摘要:

ANALYST, JANUARY 1989, VOL. 114 57 Determination of Virginiamycin in Combination With Chlortetracycline in Feeds by Anion-exchange Chromatography Hussein S. Ragheb and Steven Ridge Biochemistry Department, Purdue University, West Lafayette, IN 47907, USA Aftab Din Smith Wine Animal Health Products, Applebrook Research Center, West Chester, PA 19380, USA Standard additions experiments for chlortetracycline hydrochloride and virginiamycin at a ratio of 16 : 1 showed a positive bias for both the plate and turbidimetric methods. The bias was eliminated by anion-exchange chromatography in both the presence and absence of feed components. When chlortetracy- cline and virginiamycin pre-mixes were used for fortification of laboratory-prepared feeds at chlortetracycline to virginiamycin ratios of 8 or 16 : 1, the recovery of virginiamycin, without treatment, was 93-100% using the plate assay and 181-236% with the turbidimetric method. The corresponding values obtained using anion-exchange chromatography were 89-99 and 85-91 %, respectively.The anion-exchange technique is necessary if the turbidimetric method is used for the assay. Keywords: Virginiamycin determination; microbiological assay; chlortetracycline interference; anion- exchange chromatography Virginiamycin is a polypeptide antibiotic with two chief components, factor M and factor S, which show marked biological synergism. It is used at a concentration of 27.5 p.p.m. to control swine dysentery. If chlortetracycline (CTC) is added with virginiamycin to the feed at a level of 220-240 p.p.m., it can be used for the treatment of swine enteritis, prevention of disease during stress and also for growth promotion and feed efficiency, Currently there are two plate methods'J and one turbi- dimetric methodi available for the determination of vir- giniamycin in feeds.The first plate method is a disc assay using M . luteus in which the feed sample is extracted with citric acid - acetone. 1 The second plate method is a cylinder assay based on inhibition of the same test organism, but, in this instance, the feed samples are extracted with methanol - pH 2.5 phosphate buffer.? The turbidimetric method is based on inhibition of S. faecium and extraction of the feeds with methanol.3 If CTC is present in combination with virginia- mycin, it would be desirable to determine the extent of the interference of the former in the determination of virgini- amycin.The object of this work was to investigate whether this interference could be eliminated. The differences in the solubility profiles of the two antibiotics, the use of a solid bonded-phase for their separation and ion-exchange chromat- ography were investigated. The only successful method was anion-exchange chromatography. Experimental Assay Organisms For the plate method, stock cultures of Micrococcus luteus ATCC 9341 were maintained on Difco Penassay Seed Agar slants (Difco Laboratories, Detroit, MI, USA). After incuba- tion for 24 h at 30"C, the growth was washed with sterile isotonic saline solution, spread over the surface of the same medium (250 ml) in a Roux bottle and incubated for 18 h.The surface growth was then washed with 25ml of sterile saline and was used as an inoculum for the diffusion method. The suspension is stable for about 2 weeks if kept in a refrigerator. The inoculum for the turbidimetric assay using Streptococcus faecium ATCC 8034 was prepared on the same medium. However, this inoculum can only be used once and is unstable even if refrigerated. Media and Reagents Difco Yeast Beef Agar was used for the plate assay. Each plate contained 10ml of the inoculated medium. For the turbidimetric assay, Difco Penassay broth was inoculated with the test organism. A pH 8 phosphate buffer was prepared by dissolving 16.73 g of anhydrous K2HP04 and 0.523 g of anhydrous KH2P04 in 1 1 of water. An anion-exchange resin, AGMP-1, 10G200 mesh, chloride form (Bio-Rad, Richmond, CA, USA) was used.An Econo column, 30 x 1.0cm i.d., cross-sectional area, 0.785cm2 (Bio-Rad), was filled to a height of about 6.5cm with the resin material suspended in methanol. Each column was washed with two 25-ml portions of 20% methanol - pH 8 buffer. For some experiments, the resin material was packed in either pH 8 buffer, 10% methanol - pH 8 buffer or 50% methanol - pH 8 buffer. In other experiments, the pH of the buffer was changed. Standard Solutions An appropriate amount of a virginiamycin reference standard (SmithKline Animal Health Products, West Chester, PA, USA) was dissolved in methanol to obtain a solution containing 1000 pg ml-1. This stock solution is stable for 4 weeks if kept in a refrigerator.From the stock solution, working standard solutions containing 0.20, 0.30, 0.40, 0.50, 0.60, 0.80 and 1 .0 pg ml-1 of virginiamycin were prepared in 20% methanol - pH 8 buffer. For the plate assay, concentra- tions of 0.20,0.30,0.40,0.50 and 0.60 pg mi-1 were used. The other standard solutions were used for the turbidimetric method. Assay Procedure For the plate method, the equipment and assay design were the same as those used by the A0AC.J Stainless-steel cylinders (6 mm i.d.) or cups punched in agar (6 mm i.d.) were each dosed with 0.2 ml of the standard or assay solution. Each plate contained three cylinders or cups with 0.40 pg ml-1 of virginiamycin (reference concentrations) and the other three cylinders (or cups) were used for the standard or unknown.Three replicate plates were prepared for each level of the standard and for each unknown solution. The plates were incubated at 37°C for 18 h and the zones of inhibition were measured (fO.l mm). Plots of log c versus diameter (mm)58 ANALYST, JANUARY 1989. VOL. 114 were used to measure the unknown potency. For the turbidimetric assay, the equipment and methods were as reported previously.5 Each tube contained 1ml of the standard or unknown and 5 ml of the inoculated broth. Three replicate tubes were prepared. The zero level on the standard response line corresponded to 1 ml of 20% methanol - buffer and 5 ml of broth. After inoculation, tubes, the contents of which were similar to those of the zero level, were refrigerated and used as a control for calibration of the spectrophotometer.All other tubes were incubated at 37°C for 3-4 h. When incubation was complete [zero level reaches 25-30% trans- mittance ( r ) ] , the tubes were shaken (by inversion), allowed to stand for Smin, then measurements taken at 600nm. Calibration graphs were constructed by plotting log per cent. T versus c. Points were connected by straight lines and the unknown potency was measured from the graph. Sample Extraction On the day before the assay, the feed samples (laboratory- prepared) were weighed; for samples containing 20-30 p.p.m. of virginiamycin, the mass taken was 40 g. The samples were pre-washed with hexane, allowed to dry, extracted with methanol (the volume of methanol used was twice the sample mass) and an equal volume of pH 8 buffer and shaken for 30 min.The samples were then centrifuged at 1500 rev min-1 for 10 min and filtered through Glass Microfibre filter-paper [GFIA, 12.54 cm (Whatman International, Maidstone, UK)]. The resulting extracts were stored in a refrigerator overnight in tightly stoppered flasks to minimise evaporation of methanol. An aliquot of the sample extract (20ml) was adjusted to pH 8.2 with NaOH solution (if necessary) and diluted to 50ml with pH 8 buffer. This procedure decreases the methanol concentration to 20%. For experiments invol- ving chromatography, five 5-ml aliquots of the sample were allowed to pass through the column under gravity. The initial eluate containing the virginiamycin was collected in a 100-ml calibrated flask. The column material was washed further with 20% methanol - pH 8 buffer until the contents of the flask had been made up to the mark.Further dilutions with 20% methanol - pH 8 buffer were made to give a theoretical concentration of 0.5 pg ml-1 of virginiamycin. For experi- ments not involving chromatography, the feed extract was diluted so that the final concentration of methanol was 20% and that of virginiamycin was 0.5 pg ml-1. For the standard additions experiments in the presence of the feed, the same technique was used; standard solutions were added to the blank swine feed on the day before the assay. Laboratory-prepared Feeds Swine feed with the following composition was used: crude protein, 28%; crude fat, 5%; crude fibre, 5%; calcium, 12-14% ; phosphorus, 4.5% ; and NaCl 6 7 .2 % . In prelimi- nary studies directed towards elimination of the interference by CTC, blank swine feed was used for the standard additions experiments. When the method had been developed fully, the feed was fortified with a virginiamycin pre-mix [stafac 22 (10 g lb-1) and CTC pre-mix (SO g lb-I), kindly donated by SmithKline Animal Health Products]. Three swine feeds were prepared, each containing 27.5 p.p.m. of virginiamycin. Feed A contained only virginiamycin, feed B virginiamycin and 220 p.p.m. of CTC and feed C virginiamycin and 440 p.p.m. of CTC. All feeds were ground so that they passed through a 1-mm screen, then mixed in a twin-shell blender for 2 h and refrigerated. Results and Discussion Initial experiments designed to investigate whether differ- ences in the solubility profiles of CTC and virginiamycin Table 1. Effect of methanol on anion-exchange chrornatography of virginiamycin.All experiments were performed without feed extract Virginiamycin recovered, '/" Turbidimetric Diffusion assay* assay _. NTI- AExt NTT AEx<- A. 10% methanol - p H 8 buffer$ 96.7 63.3 97.0 59.0 B, 20% methanol - pH 8 buffer3 98.0 94.0 95.8 89.0 C, 50% methanol - pH 8 buffer$ 97.0 97.0 95.3 96.7 * The theoretical concentration of virginiamycin was 0.40 v ~ g ml-1 for the diffusion assay and 0.60 pg ml I for the turbidimetric assay. i- NT = no treatment; AEx = anion-exchange chromatography. B n = 6 . Eluent $ n = 3 . Table 2. Effect of anion-exchange chromatography on virginiamycin recovery in the presence of chlortetracycline with and without feed.Anion-exchange chromatography was used in all experiments Virginiamycin found/ug ml- Diffusion assay' Turbidimetric assay' NF-I- WFt NFt WFt Average . . . . . . 0.39 0.39 0.56 0.57 Kecovery, Yo . . . . 92.9 92.9 93.3 95.0 Standard deviation, 'Yo . . 4.87 4.11 2.97 7.89 No. ofdeterminations . . 10 12 10 14 * The theoretical concentrations o f virginiamycin and CTC-HCI were 0.42 and 6.72 pg ml-1, respectively, for the diffusion assay and 0.60 and 9.6 pg ml 1 , respectively, for the turbidimctric assay. 5- NF = no feed present; WF = with swine feed. The final feed dilution was 1 + 20 for the turbidimetric as\ay and 1 + 29 for the plate assay. The feed dilution before chromatography was 1 + 10 for all assays. would result in removal of the former were unsuccessful.Blank feeds fortified with 27.5 p.p.m. of virginiamycin and 220p.p.m. of CTC were pre-washed with either dichloro- methane, benzene or carbon tetrachloride. However, only partial recovery of virginiamycin was obtained. Solid bonded- phase materials were then investigated. Sharma et al.6 used a CIS material for reversed-phase separation of tetracycline antibiotics. However, we were unable to separate chlortetra- cycline hydrochloride (CTC-HC1) from virginiamycin in spite of using several types of bonded material. For example, both antibiotics were eluted in 20% methanol fractions using diamino, quaternary amine, diol and aromatic sulphonate bonded-phases and in 60% methanol fractions using a CI8 bonded-phase. Sharma et al.' reported greater success in the separation of CTC from biological samples using an anion- exchange technique. Two factors pertaining to this technique needed to be investigated: (i) the optimum pH and (ii) the effect of the organic solvent (methanol) on the recovery of virginiamycin and its separation from CTC.For this investiga- tion standard additions experiments were performed in the absence of a feed matrix. In all experiments virginiamycin was detected by measuring its inhibitory activity against M . luteu5 and S. faecium. The results showed t'hat in the absence of methanol both antibiotics were retained on the anion- exchange resin. When the methanol concentration was set at lo%, the virginiamycin recovery was poor (about 60"/0). At 2&50% methanol, there was a marked improvement in the virginiamycin recovery (Table 1).A methanol concentration of 20% was preferred for the assay because solutions of higher methanol concentrations tended to seep out of the cylinders or cups during the diffusion assays. Further, a higher methanol content resulted in longer incubation periods (approximately 5 h) for the turbidimetric assay. Table 1 shows the lowerANALYST, JANUARY 1989. VOL. 114 59 Table 3. Comparison between no treatment and anion-exchange chromatography for the determination of virginiamycin in laboratory-prepared feeds Virginiamycin found, p.p.m. Theoretical concentration. p.p. m. Diffusion assay Turbidimctric assay Feed* Virginiamycin CTC-HCl NTt AExt "14- AExt A . . . . . . . . 27.5 0 Averaged: 26.4 26.5 23.6 26.2 Recovery, "/o 96.0 96.1 85.8 95.3 Recovery.Yo 93.2 89.1 181.1 85.1 C . . . . . . . . 27.5 440.6 Average: 28.9 25.8 64.9 25.0 Recovery, '7L 105. 1 93.8 235.7 9.07 B . . . . . . . . 27.5 220.3 Average$- 25.7 23.5 39.9 23.4 The basic feed was a commercial blank swine feed. + NT = no treatment; AEx = anion-exchange chromatography $- Number of determinations = 4 for each feed. Table 4. Student's t-test evaluation of laboratory-prepared feeds t value* Feed Testofsignificance A DF R DF C DF NTtandAExt . . 0.028 6 1.826 5 5.2163 4 between N T and AEx . . . . . . 1.959 6 7.098t 6 15.625;: 6 For NT between plate and turbidimetric assay . . . . . . 2.436 6 6.844$ 6 12.899t S For AEx between plate and turbidimetric assay . . . . . . 0.186 6 0.037 5 0.635 5 t NT = no treatment: AEx = anion-exchange chromatography. $ Significantly different at p = 0.05.For plate assay between For turbidimetric assay f depend4 on the number of degrees ot freedom (DF). recovery of virginiamycin (89%) obtained with the turbidi- metric method using 20% methanol compared with that obtained using 50% methanol (96.7%). The plate assays did not produce such a large difference in recovery values for the two methanol concentrations. As solutions for the diffusion assays were always prepared from solutions used for the turbidimetric assay, the lower virginiamycin recovery detected turbidimetrically was probably due to assay variability. Studies on the effect of pH (range 3-9) showed that in strongly acidic systems (pH 3), virginiamycin lost some of its microbiological activity.When 20% methanol - p1i 5 buffer or 20% methanol - pH 6 buffer was used, virginiamycin could not be separated from CTC as evidenced by a positive bias in the virginiamycin recovery (approximately 200%). With 20% methanol - pH 7 buffer, a marked improvement in the virginiamycin recovery was obtained with the plate method, whereas the turbidimetric assay still showed some interference (approximately 110-115%). This indicated that some CTC activity was present in the virginiamycin fraction. S. faeciurn, which is used for the turbidimetric assay, is more sensitive towards CTC than M . luteus, which is used for the plate assay. The best separation of virginiamycin from CTC-HCl was obtained using 20% methanol - pH 8 buffer. These results are consistent with the dissociation constants of CTC.8 At a strongly acidic pH, the CTC molecule is fully protonated as a singly charged cation. Between pH 3.3 and 7.5, CTC exists predominantly in the zwitterion form with the dimethylamino group in the 4-position protonated and the OH group in the 3-position ionised.As the p1-I increases from 8 to 9, a marked dissociation of the dimethylammonium cation occurs and the molecule becomes fully sorbed on to the resin material. However, virginiamycin passes through the column, probably because it has the same charge as the functional group o f the resin. At this point, it should be mentioned that the recovery of virginiamycin, when present alone, was consistent (about 95%) at pH 6-43. At pH 6 larger zones of inhibition were obtained with the diffusion assay and shorter incubation times using the turbidimetric method.However, this pH was not suitable for separation of the two antibiotics. Table 2 shows the results of the standard additions anion-exchange experiments in both the presence and absence of the feed extract. The virginiamycin recovery was about 93-95% of the theoretical value and was not affected significantly by the type of feed matrix or method of determination used. When no anion-exchange chromato- graphy was performed, a mixture of virginiamycin and CTC-HCl showed a positive bias in the recovery result (approximately 160%) of the theoretical value) using the diffusion assay and an even larger bias (approximately 180-200% of the theoretical value) using the turbidimetric method.It should be noted that, with the plate assay. anion-exchange chromatography was successful in separating the two antibiotics even when the mixture contained a 40-fold excess of CTC-HCI over virginiamycin. For concentrated feed extracts, however, complete separation of virginiamycin from CTC-HC1 was not possible. Hence, when a feed diluted 1 + 4 (i.e., each gram of feed was diluted four-fold) was passed through the resin material, some CTC activity passed through the column. Such concentrated feed extracts are likely to be encountered if the theoretical virginiamycin concentration is 5.5 p.p.m. It appears that the presence of certain feed components in high concentrations changes the chromato- graphic behaviour. Experiments are currently in progress to resolve this problem.Table 3 shows that, for laboratory-prepared feeds: no significant difference between no treatment and anion- exchange chromatography was observed for feed B (CTC to virginiamycin ratio = 8) using the plate method. For feed C (CTC to virginiamycin ratio = 16) anion-exchange chromato- graphy was found to be necessary. For the turbidimetric assay, a high positive bias was observed when the CTC to virginia- mycin ratio was 8 or 16. A statistical evaluation of the significance of differeflces between these results (Student's t-test) confirmed the above finding (Table 4). It is interesting to note that the difference in the values of t between the turbidimetric and plate assays for all feeds was small, indicating that the results belonged to the same population.This lends more credence to the anion-exchange technique. There was one instance where a difference between the standard additions experiments and results for laboratory- prepared feeds was observed. The difference occurred when no treatment was applied and when both antibiotics were assayed by the plate method. The results for the standard additions experiments showed a positive bias (approximately 112% of the theoretical value) when the ratio of CTC to virginiamycin was as low as 2. For laboratory-prepared feeds60 ANALYST, JANUARY 1989, VOL. 114 there was no necessity for anion-exchange chromatography of feed B. It appears, therefore, that a different response is obtained using CTC-HC1 (standard additions experiments) as opposed to a pre-mix containing CTC activity (laboratory- prepared feeds).Premixes are known to contain several isomers of CTC in addition to a calcium complex; these are probably inactive or have little activity against M . luteus, which was used for the plate assay. On the other hand, anion-exchange chromatography was found to be necessary for feed C using the plate method and for feeds B and C using the turbidimetric method (Table 4), probably because S. faeciurn is sensitive to many forms of CTC compounds. It should be noted that feeds vary widely in their composi- tion. Hence, if a sample that is known to contain both virginiamycin and CTC is received in the laboratory and the plate method is used for its analysis, it should first be assayed without employing chromatography. If the virginiamycin results show a positive bias greater than 10%, then the assay is repeated using the anion-exchange technique. If the turbi- dimetric method is chosen as the assay method, chromato- graphy is necessary. Further, although our interest was mainly in swine feed containing 27.5 p.p.m. of virginiamycin in combination with 22&440 p.p.m. of CTC, preliminary experi- ments with poultry feed containing both antibiotics showed that the anion-exchange technique was successful in minimis- ing the interference by CTC in the virginiamycin assay. References 1. Katz, S. E., Katz, J . M . , Miller, J . A , , Wang, R.. and Shapiro, R . J . . J . Axvoc. Off. And. Chem., 1984, 67, 569. 2. Ragheb, H. S., Black, L. J . . and Waisncr, D. J . , J . Assoc. Ojf. Anal. Chem., 1979, 62. 671. 3. Brennecke, D. M., Reed, S. J . , and Rarkate. J . A., J. A.c.soc. Ofy. Anal. Chem., 1981, 64, 319. 4. “Official Methods of Analysis of the Association of Official Analytical Chemists,” Fourteenth Edition. AOAC, Washing- ton, DC, 1984, p. 813. 5 . Ragheb, H. S., J . Assoc. Ojf. Atid. Chem., 1977, 60, 1119. 6. Sharma, J. P., Perkins, E. G., and Bevill, R. F., J . Chronzat- ogr., 1977, 134, 441. 7. Sharma, J . P.. Kovitz, G. D., Perkins, E. G., and Bevill, R. F.. J . Pharm. Sci., 1977. 66, 1319. 8. Sharma, J. P.. and Bevill, R. F.. J . Chromatogr., 1978, 166. 213. Paper 8l01332B Received April 5th, 1988 Accepted July 27th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400057

