摘要:

952 Analyst Augast 1983 Vol. 108 $9. 952-958 Extraction and Spectrophotometric Determination of Titanium(1V) with N-Phenyllaurohydroxamic Acid and Phenylfluorone H. Dasaratha Gunawardhana" Trace Analysis Research Centre Chemistry Defiartment Dalhousie University Halifax Nova Scotia B3H 4 J1, Canada N-Phenyllaurohydroxamic acid reacts with titanium(1V) in 9-10 M hydrochloric acid to give a complex that is completely extractable into solvents such as hexane and chloroform. The chloroform extract of the titanium complex on second extraction from a dilute hydrochloric acid medium (0.1-0.5 M) in the presence of phenylfluorone and isoamyl alcohol, forms an intensely coloured complex possessing an absorption maximum at 540 nm. Even though the molar absorptivity of the complex under optimum conditions a t 540 nm is 2.33 x 105 1 mol-l cm-l the measurements are more precise at 560 nm with a molar absorptivity of 1.23 x lo5 1 mol-l cm-l.The system obeys Beer's law for up to 0.4 p.p.m. of titanium(1V). Considerable amounts of many cations and anions including a 350-fold molar excess of iron(II1) can be tolerated. Interference from zirconium(1V) can be mitigated or even eliminated by the addition of fluoride ions. The method can be applied to the determination of titanium present a t 10 parts per log. Keywords Titanium ( I V ) determination ; spectrophotometry ; N-phenyllauro-hydroxamic acid ; PhenylfEuorone ; liquid - liquid extraction The development of highly sensitive and selective spectrophotometric methods for titanium is essential because it is difficult to obtain a high sensitivity1 using flame atomic-absorption spectrometry owing to the formation of a highly stable oxide species in the flame and atomic-absorption spectrometry with electrothermal atomisation generally involves expensive instru-mentation not available in all laboratories.N-Phenylbenzohydroxamic acid and its analogues2 are used extensively as reagents for the spectrophotometric determination of titanium. Several worker^^-^ have studied the enhance-ment of sensitivity by the addition of thiocyanate. However the sensitivities are still not high enough for the determination of titanium present in the 10-100 p.p.b. (parts per log) range. Pilipenko and co-workers6-* reported the use of phenylfluorone (9-phenyl-2,6,7-tri-hydroxy-3H-xanthen-3-one) (PF) and an alcohol in conjunction with N-phenylbenzohydrox-amic acid and with N-phenylfurohydroxamic acid to achieve enhanced intensity of the colour The maximum sensitivity so far achieved corresponds to a molar absorptivity of 9.0 x lo4 1 mol-l cm-l but there is noticeable interference from a four-fold molar excess of iron(III),' which commonly occurs in conjunction with titanium.N-Phenyllaurohydroxamic acid (phenyldodecanohydroxamic acid) (PLHA) reacts with titanium (IV) and with the addition of PF forms an intensely coloured complex. This paper reports a detailed investiga-tion of the conditions affecting the formation and extraction of the titanium(1V) - PLHA - PF complex and the application of the system to the determination of trace amounts of titanium with virtually no interference from iron.Experimental Apparatus A Pye Unicam SP 8000 recording ultraviolet - visible spectrophotometer was used for absorptiometric measurements. A Fisher Accumet Model 320 pH meter fitted with a glass and Hg - Hg,Cl electrode assembly and a Varian Model 475 atomic-absorption spectro-photometer equipped with Perkin-Elmer HGA-22000 graphite furnace were also used. * Present address Centre for Analytical Research and Development Department of Chemistry Uni-versity of Colombo P.O. Box 1490 Colombo-3. Sri Lanka GUNAWARDHANA 953 Reagents Analytical-reagent grade chemicals were used throughout. All solvents were purified and distilled prior to use. PLHA was prepared from BDH lauric acid using a previous procedurelo applied to a mixture of fatty acids.However the final product was recrystallised from an ethanol - water mixture and its purity was checked by a melting-point determination,2 elemental analysis and thin-layer chromatography on silica gel with chloroform - hexane (1 + 1) as the mobile phase. Analytical-reagent grade phenylfluorone (Eastman) was used without further purification. A stock solution of titanium(1V) was prepared by fusing high-purity titanium(1V) oxide with potassium hydrogen sulphite. After completion of the fusion the residue was dissolved in 5 cm3 of concentrated sulphuric acid and water and finally diluted to 1 1 with water. The final solution of approximately 100 p.p.m. was standardised by an EDTA back-titration procedures with bismuth and xylenol orange indicator in the presence of peroxide.Extraction Procedure The following procedure was used to obtain the extraction graphs in Fig. 1. A titanium(1V) solution of appropriate concentration (9.0 cm3) in hydrochloric perchloric or sulphuric acid, was mixed with 1.0 cm3 of water or 0.1 M sodium chloride solution. This solution was equili-brated for 3-5 min with 10.0 cm3 of 0.01-0.05~0 m/V PLHA in hexane. The concentration of titanium in the aqueous phase was determined by atomic-absorption spectrometry with electrothermal at0misation.l A similar procedure was used for iron(III) and the concentra-tion of iron in the aqueous phase after the extraction was determined by flame atomic-absorption spectrometry. In order to avoid complications no buffers were used at moderate and high pH values.Recommended Procedure for Spectrophotometric Determination of Titanium( IV) First extraction Transfer an aliquot of sample solution containing up to 0.4 pg of titanium(1V) into a stop-pered bottle or a separating funnel and dilute to 10.0 cm3 with concentrated hydrochloric acid, so that the final concentration of hydrochloric acid is 9-10 M. Add 10.0 cm3 of 0.01-0.05~0 m/V PLHA in chloroform and equilibrate for 3-5 min. Second extraction Withdraw 5.0 cm3 of the organic phase from the first extraction and transfer it into a separat-ing funnel. Add 2.0 cm3 of about 0.025% m/V PF in isoamyl alcohol then add 5.0 cm9 of 0.1-0.5 M hydrochloric acid and equilibrate for 1 min. Measure the absorbance of the organic phase a t 560 nm against a reagent blank.The solution of PF in isoamyl alcohol was prepared by mixing 1.0 cm3 of 0.3% m/V PF in dimethylformamide with 10.0 cm3 of isoamyl alcohol. In this way it was possible to achieve a greater solubility of PF in isoamyl alcohol. This preparation was done daily. The pKa value of PLHA in 50% V/V aqueous ethanol was determined spectro-photometrically as 10.35. Results and Discussion Extraction of Titanium( IV) and Iron(II1) The liquid - liquid extraction behaviour of titanium(1V) and iron(1II) with PLHA in hexane is shown as the percentage extracted veysaus pH and molarity of hydrochloric acid in Fig. 1. Both titanium(1V) and iron(II1) show the theoretically expected sigmoidal curve+ when the extraction is carried out at moderate acidities and constant ionic strength (p = 0.1).The decreasing extraction at pH 8 [Fig. l(A)] can be attributed to the ease of hydrolysis of titanium(1V). Nevertheless the extraction in hydrochloric acid medium results in an increase in extraction reaching complete extraction when the molarity of hydrochloric acid is >9. The unique behaviour of titanium is attributed to the formation of a species con-sisting of titanium(IV) PLHA and chloride ions.1° The maximum percentage extraction o 954 GUNAWARDHANA EXTRACTION AND SPECTROPHOTOMETRY OF TI(IV) Analyst VoZ. 108 90 8 =- 70 2 fi 50 0) + Q 3 CI 30 a 10 i 10.0 3.0- 1.0 1.0 3.0 5.0 7.0 9.0 11.0 HCVM PH Fig. 1. Extraction graphs for (A) titanium and (B) iron showing distribution as a function of acidity for PLHA in hexane as extractant; p = 0.1 with sodium chloride only in the pH scale.titanium achieved in the presence of sulphuric acid medium was 12%. The inability of sulphuric acid medium to exhibit a complete extraction is evidence for the presence of chloride ions in the extracted species. The extraction graphs in Fig. 1 illustrate further the ability of PLHA to extract titanium selectively from concentrated hydrochloric acid media. Absorption Spectra and the Composition of the Extracted Species Ti(1V) - PLHA -PF The intensity of the colour of the extracted species Ti(1V) - PLHA - PF is low in hexane and, therefore chloroform was used for spectrophotometric studies. Job’s method of continuous variation was applied to Ti(1V) - PLHA by keeping the concentration of PF in an excess.The Job plot [Fig. 2(A)] obtained in this manner suggests the formation of a 1 2 complex between titanium(1V) and PLHA.12 The second Job plot [Fig. 2(B)] was obtained by keeping the stoicheiometric ratio of titanium(1V) to PLHA as 1 2 which reveals the stoicheiometry of 0.3 0.1 0’ 0.1 0.3 0.5 0.7 0.9 [Ml/([MI + [Ll) Fig. 2. Continuous variation plot of (A) Ti(IV) -PLHA; (B) [Ti(IV) 2 PLHA] - PF. Total concentration, 0.1 pmol per 5 cms; and solvent chloroform - isoamyl alcohol August 1983 WITH N-PHENYLLAUROHYDROXAMIC ACID AND PHENYLFLUORONE 955 the complex as Ti(1V) PLHA PF as 1 2 1. Greater curvature of graph (A) compared with graph (B) reveals a stronger bonding between titanium(1V) and PF. Absorption spectra of the pink - violet chloroform - isoamyl alcohol extract of titanium(1V) - PLHA - PF complex [Fig.3(A)] show a prominent peak at 540 nm against a pure chloroform blank. The spectrum of the reagent blank (PLHA - PF) possesses two bands at 462 and 492 nm [Fig. 3(B)] caused mainly by PF because PLHA does not absorb in the visible region. A slight change in the shape of the spectrum of the complex leading to a broad band in the region 550-610 nm was observed when the concentration of PF was raised above the stoicheio-metric value. Two peaks at 462 and 492 nm in the spectrum [Fig. 3(C)] can be attributed to the presence of an excess of PF in the medium. When the concentrations of the two ligands remain at the stoicheiometric value the intensity of the colour (molar absorptivity €540 = 1.5 x lo4 1 mol-1 cm-l) is too low for analytical purposes.Maximum intensity of the colour ( E ~ ~ = 2.33 x 105 1 mol-1 cm-1) was achieved when PLHA and PF are present in 130-fold and 15-fold molar excesses respectively with respect to titanium. The absorption spectra of the complex Ti(1V) - PLHA - PF under the maximum intensity conditions were also characterised by a prominent peak at 540 nm against a reagent blank [Fig. 4(A)]. However this peak gradually disappears with time leaving behind a hump at 540 nm which leads to the reappearance of another band in the region 510-520 nm [Fig. 4(B) and (C)]. The intensity of the second band increases with time giving an end absorption, comparable to the absorption spectra of the complex against the pure solvent [Fig. 4(D)].However on observation after 12 h of extraction the spectra of the same complex against a freshly prepared reagent blank revealed a peak at 540 nm identical with that in Fig. 4(A). This suggests that the complex remains stable even after 12 h while the reagent blank 0.3 E e 2 0.2 n a 0 0.1 350 450 550 650 Wavelengthhm Fig. 3. Absorption spectra against pure solvent blank. (A) Ti(1V) -PLHA - PF complex Ti = 20 p ~ , PLHA = 40 p~ and PF = 20 p ~ ; (B) PLHA - PF complex PLHA = 40 p~ and PF = 20 p ~ ; (C) Ti(1V) - PLHA -PF complex Ti = 16 p ~ PLHA = 32 p~ and PF = 24 p ~ ; (D) Ti(1V) -PLHA complex in hexane Ti = 90 p~ and PLHA = 410 p ~ . 1 .o 0.8 8 0.6 0 e 2 a 0.4 a 0.2 500 550 600 Wavelengthhm Fig. 4. Absorption spectra of Ti(1V) - PLHA - PF complex extracted into chloroform -isoamyl alcohol (5 + 2) under the maximum intensity conditions.Ti = 0.25 p.p.m. PLHA = 0.01% m/V and PF = 0.008% m/V. (A) Within 3 min of extraction against a reagent blank; (B) after 60 min of extrac-tion against a reagent blank; (C) after 90 min of extraction, against a reagent blank; (D) spectra of the complex against a solvent blank; and (E) reagent blank veYsz4.s pure solvent 956 GUNAWARDHANA EXTRACTION AND SPECTROPHOTOMETRY OF TI(IV) Analyst Vo,?. 108 deteriorates with time. The high absorption of the reagent blank [Fig. 4(E)] at 540nm is attributed to this behaviour. It has been observed that the branch of the spectrum corresponding to the region 560-600 nm remains unaltered with respect to the changes in parameters such as time different reagent blanks and the volume of isoamyl alcohol within a reasonable range.A significant change in absorbance of this portion of the spectrum was observed only with changing concentration of titanium. Therefore 560nm can be considered as the most suitable wavelength for analytical application although €560 = 1.23 x lo6 1 mol-l cm-l which is lower than that at 540 nm. Effect of Isoamyl Alcohol Ethanol and Methanol By changing the concentration of isoamyl alcohol in the chloroform extracts used for the second extraction over the range 2040y0 V / V it was observed that the absorbance at 560 nm remained constant. A decrease in absorbance was observed above 40% probably due to the increasing volume of the organic phase.Below 20% incomplete solubility of the complex causes a decrease in absorbance. The value of the absorbance remained unchanged on substituting 30% of absolute ethanol for isoamyl alcohol; however the same percentage of methanol gave a cloudy organic phase that was not suitable for absorptiometric studies even after prolonged centrifuging. Effect of the Concentration of Reagents (PLHA and PF) Absorbances at 560 nm remain constant when the concentration of PLHA in chloroform is maintained between 0.02 and 0.05y0 m/V. This corresponds to a molar excess of PLHA in the range 130-330-fold. Higher and lower PLHA concentrations lead to a decrease in absorbance. Very high concentrations of PLHA are undesirable probably owing to the increased interaction between PLHA and PF resulting in a decrease in the interaction with titanium(1V).At low PLHA concentrations a decrease in absorbance was observed probably due to the incomplete extraction of titanium(1V). It has been reported8 that with N-phenylfurohydroxamic acid (PFHA) a 720-fold molar excess is required for complete complex formation. In order to ensure complete fixing of titanium as a complex only a 130-fold molar excess of PLHA is required. This reveals that the Ti(1V) - PLHA bond is more stable than that of Ti(1V) - PFHA. The absorbance at 560 nm remained unaltered on changing the concentration of PF from a 15-fold to a 60-fold excess. A decrease in absorbance at lower concentrations and an increase in absorbance at higher concentrations have been observed. Even though higher concentra-tions of PF (above 60-fold) gave higher absorbances against reagent blanks an instability of the organic phase producing cloudiness was observed within a short time.This led to a precipitate after 60 min. This can be accounted for by the further formation of Ti - PF complex which in general is not soluble in the aqueous phase and also is not extracted by the chloroform - isoamyl alcohol organic phase.8 The solution of PF in dimethylformamide was stable and remained unaffected 2 weeks later. Effect of the Concentration of Hydrochloric Acid in the Second Extraction The concentration of hydrochloric acid was varied from 0.01 to 4.0 M and the concentration was determined by titrating with a standard solution of sodium carbonate (0.05 M) using methyl orange as indicator.It was observed that the Ti-PLHA -PF complex can be extracted over a wide hydrochloric acid concentration range and the absorbance at 560 nm remains virtually constant over the acidity range extending from 0.1 to 1.0 M. Even though the absorbances at 560 nm are higher with a lower concentration of hydrochloric acid and also with water a rapid change in the shape of the spectrum accompanied by a rapid deterioration of the reagent blank as well as the complex has been observed. Higher concentrations of hydrochloric acid ( 0 . 5 ~ ) retard the deterioration and the shape of the spectrum was un-affected even after 12 h. With increasing concentration of hydrochloric acid beyond 1 .O M the organic phase became more yellowish in appearance a change that was accompanied by a decrease in absorbance at 560 nm against a reagent blank.A breakdown of the Ti(1V) - PLHA - PF complex wa Aagast 1983 WITH N-PHENYLLAUROHYDROXAMIC ACID AND PHENYLFLUORONE 957 observed at a 4 M concentration of hydrochloric acid giving a yellow organic phase possessing a zero absorbance at 560 nm against a reagent blank. The absorption spectrum of this yellow organic phase against a solvent blank was identical with that of Fig. 4(E) indicating the non-existence of the titanium complex under these conditions. Verification of Beer’s Law The system obeyed Beer’s law at 560 nm up to 0.4 p.p.m. of titanium. A negative deviation has been observed at higher concentrations. The minimum detection limit lies in the region of 10 p.p.b. of titanium.The molar absorptivity at 560 nm was 1.23 x lo5 lmol-l cm-l. A plot of Beer’s law was a highly precise reproducible straight line calibration graph that passed through the origin. Effect of Diverse Ions The system can tolerate a 350-fold molar excess of iron(III) a 4-fold excess of vanadium(V), a &fold excess of molybdenum(VI) a 3-fold excess of tungsten(V1) and a 200-fold excess of fluoride. The inter-ference from zirconium is significant because of the co-occurrence of zirconium in many titanium-bearing minerals. The presence of zirconium produced cloudiness initially in the organic phase but with a subsequent precipitation of a red complex. The attempted applica-tion of an excess of PLHA ascorbic acid hydroxylammonium chloride and tin(I1) chloride in both the first and the second extractions in order to eliminate the interference from zirconium, became futile.As the system can tolerate a 200-fold molar excess of fluoride the ability of zirconium(1V) to form stable complexes13 with fluoride more readily than titanium was taken into consideration. It was observed that in the presence of a 200-fold molar excess of fluoride (approximately 100 p.p.m.) the system was capable of tolerating a 50-fold molar excess of zirconium without affecting the over-all sensitivity appreciably. In fact this is almost 100 times the amount of zirconium that would otherwise have been tolerated by the system without any fluoride ions. The tolerance limit for both niobium(V) and zirconium(1V) is 0.5-fold. Application of the Method The proposed spectrophotometric method was applied to the determination of titanium in a rock sample.About 0.1000 g of accurately weighed finely powdered rock sample (SGR-1) was fused with potassium hydrogen sulphate(5 g) and diluted to 250.0 cm3 with concentrated hydrochloric acid and water so as to maintain the final acid concentration of about 10 M with respect to hydrochloric acid. The results obtained were comparable to those obtained by neutron activation analysis and atomic-absorption spectrometry with electrothermal atomisa-tion using the standard additions method1* to eliminate matrix interferences. The percent-ages of titanium determined were as follows neutron activation analysis 0.15 and 0.14; atomic-absorption spectrometry 0.15; and this method 0.14 & 0.006 (G = 3).Highly sensitive and selective spectrophotometric methods of this nature will find many advantages in the determination of unbound or “free” concentrations of titanium(1V) ,16 provided that the concentration of bound species of titanium is unaffected by 10 M hydro-chloric acid and PLHA in chloroform. The author thanks Professor D. E. Ryan Director TARC Dalhousie University for kind advice and encouragement. Thanks are also extended to the Canadian International Develop-ment Agency (CIDA) for the financial assistance to the collaborative programme between TARC Dalhousie University and the Centre for Analytical Research and Development (CARD) University of Colombo Sri Lanka. References 1. 2. 3. 4. Studnicki M. Anal. Chern. 1980 52 1762.Majumdar A. K. “N-Benzoylphenylhydroxylamine and its Analogues,’’ Pergamon Press Oxford, Ni Che-ming and Liang Shu-chuan Sci. Sin. 1963 12 615. Afghan B. K. Marryatt R. G. and Ryan D. E. Anal. Chiwa. Acta 1968 41 131. 1972 pp. 122-157 958 GUNAWARDHANA Savariar C. P. and Joseph J. Anal. Chim. Acta 1969 47 347. Pilipenko A. T. Shpak E. A, and Zul’figarov 0. S. Zh. Anal. Khim. 1974 29 1074. Pilipenko A. T. Shpak E. A. and Zul’figarov 0. S. Zh. Anal. Khim. 1975 30 1009. Pilipenko A. T. Shpak E. A. and Eremenko M. V. Zh. Anal. Khim. 1975 30 1535. West T. S. “Complexometry with EDTA and Related Reagents,” Third Edition BDH Chemicals, Gunawardhana H. D. and Asinvatham D. Indian J . Chem. 1982 21 338. Stary J. “The Solvent Extraction of Metal Chelates,” Pergamon Press Oxford 1964 p. 22. Abbasi S. A. Int. J . Environ. Anal. Chem. 1982 11 1. Sharp D. W. A. Editor “Transition Metals-Part 1,” MTP International Reviews of Science, Elwell W. T. and Gidley J. A. F. “Atomic Absorption Spectrophotometry,” Second Edition, Robert F. B. and Kratochvil B. Anal. Chem. 1980 52 546. Poole 1969 p. 210. Volume 5 Butterworths London 1972 p. 285. Pergamon Press Oxford 1966 p. 74. Received January 14th. 1983 Accepted February 23rd 1983 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15

