# Fileset

[Angew Chem Int Ed - 2024 - Holstein - Reduction‐Induced Self‐Propelled Oscillatory Motion of Perylenediimides on Water.pdf](https://mdr.nims.go.jp/filesets/8c0b900d-bd86-45ed-92c3-dd134aa7458a/download)

## Creator

[Lara Rae Holstein](https://orcid.org/0009-0008-6412-5108), Nobuhiko J. Suematsu, [Masayuki Takeuchi](https://orcid.org/0000-0002-0207-0665), [Koji Harano](https://orcid.org/0000-0001-6800-8023), Taisuke Banno, [Atsuro Takai](https://orcid.org/0000-0003-3457-3352)

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Reduction-Induced Self-Propelled Oscillatory Motion of Perylenediimides on Water](https://mdr.nims.go.jp/datasets/f99dfc7f-48c2-49db-a0b7-05705106c9e5)

## Fulltext

Reduction‐Induced Self‐Propelled Oscillatory Motion of Perylenediimides on WaterNon-equilibrium ProcessesReduction-Induced Self-Propelled Oscillatory Motion ofPerylenediimides on WaterLara Rae Holstein, Nobuhiko J. Suematsu,* Masayuki Takeuchi, Koji Harano,Taisuke Banno, and Atsuro Takai*Abstract: The emergence of macroscopic self-propelledoscillatory motion based on molecular design hasattracted continual attention in relation to autonomoussystems in living organisms. Herein, a series of perylene-diimides (PDIs) with various imide side chains wasprepared to explore the impact of molecular design andalignment on the self-propelled motion at the air-waterinterface. When placed on an aqueous solution contain-ing a reductant, a solid disk of neutral PDI was reducedto form the water-soluble, surface-active PDI dianionspecies, which induces a surface tension gradient in thevicinity of the disk for self-propelled motion. We foundthat centimeter-scale oscillatory motion could be elicitedby controlling the supply rate of PDI dianion speciesthrough the reductant concentration and the structure ofthe imide side chains. Furthermore, we found that theonset and speed of the self-propelled motion could bechanged by the crystallinity of PDI at the water surface.This design principle using π-conjugated molecules andtheir self-assemblies could advance self-propelled, non-equilibrium systems powered by chemical energy.IntroductionAnimate objects transduce chemical energy into mechanicalforce, from which emerge autonomous rhythmic motionsand spatial patterns, such as heartbeat and circadian activity.These autonomous behaviors are realized by a network ofchemical reactions in a dissipative system, where hierarch-ical self-organized supramolecular assemblies convert mo-lecular-level events into macroscopic ones.[1] In the last fewdecades, research on supramolecular systems chemistry thatinvolves chemically-fueled reaction cycles has advancedtremendously, leading to the development of syntheticsystems that exhibit oscillatory events of, for instance, self-assembling processes and color changes.[2] However, itremains a formidable challenge to explore synthetic systemscapable of eliciting autonomous oscillatory motion at largelength scales.Meanwhile, inanimate, self-propelled objects have beenextensively studied as a type of active matter, from organicto inorganic materials, some of which exhibit oscillatorymotions.[3,4] Representative examples include organic solids,such as camphor, gels, and droplets containing surfactantsthat spontaneously move at the air-water or oil-waterinterface due to the Marangoni flow which is attributed toan anisotropic surface tension gradient in the vicinity of theobjects.[5,6] It has also been reported that the continuous self-propelled motion of organic objects by the Marangoni flowis perturbed by a chemical reaction, resulting in oscillatorymotion.[6] In the case of organic objects on the water surface,the driving force for the self-propelled motion is reduced bythe chemical reaction; however, few systems are known inwhich the reaction product generates the driving force.Furthermore, most of these materials have been used intheir bulk state consisting of random molecular orientations,paying little heed as to how consistent design of molecules,self-assembling processes, and chemical reactions may becombined to control macroscopic oscillatory motion.To address these issues, we have employed a series ofperylenediimide (PDI) molecules as motifs for self-propelledobjects by taking advantage of a few key characteristics.Firstly, although PDI has a rigid and hydrophobic π-conjugated backbone, its hydrophilicity can be increased byintroducing hydrophilic imide side chains.