Photofunctions of Dye-Clay Hybrids: Recent Developments

Abstract

Precise design of hybrid nanostructures based on dyes in hybrid materials toward controlled photochemical reactions and novel photoinduced phenomena is overviewed with the emphasis on the recent developments. Various clays and clay minerals with different origins and characteristics have been used as hosts to control the location, orientation, and aggregation as well as the dynamic states (rotation and diffusion) of the dyes. The designed nanostructures affect photochemical properties such as efficiency, selectivity, and the rate of some photochemical reactions. Using the photochemical reactions in nanospaces, unique photoinduced phenomena such as nanostructural/morphological change and adsorption/desorption triggered by irradiation have been found.

1 Introduction

Photochemical reactions in heterogeneous systems may differ significantly from analogous reactions in homogeneous liquids/solutions or gas phases [1,2,3,4,5,6]. Important roles of the media/supports to control such parameters as the reaction rates/yields and product selectivity have been recognized so far, so that various photofunctional hybrids have been designed by organizing molecular species in/on solid surfaces. The location (proximity), orientation, association/aggregation, as well as the freedom (rotation and diffusion) of molecules on the surfaces or in the solids vary depending on the host-guest interactions at the interface to affect the characteristics and functions. The molecular and supramolecular designs (by the selection of host and guest and their composition and the additives) have been done using nanospace materials such as zeolites, mesoporous silicas, MOFs, COFs, and layered materials for organizing molecular and polymeric photofunctional species, and unique/useful photofunctions of the resulting hybrids have been reported [7,8,9,10,11]. Materials with defined (ordered) nanospaces have advantages as supports to accommodate guest species because their structure-property relationships will provide indispensable information on designing materials with controlled properties [12]. Spectroscopic properties, which are very sensitive to the environment, of the immobilized species, have given insights to the nanoscopic structures of the host-guest systems where conventional instrumental analysis does not have access [11,12,13,14,15,16]. By utilizing photoprocesses, one can obtain such information as distribution [17,18,19,20], orientation [21, 22], and mobility [23] of the guest species on/in nanospaces.

In this chapter, among possible host-guest systems, the studies on the organization of photofunctional species on/in clays (more accurately clay minerals and their synthetic analogs) will be summarized with the emphasis on the developments in the last two decades. The attention will mainly be focused on the role of the nanostructures, which directly and indirectly correlate the photofunctions of the guest species and the host-guest systems (Fig. 1).

Fig. 1

Possible effects of dye-smectite hybridization

2 Characteristics of Dye-Clay Hybrids for Photochemical Studies and Photofunctional Materials

Smectites have been used most extensively for a wide range of application including environmental and biomedical ones [24,25,26,27,28]. Smectites are a type of swellable 2:1 type layered clay minerals and consist of negatively charged silicate layer and charge compensating interlayer cations which are exchangeable [29,30,31]. The negative charge in the layers is generated by isomorphous substitution of framework metal cations with similar size and lower valency, and to compensate this negative charge, metal cations such as sodium and calcium occupy the interlayer space. The terms of bentonite and montmorillonite have often confused, bentonite is a term of a natural resource, and montmorillonite is the name of a clay mineral. The amount as well as the site of the isomorphous substitution influences the surface and colloidal properties of smectites. Impurities present both within the structure and on the particle surface, and elements and their amounts vary depending on the source of the clay minerals. Synthetic analogs of smectites, i.e., hectorite (Laponite, Rockwood Ind. Co. and Sumecton SWF, Kunimine Ind. Co.) [32], saponite (Sumecton SA, Kunimine Ind. Co.) [33], and swelling mica (sodium-fluor-tetrasilicic mica, TSM, Topy Ind. Co. and others) [33], do not contain colored impurities so that they are advantages for the photochemical studies. In addition to the commercially available ones, synthetic analogs of smectite have been prepared in the laboratory and used for the adsorption of dyes [25, 34]. The interlayer cation, which compensates the negative charge of the silicate layer, is exchangeable by the reactions in suspension and in solid state [35,36,37].

Cation exchange with interlayer exchangeable cations and the adsorption of polar molecules by ion-dipole interactions with interlayer cations and/or hydrogen bonding with the surface oxygen atom of the silicate sheets are known driving forces for the intercalation [38,39,40,41,42,43,44,45,46,47,48,49,50,51,52]. One of the characteristic features of smectites is the possible surface modification. Nanoporous pillared smectites have been obtained using inorganic particles and small organic cations as pillars [53,54,55]. Organophilic modification has been conducted by the cation exchange with cationic surfactants of various structures (Fig. 2) [26, 33, 56,57,58,59,60,61]. Due to the variation of the layer charge density and the molecular structures of the surfactants, host-guest systems with controlled microstructures and properties have been obtained.

Fig. 2

Examples of organic cationic dyes and schematic drawing for possible arrangements of alkylammonium cations [62, 63]

Layered alkali silicates, namely, magadiite (Na₂Si₁₄O₂₉∙nH₂O) and octosilicate (Na₂Si₈O₁₇∙nH₂O), which are characterized by the higher layer charge density than smectites, have also been used as hosts to construct photofunctional dye-silicate hybrids. The reactions of the layered alkali silicates involve covalent attachments through the reactions with the surface silanol groups [20, 64]. On the other hand, layered double hydroxides (LDHs) are composed of positively charged brucite-type layers of mixed-metal hydroxides and exchangeable anions located at the interlayer spaces, which compensate for the positive charge of the brucite-type layers [65]. Due to the structural characteristics and compositional variation, the application of LDHs in such areas as adsorption/separation of ions [65,66,67,68,69], catalysis [66,67,68], polymer additives [66,67,68], and medical and biochemical uses [66, 68] has been proposed so far. The chemical composition of the LDHs is expressed as [M(II)_1-XM(III) _X(OH)₂][Aⁿ⁻_X/n]^X− where M(II) = Mg, Co, Ni, etc.; M(III) = Al, Cr, Fe, etc.; and A is an interlayer anion such as CO₃²⁻ and Cl⁻.

In addition to the crystalline structures, the particle size and its distributions of layered solids are key issues in order to achieve optimum performance of layered solids and their intercalates; accordingly, attention has been paid for the powder morphology during the syntheses as well as classification [27, 70, 71]. Powders [36, 72], suspensions [73, 74], and thin films [55, 75,76,77] have been used for the evaluation of the photoprocesses, as well as for other application [78]. One of the unique and attractive properties of smectites is their spontaneous swelling in water. Platy particles pile up with their ab plane parallel to the substrate to form a film when the suspension is evaporated on a flat substrate [55, 75, 79]. The preparation of thin films by the Langmuir-Blodgett technique (LB technique) from exfoliated platelets of clays has also been reported [80, 81]. Inorganic-organic multilayered films have also been prepared via alternate adsorption of a cationic species and an anionic sheet of an exfoliated layered solid (layer-by-layer deposition technique, hereafter abbreviated as LbL technique) [82,83,84,85,86].

3 Surface Modification

In addition to the structural and compositional variation of smectites and other clay minerals (Table 1), the possible surface modification with organic/inorganic cations and polymers makes the variation of the material more versatile (Fig. 2) [87]. Long-chain alkylammonium ions have been studied most extensively in the chemistry of organophilic smectites, and the practical application of the organophilic smectites as adsorbents [25, 88] has been extensively reported. Phospholipids have been utilized for the construction of environmentally benign organoclay [89, 90]. Several nonionic surfactants have also been used for the surface modification of smectites [75, 91,92,93]. Intercalation of alkylammonium ions with more complex structures into layered silicates to precisely design hydrophobic nanospace, and, recently, flexibility of the interlayer surfactant aggregates has been discussed based on quasi-elastic neutron scattering data [57].

Table 1 Abbreviation of specific clays discussed in this chapter

Microporous and mesoporous solids have been obtained by crosslinking the nanosheets. The pioneering example is the pillaring with polyoxocations (e.g., [AlO₄Al₁₂(OH)₂₄(H₂O)₁₂]⁷⁺) [94, 95]. Nanoporous solids composed of silicate layers and metal/metal oxide finite particles have been prepared [96,97,98,99,100]. The microporous solids composed of silicate nanosheet and small organoammonium cations (e.g., tetramethylammonium ion, TMA) have been prepared and used for the separation/sensing and other functional materials [101,102,103,104,105,106,107]. The adsorptive properties of smectites modified with aliphatic and aromatic ammonium ions have also been investigated [101, 108, 109]. Pore size and porosity are controlled by selecting pillaring agents.

Synthetic clays by hydrothermal method have been developed as summarized in Table 1 and become necessary products for clay research thanks to the high purities and regular chemical equations. The careful syntheses of the clay minerals have been done to control the size of the silicate layers and layer charge densities. The research about the ion exchange led to the development of organically modified clays, which enabled to intercalate non-polar organic compounds into the interlayer spaces [110, 111]. The LB and LbL techniques have been used to fabricate thin films, while the effective swelling and subsequent evaporation of organically modified clays [76] led the thick films with improved quality.

