Influence of electronic vs. steric factors on the solid-state photochromic performances of new polyoxometalate/spirooxazine and spiropyran hybrid materials

Abstract

Two new photochromic hybrid organic–inorganic supramolecular materials (SN)₂[W₆O₁₉] (SN₂W₆) and (SP)₂[W₆O₁₉]·0.5CH₃CN (SP₂W₆) have been elaborated by combining for the very first time an isopolyoxotungstate [W₆O₁₉]²⁻ with both photoswitchable cationic spirooxazine (SN) and spiropyran (SP), via electrostatic interactions. (SN)₂[Mo₆O₁₉] (SN₂Mo₆), a third new photochromic compound which contains the [Mo₆O₁₉]²⁻ unit, has been also designed. Their structure was solved by single-crystal X-ray diffraction analysis, and their solid-state photochromic properties in ambient conditions were thoroughly investigated by UV-visible diffuse reflectance spectroscopy. SN₂W₆ and SN₂Mo₆ have been proved to be isostructural while SP₂W₆ is isostructural with its already reported counterpart SP₂Mo₆. In the SN₂M₆ and SP₂M₆ series, the SN and SP molecules are in very close contact with the POM units, which a priori should dramatically limit their photoisomerization in the solid state. However, this strongly varies with the nature of the metal in the [M₆O₁₉]²⁻ unit. Indeed SN₂W₆ and SP₂W₆ exhibit remarkable photochromism under low-power UV excitation (λ_ex = 365 nm) while SN₂Mo₆ and SP₂Mo₆ develop very limited color-change effects under similar UV exposure. These results have unambiguously highlighted that the efficiency of the SN and SP photoisomerizations in these hybrid systems is much more impacted by an electronic factor resulting from the intrinsic optical absorption properties of both SP and POM components than by a steric factor induced by crystal packing.

---

1. Introduction

In recent years, considerable efforts have been made to elaborate efficient solid-state photochromic materials that exhibit reversible light-induced switching between two optically distinguishable states.¹ The ability to control their optical properties with light is of great interest in a wide range of technological and marketable applications, such as smart pigments for paints and windows,² optical switches,³ and more recently, 3D high-density optical data storage.⁴ For all these listed applications, ideal materials should develop efficient photoactivity in the solid state and in ambient conditions (strong coloration contrast, optical bistability, fast photoswitch rates, high cyclability). Currently several organic molecules possess attractive photoswitchable properties but many do not meet all the criteria. For example, the classes of spiropyrans⁵ (SP) and spirooxazines⁶ (SN) have probably been the most widely studied in these last fifty years, because of their high optical performances and their synthetic accessibility. Such molecules absorb UV irradiation in the 350–400 nm range, and exhibit efficient ring-opening/ring-closure processes between a nonplanar colorless closed form and a coloured opened merocyanin (MC) form (Fig. 1a). MC exhibits a broad absorption band in the visible domain and can be converted back into the closed-ring form by visible-light irradiation or heating. These reversible photoisomerizations are generally highly efficient in solution or in crystalline state but only at very low temperature.⁷ Actually, in most cases, SP and SN solid-state photochromism is weak or totally annihilated in ambient conditions, which strongly limits the incorporation of such molecules into efficient photoadressable solid devices. Last decades, several innovative strategies were successfully managed to increase the photochromic activities of spiro-derivative molecules in the solid state. For example, the embedding of spiropyrans and spirooxazines into the cavities of porous organic polymers or sol–gel host matrixes leads to amorphous soft materials with strong photoresponses.⁸ The elaboration of crystallized hybrid systems is an alternative approach which, besides, allows the X-ray structural determination of the spiro-derivative/host matrix interface, opening the way to the establishment of fine structure/property relationships.

Fig. 1 (a) Photoinduced equilibriums involving the cationic spiropyran (SP) and spirooxazine (SN). (b) Polyhedral representation of the [M₆O₁₉]²⁻ (M = Mo, W) Lindqvist-type polyoxometalate (gold sphere: oxygen).

Recently, we evidenced that the coupling with polyoxometalates (POMs) constitutes an attractive opportunity to reach such crystallized hybrid materials, and to drastically improve the solid-state photochromic performances of spiro-derivatives at room temperature. POMs are described as soluble metal-oxides of early transition metals (usually W, Mo, V) in high oxidation states.⁹ The huge diversity of their compositions, structures, charges and sizes make them relevant for applications in numerous fields in optic, which include photochromism,¹⁰ electrochromism,¹¹ NLO,¹² and photoluminescence.¹³ Focusing on photochromism, it has been shown that the assembly of POMs with SP molecules can in some cases exalt the SP → MC photoconversion because of the strong stabilization of the zwitterionic MC form in polar POM frameworks. Noticeably, innovative bistable materials with remarkable photoresponses were obtained from spiropyrans and spirooxazines initially nonphotochromic in their crystalline state.¹⁴ Two new classes of hybrid materials were conceived from two complementary synthesis strategies. The covalent approach consisted in grafting via peptidic bonds one or two neutral SP molecules onto an Anderson-type {MnMo₆O₁₈} platform.¹⁵ The first multi-photochromic single molecule incorporating one SP and one SN fragments onto a POM core has been even successfully obtained.¹⁶ In a second ionic approach, supramolecular SP/POM assemblies were designed by associating a cationic spiropyran (SP) with different polyoxomolybdates via electrostatic interactions.¹⁷ Strikingly, the SP ⇆ MC photoconversion efficiency strongly varies in such ionic hybrid systems. For example, (SP)₄[Mo₈O₂₆]·2CH₃CN and (SP)₃(NH₄)[Mo₈O₂₆] are remarkably photochromic, while (SP)₂[Mo₆O₁₉]·0.5CH₃CN (SP₂Mo₆) exhibits a very limited UV-induced color-change effect.^17a

Overall, the photochromic performances of the SP/POM assemblies are governed by both steric and electronic factors. The first one reflects the relative arrangement of the SP cations and the POM units in the frameworks, while the second one refers to the relative positioning of their absorption bands in wavelength space. For example, the weak photochromic performances of SP₂Mo₆ were assumed to result from a negative combination of these two factors. In contrast with best photochromic systems, the SP cations and the [Mo₆O₁₉]²⁻ units are strongly linked via π–π stacking interactions. This specific arrangement dramatically limits the free volume around the SP molecules and should disadvantage their photoisomerization. In addition, the absorption bands of the [Mo₆O₁₉]²⁻ anion strongly overlap those of the SP cation in the UV domain. Consequently, [Mo₆O₁₉]²⁻ should dramatically absorb the UV excitation energy needed to activate the ring-opening process of the spiropyran.

