摘要:

Auzal.yst, October 1996, Vol. I21 (1485-1488) 1485 Analytical Performance Testing of an Atrazine lmmunoassay System Sean D. W. Combera, Chris D. Watts“ and Barbara Youngb a WRc Medmenham, Henley Road, Medmenham, Marlow, Buckinghamshire, UK SL7 2HD Millipore Inc., Bedford, M A 01 730, USA A rigorous performance evaluation of an enzyme immunoassay (EIA) kit for the determination of atrazine in water samples was undertaken. Eleven individual batches of samples containing standards and spiked drinking waters were analysed and precision, bias and limit of detection were measured using statistical analysis. The technique was shown to be capable of achieving performance criteria (a total standard deviation of less than 5% or 2.5 ng, whichever is the greater) demanded of modern analytical systems and achieved a limit of detection of 9.2 ng 1-1.There was no statistically significant bias measured for drinking water samples. Interference tests showed that the atrazine immunoassay was not significantly affected in the pH range 4.0-8.0 or by drinking water matrix components (anions, cations and chlorination by-products), even at their maximum allowable concentrations. There was a small extent of cross-reaction with simazine and atrazine degradation products, but given the persistence of atrazine, through its resistance to hydrolysis, breakdown products are likely to be present at much lower concentrations than the parent compound in drinking water. Simazine may potentially be more problematic, so it would be prudent to monitor a proportion of samples for simazine to determine the extent to which this may be contributing to the ‘atrazine’ measured in drinking water samples using the EIA kit.Keywords: Atrazine; enzyme immunoassay; drinking water; tap water Introduction Enzyme immunoassay (EIA) has been used for many years in the field of biochemistry for the qualitative and quantitative analysis of numerous compounds for which antibodies can be raised. More recently, more sensitive immunoassay kits have been developed for analytical purposes, and in particular, for the determination of pesticides in natural waters (see, for example, refs. 1-3). Immunoassay has certain advantages over conven- tional instrumental methods such as GC and MS by being relatively cheap, simple to use and offering the possibility of use in the field.In addition, only small sample sizes are used and no solvents are needed, which offers a distinct advantage over conventional analytical methods that require the purchase, storage, handling and disposal of large amounts of solvents, with all of the associated health and safety risks. Immunoassay has therefore found a niche market for the determination of pesticides at relatively high concentrations and also for general screening purposes. The main principle of immunoassay (IA) is the biochemical reaction between an antibody and an antigen, which for pesticide analysis is a pesticide. The IA kit is supplied with antibodies which have been raised to react selectively with compounds that resemble the specified pesticide, which in this case is atrazine.The antibodies are prepared by the IA kit manufacturer through stimulation of a mammalian immunosy- stem by a pesticide-hapten compound, which combines the functional group of the pesticide and the high molecular mass required to stimulate antibody formation. Once prepared in this manner, the antibodies will respond to atrazine. The kit also includes a labelled form of the pesticide, often termed the ‘conjugate.’ The label is normally an enzyme, hence enzyme- linked immunosorbent assay (ELISA). In the assay, atrazine in the water sample competes with the conjugated atrazine for a limited number of binding sites on the specific antibodies, which are immobilized on the walls of a small cell. The presence of unlabelled atrazine in the sample results in less label being bound to the antibody in the first stage of the assay.In the next step, unbound atrazine and conjugate are removed through washing steps. Finally, the amount of bound labelled atrazine is determined by reaction of the enzyme with a reagent (termed ‘chromogen’) and photometric determination of the product. A consequence of this principle is that the response is inversely proportionally to the amount of determinand in the water sample. Some of the earlier IA kits for atrazine determination were highly susceptible to cross-reactivity with compounds of similar structure (e.g., other triazines and degradation products of atrazine), although for certain screening purposes this may have been an advantage. Recent developments. however, have improved both the sensitivity and selectivity of immunoassays to the extent that some commercially available kits now offer a performance similar to that obtainable from more conventional analytical techniques.3 In most cases, the performance testing carried out on ETA kits has not been extensive and more rigorous testing would provide a better indication of how well they compare with conventional analytical methods.This paper describes the performance testing of a commercially available atrazine EIA kit for precision, bias, limit of detection and the effect of interferents (e.g., pH, anions, metals, cations, breakdown products and surfactants), using an automated 1A instrument (Bio-Tek ELs I 000). Experimental Apparatus An automated Bio-Tek ELs 1000 immunoassay analy ser combined with Envirogard high-sensitivity atrazine immu- noassay kits (Millipore, Bedford, MA, USA) was used to perform all of the tests.Chemicals All IA chemicals (conjugate, substrate and stop solutions) were supplied by Millipore. Distilled, de-ionized water (DDW) was used throughout the performance testing. Two different sources of atrazine standards were used for the testing (Millipore and1486 Analyst, October 1996, Vol. I21 British Greyhound Chromatography and Allied Chemicals, Birkenhead, UK). Working standard solutions used in the analysis were prepared freshly on a weekly basis from concentrated stock standard solutions. The stock standard solutions were stable over the duration of the testing ( I month) and the working standard solutions showed no sign of deterioration over the course of 1 week.Performance Testing Performance testing was undertaken using a protocol previously described,s which has been adopted by the Drinking Water Inspectorate as the water industry standard. The following regime was used for the analytical system performance tests. A ‘normal’ calibration was run, and then duplicate determi- nations of the following test samples: (i) blank sample; (ii) a standard solution at 10 ng 1- I ; (iii) a standard solution at 90 ng I-’; (iv) drinking water at 20 ng 1 - I ; and (v) drinking water at 80 ng 1-I. Note: although the standard solutions (ii) and (iii) are described as ‘standards’ (i.c.., they are prepared in a blank matrix), they should more properly be thought of as ‘test samples’.i.e., they are not standards in the sense of being used in any way for calibration purposes. These samples were prepared freshly for each batch of analysis from a different stock standard solution. The degree of imprecision caused by sample preparation was known (before the tests) to be small in relation to the analytical variation, by a within-batch compari- son of one sample with several separately prepared portions of nominally the same sample. Samples (iv) and (v) were spiked with the same stock as samples (ii) and (iii), then analysed in duplicate (with results expressed as the mean of the two determinations) for 1 1 batches of analyses, with the objective being to test the method performance with respect to the drinking water quality standard of 100 ng 1- I .A batch is defined as a set of results for which a given calibration is applied (see above). Each duplicate determination made on the above test samples was also duplicated. This allowed the effect on precision of defining a test result as a mean of two measurements to be assessed. Interference Testing A series of interference tests were undertaken to examine how the method performed under extreme conditions where other water solutes, which may be interferents, were present at the UK’s Maximum Allowable Concentration (MAC) in drinking water. DDW samples were spiked with each set of interferents and 0,20 and 80 ng I-’ of atrazine; a set of samples without the interferent present was also run in each batch for comparative purposes. The following solutions were prepared: Metals (MAC) Anions Cations Humic acids Chlorination by-products S urfactants PH Atrazine degradation products Other triazines (used in UK) Fe 0.2 mg 1-1, Cu 3 mg 1 - 1 , Ni 50 pg 1-1, Zn 5 mg 1-1, Mn 50 pg I-’ C1- 200 mg 1-1, N03Z- 50 mg 1-I.Sod2- 250 mg 1-1 Ca2+ 250 mg 1-1, Mg*+ 50 mg 1-’ Extracted River Thames water at 1 mg-I, pH 4.8 and 20 mg I-I, pH 4.5 Chloramine mixture 50 pg 1-I, chloroform 50 pg 1- l , trichloro- acetic acid and dichloro- acetic acid 50 pg I-’ Linear alkyl sulfonates and 4-nonylphenol at 200 pg 1-I pH 4.0, 6.0, 8.0 Hydroxyatrazine, desethyl- atrazine, deisopropylatrazine at 0.1 pg I-’ Simazine (0.1 pg 1-I) The mean and standard deviation of each solution with and without the interferent present (four replicates) and the difference between them, were calculated.Using the f-test at the 95% confidence level and the standard deviation of the unamended sample, it was possible to calculate the mean concentration difference required to result in it being signifi- cantly different from the unamended sample. The actual mean difference was then expressed as a percentage of the theoretical value to ‘fail’ [to ‘fail’ being defined as (mean difference/ standard deviation)/to,o5]. Results and Discussion Preliminary results showed that for determinations in the range el 00 ng 1-1 of atrazine, a significantly better performance could be obtained by restricting the calibration range to 0-1 10 ng 1-1, rather than the 0-500 ng 1-1 range recommended by instructions included with the kit. At concentrations above I 10 ng 1-1, the slope of the calibration curve was very shallow with a consequent decrease in precision.Cali bration using solutions -~~ - - Table 1 Precision data for atrazine“ Atrazine addedfng I-‘ DDW DDW DDW Drinking Drinking Drinking Parameter +0 +10 +90 water +0 water +20 water +80 Over-all mean 1.51 S within I .80 S between 2.75 S total 3.29 Fo 0 5 1.72 FC.,,‘ 1.73 Estimated d.f. I 3 Limit of detection 9.16 ng I-’ Precision Target S total 2.5 assessment 10.28 1.19 0.84 1.46 2.5 1.60 0.34 18 Pass 94.64 2.21 4.33 4.86 4.7 1.75 1.05 12 Pass -0.74 1.22 1.71 2.09 2.5 1.72 0.70 14 Pass 18.86 2.2s I .66 2.80 2.5 1.60 1.25 18 Pass 80.05 4.47 1.18 4.62 4.0 1.57 1.33 21 Pass * S within = within-batch standard deviation; S between = between-batch standard deviation; S total = total standard deviation; target S value calculated as 2.5% of mean or 2.5 ng, whichever is the greater; Fo.os taken from appropriate statistical tables; Fcalc derived from equation in Gardner;5 estimated d.f.= Estimated degrees of freedom; precision assessment = Pass if Fcalc < Fo.os.Analyst, October IY96, Vol. I21 1487 containing 0, 10,40,60, 100 and 1 10 ng I- I of atrazine ensured good precision across the whole range, whilst still encompass- ing the concentration of interest (the maximum allowable concentration of an individual pesticide in drinking water of 100 ng I--'). The data were processed using statistical tests of precision, bias and recovery as described by Gardner.5 Performance Tests Results obtained for the precision tests are displayed in Table 1.All concentrations were calculated using Kineticalc software installed on the ELs 1000. The results in Table I show that the method met the performance requirements of having a total standard deviation on spiked samples of not significantly worse than 5% or 2.5 ng, whichever is the greater. The limit of detection was 9.2 ng I - ' , which also satisfied the performance requirement of less than 10 ng l-I..The atrazine assay kit did not exhibit any significant bias (recoveries were 100% f standard deviation) for drinking water spiked at both the 20 and 80 ng I-' level5 (Table 2). Interference Tests The results of calculation of mean differences between measured concentrations of atrazine in samples with and without a particular solute are presented in Table 3. A value of less than 100% demonstrates that the amended sample (with a particular solute or mixture of solutes added) is statistically indistinguishable from the unamended sample.Any value above 100% signifies that the presence of a potential interferent causes a bias in the analytical result. Obviously, the greater the number, the more effect the interferent has on the analysis. The matrix components typically found in a drinking water (e.g., anions, cations and chlorinated by-products) do not cause a significant effect even at their MAC. There is a similar lack of interference caused by varying the pH between 4 and 8, which encompasses most UK drinking waters. Very high levels of humic acids cause an interference effect ( 10 ng 1- I positive bias) particularly at lower analyte concentrations, but such high concentrations of huinics are unlikely to be present in drinking waters.Such effects from humicc are well d o c i ~ m e n t e d ~ ~ ~ - ~ and may result from the inherently low pH of such samples. Only in raw upland waters are levels of humics likely to cause any problems with the assay and these may be overcome using standard addilions or pH correction (often recommended in immunoassay procedures). Although the presence o f surfactants at their MAC causes a measureable interference at low atrazine concentrations. these are extreme values which are unlikely to be encountered in most tap waters. The susceptibility of inimunoassay to cross-reactivity with chemicals of similar structure manifested itself as a positive bias for atrazine in the presence of simazine and atrazine degradation products.Data supplied by Millipore for cross-reactivity of the atrazine high-sensitivity plate kit revealed that at concentrations of 1.5. 3 and 2 pg 1-1, simazine, 6-hydroxyatra~ine and de- ethylatrazine, respectively, have been shown to produce the same response as seen for 0.1 pg 1- I atraLine. This equates to a least detectable concentration of SO ng 1- 1 for de-ethylatrazine, 40 ng 1- I lor simaLine and 7 ng 1- I for 6-hydroxyatrazine. The persistence of atrazine in water, particularly to hydrolysis, means that breakdown products should not be present at significant concenlrations. Simazine, however, is still widely used in the UK, so it would be necessary to monitor samples for simazine to ensure that concentrations are insignificant.Conclusions The use of immunoassay as an analytical tool has developed considerably over the last few years, to the stage where it is now a viable alternative to conventional chromatographic methods for certain determinands. The performance data prescnted in this paper demonstrate that this particular immunoassay kit for the determination of atrazine in drinking water is capable of providing a precision better than the target set by the UK's Drinking Water Inspectorate for drinking water analysis (i.e., total standard deviation better than 5% or 2.5 ng, whichever is greater), 100% recoveries and a sub- 10 ng I- I limit of detection. Indeed, these peformance data are better than those obtained from certain conventional analytical techniques.g.9 The method offers several other advantages: capital and running costs are less than for chromatographic and mass spectrometric equip- ment, the lack of sample pre-treatment decreases the potential for contamination, the sample sizes required are much lower and chemical solvents, with their accompanying health, safety and cost implications, are not required.The user does need to be aware of potential cross-reactions with other triazines, but the constant development of immunoassay should result in greater specificity with later generation kits. From the above data, it can be seen that immunoassay now offers the analyst an alternative to chromatographic methods for the determination of atrazine in drinking waters.The authors thank Barbara Young and Linda Dohrman of Millipore and Yves Cohnen of BioTek for their help and advice during the course of this work and Millipore for funding the work. Table 3 Results of atrazine interference tests lnterferent effect (%) Interferent At +20 ng I-' Anion, 71 Cations 95 Huniics (1 mg I-') 41 Chlorination by-products 70 Humics (20 mg 1 I ) 514 pH 4.0 92 pH 6.0 76 pH 8.0 0 S urfac t an t s 24 I Simazine 250 Atrazine breakdown products I86 Heavy metals 88 At +80 ng I-' 23 44 7 71 104 51 29 98 45 115 290 78 Table 2 Recovery data for the atrazine test Standard error on Mean mean 95 % Expected I ecovery/ recovery/ confidence recovery/ Recovery Recovery Sample ng I-' ng I-' limit ng 1-1 assessment Tap + 20 ng 1-1 19.55 2.97 1.62 19.98 97.85 Pass Tap + 80 ng 1 - 1 80.5 1 5.02 2.74 79.73 100.98 Pass1488 Analyst, 0ctohc.r- 1996, Vol. 121 References 7 Ruppert, T. W.. Weil, L., and Niessner, R., in Irrimunoassays in Food Ferguson, B. S.. Kelsey, D. E.. Fan, T. S., and Bushway, R. J., Sci. Total En\~ii.on.. 1993, 132, 415. Thuniian. E. M., Meyer, M., Pomes M., Perry, C. A., Schwab, A. P., A n d . Chcni., 1990, 62, 2043. Inmiiinoc.l7e~~iit~ul Methods for Envirwmcntul Analysis, ed. Van Emon, J. M., and Mumma, R. O., American Chemical Society, Washington DC, ACS Symp. Ser. no. 442, 1990. Watts, C. D.. and Hegarty. B., Pui.e Appl. Clicin., 1995, 67, 1533, aiid references cited therein. Gardner. M. .I., A I M L I I I ~ I U ~ 0 1 1 Analjtic~al Quulitj Conii-olfiw tfze Water. I n d r i s f i ~ j (NS-30). revised edition, WRc, Marlow, 1989. Harrison. R . O., Gee. S. .I.. and Hammock, B. D., ACS Symp. Sei.., 1988. No. 379, 3 16. Analy~is, ed. Noms, B. A,, and ClifCord, M. N., Elsevier Applied Science, London, 1992. 8 Department of the Environment, Welsh Office. Grrrduunr c on Sufegiaal-ding tlze Qitalitv of Public Water Siipplrc)c, H.M. Stationery Office, London, 1989. NAMAS. Ar i reditation Rrqiurerrwntr joi Suniplrng and TtjL\ting rn Ac ( o r danc e M rtiz the Di inXing Wutei Tcrtrng Sptlcifi( atroir (DWTS). NIS 70, NAMAS Executive, National Physical Laboratory. Tedding- ton, Middlesex. 1994. 9 P U ~ I ~ T 6102S94B Received April 24, 1996 Accepted July 2 , 1996