4 Photophysics of Dye-Clay Hybrid Systems

4.1 Changes in the Absorption Properties, Color Change, and Stability

4.1.1 Effects of Host-Guest Interactions

The interactions between lone pair of the oxygen atom of silicate layer of clays and π-electron of dyes were proposed to affect for relatively planar structured dyes such as crystal violet [38, 39, 45,46,47], rhodamine B [48], pyronin Y [49], thiazines [50,51,52], and methylene blue [40,41,42,43,44, 112, 113]. The adsorption of pyronin Y (Scheme 1) has been examined in aqueous suspensions of smectites (Wyoming bentonite from Wards Natural Establishment Inc., a natural bentonite from Ünye, Turkey and LP-XLG) [49, 114]. Depending on the concentration and the dye-clay ratios, the absorption shifted, which has been explained as a result of the interactions between pyronin Y and silicate layer as well as the dye aggregation, although this latter effect will be exhaustively explained in Sect. 4.1.2. Thiazine dyes such as thionine [51], methylene blue [41, 42, 51], and tetraethyl thionine [51] (Scheme 1) showed hypsochromic shifts in clay suspensions [41, 51]. Thionine had absorption at 595 and 560 nm in an aqueous solution, which were attributed to a π-π^∗ transition of monomer and the absorption of H-dimer (see further details on aggregation types in Sect. 4.2.1), respectively [51]. By the addition of a Na-montmorillonite, a new absorption appeared at 528 nm, which was proposed to be caused by the interactions between the π-electron of thionine and the oxygen of the silicate layers, and another new absorption at 690 nm appeared, which was ascribed to J-aggregate (consult Sect. 4.1.2). Methylene blue [40, 115] and tetraethyl thionine [51] also showed hypsochromic shifts of monomer and a new absorption appeared by the addition of a montmorillonite. In LP-XLG, the absorption of thionine and tetraethyl thionine did not shift compared to the solution [51]. It was thought that the thionines had weaker π-interactions with the silicate layer than those with the bentonites. In the presence of a vermiculite (obtained from Zonolite), the absorption of the dimer increased and that of the monomer decreased, while the absorption shift was not observed [51]. The authors thought that the limited interlayer expansion for the vermiculite restricted the molecular conformation of thiazines to interact with the vermiculate. Thus, the hypsochromic shifts of the absorption of the dimers of pyronin and thiazines in the clay suspensions were observed by the interactions between lone pair of oxygen atom of silicate layer and π-electron of dyes. It suggested that the pyronin formed dimer in an aqueous solution, while it was de-aggregated and adsorbed in the interlayer space of the clays. The polarized IR spectra of the film of pyronin Y in LP-XLG and the Wyoming montmorillonite [49] suggested that the interactions between the oxygen of the layered silicates and the π-electron of the dyes made the molecular orientation of the dyes parallel to the silicate layer.

Scheme 1

Molecular structures of pyronin Y and thiazines

Stability of dyes is one of the prerequisites for the practical application, and the dye-clay interactions have been expected to play a role in it [116,117,118,119,120,121,122,123,124,125,126]. The improvements of the stability of the dyes upon irradiation and heating have been seen in (1) cationic dyes with smectites, (2) anionic dyes with layered double hydroxides (LDHs), and (3) nonionic dyes with organically modified smectites and LDHs. The improvements of the stabilities of the dyes by the hybridization with the layered materials were explained as results of reducing the intensity of the incident light by the absorption and scattering with the host [127,128,129], electronic stabilization of the dye [130,131,132], and suppressed gas diffusion mainly oxygen in the hydrophobic environment [111, 133,134,135,136,137,138,139].

An example of the improved chemical stability of the anthocyanin by the interactions with SA was shown by the color change upon exposure to acidic and basic atmospheres repeatedly [102]. Such natural dyes as β-carotene, anthocyanin, carmine, annatto, and carthamus yellow were adsorbed on a hydrotalcite [111, 139] and organically modified montmorillonite [110, 111]. Carminic acid, safflomin, and norbixin (Scheme 2) are the main components of carmine, carthamus yellow, and annatto, respectively [132]. When interacted with the hydrotalcite, the absorption spectra of carmine showed bathochromic shift as shown in Fig. 3a. The absorption shift was explained by the electrostatic interactions between carmine and the hydrotalcite and the planar molecular conformation induced by the adsorption on the hydrotalcite. As a result, the photostability was improved. Carthamus yellow showed the same trend as carmine. On the other hand, the stability of annatto was not affected in the presence of the hydrotalcite, suggesting the adsorption of annatto at the external surface of the hydrotalcite. The planar conformation was induced by the intercalation and was thought to contribute to the photostability of the dyes through the decrease of the lifetime of the excited state by the effective internal conversion. It is known that a rate constant of internal conversion is limited by Franck-Condon factor which is the overlap integral of vibrational parts of wave functions in the states before and after the transition [140]. Carmine and carthamus yellow were reported to have planar molecular conformations in hydrotalcite, expecting similar molecular conformation in excited state to that of the ground state. The conformation similarity increased the Franck-Condon factor and increased an effective internal conversion from the excited states. The opposite phenomenon (the dyes adsorbed on the external surface of smectites exhibited a decreased internal conversion rate) was also reported as discussed in Sect. 4.2.1. The change of the molecular conformation of the adsorbed dyes in the excited states was allowed on the external surface, while the molecular vibration was suppressed if compared with those in solutions.

Scheme 2

Molecular structures of dyes, which were reported to be stabilized by the host-guest interactions with clays

Fig. 3

(a) Diffuse reflectance UV-vis spectra of carmine intercalated in the hydrotalcite at the amount of 0.025 g/g (dotted line), 0.46 g/g (thin line), and 1.85 g/g (thick line). (b) Change of the absorbance during the irradiation for carmine intercalated in the hydrotalcite with the amount of (a) 1.85 g/g and (b) 0.025 g/g, (c) carmine mixed with SiO₂, and (d) carmine mixed with the hydrotalcite (Reproduced from the reference [132] with permission)

Photodecomposition of trifluralin (Scheme 2) was reported as a cyclization between an alkylamino group and one of two nitro groups to form an imidazole ring [141]. The adsorption of trifluralin onto SWy-1 suppressed the molecular motion to form the imidazole ring improving the photostability [142].

Rhodamine 6G (R6G in Scheme 2) was intercalated into smectites, KF, SA, and synthetic hectorites (HE and LP-RD) [143]. The stability of R6G against irradiation was remarkably improved on KF in both of a suspension and a film. As shown in Fig. 4, the emission intensity of R6G depended on the clays, and there is a linear correlation between the photoluminescence intensity and the dye stability, suggesting the quenching of the excited state of R6G was the key parameter to determine the stability. The stability of R6G was substantially modified upon the adsorption onto KF, where the energy transfer from R6G to KF led shorter lifetime of the excited state of R6G.

Fig. 4

(a) Photoluminescence spectra and (b) the relationship between the photodecolorization rate constant and the photoluminescence intensity (black squares) and the photoluminescence quantum efficiency (red squares) of R6G in HE, SA, LP-RD, and KF suspensions (Reproduced from the reference [143] with permission)

Improvement of the stability of R6G upon visible light irradiation was also reported for a polymer-smectite intercalation compound [143, 144]. R6G was intercalated in SA with poly(vinyl pyrrolidone) (PVP) to obtain a film, which the decoloration suppressed upon visible light irradiation over the R6G in SA or in PVP [144]. It was thought that the suppressed oxygen diffusion in the layered structure of SA-PVP contributed to the observed stability.

Another important process is charge transfer (CT) interactions from dyes to layered materials which induce bathochromic shift of absorption spectra [145, 146]. Hybridization of biphenyl with a synthetic hectorite (laponite)-induced absorption at 320 nm attributed to the CT transition, where the biphenyl acted as an electron donor and electron-deficient sites or Lewis acid sites of laponite were electron acceptors [147]. The CT interactions induced a triplet state of biphenyl by recombination from the CT state, and subsequent phosphorescence at 480 nm was observed at 130°C which was stronger than the fluorescence.

Anthraquinone-2-sulfonic acid was hybridized with MgAl-LDH (Mg_0.65Al_0.35(OH)₂(CO₃)_0.01), and the hybrid showed photoinduced reduction of the anthraquinone in formamide as shown by the color change from colorless to red [148]. The red color returned to the initial colorless in the dark. The mono-anionic and di-anionic anthraquinones with different lifetime were observed (13.9 and 16.9 min). The solvent was thought to act as the electron donor, and a surface of the MgAl-LDH provides a high pH environment to stabilize the anionic anthraquinone.

Another example is the diverse coloration of retinal Schiff base by the interactions with different smectites. Retinal in rhodopsin as a photoreceptive unit exists as a protonated Schiff base in a cis-isomeric state [149], and the rhodopsin provides three different environments for the retinal Schiff base to give blue (λ_max = 425 nm), green (λ_max = 530 nm), and red (λ_max = 560 nm) colors with broad absorption [150]. The similar absorption changes of the retinal Schiff base were observed by mixing the retinal Schiff base with three montmorillonites, Bengel Bright 11 obtained from Wyoming, USA (Hojun Ind. Co., Japan) (479 nm), a montmorillonite obtained from Mikawa, Japan (503 nm), and Bengel A obtained from China (Hojun Ind. Co., Japan) (532 nm) [151]. This is a rare example for the retinal Schiff base to show the color variation after the isolation from the protein.

4.1.2 Effect of the Dye Aggregation

Aggregation of such dyes [152,153,154,155,156] as acridine orange [157,158,159], methylene blue [23, 40, 160], azobenzenes [62, 161,162,163,164,165,166,167,168,169], merocyanines [170, 171], rhodamines [172,173,174,175,176,177,178], nile blue A [179], and porphyrins [180,181,182] has been reported to be induced by the interactions with layered materials [183, 184]. According to Kasha’s molecular exciton theory [185], J- and H-aggregates are distinguished by an angle between the line connecting centers of the dyes and the long axis of the dye molecule (α in Fig. 5). When the angle α is larger than 54.7° as shown on the left side in Fig. 5, the transition from S₀ to S₂ is allowed, it is called H-aggregate, and it’s characterized by a hypsochromic shift in the absorption band. When the angle α is smaller than 54.7° as shown on the right side in Fig. 5, the head-to-head aggregate is stabilized, and the transition from S₀ to S₁ is allowed. The aggregate shows a bathochromic shift (redshift) upon aggregation and is J-aggregate. Some dyes were intercalated into layered materials as monomolecular or bimolecular layers. The tilt angle between the silicate layer and dyes transition moment may cause shifts in the absorption spectra.

