摘要:

Analyst, August 1996, Vol. 121 (1 107-1 I1 0 ) 1107 Bifunctional Cryptand Modifier for Capillary Electrophoresis in Separations of Inorganic/Organic Anions and Inorganic Cations Chyow-San Chiou and Jeng-Shong Shih* Department of Chemistry, National Taiwan Normal University, Taipei, Taiwan I1 71 8, Republic of China The macrocyclic polyether cryptand-22 was employed as a bifunctional modifier to separate cations and anions via capillary electrophoresis. The remarkable cation complexing ability of cryptand-22 resulted in an improvement in the separationhesolution of several cations, e.g., K+, Na+ and NH4+ ions, by capillary electrophoresis. Cryptand-22 was protonated as cryptand-2H2+ at pH < 7 and the protonated cryptand-22 was used as a capillary surface modifier to separate various organichorganic anions by capillary electrophoresis.This modifier treatment resulted in the electro-osmotic flow (EOF) and electrophoretic mobility of the anions being in the same direction, thereby giving a greater column efficiency. At pH < 7, cryptand-22 is protonated and its behaviour is similar to that of quaternary ammonium salts which can form complexes with anions; in addition, the EOF is reversed. The effect of the eluent anion on the velocity of the EOF was also investigated; the maximum velocity of the EOF was obtained by using acetic acid as the eluent. The capillary electrophoresis system with the protonated cryptand modifier was successfully applied to the separation and determination of various inorganic anions, e g . , I-, NO3-, SCN-, Br03-, 103- and Cr042-, and various carboxylate anions, e g ., p-nitrobenzoic acid, p-chlorobenzoic acid, p-hydroxybenzoic acid, benzoic acid and terephthaiic acid. Keywords: Cryptand; capillary electrophoresis; modifier; cations; anions Introduction The determination of inorganic/organic anions and cations is of importance to the environmental sciences and chemical in- dustry. Ion chromatography (IC) has conventionally been used to determine inorganic and organic anions by utilizing a conductivity detector14 or, for a limited number of anions, a UV detector.5.6 Capillary electrophoresis (CE) is complemen- tary to IC and has been used to determine inorganic and small organic ions.7-13 CE is attractive owing to its short analysis time, large peak capacity8 and minimal solvent and sample requirements.Optimum analysis time and resolution in CE require control of the magnitude and the direction of the electro- osmotic flow (EOF). Surfactants are among the most widely used buffer additives in CE. Numerous types of surfactants can be used in CE (i.e., anionic, cationic, zwitterionic or non-ionic). At concentrations below the critical micellization concentration (c.m.c.), monomeric ionic surfactant molecules can adhere to the capillary wall and modify the EOF. Depending on the surfactant charge, the EOF can be increased, reduced or reversed. EOF reversal,I4 for example, can be obtained by adding a cationic surfactant such as cetyltrimethylammonium bromide (CTAB)'S to the buffer. The cationic surfactant monomers adhere to the wall through ionic interactions.The positive charge on the wall results from hydrophobic inter- actions of free surfactant monomers with those adhering to the wall. Alternatively, for cationic surfactants both below and above their c.m.c., relative changes in analyte electrophoretic mobility were attributed to ion-pairing and interactions with micellar aggregates, respectively. An alternative approach, which avoids the complex dynamics of micellar systems, involves using high relative molecular mass polyelectro- lytes. 16717 Such polyelectrolytes should provide an ion-ex- change equilibrium to control relative ion migration. Another approach that can be adopted, when the mobilities of the free cations are similar, is to add a reagent to complex the sample cations partially and, thereby, increase the differences in effective mobility.'* In the same way as polyelectrolytes, macrocyclic pol yethers, e.g., cryptands (the molecular structure of cryptand-22 is shown in Fig.1) and crown ethers, can form remarkable complexes with cations and can be expected to be potential modifiers for the separation of ions by CE. Crown ethers have been demonstrated to be good CE modifiers to separate metal ions.I9 However, crown ethers cannot form complexes with anions and hence cannot be used as CE modifiers to separate anions. In contrast, cryptands not only form complexes with cations at pH >7, but also form stable complexes with anions at pH <7 in which the cryptand can be protonated as cryptand-2H2+. With the assistance of an electrostatic effect, the molecule will be adsorbed on the wall and, in turn, will have consequences for the EOF.Furthermore, the protonated cryptand (cryptand-2H2+) will have various interactions with the analyte anions and hence facilitate the separation. This aim of this work was to separate inorganic/organic anions and inorganic cations with the aid of a cryptand. Cryptand-2H2+ is expected to act as a capillary column wall modifier so that the EOF direction corresponds to that of the electrophoretic mobility of the analyte anions, A * To whom correspondence should be addressed. Fig. 1 Molecular structure of cryptand-22.1108 Analyst, August 1996, Vol. 121 leading to a shortening of the analysis time. Also, the different types of interaction among the anions and cations should promote better separation.Experimental Instrumentation For CE a Jasco (Tokyo, Japan) 890-CE instrument and a Jasco 870-CE UV detector were used. Data acquisition was carried out with a laboratory-built ADC computer system; the data acquisition rate was 5 points s-- 1 . Conventional fused-silica capillaries (id 75 pm, od 360 pm, end-to-end length 80 cm, end- to-window length 60 cm) were obtained from Polymicro Technologies (Phoenix, AZ, USA). Analyte zones were de- tected in the direct and indirect mode via UV absorbance at 214 and 254 nm (deuterium lamp). Samples were introduced hydrostatically by elevating the sample vials. Chemicals Cryptand-22, 4-aminopyridine, KN03, KI, K2Cr04, KSCN, KBr03, NaN03, NH4Cl and KI03 were obtained from Merck (Darms tadt, Germany).p-N i trobenzoic acid, p - hydroxybenzoic acid and p-chlorobenzoic acid were purchased from Nacali Tesque (Kyoto, Japan). Terephthalic and benzoic acids were purchased from Sigma (St. Louis, MO, USA). All other chemicals were of analytical-reagent grade from several suppliers. De-ionized water from a Milli-Q system (Millipore, Bedford, MA, USA) was used to prepare all buffers and sample solutions. Electrophoretic Procedures Capillaries were conditioned with the separation electrolyte for 30 min before the first run and for 2 min between runs; this has been shown to lead to improved migration time reproduci- bility.*O Resolution (Rs) was calculated from Rs = 2Atm,,/(W1 + W2), where Atrllrg is the difference in the net migration time between zones I and 2, and Wl and W? are their respective base widths (in time units).Dimethyl sulfoxide (DMSO) was added to the samples as a neutral marker for electrophoretic mobility determination. The electro-osmotic mobility (pCo, in cm2 V-I s-I) was calculated by the following equations: Peo = jcilJ(trnv> k e = Pohs - keo where Id and I, are the length of the capillary to the detector and the total length of the capillary, respectively, V is the running voltage and t, is the migration time of the neutral marker. The electrophoretic mobility of the analyte, pe, was obtained by subtracting peo from the observed mobility, Fobs. Results and Discussion In order to make cryptand-22 function in a similar way to a quaternary ammonium salt in terms of a capillary column wall modifier, the first step is to protonate the cryptand. It was envisaged that, for CE buffer containing cryptand-22, the use of various acids to adjust the pH values of the solution for protonation of the cryptand might cause different effects towards the EOF.Therefore, the effect of various inorganic and organic acids on the EOF was investigated; the order of the EOF was CH3COOH (EOF = 4.82 X cm2 V-I s-I) > HC1 (4.12) > H2S04 (3.00) > H3P04 (2.23). The mobility of the EOF of the organic acid part was almost identical and had a faster EOF than the inorganic part. In general, one cryptand- 2H2+ will attract two univalent anions. The carboxylate anions C3HTCOOH (4.71) > C2HSCOOH (4.60) > HCOOH (4.35) > are larger than inorganic anions.Therefore, the interaction with cryptand-2H2+ will be reduced, thereby increasing the inter- action between cryptand-2H2+ and the capillary wall and leading to a higher reversed EOF. Of the inorganic acids, e.g., HCl, H2S04 and H3P04, the EOF was greatest in HC1 and smallest in H3P04. A possible reason for this might be that the univalent C1- ion has less interaction with cryptand-2H2+, thus making it easier for cryptand-2H2' to adhere to the negatively charged surface of the column wall so that the EOF is reversed with a higher velocity. On the other hand, owing to the stronger electrostatic effect with trivalent phosphate, cryptand-2H2+ is unable to interact effectively with the anions on the column surface, thereby resulting in a smaller EOF. Fig. 2 displays the effect of pH on the EOF.When the buffer contained 1 X mol dm-3 cryptand-22 at pH 8.0, a sufficient amount of cryptand-2H2+ was available to effect adsorption on the column wall, and subsequently to reverse the EOF. The lower the pH value, the more rapid was the EOF turning velocity. The EOF was stable within the pH range 4-6. At this stage, the interaction between the capillary column surface and cryptand-2H2+ was saturated. The effect of cryptand concentration on the EOF velocity at pH 7.0 and 9.0 was also investigated. The experimental results showed that, at pH 7.0, the EOF was reversed when the cryptand concentration was 1 X mol dm-3; increasing the amount of cryptand-22 further did not significantly influence the EOF. At pH 9.0, increasing the concentration of cryptand slowed down the EOF, but even at a cryptand concentration of 1 X mol dm-3 the EOF was not reversed.This indicates that certain cryptands are not protonated at pH 9.0. The CE system was employed to separate the following inorganic anions: I-, Nos-, SCN-, BrO?-, Cr042- and IO3-. The reason for selecting these anions was that they exhibited absorption in the UV range. Without cryptand-22 in the CE buffer, the EOF was in the opposite direction to the electro- phoretic mobility of the anions, and had a lower velocity compared with the electrophoretic mobility velocity, except for IO3-. That is, with the reversed voltage method, the electro- pherogram only showed five peaks (excluding IO3-) and the analysis time was prolonged. Fig. 3 presents the relationship between the concentration of cryptand-22 and the electro- phoretic mobility of various anions.The results indicated that increasing concentrations of cryptand-22 led to a decrease in the electrophoretic mobility velocity of the anions. Also, Cr042- showed the largest effect, probably because Cr042- had the strongest electrostatic effect with cryptand-2H2+, which re- sulted in the greatest interaction. As shown in Fig. 4, with 1 X mol dm-3 cryptand, approximately 6 min was necessary to obtain the electropherogram of the six anions. In another experiment, 35 min was necessary and only five peaks (103- excluded) were observed in the absence of the cryptand. However, owing to a decrease in the EOF and its reversal so that 6 , , / I -a 3 4 5 6 7 8 9 1 0 1 1 PH Fig.2 adjusted by use of acetic acid. Effect of pH on EOF in 10 mmol dm-3 cryptand-22; pH wasAnalyst, August 1996, Vol. 121 1109 the EOF was in the same direction as the electrophoretic mobility of the anions, accompanied by an increase in the cryptand-2H2+ concentration, the total separation time was reduced. For organic anions, without cryptand-2HZ' in the CE buffer, the electropherogram obtained is shown in Fig. 5. p-Ni- trobenzoic acid and p-chlorobenzoic acid could not be separated from each other; in addition, the latest peak was terephthalic acid. The relationship between the cryptand concentration and the resolution (R) of p-nitrobenzoic acid and p-chlorobenzoic acid was investigated. The results showed that the R value of p- nitrobenzoic acid and p-chlorobenzoic acid reached a maximum with a concentration of cryptand-22 between 1 X 10-3 and 1 X mol dm-3.Within this concentration range, the two substances could be separated at pH 7.0. The pH was set at 7.0 to ensure that these substances existed as anions. p-Nitrobenzoic acid could be separated from p-chlorobenzoic acid within a certain cryptand-22 concentration range primarily because the nitro group had a stronger electron withdrawing effect and made the net negative charge of the anion less than that of p-chlorobenzoic acid. This subsequently resulted in less obstruction from cryptand-2H2+. Fig. 6 shows the electrophero- gram of the five organic anions at pH 7.0. The anions were -* x -14 I I I I 1 I 0 1 2 3 4 5 6 Concentration of cryptand/mol dm-3 Fig.3 Plot of electrophoretic mobility versus the concentration of cryptand-22. Experimental conditions as for Fig. 2 (at pH 7.0). 1, I-; 2, NO3-; 3, SCN-; 4, Cr04*-; 5 , Br03-; 6, 103- ( X rnol dm-3). 300 250 > 0 c .- $ 200 c c - 150 100 0 2 4 6 Ti m e/m i n Fig. 4 Electropherogram of six inorganic anions in 10 mmol dm-3 cryptand-22 at pH 7.0 (adjusted with acetic acid). Voltage was -20 kV. Concentration of anions was I W 4 rnol dm-3. 1, I-; 2, NO3-; 3, SCN-; 4, Br03-; 5 , Cr04z-; 6, I03-. completely separated and the migration order was entirely different from that in the absence of cryptand-22. Because the analyte cations do not absorb in the UV/VIS range, they were detected by indirect absorption. Tn order to achieve indirect UV/VIS detection, an ion, called a visualization agent, which absorbs light in the UV/VlS range, is added to the background electrolyte to create a high background signal.When the analyte ions displace the visualization agent ions from the buffer, the analyte ions are indirectly detected by the change in the background signal. The highest sensitivity is achieved for analyte ions having an effective mobility close to the mobility of the absorbing ion.21 In order to ensure that part of the cryptand was not protonated, the running electrolyte was maintained at pH 9.0. The UV-absorbing agent selected should have a high pK, in order to achieve the same electrophoretic direction as the cations below pH 9.0. The pK, of 4-aminopyridine is 9.1 1; therefore, sensitive detection can be achieved even at a pH of 10.In addition, this compound could act as a UV- absorbing agent at a wavelength of 254 nm.22 The charge density and ionic diameter of K+, Na+ and NH4+ are different. Therefore, the interaction between these cations and cryptand-22 is not identical. This phenomenon will increase the resolution of the cations. As Fig. 7(a) shows, Na+ and NH4+ could not be separated efficiently when no cryptand had been added. In Fig. 7(h), with the addition of 1 X mol dm-3 300 3 501 I I I I 1 0 2 4 6 8 1 0 1 2 1 4 Time/min Fig. 5 Electropherogram of five organic anions in 10 mmol dm-' CH3COONa at pH 7.0 (adjusted with acetic acid). Voltage was 20 kV. Concentration of acids was 10-4 rnol dm-?. 0, Dimethylsulfonic acid; 1 , p - hydroxybenzoic acid; 2, p-chlorobenzoic acid; 3, p-nilrobenzoic acid; 4, benzoic acid; 5 , terephthalic acid.600 700 1 > c. -z 500 400 300 200 W .c. i 4 3 100 -- 6 8 10 12 14 16 18 20 Time/min Fig. 6 Electropherogram of five organic anion5 in 0.5 mmol dm-? cryptand-22 at pH 7.0 (adjusted with acetic acid). Voltage was -20 kV. Concentration of acids was 10-4 rnol dm-3. 1, p-Hydroxybenzoic acid; 2, p-chlorobenzoic acid; 3 , p-nitrobenzoic acid: 4, benzoic acid; 5 , terephthalic acid.1110 Analyst, August 1996, Vol. 121 450 400 cryptand, the resolution of K+, Na+ and NH4+ is better than in Fig. 7(a). Obviously, addition of cryptand increases the difference between the cations. As Fig. 7(c) shows, by adding aqueous 15% methanol, the resolution is further improved. The addition of methanol decreases the solvation of the cations in water and increases the interaction between the cations and the cryptand.Adding acetonitrile also had the same effect. The addition of organic solvent to the CE buffer will change the viscosity of the solution and affect the velocity of the EOF. In this work, the effect of methanol and acetonitrile on the EOF and the electrophoretic mobility of the anions and cations was investigated. The experimental results revealed that methanol and acetonitrile have very different effects on the EOF at pH 7.0. Methanol decreased the velocity of the EOF; however, acetonitrile increased the velocity. This can be explained by the hydrogen bonding between methanol and cryptand-2H2+, which leads to less interaction on the capillary column wall and, accordingly, to a decrease in the positive charge on the wall, - - I I I I 600 ( b ) - - 5 550 - 4- 450 - 400 1 1 2 350 ' I 1 I 1 1 2 3 4 5 6 650 I I I I 1 500 1 - 450 - J - 5 6 7 8 9 10 Ti me/m i n Fig.7 Electropherogram of three inorganic cations. Experimental condi- tions: 4-aminopyridine, 5 X 10-3 mol dm-3, pH 9.0 (adjusted with acetic acid). Voltage was 20 kV. (a), Non-cryptand; (h), adding 10-4 rnol dm-3 cryptand; (c), adding mol dm-3 cryptand and 15% methanol. 1, K+; 2, Na+; 3, NH4+ ( mol dm-'). resulting in a decrease in the EOF. A possible explanation for the increase in the velocity of the EOF on adding acetonitrile is that the solvation of cryptand-2H2' in water decreases in the presence of acetonitrile, while reinforcing the interaction between cryptand-2H2+ and the anions on the capillary column wall.This consequently causes an increase in the velocity of the EOF. At pH 9.0, the addition of organic solvent also affects the velocity of the EOF in the same way as at pH 7.0; however, the trend in the change is slow. For inorganic/organic anions and inorganic cations, the experimental results indicate that, with the addition of an organic solvent to the CE buffer, the electrophoretic mobility of all the ions tends to be reduced. This is because the organic solvent decreases the solvation of the ions in water and increases the interaction between the ions and the cryptand. The authors express their appreciation to the National Science Council of the Republic of China in Taiwan for financial support. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Okada, T., and Kuwamoto, T., Anal.Chem., 1985, 57, 258. Huang, X., Luckey, J. A., Gordon, M. J., and Zare, R. N., Anal. Chem., 1989, 61,766. Cassidy, R. M., and Elchuk, S., Anal. Chem., 1982, 54, 1558. Knight, C. H., Cassidy, R. H., Recoskie, B. M., and Green, L. W., Anal. Chem., 1984, 56, 474. Jackson, P. E., and Haddad, P. R., J . Chromatogr., 1988, 439, 37. Cassidy, R. M., and Elchuk, S., J . Chrornatogr. Sci., 1983, 21, 454. Devevre, O., Putra, D. P., Botton, B., and Ganbaye, J., J.Chromatogr. A, 1994,679, 349. Jones, W. R., and Jandik, P., J. Chromatogr., 1991, 546, 445. Buchberger, W., and Haddad, P. R., J . Chromatogr., 1992, 608, 59. Cousins, S. M., Haddad, P. R., and Buchberger, W., J . Chromatogr. A , 1994, 671, 397. Jones, W. R., and Jandik, P., J. Chromatogr., 1992,608, 385. Chen, M., and Cassidy, R. M., J . Chromatogr., 1993, 640,425. Lee, Y. H., and Lin, T. L., J. Chromatogr. A , 1994,675, 227. Foret, F., Krivankava, L., and Bocek, P., Capillary Zone Electro- phoresis, VCH, New York, 1993, ch. 6. Lee, T. H., and Lin, T. I., J . Chromatogr. A, 1994, 680, 287. Stathakis, C., and Cassidy, R. M., Anal. Chem., 1994, 66, 2110. Terabe, S., and Isemura, T., Anal. Chem., 1990, 62, 652. Shi, Y., and Fritz, J. S., J. Chromatogr., 1993, 640, 474. Shi, Y., and Fritz, J. S., J . Chromatogr. A, 1994, 671, 429. Smith, S. C., Strasters, J. K., and Khaledi, M. G., J . Chrornatogr., 1991, 559, 57. Forest, F., Fanali, S., and Ossicini, L., J. Chromatogr., 1989, 470, 299. Beck, W., and Engelhardt, H., Fresenius' J . Anal. Chem., 1993,346, 618. Paper 6/01 685E Received March I I , I996 Accepted May 15, 1996