To date, the influence of both electronic and steric factors on the photochromism efficiency of the SP/POM assemblies have not been uncorrelated because both factors vary in all reported materials. Herein, the main goal of the present work is to specifically investigate the effect of the electronic factor, everything else being equal. For a given topology, the absorption bands of the POMs vary in energy with the nature of the metal.¹⁸ Then, the substitution of Mo by W in the POM units appears as a very useful tool to tune the electronic factor without any structural change. Furthermore, the combination of spiropyrans and spirooxazines with isopolyoxotungstates using the ionic approach has never been explored so far. With this aim, we report the synthesis of a new supramolecular system (SP)₂[W₆O₁₉]·0.5CH₃CN (SP₂W₆) which is built upon the same SP cation (Fig. 1a) associated with the [W₆O₁₉]²⁻ unit. Noticeably, SP₂W₆ is isostructural with SP₂Mo₆, what allows keeping constant the steric factor.

To extend our study to other spiro-derivatives, we have also designed two new isostructural hybrid systems (SN)₂[W₆O₁₉] (SN₂W₆) and (SN)₂[Mo₆O₁₉] (SN₂Mo₆) which combine for the very first time POMs units and a rare cationic spirooxazine (SN) derivative (Fig. 1a). The high photochromic performances of spirooxazines are well-recognized and have found applications in ophthalmology, being used for example in Orgavert™ sunglasses and in PHOTOLITE™ photochromic lenses.^5a It has also been noticed that while the structure of SN and SP are close, their photochromic properties can differ both on a photogenerated hue and a coloration rate point of view.¹⁶ The structural characterizations of SP₂W₆, SN₂W₆, and SN₂Mo₆ are thus reported. Besides, their optical properties are widely discussed and compared with those of SP₂Mo₆. This study reveals that, in strong contrast with their Mo counterparts, SN₂W₆ and SP₂W₆ exhibit remarkable solid-state photochromism in ambient conditions.

2. Results and discussion

2.1 Synthesis and structural description

(SP)₂[W₆O₁₉]·0.5CH₃CN (SP₂W₆), (SN)₂[W₆O₁₉] (SN₂W₆), and (SN)₂[Mo₆O₁₉] (SN₂Mo₆) all contain the Lindqvist-type [M₆O₁₉]²⁻ anion¹⁹ (M = Mo or W) (Fig. 1b) which is built upon six edge-shared [MO₆] octahedra. Yellow-ochre powder of SP₂W₆ has been isolated from a method derived on the synthesis of (SP)₂[Mo₆O₁₉]·0.5CH₃CN (SP₂Mo₆),^17a starting from (NBu₄)₂[W₆O₁₉] and two equivalents of SP(NO₃) dissolved in acetonitrile. Orange single crystals have been also obtained using slow diffusion of ethanol into the filtrate. SN₂W₆ and SN₂Mo₆ have been synthesised starting from (NBu₄)₂[M₆O₁₉] and two equivalents of SN(NO₃) dissolved in DMF. Orange-red single crystals of SN₂W₆ and red single crystals of SN₂Mo₆ have been directly obtained from the solutions using slow diffusion crystallization techniques. Elemental analyses, TGA/DSC measurements (Fig. S1, ESI†), and IR spectroscopy analyses (Fig. S2, ESI†) show that these two latter compounds are free from crystallized DMF molecules. Importantly, the coupling of the SP and SN molecules with the [M₆O₁₉]²⁻ anion has a beneficial effect on the thermal resistance of the photoswitchable organic molecules. SN₂W₆ and SN₂Mo₆ decompose at 245 °C and 250 °C respectively, whereas SN(NO₃) is thermally stable up to 190 °C only. Similarly, SP₂W₆ and SP₂Mo₆ decompose at 240 °C and 255 °C respectively, while SP(NO₃) is thermally stable up to 200 °C only.

Crystal structures of SP₂W₆, SN₂Mo₆, and SN₂W₆. Single-crystal X-ray diffraction analysis of SP₂W₆ reveals that this new compound is isostructural with SP₂Mo₆. SP₂W₆ is built upon isolated [W₆O₁₉]²⁻ units, SP cations and crystallized acetonitrile molecules that are assembled in a supramolecular framework. The [W₆O₁₉]²⁻ entities are well surrounded by the SP cations which interact with the POM surfaces via electrostatic interactions and C–H⋯O contacts (Fig. 2a). In particular, the planar methylpyridinium rings of two SP ions and the oxygen facets of the [W₆O₁₉]²⁻ unit perfectly face each other with a very low dihedral angle (α = 4.0°), and the shortest C⋯O distance of 3.389(3) Å. These values are quite comparable with those observed for SP₂Mo₆ (d_C⋯O = 3.372(5) Å, and α = 4.4°).

Fig. 2 Mixed polyhedral and ball-and-stick representations of the crystal packing in (a) SP₂W₆ and (b) SN₂W₆ showing the interactions between the π-donor methylpyridinium groups of the SP and SN cations and the [W₆O₁₉]²⁻ oxygen facets. (blue octahedra = WO₆, grey spheres: carbons of SP and SN cations; dark spheres: carbons of acetonitrile molecules; green sphere: nitrogen; gold sphere: oxygen). N⋯O interactions between acetonitrile molecules and [W₆O₁₉]²⁻ units in SP₂W₆ are displayed as dotted lines. The hydrogen atoms are omitted for clarity.

