ISSN:0003-2654    

DOI:10.1039/AN98005FX025

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BX027

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005FP077

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BP083

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

July 1980 The Analyst Vol. 105 No. 1252 Some Observations on the Determination of iron by Atomic-absorption Spectrophotometry Using Air = Acetylene Flames K. C. Thompson* and K. Wagstaff Severn-Trent Water A uthority, Malvern Regional Laboratory, 14 1 Church Street, Malverpz, Worcestershire, WR14 2AN Calibration graphs for iron solutions in fuel-rich air - acetylene flames have been shown to exhibit inflections and even maxima; the response for a given standard can also depend on the age of the solution. Under these flame conditions iron(I1) and iron( 111) gave significantly different calibration graphs. These effects disappeared in a fuel-lean air - acetylene flame, but this resulted in a considerable loss in sensitivity. In fuel-rich flames low levels (0.5-5 pg ml-l) of silicon were found to enhance the iron response significantly.For a given concentration of silicon the degree of enhancement increased with increasing iron concentration. The presence of 100 pgml-1 of calcium was found to suppress the response slightly at the 0.5 pg ml-l iron(II1) level whilst giving a very significant enhancement a t the 5 pgml-l iron(II1) level. In the fuel-lean flame up to 10 pg ml-1 of silicon or 100 pg ml-1 of calcium had a negligible effect on iron(II1). The fact that anomalous calibration graphs, age of solution effects and inter-element effects of silicon and calcium were not observed in the presence of 2% m/V ammonium perchlorate was taken to indicate that a condensed phase effect was responsible for the observed phenomena. This study clearly shows that for iron determinations in the absence of ammonium perchlorate, fuel-rich flame conditions that correspond to maximum iron sensitivity should not be used and fuel-lean flame conditions that result in a decreased sensitivity are recommended.Keywords : Iron determination ; atomic-absorption spectvophotovtzetry ; anomalous calibration graph shapes ; air - acetylene flame Comparability studies on the determination of iron using atomic-absorption spectrophoto- metry with an air - acetylene flame have often shown poor agreement between laboratories. It has been reported that inter-element effects for iron are minimised if an “oxidising flame” is used1-6 and that interference effects are far more pronounced in a sulphate matrix than a nitrate or chloride m a t r i ~ .~ ~ ~ ~ * This study was initiated after finding that erroneous results were sometimes observed for iron quality control solutions when fuel-rich flames, with the acetylene flow adjusted for maximum iron sensitivity, were used. However, satisfactory results were obtained in fuel-lean flames. Both the control chart solutions and the routinely used iron standards contained 0.5-5 pg ml-l of iron(II1) in a 1% V/V nitric acid (70% mi?) matrix. Hence, a simple inter-element effect could not easily account for the discrepancies observed in the fuel-rich flames. When large batches of samples are routinely analysed it is often convenient to select similar (i.e., compromise) operating conditions for a number of elements. Alternatively, many operators optimise the flame conditions for maximum sensitivity for each element.However, these procedures can lead to significant errors for certain elements. For example, a previous studyg has shown that calibration graphs for chromium in the luminous air - acetylene flame, optimised for maximum sensitivity, can exhibit both inflections and distinct maxima. These disappeared when a non-luminous flame was used, but a significant loss in sensitivity (typically 50%) was observed. * Present address : Yorkshire Water Authority, Southern Division, Pollution Control Laboratory, Blackburn Meadow Sewage Works, Alsing Road, Tinsley, Sheffield, S9 1HF. 641642 THOMPSON AND WAGSTAFF: SOME OBSERVATIONS ON THE Analyst, vol. 105 This study illustrates the critical effects of flame conditions, burner height and oxidation state on the shape of the iron calibration graphs when the matrix is 1% V/V nitric acid (70% m/m) (this is a commonly used matrix for water analysis). I t also demonstrates that low concentrations of silicon can have a very significant enhancement effect on the iron response in fuel-rich flames.Experimental Apparatus A Varian Techtron AA6 instrument equipped with the standard 100-mm slot length high- solids air - acetylene burner and a Varian iron hollow-cathode lamp were used. Additional confirniatory results were obtained using Varian-Techtron 1200, Instrumentation Labora- tories 257, Pye SPS and Perkin-Elmer 280 instruments, all equipped with a 100-mm slot length air - acetylene burner. It should be noted that all results given in the text refer to the Varian AA6 instrument unless otherwise stated.All solutions were prepared in 100-ml calibrated glass flasks, except for the solutions that were used to obtain the results given in Table 111, for which 50-ml calibrated polypropylene flasks were used. 0.2 a, c rn f 2 Ll 0.1 0 1 2 3 4 5 Iron concentration/pg ml-' Fig. 1. Calibration graphs for iron(I1) (0) and iron(II1) (0). Varian AA6 instrument, burner set 5 mm below grazing incidence. Slightly fuel-lean flame (the reference conditions). Iron(II1) standards 24 h old, iron(I1) standards 1 h old. Reagents Water. Iroa(ll1) stnndard s o l d o n , 1000 pg ml-? LSDH Chemicals. Ascorbic acid solution, 10000 pg ml-l. Iron(l1) standard solution, 1000 pg ml-l (containing 200 pg ml-l of ascorbic acid).Freshly distilled water was used for the preparation of all solutions. Iron(II1) nitrate dissolved in 1 M nitric acid. Dissolve ascorbic acid (1.00 g) in 100 ml of water. Dissolve ammonium iron(I1) sulphate hexahydrate (1.755 g) in 150 ml of water containing 2.5 ml of nitric acid (70y0 m/m) and 5 ml of 10000 pg ml--l ascorbic acid solution. Dilute the solution to 250 ml in a calibrated flask using distilled water. Iron solutions (20 pg ml-l) were prepared from both standard solutions and were found to give similar (*1%) absorbance values in the dinitrogen oxide - acetylene flame. Nitric acid, 70% nz,ht. This solution con- tained less than 2 pg nil-1 of silicon. Acetylene. British Oxygen Company Limited. Cylinder pressure between 50 and 100% of normal capacity.Calcium chloride solution, 1 M. BDH Chemicals. This solution contained less than 5 pg ml-l of silicon. Szlzcon standard solution, 1000 pg ml-l. Fjsons Limited. Sodium silicate dissolved in water. Both of the 1000 pg nil-l iron standards contained less than 2 pg ml-l of silicon. Low-in-chromium grade (BDW Chemicals).Juh, 1980 Operating Conditions The operating conditions used were as follows: lamp current, 5 mA; wavelength, 248.3 nm; spectral band pass, 0.2 nm; air flow, as recommended in manufacturer’s handbook; opera- tion mode, absolute absorbance mode; integration period, 3 s; and calibration standard matrix, 1% V/V nitric acid (70% m/m). The diluted iron(I1) standards also contained 100 pg ml-l of ascorbic acid in order to minimise oxidation to iron(II1).It was realised that this could possibly introduce a matrix effect, but it was felt that it was important to ensure that the iron(I1) standards would be stable over the measurement period. The burner heights given in the text were related to the grazing incidence position. This was defined as the burner height setting that, without a flame, resulted in the burner grid just intercepting the light beam to give an absolute absorbance reading of 0.01. I t is important to appreciate that the burner height and flame conditions are inter-related in that the flame becomes more fuel-lean at increasing distances above the burner slot because of secondary air entrainment. All measurements were the average of two 3-s integrations and a blank measurement was made between every sample measurement.Each reading was calculated by averaging the blank readings immediately prior to and immediately subsequent to each sample measurement and subtracting this value from the mean of the two sample readings. DETERMINATION OF IRON BY AAS USING AIR - ACETYLENE FLAMES 643 0.3 a, 0.2 2 0, 0.1 2 0 1 2 3 4 5 Iron concentration/pg mI-’ Fig. 2. Calibration graphs for iron(I1) (0) and iron(II1) (0). Varian AA6 instrument, burner set 3 mm below grazing incidence. Iron(II1) standards 24 h old, iron(I1) standards 1 h old. Slightly fuel-rich flame. Reswl t s Shape of Calibration Graphs This section illustrates some typical calibration graphs that were observed in the deter- mination of iron under various flame conditions. All solutions were prepared in 100-ml calibrated glass flasks.Figs. 1 4 show the calibration graphs obtained with the Varian AA6 instrument under various flame conditions for iron(I1) and iron(II1). Fig. 1 shows that in a slightly fuel-lean flame with the burner set to 5 mm below grazing incidence (this corresponds to the routine operating conditions used for iron at the Malvern Regional Laboratory) the calibration graphs for iron(I1) and iron( 111) were effectively coincident. The calibration graphs obtained using these operating conditions were reproducible from day to day and these conditions were designated as the “reference conditions.” The coincidence of the two graphs also shows that the presence of 100 pg ml-l of ascorbic acid in the iron(I1) standards did not result in a matrix effect with these operating conditions.Figs. 2-4 were obtained at a burner height setting of 3 mm below grazing incidence, where the maximum obtainable iron sensitivity was observed. This burner height setting was selected for this part of the study because many users of atomic-absorption instruments644 THOMPSON AND WAGSTAFF: SOME OBSERVATIONS ON THE Analyst, VoZ. 105 0.3 a C 2 0.2 si n a 0.1 0 1 2 3 4 5 I ron conceritration/pg m1-l Fig. 3. Calibration graphs for iron(I1) (0) and iron(II1) (0). Conditions as for Fig. 2 except fuel-rich flame (acetylene flow set for maximum transmission of radiation through flame). select operating conditions for maximum sensitivity. Fig. 2 shows that for a slightly fuel- rich flame the iron(I1) and iron(II1) graphs differ significantly.A smaller but still significant difference was observed in the slightly fuel-lean flame at this burner height setting. Fig. 3 shows that for a fuel-rich flame the iron(II1) graph exhibits anomalous behaviour between 3 and 4 pg ml-l, whereas this was not observed with the iron(I1) graph. These flame con- ditions correspond to maximum transmission of radiation through the flame and maximum iron sensitivity over the range 0 4 pg ml-l. ! 