review zeolite photochemistry: impact of zeolites on...

Journal of Photochemistry and Photobiology C: Photochemistry Reviews4 (2003) 19–49

Review

Zeolite photochemistry: impact of zeolites on photochemistryand feedback from photochemistry to zeolite science

Shuichi Hashimoto∗Chemistry Department and Advanced Engineering Courses, Gunma College of Technology, 580 Toriba-machi, Maebashi, Gunma 371-8530, Japan

Received 23 September 2002; received in revised form 3 February 2003; accepted 3 February 2003

Abstract

The past decade has witnessed increasing activity in photochemistry within zeolites. This article critically reviews the topics of suchactivities and provides some comments for future studies. Particular emphasis is placed on transient spectroscopic studies, which haveshown to be a powerful technique for mechanistic investigations in solid systems. As a host material for organic species, zeolites areanother promising candidate for modifying the photophysics and photochemistry of a given species. Besides, photochemical reactionscan be pursued in zeolites and provide product distributions considerably different from those in solutions. Accordingly, photochemistryhas enjoyed a great benefit from zeolite materials and zeolite science. In return, photochemical techniques can provide pieces of usefulinformation on the fundamental issues of zeolite science such as adsorption and diffusion of molecules within the zeolite. Thus, we canexpect further development in this area if this cooperative relationship continues. In this article, we mainly describe mechanistic aspect ofphotochemical reactions. Issues associated with experimental techniques are also reviewed.© 2003 Japanese Photochemistry Association. Published by Elsevier Science B.V. All rights reserved.

Keywords:Zeolite; Photochemistry; Adsorption; Diffusion; Transient

Contents

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202. Materials, property and methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.1. Structure and chemical properties of zeolites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.2. Adsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.3. Diffusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.4. Sample preparation protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.5. Diffuse reflectance spectroscopy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3. Molecular photophysics and photochemistry in zeolites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.1. Confinement effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.2. Interaction with cationic sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.3. Interaction with the zeolite framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4. Intracage reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.1. Excimer and dimer cations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.2. Electronic energy transfer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384.3. Charge transfer and electron transfer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

5. Reactions through intercage migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

∗ Tel.: +81-27-254-9281; fax:+81-27-254-9022.E-mail address:[email protected] (S. Hashimoto).

1389-5567/03/$ 20.00 © 2003 Japanese Photochemistry Association. Published by Elsevier Science B.V. All rights reserved.doi:10.1016/S1389-5567(03)00003-0

20 S. Hashimoto / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 4 (2003) 19–49

1. Introduction

Zeolites have been an object of scientific research and amaterial beneficial to mankind for more than 200 years sincetheir discovery in 1756. However, it was not until 10–15years ago that zeolites attracted the keen interest of photo-chemists who wanted to use them for their research. Sincethen, an active interplay of zeolites with photochemistry hasstarted[1–11]. Photochemists are most interested in control-ling chemical reactions with the aid of zeolites and othersupramolecular assemblies, aimed at constructing artificialphotosynthetic systems, controlling chirarity and inventingnanoscale advanced materials. Zeolites are found to be par-ticularly useful for such purposes since they host variousmolecules including organics in their cavities and channels,and such inclusions have often been shown to modify thenormal solution photochemistry of a given species. The con-finement of molecules within the constrained space and thecatatytic activity of the surface adsorption sites were con-sidered decisive for product distributions unique to zeolites.On the other hand, zeolite science covers a broad spectrum,from synthesis to structural characterization, adsorption andtransport phenomena, and applications such as catalysts andother classical uses (adsorbents or desiccants, water soft-eners, soil improvers, and so forth)[12–14]. Furthermore,zeolites have been recognized as new materials for opticaland electronic applications, such as lasers, quantum dots andsecond harmonic generation, in recent years[15–18]. Thesedays, zeolite science as materials science is getting closerand closer to photochemistry than ever before.

Despite their high interest and anticipation, many photo-chemists who have worked in the gas phase, in transparentsolutions including micelles and colloids, and on single crys-tals are still skeptical about zeolites and hesitate to utilizethem for their research mainly because of distrust of the re-liability of the systems as well as that of the spectroscopiesto deal with them. The following questions are frequentlyasked. How much are we sure about the postulate that thelocation of adsorbed molecules is inside the cages and chan-nels of zeolites? Can you obtain a homogeneous distributionof guest molecules among the cavities rather than a hetero-geneous one? Are the transient absorption spectra obtainedby a diffuse reflectance technique reliable enough? Zeolitesare, of course, not an ideal system perfectly suited for con-trolling chemical reactions; rather, they bear some draw-backs, but we believe there are some ways to utilize them totheir maximum capability for our purposes if we are fullyaware of their physical and chemical properties. Along thisline, we try to relieve the uneasiness brought about from theskepticism of many photochemists by carefully examiningthe experiments carried out with great efforts to show thatzeolites are certainly usable even for sophisticated research.

This article critically reviews the photochemical studieswithin zeolites and summarizes the progress made duringthe past 10 years with the goal of providing useful infor-mation to photochemists regarding the advantages and lim-

itations of zeolites for photochemical use. Particular em-phasis is placed on the mechanistic studies where the tran-sient absorption spectroscopic method, the most powerfulway to look at reaction mechanisms, is employed, and thecurrent understanding to the field will be presented. Addi-tionally, consideration is given to issues to overcome in thenear future. Although a number of review articles dealingwith photochemical studies in zeolites have been publishedin the past few years[1–11], most articles cover rather spe-cific subjects such as organic photochemistry, charge trans-fer (CT), confinement, the antenna effect, radical reactions,etc. We try to organize diverse subjects relevant to zeolitesfrom the mechanistic point of view revealed by using thetransient spectroscopy.

2. Materials, property and methodology

2.1. Structure and chemical properties of zeolites

Zeolites are microporous crystalline aluminosilicate ma-terials, with alkali or alkaline earth metals as counterions[12–14]. As pictorially represented inFig. 1, they consist ofa framework of [SiO4]4− and [AlO4]5− tetrahedra, linkedto each other at the corners by sharing their oxygens. Thisleads to the following general formula:

(M+)x [(AlO2−)x(SiO2)y ] · mH2O

Fig. 1. Framework structure of sodalite-based zeolites.

S. Hashimoto / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 4 (2003) 19–49 21

In this formula, M+ represents the alkali metal cation. Thetetrahedra make up three-dimensional network structures toform channels and cages or cavities of discrete size.

In aluminum-containing zeolites, negative charges gen-erated in the framework originating from the differencein charge between the [SiO4]4− and [AlO4]5− tetrahedramust be balanced by the charge-compensating cations.These cations can generally be exchanged by conventionalmethods of ion exchange. The cations and water moleculespresent are located in the cages (cavities) and channels.Several types of zeolites are frequently employed in pho-tochemical research: their common names are X, Y, A,mordenite, L, ZSM-5, silicalite, MCM-41, to name a few.In the crystallographical classification[19], X and Y belongto FAU (faujasite), A to LTA, L to LTL, and ZSM-5 andsilicalite to MFI. These three-letter framework type codeswere assigned by the Structure Commission of the Interna-tional Zeolite Association (http://www.iza-structure.org/).The chemical composition of zeolites is not necessarilylimited to the aluminosilicate: aluminophosphate-based ze-olites with the AlPO4 lattice and molecular sieves composedof other elements have been synthesized and exhibited cat-alytic properties distinct from those of aluminosilicate-basedzeolites[14].

The structure of faujasite zeolites (zeolites X and Y)is cubic and built from sodalite cages (�-cages, 0.66 nmin diameter, with an entry aperture of 0.21 nm) connectedvia the double 6-membered ring (D6R) frame. Note thatthe entry aperture of the sodalite cage is too small for theoxygen molecule to enter; however, water molecules areknown to go in. Zeolites X and Y have different frame-work Si/Al ratios: 1.0 < Si/Al < 1.5 for zeolite X and1.5 < Si/Al < 3 for zeolite Y. Interestingly, no faujasitewith Si/Al ratio of less than 1.0 has been prepared to datebecause of the unstable framework structure (Loewenstein’srule [20]). These zeolites form a three-dimensional networkof nearly spherical supercages of about 1.3 nm in diame-ter connected tetrahedrally through 0.74 nm windows (di-amond lattice). The supercage concentration in zeolite Y,with Na+ ions as the charge-compensating cation (Na+-Y),

Fig. 2. Framework structure of zeolite L.

is estimated to be approximately 6× 10−4 mol/g on the ba-sis of the crystal structure. The charge-compensating cationsare mobile and distributed among several types of sites:SI (both in X and Y) in the center of the D6R cage, SI ′(both in X and Y) inside the sodalite cage, SII (both inX and Y) on top of the hexagonal faces of the sodalitecages facing into the supercage and SIII (only in X) onthe four-ring on the supercage wall. Only cations at SIIand SIII are accessible to the organic molecules adsorbedwithin a supercage. A de-aluminated zeolite[21,22] canbe prepared through various post-synthesis modifications.Ultra-stable zeolite Y (USY), for example, de-aluminatedby hydrothermal treatment and subsequent extraction of theextra-framework aluminum, has fewer charge-neutralizingcations, resulting from the decrease in the aluminum con-tent and consequently provides less polar environment insupercage networks than Al-containing zeolites do.

Zeolite A is a synthetic material, the structure of which isalso composed of sodalite cages similar to faujasite but con-nected through double 4-membered ring (D4R) of [SiO4]4−and [AlO4]5−. By this connection, three cages are present:D4R, sodalite cage, and�-cage (1.1 nm in diameter witha 0.41 nm opening). The structure of zeolite A is also cu-bic, having a Si/Al ratio of 1.0, with alternating Si and Alatoms in the framework. Aromatic guest species are hardlyadsorbed inside the�-cage because of size restrictions, andonly external surfaces are available for the adsorption ofsmall quantities.

Pentasil (5-membered ring) units as building blocks, incontrast to the sodalites, constitute channel-type zeolitessuch as ZSM-5 and mordenite. ZSM-5 and its high-silicaanalogue silicalite possess a structure consisting of two in-tersecting channel systems, with the vertical channel be-ing elliptical (0.57 mm× 0.51 nm), whereas the horizontalchannels have nearly circular 0.54 nm× 0.56 nm openings.On the other hand, the channel system in mordenite is aone-dimensional 12-membered ring system with an openingof 0.65 nm×0.70 nm. Of other channel-type zeolites, zeoliteL has a unidimensional channel system as in the mordenite,formed by 6- and 4-membered rings (Fig. 2).

http://www.iza-structure.org/


The main channels of zeolite L are made by the stackingof sections with a length of 0.75 nm in thec-direction. Thesections are joined by shared 12-membered ring windowshaving a free diameter of 0.71 nm. These rings make upthe narrowest parts of the main channel. The largest freediameter is about 1.3 nm and lies midway between the12-membered rings. In zeolite L, five positions are found for

the extra-framework cations, i.e. sites A, B, C, D and E. Onlythe cations in site D are exchangeable at room temperature.Additionally, the synthesis of mesoporous (extra-large-pore)channel-type aluminosilicates, which allow the inclusion ofmolecules larger than that possible with faujasite zeolites,has been reported recently[23]. The crystalline aluminosil-icate MCM-41 is formed from an array of parallel hexago-nal channels whose diameter can be controlled from 2.0 to6.0 nm by varying a template during its synthesis.

These zeolites present an interesting contrast in topol-ogy, as well as in the sizes of pores or channels and oftheir openings from an entrapment point of view—adsorb-ing molecules inside the zeolites in a size-selective,shape-selective manner.

Zeolite can be regarded as a solvent thatdissolvesor dis-perses molecules into pores and channels, similar to solventcages[4]. The solvation-like interaction can be expectedfor zeolite host–guest molecule pairs, and the negativelycharged framework, and the mobile cations combine to pro-duce an electrostatic field akin to solvent polarity inside thecavities where the molecules reside. The electrostatic fieldstrength (polarity) in zeolites has been suggested to be ex-tremely high[24–26]and is considered to be due to the factthat the cations exposed at the center of the supercage beingonly partially shielded. The electric field strength is depen-dent both on cation size and on Si/Al ratio. For example, thesmall Li+ ion induces a stronger field in its proximity thanthe larger Cs+ ion. Also, cations in zeolite Y exhibit higherfields than those in zeolite X. The effects of the fields, beingakin to those of solvent polarity, have been explored witha number of fluorescence probes incorporated into zeolites[27–31]. Despite the similarity of zeolite pores to solventshells, zeolite pores are rigid and distinctly shaped, in con-trast to the soft and featureless solvent shells. Examinationof physical and chemical properties characteristic to the ze-olite is appropriate to gain some insight into the active roleof the host material.

Zeolites have amphoteric properties, and the existenceof acid and base sites is well documented[32–37]. The

bridging Si–OH· · · Al group is usually referred to as aBrønsted acid site in zeolites. On the contrary, no basicframework OH group has been reported to exist. Zeo-lites contain Lewis as well as Brønsted acid sites. Thethree-coordinated aluminum sites on the framework andnon-framework Al sites are normally considered to be Lewissites.

Additionally, charge-compensating cations act as Lewisacids, while the framework oxygens represent a base. In par-ticular, the oxygens adjacent to Al (Si–O–Al oxygens) aremore basic because of a larger negative charge on oxygen.

The Lewis acidity is usually connected to an electron-accepting property, and the basicity to an electron-donatingproperty. Thus, the zeolites behave both as electron donorsand as acceptors of moderate strength to the guest species,depending on the adsorption site. As a rule, the acid strengthor the electron-accepting ability increases with increasingSi/Al ratio and for zeolites with smaller alkali metal cations.For example, Li+-Y is more acidic than Cs+-X. On theother hand, the basic strength (electron-donating ability)of the zeolites increases with decreasing Si/Al ratio andfor zeolites with larger alkali metal cations. In this regard,Cs+-X is more basic than Li+-Y. Accordingly, the chemicalproperties of the zeolites can be fine-tuned in terms of com-position, cations and framework structure. The resultingproperties can be assessed on the basis of Sanderson’s elec-tronegativity: the electronegativity of the zeolites calculatedaccording to the Sanderson’s equalization principle[38]has been used as a good measure of the electron-acceptingability of zeolites. Likewise, Sandarson’s partial charges ofthe framework oxygen atoms serve as viable measures ofthe framework donor strengths[39]. More recently, an alter-native computational approach to the basicity of frameworkoxygens has been developed based on quantum chemicalcalculations[40].

The remarkable structural features and physicochemicalproperties of zeolites are expected to provide new aspectsof photophysical and photochemical processes that can bemonitored by employing transient spectroscopy and othertechniques.


2.2. Adsorption

Adsorption is one of the fundamental issues in zeolitescience[12–14]. Fig. 3represents an adsorption isotherm ofN2 in zeolite Na+-Y (Tosoh Lot. #3001) at 77 K measuredin our laboratory.

As is clear from the figure, adsorption of N2, whichis easily incorporated into the supercages, occurs in twostages with increasing equilibrium pressure. The initial, verysteeply rising part implies that N2 is being adsorbed into thesupercages, and the following flat part, where the adsorptiontakes place very slowly with increasing pressure, accountsfor N2 molecules covering the outer surfaces of the zeoliteat a monolayer level. In this case, more than 90% of theN2 molecules were adsorbed in the supercages. This curvegives a basic idea of why zeolite scientists think a priori thatadsorption takes place mostly in the supercages and that themolecules first go into the supercages. A simple explana-tion for the result is that most of the molecules are adsorbedinto the supercages, because the volume of the supercagesdominates, and the adsorption takes place first in the cagesmainly because of a big gain in entropy.

Size exclusion is a basic guideline for adsorption ofmolecules in zeolites. Here, we give a nice example of thesize exclusion in zeolites. The maximum quantities thatcould be adsorbed in Na+-X zeolite fromn-hexane solutionat 296 K were compared for anthracene and the anthracenederivatives, 9-methylanthracene (MA), 9,10-dimethylanth-racene (DMA) and 9,10-diphenylanthracene (DPA)[41].

The results are shown in the parentheses: anthracene wasadsorbed to the extent of nearly one molecule/cage, and theslightly larger MA was adsorbed to an extent slightly lessthan that of anthracene. However, an abrupt decrease in ad-sorbed quantity was observed for DMA and DPA both ofwhich apparently have a size larger than the aperture of thesupercage in the Na+-X zeolite. Thus, the adsorbed amountsof DMA and DPA may represent the molecules located onthe outer surfaces of the zeolite. The measurement of themaximum quantity adsorbed can give an idea about the lo-cation of molecules, i.e. whether inside or outside. Note thatthe maximum amount of anthracene adsorbed on Na+-A ze-olite is well less than 1× 10−6 mol/g. Another example isa CT complex of arene donors formed in methylviologen

Fig. 3. Adsorption and desorption isotherm curves of N2 in zeolite Na+-Yat 77 K: ( ) adsorption curve; () desorption curve. Both curves aresuperimposed, indicating that the adsorption and desorption processes arecompletely reversible.

(MV2+) exchanged zeolites. Yoon et al. found that the ap-pearance of the CT absorption bands strongly depend on themolecular dimensions of arene donors and they concludedthat the complexes form inside the cages and channels ina size-selective, shape-selective manner[42–45]. These twocases show that adsorbed organic molecules reside mainlyinside the zeolites if they can meet the requirement of sizelimitation imposed by the entry apertures to the cages.

As mentioned in the case of N2, the process of adsorp-tion is a dynamic one involving diffusion of adsorbates.A pictorial representation of this process is given by fluo-rescence spectroscopy[46]. A fixed quantity of anthracene(1.3× 10−4 mol/g) was adsorbed on Na+-Y zeolite by sub-limation at 150◦C for 27 h in an evacuated container, fol-lowed by the measurement of the emission spectra of thesamples with time. In one experiment, a sample was left atroom temperature, while in the other case, another samplewas kept in a drying oven at 150◦C. Fig. 4 compares theemission spectra of these samples.

