J. MATER. CHEM., 1994, 4( l), 145-151 New Fluoroionophores for Alkali-metal Cations based on Tetrameric Calixarenes Consuelo Perez-Jimenez, Stephen J. Harris and Dermot Diamond* School of Chemical Sciences, Dublin City University, Dublin 9, Ireland The synthesis and fluorescence behaviour of two novel fluorescent calix[4]arenes containing four anthracene moieties on the lower rim are described. The calixarenes exhibit an optically selective fluorescent response to complexation with group 1 metal cations. Interesting optical responses to Li', Na+ and K+ have been found with the tetraester derivative and high affinity for Na' ions with the tetraamide derivative. They may prove to be useful components for the fabrication of optical sensors for the determination of these ions.'H NMR studies enable the molecular basis for differences in the optical behaviour of the two ligands to be understood at least partially, and perhaps explain why the tetraamide calixarene has better potential as an Na' sensor material. There is currently considerable interest in the development of optical sensors for chemical analyses especially for the selective determination of clinically important species such as lithium, sodium and potassium. The inherently sensitive nature of fluorescence signalling makes it an attractive option, particu- larly when combined with a selective complexing process involving the target species.' Several groups have recently reported fluorescent alkali- metal sensors derived from macrocyclic compounds such as cyclic and non-cyclic pol yet her^,^,^ crown ethers4 and cryp- tand~.',~ Calixarenes, another major class of macrocyclic receptor molecule^,^'^ have recently evoked some interest in the area of fluorogenic sensors.Calixarene derivatives incorpo- rating ionophoric functional groups linked to the phenolic oxygen atoms exhibit excellent properties as neutral receptors for metal ions. In particular, they form stable complexes and show sharp size-related selectivity because of their well preor- ganized ionophoric group^.^-" More recently, Sat0 and Shinkai12 found very interesting luminiscence properties of modified calix [4] arenes. This led us to prepare fluorescent calixarenes13 by incorporation of aromatic fluorophores at sites adjacent to the polar ionophoric cavity.It was anticipated that the inclusion of a positively charged species in the cavity would result in sufficient perturbation of the fluorophoric groups to enable complexation to be monitored using fluor- escence, as demonstrated recently with pyrene based system^.'^ We now report the synthesis and spectroscopic study of new calix [41 arene derivatives (I and 11), containing four anthracene moieties on the lower rim, designed to combine the specific complexing ability of the tetramer and amide" series for alkali-metal ions with the photophysical behaviour of the anthracene ring.20 Experimental Instrumentation The structures of the final products and intermediates were elucidated by a variety of spectral methods.IR spectroscopy was carried out using a Perkin-Elmer 9836 infrared spectro- photometer and proton ('H) NMR spectroscopy was carried out using a Bruker AC 400 nuclear magnetic resonance spectrometer. Fluorescence spectra were recorded on a Perkin- Elmer LS-5 luminescence spectrometer equipped with a red- sensitive Hamamatsu R928 detector and a thermostatted cell compartment. Materials The calix[4]arenes I and I1 were used as reagents for Auor- escence measurements. Extra-pure-grade lithium, sodium and potassium perchlorates and thiocyanates were obtained from Aldrich. The solvents chloroform (Riedel De Haen), tetra- hydrofuran (THF)and methanol (Fluka) were used without further purification. Tetraester I and tetraamide I1 were prepared (see Scheme 1) from the known p-tert-butylcalix [4] arenetetraacetic acid.21 The tetraacid 3 was treated with thionyl chloride to afford the