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST. JANUARY 1989, VOL. 114 Time-dependent Calibration Graphs in Electrothermal Atomisation Atomic Emission Spectrometry Robert E. Leyon" and James A. Holcombet Department of Chemistry, University of Texas at Austin, Austin, TX 78712, USA An automatic quantitative method has been developed using a family of calibration graphs, each measured at a different time during atomisation in a graphite furnace. Even at high analyte concentrations useful data can be obtained from the falling portion of the transient signal when the atom density has fallen sufficiently from its peak value. The method is demonstrated with emission from sodium atoms produced in a graphitefurnace and excited at the high temperatures therein. The method avoids calibration graph non-linearities caused by self-absorption and permits automatic analyses over a wide range of concentrations without sample dilutions or instrumental adjustments.Keywords : Atomic emission spectrometry; electrothermal atom isa tion; multiple calibration graphs; automated analysis Electrothermal atomisation atomic emission spectrometry (ETA-AES) using modern pulse-heated atomisers was first reported in 1974 by Massmann and Gucer,' although emission from atoms in a heated graphite tube had been studied by King2.3 early in this century. A great deal of practical and theoretical work was carried out by Ottaway and co-workers and a review of their work up to 1979 has been published.4 Both aspects of ETA-AES are covered briefly in annual reviews of emission spectroscopy.5 As a quantitative analytical method ETA-AES has been applied to a variety of practical problems but the range of the method was usually limited to two or three orders of magnitude in concentration because of curvature of the calibration graphs.This also occurred with probe,h.7 platform7-11 and constant-temperature12 atomisa- tion and regardless of whether peak height or peak area was measured as the dependent variable. This limited working range is largely the result of curvature or rollover at high concentrations. This is usually attributed to self-absorption and self-reversall3 although collisional quenching of excited states14 and slow atomisation13 may be important in some instances. Dilution or pre-concentration methods are usually employed to bring unknowns into a more optimum ( e . g ., more sensitive) range of a working curve, at the cost of additional time, complexity and risk of contamina- tion or loss. In order to extend the upper working limit of a spectro- metric method that exhibits a plateau or rollover in the calibration graph, a number of special techniques can be used. Woodriff et d.15 suggested using the width of a transient graphite furnace atomic absorption peak at 40% of the maximum signal as a measure of the amount of analyte. Another approach is to make the measurement at a wavelength away from the absorption or emission maximum. For example, using a wavelength-modulated continuum- source instrument, O'Haver and co-workerslh,17 developed an atomic absorption method using multiple calibration graphs. They constructed a separate calibration graph at each of several wavelengths across the modulation interval, with lower sensitivity resulting at wavelengths furthest from the centre of the absorption line.Although each log - log calibration graph covered a linear range of only about two orders of magnitude, the entire family of graphs extended the working range to about five orders of magnitude. When wavelength modulation is used for background correction in emission work, curvature can still result at high concentrations ,. Present address: Department o f Chemistry, Dichinson College, T To whom correspondence should be addressed. Carlisle, PA 17013, USA. because measurable emission may exist at all wavelengths within the modulation interval and the correction may no longer be accurate.10 Further, working at wavelengths c?f lower sensitivity does not necessarily solve the self-absorption problem because measurements are still being made at high atom densities. A second possible approach to the problem is to minimise self-absorption by making measurements when atom densities are low.Tn this paper we use a family of calibration graphs as did Harnly and O'Haver,l6 but obtained at different times rather than at different wavelengths. During the course of a furnace firing, the atom density increases and then decreases so that low atom densities are found at times earlier and later than that of the peak signal. Thus a family of calibration graphs obtained, for example, at times after the peak should provide usable measurements on high-concentration solutions while avoiding high atom densities.The use of multiple calibration graphs has been reported previously18 but in the context of ordinary absorption spectrometry and with the emphasis on statistical criteria for the choice of a single graph to use for a given unknown. In a similar application, Olsen et al.19 constructed a single calibration graph from several points taken from a single analyte peak in flow injection analysis. The shape of the emission peak for a given element depends on the heating rate of the furnace, which controls the change in atom population with time and the excitation process occurring in the gas phase, and the rate of loss of analyte atoms by diffusion and expulsion. As long as these factors remain unchanged, information can be obtained from any part of the emission peak.This approach assumes that the form of the dependence of excited atom concentration on time does not change with analyte concentration, that is, there are no complications due to changes in atomisation kinetics, conden- sation and re-atomisation, excitation, etc. Fig. 1 shows a series of typical emission versus time traces for different concentrations of analyte. For clarity, the traces have been drawn without the noise that would actually be present. With a high-speed data acquisition system and a suitable computer program, data can be taken at a series of times in each firing and a calibration graph can be constructed at each time. The highest concentrations are enough to produce an emission signal that saturates the measuring circuitry used in this study for a period of time before returning to a measurable level.Even though the peak emission signal has no significance with the present configura- tion when the analogue to digital conversion (ADC) system is saturated, there is still useful information in the descending part of the signal. The upper concentration limit of this62 ANALYST, JANUARY 1989, VOL. 114 t > v) C a, c C 0 vl v) LI .- c .- .- ._ E w J I I , I ! 1 I 0 1 2 3 4 5 6 t s Fig. 1. Family o f emission peaks from sodium at 589.0 nm and concentrations (mg 1-1): A, 0.00371; B, 0.0185; C, 0.0927; D, 0.371; E. 0.927; F, 1.8.5; and G, 3.71 method is determined not by the measuring electronics or curvature in working curves but by how long the furnace can be kept at the selected atomisation temperature, how long data can be recorded or the appearance of memory effects in the furnace.Experimental To maximise the viewing area within the furnace volume while minimising the amount of blackbody radiation entering the optical system, a large diameter furnace was deemed most appropriate. The graphite furnace was prepared from a standard, pyrolytic graphite coated tubular furnace (Part No. 44192-01, Instrumentation Laboratories). A 1.6 mm i.d. hole was drilled for sample introduction. Typical furnace dimen- sions were 4.7 mm i.d. X 6.3 mm o.d., and each was cut to 13 mm in length. It was mounted in a CRA-90 (Varian Associates) workhead using Varian electrodes; the curvature at the ends of the electrodes was machined to match the larger diameter of the Instrumentation Laboratories furnace to provide better electrical contact and thus more uniform heating. The workhead was enclosed in an anodised alumin- ium box*() which served to rninirnise both entrainment of air and internal reflections of the continuum emitted by the furnace wall.To reduce the amount of furnace wall radiation falling on the detector further, a mask with a 2.4 mm diameter aperture hole was mounted on the workhead in front of the furnace, and another, larger, mask was placed between the enclosing box and the two-mirror optical system to prevent radiation from reaching the entrance slit directly from the furnace without being reflected from the mirrors. The dry, ash and atomise steps were controlled by the CRA-90 power supply.The atomisation cycle was optimised by the simplex method21 using ash time, ash temperature, atomisation temperature, hold time and heating rate as the variables and the net intensity of sodium emission as the parameter to be maximised. A rather broad and noisy maximum was found in the response surface. The following values were used for all subsequent work: dry, 10 s at 200°C; ash, 10 s at 700 "C; atomise, at 2500 "C using a heating rate of 800°C s-1 and a hold time of 5 s. These temperatures are the nominal settings on the CRA-90 controller. Because of the larger furnace used, the factory-calibrated power settings resulted in lower actual temperatures. The hold time and heating rate were the maximum available values. Alignment of the furnace was critical in spite of the masks and inherently high spatial resolution of the optical system.The two-mirror optical system has been described else- where.22 It consists of two spherical concave mirrors arranged in an over-and-under symmetric arm configuration and allows high spatial resolution without sacrificing light collecting ability. To minimise internal reflection in transmitting optical components no lenses or windows were present in the optical train. A 0.4-mm horizontal slit was placed in the tangential image plane and the 0.5-mm monochromator entrance slit was located in the sagittal image plane. A 0.5-m Ebert monochro- mator (Jarrell-Ash, Model 82-516) with a band pass of 0.8 nm and an R292 photomultiplier tube (Hamamatsu TV) com- pleted the optical system.A trigger signal to begin data acquisition was sent to the digital waveform recorder (Model 805, Biomation, Cupertino, CA, USA) by the CRA-90 controller at the start of the atomisation phase. The output from the detector was capacit- atively filtered before being sent to the waveform recorder where it was digitised and stored as 2048 eight-bit words. The response time of the system was calculated to be 6.8 ms and data were collected every 5 ms, giving a total recording time of approximately 10 s, which was generally sufficient to record the entire atomisation signal. After a firing, the stored data were transferred from the transient recorder into a Z80-based microcomputer (Vector Graphics) running CP/M-based BASIC software. A 10-ml Hamilton syringe with a length of PTFE tubing fastened over the needle was used to deliver the 2-ml sample aliquot into the furnace.The sample was held entirely in the tubing and did not come into contact with any part of the syringe or needle. The syringe was supported in a brass holder on a guide rod for maximum reproducibility of sample introduction. All chemicals were of analytical-reagent grade. Solutions were prepared from concentrated (37k1000 mg 1-1) stock solutions using dc-ionised water and calibrated glassware that had been stored in dilute nitric acid. Nitrogen was used as the sheath gas. Results and Discussion The graphite furnace used in this work is of the same general design as the Varian CRA-61/63/90 series, but is larger. At least one study13 found no observable atomic emission signal from atoms in a CRA-63 furnace and attributed this negative result to the relatively short residence time, which has been estimated to be about 1 s for a number of metals.24 In contrast, other w0rkers2~,26 did observe an emission signal from the mini-Massmann-type Varian furnaces.However, as noted previously, the larger furnace generally provided an over-all improvement in the signal to background ratio. Visual observation of this modified furnace indicated that it was being heated uniformly enough to prevent condensation at the ends, thus minimising memory effects and analyte loss by condensation. The behaviour of several elements was surveyed in the early stages of this work. Sodium was chosen for intensive study because it gave a single emission peak during atomisation with this system and exhibited good analytical sensitivity.The 589.0-nm line was monitored in all instances. Fig. 2 shows a trace of the background continuum radiation emitted at 589.0 nm by the furnace and a typical emission signal for sodium. Algorithm and Data Handling The algorithm used by the computer program is outlined in Table 1. A series of firings is made using standards covering a concentration range of four orders of magnitude or more,ANALYST, JANUARY 1989, VOL. 114 t 63 n I I 0 1 2 3 4 5 6 tl s Fig. 2. A. Background emission signal from the empty furnace at 589.0 nm; B, sodium emission signal at 589.0 nm with 2 ml of a 0.0185 p.p.m. solution Table 1. Algorithm used in autocalibration program I. Run series of standards A .Correct for background emission B. Measure net emission at specified times C. Save data in an array of log(net emission) for each concentration and specified time 11. Make calibration graphs A . Display plot of log(net emission) versus log concentration B. Fit linear, quadratic and cubic graphs to data points C. User chooses desired graph D. Store array of fitted points 111. Analyse unknowns A. Correct for background emission B. Measure log(net emission) at specified times C. Determine which calibration graphs are applicable D. Use interpolation to find unknown concentrations E. Print concentrations found depending on the detection limit (lower limit) and how long the graphite furnace can be held at the atomisation tempera- ture (upper limit). The 2048 data points are averaged in groups of four and the furnace background signal is subtracted from these data to obtain the final time-dependent emission signal.The data are then smoothed by a 17-point cubic Savitsky - Golay filter27 and the points corresponding to 200-ms intervals during the emission peak are stored in a two-dimensional array with indices keyed to concentration and time. After all the concentration and intensity data have been gathered, the computer prepares a series of calibration graphs, one for each time (and one for the peak maximum up to the concentration that saturates the ADC). Each is displayed graphically. The program permits removal of outliers or rejection of the entire graph, a procedure recom- mended by de Krenk et ~1.28 as part of the intelligent use of automatic calibration functions.In this work an outlier was defined as a point whose residual was more than five times larger than the next-largest residual. After the user is satisfied that only valid data remain, the program fits first-, second- and third-order polynomials to the data, using a least-squares polynomial fit,29 and displays the experimental intensity values, the calculated intensities and the residuals. Then the user chooses one of the fitted graphs by specifying the order of the graph to be stored. Here again the residuals are used as a guide by using their signs and the sum of their squares. When a fitted graph is chosen, the computer stores the points on the fitted graph rather than the experimental data points or the 2 3 Nitrogen flow-rateil min-l Fig.3. Effect of sheath gas flow-rate on height of the sodium emission peak. 0, Gas entering at workhead, immediately beneath furnace and A, gas entering at bottom of box enclosing furnace coefficients of the polynomial, which makes it possible to use a simple interpolation rather than having to find the root of an equation when calculating an unknown concentration. To analyse an unknown, the blank is fired and the data are stored for future use. Then the unknown is fired under the same conditions used for the standards, and the computer measures and stores the emission signal at the specified times as before. For each time it then retrieves the emission signal and uses an interpolation routine29 to compute the concentra- tion from the appropriate working graph.After all the times have been used the program prints the concentrations it has calculated. Sheath Gas As the windows on the enclosing box were removed, some sheath gas flow was required during atomisation to prolong the tube life. It was possible to route the nitrogen sheath gas to the workhead either through the screen directly under the furnace, or simply through a hole in the side of the box that enclosed the workhead. Experiments showed that the emis- sion signal was larger when the gas entered through the box. Another set of experiments showed that even with the gas entering the side of the box, the flow-rate affected the peak height. The results are shown in Fig. 3. The lines are those obtained by a linear least-squares fitting procedure.Stopping the nitrogen flow during atomisation did not result in higher signal intensity. A constant flow-rate of 1.2 1 min-1 was used in all subsequent work. Background Furnace Emission In addition to analyte emission, some continuum radiation from the incandescent furnace walls reaches the detector. This background contributes to the noise level and, if not constant, can lead to analytical errors. The simplest approach for background correction involves firing an empty furnace, or one containing a method blank, and subtracting this signal from that obtained with a sample present. The background must be reproducible for this approach to be accurate though it has been used successfully in certain situations .3°-34 Wavelength modulation, a technique first developed for64 ANALYST, JANUARY 1989, VOL.114 atomic absorption work, has also been used in several laboratories.7.+11,23~2,35 It allows relative errors in the range of 30% or less to be achieved as long as the emission line is narrower than the wavelength modulation interval. Brinkman and Sacks36 used a predecessor of this method involving a stationary silica plate placed just before the exit slit of the monochromator, and two detectors to measure independently the signals at two wavelengths. The simplest way to minimise the effect of continuum emission is to limit the amount of light that falls on the detector, either by using a high-dispersion monochromator or by viewing only a small volume of the furnace. The optical system used in this work has been shown to possess high spatial resolution22 but still it was necessary to use a mask between the furnace and the entrance slit of the monochroma- tor and to correct for the radiation that was scattered37 by the gases in the furnace. In this study two methods for correcting for furnace emission were used. In the first method, the blank emission signal is recorded separately, stored and subtracted from the measured signal point by point.This requires a reproducible blank and is somewhat time consuming unless the correction is made only at the times specified for the data to be saved. In the second method, the beginning and end of the emission peak are located and the slope and intercept of the straight line connecting them are calculated. Then the corresponding point on this line is subtracted from the measured signai at each of the specified times.This procedure does not require a separate measurement of the blank but does have some disadvantages. The base line changes with variations in peak width and hence with analyte concentration. This results in a slightly different base line for each standard and leads to some curvature in the calibration graphs. Because the end of the signal is defined as the point where the measured signal levels off or begins to increase again and so is not found at high concentrations, use of this correction method limits the analytical range that can be used. Sodium Results Sodium standards covering the range 0.0037-11 p.p.m. were prepared. A family of calibration graphs, constructed from the emission traces in Fig.1, is shown in Fig. 3. As this figure is for illustration only, the lines in it have been fitted by eye. Ail graphs are plotted using logarithmic co-ordinates.38 After constructing the family of calibration graphs, severaI standard solutions were used as unknowns. The results for the analysis of these solutions, using the background correction methods indicated, are given in Table 2. The concentration found represents the mean value for the various time delays in a single firing; more than one value is given when replicate analyses were made. The number of concentration results ( N ) varies because of outliers and differences in the number of results returned by the computer program. It appears that point-by-point correction gives better accuracy and worse precision than the base-line method, but the differences are not great. When using point-by-point correction, accuracy is improved by making frequent re- measurements of the blank signal.Effect of Atomisation Process It is well known that the mechanism of gas-phase atom formation in the graphite furnace is strongly dependent on the chemical and physical conditions of the analysis. For graphite furnace atomic absorption, quantitative analytical methods that make use of the peak height depend most strongly on all these conditions remaining constant for standards and un- 1 -3 - 2 - 1 0 Log([Nal/mg I-') Fig. 4. 1, using data at: A, 1.0; B, 1.5; C. 2.0; D. 2.5; and E, 3.0 s Family of calibration graphs obtained from the traces in Fig. Table 2.Quantitative results for sodium Point-by-point background correction" Actual Concentration concentration/ found/ Relative mg 1-1 mg 1 - I Nri: error, 76 RSD, O/o 0.00742 0.00834 6 + 12.4 10.5 0.00927 0.018s 0.037 1 0.0374 0.0396 0.0413 0.0927 0.0949 0.185 0.163 0.173 0.371 0.414 0.742 0.825 1.11 0.991 1.26 1 .8S 1.60 1.896 4 4 S 10 14 1s 9 9 18 18 1s 15 +0.8 +6.6 +11.3 +1.9 -11.9 -6.2 +11.6 +11.2 - 10.8 + 13.9 -13.5 +2.2 16.1 10.1 16.2 9.9 16.3 13.8 9.4 5.6 9.1 5.7 7.8 4.6 Straight line background correction? Concentration found/ mg 1 - 1 N t 0.00940 2 0.0224 3 0.022s 3 0.0161 2 0.0368 4 0.0414 4 0.0366 3 0.0801 5 0.220 4 0.163 3 Relative error.% +1.4 +21.1 +21.6 - 13.0 -0.8 +11.6 -1.4 - 13.6 + 18.9 -11.9 RSD, Yo 6.0 3.9 6.8 2.6 3.1 3.8 13.0 24.1 10.0 8.2 * Correction made by subtracting a previously recorded blank firing..i- Correction made by subtracting a tangent base line. $ N is the number of concentration results generated from a single firing. 5 A new7 blank was recorded immediately prior to this firing.ANALYST, JANUARY 1989, VOL. 114 65 knowns. Because of the strong temperature dependence of the emission signal intensity in ETA-AES, both height and area of the signal will exhibit a significant dependence on the temporal precision of atom formation. Any change in the atomisation process will change both the shape and the position of the emission signal in time. The net effect will be a change in the magnitude of the signal at a given time. As with other ETA-AES methods reported so far, this will introduce a major error into the present method, one that can be corrected only by preparing an entirely new family of calibration graphs that ensures matrix-matched standards and samples.Atomisation in a “constant-temperature” furnace might minimise some of these variations and may be especially helpful in reproducing the blank emission signal. A narrow peak (i.e., one resulting from rapid atomisation) formed at a maximum temperature is optimal for our method of time- dependent calibration graphs. This permits the greatest sensitivity while allowing the use of the decaying portion of the signal to extend the upper concentration limit of the tech- nique. This shape can most readily be obtained with high heating rates and analytes that can be stabilised on the graphite surface.References Massmann, H., and Gucer, S . , Spectrochim. Acta, Part R , 1974, 29, 283. King, A. S . , Astrophys. J . , 1905, 21. 236. King, A. S . , Astrophys. J . , 1908, 27, 353. Ottaway, J . M., Hutton, R. C . , Littlejohn, D . , and Shaw, F., Wiss. Z. Karl-Marx- Univ. Leipzig, Math. -Naturwiss. Reihe, 1979, 28(4), 357. Keliher, P. N., Boyko, W. J . , Clifford, R . H., Snyder, J . L., and Zhu, S. F., Anal. Chem., 1986, 58, 335R. Giri. S. K., Littlejohn, D., and Ottaway, J . M., Analyst, 1982, 107, 1095. Marshall, J . , and Ottaway, J . M., Talanta, 1983, 30, 571. Gregoire. D. C . , and Chakrabarti, C. L., Spectrochim. Acta, Part B , 1982, 37. 625. Ottaway, J . M., Bezyr, L.. and Marshall. J . , Analyst, 1980, 105, 1130. Marshall, J . , Littlejohn, D., Ottaway, J .M., Harnly, J . M.. Miller-Ihli, N. J . , and O’Haver, T. C., Analyrt, 1983, 108, 178. Ottaway, J . M., Bezur. L., Fakhral-Aldeen, R., Frech, W., and Marshall, J . , in Braetter, P . , and Schramel, P., Editors. “Trace Element Analytical Chemistry in Medicine and Biol- ogy. Proceedings of the First International Workshop, Neuher- berg. FRG, April 1980,” Walter de Gruyter, Berlin, 1980. p. 575. Baxter, D. C., Frech, W., and Lundberg, E., Analyst, 1985. 110,475. Marshall, J., Littlejohn, D., Ottaway. J . M., Miller-Ihli* N. J . . O’Haver, 2 . C., and Harnly, J . M.. Specrrochim. Acta, Part R , 1984, 39, 321. Li, G.. Guangpuxue Yu Guungpu Fenxi, 1985, 5, 31; Chem. Abslr., 1985, 103. 134011h. Woodriff. K., Marinkovic. M., Howald, R. A., and Eliezer, I., Anal.Chem.. 1977, 49, 2008. Harnly, J. M., and O’Haver, T. C . , Anal. Chem., 1981. 53, 1291. Miller-Ihli, N. J . , O’Haver, T. C . , and Harnly, J . M., Anal. Chem., 1984, 56, 176. Mitchell, D. G., Mills. W. N . , Garden, J . S . , and Zdeb. M.. Anal. Chern.. 1977, 49, 1655. Olsen. S . , RfiiiCka, J . , and Hanscn. E. H . , Anal. Chim. Acta, 1982, 136, 101. Salmon, S. G . , Davis, R. H., Jr., and Holcombe, J . A . , Anal. Chem., 1981, 53, 324. Nelder, .I. A . , and Mead. R., Comput. J . , 1965, 7, 308. Salmon, S G . , and Holcombe, J. A . , Anal. Chem., 1978, 50, 1714. Epstein, M. S . , Rains, T. C., and O’Haver. T. C., Appl. Spectrosc., 1976, 30, 324. Sturgeon, K. E., Chakrabarti, C. L., and Bertels, P. C., Anal. Chem., 1975, 47, 1250. Matousek, J. P., and Smythe, L.E . , Appl. Spectrosc., 1978,32, 54. Littlejohn, D., and Ottaway, J . M., Anal. Chim. Acta, 1979, 107, 139. Savitsky, A., and Golay, M. J . E., Anal. Chem., 1964, 36, 1627. de Krenk, H., Pronk, H. F., and Wijker, J . , Spectrochim. Acta, Part B , 1983, 38, 831. Shoup, T. E., “Numerical Methods for the Personal Com- puter.” Prentice-Hall, Englewood Cliffs, NJ, 1983. Shaw. F., and Ottaway. J . M., Anal. Lett., 1975, 8, 911. Littlejohn, D., and Ottaway, J . M., Analyst, 1977, 102, 393. Hutton, R . C., Ottaway. J . M., Rains, T. C., and Epstcin, M. S . , Analyst. 1977, 102, 429. Ebdon, L., Hutton, R. C.. and Ottaway, J . M., Anal. Chim. Acta, 1978, 96, 63. Ottaway, J . M.. and Shaw, F., Anal. Chim. Acta, 1978,99,217. Epstein, M. S . , Rains, T. C . , Brady, T. J., Moody, J .R . , and Barnes, I. I > . , Anal. Chem., 1978, 50, 874. Brinkman. D. W., and Sacks, R. D., Anal. Chem., 1975, 47, 1723. Littlejohn, D., and Ottahay, J . M., Analyst, 1977. 102, 553. Goode, S. K., and Northington, J . W.. Appl. Spectrosc., 1979, 33, 12. Paper 8100266E Received January 26th, I988 Accepted September 28th, I988 1 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. Conclusions This work demonstrates the usefulness of time-dependent calibration graphs using emission intensity data from the falling side of the intensity versus time profile. The system provide s auto - r an gi ng capabilities with no s am pl e man i p u 1 a- tions and minimum instrument adjustments even for samples of high concentration. No manual data transcribing, plotting, interpolation or calculation is required over a range of at least 3.5 orders of magnitude in concentration. Any additional improvements in tube design, optical system or detector circuitry could certainly lower the limits of detection and further extend the dynamic range.The present protocol requires only that an aliquot of the sample be introduced into the furnace with a single firing of the atomiser. The accuracy obtained is comparable to other methods. No special accessories are required but modifications of, or additions to, existing data collection routines need to be made. In principle, the method can be applied to any analytical technique that produces a transient signal. This list includes graphite emission or absorption, continuous-flow methods, chromatographic methods and flame spectrometry if the sample is introduced as a plug (e.g., in hydride generation or discrete nebulisation of millilitre sample volumes). The largest source of error for ETA-AES is the reproduci- bility of the blank. When a blank was recorded just prior to an unknown the accuracy of the result was much higher than if the blank had been recorded six to eight shots earlier. As noted previously, wavelength modulation techniques could be used t o reduce this uncertainty significantly. The advantages of the method include those inherent in all ETA-AES techniques (e.g., applicability to small sample volumes, high sensitivity and potential for multi-element analysis). The added feature discussed in this work is the convenient establishment of multiple calibration graphs and the accompanying increase in dynamic range and avoidance of extra manipulation and re-analysis. Unlike single measure- ments made of the peak height, this approach affords the opportunity of maximising the use of the intensity information that is available from a single firing of the atomiser. Disadvantages include the susceptibility to matrix effects present with graphite furnace atomisation as well as problems common to most emission methods (e.g., sensitivity to changes in heating and excitation conditions). 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. This work was supported by grant CHE84-09819 from the National Science Foundation. R. E. L. thanks Dickinson College for sabbatical leave during which this work was performed.