ISSN:0003-2654    

DOI:10.1039/AN9962101485

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, October- 1996. Vol. 121 (1489-1494) 1489 Optical Nitrite Sensor Based on a Potential-sensitive Dye and a Nitrite-selective Carrier A membrane responsive to nitrite has been developed which is composed of plasticized PVC, the anion carrier benzylbis(tripheny1phosphine) palladium(rI)chloride, and the potential-sensitive dye (PSD) rhodamine B octadecyl ester perchlorate. On exposure to nitrite, fluorescence intensity increases, while the wavelengths of both the excitation and emission maxima remain unchanged. The sensor membrane exhi bits its highest sensitivity to nitrite in the 5 to 5000 mg I-' range, and the detection limit is 0.5 mg 1-1. The signal change on exposure to 100 mmol l-1 nitrite is as high as +95%. The effect of pH is significant: from pH 5.0 to pH 9.0 and in the absence of nitrite, the fluorescence intensity changes almost linearly by around -9% per pH unit.In addition, the sensor is cross-sensitive to pH: the relative signal change from plain buffer to 1 mmol 1- nitrite is smaller by 65% at pH 9.0 than at pH 5.0. The selectivity coefficients relative to nitrite were determined by the separate solution method at pH 7.13 and were found to be 8 x 10--3 for nitrate, 1.6 x 10-3 for chloride, 8 X 10- for hydrogencarbonate, and 3 X 10-4 for sulfate. The lifetime of the sensor membrane is limited by leaching of the PSI), which is in the range of 1-3% h l . Keywords: Optode: nitrite sensor; anioii sensei-; j-luowsceiice; polcri-ity pi-ohe membrane. In order to maintain electroneutrality. a proton i \ co- extracted into the membrane, where it protonates a pH indicator dye contained in the polymer membrane.On protonation, the dye undergoes a change in either absorption or fluorescence. Respective sensors have been presented for carbonate,3 chlo- ride,4,5 iodide,5 nitrate"-8 and cyanide.') Recently, a fibre optical nitrite sensor has been described which is based on electro- polymerized cobaltporphyrin films and does not require the addition of indicator dyes for optical transduction. 1 0 We present a sensing scheme for anions that is based on our previous work on anion sensing using potential-sensitive dyes (PSDs).I1 The signal change is the result of the change of the micro-environment of the PSD at the sample-\ensor interface. and thiy is measured using a dye which ha5 optical properties that re\pond to a change in the polarity of its micro- environment.PSDs have so far been used for sensing of nitrate based on the rather unselective anion exchanger tridodecylmethyl ammo- nium chloride (TDMACI). I In this paper. the feasibility of selectively sensing anions using a PSD in combination with the nitrite carrier benzylDis(triphen~1phosphine) palladium(r1) chlo- ride (BPP) is demonstrated. A solid-state nitrite-sensitive membrane was prepared and inve\tigated in term\ of signal change, sensitivity. stability, limit\ of detection. and the selectivity for nitrite over other anion$ foirnd i n drinking water. Introduction There is a need for suitable methods for the determination of nitrite in water, owing to the important role of nitrite as a precursor in the formation of N-nitrosamines, many of which have been reported to be potent carcinogens, and its importance in indicating the level of organic pollution in water.So far, nitrite has been determined mainly via chromatographic, optical. and electrochemical methods. Chromatographic meth- ods are based on the use of ion-exchange columns for separation of anions along with detection lia UV-absorption, refractive index, or conductivity. Spectrophotometric methods are based on the formation of an azo dye by diazotation of an aromatic amine with nitrite and coupling of the diazonium cation with an aromatic amine or phenol. This scheme forms the basis for the commercially available colour test strips, but unfortunately cannot be applied to continuous sensing.In order to meet the need for continuous monitoring, electrochemical sensors have been developed. These arc based 011 the use of selective anion carriers contained in lipophilic polymeric matrices (such as plasticized PVC) coated onto the surface of a potentioinetric electrode. Anion carriers have also been used in optical approaches. In this scheme, the anion carrier extracts the anion into a sensor Experimental Chemicals Rhodamine B octadecylester perchlorate (RBQE) was obtained from Lambda Fluoreszenz-Technologie (Graz, Austria) and was used as received. PVC (high molecular weight), bis- (2-ethylhexyl)-sebacate (DOS), o-nitrophenyl octylether (NPQE) and tetrahydrofuran (THF) were obtained from Fluka AG (Buchs, Switzcrland). Poly(viny1 chloride-co-vinyl acetate- co-2-hydroxypropyl acrylate) exhibiting a vinyl content of 8 1 % m/m, a vinyl acetate content of 4% m/m and a 2-hydroxypropyl acrylate content of 15% m/m was obtained from Aldrich (Steinheim, Germany).The nitrite selective carrier BPP was from Aldrich. Unless stated otherwise, 20 mmol I-' sodium phosphate buffer of pH 7.13 was used as the plain buffer and sodium salts of nitrite, nitrate, chloride, hydrogencarbonate, and sulfate were added. Ail buffer components were of analytical grade. Double-distilled water was used throughout. Preparation of the Nitrite-sensitive Membrane A solution was obtained by dissolving 2.5 mg of PVC or PVC copolymer, 5.0 mg of the plasticizer. 0.3 mg of RBOE and the respective amount of BPP in 1.5 ml of THF (Table I).A dust-I490 Anulyst, October- 1996, Vol. I21 free 12 X 50 mm 175 pm polyester foil (Mylar, type GA- 10, Du Pont, Wilmington, DE, USA) was placed in a desiccator containing THF. Then, 0.2 ml of the sensor solution were added onto the support. The membrane wac left in the desiccator for 30 min and the resulting membrane was placed in ambient air for drying. The thickness of the films was in the order of 2-4 pm (as calculated from the volume employed for spreading), and this resulted in an optical signal with a S/N of typically 300-500. Apparatus Fluorescence excitation and emission spectra a s well as response curves of the sensing membranes were measured on an Aminco (Rochester, NY, USA) SPF 500 spectrofluorimeter equipped with a 250 W tungsten halogen lamp as a light source and linked to an HP 98 15A desk calculator (Hewlett-Packard, Avondale, PA, USA) and a red sensitive detector. Response curves were recorded by placing the membranes in a flow- through cell to form one wall of the cell.Excitation light hit the sensor membrane from outside (after passing the glass wall of the flow cell and the polyester support), and fluorescence was detected at an angle of 55" relative to the incident light beam (Fig. 1). Buffer solutions and buffered sample solutions were pumped through the cell at a flow rate of I .S ml min-1. When studying the response of the sensing membranes, excitation and emission wavelengths were set to 550 and 590 nm, respectively. The absorption spectra of the sensor membranes were measured on a Shimadzu (Kyoto, Japan) UV-2 10 1 -PC photometer.All experiments were performed at 22 f 2 "C. Results Choice of Indicator The sensing scheme used in this work is based on the use of lipophilic derivatives of rhodamine B which dissolve very well in polymeric matrices because of their high solubility in organic Table 1 Composition of anion seiisor membranes MI-M5 Metnbrane Dye Polymer BPP Plasticizer M1 KBOE PVC 200 mol% NPOE M2 RBOE PVC 100 mol% NPOE M3 RBOE PVC 40inol% NPOE M4 RBOE PVC 40mol% DOS MS KBOE PVC-Co 40tnol% NPOE MI Fig. 1 Optical arrangement for measurements with the flow-through cell: D, detector; FC. flow-through cell; MI, mirror; MC, monochromator; SM, sensor membrane consisting of the polymer support and the nitrite sensitive coating; S, sample solution.solvents, plasticizers, and plasticized polymers. The fact that RBOE remains highly fluorescent in plasticized polymers is in contrast to most other PSDs such as aminostyrylpyridinium salts, merocyaiiines and derivatives of acridine orange,] and results in good S/N (Table 2). Esters of rhodamine B are preferred over other PSDs such as aminostyrylpyridinium dyes and carbocyanines because they exhibit properties that render them advantageous in being highly photostable, easily accessible, and highly fluoreccent. The excitation maximum of RBOE is at around 560 nm. This makes it compatible with the green light-emitting diode (LED) which is another advantage because LED\ represent a preferred light source in optical sensor technology. Choice of Carrier In recent years, a wide range of selective anion carrier\ has become available. Most of them are based on metalloporphy- rins.metallophthalocyanines, metallocorrins or alkyltinorganic compounds. In all cases, the anion selectivity depends on the nature of the central metal ion and its complexation with specific anions. In general, these metalorganic compounds also complex hydroxide ions, which restricts the applicability of anion sensors to pHs below 6.0. Recently, the use of a selective anion carrier (BPP) in ISEs has been reported which exhibits very low cross-sensitivity to pH.15 In addition, the selectivity and sensitivity of the resulting ISE could be improved by the addition of up to 30 mol% of the lipophilic quaternary ammonium salt TDMACl (relative to BPP). This is critical for the presented approach because the dye which is used for the optical transduction (RBOE) is a quaternary ammonium salt as well.Finally, BPP is advantageous over metalloporphyrins because it does not absorb significantly above 400 nm when embedded in the sensor membrane and, therefore, does not quench the fluorescence of RBOE. Choice of Polymer Probably the best known material for use in ISEs and optodes is plasticized PVC. It forms fairly stable sensor layers and acts as a good solvent for both lipophilic indicator dyes and ion exchangers or selective ion carriers. On the other hand, the mechanical stability of films is low and a large fraction of plasticizer (usually 66%) is required to obtain fast and stable response. The plasticizer also limits the shelf lifetime because it slowly evaporates or diffuses out.In addition, toxic solvents are required to manufacture the sensor layers. Despite these shortcomings, plasticized PVC has found widespread applica- tion. Currently, copolymers of PVC have been applied for anion sensing as well and are advantageous over PVC because the toxic solvent THF can be replaced by ethyl acetate. However, the matrix affects the selectivity pattern of the anion sensors in that it can decrease the selectivity.12 Table 2 Excitation, emission, and absorption maxima (in nm) and corresponding relative intensities of a 40 Imol ethanolic solution of RBOE diluted 100-fold with the respective solvent Relative fluorescencc intensity/ arbitrary Absorption max.; Solvent Excitation Emission units Abborbance ( A ) Ethanol 559 579 8.1 556; 0.033 DOS 563 S 84 3.2 561; 0.025 NPOE 563 583 10.0 n.d.' Water 554 574 0.1 573; 0.01 I n.d.Not determined because of strong intrinsic absorption of NPOE.Analyst, October 1996, Vol. 121 149 1 Sensor Response On exposure to buffer solutions containing nitrite, membrane M1 shows a distinct response in giving an increase in fluorescence intensity (Fig. 2) which is fully reversed on exposure to plain buffer. The absorption spectra not only show that the absorption of RBOE does not change but also that there is a large change in the absorbance of BPP, albeit at 342 nm which is disadvantageous (Fig. 3). A calibration plot was established from the relative signal changes, and this is shown in Fig. 4. The increase in the relative signal of M1 from 0 to 100 mmol 1-1 nitrite was as high as +110%.The limit of detection for nitrite was 10 pmol (equivalent to 0.5 mg I-l). The forward and reverse response times t9.5 (for 95% of the total signal change to occur) were in the range of 1 &I 5 min. Contacting the membrane alternatively with plain buffer of pH 7.13 and the same buffer containing 1 mmol I-' nitrite at a flow rate of 1.5 ml min-1 resulted in a decrease in signal by 2-3% h-1. Due to leaching, the operational lifetime of M 1 was around 12 to 24 h, and frequent 450 500 550 600 650 700 wavelengthhn Fig. 2 Fluorescence excitation (and corresponding emission) spectra of MI on exposure to: A, plain 20 inmol 1-' phosphate buffer; B, I inmol I--' nitritc; C, 10 mmol I-' nitrite; and D, 100 mmol I - ' nitrite, all at pH 7.13.I 320 395 470 545 620 wavelengthhm Fig. 3 Absorption spectra of M1 on exposure to: A, plain 20 mmol I-' phosphate buffer; B, 10 mmol 1 - I nitrite; and C, 100 mmol 1 - I nitrite, all at pH 7.13. t I I I I I I I I I Fig. 4 Work function of the nitrite sensor M1, and respective plots for potentially interfering anions occurring in drinking water. A, Nitrite; B, nitrate; C, chloride; D, hydrogencarbonate: and E, sulfate. recalibration was required. The shelf lifetime of the membranes exceeded 3 months when they were stored in the dark at room temperature. Similar behaviour was observed for the other membranes plasticized with NPOE. In order to investigate the effect of the plasticizer on the response, NPOE was replaced by the less polar DOS.The resulting sensor membrane showed similar relative signal changes and sensitivity towards nitrite when compared with the NPOE membranes. Both the forward and reverse reponse times remained unchanged. However, the stability of the DOS-based sensor membrane in terms of signal drift and leaching was inferior. In a previous investigation on the effect of' polymeric matrices on the response of an RBOE-based nitrate sensor, similar effects have been found.12 Effects of pH The fluorescence of RBOE in membranes MI-M5 decreased almost linearly by around 9% on increasing the pH by one unit (Fig. 5). From pH 5.0 to 9.0, the fluorescence intensity decreased by -38%. Additionally, the magnitude of the relative signal change (e.g., from plain buffer to 1 mmol l-' nitrite) was pH-dependent and was lower, by around 65% at pH 9.0 than at pH 5.0 (Fig.5 ) . t I 1 1 I 9 8 7 6 5 4 PH Fig. 5 pH-Dependence of the horescence intensity of sensor membrane M1 in: A, plain phosphate buffer; B, phosphate buffer containing 1 mmol 1-1 nitrite. Selectivity The selectivity of sensor membrane M1 for anions is in clear contrast to the Hofmeister pattern. Anions occurring in drinking water in higher concentrations include chloride, sulfate, hydro- gencarbonate, and nitrate, and the response to those was investigated in more detail. Fig. 4 shows the relative signal changes caused by nitrate, chloride, sulfate, and hydrogen- carbonate in comparison with nitrite. selectivity coefficients relative to nitrite were determined by the separate solution method and are shown in Table 3.The best selectivity for nitrite over nitrate was obtained with M2 which exhibits a 1 : 1 ratio of RBOE to BPP. M2 also showed the best selectivity for nitrite over chloride and sulfate. However, the selectivity over chloride was clearly too small to carry out measurements in sea-water or human blood. For such sensor applications, a more selective Table 3 Selectivity factors (log K(,pc*) of sensor membranes M 1-M5 relative to nitrite for nitrate, chloride, hydrogencarbonate and sulfate as determined by the separate solution method (SSM) at pH 7.13 Membrane M1 M2 M3 M4 M5 Nitrate 1.7 2.1 1.7 1.3 1.9 Chloride 2.6 2.8 2.8 2.2 2.1 Sulfate >3.5 >3.s >3.s >3.s >3.5 A Log Kop, is the log of the ratio of the concentrations of interferent and nitrite giving the same signals.Hydrogencarbonate 3.2 3.1 3.2 3.1 3.11492 Analyst, Ortohei- I996, Vol. I21 ilasticized 'VC carrier is required. The selectivity of sensor membrane M4 for nitrate and chloride was worse than that of the NPOE-based membranes which is in agreement with reported findings (Table 3).15 MS exhibited a selectivity behaviour similar to MI-M3; however, the selectivity for nitrite over chloride was slightly smaller . aqueous phase (phosphate buffered) Discussion Sensing Scheme of the Anion-sensitive Membranes Based on RBOE RBOE is a solvatochrotnic dye whose fluorescence intensity rather than the maxinia of the excitation and emission plasticized PVC A wavelengths are affected by changes of polarity.The fluores- cence of the dye is stronger in moderately polar solvents (such as plasticized PVC) than in water (Table 2).16,'7 RBOE exhibits lower fluorescence (around 40 times) in water than in solvents such as ethanol or plasticizers. This is due to the formation of non-fluorescent dimers caused by hydrophobic interactions between the long alkyl chains of RBOE in water. Furthermore, the absorption in water is approximately half of the absorption in DOS. These facts indicate the formation of dimers in water which are non-fluorescent and, thus, not visible in the fluorescence spectrum. Similar results have been observed when comparing the optical properties of RBOE in water and ethanol. 6 aqueous phase (phosphate buffered) polyester support rhodamine B benzylbis(tripheny1phosphine) palladium(1 I) chloride watery phase in U plasticized PVC f octadecylester weak fluorescence aqueousphase containing nitrite B -0 strong fluorescence Fig.6 membrane phaw on exposure to nitrite. Schematic representation of the nitrite-sensitive membrane, and the motion of RBOE from A, the more aqueous interphase to B, the more lipophilicAnalyst, October 1996, Vol. 121 1493 Changes in the counterion of RBOE change the solubility of RBOE in the plasticizer and affect the dissociation of dye dimers to monomers, causing the fluoresence to increase. RBOE is a cationic lipophilic dye which due to its amphiphilic structure, acts as a carrier for anions in a fashion similar to the well known lipophilic quaternary ammonium ions. Conse- quently, the RBOE cation by itself causes a response to lipophilic anions by extracting them into the organic phase (similar to an anion exchange catalyst).’ 1 In the present case, an interaction different to the previous case takes place. Opposite to the RBOE-based anion sensor whose selectivity corresponds to the lipophilicity of anions (Hofmeister pattern), BPP selectively reacts with nitrite.The neutral carrier BPP preferably binds nitrite to form a negatively charged complex15 which, in turn, forms an ion pair with cationic RBOE. The ion pair is more lipophilic than cationic RBOE alone and fluorescence increases. A schematic repre- sentation of this extraction mechanism is shown in Fig. 6. Effect of the Polymer Matrix on the Response Plasticized PVC membranes undergo a substantial water uptake during conditioning in aqueous solutions.Recent investigations on bulk membranes of plasticized PVC have shown a water-rich region of a thickness of 50 pm on exposure of the bulk membrane to water. ‘8 The sensor membranes presented here do not exhibit thicknesses higher than 5 pm. Consequently, there is a significant water content in these sensor membranes after equilibration in buffer. Plasticizers and water form a micro- emulsion, and this formation is known to be enhanced by the presence of surface active components.19 RBOE is an amphi- philic compound and is therefore surface active. The micro- emulsion contains water domains of different sizes and provides a large interfacial area.20 This explains both the relatively fast response of the sensor membranes and the large signal changes, because the signal changes are not restricted to the dye bound to the surface of the sensor layer (see Fig. 6).Effect of the Charge of PSDs and Anion Carriers on the Response To date, most of the selective anion carriers known are either metalloporphyrins or tin organic compounds. Their selectivity is caused by the interaction of the central metal atom of the organic molecule with the anion.4.5 However, most of these carriers are highly cross-sensitive to hydroxide and, conse- quently, cannot be operated at pH values above 6.0. In order to reduce the cross-sensitivity of the anion carriers and to improve selectivity, lipophilic cationic and anionic sites have been added to the solvent polymeric membranes.The nitrite carrier aquocyanocobalt(111)-hepta(2-phenylethyl)cobyrinate perchlor- ate has been combined with quaternary ammonium salts and borates in order to improve its selectivity.*l When this so-called ‘charged carrier’, which is positively charged in the un- complexed form, but neutral in the complexed form, was used together with borate, the selectivity for nitrite over interferent anions was enhanced. Addition of quaternary ammonium salts to the selective anion carriers, however, resulted in a selectivity identical to the Hofmeister pattern. The second type of anion carrier is the ‘neutral carrier’ such as chloro(5,10,15,20-tetraphenylporphyrinato) cobalt(111) which is neutral in the uncomplexed form but negatively charged in the complexed form. This carrier was reported to maintain its selectivity if used together with quaternary ammonium salts and to lose it if combined with borates.22 Most of the PSDs including RBOE are quaternary ammonium salts; if combined with charged anion carriers, their selectivity is lost.However, if cationic PSDs are combined with neutral anion carriers, the selectivity of the resulting sensor membrane should remain. Indeed, the use of cationic PSDs together with charged anion carriers such as chloro(octaethy1porphyrinato) indium(II1) or trioctyltin chloride results in unselective anion sensor mem- branes.14 Consequently, the combination of a cationic PSD with a neutral anion carrier such as BPP results in an anion-selective sensor membrane. Effect of the RBOE-to-BPP Ratio on the Response Badr et a1.15 have shown that the ratio of BPP to the cationic additive (TDMAC1) affects the response. At a concentration of around 30 mol% TDMAC1, both the cross-sensitivity to pH and the selectivity of the electrochemical nitrite sensor to inter- ferents were reported to be optimal.In the present case, however, a change of the ratio had no effect on the cross- sensitivity to pH (Fig. 5 ) and only minor effects on the selectivity to interfering anions (Table 3). As a consequence, a tailoring of the sensor material is not possible. On the other hand, a change in the ratio caused by leaching of the components did not shift the calibration plot. Conclusion Ion-exchange or co-extraction based optodes are an established and theoretically well-described group of optical sensors.However, due to the mechanism, even slight changes in pH cause large errors in analyte determination. In contrast to this type of optode, PSD-based optodes respond to changes of the microenvironment of the dye rather than to protonation- deprotonation of a pH indicator. Therefore, they exhibit significantly lower cross-sensitivity to pH than optical sensors based on the ion-exchange-co-extraction mechanism. How- ever, an empirical relation between fluorescence changes and analyte ion concentration has to be developed, rather than a mathematical model which is valid for ion-exchange or co- extraction based optodes. The most serious disadvantage of PSD-based ion sensors is their limited lifetime due to significant leaching of the indicator dye. This work was supported by the Austrian Science Fpundation within project S5701-PHY and P10,389-CHE which is grate- fully acknowledged.References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Schulthess, P., Ammann, D., Krautler, B., Caderas, C., Stepanek, R., and Simon, W., Anal. Chem., 1985,57, 1397. Li, J.-Z., Pang, X.-Y., and Yu, R.-Q., Anal. Chim. Acta, 1994, 297, 437. Mod, W. E., Seiler, K., Lehmann, B., Behringer, C., Hartman, K., and Simon, W., Pure and Appl. Chem., 1989,61, 1613. Tan, S. S. S., Hauser, P. C., Wang, K., Fluri, K., Seiler, K., Rusterholz, B., Suter, G., Knittli, M., Spichiger, U. E., and Simon, W., Anal. Chim. Acta, 1991, 255, 35. Wang, E., and Meyerhoff, M. E., Anal. Chim. Acta, 1993, 283, 673. Tan, S. S. S., Hauser, P. C., Chaniotakis, N. A., Suter, G., and Simon, W., Chimia, 1989, 43, 257. Lumpp, R., Reichert, J., and Ache, H. J., Sens. Actuators, B, 1992,7, 473. Hauser, P. C., and Tan, S. S. S., Analyst, 1993, 118, 991. Bachas, L. G., and Freeman, M. K., Anal. Chim. Acta, 1990, 241, 119. Yang, S. T., and Bachas, L. G., Talanta, 1994,41, 963. Mohr, G. J., and Wolfbeis, 0. S., Anal. Chim. Acta, 1995, 316, 239. Mohr, G. J., and Wolfbeis, 0. S., Sens. and Actuators, B, submitted. Kawabata, Y., Tahara, R., Kamichika, T., Imasaka, T., and Ishibashi, N., Anal. Chern., 1990, 62,..2054. Mohr, G. J., Lehmann, F., Ostereich, R., Murkovic, I., and Wolfbeis, 0. S., Fresenius’ J. Anal. Chem., in the press.1494 Analyst, October 1996, VoE. 121 15 16 17 18 19 Badr, I. H. A., Meyerhoff, M. E., and Hassan, S. S. M., Anal. Chem., 1995, 67, 2613. Nakashima K., and Fujimoto Y., Photochem. Photohiol., 1994, 60, 563. Nakashima K., Fujimoto Y., and Anzai T., Photochem. Photohiol., 1995, 61, 592. Li. Z., Li, X., Petrovic, S., and Harrison, D. J., Anal. Methods Instrum., 1993, 1, 30. Rees, G. D., and Robinson, B. H., A h . Mater. (Weinheim, Fed. Repuh. Ger.), 1993, 9, 608. 20 Wolfbeis, 0. S., Sens. Actuators, B, 1995, 29, 140. 21 Schaller, U., Baker, E., Spichiger, U. E., and Pretsch, E., Anal. Chem., 1994,66, 391. 22 Bdkker, E., Malinowska, E., Schiller, R., and Meyerhoff, M. E., Talanta, 1994, 41, 88 1. Paper 61041 51 E Received June 13, 1996 Accepted July 29, 1996