0.2 +? s n a 0.1 0 1 2 3 4 5 iron concentration/pg ml-’ Fig. 4. Calibration graphs for iron(I1) (0) and iron(II1) (0). Conditions as for Fig. 2 except luminous flame. The points on this graph ha.ve been joined together by straight lines as i t was not considered possible to portray the graph accurately by means of a curve.Fig. 4 shows the calibration graph for a luminous flame and it can be seen that both the iron(I1) and iron(II1) graphs exhibit inflections and distinct maxima. There are also large differences between the iron(I1) and iron(II1) calibration graphs. These results were similar to those observed in an earlier study on the determination of chromium in the air - acetylene flame.g The effect was reduced if the burner was lowered but could still be observed withJluh, 1980 DETERMINATION OF IRON BY AAS USING AIR - ACETYLENE FLAMES 645 the burner set 7 mm below grazing incidence. If the burner was raised to grazing incidence the effect was more pronounced. The effect was also observed on a Perkin-Elmer 280 instrument; the calibration graphs obtained for iron(II1) using a slightly fuel-lean flame and a flame on the verge of luminosity, at the burner height setting corresponding to maximum obtainable sensitivity for a 3 pg ml-1 iron standard, are shown in Fig.5. It can be seen that similar behaviour to that of the Varian AA6 was observed with this instrument. It should be stressed that the blank solution [1% V/V nitric acid (70% m/m)] was run between each standard and checks for calibration drift were frequently applied during the preparation of the calibration graphs. 0.3 0) C -e 0.2 a a 0.1 0 2 4 6 8 10 ~ r o n concentration/pg mt-’ Fig. 5 . Calibration graphs for iron(II1). Perkin- Elmer 280 instrument, burner set 6 mm below grazing incidence. Standards 1-week old. Slightly fuel-lean flame (0). Fuel-rich flame on the verge of luminosity (0). The points on this latter graph have been joined together by straight lines as it was not considered possible to portray the graph accurately by means of a curve.Points A and B refer to standards that were 2 weeks old. It can be seen from Figs. 1 4 that for a 3.5 pg ml-l iron(II1) solution maximum sensitivity was observed in the fuel-rich and luminous flames (Figs. 3 and 4) and that the corresponding absorbance reading observed using the “reference conditions” (Fig. 1) was significantly lower. The responses observed in the fuel-rich and especially in the luminous flames were depen- dent on the age of the standards, and this effect is illustrated in Tables I and I1 and also in Fig. 5. The results obtained under the “reference conditions” were always reproducible from day to day and did not appear to depend on the age or the oxidation state of the standards.Table I shows that the response of 5 pg ml-l iron(I1) and iron(II1) standards, on that occasion, did markedly depend on the age of the solutions when measured in fuel-rich flames. Table I1 shows that on another occasion the response of two different 5 pg ml-l iron(II1) solutions prepared 4 days apart varied much less in fuel-rich flames, but that the corresponding 0.5 pg ml-l iron(II1) solutions did show a significant variation in response. The variations between the corresponding standards appeared to be less significant at the lower burner height setting. Table I1 also demonstrates that the optimum response for a 0.5 pg ml-1 iron solution tends to occur in a more fuel-rich flame than that for a 5 pg ml-1 iron solution.This difference indicates that the degree of atomisation is concentration dependent in fuel- rich flames. On yet another occasion two 5 pug ml-l iron(II1) standards were prepared within a period of 15 min and when measured 2 h later in a flame 011 the verge of luminosity gave absorbances of 0.220 and 0.345. In the slightly fuel-lean flame the corresponding absorbances were 0.247 and 0.244, respectively. It should be re-emphasised that all solutions contained 1% The behaviour of the standards in fuel-rich flames was unpredictable.646 THOMPSON AND WAGSTAFF: SOME OBSERVATIONS ON THE Analyst, T/ol. 105 V/V.nitric acid (70% m/m), so that the results cannot be caused by hydrolysis of iron(II1) species.The anomalously shaped calibration graphs have been observed by the authors on many occasions using the laboratory’s Varian Techtron AA6 and 1200 instruments and were also observed on Perkin-Elmer 280, Pye SP9 and Instrumentation Laboratory 257 instruments. Since observing these effects other workers have been approached (see Acknowledgements) and they have confirmed these anomalous-shaped calibration graphs in fuel-rich flames using other lamps and spectrophotometers, vix., Perkin-Elmer 290B, Perkin-Elmer 305B, Perkin-Elmer 560 and Instrumentation Laboratory 351. Effect of Silicon on Iron It was thought that some of the observed effects could be due to leaching variable amounts of silicon species from the walls of the glass flasks. It has been reported that the presence of silicon significantly depresses the iron signa1.4~6~7~10~11 In fact, Platte and Marcy4 observed a 25% depression of the signal from a 10 pg ml-l iron(II1) solution in the presence of 5 pg ml-1 of silicon and an 18% depression in the presence of 100 pg ml-l of silicon.Unfortunately, the air - acetylene flame conditions used were not specified. A simple suppression of the signal by silicon was not thought to account for the effects observed, because some of the older iron solutions (see Tables I and 11) gave significantly higher absorbance values in fuel- rich flames than the equivalent freshly prepared solutions, whereas others gave lower readings. TABLE 1 EFFECT OF THE AGE OF 5 pg m1-l IRON(I1) AND IRON(II1) SOLUTIONS ON ABSORBANCE SIGNALS OBSERVED UNDER VARIOUS FLAME CONDITIONS Notes for Tables I-V.All solutions contained 1% V/V nitric acid (70% yn/yn). The iron(I1) solutions contained 100 pg ml-l of ascorbic acid. All measurements were taken with the Varian AA6 instrument. In Tables I, 111, IV and V the burner was set 3 mm below grazing incidence, except for the reference conditions, where it was set 5 mm below grazing incidcmcc. Except for Table I11 all solutions were prepared in calibrated glass flasks from 50 pg ml-l iron standards prepared from 1000 pg ml-1 iron standards. The 50 pg ml-l solutions were always freshly prepared for each new set of calibration standards. Calcium was added as calcium chloride and silicon was added as sodium silicate. Tables I and I1 wcre obtained using a different nebuliser from Tables 111, 1V and V.Oxidation Age of (-------L > state solution/h 3.0 (a) 3.5 (b) 3.75 (c) 4.0 (d) 4.25 (e) I11 1 0.236 0.300 0.301 0.223 0.159 I11 48 0.233 0.319 0.354 0.328 0.248 I1 .1 0.236 0.328 0.361 0.328 0.176 I1 24 0.237 0.323 0.355 0.273 0.135 Acetylene flow-meter reading, arbitrary units* * Flame conditions : (a) slightly fuel-lean (reference conditions) ; (b) slightly fuel- rich; (c) fuel-rich; (d) verge of luminosity; (e) luminous. However, the addition of small amounts of silicon (added as sodium silicate) was found to have a significant enhancement effect on the response of iron(II1) solutions in fuel-rich flames. Table I11 demonstrates this effect and shows that under the “reference conditions” the addition of up to 10 pgml-l of silicon has a negligible effect, but that as little as 0.5 pg ml-l of silicon can have a significant enhancement effect in fuel-rich flames. The effect was much more pronounced at the 5 pgml-l than at the 0.5pgml-l of iron level.The addition of up to 20 pg ml-l of sodium (as sodium chloride) exerted a negligible effect, which indicated that the observed effect. was due to silicon rather than sodium. In order to minimise silicon contamination, the solutions used in Table I11 were stored in 50-ml cali- brated polypropylene flasks. These were prepared using adjustable micropipettes, fitted with polypropylene tips, directly from a standa,rd solution of 1000 pg ml-l of iron(II1) as iron(II1) chloride in 1 M hydrochloric acid contained in a polyethylene bottle (Fisons).Hence, it would appear that the variation in response of iron standards with respect to age observed in fuel-rich flames may be a t least partly attributable to leaching of siliconJuly, 1980 DETERMINATION OF IRON BY AAS USING AIR - ACETYLENE FLAMES 647 species from glassware. The amount of silicon leached will depend, to some extent, on the previous history of the flask. Twenty iron(II1) standards that had been prepared a week or more previously in glass flasks were analysed to determine reactive silicon using the standard ammonium molybdate method.12 All of the molybdate-reactive silicon levels found were below 0.2 pg ml-l, which indicates that other unknown factors (e.g., polymeric non- molybdate-reactive silicon species) may also be responsible for the anomalous calibration graphs observed in fuel-rich flames.This is further substantiated by the fact that the response of iron( 111) standards prepared in polypropylene flasks still exhibited some variation with respect to time of standing when measured in fuel-rich flames. A set of six 5 pg ml-1 iron standards prepared over a 48-h period exhibited differences in response of up to about 15% in a flame on the verge of luminosity. The responses were noticeably less reproducible for solutions that were nebulised within 30 s of preparation. TABLE I1 EFFECT OF AGE AND BURNER HEIGHT ON THE ABSORBANCE SIGNALS OBSERVED FOR 0.5 AND 5 pg m1-l IRON(II1) SOLUTIONS Acetylene flow-meter reading, arbitrary units 2.5 3.0 3.5 3.75 4 4.25 2.5 3.0 3.5 3.75 4 4.25 2.5 3.0 3.5 3.75 4 4.25 Flame conditions Fuel-lean Slightly fuel-lean Slightly fuel-rich Fuel-rich Verge of luminosity Luminous Fuel-lean Slightly fuel-lean Slightly fuel-rich Fuel-rich Verge of luminosity Luminous Fuel-lean Slightly fuel-lean Slightly fuel-rich Fuel-rich Verge of luminosity Luminous Notes as for Table I.Absorbance signal* A - Burner height below grazing (--*-------, incidence/mm 1 h t 96 ht 1 0.277 0.276 0.301 0.300 0.307 0.304 0.271 0.247 0.159 0.152 0.106 0.102 3 0.204 0.206 0.276 0.276 0.306 0.305 0.292 0.270 0.194 0.189 0.140 0.139 5 0.155 0.158 0.238 0.239 0.284 0.287 0.294 0.290 0.234 0.229 0.153 0.150 5 pg ml-1 iron(II1) 0.5 pg ml-1 iron(II1) ' -- 96 ht N.R. N.R. 0.038 0.039 0.036 0.030 0.021 0.030 0.035 0.037 5 0.037 5 0.037 0.017 0.025 5 0.029 5 0.033 0.034 5 0.034 1 h t 0.029 5 0.034 0.036 5 0.036 5 0.023 0.015 5 0.021 0.030 0.034 5 0.035 0.026 5 0.023 0.016 0.025 0.028 5 0.031 0.029 0 0205 * N.R.