Introduction

Precise modulation of materials’ properties has long been a target pursued by materials scientists. In the field of crystalline materials, crystal engineering strategies to elicit specific structure based on multicomponent systems such as polymorphs, cocrystals, and solvates are now well established1,2,3,4,5,6,7,8. These approaches are effective in significant alteration of the crystal structure, and thus usually lead to notable changes in properties, which however, usually makes preparation of a material with a specific value of certain physical or chemical characteristics challenging. Solid solutions (sometimes referred to as mixed crystals or molecular alloys), on the other hand, enable access to variable stoichiometric ratios which—provided they form across a wide range of compositions—could be applied to elicit certain properties at will9,10,11. It has been shown (for examples, see refs. 12,13,14,15,16,17) that with solid solutions the molecular packing can be varied depending on the composition with a limited effect on the (averaged across the crystal) unit cell parameters. This implies that properties that depend directly on the molecular packing and interactions will be affected by the composition15,18. Previous studies have provided only a few cases of solid-state photochemical preparation of products obtained by reaction between the components19,20,21. Photomechanical crystalline materials with smart functionality have been widely studied due to their immense potential as light-to-mechanical energy converters22,23,24,25,26,27,28,29,30, and the modulation of their properties has become one of the key research challenges31. This was previously accomplished with controlling the bending direction by using cocrystals32, and similarly, by capitalizing on polymorphism33,34,35,36. Since many of the photomechanical effects in the solid state are based on photoreactions37,38, solid solutions could provide an alternative, and possibly, advantageous approach to control the response from smart crystals. Motivated by these guiding principles and aiming to explore the effectiveness of the solid solution strategy, here we report a set of solutions of two anthracene derivatives, 9-anthraldehyde (9AA) and 9-methylanthracene (9MA) (Fig. 1a)39,40. The different fluorescence of the two crystals and the excellent mechanical flexibility of the crystals of 9AA provide basis for concomitant modulation of the emissive and mechanical properties. We address the question of exerting a control over a collection of properties that depend on the electronic and geometric structure—emissive properties, reactivity, elasticity, and photomechanical response. Within a more general context, this study sets the basis for the design and preparation of smart crystalline materials with quantitively pre-determined properties by using solid solutions instead of pure organic crystals, and provides an alternative access to tunable smart crystalline materials.

Fig. 1: Chemical structures, crystal structures, and composition of the pure components and the solid solutions.
figure 1

a Chemical structures and approximate size of 9AA and 9MA defined by van der Waals surfaces, including the hydrogen atoms. b Correlation between the feed molar ratios (x) and actual doping ratios (y) of the two components in the mixed crystals, the latter determined based on the 1H NMR spectra (solvent: CDCl3). c PXRD patterns of products prepared by cooling of the crystallization solutions (the coefficients in the formulas represent the actual compositions). d−g Molecular packing, viewed along the a axis in the crystals 9MA (d), (9AA)0.5(9MA)0.5 (e), and along the b axis in the crystals 9AA (f), and (9AA)0.9(9MA)0.1 (g).

Full size image

Results and discussion

The mixed crystals were prepared by cooling mixed ethanol/water (2 : 1 v/v) solutions, as shown in Supplementary Fig. 1. Flaky crystals (similar to the crystals of pure 9MA) were obtained when the feed molar ratio (x) of 9AA was x ≤ 50%, and rod-like crystals (similar to the crystals of pure 9AA) were obtained at x ≥ 70%. At x = 60%, two types of crystals with different morphology and fluorescence were obtained concomitantly. We note that the feed ratio (x) is not always consistent with the actual doping ratio (y), and the actual ratio y was determined from the nuclear magnetic resonance (1H NMR) spectra (Supplementary Fig. 2, several single crystals were dissolved in deuterated chloroform to record the spectra), where the two types of crystals obtained at x = 60% were manually separated by using the difference in the color of their fluorescence (yellow and orange). The value of y was calculated based on the characteristic peaks of 9MA and 9AA (for 9MA, δ(CH3) = 3.0 ppm; for 9AA, δ(aldehyde) = 11.4 ppm), and the results are shown in Fig. 1b and Supplementary Table 1. Figure 1b shows that the actual ratio y of the mixed crystals that had 9MA and 9AA as the major component (the host), respectively, are linearly correlated with the feed ratios x. The actual ratio y for the two types of crystals obtained at feed ratio x = 60% are 51.3% and 80.0%, respectively. These values can be considered the doping limits for the mixed crystals with 9MA and 9AA as the host, respectively. We confirmed the phase composition by powder X-ray diffraction (PXRD) and differential scanning calorimetry (DSC)41. As shown in Fig. 1c, the characteristic peaks of the mixed crystals at y ≤ 51.3% and y ≥ 80.0% are consistent with those of 9MA crystals (peaks at 2θ = 9.6° and 15.0°) and 9AA crystals (10.1°), respectively. The DSC results indicate that the melting point decreases with increasing concentration of the guest (defined as the molecule in smaller amount) regardless of whether 9AA or 9MA acts as the host (Supplementary Fig. 3)42,43.

