A building-block design for enhanced visible-light switching of diarylethenes

Current development of light-responsive materials and technologies imposes an urgent demand on visible-light photoswitching on account of its mild excitation with high penetration ability and low photo-toxicity. However, complicated molecular design and laborious synthesis are often required for visible-light photoswitch, especially for diarylethenes. Worse still, a dilemma is encountered as the visible-light excitation of the diarylethene is often achieved at the expense of photochromic performances. To tackle these setbacks, we introduce a building-block design strategy to achieve all-visible-light photochromism with the triplet-sensitization mechanism. The simply designed diarylethene system is constructed by employing a sensitizer building-block with narrow singlet-triplet energy gap (ΔEST) to a diarylethene building-block. A significant improvement on the photochromic efficiency is obtained as well as an enhanced photo-fatigue resistance over those under UV irradiation. The balance between the visible-light excitation and decent photochromism is thus realized, promoting a guiding principle for the visible-light photochromism.


Review 1:
The manuscript details the development of a donor-acceptor photochromic dyad where the donor is a sensitizer that allows for a fully visible light photoswitch based on triplet-triplet ET.
The experiments are well realized with adequate controls to determine the mechanism and with definite demonstration of the application of the system including in a solid-support.
The general interest is large and the area of research is of considerable novelty. I recommend publication in Nature Communication after some minor improvements. Main Text, Page 9-10: To evaluate the high efficiency of the visible-light photochromism, quantum yields and conversion ratios (Table 1 and Supplementary Table 2 Eq. 2 Equal to: Where α 1 , α 2 are the photocyclization conversion ratios in PSS state at the given wavelength of λ 1 , λ 2 , respectively. The absorbance A measured at any particular wavelength λ of a mixture of open and closed forms was introduced, while the overall concentration C o + C c is constant and the formula can be obtained as follows: Where α 1 , α 2 are the photocyclization conversion ratios in PSS state at the given wavelength of λ 1 , λ 2 , respectively.    Main Text, Discussion: ... Third, the modular building of DAE-sensitizer system can also be directed by various means, e.g. supramolecular self-assembly 44 and polymer chemistry 45 , which may inspire R-9 further design of water-soluble, all-visible-light triggered DAEs. Under such circumstance, a wide research direction might be open to biomaterials and bio-technologies operating with visible-light photoswitches, such as super-resolution imaging 46 , in which visible-light switching and water-solubility are highly demanded. Current diarylethene system still needs to be improved as the visible light wavelength used in this work (λ = 420 nm) doesn't reach the best wavelength range for materials and biological applications (biological window: 600-1200 nm). Delicately selecting diarylethene cores with lower triplet energy levels and corresponding narrow ΔE ST sensitizers is to be carried out to achieve red-light triggered photochromism that enables deeper tissue penetration during bio-imaging and modulation, which would be our next work. erased from the same filter paper with DAE/DT using different masks with 420 nm irradiation for 7 min and > 550 nm irradiation for 10 min, respectively.

Main Text, Page 16:
…Meanwhile, a vague photoswitchable patterning was obtained by soaking the filter paper with a mixed solution of DAE-1o and DT (1 mM, molar ratio = 1:1, Supplementary Fig. 14).
The much attenuated photochromism of DAE-1o/DT in filter paper is probably due to the reduced mobility in solid state that significantly affects the intermolecular TTET process. The Indeed, the measured values may be more valid than the calculated ones. Regretfully, it is not fully accessible in our system. Therefore, we still employed the calculated values for both DAE-1o and DT. Thanks again for your advice.

Q4: While the PL lifetime of DAE-DT dyad was measured and discussed, no static PLQY of it is not reported in this work. I would suggest the PLQY measurements of DAE-o-DT and DAE-c-DT. If fluorescent, its switching behavior should also be discussed.
A4: Many thanks for your constructive suggestion. We have measured the visible-light photoswitchable emission performances of our DAE  The values of k rISC are obtained according to the mentioned formulas (see Table R1 below). Compared to the bald DT, the dyad DAE-o-DT has a much lower k rISC because of the competitive TTET process. We also evaluated k rISC of DAE-o-DT at different temperature.
There is no noticeable change for the luminescence delayed spectra neither in the solution at 350 K nor in the PMMA film at lower temperatures (see Fig. R3-4), which showed that negligible change of k rISC occurred from 100 K to 350 K.