= not recorded. t Age of solution. Effect of Calcium on Iron The authors' laboratory is concerned primarily with the analysis of non-saline natural water samples. Calcium is the main cationic matrix species of most of these samples, so a brief study was made of the effect of 100 pg ml-l of calcium (added as calcium chloride) on the response of the iron(II1) standards contained in calibrated glass flasks. This level of calcium corresponds to a moderately hard water. The presence of a matrix of high boiling-point could be expected to have a significant effect on the shape of the calibration graph in fuel- rich flames.The results shown in Table IV show that 100pgml-l of calcium has a negligible effect under the reference conditions, but has a pronounced effect on the shape of the calibration graph in fuel-rich flames. At an iron concentration of 0.5 pg ml-l a slight suppression was observed, whereas at the higher iron concentrations of 4 and 6.5 pgml-l considerable enhancements were found in the fuel-rich flames. The interference effect of 100 pg ml-l of calcium is markedly dependent on the iron(II1) concentration. It should be noted that if the acetylene flow was increased by 10% from the reference flow-rate setting, the presence of 100pgml-l of calcium then exerted a significant enhancement at iron con- centrations above about 3 pg ml-l.648 THOMPSON AND WAGSTAFF: SOME OBSERVATIONS ON THE AnaZyst, voz.105 TABLE I11 EFFECT OF SILICON ON THE IRON(III) ABSORBANCE SIGNALS Notes as in Table I. All solutions were prepared in polypropylene flasks using adjustable micro- pipettes with polypropylene disposable tips directly from the 1000 pg ml-l iron(II1) standard (see text). Solutions were approximately 20 min old. Flame conditions* Silicon concentration/ pg ml-1 0 0.5 1 2 5 10 Reference conditions (a) ------ 7 0.5 pg ml-I 5 pg ml-l iron( 111) iron( 111) 0.028 0.277 0.027 0.276 0.028 0.276 0.028 0.280 0.028 0.283 - 0.285 Verge of luminosity (4 7 - h - 0.5 pg ml-1 5 pg ml-1 iron( 111) iron( 111) 0.046 0.285 0.044 0.320 0.045 0.362 0.049 0.447 0.048 0.461 __ 0.456 Luminous (e) r-- - 0.5 pg ml-I iron( 111) iron (111) 0.038 0.152 0.040 0.170 0.046 0.249 0.048 0.392 0.047 0.466 - 0.458 5 pg ml-1 * See footnote to Table I.Effect of 2% m/V Ammonium Perchlorate on Iron In order to attempt to explain some of the results, the effect of ammonium perchlorate was tested. A previous study13 had already shown that the addition of 2% m/V ammonium perchlorate to all solutions and standards overcame both anomalous calibration graph shapes and numerous inter-element effects observed with chromium determinations in fuel-rich air - acetylene flames. Table V demonstrates the effect of adding 2:& m/V ammonium perchlorate to a range of iron(II1) standards. It can be seen that in the presence of ammonium perchlorate the iron(II1) absorbance signals are almost directly proportional to concentration, up to 6.5 pg ml-l of iron(III), even in the luminous flame.The response in fuel-rich flames did not appear to depend on the age of the solutions. The effect of the addition of 100 pug ml-1 of calcium and 10 pg ml-l of silicon to 4 and 6.5 pg ml-l iron(II1) solutions in fuel-rich flames was virtually eliminated in the presence of ammonium perchlorate. In the authors' opinion, the routine use of 2% m/V ammonium perchlorate for the determination of iron in fuel-rich flames, optimised for maximum iron sensitivity, is not as attractive as the simple expedient of using a fuel-lean flame. TABLE IV EFFECT OF 100 pg ml-l OF CALCIUM ON THE IRON(III) ABSORBANCE SIGNALS Notes as in Table I. Solutions were 20 h old. Flame conditions* Reference conditions (a) Iron(II1) (-A=-------- -7 concentration/ 0 pg ml-l 100 pg ml-1 pg ml-1 Ca Ca 0.5 0.028 0.028 1.5 0.084 0.083 4 0.217 0.216 6.5 0.341 0.341 Verge of luminosity Luminous (d) (el A 7 - h v r -3 0 pg m-1 100 pg ml-1 0 pg ml-l 100 pg ml-1 Ca Ca Ca Ca 0.042 0.039 0.042 0.030 0.111 0.126 0.070 0.101 0.241 0.341 0.153 0.299 0.352 0.533 0.213 0.406 * See footnote t o Table I.Discussion The marked decrease in the degree of atomisation observed with increasing iron concentra- tion in fuel-rich flames (see Tables I1 and V) was thought to be caused by a condensed phase effect ,'v8 the degree of atomisation decreasing with increasing clotlet size. This was attributed to the relative involatility of iron (boiling-point 2 800 "C1*) and iron(I1) oxide (melting-pointJzdy, 2980 DETERMINATION OF IRON BY AAS USING AIR - ACETYLENE FLAMES 649 1370 ‘C1*). It was not thought to be due to the stability of iron(I1) oxide (dissociation energy 4.0 eV8) because the effects were not observed with the more volatile element lead, which has a similar oxide dissociation energy of 4.1 eV.8 The lead calibration graph did not exhibit any inflections or maxima in the luminous flame even with the burner height set for grazing incidence.The response of the lead standards did not appear to depend on the age of the solutions even in the luminous flame. Batch TABLE V EFFECT OF 2% m/V AMMONIUM PERCHLORATE ON THE IRON(III) ABSORBANCE SIGNALS Notes as in Tab12 I. Flame conditions* r A -7 Reference conditions Verge of luminosity Luminous (a) (4 (el Iron(II1) Calcium or silicon ,--.A--, ,--A-, r-~-----, concentration/ concentration1 0% mlV 2.0% mlV 0% m/V 2.0% m/V 00,; m/V 2.0 mlV yg ml-l yg ml-l NH4C10, NH,CIO, NH,CIO, NH,CIO, NH,C10, NH,C104 Batch 1 (solutions 22 h old) .. 0.5 1.5 4.0 4.0 6.5 6.5 (solutions 0.5 h old) . . . . 4.0 4.0 4.0 4.0 6.5 6.5 6.5 6.5 Batch 2, tun 1 week after batch 1 * See footnote to Table I. 0 0 0 0 100 (Ca) 100 (Ca) 0 100 (Ca) 5 (Si) 100 (Ca) + 10 (Si) n 1ii0 (Ca) 5 (Si) 100 (Ca) + 10 (Si) 0.027 0.081 0.208 0.209 0.330 0.330 0.207 0.208 0.319 0.325 - - - - 0.026 0.045 0.044 0.044 0.054 0.079 0.102 0.130 0.070 0.132 0.207 0.237 0.343 0.156 0.354 0.206 0.335 0.346 0.298 0.356 0.330 0.316 0.540 0.213 0.537 0.328 0.531 0.544 0.494 0.557 0.200 0.196 0.364 0.160 0.370 0.198 - 0.358 - 0.372 0.203 0.330 0.365 0.350 0.382 0.200 - 0.369 - 0.374 0.318 0.306 0.571 0.220 0.565 0.317 - 0.570 - 0.574 0.322 0.507 0.572 0 537 0.583 0.319 - 0.57‘3 - 0.585 The marked increase in signal observed for iron(II1) in fuel-rich flames in the presence of 0.5-5 pg ml-l of silicon is difficult to explain.Most significant inter-element effects in flame atomic-absorption spectrophotometry are normally observed at much higher interferent to analyte It is thought that in the presence of small amounts of silicon the clotlets in the fuel-rich flame undergo more rapid atomisation reactions with the flame gases than in the absence of silicon. At the 5 pg ml-l of iron(II1) level the degree of enhancement for a given silicon concentration (see Table 111) is significantly greater than at the 0.5 pg ml-l of iron level.At the 5 pg ml-l level the clotlets formed will be larger and consequently will take longer to reach equilibrium with the flame gases. Thus, the enhancement observed with silicon could be expected to increase with increasing iron concentration. In the presence of ammonium perchlorate explosive decomposition is thought to occur during clotlet formation in the lower regions of the flame. This will result in the formation of smaller clotlets, which then reach much more rapid equilibrium with the flame gases. This latter equilibrium is effectively unaffected by silicon and calcium in fuel-rich flames. Conclusions All of the anomalous responses for iron reported above were observed in a 1% V/V nitric acid (70% m/m) matrix, and when this matrix is used for the determination of iron it is strongly recommended that due regard should be paid to the following points.1. For the routine determination of iron a slightly fuel-lean flame should be used. How- ever, this will result in a significant decrease in sensitivity (up to 40%) when compared with the maximum sensitivity that can be observed in fuel-rich flames (this loss of sensitivity does not, however, impair the detection limit by a similar amount, as the “noise” is reduced on changing to a more oxidising flame). The burner height should be set lower (2-3 mm) than that typically used for elements such as cadmium and lead. 2. Although the addition of 2y0 m/V of ammonium perchlorate to all samples and standards effectively overcomes all of the anomalous effects observed in the fuel-rich flames, it is felt that the use of the slightly fuel-lean flame is a simpler procedure for routine iron determinations.650 THOMPSON AND WAGSTAFF 3.The determination of iron in the fuel-rich air - acetylene flame optimised for maximum sensitivity without the addition of ammonium perchlorate is not to be recommended. The calibration graph under such conditions can exhibit inflections and even maxima, which can easily be missed unless a large number of standards are employed, and the response for corresponding iron(I1) and iron(II1) standards differed significantly in many instances. It should be noted that some microprocessor-controlled atomic- absorption instruments will accept only two or three standards for setting up the calibration graph and under these operating conditions inflections or even maxima in the calibration graphs can easily be overlooked.4. The response of iron(I1) and iron(II1) standards can also be markedly dependent on the age of the solutions when measured in fuel-rich flames without the addition of ammonium perchlorate. This was thought to be due partly to trace amounts of silicon species leaching from glassware. The relative response of standards of different ages cannot be predicted. Hence, any testing of inter-element effects on a single concentration of iron in fuel-rich flames is pointless as the effects can be dependent on the age of the solution, the previous history of the glass storage vessel and the iron concentration. 5. The calibration graph should occasionally be checked using a t least ten standards. If the samples are likely to contain iron, iron(I1) and iron(II1) standards should be prepared in both oxidation states in order to check that coincident calibration graphs for the two oxidation states are observed. Some of the quality control solutions should be prepared by making standard additions to typical sample solutions. The authors acknowledge the confirmatory work carried out by Mr. B. Farey of the Thames Watei- Authority, Dr. P. Jackson of the Severn-Trent Water Authority and Mr. A. Stacey of the Yorkshire Water Authority. The authors also thank Mr. W. F. Lester, Director of Scientific Services, Severn-Trent Water Authority, for permission to publish this work. References 1. 2. 3. 4. 5. 6. 8. 9. 10. 11. 12. 13. 14. Allan, J . E., Spectvochim. Acta, 1959, 10, 800. Curtis, K. E., Analyst, 1969, 94, 1068. Price, W. J., and Roos, J . T. H., Spectvochim. Acta, 1971, 26B, 279. Platte, J . A., and Marcy, V. M., A t . Absovpt. Newsl., 1965, 4, 289. Relcher, C. B., Anal. Chiun. Acta, 1972, 62, 87. Analytical Methods Committee, Analyst, 1978, 103, 643. Thompson, K. C., and Reynolds, R. J., “Atomic Absorption, Fluorescence and Flame Emission Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,” Academic Thompson, K. C., Analyst, 1978, 103, 1258. Roos, J . T. H., and Price, W. J., Analyst, 1969, 94, 89. Ferris, A. P., Jepson, W. B., and Shapland, R. C., .4vzalyst, 1970, 95, 574. Department of the Environment, “Analysis of Raw, Potable and Waste Waters,” HM Stationery Thompson, K. C., and Wagstaff, K., Analyst, 1979, 104, 224. Kaye, G. W. C., and Laby, T. H., “Tables of Physical and Chemical Constants,” Thirteenth Edition, Spectroscopy, A Practical Approach,” Charles Griffin, London, 1978. Press, London, 1974. Office, London, 1972. Longmans, London, 1966. Received November 9th, 1979 Accepted Janzuavy 21st, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500641

