摘要:

476 Analyst, April, 1983, Vol. 108, pp. 476-480 Volatilisation of Zirconium, Vanadium, Uranium and Chromium Using Electrothermal Carbon Cup Sample Vaporisation into an Inductively Coupled Plasma Kin C. Ng and Joseph A. Caruso" Department of Chemistry, University of Cincinnati, Cincinnati, OH 4522 1, USA Zirconium, vanadium, uranium and chromium react with ammonium chloride (7% m/V) when heated in an electrothermal carbon cup to form their corresponding chlorides. These metal chlorides are subsequently vaporised into an inductively coupled plasma for optical emission spectroscopy. The preferential halide formation of these refractory elements in the electrothermal carbon cup has allowed their determinations to proceed with sub-nanogram detection limits and adequate precision of about 6% relative standard deviation for 5-pl samples.Linear dynamic ranges span about three orders of magnitude. Keywords : Metal chloride formation ; electrothermal carbon cup vaporisation ; inductively coupled plasma ; optical emission spectroscopy In a previous report ,1 microlitre sample introduction into an inductively coupled plasma (ICP) was described. A 10-p1 sample was dried and vaporised with an electrothermal carbon cup and the sample vapour was carried by a stream of argon gas into the ICP for optical emission spectroscopy. Detection levels, approximately an order of magnitude lower than those using the conventional solution nebulisation - ICP, were found for 21 elements. Performance comparisons were made between tantalum-overcoated pyrolytic graphite cups and the stan- dard pyrolytically coated cups.However, the 21 elements determined were relatively volatile. Preliminary studies with the less volatile chromium, vaporised in both the tantalum coated and regular cups, produced severe memory effects.l This has been attributed to the possible reaction between chromium and the carbon cup surface to form refractory compounds. Although a carbon cup power supply could be utilised that would provide enough energy to vaporise these compounds, the sublimation of carbon at high temperatures precludes this approach. Kirkbright and Snook2 have met with the same difficulty in determining refractory elements using a similar system. These workers, however, were able to determine refractory elements successfully by employing halogen - argon gas mixtures in the sheathing gas of the graphite rod and the plasma injector gas.Volatilisation of elements of interest through halide formation has been utilised for sample introduction into a low-power microwave-induced A comprehensive review on volatilisation5 and a report on halogenation6 of trace elements for atomic spectrometry have appeared in the literature. In this study ammonium chloride was added to the sample solution to promote analyte volatilisation through halide formation from the carbon cup. This method was applied to the determination of zirconium, vanadium, chromium and uranium. Experimental Conditions and Procedures The experimental operating conditions and procedures employed were the same as those reported earlier,l except that 5-p1 rather than 10-pl samples were used.The carbon cup power supply settings for zirconium, vanadium and chromium in ammonium chloride solutions were as follows: dry at 1.2 V for 45 s, ash at 1.2 V for 10 s and atomise at 9.8 V for 1.5 s; for uranium in ammonium chloride solutions, dry at 1.2 V for 45 s, ash at 1.2 V for 10 s and atomise at 12.3 V for 1.5 s. When no ammonium chloride was added to the analyte solutions, the dry setting was 1.2 V for 45 s, the ash setting a t 3.1 V for 10 s and the atomisation was set at the maximum (13.3 V) for 1.5 s. * To whom correspondence should be addressed.NG AND CARUSO 477 Reagents Analytical-reagent grade chemicals were used with 1 000 p.p.m. single element stock solutions prepared by dissolviiig ZrO(N0,),.2H20 and UO2(NO,),.6H,O in distilled, de-ionised water.Vanadium in the form of V,O, and chromium as K,Cr,O, were commercially available as 1000 p.p.m. atomic-absorption standards (Fisher Scientific Co.). A 10% m/V ammonium chloride solution was prepared in distilled, de-ionised water. Multi-element solutions containing the appropriate percentage of ammonium chloride were prepared by dilution from these stock solutions. Results and Discussion Refractory elements have been determined using the electrothermal carbon cup for sample introduction into the ICP. When using the pyrolytically coated carbon cup for these elements with the “atomise” setting at the maximum (13.3 V), 10 p.p.m. of chromium showed a severe memory effect and vanadium showed a poor detection limit and memory with a new pyro- lytically coated cup.After approximately 15 firings, a 10 p.p.m. vanadium signal was pro- gressively degraded to no signal, which was attributed to the progressive loss of the pyrolytic carbon coating upon the successive firings of the cup at high temperature. For zirconium, no signal was detected with a 1 p.p.m. sample. Uranium, although refractory, was determined with good precision. The relative standard deviation (RSD) was 3.2% for five determinations of a 0.1 p.p.m. sample. However, the sensitivity for uranium is low, as shown in Fig. 1, line A. The “atomise” setting of the power supply was set at the maximum. With 5 p1 of the 10 p.p.m. samples, chromium still exhibited its severe memory effects. No signal was seen for zir- conium, vanadium and uranium as these elements may form refractory compounds with the tantalum carbide cup surface. The experiments were repeated with the tantalum carbidised pyrolytically coated cup.k 1 0 3 c lo* g 10’ C .w .- 0 v) .- .- c C 100 ~ ~ 1 1 1 1 1 ~ K 0.01 0.1 1 10 100 Uranium concentration, p.p.m. Fig. 1. Uranium emission response: A, uranium in water; and B, uranium in 7% (m/V) ammonium chloride solution. Chloride Formation and Volatilisation It is assumed that the temperature at which chloride is available from the added chloride- containing compounds is sufficient for volatile metal chloride formation., Therefore, the choice of a compound with an easily available chloride should be one that allows for sample drying without the loss of analyte in this step.Ammonium chloride is such a substance as it is soluble in water and has a sublimation temperature of 340 OC., Hence, it was chosen as the agent to provide the chloride. In the preliminary study, a sample of 1 p.p.m. of zirconium in 7% m/V ammonium chloride solution was dried at 1.2 V for A5 s, ashed at 2.5 V for 10 s and atomised at 10 V for 1.5 s. Analyte emission was observed during this ash step while the remaining analyte emission appeared during atomisation. I t is believed at this ash tempera- ture (corresponding to 2.5 V), chlorides have been released from ammonium chloride and have volatilised some of the analytes. Therefore, to avoid any analyte loss at the ash stage, it was set at the same voltage as that of the dry stage. Hence, for all later experiments, samples478 NG AND CARUSO: VOLATILISATION OF ZR, V, U AND CR Analyst, VoZ.108 were simply dried and then atomised at high temperatures. Chloride from ammonium chlor- ide in the gas phase at the elevated temperatures reacts with analytes to form volatile chlorides that are subsequently carried by the argon injector gas into the ICP. The melting- and boiling- points for the refractory elements and their corresponding carbides and chlorides are listed in Table I. Most of the small amounts of analyte evaporates between its melting- and boiling- It can be seen that the resulting chlorides lie below a boiling or sublimation tem- perature of 1000 "C. Therefore, the power supply may then be adjusted to give sufficient current to sustain vaporisation of these metal chlorides.The concentration of ammonium chloride used was optimised for potential multi-element determination of these elements. Its concentration in multi-element solutions (containing 1 p.p.m. each of zirconium, vanadium, uranium and chromium) was varied from 1 to 9% m/V with 1% increments. Some elements have shown multiple peaks from greater than 7% ammonium chloride solutions. The compromise optimal concentration was determined to be With the presence of ammonium chloride in the analyte solutions, similiar signal responses were obtained using tantalum carbidised pyrolytically coated and non-carbidised pyrolytically coated cups. Therefore, the non-carbidised pyrolytically coated cups were used for all chloride formation and volatilisation experiments.7%. TABLE I MELTING- AND BOILING-POINTS ("C) Corresponding carbide 7 Corresponding chloride : Element M.p. B.p. M.p. B.p. boiling/sublimation temperature/"C Zirconium . . . . 1852 4 377 3 540 5 100 < 500 Vanadium . . . . 1890 3 380 2 810 3 900 < 500 Chromium . . . . 1867 2 672 1890 3 800 < 600 Uranium . . .. 1132 3 818 2 360 4 370 < 1 000 Carbon . . . . 3652 4 827 (subl.) Precision intensity was measured. However, improved precision was found with this decontamination firing. Sharp, single peak analyte signals resulted from the volatilisation technique, hence the peak No memory was seen in the decontamination firing between samples. Improved precision 6 s H Time 4 Fig. 2. Typical signal reproducibility using 7% (m/V) ammonium chloride solution volatilisation (60 ng of vanadium).6 s I H Time d Fig. 3. 10 p.p.m. of chromium volatilised in 7 yo (m/ V ) ammonium chloride solution (arrows point a t the second response peaks which show memory).April, 1983 USING ELECTROTHERMAL CARBON CUP VAPORISATION INTO AN ICP 479 was also obtained when samples were determined from low to high concentrations. All the experiments were done under these two conditions, Fig. 2 (50 ng of vanadium) represents a 5.5% RSD and is typical of the precision available with the volatilisation method. Chromium (10 p.p..m.) shows a second peak as seen in Fig. 3. This second peak carries a memory in successive determinations. However, it does not interfere significantly with the signal pre- cision (5.8% RSD). At lower concentrations, chromium does not show this second peak, suggesting that the volatilisation technique becomes inefficient at increasing analyte concentra- tions, In fact, 100 p.p.m.solutions of chromium, zirconium and uranium show multiple peaks and their responses are non-linearly related to those at the lower concentrations. Detecti0.n limits The detection limit is defined as the analyte concentration in 7% m/V ammonium chloride solution that gives a net emission intensity equal to twice the standard deviation of the 7% m/V ammonium chloride background emission intensity. The values are at parts per billion (ng) levels and they are compared with those of the conventional nebulisation - ICP system in Table 11. Recorder signal tracings for blank and analyte emissions are shown in Fig.4. TABLE I1 DETECTION LIMITS 7% ammonium chloride volatilisation ; Solution nebulisation - electrothermal ICP*, vaporisation - ICP Element Wavelength/nm p.p.m. (5 pl), p.p.m. Zirconium . . 339.2 0.007 7 0.003t Vanadium . . 437.9 - 0.05t Chromium . . 357.9 0.023 0.04t Uranium .. 386.0 0.25 0.006t Uranium .. 386.0 0.25 0.000 65 Vanadium .. 309.3 0.005: - 7% ammonium chloride volatilisation ; electrothermal vaporisation - ICP/ ng 0.02t 0.3t 0.2t - 0.03t 0.003§ * Taken from reference 10. t DL (detection limit) = 2a of 7% ammonium chloride background emission. $ Best value reported in reference 10. S LQL (lowest quantifiable level) value (without volatilisation) = 0.62 mm expressed as intensity. See reference 1. Uranium, without the addition of ammonium chloride, gives a low detection level owing partly to the lack of background emission from ammonium chloride.Its detection limit corresponds to a peak emission intensity equal to one quarter of one block (0.62 mm expressed as intensity) on the strip chart paper.l Therefore, the addition of 7% m/V ammonium chloride has improved the sensitivity (Fig. 1) whilst sacrificing the detection level for uranium (Table 11). Linearity The linear dynamic ranges, using the ammonium chloride volatilisation technique for zirconium, vanadium and chromium are shown in Fig, 5 ; uranium is shown in Fig. 1. Each point in Figs. 1 and 5 is an average of five determinations. The linear regression correlation coefficient was calculated using all the determined values and the values are Zr 0.9975, V 0.9943, Cr 0.9983, U 0.981 9 and U, without volatilisation, 0.9736, indicating acceptable linearities for zirconium, vanadium and chromium.However, marginal improvement is obtained for uranium using this volatilisation technique. Conclusion Volatilisation of refractory elements to their chlorides has allowed zirconium, vanadium and chromium to be determined with the electrothermal carbon cup sample vaporisation - ICP combination. The addition of 7% m/V ammonium chloride to the sample solution has proved to480 \ NG AND CARUSO c - 0 J A 6 s . Fig. 4. Recorder tracings of typical emission signals : A, de-ionised distilled water blank at the uranium wavelength ; B, uranium in de-ionised distilled water; C, 7% (m/V) ammonium chloride solution background at the uranium, zirconium, vanadium and chromium wavelengths ; D, uranium, zirconium, vanadium and chromium in 7% (m/V) ammonium chloride solution.lb O 0.01 0:1 ; ;o 100’ Concentration of metal, p.p.m. Fig. 5. Linearity ranges of about three orders of magnitude obtained for A, zir- conium; B, chromium; and C, vanadium volatilised in 7% (m/V) ammonium chloride solution. be a convenient source of chlorides for in situ formation of metal chlorides in the carbon cup. Uranium, although refractory, has reasonable detectability without chloride formation. The authors are grateful for National Institute of Occupational Safety and Health grant support by grant number OH-00739. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Ng, K. C., and Caruso, J . A., Anal. Chim. Acta, 1982, 143, 209. Kirkbright, G. F., and Snook, R. D., Anal. Chem., 1979, 51, 1938. Skogerboe, R. K., Dick, D. L., Pavlica, D. A., and Lichte, F. E., Anal. Chem., 1975, 47, 568. Alder, J. F., and da Cunha, M. T. C., Can. J . Spectrosc., 1980, 25, 32. Bachmann, K., Talanta, 1982, 29, 1. Kantor, T., Hanak-Juhai, E., and Pungor, E., Spectrochim. Acta, Part B, 1980, 35, 401. Weast, R. C., Editor, “Handbook of Chemistry and Physics,” Fifty-ninth Edition, Chemical Rubber Company, Cleveland, OH, 1978. Rautschke, R., Amelung, G., Nada, N., Boumans, P. W. J. M., and Maessen, F. J. M. J., Spectrochim. Acta, Part B, 1975, 30, 397. Kantor, T., and Pungor, E., J . Therm. Anal., 1974, 6, 521. Winge, R. K., Peterson, V. J., and Fassel, V. A., Appl. Spectrosc., 1979, 33, 206. Received September lst, 1982 Accepted October 13th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800476

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst, April, 1983, VoE. 108, pp. 481-484 Electrothermal Atomisation Atomic-absorption 481 Spectrophotometric Determination of Chromium(V1) in Urine by Solvent Extraction Separation with Liquid Anion Exchangers Claudio Minoia Centro Ricerche di Fisiopatologia e Sicurezza del Lavoro, Fondazione Clinica del Lavoro, Univevsith di Pavia, Pavia, Italy Am brog io Mazzucotelli" Istituto di Petrografia, Universith di Genova, Genoa, Italy Alessa ndro CavaI leri Cattedra di Medicina del Lavoro, Universith di Modena, Modena, Italy and Vincenzo Minganti Istituto di Chimica Generale ed Inorganica, Universith di Genova, Genoa, Italy An electrothermal atomic-absorption spectrophotometric determination of chromium(V1) in urine samples is described. The separation of these ions from the biological matrix by using high relative molecular mass amines (such as Amberlite LA-1 or LA-2 liquid anion exchangers) is also reported.Keywords : Hexavalent chromium ; liquid anion exchangers ; electrothermal atomisation atomic-absorption spectrometry ; urine analysis Chromium exists mainly in two oxidation states in solution, chromium(II1) and chromium(V1). The first is the more stable and exhibits a tendency to form inert comp1exes.l Chromium, which is highly toxic, is used as a corrosion inhibitor and finds its way into waste waters from cooling towers. An understanding of the chemical separation and oxida- tion states of trace elements in biological and geochemical environments is very important. The presence of chromium in trace amounts in ground or sea waters has an influence on aquatic life.The environmental consequences of long-term exposure to high concentrations of these substances are uncertain, but chromium seems to interfere with the enzymatic sulphur uptake of cells2 and affects the lungs, liver and kidneys.3~~ Although there have been no reports of the oral toxicity of tervalent chromium, it is considered necessary for the maintenance of a normal glucose tolerance f a c t ~ r . ~ ~ ~ In view of the above, the separation and determination of chromium at trace levels has received considerable attention. Atomic-absorption spectrophotometry with electrothermal atomisation is the most common technique used for the determination of total chromium in u ~ i n e . ~ ~ ~ Direct analysis of biological samples is widely used in monitoring exposed workers but it is unsatisfactory for the determination of trivalent or hexavalent chromium.e A separate determination is possible employing the traditional spectrophotometric 1,ti-diphenyl- carbazide method, but this procedure is subject to interference from other ions and requires mineralisation of the sample.9 SlavinlO reported that only hexavalent chromium is chelated in ammonium tetramethylene dithiocarbamate (ammonium pyrrolidine dithiocarbamate, APDC) and extracted in isobutyl methyl ketone (IBMK).In a previous paper we reported data on the determination of chromium(V1) in urine samples after chelation with APDC, extraction with IBMK and atomic-absorption measure- ment using a graphite furnace.8 The extraction of chromium(V1) by various high relative molecular mass amines from acidic solutions has also been re~0rted.ll-l~ Sastri and co- w o r k e r ~ ~ ~ ~ ~ ~ showed that peroxychromic acid is extracted quantitatively into trioctylamine Present address: Istituto di Chimica Generale ed Inorganica, Universith di Genova, Genoa, Italy.* To whom correspondence should be addressed.482 MINOIA et al. : ELECTROTHERMAL AAS OF CR(VI) IN URINE Analyst, VoZ. 108 (TOA) or Aliquat 336 with benzene as diluent. The same liquid anion exchanger has been applied to sea-water analysis for chromium speciation.16 There appear to have been no applications of high relative molecular mass amines either to biological materials or to urine analysis for chromium(V1). In this paper we report results obtained in the determination of chromate and dichromate ions in urine samples after an ion-exchange extraction with Amberlite LA-1 or LA-2 diluted in IBMK, stripping in the same solution with 6 M hydrochloric acid and subsequent determination by graphite furnace atomic-absorption spectrophotometry.100 200 300 I -- -, Experimental Apparatus and Reagents A Perkin-ElmerJ Model 603 atomic-absorption spectrophotometer equipped with a Perkin-Elmer, Model 56, strip-chart recorder and an HGA-76B graphite furnace was used. Pyrolytically coated graphite tubes were used in the furnace. Chromium was determined at 357.9 nm. A chromium hollow-cathode lamp was operated at 30 mA and a slit setting of 4 (0.7 nm) was used. The atomisation programme was as follows: drying for 40 s at 120 OC, charring for 40 s at 500 "C and atomisation for 8 s a t 2700 "C (gas stop flow).Argon was used as the purge gas a t a flow-rate of 1.0 1 min-l. Chromium stock solutions were prepared by dissolving 0.2828 g of potassium dichromate in 1000 ml of doubly distilled water. The liquid anion-exchange solution was prepared by adding 100 ml of Amberlite LA-1 (BDH Chemicals) to 50 ml of 6 M hydrochloric acid, stirring continuously and diluting to 250 ml with IBMK in a calibrated flask. All reagents were of analytical-reagent grade. Procedure Immediately after the collection of urine, 1 ml of sample and 1 ml of liquid anion-exchange solution were pipetted into a stoppered polyethylene tube. After mixing for 2 min with a mechanical shaker, the tube was centrifuged for 10 min at 2500 rev min-l.A 20-4 volume of the upper layer was pipetted directly into the atomic-absorption furnace. Standards were prepared by adding known amounts of chromium(V1) to urine obtained from normal subjects. A reagent blank was run through all the steps in the procedure together with the samples. Results and Discussion In order to test the extraction of chromium(VI), two calibration graphs were prepared by analysing aqueous solutions containing, respectively, 100-200 and 300 pg 1-1 of chromium- (111) (as CrC1,.6H20) and the same concentration of chromium(V1) (as K2Cr20,). These 0.75 8 -e 2 2 c 0.50 0.25 2 100 Fig. 1. Electrothermal atomic-absorption peaks of (a) trivalent chromium and (b) hexavalent chromium after extraction procedure from aqueous standard solutions.Numbers on peaks are chromium ion con- centrations (p.p.b.),April, 1983 BY SOLVENT EXTRACTION LIQUID ANION EXCHANGERS 483 TABLE I CALIBRATING PARAMETERS FOR CHROMIUM(VI) EXTRACTION FROM AQUEOUS SOLUTIONS Cr(V1) added/pg 1-1 Cr(V1) found/pg 1-l -- .. .. 25 26 2 2 f 3 2 2 f 3 .. .. 26 26 2 0 f 5 2 1 f 4 .. 26 26 2 4 f 1 2 3 f 2 Diluent LA- 1 LA-2 LA- 1 LA-2 22 IBMK .. standard solutions were treated with Amberlite LA-1 and IBMK as described under Pro- cedure for urine samples. The graphite furnace atomic-absorption analyses indicated complete quantitative recovery of chromium(V1) whereas chromium(II1) seemed to be absent from the extracted phase, as shown in Fig. 1. The ion-exchange extraction of chromium(V1) from aqueous standard solutions was carried out with different liquid resins (Amberlite LA-1 and LA-2) dissolved in different organic solvents (benzene, carbon tetrachloride and IBMK), as shown in Table I.The two resins showed equivalent performance. Evaluation of the extraction parameters was then performed on urine samples by testing the extraction power by adding the hydrochloric acid stripping solution before or during the addition of IBMK dissolved Amberlite solution (Table 11). TABLE I1 TESTS PERFORMED ON STRIPPING PARAMETERS Cr(V1) found/pg 1-l Cr(V1) added/ \ Sample* PiZ 1-1 CeH, cc1, IBMK A .. .. .. 25 2 0 f 6 2 2 f 5 24 f 1 B .. .. .. 25 12 f 4 9 * : s 15 f 3 * A = urine + LA-1 - 6 M HC1- IBMK; B = urine + LA-1 - IBMK + 6 M HC1. The strong reducing power with respect to chromium(V1) seems to be due to the action of the biological matrix with time; as shown in Fig.2, a decrease in recovery of chromium(V1) starts a few hours after the collection of the sample. This effect may be due to the reduction of chromium(V1) to chromium(III), which is not extracted by the suggested procedure as shown above. 100 75 $? s 50 a 25 I 2 6 10 Time/h Fig. 2. Recoveries of hexavalent chromium from a urine sample spiked with 10 pl of 100 p.p.b. Cr(V1) solution.484 MINOIA, MAZZUCOTELLI, CAVALLERI AND MINGANTI TABLE I11 ACCURACY OF DETERMINATION OF CHROMIUM(VI) IN URINE SAMPLES Recovery f s.d., yo Cr(V1) added/ I A i 10 10 94.7 f 1.1 99.4 f 0.4 10 16 97.2 f 0.8 100.1 f 0.3 10 20 96.9 f 1.6 100.3 f 0.2 10 26 96.2 & 1.8 100.1 f 0.3 No. of samples CLg 1-1 APDC - IBMK LA- 1 The accuracy of chromium(V1) recovery is summarised in Tables I11 and IV.The detec- tion limit is 0.1 pg 1-l. As shown, the results obtained were better than those obtained earlier by the APDC - IBMK method.a The better recoveries with the proposed method are probably due to the smaller number of manipulations required. TABLE IV PRECISION OF DETERMINATION OF CHROMIUM(VI) IN URINE SAMPLES Variation, % Cr(V1) found/ , A \ No. of samples Pg 1-1 APDC - IBMK LA- 1 10 6 8.2 6.4 10 12 7.2 3.7 10 18 9.7 4.1 10 37 4.9 3.2 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Rao, V. M., and Sastri, M. N., Talanta, 1980, 27, 771. Pankow, J . F., and Janauer, G. E., Anal. Chim. Acta, 1974, 69, 97. Valkovic, V., “Trace Elemental Analysis,” Taylor and Francis, London, 1975. Forstner, U., and Muller, G., “Schwermetalle in Flussen und Seen,” Springer Verlag, Berlin, 1974. Schwartz, K., and Mertz, W., Arch. Biochem. Biophys., 1959, 85, 292. Nise, G., and Vesterberg, O., Scand. J . Work. Environ. Health, 1979, 5 , 404. Nomiyama, H., Yotoriyama, M., and Nomiyama, K., Am. Ind. Hyg. Assoc. J., 1980, 41, 98. Minoia, C., Colli, M., and Pozzoli, L., At. Spectrosc., 1981, 2, 163. Beyermann, K., Fresenius Z . Anal. Chem., 1962, 190, 4. Slavin, W., At. Spectrosc., 1981, 2, 9. Nagakawa, G., Nippon Kagaku Zasshi, 1980, 81, 1533. Smith, E. L., and Page, J. E., J . SOC. Chem. Ind., London, 1948, 67, 48. Moore, F. L., Anal. Chem., 1958, 30, 908. Sastri, M. N., and Sundar, D. S., Anal. Chim. Acta, 1966, 33, 340. Sastri, M. N., and Rao, T. S. R. P., J . Inorg. Nucl. Chem., 1968, 30, 1727. De Jong, G. J., and Brinkman, U. A., Anal. Chim. Acta, 1978, 98, 243. Received September 3rd, 1982 Accepted November 8th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800481

