ANALYST, DECEMBER 1991, VOL. 116 1211 lon-selective Electrode Studies on Novel Organic Molecule Sensors* J. D. R. Thomas School of Chemistry and Applied Chemistry, University of Wales College of Cardiff, P.O. Box 912, Cardiff CFI 3TB Researches on prospects for novel ion-selective electrodes, based on organic molecule sensors, are described. The organic molecules are large crown ethers extending from bis( metaphenylene)-26-crown-8 to bis(metaphenylene)-38-crown-l2, small crown ethers, bis-crown ethers, and acyclic polyethers consisting of diphenyl ethers of tetraethylene glycol and receptor molecules of planar and tetrahedral tripodal types. Keywords: Acyclic polyether sensor; crown ether ion sensor; ion-selective electrode; neutral carrier electrode The various organic molecules based on structures containing ethoxylate (-CH2CH20-) units owe their function as poten- tiometric ion sensors to the ability of the ethoxylate structure to either crown, or coil around, or clasp analates.Such complexation is facilitated by the organic molecules contain- ing the ethoxylate units carrying a sequence of localized charges of sufficient energy to form ion-dipole bonds with appropriate cations. These charges are frequently com- plemented by features of other substituents within the molecule, such as phenolic, carboxylic and aromatic ring substituents. Overall, the conformation of the molecule permits a solvation type shell around the cation effectively to replace the ion hydration shell. The charged cationic complex thus formed is electrically balanced by anions.The impetus for studies on structure with polyethoxylate units as ion sensors is attributable to the observation of Moore and Pressman’ that valinomycin is capable of actively trans- porting potassium across rat mitochondria membranes. This led to the very successful valinomycin sensor for potassium.2-3 Also relevant is the description by Pedersen4 of the function of crown ethers in promoting the dissolution of salts in which they arc otherwise insoluble. However, it is to be noted that cation adducts of polyethylene glycols had already been known for several years,5-7 and it was this category of ethoxylates that was first exploited in ion-selective electrodes (TSEs). The special properties of polyethylene glycols in this application are attributable to the ethoxylate units (EOUs) in the polyethoxylate complex assuming a tight helical conforma- tion of appropriate ring radius, for holding the ion in a ‘cage’ of oxygen atoms of the EOUs by ion-dipole interaction.8 For the particular instance of barium ions complexing with a nonylphenoxypolyethoxylate (NP) [Antarox (or Igepal) CO- 880 with 30 EOUs], 12 EOUs are involved in holding the Ba*+ ions in a tight helical arrangement with a ring radius of about 1.3 A, the cage for the barium ions being formed by the 12 oxygens in two loops of 6 EOUs each.8 For the crown ethers, some of the earliest ISE studies were by Rechnitz and Eya19 on dicyclohexyl- and dibenzo-18- crown-6 ligands.Shortly afterwards, Petranek and Ryba’O.1’ assessed various types of crown-6 and larger crown com- pounds for sensing potassium.They favoured dimethyldiben- zo-30-crown-10 with dipentyl phthalate in poly(viny1 chloride) (PVC) as the best of the series. Although these and later studies on crown ether sensors showed that the selectivity for potassium does not exceed that by valinomycin, researches on polyethers as potentiometric sensors have yielded many interesting features and new directions in sensing. This paper outlines studies in Cardiff on attempts to assess and exploit additional molecular systems based on EOUs as potentio- * Presented at thc Royal Socicty of Chemistry 150th Anniversary Congress, Imperial College, London, 8th-11th April, 1991. metric ion sensors in PVC matrix membranes with appropriate plasticizing solvent mediators.The