The isostructural compounds SN₂W₆ and SN₂Mo₆ are built upon [M₆O₁₉]²⁻ blocks and SN cations (Fig. 2b). The structure of SN₂W₆ has been arbitrary chosen to describe the crystal packing of this series. The [W₆O₁₉]²⁻ anions are well separated from each other in a cage-like supramolecular organic framework delimited by the SN cations (the shortest O⋯O distance between two adjacent [W₆O₁₉]²⁻ units is 5.67 Å). The methylpyridinium group of four SN cations are quite parallel with the oxygen facets of the [W₆O₁₉]²⁻ block and they interact together in a large contact zone, with a short interplanar distance (shortest C⋯O distance of 2.970(5) Å), and a low dihedral angle (α = 7.5°). These values match very well with those observed for SN₂Mo₆ (d_C⋯O = 2.954(1) Å, and α = 7.0°).

To sum up, the structural studies of the SP₂M₆ and SN₂M₆ series well evidence that the spiro-molecules undergo very strong geometrical constraints, due to their very close proximity with the [M₆O₁₉]²⁻ units, which should a priori be highly unfavourable for efficient photoisomerization. Nevertheless in each series, the organic cations interact with the POMs in a comparable way, enabling to maintain constant the steric factor, and to specifically investigate the influence of the nature of the metal in the POM unit, and thus the electronic factor, on the solid-state photochromic performances.

2.2 Optical properties

Optical properties before UV irradiation. The optical properties of SP₂W₆, SP₂Mo₆, SN₂W₆ and SN₂Mo₆ before UV irradiation have been investigated in ambient conditions by diffuse reflectance spectroscopy of microcrystalline powders. The main optical characteristics of the four hybrid systems are gathered in Table 1. Fig. 3 displays the absorption spectra of SP₂W₆ and SP₂Mo₆ compared with those of the iodide salt of the SP cation SP(I) and the (NBu₄)₂[M₆O₁₉] salts that contain similar [M₆O₁₉]²⁻ units without SP entities. As previously discussed,^17a the absorption spectrum of SP₂Mo₆ (Fig. 3a) in the UV domain is the sum of the absorption bands of the SP cation (for which the low-energy absorption is located at λ_max^(spiro) = 350 nm), and the broad ligand-to-metal charge-transfer (LMCT) transitions of the [Mo₆O₁₉]²⁻ anion¹⁸ (λ^(POM) = 269 and 327 nm). The absorptions of both POM and SP moieties exhibit a significant degree of overlapping which can be well assessed by introducing the Δλ parameter defined as Δλ = λ_max^(spiro) − λ_max^(POM), with λ_max^(POM) the low-energy LMCT transition centred on the POM unit. SP₂Mo₆ is then characterized by a very small Δλ value (23 nm). In strong contrast, for SP₂W₆ (Fig. 3b), the LMCT transitions of [W₆O₁₉]²⁻ are located at much higher energies than those of [Mo₆O₁₉]²⁻ i.e., λ^(POM) = 220 nm (not shown in Fig. 3b) and 280 nm,¹⁸ and then, the Δλ value significantly increases to reach 70 nm. Furthermore, the [W₆O₁₉]²⁻ anion does not absorb at λ_max^(spiro) = 350 nm. A similar effect is observed in the SN₂M₆ series (Fig. S3, ESI†). Indeed, Δλ = 48 and 95 nm for SN₂Mo₆ and SN₂W₆, respectively. These values are higher than those of SP₂Mo₆ and SP₂W₆ because SN absorbs at a higher wavelength (λ_max^(spiro) = 375 nm) than SP. In summary, the electronic factor is strongly modified by substituting Mo by W in the SP₂M₆ and SN₂M₆ series. This leads to a better separation in energy between the absorption bands of both organic and inorganic components in the UV domain. Consequently, in marked contrast with its Mo counterpart, the [W₆O₁₉]²⁻ anion does not compete with SP and SN to absorb the excitation UV energy at λ_ex = 365 nm needed to activate their ring-opening processes.

Table 1 Characteristic optical parameters of the SP₂M₆ and SN₂M₆ series

	SP2Mo6	SP2W6	SN2Mo6	SN2W6
a Lower energy POM LCMT transition.b Δλ = λmax^(spiro) − λmax^(POM).c Range of wavelength for the charge-transfer transition.d Color of powdered sample before UV irradiation.
λmax^(POM)^a (nm)	327	280	327	280
λmax^(spiro) (nm)	350	350	375	375
Δλ^b (nm)	23	70	48	95
λCT^c (nm)	430–540	400–510	430–590	420–560
Color^d	Brown-ochre	Yellow-ochre	Red-ochre	Brown-ochre

---

Fig. 3 Normalized Kubelka–Munk transformed reflectivity spectra before UV irradiation of (a) SP₂Mo₆ (black), SP(I) (magenta), and (TBA)₂[Mo₆O₁₉] (blue), and (b) SP₂W₆ (black), SP(I) (magenta), and (TBA)₂[W₆O₁₉] (blue). The Δλ parameter is defined as Δλ = λ_max^(spiro) − λ_max^(POM), with λ_max^(POM) the low-energy LMCT transition of the POM unit.

The absorption spectra of SP₂W₆, SN₂W₆ and SN₂Mo₆ also exhibit an additional weak absorption band in the range 400–600 nm (Table 1), which is responsible for the color of powdered materials. This absorption is ascribed to an intermolecular Charge-Transfer (CT) transition between the electron-rich “π-donor” spiro-derivative and the electron-poor [M₆O₁₉]²⁻ “acceptor”, due to their close proximity in the solid-state (see above).^17a The CT band totally disappears when the samples are dissolved in highly polar solvents such as DMSO, and the resulting spectra exactly match the sum of the absorption spectra of isolated organic and [M₆O₁₉]²⁻ ions (Fig. S4, ESI†). For a given electron-donor organic molecule, the energy of the CT transition increases with that of the LMCT band of the associated POM unit.^17a Then, λ_CT are systematically blue-shifted in SN₂W₆ and SP₂W₆, and the color of these two powdered samples are less pronounced than those of their Mo counterparts (Fig. S5, ESI†), enabling us to speculate a better coloration contrast during the photochromic process.