Initially, a broad and structureless band centered at 500 nmascribable to the excimer emission is dominant, in additionto a structured band assignable to the monomer emission


Fig. 4. Emission spectra of 1.3×10−4 mol/g anthracene adsorbed by sub-limation on Na+-Y zeolite at 150◦C in vacuum: (A) spectrum measuredfollowing 27 h of heating at 150◦C; (B) spectrum measured after a weekof heating at 150◦C; (C) spectrum of a sample kept at room temperaturefor 2 months following the same treatment as that of A.

(A). When kept at 150◦C for a week, the contribution ofthe excimer emission decreased appreciably (B), indicatingthat the anthracene molecules initially accumulated on theexternal surfaces of the zeolite particles, thus giving rise toexcimer emission, diffused into the particles, resulting in themonomer emission. On the other hand, when kept at roomtemperature, the reduction in the contribution of the excimeremission took place, but only very slowly (C), suggestingthat the diffusion of anthracene molecules from the exteriorsurface to the interior of the particles is a very slow processat room temperature. Accordingly, the fluorescence spectro-scopic method demonstrates nicely that it can indicate thelocation of adsorbed molecules on zeolites and also tracetheir diffusional motions from outside to inside the zeolitecrystal driven by the concentration gradient.

Adsorption on the external surfaces of zeolites was ex-amined by Turro’s group using aromatic ketones such asdibenzylketones (DBKs) which have a larger affinity forthe surface than plain aromatics because of the polar –C=Ogroup [47,48]. MFI zeolite was employed, whose channelaperture has a smaller size than these molecules, such thatthey cannot enter the channels. Under these circumstances,the adsorption isotherm was measured, and analysis basedon the Langmuir isotherm assuming monolayer coveragewas applied. However, a Langmuir analysis that assumedonly one type of adsorption site failed to reproduce the ex-perimental adsorption isotherm curves; instead a Langmuirisotherm analysis based on two sites of differing binding

energies was successful in fitting the experiment. The resultof the analysis is consistent with a model in which there aretwo sites, i.e. holes (apertures to the channels) and externalframework surfaces; the former were found to have a muchstronger affinity for the ketones than do the latter.

Based on the examples given here, we can have a roughimage of the adsorption of molecules in zeolites. In partic-ular, it is very important to take into account the dynamicnature of the adsorption, i.e. the thermal diffusion of ad-sorbed species. We again emphasize that the diffusional mo-tion tends to bring molecules into the cages and channelsfrom the exterior surface. Also, adsorption on the exteriorsurfaces of zeolites is not necessarily a major drawback,since another type of interesting chemistry can be broughtabout on these surfaces.

Among the issues of adsorption and the distribution ofguest species in zeolite cavities, adsorption sites withinzeolites are of particular importance. Adsorption sites areusually associated with the extra-framework cations andframework oxygens. Of these, the primary adsorption sitesfor small aromatic molecules are considered to be locatedat the cation sites. For example, it has been established ex-perimentally (IR[49–52], Raman[53,54], NMR [55–59],neutron diffraction[60,61], small-angle neutron scattering[62,63]) as well as by computer simulation[64,65] thatbenzene can be localized at two distinct binding sites,cations and windows, within the Na+-Y zeolite supercage.In SII binding site (four per supercage), benzene is faciallycoordinated to a supercage 6-ring, 0.27 nm above Na+. Inthe window site (two per supercage, shared) benzene iscentered in the 12-ring window, 0.3 nm from SII . Becauseof the strong Na+-benzene interaction, binding site SII ismuch more stable than the window site. The distribution isconsidered highly heterogeneous[66] which is reminiscentof the situation within micelles[67]. The big difference isthat, while in micelles the equilibrium distribution is at-tained very quickly because of the high mobility of guestmolecules, this is not the case in zeolites due to very slowdiffusion [68]. Experimentally, NMR spectroscopy is usedto directly probe the guest populations for a high-pressureatomic gas occluded in the zeolite cavities[69,70]. As anexample, the distribution of xenon atoms in Na+-A zeolitewas described by Poissonian statistics for mean occupanciesless than about three Xe atoms per�-cage. At higher load-ings, the finite volume of the atoms becomes significant, asreflected by an excellent fit of the129Xe NMR data, with abinominal distribution for the value of the mean occupancy,n (n = [guest]/[cage]) up to five, with a maximum of sevenxenon atoms per�-cage. Then, the hypergeometric distribu-tion comes into play, but even this fails at very high loadings,predicting a larger variance than that actually observed.Thus, even for the simplest case of rare gas atoms, whichhave the least interaction, both mutually and with the frame-work, the distribution among the cages not very simple.

For large aromatic species such as naphthalene[71–73],anthracene[41,74,75], and pyrene[66,76,77], it has been


shown that the emission spectra in Na+-Y or Na+-X areloading-dependent: the contribution of excimer emission in-creases at the expense of monomer emission as the loadingincreases, even at very low mean occupancy(n 1). Thisis considered to be an indication of a heterogeneous distri-bution within the zeolites, because the excimer emission isconsidered to arise from more than one molecule occupyingthe same cage. Detailed analysis has not been carried out toelucidate the distribution law of these molecules, because itis imagined to be no easy task. In essence, one should beaware in zeolites that an average occupancy of unity doesnot mean that most supercages contain one guest species onthe average but rather that most cages are empty, with somebeing multiply-occupied[78].

2.3. Diffusion

The transport of molecules within the intracrystalline cagenetwork of zeolites is important, both in influencing chemi-cal reactions of the guest species and for the design and ap-plication of shape-selective catalysts[79,80]. In usual cases,the molecules introduced externally into zeolites are notfixed at particular sites, but rather migrate by executing ahopping motion from one adsorption site to another withinthe same cage and occasionally to other cages through oneof the connecting windows, unless they are immobilized bya “ship-in-a-bottle” synthesis[81–83]. Typically, intercagehopping assumes a relatively high activation barrier and is aslow process, allowing the guest molecules to stroll aroundthe adsorption sites within a given cage before they jumpout [84–86] (Fig. 5).

Reactions controlled by the transport of guest moleculesthrough the zeolite framework are of particular importancein that these can involve triplet states and radicals that mayhave long lifetimes. On the contrary, at short times (ps tons), reactions are static and occur between reactants locatedrelatively close to each other.

Depending on the similarities in the sizes of the moleculesand cages, the intracrystalline diffusivity can vary overa range of about 12 orders of magnitude, from 10−8 to10−20 m2/s. The values are obviously much smaller thanthose obtained in solutions, being constrained by the ge-ometries of pores. Of particular interest is benzene, whose

Fig. 5. Pictorial representation of the diffusion of molecules in a zeolite particle. The diffusional motion is classified as two types according to the natureof the activation energy: (1) activation energy for intercage jump and (2) activation energy for intracage jump.

diffusivity in Na+-Y and Na+-X was thoroughly investi-gated both experimentally and theoretically. Among var-ious experimental techniques, the pulsed field-gradientNMR (PFG NMR [87,88]) and quasi-elastic neutron scat-tering (QENS [89,90]) methods have gained popularity.Regarding the theoretical approaches, molecular dynamics[91–93] and Monte Carlo simulation[84–86] techniqueshave been carried out. According to a recent QENS study[90], the values of the self-diffusion coefficients of benzeneat 300 K are 3.1 × 10−11 m2/s in Na+-X (Si/Al = 1.23)and 4.6× 10−13 m2/s in Na+-Y (Si/Al = 2.43) for loadingsof 1.5–2 molecules per supercage. At room temperature,the diffusivity is almost 2 orders of magnitude lower inNa+-Y than in Na+-X. This reflects the strong influenceof cation distribution at this temperature. The activationenergy for diffusion in Na+-Y, 33.7 kJ/mol, is about twicethat in Na+-X, 17 kJ/mol, at low loading. The diffusivitiesare known to decrease with increasing loading[90].

The meaning of the self-diffusion coefficient,D is trans-lated to the mean-square jump distance,〈r2〉 according tothe Einstein expression in three dimensions[85]:

〈r2〉 = 6Dτ (1)

where τ represents the time required for a jump.Eq. (1)predicts that, in Na+-Y, in which the cage-to-cage jump dis-tance is 1.1 nm, the average distance traveled by benzene isestimated to be 1.5 cages in 1�s and 48 cages in 1 ms atroom temperature, while in Na+-X, it will be 12 cages in1�s and 400 cages in 1 ms. Thus, for species with long life-times, intercage migration is definitely an important factoraffecting kinetics and mechanism of reactions.

Only a few studies have been carried out so far on theintrazeolite diffusivity of aromatics larger than benzene. Forexample, the diffusion coefficient of naphthalene in Na+-Xand Na+-Y below 350 K was reported to be∼10−16 m2/s[94]. We have recently obtained the values ofD for azulene,ferrocene and anthracene, ranging from 10−15 to 10−16 m2/s,sufficiently smaller than the value of benzene in Na+-Y,based on the triplet quenching experiment at ambient tem-perature[68]. It is highly possible that the cage-to-cage dif-fusivity of the guest species is dependent on the molecularsize because of constraints on the diffusive motion imposedby the windows and walls of the zeolite. It has also been


pointed out that the intercage hopping dynamics of benzeneare largely affected by the adsorption interaction with thehost zeolite: a self-diffusion coefficient as small as 10−18 to10−19 m2/s was measured for benzene in Ca2+-Y, in whichbenzene is strongly held[95].

2.4. Sample preparation protocol

Several different methods have been applied for the ad-sorption of aromatic molecules into zeolites, including subli-mation at elevated temperatures under evacuated conditions[96–98], stirring zeolite powders in solutions of aromaticmolecules followed by the removal of solvents with evacua-tion treatments[28], and mixing solid aromatics with zeolitepowders in a mortar[99]. Different handling in the proce-dures for preparing samples as well as the different methodof sample preparation sometimes induces contradicting ob-servations even for similar systems. The major cause of thediscrepancies observed is associated with the various hydra-tion levels of the zeolites: water could easily be introducedinto the zeolites, which have an extremely hygroscopic na-ture, in the sample preparation procedure.

For example, Thomas and co-workers[100,101]observedin the transient absorption spectra the formation of the rad-ical cations of aromatic molecules and trapped electronsin the form of Na43+ absorbing at 520 nm in Na+-Y and550 nm in Na+-X upon laser excitation of the aromatics.On the other contrary, Gessner and Scaiano[102] failedto confirm the presence of Na4

3+ as a counterpart of theradical cations in a similar experiment. Both groups em-ployed the solution method for adsorption. As was pointedout later in the study from our laboratory[72], this con-tradiction is due to the fact that the absorption spectrumof the trapped electrons are considerably affected by thepresence of various amounts of water molecules within thezeolites. Their absorption peak is shifted to longer wave-lengths in the presence of large quantities of co-adsorbedwater, which might be introduced in the sample preparationprocedures.

In other studies, both fluorescence and triplet decay ratesof aromatics were reported to be significantly affected bythe presence of water in zeolites. Ramamurthy et al.[29]reported that the fluorescence decay rate of pyrene adsorbedinto Na+-Y is smaller with increasing hydration level. Hy-dration may cause the liberation of the guest moleculesfrom the host zeolite by the action of the water molecule,which has a much stronger interaction with the zeoliteand can lie preferentially on the zeolite walls. It has beenshown that the enhanced non-radiative decay rates in theadsorbed molecules give rise to the short lifetimes of the S1state[103,104]. Another cause of the short lifetimes in thedry zeolites may be the self-quenching of fluorescence byground-state species, since pyrene and other aromatics withrelatively large molecular size tend to form aggregates atthe immediate surface of the zeolites, and these aggregatescan dissociate in the hydrated zeolites[28,29]. Our study

Fig. 6. Decay constants of triplet-state anthracene plotted as a functionof the quantities (cm3/g) of various solvents for samples with anthraceneloading level of 1.0 × 10−5 mol/g in Na+-Y (pore volume: 0.378 cm3/g)at similar laser intensities. Symbols: (�) H2O; (�) D2O; (�) methanol;(×) pyridine; (�) acetonitrile; (�) n-hexane; (�) dehydrated.

showed that the fluorescence decay rate of anthracene isgreater with increasing loading levels in dry Na+-Y zeo-lite [41]. Excimer formation is largely responsible for theself-quenching of the fluorescence of anthracene in Na+-Xand Na+-Y.

The long triplet lifetimes of aromatic species were foundto be altered dramatically at various hydration levels inthe zeolites[105]. Fig. 6 depicts the decay constants oftriplet-state anthracene plotted as a function of the quantityof co-adsorbed water and other solvents.

In dehydrated Na+-Y (<0.4 wt.% water content), witha triplet decay constant of 60± 5 s−1, the value compara-ble to that observed in a rigid plastic matrix, poly(methylmethacrylate), was obtained[106,107]. With increasingwater content, the decay constant increased once to a valuethat is about 3 orders of magnitude larger than that in thedehydrated zeolite and then again decreased to a value moreor less similar to the original one in the presence of water inquantities sufficient to fill the total pore volume. This pecu-liar behavior of the decay constant of triplet state anthracenewas dependent both on the loading level of anthracene andon the intensity of laser light employed for excitation. Theobservation was rationalized by the different mobilities ofanthracene within the supercage networks of the zeolitebeing strongly dependent on the water content, leading tothe self-quenching and triplet–triplet annihilation. The in-tercavity motion of anthracene as well as the intracavitymotion as a function of water content initially increasesbecause of the cancellation of adsorption interactions ofthe guest anthracene with the zeolite walls, followed by asignificant decrease due to blocking of access to the neigh-boring cages and the reduction in the free volume inside thecages by water molecules. As is evident inFig. 6, solvents


Fig. 7. Absorption and emission spectra (uncorrected,λex = 355 nm) ofanthracene adsorbed into dehydrated (water content:<0.4 wt.%) zeolitesand in solution: (A) 5.0× 10−6 mol/g in Na+-Y; (B) 5.0× 10−6 mol/g inin the presence of 0.12 cm3/g n-hexane in Na+-Y; (C) 5.0×10−6 mol/g inthe presence of 0.24 cm3/g n-hexane in Na+-Y; (D) 1.0 × 10−4 mol/dm3

in n-hexane solution. The absorption spectra in (A)–(C) are representedby Kubelka–Munk units, while the spectrum in (D) is represented byabsorbance.

that have a strong affinity towards zeolite frameworks (al-cohols, acetonitrile, pyridine, etc.) have a similar but lesspronounced effect on the triplet lifetime of anthracene, butthose with weak interactions with the zeolite (n-hexane,benzene) have practically no effect.

Solvent molecules employed for the adsorption of aro-matic species, usually alkanes such asn-hexane, whichremained in samples after preparation due to insufficientevacuation treatment, affected the photophysical behaviorof guest species in a different way[74]. Although they donot appear to interfere appreciably with the adsorbed stateof guest aromatics on the interior surface of the zeolite,they do form complexes with the guest aromatics. Such anexample is shown inFig. 7 for the fluorescence spectra ofanthracene adsorbed in Na+-Y in the presence ofn-hexane.

In this case, a gradual red shift in the fluorescence spectrawith increasing amount ofn-hexane is indicative of com-plexes between anthracence andn-hexane at various molec-ular ratios. In spite of this fact, it was found that solventssuch asn-hexane do not affect dramatically the fluorescencelifetimes or triplet decaytimes of guest aromatic species,which makes it difficult to perceive them even when presentin samples.

These examples call for the necessity of taking extremecare in removing both water as a contaminant from sol-vents and/or moisture, and solvent molecules themselvesemployed for adsorption when aromatic species are adsorbed

conventionally using solvents. For removing water as wellasn-hexane or dichloromethane, evacuation for more than12 h at less than 0.1 Pa at an elevated temperature of about100◦C is recommended unless thermal reactions or evapo-ration of the guest species is a problem; this treatment gavea water content of less than 0.4 wt.% for Na+-Y. Note thatevacuation at room temperature even for a prolonged periodis not very effective for removing water andn-hexane. Con-sidering the lower reliability of the dehydration level in theadsorption method from solution, the adsorption through agas phase diffusional process in vacuum appears advanta-geous. However, this adsorption method tends to end up witha heterogeneous distribution of guest molecules, and heatingat an appropriate temperature for a prolonged period (usu-ally 2–3 weeks and sometimes even more than 1 month) isnecessary to allow guest species to diffuse through the entirecage network. To avoid the problems of solvent remainingin the samples and of possible damage due to heating, fluo-rocarbons such as perfluorohexane can be used as a solventif the molecules of interest are sufficiently soluble[108].

It should also be noted that the glass sealing of the samplesis absolutely necessary to protect them from contaminationby moisture. Most strikingly, unsubstituted aromatics suchas phenanthrene, anthracene, biphenyl, naphthalene, pyrene,etc., adsorbed into Na+-X and Na+-Y zeolites form crystalsin the presence of water and other solvents bearing strongaffinity for the zeolites in large quantities, which can nearlyfill the total pore volume[74,109]. Similar phenomena wereobserved in zeolite L, too[110]. This is a serious prob-lem because zeolite samples adsorbed with these aromaticspecies eventually produce crystals by absorbing moisturefrom air and no longer keep the guest species in the originaladsorbed state when left without being sealed.

Thus, we should point out that some photochemical stud-ies in zeolite systems might have problems due partly toinappropriate dehydration procedures and partly to insuffi-cient isolation method from air, e.g. use of septum seals afternitrogen purging of samples. These operations may cause apeculiar observation of aging of guest species in a period ofweeks and months due in part to slow evaporation of sol-vents used for adsorption and in part to penetration of waterinto the cages from the air[111]. The exact mechanism ofthe aging is not clear, but continuous diffusion and eventualcrystallization of the guest molecules caused by co-adsorbedwater seems to be responsible for the observed changes inthe fluorescence spectra of pyrene in zeolites. Thus, the re-mark can be made that a critical view should always be takenin reading the literature related to zeolite photochemisty asto the methods and handling in sample preparation proce-dures, since no standard procedures generally applicable tomany guest molecules are available at this moment.

