ISSN:0003-2654    

DOI:10.1039/AN98207FX029

出版商:RSC    

年代:1982

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98207BX031

出版商:RSC    

年代:1982

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98207FP077

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

x August, 1982 SUMMARIES OF PAPERS I N THIS ISSUESmdy of Some Systematic Errors During the Determination of theTotal Selenium and Some of Its Ionic Species in Biological MaterialsA method using nitric acid alone for the digestion of biological materials forthe determination of total selenium or of some of its ionic species leads toerroneous results because of incomplete mineralisation of some organicselenium compounds, including the selenonium derivatives, the main meta-bolites of the element in urine, and a selenoamino acid present in plant andanimal tissues; such a method also suffers from many interferences owing tothe presence of incompletely digested organic constituents of the sample.The use of perchloric acid under certain conditions and of sulphuric acid, inconjunction with nitric acid, leads to an excellent recovery of the seleniumadded to biological materials or included in organic selenium compounds.The nitric acid method, the only one which affords a selective determinationof some selenium ionic species in aqueous solution, has been critically examinedbut its efficiency cannot be improved; all of the parameters investigated arehelpful, however, in choosing an adequate digestion procedure for thedetermination of the total selenium content of a complex sample.Keywords : Selenium determination ; biological samples ; organic seleniumJEAN NRVE, MICHEL HANOCQ, LaOPOLD MOLLE and GISfCLELEFEBVREUniversity of Brussels, Institute of Pharmacy, Campus Plaine 205/ 1, B-1050 Brussels,Belgium.Analyst, 1982, 107, 934-941.compounds ; atomic-absorption spectrophotovnetryDirect Determination of Arsenic in Coal by Atomic-absorptionSpectroscopy Using Solid Sampling and Electrothermal AtomisationA method using graphite furnace atomic-absorption spectroscopy for thedetermination of arsenic in coal is described.The analysis may be performedon either a solution of coal or, preferably, by direct atomisation of a slurry ofpowdered coal. Dissolution of coal is brought about using a mixture of nitricacid, hydrofluoric acid and hydrogen peroxide. Both procedures requirematrices of nickel nitrate (volatilisation suppressant) and magnesium nitrate(ashing aid). The arsenic contents of a selection of coals determined by thesolution method were in good agreement with certified values and resultsobtained by other analytical procedures. The solid-sampling technique gavethe best results when powdered coal standards of known arsenic content wereused for calibration. The analytical characteristics of the method comparefavourably with the existing British Standard method, which was shown togive low recoveries for certain coals. The over-all analysis time for theproposed slurry technique is 10niin and the detection limit is less than0.1 yg g-1.Keywords ; Arsenic determination ; coal analysis ; atomic-absorption spectro-scopy ; electrothermal atomisation ; solid samplingL. EBDONDepartment of Environmental Sciences, Plymouth Polytechnic, Drake Circus,Plymouth, Devon, PL4 8AA.and W. C. PEARCENational Coal Board, East Midlands Regional Scientific Department, Station Street,Mansfield Woodhouse, Nottinghamshire.Analyst, 1982, 107, 942-950

ISSN:0003-2654    

DOI:10.1039/AN98207BP083

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

AUGUST 1982 The Analyst Vol. 107 No. 1277 The Heterogeneous Composition of Pharmaceutical-grade Nystatin A. H. Thomas, P. Newland and N. R. Sharma Division of Antibiotics, National Institute f o r Biological Standards and Control, Holly Hill, London, N W3 6RB Samples of nystatin from diff erent manufacturers have been analysed bio- logically and chemically. Examination by high-performance and thin-layer chromatography has shown that pharmaceutical-grade nystatin is a complex mixture. The composition of nystatin varied with the source of manu- facture. The existing pharmacopoeial specifications were unable to detect these wide differences in composition. For the atypical samples there was a significant difference between the potency estimates obtained with S. cerevisiae and C .tropicalis; no such differences were apparent with the other nystatin samples examined. Keywords Nystatin composition ; high-pevformance liquid clzromntography ; thin-layer chromatografihy ; microbiological assay Nystatin, a tetraene antifungal antibiotic, is known to be a complex mixture of closely related constituents. Thin-la yer chromatography (TLC) and high-perf ormance liquid chromatography (HPLC) have shown samples of nystatin currently available to be even more heterogeneous, than had been previously reported. Pharmaceutical-grade nystatin is considered to consist largely of nystatin A,.4 The structure of nystatin A, has been eluci- dated5 (Fig. 1) in addition to the glycosidally linked mycosamine; nystatin A, and polyfungin B contain digitoxose as an additional neutral sugar.6 No information could be found about nystatin A,.Polyfungin contains four tetraene antibiotics: nystatins A,, A, and A, and polyfungin B.2 Samples of nystatin and polyfungin have been examined to determine whether the atypical samples of nystatin comply with the specifications of the British Pharmacopoeia and to determine whether the pharmacopoeial tests would differentiate these heterogeneous materials. H F O O o H OH OH OH OH 16 HO OH 17 COOH (3-43 H3C H Fig. 1. Nystatin A, (C,,H,,NOI,, relative molecular mass 926). Experimental Materials Samples of pharmaceutical-grade nystatin currently available, representing the principal manufacturers from China, Hungary, Italy, USA and USSR, were examined. Samples of amphotericin A, lucensomycin, natamycin and polyfungin were generously provided by The Squibb Institute (Princeton, N J, USA), Farmitalia Carlo Erba (Milan, Italy), Gist Brocades 849850 THOMAS et al.: HETEROGENEOUS COMPOSITION Analyst, Vol. I07 (Delft, The Netherlands) and Instytut Prezemyslu Farmaceutycznego (Warsaw, Poland), respectively. Analytical-reagent grade solvents and reagents were obtained from BDH Chemicals (Poole) . - Microbiological Assay The suggested diffusion assay described in the British Pharmacopoeia4 using Saccharomyces cerevisiae NCYC 87 (NCTC 10716) was used except that the medium was supplemented by the addition of 2% m/V of sodium chloride. The samples were dissolved in dimethyl sulphoxide to give a solution of 4000 IU ml-l; dilutions were made in phosphate buffer containing 5% V/V of dimeth~lformamide.~ The concentration of dimethyl sulphoxide present in the assay solutions did not produce any zone of inhibition.Some samples were assayed using Candida tropicalis (NCYC 1393) [this strain was kindly provided by Dr. M. Dubost of Rh8ne-Poulenc (Seine-sur-Vitry, France)] ; the same diffusion assay was used. The assay medium was inoculated with 1.0 x 105-2.0 x lo5 colony-forming units ml-1 of the assay organism. The concentrations of the assay solutions were 200, 100 and 50 IU ml-1 with S. cerevisiae and 400, 200 and 100 IU ml--l with C. tropicalis. The assays were carried out in Petri dishes, the solutions (8 p1 of each) being applied automatically from syringe needles on to the agar layer. Each assay was designed to contain sufficient information to provide from its own internal evidence a measure of the potency of the “test” in terms of the “standard” and the fiducial limits to that measurement. Three dose levels of the “standard” and three dose levels of the “test” were tested simultaneously, i.e., on each Petri dish.Six replicate dishes were used for each assay. The assays were in terms of the British Standard for Nystatin, which is identical with the proposed second international standard for nystatin. The Petri dishes were incubated in the dark at 32 “C for 18 h, after which the areas of the zones of inhibition of growth were measured in arbitrary units and recorded with an image analyser (Optomax, Micro Measurements, Saffron Walden). Thin-layer Chromatography Pre-coated silica plates (silica gel 60, 20 x 20 cm, 0.25 mm thick; E.Merck, Darmstadt, Germany) were developed with the lower phase chloroform - methanol - water (20 + 22 + 10). Before use the plate was pre-run in the mobile phase and dried at 100 “C for 30 min. Methanolic solutions of the samples (3 mg ml-l) were freshly prepared and 2 p1 were applied to the plate. The plate was placed in a saturated chromatography tank so that it was not in contact with the solvent and left for 30 min to become impregnated with the vapour of the mobile phase. The plate was developed over a 15-cm path in the same direction as the pre-run. The separated components were detected after spraying the air-dried plate with a mixture of 5% V/V anisaldehyde and 50/, V/V sulphuric acid (sp.gr. 1.84) in methanol and heating at 100 “C for approximately 5 min. For bioautographic detection the chromatoplate was dried at room temperature in the dark for 2 h and a 4-5 mm thick layer of solidified assay medium inoculated with either C. tropicalis or S. cerevisiae was placed in contact with the surface of the plate. Biological activity was revealed as clear zones of inhibition of growth after overnight incubation at 32 “C. High-performance Liquid Chromatography The apparatus consisted of two reciprocating pumps, a gradient controller (Constametric I, Constametric I1 G and a Gradient Master, Model 1601; Laboratory Data Control) and a variable-wavelength spectrophotometer (CE272 ; Cecil Instruments Ltd.) fitted with a 75-pl flow-through cell. A pre-column (35 x 4 mm i.d.) and a column (100 x 4 mrn i.d.) were packed with MOS Hypersil (Shandon Southern Products), 5 and 3 pm, respectively. The mobile phase was composed of methanol and sodium acetate buffer (pH 5.8); it was filtered through a glass microfibre filter and de-gassed prior to use.The gradient used was provided by a starting solution of 55.0% l’/V methanol in 0.005 M acetate buffer and a terminating solution of 70.0% V/V methanol in 0.005 M acetate buffer; the pH of both mixtures was 6.8 The gradient elution profile was a linear increase in methanol concentration from 55.0 to 58.75% V/V in 15 min, then held for 10 min, followed by a linear increase from 58.75 to 67.0% V/V methanol in 10 min and continued for a further 15 min. The flow-rate was 0.1.August, 1982 OF PHARMACEUTICAL-GRADE NYSTATIN 851 1 ml min-l and the pump outlet pressure was 3200 lb i r 2 .The samples were dissolved in methanol at concentrations of 3-5 mg ml-l and 5-pl volumes were injected on to the column. The column was re-equilibrated with the starting solution in the course of 5 min before injecting the next sample. The eluate was monitored at 304 nm with a sensitivity of 0.5 a.u.f.s. The content of nystatin A, was determined from the area of the peak due to nystatin A,, expressed as a percentage of the total area of all the peaks derived from the sample being examined. Absorbance Each sample (100 mg) was dissolved in a mixture of 5 ml of glacial acetic acid and 50 ml of methanol and diluted to 100 ml with methanol. The solution was diluted 1 + 99 with methanol and examined between 210 and 350nm.The nystatin and polyfungin solutions showed four absorption maxima at 230, 291, 304 and 318 nm, and a shoulder at 280 nm. The ratios of the absorbances at the maxima at 230, 291 and 318 nm to the absorbance at the maximum at 304 nm, and the ratio of the absorbance at the maximum at 230 nm to the absorbance at the shoulder at 280 nm were calculated. The Ai2i value at 304 nm was also calculated. Results and Discussion All of the samples of nystatin and polyfungin gave a positive reaction to the two identifica- tion tests described in the British Pharmacopoeia. None of the samples could be differenti- ated by the colour produced by reaction with one drop of either concentrated sulphuric acid or concentrated hydrochloric acid.All of the samples complied with the light absorption test of the British Pharmacopoeia. The small-ring tetraenes natamycin and lucensomycin also complied with this test (Table I). TABLE I AT THE SHOULDER AT 280 nm AND RATIOS OF THE ABSORBANCES AT THE MAXIMA AT 291, 318 AND 230 nm TO THE ABSORBANCE AT 304 nm FOR SAMPLES OF NYSTATIN (N), POLYFUNGIN (P), AMPHOTERICIN A (Amph. A), LUCENSOMYCIN (Luc) AND NATAMYCIN (Nat) RATIO O F THE ABSORBANCE AT THE MAXIMUM AT 230 nm TO THE ABSORBANCE Ratio of absorbances* Sample N2 . . N6 . . N9 . . N10 .. N11 .. N12 . . N13 . . N7 . . N8 . . N14 . . N15 . . N16 . . N17 . . N18 . . N19 . . N20 . . N3 . . N5 . . N1 . . N4 . . P1 .. P2 . . Amph. A Luc .. Nat . . . . . . .. .. . . . . . . . . . . . . .. .. . . .. . . . . . . . . . . . . .. .. .. . . .. . . . . .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . 7 230 nm 280 nm 1.14 0.90 1.15 0.97 0.83 0.88 0.98 0.98 0.93 0.95 0.89 0.86 0.92 0.93 1.03 1.04 0.98 1.08 0.87 0.95 1 .oo 1.04 0.97 0.65 0.68 291 nm 304 nm 0.65 0.67 0.67 0.66 0.65 0.66 0.65 0.66 0.67 0.68 0.68 0.66 0.66 0.68 0.66 0.69 0.66 0.65 0.66 0.66 0.67 0.66 0.66 0.64 0.63 318 nm 304 nm 0.86 0.89 0.88 0.89 0.89 0.89 0.89 0.90 0.89 0.88 0.90 0.89 0.89 0.90 0.88 0.89 0.89 0.89 0.89 0.92 0.90 0.89 0.90 0.92 0.91 7 230 nm 304 nm 0.38 0.33 0.47 0.34 0.27 0.29 0.34 0.35 0.35 0.41 0.35 0.31 0.35 0.37 0.38 0.45 0.37 0.39 0.33 0.33 0.37 0.35 0.35 0.21 0.21 * Official specifications for ratio of absorbance : 230 nm/280 nm: between 0.90 and 1.25 (CFR)’; 291 nm/304 nm: between 0.61 and 0.73 (BP)4; and 318nm/304nm: between 0.83 and 0.96 (BP).4852 THOMAS et al.: HETEROGENEOUS COMPOSITION Analyst, Vol. 107 However, these tetraenes may be distinguished from the large-ring nystatin by determining the ratio of the absorbance at 230 nm to that at 280 nm7 (Table I). The diene chromophore in nystatin is responsible for the maximum at 230nm; the diene is absent from the small- ring tetraenes. The present pharmacopoeia1 test would be improved by the inclusion of such a requirement or by the additional test that the ratio of the absorbance at the maximum of 230 nm to that at the maximum of 304 nm should be not less than 0.27. This would then exclude natamycin and lucensomycin. As previously demonstrated, TLC distinguished nystatin from natamycin and l~censomycin,~ and also showed the variation in the composition of the samples of nystatin and polyfungin (Fig.2). Nine different components were detected of which at least four were biologically active against the two test organisms. Most of the samples showed one principal component (nystatin A,) and two lesser components. The composition of some of the samples was markedly heterogeneous, consisting of as many as six components. The two samples of polyfungin consisted mainly of nystatin A, with at least four other components. The major constituents in the atypical samples of nystatin were tentatively identified as nystatins A, and A, and polyfungin B by reference to the components found in polyfungin.2t* HPLC confirmed the heterogeneity of the atypical nystatin samples.This method is the first to demonstrate unequivocally the difference between samples of nystatin from different sources of manufacture., HPLC was also used to determine the relative content of nystatin A, in each sample (Table 11). The two samples of polyfungin bore a closer resemblance t o commercial nystatin than did the atypical nystatin samples, which did not contain more than 12% of nystatin A,. The samples have been listed by source of origin (Table II), and their estimated potencies, A;zk values at 304 nm and the relative contents of nystatin A, determined by HPLC are shown. Valid bioassays were obtained with both test organisms; this was surprising as some TABLE I1 COMPARISON OF SAMPLES OF NYSTATIN (N) AND POLYFUNGIN (P) FROM DIFFERENT SOURCES Potencies determined by agar diffusion assay using Saccharomyces cerevisiae and Candida tropicalis, A;% a t 304 nm and the relative content of nystatin Ai deter- mined by HPLC.The precision of the biological assays met the requirement of the BP monograph for nystatin; the fiducial limits of error ranged from -& 1.3 to 54.6%. Source A .. .. B . . . . c .. . . Sample N2 N6 N9 N10 N11 N12 N13 N7 N8 N14 N15 N16 N17 N18 N19 N20 D .. .. N3 N5 E .. .. N l t N4 F . . . . P1 P1 Potency/IU mg-l* f--7---------- S. cerevzszae C . tropicalis 3000 (3093) - 4260 (4421) - 2950 (3029) 2637 (2742) 6 088 6 365 5 680 6012 4 495 4 599 5421 - 5 278 5 854 4 693 - 5 259 - 5051 3228 (3312) 3890 (4014) 2150 (2238) 5 302 3272 (3314) 4 707 2570 (2657) 4354 (4549) 2729 (2792) 4207 (4400) 2191 (2232) 4268 (4447) 4317 (4386) 4617 - 4 884 4 884 5519 - 5 657 - 5 726 6316 A ;y: at 304 nm 530 800 405 945 770 790 740 830 830 630 640 625 640 600 500 570 720 735 7 90 950 740 755 Nystatin A,, % 57.79 62.02 55.98 67.38 78.41 59.81 65.03 64.90 61.43 11.95 12.20 6.17 8.22 6.61 7.50 3.80 63.35 69.06 71.01 73.43 40.60 61.40 * The values in parentheses are the upper fiducial limits of the estimated potency.t Sample used as the standard for the biological assay: British Standard for Nystatin.852 THOMAS et al. : HETEROGENEOUS COMPOSITION Analyst, Vol. 107 However, these tetraenes may be distinguished from the large-ring nystatin by determining the ratio of the absorbance at 230 nm to that at 280 nm7 (Table I).The diene chromophore in nystatin is responsible for the maximum at 230nm; the diene is absent from the small- ring tetraenes. The present pharmacopoeia1 test would be improved by the inclusion of such a requirement or by the additional test that the ratio of the absorbance at the maximum of 230 nm to that at the maximum of 304 nm should be not less than 0.27. This would then exclude natamycin and lucensomycin. As previously demonstrated, TLC distinguished nystatin from natamycin and l~censomycin,~ and also showed the variation in the composition of the samples of nystatin and polyfungin (Fig. 2). Nine different components were detected of which at least four were biologically active against the two test organisms. Most of the samples showed one principal component (nystatin A,) and two lesser components. The composition of some of the samples was markedly heterogeneous, consisting of as many as six components.The two samples of polyfungin consisted mainly of nystatin A, with at least four other components. The major constituents in the atypical samples of nystatin were tentatively identified as nystatins A, and A, and polyfungin B by reference to the components found in polyfungin.2t* HPLC confirmed the heterogeneity of the atypical nystatin samples. This method is the first to demonstrate unequivocally the difference between samples of nystatin from different sources of manufacture., HPLC was also used to determine the relative content of nystatin A, in each sample (Table 11). The two samples of polyfungin bore a closer resemblance t o commercial nystatin than did the atypical nystatin samples, which did not contain more than 12% of nystatin A,.The samples have been listed by source of origin (Table II), and their estimated potencies, A;zk values at 304 nm and the relative contents of nystatin A, determined by HPLC are shown. Valid bioassays were obtained with both test organisms; this was surprising as some TABLE I1 COMPARISON OF SAMPLES OF NYSTATIN (N) AND POLYFUNGIN (P) FROM DIFFERENT SOURCES Potencies determined by agar diffusion assay using Saccharomyces cerevisiae and Candida tropicalis, A;% a t 304 nm and the relative content of nystatin Ai deter- mined by HPLC. The precision of the biological assays met the requirement of the BP monograph for nystatin; the fiducial limits of error ranged from -& 1.3 to 54.6%.Source A .. .. B . . . . c .. . . Sample N2 N6 N9 N10 N11 N12 N13 N7 N8 N14 N15 N16 N17 N18 N19 N20 D .. .. N3 N5 E .. .. N l t N4 F . . . . P1 P1 Potency/IU mg-l* f--7---------- S. cerevzszae C . tropicalis 3000 (3093) - 4260 (4421) - 2950 (3029) 2637 (2742) 6 088 6 365 5 680 6012 4 495 4 599 5421 - 5 278 5 854 4 693 - 5 259 - 5051 3228 (3312) 3890 (4014) 2150 (2238) 5 302 3272 (3314) 4 707 2570 (2657) 4354 (4549) 2729 (2792) 4207 (4400) 2191 (2232) 4268 (4447) 4317 (4386) 4617 - 4 884 4 884 5519 - 5 657 - 5 726 6316 A ;y: at 304 nm 530 800 405 945 770 790 740 830 830 630 640 625 640 600 500 570 720 735 7 90 950 740 755 Nystatin A,, % 57.79 62.02 55.98 67.38 78.41 59.81 65.03 64.90 61.43 11.95 12.20 6.17 8.22 6.61 7.50 3.80 63.35 69.06 71.01 73.43 40.60 61.40 * The values in parentheses are the upper fiducial limits of the estimated potency.t Sample used as the standard for the biological assay: British Standard for Nystatin.August, 1982 OF PHARMACEUTICAL-GRADE NYSTATIN 853 of the samples have such a different composition from the standard (Nl). The minimum potency for pharmaceutical-grade nystatin is that the upper fiducial limit of error of the estimated potency should not be less than 4400 IU mg-l; when assayed with S. cerevisiae most of the samples were found to meet this specification. However, when assayed with C. tropicalis the potency estimates of the samples containing less than 12% of nystatin A, were significantly lower than the results obtained using S .cerevisiae. There was no marked difference between the potency estimates obtained with the two assay organisms of the samples containing mainly nystatin A,. C. tropicaZis has been used outside the UK routinely for the assay of nystatin and it has the advantage that the edge of the zone of inhibition is sharper with this organism than with S. cerevisiae. The comparison of the two organisms was made on the same day, using the same solutions of “standard” and “test” for the two assays. The results obtained were reproducible. Only the atypical samples of nystatin (containing less than 12% of nystatin A,) were found to give consistently different results with the two assay organisms, the result obtained with C. tropicalis always being significantly lower. Bioautography of the chromatoplates showed that the principal components present in the atypical samples were less active than nystatin A, against C.tropicalis. Although the A;?; value may be used as a measure of the tetraene content of nystatin, no minimum value has been suggested in an official monograph. A research standard from a typical production lot of nystatin had a value of 866,5 but the lack of correlation between the Ai2$ value and potency, especially with degraded material,g and the marked hetero- geneity of some samples question the value of such a parameter. Nevertheless, samples with a low A;:& value (640 or less) did include all of the atypical samples and the low potency materials. A suitable minimum value based on the examination of a number of production batches of nystatin would be a useful check on the quality of the material. None of the current official identification tests for nystatin will differentiate it from other tetraene antibiotics or show that it consists largely of nystatin A,.The light absorption test can be improved to exclude natamycin and lucensomycin and the determination of the A!:k value at 304nm can be used as a qualitative test. The TLC method will identify nystatin and indicate its composition. Nystatin A, contains the amino-sugar mycosamine; the neutral sugar L-digitoxose has been isolated from nystatin A, and polyfungin B.6 An identification test for pharmacopoeial-grade nystatin based on the isolation and identification of these two carbohydrates was tried. Two extraction procedures were tested6,lo and, although digitoxose was recovered from the atypical samples of nystatin, the test was not sufficiently reproducible to be recommended as an official test.The only means of separating the components of nystatin and amphotericin A was by HPLC using a complicated gradient elution profile. Amphotericin A is unlikely to be produced on a commercial scale, so a simple linear gradient could be used to examine samples of nystatin. At present, we are using HPLC with a linear gradient increasing from 56.5 to 70% of methanol during 30 min for the assay of nystatin A, in samples of nystatin. Such a method would be the most suitable for assessing the quality of pharmaceutical-grade nystatin. The reduced activity exhibited by the atypical samples of nystatin against C. tropicalis suggests that their activity against some clinically important fungi and yeasts should be determined.Conclusions Pharmaceutical-grade nystatin has been shown to be a complex mixture having a very varied composition. The existing pharmacopoeia1 specification is inadequate to detect the wide differences in the composition of nystatin produced by different manufacturers. A method to confirm that pharmaceutical-grade nystatin has a composition similar to that previously used and perhaps even to determine the content of nystatin A, will be required if it is found that such material is a more effective antifungal agent than the atypical material. References 1. 2 . Shenin, Y., Kotienko, T. W., and Ekzemplyarow, 0. N., Antibiotiki, 1968, 13, 387. Porowska, N., Halski, L., Plociennik, Z . , Kotiuszko, D., Morawska, H., Kowszyk-Gindifer, Z . , and Bojarska-Dahling, H., A d v . Antimicvob. Antineoplast. Chewother., 1972, 1, 1031.854 3. 4. 5. 6. 7. 8. 9. 10. THOMAS, NEWLAND AND SHARMA Thomas, A. H., Newland, P., and Quinlan, G. J . . J. Chrornatogr., 1981, 216, 367. “British Pharmacopoeia 1980,” HM Stationery Office, London, 1980, p. 313. Michel, G. W., Anal. Projiles Drug Subst., 1977, 6, 341. Zielinski, J., Jereczek, E., Sowinski, P., Falkowski, L., Rudowski, A., and Borowski, E., J . Antibiot., “Code of Federal Regulations,” Title 21 (1979), Part 449, US Government Printing Office, Washing- Plociennik, Z., Halski, L., and Kowszyk-Gindifer, Z., Chem. Anal. (Warsaw), 1973, 18, 151. Thomas, A. H., Analyst, 1976, 101, 321. Brik, H., J. Antibiot., 1976, 29, 632. 1979, 32, 565. ton, DC, 1980, p. 593. Received February 3rd. 1982 Accepted March llth, 1982