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst April 1983 Vol. 108 pp. 485-491 485 Use of Laser Raman Spectrometry for a Quantitative Study of the Urea Synthesis under Process Conditions Part I. A Feasibility Study Martin van Eck Johannes P. J. van Dalen and Leo de Galan Laboratory for AnaZytical Chemistry University of Technology Jaffalaan 9 2628 B X Delft The Nethevlands The feasibility of laser Raman spectrometry for the in situ determination of the chemical composition during the synthesis of urea under process conditions has been studied. Raman bands suitable for quantitative analysis were found for all components. The use of an internal standard appears to be essential. The influence of temperature and pressure upon molar intensities has been studied for model components and as a result for most components of the urea synthesis no effect of pressure is expected.By contrast the Raman spectra of all components will be affected by a change of temperature. Keywords Laser Raman spectrometry ; in situ analysis ; uyea synthesis Many chemical processes proceed at non-ambient conditions of elevated temperature and pressure. The development and optimisation of process conditions benefit from a knowledge of the exact composition of the reaction mixture. Normally this is achieved by sampling and off -line analysis. Circumstances are conceivable however where this technique fails and in situ analysis must be used. An example is the synthesis of urea. Urea is formed in a two-step reaction: . . . . * ' (1) . . ' * (2) 2NH + CO d - H,NCOONH, H,NCOONH A - H,NCONH + H,O Normal synthesis conditions are that the temperature range is 170-210 "C and that the pressure range is 15-25 MPa.l A change of pressure or temperature will give a considerable change in the composition of the reaction mixture.Hence sampling and off -line analysis may produce unreliable results and in situ analysis is to be preferred for the determination of all reaction components. For this purpose spectroscopic methods are suitable. Ultraviolet spectros-copy would give data for carbonyl compounds and for ammonia but not for water. By con-trast infrared spectroscopy is impossible because a major part of the spectrum will be obscured owing to the presence of water in the mixture. On€y laser Raman spectrometry (LRS) seems feasible. The visible radiation used to excite and detect Raman transitions can be easily directed to a measuring cell.Water is an acceptable solvent and all components including water give a characteristic Raman band. The advantages of LRS for quantitative analyses have been shown for gases, liquids3 and solids. Better analytical results are to be expected if relative intensities are used instead of absolute ones as Raman intensities are affected by many experimental factors5-' The standard to which the intensities are related must be added to the mixture because no solvent is used in the urea synthesis. The suggestion of Tunnicliff and Jones8 of eliminating a number of the experimental factors in quantitative Raman analysis by diluting the sample cannot be used if in sit% analysis is desired. A significant improvement of the signal to noise ratio is possible if multiple scan and signal averaging tech-niques are used.Another advantage of computerised data processing is the possibility of curve fitting which is important when band overlap occurs. In situ analysis implies the need to study the influence of temperature and pressureg on chemical composition but before such conclusions can be drawn it must be verified that the temperature and pressure have no or a known influence upon the spectra produced by the components. Common spectroscopic techniques cannot be used in this instance. Generally the Raman effect is weak and the detection limits are rather high 486 VAN ECK et al. LASER RAMAN SPECTROMETRY FOR STUDY Analyst VoZ. 108 Experimental Apparatus The Raman spectrometer was composed of a scanning double-grating monochromator the Jobin Yvon Ramanor HGBS with a Coherent CR-2 laser as the light source.The 514.5-nm line of the argon-ion laser at a power of 1 W was used. Detection was carried out photo-metrically using a Hamamatsu R376 photomultiplier tube a d.c. amplifier and a recorder. Integrated band areas were measured with a Shimadzu Chromatopac CElB integrator. The conventional 90" geometry was applied to collect the scattered radiation (Fig. 1). Double monoc h rom ator - Recorder Amplifier Computer - w I I-t High pressure cell with Lenses sapphire Mirrors windows Fig. 1. Laser Raman spectrophotometer. Cells As Raman spectra are obtained in the visible part of the spectrum optical materials can be used that transmit radiation in this region.Hence in most instances cells or cell windows of ordinary glass can be used. The experiments at atmospheric pressure and ambient or elevated temperatures were carried out in a cell entirely made of glass. The cell was heated electrically with resistance wire the contents were stirred and the temperature was controlled with a thermocouple. This cell suitable for pressures of up to 400 MPa has a stainless-steel housing and three sapphire windows. Owing to corrosion of the stainless-steel housing the cell could not be used for mixtures containing ammonia. Sapphire like glass has the quality of transmitting visible radiation but it has a much higher resistance against corrosion and pressure. A disadvantage of sapphire is its birefringence because the orientation of the For the experiments at elevated pressure a Nova Swiss cell was used.The pressure in the cell was generated hydraulically. 1 .o x x x x x x x X A ' 90 180 270 360 Position of entrance window/" Fig. 2. Variation of intensity ratios of CCl bands with the position of the sapphire entrance window of the high pressure cell. A Ratio and B ratio of Imlr 220cm-1 to of I ~ 1 . r 22ocm-1 to ICCI~ 818 em-'; ICCI~ 460 em-1 April 1983 OF UREA SYNTHESIS UNDER PROCESS CONDITIONS. PART I 487 windows may influence tlie polarisation properties of the incident and scattered radiation and, hence the intensity of some Raman bands. Fig. 2 shows the influence of the orientation of the entrance window on the intensity ratios of different tetrachloromethane bands.As the 220 and 318 cm-l bands have the same depolarisation ratiolo and are affected in the same way by the position of the entrance window their intensity ratio reniains constant. The 460 cm-l band however is completely polarised and rotation of the cell window changes the intensity of this band significantly as is shown by the change in the ratio of the intensity of the 220 cm-l to the intensity of the 460 cm-l band. To obtain reproducible results it is therefore mandatory to ensure a fixed position of the sapphire cell windows. Results and Discussion Spectra Fig. 3 shows the spectra of aqueous solutions of urea ammonium carbamate and a mixture of the two components; spectrum D was made under process conditions of temperature, 170 O C pressure 25 MPa and the molar ratio of ammonia to carbon dioxide 2.The carbamate spectrum shows a strong band at 1035cm-l (C-N stretching) suitable for a quantitative determination. The shoulder a t the high-frequency side of this band does not arise from carbarnate but from carbonate," formed in the reaction of carbamate with water. A band a t the low-frequency side a t 1020 cm-l would indicate the presence of hydrogen carbonate. These bands will probably be absent under the very strong alkaline conditions that prevail during the urea synthesis. I t has a strong band a t 1005 cni-' (C-N stretching) and although this band overlaps partially with the 1035 cm-l band of carbarnate both bands can be determined readily with computer-ised data processing.The very broad band in these spectra situated in the 3000-3600 cm-' region can be assigned to the 0-H stretching vibrations14 of the water that was used as a solvent. Superimposed on this band are some weak bands that can be assigned to the N-H stretching vibrations14 of urea and carbamate. These features are also encountered in tlie spectrum of a 25% aqueous solution of ammonia (Fig. 4). Besides the bands assigned to the 0-H and N-H stretching vibrations this spectrum shows weak bands at 1100 and 1650 cm-l arising from the bending vibrations of NH,.15-17 Carbon dioxide is a three-atom molecule with a centre of symmetry and hence the Raman spectrum should show only one band assigned to the symmetrical C-0 stretching ~ i b r a t i 0 n . l ~ This very strong band is situated a t 1388 cm-l.Owing to Fermi resonance a second band is present a t 1286 cm-l. The Raman spectrum of urea is well k n ~ w n . ' ~ ~ ' ~ 1 L 1000 I b) C) 1000 1000 Avlcrn-' 1000 Fig. 3. Raman spectra of (a) urea solution; (b) ammonium carbamate solution ; (c) urea + ammonium carbamate solution; and ( d ) synthesis mixture a t process conditions NH,/CO = 2, T = 170 "C and P = 25 MPa 488 Analytical Characteristics VAN ECK et al. LASER RAMAN SPECTROMETRY FOR STUDY Analyst Vol. 108 The absolute intensity ( I ) of a Raman band can be given by . . - * (3) I = AsZL[C] . . where A is a constant sZ is the Raman scattering cross-section L is the factor for the local field correction18 and [C] is the concentration of the scattering compound.Hence absolute intensi-ties will be directly proportional to the concentration only if A sZ and L are independent of the concentration. Fig. 5 shows the intensity of the 1005 cm-l band of urea in aqueous solution as a function of the concentration with and without the addition of acetonitrile (ACN) as an internal standard. It is obvious that the absolute intensity values show a strong scattering mainly due to cell repositioning. Fig. 4. Raman spectrum of a 26% ammonia solution. Fig. 5. Graph of intensity of the 1005 cm-1 band of urea versws concentration of urea. A Absolute intensity; and B intensity of urea band relative to intensity of the 920cm-1 band of ACN added as internal standard. If relative intensities are used instead of absolute intensities the coefficient of variation for each individual point found with repetitive scans is considerably larger 5% instead of 3%.As the points lie much closer to a straight line however ACN appears to be effective as an internal standard. The detection limits based on a signal to noise ratio of 2 for urea and ammonium carbonate in aqueous solution are 4 x mol l-l respectively. For the present purpose these figures are adequate because at process conditions the carba-mate formation proceeds to completion and 50-80~0 of the carbamate is converted into urea.l and 5 x Quantitative Analysis Under Non-ambient Conditions The synthesis of urea however proceeds at elevated pressure and temperature and therefore calibration at ambient conditions and extrapolation to process conditions are possible only if the effect of pressure and temperature on the spectra is analysed.A direct study of these effects on the synthesis components is impossible because a shift of the equilibria may occur. Hence model compounds were chosen which are stable with respect to temperature and pressure. The above data were obtained at atmospheric pressure and ambient temperature. Pressure Effects Increasing pressure will give an increase in the Raman intensities owing to a decrease of volume and an increase in concentration. This effect will be very pronounced for gases less for liquids and negligible for solids owing to the difference in compressibility. If only the compressibility is taken into account molar intensities remain constant and an internal stand-ard eliminates the effect of compression April 1983 OF UREA SYNTHESIS UNDER PROCESS CONDITIONS.PART I 489 Literature data indicate that a change of pressure may result in a shift of the Raman f r e q u e n c i e ~ ~ ~ ~ ~ ~ while no change of the molar intensities is observed. If the change of pressure causes a change of the polarisability tensor a significant effect not only on the frequency but also on the molar intensity may be expected. This appears to be the case for liquid and fluid ammonia.15 The effect of pressure found for this compound was explained by assuming that the pyramidal height of the ammonia molecule is sensitive to pressure changes. A change of the pyramidal height results in a change of the polarisability tensor leading to a change of the molar scattering intensity.Fig. 6 shows the effect of pressure on the intensity of bands of pure ethanol acetone and 1,4-dioxan. A correction was made for the increase in intensity due to compression approxi-mately 3% in the range from 0.1 to 30 MPa at 293 K. Fig. 6 shows that up to 30 MPa the press-ure does not affect the Raman intensities of these bands neither was a change of band fre-quencies observed. Hence the exceptional behaviour found for ammonia15 is not encountered for these compounds nor is it expected for a compound like urea. This effect turns out to be different at different temperatures. 90 110 100 90 0 10 20 30 Pressu re/M Pa Fig. 6. Variations of Raman band heights of some model compounds with pressure. (a) Acetone 790 cm-l; (b) ethanol 890 cm-l; and (G) l,li-dioxan 840 cm-l.Temperature Effects The factors that have to be taken into account if the effect of temperature on the intensity of Raman bands is considered are thermal expansion local field effect population factors and the polarisability and hyperpolarisability tensors. It is obvious that thermal expansion will lead to a decrease in the observed Raman intensities, owing to a decrease in concentration. The appropriate correction is simple and is required only if absolute intensities are measured. Relative intensities ratioed to an internal standard are not affected by thermal expansion the change in concentration is equal for the compound to be measured and the standard. The correction factor for the local field effect (L) will also be influenced by a change of temp-erature as the index of refraction is a function of the temperature.According to the equation for (L) proposed by Eckhardt and Wagner,l* for pure water a decrease of the Raman intensity of 5% will be found if the temperature increases from 293 to 373 K. Assuming that the index of refraction does not change significantly over the range of the Raman spectrum the intensi-ties of all Raman bands will be influenced to the same extent. Hence the use of an internal standard compensates for thermal expansion as well as for the change of the correction factor for the local field effect. The temperature enhancement of the intensity of Stokes Raman bands due to a change of the population factors is given by21 I(T) 1 -exp(-hcv,/kT,) '.(4) - -I(T,) - l-exp(-hcvr/kT 490 VAN ECK et aZ. LASER RAMAN SPECTROMETRY FOR STUDY Analyst VoZ. 108 where or is the Stokes Raman shift and h c and k are fundamental physical constants. This equation predicts that the increase of I with T is more pronounced as the band lies closer to the Rayleigh band. For example the intensity of a band at 3000 cm-1 increases by less than O.lyo if the temperature increases from 293 to 523 K whereas a band at 1000 cm-l would increase by 6%. Relative intensities are affected too because the band of the standard must have a Raman shift different from that of the compound which has to be measured. In all instances however this effect can be calculated easily and the appropriate correction applied. The Raman scattering cross-section of a band is a function of the polarisability and the hyperpolarisability22 so that * * (5) st = C(B + AF) where B is a function of the polarisability tensor only and A is a function of the polarisability and the hyperpolarisability tensor.is the averaged electric field operating on the scattering molecule and caused by the electric fields of all molecules present in the solution. As is a function of the temperature S-2 will also be a function of the temperature. The significance of this effect is determined by the magnitude of AF which depends on the hyperpolarisability of the scattering compound and on the dipole moments of solvent and solutes and their con-centrations. This effect is difficult to predict and must be analysed experimentally.Except for the population factor all factors mentioned will result in a decrease of the absolute Raman intensity with increasing temperature. 120 100 ij E E 2 80 c. -0 20 40 60 80 100 120 TemperaturePC Fig. 7. Variation of band parameters of the 1020 cm-l band of 1,2-A Width ; B measured area corrected dichlorobenzene with temperature. for thermal expansion; C measured area; and D height. Fig. 7 shows the effect of temperature on the band parameters of the 1020 cm-1 band of 1,2-dichlorobenzene. An increase of the band width is found with increasing temperature. This is no compensation however for the decrease in band height. As a result the band area decreases by 13% over the range from 273 to 373 K. When corrected for the concentration decrease of 11% owing to thermal expansion and the population factor increase of 1% the corrected band is practically constant over the temperature range considered.In Fig. 8 the effect of temperature on the relative intensity of the 1005 cm-1 band of urea in an aqueous solution is shown. ACN was used as an internal standard. A decrease of the relative intensity of 3.5% is found in the range from 273 to 373 K and due to the population factors of urea and ACN a decrease can be predicted of 1%. Hence 2.5% must be attributed to a change of the ratio of QU, to QACN because the influences of the local field effect and the thermal expansion are eliminated in the band ratio April 1983 OF UREA SYNTHESIS UNDER PROCESS CONDITIONS. PART I 491 1 1 20 40 60 80 100 TemperaturePC Fig.8. Variation of ratio of lures loo5 cm-i to IACN 920 om-1 with tempera-ture. Full line was calculated through the points broken line is found if a correction is applied for the variation of the population factors with tempera-ture. Conclusions Raman intensities are affected by a number of instrumental and fundamental factors. The use of an internal standard may compensate for many of the instrumental factors but not for all as was shown by rotating the sapphire entrance window of the high pressure cell. Hence, to avoid erroneous results reproducible window mountings are a necessity. Measuring relative intensities instead of absolute intensities results in better accuracy but as a consequence the use of a scanning system will also result in an increase of the standard deviation.Except for ammonia no effect of pressure on the relative intensities of the compounds present in the urea synthesis is expected. The effect of pressure on the molar intensity of ammonia will need to be studied at various temperatures to make the necessary corrections possible. A new cell will be designed enabling measurements to be made under process conditions without having corrosion or polarisation problems. This effect will be reduced partially by the use of an internal standard but corrections have to be made with, for example the population factor. Laser Raman spectrometry seems feasible for quantitative in situ analysis of the urea synthesis mixture at process conditions if irreproducibility due to trivial instrumental factors is avoided a proper internal standard is used and the necessary corrections for the temperature effect (and in the example of ammonia for the pressure effect) are made.All bands in the spectrum of the mixture will be affected by temperature. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Lemkowitz S. M. Thesis University of Technology Delft 1975. Diller D. E. and Chang R. F. AppZ. Spectrosc. 1980 34 411. Riddell J. D. Lockwood D. J. and Irish D. E. Can. J . Chem. 1972 50 2951. Bus J. Anal. Chem. 1974 46 1824. Bernstein H. J. and Allen G. J . Opt. SOC. Am. 1955 45 237. Long D. A. Milner D. C. and Thomas A. G. Proc. R. SOC. London Ser. A 1956 237 197. Rea D. G. J . Opt. SOC. Am.1959 49 90. Tunnicliff D. D. and Jones A. C. Spectrochim. Acta 1962 18 579. Irish D. E. Jarv T. and Ratcliffe C. I. AppZ. Spectrosc. 1982 36 137. Murphy W. F. Evans M. V. and Bender P. J . Chem. Phys. 1967 47 1836. Oliver B. G. and Davis A. R. Can. J . Chem. 1973 51 698. Saito Y. Machida K. and Uno T. Spectrochim. Acta Part A 1971 27 991. Duncan J . L. Spectrochim. Acta Part A 1971 27 1197. Herzberg G. “Infrared and Raman Spectra of Polyatomic Molecules,” Van Nostrand New York, Buback M. and Schulz K. R. J . Phys. Chem. 1976 80 2478. Gardiner D. J. Hester R. E. and Grossman W. E. L. J . Raman Spectrosc. 1973 1 87. Schwartz M. and Wang C. H. J . Chem. Phys. 1973 59 5258. Eckhardt G. and Wagner W. G. J . Mol. Spectrosc. 1966 19 407. Walrafen G. E. J . Chem. Phys. 1971 55 5137. Whalley E. Proc. 4th Int. Conf. High Pressure Kyoto Japan 1974; Rev. Phys. Chem. Jpn. Special Placzek G. “Marx Handbuch der Radiologie,” 1934 6 205. Koike J. Suzuki T. and Fujiyama T. Bull. Chem. SOC. Jpn. 1976 49 2724. 1945. Issue 1975. Received March 22nd 1982 Accepted October 4th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800485