[7] Therefore, we[*] L. R. Holstein, M. Takeuchi, A. TakaiMolecular Design and Function Group, National Institute forMaterials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, JapanE-mail: TAKAI.Atsuro@nims.go.jpL. R. Holstein, M. TakeuchiDepartment of Materials Science and Engineering, Faculty of Pureand Applied Sciences, University of Tsukuba, 1-1-1 Tennodai,Tsukuba, Ibaraki 305-8577, JapanN. J. SuematsuSchool of Interdisciplinary Mathematical Sciences; Graduate Schoolof Advanced Mathematical Sciences; Meiji Institute for AdvancedStudy of Mathematical Sciences (MIMS), Meiji University, 4-21-1,Nakano, Tokyo 164-8525, JapanE-mail: suematsu@meiji.ac.jpK. HaranoCenter for Basic Research on Materials, National Institute forMaterials Science (NIMS), 1-1 Namiki, Tsukuba, Ibaraki 305-0044,JapanT. BannoDepartment of Applied Chemistry, Faculty of Science and Technol-ogy, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama,Kanagawa 223-8522, Japan© 2024 The Authors. Angewandte Chemie International Editionpublished by Wiley-VCH GmbH. This is an open access article underthe terms of the Creative Commons Attribution Non-CommercialNoDerivs License, which permits use and distribution in any med-ium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.AngewandteChemieResearch Articlewww.angewandte.orgHow to cite: Angew. Chem. Int. Ed. 2024, 63, e202410671doi.org/10.1002/anie.202410671Angew. Chem. Int. Ed. 2024, 63, e202410671 (1 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbHhttp://orcid.org/0009-0008-6412-5108http://orcid.org/0000-0001-5860-4147http://orcid.org/0000-0002-0207-0665http://orcid.org/0000-0001-6800-8023http://orcid.org/0000-0002-7480-4242http://orcid.org/0000-0003-3457-3352https://doi.org/10.1002/anie.202410671envisioned that the interfacial properties of aqueous solu-tions, which govern the Marangoni effect, could be system-atically tuned by the molecular design. Secondly, PDI isknown to react with a reductant to form anionic species inwater.[2g,8] Because the anionic species are more water-soluble than the neutral forms, the Marangoni flow can bealtered by redox reactions. In addition, PDI is a well-knownbuilding block for supramolecular self-assembly,[9] therefore,the effect of molecular alignment of self-propelled objects atthe interface on the autonomous motion could be explored.Considering these characteristics of PDIs, we report hereinfor the first time autonomous, centimeter-scale oscillatorymotion of PDI solid disks induced by a reduction reaction atthe air-water interface, with no background reactions, asillustrated in Figure 1. Importantly, we found that thismacroscopic oscillatory motion was significantly affected bythe imide side chains and the molecular alignment at thesurface of the PDI disk.Results and DiscussionSyntheses of PDI Derivatives and Their Redox Properties.We designed a series of PDIs (PDI 1–5) havingoligo(ethylene glycol) chains with different terminal moi-eties at the imide positions, as shown in Figure 1. Comparedto PDI 1 and PDI 3, consisting of triethylene glycol chains,PDI 4 with longer hexaethylene glycol chains is morehydrophilic, while PDI 2 with shorter diethylene glycolchains and PDI 5 with bulky tert-butoxycarbonyl (BOC)groups are more hydrophobic. All the PDIs were synthe-sized by a condensation reaction between 3,4,9,10-perylene-tetracarboxylic dianhydride and the corresponding aminesfollowing a minor modification to a previously reportedmethod.[10] Detailed experimental methods and character-ization data (1H NMR, 13C NMR, and high-resolution massanalyses) are shown in Figures S1–S10 in the SupportingInformation.First, the redox properties of these PDIs were studied.Although PDI 1–3 were hardly soluble in water, theydissolved in a N-cyclohexyl-2-aminoethanesulfonic acid(CHES) buffer solution at pH=9.4�0.2 containing excesssodium hydrosulfite (Na2S2O4), a known reducing agent,yielding a reddish-purple solution. PDI 4 was soluble inwater both in the absence and presence of Na2S2O4. Thisobservation coupled with the UV/Vis absorption spectra ofthe resulting solutions of PDI 1–4 in the presence ofNa2S2O4 (Figures 2 and S11a) were in good agreement withreports of analogous PDI dianion species, [PDI]2� .[8a,11] Instark contrast, PDI 5 remained almost insoluble in waterdespite the presence of Na2S2O4, probably owing to thehydrophobic and sterically hindered BOC groups at theterminal of the imide side chains.It is known that [PDI]2� reacts with dioxygen in air torevert to its original neutral state.