ISSN:0003-2654    

DOI:10.1039/AN9891400061

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 Indirect Determination of Tungstate in Rat Tissues by Atomic Absorption Spectrometry 67 Debasis Chakraborty and Arabinda K. Das Department of Chemistry, University of Burdwan, Burdwan 773 104, India An indirect method is described for the determination of tungsten as tungstate in tissue samples by atomic absorption spectrometry (AAS). Tungstate forms a stable ion-association complex [Fe(dipy)#+W04*- (dipy = 2,2'-dipyridyl) in acidic solution, which can be extracted into chloroform with an efficiency of higher than 98%. The extract can be analysed for iron (and hence indirectly for WO&) by flame AAS after stripping back into 60% perchloric acid. The calibration graph is linear up to 19 p.p.m. of W04*- and the limit of detection is 0.17 p.p.m.Many foreign ions do not interfere and the method has been applied successfully to the determination of tungstate in rattus norvegicus tissue samples. Keywords: Indirect tungstate determination; atomic absorption spectrometry; tissue samples The determination of tungsten in rat tissue is of importance, because tungstate causes a significant excretion of molyb- denum, which results in both reduced growth and capacity to oxidise xanthine, and leads to a high mortality rate.' Tungsten can be determined by atomic absorption spectrometry (AAS) using a high-temperature N 2 0 - C2Hz flame, but the sensitivity of this direct determination*-13 is poor. In addition to the low sensitivity of the method, serious interferences4.5 have been encountered and on occasion solvents have been used that are unsuitable for aspiration into flames.s.10 However, the sensi- tivity has been improved by carrying out the solvent extrac- tionh.9.12 of tungsten followed by aspiration of the solvent directly into the flame. An indirect AAS determination involving the formation of a PbW04 precipitate and the subsequent determination of the excess of Pb has been reported previously. 11 This paper describes a more rapid and sensitive AAS method for the determination of tungstate based on the following principle: in acidic solution, tungstate and Fell form a stable complex with 2,2'-dipyridyl (dipy), extractable in chloroform.The FeII in the extract can be determined by flame AAS after stripping back into the aqueous phase with 60% perchloric acid.The tungstate concentration is then related to the amount of Fe" present. The method has been used successfully for the determination of tungstate in rattus n o r ~ ~ g i c u s tissue samples. Experimental Absorbance measurements were performed with a Shimadzu Model 646 atomic absorption spectrometer, using the follow- ing instrumental conditions: Fe lamp current, 9 mA; wavelength, 248.3 nm; slit width, 0.19 nm; CzH2 flow-rate, 2.5 1 min-1; air flow-rate, 10 1 min-1; and burner height, 4 mm. Values of pH were obtained with a Sambross Model 335 digital pH meter. Reagents Zron(ZZ) solution, 1900 p.p.m. of Fe". The stock solution was prepared by dissolving analytical-reagent grade am- monium iron(I1) sulphate (S.M., Baroda, India) crystals in 450 ml of 5% sulphuric acid and making up to the mark in a 500-ml calibrated flask with doubly distilled water after the addition of 0.5 g of NHzOH.HCl.The solution was stan- dardised titrimetricallyl5 and working solutions were prepared by serial dilution with doubly distilled water. Tungstate solution, 3000 p. p. m. of W042-. The stock solution was prepared by dissolving analytical-reagent grade Na2W04.2H20 (Merck, Darmstadt, FRG) in doubly distilled water. It was standardised gravimetricallyls as BaW04 and diluted as required. Acid digestion mixture. Concentrated sulphuric acid, 50% nitric acid and 72% perchloric acid (4 + 5 + 5 ) was used. All other reagents were of analytical-reagent or general purpose- reagent grade. Procedure The following solutions were added to a separating funnel in the following order: 1 ml of a solution containing 150 p.p.m.of Fe", 0.6 ml of 0.06% dipy solution, 9 ml of the sample solution (for the blank, 9 ml of process blank was used), 1 ml of methanol, 0.5 ml of pH 1.0 buffer (NH4Cl - HCl solution) and 5 ml of chloroform. The mixture was shaken for 40 s and left to stand for 2 min in order that the extraction equilibrium be established. The aqueous phase was discarded and the organic phase was washed twice with 2 ml of doubly distilled water and the washings were discarded. The organic phase was then treated with 5 ml of 60% HC104 and the aqueous phase made up to 10 ml after adding 100 pg of an A13+ solution (following this procedure the sensitivity to iron was increased16). The absorbance was then measured after aspirating the aqueous phase into an air - CzHz flame using doubly distilled water as reference.Procedure for the Determination of Tungstate in Rat Tissues The proposed method was applied to the determination of tungstate in the liver and kidney of rattus norvegicus. Fresh, whole kidney samples (0.5-0.7 g) and fresh whole liver samples (2.2-3.8 g) were weighed into a 500-ml Kjeldahl flask and 14 ml of the acid digestion mixture17 and a few glass beads were added. The flask was heated gently at 50 "C using an electric temperature-controlled heating mantle. To prevent charring, small portions of 50% nitric acid - 72% perchloric acid (1 + 1) were added until the solution became clear. The flask was then heated at 70 "C in order to concentrate the solution and heating was continued until all the white fumes had been expelled from the flask.The solution was allowed to cool at room temperature (30 "C), diluted slightly and the digest made alkaline by cautious addition of 10.0 M NaOH solution. The solution was then boiled vigorously at 120 "C for 30 min. The resulting mixture was again cooled at room temperature and the pH was adjusted with dilute nitric acid to 0.7-1.5. The digest was transferred quantitatively into a 10-ml calibrated flask and diluted to the mark with doubly distilled water. Aliquots of this solution were analysed by the proposed method.68 ANALYST, JANUARY 1989, VOL. 114 extraction remained constant when 0.5-1 ml of a 0.06% adduct solution was added. Hence a volume of 0.6 ml was used. The absorbance of the blank solution was unaffected by the addition of dipy to the aqueous phase.Process Blank To prepare the process blank, 4 ml of concentrated H2S04, 5 ml of 50% HN03 and 5 ml of 72% HC104 were placed in a 500-ml Kjeldahl flask without a tissue sample and the final volume of the solution (after following all the sample digestion steps) was made up to 10 ml. Results and Discussion Kinetics of Extraction The extraction of the tungstate complex into chloroform was investigated by shaking 5 ml of solvent vigorously with the aqueous phase for different periods of time. Experiments showed that shaking for 30 s was sufficient for complete extraction. The time necessary to achieve the extraction equilibrium was studied using chloroform solution at pH 1 .O. Extraction with this solvent proved to be rapid; equilibrium was attained in about 2 min for both the known solution and the unknown tissue samples.Effect of pH The pH of the solution strongly affected the formation and extraction of the ion-association complex. The percentage of extraction was constant and at a maximum in the pH range 0.7-1.5. Choice of Solvent System Various solvent systems were investigated as possible ex- tractants of the ion-association complex. The following seven systems gave positive results: chloroform (98.3%); chloro- form - butan-l-ol(1 + 1) (78.9"/0); 1.2-dichloroethane (8.1%); isobutyl methyl ketone (TBMK) (6.3%); amyl methyl ketone (.5.0"%); butan-1-01 (4.3%); and ethyl acetate (1.2%). From the results it was deduced that under the conditions used chloroform was the best solvent because it was highly selective and yielded the highest extraction efficiency.Choice of Stripping Agent The [Fe(dipy)3]2+W042- ion-association complex extracted into chloroform was stripped back into the aqueous phase by means of various acids of different strengths, e.g., HCI, HN03. HzSOj and HC104. The highest sensitivity was achieved using HC104 for a fixed amount (12 p.p.m.) of W04'-. Complete back extraction was achieved when the HC104 acid strength was .5&72%. Variation of Polarity of the Medium The polarity of the aqueous phase was varied by the addition of 0.1-2 ml of methanol. It was found that the addition of 1 ml of methanol to the aqueous phase produced the highest extraction, viz., the highest sensitivity for a fixed amount of W042- (12 p.p.m.).Hence a volume of 1 ml was used in the procedure. Choice of Adduct Using l,l0-phenanthroline (phen) and dipy, the Fe(phen)32+ and Fe(dipy)32+ systems, respectively, were investigated for the extraction of tungstate into chloroform. The results indicated that [Fe(di~y)~]2+W0~2- was the better system because it had a higher extraction efficiency for W042- (12 p.p. m.) . Effect of Adduct Concentration Increasing the concentration of dipy in the aqueous phase increased the sample extraction, whereas the net sample Effect of Foreign Ions The determination of 60 pg of W042- is possible (within an error of ?2%) in the presence of the following ions: Br03- (%-fold excess); SO3*- (37-fold excess); citrate (32-fold excess); F- (27-fold excess); Se032- (26-fold excess): NO2-, NO3-, Br- and EDTA (23-fold excess); CH3COO- and B40+ (20-fold excess); C1- and S2- (17-fold excess); S2032- (16-fold excess); oxalate and SCN- ( S f o l d excess); As043- (12-fold excess); Ag+ and Co2+ (ll-fold excess); P o p , Mn2+, Ni2+, Cd2+ and A13+ (10-fold excess); tartrate, 103- and SO& (%fold excess); Mo042-, Ca2+ and U02*+ (&fold excess); Cu2+, Ti", Hg2+, Ba2+, Zn2+ and V03- (7-fold excess); T- and Zr02+ (&fold excess); and Pb2+ (3-fold excess).Clearly, most cations and anions are tolerated in a large excess in the determination of 60 pg of W042-. Composition of Extracted Species In order to determine the composition of the extracted species formed under the experimental conditions, the general procedure was applied to a series of solutions having fixed dipy and tungstate concentrations and an increasing concen- tration of Fe".The molar ratio was determined to be 1.02, so presumably the extracted species is [Fe(di~y)~]2+W0~2-. Analytical Figures of Merit The calibration graph is linear up to 19 p.p.m. of W042- and the limit of detection (30)IX is 0.17 p.p.m. The sensitivity of the method is 0.25 p.p.m,, which is better than that obtained using any of the earlier direct AAS methods. The relative standard deviation for ten determinations of 1.0 p.p.m. of Table 1. Tungstate in rat tissue samples Sample Tissue 1 Liver 2 Liver 3 Liver 4 Liver 5 Liver 6 Liver No. analysed Kidney Kidney Kidney Kidney Kidney Kidney T ~ t a l [ W O ~ ~ - ] i 0.12 0.64 0.14 0.80 0.09 0.39 0.16 0.86 0.17 0.69 0.13 0.67 I-lgg-' Table 2.Recovery of WOj2- added to tissue samples. Amount taken. 9 ml Amount of Amount of Sample addedipg foundipg Recovery. Yo wo,2 WO,' Liver-1 . . . . 0 10 20 Kidney-3 . . 0 20 30 Liver-S . . . . 0 30 40 0.34 - 9.82 9s 20.17 99 0.29 - 19.65 97 31.48 103 0.38 - 32.22 1 06 40.76 101ANALYST, JANUARY 1989, VOL. 114 WO.+- is 1.22% and that for 15 determinations of 1.0 p.p.m. (after standard additions) of W042- in kidney solution (sample 2) is 1.35%. Application The results of the determination of tungstate in various liver and kidney samples are given in Table 1. Apparent recoveries of 95-106% were obtained for the determination of W042- in three tissue samples spiked with 10-40 pg of W04’- (Table 2). Conclusion This paper describes the development of a solvent extraction method for the determination of tungsten as tungstate.The technique is based on the solvent extraction of tungstate in acidic solution by chloroform and the subsequent measure- ment of the absorbance of iron, rather than tungsten. Because of the poor sensitivity given by the direct determination of tungsten by AAS. this method is not widely used, whereas a high sensitivity is achieved using the present indirect method. Also, although the indirect AAS determination of tungstate is made possible by precipitation of tungstate as PbW04, serious interferences may arise owing to the co-precipitation of other anions. The indirect method described here is highly sensitive, rapid, simple and free from interference from many foreign ions.Therefore, W042- can be determined accurately without using a masking agent; this is a unique advantage of the present method. In addition, the proposed method is suitable for the routine laboratory determination of tungsten as tungstate, providing an alternative to existing methods,lg viz. , spectrophotometry, gravimetry, titrimetry, spectrography, electrolysis, kinetic, X-ray fluorescence and neutron activa- tion analysis. One of the authors (D. C.) thanks the UGC, New Delhi for s po n so r s hi p . 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 69 References Westerfeld, W. W., and Richert. D. A., A n n . N . Y. Acad. Scl., 1954, 57, 896. Amos, M. D., and Willis, J . B., Spectrochim. A m , 1966, 22, 1325. Quin, B.E . , and Brooks, R. R . , Anal. Chim. Acta, 1973. 65, 206. Saltykova. A. M.. Davidovich, N. K.. and Melarned, S . H. G . , Z h . Anal. Khim., 1973, 27. 1216. Kooney, R. C., and Pratt, C. G., Analyst, 1972. 97. 300. Dharma Rao, P . , A t . Absorpt. Newsl., 1970, 9. 131. Husler, J . , A t . Absorpt. Newsl., 1971. 10. 60. Morrow, R . W., Rep. At. Energy Comm. U.S., Report No. Y-1812, 1972; Chem. Abstr., 1972, 77, 56054~. Korrey, J . S, and Goulden, P. D., A t . Ahsorpt. News1 , 1975. 14, 33. Yudelvich, I. G., and Shaburova, N . P., Chem. Anal. (Warsaw), 1974, 19, 941. Tindall, M. F . , A t . Absorpt. Newsl., 1977, 16, 37. Musil, J . , and Dobezal, J . , Anal. Chirn. Acta, 1977. 92, 301. Henrion, G., Gelbrecht, J . . Hoffman, T., and Marquardt, D . , Z . Chem., 1981, 21. 192. Chong, R . W., and Boltz, D . F., Anal. Lett.. 1975, 8, 721. Bassett, J . , Denney, R. C., Jeffery, G. H., and Mendham, J . , Editors, “Vogel’s Textbook of Quantitative Inorganic Analysis Including Elementary Instrumental Analysis,” Fourth Edition, ELBS and Longmans, New York, 1978, pp. 360 and 486. Ottaway, J. M., Coker, D. T., and Singleton, B., Tafanta, 1972, 19, 787. Snell, F. D., and Snell, C. T., “Colorimetric Methods of Analysis Including some Turbidimetric and Nephelometric Methods,” Volume 11, Third Edition, Van Nostrand, New York, 1949, p. 464. Long, G . L., and Winefordner. J . D., Anal. Chern., 1983, 55, 712A. Rodriguez-Vazquez, J . A , , Talanta, 1978, 25, 299. Paper 8102648C Received July 4th, 1988 Accepted August 22rzd, I988

ISSN:0003-2654    

DOI:10.1039/AN9891400067

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 113 71 Artifactual Chemiluminescence in the Determination of Nitrogen Monoxide in the Vapour Phase of Mainstream Cigarette Smoke Peter R. Houlgate and William H. Evans Department of Trade and Industry, Laboratory of the Government Chemist, Queens Road, Teddington, Middlesex TWII OLY, UK Evidence is presented for the artifactual chemiluminescence from olefins in the determination of nitrogen monoxide in the vapour phase of mainstream cigarette smoke. The interference may be quantified by measuring the residual chemiluminescence remaining in the vapour phase of mainstream cigarette smoke after 24 h. For the nitrogen monoxide analyser used, corrections to the yield of most cigarette brands retailed in the UK were found t o be about 1.5 pg of nitrogen monoxide per puff, each machine puff being defined by inter nat i o na I sta nda rds.Keywords : Interferences; nitrogen monoxide; mainstream cigarette smoke; chemiluminescence analysis It has been established that the nitrate content of tobacco is one source of the nitrogen oxides present in cigarette smoke. 1.7 The thermal decomposition of alkali and alkaline earth metal nitrates yields nitrogen monoxide (NO) and nitrogen dioxide (NO:,). The equilibrium shown in equation (1) is temperature dependent with the dissociation to NO being complete at temperatures >A20 "C.3 At a burning cigarette tip (70&9OO0C), any NO2 formed should be con- verted into NO. Repeated investigations+" have indicated that the concentration of NO2 in fresh cigarette smoke is minimal, confirming these basic considerations.2N07 e 2 N O + 0 7 . . . . (1) Once formed. NO is drawn through the tobacco rod, cooling rapidly, and is diluted with increasing concentrations of air giving favourable conditions for oxidation of the monoxide according to the series of reactions shown in equation (2). . . (2) 2NO + 0: * [NO]:, + O7 fast\ 2N02 e Nz04 and subsequently N 2 0 4 + H:,0 + HN02 + FINO3 3HN02 + HNO3 + 2 N 0 + H2O At ambient temperatures the tetroxide dimer undergoes approximately 16% dissociation to the dioxide, although the latter is normally reported to be the primary oxidation product. The rate of oxidation of NO has been studied in detail. Third-order kinetics apply, which in the presence of an excess of oxygen becomes pseudo second-order.The reported rate constant is 7.3 X 109 cm" mol-2 s - J at 20 OCJO; this reaction rate has been confirmed in dilute mainstream cigarette smoke (MCS).h.ll Hence the measurement of NO in the vapour phase of MCS requires the immediate analysis of each machine puff. Because MCS can be monitored on a puff to puff basis, using a manifold linear or rotary smoking machine, cheniilumi- nescence analysis has superseded the earlier spectrophotome- tric methods for the measurement of NO in MCS. The mechanism o f the reaction between NO and ozone is given by the series of reactions shown in equation (3) in which NOz" is the excited state emitting radiation in the range 600-3000 nm with an emission maximum at 1200 nm. Only a fraction of the NO molecules are excited to NO2* and quenching with a species M allows only a fraction of the NOz" to emit radiation.In practice, equalisation of this quenching for the standard and sample atmospheres is ensured by measurement in a nitrogen flow, although it has been demonstrated that carbon dioxide and water vapour are more effective quenching agents than nitrogen,13>l4 and both are present at high concentrations in the vapour phase of MCS. It has also been shown that this additional quenching can be contained by a rapid 2 : 1 dilution of the vapour phase of MCS with nitrogen.8 However, this quenching remains a function of the design of the NO analyser. . . . . ( 3 ) 1 NO + 0 3 + NO:," + 0 2 NO + 0 3 + NO2 + 0 2 NO2" +NO,+hv NO,"+M+NO,+M+heat It has been established using chemiluminescence analysis that NO yields in MCS are related to native or added nitrate levels present in the cigarette tobacco.'l31s Increasing levels of volatile amine bases in the tobacco do not contribute to the MCS yield of NO,'j.IS but these sources of nitrogen do contribute to the sidestream cigarette smoke (SCS) yield.9 For both MCS and SCS, when the relationship between NO yield and nitrate content of the cigarette tobacco is extrapo- lated to a zero nitrate content, a large residual yield of NO remains. Also, it has been found that the SCS from cigarettes made from nitrogen-free cellulose is about one seventh of these residual SCS yields.confirming that amine bases are the precursors of NO, whereas the MCS yield of NO is identical with that obtained from extrapolation to zero nitrate for cigarettes with various, or added, nitrate levels.9 It has been suggested that the residual source of NO in MCS arises from the oxidation of atmospheric nitrogen at the elevated temperatures found at the burning tip of a cigarette.This explanation seems unlikely in the absence of an electrical discharge. It has been observed in this laboratory that 24 h after collecting the vapour phase from each cigarette brand in a polythene bag, a relatively constant apparent NO response is obtained. If pseudo second-order kinetics for the oxidation of NO are obeyed, then decay should proceed to below recognisable levels. The final product of the oxidation, i.e., nitric acid, would not be expected to give a positive response with ozone. The residual NO response from nitrogen-free cellulose cigarettes and that observed after 24 h may represent artifactual chemiluminescence, not previously defined, from a component, or components, of the vapour phase of MCS.Investigations into the delineation of these interferences are described in this paper. Experimental Apparatus Automalic smoking machine. A Borgwaldt Model RM20CS automatic rotary smoking machine was used according to IS072 ANALYST, JANUARY 1989, VOL. 114 3308: 1986. This machine is a restricted smoker and should be adjusted to produce a bell-shaped puff profile with a puff volume of 35 k 0.25 ml, a puff duration of 2 -t 0.05 s and a puff frequency of 60 k 0.5 s during smoking. Nitrogen monoxide analyser. A ChemLab chemilumi- nescent analyser was used. This has a facility for introducing the vapour phase into the reaction cell via a gas sampling valve with a 1-ml capacity sampling loop.This sample is then diluted with nitrogen and ozone-enriched oxygen at the maximum available flow-rate of 2.5 ml 5-1 to equalise the quenching reaction between the sampling standards and the cigarette smoke. Other chemiluminescent reactions, e.g., from ozone - oxygen mixtures or from the reaction of ozone with other smoke constituents, were excluded by using a red transmission filter. Type 2030, to remove radiation with a wavelength below 650 nm, measurements being performed in the range 650-950 rim. The gas sampling valve of the analyser can be operated manually for calibration or for experimental purposes regard- ing interference.or automatically, using a signal from the smoking machine. Integrator. An LDC Model 308 integrator was used to integrate peak heights whether measured manually or on a puff to puff basis. Materials Carbon dioxide, oxygen and oxygen-free nitrogen were supplied by BOC and ethylene, propylene and hydrogen sulphide by Matheson. Nitrogen monoxide. Certified standard mixtures in nitrogen with concentrations of 100, 400, 700 and 1000 v.p.p.m. of analyte, Hy-line grade supplied by BOC (throughout this paper v.p.p.m. denotes gas concentrations as parts per million by volume). Carbon monoxide, 8% VIV (Alpha grade, BOC). Certified standard mixture. Cigarettes. Popular UK brands together with Kentucky reference cigarettes IR3 and TR4F were used. Before machine smoking, the cigarettes were placed in an environmental cabinet for 48 h with the atmosphere maintained at a temperature of 22 t 1 "C and at a humidity of 60 k 2% ( I S 0 3402 : 1978).All cigarettes were inserted to a depth of 9 mm in the smoking machine labyrinth holders and smoked to the following butt lengths: plain cigarettes, 20 mm; filter cigarettes S75 mm in length, overwrap length plus 3 mm; filter cigarettes >7S mm in length, overwrap length plus 5 mm; TR3, 30 mm: and IR4F. 35 mm. Procedure The NO analyser was calibrated manually by taking the average of 6 1 0 integrated peak height readings for each standard. repeating after each series of experimental readings and averaging for each standard. Gaseous atmospheres were prepared by venting into evacuated polythene bags.Using the smoking machine piston ~ gaseous mixtures were also prepared by venting into evacu- ated polythene bags. The responses were measured manually and averaged for 6-10 applications. The vapour phase of MCS was obtained by smoking ten cigarettes of one brand placed in alternate ports of the Borgwaldt rotary smoking machine and collecting the vapour in polythene bags after filtration through a 44-mm filter pad held in a Cambridge filter-holder. Responses after 24 h (or thereafter) were measured manually and averaged for ten applications. The MCS yield was obtained similarly, using a smoking machine, by integration of the NO response on a puff to puff basis. All responses were converted into concentration (v.p.p.m.) via the regression slope for calibration and calculated at standard conditions of temperature and pressure (STP). Results and Discussion Other Sources of Chemiluminescence in the Determination of Nitrogen Monoxide It has been reported that some of the components found in the vapour phase of MCS.e.g., nicotine, ammonia, methane, methanol, acetaldehyde, acrolein, benzene. hydrogen sul- phide, pyridine, halogenohydrocarbons, nitromethane and nitrobenzene, do not interfere in the chemiluminescent reaction of nitrogen monoxide with ozone. Any radiation emitted by these components was at wavelengths below 600 nm and was filtered before detection.8 However, it has also been recognised that a chemilumi- nescent reaction occurs between ethylene and ozone,16 and further investigation has indicated that the emission below 650 nm due to higher olefin homologues, e.g., cis-but-2-ene and trans-but-2-ene, may be ten times greater than that due to ethylene.17 The reaction of 5 v.p.p.m.of ethylene with ozone has been reported as giving no measurable chemiluminescent response at wavelengths >650 nm,lx whereas olefins give responses 100 times lower than that of NO above 590 nm.19 In addition, the chemiluminescent response of olefins with ozone in the presence of hydrogen sulphide has been reported to be enhanced 50-fold when measured over the sensitive range of a photomultiplier.20 Hydrogen sulphide is present in the vapour phase of MCS at levels of 18-50 pg per cigarette (equivalent to 40-1 10 v.p.p.m.).2* Atmospheres of each major component of the MCS vapour phase were prepared in polythene bags, mixed with air and measured according to the described procedure. With the NO analyser used in this work, no positive chemiluminescent response was obtained with carbon monoxide, carbon dioxide or ammonia, or with various paraffinic hydrocarbons and aldehydes or methyl nitrate (formed in situ from methanol and NOz).Conversely, olefins such as ethylene (b.p. -102.4 "C) and propylene (b.p. -47.7 "C) gave positive responies. Each is preyent in the MCS vapour phase at a level of about 250 pg per cigarette,22 and the total of the four C4 alkenes (b.p. -6.6 to +3.7 "C) is about 150-160 pg per cigarette. These values are equivalent to ca. 640 v.p.p.m. of ethylene, ca. 420 v.p.p.m. of propylene and ca. 200 v.p.p.m. of C4 alkenes at STP. The concentration5 for the first two olefins gave responses of apparent NO which were difficult to measure accurately.At the higher concentra- tions given in Table 1, the response for ethylene appeared to Table 1. lnterfcrence from gaseous components of MC'S in the chemilurninescent measurement o f N O Concentration, Apparent concentration Component v.p.p.m. ofNO at STP,v.p.p.m. Acetaldehyde . . . . Methanol: . . . . . . Ammonia . . . . . . Methane-ethanc . . . . Chrbonrnonoxidc . . . . Carbondioxide . . . . Ethylene . . . . . . Propylene Hydrogen sulphide 1000 8000 10 000 20 01)o 80 000 200 000 SO00 10 000 20 000 1000 2000 SO00 10 000 20 000 500 800 1500 N D = not detected. + In the presence of N20d and measured after 24 h. ND' N D N D N D ND ND 36 68 I35 22 37 82 15 1 282 25 49 82ANALYST, JANUARY 1989, VOL.114 73 be linear, with an interference factor of 6.9 x 10-3 v.p.p.m. of NO at STP. Propylene gave a non-linear response, decreaving as the concentration (STP) decreased, which can be defined by the following power regression ( Y = +0.9996): Apparent NO (v.p.p.m.) = 0.0541 propylene (v.p.p.m.)O X(,19 Using this equation, the interference factor at, for example, 320 v.p.p.m. of propylene was calculated to be 23.5 x 1 0 - 3 v.p.p.m. of NO. The interference factors given in Table 2 were obtained by maintaining the ethylene atmospheres in air and by carrying out measurements on consecutive days. Further, after 24 h, when complete oxidation of NO might be expected, atmospheres of ethylene and NO in air yielded an interference factor of 6.7 x 10 v.p.p.m. of NO, similar to that obtained in the absence of NO (Table 2).It would appear that the decrease of the interference factor with time is caused by diffusion losses, and this would apply equally to the aged vapour phase of MCS retained for 24 h or longer. Hydrogen sulphide yielded an average interference factor of 56 x lo-? v.p.p.m. of NO at STP (Table 1). For 4CL110 v.p.p.m. of NO in the MCS vapour phase, NO yields were enhanced by only 1-3 pg of apparent NO per cigarette. However, using the described analyser, the enhancement of olefin chemiluminescence by hydrogen sulphide, which is noted for its wider range of wavelength measurement, was evident in the range 65CL950 nm (Table 3). Hydrogen sulphide is extremely reactive and may combine with other components of the MCS vapour phase during chemiluminexence, in preference to the olefins.The MCS vapour phase obtained from ten cigarettes of a single brand, machine smoked as described under Procedure, was analysed for its average hydrogen sulphide content using Draeger tubes attached to the exhaust vent of the chemiluminescent reaction cell of the NO analyser, and with the ozone generator snitched off. The use of tubes (25 cm x 6 mm i.d.), loosely packed with glass wool, on to which 1 ml of a saturated basic lead acetate solution had been absorbed, and interposed on the suction side of the rotary smoking machine piston, reduced the hydrogen sulphide concentration in the MCS of the cigarette brand studied from 40 to <5 v.p.p.m. Table 2.Decrease in ethylene interference with time Interference factor" x Ethylene concentration. v,p.p.m. Average interference Time/h 5000 10000 20000 factor x 10-3 0 7.2 6.8 6.8 6.9 24 6.4 6.5 6.4 6.4 48 5.9 5.0 5.8 5.6 72 4.9 5.1 5.0 5.0 Calculated at STP. The NO yields in the MCS vapour phase of three cigarette brands were compared, using lead acetate - glass-wool absorp- tion tubes and similar tubes containing glass-wool saturated with 1 ml of water only. The results, for NO and apparent NO per cigarette, are given in Table 4; they indicate that there are no statistically significant differences in yields when the hydrogen sulphide is almost completely removed. It was concluded that hydrogen sulphide does not enhance the artifactual chemiluminescence from olefins in the MCS vapour phase.Quenching in the Chemiluminescent Measurement of Nitrogen Monoxide Two gaseous products of MCS that are present in high concentrations are carbon monoxide (2-6% VIV) and carbon dioxide (9% VIV),22323 neither of which gave a direct chemiluminescent response with ozone (Table 1). Neverthe- less. the quenching of NO chemiluminescence may occur with carbon dioxidel3.14 and this may be a reflection of the type of NO analyser being used. Atmospheres of NO in N2 and C 0 2 were prepared by venting into evacuated polythene bags, taking care to elimi- nate air during preparation. Recovery ratios for these mixtures are shown in Table 5 . Linear regression of a plot of recovery ratios veYsus carbon dioxide concentration ('/o VIV) gives the following equation ( Y = -0.9546): Recovery ratio = 0.979 - 0.00S787(C02, Yo VIV) The value of the intercept, 0.979, provides evidence for oxidation during the preparation of the C02 - NO atmos- pheres, but the slope of the graph suggests that, with the NO analyser used, the quenching of the NO response is 5.2% for C 0 2 levels of ca.9% VIV in the vapour phase of MCS. For carbon monoxide tested in the range 2-6% V!V, the average recovery ratio of 0.974 (Table 5 ) is the same as the recovery ratio intercept for zero C 0 2 and hence it can be concluded that there is no quenching of the NO chemilumi- nescence. It was found that the degree of quenching of the olefin chemiluminescence with 10% VIV C02 was 5.8% for lO000 v.p.p.m. of ethylene and about 5.9% for 5000 and 10000 v.p.p.m.of propylene. It seems likely that C02 will quench the artifactual olefin chemiluminescence to a similar extent to that found for NO. Artifactual Chemiluminescence in Aged Mainstream Cigarette Smoke To ensure that the oxidation of NO proceeded according to pseudo second-order reaction kinetics for the equipment used, atmospheres of NO in N2 and air were prepared to give an oxygen content of 15.1% VIV, typical of the level in MCS, Table 3. Enhancement of olefin chemiluminescence by hydrogen sulphide Hydrogen sulphide, v.p.p.m. 0" 20 40 60 80 100 Apparent NO, v.p.p.m. t Olefin Con ce n t r a t i on , v . p . p . m . Ethylene . . . . . . 1000 (6.9)* ~ 1 3 . 0 15.0 18.6 26.2 32.1 Propylene . . . . . . 500 (11.S)* 16.2 23.2 29.2 36.8 45.9 Enhancement ratio - Ethylene .. . . . . 1000 z 1 . 9 2.2 2.7 3.8 4.6 Propylene . . . . . . 500 - 1.4 2.0 2.5 3.2 4.0 * Expected apparent NO concentration from olefin. f Calculated at STP and corrected for artifactual chemiluminesccncc from hydrogen sulphide.74 ANALYST, JANUARY 1989, VOL. 114 Table 4. Comparison of NO yields with and without the removal of hydrogen sulphide NOipg per cigarette$ Hydrogen Hydrogen Tar sulphide sulphide yield?: not removcd removed Cigarette mg per Puffs per type" cigarette cigarette Mean SD Mean SD rsv . . . . 14 11.1 83 3.3 79 4.9 KS . . . . . . 11 10.1 139 2.7 137 5.8 Regular . . . . 13 8.4 487 20.5 482 33.6 IS = international size; V = ventilated; and KS = king size. t Tar yield is particulate matter, water and nicotine free. i: Mean and SD of six determinations.Table 5. Quenching effect of carbon monoxide and carbon dioxidc on the N O response Recovery Concentration. ratio Interferent Yo VW NO, v.p.p.m. for N O Carbon dioxide Carbonmonoxide . . 4.8 4.8 9.1 13.0 16.7 20.0 20.0 33.0 2.0 4.0 6.0 6.0 96 0.989 950 0.948,0.943 950 0.928 950 0.900,0.886 2180 0.867,0.873 670 0.868 950 0.870 950 0.800 660 0.980 490 0.976 520 0.968 560 0.972 Table 6. Decay o f nitrogen monoxide with time Concentration of NO." v.p.p.m. Decay ratci'/ Timeimin Measured Theoretical cmh mol- s I 0 15 30 60 120 180 240 360 1430 305 140 92 54 31 26 18 15 7 - 144 91 55 28 21 16 11 3 7.58 7.35 7.47 6.70 6.20 6.58 5.41 3.12 * At STP. + The calculated theoretical average value is 7.34 x 10" cmh mol-2 s-1. and the NO concentration was measured periodically for up to 24 h.These atmospheres were contained in polythene bags that had been conditioned to the vapour phase of MCS. The exercise was repeated three times and the average results obtained are presented in Table 6. The expected reaction kinetics for the oxidation of NO were obeyed for the first 60 min but thereafter the rate of oxidation slowed. After 24 h there was a residual level of 7 v.p.p.m. of NO which remained constant for the next 48 h. Thi5 was some 4 v.p.p.m. of NO greater than expected after 24 h and reflected the value of the measurement blank. Ten cigarettes of a single brand were smoked on a Borgwaldt rotary smoking machine and the vapour phase of the MCS was collected in a polythene bag. The initial concentration wa5 not measured because collection lasted about 10 min and oxidation would not permit a meaningful initial NO yield to be obtained.The oxygen concentrations in these atmospheres were monitored and averaged 14.8% VIV. After 6, 23. 48. 72 and 96 h the residual responses were Table 7. Decline in the apparent response to nitrogen monoxide in aged vapour-phase MCS Apparent concentration of NO, v.p.p.m.? Tar Cigarette yield/ After After After After After type* mg 6 h 24h 48h 72h 96h Regular$ . . . . . . 12 36 29 26 19 18 KSV . . . . . . . . 13 42 28 19 17 16 Regular . . . . . . 13 50 29 30 25 19 KSV . . . . . . . . 13 58 45 39 37 31 TSV . . . . . . . . 14 51 37 31 25 21 KSS . . . . . . . . 15 54 37 34 27 22 KS . . . . . . . . 15 58 43 36 32 26 RegularP . . . . . . 17 37 22 21 17 11 Mini .. . . . . . . 17 38 29 21 17 16 KS . . . . . . . . 17 49 31 24 18 13 RegularP . . . . . . 18 46 25 21 16 14 RegularP . . . . . . 25 48 24 26 19 14 Average: . . . . . . - 48.1 31.6 27.3 22.4 18.4 Range: . . . . . . 12-25 36-58 22-45 19-39 16-37 11-31 NOdecayB: 15 7 7 7 - . . . . . . Ethylenel: . . . . . . 35 32 28 25 - Lo w-tar ventiluted cigarette brands- EMKSV . . . . . . 4 32 23 19 13 10 KSV . . . . . . . . 9 28 19 16 1 1 10 KSV . . . . . . . . 9 30 20 15 13 4 Average: . . . . . . - 30 20.7 16.7 12.3 8.0 Range: . . . . . . 4-9 28-32 19-23 15-19 11-13 4-10 ventilated; and P = plain. * EM = extra mild; KS = king size; IS = international size: V = -t At STP. 3: These cigarettes would be expected to yield a vapour-phase mainstream smoke containing 7&290 v.p.p.m.o f NO; those contain- ing air-cured tobacco having yields o f 12 and 15 mg of tar would be expected to yield vapour phases containing 1000 and 500 v.p.p.m. of NO, respectively. $ Refer to Table 6. r[ Value for 5000 v.p.p.m. of ethylene calculated from the average values given in Table 2; the value after 6 h is the zero-hour response. measured as v.p.p.m. of NO. This procedure was repeated for various types of cigarette brands retailed in the UK and the results, corrected to STP, are shown in two series in Table 7. The second series lists low tar ventilated brands only. Also shown in Table 7 are the levels of NO found experimentally after oxidation for the relevant periods of time and the decrease in the apparent NO concentration with time for SO00 v.p.p.m.of ethylene (from the average values given in Table An examination of these results indicated an average level of apparent NO for the cigarette brands of about 32 v.p.p.m. after 24 h compared with the average of 21 v.p.p.m. for low-tar ventilated brands. Measurements from similar atmo- spheres, after 24 h, from a range of different cigarette brands confirmed these averages for each series with values of 34 and 20 v.p.p.m., respectively (Table 8). For each series of cigarette types the response towards apparent NO decreased at a rate of 3.6-4.4 v.p.p.m. per 24 h, which was similar to the decrease in the equivalent apparent NO from 5000 v.p.p.m. ethylene atmospheres (3.7 v.p.p.m. per 24 h) and was attributed to diffusion losses. To correct for the artifactual chemiluminescence present at 0 h required extrapolation of these losses from the artifactual chemiluminescence measured after 24 h.The MCS vapour phase from king size filter cigarettes collected in a polyethylene bag and sampled immediately contained 680 v.p.p.m. of ethylene, 480 v.p.p.m. of propylene 2)-ANALYST, JANUARY 1989, VOL. 114 75 Table 8. Apparent concentration of NO (measured at STP) in 24 h aged vapour-phase mainstream smoke from other cigarette brands. Abbreviations are as given in Table 7 Cigarette type * ~ KS KS KS RV IS ISV KS P Average Taryieldlmg . . 10 11 13 14 14 14 17 2.5 - Residual N0.v.p.p.m. 32 39 29 39 42 30 31 30 34.0 Range of apparent residual NO: 29-42 v.p.p.m. Low-tar ventilated cigarettes- UMKSV KSV KSV KSV Average - Taryieldimg .. . . 4 9 9 9 Residual NO, v.p.p.m. 18 17 22 20 19.5 Range of apparent residual NO: 17-22 v.p.p.m. These cigarettes would be expected to yield a vapour-phase mainstream smoke containing 120-280 v.p,p.m. of NO; those containing air-cured tobacco having yields of 10 and 25 mg of tar would be expected to yield vapour phases containing 760 and 660 v.p.p.m. of NO, respectively. and about 200 v.p.p.m. of combined C4 alkenes (at STP), as measured by headspace capillary gas - liquid chromatography. After 24 h this atmosphere contained 580 v.p.p.m. of ethylene, 430 v.p.p.m. of propylene and gave similar responses to the original response for the C4 alkenes. Ethylene and propylene would be expected to contribute about 15 v.p.p.m. of apparent NO to the artifactual chemiluminescence measured after 24 h.The remaining apparent NO originates from other volatile olefins, e.g., C4 alkenes, which give a high chemiluminescent response with ozone. 17 Therefore, in order to determine NO in the vapour phase of MCS it was necessary to measure the artifactual response by collecting the vapour phase for each cigarette brand examined once and measuring this response after 24 h. If this response is x v.p.p.m. (at STP), then the correction to be applied for artifactual chemiluminescence is (x - 7 + 4) v.p.p.m. for the procedure used in this laboratory with the present analyser. This value should be corrected further by a factor of 1.055 to account for quenching by CO2. Hence, for the measured averages of 33 and 20 v.p.p.m. in Tables 7 and 8, obtained after 24 h, the calculated yields should be reduced by 1.48 and 0.84 pg of NO per puff per cigarette.This gives a corresponding correction of between 6.3 and 17.8 pg of NO per cigarette for cigarette brands with puff counts in the range 7.5-12.0. which can be compared with an average yield of some 60 pg of NO per cigarette for cigarette brands manufactured in the UK. Conclusion The artifactual chemiluminexence of olefins with ozone interferes with the determination of nitrogen monoxide in the vapour phase of MCS. For the NO analyser used in this work, the maximum artifactual response is equivalent to 18 ug of NO per cigarette. This is lower than the residual yields of 25-50 pg of NO observed previously from cellulose nitrogen-free cigarettes. However, it seems likely that the latter residual yields originate wholly, or in part, from the artifactual chemiluminescence of olefins with ozone, the extent of which may depend on the type of NO analyser used for measure- ment.This work was funded by the Department of Health and Social Security and is published with the permission of the Government Chemist. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. References Neurath, G., and Ehmke, H . , Beitr. Tabakforsch., 1964. 2, 333. Broaddus, G. M., York, J . E.. and Moseley, J . M., Tob. Sci.. 1965, 4, 149. Latimer, W. M., and Hildebrand, J . H., “Reference Book o f Inorganic Chemistry,” Third Edition, Macmillan, New York, 1964, p. 206. Norman, V., and Keith, C. H.. Nature (London), 1965, 205, 915.Barkemeyer, H., and Seehofer, F.. Beitr. Tubukforsch., 1968, 7, 278. Sloan, C. H.. and Kiefer, J. E., Tob. Sci., 1969. 13. 180. Neurath, G. B., Dunger, M., and Pein, F. G.. IARC Scientific Publication No. 14, International Agency for Research on Cancer, Lyon, 1976, p. 215. Jenkins, R. A.. and Gill, B. E . , Anal. Chem., 1980, 52, 925. Umemura, S . . Muramatsu, M., and Okada. T., Beitr. Tubak- forsch., 1986, 13, 183. Baulch, D. L.. Drysdale, D. D., Home, D. G., and Lloyd, A. C . , “Evaluated Kinetic Data for High-temperature Reac- tions: Volume 1, Homogeneous Gas-phase Reactions of the H, - O2 System,” Butterworths, London, 1972. Borland, C. D. R., Chamberlain, A. T., Higcnbottam. T. W., Barber, R. W., and Thrush. B. A.. Beitr. Tabakforsch., 1985, 13, 67. Steffenson, D. M., and Stedman, D. H.. Anal. Chrm., 1974, 46, 1704. Klimisch, H. J . , and Kircheim, E.. 2. Lehensm. Unters. Forsch., 1977, 163, 48. Matthews, R . D . , Sawyer, R. F.. and Schefer, R . W., Environ. Sci. Technol., 1977. 1 1 . 1002. Norman, V., Ihrig, A. M., Larson, T. M.. and Moss, B. I-., Beitr. Tabakforsch., 1983, 12, 55. Kederbragt, G. W., van der Horst, A . , and van Duijn, J . , Nature (Londotz), 1965, 206, 87. Kummer, W. A , , Pitts, J . N., and Steer, R. P., Environ. Sci. Technol., 1971. 5 . 1045. Fontijn, A , , Sabadell. A . J., and Ronco, R . J . , Anal. Chem., 1970, 42, 575. Sigsby. J . E . , Black, F. M., Bellar, T. A , . and Klosterman. D . L., Environ. Sci. Technol., 1973, 7, 51. Wiese, A. H . , Henrich, K. K., and Schurath, W., Environ. Sci. Technol., 1979, 13, 85. Collins, P. F . , and Williams, J . F., Heitr. Tubukforsch., 1979. 10,24. Wynder. E. L., and Hoffmann, D . , “Tobacco and Tobacco Smoke,” Academic Press, New York. 1967. pp. 442-446. Brunnemann, K. D., and Hoffmann, D., 1. Chromatogr. Sci., 1974, 12, 70. Paper 81020063 Received May 20tt1, 1988 Accepted August 25th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400071

