A multicolor photochromic gel based on temporal separation of photochemical reactions

Abstract

Photoswitching among several states is essential for the development of photoresponsive materials showing multiple functions. Here, we developed multistate photochromism under a single light source with constant wavelength and intensity, based on the new concept of temporal separation of photoreactions.

---

New concepts

Optical switching among several states is important for advanced photoresponsive materials exhibiting multiple functions. Such switching has ever been achieved by photochromic reactions dependent on excitation wavelength and intensity, where different excitation conditions can selectively excite different molecular species or populate different excited states, leading to distinct photochemical responses. However, such photochromism requires switching of the light sources, posing complexity in applications. Here, the concept of temporal separation of photochemical reactions has been proposed for achieving multistate photochromism under a single light with constant wavelength and intensity. This concept combines a photochromic clock reaction, where the photoproducts start to accumulate some induction period after the beginning of photoirradiation, and a general photochromic reaction, to temporally separate the two photochromic reactions based on the induction period. In this work, photoreduction of anthraquinone was used as the photochromic clock reaction, showing an oxygen-depletion-based induction period before photoinduced coloration, while a diarylethene derivative was used as the general photochromic compound. A supramolecular gel medium was utilized to control oxygen transfer from the air, giving a well-defined induction period and on-demand decoloration of anthraquinone. The concept of temporal separation will be available not only for other photochromic reactions but also for various photochemical reactions, leading to the development of temporally programmed photofunctional materials.

Photochromic reactions, allowing optical switching of the properties of molecules, have been a basis for developing various photoresponsive materials.^1–4 General photochromic reactions enable switching between two states. Combinations of several photochromic reactions and/or advanced molecular designs have achieved switching among three or more states by changing excitation wavelengths^5–10 and intensities,^11–14 or using additional stimuli such as acids and metal ions,^15,16 resulting in multicolor or multistate photochromism. However, such systems require multiple stimuli, including light sources with different wavelengths and intensities. While this feature offers arbitrary control over multiple states, it also leads to complex operations in applications.

In this context, we propose a temporal separation of several photochromic reactions under light irradiation with constant wavelength and intensity. In a typical photochromic reaction shown in Fig. 1a, the reaction rate is the highest at the beginning of photoirradiation, similarly to other photochemical reactions, and the reaction finally converges to the photostationary state (Fig. 1b). On the other hand, in some cases relating to reaction inhibitors and cooperative reactions, an induction period appears between the beginning of photoirradiation and the substantial start of photochromic reactions (Fig. 1c).^17–19 We call such reactions “photochromic clock reactions”,^20,21 likening them to clock reactions.^22–24 Combining typical photochromic and photochromic clock reactions, two or more photochromic reactions would be temporally separated (Fig. 1d). As a result, the system state is switchable in a multistep manner under a single constant light, depending on the photoirradiation time. This behavior leads not only to multicolor photochromism but also to temporally programmed photoresponsive materials switching their functions autonomously under constant light.

Fig. 1 (a) Typical photochromic reaction scheme. (b)–(d) Color change dynamics under constant light of (b) general photochromic reactions, (c) photochromic clock reactions, and (d) temporally separated photochromic reactions.

We have previously reported photochromic clock reactions based on the photoreduction of anthraquinone (AQ, Fig. 2a) by controlling oxygen penetration from the gas phase with a supramolecular gel.²⁰ AQ is colorless at first but absorbs ultraviolet (UV) light in the presence of reducing agents such as triethylamine (TEA) to generate the yellow reduced forms (the radical anion AQ˙⁻ is mainly formed in strongly basic ethanol (EtOH)). The reduced forms react with molecular oxygen to recover the original oxidized form.

Fig. 2 (a) and (b) Molecular structures of (a) anthraquinone (AQ) and (b) a diarylethene (DE) derivative. (c) Schematic illustration of multicolor photochromism in the DE–AQ gel. (d) Expected reaction dynamics in the two-step coloration. (e) Molecular structure of the gelator, LBG.