2.5. Diffuse reflectance spectroscopy

Since first implemented in 1981 by Wilkinson and Kessler[112], the diffuse reflectance laser photolysis technique


Fig. 8. Schematic diagram for the diffuse reflectance laser photolysissetup.

has gained popularity for surveying the transient species invarious non-transparent solids such as organic microcrys-tals, powdered solid polymers and inorganic semiconductorpowders[113–117]. The technique has been successfullyapplied to photophysical and photochemical investiga-tions of molecules adsorbed on solid supports, includingsilica gel, alumina and zeolites[3]. The optical arrange-ment for experiments and the methods of data analysishave been established by Wilkinson’s group[113–115].The time resolution of detection systems was initially lim-ited to the microsecond region but was later extended tothe picosecond and femtosecond domains by utilizing apump–probe method in which an ultra-short white-lightcontinuum generated by focusing a fundamental laser beaminto H2O or D2O was employed as a monitor light source[118–120].

Fig. 8 shows a setup for a nanosecond time-resolveddiffuse reflectance laser photolysis apparatus built in ourlaboratory.

The third or fourth harmonics of a pulsed Nd:YAG laser(8 ns) were used as an excitation light source, and a bulb-typepulsed Xe lamp (∼50�s) or Xe short-arc lamp was usedfor a monitor light source. The laser beam was irradiatedat right angles to a powder sample contained in a 2 mmoptical path-length Suprasil cell. The monitor light beamwas illuminated at about 35–40◦ with respect to the excit-ing beam, and the diffusely reflected light, which is dis-tinct from the specularly reflected light from the surface ofthe sample cell, was collected with a lens (Ø 40 mm,f =100 cm) and introduced into a polychromator. The light sig-nal (400–900 nm) was detected either with a photomultiplieror a gated photodiode array detector with an image inten-

sifier. The time resolution was limited to the microsecondrange for the detection with a photomultiplier, because thediffusely reflected light was so weak in intensity that thephotomultiplier need to be wired for full-stage (nine dyn-odes for R928 (Hamamatsu)) for amplification, and a signalcable need to be hooked up with a load resistor larger than500�; this sacrifices the time resolution significantly. It isalso the case in the photomultiplier detection system for thesolid samples that the early stages of the event are disturbedby the scattered light from the laser pulse and/or intense flu-orescence emission. On the other hand, the detection witha gated diode array using a pump–probe method allowedthe time resolution to reach 5 ns by eliminating spontaneousemission and scattered light from the laser. In this system,the detection wavelength was extended to the near-IR re-gion (900–1600 nm) by using a red-sensitive InGaAs detec-tor [71].

With the diffusely reflected light from powdered sam-ples being used as the monitoring light, the transient ab-sorption spectra in the diffuse reflectance laser photolysisshould be obtained with a treatment different from that forspectra in ordinary transmittance mode measurements. TheLambert–Beer law is not applicable to represent the tran-sient absorption, because the diffusely reflected light consistsof a group of light beams with various optical pathlengthsfrom the sample. Instead, the Kubelka–Munk function hasbeen applied to describe the steady-state absorption in thereflectance mode[121]:

F(R) = (1 − R)2

2R= k

s

whereF(R) represents the Kubelka–Munk remittance func-tion; R, the diffuse reflectance from a sample;k, the absorp-tion coefficient; ands, the scattering coefficient. Thus, theplots of F(R) versus wavelength represent the absorptionspectra if the scattering coefficient is wavelength- inde-pendent. The Kubelka–Munk function has been extendedto time-resolved spectroscopy, and the transient absorp-tion intensities were expressed in terms of absorption (%)[115–118]:

absorption(%) = 100×(

1 − R

R0

)

in which R0 denotes the intensity of the diffusely-reflectedlight of the probe pulse with excitation, andR, that withoutexcitation.

It is important to note that the value of absorption (%)is not always a linear function of the concentration of thetransient species produced in the sample, and a linear re-lationship holds only at small values of absorption (%)[115,120]. The saturation phenomena of absorption (%) aredependent on the value ofk/s. For the samples with smallvalues ofk/s, the deviation from linearity is appreciableeven at a few percent of the absorption (%) value, whereasfor large values ofk/s, linearity holds at more than 15%.


The value ofs is larger for particles with smaller size andincreases linearly with the average surface area of particlesper unit volume of the sample, since the scattering of lighttakes place at interfaces between particles and air and thenumber of scattering events per unit volume in the powderis proportional to the surface area. In adsorbed systems, thescattering of light mostly results from host supports suchas zeolites, ands is approximated by the value for the hostmaterial. Thus, the guest species is required to have a largevalue ofk or ε (k = 2.303εc, whereε is the molar absorp-tion coefficient andc is the molar concentration) to allowa good linear relationship between absorption (%) versusthe concentration of the transient species up to a relativelylarge value of absorption (%). In actual experiments, it isrecommended that the linearity of absorption (%) versusthe intensity of laser light for excitation first be checkedfor a singlet or triplet excited-state species, which can begenerated as a linear function of laser intensity. Then, oneshould check the laser intensity-dependent behavior of agiven species before kinetic analyses for the species areperformed.

A precaution is necessary to prevent sample damagedue to repeated irradiation of laser light: a colored spotmay sometimes develop after several shots. No perma-nent remedy is available for solid samples, although aflow method can be used for solutions. Either shakingsamples after every few shots or changing the position ofirradiation on samples routinely can minimize the changein reflectance caused by the colored sample solids. It isalso important that the excitation intensity be as low aspossible for the least possible damage of samples. No-tably, monitor light occasionally causes an artifact in thedecay of transient species; two examples are given herefor such cases. In the first case, the transient absorptionsignal of a trapped electron, Na4

3+ generated in zeoliteNa+-X was followed by a continuous Xe-arc lamp throughIR-cut-off filters and a 10 cm-long water filter[122]. Thefilters are supposed to prevent the heating of the sample.It was found that the decay rate is dependent on the in-tensity of the monitor light; the decay rate is low whenthe monitor-light intensity was decreased using neutraldensity filters. In contrast, the decay rate was very slowwhen monitored with a pulsed Xe lamp with a durationof 10�s. The phenomenon observed is ascribed to thephotobleaching of trapped electrons. Similar phenomenaare well-investigated for trapped electrons produced by�-irradiation in low-temperature matrices[123–125]. In thesecond case, the decay rate of the triplet–triplet absorptionsignal of anthracene in the Na+-Y zeolite was found to de-pend on the intensity of monitor light, and again the decayrate was lower as the light intensity was decreased (Fig. 9)[68].

In this case, the temperature rise in the sample causedby the small amount of heat generated by irradiation mustbe responsible. The two examples tell us the importance oftaking extra care about the intensity of the monitoring light

Fig. 9. Monitor light intensity-dependent decay curves of triplet-state anthracene in dehydrated Na+-Y zeolite. Loading: 5.0×10−6 mol/g.Relative intensity of monitor light is indicated in perccentage.The lower decay rate was observed with decreasing monitor lightintensity.

when working with a transient species with a long lifetimein solid samples. Preferably, the use of monochromatic mon-itoring light is recommended.

Transient absorption spectroscopy in terms of diffuselyreflected light was found to carry a certain limitation in thetemporal resolution of transient absorption signals in thefemtosecond time domain[120]. This is because ultra-short(much less than the picosecond scale) light pulses of bothexcitation and probe light are temporally broadened due tonumerous times of refraction, reflection and diffraction inlight-scattering materials before arriving at the sample sur-faces[126]. Thus, the time profiles of transient absorptiondepends on the optical properties of the samples, namelyk and s. In the case of smallk for a given value ofs, theconcentration of excited species near the sample surfacegradually increases with time, and its depth profile changesup to several tens of picoseconds. This phenomenon leadsto a decay time much longer than the lifetimes of a giventransient species, given that the lifetimes are less than afew picoseconds. On the other hand, for a sample witha largek value, excited species are populated only at ornear the surface due to the limited penetration depth ofthe excitation light, and the temporal change in the depthprofile is completed within a few picoseconds. These fea-tures enable the decay of absorption (%) to correspondfaithfully to the concentration of the transient species. Ineffect, the temporal resolution is less than a few picosec-onds, equivalent to that of transmittance mode experimentsunder an appropriate optical condition, such as a largeextinction coefficient at the excitation wavelength with asmall scattering coefficient. It should be borne in mind,however, that the time resolution might be extended to


several tens of picoseconds under inappropriate opticalconditions.

3. Molecular photophysics and photochemistry inzeolites

3.1. Confinement effect

Spatial confinement in the cavities of zeolites is expectedto provide molecules with geometric restrictions due to theframework, which may lead to unusual photophysics and/orphotochemistry of the guest species. This idea was tried

with a molecule whose size is similar to the dimension ofthe entry aperture, and thus the host/guest complex was ex-pected to fit in very tightly. Alternatively, in order to inducea pronounced confinement effect, a bulky molecule that can-not enter through a pore opening was incorporated with theship-in-a-bottle synthesis[81–83]. Further extension of theidea of confinement has led to the concept ofelectronic con-finementin which it was anticipated that the molecular or-bitals of the guest species undergo severe deformation dueto the geometrical constraints imposed by the zeolite walls.A number of research efforts have been carried out on stericconfinement. Another important aspect of the confinement isthat zeolite cavities may prevent or restrict the approach ofguest molecules, in particular, reactive intermediates. Here,we review from the two aspects, static and dynamic, con-finement within zeolites.

Cationic dyes are a good example of how monomer–dimer(or aggregate) equilibria are dramatically affected by theconfinement in zeolites[127,128]. Thionine (TH+), whichis smaller in size than the channel diameter, was found tobe incorporated in the monomer form in the main channelsof zeolite K+-L via ion exchange; however, methylene blue(MB+) which is larger than the channel diameter, couldnot enter the channels but instead remained on the exteriorsurfaces, forming dimers and aggregates.

The spatial restriction by the channel walls prevents theformation of parallel dimers (H-type aggregates) inside thechannel. Most interestingly, Carzaferri’s group showed that

TH+ molecules first adsorb on the outer surfaces in thedimeric form and slowly enter the channels in a monomericform [127]. This clearly shows that the spatial regulation bythe channel walls depends on the guest species to be incor-porated. Since the cationic dyes have a tendency to form ag-gregates in solution and these aggregates cause photodegra-dation of dye molecules, incorporation into zeolites is oneof the promising ways to improve the photostability of dyes.

Association of the methylviologen radical cation (MV•+)was investigated in zeolites by Yoon and co-workers[129].The dimerization of MV•+ was observed, from the mea-surement of absorption spectra, exclusively in zeolite L andnot in the large-pore zeolite Y or in narrow-channel ZSM-5when incorporated into dehydrated zeolites.

This finding, which is in striking contrast with the obser-vation of TH+ in zeolite L is due to the smaller short-axislength for MV•+ than for TH+. As observed for MV•+,neutral molecules such as biphenyl, naphthalene and an-thracene were also found to form a ground-state dimer (pos-sibly with sandwich-type configuration), which gives riseto excimer emission on photoexcitation within the straightchannels of zeolite L[75]. This fact suggests that moleculeswith a small lateral width are forced to lie along the channelwall within zeolite L owing to a monodirectional geometricrestriction. Molecules that have a short-axis, of size similarto that of MV•+, can fit into the channel, forming a sand-wich structure. This geometry is expected to bring about arather tight fit, and indeed, a remarkably strong interannularinteraction was observed between two anthracene moieties,as evidenced by the electronic absorption spectrum. Overall,the channel-type zeolite L exhibits larger structural regula-tion for adsorbed species over zeolite Y with large sphericalpores.

The tight fit of molecules into the zeolite pores andchannels enables the observation of room temperature phos-phorescence (RTP). Although Ramamurthy et al. reportedpreviously[99,130], the remarkable enhancement of phos-phorescence from aromatic species adsorbed in heavy atom-exchanged zeolites, as will be described in more detail

in Section 3.2, RTP was observed in zeolites withoutthe assistance of heavy-atom perturbations. A study fromour laboratory showed that 9-ethylcarbazole (ECZ) and


other aromatic species, e.g. naphthalene, phenanthrene,1,2,4,5-tertacyanobenezene, etc. give rise to phosphores-cence emission with an appreciable intensity in K+-Lzeolite at room temperature despite the fact that RTP isextremely weak in zeolite Y, which has large sphericalpores[75].

The phosphorescence lifetime of ECZ in K+-L was mea-sured to be 2.6 s at 296 K, compared to 7.6 s at 77 K, andthis long lifetime suggests that the rate of radiationlessdeactivation from T1 is significantly reduced. Thus, it wasconcluded that the tight fit into the channels and resul-tant increased structural rigidity are largely responsiblefor RTP. Earlier than our work, Scaiano and co-workersreported the observation of RTP from acetophenone and�-phenylpropiophenone incorporated into in silicalite[131],an all-silica zeolite with narrow channels. Since aromaticketones have an intersystem crossing quantum yield of al-most unity, the observation of RTP appears much easier.Nevertheless, the steric restrictions imposed on the ketonesin the channels of silicalite are very important for the obser-vation that pioneered RTP in zeolites. A similar observationof RTP was reported for naphthalene in ZSM-48 (pure silicazeolite having unidimensional channels with dimensionsof 0.57 nm × 0.51 nm) and 2,3-diazabicyclo[2.2.1]hept-2-ene (DBH) in silicalite and Theta-1 (high-silica zeo-lite with Si/Al = 30 and monodirectional channel-typezeolite with an oval channel of 0.46 nm × 0.57 nm)[132].

Marquez et al. proposed an idea that the influence ofthe cavity dimensions on the electronic structure of someguest organic molecules incorporated within zeolites canbe related to the quantum confinement concept[132–134].This picture was correlated with an “electronic confine-ment effect” in which the electron density of the guestis constrained to be mainly localized within the zeolitecavity as a result of strong short-range repulsion with theelectrons of the zeolite walls[135]. This idea was firstapplied for the interpretation of the absorption and fluores-cence spectroscopic properties of anthracene incorporatedin zeolites with various pore sizes (Na+-Y, Na+-mordenite,Na+-ZSM-5 in the order of decreasing pore size)[133]. A

larger bathochromic shift of the 0-0 transition of the elec-tronic absorption spectrum in zeolites with smaller poresizes was directly ascribed to the electronic confinementeffect. The concept of electronic confinement is based ona molecular orbital (MO) treatment. Later, these workersobtained the same conclusion from the absorption and emis-sion spectroscopic observation of naphthalene[132] andDBH [134] by applying the same analysis. Notably, theyincorporated these molecules in pure silica zeolites, whichcontain few cations, in order to study exclusively the influ-ence of the confinement effect, avoiding the interference ofelectrostatic effects due to framework and extra-frameworkcations.

A few questions can be raised regarding this work. Theidea of electronic confinement is fascinating if solid ex-perimental proof is obtained by using well-suited guestmolecule/zeolite pairs with carefully executed procedures.First of all, with normal adsorption techniques[96–99], it isextremely difficult to incorporate molecules externally intozeolites with such narrow spaces that they may provide alarge enough constraint on the guest species to bring aboutdeformation of its MOs. They will be simply expelled fromthe pores and channels and may stick on the outer surfaces.In this sense, anthracene may not be very appropriate forsuch a purpose. Anthracene molecules are safely assumed togo into the cavities of zeolites X and Y but scarcely go intothe small channels of Na+-ZSM-5. Besides, research fromour laboratory showed that anthracene forms a ground-statedimer, which gives an appreciably distorted absorptionspectrum even at very low loadings due to excitonic inter-annular interaction forced by the channel walls and affordsa broad, red-shifted excimer emission spectrum in the nar-row channels of zeolites K+-L and Na+-L [75]. Rememberthat the size of the entry aperture of the channel for zeoliteL is similar to that of mordenite. Thus, we can observe amore red-shifted 0-0 band of the absorption spectrum in ze-olite L than in Na+-Y, but for anthracene this is not due tothe electronic confinement effect. The tendency of formingaggregates on the surfaces of zeolites with small cavitiesalso makes it difficult to detect the minute differences in thespectra dependent on the zeolite type. A further concern isthe formation of anthracene crystals, which give red-shiftedabsorption and emission spectra[74,110]. They can easilybe formed when samples adsorbed with anthracene are leftin the open air due to hydration. A similar discussion mayapply to naphthalene and DBH because of the very poorspectra, which make it difficult to be convinced by the ar-guments given in the literature. Extra effort in experimentalwork is needed to verify the idea of electronic confinementby zeolites.

Investigations aimed at pursuing large confinement effectswere carried out with bulky molecules incorporated using theship-in-a-bottle synthesis. Garcıa and co-workers[136] stud-ied the photophysical properties of 2,4,6-triphenylpyrylium(TPP+) and substituted tritylium ions encapsulated withinY, �, and MCM-41 zeolites.


The effect of a tight fit was observed as simultaneous emis-sion of fluorescence (λmax = 470 nm) and RTP (λmax =560 nm) of TPP+ within H+-Y and La+-Y, whereas onlyfluorescence was detected for TPP+ in extra-large-poreMCM-41 or on silica. The characteristic T1–Tn absorptionspectrum of the TPP+ was detected with transient absorp-tion measurements. For TPP+/La+-Y, the maximum ofthe broad 450–600 nm absorption was somewhat shiftedto longer wavelengths (peaking at 550 nm) with respect toTPP+/MCM-41. They also noted an increase in the extinc-tion coefficient for the band in zeolite Y compared withthat in MCM-41 or the spectrum in solution. Althoughthey ascribed these changes to molecular orbital distortionsresulting from the confinement of the molecule within therestricted space of the cavity, the effect of confinement isnot very dramatic, judging from their result.

