ISSN:1359-7337    

DOI:10.1039/AC99633FX050

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

ANCOFE 33( 10) 353-392 (1 996) OCTOBER 1996IllCOMMUNICATIONS 353357I36136536737137538 1385AnalyticalC o m m u n i cat i o nsFormer I y Analytical ProceedingsCONTENTSSorption Properties of Styrene-Divinylbenzene Macroreticular Porous Polymers-Salwa KPoole, Colin F. PooleReversed-phase Liquid Chromatographic Method for the Determination of pAminobenzoic Acidand Anthranilic Acid in Urine: Application to the Investigation of Exocrine PancreaticFunction-Abdul Rob, Wayne H. Bradbury, Alexander R. W. ForrestSolid-phase Microextraction Combined With Electrochemistry-Feng Guo, Tadeusz Gorecki,Donald Irish, Janusz PawliszynSelective Fluorimetric Recognition of Dihydrogen Phosphate Over Chloride Anions by a NovelRuthenium(~) Bipyridyl Receptor Complex-Paul D.Beer, Roger J. Mortimer, Fridrich Szemes,John S. WeightmanDetermination of Carbohydrates by Flow Injection With Direct ChemiluminescenceDetection-lrena B. Agater, Roger A. Jewsbury, Kath WilliamsAnalysis of Dissolved Gases by Headspace Sampling Gas Chromatography With Column andDetector Switching. Preliminary Results-Pierre-Marie Sarradin, Jean-Claude CapraisIdentification of Hydrogen Peroxide as the Autoxidation Product ofN-phenyl-2-propyl-3,5-diethyl-l,2-dihydropyridine-Declan P. Raftery, Malcolm R. Smyth,Raymond G. Leonard, Brendan J. Kneafsey, Martin C. BrennanCalculation of Intensities of Molecular Interferences in GD-MS: Application to Analysis ofAluminium Alloys-Vladimir D. KurochkinLow Absorbance Differential Spectrophotometry-Zhao Shanlin, Li Ping, Liu Guomin, WuLixiang, Zhang Qikai389 Cumulative Author Index390 Technical Abbreviations and Acronyms391 Conference Diary//)- THE ROYALSOCIETY OFlnfO~l?latiOnServices Cambridge, EnglandTypeset and printed by Black Bear Press LimitedAnalytical CommunicationsINSTRUCTIONS TO AUTHORSAnalytical Communications is the analytical communications journal of theRoyal Society of Chemistry and publishes communications on all aspects ofthe theory and practice of analytical chemistry, fundamental and applied,inorganic and organic, including chemical, physical, biochemical, clinical,pharmaceutical, biological, automatic and computer-based methods. Com-munications on new techniques and instrumentation, detectors and sensors,and new areas of application, with due attention to overcoming limitationsand to underlying principles, are all equally welcome.There is no page charge for papers published in Analytical Commu-nications.The following types of papers will be considered.Communications, which must describe novel work and be of immediatescientific significance.These articles should be restricted to 2 4 journalpages in length (up to 10 A4 pages of typescript), including figures, tablesand references. Communications are usually published within 10 weeks ofreceipt. They are intended for brief descriptions of work that has progressedto a stage at which it is likely to be valuable to workers faced with similarproblems. A fuller paper may be offered subsequently, if justified by laterwork, to The Analyst or JAAS.Communications will be examined by twoindependent referees.Highlights, these are mini-reviews or summaries of research in a welldefined, specific topic area covering approximately the last 12 months. (i)Given topics should review work no more than approximately 12 monthsold. (ii) Articles should cover only the most interesting/significantdevelopments in that specific subject area, e.g., sample preparation forcapillary HPLC or sensors for pesticides analysis. (iii) The articles shouldbe highly critical and selective in referencing published work (perhapsreferring only to four or five papers if necessary). (iv) A small amount ofspeculation (one or two paragraphs) of possible future developments mayalso be appropriate in the Conclusions section.(v) Articles should be brief2-3 journals pages (up to 7 double-spaced, typed A4 pages), contain nomore than 1 or 2 tables and preferably no figures.Copyright. The whole of the literary matter (including tables, figures,diagrams and photographs) in Analytical Communications is Royal Societyof Chemistry copyright and may not be reproduced without permission fromthe Society or such other owner of the copyright as may be indicated.Papers that are accepted must not be published elsewhere except bypermission. Submission of a manuscript will be regarded as an undertakingthat the same material is not being considered for publication by anotherjournal in any language. All authors submitting work for publication arerequired to sign an exclusive copyright licence.All submissions should beaccompanied by a completed form, without which publication cannotproceed.Manuscripts. Papers should be typewritten in double spacing on one sideonly of the paper. Copies of any related, relevant, unpublished material andraw data should be made available on request. Each table and illustrationshould be on a separate sheet at the end of the text; three copies of text andillustrations should be sent to the Managing Editor, Analytical Cornrnu-nications, The Royal Society of Chemistry, Thomas Graham House,Science Park, Milton Road, Cambridge CB4 4WF, and a further copyretained by the author.Administration and Publication Procedure. Receipt of a contributionfor consideration will be acknowledged immediately by the EditorialOffice.The acknowledgement will indicate the paper reference numberassigned to the contribution. Authors are particularly asked to quote thisnumber on all subsequent correspondence.All communications are sent to two referees, whose names are notdisclosed to the authors. On the basis of the referees’ reports, the ManagingEditor decides whether the paper is suitable for publication, eitherunchanged or after appropriate revision. This decision and relevantcomments of the referee are sent to the author.When rejection of a paper is recommended, the Editor informs the author,Authors will receive formal notification when papers are accepted forand returns the top copy of the manuscript.publication.Proofs. The address to which proofs are to be sent should accompany thepaper.Proofs should be carefully checked and returned immediately (byfirst class mail, air mail, express mail or fax). Particular attention should bepaid to numerical data both in the tables and text.Offprints. Fifty offprints of each paper are supplied free of charge.Additional reprints can be purchased.ANALYTICAL JOURNALS 1996Published by The Royal Society of ChemistryThe Analyst12 issues a year plus indexEEA f487.00 USA $923.00Rest of World f499.00ISSN 0003-2654Analytical Communications12 issues a year plus indexEEA f 191 .OORest of World f 195.00ISSN 1359-7337USA $361 .OOjournal of Analytical AtomicSpectrometry (JAAS)12 issues a year plus indexEEA f599.00 USA$1136.00Rest of World f 61 4.00ISSN 0267-9477Analytical Abstracts12 issues a yearEEA f 570.00 USA $1067.00Rest of World f577.00ISSN 0003-2689SPECIAL PACKAGES (Non-RSC members only)The Analyst, Analytical Abstracts and AnalyticalCommunicationsJournal Ref.No. 0000-01 24EEAf1123.00 USA $2129.00RestofWorld f1151.00The Analyst and Analytical CommunicationsJournal Ref. No. 0000-01 40EEAf610.00 USAfll56.00Rest of World f625.00The Analyst and Analytical AbstractsJournal Ref. No. 0000-01 32EEA f951 .OO USA $1804.00Rest of World f975.00ORDERINGNon-RSC members should send their orders to:The Royal Society of Chemistry, Turpin Distribution Services Limited, Blackhorse Road, Letchworth, Herts SG6 1 HN, United KingdomTel: +44 (0)1462 672555 Fax: +44 (0)1462 480947RSC Members should obtain members’ prices and send their orders to:Membership Administration Department, The Royal Society of chemistry, Thomas Graham House,Science Park, Milton Road, Cambridge CB4 4WF, United KingdomTel: +44 (0)1223 420066 Fax: +44 (011 223 423623Turpin Distribution Services Limited is wholly owned by The Royal Society of ChemistryInformationServices

ISSN:1359-7337    

DOI:10.1039/AC99633BX052

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Vol33 (353-356) 353 Sorption Properties of Styrene-Divinylbenzene Macroreticular Porous Polymers Salwa K. Poole" and Colin F. Poole".b* a ZenecalSmithKline Beecham Centre for Analytical Chemistry, Imperial College of Science, Technology and Medicine, South Kensington, London, UK SW7 2AY. E-mail c.poole@ic.ac.uk Department of Chemistry, Wayne State University, Detroit, MI 48202, USA The sorption properties of some styrene-divinylbenzene macroreticular porous polymer sorbents for solid-phase extraction are characterized using the solvation parameter model.Dispersion interactions are generally responsible for retention with a small contribution from hydrogen-bond interactions that may arise from the presence of polymer impurities. Lone pair-lone pair electron repulsion is a common mechanism reducing retention. Interactions of a dipole-type are not significant for styrene-divinylbenzene sorbents while these play a significant role in retention on Tenax sorbents, highlighting the complementary character of these two porous polymers.The solvation parameter model can also be used to characterize retention of styrene-divinylbenzene sorbents in non-aqueous solvents and to predict conditions resulting in maximum sample concentration in the recovery step in solid-phase extraction. Macroreticular porous polymers are widely used in solid-phase extraction for the isolation of organic compounds from air and water, and in gas, liquid and supercritical fluid chromatography as stationary phases.' Porous polymers based on the co- polymerization of styrene and divinylbenzene are the most common type, although others based on ethylvinylbenzene, acrylonitrile, acrylic esters, vinylpyrrolidone, etc., are well known.The choice of experimental conditions allows rigid or soft beads with different properties to be prepared, in narrow particle size ranges, from non-porous to about 600 nm pores, surface areas from about 2 to 1000 m2 g-l, and with different monomer chemistries. It is this range of properties that enables them to be used for so many divergent applications in analytical chemistry.There are a number of empirical guides that can be used to match the sorption properties of individual porous polymers to particular applications and a very large applications data base is available. 1 Modified versions of the McReynolds ' phase constants system, Snyder's selectivity triangle, and the relative retention of ethylene, acetylene and carbon dioxide, have been proposed as methods to standardize sorbent selectivity of porous polymers.24 These methods are inadequate on theoret- ical and practical grounds."6 More recently, Abraham and co- workers have used the solvation parameter model to character- ize the sorption properties of and Poole and co-workers the sorption properties of Tenax sorbents.9 Poole and Poole used the solvation parameter model to explain the influence of solvent effects on the breakthrough volume of porous polymer particle-loaded membranes, used for the solid- * To whom correspondence should be addressed.phase extraction of organic compounds from water containing 1% (v/v) of various organic solvents which were added to improve sample processing properties. l o Only recently has experimental data for porous polymers of the styrene- divinylbenzene type, suitable for analysis by the solvation parameter model, become available, and is the subject of this communication.11912 Results and Discussion The form of the solvation parameter model suitable for characterizing the retention properties of porous polymer sorbents with sorption occurring from the gas phase is given by eqn.( 1 ) F (1) where SP is some free energy related property of the system, such as the specific retention volume, distribution constant, or the breakthrough volume; R2 is the solute's excess molar refraction; n: is the ability of the solute to stabilize a neighbouring dipole by virtue of its capacity for orientation and induction interactions; a? and py are parameters characteristic of the solute's effective hydrogen-bond acidity and hydrogen- bond basicity, respectively; and log ,516 is the distribution constant for the solute between a gas and n-hexadecane at 298 K.The solute's excess molar refraction is usually available by simple arithmetic calculation; the other solute descriptors are parameters derived from equilibrium measurements for com- plexation and partition processes with values available for over 2000 compound~.~33~~ The system constants in equation (1) are unambiguously defined: the r constant refers to the capacity of the sorbent for interaction with solute n- or n-electrons; the s constant to the sorbent's capacity for dipole-dipole and dipole- induced dipole interactions; the a constant characterizes the sorbent's hydrogen-bond basicity (because a basic sorbent will interact with an acidic solute); the b constant characterizes the sorbent's hydrogen-bond acidity; and the 1 constant incorpo- rates contributions from solvent cavity formation and dispersion interactions. The system constants are determined from experi- mental values for the observed parameter, SP, for a group of solutes of known properties, sufficiently varied to define all interactions in eqn.(1), and of sufficient number to establish the statistical validity of eqn. (1), by multiple linear regression analysis.Table 1 summarizes the fit of the solvation parameter model to several sets of experimental data taken from the references indicated in the table. The data of Pankow et a1.12 is the most useful for styrene-divinylbenzene sorbents and allows the system constants to be determined as a function of temperature. The system constants vary linearly with temperature over the range 70 to 90 "C, but if the data point at 20 "C is included, a SP = c + rR2 + s$' + aa,H + b(Jy + I log L16354 Analytical Communications, October 1996, Vol33 quadratic fit is superior [R* is generally > 0.99 except for the a constant on sorbent 3M(a)] and is similar to trends observed in the temperature dependence of the system constants in GLC, for which more extensive data over wider temperature ranges are a ~ a i l a b l e .~ ~ ? ~ ~ The data at 20 "C is typical of normal sorbent trapping conditions and will be all that is interpreted here. The styrene-divinylbenzene polymers differ primarily in surface area (and therefore presumably pore structure) with the 3M(a) polymer having a surface area of 350 m2 g-l, 3M(b) 880 m2 g.--l, and Chromosorb 106 700-800 m2 g-1. The main contribution to retention is dispersion, as represented by the 1 constant with lone pair-lone pair electron repulsion reducing retention (the r constant is negative for three of the four sorbents).Dipole-type interactions are not significant (the s constant is zero). The general structure of the sorbent contains no functional groups to act as a hydrogen-bond acid, yet some capacity for hydrogen-bond interactions is identified by the model.All four styrene-divinylbenzene sorbents are weak hydrogen-bond acids (b constant) while all but the 3M(a) polymer are weak hydrogen-bond bases (a constant). This would suggest that the polymers contain unanticipated polar impurities in their structure introduced during synthesis or during thermal conditioning and use.Their general influence on retention will be discussed subsequently. Available data for Porapak R, and some representative data for Tenax polymer^,^ is included in Table 1 for comparison with the styrene-divinylbenzene sorbents. Porapak R is a poly- (vinylpyrrolidone) polymer with a high selectivity for hydro- gen-bond acids (large a constant), as would be anticipated from its structure, although the fit of the data to the model is only modest, in part due to the limited variety and number of the solutes reported.Retention on Tenax GC is dominated by dispersion, but in contrast to the styrene-divinylbenzene sorbents, a significant contribution results from dipole-type interactions (large s constant). This shows a fundamental difference in the capacity of the styrene-divinylbenzene and Tenax sorbents for polar interactions and is an illustration of their complementary sorbent properties.The contribution of the various intermolecular interactions to the retention of four varied solutes on the two styrene- divinylbenzene polymers 3M(a) and 3M(b) and Tenax GC is summarized in Table 2. The (c + llog ,516) term is taken as a rough measure of the cavity and dispersion contributions to retention and is clearly dominant and sensibly increases with the surface area for the two styrene-divinylbenzene sorbents.The very significant contribution of dipole-type interactions on Tenax is represented by the large contribution of sn2H to the retention of benzyl cyanide and benzyl alcohol.The sorbent Table 1 System constants for porous polymer sorbents at different temperatures System constants* Sorbent TemperaturePC Styrene-Divinylbenzene- 3M(a) 20 70 75 80 85 90 3Wb) 20 70 80 85 90 Gallant et al. 20 Chromo- sorb 106 20 Poiy(viny1pyrroiidone)- Porapak R 20 C - I .92 (0.13) -2.73 (0.10) -2.81 (0.12) -2.88 (0.11) -2.89 (0.12) -2.92 (0.09) (0.20) -2.60 (0.07) (0.09) -2.82 (0.08) (0.08) (0.21) -1.87 -2.77 -2.90 -2.46 - 1.69 (0.16) - 1.75 (0.28) Poly(2,6-diphenylphenylene oxide)- TenaxGC 20 -2.54 (0.12) 20 -3.28 (0.1 1) r -0.70 (0.13) -0.38 (0.10) -0.32 (0.1 1) -0.32 (0.10) -0.26 (0.11) -0.22 (0.08) -0.75 (0.22) -0.33 (0.07) -0.33 (0.08) (0.07) (0.07) 0.44 (0.23) -0.29 -0.26 -1.17 (0.13) -0.67 (0.27) -0.61 (0.19) -0.36 (0.18) S a 0 0.54 (0.19) 0 0.57 (0.16) 0 0.64 (0.19) 0 0.61 (0.17) 0 0.61 (0.18) 0 0.56 (0.14) 0 0 0 0 0 0 0 0 0 0 0 0.85 (0.21) 0 0.67 (0.25) 0.37 2.02 (0.21) (0.51) 1.26 0 (0.21) 0.70 0 (0.18) h 0.85 (0.12) 0.80 (0.10) 0.68 (0.11) 0.69 (0.10) 0.66 (0.1 1) 0.63 (0.08) 0.91 (0.22) 0.59 (0.07) 0.62 (0.08) 0.60 (0.07) 0.60 (0.07) 0.65 (0.24) 0.83 (0.17) 0 0 0.43 (0.16) 1 1.21 (0.05) 1.01 (0.04) 1 .oo (0.05) 0.98 (0.04) 0.95 (0.05) 0.93 (0.04) 1.27 (0.08) 1.03 (0.03) 1.02 (0.04) 0.99 (0.03) 0.98 (0.03) 1.29 (0.07) 1.51 (0.05) 1.19 (0.lo) 1.39 (0.04) 1.41 (0.04) R 0.994 0.993 0.991 0.993 0.991 0.995 0.984 0.997 0.996 0.996 0.995 0.978 0.993 0.975 0.975 0.990 SE 0.067 0.06 0.07 0.06 0.06 0.05 0.12 0.04 0.05 0.04 0.04 0.26 0.15 0.29 0.34 0.20 F 238 224 17 1 214 164 273 125 717 400 416 363 175 370 53 647 513 Statistics' Reference n for data 16 12 17 17 17 17 17 16 12 17 17 16 15 37 12 27 15 16 16 104 17 51 18 * Numbers in brackets are the standard deviation in the coefficient.' R = correlation coefficient; SE = standard error in the estimate; F = F- statistic; n = number of solutes.Analytical Communications, October 1996, Vol33 355 Table 2 Contribution of intermolecular interactions to retention on porous polymers at 20 "C Solute Polymer Propylbenzene 3M(a) 3Wb) Tenax GC 3M(b) Tenax GC 3M@) Tenax GC Benzyl alcohol 3M(a) 3Wb) Tenax GC Di-n-butyl ether 3M(a) Benzyl cyanide 3M(a) Intermolecular interactions c+l log L'6 3.189 3.508 3.327 2.819 3.118 2.903 2.958 3.265 3.062 3.178 3.496 3.314 rR2 -0.421 -0.453 -0.366 0 0 0 -0.5 17 -0.557 -0.450 -0.606 -0.602 -0.487 SJt? 0 0 0.63 1 0 0 0.316 0 0 1.401 0 0 1.098 a@ 0 0 0 0 0 0 0 0 0 0.179 0 0 Solute descriptors used in the calculation Log L'6 R2 Jt: a: E Propy lbenzene 4.230 0.604 0.50 0 0.15 Di-n-butylether 3.924 0 0.25 0 0.45 Benzyl cyanide 4.039 0.742 1.11 0 0.33 Benzyl alcohol 4.22 1 0.803 0.87 0.33 0.56 bPF 0.128 0.136 0 0.384 0.409 0 0.282 0.300 0 0.478 0.509 0 Predicted log vg/ 1-1 g 2.896 3.191 3.592 3.203 3.527 3.219 2.723 3.008 4.013 3.229 3.403 3.925 hydrogen-bond basicity of 3M(a) contributes to the retention of benzyl alcohol but at a level which is still quite small and less important than the hydrogen-bond acidity of the styrene- divinylbenzene sorbents.The results in Table 2 may seem contrary if a solute focus is used to explain retention properties, as is commonly the case.For example, in the case of benzyl alcohol, its capacity for hydrogen-bond interactions (partic- ularly hydrogen-bond acidity) is of minor importance compared with other interactions, and viewing the properties of benzyl cyanide in terms of its dipolarity makes no sense in under- standing its retention on the styrene-divinylbenzene sorbents, since these sorbents have no capacity for interactions of this kind.Sorbent selection must be based on first characterizing the capacity of the sorbent for specific intermolecular interactions and then using that information to predict solute retention, as we have done for a few compounds as an illustration in Table 2. A comprehensive treatment of the retention properties of a styrene-divinylbenzene sorbent [similar to 3M(a)] for the extraction of solutes from water is available elsewhere.lo Two features are of particular note: retention is dominated by the ease of cavity formation in the solvated sorbent and the strong hydrogen-bond acidity of water; and the uptake of the sample processing solvent by the sorbent produces large changes in the phase ratio and sorbent selectivity.Chambers and Fritz have recently provided retention data for a number of varied solutes on a styrene-divinylbenzene sorbent in non-aqueous solvents. 1 The solvation parameter model, eqn. (2), is fitted to this data in Table 3. For condensed phases solute transfer occurs with an approximate cancelling of dispersion interactions and thus eqn.(1) is modified to take account of this by using the characteristic volume (V,) as the solute descriptor for the cavity term and m as the system constant representing the difference in the ease of cavity formation between the solvated sorbent and the sol- vent.9*10.21,22 The fits to the model are reasonable given the (2) small capacity factor values observed with the three solvents and the accepted experimental difficulty in making such measurements accurately. For acetonitrile, methanol and etha- nol the ease of cavity formation in the solvated polymer (rn constant) and the capacity of the solvated sorbent for electron lone pair-lone pair interactions ( r constant) favour retention. Hydrogen-bond acid or base interactions diminish retention, and a major contribution to the selectivity difference between SP = c + mVJ100 + rR2 + m y + aa? + bpy Table 3 Sorption properties of a styrene-divinylbenzene sorbent in organic solvents System constant Acetonitrile" Methanol* Ethanol* m 0.59 (0.07) 0.66 (0.06) 0.32 (0.06) r 0.53 (0.08) 0.38 (0.05) 0.37 (0.07) -0.46 (0.12) 0 0.27 (0.10) a -0.50 (0.06) -1.03 (0.05) -1.17 (0.05) S b -0.76 (0.12) -0.82 (0.08) -0.87 (0.10) C -0.74 (0.12) -0.44 (0.08) -0.42 (0.07) Statisticst- R 0.960 0.978 0.986 SE 0.08 0.07 0.07 F 58 146 154 n 30 30 28 t See Table 1 for description of terms.* Numbers in brackets are the standard deviations in the coefficients. the two alcohols and acetonitrile is seen in their different capacity to act as hydrogen-bond bases (a constant). Methanol and ethanol solvated sorbent has virtually identical Y, b, and c constants, from which it can be inferred that the phase ratio of the solvated sorbents and capacity for hydrogen-bond acid and electron lone-pair interactions are about the same.The two solvents differ primarily in their cohesive density, hydrogen- bond basicity, and capacity for dipole-type interactions. Acet- onitrile is most potent in its capacity to compete with the solvated sorbent in dipole-type interaction (s constant), while the solvated sorbent can compete effectively with methanol and ethanol for these interactions.Since pure organic solvents are commonly used to recover trapped analytes from porous polymer sorbents after extraction the solvation parameter model can be used to select the appropriate solvent to provide the largest concentration factor for an analyte.References 1 2 3 4 Poole, C. F., and Poole, S. K., Chromatography Today, Elsevier, Amsterdam, 199 1. Castello, G., and D'Amato, G., Chromatogruphia, 1987, 23, 839. Hepp, M. A., and Klee, M. S., J . Chromatogr., 1987,404, 145. Castello, G., and D'Amato, G., J . Chromatogr., 1983, 254, 69.356 Analytical Communications, October 1996, Vol33 5 6 7 8 9 10 11 12 13 14 Poole, C.F., and Poole, S. K., Chem. Rev. (Washington, D. C.), 1989, 89, 377. Poole, C. F., Kollie, T. O., and Poole, S. K., Chromatographia, 1992, 34, 28 1. Abraham, M. H., and Walsh, D. P., J . Chromatogr., 1992, 627, 294. Grate, J. W., Abraham, M. H., Du, C. M., McGill, R. A., and Shuely, W. J., Langmuir, 1995, 11, 2125.Poole, C. F., Poole, S. K., Seibert, D. S., and Chapman, C. M., J . Chromatogr. B, Biomed. Appl., 1996, in the press. Poole, S. K., and Poole, C. F., Analyst, 1995, 120, 1733. Chambers, T. K., and Fritz, J. S., J . Chromatogr. A, 1996, 728, 271. Pankow, J. F., Luo, W., Isabelle, L. M., Hart, K. M., and Hagen, D. F., J . Chromatogr. A, 1996, 732, 317. Abraham, M. H., Chem. Soc. Revs., 1993, 22, 73. Abraham, M. H., Andonian-Haftvan, J., Whiting, G. S., Leo, A., and Taft, R. S., J . Chem. Soc., Perkin Trans. 2, 1994, 1777. 15 16 17 18 19 20 21 22 Taylor, D. G., Kupel, R. E., and Bryant, J. M., NZOSH Publication 77-1 85, US Department of Health, Education and Welfare, Cin- cinnati, OH, 1977. Stanetzek, I., Geise, U., Schuster, R. H., and Wunsch, G., Am. Ind. Hyg. Assoc. J., 1996, 57, 128. Pankow, J. F., Anal. Chem., 1988, 60, 950. Brown, R. H., and Purnell, C. J., J. Chromatogr., 1979, 178, 91. Poole, C. F., and Kollie, T. O., Anal. Chim. Acta, 1993, 282, 1. Poole, S. K., Kollie, T. O., and Poole, C. F., J. Chromatogr., 1994, 664, 229. Seibert, D. S., and Poole, C. F., Chromatogruphia, 1995, 41, 51. Seibert, D. S., Poole, C. F., and Abraham, M. H., Analyst, 1996,121, 511. Paper 6f045SI K Received July I , 1996 Accepted August 19,1996