ISSN:0003-2654    

DOI:10.1039/AN9830800952

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst August 1983 Vol. 108 pp. 959-965 959 Spectrophotometric Determination of Rhodium( 111) in Aqueous and Alcoholic Media Using 2-Thiobarbituric Acid Basilio Morelli Universitb degli Studi di Bari Dipartimento di Chimica Via Amendola 173 70126-Bari Italy A spectrophotometric study of the rhodium(II1) - 2-thiobarbituric acid system is presented. Rhodium(II1) forms 1 2 and 1 4 complexes with 2-thio-barbituric acid in water and 98% ethanol respectively. Conformity to Beer's law at 327 nm was observed for up to 14 pg ml-l of rhodium in water and for up to 12 pg ml-1 in 98% ethanol and the detection limits were 0.24 and 0.34 pg ml-l of rhodium respectively. Molar absorptivities a t 327 nm in aqueous and alcoholic media were 1.04 x 104 and 1.08 x lo4 1 mol-l cm-l and Sandell's sensitivities were 0.009 8 and 0.009 5 pg cm-2 respectively.The tolerance of the system to platinum metals and other common cations is reported the method is evaluated and a comparison with the main colorimetric methods for rhodium determination developed in recent years is presented, Keywords Rhodium determination ; 2-thiobarbiturac acid ; sfiectrophotometry Recently barbituric and thiobarbituric acids and a large number of their derivatives have been frequently employed in pharmacological appli~ationsl-~ but their use as analytical reagents has been On investigating the analytical capabilities of 2-thiobarbituric acid (TBA) we observed that it can be a useful spectrophotometric In this paper a spectrophotometric study on the rhodium(II1) - TBA system is described and the behaviour of the system is compared in both aqueous and alcoholic media.The method has been found to be sensitive and accurate for the determination of rhodium in both media and Beer's law was obeyed over the whole concentration range tested. The molar ratio of rhodium to TBA has been demonstrated by the molar ratio method Sandell's sensitivities were determined and the effects of temperature, heating time and acidity and the tolerance to platinum metals and other foreign ions have been investigated . Experimental and Results Reagents All reagents were of analytical-reagent grade. Rhodium(III) ' standard solution. M. 2-Thiobarbituric acid solution 3 x M. Prepared in the usual ~ a y . ~ - ~ Foreign ion solutions. These contained 2-10 mg ml-l of the ion.8 Buer 0.2 M acetic acid - 0.2 M sodium acetate solution.Stock solutions of rhodium(II1) were prepared by dis-The working solving rhodium(II1) chloride trihydrate in water or in absolute ethanol. rhodium(II1) solutions were about 3.6 x Apparatus previous The apparatus for absorbance and pH measurements was the same as that described in Effect of Heating on Development of Colour A preliminary investigation was carried out to determine the effect of heating on the forma-tion of the complex in both aqueous and alcoholic media. The samples were placed in a water-bath at 35 50 65 and 75 "C for periods of 10 20 35 50 55 and 60 min. Maximum absorbance was obtained after heating for about 50 min at 75 "C for both media while ap-proximately 98% (in water) and 97% (in 98% ethanol) of the colour intensity was develope 960 MORELLI SPECTROPHOTOMETRY OF RH(III) IN AQUEOUS Analyst VoZ.108 after heating for 35 min at 75 "C. A development time of 60 min at 75 "C was adopted in subsequent investigations to ensure complete complexation. At lower temperature the rate of colour development becomes progressively slower ; there is no appreciable formation of the coloured complex at room temperature. Colour-developed solutions measured after different heating times and at different tempera-tures showed no change in the maximum absorption wavelength. Fig. 1 (a) (aqueous medium, pH 4.7) and Fig. l ( b ) (alcoholic medium pH 2.7) show the increase of absorbance with time of heating at different temperatures it is evident that a constant absorbance value is achieved after the samples have been standing at 75 "C for no longer than 50 min.The graphs in Fig. l ( a ) and l ( b ) were obtained with samples containing 44 and 47 pg of rhodium per 5 ml of solution respectively as described under Procedure. Om2 O-11 /I---- 500c 35°C 1 I I I 10 30 50 Tim e/m i n Fig. 1. Effect of heating time and temperature on absorbance. (a) Aqueous medium 44 pg of Rh(II1) per 5 ml and pH 4.7; (b) 98% ethanol medium 47 pg of Rh(II1) per 5 ml and pH 2.7. Reference reagent blank 327 nm. Effect of pH In order to determine the effect of acidity on the formation of the Rh(II1) - TBA complex, the usual procedure was followed.8 The results obtained in aqueous medium are different from those obtained in 98% ethanol.Typical graphs of absorbance versws pH are shown in Fig. 2(a) (water) and 2(b) (98% ethanol). In an aqueous medium the absorbance was found to be maximum and constant from pH 4.4 to 6.4; in an alcoholic medium it was maximum and constant up to an apparent pH of 3.8 (it was not possible to make measurements of absorbance in ethanol a t pH > 5 owing to opalescenc August 1983 AND ALCOHOLIC MEDIA USING 2-THIOBARBITURIC ACID 961 2 4 6 DH 8 Fig. 2. Effect of acidity on absorbance. (A) Aqueous medium 40 pg of Rh(II1) per 5 ml; (B) 98% ethanol medium 40 pg of Rh(II1) per 5 ml. Reference reagent blank 327 nm. of a finely divided precipitate which appeared in solution a few minutes after sample prepara-tion). For these reasons all further measurements were made in the optimum pH range i e ., at pH 4.7 (by means of acetic acid - acetate buffer solution) in water and at pH < 3.8 in 98% ethanol. Procedure The procedure for the determination of rhodium(II1) using TBA was as follows to a 2-ml aliquot of 3.0 x 1 0 - 2 ~ TBA aqueous or alcoholic solution in a 5-ml calibrated flask were added 2.5 ml of acetic acid - acetate buffer solution (for the experiments in aqueous medium) or 100 p1 of distilled water (for an alcoholic medium); alternatively x p1 of 4 M sodium hy-droxide solution plus 100-x pl of distilled water were added which adjusts the pH to the desired value without changing the ethanol to water ratio of the samples). A few microlitres of 3.6 x lov3 M rhodium(II1) aqueous or alcoholic standard solution were then added and the resulting solution was made up to volume with distilled water or absolute ethanol.The mixture was heated in a water-bath at 75 "C for 60 min the solution rapidly cooled to room temperature and its absorbance was measured at 327 nm against a reagent blank.8 The absorbance of solutions was stable for a t least 1 h. Absorbance Spectra The absorbance spectrum of a solution of Rh(II1) - TBA at pH 4.7 in aqueous medium (47 pg of rhodium per 5 ml) obtained by following the Procedure is shown in Fig. 3. No differences were observed in the shape of spectra obtained in 98% ethanol. The spectrum of the Rh(II1) - TBA system exhibits a maximum at 327 nm. The absorbance spectra were scanned as described in an earlier paper.* 340 360 380 400 Wavelengthhm Fig.3. Absorption spectrum of Rh(II1) - 2-thiobarbituric acid system. Aqueous medium 47 pg of Rh(II1) per 5 ml and pH 4.7. Reference reagent blank 962 MORELLI SPECTROPHOTOMETRY OF RH(III) IN AQUEOUS Analyst VoZ. 108 Rh3+ TBA molar ratio Fig. 4. Molar ratio of the Rh(II1) - 2-thiobarbituric complex determined by the molar-ratio method. Reference (a) Aqueous medium pH 4.7. 98% ethanol. Reference acetate buffer. (b) 98% ethanol medium pH 2.8. 2-Thiobarbituric acid concentration 6.4 x lo-’ M; wavelength 327 nm. Calibration Graphs To obtain the calibration graphs for the determination of rhodium the described procedure was followed under optimum conditions. Conformity to Beer’s law is observed up to 14 and 12 pg ml-l of metal ion in aqueous and ethanolic medium respectively; the molar absorptivi-ties at 327 nm calculated from the linear regression coefficients are 1.04 x lo4 and 1.08 x lo4 1 mol-l cm-1 and the relevant sensitivities of the reaction for rhodium according to Sandell’s calculation were determined as 6eing 0.0098 and 0.0095 pg cm-2 per 0.001 absorb-ance unit respectively.Composition of the Rh(II1) - TBA Complex From the spectrophotometric data the molar ratio of rhodium(II1) to TBA was found to be 1 2 and 1 4 in both water and The molar ratio method was used in the optimum pH range. TABLE I TOLERANCE OF THE RHODIUM(II1) - 2-THIOBARBITURIC ACID SYSTEM TO DIVERSE IONS All solutions contained 56 pg of rhodium per 5 ml. The tolerance to a foreign ion was taken as the largest amount that gave an absorbance of not more than 1% absolute different from that of rhodium alone.Foreign ion Li(1) . . K(1) . . Tl(1) . . Sr(I1) . . Ba(I1) . . Mn(I1) Fe(II1) . . Ni(I1) . . Cs(1) . . : ;$;:I) :: Co(I1) . . Cu(I1) . . r Aqueous medium 168 162 210 42 11 131 91 2 600 4 68 3.6 4.6 118 608 0.2 98% ethanol ’ medium 101 100 143 10 2.4 6 2 33.6 4.6 3 1.6 1.8 168 32 0.7 Amount tolerated relative to rhodium % Foreign ion 2#) : z$;;\ :: Cd(I1) . . Sn(I1) . . As(II1) . . Sb(II1) . . Bi(II1) . . Ru(II1) . . Pd(I1) . . Ir(II1) . . Os(VII1) Pt(I1) . . Amount tolerated relative to rhodium yo r Aqueous medium 26 412 11 6 10 30 13 83 88 1.6 4 18 62 4.6 98% ethanol ’ medium 10 470 8 24 18 37 4 42 4 0.6 6 17 4 3.Augzcst 1983 AND ALCOHOLIC MEDIA USING 2-THIOBARBITURIC ACID TABLE I1 STATISTICS FOR RHODIUM(III) DETERMINATION BY 2-THIOBARBITURIC ACID 963 No. of standard specimens n = 15; level of significance p = 0.01; and Student’s t = 3.01. Correlation Angular coefficient/ Detection limit/ Medium y-Intercept coefficient ml cm-l pg-l Variance pg ml-l Aqueous . . 0.004 0.999 0.101 0 7.4 x 10-6 0.24 98% ethanol . . . . 0.0004 0.998 0.1048 1.6 x 10-4 0.34 98% ethanol. typical experimental plots obtained at 327 nm with samples containing 5.4 x increasing amounts of rhodium as required. Fig. 4(a) (aqueous medium pH 4.7) and Fig.4(b) (98% ethanol pH 2.8) show M TBA and Effect of Foreign Ions The extent of interference on this spectrophotometric method by the other platinum metals and by common cations was determined at 327 nm as in previous work.* Substances tested and tolerances are listed in Table I. In an alcoholic medium with samples containing concentrations of Tl(I) Mg(II) Ca(II), Sr(II) Ba(I1) and La(1) of 10-20 times higher than those reported in Table I the formation of a finely divided precipitate was observed ; filtration to remove the precipitate was ineffective because it introduced large negative errors in the absorbance (presumably due to adsorption of the complex on the precipitate). In general the tolerance of the Rh(II1) - TBA system to diverse ions is higher in water than in ethanol.Statistical Evaluation of the Method and Conclusion The statistical analysis of experimental results allows a critical evaluation of the proposed method and an objective comparison between the behaviour of the Rh(II1) - TBA system in 35 25 $! u’ 15 5 2 6 10 [Rh3+l/yg mi-’ Fig. 5. Variation of confidence limits at different levels of significance ( p ) in the form of uncertainty (yo) on concentration in A ( p = 0.05) and B ( p = 0.01) aqueous media; and C ( p = 0.05) and D ( p = 0.01), 98% ethanol 964 MORELLI SPECTROPHOTOMETRY OF RH(III) IN AQUEOUS Analyst VoZ. 108 water and in ethanol. In both instances the excellent linearity of the calibration graphs and the conformity to Beer's law is clear from the high values of the correlation coefficients and from the values of the intercepts on they axis (absorbance) which were close to zero.Statistical analysis of the absorbance versysus concentration graphs allows detection limits (DL) and variances (S;) to be calculated.8-1° At the p = 0.01 level of significance the value calculated in both media are as follows in water DL = 0.24 pg ml-l and So2 = 7.4 x and in 98% ethanol DL = 0.34 pg ml-1 and The DL calculated for the rhodium determination in aqueous medium is lower than that found in ethanol; this is because the higher value of the angular coefficient of the calibration graph in ethanol is insufficient to compensate for the greater scattering of experimental points, i.e. the higher value of variance. The values calculated indicate that the proposed method allows the detection of rhodium at very low levels with good precision in both media.Full results of the statistical analysis are listed in Table 11. The mathematical statistics also allow the determination of confidence limits,8-10 which are shown in Fig. 5 in the form of s o 2 = 1.6 x 10-4. TABLE I11 COMPARISON OF SPECTROPHOTOMETRIC REAGENTS FOR RHODIUM(III) Range of Molar concentration1 absorptivity1 Reagent wg mi-' 1 mol-l cm-l 2-Thiobarbituric acid-Thiosalicylamide . . . . 1.25-4.5 7.01 x 10' In water . . . . . . 0.24-14 1.04 x 104 In 98.x ethanol . . 0.34-12 1.08 x 104 ChromeAzurolS 0.7-70 1.2 x 104 1-(2-Pyridylazo)naphth-2-01 . . 25-15 1.1 x 104 1-(2-PyridyIazo)naphth-2-01 . . Up to 40 4.66 x lo4 4.04 x 104 4-(2-Pyridylazo)resorcinol .. Up to 12 1.62 x lo6 4-(2-Pyridylazo)resorcinol . . 2.06-4.95 7 x 10' Nitroso Rsalt 2.06-10.4 8.35 x 10' 1-(1-Thiazolylazo)naph~~-2-01~ * 0.8-4 9.8 x 10' Tropolone Up to 9 1.17 x 104 1,2,4-Triazoline-3-thione . . 1.5-13 3.2 x 104 acid . . . . . . 0.05-1.4 6.5 x 104 acid 0.1-2 3.5 x 104 5-Sulphoallthiox' ' 0.04-0.8 7.4 x 108 pH3 0.4-4.2 1.1 x 104 8-Alkylthioquinoline (Allthiox) 2-25 wg per 25 ml 7 x 10' Sulphochlorophenolazo-rhodanine-In sulphuric acid - acetic In phosphoric acid - acetic Nitroso R salt-pH6 Di-o-to1 ylthiour; . . . . . 0.2-16 1.2 x 104 1.47 x 104 Diphenyldithiophosphinic acid 1.2-8.4 1.11 x 104 E thylenediaminetetraacetic acid . . . . . . . . Up to 150 6.02 x loL Cyclohexanediamine tetraacetic acid .. Up to 150 8.3 x lop Eriochrome Cya&e R -a-Benzil monoxime 10-500 1.2 x 10s p-( Dimethylamino) benzylidenk: rhodanine . . . . . 0.3-5 wg per 25 ml 9.9 x 104 N-a-Pyridyl-N'-benzoyl thio-urea . . . . . . . . 0.69-6.9 9.54 x 108 N-Phenyl-N'-( u-pyridyl) thio-urea . . . . . . . . Up to 70 wg per 2.86 x lo8 Phenanthrenequinone mono-semicarbazone . . Up to 16.46 5.55 x 10' 1-Phenyl-3-thiobenzoylthio-cetylpyridinium bromide . . 0.2-3.2 1.2 x 104 5ml carbamide . . . . . . Up to 5 3.1 x 104 Wavelength/ nm 327 327 380 570 590 594 640 517 510 520 630 400 420 302 510 510 430 445 480 400 360 360 356 630 435 535 360 335 480 325 Sandell's sensitivity1 wg cm-* Heating and comments Reference 0.0098 60 rnin at 75 "C This work 0.0095 60 rnin at 75 "C This work 0.0146 10 min in a hot bath fextrac- 11 0.008 6 0.009 3 0.002 2 0.002 5 0.000 6 0.014 7 0.012 3 0.0105 0.014 7 0.009 0 0.003 2 tion with isobutyl 'methyl ketone - ethanol) 45-50 min in a boiling water-bath (extraction with tri-chloromethane) 4 h on a steam-bath (extrac-tion with trichloro-methane) 4 h on a steam-bath 90 min at 100 "C 60 min boiling Extraction with trichloro-methane 30-60 rnin at 60 f 5 "C Extraction with pyridine in trichloromethane --0.001 6 30-60 rnin at 65-75 "C 0.002 9 0.013 9 0.0094 Brief heating (extraction with tributylamine in trichloro-methane) 0.008 6 0.0070 Extraction with 1,2-dichloro-ethane methane 0.009 3 Extraction with tetrachloro-0.1709 -0.1239 -0.008 6 Ternary complex 0.0857 10 min a t 95 "C 0.0010 60 min at 150 "C (extraction with nitrosobenzene) 0.010 2 Extraction with trichloro-methane 0.0359 Extraction with butanol 0.0185 Extraction with organic sol-0.003 3 Extraction with organic sol-vents vents 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 27 28 29 30 31 32 33 3 August 1983 AND ALCOHOLIC MEDIA USING 2-THIOBARBITURIC ACID 965 percentage uncertainty on concentration Le.as AC/C% for both p = 0.01 (99% probability) and p = 0.05 (95% probability) levels of significance. This allows a direct calculation of the relative uncertainty on concentration over the full range of concentrations tested.In conclusion the determination of rhodium in aqueous medium seems a little more advis-able with respect to the determination in ethanolic medium (in spite of the higher value of molar absorptivity observed in the latter) for the following reasons larger useful concentration range ; lower detection limit ; lower variance ; greater precision ; and greater tolerance to foreign ions. However ethanolic medium allows the determination of rhodium at lower pH values (pHe3.8) with good sensitivity therefore extending the useful pH range over which TBA may be used. Table I11 allows a comparison with the main colorimetric methods for rhodium(II1) developed in recent years. The table shows that TBA ranks among the more sensitive reagents; moreover the present method is shorter and far simpler than many other sensitive methods which require longer sample standing times at high temperature and/or prior extraction of the complex with organic solvents (making the procedure more tedious and time consuming) and are utilisable in narrower concentration ranges.The method is accurate over the concentration range tested and the tolerance of the rhod-ium(II1) - TBA system to a large number of foreign ions is good particularly with respect to other platinum metals which have been major interferents in other known methods. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. 26. 27. 28. 29. 30. 31. 32. 33. 34. References Srivastava V.K. Satsangi R. K. Shankar K. and Kishor K. Pharmazie 1981 36 252. Dhasmana A. Barthwal J. P. Pandey B. R. Ali B. Bhargava K. P. and Parmar S. S. J . Agarwal J. C. Gupta Y . K. Bhargava K. P. and Shankar K. Indian J . Chem. 1981 20 714. Diamantatos A. Anal. Chim. Acta 1978 98 315. Dvinskaya L. M. and Nikiforova L. N. Izuch. Lipidnogo Obmena S-Kh. Zhivotn. 1980 37. Nakashima K. and Akiyama S. Chem. Pharm. Bull. 1981 29 1755. Morelli B. Analyst 1982 107 282. Morelli B. Analyst 1983 108 386. Morelli B. Analyst 1983 108 869. Nallimov V. V. “The Application of Mathematical Statistics to Chemical Analysis,” Pergamon Sur K. and Shome S. C. Anal. Chim. Acta 1969 48 145. Saxena K. K. and Dev A. K. Indian J . Chem. 1969 7 75. Busev A. I. Ivanov V. M. Gorbunova N.N. and Gresl V. G. Tr. Kom. Anal. Khim. Akad. Kodama K. Nagoya-Shi Kogyo Kenkyusho Kenkyu Hokoku 1969 42 48. Kodama K. and Kodama N. Nagoya-Shi Kogyo Kenkyusho Kenkyu Hokoku 1969 42 51. Shrivastawa S. C. Garg V. C. and Dev A. K. Indian J . Appl. Chem. 1969 32 223. Rollins 0. W. and Oldham. M. M. Anal. Chem 1971 43 146. Ivanov V. M. Busev A. Gresl V. G. and Zagruzina A. N. Zk. Anal. Khim. 1971 26 1553. Dedkov Y . M. Lozovskaya L. V. and Slotintseva M. G. Zh. Anal. Khim. 1972 27 512. Rizvi G. H. Gupta B. P. and Singh R. P. Mikrochim. Ada 1972 459. Radushev A. V. and Prokhorenko E. N. Zk. Anal. Khim. 1972 27 2209. Propistsova R. F. and Savvin S. B. Zh. Anal. Khim. 1973 28 1768. Dedkov Y . M. and Slotintseva M. G. Zh. Anal. Khim. 1973 28 2367. Markova L. S. Savostina V. M. and Peshkova V. M. Zh. Anal. Khim. 1974 29 1378. Rakovskii E. E. Shvedova N. V. and Berliner L. D. Zh. Anal. Khim. 1974 29 2263. Kabanova L. K. Solozhenkin P. M. and Usova S. V. Izv. Akad. Nauk Tadzh. SSR Otd. Fiz.-Issa Y . M. and Issa F. M. Fresenius 2. Anal. Chem. 1975 276 72. Duchkova H. Cermakova L and Malat M. Anal. Lett. 1975 8 115. Savostina V. M. Shpigun 0. A. Parmenova V. A. and Peshkova V. M. Zh. Anal. Khim. 1977, Pangarova R. Mosheva P. and Topalova E. Dokl. Bolg. Akad. Nauk. 1977 30 859. Das D. K. Mazumdar M. and Shome S. C. J. Indian Chem. Soc. 1977 54 779. Usatenko Y . I. Meshchervakova N. R. and Pedan V. P. Zh. Anal. Khim. 1979 34 1211. Kamil F. Sindhwani S. K. and Singh R. P. Rev. Roum. Chim. 1980 25 875. Uttarwar R. M. and Joshi A. P. J . Indian Chem. SOL 1981 58 898. Heterocycl. Chem. 1981 18 635. Press Oxford 1963. Nauk SSSR 1969 17 360. Mat. Geo1.-Khim. Nauk 1974 53. 32 656. Received February 7th 1983 Accepted March 81h 198