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

492 Analyst April 1983 Vol. 108 pp. 492-504 Distribution of Zinc Amongst Human Serum Proteins Determined by Affinity Chromatography and Atomic-absorption Spectrophotometry John W. Foote and H. Trevor Delves Chemical Pathology and Human Metabolism Medical Faculty of the University of Southampton South Laboratory and Pathology Block Level D Southampton General Hospital Trernona Road Southampton, so9 4XY Affinity chromatography for albumin has been coupled with electrothermal atomic-absorption spectrophotometry to determine the distribution of zinc between albumin and globulin ligands in normal human serum. The pro-cedure is both simple and rapid and requires only 400 p l of serum for duplicate analyses. There is no alteration in the distribution of zinc between albumin and the globulins during the separation process and the total recovery of zinc from the column is quantitative 98.6%.Albumin-bound zinc and globulin-bound zinc are determined with relative standard deviations of 4.5 and 5.9% respectively. The distribution of zinc obtained is in very good agreement with that found using more complex techniques. Keywords Zinc determination ; serum proteins ; aflnity chromatography ; atomic-absorption spectrophotometry ; kinetic immunoturbidimetry Abnormally low serum zinc concentrations are encountered in a wide variety of clinical condi-tions in which no deficiency of zinc is suspected,l thereby limiting the value of such determina-tions in the detection or assessment of deficiency of this metal. While these findings may on occasion relate to a redistribution of zinc from serum to the liver and other tissues owing to changes in steroid metabolism1 or following the release of protein factors from certain white blood cellsj2 there is little doubt that they are often the consequence of the fact that about 98% of the zinc in serum is bound to protein^,^ principally albumin and a2-macroglobulin,4 and these are themselves subject to changes in concentration in a broad spectrum of diseased states.Consideration of both the concentration and the zinc content of the individual binding species is therefore required for the complete comprehension of an abnormal serum zinc determination. Several techniques have been proposed for the separation of zinc-binding serum proteins including salt fractionation,5,6 electrophoresis,6J anion-exchange chrornat~graphy,~~~~~ gel filtrati0n,~s~s8-~1 sucrose density-gradient centrifugation12 and polyethylene glycol precipita-tion.13 The use of some of these techniques is however associated with disruption of the interaction between zinc and its carrier proteins whilst others are time consuming or require complex equipment that restricts their use to specialised laboratories.There remains a need for a method for the separation of zinc-binding serum proteins that is both accurate and com-plete does not result in the displacement of zinc from its carrier proteins and is sufficiently simple and rapid to allow its application to clinical investigation. We have used pseudo-ligand affinity chromatography of serum proteins on immobilised Cibacron Blue FSG-A to develop a method for the rapid separation of albumin-bound zinc from that bound to or,-macroglobulin.Two fractions are collected the first containing a2-macroglobulin together with other proteins that elute spontaneously from the Cibacron Blue column the second containing albumin which is eluted with thiocyanate. These fractions are analysed for zinc by atomic-absorption spectrophotometry with electrothermal atomisation without further preparation and protein-related matrix interferences upon zinc sensitivity that are encountered in both fractions are readily overcome by the incorporation of bovine albumin into the calibrating standards. Albumin and a,-macroglobulin determined by kinetic im-munoturbidimetry show a virtually complete and highly reproducible separation that is obtained within 30 min.Zinc losses from the albumin immobilised by the resin are negligible and no undue contamination or loss of zinc results from the performance of this method. These features together with the technical simplicity of the chromatographic procedure make the method suitable for the clinical investigation of zinc transport in human serum FOOTE AND DELVES Experimental Affinity Chromatography 493 Ajq5aratus The following apparatus was used an analytical glass column 250 x 6 mm i.d. Altex Type 252-00 fitted with an adjustable plunger Altex Type 252-05 (Anachem Ltd. Luton); a sample valve consisting of two four-way slider valves Altex Type 201-02 (Anachem Ltd,), fitted with a laboratory-made 200-4 sample loop; a Varioperpex I1 Pump 2120 (LKB Produkter AB Broma Sweden) ; a 2138 Uvicord S fitted with filters for use at 280 nm and a 2138-100 flow-cell incorporating a 5-mm light path (LKB Produkter AB) as ultraviolet (UV) monitor; and PTlX tube 0.3 mm i d .Altex 200-34 and couplers (Anachem Ltd.). Samples of whole blood (3-4 nil) collected by venepuncture from an antecubital vein using disposable plastic syringes and stainless-steel needles (Gillette Surgical Isleworth) were trans-ferred into plain 10-ml glass bottles (Labco Marlow) that had been cleaned by immersion in dilute nitric acid (1 + 19) for 2 h rinsed six times with de-ionised water and then dried. For certain experiments that required the use of plasma a portion of the blood collected by venepuncture was transferred into 2-ml polycarbonate tubes [Teklab (ML) Ltd.Sacriston, Durham] containing dried lithium heparin which was known to be free of zinc. Serum or plasma separated by centrifugation was stored in plain 2-ml polycarbonate tubes at -20 "C until required for analysis. Polycarbonate tubes of 10-ml volume [Teklab (ML) Ltd.] required for the collection of column fractions were cleaned by immersion in 2 mM disodium ethylenediaminetetraacetate solution for 1 h before being rinsed six times with de-ionised water and shaken dry. Reagents All reagents were of AnalaR grade (BDH Chemicals Ltd. Poole) unless otherwise stated. Chelex 100 (-400 mesh). Ethylenediaminetetraacetic acid disodium salt. Hydrochloric acid 11 M. Aristar grade. Nitric acid 15 M. n-Octanoic acid.Specially-pure grade. Blue Sepharose CL-6B. Sodium chloride. Sodium hydroxide 4 M AVS volrumetric solution. Sodirum dihydrogen orthophosphate dihydrate. Disodium hydrogen orthophosphate. Sodium t hiocy anate. Tris (hy droxy met hyl) met hy lamine. Bio-Rad Laboratories Watford. Pharmacia Fine Chemicals Hounslow. Procedrure Preparation of reagents. Dissolve 121 g of tris(hydroxymethy1)methylamine (Tris) and 58 g of sodium chloride in 1 1 of de-ionised water to obtain a stock solution containing 1 M Tris and 1 M sodium chloride. Purify this solution by passage down a column of Chelex 100 pre-pared in the sodium form as previously described.14 Prepare daily a starting buffer containing 0.05 M sodium chloride in 0.05 M Tris - hydrochloric acid adjusted to pH 7.4 by diluting the purified stock solution 1 + 19 with de-ionised water and titrating to pH 7.4 with 5 M hydro-chloric acid, Prepare an eluting buffer containing 0.2 M sodium thiocyanate in 0.05 M phosphate buffer at pH 7.4 by dissolving 16.2 g of sodium thiocyanate 1.53 g of sodium dihydrogen orthophosphate and 5.71 g of disodium hydrogen orthophosphate together in 1 1 of de-ionised water.Similarly, prepare a regeneration buffer containing 0.05 M sodium octanoate in 0.05 M phosphate buffer at pH 7.4 by adding 7.9 ml of n-octanoic acid and the same amounts of the phosphate salts stated previously to 800 ml of de-ionised water. Titrate to pH 7.4 with 4 M sodium hydroxide solution and subsequently adjust the volume to 1 1 with de-ionised water. Purify the eluting and regeneration buffers by passage down a Chelex 100 column.Swell 1 g of Blue Sepharose CL-6B in 50 ml of de-ionised water for 15 min and then wash the resin four times over a sintered-glass filter using 250 ml of de-ionised water on each occasion. Store the swollen resin at 4 "C until required for use 494 Analyst Vol. 108 Connect the pump sample valve column and UV monitor in series and arrange to collect fractions at the output of the UV monitor. Clean the apparatus by flushing dilute nitric acid (1 + 19) through the system for 2 h followed by de-ionised water for a further 24 h. Pack the prepared Blue Sepharose CL-6B to a height of 6 cm in the de-contaminated glass column and reduce the dead-space to zero using the adjustable plunger. Equilibrate the resin with 10 bed-volumes of starting buffer.Columns may be maintained in continuous use for up to 5 d without replacement of the resin, Load a 200-pl serum sample into the sample loop and introduce it to the column in a stream of starting buffer at a flow-rate of 0.4 ml min-*. Collect the column effluent as a single fraction (fraction I) until its absorbance at 280 nm falls below 0.04 absorb-ance unit. This fraction contains &,-macroglobulin with other plasma proteins which elute spontaneously from the column. Replace the starting buffer with eluting buffer and continue to collect the column effluent as a second single fraction (fraction 11) until its absorbance is again less than 0.04 unit. This fraction contains albumin almost completely isolated from other serum proteins.Prepare the column for further use by reversing the direction of flow and passing 2 bed-volumes of regeneration buffer up the column to remove residual albumin and trace amounts of lipoprotein that is immobilised by Cibacron Blue FSG-A but is not completely eluted by 0.2 M thiocyanate s01ution.l~ After re-equilibration with a further 10 bed-volumes of starting buffer the column is ready for re-use. FOOTE AND DELVES DISTRIBUTION OF ZN AMONGST Preparation of the apparatus. Protein separation. Perform two separations on each sample of serum. Atomic-absorption Spectrophotometry Apparatus A Perkin-Elmer Model 2380 atomic-absorption spectrophotometer was used with a Perkin-Elmer hollow-cathode lamp. The instrument was fitted with an AS-40 auto-sampler and an HGA-500 graphite furnace using standard (non-coated) furnace tubes.All sample tubes auto-sampler cups and other glassware were cleaned by immersion in dilute nitric acid (1 + 19) for 2 h and were then rinsed six times with de-ionised water. Addi-tionally auto-sampler cups were filled with 2 mM disodium ethylenediaminetetraacetate solution for 1 h and were rinsed a further six times with de-ionised water. Pipette tips were washed individually with a single rinse of 5 M hydrochloric acid followed by 10 rinses with de-ionised water immediately before use. Reagents All reagents were of AnalaR grade (BDH Chemicals Ltd.) unless otherwise stated. Bovine albumin fraction V . Chelex 100 (-400 mesh). Bio-Rad Laboratories. Hydrochloric acid 11 M. Aristar grade.Nitric acid 15 M. Sodium chloride. Zinc chloride solzdion. Sigma London Chemical Co. Poole. Used for atomic spectrophotometry and containing 5 mmol 1-1 of zinc. Procedure Prepare a solution containing 40 g 1-1 of bovine albumin in 0.14 M sodium chloride solution by dissolving 400 mg of albumin in 10 ml of saline solution, containing 8.19 g 1-1 of sodium chloride. Add 3 ml of Chelex 100 resin swollen in the sodium form and mix gently for 2 h. After centrifugation transfer the supernatant liquid to a zinc-free tube for use as a working albumin solution. Prepare a stock zinc standard solution containing 200 p~ zinc in 0.5 M hydrochloric acid as preservative by diluting 2 ml of the zinc chloride standard solution (BDH Chemicals Ltd.) and 2.5 ml of concentrated hydrochloric acid to 50 ml with de-ionised water.Make aqueous working standards by adding 0,0.5,1.0 1.5 and 2.0 ml of the stock zinc stan-dard solution to 0.1 ml of concentrated hydrochloric acid and making the volume of each up to 20 ml with de-ionised water. These solutions contain 0 5 10 15 and 20 PM zinc in 0.05 M hydrochloric acid respectively. Preparation of bovine albumin. Preparation of calibrating standards April 1983 SERUM PROTEINS BY AFFINITY CHROMATOGRAPHY AND AAS 495 Prepare working standards for analysis by adding 75 p1 of the working albumin solution and 75 pl of the aqueous working standards containing 0-20 p~ zinc to 1.73 ml of starting buffer for fraction I or 1.73 ml of eluting buffer for fraction 11 and mixing well. These calibrating standards contain 0 0.2 0.4 0.6 and 0.8 p~ zinc.Analyse the calibrating standards using the instrumental settings given in Table I. Using the resultant standard graphs and the same instrumental conditions determine the concentration of zinc in fractions I and 11. The albumin-associated and the a,-macroglobulin-associated zinc concentrations are obtained from the following Zinc analysis of protein fractions. equations : [Zn]a2-Mncroglobulin = [Zn]Fraction I and [ZnlAlbu~h = [znlFractiOn I1 x where I' is the fraction volume. TABLE I CONDITIONS FOR ATOMIC-ABSORPTION SPECTROPHOTOMETRY Model 2380-Wavelength 213.9 nm; lamp current 15 mA; slit width 0.7 nm (low); deuterium background correction; integration time 5.0 s; and peak height measurement in absorbance mode.Injection volume 5 pl; and replicates x 2. Precision is enhanced if the sample tip is wiped after every 30-45 min of use with a tissue dampened with acetone. Although not seen with aqueous solutions the presentation of samples in 0.05 M phosphate buffer is associated with an auto-sampler memory effect, which is obviated by the use of dilute hydrochloric acid (1 + 99) as the auto-sampler tip flushing fluid. A S-40-HGA -500-Internal gas flow-rate Step Temperaturel'C Ramp time/s Hold time/s (argon)/ml min-l Dry1 * . ,. 90 20 Dry2 110 10 Ash 1 500 20 Ash 2 . . 500 1 Atomise . . 2 100 1 Clean . . 2 500 1 Kinetic Immunoturbidimetry Apparatus [Walter Sarstedt (UK) Ltd. Leicester]. An 8600 Reaction Rate Analyser (LKB Produkter AB) Reagents All reagents were AnalaR grade (BDH Chemicals Ltd.) Ammonia solution sp.gr. 0.88. Fisons Scientific Ltd., Ammonium chloride. 10 300 10 300 6 300 4 60 6 60 6 60 was used with 10 mm i.d. cuvettes unless otherwise stated. Loughborough. Ammonium d i hydrogen ort hop hosp hate. Diammonium hydrogen ort hop hosp hate. Polyethylene glycol 6000. Laboratory-reagent grade. Sheep anti-(human albumin) antiserum. Sheep anti- (human a2-macroglo bulin) antiserum. Standard human serum ORDT 02/03. Nephelometric grade Seward Laboratory London. Nephelome tric grade Seward Laboratory. Hoechst Pharmaceuticals Hounslow 496 Procedwe FOOTE AND DELVES DISTRIBUTION OF ZN AMONGST Analyst Vol. 108 The method described here is a modification of that reported by Price and Spencer.16 Preparation of reagents.Prepare a phosphate buffer reagent containing 0.1 M ammonium chloride in 0.05 M phosphate buffer adjusted to pH 7.0 by dissolving 5.35 g of ammonium chloride 4.03 g of diammonium hydrogen orthophosphate and 2.24 g of ammonium dihydrogen orthophosphate in 800 ml of de-ionised water. Titrate to pH 7.0 with ammonia solution and adjust the volume to 1 1 with de-ionised water. Make a polyethylene glycol reagent containing 40 g 1-1 of polyethylene glycol in phosphate-buffer reagent by adding 40 g of polyethylene glycol to 800 ml of the prepared phosphate-buffer reagent. When the material has been dissolved adjust the volume to 1 1 with further phosphate-buffer reagent. De-gas the solution thoroughly before use. Preparation of antisera.Dilute both the anti-(human albumin) and the anti-(human a,-macroglobulin) antisera 1 + 8 with polyethylene glycol reagent. Allow the diluted antisera to stand at room temperature for 30min and then centrifuge at 3000 revmin-l for 5min to remove any precipitates. Produce calibrating standards for albumin determina-tions by diluting the Hoechst reference serum 1 + 200,300,400,600 and 1200 with phosphate-buffer reagent. Similarly prepare standards for a,-macroglobulin determinations by diluting the reference material 1 + 8 12 16 25 and 50. Calculate the concentrations of albumin and a,-macroglobulin resulting from these dilutions using the specifications supplied with the batch of reference serum in use. Preparation of samples. Samples must be diluted to bring the protein concentrations into the analytical range of the system.Using phosphate-buff er reagent dilute the plasma samples 1 + 300 for albumin and 1 + 25 for a,-macroglobulin determinations respectively. Place 1 .O ml of the polyethylene glycol reagent in a measuring cuvette and add 25 pl of the pre-diluted sample or calibrating standard for albumin determinations or 75 p1 of the pre-diluted sample or calibratihng standard for the a,-macroglobulin determinations. Use the following instrumental settings wave-length 340 nm; reaction course increase; reaction time 60 s; delay function off; injection volume 125 p1. Analyse the calibrating standards by injecting antiserum to initiate the reaction and plot the peak height obtained 60 s after the injection against the protein con-centration.Analyse the prepared samples in the same way reading the peak heights meas-ured after 60 s directly against the calibrating graphs previously obtained to determine the albumin and a,-macroglobulin concentrations. Multiply albumin concentrations by 301 and a,-macroglobulin concentrations by 26 to obtain the concentrations present in the original serum samples before dilution. Preparation of calibrating standards. Determination of albzGrnin and a,-macroglobulin. Results and Discussion Kinetic Immunoturbidimetry Sensitivity and precision Detection limits calculated to be those concentrations that were three times the standard deviation of 10 replicate analyses at the blank level were 90 and 420 ng per 1.0 ml measuring volume for albumin and a,-macroglobulin respectively.When the procedures for sample preparation given previously are used these values represent serum concentrations of 1.1 g 1-l for albumin and 0.14 g 1-1 for a,-macroglobulin being 3 4 % of the normal concentrations of the proteins in human serum The precision of the method has been assessed for both proteins by replicate analyses at two levels (Table 11) worst-case relative standard deviations of 3.1 and 2.9% being obtained for albumin and a,-macroglobulin respectively. TABLE I1 PRECISION OF PROTEIN DETERMINATIONS Albumin a2-Macroglobulin Parameter 77 Number of analyses 10 10 10 10 Mean concentrationlg 1-l . . . . 20.9 51.5 1.65 3.30 Standard deviationlg 1-1 . . I . . . 0.45 1.64 0.047 0.063 Rangelg 1-l .. . . . 20.4-21.4 49.9-54.8 1.60-1.74 3.22-3.38 Relative standard deviation yo . . . . 2.2 3.1 2.9 1. April 1983 SERUM PROTEINS BY AFFINITY CHROMATOGRAPHY AND AAS 497 A ccwacy Determinations of a,-macroglobulin by kinetic immunoturbidimetry (KIT) have been com-pared with determinations by radial immunodiffusion (RID) in 100 serum samples (Fig. 1). The data obtained indicate an excellent agreement between the methods the regression being as follows : [a,-macroglobulin],, = 1.04 [a,-macroglobulin] RID - 0.02 g 1-1 with an associated correlation coefficient of 0.94. 1 2 3 4 5 6 ar2-Macroglobulin by radial immunodiffusion/g I-' Fig. 1. Comparison of a,-macroglobulin determinations in 100 serum samples by radial immunodiffusion and kinetic im-munoturbidimetry.The solid line is the line of identity and the broken line is the regression through the data. 10 20 30 40 50 Albumin by bromocresol green dye bindinglg I-' 1 Comparison of albumin deter- Fig. 2. minations in 40 serum samples by the bromocresol green dye binding method and by kinetic immunoturbidimetry. The solid line is the line of identity and the broken line is the regression through the data. For albumin determinations by kinetic immunoturbidimetry have been compared with determinations by the bromocresol green dye-binding method in 40 sera (Fig. 2) as this tech-nique is established in a considerable number of clinical laboratories. Despite the excellent correlation coefficient obtained r = 0.95 the regression of [albumin] KIT = 1.17 [albumin] RID - 7.48 g 1-1 indicates the presence of a significant bias that is characteristic of comparisons between specific immunochemical or electrophoretic techniques for albumin determination and the bromocresol green meth0d.l' The dye-binding technique both under estimates high albumin concentrations and more significantly over estimates low albumin concentrations1' owing to a non-specific binding of a- and /3-globulins in addition to albumin.lB However, hypoalbuminaemia is a feature of several illnesses that are associated with low serum zinc concentrations in which the relationship between zinc and albumin its major serum-carrier protein may be of particular importance in the detection of a deficiency state of the metal.All albumin concentrations reported here have been determined using the immunoturbidimetric technique described as immunochemical assays are accurate at all the albumin concentrations encountered in clinical practice in both normal and pathological ~amp1es.l~ Atomic-absorption Spectrophotometry The presence of protein-related matrix interferences resulting in enhanced zinc sensitivity in both fraction I and fraction I1 precluded the use of calibrating standards prepared in the relevant buffer.When using such standards zinc recoveries determined by standard additions and expressed as mean -+ standard error (No. of tests) were 148 & 1.8% (4) in fraction I and 157 & 1.8% (4) in fraction I1 (Fig. 3). However the incorporation of bovine albumin into the calibrating standards using the procedure stated previously readily overcame these inter-ferences resulting in zinc recoveries of 100 & 1.8% (4) and 102 & 1.3% (4) in fractions I and 11 respectively (Fig.3) 498 FOOTE AND DELVES DISTRIBUTION OF ZN AMONGST Analyst VoZ. 108 0 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 Zinc added/pmol I-' Fig. 3. Zinc recoveries determined by standard additions to (a) Fraction I and (b) Fraction 11. Least-squares regression equations for the data obtained using calibrating standards prepared in buffer alone (A) show enhanced zinc sensitivity, y = 1.48% + 0.34 andy = 1.57% + 0.49 in Fractions I and 11 respectively. However the data obtained using calibration standards incorporating bovine albumin (B) give regression equations of y = 1.00% + 0.29 for Fraction I andy = 1 .0 2 ~ + 0.36 for Fraction I1 indicating quantitative recoveries of zinc. Data shown are mean values with 95% confidence limits. Protein-related matrix interferences in the fractions obtained following the chromatographic separation of serum have either not been considered at all by previous workers or have been dismissed in view of the