Fig. 5

Schematic representation of the relationship between the dye arrangement and energy level change by molecular aggregation

The effects of the length of the alkyl chain of the guest molecules were shown to affect the stability of aggregates in layered materials [159, 171]. Equilibrium constants K_b (M⁻¹) of the adsorption of N-alkylated acridine oranges [158], whose alkyl chains were methyl to tetradecyl (Scheme 3), onto KF were estimated [159]. The rate constant of disaggregation k_m (M⁻¹ s⁻¹) was estimated by the change in the absorption change of the monomer. As shown in Fig. 6, the second-order rate constants k_m increased by increasing the length of the alkyl group for short alkyl chains (number of C atoms up to 4), and the opposite behavior was observed for large alkyl chains (C atoms > 4). The variation of k_m was thought to be due to a compromise between the steric repulsion and the hydrophobic attraction between the alkyl chains, respectively.

Scheme 3

Molecular structures of dyes which formed aggregates in layered materials

Fig. 6

Dependence of the second-order rate constant, k_m, on the alkyl chain length of 1–14 (Reproduced from the reference [159] with permission)

By freeze-drying the suspensions, auramine O (Scheme 3) was hybridized with three montmorillonites SYn-1, SAz-1, and SWy-1 obtained from Source Clays Repository of the Clay Minerals Society [63]. The basal spacings of SYn-1 and SAz-1 did not change by the hybridization, while the absorption due to H-aggregate was observed. The authors proposed that auramine O was adsorbed on the external surface of SYn-1 and SAz-1 as H-aggregates. The absorption and the emission of J- and the H-aggregates of auramine O were observed for SWy-1 depending on the dye loading.

Structure of the intercalation compounds of an amphiphilic cationic azobenzene (AZC₂N⁺C₂OH, in Scheme 3) KF was proposed from the basal spacing (1.86 nm, corresponding to the gallery height of 0.9 nm) and the bathochromic shift of the visible absorption spectra, which was thought to be due to the head-to-tail orientation [62]. Considering the molecular size of AZC₂N⁺C₂OH and the observed bathochromic shifts of the visible absorption spectrum, it was proposed that AZC₂N⁺C₂OH formed J-aggregate as monolayer or bilayer as shown in Fig. 2. The absorption spectrum of AZC₂N⁺C₂OH intercalated in magadiite showed a hypsochromic shift, indicating the formation of H-aggregate, which has a gallery height of 1.57 nm [165, 186]. It was considered that the larger layer charge density of magadiite led the AZC₂N⁺C₂OH with the higher tilt angle to the silicate layer than that in KF to lead the H-aggregate (head-to-tail dimer).

The absorption spectra of methylene blue (Scheme 1) in the aqueous dispersions of smectites, whose layer charge densities were reduced by Hofmann-Klemen effect, was investigated [23, 115]. The phenomenon called Hofmann-Klemen effect is attributed to the migration of interlayer lithium ions to the vacancies in the octahedral sheet of smectites by heating. The absorption of J-aggregate decreased, and the dimer increased by reducing the layer charge density of the montmorillonite (obtained from Apache Country, Arizona) [115]. The same trend was seen in montmorillonites obtained from San Diego Country, California, and Horní Dunajovice, Czech Republic; a beidellite obtained from Stebno, Czech Republic; and a smectite obtained from Grand Country, Washington.

Dialkylated spiropyrans (Alkyl-SP-1 and Alkyl-SP-2 in Scheme 3) were intercalated into a film of a didodecyldimethylammonium exchanged montmorillonite, and the aggregation of photochemically formed merocyanines was investigated [171]. A photomerocyanine formed from 6-NO₂-SP has a characteristic absorption at 552 nm, while those of Alkyl-SP-1 and Alkyl-SP-2 absorbed at 493 and 617 nm, which were attributed to H- and J-aggregates, respectively.

Aggregation of dyes in the interlayer space has been reported. The emission of dyes is weaker for H-aggregate due to that the transition dipoles weaken each other [187, 188], while the emission of J-aggregate is intense, and the Stokes shift is smaller than those of the monomer [189,190,191]. Aggregates of pseudoisocyanine (PIC in Scheme 4) have been studied from 1935 [192] and those in layered materials have been reported from 1996 [193]. PIC has absorption at 490 and 520 nm in an aqueous solution, and the absorption of J-aggregate at 577 nm is seen only at high concentration [194]. An aqueous suspension of KF (10 mg/L) with PIC (5 × 10⁻⁶ M) showed absorption at 570 nm, which was attributed to the J-aggregate as shown in Fig. 7. The absorption of J-aggregates increased with increasing the concentration of KF. The aggregation of PIC was investigated using montmorillonites (SWy-1, SAz-1, and SYn-1 supplied from Source Clays Repository of the Clay Minerals Society and KF), SA, synthetic hectorites (SWN supplied from Coop Chemical and LP-RD), TSM [195,196,197,198,199,200], and magadiite [201] to find H-aggregate in SWy-1, SWN, LP-RD, and SA [197, 199, 202, 203]. For TSM, KF, SAz-1, and SYn-1, the absorption of J-aggregate was observed, and the absorption of the H-aggregate increased with the increase of the loading amount of PIC. There was no clear correlation between the cation exchange capacity (CEC) and the types of aggregates. It was proposed that the particle size of the clays can be a factor to determine the aggregation. The broadening of the absorption spectra of PIC due to the aggregation was observed for magadiite and dodecyltrimethylammonium (C₁₂TMA)-exchanged magadiite [201]. Because, in magadiite, the absorption shift was smaller and the Stokes shift was larger than those of J-aggregate observed in other clays, PIC formed aggregates in magadiite, while they were not the H- and the J-aggregates.

Scheme 4

Molecular structures of cyanines adsorbed on clays

Fig. 7

Absorption spectra of the (a) 5 × 10⁻⁶ M PIC aqueous solution and aqueous mixtures containing (b) 20, (c) 10, (d) 5, and (e) 1 mg of KF and 100 mL of 5 × 10⁻⁶ M PIC aqueous solution (Reproduced from the reference [193] with permission)

The H- and the J-aggregates of PIC were switched by swelling of host SA [204]. A SA film with 46.4 meq/100 g (70%CEC) of PIC had absorption of the monomer and a shoulder of the H-aggregate. By adding DMSO, the absorption of the J-aggregate increased, and the monomer and the H-aggregate decreased. The absorption spectrum returned to the initial shape by removing DMSO by washing with ethanol and subsequently drying.

Aggregation of rhodamine 6G (R6G, Scheme 2) in the films of Wyoming montmorillonites [176, 205] and laponites [172, 174,175,176] was investigated. R6G had absorption at 527 nm in an aqueous solution, [172] and, by adding the montmorillonite, absorption at 534 nm due to J-aggregate and two shoulders at around 500 and 470 nm appeared [205]. The absorption at 500 nm increased, and the shift of the monomer absorption from 527 to 538 nm was observed in a laponite film, which corresponded to the formation of the H- and the J-aggregates [172]. H- and J-aggregates were observed in the absorption and emission spectra of the film of a hexadecyltrimethylammonium (C₁₆TMA) exchanged KF [206]. Effective quenching of the emission from the H-aggregate [207] was observed for auramine O in SYn-1 powder [63], PIC in LP-RD powder [196], and merocyanine 540 in a bentonite (obtained from Ordu/Ünye in Turkey) and the bentonite exchanged with C₁₆TMA [208].

4.2 Changes in the Photoluminescence Properties

4.2.1 Effects of the Host-Guest Interactions

Adsorption on the external surface and in the interlayer space of layered materials suppresses molecular motion of dyes and induces a planar conformation [183, 209,210,211,212,213,214,215,216,217]. Three different effects of electrostatic interactions and the suppression of the molecular motion on the emission spectra have been proposed: (1) the bathochromic shift of emission by electrostatic interactions, (2) smaller Stokes shift than that in solutions owing to the similar molecular conformation of the excited and ground states inducing hypsochromic shift and increase intensity of the hypsochromic emission, or (3) decrease of the internal conversion rate due to the fixation of the molecular conformation in the excited state on the external surface. These effects were reported for porphyrins [17, 218,219,220,221,222], auramine O [223], and triphenylbenzene derivatives [224].

Porphyrins have two π-π^∗ transitions known as Soret band at around 400–500 nm and Q-band at around 500–700 nm. The Soret band and the Q-band of tetraphenylporphine (TPP, Scheme 5) and Fe(III)-TPP shifted by the adsorption on a purified Wyoming bentonite from 416 to 445 nm and 620 to 664 nm, respectively [218]. The shift of the absorption was explained by the coplanar structure of the porphyrin moieties and the phenyl substituents on the silicate layer. The relationship between the shift of the Soret band and the molecular structure of porphyrins has been studied to find that the matching of the intercharge distance of the porphyrins and the distance of the adjacent negative charge of the silicate layer was important [17, 219,220,221]. The shifts of the Soret bands of p-TMPyP, m-TMPyP, and o-TMPyP (Scheme 5) on SA were 30, 12, and 6 nm compared to their aqueous solutions [221]. The shifts were larger when the intercharge distance of the porphyrins (p-TMPyP, 1.05; m-TMPyP, 0.99; and o-TMPyP, 0.88 nm) was closer to the distance of the adjacent negative surface charge (1.19 nm). The planar structure of the three porphyrin derivatives on SA was proposed to be a reason for the shift of the fluorescence to longer wavelength region as shown in Fig. 8 [225, 226]. Fluorescence quantum yields and rate constants of the internal conversion of [Sb^V(TPP)(OH)₂]⁺ and [Sb^V(DMPyP)(OH)₂]³⁺ (Scheme 4) increased by the adsorption because the excited states of the two derivatives had more similar structures to the ground states than those in solutions. Among the three tested porphyrins, [Sb^V(TMPyP)(OH)₂]⁵⁺ adsorbed on SA had the fluorescence quantum yield same as that in an aqueous solution, while the internal conversion rate was lower than those of the others, suggesting the suppression of the molecular motion to keep the molecular structure in the excited state by the stronger electrostatic interactions with SA if compared with [Sb^V(TPP)(OH)₂]⁺ and [Sb^V(DMPyP)(OH)₂]³⁺.