ISSN:0003-2654    

DOI:10.1039/AN9820700849

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

Analyst, August, 1982, Vol. 10'7, pp. 855-866 855 Gas-chromatographic Determination of Concentrations of Trace Amounts of 46 Odorants Prepared in Air in a 10-m3 Stainless-steel Odour Test Room Yasuyu ki Hoshi ka * Aichi Environmental Research Centre, 7-6 Nagare, Tsuji-machi, Kita-ku, Nagoya-shi, A ichi 462, Japan and Giichi Muto Saitarna Institute of Technology, 1690 Fusai-ji, Okabe-machi, Osato-gun, Saitarna 369-02, Japan The determination of the concentrations of trace amounts of 46 odorants prepared at odour threshold levels (in parts per billion) in air in a 10-m3 stainless-steel odour test room was investigated by gas chromatography using two pre-concentration methods, namely cold trapping with liquid oxygen and room-temperature adsorption trapping with porous polymer beads, such as Tenax-GC or graphitised carbon black coated with a small amount of stationary phase (SP-1000 + orthophosphoric acid) (0.3 + 0.3% on Carbo- pack B, 60-80 mesh).The 46 odorants tested were four sulphur compounds, ten lower aliphatic carbonyl compounds, an aromatic hydrocarbon, seven lower aliphatic mono-alcohols, eleven phenols, six lower aliphatic carboxylic acids and seven indoles. The recoveries of the odorants having boiling- points of less than about 150 "C were quantitative, but those having boiling- points of more than 160 "C gave recoveries of about 50%, except for the phenols, which had much lower recovery levels. Keywords : Odour test room ; odorant determination ; pre-colum?z concentra- tion ; air analysis ; gas chromatography The data for odour threshold values of odorants in air in odour test rooms vary widely, depending on factors such as the test procedure and technique, the observer and the purity of the odorants.The test odorants used must be of high purity, and a static air dilution system utilising an air dilution medium that has a low background level of odours must be employed. The determination of the odour threshold value of an odorant in the air is subject to potentially high errors resulting from adsorption of the odorant on to the surface of the apparatus and from diffusion losses. Of the experimental parameters used when measuring odour threshold values, the determination of actual concentrations [in parts per million or parts per billion ( l o g ) ] of odorants in odour test room air is one of the most import ant.Leonardos et aZ.l used an odour test room which had a volume of 13.2 m3 to determine the odour threshold values of 53 odorants. Smith and Hochstettler2 reported the determination of odour threshold values in air using carbon-14 labelled compounds and scintillation counting to monitor concentrations. Mills et aZ.3 also used an odour-free room for quantitative odour measurements. However, in their work, the actual concentrations of the odorants in the odour test room were not determined by reliable methods. Fluck4 used an odour test room that had a volume of approximately 28.2 m3 and a low odour background to determine the odour threshold of phosphine. Concentrations of the phosphine were determined by using detector tubes.Whisman et aL5 evaluated ethanethiol and tetrahydrothiophene as odorants in propane, measuring concentrations in air in four testing modes by gas chromatography. However, there are few reports on the determinations of the actual concentrations of a wide variety of odorants in an odour test room and this paper describes the determination of the actual concentrations of 46 odorants prepared in air in a 10-m3 stainless-steel odour test room. * Present address : Department of Hygiene, Shinshu University, School of Medicine, 3-1-1 Atahi, Matsumoto-shi, Nagano 390, Japan.856 HOSHIKA AND MUTO: GC DETERMINATION OF TRACE AMOUNTS Analyst, VoZ. 107 Experimental Odour Test Room This test room (Fig. 1) is of the five window type, and was installed in Aichi Environ- mental Research Centre in 1972. It is made of stainless steel (suspended stainless-steel ceiling, polished, No. 300) and has a volume of approximately 10 m3 (2 100 x 2200 x 2 160 mm).It is equipped with a mixing fan (1 720 rev min-l), ducts and activated carbon filters for removing odorised air (ca. 30 m3 min-l) and for introducing a controllable odour- free background of air (2830 m3 h-l) that serves as a diluent. Incoming air that is used to purge the room is passed through a bank of activated carbon filters and circulated through the antechamber and the test room by a booster. The working area serves as a buffer between the odour test room and the external equipment and is also used to acclimatise the member of the odour panel to the relatively odour-free conditions prior to each test exposure.Reagents purity) were obtained from Matheson Gas Products, East Rutherford, N J, USA. Hydrogen sulphide (of 99.6% minimum purity) and methanethiol (of 99.5% minimum Standard ' 2.060 m 2.160 rn _ I (1 1 (2) F - Aiy in i - Air out Fig. 1 . 1, Side table: 2, working area (6400 x 5400 x 2700 mm); 3, experimental table (2400 x 1500 mm); 4, scullery: 5, working area for odour test room (3200 x 5400 x 2700mm); 6, odour test room (2100 x 2200 x 2160 mm, about 10 m3, suspended stainless-steel ceiling, polished, No. 300) ; 7 , store room (3250 x 5400 x 2 700 mm) ; 8, blower room (for supply a t 2 830 m3 h-' and for exhaust a t 5.0 m3 min-l in odour test room and a t 28.5 m3 min-1 in working area). ( b ) Schematic diagram of- 10-m3 stainless-steel odour test room (five- window type) : (1) side view; and (2) front view.A, Mixing fan ( 1 720 rev min-') ; B, fluorescent lamp (30 W, two per room) ; C , sprinkler, steam, 0.5 kg ~ m - ~ , 1 1 min-l, 16 chips per one room; D, odour test room doorway; and E, sampling port. (c) Schematic diagram of air flow in 10-m3 stainless-steel odour test room. A, Activated carbon filter box; B, booster (air supply fan, 2830 m3 h-l); and C, exhaust blower (for odour test room, 5.0 m3 min-l). (a) Schematic diagram of odour test room and other rooms.August, 1982 OF 46 ODORANTS PREPARED IN AN ODOUR TEST ROOM 857 solutions of each were prepared by dissolving the pure gas (50 ml) in 50 ml of ethanol or benzene. The hydrogen sulphide solution was standardised by iodimetric titration, and the methanethiol solution was standardised by titration with standard silver nitrate solution.Dimethyl sulphide (99.9%) and dimethyl disulphide (99.9%) were obtained from Wako Pure Chemical Industries Ltd., Osaka, Japan, and standard solutions containing about 2-100 pg pl-1 were prepared by dissolving the compounds in ethanol. Acetaldehyde (purity 99.5%) was obtained from E. Merck (Darmstadt, West Germany), propanal, propenal, but anal, 2-met h ylpropanal, pent anal, 3-met hylbut anal and pent an-3-one from Tokyo Kasei Kogyo, Tokyo, Japan, and butan-2-one from Katayama Chemical Industries Ltd., Osaka, Japan. All reagents were of guaranteed or analytical-reagent grade. -4 standard solution containing a mixture of the ten lower aliphatic carbonyl com- pounds was prepared by dissolving 100 p1 of each in 1-10 ml of distilled water.Styrene, minimum assay 95%, stabilised with 0.003-0.004~0 of 4-tert-butylcatechol (chemical grade) was obtained from Wako Pure Chemical Industries Ltd. A standard solution was prepared by dissolving 100 pl in 1 nil of ethanol. The seven lower aliphatic mono-alcohols were obtained from PolyScience Corp., Gross Point Road, Niles, IL, USA, Kit No. 11A. A standard mixture containing 0.1 ml of each reagent was prepared. The eleven phenols were obtained from Wako Pure Chemical Industries Ltd. or Tokyo Kasei Kogyo, and a standard solution was prepared by dissolving 1 g of each in 100 ml of ethanol. The six lower aliphatic carboxylic acids were obtained from PolyScience Corp., Katayama Chemical Industries Ltd.or Tokyo Kasei Kogyo. A standard solution containing 0.2 ml of each acid was prepared. The seven indoles were obtained from Aldrich Chemical Co., Milwaukee, WI, USA, and a standard solution was prepared by dissolving 17.6-100mg of each compound in 10ml of ethanol. Standard Sample Gas Preparation Method in the 10-m3 Stainless-steel Odour Test Room All of the five windows of the test room were closed and the sampling port [E in Fig. l(b)] was sealed with a silicone-rubber septum. The exhaust dumper was opened fully, and the air inlet dumper was half opened. The mixing fan was then turned off, and the air in the test room was allowed to come to rest. Upon completion of the analysis the test room and ante- chamber were purged for 20 min with air that had been filtered through activated carbon.The booster fan was then turned off and the dumpers were closed for 2min, after which 0.2-50 1 of the air in the odour test room (background gas sample) was transferred via the sampling port directly on to either the cold-trapping pre-column, which was cooled with liquid oxygen, or the adsorption-trapping pre-column, which was at room temperature (Fig. Samples of the standard solutions were introduced into the test room through a door, D [Fig. l(b)]. The standard solutions (3-600 pl) were injected via microsyringes (10, 100, 500 or 1000 pl, as appropriate) on to watch-glasses (9 cm diameter), which were placed in the centre of the odour test room. As soon as the door had been closed the watch-glass was heated at 100 "C for the lower aliphatic mono-alcohols, or at 200 "C for the other odorants, except for the sulphur compounds, which were left at room temperature. The mixing fan was turned on for 2 min then 0.2-50 1 of the sample gas was transferred via the sampling port directly on to the pre-columns using a vacuum pump (Mini-Vac, hlodel PS-5, Yamato, Japan, maximum flow-rate 5 1 min-l), the sample flow-rate being monitored by a gas meter (T-3 Dry Test gas meter, Chubushinagawa, Seisakusho, Japan, 1 1 rev-l).2). The air in the test room served as the dilution medium. Operating Conditions for Gas Chromatography The gas chromatograph used was a Shimadzu, Model GCSAP,FFp, equipped with a flame- ionisation detector (FID), a flame-photometric detector (FPD) and a digital integrator (Shimadzu, Model ITG-2A) for quantitative analysis and retention time determinations. The operating conditions are given in Table I.858 HOSHIKA AND MUTO: GC DETERMINATION OF TRACE AMOUNTS Analyst, VoZ.I07 Fig. 2. Schematic diagram of gas sampling procedure in 10-m3 stainless-steel odour test room. 1, Ceiling of odour test room; 2, silicone septum, No. 14; 3, Pyrex glass tube, 10 cm x 5 mm i.d. x 7 mm 0.d.; 4, PTFE tube, 5 cm x 1 mm i d . ; 5, pre-column for cold trapping; 6, liquid oxygen bottle; 7, vacuum pump (Mini-Vac, Model PS-05, Yamato, Japan, maximum rate 5 1 min-l) ; 8, gas meter (T-3 Dry Test gas meter, Chubushinagawa Seisakusho, Japan, 1 1 rev-l) ; and 9, pre-column for adsorption trapping. Sample vapours of the four sulphur compounds, the ten lower aliphatic carbonyl com- pounds and the seven lower aliphatic mono-alcohols were collected directly in the pre- column by cold trapping.The sampling rate was about 0.2 1 min-l. The sample vapours of the 11 phenols and the six lower aliphatic carboxylic acids were collected directly in the adsorption trapping pre-column at room temperature (ca. 25 "C) at a sampling rate of about 0.2 1 min-l. For the seven indoles, 50 1 of the vapour were collected at a rate of 5 1 min-l on Tenax-GC packing material contained in a pre-column (3 cm x 8 mm i.d., glass). This packing material was then transferred into the analytical pre-column (14 cm x 4 mm i.d., glass). Identification of all of the odorants was carried out by comparison of their retention times with those of the authentic compounds, and the concentrations of the odorants were obtained from calibration graphs prepared using the authentic compounds.The gas-chromatographic column packings were purchased from Wako Pure Chemical Industries Ltd. Some of the details of the methods for injecting samples concentrated by cold and adsorp- tion trapping into the gas chromatograph have been reported in previous Sensory Testing This was carried out by sniffing directly the odorous gases prepared in the odour test room. A trained panel of four members from a division of the Aichi Environmental Research Centre was used on alternative days for determining odour quality and odour intensity. Results and Discussion Typical Gas Chromatograms for 46 Odorants in the 10-m3 Stainless-steel Odour Test Room Air Fig.3 shows typical gas chromatograms of the four sulphur compounds and the back- ground air sample obtained from the odour test room. As shown in Fig. 3(a), four sulphur compounds, i.e., carbonyl sulphide, hydrogen sulphide, carbon disulphide and sulphur dioxide, were detected in the background air sample. Hoshika et aL8 have reported the gas-chromatographic determination (using an FPD) of trace concentrations of these four sulphur compounds at parts per billion levels in urban air (in the Nagoya area) using theTABLE I OPERATING CONDITIONS FOR GAS CHROMATOGRAPHY Conditions. . . . Compound . . Pre-concentration Packing - Main column . . 1 2 7 4 5 6 7 8 . . Sulphur compounds Carbonyl compounds Acetaldehyde Styrene Alcohols Phenols Carboxylic acids Indoles .. Cold trapping Cold trapping Cold trapping Adsorption trapping Cold trapping Adsorption trapping Adsorption trapping Adsorption trapping 0 r . . ‘ 20 r o /o TCEP on Shimalite (AW. DMCS). 5% TCEP on Carbopack B, 60-80 mesh 0.476 Triton X-100 on Carbopack B, 60-80 mesh 594 SP-1200 + 1.75y0 Reritone 34 on 0.5% PEG-1500 + 0.2y0 KOH on Carbopack B, 60-80 mesh 0.1% SP-1000 on Carbopack C, 80-100 mesh 0.3% FFAP + 0.3% H,PO, on Carbopack B, 60-80 mesh 1% Silicone XE-60 W on (AW, Chrornosorb DMCS), 80-100 mesh 60--Xb mesh ” Chromosorb W (AW, DMCS), 60-80 mesh Tenax-GC, 60-80 mesh Pre-column . . (a) 25% TCEP on Shimalite (AW, mesh (6) Tenax-GC, 60-80 mesh DMCS), 60-80 25% TCEP on Shimalite (AW, mesh DMCS), 60-80 25% TCEP on Shimalite (AW, DMCS), 60-80 mesh 2% NaOH on glass beads, 30-60 mesh Tenax-GC, 60-80 mesh 0.3% SP-1000 + 0.3% H,PO, on Carbopack B, 60-80 mesh Tenax-GC, 60-80 mesh Column size- Main column (glass) Pre-column (glass) . .3 m x 3mm i.d. . . ( a ) 31 cm x 4 mm ( b ) 18cm x 4mm i.d. i.d. 1.5 m x 3 mrn i.d. 31 cm x 4 mm i.d. 1.5 m x 3 mm i.d. 31 cm x 4 mm i.d. 1.5 m x 3 mm i.d. 31 cm x 4 mm i.d. 3 m x 3 mrn i.d. 18 cm x 4 mm i.d. 1.75 m x 3 mm i.d. 18 cm x 4 mm i.d. 1.5 m x 3 mm i.d. 18 cm x 4 nun i.d. 3 m x 3 mm i.d. 3cm x 8mmi.d. and 14 cm x 4 mm i.d. Temperature/T- Main column . . Pre-column . . ‘75. no 70 -183 to 100 (2.5 min) 150 65 50 1000 FID 75 - 153 to 100 (’2.5 min) 150 65 50 1000 FID 80,95 25 to 200 (24 s) 135 - 183 to 100 (2.5 min) 220 25 to 250 (35 s) 250 200 25 to 200 (24 s) 170 25 to 280 (50 s) .. (a) --18:3’t0 ioo (2.5 min) ( b ) 25 to 800 (24 s) Injection and detector 150 (200) 150 150 250 200 Carrier gas- Nitrogm/ml min-l . . 