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, July, 1980, Vol. 105, $9. 651-656 651 Determination of Caesium in River and Sea Waters by Electrothermal Atomic-absorption Spectrometry. Interference of Cobalt and Iron* P. Frigieri, R. Trucco, I . Ciaccolini and G. Pampurini Centro Infovmazioni Studi Espevienze, P.O. Box 3986, 20100 Milan, Italy For the enrichment or the simple recovery of caesium from river and sea waters, selective inorganic exchangers were considered. Ammonium hesa- cyanocobalt ferrate (NCFC) was chosen because it can be used in strongly acidic solutions (with the exception of concentrated sulphuric acid), Cacsium is fully retained by the NCFC chromatographic column and can then be recovered by dissolution in hot sulphuric acid. The solution is then diluted and analysed, either directly or following caesiuni separation, by atomic- absorption spectrometry.To check the reliability of the analytical procedure, a series of cxperirnents were carried out in which the possible interfering species were added to the aqueous caesium solution prior to analysis. The well known ionic interference in flame atomisation processes caused by magnesium, calcium, strontium and metals was investigated by electrothermal atomisation measurements. The experimental data showed that this effect does not occur even when these elements are present in concentrations of the order of thousands of parts per million. However, strong interferences from iron and cobalt were observed. Keywords: Electvothevmal atomic-absovption spectvometvy; caesiwm detev- mination ; cobalt and ivon intevfevemes ; spectval intev fevence ; caesium envichment Nuclear activities, such as electricity production by nuclear power plants, involve the occasional release of small amounts of radionuclides into the environment.Among the possible diffusion paths of such discharges rivers often play a role of paramount importance. Studies on the distribution of radionuclides between river water and the suspended solid materials or sediments can be carried out using radioactive tracer techniques, but they also require analytical methods suitable for the characterisation of the different components of the system. The caesium concentration in these matrices is very low, so that, in addition to a sensitive analytical method, it was necessary to make use of an enrichment technique to bring the caesium concentration within the scope of the analytical method.Atomic-absorption spectrometry was selected as the analytical method because of its sensitivity, speed and ease of operation. For the enrichment or even for the simple recovery of caesium from different solutions, selective inorganic exchangers were considered. Ammonium hexacyanocobalt ferratel (NCFC) was chosen because it can be employed in strongly acidic solutions (with the exception of concentrated sulphuric acid2). The use of NCFC is particularly advantageous for the recovery of caesium from solutions obtained on acid dissolution of mineral or biological materials. Caesium is fully retained by a chromatographic column of NCFC and can then be recovered by dissolution of the NCFC in hot 12 M sulphuric acid.The solution is then evaporated to form a paste, cooled and distilled water is added. When columns of NCFC supported in silica gel are used, the dissolution of the NCFC is made with hot 18 M sulphuric acid with continuous stirring. The solution contains an insoluble residue, which must be washed after filtering to ensure complete recovery of the NCFC. Finally, the solution is diluted with water to give a clear solution, 0.1 &I in sulphuric acid, which is used for the spectrometric analysis. * Presented at the XXI Colloquium Spectroscopicum Internationale, Cambridge, July 1st-6th, 1979.652 FRIGIERI et d.: DETERMINATION OF CAESIUM IN Analyst, vd. 105 Spectrometric Analysis i'ery small amounts of caesium in aqueous solution can be analysed by absorption or emission spectrometry.Atomic-absorption determinations can be performed by flame or electrothermal atomisers. On the other hand, flame emission determinations can be carried out using pre-mixed or total-combustion burners. The following experimental detection limits were found for aqueous solutions: 8 pg 1-l, by flame emission using a total-combustion burner, with an oxygen - acetylene flame; 3 pg l-l, by flame emission using a pre-mixed burner, with an air - hydrogen mixture; 5Opgl-l, by atomic absorption using a flame atoiniser, with an air - acetylene flame; 1 pg l-l, by atomic absorption using an electro- thermal atomiser (Massmann type). Results for solutions with sulphuric acid concentrations ranging between 1 and 0.1 M show detection limits varying with the acid concentration; l O , ~ g l - ~ of caesiuin are detectable in 1 M sulphuric acid and 5 pg 1-1 of caesium are detectable in 0.1 M sulphuric acid.An almost linear graph of absorbance versus caesium concentration up to 0.25 absorbance unit is obtained. A Perkin-Elmer HGA-70 graphite furnace electrothermal atomiser was used, working for 60 s at 100 O C , 30 s at 750 "C and 2500 OC, with a Varian hollow-cathode lamp source, 20 mA current (caesium analytical line at 852.11 nm), a Varian AA5 spectrometer, modified by a Jarrell -Ash 0.5m monochromator and a Hamamatsu R406 photomultiplier, working with a 600 pm x 2 mm slit. A Pye Unicam SP 1900 spectrophotometer, working with a 470-pm slit, was sometimes used as an alternative to the Varian instrument.The most sensitive technique was selected for caesium analysis. Interference Studies The literature on caesium determination 'by atomic spectrometric techniquess-17 is essentially concerned with flame emission and. absorption methods. Information on the possible interferences from matrix or inter-elemental effects in electrothermal atomisation processes is difficult to obtain. To check the reliability of the analytical procedure, a series of determinations were carried out on samples in which the possible interfering species had been added to the aqueous caesium solution. The well known ionic interference in flame atomisation processes caused by magnesium, calcium, strontium and the alkali metals was investigated in the electrothermal atomisation measurements.The experimental data showed that this effect does not occur even in solu- tions where such elements were present in concentrations of the order of thousands of parts per million. Another possible interference effect, i.e., light scattering by sulphuric acid vapour during atomisation of the samples obtained by NCFC dissolution, was avoided by using a suitable thermal pre-treatment procedure. When the effects of the other chemical species from NCFC dissolution were checked, the experimental results showed a very strong spectral interference on caesium absorption measurements. In order to discover the source of the interference the ionic species derived from NCFC dissolution were considered. Interference from iron ions was investigated by the addition of non-complexed iron ions to caesium standard solutions. The results showed that iron depresses the caesium absorption signal. As shown in Fig.1, the depression depends on the ratio of the iron to caesium concentrations, with the instrumental response reaching its minimum value when the ratio is about 5000 to 1 ; at higher ratios, the signal increases with increasing iron to caesium concentration ratio. The interference behaviour suggests that the caesium absorption signal is affected by both the molecular absorption and the emission bands of iron. If the molecular iron com- pound has an emission band superimposed on to the caesium absorption line, the ground- state atoms can absorb the molecular-decay photons which are in competition with those from the source.This is a possible explanation for the depression of the caesium absorption signal with the increase in iron concentration, at least qualitatively. The enhancement of the caesium absorption signal can be ascribed to simultaneous absorp-July, 2980 RIVER AND SEA WATERS BY ELECTROTHERMAL AAS 653 tion by both caesium and iron. This becomes apparent only when the concentration of the iron is high enough to be measured. However, if the same behaviour is considered, taking into account the quantitative effect of each parameter, the explanation suggested above is perhaps less satisfactory. The emission from a molecular iron compound falling within the absorption band width of the caesium atoms will almost certainly be very small compared with emission from the caesium hollow-cathode lamp.It does not seem possible, therefore, that the molecular emission alone is responsible for the reduction of the ground-state atomic population of caesium. This disagreement is particularly evident at the lower interferent to analyte ratios where the depressive interference is most pronounced. Consequently, for a satisfactory explana- tion of the phenomenon other effects, such as atomisation efficiency, must be considered. 0.2 0, C -2 2 a 0.1 a 0 200 400 600 800 1000 Iron concentrationhng I-’ Fig. 1. Interference by iron on the caesium absorption measurements by electrothermal atomic-absorption spectro- metry. Concentration of caesium (pg 1-1) : A, 0; B, 10; C , 20; D, 30; E, 50; F, 80; G, 100. The broken line connects the experimental points relevant to an iron to caseium concentration ratio of 5 000.The presence of an iron emission band should be detected by emission measurements on solutions containing only iron, and an iron absorption band should be detected by absorption measurements made over the caesium resonance line on the same iron solutions. Using pre-mixed burners (air - acetylene, air - hydrogen and air - dinitrogen oxide mixtures) and total-combustion burners (oxygen - hydrogen mixture) the iron emission bands were found by successive scannings in the spectral range between 850 and 854nm. The emission intensity of iron solutions at 852.1 nm (the caesium resonance line) was measured to estimate the emission band strength. The signals for iron in concentrations greater than 10 mg 1-1 are strong enough to allow the quantitative determination of iron at that wavelength.The intensity of the emission band does not vary when use is made of filters that absorb reflections of higher order spectra; this band may be the Fe-0 emission band, the head of which is located as 857.8 nm as pointed out by other authors.ls Using varying iron concentrations, the iron molecular absorption band (for which there are no references in the literature) was detected by electrothermal absorption measurements at the caesium resonance line. The absorbance measured is proportional to the amount of atomised iron, so that it is possible to determine up to 30 pg of iron in the sampled volumes of either 14.1 M hydrochloric acid or 0.25-0.025 N sulphuric acid solutions.The presence and the extent of iron absorption bands were estimated by measurements at the 853.6-nm emission line of a boron hollow-cathode lamp. Iron is determined with almost the same sensitivity in this instance as at the 852.1-nm caesium line.654 Analyst, VoZ. 105 A series of tests were carried out analagous to those used to study the iron interference. The results showed the presence of emission and absorption molecular bands of cobalt chemical species at wavelengths around the 852.1-nm caesiuin absorption line. Measurenients of either emission or absorption intensities at the caesium resonance wavelength permit calibration graphs to be prepared for cobalt in solutions that do not contain caesium. An emission detection limit of 250 pg 1-1 and an absorption sensitivity limit of 100 mg 1-1 were obtained for cobalt.The effect of the cobalt interference on the spectral response of caesium is shown in Figs. 2 and 3. The graphs show the results obtained for solutions of various concentrations of cobalt and caesium. FRIGIERI et aZ. : DETERMIXATION OF CAESIUM IN Further possible interferences from the NCI'C dissolution were then studied. o, 0.3 C $ II 2 0.2 0.1 0 10 30 50 70 Cobalt concentration/mg 1-1 Fig. 2. Interference by cobalt on the caesium absorption measurements by electrothermal atomic-absorption spectro- metry. Concentration of caesium ( p g 1-l) : A, 10; B, 30; C, 50; D, 100. 0, C m .Q, 0.1 II a a 0 500 5 000 Cobalt concentration/mg I- ' Interference by cobalt on the caesium absorption measurements by electrothermal atomic-absorption spcctro- metry.Concentration of caesium (pg 1-l) : A, 10; B, 30; C, 50; D, 100. - 100 - Fig. 3. The presence of iron and cobalt in samples from NCFC dissolution prevents the direct analysis of caesium, unless it is possible to prepare standard samples by dissolution of the same amounts of NCFC containing similar concentrations of caesium to those expected in the samples to be analysed. The suitability of this method was investigated by analysing solutions with caesium con- centrations ranging from 10 to 30 pg 1-1 and IXCFC concentrations ranging from 1000 to 7 500 mg 1-l. The results showed a 10% systematic error at 2 500 mg 1-I of NCFC increasing to 40% at 7500 mg 1-l. The precision of the absorption measurements, expressed as the standard deviation derived from ten repeated measurements on the same sample, varied between 50/, and 8% and the precision of the rrietliod overall was 1.7 pg 1-1 at the 20 pg 1-1 of caesium level and 1.8 pg 1-1 at the 30 pg 1-1 of caesium level.It appears, therefore, that, although a direct analysis can be made by atomic-absorption spectrometry, the method requires such restrictive conditions that it is more convenient to remove iron and cobalt by a preliminary chromatographic separation. It is likely that the use of a spectrometer with a background corrector working in the visible region which would compensate for the enhancement due to molecular absorption might improve the results obtained by direct analysis. Caesium Analysis In view of the above considerations caesium was fixed on an NCFC column, following Caesium is retained on a preliminary separation from iron and cobalt, and then assayed.strong cation-exchange resin from which iron and cobalt are eluted as anionic complexes.July, 1980 Procedure Sample dissolution RIVER AND SEA WATERS BY ELECTROTHERMAL AAS 655 Approximately 3 or 4 g of the exchanger (NCFC supported on silica gel) are added to The beaker is then covered and The heating must be continued until the NCFC is com- The resulting solution is a clear red - violet colour and there is a white 20 ml of concentrated sulphuric acid in a 250-ml beaker. heated to 250 "C on a hot-plate. pletely dissolved. precipitate of silica gel. Filtration and conditioning The solution is cooled and filtered through glass-wool wetted with concentrated sulphuric acid.The glass-wool is then washed with 70-80 ml of concentrated sulphuric acid, which is then collected in a 250-ml beaker and fumed to dryness at 250 "C until the evolution of white fumes is complete. The dried residue is warmed slightly and dissolved in 100 ml of distilled water. After dissolution the volume is increased to 150 ml with distilled water and 20 ml of 0.1 M EDTA solution are then added. The pH of the resulting solution is adjusted to 5 with tribasic sodium citrate (5% solution). The solution is then allowed to stand for at least 1 h. Chromatographic procedure The resulting solution obtained as above is passed through a column, with an internal diameter of 1.5 cm, containing 5 ml of Dowex 50-X8 (50-100 mesh) and previously con- ditioned into the Na+ form, with a flow-rate of 3-4 ml min-l of a 20% sodium chloride solution.The resin is washed twice with 50-ml portions of distilled water. The caesium retained on the column is then eluted with 200 ml of 1 M sulphuric acid at a flow-rate of 3 4 ml min-l. Treatment of the elution solution To remove the sodium eluted with the caesium, the eluate is evaporated and fumed to dryness. A few millilitres of distilled water are then added and the solution is then diluted with 120 ml of 95% ethanol and the resulting solution is warmed slightly to coagulate the sodium sulphate precipitate. The solution is cooled and filtered through a 5-pm porous PTFE Millipore filter. The filter and the beaker are washed with 20ml of 95% ethanol.The ethanoholic solution is then evaporated to 1-2ml and diluted to 100ml with water. The pH of the final solution should be between 1.4 and 1.8. Caesium spectrometric analysis The solution obtained from the separation process is analysed by atomic-absorption spectrometry using the following operating conditions : thermal pre-treatment, 60 s at 100 "C and 30 s at 750 "C; electrothermal atomisation in a Perkin-Elmer HGA-70 graphite furnace for 3 s at 2500 "C; spectrometric analysis using either a Varian AA5 or a Pye Unicam SP 1900 spectrometer. The calibration graphs are obtained using an NCFC inorganic exchanger with standard solutions of caesium and assayed by the procedure described above. Table I shows the accuracy and the reproducibility of the method for the assay of various amounts of caesium on NCFC.Conclusions The determination of trace amounts of caesium in natural matrices is necessary for studying the diffusion of caesium radionuclides into the environment following discharges from nuclear power plants. While the determination of the radioactive isotopes of caesium, either in the original samples or in the inorganic exchangers used for caesium enrichment, is not difficult, the quantitative analysis of total caesium is less easy because of the low concentrations involved. The use of selective inorganic exchangers allows trace amounts of caesium to be concentrated sufficiently for determination by direct atomic-absorption analysis, although decomposition products from the exchanger can give rise to interferences.656 FRIGIERI, TRUCCO, CIACCOLINI AND PAMPURINI TABLE I REPRODUCIBILITY AND ACCURACY OF THE SPECTROCHEMICAL PROCEDURE FOR THE QUANTITATIVE DETERMINATION OF CAESIUM RETAINED BY NCFC INORGANIC EXCHANGER Caesiumlpg I Expected 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Found 9.0 11.0 9.5 11.5 10.5 10.0 10.0 10.5 11.0 Average .. . . . . 10.3 Coefficient of variation, yo.. 8 Expected 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 Found 46.0 52.5 53.0 50.0 45.0 47.0 52.0 51.0 53.0 Expected 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 -1 Found 101.0 108.0 111.0 103.0 98.0 98.5 92.5 92.5 49.9 100.6 6 7 An analytical procedure employing a preliminary chromatographic separation on a strong cation-exchange resin proved to be the most convenient, versatile and reliable method to meet our objectives. Part of this work was sponsored by ENEL. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Petrow, H. G., and Levine, H., Anal. Chew., 1967, 39, 360. Cosmai, G., Croce, E., Pioltini, F., and Trucco, I<., Nota Tecnica, CISE-77.055, 1977. Mountjoy, W., U.S. Geol. Sztrv. Pvot. Pap., 1968, 600B, B119. Pouget, R., Rapp. C.E.A., 1963, 2285. Horstman E. L. Anal. Chem., 1956, 28, 1417. Vajnshtejn, E. E., Geokhimiya, 1961, 4, 362. Osterried, O., 2. Anal. Chem., 1964, 199, 260. RubeSka, I., Collect. Czech. Chem. Cornwun., 1965, 30, 1731. Folsom, T. R., Feldman, C., and Rains, T. C., Science, 1964, 144, 538. Feldman, C., and Rains, T. C., Anal. Chem., 1964, 36, 405. Folsom, T. R., Sreekumaran, C., Weitz, A. E., Jr., and Teunant, D. A., A p p l . Spectrosc., 1968, 22, Folsom, T. R., “Reference Methods for Marine Radiation Studies,” I.A.E.A., Vienna, 1970. Allen, A. J . F., Anal. Chim. Acta, 1972, 59, 111. Govindaraju, F., Hermann, R., Mevelle, G., and Chouard, C., A t . Absovpt. Newsl., 1973, 12, 73. Langmyhr, F. J., 2. Anal. Chew., 1973, 264, 122. RubeSka, I., Chew. Listy, 1973, 67, 1197. Felsom, T. K., Hansen, N., Parks, G. J., Jr., and Weitz, W. E., Jr., Appl. Spectrosc., 1974, 28, 345. Gaydon, A. G., “The Spectroscopy of Flames,” Second Edition, Halsted Press, London, 1974, p. 356. Received Septembev 17th, 1979 Accepted January 21st, 1980 109.