ISSN:0003-2654    

DOI:10.1039/AN9962101489

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, October 1996, Vol. 12 I (1495-1499) 1495 Polymeric Membrane Salicylate-sensitive Electrodes Based on Organotin(iv) Carboxy lates Dong Liu, Wen-Can Chen, Guo-Li Shen and Ru-Qin Yu* Department (fl Chemistry and Chemicul EnRineering, H u n m University, Changsha, 410082, China Selectivity properties were established for membrane electrodes prepared by incorporating tribenzyltin carboxylates in plasticized polymeric membranes. These electrodes display high selectivity for salicylate with respect to many common anions. An electrode prepared with tribenzyltin octoate, using o-nitrophenyl octyl ether (0-NPOE) as the plasticizer, possesses the best potentiometric response characteristics, including a fast response time. It shows a linear response towards salicylate ions over the concentration range 0.1-5 X 10F mol I-' with a slope of -57.5 mV decade-' in buffer solutions of pH 5.5.The behaviour of the electrode is considerably influenced by the plasticizer employed and the optimum response appears to result when o-NPOE is present. The electrode was applied to the determination of salicylate in pharmaceutical and urine samples. Keywords: Salic,ylate-s~nsitive electrode; tribenzyltiiz c*al-hoxylates; ionophore Introduction Highly selective polymeric membrane electrodes are now routinely used for the in situ determination of various cations. However. the development of similar devices by using conventional ion exchangers as the carriers for the detection of anions was difficult. Usually all these electrodes exhibit roughly the same selectivity sequence, i.e., Clod- > SCN- > I- =: Sal- > NO3- > NO2- > C1- > SO4'-- (Hofmeister sequenced).Since the pioneering work of Simon's groups,6 using vitamin €3 12 derivatives and trioctytin chloride as anion carriers for the preparation of polymeric membrane electrodes that show selectivity patterns different from the Hofmeister series, significant effort has recently been placed on the development of ion-selective electrodes (ISEs) by the use of organometallic species and metal-ligand complexes as mem- brane-active c~mponents.~-l The electrode selectivity, in these cases, is not governed by simple anion lipophilicity rather than by a specific chemical interaction between the organometallic species in the membrane phase and the anions in solution. In other words, the nature of the central metal and the structure of the carrier will play an important role in the realization of the selectivity pattern. A salicylate-selective electrode based on a quaternary ammonium salt l6 has been proposed for some practical purposes.However, the sensor suffered significant interfer- ences from a wide range of other anionic species, including a number of common physiological anions, and thus cannot be utilized for direct measurements in biological samples. A systematic search for salicylate-sensitive electrodes has been undertaken in this laboratory. After discovering a salicylate- A To whom correjpondence 5hould be addre5xd. sensitive electrode based on it tin(1v)-phthalocyanine com- plex,l7 which has a similar Structure to tetraphenylporphyrin- tin(rv) dichlorides reported by Chaniotakis et al.,lx it has been found that tribenzyltin carboxylates, which have a radically different structure to the salicylate-sensitive carriers reported co far, possess promising potentiometric response characteristics for salicylate ions.In this paper, the use of tribenzyltin carboxylates as salicylate ionophores is reported together with their potentiometric response characteristics and preliminary application to the determination of salicylate in human urine samples and pharmaceutical analysis. Experimental Reagents All aqueous solutions for the potentiometric measurements were prepared with distilled, de-ionized water and salts of the highest purity available. Working standard solutions were freshly prepared by accurate dilution from a stock standard solution stored in an amber-coloured bottle.High molecular mass PVC powder of chromatographic grade and dinonyl sebacate (DNS), dibutyl phthalate (DBP) and tetrahydrofuran (THF) of analytical-reagent grade were purchased from Shang- hai Chemicals (Shanghai, China) and used without further purification. To ensure high purity of o-nitrophenyl octyl ether (o-NPOE) for measurements in liquid membranes, it was prepared according to the reported procedure. '9 Synthesis of the tribenzyltin carboxylates (Fig. 1) was carried out according to the method of Xie et a1.2" by reaction of bis(tribenzy1tin) oxide with the corresponding carboxylic acid. Elemental analysis and melting-points were used to identify the compounds (Table 1).Apparatus All potentiometric measurements were carried out at 20 "C on a Model 90 1 Microprocessor Ionalyzer (Orion, Cambridge, Fig. 1 Structure of the carriers studied. R = I , H; 2, C1H7; 3, C7HIs; 4, C6H5; .5,4-CH3CbH4; 6, 4-ClC6H4.14% Aiialyst, Ortoher 1996, b J d . 121 MA, USA). Plasticized PVC membranes were prepared accord- ing to the reported method.'l.'2 The composition of these membranes was 1.5% m/m of the ionophore, 64.0% m/m of the plasticizer and 34.5% m/m of PVC. A solution containing 0.01 mol I - l sodium salicylate and 0.1 rnol I-' potassium chloride was used as the internal filling solution and a saturated calomel electrode was used as the reference electrode. The electrode cell for potential measurements was Hg I Hg2C12 1 KCI (satd.)llsample solution I membrane I 0.01 rnol 1-I Nasal, 0.1 mol 1-' KClIAg IAgCl Procedures The calibration solutions were buffers 0.01 rnol I-' in H3P04 and the pH was adjusted with NaOH solution.The calibration of the electrodes was carried out by adding, while stirring, appropriate amounts of aliquots of standard salicylate solutions of different concentrations to a beaker containing 30.0 ml of buffer. The potentiometric selectivity coefficients of the electrodes for salicylate with respect to other anions were determined by the separate solution method in 0.1 rnol 1 - 1 solutions of the corresponding sodium salts. The solutions were buffered to pH 5.5. The single-ion activities were calculated by use of the extended Debye-Huckel equation.For the preparation of pharmaceutical samples containing salicylate for potentiometric measurement, tablets of aspirin or aspirin-phenacetin-codeine (APC) were finely powdered. A precisely weighed sample of the powder containing approxi- mately 0.5 g of acetylsalicylic acid was refluxed with 30 mi of 0.5 rnol 1-1 sodium hydroxide for 1 h. After being filtered, the solution was adjusted to pH 5.5 with sulfuric acid and then transfered into a 250 in1 calibrated flask and diluted to volume with phosphate buffer (pH 5.5). An aliquot of 30 ml of this solution was pi petted into the measuring cell for emf recording. A pharmacopoeia1 procedure'? was used as the reference method for the assay of aspirin contents. This method is based on the conventional acid-base titration of acetylsalicylic acid, using phenophthalein as the indicator and keeping the tem- perature below 10 "C to avoid the hydrolytic decomposition of aspirin.Results and Discussion Potentiometric Response Characteristics of Electrodes The potentiometric response characteristics of the PVC niem- brane electrodes incorporating tribenzyltin carboxylates to- wards salicylate in buffer solutions of pH 5.5 are shown in Fig. 2. All the electrodes are sensitive towards salicylate to some extent. The electrode containing carrier 3 exhibits a nearby Nernstian slope of -57.5 mV decade - 1 for the linear response concentration range 0.1-5.0 X 10-6 moll-1, whereas membrane electrodes based on carriers 4-6 show poor potentio- Table 1 Physical properties and analytical data for the carriers Elemental analysis: found (calculated) Carrier Col O L I ~ MpPC C ( 9 ) H (%) 1 Whte 141 60.21 (60.44) 5.09 (5.07) 3 Colourless 66-67 65.01 (65.09) 6.63 (6.74) 4 Wt11te 102-103 65.69 (65.54) 5.02 (5.03) 5 White 96-97 65.99 (66.08) 5.27 (5.3 I ) 6 Colourless 105 61.12 (61.14) 4.50 (4.57) 2 White 1 11-1 12 62.56 f62.66) 5.89 (5.84) metric responses with an average slope of about -30 mV decade-1.The electrode based on carrier 3 shows slightly superior potentiometric characteristics when compared with tin(1v)-phthalocyanine, reported elsewhere. 17 The disparity of these electrodes in potential response characteristics may derive from the differences in their molecular structures. Generally, it is thought that, with sensing materials of the metal-ligand complex type.the selective potential response towards anions comes from the interaction between the central metal and the anions sensed. The differ- ences in the molecular structures of the ionophores, on the other hand, may also have a significant influence on the response characteristics of corresponding electrodes. Different hydro- carbyls attached to the carboxy group in the organotin derivatives involved, for instance, may have a substantial influence on the properties of the compounds and the behaviour of the electrodes prepared. For carriers 4-6, containing benzene rings, the response characteristics seem to be influenced by the steric effect of the benzene ring hindering the interaction between the central metal Sn(iv) and the anion sensed. For carriers 1-3, such a hindrance should be diminished because of the linear structure of the hydrocarbyls involved.The lipophilicity of the ionophores also plays a role, although its effect is not so significant. For electrodes based on carriers 1-3, only a minor difference in potentiometric response characteristics was observed: the linear response limits for carriers 1 , 2 and 3 are 1 .0 X 10--5, 8.0 X 10-6 and 5.0 x 10-" mol 1-1. respectively, with a slope of approximately -57 mV decade--- 1 in each instance. The stability and reproducibility of the carrier 3-based electrode were also tested. Its potential drift was within 0.8 mV when the electrode was dipped in a solution containing 1 0 -3 rnol I-' Nasal for 2 h. The slope of the electrode was reproducible to within 1.0 mV decade-l over a period of 2 months.The conditions for the storage of the electrodes were evaluated. It was found that the best results were obtained when the electrodes were kept at room temperature, conditioned in 10-4 rnol 1-1 Nasal between experiments. Under these conditions the fluctuation of the starting potential was found to be within +5 mV. The effect of variations in membrane composition on the electrode performance was studied (Table 2) An increase in the concentration of the carrier in the membrane phase is beneficial for obtaining any electrode slope closer to the theoretical value. ,4t least 1.5% m/m of carrier must be present in the membrane phase to obtain electrodes with a more or less normal response slope. When the concentration of carrier reached 3% m/m, the resulting membranes were found to be heterogeneous or non- transparent and the potential readings for these electrodes in the 100 0 > E Lrl --.-100 -200 -6 -5 -4 -3 -2 - 1 Log Li Fig. 2 carriers depicted in Fig. 1: I , a: 2, b: 3, c: 4, d; 5. e; and 6, f. Potentioinetric response to salicylate for the electrodes based o nAnalyst, Octohel- 1996, Vol. 12 1 1497 low concentration range were fairly unstable. When the o- NPOE to PVC ratio was altered from 1 : 2 to 2 : 1, the resulting membranes possessed the lowest detection limit. Taking all these results into account and on the basis of solubility considerations, the optimum carrier-plasticizer-PVC composi- tion was found to bc 1.5:64: 34.5 m/m and most of the electrodes tested in this study had this composition.Selectivity Possible interferences from a number of monovalent anions (Cl-, Br-, I - , C104-, NO3-, NO2-, SCN-, acetate, benzoate, citrate and lactate) were studied. Table 3 gives the selectivity coefficient values for electrodes containing carriers 1-3 with respect to various anions tested. In general, these three electrodes display remarkable selectivity for salicylate over common anions. This may be interpreted in terms of the special interaction between salicylate ions and carrier incorporated in the tnetnbranes. Thiocyanate and benzoate ions seem to tend to interact with the organotin carriers to some extent, showing some interference i n the salicylate detennination. When compared under similar experimental conditions, the selectivity characteristics of organotin carboxylates are slightly superior to those of tin(rv)-phthalocyanine reported elsewhere.I7 Effect of pH on Response Characteristics of Electrodes The potentiometric response of the organotin-based electrodes tested was found to be sensitive to pH changes (Fig. 3). The pH dependence of these electrodes was tested by measuring calibration curves in buffer solutions at various pH values. Fig. 4 illustrates the results obtained with the electrode containing carrier 3, which show that there was essentially no difference in the slopes of the calibration curves for these buffer systems. However, as the pH increases, the detection limits (determined according to the IUPAC recommendation) of the calibration curves for salicylatc deteriorate.This is accom- Table 2 Variation of potentiometric response characteristics of electrodes using tribenzyltin octoute as the carrier with different membrane composi- tions, measured in phosphate buffer solution (pH 5.5) Proportion of component (% d m ) Carrier o-NPOE PVC 0. I 50 49.9 0.5 50 49.5 I .5 50 48.5 3.0* 45 52 I .5 33 65.5 1 .5 65 33.5 Detection limit/ lnol 1 ' 5 x 1 0 2 x 10-5 7 x 10-6 4 x 1 0 1 x 1 0 - 5 I x 10- 6 SlopeImV decade-- I -32.4 -48.6 -56.4 -54.7 -56.6 -57.5 . Potentiometric readings in the solutions of concentration < 11701 I were unstable for this ineinbrane composition. panied by a decrease in the starting potential (recorded with the cell before any addition of the salicylate standard solution). The experiments with electrodes containing carriers 1 and 2 showcd results similar to those for carrier 3.This behaviour can be explained in terms of the increased interference from OH-. As a consequence, the detection limits of the electrode in Fig. 4 were equal to 1.0 X 10-6,2.7 X and 9.8 X when the electrode was exposed to calibration solutions buffered to pH 5.0, 6.5 and 8.0, respectively. Dynamic Response Characteristics Response time is an important factor for an ISE. Although a tin(rv)-tetraphenylporphyrin-based polymeric membrane elec- trode has been reported's with high selectivity, the response time seems slow, ranging from 2 to 10 min depending on the concentration of salicylate in the samples. The practical response times of organotin compound-baced electrodes were studied by rapidly changing the salicylate concentration in stirred solutions.More or lcss stable potential readings can be obtained within 10-15 s, depending on the mixing efficiency. The actual potential \wsus titne trace showed that 90% of the expected response can generally be obtained within 30 s for solutions with salicylate levels higher than 1 OW3 mol I- I , and within 1 min for lower concentrations (Fig. 5). In the presence of a background electrolyte, c..,q., 0.1 mol I-' KNOT, the response time would be even faster and the potential readings remain stable. Preconditioning of the membranes is beneficial for achieving a fast response titne. A freshly preparcd membrane used directly after preparation without treatment gives a slow response time of 2-3 min. After exposure to 10- mol 1-1 Nasal solution for 3 h, the same electrode gives a normal response time of around I min.Influence of Plasticizer The plasticizer used seems to have a significant influence on the response of the electrode. When o-NPOE was used as the plasticizer, the detection limit for salicylate and the selectivity coefficients of the electrode for carrier 3 were better than thosc for the DBP-based electrode. The electrode employing DNS shows hardly any potentiometric response towards salicylate (Fig. 6). The observed phcnoinena might be related to the existing forms of the ionophores molecule involved. It has been noted that substituted organotin carboxylates may have two forms (Fig. 7): the single molecular form and the commonly existing carboxylate-bridged polymer (B).When dissolved in a polar solvent, according to Yang et d.,24 the polymer tends to depolymerize and return to the single molecular form. For the solvent polymeric membrane electrodes, the sensitive mem- branes are in fact liquid membranes, with ionophores dissolved in a plasticizer as the solvent. For substituted organotin carboxylates, the form in which it exists will be determined by the polarity of the solvent or the plasticizer in membrane. When the plasticizer is o-NPOE, which has a comparatively strong Table 3 Potentiometric selectivity coefficients. log Kl:,',:,, for the electrodes, determined using the separate solution method in 0.01 in01 I - l phosphatc buffer solution (pH 5.5) at an anion concentration of 0.1 mol 1 Interfering ion, j Carrier C104- SCN- I- Br- NOT- NO-, C 1 ~ Acetate Lactate Citrate Benmate I -2.98 -1.96 -3.66 -3.92 -4.i7 -3.00 -4.52 -3.76 -2.46 -3.31 -1.21 2 -3.02 -2.01 -3.79 -3.84 -4.13 -2.94 -4.79 -3.77 -2.52 -3.47 -1.28 3 -3.07 -2.14 -3.78 -3.95 -4.20 -3.12 -4.85 -3.89 -2.95 -3.40 -1.26 A" 2.24 I .04 0.62 -1.17 -0.95 -1.82 -1.96 -2.15 -1.94 -1.56 -1.01 * The niembrane was prepared with 5(h m/m Aliquot 336-Sal, 30% PVC 65% DBP.1498 Anulyst, October 1996, V d .121 polarity, the single molecular form of the ionophore would be dominant. In contrast, when employing DNS, the polarity of which is the weakest among the three mediators studied, the polymeric structure might be dominant. The relatively high molecular mass of the polymer will reduce the mobility of ionophore in the membrane phase, causing the potentiometric response to diminish for particular membranes.Prelim in ary Applications Polymeric membrane salicylate-sensitive electrodes may find applications in a variety of fields. To demonstrate such applications, the electrode based on ionophore 3 and employing o-NPOE as the plasticizer was used in assaying the content of 300 200 100 > E G o -100 carrier 1 carrier2 A carrier3 -200+' 1 ' ' ' 1 ' 1 ' I 1.0 3.0 5.0 7.0 9.0 11.0 13.0 PH Fig. 3 pH response of electrodes containing organotin carriers. 2oo 100 i l s o c 4 - 100 -200 -6.0 -5.0 -4.0 -3.0 -2.0 -1.0 Log Li Fig. 4 pH value\. Calibration graphs for electrode incorporating carrier 3 at different 40 vl . 30 G 20 1 0 : ' 1 ' I ' I I ' I -6.0 -5.0 -4.0 -3.0 -2.0 -1.0 Log Ll Fig.5 Response time (90%) of electrode based on can-ier 3. acetylsalicylate acid in tablets and in the determination of salicylate in human urine samples. Aspirin is usually dispensed alone or together with phenace- tin and codeine as APC tablets. They are commonly used as effective analgesic and antipyretics. The content of acetylsal- icylic acid in tablets is conventionally assayed by titrimetric analysis of the hydrolysed product, salicylate. Many other methods based on chromatography, spectrophotometry and enzymic methods have also been de~eloped,25-~7 but most of them require lengthy clean-up of the sample. A direct potentiometric method using ISEs seems to be promising with the advantages of being simple and fast. The tablets were treated according to the above-described procedure and the electrode was used to determine the concentration of salicylate as the hydrolysed product of aspirin in samples by employing the standard additions method.The results obtained with potentio- metric method compared with those of Chinese Pharmacopoeia (C.P.) standard procedure2? and the label values are shown in Table 4. The results obtained by the two methods agree with the label values very well, but the C.P. procedure is tedious for practical application. Polymeric membrane electrodes for salicylate measurements employing quaternary ammonium salts as the carriers have been proposed previously. 16,28 These earlier versions of salicylate probes were, unfortunately, not suitable for clinical purpose they lacked adequate selectivity over common anions.Because of the high selectivity of the described electrodes over a number of physiological anions, including chloride, it makes the use of the sensor in the analysis of physiological samples feasible. The electrode was used to determine salicylate in human urine. All samples can easily be prepared simply by tenfold dilution of 1 > E 4 . - 1 00.0 0.0 00.0 -b Log (1 -200.0 I . l . l . l ' / -6.0 -5.0 -4.0 -3.0 -2.0 -1.0 Fig. 6 with: 1, o-NPOE; 2, DNS; and 3, DBP. Potentiometric response of carrier 3-based electrodes plasticized R (A) (B) Fig. 7 Two forms of the tribenzyltin carboxylate molecule. -~~ ~ Table 4 Determination of aspirin in tablets (%) Tablet Label Potentiomctry " C.P. method APC 41.15 40.98 40.77 Aspirin 80.62 80.30 80.84 Mean values for three separate determinations.Analyst, October 1996, Vol.121 1499 Table 5 Comparison of results of electrode and spectrophotometric methods for the determination of salicylate concentration (mmol 1-1) in urine samples. Samples were diluted tenfold with 0.05 moll-' phosphate solution (pH 5.5) Measurement No. 9 10 I 1 12 Method 1 2 3 4 5 Electrode 0.82 0.78 0.68 0.97 1.02 Spectrophotometry 0.86 0.75 0.66 0.94 1.07 13 14 15 urine with 0.05 moll-' phosphate buffer solution (pH 5.5). The results are shown in Table 5. A comparison with the results obtained by spectrophotometry29 (which is based on the Trinder reaction, in which sample salicylate reacts with Fe3+ ions to form a coloured complex in acidic solution) showed good agreement, indicating that the application of organotin com- pound-based electrodes to physiological samples is feasible.This work was supported by the National Natural Science Foundation of China. References 1 2 3 4 5 6 Oesch, U., Ammann, D., and Simon, W., Clin. Chem., 1986, 32, 1448. Meyerhoff, M. E., and Opdycke, W. N., Adv. Clin. Chem., 1986, 25, 1. Thomas, J. D. R., Anal. Chim. Acta, 1986, 180,289. Hofmeister, F., Arch. EAP. Pharmakol., 1888, 24, 247. Schulthess, P., Ammann, D., Simon, W., Caderas, C., Stepanek, R., and Krautler, B., Helv. Chim. Acra, 1984, 67, 1026. Wuthier, U., Pham, U. V., Zund, R., Welti, D., Funck, R. J. J., Bezegh, A., Ammann, D., Pretsch, E., and Simon, W., Anal. Chem., 1984, 56, 535. Hodinar, A., and Jyo, A., Anal. Chem., 1989, 61, 1171. Daunert, S., Wallace, S., Floride, A., and Bachas.L. G., Anal. Chem., 1991,63, 1676. 7 8 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Wang, E., and Meyerhoff, M. E., Anal. Chim. Acta, 1993, 283, 673. Blaor, T. L., Allen, J. R., Daunert, S., and Bachas, L. G., Anal. Chem., 1993,65, 2155. Li, J. Z., Wu, X. C., Yuan, R., Lin, H. G., and Yu, R. Q., Analyst, 1994,119, 1363. Gao, D., Gu, J., Yu, R. Q., and Zheng, G. D., Analyst, 1995, 120, 499. Glazier, S. A., and Arnold, M. A., Anal. Chem., 1991, 63, 754. Rothmaier, M., and Simon, W., Anal. Chim. Acta, 1993, 271, 135. Bakker, E., Malinowske, E., Schiller, R. D., and Meyerhoff, M. E., Talanta, 1994, 41, 881. Papazoglou, A. M., Diamandis, E. P., and Hadjioannou, T. P., Anal. Chim. Acta, 1984, 159, 393. Li, J. Z., Pang, X. Y., Gao, D., and Yu, R. Q., Talanta, 1995, 42, 1775. Chaniotakis, N. A., Park, S. B., and Meyerhoff, M. E., Anal. Chem., 1989, 61, 566. Horning, E. C., Org. Synth., Cull. Vol., 1955, 3, 140. Xie, Q. L., Xu, X. H., and Zhang, D. K., Acta Chim. Sin., 1992, 50, 508. Moody, G. J.. Oke, R. B., and Thomas, J. D. R., Analyst, 1970, 95, 910. Craggs, A., Moody, G. J., and Thomas, J. D. R., J . Chem. Educ., 1974, 51, 541. Pharmacopoeia Committee of the Ministry of Health of China, Chinese Pharmacopoeia, Chinese Health Press, Beijing, 1990, vol. 2, Yang, Z. Q., Xie, Q. L., and Zhou, X. Z., Acta Chim. Sin., 1995, 33, 721. Yon, K., and Bittikofer, J. A., Clin. Chem., 1984, 30, 1549. Walter, L. J., Biggs, D. F., and Coults, R. T., J. Pharm. Sci., 1984,63, 1754. Dadgar, T., Climax, J., Lambe, R., and Darragh, A. T., J . Chroma- togr., 1985, 342, 315. Choi, K. K., and Fung, K. W., Anal. Chim. Acta, 1982, 138, 385. Trinder, P., Biochem. J . , 1954, 57, 301. p. 4. Paper 6102880B Received April 24, 1996 Accepted June 3,1996