Solid-state photochromic properties. First, the photochromic properties of SP₂W₆ and SP₂Mo₆ in ambient conditions have been compared (Fig. 4). While the color-change effect of SP₂Mo₆ is very limited under low-power near UV excitation (λ_ex = 365 nm),^17a SP₂W₆ exhibits a remarkably improved photochromic response under similar UV exposure (Fig. 4a). The color of the yellowish-ochre powder gradually shifts to purple-brown with a high coloration contrast and a fast coloration speed. Similarly as observed for SP₂Mo₆, the color change of SP₂W₆ deals with the appearance of the MC form of the SP cation that is optically characterized by an absorption band arising at λ_max^(MC) = 590 nm (Fig. 4b). This quite evidences that, even if both materials are isostructural, the SP → MC photoisomerization is much more efficient in SP₂W₆ than in SP₂Mo₆. This can be explained considering that in the case of SP₂Mo₆, most of the incident UV-light flow is strongly absorbed by the POM unit, and the resulting incident UV-light flow, available for the SP → MC photoisomerization, is much more limited than for SP₂W₆. Thus, the concentration of the photogenerated MC form is weaker in the hybrid polyoxomolybdate, and consequently, its color-change effect is dramatically affected. Furthermore, let us also noticed that the [M₆O₁₉]²⁻ units do not contribute to the photogenerated coloration because, in contrast with the known photochromic polyoxometalate/organoammonium hybrid materials,¹⁰ their photoreduction does not occur in SP/POM assemblies.

Fig. 4 (a) Photographs of powders of SP₂W₆, and SP₂Mo₆ at different time (in min) during the coloration process under UV light (λ_ex = 365 nm). (b) Temporal evolution of the absorption at 590 nm for SP₂W₆ after 0, 0.166, 0.333, 0.5, 1, 2, 3, 4, 5, 7, 10, and 30 min of 365 nm-UV irradiation. (c) Experimental (point) and fitted (line) Abs⁵⁹⁰(t) vs. t plots for SP₂W₆ (■), and SP₂Mo₆ (●).

The solid-state photochromic properties of SN₂W₆, and SN₂Mo₆ in ambient conditions exactly follow the same trend (Fig. 5). Noticeably, the brown-ochre powder of SN₂W₆ gradually changes to dark-green under UV excitation at λ_ex = 365 nm (Fig. 5a). The photochromic effect that is detectable with naked eyes after only 2 s, finally reaches a saturation level after only 10 min, revealing a very fast photoresponse and a remarkable coloration contrast, even more pronounced than those of SP₂W₆. The intense photogenerated absorption band of the MC form of the SN cation is located at λ_max^(MC) = 620 nm in SN₂W₆ (Fig. 5b). This band is also observable in the optical spectrum of SN₂Mo₆ (Fig. S6, ESI†) but its intensity is much weaker. This latter compound develops very limited photochromic performances, and the color-change effect remains hardly detectable by naked eyes.

Fig. 5 (a) Photographs of powders of SN₂W₆, and SN₂Mo₆ at different time (in min) during the coloration process under UV light (λ_ex = 365 nm). (b) Temporal evolution of the absorption at 620 nm for SN₂W₆ after 0, 0.166, 0.333, 0.5, 1, 2, 3, 6, and 10 min of 365 nm-UV irradiation. (c) Experimental (point) and fitted (line) Abs⁶²⁰(t) vs. t plots for SN₂W₆ (■), and SN₂Mo₆ (●).

The coloration kinetics of the four hybrid systems have been quantified at room temperature by monitoring the photogenerated absorption at λ_max^(MC) (Abs^λ_max(t)) as a function of the UV irradiation time t. The Abs^λ_max(t) vs. t plots have been adequately fitted using a biexponential rate law^16,17a (Fig. 4c and 5c for the SP₂M₆ and SN₂M₆ series, respectively), and the kinetics parameters are gathered in Table S1, ESI†. The photocoloration rates of the compounds can be well compared by considering their half-life time (t_1/2) namely the UV irradiation time required for Abs^λ_max(t) to reach half of its maximum value. SN₂W₆ exhibits the most effective photocoloration process, much more intense and faster than the three other compounds. Importantly, its photocoloration rates (t_1/2 = 0.8 min) is twenty times faster than that of SN₂Mo₆ (t_1/2 = 16.0 min). In the SP₂M₆ series, the photocoloration rate is about eleven more times faster for SP₂W₆ (t_1/2 = 1.3 min) than for SP₂Mo₆ (t_1/2 = 15.0 min). This quite evidences that the SP and SN photoisomerization rates are drastically improved by substituting [Mo₆O₁₉]²⁻ by [W₆O₁₉]²⁻ in both series, despite the strong steric constraints applied to the spiro-derivative molecules by the POM framework.

The fading kinetics of SN₂W₆ and SP₂W₆ have been also investigated at room temperature to evaluate the efficiency of the MC → SN and MC → SP back conversions in the dark (thermal fading), and under visible-light irradiation at 630 and 590 nm, respectively, i.e. with energies corresponding to the absorption band of the MC forms of the SN and SP ions. The Abs^λ_max(t) vs. t plots of the coloration and fading processes are displayed in Fig. 6a for SN₂W₆ and S7, ESI† for SP₂W₆. Decays have been well fitted using a biexponential rate law, (see Table S2, ESI† for the detailed fading kinetic parameters). The fading rates have been compared by considering their fading half-life time. Under visible-light irradiation, SN₂W₆ and SP₂W₆ exhibit very fast and full bleaching processes. Very interestingly, the fading rates are comparable with the coloration ones (for the bleaching process, t_1/2 = 0.6 and 0.2 min for SN₂W₆ and SP₂W₆, respectively). In strong contrast, both hybrid systems develop very limited thermal fading processes. Indeed, the maximum of absorption loss after 2 hours reaches only 6% for SN₂W₆ and 8% in the case of SP₂W₆, revealing that after switching off the UV irradiation, most of the MC forms remain stabilized in the polar hybrid POM framework. Hence, SN₂W₆ and SP₂W₆ are quasi-bistable at room temperature.

Fig. 6 (a) Comparison of the temporal evolutions of the absorbance at 620 nm for SN₂W₆ under red light (λ_ex = 630 nm) (▲) and in the dark (●) for a sample initially irradiated for 10 min with UV light (λ_ex = 365 nm) (■). The dashed line shows the absorbance value just before switching off the UV light. (b) Evolution of the absorbance monitored at 620 nm for SN₂W₆ during successive coloration/bleach cycles at room temperature. In each cycle, the sample was irradiated for 1 min with UV light (λ_ex = 365 nm), and then for 10 min with red light (λ_ex = 630 nm). The red line corresponds to the linear fit of Abs⁶²⁰(n) vs. n (R² = 0.965).