dilution of the proteins that results from serum chromatography.11 The data shown here however demonstrate significant matrix interferences that result from the presence of the serum globulins at a dilution of 1 + 13 in fraction I and serum albumin at the higher dilution of 1 + 39 in fraction 11. These matrix effects are similar to those we have observed in 1 + 25 dilutions of whole serum using similar electrothermal conditions which could also be overcome by the incorporation of bovine albumin into the calibrating standards provided that standard (non-coated) furnace tubes were used.14 Here we found that the albumin content of the calibrating standards required to overcome the matrix interferences was not critical as the concentrations of albumin in the working albumin solution used for the additions could be varied between 4 and 40 g 1-1 without any significant effect on performance.We also found that the presence of proteins in pyrolytically coated furnace tubes resulted in suppression of the zinc signal that could not surprisingly be overcome by the simple addition of albumin to the calibrating standards. All zinc determinations reported have therefore, been performed in standard furnace tubes using the electrothermal conditions given in Table I and calibrating standards containing bovine albumin.Affinity Chromatography Protei f i separation The parent serum and the two fractions obtained following its separation under the previously described conditions have each been examined for their protein content by crossed-immuno-electrophoresis (Fig. 4). All of the major protein constituents of the serum were recovered in one or other of the two fractions except a,-lipoprotein that is not however suspected to be a zinc-binding ligand. The majority of the protein species were eluted from the column spontaneously under the starting conditions and these included the zinc metalloprotein a2-macroglobulin and the proteins transferrin and immunoglobulin-G that have been proposed as minor serum zinc ligands.'98J1 Each of these three proteins was identified exclusively in fraction I [Fig.4 ( b ) ] . Only negligible amounts of albumin were identified in fraction I Fig. 4. Crossed immunoelectrophoresis of (a) the parent serum (b) fraction I and (c) fraction 11. Abbreviations alb albumin ; arL a,-lipoprotein ; a,S a,-acid glycoprotein ; a2M a2-macroglobulin ; /3L, fl-lipoprotein ; IgG immunoglobulin G ; and Tf transferrin. [to face Page 49 April 1983 SERUM PROTEINS BY AFFINITY CHROMATOGRAPHY AND AAS 499 whereas this protein was the major constituent of fraction 11 which otherwise contained only a,-acid glycoprotein and /3-lipoprotein together with a number of other species that could not be reliably identified but were present in trace amounts only [Fig.4(c)]. These data show that pseudo-ligand affinity chromatography for albumin on Blue Sepharose CL-6B results in a much more complete separation of albumin from other serum proteins than that achieved with methods that discriminate between proteins on the basis of either electrical charge or molecular size. This separation is obtained using a flow-rate of 84 ml cm- h -l and a bed of resin that is no more than nine times the volume of the sample permitting column geometry that gives rise to very rapid separations with the least possible dilution of the sample. It is important to note that the separation of the serum proteins obtained with Cibacron Blue FSG-A varies with the extent of substitution of the agarose matrix by the dye and the pattern we have reported here results from the use of a resin of relatively low substitution (2 pmol of dye per millilitre of swollen More highly substituted gels incorporating 4.5 pmol of dye per millilitre of gel, as with some proprietary preparations retain haptoglobin and a,-macroglobulin which are released only in the presence of a salt gradient.,l Protein recovery Albumin and a,-macroglobulin recoveries calculated using kinetic immunoturbidimetric determinations of the two proteins in the column effluent were 84 & 1% (10) [mean & 1 standard error (n)] and 99 & 2% (lo) respectively.Cibacron Blue FSG-A binds human albumin more tightly than it does albumins of other speciesl5 and the conditions required for the elution of the human protein are consequently more harsh and may result in its denatura-tion.Although an excellent albumin recovery of 97 2% (10 )was obtained following elution with 0.05 M sodium octanoate the use of this eluent resulted in the complete dissociation of albumin from its bound zinc which was retained in the column and could not itself be recovered without the use of the chelating agent EDTA. The elution of albumin with 0.2 M sodium thiocyanate solution results in an acceptable compromise between desorption efficiency and the avoidance of denaturing column conditions. The incomplete recovery of albumin thereby obtained has not proved a problem in practice because a complete recovery of the albumin-associated zinc has consistently been obtained using this agent (see Blue Sepharose chromato-graphy of purified albumin and also Fig.6). Interaction of Blue Sepharose CL-6B and zinc Blue Sepharose CL-6B consists of a sulphonated polyaromatic dye Cibacron Blue F3G-A, which is covalently coupled to a cross-linked agarose gel. Ionic interactions between proteins and the sulphate groups of the dye contribute to the properties of the resin at neutral pH and raise the possibility of similar interactions between the resin and the zinc present in serum samples.21 To investigate this possibility 200-4 samples containing 5 C,~M zinc chloride in aqueous solution were applied to a 60 x 6 mm diameter bed of Blue Sepharose CL-6B at a flow-rate of 0.4 ml min-l. Less than one third of the zinc was recovered when the resin was equilibrated with de-ionised water demonstrating the ability of the resin to act as a cation-exchanger under conditions of low ionic strength.However when the resin was equilibrated with starting buffer (0.05 M sodium chloride in 0.05 M Tris - hydrochloric acid pH 7.4) the recovery of zinc was quantitative 101 2.6% (6) [mean & 1 standard error (n)] and moreover the elution profile of zinc took the form of a prompt and symmetrical peak (Fig. 5 ) . These observations are consistent with previous reports on the behaviour of Cibacron Blue FSG-A in media of differing ionic strength,Z1 and are taken to indicate the absence of any significant interaction between the resin and zinc under those column conditions that are operative during the passage of a,-macroglobulin transferrin and immunoglobulin-G through the resin.Blue Sepharose chromatography of purijed albumin Investigations of the distribution of zinc amongst the serum proteins clearly requires a method for protein separation that does not disturb the specific zinc - protein relationships if accurate results are to be obtained. Because the precise molecular mechanism by which Cibacron Blue FSG-A binds albumin is not known its effects on the albumin - zinc interaction cannot be predicted. Accordingly 200-4 samples containing 40 g 1-1 of human albumin (Human albumin “essentially globulin free,” Sigma London Chemical Co. Poole) an 600 FOOTE AND DELVES DISTRIBUTION OF ZN AMONGST ArtaZyst VoZ. 108 8 pmol l-1 of zinc have been submitted to chromatography on a 60 x 6 mm diameter column of Blue Sepharose CL-6B.In some experiments the starting buffer was 0.05 M sodium chloride in 0.05 M phosphate buffer adjusted to pH 7.4 whereas in other experiments 0.05 M sodium chloride in 0.05 M Tris - hydrochloric acid also adjusted to pH 7.4 was used. In all other respects the column conditions and procedure were those previously given for the separation of serum. The result obtained with the phosphate starting buffer was surprisingly poor only 68 & 1% (6) [mean 1 standard error ( r t ) ] of the zinc was found with the albumin in fraction I1 while a further 28 & 2% (6) was found in fraction I. There is no doubt that this distribu-tion of zinc resulted from the dissociation of the metal from the albumin to which it was originally bound because the molecular absorbance at 280 nm of the columr effluent indicated that all of the protein eluted in fraction 11.The over-all recovery of zinc of 96 & 2% (6) excluded any exogenous loss or gain of the metal. 0 5 10 15 Time/min Fig. 6. Elution profile of free zinc from Blue Sepharose CL-6B. Bed dimensions 60 x 6 mm diameter; eluent 0.05 M sodium chloride in 0.06 M Tris - hydrochloric acid, pH 7.4; and flow-rate, 0.4 ml min-l. A good recovery of zinc 101 5 2% (6) was also obtained with the Tris - hydrochloric acid starting buffer and furthermore 99 & 2% (6) of the zinc was recovered with the albumin in fraction 11. It is not possible to say whether the small amount of zinc found in fraction I, <2% was due to a leakage through the column of albumin with its bound zinc to the dissociation of zinc from albumin that was retained in the column or to exogenous contamina-tion.Nevertheless this is an acceptable result which supports the earlier observations of Smith et aZ.22 who applied 65Zn - rat-albumin complexes to a Blue Sepharose column and recovered 94y0 of the 65Zn still in association with albumin. However when these workers subjected g5Zn - rat-albumin complexes to ion-exchange chromatography with DEAE-sephadex they found a significant dissociation of zinc from albumin with the displaced zinc being eluted coincidentally at a chloride concentration that under similar conditions would have eluted transferrin from serum.22 Protein - zinc equilibria may also be disturbed during gel filtration. Chesters and Will1* separated 65Zn-labelled serum on Sephadex GlOO and found substantial exchanges between the zinc in the sample and that bound to the column matrix.These effects were minimised only after the resin had been pre-equilibrated with the sample.1° Hence the separation achieved with Blue Sepharose affinity chromatography and Tris -hydrochloric acid as described is clearly superior to those separations obtained by anion-exchange chromatography or gel filtration as there is no loss of zinc from albumin which is the only serum zinc ligand immobilised by the resin and there is negligible affinity of the resin for zinc April 1983 SERUM PROTEINS BY AFFINITY CHROMATOGRAPHY AND AAS 501 Elzction pvojle of zinc found during serum separation on Blue Sepharose CL-6B The elution profile of zinc and the molecular absorbance profile of the column effluent at 280 nm wlkh were obtained during the separation of serum using the described procedure are shown in Fig.6. The absorbance profile shows two peaks the first being due to the serum globulins that elute spontaneously from the column and are collected in fraction I the second being due to the elution of albumin that is collected in fraction 11. Zinc is also eluted as two discrete peaks that relate temporarily to the elution of the globulins and albumin respectively. Globulin-bound zinc (fraction I) is completely separated from albumin-bound zinc (fraction 11). The elution of zinc is virtually completed within 30 min. Typical blank runs through the column gave zinc concentrations of 10 nmol 1-1 in both fractions.- Fraction I ,I Fraction II 1.2- I I1 1 1.0 0.6 9 Q e 0.4 m d 0.2 2 0 0 5 10 15 20 25 30 Time/m i n Fig. 6. Elution profile of (-4) zinc and (B) the molecular absorbance profile a t 280 nm obtained during the separation of serum on Blue Sepharose CL-6B. Bed dimensions 60 x 6 mm diameter; flow-rate 0.4 ml min-l; and sample volume 200 yl. Precision The precision of the method has been assessed by ten replicate pairs of analyses of a normal serum. The relative standard deviations of 5.9 and 4.5% which were obtained for the zinc found in fractions I and 11 respectively are acceptable in view of the procedures involved and are adequate for clinical investigation. These data are shown in Table 111. TABLE I11 PRECISION OF THE SEPARATION AND DETERMINATION OF THE ZINC SPECIES I N NORMAL SERUM Parameter Globulin zinc Albumin zinc Number of determinations .. . . 10 pairs 10 pairs Mean zinc concentration/pmoll-l . . 2.5 15.1 Rangelymol 1-1 . . . . 2.4-2.7 13.616.0 Standard deviation/ ymol 1-1 . . 0.15 0.68 Relative standard deviation yo . . . . 5.9 4.5 Zinc distribution in normal serum The distribution of zinc in the sera of eight male and eight female adults has been deter-mined. All the subjects were in good health and none had taken any medication during the month prior to blood sampling. The data obtained are given in Table IV together with the total concentrations of zinc that were measured in the sera by atomic-absorption spectro-photometry with electrothermal at0misation.1~ The summation of the concentrations of zinc found in the two fractions agreed well with the total zinc concentrations measured in the ser 502 FOOTE AND DELVES DISTRIBUTION OF ZN AMONGST Analyst VoZ.108 and the total recovery of zinc was quantitative 98.6 & 3.9% (16) [mean & 1 standard devia-tion (n)] with a range of 94-1040/,. The concentration of zinc associated with proteins other than albumin was 2.4 & 0.3 pmol 1-1 (16) and that bound to albumin was 10.0 &- 2.0 pmoll-1 (16) representing 19 -+ 3.1% (1624%) and 81 & 3.1% (7646%) of the total zinc respectively. There was no significant correlation between the concentration of a,-macroglobulin and the globulin-bound zinc or between the concentration of albumin and the albumin-bound zinc, r = 0.07 and 0.20 respectively.Neither was there a significant correlation between the con-centration of albumin and the total serum zinc r = 0.17. TABLE IV ZINC DISTRIBUTION IN NORMAL SERUM F F F F F F F F M M M M M M M M Age 47 32 39 . . 57 43 22 28 37 22 28 . . 30 . . 26 33 41 31 . . 31 Zinc/ pmol 1-1 Non-albumin Albumin Total fraction fraction serum 2.0 7.0 8.9 2.4 8.3 10.4 2.3 7.7 9.6 2.5 9.8 11.8 2.0 8.0 10.0 2.7 10.6 13.0 2.5 8.1 10.7 2.1 9.6 12.5 2.2 10.4 12.5 2.2 10.8 13.8 2.9 13.1 16.6 2.2 9.4 12.1 2.4 9.4 11.6 2.5 14.0 17.5 2.2 11.0 13.8 2.7 12.5 16.2 A I \ Zinc recovery yo 101 103 104 104 100 102 99 94 101 94 96 96 102 95 96 94 a ,-Macroglobulin / 2.1 1.7 2.1 1.9 2.0 2.0 2.4 2.0 2.1 1.3 1.8 1.7 1.9 1.4 1.7 1.8 g 1-1 Albumin/ g 1-1 37 39 37 42 39 46 42 39 44 40 39 45 47 47 43 41 The distribution of the serum zinc we have found is similar to that obtained by Giroux et using polyethylene glycol precipitation and interestingly it is also similar to that found by Boyett and Sullivan7 who used Pevikon-Geon Block Electrophoresis (Table V) .The excellent agreement between the value we have found for non-albumin-bound zinc and that obtained for a,-macroglobulin-bound zinc by Giroux et aZ.23 is of particular interest as it sug-gests that the zinc in serum is almost exclusively associated with either albumin or a,-macro-globulin and that any zinc bound to other proteins is of little quantitative significance.In agreement with previous investigators we have found a remarkably constant concentration of zinc associated with globulin (2.0-2.9 pmol l-1) a feature that has been recorded in the sera of both healthy subjects7 and patients with a variety of diseases.l The constancy of the globulin-bound serum zinc has led to speculations on the presence of a physiological control of this zinc component.' TABLE V COMPARISON OF THE DISTRIBUTION OF SERUM ZINC OBTAINED IN THE PRESENT STUDY AND THAT REPORTED FOLLOWING POLYETHYLENE GLYCOL PRECIPITATION^^ AND PEVIKON-GEON BLOCK ELECTROPHORESIS' Affinity chromatography Polyethylene glycol Pevikon-Geon Block Parameter (present study) n = 16 precipitation n = 28 Electrophoresis n = 10 Total-serum zinc/ Albumin-bound zinc/ Non-albumin-bound pmol 1-1 .. 12.6 f 2.6* 12.9 f 1.4* 11.6 f 2.0* pmol l-1 . . * . 10.0 f 2.0 10.5 f 1.2 9.8 f 1.6 zinc/pmoll-' . . 2.4 f 0.3 2.5 f 0.5t 2.1 f 0.4 * Mean f 1 standard deviation. t a*-Macroglobulin-bound zinc April 1983 SERUM PROTEINS BY AFFINITY CHROMATOGRAPHY AND AAS 503 The zinc distribution that we have found does not agree with the frequently quoted assign-ment of 3040y0 of the serum zinc to a,-macroglobulin by Parisi and Valee.* However, according to these investigators a,-macroglobulin contains 4.9-1 1.8 pmol of zinc per gram of protein which is an impossibly high figure when related to a normal serum a,-macroglobulin concentration of 2.0 g 1-1. These values which are certainly artefactual may have resulted from contamination during the isolation of a,-macroglobulin by anion-exchange chromato-graphy which has been shown to result in the displacement of zinc from albumin,, or from the use of heparinised plasma samples that are unsuitable for use in the investigation of zinc binding to proteins in blood (see Suitability of ser2cm and plasma samples).Suitability of serzlm and plasma samples Chesters and Willlo have investigated the distribution of zinc in porcine serum and porcine plasma prepared with either fluoride or heparin. Whereas there was no difference in the distri-bution of zinc in protein fractions obtained from serum or fluoride-treated plasma the use of heparin as an anticoagulant was associated with a significant displacement of zinc from albumin to the globulins.10 We have therefore investigated the distribution of zinc in both serum and heparinised plasma prepared from the same samples of human venous blood.The data obtained are shown in Table VI. Although the recovery of zinc was quantitative with either TABLE VI ZINC DISTRIBUTION OBSERVED IN SERUM AND HEPARINISED PLASMA PREPARED FROM IDENTICAL SAMPLES OF THE BLOOD OF FIVE SUBJECTS Globulin Subject fraction 1 2.5 2 5.0 3 4.6 4 9.4 5 4.4 Plasma zinc/pmol l-f Albumin -+ Albumin globulin fraction fractions 5.7 8.2 4.4 9.4 4.7 9.3 2.8 12.2 7.5 11.9 Globulin fraction 2.0 2.3 2.0 2.2 2.2 Serum zinc/ pmol 1-1 A 7 Albumin + Albumin globulin Total fraction fractions zinc/pmoll-l* 7.0 9.0 8.9 7.7 10.0 9.6 8.0 10.0 10.0 9.4 11.6 12.1 10.8 13.0 13.8 * Determined in serum samples sample there was a variable but significant shift of zinc from albumin to the globulins with heparinised plasma.These data together with those of Chesters and Will,l0 indicate that heparinised plasma is an unsuitable sample for the investigation of the transport of zinc in blood. Conclusions Pseudo-ligand affinity chromatography on Blue Sepharose CL-GB coupled with atomic-absorption spectrophotometry using electrothermal atomisation has proved suitable for the investigation of zinc transport in human serum. No manipulation of the sample is required prior to its separation on the Blue Sepharose column thereby minimising the risk of contamina-tion. The volume of serum required for analysis 200 pl allows the application of the tech-nique to the investigation of children and the separation time is only 30 min.Albumin-bound zinc globulin-bound zinc and the total-serum zinc are each obtained by direct measurement, an advantage over other methods such as the polyethylene glycol precipitation technique in which the a,-macroglobulin-bound zinc value is derived from the difference between the albumin-bound zinc and the total-serum zinc. Zinc loss or contamination by exogenous zinc, is therefore readily detected during each analysis. An automated system has been developed in our laboratory using two identical columns to allow the continuous processing of samples with the best possible precision while columns are taken out of service alternately for regenera-tion.Details can be made available on request. There is little doubt that the globulin-bound zinc concentrations that are obtained using the system described represent sufficiently close approximations of the a,-macroglobulin-bound zinc to be of clinical usage. The scope of ou 504 FOOTE AND DELVES system may possibly be extended to allow a more fundamental investigation of zinc transport in serum by using the more complex separations that are obtained with resins that are more highly substituted than Blue Sepharose CL-6B. The properties of these and the conditions required for their use have been recently reviewed by Gianazza and Arnaud.20s21 We thank Professor B. E. Clayton for her interest and encouragement. We are grateful to BDH Chemicals Ltd.Poole and to the Wessex Regional Health Authority’s Research Com-mittee for financial assistance given to one of us (J.W.F.). Our thanks are also due to Dr. C. R. Lowe for helpful discussions regarding the use of Blue Sepharose CL-6B; Dr. R. B. Ellis for similarly helpful advice regarding the automation of the chromatographic procedure and for the generous loan of chromatographic apparatus; Dr. P. G. Shakespeare for the crossed-immunoelectrophoretograms shown in Fig. 4; Mrs. B. Lloyd for advice on the development of the immunoturbidimetric analyses; Dr. A. M. Ward for the determinations of cc,-macro-globulin by radial immunodiffusion; and to the staff of the N.H.S. Chemical Pathology Laboratory of the Southampton General Hospital for the determinations of albumin by bromocresol green dye binding.The atomic-absorption equipment used in this investigation was on loan from the Perkin-Elmer Corporation Nonvalk CT USA and Bodenseewerke, West Germany. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. References Falchuk K. H. N . Engl. J. Med. 1977 296 1129. Lekakis J. and Kalofoutis A. Clin. Chem. 1980 26 1660. Giroux E. L. and Henkin R. I. Biochim. Biuphys. Acta 1972 273 64. Parisi A. F. and Vallee B. L. Biochemistry 1970 9 2421. Vikbladh I. Scand. J . Clin. Lab. Invest. Suppl. 2 1951 3 1. Dennes E. Tupper R. and Wormel A. Biochem. J. 1962 82 466. Boyett J. D. and Sullivan J. F. Metabolism 1970 19 148. Evans G. W. and Winter T. W. Biochem. Biophys. Res. Commun. 1975 66 1218. Dawson J. B. Bahreyni-Toosi M. H. Ellis D. J. and Hodgkinson A. Analyst 1981 106 153. Chesters J. K. and Will M. BY. J. Nutr. 1981 46 111. Gardiner P. E. Ottoway J. M. Fell G. S. and Burns R. R. Anal. Chim. Acta 1981 124 281. Song M. K. and Adham N. F. Clin. Chim. Acta 1979 99 13. Giroux E. L. Biochem. Med. 1975 12 258. Foote J. W. and Delves H. T. Analyst 1982 107 1229. Leatherbarrow R. J. and Dean P. D. G. Biochem. J. 1980 189 27. Spencer K. and Price C. P. Clin. Chim. Acta 1979 95 263. Webster D. Bignell A. H. C. and Attwood E. C. Clin. Chim. Ada 1974 53 101. Webster D. Clin. Chim. Ada 1974 53 109. Slater L. Carter P. M. and Hobbs J. R. Ann. Clin. Biochem. 1975 12 33. Gianazza E. and Arnaud P. Biochem. J. 1982 201 129. Gianazza E. and Arnaud P. Biochem. J. 1982 203 637. Smith K. T. Failla M. L. and Cousins R. J. Biochem. J. 1979 184 627. Giroux E. L. Durieux M. and Schechter P. J. Bioinorg. Chem. 1976 5 211. Received October 13th. 1982 Accepted November 17th. 198