ISSN:0003-2654    

DOI:10.1039/AN9962101495

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Surface Plasmon Resonance of Self-assem bled Pht halocyanine Monolayers: Possibilities for Optical Gas Sensing Tim R. E. Simpson,a Michael J. Cook,” Michael C. Petty,h Stephen C. Thorpec and David A. Russella** a School cf Chernicul Sciences, University of East Anylia, Norxich, UK NR4 7T.l Dur-ham, Dui-lzam. UK DHI 3LE Centse for Molecxlat- Elt.c*ti-onic~s, School of Engineering, Univci*sity of Health and Sujcty E.\-ec*utiiv, Broad tune, Sliefli‘eld, UK S3 7HQ A diphthalocyanine disulfide (Pc) molecule has been deposited as a monolayer on gold-coated substrates through the process of self-assembly. To establish the molecular orientation of the Pc molecule on the gold surface the two complementary techniques of transmission IR and reflection absorption IR (RAIR) spectroscopies were used.The appearance of IR absorption bands associated with the Pc nucleus in the transmission spectrum, and their absence in the RAIR spectrum, suggests that the Pc self-assembled monolayer (SAM) is orientated with the macrocycle parallel to the metal surface. The Pc SAM has been used in conjunction with surface plasmon resonance (SPR) to establish the utility of combining these techniques for optical gas sensing. The SPR reflectivity curves for the gold substrate and the Pc SAM on the gold substrate have been obtained. On exposure of the Pc SAM to the environmentally important NO2 gas, changes of the reflectivity signal were obtained in proportion to the concentration of the analyte gas. The results obtained show that the monolayer deposition technique of self-assembly is an ideal method for the production of chemically sensitive substrates which can be combined with surface plasmon resonance for the optical sensing of gaseous species.Keywords: Surfuce plasnion resonunce; self-ussevlzhled nionolayers; phthaloc-yanine; optical sensor; nitrogen dio-xide Introduction The formation of organic monolayer? on surfxes through the process of self-assembly is now a well established procedure. In particular, two forms of self-assembled monolayers (SAMs) have been extensively characterized.’ The first involves the formation of a covalent bond between an organic amphiphile and the substrate. This type of SAM is typified by the organosilicon on a hydroxylated surface, either silicon or glass, system.2 The second type is formulated between an amphiphile containing a thiol or disulfide moiety and a metal surface.? The bond formed in this second type of SAM is considered to be somewhat ionic in nature as a thiolate species is formed between the metal, typically gold or silver, and the amphiphile.4 Numerous spectroscopic techniques have been applied to the characterization of these SAMs including surface plasmon resonance (SPR).SPR is a method by which photons can be coupled to the surface plasmons at a metal-dielectric interface. _ _ ~ _ _ _ _ _ _ _ _ _ _ _ ~_ - ~~ * To whom coric\pondencc 4iould be dddre\\ed The production of highly reproducible monolayer structures on a metal surface by self-assembly thu\ creates an ideal interface for the generation of surface plasmons. SPR has been used for the characterization of monolayer thickness of a SAM5 and for the measurement of‘ protein interactions” with SAMs at the metal surface.For the development of optical sensors the production of a monolayer film should provide a fast response for a particular analyte as there should be minimal diffusion effects within the monolayer. Accordingly, SAMs may prove to have advantages over multilayer structured chemical sensors such as Langmuir-Blodgett (L-B) films. We have recently reported the synthesis and characterization of a series of highly substituted phthalocyanine (Pc) molecules each containing either a thiol or disulfide moiety to facilitate the formation of SAMs on gold surface^.^ A preliminary reflection absorption IR (RAIR) spectroscopic study of one of these molecules, a diphthalocyanine disulfide, on a gold-coated optical waveguide established that this spectroscopic method was sufficiently sensitive to detect a Pc SAM film.8 Addition- ally, we have formed a Pc SAM, using a thiol-terminated Pc macrocycle, on a gold-coated optical waveguide.The thiol Pc SAM sensor was used to monitor the concentration of NO2 gas by measurement of evanescent wave excited fluorescence intensity.9 In this current study complementary RAIR and transmission 1R spectroscopic characterization of the diph- thalocyanine disulfide (I), Fig. 1, on a gold-coated glass or silicon substrate has been performed to establish the orientation of the Pc nucleus in relation to the metal surface. The Pc SAM has then been used in a preliminary study to investigate the potential of combining self-assembly technology with SPR for the optical sensing of NO2 gas detected through changes of the surface plasmon reflectivity signal. Experimental Reagents The synthesis and characterization of the 1,1’,4,4’,8,8’,11,1 1’,15,15’,18,18’,22,22’-tetradecakishexyl- 25,25’-(3,3’-dithiodipropy1)diphthalocyanine { (ChHI 3)7P~(CH&3 }2 (1) used in this work has been reported elsewhere.7 All other chemicals were purchased from Aldricli (Milwaukee, WI, USA) unless otherwise stated, and used as received.Preparation of Substrates for IR Transmission Characterization High-purity silicon (Advent Research Materials, Halesworth, Suffolk, UK; orientation 100, purity 99.999%, thickness 0.56I 502 Analyst, October 1996, Vol.I21 mm, one side etched, one side lapped) was cut into two substrates, approximately 35 mm X 25 mm. The two substrates were cleaned in a solution containing SO ml of Millipore water, 10 ml of 30% hydrogen peroxide and 10 ml of 25% ammonia solution at 80 "C for 5 min. The silicon substrates were then cleaned in a solution containing 60 ml of pure water, 10 ml of 30% hydrogen peroxide and 10 ml of hydrochloric acid at 80 "C for 5 min.10 (The hydrogen peroxide, ammonia solution and hydrochloric acid were all Aristar grade and were purchased from Merck, Poole, Dorset. UK). The substrates were then rinsed with Millipore water and dried in a stream of refluxing propan-2-01 (AnalaR grade) vapour. The silicon substrates were then coated with chromium, approximately 20 nm thick (1 6-20 mesh powder, 99.995% purity, Johnson Matthey, Royston, Hertfordshire, UK) to ensure adhesion of the gold to the silicon surface, by thermal evaporation under vacuum.A gold film of approximately 100 nm thickness (99.995% purity, Johnson Matthey) was thermally evaporated on top of the chromium layer. A SAM of 1 was deposited onto one of the substrates (see below) while the other acted as a reference substrate. Preparation of Substrates for RAIR Spectroscopic Characterization Glass microscope slides (Merck) were used as the SAM substrate. The slides were cleaned using a solution of aqueous KOH in methanol (100 g of KOH, AnalaR grade, was dissolved in 100 ml of Millipore water and then diluted with methanol, Distol grade, to a total volume of 250 ml).The slides were rinsed in fresh Millipore water and then dried in a stream of refluxing propan-2-01 (AnalaR grade) vapour. The cleaned dry slides were then coated with approximately 500 nm of gold by thermal evaporation under vacuum. Both IR and RAIR spectra were obtained using a Bio-Rad FTS 40 Fourier-transform infrared spectrometer (MCT detec- tor). RAIR spectra were recorded at an incidence angle of 85" using a Spectra-Tech FT85 grazing incidence reflection ac- cessory. Formation of Pc SAMs on the Gold-coated Substrates The Pc SAM was formed by immersing a freshly prepared gold- coated substrate into a solution of the Pc ( = 3.0 X 10-4 mol din-3) in cyclohexane (spectroscopic grade). To ensure the formation of well organized self-assembled films, the gold- coated substrates were left in the Pc solution for a period of 24 h.10 Reference (blank) substrates were placed in cyclohexane for an equivalent amount of time. Both Pc SAM and reference substrates were washed with cyclohexane and dried in an oven at 25 "C for 2 h. The substrates were stored in clean sample jars. Interaction of NO2 with the Diphthalocyanine Disulfide : VVIVIS Absorption Data UV/vTS absorption spectra were recorded using a Hitachi U3500 spectrophotometer from a 1 X lo-' mol dm-3 solution of the diphthalocyanine disulfide, 1, in cyclohexane. To measure the interaction of NO2 with the Pc molecule, 20 ml of SO Torr (1 Torr = 133.322 Pa) NO2 in air (at STP) was bubbled through the Pc solution for 5 s using a gas syringe and an absorption spectrum was again recorded.Preparation of Self-assembled Pc Monolayers for Surface Plasmon Resonance Glass microscope slides were washed and then polished using a 1 % Decon-90 in Millipore water solution. The glass slides were then sonicated in fresh Millipore water for 30 min. The slides were rinsed in fresh Millipore water and blown dry in clean air. Gold films were thermally evaporated at a pressure of =: 10-5 Ton [Edwards (Crawley, West Sussex, UK) E306A vacuum c6H13Q-\ , 4c6H13 Fig. 1 gas through a solution of 1). Structure of the diphthalocyanine disulfide (I) and the structure of the phthalimide derivative (2) (the proposed oxidation product after bubbling NO2Analyst, October 1996, Vol. 121 1503 evaporator] onto the clean microscope slides. The gold films were coated to a thickness which gave a good SPR profile by experiment. The SPR curves were analysed by a curve-fitting programme and the thickness of the gold was calculated to be =42 nm.The fresh gold films were immersed in a solution of = 3.0 X 10-4 mol dm-3 of the Pc molecule in cyclohexane (spectroscopic grade) ior 24 h. The substrates were then washed in fresh cyclohexane and blown dry in clean air. Gas Sensing of NO2 Using Surface Plasmon Resonance The experimental configuration used to generate the SPR profiles of the Pc SAMs has been reported previously1 1,12 and is shown in Fig. 2. The Pc SAM glass substrate was clamped, with its uncoated side, to one side of a glass prism. Index-matching fluid (refractive index = I .497 at 20 "C, Optaball Radiall, Fibre Optic Centre, Rochester, Kent, UK) was used to link the prism and the glass substrate optically.The prism-glass waveguide was mounted on a computer-controlled rotating table which could rotate the prism by 0 whilst rotating the detector at 20. y- Polarized light from an He-Ne laser (Melles Griot, Aldershot, Hampshire, UK) at h = 632.8 nm was used to generate the SPR. A beam splitter was used to reflect a small portion of the incident radiation onto a reference detector to compensate for any fluctuations in laser intensity. The output signals from the photodiode detector and from a reference detector were digitized and stored in a personal computer which generated each SPR profile. SPR profiles were obtained from blank gold film substrates in order to measure the changes in the reflectivity due to the Pc SAM.The response of the Pc SAM to varying concentrations of NO2 was measured using the experimental system shown in Fig. 2. The glass substrate was mounted on the prism and a Perspex gas cell was clamped onto the surface of the Pc monolayer. The SPR profile of the film was measured by recording the ATR signal of the He-Ne laser as the angle of incidence was varied. In this experimental configuration the prism was fixed while the laser and detector were mounted on rotating stages which were controlled manually using mic- ronietre screws. Again a portion of the incident light was ratioed with the detector output to compensate for any fluctuations in laser intensity. The laser and detector were set at the angle which corresponded to the steepest part of the SPR curve at angles less than the SPR minimum.A gas blender (Model 850, Signal Instruments, Camberley, Surrey, UK) was used to provide varying concentrations of NO2 in oxygen-free nitrogen (OFN). The NO2 gas was supplied in a cylinder (BOC, Guildford, Surrey, UK) at a concentration of 1000 ppmv Fig. 2 The instrumentation used to generate the SPR profiles and study the effects of NO2 gas on a SAM of the diphthalocyanine disulfide deposited on a gold-coated optical slide. blended in nitrogen. Therefore the gas used in this study was a mixture of nitrogen oxide gases with NOz the major component. The Pc SAM was flushed with OFN and the SPR signal monitored over time. This measurement provided a baseline for the measurements with N02.After each NO2 measurement the gas cell was flushed with OFN to allow the Pc SAM to recover to the baseline SPR response. Reference gold-coated slides were subjected to NO2 in OFN in order to obtain a background response. Results and Discussion Fig. 3 shows the UV/VIS absorption spectra from the diphthalocyanine disulfide, in cyclohexane, before and after interaction with NO2. Prior to interaction with NO2 the absorption spectrum shows the characteristic phthalocyanine Q bands (&,lax = 692 and 725 nrn), the splitting being typical of a metal-free Pc centre. Bubbling the NO2 gas through the Pc solution caused the latter to change from blue to colourless, the absorption spectrum showing the complete removal of the Pc Q bands. The solution colour change appeared permanent, with no reappearance of the characteristic Pc spectrum after 1 week.The colourless solution was analysed using CC-MS and the principal oxidation product gave an mlz of 3 15, which has been tentatively assigned to the phthalimide structure 2 shown in Fig. 1. Fig. 4 shows the IR transmission and Fig. 5 the RAIR spectra obtained for the SAM of 1 on the gold-coated silicon wafer and gold-coated glass slide, respectively. The transmission spec- trum shows the IR absorption bands of the alkyl CH stretches (CHZ, v,, at 2931 cm-l, v, at 2847 cm-I; CH3, v,, at 2959 cm--l, v, at 2871 cm-I), a weak band at 3300 cin-l tentatively assigned to the NH stretch, the aromatic CH stretches in the broad envelope between 3030 and 3100 cm-I [Fig. 4(a)] and the ring vibrations in the 'fingerprint' region of the IR spectrum (out-of-plane ring C-H deformation at 884 cm-', ring mode involving skeleton and N-H at 1023 cm-1, in-plane C-H deformation at 1076 cm-1, C=N stretch at 1093 cm-1, C-H deformation or C=N stretch at 1149 cm-I, C-N stretch at 1270 cm-1, isoindole stretch at 1433 cm-1 and N-H in plane deformation at 1498 cm-l) [Fig.4(b)]. However, in the RAIR spectrum only the alkyl CH stretches (CH2, v,, at 29 17 cm- I , v, at 2850 cm-1; CH3, v,, at 2956 cm-I) are evident (Fig. 5). With consideration of the metal surface selection rule these two spectra give complementary data with regard to defining the orientation of the diphthalocyanine molecule on the metal surface. The presence, or absence, of a particular absorption I 1 500 6 0 700 800 Wavelengthlnm Fig.3 UV/VIS absorption spectra from the diphthalocyanine disulfide in cyclohexane before (solid line) and after (dashed line) interaction with NO2 gas.I504 Ancxlyst, October 1996, Vol. 121 0.100 band is indicative of whether the nucleus of the Pc molecule is parallel or perpendicular to the metal surface. As such, the presence of the aromatic CH stretch, the NH stretch and the ring vibrations in the transmission spectrum, and their absence in the RAIR spectrum, suggest that the Pc nucleus is orientated parallel to the gold wrface with the alkyl chains tilted in relationship to both the parallel and perpendicular planes of the metal surface. Such an orientation would suggest that the Pc molecule used in this study self-assembles in a different manner from that observed for a thiol-derivatized Pc.9 Such a difference can be readily attributed to the different peripheral substituent groups of the two Pc macrocycles which are likely to result in a different packing of the molecules on the gold surface.It should be noted that the two substrates, i.e. the glass and silicon, do not affect the self-assembly packing arrangement of the Pc molecule. A RAIR spectrum of 1 on a gold-coated (500 nm) silicon substrate gave a similar spectrum, in the 4000-2300 cm-I region, to that seen on the gold-coated glass substrate (a full RAIR spectrum on gold-coated silicon could not be -. 1 0.01 20 0.01 IS 0.010s 1 I I- - I/- -7-- 4 L- ~- 3300 3200 3100 3000 2000 2800 2700 1023 p, I 0 00x1 I 0 0082 - -I ,~ i I S 0 0 1100 1300 1200 I100 1000 9’00 Wavenumber/cm I Fig.4 wafer. (0) Spectrum between 3350 and 2700 cm between 1600 and 820 c n - I, the ‘fingerprint’ region. l’ran5inission IR spectra from the Pc SAM on a gold-coated silicon and ( h ) spectrum _ ~ - _ _1 0 020 i7- -- ? 3 0 0 3200 31G- - 3000 2900 2800 Wavenurnbedcm- ’ Fig. 5 between 3350 and 2700 cm- I. RAIR Spectrum of the Pc SAM on a gold-coated glass slide obtained as the lower regions of the spectrum were obscured by the occurrence of interference patterns through the partial transmission of the probe beam). Fig. 6 shows the SPR reflectivity curves for a gold-coated slide and that of a slide after deposition of the SAM of 1. It can be seen that the shape of the reflectivity curve changes on deposition of the SAM, with the curve becoming shallower and broader in appearance with a slight increase in the angle of resonance minimum owing to the finite thickness ol’ the self- assembled monolayer.Experiments were conducted on the reference substrates to establish changes of the SPR reflectivity signal in response to gases flowing over the gold surface. On exposure to a flow of OFN the reflectivity signal decreased, reaching a new equilibrium value. However, the original signal recovered, over 2 h, when the OFN was doped with 1000 ppmv NO2. This result implies that the flow of pure OFN removes molecular species, possibly water molecules, from thc gold surface and that the vacant sites are subsequently filled by physisorbed NO2 molecules. The gold-coated slide supporting the Pc SAM was similarly investigated but showed contrasting behaviour. In this instance there was no change in the SPR reflectivity when pure OFN was passed over the SAM surface, but the reflectivity signal changed when NO? was introduced.The results imply that the observed change with NO2 is attributable to an interaction with the Pc monolayer and not with the gold underlayer. The change of the intensity of the SPR mininiuni on interaction of gaseous NO? with the Pc SAM was monitored, at a fixed angle of incidence, as a function of NO2 concentration. Fig. 7 shows the changes in reflectivity for the Pc SAM upon I 1 00 100 300 300 500 Timc/min Fig. 7 Changes of SPR reflectivity over time aftcr interaction of the Pc SAM to varying concentrations ol”Oz gas. Two concentrations of gas were used, i.e.300 and 1000 ppmv. Signals were returned to baseline values alicr purging the Pc SAM sensor with oxygen-free nitrogen. (The experiment was terminated before the final measurement returned to the baseline v a1 ue . }Amlyst, October- 1996, Vol. 121 1 50s exposure to two concentrations of NO? (~lir. 300 and 1000 ppmv) as a function of time. The response time and recovery time of the Pc SAM sensor to the electron acceptor NO? gas was found to be dependent on the gas concentration. For example, for 300 pprnv NO? gas the response time was 56.4 min (recovery time on purging with nitrogen was 94.8 min) while for a sample of 1000 ppmv NOz gas the response time was 4.8 min (the recovery time was 207 min). Other studies with Pc molecules, formulated as L-B films, used for SPR gas sensing have also shown this concentration dependence.' I , l i It has already been established using UV/vlS spectroscopy that at high NO? concentrations the phthalocyanine 1 breaks down into phthalimide derivatives.At a concentration of 1000 ppinv no evidence of such oxidation was obtained from the gold surface as determined by RL41R spectroscopy. Additionally, the reversible nature of the SPR signal on interaction of the NO2 even at the 1000 ppniv level suggests that a chemisorption interaction of the NO? gas with the Pc molecule is occurring rather than a permanent oxidation reaction with the macrocycle. This suggests that the Pc SAM film exhibits similar physico- chemical behaviour towards NO2 gas as that shown by previously reported vacuum-sublimed phthalocyanine films.l4 The preliminary data shown in Fig. 7 suggest that the NO:! sensor may give a reproducible signal as similar changes of reflectivity intensity were obtained for two separate measure- ments at 1000 ppmv. For rneasuretncnt of NO2 at the occupational hygiene levels a limit of detection of at least 3-5 ppmv would be required. Previous conductiometric studies have shown that a highly substituted Po structure. similar to the Pc used in this work, imparts a degree of selectivity towards the targel analyte, ' 5 while the use of metallated Pc molecules, particularly cobalt-, lead- and copper-Pc, l 6 significantly improves the sensitivity of such electrochemical sensors. The addition of such metals to our highly substituted diph- thalocyanine molecule could be readily achieved, although we have yet to determine whether such metallated Pcs would enhance the sensitivity of this optically based sensor system.The equilibrium response time of this sensor system may be improved by the use of other peripheral side chains which should alter the packing arrangement of the SAM on the gold surface. To enhance the speed of recovery it should be possible to desorb the gaseous NO. by hcating the Pc SAM substrate. The gold coating on the substrate could easily fxilitate such he at i n g . In conclusion therefore, the research presented in this paper has shown that a low molecular mass gas can be readily detected by surface Flasmon resonance using a specifically designed self-assembled phthalocyanine monolayer possibly offering a novel form of chemical sensing technology.The authors thank the EPSRC for a studentship for T.R.E.S., the Health and Safety Executive (Sheffield) for partial support of this work and the EPSRC Mass Spectrometry Service (Uni- versity of Swansea) for the GC-MS data reported. References 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 IS 16 ULman, A., An Iiiti.odirc~tion to Ultrutiiin Oiypnic Filnis ,fi.oiqi LarzSmuir.-R(od~~~,tt tu Self nihly, Academic Press, Boston. I99 1. Sagiv. J., J . Am. Ciioni. S o c . . 1980, 102, 92. Nuzzo, R. G., and Allara, D. L., J . Am. C'hrni. So(.., 1983. 105, 448 I . Li, Y.. Maung, J., McIver Jr.. R. T., and Hemminger, J . C., .I. Am. Ciwni. Soc., 1992, 114, 2428. Hanken. D. G.. and Corn. R. M., Ariul. Chcni., 1995, 67, 3767. Sigal, G. B., Bamdad, C.. Barberis, A.. Strominger. J., and Whitesides, G. M., Ancrl. Cheni., 68, 490. Chambrier. I., Cook, M. J.. and Russell, D. A,, Syizt/?c.sis. 1995, 1283. Simpson, T. R. E., Russell, D. A.. Chambrier. I., Cook, M. J.. Horn, A. R., and Thorpe, S. C., Sens. Actuutors B., 1995, 29, 353. Simpson, T. R. E.. Revell, D. J.: Cook, M. J., and Russell, D. A., Ltrnginuir, 1996, submitted. Bertilsson, L.. and Liedberg, B., Lnii,qnruir-, 1993, 9, 141. Lloyd, J. P., Pearson, C.. and Pclty. M. C., Thin Solid Filnix, 1988, 160, 33 1. Zhu, D. G.. Petty. M. C., and Harris, M., Sens. A~~tircitors B. 1990, 2. 265. Vukusic, P. S.. and Sambles, J. R . , Tliiri Solid Fi1ni.s. 1992, 221, 31 I. Rott, B., and Jones, T. A,, Sens. Actuaturs, 1984, 5. 42. Cook. M. 5.. .I. M u t ~ ~ r . Cheni, 1996, 6, 677. Snow, A. W., and Bar gel-, W. R., in Phthaloc~yaniiics-Pi-opri.ties uiid Applicutiuris, ed. I,c~notf, C. C., and Lever, A. B. P., VCH, New York. 1989, pp. 341-392. pp. 237-304.