Finally, the robustness of SN₂W₆ and SP₂W₆ for repeatable “recording–erasing” processes at room temperature was investigated by monitoring the evolution of Abs^λ_max of samples alternatively irradiated under UV light and visible light. We managed to perform a long study of cyclability by following the variation of the absorption on approximately fifty cycles, what to our knowledge, was never reported for powdered photochromic samples so far. In the case of SN₂W₆, the evolution of Abs⁶²⁰(n) vs. n, i.e., the number of coloration/fading cycles, reveals that this hybrid system shows a good cyclability with a high coloration contrast (Fig. 6b). For each cycle, the bleaching process is full, but a fatigue is observed during the successive photocoloration processes. The photogenerated absorption progressively decreases with n, and the absorption loss is about 33% at the end of fifty cycles. Surprisingly, the decrease of Abs⁶²⁰(n) with n, can be fitted with the empirical linear relationship described in eqn (1). This relationship allows predicting that the variation of absorption between both colorless and colored states should be zero at the end of 128 cycles.

Abs⁶²⁰(n) = −0.0045n + 0.5751 (1)

The photocoloration fatigue is more marked for SP₂W₆, and the absorption loss reaches about 50% at the end of fifty cycles (Fig. S7b, ESI†). Similarly, Abs⁵⁹⁰(n) linearly decreases with n according to eqn (2), revealing that the color-change effect should disappear at the end of 98 cycles.

Abs⁵⁹⁰(n) = −0.0040n + 0.3916 (2)

At first sight, the poorer cyclability of SP₂W₆ could result from the presence of the crystallized acetonitrile molecules which is expected to introduce some flexibility in the hybrid framework.

3. Conclusion

In conclusion, the two first supramolecular hybrid isopolyoxotungstates SN₂W₆ and SP₂W₆ have been successfully designed by assembling the Lindqvist-type [W₆O₁₉]²⁻ unit with cationic spirooxazine and spiropyran. Noticeably, while the SN and SP ions undergo strong steric constraints in the frameworks, these systems exhibit remarkable solid-state photochromic properties in ambient conditions, which include high photocoloration contrasts, fast coloration and fading rates, and good cyclabilities. Besides, SN₂W₆ and SP₂W₆ are quasi-bistable systems at room temperature, and they possess high light-driven “recording–erasing” potentialities. Strikingly, their photochromic performances are drastically improved compared to SN₂Mo₆ and SP₂Mo₆, their isostructural Mo counterparts, revealing that the photoisomerization of the spiro-molecules is much more impacted by the electronic factor than by the steric one. In strong contrast with [Mo₆O₁₉]²⁻, the high-energy LMCT transition of [W₆O₁₉]²⁻ does not overlap the absorption of the SN and SP cations in the UV domain, and both organic and inorganic components do not compete to absorb the excitation UV light.

The findings may open a new pathway for the design of supramolecular SP or SN/POMs assemblies with outstanding solid-state photochromic properties. Importantly, they quite evidence that efficient photochromic hybrid systems can be designed even if their crystal packings are a priori highly unfavourable for the photoisomerization of the spiro-molecules. In addition, the substitution of Mo by W is a very useful tool to favourably tune the electronic factor. We are currently working to characterize the structure of SP₂W₆ and SN₂W₆ in their colored forms using single-crystal X-ray photodiffraction analyses. Exploitation of the remarkable photochromic properties of SN₂W₆ and SP₂W₆ is also now under investigation. In particular, their integration into organic polymers or as thin film to elaborate photonic devices and sensors will be considered. This kind of immobilisation is also expected to increase the resistance to “fatigue”.²⁰ In addition, the synthesis of new supramolecular assemblies built upon other isopolyoxotungstates varying in shape, size and charge are currently under study. Finally, new hybrid systems, based on polyoxotungstates and photoactive organic dyes associating both photochromism and fluorescence are also underway.

4. Experimental section

Synthesis

SP(NO₃),²¹ SN(I),²² (NBu₄)₂[W₆O₁₉]²³ and (NBu₄)₂[Mo₆O₁₉]²⁴ have been synthesized according to the reported procedures.

SN(NO₃). SN(I) (1 g, 2.1 mmol) was dissolved in a cosolvent MeOH/CH₃CN (30/25 mL) (solution 1). AgNO₃ (360 mg, 2.1 mmol) was dissolved in MeOH (15 mL) (solution 2). Solution 2 was added dropwise to solution 1 under stirring leading to the precipitation of a yellow solid of AgI. The solution was filtered through a short plug of celite to afford an oily solution after removal of the solvent under reduced pressure. The solution was then precipitated in ether and the brown powder obtained was washed with ether and dried in air. SN(NO₃) was obtained in 40% yield. ¹H NMR (300 MHz, CDCl₃) δ (ppm) 9.82 (d, 1H), 9.60 (d, 1H), 8.30 (d, 1H), 8.1 (dd, 1H), 7.92 (s, 1H), 7.7 (d, 1H), 7.11 (d, 1H), 6.97 (t, 1H), 6.62 (d, 1H), 4.84 (s, 1H), 2.77 (s, 3H), 1.40 (s, 3H), 1.36 (s, 3H).

