing dichloromethane extraction technique13 yielded an average 260% increase in efficiency, expressed as actual peak area, for Table 3. Replicate analyses of a mid-estuarial sediment sample Aromatic hydrocarbon Foundipg kg- 1 RSD, Yo Benzene . . . . . . Toluene . . . . . . Ethylbenzene . . . . m-Xylene . . . . . . o-Xylene .. . . . . Cumene . . . . . . Propylbenzene . . . . 1,3 ,S-Trimethylbenzene 1,2,4-TrimethyIbenzene 1,2,3-Trimethylbenzene . . . . . . 8.4 . . . . . . 49.3 . . . . . . 7.1 . . . . . . 4.7 . . . . . . 2.2 . . . . . . 1.2 . . . . . . 0.6 . . . . . . Not detected . . . . . . 0.5 . . . . . . 0.5 1.9 2.0 1 .5 1 . 1 1.2 1.1 1.9 1.8 1.8 - benzene, toluene and ethylbenzene and an increase of >I%% for the remaining seven aromatic compounds. The limit of detection of the method was defined as recommended by Grob and Kaiser14 and Kolb et aZ.,25 i.e., as the smallest amount of sample that will cause a measurable signal (e.g., twice the noise) over the noise signal. This is also known as the minimum detectable level (MDL). The detector specificity is the ratio of the detector response for a contami- nant to that for the desired component.Using this definition, the MDL -1alues were as follows: benzene, 0.009; toluene, 0.01: ethylbenzene, 0.07; 0- and m-xylene, 0.08; cumene, propylbenzene and 1,3,5-trirnethylbenzene, 0.17-0.23; and 1,2,4-trimethylbenzene and 1,2,3-trimethylbenzene, 0.33- 0.36 ug kg-1. Experiments were carried out to determine whether the natural salt content in the sediments would contribute to an unquantifiable "salting-out"effect. Sea water salinity values overlying the sediments varied from 10 to 33.6 parts per thousand at the seaward end of the Southampton water estuary. Although it was necessary to employ the salting-out effect in the method (which leads to an increase in the concentration of non-polar or low polarity compounds in the vapour phase by the addition of a soluble electrolyte to the liquid phase), interstitial water in the pores of the sediment will contain salt.Consequently, one of two identical specimen samples was analysed for its volatile aromatic content. A prepared blank sediment was spiked to give an equivalent concentration of benzene, toluene, o-xylene and 1,2,3- trimethylbenzene. The second specimen sample and the spiked blank were spiked further by "known additions'' of these four key aromatic hydrocarbons. On analysis, the resulting data, corrected for the mass of sediment taken, yielded results that were within 2.5% for all components, indicating that the original specimen samples were not contributing a salting-out effect at levels that were sufficient to interfere with the calibration of the method.Further analyses on samples taken from high- to low-salinity regions provided similar results. It was concluded that the comparatively small contribution of "salt" from the sediment relative to the large volume of water added, i.e., 500 ml, effectively negates any salting-out effect that may begin to influence the analysis at signficantly greater sediment volumes or smaller water volumes. Investigations were conducted into the effect of equilibra- tion-bath temperatures on peak area recovery data. Standards were equilibrated at 10 "C intervals from 30 to 90 "C for 45 k 1 min before analysis. The recovery was optimised for benzene, as this is the component of greatest environmental concern. The maximum recovery of benzene and toluene was achieved at 80 "C, although the recovery of the remaining compounds levelled off between 55 and 65 "C.The temperature was then varied at 2 "C intervals, i.e., 78, 80, 82 and 84 "C, respectively. From the plots of log(peak area) versus the reciprocal of the absolute temperature, changes in peak area of between 2 and 3% resulting from a 2 "C change in temperature were observed (2.6% for benzene and <2% for the other aromatic hydrocar- bons). These results are in broad agreement with those ofANALYST, JANUARY 1989, VOL. 114 Croll et ~ 1 . 