134 Volume 59, Number 1, 2005 APPLIED SPECTROSCOPY0003-7028 / 05 / 5901-0134$2.00 / 0q 2005 Society for Applied Spectroscopy

Absorption and Fluorescence Spectroscopic Study onComplexation of Oxazine 1 Dye by Calix[8]arenesulfonate

MIKLOS KUBINYI,* TAMAS VIDOCZY, OLIVIA VARGA, KORNEL NAGY, andISTVAN BITTERDepartment of Physical Chemistry, Budapest University of Technology and Economics, 1521 Budapest, Hungary (M.K., O.V.);Chemical Research Center, Hungarian Academy of Sciences, 1525 Budapest, P.O.B. 17, Hungary (T.V., K.N.); and Departmentof Organic Chemical Technology, Budapest University of Technology and Economics, 1521 Budapest, Hungary (I.B.)

It has been established by combined absorption and fluorescencemeasurements that the cationic dye Oxazine 1 (OX) and the poly-valent anionic host calix[8]arenesulfonate (SCA8) form two com-plexes in simultaneous reactions: OX 1 SCA8 ↔ OX·SCA8 (1), andOX·SCA8 1 OX ↔ OX2·SCA8 (2). The equilibrium constants forthe two reactions, as functions of the ionic strength (I), and theabsorption and fluorescence spectra of the two complex species havebeen determined by a least-squares fitting method from the exper-imental data. The variations of the binding constants with the ionicstrength could be described on the basis of Debye–Huckel theory.The equilibrium constants are large; their values extrapolated to I5 0 are K(1) 5 5.5 3 106 M21 and K(2) 5 4.4 3 105 M21. Thefluorescence of OX undergoes a strong static quenching upon com-plexation. These results indicate that the complexes are held to-gether by strong electrostatic forces. The addition of non-fluorescenttetramethylammonium chloride to OX–SCA8 mixtures results in adramatic fluorescent enhancement, which demonstrates the poten-tial applicability of this supramolecular system in fluorescence as-says.

Index Headings: Absorption; Fluorescence; Supramolecular com-plex; Calixarenesulfonates; Dye probes; Partial least squares; PLS.

INTRODUCTION

The understanding of optical signals provided by fluo-rescent dye probes about their microenvironment in com-plex chemical and biological systems requires studies onthe behavior of these dyes in model systems, like mi-celles, vesicles, and complexes with macrocyclic supra-molecular hosts.1 Among these the macrocyclic hosts aremodels for nanocavities of various size, shape, and hy-drophobicity. Cyclodextrins (CDs) and calixarenesulfon-ates (SCAs) are particularly significant hosts in thesestudies, with respect to their water solubility and variablering size.2,3 For the encaged dyes the neutral CDs and themultiply charged SCAs represent two different types oflocal environment: the dye–CD adducts are held togetherprimarily by hydrophobic interaction, whereas the dye–SCA adducts are held together by electrostatic forces.4

Comparative studies on the complexation behavior ofthese two types of macrocycles concluded that cationicdyes form very stable complexes with the anionic SCAsin water. The equilibrium constants of the formation of1:1 complexes fall most often in the 105–107 M21 range,whereas the corresponding values are typically 102–104

M21 for cationic dye–CD systems.2,5,6

The determination of the equilibrium constant Ka from

Received 13 May 2004; accepted 18 August 2004.* Author to whom correspondence should be sent. E-mail: kubinyi@

mail.bme.hu.