(SN)₂[Mo₆O₁₉] (SN₂Mo₆). (NBu₄)₂[Mo₆O₁₉] (220 mg, 0.16 mmol) was dissolved in DMF (9 mL). The solution (solution 1) was stirred for a few minutes at room temperature. In addition, a second solution (solution 2) was obtained by dissolving SN(NO₃) (150 mg, 0.37 mmol) in DMF (5 mL) at room temperature. Solution 2 was added dropwise to solution 1 under vigorous stirring. The mixture was stirred at 40 °C for two hours, kept at room temperature, and then filtered. Crystallization of the product by diffusion of ethanol into the filtrate gave suitable single crystals for X-ray diffraction. Yield in Mo: 64%. Anal. calcd for C₄₄H₄₄O₂₁N₆Mo₆: C, 33.42; H, 2.78; N, 5.32. Found: C, 33.38; H, 2.80; N, 5.29. FT-IR (KBr, cm⁻¹): SN cations 1629 (m), 1604 (s), 1584 (s), 1531 (m), 1487 (s), 1467 (m), 1441 (m), 1425 (m), 1394 (s), 1368 (m), 1349 (sh), 1306 (m), 1284 (s), 1237 (m), 1189 (w), 1169 (m), 1147 (m), 1130 (m), 1119 (w), 1108 (sh), 1079 (m), 1055 (s), 1023 (m), 1003 (w); νMo=O, νMo–O–Mo 959 (vs), 928 (sh), 912 (w), 892 (w), 796 (vs), 751 (s), 682 (w), 655 (w), 623 (sh), 600 (m), 577 (m), 538 (m), 522 (sh), 477 (sh), 463 (m), 437 (m). From ATG/DSC measurements, SN₂Mo₆ is stable up to 280 °C.

(SN)₂[W₆O₁₉] (SN₂W₆). (NBu₄)₂[W₆O₁₉] (305 mg, 0.16 mmol) was dissolved in DMF (8 mL). The solution (solution 1) was stirred for a few minutes at room temperature. In addition, a second solution (solution 2) was obtained by dissolving SN(NO₃) (200 mg, 0.49 mmol) in DMF (5 mL) at room temperature. Solution 2 was added dropwise to solution 1 under vigorous stirring. The mixture was stirred at 40 °C for one hour and a half, kept at room temperature, and then filtered. Crystallization of the product by diffusion of ethanol into the filtrate gave suitable single crystals for X-ray diffraction. Yield in W: 46%. Anal. calcd for C₄₄H₄₄O₂₁N₆W₆: C, 25.19; H, 2.10; N, 4.01. Found: C, 25.15; H, 2.07; N, 4.06. FT-IR (KBr, cm⁻¹): SN cations 1628 (w), 1601 (m), 1583 (m), 1529 (m), 1487 (s), 1466 (m), 1440 (m), 1424 (m), 1395 (m), 1366 (m), 1347 (sh), 1324 (w), 1305 (m), 1283 (s), 1260 (sh), 1236 (m), 1188 (w), 1168 (m), 1144 (w), 1128 (m), 1119 (w), 1106 (w), 1077 (m), 1056 (m), 1024 (m); νW=O, νW–O–W 980 (vs), 927 (m), 907 (w), 890 (w), 810 (vs), 752 (s), 679 (m), 670 (m), 653 (sh), 632 (sh), 623 (sh), 583 (m), 537 (m), 520 (w), 443 (s). From ATG/DSC measurements, SN₂W₆ is stable up to 260 °C.

(SP)₂[W₆O₁₉]·0.5CH₃CN (SP₂W₆). (NBu₄)₂[W₆O₁₉] (305 mg, 0.16 mmol) was dissolved in acetonitrile (10 mL). The solution (solution 1) was stirred for a few minutes at room temperature. In addition, a second solution (solution 2) was obtained by dissolving SP(NO₃) (200 mg, 0.49 mmol) in acetonitrile (6 mL) at room temperature. Solution 2 was added dropwise to solution 1 under vigorous stirring leading to the slow precipitation of an ochre powder of SP₂W₆. The mixture was stirred at 40 °C for one hour and a half, kept at room temperature, and then filtered. The powder was washed with acetonitrile and ethanol, and dried in air. Yield in W: 53%. Crystallization of the product by diffusion of ethanol into the filtrate gave suitable single crystals for X-ray diffraction. Anal. calcd for C₄₁H_47.5O₂₁N_4.5W₆: C, 24.09; H, 3.32; N, 3.08. Found: C, 24.15; H, 3.34; N, 3.12. FT-IR (KBr, cm⁻¹): SP cations 1608 (m), 1486 (s), 1472 (m), 1454 (m), 1433 (w), 1386 (w), 1362 (m), 1311 (s), 1297 (s), 1279 (w), 1252 (w), 1236 (w), 1207 (sh), 1190 (w), 1168 (w), 1047 (w), 1017 (w), 1104 (w), 1057 (m), 1025 (m); νW=O, νW–O–W 977 (vs), 926 (m), 915 (m), 878 (w), 810 (vs), 765 (m), 707 (m), 661 (w), 585 (m), 552 (w), 530 (w), 445 (s). From DSC/TGA measurements, SP₂W₆ loses the acetonitrile molecules above 180 °C.

X-ray crystallography

Data collection was carried out by using a Siemens SMART three-circle diffractometer for SN₂Mo₆ and by using a Bruker Nonius X8 APEX 2 diffractometer for SN₂W₆ and SP₂W₆. Both were equipped with a CCD bi-dimensional detector using the monochromatized wavelength λ (Mo Kα) = 0.71073 Å. Absorption correction was based on multiple and symmetry-equivalent reflections in the data set using the SADABS program²⁵ based on the method of Blessing.²⁶ The structures were solved by direct methods and refined by full-matrix least-squares using the SHELX-TL package.²⁷ Crystallographic data are given in Table 2. CCDC 1414602–1414604 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via http://www.ccdc.cam.ac.uk/data_request/cif.†

Table 2 Crystallographic data

	SN2Mo6	SN2W6	SP2W6
Empirical formula	C44H44N6O21Mo6	C44H44N6O21W6	C41H47.5N4.5O21W6
Formula weight [g]	1568.49	2095.95	2042.43
Temperature [K]	298	298	298
Crystal system	Monoclinic	Monoclinic	Monoclinic
Space group	P21/n	P21/n	P21/n
a [Å]	14.3898(6)	14.410(3)	10.4923(2)
b [Å]	12.7305(5)	12.743(3)	15.1018(3)
c [Å]	14.4617(7)	14.549(3)	16.1338(2)
β [°]	111.279(2)	111.339(10)	104.6525(6)
V [Å^3]	2468.6(2)	2488.5(9)	2473.30(7)
Z	2	2	2
ρcalc [g cm^−3]	2.110	2.797	2.743
μ [mm^−1]	1.566	13.899	13.979
Data/parameters	7214/353	7253/353	7227/340
Rint	0.0353	0.0307	0.0489
GOF	0.959	1.032	1.119
R1 (>2σ(I))	0.0264	0.0208	0.0320
wR2	0.0600	0.0563	0.0908