2 ~ ) and suggest that temperature stability to within +2 “C is acceptable. It was calculated that the water-bath thermostat unit was capable of operating within these limits. Owing to the relatively high combined mass of the sample and vessel. stable equilibration was not fully achieved until a minimum of 30 min had expired. Equilibration times of between 35 and 60 min were found to provide repeatable recoveries.To achieve the best compromise between stable equilibration and total analysis time, an equilibration time of 45 min was selected as the optimum. Two further aspects of the technique were investigated, viz., the effect of varying the sample to water volume ratio with respect to mass peak area recovery and the effect of performing repeated injections from the same vessel, known as multiple headspace extraction (MHE). It was found that an increase in the mass of sediment relative to the volume of water added enriched the concentra- tion of organic vapour in the headspace, mainly owing to the high mass and a secondary, smaller, salting-out effect.However, estuarine samples actually contain a wide range of organics and this is reflected in the more complex chromato- grams obtained. Therefore, the resulting peak area is much larger than that necessary for accurate quantification at the 1 .0 pg kg-1 level. Frequently, complex chromatograms require the use of GC - MS to elucidate the identities of the organic components. Although we have used GC - MS in conjunction with headspace analysis, this is not necessary for complete quantification and identification using the given method parameters. The use of MHE, as described by McAuliffe,26 allows the analyst to define the distribution coefficient of a particular compound or range of compounds.It provides a qualitative mechanism €or the identification of trace amounts of com- pounds as part of a primary analysis. The nature of the distribution coefficient indicated by successive equilibrations provides information on the behaviour of the compound. Volatile aromatic hydrocarbons have a low distribution coefficient compared with alkanes and cycloalkanes, enabling them to be recovered quantitatively after three equilibrations. The application of MHE to a variety of sample matrices, e.g., in the analysis of halogenates and aromas has been discussed by Kolb et al.27 The use of MHE in this method is therefore limited to a theoretical appreciation. Our results indicated that a second 2.0-ml aliquot could be withdrawn from the headspace with no major alteration in the peak recovery within 1 standard deviation.However, removal of a third aliquot resulted in a reduction in the benzene peak area of at least 5%. We would not therefore recommend the use of MHE beyond a second injection from the same headspace sample. However, the analysis of headspace aliquots (single head- space extraction) from replicate samples generated reprodu- cible data to within an RSD of 2% (four replicates). Each aliquot was injected by separate analysts in both a University and an external environmental laboratory, including an operator with no prior experience of gas chromatography or syringe handling. The data obtained are shown in Table 3. Conclusions The comparative simplicity of the headspace vessel and the sampling preparation technique minimise the problems asso- ciated with handling bulk sediment samples.The technique is designed to circumvent contamination and component losses by sampling up to the moment of headspace injection. Samples can be taken. prepared and analysed by relatively inexperienced staff with the minimum of training. Any sources of error are negated by avoiding the use of solvents or ancillary sample handling steps. The sensitivity and precision of the method are superior to those of an existing standard solvent extraction method. 13 Temperature control, equilibration time and syringe sampling techniques are the nucleus of the method 51 and accurate reproduction of the method parameters will maintain the integrity of the method. The proposed method is now in use in the Exxon Chemical Company Environmental Laboratory, where it is applied to the rapid sampling and analysis of sludges, muds and solid wastes. Over 150 coastal and estuarine samples have also been analysed successfully using the method, and valuable environ- mental data on the flux of key aromatic compounds have been obtained.Additional experiments have also led to a broaden- ing of the scope of the method and to its use in the simultaneous determination of other organic compounds including alkanes and cycloalkanes. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. References Clark, R. B., “Marine Pollution,” Clanden Press. Oxford, 1986, p. 173. Froelich, P. N., Limrzol. Oceunogr., 1980, 25, 564.Page, D . S., Foster, J. C.. Fickett, P. M., and Gilfillan. E . S., Mar. Pollut. Bull., 1988, 19, 107. Law, R. J . , Mar. Pollut. Bull., 1981, 12, 153. Hunt, J . M., Huc, A. Y., and Whelan, J . K., Nature (London), 1980, 288, 688. Readman, J . W., Preston, M. R., and Mantoura, R . F. C . , Mar. Pollut. Bull., 1986, 17, 298. “Benzene: Control of Toxic Substances in the Atmosphere,” OECD Environment Monograph No. 5 , OECD, Paris, 1986. “Evaluation of Benzene Toxicity in Man and Animals,” DGMK Project 1 7 6 6 , German Society for Petroleum Sciences and Coal Chemistry, Hamburg, 1980. Laskin, S., and Goldstein, B. D., “Benzene Toxicity,” Ameri- can Petroleum Institute Report, Hemisphere Publishing, Washington, DC. 1977. Grob, R . L., in Grob. R . L., Editor, “Modern Practice o f Gas Chromatography,” Second Edition, Wiley, New York, 1985, Chapter 10. Knap. A. H., Le B . Williams, P. J . , and Tyler, I.. Nature (London), 1979. 279. 517. Murray, D. A. J . , J. Chromatogr., 1979, 177. 135. Venkaste, M. I., Ruth, E., and Kaplan, I. R . , Mar. Pollict. Bull., 1986, 17, 554. Grob, R. L., and Kaiser, M. A., “Environmental Problem Solving Using Gas and Liquid Chromatography,” Journal of Chromatography Librury , Elsevier, Amsterdam, 1982, Vol- ume 21, p. 91. “Determination of Very Low Concentrations of Hydrocarbons and Halogen Hydrocarbons in Water 1984-1985,” HM Stationery Office, London, 1985. Grob, K., and Zurcher, F., J. Chromatogr., 1976, 117, 285. Pizzie, R . , PhD Thesis, University of Southampton, 1984. Hunt, J. H., and Whelan. J . K., Org. Ceochem., 1979, 1,219. Otson, R., Williams, D. T., and Bothwell. P. D . , Environ. Sci. Technol., 1979, 13, 936. Croll, B. T., Sumner, M. E.. and Leathard, D. A , , Analyst, 1986, 111, 73. Burke, R. A , , Jr., Brookes, J . M., and Sackett, W. M.. Geochim. Cosmochim. Acta, 1981, 45, 627. Hunt, J . H . , and Whelan, J . K., Geochim. Cosmochim. Acta, 1980.44, 1767. CONCAWE Report No. 8/86. CONCAWE (“Oil Companies International Study Group for the Conservation o f Clean Air and Water”), The Hague, The Netherlands, 1986. Lee, H.-B., Hong-You, R. L.. and Chau, A.S.Y.. Analyst, 1986, 111, 81. Kolb. B.. Kraub, H . , and Aucr. M., “Pcrkin-Elmer Applica- tions Paper 21/1978/I-ISA-21,** Perkin-Elmer, Buckingham- shire, 1978. McAuliffe, C.. Chemrech, 1971, January, 4. Kolb, B., Pospisil, P., and Auer, M., Clzromatogruphia, 1984, 19. 113. Paper 8103 181 I Received August 3rd, I988 Accepted September 22nd, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400047

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 53 Development and Optimisation of a High-performance Liquid Chromatographic Assay for Tioconazole and its Potential Impurities Part 11.* Selection of Detection Conditions for Potential Impurities Adrian G. Wright University of Bradford, Bradford BD7 1 DP, UK John C. Berridget Pfizer Central Research, Sandwich CT13 9NJ, UK Anthony F. Fell University of Bradford, Bradford BD7 1 DP, UK An optimised high-performance liquid chromatographic separation developed for the assay of tioconazole and its potential impurities has been applied to real-world samples where tioconazole is in excess. As severe peak tailing interferes with the assay of two of the impurity peaks, changes in detection wavelength have been examined as a means to discriminate between this interference.The resulting enhancement of resolution has been exploited in the optimisation of analysis time. Keywords: Tioconazole; detection wavelength optimisation; United States Pharmacopeia The resolution, R,, between two adjacent chromatographic peaks is given by: k’ 1 + k’ R =’-. ( a - l) fl.