ISSN:0003-2654    

DOI:10.1039/AN9962101107

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1111-1114) 1111 Development of a Tube Enzyme lmmunoassay for ‘On-site’ Screening of Urine Samples in the Presence of P-Agonists Willem Haasnoota, Lucia StreppeP, Geert Cazemiera, Martin Saldenb, Piet Stoutena, Martien Essersa and Piet van Wichenb a State Institute for Quulity Control of Agrkultural Products (RIKILT-DLO), P.O. Box 230, 6700 AE Wageningen, The Netherlands h Euro-Diagnostica BV, P.O. Box 5005, 6802 EA Arnhem, The Netherlands An on-site screening test for the detection of P-agonistic drugs in urine was developed. The test is based on the principle of an enzyme immunoassay in polystyrene tubes. Results can be obtained by visual interpretation or by measurement with a differential photometer. The total time required to perform the test for a set of samples (five samples + one cut-off standard) is about 20 min (visual interpretation) with an additional 2 min for an instrumental interpretation. Owing to its speed and simplicity, the test can be performed in slaughter- and farmhouses.In the tube test, a mixture of antibodies raised against clenbuterol and salbutamol is used, which makes this test sensitive towards a range of P-agonists (multi-test). In this study, the test was applied to the screening of 269 bovine urine samples. Bovine urine samples with a level of 3 ng ml-1 of clenbuterol and higher were found positive with this on-site test. Owing to fewer matrix effects, a lower level (1 ng ml-l) could be detected in calf urine. The detection of positive samples at the place of sampling can result in more effective control of the illegal use of P-agonists.Keywords: P-Agonists; on-site screening; tube test; enzyme immunoassuy; urine Introduction According to EC document 86/469/EEC,l which bans the use of any drug to improve animal growth, P-agonists have to be included in the control programme of all EU member states. Within The Netherlands, different sample materials (urine, faeces, cattle feed, tissues, hair, eyes, etc.) are offered for analysis for the presence of P-agonists. Of these samples, urine, which can be taken in both farm- and slaughterhouses, is still the most frequently analysed material. The amount of samples analysed within the Dutch Ministry of Agriculture Nature Management and Fisheries for the presence of (3-agonists is about 20000 per year.To manage all these samples, a two-stage control programme, using microtitre plate enzyme immunoasssays2~3 (EIAs) for screening and GC- MS for confirmation4.5 is applied. For screening, two types of microtitre plate EIAs (clen- buterol- and P-agonist-EIA) are used. In the clenbuterol-EIA, antibodies raised against clenbuterol-BSA are used while the p- agonist-EIA is based on a mixture of anti-clenbuterol and anti- salbutamol. In both EIAs, salbutamol hemisuccinate-horse- radish peroxidase (Sal-HS-HRP) is used as the enzyme conjugate. The clenbuterol-ETA and the P-agonist-EIA show high cross-reactivities towards several P-agonists (see Table 1). With both EIAs, urine samples can be analysed after a minor sample preparation (direct method), i.e., a five-fold dilution in buffer and pH adjustment (pH 7 f 0.5).Applying these direct screening methods, blank urine samples show blank values (background) that differ with the age of the sampled animals.3 Blank calf urine samples show blank values of < 0.5 ng ml-I whereas the blank values of bovine urine samples can be as high as 3 ng ml-1. These blank values can be reduced by applying a clean-up procedure, e.g., extraction with is~butanol.~ Using the direct EIAs, the concentrations found in urine samples obtained from animals treated with clenbuterol are, in general, higher (2-3 times) than those found by GC-MS.3 These higher results can only partly be explained by the measured background. Another reason might be the presence of metabolites of clenbuterol.Recently, the presence of polar metabolites in bovine urine (clenbuterol arylhydroxylamine and arylsulfa- mate) was confinned.6 Applying isobutanol extraction for clean-up, the results obtained with the EIAs and GC-MS are of the same order, which can be explained by a reduced matrix effect and probably also by non-extracted polar metabolites of clenbuterol. The above screening and confirmation methods are suitable for laboratory use only. Inspection services in The Netherlands prefer to perform a first screening at the place of sampling (farm- or slaughterhouses). A first on-site impression of the presence or the absence of 13-agonists can result in more effective control. The ‘on-site’ screening assay described here is based on the same principle as the competitive microtitre plate enzyme immunoassay (P-agonist-EIA).Changes have been made in the format (tubes instead of microtitre plates), sensitivity, speed, sample preparation, sample volume and interpretation (visual or Table 1 Cross-reactivity of the clenbuterol- and (3-agonist-EIA towards several P-agonists Percentage of cross-reactivity P- Agonist Clenbuterol Salbutamol Bromobuterol Cimbuterol Mapenterol Mabuterol Tulobuterol Clenpenterol Carbuterol Terbutaline Pirbuterol Cimaterol * Not determined. Clenbuterol- EIA 100 6 100 60 80 70 2 nd* 5 4 4 6 P- Agonist- EIA 100 100 1 00 75 70 60 50 50 40 40 30 101112 Anulwt, August 1996, Vol. 121 with a portable photometer). The tube test kit was developed for direct application to urine samples. The results obtained with the tube test were compared with those obtained by GC-MS.Experimental Materials Clenbuterol hydrochloride was obtained from Sigma (St. Louis, MO, USA) and salbutamol sulfate and terbutaline sulfate from Bufa-Chemie (Castricum, The Netherlands). All other P- agonist standards were obtained from R. Schilt (RIKILT-DLO, Wageningen, The Netherlands). Horseradish peroxidase grade I (EC 1.1 1.1.7) was obtained from Boehringer (Mannheim, Germany). MaxiSorp StarTubes (75 X 12 mm) were obtained from Nunc (Roskilde, Denmark). Antibodies raised against clenbuterol- and salbutamol-BSA and salbutamol-peroxidase conjugate were prepared as described previously.3 N,O-Bis- (trimethylsily1)trifluoracetamide (BSTFA) and N-methyl-N- (tett-butyldimethylsilyl)trifluoracetamide (MTBSTFA) were purchased from Pierce (Rockford, IL, USA).Equipment A differential photometer with a 450 nm filter (Artel, Windham, ME, USA) was used for the instrumental interpretation of the tube test. For confirmatory analysis, an HPMSD Model 5971A mass spectrometer equipped with a Model 5890 gas chromatograph (Hewlett-Packard, Avondale, PA, USA) and operating in the multiple-ion detecton (MID) mode was used. Ammonia was used as the reagent gas and GC separation was achieved on a DB5 capillary column (J&W Scientific, Folsom, CA, USA) (30 m X 0.25 mm id, film thickness 0.25 pm). The GC oven was programmed from 1 10 to 280 "C at 10 "C min-I and held at 280 "C for 5 min. Tube Enzyme Immunoassay The tube EIA test kit (5065BAGT; Euro-Diagnostica, Arnhem, The Netherlands) contains a test-tube rack (for six tubes), disposable droppers, 12 pre-coated StarTubes and six dropper vials containing: lyophilized Sal-HS-HRP (vial B), salbutamol standard solution (3 ng ml-I; cut-off standard) (vial C), a positive control standard (50 ng ml-1 clenbuterol) (vial D), one- component substrate solution (TMB) (vial E), dilution buffer (vial F) and stop solution (2 mol 1-I sulfuric acid) (vial G).Test kit procedure First, the lyophilized enzyme conjugate [SAL-HS-HRP (vial B)] is dissolved by the addition of 4 ml of dilution buffer (vial F). The number of StarTubes needed [amount of samples plus one for the control standard (cut-off standard)] are placed in the tube rack. To the tube placed at the first position of the tube-rack (marked 'S'), two drops of the standard solution [cut-off standard (vial C)] are added.To the other tubes (positions marked 1-5) two drops of the urine samples, are added using disposable droppers. Four drops of the reconstituted enzyme conjugate (vial B) are added to each tube and the whole is mixed by shaking the rack for 10 s. After 10 min at room temperature, the tubes are washed five times with cold tap water as follows: the tubes are emptied by inverting the rack, the tubes are filled with cold tap water, the water is removed and this procedure is repeated four times. After the washing procedure, water droplets are removed by blotting the open ends of the inverted tubes on clean absorbent paper. Six drops of substrate solution [TMB (vial E)] are added and the rack is shaken for 10 s and left at room temperature for 5-10 min for colour development.Visual interpretution. The colours in the sample tubes are compared with the colour of the cut-off standard (tube 'S'). Samples as dark as or darker than the cut-off standard are considered negative. Samples showing less colour intensity than the cut-off standard are considered suspect. Instrumental interpretation. Five drops of stop solution (vial G) are added to each tube and, after shaking the rack for 2 s, the absorbance is measured at 450 nm using an empty StarTube to blank the photometer. Samples with an absorbance equal to or higher than that of the cut-off standard are considered negative and samples with lower absorbance as suspect. GC-MS procedure The procedure described by van Rhijn et al.5 was used.Buffered urine was applied to an immunoaffinity column containing immobilized antibodies raised against clenbuterol coupled to BSA. The column was washed and P-agonists were eluted with methanol-0.1 mol I- acetic acid (pH 2.75) (70 + 30 v/v). The resulting extract was applied to a CIS cartridge column. The column was washed and P-agonists were eluted with methanol- acetonitrile (85 + 15 v/v). The eluate extract was split into two equal parts and evaporated under nitrogen. One of the parts was derivatized using BSTFA to obtain the TMS derivative and the other with MTBSTFA to obtain the tBDMS derivative. After derivatization, the sample aliquots were recombined and used for GC-MS analysis in the chemical ionization mode. Samples Bovine urine samples (n = 249) were obtained from the General Inspection Service (Kerkrade, The Netherlands).Lyophilized blank bovine urine samples (n = 20) were obtained from the bank of reference blank samples prepared by the European Community Reference Laboratory (CRL) of the National Institute of Public Health and the Environment (RIVM, Bilthoven, The Netherlands). Results and Discussion The tube test was developed according to the principle of a competitive enzyme immunoassay. Just as in the commercially available microtitre plate EIAs (clenbuterol- and (3-agonist- EIA; Euro-Diagnostica), the enzyme conjugate is lyophilized. This lyophilized conjugate is stable for several months when stored at 4-8 "C. Just before use, this lyophilized conjugate is dissolved in buffer.This dissolved conjugate is stable for about 7 days at 4-8 "C. The test had to be made suitable for on-site application to urine samples directly (no sample preparation). Therefore, changes had to be made in the format (tubes instead of microtitre plates and droplet vials instead of pipettes), sensitivity, sample volume, speed and interpretation of results (visual or with a portable photometer). Sensitivity To obtain a wide-range on-site test for the detection of several (3-agonists, the reagents used for the 6-agonist-EIA were chosen to prepare the tube test. To make the tube test as simple as possible, urine had to be introduced without any pre-treatment and the sensitivity had to be adjusted. The calibration graph used in the microtitre plate P-agonist-EIA has a range of 0.1-5 ng ml- salbutamol (Fig.1). Using this EIA with SO pl portions of fivefold times diluted urine samples (10 p1 portions of urine in the test), the range of detection in urine is 0.5-25 ng ml-1. In the tube test, two drops (80 pl) of undiluted urine samples were used. The tube test system was calibrated to give a sensitivity in a suitable range (0.5-50 ng mi-1; see Fig. 1 ) for direct application to urine samples. In the microtitre plate EIA, antibodies are added to the assay as late as possible, i.e., after the introduction of samples/Analyst, August 1996, Vol. 121 1113 standards and enzyme conjugate. To simplify the tube test, the tubes were pre-coated with the antibodies and samples/ standards were added to the test first, followed by the enzyme conjugate. Compared with the microtitre plate EIA, the binding of the analytes to the antibodies is favoured in the tube test.Reversing the order of addition, i.e., first the enzyme conjugate and then the samples/standards, resulted jn a much lower sensitivity. Hence, the prescribed order of addition of samples and reagents is essential. In the microtitre plate EJA, a calibration graph is constructed by analysing a range of six different standard solutions of salbutamol. In the tube test one calibration standard had to be sufficient. After several experiments with positive and negative urine samples, a standard solution of 3 ng ml-1 of salbutamol was selected as the calibration standard (cut-off standard). Compared with just buffer (Bo), this cut-off standard showed a BIBo of 57 k 7% (n = 18).Urine samples with greater or the same colour as this standard were considered to be negative and urine samples with less colour were considered to be suspect. In addition to this calibration standard, a positive control standard (50 ng ml- * clenbuterol) was added to the test-kit. This positive control was added to the tube test kit to establish its performance. Compared with just buffer (Bo), this positive control showed a BIB0 of 33 & 1% ( n = 5). Speed The time of the first incubation step in the tube test procedure, i.e., after the introduction of sample/standard followed by the enzyme conjugate, was varied from 0.5 to 30 rnin (see Fig. 2). 100 1 0.05 0.1 1 10 100 Concentration/ng ml-’ Fig. 1 tube test ( A ).Calibration graphs obtained with the (3-agonist-EIA (a) and the 2.50 2.00 a, 0 $ 1.50 e % 2 1.00 0.50 0.00 + * + + + + + + + A A ++ A * A a b 0 6 12 18 24 30 Incubation time/min Fig. 2 Time of the first incubation step in the tube test plotted against the absorbance measured at 450 nm. +, Two drops of water; and A, two drops of cut-off standard (3 ng inl-1 salbutamol). As shown by adding a blank (two drops of water) and the cut-off standard (3 ng ml-1 salbutamol) to the test, 50% of the maximum bound enzyme conjugate (50% maximum colour intensity) was found after an incubation time of only 0.5 min. Maximum colour intensity was achieved after a 10 min incubation and this incubation time was chosen in the tube test protocol. The first incubation is followed by a five-times washing procedure (running tap water) and the addition of a chromogen/substrate solution.Within 5 rnin after the addition of chromogen/substrate solution, test results can be obtained visually (blue colour). The total time to perform the test using visual interpretation is about 20 min. Instrumental Interpretation The visual interpretation can be followed by an instrumental interpretation using a portable differential photometer. After the addition of five drops of stop solution, the intensity of the yellow colour can be measured at 450 nm. In the following experiments, the absorbances of samples and standards (cut-off standard and positive control) were measured in the right-hand well of the photometer against air (empty left-hand well). To compare the results between different tube test kits, the results of samples were calculated as percentage of the cut-off standard using the following equation: absorbance of sample/absorbance of cut-off standard X 100% = percentage of the cut-off standard According to this equation, negative samples will have values > 100% and positive samples < 100%.Urine Samples Twenty blank reference bovine urine samples were analysed with the tube test and values (percentages of the cut-off standard) between 125 and 185% were found (mean 150%; Fig. 3). Thus, all samples were indeed found to be negative. The highest percentages (lowest background) were found in calf urine (sample Nos. 5 , 13, 14, 19 and 20). Such age-dependent blank values of bovine urine samples in direct immunoassays have been reported previously.3 Urine samples ( n = 56) in which clenbuterol was detected by GC-MS were analysed with the tube test also and the results were compared (Fig.4). Of these samples, eight were found to be negative (false negative) with the tube test. The concentra- tions of clenbuterol found by GC-MS in these samples ranged from 1.3 to 2.4 ng ml-1. Six mabuterol-containing urine samples (GC-MS values ranged from 2.0 to 2.5 ng ml-1 mabuterol) were analysed by the tube test, which resulted in percentages of the control standard 200 Y- 140 c g 120 a, + + + + + + + + + LL 100 1 2 3 4 5 6 7 8 9 1011121314151617181920 Urine sample Fig. 3 Results obtained with the tube test in the analysis of 20 blank bovine urine samples. The dashed line indicates the mean blank value ( I 50%).1114 Analyst, August 1996, Vol.121 of 102-1 1 1 % (false negatives). The tube test results were found to be positive for urine samples containing 4.5 and 20 ng ml-1 mabuterol (80 and 67% of the cut-off standard, respectively). These results correspond with the cross-reactivity found for mabuterol (60%, see Table I). The tube test was applied to 185 bovine urine samples which were found to be negative by the screening methods applied in the laboratory (clenbuterol- and (3-agonist-EIA). The tube test results ranged from 200 to 83% of the cut-off standard. Of these samples, 18 (9.7% of the total samples) showed a response of < 100% (99-83%) of the cut-off standard (false positive). To decrease the percentage of false positives, the cut-off level can be changed by using, for instance, a level of 80% of the cut-off standard, which corresponds to about 4 ng ml-1 of clenbuterol (see Fig.4). On the other hand, for calf urine samples (with lower blank values), the cut-off level can be changed to 120%, which corresponds to a level of 1 ng ml-1 clenbuterol (see Fig. 4). In The Netherlands, the tube test is applied by the General Inspection Service at farmhouses as a pre-screening of urine samples to detect the possible presence of animals treated with CJ-agonists. Being aware of the lower sensitivity of the tube test compared with the methods applied in the laboratory, all samples (negatives and suspected positives) are re-analysed in the laboratory. The tube test results are used to locate treated animals directly at the farm and, if suspected samples are found, to increase sampling.Conclusions The (3-agonist tube test can be used for on-site application (farm- and slaughterhouses) to urine samples. Results can be U I 1 10 100 Concentration/ng mi-’ Fig. 4 Tube test results (expressed as percentages of the cut-off standard) compared with the results obtained by GC-MS. obtained by visual interpretation within 20 min. The tube test can be used to detect clenbuterol and mabuterol in bovine urine at a level of 3 4 ng ml-I and higher. Owing to the lower background, the detection limit in calf urine is lower (about 1 ng ml-l). Owing to the application of a mixture of antibodies raised against clenbuterol and salbutamol, the tube test detects a range of CJ-agonists (multi-test).In general, the sensitivity is lower than that of the methods applied in the laboratory (1 ng ml- I ) and therefore the on-site test can only be used for pre- screening. The detection of p-agonists at the place of sampling can result in more effective control of the illegal use of (3- agonis ts. The authors thank R. Schilt (RIKILT-DLO, Wageningen, The Netherlands) for supplying the CJ-agonist standards and Leen van Ginkel (CRL; RIVM, The Netherlands) for supplying the blank reference urine samples. References Council of the European Communjties, Off. J . Eur. Commun., 1986, L275, 36. Haasnoot, W., Hamers, A. R. M., van Bruchen, G. D., Schilt, R., and Frijns, L. M. H., in Food Safety and Quality Assurance: Applications of Immunoassay Systems, Proceedings of an International Con- ference, Bowness-on-Windermere, Cumbria, UK, March 19-22, 2991, ed. Morgan, M. R. A., Smith, C. J., and Williams, P. A., Elsevier, London, 1992, p. 237. Haasnoot, W., Cazemier, G., Stouten, P., and Kemmers-Voncken, A., in Residue Analysis in Food Safety: Applications of Immunoassay Methods, ed. Beier, R. C., and Stanker, L. H., American Chemical Society, Washington, DC, 1996, p. 60. Schilt, R., Haasnoot, W., Jonker, M. A., Hooijerink, H., and Paulussen, R. J. A., in Euro-Residue Conference, Proceedings of an International Conference, Noordwijkerhout, The Netherlands, May 21-23, 1990, ed. Haagsma, N., Ruiter, A., and Czedik-Eysenberg, P. B., Rijksuniversiteit Utrecht, Faculteit der Diergeneeskunde, Utrecht, 1990, p. 320. van Rhijn, J. A,, Traag, W. A., Heskamp, H. H., J. Chromatogr., 1993,619,243. Zalko, D., Bories, G., and Tulliez, J., Euro-Residue 111 Conference, Proceedings of an International Conference, Veldhoven, The Nether- Zands, May 6-8, 1996, ed. Haagsma, N., and Ruiter, A., Rijksuni- versiteit Utrecht, Faculteit der Diergeneeskunde, Utrecht, 1996, p. 993. Paper 6/01 654E Received March 8, 1996 Accepted May 23,1996

ISSN:0003-2654    

DOI:10.1039/AN9962101111

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1115-1118) 1115 Near-infrared Optical Detection of Acids in Atmospheric Air by Phthalocyanine Dyes in Polymer Films Luis E. Norena-Franco and Frank Kvasnik Department of Instrumentation and Analytical Science, UMIST, P.O. Box 88, Manchester, UK M60 1 QD Recently available phthalocyanine dyes with absorption bands in the NIR spectral region were investigated as reagents for the detection of vapour leaks of industrial acids such as hydrochloric and acetic acid. The dyes ProJet 830NP and Pro Jet 860NP (Zeneca Specialities) were entrapped in silicone polymer films. The response of the sensing films to the presence of acids was assessed by absorption spectrometry. The acid concentration in the test atmospheres was approximately 4% m/m. The long-term stability of the phthalocyanine dyes and the reversibility of their optical properties after exposure to acid vapours make these materials suitable for optical sensor applications.Keywords: Optical sensor; chemical sensor; near-infrared; indicator dyes; phthalocyanine dyes; silicone polymers Introduction Optical chemical sensors have been the focus of intense research efforts and numerous chemical sensor systems have been developed in recent years. ,2 In many of these systems, the sensing element is a reagent that changes its optical properties in response to the presence of an analyte. The sensing reagent is frequently an indicator dye that has to be immobilized on or entrapped within a suitable matrix. Fibre-optic point and distributed chemical sensors can be located close to potential sources of leaks and therefore experience high concentrations of an analyte.The placement of such sensors can be in remote environments that are hostile or difficult to access and consequently the transmission of the optical signal over long lengths of optical fibre is frequently required. Most of the work on optical chemical sensors is carried out in the visible spectral region where the chemical and optical properties of indicator dyes are well known. However, most optical fibres are silica-cored and have a minimum optical attenuation not in the visible but in the NIR spectral region. NIR-absorbing phthalocyanine dyes are employed in a wide variety of high-technology appli~ations.~-S They are chemic- ally, thermally and optically stable, have narrow and intense absorption bands and the background information on the general aspects of their chemistry and property-structure relationships is available.68 These properties of phthalocyanine dyes appear to be ideal for optical chemical sensing applica- tions; however, to our knowledge, phthalocyanine dyes have not been used for sensing before.The general structure of a metal-based substituted phthal- ocyanine is shown in Fig. 1. The basic chromophore of phthalocyanine is the cyclic conjugated 16-centre ring system consisting of four bridged pyrrole units linked by nitrogen atoms and it contains a total of 18 n electrons. The chromo- phoric structure belongs to the aromatic classification since it possesses (4n + 2)n electrons in the Hiickel series, and such molecules have the n electrons strongly delocalized around their perimeter.The strong absorption bands of phthalocyanine dyes are due to n-n* electronic transitions. Non-substituted phthalocyanines absorb in the visible region. However, the addition of benzene groups or multiple electron-donating substituents moves the absorption bands from the visible to the NIR spectral region by increasing the length of the conjugation and the electronic delocalization. For fibre-optic sensing applications the reagents need to be immobilized in a suitable matrix. The properties of the matrix must be such that the characteristic absorption bands of the indicator dye are not lost, a uniform distribution of the indicator dye is achieved and analytes may rapidly permeate through it.Further, the refractive index of the matrix should be lower than that of silica so that it may be used as a cladding for silica-cored optical fibres in the manufacture of sensing optical fibre guides. The phthalocyanine dyes supplied are only soluble in non-polar solvents, and therefore a non-polar polymer or a polymer of low-polarizability would be the best choice for the immobiliza- tion matrix. Silicone polymers have an amorphous structure that gives low light scattering and a low optical attenuation up to a wavelength of 1600 nm. The silicone polymers also have the highest reported permeability to chemical vapours of any organic polymer, which would give a rapid transducer response. Therefore, a clear, flexible, cross-linked poly(dimethylsi1ox- ane) polymer was selected as the matrix for immobilization of the phthalocyanine dyes.This paper reports the optical characterization of the phthalocyanine dyes ProJet 830NP and ProJet 860NP, which have never previously been used in optical chemical sensors. Initially solution studies were carried out in order to ascertain the potential of the dyes for sensing applications. Results are presented for the two dyes in dichloromethane under neutral and acidic conditions. The preparation of sensing films containing the dyes in cross-linked poly(dimethylsi1oxane) is detailed, and results are presented for the spectra of the sensing films before and after exposure to vapour of hydrochloric and acetic acid. D \ N cu N Fig. 1 General structure of a metal-based substituted phthalocyanine.1116 Analyst, August 1996, Vol.121 Experimental Apparatus Absorbance measurements were performed with a Cary 2300 UV/VIS/NIR spectrophotometer (Varian, Sunbury-on-Thames, UK). In the NIR region, the spectrophotometer has a specified wavelength accuracy of 0.8 nm and a resolution of 0.35 nm. The sample cuvettes used were made of silica glass with optical path length 10 mm and capacity 4 ml. Reagents The copper-based phthalocyanines ProJet 830NP (M, 1920) and ProJet 860NP (M, 2250) were obtained from Zeneca Speciali- ties (Blackley, Manchester, UK). Poly(dimethylsi1oxane) (aver- age M, 16 500) was obtained from Fluorochem (Old Glossop, Derbyshire, UK), tetraethylorthosilicate and tin(r1) 2-ethylhex- anoate from Aldrich (Gillingham, Dorset, UK) and dichloro- methane, propan-2-01, 1 mol 1-1 hydrochloric acid, 36% m/m hydrochloric acid, 99.8% m/m acetic acid and 1 moll-’ sodium hydroxide solution were obtained from Fisons (Loughborough, UK) .Procedures Dyes in solution ProJet 830NP and ProJet 860NP dyes were dissolved in dichloromethane to yield 5.2 X and 4.4 X 10-6 mol 1-I solutions, respectively. Acidified dye solutions were prepared by adding separately 20 p1 of acetic acid, 200 pl of acetic acid, 20 pl of 1 mol 1 - l hydrochloric acid, 200 pl of 1 mol 1-1 hydrochloric acid and 20 pl of 36% m/m hydrochloric acid to 3.5 ml of solution of each dye. The absorption spectrum of the dyes in solution was recorded over the wavelength range of 600 to 1200 nm. Dyes in films The silicone elastomer was a cross-linked silanol-terminated poly(dimethylsi1oxane). This can be prepared by cross-linking with a tetraalkoxysilane and a tin(I1) organometallic salt as catalyst.For the work reported here, 10 g of the silicone polymer was prepared from 8 g of poly(dimethylsiloxane), 2 g of tetraethyl orthosilicate, 3 g of propan-2-01 and 0.1 g of tin(I1) 2-ethylhexanoate. The dye-polymer formulations were pre- pared by adding 3 ml of dye in dichloromethane solution to 10 g of polymer preparation at the beginning of the cross-linking process. The sensing films were prepared by dip-coating glass slides manually in the dye-polymer mixture and leaving the slide to dry and harden overnight in air at room temperature. The total thickness of the sensing films was calculated by spectrophotometric interference to be in the range 16-18 pm.The sensing films were exposed to chemical vapours in a sealed flask whose atmosphere contained approximately 4% m/m of each acid in air. This acid concentration was considered to represent a typical low exposure level of a sensor next to a leak source. The concentrations of the test atmospheres were checked by absorption in standard sodium hydroxide solution followed by back-titration with standard hydrochloric acid solution. Each sensing film was supported in a sample cuvette so that the film would be perpendicular to the light path of the spectrophotometer. The absorption spectra of the films were recorded over the wavelength range 500-1200 nm before and after exposure to acid vapour. The response time of the sensing films to hydrochloric and acetic acid vapour was measured with the Varian Cary 2300 spectrometer in the absorbance versus time mode. The wave- length was set to 8 10 nm for the ProJet 830NP-silicone sensing film and to 840 nm for the ProJet 860NP-silicone sensing film.The closed circuit test atmospheres were circulated by a small pump from the flask through the cuvette placed inside the spectrometer. Results and Discussion Dyes in Solution The NIR absorption maximum of ProJet 830NP in dichloro- methane is observed near 810 nm and that of ProJet 860NP in dichloromethane near 840 nm. Their molar absorptivities were 7.3 X lo4 and 9.3 X 104 1 mol-1 cm-l, respectively. These strong NIR absorptions are the result of both n-n* electronic transitions and the extended conjugation of the basic chromo- phore.The absorption peaks of the dyes were asymmetric with a shoulder at a shorter wavelength. These shoulders are characteristic features of the spectra of the phthalocyanine dyes and are due to distinct vibrational transitions.3.6-8. The unsubstituted copper phthalocyanine has an absorption maximum between 657 and 678 nm and a molar absorptivity337 between 10 x 104 and 22 x 104 1 mol-1 cm-1. The large shifts of the absorption band in the substituted copper phthalocyan- ines, ProJet 830NP and ProJet 860NP, are achieved by enlarging the chromophoric system of the unsubstituted copper phthalocyanine with a large number of aryl substituents. The main difference between the two ProJet dyes is that ProJet 830NP has amino groups whereas ProJet 860NP has thio groups in the aryl substituents.The presence of different groups in aryl substituents probably accounts for the observed differences in the molar absorptivities and in the position of the absorption maxima of the two ProJet dyes. In solution phthalocyanine dyes are very stable and only react with relatively high acid concentrations. Fig. 2(a) shows the 0.5 0.4 0.3 0.2 0.1 : 0.0 f2 600 700 800 900 1000 1100 1200 0 2 0.5 ( b ) 0.4 0.3 0.2 0.1 0.0 t I I I I I 1 600 700 800 900 1000 1100 1200 hlnm Fig. 2 Absorption spectra of (a) ProJet 830NP and (h) ProJet 860NP in dichloromethane: A, original; B, after addition of 20 p1 of acetic acid; and C, after addition of 200 p1 of acetic acid.Analyst, August 1996, VoE. I21 1117 absorption spectrum of ProJet 830NP in solution before and after the additions of acetic acid.The addition of 20 pl of the acid caused a - 10% decrease in the height of the absorption peak near 8 10 nm and the addition of 200 pl of the acid caused a -25% decrease in the height of the original peak. Similar decreases in the height of the 810 nm absorption peak were obtained with the additions of 20 and 200 pl of 1 mol 1-1 hydrochloric acid solution. Fig. 2(b) shows the absorption spectrum of ProJet 860NP in solution before and after the additions of acetic acid. The addition of 20 pl of the acid caused a - 10% decrease in the height of the absorption peak near 840 nm and the addition of 200 pl of acetic acid caused a ~ 2 0 % decrease in the height of the original peak.Again, the response to 1 mol 1-1 hydrochloric acid was similar to the response to acetic acid. Fig. 3(a) shows the absorption spectrum of ProJet 830NP in dichloromethane before and after the addition of 36% m/m hydrochloric acid. This addition causes a shift of the absorption spectfum, which also appears much broader. The shape of the absorption spectrum changes from a single maximum absorp- tion peak with a shoulder to two separate absorption peaks. The height of absorption at near 810 nm is decreased by - 50%. The new peaks are located near 830 and near 950 nm. Fig. 3(b) shows the absorption spectrum of ProJet 860NP in dichloro- methane before and after the addition of 36% m/m hydrochloric acid. Again, this addition results in the shift of the spectrum, a broadening of the absorption band and the appearance of two peaks.The height of absorption near 840 nm is decreased by slightly over 55%. The new peaks are located near 880 and near 975 nm. In both cases the separation of the 'new' peaks corresponds approximately to the separation of the main peak and the shoulder in the original spectrum. 0.5 3 g 0.0 I, 600 700 800 900 1000 1100 1200 0.4 4 A 0.3 0.2 0.1 The spectral changes shown in Fig. 2 may be attributed to protonation of either central nitrogen atoms with lone electron pairs, as shown in Fig. 4(a), or outer substituents, as shown in Fig. 4(b). This protonation modifies the delocalized electronic distribution, thus decreasing the mobility of the JC electrons throughout the conjugated chromophoric system. This results in a decrease in the height of the original absorption band.The spectral changes shown in Fig. 3 are unlikely to be caused by a single mechanism. The relative separation of different features in these spectra suggests that the protonation of the molecule, which decreases the height of the absorption peak, is accom- panied by a bathochromic shift of the whole spectrum. The excited state of the phthalocyanine molecules is more polar than their ground state because their n electrons migrate from the centre of the molecule towards the outside during the excitation3 and therefore bathochromic shifts are likely to be observed when hydrogen increases the polarization of the phthalocya- nines. Both dyes remain protonated while in acidic conditions. Returning the solution to neutral pH conditions by the addition of sodium hydroxide in solution reverses the protonation and returns the absorption spectra to the original forms.Dyes in Films In the preparation of the dye-containing films, dichloromethane, an aprotic non-polar solvent, evaporates out of the film and the polymer must act as a new solvent for the dye. Phthalocyanine dyes and poly(dimethylsi1oxane) are both non-polar materials and therefore they are compatible with each other. They formed clear, uniform, dye-solvent-polymer solutions and later clear uniform films. Fig. 5(a) shows the absorption spectra of the ProJet 830NP- silicone film before and after exposure to vapour of hydrochlo- ric and acetic acid. Fig. 5(b) shows the absorption spectra of the ProJet 860NP-silicone film before and after exposure to vapour of hydrochloric and acetic acid.The characteristic absorption bands of the phthalocyanine dyes are not lost when they are entrapped in the silicone elastomer and they still react to the presence of acids. Therefore, the silicone elastomer is a suitable matrix for both dyes. The spectral features near 650 and 1185 nm are generated by inherent absorption of the undoped silicone matrix, as shown in Fig. 5(c). Fig. 5(a) shows a -80% decrease in the intensity of the original absorption band of ProJet 830NP near 810 nm after exposure to hydrochloric acid vapour and a -45% decrease after exposure to acetic acid vapour. Fig. 5(b) shows a ~ 7 0 % reduction in the intensity of the original absorption band of ProJet 860NP near 840 nm after exposure to hydrochloric acid and -40% after exposure to acetic acid.The absorption peaks near 950 and 975 nm in the spectra of the ProJet 830NP-silicone and ProJet 860NP-silicone films, respectively, are barely seen 0.0 I I I 1 I I 1 600 700 800 900 1000 1100 1200 hlnm Fig. 3 Absorption spectra of ( a ) ProJet 830NP and (b) ProJet 860NP in dichloromethane: A, original; and B, after addition of 20 pl of 36% m/m hydrochloric acid. Fig. 4 basic chromophore and (h) protonation of outer substituents. Protonation of phthalocyanine molecules. (a) Protonation of the1118 Analyst, August 1996, Vol. 121 0.065 0.065 ( b ) 0.050 0.055- 0.055 - 0.040 - a 0.050 - 0.050 - 0.035 - 0.045 I f I 1 f 1 0.045 I I I I 1 0.030 I I I 1 I 600 700 800 900 1000 1100 1200 600 700 800 900 1000 1100 1200 600 700 800 900 1000 1100 1200 h/nm Fig.5 vapour; and C, after exposure to hydrochloric acid vapour. (c) Undoped silicone matrix. Absorption spectra of (a) ProJet 83ONP-silicone sensing film and (b) ProJet 860NP-silicone film: A, unexposed; B, after exposure to acetic acid B 0.055 I < B 0.050 I I I I I 0 1 2 3 4 5 Time/mi n Fig. 6 Response times of ( a ) ProJet 830NP-silicone sensing film and (h) ProJet 860NP-silicone sensing film to A, acetic acid vapour and B, hydrochloric acid vapour. after exposure to hydrochloric acid vapour because of the small optical path through the sensing films. Fig. 6(a) shows the time dependence of the absorbance of the ProJet 830NP-silicone film at 810 nm to acid vapour. The response time was measured as the time elapsed between 10 and 90% of the full value, and was found to be approximately 11 s and 2 min for hydrochloric and acetic acid vapour, respectively.Fig. 6(h) shows the time dependence of the absorbance of the ProJet 860NP-silicone film at 840 nm to hydrochloric and acetic acid vapour. In this case, the response time of the film was found to be approximately 40 s and 3 min for the hydrochloric and acetic acid vapour, respectively. The difference in the speed of the protonation reaction is probably caused by different outer substituents of the basic phthalocyanine chromophoric system. ProJet 830NP has aryl- amino outer substituents whereas ProJet 860NP has arylthio outer substituents. Although the non-electronegative character of the two groups hardly differs, the arylamino group is more basic than the arylthio group and reacts more easily with hydrogen ions.However, other effects such as differential solvation of the two dyes in silicone could also affect the response time. The spectral changes generated by exposure to acidic vapours are reversed by ambient air in a few hours. The longer reverse response time in ambient air is due to both the rate of deprotonation of the phthalocyanine molecule and the rate of proton desorption from the polymer matrix. The ProJet 830NP- and ProJet 860NP-silicone sensing films were found to be stable for over 3 months. This indicates long-term stability of the sensing films and contrasts with a particular problem of most other dyes in that they tend to fade, particularly when they are immobilized in any solid medium.9 This particular problem is further accentuated for NIR-absorbing dyes.The sensing films have shown a response to acids, but no response to concentrated ammonia vapours could be measured. The films were also found to be insensitive to temperature and humidity fluctuations in the open laboratory environment. The work reported here has shown that the cross-linked poly(dimethylsi1oxane) is an appropriate medium for immobi- lization of phthalocyanine dyes and that the newly prepared ProJet 830NP-silicone and ProJet 860NP-silicone sensing films absorb in the NIR region. The sensing films containing ProJet 830NP responded faster to the presence of acidic vapour and gave large absorbance changes. Therefore, these films are better suited for the optical sensing of acids in the atrnos- phere. We thank J. Campbell of Zeneca Specialities, (Blackley, UK) for providing the speciality ProJet 830NP and ProJet 860NP phthalocyanine dyes, and CONACYT of Mexico for the provision of the financial support for a PhD studentship for L.E.N.-F. References 1 2 3 4 5 6 7 8 9 Paper 6J02015A Received March 22,1996 Accepted April 26, 1996 Fiber Optic Chemical Sensors and Biosensors, ed. Wolfbeis, 0. S., CRC Press, Boca Raton, FL, 1990. Lieberman, R. A., Sens. Actuators B, 1993, 11, 43. Infrared Absorbing Dyes, ed. Matsuoka, M., Plenum Press, New York, 1990. Gregory, P., High-Technology Applications of Organir- Colorants, Plenum Press, New York, 1991. Fabian, J., Nakazumi, H., and Matsuoka, M., Chem. Rev., 1992, 92, 1197. Fabian, J., and Hartmann, H., Light Absorption of Organic Colorants, Springer, Berlin, 1980. Griffiths, J., Colour and Constitution of Organic Molecules, Academic Press, London, 1976. Gordon, P. F., and Gregory, P., Organic Chemistry in Colour, Springer, Berlin, 1983. Griffiths, J., in Developments in The Chemistry and Technology of Organic Dyes, ed. Griffiths, J., Blackwell, Oxford, 1984, pp. 16-30.