ISSN:0003-2654    

DOI:10.1039/AN9830800959

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

966 Analyst August 1983 Vol. 108 pp. 966-970 Rapid Spectrophotometric Determination of Saccharin in Soft Drinks and Pharmaceuticals Using Azure B as Reagent P. G. Ramappa" and Anant N. Nayak Department of Post-Graduate Studies and Research in Chemistry University of Mysore Manasa Gangotri, Mysore-670006 India Saccharin reacts quantitatively with Azure B in disodium hydrogen ortho-phosphate - citric acid buffer forming a blue product that can be extracted into chloroform. Beer's law is valid over the concentration range 2-68 pg ml-l of saccharin and the molar absorptivity is 2.4 x lo3 1 mol-l cm-l. Reasonable amounts of ingredients that are likely to be present in soft drinks and pharma-ceuticals do not interfere. Recoveries of saccharin from soft drinks and pharmaceuticals were satisfactory.Keywords Saccharin determination ; spectrophotometry ; soft drinks ; pharma-ceuticals ; Azure B Saccharin has been widely used in medicine and in a variety of food products as a non-nutritive sweetener. There is an extensive literature on the determination of saccharin in food bever-ages. Gravimetry,l polarographyJ2 thin-layer chromatography,3 gas - liquid chromato-graph~,*-~ ion-selective electrode analysisI8 infrared ~pectrometry,~ ultraviolet spectrometrylO and molecular emission cavity analysisll are some of the techniques that have been used, which are tedious and time consuming. The simple technique of high-performance liquid chromatography (HPLC) is widely used for this determination12-14 but requires costly equip-ment. Tris(1,lO-phenanthroline) - iron(I1) chelate,15 phenosulphonaphthalein,la methylene blue,17 phenothiazinels and chlorophenothiazinel9 have been proposed for the spectrophoto-metric determination of saccharin but require vigorous reaction conditions and a longer reaction time.For example phenothiazinels requires the use of a copper(I1) catalyst ethanol and heating at 70 "C for 50 min before extraction into xylene. In this paper we describe the development of a simple but accurate spectrophotometric method for the determination of saccharin in soft drinks and pharmaceutical products. The proposed method offers advantages over HPLC methods in that it requires no costly equipment and it can be easily automated. Experimental Apparatus A Beckman Model DB spectrophotometer with matched 1-cm silica cells was used for absorbance measurements.A digital pH meter Model L1-120 (Elico) was used for pH measurements. Reagents All reagents were of analytical-reagent grade and were used without further purification. Doubly distilled water was used throughout. Azzlre B solzltion. A 0.05% aqueous solution of Azure B (Eastman Kodak) was prepared. Saccharin solzltions. A stock solution of saccharin was prepared from the sodium salt (Boots India) by dissolving 100 mg of saccharin in 100 ml of doubly distilled water and standardising.20 The stock solution was further diluted to give a standard solution. Fresh McIlvaine buffer solutions in the pH range 2.2-8.0 were prepared daily from 0.2 M disodium hydrogen orthophosphate and 0.1 M citric acid.21 Disodizcm hydrogen orthophosphate - citric acid bzlfler.* To whom correspondence should be addressed RAMAPPA AND NAYAK 967 Procedure Preparation of calibration graph Transfer an aliquot of the standard solution containing 20-680 pg of saccharin into a 50-ml separating funnel. Add 5 ml of disodium hydrogen orthophosphate - citric acid buffer of pH 5 and 2 ml of o.05y0 Azure B solution and dilute to 10 ml with water. Mix the solution and add 9 ml of chloroform. Shake the mixture for about 2 min and allow it to settle for about 1 min. Transfer the chloroform extract into a 10-ml calibrated flask and dilute to 10 ml with chloroform. Dry the extract with 0.5 g of anhydrous sodium sulphate and measure the absorbance at 685 nm against a reagent blank prepared under similar conditions.Plot the absorbance against the concentration of saccharin to obtain a calibration graph. Determination of saccharin in soft drinks De-carbonate the soft drink by repeated shaking and pouring from one beaker to another. Transfer the solution into a 100-ml separating funnel. Add 5 ml of 10% sulphuric acid and extract the mixture twice with 30ml of diethyl ether. Discard the lower aqueous layer. Extract the upper ether layer twice with 10 ml of 2% sodium hydrogen carbonate solution. Discard the ether layer. Acidify the aqueous layer with 5% hydrochloric acid and then extract twice with 25 ml of diethyl ether into a conical flask. Evaporate all of the ether from the extract on a hot water-bath. Dissolve the saccharin residue in 10 ml of water and transfer the solution completely into a 100-ml calibrated flask.Dilute the solution to the mark with water. Proceed as described under Preparation of the calibration graph The recovery was tested by adding aliquots of standard solution containing 1.25 mg of saccharin to 50-ml portions of de-carbonated saccharin-free soft drinks. Determination of saccharin in pharmaceutical products Weigh accurately 20 saccharin tablets and grind them to a fine powder in an agate mortar. Transfer an exactly weighed amount of the powder containing 0.5-17.0 mg of saccharin into a 250-ml calibrated flask. Dissolve the powder in doubly distilled water and make up to the mark. Take an aliquot of the diluted solution and determine the saccharin content following the recommended procedure.For the analysis of syrup containing saccharin place the re-quisite volume of the syrup in a 250-ml calibrated flask dilute to the mark with doubly distilled water and proceed as described under Preparation of the calibration graph. Results and Discussion Azure B is an important thiazine dye proposed for the extraction - photometric determina-tion of rhenium22 and antim0ny.~3 It is soluble in water giving a blue solution and is almost insoluble in chloroform. A detailed investigation of the reaction between Azure B and saccharin in various acidic and buffer media showed that Azure B reacts instantaneously with saccharin in disodium hydrogen orthophosphate - citric acid buffer at room temperature, forming a blue product that can be extracted into chloroform.The colour of the chloroform extract showed no loss of absorbance even after standing for 70 h. The optimum pH range for the formation of the blue product was 2.8-6.6. At pH less than 2.8 the reagent is slightly extracted into chloroform. Above pH 7.0 Azure B gives a violet solution that does not react with saccharin. A buffer medium of pH 4.5 was conveniently chosen for further work. Various organic solvents were tried for the extraction of the blue compound from the aqueous solution. The most suitable solvents were chloroform and dichloroethylene. Chloro-form was selected because of its lower volatility and low miscibility in the aqueous phase. A single extraction was sufficient to remove saccharin completely from the aqueous phase. Spectral Characteristics The absorption spectra of the blue product and reagent blank are shown in Fig.1. The maximum absorption of the blue compound in chloroform is at 680-688 nm. The absorption spectrum of the reagent blank under similar conditions shows little absorbance at and near the wavelength of maximum absorption and therefore the analytical conditions are excellent. All subsequent studies were carried out at 685 nm 968 RAMAPPA AND NAYAK SPECTROPHOTOMETRY OF SACCHARIN Analyst Vd. 108 L 2 a 0.2 P 0 450 550 650 750 Wavelengthlnm Fig. 1. Absorption spectra of A coloured product of saccharin - Azure B; and B Azure B reagent blank. Saccharin concentration, 40 pg ml-l; and Azure B concentration, 0.05%. Beer's law is valid over the concentration range 2-68 pg ml-l of saccharin.The optimum The molar absorptivity is 2.4 x lo3 1 mol-l cm-l. range (Ringbom plot24125) is 9-66 pg ml-l. Effect of Reagent Concentration The effect of reagent concentration was investigated by measuring the absorbance at 685 nm of chloroform extracts obtained by the reaction of 25 pg ml-l of saccharin with various amounts of Azure B. A l-fold molar excess of Azure B was required for maximum absorbance; hence 1 ml of 0.05% reagent solution suffices for less than 68 pg ml-l of saccharin. The colour development in the aqueous phase was virtually independent of temperature up to 80 "C. The aqueous phase was cooled to room temperature (27 "C) and the blue compound was then extracted into chloroform. Effect of Concomitant Substances and Application to Soft Drinks Tablets and The effect of some foreign substances that are likely to be present in soft drinks and pharma-ceuticals was studied.Different amounts of substances were added to 25 pg ml-l of saccharin and the colour was developed and extracted into chloroform following the recommended pro-cedure. An error of & 2% in the absorbance reading was considered tolerable. The toler-ance limits for the concentrations of various foreign substances given in Table I show that Syrup TABLE I EFFECTS OF VARIOUS SPECIES ON THE DETERMINATION OF 25 pg ml-l OF SACCHARIN Tolerance limit*/ Species added pg ml-I Ascorbic acid . . 100 Barbitone . . 2 000 Benzoic acid . . 250 Citric acid. . 5 000 Dextrose . . 6 000 Gelatin . . 5 000 Gum acacia . 10000 4-Hydroxybenzoic acid .. 300 Lactose . . 4 500 Maltose . . 6 000 Species added Reserpin Sodium hydrogen carbonate Sodium alginate ,. Sorbic acid . . Starch . . Sucrose . . Stearic acid . . Tartaric acid . . Talc . . Tolerance limit*/ pg ml-1 300 1600 750 500 3 600 7 000 800 7 600 4 500 * Amount causing an error of less than 3% August 1983 I N SOFT DRINKS AND PHARMACEUTICALS WITH AZURE B 969 TABLE I1 RECOVERY OF SACCHARIN FROM VARIOUS SOFT DRINKS Soft drink Torino (orange drink) Torino (lemon drink) Zaffa (mango drink) Ba-jal (orange drink) Ba-jal (cola) . . Bikojoy (orange drink) Bikojoy (cola) . . Double Seven (cola) Amount of saccharin added 25 pg ml-l. Saccharin found*/ 24.25 97.00 25.74 102.96 23.72 95.68 24.75 99.00 24.96 99.84 23.91 95.64 23.36 93.44 24.01 96.80 pg ml-l Recovery % Relative standard deviation yo 1.21 1.24 1.46 0.92 1.04 1.81 1.60 1.12 * Average of five determinations.saccharin can be selectively determined in the presence of many substances that are likely to be present in soft drinks and saccharin tablets. None of the eight soft drinks available in Mysore contain saccharin so a known amount of saccharin was added to the samples. Sucrose present in the soft drink was not extracted into diethyl ether. Only saccharin was extracted into the ether which was subsequently analysed for its saccharin content. The results presented in Table I1 show a recovery of 93.44-102.96~0 of saccharin. The proposed method was successfully applied to the determination of saccharin in pharma-ceutical preparations.The results of the assays of tablets and a syrup presented in Table I11 show a recovery of 98.4-104.7% of saccharin. TABLE I11 DETERMINATION OF SACCHARIN IN COMMERCIAL PHARMACEUTICAL PREPARATIONS Amount of saccharin per tablet or d / m g Average mass of Found; Recovery, Preparation Sample tabletlmg Label clam % Tablets . . . . Sweetex 60.176 12 12.57 104.7 Madhurin 50.55 12 12.43 103.5 Pellet . . . . Sweetex 12.98 13 13.02 100.4 syrup . . Sweetex - 250 246.0 98.4 * Average of five determinations. One of the authors (P.G.R.) is grateful to the University of Mysore for the award of a research grant. 1. 2. 3. 4. 6. 6. 7. 8. 9.10. 11. 12. 13. References Oakley M. S. J . Assoc. Off. Agric. Chem. 1947 30 492. Lasheen A. M. Proc. Am. Soc. Hod. Sci. 1966 77 135. Korbelak T. J . Assoc. Off. Anal. Chem. 1969 52 487. Conacher H. B. S. and O'Brien R. C. J . Assoc. Off. Anal. Chem. 1970 53 1117, Ratchik E. S. and Viswanathen V. J . Pharm. Sci. 1975 64 133. Robert J. J . Assoc. Off. Anal. Chem. 1971 54 1140. Koenig H. Fresenius 2. Anal. Chem. 1971 225 123. Hazemoto N. Kamo N. and Kobatake Y. J . Assoc. Off. Anal. Chem. 1974 57 1205. Coppini D. and Alabasini A. Mitt. Geb. Lebensmittelunters. Hyg. 1968 59 239; Chem. Abstr., Hussein M. M. Jacin H. and Rodriguez F. B. J . Agric. Food Chem. 1976 24 36. Belcher R. Bogdanski S . L. Sheikh A. R. and Townshend A. Analyst 1976 101 562. Szokolay A.M. J . Chromatogr. 1980 187 249. Woodward B. B. Heffelfinger G. P. and Ruggles D. I. J . Assoc. Off Anal. Chem. 1979 62 1011. 1969 70 76503g 970 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. RAMAPPA AND NAYAK Leuenberger U. Gauch R. and Baumgartner E. J. Chromatogr. 1979 173 343. Yamomoto Y. Kumamaru T. Hayashi Y. Yamate M. Kobayashi T. and Tanaka R. J . Pharm. Fernandez-Flores E. Johnson A. R. Leber B Larry D. and Lerner S. J . Assoc. 08. Anal. Beltagy A. Y. Rida S. M. and Issa. A. Pharmazie 1974 29 64. Tanaka A. Nose N. Suzuki T. Kobayashi S. and Watanabe A. Analyst 1977 102 367. Ramappa P. G. and Sanke Gowda H. Fresenius Z . Anal. Chem. 1978 292 413. “British Pharmacopoeia 1973,” HM Stationery Office London 1973. Britton H. T. S. “Hydrogen Ions,” Volume 1 Chapman and Hall London 1955 p. 356. Tarayan V. M. and Vartanyan S. V. Dokl. Akad. Nauk Arm. SSR 1968 47 214. Tarayan V. M. Ovsepyan E. N. and Ekimyan M. G. Uch. Zap. Erevan. Univ. Estestv. Nauk, Ayres G. H. Anal. Chem. 1949 21 652. Ringbom A. 2. Anal. Chem. 1938 115 332. SOC. Jpn. 1968 88 28. Chem. 1973 56 1411. 1972 1 73. Received January 6th 1983 Accepted March 7th. 198

ISSN:0003-2654    

DOI:10.1039/AN9830800966

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst August 1983 Vol. 108 @. 971-977 971 Interfacing an Automatic Elemental Analyser with an Isotope Ratio Mass Spectrometer the Potential for Fully Automated Total Nitrogen and Nitrogen-15 Analysis Thomas Preston Scottish Universities Research and Reactor Centre N.E.L. Estate East Kilbride Glasgow G75 OQ U and Nicholas J. P. Owens Natural Environment Research Council Institute for Marine Environmental Research Prosfiect Place The Hoe Plymouth PL1 3DH An interface between an automatic nitrogen analyser and an isotope mass spectrometer is described. The interface enables total nitrogen analyses and nitrogen isotope ratio measurements to be made semi-automatically at the rate of 12 h-l. The precision of the measurements was at the detection limit of the mass spectrometer (0.00026 atom-% 15N; f 0 .7 r ’ / ~ ~ ) . No significant cross-contamination between samples was observed (< 0.02% carry over). The potential for full automation and extension of the interface for other stable isotope measurements is discussed. Keywords Nitrogen- 15 isotope mass spectrometry ; automatic nitrogen analysis ; elemental gas chromatograph - mass sfiectrometer interface There is increasing interest in the application of stable isotopes in agricultural science clinical science and aquatic biology. This is especially so for nitrogen tracer studies. Typical experiments generate large numbers of samples and require analytical techniques that are both precise and rapid. Mass spectrometry (MS) is the favoured technique for isotope analysis. The majority of nitrogen samples for MS are processed by Kjeldahl digestionf to ammonium for total nitrogen analysis; ammonium is then oxidised to nitrogen gas by the Rittenberg technique,2 for isotope analy~is.~ This process is time consuming and requires careful super-vision at every stage in order to avoid nitrogen loss contamination or incomplete recovery that can lead to isotope fractionation.An additional problem is that without modification, Kjeldahl digestion does not reduce oxidised nitrogen species. This may lead to loss of accur-acy for example during the analysis of plant tissue that may contain significant amounts of nitrate . The main alternative to the Kjeldahl - fittenberg technique is Dumas comb~stion.~ This technique has many advantages over Kjeldahl digestion not least because Dumas combustion is a single-step process producing nitrogen by the oxidation of organic matter.Increased use of Dumas combustion for nitrogen isotope analysis was predicted by Bremne~-.~ Mass spectrometry was used earliers to detect products of incomplete combustion in a Dumas apparatus and Holt and Hughes’ studied the decomposition of diazo compounds by isotope labelling using Dumas combustion and mass spectrometry. Laboratory-built Dumas combustion apparatus that has been developed for nitrogen isotope analysis is as sophisticated as commercially developed equipment.*-1° Use of less sophisticated apparatus, which is nevertheless suitable for carbon-isotope analysis in organic material may lead to inaccuracies in nitrogen isotope analysis owing to the presence of small amounts of carbon monoxide hydrocarbons at m/z 28 and m/x 29 (shown by elevated m/x 27 peaks) and nitrogen oxides at m/x 30.11 Barsdate and Dugdale12 were the first to modify a commercial nitrogen analyser for nitrogen-isotope analysis.The potassium hydroxide trap of a Coleman nitrogen analyser13 was re-placed with a liquid-nitrogen trap to remove the carbon dioxide carrier and pure nitrogen was expanded into the mass spectrometer. This method was later refined to increase sample throughput and precision.14J5 Modern designs of nitrogen analysers with helium carrier and oxygen pulse (see later the Carlo Erba ANA and also Tsuji et aZ.16) offer some advantages ove 972 PRESTON AND OWENS AUTOMATIC ANALYSER WITH ISOTOPE Analyst VOZ.108 the latter method. Oxidation conditions can be optimised for each sample batch by varying the oxygen pulse; products can then be checked by mass spectrometry. Because samples are always accompanied by a large excess of helium there are no leakage problems; the range of sample sizes may thus be varied considerably. Memory effects are also reduced as the analyser is continually swept with carrier gas. Furthermore the carrier does not interfere with nitrogen isotope analysis. Continuous or batch methods can be used to interface a nitrogen analyser [essentially a gas chromatograph (GC)] with MS. Batch methods enable more precise isotope-ratio analysis but require sample trapping; continuous interfaces are simpler and do not interrupt GC operation. A simple MS inlet system the continuously sampling capillary can monitor a GC effluent.The capillary is crimped to limit pressure in the MS and the excess GC effluent flows to waste. A continuous interface that reduces carrier flow into the MS such as a jet separator is commonly used in GC - MS. A separator interface has been used in the analysis of nitrogen isotope ratios in the effluent of a Carlo Erba 1104 elemental analyser with an organic MS.17 However carrier separators are inefficient at low masses and introduce an isotope-fractionation effect because their operation is mass dependent. The performance of peak switching single collector mass spectrometers with changing sample pressures is also limited because there is often insufficient time for accurate measurement of isotope ratios.Metzger17 obtained a reproducibility of &- 0.01 atom-% 15N for isotope-ratio measurement with an elemental analyser coupled to an organic MS. The greater sensitivity of electron multi-plier detector MS can be overcome in part by the operation of electrometer detector MS at higher pressures. The limitations of single collector techniques may be overcome by the use of a double electrometer magnetic sector MS with electronic ratio measurement. This paper describes a method of nitrogen isotope measurement by means of an automatic nitrogen analyser interfaced to a double collector mass spectrometer. Experimental Elemental Gas Chromatograph (EGC) A Carlo Erba (Milan Italy) ANA 1400 automatic nitrogen analyser was used throughout this work. The mode of operation is based on the Dumas principle with high temperature “flash” combustion.The carrier gas used was C. P. grade helium (BOC Special Gases London) at 60 ml min-l. The oxygen used was argon-free oxygen (BOC) with a 5-ml loop. The analyser has a 24 seat automatic sampling manifold capable of accommodating solid or liquid samples. One complete analytical cycle takes 5 min. The sample size range is between 0.18 and 180 pmol of N with an accuracy of &- 0.2% at 35 pmol of N (Carlo Erba). Mass Spectrometer (MS) The instrument used was a V.G. Isogas (Middlewich Cheshire) MM 622 isotope mass spectrometer. The MM 622 is a single inlet double-collector instrument with a gas tight, electron impact ionisation source. The electron energy was 70 eV and filament current 200 pA.The standard instrument has an internal precision of The all-metal inlet was modified from the original to that shown in Fig. 1. Although the standard MS sensitivity would be suitable for the normal EGC sample size, approximately 50 pmol of N, an increase in MS sensitivity was desirable for the samples used during this study (1-10 pmol of N,). Methods considered were increasing filament current, purchasing a high sensitivity ion source or high-precision electronics (for improved precision at low peak heights rather than greater sensitivity) or using a greater sample leak rate into the ion source. The last method has proved an effective and economical alternative. The 800mm x 0.15 i.d. capillary of the standard inlet system was replaced by an 800 mm x 0.35 mm i.d.stainless-steel capillary and crimped to give an analyser pressure of 0.76 x Torr helium (measured by ion gauge a standard fitting on the mass spectro-meter used). This ensured that the true analyser pressure never went above Torr, satisfying ion mean free path and peak broadening considerations (see later). Because the MS pumping is by a 170 1 s-1 oil diffusion pump. 0.00026 atom-% I5N & o.7°/o at natural abundance.* * O l a o = [(g sample - 2 standard)/$ standard x 1000. 28 2 v 4 Fig. 1. EGC interface and MS inlet system. ANA automatic nitrogen analyser (EGC) ; 1-8 high vacuum valves; VR variable-volume steel bellows reservoir 10-70 ml; MS capillary leak to mass spectrometer; R rotary pump; SL sample loop (80 ml) ; CF cold-finger (optional) ; EL exhaust loop &-in stainless-steel tubing except EL (PTFE) .0 5 I I L l rJ L A Procedure To operate the EGC alone close 2 3,4 5 7 and 8 open 1 (see Fig. 1). The EGC - MS may be operated in either one of two modes. (i) Continuous mode close 1 open 2 4 7 and 8. Measure the 29/28 ratio at maximum or pre-determined sample pressures. (ii) Batch mode initially the sequence of valve operation is as the continuous mode but check that VR is fully evacuated. Monitor m/z 28; when the majority of the sample is in the sample loop (if SL is 80 ml and the EGC carrier is at 60 ml min-l 99% of the sample is con-tained within SL when m/z 28 has reduced to 5% maximum) open 1 close 2 and 8. Open 3 and fully expand VR; close 7; monitor m/z 28 and adjust VR to give the required value; measure the 29/28 ratio.Open 7 and evacuate inlet by opening 5 (6 is normally left open); when evacuated close 5 and 3 open 8 and 2 close 1. I t is important to operate valves in the sequence described to minimise perturbations to the EGC carrier flow. bl 0 4 0 I 4 C) .-0 4 Time/mi n Fig. 2. (a) EGC trace; (b) m/z 28 from MS in continuous mode; and (c) m/z 29 from MS in continuous mode. All samples 3.6 pmol of nitrogen. 0 5 10 Time/min Fig. 3. The m/z 29 and 29/28 ratio traces produced by the consecutive analysis of three replicate urea stand-ards (3.6 pmol of nitrogen) 974 PRESTON AND OWENS AUTOMATIC ANALYSER WITH ISOTOPE Analyst VoZ. 108 Results and Discussion Peak Resolution Fig. 2 shows the peak shapes produced by the EGC and MS operating in the continuous mode.The peaks are essentially similar and show little broadening. The m/z 29 and m/z 29/28 ratio display is sufficiently flat-topped to be suitable for isotope-ratio measurement. These traces compare with GC - MS isotope-ratio traces modelled by Matthews and Hayes.lg Fig. 3 also shows that samples can be analysed consecutively within the analytical cycle time of the EGC thus enabling 12 total nitrogen and nitrogen isotope analyses to be completed in 1 h. Fig. 4 shows the 29/28 ratio for marine phytoplankton samples artificially enriched with 15N. Again the trace shows that samples may be analysed consecutively within the analytical cycle time of the EGC and that there is good reproducibility between replicates. 0.0 (a) 0 .- CI 2 0.02 -a0 cy cv 0.04 -I I I I I I 0 10 20 30 40 50 Time/min Fig.4. A 29/28 ratio trace for (a) unenriched marine phytoplankton ( b ) (c) and (d) marine phytoplankton with increasing 15N enrichment. Table I shows typical m/z 29 peak heights for a range of nitrogen concentrations. The response of the MS is reproducible and with appropriate calibrations m/z 28 peak height may be used as an alternative to the EGC determination of the nitrogen content of samples. Background Also shown in Table I is the background m/z 28 and m/z 29 peak heights of the MS and the range of blank values obtained for the EGC. There is clearly a significant background con-tribution (approximately 0.1 pmol of N,) from the EGC carrier and the oxygen used to aid sample combustion.No attempt was made to purify these gases but a reduction in the blank would enable lower nitrogen content samples to be analysed. For low nitrogen content work (1 pmol of N,) and depending upon the sample type the oxygen-loop size may be reduced without decreasing the efficiency of the combustion ; combustion efficiency for specific samples should be checked by MS scanning. TABLE I BACKGROUND m/Z 29 PEAK HEIGHTS AND m/Z 29 PEAK HEIGHTS FOR A RANGE OF NITROGEN STANDARDS Sample m/z 29 peak height/A x 10-la MM 622 background . . 0.012-0.018 EGC carrier blank . . 0.81-0.93 EGC carrier + 6-do injection . . 2.472.8 Urea standards/pmol N,-1.4 . . 23.09-24.11 1.8 . . 30.69-30.6 1 3.6 . . 69.99-60.11 2.7 . . 46.19-46.2 1 4.3 .. 72.99-73.1 Atxgtxst 1983 RATIO MS POTENTIAL FOR AUTOMATED N AND 15N ANALYSIS 975 Memory Effects Memory effect in the MS inlet during batch operation can be removed by pumping between samples. The memory in the EGC and MS between consecutive samples analysed in the continuous mode was tested by comparing the blank m/z 29 peak heights before and after the analysis of a sample of both high nitrogen content (40pmol of N,) and 15N enrichment (approximately 5 atom-% 15N). Immediately following this sample the peak carry over on the next three blank determinations was equivalent to 0.0160 0.0006 and 0.0000% of the test sample. This effect was reproducible and therefore presents little problem particularly if samples of similar nitrogen content and enrichment are analysed consecutively.Precision Table I1 shows that at constant pressure sample reproducibility is as good as the internal precision of the MS (& 0.00026 atom-% 15N & 0.70/00) at natural abundance. A “pressure effect,” which gives rise to variations in the measured isotope ratio as a function of sample pressure causes a loss of precision in the continuous mode [Table II(a)]. TABLE I1 REPRODUCIBILITY OF 29/28 RATIO ANALYSIS OF ZERO ENRICHMENT UREA STANDARDS Range of Oleo 15N obtained for standards analysed a t a standard pressure* (3.5 pmol of N,; n = 6) Variations of oleo 15N obtained for a range of sample size; three replicate determinations on 1.75 pmol N and 7 pmol of N (50% and 200% of Mode of operation standard pressure*) (a) Continuous ratio measurement at maximum m/z 28.. f0.7 (b) Continuous ratio measurements as mlz 28 passes pre-determined level . . f0.7 (c) Batch analysis ratio measurement with mlz 28 adjusted to standard pressure* . . f0.7 * m/z 28 = 0.7 x lo-* A. f28.8 f 1.4 f0.7 The pressure effect is of the order of 0.01 atom-% 15N & 28.8O/,, at natural abundance, over a four-fold range of sample pressures. The effect may be corrected for by a model. However if no correction is made the precision obtained may still be within the limits required for some isotope enrichment experiments especially if samples and standards of similar nitrogen content are to be analysed. The pressure effect in the continuous mode may be minimised by measuring the 29/28 ratio at pre-determined pressures; sample results may then be compared with standards at similar pressures as is usual with single inlet MS.Table II(b) shows results obtained by recording the 29/28 ratio manually as sample pressure passed pre-determined m/z 28 peak heights. The precision obtained by this method (& 0.0005 atom-% 15N & 1.4°/00 at natural abundance) is suitable for enrichment experiments. Measurements of the 29/28 ratio at pre-determined peak heights as the major ion peak rises and falls can also demonstrate and by averaging eliminate any chromatographic resolution in the EGC of the isotopic species. For the most precise work [& 0.7°/,, Table II(c)] the batch mode of analysis is used to remove pressure effects. In this instance after trapping in a sample loop a variable volume bellows is used to adjust samples and standards to a constant major peak height.The reproducibility of ratio measurements in the batch mode may theoretically be reduced by peak broadening effects caused by the variation of total pressure (helium pressure) within the MS. This is caused by the use of the variable volume to bring nitrogen pressure to a standard value. Peak broadening was investigated but was less than the detection limit of the instru-ment over a four-fold pressure range. Analysis at very large pressure variations showed that peak broadening (or the abundance sensitivity correction20) varies by less than 0.lo/, for a four-fold pressure change. For a discussion of peak broadening effects in this type of MS see reference 21 976 PRESTON AND OWENS AUTOMATIC ANALYSER WITH ISOTOPE Analyst VoZ.108 The performance of the EGC - MS in the analysis of 15N-enriched standards is shown in Table 111. The data show that for a range of standards with nominal 15N enrichments of 0.5 atom-% 15N to 10 atom-% 15N there was a mean (‘yt = 30) coefficient of variation (CV) of 0.34%. This compares with a CV of 0.07% at natural abundance (see Table 11) and together with the slope of 1 (P <0.001) for the expected and obtained 15N enrichments indicates that there is no significant deviation from linearity of the MS collectors and amplifier circuits. TABLE I11 MEASURED 15N ATOM-% OF A RANGE OF NOMINAL 15N STANDARDS Measured atom-% 15Nt Nominal atom-% 15N* ‘2 (n = 6) Standard deviation CV ./d 0.5 0.506 0.002 0.40 1 .o 0.983 0.003 0.31 2.0 1.944 0.013 0.67 5.0 6.149 0.009 0.17 10.0 10.079 0.015 0.16 * Standards obtained from Los Alamos National Laboratory NM.Corrected for carrier and oxygen blank. Although unnecessary for this work it may be desirable to remove the excess of carrier pressure from samples for more accurate batch analysis or for work with an MS that has a greater peak broadening problem. A cold trap (CF Fig. 1) can be used in place of the sample loop. For nitrogen work this could be a cryosorption trap filled with a molecular sieve at liquid-nitrogen temperature. A cold trap would allow removal of helium pressure by pumping and also give an increase in sensitivity as it would have a smaller volume than the sample loop. A cryosorption system was not used during this study because the sample loop performed well and had certain advantages.Sample loop timing problems were overcome by using the continuous sampling capillary to facilitate trapping. Disadvantages of a cryosorption system include an increased and variable carrier blank contribution because a larger volume of carrier would be sampled than in an ambient temperature sample loop. Also chances of sample contamination by interfering gases that have been separated chromatographically by the EGC would be increased. A cryosorption system would also be more difficult to automate and may require a longer cycle time than the 5 min of the EGC used in this study. Future Development Although the system described above is semi-automatic various stages of further automation are possible.Automatic recording of isotope ratios at pre-determined major peak heights would reduce the pressure effects apparent in the continuous sampling mode [see Table 11 ( b ) ] . An important feature of the continuous mode is its potential for greater sensitivity. In the system described above approximately 1% of the sample enters the MS 99% flows to waste and serves no function. In conventional batch analysis the excess of sample serves to facilitate handling and reduce leakage problems in vacuum systems and to maintain viscous flow conditions and constant pressure in the MS inlet. The full sensitivity of the continuous inter-face used could be realised at capillary GC flow-rates (< 1 ml min-1 of helium) allowing direct coupling to the MS source and precise analysis of as little as 10-100 nmol of N (1 nmol of N, if the carrier blank is reduced).This would provide an alternative and more precise isotope-analysis system for GC effluents than the isotope ratio monitoring apparatus (IRM - GC -MS) described by Matthews and Hayes.lg For a high precision batch system automatic vacuum values and variable reservoirs are available that would permit fully automatic isotope analysis allowing microprocessor control and routine use by unskilled operators. Precision was limited in the system described by the ratio-measurement electronics of the MS ; this would be improved by using the high-precision version of the MM 622 or a double-collector dual-inlet MS (differential MS with helium diluted reference gas).The cycle time (5 min) of the EGC used above is too short for the usual differential MS sequence (approximately 8 min) ; however the cycle time of the similar Carlo Erba 1106 C N H 0 and S elemental analyser is sufficient. An interferace of the typ August 1983 RATIO MS POTENTIAL FOR AUTOMATED N AND 15N ANALYSIS 977 described (either continuous or batch depending upon the precision required) between an elemental analyser and an isotope MS would therefore provide the potential for a versatile, isotope-analysis system that could be used for the rapid analysis of 15N 13C lSO 34S and H/D in any sample that can be handled by the EGC. Conclusions An interface between an automatic nitrogen analyser and a dual collector isotope MS enables the precise and rapid analysis of total nitrogen and nitrogen isotope ratios to be per-formed.Isotope-ratio measurement may be made in one of two modes in a simple continuous interface at pre-determined sample pressures and by batch analysis depending upon the level of precision required. The interface described is semi-automatic but can be modified to a fully automated system which with alternative instrumentation may be used as the basis of isotope ratio analysis apparatus for a wide range of light gas stable isotopes. We thank the Director of IMER for his encouragement and for the use of laboratory Thanks are also due to Dr. A. E. Fallick for helpful discussions during the prepara-Dr. B. B. McInteer kindly provided the Los Alamos 15N standards. facilities. tion of the manuscript.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. References Kjeldahl J. 2. Anal. Chem. 1883 22 366. Sprinson D. B. and Rittenberg D. J . Biol. Chem. 1949 180 707. San Pietro A. in Colowick S. P. and Kaplan N. O. Editors “Methods in Enzymology IV,” Steyermark A. “Quantitative Organic Microanalysis,” Second Edition Academic Press New York, Bremner J. M. in Black C. A. Editor “Methods of Soil Analysis,’’ Agronomy Volume 9 Part 2, Van Meter R. Bailey C. W. and Brodie E. C . Anal. Chem. 1951 23 1638. Holt P. F. and Hughes B. P. J . Chem. SOC. 1953 1666. Agricultural Research Council Letcombe Laboratory Wantage Oxford. Fiedler R. and Proksch G. Anal. Chim. Acta 1972 60 277. Wada E. Tsuji T. Sairo T. and Hattori A. Anal. Biochem. 1977 80 312. Preston T. and Griffiths H. unpublished work. Barsdate R. J. and Dugdale R. C. Anal. Biochem. 1965 13 1. Sternglanz P. D. and Kollig H. Anal. Chem. 1962 34 644. Desaty D. McGrath R. and Vining L. C. Anal. Biochem. 1969 29 22. Pavlou S. P. Friederich G. E. and MacIsaac J. J. Anal. Biochem. 1974 61 16. Tsuji O. Masugi M. and Kosai Y. Anal. Biochem. 1975 65 19. Metzger J. Fresenius 2. Anal. Chem. 1978 292 44. Schoeller D. A. and Hayes J. M. Anal. Chem. 1975 47 408. Matthews D. E. and Hayes J. M. Anal. Chem. 1978 50 1465. Deines P. Int. J. Mass. Spectrom. Ion Phys. 1970 4 283. Fallick A. E. and Baxter M. S. Int. J . Mass. Spectrom. Ion Phys. 1977 25 155. Academic Press New York 1957 pp. 473-488. 1961. American Society of Agronomy Madison WI 1965 pp. 1256-1286. Received December 13th 1982 Accepted Ma~ch 4tk 198