We determined the structures of the mixed crystals (9AA)0.5(9MA)0.5 and (9AA)0.9(9MA)0.1 (CCDC deposition numbers 2368690 and 2389727). As can be inspected from Fig. 1d–g and Supplementary Table 2, the crystals (9AA)0.5(9MA)0.5 and (9AA)0.9(9MA)0.1 have cell parameters and molecular packing that is very similar to those of 9MA and 9AA crystals, respectively. As expected, the overlap of the substituents appears as a static disorder that was resolved in the refinement (Supplementary Fig. 4, 5 and Tables 3, 4); the contribution of 9MA in (9AA)0.5(9MA)0.5 and (9AA)0.9(9MA)0.1 was refined to 0.5 and 0.1, respectively, values that are consistent with the 1H NMR results. The 9MA crystals prepared in this study have head-to-tail stacked molecules, and the 9AA crystals have head-to-head stacked molecules, which leads to head-to-tail and head-to-head stacking in (9AA)0.5(9MA)0.5 and (9AA)0.9(9MA)0.1, respectively.

As shown in Fig. 2a–d, the emission of mixed crystals is different from those of the pure components. 9MA emits green fluorescence (λem = 520 nm, ΦF = 19.5%, τ = 62.2 ns), and 9AA emits yellow fluorescence (λem = 557 nm, ΦF = 4.3%, τ = 7.0 ns). Taking two of the mixed crystals as example, the emission of (9AA)0.5(9MA)0.5 is yellow (λem = 568 nm, ΦF = 0.8%, τ = 3.3 ns), while that of (9AA)0.9(9MA)0.1 is orange (λem = 582 nm, ΦF = 2.3%, τ = 5.3 ns, Supplementary Fig. 6). Thus, for a 9AA crystal doped with 9MA, such as (9AA)0.9(9MA)0.1, there is red-shift in both the absorption and emission spectra compared to the pure compound (Fig. 2e). This shift was rationalized by time-dependent density functional theory (TDDFT) calculations on binary molecule pairs 9MA-9MA, 9AA-9AA and 9AA-9MA used as models and the B3LYP/def2-TZVP method44,45. Molecular pairs were extracted from the respective crystal structures and only the hydrogen atom positions were optimized (Supplementary Fig. 7). As shown in Supplementary Tables 5–7 and Supplementary Data 1–3, the S0 → S1 transition of the molecular pair 9AA-9MA (2.46 eV) has a lower energy gap compared to 9MA-9MA (2.97 eV) and 9AA-9AA (2.56 eV), a result which is consistent with the observed redshift. Supplementary Fig. 8 shows the fluorescence spectra of solid solutions with different compositions. The results show a gradual red-shift of the maximum emission wavelength compared to the monomer as the concentration of the guest (defined as the molecule in smaller amount) increases. The relationship between the composition and maximum emission wavelength indicates that the preparation of solid solutions is an effective strategy for tuning of the emission maximum.

Fig. 2: Emissive and mechanical properties.
figure 2

a−e Images of 9MA (a), (9AA)0.5(9MA)0.5 (b), (9AA)0.9(9MA)0.1 (c), and 9AA (d) crystals recorded under visible (top) and 405 nm (bottom) light. e Fluorescence emission spectra (solid lines) and diffuse reflectance spectra (dashed lines) of 9MA, 9AA, (9AA)0.5(9MA)0.5, and (9AA)0.9(9MA)0.1. f Variations in the elastic modulus and hardness with compositions. g, h Elastic bending of 9AA (g) and (9AA)0.9(9MA)0.1 (h).