ISSN:0003-2654    

DOI:10.1039/AN9962101115

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1119-1122) 1119 Cyclodextrin-based Optosensor for the Determination of Riboflavin in Pharmaceutical Preparations Zhilong Gonga3b and Zhujun Zhanga,* Department of Chemistry, Shaanxi Normal University, Xian, 71 0062, China Department of Chemistry, Nanjing University, Nanjing, 21 0008, China A flow-through optosensor for riboflavin was developed in conjunction with a flow injection system and using immobilized P-cyclodextrin as the sensing agent. The detection limit for riboflavin was 9 ng ml-1 with a relative standard deviation of 1% for the determination of 0.05 pg ml-1 (n = 7) of riboflavin. The recommended method was successfully tested for determination of riboflavin in pharmaceutical preparations. Keywords: Riboflavin determination; cyclodextrin; optosensor; fluorimetry; flow injection Introduction Cyclodextrins (CDs) are sugar molecules having the structure of a hollow truncated cone with a hydrophobic cavity.Their complexation ability has been attributed to four factors: (1) Van der Waals interactions, (2) hydrogen bonding, ( 3 ) displacement of high-energy water molecules from the cavity and (4) release of the strain energy of the CD on inclusion of a guest molecule. The hydrophobicity of the guest molecule also plays a key role in the stability of the complex. The stoichiometry of the complexes is not necessarily 1 : I ; instances of 2 : 1 and 1 : 2 complexes are known.1.2 Fluorescence is enhanced by inclusion complexation of many fluorophores with cyclomaltoheptaose (P-cyclodextrin; P-CD).Inclusion of the analyte (inclusate) in the cavity of p-CD places it in a more hydrophobic environment where it becomes protected from bulk solution quenchers (e.g., water). The extent of complexation (i.e., selectivity and sensitivity) is strongly dependent on the analyte geometry and functional group orientation. Many methods for the determination of riboflavin have been described, including spectrophotometric,'-5 HPLC,6,7 fluori- metrics.9 and chemiluminescence methods.10 To the best of our knowledge, no reports exist on the application of a CD-based flow-through optosensor for the determination of riboflavin. In this paper, a flow-through optosensor is reported that uses immobilized CD as a reagent for the semi-selective and sensitive determination of riboflavin in aqueous media, combin- ing the selectivity of the CD with enhancement of the fluorescence of the complexed molecule.The experimental results show that the sensor has a higher sensitivity and selectivity than the methods mentioned above, and requires only simple, conventional instrumentation and a simple procedure. Experimental Reagents A stock standard solution of riboflavin (0.5 mg ml-1) was prepared by dissolving 50 mg of riboflavin (Shanghai No. 1 * To whom correspondence should be addressed. Chemicals Factory, Shanghai, China) in an appropriate volume of water and diluting to 100 ml with water. Working standard solutions of riboflavin were prepared by dilution. p-Cyclodextrin (Sigma, St. Louis, MO, USA) was stored in a vacuum desiccator. Silica gel (100-120 mesh) was dried in vacuo at 150 "C for 6 h and stored in a desiccator.All other reagents were of analytical-reagent grade and distilled, de-ionized water was used throughout. Tabellae vitamini B2 and Tabellae vitamini B compositae tablets were purchased from the local market and stored in a refrigerator. The carrier solution used in the flow injection (FI) experi- ments was 0.1 moll-' NaH2P04-0.1 moll-' Na2HP04 buffer solution (pH 6.5). Instrumentation Fluorescence emission measurements were carried out with a Perkin-Elmer LS50 luminescence spectrometer, which employs a xenon-pulsed (10 s half-width, 50 Hz) excitation source. The excitation and emission slits were set at 5 and 10 nm, respectively, throughout. Fig. 1 illustrates the simple optosensing FI manifold used.A conventional Perkin-Elmer L225 1247 flow cell (25 pl volume) was used. Glass-wool was placed at the bottom of the flow cell, to prevent the displacement of particles by the carrier. The CD- immobilized silica gel (20 pl) was loaded with the aid of a syringe and the other end of the flow cell was kept free, The cell was then connected to the flow system. A three-channel peristaltic pump was used to generate flowing streams. Two six-way valves were used for sample introduction (valve A in Fig. 1) and for elution of the complexed molecules (valve B). Preparation of P-CD-immobilized Silica Gels P-Cyclodextrin was immobilized on the silica gel by a method described previously.11 A 6 g amount of silica gel (SG), which had been dried in vacuo at 150 "C for 6 h, was added to 150 ml of 10% solution of (3-aminopropyl)trimethoxysilane in dry toluene.The slurry formed was refluxed under nitrogen for 20 h with continuous Peristaltic Pump Sample I, Excitation I Carrier I immobilized p-CD Fig. 1 Optosensing FI manifold.1120 Analyst, August 1996, Vol. 121 stirring. The amine-modified SG thus obtained was filtered, washed with toluene, acetone and methanol in that order and then dried in vucuo at 100°C for 2 h. A 2 g amount of p-CD was dried in a vacuum desiccator overnight and then treated with 4.0 g of p-toluenesulfonyl chloride in 50 ml of dry pyridine at room temperature for 6 h to tosylate the primary hydroxyl groups. The resulting CD tosylate, after being dried in ~~ac'uo at 50 "C for 2 h, was added to a slurry prepared from 2 g of the amine-modified SG and 100 ml of dry pyridine.This mixture was continuously stirred at 70 "C for 40 h to give amine-type SG with chemically bonded CD. This was filtered, washed with pyridine, acetone and methanol and dried in ~?acuo at 60 "C for 2 h and was then ready for use. General Procedure Samples or standards (2 ml) were injected via valve A and pumped through the flow system. Riboflavin passed through the flow cell where it complexed with (3-CD. The high fluorescence of the complex on the (3-CD-immobilized silica gels was measured at the spectral maxima, A,, = 468 nm, A, = 52 1 nm. Once the fluorescence measurement had been made, 2 ml of methanol-water (60 + 40 v/v) were injected via valve B (to elute the riboflavin complexed with (3-CD), then 2 ml of buffer solution were pumped through the flow cell before proceeding with the next sample injection. For injection in this flow-through sensor, standards and samples were prepared as follows.An appropriate aliquot of the riboflavin standard solution (or the sample) was transferred into a 25 ml calibrated flask, 5 ml of 0.1 mol 1-' NaH2P04-0.1 mol I-' Na2HP04 (pH 6.5) buffer solution were added and the solution was diluted to volume with water. Sample solutions were prepared by directly dissolving Tabellae vitamini B2 and Tabellae vitamini B compositae tablets in an appropriate volume of water without any pre- treatment and diluting to volume. The flow rate used was 2.0 ml min-1 throughout. Reagent blanks were prepared and measured following the same procedure.Results and Discussion Spectral Characteristics The fluorescence spectra of riboflavin under different experi- mental conditions were measured using the flow cell as shown in Fig. 2. It can be seen that the maximum excitation and emission wavelengths of riboflavin in aqueous solution and in (3-CD solution and complexed with (3-CD were all 468 and 521 nm, respectively. The fluorescence of riboflavin in (3-CD solution is only slightly stronger than that in aqueous solution. Complexation with immobilized CD resulted in a stronger fluorescence signal, about 3 4 times greater than that in aqueous solution, owing to the accumulation of riboflavin on the immobilized fl-CD and the change of the environment for the riboflavin. When riboflavin solution passed through the flow cell it complexed with the immobilized (3-CD, rigidity was achieved and water was removed from the environment of riboflavin.Hence a more favourable environment results in a strongly fluorescent complex. Many factors may account for the complexation, including hydrophobicity, dimensions, manner of complexation, structure and characteristics.2 The most important factor is that riboflavin is a very polar molecule. Effect of pH The effect of pH on the fluorescence intensity was studied in the range 5.5-8.0. The maximum fluorescence intensity was obtained at pH 6.5. Therefore, 5.0 ml of 0.1 mol I-' NaH2P04- 0.1 moll-1 Na2HP04 (pH 6.5) buffer solution per 25 ml of final solution was selected as the optimum. Effect of Eluent Elution of the complexed riboflavin from (3-CD immobilized on silica gel was studied by using mixtures of methanol and water with different volume proportions from 10 + 90 to 70 + 30.With increase in the methanol content the volume of eluent needed decreased. At ratios up to 55 + 45, the volume of eluent needed remained constant at about 2 ml. Finally, 2 ml of the eluent of 60 + 40 (v/v) was selected. It was verified that 2 ml of this mixture completely washed the complexed riboflavin out of the (3-CD immobilized on silica gel without influencing the stability of the immobilized p-CD, and allowed the re-use of the latter. Effect of Flow Rate on Sensitivity The effect of the flow variables on the sensitivity and precision of the measurements was assessed by measuring the fluores- cence of riboflavin solution at the same concentration at different flow rates from 0.5 to 2.5 ml min-J.The response curves of riboflavin at different flow rates are shown in Fig. 3. It was found that the sensitivity and precision were improved with an increase in flow rate. The fluorescence signal was slightly affected when the flow rate was up to 2.0 ml min-I. Finally, 2.0 ml min-I was selected. 300 1 1' -1 G "400 450 500 550 600 Wavelengthhm Fig. 2 Fluorescence spectra of riboflavin. I , I', Complexed with fi-CD: 2, 2', in /3-CD solution; and 3, 3', in aqueous solution. Riboflavin concentra- tion: 4 X 10-6 mol 1-1. 100 40 20 0 I I I 0 30 60 90 120 150 180 Time/s Fig. 3 Riboflavin concentration: 2.1 x 10-6 mol I-'. Effect of flow rate: A, 2.5: B, 2.0; C, 1.0; and D, 0.5 ml min-l.Analyst, August 1996, Vol.121 1121 Effect of Sample Volume on Sensitivity One of the main advantages of the sensor is the potential increase in sensitivity with increase in the sample volume taken for analysis. This effect can be assessed by measuring the fluorescence intensity of immobilized (3-CD with different volumes of solutions containing the same concentration of riboflavin passed through the flow cell, Plots of fluorescence intensity versus sample volume showed an increase in fluorescence signal with increase in sample volume, tending asymptotically to a constant fluorescence value above a certain volume (8 ml in this study). Finally, a 2 ml sample volume was selected as a compromise between sensitivity and sample throughput.Response-Time Curve and Reversibility Fig, 4 shows the response curve of riboflavin on the silica gel with immobilized p-CD. It can be seen that the response is a dynamic process. The fluorescence intensity increases with time. The steady-state value (i.e., the highest fluorescence intensity) was reached after about 4 min. The measurement was made at an early stage, before the injected solutions reached constant fluorescence, which shortened the measurement time. Using 2 ml of methanol-water (60 + 40 v/v), the complexed riboflavin with immobilized (3-CD can be eluted quickly and completely without damaging the immobilized CD. The reversibility is excellent, as shown in Fig. 5. It takes about 3 min to complete one measurement, including the time of injection of the eluent and buffer solution.loo0 I A 900 1 600 500 400 300 200 100 0 c / B Time/min Fig. 4 Response curves of riboflavin. Concentration: A, I .32 x 10-5; B, 9.15 X lop6; C, 5.53 X 10-6; D, 1.93 X 10-6; and E, 5.7 X 10-7 mol 1--l. 300 k 250 B 200 c .- LI) != a, a, ._ 2 150 $ 100 3 - U 50 I I I I I A 50 l i 0 150 200 250 300 350 400 Time/s Fig. 5 10-6 moll-'. Elution curves and reversibility. Riboflavin concentration: 5.1 X Stability of the Sensor When at room temperature the sensor was found to be very stable for up to 2 months (the maximum time tested), this being assessed by measuring the solution of riboflavin at the same concentration every 6 days for 2 months. No (3-CD was found to be damaged and the fluorescence did not decrease. Selectivity of the Sensor In the selectivity study, a foreign species was considered not to interfere if it caused an error of less than 5% in the Table 1 Effect of other substances on the determination of 1 pg ml-1 of riboflavin Substance - K' c1- Ca2+ Mg2' Zn2+ Fe3f ~ 1 3 + Ascorbic acid Thiamine Pyridoxal Cyanocobalamin FMN FAD 1,lO-Phenanthroline Concentration/ pg ml-1 5000 5000 50 40 100 50 1 1' 0.5 - 20 500 10 1 1 100 Recovery* 100.0 99.3 99.5 98.7 92.3 99.2 95.0 12.8 43.1 90.5 95.1 95.5 94.9 192.5 173.6 102. I * Mean of three determinations.' 10 pg ml-I 1,lO-phenanthroline added. 250 200 2. LI) a, a, c + .- 150 .- z 100 E h 3 50 20 50 80 110 140 170 200 I I I I I I 850 650 450 250 0 10 20 30 40 [Rib~flavin]lO-~ mol I-' Fig. 6 2 , 2 x 10-6-2 x 10 Calibration graphs for riboflavin: 1, 2 X 10-7-4 X lop6 mol 1-I; moll-].Table 2 Results of determination of riboflavin in pharmaceutical prepara- tions Riboflavin/mg per tablet Sample Stated Found* Tabellae vitamini B2 5.0 4.92 (f0.9%) Tabellae vitamini B compositae 1.5 1.55 (+1.2%) * Mean of three measurements (rt RSD).1122 Analyst, August 1996, Vol. 121 determination of 1 pg ml-1 of riboflavin. The results are summarized in Table 1. Because CD complexation is fairly non- specific for small and relatively hydrophobic organic mole- cules, many other substances including other vitamins can also compete for the binding sites on the immobilized CD and interfere with the determination of riboflavin. However, the interferences from Fellr and other inorganic ions are due to their effect on the emission of riboflavin.As shown in Table 1, the main interferences were produced by Fell', vitamin C, FMN and FAD. It was found, however, that the quenching effect of Fell' could be partly eliminated by masking it with 1,lO-phenan- throline. Response to Riboflavin The analytical figures of merit of the proposed sensor were evaluated. Calibration graphs (shown in Fig. 6) were prepared from the results of triplicate 2 ml injections of the corresponding riboflavin standard solutions. The linear range is 2 X 10-7-2 x mol I-' with a correlation coefficient of 0.9997. The detection limit, defined as three times the standard deviation of the blank, is 9 ng ml- * and the relative standard deviation is 1 % for seven determinations of 0.05 pg ml-1 of riboflavin. Sample Analysis Following the procedure detailed under Experimental, the proposed sensor was applied to the direct determination of riboflavin in Tabellae vitamini B2 and Tabellae vitamini B compositae tablets. The stated composition of the latter was riboflavin 1.5, thiamine 15, pyridoxal 1.5, nicotinamide 22.5 and sodium pantothenate 0.75 mg per tablet.The former was stated to contain 5.0 mg per tablet of riboflavin. The results are given in Table 2. This work was supported by the National Natural Science Foundation of China. References 1 2 3 4 5 6 7 8 9 10 11 Szejtli, J., Cyclodextrins and Their Inclusion Complexes, Akademiai Kiad6, Budapest, 1982. Blyshak, L. A., Dodson, K. Y., Patonay, G., Warner, I.M., and May, W. E., Anal. Chem., 1989, 61, 955. Berzas Nevado, J. J., Rodriguez Flores, J., and Villasenor Llerena, M. J., Fresenius' J . Anal. Chem., 1994, 350(10-ll), 610. Perez-Ruiz, T., Martinez-Lazazo, C., Tomas, V., and Val, O., Analyst, 1994, 119, 1199. Zhebentyaev, A. I., and Duksina, S. G., Farmatsiya (Moscow), 1986, 35(2), 16. Greenway, G. M., and Kometa, N., Analyst, 1994, 119, 929. Barna, E., and Dworschak, E., J . Chromatogr. A, 1994, 668(2), 359. Shao, X. F., and Zhang, Y. H., Guangpuxue Yu Guangpufenxi, 1994, 14(2), 126. Zhao, Y. B., Yu, Z. X., Guo, X. Q., Xu, J. G., and Chen, G. Z., Fenxi Huaxue, 1992, 20(11), 1261. Perez Ruiz, T., Martinez Lozano, C., Sanz, A., and Tomas, V., Analyst, 1994, 119, 1825. Fujimura, K., Ueda, T., and Ando, T., Anal. Chem., 1983, 55, 446. Paper 5108415F Received December 19, I995 Accepted May 8,1995