---

Physical measurements

Elemental analyses of the solids were performed by the “Service de microanalyses ICSN CNRS, in Gif sur Yvette (France), and by the “Service d'Analyse du CNRS”, in Vernaison (France). FT-IR spectra were recorded in the 4000–400 cm⁻¹ range on a BRUKER Vertex equipped with a computer control using the OPUS software. Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) were performed by flowing dry argon with a heating and cooling rate of 5 °C min⁻¹ on a SETARAM TG-DSC 111 between 20 and 800 °C. Diffuse reflectance spectra were collected at room temperature on a finely ground sample with a Cary 5G spectrometer (Varian) equipped with a 60 mm diameter integrating sphere and computer control using the “Scan” software. Diffuse reflectance was measured from 250 to 1550 nm with a 2 nm step using Halon powder (from Varian) as reference (100% reflectance). The reflectance data were treated by a Kubelka–Munk transformation²⁸ to obtain the corresponding absorption data. The photocoloration and fading kinetics were quantified by monitoring the temporal evolution of the photogenerated absorption Abs^λ_max(t) defined as Abs^λ_max(t) = −log(R^λ_max(t)/R^λ_max(0)), where R^λ_max(t) and R^λ_max(0) are the reflectances at the time t and at t = 0, respectively. For the coloration kinetics, the samples were irradiated with a Fisher Bioblock labosi UV lamp (λ_ex = 365 nm, P = 6 W) at a distance of 50 mm. Abs^λ_max(t) vs. t plots have been fitted according to a biexponential rate law Abs^λ_max(t) = (A₁ + A₂) − A₁ exp(−k₁^ct) − A₂ exp(−k₂^ct), with k₁^c and k₂^c the extracted coloration rate constants. For the bleaching processes, the samples were first irradiated under 365 nm-UV excitation for 10 minutes in the case of SN₂W₆, and for 20 minutes in the case of SP₂W₆, times for which the photoinduced absorptions are quasi-saturated. Then, the compounds were put under Thorlabs LED Array light sources (590 nm to 1.4 W cm⁻² and 630 nm to 2.4 W cm⁻²) at a distance of 100 mm. The bleaching kinetics were determined at room temperature by monitoring the temporal decays of Abs^λ_max(t) of samples once irradiated Abs^λ_max(t) vs. t plots have been fitted according to a biexponential rate law Abs^λ_max(t) = (A₀ − A₁ − A₂) + A₁ exp(−k₁^ft) + A₂ exp(−k₂^ft), with k₁^f and k₂^f the extracted fading rate constants.

Acknowledgements

This work was supported by the CNRS, the Ministère de l'Enseignement Supérieur et de la Recherche, the ANR-11-BS07-011-01 BIOOPOM, and the LUMOMAT project supported by the Région des Pays de la Loire.