The 9-(4-methoxyphenyl)xanthylium ion (AnX+), abulky cation, has also been prepared by a ship-in-a-bottlesynthesis within the supercages of zeolites Y and� [137].

No fluorescence has been observed for AnX+ at room tem-perature in solution, and this anomaly was rationalized as-suming a deactivation by intramolecular quenching of thexanthylium moiety by the anisyl substituent occurring in atwisted conformation. RTP was observed from AnX+ in thezeolites. Besides, the triplet-state of AnX+ was observedin the zeolites as a long-lived species (tens of microsec-onds) with transient absorption spectroscopy, while it hasnot been observed in solution. These results indicate that theintramolecular quenching of the xanthylium moiety by theanisyl substituent is less effective in the restricted environ-ment provided by the zeolite cage.

The confinement effect was stimulated with the aid ofco-adsorbed solvent molecules within zeolites. Scaianoet al. [138] found that the incorporation of large concen-trations of pyridine leads to an increased lifetime of tripletp-methoxy-�-phenylpropiophenone adsorbed in Na+-Y.

In the case ofp-methoxy-�-phenylpropiophenone, the tripletlifetime is normally determined by intramolecular quench-ing by the�-phenyl ring. The observed enhancement of thetriplet lifetime was regarded as an indication of reduced abil-ity to undergo conformational change as the pyridine occu-pancy of the supercages increased. Thus, the extent of thecompact environment found by guest molecules upon ad-dition of large amounts of co-adsorbed molecules and anincreased local viscosity within the supercages of the ze-olite provided by the co-adsorbed solvents were suggestedto slow the intramolecular quenching process. It should benoted that the effect of co-adsorbed solvent molecules iscritically dependent on the nature and quantity of solventsintroduced and a certain precaution is necessary, becausethe solvents usually increase the mobility of guest specieswithin the zeolites[105,139].

As an example of the dynamic confinement effect (re-striction of movement), Turro’s group investigated thecombination reaction of benzyl radicals and diphenyl-methyl radicals, produced by the photolysis of DBKs and1,1,3,3-tetraphenylacetone on MFI zeolites[47,48,140].

The parent molecules are too large to enter the channelsand are forced to reside on the external surfaces. Two typesof adsorption sites, a capping position on the channels orhole (site I) and the framework surface (site II) are assumedbased on the Langmuir isotherm study of adsorption. TheESR and fluorescence spectroscopic studies as well as prod-uct analyses revealed that the radicals with a smaller sizethan the channel diameter can diffuse into the channels,while those too bulky to fit in the channel diffuse on theouter surface, leading to combination forming stable prod-ucts. Here, the benzyl radicals result in the combination re-action both on the external surfaces and within the channels;on the contrary, a portion of the diphenylmethyl radicals aretrapped within the channels and are found to persist for manyweeks.

Another piece of research in Turro’s group also discoveredthat a diphenylmethyl radical (DPM•) generated by photol-ysis in a MFI zeolite, the sodium form of LZ-105, is ex-tremely persistent and reacts reversibly with oxygen to forma peroxy radical (DPMO2•) at room temperature[141].


The DPM• was detected both by an ESR spectrum and by afluorescence spectrum. DPM• is rendered persistent througha “supramolecular steric effect” (translated as dynamic con-finement effect), which severely inhibits the approach of twoguest species, leading to radical–radical reactions.

3.2. Interaction with cationic sites

In their pioneering work, Ramamurthy et al.[99,130]reported that external heavy-atom effects of associatedcations on the guest allocated within zeolite supercagesincrease remarkably the efficiency of intersystem cross-ing, leading to triplet states. In some cases, this externalenhancement can be comparable to effects caused by cova-lently bonded heavy atoms. For example, the phosphores-cence of aromatic species such as naphthalene incorporatedinto Rb+-X and Cs+-X at 77 K was observed, while onlythe fluorescence was detected in Na+-X and Li+-X. Amuch greater effect of the charge-compensating cationswas observed in Tl+-X, in which even RTP was observedfor the incorporated aromatics. Similar systems were in-vestigated with transient absorption spectroscopy in ourlaboratory[142], and a more elaborate picture of the inter-action between Tl+ ions and guest aromatic species wasobtained.

Transient absorption measurements were carried out foranthracene, phenanthrene and naphthalene incorporated intoNa+-X, Cs+-X (68% Cs+) and Tl+-X (98% Tl+), and therecorded spectra are depicted inFig. 10.

The spectra in both Na+-X and Cs+-X are quite similarto those of triplet–triplet (T1–Tn) absorption in solutions.Note, however, that the absorption intensities in Cs+-X aremuch greater for the samples with a similar loading levelat a constant excitation energy than those in Na+-X due toheavy atom effect. Strikingly, the transient absorption spec-tra in Tl+-X are different from those in Na+-X and Cs+-X:remarkably broad and featureless spectra with a red shiftare obvious. Nevertheless, the transient spectra in Tl+-Xare assigned to the T1–Tn absorption of these aromatics onthe basis of a similarity of the decay profiles of transientabsorption and phosphorescence. Phosphorescence intensi-ties are considerably higher in Tl+-X, although the phos-phorescence spectra in Tl+-X are not much altered fromthose in Cs+-X: slightly broader envelopes and small redshifts in Tl+-X were noted. These observations are rational-ized on the grounds that triplet complexes of the aromaticspecies with the charge-compensating cations are formed inthe Tl+-X system.

The triplet complexes were observed in zeolites, but notin solutions. In solution studies[143–145], it was demon-strated that the fluorescence quenching of a number of aro-matic hydrocarbons by metal ions including Tl+ and Ag+,leads to the enhanced formation of the triplet-state of thearomatic species. Among the quenchers, the quenching rateconstants of Tl+ and Ag+ were exceptionally larger thanthose of the pure heavy-atom quenchers, Cs+ and Xe. Thequenching mechanism due to non-fluorescent complex for-mation, followed by a rapid intersystem crossing, was sug-gested, but no experimental proof has been obtained. Thus,the triplet complexes detected with the transient absorptionspectra are considered as a precursor that was sought inprevious solution studies[143–145]. The observation of thetriplet complexes in Tl+-X is one of the intriguing examplesthat demonstrate the usefulness of the zeolite that geomet-rically fixes labile species in solution.

Scaiano and co-workers[146] demonstrated the differ-ent effects of charge-compensating cations on the excited-state behavior of guest species; the interaction of Brønstedand Lewis acid charge-compensating cations of zeolite with1,1-diphenyl-2-propane (DPA) and 1,1,3,3-tetraphenyl-2-propane (TPA) can divert their normal photochemicalbehavior, which is characterized by the formation ofdiphenylmethyl radicals (DPM). Laser photolysis of DPAincluded in Na+-Y and Na+-X zeolites generated a transient

Fig. 10. Transient absorption spectra of aromatics incorporated into Na+-X(top), Cs+-X (68 at.% of Cs) (middle), and Tl+-X (97% Tl, bottom) zeo-lites at a delay of 1�s upon excitation with a laser pulse: (A) anthraceneexcited at 351 nm; (B) phenanthrene excited at 351 nm and (C) naphtha-lene excited at 248 nm.


absorption that is reasonably close to that reported fordiphenylmethyl radical (DPM•) in solution:

Ph2CHCOCH3hv→Ph2CH• + •COCH3

Contrastingly, the laser photolysis of DPA in a de-aluminatedand H+-exchanged Y zeolite (HDY) led to a transient ab-sorption consisting mainly of DPM carbocation (DPM+)with a minor contribution of DPM•. HDY has been indi-cated to possess a nearly uniform distribution of the strengthof the Brønsted acid sites, and the ground-state absorptioncharacteristics of incorporated DPA suggested the absenceof an n→ �∗ band due to the C=O chromophore owing toprotonation at the acid sites. Thus, a mechanism in whichDMP+ was generated directly from the excited states of theconjugate acid of the DPA was suggested.

Enhanced formation of DPM+ relative to the yield ofDPM• was noted for the photoexcitation of DPA adsorbedin Zn2+-Y that contains only Lewis acid sites. This findingsuggests that the modification of the excited-state reactiontakes place regardless of the Lewis or Brønsted acid natureof the zeolite sites.

Close inspection of fluorescence spectra dependent on thedegree of hydration as well as on the presence or absenceof cations revealed the notable interaction of cationic siteswith aromatic species such as anthracene in Na+-Y zeo-lite [74]. This is reminiscent of the “cation–� interaction”,which has attracted much attention from a theoretical pointof view. Several examples of the cation–� interaction havebeen given[147–149], including the clustering of alkalimetal ions with aromatics and olefins in the gas phase, thecomplexation of cations with calixarenes in solution, andthe binding of proteins with cationic ligands or substratesin biological systems. This concept also applies to the na-ture of the adsorption of aromatics into zeolites, as has beenshown for benzene[49–65]. Fig. 11 compares absorptionand emission spectra of anthracene adsorbed at low loadingin Na+-X, Na+-Y, USY and inn-hexane solution.

Apparently, the spectra in Na+-X and Na+-Y are broad-ened compared with those in siliceous USY andn-hexanesolution, and this is due to the presence of the cation.Additionally, Fig. 12 shows the changes in the spectra ofanthracene in Na+-Y on addition of various quantities ofwater, which interacts with the zeolite, both the frameworkand cations, more strongly than anthracene does.

On increasing the quantity of water, the emission spectralbands sharpen and finally resemble those in USY. Therefore,the broadening of the anthracene emission spectra coincideswith the strength of the cation–� interaction. For pyrene, thebroadening of the emission spectrum in the zeolite Na+-X

Fig. 11. Absorption and emission spectra (uncorrected,λex = 355 nm) ofanthracene adsorbed into dehydrated (water content:<5×10−4 mol/g) ze-olites and in solution: (A), 5.0×10−6 mol/g in Na+-X; (B) 5.0×10−6 mol/gin Na+-Y; (C) 5.0 × 10−6 mol/g in USY; (D) 1.0 × 10−4 mol/dm3 inn-hexane solution. The absorption spectra from (A) to (C) are repre-sented by Kubelka–Munk units while the spectrum in (D) is representedby absorbance.

Fig. 12. Absorption (Kubelka–Munk units) and emission spectra (uncor-rected,λex = 355 nm) of 1.0×10−5 mol/g anthracene adsorbed into Na+-Ywith various levels of hydration: (A) dehydrated (<5 × 10−4 mol/g ofwater); (B) 0.04 cm3/g hydrated; (C) 0.08 cm3/g hydrated; (D) 0.17 cm3/ghydrated.


and Na+-Y is appreciable, and this was also indicative ofthe cation–� interaction[150].

A more general consequence of the presence ofcharge-compensating cations can be related to the electro-static field within the zeolite. Light charge-compensationcations such as lithium ions have been suggested to exerta large effect due to their small size, giving rise to highcharge density[5,7]. This effect of lithium ions has beendescribed as the “light atom effect” by Ramamurthy andco-workers who utilized this effect for the control of pho-tochemical transformations within zeolites[5,7]. Also, thegeneral trend of the electrostatic field generated by thecharge-compensating cations within the zeolite has been ex-plored with fluorescence probes that normally sense solventor polarity effects[27–31].

Cozens et al. observed that the xanthyl radical under-goes an ET with O2 to generate the xanthylium cation inzeolite Y [151]. The ET yield was strongly dependent onthe charge-compensating cation: the yield is much higherin Li+- and Na+-Y than in Rb+- and Cs+-Y. As shown inscheme, the reaction was assumed to involve an intermedi-ate, a peroxy radical:

They considered that Li+ and Na+ provide more stabiliza-tion to the O2

•− than the larger cations such as Rb+ and Cs+.Thus, the electrostatic field around the charge-compensatingcations may play a decisive role in the feasibility of ET fromthe radical to O2.

3.3. Interaction with the zeolite framework

Interaction of guest species with the zeolite framework ismainly associated with the electron-donating nature of thezeolite. As such, there exists an increasing understandingthat the zeolite framework is more active than expected be-fore. Here, we give a few examples.

Thomas and co-workers[152,153]explored the reactionof photo-excited pyrene incorporated into a basic zeolite,a partially cesium-exchanged X and Y zeolite (Cs+-X andCs+-Y). At the low loading level of 5×10−6 mol/g, the laserphotolysis experiment afforded a transient absorption spec-trum consisting of the band of the radical cation of pyrene(Pyr•+) and that of the radical anion of pyrene (Pyr•−).These workers assumed that the ET from the basic site ofthe zeolite to the excited singlet state(1Pyr∗) of pyrene isresponsible for the formation of Pyr•−. In their system, thesimultaneous formation of Pyr•+ and Pyr•− suggested apossible ET from pyrene in an acidic environment to pyrene

in a basic environment on photo-excitation as an alternativemechanism. However, they excluded this possibility on thegrounds that ET requires both donor and acceptor to be inclose proximity, whereas the sufficiently low loading levelof the pyrene employed implies that a great majority of thepyrene molecules are located in different cages. Thus, theirpoint is that Pyr•− was produced through the ET from thebasic site in the zeolite to1Pyr∗ independently from Pyr•+,which is formed by the ET from1Pyr∗ to the acid site withinthe zeolite.

Factors affecting Pyr•− formation were investigated. Ahigh yield of Pyr•− was obtained in zeolite X with respectto that in zeolite Y. The yields of Pyr•− in both zeolites fol-low the same sequence of Cs+ > Rb+ > K+ > Li+ for thecharge-balancing cation-exchanged zeolites. These resultssuggest that the basic nature of the zeolite is responsiblefor the reaction. In addition, the formation of Pyr•− wassuggested to follow a single-photon process on the basis ofthe linear changes of the yield of Pyr•− as a function of ex-citation light intensity in K+- and Cs+-exchanged X and Yzeolites. This observation is considered to be the foundationfor the assumption of a photoinduced ET from the zeolites

to 1Pyr∗. Some concern can be raised over the mechanismof Pyr•− formation: (1) Is it not possible that photoejectedelectrons from1Pyr∗ formed with a single-photon mecha-nism participate in the formation of Pyr•−? (2) Is the ion-ization energy or oxidation potential of the zeolite sufficientto cause the photoinduced ET leading to the formation ofPyr•−? These questions remained to be answered experi-mentally.

Keavan and co-workers[154] observed the photoreduc-tion of methylviologen (MV2+) ion-exchanged in Na+-Y,Na+-A and Na+-X, at 77 K. Also, Scaiano, Garcıa andco-workers[155] showed that faujasites and other zeolitescontaining MV2+ behave as electron donors on excitationof the absorption bands of MV2+, giving rise to MV•+ atambient temperature. In their study, various alkali metalion-exchanged zeolites were employed. The observationof MV•+ in the transient absorption spectra and conven-tional diffuse reflectance spectra after photoirradiation ledto a picture of photon absorption promoting MV2+ to itsexcited-state, from which it abstracts one electron from theoxygens of the framework, generating the oxygen radicalcenter and MV•+.

Furthermore, work from our laboratory[156] showeda clear indication of the electron-donating property of ze-olites in the photoinduced ET reaction. Strong electron


Fig. 13. Transient absorption spectra (λex = 266 nm) of TCNB in faujasitezeolites with delays at 100 ns (—) and 10�s ( ): (A) 5.0×10−5 mol/g inUSY (laser power: 12.0 mJ/cm2); (B) 5.0 × 10−5 mol/g in Na+-Y (laserpower: 8.8 mJ/cm2) in the presence of 2 wt.% water; (C) 4.2×10−5 mol/gin Cs+-Y (laser power: 8.8 mJ/cm2) in the presence of 2 wt.% water.

acceptors such as 1,2,4,5-teracyanobenzene (TCNB) andother cyanoaromatics were incorporated within the zeolites,and their redox properties in the excited-state were pursuedwith the laser photolysis experiment.

Three types of faujasite zeolites with increasing order ofelectron-donating ability, i.e. USY< Na+-Y < Cs+-Y(60 at.% of Cs+), were employed as host materials.Fig. 13compares the transient absorption spectra upon excitation ofTCNB at 266 nm in the three zeolites under similar experi-mental conditions.

In USY, the transient absorption bands peaking at 380and 615 nm were assigned to those of triplet-state TCNB,whereas in Na+-Y and Cs+-Y, an additional new band as-cribable to the radical anion of TCNB, TCNB•− was ob-served with a peak at 474 nm. Note that a linear relationshipwas obtained between the yield of TCNB•− and the laserpower applied both in Na+-Y and in Cs+-Y; thus, the for-mation of TCNB•− was inferred to follow a single-photonprocess. This finding suggested the occurrence of ET in theexcited-state of TCNB in the zeolites with sufficiently basicnature.

Fluorescence decay curves measured for TCNB in thethree zeolite systems are depicted inFig. 14. Apparently,a faster decay was observed in zeolites with stronger ba-sic sites, which indicates the occurrence of fluorescence

Fig. 14. Fluorescence decay curves of 5.0×10−5 mol/g TCNB in variouszeolites excited at 300 nm and monitored at 340 nm: (a) in USY; (b) inNa+-Y; (c) in Cs+-Y and (d) lamp profile.

quenching for TCNB adsorbed in zeolites with electron-donating nature. Thus, it was concluded that the ET fromthe electron-donating sites within the zeolites, Na+-Y andCs+-Y to the excited singlet state of TCNB(1TCNB∗) wasunambiguously observed:

1TCNB∗ + zeolite→ TCNB•− + zeolite•+

Additionally, in Cs+-Y, the time evolution of the transientabsorption signal of TCNB•− showed an indication of theET from the zeolite to3TCNB∗:

3TCNB∗ + zeolite→ TCNB•− + zeolite•+

Thus, the absorption signal of TCNB•− evolved in twostages, a fast rise within the duration of the laser pulse anda slow rise in the microsecond time scale, which coincideswith the decay of the triplet signal. Remember that the oxi-dation of zeolites is considered to take place at specific sitessuch as the framework oxygens at the 12-membered ring.