ISSN:1359-7337    

DOI:10.1039/AC9963300353

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Vol33 (357-359) 357 Reversed-phase Liquid Chromatographic Method for the Determination of pAminobenzoic Acid and Anthranilic Acid in Urine: Application to the Investigation of Exocrine Pancreatic Function Abdul Rob=, Wayne H. Bradburyb and Alexander R. W. Forrest" a Department of Biological and Chemical Sciences, University of Essex, Wivenhoe Park, Colchester, UK C04 3SQ b Department of Chemical Pathology, Royal Free Hospital, London, UK NW2 3QJ UK SlO 2JF Department of Clinical Chemistry, Royal Hallamshire Hospital, ShefSield, Exocrine pancreatic function can be assessed by determining the recovery in urine of p-aminobenzoic acid (PABA) following the oral administration of (N-benzoyl-L-tyrosy1amino)benzoic acid (bentiromide).However, misleadingly abnormal results often occur in patients with abnormal intestinal, liver and renal function.Co-administration of 14C-PABA and the determination of the ratio of the recoveries of PABA to 14C-PABA, the PABA excretion index (PEI), has been used to eliminate these misleading results. Unfortunately, this has the disadvantage of exposing patients to radioactivity.We report the development and optimization of a modified test in which anthranilic acid, the ortho isomer of aminobenzoic acid, is substituted for the radiolabelled PABA. PABA and anthranilic acid in urine are determined by a C18 reversed-phase HPLC method. Peak detection is carried out at 254 nm. The method is specific for PABA and anthranilic acid, pTecise and fairly rapid.The PABA excretion index values in patients with pancreatic steatorrhoea (0.18-0.43) determined using the new liquid chromatographic method were clearly separated from those in healthy controls (0.63-1.14). After oral administration, the synthetic peptide (N-benzoyl-L- tyrosylamino) benzoic acid (bentiromide) is h ydrol ysed by pancreatic chymotrypsin (EC 3.4.21.1) to benzoyltyrosine and p-aminobenzoic acid (PABA) (Fig.1). The determination of the proportion of PABA recovered in urine following oral admini- stration has been shown to distinguish between normal subjects and those with chronic pancreatitis. However, low recoveries of PABA may also occur in patients with gastrointestinal mucosal disease, liver disease or renal disease.2 This is because the recovery of PABA in urine depends not only on chymo- trypsin hydrolysis but also on the absorption, conjugation and excretion of PABA.The specificity of the test in the diagnosis of pancreatic disease may be improved by performing a second test, replacing oral bentiromide by PABA, to correct for abnormal PABA absorption, metabolism and renal handling.3 This two-day test is inconvenient for clinical use.A single day test where a tracer dose of I4C-PABA is administered with the bentiromide, has been in~estigated.~.5 Unfortunately, this introduces the dis- advantage of administering radioactivity to patients and re- quires access to a liquid scintillation counter. We report the development and optimization of a modified test using the ortho isomer of arninobenzoic acid (anthranilic acid) (Fig.2) in place of W-PABA, with the determination of both compounds by a new reversed-phase high-performance liquid chromatographic method. Experimental Apparatus We used a high-performance liquid chromatography (HPLC) system fitted with a Novapak C18, 5 pm radial compression N-benzoyl-L-tyrosyl-p -aminobenzoic acid (Bentiromide) HZ0 a-chymotrypsin t p-Aminobenzoic acid (PABA) Bentoyl tyrosine Fig.1 Synthetic peptide (N-benzoyl-L-tyrosy1amino)benzoic acid (bentir- omide) is hydrolysed by pancreatic chymotrypsin (EC 3.4.21.1) to benzoyltyrosine and p-aminobenzoic acid (PABA). PABA is urine is determined by reversed-phase high-performance liquid chromatography in a procedure requiring no extraction.358 Analytical Communications, October 1996, Vol33 cartridge with M45 pump system and a Model No.440 detector (Waters Associates, Northwich, Cheshire, UK). Reagents Bentiromide tablets containing 333 mg of (N-benzoyl-L- tyrosy1amino)benzoic acid (bentiromide) were obtained from Roche Products Ltd., Welwyn Garden City, Hertfordshire, UK. Anthranilic acid (orthoaminobenzoic acid) and m-hydroxy- benzoic acid were obtained from Sigma Chemical Co., Poole, Dorset, UK.Gelatine capsules containing 1 13 mg of anthranilic acid were prepared by the Royal Hallamshire Hospital phar- macy, Sheffield, UK. Other chemicals were obtained from Merck Ltd., Lutterworth, Leicestershire, UK. The mobile phase (2.8% v/v) was prepared by dissolving 25 ml of methanol (HPLC grade) and 5.5 g of sodium acetate trihydrate in 900 ml of distilled water.The pH was adjusted to 3.0 with glacial acetic acid before the addition of MEOH because pH adjustment in the presence of organic solvent is unsatisfactory. Administration of Bentiromide The test was performed on a total of 19 patients with pancreatic steatorrhoea. All gave informed consent to the procedure. Also, we obtained consent from volunteers and the patients with no evidence of pancreatic disease.The consent form used and the experimental protocol were approved by the hospital ethical committee. After an overnight fast, the subjects emptied their bladders and then orally received 3 X 333 mg of bentiromide tablet and 3 X 113 mg of anthranilic acid capsule with 300 ml of water. Urine was collected for 6 h.The subjects were encouraged to drink during the collection period but were requested not to eat for 4 h. The total volume of each specimen was recorded. If the volume was less than 1 1 the urine was diluted to 1 1 with distilled water. Aliquots of urine were stored at -20 "C. Chromatographic Conditions One millilitre of urine was added to 1 ml of 8 moll-1 sodium hydroxide containing 1 g 1- 1 of rn-hydroxybenzoic acid (Fig.2) as internal standard in a screw-topped glass tube. The tube was heated to 120 "C for 1 h. After cooling, 50 1-11 of the hydrolysate were added to 950 1-11 of 0.25 mol 1-1 acetic acid, mixed, and centrifuged at 5000 rev min-1 for 5 min. Clear supernatant (25 pl) was injected onto the CIS reversed-phase HPLC column. Chromatography was carried out at room temperature using methanol as a mobile phase in 40 mmol 1-1 acetate buffer.The flow rate was 2 ml min-1 and detection was carried out at 254 nm. Results and Discussion The chromatograms obtained under different conditions are shown in Fig. 3. For each batch, standard curves were constructed by plotting the peak height ratio (height of A QH* Q"" kOOH kOOH o-aminobenzoic acid m-hydroxybenzoic acid (Anthranilic acid) Fig.2 Chemical structures of anthranilic acid and m-hydroxybenzoic acid. Anthranilic acid is used as a substitute for 14C-PABA in the exocrine pancreatic function test. m-Hydroxybenzoic acid is used as an internal standard during PABA determination using a reversed-phase liquid chromatographic system. compound : height of internal standard) versus concentration for both PABA and anthranilic acid.The recovery of PABA and anthranilic acid, as a proportion of the administered dose, was also calculated for each subject. A PABA excretion index (PEI) (% recovery of PABA : % recovery of anthranilic acid) was subsequently determined for each subject (Table 1). It should be noted that it is more meaningful to report the PEI rather than only the recoveries of p-aminobenzoic acid or anthranilic acid because misleading PABA results often occur in patients with abnormal renal and intestinal malfunction.2 Fig.3(a) demon- strates the chromatogram obtained for the pre-test urine sample (blank urine); Figs. 3(b) and (c) illustrate the chromatograms recorded from control subjects and patients with pancreatic disease, respectively.PABA determination in urine has traditionally been carried out by calorimetric methods based on the reactions of aromatic amines such as the Bratton and Marshall reaction.6 These methods frequently give falsely high results with the patient taking certain drugs including paracetamol, ibuprofen and sulfanilamide.7-9 Thus, it is essential to use an alternative procedure which would be interference free.The increased specificity of high-performance liquid chromatography might obviate this problem. The determination of PABA and anthra- nilic acid by high-performance liquid chromatography could open the possibility of correcting for inter-patient variation in intestinal absorption, hepatic and renal handling of PABA without the requirement for '4C-PABA administration.Whilst anthranilic acid is not an approved drug, it is lawful in the United Kingdom for a registered medical practitioner to P - 0 2 4 6 8 1 0 Retention time/min Fig. 3 Chromatograms of: (a), pre-test (blank) urine sample of healthy volunteer; (b), post-test urine sample from control healthy subject; and (c), post-test urine specimen from patient with pancreatic steatorrhoea.P = p - Aminobenzoic acid; IS = internal standard (m-hydroxybenzoic acid); A = anthranilic acid (o-aminobenzoic acid). Column, C18 Novapak radial compression cartridge; mobile phase, 2.8% v/v methanol in 40 mmol 1-I acetate buffer, pH 3.0; flow rate, 2 ml min-I; detector, 254 nm. Table 1 The recovery of PABA and anthranilic acid, as a proportion of the administered dose, was calculated for each subject.A PABA excretion index (% recovery of PABA: % recovery of anthranilic acid) is presented for different conditions PABA excretion Condition index (PEI) Patients with no evidence of pancreatic disease n = 10 Normal subject n = 10 0.63-1.14 0.63-1.14 Pancreatic steatorrhoea n = 19 0.18-0.43Analytical Communications, October 1996, Vol33 359 administer it to individual, named patients in the manner we have described.The limit of detection for the present HPLC method is calculated by determining the concentrations of PABA and anthranilic acid resulting in a signal 3X the mean noise level from 10 random urines. This was 6 mg 1-1 for PABA and 12 mg 1- 1 for anthranilic acid. The PABA and anthranilic acid standard curves were linear through the range 0-500 mg 1-I.The within-run and between-run precision was evaluated by the repeated analysis of a single sample (Table 2). Pre-test, first passed morning samples of urine from each of the subjects in the study were collected and assayed. No significant interfering peaks were found from various commonly used drugs in the areas of either PABA, anthranilic acid or the internal standard (Table 3), except for the drug sulfanilamide.The peak for sulfanilamide was close to the peak of PABA. Thus, care must be taken in interpreting PABA results in patients taking sulfanilamide drug formulations. Extensive literature searches have failed to uncover data that suggests that anthranilic acid has any significant toxicity in animals.10 This is the first reported method for the simultaneous determination of PABA and anthranilic acid.Our method is relatively precise and specific. No peaks that might potentially interfere in the assay were found in the chromatograms obtained from either pre-test urine samples from those participating in the study or in those samples collected randomly from hospital in-patients (Fig. 3).PABA excretion index values in healthy volunteers (0.63-1.14) were clearly separated from those in patients with pancreatic steatorrhoea (0.18-0.43) (Table 1). The confirmation of the pancreatic steatorrhoea in the participating patients was also revealed from faecal fat measurement ( > 20 mmol 1- I ) and endoscopic retrograde pancreatographic (ERP) investigation (data not shown). The healthy volunteers (n = 10) (age range 26-47 years, medical and laboratory staffs) and patients with no evidence of pancreatic disease (n = 10) (judged clinically), who partici- Table 2 The within-run and between-run precision for PABA and anthranilic acid Anthranilic Experimental PABA acid Within-run n = 10 4.9 6.6 condition (%CV) (%CV) Day today run n = 10 3.0 4.5 Table 3 Results of testing various drugs for possible interferences, during chromatography, from PABA, anthranilic acid and the internal standard.No significant interfering peaks in the areas of either PABA, anthranilic acid or the internal standard by these drugs is represented as -; slight interference is indicated by + Anthranilic Internal Drugs PABA acid standard Paracetamol - - - - - Ibuprofen - Digoxin - Chloropropamide - - - - - Frusemide - - - Atenol - - - Sulfanilamide - - Sulfadiazine - - + - pated in the study made no complaint of any side affects such as dizziness, nausea or gastrointestinal disorders due to consump- tion of anthranilic acid.The range of PABA Excretion Index (PEI) for both groups (volunteers and patients with no sign of pancreatic disease) are the same, despite the fact that they are different subjects (Table 1).As anthranilic acid is only available as a chemical, the first step for its wide use in pancreatic disorder assessment would be to formulate the chemical as a capsule. In this respect, the results described in this paper would help to achieve this objective. We have found that the optimum dose of anthranilic acid needed for the test is 339 mg and this could be used as a guide for the amount needed in the formulation of the capsule.The optimum dose of anthranilic acid implies the amount used in the test which is not toxic to the subjects. Furthermore, the analytical procedure which we have developed not only can be used as a novel means for the determination of anthranilic acid in urine, but also to check the purity of the chemical prior to formulating it as capsules.The clinical usefulness of the new method needs to be fully tested by running a study simultaneously with one of the established techniques (e.g., radiolabelled PABA method) and results analysed using suitable statistical tests. Conclusions This paper has reported the development and optimization of a new liquid chromatographic method for the determination of PABA and anthranilic acid.The use of anthranilic acid as a substitute for "T-PABA is welcomed by hospital patients, since there is no risk of contamination by radioactivity. Furthermore, the analytical procedure, which we have described for the determination of PABA and anthranilic acid, is free from interferences from commonly present conjugated drugs in urine samples.The work described in this paper has, without doubt, laid down the foundation for further exploration of the use of anthranilic acid as a substitute of "C-PABA with the possibility of industrial scale formulation of this chemical as capsule. We are grateful to Dr. K. Cleur for supplies of Bentiromide and the pharmacy staff of the Royal Hallamshire Hospital for preparing the anthranilic acid capsules.The chromatographic system in the present research was purchased with a grant from the trustees of the former United Sheffield Hospitals. We would like to thank Dr. C. D. Holdsworth and Dr. J. R. Worters of the Gastroenterology Department, Royal Hallamshire Hospital, for allowing their patients to participate in the study. References 1 2 3 4 5 6 7 8 9 10 Furuya, K. N., Clin. Biochem., 1995, 28, 531. Chaloner, C., Clin. Chim. Acta, 1995, 233, 89. Pemberton, P. W., Clin. Chim. Acta, 1991, 199, 253. Mitchell, C. J., Field, H. P., and Losowsky, M. J., Br. Med. J., 1981, 282, 1751. Braganza, J. M., Br. Med. J., 1984, 289, 562. Bratton, A. G., and Marshall, E. K., J . Biof. Chem., 1935, 128, 537. Braganza, J. M., and Herman, K. J., Clin. Chim. Acta, 1983, 130, 339, Ito, S., and Imai, Y., Clin. Chem., 1982, 28, 323. Berg, J. D., and Lawson, N., Ann. Clin. Biochem., 1985, 22, 586. Lindeman, S. V., Science (Washington, D.C.), 1990, 39, 2383. Paper 6f05147B Received July 23, 1996 Accepted August 29, 1996