ISSN:0003-2654    

DOI:10.1039/AN9962101119

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 ( I 123-1126) 1123 Acetylcholinesterase Amperometric Detection System Based on a Cobalt(ri) Tet raphenyl porphyri n-mod if ied Elect rode Qing Deng and Shaojun Dong* Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China An acetylcholinesterase (AChE) activity detection system was fabricated based on the electrocatalysis of cobalt(1r) tetraphenylporphyrin of the electrooxidation of thiocholine chloride, which is the product of the hydrolysis of acetylthiocholine chloride by AChE. A simple modified method was used to form the base electrode. AChE was cross-linked on the base electrode by glutaraldehyde. The optimum working conditions are discussed and the characteristics of the detection system are evaluated.Keywords: Acetylcholinesterase; amperometric detection; tetraphenylporphyrin cobalt(r1)-modified electrode Introduction With the current widespread concern about human health and environmental conditions, more and more researchers are showing great interests in the enzyme acetylcholinesterase (AChE) because of its important role in the metabolism of neurotransmitters and its blockage by some inhibitors (organo- phosphorus and carbamate pesticides). Therefore, it is im- portant to fabricate a system to detect AChE activity. Various transducers have been combined with AChE as a biological recognition element to form this detection system, such as pH electrodes233 and spectroph~tometric,~ fluorimetric,5 voltam- metric637 and piezoelectric* systems.Among these, ampero- metric detection of the acetyl- or butyrylthiocholine hydrolysis process, catalysed by AChE, is the main strategy. The reaction scheme is as follows? AChE CH3COOH + HSCH2CH2N+ (CH3)3C1- CH3COSCH2CHZN' (CH3)2C1- + H20 - The product of the enzyme reaction, thiocholine, can be anodically oxidized on a platinum,1° carbon9,' or mercury electrode? However, the large overvoltage makes them incon- venient for this purpose.12 Similar problems in the detection of other sulfhydryl compounds were overcome by the use of a variety of carbon-based electrodes chemically modified with cobalt phthalocyanine (CoPc),13J4 tetracyano-p-quinodime- thane (TCNQ) and 1, l'-dimethylferrocene.15 Recently, CoPc- 6-19 and TCNQ20 modified carbon-paste electrodes for the detection of thiocholine have been described.The overvoltage can be reduced by 300-400 mV. We found that cobalt(r1) tetraphenylporphyrin (CoTPP) also showed high electrocata- lytic activity towards the oxidation of thiocholine. In this paper, a sensitive and fast-responding AChE activity detection system based on a CoTPP-modified electrode is reported. The optimum working conditions and characteristics of the system were evaluated. * To whom correspondence should be addressed. Experimental Reagents AChE (EC 3.1.1.7, 280 U mg- l, from electric eel), acetylthio- choline (ATCh) chloride and butyrylthiocholine (BTCh) chlo- ride were purchased from Sigma (St. Louis, MO, USA). CoTPP was synthesized according to the literature.21 BSA was the product of Shanghai Institute of Biochemistry, Chinese Acad- emy of Sciences (Shanghai, China). All other reagents were of analytical-reagent grade.Instrumentation and Procedures Chronoamperometric measurements were made with a BAS 100 B/w electrochemical analyser (Bioanalytical Systems, West Lafayette, IN, USA). The cyclic voltametric curves were obtained using an EG&G Princeton Applied Research (Prince- ton, NJ, USA) Model 370 electrochemical system in conjunc- tion with a Gould (Shenyang, Liaoning, China) Model 6000 x-y recorder. All studies were carried out in a conventional three-electrode voltammetric cell containing S ml of 0.05 mol 1-l phosphate buffer (pH 7.8) and magnetically stirred at 300 rpm. The enzyme working electrode was a CoTPP-modified glassy carbon electrode coated with a cross-linked AChE layer.A platinum foil electrode was used as the auxiliary electrode, All potentials presented here are relative to an Ag/AgCl (saturated KC1) reference electrode. The potential of the working electrode was set at +0.25 V. Preparation of AChE Electrode A glassy carbon electrode was polished with a-A1203, rinsed with distilled water and dried at room temperature; this electrode was called GCo. A 10 pl volume of a 10-3 mol 1-1 solution of CoTPP in dimethylsulfoxide was dropped on the surface of GCo. After 15 min, the electrode was cleaned with distilled water and dried at room temperature. This electrode was called CoTPP/GC. A 0.5 mg amount of AChE and 2 mg of BSA were dissolved in SO pl of phosphate buffer (pH 7.5) and mixed thoroughly, then 15 p1 of 2.5% glutaraldehyde were added to the solution and mixed.A 5 pl volume of the enzyme solution was dropped on the CoTPP/GC surface. The enzyme electrode was kept in a refrigerator(4"C) overnight; this electrode was called AChE- CoTPP/GC. If the enzyme solution was dropped on the surface of a bare GCo, then the enzyme electrode was called AChE/GC. All the enzyme electrodes were stored in phosphate buffer solution at 4 "C when not in use. Results and Discussion Electrocatalysis of thiocholine by CoTPP The hydrolysis of ATCh or BTCh chloride or iodide can be catalysed by AChE. The products, thiocholine chloride or1124 Analyst, August 1996, Vol. 121 iodide, are electroactive. Their oxidation current can be recorded at a potential of +0.7 to +1.0 V on a glassy carbon electrode.Obviously, this potential is too high to be interfered with for detection. Moreover, iodide ions would also be oxidized at this potential. Hence ATCh or BTCh iodide is not suitable as a substrate. We found that CoTPP could be used as a good catalyst for the oxidation of thiocholine. Fig. 1 shows the cyclic voltammetrics (CV) curves for the AChE-CoTPPIGC electrode in a solution containing 0 or 0.8 mmol 1-1 ATCh. When the potential scan was initiated from 0 V to a switching potential of +0.7 V, no peak current was observed at any time in the solution without ATCh. When the enzyme electrode was dipped into the 0.8 mmol 1-l ATCh solution for 10 min, an obviously catalytic current was obtained. The peak potential (E,) was +0.25 V.This Ep value was constant although the peak current was changed with different scan rates and ATCh concentrations. Fig. 2 shows the hydrodynamic voltammograms for the GCo, CoTPP/GC, AChE/GC and AChE-CoTPP/GC electrodes. The responses observed at the GCo and CoTPP/GC electrodes to ATCh were equal to the baseline currents where ATCh was electroinactive (curves A and B). When there was only AChE and no CoTPP on the electrode surface, the enzyme catalysed the hydrolysis of ATCh, and the product can be detected at +0.7 V although the peak current is obtained at +0.9 V from the CV curve (not shown). The voltammogram for the AChE-CoTPP/ GC electrode exhibits a peak rather than a normal plateau appearance (curve D). This effect has been noted in numerous earlier studies involving CoPc catalysis.22J3 The reason for this was the irreversible passivation of the electrocatalyst at potentials high enough to oxidize the phthalocyanine ring.Here a similar effect, was observed although the potential of irreversible oxidation of the TPP ring is more positive than that of the Pc ring. The presence of chloride ions in the solution from the substrate decreases the overpotential of TPP ring oxidation by a chloride-TPP binding reaction.24 This effect can be seen more clearly from the CV curves. The more positive the switching I " ' " " 0.0 0.1 0.2 0.3 0.4 0.5 0 6 0.7 € 1 V vs. AgIAgCI Fig. 1 Cyclic voltammograms of the AChE-CoTPP/GC electrode in solution (pH 7.8) in the absence (dashed line) and presence (solid line) of 0.8 mmol 1-I ATCh.Scan rate, 50 tnV s-1. potential, the more the catalytic current decreases in the second scan. The electrocatalysis mechanism of thiocholine oxidation using a CoPc-modified carbon electrode was assumed to involve electrooxidation of Co' to Co" at potentials around +0.3 V. Co' was produced by chemical reaction between thiocholine and CoPC. It seems that the same reaction mechanism is also valid for CoTPP. However, a higher response can be seen at potentials more positive than +0.6 V in curve D. This means that the thiocholine produced in the enzyme reaction on an AChE- CoTPP/ GC electrode also penetrated to the glassy carbon surface and was oxidized there. For analytical purposes, we are only interested in the catalytic current, hence +0.25 V was selected for further amperometric detection using AChE-CoTPP/GC. Moreover, a lower potential is helpful in obtaining higher selectivity.Optimization of Experimental Conditions The effect of pH was studied between pH 5 and 10 in phosphate buffer solution. The response increased dramatically from pH 5 to 8. The catalytic current at pH 10 was nearly the same as that at pH 8. This is in agreement with the literature's with CoPc as a catalyst. For optimum sensitivity, pH 8.5 should be chosen. Taking account of future use as a biosensor to detect pesticides, pH 7.8 was selected because most organophosphorus pesticides will be degraded more easily in alkali solution. Studies of the influence of temperature on the bioelectro- catalytic process were performed. The results show that with increase in temperature from 25 to 50 "C, the steady-state currents increase.Above 55 OC, the responses decrease dramatically. This is a typical result for an enzyme reaction. The highest response was obtained at 54 "C. However, this temperature is so high that it affects the lifetime of the enzyme, and therefore room temperature (27 "C) was conveniently adopted. The apparent activation energy of the rate-controlled process was calculated according to 25 In I = C - EIRT 0.8 L 0.6 < 3 0.4 -.. - 0.2 0.0 -01 00 01 0 2 03 0 4 0 5 0 6 0.7 0.8 E / V vs. AgIAgCI Fig. 2 Hydrodynamic voltammograms for the different electrodes to 0.1 mmol 1-' ATCh. Electrode: A, GC,; B, CoTTP/GC; C, AChE/GC; D, AChE-CoTPP/GC,Analyst, August 1996, Vol. 121 1125 1.0 0.8 0.6 0.4 0.2 0.0 where I is the steady-state current, C is a constant independent of the temperature, R is the universal gas constant, and T is the absolute temperature and E is the apparent activation energy with a calculated value of 43.3 kJ mol-l.The temperature coefficient of the current was obtained as 3.1% OC-'. - - - - - - Response of Detection System The effect of substrate concentration on the signal was obtained by performing fixed injections of standard ATCh or BTCh solution into a stirred solution of the supporting electrolyte in the voltammetric cell. Fig. 3 illustrates the resulting chrono- amperometric curve for ATCh with an increase of 0.027 mmol 1-I in every step. It is obvious that the steady-state current was reached within 15 s between steps. Fig.4 shows the calibration curves for the AChE-CoTPP/GC electrode with ATCh and BTCh. Obviously, both ATCh and BTCh can be hydrolysed by AChE, but the sensitivity of the enzyme electrode to ATCh is 1.26 yA 1 mmol-1, which is nearly ten times higher than that to BTCh. In the concentration range 5 X 10-6-8.0 X mol ]-I, the steady-state current increases linearly with increase in ATCh concentration. The linear range of the AChE-CoTPP/GC electrode to ATCh is slightly wider than that to BTCh. The detection limit for ATCh is 0.8 ymol 1-I whereas that for BTCh is only 8 ymol 1-l. Fig. 5 shows the Eadie-Hofstee plots of the calibration curves. The plot of I versus I/c for ATCh has a break point [Fig. 5(a)] which indicates that the rate is controlled both by diffusion of the substrate and the kinetics of the enzyme.The vertical O.Oo0 - -0.250 -0.5001 \ -0.750j -1.000 i I ::::,:I -2.250 lj 1 -2.500 -1 0.00 100 200 300 400 500 600 Time/s Fig. 3 Chronoamperometric curve for the AChE-CoTPP/GC electrode under optimum conditions. The concentration of ATCh was increased by 0.027 mmol I-' in each step. Potential, +0.25 V; temperature, 27 "C; solution, 0.05 mol I-' phosphate buffer (pH 7.8). 0.4 1 / t d portion of Fig. 5(a) is attributed to diffusion limitation, that is, at low ATCh concentration the diffusion of ATCh through the enzyme film is the rate-limiting process and at high ATCh concentration the enzyme reaction becomes rate-limiting, where the substrate concentration at the enzyme is higher. The apparent Michaelis-Menten constant is calculated as 5.22 mmol 1- according to the non-linear regression curve-fitting procedure. However, the response to BTCh is totally controlled by enzyme kinetics because the affinity of AChE to BTCh is not as high as it is to ATCh.The substrate concentration at the electrode surface is always sufficient to saturate the enzyme. The apparent Michaelis-Menten constant is 1.19 mmol 1-1 from the slope of the plot [Fig. 5(6)] and was verified by the non-linear regression curve-fitting procedure. This is in ac- cordance with data in the literature. l 7 The repeatability and reproducibility of the sensor were also determined. The repeatability of one electrode to determine 0.2 mmol l-1 ATCh was fairly good. The RSD was less than 2%. In contrast, the reproducibility of seven electrodes was not as good as the repeatability of one electrode from the calculated RSD value (less than 15%).Therefore, it is necessary to calibrate each electrode before detection using standard solutions. The enzyme electrode can be stored for 3 months. In order to check the storage stability, the enzyme electrode was used at regular intervals under the same conditions. In the first 15 d, the decrease in enzyme activity was only 5%. The response remained at 85% after 1 month and the activity of the enzyme decreased to 60% after 60 d. It still can be used after calibration. I I I 1.4 1.0 1.1 1.2 1.3 ?i k 0.20 r 0.15 0.10 0.05 0.00 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 c/mmol Fig. 4 Effect of ATCh and BTCh concentration on signals of the AChE- CoTPP/GC electrode.Experimental conditions as in Fig. 3. 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 //c/pA dm3 mmol-' Eadie-Hofstee plots for the AChE-CoTPP/GC electrode. ( a ) ATCh Fig. 5 and (b) BTCH. Experimental data from Fig. 4.1126 Analyst, August 1996, Vol. 121 The financial support of the National Natural Science Founda- tion of China is gratefully acknowledged. References 1 2 3 4 5 6 7 8 9 10 Tran-Minh, C., Ion-sel. Electrodes Rev., 1985, 7, 41. Tran-Minh, C., Pandey, P. C., and Kumaran, S., Biosens. Bio- electron., 1990, 5 , 461. Kumaran, S., Meir, H., Danna, A. M., and Tran-Minh, C., Anal. Chem., 1991,63, 1914. Leon-Gonzalves, M. F., and Townshend, A., Anal. Chim. Acta, 1991, 236, 267. Guibault, G. G., and Kramer, D. N., Anal. Chem., 1965, 37, 1675.Kumaran, S., and Tran-Minh, C., Electroanalysis (N.Y.), 1992, 4, 949. Medyantseva, E. P., Buduikov, G. K., and Babkina, S . S., Zh. Anal, Khim., 1990, 45, 1386. Guibault, G. G., and Ngwainbi, J. N., NATO ASZ Ser. C , 1989, 226, 187. Stoytcheva, M., Electroanalysis (N.Y.), 1995, 7, 560. Gruss, R., Scheller, F., Shao, M. J., and Liu, C. C., Anal. Lett., 1989, 22, 1159. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ~ Martorell, D., Cespedes, F., Matrinez-Fabregas, E., and Alegret, S., Anal. Chim. Acta, 1994, 290, 343. Halbert, M. K., and Baldwin, R. P., Anal. Chem., 1985, 57, 591. Qi, X., Baldwin, R., Li, H., and Guarr, T. F., Electroanalysis (N.Y.), 1991, 3, 119. Wring, S., Hart, J. P., and Birch, B. J., Analyst, 1989, 114, 1563. Kulys, J., and Prungiliene, A., Anal. Chim. Acta, 1990, 243, 287. Skladal, P., Anal. Chim. Acta, 1992, 269, 281. Skladal, P., and Mascini, M., Biosens. Bioelectron., 1992, 7, 335. Hart, J. P., and Hartley, I. C., Analyst, 1994, 119, 259. Hartley, I. C., and Hart, J. P., Anal. Proc., 1994, 31, 333. Kulys, J., and D’Costa, E. J., Biosens. Bioelectron., 1991, 6, 109. Lin, X. Q., and Kadish, K. M., Anal. Chem., 1985, 57, 1498. Santos, L. M., and Baldwin, R. P., Anal. Chem., 1987, 59, 1766. Halbert, M. K., and Baldwin, R. P., Anal. Chem., 1985, 57, 591. Kadish, K. M., Lin, X. Q., and Han, B. C., Znorg. Chem., 1987, 26, 4161. Palmer, 1981, p. T., Understanding 104. Enzymes, Ellis Horwood, Chichester, Paper 6101564F Received March 5,1996 Accepted April 24, 1996