types of structures dis- cussed consist of: (i) large crown ethers for sensing diquat (DQT), paraquat (PQT) and guanidinium; (ii) small crown ethers for lithium; (iii) bis-crown ethers for alkali metal cations; and ( i v ) acyclic structures, consisting of ‘clasp’ type diphenyl ethers of tetraethylene glycol and ‘scorpion’ or ‘grasp’ types based on planar and tetrahedral tripodal receptor molecules for alkali and alkaline earth cations, and guani- dinium. Large Crown Ethers as Sensors Diquat and Paraquat The way in which a selection of crown ether type neutral carrier molecules for ion-sensing can be helped by structural studies on interactions between the carriers and ions has been well demonstrated by Stoddart and co-workers in Sheffield (now at Birmingham), and Williams in London. Their work showed interesting features on the interactions between macrocyclic pol yether molecules and cations according to three types of bonding (for relevant references, see ref.12): (i) coordination via [N-H . . . 01 hydrogen bonds as in complexation between polyether and primary alkylammo- nium ions; (ii) coordination through [C-H . . . 01 linkages to the DQT dication, alkylphosphonium cations, etc. ; and (iii) metal cation coordination to oxygen atoms of the crown ether compound. The structural information obtained by Stoddart and co-workers has been directed in Cardiff12 to the development of ISE sensors for DQT and PQT dications based on dibenzo-30-crown-10 (DB30C10). This is an important appli- cation in view of the use of DQT and PQT as contact herbicides.The maximum stability’3-16 for dibenzo-3n-crown-n (DB3nCn) complexes with DQT occurs when n = 10 (Table 1) while the X-ray structure shows the plane of the DQT molecule to be enclosed in a U-shaped cavity14 formed by DB30C10. The complexation involves three types of Table 1 Stability constants ( K , ) and free energies of formation (AC,) of DQT.DB3nCn complexes in acetone* Crown cther K,ldm3 mol-1 AGfIkJ mol-1 DB27C9 4.1 x 102 15.0 DB30C10 1.8 x 104 24.3 DB33C11 1.1 x 104 23.1 DB36C12 2.0 x 103 18.9 DB30ClO(-OCH2)2Ph - 31.0 * Data from refs. 14, 15 and 17.1212 ANALYST, DECEMBER 1991, VOL. 116 H CH3 H CH3 lo -0-0-0-0 wowowo~o a: DB3nCn DB-crown 0 1 BMP-crown BPP-crown Fig. 1 Chemical structures (top) of (a) 4,4'-dipyridinium, (b) paraquat and (c) diquat, showing some dimensions, and structure (bottom) of dibenzo-type 3n-crown-n crown ethers, showing the positions of the benzene 0,O-disubstitution in dibenzo (DB), bis(metapheny1ene) (BMP) and bis(parapheny1ene) (BPP) crown ether derivatives (from ref.17) DB30C10 (host)-DQT (guest) interactions: (i) DB30C10 catechol-oxygen electrostatic interaction with the positively charged nitrogen atoms in DQT. For this, the crown ether catecholo-0 separation (2.6 A) and N-N separation in DQT (2.8 A) are similar (Fig. 1) so that the catechol oxygens are neatly directed above and below the DQT nitrogens in the [DQT.DB30C10]2+ complex; (ii) DB30C10 benzene ring x-electron charge transfer to the electron deficient DQF+ ; and (iii) hydrogen bonding between H6 and H6' (Fig.1) with oxygen atoms in the DB30C10 framework, and as mentioned above. A wide ranging study was then undertaker1,~3>*~ during which various crown ethers, namely, DB30C10, bis-meta- phenylene-32-crown-10 (BMP32C10), bis-metaphenylene-38- crown-12 (BMP38C12), bis-paraphenylene-34-crown-10 (BPP34C10), bis-paraphenylene-37-crown-l l (BPP37Cll) and dinaphthalene-36-crown-10 (DN36C10) were tested for their efficacy as sensor systems for DQT and PQT and their conformational behaviour in their complexation with the two dications and with the related 4,4'-dipyridinium (4,4'-DPy). Also studied18 was the use of ion-pairing reagents in ISEs for the dications, these being related also to searches for the best type of anion excluder for use in PVC ISE type membranes with the crown ether