ARTICLE https://doi.org/10.1038/s41467-019-11885-4 OPEN An all-photonic full color RGB system based on molecular photoswitches Gaowa Naren1, Chien-Wei Hsu1, Shiming Li1, Masakazu Morimoto2, Sicheng Tang 3, Jordi Hernando4, Gonzalo Guirado 4, Masahiro Irie2, Françisco M. Raymo3, Henrik Sundén 1 & Joakim Andréasson 1 1234567890():,; On-command changes in the emission color of functional materials is a sought-after property in many contexts. Of particular interest are systems using light as the external trigger to induce the color changes. Here we report on a tri-component cocktail consisting of a fluorescent donor molecule and two photochromic acceptor molecules encapsulated in polymer micelles and we show that the color of the emitted fluorescence can be continuously changed from blue-to-green and from blue-to-red upon selective light-induced isomerization of the photochromic acceptors to the fluorescent forms. Interestingly, isomerization of both acceptors to different degrees allows for the generation of all emission colors within the red- green-blue (RGB) color system. The function relies on orthogonally controlled FRET reactions between the blue emitting donor and the green and red emitting acceptors, respectively. 1 Chemistry and Chemical Engineering, Chemistry and Biochemistry, Chalmers University of Technology, 41296 Göteborg, Sweden. 2 Department of Chemistry and Research Center for Smart Molecules, Rikkyo University Nishi-Ikebukuro 3-34-1, Toshima-ku, Tokyo 171-8501, Japan. 3 Laboratory for Molecular Photonics, Department of Chemistry, University of Miami, 1301 Memorial Drive, Coral Gables, FL 33146-0431, USA. 4 Departament de Química, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Spain. Correspondence and requests for materials should be addressed to J.A. (email: a-son@chalmers.se) NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 S timuli-responsive multicolor luminescent materials attract display absorption only at wavelengths shorter than 450 nm and extensive attention due to their potential applications in areas show no or only minute overlap with the blue per emission such as optical memory systems, bioimaging, sensors, white- (Supplementary Fig. 2). UV light exposure isomerizes the DAEs light generation, encryption/decryption protocols, and anti- to the closed isomers, displaying absorption also in the visible counterfeiting1–9. Additive mixing, that is, changes in the overall region of the spectrum. The absorption of both DAEg(c) and emission color of a sample achieved by mixing two or more indi- DAEr(c) overlaps substantially with the blue emission from per, vidual fluorophores in different ratios has been successfully implying that they can act as quenchers of the per emission in applied10–14, whereas other approaches imply exposing the sample FRET processes. The FRET processes would not only imply to physical or mechanical stimuli such as acid/base, ions, solvent quenching of the blue emission from per, but also sensitization of vapor, and grinding/smearing15–26. The major downside with all the fluorescence from the respective DAE(c) isomer: Green the above mentioned efforts is that they require physical access to emission from DAEg(c) and red emission from DAEr(c). Thus, the sample. Thus, closed systems are excluded, as is remote three extreme situations can be identified: (i) No UV-induced operation. Using light instead as the external stimulus eliminates the isomerization implies both DAE derivatives in the open isomeric need for physical access. It is also waste-free, non-invasive, and can form (DAEg(o) and DAEr(o)). Per is the sole fluorophore, and be delivered with extremely high spatiotemporal precision (where blue fluorescence is emitted. (ii) UV-induced isomerization of and when). Several, at least partly, photonically driven systems have only DAEg(o) to yield DAEg(c) and DAEr(o). FRET from per to been devised to display a large range of attainable changes in the DAEg(c) results in green emission from DAEg(c). (iii) Isomer- emission color27–49, often by the inclusion of photochromic ization of only DAEr(o) to yield DAEg(o) and DAEr(c). FRET molecules (molecular photoswitches). Most appealing would be an from per to DAEr(c) results in red emission from DAEr(c). In all all-photonic system that allows for full-color reproduction, that is, other intermediate situations (partial isomerization of one or both all colors in the visible spectrum can be generated from one and the DAE:s and partial FRET reactions), the overall emission will be a same sample. This communication describes such a system. The blend between red (R), green (G), and blue (B), implying that full function relies on orthogonally controlled FRET (Förster Resonance color reproduction in the RGB system would be possible. Energy Transfer) processes between a blue emitter (perylene) and Key to the