The absorption bands of CT complexes formed within ze-olites were also affected by the electron-donating nature ofthe zeolites[157,158]. A larger hypsohochromic shift wasobserved for the peak position of the CT absorption bandin the zeolite with increasing electron-donating ability. Ad-ditionally, ground-state species such as MV2+ [158] and


I2 [159] incorporated in zeolites exhibited a similar phe-nomenon, depending on the basic nature of the zeolites, withthe trend, however, in reverse. On going to greater basicityof the zeolite, a larger bathochromic shift was observed forMV2+ while a hypsochromic shift was observed for I2.

The example given thus far in this section is associ-ated with a CT interaction in the excited-state of the guestspecies, leading to ET from the zeolite framework to theguest species. Cozens and co-workers showed anothertype of participation of the zeolite host in reaction: thezeolites react by nucleophilic attack with a carbocation,4-methoxycumyl cation, depending on their basicity[160].They investigated the reactivity of this carbocation, usingnanosecond laser photolysis in non-acidic zeolites.

The carbocation has a relatively short lifetime, in the�s toms range, in dry alkali metal cation-exchanged zeolites. Thelifetime decreases in a systematic manner as the cation is var-ied from Li+ to Cs+ and with decreasing Si/Al ratio and isthus strongly dependent on the basicity of the zeolites. Thisfinding was attributed to the direct interaction between thecarbocation and the zeolite host, which may act as a nucle-ophile, adding to the carbocation to generate a carbocationadduct (framework-bound alkoxy species). The reactive na-ture of zeolites toward the carbocation, akin to nucleophilessuch as water and methanol, was thus revealed.

4. Intracage reactions

In the recent few years, the investigation of photochemicalreactions between more than one guest species, e.g. A andB both incorporated in the zeolite, increased significantly, inaddition to the photophysical and photochemical studies ofindividual molecules. Here, we take a look at a few examplesand assess the achievements of those studies.

4.1. Excimer and dimer cations

One way to look at intracage reactions is to exploit theheterogeneous distribution of molecules among the cages,e.g. increasing the loading level leads to the multiple oc-cupancy of the cage. Excimer formation of pyrene wasinvestigated as early as 1984[76] and also later in Na+-X

and Na+-Y [66,77,150]. At high loadings, the broad andfeatureless emission observed at the expense of monomerfluorescence was ascribed to the excimer emission. It wasconcluded that the excimer emission arises on excitationof dimeric species in doubly occupied cages, since no risewas detected on pulsed excitation in the nanosecond timeregime. A concern about pyrene is that crystalline pyrenegives excimer-like emission, and the microcrystals caneasily be formed with inappropriate sample preparationprocedures. In particular, the use of a rotary evaporator forremoving solvent in the early study[76] might have causedthe formation of crystals on the outer surface of the zeolitesif unadsorbed pyrene were left in solution. Additionally,incomplete isolation of the samples from the atmosphere

may lead to subsequent crystallization, as mentioned inSection 2.4. For this reason, the investigation with an-thracene or naphthalene is advantageous because the crys-tals of these species give no excimer-like emission[161].

The excimer emission of anthracene was observed inzeolites Na+-X and Na+-Y by our group[41,74,75]. Theobservation in the stationary experiment is similar to that ofpyrene, but the dynamics characteristic to the zeolites areof interest. The decay rate of the monomer fluorescence isdependent on the loading level: the decay is faster at higherloading levels. On the other hand, the rise in the excimeremission is very fast, on the ps time scale, indicating a sig-nificantly fast formation process. One possibility is energymigration from the monomer singlet state to the variousconfigurations or overlapping structures of dimeric speciesresiding in the neighboring cages. The other possibility isto assume a purely dynamic process: fast quenching of theexcited singlet state of anthracene by another anthraceneoccupying the same cage. Since the absorption spectrumof anthracene was not altered depending on the loadinglevel, the formation of a ground-state dimer is not evidentin Na+-Y. In contrast, an appreciably deformed absorp-tion spectrum was obtained in channel-type K+-L zeolite,suggesting the formation of the ground-state dimer[75].Whether the excimer formation is static or dynamic is stillnot clear for anthracene in Na+-Y.

The excimer emission was observed at high loadingsfor naphthalene[72,73,75] in Na+-Y and K+-L, and forbiphenyl [75] in K+-L. The loading-dependent excitation


spectra clearly suggested that the excimer formation is staticfor these species in the zeolites. Two noteworthy picturesemerge from the spectroscopic evidence suggesting thatthe configuration of excimer in Na+-Y should be differentfrom that in K+-L: first, a fully overlapped structure is ex-pected in Na+-Y in contrast to a partially overlapped one inK+-L. Second, the excimer emission of biphenyl observedin highly loaded K+-L was totally unobserved in Na+-Y. Inthese cases, the matching of pore size and shape with theguest species is considered decisive.

The observation of dimer cations is another example ofintracage processes. A study in our laboratory[71,72] re-vealed the transient absorption spectra of naphthalene to bedramatically dependent on the loading levels in hydratedNa+-X and Na+-Y (Fig. 15).

The hydrated zeolites were employed, because the tran-sient absorption band due to Na4

3+, which strongly inter-fere with other bands, can be suppressed. At the low load-ing level of 1.0×10−5 mol/g ([naphthalene]/[supercage]=0.017), the transient spectra were obtained that were ascrib-able to the superposition of triplet state naphthalene, theradical cation of naphthalene (Nap•+) and trapped electronsin the form of both Na32+ and Na2+ (λ∼700–900 nm). Atloading level 50 times higher (5.0×10−4 mol/g), absorptiondue to the naphthalene radical cations was greatly reduced,and a new broad, featureless absorption band appeared witha peak at 590 nm and a new band also appeared with aλmaxat 1100 nm. These absorption bands are reminiscent of thatof naphthalene dimer cations, which are formed by the asso-ciation of Nap•+ and naphthalene (Nap) in the ground-state:

Nap•+ + Nap→ (Nap)2•+

Dimer cations are characterized by their charge-resonance(CR) absorption band, which corresponds to the transi-

Fig. 15. Transient absorption spectra of naphthalene adsorbed in hy-drated Na+-Y zeolite. at various delay times (λex = 266 nm): (A)1.0 × 10−5 mol/g naphthalene (13 mJ/cm2); (B) 5.0 × 10−4 mol/g naph-thalene (13 mJ/cm2).

tion between two split highest occupied molecular orbital(HOMO) levels generated by the interaction between aradical cation and a neutral chromophore, and this bandis located at a significantly low energy (1100 nm) with re-spect to the visible dimer band (the 590 nm band for Nap).The dimer cations have been observed for naphthalenein �-irradiated low-temperature glassy matrices[162,163]and pulsed-electron irradiated solutions at high concen-trations at room temperature[164,165]. They were alsoreported in radiation-treated or photoirradiated solutionsof naphthalene-pendanted polymers at room temperature[166,167]. The naphthalene dimer cations in Na+-X andNa+-Y are significantly longer lived than those in solutionsat room temperature, and this is apparently due to the effectof confinement within the cages. The constraint providedby the zeolite medium on the guest species is expectedto be analogous to that in low-temperature glasses or intethered systems such as polymers, in which restriction ofboth translational and orientational motions of molecules isimposed.

The formation of dimer cations of other aromatic specieswithin the zeolite Na+-Y has not been demonstrated. Thiswas ascribed to the smaller stabilization energy of the dimercations of other aromatics compared with that of naphtha-lene. Very recently, the CR-band of the dimer cations ofanthracene (λmax = 1500 nm) were observed on excitationof the ground-state dimer formed in K+-L [168]. The re-sult suggests that it is important to hold the dimer pairs verytightly in matrices to stabilize the dimer cations. It is intrigu-ing to generate dimer cations at room temperature utilizingthe zeolite framework.

4.2. Electronic energy transfer

Carzaferri’s group devoted themselves to the study of en-ergy transfer and energy migration in dye-doped zeolites,aiming at the construction of artificial antenna systems mim-icking chlorophylls[169–173]. They synthesized cylindricalzeolite L crystals ranging in length from 30 nm to 3�m andincorporated luminescent dyes into the channels of the crys-tal. A cylinder of 600 nm diameter, for example, consistsof about 100,000 parallel-arranged channels. Red-emittingdye molecules (oxonine, Ox+) were put at each end of thechannels filled with green emitting dyes (pyronine, Py+).Irradiation of light of appropriate wavelength on the

cylinder caused the absorption only by Py+. The excitationenergy then migrates between Py+ molecules along the axis


of the crystal, and is eventually trapped by Ox+, resultingin the emission of red light from Ox+. Due to the short dis-tances and the ordering of the electronic transition dipolemoments of the dyes with respect to the channels, the excita-tion energy was considered to be transported by Förster-typeenergy migration preferentially along the axis of the cylin-drical antenna to a specific trap. The main role of zeoliteframework is to provide the desired geometry for arrangingand stabilizing the incorporated dyes.

Calzaferri’s group investigated this energy transfer systemrather extensively[169–173]. For instance, they investigatedthe orientation of the S0 → S1 transition dipole moment ofthe molecules Py+ and Ox+, which have transition dipoleswhose orientations are parallel to the molecule’s long axis,incorporated in the channels of zeolite L crystals by meansof fluorescence microscopy and single-crystal imaging. Theresult indicated a cone-shaped orientation distribution ofthe transition dipole moments with a half-opening angle of72◦ around the crystal axis for Ox+. This in not consistentwith the structural model according to which the maximumcone half-angles possible for Py+ and Ox+ are 30 and40◦, respectively. The discrepancy between the geometricconsideration and the results of the fluorescence measure-ments was explained by assuming that these species areexposed to a considerable electrostatic field in the zeolitechannels. Other topics related to the energy transfer indye-loaded zeolite L systems can be found in the literature[169–173].

Another strategy for realizing an energy transfer-basedantenna effect is to exploit rare earth ion-exchanged zeo-lites, mimicking rare earth chelates[174], in which ligandsfirst absorb light, followed by energy transfer to the rareearth ion. As such, the group of Garcıa and co-workersattempted to prepare rare-earth complexes of Eu3+ in-troduced as the charge-compensating cations with lig-ands, 2,2′-bipyridine (bpy), benzoiltrifluoroacetone (btfa),1,10-phenanthroline (phen) and dipicolinic acid (dpic) inzeolites, Y, mordenite and ZSM-5[175]. Formation of thecomplexes was assessed by various indirect methods. Itwas found that the complexes formed in the zeolites ap-peared deficient in the number of ligands compared to thosein solution, presumably due to steric constraints from theframework. Upon laser excitation at 266 nm, the zeolitessamples exhibited emission at room temperature and theluminescence intensity increased notably upon complexa-tion. The emission decay did not follow simple first-orderkinetics, but apparently the luminescence is longer livedinside the zeolite than in solution. These authors set ahigh value to the ability of the zeolite to immobilize thecomplex to such an extent that the emission is more than10-fold longer lived than that recorded for the same typeof complex in solution. Their pioneering work applied therare earth-exchanged zeolites to photochemical use and ob-served the sensitization of the emission of rare earth ionson excitation of organic ligands. Nevertheless, their analy-sis was not in-depth. For example, they observed that the

luminescence decays more rapidly as the Eu3+ content ofthe zeolite increases, but, they did not pursue an explana-tion. It might be interesting if they can detect the excitonicinteraction within zeolites. In addition, they observed thatthe lifetime of emission increases with the amount of theligand incorporated; Eu3+-doped zeolites with no ligandare much shorter lived. Furthermore, they observed thatthe nature of the ligand influences the decay of the lumi-nescence. However, no explanation was given for theseremarkable observations. They should have looked at therise of emission to see if they could detect the energytransfer process from the ligands to Eu3+and determine theenergy transfer rate constant. Also, it is necessary to deter-mine whether the energy transfer is a reversible process orirreversible one, because the emission lifetime of the rareearth ion is affected by the excited-state lifetimes of theligands, usually triplet state lifetimes if the energy transferis reversible.

With an idea similar to that of Garcıa’s group, Wada et al.made a considerable effort to increase the emission quantumyield of Nd3+, which rarely emits in fluid media due to thefast relaxation of its excitation energy through non-radiativeprocess and collisions[176]. They fabricated a complex ofNd3+ with bis (perfluorooctylsulfonyl)amine as a ligand inzeolite Y. Additionally, they replaced the –OH groups on thesurface of the zeolite cages

with –OD groups since it was expected that the –OH groupswith high vibrational frequency coordinated to Nd3+ shouldinduce rapid de-excitation of the excited-state of Nd3+. Aremarkably increased emission quantum yield of about 0.1was obtained. They attributed the success mainly to the sup-pression of the relaxation of the excitation energy of Nd3+by the low-vibrational zeolite cage wall and the energy mi-gration through collisions by locating Nd3+ separately indifferent cages.

More recently, a study from our laboratory[177] inves-tigated the energy transfer between guest aromatic speciessuch as benzophenone and naphthalene, and a rare earthcharge-compensating cation, Tb3+, within the zeolites Yand L by the time-resolved measurement of emission fromthe 5D4 excited-state of Tb3+. Although these moleculesdo not have the ability to form a complex with Tb3+ in the


ground-state, a remarkable sensitization of Tb3+ lumines-cence was observed, presumably due to a supramolecular ef-fect. The observation of the rise and decay of the emission ofTb3+, together with the measurement of triplet decay kinet-ics enabled the determination of the energy transfer rate con-stant for the naphthalene/zeolite Y system. The study showedthat, while irreversible energy transfer takes place fromtriplet benzophenone (actually, the protonated form in thezeolite) to Tb3+, reversible energy transfer occurs betweentriplet-state naphthalene and Tb3+, where the separation be-tween triplet level and the5D4 level of Tb3+ is small. The re-markable role of the zeolite is ascribed to the enforcement ofthe formation of a complex between the aromatic species inthe excited-state and Tb3+. Additionally, in highly loaded ze-olite L, the excitonic interaction of Tb3+ was observed in theexcited-state.

Although not tried so far, sensitization of rare earthemission might be possible via energy transfer upon ex-citation of the matrix, i.e. the zeolite framework. Such apossibility has already been realized in several other sys-tems. For instance, the energy transfer from an excitedmatrix to rare earth ions, leading to sensitized emission,has been observed in a Eu3+-doped silica gel/CdS nanopar-ticle composite[178], in the two-phase structure of or-dered TiO2 nanoparticles doped with Eu3+ [179], and inEu3+-doped thin films of GaN on Si substrates[180], forinstance. The literature dealing with these relatively newtypes of luminescence sensitization of rare earth ions hasbeen surveyed by Calzaferri and co-workers very recently[181].

4.3. Charge transfer and electron transfer

CT and ET have attracted considerable attention dueto their practical application for energy storage utilizingcharge separation and to the fundamental nature of thesereactions in chemistry. Microheterogeneous supports suchas zeolites have potential use for aiding the charge sepa-ration process by impeding energy wasting back-reactions.Aluminosilicate zeolites have shown considerable promisefor promoting the stabilization of photochemically gener-ated redox species. The arrangement of cages and channelsin these crystalline zeolites allows for the placement ofmolecules in well-defined, unique spatial arrangements.Here, we examine how zeolites work for CT and ETphenomena.

Adsorption of molecules with electron-donating natureand those with electron-accepting nature within the zeoliteshas led to the formation of intracage CT complexes due tothe attractive force between the donor–acceptor (DA) pairs.The area was pioneered by Yoon et al. by incorporatingarene donors into various pyridinium cation-exchanged ze-olites [42–45]. Later, the incorporation of a neutral form ofboth donors and acceptors was found to induce the CT com-plexes[157,182]. Although zeolites impose restrictions onthe size and shape of molecules to be incorporated, the for-

mation of CT complexes inside the cavities is unconstrainedby the zeolite framework, and the complexes exhibit prop-erties similar to those formed in solution. For instance, theposition of the CT absorption peaks is represented by a lin-ear function of the ionization energy of the donors or theelectron affinity of the acceptors, as predicted by Mullikentheory. Nevertheless, the properties of these CT complexeswere somewhat affected by various factors: the frameworkstructure, the degree of hydration, and the basic nature ofthe zeolite.

The excited-state properties of the CT complexes are ofconsiderable interest, because the investigation can be madeof the fundamental aspects of photoinduced ET reactionswithin zeolites. The excitation of the CT complexes of neu-tral DA pairs normally leads to the CT state (D+A−), fol-lowed by an ultra-fast charge-recombination (CR) reactionor a back-electron transfer (BET) process[183–186]. Yoonand co-workers observed that the photoexcitation of theMV2+–arene and other pyridinium–arene CT complexes en-capsulated in Y leads to the evolution of a broad transientabsorption due to the superposition of the bands of MV•+and arene•+ within the duration (25 ps) of the laser pulse[187,188]. The decay rate (BET process) is at least 10 timesslower than that in acetonitrile solution (k > 4 × 1010 s−1).Since the transient absorptions lasted more than 4 ns, theyalso investigated the behavior on longer time scales withnanosecond laser excitation. The combined picosecond andnanosecond kinetic data establish that there are at least threekinetically distinguishable decay phases of the transients, i.e.one decay process monitored on the picosecond time scalewith a lifetime between 45 ps and 1.2 ns, one process moni-tored on the microsecond time scale with a half-life between1.6 and 130�s, and one process with a lifetime greater than1 ms.