ISSN:0003-2654    

DOI:10.1039/AN9830800492

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analjst April 1983 VoL 108 pp. 505-509 505 Spectrofluorimetric Determination and Thin-layer Chromatographic Identification of Selenium in Teresa Moreno- Dominguez Concepcion Garcia- Moreno and Abel Marine-Font Department of Bromatology Toxicology and Chemical Analysis Faculty of Pharmacy University of Salamanca, Salamanca Spain A method is described for the determination of selenium in foods. Digestion of the samples and fluorimetric determination are based on the method of Michie et al. with minor modifications. To confirm the results from a qualitative point of view a new thin-layer chromatographic procedure is proposed. Keywords Selenium determination ; food analysis ; spectrojluorimetry ; thin-layer chromatography ; 2,3-diaminonaphthalene Spectrofluorimetry is a commonly used method of choice for the determination of selenium in foods 2,3-diaminonaphthalene (DAN) being the fluorescence reagent that gives the best result^.^-^ Taking into account that the samples must be previously digested and that under such conditions selenium derivatives that are easily volatilised may be formed the first prob-lem to be dealt with is that of possible losses of selenium in this process.To avoid such losses the influence of a reflux condenser was studied operating on the same sample under different conditions (a) without a condenser; (b) with a condenser during the digestion process itself but without it in the following steps of elimination of nitric and perchloric acids; and (c) with a condenser in the digestion step and the other steps.Procedure (b) is the most suitable. In the digestion process different reagents are introduced which seem to influence the fluorescence intensity of the selenium - DAN complex; this led us to consider the need to sub-ject the standard selenium solutions to the digestion process as other workers have d ~ n e . ' ~ ~ To confirm the identification of selenium it was thought convenient to carry out verification by thin-layer chromatography on the solutions with which the spectrofluorimetric study was performed. Even though the reaction between selenium and DAN seems to be fairly elective,^^^-^ thin-layer chromatography confirmed the absence of interferences. In this paper a method is described for the determination of selenium in foods. Digestion of the foods and fluorimetric determination are based on the method of Michie et aZ.,l with minor modifications.A novel procedure is proposed for the identification of selenium by thin-layer chromatography which is thought to confer even greater reliability on the results. Experimental Apparatus Spectrojuorimeter. Aminco SPF 125. Rotovapor. Buchi R.A. with W-240k bath. pH meter. Radiometer pH-M-62. Thin-layer chromatographic plates. For five 200 x 200 mm plates homogenise 15 g of cellulose powder MN-300 and about 100 ml of water in a high-speed blender. Activate the plates at 110 "C for 30 min before use. Heating mantles. Selecta. Reagents EDTA solutiort 0.02 M. DAN reagent. Dissolve 0.05 g of DAN in 50 ml of 10% V/V sulphuric acid. Prepare im-mediately before use and purify by shaking with four 10-ml volumes of cyclohexane discarding the organic layer each time 506 MORENO-DOM~NGUEZ et al.SPECTROFLUORIMETRY Analyst Vol. 108 Dissolve 37 g of glycine in water add 20 ml of concentrated hydrochloric and dilute to 500 ml with water.l Transfer a Titrisol ampoule (Merck) containing 1.000 & 0.002 g of selenium (SeO,) into a 1-1 calibrated flask and dilute to volume with 10% V/V sulphuric acid. (a) 1 pg ml-l solution; dilute the stock solution with 10% V/V sulphuric acid. (b) 0.01 pg ml-l solution; prepare fresh daily from solution (a) by dilution with doubly distilled water. Glycine - hydrochloric acid bufer p H 2.4. Seleniwn stock solution 1 mg ml-l. Selenium working solutions. Procedure Sample digestion Weigh between 1 and 5 g of sample depending on the approximate level of selenium thought to be in the food in question into a 250-ml Kjeldahl flask.Add 5 ml of concentrated nitric acid 10 ml of 70% m/m perchloric acid 10 ml of doubly distilled water and leave to stand for 12-16 h at room temperature. Fit a reflux condenser to the Kjeldahl flask add 5 ml of 70% perchloric acid and boil until yellow nitrous vapours are no longer evolved. Successively add 5-ml volumes of concentrated nitric acid until total destruction of the organic matter has taken place (the nitric acid is added to compensate for the losses of oxidising mixture and to avoid carbonisation). Once a clear and transparent solution has been obtained cool and remove the condenser. Add 5 ml of doubly distilled water and 10 ml of concentrated sulphuric acid boil (a yellow colour will appear) until the solution becomes clear again cool add 15 ml of doubly distilled water and heat again until the appearance of white fumes of sulphur trioxide.In this way it is possible to eliminate trace amounts of nitric and perchloric acid. Flz4orimetry It is well known that only selenium(1V) forms a complex with DAN.lJO However owing to the strongly oxidising conditions of the digestion process part of the selenium may be found in the form of selenium(VI) which must be reduced to selenium (IV) before proceeding to the formation of the complex. This is carried out as follows. Add to the digestion liquid 1 ml of 30% m/V hydrogen peroxide and boil for 10 min. Cool and repeat the treatment until fumes of sulphur trioxide appear again; this ensures the com-plete reduction of the selenium.Cool and transfer the contents of the Kjeldahl flask into a 100-ml calibrated flask. Wash the Kjeldahl flask with doubly distilled water add the wash-ings to the digestion liquid and dilute to 100 ml. Transfer the contents of this flask into a 250-ml beaker. Add 10 ml of 0.02 M EDTA solution. Using a pH meter adjust the pH to 2.4 initially using concentrated ammonia solution and then 5 N ammonia solution. Add 25 ml of glycine -hydrochloric acid buffer. Add 2 ml of freshly prepared DAN reagent and heat in an oven at 60 "C for 1 h. Cool the beaker rapidly in crushed ice transfer the contents into a 250-ml separating funnel, add 5 ml of cyclohexane and shake vigorously for 90 s.The selenium - DAN complex is extracted into the organic phase. The first aqueous frac-tion that leaves the separating funnel is discarded and the second fraction comprising the whole of the organic phase and part of the aqueous phase is collected in test-tubes allowing the two phases to separate well. Using the cyclohexane phase measure the fluorescence with a maximum excitation wave-length of 360 nm and a maximum emission wavelength of 520 nm. The results are expressed in terms of the amount of selenium in the solution read from a calibration graph that should be constructed for each determination or set of simultaneous determinations. The excitation and emission spectra of the organic phase were measured to ensure that they coincided with those of the fluorophore formed from standard selenium.Preparation of calibration graph Take aliquots of the selenium working solutions with 1.00 (solution a ) and 0.01 pg ml-l of selenium (solution b) so as to contain 0.05 0.25 0.50 0.75 1.00 1.25 and 1.50 pg of seleniu April 1983 AND TLC IDENTIFICATION OF SE IN FOODS 507 and prepare a blank free from selenium. Subject these solutions and the blank to the same process as described for the samples including digestion formation and extraction of the selenium - DAN complex and fluorimetric determination. T hin-lay er chromatography After carrying out the fluorimetric determination concentrate the cyclohexane phase in a Rotovapor at 60 "C until the residue is nearly dry. Add 0.5 ml of cyclohexane and use this solution for thin-layer chromatography.On the same thin-layer chromatographic plate spot a standard selenium sample and a standard and sample superimposed (co-chromatography) . Develop the plate with ethanol - 25% ammonia solution (70 + 30). Upon exposure to UV light the Se - DAN complex gives a pink fluorescence at 360 nm. In all instances in which a fluorimetric response was registered it was possible to confirm the presence of selenium by thin-layer chromatography. TABLE I REPRODUCIBILITY OF METHOD APPLIED TO NINE SELENIUM DETERMINATIONS ON A SINGLE BATCH OF BROWN RICE Sample No. Seleniumlpg g-l 1 0.07 2 0.07 3 0.08 4 0.07 5 0.07 6 0.08 7 0.08 8 0.09 9 0.08 Mean * . 0.076 Standard deviation . . 0.207 Relative standard deviation . . 9.22% Results and Discussion In order to verify the reproducibility of the method selenium was determined nine times on different days in the same batch of brown rice (Table I).The relative standard deviation was 9.22%. To determine the recovery 0.05 0.08 0.10 0.15 and 0.20 pg g-l of selenium were added to samples of the same rice and four assays were carried out for each amount of selenium added. The results are given in Table 11. The average recovery of the 20 assays was 91.32%. TABLE I1 RECOVERY OF SELENIUM ADDED TO DIFFERENT SAMPLES OF THE SAME BROWN RICE Selenium Recovery % Average addedlpg g-l I A- recovery yo 0.05 80.0 100.0 75.0 100.0 88.75 0.08 100.0 88.0 75.0 87.5 87.62 0.10 100.0 110.0 70.0 100.0 95.00 0.15 106.0 107.0 74.0 74.0 90.25 0.20 125.0 75.0 100.0 80.0 95.00 Over-all average 91.32 The method was applied to different foods of animal and vegetable origin and satisfactory results were obtained.ll The method has been applied to blood samples with equally satis-factory results.12 I t seems of importance to point out that in all of the chromatograms both those correspond-ing to the selenium standard and those of the real samples two stains appeared on illuminating the plates with UV light at 360 nm one of which fluoresced pink (corresponding to the selenium - DAN complex) and the other green.The latter also appeared in the spot corre-sponding to the blank. In order to confirm whether the substance or substances responsibl 508 ---- --_ ('-B-I f0 c') -. <-%J (0) <E) 1 2 3 - - ->. u) CI .-for this green stain interfere in the fluorimetric determination the stains were removed eluted with 5 ml of cyclohexane and their fluorescence spectra measured.Fig. 1 shows the chroma-togram used for elution and Figs. 2 and 3 show the corresponding fluorescence spectra. The spectra corresponding to the substance or substances that fluoresce green in the chroma-togram (Fig. 1) do not interfere in the spectrofluorimetric determination of selenium as there is no fluorescence in the region of the spectrum of the selenium - DAN complex (Figs. 2 and 3). Also the spectra belonging to the pink stains of the standard and of the problem sample (Figs. 2 and 3) coincide and they also correspond to that of the unchromatographed selenium -DAN complex. 300 335 400 450 520 590 Wavelengthlnm Fluorescence emission spectra for the removal in cyclohexane of substances shown in the chromatogram in Fig 1 and the standard selenium - DAN complex not chroma-tographed previously 1 substance A (in the chromatogram gives a green fluorescence) ; 2 substance B (in the chromato-gram gives a green fluorescence); 3 substance C (in the chromatogram gives a green fluorescence) 4 substance D (in the chromatogram gives a pink fluorescence) ; 5 substance E (in the chromatogram gives a pink fluoresceme); and 6, standard selenium - DAN complex not chromatographed previously.Fig. 3 April 1983 AND TLC IDENTIFICATION OF SE IN FOODS References 509 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Michie N. D. Dixon E. J. and Bunton N. G. J . Assoc. Off. Anal. Chem. 1978 61 48. Watkinson J. H. Anal. Chem. 1966 38 92. Hall R. J. and Gupta P. L. Analyst 1969 94 292. Whetter P. A. and Ullrey D. E. J . Assoc. Off. Anal. Chem. 1978 61 927. “Official Methods of Analysis of the Association of Official Analytical Chemists,” Twelfth Edition, Association of Official Analytical Chemists Washington DC 1975 p. 455. Clarke W. E. Analyst 1970 95 65. Allaway W. H. and Cary E. E. Anal. Chem. 1964 36 1359. Lott P. F. Cukor P. Moriber G. and Solga J. Anal. Chem. 1963 35 1159. Cukor P. Walzcyk J. and Lott P. F. Anal. Chim. Acla 1964 30 473. Parker C. A. and Harvey L. G. Analyst 1962 87 558. Moreno-Dominguez T. Tesina Exp. Licenc. Salamanca 1982. Mateos-Notario P. personal communication. Received July 26th 1982 Accepted November 2nd 198