[8a,c] However, when anaqueous solution of PDI 1 in the presence of excess Na2S2O4in an open cuvette was stirred at 600 rpm at 25 °C, less than10% oxidation of [PDI 1]2� was observed within 60 min, asshown in the inset of Figure S11b.[12] Similar results werealso obtained for PDI 2–4 under the same conditions.Therefore, the re-oxidation of [PDI]2� during the self-propelled motion, as will be discussed in the next section,did not need to be considered. Additionally, electrochemicalmeasurements (cyclic voltammetry) of PDI 1–5 showed thatthe first and second reduction potentials were almostidentical among the PDIs (Figure S12). Thus, the electroniceffects of the terminal structures of the oligo(ethyleneglycol) imide side chains on the stability of [PDI]2� may benegligible.Figure 1. Schematic illustration of the reduction reaction of PDI 1–5 at the water surface to form their dianion counterparts, which inducesMarangoni flow.Figure 2. UV/Vis absorption spectra of PDI 1 (60 μM) in the presenceof Na2S2O4 (46 mM) in a CHES buffer solution under air at 25 °C. Theinset shows a photograph of the sample solution.AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (2 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=Swimming and Resting Motion of PDI 1. When a bulksolid of PDI 1 was floated on a CHES buffer solution (pH=9.4�0.2) containing Na2S2O4 (23 mM), periodic movementsof swimming at a speed of less than 2 mms� 1 and restingwere observed for over 30 s as shown in Movie S1.[13] Uponcloser inspection, a dark purple solute corresponding todissolved [PDI 1]2� could be seen diffusing from the bulksolid, eventually accumulating below the interface beforethe motion ceased. Similar behavior was also observed for abulk solid of PDI 3 (Movie S2). To study such macroscopicrhythmic motion in detail, we prepared PDI disks of uniformsize (2 mm diameter with 0.5 mm thickness) to eliminatedifferences in motion dependent on the shape of the solid.[14]Because the PDI 1 solid itself was too brittle to be moldedby pressing, it was mixed with a polystyrene-isopreneelastomer and molded to obtain PDI 1 elastomer disks (seeSupporting Information for details). When placed on aCHES buffer solution containing Na2S2O4 (23 mM) in aPetri dish (diameter: 60 mm, solution depth: 3 mm), thePDI 1 disk exhibited oscillating self-propelled motion at amaximum speed of 2.0 mms� 1 and a frequency of 0.14 s� 1 onaverage for over 300 s before coming to rest (Figure 3c andMovie S3). Unreacted PDI 1 still remained in the disk afterthe disk was stopped, indicating that the amount of PDI 1(ca. 2 mg per disk) was sufficient and does not affect theself-propelled motion. The appearance of oscillatory motionof the PDI 1 elastomer disk was similar to that of the PDI 1bulk solids, although the frequency and speed were differ-ent. Such oscillatory motion was not observed for a disk ofpristine polystyrene-isoprene elastomer (Movie S4). Addi-tionally, no motion of the PDI 1 disk was observed in theabsence of the Na2S2O4, as shown in Movie S5. These resultsindicate that the motion was induced by the formation of[PDI 1]2� at the solid-liquid interface. Note that we alsocompared the self-propelled motion of the PDI 1 disk atdifferent depths of buffer solutions in a Petri dish (depths of3 mm and 6 mm) and found no significant difference in thespeed and duration. This result indicates that the maindriving force of the PDI disks is the Marangoni flow, ratherthan other types of flow such as density-driven flow.Spatiotemporal dynamics of the PDI 1 disk could bealtered by varying the concentration of Na2S2O4. At lowreductant concentrations, such as 2.3 mM, there was almostno movement (Figure 3a); when increased above 10 mM,slight but distinct self-propelled motion began to emerge(Figure 3b). Discontinuous intermittent motion was ob-served at both 23 mM and 46 mM, albeit with differentbehavioral consequences depending on the concentration.Unlike the slow, consistent oscillations over 300 s observedwith 23 mM (Figure 3c), the 46 mM case saw significantlyfaster (maximum speed of 19 mms� 1) yet shorter-lived, diskspeeds that ceased within the first 100 s (Figure 3d). After aPDI 1 disk on a Na2S2O4 buffer solution (46 mM) came to ahalt, a second PDI 1 disk was added. No motion wasobserved for this second