ISSN:0003-2654    

DOI:10.1039/AN9830800971

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

978 Analyst August 1983 Vol. 108 pp. 978-983 Determination of Phosgene in Methylene Chloride After Cyclisation With a 2-Hydroxyamine and Gas Chromatography With N it rog en -sel ective Detection Olle Gyllenhaal Analytical Chemistry AB Hassle S-431 83 Molndal Sweden A convenient and simple method for the determination of phosgene in methylene chloride has been developed. An aliquot of the sample is mixed with a solution of metoprolol or 2-aminophenol in an excess. After a reaction time of 10min the solution is taken to dryness and the cyclic derivative analysed by gas chromatography with nitrogen-selective detection. The precisions (relative standard deviation) at the 40 and 10 ng ml-l levels were 3.9 and 7.3% for metoprolol and 3.3 and 3.6% for 2-aminophenol respec-tively.The absolute yields at the 40 ng ml-l level were 91 and 95% respec-tively. The present limit of detection is approximately 1 ng ml-l. Phosgene appeared in methylene chloride (> 2 ng ml-1) stabilised with 20 p.p.m. of amylene within 3 d if stored in clear glass in the presence of daylight. After 15 d the 100 ng ml-l level was reached. Keywords Phosgene determination ; 2-hydroxyamines ; gas chromatography ; nitrogen-selective detection ; methylene chloride Phosgene (carbonyl dichloride) is a commercially important starting material in the production of polycarbonates dyes and aromatic isocyanates. Minor amounts are consumed in the synthesis of pharmaceutical products and plant protection agents. Phosgene is synthesised from chlorine and carbon monoxide in the presence of a suitable catalyst.It can also be formed by thermal decomposition of chlorinated hydrocarbons such as carbon tetrachloride and trichloroethylene or by oxidative degradation of such solvents in the presence of light. The latter process may take place both in the pure solvent and in the atmosphere. Most of the earlier methods for the determination of phosgene are colorimetric e.g. based on reactions with 4-(4'-nitrobenzy1)pyridine and N-benzylanilinel or aniline.2 Also iodine liberated from the reaction with potassium iodide in acetone can be used and has been adapted to the determination of phosgene in chlorinated solvent^.^ Over the last decades gas chroma-tography has been used increasingly for phosgene determination in process mixture^^-^ and in the atm~sphere.~-lO For low concentrations of intact phosgene the detector of choice is the electron-capture In the course of developing a method for the /3-blocking drug metoprolol using phosgene as a derivatising agentll it was thought that it would be of interest to investigate whether this reaction could be reversed and used for the determination of phosgene itself.Methylene chloride was chosen as the sample as it is commonly used for the extraction of amines and preliminary experiments showed that significant amounts of phosgene could be formed easily. Phosgene present in the sample is cyclised with an excess of metoprolol or 2-aminophenol. After evaporation of the organic phase the residue is dissolved in a minute volume of ethyl acetate and analysed by gas chromatography with nitrogen-selective detection.Experimental Instrument A Varian 3700 gas chromatograph with a thermionic (nitrogen - phosphorus selective) detector was used. The glass columns (120 and 60 x 0.2 cm i.d.) were filled with 3% Hi-EFF-8BP on Gas-Chrom Q 100-120 mesh (Applied Science Laboratories State College PA USA) and 3% Carbowax 20M on the same support (Ohio Valley Speciality Chemical Marietta OH, Presented in part at a meeting of the Scandinavian section of the Chromatography Discussion Group, held in May 1982 GYLLENHAAL 979 USA) respectively. The nitrogen carrier gas flow-rate was 45 and 30 ml min-l respectively. The column oven temperature was 240 "C for metroprolol and 200 "C for 2-aminophenol. The injector and the detector were maintained at 250 and 300 "C respectively.Reagents and Chemicals Phosgene 2 M in toluene. Amylene (2-methylbut-2-ene. Aldrich (Milwaukee WI USA). Methylene chloride. Merck (Darmstadt G.F.R.) pro analisi grade stabilised with 20 p.p.m. of amylene. Methanol and ethyl acetate. Merck (Darmstadt G.F.R.) pro analisi grade. Hexane. 2-Aminophenol and 2-amino-4-methylphenoE. Fluka. By reacting the latter compound with phosgene a suitable gas chromatographic marker was obtained (5-methylbenzoxazoline-2-one) for the determination of phosgene with 2-aminophenol as reagent. Metoprolol (1-isopropylamino-3- [4- (2-methoxyethyl)phenoxy]propan-2-ol tartrate } 3-isopropyl-5-[ 4- (2-methoxyethyl)phenoxymethyl] oxazolidin-2-one and 1 -isopropylamino-3- [4- (2-ethoxymethyl)phenoxy]propan-2-ol.Synthesised at AB Hassle (Department of Organic Chemistry). 2-Benzoxazolinone. Synthesised at AB Hassle but can also be obtained commercially e.g., from Aldrich. Fluka (Buchs Switzerland) purum grade. Rathburn Chemicals (Walkerburn UK) HPLC grade. Methods for the Determination of Phosgene Reaction with metoprolol as reagent The sample (1 ml) was mixed with 1 ml of a 10 pg ml-1 solution of a metoprolol in hexane. The metoprolol solution also contained 100 pg of the marker. After 10 min the reaction mixture was mixed with 0.5 ml of buffer pH 5 (ionic strength p = 0.1) and vibrated for 30 s. Most of the organic phase was then removed and taken to dryness by a stream of nitrogen. The evaporation residue was dissolved in 25 p1 of ethyl acetate and 2-pl aliquots were analysed- by gas chromatography with nitrogen-selective detection after separation on 3% Hi-EFF-8BP.Reaction with 2-aminophenol as reagent The sample (1 ml) was mixed with 1 ml of a 5 pg ml-l solution of 2-aminophenol in methy-lene chloride also containing 140 ng ml-l of the marker. After 10 min the solvent was removed by evaporation and the residue dissolved in 25 pl of ethyl acetate. Gas chromatographic separation was on 3% Carbowax 20M prior to nitrogen-selective detection. Standard samples of phosgene in methylene chloride were prepared from fresh dilutions of the 2 M solution diluting to 400 ng ml-1 with sodium sulphate-dried methylene chloride. Volumes of 10-2OOpl were then added to 1-0.8-ml aliquots of methylene chloride and a known amount of the appropriate marker.The samples were then analysed immediately by either of the two procedures. Determination of actual phosgene concentration in the 2 M toluene solution Triplicate 50-p1 samples were hydrolysed in 25 ml of de-ionised water in 250-ml Erlenmeyer flasks. The chloride ion content was then determined by potentiometric titration with 0.1 M silver nitrate solution. The concentration of phosgene was calculated assuming that all of the chloride ions emanated from phosgene. Potassium chloride (pro analisi) was used to calibrate the silver nitrate solution. Results and Discussion Choice of Derivatising Agent To complement the use of the pharmaceutical compound metoprolol as the reagent other alternatives were investigated. Attempts to cyclise 2-aminoethanol and 2-ethylamino-ethanol were not successful.No main product was obtained but several peaks were observed in the gas chromatograms. One possible explanation for this is that oligo- or polymeric pro-ducts are also formed instead of a single cyclic derivative. 2-Aminophenol was found to react to give a single derivative and was selected as an alternative to metoprolol 980 GYLLENHAAL DETERMINATION OF PHOSGENE IN METHYLENE Analyst VoZ. 108 I - -- Q1 a ? I I I I 5 10 20 50 1 2.5 5 Amount of metoproloVpg Amount of 2-aminophenoVpg Reaction yield as a function of the amount of reagent, which was (a) metoprolol and (b) 2-aminophenol. The reaction was performed by mixing 1 ml of methylene chloride and 400 ng of phosgene with (a) 1 ml of hexane containing metoprolol and marker and (b) 1 ml of methylene chloride containing 2-amino-phenol; caffeine was only used with 2-aminophenol before the appropriate marker had been synthesised.Fig. 1. a Time/m in Fig. 2. Reaction yield as a function of time using 2-aminophenol as reagent. Methylene chloride (6 ml) containing 5 pg .ml-l of 2-aminophenol and caffeine as marker was mixed with 6 ml of 400 ng ml-l of phosgene in methylene chloride. At intervals 0.6-ml aliquots were withdrawn and mixed with 0.5 ml of methanol. After evaporation the residue was dissolved in 26 pl of ethyl acetate. Derivatisation Conditions The amount of metoprolol base required to derivatise 400 ng of phosgene was found to be at least 2 pg [Fig. l ( a ) ] that is approximately a three-fold excess.It is assumed that part of the excess of metoprolol acts as an acid scavenger. Thus it is recommended to use an ade-quate excess of metoprolol as this will remove any acid already present in an unknown sample. The time course for the selected concentration (10 pg ml-l) was constant from 0.5 to 60 min. The amount of 2-aminophenol required to derivatise the same amount of phosgene is shown in Fig. l ( b ) . The amount selected was 5 pg and contrary to the metoprolol - phosgene reaction the reaction is not instantaneous in this instance (Fig. 2). A constant yield was obtained from 7.5 to 210 min. A reaction time of 10 min was used with both reagents. 9 6 3 0 9 6 3 c Time/m in Fig. 3. Gas chromatograms for the determination of phosgene in methylene chloride with metoprolol as reagent.(a) Sample with 6.6 ng ml-1 of phosgene; (b) blank sample with marker added. Chromatographic conditions were as follows 3% Hi-EFF-8BP column a t 2&0 "C; P phosgene as 3-isopropyl-5-[4-(2-methoxyethyl)]phenoxymethyloxazolidin-2-one; and M, marker same as above but 2-ethoxymethyl August 1983 Purification of the Reaction Mixture The remaining excess of metoprolol had to be reduced to avoid chromatographic disturb-ances. This was accomplished by a brief buffer (pH 5 ) wash of the organic phase prior to the evaporation step. No such precautions were necessary with the excess of 2-aminophenol. This might be due to the lower chromatographic ability of the phenol group. CHLORIDE AFTER CYCLISATION AND GC WITH N SELECTIVITY 981 M 5 3 0 5 3 0 Ti me/mi n Fig.4. Gas chromatograms for the determination of phosgene in methylene chloride using 2-aminophenol as reagent. (a) Sample with 5 ng ml-l of phosgene and (b) blank sample with marker added. Conditions as follows column 3% Carbowax 20M a t 200 "C; P phosgene as benzisoxazolin-2-one; and M marker (6-methylbenzisoxazolin-2-one). Phosgene concentrationlng ml- ' Fig. 5. Standard graph for the determination of phosgene in methylene chlor-ide using metoprolol as re-agent prepared as described under Experimental. Each point represents the mean of duplicate samples. Linear regression slope 0.027 ml ng-l intercept 0.019 and correlation coefficient 0.999 8. Gas Chromatography As discussed elsewherell the oxazolidine derivative of metoprolol required a polar stationary phase such as Hi-EFF-8BP (cyclohexanedimethanol succinate) in order to chromatograph properly.The column temperature of 240 "C is near to that of the recommended maximum operation temperature of this phase (250 "C) but so far no problems with short column life or detector contamination have been experienced. Fig. 3 shows the chromatograms from a blank sample and a sample found to contain 6.5 ng ml-l of phosgene. The separation is not com-plete between the derivative and the marker but is sufficient for quantitative purposes. Carbowax 20M was found to be an alternative stationary phase for the 2-aminophenol derivative. Chromatograms with this phase are shown in Fig. 4. No reacting internal standards were available.Instead markers were used with almost identical structure and suitable chromatographic retention characteristics (Figs. 3 and 4). Standard Graphs Precision and Absolute Yield Standard graphs for the determination of phosgene in methylene chloride were prepared in the range 4-80 ng ml-l (nominal concentration). An example of a standard graph is given in Fig. 5. In this example metoprolol was used as the reagent. The intercept of the graph i 982 GYLLENHAAL DETERMINATION OF PHOSGENE I N METHYLENE Ana&St VOi?. 108 partly a technical one owing to the incomplete separation of the marker (compared with Fig. 3) and how the base line was drawn. Blanks were sometimes encountered when 2-aminophenol a very polar reagent was used as the reagent and were due to the formation of phosgene in the methylene chloride which was used as the solvent.Using metroprolol as the reagent the precision in the determinations of phosgene at the 40 and 10 ng ml-l levels were 3.9 and 7.3y0 respectively (n = 8). The somewhat higher values obtained as compared with the determination of metoprolol with phosgenell were felt to be due to the absence of a reacting internal standard and the volatility and reactivity of phosgene itself. The yields at the 40 ng ml-l level were 91% (metoprolol) and 95% (2-aminophenol). Around 6% is lost in the transfer step after the buffer wash. The yields were calculated after com-parison with reference solutions of the pure derivatives. Diluted solutions of phosgene were generally used immediately as the concentration of phosgene gradually declined.A 400 ng ml-l solution in dry methylene chloride lost 15% over the first 24 h. After 1 week the loss was about 40%. With 2-aminophenol as the reagent the precision was 3.3 and 3.6y0 respectively. Formation of Phosgene in Methylene Chloride At an early stage of this work a screening of methylene chloride samples for the presence of phosgene was conducted. The presence of phosgene in several of these samples in high con-centrations prompted the more continuous monitoring of methylene chloride stored under different conditions. The results are shown in Fig. 6. Reference samples were taken from methylene chloride (pro analisi) and methylene chloride that had been glass distilled. Both samples were stored in brown glass containers in a fume hood.They both contained about 0.5 ng ml-l of phosgene. This basal level was constant for at least six weeks. From Fig. 6, it can be seen that methylene chloride stored in clear glass flasks on the laboratory bench top contain considerable amounts of phosgene after only a few days although the flasks were protected from direct sunlight. The faster formation of phosgene in distilled methylene chloride (Fig. 6) may be due to the absence or lower concentration of water which reacts with L l E 0 0 0 0 I 1 1 5 10 15 L O 0 Time/d Fig. 6. Concentration of phosgene as a function of storage time in the presence of daylight. 0 Distilled; 0 pro analisi grade not distilled with 20 p.p.m. amylene; and A pro analisi grade not distilled with an extra 200 p.p.m.of amylene added (220 p.p.m. in all). Samples were taken each morning and analysed in parallel with pro analisi grade non-distilled and distilled methylene chloride that had been protected from light. Metoprolol was used as reagent August 1983 CHLORIDE AFTER CYCLISATION AND GC WITH N SELECTIVITY 983 initially formed phosgene in the non-distilled methylene chloride. The extra addition of 200 p.p.m. of the stabiliser amylene does not prevent the formation of phosgene but after 15 d the level was lower (Fig. 6). At the higher sample concentrations of phosgene the sample volume was reduced to avoid too much consumption of the reagent by hydrogen chloride. The formation of phosgene still seems to reach a steady state (Fig. 6). These findings show that even after a short period of time on the bench top significant amounts of phosgene can be formed in methylene chloride.The actual amounts will mainly pose methodological hazards i.e. it will react with the analyte of interest. An example of such an instance has been described recently.12 The health hazards should not be neglected either as in one instance levels above 10 pg ml-l were observed. In the course of the preparation of this manuscript a similar approach for the determination of formaldehyde was pub1i~hed.l~ Gaseous formaldehyde was adsorbed on to a Chromosorb 102 column impregnated with N-benzylethanolamine. The oxazolidine formed was then desorbed and determined by capillary gas chromatography with flame-ionisation detection. I am greatly indebted to Professor Jorgen Vessman for valuable support and encouragement. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References Noweir M. H. and Pfitzer E. A. J . Am. Ind. Hyg. Assoc. 1971 32 163. Crummet W. B. and McLean J. D. Anal. Chem. 1965 37 424. Czenvinski W. and Vieweger H. Chem. Anal. (Warsaw) 1966 11 923. Basu P. K. King C. J. and Lynn S. J . Chromatogr. Sci. 1972 10 479. Baiker A. Geisser H. and Richarz W. J . Chromatogr. 1978 147 453. Kuessner A. J . Chromatogr. 1981 204 159. Preistly L. J. Jr. Critchfield F. E. Ketcham N. H. and Cavender J. D. Anal. Chem. 1965 37, Dahlberg J. A. and Kihlman I. B. Acta Chem. Scand. 1970 24 644. Jeltes R. Burghardt E. and Breman J. Br. J . Ind. Med. 1971 28 96. Esposito G. G. Lillian D. Podolak G. E. and Tuggle R. M. Anal. Chem. 1977 49 1774. Gyllenhaal O. and Vessman J. J . Chromatogr. 1983 273 129. Johansson L. and Vessman J. J . Chromatogr. 1982 239 323. Kennedy R. E. and Hill R. H. Jr. Anal. Chem. 1982 54 1739. 70. Received February 3rd 1983 Accepted February 25th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800978