neutral ligand.The best electrodes for DQ-T are basedl3J7 on DB30C10 plus DQT-2TPB (TPB = tetraphenylborate) with either 2- nitrophenyl phenyl ether (NPPE) or 2-nitrophenyl octyl ether (NPOE) as solvent mediator [Fig. 2(a)] but good electrodes were also obtained with just DQT-2TPB and the solvent mediators [Fig. 2(3)]. The other crown ethers studied gavel7 good ISE properties for 4,4'-DPy of calibration slope between 33 and 41 mV decade-', extending down to 1.7-6 pmol dm-3. However, the expectation in this study18 was the discovery of a PQT ISE based on BPP34C10 as a complex exists19 (of stability constant 730 dm3 mol-l) through the 0-0 separation (5.5 A) in this host being not too far removed from the N-N separation of 7.0 8, in PQT [Fig.1(6)]. The expectation was 8 6 4 2 0 -2 -4 I l I I I I 1 1 1 I I 1 m -,g 5 a0 Y m 4 3 -J 2 1 0 -1 -2 -3 DOT POT Mg Ca Ba Li Na K NH4 Gu PhNH, Et2NH2 B Fig. 2 (a) Effect of plasticizing solvent mediator on the selectivity of diquat ISEs based on DB30C10 plus DQT.2TPB. (b) Effect of plasticizing solvent mediator on the selectivity of diquat ISEs based on DQT-2TPB without crown ether. 0, NPOE; El, NPPE; A, DOPP; A, dinonyl phthalate (DNP); and M, DBP. Lines joining points connect the various solvent mediator types (from ref. 17) not realized,17 probably because the [PQT.BPP34C10]2+ complex is considerably less stable than [DQT-DB30C10]2+ (stability constant 18 000 dm3 mol-1) and, therefore, much too weak to exhibit a potentiometric function. A wide range of anion systems, namely, phosphorus hexafluoride (PF6) , anthraquinone-2-sulphonate (AS), octyl sulphate (OS), picrate (PIC) , dipicrylaminate (DPA), Diam- ine Green B (DGB), tetraphenylborate (TPB) and tetrakis-4- chlorophenylborate (T4ClPB) have been studied as possible ion-pairing agents for use in ISE membranes for PQT, DQT and 4,4'-DPy, but only PF6, TPB and T4ClPB gave good electrodes.18 The PF6 electrodes were not selective, thus leaving just the TPB and T4CIPB systems as suitable for exploitation.In this respect, electrodes based entirely on DQT-2TPB or PQT.2TPB had already been found12 to give excellent ISEs with appropriate solvent mediators. The calibration response of a DQT.2T4CIPB based ISE for DQT during an appraisal study20 was found to be stable for 55 d with a near-Nernstian slope and detection limit in the pmol dm-3 range.The electrode was of fast response (3 s at 1 mmol dm-3 and 25 s at 1 pmol dm-3) and was useable at pH 2-12 over a sample temperature range of 2-50 "C. Samples could be analysed by the standard additions method with about a -5% error and a precision of 7-8%. For comparison, DQT was also determined by titration with sodium tetra- phenylborate using DQT.2TPB and tetrabutylammonium tetraphenylborate (TBAeTPB) ISEs as sensors. Here, the errors were 12-20% for DQT analysis in de-ionized water, sodium chloride solution or simulated serum.** Guanidinium The importance of guanidinium in the biological and medical fields has led to interest in ISE development. For this, the guanidinium cation, [(H2N)3C]+, can complex with crownANALYST, DECEMBER 1991, VOL.116 1213 ethers of between 18 and 33 ring atoms.21-23 The 27-mem- bered ring is the most selective towards the guanidinium cation (Gu+), the stability constant trend being K > Gu > alkylammonium > other metal cations.21-23 X-ray and nuclear magnetic resonance (NMR) spectroscopy data21-24 support the [NH . . . 