ISSN:0003-2654    

DOI:10.1039/AN9962101123

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1127-1131) 1127 Adsorption of Trace Metals From Sea-water Onto Solid Surfaces: Analysis by Anodic Stripping Voltammetry Vlado CuculiCa and Marko Branicaath a Centre for Marine Research-Zagreb, Ruder BoSkoviC Institute, POB 101 6 , 10000 Zagreb, Croatia 52425 Jiilich, Germany Institute for Applied Physical Chemistry (IPC), Research Centre Jiilich (KFA), The rates and equilibria of the adsorption of dissolved Cd", Pb" and Cu" (at concentration levels of 2 X 10-8 and 8 X moll-l) from sea-water and from a 0.55 mol 1-1 NaCl model solution onto electrochemical glass and quartz cells and Nalgene [fluorinated ethylene-poly(propy1ene)l sample bottles with and without added glass beads at pH 6.2 and 8.1 were measured by differential-pulse anodic stripping voltammetry.Lead(I1) shows higher adsorption than Cu", whereas no adsorption of Cd" is observed. Nalgene is the most suitable material for samplers and storage bottles, whereas quartz is the best material for the electroanalytical vessels. The maximum surface covering concentrations of Pb", (r,) (from sea-water at pH 8.1) on the surfaces of a quartz cell, a glass cell, a Nalgene bottle and glass beads were found to be 2.0 X 10-l1, 3.1 X 10-l1, 2.0 X X 10-11 mol cm-2, respectively. The maximum Pb capacities of the glass and quartz cells with the electrode assembly were calculated to be 2.3 XlO-9 and 1.5 X 10-9 mol, respectively. A procedure is proposed for the measurement of the trace metal capacity of the cell and the electrode assembly used in the experiments, for the determination of the metal concentration in natural samples.Keywords: Adsorption; isotherm; cadmium(ii); lead(i1); copper(ir); sea-water; glass; quartz; fluorinated ethylene-poly(propy1ene) and 2.6 Introduction Various electroanalytical techniques, mainly differential-pulse anodic stripping voltammetry (DPASV), 1-5 have been used for the determination of trace metals at ppb levels in model solutions or in natural water samples. However, an important problem occurs in the actual electrochemical determination of dissolved metals at low natural levels, namely, their adsorption onto sampling, storing and measuring vessel surfaces results in experimental errors that are not negligible.6.7 The adsorption has been studied by several workers in order to explain the adsorption processes of trace metals on electrochemical cell materials.Some workers also included the particulate matter.*-l3 In order to understand these processes under natural sea-water conditions it is necessary to study model solutions with higher metal concentration levels (> 10-8 mol 1-I). However, in most papers, the experiments were performed in model solutions or, if performed in natural sea- water samples, at significantly higher metal concentrations (10-7-10-5 moll-I). The aim of this research is primarily oriented towards natural sea-water samples and a study of the adsorption of dissolved Cd", Pb" and Cu" ions at low concentration levels onto quartz and glass cells, and Nalgene bottles with and without added glass beads. The adsorption isotherms of Pb" for different adsorbing surfaces are also presented.These experiments are a step towards finding the most convenient vessel material that will minimize loss of the dissolved trace metals from the natural water sample. A procedure for the characterization of the cell and electrode assembly in terms of cell Pb capacity is also given. Experimental Instrumentation and Reagents All voltammetric measurements were performed with a PAR 174 polarographic analyser, (Princeton Applied Research, Princeton, NJ, USA), connected to a Hewlett-Packard (Avon- dale, PA, USA) 704514 X-Y recorder. pH was measured with an Orion Research (Cambridge, MA, USA) pH-meter. A constant temperature of 25 "C was maintained during the experiments by a HAAKE D8 thermostat (HAAKE Mess-Technik, Karlsruhe, Germany).The electrochemical vessels were quartz and glass cells (both 60 ml, adsorbing area about 50 cm2) with a corresponding universal cap (Metrohm, Herisau, Switzerland, No. 6.1414.0 10). The adsorption experiments were performed in a 1 1 Nalgene bottle [fluorinated ethylene-poly(propy1ene)l with an adsorbing area of approximately 580 cm2. Glass beads were used as an additional adsorptive surface (area of one bead about 0.24 cm2; diameter 0.3 cm). The working electrode was a hanging mercury drop electrode (HMDE) (Metrohm, No. 6.0335 .OOO), the reference electrode was an Ag/AgCl electrode (saturated NaC1) and platinum wire served as the counter electrode. Stirring of the solution was performed with a 'turbo' stirrer constructed in this laboratory, with a constant speed of 2000 rev min-1.The capillary of the working electrode, together with the reference and counter electrodes, the stirrer and the N2 tube comprised the electrode assembly with a surface area of 24 cm2. Pure N2 was saturated with C02 by passing it through wash bottles [the first contained a buffered solution of MgC03(s), Na2B4O7.10H2O (each 10.4 g 1-l) and 0.1 mol 1-1 KH2P04 and the second contained natural sea-water] maintain- ing the natural pH of sea-water (8.1) in the reaction vessels. The pre-electrolysis was performed for 5 min with stirring and for 15 s without stirring. The pre-electrolysis potentials were -0.8 V for the adsorption experiments and -0.6 V for the isothermal measurements; the pulse amplitude was 50 mV, the scan rate 10 mV s-l and the drop time 0.5 s.All solutions were prepared from analytical-reagent grade chemicals. Stock solutions of Cd", Pb" and Cu" were prepared by dissolution of an appropriate amount of their nitrate salts (all from Merck, Darmstadt, Germany) in doubly distilleG water. The sea-water sample was taken from Jadrija (near Sibenik, Croatia), with a pH of 8.1 and a salinity, S, of 38%~~.1128 Analyst, August 1996, Vol. 121 0.4 pH=6.2 0.2 1 5 0 . 0 Procedure Adsorption measurements Prior to the experiments, the experimental vessels (quartz cell, Nalgene bottle) and glass beads were thoroughly washed. The quartz cell was washed with chromic acid, then with 10% nitric acid and finally with doubly distilled water; the Nalgene bottle was washed with 50% nitric acid and then with doubly distilled water.Glass beads were boiled for 1 h in concentrated nitric acid, then for 1 h in concentrated hydrochloric acid and finally washed with doubly distilled water.'" All measurements were performed at pH 8.1 and 6.2, achieved by the addition of 1 X lo-' moll-' nitric acid, whereas in 0.55 moll-I NaCl solution pH values were regulated by the addition of borate buffer. Adsorption onto the quartz and glass cells and the electrode assembly was studied for natural sea-water and NaCl solution. A mixture of 2 X mol I-' each of Cd", Pb" and Cu" was added to 40 ml of electrolyte in the quartz cell. Measurements in theglass cell were performed with only natural sea-water as the electrolyte. For the first seven values, the anodic stripping voltammetric peak current was recorded every 8 min and subsequently every 30 min until the adsorption equilibrium was achieved.Adsorption onto the surface of the Nalgene bottle was measured for 500 ml of sea-water to which 8 X 10-8 moll-1 of each dissolved metal ion had been added. The bottle was rotated in order to mix the solution. At the same time, in a quartz cell, 30 ml of sea-water (at pH 2.5) were prepared, and a 10 ml aliquot of the solution from the bottle was then added to the quartz cell. The anodic peak current was recorded, representing the concentration of dissolved metal at zero time, t = 0 min. The electrochemical cell was then washed and the whole procedure was repeated with a further aliquot from the adsorption vessel until adsorption equilibrium was achieved.The same experimental procedure was also performed with the addition of glass beads at two levels with approximately (i) 3050 beads (total area 725 cm2) and (ii) 9150 beads (total area 2175 cm2). - - pH=6.2 " " " " ' Adsorpion isotherm of the Nalgene bottle The procedure for adsorption onto the surface of the Nalgene bottle and glass beads excludes adsorption on the cell. This was achieved by keeping the cell supporting electrolyte at pH 2.5 where no adsorption was observed. The construction of the isotherms was performed by using the following equations: 12 0.4 0.2 pH=8.1 I l l Adsorption isotherms of Pb'I The Langmuir isotherms were constructed from the results obtained in the glass cell, quartz cell, Nalgene bottle and Nalgene bottle with glass beads of adsorptive area 725 cm2 for natural sea-water (at pH 8.1; S = 38%0).The volume of the electrolyte in the cells was 60 ml and in the Nalgene bottle 500 ml. Prior to all experiments, the entire equipment was thoroughly washed in order to remove any residual trace metals. 14 - pH=8.1 ~ Adsorption isotherms of the electroanalytical cells Lead(I1) was added to 60 ml of sea-water in the range from 6 X 10-9 to 2 X moll-1. The first measured peak current, I(,)", corresponds to the C(l)o, i.e., 6 X 10-9 moll-' of Pb2+ added. After 3 h (the equilibration time determined from the adsorption procedure described above), adsorption equilibrium occurs, and the peak current i( 1 )33, represents the equilibrium concentration of Pb2+ in the solution. Z(l)o - I(,)..is the adsorbed Pb2+ value, C(l)ads. The numerical value of a further added amount of Pb2+ was summarized with the previous numerical vlue of C(lIm measured in solution. The peak current obtained is equivalent to the new concentration of Pb2+ in the solution, i(2)o. The addition of Pb" to the solution up to a concentration of 2 X 10-7 moll-l was repeated every 3 h until sufficient points for the construction of the Langmuir isotherm had been obtained. c(i)o - c ( i ) m = C(i)ads (2) where r(,) are the surface covering concentrations, C(i)o are the added concentrations of Pb2+, C(i).. are the equilibrium concentrations of Pb2+ in the solution, A is the area of the electroanalytical vessel surface (cm2) and V is the volume of the electrolyte (1). The abscissa represents the equilibrium concen- trations achieved after 3 h of equilibration, [C(i,,/mOl l-l] and the ordinate represents the surface covering concentrations (Tho1 cm-2).Results and Discussion Adsorption of Cd", Pb" and Cu" Onto the Quartz Cell and Electrode Assembly The adsorption of Cd", Pb" and Cu" onto the quartz cell and electrode assembly from natural sea-water and the model solution of 0.55 rnol 1-1 NaCl at two pH values (8.1 and 6.2) was measured. The time dependence of the concentration of dissolved Pb" in the electrolyte is shown in Fig. I (A). The ratio ii : il is the anodic peak current normalized to the first measured value ( i l , where t = 0). Adsorption equilibrium at pH 8.1 was reached after approximately 3 h from sea-water and after only 45 min from the NaCl solution.At equilibrium, 48% of Pbrl was adsorbed from sea-water and about 55% from the NaCl electrolyte, whereas at pH 6.2 41% of Pb" was adsorbed from sea-water and 35% from the NaCl solution. The stronger adsorption of trace metals at higher pH values is in good agreement with the results of previous studies.6-9.13 A longer period is necessary for establishment of the adsorption equilib- 1 .o 0.8 0.6 0.4 0.2 a- 0.0 1 .o 0.8 0.6 0.4 0.2 0.0 . \- Fig. 1 [Pb*+]=2~10-~ mol 1-l [Cu2+]=2x1 o-* mol I-' (A) seawater (S=38%0) (B) . seawater (S=38%0) 9 0.55 mol I-'NaCl = 0.55 mol I-'NaCI current on the time of adsorption in the quartz cell at pH values b.2 and 8.1. Initial concentrations of Cd", Pb" and Cu" were 2 X lo-* mol 1-1. S = 38%0.Analyst, August 1996, Vol.121 1129 riurn in natural sea-water, probably because of its complex composition and also its higher ionic strength6 ( = 0.7 moll-1). Competition between ions naturally present in sea-water (such as Mg2+ and Ca2+) and the added metals also affects the equilibration time of the adsorption. The ionic strength of sea- water is higher than that of the 0.55 mol 1-1 NaCl solution; hence, the concentrations of Na+ and K+ in sea-water are much greater than those of the trace metals, and therefore these ions are able to compete efficiently for active surface sites.6 For 0.55 moll-' NaCl, the adsorption curve shows a more regular shape than that for natural sea-water which is in the form of a 'broken' line. These observations are also applicable to the complex competition with sea-water constituents.The diffusion coefficient, D , is also a parameter that influences the time at which adsorption equilibrium is achieved, as well as the shape of the adsorption curves. It has been shown that diffusion affects the adsorption phenomena.7 A difference between the diffusion coefficients in the natural sea-water and in the-model solution is evident, and is caused by the trace metal speciation in the two systems.15 The mobility of Pb" species is lower in natural sea-water than in 0.55 moll-' NaCl, where Pb" is mainly present as the Pb2+ ion and PbCl,, whereas in sea- water it also occurs as an organic and/or inorganic complex.'6,'7 This fact probably affects the over-all adsorption processes. The amount of adsorbed Pb'I is similar for both electrolytes (natural sea-water and the model solution of 0.55 moll-' NaC1) and at both pH values (Table 1).This result indicates that the cell and electrode assembly adsorbing capacity for Pb" from natural sea-water and 0.55 moll-' NaCl solution is similar and could be the subject of further investigations. Fig. 1(B) shows the rate of the Cu" current decrease. The shape of the curves is similar to that of the corresponding curves for Pb. The conclusions concerning Pb" adsorption can be partially used to explain Cull adsorption. The amount of metal adsorbed is different. The loss of dissolved CuT1 is lower than for Pb". At pH 8.1, the losses of Cull are 25 and 23% for sea-water and 0.55 moll- NaCl solution, respectively. At pH 6.2 the loss of Cu" is 15% for both electrolytes (Table 1).The very similar extent of adsorption indicates that Cu" is present in the same ionic form in both the 0.55 mol 1-1 NaCl solution and natural sea-water. At the lower pH value (6.2), the shape of the curve is the same for the NaCl solution and sea-water. The lower adsorption of Cu" species in comparism with Pb" could be explained by a smaller extent of hydrolysis.",I2,'~20 The major inorganic Cu" species in natural sea-water at 25 "C and salinity S = 35%0 are: CuC03 (73.8%), Cu(CO&- (14.2%), &OH* (4.9%) and Cu2+ (3.9%).17 The lower amount of Cu adsorbed in Table 1 Percentage of Pb", Cull and Cd" adsorbed onto different adsorbing vessels. Electrolyte, natural sea-water (T = 25 "C; S = 38%0). Initial metal concentration in quartz cell, 2 X 10-8 moll-1; initial metal concentration in Nalgene bottles, 8 X 10-8 mol 1-1 Pb" cu" Cd" Adsorbing PH PH PH PH PH PH Quartz cell + 41.0 48.0 15.0 25.0 - - Glass cell + Nalgene bottle 1.1 1.5 1.5 1.6 < 1 - Nalgene Nalgene vessel 6.2 8.1 6.2 8.1 6.2 8.1 assembly 35.0* 5S.O* 15.0* 23.0* - assembly - 52 - so - bottle +1' 14.0 52.0 29.0 42.5 < 1 < 3 bottle +2+ 21.0 88.0 38.0 52.0 < 1 <4 * Electrolyte, 0.55 mol 1-' NaCl.+ + I = 725 cm* and +2 = 2175 cm* area of added glass beads. comparison with Pb cannot be ascribed to the competition between Cd, Cu and Pb species for the free sites on the adsorptive surface. This is illustrated in Fig. 2(A) and (B). The amount of Cu adsorbed is virtually the same with or without Cd and Pb species in the measuring system.The same situation was observed for the Pb adsorption measurements in the quartz cell at pH 8.1. It should be noted that after the addition of Cu" to sea-water, Cu" reacts with the organic compounds present21 and/or adsorbs onto the cell walls and electrode assembly. From our results we cannot distinguish between these two cases. The results for Cd" show that there is virtually no adsorption from sea-water and 0.55 mol 1-1 NaCl at both pH values (8.1 and 6.2) in comparison with Pb" and CuT1 ions. This can be explained by the fact that the main Cd" species in sea-water and the NaCl model solution are dissolved Cd"-chloro com- ple~es,2~.23 and, according to other workers, Cd-chloro com- plexes do not adsorb in contrast to free Cd ions.24,25 Adsorption of Cd", Pb" and Cu" onto the Glass Cell and Electrode Assembly The adsorption of Cd", Pb" and Cull onto the glass cell and electrode assembly from natural sea-water at pH 8.1 was measured. Adsorption equilibrium was reached after 90 and 45 min for Pb and Cu, respectively. At equilibrium, about 52% of Pb" and 50% of Cu" were adsorbed (Fig. 3).No adsorption of Cd2+ was observed. A similar situation occurred in the glass cell as in the quartz cell when the metal examined was with the other (A) Co(Pb2') = 2 ~ 1 0 ~ rnol I-' 1.0 It ;,Pb,wit;C;andCu , I , , + only Pb Umin 0 0 20 40 60 80 100120140160180200220 (B) CO(Cu") = 2 ~ 1 0 ~ rnol I-' 1 .o 0.8 0.6 -e- Cu with Cd and Pb -t only Cu 0.2 0'4 t t 0 ' ' ' I ! ' I " I " 0 20 40 60 80 100120140160180200220 Umin Fig.2 Dependence of Pb" (A) (with and without Cd" and Cu") and Cu" (B) (with and without Cd" and Pb"), normalized DPASV peak current on the time of adsorption in the quartz cell at pH value 8.1. Initial concentrations of Cd", Pb" and Cu" were 2 X 10-8 mol 1-1. S = 38%0.1130 Analyst, August 1996, Vol. 121 _____ two metals or present alone. This fact probably does not affect the adsorption process as shown previously in Fig. 2. The amount of Pb" adsorbed onto the glass cell and electrode assembly was calculated from the experimentally determined adsorption isotherm data for the glass cell using eqns. (1) and (2) and was similar to the experimentally measured values. The data obtained represent the adsorption equilibrium at pH 8.1. It was calculated that 3.9 X lo-* mol 1-1 of Pb" is adsorbed, which is about 52% of the initial concentration of the added Pb concentration in the glass cell. This result agrees with the results obtained in the quartz cell (48%) and in the Nalgene bottle with and without glass beads.This confirms the strong adsorption ability of glass and quartz for the Pb" in sea-water. Adsorption of Cd", Pb" and Cu" Onto the Nalgene Bottle The adsorption of trace metals from natural sea-water at pH 8.1 and 6.2 onto the Nalgene bottle and with added glass beads was measured. Table 1 shows the results obtained. A total of 1.5% of Pb ions was adsorbed at pH 8.1. Lower values of adsorption onto the Nalgene bottle than onto the cells and electrode assembly were obtained. The Nalgene material adsorbs smaller amounts of trace metals than the quartz cell.Obviously, Nalgene is a much more suitable material for sampling, storing and measuring trace metals in sea-water. The adsorption of Pb'I and CU" is more pronounced (Table 1) on increasing the amount of glass beads (Fig. 4). This can be explained by the strong adsorptive ability of the glass surface.7-9 1 .o 0.8 0.6 . -- a 0.4 0.2 0 (A) Co(Pb2+) = 2x1 O4 rnol I-' t -8- Pb with Cd and Cu + only Pb 0 15 30 45 60 75 90 105 120 ffmin 1 .o 0.8 0.6 . -- a 0.4 0.2 0 -8- Cu with Cd and Pb + only c u 0 15 30 45 60 75 90 105 120 ffmin Fig. 3 Dependence of Pb" (A) (with and without Cd" and Cu") and Cu" (B) (with and without Cd" and Pb"), normalized DPASV peak current on the time of adsorption in the glass cell at pH value 8.1.Initial concentrations of Cd", Pb" and Cu" were 2 X 10-8 rnol I-l.S = 38%. ~ Adsorption Isotherms of Pb" From the adsorption results presented above, Pb" was chosen as the best model metal ion for the construction of the adsorption isotherms using natural sea-water (S = 38%0, 25 "C and pH = 8.1). The procedure for the construction of the isotherms for the cells and Nalgene bottle was as described under Experimental. The results obtained virtually follow the Langmuir isotherm (Fig. 5). It is obvious that the electroanalytical glass cell and electrode assembly has the highest maximum surface covering concentration (Fa). The glass cell and electrode assembly give r, = 3.1 X lo-" rnol cm-2. Hence, the electrochemical glass cell is not very suitable for the determination of trace metals in natural sea-water.However, actual measurements at the natural pH values are of importance because of trace metal speciation. Therefore, the measurements should be performed in an electrochemical system made of a material with negligible or, at least, controlled adsorption. For the quartz cell and electrode assembly, r, is 2.0 X 10-11 mol cm-2, which is nearly ten times higher than I?, for the Nalgene bottle, viz., 2.0 X 10-12 mol cm-2. However, it is less than the adsorption onto the glass 0.9 0.8 0.7 0.6 0.5 0.4 0.3 .-- 0.1 0.0 o.2 2- 0.5 0.4 0.3 0.2 0.1 0.0 0.0 --J 0 25 50 75 I00125150 175 200 0 25 50 75 100125150 I75 200 ffmin tlmin Fig. 4 Dependence of the Cd", Pb" and Cu" normalized DPASV peak current on the time of adsorption from natural sea-water in Nalgene bottle with (A) 725 cm2 and (B) 2175 cm2 area of glass beads at pH values 6.2 and 8.1.Initial concentration of trace metals were 8 X 10-8 rnol 1-1. glass cell + electrode assembly nalgene bottle + glass beads (adsorptive area = 725 cm2) quartz cell + electrode assembly / J nalgene bottle - - I I I - - 0,o I I I 1 I I I 1 , I 0 2 4 6 8 10 12 14 16 18 20 22 C,(Pb2+)x106/mol I-' Fig. 5 8.1; temperature, T, 25 "C. Adsorption isotherms of Pb" measured in seawater, S = 38%0 at pHAnalyst, August 1996, Vol. 121 1131 surface or the Nalgene bottle with glass beads, for which r, is 2.6 X 10-11 mol cm-2. Characterization of the Metal Capacity of the Cell and Electrode Assembly The results show that the procedure described here is the most convenient for the characterization of the adsorption of the cell and electrode assembly used in an experiment.Once con- structed, the adsorption isotherm can be used in further experiments by calculating the correct results obtained for the concentration of the dissolved metal ions in the solution. These corrections are particularly important when the metal complex- ing capacity is determined. In these types of measurements, it is important to distinguish the metal capacity of the cell and electrode assembly from the complexing capacity of the natural water sample.26-2* The maximum metal capacity of our glass cell and electrode assembly for Pb" obtained with a 2 x 10-7 mol -1-1 Pb" solution was calculated by multiplying the maximum surface covering concentrations by the area of the cell and of the electrode assembly (50 cm2 + 24 cm') and was found to be 2.3 X 10-9 mol; for the quartz cell and electrode assembly the metal capacity was 1.5 X 10-9 mol.For lower concentrations of the Pb" in the solution, the metal capacity of the cell can be calculated from the isotherm found. From the results of the Pb" adsorption isotherm, a relation- ship between the amount adsorbed and the adsorptive area of the glass cell with the electrode assembly and of a Nalgene bottle with glass beads of area 725 cm2 was calculated. The amount of Pb" adsorbed in a Nalgene bottle with glass beads is 8.3 times higher than in the glass cell with the electrode assembly (1.8 X 10-8 mol1-1/2.3 X 10-9mol1-' = 8.3). The surface area of the glass beads is 9.8 times larger than the area of the glass cell and the electrode assembly (725 cm2/74 cm2 = 9.8). The difference between the area ratio and the amount adsorbed ratio (9.8 and 8.3) is probably due to a strong adsorption on the Plexiglas stirrer in the glass cell.29 The area of the Nalgene bottle can be neglected because the amount adsorbed on the bottle is less than 7% of the amount of Pb adsorbed on the glass beads.Conclusions Fluorinated ethylene-poly(propy1ene) (Nalgene) was found to be the most suitable material for samplers and storage bottles for the determination of trace metals in natural sea-water samples. Quartz was the best material for the electroanalytical vessels (cells). The glass electrode assembly showed a strong ad- sorptive ability, hence, it should be replaced with quartz parts as much as possible.The calculated values of the maximum cell and electrode assembly Pb capacities for glass and quartz are 2.3 X 10-9 and 1.5 X 10-9 mol, respectively. We thank Dr. I. Piieta for very helpful discussions. This work was supported by the Ministry of Science and Technology of the Republic of Croatia and in the framework of the EUREKA- ELAN1 EU-493 (EUROMAR) project. Financial support from the International Bureau of KFA, Jiilich, within the bilateral agreement between Germany and Croatia, is gratefully ac- knowledged. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Imtrumental Methods in Electrochemistry, Southampton Elec- trochemistry Group, Wiley, New York, 1985.Bond, A. M., Modern Polarographic Methods I n Analytical Chem- istry, Marcel Dekker, New York, 1980. Crow, D. R., Polarography of Metal Complexes Academic Press, London, 1969. Marine Electrochemistry, A Practical Introdaction, ed. Whi tfield, M., and Jagner, D., Wiley, London, 198 1. Bard, A. J., and Faulkner, L. R., Electrochemical Methods- Fundamentals and Applications, Wiley, New York, 1980. Davison, W., de Mora, S. J., Harrison, R. M., and Wilson, S., Sci. Total Environ., 1987, 60, 35. Petrie, L. M., and Baier, R. W., Anal. Chim. Acta, 1976, 82, 255. Diaz-Cruz, J. M., Esteban, M., van den Hoop, M. A. G. T., and van Leeuwen, H. P., Anal. Chem., 1992, 64, 1769. Diaz-Cruz, S., Diaz-Cruz, J. M., and Esteban, M., Anal. Chem., 1994, 66, 1548. Kozar, S., Bilinski, H., Branica, M., and Schwuger, M. J., Sci. Toral Environ., 1992, 121, 203. PlavSii, M., Kozar, S., KrznariC, D., Bilinski, H., and Branica, M., Mar. Chem., 1980, 9, 175. Bilinski, H., Kozar, S., and Branica, M., in Colloid andfnterface Sci., ed. Kerker, M., Academic Press, London, 1976, pp. 211-231. Vuceta, J., and Morgan, J. J., Environ. Sci. Terhnol., 1978, 12, 1302. Goldberg, E. D., Koide, M., Bertine, K., Hodge, V., Stallard, M., MartinEiC, D., Mikac, N., Branica, M., and Abaychi, J. K., Appl. Geochem., 1988,3.561. Leeuwen, H. P., Sci. Total Environ., 1987, 60, 45. Riley, J. P., and Chester, R., Introduction to Marine Chemistry, Academic Press, London, 5th edn., 1979. Millero, F. J., and Sohn, M. L., Chemical Oceanography, CRC Press, Boca Raton, FL, Ann Arbor, MI, London, 1992. Sillen, L. G., and Martell, A. E., Stability Constants, Chemical Society, London, 1964. Batley, G. E., and Florence, T. M., J . Electroanal. Chem., 1976, 72, 121. Stumm, W., and Morgan, J. J., Aquatic Chemistry, Wiley, New York, 1970. Florence, T. M., and Batley, G. E., J . Elrctroanal. Chem., 1977, 75, 791. Bar& A., and Branica, M., J . Polarogr. Soc., 1967, 13, 4. Bubii, S., and Branica, M., Thalassia Jugosl., 1973, 9, 47. Millward, G. E., Environ. Technol. Lett., 1980, 1, 394. Benjamin, M. M., and Leckie, J. O., Environ. Sci. Technol., 1982,16, 162. van den Berg, C. M. G., Mar. Chem., 1982, 11, 323. Kramer, C. J. M., Mar. Chem., 3985, 18, 335. Omanovii, D., Piieta, I., Peharec, Z., and Branica, M., Mar. Chem., 1996, in the press. OmanoviC, D., Peharec, i., Magjer, T., LovriC, M., and Branica, M., Elec-troanalysis, 1994, 6, 1029. Paper 6/01 147K Receiwd February 16, 1996 Accepted April 29,1996

