A bioinspired sequential energy transfer system constructed via supramolecular copolymerization

Sequential energy transfer is ubiquitous in natural light harvesting systems to make full use of solar energy. Although various artificial systems have been developed with the biomimetic sequential energy transfer character, most of them exhibit the overall energy transfer efficiency lower than 70% due to the disordered organization of donor/acceptor chromophores. Herein a sequential energy transfer system is constructed via supramolecular copolymerization of σ-platinated (hetero)acenes, by taking inspiration from the natural light harvesting of green photosynthetic bacteria. The absorption and emission transitions of the three designed σ-platinated (hetero)acenes range from visible to NIR region through structural variation. Structural similarity of these monomers faciliates supramolecular copolymerization in apolar media via the nucleation-elongation mechanism. The resulting supramolecular copolymers display long diffusion length of excitation energy (> 200 donor units) and high exciton migration rates (~1014 L mol−1 s−1), leading to an overall sequential energy transfer efficiency of 87.4% for the ternary copolymers. The superior properties originate from the dense packing of σ-platinated (hetero)acene monomers in supramolecular copolymers, mimicking the aggregation mode of bacteriochlorophyll pigments in green photosynthetic bacteria. Overall, directional supramolecular copolymerization of donor/acceptor chromophores with high energy transfer efficiency would provide new avenues toward artificial photosynthesis applications.

concept of energy transfer. Controlled energy transfer is of great importance for all kind of concepts. I like to strongly recommend publication, but a few points need to be improved before the manuscript can be accepted for publication.
1) The term Light Harvesting System is a bit misleading for this reviewer. It is for me a cascade energy transfer. Although it is with a very high yield -worth publishing in Nat. Comm. -but it does not harvest the light in some kind of reaction. Maybe, it is well accepted, but the general audience will probably expect more. The same type of nomenclature is used in the introduction. I would suggest the authors to make clearer what are the natural systems and what are mimics and where do they differ.
2) The color coding in the figures is changing and makes following the manuscript difficult. See e.g. 3) The authors are using the model of Van der Schoot, many other are using today Mass Balance Models as developed by Markvoort and ten Eikelder (J. Phys. Chem. B. 2012, 116, 5291-5301). 4) Compound 3 is different from 1 and 2. Not only doesn't it show CD, also the cooperativity is much less. It comes close to an isodesmic pathway. 5) Maybe the authors can say something about the thermodynamic parameters of the copolymerization -especially when 3 is mixed with 2 and/or 1 and 2.
Reviewer #3 (Remarks to the Author): In the current paper, the authors reported a novel artificial supramolecular SET-LHSs inspired, constructed via the supramolecular copolymerization of σ-platinated acenes. By comparing the experimental measurement with DFT and TDDFT calculation results, the authors showed that through minor structural variations of acenes, the transitions of vibronic absorption and emission bands could be readily tuned from the visible to NIR region. This result sounds reliable. However, the reviewer pays attention that in the "Result, Spectroscopy of 1-3 in monomeric state" section, the maximum peak in both of the Absorption spectra (Fig2 a) and Electronic transition spectrum (Supplementary Fig 1~3) are not located at the HOMO-LUMO transition region. These data show in the plot but lack discussion in the main text. What the excited state (S2, S3 ?) and transition (HOMO-1→LUMO+1) are related to these peaks? Are those peaks also captured in the experimental signals? Although these peaks appear at higher energy levels, they are still within the high Violet to low UV wavelengths region and may affect the reported spectral red-shifts phenomenon. I would suggest the authors provide some discussion about those peaks. Other than that, the paper is in good shape and suitable for publishing in Nature Communications. Thank the reviewer 1 for the kind comment on this manuscript. According to the reviewer's suggestion, we have supplemented the energy transfer properties of the ternary species 1/2/3 in the "good" solvent CH2Cl2, in which the monomeric state dominates. As shown in Figure R1a, when monomers 1-3 are dissolved in CH2Cl2, the energy transfer efficiencies (ΦET) for 1 and 2 are determined to be 37.1% and 16.7%, respectively. They are significantly lower than those in the "bad" solvent MCH, in which the supramolecular copolymeric state dominates (ΦET: 92.4% for 1 and 92.5% for 2, Figure R1b]. Moreover, no fluorescence emission of 3 (λmax = 770 nm as shown in Figure R1b) is observed for the ternary species 1/2/3 in CH2Cl2 due to the failure of cascade energy transfer. Hence, it is 2 evident that supramolecular copolymerization is necessary for the high energy transfer efficiency of the ternary species 1/2/3. Figure R1. Energy transfer experiments of the ternary species 1/2/3 (c: 8.0 × 10 −5 mol L -1 for 1, 1.6 × 10 −5 mol L -1 for 2, and 8.0 × 10 −6 mol L -1 for 3): a) in CH2Cl2, and b) in MCH.
2. Reply to the second comment made by Reviewer 1 "In the last part: photoresponsiveness of each compound toward 460 nm and 525 nm are suggested be studied in more detail to give more information." According to the reviewer's comment, we have supplemented time-dependent absorption signal changes of supramolecular homopolymers derived from the individual compounds 1-3 upon exposure to 460 nm and 525 nm LED lamps. For 1, the endoperoxidation photochemical reaction follows the first-order kinetics with a rate of 0.00265 s −1 under 460 nm light irradiation (Figures R2a,d). In comparison, they exhibit poor photooxidation capability when exposed to 525 nm light irradiation, as reflected by slight reduction of the anthracene's absorption intensity ( Figure R3a). In terms of 2, the photooxygenation reaction follows to the second-order kinetics ( Figure     Thank the reviewer for the kind comment! We agree with the reviewer that the "light harvesting" is not an accurate concept in the current study, despite that many relating literatures in the energy transfer field use this term. Hence, we have removed the phrase "light harvesting" in the title. In addition, this term has been also replaced by "sequential energy transfer" or "energy transfer" in the main text and supplementary information.

