摘要:

538 Autalyst, June, 1979, Vol. 104, PP. 538-544 Determination of Nitrogenous Gases Evolved from Soils in Closed Systems C. J. Smith and P. M. Chalk School of Agriculture and Forestry, University of Melbourne, Parkville, Victoria 3062, Australia A simple method is described for determining nitrogen oxide and nitrogen dioxide, evolved from soils, in closed systems. These gases are absorbed by an acidic solution of potassium permanganate, and the resulting nitrate is determined by a steam distillation method. Excess of permanganate is reduced with iron(I1) sulphate and neutralised with sodium hydroxide solution. Ammonium in solution is removed by distillation with magnesium oxide, and nitrate is determined by distillation after reduction to ammonium by Devarda's alloy. Nitrogen and dinitrogen oxide evolved from soils are measured using gas chromatography on a single 0.61-m column of molecular sieve SA, tempera- ture programmed to 250 "C a t 39 "C min-l, after an initial period of 1 min at 35 "C.A complete analysis requires 19.5 min, and 2 pg of nitrogen can be determined quantitatively for each gas. Keywords ; Acidic fiermanganate ; gas chromatogvaphy ; nitrogenous gases ; steam distillation ; soils Both indire~tl-~ and direct2YH evidence indicate that gaseous forms of nitrogen can be lost from soil during the nitrification of ammonium or ammonium-forming fertilisers by soil micro-organisms. It appears that evolution o:E nitrogen, dinitrogen oxide and nitrogen oxide or its oxidative derivative, nitrogen dioxide, can occur, resulting in poor fertiliser efficiency.Methods suitable for measuring such losses from soil in closed systems have not been firmly established. A review of methods developed for work in other areas9 reveals many diverse techniques. Of these, it appears that nitrogen oxide and nitrogen dioxide would be determined most readily by chemical means, and nitrogen and dinitrogen oxide by gas chromatography. Nitrogen oxide and nitrogen dioxide can be albsorbed in an acidic oxidising reagentg and determined as nitrate. A variety of oxidants and methods of nitrate determination have been used. The determination of soil nitrate by steam distillation using magnesium oxide and Devarda's alloy as reductant is a widely accepted method for neutral salt extracts.10 It would therefore be advantageous if this method could be modified and used.The gas-chromatographic separation of nitrogen and dinitrogen oxide must allow for the presence in the closed system of the oxygen and carbon dioxide required by the autotrophic nitrifying bacteria. Dual columns have generally been used to resolve these four gases. Disadvantages include the need for dual injectjonsll or sample splitting12 for columns in parallel, or the need for effluent switchingls and polarity reversaP3-16 for columns in series. Separation has been achieved on single columns of a molecular sieve16 or charcoall7 using temperature programming, or on a long (5.5 m) column of Porapak Q using a small samplels or low temperature.lQ The purpose of this paper is to report methods developed in our laboratory for determining nitrogenous gases evolved from soils in closed systems.A rapid steam distillation procedure for nitrogen oxide and nitrogen dioxide absorbed by acidic permanganate solution, and a convenient gas-chromatographic procedure that permits the resolution of nitrogen, dinitrogen oxide, oxygen and carbon dioxide on a single column, are described. Experimental Apparatus and Materials Steam distillation apparatus. Distillation $asks, 250 ml. Side-arm type, as described by Bremner and Keeney.lo Microbwette, 5 ml. Becker gas chromatograph. Packard, Model 410. As described by Bremner and Keeney.10 Metrohm, piston type, 0.005-ml graduations.SMITH AND CHALK 539 Automatic linear temperature programmer. Thermal conductivity detector. Mass $ow controller.Strip-chart recorder. Columns. Column packing material. Carrier gas. Helium; Commonwealth Industrial Gases, pure compressed. Gas syringes, 25, 250 and 1000 pl. Precision Sampling, Series A2 Pressure-lok. Calibration gases. Standard septum bottles, capacity 125 ml. Precision Sampling, borosilicate glass. Mininert valves. Packard, Model 757. Packard, Model 702. Packard, Model 784. Riken Denshi, single pen. Stainless steel, 0.61 m x 5 mm i.d. Molecular sieve 5A, 100-120 mesh. Nitrogen, dinitrogen oxide, nitrogen oxide, nitrogen dioxide, oxygen, carbon dioxide and carbon monoxide ; Commonwealth Industrial Gases, pure compressed. Precision Sampling, for standard septum bottles. Reagents Potassium fiermanganate solution, 0.2 M in 0.5 M sulphuric acid, Iron(I1) sulphate solution, 0.5 M in 1 M sulphuric acid.Sodium hydroxide solution, 2 M. Magnesium oxide, heavy. Prepared as described by Bremner and Keeney.lo Boric acid - indicator solution. Devarda's alloy powder. Merck. Sulphuric acid, 0.0025 M. Standard solution. Ammonium sulphate solutions. Four standard solutions containing 49-505 pg ml-l of Potassium nitrate solutions. Five standard solutions containing 99-991 pg ml-l of Ammonium sulphate - potassium nitrate solutions. Three standard solutions containing Prepared as described by Bremner and Keeney.lo nitrogen. nitrogen. 99494 pg ml-l of ammonium-nitrogen and 148-715 pg ml-1 of nitrate-nitrogen. Chemical Procedure Prepare the distillation apparatus for use by flushing with steam for 10 min. Adjust the distillate collection rate to 7-8 ml min-1.Add 5 ml of boric acid - indicator solution to a 60-ml Erlenmeyer flask and place it under the condenser. Transfer 3-5ml of potassium permanganate solution containing up to 1 mg of nitrogen into a 250-ml side-arm distillation flask. Rinse the container with 3 ml of iron(I1) sulphate solution and two 5-ml portions of distilled water and transfer the rinsings into the distillation flask. To determine ammonium, neutralise the solution with 5-6 ml of 2 M sodium hydroxide solution and add 0.6g of heavy magnesium oxide with a calibrated spoon. Attach the flask to the apparatus. Begin distillation and continue until 35ml of distillate have been collected, after 4-5min. Titrate the distillate from a green to a permanent, faint pink colour with standard 0.0025 M sulphuric acid.To determine nitrate, remove the stopper from the side-arm of the flask and add 1.2 g of Devarda's alloy with a calibrated spoon. Replace the stopper and follow the distillation and titration procedure described previously. Determine ammonium and nitrate in a reagent blank and subtract the result from the sample titre. Calculate ammonium-nitrogen or nitrate-nitrogen on the basis of 1 ml of 0.0025 M sulphuric acid = 70 pg of nitrogen. Gas-chromatographic Procedure Prepare the gas chromatograph for use by setting the following controls. Carrier gas Cylinder outlet pressure .. .. .. 400kPa Mass flow controller outlet pressure . . .. 200kPa Flow-rate (approximate) .. .. .. 50mlmin-l Temperatures Injection ports . . .. .. .. ..90°C Detector . . .. .. .. .. .. 250°C540 SMITH AND CHALK: DETERMINATION OF NITROGENOUS Analyst, Vd. 104 Columrt tem+eratu.re (programme Initial period . . .. .. .. .. Temperature rise .. .. . . .. Final period . . .. .. .. .. Cooling period . . .. .. .. .. Stabilising period. . .. .. .. .. Bridge current . . .. .. .. .. Detector Signal attenuation . . .. .. .. 1 min at 35 "C 39 "C min-l for 5.5 min 5 min at 250 "C 6 min 2 min 200 mA (250 mA maximum at 250 "C for helium carrier gas) Variable between x l and x16 Recorder Range .. .. .. .. .. 5mV Chart speed' . . .. .. .. .. 30mmmin-l Calibrate the Rotaineters of the mass flow controller by adjusting the needle valves to vary the flow-rates, which are measured with a soap-film flow meter at the detector outlets.Equalise the flow-rates in the reference and separating columns by setting the Rotameters. Determine the optimum flow-rate of the carrier gas by determining the minimum height equivalent to a theoretical plate for the oxygen or nitrogen peaks, for flow-rates varied between 15 and 75 ml min-1. Quantitative determination of nitrogen or dinitrogen oxide is based on the measurement of peak area when known amounts of pure gas are injected. Results and Discussion Chemical Procedure Several major considerations are essential to the successful application of the proposed procedure to the determination of nitrogen 0xid.e and nitrogen dioxide evolved from soils. Because the method is based on the determination of nitrate formed in the acidic perman- ganate solution, quantitative recovery of nitrate is required.As nitrate is determined by reduction to ammonium, ammonium in the solution arising from absorption of ammonia must be removed quantitatively beforehand. Absorbed ammonia must not interfere in the determination through oxidation of ammonium in acidic permanganate solution to nitrate during the period of gas absorption. The capacity of the solution to absorb and oxidise nitrogen oxide and nitrogen dioxide evolved from soils must not be exceeded. Quantitative recovery of 251 pg of ammonium-nitrogen .added to acidic permanganate solution is obtained when 0.4-1.0 g of magnesium oxide is used (Table I). Accurate weighing of the magnesium oxide is therefore not necessary and it is recommended that a spoon calibrated to hold approximately 0.6 g be used.Nitrate added to acidic permanganate solution (247 Jug of nitrogen) is recovered quantitatively when 1.0-1.6 g of Devarda's alloy powder is added following distillation with magnesium oxide (Table 11). Again, a calibrated spoon holding approximately 1.2 g can be used for convenience. Using the developed procedure, quantitative recovery of a range of additions of ammonium TABLE I EFFECT OF THE AMOUNT OF MAGNESIUM OXIDE ON THE RECOVERY OF 251 Jug OF 5 ml of 0.2 M potassium permanganate in 0.6 M sulphuric acid were treated with 1 ml of ammonium sulphate solution. AMMONIUM-NITROGEN ADDED TO ACIDIC PERMANGANATE SOLUTION Amount of Amount of ammonium-nitrogen Coefficient of magnesium oxidelg recovered* /pg variation, yo Recovery, % 0.2 244 0.2 97 0.4 248 0.6 99 0.6 250 0.3 100 0.8 249 0.2 99 1.0 250 0.2 100 * Mean of four replicate determinations.June, 1979 GASES EVOLVED FROM SOILS I N CLOSED SYSTEMS 541 TABLE I1 EFFECT OF THE AMOUNT OF DEVARDA’S ALLOY POWDER ON THE RECOVERY OF 247 pg OF NITRATE-NITROGEN ADDED TO ACIDIC PERMANGANATE SOLUTION 5 ml of 0.2 M potassium permanganate in 0.5 M sulphuric acid were treated with 1 ml of potassium nitrate solution.Amount of Amount of Devarda’s nitrate-nitrogen Coefficient of alloy powderlg recovered*/pg variation, yo Recovery, yo 0.4 7 22 3 0.8 240 1.0 97 1.0 246 0.9 100 1.2 246 0.7 100 1.6 246 0.5 100 * Mean of four replicate determinations. or nitrate added alone to acidic permanganate solution is achieved (Table 111). The results illustrate that the method is precise and accurate, When ammonium and nitrate are added together, quantitative recoveries of each are obtained during a period of standing of up to 10 d (Table IV).These results confirm the absence of mutual interference over a range of nitrogen additions. Over a period of 96 h, the absorption and oxidation of nitrogen oxide by 3 ml of acidic permanganate solution in a vial (surface area 0.8 cm2) is quantitative only up to 29 pg of nitrogen added (Table V). In contrast, nitrogen dioxide is readily absorbed, with up to 500 pg of nitrogen being quantitatively recovered in solution under the same conditions. Nitrogen oxide may be partially absorbed as nitrogen dioxide (2N0 + 0, = 2N0,). How- ever, as sufficient oxygen was present and nitrogen dioxide is readily absorbed, the oxidation must be relatively slow.The rate of absorption and oxidation of nitrogen oxide (5NO + 3Mn04- + 4H+ = 5N03- + 3Mn2+ + 2H,O) is limited by the surface area of the acidic permanganate solution. It was found that 3 ml of solution would quantitatively absorb 200 pg of nitrogen oxide-nitrogen in 96 h when it was placed in the bottom of the standard septum bottle (surface area 15 cm2). The limitation of nitrogen oxide absorption by use of 3 ml of solution in a vial does not appear to have been a problem in practice, Nitrogen oxide has not been measured in the atmosphere above soil by gas chromatography, using the method described in this paper (2 pg of nitrogen oxide-nitrogen can be determined with good precision). The highest amount of nitrogen measured in the permanganate solution was 400pg evolved from soil over 8 d.It is recommended, however, that the surface area of the permanganate solution TABLE I11 RECOVERY BY THE DEVELOPED PROCEDURE OF AMMONIUM-NITROGEN OR NITRATE-NITROGEN ADDED ALONE TO ACIDIC PERMANGANATE SOLUTION Amount of Amount of nitrogen nitrogen Coefficient of Component added*/pg recoveredtlpg variation, yo Recovery, y-, Ammonium-nitrogen . . .. 49 48 2.7 98 101 101 0.9 100 251 25 1 0 100 505 503 0.1 100 Nitrate-nitrogen . . .. .. 99 97 2.4 98 247 247 0.2 100 493 484 0.2 98 738 729 0.1 99 991 975 0.1 98 * 5 ml of 0.2 M potassium permanganate in 0.5 M sulphuric acid were treated with either 1 ml of ammonium sulphate solution or 1 ml of potassium nitrate solution containing the amounts of nitrogen indicated.t Mean of three replicate determinations.542 SMITH AND CHALK: DETERMINATION OF NITROGENOUS AnaZyst, Val. 104 TABLE IV EFFECT OF STANDING TIME ON THE RECOVERY BY THE DEVELOPED PROCEDURE OF AMMONIUM-NITROGEN AND NITRATE-NITROGEN ADDED TOGETHER TO ACIDIC PERMANGANATE SOLUTION Amount of nitrogen added*/ p g r A I - 3 d 7 d 10 d Added as Addedas p----h---q r - ~ - , NH4+ NO,- NH4+ NO,- NH,+ NO,- NH,+ NO,- 99 148 99 99 97 98 99 99 251 364 98 99 98 99 97 99 494 715 100 100 100 100 100 100 3Xecovery after standing,? % * 5 ml of 0.2 M potassium permanganate in 0.5 M sulphuric acid were treated with 1 ml of ammonium sulphate - potassium nitrate solution containing the amounts of nitrogen indicated. t Standing a t 30 OC. Mean of three replicate (determinations.be maximised, as situations may arise where t:he rate of nitrogen oxide evolution may be high. Ammonium and nitrate in an acidic permanganate solution have previously been deter- mined by steam distillation.20 Ammonium is determined by distillation with magnesium oxide, and nitrate is determined by distillation after reduction to ammonia by iron(I1) hydroxide. The reduction is catalysed by silver sulphate, and iron(I1) hydroxide is formed by treatment of iron(I1) sulphate with magnesium oxide. The procedure involves volu- metric dilution of an acidic permanganate solution, and distillation of separate aliquots for ammonium, and nitrate + ammonium, nitrate being determined by difference. This pro- cedure therefore requires more work than the proposed procedure.More important, the determination of nitrate by difference can be criticised because the experimental error is the sum of the errors associated with the two determinations. In the developed procedure, heavy magnesium oxide is used instead of sodium hydroxide to make the neutralised solution alkaline, in order to avoid loss of sample caused by excessive foaming during subsequent distillation with Devarda's alloy. This problem has been encountered in the development of steam distillation methods for determination of inorganic nitrogen (nitrite + nitrate) in alkaline permanganate solution^.^^-^^ Glass-wool can be used in the spray trap of the distillation apparatus to prevent foaming 1oss,21s22 but the need to replace it and to clean the trap after each distillation is undesirable.More recent method^^^^^^ have employed other reduct ants. TABLE: V ABSORPTION OF NITROGEN OXIDE OR NITROGEN DIOXIDE BY ACIDIC PERMANGANATE SOLUTION I N 96 h Compound Nitrogen oxide . . . . Nitrogen dioxide . . .. Amount of gas inj ected*/ p1 10 15 50 75 175 250 350 Amount of nitrogen injected/p.g 6 9 29 43 101 143 200 Amount of nitrogen absorbedt/pg 6 9 29 34 67 67 103 Coefficient of variation, yo 10.2 6.9 2.6 4.5 1.5 0.7 2.4 Recovery in solution, yo 100 100 100 79 66 47 52 275 158 156 0.7 99 450 258 260 1.3 101 650 373 374 2.2 100 900 517 517 0.8 100 * Atmospheric pressure, 25 "C. The gas was injected into a standard septum bottle sealed with a t Absorbed by 3 ml of 0.2 M potassium permanganate in 0.5 M sulphuric acid contained in a 4-ml vial Mininert valve.placed inside the standard septum bottle. Mean of three replicate determinations.June, 1979 GASES EVOLVED FROM SOILS IN CLOSED SYSTEMS 543 The developed procedure could be modified to provide a convenient method for deter- mining inorganic nitrogen in alkaline permanganate solutions used for absorbing nitrogen oxide and nitrogen dioxide evolved from soils. As ammonia is not absorbed, only a single distillation is necessary. The alkaline permanganate solution could be neutralised with acidified iron(I1) sulphate, prior to the addition of magnesium oxide and Devarda's alloy together. Otherwise the method would remain the same. Although useful for some purposes, alkaline permanganate cannot be used in closed-system soil studies involving nitrification , because of absorption of carbon dioxide. Gas-chromatographic Procedure Complete resolution of oxygen , nitrogen, dinitrogen oxide and carbon dioxide is achieved under the chromatographic conditions specified (Fig.1). Retention times relative to carbon monoxide (reference compound) are indicated. In the absence of oxygen, nitrogen oxide can also be separated on this column. Argon is not separated from oxygen on this column, but if necessary it could be eliminated from the artificial soil atmosphere by using argon-free oxygen. Time/min Fig. 1. Chromatogram of gas mixture (relative retention times in parentheses). The total retention time is approximately 8 min, but an additional 3.5 min are required in order to ensure total elution of carbon dioxide, which exhibits tailing.A cooling period of 6 min and a stabilising period of 2 min are also required, making a total analysis time o f 19.5 min. The retention and elution times compare favourably with those in other published methods of temperature-programmed separation of these gases on a single column. For example, the total retention time is approximately 22 min for a 3-m column of molecular sieve 5A, temperature programmed to 400°C after an initial period of 3min at room temperature.16 For a 3-m column of acid-washed coconut charcoal, temperature pro- grammed to 175 "C after an initial period of 5 min at 40 "C, the total elution time is approxi- mately 12 min.17 The precision of the method permits the quantitative determination of 2 pg of nitrogen as dinitrogen oxide or nitrogen (Table VI).The sensitivity could be increased, if necessary, by increasing the cell current to 250 mA, or by decreasing the recorder range below 5 mV. For qualitative work based on peak-height measurements, 0.07 pg of dinitrogen oxide- nitrogen or 0.04 pg of nitrogen can be detected (2.5 mm response) using a cell current of 200 mA and a recorder range of 1 mV. This is similar to the sensitivity obtained by Burf ord .13 An important advantage of the method is the use of a single, short column for separation. The temperature programme is pre-set, and will function automatically following injection and initiation. This is an advantage as the apparatus can be left unattended. A stable base line is obtained through the use of mass flow controllers to stabilise flow-rates.544 SMITH AND CHALK TABLE VI EFFECT OF AMOUNT OF NITROGEN ON THE PRECISION OF THE DETERMINATION OF NITROGEN AND DINITROGEN OXIDE Amount of nitrogen & due to variation (yo) for inj ected*/pg Responsef (mm2) Coefficient of Injected Injected r-A---7 7- asN, asN,O N2 N2O N2 N,O 2 2 24 17 5.7 9.4 3 3 89 73 4.3 6.0 6 7 191 22 1 1.3 2.0 9 14 36 1 698 1.3 0.9 * l-ml sample.t Mean of five replicate determinations. Attenuation x 1. Recorder range 6 mV. Peak area was determined manually as the triangular area enclosed by peak ta:ngents and peak base line. The only single column alternative to a temperature-programmed method is the use of long columns of a material such as Porapak Q. To obtain resolution of oxygen and nitrogen, a small sample and possibly a low-temperature: accessory would be req~ired.l*,~~ A small sample would then require the use of a more sensitive detector such as the helium ionisation detector.The disadvantage is that this detector is not a common item of equipment on a gas chromatograph, unlike the thermal conductivity detector. Another alternative to the proposed method is the use of dual columns in series. The advantage is a shorter analysis time,l4 but some negative peaks will be obtained. This would require manual polarity reversal during separation, expensive electronic integration or use of the middle of the chart as the base line, thereby decreasing the available range. This research was supported, in part, by a Commonwealth Postgraduate Research Award to the senior author. 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. References Gerretsen, F. C., and de Hoop, H., Can. J . Microbiol., 1957, 3, 359. Wagner, G. H., and Smith, G. E., Soil Sci., 1958, 85, 125. Soulides, D. A., and Clark, F. E., Proc. Soil Sci. Soc. Am., 1958, 22, 308. Clark, F. E., Beard, W. E., and Smith, D. H., PYOC. Soil Sci. SOC. Am., 1960, 24, 50. Khan, M. F. A., and Moore, A. W., Soil Sci., 1968, 106, 232. Schwartzbeck, R. A., MacGregor, J. M., and Schmidt, E. L., Proc. Soil Sci. SOC. Am., 1961, 25, 186. Meek, B. D., and MacKenzie, A. J., Proc. Soil Sci. Soc. Am., 1965, 29, 176. Steen, W. C., and Stojanovic, B. J., Proc. Soil Sci. SOG. Am., 1971, 35, 277. Cheng, H. H., and Bremner, J. M., in Black, C. A., Evans, D. D., White, J. L., Ensminger, L. E., and Clark, F. E., Editors, “Methods of Soil Analysis,” American Society of Agronomy, Madison, Wisc., 1965, Part 2, p. 1287. Bremner, J. M., and Keeney, D. R., Analytica Claim. Acta, 1965, 32, 485. Herbert, R. A., and Holding, A. J., J . Chromat. Sci., 1972, 10, 174. Smith, K. A., and Dowdell, R. J., J . Chromat. Sci., 1973, 11, 655. Burford, J. R., J . Chromat. Sci., 1969, 7 , 760. Bennett, D., J . Chromat., 1967, 26, 482. Blackmer, A. M., Baker, J. H., and Weeks, M. E., Proc. Soil Sci. SOC. Am., 1974, 38, 689. Graven, W. M., Analyt. Chem., 1959, 31, 1197. Stevenson, F. J., and Harrison, R. M., Proc. Soil Sci. Soc. Am., 1966, 30, 609. Wilhite, W. F., and Hollis, 0. L., J . Gas Chromai!., 1968, 6, 84. Douglas, L. A., and Bremner, J . M., Soil Biol. Biochem., 1971, 3, 209. Bundy, L. G., and Bremner, J. M., Commun. Sod. Sci. Plant Anal., 1973, 4, 179, Kieselbach, R., Ind. Engng Chem., 1944, 16, 764. Kieselbach, R., Ind. Engng Chem., 1944, 16, 766. Tedesco, M. J., and Keeney, D. R., Commun. Soil Sci. Plant Anal., 1972, 3, 339. Nelson, D. W., and Bremner, J. M., Soil Biol. Biochem., 1970, 2, 203. Received July 12th, 1978 Accepted December 29th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400538