ISSN:0003-2654    

DOI:10.1039/AN9962101127

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1133-1136) 1133 Kinetic Fluorimetric Determination of Gliadins in Foods Belen Gala, Agustina Gomez-Hens and D. Perez-Bendito Department o j Analytical Chemistry, Faculty of Sciences, University of Ccirdoba, E-14004, Ccirdoba, Spain Kinetic methodology was applied for the first time to the determination of gliadin proteins by using a stopped-flow mixing technique. The method is based on two simultaneous processes: the reaction between gliadins and sodium dodecyl sulfate and the elimination of the quenching caused by this surfactant of the fluorescence of Cresyl Violet. Thus, the increase in fluorescence intensity with time is directly related to gliadin concentration. The use of this oxazine dye allows dynamic fluorescence measurements at long wavelengths, which avoids potential interferences from the sample matrix. The reaction rate is measured within 5 s, so the method is very suitable for the routine determination of gliadins in food samples.The dynamic range of the calibration graph was 0.5-50 yg ml-1 and the LOD was 0.25 yg ml-l. The RSD was 1.6%. The method was applied to different food samples and the analytical recoveries were 88-107 %. Keywords: Kinetic methods; gliadins; fluorescence; food samples; stopped-flow method Introduction Prolamin fractions of cereals such as wheat, barley and rye cause coeliac disease or gluten-sensitive enteropathy in humans, which gives rise to characteristics changes in the intestial mucosa: loss of villous structure, degenerative changes in the epithelial cells and adverse effects on the nutrient absorption functions.The incidence of this disease is higher in children than in adults and varies from one geographical area to another. The mean frequency in Europe ranges between 0.05% and 0.1 % although it reaches 0.3% in the west of Ireland. The prolamin fraction of wheat, which is a primary food source in humans, is named gliadin and constitutes about one third of the gluten, glutenin being the remaining two thirds. Both fractions can be easily separated on the basis of the solubility of gliadins in 70% ethanol where glutenins are insoluble. 1 The use of electro- phoretic techniques2 has established that a single variety of wheat may have over 40 different gliadin proteins with molecular masses of 30 000-80 000. These proteins have an exceptionally high content of glutamine and proline compared with other proteins.They also contain low levels of aspartate and glutamate so that there is an over-all scarcity of charged amino acids. This explains the very low ionic character of gliadins and their lack of solubility. Gluten characterization is of great interest since variation in structure is associated with wheat quality. Although this characterization is difficult owing to its insolubility in water or salt solutions, heterogeneity and tendency to aggregate, ad- vances in electrophoretic and chromatographic techniques have allowed the separation of gluten protein^.^ Thus, the best separations of gliadin proteins to date have been by two- dimensional electrophoresis by combining isoelectric focusing (IEF) with sodium dodecyl sulfate polyacrylamide gel electro- phoresis (SDS-PAGE).4 While size does not influence mobility in IEF, proteins are turned into random coils and charge differences are eliminated by bound SDS.This allows separa- tions in SDS-PAGE almost totally on the basis of size. Initial studies of the capillary electrophoresis of wheat gliadins have been reported recently.5.6 Reversed-phase LC complements other methods since gliadins are separated by surface hydro- phobicity and not by size or charge.7 In addition to the complexity of the separation, a second also difficult step in gliadin characterization is the interpretation of the electro- phoretic and chromatographic data. Different statistical meth- ods have been applied for this purpose.3 However, from the viewpoint of the clinical sensitivity to gliadins, the characterization of these proteins is not required and the main interest is the availability of methods that allow their simple and sensitive determination in foods, since the treatment of coeliac disease consists in the use of gluten-free diets. For this purpose, several methods based on heterogeneous immunoassay techniques such as radioimmunoassay8 and enzyme immunoassay9-15 have been reported, which show good selectivity but require 2-6 h.This paper describes an inexpensive method for the detenni- nation of gliadins which avoids the use of immunoreagents and is faster than the immunoassay methods reported for these proteins. Gliadin concentration is directly related to the rate of release of Cresyl Violet from its interaction with SDS so that the reaction is followed through the appearance of the fluorescence of this dye (hex 585 nm, he, 627 nm).The high initial rates obtained require the use of a stopped-flow mixing technique, which also allows the rapid acquisition of analytical data, partial automation of the method and its simple application to routine analysis. This is the first time that a kinetic fluorimetric determination of gliadins has been reported. A previous systematic study of the fluorescent behaviour of various oxazine and thiazine dyes, which emit fluorescence in the range near 600 nm, showed that it is considerably affected by the presence in the solution of an anionic surfactant such as SDS:16 whereas the fluorescence is quenched by the presence of monomers of SDS (when the SDS concentration is lower than its c.m.c.), it increases when the solution contains micelles of SDS so that the fluorescence intensity is higher than in the absence of SDS.It was suggested that there is an interaction between the negatively charged SDS monomers and the positively charged azine dye, which affects the electronic distribution of the latter and produces the loss of fluorescence. However, when SDS micelles are formed, they protect the fluorescence of the dye from non-radiative processes, this effect prevailing over the electrostatic effect caused by the surfactant monomers. The study of the effect of different compounds on the fluorescence of these systems showed that the presence of a protein does not affect the fluorescence of the azine dye in the absence of SDS or in the presence of SDS micelles, but avoids the quenching effect of SDS monomers.This is ascribed to the fact that the denaturant effect of SDS on the protein is stronger than its interaction with the azine dye, so that this is released and its native fluorescence appears. By using Cresyl Violet as an azine dye, lysozyme was directly determined in different1134 Analyst, August 1996, Vol. 121 pharmaceutical samples containing other drugs such as antibiot- ics and anaesthetics.16 The presence of potentially interferent fluorescent species does not affect the selectivity of the method as they emit fluorescence at lower wavelengths than the azine dye. Experimental Instrumentation An SLM-Aminco (Urbana, IL, USA) Model SOOOC lumines- cence spectrometer, equipped with a 450 W xenon arc source and an R928 photomultiplier tube, was used.The instrument was furnished with an SLM-Aminco Milliflow stopped-flow module, which was fitted with an observation cell of 0.2 cm pathlength and was controlled by the associated electronics, the computer and a pneumatic syringe drive system. The solutions in the stopped-flow module were kept at a constant temperature of 25 "C by circulating water from a thermostated tank. Reagents All chemicals were of analytical-reagent grade. A stock standard solution (100 pg ml-1) of gliadin (Sigma, St. Louis, MO, USA) was prepared by dissolving the appro- priate amount in 2 ml of 0.1 mol 1-1 SDS solution and diluting to 100 mi with distilled water.Aqueous solutions of Cresyl Violet acetate (Sigma) (9.3 X mol 1-I) and SDS (Merck, Darmstadt, Germany) (0.1 mol l-l), were also prepared. A borate buffer solution (0.2 mol 1-I) was prepared from boric acid and adjusted to pH 10.0 with sodium hydroxide solution. Procedure One of the two 5 ml reservoir syringes of the stopped-flow module was filled with a previously prepared solution contain- ing gliadins at a final concentration of 0.5-50 pg ml-1, borate buffer (1.75 X 10-2 rnol 1-1) and SDS (1.8 X 10-3 mol 1-l). The other syringe was filled with a solution containing Cresyl Violet acetate (2 X 10-5 mol 1-I), borate buffer (1.75 X 10-2 mol 1-1) and SDS (1.8 X 10-3 mol 1-1). After the two 2 ml drive syringes had been filled, 0.04 ml of each solution was mixed at a flow rate of 20 ml s- in the mixing chamber in each run.The variation of the fluorescence intensity with time throughout the reaction was monitored at he, 585 nm and hem 627 nm for 10 s using the slow kinetic acquisition mode of the instrument. All measurements were carried out at 25 "C. The data were processed by the computer, furnished with a linear regression program for application of the initial-rate method. The reaction rate was determined in about 5 s and each standard or sample was assayed in triplicate. The blank signal was negligible. Deternzination of gliadin in food samples Each sample (0.1 g) was extracted sequentially and twice with water (30 ml) and 0.4 moll-' sodium-chloride solution (30 ml) to remove water-soluble albumins and salt-soluble globulins, respectively.Each extraction was carried out for 10 min at room temperature and was followed by centrifugation. The residue was extracted twice for 10 min at room temperature with a 70% aqueous ethanol solution (30 ml) to separate the gliadin fraction from the glutenin fraction, which remains in the residue. After centrifugation, the supernatant liquid was evaporated by heating gently. The residue was dissolved in 0.1 moll- 1 SDS (1 ml) and diluted to 50 ml with distilled water. Finally, a volume of this solution was treated as described above. The standard additions method was used for each analysis. Results and Discussion As indicated above, the control of the presence of gliadins in gluten-free foods is of great interest for patients with coeliac disease.Taking into account the lack of simple, inexpensive and automatic methods for gliadin determination, the usefulness of a new approach involving a stopped-flow mixing technique and long-wavelength fluorescence measurements was investigated. Several oxazine and thiazine dyes were tried as reagents for this purpose: Cresyl Violet (Aex 585 nm, A, 627 nm), Toluidine Blue (A,, 623 nm, hem 664 nm), Azure A (Aex 622 nm, A,, 673 nm), Azure B (Aex 647 nm, A, 673 nm) and Nile Blue (Aex 624 nm, A, 675 nm). The fluorescence of all these dyes is quenched by the presence of SDS monomers. However, by placing the dye and SDS in a syringe of the stopped-flow module and gliadins in the other, the corresponding kinetic curves can be obtained, as Fig.1 shows. These dynamic signals could be ascribed to the fact that the surfactant has a higher affinity for gliadins than for the dye, so that the presence of gliadins in the reaction mixture releases the dye, which retrieves its native fluorescence. Thus, the slope obtained corresponds to the rate of the displacement reaction caused by the gliadins, which it has been verified is proportional to their concentration. As can be seen in Fig. 1, the best initial rate was obtained when Cresyl Violet was used. This could be ascribed to the fact that the protons of the amino group of Cresyl Violet are not substituted by methyl or ethyl groups as in the other dyes, which allows it to react more easily. The reaction is very fast and the initial rate can be measured in 5 s.The main limitation of the application of the method in food analysis is its lack of selectivity since the presence of other proteins causes a positive interference. Hence, the determina- tion of gliadins in foods requires the previous separation of water-soluble albumins and salt-soluble globulins and the extraction of gliadins from the residue with a 70% aqueous ethanol solution, which allows their separation from glutelins.1 Effect of Reaction Variables The system was optimized by altering each variable in turn while keeping all others constant. All reported concentrations are initial concentrations in the syringes (viz., twice the actual concentrations in the reaction mixture at time zero after mixing). Each kinetic result was the average of three measure- ments.Those values yielding the minimum RSD for the initial rate under conditions where the reaction order with respect to the variable concerned was zero or near zero were taken as optimum. The effect of pH was studied over the range 2.0-12.5 by using different amounts of hydrochloric acid and sodium hydroxide solutions [Fig. 2(a)]. The initial rate was maximum and independent of pH over the range 9.5-1 1.5. This optimum pH range was similar to that found for lysozyme determination,16 where the decrease of the initial rate at higher pH values was 10 20 30 Tim e/s Fig. 1 Kinetic curves obtained at pH 10 for gliadins (10 pg ml-I) in the presence of SDS (1.5 X 10-3 mol I-') and A, Cresyl Violet ; B, Azure A; C, Azure B; D, Nile Blue and E, Toluidine Blue.[Dye] = 2 X moll-I; [borate buffer] = 3.5 X 10-2 mol 1-1.Analyst, August 1996, Vol. 121 1135 ascribed to the fact that the interaction between SDS and lysozyme changes when the isoelectric point of the protein (1 1 .O), is reached. However, the study carried out with gliadins has shown that their interaction with SDS also takes place when the pH is greater than their isoelectric points, which were reported17 to be between 5 and 8. In fact, the electrophoretic separation of gliadins in the presence of SDS18 is carried out at pH 8, where the surfactant is bound to proteins and eliminates the difference in charge between them. Thus, the decrease in the initial rate at high pH values could be better ascribed to the loss of fluorescence of the reagent in a strongly basic medium.A borate buffer (pH 10) was used to adjust the pH of the samples. The study of the effect of the concentration of this buffer showed that the initial rate was independent of this variable between 1.2 X 10-2 and at least 2.5 X 10-2 mol 1-1. The effect of SDS concentration is shown in Fig. 2(b). The initial rate remained constant over the range 1.6 X 10-3-2.0 X 10-3 mol 1-1. The blank signal was negligible over this range but increased at higher concentrations, so that the difference between the sample and blank was decreased and reached zero when the SDS concentration was 2.7 X 10-3 mol I-'. Although the c.m.c. reportedlg for SDS is 8.1 X mol I - l , using a stalagmometer we found that it decreases in the presence of 30 pg ml-' of gliadins and 1.75 X lo-* moll-' borate buffer, obtaining a value of 1.7 X rnol I-l, which is within the optimum SDS range obtained for the initial rate.These results show that the interaction between Cresyl Violet and SDS is modified when the micelles of surfactant are formed. Fig. 2(c) shows that the Cresyl Violet concentration does not affect the initial rate over the range 1.3 X 10-3-2.3 X 10-3 mol 1-I, decreasing at lower or higher concentrations. The study of the reagent distribution showed that the initial rate was higher when SDS and borate buffer were placed in both syringes. The variation of the temperature in the range 15-35 "C had no effect on the initial rate, the blank signal being negligible. However, the initial rate of the sample increased slightly at higher temperatures, also increasing the blank signal.Thus, a temperature of 20 "C was chosen in order to avoid subtraction of the blank. Under the optimum working conditions, the initial slopes of the kinetic curves were consistent with a first-order dependence on gliadin concentration. As the reagents exhibited a pseudo- zero order, the following kinetic equation can be proposed: v = k[gliadins], where v is the reaction rate and k is the conditional rate constant. Features of the Proposed Method The kinetic curves obtained at different gliadin concentrations were processed by using the initial-rate method. Two in- strumental conditions were used to obtain the linear range of the calibration graph, thus obtaining two linear portions, namely 0.5-5 and 5-50 pg ml-1, which conformed to the equations v = 23.3C + 4.8 and v = 3.2C + 12.4, where v is the initial rdte (s-l) and C is the gliadin concentration.The r value was 0.999 (n = 5) and 0.998 (n = 7), respectively. The LOD, calculated according to IUPAC's recommendations,20 was 0.25 pg ml-1. The precision of the proposed method was determined at two gliadin concentrations, 5 and 20 pg ml-1; the RSDs obtained (n = 11) were 1.6 and 1.5%, respectively. Finally, another salient feature of this method is speed: the initial rate for each sample can be obtained within 5 s, which makes the method applicable to automatic routine analyses. Such a short measurement time precludes the use of the batch technique and justifies the use of the stopped-flow mixing technique, which in addition mini- mizes reactant manipulations and ensures a high sample throughput. However, the application of the method to the analysis of samples containing other proteins in addition to gliadins requires their separation as they cause a positive interference.Applications The proposed method was applied to the analysis of several gluten-free food samples which contained other proteins such as albumin and casein. These samples were chosen in order to check the effectiveness of the extraction procedure used to remove these proteins, which interfere in the direct determi- nation of gliadins. Each sample was analysed according to the procedure described above and by using the standard additions method. The best results were obtained using 70% aqueous ethanol without prior removal of lipids.None of these samples gave an analytical signal. Table 1 summarizes the analytical recoveries obtained by adding three different amounts of gliadins to each sample; these recoveries ranged from 94 to 107%, with a mean of 100%. The method was also similarly applied to the determination of gliadins in different food samples containing wheat gluten. Table 2 summarizes the concentrations and analytical recov- eries obtained. These were calculated by adding three different amounts of gliadins to each sample and subtracting the results obtained for similarly prepared unspiked samples. The recov- eries were 88-105% (mean 99%). The over-all process, including sample preparation, takes about 90 min, the greater part being spent on the separation step, as the time required for the stopped-flow measurement is very short, However, the special features of gliadins entail that the methods described for their determination are generally slow.Thus, immunoassay methods take between about 2 and 6 h9-'5 1 2 ;1 10 PH 20 10 1 2 3 4 1 2 3 [SDS]/mmol 1-' [cresyl violet]/ 10- mmol1~' Fig. 2 Effect of pH (a) and SDS (h) and Cresyl Violet (c) concentrations on the gliadins/SDS/Cresyl Violet system. 10 yg ml-1 gliadins; 1.8 X lo-' rnol 1-1 SDS and 5 X 10-6 moll-] Cresyl Violet in (a); 1.75 X 10W2 mol 1-1 borate buffer (pH 10) in (b) and (c); 2 X mol 1-1 Cresyl Violet in (6); and 1.8 X lo-' rnol 1--' SDS in (c).1136 Analyst, August 1996, Vol. 121 Table 1 Recovery of gliadins added to gluten-free foods Table 2 Determination of gliadins in foods Sample* 1 2 3 4 5 7 Added/ 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 mg g-' Foundtl 23.5 i 0.5 4 7 f 2 71 f 2 23.5 f 0.2 52.0 f 0.6 7 4 i 1 24.5 f 0.6 51 f 2 7 6 f 2 26f 1 52.5 f 0.2 77.5 f 0.2 24.6 f 0.5 51.2 f 0.1 75.6 f 0.2 25.9 k 0.3 53.3 f 0.3 77f 1 25.5 f 0.2 47.8 f 0.3 73.3 f 0.7 mg g-' Recovery (%I 94 94 95 94 104 99 98 102 101 104 105 103 98 I02 101 1 04 107 103 I02 96 98 * Name (composition; trade mark, manufacturer): 1, corn starch (Maizena, CPC-Espaiia SA); 2, gluten-free flour (wheat and corn starchs, casein, sugar, soya flour, egg albumin, calcium carbonate, ascorbic acid; Sing16 Jaime Pedr6 SA); 3, food colouring [tartrazine (E-102), Sunset Yellow FCF (E- 110) and corn flour; Granja San Francisco, Nutrexpa SA]; 4, vanilla ice cream (skimmed milk powder, sugar, butter, emulgent, mono- and diglycerides, carrageenan, curcumin and aroma; Frigo SA); 5 , baby food, gluten-free (corn, rice, tapioca, soya; Sandoz Nutrition Multicereales, Sandoz); 6, yogurt (Danone); 7, cocoa powder (cocoa, lactose, corn starch, skimmed milk, vegetable fibre, lecithin, dicalcium phosphate, aspartame, aroma, salt; Cola Cao Light Nutrexpa SA).t Mean of three determinations f s. and some of them require a previous extraction with 70% ethanoL'710 Conclusions The results obtained in this work clearly show the usefulness of the kinetic method described to control the presence of gliadins in foods. It is a valid alternative to existing immunoassay methods reported for this purpose since, although it is less selective with regard to other proteins, it avoids the use of expensive immunoassay reagents. After the samples have been treated for the separation of the gliadin fraction from other proteins, the use of the stopped-flow mixing technique allows the manipulations involved in the determination step to be reduced and the measurements are obtained shortly after mixing.Hence the method can be easily adapted to the routine determination of gliadins in foods. Also, the use of dynamic fluorescence measurements at long excitation and emission wavelengths could avoid the potential background signal from the matrix, which generally has a static character and occurs at lower wavelengths. On the other hand, the probability of non- radiative quenching processes is small owing to the short fluorescence lifetime of Cresyl Violet.The authors are grateful to the CICyT (Comisi6n Interminis- terial de Ciencia y Tecnologia, Project PB9 1-0840) for financial support. References 1 2 Osborne, T. B., The Proteins ofthe Wheat Kernel, Carnegie Institute, Washington, DC, 1907, Publ. No. 84. Bietz, J. A., and Wall, J. S., Cereal Chem., 1972, 49, 416. Sample* 1 2 3 4 5 6 7 Gliadin contenttl 3822 mg g-' 17-t- 1 23 + 2 19f I 18f 1 12.6 rfr 0.4 7.5 f 0.5 Recovery Added/ 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 25 50 75 mg g-' Found?/ 26f 1 5 2 f 3 7 6 f 2 26f 1 52.0 f 0.1 78.4 f 0.5 23.0 f 0.4 4 8 f 4 72+4 23.0 f 0.7 49+ 1 72+3 26.0 f 0.8 4 8 f 1 7 8 f 2 21.9 & 0.3 49.6 + 0.2 70.1 f 0.2 25.1 f 0.2 4 9 f 2 7 8 f 2 mg g-' * Name (composition; trade mark, manufacturer): Recovery (%I 1 04 i04 101 104 i04 105 92 96 96 92 98 96 104 96 I 04 88 99 93 100 98 104 I , wheat flour (Gallo, Comerciaf Gallo SA); 2, wheat and rice flours [wheat, rice, tartaric acid (E-334), tartrazine (E-102) and Plnceau 4R (E-124); Yolanda, Industrias Ldpez Caballero SA]; 3, white bread; 4, biscuit (wheat flour, sugar, sodium and ammonium hydrogencarbonate, lecithin, aroma, sodium hydrogensulfite; Galletas Maria, Cuetara SA); 5 , breakfast cereal (wheat, honey; Pascual); 6, baby food (wheat, corn, barley, rye, rice, sugar, maltodextrin, lecithin, vitamins, vanilla; Nutriben crecimiento 5 cereales, Alter SA); 7, cocoa powder (cocoa, sugar, wheat flour, malt extract, dicalcium phosphate, aroma, salt; Cola Cao, Nutrexpa SA).7 Mean of three determinations f s. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Bietz, J. A., and Simpson, D. G., J. Chromatogr., 1992, 624, 53. Payne, P. I., Holt, L. M., Jarvis, M. G., and Jackson, E. A., Cereul Chem., 1985,62, 319. Werner, W. E., Wiktorowicz, J. E., and Kasarda, D. D., Am. Lab., 1994,26,32NN. Bietz, J. A., and Schmalzried, E., Food Sci. Technol., 1995, 28, 174. Lookhart, G. L., and Albers, L. D., Cereal Chem., 1988, 65, 222. Ciclitira, P. J., and Lennox, E. S., Clin. Sci., 1983, 64, 655. Meier, P., Windemann, H., and Baumgartner, E., 2. Lebensm.- Unters.-Forsch., 1985, 180, 467. Fritschy, F., Windemann, H., and Baumgartner, E., 2. Lebensm.- Unters.-Forsch., 1985, 181, 379. Skerritt, J. H., J. Sci. Food Agric., 1985, 36, 987. Troncone, R., Wale, M., Donstiello, A., Fanis, E., Rossi, G., and Auricchio, S., J . Immunol. Methods, 1986, 92, 21. Freeman, A. R., Galfre, G., Gal, E., Ellis, H. J., and Ciclitira, P. J., J. Immunol. Methods, 1987, 98, 123. Freeman, A. R., Galfre, G., Gal, E., Ellis, H. J., and Ciclitira, P. J., Clin. Chim. Acta, 1987, 166, 323. Skerritt, J. H., and Hill, A. S., J. Agric. Food Chem., 1990, 38, 1771. Gala, B., G6mez-Hens, A., and Pkrez-Bendito, D., Tulanta, in the press. Wrigley, C . W., J . Chromatogr., 1968, 36, 362. Ng, P. K. W., Slominski, E., Johnson, W. J., and Bushuk, W., Cereal Chem., 1989, 66, 536. Cline Love, L. J., Habarta, J. G., and Dorsey, J. G., Anal. Chem., 1984,56, 1133A. Long, G. L., and Winefordner, J. D., Anal. Chem., 1983, 55, 712A. Paper 610 I661 H Received March 8,1996 Accepted June 3 , I996