- s 4 a where a is the selectivity factor between the two compounds, N is the column plate count and k‘ is the average capacity factor for the two peaks. For these studies the column is specified and N is, therefore, fixed. The selectivity factor, a, may be optimiscd with respect to column chemistry and mobile phase composition.’ As it is assumed that changing the proportion of the aqueous phase does not affect the selectivity between two peaks, increasing the average capacity factor, k ’ , can be exploited to improve resolution.However, the degree of improvement becomes progressively less significant once k’ exceeds The principal variables influencing the separation of tioco- nazole and its potential impurities (Fig. 1) have been identified and optimised with respect to selectivity in previous work.3 For convenience, the sample used for method develop- ment contained all components (tioconazole and the four potential impurities; each between 40 and 80 pg ml-1) at comparable levels. However, this does not reflect the real- world situation where, under United States Pharmacopeial (USP) regulations,4 the bulk drug must contain a minimum of 97% tioconazole. As the impurities are present at such low levels the detection sensitivity required is high.In fact, real-world samples analysed using the optimised separation conditions, and under these constraints of high sensitivity, yield chromatograms where tailing of the tioconazole peak is so extensive as to interfere with the determination of related compounds B and C (Fig. 2). Clearly the sloping base line would not be acceptable for a routine assay of these compounds. A solution to this interference problem was sought by optimising the detection conditions to increase the discrimination between tioconazole peak tailing and the signals for the two impurity peaks. It is common practice to adopt a single detection wavelength which fulfills the twin requirements of satisfactory sensitivity for all peaks of interest and maximum discrimination against interference due, for example, to peak tailing.However, few reports on the detailed assessment and validation of this strategy have appeared in the applications literature. The possibilities of increasing the average capacity factor for enhancing resolution and overcoming interference of the tioconazole peak were also examined. /-N CH- I cHz-Nd 0 cld \ I R Tioconazole R = \ = cHzo Related compound A 1 - { 2-[ (3-t h ien yl )met hoxyl-2- (2,4-dichlorophenyl)ethyl}imidazole \ Related compound B 1 - { 2 - [ ( 2,5- d i c h I o ro -3-t h i e n y I 1 m e t h ox y 1 -2 - (2,4-d i c h I o ro p h en y I ) et h y I } i m id azo I e R = CI I \ Related compound C 1 -{ 2-[(5-bromo-2-chloro-3-thienyl)methoxy]- Br 2-(2,4-dichlorophenyI)ethyl}imidazole R = cHh cI s Compound D (hydrolysis product) 1 -( 2,4-d i c h I o ro p h e n y 11-24 i mi dazo I - 1 - y I )et h a no I R = H * For Part I of this series see reference 3.-: To whom correspondence should be addressed. Fig. 1. and hydrolytic degradation product D Structures for tioconazole, related compounds A, B and CANALYST, JANUARY 1989, VOL. 114 A 0 20 timin Fig. 2. Chromatogram for bulk tioconazole run with optimum mobile phase and a detection wavelength of 220 nm. Interference between tioconazole peak tail and related cornpounds B and C is extensive. Eluent, (methanol - acetonitrile (70 + 30)] - pH 4 triethylamine phosphate buffer (0.05 M) containing 1-octanesulphonic acid (0.025 M) (54 + 46 VW) (this is the optimum eluent from reference 3); flow-rate, 1 .5 ml min- 1 The examination of wavelength to enhance discrimination between tioconazole and compounds B and C was primarily concerned with locating the best compromise wavelength which provided good discrimination while still being able to detect <1% mim of each of the potential impurities.Experimental A Hewlett-Packard 1040A diode array detector, a Hewlett-Packard 8SB microcomputer, a Hewlett-Packard 7470A plotter (Hewlett-Packard, Wokingham, UK) and an LDC constaMetric 3000 pump (LDC UK, Stone, UK) were used. The sample injection valve was a Rheodyne 7010 fitted with a 20-pl loop (Alltech Associates, Carnforth, UK). The column was 5-pm Hypersil phenyl (150 x 4.6 mm i.d.) (Technicol, Stockport, England). Mobile phases were pre- pared from HPLC-grade solvents (Rathburn Chemicals, Peebles.UK), HPLC-grade 1-octanesulphonic acid (Fisons, Lough borough, UK) and reagent-grade trieth ylamine (Hop- kin and Williams. Chadwell Heath, UK). The buffer pH was adjusted using reagent-grade phosphoric acid (BDH, Poole, UK). The flow-rate used throughout was 1.5 ml min-1. Tioconazole, the three related impurities and the degradation product (Fig. 1) were supplied by Pfizer Central Research, Sandwich, UK. Results and Discussion The UV - visible spectra for tioconazole and related com- pounds B and C (Fig. 3) were assessed for differences which could be exploited. The spectra for B and C were essentially identical and therefore only one spectrum was compared with that of tioconazole. Differences between the spectra were assessed by calculating the ratio of absorbance for B to that for tioconazole at the same wavelength.Plotting this ratio over the wavelength range of interest (220-270 nm) permitted recognition of the wavelength producing maximum discrimi- nation between compounds B and C and tioconazole (Fig. 4). A graph of ratio against wavelength revealed improved discrimination between tioconazole and compounds B and C above 260 nm, while still retaining a significant absorbance for 190 240 300 190 240 300 Wavelengthinm Fig. 3. (2), B (3) and C (4), and hydrolytic degradation product D ( 5 ) UV - visible spectra for tioconazole ( I ) , related compounds A 220 230 240 250 260 270 Wavelength/nm Fig. 4. Ratio of absorbance for compound B and absorbance tor tioconazole plotted against wavelength.Maximum discrimination is revealed by an increase in ratio above 250 nm these potential impurities. A study of the UV - visible spectra for potential impurities A and D (Fig. 3) indicated only low UV absorbance at 260 nm, which presented difficulties for the quantification of the components. Detection wavelengths above 260 nm were observed to yield greater discrimination between tioconazole and compounds B and C, but also caused a further reduction in the signals for A and D, which were already very low. As a result 260 nm was investigated as the detection wavelength of choice. An overriding consideration in any final decision on detection wavelength was the ability to detect and measure low levels of all four potential impurities. It was considered undesirable to develop a method which minimised the interference from tioconazole so that B and C could be measured, if as a consequence A and D could not be detected.On transferring the separation to the 15-cm column it was found that the optimum separation conditions reported in the earlier study7 yielded a longer analysis time than was desirable. The analysis time was therefore adjusted by decreasing the proportion of aqueous buffer. The proportions of methanol and acetonitrile were kept constant relative to each other (70 + 30) to avoid any loss of selectivity. Studies revealed that changes in the proportion of buffer ranging between 40 and 55% had only a marginal effect on the selectivity between tioconazole and compounds B and C. Two mobile phases were studied to assess the enhancement of resolution between tioconazole and compounds B and C that could be achieved by increasing the average capacity factor.ANALYST. JANUARY 1989, VOL.114 55 Table 1. Regression data for spiked tioconazole standards and external standards for the two mobile phases studied Correlation coefficient Intercept Mobile phase* Standard+ Impurity 220 nm 260 nm 220 nm 260 nm 1 ES A B C D 1 BULK A B C D 2 ES A B C D 2 BULK A B C D 3 . 9 9 4 >0.997 0.010 0.013 0.004 -0.023 H . 