01 arrangement of hydrogen bonds to yield stable complexes of 1 mol host to 1 mol guest in most instances and, occasionally, 2 : 3, depending on the solvent and guanidinium salt . In initial studies25326 on crown ether based ISEs. for guanidinium, the most suitable system was based on dibenzo- 27-crown-9 (DB27C9) using dibutyl phthalate (DBP) in PVC. Dioctyl adipate (DOA) and dioctyl sebacate (DOS) plasticiz- ing solvent mediators with the same sensor yielded the next best electrodes,26 while NPOE, NPPE and dioctylphenyl- phosphonate (DOPP) proved to be unacceptable as solvent mediators.It is to be noted that functional Gu+ ISEs may be obtained with just guanidinium tetraphenylborate as sensor with either DBP or DOPP as solvent mediator, each with good all-round selectivity.26 The pH interference-free range (4-10) was narrower for the electrodes without crown ether26 than those with crown ether, when the pH range was 3-12. In a consideration27 of the relative merits of DB27C9, DB30C10 and bis( metaphenylene)-26-crown-8 (BMP26C8) as sensor, the 26-membered ring system of BMP26C8 was deemed to have the optimum crown ether ring size for sensing guanidinium. This worked well with either DBP or DOA as solvent mediator, and functioned both with and without either guanidinium tetraphenylborate or guanidinium tetrakis(4- chloropheny1)borate as the anion excluder.Hence, the metaphenylene component of the crown ether led to an optimum crown to improve on the previous best 27-membered system. The electrodes were functional over several weeks (>16), while the main interferents were pyridine, tetraethy- lammonium and PQT.27 Small Crown Ethers as Sensors Lithium is an effective therapeutic agent for the treatment of manic depression. Hence, a great deal of attention has been given to the development of lithium ion sensors. The analytical constraint is the clinical requirement of maintaining the lithium level between about 0.5 and 1.0 mmol dm-3; this being in a background of approximately 0.15 mol dm-3 sodium means that the sodium level is about 200 times the planned lithium level and 1400 times the lowest lithium level.Therefore, a successful lithium ion sensor should have sufficient selectivity for lithium to discriminate against large sodium backgrounds. In the development of lithium ISEs, the most encouraging results have been obtained for sensor membranes consisting of neutral carrier molecules, admixed with compatible solvent mediators in PVC matrices. The essential guideline is that the neutral carrier must preferably have 4-5 coordinating sites, and that the ligand should be sufficiently flexible to allow a fast exchange, despite the rigidity in arrangements of coordination sites.These points are important in relation to reversibility for the electrode. Of course, lithium being small with a high charge density tends to be highly solvated; hence, a suitable neutral carrier should be capable of stripping the water of hydration off the lithium ion (the enthalpy of hydration of lithium is high, namely, 510 kJ mol-1). In relation to the above, research on the development of crown ethers for lithium ISEs has centred on 12-crown-4 through to 16-crown-4 type molecules.28 The lithium ion (0.68 A diameter) must fit snugly into the polyether cavity, and the crown ethers that fit such a specification range from 12-crown- 4 to 14-crown-4. Others, such as 15-crown-4 and 16-crown-4 have larger cavity diameters and are selective for larger cations.The 12-crown-4 ether (Fig. 3) with its 'ideal' cavity size for lithium gave an electrode29 of near-Nernstian response at R W- n (4) Fig. 3 Some crown-4 molecules studied as lithium ion s e n ~ o r s ~ , ~ * 1 x 10-5-1 x 10-4 mol dm-3, but was of poor selectivity, particularly with respect to sodium. A later investigation30 modified the 12-crown-4 by having a substituent R (-CH20C18H37 or -CH20C( O)CI7H35) designed to hinder the formation of sandwich complexes with the larger alkali metal cations. This proved to be unsuccessful in that the kR,'Na values deteriorated from 0.12 for the simple 12-crown-4 material to 81 and 280 for the respective derivatives. Of a comprehensive range of crown-4 derivatives studied by Kitazawa et aZ.30 covering 12-, 13-, 14-, 15- and 16-membered rings, it was found that lithium ion selectivity was