65 Hydrogen/ml min-l . . 40 Air/ml mir1-l . . . . 40 Detector . . . . . . FPD 65 50 1000 FID 65 50 1000 FID 37 50 1000 FID 65 50 1000 FID 50 50 1000 FID860 HOSHIKA AND MUTO: GC DETERMINATION OF TRACE AMOUNTS AndySt, ~ O I ? . 107 0 5 0 5 Ti m e m i n 0 5 10 0 5 10 15 Ti m el ni in Fig. 3. Typical gas chromatograms of ( a ) background sample air, (b) four sulphur compounds, (c) background sample air and ( d ) dimethyl disulphide in air in 10-m3 stainless- steel odour test room. (a) Background sample air: 1, carbon dioxide; 2, carbonyl sulphide; 3, hydrogen sulphide: 4, carbon disulphide; and 5 , sulphur dioxide.FPD sensitivity, 4 x 10; concentrated volume, 0.2 1. Other gas-chromatographic conditions as in Table I, l ( a ) . (b) Sulphur compounds: 1, Carbon dioxide: 2, carbonyl sulphide; 3, hydrogen sulphide (detected concentration 32 p.p.b.) ; 4, carbon disulphide; 5, methanethiol (5 p.p.b.) ; and 6, dimethyl sulphide + sulphur dioxide (62 p.p.b. as dimethyl sulphide). FPD sensi- tivity, 32 x 10; concentrated volume, 0.1 1. Other gas-chromatographic conditions as in Table I, l ( a ) . (c) Background sample air: 1, carbon dioxide; 2, carbonyl sulphide; 3, hydrogen sulphide ; 4, carbon disulphide; and 5 , sulphur dioxide. FPD sensitivity, 16 x 10; concentrated volume, 11. Other gas-chromatographic conditions are as in Table I, l ( b ) . (d) Dimethyl disulphide: 1, carbon dioxide; 2, carbonyl sulphide; 4, carbon disulphide ; 5, sulphur dioxide ; 6, dimethyl disulphide (detected concentration 19.5 p.p.b.).FPD sensitivity, 16 x 10; concentrated volume, 0.5 1. cold-trapping method with liquid oxygen cooling. The detected concentration ranges and average values for the four sulphur compounds were as follows: carbonyl sulphide, 0.4-2.9 p.p.b. (average 1.2 p.p.b) ; hydrogen sulphidt:, 0.5-24 p.p.b. (average 2.4 p.p.b.) ; carbon disulphide, 0.1-75 p.p.b. (average 2.7 p.p.b.) ; and sulphur dioxide, 2.0-80 p.p.b. (average 22 p.p.b.). The percentage recoveries of the four sulphur compounds are listed in Table 11. The recoveries of hydrogen sulphide, methanethiol and dimethyl sulphide were quantitative, whereas that of dimethyl disulphide was about 60%.A faint odour of rotten eggs or decayed cabbage was perceived by olfaction in the odour test room. Fig. 4 shows typical gas chromatograms of the ten lower aliphatic carbonyl compounds and the background air sample obtained from the odour test room. It can be seen that the carbonyl compounds were completely separated within about 20 min, except for pentanal and pentan-3-one, the peaks of which overlapped. However, propanal, propenal and acetone, each of which contain three carbon atoms, and also 2-methylpropanal were separated completely within about 6 min. The percentage recoveries of the ten compounds are listed in Table I1 and can be seen to be quantitative. A strong, pungent odour was perceived by olfaction in the odour test room. Fig.5 shows typical gas chromatograms of acetaldehyde and the background air sample obtained from the odour test room. As shown in Fig. 5(a), acetaldehyde was present in the background air sample. The detected concentration was 13.8 p.p.b. Hoshikag has reported ranges and average concentrations of acetaldehyde in 13 samples of urban air (from the Nagoya area) of 1.5-9.6 p.p.b. and 4.7 p.p.b., respectively.August, 1982 OF 46 ODORANTS PREPARED IN AN ODOUR TEST ROOM TABLE I1 861 RECOVERY DATA FOR ODORANTS IN ODOUR TEST ROOM AIR Number of replicate analyses performed, 5-6; temperature, 25 & 1 "C; and relative humidity, 50 & 10%. Compound Sulphur compounds- Hydrogen sulphide Methanethiol Dimethyl sulphide Dimethyl disulphide Carbonyl compounds- Acetaldehyde Propanal Propenal Acetone 2-methyl propanal Butanal Ethyl methyl ketone 3-Methylbu tanal (pentanal + diethyl ketone) Acetaldehyde Styrene A Icohok- Ethanol Propan-1-01 Propan-2-01 Butan-1-01 Butan-2-01 Pentan-1-01 Pentan-2 -01 Phenols- Phenol o-Cresol m-Cresol p-Cresol o-Ethylphenol p-E thylphenol 2,6-Xylenol 2,5-Xylenol 2,3- + 3,5-Xylenol 3,4-Xylenol Concentration, p.p.b.Average Coefficient GC Concentration of standard solution ,-'--, recovery, of variation, conditions and amount taken Calculated Found 70 Yo - - 1 (a) and ( b ) 800 ng pl-I in water, 600 p1 34.4 32.9 95 - 2 700 ng pl-l in benzene, 40 pl 5.6 4.8 85 12 500 ng p1-1 in benzene, 120 p1 58.9 62.4 106 100 vg pl-1 in benzene, 10 pi 29.7 19.1 64 - 86.3 77.4 1.2 89.7 1.6 3.1 68.0 74.0 f 2.3 109 73.0 66.3 65.6 f 3.9 98.9 5.9 56.3 f 1.1 104 2.0 56.9 * 0.7 103 1.2 5554:; 68.0 67.7 f 3.7 99.6 5.5 44.2 44.0 f 2.5 99.5 5.7 (45.9 + 45.9) 93.2 & 4.8 102 5.2 100 p1 ml-I in water, 6 p1 25.9 23.5 f 0.3 90.7 1.3 100 ~1 ml-' in ethanol, 20 pl 42.6 42.0 f 3.4 98.6 8.1 - - - 100 wl per 10 ml in water, 200 p~ 41.9 41.6 0.1 99.3 2.4 32.6 30.4 * 0.0 93.3 0.0 31.9 30.4 f 0.1 95.3 0.3 5 Standard mixture containing 0.1 ml 26.7 23.1 0.0 86.5 0.0 of each alcohol, 7 pl 26.5 22.5 k 0.0 84.9 0.0 22.6 18.9 f 0.0 83.6 0.0 22.5 17.2 i 0.0 76.4 0.0 103 14.0 f 1.4 13.6 10.0 90.0 12.4 1.1 13.8 8.9 90.0 9.6 & 1.0 10.7 10.4 90.0 12.6 f 1.1 13.8 8.9 79.8 6.6 * 0.7 8.3 10.6 79.8 9.5 0.6 11.9 6.3 79.8 14.3 f 1.2 17.9 8.4 79.8 9.6 i 0.7 12.0 7.3 79.8 10.5 f 1.1 13.2 10.5 79.8 7.1 f 0.7 8.9 9.9 6 1 g per 100 ml in ethanol, 400 pl Caiboxylic acids- > 21.3 23.8 -C 5.5 112 22.3 16.4 17.3 4.3 105 24.9 10.6 f 1.2 79.1 11.3 7 Standard mixture containing 0.2 ml tg:; 7.8 -j= 0.8 58.6 10.3 of each acid, 3 p1 11.1 7.0 -j= 0.2 63.1 2.9 11.2 4.2 & 0.3 37.5 7.1 Acetic acid Propanoic acid 2-Methylpropanoic acid Butanoic acid 3-Methylbutanoic acid Pentanoic acid lndoles- Indole 0.43 0.27 f 0.01 62.8 3.7 3-Methylindole 0.33 0.19 & 0.01 57.6 5.3 2-Methylindole 0.32 0.18 & 0.02 56.3 11.1 8 17.6-100 mg per 10 ml in ethanol, 0.42 0.24 & 0.02 57.1 8.3 0.19 0.13 f 0.01 68.4 7.7 0.41 0.29 f 0.02 70.7 6.9 5-Methylindole 1,2-Dimethylindole 2,3- + 2,5-Dimethyl- * From Table I.indoles Fig. 5(b) shows a typical gas chromatogram illustratifig the disappearance of the acetalde- hyde peak with the use of a water trap (2 "C).Fig. 5(c) shows a typical gas chromatogram of acetaldehyde from the odour test room. The detected concentration was 35.9 p.p.b. The percentage recoveries of acetaldehyde are listed in Table I1 and can be seen to be quantitative. Fig. 6 shows typical gas chromatograms of styrene and the background air sample obtained from the odour test room. The percentage recoveries of styrene are listed in Table I1 and can be seen to be quantitative. A faint, solvent-like odour was perceived by olfaction in the odour test room. A faint pungent odour was perceived by olfaction in the odour test room. The detected concentration was 40.2 p.p.b.862 HOSHIKA AND MUTO: GC DETERMINATION OF TRACE AMOUNTS Analyst, Vol.107' I 3 10 20 Tirne/mi n Fig. 4. Typical gas chromatograms of (a) background sample air and (b) 10 lower aliphatic carbonyl compounds in air in 10-m3 stainless-steel odour test room. (a) Background sample air: FID sensitivity, 2 x lo2; concentrated volume, 1 1. (b) Lower aliphatic carbonyl compounds : 1, acetaldehyde (detected concentration 75.8 p.p.b.); 2, propanal (73.0 p.p.b.); 3, propenal; 4, acetone (62.8 p.p.b.); 5, 2- methylpropanal (54.1 p.p.b.) ; 6, butanal (50.2 p.p.b.) ; 7, butan-2-one (54.1 p.p.b.) ; 8, 3-methylbutanal (42.9 p.p.b.) ; and 9, pentanal + pentan-3-one (90.5 p.p.b.). FID sensitivity, 2 x lo2; concentrated volume, 1 1. Other gas-chromatographic con- ditions as in Table I , 2. u 0 2 4 6 0 2 4 6 Ti me/m in L 0 2 4 6 Fig. 5. Typical gas chromatograms of background sample air (a, b) and acetalde- hyde (c) in air in 10-m3 stainless-steel odour test room. (a) Background sample air: 1, acetaldehyde (detected concentration 13.8 p.p.b.). FID sensitivity, 2 x lo2; con- centrated volume, 11.( b ) Background sample air: concentrated volume 1 1, but passed through a water trap (2 "C). FID sensitivity, 2 x lo2. (c) Acetaldehyde: 1, acetaldehyde (detected concentration 35.9 p.p.b.). FID sensitivity, 2 x lo2; con- centrated volume, 1 1. Other gas-chromato- graphic conditions as in Table I, 3.August, 1982 OF 46 ODORANTS PREPARED I N AN ODOUR TEST ROOM 863 I I 0 5 10 Time min Fig. 6. Typical gas chroma- tograms of (a) background sample air and (b) styrene in air in 10-m3 stainless-steel odour test room. (a) Background sample air: FID sensitivity, 16 x lo2; concentrated volume, 1 1.(b) Styrene: large peak (detected concentration 40.2 p.p.b.). FID sensitivity, 16 x lo2; concentrated volume, 11. Other gas-chromatographic con- ditions as in Table I, 4. L 0 5 10 Time)min -I- 15 Fig. 7. Typical gas chromato- grams of (a) background sample air and (b) seven lower aliphatic mono-alcohols in air in 10-m3 stainless-steel odour test room. (a) Background sample air: FID sensitivity, 4 x lo2; concentra- ted volume 1 1. ( b ) Lower ali- phatic mono-alcohols : 1 , ethanol (detected concentration 42 p.p.b.) ; 2, propan-2-01 (30p.p.b.) ; 3, propan-1-01 (31 p.p.b.); 4, butan-2-01 (22 p.p.b.) ; 5, butan- 1-01 (23 p.p.b.); 6, pentan-2-01 (17 p.p.b.); and 7, pentan-1-01 (18 p.p.b.). FID sensitivity, 4 x lo2; concentrated volume, 1 1. Other gas-chromatographic conditions as in Table I, 5.Fig. 7 shows typical gas chromatograms of seven lower aliphatic mono-alcohols and the background air sample obtained from the odour test room. The percentage recoveries are listed in Table I1 and again can be seen to be quantitative. A faint, sweet odour was perceived by olfaction in the odour test room. Fig. 8 shows typical gas chromatograms of the 11 phenols and the background air sample obtained from the odour test room. Complete separation occurred within 18 min with no tailing, except for 2,3- and 3,5-xylenols, the peaks of which overlapped. I t is particularly noteworthy that 0-, m- and 9-cresols were separated completely within about 7 min. The percentage recoveries of these 11 phenols are listed in Table 11, and show the recoveries of these compounds to be poor.,4 strong odour of disinfectant was perceived by olfaction in the odour test room. Fig. 9 shows typical gas chromatograms of the six lower aliphatic carboxylic acids and the background air sample obtained from the odour test room. The percentage recoveries of acetic acid, propanoic acid and 3-methylpropanoic acid were quantitative ; those of butanoic acid and pentanoic acid were low (Table 11). A strong rancid or goaty odour was perceived by olfaction in the odour test room.864 HOSHIKA AND MUTO : GC DETERMINATION OF TRACE AMOUNTS Analyst, VoZ. 107 0 10 20 Ti me, m i n Fig. 8. Typical gas chromatograms of (a) background sample air and (b) 11 phenols in air in 10-m3 stainless-steel odour test room.(a) Background sample air: FID sensitivity, 2 x 102; concentrated volume, 1 1. (b) Phenols: 1, phenol (detected concentration 15.5 p.p.b.); 2, o-cresol (13.7 p.p.b.); 3, m-cresol (12.2 p.p.b.); 4, p-cresol (13.5 p.p.b.); 5, p-ethylphenol (8.4 p.p.b.); 6, o-ethylphenol(9.6 p.p.b.) : 7, 2,6-xylenol (14.6p.p.b.) ; 8,2,5-xylenol(lO.4 p.p.b.) ; 9, 2,3-xylenol + 3,5-xylenol (10.8 p.p.b.); and 10, 3,4-xylenol (7.4 p.p.b.). FID sensitivity, 2 x lo2; concentrated volume, 1 1. Other gas-chromatographic conditions as in Table I, 6. 0 5 10 Ti me, m i n Fig. 9. Typical gas chroma- tograms of (a) background sample air and (b) six lower fatty acids in air in 10-m3 stainless-steel odour test room. (a) Background sample air: FID sensitivity, 8 x lo3; concentra- ted volume, 1 1.(b) Lower fatty acids: 1, acetic acid (detected concentration 20.3 p.p.b.) ; 2, propanoic acid (14.8 p.p.b.); 3, 3-methylpropanoic acid (9.0 p.p.b.); 4, butanoic acid (5.5 p.p.b.) ; 5, 3-methylbutanoic acid (6.5 p.p.b.) ; 6, pentanoic acid (4.3 p.p.b.). FID sensitivity, 8 x lo3; concentrated volume, 1 1. Other gas-chromatographic conditions as in Table I, 7. Fig. 10 shows typical gas chromatograms of the seven indoles and the background air sample obtained from the odour test room. The seven indoles were separated within about 14 min without tailing, except for 2,3- and 2,5-dimethylindoles, whose peaks overlapped. The percentage recoveries of the indoles were about 60-70y0 (Table 11). A strong faecal odour was perceived by olfaction in the odour test room.Percentage Recovery versus Boiling-point The percentage recoveries of the 46 odorants prepared in the 10-m3 stainless-steel odour test room air are listed in Table 11. Replicate determinations of the concentrations gave coefficients of variation of less than 11%, except for acetic acid and propanoic acid, which had values of more than 20%.A ugust , 1982 O W $ 0 5 10 15 Time' mi n Fig. 10. Typical gas chro- matograms of (a) background sample air and ( b ) seven indoles in air in 10-m3 stainless-steel odour test room. (a) Back- ground sample air: FID sensi- tivity, 4 x 103; concentrated volume, 50 1. (b) Indoles: 1, 1,2-dimethylindole (detected concentration 0.14 p.p.b.) ; 2, indole (0.28 p.p.b.) ; 3, 3-methyl- indole (0.20 p.p.b.) ; 4, 5-methyl- indole (0.26 p.p.b.) ; 5, 2-methyl- indole (0.20 p.p.b.); 6, 2,3- dimethylindole + 2,5-dimethyl- indole (0.31 p.p.b.). FID sensi- tivity, 4 x lo3; concentrated volume, 50 1. Other gas-chro- matographic conditions as in Table I, 8. OF 46 ODORANTS PREPARED I N AN ODOUR TEST ROOM 10 I 0 0 0 0 0 0 0 0 O O 0 0 0 O O 0 0 0 0 0 0 0 0 00 0 0 0 865 Fig. 11. Percentage recovery vevsus boiling-point.866 HOSHIKA AND MUTO The relationship between the percentage recoveries of the 46 odorants and their boiling- points is shown in Fig. 11. Percentage recoveries of the odorants that had boiling-points of less than ca. 150 “C were quantitative, but odorants that had boiling-points of more than 160 “C gave recoveries of only about SO%, except for the phenols, which gave even poorer results. The authors thank Dr. K. Yoshimoto, Aichi Environmental Research Centre, for useful suggestions. 1 . 2. 3. 4. 5 . 6. 7 . 8. 9 . References Leonardos, G., Kendall, D., and Barnard, N., J . Air Pollut. Control Assoc., 1969, 19, 91. Smith, H. O., and Hochstettler, A. D., Environ. Sci. Technol., 1969, 3, 169. Mills, J . L., Walsh, R. T., Leudtke, K. D., and Smith, L. K., “Quantitative Odor Measurement,” paper presented at the Air Pollution Control Association 56th Annual Meeting, Sheraton- Cadillac Hotel, Detroit, MI, June, 1963. FIuck, E., J . Air Pollut. Control Assoc., 1976, 26, 795. Whisman, M. L., Goetzinger, J. W., Cotton, F. O., and Brinkman, D. W., Environ. Sci. Technol., Hoshika, Y., Analyst, 1981, 106, 166. Hoshika, Y . , Analyst, 1981, 106, 686. Hoshika, Y., Kozima, I., Koike, K., and Yoshimoto, K., Bunseki Kagaku, 1974, 23, 1393. Hoshika, Y . , J . Chromatogr., 1977, 137, 455. 1978, 12, 1285. Received December 29th, 1981 Accepted March l l t h , 1982