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst, June, 1979, VoZ. 104, $9. 545-551 545 Rapid Method for Determining Fluoride in Vegetation Using an Ion-selective Electrode Alberto Enrique Villa Environmental Research Department, A L UAR Aluminio Argentino SA I C , 9120 Puerto Madryn, Chubut, A rgentina Fluoride was extracted from dried vegetation by stirring with 0.1 N perchloric acid for 20 min at 20 "C. The fluoride content was determined in this extract (pH 1) using the method of standard additions, thus eliminating the need to de-complex fluoride prior to analysis. The presence of up to 2.0% of silicon, 0.06% of iron, 0.1% of aluminium, 0.7% of magnesium and 1.2% of calcium did not result in any interferences and recoveries of 98-102y0 were obtained. The fluoride contents of standard samples determined by this method were highly correlated (r = 0.999) with those obtained by reference methods over the range 4-2000 pg g-l of fluoride in the dry matter.Keywords ; Fluoride determination ; jhoride ion-selective electrode ; fluoride an vegetation ; perchloric acid extraction. A study of the influence of the emissions from an aluminium smelter on the surrounding terrestrial and marine environment has stimulated research into the methods used for the determination of fluoride in samples of sea water, vegetation and certain other materials1 The methods currently used for the determination of fluoride are complex and time con- surning2 Since the introduction of the fluoride ion-selective electrode (FISE) ,3 many papers have been devoted to the development of rapid, simple methods for determining fluoride and an excellent review has been published.4 In this paper, a re-examination of current methods using the fluoride electrode to deter- mine fluoride in vegetation is made, and a rapid technique, which circumvents certain limitations of these methods, is reported.Efforts to improve the speed and simplicity of analytical methods for the determination of fluoride in vegetation have resulted in the use of direct extraction of dry, pulverised or ashed tissue, followed by potentiometric analysis. Different techniques have been used, including acid or acid - base extraction at room temperatures' or acid or base digestion at an elevated temperature.+1° The main criticism of direct extraction has been its relatively poor accuracy as simple extraction procedures, which do not involve alkali fusion, have occasionally resulted in low value^.*^^^^^^ However, if the methods are re-examined and certain modifications are introduced, it can be shown that the combination of an acid-extraction step and use of the method of standard additions at pH 1 yields a simple, fast and reliable method for the determination of fluoride in dried vegetation over the range 4-2000 pg g-l.Experimental Apparatus The potential measurements were made using an Orion Research, Model 801-A, digital pH - millivolt meter equipped with an Orion, Model 94-09A, fluoride ion-selective electrode and an Orion, Model 90-01, reference electrode. The pH measurements were made using an Orion, Model 91-02, glass electrode. X-ray fluorescence data were obtained with a Philips, Model PW1450, X-ray fluorescence spectrometer.Reagents sodium fluoride (dried at 110 "C for 2 h) in de-ionised distilled water and dilute to 1 1. (sp. gr. = 1.52) to 1 1 with de-ionised distilled water. Fhoride standard solution, 2000 mg 1-l. Dissolve 4.421 g of analytical-reagent grade Perchloric acid, 0.1 N. Dilute 10.3 ml of 62% analytical-reagent grade perchloric acid Plastic laboratory ware is recommended for storage and handling of reagents and samples.546 Analyst, VoZ. 104 Procedure Dry the foliage for 24-48 h at 80 "C, then grind it to pass through a number 60 sieve and store in clean, dry, tightly closed plastic bottles. Rotate the bottle to mix the sample thoroughly before removing aliquots. Weigh about 1.0 g of powdered sample accurately (or 0.25 g if the sample contains more than 100 pg g1 of fluoride) and place it in a 100-ml wide-mouth plastic container.Add 25 ml of 0.1 N perchloric acid and stir magnetically for 20 min. Add an additional 25 ml of 0.1 N perchloric acid and insert the fluoride and reference electrodes while continuing to stir. Wait for reading E, to stabilise (usually not more than 5 min). Add 0.2 ml of the fluoride standard solution (2000 mg 1-l) from a microburette or a 0.250-ml Hamilton-type syringe. Wait for a. stable reading, E,. Determine the fluoride concentration in the sample using equation (4) (see Theory). Measurements should be carried out at constant temperature (&2 "C). VILLA : RAPID METHOD FOR DETERMINING FLUORIDE Discussion of Current Methods Calibration Graph Procedure Calibration graphs for the determination of fluoride concentration in vegetation have been almost always used in this type of a n a l y s i ~ ~ ~ 6 J ~ ~ ~ ~ as this procedure is rapid and straight- forward.There are, however, certain limitations associated with this technique. (i) The final result is obtained by graphical interpolation. This procedure is an additional source of error, especially with a semi-logarithmic graph. (ii) Unless the reagents used are specifically low in their fluoride content, and unless the solubility of the lanthanum fluoride membrane of the electrode is low,l4 it is not uncommon to find a curvature in the calibration graph at low concentrations.5~~~-~6 A major portion of this curvature has been ascribed to fluoride contamination in the reagents.13J6 The addition procedure outlined in this work minimises this problem because only perchloric acid containing negligible amounts of fluoride is used.Spiking the low concentration standards, as recommended in the AOAC method,12 is helpful in overcoming this situation but this procedure has an adverse effect on the detection limit, rendering samples with low fluoride content difficult to analyse with acceptable precision. (iii) As the fluoride ion-selective electrode measures the activity of the fluoride ions, the use of an ionic strength - de-complexing buffer (ISDB) is essential in procedures that rely on a calibration graph. The ISDB used must have the property of de-complexing fluoride completely from any compound formed with interfering ions such as magnesium, silicon, aluminium or iron, leaving the total fluoride present as free fluoride ion, otherwise the measurement would result in low values.This type of result has been associated with low recoveries when using direct extraction proced~ires,~J~J~ but in many instances it could be due to the use of an inadequate ISDB system or to abnormally high amounts of interfering ions in the sample.1° (iu) It is common to find 1- or 2-mV drifts in the calibration data after relatively short periods. Hence, frequent re-calibration is necessary to maintain accuracy; the method of standard additions minimises this problem. Standard Additions Procedure Baumannls has already described a standard additions method for the determination of total fluoride in the presence of fluoride cornlplexing cations.This technique has been applied to the determination of fluoride in wine:s,lg sea water17p20 and mineral water.,l Bakers employed a standard additions technique for vegetation after an alkali-fusion procedure and more recently Galloway et aL9 measured fluoride in foliage using a standard additions procedure in acidic solution after dige:;tion with 2 N potassium hydroxide solution. Preliminary work in this laboratory indicates, however, that for certain vegetation samples the procedure proposed by Galloway et al. yields a finely divided precipitate (probably silica) that renders the ion-selective electrode performance unreliable. This fact has been recognised.22 The usual mode of operation15 involves the determination of fluoride at pH 5.5 in order t o avoid association of fluoride with hydrogen ions.However, it has been found in theJune, 1979 I N VEGETATION USING AN ION-SELECTIVE ELECTRODE 547 present work that the determination of fluoride can be conveniently carried out at pH 1, at which only about 1% of the fluoride is present as free fluoride ions because a linear behaviour of the electrode is obtained down to approximately 8 x lobs M and as FH is the predominant complex species under these conditions, interference from other cationic species is negligible (see Results and Discussion). The use of a standard additions procedure at low pH circumvents several limitations associated with calibration techniques as under the conditions used in this work the fraction of free fluoride ions remains constant during the addition procedure, allowing the deter- mination of total fluoride extracted from the sample as discussed in the following section.Theory The initial potential of a test solution is adequately described by the equation El = Eo - s log x, V?) ~ .. . where El observed potential, Eo = potential due to the nature of the electrodes and their contact with the external solution, S = electrode slope or Nernst factor, the theoretical value at 22 "C being 58.5 mV per decade of change in activity, X, = molar fraction of free fluoride ions, 7, = fluoride activity coefficient in the test solution, M , = mass of fluoride in the test solution and V , = initial volume of the test solution. After a known amount of fluoride has been added to the test solution, a new potential E, is obtained: where M , = mass of fluoride added, y = correction factor for the change in the initial volume after the addition and all the other symbols have already been defined.When equation (2) is subtracted from equation (l), the difference between the final and initial potential, AE, is represented by the equation If the ratio X,y2/Xlyly were unity, a simple rearrangement of equation (3) yields an expression for the total fluoride concentration in micrograms per gram : .. .. * * (4) Ma m[logl (AEIS) - 13 ' ' c, = where m = dry mass of sample in grams. Equation (4) was employed throughout this work to obtain the fluoride concentrations reported as Proposed method data in Tables I, 11, IV and V.However, in order to make any valid simplification of equation (3), which is the basic expression to be applied to every known addition measurement, a careful analysis of the actual values of X i , yi and y has to be made. In the proposed method, the ratio y2/yl was assumed to be unity, as little change in the ionic strength would result from the addtion of 0.2 ml of standard solution to 50 ml of test solution. For the same reason, y in equation (3) can be taken as unity within experimental error (its theoretical value would be 1.004). The evaluation of the free fluoride fraction X , is not straightforward. However, as Baumannl8 pointed out, with a high hydrogen to fluoride ratio, as is obtained at pH 1, the fluoride complex species is predominantly HF,23 provided that the concentrations of other interfering cations are not excessive. If this is not the case, a separate determination of these concentrations should be made and the ratio X,/X, evaluated by a method previously 0utlined.~4548 VILLA : RAPID METHOD FOR DETERMINING FLUORIDE AnaZyst, VoZ.104 It was found that the average pH (initial value 1.15,J remained constant after addition of the fluoride standard solution (0.2 ml), This suggests that the hydrogen-ion activity remains constant throughout the measurement procedure, showing that the fraction of free fluoride ions is also constant and that the ratio X2/XI may ‘be taken as unity. Thus, equation (4) is a valid simplified form of equation (3), under the present experimental conditions. Several analyses were made with a pH electrode also in the test solution.Results and Discussion The fluoride contents of eight standard samples of vegetation were determined using the proposed method and the values compared with those found in a previous collaborative study.25 The results for the two sets of data, which were highly correlated (Y = 0.999,), are shown in Table I. An independent check of the proposed method was carried out at the Boyce Thompson Institute for Plant Research,. The results shown in Table I1 correspond to the comparison of data obtained using two widely accepted pro~edures~~9~6 with those from the proposed method. Comparable results are obtained with most of the samples tested. The determination of fluoride in a camellia sample by the proposed method and the AOAC procedure12 yields lower values than the ASITM method.27 We are presently studying whether the problem is due to organic fluorides, or a complexing condition. TABLE I COMPARISON OF ACCEPTED VALUES FOR STANDARD VEGETATION SAMPLES WITH THOSE OBTAINED 13Y THE PROPOSED METHOD Fluoride contentlpg 8 - I Sample Alfalfa and grass Alf a1 f a ..Pine needles . . Orchard grass Pasture grass Alf a1 f a .. Sorghum .. Apple . . .. Reference method* Proposed methodt . . . . 365.4 f 52.6 357 .. . . 73.3 f 12.2 69.7 f 3.2 .. . . 53.1 f 10.9 47.2 f 2.7 .. . . 38.6, f 6.1 28.6 f 1.5 .. . . 32.6 & 6.0 36.7 f 2.3 .. . . 27.7 f 6.4 26.7 f 0.4 . . .. 5.9 f 3.2 5.7 & 0.5 .. .. 3.9 f 2.7 5.6 f 0.3 * The values quoted are the means of the results obtained with several t The standard deviations shown after the means were based on three Only one determination on alfalfa and grass sample was different analytical procedures (see reference 25). observations.made. TABLE I1 COMPARISON OF THE FLUORIDE CONTENTS OF VEGETATION SAMPLES BY THREE DIFFERENT METHODS Fluoride content*/pg g-l 7- ASTM semi-automatell method2s Alfalfa and grass . . .. 334 f 4 Apple . . .. .. .. 77.6 f 2.3 Alfalfa . . .. .. .. 66.3 f 1.2 Pine needles . . .. .. 42.3 f 0.8 Orchard grass . . .. .. 33.9 f 1.3 Pasture grass . . .. .. 29.2 & 1.2 Alfalfa . . .. .. .. 4.4 f 0.2 Sorghum . . .. .. 2.5 & 0.7 Mixedsward . . .. .. 1175 f 8 Camellia .. .. .. 390 f 4 AOAC Official First Action methodlB 374 f 5 71.3 f 1 46.8 f 0.8 35.9 j, 0.5 25.3 & 0.6 20.7 f 1.1 4.0 f 0.1 2.0 f 0.1 1296 f 41 318 f.1 Proposed method 367 f 7 82.3 f 2.5 52.4 f 3.8 33.1 f 1.2 33.8 f 0.6 22.7 f 0.6 4.3 & 0.1 2.3 & 0.2 1293 f 40 225 f 6 * The three sets of data were obtained at the Boyce Thompson Institute for Plant Research.June, 1979 IN VEGETATION USING AN ION-SELECTIVE ELECTRODE 549 TABLE I11 CONCENTRATION OF CERTAIN ELEMENTS IN SAMPLES OF VEGETATION Element, yo dry matter I A \ Sample Si Fe Al Mg Ca Unwashed mixed sward . . . . 2.88 0.1 0.7 0.6 0.7 Washed mixed sward . . . . 1.98 <0.05 <0.1 0.7 1.2 White clover. . .. .. . . 0.8 <0.06 (0.1 0.2 - Pine needles . . .. .. . . 0.5 t0.06 <0.1 0.4 0.6 In order to assess the effect of cation concentrations in vegetation on fluoride measure- ments, the silicon, iron, aluminium, magnesium and calcium contents were determined using an X-ray fluorescence procedure with an internal standard.28 Some results are shown in Table 111.The concentrations found for each element were typical for the species, with the exception of silicon; this was rather high in the mixed sward samples. However, the data in Table Iv and V show that the high silicon content in these samples had little effect on the fluoride concentration, as determined by the proposed method. The values obtained compare favourably with those found using the AOAC method (Table 111) and the reproducibility, as indicated by the standard error, was comparable. The lower values obtained with the alkali-fusion method13,26 may reflect large blank corrections ; greater variability between replicate determinations resulted in higher standard errors for this method.TABLE IV FLUORIDE CONTENTS OF SAMPLES OF VEGETATION OBTAINED BY DIFFERENT METHODS Fluoride contentlpg g-l Alkali fusion13 followed by AOAC Official First potentiometric f A I analysis Proposed method 449 f 75 Sample* Action method12 Pine needles . . .. .. .. 1752 f 47 1417 f 120 1701 f 60 485 & 30 Chuquiraga avellanedaet . . .. 478 & 18 Mixed sward, 1 .. .. .. 1202 f 65 1085 f 128 1231 f 48 L - 1245 & 60 Mixed sward, 3 .. .. .. 1025 f 50 1011 f 72 1150 f 37 Mixed sward, 2$ . . .. .. * With the exception of mixed sward 3, which is a washed sample, all the other samples were unwashed. t This is a xerophitic native shrub that is widely distributed in the area under study.Analysis was carried out after adding 0.2 M citric acid. Because of the existence of polyvalent interfering cations, it is not easy to measure fluoride in unwashed vegetation samples by means of calibration graph procedures as the total amount TABLE V COMPARISON OF THE RECOVERIES OF FLUORIDE OBTAINED WITH ALKALI FUSION AND WITH THE PROPOSED METHOD Fluoride/pg g-1 A r * Alkali fusion methodla Proposed method A * I > r I Recovery, Recovery, Sample Present Added* Foundt % Present Added* Foundt yo Washedmixedsward .. .. .. 1105 f 76 1000 2008 98.4 1160 f 37 1000 2120 98.6 ChuquiragaavellanedaG .. . . 27.3 & 26 40 63.1 93.8 30.8 & 2.0 40 72.6 102.4 Whiteclover . . . . .. . . 119 f 6 60 172 96.1 123 f 3 60 181 98.9 The fluoride was added as sodium fluoride.t The standard deviations shown after the means were based on three observations.550 VILLA: RAPID METHOD FOR DETERMINING FLUORIDE Analyst, VoZ. 104 of these cations is difficult to evaluate. However, from the point of view of efficient pro- tection to livestock it is important to know what the total fluoride is in unwashed forage. The present method seems to fulfil this requirement. In order to test the possibility of cationic interference and also to check the accuracy of the proposed procedure, some vegetation sampl'es were spiked with sodium fluoride prior to extraction. The recoveries were then determinied and compared with corresponding results obtained employing a recently published te~hnique.1~ As shown in Table V the proposed method gave quantitative recoveries of fluoride even in the presence of relatively high amounts of silicon.Data obtained by the alkali-fusion technique13 were slightly lower. This might be due to the small amounts of fluoride that are retained in the filtration step; when the mixture of silicates and amphoteric hydroxides retained on the filter-paper at pH 8-9 is washed with TISAB,13 small amounts of fluorides can be detected in the filtrates. The use of strong phosphoric acid has been suggested in order to eliminate cationic inter- ferences.SJ8 However, it has been found that several different analytical-reagent grade phosphoric acids yield significant fluoride contamination, which lowers the precision, especially with samples of low fluoride content; perchloric acid was found to perform well as the extractant without contributing appreciably to background fluoride.The addition of an equal volume of 0.2 M citric acid to the acid extract before carrying out the fluoride determination by means of FISE does not change significantly the values obtained as shown for sample (2) of unwashed mixed sward in Table IV. This suggests that the influence of cationic interferences may be negligible and that analysis of unwashed forage can be reliably accomplished by means of the proposed method. The possibility that certain fluorosilicates arid organofluoride compounds are not com- pletely extracted into an acidic medium has been ~uggested.~' The results obtained in this work seem to indicate that this problem is not important in the species investigated, with the possible exception of camellia.Nevertheless, before employing the proposed method for routine analyses it would be advisable to carry out a comparison of the results obtained with the proposed method and a reference procedure for each species to be studied. FISE behaviour was checked by constructing calibration graphs using standards prepared in 0.1 N perchloric acid solution. The absence of appreciable curvature when the fluoride concentrations were reduced to about 1 x lo-' M agrees with previous findings.24 In order to ensure that the value for S used in equation (4) was correct and consistent, calibration graphs were run every week. At 23 -J-- 2 "C, the value of S was found to be 57.9 -J-- 0.4. Conclusions The method of standard additions to a perchloric acid extract, as reported in this paper, (i) (ii) It is simple, avoiding distillation, fusion and filtration.(iii) The blank correction for samples containing more than about 10 pg g1 of fluoride (iv) Linear behaviour of the FISE in the sub-micromolar range is achieved at pH 1. (v) Because of (iii) and (iv), samples with a very low fluoride content can be measured (vi) The accuracy of this method is at least a s good as that of other accepted procedures. (vii) It is insensitive to interference by polyvident cations over a wide range of concentra- tions. Consequently, fluoride determinations of unwashed vegetation samples can be accomplished reliably. (viii) The use of ISDB systems is unnecessary. (ix) This procedure can be easily adapted .for the determination of fluoride in several other materials; this will be discussed in a separate paper. has several advantages when compared with other methods in use.It is rapid, the whole procedure taking about 30 min. is negligible. without appreciable decrease in precision. The author is indebted to the Boyce Thompson Institute for supplying vegetation samples and to L. Heller for performing the analyses reported in Table I1 and for his permission to publish these data. Helpful discussions with J. Ares and support and encouragement fromJune, 1979 I N VEGETATION USING AN ION-SELECTIVE ELECTRODE 551 A. Vidoz are acknowledged. Thanks are given to S. Baggio for his help with X-ray measure- ments. Analytical assistance from G. Mondadori, L. Gomez and C. Lena is gratefully appreciated.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. References Ares, J., .I. Air Pollut. Control Ass., 1978, 28, 344. Willard, H. H., and Winter, 0. B., Ind. Engng Chem. Analyt. Edn, 1933, 5, 7. Frant, M. S., and Ross, J. W., Jr., Science, N.Y., 1966, 154, 1553. Jacobson, J. S., and Weinstein, L. H., J . Occup. Med., 1977, 19, 79. Gyoerkoes, T., and Baretincic, J. M., J . Metals, N.Y., 1970, 22, 489. Jacobson, J, S., and Heller, L. I., “Selective Ion Electrode Analysis of Fluoride in Vegetation,” in Englund, H. M., and Berry, W. T., Editors, “Proceedings of the Second International Clean Air Congress,” Academic Press, New York, London, 1971, p. 459. Jones, H. C., and McLaughlin, S. B., “A Simple Method for Measuring Fluorides in Foliage Using the Fluoride Specific Ion Electrode,” Preprint, Meeting of the Working Group on Fume Damage, International Union of Forest Research Organization, Gainesville, Fla., March 14-21, 1971. Baker, R. L., Analyt. Chem., 1972, 44, 1326. Galloway, H. L., Sheaf, R. E., and Skaggs, C. H., Am. Ind. Hyg. Ass. J., 1975, 36, 721. Johnson, M., Fluoride, 1976, 9, 54. Jacobson, J. S., and Heller, L. I., J . Ass. Off. Analyt. Chem., 1975, 58, 1129. J . Ass. Off. Analyt. Chem., 1975, 58, 384. McQuaker, N. R., and Gurney, M., Analyt. Chem., 1977, 49, 53. Midgley, D., Analyt. Chem., 1977, 49, 1211. Orion Fluoride Electrode Instruction Manual, Orion Research Inc., Cambridge, Mass., 1976. Bagg, J., Analyt. Chem., 1976, 48, 1811. Warner, T., Science, N.Y., 1969, 165, 178. Baumann, E. W., Analytica Chim. Acta, 1968, 42, 127. Martin, C., and Brun, S., Trav. SOC. Pharm. Montpellier, 1969, 29, 161. Rix, C. J., Bond, A. M., and Smith, J. D., Analyt. Chem., 1976, 48, 1236. Liberti, A., and Mascini, M., Analyt. Chem., 1969, 41, 676. Jordan, D. E., J . Ass. Off. Analyt. Chem., 1970, 53, 447. Brosset, C., Svensk Kem. Tidskr., 1942, 54, 156. Baumann, E. W., Analytica Chim. Acta, 1971, 54, 189. Jacobson, J. S., and McCune, D., J . Ass. Ofi. Analyt. Chem., 1972, 55, 991. “Annual Book of ASTM Standards,” American Society for Testing and Materials, Philadelphia, 1973, Cooke, J. A,, Johnson, M. S., and Davison, A. W., Envir. Pollut., 1976, 11, 257. Jenkins, R., and de Vries, J. L., “Practical X-Ray Spectrometry,” Second Edition, Springer Verlag, Received July loth, 1978 Accepted November 16th, 1978 Method D 3270-73T. Berlin, 1972, p. 132.