overall function described above is selective, or two diarylethene (DAE) photoswitches emitting green and red light, orthogonal, isomerization of the open to the closed forms, respectively. Selective isomerization of the two photoswitches results DAEg(o) → DAEg(c) and DAEr(o) → DAEr(c). As indicated in in blue-to-green and blue-to-red emission color changes in a color- Fig. 1, we have selected 242 and 415 nm for the respective correlated3 fashion, whereas isomerization of both photoswitches reactions. From Supplementary Fig. 2, it is seen that DAEg(o) allows for the generation of all emission colors within the red- shows no detectable absorption at 415 nm, so that this green-blue (RGB) color system. wavelength is expected to trigger exclusively the DAEr(o) → DAEr(c) closing reaction. Although DAEg(o) is the stronger absorber at 242 nm, light of this wavelength is also absorbed by Results DAEr(o). However, as the isomerization quantum yield for the System design. Figure 1 shows the structures and the iso- DAEg(o) → DAEg(c) reaction is substantially higher than for merization scheme for perylene (per) and the two photochromic DAEr(o) → DAEr(c) (0.50 vs. 0.0005 in acetonitrile), 242 nm DAE derivatives50. exposure is expected to be virtually specific for DAEg(o) → The absorption spectra of all species are shown in Supple- DAEg(c). The dramatic difference in the isomerization mentary Fig. 2. We refer herein to the two DAE derivatives as quantum yields is ascribed to the formation of an intramole- DAEg and DAEr, the g and the r referring to the emission color cular charge transfer (ICT)-state for DAEr(o) in polar of the closed isomers (green and red, respectively). Indicated are environments, outcompeting the isomerization reaction51. It also the FRET processes responsible for the color changes. The is acknowledged that such inefficient isomerization process open forms of the DAE photoswitches, DAEg(o) and DAEr(o), requires extensive light exposure. However, the extremely good photo-stability of the system (see below) implies that the long- F F F F term performance is still not affected. DAEg(o) F F F F DAEg(c) For FRET to occur efficiently, the distance between the donor F F λ = 242 nm F F and the acceptor should be shorter than the critical Förster radius Et (Vis) Et R0, the distance at which the efficiency of the FRET reaction is I S Et S I I S Et S I 50%. Based on a per fluorescence quantum yield of 0.85 in the O O O O O O O O polymer micelles and assuming κ2 = 2/3 (freely rotating chro- FRET (blue-to-green) mophores) R0-values of 45 and 51 Å were determined for the per- DAEg(c) and per-DAEr(c) FRET-pairs, respectively (see Table 1 per FRET (blue-to-red) DAEr(o) F F F F Table 1 Photophysical properties of the studied compounds. F F F F DAEr(c) λ = 415 nm Data determined in the polymer micelles, unless F F F F otherwise noted Et Et (Vis) S Et S S Et S S O O O O S S O O O O S Compound λmax, abs λmax, em ΦF Φisoa, b R0 (Å)c (nm)a (nm) Fig. 1 Structures and isomerization scheme of the compounds used in this Per 434 443 0.85 – – study. The shaded arrows indicate the FRET (Förster Resonance Energy DAEg(c) 435 504 0.23 0.50 45 Transfer) processes and the accompanying changes in the emission color. DAEr(c) 508 606 0.43 5 × 10−4 51 Please note that the photoinduced ring-opening reactions DAE(c) → DAE(o) aData in acetonitrile induced by visible light (Vis) are too inefficient to influence the isomeric bIsomerization quantum yield for the ring-closing reactions DAE(o) → DAE(c) cCritical Förster distance for the FRET process with per as the donor and DAE(c) as the acceptor distributions with the herein applied irradiation wavelengths 2 NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 ARTICLE 1.0 a Emission intensity (norm.) 0.8 0.6 41 2 nm 5n 24 m 0.4 0.2 0.0 b Emission intensity (a.u.) 0.8 Fig. 2 Schematic depiction of the micelle cocktail approach used in this 0.6 study. Curved arrows indicate the orthogonally controlled FRET processes, responsible for the emission color changes 0.4 for a compilation of relevant photophysical data). Constraining 0.2 the three chromophores by covalent means would indeed allow for donor–acceptor (D–A) distances shorter than these R0-values. 