The authors have proposed several factors to rationalizethe remarkably long lifetimes of the transients. It is thoughtthat the electrostatic field within the zeolite supercage, aswell as the adsorption effects of ionic species at the polaraluminosilicate surface, are factors that contribute to theslowing of the BET. Coulombic interactions between thepositively charged transients and the negatively chargedframework may explain the incomplete or retarded BETcompared with the homogeneous solution. The positivelycharged transients may be pulled further apart from one an-other by interacting with the negatively charged framework.Furthermore, the authors think that the negatively chargedmicroenvironment of the zeolite may also influence thereduction and oxidation potentials of the electron acceptorand electron donor, respectively. The CR decays of excitedCT complexes consisting of 1,2,4,5-tetracyanobenzene andarene donors[157] were pursued by means of a femtosec-ond laser photolysis technique. Significantly slower decaykinetics than those in solutions of polar solvents wererecorded in the zeolites (Fig. 16).

For the CR decays, the restricted conformational relax-ations in the excited-state CT complexes adsorbed in the


Fig. 16. Time-dependent decays of transient absorption in dehydrated (A–D) and 1.0×10−2 mol/g hydrated (E–G) Na+-Y: (A) TCNB-naphthalene; () at457 nm (TCNB•−); ( ) at 682 nm (Nap•+); (B) TCNB-phenanthrene; () at 460 nm (TCNB•−); (C) TCNB-pyrene; () at 452 nm (TCNB•− and Pyr•+); (D)TCNB-anthracene; () at 740 nm (Ant•+); (E) TCNB-naphthalene; () at 460 nm (TCNB•−); ( ) at 685 nm (Nap•+); (F) TCNB-phenanthrene; () at460 nm (TCNB•−); (G) TCNB-anthracene; () at 735 nm (Ant•+).

zeolites were considered to be responsible for the remark-ably slow processes. Thus, two types of mechanisms wereproposed for the slow relaxation of the excited CT com-plexes: either the electrostatic field inside the zeolite or thespecific adsorption interaction between the guest species andthe adsorption sites in the zeolite. At the present stage, itis difficult to pinpoint which is the predominant factor, be-cause it seems to depend on the chemical nature of both theguest molecules and that of the zeolite.

A laser photolytic investigation of the CT complex be-tween cyclobis-(N,N′-paraquat-p-phenylene) macrocycle(CV4+) and 1,4-dimethoxybenzene (DMB), characterizedby a broad CT band atλmax of 475 nm, encapsulated withinthe supercages of zeolite Y was carried out by Garcıa andco-workers[189].

On excitation at 266 nm, the wavelength at which the vio-logen species embedded in CV4+ mainly absorbs the lightand thus away from the location of the CT band, the yieldof ET to produce viologen radical cation (CV•(3+)) de-creased dramatically compared with the system containing

no DMB. Note here that the excitation of viologens resultsin the abstraction of electrons from the zeolite host, asmentioned inSection 3.3. In contrast, 532 nm excitationallowed the detection of the CV•(3+) radical cation andthe dimer radical cation of DMB, (DMB)2

•+, as long-livedtransients whose decay was not complete after hundredsof microseconds following laser excitation. The result wasconsidered to indicate that the initially complexed DMB•+migrates out of the macroring and becomes stabilized byinteraction with an electron-rich neutral DMB, formingthe dimer cation. Since this was not the case in acetoni-trile solution, due to a fast back reaction, these authorsconsidered that the photochemical behavior is distinctiveof restricted media. The study demonstrated that stabiliz-ing the product of ET by opening a new pathway is oneof the remedies for retarding BET. Note that the pho-toinduced formation of the dimer cations on excitation ofCT complexes was once sought by Yoon and co-workers,but they failed to observe this by exciting CT complexesof arene donors in pyridinium ion-exchanged zeolites[188].

Cozens and co-workers explored ET and charge migra-tion between electron donors and acceptors encapsulatedwithin dry Na+-Y zeolites using nanosecond laser pho-tolysis [190]. Photoexcitation of chloranil (Chl) in Na+-Ycontaining co-adsorbed 4,4′-dimethxybicumene (DMBC) asan electron donor generated the long-lived radical anion(Chl•−) and 4-methoxycumyl cation, resulting from the de-composition of DMBC•+,


while photoexcitation oftrans-anethole (An) in Na+-Ycontaining co-adsorbed 1,4-dicyanobenzene (DCNB) as anelectron acceptor produced An•+, DCNB•− and trappedelectrons (e.g. Na43+, Na3

2+) in zeolites. No evidence forthe formation of CT complexes was found in either system.

The authors proposed that the fast generation and migrationof electrons and holes within the cavities of dry Na+-Y leadsto long-lived charge-separated species, rather than the directredox reactions between donor and acceptor pairs. Note that,in the absence of the quencher, both Chl•− and An•+ wereproduced on photoexcitation by the reaction with the zeolitehost, and they are long-lived. The big disadvantage of thisexploration is a lack of quenching data to support their pro-posal, due to the non-emissive nature of the excited states.It should be determined whether or not the singlet and/ortriplet excited states indeed react with these quenchers. Inthis case, a picosecond or femtosecond transient absorptionstudy is needed. An alternative simple explanation for thedata is that An•+/DCNB•− pairs are long-lived, because thetrapped electrons in dry zeolites are essentially unreactivetoward guest aromatic species, as was shown in the previousstudy [72], and Chl•− is long-lived, because the oxidizedcounterpart of the ET, DMBC•+ decomposes to a more sta-ble carbocation and radical.

Photoinduced electron transfer (PET) between tris-bipyridyl ruthenium(II) (Ru(bpy)32+, bpy: 2,2′-bipyridine)–bipyridinium acceptors has been extensively investigatedas a prototypical system to demonstrate the possibilitiesfor long-term charge separation. Mallouk and Kim demon-strated that PET readily takes place from Ru(bpy)3

2+ on theexternal surfaces of zeolites (Y, L, and mordenite) to MV2+

ions incorporated within the pores[191]. The PET takesmuch more readily whenN,N′-ethylene-2,2′-bipyridinium(2DQ2+) is tethered, as the electron relay, to one of thebpy ligands of the Ru(II) complex and when relay is in-tercalated into the channels of L incorporating viologen

acceptors[192]. The results show that the spatial organiza-tion of the photosensitized donor on the external surfaces andcorresponding acceptor on the internal surfaces of the zeoliteleads to much longer-lived charge separation, as comparedto those in which DA pairs are placed in the neighboringcages of Y[193] or in homogeneous solution.

Dutta and co-workers carried out a diffuse reflectancelaser photolysis investigation of PET from Ru(bpy)3

2+ syn-thesized in zeolite Y supercages to various ion-exchangedbipyridinium acceptors (with differing reduction poten-tials) residing in neighboring supercages[194]. Since thesize of Ru(bpy)32+ does not allow its exit from the su-percage, ET over a distance longer than the separationbetween adjacent cages (0.5–2 nm, depending on the ad-sorption sites within the cages and the relative orientationsof the molecules) may not be possible during the lifetimeof Ru(bpy)32+∗ (0.67�s). Although the measurement ofthe rate of forward ET was not possible due to the timeresolution of the detection system (lower limit >107 s−1),the rate constants for the BET were measured to be on theorder of 104 s−1 at wavelengths between 360 and 390 nm,representing the bleaching signal of Ru(bpy)3

2+ and thedecay signal of the bipyridinium radical ions due to BET.A significant observation is the formation of long-livedcharge-separated species at high loadings of bipyridiniumacceptors, and this was ascribed to the presence of a route


for charge propagation by electron hopping via denselypacked viologen species to compete with the BET. Theship-in-a-bottle synthesis of Ru(bpy)3

2+ in zeolite Y wasa very successful example of how nicely the crystallinezeolite allows for placement of molecules in well-definedand unique spatial arrangements within the cages[82,83].It seems, however, that there still remains room for in-creased sophistication of the arrangement of moleculeswithin the zeolite, because the experimental observationsare still complicated and it is difficult to extract concreteconclusions.

Bossmann et al. demonstrated that TiO2 nanowires in-corporated within the interior of Y also serve as an electronrelay between the zeolite-encapsulated Ru(bpy)3

2+ and asize-excluded Co(III) complex placed in solution[195]. Asimilar system was investigated by Garcıa and co-workerswith the transient absorption spectroscopy[196]. Samplesof zeolite Y containing 2,4,6-triphenylpyrylium (TPP+) orRu(bpy)32+ were prepared, and their photophysical andphotochemical properties were compared with the samesamples containing encapsulated nanoscopic TiO2 clustersas relays. Upon irradiation of TPP+ and Ru(bpy)32+ in thepresence of TiO2, the transient generation of the pyranylradical (TPP•) and Ru(bpy)33+ were observed. These largesensitizers are considered to remain immobilized insidethe cavities. Based on a time-resolved fluorescence study,these authors concluded that the quenching of TPP+∗ orRu(bpy)32+∗ by TiO2 is due to a static mechanism, tak-ing place instantaneously after dye excitation, suggestingthat only those excited photosensitizer molecules in di-rect contact with TiO2 clusters are quenched due to PETfrom Ru(bpy)32+ to TiO2 (electron injection) and alsofrom TiO2 to TPP+ (hole injection). Cases of the formertype are well known, but those of the latter type are lessdocumented.

Quite recently, Yoon and co-workers reported the PETfrom Ru(bpy)32+ encapsulated in K+-exchanged zeoliteY to the size-excluded and externally placed viologen(N-[3-(dicyclohexylmethyl)oxypropyl-N′-methyl-4,4′-bipy-ridinium [DHC-MV2+]] in acetonitrile [197]. Notably, K+ions were liberated from the zeolite to solution during in-terfacial ET from Ru(II) complexes to DHC-MV2+; theyform strong host–guest complexes with added crown ethermolecules in solution, which leads to retardation of thereverse flow of cations, and hence, the charge-balancingelectrons, from the solution to the zeolites.

Despite the absence of electron relay in the zeolite, theyield of DHC–MV•+ reached more than∼50 times theamount of Ru(bpy)32+ situated in the outermost supercages.

This is attributed to the photosensitized electron pumpingfrom the zeolite framework to the viologen by the outer-most Ru(bpy)32+ ions. This finding is surprising, becausesacrificial electron donors usually necessary for PET inRu(bpy)32+–DHC–MV2+ pairs are totally absent in thezeolite, suggesting that the zeolite framework serves as anunlimited source of electrons. Several points need to beclarified in this system. First, it is not clear what is themechanism of electron pumping in the zeolite or whichpart of the zeolite is responsible for providing electrons toRu(bpy)33+ to regenerate Ru(bpy)3

2+. Some concern maybe raised that the zeolite framework may undergo decom-position after giving up electrons to compensate for the lossof K+ ions and electrons. Second, in order to explain thehighly efficient PET in the zeolite, these authors consideredthe injection of electrons from Ru(bpy)3

2+∗ to the conduc-tion band of the zeolite, followed by the abstraction of theelectrons by DHC–MV2+ at the terminus of the frameworkas an alternative mechanism to the ET between DA pairs indirect contact. If this is the case, Ru(bpy)3

2+∗ should un-dergo quenching by the zeolite even in the absence of violo-gen; we suppose that this may not be the case, based on thelifetime data. Zeolites can be regarded as semiconductors,but the exact nature has not been characterized at this mo-ment[198]. Third, although Yoon and co-workers made anamazing observation of thermal ET also taking place fromthe zeolite framework to DHC–MV2+, the exact reducingagent is unidentified. In spite of these ambiguities, Yoon’sgroup showed that PET in zeolites is highly promising inattaining extremely high yields of charge separation.

The study of ET and related subjects will continue to be anactive area of research in zeolite photochemistry, partly fromthe perspective of promoting charge separation and partlyfrom that of the remaining issues to be solved by researcheffort.

5. Reactions through intercage migration

Intracage reactions such as excimer formation, charge-and electron-transfer, and singlet energy transfer, are suf-ficiently fast processes taking place within the ps and nstime scale. On the contrary, triplet–triplet energy transferand radical–radical combination in zeolites are very slowand bear peculiar decay kinetics distinct from first- orsecond-order reactions, because the reactants have to travelthrough many cages before getting access to a reactingpartner, and, more importantly, the distribution of reactantsis heterogeneous. These reactions were investigated experi-mentally and analysis was given to the results based on thetheoretical model.

Scaiano et al.[199] examined the quenching of tripletxanthone (Xan) by 1-methylnaphthalene (MeNap) throughthe triplet energy transfer mechanism in the faujasiteNa+-Y, with a diamond lattice structure of supercagenetworks.


They focused the reaction yields within 100 ns, wherethe cage-to-cage molecular transport was considered neg-ligible, and thus only quasi-static energy transfer wasassumed. In this case, 100 ns was the limitation of thedetection system used for the transient absorption mea-surements. Accordingly, the initial yield of3Xan∗ in thedecay curves was observed to decrease with increasingloading level of the quencher, MeNap. Three models weretested for interpreting the variations in triplet yield withquencher concentration: (1) donor and acceptor must sharethe same cage for the energy transfer to occur; (2) en-ergy transfer can occur in the same cage and in any of itsfour immediate neighbors (five supercages); and (3) trans-fer is also possible in second-nearest neighbor cages (fora total of 17 supercages). The experimentally observedyields of 3Xan∗ as a function of quencher occupancy, i.e.[MeNap]/[supercage] was best fitted by a equation which isbased on model 2. Thus, Scaiano et al. concluded that thesphere for energy transfer includes the original excitationsupercage and its four nearest-neighbor supercages, butno more.

The study gave some insight into the effective range oftriplet–triplet energy transfer, i.e. about 1 nm in the zeolitesystem. Yet, too simplified a model was used for the descrip-tion of the energy transfer in such a complicated system.For example, the experimental data points appear to devi-ate on the side of larger values from the simulated curve asthe occupancy of the quenchers increases. Apparently, 100%quenching cannot be achieved at an occupancy of unity, eventhough the model requires this condition. This is because themodel counts the probability of the absence of a quencherin any five supercages, not necessarily the cages consist-ing of one with the triplet probe and the four immediatelyneighboring cages. More elaborate analysis based on a so-phisticated model is necessary for resolving an interestingbut challenging problem of energy transfer distance withinthe zeolite cage networks. Also, the time-dependent decaycurves of3Xan∗ should be fully analyzed, which may pro-vide useful information on the mechanistic aspect charac-teristic to the zeolite, since the assumption of 100 ns for thetime limitation to which the quasi-static quenching applieshas no solid ground.

Quenching of photoexcited triplet-state anthracenethrough triplet–triplet energy transfer mechanism by a fewenergy acceptors, including azulene, ferrocene and alsoanthracene in the ground-state, was investigated in dehy-drated Na+-Y zeolite with a diffuse reflectance transientabsorption technique by our group[68,200].

At sufficiently the low loading levels of≤5.0×10−6 mol/g,the decay of triplet-state anthracene was approximated by asingle exponential function with a decay constant of 54±5 s−1. Upon addition of the quenchers, dynamic quenchingof triplet-state anthracene was observed. A faster decay wasobserved with increasing loading level of quenchers, and thedecay curves were non-exponential (Fig. 17).

The probe molecules are expected to either self-deactivateor undergo quenching through collision with another dopedspecies (a quencher) continuously migrating within the cagenetworks, since the triplet–triplet energy transfer proceedsmainly through a collisional mechanism due to exchangeinteraction [201]. The general method of time-resolveddiffusion-mediated excited-state deactivation has alreadyfound application for probing dynamical and structuralproperties of a variety of microheterogeneous media, suchas colloids[67]. The kinetics of the excited-state quench-ing in zeolites are most conveniently described in terms ofthe CTRW model[202–204]. In its simplest but still quiterealistic formulation, it assumes that quenchers performindependent random walks that involve unbiased jumps tonearest-neighbor lattice sites, and the waiting time distribu-tion function in each site is exponential. The lattice site isnormally associated with a zeolite supercage, but general-ization to true adsorption sites is possible. Supercages areassumed to be occupied by reactants at random, accordingto the Poisson law. The reaction is assumed to take placewhenever the probe and the quencher are in the same su-percage. For open zeolite structures, quenching may alsooccur when the probe and the quencher reside in neighbor-ing supercages, as was pointed out by Scaiano et al.[199].The final expression for the experimentally observableexcited-state survival probabilityΦ(t) is given by[200]

−ln Φ(t) = k0t + nR(t), (2)

where k0 is the excited-state self-decay lifetime,n is theaverage number of quenchers per supercage ([quencher]/[supercage], typically,n 1), andR(t) is the time-dependentreaction flux. The analytical expression forR(t) includeskq, the pseudo-first-order rate constant for the intracagequenching;km, the rate constant for the intercage migration;andk′

q, the rate constant for quenching from the neighboringsupercage (Fig. 18).

It follows from Eq. (1) that the function

F(t) = −R(t) = 1

n[ln Φ(t) + k0t ] (3)

must be universal for any quencher concentration, as long aslow quencher concentrations are employed(n 1). F(t) can


Fig. 17. (a) Experimental decay curves of triplet-state anthracene in dehydrated Na+-Y zeolite at various loading levels of azulene: (A) 0 mol/g; (B)7.6× 10−6 mol/g; (C) 1.8× 10−5 mol/g; (D) 3.0× 10−5 mol/g. The loading level of anthracene was 5.0× 10−6 mol/g, and the excitation wavelength was355 nm, while the monitoring wavelength was 420 nm. The ordinate represents the normalized absorption measured in the diffuse reflectance mode. Theabsorption is defined by�I/I0 (I0 is the reflectance intensity before laser excitation, and�I is the change in the reflectance after excitation), which isproportional to the concentration. The values of�I/I0 at t = 0 are always less than 10%. The solid lines indicate the fitting based on the model. (b)Triplet–triplet absorption decay of anthracene in dehydrated Na+-Y zeolite excited at 355 nm and monitored at 420 nm, at various loading levels: (A)5.0 × 10−6 mol/g; (B) 1.0 × 10−5 mol/g; (C) 2.0 × 10−5 mol/g; (D) 1.1 × 10−4 mol/g.

be thought of as the decay function in the presence of just onequencher and in the absence of self-deactivation.Eq. (3)wasused as a test of the applicability of the present treatment tothe experiments.Fig. 19a and bshows the plots according toEq. (3), universal plots for two quenching systems, azuleneand ground-state anthracene, in dehydrated Na+-Y zeolite.