Scheme 5

Molecular structures of porphyrin derivatives and Sb-porphyrin complexes adsorbed on clays (Reproduced from the reference [225] with permission)

Fig. 8

Fluorescence spectra of p-TMPyP (A), m-TMPyP (B), and o-TMPyP (C) with clay (solid line) and without clay (dashed line) (Reproduced from the reference [226] with permission)

The fluorescence of methyl viologen (MV²⁺) shown in Scheme 6 in hectorite and montmorillonite was observed as a result of the photoinduced charge transfer from the silicate layers to MV²⁺ [227]. Change of the emission spectra of MV²⁺ by the adsorption on layered materials has been studied [227]. MV²⁺ and viologen derivatives have been intercalated into montmorillonites [227,228,229,230,231], saponite [232, 233], hectorite [227, 231], nontronite [227, 231], vermiculite [229], titanates [234], niobates [235,236,237,238], and layered zirconium phosphate [239, 240]. The absorption spectra of MV²⁺ showed bathochromic a shift by the intercalation into smectites (KF, SA, and LP-XLG) [233], the layered zirconium phosphate/phosphonates [239, 240], and niobate [235,236,237,238]. The intercalated hybrid showed reversible color change to blue by the photoinduced charge transfer from the host. Non-emissive MV²⁺ showed fluorescence in an aqueous suspension of a hectorite (San Bernardino) and a montmorillonite (Clay Spur, Wyoming), and this phenomenon was explained by the planar structure of MV²⁺ on the silicate layer [231]. This phenomenon can be explained as “adsorption-induced emission.” The emission intensity of MV²⁺ increased by the decrease of the loading amount (by the increase of the concentration of the clays). The self-quenching of fluorescence of MV²⁺ on the clays was proposed as a reason of the weaker emission at the high loading amount.

Scheme 6

Molecular structures of dyes which form excimer in the interlayer space

Absorption at 410 and 450 nm appeared in powder of pyrene hybridized with LP-RD by UV irradiation. Electron spin resonance experiment with and without O₂ (triplet quencher) indicated that the absorption was from a triplet state of radical cation of pyrene [241]. The radical cation of pyrene was thought to be generated by the photoinduced charge transfer from pyrene to LP-RD. The color of the hybrid was different depending on the pre-activation temperature of LP-RD. The intensity of the ESR signal was higher when the pre-activation temperature was lower, suggesting the interactions of the surface water/OH groups with the adsorbed pyrene.

4.2.2 Excimer Formation

Pyrene has been used as a probe to investigate the surface chemistry [242,243,244,245,246,247] and refractive index [248] of clays as well as the surface modification with alkylammonium surfactants [249,250,251,252]. Excimer emission of pyrenemethylamine (Scheme 6) was enhanced by adding LP-XLS to an aqueous solution of pyrenemethylamine, suggesting change in the intermolecular distance between the adjacent pyrenemethylamine molecules [253]. As shown in Fig. 9, co-adsorption of N-isopropylacrylamide further enhanced the excimer emission, suggesting that N-isopropylacrylamide reduced the intermolecular distance of pyrenemethylamines to induce effective intermolecular interactions.

Fig. 9

Schematic illustration of the adsorption of pyrenemethylamine and N-isopropylacrylamide on LP-XLS and the corresponding normalized emission spectra. (a) Pyrenemethylamine in the aqueous solution; (b) pyrenemethylamine adsorption on LP-XLS; (c) co-adsorption of N-isopropylacrylamide on LP-XLS. The spectra a, b, and c correspond to the conditions A, B, and C (Reproduced from the reference [253] with permission)

Packing of DPDP [254] (Scheme 6) changed by the swelling of SA with DMSO [255]. DPDP was intercalated into SA film with various loading amounts from 0.033 to 49.7 meq/100 g (from 0.05 to 75%CEC). At the loading from 0.033 to 26.5 meq/100 g, the emission of the monomer at 482 nm decreased, while new emission at 588 nm of excimer appeared by the increase of the concentration. The gallery height of the hybrid increased from 0.48 to 0.93 nm by the swelling of SA with DMSO and the excimer emission enhanced, suggesting that the swollen SA provides the interlayer expansion large enough for the π-π stacking of DPDP.

4.2.3 Energy and Electron Transfer Between Molecules Adsorbed on Clays

The construction of photocatalyst by energy transfer in the interlayer space is a topic of interest [256,257,258,259,260,261,262]. Adsorption on layered materials is a way to concentrate dyes and to control the proximity, which resulted in efficient energy [19, 263,264,265,266,267,268,269] and electron transfer [18, 31, 270, 271]. Expected effects of hybridization of dyes and layered materials are (1) efficient quenching of the excited state of the dyes for photostabilization, (2) separation of the photosensitizer from photocatalyst to avoid the photocatalytic decomposition of sensitizer, and (3) directional energy and electron transfer by locational and orientational design.

A possible way to improve the photostability is quenching the photoexcited state by energy transfer as discussed in Sect. 4.1. The stability of herbicides [272,273,274], bioresmethrin [275], and norflurazon [276] was improved by energy or electron transfers to methyl green co-adsorbed on montmorillonite and to thioflavin T on SWy-1. The separation of the photosensitizer from the photocatalyst (anatase particle) by smectite nanosheet induced durability of the photosensitizer. The photocatalytic oxidation of benzene to phenol was done in an aqueous suspension of anatase with the tris(2,2-bipyridine)ruthenium(II) (designated as [Ru(bpy)₃]²⁺)-synthetic saponite, resulting in a high yield of benzene decomposition and selectivity of phenol. The hybrid was processed as a film to be used as the photocatalyst layer to obtain a photocatalytic flow reactor [277, 278].

The fluorescence of [Ru(bpy)₃]²⁺ was quenched by SO₂ gas [107] suggesting a possible gas sensor application. The fluorescence of [Ru(bpy)₃]²⁺ was quenched by the MV²⁺ co-adsorbed on smectites, and the quenching efficiency was higher when smectites with larger particle sizes (e.g., TSM) was used if compared with those on clays with smaller particle sizes (e.g., SA and LP-XLG) [279]. Such quenchers as cyanine, MV²⁺, and anthraquinone derivatives were co-adsorbed with cyanine dyes on laponite RDS (Southern Clay Products, Inc.) to quench the fluorescence of the J-aggregate of the cyanine [280].

MV²⁺ (Scheme 7) in the interlayer space of smectites acted as an electron acceptor [232, 233, 281,282,283,284]. It was reported the photoinduced electron transfer from TPP (Scheme 5) to MV²⁺ under visible light irradiation to form the TPP radical cation and MV⁺ radical cation [281]. The color of MV²⁺ intercalated in a hectorite-like layered silicate changed by the adsorption of N,N-dimethylaniline and 2,4-dichlorophenol [285], and the color depended on the layer charge density and the concentration of the adsorbents. The hectorite-like layered silicate was synthesized from LiF, Mg(OH), and SiO₂ sol with various ratios to vary the layer charge density [286]. The color of the hybrids of the hectorite and MV²⁺ changed from yellow to purple green and purple by the adsorption of N,N-dimethylaniline depending on the layer charge density and the increase of the concentration of N,N-dimethylaniline. Color changed from yellow to orange was seen by the adsorption of 2,4-dichlorophenol. The color of the hybrids with N,N-dimethylaniline was explained to the formation of methyl violet, and that with 2,4-dichlorophenol was attributed to the charge-transfer complexes.

Scheme 7

Molecular structures of dyes exhibited energy and electron transfer on layered materials

Effects of the molecular location and orientation on the energy transfer between adjacent molecules have been investigated in porphyrin-smectites systems [19, 211, 260, 264, 266, 284]. Energy transfer from Zn complexed aniline-substituted porphyrin (ZnTMAP, Scheme 7) to pyridine-substituted porphyrin (p-TMPyP, Scheme 5) was investigated in SA suspension by monitoring the contribution of the fluorescence from ZnTMAP excited at 428 nm, which matched the absorption of Soret band of ZnTMAP [287]. When the two porphyrins adsorbed on SA individually and subsequently mixed, the contribution of the fluorescence intensity of ZnTMAP was the same in the concentration region of the porphyrins to 6.6 meq/100 g (10%CEC, for each), while the contribution decreased in the higher concentration region (Fig. 10). When the two porphyrins were co-adsorbed on SA, the contribution of the fluorescence intensity of ZnTMAP was decreased from 0.6 to 0.1 with increase of the concentration of the two porphyrins to 0.066 meq/100 g (0.1%/CEC for each), and then, the contribution of ZnTMAP gradually decreased and become negligibly small at 29.8 meq/100 g (45%CEC for each) as shown in Fig. 10. These results suggested that the energy transfer between SA layers was major in the suspension prepared with SAs that adsorbed porphyrins individually, while the intralayer energy transfer was major in the suspension prepared with SA with porphyrins by the co-adsorption.

Fig. 10

Contribution of ZnTMAP for fluorescence at various porphyrin loading levels for independent adsorption (squares) and co-adsorption (circles) (Reproduced from the reference [287] with permission)

The sequential energy and electron transfers from 2-acetylanthracene to Zn-p-TMPyP to DNPV²⁺ (Scheme 6) on SA were reported [288]. 2-Acetylanthracene was encapsulated in octaamine to be adsorbed on SA with 1:1 ratio of Zn-p-TMPyP. In a suspension of SA with 2-acetylanthracene in octaamine and Zn-p-TMPyP, the decrease of fluorescence of 2-acetylanthracene and the increase of that of Zn-p-TMPyP compared to suspensions were seen, suggesting the energy transfer from 2-acetylanthracene in octaamine to Zn-p-TMPyP. The shorter lifetime of Zn-p-TMPyP in a suspension of SA with DNPV²⁺ than that in a suspension without DNPV²⁺ suggested the electron transfer from Zn-p-TMPyP to DNPV²⁺.