0.0 However, a covalent approach would imply that the three units 450 500 550 600 650 700 750 800 would have a fixed stoichiometric relation in the structure, and a Wavelength (nm) loss of flexibility in the overall approach. Moreover, the synthetic Fig. 3 Emission spectra in the polymer micelles. a Normalized emission procedures would be highly complex and tedious. Instead, we spectra of per (blue line), DAEg(c) (green line), and DAEr(c) (red line) in opted for a self-assembling cocktail strategy in which the the polymer micelles. b Emission spectra of the tri-component micelle monomeric chromophores are encapsulated in polymer micelles, cocktail before UV irradiation (blue line), after exposure to 242 nm light offering a much larger degree of flexibility as for the composition (4 nm spectral bandwidth, 10 min irradiation time, green line) and after of the system, which in turn allows a very convenient means of exposure to 415 nm light (4 nm spectral bandwidth, 2.5 h exposure time, changing both the average D–A distances and the stoichiometry. red line). All spectra in panel (b) were recorded upon excitation at 423 nm. The overall approach is schematically depicted in Fig. 2. The relative emission intensities determined by spectral integration in (b) In addition to the FRET processes indicated in Fig. 2, DAEg(c) are 1.3:4.4:3.7 (blue:green:red), that is, the system displays fairly small and DAEr(c) also constitute a FRET pair with R0 = 45 Å. This intensity variations accompanying the emission color changes observation, however, imposes no obstacles to the intended overall function of the system as long as DAEg can be isomerized with a high degree of selectivity. and DAEr(o) (1.1 μM). The blue spectrum was recorded after excitation of the sample at 423 nm. The spectrum is virtually Spectroscopic characterization. All three chromophores display identical to that for per alone, apart from a weak tail at the red poor solubility in water, but are dissolved in the presence of the end of the spectrum. This tail is ascribed to small amounts of the ST-7-8 copolymer52 (see Supplementary Fig. 1 for the structure, closed, fluorescent isomers of the DAE units present already at polymer 1b in ref. 52) in aqueous solution by encapsulation into the beginning of the experiment. Comparing the two per emission the hydrophobic interior of the polymer micelles, positioning intensities (Supplementary Fig. 3), it is seen that the intensity them at distances well suited for efficient FRET reactions to dropped by ca. 70% in the cocktail, tentatively ascribed to occur53. Given the bulk concentrations in our experiments quenching by electron transfer reactions. The redox data of all ([per] = 0.45 μM, [DAEg] = 3.7 μM, [DAEr] = 1.1 μM, and three compounds reveal that such reactions are thermodynami- [micelles] = 0.31 µM), each micelle contains on the average 1 per, cally favorable (see Supplementary Note 3). The dynamic nature 12 DAEg, and 4 DAEr molecules. This implies average nearest of the quenching was established by time-resolved single photon D–A distances compatible with efficient FRET reactions to occur, counting (SPC) experiments, where the fluorescence lifetime of which is also supported by the experimentally obtained FRET per decreased from 4.8 to 1.0 ns upon addition of the DAE open efficiencies of around 0.9 for both the per-DAEg(c) and the per- isomers. DAEr(c) bi-component cocktails (see Supplementary Note 4 As anticipated, exposing the sample to light at 242 nm triggers together with Supplementary Figs. 7 and 9). isomerization of mainly DAEg(o) to DAEg(c), and the Figure 3a shows the normalized emission spectra of per alone, concomitant FRET reaction from per to DAEg(c). This is DAEg(c) alone, and DAEr(c) alone in the polymer micelles. The apparent by comparing the green spectrum in Fig. 3b (λexc = fluorescence quantum yields in this environment were deter- 423 nm) to those of Fig. 3a: The overall emission is dominated by mined to be 0.85 for per, 0.23 for DAEg(c), and 0.43 for DAEr(c). DAEg(c) at the expense of per. Minor emission is also seen Please note that the quantum yields for the DAE derivatives are from DAEr(c), showing that the 242 nm light is not selective for dramatically higher compared to all other compounds from the DAEg(o) to 100%. If the cocktail sample is instead exposed to common families of photoswitches, typically displaying fluores- 415 nm light, the red spectrum is recorded upon excitation at cence quantum yields lower than 0.05 in solution. Photographs 423 nm. Apart from minor spectral features from per and of the cuvettes under UV irradiation as well as the CIE DAEg(c) at shorter wavelengths, it is clear that the main emitter coordinates of the emission from the respective compound are in this case is DAEr(c). FRET being the dominating quenching shown in Fig. 4. mechanism of per is established by comparing the excitation Figure 3b shows the emission spectra from the tri-component spectra with the corresponding absorption spectra (see Supple- cocktail, containing originally per (0.45 μM), DAEg(o) (3.7 μM), mentary Fig. 15 and Supplementary Note 5). The CIE NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 CIE 1931 a 520 0.8 540 560 0.6 500 d 580 c h m y k n 0.4 g 600 f j I b 620 i e 0.2 a 480 0.0 460 0.0 0.2 0.4 0.6 0.8 x Fig. 5 Full-color reproduction. CIE coordinates for the emission from the tri- component micelle cocktail after isomerization of both DAE derivatives to different extent. Please note that