ISSN:1359-7337    

DOI:10.1039/AC9963300357

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Val 33 (361 -364) 36 1 Solid-phase Microextraction Combined With Elect roc hem istry Feng Guo, Tadeusz Gorecki,? Donald Irish and Janusz Pawliszyn* Department of Chemistry and Waterloo Centre for Groundwater Research, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1 Solid-phase microextraction combined with electrochemistry (SPME/EC) for trace level mercury determination is reported.An SPME/EC fibre was made of a carbon steel wire with a 30 pm gold coating. Mercury(I1) ions were electrochemically extracted from aqueous solution, and desorbed with a dedicated desorption system, then detected by ion-trap GC-MS. Mercury(@ ions in aqueous solution, and mercury vapour in gas, were detected. Inorganic mercury and organomercury compounds were differentiated.Solid-phase microextraction (SPME) is a new and promising sample preparation technology. The process of SPME can be described as the extraction of analytes from a sample matrix to the coating on a miniature cylindrical fibre by intermolecular forces or other forces, followed by desorption and detection by various instruments, including GC-MS,' GC-FID,* HPLC3 and spectroscopy.4 So far SPME has been mainly used for the analysis of organic compounds.1.536 SPME has also been applied for the analysis of Pb"7 and Hg"8 after derivatization, and Bilr19 by using ion- exchange resin coatings. SPME combined with electrochemistry was first proposed in a review.10 The idea of SPMElelectrochemistry (SPMEEC) is that analytes are electrochemically oxidized or reduced to desirable derivatives on microelectrodes with conducting coatings. The derivatives are extracted in situ by the coating through partitioning. It is the purpose of this communication to illustrate the feasibility of combining SPME with electro- chemistry in order to broaden the scope of SPME applications to analytes that are not normally amenable to analysis by this method. In this research, mercury and mercury compounds have been chosen as the first species for measurement by SPMEEC.As a consequence of its high toxicity, interest in the detection of Hg in environmental materials continues. CVAAS and differential pulse anodic stripping voltammetry (DPASV) are the most popular methods for mercury determination. 1,12 However, they both suffer from interferences from other species (e.g., volatile compounds which absorb at 253.7 nm and certain inorganic species for CVAAS,13,14 transition metal cations in DPASV15). Gas chromatography and reversed-phase LC and HPLC, coupled with different detectors, such as CVAAS,l6 ECD17 or MS,'* have been successfully applied in organomercury speciation.Some of these methods are also capable of determining inorganic mercury after alkylation,19 but the methods involve time-consuming work-up procedures.Chro- matography is not very suitable for the determination of the total amount of mercury present. * To whom correspondence should be addressed. + On leave from the Faculty of Chemistry, Technical University of Gdansk, Poland. In SPMEEC, mercury is first extracted into a 10 pm gold coating on a 140 pm od carbon steel electrode by electrodepo- sition at a preset potential for an appropriate time.The gold coated steel fibre with Hg is then placed in a dedicated injector of a GC-MS system, and the mercury is thermally desorbed and subsequently analysed. Inorganic mercury has been reported to be reduced on a gold microelectrode at -0.2 V (SCE), and the half-wave potentials of organomercury compounds, such as methyl, ethyl, n-propyl, n-pentyl, and phenyl mercury, were reported to cover a narrow range from -0.53 to -0.61 V (SCE).20 It is expected therefore that inorganic mercury can be differentiated from organomer- cury compounds, but difficulties in speciation will exist for the latter.The wet digestion method can be used to mineralize organomercury compounds, so that inorganic mercury(11) can be released and then detected.Therefore, inorganic mercury and total mercury can also be determined by SPME/EC. SPME/EC not only reduces Hg, but also preconcentrates it in the same step, In addition, volatile compounds which interfere in AAS are expected to cause no interferences in this method.Compared with DPASV, analytes after electrodeposition are then analysed by a very sensitive and selective MS. Experimental Reagents KN03 solution (1 .O moll-1) was prepared by dissolving KN03 (Caledon Laboratories, Georgetown, Ontario, Canada) in ultrapure water (Nanopure, Barnstead, USA). Mercury(@ standard solution (10.0 ppb) was prepared by spiking 0.0005 moll-' HgN03 (Sigma, St.Louis, MO, USA) in 1000 ml of 1.0 mol 1-1 KN03. A Hg-EDTA complex was formed by spiking 25 ml of 50 ppb Hg" solution with 1 ml of 0.01 moll-1 EDTA solution. Apparatus SPMEIEC device Owing to its good mechanical properties, a gold coated carbon steel wire (140 pm od) was used. The electrochemical SPME device was constructed by modifying a Hamilton 7 105 syringe using the procedure described in detail elsewhere.*l The only difference was that stationary phase coated fibre was replaced by a gold coated wire.Electrochemical extraction cell As shown in Fig. 1, a three electrode cylindrical minicell of 3.5 cm id and 6.0 cm in height, accommodating 5 ml of the sample, was used to facilitate transport of the ions to the working electrode. The SPMEEC device acted as the working electrode, a platinum wire was the counter electrode, and an Ag/AgCl standard electrode was the reference electrode.The working electrode was closely surrounded by the counter electrode wire362 Analytical Communications, October 1996, Vol33 at a distance of about 1.5 cm. This enabled efficient transport of Hg" ions to all points of the working electrode surface. The reference electrode was connected with the main cell by a short Teflon capillary, whose tip was located about 0.5 cm from the working electrode. In this way, the IR potential drop was minimized.A potentiostat (Pine Instrument, Grove City, PA, USA) was used to keep the working electrode potential at -0.2 V (versus Ag/AgCl standard electrode). Magnetic stirring was used.For more details, refer to the Results and Discussion. Dedicated desorption system The schematic of the dedicated desorption system22 is shown in Fig. 2. Besides the main body of the injector, other components of this system included a dc supply, an ac supply and relay, a capacitor and a switch, a resistance digital meter and a voltage digital meter (not shown).The dc power supply consisted of an autotransformer and an ac/dc bridge. A 1000 pF/150 VDC Fig. 1 SPME/EC minicell. 1, Reference electrode, Ag/AgCl standard electrode; 2, 100 ml plastic syringe; 3, Teflon capillary (2 mm); 4, 8 ml Teflon vial; WE, working electrode, SPME/EC device; CE, platinum wire. 1 A oms-to Ion-Trap MS Fig. 2 Capacitive discharge desorption system. 1, Syringe; 2, electric connection I; 3, injector body; 4, steel wire; 5 , gold coating; 6, electric connection 11; 7, transfer line; 8, capacitor; 9, relay; 10, butt connector.capacitor was used for the capacitive discharge. The electrical connections were the SPMEEC device and some purpose- designed stainless steel tubing, to which one end of the transfer line was glued. One end of the stainless steel tubing had a constriction through which the gold coated wire could not go, but the carrier gas could.The constriction assured a good electric contact between the wire and the tubing. The transfer line was placed between the desorption system and the ion-trap MS. Most of the transfer line was kept in the GC oven at 100 "C. GC-MS (ion-trap) and transfer line GC-MS (ion-trap) Saturn I1 (Varian) was used.Segment RF settings were set to 250, equivalent to storage mass of 40 u. The AGC target was set at 25 000. The He carrier gas pressure in GC was set at 20 psi (1 psi = 6894.76 Pa). The GC oven temperature was set at 100 "C. Deactivated fused silica tubing and a 2 m X 0.1 mm x 0.5 pm SPB-1 column (Supelco, Bellefonte, PA, USA) were used as transfer lines.Procedure Mercury vapour in gas detected by SPME After a drop of mercury (about 1 mm in diameter) was added to a 20 ml vial, the vial was capped for 20 min. The gold coated wire was withdrawn into the syringe needle by retracting the plunger prior to piercing the septum of the vial. It was then exposed to the gas phase of the vial for 5 min by depressing the plunger. The plunger was retracted before the needle was removed from the vial, to avoid damaging the coating.The needle of the syringe was inserted through the septum of the dedicated injector. Then the plunger was depressed again, until the gold coated wire touched the stainless steel tubing connection, as in Fig. 2 (6). After applying the capacitive voltage pulse of 80 VDC to the wire, Hg was desorbed and analysed.Before the syringe was removed from the injector, the wire was again withdrawn into the needle. Mercury(zz) in solution detected by SPMEIEC Prior to use, the gold coated wire was cleaned at +0.8 V in 0.1 moll-1 HC104. All potential impurities and traces of Hg from previous analyses were oxidized at this potential and removed from the coating. Following that, the device was installed in a three-electrode cell with 10 ppb Hg" solution. The gold coated wire was exposed to the sample, and connected as the working electrode (-0.2 V versus Ag/AgCl) to the potentiostat.After 10 min, the power was turned off and the device was transferred to the dedicated GC/MS injector port. Desorption was performed in the same manner as for free mercury determination. Diferentiation of free inorganic Hg" and complexed Hg" by SPMEIEC SPME/EC was performed on a solution containing 50 ppb Hg" and excess EDTA.Two working electrode potentials were examined, -0.6 and -1.15 V versus SCE. Results and Discussion Metallic Mercury in Gas Detected by SPME Mercury desorption Before the start of the capacitive discharge desorption process, the voltage across the capacitor and the resistance between the gold coated wire and the stainless steel tubing connection were monitored by a voltmeter and a resistance meter, respectively, to assure repeatable desorption conditions.The discharge of theAnalytical Communications, October 1996, Vol33 363 capacitor, lasting only milliseconds, resulted in a rapid temperature rise of the gold coated wire, so that mercury should be dissociated from the amalgam very quickly. Owing to lack of the proper means, the temperature of the wire during the discharge could not be measured accurately.A rough estimate based on the glow of the wire during the heating pulse indicates that the temperature could reach as high as 800 OC, depending on the capacitor voltage.As the background was very pronounced at such high temperatures, in the experiments presented the temperature was limited to around an estimated 500 "C. Determination of mercury vapours This experiment was performed to verify the feasibility of the proposed method of mercury determination. A large mercury peak was obtained by exposing the gold coated wire to HgO vapours. Since mercury has relatively high vapour pressure, its concentration in the headspace of the vial was also high.Gold has very high affinity to mercury, so a large amount of HgO can be absorbed by the gold coating. The capacitive discharge desorption system produced very high and sharp mercury peaks. Compared to peaks produced by the conventional SPI injector (Septum Programmable Injector, Varian Palo Alto, CA, USA), they were about 100 times higher (total ion current).Besides, there was a large carryover of mercury (over 50%) when using SPI. The carryover when using the capacitive discharge desorption system was significantly smaller ( < 30%). Because strong interaction exists between mercury atoms and gold atoms in the amalgam, heat provided at the highest temperature of the SPI desorption system, 300 OC, was not enough to release mercury out of the amalgam matrix rapidly.By using the capacitive discharge desorption system, the wire was heated to a high temperature in a very short time, which released mercury out of the wire more effectively. Carryover could be further reduced in the dedicated system by redesigning the injector in such a way that all the elements exposed to high temperature are temperature resistant.It should be noted, however, that carryover does not constitute a problem as long as it is repeatable. Traces of Hg" left in the gold coating after thermal desorption were removed during the conditioning step, cleaning at +0.8 V (SCE) for 5 min in HC104, before the next extraction. The blank after such procedure did not reveal any peaks.Inorganic Mercury Ions in Solution Detected by SPMEIEC Electrochemical reduction of mercury in solution At -0.2 V (Ag/AgCl), Hg" ions could be readily reduced and absorbed by the gold coating forming an amalgam. Since the level of Hg" in the solution was very low (10 ppb), the electrochemical reduction of mercury was diffusion controlled.The cell content was stirred vigorously with a magnetic stirrer. Fast stirring can decrease the thickness of the electrical double layer, increasing the concentration gradient in the diffusion layer and the ion transfer rate, thus accelerating Hg deposi- tion. In Fig. 3, the first peak is the HgO peak after desorption from the gold coated wire. The mass spectrum of this peak is the characteristic mercury isotope mass spectrum with correct isotopic abundances. The second peak comes from interfering substances from the column and butt connector.For details about these interfering substances, refer to the next section. Choices of transfer line Surface deactivated fused silica tubing was first examined as the transfer line, in order to avoid adsorption of the desorbed mercury on the surface active sites.However, the mass spectrum of the peak obtained was not that of pure mercury. This indicated that the heat produced by the capacitive discharge not only stripped mercury off the gold coating, but also caused partial decomposition of some of the components present in the high temperature zone (the ferrule of the butt connector, made of polyimide, and possibly the transfer line, one end of which was directly glued to the stainless steel connection, as shown in Fig.2). Raw fused silica tubing was used then. However, the pure mercury mass spectrum was also not obtained. Consequently, an SPB-1 narrow bore column was tried to separate the decomposition products from Hgo. As shown in Fig. 3, two peaks were obtained in the total ion current chromatogram.The first peak was mercury with the correct abundances in the mass spectrum. It eluted in the dead time, because the column coating did not retain metallic mercury. The second peak was the substances released from the column and butt connector. Oven temperature Most of the transfer line was contained in the GC oven. Its temperature showed significant influence on separation of mercury and the interfering substances.For the 2 m column used, the resolution between mercury and the peaks of interfering substances decreased significantly with increase of temperature. At 300 OC, a single peak was observed. At 100 OC, mercury and the decomposed substances were well separated (Fig. 3). At temperatures lower than 100 OC, the liberated mercury was trapped in the system.It was also found that the He carrier gas flow rate and AGC target influenced the sensitivity and resolution of the mercury peak. A He carrier gas pressure of 20 psi and an AGC target of 25 000 yielded the optimal sensitivity and resolution. Detection limit The LOD, defined as the Hg" concentration in the aqueous sample producing an S/N of 3, was estimated on the basis of the analysis of an 800 ppt Hg" solution.The estimated value was =500 ppt for a 10 min extraction. It should be possible to further improve the LOD by using longer extraction times. Differentiation of Free Inorganic Mercury and EDTA-complexed Mercury by SPMEIEC SPMEEC of base electrolyte containing EDTA was first performed to determine the blank.No Hg" peak was observed. The solution was then spiked with Hg" to a concentration of 50 100 - T 2 TO7 i7j 0 3 100 200 300 400 500 0:lO 0:20 0:30 0:40 050 Time/s Fig. 3 Total ion current chromatogram obtained for 10 ppb Hg" solution, using the capacitive discharge desorption system. Arrow indicates the beginning of the capacitive discharge process. The two block graphs are the corresponding mass spectra of the two peaks.364 Analytical Communications, October 1996, Vol33 ppb.Extractions at -0.6 and -1.15 V (SCE) did not produce any measurable amounts of Hg, which proves that speciation between free and complexed mercury can be easily achieved by SPME/EC. Conclusions SPME/EC coupled with ion-trap GC-MS successfully detected inorganic mercury ions in solutions, and enabled differentiation of free inorganic mercury and complexed mercury.Direct detection of metallic mercury in gas was also possible. This study shows the possibility of applying SPME combined with electrochemistry to trace level analysis. In a future study, a conducting polymer will be used as a sorbent, which will extend SPME/EC to organic compound analysis.References 1 Potter, D., and Pawliszyn, J., Environ. Sci. Technol., 1994, 28, 298. 2 Arthur, C., and Pawliszyn, J., Environ. Sci. Technol., 1992, 26, 979. 3 Chen, J., and Pawliszyn, J., Anal. Chem., 1995, 67, 2530. 4 Wittkamp, B., and Tilotta, D., Anal. Chem., 1995, 67, 600. 5 Potter, D., and Pawliszyn, J., J. Chromatogr., 1992, 625, 247. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Buchholz, K., and Pawliszyn, J., Anal. Chem., 1994, 66, 160. Gbrecki, T., and Pawliszyn, J., Anal. Chem., in the press. Cai, Y., and Bayona, J. M., J. Chromatogr. A, 1995,696, 113. Otu, E., and Pawliszyn, J., Mikrochim. Acta, 1993, 112, 41. Pan, L., Adams, M., and Pawliszyn, J., Anal. Chem., 1995, 67, 4396. Zhang, Z., Yang, M., and Pawliszyn, J., Anal. Chem., 1994, 66, 844 A. Minagawa, K., and Takizawa, Y., Anal. Chim. Acta, 1980, 115, 103. Gill, G., and Fitzgerald, W., Mar. Chem., 1987, 20, 227. Lindstedt, G., Analyst, 1970, 95, 264. Rievaj, M., and Bustin, D., J. Electroanal. Chem., 1974, 50, 379. Sarzanini, C., and Sacchero, G., Anal. Chim. Acta, 1994, 284, 661. Evans, O., and McKee, G., Analyst, 1988, 113, 243. Drabaek, I., and Carlsen, V., Talanta, 1990, 37, 89. Bergdahl, A., and Schuts, A., Analyst, 1995, 120, 1205. Mairanovskii, G., Russ. Chem. Rev., 1976, 45, 298. Arthur, C., and Pawliszyn, J., Anal. Chem., 1990, 62, 2145. Gbrecki, T., and Pawliszyn, J., Anal. Chem., 1995, 34, 3265. Paper 61051 75H Received July 24, 1996 Accepted August 19, I996