Full size image

The solid solution strategy to change the composition of crystals could provide means to modulate the mechanical properties16,46. The most developed facets were indexed by the preferred orientation in the PXRD patterns (Supplementary Fig. 9)47. In agreement with previous studies, the main facet of 9MA and (9AA)0.5(9MA)0.5 is (100), while for 9AA and (9AA)0.9(9MA)0.1 it is (001)39,48. The elastic modulus (E) and hardness (H) on these facets were obtained by nanoindentation (Supplementary Fig. 10, Supplementary Table 8). The variations of E and H with composition are shown in Fig. 2f. Regardless of whether 9AA or 9MA acts as the host, both the stiffness and hardness increase with increasing concentration of the guest (defined as the molecule in smaller amount). This observation corresponds with the solid-solution-related strengthening49,50, where the guest molecules act as point defects in the host lattice, causing local structural deformation, and making the crystal stiffer and harder. The similarity in molecular packing probably allows crystals of (9AA)0.9(9MA)0.1 to be mechanically flexible, similar to the 9AA crystals (Fig. 2g, h, Supplementary Movies 1, 2).

Given that the distance between the anthracene groups in 9MA is 3.859 Å and therefore less than the required distance (4.2 Å) for a photoreaction to occur (Supplementary Fig. 11)51,52, [4 + 4] solid-state photodimerization reaction is feasible. In (9AA)0.5(9MA)0.5, the (average) distance between the anthracene groups is reduced to 3.756 Å (Supplementary Fig. 11), and also satisfies the topochemical reaction condition. Accordingly, upon exposure of (9AA)0.5(9MA)0.5 powder to 405 nm radiation (300 mW cm−2), new peaks appeared in the PXRD pattern at 2θ = 10.1°, 11.8°, 13.4°, 23.5°, and 25.7°, indicating the generation of a new crystalline phase (Supplementary Fig. 12). Unlike the 9MA-9MA homodimer (head-to-tail) obtained after irradiation of 9MA crystals, in (9AA)0.5(9MA)0.5 the randomness of 9AA and 9MA molecules at the same lattice site implies that the heterodimer 9AA-9MA (head-to-tail) will also be formed (Fig. 3a) in addition to the homodimers 9AA-9AA and 9MA-9MA (head-to-tail). If we assume an equal probability for photodimerization of adjacent molecules in (9AA)0.5(9MA)0.5, then the theoretical content of the three binary molecule pairs is 9AA-9AA : 9AA-9MA : 9MA-9MA = 0.244 : 0.500 : 0.256 (Fig. 3b; for details of the calculation, see Supplementary Fig. 13). The outcome of the photodimerization was monitored by 1H NMR before and after irradiation (405 nm, 300 mW cm-2, 40 min). The maximum temperature of the crystals at this irradiation power remained below 35 °C (Supplementary Fig. 14) to avoid side reactions and sublimation. As shown in Fig. 3c, the characteristic peaks of the 9MA-9MA dimer at 3.91 ppm (corresponding to the hydrogen opposite the methyl group in the 9MA-9MA dimer), and the 9AA-9AA dimer at 10.0 ppm and 5.42 ppm (with 5.42 ppm representing the hydrogen opposite the aldehyde group in the 9AA-9AA dimer), were determined by reactions of the crystals of 9MA and 9AA, and it was concluded that the head-to-head or head-to-tail dimers do not affect the position of the characteristic peaks. After irradiation of the crystals of (9AA)0.5(9MA)0.5 at 405 nm, new peaks appeared at 4.00 ppm, 5.30 ppm, and 10.11 ppm, consistent with generation of the 9AA-9MA heterodimer. Among them, the areas of the peaks at 4.00 ppm and 5.30 ppm remained essentially the same for different irradiation times (Supplementary Table 9). Therefore, the peaks at 4.00 ppm and 5.30 ppm were identified as characteristic peaks from the hydrogen opposite the methyl group and the hydrogen opposite the aldehyde group in the heterodimer, respectively.