ISSN:0003-2654    

DOI:10.1039/AN9790400545

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

552 Analyst, June, 1979, Vol. 104, pp. 552-559 Automated Catalytic Method for the Routine Determination of Molybdenum in Plant Materials B. F. Quin and P. H. Woods Winchmore Irrigation Research Station, Ministry of Agriculture and Fisheries, Private Bag, Ashburton, New Zealand Molybdenum is determined by its catalytic effect on the liberation of iodine from iodide by hydrogen peroxide. The detection limit (twice the standard deviation of the blank) is 0.01 p.p.m. in plant material, using a 0.25-g sample. Interference from iron is eliminated by preventing any reduction to iron(I1) before complexation with fluoride. High concentrations of salts, other metals and phosphate do not interfere, and agreement with established routine methods is very good. Keywords : Molybdenum determination ; plant material analysis ; automated catalytic analysis Despite the availability of a number of methods for the routine determination of trace amounts of molybdenum in plants, there is still a need for the development of a method that is superior in accuracy, precision and speed at the levels commonly encountered.The dithiol spectrophotometric procedure, originally published by Hamencel and later adapted for plant analysis by Piper and Beckwith,2 still appears to be the most widely used method for routine determinations on plant material. However, despite procedural improvements introduced by Bingley3 and Quin and brook^,^ both the accuracy and precision of this method deteriorate at levels approaching 0.1 p.p.m., which is the critical level for many plants. Further, the sample size necessary to achieve the required sensitivity with reasonable reproducibility (1-2 g) creates difficulties in pot and small plot trials.The application of airect atomic-absorption and spectrographic methods is restricted by poor detection limits caused by the refractory nature of molybdenum oxides. Although the sensitivity can be enhanced by solvent extraction, this is achieved at the expense of the speed of the analytical procedure. In an effort to surmount these problems, a study was made of the automated catalytic method described by Hadjii~annou.~ This method, in which molybdenum is determined by measuring its catalytic effect on the liberation of iodine from potassium iodide by hydro- gen peroxide, is extremely sensitive (less than 0.0002 pg ml-l in solution), but is difficult to perform manually, as the reaction is time dependent and sensitive to minor changes in reagent concentrations. Fuge6 developed Hadjiioannou's method, and described an automated procedure for the determination of molybdenum in geological and biological samples.However, this pro- cedure suffers from interference by many other elements, which severely limits its application. Bradfield and Stickland' studied the effect of elements present in dried plant material on the reaction, and described a rapid method for the determination of molybdenum in the range 0.01-0.1 pg ml-l or 0.1-1.0 p.p.m. in plants. Our studies of this procedure led us to the conclusion that the method was still not sufficiently developed for use in routine analysis, especially where a range of plant species and interferent concentrations were encountered.In particular, the extent of interference from iron could not be predicted even if iron analyses were carried out. Eventually, streamlining of the sample digestion procedure, coupled with a more complete understanding, and therefore control, of the major inter- ferences, enabled a procedure to be developed that permits the rapid, accurate and acceptably reproducible determination of molybdenum in all plants on a routine basis. Experimental Equipment Automatic analysey The manifold used is shown in Fig. 1. A, full Chemlab Automatic Continuous FlowQUIN AND WOODS 553 System was used, incorporating a CS40 sampler, CPP30 proportioning pump, short path length (15 mm) spectrophotometer module and chart recorder.The sampler was converted into a twin-probe concentric ring system, which, in conjunction with a second spectrophoto- meter module and use of a two-pen chart recorder, permitted the simultaneous determina- tion of molybdenum and total nitrogen in their respective digests. Pump 3.4 3.9 Waste 4 fy 5 min Wash Vacuum I De-bubbler, 2.9 Sample 0.8 KI Waste y ' $1 \ Waste f l - 2.9 Flow cell Waste - Spectrophotometer Recorder 420nm 15mm Fig. 1. Flow diagram for automated analysis : SMC, single-mixing coil; and TDC, time-delay coil. Glassware Glassware (borosilicate) was subjected to routine cleaning with detergent and rinsed with distilled water. Jena test-tubes (20 x 150mm) that had previously been used several times for total nitrogen determinations gave lower and more constant blank values than new, acid-cleaned tubes.Reagents Analytical-reagent grade chemicals and distilled water (glass or stainless steel) were used. Hydrochloric acid for digestion, 2 N. Hydrogen peroxide f o r digestion, 100-volume. Hydrogen peroxide solution. Add 12 drops of Brij 35 solution (20% m/V). Hydrochloric acid for sample dissolution and azttomatic analyser wash, 0.125 N. Ammonium fluoride solution, 0.25% m/V. per litre. Dilute 0.65 ml of 100-volume hydrogen peroxide to 1 1. Add 12 drops of Brij 35 solution (20% m/V)554 QUIN AND WOODS: AUTOMATED CATALYTIC METHOD FOR THE ROUTINE Analyst, Vd. 104 Potassium iodide solution, 0.5% m/V. Standard molybdenztm solution, 100 pg ml-l.Dissolve 150 mg of molybdenum(V1) oxide in 10ml of 0.2 N sodium hydroxide solution, dilute to about 200 ml, add 25 ml of 0.1 N hydrochloric acid and make up to 1 1 with water. Prepare a working standard by diluting 1 ml of the concentrated standard to 100 m:l with water. From this, prepare solutions containing 0.01-0.20 pg ml-l of molybdenum in 0.125 N hydrochloric acid. Preparation of Sample For a single determination, a mass of sample of 0.25 g in a final volume of 2.5 ml is sufficient. However, the use of a 1-g sample permits the determination of other metals in the remaining sample solution. Weigh 1 g (0.25 g) of oven-dried plant material (less than 60 mesh) into a silica crucible and ash it overnight in a muffle furnace at 550 "C. Transfer the cooled ash into a 20 x 150 mm borosilicate test-tube and add 4 ml (2 ml) of 2 N hydrochloric acid, 1 ml (0.5 ml) of 100-volume hydrogen peroxide, and one anti-bumping granule.Take to dryness, for example in an aluminium block drilled with 21 mm diameter holes as described by Quin and Woods.* Prepare blank tubes in a similar manner. Add 10 ml (2.5 ml) of 0.125 N hydrochloric acid to the residue, heat at 60 "C for 30 (15) min, mix thoroughly and re-heat at 60 "C for a further 20 (10) min. Allow to stand overnight, then determine the molyb- denum content by the catalytic method on the automatic analyser. Determination of Molybdenum Set the gain on the spectrophotometer module in the range 2-3, and the recorder at 50 mV, to cover a range of 0-2 p.p.m. in plant material. Set the base line at zero on the recorder and set the sampler at a rate of 35 per hour with 35- and 70-s sample and wash phases, respectively. A series of one blank, a full set of standards (0, 0.1, 0.2, 0.5, 1.0 and 2.0 p.p.m.), 30 samples, a blank and two standards (0.1 and 0.5 p.p.m.) was found to be convenient using a 40-tube sampler.Read the peak heights on the recorder, construct a calibration graph and calculate the molybdenum concentrations in the samples. Dilute off-scale samples 5-folld with 0.125 N hydrochloric acid and re-run, rather than using higher standards and lower scale expansion (see Results and Discussion). Switch the system on 15 min before use. Results and Discussion Method of Digestion The hydrochloric acid - hydrogen peroxide digestion of the dry-ashed material seems preferable to the nitric acid - perchloric acid mixture' because it is rapid and amenable to the handling of large sample batches.The ladter method requires, when using the same glassware, longer drying periods at higher temperature in order to avoid the retention of small amounts of perchloric acid in the test-tube, which have a marked effect on the final pH of the sample solution and, as a result, on the reaction rate, The incorporation of hydrogen peroxide with the hydrochloric acid ensures that all of the iron contained in the sample is oxidised to iron(III), the importance of which is discussed later. Automatic Analyser Manifold The manifold (Fig. 1) differs from that used by Bradfield and Stickland' in the following respects: (i) de-bubbling of the sample tube to prevent inter-sample bubble entering the flow cell, (ii) less pull on the flow cell, (iii) a shorter time delay and (iv) interchange of the order of addition of the potassium iodide and hydrogen peroxide solutions.Alterations (i) and (ii) were a consequence of the design of the Chemliab flow cell. A shorter time-delay coil was used because the increased sensitivity obtained with a longer time delay was offset by reduced accuracy. The order of addition of potassium iodide and hydrogen peroxide was changed to control interferences, as described later. Sensitivity, Detection Limits and Reproducibility The sensitivity of the reaction for the determination of molybdenum depends on the difference in the rates of the catalysed and uncatalysed reactions.These rates are functionsJune, 1979 DETERMINATION OF MOLYBDENUM IN PLANT MATERIALS 555 of time and the concentrations of acid, iodide and peroxide. In contrast to the statement that the concentrations used give maximum sensitivity,' we found that the largest increment between the catalysed and uncatalysed reaction occurs at an acid concentration closer to 0.2 N (Fig. 2), but the sudden change in the rate of increase at about this acid concentration, concurrent with much poorer reproducibility, were sufficient cause to retain an acid con- centration of 0.125 N, particularly as the sensitivity is in any event more than adequate. A calibration graph of peak height on the recorder against molybdenum concentration in the range 0-0.2 pg ml-l, obtained using a 15 mm path cell in a Chemlab spectrophotometer set at gain 2.5, with a 420-mm filter and a recorder setting of 50 mV, was a straight line passing through the origin.70 L 60- 7 L 8 5 0 - 40- + r m .- 2 30- Y 2 20- - I I I I 0 0.1 0.2 0.3 0.4 0.5 0.6 Hydrochloric acid concentrat i 0 n/ N Fig. 2. Effect of acid concentration on rate of catalysed and uncatalysed reaction: 0, no molybdenum; and A, 0.05 pg ml-l of molybdenum. Doubling the concentration of the potassium iodide from 0.5 to 1.0% m/V increased the sensitivity, but it was found that the calibration graph became unpredictable (gradient initially increasing, then decreasing) at absorbances above 0.5 (corresponding to 1.5 p.p.m. in plant material). This effect was accompanied by a yellowish discoloration in the solution, possibly caused by the formation of a fluoro complex, as the effect was not noticed in the absence of fluoride.The use of a 0.5% m/V iodide solution gave a suitable working range of 0-2 p.p.m. in plant material, and a &fold dilution if required should enable most samples with higher molybdenum contents to be determined. A value of twice the standard deviation of the blank was used as a measure of the detection limit. Ten blank digests gave a mean of 0.009 p.p.m. with a standard deviation of 0.005 p.p.m., giving a detection limit of 0.01 p.p.m. in plants (Table I). The lower blank values obtained compared with those reported by Bradfield and Stickland' suggest that the test- tubes used for the nitrogen determination contained lower levels of extractable molybdenum as a result of the higher temperature (340 "C) to which they had been subjected in the sulphuric acid digestion.As shown in Table I, the reproducibility was satisfactory. The variability between repli- cates of the same sample solution was minute, showing that it was due entirely to sampling and digestion. Interferences other metals and phosphate. Interferences can be classified into four groups, viz., ionic strength of the solution, iron,$5 6 QUIN AND WOODS: AUTOMATED CATALYTIC METHOD FOR THE ROUTINE Analyst, VOl. 104 TABLE I ANALYTICAL RESULTS FOR REPLICATE MOLYBDENUM DETERMINATIONS Standard Individual results, p.p.m. Mean, deviation, Sample (10 determinations) p.p.m. p.p.m. Blank . . .. . . 0.009, 0.011, 0.010, 0.008, 0.006, 0.009 0.005 0.010, 0.008, 0.009, 0.008, 0.010 Ryegrass - clover .. 0.08, 0.08, 0.08, 0.11, 0.10, 0.090 0.013 0.10, 0.08, 0.08, 0.10, 0.11 Wheat (leaves) . . 0.27, 0.27, 0.27, 0.28, 0.31, 0.290 0.020 0.32, 0.31, 0.30, 0.30, 0.27 Lucerne .. . . 2.96, 2.88, 3.04, 3.04, 2.92, 2.93 0.066 2.92, 2.80, 2.88, 2.92, 2.92 Ionic strength of solutions Hadjiioannod found that large deviations in ionic strength (with zinc chloride) did not affect the rate of reaction. However, both Fuge6 and Bradfield and Stickland7 found that high concentrations of salt depressed the rate of reaction. Fortunately, the effects of individual cations did not appear to be additive, and Bradfield and Stickland' reported no interference from solutions containing 0.4% of potassium, 0.24% of calcium and 0.05% of magnesium, equivalent to a digest of a plant sample containing the high concentrations of 4% of potassium, 2.4% of calcium and 0.5% of magnesium.However, higher concentra- tions of any of these three cations did cause up to 25% depression, calcium and magnesium having the most marked effect. Using the proposed procedure, no interference was found from solutions containing up to 0.5% of potassium, sodium, calcium or magnesium as their chlorides (equivalent to 5% of each in plant material). Higher concentrations of sodium and potassium also had no effect, while calcium and magnesium chlorides slightky enhanced the rate of reaction. We believe that these interferences are actually due to a chloride - iodine interaction, as sulphate salts did not affect the reaction.iron Iron also catalyses the liberation of iodine from iodide by hydrogen peroxide, and Bradfield and Stickland7 reported that 20pgml-l of iron(III), equivalent to 200 p.p.m. in plant material, gave an absorbance equivalent to 0.1 pg ml-l of molybdenum (1 p.p.m. in plant material), but the addition of ammonium fluoride to complex iron(II1) reduced the inter- ference to an equivalent of 0.007 pg ml-l, or 0.07 p.p.m. in plant material. Our investi- gations, however, showed that, when using their procedure, iron interference was not consistent. The interference from a given concentration of iron changed from day to day and, more important, from sample to sample. As a result, the procedure could not be used for the routine determination of molybdenum in plant samples containing more than about 100 p.p.m.of iron. As iron concentrations coinmonly fall in the range 50-300 p.p.m., this interference seriously limits the usefulness of the method. Further investigations led us to the conclusion that the variability in iron interference was a result of partial reduction of iron(II1) to iron(I1) in the reagent stream, before all of the iron(II1) is complexed with fluoride. Although the oxidising digestion results in virtually all of the iron in the sample solution being present as iron(III), the addition of iodide to the sample before peroxide allows some reduction to take place, and Fuge6 had demonstrated that iron(I1) gave a 6-fold increase in absorbance over iron(II1). We found that the extent of the interference changed markedly with minute changes in the relative concentrations of the iodide and peroxide, and the presence of certain ions in the sample solutions, particularly phosphate, affected the extent of the interference. The simple solution to this problem was to change the order of addition of the reagents, so that the peroxide was added prior to the iodide.This had the effect of ensuring that all iron remained as iron(II1) and, as a result, an iron concentration of 100 pg ml-1(1000 p.p.m. in plant material) gave an absorbance equivalent to only 0.001 pg ml-l of molybdenumJune, 1979 DETERMINATION OF MOLYBDENUM I N PLANT MATERIALS 557 (0.01 p.p.m. in plant material). When the order of addition was changed, iron(I1) solutions actually caused a depression in the base line, presumably because of the utilisation of some peroxide in oxidising this iron.For this reason, it was important to use an oxidising digestion. Titanium , vanadium , tungsten and chromium However, several factors combine to eliminate any significant interference from these metals : (i) all four are usually present in plant material at concentrations similar to, or less than, that of molybdenum; (ii) molybdenum was found to have a much greater catalytic effect on the reaction, viz., 10, 5, 5 and 10 times greater than the four metals, respectively; (iii) although changing the order of addition of iodide and peroxide did not have a marked effect on the interference from these metals, the ammonium fluoride , originally added to complex iron(II1) , reduced interference from all four metals, by a factor of about five in each instance; (iv) interference from solutions of titanium and vanadium (the two metals most likely to be present in concentrations sufficient to cause interference) was greatly reduced (50-fold) by allowing the sample solutions to stand overnight, presumably as a result of hydrolysis and pre~ipitation.~ Standing overnight also clarified the solution without the need for centri- fugation and, as it did not interfere with productivity, it was adopted as routine. These four metals, like iron, also catalyse the oxidation of iodide by peroxide. Phosphate Bradfield and Stickland' reported that phosphate reduced the catalytic effect of both molybdenum and iron.We found that adding the peroxide prior to the iodide removed this effect (Fig.3). The explanation of this effect is not readily apparent. Comparison with Other Methods The real test of any routine analytical method is how it performs against established methods. We compared results for pasture analysis with those from Ruakura Agricultural Research Centre, Hamilton, New Zealand, where in excess of loo00 plant samples are analysed for molybdenum annually, by the dithiol spectrophotometric procedure of Bingle~.~ The results (Fig. 4) were satisfactory, particularly in view of the purposely chosen range of low concentrations. Another batch of samples, which included a wide range of plant species, were compared (where sample size permitted) by three methods: (i) the catalytic method described in this E - 0 200 400 600 800 Phosphorus concentration (as NaH,PO,l/pg ml-' Fig.3. Effect of phosphate on the reaction: 0, peroxide added prior to iodide; and A, iodide added prior to peroxide, both in the presence of 0.05 p g ml-l of molybdenum. 0 Molybdenum in plant material, p.p.m. (Ruakura) Fig. 4. Regression of molybdenum in plant material (this method) against the adaptation of the dithiol spectro- photometric procedure of Bingleya used at the Ruakura Agricultural Research Centre, r = 0.998.558 QUIN AND WOODS: AUTOMATED CATALYTIC METHOD FOR THE ROUTINE Analyst, VoZ. 104 paper, (ii) the dithiol procedure used at Invermay Agricultural Research Centre3 and (iii) an adaptation of the method of Quin and Brooks4 used at Winchmore Irrigation Research Station. Firstly, using a similar digestion pro- cedure to that used in the catalytic method, as the height of the muffle furnace prevented efficient ashing in 20 x 150 mm test-tubes, and any trace amounts of organic matter remain- ing interfered in the measurement of low rnolybdenum concentrations.Secondly, the addition of potassium iodide, used to control copper(I1) interference by reducing it to copper(I), was omitted, as copper concentrations of less than 20 p.p.m. in plant material did not interfere. Agreement between all three methods was good (Fig. 5) and the increased productivity of the automated catalytic method described in this paper is considerable (Table 11). Further, the automated method relquires less skill and experience for acceptable results to be obtained. The adaptations involved the following.1 * i * * 0 ' I e m X x * r". 0 2 4 6 8 0 2 4 6 8 10 Molybdenum in plant material, p.p.m. (Invermay) Molybdenum in plant material, p.p.m. (Winchmore) Fig. 5. Regression of molybdenum in plant material (this method) against: (a), the dithiol spectro- photometric method of BingleyS used a t the Invermay Agricultural Research Centre, Y = 0.993; (b), the adaptation of Quin and Brooks4 procedure used a t Winchmore Irrigation Research Station, Y = 0.989. a, Samples common to both regressions; and x , samples used in one regression only. Conclusions Several small but significant changes made to the procedure described by Bradfield and Stickland' have combined to produce a rapid automated method with sufficient sensitivity, TABLE I1 PRODUCTIVITY COMPARISOK FOR VARIOUS METHODS Results are given in man hours per 100 samples. Method Determination of Weighing and molybdenum digestion concentration Total This work . . .. .. . . 2.5 1* 3.5 Dithiol (Ruakura procedure3) . . .. 6 6 11 Dithiol (Quin and Brooks4) . . . . 1.5 (2.5t) 4 5.5 (6.5t) * Includes reagent preparation and chart reading only; automatic analyser running time does t Using digestion procedure described in this paper. not involve technician time.June, 1979 DETERMINATION OF MOLYBDENUM IN PLANT MATERIALS 559 reproducibility and accuracy to meet the demands of both routine and research analysis. A digestion more suitable to the analysis of large batches of samples has been developed and the problems with interferences from high salt concentrations, iron, other metals and phosphate have been overcome. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Hamence, J. H., Analyst, 1940, 65, 152. Piper, C. S., and Beckwith, R. S., J . SOC. Chem. Ind., Lond., 1948, 67, 374. Bingley, J. B., J . Agric. Fd Chem., 1959, 7 , 269. Quin, B. F., and Brooks, R. R., Analytica Chim. Acta, 1975, 74, 75. Hadjiioannou, T. P., Analytica Chim. Acta, 1966, 35, 360. Fuge, R., Analyst, 1970, 95, 171. Bradfield, E. G., and Stickland, J. F., Analyst, 1976, 100, 1. Quin, B. F., and Woods, P. H., Commun. Soil Sci. PI. Analysis, 1976, 7 , 415. Cotton, F. A., and Wilkinson, G., “Advanced Inorganic Chemistry,” Third Edition, Wiley, New Received November 21st, 1978 Accepted December 5th, 1978 York, 1972, pp. 810 and 821-822.