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

984 Analyst August 1983 Vol. 108 pp. 984990 Simultaneous Gas-chromatographic Analysis Lower Fatty Acids, of Phenols and Indoles in Faeces and Saliva Using a Fused Silica Glass Capillary Column Yasuyuki Hoshika and Ninzo Murayama Japan Deflartment of Hygiene Shinshu University School of Medicine 3-1-1 A sahi Matsumoto-shi Nagano 390, The simultaneous gas-chromatographic separation of a mixture of 14 lower fatty acids 11 phenols and 7 indoles was performed by using a fused silica glass capillary column of Carbowax 20M (50m x 0.2mm i.d. Carbowax 2OM deactivated). Complete separation of the mixture was obtained except for the peaks of phenols and o-cresol o-ethylphenol and 3,5-xylenol and pelargonic acid 2,3-xylenol and 1,2-dirnethylindole whose peaks overlapped, and 2- and 3-methylindoles which were poorly separated.The optimum conditions are as follows column temperature held for 1 min at 100 "C; column oven heated at 4 "C min-1 from 100 to 220 "C maintained at 220 "C for 9 min for standard compounds or 29 min for sample specimens then cooled to 100 "C; and carrier gas (nitrogen) flow-rate 0.97 ml min-l. The method was applied to the analysis of the lower fatty acids phenols and indoles in cat and human faeces and non-smoker saliva. Keywords Gas chromatography; fused silica glass capillary column ; faeces analysis; lower fatty acids phenols and indoles The simultaneous analysis of lower fatty acids phenols and indoles is often required in the organic analysis of human body fluids drugs and their metabolites in biological samples foods, cigarette smoke and in environmental pollution.The acids have varied toxic effects such as an irritant effect on the skin eyes and mucous membranes,l and therefore they are of consider-able health and environmental concern. The acids are also present in the breath and serum of patients with hepatic coma2 and in the breath of patients with cirrhosis of the liver,3 are inter-mediates and end-products in bacterial fermentation and occur as metabolites of alcohols via aldehydes .4 s5 It is well known that the phenols are metabolites of benzene and its alkyl derivatives6s7 and the free and conjugated phenols are present in blood organs urine saliva sweat and faeces.* Indoles are interesting compounds as metabolites of tryptophan in cerebrospinal fluid brain, plasma and urineg and are present also in ovine plasma and rumen fluidlo and in human saliva.ll The co-occurrence of the lower fatty acids phenols and indoles has been deter-mined in cigarette smoke and its condensate12-14 and in accumulated liquid poultry manure15 and liquid swine manure.16-18 Further these compounds have low odour threshold values of below 1 p.p.b.(part per lo9) in air. Therefore the simultaneous determination of these compounds in odour pollution analysis is an interesting problem. Gas chromatography (GC) is a highly sensitive and specific technique for the determination of these compounds. Direct GC of these compounds in the free form at low concentrations has been limited by adsorption and decomposition in the column ghosting phenomena peak tailing and poor separability of many isomeric compounds.In general when packed columns are employed the simultaneous GC separation of mixtures of lower fatty acids phenols and indoles is difficult because the fatty acids have relatively low boiling-points whereas the phenols and indoles have higher boiling points and have many isomers. The simultaneous GC analysis of trace amounts of the lower fatty acids (from C2) phenols and indoles using glass capillary columns (high-resolution columns containing FFAP and PEG 20M) with a Tenax-GC pre-column has been reported previ0usly.1~ The method has been applied to the determination of lower fatty acids phenols and indoles in Japanese cigarette smoke HOSHIKA AND MURAYAMA 985 Three kinds of fused silica glass capillary columns (Carbowax 20M VFA and methylsilicone; 50 m x 0.2 mm i.d.) of high resolution were used to achieve the simultaneous GC analysis of low concentrations of the lower fatty acids phenols and indoles in normal fresh urine from a Japanese male and the peaks of typical phenols were also identified as their bromophenols, which were prepared by the reaction with bromine.20 Jellum2I has reviewed the profiling of human body fluids in healthy and diseased states using GC and mass spectrometry.In this work a fused silica glass capillary column of high resolution was used to achieve the simultaneous GC analysis of the lower fatty acids phenols and indoles in Persian cat and human faeces and human saliva. Experimental Reagents Fourteen lower fatty acids eleven phenols seven indoles and ethanol were obtained from PolyScience (Niles IL USA) Aldrich (Milwaukee WI USA) Katayama Chemical Industries, solvent 1 17 6 18 I 22 f 26 27 0 10 20 30 40 Time/min Fig.1. Typical gas chromatogram of a standard mixture of 14 lower fatty acids 11 phenols and 7 indoles using a fused silica glass capillary column (50 m x 0.2 mm i d . ) containing Carbowax 20M ; other GC conditions as under Experimental section except for FID sensitivity 2 x 102. Peaks 1, CH,COOH (156 ng) ; 2 C,H,COOH (195 ng) ; 3 iso-C3H7-COOH (221 ng) ; 4 C,H,COOH (221 ng) ; 5 iso-C4HgCOOH (260 ng) 6 C4HsCOOH (260 ng) ; 7 iso-C,H,,COOH (312 ng) ; 8 C,H,,COOH (312 ng) ; 9 2,6-C6H3(CH,),0H (650 ng) ; 10, OH (650 ng) ; 12 C,H,,COOH (1274 ng) ; 13 p-C6H,(C,HS)OH (650 ng) ; 14 2,5-C6H,(CH,),0H (650 ng) ; 15 p-C,H,(CH,)OH (659 ng) ; 16 m-C,H,(CH,)OH (650 mg) ; 17 2,3-C6H,(CH,),0H (650 ng) + 1,2-C8H5(CH3),N (189 ng) ; 18 C8H,,COOH (1 040 ng) + o-C6H4(C,H5)OH (650 ng) + 3,5-C6H,(CH3),0H (650 ng) ; 19 3,4-C,H,(CH3),0H (650 ng) ; 20 CgHlSCOOH (1 105 ng) ; 21 C,,H21COOH (1 235 ng) ; 22 C,H,N (338 ng) ; 23 3-C8H6(CH,)N (293 ng) ; 24 2-C8H6(CH3)N (280 ng) ; 25, C,,H,,COOH (1 300 ng) ; 26 5-C8H,(CH3)N (371 ng) ; 27 2,3-C,H,(CH,),N (163 ng) ; and 28 2,5-C8H5(CH3),N (234 ng).C6H&OOH (338 ng) ; 11 C6H5OH (650 ng) + O-C&(CH,) 986 HOSHIKA AND MURAYAMA GC OF FATTY ACIDS PHENOLS AND INDOLES Analyst vd. 108 (Osaka Japan) Tokyo Kasei Kogyo (Tokyo Japan) and Wako Pure Chemical Industries (Osaka Japan). All reagents used were guaranteed- or analytical reagent-grade chemicals and were used without further purification.Standard solutions of the reagents were prepared by dissolution in distilled water or ethanol to give concentrations of about 0.25 x mol per 10 ml of solvent. The volume injected into the gas chromatograph was usually 0.2-5 pl with a 10-p1 Hamilton (Reno Nevada) microsyringe (701-N) . 0.025 x loA4 or 0.0025 x Apparatus The gas chromatograph used was a Shimadzu Model GC5AP,F (dual column system) equipped with on-column injection a flame-ionisation detector (FID) and a digital integrator (Shimadzu Model ITG-2A) for the determination of retention times and quantitative analysis. A fused silica glass capillary column (WCOT-type 50 m x 0.2 mm i.d.) made by Hewlett-Packard (Avondale PA USA) containing Carbowax 20M (Carbowax 20M deactivated) was obtained from Yokogawa-Hewlett-Packard (Tokyo Japan).Gas Chromatography The fused silica glass capillary column was pre-conditioned at 200 "C for 1 h with a constant flow of nitrogen (ca. 1 ml min-l) before being connected to the FID. The chromatographic conditions were as follows stationary phase Carbowax 20M ; column temperature held for 1 min at 100 "C; column oven heated at 4 "C min-l from 100 to 220 "C, maintained a t 220 "C for 9 min for standard compounds or for 29 min for sample specimens, and then cooled to 100 "C; injection port and detector temperatures 250 "C; carrier gas (nitrogen) flow-rate 0.97 ml min-l; purge gas (nitrogen) flow-rate 40 ml min-l; splitting ratio 1 10; and air and hydrogen flow-rates for the FID 1.0 1 min-l and 50 ml min-l respect-ively.The nitrogen carrier gas was purified by using molecular sieve 5A 60-80 mesh (40 cm x 6 mm i.d. stainless-steel column) and was further purified by using 40% sodium hydroxide on Solvent 1 3 I i 21 0 10 20 30 40 50 Timelmin Fig. 2. Typical gas chromatogram of lower fatty acids phenols and indoles isolated from FID sensitivity 4 x los; the peak the ethanol concentrate of faeces of a Persian cat. numbers and other GC conditions are the same as in Fig. 1 August 1983 987 Chromosorb W AW 60-80 mesh (7 cm x 5 mm i.d. glass column which was pre-conditioned at 250 "C for 4 h with a nitrogen carrier gas flow-rate of 0.25 1 min-l). IN FAECES AND SALIVA WITH A FUSED SILICA CAPILLARY COLUMN Sample Preparation Methods The samples of fresh faeces were obtained from a Persian cat (4 years old) and a normal Japanese male (36 years old).The samples of the faeces solution were obtained by mixing the faeces (2 g) and ethanol (100 ml) vigorously and the solutions obtained were filtered. A 25-ml volume of filtrate was concentrated to 2 ml by bubbling throughout nitrogen carrier gas at a constant flow-rate (0.2 1 min-l at 25 "C). Sample concentration was required because the method was of low sensitivity and the sensitivity was limited by dilution of the sample with the solvent. The limiting volume for injection into the GC system was 5 pl; this volume of the concentrated sample filtrate was injected with a 10-p1 Hamilton microsyringe into the gas chromatograph.The sample of human saliva was obtained from a normal non-smoker (36-year-old male) and 5 p1 of the sample were injected directly. Results and Discussion Fig. 1 shows a typical gas chromatogram obtained with a standard solution of 14 lower fatty acids 11 phenols and 7 indoles by use of the Carbowax 20M fused silica glass capillary column. The time required for analysis was less than 40 min. TABLE I RECOVERIES OF 28 PEAKS FROM ETHANOL STANDARD SOLUTION AND CORRECTED CONCENTRATIONS DURING CONCENTRATION OF THE LOWER FATTY ACIDS, PHENOLS AND INDOLES IN TWO FAECES SAMPLES Peak No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 16 16 17 18 19 20 21 22 23 24 25 26 27 28 Recovery, 36.7 47.8 36.4 51.8 63.3 73.2 80.9 63.0 70.9 54.5 70.1 76.7 74.5 77.0 72.3 70.2 70.6 80.1 81.3 87.6 90.8 78.6 88.4 75.5 94.4 86.8 70.8 % 100 Concentration/pg g-lt Concentration in A I human saliva/ Persian cat faeces Human faeces ng pl-l 1607 2 034 63 784 1620 87 277 422 33 1636 523 55 1 11 267 18 511 181 56 1 1 29 90 311 260 42 7 1 29 90 * Recoveries in the concentration of the standard solution (0.48-4 ng pl-l) from 25 to 2 ml by the bubbling method with a constant flow of nitrogen carrier gas (0.2 1 min-l) at 25 "C for 8 h.t These values were corrected according to the percentage recoveries in the concentration procedure 988 HOSHIKA AND MURAYAMA GC O F FATTY ACIDS PHENOLS AND INDOLES Analyst voz.108 The separation of the lower fatty acids was complete with no tailing; however that of the phenols was incomplete especially for phenol and o-cresol for o-ethylphenol 3,5-xylenol and pelargonic acid and for 2,3-xylenol and 1 ,2-dimethylindole whose peaks overlapped. The separation of the indoles was complete except for 2- and 3-methylindoles. The separation of the peaks of +-ethylphenol,2,5-xylenol 9-cresol and m-cresol was better than that obtained previously19 using a PEG 20M G-SCOT glass capillary column (20 m x 0.28 mm i.d.). However the separation of the peaks of 2,3-xylenol and 1,2-dimethylindole and of o-ethylphenol 3,5-xylenol and pelargonic acid was poorer than that obtained with a PEG 20M G-SCOT glass capillary column. Fig. 2 shows a typical gas chromatogram of the lower fatty acids phenols and indoles isolated from the ethanol concentrate of the Persian cat faeces.At least 90 peaks were detected. The main group of compounds in the faeces sample was lower fatty acids the main component being acetic acid. The concentrations of these compounds detected are listed in Table I. The recoveries (36.4-100%) of the 28 peaks produced from the standard solution (0.484 ng pl-l) following the concentration procedure with a reduction in volume from 25 to 2 ml by bubbling nitrogen carrier gas at a constant flow-rate (0.2 1 min-l) at 25 "C for 8 h were estimated from a com-parison of the peak area with those obtained by the direct injection method of corresponding amounts of the standard solution into the gas chromatograph.In this recovery test the peaks of lauric acid and 2,5-dimethylindole were stable and therefore these peaks may be suitable as standard compounds for quantitative detection. This test is required in order to provide preliminary information on the oxidative degradation photolysis and evaporation loss of the compounds. Fig. 3 shows a typical gas chromatogram of the lower fatty acids phenols and indoles isolated from an ethanol concentrate of faeces from a human male. The main component was acetic acid. The corrected concentrations of acetic acid propionic acid and isobutyric acid were c. W 0 v) --2 \ 1 5 1 L 10 20 30 40 50 Time/min Fig. 3. Typical gas chromatogram of lower fatty acids phenols and indoles isolated from FID sensitivity 4 x lo3; the peak the ethanol concentrate of faeces of a human male.numbers and other GC conditions are the same as in Fig. 1 August 1983 IN FAECES AND SALIVA WITH A FUSED SILICA CAPILLARY COLUMN 989 1 A kl r 0 10 20 30 40 Time/min Fig. 4. Typical gas chromatograms of lower fatty acids phenols and indoles obtained from a human saliva (non-smoker) by (A) direct injection method and (B) saliva sample treated with 0.1 N NaOH solution (pH 13). FID sensitivity 4 x lo3; the peak numbers and other GC conditions are the same as in Fig. 1. double those in the Persian cat faeces and the corrected concentrations of butyric acid and valeric acid were 24-50 times higher than those in the Persian cat faeces. Indole and 3-methylindole (skatole) were minor components in these faeces samples.The odour characteristics (quality and intensity )of these faeces concentrations were faecal and very strong. Analytical data on the faeces of several animals vix. pig poultry cow horse and others have been widely reported. However there have been few reports on Persian cat faeces. Fig. 4 shows a typical gas chromatogram of the lower fatty acids phenols and indoles ob-tained from human (non-smoker) saliva by the direct injection method. As shown in chroma-togram in Fig. 4A acetic acid propionic acid and isobutyric acid were retained more than in the standard solutions using ethanol as solvent. Fig. 4B shows the data for the saliva sample treated with 0.1 N sodium hydroxide solution (pH 13). The peaks of acetic acid propionic acid and some phenols have completely dis-appeared.The main component of the compounds in the sample was propionic acid. Wursch et reported the inhibition of salivary cc-amylase production by maltitol and maltotriitol and their influence on acid production in human dental plaque. Tenovuo et aZ.23 reported that the inhibition of dental plaque acid production by the salivary lactoperoxidase antimicrobial system. However there have been few reports on the simultaneous GC analysis of the lower fatty acids phenols and indoles in human saliva. In this study the fused silica glass capillary column was employed to expand the range of individual chemicals that can be quantitated. Conclusion The simultaneous GC analysis of a mixture of lower fatty acids phenols and indoles in biological samples using a fused silica glass capillary column (Carbowax 20M deactivated, 60 m x 0.2 mm i.d.) has been demonstrated.The proposed method appears to be capable of giving rapid and good separations of these compounds in complex biological samples etc. The authors thank Dr. K. Yoshimoto Aichi Environmental Research Centre for useful suggestions 990 HOSHIKA AND MURAYAMA References Fassett D. W. in Patty F. A. Editor “Industrial Hygiene and Toxicology Volume 11 Toxi-Annison E. F. Hill K. J. and Lewis D. Biochem. J . 1957 66 629. Chen S. Mahadevan V. and Zieve L. J . Lab. Clin. Med. 1970 75 622. Lundquist F. in Trenolieres J. Editor “Alcohols and Derivatives,” Volume I. Pergamon Press, Oxford 1970 p. 95. Sardesai V. M. Editor “Biochemical and Clinical Aspects of Alcohol Metabolism,” Charles C.Thomas Springfield IL 1969 p. 319. Bakke 0. M. and Scheline R. R. Toxicol. Appl. Pharmacol. 1970 16 691. Williams R. T. in “Dextoxication Mechanisms. The Metabolism and Detoxication of Drugs, Toxic Substances and Other Organic Compounds,” Second Edition John Wiley New York 1959, p. 188. Deichmann W. B. and Keplinger M. L. in Patty F. A. Editor “Industrial Hygiene and Toxi-cology Volume 11 Toxicology,” Interscience New York 1962 p. 1363. Anderson G. M. and Purdy W. C. Anal. Chem. 1979 51 283. Norheim G. and Rygge J. J . Chromatogr. 1978 154 291. Hoshika Y. J . Chtromatogr. Sci. 1981 19 444. Hoffmann D. and Rathkamp G. Anal. Chem. 1970 42 366. Grob K. Helv. Chim. Acta 1968 51 718. Grob K. and Voellmin J. A. J . Chromatogr. Sci. 1970 8 218. Burnett W. E. Environ. Sci. Technol. 1969 3 744. Yasuhara A. and Fuwa K. Bull. Chem. SOC. Jpn. 1977 50 731. Yasuhara A. and Fuwa K. Bull. Chem. SOC. Jpn. 1977 50 3029. Yasuhara A. and Fuwa K. Agrtc Biol. Chem. 1980 44 2379. Hoshika Y. J . Chromatogr. 1977 144 181. Hoshika Y. in Jaeger H. Editor “Glass Capillary Gas - Liquid Chromatography Clinical and Jellum E. J . Chromatogr. 1977 143 427. Wursch P. Koellreutter B. and Delvedovo S. Int. J . Vitamin Nutr. Res. 1981 51 197. Tenovuo J. Mansson-Rahemtulla B. Pruitt K. M. and Arnold R. Infect. Immun. 1981 34 208. Received January 31st 1983 Accepted March 17th 1983 cology,” Interscience New York 1962 p. 1771. Pharmacological Analysis,” Marcel Dekker New York in the press. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23