Notes and references

1. (a) Y. Hirshberg and E. Fisher, J. Chem. Phys., 1955, 23, 1723; (b) H. Bouas-Laurent, Photochromism: Molecules and Systems, Elsevier, Amsterdam, 2003.
2. C. Bechinger, S. Ferrere, A. Zaban, J. Sprague and B. A. Gregg, Nature, 1996, 383, 608.
3. B. L. Feringa, Molecular Switches, Wiley-VCH, New York, 2001.
4. (a) M. Irie, Chem. Rev., 2000, 100, 1685; (b) S. Kawata and Y. Kawata, Chem. Rev., 2000, 100, 1777; (c) J. A. Delaire and K. Nakatani, Chem. Rev., 2000, 100, 1817; (d) C. C. Corredor, Z.-L. Huang, K. D. Belfield, A. R. Morales and M. V. Bondar, Chem. Mater., 2007, 19, 5165; (e) Y. Yokoyama, Chem. Rev., 2000, 100, 1717.
5. (a) M. Natali and S. Giordani, Chem. Soc. Rev., 2012, 41, 4010; (b) V. I. Minkin, Chem. Rev., 2004, 104, 2751; (c) G. Berkovic, V. Krongauz and V. Weiss, Chem. Rev., 2000, 100, 1741.
6. (a) B. Schaudel, C. Guermeur, C. Sanchez, K. Nakatani and J. A. Delaire, J. Mater. Chem., 1997, 7, 61; (b) J. Biteau, F. Chaput and J.-P. Boilot, J. Phys. Chem., 1996, 100, 9024; (c) W. Yuan, L. Sun, H. Tang, Y. Wen, G. Jiang, W. Huand, L. Jiang, Y. Song, H. Tian and D. Zhu, Adv. Mater., 2005, 17, 156; (d) R. A. Kopelman, S. M. Snyder and N. L. Franck, J. Am. Chem. Soc., 2003, 125, 13684.
7. J. Harada, Y. Kawazoe and K. Ogawa, Chem. Commun., 2010, 46, 2593.
8. (a) B. Osterby, R. D. McKelvey and L. Hill, J. Chem. Educ., 1991, 68, 424; (b) D. Levy, Chem. Mater., 1997, 9, 2666; (c) C. Sanchez, B. Lebeau, F. Chaput and J.-P. Boilot, Adv. Mater, 2003, 15, 1969; (d) M.-Q. Zhu, L. Zhu, J. J. Han, J. K. Hurst and A. D. Q. Li, J. Am. Chem. Soc., 2006, 128, 4303; (e) I. Khale, O. Tröber, S. Trentsch, H. Richter, B. Grünler, S. Hemeltjen, M. Schlesinger, M. Mehring and S. Spange, J. Mater. Chem., 2011, 21, 5083.
9. (a) Polyoxometalate Chemistry: From Topology via Self-Assembly to Applications, ed. M. T. Pope and A. Müller, KLUWER, Dordrecht, Netherlands, 2001; (b) Polyoxometalate Chemistry for Nano-Composite Design, ed. T. Yamase and M. T. Pope, KLUWER, New York, 2002; (c) Polyoxometalate Molecular Science, ed. J. J. Borras-Almenar, E. Coronado, A. Müller and M. T. Pope, KLUWER, Dordrecht, Netherlands, NATO Science Series II, Vol. 98, 2003; (d) D.-L. Long, E. Burkholder and L. Cronin, Chem. Soc. Rev., 2007, 36, 105; (e) A. Dolbecq, E. Dumas, C. R. Mayer and P. Mialane, Chem. Rev., 2010, 110, 6009; (f) U. Kortz, A. Müller, J. V. Salgeren, J. Schnack, N. S. Dalal and M. Dressel, Coord. Chem. Rev., 2009, 253, 2315; (g) P. Gouzerh and A. Proust, Chem. Rev., 1998, 98, 77; (h) D.-L. Long, R. Tsunashima and L. Cronin, Angew. Chem., Int. Ed., 2010, 49, 1736.
10. (a) T. Yamase, Chem. Rev., 1998, 98, 307; (b) L. Yang, Z. Zhou, P. Ma, J. Wang and J. Niu, Cryst. Growth Des., 2013, 13, 2540; (c) Y. Wang, P. Ma and J. Niu, Dalton Trans., 2015, 44, 4679; (d) Y. Chen, G. Yu, F. Li, C. Xie and G. Tian, J. Mater. Chem. C, 2013, 1, 3842; (e) X. Luo and C. Yang, Phys. Chem. Chem. Phys., 2011, 13, 7892; (f) K. Hakouk, O. Oms, A. Dolbecq, H. El Moll, J. Marrot, M. Evain, F. Molton, C. Duboc, P. Deniard, S. Jobic, P. Mialane and R. Dessapt, Inorg. Chem., 2013, 52, 555; (g) R. Dessapt, M. Collet, V. Coué, M. Bujoli-Doeuff, S. Jobic, C. Lee and M.-H. Whangbo, Inorg. Chem., 2009, 48, 574.
11. (a) S. Liu, D. G. Kurth, H. Möhwald and D. Volkmer, Adv. Mater., 2002, 14, 225; (b) G. Gao, L. Xu, W. Wang, W. An, Y. Qiu, Z. Wang and E. Wang, J. Phys. Chem. B, 2005, 109, 8948; (c) L. Jin, Y. Fang, P. Hu, Y. Zhai, E. Wang and S. Dong, Chem. Commun., 2012, 48, 2101.
12. (a) L. Xu, E. Wang, Z. Li, D. G. Kurth, X. Du, H. Zhang and C. Qin, New J. Chem., 2002, 26, 782; (b) J.-D. Compain, P. Mialane, A. Dolbecq, J. Marrot, A. Proust, K. Nakatani, P. Yu and F. Sécheresse, Inorg. Chem., 2009, 48, 6222.
13. (a) L. Xu, H. Zhang, E. Wang, D. G. Kurth and Z. Li, J. Mater. Chem., 2002, 12, 654; (b) H. Li, W. Qi, W. Li, H. Sun, W. Bu and L. Wu, Adv. Mater., 2005, 17, 2688.
14. A. Parrot, G. Izzet, L.-M. Chamoreau, A. Proust, O. Oms, A. Dolbecq, K. Hakouk, H. El Bekkachi, P. Deniard, R. Dessapt and P. Mialane, Inorg. Chem., 2013, 52, 11156.
15. O. Oms, K. Hakouk, R. Dessapt, P. Deniard, S. Jobic, A. Dolbecq, T. Palacin, L. Nadjo, B. Keita, J. Marrot and P. Mialane, Chem. Commun., 2012, 48, 12103.
16. A. Saad, O. Oms, J. Marrot, A. Dolbecq, K. Hakouk, H. El Bekkachi, S. Jobic, P. Deniard, R. Dessapt, D. Garrot, K. Boukheddaden, R. Liu, G. Zhang, B. Keita and P. Mialane, J. Mater. Chem. C, 2014, 2, 4748.
17. (a) K. Hakouk, O. Oms, A. Dolbecq, J. Marrot, A. Saad, P. Mialane, H. El Bekkachi, S. Jobic, P. Deniard and R. Dessapt, J. Mater. Chem. C, 2014, 2, 1628; (b) J.-D. Compain, P. Deniard, R. Dessapt, A. Dolbecq, O. Oms, F. Sécheresse, J. Marrot and P. Mialane, Chem. Commun., 2010, 46, 7733; (c) J.-D. Compain, P. Mialane, J. Marrot, F. Sécheresse, W. Zhu, E. Oldfield and A. Dolbecq, Chem.–Eur. J., 2010, 16, 13441.
18. J. Li, J. Cluster Sci., 2002, 13, 137.
19. D. Hagrman, P. J. Hagrman and J. Zubieta, Angew. Chem., Int. Ed., 1999, 38, 3165.
20. R. Klajn, Chem. Soc. Rev., 2014, 43, 148.
21. P. Mialane, G. J. Zhang, I. M. Mbomekallé, P. Yu, J.-D. Compain, A. Dolbecq, J. Marrot, F. Sécheresse, B. Keita and L. Nadjo, Chem.–Eur. J., 2010, 16, 5572.
22. S. Benard and P. Yu, Chem. Commun., 2000, 65.
23. C. Sanchez, J. Livage, J. P. Launay and M. Fournier, J. Am. Chem. Soc., 1983, 105, 6817.
24. W. G. Klemperer, Inorg. Synth., 1990, 27, 74.
25. G. M. Sheldrick, SADABS; program for scaling and correction of area detector data, University of Göttingen, Germany, 1997.
26. R. Blessing, Acta Crystallogr., Sect. A: Found. Crystallogr., 1995, A51, 33.
27. G. M. Sheldrick, SHELX-TL Version 5.03, Software Package for the Crystal Structure Determination, Siemens Analytical X-ray Instrument Division, Madison, WI, USA, 1994.
28. P. Kubelka and F. Munk, Z. Tech. Phys., 1931, 12, 593.

---

Footnote

† Electronic supplementary information (ESI) available: Complementary Kubelka–Munk transformed reflectivity and absorption spectra, UV-vis absorption spectra in solution, photocoloration and fading kinetics parameters. CCDC 1414602–1414604. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra15860e

---