ISSN:0003-2654    

DOI:10.1039/AN9790400552

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

560 Analyst, June, 1979, Vol. 104, pp. 560-565 Determination of the Prostaglandin F,, Content of Pharmaceutical Preparations with Triangle Programmed Bromimetric Titration in Flowing Solutions Zs. FehBr, G. Nagy, K. T6th and E. Pungor Institute for General and Analytical Chemistry, Technictzl University, Budapest, Hungary and A. T6th Chinoin Pharmaceutical and Chemical Works, Budapest, Hungary A survey of the different methods for prostaglandin analysis is given. The use of bromine as a reagent for the accurate determination of prostaglandin F2a is indicated from its chemical structure. Several reasons, however, hinder the use of classical bromimetry. In this paper the application of a new analytical method, the so-called triangle programmed titration technique, is described for prostaglandin analysis.This method permits the simple and effective use of bromine as a reagent by performing the titration in a continuous-flow system. The reagent is generated coulometrically during the titration. Methods are described for the determination of the prostaglandin FZac content of different pharmaceutical preparations. Keywords : Prostaglandin Fza determination ; coulometry ; biamperometry ; triangle +rogrammed bromimetric titration ; flow-through analysis The prostaglandins are a group of biologically active compounds that have evoked con- siderable interest in recent years. Research :has centred on their biosynthesis, synthetic preparation and pharmacological activity. All areas of prostaglandin research require a generally applicable method for the reliable determination of a given member of the prosta- glandin fami1y.l Various chromatographic procedures have been investigated such as gas - liquid chroma- tography of tert-butyldimethylsilyl ethers,2 of tetrakistrimethylsilyl ethers3 and of tri- methylsilyl ethers and trimethylsilyl esters4 olf the prostaglandins, gas - liquid chromato- graphy of the derivatives coupled with mass spectrometric analysis being the technique most generally applied."' Most of the spectrophotometric methods that have been described require some chemical reaction of the prostaglandin molecule.8~Q Other analytical procedures adopted include radioanalytical methods involving isotope dilution,1° radioimmunoassayll and enzymatic assay.12 Bioassayl3 is the preferred method because it provides high sensitivity by observing the effect of prostaglandin on blood pressure arid on the gastrointestinal organs.This paper describes a simple, effective and novel method for the determination of the prostaglandin F,, (PGF,,, 7-{3a,5a-dihyclroxy-2/3- [ (3S)-3-hydroxy-trans-l-octenyl]-la- cyclopentyl)-cis-5-heptenoic acid) content of experimental and commercial pharmaceutical products. The method combines the advantages of the flow-through analysis channel principle with the precision and reliability of titrimetric analytical methods.14J5 Electrically generated bromine is used as titrant.FEHBR, NAGY, T ~ T H , PUNGOR AND T ~ T H 561 In triangle programmed titrations, the sample solution flows in the analysis channel at a constant rate and at a certain point in the channel the sample stream meets the reagent stream.The progress of the chemical reaction is indicated by using an appropriate detector placed downstream in the solution-carrying tube. It is obvious that a titration can be carried out only if addition of the reagent is made according to an appropriate reagent addition rate - time programme. In providing easily evaluated titration graphs the triangle-shaped reagent addition rate - time programme proved to be advantageous. In this instance the flow-rate of the reagent is increased linearly from zero to its maximum value achieved at time instant r , after which it is decreased at the same rate as it was increased. Accordingly, two titration graphs will be obtained, connected to each other, by recording the detector signal versus time.The evaluation of the results is made on the basis of the titration graphs, taking the time period between the two equivalence points as readings: a Q = 2 r - & x X v .. .. .. .. * (1) where Q is the time interval between the appearance of the two equivalence points; 27 is the duration of the current programme; a and b are stoicheiometric constants; n is the slope of the current programme; c is the concentration of the sample solution; and V is the flow- rate in the analysis channel. The method, being based on an actual titration procedure, gives highly reliable results. The theoryl4~l5 and applicationl6-l9 of triangle programmed titrations have been published elsewhere. Experimental Apparatus The triangle programmed titrations using bromine obtained by current-programmed electrolysis were carried out in an apparatus, the schematic design of which is shown in Fig.1. The electrolysis cell (1) in the first part of the flow-through analysis channel has two 9 t 13 Fig. 1. Experimental set-up used for triangle programmed bromimetric titrations.562 FEHkR et al. : DETERMINATION OF THE PROSTAGLANDIN FZcr Analyst, Vd. 104 compartments used for the generation of the bromine. There is a spiral-shaped platinum- wire electrode in each compartment forming tihe working electrode (2) (which is the anode) in one and the auxiliary cathode (3) in the other. A dialysis membrane (4) separates the two half-cells. The reagent generating base solution and the auxiliary solution (0.5 M potassium bromide in 1 M sulphuric acid) are supplied by using a peristaltic pump (5).A current generator (7), Radelkis OH 405, controlled by a function generator (6), Philips PM 5168, operating in the single-shot isosceles triangle mode is used for the reagent genera- tion. The solution stream through the anodic half-cell, which carries the bromine reagent, meets the flowing sample solution, supplied by ix peristaltic pump (9), in the small drip vessel (8). The sample is immediately mixed with the reagent stream in this vessel and the titration reaction starts. Solution segments, the compositions of which depend on the sample flow- rate, the reagent flow-rate and the progress of the titration reaction, enter the detector section of the analysis channel after a certain.time delay. When titrating with bromine a voltammetric (biamperometric) detector cell, (10) containing two similar platinum-wire electrodes is used. The cathode (12) of the detector cell indicates the concentration of the excess of bromine in the streaming solution voltammetrically, while the anode (11) serves as a quasi-reference electrode as a high and almost constant bromide-ion concentration exists in the flowing solution. In order to obtatin a titration graph, a constant potential of 0.5 V is applied between the two platinum electrodes and the current is recorded versus time with a polarograph, Radelkis Type OH-102 (13). The drip vessel performs the important function of electrically isolating the generating and detecting parts of the system. Chemicals All of the chemicals used were of analytical-reagent grade.The prostaglandin standards and the pharmaceutical products containing prostaglandin F,, were generously supplied by the Chinoin Pharmaceutical and Chemical Works. Measuring Procedure The first step in an analysis is the preparation of the calibration graph. For this purpose, standard solutions are prepared, the concentrations of which are in the range of the sample concentration expected. The standard solutions are introduced, using the sample-carrying peristaltic pump, one after the other into the analysis channel while reagent generating base electrolyte and auxiliary solutions are streamed through the electrolysis cell. A potential difference of 0.5 V is adjusted on the polarograph and connected to the two platnium electrodes of the detector cell.The current flowing through the detector cell is then recorded as a function of time. The titration graphs for the prostaglandin standard solutions are recorded using the same values for the parameters 27 and as used for determination on I Fig. 2. Triangle programmed bro- mimetric titration graph of prosta- glandin F2% sample solution. C = 0.86 mA. 4.6 X M; 27 = 200 S; and i,,,. =June, 1979 CONTENT OF PHARMACEUTICAL PREPARATIONS 563 the samples. Similarly, while pumping the different sample solutions through the analysis channel at the same rate, the sample solutions are titrated with the bromine reagent. The values of Q are determined from the titration graphs of the standard solutions (Fig. 2) and a Q v e i m s concentration calibration graph is drawn and is used to obtain the concentration values of the sample solutions from the corresponding Q values determined from the titration graphs.The automatic evaluation of the bromimetric triangle programmed titrations can easily be carried out using a small desk computer. Descriptions of the evaluation method, the apparatus and the results obtained were published e1~ewhere.l~ Results and Discussion Fig. 2 shows a bromimetric triangle programmed titration graph obtained for 4.6 x 10-5 M prostaglandin F,, solution. The parameters of the reagent addition programme are given, together with the method of determining the value of Q. The recorded trace of the detector signal shows a periodic oscillation. It is well known that the voltammetric signal is strongly influenced by the flow-rate of the solution and as, further on, the peristaltic pump and the drip vessel induce a periodical change in the flow- rate the reason for the oscillation is obvious. In order to suppress the oscillation to some extent a buffer vessel (14) is connected in the line.However, as can be seen from Fig. 2, the oscillation does not interfere seriously with the determination of the value of Q. Fig. 3 shows a calibration graph for prostaglandin F,, obtained by plotting the Q values as a function of the concentrations of the different standard solutions. The Q - c dependence is rectilinear, as is expected from theoretical considerations. 100 i 27 = 200 s E CY’ $ 50 t 01 I I I I I - 2 3 4 5 6 7 8 9 Concentration of PGF2,/1 0-5M 27 = 200 s; and imax.= 0.86 mA. Fig. 3. Triangle programmed bromimetric calibration graph of prostaglandin Fag. From equation (1) it can be seen that by changing the experimental parameters the slope of the calibration graph can be influenced in order to adjust it to the requirements of the analytical problem to be solved. The flow-rate of the sample solution and the properties of the reagent addition triangle programme (27, imrtx.) are the most effective parameters for this purpose. It can also be seen from equation (1) that the stoicheiometric constant of the titration reaction can be simply calculated on the basis of the equation or the titration graph. From the slope of the calibration graphs the a/b stoicheiometric ratios obtained were of the order of 0.5 (0.47, 0.55 and 0.53).This means that one prostaglandin molecule reacts with two bromine molecules under these circumstances, suggesting that bromine addition during the titration takes place across the two double bonds of the two side-chains. The bromine - chlorine reagent (BrC1) is frequently used for bromine addition reactions and we carried out triangle programmed titrations employing this reagent. The sulphuric acid in the reagent-generating solution was replaced with hydrochloric acid. No change was observed in either the slope of the calibration graph or in the titration graph, indicating that no change occurs in the stoicheiometry of the titration reaction.564 FEHBR et d. : DETERMINATION 01F THE PROSTAGLANDIN F,, Analyst, Vd. 104 In titrations carried out in flowing solutions a constant time period is allowed for the chemical reaction to take place.This time is determined by the flow conditions. As the reaction time is constant, even relatively slow reactions can be used for analyses carried out with the triangle programmed technique. The classical bromimetric direct titration of prostaglandin does not give reproducible results because the bromination reaction proceeds in several steps and is kinetically compli- cated. A method depending on the back-titreition of the excess of bromine added to the sample is lengthy, requires strictly controlled conditions and suffers from the general errors of a back-titration method. However, triangle programmed titration has a well defined reaction time and the reaction products of the titration corresponding to each titration step leave the system, avoiding the errors caused by further bromine-consuming reactions.The effect on the analysis of other constituents in the pharmaceutical preparations was also studied. Formulations identical with the sample preparations but omitting the prostaglandin were prepared. These preparations were dissolved in the same volume of water as the sample and the triangle programmed titration of this blank solution was carried out. No noticeable change in the measured Q values was observed on alternating the blank solution with distilled water in the sample stream, showing that none of the non-prostaglandin ingredients of the preparation consumed any bromine during the reaction period.In addition, varying amounts of prostaglandin F,, were dissolved in the blank solutions and the resulting solutions compared with the prostaglandin standard solutions. It was con- cluded from these two experiments that the inactive (non-prostaglandin) constituents of the preparations examined had no significant interfering effect on the prostaglandin F,, deter- mination when carried out by the triangle programmed bromimetric titration method using the apparatus described. The prostaglandin content of the experirnent(a1 preparations was determined as follows. The preparation was quantitatively dissolved and diluted to 25 or 50ml with water, depending on the amount of drug (stirring the mixture for 10 min at 25 "C gave complete dissolution), was pumped through the analysis channel of the apparatus at a flow-rate of 2.31 x 10-2 ml s-l and the triangle programmed titration was carried out (27 = 200 s and i,,,.= 0.86 mA). From the value of Q obtained by using the calibration graph prepared previously with standard prostaglandin F,, solution and using the same values of 27, imax. and streaming rate, the prostaglandin concentration and hence the prostaglandin F,, content of the sample preparations were determined. A scatter of the results of &1% was observed, regardless of whether pure solutions of prostaglandin or solutions prepared by dissolving experimental preparations were analysed. It can be concluded, therefore, that the reproducibility of the measured results is basically determined by the uncertainties of the graphical evaluation.When the dissolution of the medicament is effected with solvents other than distilled water, the effect of the solvent (e.g., ethanol) on the prostaglandin determination must be studied. When preparations of new composition are to be analysed, the effect of the other ingredients on the prostaglandin determination must also be studied. It can happen that the auxiliary component or the solvent used for the dissolution has a slight effect on the determination of the prostaglandin content of the medicament. In most instances this does not hinder the applicability of the described method. In order to obtain accurate results, however, the calibration must be carried out with standard prosta- glandin solutions containing the auxiliary components in the same concentration that they have in the sample solution.Similarly, if the solvent of the sample solution shows inter- fering effects then the same solvent must be used to prepare the standard solutions for the calibration. Referelices 1. 2. 3. 4. 5. 6. Oesterling, T. O., Morozowich, W., and Roseman, T. J., J. Pharm. Sci., 1972, 61, 1861. Kelly, R. W., and Taylor, P. L., Analyt. Chem., 1976, 48, 466. Szederkhyi, F., and KovAcs, G., Prostaglandins, 1974, 8, 285. Albro, P. W., and Fishbein, L., J . Chrornat., 1969, 44, 443. Maclouf, J., Rigand, M., Durand, J., and Chebroux, P., Prostaglandins, 1976, 11, 999. Lincoln, F. H., Axen, U., Green, K., Ohlsson, H , and Samuelsson, B., Analyt. Lett., 1976, 9, 187.Jzcne, 1979 CONTENT OF PHARMACEUTICAL PREPARATIONS 565 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Cory, H. T., Lascelles, P. T., Millard, B. J., Snedden, W., and Wilson, B. W., Biomed. Mass Takeguchi, C., Kohno, E., and Sih, C . J., Biochemistry, N . Y . , 1971, 10, 2372. Gantt, C. L., Kizlaitis, L. R., Thomas, D. R., and Greslin, J. G., Analyt. Chem., 1968, 40, 2190. Aizana, Y., and Yamada, K., Prostaglandins, 1976, 11, 43. Jaffe, B. M., Smith, J. W., Newton, W. T., and Parker, C . W., Science, N . Y . , 1971, 171, 494. Anggard, E., Matschinsky, F. M., and Samuelsson, B., Science, N . Y . , 1969, 163, 479. Bergstrom, S., Carlson, L. A., and Weeks, J. R., Pharmac. Rev., 1968, 20, 1. Nagy, G., Tbth, K., and Pungor, E., Analyt. Chem., 1975, 47, 1460. Nagy, G., FehBr, Zs., Tbth, K., and Pungor, E., AnaZytica Chim. Acta, 1977, 91, 87. Nagy, G., FehBr, Zs., Tbth, K., and Pungor, E., Analytica Chim. Acta, 1977, 91, 97. Nagy, G., FehCr, Zs., Tbth, K., and Pungor, E., Andytica Chim. Acta, 1978, 100, 181. Tbth, K., Nagy, G., FehBr, Zs., and Pungor, E., 2. Analyt. Chem., 1976, 222, 379. Nagy, G., Lengyel, Z., FehCr, Zs., T6th, K., and Pungor, E., Analytica Chim. Acta, 1978, 101, 261. Received October 17th, 1978 Accepted December llth, 1978 Spectrom., 1976, 3, 117.

ISSN:0003-2654    

DOI:10.1039/AN9790400560

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

566 SHORT PAPERS Analyst, Jzcute, 1979 Determination of Phenindione Using Organic Brominating Agents A. Abou Ouf, M. 1. Walash, M. Rizk and F. Belal Faculty of Pharmacy, Mansoura University, Mansoum!, Egypt Keywords : Phenindione determination ; brominaiion ; titrimetry Phenindione (2-pheny1indane-lJ3-dione) is an anticoagulant , which can be determined by a bromination procedure according to the British Pharmacopoeia1 and the National Formulary,2 or by spectroph~tometric,~ oxidation J4 polarographic5 and oscillometric6 methods. In dosage forms, phenindione is assayed by am ultraviolet absorption procedure192 in which the medium is either 0.1 M sodium hydroxide solution or toluene - ethanolic potassium hydroxide solution, or by a simple colorimetric procedure' based on the orange colour of phenindione in an alkaline medium.Owing to the importance of the therapeutic effect of phenindione, the National Formulary2 also recommend the ultraviolet absorption method The purposes of the present investigation were to develop a simple assay for phenindione using N-bromosuccinimide or dibromohydantoin as titrant , and to apply the procedure to the analysis of phenindione tablets. for measuring uniformity. :* Experimental Reagents Dibromohydantoin. The compound was synthesised according to Bury et aZ.8 and was recrystallised from hot water. A 0.005 M solution was prepared by dissolving 1.43 g in hot water, cooling, then diluting with water to 11. The solution was freshly prepared before use and standardised iodimetrically. N-Bromosuccinimide.Freshly recrystallised. A 0.01 M solution was prepared by dissolving 1.78 g in the minimum amount of warm water and diluting to 1 1 with water. The solution was freshly prepared before use and standardised iodimetrically. Phenindione. Phenindione B.P. and the tablet form, Dindevan (50 mg) , were obtained from commercial sources. Sample solution. A 100-mg amount of pure drug was dissolved in the minimum amount of 0.1 M sodium hydroxide solution and diluted to 100 ml with the same solution. Stoicheiometric Study The quantitative nature of the reaction between phenindione and the brominating reagents was checked by titrating phenindione solution (0.005 M) with the reagent solutions. The stoicheiometry was calculated to be 1 : 1 and 1: 2 for dibromohydantoin and N-bromo- succinimide , respectively.Procedure A for the Pure Drug An accurately measured volume containing :?-15 mg of phenindione was transferred into a 100-ml beaker and titrated with either dibromohydantoin or N-bromosuccinimide with continuous stirring until the red colour of the solution had completely disappeared. The titration was also followed potentiometrically using a combination of platinum and calomel electrodes. Procedure B for Tablets An accurately weighed amount of powder, containing approximately 100 mg of phenindione, was extracted with three Twenty tablets were weighed and ground to a fine powder.SHORT PAPERS 567 successive 25-ml portions of 0.1 M sodium hydroxide solution and the extracts were filtered into a 100-ml calibrated flask. The residue was washed with small portions of the sodium hydroxide solution and the washings were used to dilute the extracts to the mark.An aliquot of this solution (10 ml) was titrated as described in procedure A. Results and Discussion The dibromohydantoin and N-bromosuccinimide give the dibromo derivative of the drug, and the equivalence point in the titration therefore corresponds to the addition of 1 mol of dibromohydantoin or 2 mol of N-bromosuccinimide per mole of phenindione. The bromination is presumed to involve addition to the enol form, and the change in colour from red to colourless may be due to the breakdown of the conjugated s y ~ t e m . ~ The titrations were carried out in sodium hydroxide solution. During the titra- tions the red colour of the solution decreased in intensity and disappeared entirely at the end-point.It appears, therefore, that a simple procedure could be devised in which this self-indicating property of phenindione might be utilised for its determination. This visual titration method produces results that are in good agreement with those of the potentio- metric procedure, which of course depends upon the establishment of the Br, - Br- system. TABLE I RESULTS FOR DETERMINATION OF PHENINDIONE The recovery by the official method’ is 100.02 f 0.76%. Results given are the means of at least three determinations. Recovery with dibromohydantoin, % Amount taken/mg 5 8 11 12 15 Mean Standard deviation Visual 99.9 98.3 99.9 98.9 99.8 99.36 0.725 1 Potentiometric 97.6 100.3 98.79 98.9 99.1 98.8 0.96 Recovery with N-bromosuccinimide, yo -C 99.9 99.9 98.3 99.9 99.8 98.79 98.9 100.8 99.4 99.1 99.3 99.7 0.725 0.785 The results of procedures A and B are shown in Tables I and 11, respectively. Slightly higher recoveries were obtained from tablets using the proposed methods than with the official meth0d.l The results show that the proposed procedures are quantitative and suitable for the microdetermination of phenindione. TABLE I1 RESULTS FOR DETERMINATION OF PHENINDIONE IN TABLETS Dindevan tablets (Evans Nile Co.).Results given are the means of at least five determinations. Dibromohydantoin N-Bromosuccinimide Official method‘ r ----7 P mg found/mg Recovery, yo found/mg Recovery, yo found/mg Recovery, yo 60 49.05 98.1 f 0.63 49.45 98.9 f 1.0 48.87 97.7 f 0.3 Labelled amount/ Amount Amount Amount References 1. 2. 3. 4. “British Pharmacopoeia 1973,” HM Stationery Office, London, 1973, p. 356. “National Formulary,” Fourteenth Edition, American Pharmaceutical Association, Washington, Mendes, M. M., Neto, H. R., Duorta Silva, N. C., and Goncalves, J. B., Bull. SOC. Chim. Biol., 1964, Sharp, L. K., J . Pharm. Pharmac., 1965, 7 , 177. D.C., 1975, p. 560. 46, 785.568 SHORT PAPERS Analyst, Vol. 104 5. 6. 7. 8. 9. Jacbson, E., and Kelven, H. K., Analytica Chzm. A d a , 1972, 62, 405. Szabo, A. Z., Pungor, E., and Ponzanska, T., -4nalyt. Lett., 1972, 4, 261. Bose, B. C., and Vijagvorgiga, R., J . Pharm. I’harmac., 1962, 14, 58. Bury, S., Kwit, D., and Jawarska, R., Polish .Pat., 51 580 (CI C07d), Aug. 24th, 1966, Appl. June Atherden, L. M., “Bentley and Driver’s Text 1300k of Pharmaceutical Chemistry,” Eighth Edition, 9th. 1964; Chem. Abstr., 1967, 67, 90811a. Oxford University Press, London, 1969, p. 595. Received November 27th, 1978 Accepted January 16th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400566