ISSN:0003-2654    

DOI:10.1039/AN9962101501

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, Octoher 1996, Vol. 121 (1507-1.510) 1 507 Determination of Formaldehyde in Air by lo n-exc I us i o n and lo n-exc h a ng e Chromatography With Pulsed Am perometric Detection ~ ~ Yilin Shi and Brian J. Johnson Department of Cheniistry, University of Nevada, Las Vegas, 4505 S . Maryland Parkway, Las Vegas, NV 89154-4003, USA A method is presented for the determination of formaldehyde in air sample extracts containing aqueous sodium hydrogensulfite. Utilizing the unique properties of its hydrogensulfite complex, formaldehyde is separated from other sample components by ion-exclusion and ion-exchange chromatography, then selectively detected by amperometry at a silver electrode. Optimum sensitivity of detection was found to occur at +0.10 V versus a silver wire reference electrode using a strongly basic background electrolyte.Using ribose as an internal standard, a linear response (r2 > 0.99) was observed for aqueous formaldehyde concentrations in the range 0.02-10.0 mg 1-I; detection limits of < 1 ng for formaldehyde were obtained using a 50 pl sample loop. The short-term reproducibility was better than 5 % (as RSD). Analysis of laboratory air by collection in impingers containing aqueous NaHS03 yielded results consistent with previous literature values. Keywords: Foi-maldehyde determination; ion exclusion; ion exchange; pulsed amperometric detection; hydroxymethanesuljonate Introduction Formaldehyde is a ubiquitous compound in the environment, occurring both naturally and due to anthropogenic inputs. In the atmosphere, formaldehyde is a crucial intermediate in the oxidation of natural methane and non-methane hydrocarbons and is an important component of photochemical smog2 Formaldehyde is one of the largest volume production chem- icals,3 having many commercial uses.Studies in industrial environments4 have demonstrated a statistical link between formaldehyde exposure and incidence of lung cancer, although possible confounding variables were not eliminated. Concentra- tion information for formaldehyde in various matrices is therefore needed in a wide variety of technical, industrial and human health-related applications. While many methods are available for formaldehyde deter- mination, all have potential drawbacks. Most of the methods based on the formation of a colour reagent have chemical interferences, e g ., ozone for the 2,4-dinitrophenylhydrazine (2,4-DNPH) methods (this topic has been addressed for a related fluorimetric method, however),6 nitrate, nitrite, phenol, ethanol, etc., for the chromotropic acid method7 and ethanal, propenal, sulfite, etc., for the pararosaniline method.* In addition to the interferences, all three methods generate large amounts of toxic waste, and the 2,4-DNPH method is prone to high blank values. A method based on oxidation of the formaldehyde to formate by H202, followed by ion chromato- graphic analysis of the formate, suffers not only from the obvious possible contamination by formic acid in the air but also from low recoveries of formaldehyde from various collection substrates.') Detection of formaldehyde by an amperometric detector after chromatographic separation would appear to be an appealing option for a sensitive and selective determination, but initial studies using platinum electrodes in acidic solution yielded detection limits of only about 1.0 mg I-' HCHO, with ethanol and methanol interference.10 Oxidation of HCHO on platinum surfaces is complicated by the formation of a stable CO complex that can lead to poor sensitivity.]' The chromatographic separation of formaldehyde from other polar organic molecules and hydrogensulfite using aqueous eluents is challenging, but is highly desirable to maintain compatibility with amperometric detection and to minimize the production of hazardous waste. Using both ion-exclusion and ion-exchange processes, the properties of the HCHO-S'" complex [hydroxymethanesulfonate (HMSA)] and detection at a silver electrode in basic solution, we have developed and tested a laboratory method for the determination of formal- dehyde that is highly selective, sensitive and linear over a large dynamic range.The method has been applied to some preliminary measurements of HCHO concentrations in indoor air, but the analytical methodology could also be used for other matrices. Experimental Apparatus Chromatographic separations were performed with an isocratic Dionex (Sunnyvale, CA, USA) Qic ion chromatograph vari- ously equipped with one of the following columns or a combination of two of the columns (described below): Dionex AS4A (25 cm, low-capacity strong anion exchanging), Dionex AG 10 (5 cm, high-capacity strong anion exchanging), Dionex AS 1 (25 cm ion exclusion) and Phenomenex Rezex RFQ ( 10 cm ion exclusion).Eluents were stored in polyethylene containers under a nitrogen or helium atmosphere. As shown in Fig. 1 and explained in the Results and Discussion section, modifications to the instrument configuration were subsequently incorpo- rated. Injector ri Acidic eluent 1 P y p I Basic eluent Fig. 1 Table 2. Formaldehyde analytical system. Operating conditions are given inDetection was accomplished with a Dionex pulsed amper- ometric detector (PAD) using both platinum and silver working electrodes and a silver wire reference electrode. ‘The PAD allows the addition of oxidative aiid reductive pulses following the application of the analytical potential to prevent unwanted side-products from accumulating on the electrode and impeding electrode performance.12 Reagents High-purity water (low organic, 18 mQ cm; Barnstead Dubuque, IA, USA) was used for the preparation of all eluents and reagent solutions. The H2SO4 used in eluents was a highly purified trace metal grade (Spectrum Chemical, Gardena, CA, USA). as was the HN03 (Ultrex; J. T. Baker, Phillipsburg, NJ, USA). NaOH (Matheson, Houston, TX, USA) was prepared as a 50%) m/m aqueous solution to precipitate the Na2C03 before diluting to the appropriate strength for use as an eluent. Formaldehyde working standard solutions were prepared from a nominal 1000 mg 1- stock standard solution (standardized by sodium sulfite titration’?) prepared by dilution of formalin (37% m/m HCHO) (Mallinkrodt, Paris, KY, USA).All other reagents were of aiialytical-reagent grade and werc used as received. Samples Except where noted otherwise, all samples were treated with excess NaHSO? (100400 mg I-]) to preserve the formal- dehyde as hydroxymethanesulfonate. Ambient air samples were collected in impingers containing 400 mg 1 - 1 NaHSO? and 4.0 mg 1-’ ribose (2-10 ml liquid volume); flow rates were typically 0.4 1 min-1. (The ribose was used as an internal standard, which is described in the Results and Discussion section). The collection efficiency was evaluated by sampling through two impingers in series. No detectable amounts of formaldehyde were found in the second impinger for typical laboratory air, indicating nearly quantitative collection.Separation Experiments The separation of formaldehyde from other sample components and from potential interferents was first investigated using single columns with a variety of eluents. In the early experiments, solution mixtures containing HCHO, HCOOH, HS07-, Br- and C2O4’- were used to characterize the separation. With the AS4A column, the following eluents were used: 1.8 X lo-? mol I-’ Na2C03-1.7 X mol 1-I NaHCO?, potassium hydrogenphthalate (KHP)-potassium so- dium phthalate (KNaP) (three combinations in the range 10-4--10-3 mol I - I ) and boric acid (0.5-1.0 X lo-? mol 1- range)-NaCI (0.5-1.0 X lo-“ mol 1- I range). With the AGIO column, 5.0 X 10-2 rnol 1-1 NaOH was used, along with some mixed eluents containing NaOH, Na2C03 and NaN03 (added as HN03).Acidic eluents, iiicluding HCl, HN03 and H2S04, were u5ed for the AS1 and Rezex RFQ ion-exclusion columns; mixtures of these acids and their \odium salts ( e g . , HN03 and NaNO?) were also investigated. Detection Experiments Using selected eluents from the separation experiments, studies of amperometric detection by oxidation of formaldehyde were conducted using bath platinum and silver electrodes. The oxidation potentials for formaldehyde were optimized for the best response for each of the various eluent-electrode combina- tions; both pulsed and non-pulsed data acquisition cycles were investigated. System Experiments As described in the Results and Discussion section, considera- tion of the results of the separation and detection experiments led to the experimental design depicted in block form in Fig.1 . The proper eluent concentrations to obtain optimum separations of formaldehyde from sample components and sensitive detection of formaldehyde were refined on the basis of the previous experiinen ts. Po ten tial interfering compounds were investigated, including ethanol, benzaldehyde, 2-methylpro- panal, methanol, ethanol, formic acid and sodium oxalate. Candidate compounds for use as internal standards were investigated to improve the short-term analytical reproducibil- ity (c.g., by compensating for detector drift) and to provide correction for evaporation of impinger solutions due to air sampling. Because of their absence in the analytical samples under consideration and their response to the detector, the sugars xylose, glucose and ribose were tested for their suitability .Results and Discussion Separation Experiments Using the AS4A ion-exchange column, it was possible to separate formaldehyde from the strongly acidic anions (includ- ing CI-, NO?-, S042- and using the C032--HC03- eluent. Here, the HMSA complex is converted on the column into formaldehyde and sulfite, with the sulfite being retained and the non-ionic formaldehyde being essentially non-re- tained. l4 However, weakly ionized species such as HCOOH (i.e., a potential interferent) are not separated from formal- dehyde with this eluent. The use of weaker eluents (e.g., boric acid-NaC1 and KHP-KNaP) effects this separation, but severe peak tailing is encountered, and the strongly acidic anions have very long retention times.Further, no separation between formaldehyde and other non-ionic organic compounds (includ- ing potential interferents) is possible. From the experiments conducted with the AS 1 ion-exclusion column, it was determined that the HMSA complex does not decompose into formaldehyde and hydrogensulfite under the separation conditions employed. When hydrogensulfite is present, formaldehyde injected onto the column is retained, but when excess hydrogensulfite is present, the HMSA anion is excluded by the negatively charged stationary phase and elutes with the void volume. The HMSA anion, being resistant to oxidation,I” could not be detected by amperometry, but its presence was inferred from fraction collection followed by the chromotropic acid colorimetric test? However, formic acid and other carboxylic acids are readily separated from the HMSA anion, as are alcohols and sugars.Detection Experiments Table I contains a suminary of the results for the detection of formaldehyde under various conditions. E l refers to the analytical potential during which the current measurement is taken and E2 and E3 are reductive and oxidative pulses intended to clean the electrode surface; in all cases, t l (duration of the E l pulse) was 240 ms, t2 was 60 ms and t3 was 60 ms. The important points to note are that (1) the best detection for formaldehyde is clearly using the silver electrode with a basic eluent and (2) as noted above, formaldehyde should not be complexed with hydrogensulfi te if amperometric detection is desired.Fortunately, these two requirements are compatible; the HMSA complex is unstable in alkaline solution.Analyst, October 1996, Vol. 121 1509 System Experiments Optimization of separation Because neither the ion-exchange nor the ion-exclusion process alone can accomplish the separation of formaldehyde from all sample components and potential interferents, the separation modes were used in tandem, as illustrated in Fig. 1. To effect a more rapid analysis, the 25 cm AS1 and AS4A columns were replaced with the 10 cm Rezek RFQ and S cm AGlO columns, respectively. Characterization studies of the shorter columns were conducted before the system experiments commenced. 16 From consideration of the previous separation and detection experiments, and refinements based on further optimization of system performance, the conditions in Table 2 were estab- lished.Repr-oducihility In combination with the optimization of separation experi- ments, different eluents were investigated for their impact on detection. In particular, mixed NaOH-Na2C03 eluents were used instead of the pure NaOH eluent indicated in Table 2. The presence of carbonate was found to contribute to a steady decline in the detector response. When carbonate was removed from the eluents, the downward drift in response was slowed considerably but not eliminated. The decreased response is apparently due to slow oxidation of the silver working electrode; the electrode surface turns noticeably darker after a long period (e.,?., time-scale of hours) of operation.Compensation for the remaining small detector drift was accomplished by using an internal standard. Because of their positive response to the detector-eluent combination, their absence in the samples under investigation (e.g., air extracts) and their stable and non-toxic nature, the sugars glucose, xylose Table 1 Detection of formaldehyde by pulsed amperometry Working electrode Eluent Peak response/ E ~ l v E ~ / v E ~ / v nA mg-I 1 Pt 1 x lo-' mol 1-I HC1 +0.40 +1.25 -0.10 0.33 4 X 10-4 mol 1-1 KHP-4 x 10-4 mol 1-1 KNaP +0.40 +1.25 -0.10 0.87 1 X lo-' mol 1-1 Ag 1 X lop7 moll-] 1 X lo-? moll-' 1 x 10-3 moll-] NaOH +0.40 +1.25 -0.10 0.34 HNOT +0.10 +0.09 -1.15 0 NaOH +0.10 +0.09 -1.15 60 - NaOH +0.10 - 60 Table 2 Standard operating conditions for HCHO determination Sepuration- Column 1 Column 2 Dionex AGlO Eluent 1 1 .O X lo-? moll- H2S04 Flow rate I 1.0 ml min-I Eluent 2 2.0 x 10-2 moll-' NaOH Flow rate 2 0.6 ml min-1 Detection- Electrode Silver Potential +o.10 v* Rezex RFQ ion exclusion * Versus a silver wire reference electrode. Pulsed or non-pulsed (see Table I). and ribose were investigated for their suitability. Of these candidate compounds, ribose proved to be the most suitable in terms of its separation from other sample components. Fig. 2 illustrates a typical chromatogram. When the silver electrode is freshly polished and the ribose internal standard is used, the system can function for several hours under the standard operating conditions (Table 2) before the electrode needs to be repolished.Replicate analyses typically yield RSDs of less than 5% (Table 3). Interferences Solutions of various representative compounds at SO mg I-' (e.g., 2-3 orders of magnitude greater than typical analyte concentrations) were run on the system, including methanol, ethanol, ethanal, benzaldehyde, 2-methylpropanal, formic acid and sodium oxalate. There was no response to any of the compounds except formic acid, and its response was at least 400 times lower on a molar basis than that for formaldehyde. Although separated from formaldehyde, formic acid in high concentrations could conceivably interfere with the current method owing to incomplete resolution from the internal standard peak. Although not addressed in this study, the selection of a new internal standard would correct this potential interference.The detector responds weakly to sulfite, but it is completely separated from formaldehyde (Fig. 2). Besides formaldehyde, the only compounds found to date to give an appreciable response are the sugars, and they are easily separated from formaldehyde (e.g., ribose in Fig. 2). Air sample results Least-squares analysis of a typical calibration experiment yielded the response curve R = 16.SC + 0.30 over the concentration range 0.02-10.0 mg 1-I with r2 = 0.9999, where R is the formaldehyde to ribose peak-height ratio and C is the formaldehyde concentration in mg I-'. Table 3 gives the analytical results for measurements of formaldehyde concen- trations in air in a chemistry laboratory and in an anatomy laboratory where formaldehyde was being used for the preservation of biological samples. The range of concentrations 1 b .5 nA I I I I I 3 6 9 12 1s Retentiodmin Fig. 2 Chromatogram of 0.1 mg 1-1 HCHO in 400 mg 1-1 HS03- with 4.0 mg 1-1 ribose as internal standard. Peak I , HCHO: peak 2, ribose; peak 3, sulfite. Conditions are given in Table 2.1510 Analyst, October 1996, Vol. 121 Table 3 Measured formaldehyde concentrations in air Location Date UNLV Chemistry research laboratory 5/10/94 511 9/94 6/29/94 711 3/94 UNLV Anatomy teaching laboratory 6/29/94 6130194 6130194 71 12/94 711 3/94 71 14/94 912 1/94 912 1/94 912 1/94 912 1 I94 * In aqueous sample extract. + In air. Air volumell 3.0 3 .0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 Solution volumelml 2.0 2.0 2.0 2.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 HCHO concentration*/ P.!2 I-' 3 0 f 2 2 1 f l 86+ 1 76+2 32+4 56+2 39f 1 l l 0 f l 102f4 6 4 f 3 88+3 60+ 1 66+2 66+2 HCHO concentration'/ 20* 1 14f 1 14+ 1 13+ 1 27+2 4 7 f 2 32+ 1 92f 1 85 k 3 53+3 73+3 50f 1 55+2 5 5 f 2 I.18 m-? (13-92 pg m-3) compares favourably with the range 18-500 pg m--3 measured in various industrial environment^.'^ Conclusions The proposed method displays excellent selectivity and very good sensitivity for formaldehyde, and does so using much less toxic reagents than most current methods.Although designed for the analysis of air samples, the method should be applicable to other matrices (e.g., food and biological extracts). No problems with long-term reproducibility have been encoun- tered, but repolishing of the silver electrode is required when detection has degraded (usually after 4-6 h of operation).With a different collection technique,ls the method could be used for the determination of formaldehyde in clean air in conjunction with isotopic measurements. Further studies, including com- parison with other techniques, are planned. References Logan, J . A., Prather, M. J., Wofsy, S. C., and McElroy, M. B., J . Geophys. Res., 1980, 86, 7210. Druzik, C. M., Grosjean, D., Van Neste, A., and Parmar, S. S., Int. J . Environ. Anal. Chem., 1990, 38, 495. Kirschner. E. M.. Chem. Eng. News, 1996, 74 (15), 16. Acheson, E. D., Barnes, H. R., Gardner, M. J., Osmond, C., Pannett, B., and Taylor, C. P., Lancet, 1984, i, 61 1. 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 Smith, D. F., Kleindienst, T. E., and Hudgens, E. E., J . Chromatogr., 1989,483,431. Rodler, D. K., Nondek, L., and Birks, J. W., Environ. Sci. Technol., 1993,27, 28 14. Altshuller, A. P., Miller, D. L., and Sleva, S. F., Anal. Chem., 1961, 33, 621. Miksch, R. R., Anthon, D. W., Fanning, L. Z., Hollowell, C. D., Revzan, K., and Glanville, J., Anal. Chenz., 1981, 53, 21 18. Lorrain, J. M., Fortune, C. R., and Dellinger, B., Anal. Chenz., 1981, 53, 1302. Rocklin, R. D., Adv. Cheni. Ser. 1985, No. 210, 13. Okamoto, H., and Tanaka, N., Electrochim. Acta, 1993, 38, 503. Hughes, S., Meschi, P. L., and Johnson, D. C., Anal. Chim. Ac,ta, 1981, 132, 1. Walker, J. F., Formaldehyde, Reinhold, New York, 1964, p. 486. Murphy, A. P., Boegli, W. J., Price, M. K., and Moody, C. D., Environ. Sci. Technol., 1989, 23, 166. Boyce, S. D., and Hoffman, M. R., J . Phys. Chem., 1984, 88,4740. Shi, Y., Master's Thesis, University of Nevada. Las Vegas, 1995, Levin, J . O., Lindhal. R., and Andersson, K., in DifJusive Sampling: an Alternative Approach to Workplace Air Monitoring, ed. Berlin, A., Brown, R. H., and Saunders, K. J., Royal Society of Chemistry, London, 1987, pp. 345-350. Johnson. B. J., and Dawson, G. A., Environ. Sci. Technol., 1990,24, 898. Johnson, B. J., and Dawson, G. A., Analyst, 1990. 115, 1153. pp. 21-28. Paper 6103072F Received May I , 1996 Accepted June 24, I996