dramatically enhanced for the 14- and 15-crown-4 derivatives with keT,tNa being in the range 0.032-0.062, but deteriorated for larger ring sizes.Of course, it is important to realize that selectivity ultimately depends on the solvent mediator, and several have to be investigated before dismissing the prospects of any single sensor. For the crown ethers mentioned above, the mediator was NPOE.30Jl A benzo-14-crown-4 derivative (Fig. 3) proved to be disappointing32 as did the corresponding material without the right-hand phenoxy group (unpublished work). The ISEs from both gave similar lithium ion responses, but the selectivity was inadequate for overcoming the sodium back- ground of blood serum.32 The sodium uptake by the mem- brane containing this derivative during permeation experi- ment@ with radiotracers was deemed to be mainly due to PVC-solvent only contributions.Sensing With Bis-crown Ethers A reason for the poor selectivity of the 12-crown-4 ether (Fig. 3) towards lithium is the possibility of sandwich compound formation between two 12-crown-4 ether ligands and one sodium i0n.29 This leads to prospects for bis-crown ethers as ion sensors. There has been enquiry in this direction,34-3* for example, Shono and co-workers35~36 have demonstrated systems with good potassium ion selectivity for some bis(l5- crown-5) derivatives, and one containing a dodecyl link exhibits good lipophilicity and longer lifetime for the resulting ISEs.36 Further studies in Cardiff37 on bis( 1,4,7,1O-tetraoxa- dodecan-2-ylmethyl) 2-dodecyl-2-methylmalonate ( 5 ) and bis-( 1,4,7,10 , 13-benzopentaoxacyclopentadecin- 15-ylme th yl) heptanedioate (6) (Fig.4) show the former to be selective to sodium and the latter to potassium, both being best used with NPOE plasticizing solvent mediator and potassium tetrakis-4- chlorophenylborate anion excluder in PVC matrices.1214 0 *C cY C- /ClZHZ5 C //O ANALYST, DECEMBER 1991, VOL. 116 (7) R=H (8) R=CHZPh Fig. 4 Bis( l74,7,10-tetraoxadodecan-2-ylrnethyl) 2-dodecyl-2-methylmalonate (5) and bis( 1,4,7,10,13-benzopentaoxacyclopentadecin-15- ylmethyl) heptanedioate (6). Also shown is 1,ll-bis(2-hydroxy-5-formylphenoxy)-3,6,9-trioxaundecane (7) and its benzyl ether (8) Although the above electrodes offer promising alternatives to glass electrodes for sodium and to valinomycin electrodes for potassium, the data for measurements of the ions in blood serum demand further research in order to improve the correlations with flame photometric measurements.37 The molecules shown in Fig.4 have also been assessed as possible uranyl ion sensors.38 The pimelate failed to respond, while the malonate gave a steady uranyl response over the 18 d study period with a near-Nernstian slope. The solvent mediator in each instance was NPOE. The fact that the malonate-based ISE failed to respond after storage for 5 months was attributed mainly to minor radiation damage.38 Sensors Based on Ethoxylate-type Acyclic Structures The most exciting acyclic polyether structures studied at Cardiff for their scope in potentiometric sensing are clearly the various polyalkoxylate types, and especially those based on nonylphenoxypolyethoxylate (Antarox types) with different numbers of alkoxylate units, and referred to earlier in this paper.These were initially developed for their barium ion sensing qualities, but a bonus was the ability to sense alkoxylates themselves. Further, the inflection in the calibra- tion provides a facility for measuring the critical micellization concentration (CMC) of alkoxylate-type non-ionic surfac- tan t s .39-41 With regard to CMC, the inflections observed in the e.m.f. versus log[alkoxylate] graphs are definite and the breaks40 compare favourably with literature CMC data42-44 and agree with a derived equation.42 log (106 x CMC) = An + B (1) where A and B are constants for a particular hydrophobic group and n is