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

568 SHORT PAPERS Analyst, Vol. 104 Application of Difference Spectrophotometry to the Determination of Dipyrone M. Abdel-Hady Elsayed,* H. Abdine and M. E. Abdel-Hamid Department of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, University of Alexandria, A lexandria, Egypt Keywords : Dipyrone determination ; diflerence slbectrophotomeiry Various colour reactions have been utilised for the determination of dipyrone [sodium N- ( 1,5-dimethyl-3-oxo-2-phenylpyrazolin-4-yl)~-N-me thylaminomethanesulphonate monohy- drate] in pharmaceutical preparations. These reactions were developed by the use of /3-naphthoq~inone-4-sulphonate,~ 4-dimethylaminocinnamaldehyde,2 quinone,3 @diazo- benzenesulphonic acid,4 uranyl nitrate,5 diazotisation with subsequent alkalinisation,e citral in the presence of trichloroacetic acid,7 chrornotropic acid in concentrated sulphuric acid8 and neotetrazolium chl~ride.~ These methods, being based on the functional group reaction, suffer from a lack of specificity.Moreover, such methods are not directly applicable, as the conditions for the colour formation must be worked out. The fluorimetric method developed by treating dipyrone with P-benzoquinonedichlor- amine,1° or spectrophotometric measurement iin the ultravioletll or infrared region12 are all sensitive to any diverse substances that usually exist in pharmaceutical formulations. The other methods of assay by polarography,13 potentiometry,14 complexometry~~ and iodimetry16 either are not specific or are insufficiently accurate. In this work spectral changes of dipyrone induced by pH change or oxidation with bromine have been utilised in the application of the Apj methodl7*l8 and the AA method,lS respec- tively.Experimental Materials and Reagents All reagents were of analytical-reagent grade, unless stated otherwise. Dipyrone powder. The Alexandria Company for Pharmaceutical and Chemical Industries, Alexandria, Egypt. Dipyrone tablets. Hoechst Orient SAA Cairo, 0.5 g. Batch No. 406; average mass = 0.71 g. Dipyrone ampoules. Hoechst Orient SAA Cairo, 1.0 g per 2 ml. Batch No. 400. Methanol. Spectroscopic grade. Bromine - methanol solution, 2% VlV. Sulphuric acid, 0.1 N. Sodium carbonate solution, 0.025 M. Sodium hydrogen carbonate solution, 0.025 M. The instrument used was a Prolabo photoelectric spectrophotometer with 1-cm silica cells.* Present address : Department of Pharmacy, University of Nigeria, Nsukka, Nigeria.June, 1979 SHORT PAPERS 569 Preparation of Standard Calibration Graphs For the Ap2 method Two sets of solutions were prepared so that each set contained 0.8, 1.2, 1.6, . . ,, 3.2% m/m of dipyrone powder. One set was prepared by dilution with 0.1 N sulphuric acid and the other by dilution with a mixture of 0.025 M sodium carbonate solution and 0.025 M sodium hydrogen carbonate solution (pH 10). The absorbances of each solution, in acidic and in alkaline media, were measured at equally spaced wavelengths and the coefficient p 2 was calculated. The coefficient, A$,, for each concentration of dipyrone (9, of an acidic solution -9, of an alkaline solution) was then calculated.For the AA method Serial volumes of 1.0,2.0,3.0, . . ., 8.0 ml of the standard solution (0.1 mg ml-l of dipyrone powder) were transferred into a set of 25-ml calibrated flasks and diluted to volume with methanol (solution S). Similar aliquots were transferred into a second set of 25-ml cali- brated flasks, each containing about 10 ml of methanol. A freshly prepared solution of bromine in methanol (2% V/V) was added dropwise, with swirling, to each flask. A transient blue colour appeared. Addition of the bromine solution continued until the solutions acquired a faint yellow tinge. The flasks were set aside at room temperature (25 "C) until the solutions became colourless. Each flask was diluted to volume with methanol (solution B).The absorbance of dipyrone in methanol (solution S) was measured against a blank of a solution of corresponding concentration of the oxidised derivative (solution B) at 232 nm. For the A method The absorbance of dipyrone in methanol (solution S) was also measured against a blank of methanol at 235 nm. Assay for Pharmaceutical Preparations Dipyrone tablets An accurately weighed amount of powdered and well triturated tablets, equivalent to one tablet, was extracted and suitably diluted as described for the preparation of the calibration graphs. Spectrophotometric measurements were carried out as previously described. Dipyrone ampoules volume of exactly 1 .O ml was suitably diluted. carried out as described above. The contents of five ampoules were mixed in a small, dry conical flask.A measured Spectrophotometric measurements were Results and Discussion Apz Method The quadratic polynomial P, was chosen for the application of the APj method, as the segment XZ in the absorption spectrum of dipyrone in 0.1 N sulphuric acid (Fig. 1) is quadratic. Eight-point orthogonal polynomials were used to give maximum information about the over-all spectrum. The quadratic coefficient 9, is calculated according to the following equation17 : where the figures in parentheses are the constant P, given in standard books.20-22 The absorbances of the solutions of different concentrations prepared in each set were measured over the wavelength range 236-278 nm at 6-nm intervals. This wavelength range was chosen for two reasons: firstly, the corresponding p , value is a maximum [Fig.1 (.-.- *)I, and secondly, Aq, (where Aq2 = Afi22/168) for a solution of 2.0% m/m, exceeds570 SHORT PAPERS Analyst, Vol. 104 0.140.18 In this instance the coefficient of variation of Afi2 can be considered to be less than unity. The relationship between Afi2 and concentration is rectilinear in the range 0.8-3.2% m/m. The corresponding calibration graph can be described by the following regression equation23 : Ap2 x lo3 = -0.3357 - 6.353C .. .. (1) where C is the concentration. 220 240 260 280 300 Wavelengthhn Fig. 1. (a) Absorption spectra of dipyrone (2% m/m). Solvents: 0.1 N €E,SO, (- - -) and 0.025 M Ap2 curve from (a). The segments X Z and XZ extend from 230 to 280 rim. C0,2--HCO,- buffer of pH 10 (- 1. AA Method The spectrum of dipyrone in methanol exhibits a strong absorption at 235nm and a shoulder a t 260-275 nm (Fig.2). On adding bromine the methanesulphonate radical is oxidised to sulphate ion. The oxidised dipyrone in methanol is very stable, as there was no change in absorbance for at least 24h. The bromine-oxidised dipyrone displays no bathochromic shift but induces a hypochromic effect at 230-245 nm. Therefore, the AA graph exhibits strong maximum absorbance at 232 nm. Moreover, AA measuirement at this wavelength fulfils the require- ments for the application of the AA method, because A232 for the methanolic solution plus A232 for the oxidation product does not exceed 1,24 and the AA value is approximately 0.4 for the 2% m/m solution.25 8 0.4 C m e v) 2 0.2 0.0 1 I I I I I I I 220 240 260 280 300 Wave I en gt h /nm Fig. 2.Absorption spectra of dipyrone (- - -) and oxidised dipyrone (-), 2% m/m in methanol, and A A curve therefrom.June, 1979 SHORT PAPERS 571 Within a dipyrone concentration range of o.4-3.2y0 m/m, a graph of AA versus concentra- tion is linear. The calibration graph is described by the following equation: AA232 = -0.0023 + 0.2294C . . .. .. * - (2) Conventional Spectrometric Method Applying the conventional spectrophotometric method, Beer's law is followed within a dipyrone Concentration range of 0.P3.2y0 m/m in methanol. Using the method of least squares the data of absorbance measurements at 235 nm give the following regression equation : A,,, = -0.004 + 0.2841C . . .. .. * - (3) Applications The AA method has found wide application in the assay of pharmaceutical compounds containing pH-sensitive auxochrome.When the pH-induced spectral changes do not fulfil the requirements for the application of the AA method utilisation of the A$, method is the alternative. However, spectral changes, through which the AA method can be safely applied, are induced by oxidation (Fig. 2). The validity of both methods, represented by equations (1) and (2), was tested by analysing laboratory-made tablets (Table I) and comparing the results with results from the conventional spectrophotometric method, equation (3). The applicability of the different methods to the assay of the commercial dosage forms was checked (Table I). TABLE I ASSAY RESULTS FOR DIPYRONE PREPARATIONS Results are percentages of nominal dipyrone content.Mean dipyrone content, +- standard deviation r Preparation AA method A method Ap, method Laboratory-made tablets . . . . . . 100.32 f 0.198 (n = 7) 101.85 &- 0.487 (n = 7) 100.20 f 1.152 (n = 5) I I I J T Y t-value = 7.66* (2.179) t-value = 3.42* (2.228) Commercial tablets . . . . . . . . 101.39 & 0.703 (n = 5) 103.79 f 0.935 (n = 5) 101.20 & 0.963 (n = 5 I I t I V Y t-value = 4.592 (2.306) t-value = 4.32' Ampoules .. .. .. .. .. 98.38 -j= 0.205 (n = 5) 101.17 f 0.562 (n = 5) 97.75 f 1.342 (n = 5) \ I I J T T t-value = 10.38* t-value = 5.24* * Calculated t-values. The figures in parentheses are theoretical t-values a t 0: = 0.05. Applying the t-test in the comparison of the three methods, the calculated t results exceed the theoretical t results.Consequently, the null hypothesis is rejected and the difference between the results of the A method and the proposed methods is significant, with the former being less accurate. The mean percentage of the nominal dipyrone content of the ampoules is either 98.38 (AA method) or 97.75 (A?, method), compared with 101.17 ( A method). (i) The absorbance due to the interference from diverse components that usually exist in pharmaceutical preparations, giving a positive error on the application of the A method, with a consequent decrease in its accuracy (Table I, laboratory-made tablets). (ii) The results from the AA method and the A?, method are statistically comparable, as the calculated t-value (2.01) does not exceed the theoretical t-value (2.306 a t a = 0.05), whereas the results from the A method are not as accurate as those from the proposed method.This conclusion is based on: Conclusions Utilisation of the AA and A$, methods in the determination of dipyrone in tablets or ampoules proved to be useful in correcting for the absorbance contribution from the excipi- ents. This increases the specificity of these methods. Moreover, the proposed methods are672 SHORT PAPERS Analyst, Vol. 104 simple to use for routine and control analysis. The lower standard deviation of the results from the AA method compared with the Ajj2 method gives a consequent increase in the reproducibility. This is because the absorbance is measured at only one wavelength (Amax.) in the AA method, but absorbances are measured at different wavelengths in the A$, method (eight wavelengths for the alkaline solutions and also for the acidic solutions).1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. References Maggiorelli, E., and Conti, L., Farmaco, Ed. I’rat., 1960, 15, 179; Chem. Abstr., 1960, 54, 13549. Kato, K., Umed, M., and Tsubota, S., Yakumigaku, 1964, 24, 116; Chem. Abstr., 1964, 61, 15936. Yavorskii, N. P., Koval, U. S., and Starushohenko, N. M., Farmatsevt. Zh. (Kiev), 1965, 20, 13; Drygieniec, D., and Balkowska, J., Farmacja Pol., 1966, 22, 753; Analyt. Abstr., 1968, 15, 1019. Grabowicz, W., Acta Pol. Pharm., 1967, 24, 315; Analyt. Abstr., 1968, 15, 5589. Adam, W., Acta Pol. Pharm., 1967, 24, 161; Analyt. Abstr., 1968, 15, 7544.Shingbal, D. M., and Shirsat, P. D., Indian J . Pharm., 1975, 37, 94. De Mucciarelii, S. I. L., and Ascheri, S. M., Rezrta Asoc. Bioquim. Argent., 1975, 40, 79; Analyt. Abstr., Vasileva-Aleksandrova, P. S., and Shishmanov, P. S., Analyt. Chem., 1975, 47, 1432. Uno, T., and Yamamoto, M., Japan Analyst, 1973, 22, 1420. Silveira, C., and Barrosa, M. T., Revta Port. Farm., 1964, 14, 96; Chem. Abstr., 1964, 61, 14470. Naobumi, O., Yakugaku Zasshi, 1965, 85, 1001; Chem. Abstr., 1966, 64, 4870. Roushdi, I. M., Soliman, S. A., and Sadek, M. H., U.A.R. J . Pharm. Sci., 1971, 13, 347. Murata, T., and Ochiai, T., Kumamoto Pharm. Bull., 1959, 4, 15. Kalejs, O., and Volkova, M. E., Aptech. Delo, 1960, 9, 45; Chem. Abstr., 1960, 54, 21654. “State Pharmacopoeia of USSR X,” Ministry of Health, Moscow, 1968. Glenn, A. L., J . PharrPz. Pharmac., 1963, 15, Suppl., 123T. Abdine, H., Wahbi, A, M., and Korany, M. A , J . Pharm. Pharmac., 1972, 24, 518. Aulin-Erdtman, G., Chemy Ind., 1955, 74, 581. Davies, 0. L., “Design and Analysis of Industrial Experiments,” Second Edition, Oliver and Boyd, Fisher, R. A., and Yates, F., “Statistical Tables for Biological, Agricultural and Medical Research,” Milne, W. E., “Numerical Calculus,” First Edition, Princeton University Press, Princeton, 1949, Bauer, E. L., “Statistical Manual for Chemists,,” Academic Press, London, 1971, pp. 61 et seq. Juenjo, G. M., and Glenn, A. L., Chemy Ind., 1956, 75, 813. Twyman, F., and Lothian, G. E., Proc. Phys. SOC., 1933, 45, 463. Chem. Abstr., 1966, 64, 9512. 1976, 30, 3E33. London, 1956, pp. 344 et seq. Fourth Edition, Oliver and Boyd, Edinburgh, 1953, pp. 80 et seq. pp. 265 et seq. and 375. Received September 26th, 1977 Amended May 9th, 1978 Accepted November 24th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400568