tetraacid chloride 4 which was treated with 9-anthra- cene methanol in THF containing pyridine to furnish the tetraester I (m.p.105.5-109.0"C) and with 1 4 I * kH2-C-OEt II I 0 2 ii OCH2-C-OH II0 3 CH2RH 0 IR=O 11 R = NCH3 Scheme 1 Reagents (i) BrCH2C02CH2CH, K2C03 in acetone; (ii) KOH, H20,EtOH, HC1; (iii) SOC1,; (iv) pyridine, THF 9-(methylaminomethy1)anthraceneto furnish the tetraamide I1 (m.p. 158.0-161.0 "C), respectively. The analyses of the structures of the products and intermediates by spectroscopic methods were found to be consistent with the predicted structures.The analytical data for the calix [41 arene deriva- tives follow. Tetraester Z Yield 65%; Found: C, 80.00; H, 6.42. Calc. for C,,,H1,,0,2: C, 81.92; H, 6.38; IR spectrum (KBr) vmaX/cm-': 1755 (C=O). 'H NMR (400 MHz; solvent CDCl,; standard TMS) dH 1.05 (36 H, s, CMe,), 2.85 (4 H HB, d, ArCH,Ar), 4.65 (4 H HA, d, ArCH2Ar), 4.70 (8H, s, OCH2C02), 6.10 (8 H, s, CH,anthr), 6.60 (8 H, s, ArH), 7.31-8.30 (36 H, m, anthracene). Tetraamide ZZ Yield 60%; Found: C, 79.31; H, 6.86; N, 3.00. Calc. for Cl16Hl,6N40,: C, 78.98; H, 6.69; N, 3.15. IR spectrum (KBr) vmaX/cm-': 1658 (C=O). NMR (400 MHz; solvent CDCl,; standard TMS) 6, 1.12 (36 H,s, CMe,), 2.60 (12 H, s, NCH,), 3.20 (4 H HB, d, ArCH2Ar), 5.05(8 H, s, OCH,CON), 5.18 (4 H HA, d, ArCH,Ar), 5.52 (8 H, s, CH,anthr), 6.83 (8 H, s, ArH), 7.20-8.40 (34 H, m, anthracene).Effect of Complexation on Fluorescence Emission Solutions of ligand I (5 x mol drn-,) and ligand I1 (5 x lop5mol drnp3) were made up in CHCI, and THF, respectively. 5 cm3 aliquots of these solutions were taken and incremental concentrations of alkali metals were added. During the work with ligand 11, access to the fluorescence instrument was restricted and samples had occasionally to be stored for several days. Routine stability checks showed that with CHC1,-based experiments, the emission spectra changed significantly suggesting that the complex had decomposed. However, this problem did not occur with THF-based experi- ments, and it was noted that there was no discernible change in the emission spectrum after storage of the complex for periods of up to one week.Furthermore, it was established that the trends in emission behaviour observed on com-plexation with both ligands were independent of whether THF or CHC1, was used as the ligand solvent, whether the source of the metal ions was SCN- or ClO,, or whether the salt was dissolved in methanol or water. Hence although different solvents were used for fluorescence experiments with ligands I and 11, valid comparisons can be made for the two sets of results. Methanolic solutions of lithium, sodium and potassium thiocyanates were used with ligand I, whereas aqueous solu- tions of metal perchlorates were added in the case of ligand 11.As mentioned above, comparative experiments with both ligands showed the same general trends whether aqueous or methanolic solutions of the metal salts were used, and as very small volumes of these salts were used as sources of the metal ions in the titration experiments, this is not unexpected. NMR experiments in all cases employed CDC1, as the solvent. The fluorescence-intensity changes of ligand solutions upon stepwise addition of alkali-metal salts were recorded from 300 to 700 nm. Quartz glass cells (1 x 1 cm2) were used for each measurement. Other measurement conditions are given in the figure legends. Effect of Complexation on 'H NMR Spectra The complexing ability of the new fluorescent calix [41 arenes synthesized was determined by 'H NMR titration experiments.The 'H NMR spectra were obtained from a 5.1 mmol dm-, solution of the ligands in CDC1, (Aldrich). The salts used for J. MATER. CHEM., 1994, VOL. 4 these experiments were thiocyanates in CD,OD (Aldrich). 