ISSN:0003-2654    

DOI:10.1039/AN9830800984

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst August 1983 Vol. 108 $9. 991-996 991 Ionic Polymerisation as a Means of End-point lndicatio,n in Non-aqueous Thermometric Titrimetry Part XI." The Reaction Mechanism of lodimetric End-point Indication and an Evaluation of a Copolymerisation Indicator Reaction Edward J. Greenhow and G. Louis Jeyarajt Department of Chemistry Chelsea College University of London Manresa Road London S W3 6LX The mechanism of the reactions marking the end-point when ethyl vinyl ether is used as the indicator reagent in the thermometric titration of iodine-reactive analytes with solutions of iodine in dimethylformamide has been investigated by capillary gas chromatography - mass spectrometry. Three major constituents ICH,CH (OEt) CH,CH (OH) (OEt) I [CH,CH (OEt)]&H,CH-(OH) (OEt) and I[CH,CH(OEt)],CH,CH(OH) (OEt) have been identified, confirming the occurrence of a polymerisation process initiated by the iodonium ion.The hydroxy group in the molecules arises from the use of aqueous thiosulphate to terminate the polymerisation. An improvement in end-point sharpness using this catalytic thermometric procedure is effected by using a mixture of ethyl vinyl ether and 1,3-dioxolane as the indicator reagent instead of ethyl vinyl ether alone. The temperature rise in this copolymerisation reaction is greater than in the ethyl vinyl ether homopolymerisation and the end-point inflection is sharper. Keywords Non-aqueous thermometric iodimetry ; end-point indication ; ethyl vinyl ether ; 1,3-dioxolane ; copolymerisation In this work the mechanism of the iodine - ethyl vinyl ether reaction used to indicate the end point in non-aqueous iodimetric titrimetryl has been investigated using capillary gas chroma-tography - mass spectrometry (capillary GC - MS) to analyse the products resulting from the iodine-initiated ethyl vinyl ether reaction.A knowledge of the structure of these products should throw some light on the nature of the indicator reaction and its limitations. Of the halogens iodine is the only one that has proved useful as an initiator for cationic polymerisation. It initiates the polymerisation of vinyl ethers,2 $-meth~xystyrene,~ N -vinylcarbazole and less effectively styrene4 and acenaphthalene.5 Berthelots first reported on the use of iodine as an initiator for the polymerisation of styrene and later Wislicenus7 reported that a violent reaction occurred on addition of iodine to ethyl vinyl ether to yield a polymer of low relative molecular mass.There are differing views about the mechanism of the iodine-initiated polymerisations. Eley and Richards* showed that ions and not as reported earlier,g free radicals were the active intermediates in the polymerisations. More recently, Ledwith and SherringtonlO re-investigated the iodine-initiated polymerisation of alkyl vinyl ethers and suggested that in dichloromethane solution the active catonic initiators arise from initially formed 1,2-diiodoalkoxyethanes. On the other hand Heublein and Helbigll proposed that hydrogen triiodide is the initiator at low temperatures for these latter reactions. Evi-dence for the suggested reaction mechanisms has been obtained from kinetic and spectroscopic measurements.From spectroscopic studies it is difficult to distinguish among charge-transfer complexes diiodides and monoiodides and as recently as 1979 Janjua and Johnson12 were of the opinion that the evidence so far obtained to explain the mechanism of the iodine-initiated polymerisation of alkyl vinyl ethers was far from conclusive. Iodine is known to initiate the polymerisation of cyclic ethers as well as the substituted olefins noted above. It will for example initiate the copolymerisation of tetraoxane13 and trioxane14 with 1,3-dioxolane. However when these reactions were evaluated as possible indicator reactions in non-aqueous iodimetric titrations they were found to occur too slowly a t * For Part X see Analyst 1979 104 801.t Eresent address Department of Chemistry] Illinois State University] Normal IL 61761 USA 992 GREENHOW AND JEYARAJ IONIC POLYMERISATION FOR END-POINT Analyst VOl. 108 ambient temperature to be of practical value but it was found that an exothermic copoly-merisation reaction occurred rapidly at ambient temperature when iodine in solution in dimethylformamide was added to a mixture of 1,3-dioxolane and ethyl vinyl ether and this reaction has been investigated. Experimental Reagents Laboratory-reagent grade dimethylformamide (DMF) was distilled and dried over molecular sieve 4A before use. Ethyl vinyl ether (laboratory-reagent grade) was distilled to remove higher boiling residues before use. AnalaR-grade iodine was used as received.Dimethyl sulphoxide dichloromethane styrene 9-methoxystyrene caprolactam trioxane y-butyro-lactone and angelica lactone were of laboratory-reagent grade and were used as received. 1,S-Dioxolane was prepared from ethane-1 ,2-diol and ~araforma1dehyde.l~ Apparatus Thermometric titrimetry beakers (capacities 14 and 65 ml) instead of foam-insulated titration flasks. The automatic apparatus described in Part I I P was used but with unsilvered Dewar Capillary GC - MS used for the analysis of the ethyl vinyl ether reaction products. used to calibrate the mass spectrometer which was operated at 70 eV. column was coated with OV-1. A VG Micromass l6F mass spectrometer combined with a Pye 104 gas chromatograph was Perfluorotributylamine was The 15-m capillary Procedure Investigation of the reaction mechanism To 20 ml of DMF and 20 ml of ethyl vinyl ether in a 65-ml Dewar beaker add a 0.1 M solution of iodine in DMF from the motor-driven syringe of the titration apparatus at 0.2 ml min-l and record the temperature - titrant volume curve on the millivolt chart recorder (500-mV scale) at a chart speed of 600 mm h-l.Continue to add titrant after the end-point inflection is seen until the reaction mixture begins to bubble and then add 5 ml of 0.1 M sodium thiosulphate solution to quench the reaction and remove the excess of free iodine. Transfer the reaction mixture into a 250-ml separating funnel and separate the organic layer from the aqueous layer containing most of the DMF. Add 20 ml of dichloromethane to the organic layer and extract this layer three times with 50 ml of distilled water to remove substantially all of the remaining DMF.Samples of the washed organic layer and of this layer diluted further with dichloromethane to concentrations appropriate for capillary columns are analysed by direct injection into the column. The temperature of the column is maintained at 50 "C during the injection and then increased to 250 "C at 20 "C min-l. Capillary GC - M S . Evahation of the ethyl vinyl ether - 1,3-dioxolane mixed indicator Prepare a 0.1 M solution of isoquinoline in DMF and pipette 1 ml of this into the 14-ml titration beaker. Add appropriate amounts of ethyl vinyl ether and 1,3-dioxolane and titrate the mixture with 0.1 M iodine in DMF. Add the titrant at 0.17 ml min-l and record the temperature and titrant volume as in Investigation of the reaction mechanism.Results and Discussion Fig. 1 shows the gas chromatogram obtained from a 0.25-p1 amount of the washed dichloro-methane extract of the ethyl vinyl ether reaction product further diluted 1 + 1000withdichloro-methane. The peaks shown in the chromatogram are those obtained after the solvent and unreacted monomer had been eluted from the column. The four major peaks A. B C and D, were analysed on-line in the mass spectrometer by irradiation at 70 eV. Fragments with m/e values of 127,103 75,73,47 and 45,73 being the base peak were observed in the mass spectr August 1983 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART XI 993 45 47 Fig. 1. Capillary gas chromatography of the product formed in the iodine-catalysed reaction of ethyl vinyl ether in dimethylformamide.The four major peaks (A B C D) were analysed on-line in the mass spectrometer by irradiation a t 70 eV. 103 75 143 171 187 233 I . 100 50 0 50 100 150 0 50 100 Mass to charge ratio 150 200 250 103 5 100 150 200 250 Mass to charge ratio Fig. 2. product. Mass spectra of fractions from the capillary gas-chromatographic separation of the reaction (a) Fraction A; (b) fraction B; and (c) fraction D. of A B find D (Fig. 2). The m/e of 127 indicates that iodine is a constituent of these three chromatographic fractions and this was confirmed as a constituent element in the original dichloromethane extract by oxygen flask combustion.The fragmentation pattern of fraction C in the chromatogram showed it to be completely different from fractions A B and D in stmctural type and to have a base peak m/e value of 99. This fraction has not been identified. Making the assumption that A B and D are oligomers of ethyl vinyl ether a detailed considera-tion of the possible fragmentation and rearrangement processes suggests that the observed m/e values 103 75 73 47 and 45 correspond to the following fragments 994 GREENHOW AND JEYARAJ IONIC POLYMERISATION FOR END-POINT Azalyst VoZ. 108 + + + + HC //OH CH3CH=OC2H5 &OH CH ,CH = OH + \OH HC 4OCZH5 \OC,H \OC,H, m/e 103 mle 75 mle 73 mje 47 mle 45 and the following structures are proposed for A B and D on the basis of the interpretation of the mass spectra: /OH /OH OC2H5 0C2H5 CHCH CH 1 c H ~ ~ ~ 7 ~ H \ 0 c 2 H 5 /OH \OC,H 3 \OC,H, ICH,CHCH,CH I A B D Molecular ion peaks corresponding to the structures proposed for A B and D were not observed ; this is not unexpected in the mass spectra of aliphatic halides especially iodine compounds.17 Loss of iodine from the parent ions of the proposed structures for A and B would result in frag-ments with m/e values of 161 and 233 respectively and these are seen in Fig.2(n) and ( b ) . The highest observed m/e value 259 in the mass spectrum of D [Fig. 2(c)] is believed to result from the loss of both iodine and ethanol from the parent ion. Formation of A B and D would be expected to occur by the following polymerisation processes : Initiation + 1s-I,- + CH,=CH(OC,H,) -+ ICH,CH(OC,H,) 13- .. Propagation : + CH2= CH(e1ectrode + ICH,CH 13- > ICH,CHCH,CH 13-I I \ OC2H5 OC2H5 2CH2 = CH(OC,H,) A' - - (2) 3CH2 =CH(OC2H,) C' B' Termination of the reaction chains shown in equation (2) occurs when the aqueous sodium thiosulphate solution is added. Siggial* has shown that iodine combines with ethyl vinyl ether in the presence of an alcohol according to the following equation : CH2=CH(OC2H5) + I + ROH + ICH,CHOR + HI I . . - * (3) 0C2H5 Therefore on addition of aqueous thiosulphate to the reaction mixture it might be expected that water would undergo a similar reaction with the intermediates A' B' and D' to yield A B and D (above) while the thiosulphate reduces the 13- to 31-: ICH,CH-CH,CHOH + HI + 31- aq.Na2S,0 ICH,CH-CH,ZH 13-' - (4) I I I I OC2H5 OC2H5 OC2H5 OC2H5 A' August 1983 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART XI 995 Further polymerisation according to the propagation reaction (2) would of course lead to the formation of polymers of higher relative molecular mass but these would not be sufficiently volatile to determine by GC. In catalytic thermometric titrimetry it is preferable that the titrant - catalyst be present in the titration solution as “free” ions as distinct from ion pairs such as 1+13- in order to achieve rapid indicator reactions and therefore sharp end-point inflections. In this work an evaluation of solvent systems has shown that the ethyl vinyl ether reaction occurs rapidly in the presence of polar solvents such as DMF and dimethyl sulphoxide but slowly in the presence of non-polar solvents such as toluene.Iodine will yield “free” iodonium ions in the polar solvents but exists mainly as molecular iodine in toluene and it is clear that the indicator reaction is initiated more effectively by iodonium ions than by molecular iodine. I t is probable that the cation initiating the polymerisation in the polar solvent is a solvated iodonium ion formed for example with DMF by the following reaction: I+ . . I 21 + HCON(CH,) + HCON (CH3)2 13- . . The results confirm that the indicator reaction in DMF is a cationic polymerisation initiated by the iodonium ion. Copolymerisation of ethyl vinyl ether with lJ3-dioxolane As noted in the introduction at ambient temperature iodine-catalysed polymerisation of mixtures of 1,3-dioxolane with trioxane and tetraoxane occurs too slowly to be of interest as the basis of indicator reactions in catalytic thermometric titrimetry.Mixtures of ethyl vinyl ether with angelica lactone butyrolactone caprolactam p-methoxystyrene and trioxane were similarly ineffective for this purpose but the mixture of ethyl vinyl ether with 1,3-dioxolane did undergo an exothermic reaction on addition of the solution of iodine in DMF. This copolymerisation of dioxolane with ethyl vinyl ether has been compared systematically with the homopolymerisation of ethyl vin7electrodes an indicator reaction. Blank titrations have been carried out to study the effect on end-point sharpness of changing the ratio of ethyl vinyl ether to 1,3-dioxolane in the absence [Fig.3(a)] and presence [Fig. 3(b)] of DMF. In Fig. 3(a) addition of 1,3-dioxolane to ethyl vinyl ether is seen to improve the end-point sharpness in the subsequent titration with iodine although the blank titration volume is increased significantly. A similar effect is shown in Fig. 3(b) when a constant volume of DMF is included in the reagent mixture. In both parts of the figure it can be seen that end-point sharpness increases with increase in content of 1,3-dioxolane over the range of concentrations considered. In the absence of 1,3-dioxolane the effect of changing the ethyl vinyl ether DMF ratio from 1 1 to 2 1 [liig. 3(b) A-C] can be seen to be small but the end-point sharpness shows a slight decrease with the increase in content of ethyl vinyl ether.Volume of 0.1 M iodine reagent/ml 4 Fig. 3. (a) Effect of ethyl vinyl ether 1,3-dioxolane ratio on end-point sharpness in the blank titration with iodine. The ratios of ethyl vinyl ether to 1,3-dioxolane (ml) were as follows (A) 1 1; (B) 1.5:0.5; and (C) 2 O . (b) Effect of dimethylformamide ethyl vinyl ether 1,3-dioxolane ratio on end-point sharpness in the blank titration with iodine. The ratios of DMF ethyl vinyl ether 1,3-dioxolane (ml) were as follows (A) 1 1 O ; (B) 1 1.5:O; (C) 1 2 0 ; (D) 1 l l ; and (E) 1:1.5:0.5 996 GREENHOW AND JEYARAJ LLJJ) Volume of 0.1 M iodine reagenvml Fig. 4. Titration of isoquinoline for the compari-son of ethyl vinyl ether and ethyl vinyl ether - 1,3-dioxolane as indicator reagents. Titrant 0.1 M iodine in acrylonitrile; titrand (A) 0.5 2,0,0; (B) 0.5, 1.6 0.5 0; (C) 0.5 2,0 7.5; (D) 0.26 1.5 0.5 3.8; and (E) 0.5 1.5 0.5 7.5 of DMF ethylvinyl ether 1,3-dioxolane (all ml) and isoquinoline (mg) respectively.In Fig. 4 ethyl vinyl ether is compared with the mixed indicator reagent in the titration of isoquinoline. The end-point is sharper when the latter system is employed although with this system there is a greater increase in temperature before the end-point inflection is seen. The main advantage of employing the mixed indicator is in the ease in which the end-point can be located from the more clearly defined inflection. In the titration of the isoquinoline acrylo-nitrile has been used instead of DMF as the titrant solvent because its use results in lower values for the blank titration volume.The calculated reaction stoicheiometries about 4 atoms of iodine per molecule of isoquino-line are in line with values obtained for nitrogen bases in the earlier paper on thermometric iodimetric titrimetry-l The reaction mechanism for the copolymerisation would be expected to be similar to that for the hoinopolymerisation of ethyl vinyl ether. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Greenhow E. J. and Spencer L. E. Analyst 1974 99 82. Eley D. D. and Saunders J. J . Chem. SOC. 1952 4167. Okamura S. Kanoh N. and Higashimura T. Makromol. Chem. 1961 47 19. Trifan D. S. and Bartlett P. D. J . Am. Chem. SOC. 1959 81 5573. Giusti P. Puce G. and Andruzzi F. Makromol. Chem. 1966 98 170. Berthelot M. Bull. SOC. Chim. Fr. 1866 2 6294. Wislicenus J. Justus Liebigs Ann. Chem. 1878 192 113. Eley D. D. and Richards A. W. Trans. Faraday SOC. 1949 45 436. Chalmers W. Can. J . Res. 1936 7 464. Ledwith A. and Sherrington D. C. Polymer 1971 12 344. Heublein G. and Helbig M. J . Prakt. Chem. 1972 314 1. Janjua K. M. and Johnson A. F. J . Sci. Ind. Res. 1979 23 109. Morita Y. Ishigahi I. Kumakura M. Watanabe Y. and Ito A. J . Appl. Polym. Sci. 1979 23, Panaitov I. and Dimitov I. Izv. Otd. Khim. Nauk Bulg. Akad. Nauk 1969 2 87. Greenhow E. J. personal communication. Greenhow E. J. and Spencer L. E. Analyst 1973 98 98. Budzikiewicz H. Djerassi C. and Williams D. H. “Mass Spectrometry of Organic Compounds,” Holden-Day San Francisco 1967. Siggia S. Anal. Chem. 1948 20 762. 3395. NOTE-References 1 and 16 are to Parts V and I11 of this series respectively. Received February 2nd 1983 Accepted March 15th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800991