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

672 SHORT PAPERS Analyst, Vol. 104 Spectrophotometric Determination of Cobalt(l1) with 2,2'-Pyridil Bis( 2-q uinolyl hyd razone) H. Kulshreshtha, R. B. Singh and R. P. Singh Department of Chemistry, University of DeZhi, DehZi-l:10007, India Keywords : 2,2'-Pyridil Bis (2-quinolylhydrazone) reagent ; cobalt determina- tion ; alloy analysis ; spectrophotometry In recent years, many nitrogen-containing heterocyclic hydrazones have been prepared and explored extensively for their potential as analytical reagents1-' As the introduction of a quinoline ring into the hydrazide molecule results in an increase in the sensitivity of the reagent, we have synthesised a new reagent, 2,2'-pyridil Bis( 2-quinolylhydrazone) (PBQH) (I), with this characteristic. Its application to the determination of micro amounts of cobalt is described in this paper.In aqueous solution, PBQH reacts with cobalt(I1) to form an orange - yellow complex (Amsx. 510 nm). This complex, on addition of a strong acid, e.g., perchloric acid, shows a bathochromic shift from 510 to 530 nm. This pink complex is stable even in 5 N perchloricJune, 1979 SHORT PAPERS 573 acid medium. As all other transition metal complexes formed by PBQH are decomposed in strongly acidic media, this property can be utilised to form a selective method for the determination of cobalt (11). PBQH, I Experimental Apparatus spectra. A Unicam SP600 spectrophotometer with 1-cm silica cells was used for recording the A Beckman Expandomatic SS-2 pH meter was used for pH measurements. Reagents PBQH reagent. 2,2’-Pyridil (di-2-pyridylglyoxal) (2.1 g) and 2-hydrazinoquinoline (3.18 g) were heated under reflux with a small amount of toluene-4-sulphonic acid (2-5 drops) in dry benzene under nitrogen in a Dean and Stark apparatus.After 15 h the benzene solution was concentrated, cooled and washed with dry diethyl ether. The product was precipitated as yellow crystals, which were shown by thin-layer chromatography to be homogeneous. Elemental analysis confirmed the synthesis (calculated for C,,H,,N8, C = 72.86, H = 4.48 and N = 22.66%; found, C = 72.6, H = 4.48 and N = 22.6%). An approximately M solution in 95% ethanol was prepared and was stored in an amber-glass bottle. This solution was stable for several weeks. Cobalt(I1) solution, 0.1 M. Prepared by dissolving cobalt metal in perchloric acid and diluting to volume.The resulting solution was standardised titrimetrically.8 Sodium hydroxide solution, 1 N. Perchloric acid solution, 10 and 1 N. All other solutions were prepared with analytical-reagent grade chemicals and doubly distilled water. Procedure Determination of cobalt in highly acidic media To a suitable aliquot containing 2-20 pg of cobalt, add 2 ml of 10-2 M PBQH solution, Add sufficient 10 N perchloric acid to give a solution with a total perchloric acid content equivalent to 3 N. Dilute to 10.0 ml with water and measure the absorbance at 530 nm against a reagent blank. Calculate the cobalt content of the sample. Determination of cobalt in moderately acidic media To a suitable aliquot containing 2-18pg of cobalt, add 2ml of 1 0 - 2 ~ PBQH solution.Adjust the pH to 3.0 by using 1 N perchloric acid or 1 N sodium hydroxide solution. Dilute to 10.0 ml, maintaining a 50% (V/V) ethanol - water medium, and measure the absorbance at 510 nm against a reagent blank. Results and Discussion Spectral Behaviour of the PBQH - Cobalt(I1) Complex A study was carried out on the complexation of PBQH with cobalt(I1) over a wide range of acidity. The orange - yellow complex showed maximum absorbance at 510 nm through- out the pH range 2-12, and a bathochromic shift from 510 to 530nm on the addition of perchloric acid. This pink complex is stable in a perchloric acid medium of concentration Up to 5 N.574 SHORT PAPERS Analyst, VoE. 104 Effect of PBQH Concentration and Stability of the Complex For complete complex formation three times the theoretical amount of ligand is sufficient under either set of conditions.The reagent does not absorb at the wavelength of maximum absorption of either of its cobalt complexes. This is advantageous because the excess of reagent is not critical and a blank is necessary only to check the purity of the reagents. Both the complexes are stable for at least 24 h. Characteristics of the Complex Orange - yellow complex Beer’s law is obeyed up to 1.92 p.p.m. of cobalt. The optimum concentration range, evaluated by Ringbom’s method: is 0.24-1.68 p.p.m. Sandell’s sensitivity10 is 0.001 7 pg cm-2 and the molar absorptivity is 3.3 X: lo4 1 mol-l cm-l at 510 nm. Pink complex evaluated by Ringborn’s method: is 0.24-1.92 p.p.m.and the molar absorptivity is 3.8 x lo4 1 mol-l cm-l at 530 nm. Beer’s law is obeyed up to 2.16 p.p.m. of cobalt. The optimum concentration range, Sandell’s sensitivity is 0.001 5 pg cm-2 Composition of the Complexes variation,ll shows that the metal to ligand ratio is 1 : 2. The composition of both complexes, as determined by Job’s method of continuous Effect of Other Ions Synthetic solutions containing a known a.mount of cobalt and various amounts of other ions were prepared and the recommended general procedure for the determination of cobalt was followed. An error of &2% in absorbance reading was considered tolerable. There was no interference from the following ions under neutral or acidic conditions: bromide, sulphite, sulphate, nitrite, nitrate, iodide, chloride (3 000 p.p.m.each) ; fluoride, thiourea, borate, citrate, tartrate, acetate, phosphate, oxalate (1 000 p.p.m. each) ; thio- cyanate and thiosulphate (500 p.p.m. each). Calcium(I1) , strontium(II), barium(II), magnesium( 11) , lead( 11) (800 p.p.m. each) ; aluminium(III), bismuth(III), molybdenum(V1) (400 p.p.m. each) ; beryUium(II), zinc(II), cadmium(II), mercury(II), platinum(IV), osmium(VIII), iridium(III), rhodium(II1) (200 p.p.m. each) ; uranium(V1) , manganese( 139, titanium( IV) , silver( I), tin(II), gold( 111) , ruthenium(II1) (100 p.p.m. each) ; nickel(II), copper(II), iron(II), vanadium(V) and palladium(I1) (50 p.p.m. each) could be tolerated by increasing the amount of perchloric acid from 3 to 5 N. EDTA, persulphate and cyanide interfered seriously.Practical Applications Cobalt has been determined in alloys by using PBQH in a highly acidic medium. The solutions of the alloys were prepared as described by Singh et aZ.12 The results of the analysis of two alloys are summarised in Table I. TA.BLE I DETERMINATION OF COBALT IN ALLOYS Cobalt content, yo Number of Relative standard Alloy zGz--- Found determinations deviation, yo Monel wire .. .. 0.61 0.60 8 Nilo-K wire . . .. 17.4 17.6 8 4.0 1.1 Conclusion This method using PBQH as a reagent cornpared favourably with other sensitive methods The method is simple, sensitive and selective. known for the determination of cobalt.June, 1979 SHORT PAPERS 575 The spectrophotometric determination can be carried out over a wide pH range (2-12) or in 2-5 N perchloric acid medium even in the presence of large amounts of other ions.There is no need for an extraction step, reaction is instantaneous and the complexes formed are stable. Large excess amounts of ligand do not interfere, which is an advantage over the nitrosonaphthol reagents. The sensitivities of different methods for the determination of cobalt are compared in Table 11. TABLE I1 SENSITIVITIES OF METHODS FOR THE SPECTROPHOTOMETRIC DETERMINATION OF COBALT(II) Sensitivity1 Reagent pg cm-a Benzil mono(2-pyridy1)hydrazone. . .. .. . . 0.0020 Bend bis(2-pyridy1)hydrazone . . .. .. . . 0.0128 o-Hydroxybenzaldehydeisonicotinyl hydrazone . . 0.0029 Nitroso-R-salt . . .. .. .. .. . . 0.0019 0.004 2 o-Nitrosoresorcinol . . .. .. .. .. . . 0.0025 Dithioxamide .... .. .. .. . . 0.0046 2,2’-Dipyridylketoxime . . .. .. ,. . . 0.0029 2,3-Quinoxalinedithiol . . .. .. . . . . 0.0016 0.006 4 Phenanthrenequinone monoxime . . .. .. . . 0.0033 2-Nitroso-5-dimethylaminophenol. . .. .. . . 0.0012 2-Nitroso-5-diethylaminophenol . . .. .. . . 0.0011 PBQH,pH = 3 .. .. .. .. . . 0.0017 PBQH, 3 N HC10, medium’ * . . .. .. . . 0,0015 Anla*./nm 478 53 1 420 420 520 430 400 388 505 598 420 445 442 510 530 Reference 4 6 7 10 10 10 13 14 15 15 16 17 17 Proposed method Proposed method The authors thank the C.S.I.R., New Delhi, India, for the award of a junior research fellowship to one of them (H.K.). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. References Heit, M. L., and Ryan, D. E., Analytica Chim. Acta, 1966, 34, 407. Singhal, S. P., and Ryan, D. E., Analytica Chim. Acta, 1967, 37, 91. Lions, F., and Martin, K., J. Am. Chem. SOC., 1958, 80, 3888. Pflaum, R. T., and Tucker, E. S., Analyt. Chem., 1971, 43, 65. Going, J. E., and Pflaum, R. T., Analyt. Chem., 1970, 42, 1098. Schilt, A. A., Wu, J . F., and Case, F. H., Talanta, 1975, 22, 915. Vasilikiotis, G. S., and Kouimtzis, Th. A., Microchem. J., 1973, 18, 65. Pribil, R., Talanta, 1959, 3, 91. Ringbom, A., 2. Analyt. Chem., 1939, 115, 332. Sandell, E. B., “Colorimetric Determination of Traces of Metals, ’’ Third Edition, Interscience, New York, 1959, pp. 83 and 414. Job, P., Annls Chim., 1928, 9, 113. Singh, R. B., Jain, P., Gag, B. S., and Singh, R. P., Analytica Chim. Acta, 1979, 104, 191. Jacobs, W. D., and Yoe, J . H., Analytica Chim. Acta, 1959, 20, 332. Holland, W. J., and Bozic, J., Talanta, 1968, 15, 843. Ayres, G. H., and Annand, R. R., Analyt. Chem., 1963, 35, 33. Trikha, K. C., Katyal, M., and Singh, R. P., Talanta, 1967, 14, 977. TBei, K., and Motomizu, S., Analyst, 1976, 101, 497. Received November 15th, 1978 Accepted December 12th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400572

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