ISSN:0003-2654    

DOI:10.1039/AN9962101507

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Aiialyst, Octohei- 1996, Vol. 121 (151 1-1514) 151 1 Simultaneous Determination of Urinary Zinc, Cadmium, Lead and Copper Concentrations in Steel Production Workers by Differential-pulse Anodic Stripping Vol tam met ry The determination of toxic metals in urine is an important clinical screening procedure. In this study, differential-pulse anodic stripping voltammetry on a hanging mercury drop electrode was used for the simultaneous determinations of zinc, cadmium, lead and copper in the urine of 23 production and 23 quality control workers in a steel production plant and their matched normal controls. The urine specimens were pre-treated with a mixed acid solution and Analytical Products Group set-point laboratory standards were used to check the analytical accuracy. The results indicated that the urinary zinc, cadmium, lead and copper levels of the production and quality control workers are significantly higher than those of the controls.The possible connection of these elements with the etiology of disease is discussed. 'The results also show the need for immediate improvements in workplace ventilation and industrial hygiene practices. Keywords : Steel y rodi r c. tioiz n w h cix : u'iflei-t.n tial-17 ii 1s e anodic sti-ippiiig ~1oltc~~~iiricti:y; utoniic uhsoiptiori spctimzetry; imny nictals; wine analysis Introduction The biological monitoring of toxic metals in urine has become a matter of wide interest owing to the toxicity of these metals and their influence in controlling the course of biological processes.' It is because of its speed, simplicity, low cost, high sensitivity and ability to determine a number of metals simultaneously' 5 that differential-pulse anodic stripping vol- tamiiietry (DPASV) has been widely used for measuring toxic metals in various matrices such as body fluids.In DPASV, metal ions are reduced and amalgamated at a hanging mercury drop electrode (HMDE) or a mercury-film electrode (MFE) during pre-electrolysis at a suitable applied potential. The reduced amalgamated metals are then reoxidized by means of a potential ramp imposed between the working electrode (HMDE) and a platinum rod electrode. The HMDE is widely used as a working electrode in anodic stripping voltammetry (ASV).h ASV analysis has been shown to be applicable to the simultaneous determination of metals in urine, and advances in microcomputer technology have made ASV more powerful. In order to decrease the effects of water- soluble proteins and intermetallic interferences in the voltam- metric determination, a variant of the standard addition calibration procedure for determinations by ASV was adopted in order to eliminate the background current.7 In this method, two deposition times and two standard additions were used in the determination of unknown concentrations. In DPASV, the background currents are eliminated electronically.This has led to substantial improvements in both accuracy and precision. Kaohsiung is the most important industrial area in Taiwan, and it is also the most affected by air pollution and industrial wastes (most of the pollutants being heavy metals) which are detrimental to human health.The aim of this study was to examine trace elements (Zn, Cd, Pb and Cu) in the urine of steel production workers and quality control workers to evaluate the degree of their exposure in these working environments. These data can provide guiding references to occupational diseases and for pollution control. Experimental Apparatus A Milestone (Bergamo, Italy) microwave digestion system was used for sample digestion. DPASV was performed with a Metrohm 646 VA processor fitted with a 647 VA stand. Its central element is the multi-mode electrode (MME), which combines the dropping mercury electrode (DME) and the hanging mercury drop electrode (HMDE) in a single unit. A rotating disc electrode (RDE) can also be fitted in the sland. A salt bridge filled with 3 mol 1- I potassium chloride served as a link between the reference electrode and the working electrode (HMDE) and a platinum rod was used as the auxiliary electrode.Dissolved oxygen was removed from the urine samples by purging with purified nitrogen (99.999%) through the mcasur- ing vessel for 5 min. During the experiments, nitrogen was passed over the solution to prevent oxygen interference. All glassware and polyethylene bottles were soaked in 2 mol 1 - 1 nitric acid for at least 7 d, rinsed several times with de-ionized water, soaked in de-ionized water and finally soaked in 0.1 moll- hydrochloric acid before use. The measuring vessel and capillary were treated with dimcthyldichlorosilane at regular intervals.Reagents and Standard Solutions All acids were of Suprapur grade (Merck, Darmstadt, Ger- many). Sodium acetate and mercury were of analytical Suprapur grade (Merck). De-ionized water (1 8.2 MQ cm- I ) , prepared with a Millipore (Bedford, MA, USA) Plus Milli-Q ultra-pure water system, was used throughout. Stock standard solutions containing 1000 ppm of Zn, Cd, Pb and Cu were prepared from Merck Titrisol standard solutions ( I .000 k 0.002 g) by adding 5.0 in1 of concentrated nitric acid and then diluting to 1 1 with de-ionized water. Working standard solutions (1 ppm) were prepared daily by appropriate dilution arid stored at 4 "C. Mixed standards for the standard additions method were Zn 25, Cd 2.5, Pb 5.0 and Cu 2.5 pg I--'. Other solutions being used were acetate buffer solution (pH 4.64), 3 mol I-] potassium chloride solution and 0.1 mol I lead ion standard solution.Sample Pre-treatment The urine specimens were filtered through a 0.45 X lo--" m Millipore membrane filter and concentrated nitric acid was1512 Anulyst, Octohet- 1996, Vol. 121 added to the aliquot for future analysis. The acidified urine specimen was refrigerated at 4 "C for no longer than 2 weeks or frozen at -20 "C for a longer storage period prior to analysis. Microwave dissolution is particularly suitable for the rapid preparation of samples for instrumental analysis and its uses for preparing analytical solutions for AAS and DPASV are well documented.8 Volumes of 5 ml of satnple and 10 ml of a mixed acid solution were transferred into a I25 ml pressure-resistant PFTE bottle.The sample was digested either at 300 W for 4 min or at 600 W for 2 min, to move the interfering matrix within the samples. The digested solution was evaporated almost to dryness to remove excess acid and then diluted to 25 ml with de- ionized water. Voltammetric Procedure Before the analysis of sample solutions, the accuracy of the 646 VA processor was checked with a lead standard solution prepared by mixing 1.0 ml of 3 mol 1-l KC1 solution, 20 ml of de-ionized water and 20 yl of 0.1 mol I-' Pb standard solution. Sequential simultaneous determinations of zinc, cadmium, lead and copper in urine samples of workers were then performed by DPASV. The optimum experimental conditions were established as follows: the potential was swept using differential-pulse modulation (DPASV) with a pulse rate of 3.33 s-l, a scan rate of 10 mV s-I and a pulse amplitude of 50 mV; two repetitions were made.The standard additions technique was used to give the concentrations of zinc, cadmium, lead and copper simultaneously when a sweep potential was applied between - 1.150 V and 160 mV (for zinc - 1.150 V to -800 mV, for cadmium -800 mV to -450 mV, for lead -500 mV to -200 mV and for copper -200 mV to 160 niV). To perform an analysis sample, 1 nil of the pretreated sample was pipetted into a measuring vessel and the pH was adjusted to 4.5 by addition of 2 ml of acetate buffer. De-ionized water was then added to give a total solution volume of 20 ml. The solution was then de-aerated with pure nitrogen for 5 min, followed by deposition at the HMDE.DPASV was performed and the results of the DPASV of the samples, together with two subsequent standard additions, were recorded. Further, Standard Trace Metals 7879 Level IT [Analytical Products Group (APG), Belpre, OH, USA) were used to check the analytical reliability, following the same procedures, and all experiments were conducted at ambient laboratory temperature (25 "C). Results and Discussion Our goal was to accomplish the simultaneous multi-element analysis of urine in a routine clinical laboratory situation. To assess the reliability of the DPASV approach, we evaluated critical factors such as detection limit, range of calibration, cost, accuracy and precision, which might affect the analysis of urinary Zn, Cd, Pb and Cu in exposed workers.Analytical Reliability The precision and accuracy of the analytical method were checked with APG Standard Trace Metals 7879 Level 11. Table 1 lists the analytical data obtained by DPASV, indicating that this method was reliable for analysis. Influence of Sample pH Various pH conditions had been reported for ASV. Kemula and Kubiky acidified the sample to approximately pH 1 with hydrochloric acid prior to the analysis, whereas Copeland et u1.Io preferred an acetic acid-acetate buffer for analysis. Franke and De Zeeuw I I used a combination of these approaches with different reagents. Lund and Eridsen1' used a sodium acetate buffer (pH 6.5) with some samples but could not obtain satisfactory results. In this study, the concentrations of Zn, Cd, Pb and Cu in the urine of workers were measured in the pH range 1.18-5.92, and the optimum pH for analysis was found to be 4.5.Copper was shown to interfere with the analysis at pH 2 2 . Standard Additions Method The voltammograms of Zn, Cd, Pb and Cu determinations are shown in Fig. 1. The results indicate that analysis of the samples by the standard additions method could achieve high sensitivity (slopes: Zn 2.276, Cd 3.637, Pb 6.876 and Cu 2.452 pg 1LA-I). Metal Concentrations in Urine Specimens Urine specimens from 23 production workers, 23 quality control workers and 23 unexpoTed normal controls were analysed by DPASV. The difference between the results obtained in this study and those in the literature was analysed by the Mann-Whitney test, using the p value as a measure of significance.Regrecsion relationships, using FAAS- or ETAAS-measured data as the independent variables and DPASV-measured data as the dependent variables, are shown in Table 2. Agreement between the results was good for Cd, Cu, Zn and Pb. Table 3 shows that the range of urinary zinc levels in normal controls was 271.5-650.9 yg I-', with a mean value of 433.7 k 122.8 yg 1-I. The normal urinary zinc values of 500 yg 1- reported by Kimberly and Paschal17 and 361 f. 228 yg 1 1 reported by Abdulla and Chimielnickal4 were in good agree- ment with our results. The urinary zinc levels of quality control workers (Table 3) and production workers were 8 18.4 * 238.1 and 1012.1 k 393.9 yg 1-1, respectively. Both results were significantly higher than that of the present normal controls (17 <0.01).The exposed mean values reported by Abdulla and Chimielnicka14 (524 k 185 yg 1- 1 ) (p < 0.01) and Kimberly and Paschal'? (700 yg 1-I) were in good agreement with those for quality control workers but lower than those for production workers. However, Adeloju et al.Is reported that the average urinary zinc level was 1500 k 500 pg 1- I , a value higher than ours (p <0.01). The normal range of urinary rinc value is 140-800 yg 1- I . There were 7 (30%) and 17 (74%) persons with values above 800 yg 1-1 in the quality control workers and production workers, respectively. Two production workers had values less than 140 pg 1- I . The range of urinary cadmium levels in normal controls was 0.68-6.91 pg l - l , with a mean value of 3.49 k 2.1 I pg l-l, which was in good agreement with our previous work16 and with that of Copeland et ~ 1 .~ ~ ) (2.64 k 2.45 pg I-'). The levels of urinary cadmium level in quality control workers and produc- - - - Table 1 Accuracy of Zn, Cd, Pb and Cu determination in Standard Trace Metals 7879 Level II* Zn/pg 1 - I Cd/pg 1- ' Pb/@ 1- ' cu/pg 1-1 Measured value 139.5 k 13.6 116.9 k 1.2 222.3 f 18.1 118.8 f 5.7 Certified value 139.4 k I 1.6 117.4 k 7.7 223.4 k 19.9 122.5 It 6.4 '' Each value is the mean k s of six runs by DPASV.Anulyst, October 1956, Vol. 121 1513 tion workers were 7.79 k 2.39 and 9.67 k 5.08 pg 1-1, respectively. Both results were significantly higher than that of normal controls (p < 0.0 I). This is in good agreement with our previous report17 and the exposed mean values reported by Golimowski et ~ 1 .~ 7 (8.3 k 0.8 pg 1-I) and Ong et ~ 1 . ' ~ (> 2 year exposure, 9.3 pg 1-I). The reported exposed mean value reported by LinI') (15.2 f 4.8 pg I-') was significantly higher than that in the present study (p <0.01). A urinary cadmium concentration of G20 pg I--] is generally considered as normal. In this study, only three production workers showed cadmium concentrations in excess of the normal level, i.e., 13% showed >20 pg 1-1. 2.50 2.25 2.00 1.75 1.50 5 1.25 : 1.00 .75 .50 .25 00 -1 250 1 225 200 : 2 1251 50 751 0 Zn 250 225 200 175 150 5 125 100 75 50 25 00 -800 -; a - The uptake of cadmium following environmental or occupa- tional exposure results in a gradual accumulation in the liver and kidney.with the eventual production of kidney dysfunction? It is also linked with respiratory ailments, hypertension, and damage to bones.21 Itai-Itai disease and paralysis. In this study, among the three production workers with cadmium concentra- tions >20 pg 1-1, one had proteinuria and abnormal liver function, another had hypertension arid the third had proteinuria with hypertension. However, none of the workers in our study had Itai-Itai disease or paralysis. The range of urinary lead levels in normal controls was 3.7-56.7 pg 1-l, with a mean value of 33.0 k 15.7 pg 1-- I , which Pb Cd 5 -450 250 1 cu -200 EImV -1 60 Fig. 1 Voltammogram of Zn, Cd, Pb and Cu in thc samc urine sample by the standard additions method. Table 2 Regression relationships for urine samples analysed by AAS (x) and DPASV (y) Normal controls' Quality control Production workers' urine ( n = 23) workers' urine ( n = 23) urine ( n = 23) Metal Regression equation I' Regression equation I' Regression equation I' Zn y = 0.867.~ + 53.656 0.99 y = 0.879~ + 88.573 0.99 y = 0.959~ + 48.83 0.993 Pb = 0.985~ + 0.873 0.9X9 y = 0.91 1s + 3.795 0.972 y = 0.965 + 1.432 0.994 Cd V = 1.025~ - 0.046 0.997 J = 0.982~ + 0.189 0.996 y = 0.971~ - 0.399 0.996 cu J = 0.964~ + 0.574 0.996 y = 0.989~ + 0.931 0.996 y = 1.018~ - 0.098 0.9961514 Atiulyst, October 1996, Vol.121 Table 3 Metdl concentration\ ' in urine $pecimen5 determined by DPASV Group Znlpg 1 ~ Norinal control 433.7 f 122.8 (27 I .5-650.9) 818.3 f 238.1' Production worker\ 1012.1 f 393.91 Detection limit 0.207 Quality control workers (448.9-1 320) (128.1-1575) Cdlpg I- 1 3.49 f 2.1 1 (0.68-6.9 1 ) 7.79 f 2.39' (3.30-10.56) 9.67 k 5.08' (3.49-23.12) 0.064 Pblpg I-' 33.0 f 15.7 (3.7-56.7) 47.5 * 7.4' (3.5.9-66.4) 52.0 f 18.5' ( 28.9-8 5.6) 0.033 Culpg I-' 12.9 k 4.6 34.1 k 11.6' (15.9-62.3) 37.2 f 18.0' ( I 6.5-77.1 ) (6.7-27.3) 0.04 1 ' Each v;llue is the iiieitn k s with the range shown in parentheses. Each is con1piired with thc corresponding normal control value by Student's t-tcst, 1' < 0.01.were in good agreement with our previous study1" and the value reported by Burguera ct u/.?? (39.0 k 8.3 pg I-'). The urinary lead values of quality control workers and production workers were 47.5 f 7.4 and 52.0 f 18.5 pg 1 - 1 .re5pectively. Both were 5ignificantly higher than that of the normal controls (p < 0.01). The result was in good agreement with our previous study16 and the exposed mean value of lead reported by Lini0 (58.8 k 2 I .3 pg I 1). The expmed mean value reported by Burguera et ~ 1 1 . 2 ~ (71.7 k 26.3 pg I-I) was higher than that in the present study (p < 0.01 ). The method de\cribed above has proved to be suitable for this work because it covers the range critical for lead poisioning in humans, a\ indicated in the literature:' 80 or 65 pg 1- 1 is normal, 80 or 65-150 pg 1-1 acceptable, 150-200 pg 1-1 excessive and > 250 pg I-' dangerous. Normal levels in urine are < 80 pg I-'. In this study only four production workers showed values in excess of normal level, i.e..17% showed > 80 pg I-'. The effect\ of lead toxicity include impaired blood synthesis, renal failure, gastrointeqtinal dy\function, hypertension, periph- eral neuropathy and brain damage. In this study, among the four production worker5 with value greater than 80 pg 1-l- two had hypertension, one had proteinuria and one had both Iiyper- tension and proteinuria. However, none of the workers had brain damage. The range of urinary copper levels in normal controls was 6.7-27.3 pg 1 - 1 with a mean value of 12.9 k 4.6 pg 1 1 , which were in good agreement with the reported data by Marshall and OttawayZi (10-30 pg I-I), Abdulla and Chmielnickai4 ( 1 2.0 f 7.5 pg 1 1) and Dubez-' ( 1 1.6 5 9.2 pg 1- I ) . The urinary copper levels in quality control workers and production workers were 34.1 f 11.6 and 37.2 f 18.0 pg I 1, respectively.Both were significantly higher than that of the present normal controls (p < 0.01). The result was in good agreement with the exposed mean copper level reported by Marshall and OttawayZ3 (38.9 k 33.1 pg 1 I ) . A tolal copper content in urine of d 15.5 pg I-' is generally considered as normal. In the present study, all of the workers showed concentrations in excess of the normal level. Copper poisoning leads to a variety of toxic effects, such as hepatic necrosis. gastrointestinal bleeding, azotaemia, hae- maturia, hypotension, coma and death.25 In this study, two worker\ had hypotension which may be related to copper poisoning. However, none of the workers had Wilson's disease, haematuria, azotaemia or hepatic necrosis.Conclusions The results showed that the levels of zinc, cadmium, lead and copper in the production workers and the quality control workers were all significantly higher than those in the normal controls. The results indicate the need for an immediate improvement in workplace ventilation and industrial hygiene practice. The DPASV method is not only cheaper and simpler but also more rapid and accurate than AAS. Hence the procedure provides important potential for the surveillance of occupa- tionally exposed persons and extended ecotoxicological base- line studies in man. This work was supported by Kaohsiung Medical College, Taiwan. References 1 2 3 4 5 6 7 8 9 10 I I 18 19 20 21 22 23 24 25 Bond, A. M., and Reust, J. B., Anal. ('hini.Artu, 1984, 162, 389. Staubcr, J. Id.. and Florence, T. M., Anul. C'Iiirn. Actu, 1990, 237, 177. Nurnberg, H . W., A n d . Chini. Actu, 1984. 164, 1. Daniele, S.. Raldo, M. A., Ugo. P., and Mazzocchin, G. A., A w l . Cliini. Acfu, 19x9, 219, 19. Pan, 'r. C., Horng, C. J., Lin. S. K., Lin. T. H. and Huang, C. W.. B i d . Truc~ Eleni. Rrs., 1993, 38, 233. Onar, A. N., and Teniizer, A., A/ztr/y,si, 1987, 112, 227. Whang, C. W.. Page, J. A., Van Loon. G., and Griffin, M. P., Anul. Chcm.. 1984, 56, 539. Blust, R., van der Linden, A., and Decleir, W., At. Spwtrosc., 1985, 6, 163. Kemula, W.. and Kubik, Z., Nutui-c (London), 1961. 189. 57. Copeland. T. R.. Christie, J. H., Osteryoung, R . A.. and Skogerboe, K. K., AIIN/. Circm., 1973, 45, 2 17 1. Franke, J. P., and De Zeeuw, R. A.. Phuiw. Wcvkhl., 1976, 111, 725. Lund. W., and Eridxen, R., A n d . Chirri. Actu, 1979, 107, 37. Kimberly, M. M., and Paschal, D. C., Anul . Chiin. Actci. 1985, 174, 203. Abdulla, M., and Chniielnicka, J., Biol. Tr-utc.e Elcm. Rcs., 1990, 23, 25. Adeloju. S. B., Bond, A. M., and Briggs, M. H., Atzrxf. Chrm., 1985, 57. 1386. Pan. T. C.. Horng, C. J., and Lin, S. R., Kuohsiuncq .I. Mod. Sci.. 1993, 9, 643. Golimowski, J., Valents. P., Stoeppler, M., and Numberg, H. W.. Tuluntu, 1979. 26, 649. Ong, C. N., Chua. L. H., Lee, B. L., Ong, H. Y.; and Chia, K. S., .I. Atrul. Toxicol., 1990. 14, 29. Lin, S. M., Antxl. Sci.. 1991, 7, 155. Lauwerys. R., in The Ciwuiistry, Biocheruiistry trnd BioLo<qy of Cudniium, ed. Webb, M., ElsevierINorth-Holland, New York, 1979, Nomiyama, K., Sci. Totul Environ., 1980, 14, 199. Burguera, J. L., Burguera, M., Cruzo, L. L., and Naranjo, 0. R.,Anul. Chini. Actu. 1986, 186, 273. Marshall, J., and Ottaway, J. M., Tulutztu, 1983, 30, 571. Dube, P., At. Sprc.ti.osol-., 1988, 9, 55. Prasad, A. S . , Truce Elements urid It-on in Huniun Metuholism. Plenum Press, New York, 1978. pp. 433-1153. Puper 6103033E Rec.eiid April 30, 1996 Accepted July 3, 1966