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 77 Micellar Enhancement of Benzodiazepine Fluorescence M. de la Guardia Cirugeda and F. Rodilla Soriano Department of Analytical Chemistry, Faculty of Chemistry, University of Valencia, 50 Dr. Moliner Street, Bu rjaso t, Va len cia, Spa in The interactions that exist between benzodiazepines and surfactants provide micellar enhancement factors for their fluorimetric determination in the range 1.2-6.5, depending on the nature of both the benzodiazepine and the surfactant. A series of benzodiazepines and anionic surfactants were treated topologically to determine the influence of each benzodiazepine substituent on the basic benzodiazepine structure and the influence of both the hydrophobic moiety of the surfactant and its counter ion on the sensitisation process.Sensitisation parameters were used to quantify the effect of the chemical structures of both surfactants and drugs on their interaction. Keywords : Benzodiazepin es; flu0 rime tric determination; micellar enhancement; an ionic surfactan ts; topological description Benzodiazepines are important psychotherapeutic agents that act on the central nervous system and they are widely used for the treatment of anxiety, insomnia and epileptic convulsions owing to their low toxicity and limited side effects.'-3 These compounds have been identified and determined in pure drugs. medicines, biological fluids and tissues using a variety of methods.+q The fluorimetric determination of benzodiazepines can be carried out in one of three ways: (i) by direct measurement of their fluorescencel".ll; (ii) by prior hydrolysis~2--~4; or (iii) by prior reaction with various substances to form highly fluores- cent derivatives.ls.16 The limited development of simple methods for the fluorimetric determination of benzodiazepines is possibly due to the low sensitivity of their natural fluorescence.For this reason, the fluorimetric study of the interaction of these compounds with micellar media could prove interesting as a means of increasing their fluorescence quantum yield. However, little work has been reported on the fluorirnetric determination of psychotherapeutic drugs in micellar media and an initial study concerning anorexics, tranquillisers and other drugs is currently in progress.17.18 The addition of a surfactant to a chemical system causes significant changes in behaviour.19-21 Such changes depend on the surfactant concentration and occur at concentrations higher than the critical micellar concentration (CMC) ,22 which indicates that they are the result of micellar formation and not of interactions of the type that occur in co-ordination compounds and ion pairs. At present there is only limited knowledge of the factors governing the interaction of chemical systems with micelles and, in fact, the only norm that exists is that of charge compatibility between the micelle and the analyte.23 However, this explains neither the interaction that is possible with non-ionic micelles nor why the same types of surfactant can cause notably different effects. Hence, investigations into the interactions that exist between different surfactants and a particular system, or analogous systerns,2426 could help in establishing the conditions that result in the best analytical characteristics and a set of rules for predicting the sensitisation conditions of other chemical systems.The aim of the work described in this paper was to study the interaction of different benzodiazepines with various surfac- tants in order to establish more sensitive methods for the determination of benzodiazepine tranquillisers using micellar media. In the fluorimetric investigation of the interaction of benzodiazepines with anionic surfactants, we have tried to establish the influence of both benzodiazepine substituents and surfactant structure on the sensitisation process.Theory Topological and group contribution models27-'] have been used in chemical engineering to establish the laws of variation of the physical and chemical properties of a homologous series of compounds and, more importantly from a technical point of view, to predict the behaviour of unsynthesised compounds and to establish the properties of mixtures of compounds.32.-'' From a chemical - analytical point of view, we have applied the topological treatment to a series of ethylene oxide condensate surfactants in order to establish the laws of variation of both their physical34 and spectral's properties with the average degree of condensation. IIence, graphs were constructed for the characterisation of these compounds. In this study a topological treatment was used to quantify the influence of the structures of both anionic surfactants and benzodiazepines on the sensitisation of their fluorescence in micellar media.It was assumed that the molecule of an anionic surfactant consists of two components: a strongly hydrophobic moiety, A, and a counter ion, C. Any property PTi of a surfactant T,, which depends on the surfactant structure, can be separated into two contributions: PTi = SPA, + SPC, . . . . . . (1) where SPAi and SPC, are the sensitisation parameters (SPs) corresponding to each part of the molecule. In order to study the effect of surfactant structure on the fluorescence sensitisation of a benzodiazepine it is necessary to establish a variable that quantifies the process and to establish its value for a sufficiently wide range of systems under the same conditions. For this purpose the micellar enhancement factor (MEF) of the benzodiazepine, defined as the ratio of the slopes of the calibration graphs obtained in the presence and absence of a surfactant, was used.It can be assumed that the MEF obtained in the presence of a given surfactant depends on the hydrophobic moiety and its counter ion. Hence. for a species T , , MEFT, = SPA, + SPc, . . . . . . (2) For a group of surfactants Ti having x distinct structures of the hydrophobic moiety and y different counter ions, x + y = i , . . . . . . . (3) Hence it can be established that CMEF.T, = :SPAI + CSPC, . . . . (4) 1 J ' Equation (4) describes a system of i independent equations with i unknowns when there are i representative surfactants having the same number of possible structures for the hydrophobic moiety and for the counter ion.This system can be described as a combination of matrices consisting of a78 ANALYST, JANUARY 1989, VOL. 113 vector of the experimental MEF values, a vector of SPs and a topological matrix describing the structure of the different surfactants used: . . ( 5 ) where M is a square matrix of i rows and i columns. The iz components comprising the topological matrix correspond to the different hydrophobic groups and counter ions belonging to the group of molecules studied. Using binary notation to indicate either the absence or presence of a determinate part in the molecule, the vectors (the rows of the matrix) describe each of the surfactants in relation to the whole.Because of the experimental errors inherent in the determi- nation of the MEF values, equation ( 5 ) does not have a unique solution as it does not represent a compatible and determinate system. Therefore, it is necessary to adopt some method of rough calculation. In this work the optimum SPs were calculated from the experimental MEF values for a series of well-characterised surfactants in order to minimise the differ- ence, established by a least-squares method, between the MEF values calculated from the SPs and the experimentally obtained values. To calculate the parameters corresponding to the sensitisa- tion of the benzodiazepine fluorescence by anionic surfac- tants, a semi-logarithmic method was used to establish the following relationship: .. (6 The effect of eight anionic surfactants on the fluorescence of diazepam was investigated. The surfactants, which provided Table 1. Topological description of the anionic surfactants for determining the influence of their structure on miccllar enhancement of diazepam fluorescence Hydrophobic Counter Surfactant part ion SDS TLS SLES DSS AIS ALES SAPES scu-0s . . . . . . A, A, A, A2 A , A3 A4 A5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 1 0 (1 (1 0 1 0 0 1000000 1 0 0 1 0 0 1 0 0 0 1 0 0 0 1 0 0 10000010 0 0 1 0 0 0 10 00 0 1 0 100 0000 1 100 - Table 2. Benzodiazepine structures R1 R2 /J&5kc=N/-CH-bR3 Benzodiazcpinc R, R? Diazepam . . CH, C1 Nitrazepam . . H NO? Temazepam . . CH, c1 Oxazepam . . 11 CI Lorazepam .. H Cl Prazepam . . CH7C3H5 C1 Clonazepam . . H NO? R3 H H H CI H H CI R4 H H H H OH 0 H OH different hydrophobic structures and three different counter ions were the following: lauryl sulphate, Al; decyl sulphate, A2; lauryl ether sulphate, A3; alkyl phenol ether sulphate, A,; a-olefin sulphonate, A,; Na+, C , ; NH4+, C,: and triethanol- amine, C3. The topological description of these surfactantc and an expression of equation (6), adapted to the studied population, are given in Table 1. A method, analogous to that described above, can be used to quantify the effect of benzodiazepine structures on their interaction with the anionic micelles. In this instance, after determining MEFs for different benzodiazepines in the presence of the same surfactant, equation ( 5 ) or (6) can be used to obtain SPs corresponding to each possible structure of the substituents on the benzodiazepine nucleu\.For this study seven benzodiazepines, differing only in the nature of the substituents R1, R2, R3 and R1, were used (see Table 2). Hence seven SPs were defined corresponding to the R I substituents (SPA), to the R2 substituents (Cl, SP,; and NO,, SP,,), to the R? substituents (H, SPc.; and C1, SPcr) and to the R4 substituents (H, SP,; and OH, SPDt). The three different R I substituents are distinguished according to the number of bonds or CH groups present, viz., a for H, 2a for CH3 and 5a for CH2C3Hi. This procedure simplifies the situation but can be justified on the basis that all the R1 substituents are of an analogous type and do not cause an excessive increase in molecular polarity.ANALYST, JANUARY 1989, VOL.114 79 Table 3. Topological description o f the benzodiazepines for detcrmin- ing the influence of their structure on fluroresccncc sensitisation in micellar media Benzodiazepine R I Diazepam(D) . . 2a Prazepam (P) . . 5a Nitrazepam (N) U Clonazepam (C) a Oxazepam(0) . . u Lorazepam (L) . . u Teniazepam (T) 2U MEFD MEF, MEFN MEF,. MEF, MEFL MEFo - - h = CI; h' = NO?. f c = H; c' = C1. < d = H: d' = OH. R? * b b b' b' b h h Rli. c c' c C' c C C' R4$ d d d d d' d' d' 2 1 0 1 0 1 0 5 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 1 0 2 1 0 1 0 0 1 1 1 0 1 0 0 1 1 1 0 0 1 0 1 - Hence. it can be assumed that the effect of each of these substituents on the sensitisation of the fluorescence is defined by an SP and that the over-all effect of a molecule on the sensitisation process is described by the sum of the contribu- tions of each factor.This assumption is reasonable as it was not intended to establish the contribution of each substituent to the fluorescence quantum yield of the molecule (viz., the activity of the fluorogenic groups) but rather to establish their contribution to the interaction between micelles and benzo- diazepine molecules. In addition to being dependent on the respective charges,2-7 this interaction is also highly conditioned by steric factors.25 Accordingly, the benzodiazepine MEF (MEFB) in the presence of a determinate type of micelle for a particular compound is given by: MEFH = aSPA + (SPR or SPBr) + (SPc or SPc8) + (SPD or SPDj) (7) where N corresponds to the number of groups in the substituent R1.The remaining parameters depend on the structure of the benzodiazepine studied. Based on previous considerations, a topological description of each of the compounds studied was established that made use of binary notation to indicate the presence or absence of determinate substituents R2, R3 and R4 and a relative criterion to ascertain the number of associated groups on the R , ( a ) substituent. A procedure, similar to that used for considering the contribution of the surfactant structures to the MEF (Table 1), was used to define the topological matrix given in Table 3. The product of this matrix and the SP vector is the set of MEFs corresponding to each of the benzodiazepines in the presence of each surfactant. Experimental Apparatus A Shimadzu RF-520 dual-beam spectrofluorimeter equipped with a xenon lamp and a U 135 S recorder was used for the spectrofluorimetric study and a Shimadzu UV-240 spectro- photometer equipped with a Model PR-1 recorder for the absorbance measurements.In both instances 1-cm quartz cells were used. Values of pH were determined using a Crison Model 501 digital pH meter with a precision of kO.01 unit and an HP 83 computer was used to carry out data adjustments. Table 4. Effcct of anionic surfactant on the fluorescence of diazepam Surfactant SDS . . . . . . . . TLS . . . . . . . . SLES . . . . . . . . DSS . . . . . . . . A LS . . . . . . . . ALES . . . . . . . . SAPES . . . . . . sa-0s . . . . . . . . MEF" SDf 6.2 0.2 6.5 0.1 4.8 0.4 4.8 0.3 4.3 0.2 3.1 2.5 - 2.1 - - The MEF is determined from the quotient of the experimental graphs obtained for the fluorescence of solutions containing 1-10 p.p.m. of diazepam both in the presence and absence of each surfact ant assayed.T SD = standard deviation for three independent determinations of MEF. Reagents Solutions of the following surfactants were used: Triton X-100 (a tert-octylphenol - ethylene oxide condensate) purchased from Probus; Nemol K-38 and Nemol K-3030 (nonylphenol- ethylene oxide condensates) purchased from Mass6 y Carol; cetyltrimethylammonium bromide (CTAB) purchased from Merck; sodium lauryl sulphate (SDS) purchased from Fluka: and sodium decyl sulphate (DSS), ammonium laurpl sulphate (ALS), triethanolamine lauryl sulphate (TLS), sodium lauryl ether sulphate (SLES), ammonium lauryl ether sulphate (ALES), sodium alkyl phenol ether sulphate (SAPES) and sodium a-olefin sulphonate (Scw-0s) donated by Molins-Kao. Solutions of diazepam.prazepam, nitrazepam, clonaze- pam, temazepam, oxazepam, lorazepam and medazepam in sulphuric acid were used. Diazepam was supplied by the pharmaceutical service of the Hospital of Valencia University, oxazepam from the MADAUS laboratory and the remainder of the bcnzodiaze- pines were supplied by the Bromatology Department of the Faculty of Pharmacy. All these products were pure and used without further purification. General Procedure Calibration graphs were constructed under the optimum excitation and emission conditions for each system, in both the presence and absence of each surfactant. The benzodiazepine solutions (1-10 p.p.m.) were prepared by dilution of a stock solution of the drug with 1 M H2S04.The pK, of diazepam was determined spectrophoto- metrically by measuring the absorbance of solutions of equal concentration and different pH (adjusted using 0.926 M HC1, 0.5 M NaHS04 - 1 M Na2S04 and 1 M CH3COOH - 1 M CH3COONa buffers). The surfactant CMCs were determined by measuring the surface tension (a) of solutions containing different amounts of surfactant. The discontinuity in the graphs of a vs. surfactant concentration in the presence and absence of diazepam could indicate the existence of mixed micelles. Results and Discussion Interaction of Diazepam With Surfactants Diazepam (7-chloro-l,3-dihydro- 1 -methyl-S-phenyI-2H-1,4- benzodiazepin-2-one) has an excitation maximum at 360 nm and an emission maximum at 500 nm.The diazepam fluores- cence is strongest at acidic pH and decreases as the tempera- ture decreases. Hence the optimum conditions for the fluorimetric determination of diazepam were previously established to be a temperature of 15 "C and a sulphuric acid concentration of 0.5 ~ . 3 6 However, for practical reasons a temperature of 20 "C was used in this work.80 ANALYST, JANUARY 1989, VOL. 114 The addition of non-ionic surfactants, such as ethylene oxide condensates, or of the cationic surfactant CTAB, has no effect on either the intensity or the shape of the diazepam emission band. However, the addition of 1% rnl'VSDS causes an increase in the diazepam fluorescence of the order of 600"/0, without producing either a bathochromic or a hypso- chromic shift of the fluorescence band.Sensitisation of the diazepam fluorescence is achieved by the addition of other anionic surfactants (1% rnlV). Neverthe- less and, in spite of the experimental variation in the calibration graphs obtained for the same system, it can be seen that the degree of sensitisation is different for each of the surfactants studied (Table 4). Sensitisation is highest in the presence of lauryl sulphate, triethanolamine and SDS, where- as an average MEF of 4.8 is obtained by the addition of either sodium lauryl ether sulphate or sodium decyl sulphate. For the remaining surfactant systems progressively smaller increases in sensitivity are found.Determination of the Sensitisation Parameter Corresponding to Each Anionic Surfactant Structure Having shown that the surfactant structure influences the sensitisation of diazepam, we tried to establish the contribu- tion to the over-all process of the parts (A and C) comprising the surfactant molecule. Using the topological model described under Theory, SPs were determined for each of the counter ions and each of the hydrophobic moieties of the surfactants studied. Two types of data adjustment were used, viz., linear and semi-logarithmic. The SPs obtained for a series of experimen- tal values are given in Table 5 as are the relative errors ( E ) found between the MEFs calculated using these SPs and those determined experimentally. Clearly, there is less error inher- ent in a semi-logarithmic adjustment.The SPs found indicate the greater influence of the hydrophobic moiety, which might be expected if the diazepam micelle interaction is governed by electrostatic and steric fact o r s . It can be concluded that the structure of lauryl sulphate is the most suitable for the fluorimetric determination of diazepam. The structures of decyl sulphate and lauryl ether sulphate are of comparable suitability, whereas use of the more complex structures is not recommended. ~~ ~~ 'Table 5 . Senvtisation parameters for diazepam in the presence of anionic surfactmtk obtained by both linear and semi-logarithmic adjustments of a single actual (not average) value of the MEF tor each curtactant akkayed Linear SPA, = 5.17 SP,, = 3.62 SPA, = 3.7s SP,, = 1.62 SP*, = 1.26 SP,, = 0.87 SP,, = -0.83 SPc; = 1.47 Linear ME F, x p (7.1900 6.6400 4 .A600 4.4900 4.1800 3.0700 2.4900 2.1300 MEFC',, 6.0350 6.6400 4.6150 4,4900 4.3350 2.9150 2.4900 2.1300 E .'% -2.50 0.00 3.48 0.00 3.71 -5.05 0.00 0.00 Average relative error: 1.84 Semi -1ogari thin i c SPA, = 0.59 SP,, = 0.45 SPA; = 0.4s SPA, = 0.13 SPC, = 0.20 SPA, = 0.19 SP,, = 0.03 SP, = 0.24 Semi-logarithmic MEF,,, 6.1900 6.6400 3.4600 4,4900 4.1800 3.0700 2.4900 2.1300 ME F C , 1 6.1604 6.6400 4.4800 4.4900 4.2001 3.0553 2.4900 2.1300 E. 7'" -0.48 0.00 0.48 0.00 0.48 -0.48 0.00 0.00 Average relative error: 0.24 As regards the counter ion, the use of triethanolamine or sodium salts is recommended, whereas the NH4+ ion has a negative effect on the micellar interaction process.Fluorescence of Benzodiazepines in the Presence of Anionic Surfactants All the benzodiazepines studied gave similar excitation and emission spectra in aqueous sulphuric acid media (Table 6) and the addition of anionic surfactants at concentrations below the CMC caused a fluorescence enhancement of the benzodiazepine solutions although neither hypsochromic nor bathochromic shifts were observed. The fluorescence of each of the benzodiazepines studied was determined in the presence of SDS, TLS, SLES and DSS micelles. The average MEFs obtained for each system and the standard deviations (SDs) for three independent determina- tions of the MEF are given in Table 7. In the presence of 1% of various anionic surfactants the benzodiazepine fluorescence increased and MEFs were obtained in the range 1.2-6.2.Only for temazepam, in micelles of SLES, was a 20% relative decrease in fluorescence obtained. The influence of the surfactant structure on sensitisation is again confirmed and, in addition, it has been shown that the benzodiazepine structure also affects fluorescence enhance- ment. Sensitisation Parameters Corresponding to Benzodiazepine Substituents Sensitisation parameters were calculated as described under Theory using both linear and semi-logarithmic data adjust- ments. For each MEF, the corresponding SP was determined and the difference between the experimental MEFs and those calculated using SPs was quantified. Table 8 gives the SPs corresponding to various benzodiaze- pine substituents in the presence of sodium lauryl sulphate.The linear data adjustment yields lower average relative errors between the experimental and calculated MEFs and is the more reproducible as far as the SPs are concerned; only in a few instances were low, positive values changed to negative Table 6. Excitation and emission conditions for benzodiazepines in both aqueous solution and in the presence of SDS In aqueous medium I n the presence of SDS Benzodiazepiiie A,, inm A,, /nm A,, inm A,,, h m Diazepam . . 360 500 360 500 Prazepam . . 360 500 360 490 Clonazepam . . 350 460 350 460 Temazepam . . 360 485 360 485 Lorazepam . . 370 500 370 495 Nitrazepam . . 350 460 350 450 Oxazepam . . 360 500 360 490 Table 7. Average MEF values and SDs of benzodiazepines in the presence of different anionic surfactants Surfactant SDS TLS Benzodiazepine MEF SD Oxazepam .. 3.3 0.1 Prazepam . . 5.7 0.2 Nitrazepam . . 2.2 0.1 Clonazepam . . 1.6 0.1 Temazepam . . 1.6 0.3 Lorazepam . . 2.2 0.1 Diazepam . . 6.2 0.2 MEF SD 2.6 0.7 6.2 0.1 2.3 0.1 1.8 0.1 1.2 0.2 2.0 0.3 6.5 0.1 SLES DSS MEF SD MEF SD 2.2 0.2 3.0 0.7 4.2 0.4 3.5 0.2 1.7 0.2 1.6 0.2 1.4 0.1 1.2 0.1 0.8 0.0, 1.2 0.3 1.2 0.0, 1.9 0.1 _ _ _ _ _ ~ 4.8 0.4 4.8 0.3ANALYST. JANUARY 1989, VQL. 114 81 values. Nevertheless, SPs obtained by the adjustment of the total experimental MEF data were used as a good estimate of the contribution of each substituent to the interaction process. The SPs and the corresponding errors were of the same order of magnitude for each system and, in general, linear adjustment gave the best results; in all instances average relative errors for the MEFs of less than 20% were obtained. Table 9 summarises the SP data for a variety of substituents in the presence of each surfactant, using linear adjustment.Because the SPs obtained under different conditions are of the same order of magnitude, the validity of the proposed method is confirmed. From the values of the sensitisation parameters it can be concluded that, typically, the R1 and R3 substituents have a negative effect on the sensitisation process and the R2 substituents, particularly Cl, contribute positively to the sensitisation. As regards the R4 substituent, its contribution Table 8. Sensitisation parameters for benzodiazepine substituents in the presence of sodium lauryl sulphate, obtained by both linear and semi-logarithmic adjustments.Numbers and letters in parentheses indicate the different series of data used for the calculation of SPs Lineur aiijiistment- SP, . .-0.41 -0.41 -0.19 -0.34 SPB . . 6.60 4.96 4.81 4.82 SPB . . 1.88 -3.50 0.47 0.14 SPC . .-3.18 -0.75 -1.86 - 1.01 SPD . . 3.85 3.33 3.68 3.33 SP(1) E,* Yo SP(2) E,* Yo SP(3) E.A Yo SP(T) E,* O/u SPc . .-3.72 14.5 -1.19 13.7 -2.28 8.7 -1.47 12.8 SP,, . .-0.44 - 1.20 - 1.60 -0.83 Semi-logarithmic adjustment- SPA . .-5.69 -5.53 -2.85 -4.69 SPB . . 1.62 -1.69 0.82 2.46 SPB. . . 1.02 -2.40 0.28 1.87 SPC . .-1.01 0.06 - 1.07 -0.93 SPD . . 0.35 2.71 1.13 -0.59 SP, ..-1.10 19.0 -0.00 18.1 -1.14 10.5 -1.01 16.1 SP, . . 0.18 2.17 0.71 - 1.09 E = average relative error. Given as the difference between MEF values calculated using SPs and the experimental values.Table 9. Sensitisation parameters for benzodiazepine substituents in the presence of different surfactants Surfactant Sensit isat ion parameter SDS TLS SLES DSS SP, . . . . . . -0.34 -0.25 -0.33 -0.46 SPB’ . . . . 0.14 0.09 0.36 2.61 SPC . . . . . . -1.01 -3.02 -1.25 -3.20 SPC . . . . -1.47 -3.27 - 1.64 -3.37 SPB . . . . . . 4.82 4.91 3.93 6.55 S P U . . . . . . 3.33 5.35 2.97 2.52 SPD’ . . . . -0.83 0.49 -0.73 -0.62 - 8oi sign inverts on replacing H by OH, except for studies carried out in the presence of triethanolamine lauryl sulphate. In this instance an OH group as the R4 substituent contributes positively to the sensitisation, although to a lesser extent than the I-I group.Mechanism of Interaction Between Benzodiazepines and Surfactants Because significant increases in the fluorescence of benzo- diazepines have only been obtained in the presence of anionic surfactants, this indicates that the interaction is electrostatic in character.22 Nevertheless, the different MEFs obtained for each surfactant structure and for each benzodiazepine studied emphasise the importance of steric factors in the process or’ micellar interaction . It was assumed that it is known that anionic surfactants affect the dissociation constants of organic molecules by stabilising the protonated forms.2’ Hence, in order to study the interaction mechanism in depth, the pK, of diazepam in the presence of SDS was determined spectrophotometrically. The value obtained (3.5 k 0.1) was almost coincident with that found in the absence of a surfactant (3.4 k 0.1) and it was deduced that the micelles have no influence on the dissocia- tion of benzodiazepines.The effect of anionic surfactants on the absorbance of benzodiazepine solutions or on the absorption maxima of benzodiazepines in H2S04 media was also studied. The results indicate that the benzodiazepines are scarcely modified in micellar media; only for medazepam was a 25% increase in the molar absorptivity at 453 nm found. This allows the increase in the fluorescence in micellar media to be interpreted on the basis of the increase in the fluorescence quantum yield of the benzodiazepines and the absence of a simultaneous increase in the molar absorption within the excitation bands.Therefore, the origin of the sensitisation process lies in the protection of the singlet excited state of the benzodiazepine in a micellar micro-environment from non-fluorescent de-activation processes. In order to determine the possible localisation of the benzodiazepine molecules within the micelles, the CMCs of several surfactants were determined in I-12S04 both in the presence of 4.8 p.p.m. of diazepam and in its absence. The CMCs were determined from surface tension measurements (see Fig. 1). Identical CMCs were found in both the presence and absence of diazepam for SDS, SLES and DSS. Hence, it was concluded that mixed micelles do not exist and consequently the benzodiazepine molecules can be expected to remain localised in the micellar surface zone.Conclusion The studies described in this paper provided data regarding the influence of the structure of both the surfactant and the sensitised molecule on the sensitisation process. It has been shown that the determination of SPs is a valid procedure for calculating the influence of each of the benzodiazepine substituents and each of the surfactant com- ponent parts on the sensitisation process. Hence, SPs might be used to facilitate the design of surfactant structures that allow a better sensitivity and might therefore contribute to a greater k n ow 1 e dge of mice 11 ar in t e r ac t i on processes . Finally, the results indicate that more sensitive fluorimetric methods for the determination of benzodiazepines can be performed using micellar media. I 1 3 0 2 20 J DSS, ‘/o m V Fig.1. tension (m) in the presence and (a) absence of diazepam Effect of DSS concentration in sulphuric acid on thc surface References 1. 2. Schutz, H., “Benzodiazepines,” Springer-Verlag, Heidelberg, 1982. Daudon, M., Pharm. B i d . , 1977, 11, 389.82 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Mehta, A. C.. Talanta, 1984, 31, 1. Tammilehto, S . , Farm. Aikuk., 1973, 82, 1. Besner. J . G . , Caille, G., Plaa, G . , and Dugal, R., Rev. Can. Biol., 1973, 32, 241. Noggle, F. T., and Clark, C. R . , J. Assoc. Off. Anal. Chem., 1079, 62. 799. Battista, H. J . , Udermann, H . , Henning, G . , and Vycudilik, W . , Beitr. Gerichtl. Med., 1979. 37. 5 . Vasiliades, J . , J . Toxiccd. Clin. Toxicol.. 1983, 20, 23. Schuetz, H., Ebcl. S ., and Fits. H . , Arzneim.-Forsch., 1985, 35. 1015. Braun, J . , Caille. G . , and Maritn, E . A., Can. J . Pharm. Sci.. 1968, 3, 65. Baranov, V. P., Shasaiton, S. S . , Udovenko, G. V., Pavlovat, T. Y., and Kholobes, T. A , . Otkrytiyu fzobref., 1985, 17. 152. Koechlin, B. A.. and D’Arconte. L . , Anal. Biochern., 1963,5, 195. De Silva, J . A. F., Munno, N . , and Strojny, N.. Anal. Chem., 1973, 45, 665. Valentour, J . C . , Monforte, J. R . , Lorenzo, B.. and Sunshine, I.. Clin. Chem. (Winston-Salem, N.C.), 1975. 21. 1976. Kawahara, Y.. and Yamazaki, Y., Sunkyo Kenkyusho Nempo, 1983. 35, 65. Troschuetz, J . , Arch. Pharm. (Weinheim, Ger.), 1981, 314, 204. Galdu, M. V., de la Guardia, M., and Braco, L., Anulyst, 19x7. 112, 1047. de la Guardia, M., Galdu. M. V., Rodilla. F., and Salvador, A , . paper presented at the XVIII Jornadas del Comite Espanol de la Detergencia (CEDIAID), Barcelona, March 1987. Hinze, W. L.. in Mittal, K. L., Editor, “Solution Chemistry of Surfactants,“ Volume 1, Plenum Press, New York. 1979, pp. Pelizetti, E.. and Pramauro, E . , A n d . Chim. Acta, 1985, 169. 79-127. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. ANALYST, JANUARY 1989, VOL. 114 Cline Love, L. J . , Harbarta, J . G . , and Dorsey, J . G . , Anal. Chem., 1984, 56, 1132A. Diaz, M. E., and Sanz-Medel, A , , Talantu, 1986, 33, 255. Hartley, G. S . , Trans. Faraday Soc., 1934. 30, 444. Sanz-Medel, A., Fernandez Perez, M. M., de la Guardia Cirugeda, M., and Carrion Dominguez, J . L.. Anal. Chem., 1986, 58, 2161. Carrion. J. L., and de la Guardia, M.. XXXI R e u n i h Bienal R.S.E. de Quimica, Santiago de C’ompostela, 1986, Communi- caci6n 17 031 0. Braco, A., Carricin. J . L.. and de la Guardia, M. , J . Mol. Sfrucf., 1986, 143, 489. Dubois, J. E., Laurent, D . , and Aranda, A . , J . Chim. Phys., 1973, 70, 1608. Dubois, J . E., Laurent, D.: and Aranda, A.. J . Chim. Phys.. 1973, 70, 1616. Dubois. J . E., and Chretien, J . , J . Chromutogr. Sci., 1974. 12. 81 1. Dubois. J . E., fsr. J . Chem., 1975, 14, 14. Dubois, J . E . , and Panaye. A.. Bull. Soc. Chim. Fr., 1975, 1390. Tronch, J. E., Master’s Thesis, University of Valencia. 1984. Berna, A., PhD Thesis, University of Valencia, 1984. de la Guardia, M., and Galdu, M. V., paper presented at the XVII Jornados del Comite Espanol de la Dctergencia ( E D / AID), Madrid, March 1986. de la Guardia, M., Tronch, J . E., Carrion, J . L., and Aucejo. A . , J . Mol. Struct., 1986, 143, 497. de la Guardia, M., and Rodilla, F., J . Mol. Struct., 1986. 143. 493. Puper 8IO12Ol F Received March 25th, 1988 Accepted August 24th, I988 1 .