576 SHORT PAPERS Analyst, "01. 104 Determination of Osmium(Vlll) Alone or in Binary Mixtures with Some Group Vlll Cations by Potentiometric Titration of Iodide H. Khalifa, N. T. Abdel Ghani and M. S. Rizk Faculty of Science, Cairo University, Giza, Cairo, Egypt Keywords : Osmium( V I I I ) determination; potentiometric titration ; silver electrode This paper describes the application of the method of reducing higher valent cations with an excess of iodide and then back-titrating unreacted iodide with mercury(II), using silver amalgam as the indicator electrode, to the determination of osmium. Osmium tetraoxide has been used for the potentiometric determination of chromium(II1) with hexacyano- ferrate(1II)l and hypochlorite, chloride, chlorate and chlorite in succession.2 Osmium(VII1) has been determined kinetically down to the 40 p.p.m.level.3 An oxidimetric method4 for up to 10 mg of osmium(VII1) involved its conversion into osmium(1V) prior to its potentio- metric titration with cerium(1V) or Mn0,-. Attempts to standardise the yellow solution containing 1 g1-1 of osmium tetraoxide in 0.02 M sodium hydroxide by the strychnine sulphate method6 were reported to be unsuccessful. When the solution was treated with ethanol and then acidified with hydrochloric acid in order to obtain the hexachloroosmate(IV), it was possible to standardise it successfullys by Klobbie's iodimetric method.' Osmium has been determined spectrophotometricallyly.8,9 The stoicheiometry of the redox reactions involving osmium tetraoxide as the oxidant and iodide as the reductant may be predicted by considering the standard reduction potentials of the half-reactions involved.Osmium tetraoxide has been shownlO to undergo the reaction OsO, + 6C1- + 4e- + 8Hf --+ OsCg- + 4H20 E" = +l.OOV Bardin and Goncharenkoll reported that in an alkali metal hydroxide solution osmium tetraoxide was reduced to the red osmium(1V) ion, tetrahydroxodioxoosmate [Os02(OH)2,-] via the intermediate dihydroxotetraoxoosmate, which probably forms in accordance with the reaction OsO, + 20H-- -+ OSO,(OH),~- In hydrochloric acid the reduction with iodide may proceed in accordance with the reactions OsO,(OH);- + 2H+ + OsCl, + 2H20 OsO, + 6C1- + 4e- + 8H+ ---+ OsC1;- + 4H20 E" = +l.OOV 21, + 4e- + 41- E" = +0.54V OsO, + 6C1- + 41- + 8H+ ---+ 0sC.l:- + 21, + 4H,O E" = +0.46 V By use of the relationship E" = (O.O591/n)logK, where E" is the standard electrode potential, n is the number of electrons required to reduce 1 mol of osmium(VII1) to osmium(1V) and K is the equilibrium formation constant of the redox reaction, lo@ was calculated to be 31.13.Such a value indicates that the above reaction should proceed quantitatively to completion. The excess of iodide is titrated with a solution containing mercury(I1) ions. However, in the presence of chloride ions the mercury(I1) can form stable tetrachloromercurate(I1) ions, and this removes mercury from the reaction zone and thus delays or distorts the end-point. It was expected that this source of error would be removed by replacing hydrochloric acid with sulphuric acid.June, 1979 SHORT PAPERS 577 The method described in this paper is simple, rapid, sensitive and selective for the potentiometric iodimetric determination of osmium alone or in binary mixtures with elements of its Group such as cobalt(II), nickel(II), ruthenium(III), rhodium(III), palladium(I1) and platinum (IV) .Experimental Reagents The water used was doubly distilled from all-glass equipment. The chemicals were all of the highest purity available. Ethylenediaminetetraacetic acid disodium salt (EDTA) solution, 0.047 1 M. The solution was prepared in the usual way and standardised against a 0.05 M solution of zinc nitrate, which had been prepared by dissolving pure zinc oxide in nitric acid. Mercury(II) nitrate solzction, 0.0474-O.049 8 M. Prepared by dissolving mercury( 11) nitrate in dilute nitric acid (4ml of concentrated nitric acid per litre) and standardised by potentiometric titration, in the presence of a hexamine buffer, against EDTA using a silver amalgam indicator electrode.Prepared by dissolving potassium iodide in water and standardised against the mercury(I1) nitrate solution. This solution was stored in a black glass bottle. Osmium(IV) oxide solution, 0.0157 M. Prepared by dissolving 1 g of osmium tetraoxide in 250 ml of 0.03 M sodium hydroxide solution and standardised against sodium thiosulphate solution using starch as an external indicator. Platinum(I V ) chloride solution, 0.017 9 M. Prepared by dissolving platinum(1V) chloride in water and standardised potentiometrically.12 Palladium(III) chloride solution, 0.01 14 M. Prepared from palladium(II1) chloride and standardised potentiometrically.l3 Rhodium(III) chloride solution, 0.019 8 M; ruthenium(III) chloride solution, 0.022 9 M; nickel(I1) chloride solution, 0.0492 M ; and cobalt(II) chloride solution, 0,0502 M.These solutions were standardised potentiometrically. Potassium iodide solution, 0.095 3-0.099 8 M. Apparatus Silver amalgam electrode. This consisted of a spectroscopically pure silver rod 7 mm in diameter and 30 mm in length fitted in a glass tube by means of neutral polyethylene tubing and paraffin wax. The projecting end was amalgamated by immersion in pure mercury for a few minutes. This consisted of a Pyrex glass vessel fitted with a glass cover with ground-glass joints for the electrode, the salt bridge (agar-agar in saturated potassium nitrate solution) and the tip of a 10-ml burette.The other end of the salt bridge was immersed in a saturated potassium chloride solution in which the tip of a calomel electrode was immersed. The electrodes were connected to a Scalamp galvanometer (W. G. Pye, Cambridge) and a Pye student potentiometer. The solution was stirred by a magnetic stirrer (Ruhromage, No. 9986, type RH 12). Titration system. Effect of Acidity and Concentration of Reactants The effect of the acidity of the solution was studied by determining 1.00 ml of osmium solution containing 2.99mg of osmium; 5ml of iodide solution were added and various amounts of sulphuric acid, making the solution 0.25-1.5 M in respect of sulphuric acid in a total volume of 30 ml, heating to boiling, cooling and titrating unreacted iodide against mercury( 11) nitrate solution using a silver amalgam electrode as the indicator electrode.The results show that the optimum acidity ranges from 0.75 to 1.5 M sulphuric acid. How- ever, the maximum potential change of 311 mV per 0.1 ml of titrant was obtained with l M acid. With regard to the effect of the concentration of osmium, the results show that with 0.5-3.35 ml of solution (equivalent to 1.5-10 mg of osmium) in 1 M acid the maximum error is 0.5y0, indicating the accuracy of the present method. The standard deviation and error when determining 2.99 mg of osmium, in 10 identical experiments, were calculated to be 0.012 mg and 0.4y0, respectively.578 SHORT PAPERS Analyst, Vol.104 Procedure I. Osmium only metal present To a solution containing 0.6-17.92 mg of osmium add 2-8 ml of potassium iodide solution, adjust the solution to be 1 M in respect of sulphuric acid in a total volume of 30 ml and heat to boiling in a fume cupboard, cool and potentiometrically back-titrate the excess of iodide with mercury(I1) nitrate solution. 11. Binary mixtures of osmium with palladium, platinum or rhodium To an aliquot of solution containing 1.51-9.02 mg of osmium and 1 4 ml of palladium, platinum or rhodium solution, add 5-10ml of potassium iodide solution and continue as described in procedure I in order to obtain the total amount of iodide reacted with the two cations. Iodide forms with the other cation the sparingly soluble black compound PdI,, PtI, or RhI,.To another identical aliquot of the solution add 10 ml of 0.1 M sulphite solution [in order to reduce the osmium(VII1) to a state in which it does not react with iodide] add 5-1Oml of potassium iodide sol.ution and continue as described in procedure I in order to obtain the amount of iodide reacted with the Group VIII metal accompanying the osmium. The difference between these titrations gives the amount of iodide reacting with the osmium(VII1). 111. Binary mixtures of osmium with nickel, cobalt or ruthenium To an aliquot of solution containing 1.5-9.02 mg of osmium(VII1) and 1 4 ml of nickel, cobalt or ruthenium solution add 5-10ml. of potassium iodide solution and follow the procedure described in the first part of procedure I1 in order to obtain the amount of iodide reacted with osmium alone.The other metals do not react with iodide. To another identical aliquot of the solution add 5-6 ml of EDTA solution, heat to boiling (in order to prevent the reaction of osmium with EDTA), cool, add 10 ml of hexamine solution and potentiometrically titrate the unreacted EDTA to obtain the amount of EDTA reacted with nickel, cobalt or ruthenium. carried out behind a glass safety screen. Caution-It is essential that an excess of the iodide reagent is added and that the determination is Results arid Discussion Tables I and I1 give representative results obtained with experiments carried out in triplicate, by the procedure described, for osrnium alone or for analysis of its binary mixtures. TABLE I DETERMINATION OF OSMIUM(VIII) IN SOLUTION Volume of 0.0157 M osmium(VII1) solution/ml 0.20 0.50 0.75 1 .oo 1.25 1.50 2.00 2.68 3.00 4.00 5.00 6.00 1.00 1 .oo 1.00 1.00 1 .oo Volume of titrant/& - Expected Found 1.952 1.950 1.753 1.750 1.587 1.590 4.548 4.550 1.255 1.260 3.175 3.170 2.845 2.840 3.430 3.435 6.050 6.050 5.400 5.410 4.720 4.700 6.040 6.060 4.548 4.490 4.548 4.543 4.548 4.550 4.548 4.550 4.548 4.550 Amount of osmium/mg r 'Taken 0.600 1.493 2.240 2.990 3.730 4.480 5.970 8.000 9.020 11.940 14.930 17.920 2.990 2.990 2.990 2.990 2.990 1 Found 0.605 1.500 2.230 2.980 3.720 4.500 5.990 8.020 9.020 11 370 15.030 17.790 3.250 3.010 2.980 2.980 2.980 Error, 0.83 0.47 0.45 0.33 0.27 0.44 0.33 0.25 0.00 0.59 0.67 0.73 0.70 0.69 0.33 0.33 0.33 % Sensitivity at end-point/ mV per 0.1 ml 280 290 295 311 290 280 270 260 265 268 27 1 270 280 29 1 31 1 290 290June, 1979 SHORT PAPERS TABLE I1 ANALYSIS OF BINARY MIXTURES CONTAINING OSMIUM(VIII) Amount of osmium/mg 1 Taken Found 1.493 1.488 2.990 3.005 5.970 5.940 8.960 8.920 1.493 1.485 2.990 3.005 5.970 5.950 8.960 8.970 1.493 1.483 2.990 2.976 5.970 5.990 8.960 8.990 1.493 1.493 2.990 3.000 5.970 5.990 8.960 8.960 1.493 1.480 2.990 2.980 5.970 5.950 8.960 8.975 1.493 1.480 2.990 3.001 6.970 5.980 8.960 8.975 Error, 0.33 0.50 0.50 0.45 0.63 0.50 0.34 0.11 0.67 0.50 0.34 0.33 0.00 0.33 0.34 0.00 0.87 0.33 0.34 0.17 0.87 0.37 0.17 0.17 Yo Amount of other metal/mg - Other metal Taken Found Palladium 1.210 1.200 2.430 2.447 3.640 3.660 4.850 4.830 Platinum 3.490 3.480 6.984 6.980 10.480 10.600 13.970 14.000 Rhodium 2.037 2.020 4.075 4.080 6.110 6.140 8.150 8.180 Nickel 2.890 2.880 5.780 5.800 8.670 8.690 11.560 11.590 Cobalt 2.958 2.956 5.920 5.938 8.880 8.860 11.830 11.875 Ruthenium 2.315 2.320 4.629 4.640 6.944 6.960 9.258 9.280 579 Error, 0.83 0.70 0.55 0.41 0.29 0.06 0.19 0.21 0.83 0.12 0.49 0.37 0.36 0.35 0.23 0.26 0.07 0.30 0.23 0.38 0.22 0.24 0.23 0.24 % Negligible errors are involved in determining amounts of osmium(VII1) in the range 0.60- 8.00 mg.In most instances the changes in potential occurring at the end-point were large, ranging from 200 to 300 mV per 0.1 ml of titrant, thus allowing easy and accurate deter- mination of the end-points. When carrying out the analysis of binary mixtures osmium could be made unreactive with iodide by reduction with a sulphite and unreactive with EDTA by heating the mixture to boiling. These procedures have the advantages of high selectivity, the high accuracy which is associated with potentiometric methods, simplicity and rapidity. References 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Norkus, P., and Jankauskas, J., Zh. Analit. Khim., 1972, 27, 1629. Norkus, P., and Stulgiene, S., Zh. Analit. Khim.!. 1969, 24, 884. Alekseeva, I. I., Smirnova, I. B., and Yatsimirkii, K. B., Zh. Analit. Khim., 1970, 25, 639. Santrucek, J., Nemec, I., and Zyka, J., Colln Czech. Chem. Commun., 1966, 31, 2679. Ogburn, S. C., and Miller, L. F., J . Am. Chem. SOG., 1930, 52, 42. Ayres, G. H., and Wells, W. N., Analyt. Chem., 1950, 22, 317. Klobbie, E. A., Chem. ZentBl., 1898, 11, 65. Bera, B. C., and Chakrabattry, H. H., Analyt. Chem., 1966, 38, 1419. Popper, E. I., Grecu, I., Pitea, I., Chiorean, L., and Gocan, I., Rev. Roum. Chim., 1964, 9, 383. Lurie, Yu., “Handbook of Analytical Chemistry,” English translation, Mir, Moscow, 1975, p. 309. Bardin, M. B., and Goncharenko, V. P., Zh. Neorg. Khim., 1970, 15, 419. Khalifa, H., and Barsoum, B. N., Microchem. J., 1970, 15, 224. Khalifa, H., and Kotry, M. F., Microchem. J., 1968, 13, 705. Khalifa, H., and Khater, M. M., J. Chem. U.A.R., 1967, 10, 123. Received April 17th, 1978 Amended October 20th, 1978 Accepted November 24th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400576