9 9 9 >0.999 -0.016 -0.244 -0.239 0.011 >0.998 >o. 998 0.020 0.016 '0.01 1 0.040 >0.999 >0.998 -0.002 -0.280 -0.247 -0.016 ' 1 = 42% aqueous buffer and 2 = 48"h aqueous buffer. + ES = external standards (containing no tioconazole) and BULK = spiked bulk tioconazole. -0.038 0.010 0.006 -0.035 -0.034 -0.275 -0.266 0.007 -0.039 0.012 0.007 0.02 1 - 0.025 -0.304 -0.276 - 0,020 The first eluent consisted of [methanol - acetonitrile (70 + 30)j - triethylamine phosphate buffer (0.05 M) containing l-octane- sulphonic acid (0.025 M) and adjusted to pH 4 with phosphoric acid (58 + 42 VW).This eluent resulted in a retention time of 11 min for the last peak and an average capacity factor of 9.7 for B and C. The second eluent contained a higher buffer content (52 + 48 ViV) and resulted in a retention time of 13.6 min with an average capacity factor of 11.7 for B and C. The purpose of this assay was to detect and measure accurately low levels of the four potential impurities in the bulk tioconazole. Two sets of standards were run for both mobile phases to establish whether this could be achieved with a detection wavelength of 260 nm compared with the higher sensitivity available at 220 nm.One set of standards consisted of tioconazole (3.6 mg ml-'), spiked with known amounts of the four potential impurities (&55 pg ml-I), equivalent to 0-1.5% of tioconazole concentration, and made up in mobile phase. The second set of standards consisted of external standards containing the four potential impurities at identical concentrations (0-55 pg ml- I ) , but omitting tioconazole. The samples were dissolved in mobile phase to avoid the base-line disturbances, which resulted if another solvent were injected. It was assumed that these deviations resulted from disturbance of the columnhon-pairing agent equilibrium. Therefore, it was necessary to make up two sets of standards for both mobile phases.The linearity of response was assessed for standards with and without tioconazole for both eluents, at both detection wavelengths (260 and 220 nm). The 220-nm wavelength was selected for convenience, on the basis that any slight differ- ence in performance at the official USP detection wavelength (219 nm) would be marginal. Regression analysis of the results (Table 1) yielded good linearity for the two detection wavelengths with both sets of standards. The intercepts for the two detection wavelengths were not significantly different. A study of the error intervals associated with the intercepts revealed that, for any one set of standards, the overlap between the 95% error intervals for the two detection wavelengths was at least 83%, and in most instances, 100%.The results also revealed that A and D both retained a sufficient absorbance at 260 nm to be detected and quantified. A significant improvement in the discrimination between tioconazole and potential impurities B and C was observed with a detection wavelength of 260 nm compared with 220 nm (Fig. 5 ) . It was therefore concluded that a detection wavelength of 260 nm was suitable for this assay. I ' Tioconazole D B 5 10 timin Fig. 5. Overlay of chromatographic traces for detection wavelengths of (-) 220 and (--) 260 nm. Enhancement of discrimination between tioconazole and compounds B and C with a detection wavelength of 260 nm is clear, as is the loss of signal for compounds A and D.Eluent: [methanol - acetonitrile (70 + 30)] - pH 4 triethylamine hosphate buffer (0.05 M) containing 1-octanesulphonic acid (0.025 MY (58 + 42 ViV); flow-rate, 1.5 ml min-1 A comparison of the separations achieved for the two mobile phases revealed only marginal improvement in resolu- tion between related compounds B and C. The enhancement in resolution was calculated to be only 2% at the expense of an increase of 2.6 min in analysis time. As analysis time was the principal consideration during these studies, it was decided that the small increase in resolution did not compensate for the increased analysis time. As a result, the mobile phase producing the shorter analysis time was selected. The final, fully optimised assay conditions were: column, Hypersil phenyl 5-pm, 150 x 4.6 mm i.d.; mobile phase, [methanol - acetonitrile (70 + 30)] - triethylamine phosphate buffer (0.05 M) containing 1-octanesulphonic acid (0.025 M) and adjusted to pH 4 with phosphoric acid (58 + 42 V/V); flow-rate, 1.5 ml min-1; temperature, 40 "C; and detection wavelength, 260 nm.The proposed method was validated by comparison of assay results obtained with the USP and proposed procedures for a number of bulk tioconazole batches. Quantification of impur- ity levels was achieved by comparison with an external standard containing the four impurities, in eluent, at the concentrations given in Table 2. The assay results for the proposed method agreed very closely with those for the USP56 ANALYST, JANUARY 1989, VOL. 114 Table 2.Comparison o f assay results for the few potential impurities determined by the USP and the proposed methods USP method, Proposed method, Batch” Impurity % rn!m % rnirn 1 A B C D 2 A B C D 3 A B C D 4 A B C D (0.3 0.5 0.3 <0.3 0.4 0.5 4 . 