5.
Reply According to the reviewer's comment, we have kept the color coding consistent in the mentioned Figures.

6.
Reply to the third comment made by Reviewer 2 "The authors are using the model of Van der Schoot, many other are using today Mass Balance Models as developed by Markvoort and ten Eikelder (J. Phys. Chem. B. 2012, 116, 5291-5301)." According to the reviewer's comment, we have employed the mass balance model to fit the supramolecular homopolymerization process of 1-3, and acquired the thermodynamic parameters in the revised manuscript (see Table R1). The table has also been added in the main text (see Table 1 in the main text).
The thermodynamic parameters were obtained by analyzing data acquired from CD spectroscopic experiments for 1-2 and absorption spectroscopic experiments for 3. b Changes in Gibbs free energy of elongation (ΔGe) and the cooperativity factor (σ) are reported for a temperature of 298 K.

7.
Reply to the fourth comment made by Reviewer 2 "Compound 3 is different from 1 and 2. Not only doesn't it show CD, also the cooperativity is much less. It comes close to an isodesmic pathway." Thank the reviewer for the kind comment! According to Table R1, the cooperativity factor of 3 is 4.29×10 −3 , much larger than those of 1 and 2. The data reflect the lower cooperativity for the supramolecular homopolymerization process of 3. Moreover, the ∆G value for the supramolecular polymerization process of 3 at 298 K is determined to be -29.0 kJ mol -1 , which is larger than those of 1-2 (Table R1). As discussed in the main text, hydrogen bonding interactions play the prominent role for the 1-3 supramolecular homopolymerization processes. We rationalized that the additional non-covalent interactions may participate in the supramolecular polymerization process of 3, because of the presence of the naphtho[2,3-c][1,2,5]selenadiazole unit in the inner core. DFT calculations show the larger dipole moment for the heteroacene monomer 3 than those of the acene monomers 1 and 2 (μD: 1.75 Debye of 3 versus 0.87 Debye for 1 and 1.20 Debye for 2, Figure R4). It potentially forms dipole-dipole interactions between the neighboring molecules, thanks to the presence of intramolecular charge transfer interactions. When both hydrogen bonding and dipole-dipole interactions serve as the non-covalent driving forces in the supramolecular polymerization process of 3, their subtle interplay may affect the non-covalent interaction mode and thereby lead to the lower cooperativity for supramolecular polymerization.
The conclusion is validated by DFT calculations. For homo-trimers 13-23, the stacking directions of 13 and 23 exhibit slight deviation from the vertical direction (z axis). The phenomena indicate low slipped angles between the stacked monomers (slipped angles: 9.9° for 13 and 6.5° for 23, Figure R4a,b). By contrast, 33 adopt slipped conformations (slipped angles for the three possible conformers of 33 are 40.7°, 53.5°, and 48.9°, Figure   R4c). Hence, the contribution of dipole-dipole interactions may lead to the change of non-covalent stacking modes, which potentially result in the different cooperativity extent between 1-2 and 3 for the supramolecular homo-polymerization processes.