The experimentally observed universality suggests thatthe CTRW model is applicable to the quenching oftriplet-state anthracene.

Two significant findings were obtained from this tripletquenching study. First, self-diffusion coefficients were ex-tracted from the cage-to-cage migration rate constant,km forthe large aromatic molecules, anthracene, azulene, and fer-rocene in dehydrated Na+-Y. The very small values, rang-ing from 10−15 to 10−16 m2/s at room temperature, cannotbe measured by traditional techniques, i.e. PFG NMR andQENS. Thus, the triplet quenching method, which probes

Fig. 18. Quenching model in the zeolite. P∗ represents a probe molecule(triplet donor) and Q, a quencher (triplet acceptor), which migrates withinthe cage networks with a cage-to-cage jump rate constant ofkm. Two typesof quenching mechanisms are considered: first, quenching within a cage,where the probe and quencher molecules are both present (quenching rateconstant,kq), and second, that in which the quencher attacks from thefour direct neighboring cages (rate constant,k′

q).

longer time scales (ns to 100 ms), can be complementary toother established techniques in zeolite science. Second, bythe analysis of the experimental decay curves based on theCTRW model, the detailed reaction mechanism was revealedto depend on the zeolite lattice topology as well as on thedynamics of migration of guest molecules between adsorp-tion sites. In dehydrated Na+-Y zeolite, where connectingwindows are wide and the distance between adsorption sitesin the neighboring cages is small enough to allow sufficientoverlap of the wave-functions, the probes can be deactivatedby quenchers residing not only in the same cage but also inthe four cages of the first coordination sphere (seeFig. 20).

Interestingly, in hydrated zeolites, where co-adsorbed wa-ter molecules are expected to block the interconnecting win-dows between the cages, quenching from neighboring cagesis not possible. In future work, a temperature study mightbe needed, because the activation energy of diffusion can beobtained, the value of which might be discussed in compar-ison with that of the extensively studied benzene. Addition-ally, the validity of the order of the diffusion coefficientsobtained should be checked by other methods.

Johnston et al. reported an experimental and theoret-ical study of the recombination of diphenylmethyl rad-icals generated by flash photolysis of the precursors,1,1,3,3-tetraphenylacetone and 1,1-diphenylacetone inNa+-X zeolite [205]. The diphenylmethyl radical con-centration, monitored by time-resolved diffuse reflectancespectroscopy, decreased roughly linearly with the logarithmof time, indicative of a reaction rate that decreases grad-ually. The overall decay pattern depended only weakly ontemperature, precursor, or laser dose. For the first time,they applied a method based on the random walk conceptto the diffusive motion of the radicals in the zeolite lattice,a diamond structure of the linked cavities. The dominant


Fig. 19. (a) Universal plots for the quenching of triplet-state anthracene by azulene in dehydrated Na+-Y (Fig. 17a). F(t) is defined byEq. (3) in thetext. The loading level of anthracene was 5.0× 10−6 mol/g, and the mean occupancy numbers of azulene were () n = 0.012, (+) 0.029, and () 0.047.The self-decay rate constant isk0 = 50 s−1. (b) Universal plots for the experimental decay curves presented inFig. 17b. The mean occupancy numbersare ( ) n = 0.016, (+) 0.032, and (—) 0.18. The self-decay rate constant isk0 = 35 s−1.

decay mechanism is geminate recombination, as evidencedby product analysis. For short times (200 ns–10�s) andlow precursor concentrations, the rate of geminate recom-bination generated by the model yielded a satisfactoryreproduction of the observed time dependence. However,for long times, the fraction of radicals surviving gemi-nate recombination falls well below the theoretical limit of51%. They reasoned that quenching processes other thannon-geminate recombination are responsible for this behav-ior, although they did not suggest the exact mechanism. Wesuggest that, due to the high polarity of the zeolite mediumand an appreciable absorption of 266 nm light by the hostzeolite, photoionization of both the precursors and Na+-Xzeolite can produce radical cations and trapped electrons,which may interfere with the detection of the radicals[206].Nevertheless, their mathematical treatment is rigorous andconstitutes a prototype for theoretical analysis in reactionsinvolving intercage migration of molecules.

Cozens et al. reported the first detection of benzyl radicalsby the transient diffuse reflectance spectrum generated upon266-nm laser excitation of phenylacetic acid incorporated in

Fig. 20. Schematic diagram of two adjacent Na+-Y zeolite supercageswith embedded anthracene (triplet donor, left) and azulene (quencher,right). Four tetrahedrally arranged SII Na+ ions (spheres) serve as primaryadsorption sites for guest aromatics.

dry Na+-Y [207]. Laser photolysis of dibenzyl ketone, usu-ally used in solution studies to generate the radical, was notsuccessful in Na+-Y for the observation of the radical, ham-pered by other unknown transient species. Although detailedanalysis was not carried out on the decay kinetics, the decayof the benzyl radical is non-exponential, ranging from�s toms and also depends on the charge-compensating cations:the decay is faster in the order Li+-Y > Na+-Y > K+-Y >Rb+-Y > Cs+-Y. The explanation is given that the benzylradical essentially decays by a coupling reaction with a sec-ond benzyl radical, and the mobility decreases along the se-ries from Li+-Y to Cs+-Y; yet, some portion of the radicalsgenerated decays very slowly, because they are generatedwithin local domains, where diffusion is restricted. It mightbe interesting if the decay kinetics of the benzyl radical areanalyzed in terms of the CTRW model[68,200]because wecan have a better picture of the diffusional motion of theradicals in the lattice.

The elucidation of the kinetics and mechanisms of reac-tions involving intercage migration of adsorbed species isimportant for catalytic and synthetic applications of zeolites.More research effort is needed from both the experimentaland theoretical points of view, to accumulate basic data suchas diffusivity and reaction yields, which are beneficial topeople who want to use zeolites for practical applications.

6. Summary

We have shown an overview of the research activities inphotochemical studies of organic molecules adsorbed in ze-olites carried out over the past 10–15 years, mainly withtransient spectroscopic techniques. Transient spectroscopicstudies in zeolites, with their structural integrity and tunablechemical properties have brought to light a great deal of use-


ful information on the excited-state behavior of molecules,in particular, dynamic processes mechanistically differentfrom those in homogeneous solutions. This approach hashelped to clarify the zeolite medium considerably from thephotochemical point of view and has made a certain impacton the zeolite science community in terms of the types ofinformation obtainable by various experimental techniques.Research efforts utilizing zeolites as host materials also haveindicated the way of controlling the photophysics of ad-sorbed guest species. Many examples of such cases dealingwith the “control of photophysics” were given as well asthose dealing with the “control of photochemical reactionpathways”. Thus, the interplay with zeolites will continue tobe of great benefit to the photochemistry community. Fromthe viewpoint of methodology, transient spectroscopy witha diffuse reflectance detection technique still has experi-mental limitations and technical issues to be overcome forproviding experimental data with reliability similar to thatof transmittance mode spectroscopy, despite the increasingpopularity. Additionally, developing theoretical treatmentsthat describe complex kinetic behavior of excited specieswithin zeolites on the basis of a simplified picture are veryimportant. Given the efforts being carried out from both ex-perimental and theoretical standpoints, we will see furtherdevelopments in this area in the next few years.

Acknowledgements

We are grateful to the following sources of financialsupport: the Izumi Science and Technology Foundation(grant no. H13-J-62), the Iketani Science and TechnologyFoundation (grant no. 011023-A), the Japanese Ministryof Education, Culture, Sports, Science and Technology(Grant-in-Aid for Scientific Research, Priority Area 417,#14050100), and the Japan Society for the Promotion ofScience (Grant-in-Aid for Scientific Research #14550796).

References

[1] N.J. Turro, Pure Appl. Chem. 58 (1986) 1219.[2] V. Ramamurthy, in: V. Ramamurthy (Ed.), Photochemistry in

Organized and Constrained Media, VCH, New York, 1991, p. 429.[3] J.K. Thomas, Chem. Rev. 93 (1993) 301.[4] K.B. Yoon, Chem. Rev. 93 (1993) 321.[5] V. Ramamurthy, P. Lakshiminarasimham, C.P. Grey, L.J. Johnston,

Chem. Commun. (1998) 2411.[6] J.C. Scaiano, H. Garcıa, Acc. Chem. Res. 32 (1999) 783.[7] V. Ramamurthy, in: J.K. Whitesell (Ed.), Organic Molecular

Assemblies in the Solid State, Wiley, Chichester, 1999, p. 63.[8] G. Calzaferri, D. Brühwiler, S. Megelski, M. Pfenniger, M.

Pauchard, B. Hennessy, H. Maas, A. Devaux, U. Graf, Solid StateSci. 2 (2000) 421.

[9] N.J. Turro, Acc. Chem. Res. 33 (2000) 637.[10] K.B. Yoon, in: V. Ramamurthy, K.S. Schanze (Eds.), Solid State and

Surface Photochemistry, Marcel Dekker, New York, 2000, p. 143.[11] S. Hashimoto, in: V. Ramamurthy, K.S. Schanze (Eds.), Solid State

and Surface Photochemistry, Marcel Dekker, New York, 2000,p. 253.

[12] D.W. Breck, Zeolite Molecular Sieves, Krieger, Malabar, 1984.[13] A. Dyer, An Introduction to Zeolite Molecular Sieves, Wiley, New

York, 1988.[14] H. Van Bekkum, E.M. Flanigen, P.A. Jacobs, J.C. Jansen (Eds.),

Introduction to Zeolite Science and Practice, second ed., Elsevier,Amsterdam, 2001.

[15] G.A. Ozin, A. Kuperman, A. Stein, Angew. Chem. Int. Ed. Engl.28 (1989) 359.

[16] G.D. Stucky, J.E. MacDougall, Science 247 (1990) 669.[17] I. Braun, G. Ihlein, F. Laeri, U. Nöckel, G. Schultz-Ekloff, F. Schüth,

U. Veitze, Ö. Weiss, D. Wöhrle, Appl. Phys. B 70 (2000) 335.[18] F. Marlow, J. Caro, L. Werner, J. Kornaowski, S. Dähne, J. Phys.

Chem. 97 (1993) 11286.[19] C. Bärlocher, W.M. Meier, D.H. Olsen, Atlas of Zeolite Framework

Types, Elsevier, Amsterdam, 2001.[20] W. Loewenstein, Am. Miner. 39 (1954) 92.[21] H. Miessner, H. Kosslick, U. Lohse, B. Parlitz, V.-A. Tuan, J. Phys.

Chem. 97 (1993) 9741.[22] X. Yang, J. Phys. Chem. 99 (1995) 12761.[23] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge,

K.D. Schumitt, C.T.-W. Chu, D.H. Olson, E.W. Sheppard, S.B.McCullen, J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114(1992) 10834.

[24] E. Dempsey, Molecular Sieves, Society of Chemical Industry,London, 1968, p. 293.

[25] Y.-Y. Huang, J.E. Benson, M. Boudart, I & EC Fundam. 8 (1969)346.

[26] E. Preuss, G. Linden, M. Peuckert, J. Phys. Chem. 89 (1985) 2955.[27] B.H. Baretz, N.J. Turro, J. Photochem. 24 (1984) 201.[28] X. Liu, K.-K. Iu, J.K. Thomas, J. Phys. Chem. 93 (1989) 4120.[29] V. Ramamurthy, D.R. Sanderson, D.F. Eaton, Photochem. Photobiol.

56 (1992) 297.[30] N. Sarkaar, K. Das, D.N. Nath, K. Bhattacharyya, Langmuir 10

(1994) 326.[31] S. Uppili, K.J. Thomas, E.M. Crompton, V. Ramamurthy, Langmuir

16 (2000) 265.[32] T. Yashima, H. Suzuki, N. Hara, J. Catal. 33 (1974) 486.[33] J.A. Rabo, Catal. Rev. 23 (1981) 293.[34] A. Corma, Chem. Rev. 95 (1995) 559.[35] H. Hattori, Chem. Rev. 95 (1995) 537.[36] D.J. Barthomeuf, Phys. Chem. 88 (1984) 43.[37] M. Huang, S. Kaliaguine, J. Chem. Soc. Faraday Trans. 88 (1992)

715.[38] R.T. Sanderson, Chemical Bonds and Bond Energy, second ed.,

Academic Press, New York, 1976.[39] D. Barthomeuf, Catal. Rev. 38 (1996) 521.[40] R.C. Deka, R.K. Roy, K. Hirao, Chem. Phys. Lett. 332 (2000) 576.[41] S. Hashimoto, N. Fukazawa, H. Fukumura, H. Masuhara, Chem.

Phys. Lett. 223 (1994) 493.[42] K.B. Yoon, J.K. Kochi, J. Am. Chem. Soc. 111 (1989) 1128.[43] K.B. Yoon, J.K. Kochi, J. Phys. Chem. 95 (1991) 1348.[44] K.B. Yoon, J.K. Kochi, J. Phys. Chem. 95 (1991) 3780.[45] K.B. Yoon, T.J. Huh, D.R. Corbin, J.K. Kochi, J. Phys. Chem. 97

(1993) 6492.[46] S. Hashimoto, unpublished result.[47] T. Hirano, W. Li, L. Abrams, P.J. Krusic, M.F. Ottaviani, N.J. Turro,

J. Org. Chem. 65 (2000) 1319.[48] N.J. Turro, X.-G. Lei, W. Li, A. McDermott, M.F. Ottaviani, L.

Abrams, J. Am. Chem. Soc. 122 (2000) 11649.[49] A. de Mallman, D. Barthomeuf, J. Chem. Soc. Chem. Commun.

(1986) 476.[50] A. de Mallman, D. Barthomeuf, Zeolites 8 (1988) 292.[51] A. de Mallman, D. Barthomeuf, J. Chem. Soc. Chem. Commun.

(1989) 129.[52] P.J. O’Malley, Chem. Phys. Lett. 166 (1990) 340.[53] M.L. Unland, J.J. Freeman, J. Phys. Chem. 82 (1978) 1036.[54] M. Primet, E. Garbowski, M.V. Mathieu, B. Imelik, J. Chem. Soc.

Faraday Trans. I 76 (1980) 1942.


[55] L.C. de Menorval, S.B. Liu, D. Raftery, K. Takegoshi, R. Ryoo,A. Pines, J. Phys. Chem. 94 (1990) 27.

[56] R. Ryoo, S.B. Liu, L.C. de Menorval, K. Takegoshi, B. Chmelka,M. Trecoske, A. Pines, J. Phys. Chem. 91 (1987) 6575.

[57] B.F. Chmelka, J.G. Pearson, S.B. Liu, R. Ryoo, L.C. de Menorval,A. Pines, J. Phys. Chem. 95 (1991) 303.

[58] J.G. Pearson, B.F. Chmelka, D.N. Shykind, A. Pines, J. Phys. Chem.96 (1992) 8517.

[59] S.B. Liu, L.J. Ma, M.W. Lin, J.-F. Wu, T.-L. Chen, J. Phys. Chem.96 (1992) 8120.

[60] A.N. Fitch, H. Jobic, A. Renouprez, J. Chem. Soc. Chem. Commun.(1985) 284.

[61] M. Czjek, T. Vogt, H. Fuess, Angew. Chem. Int. Ed. Engl. 28(1989) 770.

[62] A. Renouprez, H. Jobic, R.C. Oberthur, Zeolites 5 (1985) 222.[63] H. Jobic, A. Renouprez, A.N. Fitch, H.J. Lauter, J. Chem. Soc.

Faraday Trans. I 83 (1987) 3199.[64] P. Demonitis, S. Yashonath, M.L. Klein, J. Phys. Chem. 93 (1989)

5016.[65] H. Klein, C. Kirschhock, H. Fuess, J. Phys. Chem. 98 (1994) 12345.[66] V. Ramamurthy, D.R. Sanderson, D.F. Eaton, J. Phys. Chem. 97

(1993) 13380.[67] A.V. Barzykin, M. Tachiya, Heterog. Chem. Rev. 3 (1996) 105.[68] S. Hashimoto, M. Hagiri, A.V. Barzykin, J. Phys. Chem. B 106

(2002) 844.[69] B.F. Chmelka, D. Raftery, A.V. McCormick, L.C. de Menorval,

R.D. Levine, A. Pines, Phys. Rev. Lett. 66 (1991) 580.[70] R.G. Larsen, J. Shore, K. Schumidt-Rohr, L. Emsley, H. Long, A.

Pines, Chem. Phys. Lett. 214 (1993) 220.[71] S. Hashimoto, Chem. Phys. Lett. 262 (1996) 292.[72] S. Hashimoto, T. Mutoh, H. Fukumura, H. Masuhara, J. Chem.

Soc. Faraday Trans. 92 (1996) 3653.[73] K.J. Thomas, R.B. Sunoj, J. Chandrasekhar, V. Ramamurthy,

Langmuir 16 (2000) 4912.[74] S. Hashimoto, S. Ikuta, T. Asahi, H. Masuhara, Langmuir 14 (1998)

4284.[75] S. Hashimoto, M. Hagiri, N. Matsubara, S. Tobita, Phys. Chem.