'H NMR titrations were performed by adding incremental amounts of MSCN in CD,OD (M=Li+, Na', K') directly to a CDCl, solution of I or I1 in an NMR tube. Results and Discussion The fluorescence spectra of both calix [4] arene derivatives exhibit a monomer emission with a fluorescence maximum around 418.0 nm (excitation 388 nm) in the case of tetraester I, and at around 415.5 nm (excitation 263 nm) in the case of tetraamide 11. However, substantial differences appeared when we examined the effect of the optical response of the ligands to addition of alkali-metal ions.When Li' or Na+ thiocyanate (in MeOH) is added to the CHC1, solution containing I (5 x lop6 mol dm-3), the fluor- escence intensity of the entire spectrum decreases markedly with increasing salt concentration (Fig. 1A). The addition of KSCN has a different effect. As shown in Fig. lB, the maxi- mum emission (ca. 418 nm) decreases with increasing KSCN concentration and the emission at 443 nm increases, with an isoemissive point at 432 nm. In contrast, the addition of alkali-metal perchlorate (in H20) to the THF solution containing I1 (5 x lop5mol dmP3) produced, in all instances, an increase of the fluorescence intensity compared to that of the free ligand, although the addition of lithium or potassium had only a slight effect compared with sodium.As Fig. 2 shows, the fluorescence intensity increased markedly in the concentration range 10-6-10-4 mol dmP3 of NaClO,, whereas relatively minor changes were in fact observed on the addition of lithium or potassium ions. Some indication of an inner filter effect is suggested by the reduction in the fluorescence emission at the lower edge of the spectrum (380-400 nm) compared to Fig. 1, where an emission peak is obtained, but the remainder of the spectrum appears unaffected. This arises from the very large extinction coefficient for anthracene at the excitation wave- length 263 nm (log E =4.2).22 Consequently, given the concen- tration of ligand employed, the absorbance of the solution will be very high, and a large proportion of the ligand will be in the ground state at any instant in time.However, as the degree of overlap between the emission and absorption spectra is very slight, and coincides only with the lower edge of the emission spectrum (380-400nm) we are confident that the emission spectra above 400 nm are essentially unaffected. In fact, it is very probable that similar results to those described in this research could be obtained with much lower ligand concentrations. The inset in Fig. 2 shows a plot of I/Z, (I= emission intensity of the complex emission at 415.5 nm, I,= emission intensity of the free ligand at the same wavelength) us. log[Na+].It is clear that there is a sensitive response on complexation, centred at log "a+] =4.3-4.4. As the concen- tration of I1 is 5.0 x mol dm-3, this is convincing evidence in support of the formation of a 1 : 1 complex with Na' ions. The fluorescence study of the two calix [4] arenes reveals significant differences between the anthracenemethyl ester I and the anthracenemethyl amide 11, ie. the substitution of the group -COO-by -CONCH,- between the polar cavity of the calixarene and the fluorescent anthracene groups pro- vokes a drastic change of the optical response to alkali-metal cations. With the tetraester I, the difference in behaviour obtained with K+ compared to that observed with Li' and Na' is not easy to interpret. To obtain some insights into the confor- mational changes occurring in the polar cavity on com-plexation with the alkali-metal cations, we carried out 'H NMR titration experiments. In the absence of NaSCN, compound I possesses the cone conformation as evidenced by J.MATER. CHEM., 1994, VOL. 4 147 280] A c 200-c Y 2nm 2nm Fig. 1 Fluorescence spectra of I (5.00x mol dmP3) in chloroform at different concentrations of NaSCN (A) and KSCN (B). [SCN-] = (u) 0, (b)1.0x lop6,(c) 4.0 x (d) 6.0 x (e) 8.0 x loP6mol dm-3. The spectra were measured with excitation at 388 nm. Unm Fig. 2 Fluorescence spectra of