ISSN:0003-2654    

DOI:10.1039/AN9962101511

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, October 1996, Vol. 121 (1515-1518) 1515 Determination of Ultratrace Amounts of Copper(i1) by Its Catalytic Effect on the Oxidative Coupling Reaction of 3-Met hyl-2-benzot hiazolinone Hydrazone With N-Et hyl-N-(2- hyd roxy-3-sulf opropyl)-3,5- dimethoxyanil ine Satomi Ohno", Norio Teshima", Tsuyako Watanabe", Hideyuki Itabashi",* Shigenori Nakanoh and Takuji Kawashimaa,? ulaboratory of Analytical Chemistry, Department of Chemistry, University of Tsukuha, Tsukuha 305, .lapan Chemic-al Institute, Faculty oj Education, Tottori University, Koyama-cho, Tottori 680, Japun A spectrophotometric method was developed for the determination of ultratrace amounts of copper(rr) based on its catalytic effect on the oxidative coupling reaction of 3-methyl-2-benzothiazolinone hydrazone with N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3,5-dimethoxyaniline to produce an intensely coloured dye (A,,, = 525 nm) in the presence of hydrogen peroxide. In this reaction, pyridine acted as an effective activator for the catalysis of copper(rr).By measuring the absorbance of the dye, copper(r1) can be determined at the 0.002-0.1 ng cm-3 (3.1 x 10-11-1.6 X 10-9 mol dm-3) level. The relative standard deviation for ten determinations of 0.06 ng ~ m - ~ of copper(r1) was 2.6%. The proposed method was successfully applied to the determination of copper(1r) in tap water and biological material. Keywords: Catalytic method; copper([[) detemination; 3-methyl-2-henzothiazolinone hydrazone; modified Trinder's reagents; N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3,5- dimethoxyaniline 1 , 10-phenanthrolinel4 acted as activators for the catalytic action of copper(i1). Modified Trinder's reagents have been proposed as effective hydrogen donors for oxidative coupling reactions of 4-amino- antipyrine or MBTH and these reactions have been applied to the determination of hydrogen peroxide.15-19 These reagents have the following advantages: ' ~ 1 6 (a) good solubility in water; (b) oxidative coupling reactions over a wide pH range; and (c) highly sensitive colour reactions.This paper describes the catalytic determination of ultratrace amounts of copper(i1) based on the oxidative coupling reaction of MBTH with modified Trinder's reagents in the presence of hydrogen peroxide. Levels of copper(I1) as low as mol dm-3 (pg cm-3 levels) can easily be determined by using N-ethyl-N- (2-hydroxy-3-sulfopropyl)-3,5-dimethoxyaniline as the Trin- der's reagent and pyridine as an activator.The proposed method was successfully applied to the determination of copper(r1) in tap water and in National Institute for Environmental Studies (NIES) CRM No. 5 Human Hair. Introduction Catalytic methods for the determination of trace amounts of copper(r1) have been reported by several workers using various indicator reactions. 1-14 Lopez et ul.5 developed a method for the determination of 25-380 ng cm-3 levels of copper(I1) by using the aerial oxidation of dimedone bisguanylhydrazone in the presence of pyridine as an activator. Casassas et al.7 proposed a method for the determination of copper(I1) at levels as low as 0.06 ng cm-3 based on the oxidation of 3-hydroxy-2-naphthoic acid by hydrogen peroxide in an ammoniacal medium.Velasco et a1.10 proposed an indicator reaction, viz., the oxidation of 3-hydroxybenzaldehyde azine by potassium peroxydisulfate in an ammoniacal medium, for the determination of copper(I1). We have also developed catalytic methods for sub-nanogram to nanogram levels of copper(u) by the oxidative coupling reactions of N,N-dimethyl-p-phenylenediamine, N-phenyl- p-phenylenediaminez and 3-methyl-2-benzothiazolinone hydra- zone (MBTH)4,14 with N,N-dimethylaniline (DMA). In these reaction systems, ammonia,'-*,4.13 2,2'-bipyridine14 and * Present address: Department of Applied Chemistry, Faculty of Engineering, Gunma University, Kiryu, Gunma 376, Japan. I To whom correspondence should be addressed. Experimental Reagents All reagents were of analytical-reagent grade and used without further purification.Water used to prepare the reagent and buffer solutions was obtained from a Milli-Q water purification system (Millipore, Tokyo, Japan). A commercially available copper solution for atomic absorp- tion spectrometry (1000 pg cm-3) (Kanto Kagaku, Tokyo, Japan) was used and working standard solutions were prepared fresh daily by serial dilution of the copper solution with 0.05 mol dm-3 sulfuric acid. A stock solution of 5 mmol dm-3 MBTH was prepared by dissolving 0.23 g of 3-methyl- 2-benzothiazolinone hydrazone hydrochloride (Tokyo Kasei, Tokyo, Japan) in 200 cm3 of water. The Trinder's reagents were obtained from Dojindo Laboratories, Kumamoto, Japan and 5 mmol dm-3 stock solutions were prepared by dissolving appropriate amounts of N-ethyl-N-(2-hydroxy-3-sulfopropyl- )aniline sodium salt (ALOS), N-sulfopropylaniline sodium salt (HALPS), N-ethyl-N-(2-hydroxy-3-sulfopropyl)-m-toluidine sodium salt dihydrate (TOOS), N-ethyl-N-(2-hydroxy-3-sulfo- propyl)-3,5-dimethylaniline sodium salt monohydrate (MAOS), N-ethyl-N-(2-hydroxy-3-sulfopropyl)-m-anisidine sodium salt dihydrate (ADOS) and N-ethyl-N-(2-hydroxy- 3 -sulfopropy l)-3,5 -dimethoxy aniline sodium salt (DAOS) in1516 Anulyst, October 1996, Vol.121 water and stored in a refrigerator. A stock solution of 0.01 rnol dm-3 DMA was prepared by dissolving N,N-dimethylani- line (Wako, Tokyo, Japan) in 0.1 mol dm-3 hydrochloric acid. A 0.05 mol dm-3 MOPSO buffer solution was prepared by dissolving 5.63 g of 3-(N-morpholino)-2-hydroxypropanesul- fonic acid (Dojindo Laboratories) in 500 cm3 of water.A 3 rnol dm-3 hydrogen peroxide solution was prepared by diluting the commercial 30% solution with water. Stock solutions of pyridine (0.1 mol dm-x), 2,2’-bipyridine (0.01 mol dm-3), 2,2’,2”-terpyridine (0.01 rnol dm-3) and 1,lO-phenanthroline (0.01 rnol dm-3) were prepared by dissolving appropriate amounts of each reagent in water or 0.1 mol dm-3 hydrochloric acid. Working solutions of these reagents were prepared by suitable dilution with water. Apparatus A JASCO (Tokyo, Japan) UVIDEC-320 spectrophotometer with 10 mm glass cells, a Horiba (Kyoto, Japan) F8-AT pH- meter and a Taiyo (Saitama, Japan) C-630 thermostat were used.Recommended Procedure To an aliquot of copper(r1) solution in a beaker, 3 cm3 of DAOS ( 5 mmol dm-3), 3 cm3 of MBTH ( 5 mmol drn-”, 3 cm3 of MOPSO (0.05 mol dm-3) and 5 cm3 of pyridine (0.1 mol dm-3) were added; the pH of the solution was adjusted to about 5.6 by adding 0.5 mol dm-3 sulfuric acid. The solution was transferred into a 50 cm3 calibrated flask and diluted to about 44 cm3 with water. The reaction was then initiated by adding 5 cm3 of hydrogen peroxide (3 mol dm-3) solution, diluting to the mark with water and mixing vigorously. The mixed solution was kept at 30 “C in the thermostat during the reaction. Exactly 7.5 min [for 0.02-0.1 ng cm-3 of copper(11)1 and/or 15 min [for 0.002-0.01 ng cm-3 of copper(rr)] after the addition of hydrogen peroxide, the reaction mixture was pipetted into a 10 mm glass cell and the absorbance at 525 nm was measured against distilled water as a blank.The net absorbance was obtained by subtracting the blank absorbance. Results and Discussion Catalytic Effect of Copperfrr) on Coupling Reaction of MBTH With Modified Trinder’s Reagents The catalytic effect of copper(1r) on colour development was examined under the following conditions: in the absence of activator, CCJI = 5 ng cm-3, CMBTH = 0.4 mmol dm-3, rnol dm-3 and reaction temperature = 30 “C. The results are shown in Fig. 1. The data for DMA are also included for comparison. The absorbance at each A,,, after 30 min is plotted as a function of the pH of the solution. As can be seen in Fig. 1, colour development was maximum in the pH range 5-7, and the coupling reactions of MBTH with DAOS, TOOS, ADOS and MAOS were more sensitive than that of DMA for the determination of copper(r1).The net absorbances (AA) at each A,,, are shown in Fig. 2. DAOS is the most sensitive reagent for this method: the value of AA is more than 30 times that of DMA at 5 min. Therefore, DAOS was selected for the procedure. CDMA or modified Trinder’s reagents = 0.4 rnmol dm-3, CH202 = 0.03 Effect of Experimental Variables The effect of reaction variables on the colour development for the uncatalysed and catalysed reactions was examined for a constant time of 5 min at 30 “C in the presence of 0.6 ng cm-3 of copper(I1). The effect of the concentrations of various ligands such as pyridine (py), 2,2’-bipyridine (bpy), 2,2’,2’’-terpyridine (tpy) ~ ~~~ and 1,lO-phenanthroline (phen) was examined in relation to their use as possible activators. The results are shown in Fig.3. Py, bpy and phen acted as activators whereas tpy acted as an inhibitor. The stability constant of the copper(I1)-tpy complex is larger than that of the copper(1)-tpy complex; tpy seemed to act as an inhibitor rather than an activator, forming a stable complex with copper(11).’4 Of the ligands tested, py was the most effective activator for the catalysis of copper(I1); the absorbance obtained in the presence of 0.01 mol dm-3 py was about three times higher than that in its absence. A 0.01 rnol dm-3 py concentration was selected for the procedure since the reagent blank remained constant over the py concentration range examined.In the presence of py as an activator, the effect of pH was examined over the range 5.0-7.0. The maximum colour development for the catalysed reaction was obtained in the pH range 5.2-6.1; the absorbance decreased rapidly outside this pH range. Since the reagent blank increased slightly at a pH of about 6.0, a pH of about 5.6 was selected for the proposed method. The effect of MBTH concentration was examined over the range 0.1-3 mmol dm-3. The absorbance increased gradually with increasing MBTH concentration up to 0.3 mmol dm-3, and maximum absorbance was obtained over the range 0.3-3 mmol dm-3. A 0.3 mmol dm-3 MBTH concentration was selected since the reagent blank remained almost constant. The effect of the DAOS concentration was examined over the range 0.06-1 mmol dm-3.The absorbance increased gradually with increasing DAOS concentration up to 0.1 mmol dm-3, and maximum absorbance was obtained over the range 0.1-1 mmol dm-3. The reagent blank remained constant over the DAOS v) P 4 0 2 4 6 8 1 0 2 4 6 8 1 0 2 4 6 8 1 0 2 4 4 8 1 0 PH 2 4 6 8 10 PH Fig. 1 (ng ~ r n - ~ ) : (e) 0; (0) 5. Other conditions as in the text. Effect of pH on the colour development. Copper(n) concentrationAnalyst, October 1996, Vol. 121 1517 - ADOS - concentration range examined. A 0.3 mmol dm--3 DAOS concentration was used for the procedure. The effect of hydrogen peroxide concentration was examined over the range g 2 5 1 vl CI vl a" 4 a 0 g 2 3 51 a" Q a v, 0 l2 e 5: T l 2 4 a * 0 Fig. 2 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Time / min 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Time / min Time course of the colour development; Ccuii = 5 ng cm-3.Other conditions as in the text. o ? f 5 . - : J 0- -3 -2 -1 -5 -4 -3 log(Cpy / mol drC3) log(Cbpy / ma1 dm-3) -6 -5 -4 -5 4 -3 log(Cphen / ma1 dmP3) log(Ctpy / rnol d ~ n - ~ ) Fig. 3 Effect of py, bpy, phen and tpy concentrations on the colour development for 0 ng cm-? (0) and 0.6 ng cm-7 (0) of copper(I1). CMBTH = 0.3 mmol dm-?. CDAOs = 0.3 mmol dm-?, C1j202 = 0.3 mol dm-?, pH = 5.6. The broken lines denote the absorbance obtained in the absence of the ligands. 0.03-0.6 mol dm-3. The absorbance increased with increasing hydrogen peroxide concentration, and an almost constant absorbance was obtained in the concentration range 0.3-0.6 rnol dm-3.A 0.3 mol dm-3 hydrogen peroxide concentration was used, taking into consideration reproducibility. Calibration Graphs Calibration graphs obtained with the proposed procedure were linear over the range 0.02-0.1 ng cm-3 (3.1 X 10--10-1.6 X lO--9 mol dm-3) of copper(1r) for a reaction time of 7.5 min and over the range 0.002-0.008 ng cm-3 (3.1 X 10-I '-1.3 X 10-I0 rnol dm-3) of copper(n) for a reaction time of 15 min. The reproducibility was satisfactory with a relative standard devia- tion of 2.6% for ten determinations of 0.06 ng cm-3 of copper(i1). Interferences The effect of foreign ions on the determination of copper(1r) at the 0.2 ng cm-? level was investigated. A +S% error was considered to be tolerable.The results are summarized in Table 1. Iron(m), vanadium(v) and arsenic(rr1) at concentrations up to 2000 ng cm-3 did not interfere with the determination of copper(I1). Chromium(v1) showed a serious positive inter- ference, the maximum tolerable concentration being 2 ng cm-3. However, this interference can be eliminated by previously reducing chromium(v1) to chromium(rl1) with hydrogen per- oxide: chromium(rn) at concentrations up to 200 ng cm-3 did not interfere with the determination of copper(i1). Application to Tap Water and Biological Material The proposed method was applied to the determination of copper in tap water and in NIES CRM No.5 Human Hair without preconcentration and separation of copper(I1). For tap water, the sample was collected after discharging tap water for about 30 min.Then, 5 cm3 of tap water were directly used for the recommended procedure. A pressare decomposition proce- dure in a double vessel digestion bomb20 was carried out for the wet digestion of Human Hair. The digested solution was diluted 100 times with water prior to measurement. Known amounts of copper(r1) were added to these samples to examine the recovery of copper(1r). The results are shown in Tables 2 and 3. The Table 1 Tolerance limits for foreign ions in the determination of 0.2 ng cm-3 of copper(I1) Tolerance limit/ ng c1n-3 2000 Ion or salt added Nal, K', Mgl', Ca", Sr", Ba", Vx, Fell', As"', NH4+, F-, Br -, C1-, SO4*-, PO4?-, C104-, C1O3- 200 Al"', Cr"', Co", Ni", Zn", Mo"', Pd", Cd", Shill, Pb" 20 SeCV, Mn" 2 Cr"' Table 2 Detergination of copper in tap waterk Sample taken/ Cu" added/ Cu found/ Cu in sample cm3 pg cm-3 pg cm--i (ppt) 5 0 18.2 182 5 10 27.7 I77 5 20 37.7 I77 5 30 46.4 164 Average: 175 If: 7 Parts-per-trillion. Collected at the University of Tsukuba.1518 Analyst, October 1996, Vol.121 Table 3 Determination of copper in NIES No. 5 Human Hair* Sample taken+/ Cu" added/ Cu found/ Cu in sample/ 5 0 19.9 15.4 5 10 32.5 17.4 5 20 40.7 16.0 5 30 52.0 17.0 cmz pg cm-3 pg cm-3 Pg g-' Average: 16.5 f 0.8 * Certified value of copper: 16.3 f 1.2 pg g-1. + The digested solution was diluted 100 titnes before measurement. recovery of added copper(I1) was found to be quantitative and the reproducibility was satisfactory. Furthermore, the analytical result for Human Hair was in good agreement with the certified value. The proposed method is suitable for the determination of picogram levels of copper in small sample volumes.We gratefully acknowledge the financial support of this study by a Grant-in-Aid for Scientific Research [No. 0829 and 0264 (N.T.), No. 06453066, 06303005 and 07304048 (T.K.)] from the Ministry of Education, Science and Culture. References 1 Nakano, S . , Sakai, M., Tanaka, M., and Kawashima, T., Chem. Lett., 1979, 473. 2 Nakano, S . , Tanaka, M., Fushihara, M., and Kawashima, T., Mikrochim. Acta, 1983, I, 457. 3 Holz, F., Fresenius' Z. Anal. Cheni., 1984, 319, 29. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Nakano, S., Ihara, H., Tanaka, M., and Kawashima, T., Mikrochim. Acra, 1985, I, 455. Lopez, F. S., Nevado, J. J. B., and Mansilla, A. E., Talunta, 1984,31, 32.5. Ceba, M. R., Sanchez, J. C. J., and Diaz, T. G., Microchem. J., 1985, 31, 340. Casassas, E., Izquierdo-Ridorsaand, A., and Puignou, L., Talunta, 1988, 35, 199. Marquez, M., Silva, M., and Bendito, D. P., Anal Lett., 1990, 23, 13.57. Hernandez, F. H., Beneto, F. J. L., and Escriche, J. M., Chem. Anal. (Warsaw), 1990, 35, 469. Velasco, A., Silva, M., and Valcarcel, M., Anal. Chim. Actu, 1990, 229, 107. Kamidate, T., Itoh, K., and Watanabe, H., Anal. Sci., 1990, 6, 769. Kawakubo, S., Katsumata, T., Iwatsuki, M., Fukasawa, T., and Fukasawa, T., Analyst, 1988, 113, 1827. Nakano, S., Hayashi, M., and Kawashima, T., Anal. Sci., 1993, 9, 69.5. Satoh, K., Iwamura, N., Teshima, N., Nakano, S., and Kawashima, T., J . Flow) Injection Anal., 1993, 10, 245. Tamaoku, K., Murao, Y., Akiura, K., and Ohkura, Y., Anal. Chim. Acta, 1982, 136, 121. Tamaoku, K., Ueno, K., Akiura, K., and Ohkura, Y., Chem. Pharm. Bull., 1982, 30, 2492. Madesen, B. C., and Kromis, M. S., Anal. Cheni., 1984, 56, 2849. Johnson, K. S., Sakamoto-Arnold, C. M., Willason, S. W., and Beehler, C. L., Anal. Chim. Acta, 1987, 201, 83. Ichiyama, A., Nakai, E., Funai, T., Oda, T., and Katafuchi, R., J . Biochem., 1985, 98, 1375. Okamoto, K., and Fuwa, K., Anal. Chem., 1984, 56, 1758. Paper 61033531 Received May 14, 1996 Accepted July 5 , I996

ISSN:0003-2654    

DOI:10.1039/AN9962101515

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, October 1996, Vol. 121 (1519) 1519 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 Measurements Reliability Versus Uncertainty for Analytical ~ The opening statement of the summary of a reccnt paper in Anal. Coninizrn.,l namely ‘Uncertainty of measurement is a parameter of quality that should accompany any measurement result’, brings into stark perspective the point raised by M. Thompson’,3 that ‘uncertainty’ casts doubt on the credibility of analytical measurements. Apart from the fact that ‘uncertainty’ does not blend with the concept of ‘quality’, insistence on making uncertainties explicit as a universal practice, as emphasised by Thompson,’ could damage the analytical profession. As queried by him ‘What would politicians, journalists arid the general public make of such a technical iwue?’ It can also lead to difficulties in law enforcement.‘Reliability’ is better related to q ~ a l i t y , ~ and i s consistent with Thompson‘s stance’ for an alternative, such as ‘reliance interval’ in place of ‘uncertainty’. Indeed, a feature in the Spring 1996 issue of the VAM Bz~llrtiri~ opens with the statement that ‘Uws of analytical data are becoming increasingly aware of the issue of reliability and confidence which can be placed in the results of analytical measurements’. Thi\ points to the quality concept being directly related to ‘reliability’, and the response by analysts, i n my view, could be based on a recognition and understanding of the factors and parameters that characterize the dispersion of values.Features on confidence issues of measurement. might then, more fittingly, be under titles like ‘New Guidance on Measurement Reliability’, particularly for analytic a1 me as uremen t s . ‘Reliability’ is easy to understand, to explain and to defend. Indeed, except for the arguably unsatisfactory concept in analytical measurements of ’Uncertainty in Measurement’ introduced by metrologists into 1SO parlance6-x, the paper1 by Pueyo, et al., could readily go under the title of ‘Expression of Reliability of an Acid-Base Titration’, i . ~ . , with ‘Uncertainty’ having been replaced by ‘Reliability’. The many instances of the use of ’uncertainty‘ in the text would be replaceable by ‘reliability’, or a variant thereof, or in some instances by the sufficiently and widely understood statiytical concept of ‘error’.Of course, the matter of reliability of analytical data is important for trade, for those who set or test for compliance with regulatory limits, and to all with responsibility for quality control, quality assurance and method development. For this purpose, the imporlant point is that due recognition is given to those parameters within the accepted d e f i n i t i ~ n ~ . ~ of ‘Un- certainly of Measurement’, namely ‘the parameter associated with the result of a measurement, that characterizes the dispersion of the values that could reasonably be attributed to the measurand’. But, the ‘uncertainty, paradigm signifies doubt and its use, apart from the criticisms made above, can undermine efforts directed at improving reliability.Although the Analytical Methods Committee of The Royal Society of Chemistry has recently considered the implications of ‘Uncertainty of Measurement’ in analytical science,9 the exercise has been directed at the method by which uncertainty is expressed amongst analytical chemists. It does not address what has been expressed? as the ‘hidden dangers’ of the use of ‘ uncertainty’. The expression of disquiet over the use of ‘uncertainty’ indicates that it is timely that the matter of ‘Expression of Uncertainty in (Analytical) Measurement’ be fully and properly addressed for revision, or, as Thompson has said’ *What is to stop analytical chemists adopting their own special synonym?’ It is also a matter that EURACHEM, IS0 and other bodies might reopen for full and open debate. References 1 Pueyo, M.. Obiols, M., and Vilalta, E., A n d Commun , 1996, 33, 205. 2 Thompson, M., Analyst, 1995, 120, I17N. 3 Thompson, M . Analy\ir E i i i o p , 1996, (January), 9. 4 Thomas, J. I>. R., AnulyJir Europu, 1996, (March/April), 19. 5 VAM Bulletin, 1996 (Spring), 14, 27. 6 International Standard\ Organisation, Irztri nutionol Vot ahulur) o f Bo\u und Gerieial T e r m in Mrti ology, ISO, Geneva, 1984 7 International Standard\ Organibation, lnlerrzalional Voc uhidut v of Bmit mid Gcwi al Ternis in Metrology, 2nd. edn., ISO, Geneva, 1993. International Standards Organisation, Giride to the E x p ewion .f Untei tairq i n Meusurenrei?t, ISO, Geneva, 1993. Analytical Methods Committee, Analyst, 1996, 120, 2303. 8 9 Paper 6106286E