ISSN:0003-2654    

DOI:10.1039/AN9800500651

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, July, 1980, Vol. 105, pp. 657-662 Automated Spectrophotometric Determination of Trace Amounts of Bromide in Water 657 G. S. Pyen, M. J. Fishman" and A. G. Hedley US Geological Survey, Mail Stop 407, Box 25046, Denver Federal Center, Denver, Colo. 80225, USA A rapid, sensitive and accurate indirect automated spectrophotometric method for the determination of bromide in natural waters is based on the catalytic effect of bromide on the oxidation of iodine to iodate by potassium permanganate in sulphuric acid solution. Twenty samples per hour can be analysed to levels of 0.01 mg 1-1 of bromide. Samples containing known added amounts of bromide gave values ranging from 94 to 110% of the con- centration calculated to be present. Bromide in several samples was deter- mined independently by a similar manual method, and the correlation coefficient was 0.98.Keywords : Bromide determinatzon ; spectrophotometry ; automation ; natuval waters An increasing number of requests for measuring the bromide concentration of natural waters are being made to US Geological Survey laboratories. At present, an indirect spectrophoto- metric technique, described by Fishman and Skougstadl and Skougstad et al.,2 is used. The method is based on the catalytic effect of bromide on the oxidation of iodine to iodate by potassium permanganate in sulphuric acid solution. When an aqueous solution of iodine and permanganate is mixed with carbon tetrachloride, the iodine is extracted into the carbon tetrachloride. If bromide is present with the aqueous iodine and permanganate for a certain time at a particular temperature prior to mixing with carbon tetrachloride, the amount of iodine remaining unreacted to iodate will decrease linearly with increasing bromide concentra- tion.Iodine in a carbon tetrachloride extract can, therefore, serve as a sensitive measure of the bromide concentration in the aqueous solution. A skilled operator is required, as the reaction temperature and the time for each reaction step are very critical. Because of the large number of samples being processed, a potentially hazardous situation also exists, as the analyst is exposed to carbon tetrachloride. An automated system is ideally suited to catalytic methods; reaction times are repro- ducible at each step of the process and the temperature is easily controlled.Normally, greater numbers of samples can be processed than by manual methods. In addition, the analyst comes into little contact with the organic solvent, because it is introduced into the system automatically. The automated method described in this paper uses the identical chemistry of the manual catalytic method. The solvent extraction apparatus is similar to that used by Carter and Nickless3 for the automated determination of trace concentrations of copper. Experimental Apparatus? The automated equipment consists of a Technicon sampler, proportioning pump, manifold, spectrophotometer, voltage stabiliser and recorder. The spectrophotometer is fitted with specially designed matched 50-mm flow cells (see Fig. 1) and 520-nm filters. The sampler is used with a 20 per hour (2/7) cam made by cutting off four sampling lobes (Fig.2) from a 60 per hour (2/1) cam. This results in a cycle time consisting of 40 s for sampling and 140 s for rinsing. To separate the two phases, a five-turn mixing coil is placed perpendicular to the manifold; PTFE tubing is A 17-turn coil containing glass beads is used as the extractor. * To whom correspondence should be addressed. f The use of brand names in this paper is for identification purposes only and does not imply endorse- ment by the US Geological Survey.658 PYEN et a,?. : AUTOMATED SPECTROPHOTOMETRIC Analyst, VoZ. 105 Rear view d Bottom view Fig. 1. Flow cell. Fig. 2. Alteration of 60 per hour (2/1) cam. inserted inside the separating coil to facilitate rapid separation.All connections are made with glass tubing or silicone-rubber sleeving. The final separator, as shown in Fig. 3, is similar to that used by Carter and Nickle~s,~ except that the top portion is enlarged to facilitate the clean separation of solvent from the aqueous phase. Three 1-1 flasks (Fig. 4) are used for the displacement of carbon tetrachloride. Connections are made with 1.0-mm PTFE tubing. A circulating System-255 cold bath (GCA/Precision Scientific) maintains the reaction temperature of the condenser coil at 10 "C. From separating coil PTF E tubing IF 1 To spectrophotorneter cell Fig. 3. Separator. Reagents To waste All chemicals were of analytical-reagent grade. Potassium bromide stock solution, 0.1 mg ml-l of bromide. bromide in demineralised water and dilute to 1 1 with water. Dissolve 0.149 g of potassium Prepare a series of standardsJuly, 1980 DETERMINATION OF TRACE AMOUNTS OF BROMIDE I N WATER 659 in the concentration range 0.01-0.20 mg 1-1 of bromide by appropriate dilution of the stock solution. Dissolve 6.32 g of potassium permanganate in demineralised water and dilute to 1 1 with water.Store in an amber-glass bottle and refrigerate. Dissolve 1.31 g of potassium iodide, dried overnight over concentrated sulphuric acid, in about 600 ml of demineralised water. Cool the solution and dilute to 1 1 with demineralised water; store in an amber-glass bottle and refrigerate. If iodine crystals form on standing, prepare a fresh solution. Dissolve 50 g of oxalic acid (H2C,0,.2H20) in demineralised water and dilute to 1 1 with water.Potassium $ermanganate solution, 0.04 M. Potassium iodide - sulfihuric acid solation, 1 mg ml-l of iodide. Add slowly, with stirring, 350 ml of concentrated sulphuric acid. Oxalic acid solution, 50 g 1-l. Extractor coil, 17 turns, 116-0101-01 PT Flow-rate/ I.d./in ml min-' Sampler IV, 20h-', 2/7 cam . __ 5-turn 'Oil 1 0.073 2.00 Sample 0.020 0.16 iodide Potassium < 1 I ~- 114-0209-01 I I Potassium 10.020 0.161 permanaganate T , I To sampler I V wash , 10.1 10 3.90 I Water waste - Spectroph - 0 I I Fig. 4. Bromide manifold. Procedure Set up the manifold system (Fig. 4) in a well ventilated hood to avoid contact with carbon tetrachloride vapour. Place the potassium iodide and potassium permanganate solutions in an ice-bath.Feed demineralised water into all reagent and sample lines, allowing sufficient time for good separation of carbon tetrachloride from the aqueous phase. Check both reference and sample flow cells to ensure that they are filled with carbon tetrachloride. If water droplets are present in the cells, the recorder pen will be deflected below the base line. Optically peak the spectrophotometer as instructed in the manufacturer's manual and then set the reversing switch under the spectrophotometer cover to position I (inverse). Set the CAL control to position 2.0 and lock, and the Display Rotary Switch to Damp 1. Feed the reagents through the system. Reaction will occur, and iodine will be extracted into the carbon tetrachloride, resulting in deflection of the recorder from full scale to the base line.Adjust the base line to read approximately 5 chart divisions and allow it to stabilise (approxi- mately 20min). Beginning with the highest standard, place a complete set of standards covering a concentration range from 0.01 to 0.20 mg 1-1 in the first positions of the sample tray, followed by a blank. Place individual standards of different concentrations in several positions of the remainder of the tray. Place oxalic acid solution in the thirty-eighth position, followed by two blanks. Fill the remainder of the tray with unknown samples and begin the analysis.660 PYEN et al. : AUTOMATED SPECTROPHOTOMETRIC Analyst, VoZ. 105 Prepare a calibration graph by plotting the height of each standard peak against its respective bromide concentration.With the CAL control in position 2.0, the peak height reading for 0.20 mgl-1 of bromide using a Technicon recorder was about 30 chart units. Calculate the bromide concentration of each sample by comparing its peak height with the calibration graph. Any base-line drift that may occur must be taken into account when calculating the height of a sample or standard peak. TABLE I COMPARISON AND PRECISION OF BROMIDE RESULTS BETWEEN AUTOMATED AND MANUAL METHODS USING STANDARD REFERENCE WATER SAMPLES (SRWS) Automated method SRWS No. 53 55 58 59 61 65 66 Mean*/ mg 1-l 0.017 0.171 0.015 0.035 0.084 0.069 0.141 Standard deviation/ mg 1-1 0.002 0.006 0.003 0.004 0.008 0.005 0.010 Relative standard deviation, 11.8 3.5 20.0 11.4 9.5 7.2 7.1 % Meant/ mg 1-l 0.015 0.137 0.012 0.034 0.076 0.068 0.153 Manual method Standard deviation/ mg 1-1 0.001 0.006 0.003 0.001 0.003 0.004 0.005 1 Relative standard deviation, 6.7 4.4 25.0 2.9 4.0 5.9 3.3 % * Values based on 20 replicate determinations. t Values based on 10 replicate determinations.Results and Discussion One analyst can process 30 samples per day by the manual catalytic method, and greater production without loss of accuracy and precision was the aim of automating this technique. To achieve reproducible results, a 3540-s sampling cycle was found to be necessary, and more than a 120-s rinse cycle was required for the signal to return to the base line between samples. Because the cams operate on a 6-min cycle, a 40-s sampling and a 140-s rinse were chosen; this required modification of a standard cam (described earlier in this paper).With this modification, 20 samples per hour can be processed. TABLE I1 COMPARISON OF RESULTS BETWEEN MANUAL AND AUTOMATED METHODS ON SURFACE WATER SAMPLES Sample No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Bromide concentration/mg 1-1 P -I Manual method Automated method 0.112 0.120 0.170 0.184 0.051 0.060 0.132 0.123 0.141 0.134 0.150 0.145 0.109 0.110 0.015 0.021 0.078 0.084 0.091 0.088 0.036 0.038 0.021 0.028 0.034 0.032 0.153 0.143 0.040 0.037 0.024 0.020 0.111 0.118 0.074 0.083 0.141 0.130 0.061 0 057July, 1980 DETERMINATION OF TRACE AMOUNTS OF BROMIDE IN WATER 661 The flow cells are specially designed, as a de-bubbler is not required. One piece of 2 mm i d . glass tubing was used, with a side-arm at an angle of 45" from the horizontal plane to facilitate the flow of organic solvent (a competent glassblower should be able to construct the cells).Because the chemistries of the methods are identical, no interference studies were under- taken. The manual method operated a t 0 "C; however, in the automated mode, temperatures above 0 "C gave better sensitivity. A problem that occurred with the increased temperature was that the interior walls of the system became coated with manganese dioxide owing to the oxidation of potassium permanganate ; this resulted in non-reproducible results. To keep the coating to a minimum, and to minimise the loss in sensitivity, 10 "C was found to be the optimum temperature. Even at this temperature, a slight coating does occur, but this does not hinder the reactions if an oxalic acid solution is passed through the system every 2 h; the oxalic acid reduces the manganese dioxide. Both Geological Survey Standard Reference Water Samples (SRWS) and natural surface- water samples were used to evaluate the automated procedure.The results were compared with those obtained by the manual catalytic procedure. The only change in the automated method was temperature. 0.20 0.18 0.16 - I - ~ 0.14 -g 0.12 5 0.10 E 1 +J +J 3 - 0.08 2 E 0.06 2 a 0.04 0.02 0 Fig. 5. 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 Bromide (manual)/mg I-' Correlation of the manual and automated methods on surface-water samples. Table I gives both the precision and comparison of results by the automated and manual methods on a number of SRWS containing bromide.Owing to a lack of data received from other laboratories, an inter-laboratory mean is not available ; however, manual results were obtained independently. Relative standard deviations ranged from 3.5 to 20.0% for the automated method and from 2.9 to 25.0% for the manual method. Except for SRWS No. 66, the results are higher by the automated method on all the other standard reference water samples. An extensive comparison study was undertaken between methods on 88 surface-water samples to check further for bias. Fig. 5 shows the plot of these data; no bias occurs and the correlation coefficient is 0.98. Table I1 gives typical examples of these data. To determine the reliability of the method, 15 surface-water samples were spiked with known amounts of bromide; the recoveries (Table 111) ranged from 94 to llOyo.662 PYEN, FISHMAN AND HEDLEY TABLE I11 RECOVERY OF BROMIDE ADDED TO SURFACE-WATER SAMPLES AS MEASURED BY THE AUTOMATED METHOD Bromide concentration/mg 1-1 Sample f A \ No.Present* Added Total Found? 128-062 128-063 128-095 1 3 5-02 1 113-010 128-073 134-207 134-209 134-21 3 134-288 135-122 152-097 155-129 157-161 159-188 0.093 0.117 0.128 0.063 0.020 0.084 0.178 0.117 0.130 0.081 0.084 0.067 0.084 0.036 0.061 0.10 0.05 0.03 0.05 0.02 0.05 0.02 0.01 0.01 0.03 0.04 0.04 0.02 0.02 0.05 0.193 0.167 0.158 0.113 0.040 0.134 0.138 0.127 0.140 0.112 0.124 0.107 0.104 0.056 0.111 0.190 0.166 0.155 0.121 0.038 0.129 0.138 0.125 0.144 0.105 0.136 0.103 0.109 0.054 0.115 Recovery, 98 99 98 107 95 96 100 98 103 94 110 96 105 !3 6 104 % * Values based on 10 replicate determinations. t Values based on 7 replicate determinations. Conclusions The automated method is less hazardous than the manual method, does not require a It is rapid, accurate skilled analyst and is ideally suited for a high-production laboratory. and precise. References 1. 2. Fishman, M. J., and Skougstad, M. W., Anal. Chew., 1963, 35, 146. Skougstad, M. W., Fishman, M. J., Friedman, L. C., Erdmann, D. E., and Duncan, S. S., “Tech- niques of Water-Resources Investigations of the U.S. Geological Survey, Methods for the Deter- mination of Inorganic Substances in Water and Fluvial Sediments,” 1979, Book 5, Chapter A l , p. 329. 3. Carter, J. M., and Nickless, G., Analyst, 1970, 95, 148. Received December 13th, 1979 Accepted March 4th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500657