4.3 Alignment of Dyes by Host-Guest Interactions: Study by Linear Polarized Light

Film of propylammonium-exchanged titanate (Na₂Ti₃O₇) was prepared by casting an aqueous suspension and used as a host of cationic dyes. Pseudoisocyanine (PIC in Scheme 7) intercalated in the titanate film showed absorption anisotropy [289] as shown in Fig. 11. The absorbance in the visible region with 5° of the angle between the film plane and the incident light was decreased by the rotation of the incident polarized light from 0° to 90°, while the small absorption change was observed with 90° of the angle between the substrate plane and the incident light. It suggests that a transition moment of PIC, which corresponds to the long molecular axis, was parallel to the substrate.

Fig. 11

Polarized visible spectra of a film of PIC-Ti₃O₇. The angle between the substrate plane and the incident light was set to (a) 5° and (b) 90° by rotating the film; a represents the angle between the polarization direction of the incident light and the rotation axis of the film (Reproduced from the reference [289] with permission)

The angles between molecular axes of the following porphyrin derivatives, p-TMPyP, trans-DMPyP, and cis-DMPyP (Scheme 4), and the silicate layer of SA were estimated by polarized visible light attenuated total reflection (polarized vis-ATR) spectra [290]. Figure 12 shows that the absorbance of p-TMPyP of s-polarized light was larger than that of the p-polarized light. A similar difference in the polarized absorption spectra was observed in trans-DMPyP and cis-DMPyP. The angles between the molecular axes of the three porphyrins and the surface of the silicate layer were estimated to be less than 5°.

Fig. 12

Polarized vis-ATR spectra of p-TMPyP with s- and p-polarized light (Reproduced from the reference [290] with permission)

Fluorescent dyes adsorbed in layered materials showed linearly polarized emission, suggesting the alignment of the dye dipole in the interlayer space [291,292,293]. A laponite film was prepared by spin-coating an aqueous suspension and was used to accommodate rhodamine 6G (R6G, in Scheme 2) by cation exchange by immersing the film in an R6G solution [206, 294, 295]. Fluorescence intensity excited by horizontally polarized light depended on the direction of the polarizer. The angle between the long axis of R6G and the silicate layer was estimated to be 28° from the difference of the fluorescence intensities observed with horizontal or vertical axes of the film.

Polarized fluorescence of rhodamine B [296], fluorescein [297], pyrene [298], cyanine [299], niflumic acid [300], bis(N-methylacridinium) [301], tris(8-hydroxyquinolate-5-sulfonate)aluminum(III) [302], polyphenylene [303], and polythiophene [304] in LDH films prepared by LbL technique was reported. The anisotropic value r as defined by Eq. (1) was derived to discuss the dye orientation from emission anisotropy.

$$ r=\frac{I_{VV}-{I}_{VH}}{I_{VV}+2{I}_{VH}} $$

(1)

In (1), I_VV and I_VH are emission intensities with vertical and horizontal directions excited by vertical light [305]. The fluorescence spectra of a carbocyanine (Scheme 4) intercalated in a MgAl-LDH [303] showed anisotropy (Fig. 13) with the r value as high as 0.8, where the value was higher than that in a solution (0.2) and nearly twice of the theoretical highest value in the case without macroscopic alignment (0.4) [299]. The r value of the carbocyanine was larger than pyrene (0.26) [298], fluorescein (0.3) [297], and fluorenone derivatives (0.25) [306], which have planar molecular structures. The orientation of the intercalated dye in MgAl-LDH was thought to induce the anisotropy.

Fig. 13

Polarized fluorescence with the intensities of I_VV and I_VH and the r value for the hybrid of the carbocyanine in MgAl-LDH (Reproduced from the reference [303] with permission)

5 Photochemical Reactions

5.1 Intramolecular Reactions Affected by Host-Guest Interactions

Molecular reactions of organic compounds interacted in clays are different from those in solutions. The molecular reactions of the organic molecules were affected by the polarity of interlayer space and/or by the suppression of molecular vibration in the interlayer space [8, 10, 307].

5.1.1 Azobenzene

Azobenzene (AZ) shows photoisomerization from trans- to cis-isomer by UV irradiation, and the cis-isomer returns to the initial trans-isomer by heat or visible light irradiation as shown in Scheme 8. The half-lives of the thermal isomerization of the cis-isomer to the trans-isomer of AZ in a benzene solution at room temperature and 60°C were 119 [308] and 1.5 h [309], respectively, in the dark following first-order kinetics. The polarity and the viscosity of the solvent affect the photoisomerization of AZ [310, 311], and the photoisomerization was suppressed in packed states [312]. Accordingly, the host-guest interactions are expected to affect the photoisomerization.

Scheme 8

Molecular structures of azobenzenes introduced in this chapter

Here, a fraction of the cis-isomer of AZ is used as a measure of photoisomerization. The fractions of the cis-isomer of AZ at the photostationary state in cyclohexane [313] and in toluene [314] solutions were 80 and 90%, respectively. Although the fraction of the cis-isomer is different depending on the substituents of the azobenzene and the wavelength of the incident light [315], the fraction of the cis-isomer in polymethyl methacrylate (PMMA) was as high as that in the toluene solution [316]. The fraction of the cis-isomer became smaller in poly(styrenesulfonate) (PSS) [317], azobenzene polymer [318], SiO₂ [316], and zeolite [313] as summarized in Table 2. The fraction (83%) of the cis-isomer of trimethylAZ was larger than that (60%) of hexamethylAZ in the methylcyclohexane-isohexane 2:1 solution [319] and that (64%) of C₁AZC₂OH (in Table 2) in a SiO₂ film prepared by a sol-gel reaction with tetraethoxysilane (TEOS), but it was smaller than that (93%) in the ethanol solution [316] (Table 2). It was explained that the steric repulsion between the azobenzene derivatives and the media is an important factor to change the molecular fraction of the cis-isomer. The azobenzene derivatives, which were introduced as building blocks of co-polymers (PEAZ and COClAZ in Scheme 8), showed a smaller fraction (65–70% for PEAZ and 50% in a swelled film for COClAZ) of the cis-isomer than that in the AZ solution of cyclohexane. The steric repulsion between the azobenzene part and the other part of the polymer suppressed the isomerization. LB films with amphiphilic azobenzene derivatives (as summarized in Table 2) have been studied [320,321,322,323,324,325,326,327,328,329,330,331,332]. The fraction of the cis-isomer was as high as 90% in the LB films of C₈AZC₃SO₃ (Scheme 8) with poly(diallyldimethylammonium chloride) [332].

Table 2 Fraction of cis-isomer of azobenzene derivatives

AZ was adsorbed on a sodium montmorillonite by solid-solid reactions [334] and from vapor [335]. Adsorption of AZ on a kaolinite from aqueous solution was also reported [217]. Photochemical studies of AZ in layered materials were initiated by Ogawa et al., and the first one was that on the AZ intercalated in an organically modified montmorillonite [334]. Then, amphiphilic cationic azobenzenes were intercalated into a montmorillonite [161], followed by the intercalation of various ionic azobenzenes (summarized in Table 3) into layered clay minerals [62, 163,164,165, 167, 186, 336,337,338,339,340], layered double hydroxide (LDH) [341], potassium hexaniobate (K₄Nb₆O₁₇) [342], and a titanoniobate [343] by ion exchange reactions.

Table 3 Molecular structures of cationic azobenzenes intercalated in smectites

The photochromism of AZ was observed in organically modified KF [334] and TSM [162, 349]. The fraction of the cis-isomer of AZ in didodecyldimethylammonium (2C₁₈2C₁N⁺)-TSM was about 80% at the photostationary state under a 500 W super high-pressure Hg lamp [162], while that in octadecyltrimethylammonium cation (C₁₈3C₁N⁺)-TSM was about 35% under a 100 W high-pressure Hg lamp [349]. The difference of the fraction of the cis-isomer is thought to be due to the molecular packing of the surfactants in the interlayer space. The thermal cis- to trans-isomerization of AZ in C₁₈3C₁N⁺-TSM took 2 days [349] which was faster than the half-life of cis-AZ in a benzene solution (5 days) [308]. The thermal isomerization of cis-AZ in the solution followed a first-order kinetics while in such polymers as poly(methyl methacrylate) (PMMA) [350] and poly(ethyl methacrylate) (PEMA) [351] and a silica gel synthesized by sol-gel method did not follow the first-order kinetics [350], indicating that the azobenzene molecules were in several environments in these solid-state materials.

The photochromism of cationic azobenzenes in montmorillonites [62, 161, 344], saponites [163, 337, 352], a fluorohectrite [337], taeniolite [163], and magadiite [164, 165, 186, 346] was reported [167]. Organically modified fluoro-tetrasilicic mica [345], montmorillonite [353], and beidellite [353] were also used. As summarized in Table 2, the fraction of the cis-isomer of AZC₂N⁺C₂OH in mesoporous silica (MCM-41) with the pore size of 3.2 nm was 70% [333], and those of AZ in a zeolite NaY (pore size: 0.74 nm) and a sodium mordenite (pore size: 0.7 × 0.65 nm) were 80 and 50%, respectively [313]. The fraction of cis-isomer of the cationic azobenzene (AZC₂N⁺C₂OH, Table 3) in magadiite at room temperature was 80%, similar to AZ in a cyclohexane solution [165]. It was thought that the structural change of azobenzenes was accommodated by the change of the basal spacing to achieve the relatively high yield of cis-isomer. The thermal isomerization of cis-isomer of AZC₂N⁺C₂OH in magadiite followed the first-order kinetics, indicating that AZC₂N⁺C₂OH was homogeneously distributed in magadiite. The fraction of cis-isomer at the photostationary state decreased in KF [161] and magadiite [165] at low temperatures. The fraction of cis-isomer of the cationic azobenzenes (C₈AZC₁₀N⁺ and C₁₂AZC₅N⁺, in Table 3) in KF at room temperature was about 50%, while cis-isomer was practically not detected at the temperature lower than 200 K, suggesting that the molecular motion was suppressed in the interlayer space of KF.