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst August 1983 Vol. 108 pp. 997-1002 997 Voltammetric Studies of Zomepirac Sodium and its Determination in Tablets by Differential-pulse Po I a rog rap hy Leslie G. Chatten Stanley Pons* and Lawrence Amankwa Faculty of Pharmacy and Pharmaceutical Sciences University of Alberta Edmonton Alberta T6G 2N8, Canada A simple diff erential-pulse polarographic method has been developed for the determination of zomepirac sodium in tablets. Britton - Robinson buffer of pH 11 .O containing 5% V/V of methanol was employed as the supporting electrolyte. Results obtained by the proposed method are in excellent agreement with those provided by both the official and the manufacturer's methods of assay. It was found that commonly used tablet excipients did not interfere in the analyses.The peak potential occurs at - 1.50 V. Keywords Zomepirac sodium determination ; diflerential-fiulse flolarografihy ; controlled-potential coulometry ; cyclic voltammetry ; tablets Zomepirac sodium [the sodium salt of 5-(4-chlorobenzoyl)-1,4-dimethylpyrrol-2-ylethanoic acid] (I) is a non-steroidal anti-inflammatory agent approved exclusively as an analgesic in the treatment of mild and moderately severe pain1 and is formulated as a 100-mg tablet. The official method of analysis in the US Pharmacopeia2 involves three extraction and two centrifugation steps prior to the final ultraviolet spectrophotometric analysis at 329 nm. The method is tedious and requires very careful sample handling which is not usually amenable to routine analysis. The manufacturer's method3 of assay is similar to that of the US Pharma-copeia.2 Other methods of assay include high-performance liquid chromatography for the determina-tion of the drug in blood plasma4 and in purity and stability tests.5 In addition a non-aqueous titrimetric method is employed by the US Pharmacopeia2 for the determination of zomepirac sodium as the raw material.Thin-layer chromatography2 and radioimmunoassays are other methods reported for the determination of this substance. In this report differential-pulse (d.p.) polargraphy has been applied to the determination of zomepirac sodium and involves the reduction of the conjugated carbonyl group of the benzoyl substituent. Only a single extraction is required prior to the polarographic analysis and none of the excipients present in the tablet interfere with the analysis.The d.p. polarographic peak height varies linearly with the concentration of the drug over the range 1 x 10-6-5 x The method is accurate sensitive and easy to apply for routine analysis. M. Experimental Apparatus and Conditions for Polarographic Analysis A Fisher Model 320 pH meter fitted with a glass - calomel electrode was employed to measure the pH values of the solutions. A PAR Model 174 polarographic analyser equipped with a drop timer (Model 172A) and a Houston Omnigraphic Model 2000 recorder were used in the analyses. A three-electrode combination was employed which consisted of a saturated calomel electrode a dropping-*Department of Chemistry 998 CHATTEN et d. VOLTAMMETRY OF ZOMEPIRAC SODIUM Analyst Vd.108 mercury electrode and a platinum wire as the auxiliary electrode. A conventional H-type cell was maintained at 25 & 1 "C and all sweeps utilised a scan rate of 2 mV s-l and a drop time of 2 s. In Britton - Robinson buffer (pH 11.0) the instrumental parameters were as follows: applied potential range -1.0 to -2.5 V; current 50 pA full-scale; height of mercury column, 75 cm; flow-rate of mercury 1.265 mg s-l; modulation amplitude set at 50 mV; and low-pass filter set at a time constant of 1 s. The instrument was operated in the differential-pulse mode. Controlled-potential Coulometry A PAR Model 173 potentiostat - galvanostat equipped with a PAR Model 377A three-electrode coulometric cell system was connected to a Hi-Tek digital integrator and digital voltmeter.A 19-ml volume of Britton - Robinson buffer pH 11.0 was placed in the coulometric cell on top of a 5-ml layer of triply distilled mercury and 1 ml of a 1 x 10-2 M solution of zomepirac sodium in methanol was added. The system was purged for 10 min with purified nitrogen. The applied potential was set at -1.8 V with a current range of 10 pA full-scale and the solution was electrolysed until the digital read-out indicated a constant but small count. The electrolysis was completed in 1 h. The process was repeated with a reagent blank consisting of 19 ml of Britton - Robinson buffer and 1 ml of methanol. Cyclic Voltammetry Cyclic voltammetric experiments at a hanging mercury drop electrode were performed with a four-component system consisting of a PAR EG & G Model 175 Universal Programmer a PAR Model 173 potentiostat - galvanostat a Houston Model 2000 Omnigraphic recorder and a PAR Model 9323 hanging mercury drop electrode fitted with a polarographic cell.Two supporting electrolyte systems were employed. In Britton - Robinson buffer pH 11.0, the parameters were as follows potential range -1.1 to -1.65 V; current range 10 pA; and the scan rate varied from 10 to 200 mV s-l. In a dimethylformamide - lithium perchlorate system the following settings were used : potential range -0.4 to -2.2 V; and the current range and scan rates were the same as in the previous system. Reagents The following reagents were used all of analytical-reagent grade barbital boric acid citric acid lithium perchlorate potassium dihydrogen phosphate dimethylformamide (DMF) , anhydrous methanol 0.2 N sodium hydroxide and 1 yo lithium perchlorate in DMF.Britton -Robinson buffers were prepared with distilled de-ionised water at intervals of 0.2-0.3 pH unit over the pH range 6.0-11.0. The reference standard zomepirac sodium (99.44%) was obtained from McNeil Laboratories (Canada) Ltd. and used without further purification. pH Dependence Studies These studies were carried out in Britton - Robinson buffer over the pH range 6.0-11.0. Preparation of Calibration Graphs Five test solutions of varying concentrations 1-5 x 10-4 M were prepared by appropriately diluting the stock solution with Britton-Robinson buffer pH 11.0. In the total sample volume of exactly 20 ml the amount of methanol was always maintained at 1 ml.All samples were purged with oxygen-free nitrogen for 10 min prior to each run and a stream of nitrogen was allowed to flow gently over the surface of the solution during the electro-reduction. Samples of each of five concentratons were run five times and resulted in a correla-tion coefficient for the graph of 0.9992. A stock solution of zomepirac sodium ( M) was prepared in anhydrous methanol August 1983 AND ITS DETERMINATION I N TABLETS BY DPP 999 Diffusion Dependence Studies M solution of zomepirac sodium. The applied potential was from -1.0 to -2.5 V and the height of the mercury column ranged from 60 to 80 cm. The mass of mercury was also obtained at each of the five heights over that range. These studies were carried out at pH 11.0 on a 5 x Analysis of Pharmaceutical Dosage Form Only one dosage form 100-mg tablets was available from the manufacturer.Twenty tablets were weighed finely powdered and an amount of powder corresponding to the mass of one tablet was accurately weighed into a 100-ml beaker. A 20-ml volume of methanol was added and the sample was stirred magnetically for 15 min. The mixture was transferred quantitatively into a 50-ml calibrated flask diluted to volume with methanol and then filtered through Whatman No. 1 filter-paper discarding the first 5 ml of the filtrate. A 1-ml volume of the filtrate was trasferred into the polarographic cell and 19 ml of Britton -Robinson buffer pH 11.0 were added. As previously described the solution was purged for 10 min with nitrogen prior to recording the polarogram.The amount of zomepirac sodium in the form of the free acid was calculated by the direct comparison method using a reference standard solution of zomepirac sodium (0.7649 x M) . Content Uniformity Test Each tablet was placed in an inde-pendent 150-ml beaker 20 ml of methanol were added and the system was allowed to stand for 5 min in order to promote disintegration of the tablets. The remaining larger lumps of tablet mass were crushed with a glass rod and the mixture was stirred magnetically for 20 min. After transferring the mixture quantitatively into a 50-ml calibrated flask the determination was continued as described in the previous section. Ten tablets were randomly selected from the sample. Macro-scale electrolysis The procedure used was similar to that for the controlled-potential coulometry except that the cell contained 200 mg of zomepirac sodium in 25 ml of 20% V/V methanol in 1 N sodium chloride solution.The applied potential remained a t -1.8 V while the reduction time was 8 h. The initial yellow colour of the solution turned colourless upon completion of the reduction and then the product, together with the supporting electrolyte was separated from the mercury and freeze-dried. The pH was adjusted to 11.0 with sodium hydroxide solution. Results and Discussion Zomepirac sodium exhibits three d.c. and d.p. polarographic waves in Britton - Robinson buffer over the pH range 6.0-1 1 .O [Fig. 1 (a) and ( b ) ] . At pH 6.0 a single well resolved wave is observed with a half-wave potential Ei of -1.28 V.This wave is pH sensitive and the E4 moves cathodically with increasing pH. The second wave appears at more negative potentials within the pH range 6.8-10.0 as illustrated by B in Fig. l ( a ) . The Et value of this wave occurs at -1.45 V and is independent of pH. At pH 11.0 the two waves shown in B merge to form one wave C with an Eh value at -1.5V. This latter wave is well resolved intense and is suitable for quantitative work. For reasons that will be discussed in detail wave A is attributed to the reduction of the keto group of the benzoyl substituent by a one-electron process to give a free-radical anion. In acidic medium the radical anion becomes pronated to give a free radical which undergoes dimerisation to the pinacol.At higher pHs, the radical anion undergoes a further one-electron reduction to the dianion which subse-quently becomes pronated to give the alcohol. This second reduction is exhibited by the second wave in B. The graph of diffusion current veystxs the square root of the corrected height of mercury column is a straight line with a slope of 0.181 8 pA cm4 and does not pass through the origin. Only one reduction peak was observed in the cyclic voltammetric experiment in Britton -The corresponding d.p. waves are presented in Fig. 1 (b) 1000 CHATTEN et a,?. VOLTAMMETRY OF ZOMEPIRAC SODIUM Analyst VoZ. 108 I I I I I I -1.1 -1.2 -1.3 -1.4 -1.5 -1.6 -1.7 -1.1 -1.2 -1.3 -1.4 -1.5 -1.6 -1.7 Applied potentialN Fig. 1. Effect of pH on (a) the sampled d.c. and (b) d.p.polarographic waves of zomepirac sodium (5 x 10-4 M) in Britton - Robinson buffer. A pH 6.0; B pH 8.0; C pH 11.0. Robinson buffer at pH 11.0 [Fig. 2(a)] where the potential peak (Ep) occurred at -1.53 V. No reverse anodic peak was observed under the sweep-rate range studied. Detailed analysis of the cathodic peak indicated a peak potential shift of 40-50 mV as the sweep rate V was varied from 200 to 10 mV s-l. The peak current varied linearly with the square root of the scan rate over this range indicating the absence of any complica.ting homogeneous reaction for the time scale of the experiment.' A plot of i/<* versm v (where i is the current) is independent of scan rate implying a predominantly diff usion-controlled process. The cyclic voltammogram of zomepirac sodium in DMF - lithium perchlorate shows three cathodic peaks and one reverse anodic peak.The E, waves occurred at - 1.75 - 1.85 and -2.08 V. The E was a t -1.98 V [Fig. 2 ( b ) ] . The coulometric analysis of zomepirac sodium indicated that two electrons per molecule were involved in the electro-reduction process. The ultraviolet absorption spectra of zomepirac sodium in an alkaline medium shows two maxima. The first exhibited a A,. at 260 nm and a loge where e is the molar absorptivity, of 3.94 while the A,. of the second was at 328 nm with loge = 4.04. The reduced product shows only one maximum at 228 nm with loge = 3.97. The concentration of the reduced product was based on the initial concentration of zomepirac sodium taken. The reduced product is unstable in acid and decomposes rapidly to give a red - pink precipitate.The height of the peak at -2.08 V was twice that of both the first two peaks. I I I I 1 1 I I I I I -1.2 -1.3 -1.4 -1.5 -1.6 -1.7 -1.6 -1.7 -1.8 -1.9 -2.0 -2.1 -2.2 Applied potentialN Fig. 2. Cyclic voltammogram of zomepirac sodium (5 x M) in (a) Britton -Robinson buffer pH 11.0 and (b) DMF - LiClO, August 1983 AND ITS DETERMINATION IN TABLETS BY DPP 1001 IR and NMR spectra were obtained on samples of the freeze-dried product from the macro-scale electrolysis. In acetone-d - D,O zomepirac sodium exhibits five peaks on the NMR (60 MHz) with delta values at S 1.6 (s 3H) 3.4 (s 2H) 3.55 (s 3H) 5.9 (s 1H) and 7.3 (quartet 4H). In the same solvent the reduced product exhibits six peaks on the NMR (200 MHz) with delta values at S 2.5 (s 3H) 3.8 ( s 3H) 3.95 (d 2H) 6.3 (s lH) 6.6 (s 1H) and 7.95 (b 4H).The IR (potassium bromide) spectrum of the reduced product of zomepirac sodium shows an absence of the carbonyl peak at about 1660-1 690 nm which is present in the IR spectrum of zomepirac sodium itself. The absence of the carbonyl group in the reduced product is also indicated by the disappearance of an ultraviolet absorption peak in the range 250-350 nm. The NMR peak at 6 6.6 due to a single proton observed for the product confirms that the carbonyl group is the site of the electro-reduction process. From the results obtained we propose the following pathway for the reduction : e- COO- Na' 2H+ . COO- Na' - CI Me The polarogram of deschlorozomepirac sodium was identical with that for the parent zomepirac which indicates that the 9-chloro atom does not participate in the electro-reduction, Table I gives the results of the assay of zomepirac sodium tablets based on the average value obtained with a sample of 20 tablets.The standard deviations of the method together with the analysis obtained by the manufacturer's quality control laboratory are presented for comparative purposes. The assay provides results that are equally comparable to those of the manufacturer at both pH values of 11.0 and 6.0. TABLE I ANALYSIS OF ZOMEPIRAC SODIUM TABLETS AT TWO pH VALUES The labels for the tablets claimed that they contained 100 mg of zomepirac sodium. Manufacturer's Diff erential-pulse method pH 6.0 pH 11.0 result Amount found (a = 20) yo .. . . 96.6 f 0.2 98.7 f 0.4 97.64 Although the d.c. wave could have been used for analytical purposes the d.p. wave is much better resolved as illustrated in Fig. 1 (a) and (b). The proposed method has the advantage of simplicity high sensitivity and rapidity. It has the same disadvantage as the official spectro-photometric method however in that certain degradation products if present might not be differentiated. Muschek and Grindel6 have reported on the pharmacokinetics and metabolism of zomepirac and they noted that the major metabolites were the glucuronide hydroxyzome-pirac and 4-chlorobenzoic acid. Only the last compound would not interfere with the analysis by the proposed method. The average of the results for determination of the average number of milligrams of zome-pirac sodium per tablet obtained with ten individual tablets is as follows d.p.polarographic method 100.7 & 1.0 mg; manufacturer's result 99.15 mg. The label claimed 100 mg 1002 CHATTEN PONS AND AMANKWA We gratefully acknowledge the samples of zomepirac sodium and its tablets as well as the deschlorozomepirac sodium which we received from McNeil Laboratories (Canada) Ltd. In addition we thank B. Speiser Department of Chemistry and E. Hall Faculty of Pharmacy and Pharmaceutical Sciences both of this University for their helpful suggestions and assist-ance with the reduction products. References 1. 2. 3. 4. 6. 6. 7. McEvoy G. K. Am. J . Hosp. Pharm. 1981 38 1293. “United States Pharmacopeia,” Twentieth Revision Mack Easton PA 1980 Third Supplement, McNeil Laboratories (Canada) Ltd. personal communication 1982. Ng K.-T. and Snyderman T. J. Chromatogr. 1979 178 241. KO C. Y. and Janicki C. A. J. Chromatogr. 1980 190 429. Muschek L. D. and Grindel J . M. J. Clin. Pharmacol. 1980 20 223. Chatten L. G. Boyce M. Moskalyk R. E. Pons S. and Madan D. K. Analyst 1981 106 366. p. 291. Received January 31st 1983 Accepted March lst 198

ISSN:0003-2654    

DOI:10.1039/AN9830800997

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst, August, 1983, Vol. 108, pp. 1003-1006 1003 Transient Potential Shifts with pH Glass Electrodes Due to Divalent Cations Colin D. Kennedy Department of Biochemistry, Kent, TN25 SAH Physiology and Soil Science, College, ( University of London), Wye, A shford, The introduction of low concentrations of Cu2f, Ni2+, Co2+, Mn2+, Zn2+, Cd2+ and Hg2+ salts into unbuffered aqueous solutions gave rise to transient potential shifts with glass pH electrodes. These shifts mimic transitory pH decreases in solution. Removal of the added cation gave rise to a transient potential shift in the opposite direction. The relative order of magnitude due to the added cations was studied. The quasi-pH shifts were shown to be particularly serious in flow cells where the vigour of mixing was less and could easily lead to the impression that an actual pH transient had occurred.The effects were shown to be due to a surface action at the glass membrane. A possible mechanism is suggested. Keywords : pH glass electrodes ; divalent cations ; flow cell; transient potentials During the course of investigations into the action of certain divalent cations on ion uptake and release by plant roots,l it was discovered that some cations gave rise to an apparent transient lowering of pH. As these effects could arise whenever these ions are introduced during the course of pH measure- ments, they could easily lead to erroneous conclusions regarding pH shifts. This is particu- larly so in situations where stirring is minimal. For this reason, these phenomena were investi- gated further and the results are reported.There are few previous reports of transient interference at glass electrodes but it has been shown2 that a Beckman cation-selective electrode in a flow cell gives a transient potential decrease of about 10 mV on lowering the Ca2+ concentration from 0.0538 to 0.0100 M. The other electrolyte present was potassium chloride M) and the pH was adjusted to 9.0 in order to minimise the effect of H+ ions. A similar response was shown by Sr2+. The direction and shape of the transient was similar to those reported here. In this work Ca2+ M) was included in all solutions as these ions are normally present in biological bathing solutions. As the monitoring of small pH changes is frequently associated with biological experiments this was considered to be justified.It is also known that molybdate, tungstate, vanadate, phosphate and arsenate can cause a drift to higher potentials with pH glass electrodes in buffered ~olutions.~ This obscured the short term monitoring of H+ efflux from the roots, The molar ratio of Ca2+ to test ion was 5 : 1. Experimental The apparatus consisted of a calibrated glass reservoir of 1 1 capacity from which solution was pumped peristaltically, via fine-bore silicone-rubber tubing, to twin flow cells in parallel. The reservoir solution was stirred continuously by a magnetic bar and the pH of the reservoir solution monitored by means of a combination electrode (Corning, Model 0031 1 201 N). Each flow cell consisted of a vertical polythene tube with an inner bore of length 16 mm and dia- meter 6 mm.The flow solution entered through a fine hole in the base. The tube widened to form a cup at the top from which solution was taken to waste by suction. A Beckman 5-mm combination electrode (No. 39505) was inserted into each cell so that the tip of the glass bulb was just clear of the bottom of the cell. All flow solutions contained potassium chloride (1.00 mM) and calcium sulphate (0.10 mM), the pH being adjusted by the dropwise addition of hydrochloric acid or potassium hydroxide solution where required. Solutions of zinc sulphate, cadmium nitrate, mercury( 11) chloride, copper(I1) sulphate, nickel(I1) sulphate, cobalt(I1) nitrate and manganese(I1) sulphate were prepared at a concentration of M and adjusted to within 0.1 pH unit of the pH of the flow solution.The test solutions were added to the flow solution reservoir and the apparent pH1004 KENNEDY: TRANSIENT POTENTIAL SHIFTS WITH Analyst, VoZ. 108 changes sensed by the combination electrodes in the reservoir and flow cells. An Orion four- channel variable-speed low-pulse peristaltic pump (Sage Instruments) was used to pump solution through the flow cells at 0.014 ml s-I, the time for solution to travel from reservoir to pH probe being about 60 s. The outputs from the three pH probes were led to a Model 33B-2 Vibron electrometer, an Orion Research, Model 901 , microprocessor ionalyser and a Beckman, Model 4500, pH meter. Outputs from these were taken to three pens of a chart recorder. The reservoir was calibrated for volume, and 600-800 ml of the solution were present on addition of the test compounds.Twin flow cells were required in order to com- pare the transients induced by the same test solution for two nominally identical electrodes. The geometries of the two flow cells were as identical as could be arranged. All solutions were at 25 "C. Results The addition of low concentrations (usually 2 x M) of salts of Mn2+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+ and Hg2+ to the well stirred reservoir solution containing potassium chloride (1.00 mM) and calcium sulphate (0.10 mM) gave rise to an immediate apparent fall in pH which rapidly returned to near the original value (Fig. 1). If the electrodes in the reservoir were transferred to a similar solution at the same pH but containing no added cation, a rapid transient peak was observed. Similar but much more prolonged transients were found for the combination electrodes in the flow cells (Fig.1). The magnitude and especially the recovery time of the transients varied between the nominally identical electrodes in the twin flow cells. However, in order to show the extent of this interference, the range of magnitudes and dura- tions obtained are given in Table I. The pH of the reservoir solution was about 5.7. In order to discover whether the potential changes indicated a real but transient pH change in the bulk solution, the metal salt was first added to a separate but identical reservoir solution to that being pumped through the flow cell. After about 15 min, when the transient potential Ti me/m in Fig.1. solution. 1, (A) and (B) transferred into 2 x added; (I) probe transferred into 2 x 10-5 M NiSO,; (J) probe transferred into 2 x (L) and (M) are a repeat of (I), (J) and (I). CusO,; (P) solution without CuSO,; (9) 2 x 10-5 M ZnSO,. without ZnSO,; (T) 2 x 10-5 M CdSO,; (U) solution without CdSO,. (m3 M) and CaSO, (lo-* M). Quasi-pH transients due to added divalent cations. 1 and 2, transients in stirred reservoir M CuSO,; (D) probe M ZnSO, M ZnSO,; (K), M M ZnSO,; (S) solution All solutions contained KC1 M ZnSO, added; (C) probe transferred into 2 x M ZnSO,; (E) and (F) are (C) and (D) repeated. 2, (G) and (H) 3 and 4, transients observed in flow cell. 3, (0) 2 x 4, (R) 2 xAugust, 1983 Ion (2 x 1 0 - 5 ~ ) CU2+ . . .. Zn2+ . . . . Cd2+ . , .. Ni2+ .. .. co2+ . . .. Mn2+ . . .. Hg2+ . . .. PH GLASS ELECTRODES DUE TO DIVALENT CATIONS TABLE I APPARENT pH TRANSIENTS DUE TO ADDED DIVALENT CATIONS Stirred reservoir r u pH fall 0.020-0.045 0.030 0.020 0.010-0.025 0.020 0.020 0.080-0.189 Recovery timels 30-300 20-60 50-60 10-60 40 40 40 7 No. of expts. 6 7 2 1 4 1 1 pH fall 0.22-0.29 0.034-0.067 0.069-0.087 0.079-0.120 0.028-0.044 0.026-0.036 0.025-0.028 Flow cell 1005 Recovery time/ min 18-35 6-20 8-20 20-25 4-15 6 6-15 1 No. of expts. 0 6 4 2 4 2 2 change registered by the pH probe in the separate reservoir had completely disappeared, the solution containing the added cation was pumped through the flow cells in place of the “refer- ence” solution. Similar transients to those observed on adding the test cation directly to the pumped reservoir solution occurred, indicating that the phenomenon is associated with the surface of the probe and not the bulk solution.Identical results were obtained when the built-in reference half-cells of the combination pH electrodes were replaced with a silver - silver chloride reference electrode making contact with the solution via a fine glass capillary containing 3 M potassium chloride solution. Hence the effects are surface phenomena associated with the H+-selective glass membrane and not with the combination cell reference electrode junction. The relative potential transients of the different metal ions were compared in two ways. Firstly, the test compounds were added to the reservoir solution, the transients recorded on the three pH probes and then the reservoir solution was changed for a similar solution without the test compound and the reverse transients (“pH” peaks) were recorded. The same procedure was then repeated for the other metal ions being tested. In the second method, the reservoir solution containing the test salt (2 x M) was replaced with a similar solution at as near as possible the came pH and containing the same concentration of a different test cation.The potential changes were recorded. It was not always possible to keep these unbuffered solutions at exactly the same pH, but the potential transients could easily be seen superimposed on a small permanent pH change. Using this method it was possible to show, for example, that a change from zinc sulphate (2 x 1 0 - 5 ~ ) to copper(I1) sulphate (2 x 1 0 - 5 ~ ) always produced an apparent downwards “pH” transient, while the reverse procedure always gave a peak.Fig. 1 shows typical examples of the two procedures for both reservoir and flow cells. Similar experiments at pH 4.5 gave the same type of transients, but they were of lower magnitude . By far the largest potential transients were produced by the copper salt. The order of effects was Cu2+ >> Zn2+ M Cd2+ M Hg2+ > Ni2+ M Co2+ m Mn2+. Iron(I1) sulphate was also tested but the rapid oxidation to iron(TI1) in air-saturated solutions made the results inconclusive. However, downwards “pH” transients were ob- served. Experiments with zinc and copper salts were also performed in 0 . 2 5 ~ sodium ethanoate - ethanoic acid buffer at pH 5.5.No potential transients were observed, possibly owing to the complexing of the cations. Discussion The type of interference described above could occur whenever a glass membrane pH probe There is no The effects could be It seems was used in solutions containing varying or added amounts of certain cations. reason to suppose that the effect is limited to the ions investigated here. particularly misleading in a flow cell where vigorous stirring may not be practicable. likely that other glass membrane ion-selective electrodes would suffer from the same type of interference. Any explanation of the observed phenomena must take into account that the potential shift is transitory and that a potential peak in the opposite direction (increasing pH) is observed on1006 KENNEDY changing the metal ion solution to a similar one with no added test ion.The speciation of the metal ions must also be considered. Although much more data are required in order to establish the mechanism of the observed effects, a feasible interpretation can be made based on recent theories that describe the adsorption of ions on to different planes parallel to the surface. The theories are applicable to silica, silicate and metal oxide surfaces. Both three- and four-plane models have been pro- posed.4 In both these models H+ and OH- ions are adsorbed on to a plane closest to the solid phase whereas ions such as Cu2+ and Cu(OH)+ are adsorbed on to a plane nearer to the bulk solution. If this idea is applied to a glass pH electrode, the inner region would represent an ion-exchange region that is fairly selective to H+ ions and determines the effective potential on the outside of the glass membrane.Further out, the glass would be less selective and ions such as Cu2+ could become adsorbed. The adsorption processes as reviewed by James and Barrow5 could be envisaged as and f SiOH + Cu2+ + f SiO-.CU2+ + H+ f SiOH + CU(H,O)~+ + 5 SiO-.CuOH+ + 2H+ The remaining waters of hydration have been omitted. The released protons may then be captured by the inner H+ selective layer, giving rise to an apparent pH drop. However, the excess of protons adsorbed on the inner layer would diffuse into bulk solution until equilibrium was attained. The rate of loss of protons would be influ- enced by the vigour of stirring.The reverse process could occur when the bathing solution was changed for one not containing copper ions. If this explanation is interpreted in terms of a hydrated layer6 in the outer region of the glass electrode, then Cu2+ would be adsorbed at the hydrated layer - solution interface (layer 2). Surface reaction would then release protons into the hydrated layer (layer l), possibly as far as the hydrated layer - dry layer boundary. This would be manifested by an apparent pH decrease, which would then dissipate as the protons diffused into solution until electrochemical equilibrium was reached. The third layer in the three-layer model would correspond to the diffuse layer between hydrated glass and bulk solution. Although this explanation is tentative, it does explain the main features of the phenomenon and is consistent with recent theories on the adsorption of metal ions on to oxide and silicate surfaces.It may also be significant that the relative order of effects shows some similarity to Irving - Williams order of complex stability’: Mn2+ < Fe2+ < Co2+ < Ni2+ < Cu2+ > Zn2+. How- ever, as pointed out elsewhere,5 the order for the extent of adsorption and also for the release of H+ ions on oxide surfaces may vary with the pH of the solution. A “pH” peak would then be expected. The author thanks the Agricultural Research Council for a Research Grant, during the tenure of which these effects came to light. References 1. 2. 3. 4. 6. 6. 7. Kennedy, C. D., and Stewart, R. A., J. Exp. Bot., 1982, 33, 1220. Rechnitz, C. A., and Kugler, G. C., Anal. Chern., 1967, 39, 1682. Goldstein, G., Wolff, C. M., and Sewing, J . P., Bull. SOG. Chim. Fr., 1971, 1195. Westall, J . , and Hohl, H., Adv. Colloid Interface Sci., 1980, 12, 265. James, R. O., and Barrow, N. J., in Laneragan, J . F., Robson, A. D., and Graham, R. D., Editors, Doremus, R. H . , in Eisenman, G., Editor, “Glass Electrodes for Hydrogen and Other Cations,” Irving, H., and Williams, R. J. P., J . Chem. Soc., 1953, 3192. “Copper in Soils and Plants,” Academic Press, North Ryde, Australia, 1981, p. 47. Marcel Dekker, New York, 1967, Chapter 4. Received January 24th, 1983 Accepted March lst, 1983