3 0.7 0.7 <0.3 0.6 0.4 - - - - (0.2 0.5 0.3 <0. 1 <0.2 0.4 0.5 <0. 1 <0.2 0.8 0.6 <0. 1 <0.2 0.7 0.4 <o. 1 * Solutions of tioconazole were prepared in eluent to give a concentration in the range 6.7-7 mg ml-1. The external standard used for the quantification contained the potential impurities at the following concentrations, dissolved in eluent: A = 40, B = 60, C = 59 and D = 39 pg ml-1. procedure currently prescribed. Note that the degradation product D is not covered in the USP monograph. Conclusions Following the earlier optimisation of selectivity by modifica- tion of mobile phase composition and selection of column chemistry,j a second strategy was adopted to optimise resolution and analysis time.In the present work, small increases in the proportion of aqueous buffer in the mobile phase (while keeping the proportions of the organic modifiers, and therefore selectivity, constant relative to each other) were found to increase the average capacity factor for peaks B and C, without significantly enhancing resolution. The improve- ment in resolution achieved in this way was only 2% for these peaks, whereas the analysis time increased by 24%, making this approach impractical. The detection wavelength selected (260 nm) permitted increased discrimination between the peak tail of tioconazole and the peaks for B and C, with consequently enhanced resolution.Therefore, the final assay conditions were opti- mised fully with respect to both selectivity and detection conditions. Good resolution was achieved between all peaks within 12 min. The assay was shown to yield good linearity for each of the potential impurities, both in the presence and absence of tioconazole, over the concentration range of interest. The proposed method and the recommended USP procedure gave comparable results for the three potential impurities that are controlled, in all four batches of bulk tioconazole examined. The assay method finally developed fulfilled all the addi- tional requirements specified. Moreover, on a purely practical level, for real-world samples the mobile phase was non- destructive towards the column.the assay time was half that for the USP assay, separation was improved (compared with a rioconazole B 0 25 timin Fig. 6. Typical chromatogram of bulk tioconazole recorded under USP condition^.^ Detection wavelength, 219 nm; eluent, acetonitrile - methanol - water (containing 2 ml of ammonia solution) (39 -t 36 + 25 V/V/V); flow-rate, 1.5 ml min-1 the USP method, see Fig. 6) and all four potential impurities could be accurately quantified simultaneously. It should perhaps be noted that with the increasing availability of multi-channel detection systems5 it would be possible, in practice, to employ two, or more, detection wavelengths simultaneously, in order to optimise further the detection sensitivity for the earlier eluting peaks of A and D. Current practice in the international compendia, however, is to recommend the use of a single detection wavelength for regulatory purposes. A. G. Wright is grateful to Pfizer Central Research for financial support. References 1. 2. Berridge, J . C., in “Techniques for the Automated Optimiza- tion of HPLC Separations,” Wiley, Chichester, 1985. Schoenmakers, P. J., in “Optimization o f Chromatographic Selectivity-a Guide to Method Development,” Journal of Chromatography Library, Elsevier, Amsterdam, 1986, Volume 35. Wright, A. G., Fell. A. F., and Berridge, J. C.. 1. Chrorn- atogr. ~ in the press. “United States Pharmacopeia XXI,” United States Pharma- copeial Convention, Rockville. MD. 1985, Suppl. 2. p. 1895. Fell, A. F.. and Clark, B. J . , Eur. Chromatogr. News, 1987, 3. 4. 5 . 1(1), 16. Paper 8102258E Received June 6th, 1988 Accepted October Ilth, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400053

出版商:RSC    

年代:1989

数据来源: RSC