the absorption or fluorescence spectra is relatively simple,if only one type—a 1:1 or a 2:1—host–guest complex isformed in the mixture, which is a reasonable assumptionfor the majority of dye–CD and dye–SCA systems. Thevalue of Ka can be estimated then with various versionsof the Benesi–Hildebrand method.7,8 It has been, however,pointed out that several cationic dyes form two or moredifferent types of complexes with CDs in simultaneousreactions, the equilibrium constants of which can be de-termined only by iterative methods.9,10 We performed pre-liminary experiments on the complexations of cationicphenoxazine and phenothiazine dyes by calixarenesulfon-ates, and the results suggested the occurrence of multipleequilibria in some cases. We chose one of these systems,namely, that consisting of Oxazine 1 guest and ca-lix[8]arenesulfonate host (OX and SCA8 in Fig. 1), fora detailed study involving the determination of the equi-librium constants and the spectra of the complexes. Anaspect of choosing OX as fluorescent guest was that theexpected results may also be interesting in materials sci-ence, where this cationic dye is applied as a fluorescentprobe to test nanocavities in porous solid matrices.11–14

Since the free acid forms of SCAs as well as theirsodium salts dissociate their phenolic protons in consec-utive steps,15 their inclusion properties are obviously af-fected by the pH. Thus, meaningful results could behoped primarily from experiments in which the pH wasset to a value for which one of the deprotonated formsdominated. The ionic strength of the medium was alsoexpected to influence the equilibria due to the ionic na-ture of the reactants. This relation was also investigatedin the present work.

EXPERIMENTAL

Oxazine 1 perchlorate was purchased from Sigma-Al-drich and used without further purification. The synthesisof calix[8]arenesulfonate started from p-t-butylca-lix[8]arene, which was debutylized by AlCl3 in toluene,16

and the obtained calix[8]arene was sulfonated by cc.H2SO4 into calix[8]arene-p-octasulfonic acid, which wasseparated from the reaction mixture in the form of its Nasalt using NaCl.17 The purity of both the host and theguest compound was checked by electrospray ionizationmass spectrometry, and no impurity was found in a de-tectable amount. (For the mass spectrometric measure-ments the Na ions were removed from the ca-lix[8]arenesulfonate salt by consecutive ion exchange de-salting steps using strong cation exchanger Dovex 50Wresin.)

APPLIED SPECTROSCOPY 135

FIG. 1. Chemical structures of Oxazine 1 (OX) and calix[8]arenesulfonate(SCA8).

FIG 2. (A) Absorption and (B) fluorescence spectra of OX–SCA8 mixtures at I 5 3.9 3 1023 M (water, 25 8C, , 5 1 cm). Initial concentrations:[OX]0 5 2 3 1026 M in each solution, [SCA8]0 5 (a) 0, (b) 4.7 3 1027, (c) 1 3 1026, (d) 2.2 3 1026, (e) 4.7 3 1026, (f) 1 3 1025, (g) 2.2 31025, and (h) 1 3 1024 M.

All the spectroscopic measurements were carried outat 25 8C. UV-VIS absorption spectra were recorded on aGBC Cintra 10E spectrometer. Fluorescence spectra weretaken by using a Perkin-Elmer 50B spectrofluorimeter,with the excitation wavelength set to 610 nm.

The distortion in the fluorescence spectra, caused byself-absorption, was corrected by using the approximateformula18

[{A 1A )/2]ex emF 5 F ·10 (1)md

where Fmd is the measured fluorescence intensity, and Aex

and Aem are the absorbances at the wavelengths of exci-tation and emission, respectively.

RESULTS AND DISCUSSION

Reaction Scheme and Equilibrium Compositions. Inaqueous solutions the Na salt of SCA8 dissociates fourphenolic protons in consecutive steps, for which the cor-responding pKa values are 3.73, 4.39, 8.07, and 10.1.19

The pH was set to 6.5 in our experiments, where thedoubly deprotonated form of SCA8 is predominant.

The reaction scheme, the values of the correspondingequilibrium constants, and the spectra of the supramolec-ular complexes were determined by evaluating the ex-perimental spectra of several series of solutions. In eachseries the total concentration of OX and the ionicstrength, I([1/2Sciz ) were held at constant values, and2

i

the total concentration of SCA8 was varied. To avoid theuncertainties arising from the presence of various metal

ions, the phosphate buffer medium was prepared fromNa2HPO4 and NaH2PO4, and the ionic strength was setby adding NaCl to the solutions. The results are illus-trated in Fig. 2, in which the absorption and fluorescencespectra obtained from one series of samples, with I 5 3.93 1023 M, are displayed.