Fig. 3: Reactivity of the pure components and the mixed crystals.
figure 3

a Schematic diagram of the possibility of existence of different molecular pairs in the solution solutions. b Models of the three possible photodimerization products. c 1H NMR spectra of 9MA, 9AA, and (9AA)0.5(9MA)0.5 before and after 405 nm irradiation (400 MHz, CDCl3). d Changes in composition during 405 nm irradiation of (9AA)0.5(9MA)0.5 crystals. e Content of each dimer in the solid solutions after 405 nm irradiation with a fixed duration of 40 min. f−i Photomechanical response and change in the fluorescence of (9AA)0.5(9MA)0.5 observed during 405 nm irradiation. j Fluorescence spectra of (9AA)0.5(9MA)0.5 after 405 nm irradiation for different time durations (the inset shows images of the crystal fluorescence). k, l Images of crystals of 9MA (k) and (9AA)0.5(9MA)0.5 (l) exposed to white light.

Full size image

The contribution to the total composition of the five possible reactants and products (9AA, 9MA, the 9AA-9AA dimer, the 9AA-9MA dimer and the 9MA-9MA dimer) was obtained and normalized based on the respective 1H NMR peaks (Supplementary Table 9), and the change in the composition during irradiation was monitored (one dimer was counted as two monomers). As detailed in Supplementary Table 10 and Fig. 3d, after irradiation of the (9AA)0.5(9MA)0.5 crystals, the content of the 9AA-9MA heterodimer is the highest (0.458) after 40 min irradiation, which is close to the highest theoretical content of the 9AA-9MA pairs (0.500). The content of the 9AA-9AA dimer (0.287) slightly exceeds the theoretically calculated content of the 9AA-9AA pair (0.244) in the absence of total (100%) conversion. This implies that the assumption that the reactivity of all molecule-pair combinations is identical is idealized, and instead, different combinations of monomers have different propensity to react. Specifically, the reaction between two 9AA monomers is more favorable than those between the other combinations. Assuming that there is a molecular stacking in (9AA)0.5(9MA)0.5 as shown in Supplementary Fig. 15, the probability of photodimerization is higher for 9AA-9AA and lower for 9MA-9MA. This conclusion is also supported by the experimental results that show that the content of the 9AA-9AA dimer (0.287) is significantly higher than that of the 9MA-9MA dimer (0.216) at similar theoretical content of the binary monomer pairs. These results demonstrate the feasibility and effectiveness of the synthesis of the heterodimer via the solid solution strategy.

Changes in the content of homo/hetero dimers are inevitable with changes in the doping ratio of 9AA. Following the aforementioned procedures, we also tested and calculated the content of each dimer for solid solutions with different doping ratios under the same irradiation conditions (405 nm, 300 mW cm−2, 40 min). As shown in Fig. 3e and Supplementary Table 11, the content of the 9AA-9AA and 9AA-9MA dimers gradually increases and that of the 9MA-9MA dimer gradually decreases with increasing 9AA doping ratio. During the irradiation and concurrent with the photodimerization reaction, the (9AA)0.5(9MA)0.5 crystals exhibit photomechanical effects, such as cracking and jumping (Fig. 3f–i, Supplementary Movie 3), as well as notable fluorescence changes and intensity enhancement of multiple peaks in the emission spectrum (391 nm, 415 nm, 439 nm and 466 nm, Fig. 3j). The peaks in the 400−500 nm region are typically attributed to unreacted isolated monomers rather than to anthracene dimers53,54. To investigate the source of the multiple emission peaks after irradiation, the fluorescence spectra of 9MA and 9AA in DMSO solution (concentration: 1 × 10–5 mol L–1) were recorded (Supplementary Fig. 16). The multiple emission peaks observed for (9AA)0.5(9MA)0.5 powder in the region 390−470 nm after irradiation with a 405 nm light for 30 s resemble those of 9MA in DMSO solution. This result suggests that this emission is due to unreacted molecules of the monomer. The quantum yields of 9MA and 9AA determined in DMSO (concentration: 1 × 10–5 mol L–1) showed that 9MA has a significantly higher quantum yield (78.8%) relative to 9AA (1.9%). This difference in emission intensity could account for the observation of the emission of only 9MA in the irradiated powder, where the emission from 9AA is probably too low. The change in overall fluorescence is illustrated by the shift across the CIE 1931 chromaticity diagram, where the coordinates of the total emission change from yellow to white to blue area of the chart (Supplementary Fig. 17).