ISSN:0003-2654    

DOI:10.1039/AN9830800505

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

510 Analyst April 1983 Vol. 108 pp. 510-514 Chromatographic Separation of Chlorinated Hydrocarbons Using Columns of Silica Gel of Varying Degrees of Porosity and Activation Vittorio Contardi Renzo Capelli Gilda Zanicchi and Marco Drago Istituto di Chimica Generale Grufifio Ricerca Oceanologica-Genova Universitd da Genova Genoa. Italy A rapid method for the separation of chlorinated pesticides such as the di-chlorodiphenyl trichloroethane (DDT) group dieldrin and from polychlorinated biphenyls by porous deactivated silica gel column chromatography is described. The complete separation and quantitative recoveries obtained allow a correct determination of these compounds. This procedure is help-ful in multi-pesticide residue analysis. Keywords ; Polychlorinated biphenyls ; chlorinated pesticides ; porous silica gel column chromatografihy When it is necessary to identify and quantify certain chlorinated pesticides such as dichloro-diphenyltrichloroethane (DDT) t etrachlorodiphenylethane (TDE) dichlorodiphen yldichloro-ethylene (DDE) aldrin dieldrin benzene hexachloride (BHC) and hexachlorocyclohexane (HCB) in the presence of polychlorinated biphenyls (PCBs) which are common in samples taken from the environment the most suitable methods are still considered to be gas chroma-tography with an electron-capture detector or gas chromatography - mass spectrometry coupling.However if such compounds are injected together into a column there is interfer-ence which results in a superimposition of characteristic peaks thereby impeding an accurate determination.Using available gas chromatographic packed columns it is in fact imposs-ible to distinguish the characteristic peaks of pesticides such as DDE and TDE from those of other PCBs. Some workers have proposed carrying out a separation of such pesticides from PCBs before analysis by gas chromatography. Among the various procedures presently used, those giving the best results are the ones that utilise column chromatography and which in particular employ silica gel as an adsorbent. Armour and Burke1 separated PCBs from DDT and its analogue and other common chlorin-ated pesticides by column chromatography using differently deactivated silicic acid - Celite, eluting PCBs with light petroleum and pesticides with a mixture of acetonitrile hexane and dichloromethane.Snyder and Reinert2 described the separation of small amounts of pesticides and PCBs utilising silica gel activated at 200 "C as adsorbent eluting the mixture with pentane and benzene. They pointed out that the efficiency of separation diminishes when the amount of chlorinated hydrocarbon introduced into the column is increased. Leoni3 separated various chlorinated and phosphorated insecticides from PCBs using silica gel activated at 130 "C and deactivated with 5% water eluting with hexane 60y0 benzene in hexane and 50% ethyl acetate in benzene. Collins et at?.* separated some PCBs HCB aldrin o,p-DDT and a part of @,@'-DDE from the remaining parts of $,p'-DDE P,$'-DDT and p,$'-TDE using silica gel activated at 500 "C and deactivated with 2.5% water and subsequently eluting with hexane and a solution of 10% diethyl ether in hexane.McClure5 obtained good separation of PCBs from DDT and analogous compounds using silica gel activated at 200 "C and balanced using a solution of 5% diethyl ether in benzene, eluting the PCBs with hexane and the chlorinated pesticides with a solution of 5% diethyl ether in benzene. The efficiency of these separations was not reported. Masumoto6 studied, in detail factors affecting the separation of PCBs and p,p'-DDE by silicic acid chromatography and pointed out the limits of Armour and Burkes meth0d.l Picer and Ahel' described a simple procedure for the separation of PCBs from DDTs dieldrin and BHC on a miniature silica gel porous column. The object of this work was to overcome the various difficulties to be found in the separation of mixtures of PCBs and chlorinated pesticides particularly DDE.These difficulties hav CONTARDI CAPELLI ZANICCHI AND DRAG0 511 been demonstrated clearly by the workers cited. Therefore we have examined the influence of various parameters such as porosity and degree of activation of the adsorbent on the column chromatographic separation of these compounds. Experimental All chemicals used were of analytical-reagent grade. The column chromatographic adsorbents used were as follows Kieselgel 40,70-230 mesh ASTM; Kieselgel 60 70-230 mesh ASTM; and Kieselgel 100 70-230 mesh ASTM all from E. Merck Darmstadt West Germany. Chlorinated hydrocarbons for standard solutions were P,$’-DDT $,p’-TDE P,P’-DDE, aldrin and dieldrin all from Riedel-De Haen Hannover West Germany all >99% purity.Two polychlorinated biphenyls (PCBs) Fenclor 54 (similar to Aroclor 1254) with 5244% by mass of chlorine and Fenclor 64 (similar to Aroclor 1260) with 61% by mass of chlorine were used. The solvents used hexane and benzene (all pro analysi Merck) were distilled and controlled by gas chromatography. Apparatus Glass chromatographic columns were of 300 mm length 10 mm i.d. with coarse fritted discs and PTFE stop-cocks at the bottom. Gas chromatography was performed with a Varian Aerograph 2240 chromatograph equipped with a tritium electron-capture detector and glass gas-chromatographic columns containing the following 4% SE-30 on Chromosorb W HP, 80-100 mesh; 3% OV-17 on Chromosorb W HP 80-100 mesh; and 5% QF-1 on HP Chromo-sorb W AW,DMCS,80-100 mesh.All were 1.8 m in length and 3 mm i.d. The flow-rate of the carrier gas (high-purity nitrogen) was 40 ml min-l. Materials and Reagents All products were from Caffaro Milan Italy. Caution-Benzene is highly toxic and appropriate precautions should be taken. 0% H-0 0.75% H90 100 75 50 25 100 75 50 25 100 100 100 75 75 75 50 50 50 10 nm 25 25 25 IE IIE IllE IE IIE IllE IE IIE IllE __ IE IIE IllE Silica gel porosity 4 nm TDE; NDDT;. dieldrin Fig. 1. Elution of aldrin dieldrin pp’-DDT,pp’-TDE,pp’-DDE and PCBs 1 + 1 V/ V (mixture of Fenclor 54 and Fenclor 64) with various degrees of deactivation and various porosities of silica gel. Eluent I (IE), 40 ml of hexane; eluent I1 (IIE) 40 ml of hexane; and eluent I11 (IIIE) 50 ml of benzene.Range 0-1.6% of water 512 Analyst VoZ. 108 Procedure Sets of chromatographic columns of three types of silica gel of different porosity were pre-pared. One series was filled with Kieselgel40 (average pore diameter 4 nm) activated for 24 h at 200 "!3 and then deactivated with increasing amounts of water 0.7 1.5 2.5 and 3.5%, respectively. A second series was filled with Kieselgel60 (average pore diameter 6 nm) and a third series was filled with Kieselgel 100 (average pore diameter 10 nm); both were activated and deactivated in the same manner as described for Kieselgel40. The deactivation of silica gel was achieved by adding the correct amount of water to small pieces of filter-paper (What-man 0/42) which were distributed in the adsorbent rather than adding the water directly on to the silica gel.The container (Petri dish) was then subjected to rotation for approxi-mately 1 h in order to achieve uniform contact with all of the silica gel particles and to ensure that the transfer of water to the adsorbent was homogeneous. The packing of the columns was carried out according to the technique of Snyder and Reinert,2 ie. dispersing the silica gel in hexane and transferring to the column the amount of adsorbent necessary to form a 200-mm layer (11 g for Kieselgel 40 8.6 g for Kieselgel 60 and 7.1 g for Kieselgel 100). A mixture of various chlorinated hydrocarbons at concentration levels found in commonly consumed fish taken from the Ligurian Sea* was placed on each column.This standard mixture was made by dissolving in 3 ml of hexane the following amounts of chlorinated pesticides andPCBs aldrin 2 x lo4 g ; dieldrin 3 x g ; $,p'-TDE, 5 x g. The mixture was eluted in each column in the following manner eluent I 40 ml of hexane; eluent 11 40 ml of hexane; and eluent 111 50 ml of benzene. The recovered eluents were concentrated to 3 ml and injected into the gas-chromatographic columns at different polarities described previously. Identity of organochlorines was assumed from the retention times on two columns. Quantification was based on peak-height measure-ment; PCBs were quantified by reference to standards mixtures of Fenclor 54 and 64 the contribution of each being calculated using properly selected peaks while solutions of $,PI-isomers were utilised as standards for DDT TDE and DDE.CONTARDI et al. CHROMATOGRAPHY OF CHLORINATED g; $,$'-DDE 10 x g ; p,$'-DDT 20 x 10-6 g; Fenclor 54 40 x g ; and Fenclor 64 40 x Silica gel porosity 3.5% H20 100 2.5% H20 75 75 100 50 50 4 nm 25 25 IE IIE IE IIE IllE 100 100 8 75 75 50 $ 50 a 25 25 d 0 6 nm IE IIE IllE IE IIE 100 I I 75 75 50 50 10 nm 25 25 IE IIE IllE IE IIE IllE Fig. 2. As Fig. 1 for 2.6 and 3.5% of water April 1983 HYDROCARBONS USING COLUMNS OF SILICA GEL 51 3 In order to verify the validity of the results that were also obtained on real samples silica gel (Kieselgel 40 deactivated with 1.5% of water) which allowed the best recoveries and complete separation between DDE and PCBs was utilised in the analysis of fish extracts and, also fortified samples.The chlorinated hydrocarbon mixtures that were used to fortify fish samples (striped mullet fished in the Genoa area) were the same that were used for studying the chromatographic behaviour of silica gel (diluted 5-fold with hexane). Using a hot extraction procedure in Soxhlet apparatus a known mass of the macerated flesh tissue (30 g) was freeze dried and extracted for 8 h with hexane. The solvent was gently evaporated under reduced pressure and the mass of the extractable organic material (EOM) was determined. An EOM mass of less than 3 g was dissolved in 3 ml of hexane and was subjected to clean-up by partition between hexane and a~etonitrile.~ The clean-up of extracts was completed by elution on a small column of Florisil with a hexane - diethyl ether mixture 96 + 4.1° The solvent was then evaporated and extracted c illected with 3 ml of hexane and chromatographed on a silica gel column in the same way as described previously.Gas-chromatographic analysis was carried out on concentrated extracts (3 ml). Results and Discussion The results obtained for the various types of silica gel are illustrated in Figs. 1 and 2 while Fig. 3 illustrates the recovery percentages obtained for the three types of silica gel for each degree of deactivation of the adsorbent. The results obtained on fortified fish samples are represented in Table I. The first column gives average values from three samples of striped mullet 14-16 cm in length 3149 g in mass. Fish extracts (3 ml) were fortified by adding an equal volume of the solution used when studying the chromatographic behaviour of silica gel at various porosities diluted 1 + 5 with hexane.An examination of Figs. 1 and 2 clearly demonstrates the influence of two parameters the porosity of the silica gel and its degree of deactivation. An analysis of the function of porosity shows that the best separation and the best recovery were obtained with the finest porosity (4 nm pore diameter). This behaviour results from the fact that the surface area of the adsorbent is inversely proportional (when granular dimensions are constant) to the average diameter of the pores. Of the three types of silica gel used Kieselgel 40 shows the greatest surface area (-500 m2 g-l) and demonstrates therefore the greater adsorption capacity.At such an elevated degree of activity we have the exclusion properties that for this degree of porosity allow the separation of compounds having a low relative molecular mass such as the 1 0% H20 0.7% H20 1.5% H20 / / / \ / 8 0 - \ 70 -60 >/' 50 50 50 K40 K60 K100 K40 K60 KlOO K40 K60 K100 -/ / 2.5% H20 3.5% H20 100 rTK 6ok-- I 1 I I I K60 K100 50' ' 5 0 ' K40 K60 KlOO K40 -. -. , PCB; aldrin; DDE; -_ -- DDT; TDE; --- ,dieldrin Fig. 3. Recovery of chlorinated hydrocarbons obtained from Kieselgel40 (K 40) Kieselgel60 (K 60) and Kieselgel 100 (K 100) with various degrees of deactivation 514 CONTARDI CAPELLI ZANICCHI AND DRAG0 chlorinated hydrocarbons which we have considered (relative molecular mass varying from Water increases the rate of migration of the various components of the mixture in the column by masking the active sites of the silica gel.Therefore when the amount of water is increased we see the appearance of more polar pesticides (such as DDTs) even in the first fractions. Such an effect is less sensitive in Kieselgel40 because fine-pore adsorbents display retention and efficiency characteristics that are less dependent on water content than wide-pore silica gels (Kirklandll). The operational conditions giving the best results on pure solutions if applied to residue analysis on real samples gave the same kind of separation but slightly lower recoveries (although very satisfactory compared with the low concentration of chlorinated hydrocarbons in the samples being considered).190 to 500). TABLE I RECOVERY OF VARIOUS CHLORINATED HYDROCARBONS FROM A FORTIFIED EXTRACT OF STRIPED MULLET The silica gel used was Kieselgel40 70-230 mesh 4 nm pore diameter deactivated with 1.5% water. The striped mullet was 14-16 cm in length and 31-49 g mass. Recovery % 1 Fraction I Fraction I1 Fraction I11 Fortified extract ,-’- 7-7 Concentrations in fish concentrations/ Trial Trial Trial Trial Trial Trial Trial Trial Trial (medium values of three trials)/g ml-l g ml-1 1 2 3 1 2 3 1 2 3 Aldrin 0.1 x 10-8 . . . . . . . . . . 6.8 X lo-* - - - 86 81 84 - - -Dieldrin 0 2 x lo-’ 10.2 x 10-8 - - - - - - 95 92 89 10.0 x 10-7 - - - - - - 97 97 96 4.0 x 10-7 - - - - - - 93 93 89 pIp’-DD?,’3.3 x lo-’ . . . .. . . . . 5.7 x 10-1 - - - - - - 90 88 87 ~,~’-TDE 2.3 x 10-7 . . . . . . . . . . p DDE 2.4 x 10-7 &is (Fenclor 54 - Fencidr 64 i’ + ij,j4.3 i. 10-7 100.1 x 10-1 - - - 101 98 96 - - -Conclusions The optimum conditions for complete chromatographic separation of various chlorinated hydrocarbons have now been identified. For p,$’-DDE with certain polychlorinated biphenyls, for which the margin of chromatographic separation is very small we were able to achieve a complete separation using Kieselgel 40 deactivated with 1 .5y0 water. Another important result is that under these experimental conditions with the exception of aldrin (90% recovery), we achieved approximately 100% recoveries. Further the method we have employed is simpler and permits better separation than the procedures cited and is completely reproducible.This method has been utilised in the determination of trace amounts of pesticide residues in samples taken from a marine environment with very good separations and recoveries (80-100%). The non-separation of aldrin from PCBs is not inconvenient in the determination unless significant amounts of PCBs with a low degree of chlorination are present. This is because the retention times of aldrin are different to those of polychlorinated biphenyls with a higher chlorine content. We express our thanks to Professor R. Ferro Direttore dell’Istituto di Chimica Generale dell’Universit8 di Genova for the useful suggestions given during this work and the preparation of the manuscript. This work was supported by the Italian Consiglio Nazionale delle Ricerche under the Programme of Researches on Oceanography and Marine Shelves. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Armour J. A. and Burke J. A. J . Assoc. Off. Anal. Chem. 1970 53 761. Snyder D. and Reinert R. Bull. Environ. Contam. Toxicol. 1971 6 386. Leoni V. J . Chromatogr. 1971 62 63. Collins G. B. Holmes D. C. and Jackson E. J. J . Chromatogr. 1972 71 443. McClure V. E. J . Chromatogr. 1972 70 168. Masumoto H. T. J . Assoc. Off. Anal. Chem. 1972 55 1092. Picer M. and Ahel M. J . Chromatogr. 1978 150 119. Contardi V. Capelli R. Pellacani T. and Zanicchi G. Mar. Pollut. Bull. 1979 10 307. “Pesticide Analytical Manual,” Volume I US Department of Health Education and Welfare Food Solomon J. Anal. Chem. 1979 51 1861. Kirkland J . J. J . Chromatogr. 1973 83 149. and Drug Administration Washington DC 1971 Section 21 1.14a. Received May 4th. 1982 Accepted October 26th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800510