I1 (5.00 x mol dmP3) in THF at different concentrations of NaC10, (a)0; (b)0.1 x lo-'; (c) 0.5 x (d) 1.0~loP5; (e) 2.0~ (f) 4.0 x lop5; (g) 6.0~lo-'; (h) 10 x lo-' mol drn-,.The spectra were measured with excitation at 263 nm. Inset shows plot of I/Z, us. [Na'], where Z=fluorescence intensity of ligand 11-Na' complex and I, =fluorescence intensity of free ligand I1 measured at 415.5nm. the splitting pattern of the ArCH,Ar methylene protons [Fig. 3(a)]. When NaSCN is added directly to the CDC1, solutions of I, the signals in the 'H NMR spectrum change greatly. With a salt :I ratio of less than 1 [Fig. 3(b)],signals for both complexed and uncomplexed ligand were present in the spectrum, indicating that on the NMR timescale, the exchange rate between the two species was slower at room temperature. Upon reaching a 1:1 stoichiometry [Fig. 3(c)] all the signals for the free ligand disappeared and an increase 8.0 6.0 4.0 210 6 Fig.3 Partial 'H NMR spectra of the fluorescent calix[4]arene I in CDCl, at 25 "C: (a)R =[NaSCN]/[I] =O; (b)R =0.5; (c) R = 1, where [I] =5.1 mmol drn-,. Aliquots from CD30D solution of 1 mcil dm-3 NaSCN were added to a CDC1, solution of I in a NMR tube. 0, 0=ArH; V,V =ArCH,Ar; El,-. =CH,anthr. in the sa1t:I ratio beyond unity produced no further spectral shifts. This finding indicates a 1: 1 stoichiometry for the NaSCN complex with I. In contrast, titration of ligand I with potassium thiocyanate (Fig.4) produced spectral changes up to the point of 1:l stoichiometry and did not show separate signals for complexed and uncomplexed ligand. As shown in Fig. 4(c), upon reaching K+ :I molar ratio of 1, the signals of the free ligand disappear and those corresponding to the K+ complex start to appear and continue to increase in intensity until a K+ :I molar ratio of 2 is reached [Fig.4(d)], whereas at the same stage with Na', all signals for the free ligand have already disappeared and an increase in the Na' :I ratio beyond unity produced no further spectral shift. Similar effects have been reported previously" and interpreted as arising from differences in the rate of formation/disassociation of the metal-ion-ligand com-plexes. In the case of potassium, the process of complex formation involves some rearrangement of the pendant polar groups in order to accommodate the larger-than-ideal size of the ion. The ILK+ complex would thus be less stable than the corresponding I-Na' complex and the net residence time of the potassium ion within the cavity shorter.Hence the I-K+ complex is not seen in the NMR spectrum until the potassium concentration is sufficiently high to enable the complex to exist for long enough to be picked up on the NMR timescale. Additionally, a comparison between the ID 1 I.,,,'I 2.0 I+6.0 4.0 6 Fig. 4 Partial 'H NMR spectra of the fluorescent calixC4larene I in CDCl, at 25 "C: (a) R= [KSCN]/[I] =O; (b) R=0.5; (c) R = 1; (d)R =2, where [I] =5.1 mmol drn-,. Symbols as Fig. 3. J. MATER. CHF.M., 1994, VOL. 4 NMR spectra of the I-metal salts reveals that the proton chemical shifts relative to the free ligand are much larger on Na' complexation than K+ (downfield shift aromatic proton: 0.47 with Na', 0.34 with K+; H, of AB methylene ArCH,Ar quartet: 0.47 with Na', 0.28 with K+ and upfield shift HA of AB methylene ArCH2Ar quartet: 0.39 with Na' and 0.25 with K').The 'H NMR data suggest that the formation of an I-Na' complex of 1:1 stoichiometry occurs. In contrast, with K', a 1:2 molar excess of ligand :metal ion is required before 100% complexation occurs. This suggests that the complexation process needs to be driven to completion by excess K+ ions. The observed quenching of fluorescence arising from com- plexation of ligand I with Lif and Na' may be the result of the I : Li+ and I :Na+ complexes being less sterically hindered than the free ligand.23 Similar perturbation differences on fluorescence have been noted by alkali-metal complexation of 2,3-naphtho-20-crown-6 (quenching) and its close relative 1,8-naphtho-21-crown-6 (enhancement).In this case, the differing fluorescence behaviour was accounted for by differ- ences in the geometrical orientation on co~nplexation.