Photomechanical response triggered by visible light is generally considered effective in maximizing the utilization of light energy. Triggering photomechanical response in the visible light range has been demonstrated through molecular design and cocrystals55,56,57. We found that contrary to 9MA crystals, which do not show mechanical response when exposed to white light, crystals of (9AA)0.5(9MA)0.5 are responsive to white light (Fig. 3k, l, Supplementary Movie 4, 5, the spectrum of the white light is shown in Supplementary Fig. 18). This is attributed to the redshift in the absorption spectrum of (9AA)0.5(9MA)0.5 compared to 9MA crystals as well as to smaller photoreactive group distances. Within a more general context, the results favor the photochemistry of solid solutions as a feasible strategy for realization of photomechanical response to visible light.

In contrast to the decrease in photoreactive group distance in (9AA)0.5(9MA)0.5, the distance between the photoreactive groups increases from 4.081 Å in 9AA to 4.207 Å in (9AA)0.9(9MA)0.1 (Fig. 4a). For photomechanical crystals based on the [4 + 4] photodimerization reaction, this further affects the speed and extent of the photomechanical response. Similarly, we calculated the theoretical ratio of binary molecule pairs with potential photodimerization in (9AA)0.9(9MA)0.1 as 9AA-9AA : 9AA-9MA : 9MA-9MA = 0.780 : 0.207 : 0.013 (Supplementary Fig. 19, head-to-head stacking). The 1H NMR spectra of (9AA)0.9(9MA)0.1 irradiated for different times were recorded, and the content of each component was calculated under the same irradiation conditions (405 nm, 300 mW cm−2, 40 min, Supplementary Table 12, 13, Fig. 4b−c). (9AA)0.9(9MA)0.1 showed significant new peaks at 10.0 ppm and 5.42 ppm after irradiation (generation of the 9AA-9AA dimer), and low-intensity peaks at 4.00 ppm and 5.30 ppm with comparable peak areas (generation of the 9AA-9MA dimer, Fig. 4b). The characteristic peak of the 9MA-9MA dimer (3.91 ppm) was not found due to the low amount of 9MA. The effect of 9MA doping ratio on the total conversion rate of photodimerization was also studied. As shown in Fig. 4d and Supplementary Table 14, the doping of 9MA results in rapid decrease in the total photodimerization conversion rate from 36.3% to approximately 8%. The results confirm that the solid-state photoreactions and the related mechanical response can be “tuned” by resorting to solid solutions.

Fig. 4: Photomechanical effect and its relation to the composition.
figure 4

a Photodimerization of 9AA and molecular stacking in crystals of 9AA and (9AA)0.9(9MA)0.1. b 1H NMR spectra of 9AA, 9MA and (9AA)0.9(9MA)0.1 recorded after 405 nm irradiation (400 MHz, CDCl3). c Changes in the composition of (9AA)0.9(9MA)0.1 during 405 nm irradiation. d Changes in total photodimerization conversion of mixed crystals irradiated for 40 min with varying 9MA doping ratio. e Definition of the parameters related to photomechanical bending of the crystal. f Comparison of photomechanical bending of 9AA and (9AA)0.9(9MA)0.1 crystals of similar size under the same irradiation conditions (300 mW cm–2, 20 s).

Full size image

Based on the successful modulation of the photodimerization reaction rate, we further verified the effect on the photomechanical response. 9AA affords slender, elongated crystals that bend towards the light source when they are exposed to 405 nm radiation39. Characteristic parameters used to quantify the photomechanical bending are the distance (d) and angle (φ) of deflection, among other parameters (Fig. 4e)58. The photomechanical bending is not only related to the intrinsic structural changes due to the solid-state photoreaction, but also to the crystal size, light intensity, and mechanical properties of the crystal such as its stiffness. Crystals of 9AA (length × width × thickness = 4.94 mm × 104.5 μm × 45.3 μm) and (9AA)0.9(9MA)0.1 (4.78 mm × 127.3 μm × 40.6 μm) were selected, exposed to 405 nm irradiation at the same power for 20 s, and observed for any bending. As shown in Fig. 4f and Supplementary Movie 6, 7, after 20 s of irradiation, a crystal of (9AA)0.9(9MA)0.1 bent less (d = 87.9 μm, φ = 1.4°) relative to a crystal of pure 9AA (d = 548.7 μm, φ = 6.8°). This outcome is attributed to the reduction in speed and extent of solid-state photoreaction, and increase in the elastic modulus. These results demonstrate the effectiveness of the solid solution approach to modulate the crystal bending, an aspect that to our knowledge has not been explored with photomechanical crystals prior to this work.