ISSN:0003-2654    

DOI:10.1039/AN9891400077

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST. JANUARY 1989, VOL. 114 83 Applications of a Simultaneous Assay of Ascorbic Acid, Dehydroascorbic Acid and Ascorbic Sulphate in Biological Materials Konrad Dabrowski and Silvia Hinterleitner Institute of Zoology, University of Innsbruck, 6020 Innsbruck, Austria A modified spectrophotometric assay for ascorbic acid and its derivatives based on their reaction with 2,4-dinitrophenylhydrazine (DNPH) is described. Using standard ascorbic acid or ascorbic sulphate solutions, together with animal tissue or compound diet extracts, the conditions for ascorbic acid degradation were determined. For the differential measurement of reduced ascorbic acid (AA), dehydroascorbic acid (dAA) and ascorbic sulphate (AS), five series of simultaneous determinations were performed. These included the use of (1) KBr03 for the hydrolysis of AS, ( 2 ) 2,6-dichlorophenolindophenol as an oxidant, (3) DNPH to form a hydrazone derivative with dAA and (4 and 5) two blanks (where ascorbate was degraded) t o correct for interfering substances.A variety of vertebrate and invertebrate tissues were examined for their ascorbate content, and the advantages of the modified procedure over currently available assays are discussed. The results suggest that the Artemia cyst is a unique material in which ascorbic sulphate is present in large amounts whereas fish tissues do not contain this form of vitamin C. Keywords: Vitamin C; ascorbic sulphate; ascorbate degradation; 2,4-dinitrophen ylh ydrazine Two types of procedure are used most commonly for the determination of ascorbic acid (AA) and its derivatives: (1) measurement of the oxidation of AA by 2,6-dichlorophenol- indophenol (DCIP); and (2) measurement of the coloured species formed by the coupling of oxidised AA with 2,4- dinitrophenylhydrazine (DNPH).In the original work of Roe and Kuether' and in subsequent reviews,2-3 Roe pointed out that numerous substances could interfere in these assays. Recently, several high-performance liquid chromatographic methods have been developed for the determination of AA. In an investigation aimed at comparing methods for the determination of AA in animal tissues, Carr et al.4 concluded that the spectrophotometric method using DNPH results in values that tend to overestimate the AA content in animal tissues compared with techniques that employ liquid chromat- ography, which provide greater selectivity than the spectro- photometric procedures.Roe and Kuetherl studied the transmittance curves of several DNPH derivatives and suggested that monosugar - DNPH compounds absorbing in the 450 nm region might interfere with this method in which the absorbance of the DNPH derivatives of dehydroascorbic acid (dAA) and diketogulonic acid is determined at 524 nm.4 It is common practice to measure the reduced form of AA from the difference between the total ascorbate (DCIP being used to oxidise AA) and the remaining DNPH-reactive compounds, which include dAA, diketogulonic acid and interfering substances.4,' However, in this instance the amount of dAA which is biologically active cannot be determined.The other approach uses dithiothreitol to convert dAA to the reduced form of AA and yields the total ascorbate.6 The calculation of the ratio of dAA to AA is possible if the reductive specificity of dithiothreitol towards dAA is assumed. During preliminary assays of vitamin C in fish tissues we observed considerable variation among the individuals of an otherwise homogeneous group. In order to evaluate the reliability of the spectrophotometric DNPH technique, a series of internal ascorbic acid standards were prepared and the results obtained suggested that the proportion of inter- fering substances is large and that it varies considerably depending on the type of vertebrate tissue used. The first part of this study was designed to measure spectrophotometrically both dAA and AA using the DNPH method; this was accomplished by using specific conditions for ascorbate degradation. The significance of interfering sub- stances could therefore be assessed and attempts made to identify them.The second part of this study attempted to investigate the hydrolysis of ascorbic sulphate (AS) by treatment with KBr03 as described by Terada et ul.7 and to examine the conditions necessary for identifying the interferences caused by con- taminants present in the sample or formed during the hydro- lysis. A range of vertebrate and invertebrate tissues was used in this assay. Experimental Materials Whole fish or separate tissues from freshly killed roach (Rutilus rutilus) or rainbow trout (Salmo gairdneri) were used. In addition, the following invertebrates were used: an oligochaete, Tubifex tubifex; a freshwater shrimp, Palae- monetes antennarius; and brine shrimp, Artemia salina, cysts.Fish diets, composed of both animal and plant ingredients, were also used; their composition has been described else- where.8 Reagents Ascorbic acid, ascorbic acid 2-sulphate (dipotassium salt), 2,4-dinitrophenylhydrazine, 2,6-dichlorophenolindophenol and an amino acid standard mixture were obtained from Sigma (St. Louis, Missouri, USA). All other reagents used were of analytical-reagent grade. Procedure Pre-weighed material was homogenised in Ultra-turax or Potter homogenisers for 1G20 s or 2-3 min, respectively, in ice-cold 250 mM HC104 containing 5% trichloroacetic acid (TCA). The homogenates were centrifuged at 27 000 g for 30 min and the supernatant liquid was transferred into freshly prepared tubes.The supernatants were used either directly or after dilution with the deproteinisation solution in order to obtain a concentration of approximately 2 pg of AA in 0.25 ml. Ascorbic acid or AS standards in cold deproteinising buffer were prepared fresh daily. Ascorbic acid was assayed by a modification of the techniques of Carr et al.4 and Terada et al.7 The proposed method, which allows correction for the absorption of interfering substances, is described below and was used unless specified otherwise under Results.84 ANALYST, JANUARY 1989, VOL. 114 Series A To 250 pl of deproteinised sample were added 25 p1 of distilled water followed by 25 pl of 0.2% DCIP to oxidise AA to dAA (step I).After incubating for 1 h at room temperature, 250 pl of thiourea reagent (2% thiourea in 5% metaphosphoric acid) were added followed by an equal volume of 2% DNPH in 12 M H,S04 (step 11). Series B To 250 pl of the sample were added 25 pl of DCIP and the mixture was incubated at room temperature for 1 h. Then 25 pl of a 1% solution of KBr03 in distilled water were added and the mixture was incubated at room temperature for a further 1 h (step I). Step I1 was identical with that described under series A. Series C' To 250 pl of the sample were added 50 p1 of distilled water and the mixture was incubated for 2 h at room temperature (step I). Step I1 was identical with that described under series A. Series L> A 2.50-1-11 sample was incubated in a sealed tube for 2 h at 90 "C.Steps I and I1 were then carried out as described under series A. Series E A 250-pl sample was treated as described in step I of seriec B and was then incubated for 1 h at 90°C. Step I1 was identical with that described under series A. The reagent blank was prepared in exactly the same manner as described under series A, except that deproteinising buffer was used instead of the sample extract. No difference was found when the reagent blank was run through any of the series (A-E). All series were incubated for 3 h at 60 "C after which 0.5 ml of ice-cold 18 M H2S04 was added. The samples were transferred into Eppendorf tubes and centrifuged at 11 300 g for 3 min. Absorbance measurements were made with a Beckman DU-6 spectrophotometer; the spectra were re- corded with the 5ame instrument.Series D and E are designed to decompose AA and dAA in real samples and to account for the background absorbance, i.e., the absorbance given by interfering substances. The concentration of reduced AA was determined from the difference between the absorbance readings of series A and C, that of dAA from the difference between series C and D and that of AS from (B-E)-(A-D). Results Ascorbic Acid Degradation The formation of ascorbate by the acid hydrolysis of AS was accomplished by the method of Mead and Finamore.9 Hydrolysis with 1 M HC1 (final concentration) for 5 min recovered 91.2% of ascorbate from sulphate, whereas hydrolysis with the deproteinising solution (TCA - HC104) led to only a 62.3% recovery. The fate of ascorbate in acidic solution was therefore examined further. Ascorbic acid and AS standard solutions in deproteinising buffer (200 pl) were added to SO pl of 5 M HC1 and the mixture was incubated at various temperatures and for various periods of time [Fig.l ( a ) ] . Short pre-incubation times increased the recovery of AA at 60-80 "C, whereas considerable losses occurred at temperatures above 90 "C for heating times greater than 5 min. Preliminary assays showed that only 10% of AS was recovered as AA when incubated with DNPH at 60 "C for 3 h. This contradicts the findings of Baker et al.10 Hence, we 0 "C 70 "C 80 "C 90 "C 100 "C 100 50 100 8 s 5 50 5 g o o z W > U 50 U 50 100 100 Fig. 1. Effects of incubation time and temperature on the rate of decomposition of AA and on the rate of hydrolysis of AS, determined as the release of AA from (a) the pure solutions and ( b ) solutions containing AA and AS as internal standards in (F) fish and (D1 and D2) diet extracts.The number at the top of each bar is the incubation time in minutes l o o r - i a , I I O O ( b i $ 0 30 60 90 120 5 0 1 i x x x X 0 30 60 90 120 1 50t x .A 50 0 L 30 60 90 120 0 30 60 90 120 Timeimi n Fig. 2. Time courses ( a ) , (c) and (e) of the AA and ( b ) , (d) and cf, of the dAA decom ositions. ( a ) and ( b ) 0, AA standard solution; and X , fish extract. $) and (d) 0, AA standard solution; X I diet extract; and A ~ 9 mM amino acid standard solution. (e) and v) 0, Fish extract; and X , internal AA standard. Shaded areas signify the destroyed ascorbate performed the hydrolysis of AS under conditions identical with those used for AA degradation.It was found that at 80 "C the degradation of AA was minimal and the recovery of AA maximal. Subsequently, internal standards of AA and AS were added to fish tissue (F) or diet (D) extracts. The absorbance of the control samples, containing no standard, was subtracted from that of the experimental samples and the results are shown in Fig. l(b) as a percentage of the initial absorbance of the standard. For AS, the theoretical absorbance, assuming a 100% recovery of AA from AS (on an equimolar basis), was taken into account. This procedure assumes that AA released during sulphate hydrolysis is not destroyed. A considerable loss of AA from the fish tissue extract was observed when the hydrolysis time was more than 5 min (82.2 and 57.4% after 10 and 15 min, respectively, compared with a 93.8% recovery after 5 min).However, for two different fish diets, no consistency in the absorbance changes was observed as the hydrolysis time was increased [Fig. l(b), top]. These absorbance readings were not corrected for background interferences and the apparent recovery of 118% of AA after hydrolysis at 80 "C for 15 min might have been due to either (1) the presence of AS in the diet and/or (2) an increase in the absorbance due to the presence of interfering substances. TheANALYST, JANUARY 198'9, VOL. 114 85 rates of decomposition of A A and dAA were determined with standard solutions, fish tissues and diet extracts (Fig. 2). Ascorbic acid was decomposed completely after heating with TCA - HC104 for 2 h at 80 "C, whereas the absorbance of the tissue extract [Fig.2(a) and ( h ) ] and the diet extract [Fig. 2(c) and ( d ) ] levelled off after 2 h. The physiological concentration of free amino acids, which is characteristic for fish tissues (9 mM, according to Dabrowski"), inhibited the degradation of AA slightly [Fig. 2(c)]. Final confirmation that AA and dAA were decomposed selectively was achieved using the internal standard procedure [Fig. 2(e) and 01. In this instance the absorbance of the fish extract levelled off and the internal AA standard disappeared after incubation for 3 h. In further assays it was found that similar decomposition of AA could be obtained by increasing the temperature to 90 "C and shorten- ing the incubation time to 1-2 h.In a further modification of the procedure we abandoned the addition of HCl as suggested by Mead and Finamore" because their method led to a higher rate of AA destruction in extracts subjected to the sulphate determination procedure, a considerable increase in the background absorbance and an increase of 5-15% in the absorbance of non-hydrolysed extracts compared with that of extracts in deproteinising solution alone. Ascorbic Sulphate Hydrolysis It was evident that hydrolysis of AS with KBr03 led to a considerable increase in the absorbance for some tissues and diet extracts. The results of AA degradation and the products formed during AS hydrolysis were identified from the absorption spectra of the DNPH derivatives.The absorption spectra of AA and of the products from the hydrolysis of AS are shown in Fig. 3 ( a ) . The maximum absorption for all three compounds occurs at 524 nm. Although the concentrations of the AA and AS hydrolysates were different, both methods, t.iz.. TCA - HCl and KBr03, gave a similar recovery of AA (8&95"/0) from pure standards. However, for animal tissue extracts. KBr03 proved to be a far more powerful hydrolysing agent. was less destructive towards AA and did not require heating. An experiment to assess the effect of various compounds on the formation of DNPH complexes was performed. Ascorbic acid standards were supplemented with glucose or glycogen (final concentration 0.5%) and subjected to KBr03 hydrolysis as described under Experimental.Fig. 3(b) shows that both glucose and glycogen can interfere if the absorbance of the DNPH derivatives is measured at 524 nm. An amino acid mixture (final concentration 9 mM) also increased the absor- bance by 1&15%. Examples of AA, dAA and AS Determination As noted previously for standard solutions, the hydrolysis of fish tissue extracts with KBr03 resulted in an increase in absorbance at 524 nm [Fig. 3 ( c ) ] . This was mainly due to compounds that have their maximum absorption in the range 4.10-460 nm, corresponding to the DNPH glucose derivative [see Fig. 3(b)]. Further, the use of DCIP [Fig. 3 ( c ) , curve A] gave rise to compounds whose maximum absorption was at 540 ntn, compared with 524 nm in the direct determination of dAA [Fig. 3 ( c ) , curve C] when DCIP was omitted.For fish tissues monitored at 524 nm, the results from the degradation of AA and dAA [Fig. 3(c), curve D] suggested that 35-35% of the total absorbance was due to interfering substances. This value corresponds to that obtained previ- ously (Fig. 2) in which the background level remained constant after prolonged incubation. Fish liver tissue contains high levels of glycogen (10% wet mass in freshwater fish) and this compound is likely to interfere in the determination of vitamin C. Tucker and HalverlZ suggested that most of the AS in trout is stored in the liver. Hence an experiment was performed to investigate this 0.25 0.20 0.15 0.10 0.05 al 6 0 n $ 0.2 II Q 0.1 0 ( a ) 418 470 522 574 1.25 1 .oo 0.75 0.50 0.25 0 440 480 520 560 600 400 440 480 520 560 600 418 470 522 574 Wavelengthinm Fig.3. ( a ) Effect of hydrolysis conditions on the absorption spectrum of the hydrolysis products. A. AA standard: B, AS hydrolysed with KBrO,; and C, AS hydrolyscd with TCA - HC1O3 buffer. The reaction mixturc is described in the text. ( h ) Effect of (A) glucose and (Bj glycogen on the absorption spectrum of the products formed after AS hydrolysis. The remaining spectra were obtained (C) after hydrolysis of pure AS solution and (D) from non-hydrolysed AS. (cj Absorbancc spectra of DNPH derivatives formed during prepara- tion of the fish extract for the determination o f ascorbate. Curves A , B, C and D refer t o the procedures described under Experimental. ( d ) Absorbance spectra of DNPH derivatives formed during, the deter- mination of ascorbate in trout liver extract.Curves A , B and D refer to the procedures described under Experimental and C to hydrolysi5 in 1 M HCI problem and to verify the application of the method. The absorbance at 524 nm was increased strongly by both methods of hydrolysis [Fig. 3 ( d ) ] , but this was exclusively due to the absorbance of interfering substances. Two other curves representing the background absorbancc are not shown in Fig. 3 ( d ) , but the subtraction of their absorbance values, reflecting AA decomposition, resulted in values identical with those for total (AA + dAA) ascorbate. The results of the application of the proposed assay to the simultaneous determination of AA, dAA and AS are shown in Fig. 4 ( a ) .At 524 nm the difference between absorbance curves A and B, and D and E is exactly the same, indicating that AS is not present in fish body extracts. The results stress further that calculation of the ratio of dAA to AA (difference between C and D, and A and C, respectively) is an important part of the determination of vitamin C in fish tissues. Brine shrimp cysts, which are known for their high AS content, were also subjected to the method [Fig. 4(b)]. Hydrolysis with HCI (curve C) resulted in a much lower recovery of AA than hydrolysis with KBr03 (curve B). No AA was detected in Arternia cysts. The content of AS in dry cysts was calculated to be 594 pg g-1, equivalent to 284 pg g-1 of AA. Tubifex tubifex is an oligochaete containing large amounts of glycogen.13 Hydrolysis of an extract of this animal with KBr03 resulted in a considerable increase in absorbance, probably due to glycogen [Fig.4(c)]. No store of AS could be detected in this animal. The total AA content in Tuhijex tubifex was found to be 21.1 pg g-1 wet mass, 65.7% being in the reduced form. No AS was found in adult shrimp and the interference from other substances was negligible [Fig. 4(c) and ( d ) ] . The total86 ANALYST, JANUARY 1989, VOL. 114 0 0 a c co e 0, 1.50 n 6 0.75 ( 440 480 520 560 I I 400 440 480 520 560 600 440 480 520 560 600 Wavelengthinm Fig. 4. ( a ) Absorbance spectra of DKPH derivatives formed during the determination of ascorbate in whole fish extract. Curves A-E refer to the procedures described under Experimental. A , Total AA; B, AS; C, dAA; and D and E , A A destroyed.( b ) Absorbance spectra of DNPH derivatives formed during the determination of ascorbate in Artenziu salina cysts. A , Refers to the procedure described under Experimental; B, extract hydrolysed in KBrO,; C , hydrolysed with TCA - HCIO, - HCI; and D. kept in TCA - HCIO, - HCI buffer at room temperature. (c) Absorbance spectra o f DNPH derivatives formed during the determination of ascorbate in Tubifex tuhifex. Curves A , B. C and D refer to the procedures described under Experimental. (d) Absorbance spectra of DNPH derivatives formed during the determination of ascorbate in freshwater shrimp. Curves A, B and D refer to the procedures described under Experimental ascorbate content was found to be 52 pg g-1 (wet mass), although 96% was in the oxidised form (dAA) [not shown in Fig.4(d)]. We also recommend, based on several assays with animal tissue, that in order to calculate the proportion of AA and dAA, the deproteinising liquid should be supplemented with 0.08% EDTA (final concentration). Further fresh tissue should preferably be used and if it is necessary to store the samples then the deproteinised extracts should be frozen rather than the animal tissue. The inclusion of internal standards (dAA and AA) in the TCA-based extraction solution supplemented with EDTA demonstrated that the recoveries of AA and dAA were 93.3-98.1 and 93.6-100.2%, respectively. The method allows the determination of 0.1 pg of ascorbate in 250 p1 of extract. Discussion Although the developed method is not rapid, it is simple and allows the accurate determination of AA in animal tissues and other biological materials.The background correction method recommended combines two separate series of sample blanks in which AA, dAA and, possibly, diketoguionic acid are degraded to give products that do not form hydrazones with DNPH. Most of the assays used previously either measure total ascorbate and do not correct for background or measure solely the reduced form, neglecting the biologically active form, dAA. Consequently, this results in a distortion of the true vitamin C content. Lau et al. 14 introduced the background correction method for the determination of AA in plant material, which is based on the catalytic oxidation of AA by copper. Wunderling et al. 6 emphasised the importance of the contaminants present in the sample, although it appears that blood is free from major interferents in the ultraviolet region.Carr et al.4 concluded that the AA values obtained for various animal tissues vary considerably depending on the method used. These workers determined only the reduced form of ascorbate and so their method should not suffer from the problem of interferences. However, in some invertebrates the concentration of AA determined with DNPH differed by -32 to +252% compared with a liquid chromatographic method. The determination of AA or AA + dAA in fishl"I6 and mammalian17 tissues has been reported but no correction was applied to account for background interferences and hence these results are subject to a certain degree of error depending on the type of tissue analysed, although they might reflect relative changes in the content of ascorbate in animal tissue introduced by various dietary treatments. Cowey et al.18 used an AA oxidase assay to measure AA in developing salmon eggs. As this method measures only reduced ascorbate, this may have led to the variation in the results, suggesting an increase in AA levels in endogenously feeding embryos. This may have been due to the omission of dAA from the analytical procedure. Differences in the methods applied and in their specificity were probably responsible for the one order of magnitude difference in the AA content found by Cowey et a1.18 and Sato et al.19 in rainbow trout embryos. Thomas and co-workers5.20 have measured AA levels in fish tissues.However, as dAA is readily reduced back to AA in fish,21 the significance of the information from these experi- ments concerning the effects of stress and intoxication on fish cannot be evaluated. Some reports have indicated a relatively high level of AS in both mammals (12.3 pg g-1 in rat liver22) and fish (198 pg g-1 in trout liver12 and 7.1 1 pg g-1 in mullet brains). On the other hand, Terada et al.7 were unable to detect any traces of AS in rat tissues 48 h after an intravenous AS injection. Our modified procedure, described in this paper, failed to detect AS in roach, common carp8 or trout liver. Tsujimura and ~o-workers23~24 demonstrated that when AS was given orally, the AA and AS levels in the liver increased in some fish species but not in others.This might suggest a difference in the utilisation, metabolism and storage of AS between fish species, but no details of the method used for the determina- tion of AS were given. Shapiro and Poon2i suggested that there should be a re-evaluation of these experiments in which AS was used, owing to decomposition of this substance during storage. Hence the antiscorbutic importance of AS in fish26 may have to be questioned. The results of the present investigation contradict previous suggestions of a high AS content in rainbow trout tissues12 and the importance of AS as a major form of storage of vitamin C in fish. 12.26 However, our results confirm the high AS contents found in resting, dehydrated eggs of brine shrimp."z7 Golub and Finamore28 found 2-3 VM AS per gram of protein in Artemia cysts, which is equivalent to approximately 460 pg g-1 of dry matter.This is a very similar value to that found in the present work. Golub and Fina- more28 also did not detect AA in Artemia cysts. Conclusion Background correction has been shown to be an effective and indispensable modification in the determination of ascorbate in various biological materials. The absorbance due to interfering substances, probably sugars, but also other com- pounds, represents a significant (30-50%) proportion of the total absorbance measured at 524 nm after coupling withANALYST. JANUARY 1989, VOL. 114 DNPH. The decomposition of ascorbate in real samples can be carried out according to the procedure described here, although it is advisable to run an extra assay to confirm these conditions when new materials are used.With the proposed modifications, the DNPH method is selective and can be used to check both the UV spectro- photometric and HPLC methods for the determination of AA and its derivatives. This work was supported by project No. S-35/04 of the “Fonds zur Forderung der Wissenschaftlichen Forschung in Ostereich.” We thank Dr. Reinhard Lackner for his assistance and suggestions throughout this study and Professor W. Wieser for reading the manuscript. 