disk, however when the originaldisk was transferred to fresh reductant solution, self-propelled motion resumed as illustrated in Figure S13.When placed on an aqueous phase containing 92 mM ofNa2S2O4, the disks moved vigorously with a maximum speedof more than 11 mms� 1 only for the first few seconds beforecoming to a stop (Figure 3e). These results indicate that therhythmic motion of PDI 1 disks are highly dependent on theconcentration of Na2S2O4, and the cessation of motion isowed to the accumulation of [PDI]2� and possible oxidationto the insoluble PDI on the surface of the buffer solution.After the PDI 1 disk had ceased moving, deposition of athin film was observed at the surface of the buffer solution,and transmission electron microscopy (TEM) and electrondiffraction confirmed that this material formed sheet-likecrystals (Figure S14), indicating that [PDI]2� dispersed at theair-water interface tend to aggregate.Autonomous Motion of the Other PDIs. We nextinvestigated how the reduction-driven self-propelled motionof PDI elastomer disks changes depending on the imide sidechain. Disks of PDI 2 and PDI 5 moved only slightly at aspeed of less than 2 mms� 1 for the first few seconds and thenstopped (Figures S15b and S15e). We infer that the lowwater solubility of [PDI 2]2� and [PDI 5]2� may haveprevented sufficient reduction reactions and diffusion at theinterface to induce self-propelled motion. A PDI 3 diskexhibited intermittent swimming motion similar to PDI 1, asshown in Figure S15c and Movie S6, and a PDI 4 disk swamslowly with ca. 0.5 mms� 1 for 15 min, as shown in Fig-ure S15d and Movie S7. The apparent color change in thesolution around the PDI 4 disk attributed to the formationof [PDI 4]2� is presumably due to the significantly higherwater solubility of [PDI 4]2� compared to the other PDIs.[15]These results suggest that designed solubility may be used tomodulate the self-propelled motion.Interfacial release of dianion species of PDI 1–4 into theaqueous phase over time was followed by UV/Vis absorp-tion spectroscopy to investigate the difference in thedissolution rate. For each of the PDI compounds, a 2 mmdiameter PDI elastomer disk was placed on the surface of astirred CHES buffer solution containing Na2S2O4 (23 mM)in a cuvette. The dissolution kinetics could be monitored bythe time profiles of the absorbance change (see Figure S16).The dissolution rate constant could be obtained by fitting tothe Noyes-Whitney equation, which is a typical expressionfor the dissolution rate at the solid-liquid interface (seeSupporting Information for details).[16] The resulting valuesare summarized in Table 1. We found that [PDI 1]2� and[PDI 3]2� had modest solubilities of 75 μM and 117 μM,Table 1: Dissolution rate constant (k) and solubility of [PDI]2� in CHESbuffer solutions, and surface tension (γ) of the solutions.Compound k/s� 1 Solubility/μM γ /mNm� 1 [b][PDI 1]2� (3.67�0.04)×10� 4 75.0�0.3 62.7�0.6[PDI 2]2� N/A[a] 0.36�0.01 64.3�1.7[PDI 3]2� (5.05�0.03)×10� 4 116.8�0.3 51.3�2.9[PDI 4]2� (5.20�0.11)×10� 4 341.5�2.8 54.6�0.1[PDI 5]2� N/A[a] N/A 69.4�0.6[a] The dissolution rate constant could not be calculated for [PDI 2]2�and [PDI 5]2� because they are nearly insoluble in water. [b] The γvalues were measured for solutions containing PDIs (60 μM) andNa2S2O4 (23 mM). The γ value of a CHES buffer solution with Na2S2O4(23 mM) was 69.6�0.4 mNm� 1.AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (3 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=respectively, while [PDI 2]2� reached a poor solubility of0.36 μM. No dianion spectral peaks could be observed forPDI 5 owing to its insolubility under these conditions. Onthe other hand, [PDI 4]2� exhibited the highest solubility of342 μM, which was consistent with the appearance of anintense reddish-purple solution as shown in Figure S15d andMovie S7. The respective dissolution rate constants followeda similar order as the solubility, although it was not possibleto accurately determine the rate constants for [PDI 2]2� and[PDI 5]2� due to their poor solubility. These results, coupledFigure 3. Top view of the trajectory (left) and speed profile (right) of PDI 1 disks on a CHES buffer solution containing various concentrations ofNa2S2O4: (a) 2.3 mM, (b) 11 mM, (c) 23 mM, (d) 46 mM, and (e) 92 mM. The color of the trajectory indicates the speed of the disk in mm s� 1.AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (4 