The absorption spectra of the individual series con-tained no isobestic points (this can also be seen in Fig.2), which is a clear indication that OX occurs in at leastthree different forms in the samples. The quantitativeanalysis of the spectra led to the conclusion that the hostand the guest form two complexes, OX·SCA8 andOX2·SCA8, in simultaneous reactions:

K1

OX 1 SCA8 o OX·SCA8 (2a)K2

OX·SCA8 1 OX o OX ·SCA8 (2b)2

The systems are characterized by the apparent equilib-rium constants

[OX·SCA8]K 5 and (3a)1 [OX]·[SCA8]

[OX ·SCA8]2K 5 (3b)2 [OX·SCA8]·[OX]

The balance equations for the reactants are

[OX] 5 [OX] 1 [OX·SCA8] 1 2·[OX ·SCA8] (4)0 2

and

[SCA8] 5 [SCA8] 1 [OX·SCA8] 1 [OX ·SCA8] (5)0 2

where [OX]0 and [SCA8]0 denote total concentrations. Ifthe values of K1 and K2 are known, the equilibrium con-centrations of the two reactants, [OX] and [SCA8], andof the two products, [OX·SCA8] and [OX2·SCA8], canbe computed from the system of Eqs. 3a, 3b, 4, and 5for any OX–SCA8 mixture. The algorithm is describedin the Appendix.

Procedure for Evaluation of Spectra. In the visiblerange, where the absorption of SCA8 is negligible, the

136 Volume 59, Number 1, 2005

FIG 3. (A) Absorption and (B) fluorescence spectra of OX and its SCA8 complexes. The fluorescence spectrum of OX·SCA8 is multiplied by10. The complex OX2·SCA8 is dark.

TABLE I. Spectral bands in the absorption and fluorescence spec-tra of OX and its complexes with SCA8.

CompoundAbsorption spectrum

lmax[nm] («max [M21 cm21])Fluorescence spectrumlmax [nm] (wmax [M21])

OXOX·SCA8OX2·SCA8

654 (1.083 3 105)667 (1.097 3 105)605 (1.074 3 105)672 (6.98 3 104)

671 (5.32 3 108)680 (1.43 3 107)

dark complex

absorbance of the ith sample at the jth wavelength canbe given as the sum of the contributions of the uncom-plexed and the two complexed forms of OX:

j jjA (cacd) 5 (« ·[OX] 1 « ·[OX·SCA8]i i iOX OX·SCA8

j1 « ·[OX ·SCA8] )·, (6)·SCA8 2 iOX2

where « , « , and « denote the absorptionj j jOX OX·SCA8 OX ·SCA82

coefficients of the respective species, and , is the lengthof the optical path.

The absorption coefficients of OX were obtained di-rectly from the spectrum of the pure dye. The absorptioncoefficients of OX·SCA8 and OX2·SCA8 as functions ofwavelength and the values of the two equilibrium con-stants were determined by least-squares fitting calcula-tions, in which one series of solutions, prepared with thesame ionic strength, was always involved. In each stepthe equilibrium concentrations of the uncomplexed andcomplexed forms of OX were computed from Eqs. 3a,3b, 4, and 5 using hypothetical values for K1 and K2.Substituting the obtained concentrations, the measuredabsorption coefficients for OX, and the hypothetical ab-sorption coefficients for OX·SCA8 and OX2·SCA8 intoEq. 6, the spectra of all the solutions belonging to thesame series were computed. The values for K1, K2

« and « were systematically varied as longj jOX·SCA8 OX ·SCA82

as the absolute error for the selected series, defined asj j 2S 5 [A (cacd) 2 A (exp)] (7)O O i i

i j

reached its minimum. Finally, the weighted average ofthe spectra obtained for OX·SCA8 and OX2·SCA8 wereaccepted as the real spectra of these species, and one

more fitting was carried out for each series of samples,in which only K1 and K2 were varied.