In this study, we introduced a diarylethene derivative (DE, Fig. 2b)²⁵ to the AQ-based photochromic supramolecular gel. The DE is originally the colorless open form (o-DE), converted to the purple closed form (c-DE) under ultraviolet (UV) light, and returns to o-DE under visible (Vis) light. As a result of the introduction of the DE, we achieved multicolor photochromism under constant UV light irradiation (Fig. 2c and d). The initial gel is colorless, but UV light irradiation turns the gel purple owing to the ring closing of the DE. Simultaneously, AQ also absorbs UV light, leading to quenching of the photoreduction by dissolved oxygen. Upon quenching, reactive oxygen species, such as singlet oxygen and a superoxide anion, are generated to be consumed by reactions with TEA, etc. Consequently, the dissolved oxygen concentration in the gel decreases along with the ring closing of the DE. Subsequently, the photoreduction of AQ starts when the dissolved oxygen is almost depleted. Under optimized conditions, the photoreduction begins after the photochromic reaction of the DE almost reaches the photostationary state. In other words, the photochromic reactions of the DE and AQ are temporally separated. The gel turns purple owing to the ring closing of the DE at first, and thereafter, the sample is changed to brown by the overlapping of yellow coloration caused by the photoreduction of AQ. Our concept of temporally separated photochemical reactions can be applied to various photochromic and other photoresponsive materials, contributing to the development of temporally programmed photofunctional materials.

The gelator, LBG (Fig. 2e),²⁶ was dissolved in an EtOH solution of DE, AQ, TEA, and sodium hydroxide (NaOH) by heating, and subsequent cooling at room temperature gave the gel sample. The air-saturated gel was colorless at first, but UV light irradiation turned the gel purple, and further continuous UV light irradiation changed it to brown (Fig. 3a and Fig. S1). We traced the coloration dynamics using time-resolved UV-Vis absorption spectroscopy and found that the behavior changed after photoirradiation for ∼2900 s, corresponding to the two-step coloration (Fig. 3b). Absorption spectral changes attributed to the formation of c-DE were observed for 0–2880 s under the photoirradiation (Fig. 3c).²⁵ The ring closing reaction of the DE was little affected by the presence of AQ, TEA, NaOH, and LBG (Fig. S2). On the other hand, absorption spectral changes assigned to the AQ˙⁻ formation were observed after 2880 s (Fig. 3d).²⁷ In other words, chemical species accumulated by photochemical reactions were switched at the border of ∼2900 s, demonstrating that the photochromic reactions were substantially separated by the photoirradiation time. As mentioned later, the border time corresponds to the time until the dissolved oxygen was almost depleted. A slight photodecomposition of c-DE was also observed along with the accumulation of AQ˙⁻. This decomposition would be triggered by the excitation of c-DE in the presence of AQ˙⁻ (Fig. S7). In addition, we mention that the coloration of AQ˙⁻ was gradually diminished from the air–gel interface, owing to the molecular diffusion of oxygen from the air.

Fig. 3 (a) Sample images (8.5 × 10⁻⁴ M DE, 8.6 × 10⁻⁴ M AQ, 5.7 × 10⁻³ M TEA, 1.8 × 10⁻² M NaOH in air-saturated EtOH, 1-mm cuvette) and (b)–(d) color change dynamics of the DE–AQ gel (2.8 × 10⁻⁴ M DE, 2.9 × 10⁻⁴ M AQ, 1.9 × 10⁻³ M TEA, 6 × 10⁻³ M NaOH in air-saturated EtOH, 1-mm cuvette) under 340-nm light irradiation (0.13 mW cm⁻²); (b) absorption change dynamics at 420, 570, and 750 nm, (c) spectral changes between 0 and 2880 s, and (d) spectral changes between 2880 and 3600 s.

The colored gel sample, which had been turned purple and finally brown by UV light irradiation, returned to the original colorless gel by combining two operations of Vis light (525 nm) irradiation and the sol–gel transition induced by heating to 50 °C and cooling to room temperature (Fig. 4). First, Vis light irradiation of the brown gel brought the ring opening of the c-DE to o-DE to leave only the yellow color of AQ˙⁻ (Fig. S1 yellow, S3). Next, oxygen supply from the air by the sol–gel transition resulted in the oxidation of AQ˙⁻, returning the sample to colorless. In contrast, when the sol–gel transition was performed at first, molecular oxygen from the air oxidized AQ˙⁻ to leave the purple color of c-DE. Subsequent Vis light irradiation converted c-DE into o-DE, recovering the original colorless state. This series of results showed that the back reactions of the two photochromic reactions are independently induced by Vis light irradiation and temperature control.