ISSN:0003-2654    

DOI:10.1039/AN9962101133

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 (1137-1145) 1137 The opinions expressed in the following article are entirely those of the author and do not necessarily represent the views of either The Royal Society of Chemistry or the Editor of The Analyst. Perspective Traceability of Measurements in Chemistry Yu. I. Alexandrov D. I. Mendeleev Institute of Metrology (IMM), 19 Moskovsky Avenue, 198005 St. Petershurg, Russia It is shown from the analysis of metrological validation of all physical quantities measured in chemistry that only for ‘concentration’ and ‘amount of substance’ is there no admitted measurement compatibility system like that for other physical quantities. Based on the fundamental notions of metrology, an attempt is made to derive a general concept of traceability and reliability of the measurement of concentration.Special attention is paid to metrological uses of standard samples and reference materials. Specific features of the quantity ‘amount of substance’ and its unit, the mole, are discussed in terms of metrology. Keywords: Traceability; validation; metrology Introduction Considerable attention has been paid recently to chemical measurement problems, particularly by leading international organizations such as EUROMET, EURACHEM, IUPAC and BTPM, which resulted in the establishment in 1993 of the Working Group for Cooperation in International Traceability in Analytical Chemistry (CITAC), the new Comitk Consultatif pour la Quantitk de Matiere (CCQM) attached to the Inter- national Committee of Weights and Measures (CIPM), and the establishment of a new EUROMET subject field, ‘amount of substance’.1-3 A wide range of problems arise when dealing with the need to ensure the accuracy and international traceability of measure- ments in chemistry at all levels and in publications devoted to the first results of the activity of the above groups. One of the problem fields is the traceability of all measurements in chemistry. Further, there are problems with the traceability of measurements in analytical chemistry. Finally, a specific problem is being advanced, i.e., the creation of an international system for traceability to the mole, similar to the traceability of measurements of physical quantities to the relevant base units in the SI system. The importance of this problem is indicated by the establishment of the Comitk Consultatif pour la Quantitk de Matigre.All the above problems are based on either well known disagreement in analytical results or dissatisfaction with the accuracy achieved. Moreover, there is a widespread opinion that in chemical analysis the mechanism of traceability has not yet been developed from routine chemical measurements to an internationally recognized level, like that created for physical quantitites. True, it is not pointed out which physical quantities are being measured when carrying out chemical analysis. From analysing information (regularly published in the VAM reviews in Analytical Communications) concerning the activity of the established Working Groups, we have concluded that their ultimate aim is the improvement of the reliability of measurements not in chemistry as a whole, but primarily the results of chemical analysis.Even though all this takes place when discussing measurement problems in chemistry, in practice it means an attempt to produce a system for establishing traceability to the mole. However, it should be noted that the precise metrological formulation of the above-considered problems is absent from papers dealing with them. What do we mean? First, the notion ‘chemical measurement’ should be defined more exactly. This term is often used when discussing the problems of traceability in chemistry. Terms related to it, such as ‘achievement of comparability in chemical measurements’, ‘traceability in chemical measurement’ and ‘traceable chemical measurements’ are used.Nevertheless, we have not met a clear and unambiguous definition of the term ‘chemical measure- ment’ itself. In our opinion, two approaches are possible when considering this notion: 1. Measurements of all known physical quantities in chem- ical research or production should be referred to chemical measurements. 2. A specific kind of measurement that is essentially different from measurements carried out in other fields should be referred to chemical measurements. When analysing scientific and technical measurements in chemistry, we see no ground for their differentiation as ‘chemical measurements’. All these measurements, including those carried out routinely, are based on measuring known physical quantitites. No doubt some specific properties are also measured in this field. However, these properties are repre- sented by corresponding physical quantities such as concentra- tion and pH.One of the basic quantities of the International System of Measurements (SI), i.e., ‘amount of substance’, with its unit, the mole, may be formally added to these specific quantities. In terms of metrology, all these quantities, although specific for chemistry, nevertheless represent the measured physical quantity in complete agreement with the definition ‘physical quantity’. As will be shown below, the only exception is the ‘amount of substance’. The failure to distinguish chemical metrology as an indepen- dent area of metrology confirms the above conclu~ion.~~~ The time that has elapsed since has demonstrated that such attempts are in vain.Proceeding from the first approach, i.e., taking into account all kinds of measurements in chemistry, it is necessary to specify the list of physical quantities associated with each of the problems. The analysis of traceability for each of these quantities should be a next step. It is hardly likely that all the measurements made in chemistry can be reduced to only measurements of the amount of substance in moles. Moreover, we do not know any device (‘molemeter’) for such measure- ments. In this paper, we would like to survey those physical quantities which are measured in chemistry and whose1138 Analyst, August 1996, Vol. 121 reliability is so inadequate at present. In addition, the metro- logical concept of the measurement compatibility of physical quantities being used in chemistry and the ways in terms of metrology to improve validity and precision in chemical analysis are also of interest.Physical Quantities Measured in Chemistry as a Whole In chemistry, measurements of mass, temperature and pressure are widely used. However, these wide-scale measurements are entirely under metrological control. The traceability of these physical quantities at the highest metrological level is made by using national primary standards. These standards are co- ordinated with each other by periodic verifications carried out by both the corresponding committees of the CIPM and national metrological laboratories. As an example of such verification, a study on the traceability of mass standards can be men- tioned.6-’ Measurements of thermal, electrical and optical quantities are also widespread.Similar systems for the traceability of the quantities ensuring a required accuracy have been created. There is a large set of physical-chemical quantities repre- senting the corresponding properties of both the individual substances and their mixtures. A typical example of such a physical-chemical quantity is pH. However, the traceability of these quantities is also established by national metrological laboratories and by corresponding commissions of When considering consistently all the known physical quantities, we can conclude that except for two quantities, all the others are under metrological control. This means that for these quantitites there exists a continuous chain of comparisons or calibrations that links the results of the measurement to national or international standards.This also implies that the uncertainty of the measurements is known at each step in the traceability chain. In analytical chemistry, measurements of virtually all known physical quantities are performed and, as mentioned above, are under metrological control. The classification of all these physical quantities was excellently set out by Danzer et al.13 When the measurements of these quantities are used for the determination of the qualitative composition of substances under investigation, these measurements are of independent values.14 In other cases, when the analysis is quantitative, their measurements play a subordinate role and serve for the determination of a physical quantity specific for the given analysis.Concentration is meant here, of course. If strucural analysis and phase composition analysis are not considered, the only specific quantity in chemical analysis is concentration.14 No system similar to those which have been developed and tested during many years for ensuring trace- ability for other SI physical quantities exists for this physical quantity, as was noted by Richter et al.,15 while it is the uncertainty of results for concentration determination that causes great anxiety. Another quantity for which not only the traceability system but also its standard are absent is the ‘amount of substance’ and its unit, the mole, introduced into the SI system in 1971 by a Resolution of the 14th Conference Genkrale des Poids et Mksures.Is this an accidental fact? To our mind, it is not, as it is caused by the specific, from the metrological point of view, features of these two quantities. Let us try to prove this statement. IUPAC.1’7’2 Concentration and Its Specific Features as a Physical Quantity Concentration is always a concrete quantity, i.e., when dealing with this quantity, it is necessary to relate it to the name of a concrete substance or compound. In this respect, it differs from other physical quantities. At the same time, it is this very feature that allows a description of the qualitative diversity of chemical compounds, elements and other specified entities. Indeed, it follows from the definition that concentration is the quantity expressing a relative content of a given component in a mixture or solution.16J7 One cannot just say ‘concentration’.It is always necessary to name the component whose concentra- tion is being dealt with. Therefore, it is impossible to create a common standard of concentration, like the standard of mass, length, time, etc.; such a standard will always represent the standard of the concentration of a given component with a very limited field of application that is conditioned by that particular mixture or solution for which the concentration of the given component was established. By means of chemical interactions, the number of components for connection with such a standard would be possible, and could be extended. However, the transfer of the size of the concentration unit is limited. Another specific feature of concentration as a physical quantity is the fact that it has not one unit, but a number of units differing in dimensions.The latter results from the well known definition of concentration, which does not specify what units should be used for characterizing the content of both the given component and the mixture (solution) as a whole. Therefore, by using with this aim the units of SI quantities such as mass, amount of substance and volume, six units of concentration are obtained. For the above reasons, each of these units is, in its turn, an SI unit. Concentration units are considered in more detail elsewhere.l*,19 It should be stressed that these features of concentration as a physical quantity make it impossible to create its common primary standard.Indeed, despite the large number of measurements of concentration in very different substances under analysis, a common primary standard for this quantity cannot be found in any country. Nevertheless, all the practical activity on the mass production of an enormous range of chemical substances, from polymers to ferrous and non-ferrous metals, indicates that there are practically no essential differences in chemical analysis results. What is the reason? The answer is that a sufficiently orderly system of traceability for measurements of concentration has spontaneously formed in analytical practice, completely different, however, from hier- archy schemes for the traceability of all the base and most of the derived SI units. Still, the reliability of chemical analysis results of complex chemical substances, of contaminants in various media, and of impurity traces in high-purity substances is unsatisfactory.Therefore, it is important to clarify the above-mentioned system of the traceability of concentration measurements and the reasons for the dissatisfaction with experimental results. Main Techniques of Traceability in Metrology In principle, there are two main techniques for the establishment of the traceability of physical measurements: centralized and decentralized (local). The first approach is based on the primary national standard and hierarchy scheme and it implies a continuous chain of comparisons or calibration linking the results of measurements to the primary national or international standard.Here the uncertainty of the measurement results is evaluated at each step in the traceability chain. This is the so-called metrological pyramid, where the unit size of a physical quantity is retraced from one metrological level to another. An example of such a pyramid is the system of the transfer of the mass unit size, kg (see Fig. 1). We would suggest the use for this type of metrological pyramid (when at each step, from the working level to the primary standard level, it is the same quantity that isAnalyst, August 1996, Vol. 121 1139 continuously compared) of the term ‘retracing’ instead of traceability. The term ‘traceability’ assumes establishing relationships first of all between a derived unit and base SI units entering in its dimension.The difference between the terms ‘retracing’ and ‘traceability’ is most clearly revealed when comparing hier- archy schemes for the units of the base quantities and the protocol of traceability (Ref. 15, Figs. 1 and 2). The second (local) approach is used when reproducing derived quantities directly by a field device, by applying the known theoretical dependence of this derived quantity and the base physical quantities entering in its dimension. Electrical calibration in calorimetry can be considered as an example of such a technique. In this case, the main condition for reproducing the admitted size of the joule is the use in this calibration of devices for measuring magnitudes of time, voltage, current or resistance that have been calibrated within the frames of hierarchy schemes for each of these physical quantitites.With this process, the traceability of the joule to the base SI units is achieved. Specific Features of Traceability in Concentration Measurements Concentration is one of many derived quantities and the reproduction of each of its units at the highest metrological level is performed by the second approach considered above. However, this is the only similarity between concentration and Metrological Hierarchy Scheme Metrological Pyramid Common Primary Standard (The National Standard) - _ - _ - - Common Primary Standard (The National Standard) L I Secondary Standard I I Secondary Standard I 5 I 0 0 0 oc t- W oc z a - Working Standard 3 I High Accuracy ILower Accuracy] Accuracy 1 FIELD MEASURING INSTRUMENTS FIELD MEASURING INSTRUMENTS Fig.1 Principle of the centralized transfer of the size of the units of physical quantities. SI-Base Units L Gravimetry Volumetry Coulometry Cryometry Optical Methods I Accuracy methods for the measurement T T Absolute Methods of Analysis Standard Sample T Local Working Standard Sample . Local Working Standard Sample t Local Working Local Working Standard Sample I Industrial Metal Industrial Materials Clinical Biological Materials Environmental Agricultural Analysis Analysis Analysis Analysis Analysis Chemicals Analysis FIELD MEASURING INSTRUMENTS Fig. 2 Principle of the localized traceability of concentration measurement in chemical analysis.1140 Analyst, August 1996, Vol. 121 other derived quantities. Whereas for other derived quantities primary national standards exist or may be created, such a standard cannot be produced in principle for concentration, as has already been indicated above.The ensuring and maintenance of the unit of concentration measurements, i.e., traceability of concentration measurements, is accomplished exclusively by the local method. An important contribution to the development of this technique has been made by NBS specialists.20-21 There are two ways to carry out the local approach. The first is to produce local hierarchy schemes, i.e., limited local metrological pyramids. They differ from the hierarchy schemes used in the first (centralized) approach; first, there is a much narrower field of application; and second, a local standard lies at the basis of the local pyramid but not a primary national standard.An example of such a local hierarchy scheme may be the system of traceability for analysing certain gases.22 Taking into account a great variety of substances in chemical analysis, a ~ number of hierarchy schemes with their own working standards may be created or already exist. Thus, whereas the centralized system is based on and characterized by only one standard (primary national standard), the specific feature of the local approach of concentration traceability is the presence of a number of standards, which, as a rule, are not related to each other. The second direction is to use standard samples at the working level without any traceability steps. Traceability in titrimetry is an example.23 Another example which underlines the local character of the systems of traceability of concentra- tion measurements is given in Refs.24 and 25. Peculiarities of the reproduction and maintenance of concentration measure- ment unity are considered in more detail in Ref. 26. Absolute Methods of Analysis as the Basis for the Reproduction of the Established Sizes of Concentration Units Let us now turn to the key problem of concentration measurement compatibility, i.e., the reproduction of its units. The reproduction of the units of concentration is metrologically most exact if it is based on an absolute method. Here we proceed from the definition of the absolute method given in Ref. 21 (p. 367): ‘By an absolute measurement of a physical quantity, such as the velocity of light, we mean determination of the value of that quantity in terms of significant fundamental units of length, mass, time, etc., and of those constant parameters that characterize the accepted system of theoretical equations that connect several pertinent quantitites’.For concentration, the following absolute methods are known. 1 Gravimetric Method The gravimetric method of analysis is based on the law of constant and multiple ratios. The method of analysis consists in estimating the mass of the precipitating products in a chemical reaction. Sometimes gaseous products of a chemical reaction after their precipitation are also determined by this method. The coupling equation for the gravimetric method of analysis for a general case when (1) aA’+ + P B Z - 3 (A,Bp), is written as where YA CWMA mAB y, = -x -.-EJ!.(2) MA,B~ ms = fraction by mass for component A [MI M-11; m = FmA,Bp = mass of component A in the mixture; mAorBp =mass of the reaction product; F = - = stoichiometric number; MA = atom mass of the component being MACYq3 = molecular mass of the reaction ms = mass of the analytical sample. &MA MAaBp an a1 y sed; product; In gravimetry, mass is the base physical quantity related to concentration that is measured. Volumetric Method The principle of gas volumetric analysis is based on the ideal gas equation. The method involves the determination of the amount of gaseous reaction product by the absorption of one of the components of the gas mixture by means of the solution of a corresponding reagent and by measuring the volume decrease of the gas mixture caused by the absorption.The coupling equation for gas volumetric analysis at a constant pressure p and temperature T is MAPVA Y* = - RTm, (3) where mA = x VA; MAP RT molar mass of component A; partial volume of component A, which is equal to the difference of the mixture volumes before and after the absorption. For a mixture of gases, the-concentration is estimated by this method, as a rule, in molar fractions (x = VA/V) [NlN-l]. However, no direct measurements of the number of moles of component A in the gas mixture or of the total number of moles are carried out. These values are calculated from the gas volumes using the ideal gas law. In volumetry, temperature and mass are the base physical quantities related to concentration which are measured.Potentiometric and Ionometric Method The principle of the measurement is based on the following: an electrode immersed in an electrolyte solution receives a potential depending on the concentration of the electrometric- ally active form of a component being analysed. Under certain conditions, this relationship obeys the Nernst law. The coupling equation for the potentiometric method of analysis in a general form is yi = f(expAO= 0 (4) where yi = concentration of component i; AE = potential difference; I = magnitude of current passing around electrode. In this case the concentration is expressed in terms of a mass concentration [L-3 MI, where L is the length. Here the physical quantity to be directly measured is the potential difference, i.e., voltage.Cryometric Method The cryometric method is based on the temperature decrease in the solid-liquid phase equilibrium under the effect of impurities present in the sample. For ideal solutions, the concentration of impurities may be estimated from the Van der Waals equation.Analyst, August 1996, Vol. 121 1141 The coupling equation for the cryometric method of analysis under certain conditions is X 2 = AAT ( 5 ) where X2 = molar fraction of all impurities [NlN-l]; A = cryoscopic constant of the main component; AT = cryoscopic temperature decrease. Although the concentration is determined by this method in molar fractions [NlN-l], the quantities that are measured by direct methods are temperature and time (dynamic method), or temperature and amount of heat energy (static method).It should be noted that this method is in no way related to direct measurements of the amount of substance in its unit (mole) as could be expected from the dimension of the given unit of concentration. Coulometric Method The principle of this method is based on the Faraday law. The method of analysis involves the measurement of the quantity of electric charge passing through an electrolyte cell in an electrochemical reaction. The coupling equation for this method assuming that the current yield is equal to 100% and that are no side reactions is where MA = molar mass of the component being analysed; Q = It = quantity of electric charge; MAQ MAI~ ZF zF mA =--- - = mass of component A in the mixture; = number of electrons taking part in the electro- z F = Faraday constant.chemical reaction; As a rule, the concentration determined by this method is in The base physical quantities to be measured by this method mass fractions [MlM-l]. are electric current, mass and time, as follows from eqn. (6). Spectral Methods Based on the Absorption of Electromagnetic Radiation The principle of the method is based on non-elastic interactions of the sample under analysis with an external source of electromagnetic radiation according to the Beer-Lambed law. The method consists in measuring the intensities of the incident radiation (Io) and the radiation passing through the absorption layer (Z,). The coupling equation is I,, = Ioexp(--cJ,Nl) (7) where 0, = absorption cross-section; l = length of the absorbing column of solution.Usually spectral methods require calibration by using standard samples of known concentration. However, if the data on the absorption cross-section are found independently from this calibration, this method may be used as an absolute method. One may also use the magnitudes of one of the base atom constants, i.e., oscillator force (f,,). For an optically thin layer, the concentration (N) is related to full absorption by ne2 mc where e and c are constants and A, = - Nfnml (9) y2 f I n - I , " . A, = J -dv I0 VA Remarks Much attention has been paid recently to the isotope dilution mass spectrometric method (IDMS), which is considered not only an absolute method for the measurement of concentra- tion21,27 but also a method that allows, according to some authors,15J8-30 direct measurement of the amount of substance in moles to be performed.We cannot agree with either the first or the second statement. First, we shall try to prove why the IDMS method cannot be considered an absolute method. The arguments to prove that its use for direct measurements of moles is unjustified will be presented below, when discussing specific features of the mole as a physical quantity. From the definition of the absolute method, it follows that the fundamental feature of such methods is that there must not be any dependence of one concentration on the other in the coupling equation. This can easily be seen from the coupling equations for the absolute methods considered above. What, then, is the situation with IDMS? Its coupling equation, used not for calculations f0rm:31 but for measurements, has- the following atom fractions of isotopes sample; atom fractions of isotopes spike; A and B in the A and B in the concentrations of the element in the sample and spike, respectively; weights of the sample and spike, respec- tively; measured isotope ratio of isotope A to isotope B.Basically, in terms of metrology, this method differs little from, e.g., a chromatographic method of analysis. Both methods use calibration by samples of known concentration and whereas in one method the ratio of the peaks area is determined, the other one uses the measured isotope ratio. At the same time, in both methods the size of the concentration unit is found not as a result of the measurements by these methods, but at the stage of the preparation of the samples that are used for the calibration. Therefore, the accuracy of measurement by the IDMS method, like any other relative method, cannot be greater than the accuracy with which the concentration in the spike was determined.It should only be noted that the concentration in the spike is determined gravimetrically, which is due to the spike preparation technique. Hence the concentration found by the IDMS method is also determined as a mass fraction [MI M- 11. Intermediate Conclusions The above brief review of the absolute methods of analysis had the following aims:1142 Analyst, August 1996, Vol. 121 1. To show the variety of methods that can be used for reproducing the units of concentration. 2. To show that one method alone cannot become the basis for creating the single national standard of concentration owing to the limitations of each of the above-considered methods. The variety of methods itself, each with its own field of application, is an additional proof of the fact that the reproduction of concentration units for a wider range of substances under analysis can be performed by the local approach only (see above).This statement has already been mentioned in Ref. 32: ‘Variations of the idealized accuracy-based measurement system (i.e., local traceability pyramid; Yu. A.) are now in use in many industries and are also being implemented in clinical and environmental analysis’. 3. To show that the choice of concrete units when reproduc- ing the units of concentration is determined by the coupling equation which lies in the theoretical basis of the absolute method.4. To show that when considering traceability of concentra- tion units, we deal not with one, but with a set of base and derived units of the SI system, and first of all with the units of mass. 5. To show that none of the absolute methods for reproduc- ing units of concentration is related to the direct measurement of the amount of substance in moles. The number of moles when mole is in the dimension of the concentration unit is determined not from direct measurements as follows from the fundamental notions of metrology, but from the calculations of other measured quantities. 6. To show that the statement which can sometimes be met that traceability in ‘chemical measurements’ to SI units is still at its initial stage does not reflect the actual metrological state of measurements in this field.This conclusion refers not only to concentration but also to all other physical quantities that are measured in chemistry. Nevertheless, despite the variety of absolute methods, it is not always possible to reproduce concentration units owing to the variety of substances to be analysed. Many hopes have been placed on interlaboratory comparisons to fill this gap. However, this method cannot be used to certify primary standard samples with respect to their degree of purity, because such comparisons are performed with analytical instruments which, in their turn, require calibration using samples with known concentrations of components. Therefore, recognizing that absolute methods have un- doubted priority in the reproduction of the units of concentra- tion, in those cases when their application appears impossible, preference should be given to the technique for the preparation of mixtures with a given concentration that have been prepared from individual components and previously undergone a multi- stage purification.This method to reproduce concentration units is metrologically correct to the extent to which these purified substances approach the absolute purity state. Nevertheless, the method allows one to avoid essential differences in the sizes of concentration units as compared with the method of inter- laboratory comparisons. This approach is not new, and analytical practice shows, in particular, the use of a so-called o-calibration whose advantages are considered in Ref.33. Main Conclusions on the Achievement of the Traceability of Concentration Measurements Summing up, it should be noted again that: 1. It is impossible to produce a common national standard of concentration responding to a variety of areas of chemical analysis. Therefore, achieving traceability in concentration measurements is possible only within the frames of the local (decentralized) approach. 2. The theoretical basis for the reproduction of the admitted sizes of the concentration units is absolute methods of analysis. It is these methods that ensure the traceability of the concentration units to all the base SI units, except the mole. 3. The primary standard samples based on high-purity substances are the key point in the transfer of the size of the concentration units.4. When the use of absolute analysis methods is impossible, it is preferable to perform the reproduction of concentration units with primary high-purity samples. Thus, the reproduction and transfer of the size of concentra- tion units which guarantee measurement compatibility are implemented in practice according to the scheme presented in Fig. 2.14 It should be noted that in its main elements, this scheme has much in common with the approach considered in Ref. 15 (Fig. 2). Reliability of the Results of Chemical Analysis As has already been noted, in a number of cases we have to deal with considerable differences in analytical results. For example, the Institute of Chemistry of High-Purity Substances of the Russian Academy of Sciences (Nizhniy Novgorod) has gained much experience in the generalization of analytical results of a large number of high-purity substances. The results show that the data obtained in two laboratories on the analysis of the same sample can disagree more than 10-fold in as many as 40% of cases.34 The main part of these divergences is caused by systematic errors due to the analytical procedures for high- purity substances.This example reveals another problem, i.e., the reliability of the results of chemical determination for substances which are difficult to analyse. Here the term ‘determination’ but not ‘measurement’ is used purposely. Looking at the stages in the determination of the concentra- tion of complex substances (first, with a very inhomogeneous distribution of its components), the following stages should be mentioned: sampling process, sampling preparation and the measurement of concentration. So far, there has been no agreement of opinion among analysts as to whether the sampling process (and, in a number of cases, sampling preparation) should be included in quantitative analysis or not.However, the point that matters for the users is not the above argument but the fact that each of these steps contributes to the uncertainty of the concentration determination. We could list many examples of the effect of sampling process and sampling preparation errors on the reliability of chemical analysis results. However, they contribute nothing new to the fact that the measurement of concentration is only one step in the process of obtaining the data on the concentration of components in the substance being analysed.Several consequences of this fact should be noted. 1. The stages of sampling process and sampling preparation make a considerable contribution to the deviation of the results on concentration determination from the true values. 2. It follows from point (1) that in these cases the application of more precise instruments does not improve the accuracy of the results of concentration determination. In such cases, it is not the accuracy of concentration measurement in the sample that should be increased, but the correctness of the techniques for the sampling process and sampling preparation. Thus, the development of methods for the quality control of analytical results acquires still greater significance.At the same time, as follows from point (2), the problem of the reliability of analytical results cannot be solved only within the frame of ensuring traceability, because these are two different problems. The solution of each of these problems requires its own metrological approaches and measuring devices.Analyst, August 1996, Vol. 121 1 I43 Metrological Applications of Standard Samples As was mentioned above (Fig. 2), standard samples are the key element in the system of the transfer of the size of concentration units. In order to check the reliability of the entire multi-stage procedure for obtaining the data on the concentrations of components in an initial substance, it is necessary to have other metrological devices, called ‘reference materials’.The differ- ence between standard samples and reference materials can be clearly seen when considering the measurement as an integral process including the object of measurement, measuring device and measurement method. Indeed, to carry out the measure- ment, one must choose what to measure, the instrument to measure it with and the method of measurement (see Fig. 3). Let us consider the metrological application of standard samples. If we suppose that a primary standard is used as a measuring device, then when transferring the unit size within the metrological hierarchy scheme, a secondary standard from the level immediately below it becomes the object of measure- ment. However, and this is very important, the object of measurement is transformed into the above-mentioned secon- dary standard not during this measuring process (I, Fig.4), but during the next one (11), in which this former object of measurement appears already as a measuring device. This qualitative transformation of the object of measurement in one measuring process (I, Fig. 4) into the measuring device when carrying out the next measuring process (11) actually represents the point of the transfer process of the size of the unit. At the same time, this transformation illustrates the metro- logical function of standard samples. Fig. 4 shows this transformation by an arrow connecting the first and second measuring processes. Metrological Applications of Reference Materials What, then, is the control of validity (reliability)? Let us consider measuring process I11 (Fig.4), in which the method of measurement should be tested. Here either a standard sample is used directly as a measuring device or a measuring device calibrated by this standard sample. This fact is reflected in Fig. 4 by an arrow directed from measuring process TI to measuring process 111. Reference material is used as an object of measurement for the control of the validity of the measurement method. Its metrological characteristics have been preliminarily estimated to a necessary degree of accuracy, e . g . , during measurement process I, which is shown by an arrow directed from the object of measurement in process I to that in process 111. The comparison of measurement results obtained for this reference material during the first and third measurement processes permits one to judge on the presence or absence of a statistically significant difference between them.Thus, one may Measurand (Object of Analysis) Fig. 3 Components of the measurement process. detect a systematic error of the measurement method used in process 111. A reference material is a material having a well stated value of one or more of its physical or chemical properties experimentally determined within stated measurement uncer- tainties, and designed for application as an object of measure- ment during the measurement process with the aim of controlling either the measurement method itself or other stages of quantitative analysis. Differences in Metrological Function Between Standard Sample and Reference Material It becomes evident from the above discussion, i.e., from the comparison of the combinations of measurement processes 1-11 on the one hand and processes 1-111 on the other, that there is a difference between the metrological procedure called transfer of size of units and the metrological procedure to detect a systematic error (quality control). One can see a distinct difference between calibration with the use of standard samples in which their uncertainty enters the overall uncertainty, and the test of a complete analytical procedure using reference materials, which acts as a criterion of the accuracy of a certain stage of chemical analysis, whose uncertainty does not enter into the measurement results.At the same time, such a comparison indicates different roles played by standard samples and reference materials in the measurement process, standard samples occupying always the place of measuring devices, whereas reference materials, on the contrary, are objects of measurement.Such a difference predetermines a difference in requirements as to the compositions of standard samples and of reference materials. Whereas standard samples must not contain impurities that affect the correctness of calibration (matrix Local Primary Standard Sample Measurand I I II. Measurement Process Fig. 4 (1-11) and reference material (1-111). Comparison of the metrological application of a standard sample1144 Analyst, August 1996, VoE. 121 effect), reference materials, in contrast, must be most close in their composition and concentrations of their components to the object being analysed.The closeness of reference materials in composition to the object under analysis makes it possible to reveal the effect of the impurities on the results of measure- ment. It is expedient to produce reference materials not only for testing measurement methods, but also for controlling stages such as the sampling process and sampling preparation. This is illustrated by the use of various NBS Copper Reference Materials at different stages in the copper materials cycle (Ref. 21, Fig. 11). The analysis of recent publications indicates an increasing role of such reference materials.35 Specific Features of the Quantity ‘Amount of Substance’ and its Unit, the Mole, From the Point of View of Metrology Let us now consider the second quantity, for whose measure- ment not a single national standard has so far been developed and for which no system for traceability to its units exist.We mean, of course, the base SI quantity ‘amount of substance’ with its unit, the mole. There are conflicting opinions on this quantity and its units. On the one hand, some scientists assume that the development of the system of traceability to the mole and its validity at the highest metrological level, as has been achieved for all the other base SI quantities, will help in achieving reliability of chemical analysis.1-3 On the other hand, there is a point of view according to which the quantity ‘amount of substance’ and its unit must not be included in the SI system, being a purely mathematical (a number of entities) and not physical quantity.36~37 Personally, I share this view,38 for the following reasons.There are two close methods for a quantitative determination, i.e., counting and measurement. The difference between them lies in the nature of the units used in each of these procedures. In measurement, these are the units of physical quantities, each of them representing a certain concrete quantitative character- istic of a measurement object. In counting, the units are the objects being counted themselves. In a number of cases, when carrying out measurements, counting is used as one of the methods of measurement, e.g., the measurement of length by counting the number of the known wavelengths, or the measurement of radioactivity by counting the corresponding particles.‘In order to use the mole as a unit for the measurement of the amount of substance, its definition forces one to use methods which count particles’.39 This is an absolutely correct conclu- sion, which, at the same time, reveals the physical nature of the mole as a counting number. It should be noted that when estimating the amount of substance, the determination of the number of particles is the only and final aim in the counting of their number. The old and the new interpretations of the mole, and the comparative analysis of these two methods for quantitative determination, are given in detail in Ref. 38. The main conclusions that follow from this analysis are that, first, the fundamental difference between these two procedures lies in the nature of their units, and second, when counting, there is no need to introduce the concepts of a physical quantity and provisional units because in this case the units are the objects being counted themselves.In this connection, of great significance is the fact that apart from the term ‘amount of substance’, there is no definition of this ‘physical quantity’.40 Indeed, if the mole is a counting number that is not related in any way to the system of physical quantities being measured, then, as follows from the counting procedure, the introduction of any definition for the math- ematical quantity ‘amount of substance’ becomes excessive. It should be noted that giving a name to a certain concrete number does not make the latter an independent physical quantity.This can be clearly seen from the comparison of names of numbers such as dozen, ream and mole. According to Ref. 41, ‘the mole, besides being ST’s unit of numerousness, is a unit especially suited to quantify very large (indeed!) numerousnesses. Other units of numerousness-non- SI-are, for instance, the dozen, the gross or the ream . . . One can see that the mole is the appropriate unit (the SI one) to quantify very large numerousnesses, while dozen, gross or ream are appropriate units to quantify small numerousnesses’. It is very difficult indeed to agree with such a statement. Moreover, to my mind, this statement is an additional argument for the fallibility of introducing the mole as the base unit into the system of measurable physical quantities.We have carried out an analysis of how the mole discharges the two main obligations as the base SI unit42 1. ensuring the traceability in the measurements of the physical quantity itself; and 2. ensuring the traceability of all derived units in whose dimensional representation this quantity is used. Neither of these functions is fulfilled by the mole or, as follows from its definition, can ever be fulfilled. Therefore, I cannot share the hopes of those who consider the mole as the key unit whose traceability system will be able to improve significantly reliability in chemical analysis. In this connection, let us consider again the IDMS method. It is difficult to agree with the authors’ statement2*-3” about the possibility of using this method in practice for reproducing the mole, for the following reasons.The mole was introduced as the unit of the base physical quantity. Hence, like all the units of other base quantities, the mole must be reproduced at the national primary standard level, its size having to be transferred to the level immediately below it by direct methods and being absolutely independent of the size of the units of other quantities. This is the fundamental concept of the SI system. Let us now consider the coupling equation for the IDMS method, which is suggested to be the basis for performing measurements of the amount of substance in moles: where N = number of particles (atoms, moles, etc.); R = experimentally measured ratio of isotopes A and B in sample (p), spike (s) and mixture (m).However, the initial coupling equation, designed for meas- urement purposes but not for calculation, is written in another form [see eqn. (10) or Ref. 21, p. 3681 that indicates the traceability carried out by the IDMS method to concentration or mass, rather than to the amount of substance. In this connection, very significant is the statement of Richter et al. (Ref. 15, p. 317), who studied the possible application of the IDMS method for ensuring the traceability of measurements of the amount of substance with the SI system: ‘The estimation of the number of particles in the spike is based upon the weighing with a calibration balance and thus directly connected with the traceability to the mass. . . The calculation of the unknown amount of substance is made in accordance with the equation’.The formation of the CCQM attached to the CIPM is very well timed, taking into account the different points of view on the problem of metrological functions of the mole. By proceeding from basic statements of metrology that should be equally applied to all the base SI quantities, without any exception, the CCQM should follow, in my opinion, one of two possible ways: 1. either to give a definition of the physical quantity ‘amount of substance’ and concrete recommendationsAnalyst, August 1996, Vol. 121 1145 as to the procedure for the development of a single standard of the mole, as has been done for all the other base SI quantities and their units; or 2. having recognized the futility of these attempts, to recommend that the next CGPM session excludes this quantity and its unit from the SI system. In this case, the practice of carrying out measurements in chemistry may serve as one of the criteria.The period of 25 years since the introduction of the mole seems to be more than enough to make the decision. At the same time, the calculations carried out both in chemistry and in other fields of industrial and scientific activities are actually based, as before, on the use of the gram- mole concept. However, one is forced to camouflage this quantity artificially as the ‘mass of the mole’. Conclusions Summing up, I would like to return to the title of this paper. I would like to hope that I have succeeded in demonstrating, although to a small degree, the futility of the attempts to produce a single system for the traceability of measurements in chemistry with the mole at its top, similar to the systems made for all the other base SI units. Ensuring the reliability of the results of measurements is the main problem not for all the measurements carried out in chemistry, but only for the determination of concentration in the chemical analysis of complex substances.However, it is for these very substances that the reliability (i.e., obtaining true values of concentration) is not determined by precision of the measurement of the analytical sample only. In this case, the main contribution to the overall uncertainty is made by uncertainties of the sampling process and sampling preparation. These uncertainties cannot be reduced by using more precise measuring devices.Therefore, when analysing objects of this kind, the main attention should be paid, first, not to the problems of traceability to the base units (because, as a rule, the appropriate level of accuracy of the measurement itself has been already achieved), but to the problems of testing the accuracy at the stages of the sampling process and sampling preparation. The creation of corresponding reference materials is one of the ways to solve this problem. The BCR and IUPAC have considerable experience in this field. The development of the international programme, e.g., within the frame of CITAC, could be the basis for promoting reliability of chemical analysis for an agreed set of complex substances. At the same time, the problem of increasing the accuracy of the concentration measurement itself should not be ignored.In this case, collaboration is necessary between national metro- logical laboratories that have at their disposal the approved absolute methods of analysis. Such a collaboration could be undertaken under the auspicies of BIPM. As to CCQM, its immediate task seems to be the solution of the problem of the position of the mole in the SI system. References 1 Lyons, J., and Quinn, T. J., Proc. Verb. CIPM, 1993, 61, 122. 2 Walker, R. F., Anal. Proc., 1994, 31, 193. 3 King, B., Analyst, 1993, 118, 587. 4 Komar, N. P., Izv. VUS’ow: Khim. Khim. Tekhnol. (Proceedings of Technical Universities: Chemistry and Chemical Technology), 1975, 18(3), 343. Juhasz, E., Meres Automat., 1976, 24(3), 107.5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Kochsiek, M., Die SI-Basiseinheiten. Definition, Entwickiung, Realiesierung, Physikalisch-Technische Bundesanstalt, Berlin, 1994, p. 13. Girard, G., Metrologia, 1994, 31(4), 317. Quinn, T. J., Metrologia, 1994, 31(4), 337. Armitage, D. R., in Proceedings of 2nd. Conference on Weighting, Calibration and Quality Standards in the 1990s, ed. Buckley, M. J., South Yorkshire Trading Standards Unit, Sheffield, 1992, p. 1. IUPAC, Recommendations 1984, Definition of pH-Scales, Standard Reference Values, Measurement of pH and Related Terminology, Pure Appl. Chem., 1985,57,531. Physicochemical Measurements: Catalogue of Reference Materials from National Laboratories, ed.Cali, J. P., Pure Appl. Chem., 1976, 48, 505. Recommended Reference Materials for the Realization of Physico- chemical Properties, ed. Marsh, K. N., Blackwell, Oxford, 1987. Danzer, K., Than, E., and Molch, D., Analytik, Systematischer Uberblick, Geest & Portig, Leipzig, 1976. Alexandrov Yu. I., in Teoretitcheskaya Metrologiya, ed. Schischkin, I. F., Standard Publishing House, Moscow, 1990, ch. 6. Richter, W., Dube, G., Keyser, U., and Spitzer, P., PTB-Mitt., 1994, 104(5), 312. Thermodynamics. Basic Notions. Terminology. A Book of Dejini- tions, Science Publishing House, Moscow, 1984, N103, p. 10 (in Russian). Chemie Lexikon, Rompp, 1993, p. 2333. Alexandrov, Yu. I., Meas. Tech., 1990, 33, 1246. MacCart, P., J. Chem. Educ., 1983,60(3), 187. Cali, J. P., Mears, T. W., Michaelis, R. E., Reed, W. P., Seward, R. W., Stanley, C. L., Yolken, H. T., and Ku, H. H., The Role of Standard Reference Materials in Measurement Systems, NBS Monograph No. 148, National Bureau of Standards, Washington, DC, 1975. Uriano, G. A., and Gravatt, C. C., CRC Crit. Rev. Anal. Chem., 1977, 6(4), 361. Kozlov, M. G., Konopelko, L. A., and Nechljudov, I. B., Meas. Techn., 1990 (7), 58. Proskurnlkov, A. A., Levtchenko, Yu. N., and Lakhov, V. M., Metrological Ensuring of Measurements of Substance Composition in Chemistry by Using Standard Samples4rief Review, Standard Publishing House, Moscow, 1987 (in Russian). Fourth International Symposium on Biological and Environmental Reference Materials (BERM-4), Fresenius’ J . Anal. Chem., 1990, 338(4). Kane, J. S., Analyst, 1993, 118,953. Alexandrov, Yu. I., Meas. Tech., 1991, 34, 98. Patterson, K. Y., Veillon, C., and O’Haver, T. C., Anal. Chem., 1994, 66( 1 S), 2829. De Bievre, P., Valkiers, S., Schaefer, F., Peiser, H. S., and Seyfried, De Bievre, P., Fresenius’ J . Anal. Chem., 1994, 350(4-5), 277. De Bievre, P., Anal. Proc., 1993, 30(8), 328. Fasset, J. D., and Paulsen, P. J., Anal. Chem., 1989, 61(10), 643A. Uriano, G. A., Bundesanst. Materialpruf., 1980, 7. Kaiser, H., Spectrochim. Acta, 1978, 33, 551. Dewjatych, G. G., Kowalev, I. D., and Jankov, S. U., High-Purity Subst., 1990, (4), 128. Analytical Methods Committee, Analyst, 1995, 120, 29. Gorbazevich, S. V., Dolinskij, E. F., and Judin, M. F., Meas. Tech., 1971, (7), 63 (in Russian). Adamson, A. W., J. Chem. Educ., 1978,55( lo), 634. Alexandrov, Yu. I., Meas. Tech., 1989, 32, 82. Fleming, J. W., Anal. Proc., 1994, 31, 249. Quantities and Units. Part 8. Physical Chemistry and Molecular Physics, ISO-31-8, 1992, No. 2, 139. Rocha-Filho, R. C., J. Chem. Educ., 1990 (2), 139. Alexandrov, Yu. I., and Yudin, M. F., Metrologia, submitted for publication. Paper 6100902F P., PTB-Mtt., 1994, 104(4), 225.