Following this final step, the agreement between thecalculated and the experimental spectra was very good:the average difference was 0.002 absorbance units. Therelative errors of the final K1 and K2 values were virtuallybelow 610%; however, it must be taken into account thatthe two equilibrium constants and the spectra ofOX·SCA8 and OX2·SCA8 are heavily correlated. Thus,we estimate the accuracy of K1 to be ;630% and of K2

to be 650%.The fluorescence spectra were also expressed as sums

of the contributions of OX, OX·SCA8, and OX2·SCA8:j j jF (cacd) 5 w ·[OX] 1 w ·[OX·SCA8]i OX i OX·SCA8 i

j1 w ·[OX ·SCA8] (8)OX ·SCA8 2 i2

The indices i and j refer again to the samples of a serieswith a fixed ionic strength and to the wavelength, re-spectively, and w , w , and w denote molarj j j

OX OX·SCA8 OX ·SCA82

fluorescence coefficients. The latter are proportional tothe absorption coefficients and to the quantum yields ofthe individual components,20 e.g., for OX:

w 5 KjI « dFj j j jOX 0 OX OX (9)

where Kj is a function of the instrumental response factorsand the geometry, I is the intensity of the excitationj

0

beam, d is the penetration depth, and F is the fluores-jOX

cence quantum yield of the dye.The fluorescence spectrum of uncomplexed OX was

known from the experiments, the spectra belonging toOX·SCA8 and OX2·SCA8, and the values of K1 and K2

were parameters in the least-squares fitting. In all respectsthe evaluation process of the fluorescence data was sim-ilar to that applied in case of the absorption data.

Absorption and Fluorescence Spectra of OX–SCA8Complexes. The absorption and fluorescence spectra ofOX and its two complexes are displayed in Fig. 3, andthe spectral data are summarized in Table I.

The shape of the absorption spectrum of OX·SCA8 israther similar to that of pure OX, with its maximum shift-ed to the red. The formation of 1:1 adducts from otherfluorescent guests, like Crystal Violet,2 Acridine Red,5 N-

APPLIED SPECTROSCOPY 137

FIG 4. (A) Variations of the fluorescence spectra of an OX–SCA8 mixture upon the addition of tetramethylammonium-chloride (TMACl ) at I 51 3 1022 M (water, 25 8C, , 5 1 cm). Initial concentrations: [OX]0 5 2 3 1026 M, [SCA8]0 5 2.2 3 1026 M in each solution, [TMACl]0 5 (a)0, (b) 5 3 1025, (c) 1 3 1024, (d) 2 3 1024, (e) 5 3 1024, (f) 1 3 1023, (g) 2 3 1023, (h) 3 3 1023, (i) 4 3 1023, and (j) 5 3 1023 M. (B)Fluorescence intensities at 671 nm.

methylpyridil substituted porphyrin,6 Auramine O,21 Bril-liant Cresyl Blue,22 and calixarenesulfonate hosts, are ac-companied by similar spectral variations.

The absorption spectrum of OX2·SCA8 contains twoseparated bands with comparable intensities. This is aconsequence of the splitting of the lowest energy S0→S1

transition, which is a common feature of dye dimers. Werecorded the spectrum of a 1 3 1024 M OX solution, andobserved the emergence of a second band around 610nm, as a sign of dimer formation. The spectrum of Meth-ylene Blue (the thia-analogue of OX) exhibits this featureat much lower concentrations.23,24

The fluorescence intensity of OX·SCA8 at its bandmaximum is about forty times lower as compared to thecorresponding value of the pure dye. The 2:1 complex ispractically dark. The reactions of the above-mentionedother cationic dyes with SCAs are also accompanied bydramatic fluorescence reduction.2,5,6,21,22