ISSN:1359-7337    

DOI:10.1039/AC9963300361

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Comnzunications, October 1996, Vol33 (36.5-366) 365 Selective Fluorimetric Recognition of Dihydrogen Phosphate Over Chloride Anions by a Novel Ruthenium(i1) Bipyridyl Receptor Complex Paul D. Beera, Roger J. Mortirnerb?* Fridrich Szemesa and John S. Weightmanb (1 Inorganic Chemistiy Laboratory, Univei-sity of O,yfoi-d, South Parks Road, Oxford. UK OX1 3QR Leicestet-shire, UK LEI I 3TU Department of Chemistry, Loughhorough University, Loughborough, A novel ruthenium(I1) tris(bipyridy1) receptor complex has been prepared and shown by 1H NMR and fluorescence emission spectroscopy to exhibit exclusive selectivity for the dihydrogen phosphate anion in preference to chloride.Stimulated by how nature utilizes negatively charged species for numerous biochemically important pathways and poly- anions for the storage and transmission of genetic information, the molecular recognition of anionic guest molecules by positively charged or electron deficient neutral abiotic organic receptor molecules is an area of intense current As part of a research programme aimed at designing new spectroscopic and/or electrochemical sensory reagents for anions of both biochemical and environmental importance we are currently investigating systems based on ruthenium(r1) tris(bipyridy1) receptors that contain diamide-substituted li- gands such as [4,4’-bis( 2-methoxyethyl)carbamoyl]-2,2’-bipyr- idine.? One of the techniques we have used to probe anion binding is fluorescence emission spectroscopy and to date we have demonstrated remarkable chloride over dihydrogen phos- phate anion selectivity with novel ruthenium(I1) bipyridyl- metallocenec4 and fluorimetric recognition of chloride anions by the ruthenium(r1) complexes [RuL1(bipy)2][PF& and [Ru(L*),][PF6]2 { bipy = 2,2’-bipyridine; Ll = [4,4’-bis(2- me thox ye th y 1)carbamoy 1] -2,2’- bipyridine } .3 In the present communication we report that the latter complexes are also responsive to the binding of dihydrogen phosphate anions and show that exclusive selectivity to dihydrogen phosphate over chloride anions can be achieved using the novel mono(1igand) ruthenium(1r) complex [RuL2(bipy)2][PF6]2 { L2 = [4,4’-bis(2- methoxyethyl)carbonylamino]-2,2’-bipyridine } .Experimental Synthesis The preparation of the 4,4’-carbonyl amide disubstituted bipyridine ligand L* has been reported.’ The new acyclic ligand 0 0 L‘ L2 * To whom correspondence should be addressed.L2, [4,4’-bis(2-methoxyethyl)carbonylamino]-2,2’-bipy~dine, was prepared via the condensation reaction of 4,4’-bis(amino)- 2,2’-bipyridine with chlorocarbonyl-2-methoxyethane. The ru- theniurn(I1) complexes [RuL1(bipy)2][PF& and [RuL2(bipy- 121 [PF& were obtained by refluxing the appropriate ligand with [RuC12(bipy)2].2H20 in ethylene glycol, followed by purifica- tion on Sephadex LH-20 and precipitation on addition of ammonium hexafluorophosphate. Spectroscopy A Hewlett-Packard HP 8452A diode-array spectrophotometer was employed for recording electronic absorption spectra.A Spex Model Fluoromax fluorescence spectrophotometer was used for recording fluorescence emission spectra, the appro- priate excitation wavelength corresponding to the A,,, of the metal-to-ligand charge-transfer (MLCT) absorption band being determined from the electronic absorption spectra.Dihydrogen phosphate and chloride anion recognition was probed by addition of their desiccated tetra-n-butylammonium salts. All measurements were conducted at 22 k 2°C using a 1 X 1 cm rectangular quartz cuvette and deoxygenated acetonitrile solu- tions.NMR spectra were obtained on a Bruker AM300 instrument using the solvent deuterium signal as internal reference. Results and Discussion Extensive 1H NMR titration investigations with chloride and dihydrogen phosphate anions showed that the receptor [RuLl- (bipy)#+ formed solution complexes of 1 : 1 stoichiometry 500 550 600 650 700 Wavelengthhm Fig.1 Fluorescence emission spectra of 10-5 mol dm-3 [RuL1(bipy)2I2+ in acetonitrile with addition of tetra-n-butylammonium chloride: A, 0; B, 5 x lO--5; C, 10 x 10-5; D, 15 x and E, 20 X mol dm-’.366 Analytical Communications, October 1996, Vo133 with both anions. Proton shifts suggested that coordination takes place through a combination of hydrogen bonding interactions with the amide (CO-NH) and 3,3’-bipyridyl protons and electrostatic attraction with the positively charged ruthenium(I1) Lewis acid centre.By contrast, [RuL2(bipy)2]*+ showed no significant proton shifts on addition of chloride anions; however, formation of a 1 : 1 solution complex with dihydrogen phosphate anions was established.With the NMR evidence in mind, anion recognition was probed using fluorescence emission spectroscopy. The low energy MLCT band of ruthenium(I1) tris(bipyridy1) complexes5 and inherent sensitivity of the technique makes fluorescence emission spectroscopy particularly appropriate for the study of anion binding, analytical applications being the ultimate aim of the research.Unlike the MLCT fluorescence intensity enhancement ob- served (Fig. 1) on binding of chloride to the acyclic ligand in [RuL1(bipy)2]2+, binding of dihydrogen phosphate quenches the fluorescence signal (Fig. 2). The lack of changes in the spectrum of the prototype [Ru(bipy)#+ on addition of excess chloride and precipitate formation in the case of dihydrogen phosphate addition further confirmed anion binding with [RuL * (bipy)2I2+.In the uncomplexed form of [RuLl(bipy)2]2+ the receptor sites on the bipyridyl group of the ligand L1 are free to rotate, thus reducing the quantum yield of the complex in contrast to [R~(bipy)~]2+. Rotation is then restricted on binding of chloride in the ligand cavity, leading to retrieval of the fluorescence and enhancement of quantum yield.In the case of dihydrogen phosphate, we postulate that the structure of [RuL1(bipy)2]*+ is distorted on anion binding, with a decrease in quantum yield. Although binding of chloride and dihydrogen phosphate anions gives opposite effects on the fluorescence intensity, stability constants (log Ks = 4.7 for both anions) determined by a literature method6 show that there is no selective thermody- namic discrimination between the two anions.As predicted from NMR spectroscopy, no changes in the fluorescence emission spectrum of [ R ~ L ~ ( b i p y ) ~ ] ~ + took place on addition of chloride anions. In contrast, dihydrogen phos- phate addition showed a substantial quenching in the fluores- cence intensity (Fig. 3), with a stability constant, log Ks = 4.2. Molecular modelling of the complexes is helpful in inter- pretation of these preliminary results.The N-H groups in the receptor sites of ligand L1 in [RuLl(bipy)2]2+ both point inwards and are ideally suited for co-operative binding of spherical anions. Furthermore, with rotation of the receptor sites a tetrahedral molecule such as the dihydrogen phosphate anion can also be accommodated. For [ R ~ L ~ ( b i p y ) ~ ] ~ + , formation of a stable complex of the spherical chloride anion cannot occur as the two amide protons in L2 point outwards and cannot bind, whereas the tetrahedral dihydrogen phosphate anion can be coordinated by ‘perching’ above the plane of the ligand.I v) w (I) 2.00E+05 3 8 1 c. x v) .- 2 1.00E+05 - 0.00 E +00 500 550 600 650 700 Wavelengthhm Fig. 2 Fluorescence emission spectra of 10-5 mol dm-i [RuLl(bipy)#+ in acetonitrile with addition of tetra-n-butylammonium dihydrogen phos- phate: A, 0; B, 5 x mol dm-3.C, 10 X 10-5; D, 15 x lo-? and E, 20 X Conclusions NMR measurements and fluorescence emission spectroscopy have revealed that in contrast to [R~Ll(bipy)~]2+, [RuLZ- (bipy)J2+ exhibits exclusive selectivity to the binding of the dihydrogen phosphate anion in preference to chloride.Usually, phosphate determination is based on the molydenum blue method, which is both complicated and time consuming.7 The development of fluorimetric receptor-immobilized chemical sensors based on such selective host-guest complexation as described here will provide a rapid and simple alternative for phosphate determination.In real sample analyses, interference from other fluorescent species is likely to be negligible owing to the low energy (high wavelength) MLCT emission band of such complexes. In addition to their development as chemical sensors, the systems described here could see applications as fluorescent indicators for the study of phosphate transport in biological systems where chloride anions will also be present.The authors thank the SERC (Molecular Sensors Initiative) for a postdoctoral research fellowship (GR/H32896) to F.S. and the Trustees of the Analytical Chemistry Trust Fund of the Royal Society of Chemistry for the award of an SAC Research Studentship to J.S.W. 6.00E+05 A 0 References Wavelengthhm Fig. 3 Fluorescence emission spectra of mol dm-3 [RuL2(bipy)*I2+ in acetonitrile with addition of tetra-n-butylammonium dihydrogen phos- phate: A, 0; B, 5 x mol dm-3. C, 10 x D, 15 x 10-5; and E, 20 x Dietrich, B., Pure Appf. Chem., 1993,65, 1457, and references cited therein. Beer, P. D., Chem. Commun., 1996, 689. Beer, P. D., Mortimer, R. J., Stradiotto, N. R., Szemes, F., and Weightman, J. S., Anal. Proc., 1995, 419. Beer, P. D., and Szemes, F., J . Chem. Soc., Chem. Commun., 1995, 2245. Juris, A., Balzani, V., Barigelletti, F., Campagna, S., Belser, P., and von Zelewsky, A., Coord. Chem. Rev., 1988, 84, 85. Bourson, J., and Valeur, B., J. Phys. Chem., 1989, 93, 3871. Official and Standardized Methods of Analysis, ed. Watson, C. A., The Royal Society of Chemistry, Cambridge, 3rd. edn., 1994. Paper 6104970B Received July 16, 1996 Accepted August 16, I996