the average number of EOUs in the molecule. Diphenyl Ethers of Tetraethylene Glycol Two derivatives of these have been studied27,32,38,45 namely, 1,l l-bis(2-hydroxy-5-formylphenoxy)-3,6,9-trioxaundecane (7) and its benzyl ether (8) (Fig.4). These can be looked upon as ‘clasping’ cations in ion sensing, and they were initially assessed as possible alkali and alkaline earth metal ion sensors in association with seven different solvent mediators.45 The response characteristics of compound (7) are dependent on the solvent mediator. Also, the systems with tris(2-ethylhexyl) phosphate and DOPP, each without a sensor, exhibited a significant response towards lithium ions.However, the over-all characteristics were slower, noisier, less stable and showed greater drift than the corresponding systems with a sensor present. Further evaluation32 as a sensor for lithium showed compound (7) to be much inferior to a lipophilic diamide. Continuing studies27 with compounds (7) and (8) yielded greater success with guanidinium ion sensing. Hence, each when used with either DBP or DOA solvent mediator is better than the previously recommended26 DB27C9 but, as already suggested above, they were not as good as the now recommen- ded27 BMP26C8. Sensor (8), when used with DBP or DOPP as solvent mediator, turned out to be a potential uranyl ISE sensor. This was largely on the premise of the marked reduction of iron(m) interference with sodium tetraphenylborate as anion excluder.Planar and Tetrahedral Tripodal Receptors Nine acyclic pol yethers, representing examples of planar and tripodal ‘scorpion-like’ molecules (Fig. 5), each with oligo- ether tails and a pair of anionic type ‘pincers’ together designed to ‘grasp’ ions, have recently been evaluated6 as possible sensors for barium ISEs. The chosen solvent media- tor was NPPE. The general performance of the resulting series of electrodes46 was inferior to the traditional ISE based on the tetraphenylborate of the barium complex with Antarox CO-880 (a nonylphenoxypolyethoxylate with 30 EOUs) . However, a general barium ion response seems to be favoured by a tetrahedral tripodal structured sensor (where X is benzyl in Fig. 5) with its design promoting good ion-dipole interac- tions, as seen in another on association constants of these acyclic polyethers with barium ions.ANALYST, DECEMBER 1991.VOL. 116 1215 C? p-x Fig. 5 studied as possible sensors for barium I S E S ~ ~ Planar and tetrahedral tripodal receptor type molecules Conclusion The studies presented demonstrate the underlying benefits in sensor development of structural-backed syntheses related to experience of sensor behaviour. The resulting comparative work has yielded promising sensors for a number of cations, including DQT, PQT, Gu+ and uranyl, and promising leads for lithium, sodium, potassium and barium. The author thanks his many co-workers for their dedication, and also Professor J. F. Stoddart (University of Birmingham) and Dr.D. J. Williams (Imperial College, London) and their co-workers for structural information and the provision of a range of materials. The various sponsors to whom acknow- ledgement has already been made in the cited references are also thanked for generous financial support, and steered by grants from the Science and Engineering Research Council under the sensors initiative for the main thrust in the research programme. 