R-18
The temperature-independent performances made us rethink about the temperature-dependent photochromism as we claimed in the context. We figured out that the lower photocyclization quantum yields in high temperature might be attributed to the decreased thermo-stability of DAE-c-DT. We then checked thermal-stability of DAE-c-DT under 350 K and found a thermally induced cycloreversion after 1 day. Furthermore, the photochromism of DAE-1c under 350 K was investigated and an obviously decreased Φ o-c = 0.27 was obtained compared to that of Φ o-c = 0.36 under 300 K (see Fig. R5 and Table R2 below). Hence, in order not to confuse the readers, we would like to remove Fig. 4d and the related statements from the Main Text.  increasing polarity (Fig. 4a). An obvious solvatochromic effect on the emission spectra was observed with a red-shift of the fluorescence peak from 444 nm in cyclohexane to 598 nm in acetone, indicating a lowering of excited energy levels with the increasing solvent polarity.
Accordingly, a declined visible-light TSP performance was determined ( Fig. 4b and Supplementary Fig. 13), as the photocyclization quantum yield dropped from Φ o-c = 0.40 in cyclohexane to Φ o-c = 0.008 in acetone (Supplementary Table 4). Though DAE-DT presents the best visible-light photochromism in cyclohexane, the relatively lower solubility may hamper its further applications. As a result, toluene was selected as the solvent in this work. Phys. Chem. A 2018, 122, 6673). Regretfully, we failed in tracing the transient absorption peak of the triplet excited state of DAE-1o (usually within a sub-nanosecond timescale) by our nano-transient absorption equipment, probably due to the lack of demanded resolution.

R-21
Reviewer 3: This paper reports a new design for diarylethene derivatives, which undergo photochromic reactions in both directions upon irradiation with visible light. The visible light sensitivity is one of challenging targets in molecular photoswitches. The present approach is of quite attractive and worth publishing. However, there are some drawbacks to this paper from the view point of photophysical chemistry. It is recommended to take into account the following comments. The apparent concentration-dependent visible-light photochromism was concluded in mixed systems. In order not to mislead the readers and make our description more convincing, we rewrote the section of "Selection of Matched DAE/Sensitizer Building Blocks" as follows:  Supplementary Fig. 3). Note that the DAE-1o alone is inert under 420 nm irradiation ( Supplementary Fig. 3  toluene and deaerated toluene, respectively. Considering the triplet life of DT alone is around 2973 ns, the oxygen quenching effect is our system is not that obvious probably due to the highly quenching efficiency from DT to DAE moiety. Similar oxygen quenching effect can be found in MLCT system, as the photocyclization quantum yield decreased from 0.44 to 0.22 in acetonitrile was reported previously in the paper (J. Am. Chem. Soc., 2008, 130, 7286).

REVIEWERS' COMMENTS:
Reviewer #1 (Remarks to the Author): I commend the authors for the improvement of the manuscript, expanding the set of experiments and even accepting that there were better interpretations with the new experiments. I believe that manuscript should be published. My only note is that ref 46 that has been added (Park et al) and is supposed to be about super-resolution imaging from photoswitching, is not relevant to the discussion at all. There are plenty of groups (Hell, Raymo, Enderlein, etc) who have published in that area, please pick a more appropriate citation.
Reviewer #2 (Remarks to the Author): In this revised manuscript, authors successfully followed my comments to make supplementary experiments and to strengthen the ideas involved in this work. Additional synthesis for the longer spacer reference compound, static PLQY measurements, and rate constants evaluation etc. provided a very consistent and supporting evidences for the proposed TTET mechanism of DTE ring cyclization. I am now fully satisfied with this revised manuscript.
Reviewer #3 (Remarks to the Author): The revised manuscript is worth publishing in Nat. Commun.

REVIEWERS' COMMENTS:
Reviewer In this revised manuscript, authors successfully followed my comments to make supplementary experiments and to strengthen the ideas involved in this work. Additional synthesis for the longer spacer reference compound, static PLQY measurements, and rate constants evaluation etc. provided a very consistent and supporting evidences for the proposed TTET mechanism of DTE ring cyclization. I am now fully satisfied with this revised manuscript.
A: Thank you very much for your comments, which really help us a lot to