ISSN:0003-2654    

DOI:10.1039/AN9962101137

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol, I21 () 40- a s 2 0 - g 6 0 - - 2 0 - 1147 ERRATUM Conducting Polymers and the Bioanalytical Sciences: New Tools for Biomolecular Communications A Review S. B. Adeloju and G. G. Wallace Analyst, 1996,121, 699 On page 702, Fig. 2 should have appeared as printed below. The printed figure, as supplied by the authors, was not the intended one. - 4 0 ' , I I I I 1 - 0 . 4 - 0 . 2 0.0 0.2 0.4 0.6 E N

ISSN:0003-2654    

DOI:10.1039/AN9962101147

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, August 1996, Vol. 121 I149 CUMULATIVE AUTHOR INDEX JANUARY-AUGUST 1996 Abdel-Aziz, Mohamed Shafei, Abraham, Michael H., 51 1 AbramoviC, Biljana F., 401, 425 AbramoviC, Borislav K., 401, 425 Abroskin, Andrei G., 419 Acedo Valenzuela, M. I., 547 Adam, S., 527 Adams, Freddy, 1061 Adeloju, S. B., 699 Aheme, G. Wynne, 329 Akhtar, M. Humayoun, 803 Al-Othman, Rashed, 601 Alazard, S., 527 Aldridge, Paul K., 1003 Alegret, S., 959 Aleixo, Luiz M., 559 Alexandrov, Yu. I., 1137 Almirall, J., 959 Analytical Methods Committee, Andrdde, Francisco J., 613 Angeletti, R., 229 AntonijeviC, M. M., 255 Appleton, Mark, 743 Aratake, Sachiko, 325 Araujo, Pedro W., 581 Arias, Juan Jose, 169 Artjushenko, Slava, 789 Bacci, Mauro, 553 Baggiani, Claudio, 939 Balasubramanian, N., 647 Bannon, Thomas, 715 Barlett, Philip N., 715 Barnabas, Ian J., 465 Barroso, C.G., 297 Barwick, Vicki J., 691 Baxter, Douglas C., 19 Baxter, Pamela J., 945 Baya, Maria P., 303 Benedetti, A. V., 541 Benmakroha, Yazid, 521 Biancotto, G., 229 Bilitewski, Ursula, 119, 863, 877 Birch, David J. S., 905 Bjorklund, Erland, 19 Blais, Jean-Simon, 483 Blanco, Marcelo, 395 Bogan, Declan R., 243 Bond, Alan M., 357 Borah, Lakhimi, 987 Boswell, Stephen M., 505 Bouhsain, Zouhair, 635 Boukortt, Sheriffa, 663 Boutelle, Martyn G., 761 Boyd-Boland, Anna A., 929 Boyd, Damien, IR Branica, Marko, I127 Brereton, Richard G., 441, 575, 581, 585,651, 993 Brinkman, Udo A. Th., 6 I , 1069 Bni, E. R., 297 Buchet, Jean-Pierre, 663 Burgot, Jean-Louis, 43 Bye, Ragnar, 201 Cai, Xiaohua, 965 Callejdn Mochon, M., 681 Cammann, K., 527 Campillo, Natalia, I043 Cao, Zhong, 259 Carbonnelle, Philippe, 663 Cardoso, A.A., 541 Cardwell, Terence J., 357 Carmona, Pedro, 105 Casajus, Rocio, 813 Casella, Innocenzo G., 249 Cassidy, Richard M., 839 CaviC-Vlasak, Biljana A., 53R Cazemier, Geert, 1 1 11 1079 573, 889 Cela, R., 297 Cepas, Juana, 49 Ceramelli, Giuseppe, 219 Cerdh, A., 13 Cerdh, V., 13 Chan, Wing Hong, 531 Chatergoon, Lutchminarine, 373 Chen, Guo Nan, 37 Chiou, Chyow-San, 1107 Chou, Shu-Fen, 71 Christian, Gary D., 601 Christie, Ian, 521 Cirovic, Drdgan A., 575, 581 Coello, Jordi, 395 Cole, S. Keith, 495 Collier, Wendy, 877 Copeland, D. D., 173 Corbella Tena, R., 459 Corti, Piero, 2 19 Cosano, J., 83 Craston, Derek H., 177 Crosby, Neil T., 691 Croteau, Louise G., 803 Crump, Paul W., 871 Crumrine, David S., 567 Cruz Ortiz, M., 1009 Cuculii., Vlado, 1127 Cullen, Michael, 75 Cullen, William R., 223 Daenens, Paul, 857 Daghbouche, Yasmina, 103 1 Dalene, Marianne, 1095, 1101 de Jong, Dirk, 61 de Jong, Gerhardus J., 61 de la Guardia, Miguel, 635, 923, de Lacy Costello, Benjamin P.J., de Oliveira Neto, Graciliano, 559 Dean, John R., 465 Demidova, M. G., 489 Demir, Cevdet, 651, 993 Deng, Jiaqi, 971 Deng, Qing, 1123 Deng, Zhiping, 671 Desai, Mohamed, 521 Desimoni, Elio, 249 Destradis, Angelo, 249 Devi, Surekha, 807 Dey, Nibaran C., 987 Dilleen, John W., 755 Dobrowolski, R., 897 Dodd, Matthew, 223 Dolmanova, Inga F., 431 Dong, Shaojun, 1123 Dreassi, Elena, 219 Dumasia, Minoo C., 651 Dumschat, C., 527 Dunemann, Lothar, 845 Dunhill, Roger H., 1089 Economou, Anastasios, 97, 1015 Eigendorf, Guenter K., 223 Eikenberg, Oliver, 119 El-shahat, Mohamed F., 89 El-Shorbagi, Abdel-Nasser, 183 Elbergali, Abdallah K., 585 Ellwood, Jo A., 575 Emara, Samy, 183 Emteborg, Hskan, 19 Endo, Masatoshi, 39 1 Eng, Jimmy, 65R Escobar, Rosario, 105 Essers, Martien, 11 11 Evans, Phillip, 793 Facer, M., 173 Fallon, Michael G., 127 Fang, Kai-Tai, 1025 Fawaz Katmeh, M., 329 Feam, Tom, 275 103 1 793 Fell, Gordon S., 189 Fernandes, Julio Cesar B., 559 Ferreira, Valdir S., 263 Fielden, Peter R., 97, 1015 Fillenz, Marianne, 761 Fitzgerald, Catherine A., 715 Fleet, lan A., 55 Forster, Robert J., 733 Forteza, R., 13 Francis, John M., 177 Frech, Wolfgang, 19, 1055 Fugivara, C. S., 541 Fukdsawa, Tsutomu, 89 Fung, Yingsing, 369 Gail, Ferenc F., 401, 425 Gala, Belkn, 1133 Galeano Diaz, T., 547 Gamble, Donald S., 289 Gao, Xiao Xia, 687 Garcia, M., 959 Garrigues, Salvador, 635, 923, Genrich, Meike, 877 Georgieff, Michael, 901 Ghosh, Anil G., 987 Giannousios, A., 413 Giersch, Thomas, 863 Giraudi, Gianfranco, 939 Glennon, Jeremy D., 127 Godinho, Oswaldo E.S., 559 Goldstein, Steven L., 901 G6mez-Hens, Agustina, 1133 Gong, Zhilong, 1 119 Gooijer, Cees, 1069 Goosens, Elise C., 61 Gordon, Derek B., 55 Goto, Nobutake, 1085 Greer, James C., 715 Grol, Michael, 119 Groves, John A., 441 Guiberteau Cabanillas, A., 547 Guiraum Pkrez, A., 681 Gurden, Stephen. P., 441 Haasnoot, Willem, 11 1 1 Hadjiivanov, K., 607 Halliwell, David J., 1089 Hammerich, Ole, 345 Hansen, Elo H., 31 Harris, Roy, 913 Harrison, Iain, 189 Hart, Barry T., 1089 Hartnett, M., 749 Hauser, Peter C., 339 Hayashi, Yuzuru, 591 Hayashibe, Yutaka, 7 Hays, Lara, 65R Hendrix, James L., 799 HemBndez-Cordoba, Manuel, Hernhndez, Oscar, 169 Hindmarch, Peter, 993 Hu, Yan, 883 Hulanicki, Adam, 133 Hyland, Mark, 705 Ibrahim, Naaim M.A., 239 Idriss, Kamal A., 1079 tnagawa, Jun, 623 IAiguez, Montserrat, 1009 Ioannou, Pinelopi C., 909 Irwin, G. W., 749 Ishida, Yasuyuki, 853 Ishihara, Masahito, 391 Isomura, Shinichi, 853 Iturriaga, Hortensia, 395 Ivanova, Elena K., 419 Iwatsuki, Masaaki, 89 Jackson, Laurence S., 67 Jaselskis, Bruno, 567 Jiang, Chongqiu, 3 17 1031 1043 Jimknez, Ana Isabel, 169 Jimknez, Francisco, 169 Jimknez-Prieto, Rafael, 563 JimCnez Sanchez, J. C.. 681 Johnson, Mark, 1075 Jurkiewicz, M., 959 Kalish, N.K., 489 Karayannis, Miltiades I., 435 Karlsson, Lars, 19 Kettling, Ulrich, 863 Kimbrough, David Eugene, 309 Kimoto, Takashi, 853 Kindness, Andrew, 205 Knoll, M., 527 Kolotyrkina, Irina Ya., 1037 Konstantianos, Dimitrios G., 909 Korda, T. M., 489 KoLik, Andrzej, 333 Kratochvil, Byron, 163 Kuznetsova, Vera V., 419 Kvasnik, Frank, I 1 15 Kwong, Daniel W. J., 531 Lan, Zhang-Hua, 21 1 Lancashire, Susan, 75 Lancia, Antonio, 789 Laurie, David, 951 Lawrence, Chris M., 755 Le, Quyen T. H., 105 1 Lee, Albert W. M., 531 Legouin, Bkatrice, 43 Lei, Chenghong, 971 Lewenstam, Andrzej, 133 Li, Hao, 223 Liang, Yi-zeng, 1025 Lightbody, G., 749 Lin, Hui-Gai, 259 Link, Andrew J., 65R Lipkovska, N. A., 501 Lison, Dominique, 663 Littlejohn, David, 189 Lonardi, S., 219 Lopes, Teresa 1.M. S., 1047 L6pez Carreto, Maria, 33R L6pez-Cueto, Guillermo, 407 Lopez-Erroz, Carmen, 1043 Lopez, Martin, 905 Lord, Gwyn A., 55 Loukas, Yannis L., 279 Lowry, John P., 761 Lowthian, Philip, 743, 977 Lowy, Daniel A., 363 Lu. Bin, 29R Lu, Changyin, 883 Lu, Xiao-Quan, 1019 Lu, Zheng, 163 Lund, Walter, 201 Luo, Yongyi, 601 Luque de Castro, M. D., 83 Lyons, Michael E. G., 715 McAdams, Eric T., 705 McAlemon, Patricia, 743 McAteer, Karl, 773 McCormack, Ashley L., 65R MacCraith, Brian D., 785, 789 McDonagh, Colette M., 785 McEvoy, Aisling K., 785 McKelvie, Ian D., 1089 MacLachlan, John, 11R McLaughlin, James A., 705 McNaughtan, Arthur, I1R Madsen, Gary L., 567 Magdic, Sonia, 929 Maines, Andrew, 435 Maj-Zurawska, Magdalena, 133 Malahoff, Alexander, 1037 Mannaert, Erik, 857 Marr, Iain L., 205 Marshall, William D., 289, 483, 8171150 Ariulyst, August 1996, Vol.121 Martin. Patricia. 495 Martinez-Fhbregas. E.. 959 Martinez-Lozano. Carmen. 477. Mason. Andi-cw J., 95 1 Msspoch, Santiago, 395. 407 Masu.jima, Tsutomu, 183 Mathiasson, Letinart. 19 Matsuda, Rieko, 59 1 Matsui. Masakazu, 105 1 Meaney, Mary. 789 Melbourne, Paul, 1075 Melios. Cristo B.. 263 Micczkowski, Jozef, 133 Miewwa. J., 897 Miha.jlovic. R., 355 Milai.ii.. Radmila. 627 Mills, Andrew, 535 MilosavI.jevi6, Emil B., 799 Mitrovik. Bojan, 627 Mizgunova. Ulyana M.. 43 1 Mo. Jin-Yuan. 1OI9 Mo, Songying, 369 Moanc, Siobhan, 779 Mocak, Jan. 357 Mohamed. Ashraf A., 89 Molina, Marina, 105 Monaf. Lela, 535 Monaghan. Joht Moollan, Roland W., 233 Moore, Andrew, 67 Moscllo.R.. 83 MotomiLu. Shoji, I OX5 Mottola, Noracio A., 21 1. 38 I Mounsey. Andrew. 955 Mulcahy. David. 127 Muller, Beat, 339 Munro, C. H., 835 Murphy. William S.. 127 Nakatnura, Masntochi, 469 Nakamura, Motoshi. 469 Nnkanishi. Masami. 853 Newton, R.. 173 Nie. Lihua. 883 Nielsen. Stel'fen, 3 1 Ndte. Joachim, 845 Norelia-Franco, Luis E.. I 1 15 Norris. Timothy, 1003 Obcndorf, Dagmar, 35 1 Obradovik, Danilo M., 401 Odman, Fredrik, 19 Ohtani. Hajime, X53 O'Keefle. Michael, 779, I R O'Kennedy. Richard, 243, 767, Olmi, Filippo, 553 Oms. M. T., 1.3 O'Neill, Robert D., 761, 773 Oniciu. Liviu. 363 Orlnndo, AndreLi. 553 Oshima. Mitsuko, I 085 Osipova, Nataliya V., 4 19 Ostaszewska, Joaitna. 133 Owen. Susan P., 465 Packham. Andrew J., 97, 1015 Papadopoulos, C.. 3 13 Paradowski, Dariusz, I33 Pardue.Harry L., 385 Parsons. Patrick J.. I95 813 Montelongo, F. ,459 29R Palel, Sunil U., 913 Patterson. Kristine Y., 983 Paulls. David A., 83 I Pawliszyn. JanusL R.. 929 Pedt-cro. Maria, 345 Perez-Bendito, Dolores. 49, 563, Pet-ex-Bustamante, J. A., 297 Pkrcz-Ponce. Amparo, 923 PhL-Ruiz, Tomas. 477, 8 13 Pet-gantis, Spiros A., 223 Perruccio, Piero Luigi, 2 19 Piggott, Nighel H., 951 Pihlar, Boris, 627 Pingarron, Jose, 345 Piperaki, Efrosini A.. I I I Piro, R. D. M., 229 Pitre, K. S., 79 Poc, Russell B.. 591 Poole. Colin F.. 5 I 1 Power, J. F.. 451 Prodrotnidih, Mamas I.. 435 Proinova, I., 607 Proskurnin, Mikhail A,, 419 Pyrzytiska, Krystyna, 77R Qu. Yi Bin, 139 Quevauviller. Ph., 83 Quinn, John G.. 767 Rader, W. Scott, 799 Rae, Bruce. 233 Raghunath.A. V., 825 Rahmani. Ali, 585 Kamanaiah. G. V., 825 Rangel, Anthio 0. S. S . , 1047 Ratcliffe, Norman M., 793 Razec, Saeid, 183 Redon. Miguel, 395 Regan. Fiona, 789 Reimcr. Kenneth J.. 223 Reinartz, Heiko W., 767 Rigby, Geraldine P., 871 Rios, Angel, I Rodriguez Delgado. M. A.. 459 RodrigueL-Medina, JosC F., 407 Rohm, Ingrid, 877 Rowell, Frederick J., 95 1 ~ 955 Rowell. Vibeke, 955 Romdom, Eduard J. E., 1069 Rubio. Soledad, 33R Kuzicka, Jaromir, 601, 945 Sadler. Peter J.. 9 13 Sakslund. Henning, 345 Salden, Martin. I I 1 1 Saleh. Ciamal A., 641 Salinas. F., 547 San Martin Fernandez-Marcote, Slinchez, M-1. J., 459 Santamaria. Fernando, 1009 Santos. Jose H., 357 Sanz. Antonio, 477 Sarabia, Luis A., 1009 Sartini. Raquel P.. 1047 Sasaki. Takayuki, 105 1 Sato, Hidetoshi, 325 Satyanarayana, K., 825 Sayama.Yasumasa, 7 Schafer, E. A., 24.3 Schieltz. David, 65R Schmid. Rolf D.. 863 1133, 33R M., 68 I Schoppenthau, Jiirg. 845 Scobbie, Emma, 575 Scudder. Kurt, 945 Sedait-a, I-lassan, 1079 Seibert, Donna S., 51 1 Sekino, Tatsuki, 853 Seviour, John, 951 Shah, Rupal. 807 Shanthi, K., 647 Shih, Jeng-Shong, 1107 Shijo, Yoshio, 325 Shiraishi. Haruki, 965 Shpigun. Lilly K., 1037 Shukla, Jyotsna, 79 Shulman. R. S., 489 Silva, Manuel, 49, 563 Siskos. Panayotis A., 303 Skarping, Gunnar, 1095, 1 101 Slater, Jonathan M., 743, 755 Slavin. Walter, 195 Sloth, Jens J . , 31 Smith. Clayton, 373 Smith, Dennis C., 53R Smith, Robert E., 67 Smith, Roy, 321 Smith, W. E., 835 Smyth, Malcolm R., 779, IR, 29R Srnythe-Wright, Denise, 505 Snell, James P., 1055 Sokalski, Tomasz, 133 SolC.s., 959 SolujiC, L.jiljana, 799 Somsen. Govert W., 1069 Spanne. Mrirten, 1095, I 101 Stathakis, Costas, 839 Stegman, Karel H., 61 Stegmann, Werner, 901 Stevenson. Derek, 329 Stone, David C., 671 Stouten, Piet, 11 1 I Strachan, David, 95 I , 955 Stradiotto, Nelson R.. 263 Streppel, Lucia, 1 1 1 1 Stuart, Iain A., 11R Stubauer, Gottfried, 35 I Subramaniatn, K., 825 Suffer, I. H. 'Mel', 309 Sukhan, V. V., 501 Suliman, Fakhr Eldin 0.. 617 Sultan, Salah M., 617 Sumod.jo, P. T. A., 541 Susanto, Joko P., 1085 Sweedler. Jonathan V., 45R Symington: Charles, 1009 Szklar. Roman S., 321 Tam: Wing Leong, 531 Tan, Yanxi, 483 Tang, Bo, 3 17 'l'ang, Shida. 195 Tegtmeier, M., 243 TepavEeviC, Sanja D., 425 Thastrup, Ole. 945 Thomaidis, Nikolaos S., 1 I 1 Thomassen, Yngvar, 1055 Thompson, Michael, 275.285, 67 1.977, 53R Thornes, R . D . , 243 Tian. Baomin, 965 Tiinperman, Aaron T., 45R Tinnerberg, Hgkan. 1095, 1 101 Tomas, Virginia, 477, 813 Torgov, V. G., 489 Townshend, Alan, 83 1 Troccoli, Osvaldo E., 6 13 Tsuge, Shin, 853 Tsurubou, Shigekazu. 105 1 Tudino, Mabel B., 613 Tyson, John D., 95 I , 955 T7ou w ara- K aray ann i , S te 1 la M . , Ubidc, Carlos, 407 Uehara, Nobuo, 325 Umetani, Shigeo, 105 1 Vadgama, Panka-j, 435, 521. 871 Vagg,elli, Gloria, 553 Valcarcel, Miguel, I , 83 Van Mol, Willy, I061 van Wichen, Piet, I1 1 1 Vassileva, E., 607 Veillon. Claude, 983 Velthorst, Nel H., 1069 Verbeek. Alistair, 233 Viles, John H., 9 I3 Villegas, Nuria, 395 Viiias. Pilar, 1043 Vos. Johannes G., 789 Vukanovii, B., 255 Walker. P. J., 173 Wallace, G. G., 699 Walsh, James E., 789 Walsh, Peter T., 575 Wang. Bin-Feng, 259 Wang. Chen, 3 17 Wang. Jin, 289, 817 Wang, Joseph, 345,965 Wang, Ke-Min, 259, 531 Wang, Shi-Hua, 259 Watanahe. Kazuo, 623 Wheals, Brian B., 239 White, P. C., 835 Whiting, Robin, 373 Wickstram, Torild, 201 Wilmot. John C.. 799 Wittmann, Christine, 863 Wood, Roger, 977 Woolfson, A. David, 71 1 Wu, Weh S., 321 Xin, Wen Kuan, 687 Xu, Xue Qin, 37 Xu. Yuanjin, 883 Yaniada, Shinkichi, 469 Yan, Xiu-Ping, 1061 Yao, Shouzhuo, 883 Yates 111. John R., 65R Yu, Ru-Qin, 259 Zagatto, Elias A. G., 1047 Ziinker, Kurt, 767 Zanoni, Maria Valnice B.. 263 Zaporozhets, 0. A,, 501 Zhang, Fan, 37 Zhang., Xiaogang, 3 I7 Zhang, Zhanen. 97 1 Zhang, Zhujun. 1 1 19 Zhi, Zheng-liang, 1 Zhou. Dao-Min, 705 Ziegler, Torsten. 119 Zolotova, Galina A., 43 1 435

ISSN:0003-2654    

DOI:10.1039/AN9962101149

出版商:RSC    

年代:1996

数据来源: RSC