There are several examples of the suppression of the trans- to cis-isomerization. In a fluorohectorite (obtained from Corning Inc.) in both suspensions and films, photoisomerization of N⁺C₁AZC₁N⁺ (Table 3) was suppressed, while AZC₁N⁺ isomerized as shown by the change in the absorption spectrum (Fig. 14) [337]. It was explained that attractive electrostatic forces between the silicate layers and the dicationic N⁺C₁AZC₁N⁺ hindered the isomerization. Photoisomerization of N⁺AZN⁺ (Table 3) was suppressed by the adsorption on SA [348]. Both trans- and cis-isomers of N⁺AZN⁺ were exchanged on SA. The cis-isomer showed photoisomerization to the trans-isomer on SA by 420 nm light irradiation with a higher quantum yield than that in an aqueous solution (without clay), while trans-isomer on SA did not show photoisomerization. The suppression of the trans- to cis- isomerization was thought to be due to the matching of the intercharge distance of the trans-isomer and that of the adjacent negative surface charge of the silicate layer. The trans-isomer interacted with the silicate layer with both of two cationic moieties and the molecular motion was suppressed. In the case of the cis-isomer, one of the two cationic moieties interacted with the surface charge, leading that the cis-form isomerized to the trans-isomer.

Fig. 14

UV-vis absorption spectra of (a) AZC₁N⁺ and (b) N⁺C₁AZC₁N⁺ in a fluorohectorite dispersion. The spectra were recorded before (solid line) and after the UV irradiation (dashed line). Difference spectra were derived by subtracting the spectrum before the irradiation from that of the irradiated solution (dotted line) (Reproduced from the reference [337] with permission)

Nonionic azobenzenes intercalated in organically modified clays and cationic azobenzenes exchanged on clays showed photochromism with the relatively high fraction of the cis-isomer at the photostationary state. When a dicationic azobenzene was adsorbed on SA, the isomerization of the trans-isomer was suppressed. By the design of the interlayer space of clays, the control of the photochromism of azobenzene such as the fraction of the cis-isomer, thermal fading speed to achieve quick response, and bistability is expected.

5.1.2 Diarylethene

Photochromism of diarylethenes is attributed to reversible photocyclization between two aryl rings [354]. Open-ring isomer of diarylethenes shows photocyclization under UV irradiation and closed-ring isomer with a planar π-system forms. Both of the open-ring and the closed-ring isomers do not show the thermal isomerization to give P-type (thermally irreversible, but photochemically reversible) photochromism. The reversibility and the durability of the diarylethenes have been investigated by Irie through molecular design [354]. The points are (1) substitution with a ring structure to suppress the cis-trans isomerization of stilbene and (2) substitution with thiophene rings to stabilize the closed-ring isomer as shown in Scheme 9. The open-ring isomer of the diarylethenes has two molecular conformations, photoinactive parallel and photoactive antiparallel conformations, and the ratio of these conformations are almost equal in solution [354]. In addition to the photoinduced oxidation of thiophen rings [355], a photoinactive isomer, which formed by the condensation of two thiophen rings, was reported as a deactivation process of diarylethenes [356,357,358,359]. In order to improve the reversibility and the yield of photocyclization, control of the molecular conformation of the diarylethene has been examined by the molecular and supramolecular designs using such intramolecular interactions as hydrogen bonds [360, 361], intramolecular steric repulsion [362, 363], as well as the host-guest interactions in nanospaces [364, 365]. Almost unity quantum yield (98%) of a ring-closing reaction of a diarylethene derivative was achieved in a hexane solution by restricting the molecular motion with intramolecular hydrogen bonds, being smaller in polar solvents (in methanol solution: 54%) [360, 361].

Scheme 9

Molecular design of diarylethenes

Restricted molecular motion by host-guest interactions is expected to affect the ring-closing reaction. A pyridine- and aniline-substituted diarylethenes (Py-DAE and An-DAE, Scheme 10) were intercalated into KF and magadiite by cation exchange [366, 367] and covalent functionalization [368, 369], respectively. An intercharge distance in the parallel conformation of Py-DAE (0.9 nm), which was shorter than that of the antiparallel conformation (1.3 nm), matched with the distance between adjacent negative charge of KF (0.9 nm) so that the parallel conformation was preferred on KF. As a result, a change of absorption owing to the photocyclization of Py-DAE in KF suspension was smaller than that in a solution (without clay) [367]. The reversibility of Py-DAE photochromism was improved by co-intercalation of dodecylpyridinium ion [366]. The formation of the parallel conformation was suppressed by the co-intercalation of dodecylpyridinium. An-DAE covalently bound on magadiite gave improved reversibility of the photocyclization compared with that in an ethanol solution [368, 369]. The molecular rotation from the antiparallel to the parallel conformation of An-DAE was suppressed by the covalent attachment to the silicate layer. Reversibility of photochromisms was also improved by restricting the molecular motion by the confinements. It was also claimed that the suppression of the parallel conformation improved the reversibility. The suppression of the generation of the photoinactive isomer was also concerned for the reversibility.

Scheme 10

Molecular structures of diarylethenes hybridized with clays

5.1.3 Spiropyran

Photochromism of spiropyran is dependent on the polarity of the molecular environment. In a non-polar environment, spiropyran (SP) is more stable than photomerocyanine (photoMC) (Scheme 11) as zwitter ionic structure [370]. 6-NO₂-MC is photochemically formed from 6-nitrospiropyran (6-NO₂-SP, in Scheme 12) in toluene showing blue color with a half-life of 5.6 s at room temperature, while 6-NO₂-MC in ethanol was red and the half-life is 17 min [371]. The phenoxyl moiety of merocyanine is protonated in acidic condition, and the protonated merocyanine shows yellow; thus, photochromism of spiropyran in the acidic condition is related to three isomers of spiropyran, merocyanine, and protonated merocyanine forms [370, 372, 373].

Scheme 11

The photochromism of spiropyran in (a) non-polar and (b) polar environment

Scheme 12

Chemical structures of spiropyrans hybridized in clays

SP, the protonated MC, and 6-NO₂-SP were intercalated into KF by mixing in a mixed solvent of methanol and water and subsequent filtration [170]. The hybrids of KF were yellow, suggesting that SP and 6-NO₂-SP converted to the protonated MC, which was stabilized by the interactions with hydroxyl groups at the edge of the silicate layer. 6-NO₂-MC generated from 6-NO₂-SP is red in a toluene suspension of HE by the UV irradiation [374], indicating that 6-NO₂-MC adsorbed on HE by the host-guest interactions.

Dihydropyrenes [375], Stenhouse salts [376, 377], and binaphthyl-bridged imidazole dimers [378, 379] showed negative photochromism (opposite behavior to normal photochromism in non-polar environment; photodecoloration by visible light and thermal coloration) (Scheme 11b). High conversion of the photoisomerization of the negative photochromic compounds with respect to normal photochromism is possible because of the absence of the visible light absorption by the photochemically formed colorless isomer. In polar environments such as in mesoporous silicas [380,381,382,383], zeolites [383,384,385], and LDHs [386, 387], MC form is thermally more stable than SP form. These characteristics were utilized to control negative photochromism [380, 381, 388]. Both of Py-SP and 6-NO₂-Py-SP showed negative photochromism on KF [389, 390]. It was thought that the cationic parts of pyridine and aniline moieties electrostatically interacted with negative charges on the silicate layers and the merocyanine forms were stabilized.

Nonionic spiropyrans, SP and 6-NO₂-SP, and a cationic spiropyran, 6-NO₂-Py-SP, showed normal photochromism in a cetyltrimethylammonium (CTA⁺) exchanged KF [170, 391, 392]. The CTAB provided the hydrophobic environment for the spiropyran derivatives to show normal photochromism [171]. The hydrophobicity of the clays is controlled by intercalating a wide variety of surfactants.

5.2 Intermolecular Reactions

The intermolecular reactions in/on nanospaces have been used for (1) stereoselective reaction and (2) trigger of photoinduced phenomena [393]. Intermolecular distance and orientation of guests in layered materials have been utilized to control the stereoselective reactions [394, 395].

5.2.1 Diels-Alder Reaction in Clay Interlayer

Diels-Alder reaction is known as a [4 + 2] cycloaddition between a π-conjugated diene and an alkene to form a six-membered ring. A major reaction product is predicted by Woodward-Hoffman rules, while some by-products are included in the products [394, 395]. Because the kinetically stable product is the major product of Diels-Alder reaction (endo rules), high temperatures are not recommended to accelerate Diels-Alder reaction with keeping the reaction selectivity. The addition of the catalyst, which does not affect the reaction selectivity, is proposed to increase the rate constant for Diels-Alder reaction. Acceleration of the dimerization of 1,3-cyclohexadiene [396] and the cycloaddition between 2,3-dimethyl-1,3-butadiene and acrolein [397] in the presence of a montmorillonite (K10) with Fe(III) was reported. A reaction yield of the dimerization of 1,3-cyclohexadiene increased to 49% even at 0°C for 10 h by adding K10 with Fe(III) with keeping the product selectivity the same compared to a reaction yield (30%) at 200°C for 20 h without K10 [398]. The yield (80% at 20°C for 3 h in water) of the cycloaddition increased to 95% at 20°C for 0.3 h by adding the K10 with Fe(III). The reaction condition was optimized to −24°C for 4 h in dichloromethane to achieve the yield of 96% with keeping the product isomer ratio. Though the role of the clay for the improved selectivity of Diels-Alder reaction was not explained clearly [396, 397], molecular packing of guests in clays are thought to contribute [176, 193].

5.2.2 [2 + 2] Photocycloaddition

According to Woodward-Hoffman rules, a [2 + 2] cycloaddition does not progress by heat and is photochemically allowed. Some of the stereoisomers are obtained by the [2 + 2] cycloaddition due to a biradical process [399]. Packing (orientation) at the initial state and the molecular conformation of the intermediates affect the stereoselective [2 + 2] cycloaddition. Three products are possible by the cycloaddition of 2-cyclohexene-1-one in a solution as shown in Scheme 13. The reaction yield of photocycloaddition of 2-cyclohexene-1-one in the benzene solution under UV irradiation for 10 h (without clay) was 25%, and the ratio of the products of anti-HH, syn-HT, and anti-HT was 8:0.1:29 (Scheme 13) [400, 401]. By adding SA, the reaction yield remarkably increased to 72%, and the ratio of the products of anti-HH, syn-HT, and anti-HT was 82:0.1:5 [400, 401]. It was proposed that 2-cyclohexene-1-one aggregated in a parallel fashion, which led the effective photocycloaddition to anti-HT. The photoluminescence of the excimer of 2-cyclohexene-1-one was observed for the clay suspension.