1. 7 . 3. 4. 5 . 6 . 7. 8. 9. 10. References Roe, J. H . , and Kuether, C. A , , J . Biol. Chem., 1943,147,399. Roe, J . H.. in “Methods of Biochemical Analysis,” Inter- science. New York, 1954. Volume 1. pp. 115-139. Roe, J . H . , in Gyogry, P ., and Pearson. W. N., Editors, “The Vitamins,” Academic Press, New York, 1967, Volume 7, Carr, R. S.. Bally, M. B., Thomas, P.. and Neff, J . M., Anal. C‘hern., 1983. 55, 1229. Thomas, P . , Bally, M., and Neff, J . M . , J . Fish Biol., 1982, 20, 183. Wunderling, M., Paul, H. H., and Lohmann, W., Bid. Chern. Hoppe-Seyfer, 1986. 367. 1047. Terada. M.. Watanabe, Y . , Kunitomo, M., and Hayashi, E., Arzal. Biochem., 1978, 84, 604. Dabrowski, K., Hinterleitner, S . , Sturtnbauer, C.- El-Fiky, N . , and Wieser. W.. Aquaculture, 1988, 72. 295. Mead. C. G.. andFinamore, F. J . , Biochemistry, 1969,S. 2652. Baker, E. M., Hammer. D. C . , Kcnnedp, J . E., and Tolbert, B. M.. Anal. Biochcm., 1973, 55, 641. pp. 27-49. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 87 Dabrowski, K., Comp. Biochem. Physiol., 1982, 72A, 753. Tucker, B. W., and Halver, J. E., J. Nutr., 1984, 114, 991. Dabrowski, K., in Tiews, K., and Halver, J . E., Editorr, “Finfish Nutrition and Fishfeed Technology,” Heinemann Verlag, Berlin, Volume 2, pp. 519-529. Lau, 0.-W., Luk, S.-F., and Wong, K.-S., Analyst, 1986, 111, 665. Hilton, J . W., Cho, C. Y.. and Slinger, S. J., J. Fish. Res. Board Can., 1977, 34, 2207. Soliman, A. K., Jauncey, K., and Roberts, R. J . , Aquaculture, 1986, 52, 1. Tsao, C. S . , Leung. P. Y., and Young, M., J . Nutr., 1986,117, 291. Cowey, C. B., Bell, J . G., Knox, D., Fraser, A , and Youngson, A . , Lipids, 1985, 20, 567. Sato, M., Yoshinaka, R., Kuroshima, R., Morimoto, H., and Ikeda, S . , Buff. Jpn. Soc. Sci. FiJh.. 1987, 53, 795. Thomas, P., J. Fish Biof., 1984, 25, 711. Yamamoto, Y., Sato, M., and Ikeda, S., Bull. Jpn. Soc. Scr. Fish., 1977, 43, 59. Mumma, R. O., and Verlangieri, A. J . , Riochim. Biophys. Acra, 1972, 273, 249. Tsujimura, M., Fukuda, T., Kasai, T., and Kitamura, S., Buff. Jpn. Soc. Sci. Fish., 1981, 47, 435. Tsujimura, M., Fukuda, T., and Kitamura, S., Buff. Jpn. Soc. Sci. Fish., 1982, 48, 1823. Shapiro, S. S . , and Poon, J . P., Riochim. Riophys. Acta, 1975, 385, 221. Halver, J. E., Smith, R. R . , Tolbert, B. M., and Baker, E. M., Ann. N . Y . Acad. Sci., 1975, 258, 81. Bond, A . D . , McClelland, B. W., Einstein, J . K., and Finamore, F. J . , Arch. Biochem. Biophys., 1972, 153, 207. Golub, A. L., and Finamore, F. J., Fed. Proc. Fed. Am. Soc. Exp. R i d . , 1972, 31, 706. Paper 81024770 Received June 22nd, 1988 Accepted August 26th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400083

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 113 89 Stopped-flow Determination of Iodide in Pharmaceutical and Food Samples M. Carmen Gutierrez, Agustina Gomez-Hens and Dolores Perez-Bendito Department of Analytical Chemistry, Faculty of Sciences, University of Cordoba, 74004 Cdrdoba, Spain The iodide-catalysed reaction between cerium(lV) and arsenic(ll1) has been studied using a modular stopped-flow system. The features of both the stopped-flow and conventional kinetic methods have been compared by using the same instrument. The stopped-flow method has a wider linear range and is faster than the conventional kinetic method and has been applied satisfactorily to the photometric determination of iodide in pharmaceutical preparations, table salt and cow's milk. The results obtained show that the stopped-flow method is simple, inexpensive and rapid and requires no sophisticated equipment.In addition, its high sampling rate makes it particularly suitable for the routine determination of iodide. Keywords: Iodide determination; stopped-flow; photometry; pharmaceutical and food samples The most widely used method for the determination of iodide is probably that based on its catalytic effect on the CeIV - As"[ system, known as the Sandell - Kolthoff reaction. 1.2 Although first reported many years ago, this method is frequently used in a large number of routine laboratories and is regarded as the prototype of kinetic determinations. In addition to its numer- ous applications,3 detailed kinetic studies of this reaction have been performed. Hence, Rodriguez and Pardue4 proposed a mechanism that accounted for their experimental findings. The kinetic data obtained for the Sandell - Kolthoff reaction are usually obtained by the conventional kinetic approach; however, the stopped-flow technique has not yet been used for this purpose.One possible reason for this is that, owing to the high cost of commercial stopped-flow instruments, the use of this technique has been limited and aimed mainly at the study of fast reactions that cannot be monitored by the conventional technique. However, the stopped-flow technique can also be used in conjunction with methods involving less rapid reac- tions. particularly with a versatile mixing module, as described in this work. and. with the advent of microcomputer-con- trolled stopped-flow systems, as a means of accomplishing automation and rapid handling of reagents for routine analysis.This stopped-flow modules can readily be coupled to any spectrophotometer or spectrofluorimeter and is much less expensive than commercial stopped-flow instruments. The usefulness of this module has been demonstrated in the determination of copper in serum0 and theophylline in pharmaceutical preparations,7 and in the resolution of per- phenazine - chlorpromazines and spermine - spermidine') mixtures. The aim of this work was to demonstrate the suitability of this modular stopped-flow system for the determination of iodide in various samples and for the automation of the method for routine analysis. The method was applied to the determination of iodide in pharmaceutical, table salt and milk samples.Experimental Apparatus A Perkin-Elmer Model Lambda 5 spectrophotometer fitted with a stopped-flow modules supplied by Quimi-Sur Instrumentation was used for kinetic measurements. The module. with an observation cell of 0.3-cm path length, was controlled by the associated electronics and it Hewlett-Pack- ard 98630 A computer. The spectrophotometer cell compart- ment was thermostated by a Peltier electronic system and the solutions in the stopped-flow module were kept at a constant temperature by means of a circulating water-bath. A Heron HD-150 muffle furnace and an ultrasonic bath (Bandelin Sonorex TK 52, 60 kHz, 80 W) were also used. Reagents All chemicals were of analytical-reagent grade. Doubly distilled water was used in all experiments.Standurd iodide solution, 1 g 1-1. Prepared by dissolving 0.1308 g of potassium iodide (Merck) (dried at 105 "C for 2 h) in doubly distilled water and diluting to 100 ml in a calibrated flask. Working solutions were prepared daily by cuitable dilution with doubly distilled water. Cerinm(ZV) solution, 1.6 X 10-2 M. Prepared by dissolving 0.6468 g of Ce(S04)2.4H20 (Merck) in 100 ml of2 M sulphuric acid. Arsenic(ZZZ) d u t i o n , 4 x 10-2 M. Prepared by dissolving 0.5196 g of NaAsOz (Merck) in 100 ml of doubly distilled water. Procedure One of the two 10-ml reservoir syringes was filled with a previously prepared solution containing 2.5 ml of the 1.6 x M cerium(1V) solution and 3 ml of 8 M nitric acid in a final volume of 10 ml. The other syringe was filled with a solution containing 2 ml of 8 M nitric acid, 0.6 ml of the 4 x 10-2 M arsenic(II1) solution and an iodide standard or sample solution in a final concentration range of between 2 and 100 pg ml-1 and a final volume of 10 ml.After the two 2-ml drive syringes had been filled, 0.15 ml of each solution was mixed in the mixing chamber in each run. The decrease in the absorbance during the reaction was monitored at 365 nm and displayed on a chart recorder operating at 20 mm niin-1. All measurements were carried out at 40 "C. The initial rate method was applied to the absorbance values, collected every 1 .S s, and these were processed by linear regression using the microcomputer. The reaction rate could be determined i n about 15 and each sample was assayed in triplicate.The blank signal was found to be negligible. Determination of Iodide in Authentic Samples Phurniaceutical samples No sample pre-treatment was needed for these analyses apart from an appropriate dilution of the sample to obtain a concentration level that has within the linear working range of the calibration graph. Diluted samples were treated as described under Procedure.90 ANALYST, JANUARY 1989, VOL. 114 Table 1. Interferences in the determination o f 40 ng ml I of iodide & 0 1.0 2.0 3.0 4.0 -20 30 40 50 60 4- [HN03]/M TPC !! - .- . ; I T I - 0.75 0.75 0.50 F-2-\ 1 1:; ky] 0.25 0 1.0 3.0 5.0 0 1.0 2.0 3.0 4.0 [CelV]/10-3 M [As~~l]/lO-3 M Fig. 1. Effect of reaction variables: (a) nitric acid concentration; ( b ) temperature; (c) cerium(1V) concentration; and ( d ) arsenite concen- tration Table sult sumples About 10-15 g of table salt were weighed accurately, dissolved in water in a 100-ml calibrated flask and diluted to volume with doubly distilled water.Suitable aliquots of this solution were analysed as described above, but, to compensate for the possible contribution of chloride ions to the analytical signal, 2 ml of 2 M hydrochloric acid were added to the standard and sample iodide solutions together with the arsenite solution and the nitric acid. In this instance, the blank signal had to be subtracted from the reaction rate values obtained for the samples. Milk samples The samples required alkaline ashinglo-11 prior to the final determination of iodide. This treatment was performed in a Pyrex tube (16 x 25 mm) as follows.Approximately 1 g of milk was accurately weighed and mixed with 1 ml of 4 M potassium hydroxide solution. After drying at 105 "C for 20 h the sample was heated at 150 "C for 30 min and at 600 "C for a further 1 h. The ash was dissolved in 10 ml of boiling water by means of an ultrasonic bath and the solution was centrifuged at 2500 g for 15 min. A 2-ml aliquot of the supernatant liquid was used for the determination of iodide as described under Procedure. Results and Discussion Effect of Reaction Variables This study was carried out by altering each variable in turn while keeping the others constant. The optimum reaction conditions chosen were those which yielded the minimum relative standard deviation for the initial rate measurements and which resulted in a reaction order in the variable concerned of zero or near to zero.All concentrations described here are the initial concentrations in the syringes (i.e.. these values are twice the actual concentrations in the reaction mixtures at time zero after mixing). Each kinetic result was the average of three determinations. As reported elsewhere4 if iodide and CeIV are mixed in the absence of A P , the catalytic activity of iodide is negated. For this reason, it was necessary to place the Ce'V solution alone in the first syringe and the mixture of As111 and iodide in the second. Because the Ce1V - As1II reaction only occurs in acidic media, three different acids were assayed (HN03, HCl and H2S04) at a final concentration of 2 M. Nitric acid was chosen because it provided better reaction rate values than the other Tolerance ratio.interferent species : iodide Species assayed 100: 1 . . . . B4073--, CI-, CIO1-, CN-, F-, COI-, NO>-. SOj2- ~ P04j-, P20,.'-. Si03'-, Ca2+, acetate, choline, glycine. saccharin, tartrate, theophylline. caffeine 50: 1 . . . . SCN-, citratc two acids. As shown in Fig. ~ ( L z ) , the reaction rate increased with the nitric acid concentration up to a concentration of 2 M , above which it remained constant, at least up to 4 M. The reaction rate increased with increasing temperature in the range 2&60 "C [Fig. l ( b ) ] . A temperature of 40 "C was selected. A previous study4 of the effect of the Ce'V and As"' concentrations on the reaction rate has shown that the rate is strongly dependent on the molar ratio of CeIV to As"'.The CeIV concentration was varied in the range 5 x 10-4-6.4 X 10-3 M in the presence of 2.4 x 10-3 M As"'. The variation in the reaction rate over this concentration range is shown in Fig. l(c). As can be seen, the reaction order in CeIv was close to zero for concentrations in the range 2 x 1W3-4.5 X 10-3 M. A 4 x 10-1 M CelV concentration was chosen because it resulted in the best linear range for the initial reaction rate. The effect of the As"' concentration was studied in the range 5 X 10-"4 X 10-3 M while the CeIV concentration was kept constant at 4 x 10-3 M. Under these conditions, the reaction rate was not affected by changes in the As111 concentration above 2 X 10-' M [Fig. l(d)]. The initial slopes of the absorbance - time curves obtained under the optimum experimental conditions indicated a first-order reaction with respect to iodide.As one of the determinations to be performed by the stopped-flow technique was that of iodide in table salt, and as it has been reported*.12.l3 that chloride ions affect the dependence of the reaction rate on the iodide concentration, we studied the effect of these ions. Sodium chloride could not be used owing to the iodide impurities present so the study was carried out with hydrochloric acid. Rodriguez and Pardue4 reported that the only effect of chloride ions was to give a plot of the reaction rate as a function of the iodide concentration, which was linear and parallel to that obtained in the akence of chloride and which had a positive intercept.Although these results correspond to 0.125 M hydrochloric acid, we obtained the same results over the concentration range 0.1-0.6 M HCI. Features of the Analytical Method Under the optimum experimental conditions described above. the initial slopes of the absorbance - time curves were linearly proportional to the iodide concentration in the range 2-100 ng ml-1. The absorbance values were taken at a rate of 40 data points per minute and the reaction rate was calculated by linear regression of the first ten values. The slope of the calibration graph was 1.12 X 10-2 min-1 ng-1 ml with a Pearson's correlation coefficient (Y) of 0.998. The detection limit, calculated according to the IUPAC recommendations,l-l was 1.0 ng ml-1 of iodide.The precision (RSD) of the method 0, = 0.05, n = 11) for 40 ng ml-1 of iodide was -t2.64%. The selectivity of the method was determined by adding different amounts of potentially interfering species, i. e.. those species that might be expected to occur with iodide in the samples assayed. As shown in Table 1 most of the species assayed are tolerated at a ratio of foreign species to iodide of atANALYST. JANUARY 1989, VOL. 114 91 Table 2. Determination of iodide in pharmaceutical preparations Table 4. Determination of iodide in milk Iodide content*/ Iodide mg ml- 1 Added1 Found+! Recovery. Sample Stated Found? ng ml-1 ng ml- 1 Y" 1< 8.66 8.64 10.0 14.0 20.0 23 42.0 43.0 10.0 14.0 20.0 311 10.0 10.3 10.0 14.0 20.0 4 I/ 10.0 10.5 10.0 14.0 20.0 5** 14.0 14.1 10.0 14.0 20.0 6 t t 40.0 42.0 10.0 14.0 20.0 9.5 13.1 20.0 9.5 14.5 20.5 9.9 14.2 19.9 10.5 14.5 20.2 9.7 13.H 20.3 10.5 14.5 20.0 95.0 96.4 100.0 95.0 103.5 102.5 99.0 101.4 99.5 105 .o 103.5 101 .o 97.0 98.5 101.5 105.0 1 02.0 1 00.0 * Expressed as potassium iodide.+ Mean of three determinations. f Elixifilin (Morrith): theophylline (0.533 g). potassium iodide (0.k66 g); excipient up to 100 ml. 3 Colircusi-yodo-tio-calcico (Cusi): sodium iodide (20 mg), potas- sium iodide (20 mg). calcium chloride (13.4 mg), sodium thiosulphate (20 mg); excipient up to 1 ml. 7 Lasa antiasmiitico (Lasa): aminophylline (10 mg), potassium iodide (10 mg). sodium saccharin (3 mg); excipient up to 1 ml. / Angiofiline (Seid): tonzylamine hydrochloride (0.50 g), sodium theophylline glycinate (1.20 g), ephedrine (0.08 g), potassium iodide (1 g).menthol (0.02 g); excipient up to 10c) ml. * * Navarroiodol (Navarro J.): potassium iodide (1.40 g), sodium thiosulphate (0.05 g); excipient up to 100 ml. +t Coliri ocul resolutivo (Frumtost Zyma): choline hydrochloride (10 mg). potassium iodide (40 mg); excipient up to 1 ml. Table 3. Determination of iodide in table salt Iodide Iodide content"/ Added1 Found*/ Recovery, Sample pgg-1 ngml-' ngml I Yo Iodised table salt . . 62.5 20.0 19.5 97.5 40.0 41.0 102.5 Iodisedtablesalt . . 100.0 20.0 19.5 97.5 40.0 39.5 98.7 Common salt . . . . 0.3 20.0 19.5 97.5 40.0 40.0 100.0 Mean of three determinations. least 100 : 1. Only citrate and thiocyanate are tolerated at a lower ratio. In order to compare the features of the stopped-flow method with those of the conventional kinetic method we checked the latter with our own instrument. The calibration graph obtained was linear in the range 2-10 ng ml-1 of iodide.This shows that the stopped-flow approach results in a wider linear range because ~ although the lower concentration level of the calibration graph is the same for both kinetic methodg, the upper concentration level is higher for the stopped-flow method. The detection limit of the conventional kinetic method was 0.7 ng ml-1 of iodide and the slope of the calibration graph was 1.60 x 10-2 min-1 ng-1 ml, with a Pearson's correlation coefficient ( Y ) of 0.995. It should be noted that the concentrations of iodide given in the stopped- flow method, as stated above, correspond to the initial concentrations in the syringe and that the actual concentration in the reaction mixture is half that used in the conventional Iodide Iodide content*/ Added1 Found*/ Recovery, Sample ngg-1 ngg-1 ngg-1 Yo Sterilisedmilk ., 142.5 100.0 93.7 93.7 Skimmedmilk . . 138.7 100.0 99.9 99.9 Freshmilk . . . . 142.5 100.0 102.5 102.5 * Mean of three determinations. kinetic method. Also, the observation cell of the stopped-flow module has a light-path of only 0.3 cm, whereas that used in the conventional method has a light-path of 1 cm. The precision was very similar for both methods. On the other hand, the reaction rate derived from each kinetic curve in the stopped-flow method was obtained in only 15 s, whereas in the conventional method about 3 min were required.Hence, it can be concluded that the stopped-flow method yields a calibration graph with a wider linear range, is faster and consumes smaller amounts of reagents than the conventional kinetic method. These features show that the stopped-flow method is particularly useful for the routine determination of iodide. Applications To demonstrate the applicability of the proposed stopped- flow method to the determination of iodide, the method was applied to the analysis of various samples, viz.. pharmaceut- ical preparations, table salt and cow's milk. The iodide-containing pharmaceutical preparations (antiasthmatics, antiseptics and eye drops) analysed required no pre-treatment apart from an appropriate dilution of the samples. The results obtained for these determinations, in addition to the data from the recovery study, are summarised in Table 2.Recovery data were obtained by adding different amounts of iodide standard to the samples and subtracting the results obtained for samples prepared in a similar way but with no iodide added. No sample pre-treatment was required for the determina- tion of iodide in table salt; however, in order to compensate for the effect of the chloride ions present in the samples, a suitable amount of hydrochloric acid was added to these and to the standards. The only effect of this modification was that the calibration graph obtained had a non-zero intercept and a slope that was parallel to that obtained in the absence of chloride. Table 3 shows the results obtained for the determina- tion of iodide in three table salt samples, two of which were iodised, and for the recovery study.The corresponding results obtained for various cow's milk samples are summarised in Table 4. In this instance, an initial alkaline ashing10.l1 step was required. The recovery values were obtained by adding 100 ng of iodide to 1 g of sample. From the results obtained for the determination of iodide in these samples it can be inferred that the proposed stopped- flow method is very simple, inexpensive and fast, and is, therefore, suitable for routine analysis. The authors are grateful to- the CAICYT (Project No. 0979184) for financial support. References 1. Sandell, E. R . , and Kolthoff, I. M.. I . Am. Chem. Soc., 1934, 56, 1426. 2. Sandell, E. B.. and Kolthoff. I. M., Mikrochim. Acta. 1937.9. 3. Pkrez-Bendito, D., and Silva, M., "Kinetic Methods in Analytical Chemistry," Ellis Horwood, Chichester, 19HX. pp, 62-64.92 ANALYSl’, JANUARY 1989, VOL. 114 3 . 5 . 6. 7. S . 9. Rodriguez, P. A . , and Pardue. H. L.. Anal. Chrrn., 1969, 41, 1369. Loriguillo, A.. Silva, M., and Perez-Bendito. D., Anal. Chim. Acta, 1987, 199, 29. Gutierrez, M. C . , GoInez-Hens, A , , and Perez-Bendito, D., Frrsenius 2. Anal. Chrm., 1987, 328. 120. Gutikrrez, M. C., Gomez-Hens, A . . and Perez-Bendito, D., Analyst. 1988. 113, 559. Gutierrez, M. C.. Gomez-Hens, A . , and Perez-Bendito, D.. Anal. Lett.. 1987. 20, 1847. Gutikrrcz. M. C.. Giimez-Hens, A . , and Pirez-Bendito, D.. Fresenius 2. Anal. Chern., 1988, 331. 642. 10. 11. 12. 13. 13. Belling, G. B., Analyst. 1983, 108, 763. Aumont, G.. and Tressol, J.-C., Analyst. 1986, 111. 831. Lein, A.. and Schwartz, N . , Anal. Chem.. 1951, 23, 1507, Deman, J . , Mikrochim. Acta, 1964, 67. Long, G. L., and Wincfordner, J . D., Anal. Chem.. 1983, 55, 712A. Paper 8101 858H Received Muy 11th) 19861 Accepted June 28th) 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400089