Some fluorescent dye–SCA complexes may have an-alytical applications, e.g., in the determination of non-fluorescent materials by competitive assay (via displace-ment of the fluorescent dye in the complex by a non-fluorescent analyte.) For testing our system from this as-pect, tetramethylammonium chloride (TMACl) waschosen as a non-fluorescent guest. The spectral variationsupon the addition of this compound to a selected OX–SCA8 system are shown in Fig. 4. As could be expected,the cation of TMACl gradually expels the dye from its1:1 complex, and the release of the dye is indicated by alarge fluorescence enhancement, demonstrating the sen-sitivity of the method.

Equilibrium Constants. For the discussion of the ef-fect of the ionic strength, the equilibrium constant forreactions 2a and 2b were expressed in the general form

log K 5 log K0 1 log gG·H 2 log gG 2 log gH (10)

where K is the apparent equilibrium constant, K0 is the‘‘true’’ equilibrium constant, and gG, gH, and gG·H denotethe activity coefficients of the guest with charge zG, the

host with charge zH, and the complex with charge zG 1zH, respectively. (H is SCA8 in the first reaction andOX·SCA8 in the second one).

In the special case when zG 5 zH, i.e., the complexformed is neutral, gG·H ø 1, whereas gG and gH can bereplaced by the mean activity coefficient gG 5 gH 5 g6.The variation of g6 with the ionic strength can be de-scribed by the extended Debye–Huckel equation:25

1/2AIlog g 5 2zz z z (11)6 G H 1/21 1 BaI

where A 5 0.5112 and B 5 0.3291 for aqueous solutionsat 25 8C, and a is the radius of the ionic atmosphere inA. At low ionic strengths BaI1/2 in the denominator canbe neglected, and in other cases the assumption Ba ø 1may be used. Substituting Eq. 11 into Eq. 10 yields

1/2AI0log K 5 log K 2 2 zz z z (12)G H 1/21 1 BaI

This means that the log K vs. I1/2 (or log K vs. I1/2/(1 1I1/2)) function is expected to be linear with a slope of22zzGzHz. This treatment was in accordance with the ex-perimental findings in the case of reactions betweenopen-chain a,v-dianions and a,v-dications, where the re-sulting supramolecular ion pair had zero net charge.26

The equilibrium constants for the binding of aromaticcarboxylate ions by a polycationic azonia cyclophane27

and for the binding of cationic amino esters28 and cationicdyes29 by large polyanionic porphyirin receptors—reac-tions with charged products, like our case—were alsofound to exhibit linear variation on log K vs. I1/2 or logK vs. I1/2/(1 1 I1/2) plots. The slopes, however, could notbe predicted solely by theoretical considerations. The dif-ficulties were due primarily to the fact that only a portionof the charged functional groups of the hosts took part inthe binding process. In those cases the slopes were relatedto the ‘‘effective charges’’ of the interacting ions.28,29

Adapting this approach to our system:

138 Volume 59, Number 1, 2005

FIG 5. Debye–Huckel plots of equilibrium constants K1 and K2 calculated from the (A) absorption and (B) fluorescence data. I1/2/(1 1 BaI1/2) isdenoted by SI in the equations.

1/2AI0log K 5 log K 2 2 zz9 z9 z and (13a)1 1 OX SCA8 1/21 1 BaI

1/2AI0log K 5 log K 2 2 zz9 z9 z (13b)2 2 OX OX·SCA8 1/21 1 BaI

where the z9 symbols stand for the effective charges.The log K1 and log K2 values obtained from the ab-

sorption and fluorescence measurements, as functions ofI1/2/(1 1 BaI1/2), are displayed in Fig. 5, where the valueof Ba ø 2 was assumed for our large organic ions. Theequations of the fitted trend lines are also given in thefigure. As can be seen, all four sets of data fit suitablyon a linear function. As could be expected, the uncer-tainties of the K2 values are relatively high, since thecomplex OX2·SCA8 occurred at low concentrations inmost of the samples. The intercepts of the lines derivedfrom the absorption data correspond to the values of K0

1

5 5.5 3 106 and K 5 4.4 3 105. Providing that the02

effective charge of OX, z 5 11, the effective charges9ox

of z 5 26.3 and z 5 25.2 are obtained from9 9SCA8 OX·SCA8

the slopes of these two lines. The equilibrium constantsand effective charges obtained from the fluorescence datashow a reasonable agreement with the above results, con-firming the validity of the model applied.