all underlying spectra were recorded upon excitation at one and the same wavelength (423 nm). Irradiation conditions: a no irradiation prior to excitation for emission readout; b 220 s. 242 nm; c 450 s. 242 nm; d 600 s. 242 nm; e 15 s. 415 nm; f 220 s. 242 nm + 30 s. 415 nm; g 450 s. 242 nm + 30 s. 415 nm; h 800 s. 242 nm + 30 s. b 520 415 nm; i 70 s. 415 nm; j 220 s. 242 nm + 120 s. 415 nm; k 450 s. 242 nm + 0.8 540 120 s. 415 nm; l 370 s. 415 nm; m 800 s. 242 nm + 120 s. 415 nm; n 2.5 h. 415 nm 560 0.6 500 580 component cocktails are collected in Supplementary Figs. 3–9. From here, it is seen that there are slight discrepancies in the y 0.4 600 620 quenching efficiencies estimated from steady-state and SPC data, the quenching efficiencies being typically higher from the steady- 0.2 state experiments. We ascribe this observation to a hetero- 480 geneous distribution of the chromophores within the micelles which results in a distribution of the nearest D–A distances. This 460 0.0 notion is strengthened by the multi-exponential decays observed 0.0 0.2 0.4 0.6 0.8 x for all situations. For the shortest D–A distances, implying a very efficient quenching, the per lifetime is simply too short to be seen Fig. 4 Emission colors and CIE coordinates. a Photos of the fluorophores in the overall decay which results in an underestimation of the alone in the micelles upon excitation at 365 nm. b Hollow circles: CIE actual quenching efficiency. coordinates for the emission from the fluorophores alone in the micelles: Having shown that the tri-component cocktail can generate per (0.136, 0.115), DAEg(c) (0.295, 0.596), and DAEr(c) (0.603, 0.385). red, green, and blue emission color, the ultimate test is to see if all Solid circles: CIE coordinates for the emission of the tri-component micelle colors within the RGB triangle can be generated by isomerization cocktail before UV irradiation (0.156, 0.139), after exposure to 242 nm light of both DAE derivatives. The cocktail was therefore exposed to a (0.323, 0.502) and after exposure to 415 nm light (0.541, 0.407). Please large number of combinations of 242 and 415 nm exposures, and note that all underlying spectra were recorded upon excitation at one and the resulting emission spectra were recorded upon excitation at the same wavelength (423 nm). The corresponding coordinates for the 423 nm. The corresponding CIE coordinates are shown in Fig. 5. sRGB color space are defined as 0.15, 0.06 (blue); 0.30, 0.60 (green); and It is very encouraging to note that the performance is as good as 0.64, 0.33 (red) anticipated—all colors within the RGB triangle can indeed be reproduced. Adding to the appeal is the fairly small variation in the overall emission intensity for red, green, and blue emission coordinates of the three emission spectra from the cocktail are obvious from Fig. 3b, as well as the acceptable absolute emission mapped in the CIE diagram in Fig. 4b. It is encouraging to note intensities, where the lowest observed value (blue light) how very close the cocktail CIE coordinates are to those of the corresponds to a fluorescence quantum yield of around 0.3. pure reference samples containing only one of the three Finally, the system was investigated with respect to photo- and fluorophores. We would like to emphasize that all emission thermal stability (data presented in Supplementary Figs. 10–14). spectra are recorded using one and the same excitation Generally, DAE derivatives are known to be very stable in both wavelength which clearly shows that the dominating reason for these respects, and this has been shown also for the sulfone the observed color changes is in fact the orthogonally controlled derivatives studied herein54–56. The classical aspect of photo- FRET reactions from per to the respective fluorescent DAE stability was examined by light-exposure for 160 min or more isomers. This is further strengthened by the results from SPC (365 or 405 nm, 0.7–1.0 mW cm−2) of the individual fluoro- experiments. As indicated above, the fluorescence lifetime of per phores encapsulated in the micelles. Per and DAEr showed no alone in the micelles is 4.8 ns. This number is decreased to 0.90 signs of decomposition as judged by the absorption and the ns and 0.51 ns in the tri-component cocktails containing mainly emission spectra, whereas DAEg experienced an 8% decrease in per-DAEg(c)-DAEr(o) and per-DAEg(o)-DAEr(c), that is, after the absorption and a 25% decrease in the emission intensity upon exposure to 242 and 415 nm, respectively. Comprehensive extensive light exposure. The reduced photostability of the latter steady-state and SPC data including also all possible bi- compound is tentatively ascribed to an iodo-induced heavy atom 4 NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 ARTICLE effect, increasing the yield of triplet formation which in turn salt solution