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, Jub, 1980, Vol. 105, fip. 663-668 663 Spectrophotometric Method for the Determination of Phenothiazines and its Application to Phenothiazine Drugs P. G. Ramappa, H. Sanke Gowda and Anant N. Nayak Department of Post-graduate Studies and Research in Chemistry, University of Mysore, Manasagangotri, Mysore 570 006, India A method for the spectrophotometric determination of four phenothiazines (promazine hydrochloride, mepazine hydrochloride, chlorpromazine hydro- chloride and prochlorperazine maleate), based on the coloured compounds formed between the phenothiazines and molybdoarsenic acid, is described. The infrared and electron spin resonance spectra of these coloured com- pounds showed that the molybdoarsenic acid oxidises phenothiazines to a radical cation with which it subsequently forms the coloured compound.The method is simple and rapid. The influence of the substrates commonly employed as excipients with phenothiazine drugs was studied. The proposed method has been applied to the analysis of commercial phenothiazine- containing preparations, the results of which are in good agreement with those obtained by the official method of the British Pharmacopoeia. Keywords : Phenothiazine determination ; spectrophotovnetry ; molybdoarsenic acid During the past 25 years, interest in mental health and the continued introduction of new phenothiazine drugs have resulted in an extensive literature on the analysis of phenothiazines. The methods available for the determination of phenothiazines are all in some way unsatis- fa~tory,l-~ and there is at present a great demand for a rapid, simple and accurate method for the routine quality control of phenothiazines.We therefore investigated the colour reaction between molybdoarsenic acid and four phenothiazines [promazine hydrochloride (PMH), mepazine hydrochloride (MH) , chlorpromazine hydrochloride (CPH) and prochlor- perazine maleate (PCPM)]. The proposed method offers the advantages of simplicity, rapidity, reasonable sensitivity and a wide range of concentrations. Further, the reaction occurs at room temperature and without the need for extraction. Experimental Apparatus A Beckman, Model DB, spectrophotometer with matched 10-mm silica cells was used for absorbance measurements. A Perkin-Elmer, Model 257, infrared spectrophotometer and a Varian E-4 X-band electron spin resonance spectrometer were used for spectral analyses of the coloured compounds.Reagents Molybdoarsenic acid was obtained from Doxa Laboratory, India. A 4% m/V aqueous solution was prepared. Stock solutions of PMH, MH, CPH and PCPM were prepared by dissolving the appropriate amount of sample in water, and were then standardised gra~imetrically.~ Working solutions were prepared as required by dilution. All other reagents were of analytical-reagent grade. Doubly distilled water was used throughout. Molybdoarsenic acid. Phenothiazines. Procedure Procedure and preparation of calibration graph CPH or 155-1 180 pg of PCPM were transferred into a series of 25-ml calibrated flasks. Various aliquots containing 100-890 pg of PMH, 100-930 pg of MH, 140-1 115 pg of A664 RAMAPPA et al.: SPECTROPHOTOMETRIC METHOD FOR Analyst, Vol. 105 10-ml volume of 4% molybdoarsenic acid solution was added to each flask, diluted to the mark with water and mixed well. The absorbance was measured at 525 nm for PMH and MH and at 540 nm for CPH and PCPM against the corresponding reagent blank. The graph of absorbance against the concentration of PMH, MH, CPH or PCPM was a straight line passing through the origin. These calibration graphs were subsequently used in the sample determinations. Analysis of synthetic mixtares containing phenothiaxines A portion of the mixture containing about 20mg of PMH, MH, CPH or PCPM was accurately weighed. Four portions of 15 ml of water were used to extract the phenothiazines from the powder before filtering the mixture.The residue after filtration was washed with 20ml of water. The filtrate and the washings were then combined in a 100-ml calibrated flask and the volume was made up with water. An aliquot of this solution was treated as described under Procedure and preparation of calibration graph. Four mixtures with the compositions given in Table I were prepared. The results are given in Table I. TABLE I RECOVERY OF PHENOTHIAZINES FROM VARIOUS EXCIPIENTS IN SYNTHETIC MIXTURES BY THE PROPOSED METHOD Amount Phenothiazine present/mg Excipients PMH .. . . 120 Talc (400 mg), magnesium stearate (500 mg), starch (450 mg), dextrose (250 mg) and gelatin (240 mg) (50 mg), starch (400 mg), gelatin (20 mg) and dextrose (30 mg) CPH .. . . 150 Talc (200 mg), magnesium stearate (175 mg), starch (300 mg) and dextrose (40 mg) PCPM .. . . 140 Talc (120 mg), magnesium stearate (150 mg), starch (200 mg), gelatin (30 mg) and dextrose (40 mg) MH . . . . 100 Talc (80 mg), magnesium stearate * Average recovery from five experiments. Recovery, 98.77 %* 99.38 98.40 98.56 Relative standard deviation, yo 0.92 1.24 1.12 1.46 Analysis of phenothiaxine drugs An amount of powder containing about 30 mg of CPH or PCPM was weighed accurately and assayed as described under Analysis of synthetic mixtures containing phenothiaxines. For the analysis of injection solutions and syrup, the appropriate volume of sample was transferred into a 100-ml calibrated flask and diluted to the mark with water. The pheno- thiazine content in the diluted solution was determined as described above.The samples were also assayed by the BP method5 for comparison, and the results are given in Table 11. Twenty tablets were weighed and powdered. Results and Discussion Molybdoarsenic acid is used for the identification and determination of various organic compounds.6-11 It oxidises PMH, MH, CPH and PCPM in weakly acidic solutions to a radical cation, which then forms a coloured compound with the unreacted rnolybdoarsenic acid. Maximum intensity of the colour is obtained instantaneously in solutions of pH 0.9-1.7 (Table 111). If the pH is below 0.9, the coloured solutions become turbid and if it is above 1.7 the colour intensity decreases rapidly. No coloured product is obtained in solutions of pH 6.0 or greater.Hence a medium of pH 1.2, obtained by mixing solutions of molybdo- arsenic acid and the phenothiazine, was selected for further studies. Spectral Characteristics MH, CPH and PCPM in the range 350-650 nm are shown in Fig. 1. The absorption spectra of molybdoarsenic acid and its coloured compounds with PMH, PMH and MH withJluZy, 1980 PHENOTHIAZINES AND APPLICATION TO PHENOTHIAZINE DRUGS TABLE I1 DETERMINATION OF PHENOTHIAZINE DRUGS IN COMMERCIAL PHARMACEUTICAL PREPARATIONS 665 Phenothiazine Sample* present Tablets- Largactil . . . . CPH Stemetil. . .. .. PCPM Largactil . . . . CPH Stemetil. . .. . . PCPM Largactil . . .. CPH * All supplied by May & Baker Ltd. t Average of five determinations. Injections- Syrup- Phenothiazine contentlmg I A 7 Results ------7 Proposed Label claim BP method methodt 25.0 24.26 24.84 5.0 5.08 4.90 25.0 24.92 24.90 12.5 12.50 12.31 5.0 - 4.94 hydrogen at C-2 form red compounds with absorption maxima at 525 nm, whereas chlorine at C-2 in CPH and PCPM causes a bathochromic effect, shifting the wavelength of maximum absorption to 540 nm.PMH, MH, CPH, PCPM and molybdoarsenic acid all have negligible absorbance at and near the wavelength of maximum absorption and therefore the analytical conditions are excellent. TABLE I11 EFFECT OF pH ON ABSORBANCE OF COLOURED PRODUCTS FORMED BETWEEN PMH, MH, CPH AND PCPM AND MOLYBDOARSENIC ACID Absorbance of coloured product of I - PH PMH MH CPH PCPM 0":; 1 All solutions became turbid 0.7 J 0.8 Turbid 0.3372 Turbid Turbid 0.9 0.3989 0.327 9 0.279 8 0.240 3 1.1 0.392 5 0.332 5 0.275 7 0.244 1 1.2 0.3925 0.332 5 0.2757 0.244 1 1.3 0.3925 0.327 9 0.275 7 0.244 1 1.4 0.3925 0.3325 0.275 7 - 1.5 0.3925 - 0.275 7 0.244 1 1.6 0.3925 0.332 5 0.275 7 0.244 1 1.7 0.392 5 0.3325 0.275 7 0.240 3 1.8 0.3925 - 0.275 7 0.232 8 1.9 0.387 2 0.3235 0.275 7 0.207 6 2.5 0.3468 0.275 7 0.1308 0.1972 4.5 0.2218 0.122 1 0.1300 0.187 1 i:: } All solutions became turbid.Infrared and Electron Spin Resonance Spectra In order to study the nature of the compound formed between PMH, MH, CPH or PCPM and molybdoarsenic acid, infrared spectra of the compounds and molybdoarsenic acid were measured in the solid phase (potassium bromide discs) in the region 6504000 cm-1. The R,NH+ group combined with halide ion, X-, in the molecule of many phenothiazines gives rise to a broad band in the 2300-2500cm-l range.12 For example, PMH shows a broad band between 2 300 and 2 540 cm-l, which disappears when it forms a compound with molyb- doarsenic acid.This broad band is due t o the side-chain in the PMH molecule, ie., -CH,-NH(CH,), combined with a C1- ion. The disappearance of this band in the coloured +666 RAMAPPA et al. : SPECTROPHOTOMETRIC METHOD FOR Analyst, Vol. 105 compound shows that the nitrogen in the side-chain is the reaction site, as is usually the case with organic bases.13J4 The position of the symmetric C-S stretching band at 740cm-l in PMH is found to be unaffected by the reaction, suggesting that the sulphur atom remains non-coordinated in the reaction product. The infrared spectra of the other phenothiazines and their coloured products with molybdoarsenic acid show similar behaviour.The infrared spectra of the products show the characteristic absorption band of molybdoarsenic acid. 0 ' 1 I I I I 350 400 450 500 550 600 650 Wavelength/nm Fig. 1. Absorption spectra of the coloured products of: A, PMH; B, MH; C, CPH; D, PCPM; and E, molybdoarsenic acid. Phenothiazine concentration = 20 p.p..m. in each instance. Molybdoarsenic acid con- centration = 6.955 x M. The electron spin resonance signals show that free radicals are formed by the oxidising action of molybdoarsenic acid on the phenothazine nucleus. The g values calculated from the electron spin resonance spectra are 2.0024, 2.0033, 2.0032 and 2.0023 for the compounds of PMH, MH, CPH and PCPM, respectively.These values are in close agreement with the g values for free electrons (2.0023). Hence the coloured product obtained when an aqueous solution of PMH, MH, CPH or PCPM is treated with molybdoarsenic acid is the molyb- doarsenate of the cationic radical formed by oxidation. This conclusion is given further support by the finding that the A,,,. found experimentally corresponded to that reported in the literature for the free radicals formed by oxidation.15 Stability of the Colours The reaction between phenothiazines and molybdoarsenic acid is almost instantaneous at room temperature (27 "C). The colour is fully developed 1 min after mixing the reagents and essentially constant absorbance readings are obtained over a period of 25, 30, 60 and 40 min for PMH, MH, CPH and PCPM, respectively.The order of addition of reagents is not critical. The absorbance is not affected by temperature in the range 5-60 "C for PMH, 5-65 "C for MH and CPH and 5-55 "C for PCPM.July, 1980 PHENOTHIAZINES AND APPLICATION TO PHENOTHIAZINE DRUGS 667 Effect of Reagent Concentration The effect of changes in the concentration of molybdoarsenic acid was studied by measuring the absorbance at the wavelengths recommended in the general procedure, for solutions containing a fixed concentration of the phenothiazine and various amounts of molybdoarsenic acid. The rate of formation and the colour intensity of the product initially increased with increasing concentration of molybdoarsenic acid, but constant absorbance readings were obtained in the ranges 2-15, 3-15, 4-15 and 4-16 ml of 4% molyb- doarsenic acid solution for PMH, MH, CPH and PCPM, respectively.A large excess of molybdoarsenic acid scarcely affected the sensitivity, and therefore 10 ml of 4% molyb- doarsenic acid in a total volume of 25 ml was used in all subsequent work. The results are shown in Fig. 2. 0.5 0.4 W m 2 0.3 +? a z 0.2 0.1 0 -~xx-x-x-x-x D , I I I I 2 4 6 8 10 12 14 Volume of 4% molybdoarsenic acid solution added/ml Fig. 2. Effect of molybdoarsenic acid concentration on the absorbance a t A,,,. of the coloured products of: A, PMH; B, MH; C, CPH; and D, PCPM. Phenothiazine concentration = 20 p.p.m. in each instance. Calibration Graphs, Range and Sensitivity Beer’s law is valid over the concentration range 2-36 p.p.m.of PMH, 2 4 0 p.p.m. of MH, 4-45 p.p.m. of CPH and 448 p.p.m. of PCPM. To evaluate the optimum range and the analytical accuracy, a Ringborn’s curvel6,l7 was drawn by plotting the percentage transmittance as ordinate against the logarithm of the concentration of phenothiazine as abscissa. The ranges as derived from the maximum slope of the curves were 4-35.6, 4-37.6, 5.644.6 and 6.2-47.2 p.p.m. of PMH, MH, CPH and PCPM, respectively. The Sandell sensitivitiesl8 of the reactions as calculated from the Beer’s law data are 53.3, 61.4, 72.5 and 85.4 ng cm-2, and the corresponding molar absorptitivites are 6.02 x lo3, 5.64 x lo3, 4.90 x 1 0 3 and 7.09 x lo3 1 mol-1 cm-1 for PMH, MH, CPH and PCPM, respectively. The calibration graphs were prepared by the procedure described above.Effect of Concomitant Substances and Applications to Phenothiazine Drugs In order to assess the possible analytical applications of the method, the effects of various ions and substances that often accompany phenothiazines in pharmaceutical formulations were studied. In these studies, various amounts of the ionic species were added to 20 p.p.m. of PMH, MH, CPH or PCPM solution in 25-ml calibrated flasks and the colours were developed exactly as described above. Ascorbic acid, sulphite and iodide ions reduce molybdoarsenic acid to molybdenum blue and therefore interfere in the determinations. The tolerance limits for the concentrations of the various ions causing an error of &2.5y0 are given in Table IV. To test the accuracy of the method, recovery experiments were performed on synthetic mixtures prepared in the laboratory.The usual tablet diluents and excipients were found not to interfere in the analysis by the proposed method (Table I). In some instances, where large amounts of interfering ions were present, preliminary extraction with diethyl ether in the presence of 5% of sodium hydroxide was carried out.668 RAMAPPA, SANKE GOWDA AND NAYAK TABLE IV EFFECTS OF VARIOUS SPECIES ON THE DETERMINATION OF 20 p.p.m. OF PMH, MH, CPH AND PCPM Species added Fluoride . . .. .. Chloride . . .. . . Bromide . . . . .. Iodide . . . . . . Nitrate . . .. .. Sulphate . . .. .. Sulphite . . .. . . Phosphate . . .. .. Carbonate . . .. .. Acetate . . .. .. Citrate .. . . .. Formate . . .. .. Oxalate . . . . .. Tartrate .. . . .. Ascorbic acid * . .. Dextrose . . .. .. Gum acacia . . . . .. Alginate . . .. .. Barbitone . . .. .. 7- PMH 350 3 000 2 000 25 2 000 2 000 4 1500 1500 500 500 350 200 500 4 6 000 80 700 4 000 Tolerance limit,” p.p.m. A MH CPH 250 100 2 500 400 2 250 800 20 10 1500 400 1750 500 6 3 1750 1200 1500 1000 450 160 500 250 400 200 160 120 450 200 3 2 6 500 4 500 100 100 800 800 3 600 3 500 PCPM- 80 400 700 8 300 400 2 1000 800 180 200 180 100 150 2 4 000 120 500 3 000 * Amount causing an error of less than 2.5%. The organic phase was then extracted with 0.5 M hydrochloric acid and the extract diluted to a suitable concentration with water. The proposed method was successfully applied to the determination of phenothiazine drugs in various pharmaceutical preparations.The results of the assays of tablets, injections and a syrup presented in Table I1 compare favourably with the quoted values, and with those obtained by the official method of the British Pharmacopoeia.5 The authors thank John Wyeth and Brother, Ltd., India; May and Baker, Ltd., India; Chemische Fabrik Promonta GmbH, Germany ; and British Pharmaceutical Laboratories, India, for supplying pure phenothiazines. One of the authors (A.N.N.) is grateful to U.G.C., New Delhi, the University of Mysore and the Government of Karnataka for the award of a teacher fellowship under the U.G.C. Faculty Improvement Programme. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Fairbrother, J. E., Pharm. J., 1979, 271. Issa, A. S., Beltagy, Y. A., and Mahrous, M. S., Talanta, 1978, 25, 710. Blazek, J., Dymes, A,, and Stejskal, Z., Pharmazie, 1976, 31, 681. Blazek, J., and Mares, V., Cesk. Farm., 1966, 15, 349. “British Pharmacopoeia 1973,” HM Stationery Office, London, 1973, pp. 107 and 388. Mitchell, D., Shaw, E. H., Jr., and Frary, G. G., Proc. S. D. Acad. Sci., 1944, 24, 108. Alt, F., and Umland, F., Fresenius 2. Anal. Chem., 1975, 274, 103. Lorant, B., Fresenius 2. Anal. Chem., 1975, 274, 125. Master, I., Stefan, M., and Budiu, T., Romanian Pat., 56,745, 1974; Chem. Abstr., 1974, 81, Hahn, H., and Wagenknecht, R., 2. Anal. Chem., 1961, 182, 343. 172464.. Sternberg, M., U.SyPat. 592,739, 1971; Chem. Abstr., 1971, 75, 97284e. Bodea, C., and Silberg, I., in Katritzky, A. R., and Boulton, A. J., Editors, “Advances in Hetero- Pascal, P., “Nouveau Traite de Chimie Minerale,” Tome XIV, Masson, Paris, 1959, p. 914. El-Dorry, H. F. A., Medina, H., and Bacila, M., Anal. Biochem., 1972, 47, 329. Meunier, J., Viossat, B., Leterrier, P., and Douzou, P., Ann. Pharm. Fr., 1967, 25, 683. Ringborn, A., 2. Anal. Chem., 1938, 115, 332. Ayres, G. H., Anal. Chem., 1949, 21, 652. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Interscience, New York, 1959, Received December loth, 1979 Accepted February 21st, 1980 cyclic Chemistry,” Volume 9, Academic Press, New York, 1968, p. 338. p. 80.