Scheme 13

Photocycloaddition of 2-cyclohexene-1-one

Stilbene shows cis-trans isomerization and dimerization. Four possible photodimers of the stilbene are obtained as shown in Scheme 14. The photochemistry of stilbazolium ion (STN⁺, Scheme 14) has been investigated to find stereoselective photodimerization [214, 402,403,404,405,406,407,408,409,410]. A higher reaction yield of photodimerization of STN⁺ (the total yield, 98%, and the cis-isomer, 14%; syn-HT, 70%, and syn-HH, 5%, respectively) under UV irradiation for 30 min in an aqueous suspension of a saponite than that in a solution (the total yield, 69%, and the cis-isomer, 67%, syn-HT, 2%, and anti-HH, 2%, respectively) was reported [402]. Cyano- and methyl-substituted STN⁺ were used to show that the hetero dimer was a major product thanks to the formation of the exciplex between the electron-donating methyl-substituted and the electron-accepting cyano-substituted stilbenes in the interlayer space of SA [404]. Recently, three stilbazolium derivatives, STN, MeSTN⁺, and C₁₆STN⁺ (in Scheme 14), were intercalated into a N-(2-(2,2,3,3,4,4,4-heptafluorobutanamido)ethyl)-N,N-dimethylhexadecan-1-ammonium bromide exchanged SA by mixing in HCl aqueous solution to examine the selectivity of the photochemical reactions [409]. Three factors, (1) electronic interactions between the negatively charges SA surface and the stilbazolium derivatives, (2) hydrophobic interactions among the long alkyl chains of the surfactants, and (3) both hydrophobic and lipophobic interactions of the perfluoropropyl moieties of the surfactant, affected the selectivity.

Scheme 14

Molecular structures of stilbene derivatives and photoreaction of trans-stilbene

A [2 + 2] cycloaddition in the interlayer space of a hydrotalcite [411] led oligomer of phenylenediacrylate with a head-to-head structure [412]. p-Phenylenediacrylate was exchanged on a hydrotalcite (Alcamac Cl supplied by Kyowa Chemicals Ltd.) and was irradiated a 300 W medium pressure Hg lamp for 6 h in the aqueous suspension under stirring. The ratio of monomer, dimer, trimer, and oligomer was 0:22:31:47 after the irradiation for 6 h. In contrast, the ratio of the monomer and the dimer in an aqueous solution of p-phenylenediacrylate (without the hydrotalcite) was 74:26, and the trimer and the oligomer were not obtained. The polymerization degree was up to 10. The oligomer was assigned to the syn-head-to-head structure as shown in Scheme 15. The oligomer of p-phenylenediacrylate was not obtained by the irradiation to the powder.

Scheme 15

Oligomerization of p-phenylenediacrylate in the interlayer space of the hydrotalcite (Reproduced from the reference [412] with permission)

The stereoselective cycloaddition was achieved by the molecular packing in the interlayer space of clays. Aspect ratio [413] and layer charge density [414,415,416,417] affected the yield and the selectivity. The selective cycloadditions were also reported in such nanospace materials [418] as zeolites [419, 420], mesoporous silicas [421], and a surfactant intercalated graphite oxide [422].

5.3 Uses of Photochemical Reactions as Trigger for Photoinduced Phenomena

5.3.1 Photoinduced Change in the Basal Spacing

The change in the basal spacing by photoirradiation is the first example of photoinduced structural change in intercalation compounds [334]. Azobenzene (AZ) shows reversible photoisomerization between a rod-like-shaped trans-isomer and a bending structured cis-isomer (Scheme 8), and a molecular size of a long axis was switched between 1.21 nm for the trans-isomer and 0.74 nm for the cis-isomer. AZ was intercalated into n-dodecylammonium exchanged KF by mechanical mixing without a solvent [334]. The basal spacing of the organically modified KF increased from 1.8 to 3.0 nm by the intercalation of AZ. The basal spacing of the hybrid increased from 3.0 to 3.1 nm by UV irradiation, which is the first example of photoinduced change in the basal spacing.

It was reported that the basal spacing of magadiite intercalated with AZC₂N⁺C₂OH (Table 3) changed from 2.69 to 2.75 nm under UV irradiation and retuned to 2.69 nm by visible irradiation [165, 186], while those of KF intercalated C₈AZC₁₀N⁺ and C₁₂AZC₅N⁺ (Table 3) did not change by the isomerization of the cationic azobenzenes in the interlayer space [161]. Basal spacings of the montmorillonite with a cation exchange capacity of 143 meq/100 g intercalated aminoAZ and 4,4′-diaminoazobenzene were simulated [423]. The simulated basal spacings with the trans-isomers of aminoAZ and 4,4′-diaminoazobenzene were 2.0 and 2.1 nm, and those with the cis-isomers were 1.8 nm. The simulation did not match the experimental observations, suggesting that the photoswitching of the basal spacing was not explained simply by the isomerization of azobenzene.

To state the point, more recently, it was reported that the basal spacing of AZC₂N⁺C₂OH-magadiite under humidity of 5% was negligibly small [346], while the change between 2.69 and 2.75 nm was observed when the reaction was conducted under ambient condition [165, 186]. It suggests that the basal spacing change was induced by the adsorption of vapors. The basal spacing of a naphthalene-substituted cationic azobenzene (NaphAZ in Scheme 16)-magadiite was observed during UV irradiation [346]. The basal spacing decreased by UV irradiation from 2.89 to 2.79 nm and did not return to the initial value, indicating that the molecular packing after the UV irradiation changed from the initial state.

Scheme 16

Molecular structure of a naphthalene-substituted azobenzene and a trifluoromethyl-substituted anionic azobenzene

The change of the basal spacing of clays was also induced by photoisomerization of a pyridine-substituted spiropyran (Py-SP in Scheme 12) [389]. Py-SP was intercalated into KF by two methods, (1) ion exchange with the interlayer sodium cation and (2) guest replacement using the ion exchange of the interlayer sodium cation with cetyltrimethylammonium bromide (CTAB) and subsequent exchange with Py-SP. The basal spacing of the hybrid prepared by the ion exchange was switched between 1.55 and 1.40 nm by 365 nm UV and 600 nm visible light irradiation, respectively. The basal spacing (1.55 nm) for the spiropyran form with a twisted molecular structure was larger than that for the merocyanine form with a planar molecular structure (1.40 nm). In contrast, the basal spacing (1.38 nm) of the hybrid prepared by the guest replacement did not change. The co-existing CTAB in KF expanded the inter-layer distance to accommodate Py-SP and Py-SP isomerized without changing the basal spacing.

The irradiation induced another change of a microscopic region. An azobenzene derivative with fluoroalkyl chain (C₃F₇N⁺AZC₆H₁₃ in Scheme 16) was intercalated in potassium hexaniobate (K₂Nb₆O₁₇) [342, 424] to obtain a spiral tube structure [424]. The interlayer distance of the tube was changed by irradiation of 368 nm UV and 463 nm visible light irradiation with a simultaneous structure change of the tube [342]. The size changes of the tube from 244 to 93 nm and to 170 nm by UV and visible irradiation, respectively, were observed (Fig. 15). As shown in Fig. 16, a bottom edge of a hybrid film of C₃F₇N⁺AZC₆H₁₃ and the niobate was slid out by UV irradiation (point A) up to 1.5 μm, and the edge returned to the initial position by the subsequent visible light irradiation (point C) [352, 425]. These reports suggest that the basal spacing change by photoisomerization of C₃F₇N⁺AZC₆H₁₃ with a nanometer scale induced the structure change with a micrometer scale.

Fig. 15

Atomic force microscopic (AFM) images of morphological change of a hybrid of C₃F₇N⁺AZC₆H₁₃ and the hexaniobate: (A) AFM top-view and (B) cross sections at the white dash lines of parallel of the short axis of the tube (Reproduced from the reference [342] with permission)

Fig. 16

3D morphology changes in the hybrid film. (a) Height profile of the hybrid film. (b) Relative distance from the reference point to the film edge vs the number of irradiation cycles. (c) Film thickness at a point E located a constant distance from the reference point vs the number of irradiation cycles (Reproduced from the reference [425] with permission)

5.3.2 Photoswitching of Wettability

Photoswitching of surface properties of layered material films [426] by functionalization with photochromic compounds is expected owing to polarity change with photoisomerizations. Photoswitching of wettability of a layered double hydroxide film (ZnAl-NO₃-LDH film) was reported by functionalization with a trifluoromethyl-substituted anionic azobenzene (CF₃AZC₅COO⁻ in Scheme 16) [427]. The film was synthesized by putting a porous anodic alumina/aluminum (PAO/Al) substrate into the solution containing zinc nitrate and ammonium nitrate [428] and the subsequent intercalation of CF₃AZC₅COO⁻. The morphology of the film of the LDH crystal seemed to have a curved hexagonal sheet structure (Fig. 17a). A water contact angle of the film was 151 ± 1°, while the contact angle decreased to 73 ± 1° by 365 nm UV irradiation (Fig. 17b, c). The contact angle was returned to the initial value by subsequent visible light (420 nm) irradiation. The cis-isomer of azobenzene has larger polarity owing to the bending structure, indicating that the surface of the film with the cis-isomer had higher polarity than that with the trans-isomer. An advantage of this system is thought to be an easy preparation method of ion exchange method [35, 36, 283], which is expected to make a film with a large area compared to other films that showed photoswitching of wettability [429, 430].