of the reference calomel electrode is separated from the makes it more vulnerable to degrading reactions. Equivalent electrochemical solution by a salt bridge ended with a frit, which is made of a ceramic material, allowing ionic conduction between the two solutions and samples displayed excellent thermal stability as no significant avoiding appreciable contamination. Argon was allowed to flow under the spectral changes were observed in the dark for 20 h (after solution during the measurements. The concentration of electroactive substances isomerization of DAEg and DAEr to the closed isomer). was between 1 and 10 mM while the supporting electrolyte concentration We would also like to emphasize the good color stability of the was 0.1 M. tri-component cocktail. Direct or FRET sensitized excitations of the DAE photoswitches would indeed trigger isomerization Isomerization reactions. The light for selective isomerization of DAEg and DAEr in the polymer micelles at 242 nm and 415 nm was provided by the 450 W xenon reactions (e.g., DAE(c) → DAE(o)) with concomitant changes lamp in the SPEX Fluorolog-3 spectrofluorometer after passage through the in the overall emission spectra57. However, due to the very low monochromator (4 nm spectral bandwidth). The exposure times were 10 min and quantum yields of these processes for the investigated 2.5 h, respectively, at photon fluxes of 2.7 × 10−11 einstein s−1 and 2.7 × 10−9 ein- derivatives51,56, no significant changes in the emission spectra stein s−1 as determined by ferrioxalate actinometry. were observed upon exposure to 423 nm light for 65 min—a light dose corresponding to the collection of 100 emission spectra at Isomerization quantum yield determination. The isomerization quantum yields this excitation wavelength. for the reactions DAE(o) → DAE(c) were determined using UV light at 365 nm from an LED (Engin LZ1 10UV00, FWHM = 11 nm, ∼1 mW cm−2 at the sample). The absorbance changes (DAEg at 450 nm, DAEr at 550 nm) as a function of Discussion irradiation time were monitored and compared to those of the reference com- pounds under identical irradiation power/geometries and corrected for the molar We have described a self-assembled all-photonic system con- absorption coefficients at 365 nm. The reference compounds used for DAEg and sisting of perylene and two molecular photoswitches from the DAEr were diarylethene derivatives 158 and 856. DAE photochromic family where the emission color can be tuned to any color in the RGB system. This is a rare example of stimuli- Fluorescence quantum yield determination. For fluorescence quantum yield responsive full-color reproduction, and it is very interesting to determination of all compounds in the polymer micelles, relative measurements note that the system can be operated without the need for phy- using a SPEX Fluorolog-3 spectrofluorometer together with standard compounds (perylene in cyclohexane, ΦF = 0.94; coumarine 153 in deaerated ethanol, ΦF = sical contact/access. All other previously described examples rely 0.38; perylene red in CHCl3, ΦF = 0.96)59. Note that the scattering from the on external stimuli of either physical or mechanical nature, micelles makes it challenging to correctly determine the absorption from the implying that open systems must be used. Accumulation of fluorophores at the excitation wavelengths. Herein, we have estimated the con- chemical waste is another problem where chemical stimuli are tribution from the scattering to the absorption spectra by a linear function between being applied. Mechanochromic systems undeniably require two points in the spectra with zero absorption and zero scattering, respectively. This procedure, yet not exact, should give comparable results to a λ−4-approach to physical contact which is very unpractical, not to say impossible, the scattering. for many of the applications foreseen for multicolor fluorescence switching, e.g., bioimaging. Our all-photonic version can be Data availability operated remotely, and with unsurpassed spatiotemporal preci- All relevant data are available from the corresponding author. sion, that is, where and when to induce the desired changes in the emission color. Received: 25 January 2019 Accepted: 7 August 2019 Methods Sample preparation. All micelle experiments were performed in aqueous solution (mQ). The samples were prepared by mixing amphiphilic copolymer ST-7-8 with References the relevant combination of each individual cocktail in dichloromethane, and film 1. Andréasson, J. & Pischel, U. Molecules for security measures: from keypad formation was promoted by evaporation of the solvent in a vacuum chamber. The locks to advanced communication protocols. Chem. Soc. Rev. 47, 2266–2279 micelles were formed by rehydration of the film for 10 min. with 2 ml mQ-water (2018). followed by filtration with 0.25 µm cellulose acetate filter. The average molecular 2. Hou, X. S. et al. Tunable solid-state fluorescent materials for supramolecular weight of the individual micelles was determined to be 580 kDa based on static light encryption. Nat. Commun. 