ISSN:0003-2654    

DOI:10.1039/AN9820700855

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

Analyst August 1982 VoZ. 107 pp. 867-871 867 Demountable Double-chamber Hollow-cathode Lamp A New Approach to the Determination of Trace Elements in Steel Bo Berglund and Bo Thelin" Sandvik A B S-811 81 Sandviken Sweden Swedish Institute f o r Metals Research Drottning Kristinas vag 48 S- 114 28 Stockholm Sweden A new type of excitation source for emission spectrocheinical analysis of steel has been constructed and studied. The source is a water-cooled demountable hollow-cathode lamp; a tubular sample forms the cathode and can be sputtered from both ends. The detec-tion limits for the elements studied aluminium arsenic boron calcium, cobalt chromium manganese molybdenum niobium nickel lead silicon, tin titanium and tungsten were found to be below 1 pg g-l (except for cobalt and silicon).It was also found that the lamp could be used to determine concentrations of up to several per cent. for some elements and the relative standard deviations were in the range 1-3%. The lamp makes a very interesting complement to the instrumental techniques normally used in steel analysis X-ray fluorescence and spark source optical emission spectrometry, as the suitability of these for the determination of trace elements is limited. Keywords Direct analysis of solid samples ; trace analysis ; steel analysis; This yields very low detection limits. spectrochenaical analysis ; hollowcathode lamp In a previous paper1 we described a study of a high-temperature hollow-cathode (HTHC) lamp combined with an echelle spectrometer for the determination of trace elements of high volatility in steel nickel-base and ferro alloys.The advantage of that excitation source is that it provides a very fast and convenient analysis procedure as the solid samples in chip or powder form can be analysed without prior dissolution. However one disadvantage of the HTHC lamp is the limitation of the number of elements suitable for determination. The elements studied1 were silver bismuth calcium lead antimony and zinc. For elements of low volatility it is more difficult to find suitable methods for the determination of micro-gram per gram levels that are as fast and as convenient as with the HTHC lamp. Spark source optical emission and X-ray fluorescence techniques are doubtless very fast and con-venient but their usefulness for determining trace elements is not always good enough.Spark source mass spectrometry is frequently used for the determination of trace elements in metals but this technique requires a highly qualified operator and the instrumentation is very expensive. Graphite furnace atomic-absorption (GF-AA) spectrometry has been used with great success for the determination of elements of high volatility directly in solid samples (cf. reference 2 and references cited therein). Since 1976 this method has been used for routine analysis of steel and nickel-base alloys at Sandvik AB.3 Attempts have also been made to extend this method to include elements of low volatility in low-alloy steel.4 Different versions of the glow discharge lamp developed by Grimm5 have also been used for the analysis of solid metal samples but they are not sensitive enough and therefore not generally suitable for the determination of trace elements.In an HTHC lamp temperatures up to 2000-2500 "C can be reached and the elements are therefore thermally vaporised and atomised. The cathode in a hollow-cathode lamp can also be cooled e.g. by water. In these low-temperature (LT) lamps the mechanism of atomisation is sputtering by positively charged ions. In the HTHC lamp helium is used to flush the lamp but in the LT lamps argon or neon is used because the sputtering is more efficient with these gases. In 1967 a paper was published by Knerr et aZ.6 describing a water-cooled hollow-cathode lamp used for the analysis of metals and glass. A version of this which we call a single-chamber hollow-cathode (SCHC) lamp was also built by us735 using the lamp body and power regulation system of the HTHC lamp.This paper describes an extension of the SCHC lamp idea to a new type of excitation source the double-chamber hollow-cathode (DCHC) lamp. * Present address Kiruna Geophysical Institute S-981 27 Kiruna Sweden A agust 1982 levels were used for the lamp. tungsten the line to background ratio became higher with a higher current. FOR DETERMINATION OF TRACE ELEMENTS IN STEEL 869 As can be seen two current It was found that for the elements arsenic niobium lead and In Table I1 the experimental parameters used are given. TABLE I1 EXPERIMENTAL PARAMETERS DCHC lamp-Flush time . . . . . . . . . . 60s Pre-burn time . . . .. . . . . . 60s Pressure Ne . . . . . . . . . . 1 torr Current . . . . . . . . . . . . 1 100 mA for arsenic niobium lead and tungsten, Power . . . . . . . . . . . . 250 I&' for arsenic niobium lead and tungsten, 700 mA for others 170 \4J for others Sample-Tube shaped 3 cm x 4 mm i.d. Spectyometev-Type . . . . . . . . . . . . Wavelength region . . * . . . . Linear dispersion . . . . . . . . Entrance and exit slits - . . . . . Integration time . . . . . . . . Integration time/line . . . . . . Integration timelmeasuring point . . . . &helle (IDES) 200-800 nm 0.16 nm mm-l a t 200 nm, 35 pm 60 s 1.5 s 0.1 s 0.32 nm mm-l at 400 nm The tubular sample is of course not as convenient as the powders or chips used in the HTHC lamp. However the precision in the drilling of the hole is not critical and we there-fore consider the sample preparation acceptable as lower detection limits can be reached.The diameter of the hole was varied and an experimental optimum of about 4 mm was found, in agreement with the theoretical value from the Debye relationship. For each analysis line detection limits were calculated using the expression 2 x 1.4 x (B x t ) l ( L - B ) (tic) Detection limit = where B counts s-l = intensity of the background; t s = integration time per line; L = total intensity of the line; and C pg g-l = concentration. This corresponds to twice the standard deviation of the background. For each element calibration graphs were constructed using the following reference materials low- and high-alloy steel JK series (Swedish Institute for Metal Research) ; high-alloy steel BCO series (AB Bofors Sweden); and high-alloy steel SDN series (Sandvik AB, Sweden).Results and Discussion In almost all instances they are less than 1 pg g-l which enables minimum concentrations of about 1-5 pg g-l (5-10 times the detection limit) to be determined in steel samples. This is acceptable for all elements. I t is interesting to note that the elements of high volatility arsenic calcium manganese, lead and tin also have very low detection limits. However they are higher than the corre-sponding detection limits that can be obtained using the HTHC lamp (0.01 pgg-l for lead and 0.0007 pg g-l for calcium).l For the metallurgist however the detection limits for lead and calcium achieved with the DCHC lamp are sufficiently low.For other elements of high volatility such as silver bismuth antimony and zinc poor results were obtained. However we believe that the experimental conditions for these elements can be improved and we intend to study this matter in the future. This is of great interest as it would be very convenient to be able to determine elements of both high and low volatilities in the same sample using the same technique. The calculated detection limits are given in Table 111 870 BERGLUND AND THELIN DOUBLE-CHAMBER HOLLOW-CATHODE LAMP Analyst VoZ. 107 TABLE I11 DATA FOR THE DETERMINATION OF TRACE ELEMENTS USING THE DCHC LAMP Element Aluminium . . Arsenic . . Boron . . Calcium . . Cobalt . . Chromium Manganese .. Molybdenum . . Niobium . . Nickel . . Lead . . Silicon . . Tin . . Titanium. . Tungsten. . Concentration 20-100 1040 20-90 range/pg g-l 7-50 50-1 000 16-1 1 000 3 000-26 000 40-2 500 3 000-8 000 90-3 300 50-120 40-100 9-30 60-19 000 2 000-4 000 Detection limit/ Mean RSD,* yo t3-l 2.6 0.3 2.3 0.1 2.8 0.9 3.1 0.03 3.4 1.1 3.6 (21% for 16 pgg-1) 1.5 3.4 0.9 2.4 (11% for 90 pg g-l) 1.7 3.9 1.7 1.7 1.6 0.1 0.3 0.2 0.1 0.7 0.04 1.6 0.3 0.2 0.2 * Means of ten replicate measurements for different concentrations within the concentration range studied. Comparing the detection limits shown in Table I11 with those normally obtained using a glow discharge lampll we find that they are much lower.The detection limits for a glow discharge lamp vary from 2 pg g-l (for boron) to 300 pg g-l (for tungsten). The detection limits obtained using the DCHC lamp are also approximately 10-20 times lower than those obtained using X-ray fluorescence and spark source emission spectrometry two techniques frequently used for the routine analysis of steel and they are 5-10 times lower than those obtained using an SCHC lamp. In Table 111 the concentration ranges for the reference samples used are shown and it is interesting to note that it is also possible to determine high concentration levels considering that the lamp was designed for the determination of trace amounts. Some calibration graphs are shown in Fig. 3. As can be seen good correlations are achieved between measured intensities and certified concentrations.As far as the precision of the measurements is concerned one can expect relative standard deviations (RSD) of about 1-3% as can be 4 I >- Y (a) ._ 2 0 25 50 75 100 0 25 50 75 100 I v I I - L3 0 50 100 150 200 0 25 50 75 100 Content iig g - ' Fig. 3. Calibration graphs for (a) aluminium 396.2 nm; (b) boron 249.8 nm; (c) lead 405.8 nm; and ( d ) tin 317.5 nm. Low-alloyed steel; and high-alloyed steel August 1982 FOR DETERMINATION OF TRACE ELEMENTS I N STEEL 871 seen in Table 111. These are better than those obtained using the SCHC lamp (24%) owing to the higher capacitance in the DCHC lamp and to the double-chamber technique, which gives more stable plasma waves.MJith this lamp design it is also easier to fill the whole sample cylinder with plasma without any delay which was sometimes a problem with the SCHC lamp. Another fact that improves the precision further is the continuous neon flow in the cylinder. This neon flow together with the sputtering procedure during the pre-burn time will keep the inside of the cylinder clean before analysis. The DCHC lamp technique for the determination of trace elements of low volatility in steel is very interesting because nowadays this is usually done using expensive and tedious wet chemical procedures. Plasma techniques (inductively coupled plasma or directly coupled plasma) have in some instances simplified trace analysis. However the samples have to be brought into solution and for the laboratory of a steel works it is of great importance to be able to analyse solid samples directly.This saves time as several alloys are difficult to dissolve within a reasonable time. The lamp is also of value because it can be combined with the HTHC lamp (the same regulation system) and with that combination most elements of interest can be determined rapidly and easily at microgram per gram levels. This study can be regarded as a pilot study as neither systematic studies to find the most sensitive emission lines nor analyses of samples using the calibration graphs were performed. Further the lamp can probably be improved and as far as the sample form is concerned it is necessary to develop a convenient procedure for preparing tubular samples before the technique can be used routinely in a steel-works laboratory.For hard alloys when drilling is impossible this might be a problem. However the results obtained so far are very promising and an extensive study of the lamp is in progress. This work was performed at the Swedish Institute for Metals Research and financed by the Swedish Ironmasters’ Association (Jernkontoret) the Swedish Board for Technical Development and Sandvik AB. The authors gratefully acknowledge Mrs. Ingrid Eklof at the Institute for Metals Research for skilled technical assistance. They are also grateful to Mr. Bengt Hammarberg Sandvik AB who manufactured the lamp and to Sandvik AB for permission to publish this paper. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Thelin B. AppZ. Spectrosc. 1981 35 302. Headridge J . B. Spectrochim. Acta Part B 1980 35 785. Backman S. and Karlsson R. W. Analyst 1979 104 1017. Sommer D. and Ohls K. Fresenius 2. Anal. Chem. 1979 298 123. Grimm W. Spectrochim. Acta. Part B 1968 23 443. Knerr G. Maierhofer J. and Reis A. Fresenius Z. Anal. Chem. 1967 229 241. Thelin B. “ Jernkontorets Forskning Series D No. 343,” Jernkontoret Stockholm Sweden 1981. Danielsson L. and Thelin B. Proceedings of the Fourth Commission Europignne d’Etude e t Chen F. “Plasma Physics,” Plenum New York 1974. Danielsson A. and Lindblom P. AppZ. Spectrosc;. 1976 30 151. De Gregorio P. Morello B. and Savastano G. Metall. Ital. 1980 3 89. d’ Application de Travaux de 1’Analyse en Siderugie Meeting Liege Belgium 1981. Received February 19th 1982 Accepted March Eth 198