ISSN:1359-7337    

DOI:10.1039/AC9963300365

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Vol33 (367-369) 367 Determination of Carbohydrates by Flow Inject ion With Direct C hemil umi nescence Detection Irena B. Agatera, Roger A. Jewsbury** and Kath Williamsb a Department of Chemical and Biological Sciences, University of Huddersfeld, Huddersfield, West Yorkshire, UK HDI 3DH UK CM6 IHD Chemlab Instruments, Chelmsford Road, Great Dunmow, Essex, The oxidation of mono- and di-saccharides by permanganate in the presence of MnI1 leads to chemiluminescence.This can be used to determine individual sugars using flow injection giving a linear response over three orders of magnitude. Calibration data for the range 10-4 to 10-1 mol dm-3 are reported for monosaccharides glucose, galactose, fructose, arabinose and xylose, for disaccharides lactose and sucrose, for sugar alcohol mannitol, for glycerol and, for comparison, ascorbic acid.The determination of sugars is an important analysis, particul- arly in the food industry, and there has been an interest in chemiluminescent methods for rapid and simple analysis. Most reported methods have been indirect, involving enzyme conver- sion to generate peroxide then followed typically by reaction with luminol.' Whilst possibly less selective, a direct method based on oxidation should be simpler and more reliable.Mono- and di-saccharides are known to be oxidized by a range of oxidizing agents in a stepwise fashion, each step giving formic acid and a smaller sugar. The precise nature of a particular reaction depends on the oxidant and conditions.As part of a study of the chemiluminescence from the components of food and drink, chemiluminescence has been observed to accompany the oxidation of carbohydrates by permanganate and peroxide and the determination of individual sugars by chemiluminescence during permanganate oxidation is reported here. It is likely that these reactions could be of use in the determination of specific sugars during the manufacture of foodstuffs and of excipients in pharmaceutical formulations for which chemiluminescence has already been reported.* Experimental Flow experiments were carried out using a system constructed in the laboratory. Tygon tubing (1.02 mm id) was used for the flowing streams which were pumped using a four channel peristaltic pump (H0733 1, Ismatec, Carshalton, Surrey, UK). Injection was by sample loop injection valve (5020, Rheodyne, Cotati, CA, USA).Two detectors were used. One was based on a published design3 and comprised coiled tubing (600 mm long) adjacent to a photomultiplier tube with an S20 photocathode (98 16B, Thorn EMI, Ruislip, Middlesex, UK) operating at 1200 V with output directly to a chart recorder (BD40, Kipp and Zonen, Delft, Netherlands). The other was a commercial chemiluminescence photodiode detector (CL 1 , Chemlab, Great Dunmow, Essex).* To whom correspondence should be addressed. Batch experiments were performed using a liquid scintilla- tion counter (Tricarb 2002, Packard, Downers Grove, IL, USA) in photon counting mode calibrated with luminescence stan- dards (Biolink, Radlett, Hertfordshire, UK).Chemiluminescent spectra were obtained using a fluores- cence spectrometer (F-4500, Hitachi, Wokingham, Berkshire, UK) with the source turned off and at a scan speed of 500 nm s-l. Absorption spectra were obtained from a UV/vlS spectro- photometer (160A, Shimadzu Corporation, Kyoto, Japan). Sugars, mannitol and MnS04-H20 (Sigma, Poole, UK), KMn04, glycerol and ascorbic acid (Merck, Poole, UK) and HZS04 (Fisher Scientific, Loughborough, UK) were used as supplied without further purification.Results and Discussion Oxidation of Sugars Using glucose as a representative sugar, a number of oxidants were investigated for chemiluminescence using a batch reac- tion. Only peroxide and permanganate oxidations were ob- served to emit light, the light emission from peroxide being very rapid whilst that from permanganate was slower.In this paper only the results from permanganate oxidation are reported. The rates of oxidation and the emission of light were found to be dependent upon conditions such as concentrations of sugar (Fig. 1) and oxidant. It is known4 that permanganate oxidations are catalysed by Mn" and the addition of Mn" increased the rate of sugar oxidation and of the appearance of chemilumines- cence.8.0x'08 1 Time Is Fig. 1 Variation in chemiluminescence with glucose concentration. Solutions (5 cm3) contain glucose (concentration as shown in rnol dm-3), H2S04 (1 mol dm-3) and KMn04 (0.001 mol dm-3).368 Analytical Communications, October 1996, Vol33 - Selection of Parameters and Flow Arrangement The aim of this study was to establish an analysis suitable for on-line analysis of sugars with good precision and a well defined linear range.The oxidation was carried out under acid conditions established using 2.0 rnol dm-3 H2SO4. The slow reaction coupled with a dependence of the rate upon sugar concentration could limit the linearity in a flow injection system so Mnil was used to catalyse the oxidation by permanganate.Increasing the Mn" concentration also increased the response improving the S/N, but the Mn" concentration was limited by the precipitation of MnO? upon mixing with permanganate at higher Mn" concentrations. This is a serious problem as precipitation can cause deposits in the detection cell affecting subsequent readings. A concentration of Mn" of 0.2 rnol dm-3 was chosen as the highest concentration for which precipitation did not occur.The flow manifold was chosen to be suitable for on-line analysis of sugars in solution. A solution containing the sugar was mixed with an acid solution of Mn" and then the permanganate oxidant (60 pl) was injected into the flowing stream. Measurement was made within 2 s of mixing.The only unstable reagent used was potassium permanganate solution. Permanganate is thermodynamically unstable to the oxidation of water and will also react with oxidizable impuri- ties. Decomposition, although slow, is catalysed by light. The dependence of the response for glucose upon permanganate concentration showed a maximum at 0.025 rnol dm-3 and at this concentration a 5% change in concentration gave a 3% change in response.The optimum flow rate was found to be dependent upon the sugar and, for comparison, a value of 1.9 cm3 min- for each stream was chosen for the results presented here. Very similar results were obtained with both the photomul- tiplier and photodiode detectors. D c7 1 0 $ cr 3 (D Calibration Curves The calibration data for a number of sugars, polyhydric alcohols and ascorbic acid are given in Table 1. Good linearity was found over the concentration range to 10-1 rnol dm-3.Reproducibility of the 10-7 rnol dm-3 calibration point was reasonable with an RSD of 16% over a period of several weeks. Mechanistic Comments Although permanganate is a common oxidant, the mechanisms of its oxidations are not well established. Under acid conditions, providing that there is excess reducing agent, the final product is Mn" but this is formed after a more rapid reaction to Mn'1'.5 Clearly the mechanism must involve intermediate oxidation Table 1 Calibration data for some sugars and related compounds for the range to 10-' mol dm-? (ascorbic acid in range lo-" to lo-' rnol dm-3) S 1 ope/m V Compound Intercept/mV mol ~ I din I Glucose Galactose Fructose Arabinose Xylose Lactose Sucrose Mannitol Glycerol Ascorbic acid 0.6 ? 0.2 -0.5 f 0.4 1.3 f 0.6 0.6 f 0.3 0.6 f 0.3 0.15 2c 0.02 0.16 k 0.07 1.2f0.7 0.5 f 0.2 0.3 f 3 983 * 9 1787 2c 22 5062 k 29 1557 f 15 1562k 13 840f 1 222f 16 I461 i 18 195 f 5 4.3 x 105+8 x 0.9998 0.9996 0.9999 0.9998 0.9998 1 .0000 0.9894 0.9995 0.9984 107 0.9868 states and both Mn04'- and Mn0& have been postulated.h.7 These are both unstable with respect to disproportionation and will react very rapidly.The catalysis of the oxidations by Mn" is consistent with the involvement of intermediate oxidation states. For the oxidation of carbohydrates by permanganate, contradictory claims for the first step being both to Mn04'- and to Mn04-i- have been presented.g.9 By comparing the rate of disappearance of permanganate with the chemiluminescence (Fig.2), it has been possible to show that the chemiluminescent step follows the reduction of Mn04-. A similar result is observed in the absence of added Mn". There is no interaction between the absorption of permanganate (three peaks between 500 and 600 nm) and the chemiluminescent emission which is at 650 f 10 nm (Fig.3). The wavelength of the emission is independent of carbohydrate and of the presence or absence of Mn". The chemiluminescent step thus appears to involve an intermediate oxidation state of manganese and since the wavelength of the light emitted is independent of reductant and differs from that of peroxide oxidation.which is at 430 k 10 nm, the excited state appears to involve or be derived from a manganese species. Conclusions This work has demonstrated that permanganate oxidation can form the basis of a simple chemiluminescent analytical method for sugars. Further studies are being conducted into the mechanism of this reaction and future work will address the determination of mixtures of carbohydrates.6 Ox1 08 b L b 'J . b. l ' l ' l * l ~ 0 0 0 0 \-*-- -----a-2z==~*-.- " ~ ~ 0 50 100 150 200 250 Time Is Fig. 2 Change i n (d) abwrbdnce md ( b ) chemiluminewence trom a mlution containing glucose (0.05 rnol dm-'1. Mn" (0.04 mol din- ;). KMnOJ (1.25 X 1 0 rnol dm-') dnd H$04 (0.75 mol dm-j). r 1200 - 1000 - % C c .- v) 800 - c 600 - - 8 = - &! 400 - 200 r 0 I , I , I , I .I . I . L] 550 575 600 625 650 675 700 725 750 Wavelength Inm Fig. 3 Chemiluminewmce ymtrum within I \ of mtxing tor oxidation of glucow by permanganate in the prewnce of Mn" Gau\\ian fit \hewn with A,,, = 650 nni and M/ = 68 nmAriu!\.ticul Coniniunic.utioiis, October- 1996, Vol 33 369 The experimental assistance of E. Charpontier, IUT, Be- sanqon.France, the loan of luminescence standards by Biolink UK and the loan of a chemiluminescence detector by Chemlab UK are gratefully acknowledged. References 1 Navas. M. J.. and JiineneL. A. M.. Food Clrcnr.. 1996. 55. 7. 2 Lopez Par. J. L., and Townshend, A.. Aiitrl. Cotimul., 1996, 33, 31. 3 Burguera. J. L., Townshend, A,. and Greenfield, S., Airrrl. Cliinr. Ar.ttr, 19x0. 114, 209. 4 Vogel. A. I., A TelthooX- of’ Qiitriitit(itiw I i i o r p r i r i c ~ Air~iI~.sis, Longmans. London. 1943. p. 335. 5 6 7 X 9 Powell, R. T., Oskin. T.. and Ganapathisubramanian, N., J. Phys. Chcnr., 1989. 93. 27 18. Zahonyi-Budo. E.. and Simandi, L. I., l m r g . Chin?. Acta, 1995, 237, 173. Lee. D. G.. Moylan, C. R., Hayashi, T.. and Brauman, J. I., J. Am. Clirnr. Soc.. 1987, 109, 3003. Lindroos-Heinanen. R.. and Virtanen, P. 0. I., Finn. Chrm. Lett., 1988. 15, 117. Gupta. K. K. S., Sanyal, A., Tribedi. P. S., and Gupta, S. S., .I. Chrm. Kc>.s.. S\’nop.. 1993, 484. Paper- 6105062J Receiljed July 22, 1996 Accepted August 22, I996