6, 6884 (2015). scattering measurements. With a total amount of 0.4 mg ST-7-8 used in each 3. Kim, D. & Park, S. Y. Multicolor fluorescence photoswitching: color-correlated preparation and a 10% loss in the filtration process, the total micelle concentration versus color-specific switching. Adv. Opt. Mater. 6, 1800678 (2018). was 0.31 µM. 4. Li, C. H. & Liu, S. Y. Polymeric assemblies and nanoparticles with stimuli- responsive fluorescence emission characteristics. Chem. Commun. 48, Spectroscopic measurements. Ground state absorption spectra were recorded on 3262–3278 (2012). a Cary 50 UV/vis spectrometer. Corrected fluorescence spectra were recorded on a 5. Praveen, V. K., Ranjith, C. & Armaroli, N. White-light-emitting SPEX Fluorolog-3 spectrofluorometer. Fluorescence lifetimes were measured using supramolecular gels. Angew. Chem. Int. Ed. 53, 365–368 (2014). a time correlated single photon counting (TC-SPC) setup. The excitation light, 6. Tian, Z. Y., Wu, W. W. & Li, A. D. Q. Photoswitchable fluorescent λexc = 405 nm, was provided at a repetition rate of 20 kHz by a 405 nm diode laser nanoparticles: preparation, properties and applications. Chemphyschem 10, (PicoQuant) and a MCP-PMT detector (10,000 counts in the top channel, 1024 2577–2591 (2009). channels). The emitted photons were collected at the magic angle (54.7°) at 7. Zhang, Y. L., Zhang, K. Q., Wang, J., Tian, Z. Y. & Li, A. D. Q. 440 nm. The measured fluorescence decays were fitted using the program FluoFit Photoswitchable fluorescent nanoparticles and their emerging applications. Pro v.4 (PicoQuant GmbH, Germany) after deconvolution of the data with the Nanoscale 7, 19342–19357 (2015). instrument response function (IRF) with FWHM∼90 ps. 8. Hildebrandt, N. et al. Energy transfer with semiconductor quantum dot bioconjugates: a versatile platform for biosensing, energy harvesting, and other developing applications. Chem. Rev. 117, 536–711 (2017). Electrochemical characterization. Solvent and supporting electrolyte: acetonitrile 9. Silvi, S., Baroncini, M., La Rosa, M. & Credi, A. Interfacing luminescent from SDS was used as received (water content <10 mg kg−1); tetrabutylammonium quantum dots with functional molecules for optical sensing applications. Top. tetrafluoroborate or tetrabutylammonium hexafluorophosphate from Sigma- Aldrich were used as a supporting electrolyte without further purification. Curr. Chem. 374, 27 (2016). A computer controlled VSP potentiostat (Versatile Modular Potentiostat) was 10. Hudson, Z. M., Lunn, D. J., Winnik, M. A. & Manners, I. Colour-tunable used to perform the cyclic voltammetry at low scan rates in the range 0.1–1.0 V s−1. fluorescent multiblock micelles. Nat. Commun. 5, 3372 (2014). An electrochemical conical cell equipped with a methanol jacket, which 11. Jiang, K. et al. Red, green, and blue luminescence by carbon dots: full-color makes it possible to set the temperature at 20 °C by means of a thermostat, was emission tuning and multicolor cellular imaging. Angew. Chem. Int. Ed. 54, used for the set-up of the three electrode system. For cyclic voltammetry 5360–5363 (2015). experiments, the working electrode was, in all cases, a glassy carbon disk of a 12. Kwon, J. E., Park, S. & Park, S. Y. Realizing molecular pixel system for diameter of 1.0 mm. It was polished using a 1 mm diamond paste. The counter full-color fluorescence reproduction: RGB-emitting molecular mixture electrode was a glassy carbon disk of 0.3 cm diameter. All potentials are reported free from energy transfer crosstalk. J. Am. Chem. Soc. 135, 11239–11246 vs. SCE isolated from the working electrode compartment by a salt bridge. The (2013). NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 13. Wang, R. B., Peng, J., Qiu, F., Yang, Y. L. & Xie, Z. Y. Simultaneous 39. Li, C. H., Zhang, Y. X., Hu, J. M., Cheng, J. J. & Liu, S. Y. Reversible three-state blue, green, and red emission from diblock copolymer micellar films: a switching of multicolor fluorescence emission by multiple stimuli modulated new approach to white-light emission. Chem. Commun. 6723–6725 FRET processes within thermoresponsive polymeric micelles. Angew. Chem. (2009). Int. Ed. 49, 5120–5124 (2010). 14. Xu, Z. et al. Supramolecular color-tunable photoluminescent materials based 40. Mo, S. Z. et al. Tunable mechanoresponsive self-assembly of an amide-linked on a chromophore cascade as security inks with dual encryption. Mater. dyad with dual sensitivity of photochromism and mechanochromism. Adv. Chem. Front 1, 1847–1852 (2017). Funct. Mater. 27, 1701210 (2017). 