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

580 Analyst, Jane, 1979 Communications Material for publication as a Communication must be on a n urgent matter and be of obvious scientific importance. Rapidity of publication i s enhanced i f diagrams are omitted, but tables and formulae can be included. Communications should not be simple claims for Priovity: this facility f o r rapid publication i s intended f o r brief descriptions of work that has progressed to a stage at which it i s likely to be valuable to workers faced with similar problems. A fuller paper may be offered subsequently, i f justified by later work. Manuscripts are not subjected to the usual examination by referees and inclusion of a Communication i s at the Editor's discretion. Gas Chromatographic - Mass Spectrometric Analysis of Polyethylene Bottle Packed Intravenous Solutions Contaminated with N-Ethylaniline from the Rubber Part of the Two-component Closure Keywords : N-Ethylaniline migration ; rubber disc ; Polyethylene plastics ; intravenous solutions ; gas chromatografihy - mass spectrometry For some years, polyethylene (PE) bottle packed intravenous solutions have been marketed in various countries.In the bottles, the intravenous solutions are contained within a continuous PE wall but, in order to prevent leakage under the conditions of clinical use, a rubber disc is sandwiched on the top of the bottle, forming a. three-layer PE - rubber - PE closure. During processing of PE at high temperatures, additives may be changed into unknown and potentially harmful substances.1 In our studies of possible contamination of PE bottle packed intravenous solutions available in Norway, am unknown peak was observed on the total ion current trace from a gas chromatograph - mass spectrometer arrangement.The electron-impact mass spectra indicated a compound of low relative molecular mass (m/e 121) with the base peak at m/e 106. The fragmentation pattern indicated the compound to be N-ethylaniline,Z and its identity was verified by comparing the retention time and fragmentation with those of an authentic sample. This major contaminant was present in amounts from 0.1 to nearly 2 p.p.m., depending on the pH of the intravenous solutions. Further investigation revealed that the co:ntamination originated from the rubber disc, the substance having migrated across the PE barrier during processing and storage of the solutions.Consequently, Norwegian health authorities have requested that this source of contamination be eliminated and a more suitable rubber disc inserted. Experimental Apparatus The scanning of mass spectra and the single-ion monitoring were carried out by using an LKB 2091 gas chromatograph - mass spectrometer. The electron-impact ion source was operated at 70 and 11 eV when scanning the mass spectra and at 20 eV when recording the single-ion traces. Quantitative determinations were carried out by using a Varian 1800 gas chromatograph equipped with a flame-ionisation detector. PL glass column (2.1 m x 2 mm i.d.) packed with 3% SE-30 on Supelcoport (80-100 mesh) was used. The injector and the detector were both maintained a t 150 "C.The temperature of the column oven was maintained at 100 "C and the nitrogen flow-rate was 30 ml min-l. Authentic N-ethylaniline and the unknown peak in the sample had the same retention times and mass spectra on the following columns: 3% SE-30, 3% Carbowax 20M and 3% QF-1. Procedures Determination of N-ethylaniline in the solutions solution of N-methylaniline in methanol) were added to 500 ml of the intravenous solution. Sodium hydrogen carbonate (0.5 g) and a scdution of internal standard (5.0 ml of a 200 mg 1-1 AfterCOMMUNICATIONS 581 mixing, the sample was extracted with 5 ml of chloroform and an aliquot of the extract (1 p1) was injected into the gas chromatograph. Calibration graphs were constructed for each sample, treating standard preparations as described above. When blanks were analysed for contaminants, no trace of N-ethylaniline was detected. Retention times were 331 s for N-ethylaniline and 228 s for the internal standard.Analysis of the rubber disc After decantation, the solvent was concentrated to about 3 ml before injection into the gas chromato- graph - mass spectrometer. A rubber disc (3 g), cut into small pieces, was refluxed for 2 h in 15 ml of ethanol. Results and Discussion The contaminant was detected in all of the samples investigated and the results of the deter- minations are given in Table I. The contents of N-ethylaniline in acidic solutions of carbo- hydrates are about 0.4-1.8 mg 1-1 (samples e-h), whereas in solutions of electrolytes the contents are only 0.1-0.2 mg 1-1 (samples a-d).When the contents in eight bottles of the same production TABLE I N-ETHYLANILINE CONTENTS OF INTRAVENOUS SOLUTIONS Solution* 0.9% sodium chloride solution 2.9% sodium chloride solution Ringer solution Ringer lactate solution 5 yo glucose solution 12.5% glucose solution 25% glucose solution 10% invertose solution N-Ethylaniline contentlmg 1-' 0.24 0.10 0.10 0.21 0.41 1.45 1.80 1.85 * 500-ml commercial samples. number of 10% invertose solutions (sample h) were analysed, a relative standard deviation of 8.4% was found. When a standard solution of 10% invertose to which had been added 1.85 mg 1-1 of N-ethylaniline was analysed, the relative standard deviation of the method was found to be 1.0% (n = 8). The selectivity of the method was controlled by using gas chromatography - single-ion moni- toring mass spectrometry.The base peak of N-ethylaniline and a major fragment of the internal standard were monitored. By this method, the N-ethylaniline content was found to be 0.25 mg 1-1 in 0.9% sodium chloride solution (sample a) and 1.85 mg 1-1 in 10% invertose solution (sample h) . In a non-sterilised sample of 0.9% sodium chloride solution, kindly provided by the producer, the content of N-ethylaniline was found to be 0.13 mg 1-l. It can be inferred that migration of N-ethylaniline took place during thermal sealing of the two-component closure and even during the subsequent heat-sterilisation process. Storage may also contribute to the contamination found. During vulcanisa- tion or in contact with acids, the accelerator may decompose to the corresponding amine.3 After diazomethane treatment of an ethanolic extract of the rubber, the mass spectrum of a minor peak on the total ion current trace indicated that S-methyl NN-ethylphenyl dithiocarbamate was p r e ~ e n t .~ Its identity was verified by comparing the retention time and fragmentation with those of an authentic sample. Consequently, it may be suggested that N-ethylaniline has been formed from an NN-ethylphenyl dithiocarbamate salt used as accelerator in the vulcanisation process. Dithiocarbamate compounds are frequently used as accelerators in rubber.582 COMMUNICATIONS Analyst, Vol. 104 Refe:rences 1. 2. 3. 4. Lichtenthaler, R. G., and Ranfelt, F., J . Chromat., 1978, 149, 553. Cornu, A., and Massot, R., Editors, “Compilation of Mass Spectral Data,” Second Edition, Volume Franck, R., and Muhlschlegel, H., in “Kunststoffe im Lebensmittelverkehr,” 24 Lieferung, Carl Onuska, F. J., and Boos, W. R., Analyt. Chern., 1973, 45, 967. I, Heyden, London, 1975, p. 45A. Heymann, Koln, 1977, p. B I1 47. Received April 3rd, 1979 National Centre for Medicinal Products Control, Sven Oftedalsvei 8, Oslo 9, Norway G. A. Ulsaker G. Teien

ISSN:0003-2654    

DOI:10.1039/AN9790400580

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

582 COMMUNICATIONS Analyst, Vol. 104 Luminescence Characteristics of Tubocurarine Chloride Keywords : Tubocurarine chloride ; luminescence characteristics (+)-Tubocurarine chloride (I) is probably the most well known alkaloid of the bisbenzyliso- quinoline group. Often present in the arrow poisons of the South American Indians, (+)-tubo- curarine chloride blocks the action of acetylch1o:line at the myoneural junction to skeletal muscle, causing paralysis, which in turn leads to cessation of respiration and death. A single paralysing dose in man (of the order of 100 pg per kilogram body mass) lasts up to 30 min. Although the response is variable, tubocurarine usually cause:; maximum effect within 5 min, and in overdoses death within 10 min of administration. (+)-Tnbocurarine chloride is used in abdominal surgery in very small doses as a complement to anaesthetics, as it causes paralysis of the abdominal muscles without stopping the natural movement of the intestines.Commercial samples of 00-dimethyl- tubocurarine, which have been shown to be OON-trimethyltubocurarine (II),l are also used to relax skeletal muscle. The trimethyl derivative is about three to ten times more potent than the parent alkaloid, is longer acting and does not cause respiratory depression. I R = H II R=CH3 R’ Because of the narrow safety margin with I(+)-tubocurarine chloride, it is essential that it should be used in a substantially pure form, in order to eliminate the danger of an overdose owing to incorrect standardisation of the product. A quantitative analytical method for (+)-tubo- curarine is therefore essential.Detection of ( + )-tubocurarine in body fluids or tissues is difficult as quaternary alkaloids are soluble in water and insoluble in the organic solvents that are generally employed to extract drugs from alkaline de-proteinised aqueous filtrates. Cohen2 has described a fluorimetric method for the analysis of tubocurarine using rose bengal as a complexing agent; however, such a technique is unlikely to be specific for tubocurarine. Previous workers3 have been unable to detect; fluorescence from tubocurarine from solutions containing up to 20pgml-l at room temperakure. We now report the first observation of emission from tubocurarine.Jzlne , I 9 79 COMMUNICATIONS 583 The fluorescence emission of tubocurarine in ethanol excited at 280 nm a t 298 K is a broad, structureless peak between 290 and 400 nm, the fluorescence maximum is 317 nm and the quantum yield is 10-2.At 77 K the fluorescence maximum shifts to 315 nm, the quantum yield increases to 0.23 and the fluorescence lifetime is 0.8 ns. Phosphorescence is also detectable between 370 and 570 nm, the phosphorescence maximum is a t 434 nm, the quantum yield is 0.14 and the phosphorescence lifetime is 0.44 s. The emission parameters of the trimethyl derivative show some significant differences. At 298 K the fluorescence maximum is a t 315 nm (4 = 1.3 x 10-2) ; a t 77 K the fluorescence maximum is blue-shifted to 309 nm (6 = 0.22, rp = 0.8 ns) and the phosphorescence maximum is a t 433 nm (&, = 9.6 x At 298 K a concentration of 0.1 pg ml-l of tubocurarine was detectable, which improved to 5 ng mI-l at 77 K, and the calibration graphs were linear up to a concentration of 0.1 mg ml-I.The marked difference between the phosphorescence lifetimes of tubocurarine and the trimethyl derivative could be utilised as a method for ensuring that samples of tubocurarine were free from the much more potent trimethyl derivative. The spectroscopic equipment used in this study has been described previously.* All emission spectra are corrected for instrumental response using a quantum counter technique.s rp = 0.94 s). References 1. 2. 3. 4. 5. Bick, I. R. C., and McLeod, L. R., J . Pharm. Pharmac., 1974, 26, 985. Cohen, E. N., J . Lab. Clin. Med., 1963, 61, 338. Metha, A. C., and Chalmers, R. A., Chernia Analit., 1972, 17, 565. Bowd, A., Hudson, J. B., and Turnbull, J. H., J . Chem. SOC. Perkin Trans I I , 1973, 1312. Melhuish, W. H., J , Opt. SOC. Am., 1962, 52, 1256. Received April 9th, 1979 Chemistry Branch, Royal Military College of Science, Shrivenham, Swindon, Wiltshire, SN6 8LA Ernest P. Gibson* James H. Turnbull * Present address : Department of Chemistry, University of Southampton, Southampton, SO9 5NH.

ISSN:0003-2654    

DOI:10.1039/AN9790400582

出版商:RSC    

年代:1979

数据来源: RSC