CONCLUSION

Our combined absorption and fluorescence measure-ments revealed that the cationic dye OX forms two stablecomplex species with the large-size anionic host SCA8in aqueous solution. There are several indications that thetwo complexes, OX·SCA8 and OX2·SCA8, are held to-gether by strong electrostatic forces: (1) the two bindingconstants, K1 and K2, are large, their values are ;103

times higher than that of the OX–b-CD 1:1 complex(1400 M21);30 (2) the variations of K1 and K2 with theionic strength can be described on the basis of the De-bye–Huckel theory; and (3) the fluorescence of OX un-dergoes a strong static quenching upon complexation.

It is more difficult, however, to suggest hypotheticalstructures for the complexes, since in aqueous solutionSCAs occur as mixtures of different conformers, whichare converting into each other at high rates.31 It may bepresumed that the binding of one OX cation stabilizes

the cone conformation. Considering the dimensions ofOX (18.3 3 9.3 3 5.6 A)30 and the cavity diameter ofSCA8 (;16 A),32 it can be concluded that in such a 1:1complex the guest is not incorporated entirely into thecavity. The OX has a protruding part carrying one of thepositively charged diethylamino groups, which interactswith the negatively charged sulfonate units of the hostelectrostatically. For the 2:1 complex a double partialcone conformation of SCA8 may be sterically favorable,in which the two half-cavities bind the two guests sepa-rately. Such arrangement has been proposed by Shinkaiet al.33 for the 2:1 complex of trimethyl-anilinium cationwith SCA8 on the basis of nuclear magnetic resonance(NMR) measurements.

ACKNOWLEDGMENT

This work was supported by grant T 42546 from the Hungarian Re-search Foundation.

1. J. R. Lakowicz, Principles of Fluorescence Spectroscopy (KluwerPress, New York, 1999).

2. Y. Liu, B.-H. Han, and Y.-T. Chen, J. Phys. Chem. B 106, 4678(2002).

3. Y. Liu, B.-H. Han, and H.-Y. Zhang, Curr. Org. Chem. 8, 35 (2004).4. R. Castro, L. A. Godinez, C. M. Criss, and A. E. Kaifer, J. Org.

Chem. 62, 4928 (1997).5. Y. Liu, B.-H. Han, and Y.-T. Chen, J. Org. Chem. 65, 6227 (2000).6. K. Lang, P. Kubat, P. Lhotak, J. Mosinger, and D. M. Wagnerova,

Photochem. Photobiol. 74, 558 (2001).7. S. M. Hoenigman and C. E. Evans, Anal. Chem. 68, 3274 (1996).8. P. Rodriguez, M. Sanchez, J. R. Isasi, and G. Gonzales-Gaitano,

Appl. Spectrosc. 56, 1490 (2002).9. Y. L. Loukas, J. Phys. Chem. B 101, 4863 (1997).