15. Feng, H. T. et al. Multicolor emissions by the synergism of intra/ 41. Nakagawa, T., Miyasaka, Y. & Yokoyama, Y. Photochromism of a spiro- intermolecular slipped pi-pi stackings of tetraphenylethylene-DiBODIPY functionalized diarylethene derivative: multi-colour fluorescence modulation conjugate. Chem. Mater. 27, 7812–7819 (2015). with a photon-quantitative photocyclization reactivity. Chem. Commun. 54, 16. Karpiuk, J., Karolak, E. & Nowacki, J. Tuneable white fluorescence from 3207–3210 (2018). intramolecular exciplexes. Phys. Chem. Chem. Phys. 12, 8804–8809 42. Ritchie, C. et al. Photochrome-doped organic films for photonic keypad locks (2010). and multi-state fluorescence. Phys. Chem. Chem. Phys. 19, 19984–19991 17. Kim, H. J., Whang, D. R., Gierschner, J., Lee, C. H. & Park, S. Y. High-contrast (2017). red-green-blue tricolor fluorescence switching in bicomponent molecular film. 43. Tian, W. G. et al. Phototunable full-color emission of cellulose-based dynamic Angew. Chem. Int. Ed. 54, 4330–4333 (2015). fluorescent materials. Adv. Funct. Mater. 28, 1703548 (2018). 18. Kim, J. H. et al. Highly fluorescent and color-tunable exciplex emission from 44. Wang, J. B., Kulago, A., Browne, W. R. & Feringa, B. Photoswitchable poly(N-vinylcarbazole) film containing nanostructured supramolecular intramolecular h-stacking of perylenebisimide. J. Am. Chem. Soc. 132, acceptors. Adv. Funct. Mater. 24, 2746–2753 (2014). 4191–4196 (2010). 19. Li, W. et al. Tunable RGB luminescence of a single molecule with high 45. Watanabe, K., Hayasaka, H., Miyashita, T., Ueda, K. & Akagi, K. Dynamic quantum yields through a rational design. J. Mater. Chem. C. 4, 1527–1532 control of full-colored emission and quenching of photoresponsive conjugated (2016). polymers by photostimuli. Adv. Funct. Mater. 25, 2794–2806 (2015). 20. Shiraishi, Y., Ichimura, C., Sumiya, S. & Hirai, T. Multicolor fluorescence of a 46. Wu, T., Zou, G., Hu, J. M. & Liu, S. Y. Fabrication of photoswitchable and styrylquinoline dye tuned by metal cations. Chem. Eur. J. 17, 8324–8332 thermotunable multicolor fluorescent hybrid silica nanoparticles coated with (2011). dye-labeled poly(N-isopropylacrylamide) brushes. Chem. Mater. 21, 21. Teng, M. J., Jia, X. R., Chen, X. F. & Wei, Y. A dipeptide-based multicolored- 3788–3798 (2009). switching luminescent solid material: when molecular assemblies meet 47. Wu, T. Q., Boyer, J. C., Barker, M., Wilson, D. & Branda, N. R. A “plug-and- mechanochemical reaction. Angew. Chem. Int. Ed. 51, 6398–6401 (2012). play” method to prepare water-soluble photoresponsive encapsulated 22. Wang, L. et al. The facile realization of RGB luminescence based on one upconverting nanoparticles containing hydrophobic molecular switches. yellow emissive four-coordinate organoboron material. Chem. Commun. 51, Chem. Mater. 25, 2495–2502 (2013). 7701–7704 (2015). 48. Zhu, L. L. et al. Unimolecular photoconversion of multicolor luminescence on 23. Yang, Y. J. et al. An organic white light-emitting fluorophore. J. Am. Chem. hierarchical self-assemblies. J. Am. Chem. Soc. 135, 5175–5182 (2013). Soc. 128, 14081–14092 (2006). 49. Zhu, L. Y. et al. Reversibly photoswitchable dual-color fluorescent 24. Yuan, C. X. et al. A pi-conjugated system with flexibility and rigidity that nanoparticles as new tools for live-cell imaging. J. Am. Chem. Soc. 129, shows environment-dependent RGB luminescence. J. Am. Chem. Soc. 135, 3524–3526 (2007). 8842–8845 (2013). 50. Irie, M. & Morimoto, M. Photoswitchable turn-on mode fluorescent 25. Zhang, Z. Y. et al. Excited-state conformational/electronic responses of saddle- diarylethene: strategies for controlling the switching response. Bull. Chem. Soc. shaped N,N′-disubstituted-dihydrodibenzo[a,c]phenazines: wide-tuning Jpn. 91, 237–250 (2018). emission from red to deep blue and white light combination. J. Am. Chem. 51. Morimoto, M. et al. A turn-on mode fluorescent diarylethene: Soc. 137, 8509–8520 (2015). solvatochromism of fluorescence. Dyes Pigments 153, 144–149 (2018). 26. Wang, J. W., Wang, N., Wu, G., Wang, S. N. & Li, X. Y. Multicolor emission 52. Tang, S., Donaphon, B., Levims, M. & Raymo, F. M. Structural implications on from non-conjugated polymers based on a single switchable boron the properties of self-assembling supramolecular hosts for fluorescent guests. chromophore. Angew. Chem. Int. Ed. 58, 3082–3086 (2019). Langmuir 32, 8676–8687 (2016). 27. Ai, Q. & Ahn, K. H. A photoswitchable diarylethene heterodimer for use as a 53. Yildiz, I. et al. Supramolecular strategies to construct biocompatible and multifunctional logic gate. RSC Adv. 6, 43000–43006 (2016). photoswitchable fluorescent assemblies. J. Am. Chem. Soc. 133, 871–879 (2011). 28. Bu, J. X., Watanabe, K., Hayasaka, H. & Akagi, K. Photochemically colour- 54. Fukaminato, T. et al. Single-molecule fluorescence photoswitching of a tuneable white fluorescence illuminants consisting of conjugated polymer diarylethene-perylenebisimide dyad: non-destructive fluorescence readout. J. nanospheres. Nat. Commun. 