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 93 Simultaneous Second-order Analogue Derivative Spectrophotometric Determination of Vanadium and Cobalt in Steels Ana 1. Jimenez, Francisco Jimenez and Juan J. Arias Department of Analytical Chemistry, University of l a Laguna, E-38204 La Laguna, Tenerife, Spain A method is proposed for the simultaneous determination of vanadium and cobalt by second-derivative spectrophotometry based on their reactions with 4-(1’H-I ’,2’,4’-triazol-3’-ylazo)-2-methylresorcinol. The method allows the determination of 0.10-1.02 p.p.m. of vanadium and 0.12-1.18 p.p.m. of cobalt in mixtures for which 10 < Cv/Cc, s 0.2 and it has been applied to their determination in alloy steels. Keywords: Vanadium; cobalt; simultaneous determination; derivative spectrophotometry; 44 I ’H- 1 ’,2’,4 I- triazol-3’- ylazo)-2-meth ylresorcinol Many heteroazo compounds have been synthesised and utilised as analytical reagents for the determination of different metal ions1.2 although their highly sensitive reactions generally show a lack of selectivity.However, both sensitivity and selectivity can be improved by using second-order analogue derivative spectrophotometry.3.4 The importance of the analytical chemistry of vanadium and cobalt in modern science is manifested by the large number of studies reported in the literature536 and the sequential determination of these two ions may prove to be relatively easy compared with their simultaneous determination. In this work, the application of derivative spectropho- tometry to the simultaneous determination of vanadium and cobalt using 4-( 1‘H-1’ ,2’ ,4’-triazo1-3’-ylazo)-2-methylresorci- no1 (TrAMR) as a chromogenic reagent is described.Experimental Apparatus Perkin-Elmer 550s and 555s recording spectrophotometers equipped with 1-cm path length glass cells and a Radiometer PHM84 digital pH meter with a glass and calomel combination electrode were used. Reagents The following reagents were used: a 10-3 M solution of TrAMR in methanol7.8; a standard 10-1 M de-polymerisedy solution of VV prepared from sodium vanadate and standar- dised volumetrically10; and a standard 10-1 M solution of C O ( C ~ O ~ ) ~ prepared from cobalt nitrate by treatment with perchloric acid and standardised complexometrically. 11 Tris - 0.1 M HC104 buffer solution (pH 8.10) was used as indicated.The ionic strength of the solutions was maintained at 0.25 M Analytical-reagent grade chemicals and de-ionised water (NaClOj). were used throughout with no further purification. Simultaneous Determination of Vanadium and Cobalt To a 25-ml calibrated flask were added 2.5 ml of a 2.5 M NaC104 solution, 3.0 ml of Tris - HC104 buffer solution (pH 8. lo), up to 6 ml of the sample solution containing 2.6-26 pg of vanadium and 3-30 pg of cobalt, 7.5 ml of methanol and 5.0 ml of 10-3 M TrAMR methanolic solution. The solution was made up to volume with de-ionised water. After 15 min the second-derivative spectra of the solutions were recorded against a reagent blank between 650 and 400 nm using the following conditions: slit width, 2 nm; scan speed, 120 nm min-1; chart speed, 30 mm min-1; and response time, 7 s.Simultaneous Determination of Vanadium and Cobalt in Steels A steel sample (0.1-1.0 g) was weighed accurately, treated with 5 ml of 6 M HCI and 5 ml of concentrated HN03, heated in a water-bath until dissolved and the solution taken nearly to dryness. The residue was redissolved and made up to 100 ml in a calibrated flask with 6 M HC1. Suitable aliquots were analysed as described above. Iron was first removed using diisopropyl ether. 12 Results and Discussion The TrAMR reacts with vanadium and cobalt over a wide acidity range forming the complex species V02H2R (pink) and V02HR (pink - violet), and CoH4R2 (pink) and CoH2R2 (red), respectively.*?13 Their most noteworthy spectro- photometric characteristics are given in Table 1.Various instrument parameters that affect the second-order analogue derivative spectra were studied, these being scan speed, slit width and response time. From previous experi- ments, 30 mm min-1 was chosen as the chart speed. It was found that: (i) a decrease in the scan speed caused broadening of the peaks obtained, which does not facilitate the resolution of the mixture; hence a scan speed of 120 nm min-1 was chosen; (ii) the slit width does not significantly alter the spectral resolution; a 2-nm slit width was selected; and (iii) the signal to noise ratio was too large for short response times and therefore a 7-5 response time was chosen. Fig. l(a) shows the zero-order spectra of solutions contain- ing the Co -TrAMR, V -TrAMR and Co - V - TrAMR complexes.The corresponding second-derivative spectra are shown in Fig. l(b). The distance AZl between the peak at 480 nm and the shoulder at 510 nm corresponds to the concentra- Table 1. Spectrophotometric characteristics of vanadium(V) - TrAMR and cobalt(I1) - TrAMR complexes in 50% V/V methanol - water; I = 0.25 M (NaClO,) E/lO- Concentration, Species Log K,,, 1 mol-Icm l hinm PH p.p.m. VOZHZR . . . . -0.52k0.02 1.75 525 5.0-5.3 - CoH,R2 . . . . -0.42+0.02 2.00 485 5.0-5.3 - V02HR . . . . -6.45 5 0.02 2.65 525 7.3-8.5 0.25-0.98 CoH,R, . . . . -1.69 f 0 . 0 2 4.60 50s 8.0-9.0 0.17-0.7 1ANALYST, JANUARY 1989. VOL. 114 94 0.5 T 0.3 0.1 450 530 610 1 I 50 530 61 0 Fig. 1. (a) Absorption and ( b ) second-derivative spectra o f the cobalt(I1) - TrAMK and vanadium(V) - TrAMR complexes. CK = 2 x M ; pH = 8.10.(1) Cco = 8 x lop6 M + Cv = 8 X 10-OM; (2) Ccc, = 8 X 10-6 M; and (3) C, = 8 X lop6 M. For the meaning of AZsee text 0.004 0.002 0 -0.002 -0.004 Fig. 2. SI 1 1 490 570 490 570 hinm 1 a 1 490 570 ond-derivative spectra of cobalt(I1) - TrAMR and vanad- ium( V I - i r ~ i v i ~ complexes. L~ = L x i u M; p n = 8 . 1 ~ . (a) (I) c ~ . ~ , = 8 x 10-6 M + C, = 8 x loph M; (2) Cco = 8 x lop6 M ; and (3) Cv = 8 X 10-6 M. ( b ) (1) Cc, = 1.4 X 10-5 M + Cv = 8 X 10-6 M ; (2) Cc(, = 1.4 x 10-5 M; and (3) Cv = 8 X 10-6 M. (c) (1) Cc(, = 8 X 10-6 M + Cv = 1.4 x 10-5 M; (2) Cco = 8 X 10-6 M; and (3) Cv = 1.4 x 10-5 M tion of cobalt, and the distance AZ3 = (AZ4 - AZ2) between the shoulder at 550 nm and the nesk at 580 nm corresnondc to the concentration of vanadium: AZ4 and AZ2 correspond to the concentrations of vanadium and cobalt, respectively, in pure solutions.This fact can be seen more clearly in Fig. 2 where the second-derivative spectra for solutions with a Cv/Cco ratio of 1.75, 1.0 and 0.57 are shown. Hence calibration graphs were obtained by plotting AZI and AZ2 values vs. the cobalt concentration and AZ4 (= AZ3 + AZ?) vs. the vanadium concentration, as can be seen in Figs. 3 and 4. These plots are linear and pass through the origin, the equations, assuming concentrations are expressed in p.p.m., being: 0.001 0 -0.001 -0.003 I 480 570 (b) . . 530 610 hinm 3.003 3.001 3 -0.001 -0.003 Fig. 3. Second-derivative spectra of ( u ) cobalt(I1) - TrAMR complex and ( b ) vanadium(V) - TrAMR complex.CR = 2 X 10-4 M; pH = 8.10. C,, = (1) 2qx 10-6; (2) 4 X 10-6; (3) 6 X 10 h ; (4) 1.0 X and (5) 1.4 X 10 M. C, = (1) 4 X 10-6; (2) 8 X 10 6 ; (3) 1.2 X lop5; (4) 1.6 x 10-5; and ( 5 ) 2.0 x 10-5 M 15 10 .. - a 5 0 0.5 1 .o Concentration, p.p.m. Fig. 4. Calibration graphs for cobalt(I1) (AZ, and AZ2) and vanadium(V) (A1J (see text) A11 = -0.266 + 12.381[C0] (Y = 0.999) . . (1) (alone and in mixtures); AZ2 = 0.039 + 1.843[C0] (Y = 0.999) . . (2) (correction for vanadium in mixtures); and A14 = 0.153 + 13.415[V] (Y = 0.999) . . (3) (in mixtures and alone, when AZ2 = 0). Equation (1) allows the determination of 0.12-1.18 p.p.m. of cobalt alone and equation (3) allows the determination of 0.10-1.02 p.p.m. of vanadium. The results of a statistical study carried out on two sets of 11 samples, one set containing 0.471 p.p.m.of Co and the other 0.407 p.p.m. of V, are given in Table 2. In addition, equations (1) and (3) allow the determination of cobalt and vanadium in their mixtures. Once the cobalt concentration is known from equation ( l ) , the value of AZ2 is calculated from equation (2) and added to AZ3, which isANALYST, JANUARY 1989, VOL. 114 95 Table 2. Statistical parameters for the determination of vanadium(V) and cobalt(I1) and their mixtures with TrAMR by second-derivative spectrophotometry. Vanadium(V) added = 0.407 p.p.m.; cobalt(I1) added = 0.471 p.p.m. Mixture Parameter Vanadium( V) Cobalt( 11) Vanadium( V) Cobalt (11) n- . . . . . . 0.401 0.480 0.400 0.476 a . . . . . . 1.01 x 10-2 2.79 X 10-2 1.68 x 10-2 2.63 x 10-2 a, .. . . . . 3.05 x 10-3 8.43 x 10-3 5.31 X 10k3 8.34 X 1 0 - 3 t cxp . . . . . . 1.96 1.06 1.31 0.60 Error, % . . . . 1.69 3.91 3.00 3.95 r:“,, . . . . . . 0.59 0.32 0.42 0.19 Table 3. Simultaneous determination of vanadium and cobalt in synthetic mixtures Vanadium added, Cobalt added, p.p.m. p.p.m. 1.019 0.118 0.917 0.236 0.815 0.353 0.815 0.220 0.713 0.471 0.611 0.589 0.61 1 0.439 0.61 1 0.439 0.407 0.659 0.204 1.061 0.204 0.220 0.204 0.220 Vanadium found, Relative error. p.p.m. Yo 1.017 -0.20 0.925 0.87 0.812 -0.37 0.802 - 1.59 0.696 -2.44 0.606 -0.82 0.606 -0.82 0.598 -2.13 0.413 1.47 0.205 0.49 0.21 1 7.84 0.219 7.35 Cobalt found, p.p,m. 0.102 0.23 1 0.328 0.207 0.465 0.570 0.457 0.457 0.683 1.007 0.239 0.247 Relative error, O/” - 13.56 -2.12 -7.08 -2.27 - 1.27 -3.22 4.10 4.10 3.64 5.09 7.95 12.27 Table 4.Interference levels of foreign ions in the simultaneous determination of vanadium and cobalt. Concentrations: vanadium, 0.407 p.p.m.; and cobalt, 0.471 p.p.m. [Ionl/Pl Ion added V$ C0-i: 500: 1 CJ06HJ2-, S2032-. N03-, I- - - F - , C I , B r + + so42 - Ca, Sr, Ba, TIr + + 50 : 1 MoVI, WV’ + + 10: 1 Fe”’,” Zn + + 5 : l Cut + + 1 : 1 Al, AulI1, Bi - - FerlI, Cu k k Cr”1 - SnII + Cd + -t + 100 : 1 C6O7H62 , C204? , P o d i - , Li, K, Mg, + - * In excess of F-. t In excess of S203’- and of ascorbic acid. + or -, interference 4%; 2, interference >5%. Table 5. Simultaneous determination of vanadium and cobalt in low-alloy steels Vanadium, Cobalt, O/O Reference material Certified Found* Certified Found* F-553 .. . . 1.077 1.088 4.934 4.985 J-5 . . . . . . 1.326 1.291 4.935 4.785 BCS-252. . . . 0.46 0.43 0.04 0.04 BCS-458. . . . 0.13 0.14 0.16 0.15 BAM-326-1 . . 0.024 0.026 0.223 0.180 Mean values of three determinations. determined from the second-derivative spectra of the mixture. The calculated value of AZ4 = AZ2 + AZ3 can then be used to calculate the concentration of vanadium from equation (3). The Cv/Cco range of application for the determination of cobalt and vanadium in their mixtures was investigated using synthetic samples containing both ions in different amounts. The results given in Table 3 show that the simultaneous determination of cobalt and vanadium can be achieved for samples in the Cv/Cco range 1 s 0 . 2 . Effect of Foreign Ions The effect of several ions on the simultaneous second-deriva- tive spectrophotometric determination of 0.407 p.p.m.of V and 0.471 p.p.m. of Co was investigated by first applying the recommended procedure to solutions containing a 500-fold (rnirzz) ratio of interfering ion to vanadium; if interference occurred, this ratio was reduced until the interference ceased. The criterion for interference was a deviation of more than k 5 % from the concentrations of vanadium and cobalt taken. The results are shown in Table 4. Iron can be tolerated up to a Ck,lCv ratio of 10 in the presence of fluoride ions and copper up to a Cc,/Cv ratio of 5 in the presence of thiosulphate and ascorbic acid as masking agents. Simultaneous Determination of Vanadium and Cobalt in Low-alloy Steels The proposed method was applied to the simultaneous determination of vanadium and cobalt in low-alloy steels, after removal of iron with diisopropyl ether. The results are shown in Table 5 , where they are compared with certified values.The authors acknowledge financial support from CAYCIT, Spain (grant No. PR.84-0794) and thank Professor F. Garcia Montelongo for many useful discussions. References 1. 2. Hovind, H . R., Analyst, 1975, 100, 769. Shibata, S., Flaschka, H. A . , and Barnard, A. J., Editors, “Chelates in Analytical Chemistry,” Volume 4, Marcel Dek- ker, New York, 1972, p. 1. 3. Iyer, K. V., Prakash, K. A., Iyer, S. G., and Veukateswarla, C., Indian J. Chem., Sect. A , 1985, 24, 168. 4. Spitsyn, P. K., Zavod. Lab., 1985, 51, 10. 5. Snell, F. D., “Photometric and Fluorimetric Methods of Analysis, Parts I and 11,” Wiley, New York, 1978, pp. 931 and 1191. Svehla, G., and Tolg, G., Talanta, 1976, 23, 755. Garcia Montelongo, F., Arias, J. J., JimCnez, F., and Jimenez, A. I., Mikrochim. Acta, 1983, 2 , 349. 6. 7.96 ANALYST, JANUARY 1989, VOL. 114 8. Arias, J. J., Jimknez, F., JimCnez, A. I., and Garcia Monte- longo, F., An. Quim., 1984, 80B, 247. 9. Doadrio, A., and Diaz, M. G., lnf. Quirn. Anal., 1973,27,247. 10. Charlot, G., “Les MCthodes de la Chimie Analytique. Analyse Quantitative MinCrale,” Masson, Paris, 1966, p. 972. 11. Schwarzenbach, G., and Flaschka, H., “Complexometric Titrations,” Methuen, London. 1969, p. 242. 12. Ashley, S. E. Q., and Murray, W. M., Jr., Ind. Eng. Chem., 1938, 10, 367. 13. Arias, J. J., Jimenez, F., PCrez Trujillo, J. P., and Garcia Montelongo, F., Quirn. Anal., 1983, 2 , 215. Paper 81023200 Received June I Oth, I988 Accepted July 22nd, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400093