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst April 1983 Vol. 108 pp. 515-520 Solid - Liquid Separation after Liquid - Liquid Extraction Spectrophotometric Determination 515 of Copper by Extraction of its l-Phenyl-4,4,6-trimethyl-(1H,4H)-pyrimidine-2=thiol into Molten Naphthalene Abdul Wasey Raj Kumar Bansal and Bal Krishan Puri" and Masatada Satake Department of Chemistry Indian Institute of Technology New Delhi 11001 6 India Faculty of Engineering Fukui University Fukui 910 Japan A selective spectrophotometric method has been developed for the determina-tion of copper after extraction of its l-phenyl-4,4,6-trimethyl-( 1H,4H)-pyrimidine-2-thiol (PTPT) complex into molten naphthalene. The optimum pH range for the extraction is 7.4-11.7. The solid naphthalene containing the copper - PTPT complex is separated by filtration and dissolved in chloro-form.The absorbance is measured at 400 nm against a reagent blank. Beer's law is obeyed in the concentration range 5.0-60.7 pg of copper in 10 ml of chloroform solution. The molar absorptivity and sensitivity are 1.23 x lo4 1 mol-l cm-1 and 0.008 11 pg cm-2 respectively. The interference of various ions has been studied in detail. Conditions have been established for the determination of copper in goat liver human hair and certain alloys. Keywords Solid - liquid separation ; spectrophotonzetry ; copper determina-tion ; biological materials analysis ; alloy analysis Several pyrimidinethiols have been synthesised and used successfully as analytical reagents.lJ In this work l-phenyl-4,4,6-trimethyl-( 1 H,4H)-pyrimidine-2-thiol has been utilised in the extraction and spectrophotometric determination of copper by the technique of solid - liquid separation after liquid - liquid extraction.The main advantages of this method are that equilibrium distribution in the two phases is attained in a few seconds owing to the high temperature (about 90 "C) and the metal chelates are dissolved merely by contact with molten naphthalene. As only a small amount of the organic solvent (2 g) is required for complete extraction the sensitivity of the method is enhanced as the whole of the organic phase may be taken for the analy~is.~ This technique is especially useful for the extraction of those metal iors which form complexes with the reagent at high temperatures4 or have low solubilities in aqueous solution.Experimental Equipment 191 atomic-absorption spectrophotometer were used. An Elico pH meter a Pye Unicam SP-700 recording spectrophotometer and a Pye Unicam Reagents Doubly distilled water and analytical-reagent grade acids and salts were used throughout, unless stated otherwise. Standard copper solution. Dissolve 0.1248 g of CuS04.5H,0 in water and dilute to 1 1. Standardise by the usual methods5 Sodiwn hydroxide solution 1 M. Acetic acid 1 M. Ammonia - amntoniztm acetate b u f e r p H 9.4. * To whom correspondence should be addressed. Mix 0.2 M ammonium acetate and 0.2 M ammonia solutions in the proportions 1 + 2 516 WASEY et al. SOLID - LIQUID SEPARATION AFTER LIQUID - Analyst VoZ. 108 Check the purity spectrophotometrically before use.Naphthalene. Chloroform. Check the purity spectrophotometrically before use. 2-MethyZ-2-thiocyano$entun-4-one. Add mesityl oxide (98 g 1 mol) to 50 ml of sulphuric acid (0.5 M) in a beaker over a period of 15 min at about 15 "C. Add a solution of ammonium thiocyanate (76 g 1 mol) in water (100 ml) rapidly keeping the temperature at about 15 "C. After stirring for 15 min allow the reaction mixture to stand separate the upper oily layer and wash it with water until free from acid. Prepared by the method of Mathes and co-w~rkers.~~~ Place 1 mol of 2-methyl-Z-thiocyanopentan-4-0ne prepared as described above in a round-bottomed flask and add an equivalent amount (1 mol) of aniline in ethanol and 2 4 drops of concentrated hydrochloric acid. Stir the mixture thoroughly and reflux it for about 3 h.After cooling the contents to room temperature remove half of the solvent under reduced pressure ; wash with water the crystalline product that precipitates immediately and recrystallise it three times from boiling ethanol to obtain colourless PTPT, m.p. 239 "C. The yield is about 140 g (90%). l-PhenyZ-4,4,6-trimeLhyZ-( lH,4H)-pyrimidine-2-LhioZ (PTPT). Prepare a 0.025'4 solution of this compound in dimethylformamide. General Procedure Measure the pH and adjust it if necessary to lie within the range 7.4-11.7 by adding 1 M sodium hydroxide solution or 1 M acetic acid then add 2 ml of buffer solution. Transfer the solution into a 100-ml round-bottomed flask and heat to 60 "C in a water-bath. Add 2 g of naphthalene, stopper the flask and continue to heat until the naphthalene melts.Remove the flask from the water-bath and shake it vigorously until the naphthalene separates as a solid mass. Repeat the melting and solidification procedure. Separate the naphthalene from the aqueous phase by filtration through a filter-paper. Dissolve the solid mass in chloroform and dilute to 10 ml with chloroform in a calibrated flask. Pour the solution on to anhydrous sodium sulphate in a beaker in order to remove the last traces of water. Place a portion of this solution in a 1-cm cell and measure the absorbance at 400 nm against a reagent blank. Prepare a calibration graph under similar conditions. To an aliquot of copper solution in a beaker add 1 ml of PTPT solution. Results and Discussion Absorption Spectra The absorption spectra of PTPT and copper - PTPT complex in naphthalene - chloroform solution were recorded against water and a reagent blank respectively (Fig.1). The copper -PTPT complex shows maximum absorption in the range 400405 nm where the absorbance of 0.6 Q) m 2 0.4 e v) 0.2 2 0.0 300 340 380 420 460 Wavelengthhm Fig. 1. Absorption spectra of PTPT and Cu -PTPT complex in naphthalene - chloroform solution containing 30 pg of Cu 1 ml of 0.02% PTPT and 2g of naphthalene; pH 9.4; shaking time 1 min; and as reference water for PTPT (A) and reagent blank for CU - PTPT (B) . 0.4 0) C ff g 0.2 2 0.0 r . A " " I I I 6 8 10 12 PH Fig. 2. Effect of pH on extraction of Cu -PTPT complex in naphthalene - chloroform solu-tion containing 30 pg of Cu ; 1 ml of 0.02% PTPT and 2 g of naphthalene; shaking time 1 min; reference reagent blank; and wavelength, 400 nm Afiril 1983 517 the reagent is negligible.All absorbance measurements were made at 400 nm in subsequent studies. LIQUID EXTRACTION SPECTROPHOTOMETRY OF Cu AS ITS PTPT Effect of pH Extraction of the complex was carried out at various pH values keeping other conditions constant. The nature of the spectral curves remained constant indicating the formation of only one complex under these conditions. The extraction was quantitative in the pH range 7.4-11.7 (Fig. 2). Effect of Concentration of Reagent volumes of reagent solutions (0.1-4.0 ml). 0.3 and 4.0 ml of the reagent solution were added (Table I).Extractions of the complex were carried out at fixed pH but with the addition of varying The extractions were quantitative when between TABLE I EFFECT OF VARIOUS PARAMETERS ON THE DETERMINATION OF COPPER Each parameter was varied in turn keeping the remaining parameters constant. The constant parameters were as follows volume of PTPT solution 1 ml; amount of naphthalene 2 g ; volume of aqueous phase 25 ml; volume of buffer solution 2 ml; shaking time 1 min; and amount of copper 30 pg. Amex. = 400 nm. Volume of Amount of 0.02 O /. PTPT naphthalene/ 0.1 0.220 0.25 0.2 0.355 0.50 0.3 0.368 0.75 0.5 0.370 1.00 1 .o 0.370 1.25 1.5 0.370 1.50 2.0 0.371 2.00 2.5 0.370 2.50 3.0 0.368 3.00 3.5 0.370 3.50 4.0 0.370 -solution/ml Absorbance g Standing timelmin 30 60 90 120 160 180 210 240 270 300 330 Absorbance 0.370 0.370 0.370 0.370 0.370 0.370 0.370 0.370 0.355 0.336 0.310 Standing time after addition of reagent/ min 5 10 15 20 26 30 -Absorbance 0.090 0.220 0.315 0.350 0.370 0.370 0.368 0.370 0.370 0.370 -Volume of aqueous phaselml 20 30 40 50 60 70 80 90 100 c Shaking time/ Absorbance min 0.370 1 0.370 2 0.370 5 0.369 10 0.370 15 0.370 20 I 25 30 -- c Absorbance 0.370 0.370 0.369 0.370 0.370 0.360 0.345 0.325 0.305 -Absorbance 0.370 0.370 0.370 0.369 0.371 0.370 0.372 0.370 -Volume of buffer solution ml 0.6 1 .o 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 (PH 9.4) / Amount of copper/ Pg 6 10 16 20 25 30 35 40 45 60 60 70 80 90 Absorbance 0.370 0.370 0.372 0.370 0.368 0.370 0.370 0.371 0.369 0.369 0.370 Absorbance 0.070 0.110 0.180 0.230 0.310 0.370 0.435 0.600 0.668 0.630 0.760 0.850 0.900 0.940 Effect of Volume of Aqueous Phase As the volume of the organic phase is small compared with that of the aqueous phase it was essential to study the effect of the volume of the aqueous phase on the extraction.When the latter was varied between 20 and 100 ml the absorption remained constant up to a volume of 60 ml. Above this volume the extraction was not quantitative (Table I) 518 WASEY ei! al. SOLID - LIQUID SEPARATION AFTER LIQUID - AfiaZyst VOZ.108 Effect of Naphthalene The amount of naphthalene added was varied from 0.25 to 4.0 g. The absorbance remained constant when the amount of naphthalene lay between 1.25 and 3.5 g. Below 1.25 g the extraction was incomplete and above 3.5 g it was difficult to dissolve naphthalene in the limited amount of solvent (10 ml) (Table I). Effect of Buffer Solution It was found that the addition of 0.5-10ml of the buffer solution caused virtually no variation in absorbance (Table I). Hence in all of the experiments 2 ml of buffer solution were used. Effect of Standing Time The absorbance of the extract was measured after various intervals of time. The colour of the complex in naphthalene - chloroform solution remained unchanged for up to 4 h (Table I).Effect of Standing Time After Addition of Reagent The copper - PTPT complex in a solution containing 30 pg of copper was allowed to stand at room temperature for between 5 and 30min for complex formation and then extracted into molten naphthalene. The absorbance remained constant in this range (Table I). Effect of Shaking Time The extraction of the complex into molten naphthalene was found to be very rapid and no change was observed in the extent of extraction when the shaking time was varied from 1 to 30 min (Table I). Beer's Law and Sensitivity The absorbances with various concentrations of copper were measured at 400 nm against a reagent blank under the optimum conditions described above. The absorbance varied linearly with the concentration of copper in the range 5.0-60.7 pg in 10 ml of chloroform solu-tion (Table I).The molar absorptivity and sensitivity were calculated to be 1.23 x 104 1 mol-l cm-1 and 0.008 11 pg cm-2 respectively (for an absorbance of 0.001). Ten replicate analyses of a sample solution containing 30 pg of copper gave a mean absorbance of 0.372 with a standard deviation of 0.0053. Effect of Diverse Ions Sample solutions containing 30 pg of copper and various amounts of different alkali metal salts or metal ions were prepared and the determination of copper was studied. The pH of the solution was adjusted to 7.4-11.7 and the general procedure was applied. The results are given in Tables I1 and 111. TABLE I1 EFFECT OF ALKALI METAL SALTS ON THE DETERMINATION OF COPPER Amount of copper 30 pg; amount of naphthalene 2 g; pH 9.4.Amount of anion Alkali metal salt added/pg L -Sodium orthophosphate . . 1000 2 000 Sodium oxalate . . . . 1000 2 000 Sodium citrate . . . . 1000 2 000 Sodium tartrate . . . . 1000 2 000 Sodium fluoride . . . . 1000 2 000 Sodium thiocyanate . . 1000 2 000 Absorbance (at 400 nm) 0.370 0.370 0.368 0.372 0.371 0.370 0.371 0.370 0.372 0.370 0.369 0.370 0.369 Amount of anion Alkali metal salt added/pg Sodium chloride . . . . 1000 2 000 Potassium bromide . . 1000 2 000 Potassium iodide . . 600 1000 Sodium acetate . . . . 1000 5 000 EDTA (disodium salt) . . 100 500 Absorbance (at 400 nm) 0.370 0.370 0.370 0.371 0.370 0.365 0.370 0.370 0.020 0.00 April 1983 LIQUID EXTRACTION SPECTROPHOTOMETRY OF CU AS ITS PTPT TABLE I11 EFFECT OF DIVERSE METAL IONS ON THE DETERMINATION OF COPPER Amount of copper 30 pg; amount of naphthalene 2 g; pH 9.4.Metal ion Added as Nia+ . . . . Chloride Co'+ . . . . Chloride Als+ . . . . Nitrate Mn*+ . . . . Sulphate U02'+. . . . Acetate ose+ . . . . Oxide 19+ . . Chloride Cra+ . . . . Nitrate Ptd+ . . . . Chloride Zna+ . . . . Sulphate - -Amount of ion added/ pg 100 600 100 600 100 600 100 600 100 200 100 300 100 600 100 200 100 600 100 200 -Absorbance (at 400 nm) 0.370 0.370 0.369 0.370 0.368 0.370 0.369 0.370 0.366 0.369 0.371 0.370 0.368 0.370 0.372 0.370 0.369 0.371 0.372 0.370 0.368 Metal ion Fea+ .. Cda+ . . * . Hga+ . . Ag+ . . Pba+ . . Bi3+ . . Sna+ . . Pd2+ . . Aua+ . . Rua+ . . Rh3+ . . ,. Added as Chloride Chloride Nitrate Nitrate Nitrate Nitrate Chloride Nitrate Chloride Chloride Chloride Amount of ion addedlpg 100 800 60 100 60 100 60 100 60 100 50 100 60 100 60 100 60 100 60 100 60 100 519 Absorbance (at 400 nm) 0.369 0.369 0.370 0.366 0.370 0.368 0.370 0.367 0.370 0.368 0.370 0.360 0.370 0.360 0.371 0.376 0.372 0.377 0.370 0.368 0.370 0.366 Determination of Copper in Goat Liver and Human Hair Samples were ashed separately at low temperature:^^ the residues were dissolved in 15 ml of concentrated nitric acid and evaporated to dryness and the residues were dissolved in 20 ml of doubly distilled water.An aliquot of each solution was then treated according to the pro-cedure described. Replicate samples were analysed by the present method and by atomic-absorption spectrophotometry. The results obtained by the two methods are compared in Table IV. TABLE IV DETERMINATION OF COPPER IN GOAT LIVER AND HUMAN HAIR Copper determined/ pg I Atomic-absorption spectrophotornetry' Sample Spectrophotometry (this work) (direct method) Goat liver-A 7.2 B 12.0 c 13.6 D 16.0 E 18.0 Humala hair-A 9.3 B 13.0 c 14.6 D 18.6 E 16.6 7.3 11.8 13.4 16.8 17.8 9.3 12.8 14.6 18.6 16.6 Determination of Copper in Alloys A 100-mg amount of alloy was dissolved in 15-20 ml of aqua regia the solution was evapor-ated to about 5 ml and 10 ml of concentrated hydrochloric acid were added.The solution was warmed to dissolve salts and transferred into a 250-ml calibrated flask and diluted to volume with water. An aliquot of this solution was placed in a beaker 1 ml of PTPT solution wa 520 WASEY KUMAR BANSAL KRISHAN PURI AND SATAKE added the pH was adjusted to 7.4-11.7 and the extraction was carried out according to the general procedure. Five determinations were made for each alloy to determine the average copper content and the results are summarised in Table V. TABLE V DETERMINATION OF COPPER IN ALLOYS Amount of copper Amount of copper Alloy Composition % takenlpg found* /pg Brass .. . . Cu 58.0 Pb 2.56 Zn 38.99 30.5 30.36 Fe 0.09 Sn 0.12 Gun metal Cu 85.0 Sn 10.0 Zn 2.5, Pb 0.25 Fe 2.0 Steel 33d (NBS) . . . . Cu 1.54 Fe 90.3, Ni 2.38 Cr 0.32 C 2.3, Si 1.63 Mn 0.68 Aluminium alloy . . . . Cu 4.0 A1 92.15, Ni 2.12 Mn 1.7 Fe 0.031 32.5 32.36 12.3 12.20 16.0 15.80 * Average of five determinations. Conclusion Liquid - liquid extraction is a popular separation technique because of its elegant simplicity, but it fails when a metal ion forms a complex with the complexing agent at high temperature or when the solubility of the metal complex is low at room temperature. The technique proposed here has not only solved this problem but has also enhanced the sensitivity of the method be-cause only 2 g of naphthalene are required for the complete extraction of the metal ion from 50-60 ml of the aqueous phase. The solid extract can then be dissolved in a small volume of organic solvent and the whole of it can be used for analysis. Metal ions that form thermally unstable complexes may interfere in liquid - liquid extraction spectrophotometric methods, but have no effect in the present technique resulting in better selectivity. One of the authors (A.W.) is grateful to the Council of Scientific and Industrial Research, New Delhi India for the award of a Fellowship. References 1. 2. 3. 4. 6. 6. 7. 8. 9. Singh A. K. Katyal M. Bhatia A. M. and Ralhan N. K. Talanta 1976 23 337. Singh A. K. Singh R. P. and Katyal M. J . Indian Chem. Soc. 1976 53 650. Fujinaga T. and Puri B. K. Talanta 1975 22 71. Puri B. K. and Gautham M. Mikrochim. Acta 1979 I 515. Vogel A. I. “A Text Book of Quantitative Inorganic Analysis,” Third Edition Longmans London, Mathes R. A. Stewart F. D. and Swedish F. Jr. J . Am. Chem. Soc. 1948 70 1462. Mathes R. A. J . Am. Chem. SOC. 1953 75 1747. Satake M. and Matsumura Y . Bunseki Kagaku 1977 26 386. Satake M. Matsumura Y. and Fujinaga T. Talanta 1978 25 718. 1969. Received July 22nd 1982 Accepted November loth 198

ISSN:0003-2654    

DOI:10.1039/AN9830800515

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst April 1983 SHORT PAPERS 521 Gas Chromatographic = Mass Spectrometric Identification of 9,lO-Epoxystearate in Human Blood Gunnar A. Ulsaker and Gerd Teien National Centre for Medicinal Products Control Sven Oftedals vei 6 Oslo 9 Norway Keywords Gas chromatography - mass spectrometry ; 9,lO-epoxystearate ; endo-genous component ; blood This investigation was undertaken in order to detect 9,lO-epoxystearate in human blood stored in a commercial poly(viny1 chloride) (PVC) bag It may be assumed that the epoxi-dised fatty acid esters being present in fairly substantial amounts in this plastic material are able to migrate from the PVC wall into the blood during storage. The 9,lO-epoxystearate represents the simplest of the epoxides present in a major amount in epoxidised soya and linseed oils both considered acceptable additives according to the European Pharmac0poeia.l So far, reports on such migration of epoxidised oils are lacking whereas migration of another lipo-philic additive viz.bis(Zethylhexy1) phthalate has been well documented (e.g. references 2 and 3). Nevertheless the detection of 9,lO-epoxystearate in the blood sample was eventually com-pleted and its presence in the PVC material confirmed; the same substance was also detected in the blood sample prior to any contact with PVC. Consequently the 9,lO-epoxystearate can also be regarded as an endogenous constituent of the blood sample. Epoxy intermediates are sometimes formed during the metabolic conversion of unsaturated compounds into more hydrophilic products.The epoxides may be enzymatically h ydrolysed through the action of epoxide hydrolase.4 Recently two epoxides of arachidonic acid were isolated by using epoxide hydrolase inhibitor in in vitro experiments with hepatic monooxygen-a ~ e . ~ In phospholipids taken from the lung tissue of rats epoxides of fatty acids have been isolated.6 The occurrence of epoxides in lung tissue could indicate that lipid peroxidation contributes to the formation of epoxides. It may be added that oleic acid is a major fatty acid in blood.' This paper describes the detection of 9,lO-epoxystearate in human blood. The identifica-tion is based upon the reduction of the ester and epoxy groups with lithium aluminium hydride and subsequent gas chromatographic - mass spectrometric analysis.* In the reduc-tion the epoxy carbonyl and carboxyl groups would be converted into hydroxy groups which cannot be distinguished from naturally occurring hydroxy groups.In order to prove the existence of the intact epoxide function in blood the sample was reduced with the correspond-ing deuteride. Further proof of the existence of the epoxy group was obtained through chromium(V1) oxide oxidation of the deuterated compound and subsequent reduction by lithium aluminium hydride prior to analysis. Experimental Apparatus The scanning of mass spectra was carried out by using an LKB 2091 gas chromatograph -mass spectrometer. The electron-impact ion source was operated at 70 eV when scanning the mass spectra. Two glass columns (1.5 m x 2 mm i.d.) packed with 3% QF-1 or 3% SE-30 both on Supelcoport (80-100 mesh) were used.The injector the separator and the ion source were all maintained at 230 "C. The temperature of the column oven was maintained at 100 "C for 1 min after injection and then increased at a rate of 4 "C min-l to 225 "C. The retention time for the reduced 9,lO-epoxystearate was 22 min on the QF-1 column and 30 min on the SE-30 column. The helium flow-rate was 20 ml min-l 522 SHORT PAPERS Reagents Analyst Vol. 108 All reagents used were of the highest available purity. Methyl 9,lO-epoxystearate. Sampling procedure. This material was synthesised according to the method given in reference 9. Blood drawn from two healthy donors was collected in CPD-Adenine (anticoagulant citrate phosphate dextrose) solution care being taken to exclude any contact with PVC or rubber materials.After collecting the blood it was immediately cooled in ice before centrifugation and extraction. Procedures Extraction of filasma A mixture of 10 ml of plasma and 20 ml of water was extracted with 75 ml of chloroform -methanol (7 + 3) then extracted with 2 x 30 ml of chloroform. The combined extracts were dried over anhydrous magnesium sulphate. After evaporation the reductions and oxidation were performed according to the methods given in references 8 and 10. Results and Discussion The mass spectrum of a small peak in the plasma eluted (QF-1 column) 4 min after the trimethylsilyl (TMS) derivative of reduced saturated stearate turned out to be the spectrum of the TMS derivative of reduced 9,lO-epoxystearate.As chemical reduction of the oxirane ring involves an about equal probability of cleavage of the two C-0 bonds the 9- and 10-ocatadecane-l-diols should be formed in about equal LiAIH4 ,/ BSA OTMSi CH~-(CH~)~CHD-CH-(CH~),-CDT-OTMS~ I I i 230 + --t-J + 305 I - - - - - -Cr03 LiAIH4 BSA I OTMSi CH3-(CH2)rCHD-CH-(CH2)7-CH~4TMSi ‘ I I 230 (--I-J I - * 303 I---\ Fig. 1. Expected reaction products from glyceryl 9,lO-epoxystearate subjected to various chemical treatments and major ct-cleavage products in the mass spectrometer respectively. R = Glyceryl; TMSi = trimethylsilyl; BSA = N,O-bis(trimethylsily1)acetamide ; m/z values are also shown April 1983 SHORTPAPERS 523 amounts. The ct-cleavage of the trimethylsilyl ethers of these two octadecanediols would give ions with intense fragments at m/z 215 229 303 and 317 (see Fig.1). In the samples these intense ions together with a weak molecular ion (0.7% of base peak) at m/z 430 indicated the presence of 9,lO-epoxystearate. The mass spectrum and the retention time were identical with those of the reduced and silylated synthesised sample. However the exact structure of this compound could not be determined from the spectra because hydrogenolysis converts the epoxide into secondary alcohols and the carboxyl end to a primary alcohol. The mass spectrum could as well arise from a mixture of 9- and 10-diols ketones or a mixture thereof, with a terminal hydroxy or aldehyde group. If 9,lO-epoxystearate was the source of the analysed fraction the a-cleavage ions expected would be m/z 215 230, 305 and 320 (see Fig.1). The mass spectrum of the deuterated compound showed the same pattern and the fragments were found at m/z 215 230 305 and 320 as expected (see Fig. 2). The molecular ion had moved to m/z 433 (plus 3 atomic mass units). The sample was therefore reduced with lithium aluminium deuteride. 100 $? .- 2 100 C 4- .-Q) .- + - 0 a 100 Fig. 2. Partial mass spectra of octa-decanediols in plasma derived by various chemical treatments. (a) LiAlH, BSA; (b) LiAlD, BSA; and (c) LiAlD, CrO, LiAlH,, BSA where BSA = N,O-bis(trimethylsily1)-acetamide. The existence of the intact epoxy and terminal carboxyl groups was further substantiated through chromium(V1) oxide oxidation and subsequent reduction with lithium aluminium hydride.lO Primary alcohols are oxidised to carboxylic acids and secondary alcohols are oxidised to ketones by chromium(V1) oxide.Therefore in-chain deuterium originated by reduction of an epoxide should not be affected by chromium(V1) oxide treatment if any deuterium - hydrogen exchange can be prevented. The oxidation and subsequent reduction should therefore remove two deuterium atoms from the terminal carbon atom leaving only the deuterium atom in the chain. The mass spectra had a weak molecular ion at m/z 431 and a-cleavage peaks at m/z 215 230 303 and 318 as predicted. However in addition to these, major peaks at m/z 229 and 317 were found. These results indicated a mixture of intact epoxide and 9- and 10-alcohols or ketones before the chemical treatments.A mixture of alcohols reduced with lithium aluminium deuteride would give major or-cleavage peaks at m/z 215 229 305 and 319. Therefore, only traces of the alcohols may be present in the sample. Correspondingly the small relative increase in the peaks at m/x 216 and 306 tolerates only traces of ketones in the sample. The partial loss of deuterium has therefore occurred during the chemical treatments. Before oxidation peaks at m/z 229 and 319 were small 524 SHORT PAPERS Analyst Vol. 108 Conclusions In view of the methods used there should be no doubt about the identity of the 9,lO-epoxystearate detected in the blood plasma. As no trace of this substance could be found when reagent blanks were analysed using the same technique and care was taken to prevent oxida-tion of the blood after collection the endogenous nature of the epoxide is indicated.It still remains an open question whether blood from a PVC bag in clinical conditions of use, receives additional amounts of epoxide as a result of migration from the PVC wall. An assessment of such contamination if any merits further investigation with the aid of suitable instrumentation and sampling techniques. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References “The European Pharmacopoeia,” Second Edition Part I Maisonneuve S.A. Sainte-Ruffine France. Jensen L. E. and March J. Arch. Pharm. Chemi Sci. Ed. 1977 5 43. Ulsaker G. A. and Hoem R. M. Analyst 1978 103 1080. Lu A. Y. H. and Miwa G. T. Annu. Rev. Pharmacol. Toxicol. 1980 20 513. Oliw E. H. Guengerich P. and Oates J. A. J . Biol. Chem. 1982 257 3771. Sevanian A. Mead J. F. and Stein R. A. Lipids 1979 14 634. Haan G. J. van der Heide S. and Wolthers B. G. J . Chromatogr. 1979 162 261. Walton T. J. and Kolattukudy P. E. Biochemistry 1972 11 1885. Findley T. W. Swern D. and Scanlan J. T. J . Am. Chem. Soc. 1945 67 412. Kolattukudy P. E. Lipids 1973 8 90. 1981 VI.1.2.1.1. Received May 25th 1982 Accepted October llth 198