ISSN:0003-2654    

DOI:10.1039/AN9820700867

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

a72 .4nalyst, Aiquust, 1982, Vol. 107, $9. 872-878 Attempts to Eliminate Interferences in the Determination of Arsenic, Antimony, Tin and Germanium by Molecular Emission Cavity Analysis with the Hydride Generation Technique E. Henden Faculty of Chemistry, University of Ege, Bornova, Izmir, Turkey Interferences in the determination of arsenic, antimony, tin and germanium by molecular emission cavity analysis with the hydride generation technique are reported. Most of the interferences on the emission of arsenic, antimony and germanium were eliminated by masking with EDTA, which, however, caused broadening of the tin peak. Suppression by Cd2+ and Zn2+ of the tin emission was more serious in the presence of EDTA. The effect of Hg2+ on the emission of arsenic, antimony and tin was eliminated by masking with potassium iodide, and that of Ag+ on the emission of antimony and tin was eliminated by separating as silver chloride.Cu2+ suppressed all four emissions and suppression of the arsenic emission was eliminated by separating Cu2+ as its hexacyanoferrate(I1) precipitate ; this precipitate strongly adsorbed antimony, tin and germanium ions. Many ions suppressed the emissions. Xeywords : Arsenic, antimony, t i n and germanium determination ; hydride generation ; molecular emission cavity analysis ; interferences Several ions interfere in the determination of arsenic, antimony, tin and germanium by means of hydride generation utilising sodium tetrahydroborate( 111) reduction.1-6 Some attempts, mostly for the determination of arsenic, to overcome these interferences have been reported.Belcher et aZ.394 eliminated the interferences of Co2+, Ni2+, Zn2+, Fe3+, Bi3+, Cd2+, Cu2+ and Ag+ (50 pg ml-l) in the determination of arsenic (10 pg ml-l) and antimony (25 pg ml-l) by making the test solutions 0.01 M in EDTA. The use of EDTA was later applied to the determination of arsenic in fish7 and bismuth in nickel-based alloys8 Yamamoto and Kumamarus suggested masking the interfering ions with potassium iodide to reduce their effects on the determination of arsenic. Kirkbright and TaddialO proposed thiosemicarbazide and 1 ,lo-phenanthroline as masking agents in strongly acidic solutions to reduce the inter- ferences of Cu2+, Ni2+, Pt4+ and Pd2+ in the determination of arsenic. This work was carried out in order to investigate interferences and their treatments in the determination of arsenic, antimony, tin and germanium by using molecular emission cavity analysis (MECA) in conjunction with a hydride generation technique.Masking the inter- ferents with EDTA was further studied at higher ratios of interfering ion to the ion of interest in the determination of arsenic and antimony. The effect of EDTA on the interferences observed in the determination of germanium and tin was also studied. Some other treat- ments of interferences are proposed. The MECA technique used in this work has been described earlier.ll Experimental Apparatus Emission was measured with a Pye Unicam SP 90A flame spectrophotometer, with a 1.4-mm slit (band width at 400 nm = 45 nm). A Varian G-2500 recorder, having a response time of 0.5 s for full-scale deflection, was connected to the output of the instrument.A stainless-steel cavity without water cooling (4 mm diameter and 10 mm deep), similar in shape to that described earlier,ll was used. Three stainless-steel tubes (one of 0.8 mm i.d. for nitrogen as the carrier gas and the other two of 0.3 mm i.d. for introduction of oxygen and hydrogen into the cavity) were connected via two drilled holes to the rear of the cavity at tangents to the cavity wall. The cavity was positioned horizontally in line with the detector by means of a cavity-holding assembly.HENDEN 873 The volatilisation system was similar to that described earlier4 with a glass reaction vessel 12.5 cm long and 1.5 cm in diameter. When studying interferences in the determination of arsenic and antimony, the arsine and stibine generated were separated on a PTFE column (13 cm x 3 mm i.d.) packed with 10% E-301 silicone gum rubber on Porapak Q support (80-100 mesh).The column was con- nected by means of PTFE tubes to the reaction vessel through a drying tube, packed with calcium sulphate powder, and to the stainless-steel tube (0.8 mm i.d.) in one of the rear openings of the cavity. The column temperature was controlled at 17 & 0.5 "C by immersion in water contained in a Dewar flask. For the study of interferences in the determination of germanium and tin, the hydrides of these elements were separated by a 55 cm x 3 mrn i.d. PTFE column packed as above and operated at 15 0.5 "C. Chemicals All reagents were of analytical-reagent grade, unless stated otherwise.Stock solutions (1 000 pg ml-l) of arsenic(III), antimony(III), tin(I1) and germanium(1V) were prepared from arsenic(II1) oxide, antimony(II1) oxide, tin(I1) chloride and germanium- (IV) oxide, respectively. The working solutions contained a mixture of either 0.4 pg ml-l of arsenic and 1.0 pg ml-l of antimony or 0.4 pg ml-l of tin and 2.0 pg ml-l of germanium in 0.1 M hydrochloric acid, unless stated otherwise. These solutions were prepared daily. For the interference studies, stock solutions of the possible interfering ions, containing 2500 pg ml-l of the element, were prepared from the corresponding elements, oxides, chlorides, sulphates or nitrates. Solutions of molybdenum and silicon were prepared from ammonium molybdate and sodium silicate, respectively.EDTA solution (0.25 M) was prepared from its disodium salt (Merck, laboratory-reagent grade). Sodium tetrahydroborate( 111) solution (5% m/V) was prepared by dissolving an appropriate amount of the powder (BDH Chemicals, laboratory-reagent grade) in 0.01 M sodium hydroxide solution. Measurement of Emission The oxygen - hydrogen - nitrogen (carrier gas) flame was kept on throughout the experi- ments. A 0.5-ml volume of the sodium tetrahydroborate(II1) solution was pipetted into the reaction vessel, which was then connected to the volatilisation system. After deaerating the system for 20s with the nitrogen carrier gas, the chart recorder was turned on. A 1-ml volume of the test solution was injected through the septumon the reaction vessel.The As0 and SbO emissions2 were recorded at 400 nm as the hydrides were eluted from the gas- chromatographic column. The SnO emission4 and the blue emission from germanium, probably arising from oxide-containing species,12 were recorded a t 490 nm in the same way. The results were the same when peak-height or peak-area measurements were used, except for tin in the presence of EDTA. However, for ease of measurement, the following results were obtained by measuring the heights of the sharp arsenic and germanium peaks, and the areas of the rather broad antimony and tin peaks, except where stated otherwise. Peak areas were measured by triangulation. The gas flow-rates are given in Table I. Effect of Various Acids on Emission The test solutions were prepared by dilution of the stock arsenic, antimony, tin and germanium solutions with distilled water and nitric acid, sulphuric acid or perchloric acid.TABLE I FLOW-RATES OF THE GASES USED IN THE MEASUREMENTS Flow-rate/ml min-' Mixtures f-- -l determined Oxygen Hydrogen Nitrogen Arsenic - antimony . . .. 55 85 55 Tin - germanium . . . . 35 90 45874 HENDEN: ATTEMPTS TO ELIMINATE INTERFERENCES IN THE AnaZyst, VoZ. 107' The final acidity of each solution was adjusted to 0.1 M with the acid used. The emission intensities obtained by injecting 1 ml of each solution were compared with those obtained using test solutions made 0.1 M in hydrochloric acid. Interference Studies The test solutions containing a mixture of 0.4 pg ml-l of arsenic, 1.0 pg ml-l of antimony and 500 pg ml-l of potential interfering element, as its suitable ion, were prepared.The emission intensities obtained by injecting 1 ml of these solutions were compared with that obtained by injecting 1 ml of the test solution containing the same amount of arsenic and antimony but no possible interfering ion. The interferences of 500 pg ml-l of elements on the emission intensities of 0.4 pg ml-l of tin and 2.0 pg ml-l of germanium were studied in the same way. All the final test solutions were made 0.1 M in hydrochloric acid. Eflect of EDTA on interferences The test solutions were made 0.05 M in EDTA. Eflect of potassium iodide on interferences Potassium iodide at a concentration of 1.6% m/V was added to the test solutions. Elimination of interference of silver(I) in antimony and tin determinations was kept in daylight for 10-15 min.natant liquid was injected. The silver chloride suspension formed by addition of hydrochloric acid to the test solution The precipitate formed was centrifuged and the super- Effect of Copper( 11) Hexacyanoferrate(I1) Precipitation on the Determinations Various amounts of potassium hexacyanoferrate(I1) were added to the test solutions containing 500 pg ml-l of copper(I1). The precipitates formed were centrifuged and the supernatant liquids were injected. Results and Discussion The relative standard deviations for the determination of 0.4 pg ml-1 of arsenic, 1 .O pg ml-1 of antimony, 0.4 pg ml-l of tin and 2.0 pg ml-l of germanium in arsenic - antimony (eleven experiments) and tin - germanium (seven experiments) mixtures were 3.8, 4.5, 4.6 and 3.8%, respectively.An interference was defined as significant if a change of more than two standard deviations in the measurements were observed. The above concentrations of arsenic, antimony, tin and germanium were well above the detection limits and in the linear parts of the calibration graphs. In order to study the effects of nitric, sulphuric and perchloric acids, 1 ml of a test solution, made 0.1 M in one of these acids, was injected. The emission intensities found did not show any significant difference from the emission intensities obtained using solutions 0.1 M in hydrochloric acid. Among the ions whose interferences were studied, AP+, Pb2+, Si4+, Th4+ and T1+ did not interfere in any of the determinations, whereas several other ions suppressed some or all of the four emissions (Table 11).Most of these interferences in the determination of arsenic, antimony and germanium were effectively eliminated or reduced by the addition of EDTA to the test solutions (Table 11). EDTA, which had no effect on the emission peaks of arsenic, antimony and germanium, caused broadening of the tin peak. The effect of EDTA on the tin emission is, therefore, described later. On reduction, most of the interfering elements gave black precipitates, which are believed to be the finely divided metal. Zirconium and zinc gave white precipitates (their hydr- oxides). In the presence of EDTA, no precipitation occurred from solutions containing Ni2+, Co2+, Cd2+, Fe3+, Zn2+ and In3+, whereas Cu2+, Hg2+ and Bi3+ still formed black precipitates and Ag+ gave a white precipitate (silver chloride) with a black tint.Molybdate formed a brown precipitate in the absence of EDTA and a dilute brown suspension in the presence of EDTA. Tungstate gave a blue and vanadate a violet (possibly V2+) solution in both the presence and absence of EDTA. In the EDTA-containing solution, zirconiumAugust, 1982 DETERMINATION OF As, SB, SN AND GE BY MECA TABLE I1 EFFECT OF INTERFERING IONS AND OF EDTA ON THE EMISSION INTENSITIES 875 Concentrations: arsenic, 0.4 pg ml-l; antimony, 1.0 pg ml-l; germanium, 2.0 pg ml-l; tin, 0.4 pg ml-1; foreign ion, 500 pg nil-l; and EDTA (where used), 0.05 M. Change in emission intensity, yo* Ion Ni2+ , . . .. . Zn2+ . . . . . . co2+ . . . . . . Cd2+ . . . . . . Fe3+ . . . . . . Bi3+ . . . . . . Ag++ t * * . . . . CU2+ . . . . . . Hg2+ . . . . . . 1n3+ . . . . . . W 6 + . . . . . . Mo6+ . . . . . . V 5 + . . . . . . Zr4+ . . .. .. Se4+ . . . . . . Te4+ . . .. . . Ti3+ . . . . . . Mn2+ . . .. .. Cr3+ . . .. . . As alone - 73 - 58 - 67 - 26 - 45 - 30 -21 - 88 - 24 - 72 - 44 - 18 - 59 - 88 As + EDTA 2 -4 -1 -5 -3 -3 -2 - 45 - 18 0 -2 4 - 51 - 68 Sb alone - 75 - 31 - 48 - 23 - 40 - 82 - 69 - 79 - 100 - 59 - 71 - 45 - 14 - 30 - 74 Sb + LDTA -5 2 -5 0 0 - 35 - 20 - 64 - 96 - 50 3 - 28 2 - 20 - 64 Ge alone - 100 - 55 - 100 - 22 - 100 - 100 - 19 - 77 - 54 Ge + EDTA -2 -4 -2 -4 -4 - 16 0 - 73 - 55 Sn alone7 - 100 - 18 - 87 - 44 - 81 - 64 - 39 - 100 - 25 - 50 - 32 - 30 - 78 - 60 - 28 - 13 - 17 * A space indicates no significant interference.t Results for tin with addition of EDTA are not given because EDTA affects the tin peak. AgCl suspension injected. still formed a white precipitate. Hence, as proposed ear lie^,^ the EDTA, where effective, prevents or slows down the reduction or precipitate formation of the interfering ions and thus prevents or reduces the adsorptive or reactive capture of the hydrides. Selenium formed a brown and tellurium a black precipitate just after the injection. These precipitates were dissolved after a few seconds following the injection and colourless solutions were formed. The precipitates are believed to be the elements, which are then probably dissolved by further reduction to selenide and telluride. Hydrogen selenide is expected to give a blue emission within the cavity,ll if formed.However, as the pH of the solution increases to 8.2 soon after the injection, hydrogen selenide, being an acid, could not leave the solution; the selenium emission was observed when the injection of the Se4+ solution, 0.1 RI in hydrochloric acid, was followed by injection of 1 ml of 2 M hydrochloric acid, after which the solution became acidic. Because the hydrogen selenide could not leave the solution under the working conditions used, the interference of selenium on the emissions used is believed to be at the volatile hydride formation and evolution step. Tellurium is believed to interfere in a similar way. Treatments to overcome some of the interferences that could not be eliminated by adding EDTA to the test solutions were investigated.The interference of Hg2+ on the arsenic, antimony and tin emissions was eliminated by adding 1.6% of potassium iodide to the test s o h tions. The interference of Ag+ on the arsenic emission was eliminated by the addition of EDTA, as described above. However, for the elimination of the suppression by Ag+ of the antimony and tin emissions it was necessary to separate the silver chloride formed from the test solution. For this purpose, the silver chloride suspension was kept in daylight for 10-15 min. The precipitate, formed partly by decomposition of the suspension, was centrifuged and the supernatant liquid injected. The suppressive effect of Ag+ on the arsenic, antimony and tin emissions was thus eliminated even without adding EDTA.Attempts to eliminate the Cu2+ interference by copper( 11) hexacyanoferrate(I1) precipita- tion are described below. Maximum tolerable amounts of the elements whose interferences could not be eliminated by the treatments described above are given in Table 111. EDTA had no effect on the reduction of Se4+ and Te4+.876 HENDEN : ATTEMPTS TO ELIMINATE INTERFERENCES IN THE Analyst, VoZ. 107 TABLE I11 MAXIMUM TOLERABLE AMOUNTS OF INTERFERING IONS IN THE DETERMINATION OF ARSENIC, ANTIMONY AND GERMANIUM Maximum tolerable amount/pg ml-l* 7 A \ Arsenic Antimony Germanium Interfering ion (0.4 pg ml-l) (1.0 p g ml-l) (2.0 pg ml-l) Bi3+, V5+, W6+ . . .. 100 cu2+ . . .. .. .. 100 200 Se4+ . . .. .. .. 250 250 20 Te4+ . . .. .. .. 4 4 50 * A space indicates no significant interference or that the interference could be eliminated by the treatments described.Effect of Copper( 11) Hexacyanoferrate( 11) Precipitation on Emission The effect of copper(I1) hexacyanoferrate(I1) precipitation on the emission intensities of arsenic (5 pg ml-l) , antimony (20 pg ml-l), tin (2.0 pg ml-l) and germanium (5 pg ml-1) is shown in Table IV. Hexacyanoferrate(I1) M) alone did not affect the arsenic, antimony and germanium emissions, but suppressed the emission of tin by 36%. This suppression niay be attributed to the tin(I1) - hexacyanoferrate(I1) precipitation, which possibly occurs when the pH of the solution increases during the reaction. When only half of the amount of Cu2+ was precipitated, all of the four emissions were suppressed by the unprecipitated CU~+, as expected. The precipitation of copper( 11) with an excess of hexacyanoferrate(I1) , while suppressing the antimony and tin emissions completely and the germanium emission strongly, did not affect the arsenic emission.Hence the procedure may be used to separate copper, antimony, tin and most of the germanium from arsenic and, therefore, may be utilised to minimise the interferences of these ions1 in the determination of arsenic by the hydride generation tech- nique. As support for the above claim, the precipitation was conducted in a solution con- taining a lower proportion of arsenic. Thus, the precipitation of 500 pg ml-1 of Cu2+ with 2 x M hexacyanoferrate(I1) was found to be effective in the separation of 0.5 pg ml-l of arsenic from 50 pg ml-l each of antimony, tin and germanium. The arsenic emission was unaffected, whereas the antimony emission was completely suppressed, the tin emission was almost completely suppressed (it was slightly above the detection limit) and the germanium emission was strongly suppressed [24mV, whereas the same amount of germanium in the absence of Cu2+ and hexacyanoferrate(I1) gave an emission too intense to be measured].The suppression of the emissions by the precipitation of Cu2+ with an equivalent or an excess amount of hexacyanoferrate(I1) may be attributed to adsorption by the precipitate or to the coprecipitation of the ions of interest. TABLE IV EFFECT OF COPPER( 11) HEXACYANOFERRATE (11) PRECIPITATION ON THE EMISSION INTENSITIES Copper(I1) concentration : 500 pg ml-l.Change in emission intensity, yo* r I Hexscyanoferrate( 11) As Sb Sn Ge concentration/M (5 p g ml-l) (20 pg ml-l) (2 pg ml-l) (5 pg ml-I) 2 x 10-3 - 34 - 69 - 70 - 33 4 x 10-37 3 3 - 73 5 1 x 10-2 5 - 100 - 100 - 72 2 x 10-2 4 - 100 - 100 -91 * Compared with emission in the absence of Cu2+ and hexacyanoferrate(I1). 7 Approximately equivalent to 500 pg ml-l of Cu2+.August, 1982 DETEXVIINATION OF As, SB, SN AND GE BY MECA 877 Effect of EDTA on the Tin Emission in the Presence and Absence of Interfering Ions The effect of EDTA on the tin emission remained constant when subjected to the following vari- ations: increase in the hydrochloric acid concentration from 0.1 to 0.5 M ; changes in the order of addition of the reagents; magnetic stirring of the solution in the reaction vessel; and use of the tin(1V) instead of tin(I1).The effect did not vary significantly with the amount of tin, whereas the peak height decreased with increasing EDTA concentration and increasing nitrogen carrier gas flow-rate (Table V). The broadening of the tin peak in the presence of EDTA may be attributed to the slow reduction of the tin in the tin - EDTA complex by the sodium tetrahydroborate(II1). The tin peak broadened in the presence of EDTA, but the peak area did not change. TABLE V VARIATION OF THE EFFECT OF EDTA ON THE TIN EMISSION WITH THE AMOUNT OF TIN, EDTA CONCENTRATION AND NITROGEN CARRIER GAS FLOW-RATE Concentration of tin/ Concentration of Nitrogen flow-rate/ Change in peak Change in peak pg ml-l EDTA/M ml min-l height, yo* area, yo* 0.4 0.01 40 - 34 4 0.05 40 - 49 1 0.01 90 - 54 2 0.05 90 - 68 -6 5.0 0.01 0.05 0.01 0.05 40 40 90 90 - 32 6 - 47 -2 - 50 3 - 63 -3 * Compared with emission using corresponding carrier gas flow-rate in the absence of EDTA.As the EDTA did not affect the area of the tin peak, it was thought it could be used in the elimination of interferences on the emission of tin. However, it was found that in the presence of EDTA (0.05 M), though 100 pg ml-l of Co2+ and Ni2+ did not interfere with the emission of tin (0.4 pg ml-l), the suppression of the tin emission by Cd2f and Zn2+ increased (Table VI). TABLE VI EFFECT OF EDTA ON THE TIN EMISSION IN THE PRESENCE OF SOME INTERFERING IONS Concentrations: tin, 0.4 pg ml-l; EDTA, 0.05 M.Nitrogen carrier gas flow-rate: 40 ml min-l. Concentration/ Peak-area Concentration/ Peak-area Ion pg ml-l suppression, "/G Ion pg ml-' suppression, yo Co2f . . . . 100 5 Zn2+ . . . . 2 21 Ni2+ . . . . 100 0 10 75 Cd2+ . . . . 10 100 20 88 125* 15 loo* 0 * Interference in the absence of EDTA. The increases in the suppression of the tin emission by Cd2+ and Zn2+ in the presence of EDTA might be due to the formation of a cadmium (or zinc) - tin - EDTA complex. The tin in this complex may be reduced by the sodium tetrahydroborate(II1) too slowly for it all to be detected.878 HENDEN References Smith, A. E., Analyst, 1975, 100, 300. Belcher, R., Bogdanski, S. L., Ghonaim, S. A., and Townshend, A., Anal. Chim. Acta, 1974, 72, 183. Belcher, R., Bogdanski, S. L., Henden, E., and Townshend, A., Analyst, 1975, 100, 522. Belcher, R., Bogdanski, S. L., Henden, E., and Townshend, A,, Anal. Chim. Acta, 1977, 92, 33. Arbab-Zavar, M. H., and Howard, A. G., Analyst, 1980, 105, 744. Aggett, J., and Aspell, A. C., Analyst, 1976, 101, 341. Flanjak, J., J . Assoc. Ofl. Anal. Chem., 1978, 61, 1299. Drinkwater, J. E., Analyst, 1976, 101, 672. Yamamoto, Y., and Kumamaru, T., Fresenius 2. Anal. Chem., 1976, 282, 139. Kirkbright, G. F., and Taddia, M., Anal. Chim. rlcta, 1978, 100, 145. Bogdanski, S. L., Henden, E., and Townshend, A., Anal. Chim. Acta, 1980, 116, 93. Pearse, R. W. B., and Gaydon, A. G., “The Identification of Molecular Spectra,” Third Edition, Received September 28th, 1981 Accepted March 5th, 1982 Chapman and Hall, London, 1965. 1. 2 . 3. 4. 5. 6. 7. 8. 9. 10. 1 1 . 12.