Reply to the fifth comment made by Reviewer 2 "Maybe the authors can say something about the thermodynamic parameters of the co-polymerization-especially when 3 is mixed with 2 and/or 1 and 2."
According to the reviewer's comment, we have supplemented supramolecular copolymerization behaviors of the binary species 2/3 via CD spectroscopic measurements.
With the gradual addition of 3 into 2 (from 0 mol% to 10 mol%, by keeping the concentration of 2 at 8.0 × 10 −5 mol L −1 ), the CD signals of 2 slightly decrease for their intensities ( Figure R5a). Depending on the CD melting curves, Te values slightly elevated with the increased ratios of 3 (from 315.2 K to 315.6 K, Figure R5b and Table R2).
Meanwhile, the ∆H values increase from -79.7 kJ mol -1 to -93.0 kJ mol -1 (Table R2). It is consistent with the supramolecular homopolymerization processes, in which the ∆H value of 3 is higher than that of 2. Nevertheless, the ∆G values are almost identical regardless of the amount of 3, because of the increase of ∆S values with the higher loading of 3. Overall, all of these phenomena exclude the self-sorting arrangement between 2 and 3, and thereby support their supramolecular copolymerization tendency. Figure R5. a) CD spectral changes of supramolecular copolymers 2/3 (c: 8.0 × 10 −5 mol L −1 for 2) in MCH upon varying the ratio of 3 from 0 mol% to 10 mol%. b) CD melting curves and nonlinear fitting for melting curves of 2/3 via the mass balance model.  (Fig 2a) and electronic transition spectrum ( Supplementary Fig 1~3) are not located at the HOMO-LUMO transition region. These data show in the plot but lack discussion in the main text. …… I would suggest the authors provide some discussions about those peaks. Other than that, the paper is in good shape and suitable for publishing in Nature Communications." Thank the reviewer for the kind comment! Compounds 1−3 display two absorption bands in the UV region: namely a low-energy broad band at 308~395 nm, together with a high-energy strong and sharp band. The detailed discussion of the spectroscopic properties is listed as follows.
The In terms of the latter bands, the maxima absorption signals locate at 282 nm (ε = 1.11 × 10 5 M −1 cm −1 ) for 1, 303 nm (ε = 1.18 × 10 5 M −1 cm −1 ) for 2, and 295 nm (ε = 1.01 × 10 5 M −1 cm −1 ) for 3. It mainly originates from high-energy transitions of (hetero)acene   11. Reply to the third comment made by Reviewer 3 "…Although these peaks appear at higher energy levels, they are still within the high Violet to low UV wavelengths region and may affect the reported spectral red-shifts phenomenon.… I would suggest the authors provide some discussions about those peaks." Thank the reviewer for the kind comment! Generally, for the donor-π-acceptor organ- HOMOs of (hetero)acenes, they are destabilized to a large extent than the LUMOs, as revealed by the larger changes of HOMOs upon platination of (hetero)acene units ( Figure