Chem. Phys. 3 (2001) 5043.[76] S.L. Suib, A. Kostrapopos, J. Am. Chem. Soc. 106 (1984) 7705.[77] X. Liu, K.-K. Iu, J.K. Thomas, J. Phys. Chem. 93 (1989) 4120.[78] S.B. Hong, H.M. Cho, M.E. Davis, J. Phys. Chem. 97 (1993) 1622.[79] J. Kärger, D.M. Ruthven, Diffusion in Zeolites and Other molecular

Sieves, Wiley, New York, 1992.[80] N.Y. Chen, T.F. Degnan, C.M. Smith, Molecular Transport and

Reaction in Zeolites, VCH, New York, 1994.[81] N. Herron, Inorg. Chem. 25 (1986) 4714.[82] W. DeWilde, G. Peeters, J.H. Lunsford, J. Phys. Chem. 84 (1980)

2306.[83] P. Laine, M. Lanz, G. Calzaferri, Inorg. Chem. 35 (1996) 3514.[84] S.M. Auerbach, L.M. Bull, N.J. Henson, H.I. Metiu, A.K.

Cheethaam, J. Phys. Chem. 100 (1996) 5923.[85] S.M. Auerbach, H.I. Metiu, J. Chem. Phys. 105 (1996) 3753.[86] S.M. Auerbach, H.I. Metiu, J. Chem. Phys. 106 (1997) 2893.[87] A. Germanus, J. Kärger, H. Pfeifer, N.N. Samlevic, S.P. Zdanov,

Zeolites 5 (1985) 91.[88] J. Kärger, H. Pfeifer, Zeolites 7 (1987) 90.[89] M. Czjzek, H. Jobic, M. Bee, J. Chem. Soc. Faraday Trans. 87

(1991) 3455.[90] H. Jobic, A.N. Fitch, J. Combet, J. Phys. Chem. B 104 (2000) 8491.[91] L.M. Bull, N.J. Henson, A.K. Cheetham, J.M. Newsam, S.J. Heyes,

J. Phys. Chem. 97 (1993) 1309.[92] J.A. Sousa Gonçalves, R.L. Portsmouth, P. Alexander, L.F. Gladden,

J. Phys. Chem. 99 (1995) 3317.[93] H. Klein, H. Fuess, G. Schrimpf, J. Phys. Chem. 100 (1996) 11101.[94] H. Lechert, W.D. Basler, J. Phys. Chem. Solids 50 (1989) 497.[95] D.E. Favre, D.J. Schaefer, S.M. Auerbach, B.F. Chmelka, Phys.

Rev. Lett. 81 (1981) 5852.

[96] J. Park, W.-K. Kang, R. Ryoo, K.-H. Jung, D.-J. Jang, J. Photochem.Photobiol. A: Chem. 80 (1994) 333.

[97] F. Marlow, D. Demuth, G. Stucky, F. Schüth, J. Phys. Chem. 99(1995) 1306.

[98] D. Brühwiler, N. Gfeller, G. Calzaferri, J. Phys. Chem. B 102(1998) 2923.

[99] V. Ramamurthy, J.V. Casper, D.F. Eaton, E.W. Kuo, D.R. Corbin,J. Am. Chem. Soc. 114 (1992) 3882.

[100] K.-K. Iu, J.K. Thomas, J. Phys. Chem. 95 (1991) 506.[101] K.-K. Iu, X. Liu, J.K. Thomas, J. Photochem. Photobiol. A: Chem.

79 (1994) 103.[102] F. Gessner, J.C. Scaiano, J. Photochem. Photobiol. A: Chem. 67

(1992) 91.[103] Y.S. Liu, P. de Mayo, W.R. Ware, J. Phys. Chem. 97 (1993) 5987.[104] Y.S. Liu, P. de Mayo, W.R. Ware, J. Phys. Chem. 97 (1993) 5995.[105] S. Hashimoto, T. Miyashita, M. Hagiri, J. Phys. Chem. B 103

(1999) 9149.[106] W.H. Melhuish, R. Hardwick, Trans. Faraday Soc. 58 (1962) 1908.[107] A.V. Buettner, J. Phys. Chem. 68 (1964) 3253.[108] E.L. Clennan, private communication.[109] V. Ramamurthy, Mol. Cryst. Liq. Cryst. 240 (1994) 53.[110] M. Pauchard, A. Devaux, G. Calzaferri, Chem. Eur. J. 6 (2000)

3456.[111] F.L. Corzens, M. Régimbald, H. Garcıa, J.C. Scaiano, J. Phys.

Chem. 100 (1996) 18165.[112] F. Wilkinson, R. Kessler, J. Chem. Soc. Faraday Trans. 1 77 (1981)

2073.[113] R.W. Kessler, G. Krabichler, S. Uhl, D. Oelkrug, W. Hagen, J.

Hyslop, F. Wilkinson, Opt. Acta 30 (1983) 1099.[114] F. Wilkinson, J. Chem. Soc. Faraday Trans. 2 82 (1986) 2073.[115] D. Oelkrug, W. Honnen, F. Wilkinson, C.J. Willsher, J. Chem. Soc.

Faraday Trans. 2 83 (1987) 2081.[116] F. Wilkinson, C.J. Willsher, Terahedron 43 (1987) 1197.[117] F. Wilkinson, G.P. Kelly, in: M. Anpo, T. Matsuura (Eds.),

Photochemistry on Solid Surfaces, Elsevier, Amsterdam, 1989,p. 30.

[118] F. Wilkinson, C.J. Willsher, P.A. Leicester, J.R.M. Barr, M.J.C.Smith, J. Chem. Soc. Chem. Commun. (1986) 1216.

[119] N. Ikeda, K. Imagi, H. Masuhara, N. Nakashima, K. Yoshihara,Chem. Phys. Lett. 140 (1987) 281.

[120] T. Asahi, A. Furube, H. Fukumura, M. Ichikawa, H. Masuhara,Rev. Sci. Instrum. 69 (1998) 361.

[121] W.W. Wendlandt, H.G. Hecht, Reflectance Spectroscopy, Wiley,New York, 1966.

[122] S. Takatani, H. Fukumura, H. Masuhara, S. Hashimoto, J. Phys.Chem. B 101 (1997) 3365.

[123] H. Hase, L. Keavan, J. Chem. Phys. 54 (1971) 908.[124] H.B. Steen, J. Moan, J. Phys. Chem. 76 (1972) 3366.[125] T. Shida, M. Imamura, J. Phys. Chem. 78 (1974) 232.[126] A. Furube, T. Asahi, H. Masuhara, Jpn. J. Appl. Phys. 38 (1999)

4236.[127] G. Calzaferri, N. Gefller, J. Phys. Chem. 96 (1992) 3428.[128] V. Ramamurthy, D.R. Sanderson, D.F. Eaton, J. Am. Chem. Soc.

115 (1993) 10438.[129] Y.S. Park, K. Lee, C. Lee, K.B. Yoon, Langmuir 16 (2000) 4470.[130] V. Ramamurthy, J.V. Casper, D.R. Corbin, D.F. Eaton, J. Photochem.

Photobiol. A: Chem. 50 (1989) 157.[131] H.L. Casal, J.C. Scaiano, Can. J. Chem. 62 (1984) 628.[132] F. Marquez, C.M. Zicovich-Wilson, A. Corma, E. Palomares, H.H.

Garcıa, J. Phys. Chem. B 105 (2001) 9973.[133] F. Marquez, H. Garcıa, E. Palomares, L. Fernandez, A. Corma, J.

Am. Chem. Soc. 122 (2000) 6520.[134] F. Marquez, V. Marti, E. Palomares, H. Garcıa, W. Adam, J. Am.

Chem. Soc. 124 (2002) 7264.[135] A. Corma, H. Garcıa, G. Sastre, P.M. Viruela, J. Phys. Chem. B

101 (1997) 4575.[136] M.L. Cano, F.L. Cozens, H. Garcıa, V. Martı, J.C. Scaiano, J. Phys.

Chem. 100 (1996) 18152.


[137] M.L. Cano, F.L. Cozens, V. Fornés, H. Garcıa, V. Martı, J.C.Scaiano, J. Phys. Chem. 100 (1996) 18145.

[138] J.C. Scaiano, M. Kaila, S. Corrent, J. Phys. Chem. B 101 (1997)8564.

[139] N.J. Turro, C.-C. Cheng, L. Abrams, D.R. Corbin, J. Am. Chem.Soc. 109 (1987) 2449.

[140] S. Jockusch, T. Hirano, Z. Liu, N.J. Turro, J. Phys. Chem. B 104(2000) 1212.

[141] T. Hirano, W. Li, L. Abrams, P.J. Krusic, M.F. Ottaviani, N.J. Turro,J. Am. Chem. Soc. 121 (1999) 7170.

[142] S. Hashimoto, N. Fukazawa, H. Fukumura, H. Masuhara, Chem.Phys. Lett. 223 (1994) 493.

[143] Y. Nosaka, A. Kira, M. Imamura, J. Phys. Chem. 85 (1981) 1354.[144] T. Nakamura, A. Kira, M. Imamura, J. Phys. Chem. 76 (1982) 3359.[145] H. Masuhara, H. Shioyama, T. Saito, K. Hamada, S. Yasoshoma,

N. Mataga, J. Phys. Chem. 88 (1984) 5868.[146] F.L. Cozens, H. Garcıa, J.C. Scaiano, J. Am. Chem. Soc. 115 (1993)

11134.[147] R.A. Kumpf, D.A. Dougherty, Science 261 (1993) 1708.[148] J.C. Ma, D.A. Dougherty, Chem. Rev. 97 (1997) 1303.[149] D.A. Dougherty, Science 271 (1996) 163.[150] K.-K. Iu, J.K. Thomas, Langmuir 6 (1990) 471.[151] F.L. Cozens, M.L. Cano, H. Garcıa, N.P. Schepp, J. Am. Chem.

Soc. 120 (1998) 5667.[152] X. Liu, K.-K. Iu, J.K. Thomas, Chem. Phys. Lett. 204 (1993) 163.[153] X. Liu, K.-K. Iu, J.K. Thomas, J. Phys. Chem. 98 (1994) 7877.[154] H.J.D. Mcmanus, C. Finel, L. Keavan, Radiat. Phys. Chem. 45

(1995) 761.[155] M. Alvaro, H. Garcıa, S. Garcıa, F. Marquez, J.C. Scaiano, J. Phys.

Chem. B 101 (1997) 3043.[156] S. Hashimoto, J. Chem. Soc. Faraday Trans. 93 (1997) 4401.[157] S. Hashimoto, N. Hagiwara, T. Asahi, H. Masuhara, Langmuir 15

(1999) 3123.[158] Y.S. Park, S.Y. Um, K.B. Yoon, J. Am. Chem. Soc. 120 (1999)

5667.[159] S.Y. Choi, Y.S. Park, S.B. Hong, K.B. Yoon, J. Am. Chem. Soc.

118 (1996) 9377.[160] M.A. O’Neill, F.L. Cozens, N.P. Schepp, J. Am. Chem. Soc. 122

(2000) 6017.[161] B. Stevens, Spectrochim. Acta 18 (1962) 439.[162] B. Badger, B. Brocklehurst, R.D. Russell, Chem. Phys. Lett. 1

(1967) 122.[163] A. Kira, T. Nakamura, M. Imamura, J. Phys. Chem. 81 (1977) 511.[164] A. Kira, S. Arai, M. Imamura, J. Phys. Chem. 76 (1972) 119.[165] M.A.J. Rodgers, J. Chem. Soc. Faraday Trans. 1 68 (1972) 1278.[166] Y. Tsujii, A. Tsuchida, S. Ito, M. Yamamoto, Macromolecules 24

(1991) 4061.[167] Y. Tsujii, A. Tsuchida, M. Yamamoto, T. Momose, T. Shida, J.

Phys. Chem. 95 (1991) 8635.[168] S. Hashimoto, unpublished result.[169] N. Gfeller, S.G. Calzaferri, J. Phys. Chem. B 101 (1997) 1396.[170] N. Gfeller, S. Megelski, G. Calzaferri, J. Phys. Chem. B 102 (1998)

2433.[171] N. Gfeller, S. Megelski, G. Calzaferri, J. Phys. Chem. B 103 (1999)

1250.[172] S. Megelski, A. Lieb, M. Pauchard, A. Drechsler, S. Claus, C.

Debus, A.J. Meixner, G. Calzaferri, J. Phys. Chem. B 105 (2001)25.

[173] G. Calzaferri, M. Pauchard, H. Maas, S. Huber, A. Khatyr, T.Schaafsma, J. Mater. Chem. 12 (2002) 1.

[174] N. Sabbatini, A. Guardigli, J.-M. Lehn, Cood. Chem. Rev. 123(1993) 201.

[175] M. Alvaro, V. Fornes, S. Garcıa, H. Garcıa, J.C. Scaiano, J. Phys.Chem. B 102 (1998) 8744.

[176] Y. Wada, T. Okubo, M. Ryo, T. Nakazawa, Y. Hasegawa, S.Yanagida, J. Am. Chem. Soc. 122 (2000) 8583.

[177] S. Hashimoto, S. Kirikae, S. Tobita, Phys. Chem. Chem. Phys. 4(2002) 5856.

[178] S.T. Selvan, T. Hayakawa, M. Nogami, J. Non-Cryst. Solids 291(2001) 137.

[179] K.L. Frindell, M.H. Bartl, A. Popitsch, G.D. Stucky, Angew. Chem.Int. Ed. 41 (2002) 959.

[180] J. Heikenfwld, M. Garter, D.S. Lee, R. Birkharn, A.J. Steckel, Appl.Phys. Lett. 75 (1999) 1189.

[181] H. Maas, A. Currao, G. Calzaferri, Angew. Chem. Int. Ed. 41(2002) 2495.

[182] S. Hashimoto, Tetrahedron 56 (2000) 6957.[183] H. Miyasaka, S. Ojima, N. Mataga, J. Phys. Chem. 93 (1989) 3380.[184] S. Ojima, H. Miyasaka, N. Mataga, J. Phys. Chem. 94 (1990) 4147.[185] S. Ojima, H. Miyasaka, N. Mataga, J. Phys. Chem. 94 (1990) 5834.[186] S. Ojima, H. Miyasaka, N. Mataga, J. Phys. Chem. 94 (1990) 7534.[187] S. Sankararaman, K.B. Yoon, T. Yabe, J.K. Kochi, J. Am. Chem.

Soc. 113 (1991) 1419.[188] K.B. Yoon, S.M. Hubig, J.K. Kochi, J. Phys. Chem. 98 (1994) 3865.[189] M. Alvaro, B. Ferer, V. Fornes, H. Garcıa, J.C. Scaiano, J. Phys.

Chem. B 106 (2002) 6815.[190] M.A. O’Neill, F.L. Cozens, N.P. Schepp, J. Phys. Chem. B 105

(2001) 12746.[191] T.E. Mallouk, Y.I. Kim, J. Phys. Chem. 96 (1992) 2879.[192] J.S. Kruger, J.E. Mayer, T.E. Mallouk, J. Am. Chem. Soc. 110

(1988) 8232.[193] P.K. Dutta, J.A. Incavo, J. Phys. Chem. 91 (1987) 4443.[194] M. Vital, N.B. Castagnola, N.J. Ortins, J.A. Brooke, A.

Vaidyalingam, P.K. Dutta, J. Phys. Chem. B 103 (1999) 2408.[195] S.H. Bossmann, C. Turro, C. Shunabel, M.R. Pokhrel, L.M.

Payawan, B. Baumeister, M. Wörner, J. Phys. Chem. B 105 (2001)5374.

[196] G. Cosa, M.N. Chretien, M.S. Galletero, V. Fornes, H. Garcıa, J.C.Scaiano, J. Phys. Chem. B 106 (2002) 2460.

[197] Y.S. Park, E.J. Lee, Y.S. Chun, Y.D. Yoon, K.B. Yoon, J. Am.Chem. Soc. 124 (2002) 7123.

[198] K. Tanaka, C-K. Choo, S. Sumi, Y. Kamitani, T. Fujii, K. Satoh,K. Fukuda, R. Nakata, J. Phys. Chem. B 106 (2002) 4155.

[199] J.C. Scaiano, N.C. de Lucas, J. Andraos, H. Garcıa, Chem. Phys.Lett. 233 (1995) 5.

[200] A.V. Barzykin, S. Hashimoto, J. Chem. Phys. 113 (2000) 2841.[201] D.L. Dexter, J. Chem. Phys. 21 (1953) 836.[202] H.G. Weiss, Aspects and Applications of the Random Walk,

North-Holland, Amsterdam, 1994.[203] G. Zumofen, A. Blumen, J. Klafter, J. Chem. Phys. 82 (1985) 3198.[204] A.V. Barzykin, M. Tachiya, Phys. Rev. Lett. 73 (1994) 3479.[205] L.J. Johnston, J.C. Scaiano, J.-L. Shi, W. Siebrand, F. Zerbetto, J.

Phys. Chem. 95 (1991) 10018.[206] J.L. Faria, R.A. McClelland, S. Steenken, Chem. Eur. J. 4 (1998)

1275.[207] F. Cozens, W. Ortiz, N.P. Schepp, J. Am. Chem. Soc. 120 (1998)

13543.

Shuichi Hashimoto has been teaching chemistryat the Gunma College of Technology. He ob-tained the BSc and MSc from Tohoku University,and the PhD in physical chemistry from TokyoMetropolitan University in 1982. From 1982 to1985, he conducted postdoctoral research at theUniversity of Notre Dame, working on photo-chemical studies in micelles with Professor J.K.Thomas. After 3 years of industrial experienceat Fujitsu Co. Ltd., he joined the Gunma College

of Technology in 1988, where he has been a professor of chemistry since1997. In 1993, he spent 10 months as a visiting scholar in Professor H.Masuhara’s laboratory at Osaka University, where he started an ultra-fastkinetic study in solid systems, including zeolites. His current researchinterest is the reaction dynamics of excited states in zeolites and othersolid materials.

review zeolite photochemistry: impact of zeolites on...

Documents