ISSN:0003-2654    

DOI:10.1039/AN9820700872

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

Analyst, August, 1982, Vol. 107, $9. 879-884 879 Assay of Ephedrine or Pseudoephedrine in Pharmaceutical Preparations by Second and Fourth Derivative Ultraviolet Spectrophotometry A. G. Davidson" and H. Elsheikh Department of Pharmaceutical Chemistry, University of Strathclyde, Glasgow, G 1 1X W Rapid second and fourth derivative ultraviolet spectrophotometric assay procedures are described for the determination of ephedrine or pseudo- ephedrine in pharmaceutical formulations. The methods have been applied successfully to Ephedrine Elixir BP, Ephedrine Hydrochloride Tablets BPI Ephedrine Nasal Drops BPC, Paediatric Belladonna and Ephedrine Mixture BPC, tablets and capsules containing aminophylline and amylobarbitone and coloured syrups containing triprolidine hydrochloride and codeine phosphate.A simple solvent extraction stage avoids interference from the co-formulated drugs and colouring agents in certain of the formulations. The specificity, accuracy and precision of the methods have been assessed. Keywords Ephedrine determination ; Pseudoephedrine determination ; ultra- violet spectrophotometry; derivative spectrophotornetry; pharmaceutical formulations Simple ultraviolet spectrophotometric methods for the assay of ephedrine or its diastereo- isomer pseudoephedrine in pharmaceutical formulations are subject to interference from other ultraviolet-absorbing components in the formulation such as co-formulated drugs and formulation excipients. The problem is exacerbated by the low absorptivity of the drugs (the at 257 nm is 9.0) typical of benzenoid substances, which necessitates the use of relatively large amounts of the formulation in the sample solutions.A variety of procedures have been published that render the spectrophotometric deter- mination of ephedrine or pseudoephedrine more specific and sensitive, including orthogonal polynomina1,l charge-transfer2 and bilateral diff erence3 spectrophotometric methods and measurement of absorbance after separation on an alginic acid column4 or a cationic-exchange resin5 and after conversion into benzaldehyde,6 serni~arbazone~ or coloured The technique of derivative spectrophotometry has proved particularly useful in eliminating matrix interference in the assay of many medicinal substances,12 including those showing benzenoid absorption spectra.l39l4 This paper describes a rapid ultraviolet spectrophoto- metric assay of ephedrine or pseudoephedrine in pharmaceutical formulations based on the measurement of the second or fourth derivative absorption spectra of samples obtained by dilution with 0.1 M hydrochloric acid or by a simple solvent extraction procedure. Experimental Reagents Ephedrine hydrochloride.Pseudoephedrine hydrochloride. Cyclohexane. Other reagents and solvents were of analytical-reagent grade. Sigma London Chemical Co. Ltd. Sigma London Chemical Co. Ltd. SpectrosoL grade (BDH Chemicals Ltd.). Spectrophotometer Second derivative spectra were recorded in 1-cm silica cells from 285 to 230 nm using a Perkin-Elmer 552 double-beam ultraviolet - visible recording spectrophotometer with Key- board Expansion Accessory operating in the second derivative mode.The spectral slit width was 2 nm, the scan speed 60 nm s-l, the response 2 s, the ordinate minimum and maximum settings were -0.25 and 0.25, respectively, and the recorder range was 1 V. Fourth derivative spectra were generated with a Hitachi Derivative Accessory operating in * To whom correspondence should be addressed.880 DAVIDSON AND ELSHEIKH : EPHEDRINE OR PSEUDOEPHEDRINE BY Analyst, VoZ. 107 the second derivative mode (mode 5 ) in series with the spectrophotometer in the second derivative mode with the same instrumental parameters as above except that the recorder range was 10 V. Standard Solutions Dissolve ephedrine hydrochloride or pseudoephedrine hydrochloride (about 300 mg, accurately weighed) in water and dilute to 100 nil.Dilute 10 ml of solution to 100 ml with 0.1 M hydrochloric acid and record the second or fourth derivative absorption spectrum with 0.09 M hydrochloric acid in the reference cell. Sample Solutions Method 1 : f o r Ephedrine Elixir BP and Ephedrin.e Nasal Drops B P C Dilute an amount of the sample containing about 30 mg of ephedrine hydrochloride, based on the stated concentration, to 100 ml with 0.1 M hydrochloric acid and record the second derivative absorption spectrum. Method 2 : f o r Ephedrine Hydrochloride Tablets BP Shake an amount of the powder containing 30mg of ephedrine hydrochloride with 0.1 M hydrochloric acid (90 ml) for 30 min. Dilute to 100 ml with water and filter the suspension through Whatman No.1 filter-paper, discarding the first 10 ml of filtrate. Weigh and powder 20 tablets. Record the second derivative absorption spectrum of the filtrate. Method 3: f o r Paediatric Belladonna and Ephedrine Mixture B P C and a commercial syrup containing pseudoephedrine hydrochloride and triprolidine hydrochloride Transfer an amount of the formulation containing 30 mg of ephedrine hydrochloride or pseudoephedrine hydrochloride into a separating funnel and add water (30 ml). Make the solution alkaline with 5 M sodium hydroxide solution (1 ml), add a saturated solution of sodium chloride (10 ml) and extract for 1 min with three aliquots (30 ml) of diethyl ether. Shake the combined ether extracts with three aliquots (30 ml) of 0.1 M hydrochloric acid. Gently shake the combined acidic extracts with cyclohexane. Transfer the aqueous layer into a calibrated flask (100 ml), dilute to volume with water and record the second derivative absorption spectrum.Method 4 : for tablets and capsules of ephedrine hydrochloride, aminophylline and amylo- barbitone Shake an amount of the powder containing 30 mg of ephedrine hydrochloride with 0.1 M hydrochloric acid (20 ml) for 30 min. Filter the extract through a No. 3 sintered-glass filter, wash the retained powder with water (10 ml) and add the washings to the filtrate. Continue the assay as described in method 3 from “Make the solution alkaline. . . .” Weigh and powder 20 tablets or the contents of 20 capsules. Method 5 : f o r a linctus containing pseudoephedrine hydrochloride, triprolidine hydrochloride and codeine phosphate Transfer an amount of the syrup containing 15 mg of pseudoephedrine hydrochloride into a separating funnel and continue the assay described in method 3 from “add water (30 ml).. . .” Treatment of the Results Measure in millimetres the amplitude 4-5 [Fig. l ( b ) ] in the second derivative spectrum or amplitude 1-2 [Fig. 3 ( b ) ] in the fourth derivative spectrum of the standard and sample solutions and calculate the concentration of ephedrine hydrochloride or pseudoephedrine hydrochloride in the sample solution and hence in the sample from the proportional relationship that exists between the measured amplitude and concentration. Results and Discussion Record the fourth derivative absorption spectrum. The interference from the formulation excipients in a simple spectrophotometric assay of A “blank” elixir (diluted 1 + 9 with 0.1 M Ephedrine Elixir BP is shown in Fig.l ( a ) .August, 1982 SECOND AND FOURTH DERIVATIVE uv SPECTROPHOTOMETRY 881 hydrochloric acid), consisting of all the ingredients of Ephedrine Elixir BP except ephedrine hydrochloride, shows considerable absorption in the 230-300-nm region due principally to the invert syrup and the colouring agent tartrazine. The absorption spectrum of the dilution of Ephedrine Elixir BP consists of the spectrum of ephedrine with its fine vibrational structure at 251.0, 256.3 and 262.2 nm superimposed on the absorption bands of the formulation matrix. Approximately 25% of the absorbance at 256.3 nm of a dilution of the elixir is due to ephedrine.The second derivative absorption spectrum [Fig. 1 ( b ) ] of ephedrine shows enhanced resolution of the fine structure. The peaks designated 2, 4 and 6 coincide, after correction for scan speed effects,15 with the absorption maxima in the zero order spectrum. The second derivative spectrum of the diluted “blank” elixir shows very little structure, particularly in the wavelength region of the fine structure of ephedrine. Discrimination against broad sDectra1 bands in favour of narrow bands is one of the advantages of derivative sDectroscoDv16 afnd in the second derivative assay of Ephedrine Elixir BP elhinates the broad absorption bands of the excipients. interflerence oflghe 240 280 320 240 280 320 Wavelengthinm Fig. 1 (a). Zero order absorption spectra of A, 0.03 nt/V ephedrine hydrochloride; B, 1 + 9 Ephedrine Elixir BP; and C, 1 + 9 “blank” elixir in 0.1 M hydrochloric acid.(b) Second order derivative spectra of A, 0.3% m/V ephedrine hydrochloride; and B, 1 + 9 “blank” elixir in 0.1 M hydro- chloric acid. Specificity To confirm the specificity of the second derivative spectrophotometric assay of ephedrine in the presence of the elixir matrix, aliquots of a solution of ephedrine hydrochloride (0.3% m/V) and of the “blank” elixir were mixed to simulate dilutions of the elixir (1 + 9) having a constant ephedrine concentration (0.03% m/V) and a varying matrix concentration of 0 to 140% of the normal analytical concentration of the sample. The results in Fig. 2 show that the formulation matrix at the specified sample concentration increases the amplitudes 5-6 and 6-7 of the second derivative spectrum of ephedrine by 6.3 and 7.1%, respectively, but does not affect the height of the other amplitudes.The amplitude 4-5 measured at 256.3 nm to its shorter wavelength satellite offers the highest sensitivity and is the amplitude of choice for the assay. A similar experiment carried out with the components of Ephedrine Nasal Drops BPC confirmed the specificity of the mzthod for this preparation. In certain other formulations, the interference by the colouring agents or co-formulated drugs was so great that isolation of the ephedrine by a simple solvent extraction procedure before the derivative measurement was necessary. The recovery of ephedrine by the extraction method was 99.2% (mean of eight measurements).The last stage of the extrac- tion procedure, washing the acidic extract with cyclohexane, removed from the acidic extract ether striations that cause small but rapid fluctuations in absorbance and that result in an unacceptable background noise in the derivative spectra. The specificity of the assay was150- 5 A A . . A & (' B A - n n - Concentration of elixir matrix as a ratio of sample concentration The effect of the elixir matrix on the second derivative amplitudes of Ephedrine Elixir BP. A, Amplitude 4-5; B, amplitude 3-4; C, amplitude 2-3; D, ampli- tude 1-2; E, amplitude 5-6; and F, amplitude 6-7. Fig. 2. also assessed by the method described above for those products requiring solvent extraction where the composition of the formulation was known or where the components in the hydro- chloric acid extract could be anticipated from a knowledge of the extraction properties of the excipients and of the co-formulated drugs.With only one exception, the measurement of amplitude 4-5 after solvent extraction provides a specific assay of ephedrine or pseudo- ephedrine that is free from spectral interference from the matrix and co-formulated drugs. Of the formulations examined, only one, a linctus containing pseudoephedrine hydro- chloride, codeine phosphate and triprolidine hydrochloride, showed interferences in the second derivative spectrum of the pseudoephedrine. As all three drugs are quantitatively extracted by the solvent extraction procedure, the acidic extract represents a three- component mixture of the three drugs as their hydrochlorides.In specificity experiments in which the effects of codeine and triprolidine on the second derivative spectrum of pseudo- ephedrine were investigated separately, the codeine was found to introduce a small but significant systematic error into the height of every derivative band. However, the inter- ference was completely eliminated in the fourth derivative spectrum. Fig. 3 shows the zero order and fourth derivative spectra of an acidic extract of the linctus obtained by the solvent extraction procedure. Although the fine structure of the pseudoephedrine is barely dis- cernible in the trough of the zero order composite absorption bands of the triprolidine and codeine, the improved discrimination in favour of the narrow spectral bands of the pseudo- ephedrine against the broad bands of the triprolidine and codeine yields the fourth derivative spectrum, which is almost identical with that of pseudoephedrine alone.The wavelengths of the peaks 2, 4 and 6 in the fourth derivative spectrum correspond with those of peaks 2, 4 and 6 in the second derivative spectrum and with the A,,,, in the zero order spectrum of pseudoephedrine. The regression equations for the amplitude 4-5 in the second derivative spectra of seven solutions of ephedrine and pseudoephedrine (0-0.6 mg ml-l) were y = 5 6 5 . 3 ~ - 0.54 and y = 585.3% - 0.28 respectively, where y mm is the amplitude and x mg ml-1 is the concentration. The regression equation for the amplitude 1-2 in the fourth derivative spectra of seven solutions of pseudoephedrine (04.6 mg ml-l) was y = 582.6% + 0.89.In a study of the specificity of the measurement of fourth derivative amplitudes of a constant concentration of pseudoephedrine and increasing concentrations of triprolidine and codeine up to 140% of their concentration at the analytical concentration of pseudoephedrine, the amplitude 1-2 was found to be completely free of interference from the co-formulated drugs and was the amplitude chosen for measurement in the assay of the linctus.August, 1982 SECOND AND FOURTH DERIVATIVE uv SPECTROPHOTOMETRY 1 .o a, 0.8 0 8 m n $ 0.6 0.4 a a 0.2 0 240 280 320 L , 240 280 3 883 0 Waveiengthinm Fig. 3. The zero order (a) and fourth order derivative (b) absorp- tion spectra of the acidic extract of the cough linctus containing pseudo- ephedrine, triprolidine and codeine.Linearity and Precision All the amplitudes in the second and fourth derivative spectra of ephedrine and pseudo- ephedrine show a proportional relationship with concentration up to at least twice that of the specified analytical concentration with correlation coefficients in the range 0.999 3- 0.999 99. The coefficients of variation of the results obtained in eight separate assays of Ephedrine Elixir BP, Paediatric Belladonna and Ephedrine Mixture BPC and the linctus containing pseudoephedrine hydrochloride, codeine phosphate and triprolidine hydrochloride, which were considered to represent the assay procedures involving second or fourth derivative measurement of sample solutions obtained by dilution or solvent extraction, were found to be 0.54, 1.1 1 and 1.03%, respectively, indicating satisfactory precision.Assay Results A number of ephedrine or pseudoephedrine formulations, either prepared extemporaneously in the laboratory according to compendia1 recipes or purchased locally, were assayed by the appropriate method. For comparison, the cornpendial formulations were also assayed by the official assay method~.l~-~O The results in Table I show that excellent recoveries of TABLE I ASSAY RESULTS Found as % of declared amount Declared Co-formulated drug and declared amount Formulation Drug* amount Ephedrine Elixir BP . . . . . . E 3 mgml-l - Ephedrine Hydrochloride Tablets BP Paediatric Belladonna and Ephedrine Mixture BPC .. . . . . Ephedrine Nasal Drops BPC . . Tablets . . . . . . . . Capsules . . . . . . . . Syrup . . . . . . . . . . Linctus . , , . . . . . .. E 30mg - . . E 1.5 mgml-l - .. E 5mg1n-l - .. E 25mg Aminophylline, 130 mg Amylobarbitone, 25 mg .. E 25mg Aminophylline, 130 mg Amylobarbitone, 25 mg . . PE 6 mg ml-' Triprolidine HCI, 0.25 mg ml-l . . PE 6 mg ml-' Triprolidine HC1, 0.25 mg ml-l Codeine phosphate, 2 mg ml-l Source? Corn Corn Ext Com Com Com Ext Corn Ext Com Com Com Corn Derivative Official method method 98.8 97.1 101.0 100.0 99.8 100.8 97.5 97.4 98.1 98.4 97.3 96.5 98.2 98.7 101.9 99.2 100.9 98.7 93.6 - 101.2 - 99.9 - 98.7 - * E = ephedrine hydrochloride ; PE = pseudoephedrine hydrochloride. t Com = commercial samples; Ext = extemporaneously prepared samples.884 DAVIDSON AND ELSHEIKH ephedrine were obtained in the extemporaneously prepared samples and that the commercial samples had levels of ephedrine or pseudoephedrine in good agreement with the label strengths and, where appropriate, with the results obtained by the official assay method.1. 2. 3. 4. 5. 6 . 7 . 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. References Abdine, H., Wahbi, A, M., and Korany, M. A., J . Pharm. Pharmacol., 1971, 23, 444. Taha, A. M., and Gomaa, C. S., J . Pharm. Sci., 1976, 65, 986. Abdine, H., El-Sayed, M. A., and Ibrahim, S. A,, Pharmazie, 1974, 29, 103. Di Fabrizio, F., J . Pharm. Sci., 1977, 66, 811. Smith, D. J., J . Assoc. OH. Anal. Chem., 1969, 52, 854. Wallace, J . E., J . Pharm. Sci., 1969, 58, 1489. Wallace, J . E., Anal. Chem., 1967, 39, 531. Manzoni, C., Boll. SOC. Ital. Farm. Osp., 1972, 18, 156; Anal. Abstr., 1973, 24, 3698. French, W. N., and Riedel, B. A., Can. J . Pharm. Sci., 1966, 1, 80. Dobrieky, J . , Proanalisis, 1969, 2, 122; Anal. Abstr., 1971, 20, 485. Lang, E., Fresenius 2. Anal. Chem., 1971, 253, 32; Anal. Abstr., 1971, 21, 4353. Fell, A. F., UV Spectrom. Group Bull., 1980, 8, 5. Fell, A. F., Proc. Anal. Diu. Chem. Soc., 1978, 15, 260. Jones, R., and Marnham, G., J . Pharm. Pharmacol., 1981, 33, 458. “Perkin-Elmer Model 552 UV - VIS Spectrophotclmeter : Operator’s Manual,” Perkin-Elmer Corp., O’Haver, T. C., and Green, G. L., Anal. Chem., 1976, 48, 312. “British Pharmacopoeia 1980,” Volume 2, HM Stationery Office, London, 1980, p. 555. “British Pharmacopoeia 1980,” Volume 2, HM Stationery Office, London, 1980, p. 764. “British Pharmaceutical Codex,” Tenth Edition, Pharmaceutical Press, London, 1973, p. 737. “British Pharmaceutical Codex,” Tenth Edition, Pharmaceutical Press, London, 1973, p. 757. Norwalk, CT. Received January 15th, 1982 Accepted April 26th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9820700879