ISSN:0003-2654    

DOI:10.1039/AN9800500663

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, July, 1980, Vol. 105, $p. 669-673 669 Spectrophotometric Method for the Determination of Procarbazine Hydrochloride in Capsules Sonia T. Hassib Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Cairo University, Cairo, Egypt An accurate, sensitive and selective method for the determination of Natulan is proposed. The method is based on the oxidation of Natulan with 0.1 N ammonium cerium( IV) sulphate solution to formaldehyde and N-isopropyl benzamide-4-carboxaldehyde. The latter is separated as its 2,4-dinitrophenyl- hydrazone, which, when treated with ethanolic potassium hydroxide solution, gives a wine red colour. The method is applicable to microgram amounts and can be used successfully in the determination of Natulan in capsules. The stoicheiometry of 1 mol of Natulan to 6 mol of cerium(1V) has been shown through a titrimetric study.Keywords : Determination of Procarbazine hydrochloride ; spectrophotometry ; ammonium ceriurvl(I V ) sulphate solution ; Natulan capsules Procarbazine hydrochloride (Natulan) , an N-methylhydrazine derivative, is used for the treatment of Hodgkin’s disease.l Few papers have described methods for the determination of this anti-cancer drug. An ultraviolet absorption method2 was used for analysing both pure and dosage forms, and a coulometric titration with electrolytically generated bromine3 has been adopted. In a polarographic determination Natulan was converted into an azo deri~ative.~ Hydrazine is known to be oxidised5 to nitrogen and ammonia [equation (l)] by ammonium cerium(1V) sulphate in 0.5 N sulphuric acid.No literature is available concerning the oxidation of alkyl and arylalkyl hydrazines using this reagent. In this work, the oxidation of Natulan with cerium(1V) has been shown to produce formaldehyde and N-isopropylbenzamide-4-carboxaldehyde. The analytical method consists in the reaction of the aromatic aldehyde with 2,4-dinitrophenylhydrazine, the isolation of the product and the measurement of the absorption of the red colour (Amax. 480490 nm) produced with alcoholic potassium hydroxide solution. It has been reported6 that when 2,4-dinitrophenylhydrazones are treated with alkali, a red colour is produced, probably as a result of resonance delocalisation. Natulan in capsules can be determined satisfactorily by this method.Experimental Materials and Reagents solvents were of spectroscopic grade. Switzerland. 50 mg of ~-(2-methylhydrazinomethyl)-N-isopropylbenzamide as its hydrochloride. All chemicals and reagents used were of analytical-reagent or pharmaceutical grade and Pure firocarbazine hydrochzoride. Pharmaceutical grade, Hoff mann La-Roche, Basle, Natulan cafisules. Hoffmann La-Roche, Basle, Switzerland, lot No. 31599, containing Apparatus Visible spectrophotometer. Ultraviolet spectrophotometer. Unicam SP600, with a 1-cm cell. Pye Unicam SP6-500 with a 1-cm cell.670 HASSIB : SPECTROPHOTOMETRIC METHOD FOR DETERMINATION Analyst, Vol. 105 Methods Titrimetric determination of Natulan A 10-ml aliquot of solution containing 5-30 mg of Natulan was placed in a conical flask, 20 ml of 0.1 N' ammonium cerium(1V) sulphate solution were added and after 10 min the solution was titrated with 0.1 N ammonium iron(1I) sulphate solution using ferroin as indi- cator.1 ml of 0.1 N ammonium cerium(1V) sulphate = 4.29 mg of Natulan A blank determination was also carried out. Characterisation of the hydrazone of N-iso~ro~ylbenxamide-4-carboxaldelzyde Approximately 100 mg of Natulan, dissolved in 20 ml of water, were treated with 25 nil of 0.1 N ammonium cerium(1V) sulphate solution. After 10 min 0.1 N ammonium iron(I1) sulphate solution was added to decompose excess of the reagent. The solution was boiled for 1.5 h to remove any formaldehyde, and any water lost by evaporation was replaced. Next, 80 mg of 2,4-dinitrophenylhydrazine dissolved in 2 N sulphuric acid were added. The temperature was raised to 50 "C and after 30 min the orange crystals were removed by filtration, washed with water, dried at 60 "C and then recrystallised from ethanol.It is a fluffy, crystalline substance, orange in colour and has a melting-point of 245-246 "C. When the experiment was repeated without evaporation of formaldehyde, the same hydrazone was obtained. Elemental analysis of the hydrazone gave the following results : carbon 55.3% and hydrogen 4.4% (calculated values : 54.98% and 4.58%, respectively). The hydrazone can be recrystallised from a large volume of ethanol. Spectrophotometric determination of Natulan A solution containing 6-10 mg of Natulan in 10ml of water was placed in a 150-ml beaker and treated with 3 4 ml of 0.1 N ammonium cerium(1V) sulphate solution.The reaction was allowed to proceed for 10 min at room temperature (20-25 "C). Excess of reagent was removed by dropwise addition of 0.05 N ammonium iron(I1) sulphate solution until the yellow colour of the cerium(1V) solution had disappeared. To this solution was added, with stirring, a solution containing 14 mg of 2,4-dinitrophenylhydrazine dissolved in 2 N sulphuric acid and the temperature was then raised to 50 "C. The solution became cloudy and the orange crystals were left to settle out at room temperature for 30 min. The precipitate was collected, washed with water and allowed to dry on the filter pump or in an oven at 50 "C. An aliquot of this solution, containing 20-160 pg of Natulan, was placed in a test-tube and evaporated to dryness on a water-bath. A 5-ml volume of 0.01 N ethanolic potassium hydroxide solution was added to the residue and the solution was transferred into a 10-ml calibrated flask and diluted to volume with ethanol.The absorbance of the red solution was measured at 485nm using ethanolic potassium hydroxide solution as a blank. The Natulan content was obtained from a calibration graph that had been prepared using the described procedure. It was then dissolved in 100 ml of chloroform in a calibrated flask. Spectrophotometric determination of Natulan in Natulan capsules An accurately weighed portion of powder from the capsules, equivalent to 60-100 mg of Natulan, was shaken with water in a 100-ml calibrated flask for 30 min. The solution was diluted to the mark with water, mixed well and then filtered through a dry filter-paper into a dry flask.The spectrophotometric determination of Natulan was then carried out on 10 ml of this solution. Results and Discussion When ammonium ceriurn(1V) sulphate solution is added to Natulan, a turbid solution results, a gas is evolved and initially a formalin odour is detected, then a bitter almond odour. The liberation of formaldehyde as a result of N-methyl group oxidation is demonstrated by the characteristic violet colour produced with the chromotropic acid test. By analogy, N-isopropylbenzamide-4-carboxaldehyde is expected to be formed as a result of N-benzylJuly, 1980 OF PROCARBAZINE HYDROCHLORIDE I N CAPSULES 67 1 group oxidation. This has been proved by the formation and characterisation of the corresponding 2,4-dinitrophenylhydrazone.In contrast to the formaldehyde hydrazone, this hydrazone is insoluble in water, orange in colour and has a higher melting-point. A stoicheiometric study of the cerium(1V) oxidation of Natulan is achieved through a titrimetric determination. Five minutes are needed for the reaction to go to completion (Table-I), so a direct titration is not feasible. Excess of 0.1 N ammonium cerium(1V) sulphate solution is back-titrated with 0.1 N ammonium iron(I1) sulphate solution using ferroin as the indicator. TABLE I DURATION OF REACTION WITH 0.1 N AMMONIUM CERIUM(IV) SULPHATE Volume of 0.1 N ammonium Amount of Natulan Duration of cerium sulphate consumedtl used*/mg reactionlmin ml 9.34 9.34 9.34 9.34 1 6 15 30 1.9 2.209 2.188 2.219 * Amount of Natulan is corrected according to the ultraviolet deter- t Each reading is the average of 3 titrations.mination. The actual amount of procarbazine hydrochloride in the samples used throughout the work is determined by measuring the absorbance of an accurately weighed sample dissolved in 0.1 N hydrochloric acid at 231 nm taking 497 as A:% (ref. 2) (Table 11). TABLE I1 SPECTROPHOTOMETRIC DETERMINATION OF NATULAN Amount of Natulan used/ Absorbance mg per 100 ml a t 231 nm 1.36 0.64 1.2 0.565 1.2 0.557 1.04 0.475 0.58 0.273 Amount of Natulan recovered/ mg per 100 ml Recovery, % 1.28 94.11 1.13 94.16 1.12 93.33 0.955 91.82 0.545 93.96 Mean . . . . 93.48 Standard deviation 0.98 Standard error . . 0.43 From the results obtained in the titrimetric determination of Natulan (Table 111), a stoicheiometry of 1 mol of Natulan to 6 mol of cerium(1V) is found and the mean recovery is 99.85%.Accordingly, a hypothetical equation can be written [equation ( Z ) ] ~ taking into coiisideration that hydrazine when oxidised with ammonium cerium( IV) sulphate under the conditions of the experiment is converted into nitrogen and ammonia.5 /CH3 3CH3 NHNHCH2 ~ C O N H C \ H + 6H20 + 18Ce4+ --+ CH3 /c H3 3CH20 + 2N2+ 2NH3 + 18Ce3+ + 18H+ + 3 0 H C o C O N H c H (2) \ CH3672 HASSIB : SPECTROPHOTOMETRIC METHOD FOR DETERMINATION Analyst, VoZ. 105 TABLE I11 TITRIMETRIC DETERMINATION OF NATULAN Amount of Natulan sample weighed/mg 5 10 10 15 20 25 30 30 25 Actual amount of Amount of Natulan in the Volume of 0.1 N ammonium Natulan found/ sample*/mg cerium(1V) sulphatelml mg Recovery, % 4.674 1.079 4.629 99.03 9.348 2.178 9.344 99.95 9.348 2.198 9.429 100.86 14.02 3.237 13.887 99.05 18.695 4.356 18.687 99.95 23.369 5.445 23.359 99.95 28.043 6.623 28.413 101.31 28.043 6.504 27.902 99.49 23.369 5.396 23.149 99.05 Mean .. . . 99.85 Standard deviation 0.81 Standard error . . 0.27 * Actual amount of Natulan is corrected according to the ultraviolet determination. In the spectrophot ometric determination, the amount of 2,4-dinitrophenylhydrazine, in moles per 1 mol of Natulan, and the time needed for complete precipitation were carefully studied (Figs. 1 and 2). Because the reaction mixture contains iron(II), iron(II1) and 3 4 Amount of 2,4-dinitrophenylhdrazine/mol Time/h Fig.1. Effect of changing the concen- Fig. 2. Effect of reaction time with 2,4-dinitro- tration of 2,4-dinitrophenylhydrazine on phenylhydrazine on absorbance. absorbance. cerium( 111) ions, which precipitate their hydroxides in alkali metal hydroxide solution, the hydrazone has to be separated from the reaction mixture by filtration and washing. Chloro- form is used to dissolve the hydrazone, because of its greater solvating capacity. Ethanol is used as a medium for the final absorbance measurement, although a red colour is formed in potassium hydroxide solution. The hydrazone has a higher A,,,. in ethanolic potassium hydroxide than in aqueous potassium hydroxide solution and a higher sensitivity. Fig. 3 r O.’ 0 ~ 400 420 440 460 480 500 520 540 560 580 Wavelength/nm Fig.3. Absorption spectra of the 2,kdinitrophenyl- hydrazine derivative of N-isopropylbenzamide-4-carbox- aldehyde : A, in aqueous potassium hydroxide solution; and B, in ethanolic potassium hvdroxide snliitinnJzcly, 1980 OF PROCARBAZINE HYDROCHLORIDE IN CAPSULES 673 shows the absorption spectra of equal amounts of the hydrazone in both aqueous and ethanolic potassium hydroxide solutions, the ethanolic solution having double the sensitivity. The colour in the ethanolic solution is stable for at least 2 h and Beer’s law is valid between 20 and 160 pg per 10 ml. This spectrophotometric method is more sensitive than the ultraviolet determination ; A:% is found to be 881 compared with 497 in the ultraviolet method. Table IV shows a mean recovery of 99.79% in the spectrophotometric determination of Natulan.TABLE IV SPECTROPHOTOMETRIC DETERMINATION OF NATULAN Amount of Natulan sample weighed/ pg per 10 ml 50 31.5 63.1 40 80 160 100 140 Amount of Natulan found/ pg per 10 ml Recovery, yo 49 98 32 101.58 64.5 102.21 40 100 78.5 98.12 158 98.75 99 99 141 100.71 Mean . . .. 99.79 Standard deviation 1.59 Standard error 0.56 Capsules give satisfactory results when the conditions described for the spectrophotometric Table V shows that the mean recovery of the added determination are carefully applied. Natulan was 98.81 yo. TABLE V SPECTROPHOTOMETRIC DETERMINATION OF NATULAN IN CAPSULES Amount of Natulan used/ Pg 15.4 30.8 21.15 20.64 25.48 Amount of Natulan found/ Pg Recovery, yo 15.25 99.02 31.25 101.46 21 99.29 20.25 98.11 24.5 96.15 Mean . . .. 98.81 Standard deviation 1.93 Standard error 0.86 Conclusions Natulan was oxidised with ammonium cerium( IV) sulphate solution to N-isopropyl- benzamide-4-carboxaldehyde, which gave a water-insoluble hydrazone with 2,4-dinitro- phenylhydrazine. This derivative was used successfully in the spectrophotometric deter- mination of Natulan by treating it with ethanolic potassium hydroxide solution and measuring the absorbance of the red solution produced. The method is more sensitive and selective than the ultraviolet determination and can be applied to Natulan in capsules. References 1. 2. 3. 4. 5. 6. Bollag, W., and Theiss, E., “Chemotherapy of Cancer,” Elsevier, Amsterdam, 1964, p. 311. Hoffmann La-Roche, Basle, personal communication. Beral, H., and Stoicescu, V., Pharm. Zentralholle, 1969, 108, 469; Chem. Abstr., 1969, 71, 94808k. Johnson, J . B., and Venturella, V. S., Bull. Parenter. Drug Assoc., 1971, 25, 239. Kumar, A., and Prasad, R. K., J . Indian Chem. SOC., 1974, 11, 366. Connors, K. A., “Text Book of Pharmaceutical Analysis,” Second Edition, John Wiley, New York, Received September 9th, 1979 Accepted January 16th, 1989 1975, p. 481.