10. C. Lee, Y. W. Sung, and J. W. Park, J. Phys. Chem. B 103, 893(1999).

11. M. C. Marchi, S. A. Bilmes, and R. M. Negri, Langmuir 13, 3665(1997).

12. F. del Monte, M. L. Ferrer, and D. Levy, J. Mater. Chem. 11, 1745(2001).

13. H. Maas, A. Khatyr, and G. Calzaferri, Micropor. Mesopor. Mater.65, 233 (2003).

14. C. Seebacher, C. Hellriegel, C. Brauchle, M. Ganschow, and D.Wohrle, J. Phys. Chem. B 107, 5445 (2003).

15. K. Suga, T. Ohzono, M. Negishi, K. Deuchi, and Y. Morita, Supra-mol. Sci. 5, 9 (1998).

16. V. Bocchi, D. Foina, A. Pochini, R. Ungaro, and G. D. Andreetti,Tetrahedron Lett. 38, 383 (1982).

17. S. Shinkai, S. Mori, H. Koreishi, T. Tsubaki, and O. Manabe, J.Am. Chem. Soc. 108, 2409 (1986).

APPLIED SPECTROSCOPY 139

18. J. R. Lakowicz, Principles of Fluorescence Spectroscopy (KluwerPress, New York, 1999), p. 54.

19. M. Sonoda, K. Hayashi, M. Nishida, D. Ishii, and I. Yoshida, Anal.Sci. 14, 493 (1998).

20. L. B. McGown and K. Nithipatikom, Appl. Spectrosc. Rev. 35, 353(2000).

21. Y. Zhang, R. A. Agbaria, and I. M. Warner, Supramol. Chem. 8,309 (1997).

22. Y. Zhang, T. H. Pham, M. Sanchez Pena, R. A. Agbaria, and I. M.Warner, Appl. Spectrosc. 52, 952 (1998).

23. M. K. Carroll, M. A. Unger, A. M. Leach, M. J. Morris, C. M.Ingersoll, and F. V. Bright, Appl. Spectrosc. 53, 780 (1999).

24. G. Zhang, S. Shuang, C. Dong, and J. Pan, Spectrochim. Acta, PartA 59, 2935 (2003).

25. G. Kortum, Treatise on Electrochemistry (Elsevier, Amsterdam,1965), Chap. V, p. 176.

26. M. A. Hossain and H.-J. Schneider, Chem. Eur. J. 5, 1284 (1999).27. H.-J. Schneider and I. Theis, J. Org. Chem. 57, 3066 (1992).28. T. Mizutani, K. Wada, and S. Kitagawa, J. Am. Chem. Soc. 121,

11 425 (1999).29. K. Wada, T. Mizutani, H. Matsuoka, and S. Kitagawa, Chem. Eur.

J. 9, 2368 (2003).30. N. Balabai, B. Linton, A. Napper, S. Priyadarshy, A. P. Sukharev-

sky, and D. H. Waldeck, J. Phys. Chem. B 102, 9617 (1998).31. J. Alvarez, Y. Wang, W. Ong, and A. E. Kaifer, J. Supramol. Chem.

1, 269 (2001).32. M. Nishida, D. Ishii, I. Yoshida, and S. Shinkai, Bull. Chem. Soc.

Jpn. 70, 2131 (1997).33. S. Shinkai, K. Araki, and O. Manabe, J. Am. Chem. Soc. 110, 7214

(1988).

APPENDIX

Calculation of the Equilibrium Concentrations. Theequilibrium concentrations can be expressed from Eqs.3a, 3b, 4, and 5 as

[OX] 5 [OX] 2 [OX·SCA8]0

2 2·[OX ·SCA8] (14)2

[OX·SCA8] 5 K ·[OX]·[SCA8] (15)1

2[OX ·SCA8] 5 K ·K [OX] ·[SCA8] and (16)2 1 2

[SCA8]0[SCA8] 5 (17)21 1 K ·[OX] 1 K ·K ·[OX]1 1 2

A substitution of the latter formula for [SCA8] into Eqs.15 and 16, and substituting the expressions obtained sofor [OX·SCA8] and [OX2·SCA8] into Eq. 14 leads to athird-order equation for [OX]. Thus, the equilibrium con-centration of OX can be calculated first, which is fol-lowed by the calculation of [SCA8] from Eq. 17, andfinally [OX·SCA8] and [OX2·SCA8] are obtained fromEqs. 15 and 16, respectively.