ISSN:1359-7337    

DOI:10.1039/AC9963300367

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996. Vol33 (371-373) 37 1 Analysis of Dissolved Gases by Headspace Sampling Gas Chromatography With Column and Detector Switching. Preliminary Results Pierre-Marie Sarradin and Jean-Claude Caprais IFREMER Centre de Brest, DROIEP, BP 70,29280 Plouzane, France This preliminary study presents a new method for the determination of dissolved gases in sea-water. After headspace extraction, gaseous compounds are separated by GC on a dual column adsorption system.The individual species are thereafter eluted to three different detectors (thermal conductivity, flame ionization and FPD) using a second switching valve. The analysis is quantitative for methane and carbon dioxide with limits of determination of 0.1 and 50 pmol I - l , respectively, and corresponding standard deviations of 4 and 7%, and has been applied to deep sea hydrothermal samples.Further optimization is necessary to allow the quantitative determination of hydrogen sulfide, nitrogen, oxygen and carbon monoxide. Deep sea hydrothermal vents are characterized by the presence of hot fluid sources with unusual chemical compositions, i.e., high temperature, low pH, high sulfide and methane concentra- tions.Particular ecosystems are associated with these vents. In the absence of light, these are based on chemosynthesis and particularly on the hydrogen sulfide and methane present in the fluid. The study of the chemical environment of hydrothermal organisms is based on biologically important compounds present in the mixing zone of the hydrothermal hot fluid and the Table 1 Concentrations of the dissolved gases in sea-water and hydro- thermal fluids Gas Hydrothermal fluid Sea-water CH4 0.001-3.4 mmol 1-1 0.4 nmol 1-1 H2S 0.5-18 mmol 1-I trace C02 2.3-285 mmol 1-1 2.4 mmol 1-1 cold sea-water.Present in the fluid are the dissolved gases CH4, H2S and C02 (Table 1). In this study, sampling was carried out using a sampler deployed from the French deep-sea submersible research vessel Nautile.Samples were difficult to obtain and had to be handled with care as other analyses were carried out on the 150 ml of water. The objective of this study was to develop a method for the determination of dissolved gases in a minimum volume of water, with a single shot and a good working range. The first part of this work focused on C%, CO2 and H2S.Table 2 presents the analytical methods generally used for the analysis of each compound. The only method existing for the simultaneous determination of CH4, C02 and H2S is the method of Childress et al.9 for the determination of high concentrations of these gases in blood. This procedure can be divided into three steps: extraction of the gases from water, separation and detection.The purge and trap method seems to be the most efficient method for the extraction of CH4, C02 and H$, because of the enrichment step. However, several unsuccessful attempts were made by us to simultaneously cryotrap CH4, COZ and H2S. The experiments were carried out using packed glass columns with different packing materials, the cryogen being liquid nitrogen.Childress et al.9 used an acid stripping method (dynamic headspace without trapping) in a purpose-built reactor; this method is only available for concentrated samples (blood) and small injection volumes. An alternative method is headspace sampling which has a good working range for CH4 and CO2. This method was tested for the complete set of gases.The complete chromatographic separation of CH4, COT and H2S is not possible on a single column. A dual column system with a switching valve must be used.9 The first column used in the study of Childress was a Poraplot Q: this separated 0 2 , N2, CO and CH4 as a single peak from COZ, H2S and water. The first peak ( 0 2 , N2, CO, CH4) was trapped on the second column Table 2 Analytical methods used in the literature Reference Compounds Matrix Method Detection Calibration Working range Sea-water Sea-water Water Sea-water Sea-w ater Sea-water Sea-water Sea-water Blood Headspace Headspace Headspace Modified MHE: Purge and trap Purge and trap Purge and trap Ultrasonic Dynamic heads pace FID* Gas standard TCDt.FID" TCD' PIDS FID" Gas standard TCDt, FID" FID" Gas standard TCDt Gas standard 3-700 ymol l-1 5-20 nmol 1-' 0.01-100 pmol 1-1 CH4 1-100 pmol l-1 DL 13 nmol l--lfl 0.02-18 nmoll-I 0.0545 nmol 1-' C02 4-20 mmol l-1 H2S 0-5 mmol 1-1 * FID, flame ionization detector.TCD, thermal conductivity detector. * MHE, multi-headspace extraction. PID, photoionization detector. 7 DL, detection limit.372 Analytical Communications, October 1996, Vol33 1.6E+05'1 1.4E+05.1.2E+05' 1 .OE+05- 8.OE+04. 6 .OE+ 04. 4.OE+04' (molecular sieve), the valve was switched and C02, H2S and water were eluted directly to the detectors. When the elution was finished on the Poraplot column, the final separation of 02, N2, CO, CH4 was carried out on the molecular sieve after a second valve switching. Childress et al.9 used a thermal conductivity detector (TCD) for the detection of the gases.To improve sensitivity more specific detectors were used in this study: a TCD for C02, N2, CO, 0 2 , a flame ionization detector (FID) for CH4 and an FPD for H2S. Species were eluted to the different detectors using a second switching valve. Experimental A schematic diagram of the complete analytical device used is presented in Fig.1. A preliminary study was carried out using 20 ml headspace flasks, with a 10 ml sample of water and 0.1 ml of 6 mol 1 - 1 hydrochloric acid under He. The flask was heated for 30 min at 70 "C under stirring. The injected headspace volume was 2.5 rnl. To reduce the analysis time, a temperature programme was used [40 "C (held for 7 min); raised at 15 "C min-I to 130 "C (held for 13 min)]. The headspace sampler used was a DAN1 HSS 86-50 with a Silicostil heated transfer line.The separation was performed using a series 8000 Fisons gas chromatograph equipped with a TCD, an FID EL 980 and an FPD and LIN 700 (linearizer). The switching valves were made of Hastelloy (Valco). The columns and pressure drop were Chrompack Poraplot Q Ultimetal 0.53 mm, 25 m and Molecular sieve 5A Plot Ultimetal, fused silica 0.32 mm, 5 m.The carrier gas was He N55 (0.49 bar). The integration software was Borwin 4 channels. Calibration was achieved either with a standard gas mixture from Scotty with Hamilton gas-tight syringes 1700 RN series or with standard solutions of sodium sulfide nonahydrate (Rectapur, Prolabo) and sodium hydrogencarbonate (pro analysi, Merck).Results and Discussion Fig. 2 presents a chromatogram obtained with this method for concentrations of 35 ymol 1- l of H2S, 200 pmol 1- of C02 and Carrier 0.49 bar Aur 0.64 bar He i___, 120 pmol l-1 of CH4. The valves were switched at 315 s (El), 440 s (E2) and 14 min (El, E2). The chromatographic resolution enabled quantitative analysis. However, the baseline for the TCD was disrupted by the temperature programme.The use of an isothermal temperature program or a mass flow controller would cancel this problem. Fig. 3 presents the calibration curves for the three com- pounds, obtained by the liquid addition of Na2S and NaHC03 and gaseous addition of CH4 to 10 ml of sea-water. The low sensitivity for H2S may be due to the adsorption of sulfide on part of the gas chromatograph, although most of the inox parts had been replaced except for the syringe needle of the headspace.FID 1 I 1 1 I\ 2.OE+&] h I[ O.OE+OO 5.00 10.00 15.00 20.00 25.00 30.00 3 HWD 4.OE+04] .W O . o E + O O L T O 10.00 15.00 20.00 25.00 30.00 1 I .2E+05 1 .OE+051 6.OE+04 4.OE+04 8'0E+041 FPD Ti me/min Chromatogram obtained for a mixture of Na2S (35 pmol I-'), C02 Fig.2 (200 pmol 1-1) and CH4 (120 pmol 1-1). I I ..Analytical Conzmunications, Octohei- 1996, Vol 33 373 The determination limits (3 X noise integration/calibration slope) were 0.1 and 50 pmol l-1 for CH4 and C02, respectively. Standard deviations ( n = 5 ) were 4% for 80 pmol 1-1 of CH4 and 7% for 800 pmol 1-1 of C02. Application The analytical device described here has been used aboard the French oceanographic vessel 'Nadir' during the HOT 96 diving 350 y = 1 .9 7 ~ -3.3 cruise undertaken by IFREMER CNRS Marine Research Unit no. 7 on the 9" and 13" N hydrothermal sites of the East Pacific Rise. Fifty water samples were collected of the various organisms present (Riftia, Alvinellids and mussels) by the French submersible Nautile using a specific vacuum-based and gas-tight sampler.Hydrogen sulfide was not quantified because of its low sensitivity. For methane and carbon dioxide concentrations of 1-5 ymol 1-1 (sea-water 0 pmol I - I ) and 900-8000 pmol 1-I (sea-water 2280 ymol I - ] ) were deter- mined, respectively. 0 50 100 150 200 [CH,]/p mol I -' 350 T * / / 0 - 8 I I 0 10 20 30 40 50 60 [Na,S]/p mol I-' 50 T I 25 30 i--.-'- 30 20 -I y = O.O129X+ 26.5 R2= 0.991 6 O J I 0 5 00 1000 1500 2000 [NaHCO,]/p mol I-' Fig.3 Calibration curves for CH4, H2S and CO?. Conclusions The headspace extraction conditions (temperature, time, pres- sure) must be further optimized to improve the efficiency and to reduce the amount of water injected. The hydrogen sulfide line must be checked to eliminate the possible remaining inox parts and to study the working conditions for the FPD (air and H2 amount in the flame) to enhance the sensitivity.The concomi- tant use of a TCD and a temperature program led to a fluctuating baseline use of a flow controller may overcome this problem. This preliminary study has focused on the analysis of CH4, H2S and COZ. Future work will involve the quantitative determination of 02, N2 and CO using the same method.This work was supported by a MAST I11 AMORES contract 950040. References Blanc, G., Boulkgue, J., and Gieskes, J. M., Occanol. Acru, 1991, 14, 33. Bange, H. W., Bartell, U. H., Rapsomanikis, S.. and Andrae, M. 0.. Global Biogeockem. Cycles, 1994, 8, 465. Hamilton, S. K., Sippel, S. J., and Melack. J. M.. Biogeoc.henri.sti.~. 1995, 30, 115. Evans, W. C.. White. L. D., and Rapp, J. B.. J . Grophys. Rrs.. 1988, 93, 15.305. Cutter, G. A., and Oatts, T. J., A n d . Chrni., 1987, 59, 7 17. Charlou, J. L., and Donval, J. P., .I. GwphFs. Res., 1993. 98. 9625. Ishibashi, J. I.. Wakita, H., Nqjiri. Y.. Grimaud. D., Jean-Baptiste, P., Gamo, T., Auzende, J. M., and Urabe. T., Earth Planrt. Sci. Lett.. 1994, 128, 183. Schmitt, M.. Faber. E.. Botz, R., and Stoffera, P., A n d . Chrni., 1991, 63, 529. Childress, J . J., Lee, R. W., Saunders, N. K., Felbeck. H.. Oros. D. R.. Toulmond, A., Desbruykres. D., Kennicutt, M. C., and Brooks, J.. Natiur, 1993, 362, 147. 1 2 3 4 5 6 7 8 9 Pnpei' 6105335A Receit3ed Jidy 30, I996 Accepted August 19, I996