ISSN:0003-2654    

DOI:10.1039/AN9962101519

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analyst, October I996, Vol. 121 1521 CUMULATIVE AUTHOR INDEX JANUARY-OCTOBER 1996 Abdcl- Aziz, Mohamed Shafei, Abraham. Michael H., 5 1 1 Abramovik, Bil-jana F., 401. 425 Abi-amovik. Borislav K.. 401, 425 Abroskin, Andrei G., 4 19 Acedo Valenzuela, M. I., 547 Adam. S., 527 Adams, Freddy. 1061 Adeloju. S. B., 699 Aheriic, G. Wynne, 329 Ahonen. Ilpo, 1253 Akhtai-, M. Humayoun, 803 Al-Othman, Kashed, 601 Alazard, S . , 527 A1dridge;Paul K., 1003 Aldstadt, Joseph H., 1387 Alegret. S.. 959 Alcixo. Lui7 M., 559 Alexandrov. Yu. I., I137 Allcgri. Davide, 1359 Almirall. J., 959 Alvinerie, M.. 1469 Analytical Methods Committee, Andrade, Francisco J.. 6 13 Andrews, M. K., 1355 Angeletti, R.. 229 AntotiijeviC, M. M., 255 Appleton, Mark, 743 Aratake. Sachiko. 325 Araujo, Pedro W., 58 I Arias, J.J., 1327 Arias, Juan Jose, 169 Armstrong Hewitt. S.. 1457 Art.jushenko, Slava. 789 Bacci, Mauro, 553 Baggiani. Claudio. 939 Balasubranianian, N., 647 Ballesteros, Evaristo, 1397 Bannon, Thomas, 7 15 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 Bendicho, C., 1479 Benedetti, A. V.. 541 Benmakroha, Yazid. 52 1 Bentsen, Kagne K., 1191 Biancotto, G.. 229 Bilitewski, Ursula, 119. 863, 877 Birch. David J. S.. 905 Birch, M. Eileen. 1183 Bjiirklund. Erland, 19 Blais, Jean-Simon, 483, 1419 Blanchflower. W. John. 1457 Blanco, Marcelo. 395 Bogan, Declan R., 243 Bond. Alan M., 357 Borah, Lakhimi, 987 Borowiak, Annette. 1247 Boswell. Stephen M., 505 Bouhsain, Zouhair. 635 Boukortt.Sheriffa, 663 Boutelle, Martyn G., 761 Bowker. Michael J.. 9 1 K Boyd-Boland, Anna A., 929 Boyd, Damien, IR Branica, Marko, 1127 Brereton. Richard G.. 441. 575, Brinkman, Udo A. Th., 61. 1069, Bristow. Anthony W. T., 1425 Brown. R. H., 1 17 I Brown, Richard C., 124 1 I079 573,889 58 I . 585, 65 I , 993. 1443 1327 Bru. E. K., 297 Buchet, Jean-Pierre, 663 Burgot, Jean-Louis. 43 Bye, Ragnar, 201 Cadogan, Aodhmar, 1463 Cai. Xiaohua, 965 Callej6n Mochhn, M., 681 Cammann. K., 527 Campillo, Natalia. 1043 Cannavan, Andrew. 1457 Cao, Zhong, 259 Capela. D., 1469 Carbonnellc, Philippe, 663 Cardoso, A. A., 541 Cardwell, Terence J., 357 Carmona. Pedro, I05 Cary. Robert A., I183 Casajus, Rocio, 81 3 Casella, Innoceiizo G., 249 Cassidy, Richard M., 839 CaviC-Vlasak, Biljana A., 53R Cazeiiiier.Geert. 1 11 1 Ccla, R., 297 Cepas. Juana, 49 Ceramelli, Giuseppe, 219 Cerdli. A,, 13 Cerd8. V., 13 Chan, Wing Hong, 53 1 Chatergoon, Lutchminarine. 373 Chen. Guo Nan, 37 Chen. Wen-Can, 1495 Chiou, Chyow-San. 1107 Chou, Shu-Fen, 7 1 Christian, Gary D.. 601 Christie, Ian, 521 Cirovic. Dragan A.. 575, 581 Coello, Jordi, 395 Cole. S. Keith, 495 Collier. Wendy. 877 Comber. Sean D. W., 1485 Cook, Michael J.. I501 Copeland, D. D., 173 Corbella Tena, R., 459 Corti, Piero, 219 Cosano, J., 83 Craston, Derek H.. 177 Creaser, Colin S., 1425 Crosby. Neil T., 691 Croteau. Louise G., 803 Crump, Paul W., 87 I Crumrine, David S., 567 Cruz Ortiz. M., 1009 CuculiC, Vlado, 1127 Cullen, Michael, 75 Cullen, William K., 223 Daae, Hanne Line, I191 Daenens, Paul, 857 Daghbouche, Yasmina, 103 I Dalene, Marianne, 1095, I 101 dc Jong, Dirk, 61 de Jong, Gerhardus J., 61 de la Guardia, M., 1327 de la Guardia, Miguel, 635. 923, dc Lacy Costello, Benjamin P.J., cie Oliveira Neto, Graciliano, 559 De Saeger, Emile. 1247 Dean, John R., 465, 85R Demidova, M. G., 489 Demir, Cevdet, 651, 993, 1443 Deng. Jiaqi, 971 Deng, Qing. 1123 Dcng. Zhiping, 67 I , 134 I Desai, Mohamed, 521 Desimoni, Elio, 249 Destradis, Angelo. 249 Devi. Surekha, 807 Dey. Nibaran C.. 987 1031 793 Dilleen. John W., 755 Dobrowolski, R., 897 Dodd, Matthew, 223 Dolmanova, Inga F.. 43 1 Dong, Shaojun, 1123 Dreassi, Elena, 2 19 Dumasia, Minoo C., 651 Dumschat, C., 527 Dunemann, Lothar, 845 Dunhill, Roger H., 1089 Dunn, Warwick B., 1435 Economou, Anastasios, 97, 1015 Eduard, Wijnand, 1 19 1, 1 197 Eigendorf, Guenter K., 223 Eikenberg, Oliver, 119 Einhorn, Jacques, 1425, 1429 El-shahat, Mohamed F., 89 El-Shorbagi.Abdel-Nasser, 183 Elbergali, Abdallah K . , 585 Eller, Peter M., I163 Ellwood, Jo A., 575 Emara, Samy, 183 Emteborg, Hgkan, 19 Endo, Masatoshi, 39 1 Eng, Jimmy, 65R Escobar, Rosario, 105 Essers, Martien, 1 1 11 Esteves da Silva, Joaquim C. G.. Evans, Phillip, 793 Fabrigs, Jean-Fraiiqois, 1257 Facer, M., 173 Fallon, Michael G., 127 Fang, Kai-Tai, 1025 Fawaz Katmeh, M.. 329 Fearn, Tom. 275 Fell, Gordon S., 189 Fertiandes, Julio Cesar B., 559 Fernandez-Suarez, A., 1469 Ferreira, I. M. P. L. V. O., 1393 Ferreira, Valdir S., 263 Fielden, Peter R., 97, 1015 Fillenz, Marianne, 761 Fiore, Amy A., 1265 Fischbach, Thomas J..1163 Fitzgerald, Catherine A.. 7 15 Fleet, Ian A., 55 Forsberg, Bertil, 126 1 Forster, Robert J., 733 Forteza, R.. 13 Francis, John M., 177 Frank, Gerhard, 130 1 Frech, Wolfgang, 19. 1055 Fugivara, C. S., 541 Fukasawa, Tsutomu, 89 Fung, Yingsing. 369 GaA. Ferenc F., 401,425 Gala, Belen, 1133 Galeano Diaz, T., 547 Gallego, Mercedes, I397 Galtier, P., 1469 Gamble, Donald S., 289 Gao, Xiao Xia, 687 Garcia-Fraga, J. M., 1327 Garcia, M., 959 Garrido Frenich, A,, 1367 Garrigues, Salvador, 635. 923, Gebefugi, Istvan, 1301 Geckeis, Horst, 1413 Genrich, Meike, 877 Georgieff, Michael, 901 Ghosh, Anil G., 987 Giannousios, A., 4 13 Giersch, Thomas, 863 Giraudi, Gianfranco, 939 Glennon, Jeremy D., 127 Godinho, Oswaldo E. S.. 559 Goldstein, Steven L., 901 I373 1031 G6mez-Hens, Agustina, 1 I33 Gong, Zhilong, 1 119 Gooijer.Cees, 1069 Goosens, Elise C., 61 Gordon, Derek B.. 55 Gbrecki, Tadeusz, I38 1 Gorner. Peter, 1257 Goto, Nobutake, 1085 Green, John D., 1435 Greer, James C., 715 Grol, Michael. 119 Groves. John A., 441 Grudpan, Kate. I4 13 Guiberteau Cabanillas, A,. 547 Guiraum Perez, A., 681 Gulmini. Monica, 140 I Gurdeii, Stephen. P., 441 Gustavsson, C. A., 1285 Haasnoot, Willem, I 1 1 1 Hadjiivanov, K., 607 Haferkamp, Heinz, 129 1 Hatkenschcid, Theo L., 1249 Hagenbjork-Guslaisson, Annika, Haloard. Kristin, 1191 HalEwell, David J.. 1089 Hammerich, Ole, 345 Hangartner, Markus, 1269 Hansen, Elo H., 31 Hansen, Erik Beck, 129 I Harper. Martin, 1265 Harris, P.. 1355 Harris. Roy, 9 13 Harrison, Iain, 189 Hart, Barry T., 1089 Hartnett, M., 749 Hasan, B.A., 1327 Hauser. Peter C.. 339 Hayashi, Yuzuru, 591 Hayashibe, Yutaka, 7 Hays, Lara, 65R Hceremans, Carola E. M.. 1273 klemingway, Michael A., 1241 Hcndrix, Janrcs Id., 799 Herninde7-C6rdoba, Manucl: Hemhndez, Oscar, I69 Hcstvik, Gete, 1261 Hietel, Bernhard, 1301 Hindmarch, Peter. 993, 1443 Hirata, Takafumi, 1407 Hoekstra-Oussoren, Sacha J. F., Honing, Maarten, 1327 Hoogenboom, Laurentius A. P.. Home, Elizabeth, 1463, 1469 Horng, Ching-Jyi, 15 11 Hosse, Monika, 1397 Hu, Yan: 883 Hulanicki, Adam, 133 Hyland, Mark, 705 Ibrahim, Naaim M. A., 239 Idriss, Kamal A., 1079 inagawa, Jun, 623 Ifiiguez, Montsxrat, 1009 Ioannou, Pinelopi C., 909 Irwin, G. W., 749 Ishida, Yasuyuki, 853 Ishihard, Masahito, 391 Isomura, Shiniclii, 853 Itabashi, Hideyuki, 15 15 Iturriaga, Hortensia, 395 Ivanova, Elena K., 4 19 Iwatsuki, Masaaki, 89 Jackson, Laurence S., 67 Jager, Maria E., 1327 Jaselskis, Bruno, 567 Jiang, Chongqiu.3 17 1261 1043 1327 14631522 Anulyst, October. 1996, Vol. 121 _.________~ Jiang, Wei, I3 I7 Jimenez. A. I., 1327 JimCnez, Ana Isabel, 169 Jimenez, F., 1327 JimCnez, Francisco, 169 Jimenez-Prieto, Rafael, 563 Jimenez Si’inchez, J. C., 681 Johnson, Brian J., 1507 Johnson, Mark, 1075 Jonsson, €3. A. G., 1279, 1285 Jurkiewicz, M.. 959 Kalish, N . K.. 489 Karayannis, Miltiades I., 435 Kawashima, Takuji, 15 15 Kennedy, D. Glenn, 1457 Kennedy, Eugene R., I163 Kenny, Lee C.. 1233 Kettling, Ulrich. 863 Kettrup, Antonius, 130 1 Khalaf, K. D., 1327 Kinibrough, David Eugene, 309 Kimoto, Takashi, 853 Kindness, Andrew, 205 Kirchner, Manfred, I269 Knoll, M., 527 Kolotyrkina.Irina Ya., 1037 Konstantianos, Dimitrios G., 909 Korda. T. M.. 489 Kozik, Andrze.1, 333 Kratochvil, Byron, 163 Krier, Gabriel, 1429 Kuznetsova, Vera V., 4 I9 Kvasnik, Frank, 1 1 15 Kwong, Daniel W. J., 531 Lan, Zhang-Hua, 2 I I Lancashire, Susan, 75 Lancia, Antonio, 789 Laurie. David. 9.5 1 Lavilla, I., 1479 Lawrence, Chris M., 755 Le, Quyen T. H.. 1051 Lee, Albert W. M., 531 Legouin, Beatrice, 43 Lei, Chenghong, 97 1 LeskovSek, Hermina, 145 I Levin, Jan-Olof, 1177, 1273 Lewenstam, Andrzej. 133 Li, Hao, 223 Liang, Y i-zeng, 1025 Lightbody, G., 749 Lima, J. L. F. C., 1393 Lin. Hui-Gai, 259 Lindahl, Roger, 1177, 1273 Lindh, C. H., 1285 Lindskog, Anne, 1295 Link, Andrew J., 65R Lipkovska, N.A., 501 Lison, Dominique, 663 Littlejohn, David, 189 Liu, Dong, 1495 Lonardi, S., 219 Lopes, Teresa I. M. S., 1047 Lopez Carreto, Maria, 33R Lopez-Cueto, Guilleimo, 407 Lopez-Erroz, Carmen, I043 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 Machado, AdClio A, S. C., 1373 McKelvie, Ian D., 1089 MacLachlan. John, 1 IR McLaughlin, James A., 705 McNaughtan, Arthur, 1 1 R 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, Marlensson. Maud, 1 177 Martin. Alice F., 1387 Martin, Patricia. 495 Martinez-Fhbregas. E., 959 Martinez Galera, M., 1367 Martinez-Lozano, Carmen, 477, Martinez Vidal, J. L., 1367 Mason, Andrew J., 95 1 Maspoch. Santiago, 395, 407 Masselon, Christophe, 1429 Masiijima. Tsutomu, 183 Mathiasson, Lennart, 19 Matsuda. Rieko, 591 Matsui, Masakam, 105 I Meaney, Mary. 789 Melbourne. Paul. 1075 Melius. Cristo B., 263 Mieczkowski, Jbzef, 133 Mierzwa, J., 897 Mihajlovic, R., 255 MilaZiE, Radmila, 627 Mills. Andrew, 535 MilosavljeviC, Emil B.. 799 Mindrup, Raymond, 138 1 MitroviC, Bojan, 627 Mizgunova, Ulyana M., 43 1 Mo, Jin-Yuan, 1019 Mo, Songying, 369 Moane.Siobhan, 779 Mocak, Jan, 357 Mohamed, Ashraf A., 89 Mohr, Gerhard J., 1489 Molina, Marina. 105 Monaf, Lela. 535 Monaghan. John J., 55 Montelongo, F. Garcili, 459 Montenegro, M. C. B. S. M., 1393 Moollan, Roland W., 233 Moore, Andrew, 67 Morales-Rubio. A., 1327 Mori. Giovanni, 1359 Mosello. R., 83 Motomizu, Shqji, 1085 Mottola, Horacio A,, 211, 381 Mounsey, Andrew, 955 Mowrer. Jacques. 1249, 1295 Mulcahy, David, 127 Muller. Beat. 339 Muller, Jean-Franqois, 1429 Munro. C. H., 835 Murphy, William S., 127 Nakamura, Masatoshi, 469 Nakamura, Motoshi. 469 Nakanishi, Masami, 853 Nakano, Shigenori, I5 15 NClieu, Sylvie, 1425, 1429 Newton, R., 173 Nie, Lihua. 883 Nielsen, Steffen, 3 I Nolte, Joachim, 845 Norefia-Franco, Luis E., I 1 15 Norris, Timothy, 1003 Noto, Hilde, 1 19 1 Nygren, Olle, 129 I Obendorf, Dagmar, 35 1 Obradovit, Danilo M., 401 Odman, Fredrik, 19 Ohno, Satomi, 15 15 Ohtani.Hajime, 853 O’Keeffe, Michael, 779, 1463, 817. 1419 813 1469, IR O’Kennedy, Richard, 243, 767. O’Lear. Christina, I265 Oliveira, CCsar J. S., 1373 Olmi, Filippo, 553 Olsen. Erik, 1155 Oms. M. T., 13 O’Neill, Robert D.. 761, 773 Oniciu, Liviu. 363 Oosten, Koos van, 1273 Orlando, Andrea, 553 Oshima, Mitsuko, 1085 Osipova, Nataliya V., 4 19 Ostaszewska, Joanna. 133 Owen, Susan P., 465 Packham, Andrew J., 97, 1015 Papadopoulos, C., 413 Paradowski, Dariusz, 133 Pardue, Harry L., 385 Park: Chang J., 131 1 Pmilla. P., 1367 Parsons, Patrick J., 195 Partridge, A. C., 1355 Patel.Sunil U., 913 Patterson, Kristine Y., 983 Paulls, David A., 831 Pawliszyn, Janusz B., 929, 138 1 Pedrero, Maria, 345 PCrcL-Bendito. Dolores, 49, 563, PCreL-Buslamantc, J. A., 297 Ptrez-Cid: B ._ 1479 PCrez Olmos, R., 1393 Pkrez-Ponce. Amparo, 923 PCrez-Ruiz, Tonii‘is, 477, 8 13 Pergantis, Spiros A., 223 Perruccio. Piero Luigi, 219 Petty, Michael C., 1501 Pfiiffli, P.: 1279, 1285 Piggott, Nighel H., 951 Pihlar, Boris, 627 Pingarron, Jose, 345 Piperaki, Efrosini A., 11 1 Piro, R. D. M., 229 Pitre, K. S., 79 Poe, Russell B., 591 Poole, Colin F., 5 1 1 Potter, Annika, I295 Power, J. F.. 451 Pramauro. Edmondo, 140 1 Prevot, Alessandra Bianco, 1401 Prodromidis, Mamas I.. 435 Proinova, I., 607 Proskumin, Mikhail A., 419 Pui. David Y. H., 1215 Piister, Thomas, 129 1 Pyrzynska, Krystyna, 77R Qi, Zhong-Cheng, 13 I7 Qu, Yi Bin, 139 Quevauviller.Ph., 83 Quinii, John G., 767 Rader. W. Scott, 799 Rae, Bruce, 233 Raghunath, A. V., 825 Rahmani. Ali, 585 Ramachandran, Gurumurthy, 122.5 Ramanaiah, G. V., 825 Rangel, Antonio 0. S . S . , 1047 Ratcliffe. Norman M., 793 Razee, Saeid, 183 Redon, Miguel, 395 Regan, Fiona, 789 Reimer, Kenneth J., 223 Reinartz. Heiko W., 767 Rigby, Geraldine P., 871 Riipinen, Hannu, 1253 Rios, A., 1393 Rios, Angel. 1 Rodriguez Delgado, M. A,, 459 Rodriguez-Medina, Jose F., 407 Rohm. Ingrid, 877 Roos, Aappo, 1253 Rowell. Frederick J., 95 1 , 955 Rowell, Vibeke, 955 Rozendom, Eduard J. E., 1069 Rubio, Solcdad. 33R Russell, David A., 1501 29R 1133, 33R Ruzicka. Jaromir, 601, 945 Sadler, Peter J., 913 Sakslund. Henning, 345 Salden.Martin. 11 1 I Saleh, Gamal A., 641 Salinas, F.. 547 San Martin Femindez-Marcote, Slinchez, M;’. J., 459 Sandstriim, Thomas, I26 1 Santaniaria, Fernando, 1009 Santos, Jose H., 357 Sanz, Antonio, 477 Sarabia, Luis A., 1009 Sartini, Raquel P., 1047 Sasaki, Takayuki, 105 I Saw, Hidetoshi. 325 Satyanarayana, K., 825 Sayama, Yasumasa, 7 Schafer, E. A,, 243 Schieltz. David, 65R Schmid. Rolf D., 863 Schnelle. Jurgen. I301 Schoeps, Karl-Olof, 1203 Schoppcnthau, Jorg, X45 Scohbic, Emma. 575 Scudder. Kurt, 945 Sedaira, Hassan, 1079 Seebaum, Dirk, 129 1 Seeber, Renato, I359 Seiberl, Donna S., 51 1 Sekino, Talsuki, 853 Sepii., Ester, I45 1 Seviour, John, 95 1 Shah, Rupal, 807 Shakoor. Oniar, 1473 Shanthi. K., 647 Shen. Guo-Li, 1495 Shi, Kenbing, 13 I 1 Shi, Yilin.1507 Shih. Jeng-Shong, 1107 Shijo, Yoshio, 325 Shiraishi, Haruki, 965 Shpigun, Lilly K., 1037 Shukla, Jyotsna, 79 Shulman, R. S., 489 Shulman, Stanley A., 1163 Si, Zhi-Kun. 1323 Sihvonen. Marja-Liisa, 1335 Sillanpitii. Mika. 1335 Silva, Manucl, 49, 563 Simpson, Tim R. E., 1501 Siskos, Panayotis A., 303 Skarping, Gunnar, 1095, 1 10 I Slater. Jonathan M., 743, 755 Slavin. Walter. 195 Slobodnik, Jaroslav, 1327 Sloth, Jens J., 31 Smith, Clayton, 373 Smith, Dennis C., 53R Smith, Robert F., 67 Smith, Roy, 321 Smith, W. E., 835 Smyth, Malcolm R.. 779, lR, 29R Smythc-Wright, Denise, 505 Snell, James P., 1055 Sokalski, Tomasz, 133 Solk, s., 959 SolujiC, Ljil-jana, 799 Somsen, Covert W., 1069 Song, Ruiguang, 1 163 Sooksamiti, Ponlayuth, 14 13 Sorvari, Jaana, 1335 Spaiine, Marten, 1095, 1101 Spear, Terry M., 1207 Stathakis, Costas, 839 Stegman, Karel H..61 S tegm aiin, Werner, 90 1 Stein, Kathrin, 131 I Stevenson. Derek, 329 Stone, David C., 671, 1341 Stouten, Piet, 1 I 1 1 Strachan, David, 951, 955 Stradiotto, Nelson R., 263 Streppel, Lucia, 1 1 1 1 Stuart, Iain A., 11K Stubauer, Gottfried, 35 1 M., 6x1Analyst, October 1996, Vol. 121 1523 Subramaniam, K., 825 Suffet, I. H. ‘Mel’, 309 Sukhan, V. V., 501 Suliman, Fakhr Eldin O., 617 Sultan, Salah M., 617 Sumodjo, P. T. A., 541 Susanto, Joko P., 1085 Sutra, J. F., 1469 Svanberg, Per-Arne, 1295 Sweedler, Jonathan V., 45R Symington, Charles, 1009 Szklar, Roman S., 321 Tam, Wing Leong, 53 1 Tan, Yanxi, 483, 1419 Tang, Bo, 3 I7 Tang, Shida, 195 Taylor, Robert B., 1473 Tegtmeier, M., 243 TepavEeviC, Sanja D., 425 Teshima, Norio, 15 15 Thastrup, Ole, 945 Thomaidis, Nikolaos S., I 1 1 Thomas, J .D. R., 1519 Thomassen, Yngvar, 1055 Thompson, Michael, 275, 285, 67 1,977, 1341,53R Thornes, R. D. , 243 Thorpe, Andrew, 1241 Thorpe, Stephen C., 1501 Tian, Baomin, 965 Timperman, Aaron T., 45R Tinnerberg, HAkan, 1095, 1101 Tom& Virginia, 477, 813 Torgov, V. G., 489 Townshend, Alan, 831, 1435 Trier, Colin, 1451 Troccoli, Osvaldo E., 613 Tsuge, Shin, 853 Tsurubou, Shigekazu, 105 1 Tudino, Mabel B., 613 Tyfon, John D., 95 I , 955 Tzouwara-Karayanni, Stella M., Ubide, Carlos, 407 Uehara, Nobuo, 325 Umetani, Shigeo, 105 1 Vadgama, Pankaj, 435, 521, 871 Vaggelli, Gloria, 553 Valcircel, Miguel, 1, 83, 1397 van Baar, Ben L. M., 1327 Van Mol, Willy, 1061 van Wichen, Piet, 11 11 Vassileva, E., 607 Veillon, Claude, 983 Velthorst, Nel H., 1069 Verbeek, Alistair.233 Viles, John H., 913 Villegas, Nuria, 395 Viiias, Pilar, 1043 Vincent, James H., 1207, 1225 435 Viscardi, Guido, 140 1 Vos, Johannes G., 789 Vukanovik, B., 255 Wahlberg, Sonny, 1261 Wake, Derrick, 1241 Walker, P. J., 173 Wallace, G. G., 699 Walsh, James E., 789 Walsh, Peter T., 575 Wang, Bin-Feng, 259 Wang, Chen, 317 Wang, Jin, 289, 817 Wang, Joseph, 345, 965 Wang, Ke-Min, 259, 531 Wang, Nai-Xing, 13 17 Wang, Shi-Hua, 259 Watanabe, Kazuo, 623 Watanabe, Tsuyako, 1515 Watts, Chris D., 1485 Welinder, H., 1279, 1285 Werner, Herbert, 1269 Werner, Mark A.. 1207, 1225 WessCn, Bengt, 1203 Wheals, Brian B., 239 White, P. C., 835 Whiting, Robin, 373 Wickstrom, Torild, 201 Wilmot, John C., 799 Witschger, Olivier, 1257 Wittmann, Christine, 863 Wolf, Kathrin, 130 1 Wolfbeis, Otto S., 1489 Wood, Roger, 977 Woolfson, A.David, 71 1 Wu, Weh S., 321 Xin, Wen Kuan, 687 Xu, Xue Qin, 37 Xu, Yuanjin, 883 Yamada, Shinkichi, 469 Yan, Xiu-Ping, 1061 Yao, Shouzhuo, 883 Yates 111, John R., 65R Young, Barbara, 1485 Yu, Ru-Qin, 259, 1495 Zagatto, Elias A. G., 1047 Ztinker, Kurt, 767 Zanoni, Maria Valnice B., 263 Zaporozhets, 0. A., 501 Zelano, Vincenzo, 1401 Zhang, Fan, 37 Zhang, Xiaogang, 3 17 Zhang, Zhanen, 97 1 Zhang, Zhujun, 1 119 Zhi, Zheng-lianp, 1 Zhou, Dao-Min, 705 Ziegler. Torsten, 119 Zolotova, Galina A., 43 1A brand new initiative by The RSC means that for the first time we are now able to offer you six of our highly respected journals at vastly discounted personal rates if you belong to an organisation that already subscribes.Just take a look at the substantial savings we are now offering: ChemComm E544/$103 2 E89/$139 Natural Product Reports &325/$615 &85/$132 Chemical Society Reviews E120/$225 E45/$69 Journal of Materials Chemistry E5 19/$984 E89/$139 Dalton Transactions E975/$1848 &99/$154 The Analyst &487/$923 &85/$132 Mendeleev Communications E195/$325 E55/$90 * With exception of Mendeleev Communications, this offer is available only to individuals working for organisations which already have a full non-member subscription at the same site. On these terms, can you afford not to subscribe today? To order please contact: The Royal Society of Chemistry, Turpin Distribution Services Ltd, Blackhorse Road, Letchworth, Herts SG6 lHN, UK Tel+44(0) 1462-672555 Fax +44(0) 1462-480947 RSC members ordering for their own personal use are entitled to a discount on most RSC publications, and should contact: Membership Administration Department, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF, UK Tel+44(0) 1223-420066 / Fax +44(0) 1223-423623 E-mail (Internet): rscl @rsc.org wwweb: http://chemistry.rsc.org/rsc I - ~ ~ ~ ~ ~ ~ - - - - - - - - - - I - - D - - - I - - - - 1 m 1 - - - ~ ~ ~ - - - Please send me further information on the following journals: Journal Name: Journal Name: You r N am e : Position: Organ isation: Address: Please return to the Sales & Promotion Department at the above Cambridge address

ISSN:0003-2654    

DOI:10.1039/AN9962101521

出版商:RSC    

年代:1996

数据来源: RSC