5, 3799 (2014). Am. Chem. Soc. 133, 4984–4990 (2011). 29. Bälter, M. et al. Emission color tuning and white-light generation based on 55. Irie, M., Fukaminato, T., Matsuda, K. & Kobatake, S. Photochromism of photochromic control of energy transfer reactions in polymer micelles. Chem. diarylethene molecules and crystals: memories, switches, and actuators. Chem. Sci. 7, 5867–5871 (2016). Rev. 114, 12174–12277 (2014). 30. Chen, J. et al. Design and synthesis of FRET-mediated multicolor and 56. Uno, K. et al. In situ preparation of highly fluorescent dyes upon photoswitchable fluorescent polymer nanoparticles with tunable emission photoirradiation. J. Am. Chem. Soc. 133, 13558–13564 (2011). properties. J. Phys. Chem. B 116, 4354–4362 (2012). 57. Diaz, S. A., Gillanders, F., Jares-Erijman, E. A. & Jovin, T. M. Photoswitchable 31. Chen, Q. et al. Multicolor tunable emission from organogels containing semiconductor nanocrystals with self-regulating photochromic Förster tetraphenylethene, perylenediimide, and spiropyran derivatives. Adv. Funct. resonance energy transfer acceptors. Nat. Commun. 6, 7036 (2015). Mater. 20, 3244–3251 (2010). 58. Takagi, Y. et al. Turn-on mode fluorescent diarylethenes: control of the 32. Diaz, S. A. et al. Water-soluble, thermostable, photomodulated color- cycloreversion quantum yield. Tetrahedron 73, 4918–4924 (2017). switching quantum dots. Chem. Eur. J. 23, 263–267 (2017). 59. Brouwer, A. M. Standards for photoluminescence quantum yield 33. Diaz, S. A., Giordano, L., Jovin, T. M. & Jares-Erijman, E. A. Modulation of a measurements in solution (IUPAC Technical Report). Pure Appl. Chem. 83, photoswitchable dual-color quantum dot containing a photochromic FRET 2213–2228 (2011). acceptor and an internal standard. Nano Lett. 12, 3537–3544 (2012). 34. Hsu, C. W., Sauvee, C., Sunden, H. & Andréasson, J. Writing and erasing multicolored information in diarylethene-based supramolecular gels. Chem. Acknowledgements Sci. 9, 8019–8023 (2018). J.A. acknowledges financial support from the Swedish Research Council VR (Grant # 35. Ishida, S., Fukaminato, T., Kim, S., Ogata, T. & Kurihara, S. Sequential red- 2016-0360), J.H. and G.G. from MINECO/FEDER through project CTQ2015-65439-R, green-blue (RGB) fluorescence color photoswitching in multicomponent and F.M.R. from The National Science Foundation (CHE-1505885). Mr. Mats Tiborn is photochromic fluorescent nanoparticles. Chem. Lett. 46, 170436 (2017). acknowledged for the photography related work. Open access funding provided by 36. Ishida, S. et al. Wavelength-selective and high-contrast multicolour Chalmers University of Technology. fluorescence photoswitching in a mixture of photochromic nanoparticles. Chem. Commun. 53, 8268–8271 (2017). 37. Kim, D., Kwon, J. E. & Park, S. Y. Fully reversible multistate fluorescence Author Contributions switching: organogel system consisting of luminescent cyanostilbene and turn- J.A. designed the study. G.N. performed all the spectroscopic experiments, in some on diarylethene. Adv. Funct. Mater. 28, 1706213 (2018). experiments assisted by C.-W. H.J.A. and G.N. interpreted all the spectroscopic data. 38. Kim, S., Yoon, S. J. & Park, S. Y. Highly fluorescent chameleon nanoparticles C.-W.H., M.M., S.T., M.I., H.S., S.-M.L., and F.M.R. contributed to the synthesis of the and polymer films: multicomponent organic systems that combine FRET and compounds. J.H. and G.G. performed the electrochemical measurements and the asso- photochromic switching. J. Am. Chem. Soc. 134, 12091–12097 (2012). ciated data interpretation. J.A. and G.N. wrote the paper. 6 NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11885-4 ARTICLE Additional information Open Access This article is licensed under a Creative Commons Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- Attribution 4.0 International License, which permits use, sharing, 019-11885-4. adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Competing interests: The authors declare no competing interests. Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless Reprints and permission information is available online at http://npg.nature.com/ indicated otherwise in a credit line to the material. If material is not included in the reprintsandpermissions/ article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from Peer review information: Nature Communications thanks the anonymous reviewers for the copyright holder. To view a copy of this license, visit http://creativecommons.org/ their contribution to the peer review of this work. Peer reviewer reports are available. licenses/by/4.0/. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. © The Author(s) 2019 NATURE COMMUNICATIONS | (2019)10:3996 | https://doi.org/10.1038/s41467-019-11885-4 | www.nature.com/naturecommunications 7