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 97 Determination of Beryllium in Beryl by Third-derivative Spectrophotometry Using Beryllon 111 and Decolorisation of Excess of Reagent* Yu-Lun Zhu and Ji-Xin Shao Institute of Rock and Mineral Analysis, Ministry of Geology and Mineral Resources, 26 Baiwanzhuang, Beijing, China Third-derivative spectrophotometry was used for the direct determination of beryllium in geological samples. The method is based on the determination of the pure complex of beryllium with Beryllon I l l after decolorisation of excess of chromogenic reagent. Keywords: Beryllium determination; beryl; derivative spectrophotometry; Beryllon 111; decolorisation of excess reagent A method for the determination of beryllium by third- derivative spectrophotometry with Beryllon I11 after decolor- isation of the excess of chromogenic reagent is described. When tin(I1) chloride and ascorbic acid were introduced into a red solution of the complex of beryllium with Beryllon I11 at ca.pH 7, the excess of Beryllon I11 was decolorised. The beryllium - Beryllon I11 complex can be retained quantita- tively by adding triethanolamine. By means of this process, the sensitivity and precision are increased and the range of linearity is wider than that of the original rnethod.l.2 As the difference in the wavelengths of maximum absorption (Ah) of the complex and reagent is extended from 50 to 170nm, derivative spectrophotometric measurements can be carried out easily. In the presence of oxalic acid, beryllium can be determined in beryl by third-derivative spectrophotometry without any of the separation procedures that were required previously.334 Experimental Apparatus The absorption and third-derivative spectrophotometric measurements were performed on a Philips PU 8800 spectro- photometer. The third-derivative spectrophotometric parameters were: scan range 600-500 nm, scan speed 10 nm s-1, band width 1 nm, span scale 5, response 5 s and horizontal scale 50 nm cm-1.The difference in height between the adjacent peak and trough was measured on the derivative spectrum. Reagents Analytical-reagent grade reagents were used throughout. Beryllon 111 solution, 0.04%. Ascorbic acid solution, 20%. Freshly prepared. Tin(II) chloride solution, 1% in 2% hydrochloric acid. Beryllium ion stock solution, 100 pg ml-1.Prepared from Prepared just before use. B e 0 (spectral purity) in 0.5% sulphuric acid. Procedure To a 25-ml calibrated flask, containing less than 1Opg of beryllium, add 1 ml of 10% oxalic acid and 1 drop of thymolphthalein indicator. Adjust the colour of the solution to blue with 10% sodium hydroxide solution, immediately add 3.0ml of 0.04% Beryllon 111 solution, mix the solution and * Presented at the Third China - Japan Joint Symposium on Analytical Chemistry, Hefei, China, May 1&13th, 1988. allow it to stand for 10 min, then add 5 ml of 50% ammonium acetate solution, 1 ml of 20% ascorbic acid and 0.20 ml of 1% tin(I1) chloride solution and stir the solution. After 20mi1-1, add 3 ml of 1 + 1 V/V triethanolamine, dilute to the mark with distilled water and shake well.The third-derivative value is measured as described earlier against water using a 1-cm cell. Results and Discussion pH Value The formation of the complex of beryllium with Beryllon 111 occurs in the pH range 10.5-1 1.5,lJ but the absorbance of the complex is highest in the pH range 6-8. The addition of 5 ml of 50% ammonium acetate solution to the system after the colour has developed enables the reaction mixture to be maintained at ca. pH 7. Effect of Excess of Chromogenic Reagent Pakalns and Flynnl used Beryllon I11 for the determination of beryllium. The range over which Beer’s law is obeyed can be extended by using a higher reagent concentration, but the absorbance of the blank against water is then high. For this reason, the concentration of Beryllon 111 that can be used is limited and Beer’s law is obeyed only from 0 to 0.18 pg ml-* of beryllium.In this work, to form the complex with 10 pg of beryllium in 25 ml of solution it was found to be sufficient to add 3 ml of 0.04% Beryllon I11 solution; however, the difference in the wavelengths of maximum absorption of the complex and reagent was small, the absorbance of the reagent blank at 520nm was 2.5A (1-cm cell) and the precision was poor. Therefore, this method required modification to eliminate the influence of the excess of chromogenic reagent. Decolorisation of Excess of Chromogenic Reagent The problem mentioned above can be solved by decolorisa- tion of the excess of chromogenic reagent,s but the determina- tion of beryllium with Beryllon 111 is carried out at ca.pH 7, which is different to that used in the earlier work.5 If tin(I1) chloride is used as a decolorant, it will be hydrolysed. In this work, the tin(I1) chloride - ascorbic acid system was tested as a decolorant. When this mixture was introduced into the red solution of the beryllium - Beryllon I11 complex, the azo link in the excess of Beryllon 111 was broken by reduction by tin(I1) chloride, the solution became colourless and the absorption maximum of the reagent shifted from 47.5 to 350nm. The reaction is shown in Fig. 1. The colour of the chelate compound was retained selec- tively, exhibiting an absorption maximum at 520nm. The98 H03S ANALYST, JANUARY 1989, VOL. 114 I H03S Red H03S I H03S Colourless Fig. 1. Reaction scheme H03S Fig. 2.Chelate compound 0.8 al C ([I +? 5: 2 0.4 300 400 500 600 Wavelengthinm Fig. 3. 20: 1 min-l Decolorisation of Beryllon I11 solution. Number of cycles, ratio of Be to Beryllon I11 was found to be 1 : 1 and the stable structure shown in Fig. 2 is proposed. The decolorisation can be achieved by adding 1 ml of 20% ascorbic acid and 0.20 ml of 1% tin(I1) chloride solution. A scan programme with cyclical overlapping was applied to record the progress and the end of the reaction (Figs. 3 and 4). By this means, the absorbance of the reagent blank was about zero at 520 nm. Hence the precision of the method is increased and this characteristic can be used for measurement of the pure complex. As the difference in the wavelength of maximum absorption of the complex and reagent is extended from 50 to 170 nm, derivative spectrophotometric measure- ments can be performed easily.Because the excess of tin(I1) chloride affects the dissociative equilibrium, it causes instability of the complex. By adding 0.8 al C ([I + 5: 3 0.4 300 400 500 600 Wavelength/nm Fig. 4. of cycles, 20; 1 min-1 Decolorisation of beryllium - Beryllon 111 solution. Number 2.0 i 1.6 0.4 0 2 4 6 8 1 0 Concentration of Be/pg per 25 ml Fig. 5. Calibration gra hs. (1) Decolorisation of excess of chro- mogenic reagent; and (27 in the presence of excess of chromogenic reagent 0.6 i al 6 0.4 e s n a 0.2 0 400 450 500 550 600 Wavelengthinm Fig. 6. 0.42 mg; (4) Ti, 0.20 mg; and ( 5 ) Beryllon I11 Absorption spectra. (1) Be, 4 pg; (2) Al, 0.53 mg; (3) Fe, 3 ml of 1 + 1 triethanolamine as a stabilising agent, the effect of tin(I1) chloride is inhibited and the colour of the beryllium complex does not change for at least 4 h.We found that the presence of ascorbic acid not only prevented hydrolysis of tin(I1) chloride but also prevented the colourless blank solution from reverting to red with an absorption peak identical with that of Beryllon 111 after adding triethanolamine. Calibration Graph for Spectrophotometric Measurements Under the conditions described above, Beer's law is obeyed from 0 to 10 pg of Be in 25 ml of solution and the linear range is twice that achieved in the original work' (Fig. 5). The reasonANALYST, JANUARY 1989, VOL. 114 99 Wavelengthinm Fig. 7. Third-derivative spectra. (1-5) As in Fig. 6; and (6) Al, 0.42 mg Table 1.Analytical results for beryllium (Yo) Pyrophosphate Sample* Proposed method method Beryl-1 . . . . . . 4.67 4.62 Beryl-2 . . . . . . 4.35,4.36,4.43,4.36, 4.44,4.47 4.36,4.47.4.32,4.33 Hsianghualite (raw ore) . . 0.353,0.363,0.367 0.357 t * Beryl, Be3Al2Si6Ol8; hsianghualite, Ca3Li2Be3(Si04)3F2. i Emission spectrometric method. for this is that the reagent is sufficient for the formation of the complex and in this instance, as the reagent is present in excess for complex formation, decolorisation reduces the absorbance of the excess of free reagent to a constant level. In the absence of decolorisation the absorbance of the excess of uncomplexed reagent shifts the peak positions to an extent depending on its concentration, resulting in a non-linear calibration graph.The ratio of the slopes of the two calibration graphs is 2.3, that is, the sensitivity is 2.3 times as high as that found in the presence of excess of chromogenic reagent. Without decolori- sation, the sample absorbance is measured against a full reagent blank with no correction for reagent consumption. In the proposed method, this error is eliminated and the sensi tivi ty increased. Inhibition of Oxalic Acid The effects of 42 foreign ions on the determination of 2-4 pg of beryllium were examined. The amounts of these ions tolerated (mg, relative error for beryllium <?4%) are as follows: Na+ 200: K+ 260; N03-,Cl- 300; PO43-,Sn1V 2.0; Ca2+, Sr2-t 0.2; Zn'+ 0.13; Cu2+, AsV, SbV, Ba2+ 0.1; Pb2+, Cd2+ 0.06; Mg", Gel+, Bi3+, Rev11 0.05: CrVI, WVI, Ag+ 0.04; Ni*+, UVI, Mo\'I 0.03; VV, I++.TP+, ZrI\.- NbV, TaV, Hg2+ 0.02; and Mn2+, SeI\', TeIk, Co2+, Au-i+ 0.01. Some foreign ions, such as AP+, Fe3+ and Ti'", at the level of a few micrograms, interfere seriously with the determination of beryllium. Owing to the addition of oxalic acid, the reaction of a large amount of AP+, Fe3+ and TiIV with Beryllon TIT is evidently inhibited (Fig. 6) and the decolorisation process is acceler- ated. Derivative Spectrophotometric Measurements The appearance of the derivative spectra of the complexes of beryllium, aluminium, iron and titanium was studied in the presence of oxalic acid (Fig. 7). It was found that the third-derivative spectra of the complexes of aluminium, iron and titanium from 600 to 530nm are similar to that of the reagent blank after decolorisation.As their third-derivative spectra are measured against a reagent blank, the values at 590 and 5.50 nm are zero; these are the wavelengths at which the peak and trough of the beryllium complex occur. It is possible to eliminate the remaining interferences from these ions as shown in Fig. 6 and to determine beryllium selectively without any separation procedure. The third-derivative spectra were plotted by taking a series of standard beryllium solutions (Be, pg per 25 ml: 0, 2.0, 4.0, 6.0, 8.0 and 10.0). The calibration graph was plotted as the difference in height between the adjacent peak and trough versus the beryllium concentration. Application For the determination of beryllium in beryl and hsianghualite (raw ore), 5.00-20.00mg of sample were weighed into a platinum crucible containing 1 g of potassium bifluoride, water was drained off and the melting agent was fused. After adding 5 ml of concentrated sulphuric acid, heating to dissolve the sample and evaporating to drive off sulphur trioxide vapour, the contents were fused on a Bunsen burner until transparent and SO or 100 ml of a solution in 5% hydrochloric acid were prepared. An aliquot of the solution was taken and beryllium was determined by the proposed procedure. The results for beryllium were consistent with those found using the pyrophosphate and emission spectrometric methods (Table 1). The relative standard deviation for a sample containing 4.37% of beryllium was 1.11% ( n = 8 , standard deviation = 0.048%). References 1. 2. Cheng, S. Z., et a f . , "Rocks and Minerals Analysis," 3. 4. 5 . Pakalns, P., and Flynn, W. W., Analyst, 196.5, 90, 300. Geological Publishing House, Beijing, 1974. p. 482. Sauerrer, A . , and Troll, G., Talanra, 1984, 31, 249. Sharma, C., and Khopkar. S. M., Anal. Chim. A m , 1985,169, 403. Zhu. Y.-L.. and Shao, J.-X., Fenxi Huaxue. 1986. 14. 410. Paper 8102276C Received June 7th, I988 Accepted July 18th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400097

出版商:RSC    

年代:1989

数据来源: RSC

摘要:

ANALYST, JANUARY 1989, VOL. 114 101 Simultaneous Spectrophotometric Determination of Mefenamic Acid and Paracetamol in Pharmaceutical Preparations Sukomal Das, Suresh C. Sharma, Santosh K. Talwar" and P. D. Sethi Central Indian Pharmacopoeia Laboratory, Raj Nagar, Ghaziabad-20 1 002, India A spectrophotometric procedure for the simultaneous determination of mefenamic acid and paracetamol in a mixture is described. Using 0.01 M methanolic hydrochloric acid as solvent, the absorbance of the mixture is measured a t 248, 279 and 351 nm. The concentration of each component can be calculated by solving two equations using two wavelengths, either 248 and 279 nm or 248 and 351 nm. Keywords: Anti-inflammatory and analgesic agents; mefenamic acid; paracetamol; simultaneous spectrophotometry; Vierordt's method Mefenamic acid in combination with paracetamol is very useful for the treatment of moderate pains connected with rheumatoid arthritis.The British Pharmacopoeia describes a titrimetric method for the determination of mefenamic acid in both the bulk drug and dosage form.' A wide range of chromatographic tech- niques have been used to determine mefenamic acid in blood and urine .*-5 Fluorimetric methods based on a cyclisation reaction of the N-phenylanthranilic acid derivative of mefe- namic acid resulting in the formation of benzoxazine or acridones have been described for its sensitive determina- tion.6-7 Other methods include titrimetry,g.g polarography,lO colorimetry,lI7'2 spectrophotometry13 and chromatography. 14 Paracetamol (acetaminophen) and its dosage forms are official in the BP," Indian Pharmacopoeia (IP)16 and USP.17 Both the BP and IP describe a titrimetric assay for the bulk drug, but a spectrophotometric method is used for the dosage form.In the USP,17 paracetamol is determined spectro- photometrically in the bulk drug but in the dosage form it is determined using various techniques including colorimetry, spectrophotometry , high-performance liquid chromatography (HPLC) and gas - liquid chromatography (GLC). Para- cetamol has also been determined by other methods.18 Para- cetarnol in combination with other analgesics has been determined by titrimetry,'" spectrophotometry,2".21 col- orimetry,22.2-' chemical ionisation mass spectrometry,24 GLC25 and HPLC.26 There is only one report27 on the determination of mefenamic acid and paracetamol in a mixture, but it involves separation before spectrophotometric measurement.This work was prompted by the need to develop a simple, fast, sensitive and accurate method for the routine analysis of this mixture. Experimental Hydrochloric acid and methanol were of analytical-reagent grade. Mefenamic Acid and Paracetamol Reference standards of mefenamic acid and paracetamol were obtained from the Central Drugs Laboratory, Calcutta, India. Dosage form Mefenamic acid - paracetamol tablets (250 mg of mefenamic acid, 500 mg of paracetamol), marketed as Lanagesic by Tata Pharma, Bombay, India, were obtained from the manufact- urer. Apparatus A Beckman Model-24 UV - visible spectrophotometer equipped with a recorder was used for all UV measurements.Met hods An accurately weighed amount of powdered tablets equiv- alent to about 100 mg of paracetamol was placed in a 100-ml calibrated flask. A 40-ml volume of 0.01 M methanolic hydrochloric acid was added to dissolve the medicament and the solution was diluted to 100 ml with the same solvent. The resulting suspension was filtered and 10 ml of the filtrate were diluted to 100 ml with 0.01 M methanolic hydrochloric acid. A 10-ml sample of the resulting solution was further diluted to 100 ml with 0.01 M methanolic hydrochloric acid. The absor- bance of this solution was measured in a 1-cm silica cell at 248, 279 and 351 nm using 0.01 M methanolic hydrochloric acid as a blank. The amounts of mefenamic acid and paracetamol were calculated using either of two sets of equations: Set A Mass (mg) of mefenamic acid = 332.2A351 .. . . (1) Mass (mg) of paracetamol = 113.0A24X-S4.4A351 . . (2) Set B Mass (mg) of mefenamic acid = 314.8A270-53.0AZ4X . . (3) Mass (mg) of paracetamol = 126.6A248-7Y.9A27Y . . (4) The derivation of these equations is discussed below. Results and Discussion In the analysis of binary mixtures containing two absorbing substances, absorptiometric principles are used. Pernarowski et al.2s-30 applied the absorbance ratio for the analysis of pharmaceuticals. Similarly, Vierordt's31 method has also been used successfully for the analysis of binary mixtures.32-34 In 1.2 a, 5 0.8 9 s: 0.4 2 0 I I 351 I 200 250 300 350 Wavelengthinm * To whom correspondence should be addressed.Fig. 1. and (B) paracetamol in methanolic hydrochloric acid (0.01 M) UV spectra of a 0.001 Yo m/V solution of (A) mefenamic acid102 ANALYST, JANUARY 1989, VOL. 114 Table 1. Recovery experiments on authentic samples containing mefenamic acid (MEF) and paracetamol (PAR) in different ratios Recovery (mean k SD*), Yo Sample No. 1 2 3 4 5 6 7 8 9 * n = 5 . Ratio, MEF : PAR 2.5 : 3.0 2.5 : 4.0 2.5 : 5.0 2.5 : 6.0 2.5 : 7.0 1.5: 5.0 2.0 : 5.0 3.0 : 5.0 3.5 : 5.0 MEF Set A Set B 98.5 + 0.41 99.3 k 0.48 97.8 k 0.38 100.0 I 0 . 5 0 97.8 + 0.41 99.9 + 0.61 98.5 k 0.31 100.2 5 0.37 98.5 I 0.49 100.0 + 0.51 99.5 k 0.50 99.8 k 0.41 99.2 I 0.46 100.0 I 0.42 99.5 k 0.38 99.6 k 0.36 101.2 k 0.36 101.0 i 0.29 PAR Set A 99.4 k 0.50 100.8 k 0.62 100.3 k 0.73 100.1 k 0.48 100.5 k 0.47 99.9 k 0.62 99.9 i 0.46 100.0 + 0.39 100.5 k 0.39 Set B 99.3 i 0.48 100.5 k 0.55 100.1 k 0.51 99.9 k 0.49 100.3 k 0.60 99.8 k 0.47 99.8 i 0.42 100.0 k 0.56 100.5 k 0.41 Table 2.Recovery experiments on standard additions to a commercial sample. Lanagesic tablets: declared content, mefenamic acid 250.0 mg, paracetamol500.0 mg. Found by set A equations, mefenamic acid 247.0 mg and paracetamol489.0 mg; by set B equations, mefenamic acid 244.8 mg and paracetamol490 mg Mefenamic acid Paracetamol Recoveredlpg Recovery, % t Sample Recoveredlpg Recovery, % $ Nd. Added/& SetA SetB SetA SetB Added/& SetA SetB SetA SetB 1 2 3 3 5 6 7 8 9 1 0 - - - - - 48.95 49.9 50.1 101.8 102.3 97.9 98.05 98.8 100.2 100.9 - 146.8 147.9 150.5 100.75 102.2 - - - - - 195.8 199.4 199.4 101.8 101.8 244.75 247.5 250.0 101.1 102.1 - - - - - - - - - - - - - - 49.7 50.6 50.5 101.8 101.6 99.4 101.4 100.8 102.0 101.4 148.5 150.0 99.7 100.6 - 149.1 - - - - 198.8 200.9 200.5 101.0 100.8 - - - 248.5 251.7 251.2 101.3 101.1 - - - - - - - - - - - - - - - - - Amount in 1-18 added to the final dilution of the preparation.t Mean recovery (k SD) of mefenamic acid: set A, 101.13 I 0.62%; sct B, 101.86 k 0.54%. $ Mean recovery (+ SD) of paracetamol: set A , 101.16 k 0.81Yo; set €3, 101.1 k 0.35%. Table 3. Results obtained with commercial samples. All samples were Lanagesic tablets (Tata Pharma, Bombay, India). Declared content per tablet: mefenamic acid 250 mg and paracetamol 500 mg Mefenamic acid found, Yo of declared content (k SD*) ~~~ - Sample No.Set A Set B 1 100.1 I 0.31 100.3 t 0.31 2 99.7 If- 0.25 98.5 k 0.46 3 101.1 t 0.40 99.8 k 0.51 4 99.3 k 0.42 98.2 5 0.56 5 99.2 + 0.37 97.9 + 0.62 6 101.7 k 0.48 99.8 + 0.28 7 100.0 I 0.27 96.7 5 0.42 8 99.1 + 0.27 100.7 k 0.64 9 100.2 i 0.32 98.0 k 0.29 10 99.8 t 0.31 99.0 I 0.59 * n = 5 . Paracetamol found, o/o of declared content (I SD*) Set A Set B 98.4 + 0.12 98.4 k 0.21 98.1 5 0.17 98.3 -t 0.19 99.5 + 0.19 99.6 k 0.21 96.5 k 0.16 96.7 2 0.17 97.8k 0.19 98.0 k 0.26 97.7 I 0.23 97.9 k 0.36 95.9 + 0.26 96.4 k 0.29 98.5 k 0.31 99.0 -t 0.32 96.3 k 0.24 96.8 & 0.29 97.3 k 0.17 97.8 k 0.17 both instances factors affecting the results included the accuracy of the absorbance measurements, the relative concentrations of the active ingredients, the nature of the pharmaceuticals and the spectral characteristics of the com- ponents of the mixture.In Vierordt's method, which was used in this study, it is necessary to select two points on the wavelength scale where the absorptivities are at a maximum. The wavelengths so chosen for either of the substances should not coincide with a sharply sloping part of the spectral curve of the other compound. The most suitable pair of wavelengths for the mixture of mefenamic acid and paracetamol would be either 351 and 248 nm or 279 and 248 nm (Fig. 1). Mefenamic acid also shows a peak at 220 nm in addition to those at 279 and 351 nm, but this was not selected because paracetamol shows a sharp slope in the spectral pattern in this region. The figures given in equations (1)-(4) were obtained by numerical substitution for the values of QI and b in equations ( 5 ) and (6), taking into consideration the dilution factor given in the method.. a ( 5 ) (A2QIl--AlQI2) (a1P2- a2PJ Concentration of mefenamic acid = . ' (6) Concentration of paracetamol = (A1P2-A2Pd (QId32-QI2Pd where Al and A2 are the absorbance values of the mixture either at 351 and 248 nm or at 279 and 248 nm, respectively; a and represent the absorptivities of mefenamic acid and paracetamol, respectively, at the relevant wavelengths. The measured absorptivities (A~''&,) of mefenamic acid at 248, 279 and 351 nm are 224.7, 355.5 and 301, respectively, whereas those of paracetamol are 885,149 and 0, respectively. These values were obtained by running reference standards of mefenamic acid and paracetamol in 0.01 M methanolic hydrochloric acid.The usual ratio of mefenamic acid to paracetamol is 1 : 2. It is therefore essential to determine the ratios at which one substance can be accurately measured in the presence of the other. Table 1 shows that the recoveries are satisfactory for different ratios of mefenamic acid to paracetamol from 2.5:7.0 to 2.5:3.0 and from 1.5:5.0 to 3.5:S.O. The recoveries of mefenamic acid using set A equations wereANALYST, JANUARY 1989, VOL. 114 103 between 97.8 and 101.2% whereas using set B equations they were between 99.3 and 101.0%. Similarly for paracetamol, with set A equations the recoveries were between 99.4 and 100.8% and using set B equations between 99.3 and 100.5°/~. The results indicate that the content of each component in the dosage form can be reliably determined by using the proposed method.The results shown in Tables 2 and 3 confirm these findings. Recoveries of standard additions to a commercial sample are given in Table 2. The recoveries are in the ranges 100.2- 101.8% (set A) and 100.9-102.3% (set B) for mefenamic acid and 99.7-102.0% (set A) and 100.6-lOl.6% (set B) for paracetamol. Table 3 gives results for the analysis of commer- cial samples using the proposed method. The results clearly demonstrate the utility of the proposed method for the determination of drug compounds in dosage forms. The validity of the method for pharmaceutical prepara- tions and the effect of interferences were studied by assaying authentic samples of the drug and common additives and excipients, e.g., lactose, dicalcium phosphate, Aerosil 400, talc. magnesium stearate, starch and gelatin.The recoveries were 1 . 2. 3. 3 . 5. 6. 7. 8. 9. 10. 11. 12. satisfactory. References “British Pharmacopoeia 1980,” HM Stationery Office, Lon- don, 1980, pp. 273 and 536. Dusci, L. J . , and Hackett, L. P., J . Chromatogr., 1978, 161, 340. Cotellessa, L.. Riva, R., Salva. P . , Marcucci, F., and Mussini, E . , J. Chromatogr., 1980, 192, 441. Demetriou. B., and Osborne, B. G . , J . Chrornatogr., 1974,90, 405. Bland, S. A., Blake, J . W., and Ray, R . S . , J . Chrornatogr. Sci., 1976, 14, 201. Dell, H. D . , and Kamp, R., Arch. Pharm., 1970, 303, 785. Schmollack, W., and Wenzel, U.. Pharmazie, 1974, 29, 583. Agrawal, Y. K., Sci. Cult., 1980, 46, 309.Hassib, S. T., Safwat, H. M., and El-Bagry. R . I., Analysf, 1986, 111, 45. Pryakhin, 0. R., Vdoviko, V. A , , Pokhmelkina, S. A., Petrenko, V. V., andGolovkin, V. A , , Khim. Farm. Zh., 1981, 15, 93; Anal. Ahstr., 1982, 43, 6E46. El-Sadek, M., Baraka, M., and Aboul Khier, A., lndian J . Pharm. Sci., 1987, 49, 97. Mahrous. M. S., Abdel-Khalek, M. M., and Abdel-Hamid, M. E . . Talanta, 1985, 32, 651. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Sastry, C. S. P., Mangla, D. S . , and Vijaya, D., Indian Drugs, 1985, 22, 653. Adams, W. M., J . Assoc. Off. Anal. Chem., 1983, 66, 1178. “British Pharmacopoeia 1980,” HM Stationery Office, Lon- don, 1980, pp. 326, 556 and 799. “Pharmacopoeia of India 1985,” Controller of Publications, Delhi, 1985, pp.359 and 360. “United States Pharmacopeia XXI,” US Pharmacopeial Con- vention, Rockville, MD, 1985, pp. 11-14, 212, 900 and 905. El-Obeid, H. A., and Al-Badr, A. A , , in Florey, K., Editor, “Analytical Profiles of Drug Substances,” Volume 14, Academic Press, New York, 1985, pp. 551-5196. Rhodes, H. J . , Denardo, J . J . , Bode, D. W., and Blake, M. I., J . Pharm. Sci., 1975, 64, 1380. McDowell, A. E . , and Pardue, H. L., J . Pharrn. Sci., 1978,67, 822. Peterkova, M., Matousova, O., and Kakac, B., Cesk. Farm., 1981, 30, 270. Ouf, A. A , , Walash, M. I., Hassan, S . M., and Elsayed, S. M., Arch. Pharm. Chem. Sci. Ed., 1978, 6, 164. Murfin, J . W., Analyst, 1972, 97, 663. Kanamori, H., HiroJhima-Ken Eisei Kenkyuyho Kenkyu Hokoku, 1982, 29, 31. Windorfer, A. J., and Roettger, H. J . , Arzneim. -Forsch., 1974, 24, 893. Corrol, M. A . , White, E. R . , and Zarembo, J . E., Anal. Chem., 1981, 53, 1111A. Issa, A. S . , Bettagy, Y. A., Kassem, M. G., and Daabees, H. G., Talanta, 1985, 32, 209. Pernarowski, M., Knevel, A. M., and Christian, J. E., J . Pharm. Sci., 1961, 50, 943. Pernarowski, M., Knevel, A. M., and Christian, J . E., J . Pharm. Sci., 1961, 50, 946. Pernarowski, M., Knevel, A. M., and Christian, J . E., J . Pharm. Sci., 1962, 51, 688. Stern, E. S . , and Timmons, C. J . , “Electronic Absorption Spectroscopy in Organic Chemistry,” Third Edition, Arnold, London, 1970, p. 212. Hassan, S. M., Metwally, M. E. S . , and Ouf, A. M. A . , J . Assoc. Off. Anal. Chem., 1983, 66, 1433. Talwar, S. K., Das, S . , and Sharma, S . C., J. Pharm. Biomed. Anal., 1986, 4, 511. Sharma, S. C., Das, S . , andTalwar, S . K . , J . Assoc. Off. Anal. Chem., 1987, 70, 629. Paper 8103151 G Received August 2nd, I988 Accepted September 20th, 1988

ISSN:0003-2654    

DOI:10.1039/AN9891400101

出版商:RSC    

年代:1989

数据来源: RSC