Fig. 4 Coloration and color erasing processes of the DE–AQ gel.

In addition, we evaluated the cyclability of the multicolor photochromism (Fig. S4). A cycle includes the brown coloration under UV light and the decoloration to yellow under Vis light and to colorless by the sol–gel transition. The sample underwent repeated coloration and decoloration at least for five cycles, demonstrating that the multicolor photochromism exhibits some cyclability.

The temporal separation degree between the ring closing of the DE and the photoreduction of AQ corresponds to the induction period until the AQ photoreduction substantially begins. We previously revealed that the induction period of the photoreduction is the duration until the dissolved oxygen is almost depleted, and the induction period was controllable by the nitrogen–air mixing ratio in the sample atmosphere and the light intensity.²⁰ Therefore, we measured the induction period of the photoreduction by changing the sample atmosphere and the light intensity. Consequently, the induction period showed a linear relationship with the partial pressure of oxygen (Fig. S5a). To be exact, the relationship should not be perfectly linear because longer induction periods bring more intense light absorption of c-DE to inhibit AQ from UV absorption. However, as shown in a simple simulation (Fig. S9), the inhibition effect would be small enough to ignore in this experimental condition. We also investigated the light intensity dependence of the induction period to obtain an inversely proportional relationship (Fig. S5b). These results affirm that the induction period in this system corresponds to the duration until the dissolved oxygen is almost depleted.

The oxygen consumption is triggered by the excitation of AQ in the presence of dissolved oxygen. When AQ is excited, the triplet excited state is formed via intersystem crossing, and subsequently, AQ˙⁻ is generated by a reaction with the triplet AQ and TEA.²⁷ The dissolved oxygen quenches both the triplet AQ and AQ˙⁻,²⁰ producing singlet oxygen and a superoxide anion. Such reactive oxygen species would react with TEA²⁸ (and possibly other chemical species like EtOH), leading to a decrease in the dissolved oxygen concentration.²⁰

We also demonstrated multicolor photopatterning in the DE–AQ gel using a single light source. We prepared a photomask composed of areas with different transmittances (%T) and illuminated the DE–AQ gel in a quartz cuvette with 340-nm light through the photomask (Fig. 5 and Fig. S10). As a result, we obtained a multicolor pattern (purple and brown) only via a single photoirradiation, corresponding to the difference in UV light intensity reaching the sample. Subsequent 525-nm light irradiation without the photomask erased the purple color to give a yellow pattern. Finally, the gel–sol–gel transition induced by heating and cooling returned the sample to the colorless original state. While such multicolor photochromic patterning generally requires photoirradiation with several light sources with different wavelengths, our strategy provides a simpler way to print multicolor patterns.

Fig. 5 Multicolor photopatterning using the DE–AQ gel. The transmittance (%T) of the photomask was measured at 340 nm.

Conclusions

We developed a multicolor photochromic gel based on the temporal separation of the photoinduced ring-closing reaction of the DE and the photoreduction of AQ. Under constant UV light, the ring closing of the DE occurred at first, followed by the photoreduction of AQ after the depletion of the dissolved oxygen. The back reactions of the DE and AQ were independently triggered by Vis light irradiation and thermally induced gel-to-sol transition, respectively. Our concept of the temporal separation of photochemical reactions will lead to the emergence of temporally programmed photofunctional materials: for instance, for application in photochromic video displays and autonomous photodriven microrobots, working under constant light.

Author contributions

Y. Nakai: data curation, investigation, validation, and writing – review and editing. Y. Nagai: conceptualization, data curation, funding acquisition, investigation, resources, software, supervision, visualization, writing – original draft, and writing – review and editing. Y. F.: data curation, investigation, and validation. Y. K.: funding acquisition, project administration, resources, supervision, and writing – review and editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting this article are available in the main text or the supplementary information (SI). Supplementary information: detailed experimental procedures, synthetic procedures, additional absorption spectra, and simulation of the reaction kinetics. See DOI: 10.1039/d5mh02246k.

Acknowledgements

This work was supported by JST, PRESTO Grant Numbers JPMJPR22N6, JSPS KAKENHI Grant Numbers JP21K05012, JP24K17749, JP24K01460, and the Sasakawa Scientific Research Grant from The Japan Science Society.