ISSN:0003-2654    

DOI:10.1039/AN9830800521

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

524 SHORT PAPERS Analyst, Vol. 108 Spectrop hotometric Determi nation of Smal I Amounts of Niobium in Steels Using Sulphochlorophenol S Zdenek Cizek and Vlasta Studlarova Central Research Institute, Skoda Ca., 31 6 00 Plze&, Czechoslovakia Keywords : Niobium determination ; steel analysis ; matrix extraction ; sulpho- chlorophenol S ; spectrophotometry Sulphochlorophenol S [2,7-di(2-hydroxy-3-sulpho-5-chlorophenylazo)chromotrop~c acid (SCPS)] has often been used for the spectrophotometric determination of niobium in inorganic materials. The sensitivity, appreciable selectivity and stability of the niobium - SCPS com- plex in acid medium are advantageous in the determination of niobium in steels and alloys. The direct determination of niobium in steels at levels above o.05y0 has been described by Savvin et dl; indirect determinations after prior separation with phytic acid or after extrac- tion of the niobium complex have also been Problems in the direct spectrophotometric determination of niobium in various types of steels with the use of sulphochlorophenol S were considered in detail in previous paper^,^^^ where optimum reaction conditions and the effects of sample matrices and alloying elements were reported.We explained interfering effects on the formation of the niobium complex caused by the presence of sulphate, phosphate, fluoride, tartrate] oxidising agents (chlorine, nitrogen oxides), reducing agents (ascorbic acid) and other metal ions. A method for the determination of niobium over a wide range of concentrations (o.=1-3y0) was proposed for a new COMECON standard.Practical experience with the direct spectrophotometric determination has shown a need for a more effective means of eliminating the effects of interfering elements (particularly iron) when the sample contains less than 0.1% of niobium. This problem was solved by the separa-April, 1983 SHORT PAPERS 525 tion of niobium from these elements by extracting them into an organic phase and spectro- photometric determination of niobium in the aqueous phase. The solvents chosen, on the basis of literature were diisopropyl ether (DIPE) and isobutyl methyl ketone (IBMK). This paper summarises the experimental data and gives a recommended method for the spectro- photometric determination of small amounts of niobium (O.OO1-O.l~o) in a wide range of plain carbon, low-alloy and high-alloy steels.Experimental Apparatus A Beckman ACTA M 7 spectrophotometer was used for spectrophotometric measurements. Reagents All reagents were of analytical-reagent grade. Sulphuric acid, sp. gr. 1.84. Nitric acid, sp. gr. 1.42. HydroJluoric acid, sp. gr. 1.12. Phosphoric acid, sp. gr. 1.75. Tartaric acid, 1 M solution. Sdphochlorophenol S solution, 0.05y0 mlV. Diisopropyl ether. Isobutyl methyl ketone. Methanol. Niobium stock standard solution, 200 pg ml-l. Dilute 1 + 3. Fluka, Buchs, Switzerland. Dissolve 0.100 g of niobium metal (>99.9%) in ca. 5 ml of hydrofluoric acid and ca 5 ml of nitric acid in a platinum dish. Add 10 ml of sulphuric acid and evaporate to fumes of sulphur trioxide for 10 min.Dissolve the residue in ca. 20 ml of 1 M tartaric acid and dilute to 500 ml with 1 M tartaric acid in a calibrated flask. This solution is stable for at least 6 months. Procedure Dissolve 0.500 g of sample in 20 ml of sulphuric acid (1 + 3) and 0.5 ml of phosphoric acid and oxidise with a few drops of nitric acid. Evaporate to fumes of sulphur trioxide for at least 10 min, rinse the walls of the beaker with a small volume of water and repeat the evaporation. This is necessary especially with steels with higher contents of carbon in order to effect complete decomposition of niobium carbide. Depending on the type of test material, alternative sample decomposition methods can be used (e.g., with hydrochloric acid and nitric acid), with a final evaporation to fumes of sulphur trioxide. After cooling the residue, add 5 ml of 1 M tartaric acid and a small volume of water.Heat until the residue has dissolved, transfer the solution into a 100-ml calibrated flask, dilute to the mark with water and mix well. Before pipetting an aliquot, filter this stock solution, if necessary. Pipette 10 ml of the stock solution into a separating funnel, add 15 ml of hydrochloric acid, 10 ml of DIPE and extract by shaking for 30s. After separating the layers, transfer the aqueous phase into a 50-ml beaker and heat on a water-bath (90-95 "C) for about 20 min. Cool and transfer the solution into a 50-ml calibrated flask, add 5 ml of methanol, dilute with water to about 45 ml and add exactly 2 ml of SCPS solution. Dilute the solution to the mark with water and allow to stand at room temperature for ca.50 min. This time can be reduced by accelerating the reaction by heating the test solution at 60 "C for 10 min. Measure the absorbance in suitable cells at 650 nm against a blank solution prepared in the same way by using a steel sample or sample of high-purity iron that is free from niobium. Determine the niobium content from a calibration graph, If the sample conta.ins more than 2% or 5% of chromium or nickel, respectively, pipette another 10-ml aliquot at the same time so that the background absorbance of chromium(II1) and nickel(I1) can be determined. Treat this secmd aliquot in the same manner as the test solution without the addition of SCPS. Measur: the background absorbance against water and subtract it from the absorbance of the test Folution.Pipette 10 ml of the stock solution into a separating funnel, add 10 ml of hydrochloric acid and 5 ml of water. Mix well, add 10 ml of IBMK and extract by Extraction with DIPE. Extraction with IBMK.526 SHORT PAPERS Analyst, Vol. 108 shaking for 30 s. Proceed as with DIPE but add 5 ml of hydrochloric acid to a 50-ml cali- brated flask before adding methanol and SCPS. If the sample contains chromium(II1) and nickel(II), measure and subtract their background as described above. The extraction with IBMK is suitable for steel samples that contain molybdenum. Calibration graph. Weigh 0.500-g samples of steel or high-purity iron that is free from niobium into six beakers and dissolve them in 20 ml of sulphuric acid (1 + 3) and 0.5 ml of phosphoric acid, oxidising with a few drops of nitric acid.Add 0, 0.5, 1, 1.5, 2, 2.5 ml of the standard niobium solution and evaporate to fumes of sulphur trioxide for at least 10 min. Proceed further using the procedure described above. Results and Discussion Extraction of Iron and Other Elements into an Organic Phase Chloro complexes of iron and some other elements are extracted by means of IBMK or DIPE with high efficiency in hydrochloric acid medium. Under certain conditions, niobium remains in the aqueous phase. The distribution of iron(III), molybdenum(V1) and niobium(V) between the organic and aqueous phases is dependent on the hydrochloric acid concentration and is shown in Fig. 1. This indicates that 5 M hydrochloric acid medium using IBMK and 7-8 M hydrochloric acid medium using DIPE are the most suitable for the separation of iron by 100 s c .- ti 50 w ; I L 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 Concentration of HCVM Fig.1. Extraction of iron(III), molybdenum(V1) and niobium(V) into (a) IBMK and Volume of aqueous phase 60 ml, volume (b) DIPE against hydrochloric acid concentration. of organic phase 6 ml and extraction time 1 min. extraction. Molybdenum is partially extracted from 5 M hydrochloric acid medium into IBMK whereas almost nothing is extracted into DIPE. None of the other alloying elements in steel such as chromium, nickel and vanadium are extracted using either solvent. Vanadate ions are reduced by extraction with these solvents to vanadyl ions, which do not react with SCPS. The extent of extraction, found by analysing various steel samples under the condi- tions described, are 98-99% and 60-70% for iron and molybdenum, respectively.After phase separation, residues of both solvents in the aqueous phases must be removed by heating them on a water-bath for about 20 min so that their negative influences on the niobium determination are eliminated. Reaction of Niobium with SCPS and Effects of Other Elements The optimum medium for the formation of the niobium - SCPS complex is 3 M hydro- chloric acid. In more acidic media (>3.5 M), the value due to the blank solution increases rapidly as SCPS begins to precipitate in solution. As the selectivity of the reaction of niobium with SCPS increases with increasing hydrochloric acid concentration, the question of the dis- solution capacity of SCPS was resolved by adding small amounts of methanol.Addition of 5 ml of methanol to the reaction medium prevents precipitation of SCPS and the value due to the blank solution remains almost constant. Hydrochloric acid medium of 4-5 M will virtuallyApril, 1983 SHORT PAPERS 527 completely eliminate the negative influence of molybdenum (1 mg of molybdenum m 2 pg of niobium), The small amount of phosphoric acid used as a masking agent during the sample decomposition is sufficient to bind zirconium in a stable phosphate complex that does not interfere in the reaction of niobium. Chromium(III), nickel(II), copper(I1) and vanadyl ions are not extracted either into IBMK or DIPE. These elements do not react with SCPS and do not interfere in the formation of the niobium - SCPS complex.However, their background absorbances must be eliminated by measuring against water and subtracting them (2% of chromium m O . O O l ~ o of niobium). Zirconium forms a complex with SCPS under almost the same conditions as niobium. Calibration Graph A calibration graph is constructed by using steel samples with standard niobium addition and carrying them through the complete procedure. It has been found that the two calibra- tion graphs obtained by using DIPE and IBMK for extraction are almost identical. The Beer - Lambert law is obeyed, under the conditions described, over a niobium range of 1-50 pg (0.02-1 p.p.m.). Determination of Niobium in Standard Steel Samples ian, Polish and US standard steel samples for niobium.are in close agreement with the certificate values. The accuracy of the method was established by analysing a series of British, Czechoslovak- The results, summarised in Table I, TABLE I RESULTS OF ANALYSIS OF STANDARD STEEL SAMPLES Niobium content, yo Quoted on test Found by Found by Standard sample* certificate IBMK extraction DIPE extraction CKD 166 .. .. 0.025 0.025 0.024 CKD 167 .. .. 0.09 0.085 0.085 CKD 168 .. .. 0.01 0.007 0.007 CKD 171 .. .. 0.035 0.033 0.034 IMZ 4.6.1 .. .. 0.017 0.016 0.016 IMZ 4.6.2 . . .. 0.045 0.043 0.043 IMZ 4.6.3 . . .. 0.059 0.056 0.058 IMZ 4.6.4 . . .. 0.101 0.100 0.102 BCS 219/2 .. .. 0.004 BCS 273 . . .. .. 0.0003 BCS 275 . . .. .. 0.035 BCS 277 . . .. .. 0.021 BCS 432 . . .. .. 0.029 0.002 0.001 0.038 0.022 0.028 0.002 0.001 0.037 0.023 0.027 NBS 19g. . .. .. 0.026 0.026 0.026 NBS lOle .. .. 0.013 0.013 0.015 British; and NBS = US. * Standard steel samples: CKD = Czechoslovakian; IMZ = Polish; BCS = The precision of the method was tested by repeated determinations of niobium in low- and The relative standard deviations were 2.5% and 1.0% for o.01470 high-alloy steel samples. and O . l O O ~ o of niobium, respectively. 1. 2. 3. 4. 6. 6. References Savvin, S. B., Pisarenko, I. D., Yurchenko, E. I., and Dedkov, Yu. M., Zh. Anal. Khim., 1966, 21, Wakamatsu, S., Bunseki Kagaku, 1969, 18, 376. Savvin, S. B., Romanov, P. N., and Eremin, Yu. B., Zh. Anal. Khim., 1966, 21, 1423. Ivanov, N., Borissova, R., and Veselinova, E., Fresenius 2. Anal. Chem., 1976, 280, 223. Ciiek, Z., H u h . Listy, 1979, 34, 357. CiBek, Z., and Doleial, J., Anal. Chim. Ada, 1979, 109, 381. 669.528 SHORT PAPERS Analyst, VoE. 108 7. 8. Marcus, Y., and Kertes, A. S., “Ion Exchange and Solvent Extraction of Metal Complexes,” Wiley- Koch, 0. G., and Koch-Dedic, G. A., “Handbuch der Spurenanalyse,” Springer Verlag, Berlin, 1974, Received July 21st, 1982 Accepted November 8th, 1982 Interscience, London, 1969, p. 950. pp. 264 and 265.

ISSN:0003-2654    

DOI:10.1039/AN9830800524

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

528 SHORT PAPERS Analyst, VoE. 108 Spectrophotometric Determination of Cobalt after Coprecipitation of its Morpholine-4-carbodithioate with Microcrystalline Naphthalene Chaman La1 Sethi, Ashok Kumar and Bal Krishan Puri" and Masatada Satake Department of Chemistry, Indian Institute of Technology, New Delhi 110016, India Faculty of Engineering, Fukui University, Fukui 910, Japan Keywords : Spectrophotometry ; cobalt determination ; morpholine-4-carbo- dithioate ; coprecipitation Beyer and co-workers have used morpholine-4-carbodithioate for the spectrophotometric and titrimetric determination of various metal ions.l-* This paper reports a rapid spectrophoto- metric method for the trace determination of cobalt using this reagent, which is sensitive and more convenient than a similar method involving extraction into molten na~hthalene.~ Experimental Equipment atomic-absorption spectrophometer were used.An Elico pH meter, a Pye Unicam SP-700/SP-500 spectrophotometer and a Pye Unicam 191 Reagents in the usual way.6 trigiano et al.' Standard cobalt solution, 10-2 M. Potassium morpholine-4-carbodithioate solution, 0.2%. Sodium acetate - acetic acid bufler, 0.2 M, p H 5.16. Ammonia solution, 4 M. Perchloric acid, 4 M. Potassium nitrate solution, 1 M. Naphthalene. Naphthalene solution in acetone, 20%. Chloroform. Check the purity spectrophotornetrically before use. Dissolve cobalt nitrate in distilled water and standardise Prepare by the method of Macro- Mix 100 ml of each solution (0.2 M). Check the purity spectrophotometrically before use. Procedure To the standard solution containing about 30 pg of cobalt in an Erlenmeyer flask add 2.0 ml of reagent solution, 2.0 rnl of buffer solution and a calculated amount of 1 M potassium nitrate solution to give a 0.1 M over-all ionic concentration in 30 ml of the aqueous phase.Swirl to mix the solution and allow it to stand for 2-3 min for complex formation to be completed. Add 2.0 ml of the 20% solution of naphthalene in acetone rapidly and shake the mixture vigorously for 1-2 min. Filter the solid mass formed through a filter-paper, wash with dis- tilled water, drain dry and dissolve in sufficient chloroform to give a volume of 10.0 ml. Transfer this solution into a dry flask and add 2 g of anhydrous sodium sulphate. Take a portion of this dried solution and measure its absorbance at 365 nm in a l-cm cell against a reagent blank prepared under similar conditions.* To whom correspondence should be addressed.April, 1983 SHORT PAPERS Results and Discussion Absorption Spectra 529 The absorption spectra of the reagent and cobalt morpholine-4-carbodithioate complex in naphthalene - chloroform solution against water and reagent, respectively, are shown in Fig. 1. The cobalt complex absorbs strongly at 365 nm, whereas the reagent absorbs negligibly at this wavelength. Optimum Conditions Coprecipitations of cobalt morpholine-4-carbodithioate were carried out at various pH values. The spectra obtained revealed the formation of only one type of complex in each instance. Determinations were quantitative over the pH range 3.0-8.8 (Fig. 2).For 30 pg of cobalt the coprecipitation of the complex was quantitative when a minimum of 1.0 ml of reagent solution and 1.0 ml of naphthalene solution were used. No effect on absorbance was observed when the concentration of potassium nitrate was varied over the range 0.05-1.0 M. A final concentra- tion of 0.1 M was selected for the recommended procedure. Quantitative recoveries were obtained when the volume of the aqueous phase did not exceed 45 nil. A shaking time of about 15 s was found to be sufficient to achieve complete extraction after the addition of naphthalene. The final solution of the complex in naphthalene - chloroform was stable for more than 24 h. 0.6 0 lu G 2 0.4 Q 0.2 0.0 \ \ \ \ \ \ \ \ \ \ \ \ \ \ B '---- A 320 350 400 450 500 Wavelengthhm PH Fig.2. Effect of pH. Wave- length, 365 nm. Other condi- tions as in Fig. 1. Fig. 1. Absorption spectra of (A) reagent solution against water and (B) cobalt morpholine-4-carbo- dithioate in naphthalene - chloro- form solution against reagent solution. Co, 30.0 pg; reagent (0.2%), 2.0 ml; naphthalene in acetone (20%), 2.0 ml; aqueous phase, 30 ml; and pH, 5.16. Calibration Graph A calibration graph was constructed for the cobalt morpholine-4-carbodithioate complex at 365 nm. Beer's law was obeyed over the concentration range 6-60 pg of cobalt. The molar absorptivity and sensitivity were calculated to be 1.396 x lo4 1 mol-l cm-l and 0.0042 pg cm-2, respectively, at 365 nm. An amount of 30 pg of cobalt in 10 ml of the final solution gave a mean absorbance of 0.710 with a standard deviation of 0.007.Effect of Diverse Ions The following foreign ions (amounts in parentheses) did not interfere in the determination of 30 pg of cobalt in 10.0 ml of the final solution: Cr(III), Pb(II), Cd(II), Hg(II), Bi(III), As(II1) and Sb(II1) (800 pg) ; Mg, Zn, Ru, Rh, W, Mo, U(VI), Au(III), Pt(IV), V(V) and Ag(1) (1000 pg530 SHORT PAPERS Analyst, Vol. 108 each). Interference due to Pd(I1) was eliminated by pre-extraction at low pH, and that due to Fe(II1) by masking using 5 ml of 5% triethanolamine solution. Sodium chloride, potassium bromide, sodium oxalate, sodium potassium tartrate, sodium acetate, sodium sulphate (35 mg each), sodium dihydrogen phosphate (40 mg), sodium thio- sulphate (21 mg), sodium citrate, potassium thiocyanate (30 mg each), sodium fluoride (23 mg) and disodium EDTA (10 mg) were tolerated.TABLE I DETERMINATION OF COBALT IN A SAMPLE OF HIGH-SPEED STEEL OF CERTIFIED COMPOSITION Certified composition: Coy 9.25; Mn, 0.40; Si, 0.35; S, 0.05; P, 0.05; Cr, 4.15; Mo, 5.50; and W, 618.5%. Amount taken Sample for analysis/ 1 10.00 2 15.00 3 20.10 4 30.15 6 35.20 No. Pg Cobalt found, yo Atomic-absorption Present method spectrophotometry f A \ 9.20 9.28 9.25 9.26 9.26 9.25 9.25 9.27 9.25 9.28 Analysis of Alloys An alloy sample (1.0 g) was dissolved in a mixture of 10-15 ml of concentrated hydrochloric acid and 0.5 ml of concentrated nitric acid. This solution was evaporated nearly to dryness, another 10 ml of hydrochloric acid were added and the mixture was diluted and filtered through filter-paper. The filtrate was diluted with water to 100.0 ml.A portion from a further dilution was subjected to the procedure already described. The cobalt content of the diluted solution was also determined by atomic-absorption spectrophoto- metry. The results obtained by both methods are compared in Tables I and 11. TABLE I1 DETERMINATION OF COBALT IN A SYNTHETIC SAMPLE Composition*: Co, 0.50; Mn, 0.80; W, 25.5; S, 4.50; P, 0.80; Mo, 18.50; Na, 7.05; and Fe, 42.35%. Sample No. Cobalt takenlpg Cobalt foundtlpg Error, yo 1 9.00 9.00 0.0 2 12.00 12.03 +0.25 3 15.00 15.05 +0.33 4 30.00 30.05 +0.17 * The stock solution was standardised by using atomic- t Average of five replicate determinations. absorption spectrophotometry. The authors thank The University Grants Commission (India) for financial assistance to one of them (C.L.S.). References 1. 2 . 3. 4. 5. 6. 7. Beyer, W., and Ott, R. D., Mikrochim. Technoanal. Acta, 1965, 1130. Beyer, W., and Likussar, W., Mikrochim. Acta, 1967, 721. Likussar, W., Beyer, W., and Wawshinek, O., Mikrochim. Acta, 1968, 735. Beyer, W., Ott, R. D., and Pokorny, G., Mikrochim. Acta, 1967, 575. Gautam, M., and Puri, B. K., Mikrochim. Acta, 1979, I, 515. Vogel, A. I., “A Text Book of Quantitative Inorganic Analysis,” Third Edition, Longmans, London, Macrotrigiano, G., Pallacani, G. C., Preti, C., and Tosi, G., Bull. Chem. Soc. Jpn., 1975, 48, 1018. 1969. Received July 26th, 1982 Accepted November loth, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800528

出版商:RSC    

年代:1983

数据来源: RSC