Fig. 17

SEM images of the top and cross section views of the hybrid of CF₃AZC₅COO⁻ intercalated ZnAl-LDH (a) and shapes of a water droplet on the hybrid film surface (b) before the UV irradiation and (c) after the UV irradiation (Reproduced from the reference [427] with permission)

5.3.3 Photoswitching of Magnetism

Photoswitching of magnetism of layered double hydroxides (LDHs) was reported by intercalation of photochromic compounds of spiropyrans [431,432,433,434,435], azobenzenes [436, 437], and diarylethenes [438,439,440]. Two main contributions of intralayer superexchange interactions between metal cations and interlayer dipole interactions between the LDH layers were claimed to control the magnetism of the LDHs [441]. Because the photochromic compounds change shapes, sizes, and polarities by irradiation, the photochromic compounds affected the magnetism of the LDHs-dye hybrids.

Magnetism of layered (C₃H₇)₄N[Fe^IIFe^III(C₂O₂S₂)₃] was explained by charge transfer at 120 and 6.5 K between Fe^II and Fe^III [432, 442, 443]. The phase transition at 120 K was attributed to the charge transfer between Fe^II (spin angular momentum S = 1/2) and Fe^III (S = 2) (the authors named high-temperature phase) to form the state with Fe^III (S = 5/2) and Fe^II (S = 0) (low-temperature phase). The layered (C₃H₇)₄N[Fe^IIFe^III(C₂O₂S₂)₃] showed ferromagnetism below 6.5 K due to the charge transfer between the Fe^III (S = 5/2) and the Fe^II (S = 0). A hybrid of [Fe^IIFe^III(C₂O₂S₂)₃] with MePy-SP (Scheme 17) [431, 433, 434] showed the charge transfer phase transition from the high-temperature phase to the low-temperature phase at 75 K and the ferromagnetic phase transition at 5 K [434]. Upon UV irradiation (365 nm), the intercalated MePy-SP isomerized to a merocyanine form (MePy-MC), and the isomerization induced the charge transfer between Fe^II and Fe^III. As a result, the charge transfer phase transition from the high-temperature phase to the low-temperature phase was not observed, and the ferromagnetic phase transition was observed at 22 K. Taking into account that an alkylammonium surfactant with a long alkyl chain stabilized the high-temperature phase and increased the ferromagnetic phase transition temperature T_C, [431] the longer molecular length of MePy-MC than that of MePy-SP increased T_C.

Scheme 17

Anionic photochromic compound intercalated into LDHs for photoswitching of magnetism

The magnetism of LDHs depended on the dipole interactions between the LDH layers which are smaller with longer interlayer distance [441]. A CoAl-LDH (Co_0.69Al_0.31(OH)₂(CO₃)_0.155(H₂O)_0.3) had spontaneous magnetization below the critical temperature T_M at 4.7 K [444]. T_M of a CoAl-LDH (Co_0.65Al_0.35(OH)₂) was switched by the cis-trans photoisomerization of the intercalated dicarboxylazobenzene (designated as AZCOO⁻) [436]. T_M of the CoAl-LDH hybrid (4.5 K) increased to 5.2 K by 355 nm UV irradiation accompanied by the basal spacing change from 20.29 to 20.18 nm, suggesting that the change in the nanostructure was the reason of the increase of T_M. The basal spacing did not revert to the initial value by visible light irradiation in acetonitrile, while it reverted by exposure of water under the visible light irradiation. As discussed in Sect. 5.3.1, vapor played an important role in the change in the basal spacing [346]. It is thought that the adsorption of the water molecules triggered the backward reaction to the initial molecular packing.

Magnetism of alkylcarboxylate intercalated layered Co and Cu hydroxides (Co₇(OH)_11.6 and Cu₂(OH)₂) depended on the length and π-conjugate system of the alkylcarboxylate surfactants due to the interlayer ferromagnetic interaction [445]. The photocyclization of diarylethene, which accompanies the switching of the π-conjugate system, affected the magnetic interactions of two nitronyl nitroxide moieties substituted to the phenylthiophene moieties of the diarylethene [446, 447]. T_C of a hybrid of a layered Co₄(OH)₇ (designated as Co-LDH) intercalated an open-ring isomer of SO₃DAE in Scheme 17 was 9 K, while that after UV irradiation increased to 20 K [439]. Although the mechanism was not reported in detail [438, 440], the photoswitch of the π-conjugate system of the intercalated SO₃DAE seemed to affect the magnetism of Co-LDH [446, 447].

The photoswitching of the π-conjugate system by the photocyclization and the photoswitching of the basal spacing caused by photochromism affected to the magnetism of the LDHs. As discussed in Sects. 5.2.2 and 5.3.1, the design of the hybrids and the photoinduced adsorption of vapor are expected to make these phenomena effective, suggesting that the elucidation and improvement of the photoinduced phenomena of the hybrids induce the effective magnetism switching.

5.3.4 Photoinduced Adsorption

As discussed in Sect. 5.3.1, adsorption of vapor during the photochromic changes induced the change in the basal spacing [346]. Adsorption of phenol onto organically modified clay from an aqueous solution was also reported [101, 282, 448, 449]. Motivated by the phenomena, photoinduced adsorption of phenol was examined [166, 167, 344]. AZC₂N⁺C₂OH and C₂AZC₂N⁺ (Table 3) intercalated KF were mixed with neat phenol. As shown in Fig. 18, gallery heights increased from 0.81 to 1.5 nm and 0.96 to 2.6 nm by mixing with phenol, respectively [344]. The gallery heights of the hybrids AZC₂N⁺C₂OH-KF with phenol increased further from 1.5 to 2.3 nm by UV irradiation. The large change of the gallery heights (0.8 nm) indicated that the phenol was intercalated into the hybrid by the UV irradiation. On the other hand, the gallery height of C₂AZC₂N⁺-KF increased from 2.6 to 2.8 nm by the UV irradiation in the presence of phenol. Because the polarity of C₂AZC₂N⁺ was thought to be smaller than that of AZC₂N⁺C₂OH from a comparison of dipole moments of cis-AZ (0.41 Debye) and 4-ethylazobenzene (0.34 Debye), the adsorbed amount of phenol in C₂AZC₂N⁺-KF intercalated with phenol was thought to be smaller than that of AZC₂N⁺C₂OH-KF. The basal spacing changes of AZC₂N⁺C₂OH-KF with the adsorbed amount of 105, 91, and 66 meq/100 g were determined as summarized in Table 4 [167]. Observation of several basal spacing with 91 meq/100 g suggested the inhomogeneous adsorption of phenol. The changes in the basal spacings of AZC₂N⁺C₂OH-smectites with varied CEC were also examined to find C₂AZC₂N⁺-clay (with CEC of 108 and 84 meq/100 g) increased by the UV irradiation, while that with 71 meq/100 g did not change. Changes of the basal spacings of KF, TSM, and SA intercalated AZC₂N⁺C₂OH in the presence of phenol by the UV irradiation are summarized in Table 4 [167]. The fraction of the cis-isomer of AZC₂N⁺C₂OH in TSM was estimated to be ca. 50%, and it was larger than that in KF (ca. 30%).

Fig. 18

The changes in the basal spacings of KF intercalated AZC₂N⁺C₂OH and C₂AZC₂N⁺; (a) before the intercalation of phenol, (b) after phenol intercalation, (c) after UV irradiation, and (d) after subsequent visible light irradiation (Reproduced from the reference [344] with permission)

Table 4 Photoinduced expansion of basal spacings of azobenzene exchanged clays

To enhance the structural change by the photoinduced adsorption, the following parameters were proposed: (1) fraction of cis-isomer at the photostationary state [167, 344], (2) amount [344], and (3) orientation [167] of azobenzene derivative (corresponding to the surface coverage).

The photoinduced adsorption of photochromic compounds was also reported recently [374, 382, 450, 451]. As discussed in Sect. 5.1.3, photochromism of spiropyran was affected by the presence of nanospace materials including mesoporous silicas and smectites. Color of 6-NO₂-MC (Scheme 12) photochemically formed from 6-NO₂-SP was blue in toluene under UV irradiation, while the color of it was red in the presence of HE [374]. Red-colored sediment was collected by storing the suspension in the dark so that the adsorbed amount of MC onto HE was followed, showing that the 6-NO₂-SP was adsorbed on HE with the first-order kinetics [374]. The adsorbed amount of 6-NO₂-MC was 2.8 mg/100 g. The photoinduced adsorption is thought as molecular migration between hydrophobic and hydrophilic phases by using the bistability of the photochromic compound.

Mesoporous silicas were used as adsorbent of photo 6-NO₂-MC, which is named as “photoinduced adsorption” [382, 450, 452]. Photoinduced migration of SA/MC between the mesoporous silicas and the organophilic clay [381, 382] was reported to lead the reversibility of the negative photochromisms. The organically modified clay was thought to host less polar 6-NO₂-SP, which formed by the negative photochromisms of 6-NO₂-MC accommodated in the mesoporous silica. The thermal coloration of 6-NO₂-SP to 6-NO₂-MC accelerated by using the mesoporous silicas with larger pore size [382] and that whose external surface was functionalized with phenyl groups [451], suggesting that the diffusion inside the pore is important and the intraparticle diffusion was more efficient than the interparticle diffusion.

6 Conclusions and Future Perspectives

Recent developments on the preparation and the photofunctions of dye-clay hybrids are reviewed. Photofunctions have been systematically controlled and discussed based on the variation of hosts, guests, and their compositions. These developments led to further understanding of the structure and composition-property relationships, which may provide advance knowledge to optimize materials’ performances. Not only the nanostructure design by using host-guest interfaces has been revised, but the search for raw materials and their synthesis of new host materials have been also done. Successful morphosyntheses of known materials to obtain well-defined particles and single crystals with narrow particle size distribution expanded the possibilities of the materials and made detailed characterization possible. Taking the advantages of the morphosyntheses, interfacial design and the developments of fabrication techniques, some dye-clay systems have been prepared as thin films. The application covers traditional pigments application to agriculture (on soil contamination), optical devices including light-emitting ones, sensors, photocatalysts, etc. Though the practical application is not seen, various unique photoinduced phenomena have been also ascribed.