Notes and references

1. Photochromism: Molecules and System, ed. H. Dürr and H. Bouas-Laurent, Elsevier, Amsterdam, 2003.
2. Photon-Working Switches, ed. Y. Yokoyama and K. Nakatani, Springer Japan, Tokyo, 2017.
3. A. Goulet-Hanssens, F. Eisenreich and S. Hecht, Adv. Mater., 2020, 32, 1905966.
4. P. Kobauri, F. J. Dekker, W. Szymanski and B. L. Feringa, Angew. Chem., Int. Ed., 2023, 62, e20230068.
5. H. Nishi, T. Namari and S. Kobatake, J. Mater. Chem., 2011, 21, 17249.
6. M. M. Lerch, M. J. Hansen, W. A. Velema, W. Szymanski and B. L. Feringa, Nat. Commun., 2016, 7, 12054.
7. Q. Zhu, J. Zuo, X. Ping, Y. Zhu, X. Cai, Z. Xiong, Z. Qian and H. Feng, J. Mater. Chem. C, 2022, 10, 8674–8683.
8. L. Köttner and H. Dube, Angew. Chem., Int. Ed., 2024, 63, e202409214.
9. D. Krappmann, A. J. Müller, E. J. Schulze, H. Maid, A. Dreuw and A. Hirsch, Angew. Chem., Int. Ed., 2025, 64, e202507999.
10. B. Balamut and I. Aprahamian, J. Am. Chem. Soc., 2025, 147, 19444–19449.
11. I. Yonekawa, K. Mutoh, Y. Kobayashi and J. Abe, J. Am. Chem. Soc., 2018, 140, 1091–1097.
12. Y. Kobayashi, K. Mutoh and J. Abe, J. Photochem. Photobiol., C, 2018, 34, 2–28.
13. K. Mutoh, K. Yamamoto and J. Abe, Photochem. Photobiol. Sci., 2022, 21, 1445–1458.
14. Y. Kobayashi and J. Abe, Chem. Soc. Rev., 2022, 51, 2397–2415.
15. Y. Lv, H. Ye and L. You, Chem. Sci., 2024, 15, 3290–3299.
16. S. K. Vegge, J. N. Lienert, M. E. Krogh, S. G. Stenspil, M. Dowds, J. K. S. Hansen, C. Warming, C. von Aufschnaiter, V. B. R. Pedersen, J. Wachtveitl, M. Cacciarini and M. B. Nielsen, Chem. – Eur. J., 2025, 31, e202501061.
17. C. A. Jones, L. E. Weaner and R. A. Mackay, J. Phys. Chem., 1980, 84, 1495–1500.
18. C. Decker and A. D. Jenkins, Macromolecules, 1985, 18, 1241–1244.
19. K. Morimoto, D. Kitagawa, C. J. Bardeen and S. Kobatake, Chem. – Eur. J., 2023, 29, 1–14.
20. S. Fujisaki, Y. Nagai, Y. Okayasu and Y. Kobayashi, Mater. Adv., 2024, 5, 1468–1472.
21. Y. Nakai, Y. Nagai, Y. Okayasu and Y. Kobayashi, Chem. Commun., 2024, 60, 6492–6495.
22. J. R. Roebuck, J. Phys. Chem., 1902, 6, 365–398.
23. J. A. Church and S. A. Dreskin, J. Phys. Chem., 1968, 72, 1387–1390.
24. A. K. Horváth and I. Nagypál, Chem. Phys. Chem., 2015, 16, 588–594.
25. M. Irie, K. Sakemura, M. Okinaka and K. Uchida, J. Org. Chem., 1995, 60, 8305–8309.
26. Y. Li, T. Wang and M. Liu, Soft Matter, 2007, 3, 1312–1317.
27. K. Hamanoue, T. Nakayama, Y. Yamamoto, K. Sawada, Y. Yuhara and H. Teranishi, Bull. Chem. Soc. Jpn., 1988, 61, 1121–1129.
28. R. S. Davidson and K. R. Trethewey, J. Chem. Soc., Perkin Trans. 2, 1977, 173–178.

---

Footnote

† These authors contributed equally to this work.

---