1 2 3 4 5 6 7 8 9 10 11 12 13 References Moore, C., and Pressman, B. C., Biochem. Biophys. Res. Commun., 1954, 15, 62. Stefanac, Z., and Simon, W., Chimia, 1966,20,436. Stefanac, Z . , and Simon, W., Microchem. J., 1967, 12, 125. Pedersen, C. J., J . Am. Chem. SOC., 1967, 89, 7017. Neu, R., Arzneim. Forsch., 1959, 9, 585. Uno, T., and Miyajima, K., Chem.Pharm. Bull., 1963,11,75. Levins, R. J., and Ikeda, R. M., Anal. Chem., 1965, 37, 370. Levins. R. J.. Anal. Chem., 1971,43, 1045. Rechnitz, G. A., and Eyal, E., Anal. Chem., 1972,44, 370. Petranek, J., and Ryba, O., paper presented at the IUPAC International Symposium on Selective Ion-Sensitive Electrodes, UWIST, Cardiff, UK, 9th-12th April, 1973, paper No. 13. Ryba, 0.. and Petranek, J., J. Electroanal. Chem., 1973, 44, 425. Moody, G. J.. Owusu, R. K., and Thomas, J. D. R., Analyst, 1987, 112, 121. Colquhoun, H. M., Goodings, E. P., Maud, J. M., Stoddart, J. F., Williams, D. J., and Wostenholme, J. B., J. Chem. SOC., Perkin Trans. 1, 1985, 2, 607. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Kohnke, F.H., Stoddart, J. F., Alwood, B. L.. and Williams, D. J., Tetrahedron Lett., 1984, 26, 1681. Kohnke, F. H., and Stoddart, J. F., Tetrahedron Lett., 1984,26, 1685. Allwood, B. L., Kohnke, F. H., Slawin, A. M. Z., Stoddart, J. F., and Williams, D. J., J. Chem. SOC., Chem. Commun., 1985, 311. Moody, G. J., Owusu, R. K., and Thomas, J. D. R., Analyst, 1988, 113, 65. Moody, G. J., Owusu, R. K., and Thomas, J. D. R., Analyst, 1987, 112, 1347. Colquhoun, H., Stoddart, J. F., and Williams, D. J., New Sci., 1986, May, 44. Moody, G. J., Owusu, R. K., and Thomas, J. D. R., Anal. Lett., 1988, 21, 1653. Kyba, E. P., Helgeson, M. K., Madan, K., Gokel, G. W., Tanowski, T. L., Moore, S. S., and Cram, D. J., J . Am. Chem. SOC., 1977,99,2564. Lehn, J.-M., Vierling, P., and Hayward, R.C., J. Chem. SOC., Chem. Commun., 1979,296. Stolwijk, T. B., Grootenhuis, P. D. J., van der Wal, P. D., Sudholter, J. W. H. M., and Kruise, L. J., J . Org. Chem., 1986, 51, 4891. Colquhoun, H. M., Goddings, E. P., Maud, J. B., Stoddart, J. F., Williams, D. J., and Wostenholme, J. B., J. Chem. SOC., Chem. Commun., 1983, 1140. Bochenska, M., and Biernat, J. F., Anal. Chim. Acta, 1984, 162, 369. Assubaie, F. N., Moody, G. J., and Thomas, J. D. R., Analyst, 1988, 113, 61. Assubaie, F. N., Moody, G. J., and Thomas, J. D. R., Analyst, 1989, 114, 1545. Gadzekpo, V. P. Y., Moody, G. J., Thomas, J. D. R.. and Christian, G. D., Ion-Sel. Electrode Rev., 1986, 8, 173. Gadzekpo, V. P. Y., and Christian, G. D.. Anal. Lett., 1983,16, 1371. Kitazawa, S., Kimura, K., Yano, H., and Shono, T., J . Am. Chem. SOC., 1984, 106,6978. Kitazawa, S., Kimura, K., Yano, H., and Shono, T., Analyst, 1985, 110, 295. Beswick, C. W., Moody, G. J., and Thomas, J. D. R., Anal. Proc., 1989, 26, 2. Harris, N. K., Moody, G. J., and Thomas, J. D. R., Analysf, 1989, 114, 1555. Lindner, E., T6th, K., Orvath, M., Pungor, E., Agai, B., Bitter, I., Toke, L., and Hell, Z., Fresenius 2. Anal. Chem., 1985,322, 157. Kimura, K., Mazeda, T., Tamura, H., and Shono, T., J. Electroanal. Chem., 1979, 95, 91. Kimura. K., Tamura, H., and Shono, T., J . Chem. SOC., Chem. Commun . , 1983,492. Moody, G. J., Saad, B. B., and Thomas, J. D. R., Analyst, 1989, 114, 15. Johnson, S.. Moody, G. J., Thomas, J. D. R., Kohnke, F. H., and Stoddart, J. F., Analyst, 1989, 114, 1025. Moody, G. J., and Thomas, J.D. R., in Non-ionic Surfactants: Chemical Analysis, ed. Cross, J.. Marcel Dekker, New York, 1986, p. 117. Jones, D. L., Moody, G. J., and Thomas, J. D. R., Analyst, 1981, 106, 439. Jones, D. L., Moody, G. J., Thomas, J. D. R., and Birch, B. J., Analyst, 1981, 106, 439. Hsiao, L., Dunning. H. N., and Lorenz, P. B., J. Phys. Chem., 1956, 60,657. Schick, M. J., Atlas, S. M., and Elrich, F. R., J. Phys. Chem., 1962, 66, 1326. Alexander, P. H. V., Moody, G. J., Thomas, J. D. R., and Birch, B.. Analyst, 1987, 112, 849. Moody, G. J., Saad, B. B., Thomas, J. D. R., Kohnke, F. H., and Stoddart, J. F., Analyst, 1988, 113, 1295. Feng, Y. P., Goodlet, G., Harris, N. K., Islam, M. M., and Thomas, J. D. R., Analyst, 1991, 116,469. Bartlett, J. S., Costello, J. F., Mehanic, S., Ramdas, S., Slawin, A. M. Z., Stoddart, J. F., and Williams, D. J., Angew. Chem. Znr. Ed. Engl., 1990, 29, 1404. Paper 1/01 721 G Received April 15th, 1991 Accepted May 5th, 1991