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

Analyst, August, 1982, Vol. 107, pp. 885-895 Synthesis and Chromogenic Properties of Phthalazinohydrazones and Spectrophotometric 885 and Analogue Derivative Spectrophotometric Determination of Micro-amounts of Nickel with 5-Methylfurfural-1-phthalazinohydrazone Hajime Ishii, Tsugikatsu Odashima and Takuji Imamura" Chemical Research Institute of Non-Aqueous Solutions, Tohoku University, Katahira, Sendai-shi, 980 Japan The synthesis of two new hydrazones is described together with the results of spectrophotometric studies of their chromogenic reactions. One of them, 5-methylfurfural-1-phthalazinohydrazone (MFPH), has been utilised for the spectrophotometric determination of nickel a t the parts per million level. On the basis that nickel(I1) reacts with MFPH to form a stable 1 : 2 (metal: ligand) complex, having absorption maxima a t 470 and 503nm in 2.4% Triton S-100 solution and the solution of the complex gives a constant absorbance in the pH range 7.2-10.4, a sensitive and practical method for the determination of nickel has been proposed.The molar absorptivity and the sensitivity for an absorbance oP 0.001 are 3 . 7 x lo4 1 mol-l cm-l and 1.6 ng cm-?, respectively, the relative standard deviation for 15.3 p g of nickel (15 replicates) being 0.7%. Application of the proposed method to the determination of nickel in standard iron and steel samples and in chemicals and further sensitisation of the method by employing analogue derivative spectrophotometry are also described. Keywords : Nickel determination ; phthalazinol~ydrazone ; spectrophotometry ; derivative spectrophotovnetry Hydrazones have been used widely for the spectrophotometric determination of metal ions, and a large number of papers have reported this.The most studied examples have been derived from 2-pyridylhydrazine, 2-quinolylhydrazine, 1-isonicotinoylhydrazine, oc-thio- naphthoic hydrazide and benzoic hydrazide, these having been reviewed by Katyal et aZ.l We reported the synthesis of benzothiazolylhydrazones and their application to the micro- determination of metals, such as In this work two hydrazones derived from 1-phthalazinohydrazine, furfural-l-phthalazinohydrazone (FPH) and 5- methylfurfural-1-phthalazinohydrazone (MFPH) , have been synthesised and their usefulness as spectrophotometric reagents has been examined from the viewpoint of extending our synthetic and analytical studies.A sensitive and practical spectrophotometric and a very sensitive analogue derivative spectrophotometric method for the determination of nickel with MFPH have been developed. cobalt8 and n i ~ k e l . ~ Experimental Reagents All solutions were prepared with distilled, de-ionised water. In 30ml of 5% hydrochloric acid dissolve 0.01 mol of l-phthal- azinohydrazine, then add 10 ml of ethanol containing 0.01 mol of the corresponding aldehyde. Heat the solution at 65-70 "C on a steam-bath for 10 min. Allow to stand overnight, filter off the resultant crystalline product with suction, recrystallise from ethanol with 5% hydro- chloric acid and dry in zlacuo. FPH solution, 2.5 x M, and MFPH solution, 1.0 x M.Prepare by dissolving corresponding amounts of each hydrazone in 10% Triton X-100 solution. These solutions are further diluted with 10% Triton X-100 solution, if necessary. All reagents used were of analytical-reagent grade unless stated otherwise. Synthesis of hydrazones. * Present address : Chiba Works, Kawasaki Steel Corporation, Kawasaki-cho, Chiba-shi, 260 Japan.886 ISHII et al. : SPECTROPHOTOMETRIC AND DERIVATIVE Analyst, Vol. I07 Prepare by dissolving 2.5 g of nickel nitrate [Ni(N0,)2.6H20] in approximately 1 O O m l of water, add 1Oml of nitric acid (1 + 1) and dilute to 500 ml with water. This solution was standardised by EDTA titration with copper - l-(2-pyri- dylazo)naphth-2-01 (PAN) as an indicator. Prepare working solutions by diluting this solution with water .Standard nickel(l1) solution. Apparatus For measurements of the absorbance and the absorption spectrum, a Hitachi 139 spectro- photometer and a Hitachi 556 dual-wavelength spectrophotometer were used, respectively, the latter being used as an ordinary double-beam spectrophotometer throughout all the measurements. To obtain the derivative spectrum, a modified Hitachi 200-0576 derivative unit composed of two analogue differentiation circuits was connected between the latter spectrophotometer's output and a Hitachi 057 X - Y recorder's input. The details of this apparatus and the principle and characteristics of the analogue derivative spectrophotometry have been described previously.lo~l1 Procedure Ordinary spectrophotometry Place a sample or standard solution containing less than 40 pg of nickel(I1) in a 25-ml calibrated flask and add 6 ml of 1.0 x lod2 M MFPH in 10% Triton X-100 solution and 5 ml of ~ / 1 5 potassium dihydrogen phosphate - ~ / 1 5 sodium hydrogen phosphate buffer solution (pH 8).Heat the solution at about 70 "C on a steam-bath for 3 min, allow to cool to room temperature and dilute to the mark with water. Measure the absorbance of the resultant solution at 503 nm, in 10-mm cells, against a reagent blank as the reference. Second derivative spectrophotometry When the nickel content of the coloured solution prepared by the procedure described above is too low to give a measurable absorbance, record the second derivative spectrum at around 500 nm on a chart recorder against a reagent blank, by using a combination of both first- and second-order differentiation circuits of No.6 and a scan speed of 150 nm min-l or No. 5 and 240 nm min-l, and measure the second derivative value (vertical distance from a peak to a trough or that from the base line to a trough of the peak). Dissolution of iron and steel samples To about 50 mg of sample add 20 ml of aqua regia and 15 ml of 60% perchloric acid, heat to decompose the sample and evaporate the mixture to dryness. After cooling to room temperature dissolve the residue in 20 ml of hydrochloric acid (1 + l), heat for a few minutes to dissolve the soluble salts, filter into a 100-ml calibrated flask, and dilute to the mark with water after cooling to room temperature. Transfer a 25-ml aliquot of the solution into a separating funnel, add 10 ml of concentrated hydrochloric acid and 20 ml of 4-methylpentan- 2-one and shake for 3 min to remove the iron(II1).Transfer the aqueous phase into an evaporating dish, heat to remove the majority of the acid, transfer the contents of the dish into a 25-ml calibrated flask and dilute to the mark with water. Use 2 ml or an appropriate aliquot of the resultant solution for the determination. Results and Discussion Identification of the Hydrazones Identification of the synthesised FPH and MFPH were carried out taking the results of their elemental analysis and the infrared and nuclear magnetic resonance spectra into con- sideration. The infrared spectra of both FPH and MFPH, measured with potassium bromide tablets, showed absorption peaks assigned to a characteristic peak of an azomethine bond (-CH=N-) at 1625 cm-l.12 In addition, the nuclear magnetic resonance spectra, measured in chloroform-d, with tetra- methylsilane as standard, also showed a peak assigned to an azomethine bond at 8.1 p.p.m.13 On the basis of these results, the FPH and MFPH synthesised are presumed to have the structures shown in Table I.The results of the elemental analyses are given in Table I.August, 1982 SPECTROPHOTOMETRIC DETERMINATION OF NI TABLE I STRUCTURES AND PHYSICAL PROPERTIES OF HYDRAZONES Analysis,* yo 887 r Compound R Melting-point/"C Yield, % C H N c1 FPH . . . . R, 205 64 55.8 3.59 19.3 12.5 (decomposes) (56.8) (4.03) (20.4) (12.9) MFPH . . . . R, 239-241 68 57.6 4.37 19.6 12.6 (58.2) (4.54) (19.4) (12.3) * Figures in parentheses indicate calculated values.Solubility of the Hydrazones solvents tested 10% Triton X-100 solution was the best. The solubilities of FPH and MFPH in various solvents are shown in Table 11. From the TABLE I1 SOLUBILITIES OF FPH AND MFPH Solvent FPH* MFPH* - - Water . . . . .. .. .. Ethanol . . . . .. .. * - + + + Butan-1-01 . . . . .. .. * * + Benzene . . .. .. .. .. Chloroform . . .. . . * . + + f 4-Methylpentan-2-one . . . . . . 1,2-Dichloroethane .. .. 0 . f 2.5 x M Triton X-100 . . - . + + 2.5 x M SDSt . . . . * . f f 2.5 x M Zephiraminet . . . . 1- + - - - * +, Very soluble; f, appreciably soluble; -, scarcely t Sodium dodecylsulphate. $ Tetradecyldimethylbenzylammonium chloride. soluble. Reactivity of the Hydrazones with Metal Ions FPH and MFPH scarcely reacted or did not react with metal ions when their solutions prepared just before use were used, but both hydrazones tended to react after brief irradiation of the solutions with ultraviolet light from a mercury lamp.This phenomenon seems to be related to a photoequilibrium between two tautomers shown in equation (1). According to a preliminary experiment, the solutions of both hydrazones needed to be irradiated using a R +:\NHc\R H .Ha R1 5 -Q "/N R2 = QCH3 400-W mercury lamp for more than 15 min or exposed to daylight for a t least several days before use. The reactivities with metal ions of both hydrazones thus prepared in 2.4%888 ISHII et al. : SPECTROPHOTOMETRIC AND DERIVATIVE Analyst, Vol. 107' Triton X-100 solution at various pH values are summarised in Tables I11 and IV.Both hydrazones are found to react with cobalt (11), copper( I) and -( 11), mercury(II), nickel(I1) and palladium( 11) to give coloured complexes having fairly large molar absorptivities. The absorption maxima of the nickel(I1) complexes are separated from those of the others and the molar absorptivity is larger in the MFPH complex than in the FPH complex. This suggested the usefulness of MFPH for the sensitive and selective determination of nickel, so that a further study concerning this was carried out. TABLE I11 REACTIVITY OF FPH WITH METAL IONS pH 10 -7 hmax./nm c/1 mol-1 cm-1 Metal ion .. .. &(I) - Co(I1) . . .. Cr(II1) . . .. Cr(V1) . . . . Cu(1) . . .. Cu(I1) . . . . Fe(I1) . . . . Fe(II1) .. .. HgP) - * .. Hg(I1) . . .. Mn(I1) . . . . Ni(I1) . . .. Pb(I1) . . . . Pd(I1) . . .. Ti(1V) . . . . V(1V) .. .. V(V) .. .. Zn(I1) . . .. Cd(I1) . . Amax./nIn - - 448 - - 455 456 425 - - 438 499 478 - - - - - - TABLE IV ~ / 1 mol-1 cm-1 - - 20 600 - - 21 500 33 900 5 700 - - 12 800 26 200 33 400 - - - - - - - - 3 000 - - 19 300 28 600 6 800 - - - - 2 400 35 900 - - - - - - 29 600 - 30 600 35 700 5 300 - - 17700 28 100 34200 - - - REACTIVITY OF MFPH WITH METAL IONS PH 4 PH 7 pH 10 hmax./nm ~ / 1 mol-1 cm-1 445 3 800 447 27 600 - - - - - - 46 1 36 000 462 38 500 449 9 200 43 1 4 300 445 7 600 503 36800 482 35 000 - - - - - - - - - - - - - - \ ~ / 1 mol-1 cm-1 Metal ion Ag(I) * * .. Cd(I1) . . .. Co(I1) . . .. Cr(II1) . . .. Cr(V1) . . .. Cu(1) . . . . Cu(I1) .. . . Fe(I1) . . . . Fe(II1) . . . . H m . . .. Hg(1I) .. .. Mn(I1) . . .. Ni(I1) . . .. Pb(I1) . . .. Pd(I1) . . .. Ti(1V) . . .. V(1V) . . .. V(V) .. .. Zn(I1) . . .. - 452 - - 462 462 453 - 37 000 - - 462 24 000 462 36 000 - - - 37 000 38 500 6 200 - - 12 000 35200 38 000 - _- - - 429 503 482 - - - - - 482 38 000 - - Absorption Spectra The absorption spectrum of the nickel(I1) - MFPH complex is shown in Fig. 1 together with that of MFPH. The complex has two absorption maxima at 470 and 503nm, the latter wavelength being preferable for the determination of nickel because at that wavelengthAugust, 1982 SPECTROPHOTOMETRIC DETERMINATION OF NI 889 the molar absorptivity is larger and the absorption due to MFPH itself is barely observed. The absorption maxima and the shape of the spectrum of the complex did not vary in the pH range 4.2-11.0 and in the molar ratio range of nickel to MFPH of 1 : 20-20: 1, which suggests that only one species of the complex is formed under these conditions.0 Effect of pH Fig. 2 shows the effect of pH on the formation of the nickel(I1) - MFPH complex, from which it is seen that the complex begins to be formed at about pH 4, a constant and maximum absorbance being obtained in the pH range 7.2-10.4. I I d l 0.4 0.3 Q, C m 0 P * 0.2 a 0.1 0 0.4 0.3 a 0 C m +! 0.2 z 2 550 --! ' 450 500 Wavelengthinm n h V " - - Fig. 1. Absorption spectra of MFPH and its nickel(I1) complex in 4.8% Triton X-100 solution. Nickel(II), 612 A, nickel(I1) - MFPH complex against reagent blank; and B, reagent blank against water.p.p.b.; n I m i i , 2.4 x 10-3 M ; PH, 8.0; 0.4 0.3 a u m e 0.2 z a a 0.1 Fig. 2. Effect of pH: nickel(II), 612 p.p.b. ; MFPH, 2.4 x M ; Triton X-100, 4.8%; wavelength, 503 nm, reference, reagent blank. Effect of MFPH Concentration M MFPH solution is required for 15.3pg of nickel [which corresponds to more than about 50 times the nickel concentration (molar ratio)] to obtain a constant and maximum absorbance. As can be seen from Fig. 3, more than 5 ml of 1 x 0 2 4 6 8 Volume of I x lo-* M MFPH added/ml Fig. 3. Effect of MFPH concentra- tion: pH 8.0; other conditions are the same as those in Fig. 2.890 Analyst, VoZ. 107 Reaction Time and Reaction Temperature Fig. 4 shows the effect of reaction time on the complex formation at various temperatures.The reaction time required for the complete complexation depends upon temperature ; at 70 "C, a constant and maximum absorbance is obtained after only 3min, whereas at 50 and 30 OC, reaction times of more than 15 and 30 min are required, respectively. ISHII et aZ. : SPECTROPHOTOMETRIC AND DERIVATIVE 0.4 al C (c1 0.35 s 2 0.3 A B xAx' / X I 1 1 10 20 30 40 50 60 7C React ion ti me/ m in Fig; 4. Effect of reaction time. Temperature: A 70 "C, Other conditions are the B 50 C, C 30 "C and D 20 "C; pH, 8.0. same as those in Fig. 2. Effect of Triton X-100 Concentration As this work deals with the complexation in the Triton X-100 solution, the effect of the Triton X-100 concentration on the absorbance was examined by varying its concentration from 1.6 to 20% and keeping the nickel(I1) concentration at 612 p.p.b.