~~ As mentioned previously, the same general pattern of quenching is obtained when perchlorate ions are substituted for thiocyan- ate, or THF is used as the reaction medium in place of chloroform, or the metal salt added in aqueous instead of methanolic solution. Hence the observed quenching is not due to the anion used or the medium in which the reaction is performed, but is a function of the metal ion-ligand interaction.Enhancement of the longer wavelength emission spectrum of ligand complexed with K + suggests that complexation with potassium has a significantly different effect on the molecular conformation of the ligand than either lithium or sodium, which may simultaneously encourage certain modes of vibrational relaxation to occur (lower energy) at the expense of other higher energy modes. In the case of ligand 11, the high affinity shown for sodium ions is probably a size-related phenomenon. Thus, the cavity defined by the four amide carbonyl groups and four phenoxy oxygens best matches the size of Na+ cations. However, the molecule may allow the inclusion of the smaller Li' cation or the larger K+ cation by a flexing movement of the pendant ligating groups and/or a change in the tilt angle of the aromatic ring, but the resulting contraction or expansion of the cavity will be energetically expensive and will lead, in relative terms, to a destabilization of the complex.Support for this interpretation was obtained by the 'H NMR study of compound I1 both in the absence and presence of MSCN (M =Li', Na+, K'). The 'H NMR data established unequivocally from the characteristic Ar-CH2-Ar (AB quartet) resonances that the compound I1 and the complexes with the three cations possess the cone conformation (Fig. 5-7). The addition of incremental amounts of NaSCN in CD,OD to 5.1 mmol dm-3 solution of I1 in CDCl, affected resonances arising from the ArCH,Ar protons, the aromatic protons (0.40 downfield) and the anthracene protons (resonance at 7.20 is displaced to 7.42, the signal at 8.20 disappears and the intensity of the signal at 8.4 increases).Thus signals for both complexed and uncomplexed ligand were present in the spectrum with a sa1t:II ratio of less than 1 [Fig. 5(b)].Upon reaching a 1: 1 stoichiometry [Fig. 5(c)] the corresponding signals for the free ligand disappeared and an increase in the sa1t:II ratio beyond unity produced no further spectral changes. In contrast, titration of compound I1 with lithium or potassium thiocyanate produced new additional spectral changes apart from those observed with sodium. Thus, as Fig. 6 and 7 show the CH,anthr proton signal, which was not affected by Na+ complexation, experi- J.MATER. CHEM., 1994, VOL. 4 I, 8.0 6.0 4.0 6 Fig. 5 Partial 'H NMR spectra of the fluorescent calix [4] arene I1 in CDCl, at 25 'C: (a)R = [NaSCN]/[ 111=0 (b) R =0.5; (c) R = 1;where [II] =5.1 mmol drn-,. Symbols as Fig. 3. enced a downfield shift of 0.13 and 0.08 for K+ and Li', respectively. Similarly, the NCH, protons were shifted down- field by 0.07 and 0.13 by K+ and Li+, respectively but showed no change with Na'. In addition, in the case of Li' com-plexation alone, the OCH,CON protons were also displaced (0.25 downfield). This supports the view that the inclusion of K' and especially Li' cations inside the cavity is followed by a more extensive conformational reorganisation of the cavity region than that observed with Na' which hence has the optimum size for best fit with the polar cavity.In further support of this view is the fact that during incremental addition of both Lif and K+ salts, the presence of a 50% excess of metal salt was required for 100% formation of the