ISSN:0003-2654    

DOI:10.1039/AN9800500669

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

674 Analyst, July, 1980, Vol. 105, $9. 674-678 Spectrophotometric Determination of Zinc with Thiocyanate and Rhodamine 6G T. Prasada Rao and T. V. Ramakrishna Department of Chemistry, I?.zdian Institute of Technology, Madras 600 036, India A sensitive and selective method is described for the spectrophotonietric determination of zinc. The method is based on the reaction of Rhodamine 6G with the tetrathiocyanatozincate(I1) anion to form a pink product that is stabilised by gelatin. The colour development is rapid and remains stable for a t least 3 h. Zinc contents as low as 0.2-5.0 pg in 10 ml of sample can be readily determined. The method is precise and has been applied to the determination of zinc in synthetic matrices and soil samples. Keywords : Zinc determination ; soils ; spectrophotometry ; thiocyanate - Rhodamine 6G A number of ternary systems based on the interaction of bis(phenanthrolinium)zinc(II) ~ a t i o n l - ~ or tetrathiocyanatozincate(I1) anion4-* with dyestuffs of opposite charge have been proposed for the extraction and spectrophotometric determination of zinc.The methods, though sensitive, are inadequate for the practical analysis of samples, as the colour systems are susceptible to interferences from other elements that are frequently encountered with zinc. However, our investigations revealed that the interaction between tetrathiocyanato- zincate(I1) with Rhodamine 6G cation in orthophosphoric acid medium can be exploited for a more rapid determination of zinc in the concentration range 0.008-0.2 p.p.m., without resorting to an extraction step.A few elements interfered but this was successfully overcome by suitable conditioning of the solution prior to spectrophotometric determination. This paper reports the results of the study of this colour system and the application of the method to the determination of zinc in synthetic matrices and soil samples. Experimental Reagents (ZnS0,.7H20) in water and dilute to 250 ml. solution with water to provide a 1 p.p.m. solution of zinc as required. schaft, Schmid & Co., Stuttgart) in water and dilute to 11. potassium thiocyanate in water and dilute to 100 ml with water. London) in hot water, cool and dilute to 100 ml with water. Zinc solution, 250 p.p.m. Dissolve 0.2748 g of analytical-reagent grade zinc sulphate Dilute appropriate volumes of this stock Rhodamine 6G solution, 0.005~0, m/V.Dissolve 0.05 g of the reagent (Chroma Gesell- Potassium thiocyanate solution, 5%) m/V. Dissolve 5 g of analytical-reagent grade Gelatin solution, 1% m/V. Dissolve 1 g of laboratory-reagent grade gelatin (0x0 Ltd., Spectrophotometer A Carl Zeiss PMQ-I1 spectrophotometer with 1- and 4-cm quartz cells was used. Procedure Transfer a suitable aliquot (up to 10 ml) of the sample solution, containing not more than 5 pg of zinc, into a 25-ml calibrated flask. Add, with mixing, 2.5 ml of 10 N ortho- phosphoric acid, 2.5 ml of potassium thiocyanate solution and 5 ml of Rhodamine 6G solution, followed by 1 ml of gelatin solution. Dilute the solution to the mark with distilled water and measure the absorbance in 4-cm cells at 575 nm, against a reagent blank.Establish the concentration by reference to a calibration graph, prepared with solutions containing 1-5 pg of zinc.RAO AND RAMAKRISHNA Results and Discussion 675 The addition of Rhodamine 6G to weakly acidic solutions of thiocyanate, in the presence of zinc, produced a pink colour in contrast to the orange - red colour of the dye. The formation of the ion-association complex of zinc was instantaneous, but the reaction product separated out gradually on standing. However, the addition of gelatin retarded the precipita- tion and permitted absorbance measurements to be made directly in the aqueous medium. Absorption Spectra Fig. 1 shows the absorption spectra of Rhodamine 6G with different molar proportions of zinc in the presence of excess of potassium thiocyanate.It is evident that the interaction between tetrathiocyanatozincate(I1) and Rhodamine 6G proceeds with a considerable batho- chromic shift and that the ternary complex shows maximum absorption at 575 nm as against that of the reagent blank at 530 nm. 0.6 An E 0.4 e, C m -!? 0.2 0 400 480 560 640 Wavelengthhm Fig. 1. Absorption spectra of zinc - thiocyanate - Rhodamine 6G (1 .O N in orthophosphoric acid : total volume 25 ml; 2.5 ml of 5% potassium thiocyanate solution; 1 ml of 1% gelatin). A, 1.0 ml of 2.1 x l o - 4 ~ Rhodamine 6G, reference water, 1-cm cells; B and C, as in A, with 1 and 2 ml of 1.05 x M zinc, respectively; D and E, spectra of B and C, respec- tively, using A as reference and 4-cm cells.Effect of pH/Acidity It was found that the ternary complex was fully formed when solutions were buffered in the pH range 1-5. Subsequent studies at higher acidities revealed that the colour system remained unaffected up to 6 N in orthophosphoric acid as against 4 N and 0.5 N with respect to sulphuric acid and hydrochloric acid, respectively. The results obtained are shown in Fig. 2. The orthophosphoric acid medium was considered beneficial in improving the selectivity of the method and therefore in all subsequent studies the solutions were adjusted to 1 N with respect to this acid. Effect of Reagent Concentrations The results of the investigation on the effect of thiocyanate and Rhodamine 6G concentra- tions showed that for constant maximum absorbance the solution should contain at least 4 ml of 2.5% potassium thiocyanate solution and 4.5 ml of 0.005~0 Rhodamine 6G solution. The fact that the reaction goes to completion only in the presence of excess of thiocyanate indicates the highly dissociative nature of the primary anionic complex formed between zinc and thiocyanate.It was also established that the addition of 0.75 ml of 1% gelatin solution was sufficient to prevent the precipitation of the coloured complex. Using optimum amounts of reagent solutions it was established that the order of addition was not critical provided that the gelatin solution was added after the addition of other676 RAO AND RAMAKRISHNA : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 105 reagents. Under these conditions, the colour development was instantaneous and no measurable change in the absorbance was noticed for 3 h.Further standing caused a slight decrease in the aubsorbance. O.* ~ 6 4 2 1 3 5 7 9 PH Acidity/N Fig. 2. Effect of pH/acidity on the formation of zinc - thiocyanate - Rhodamine 6G complex; 3.0 p g of zinc, 2 ml of 5% potassium thiocyanate solution, 6 ml of 0.005~0 Rhodamine 6G, 1 ml of gelatin, total volume 25 ml; 575 nm, 4-cm cells. HC1; 0, H,SO,; and A, H,PO,. Calibration Graph and Precision The calibration graph obtained by the recommended method was linear up to 5 p g of zinc and passed through the origin. The slope was 0.10 absorbance unit for 0.02 p.p.m. of zinc when measured in 4-cm cells. The molar absorptivity was calculated to be 8.17 x 104 1 mol-1 cm-l. The precision was established from the results of 15 determinations of a standard solution containing 3pg of zinc.The mean absorbance was found to be 0.60 with a standard deviation of 0.02 absorbance unit and a relative standard deviation of 3%. TABLE I INTERFERENCE STUDIES Interferents" Ca2+, Sr2+, Ba2+, Mg2+, Pb2+, Be2+, Mn2+, Ni2+, 2+, Cr3+, Al3+, Bi3+, La3+, Te4+, Sb5+, AS043-, UO, h0,3-, NO,-, SO,2-, SO,,-, s2032-, F-, Cl-, Br-, B40,2-, citrate, oxalate and tartrate Cu2+ (nil),. Hg2+ (3), Cd2+ (loo), Co2+ (nil), Pd2+ (nil), Fe3+ (75), Pt4+ (6), WOd2- (75) and v 0 d 3 - (300) (50), Zr4+ and Th4+ NO,- Remarks No interference Interfered by enhancing the absorbance Interfered by precipitation Interfered by oxidising Rhodamine 6G * Figures in parentheses give the tolerance limits in micrograms in the absence of masking agents.Stoicheiometry of the Complex and continuous-variation methods. 1:2. The ratio of zinc to Rhodamine 6G in the complex was established by both molar-ratio The results indicated a zinc to Rhodamine 6G ratio of These methods were unsuccessful when determining the combining ratio of zinc andJuly, 1980 OF ZINC WITH THIOCYANATE AND RHODAMINE 6G 677 thiocyanate, as no colour development took place when zinc and potassium thiocyanate were present in molar proportions. However, the equilibrium-shift method indicated the existence of a 1:4 complex, corresponding to the formation of the anionic complex species [Zn(SCN),I2-. From these results, it was concluded that the ion pair formed in the reaction was R,[Zn(SCN) J, where R represents the Rhodamine 6G cation.Interference Studies A systematic study of the influence of l-mg amounts of several anions and cations on the determination of 3 pg of zinc was carried out. Attempts to overcome the interference due to milligram amounts of various ions revealed that the effect of Cu2+, Cd2+, Hg2+, Pd2+ and Pt4+ can be eliminated by the addition of 2 ml of 5% thiourea solution. Addition of 2 ml of a 5% solution of fluoride masked the interfering effect of Zr4+, Th4+ and WO,,-. Interference due to NO2- was similarly overcome by the addition of urea. Although Fe3+ up to 1 mg did not interfere in the presence of 3 ml of a 0.1 M solution of EDTA, it was found that its influence at higher levels can be conveniently eliminated by precipitating as iron(II1) hydroxide at pH 4.0 and subjecting the filtrate to the determina- tion of zinc, It was also found that the addition of milligram amounts of Fe3+ to the sample solution, followed by its precipitation, effectively eliminated the interference of milligram amounts of Molybdenum and vanadium were found to collect readily and quantitatively on iron(II1) hydroxide when the latter is precipitated at pH 4.0 in the presence of 5 ml of a~etone.~ In this instance, the determination of zinc in the filtrate was completed after expelling the acetone by boiling, as the presence of acetone inhibited the ion-pair formation.The results are summarised in Table I. and vo43-. Only cobalt, therefore, interfered seriously in the determination. Analysis of Synthetic Samples Table I1 presents the results of the analyses of synthetic sample solutions of zinc, with various matrices, after overcoming the influence of the interferents, as described above.The results clearly show that the method can find use in the trace determination of zinc in real samples, such as sea water, chromium - nickel steels and manganese steels. TABLE I1 ANALYSIS OF SYNTHETIC SAMPLES The samples had a zinc concentration of 3.0 pg in 25 ml. Other constituents Absorbance None . . .. .. . . . . . . . . 0.60 NaCl (1 g) . . . . .. . . . . 0.60 Fe (5mg)* < Cu (1'mg)t . . . . . . . . 0.62 Fe (5 mg)* + Cr (3 mg) + Ni (2 mg) . . . . . . 0.60 Fe (5 mg)* + Mn (1 mgj . . .. . . . . . . 0.60 Fe (1 mg) + Mo (1 mg), . . . . . . . . . . 0.58 Sb (1 mg) + Sn (1 mg) .. . . . . . . . . 0.58 Cd (1 mg)t + Cu (1 mg)t + Pb (1 mg) . . .. . . 0.62 Cd (1 mg)t + Sn (I mg) + Pb (1 mg) . . . . . . 0.61 * By precipitating iron(I1f) as hydroxide at pH 4.0. t In the presence of 2 ml of 5% thiourea solution. $ By precipitating iron(II1) as hydroxide at pH 4.0 in the presence of acetone. Analysis of Soil Samples Table I11 shows the iesults for the determination of nitric acid soluble zinc and acetic acid and EDTA extractable zinc in two soil samples. The recoveries of zinc added to the soil samples and the results obtained by atomic-absorption spectrophotometry are also presented. To determine nitric acid soluble zinc, 0.5 g of finely ground soil sample, dried at 110 "C, was treated with 5 ml of concentrated nitric acid and evaporated to dryness.The residue was decomposed by heating with 50 ml of 1 + 1 sulphuric acid to fumes of sulphur trioxide.678 RAO AND RAMAKRISHNA TABLE I11 ANALYSIS OF SOIL SAMPLES Zinc determined Sample Nitric acid soluble zinc* . . Sample 1 Sample 2 Acetic acid extractable zinct . . Sample 1 Sample 2 EDTA extractable zinct . . Sample 1 Sample 2 Zinc added/ pg ml-l 0.4 1.2 2.0 0.4 1.0 2.0 0.2 1.0 0.2 1.0 0.2 1.0 0.2 1.0 - - - - - - Aliquot taken/ ml 5.0 2.5 2.0 2.0 5.0 5.0 3.0 2.0 4.0 4.0 2.0 8.0 8.0 4.0 2.0 2.0 1.0 8.0 8.0 2.0 Zinc foundlyg ml-1 CAtomic-7 Proposed absorption Recovery, 0.50 0.50 - 0.88 96 1.75 104 2.50 100 0.30 0.28 - 0.68 100 1.30 102 2.30 101 1.00 0.98 - 1.19 105 2.00 102 0.100 0.100 - 0.305 103 1.090 99 1.85 1.90 - 2.10 105 2.90 101 0.205 0.220 - 0.420 100 1.240 102 method method % * 0.5 g of soil in 250 ml of water.t 5.0 g of soil in 50 ml of water. The cooled sample was then boiled with 100 ml of water and filtered hot to separate the silica. The residue was washed with water and the filtrate and washings were made up to 250 ml with water. Suitable aliquots of this solution were taken and the concentration of zinc was determined in the presence of 1 ml each of 10% thoiurea solution and 0.1 M EDTA soh t ion. To determine extractable zinc, the soil sample was leached, as described elsewhere,lO by mixing 5 g of dried finely ground soil with 50 ml of 2y0 acetic acid (pH 2.5) or 50 ml of a 0.02 M solution of EDTA. After standing overnight, the residue was filtered off by suction through a Whatman No. 42 filter-paper and washed thoroughly using small portions of the appropriate extraction solution.The combined filtrate and washings were evaporated to a small volume and then made up to 50 ml with water. Aliquots of this solution were used for the determination of zinc in the presence of 1 ml each of 10% thiourea solution and 0.1 M EDTA solution. The results in Table I11 clearly show that the proposed method is reliable and that other ions that are extracted with zinc have no effect on the determination of zinc by the proposed procedure. One of us (T.P.R.) is grateful to CSIR, New Delhi for financial assistance. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Tananaiko, M. M., and Bilenko, N. S., Izv. Vyssh. Ucheb. Zaved. Khim. Khinz. Tekhnol., 1972, 15, Tsubouchi, M., Bunseki Kagaku, 1970, 19, 1400. Tananaiko, M. M., and Bilenko, N. S., Zh. Anal. Khim., 1975, 30, 689. Kish, P. P., Zimomorya. I. I., and Zoltov, Yu. A., Zh. Anal. Khim., 1973, 28, 252. Pilipenko, A. T., Kish, P. P., and Zimomorya, I. I., Ukr. Khinz. Zh., 1971, 37, 186. Slovak, Z,, and Pribyl, M., Collect. Czech. Chem. Commun., 1966, 31, 1742. Babko, A. K., and Chalaya, Z. I., Zh. Anal. Khim., 1961, 16, 268. Babko, A. K., and Chalaya, 2. I., Zh. Anal. Khim., 1962, 17, 286. Prasada Rao, T., and Ramakrishna, T. V., Bull. Chem. SOC. Jpn, in the press. Bear, F. E., “Chemistry of the Soils,” IBH, Oxford, 1964, pp. 345 and 493. 1643; Ref. Zh. Khim., 1973, 19GD, 7G64. Received November 26th, 1979 Accepted January loth, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500674

出版商:RSC    

年代:1980

数据来源: RSC