(parts per lo9). According to the results, no variation in the absorbance was observed over the concentration range examined. Composition of the Complex and Equilibrium Constant of the Complexation Reaction The complexation equilibrium between nickel(I1) and MFPH (HL) may be expressed as K,, = [NiL,(n-2)-] [H+]"/[Ni2+] [HL]" . . .. ' * (3) If the initial concentration of the ligand, b, is much higher than that of the metal ion, a, equation (3) can be approximated as K,, = x[H+In/(a - x)bn . . .. .. - * (4) where x is the equilibrium concentration of the complex formed. Preliminary studies having revealed that the system obeys Beer's law additively over the wavelength range of interest, it can be deduced that where eL and eC are the molar absorptivities of the ligand and the complex at a specified wavelength and A and A , are the absorbances of the solution containing the metal complexAugust, 1982 SPECTROPHOTOMETRIC DETERMIEATION OF NI 891 and of the solution of the ligand alone, respectively.From equations (4), (5) and (6), equation (7) can finally be obtained: Equation (7) implies that a t a constant pH there should be a linear relationship between the terms a/(A - A,) and l/b", from which the value of n can be estimated. A linear relation- ship was experimentally obtained when assuming n is equal to 2, as shown in Fig. 5. Hence, the complexation reaction of nickel(I1) with MFPH may reasonably be expressed as Ke 4 Ni2+ + 2HL t - ~ NIL, + 2H+ . . .. .. . . (2') K,, = [NIL,] [H+I2/[Ni2+] [HLI2 .. .. .. . . (3') The value obtained from equation (7) and the experimental data shown in Fig. 5 by least- squares treatment is lo@,, = 9.5 for 0.3 M sodium perchlorate medium a t 25 "C. The stability constant of the complex could not be estimated because the acid dissociation constant of MFPH is below 10-13, so that a reliable value could not be obtained experi- mentally. ( i i b ~ x 105(n=2) 0 10 20 3.5 B s 3.0 2.5 I 5 10 (lib? x 102(n= 1 ) Fig. 5. Plots of a / ( A - .4.) veYsus l/bn. Ni(I1): [a] == 1.04 x lop5 M. MFPH: [b] = 8.8 x 1.0 x 1.2 x 1.4 x and 1.6 x M. Triton X-100, 4.8% ; pH, 5.5; ionic strength, 0.3 (NaC10,). Stability of the Complex absorbance remained constant even after 1 h. The complex formed under the recommended conditions was very stable, so that the Calibration Graph, Sensitivity and Precision recommended procedure.A good straight-line calibration graph passing through the origin was obtained using the The equation of the line obtained by least-squares treatment was Ni (p.p.m.) = 1.5* x A . . .. .. - (8) where A is the absorbance. The optimum range for the nickel determination was 0.2- 1.6 p.p.m., the sensitivity for an absorbance of 0.001 and the molar absorptivity calculated from the equation (8) being 1.5, ng cm-2 and 3.7, x lo4 1 mol-l cm-l, respectively.892 ISHII et aZ. : SPECTROPHOTOMETRIC AND DERIVATIVE mended procedure. The results gave a relative standard deviation of 0.7%. Analyst, VoZ. 107 Fifteen standard solutions containing 15.3 pg of nickel were analysed by the recom- Interferences from Other Ions and their Removal Solutions containing 15.3 pg of nickel(I1) and various amounts of other ions were prepared and the recommended procedure for the nickel determination was followed.The results are summarised in Table V, from which it can be seen that cobalt(II), chromium(III), copper(I1) and citrate interfere with the determination when no masking agent is added, whereas the other ions scarcely, or do not, interfere. TABLE V EFFECT OF OTHER IONS The amount of nickel(I1) taken was 15.3 pg. Ion Al(II1) . . Ca(I1) . . Cd(I1) . . Cr(1II) . . Fe(II1) . . Hg(I1) . . Mn(I1) . . A m * - Co(I1) . . Cu(I1) . . Mg(I1) . . V(1V) . . .. . . .. . . .. .. f . . .. .. .. Ni found/ Pg Precipitated 15.0 15.3 15.0 17.0 12.1 16.4 15.8 15.6 15.7 15.2 15.7 15.3 Relative error, % - 2.0 0.0 -2.0 10.9 - 20.9 7.2 3.3 2.6 2.6 -0.7 2.6 0.0 - Ion V(V) .. . . Zn(I1) . . . . Br- . . . . c1- . . . . c10,- * . . . I- . . .. NO,- . . . . P04,- . . .. s 0 4 2 - . . . . s20,2- . . .. Tartrate .. Citrate . . . . SCN- .. .. Amount added 100 PQ 100 P-Lg 80 mg 36 mg 100 mg 130 mg 60 mg 100 m g 60 m g 100 mg 220 mg 80 mg 100 mg Ni found/ P*.g 15.6 15.9 15.4 15.5 15.4 15.4 15.2 15.4 15.1 15.3 15.3 15.5 10.9 Relative error, yo 2.0 3.9 0.7 1.3 0.7 0.7 0.7 - 1.3 0.0 0.0 1.3 -28.8 -0.7 Interference from cobalt(I1) of up to about 15 pg could be removed by masking it as follows. To the sample solution are added 5 nil of ammonia solution and solid sodium perborate [more than 1000 times the amount of cobalt(I1) contained in the solution (molar ratio)] followed by warming the solution in order to assist the formation of the cobalt(II1) ammine complex.The sample solution is placed in a separating funnel, to which solid ammonium thiocyanate [more than 1200 times the amount of cobalt(I1) (molar ratio)] and 4 ml of 1 M acetic acid - 1 M sodium acetate buffer solution (pH 4) are added. The solution is shaken with 10ml of 4-methyl- pentan-2-one for 2-3 min to extract cobalt as the thiocyanate complex, and then the recom- mended procedure is applied to the resultant aqueous phase, which is then free from cobalt(I1). Interferences from copper(I1) and chromium(II1) could also be removed by masking them with sodium thiosulphate and tartrate, respectively, prior to addition of MFPH solution. The recommended procedure is then applied.Interference from large amounts of cobalt(I1) was removed as follows. TABLE 1 7 1 DETERMINATION OF NICKEL IN STANDARD IRON AND STEEL SAMPLES AND IN CHEMICALS Nickel found, yo Sample Iron and steel, JSS-162 Iron and steel, JSS-163 Cobalt nitrate .. Cobalt chloride . . Proposed Atomic-absorption method method 0.31 0.30 0.11 0.11 0.071 0.067 0.069 0.065 0.069 0.069 -7 Certified value 0.31 0.11August, 1982 SPECTROPHOTOMETRIC DETERMINATION OF NI 893 Application to Actual Samples In order to confirm the usefulness of the proposed method, it was applied to the deter- mination of nickel in iron and steel samples, of low cobalt, copper and chromium contents, and in chemicals. The results are summarised in Table VI together with results obtained by atomic-absorption spectrophotometry carried out for comparison. Sensitisation by Employing Analogue Derivative Spectrophotometry Ishii and co-workers have previously reported that derivative spectrophotometry using the analogue differentiation circuit is extremely effective for sensitisation of ordinary spectro- photometry.l09l1 As an example of sensitisation, the second derivative spectrophotometric determination of nickel with MFPH is described here.Selection of conditions for the measurement of the second derivative spectrum In second derivative spectrophotometry, the second derivative spectrum of the deter- minant is recorded on a chart recorder and the second derivative value (vertical distance from a peak to a trough or that from the base line to a trough of the spectrum) is measured instead of the absorbance measurement in ordinary spectrophotometry.As the second derivative value depends upon both the time constant of the analogue differentiation circuit used and the scan speed of the spectrophotometer, these need to be selected so as to give adequately a well resolved larger peak (these bring about good selectivity and higher sensitivity in the determination) while taking sharpness (half-width) of the absorption spectrum (zeroth derivative spectrum) into consideration. With our apparatus, six kinds of differentiation circuits with different time constants are represented with circuit numbers from 1 to 6, which can be selected easily and increase the circuit number meaning to that of the time constant.In general, a larger circuit number and/or a faster scan speed are preferable for a broad absorption spectrum. In Fig. 6 the second derivative spectra of the 0.8 0.4 Q, 3 m a, - .- c .- : 0.0 8 L a U U C 0.4 m 0.8 n ( a ) I F5J6 j,6 h I 1 - 450 500 550 450 500 E Wavelengthlnm I0 Fig. 6. Influence of (a) circuit number (with scan speed 150 nm min-l) and (b) scan speed (with circuits all on No. 5) on second derivative spectra of nickel(I1) - MFPH complex solution. Ni(II), 198 p.p.b.; MFPH, 2.4 x Numerical values indicate first and second differentiation circuit numbers in ( a ) and scan speed in ( b ) , respectively. M ; pH, 7.5; Triton X-100, 4.8%; reference, reagent blank.894 ISHII d al. : SPECTROPHOTOMETRIC AND DERIVATIVE Analyst, Vd.107 nickel(I1) - MFPH complex solution measured whilst varying the circuit number or the scan speed are shown, from which a combination of circuit No. 6 and a scan speed of 150 nm min-1 or circuit No. 5 and a scan speed of 240 nm min-l are found to be preferable for the deter- mination of nickel from the viewpoints of higher sensitivity and better resolution, i.e., better selectivity . Calibration graphs The calibration graph prepared by plotting the second derivative value zleysus the nickel concentration gave a straight line passing through the origin when the peak to trough value or the base line to trough value was plotted. The equation for each graph measured with a combination of circuit No. 6 and a scan speed of 150 nm min-l was .. * - (9) Ni (p.p.b.) = 115 x D .. .. .. Ni (p.p.b.) = 210 x D . . . . . . .. . . (10) respectively, that measured with a combination of circuit No. 5 and a scan speed of 240 nm min-l being Ni (p.p.b.) = 140 x D . . .. . . .. , . (11) Ni (p.p.b.) = 236 x D . . .. .. .. . . (12) respectively, where D is the second derivative value represented by the value being con- verted into absorbance. The cali- bration graphs for equations (10) and (12) adopted the base line to trough value as the second An example of a calibration graph is shown in Fig. 7. 0.20 0.15 - Q - 0) - 8 4 8 0.10 f s 0) .- c .- U U cn 0.05 0.00 r5 5 10 15 20 Nickel concentration, p.p.b. Fig. 7. An example of the calibration graph in the second derivative spectrophotometry. Circuits, all No. 6 ; scan speed, 150 nm min-l; recorder sensitivity, x 5 ; reference, reagent blank; A, peak to trough values are plotted; and B, base line to trough values are plotted.Ni concentrations: 1, 1.98; 2, 3.97; 3, 7.92; 4, 11.9; and 5, 19.8 p.p.b.August, 1982 SPECTROPHOTOMETRIC DETERMINATION OF NI 895 derivative value. Although their sensitivities are fairly low, they are preferable where the procedure is applied to the analysis of actual samples, because the second derivative value is less affected by coexisting ions, which, by the complexation with MFPH, give derivative peaks near t o that of the nickel complex. As the second derivative value, D, of 0.1 in the calibration graphs corresponds to a signal of 10 cm on a chart in this work, nickel at the parts per billion level can be determined easily by the proposed method.Characteristics of the Proposed Methods I t reacts with many metal ions to form complexes, but the absorption band of the nickel(I1) complex is clearly distinguished from those of other metal complexes. Only chromium(III), cobalt(I1) and copper(I1) interfere with the nickel determination, but their interferences can easily be removed as described. Therefore, the proposed methods are very selective. In addition, the spectrophotometric sensitivity (1.6 ng cm-2) is superior to that (4.2 ng cm-2) of the representative method with dimethyl- glyoxime for nickel determination.14 Furthermore, by employing analogue derivative spectrophotornetry the sensitivity could be increased considerably. Therefore, it is expected that the proposed methods could be applied widely to the determination of micro-amounts of nickel in various kinds of practical samples. MFPH can easily be synthesised. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Katyal, M., and Dutt, Y., Talanta, 1975, 22, 151. Odashima, T., and Ishii, H., Nippon Kagaku Kaishi, 1973, 729. Odashima, T., and Ishii, H., Nippon Kagaku Kaishi, 1975, 83. Odashima, T., and Ishii, H., Anal. Chim. Acta, 1975, 74, 61. Ishii, H., and Odashima, T., Nippon Kagaku Kaishi, 1975, 1332. Odashima, T., and Ishii, H., Anal. Chim. Acta, 1976, 83, 431. Odashima, T., Anzai, F., and Ishii, H., Anal. Chim. Acta, 1976, 86, 231. Odashima, T., and Ishii, H., Bunseki Kagaku, 1977, 26, 678. Odashima, T., Satoh, S., and Ishii, H., Nippon Kagaku Kaishi, in the press. Ishii, H., and Koh, H., Nippon Kagaku Kaishi, 1980, 203. Ishii, H., and Satoh, K., Fresenius 2. Anal. Chem., in the press. Silverstein, R. M., and Bassler, G. C., “Spectrometric Identification of Organic Compounds,” Second Edition, John Wiley, New York, 1967 (translated by Araki, S., and Mashiko, Y., Tokyo Kagaku Dohjin, Tokyo, 1969). Tsujikawa, T., Mizuta, E., and Hayashi, M., Yakugaku Zasshi, 1976, 96, 125. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience, New York, 1959, p. 669. Received January 27th, 1982 Accepted March 16th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9820700885

出版商:RSC    

年代:1982

数据来源: RSC