new AB quartet of the calixarene-metal-ion complex (together with the disappearance of the AB quartet of the uncomplexed ligand 11). However, as a ligand I1 :ISf or Lif complex of 1:1.5 stoichiometry is highly unlikely, this prob- ably arises from an equilibrium phenomenon. Thus, excess metal ion is needed to drive complexation to completion. The same ratios are obtained from plots of I/I, us. metal-ion concentration [Fig. 2 (inset)]. From these plots, dissociation constants corresponding to the equilibrium ML+eM++L with M =Li+, Na+ and K+ and L =ligand 11, were determined + I I0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 6 Fig.6 Partial 'H NMR spectra of I1 in CDCI, at 25 "C: (u) R= [KSCN]/[II]=0.5;(b) R= l;(c) R=1.5;where [II]=5.1 mmoi dm-,. V, V, 0, 0, 0,W, as Fig. 3; 0,+=NCH,. to be 2.2~ lop5 and 1.5~lop4, 4.4~ 10-4mol dmp3, respectively. The preceding NMR data suggest that, in the case of ligand 11, the observed enhancement in fluorescence upon com-plexation with all three metal ions arises from the increased rigidity of the calixarene-metal-ion complex over that of the free ligand.23 With less freedom to interact with the surround- ing solvent molecules or for intramolecular movement, there is a corresponding increasing probability of fluorescence occurring.For the three cations studied, the largest enhance- ment was obtained with Na+, in agreement again with the NMR data which suggest that Na+ complexation confers more order on the molecule (as evidenced by the much simpler 'H NMR spectrum obtained from the complex compared with the free ligand) and least perturbation to the cavity region. In a similar way, the much smaller enhancement observed on lithium complexation may be considered a conse- quence of the higher perturbation that the inclusion of Li' causes in the polar cavity. Our results reveal that the tetraamide I1 possesses improved fluorescence behaviour over that of the tetraester I. To verify the utility of I1 as a fluorescent Na+ sensor, we have examined the effect of this ion on the fluorescence intensity in the presence of other alkali-metal ions.As shown in Fig. 8, the addition of concentrations from to 10-4mol dmF3 NaClO, causes an increase in the emission intensity which is proportional to the salt concentration. Similar behaviour to that illustrated in Fig. 8 has been also found in the presence of KClO,. To obtain further insight into the ligand I1 selec-tivity we recorded the changes in fluorescence intensity caused by the addition of incremental amounts of NaC10, to the THF solution of 11, in the presence of 10-3mol dr~i-~ of MClO, (M =Li+, K+). As Fig. 9 shows, the addition of low Na' concentration, in relation to Li' or Kf Concentration, J.MATER. CHEM., 1994, VOL. 4 --_/---Nat 820 6.7 x mot dm-38601 + 8.0 7.0 6.0 5.0 4.0 3.06 Fig. 7 Partial 'H NMR spectra of I1 in CDCl, at 25 "C: (a) R= [LiSCN]/[ I13 =0.5; (b) R =1; (c) R = 1.5; where [111=5.1 mmol dm-,. V, V, 0, 0, 0,H, as Fig.3; 0, +=NCH,; A, A= OCH,CON. Nnm Fig.8 Na' titrations of the fluorescence emission spectra of I1 (5.00x mol drnp3) in THF in the presence of lo-, mol dm-, of LiClO,. [NaClO,] =(a) 0, (b)1, (c) 4, (d) 6, (e) 10 x lop5mol dm-,. Na+ [ 5.3x 10" mol dm3li 660-1 I IIIIIIIIIII20 60 100 140 180 220 tlS Fig. 9 Effect of Na' on the fluorescence intensity of I1 (5.00x lop5mol drn-,) in THF, in the presence of lop3mol dm-3 of MC10, (M=Li+, K'). Fluorescence intensity was monitored at 415nm.Aliquots from 10-2mol dm-, of NaClO, in water were added directly to the solution of I1 and MCIO, in a cuvette. provides a marked increase on the fluorescence intensity. It should be noted that small changes in the Na+ concentration are translated into a large increase in the fluorescence intensity. This sensitivity suggests that it should be possible to make precise and accurate determinations of Na at concentrations + lower than the limit of detection of potentiometric sensor^.