Having the compatibility of 9AA and 9MA and their propensity to form homogeneous solid solutions established, we set out to prepare spatially heterogenous structures where the composition of the crystal is different in different segments59,60,61,62,63,64,65. We prepared flexible crystalline heterostructures by using crystallization on the surface of the crystals (Fig. 5a)66. To that end, some of the pre-prepared centimeter-size crystals of 9AA were immersed in a saturated ethanol-water solution (50 °C) of 9AA and 9MA in a molar ratio 0.8 : 0.2, and mixed crystal phases were allowed to grow on the surface of the immersed section by cooling. This resulted in surface-doped heterostructures where one side of the crystal has uniform composition while the other one has two components (Fig. 5b and Supplementary Fig. 20). By breaking off the surface-doped section (Fig. 5c), the internal yellow-emissive crystalline phase was revealed, which proves the formation of the heterostructure. Visualization of the surface with SEM showed that the new crystal has grown on the surface of the original crystal (Fig. 5d). Furthermore, the fluorescence emission recorded from the three sites of a single heterostructure (Fig. 5e) showed that the emission of sections 1 and 3 corresponds essentially to that of 9AA and (9AA)0.9(9MA)0.1. Importantly, we needed to determine whether the heterostructure preserved the mechanical flexibility. As shown in Fig. 5f, g and Supplementary Movie 8, three-point bending experiments confirmed that it remains undamaged with a maximum elastic strain \(\varepsilon\) = 3.1% (Supplementary Fig. 21), which is comparable to the mechanical flexibility of many organic single crystals67,68,69,70,71,72,73,74. The preparation of flexible organic heterostructures is relevant to the application of optoelectronic organic materials in flexible devices, and the method described here provides a replicable strategy to that end.

Fig. 5: Preparation, emission and mechanical flexibility of heterostructures.
figure 5

a Preparation of the crystalline heterostructures. b Fluorescent images of 9AA, (9AA)0.9(9MA)0.1 and a heterostructure. c Removal of the surface-doped segment by fracturing. d SEM image of the heterostructure binding site (the dashed line depicts the boundary of the original crystal). e Fluorescence spectra of different sites of the heterostructure. f Schematic diagram of three-point bending. g Elastic bending of the heterostructure.

Full size image

We further demonstrate the active optical waveguide of this heterostructure (Fig. 6a–d, the schematic diagram of the optical waveguide experiment is shown in Supplementary Fig. 22), and it should be noted that the emission intensities of 9AA and (9AA)0.9(9MA)0.1 are gradually weakened due to the occurrence of solid-state photodimerization during irradiation (Supplementary Figs. 23 and 24), which is actually unfavorable for the active optical waveguide. Therefore, the tip spectral data need to be monitored quickly after the excitation of the 405 nm point light source. As in Fig. 6d, the spectral intensity collected at the tip of the heterostructure decreases faster after the excitation position crosses the binding site, which suggests that defects may still exist inside the binding. The test results show that the heterostructure has the ability of active optical waveguiding. The alteration of fluorescence by doping paves the way for the application of this system in encrypted paper/ink. As is shown in Fig. 6e, the ethanol suspension of 9AA powder was coated on a filter paper and dried to obtain encrypted paper. When writing on paper using dichloromethane (DCM) dissolved with 9MA as ink, the handwriting was not easily detected under ambient light, while the orange fluorescent “2025” pattern could be found under 405 nm light (Fig. 6f). The maximum of the fluorescence emission of the non-inked section (553 nm) is closer to that of 9AA, while the inked section (585 nm) is closer to that of (9AA)0.9(9MA)0.1 (Supplementary Fig. 25). We hypothesize that the dichloromethane dissolved part of 9AA during the writing process, and a solid solution with 9MA as the guest was formed after evaporation of the solvent.