ISSN:1359-7337    

DOI:10.1039/AC9963300371

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Vol33 (375-379) 375 Identification of Hydrogen Peroxide as the Autoxidation Product of Akphenyl-2-propyl-3,5-diethyl-ly2=di hydropyridine Declan P. Rafterya, Malcolm R. Smytha,* Raymond G. Leonardb, Brendan J. Kneafseyb, Martin C. Brennanb a School of Chemical Sciences, Dublin City University, Dublin 9, Ireland Loctite Research and Development, Tallaght Business Park, Dublin 24, Ireland Recently, dihydropyridines have been used as autoxidizable materials in novel air-activated adhesives.The dihydropyridines react with air to produce peroxide species, which are capable of initiating free radical polymerization of methacrylates. It is proposed that N-phenyl-2-propyl-3,5-diet hyl-1,2-dihydropyridine undergoes autoxidation in the presence of glacial acetic acid, yielding a pyridinium ion and an unknown peroxide product.Employing a variety of analytical techniques, including polarography, spectrophotometry and an enzyme-based biosensor, it has been shown conclusively that hydrogen peroxide is generated in the autoxidation of this dihydropyridine. An aqueous extraction procedure was utilized to remove the hydrogen peroxide from an organic matrix and also to eliminate any interference from the dihydropyridine in the analysis procedure.Autoxidation normally refers to reactions of substances with molecular oxygen under ambient conditions without the intervention of a flame.l The initial products of autoxidations usually contain an intact singlet oxygen-oxygen bond, i.e., the primary product is a peroxide. Many types of molecules undergo autoxidation, the rate of autoxidation varying widely depending on the structure.The process of autoxidation has been studied extensively over the last century and many excellent reviews published reflecting the universal importance of the subject.2-4 The mechanism of autoxidation is now generally accepted to occur through free radical intermediates; thus, many autoxidations are autocatalytic in nature, since the reaction products, i.e., peroxides, are potential sources of free radicals.Autoxidation impacts on many aspects of our lives, for example from autoxidative degradation of rubbers and plastics, ageing of food and tissue, to air drying of paints and coatings. Biological oxidation-reduction reactions and metabolism of oxygen have also attracted much attention in recent decades.Central to this subject is the study of the coenzymes, in particular the coenzyme NADH which has a 1,2 dihydronico- tinamide as its central reactive group. Many researchers have studied dihydropyridine derivatives and their chemistry as synthetic models for NADH. The known chemistry of dihy- dropyridines is now extensive and a number of excellent reviews are a~ailable.57~ Dihydropyridines have found many practical applications in areas from pharmaceuticals to the rubber and polymer industries. Early dihydropyridine chemistry is dominated by the distinc- tion between the 1,2- and 1,4-isomers, a distinction which is now most conveniently made by modern physical methods.7 The predominant reaction of dihydropyridines is oxidation to * To whom correspondence should be addressed.the more stable corresponding pyridines which occurs by loss of a hydride ion from the dihydropyridine or by transfer of two electrons and one proton. Convincing evidence exists for both mechanisms, but the latter mechanism receives the broadest acceptance.7 The oxidative and thermal instability of dihy- dropyridines is well k n o w ~ p thus, most studies have been conducted on N-substituted 1,4-dihydro derivatives with elec- tron withdrawing substituents on the ring.The less stable 1,2-derivatives have received much less attention and are only stable when steric hindrance is evident. The best known example, N-phenyl-3,5-diethy1-1,2-dihydr0pyridine,~ first pre- pared by Craig et a1.,9.10 and mistakenly identified as a 1,4-isomer, has found general use as a reducing agent within the polymer and rubber industries, where it is the active component within the crude condensation product of aniline and butyr- aldehyde.Both the crude condensate and a purified distillate have been patented for use within the reactive adhesives.11 More recently, the use of dihydropyridines under acidic conditions as autoxidizable materials within air-activated adhesive compositions has been described. l-2 The dihydropyri- dines under acidic conditions react with air to produce peroxide species, which are capable of initiating free radical polymeriza- tions of methacrylates.A mechanism for the autoxidation has been proposed following that presented by Cilento13 for the autoxidation of benzyl 1,4-dihydronicotinamide (Fig.I), which proposes the generation of one mole of hydrogen peroxide and a pyridinium salt of the acid per mole of dihydropyridine oxidized.14 It has been proposed that the hydrogen peroxide then reacts with the dihydropyridine, generating free radicals (Fig. 2). Several reports describe the air-activated oxidation of dihydropyridines; however, few researchers have concentrated on the oxidation products produced alongside the pyridines.Fig. 1 Proposed DHP autoxidation mechanism with the hydrogen peroxide, in the presence of glacial acetic acid. + H202 generation of + 2H20 c 2 H 5 ~ 1 ~ ~ ~ + H202 CH3COOH I Fig. 2 Proposed secondary reaction between DHP and hydrogen per- oxide.376 Analytical Comniunications, October- 1996, Vol33 This paper reports on detailed studies of the characterization of hydrogen peroxide as the primary peroxidic oxidation product from the autoxidation of N-phenyl-3,s-diethyl- 1 ,2-di- hydropyridine (DHP) under acidic conditions.Experimental Reagents All reagents were of AnalaR grade unless otherwise stated.The DHP used in all experiments was commercial grade which was supplied by Loctite (Irl.) Ltd. (Tallaght, Ireland), and was purified by fractional distillation. All reagent solutions were prepared in de-ionized water, which was obtained by passing distilled water through a Milli-Q water purification system (Millipore, Milford, MA). The supporting electrolyte employed for polarographic studies consisted of 4.0 g saturated lithium chloride solution and was obtained from Metrohm (Herisau, Switzerland).The titanium porphyrin complex was synthesized according to the method of Inamo et al.15 and was supplied by Tokyo Kasei Industries (Tokyo, Japan). The titanium porphrin reagent (5 X 10-5 rnol dm-l) was prepared by dissolving 34 mg of the complex in 1 dm3 of 0.05 rnol dm-3 hydrochloric acid. A perchloric acid solution (4.80 mol dm-3) was prepared by diluting 68.60 cmi of 70% m/v perchloric acid solution to 100 cm7 in water.The tin(iv) oxide glass plates, 1/2 inch square and 0.38 inch in thickness, were obtained from Nippon Sheet Glass (Nippon Japan). The tin(rv) oxide layer was 600-700 nm in thickness. The horseradish peroxidase enzyme (EC 1.1 1.1.7.type IV) was supplied by Sigma (St. Louis, MO, USA). The electron mediator ferrocenemonocarboxylic acid, (3-aminopro- py1)triethoxysilane and glutaraldehyde (25% m/v solution) were all obtained from Aldrich (Poole, Dorset, UK). Procedures Extraction procedui-e A simple extraction procedure was employed which consisted of dissolving 1.00 g of the dihydropyridine in 36 cm3 of chloroform and 4 cm-l of glacial acetic acid in a 100 cm3 beaker.After exposure of this solution to air for a fixed period of time, typically 3 min, an aqueous extraction was carried out by mixing 1 cm3 of de-ionized water with 9 cm3 of the DHP- containing solution. It was expected that hydrogen peroxide, being ionic in nature, would be extracted into the aqueous layer, with the dihydropyridine remaining in the chloroform layer.Furthermore, to ensure the removal of trace organics, the aqueous extract samples were passed through a Sep-Pak C cartridge. This aqueous extraction procedure was carried out on all autoxidation samples prior to analysis, as it removed all traces of the dihydropyridine and placed the peroxide in a less hostile matrix for analysis.lodimeti-ic titration A 1 cm3 volume of the aqueous extract sample was first pipetted into SO cm3 of chloroform-glacial acid (1 + 1 vfv) solution. To this was added 2 cm3 of saturated sodium iodide solution and three pea size lumps of dry ice. The solution was covered then with parafilm and stored in the dark for 1.5 min. The dry-ice is required to exclude oxygen, which interferes with the analysis procedure, giving rise to an increase in the apparent peroxide level.After 15 min, 25 cm3 of de-ionized water was added to the solution, which was then titrated with 0.1 mol dm-3 sodium thiosulphate, with the end-point detected potentiometrically. A blank sample was prepared in an identical manner, using de- ionized water instead of the aqueous extract sample.As a means to reinforce the hypothesis that the peroxide, which was to be detected using the iodine titration method, was being generated Lia autoxidation, an aqueous extract sample was also prepared in an argon atmosphere. This involved the use of a glove box which was previously purged with argon for 30 min. It should be emphasized that this did not ensure the complete exclusion of oxygen.Also it should be noted that in this instance all dihydropyridine/glacial acetic acid samples were exposed to air (or argon where applicable) for 3 rnin before carrying out the extraction procedure. A blank prepared under the same conditions, but omitting the dihydropyridine, was also titrated and the result subtracted from the analysis samples. Polal-ogl-clphy Both aqueous extract samples exposed to air or argon for 3 min were analysed using polarography.A sample volume of 0.10 cm-7 of the extract was added to 10 cm3 of supporting electrolyte, with purging of the solution for three minutes with nitrogen prior to analysis. As a means of confirming the identification, the samples were subsequently spiked with 0.05 cm3 of a standard solution of 0.1 mol dm--7 hydrogen peroxide.Also, the reduction potential of the extract sample was compared with that of a hydrogen peroxide standard. SI,ec.trophotonieti-ic. analysis based on formation of a titanium poiplyin c-omplex Utilizing the aqueous extraction procedure detailed previously, the quantity of hydrogen peroxide generated in the autoxidation of DHP, in the presence of glacial acetic acid over time, was determined spectrophotometrically using the titanium por- phyrin reagent.A solution containing 1.00 g of DHP in 36 cm3 of chloroform and 4 cm3 of glacial acetic acid was left exposed to air. and at various times over the period of 60 min an aliquot was removed and the hydrogen peroxide isolated by means of the aqueous extraction procedure.A 250 pl portion of the aqueous extract sample, 250 pl of 4.8 rnol dm-3 perchloric and 250 pl of titanium porphyrin reagent were then mixed and allowed to stand for 5 min at room temperature. After diluting the sample to 2.50 cm3, the absorbance at 432 nm was measured (A,). A blank sample was prepared in an identical manner, using de-ionized water instead of the sample and its absorbance (A,,) measured at 432 nm. The concentration of hydrogen peroxide was determined by the difference in absorbance: AA4.12 = A h - A, Detei*niination qf hjdi.ogen peroxide itsing a tin o.x-ide biosensor.A tin(iv) oxide coated glass plate was treated successively with 10% aqueous solution of (3-aminopropyl)triethoxysilane (APTES) for 1 h at 50 O C , then with a 2.5% aqueous glutaraldehyde solution for 30 min at room temperature and finally with a citrate-phosphate buffered solution, pH 5.9, of horseradiTh peroxidase (HRP) for 30 rnin at room temperature, to obtain a HRP-modified electrode.The sensor signal was obtained as a cathodic current in 0.1 mol dm-3 citrate buffer at 30 "C containing the mediator (0.2 mmol dm-7 ferrocene- monocarbox yl ic acid).Apparatus The iodine titration experiments were carried out using a Metrohm Dosimat 665 Auto-titrator connected to a Metrohm 686 titroprocessor. A Pt and Ag/AgCl double junction electrode were employed as working and reference electrodes, re- spectively.Analytical Comniunicutions, Octohei, 1996, Vol 33 377 All polarographic determinations were carried out using an EG&G PARC (Princeton Applied Research, Princeton, NJ) Model 303 SMDE connected to a Model 384 Polarographic analyser.Ag/AgCI in saturated lithium chloride in ethanol was employed as reference electrode, with Pt and a dropping mercury electrode employed as counter and working electrodes, respectively. The spectrophotometric measurements were carried out using a Simadzu- 160A spectrophotometer using quartz cells of 1 cm pathlength.The biosensor signal was measured using an EG&G PARC 264A polarographic analyser with Pt and Ag/AgCl employed as counter and reference electrodes, respectively. The applied potential was + 150 mV \w*sus Ag/AgCl. Electrical contact with the tin(rv) oxide layer on the glass was achieved by attaching wire to the surface using silver-doped conducting epoxy resin, which was then coated in epoxy resin to maintain electrical insulation in solution.Results A first essential requirement for the investigation was the development of a method that eliminated the interference from the dihydropyridine itself, which is a strong reducing agent. Because of the organic nature of the dihydropyridine, an aqueous extraction procedure was employed.The nearly quantitative removal of the dihydropyridine in the aqueous extraction procedure was confirmed by TLC. A positive reaction for the presence of peroxide in the aqueous extract sample was obtained on TLC plates using N,N-dimethyl-p- phenylenediamine dihydrochloride, a specific peroxide spray reagent.lh The hydrogen peroxide gave rise to a typical purple response against a blue background.Iodometric Titration Quantitative determination of peroxides is the most commonly used analytical method for following the course of autoxidation reactions. I Many methods have been developed for determin- ing peroxides, but iodometric titration is the most popular.’ 1-23 Other more sensitive, less time-consuming methods, such as GC and HPLC, are not employed to the same extent due to the sensitivity of peroxides to heat and surfaces. The iodometric method is based on the assumption that sodium iodide (or any iodine salt) and glacial acetic acid when contacted with peroxide liberate iodine quantitatively, and can be then titrated against standardized sodium thiosulfate.Average titration volumes of 11.00 cm3 of 0.01 mol dm-3 sodium thiosulfate were required to neutralize the iodine liberated in the sample exposed to air ( n = 6), while in 9 \ comparison only 3.20 cm3 was required for the sample in the argon atmosphere (n = 3).Therefore, significant iodometric titration results were observed for aqueous extract samples exposed to air, indicating the presence of a peroxide in the samples, while a 70% decrease in titration values was observed for the samples which were exposed to the argon atmosphere. Taking into account that the complete exclusion of oxygen was not possible, and the extremely high affinity of the dihydropyr- idine for oxygen, the complete elimination of the peroxide in the argon sample was not expected.A blank sample which consisted of the aqueous extraction procedure in the absence of the dihydropyridine, resulted in no detectable amounts of iodine being liberated. Po larograp h y Polarography has been used frequently for the determination of hydrogen peroxide and has lower limits of detection in comparison with iodine-based titration technique^.'^-^^ Polar- ography also has the advantage of greater selectivity, being able to distinguish between hydrogen peroxide and organic per- oxides, which are reduced at more positive potentials compared with hydrogen peroxide.24.2h On analysis of an aqueous extract sample by differential- pulse polarography, one large broad peak, characteristic of peroxides, was observed with a reduction potential of - 1.16 V ~ ’ ~ ‘ I - s M s Ag/AgCl and a peak current of 4.74 X lo3 nA.On spiking the sample with 0.05 cm3 of 0.1 mol dm-3 hydrogen peroxide, the peak current increased to 7.28 X 103 nA. In the case of the aqueous extract sample, which was exposed to the argon atmosphere for comparison purposes, the observed peak was 1.18 X 103 nA, which is equivalent to a 75% decrease in response compared to the samples exposed to air.The standard reduction potential for hydrogen peroxide as determined from standards was found to be -1.18 V \’ersus Ag/AgCl, which matches closely the reduction potential for the peak observed in the ex tract samples.Spectrophotometric Analysis Using the Titanium Porphyrin Complex A wide variety of spectrophotometric2x--3” and chemilumines- cence1s.-;4-37 based techniques are available for the determina- tion of hydrogen peroxide.However, many of these techniques are affected considerably by interferences, especially from reducing agents. In this instance, the use of a specific reagent for the determination of hydrogen peroxide which was not subject to interferences from reducing agents, would be an advantage. The method chosen involved using the oxo[5,10,15,20-tetra(4- pyridyl)porphyrinato]titanium(rv) complex.3x Q \ Maximum absorbance at 432 nm Maximum absorbance at 450 nm Fig.3 Titanium porphyrin spectrophotometric reagent for the determination of hydrogen peroxide.378 Analytical Communications, October 1996, Vol33 This water soluble complex reacts specifically with hydrogen peroxide, with the quantity of hydrogen peroxide determined spectrophotometrically by the decrease in absorbance of the parent complex at 432 nm (Fig.3). This spectrophotometric reagent is very sensitive, capable of detecting hydrogen peroxide in the range 1 X 10-5 to 1 X 10-7 mol dm-3. Perchloric acid was added to all samples as hydrogen ions assist the complexation reaction. Quantitative results were determined using a standard curve which was constructed for hydrogen peroxide concentrations in the range 1 X mol dm-3 with r > 0.998.Fig. 4 shows the results obtained using the specific titanium porphyrin reagent. These results not only confirm the presence of hydrogen peroxide in the aqueous samples, but also demonstrate, as would be expected, that the concentration of hydrogen peroxide increases with the exposure time to air. In the absence of glacial acetic acid in the autoxidation mixture, no hydrogen peroxide was detectable.Also shown in Fig. 4 is the sharp decrease in the quantity of hydrogen peroxide generated when the autoxidation mixture was purged continuously with argon to reduce its exposure to oxygen. to 1 X Analysis of Peroxide Using a Tin(rv) Oxide Biosensor A biosensor may be defined as a device that recognises an analyte in an appropriate sample and interprets its concentration as an electrical signal via a suitable combination of a biological recognition system and an electrical transducer.39 The bio- logical recognition system, typically an enzyme, multienzyme system, antibody, or whole slices of mammalian or plant tissue, is responsible for the specific and sensitive recognition of the analyte and subsequent response resulting in a change in one or more physiochemical parameters associated with the inter- action.If generated in close proximity to a suitable transducer, it may be converted into an electrical signal. An enzyme-based amperometric biosensor for hydrogen peroxide was con- structed, based on work by Tatsuma, et aL40 It employs the horseradish peroxidase enzyme covalently attached to a tin(rv) oxide coated glass plate electrode.The system is based on enzymic reduction of hydrogen peroxide by horseradish peroxidase and subsequent electron transfer from the tin@) oxide electrode to the enzyme via an electron mediator. The electron mediator is required to enhance the electron transfer between the peroxidase and the electrode. Although hydrogen peroxide can be oxidized directly on a tin(1v) oxide electrode it requires a high overpotential of i-900 mV versus Ag/AgCl.Therefore any electroactive substances present as impurities in the sample are liable to interfere in the analysis. The coupling of horseradish peroxidase with a mediator allows operation of the sensor at considerably lower potentials, reducing the risk of interference from contamination in the sample.A small response of 83 nA was obtained from the blank sample, due to the presence of glacial acetic acid. With injection volumes of 0.10 cm3, the responses obtained for aqueous extract samples taken at 120 and 160 min were 258 and 240 nA, P Air Fig. 4 Quantitative analysis over time for hydrogen peroxide using the spectrophotometric titanium porphyrin reagent, showing a comparison of a sample exposed to air and a sample exposed to an argon atmosphere.respectively. These responses correspond to hydrogen peroxide concentrations of 2.40 and 2.10 x mol dm-3, re- spectively, as determined by standard additions. These concen- trations of hydrogen peroxide correlate well with the results obtained using the titanium porphyrin spectrophotometric reagent.Discussion The main aim of this work was to confirm that hydrogen peroxide was being generated via autoxidation of DHP. This was further complicated by the presence of the hydrogen peroxide in an organic matrix with a strong reducing agent, namely the dihydropyridine. The employment of an aqueous extraction procedure removed the hydrogen peroxide from the organic environment while also eliminating the possibility of any interference from the dihydropyridine.Experiments carried out using the titanium porphyrin spectrophotometric reagent, to determine the efficiency of the extraction procedure, indicated an extraction efficiency in excess of 95%. The possibility that a hydroperoxide, based on the pyridine ring, was generated in the autoxidation was discounted by the aqueous extraction proce- dure, as a hydroperoxide would remain in the organic layer and in contrast to the results obtained, no trace of peroxide would be detectable in the aqueous extract.The iodine-based titrations, however, confirmed the presence of a peroxide in the aqueous extraction samples.A 70% decrease in peroxide concentration was observed when the autoxidation mixture was exposed to an argon atmosphere, even though complete exclusion of oxygen was not possible. The iodine titration, however, lacks the sensitivity of other instrumental techniques, with factors such as dissolved oxygen and moisture having marked effects on the results obtained. In both instances the results correlated very well with the results obtained using polarography in the lithium chloride-acetonitrile electrolyte.The reduction potential for the peak from the aqueous extract samples matches closely the reduction potential for hydrogen peroxide standards. The titanium porphyrin reagent offered a specific spectrophotometric-based technique for the determina- tion of hydrogen peroxide in an aqueous environment.Results obtained using this technique confirmed the presence of hydrogen peroxide in the aqueous extract samples and also demonstrated that the steady-state concentration of hydrogen peroxide increased on prolonged exposure to air. All the various techniques employed yielded positive results to the presence of peroxide, while the greater selectivity of the polarographic, biosensor and spectrophotometric methods gives conclusive confirmation that it was in fact hydrogen peroxide.The approach taken with this work was mainly qualitative in nature, to show conclusively that hydrogen peroxide was generated in the autoxidation of the dihydropyridine. The kinetics of autoxidation reactions are very complex, with a whole series of reactions occurring simultaneously, which can be further complicated by the presence of even small quantities of transition metals.No attempt was made to investigate the kinetics of the reaction of this compound with oxygen or the factors that augment this reaction. Rather, it was envisaged that the conclusive identification of the autoxidative peroxide product would lead to greater understanding of the chemistry in areas where this compound is in use.References 1 Davies, A. G., Organic Peroxide, Butterworths, London, 1961. 2 Autoxidation and Antioxidants, ed. Lundberg, W., Interscience Publishers, 1961, vol. I. 3 Russell, G. A., J. Chem. Educ., 1959, 36, 11 1. 4 Ingold, K. U., Chem. Rev., 1961,61, 563. 5 Kuthan, J., Ind. Eng.Chern. Prod. Res. Dev., 1982, 21. 6 Eisner, U., and Kuthan, J., Chem. Rev., 1972, 72.7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Pyridine and its Derivatives, ed. Abramovitch, R. A., Wiley, New York. Krow, G., Michener, E., and Ramey, K. C., Tetrahedron Lett., 1971, 3653. Craig, D., Schaefgen, L., and Tyler, W. P., J . Am. Chem. Soc., 1948, 70, 1624. Craig, D., Kuder, A.K., and Efroymson, J., J . Am. Chem. Soc., 1950, 72, 5236. Grant, S. M., Melody, D. P., and Martin, F. R., British Patent 2,087,906A, 1982. U.S. Patent no. 5,523,347, 1996. Cilento, G., J . Chem. SOC., Chem. Commun., 1968, 1420. Kneafsey, B. J., Leonard, R. G., and Brennan, M. C., unpublished results. Inamo, M., Funahashi, S., and Tanaka, M., Inorg. Chem., 1983, 22, 3734.Knappe, E., and Petri, D., 2. Anal. Chem., 1962, 190, 386. Heaton, F. W., and Uri, N., J . Sci. Food Agric., 1958, 9, 781. Lea, C. H., Proc. R. SOC., 1931, HOB, 175. Lea, C. H., J . Soc. Chem. Ind., 1945, 64, 106. Lea, C. H., J . SOC. Chem. Ind., 1946, 65, 286. Wheeler, D. H., Oil Soap, 1932, 9, 89. Barnard, D., and Hargrave, K. R., Anal. Chim. Acta., 1951, 5,476. Ricciuti, C., Coleman, J. E., and Willits, C. O.,Anal. Chem., 1955,27, 405. Moore, P. W., in Polarographic Determination of Hydrogen Peroxide in Irradiated Water, and a Comparison with Colorimetric Methods, Australian Atomic Energy Commission, Sydney, 1965. Giguere, P. A., and Jaillet, J . B., Can. J . Res., 1948, 26B, 767. Analytical Communications, October I996, Vol33 379 - 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Willits, C. O., Ricciuti, C., Knight, H. B., and Swem, D., Anal. Chem., 1952, 24,785. Houston, R. A., PhD Thesis, Pennsylvania State University, 1966. Madsen, B. C . , and Kromis, M. S . , Anal. Chem., 1984, 56, 2849. Perez-Ruiz, T., Martinez-Lozano, C., Tomas, V., and Val, O., Analyst, 1992, 117, 1771. Frew, J . E., Jones, P., and Scholes, G., Anal. Chim. Acta, 1983, 155, 139. Shiga, M., Saito, M., and Kina, K., Anal. Chim. Acta, 1983, 153, 191. Tamaoku, K., Murao, Y., and Akiura, K., Anal. Chim. Acta, 1982, 136, 121. Clapp, P. A., and Evans, D. F., Anal. Chim. Acta, 1991, 243, 217. Shaw, F., Analyst, 1980, 105, 11. Scott, G., Seitz, W. R., and Ambrose, J.,Anal. Chim. Acta, 1980,115, 221. Van Zoonen, P., Kamminga, D. A., Gooijer, C., Velthorst, N. H., and Frei, R. W., Anal. Chim. Acta, 1985, 174, 151. Ibusuki, T., Atmos. Environ., 1983, 17, 393. Matsubara, C., Kawamoto, N., and Takamura, K., Analyst, 1992,117, 1781. Smyth, M. R., and Vos, J. G., Analytical Voltammetry, Compre- hensive Analytical Chemistry, vol. 27, Elsevier, 1992. Tatsuma, T., Okawa, Y., and Watanabe, T., Anal. Chem., 1989, 61, 2353. Paper 6105736E Accepted August I6,1996