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=with our observations of different swimming behaviors,indicate that there is an optimum region of solubility, ordissolution rate, in water where the self-propelled motion ofthe disk is maximized.We have also studied the extent of surface tensionchange upon formation of the respective [PDI]2� in buffersolutions, because the self-propelled motion of the PDI diskis induced by the Marangoni flow, which is attributed to thesurface tension gradient around the disk. As shown inTable 1, the surface tension (γ) of CHES buffer solutions ofPDI 1, PDI 2, PDI 3, and PDI 4 dianion species (60 μM)was found to be significantly lower than that of pristinebuffer solution at 69.6 mNm� 1. The γ value of a [PDI 2]2�solution at 60 μM (64.3 mNm� 1) obtained by sonication wascomparable to that of a [PDI 1]2� solution (62.7 mNm� 1).However, the low solubility at the solid-liquid interface didnot result in a sufficient surface tension gradient; conse-quently, PDI 2 disks showed little self-propelled motion.PDI 5 and its dianion could not be effectively dissolved evenby sonication. As a result, its γ value (69.4 mNm� 1) wasmeasured to be almost the same as that of the CHES buffersolution (69.6 mNm� 1). The γ values were also measured asa function of [PDI]2� concentration for PDIs 1–4. As shownin Figure S17, the γ value tends to decrease linearly withincreasing [PDI]2� concentrations. Inflection points wereobserved at 70 μM and 2 μM for [PDI 1]2� and [PDI 4]2� ,respectively, as critical aggregation concentrations werereached. These results are consistent with the TEM results(Figure S14), in which aggregate formation derived from[PDI 1]2� was observed at the air-water interface.Mechanism of Rhythmic Motion of PDIs. We havedemonstrated the interfacial reduction of PDI can beapplied to induce rhythmic self-propelled motion. Unliketypical self-propelled systems, such as camphor, where thedisk itself is composed of surface-active species, our systemrequires an activating process that converts surface-inactivePDI molecules to surface-active [PDI]2� . We attributevariations in the type of self-propelled behavior to localconcentration conditions around each disk; thus, the mecha-nism can be described by considering three main casesrelating to dianion supply. As depicted in Figure 4a, at lowersupply rates of [PDI]2� (Case 1), the net force acting on thedisk is too small to induce an observable degree of motion.At moderate supply rates of [PDI]2� (Case 2), the PDI diskobtains a sufficient driving force from the resulting surfacetension gradient. At higher supply rates of [PDI]2� (Case 3),the gradient around the disk is too low to induce motion,because all sides are saturated with [PDI]2� .We surmise that the rhythmic self-propelled motion ofPDI 1 and PDI 3 disks was observed in a state of delicatelybalanced conditions between Case 1 and Case 2. At thebeginning, the disk starts to move via a positive feedbackprocess, where the inhomogeneity of [PDI]2� induces asurface tension gradient that causes Marangoni flow(swimming).[6c] The motion of the disk also induces inhomo-geneity as a relatively higher concentration of Na2S2O4emerges in front of the moving disk as a consequence oflocal reductant consumption behind the disk. Eventually,the production rate of [PDI]2� at the front becomessufficiently high, which decreases the surface tension,resulting in restored symmetry of the forces acting on thedisk as it slows to a stop (rest). During the rest, [PDI]2�concentration increases over time and the disk starts tomove again, as depicted in Figure 4b. Here, the slightinhomogeneity of [PDI]2� might remain, thus the movementdirection becomes almost the same as the previousmotion.[17]Minimal self-propelled movement of PDI 2 with its shortdiethylene glycol side chains and PDI 5 with its bulky,hydrophobic BOC moieties can be described by Case 1,where the concentrations of their dianion species are nothigh enough to induce a sufficient surface tension gradient.By contrast, [PDI 4]2� , possessing high solubility coupledwith low surface tension values, exhibited only slow self-propelled motion, which corresponds to Case 3. Theseresults demonstrate how the parameters governing thesupply of surface-active species, such as variations in thechemical structure of the imide side chains of PDIs andreductant concentrations, can be used to control thebehavior of self-propelled PDI disks.Effect of Molecular Alignment of PDIs on Self-Pro-pelled Motion. Finally, we investigated how the molecularalignment of PDIs could affect the self-propelled behavior.As a representative example, we prepared thin films of PDI1 by drop-casting chloroform solution onto 3 mm diameter(0.2 mm thick) polyimide disks, followed by an annealingtreatment (Figure 5a; see Supporting Information fordetails).