Fig. 6: Active optical waveguide and information encryption design.
figure 6

a Illustration of the active optical waveguides under 405 nm excitation. Active optical waveguide of (b) 9AA, (c) (9AA)0.9(9MA)0.1, (d) heterostructure crystal and the spectra collected from the tip with increasing transmission distance. e Schematic of the preparation and use of encryption paper and ink. f Images of encryption paper before and after writing under ambient light and 405 nm light.

Full size image

With two substituted photoreactive anthracenes, this work introduces the well-known solid solution strategy to the field of photomechanical crystals. The solid solutions with 9MA as the host and 9AA as the host were comprehensively and systematically studied. The solid solution strategy provides a robust modulation of both fluorescence and mechanical properties. The statistically disordered nature of the solid solutions combined with the stereoselectivity of the solid-state reaction enables the preparation of heterodimers, and the strategy is designable and predictable. The content of homo/heterodimers can be modulated through the doping ratio of the solid solution. (9AA)0.5(9MA)0.5 was also shown to exhibit photomechanical response under white light irradiation, demonstrating the potential of the solid solution strategy to fully utilize light energy. The effect of doping on molecular stacking and on the rate and extent of solid-state photoreaction was utilized to successfully achieve the modulation of photomechanical bending in (9AA)0.9(9MA)0.1. The mechanical flexibility of 9AA and its low lattice mismatch with (9AA)0.9(9MA)0.1 combined with the surface doping strategy successfully led to the preparation of macro flexible heterostructures. The active optical waveguiding capability of heterostructures and the application of the 9AA/9MA solid solution system for information encryption were investigated. The results are significant for the development of solid-state photoreactions, photomechanical crystals as well as flexible organic heterostructures.

Methods

Mixed crystals preparation

Different molar ratios of 9AA and 9MA powders were milled to obtain physically mixed powders. Then 100 mg of the powders were completely dissolved in 15 mL of ethanol-water mixture (2 : 1 v/v) at 50 °C, and the solid solutions (mixed crystals) were prepared by cooling the solution at a rate 2 °C h−1 to 10 °C.

Characterization

The diffraction data were collected on a D8 Venture (Bruker, Germany) single crystal diffractometer. The Olex2 suite was used for structure solution and refinement75. The powder X-ray diffraction data were collected by using a MiniFlex600 (Rigaku, Japan) powder X-ray diffractometer with CuKα radiation (λ = 1.5417 Å) at 40 kV and 100 mA over 2θ = 2–35° at an interval of 0.02° and scanning speed of 8° min−1. The 1H NMR spectra were recorded on an Ascend 400 (Bruker, Germany) spectrometer. Complete solubility of the sample was confirmed before the measurement. The differential scanning calorimetry (DSC) measurements were carried out using a DSC 1 (Mettler-Toledo, Switzerland) at a heating rate of 10 °C min−1 under flow of nitrogen. Fluorescence spectra, lifetimes and quantum yields were measured on a FLS1000 steady-state/transient fluorescence spectrometer (Edinburgh, England).

Nanoindentation

The Bruker HYSITRON TI 980 was used to perform nanoindentation. The indentations were performed in a force-control mode, with 0.25 mN s−1 loading and unloading rate, and a peak load of 5 mN held for 10 s.

Optical waveguiding

A 405 nm-laser was used as a light source, and the fluorescence emission after excitation was recorded by using a USB4000-FL spectrometer (Ocean Optics).

Preparation of encrypted paper/ink

In total, 50 mg 9AA powder was suspended into 0.5 mL ethanol, coated on the filter paper (5.0 cm × 3.0 cm), and dried to prepare the encryption paper. 19.2 mg 9MA was dissolved in 1 mL DCM and used as ink for writing.

Computational methods

The time-dependent density functional theory (TDDFT) computations were performed at a B3LYP/def2TZVP level, by using the Gaussian 09 package76. The calculations were carried out in gas phase, and the positions of the non-hydrogen atoms (C, O) were fixed during the structure optimization. The TDDFT calculations were performed on the optimized structure.