ISSN:1359-7337    

DOI:10.1039/AC9963300375

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Analytical Communications, October 1996, Vol33 (381-384) 38 1 Calculation of Intensities of Molecular Interferences in GD-MS: Application to Analysis of Aluminium Alloys Vladimir D. Kurochkin Institute for Problems of Material Science, National Academy of Science of Ukraine, 252142, Krzhizhanovsky 3 , Kiev, Ukraine. E-mail: vku@ ipms.kiev.ua A method for calculating the concentration of molecular ions in GD-MS is reported.The method is based on measurements of effective stability factors (ESF) of molecular ions. ESFs have been measured for various types of molecular ions: combination of Ar with gases and components of the sample; components with each other and gases. An investigation was carried out, by using a cryo-cooled pin-cell, on 5 aluminium standard materials. It is shown that ESFs for different types of molecular ions lie in the small range 1 X 10-9-10-8 ppm-l, except for ArH+, hydrides, oxides and nitrides of metals. Comparison with reaction rate constants show that polarization grip of an atom by an ion is mainly responsible for more or less uniform values of ESFs. With use of this method a program for mathematical simulation of mass-spectra was developed.The relative errors of calculated intensities of interferences do not exceed the 15% level. The method provides a way for more accurate estimations of unresolved interferences and gives a comprehensive idea of the spectrum in a given region. The number of possible combinations, in the GD-MS analysis of multicomponent alloys, of discharge gas, hydrogen, oxygen, nitrogen, carbon and components of the sample, resulting in polyatomic peaks, runs into the thousands.Software of modem instruments, such as VG9000 (VG Elemental, Winsford, Cheshire, UK), offers a means for calculation of locations of interferences using the operator's experience, but owing to the large number of polyatomic peaks, it is rather difficult not to miss important combinations.Another aspect of the problem is that existing methods do not allow one to calculate intensities of interferences and the final decision depends on the operator's skill. The interference problem for liquid samples can be resolved by the combination of ion chromatography with ICP- MS as an on-line matrix separation technique,* but with the GD-MS analysis of solid samples another method must be used, based, for instance, on a calculation of the concentration of molecular ions. Rate constants for ion-molecular reactions involving gaseous components have been much studied.This concerns first of all reactions between ArH2+ and H2,2a He and H2+,7,8 ArH+ and other gases (02,N2),9J0 Kr and H3+.ll Many fewer data are available on reactions between gases and ions of metals.12 However, the problem is that these data cannot be applied directly to the GD-plasma conditions because molecular ions are characterized by loosely bound atoms and the equilibrium concentration strongly depends on operating conditions.This paper presents a method for the calculation of intensities of molecular interferences for a given sample content and plasma conditions.On the basis of this study a programme for the calculation of intensities and the location of interferences for a given resolving power of the instrument and composition of the sample was developed. Experimental Instrumentation All analyses of standard reference materials were carried out on the VG9000 glow discharge mass spectrometer using the cryo- cooled pin-cell operated at 1 kV discharge voltage and 3 mA discharge current in Ar.The VG9000 instrument is a double- focusing magnetic sector/ESA mass spectrometer equipped with Daly electronic multiplier for single ion counting and Faraday cup detector for the measurements of large ion currents. Sample Preparation The analysed samples were prepared by cutting exactly sized pins (1.5 x 1.5 X 23 mm).The pin samples were then filed to obtain a fresh surface without any sharp edges, rinsed in ethanol in an ultrasonic bath for 5 min and finally plasma etched for 20 min in a GD source prior to analysis. Standard Reference Materials The standard materials analysed in the course of this study were 5 aluminium alloys. All standards were National standards of the former Soviet Union and have been certified on 14 elements ( Mg, Si, Ti, Cr, Mn, Fe, Cu, Zn, As, Zr, Mo, Sn, Sb and Pb).Model The selection of molecular ions, which are mainly responsible for the major interferences, was made on the basis of experimental data. The following ions are recognized as being the most probable under the GD plasma conditions: Ar,X,'+, Ar,Mmi+, XnYmi+ and XnMmi+, where X and Y are components of the sample, i = 1 or 2 (the charge of the ion), n and m = 1, 2 or 3 are the number of atoms in the molecule, and M are molecules and atoms from the following incomplete list (Ar, C, N, 0, H, c2, N2,02, H2, c3,03, H3, co, co2, co3, CN, CN2, CH, CH2, COH, CH3, CH4, C2H2, NO, NO2, N03, NH, NH2, NH3, NH4, CNH, OH, H20).The model takes into account all existing isotopes of the above mentioned atoms and their abundances.Reactions have been divided into several groups according to the above listed molecule types and reagents: (i), combinations of Ar with metals; (ii), Ar with oxygen, nitrogen, hydrogen; (iii), oxides, nitrides and hydrides of metals; (iv), combinations of Ar with complex molecules; (v), compounds of metals with metals; and (vi), combinations of metals with complex molecules.Argon ions, in contrast to the neutral atoms, are rather chemically active species. Reaction of components of the382 Analytical Communications, October 1996, Vol33 sample with argon ions runs in two steps: a polarizing grip of the atom by an ion and a quantum-mechanical interaction of electrons. The cross-section of the grip is given by the formula: 13 ogrip = 2~ @- (1) where fl is the polarization ability of the molecule and p is the reduced mass of particles.Thus the constant of the grip is kgrip = 2~ V m (2) and is independent of the relative velocity of particles. This formula gives the upper limit of the rate constant. The rate constants characterize the direct reaction but equilibrium composition of the plasma depends also on the decomposition reaction rate.There is a method of calculation based on a formula by Saha but it requires thermodynamic equilibrium at least among heavy particles.14 In addition, dissociation potentials of molecular ions need to be known. This short review demonstrates the complexity of the theoretical calculation. Thus, to help solve the problem, we introduce effective stability factors (ESFs) of molecular ions that have been determined empirically by measuring the intensities of molecular ions in the mass spectrum. In definition, ESF is the coefficient of proportionality in the equation: [CDd+] = ESF (CDd) X [C] X [Did (3) where [CDd+] is the concentration of the molecular ion normalized to the matrix current (as if the molecule were present in the sample); [C] and [D] are isotopic concentrations of elements C and D in the sample.Eqn. (3) corresponds to the reaction of the following type in plasma: C+ + dD = CDd+ (4) Because the concentration of ion [C+] is proportional to the concentration of element ‘C’ in the sample, it is much more convenient to use, for the determination of ESF, the concen- tration in sample [C]. That is why these factors are called effective.The definition of ESF, to some extent, is similar to the definition of relative sensitivity factors (RSF), where the immediate correlation between concentration in sample and intensity of ion current is established. ESFs are introduced in an analogous way as a stability constant for ordinary chemical reactions, except for concentrations that are not the true concentrations of reagents in plasma but concentrations in the solid phase in mass units.In the determination of ESF for ions of type ArMe+, the concentration of Ar was also measured by normalization to the matrix current. Such an approach allowed us to avoid the complicated and unreliable determination of the concentration of reagents in plasma and appeared to be very convenient for practical applications.ESFs were measured using a set of standard samples containing various elements in the concentration range (0.1-10%). To minimize the influence of insufficiently well resolved interferences on the precision of measurements we used several methods. Firstly, standard samples in which the components of the molecular ion studied were in high concentration while other interferences could be neglected were used.Secondly, also used was another combina- tion of the isotopes of reagents, creating a molecular ion free of the interference. Naturally, ESFs for reactions involving isotopes are identical. Suppose that the ESF for AlZn molecular ion is being measured. The determination of ESF could be carried out using molecular ions AF4Zn, APZn, APZn, A168Zn and A17OZn.These ions are insufficiently well resolved from atomic ions 9lZr, 93Nb, 94Zr, g4Mo, 95Mo and 97Mo. The concentrations of Zr, Nb and Mo in standard samples in this case must be as low as possible. If some doubts are still cast upon the influence of unresolved interferences the method of additions could be applied.This method is well known from common analytical practice. The principle of the method is that several standard samples having various concentrations of element (Zn in this case) are measured. If the concentrations of unresolved interferences are negligibly small compared with that of the measured molecule the calibration curve with axis ‘concentration of component’ (Zn) and ‘concentration of molecular ion’ (AlZn) must be a straight line passing through the origin.Otherwise, the concentration of elements (Zr, Nb and Mo in our case) could be calculated from the simple known procedure. Concentrations of molecular ions can be measured with the standard software of the VG9000 in two ways: measurements in the window of a specific isotope or by arranging the window for a given molecular ion.In the first case standard software corrects the concentration for the abundance [ab(isot)] and RSF(isot) of this isotope; thus, to obtain comparable values of ESFs, the right-hand side of eqn. (3) must be multiplied by the term RSF(isot)/ab(isot). The shape of the calculated lines were approximated by the normal distribution (Gauss) curve. Results and Discussion In Table 1 ESFs for major interferences obtained from analyses of several standards are represented.It appears that ESFs are Table 1 Values of effective stability factors for some singly charged molecular ions measured in cryo-cooled pin-cell at 1 kV voltage and 3 mA discharge current Molecular ion ESF ppm-’ RSD Molecular ion ESF ppm-I RSD Ar2 ArH ArC ArN ArO ArLi ArMg ArAI Ar2Al (2-body) Ar2Al (3-body) ArTi ArCr ArFe ArNi ArCU ArZn Ar2Zn Ar2Zr 4.2E-9 1 .OE-3 2.OE-9 1 .OE-9 1 .OE-9 6.9E-9 5.4E-9 4.5E-9 5.1E-9 2.1E-17 (ppm- 3.9E-9 3 .OE-9 2.OE-9 4.2E-9 2.5E-9 7.OE-9 1.6E-9 3 .OE- 10 A C C C C B A A A -2) A A B B B B B A A Ar3 (3-body) Ar3 (2-body) XH xc XN xo XY (average) XY2 (3-body) A12 Mg2 AlSi AlMg AlCu AlZn CuFe CuZn y3x y2x3 1.4E-18 (ppm-2) 3.3E-10 1 .OE-5 1 .OE-9 1 .OE-6 1 .OE-6 5.4E-9 1.OE-18 (ppm-2) 1.3E-8 1.3E-8 1.4E-8 7.1E-9 1.3E-9 5.58-9 2.OE-9 2.5E-9 < 5.08-24 < 1.OE-26 A A C C C C B B A A A A A A B B C CAnalytical Communications, October 1996, Vol33 383 independent of the composition of the standards provided that they have the same matrix.The relative standard deviation of the mean (RSD) was calculated either from multiple analyses of a single standard or from multiple and single analyses of sets of standards.The following designations for RSD are used in Table I: A d k 5% ; B d f 10%; C S f 20%. Even a cursory examination of the rate constants of reactions between argon and other atoms calculated by means of expression (2) or taken from the literature’-10 shows that, owing to the weak dependence of polarization cross-section of p and p, their values lie in a rather small range (n.10-10-n.10-9 cm3 s-l).Our measurements also indicate more or less uniform values of ESF (within an order of magnitude) for various types of molecular ions, except for ArH+ and oxides and nitrides of metals. Molecular ions of type ArX+ or XY+ in general are characterized by loosely bound components.This is shown by experiments with a ‘hot’ cell, i.e., one without cooling by liquid nitrogen, where the added concentrations of molecules 68ZnA1+ and 40ArMn+ superposed on 95M0+ reduce by approximately 7 times compared with those in a cryo-cooled cell. This effect is mainly due to the decreasing number of 6*ZnA1+ molecules because the concentrations of other argon containing interfer- ences 40AFZn+ and 40AFCu+ reduce by three times.The concentration of molecular ions *4MgA1+ reduces in a ‘hot’ cell by 3 orders of magnitude. Thus, a ‘hot’ cell, in many cases, may be used for decreasing the intensities of interferences. L I I I I c 58.93 50.96 s 0 .- c. b97 I CI c a, ( c ) Zn 78A127 l E 4 * lE+3 lE+2 l E + l lE+B 1E- 1 1E-2 1E-3 96.88 96.93 96-99 Argon ions most firmly bind to a hydrogen atom.As has been shown by Roche15 proton affinity to Ar lies in the range 3.4-4.2 eV. This is a rather high binding energy: thus, in a ‘hot’ cell the concentration of ArH+ ions increases. However, simultaneously the concentration of hydrogen atoms increases. In both cells practically all of the hydrogen is present as the molecular ion ArH+.Table 1 indicates that ESF, for reactions of Ar ions with hydrogen, appears to be 6 orders of magnitude higher than for oxygen- and nitrogen-containing ions. As is indicated in Table 1, ESFs for different types of molecules (ArX+, XY+) depend only slightly on the type of atoms taking part in a reaction. This result and experiments with a ‘hot’ cell support the idea that polarization grip is mainly responsible for the molecular ion bond.ESF in definition correlates with the equilibrium constant of the reaction, i.e., with the ratio of forward and reverse reaction rate constants. This is seen from reactions where three-atomic molecular ions are formed. Such reactions may progress in two steps: X + + X + M = Xz++M ESF(X2’) ( 5 ) X2+ + Y + M = XzY+ + M ESF(X2Y’) (6) Three-body collision reaction results in a long lived complex ion.The third particle taking excess energy provides conditions for the creation of stable complex ions.14 ESF for summed N i60 ( b ) Hg 24Ar36 e 1 T i a 1 4 (I1 2713Cu 65 I , . . , 59.91 59 .% 59 -98 Masslu 145.87 145.95 146.04 Fig. 1 Calculated and measured mass spectra (solid line) of 51V, 60Ni, 97Mo and I46Nd for standard aluminium alloy B96 No.10. Table 2 Content of standard aluminium alloy B96 no. 10 Elements Mg A1 Si Ti Cr Mn Fe c u Zn Zr Concentration (%) 3.20 87.26 0.14 0.10 0.29 0.17 0.48 2.74 5.43 0.19384 Analytical Communications, October 1996, Vol33 reaction must be equal to the product of the ESFs of reactions (5) and (6): ESF(X2Y+ 3 body) = ESF(X2+) X ESF(X2Y') (7) As can be seen from Table 1, for two-body reactions ESF(Ar2) = 4.2E - 9 ppm-l and ESF(Ar2A1+) = 5.1E - 9 ppm-1.The product of these values is equal to 2.14E - 17 ppm-2. This figure coincides with ESF(Ar2A1+) for a three- body reaction as it must for equilibrium constants of ordinary chemical reactions. The same is true for the Ar3+ molecule.The number of molecules of type XY+ is equal to the number of combinations of rn elements taken two at a time (C,2), where m is the number of components of the sample. This is a rather large number and in this study, to simplify the model, average values of ESFs were used. Such approximation is quite sufficient for semi-quantitative analysis but sometimes error may not be permissible.Therefore, the ESFs for AlTi+ and AlCu+ molec- ular ion are three times lower than the average value shown in Table 1. Thus, in further investigations it will be desirable to improve the model, extending the list for molecules of XY+ type, at least for combinations involving matrix ions. More complex molecules of type X3Y+ and X2Y3+ are of less importance and currently it is possible to evaluate only the upper limit for ESFs.In Fig. 1 are shown experimental data and calculated mass spectra for aluminum standard reference alloy B96 no. 10. The composition of the alloy is shown in Table 2. Interferences are arranged in decreasing order from the top down. Calculated spectra are a superposition of all interferences and the real concentration of the isotope studied (a broken line on the figures).All figures were calculated for a resolution of 5000 on the VG9000 but the experiment could easily be repeated for every necessary resolution. The list of interferences is restricted by a 100 ppt level. The calculations were carried out on the assumption that the concentrations of gases 0 and N were of about 10 ppm.The ability of the mathematical model to show the shape of spectra in a given region is especially useful during deviations of the mass mark because it gives a comprehensive idea about lines in this region of spectra. In the case of resolved interferences it is relatively simple to avoid errors in the Table 3 Characterization of some unresolved interferences for aluminium alloy B96 no.10 Intensity of interference (ppm) Calculated' Isotope Molecular molecule Mass ion Sum Measured 88.9058 88.9103 0.5 1.2 1.09 88.9128 0.4 88.9150 0.2 88.9222 0.1 96.9060 96.9069 12.2 22.4 24.2 96.8978 10.2 97.9054 97.8957 0.5 0.5 2.8 interpretation of results. Application of the method to un- resolved interferences enables us more accurately to take into account the contribution of molecular ions in summed intensity.For the given sample content the region from 90 to 130 mlz, approximately, is particularly unfavourable from this point of view. This observation can be demonstrated in the determina- tion of Y and Mo (Table 3). Table 3 indicates that the peak measured as 89Y is actually a superposition of interferences and must be rejected. Isotopes 97Mo and 98M0 both show excessive intensity over the summed intensity of molecular peaks and the real concentration of molybdenum is about 2.0 ppm.This method must be used with caution if the intensity of the molecular peak is close to the measured data. In this case it is possible to attribute the peak to the isotope specified if the measured intensity exceeds an interference by a value more than relative error of the calculation.Otherwise, it is possible to determine only the upper limit of concentration. As indicated, comparison of calculated and measured spectral average relative error by this method do not exceed 15%. ESFs depend more strongly on discharge conditions than RSFs but prelimi- nary results show that measured ESFs could be applied to the analysis of alloys with another matrix.The values of ESFs, especially for complex ions, must be refined by further experiments with standard reference materials of various matrices. References 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 Feldmann, J., Anal. Commun., 1996, 33, 1 1 . Bedford, D. K., and Smith, D., Int. J. Mass Spectrom. Ion. Processes, 1990, 98, 179. Liao, C.-L., Xu, R., Flesh, G. D., Baer, M., and Ng, C. Y., J . Chem, Phys., 1990, 93, 4818. Fehsenfeld, F. C., Ferguson, E. E., and Mosesman, M., Chem. Phys. Lett., 1969, 4, 73. Bowders, M. T., and Elleman, D. D., J . Chem. Phys., 1969, 51, 4606. Smith, R. D., Smith, D. L., and Futrell, J. H., Int. J . Mass. Spectrom. Ion Phys., 1976,20, 223. Pollard, J. E., Johnson, L. K., and Cohen, R. B.,J. Chem. Phys., 1991, 95, 4894. Theard, L. P., and Huntress, W. T., J . Chem. Phys., 1974, 60, 2840. Kim, J. K., Thread, L. P., and Huntress, W. T., J. Chem. Phys., 1975, 62, 45. Lindinger, W., McFarland, M., Fehsenfeld, F. C., Albritton, D. L., Schmeltekopf, A. L., and Ferguson, E. E., J . Chem. Phys., 1975, 63, 2175. Bohme, D. K., Mackay, G. I., and Schiff, H. I., J. Chem. Phys., 1980, 73,4976. Fisher, E. R., Elkind, J. L., Clemmer, D. E., Georgiadis, R., Loh, S. K., Aristov, N., Sunderlin, L. S., and Armentront, P. B., J . Chem. Phys., 1990, 93, 2676. Giumousis, G., and Stevenson, D. P., J . Chem. Phys., 1958, 29, 294. Smimov, B. M., Ions and Excited Atoms in Plasma, Atomisdat, Moscow, Russia, 1974, pp. 141-148. Roche, A. E., Sutton, K. M., Bohme, D. K., and Schiff, H. I., J. Chem. Phys., 1971, 55, 5480. Paper 6103755K Received May 29,1996 Accepted July 17, 1996

ISSN:1359-7337    

DOI:10.1039/AC9963300381

出版商:RSC    

年代:1996

数据来源: RSC