ISSN:0003-2654    

DOI:10.1039/AN9830801003

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst August 1983 Vol. 108 pfi. 1007-1012 1007 Influence of Instability of Thiocyanate in Argentimetry and Mercurimetry Edmund Bishop David Darker Michael D. Jones Paul M. Stewart and Salah M. Sultan" Chemistry Department University of Exeter Stocker Road Exeter E X 4 4QD The silver - thiocyanate and mercury( 11) - thiocyanate reactions have been examined by precise mass titrimetry for possible errors arising from decomposi-tion of thiocyanate in nitric acid media. For silver potentiometric titration in either direction in 0.15 mol kg-l nitric acid is free from error provided steps are taken to overcome adsorption but thiocyanate cannot be stored at this acid concentration. Using iron(II1) as the indicator visual titration in the Volhard direction (thiocyanate titrant) is unaffected by time delay acid con-centration and indicator concentration ; in the reverse direction an error of the order of 0.5% is incurred and increases as time acid and indicator concentra-tion increase.For mercury(II) potentiometric titration is unsatisfactory in both directions. Standard mercury(I1) ion solutions hydrolyse in nitric acid less than 4.0 mol l-l while thiocyanate is unstable in solutions containing more than 0.05 mol 1-1 of nitric acid. Neutral thiocyanate solution and mercury(I1) solution in 4.0 mol 1-l nitric acid can be titrated in either direction using iron(II1) as the indicator but the precision is poor and time delays are not tolerated. Higher indicator concentrations favour better precision through increased ease in locating the end-point.Keywords Thiocyanate titrations ; Volhard titration ; mercury (11) - thio-cyanate titration ; iron( 111) - thiocyanate indicator; stability of mercury(II) ion and thiocyanate solutions A recent investigation of the instability in acidic media of thiocyanate and its iron(II1) com-plexl calls into possible question the reliability of end-point location by addition of iron(II1) in silver(1) or mercury(I1) titrations of or with thiocyanate. Titration of silver2 or merc~ry(II),~ in acidic solution containing iron(III) with thiocyanate could be expected to pose no difficul-ties because neither free thiocyanate nor its iron(II1) complex is exposed materially to adverse conditions saving the possibility of direct reaction of the precipitate or complex with acid or iron(II1) although fading of the initial end-point colour may be experienced; nevertheless an examination is apposite.Kolthoff and Lingane4 have examined the titration of silver with potassium thiocyanate at high precision and concluded that the principal source of error was adsorption by the precipi-tate. Volhard3 remarked on the occurrence of early end-points in the titration of mercury(II), while Kolthoff and Stenger5 protest the difficulty of assessing the titration error in the presence of mercury(I1) - thiocyanate complexes of 2 1 1 1 1 2 1 3 and 1 4 stoicheiometries and aver that the end-point cannot be sharp. Tanaka et a1.6 give a log p at 25 "C of pz 16.43 p3 19.14 and p4 21.12 which together with data for the iron complexI1 and assigned experi-mental conditions compute to an absorbance of 0.0002 cm-l at the exact equivalence point; reaction of iron(II1) with the mercury(I1) complex cannot therefore be adduced as a source of early end-points.The 1 2 mercury(I1) thiocyanate separates as a precipitate with adsorptive properties from stronger solutions (pKs = 2-47'; solubility a t equivalence 0.15 moll-l). It is in the reverse titration of thiocyanate with the cations that decomposition of thiocyanate is more likely to introduce error; none has hitherto been noted and CohnS has preferred this direction with mercury(I1). The situation has been explored by high precision mass titrimetry with exaggeration of conditions. * Present address Chemistry Department Faculty of Science King Saud University P.O. Box 2455, Riyad Saudi Arabia BISHOP et al.INFLUENCE OF INSTABILITY OF Experimental Analyst Vol. 108 1008 Reagents AnalaR or Aristar reagents were used together with quartz-distilled water.9 Silver nitrate. Approximately 0.1 mol kg-l solution in 0.1493 moll-1 nitric acid. Mercury(II) nitrate. (a) A 20-g amount of mercury metal was accurately weighed, covered with water and treated with 0.2 mol of nitric acid plus an excess to give the required final acid concentration warmed to effect solution boiled to complete oxidation to mercury(I1) and to remove oxides of nitrogen made up to 11 in a weighed flask and re-weighed. The solution was tested for mercury(1) and oxides of nitrogen. (b) A 0.1-mol amount of mercury(I1) oxide was accurately weighed treated with 0.2 mol of nitric acid plus an excess to give the required final acid concentration made up to 1 1 and weighed.(c) For investigation of hydrolysis the required amount of mercury(I1) nitrate was dis-solved in the required amount of nitric acid and made up to volume (usually 100 ml) with water to give a 0.1 mol 1-1 solution. Potassium thiocyanate. (a) For potentiometric titrations with silver an approximately 0.1 mol kg-l solution in 0.1493 moll -l nitric acid was prepared and stored under nitrogen in brown glass in darkness. (b) For titrations with mercury(II) an approximately 0.2 mol kg-l solution in water was prepared. (c) For reactivity tests a 0.1 mol 1-1 solution in nitric acid of the required concentration was prepared. Iron(Il1). (a) As an indicator in silver titrations 0.0427 mol 1-1 of ammonium iron(II1) sulphate in 0.1493 mol 1-1 nitric acid or 0.2142 mol 1-1 in 0.1493 moll-l nitric acid was used.(b) As an indicator in mercury titrations solutions with a range of concentrations from 0.10 to 1.50 mol 1-1 of iron(II1) nitrate in commonly 0.2 mol 1-1 nitric acid were prepared. Procedure An aliquot of titrand is transferred into the weighed titration vessel which is weighed again. Any necessary reagents are added the vessel weighed a third time about 95% of the required equivalence mass of the titrant is added from a burette and the vessel weighed a final time. A small squeeze-bottle with a fine glass jet is charged with titrant and weighed. This is used to complete the titration weighing after each increment for potentiometric titration or after the final increment for visual titration.An increment of 2-5 mg produces a sharp colour change in visual titrations. For potentiometric titrations a silver indicator and a glass reference indicator electrodelo are used and in this instance titrand and titrant are made up in similar concentrations of acid so that there is no change therein during titration. Magnetic stirring is suitable. Results and Discussion Silver Titrations Potentiometric titrations conducted as quickly as the response speed of the electrodes and desorption of adsorbed ions from the precipitate permitted gave a mass ratio of silver to tliio-cyanate of 1.001 44 & 0.000 16 for titration of silver with thiocyanate (the Volhard method2), and of 1.001 12 The difference is not significant at the 95% confidence level.In the absence of iron(III) therefore errors arising from adsorption exceed any from decomposition of thiocyanate because the standard deviation for the direction conducive to acid attack of the thiocyanate is identical with that of the Volhard direction for which the uncertainty has been ascribed to adsorpti~n.~ In visual titration in the Volhard direction using 1 nil of 0.0427 mol 1-1 iron(II1) in a medium of 0.1493 mol 1-1 nitric acid an early colour change because of adsorption of silver ion on the precipitate was overcome by vigorous agitation and the silver to thiocyanate mass ratio was 1.001 11 & 0.00018 while the indicator error amounted to 8 mg of thiocyanate solution on ca. 25-g aliquots. This is clearly satisfactory.0,000 16 for the reverse titration on ca. 25-g aliquots August 1983 THIOCYANATE I N ARGENTIMETRY AND MERCURIMETRY 1009 An interlocking series of titrations was conducted with (variation of time interval) x (variation of indicator concentration) x (variation of acid concentration). Time intervals were varied by setting up a titration with an initial addition of about 80% of the equivalence mass of titrant and interrupting the titration for 0-93 h before completion. The amount of indicator was varied from 1 ml of 0.04824 to 5 ml of 0.2142 moll-' iron(II1) solution and the acid concentration was varied from 0.1493 to 2.5moll-l. In none of the nine groups of variant was any marked deviation observed and over the whole series analysis gave 1E; = 1.001 31 s = 0.00028.At normal volumetric precision (0.04%) decomposition of thiocyan-ate does not perturb the Volhard titration. In the reverse titration of thiocyanate with silver in 0.1493 mol 1-1 nitric acid using 1 ml of 0.0427 mol 1-1 iron(II1) as the indicator the colour would clear from the solution leaving the precipitate coloured indicating adsorption of thiocyanate ions and requiring vigorous mixing. The mass ratio of silver to thiocyanate solution was 0.99589 0.00040 a difference from the potentiometric ratio of 13 standard deviations which is clearly significant. A further interlocking series of titrations was conducted as above. This immediately revealed extensive decomposition of thiocyanate increasing with time interval indicator concentration and acid concentration.Sufficient individual results are given in Table I to illustrate the influence of the several conditions. At low errors the deviation increases approximately linearly with time and with the amount of indicator. TABLE I MASS TITRATION OF ca. 25-g ALIQUOTS OF 0.1 mol kg-l THIOCYANATE WITH 0.1 rnol k g l SILVER ION Amount of indicator Nitric acid Initial concentration1 mol 1-1 0.149 3 0.149 3 0.149 3 0.149 3 0.149 3 0.149 3 1.0 1 .o 0.149 3 0.149 3 0.1493 0.1493 0.1493 0.1493 0.149 3 0.1493 0.149 3 1 .o 1 .o 1 .o 1.5 1.5 1.5 2.0 2.0 2.5 Volume/ ml I 1.0 1 .o 1 .o 1 .o 1.0 1.0 1 .o 1 .o 1.0 1 .o 1 .o 1 .o 1.0 5.0 5.0 5.0 1.0 1.0 1.0 1.0 1.0 1.0 1 .o 1.0 1.0 Concentration1 mol 1-1 Potentiometric reference 0.04824 0.04824 0.048 24 0.048 24 0.048 24 0.048 24 0.048 24 0.04824 0.048 24 0.048 24 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 0.241 2 increment, % of theory --90.5 86.7 86.8 89.3 83.4 79.1 79.7 79.3 79.7 79.5 80.5 83.3 78.6 78.0 61.7 77.8 --------Time interval 0 0 2 h 40 rnin 7 h 15 min 10 h 27 h 0 27 h 0 5 h 10 rnin 36 h 30 min 0 4 h 30 rnin 40 h 45 min 0 4 h 30 min 37 h 15 min 0 5 h 30 rnin 22 h 45 rnin 0 3 h 20 h 0 1 h 30 rnin 0 Mass ratio of silver to thiocyanate 1.001 12 0.99601 0.99583 0.99573 0.995 38 0.994 57 0.99529 0.99446 0.995 25 0.99535 0.99405 0.99625 0.99620 0.991 78 0.995 94 0.99405 0.984 83 0.99651 0.995 99 0.995 11 0.996 35 0.78041 0.99627 0.3 17 28 -* -* Deviation, % --0.518 -0.528 -0.538 -0.573 - 0.654 - 0.582 -0.665 -0.586 -0.577 - 0.706 - 0.487 -0.591 - 0.933 -0.517 -0.706 - 1.627 - 0.460 -0.513 - 0.600 - 0.476 -22.05 -- 0.484 -- 68.3 * Residual thiocyanate completely destroyed.Mercury( 11) Titrations Attempts at potentiometric titration confirmed earlier observations by the senior author. Slow establishment of the multi-equilibrium between the various complex species the low solubility of the 1 2 complex low electrode exchange currents mixed potentials arising fro 1010 BISHOP et al.INFLUENCE OF INSTABILITY OF Analyst Vol. 108 minute traces of mercury(I) reaction of the mercury metal film on the electrode with mercury(I1) species to give a coating of mercury(1) thiocyanate which disproportionates all contribute to results that are unsatisfactory when high precision is sought. Moreover trouble previously encountered in these laboratories in the preparation of standard ionic (perchlorate) and partly ionic (sulphate) solutions of mercury(I1) because of hydrolytic precipitation of basic salts occurred; solutions in 0.1 0.2 0.5 and 1.0 mol k g l nitric acid deposited crystal-line basic nitrate on standing. This further suggested the possibility that the precipitate formed during titration may be hydrolysed.The collected washed and dried precipitate, formed from 0.5 mol kg-l nitric acid solution was examined by infrared spectroscopy and showed strong bands at 2 100 and 720 cm-l with weak bands at 1610 890 and 910-930 cm-l, confirmative of thiocyanate. A further sample was analysed for mercury found 63.0%, required 63.33%. As the titration using a glass reference indicator electrodelo requires all reagents to have the same acid concentration for best results thiocyanate solutions in 0.842, 0.697 0.574 and 0.389 molkg-l nitric acid were prepared and tested at intervals for thiocyanate and sulphate. The two stronger acid solutions showed decomposition in 30 and 60 min respectively and total decomposition in 24 h. The two weaker solutions showed de-composition in less than 24 h.All showed production of sulphate. Clearly the two reagents in acidic media are incompatible high acid concentration causes decomposition of thiocy-anate while low acid concentration leads to hydrolysis of the mercury(I1). It is curious that hydrolysis of standard mercury(I1) solutions has not been reported in the literature; in reference 5 preparation of a standard solution from mercury gives the acid con-centration as 0.24moll-1 (p. 338) from mercury(I1) nitrate as 0.4mol1-1 (p. 332) and as 0.6 mol 1-1 for a mercury sample solution (p. 337). Further in the standardisation of thio-cyanate with silver the acid concentration is 0.4 moll-1 (p. 255). Certainly freshly prepared mercury(I1) solutions in nitric acid as dilute as 0.1 moll-1 are initially clear and can be used immediately but hydrolysis occurs within hours.To settle the point a series of 0.1 mol 1-1 mercury(I1) solutions was prepared in 0.5 0.6 0.7 0.8 0.9 1.0 1.2 1.4 1.6 1.8 2.0 2.5 3.0, 3.5,4.0,4.5,5.0 5.5 6.0 and 6.5 moll-1 nitric acid and inspected at intervals. All remained clear after 1 h a small amount of precipitate had appeared after 24 h in the solutions up to 0.8 mol l-l at 40 h dense precipitates had appeared in solutions up to 2.0 mol l-l at 6 d precipitation had extended to 3.5 mol l-l while the stronger acid solutions remained clear for several weeks although after 2 months slight precipitates had appeared up to 5.5 mol 1-1 acid. There was some slight variation between duplicates in initial time of precipitation but it would appear that the minimum nitric acid concentration required to give reasonably stable solutions is about 4.0 moll-l.This accords with earlier work in these laboratories on per-chloric and sulphuric acid solutions of mercury(I1). Supersaturation occurs with more dilute solutions but once precipitation begins the process becomes rapid and if the precipitate is filtered off precipitation continues immediately in the filtrate. Thiocyanate is unstable to strong acidlJ1 and is saidll to decompose slowly yielding carbonyl sulphide on boiling with 2 4 mol 1-1 acid and to give nitric oxide and hydrogen cyanide on warming with “dilute” nitric acid. However the reaction with nitric acid occurs in the cold, and at acid concentrations of 1.091 1.147 1.243 and 1.296 mol kg-1 the orange - pink to red colour appears within a few minutes and intensifies rapidly followed by vigorous effervescence with evolution of nitrous fumes leaving a colourless solution containing sulphate but no thiocyanate.In 1.036 0.936 0.842 and 0.697 mol kg-l acid the reaction occurs more slowly and gently with total destruction of thiocyanate within 24 h while the reaction in 0.574 and 0.389 mol k g l acid is incomplete in 48 h. A further series of solutions in 0.05,0.10,0.20, 0.30 0.40 0.50,0.55,0.575 0.60 0.625 0.650 and 0.70 moll-1 nitric acid was observed over a period of 1 month. Colour generation commenced in most of the solutions of 0.55 mol 1-1 and more within 1 h; all of these gave a precipitate of sulphur and became colourless after 72 h, giving positive tests for nitrogen oxide in the vapour and for sulphate in the solution and negative tests for thiocyanate.In 24 h the 0.40 and 0.50 mol 1-1 solutions showed decomposi-tion which later proceeded to completion. Solutions in 0.2-0.3 moll-1 acid were stable for 24 h but later showed incomplete decomposition. The solution in 0.1 mol 1-1 acid decom-posed within 1 month while that in 0.05 mol 1-1 acid remained stable for 2 months. There remained the question as to whether mercury(I1) in 4.0 mol k g l nitric acid and thio-cyanate in neutral aqueous solution could be titrated without (a) dilution by the water causin August 1983 THIOCYANATE IN ARGENTIMETRY AND MERCURIMETRY 101 1 hydrolysis of the mercury(I1) or (b) admixture with the strong nitric acid producing decom-position of the thiocyanate.A volumetric trial gave a volume ratio of mercury(I1) to thio-cyanate of 1.0121 with thiocyanate as the titrant and 1.0112 with mercury(I1) as the titrant. This being satisfactory a series of mass titrations was conducted in each direction with n = 12, and 2.0 ml of 1.50 mol 1-1 iron(II1) nitrate as indicator. With mercury(1I) as the titrant the ratio was 1.121 65 with s = 0.00146; with thiocyanate as the titrant the ratio was 1.121 44 with s = 0.00206; overall the ratio was 1.12156 with s = 0.00167 and one outlier. These results give a relative standard deviation of 0.14870 which seems to be the best precision to be expected of the titration and compares with the claimed5 accuracy of 0.3%. Under the circumstances there was no point in attempting a variation of acid concentration, but a multivariate (time interval) x (amount of indicator) test was mounted in which 97.5% of the mercury(I1) titrant was initially added with various amounts of indicator.For a time interval of 3 h a mass ratio of 1.16062 with s = 0.001 96 was obtained showing no significant variances for 1 ml of indicator in the range 0.10-1.50 mol 1-1 iron(II1). For a longer interval, the remaining thiocyanate was completely destroyed and the precipitate redissolved. In further sets in which the initial addition of mercury(I1) was to within 0.4% of equivalence, destruction of residual thiocyanate also occurred but the precipitate did not redissolve. A further series in which 1.00 ml of 0.25 0.50 0.75 1.00 or 1.50 moll-I iron(II1) was added as indicator with n = 10 was run but the titration was carried out without a break.The over-all mass ratio was 1.15997 with s = 0.00296 and no significant difference at the 95% confidence level was observed between the individual sets. The s-values however decreased steadily from 0.00742 for 0.25 moll-1 indicator to 0.00086 for 1.50 mol 1-1 iron(II1) ; 2.0 ml of the latter corresponds to the recommended amount [l ml of 2.57 mol 1-1 ammonium iron(II1) ~ulphate].~ This simply reflected the greater ease of observation of the end-point. Other indicators were examined in the titration of thiocyanate and chloride. Diphenyl-carbazide12J3 gave indistinct and muddy colours and a poor and vague colour change and is obviously unsuited to the high acid concentration in the mercury(I1) solution.Nitroso-pentacyan~ferrate(II),~~$~~ however gave satisfactory end-points in titration of chloride and thiocyanate with a 4.0mol1-1 nitric acid solution of mercury(II) but the rather complex indicator corrections15 attenuate the precision of the method. Conclusions For the silver - thiocyanate reaction the potentiometric titration is satisfactory in either direction in media of low acid concentration and there is no significant difference between directions. In the Volhard direction (thiocyanate titrant) visual indication is also satisfactory; time acid concentration and indicator concentration exert no significant influence. Adsorp-tion of silver ion by the precipitate gives premature colour changes and the final end-point colour does fade but the indicator error is no more than 8 mg of 0.1 mol k g l thiocyanate solution.In the reverse titration (silver titrant) visual indication introduces an error of the order of 0.5% at a nitric acid concentration of 0.15 rnol kg-l ; the error increases with exposure time increasing nitric acid concentration and increasing iron( 111) concentration. There is no reaction between iron(II1) and the precipitate. For the mercury(I1) - thiocyanate reaction potentiometric titration in both directions is unsatisfactory in precise work. Standard ionic mercury(I1) solutions require a nitric acid concentration of 4.0 mol 1-1 to prevent hydrolysis while thiocyanate is unstable to nitric acid concentrations in an excess of 0.05 moll-l. Nevertheless straightforward titration in either direction with thiocyanate solutions in water and mercury(I1) solutions in 4.0 mol 1-1 nitric acid with iron(II1) as the indicator is possible and gives results of limited precision.Any delay should be avoided and the ease of location of the end-point is improved by higher indicator concentrations. The complex does not react with iron(II1). References 1. 2. 3. 4. Sultan S. M. and Bishop E. Analyst 1982 107 1060. Volhard J. J . Prakt. Chem. 1874 117 217. Volhard J. Justus Liebigs Ann. Chem. 1878 190 57. Kolthoff I. M. and Lingane J. J. J . Am. Chem. SOG. 1935 57 2126 1012 BISHOP DARKER JONES STEWART AND SULTAN 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Kolthoff I. M. and Stenger V. A. “Volumetric Analysis,” Volume 11 Interscience New York, Tanaka N. Ebota K. and Murayama T. Bull. Chew. SOC. Jpn. 1962 35 124. Czakis M. Roczn. Chem. 1959 33 3. Cohn R. Chem. Ber. 1901 34 3502. Bishop E. and Sutton J. R. B. Anal. Chim. Ada 1960 22 590. Bishop E. Analyst 1952 77 672. Burns D. T. Townshend A. and Carter A. H. “Inorganic Reaction Chemistry,” Volume 2 Ellis Dubskq J. V. and Trtilek J. Mikrochemie 1933 12 315. Roberts I. Ind. Eng. Chem. Anal. Ed. 1936 8 365. VotoEek E. Chem.-Ztg. 1918 42 317. Kolthoff I. M. and Bak A. Chem. Weekbl. 1922 19 14. 1947 p. 337. Honvood Chichester 1981 pp. 107 and 109. Received January 31st 1983 Accepted March loth 198

ISSN:0003-2654    

DOI:10.1039/AN9830801007

出版商:RSC    

年代:1983

数据来源: RSC