^^,^' These results indicate that there is little observable inter- ference, with the fluorescence intensity of I1 with sodium ions in the range 1x 10-6-1 x mol dm-3, from much higher concentrations of Li+ and K'. In addition, Fig. 9 demon- strates clearly the fast dynamics of the emission response to complexation with sodium ions, with the final emission inten- sity being reached within a few seconds after addition of the metal ions.The limiting factor for this is the rate of dispersion of the sodium ions after injection rather than the kinetics of the complexation reaction. Conclusions Two ligands which exhibit a selective fluorescent response on complexation with sodium ions have been synthesized, with the amide derivative (ligand 11) exhibiting better selectivity than the ester (ligand I). 'H NMR experiments suggest that a 1: 1 complex is formed between I1 and Na+ ions but with K+ ions, a 50% molar excess of the metal ion is required for complete reaction of the ligand.The authors gratefully acknowledge financial help for C.P.J. from the Cornissib Interpartamental de Recerca e Innovacib Tecnolbgica (CIRIT), Generalitat de Catalunya (grant BE92-244), and for S.J.H. from the Irish Science and Technology Agency (EOLAS) (grant no. SC,192/319). References 1 R. A. Bisell, A. P. de Silva, H. Q. N. Gunaratne, P. L. M. Lynch, G. E. M. Maguire and K. R. A. S. Sandanayate, Chem. SOC.Rev., 1992,187. J. MATER. CHEM., 1994, VOL. 4 K. Haratini, J. Chem. SOC., Chem. Commun., 1987, 960; K. Haratini, Analyst, 1988,113, 1065. A. P. de Silva and T. Koyama, J. Chem. SOC., Chem. Commun., 1989,1183. S. Fery-Forgues, M.-T. le Bris, J.-P. Guette and B. Valeur, J. Chem. SOC., Chem. Commun., 1988,384. F. Fages, J.-P.Desvergne, H. Bouas-Laurent, P. Marsau, J.-M. Lehn, F. Kotzyba-Hibert, A.-M. Albrecht-Gary and M. Al-Joubbeh, J. Am. Chem. SOC., 1989,111,8672. 6 F. Fages, J.-P. Desvergne, H. Bouas-Laurent, J.-M. Lehn, J. P. Konopelski, P. Marsau and Y. Barrans, J. Chem. SOC., Chem. Commun., 1990,655. 7 C. D. Gutsche, Calixarenes, Royal Society of Chemistry, Cambridge, 1989. 8 J. Vicens and V. Bohmer, Calixarenes: A Versatile Class of Macrocyclic Compounds, Kluwer Academic Publishers, Dordrecht, 1991. 9 M. A. Mckervey, E. M. Seward, G. Ferguson, B. Ruhl and S. J. Harris, J. Chem. SOC., Chem. Commun., 1985, 388. 10 F. Arnaud-Neu, E. M. Collins, M. Deasy, G. Ferguson, S. J. Harris, B. Kaitner, A. J. Lough, M. A. Mckervey, M. J. Schwing-Weill and E.M. Seward, J. Am. Chem. SOC., 1989, 111,8681. 11 J.-D. van Loon, L. C. Groenen, S. S. Wijmenga, W. Verboom and D. N. Reinhoudt, J. Am. Chem. Soc., 1991,113,2378. 12 N. Sato and S. Shinkai, J. Chem. SOC.,Perkin Trans. 2, 1993,621. 13 C. Perez-Jimenez, S. J. Harris and D. Diamond, J. Chem. SOC., Chem. Commun., 1993,480. 14 I. Aoki, H. Kawabata, K. Nakashima and S. Shinkai, J. Chem. SOC., Chem. Commun., 1992,1771. 15 T. Jin, K. Ichikawa and T. Koyama, J. Chem. SOC.. Chem. Commun., 1992,499. 16 I. Aoki, T. Sakaki and S. Shinkai, J. Chem. Soc., Chem. Commun., 1992,730. 17 A. Cadogan, D. Diamond, M. R. Smyth, M. Deasy, M. A. McKervey and S. J. Harris, Analyst, 1989, 114, 1551. 18 R. J. Forster, A. Cadogan, M. Telting-Diaz, S. J. Harris and M. A. McKervey, Sens. Actuators B, 1991,4,325. 19 A. Arduini, E. Ghidini, A. Pochini, R. Ungaro, G. D. Andreetti, G. Calestani and F. Ugozzoli, J. Incl. Phenom. Mol. Recognit. Chem., 1988,6,119. 20 H. Bouas-Laurent, A. Castellan and J.-P. Desvergne, Pure Appl. Chem., 1980,52,2633. 21 R. Ungaro, A. Pochini and C. D. Andreetti, J. Incl. Phenom. Mol. Recognit. Chem., 1989,2, 199. 22 W. Kemp, Organic Spectroscopy, Macmillan, London, 1975, p. 168. 23 H. H. Willard, L. L. Merritt Jr., J. A. Dean and F. A. Settle Jr., Instrumental Methods of Analysis, Wadsworth, Belmont, CA, 1988,199. 24 L. R. Sousa and J. M. Larson, J. Am. Chem. SOC.,1977,99,307. Paper 3/02936K; Received 24th May, 1993