[18] Thermogravimetric (TG) analysis showed thatPDI 1 is thermally stable up to 310 °C (Figure S18a) anddifferential scanning calorimetry (DSC) results indicatedisotropic melting at 303 °C (Figure S18b). Crystalline do-mains were formed upon heating to 305 °C and subsequentcooling of the drop-cast PDI 1 coated polyimide disks;Meanwhile, those left at room temperature maintained amostly isotropic phase, as observed by polarized opticalmicroscopy (POM; Figure 5b). The difference in crystallinityFigure 4. (a) Schematic diagram depicting three [PDI]2� concentrationprofile scenarios. (b) Mechanism of the oscillatory motion of PDIs onwater upon reduction.AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (5 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=between the annealed and non-annealed disks was alsosupported by X-ray diffraction (XRD) measurements (Fig-ure 5c). The annealed sample displayed a peak at 2θ=4.3°(2.04 nm) which roughly corresponds to the length of PDI 1along the long axis, indicating vertical lamellar stacking.[19]On the other hand, the intensity of the respective peak ofthe non-annealed film was much smaller, indicating thatPDI 1 is disordered in the film.Distinct differences in self-propelled motion emergedwhen we compared the behaviors of annealed and non-annealed PDI 1 films on polyimide for at least 8 differentsamples (Figures 5d and 5e). Disks coated with annealedPDI 1 film displayed self-propelled motion with maximumspeeds averaging 33 mms� 1 upon contact with Na2S2O4buffer solution (Movie S8). By contrast, non-annealed disksexhibited faster self-propelled motion (44 mms� 1 on aver-age), depositing a thick, dark purple solution indicative ofthe rapid release of [PDI 1]2� around the disk (Movie S9).We also observed that the onset of motion took longer forannealed samples (34 s) than for their non-annealed coun-terparts (22 s). This observation was reinforced by UV/Visabsorption spectroscopy of [PDI 1]2� release rate (Fig-ure 5f), which was monitored by placing a PDI 1 disk on thesurface of a stirred CHES buffer solution containingNa2S2O4 (23 mM). The time profile of [PDI 1]2� generatedfrom the annealed sample displayed a considerable waitingperiod for ca. 40 s before the absorbance increased, inaddition to the lower absorbance during the first 300seconds compared to the non-annealed sample. Large, rod-like domains were observed when the annealed sample wasviewed under scanning electron microscopy (SEM) as shownin Figure S18c. Therefore, the altered self-propelled behav-ior is probably due to the higher crystallinity of the annealedsample that significantly decreased the total concentrationof [PDI 1]2� released into the bulk phase. The annealedsample exhibited minimal oscillatory motion, likely becausethe low concentration of [PDI 1]2� around the disk fellunder Case 1, where a sufficient surface tension gradient forMarangoni flow was not induced (Figure 4a). These resultsdemonstrate how molecular alignment of PDI molecules atFigure 5. (a) Schematic illustration of the assembled structure of PDI with and without annealing. (b) POM images of PDI 1 coated polyimide diskswith (left) and without (right) annealing at 305 °C. (c) XRD profiles of PDI 1 drop-cast onto polyimide with (red line) and without (blue line)annealing. (d) Typical trajectory and (e) The corresponding speed profiles of PDI 1 disks with and without annealing on a CHES buffer solutioncontaining Na2S2O4 (23 mM). The color of the trajectory indicates the speed of the disk in mm s� 1. (f) Time profiles of the absorbance changes at538 nm upon addition of PDI 1 disks with (red crosses) and without (blue squares) annealing on a stirred CHES buffer solution containingNa2S2O4 (23 mM).AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (6 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=the interface with water, even a partial alignment, can beused to modulate macroscopic self-propelled motion byinfluencing the supply rate. The effect of longer-range,anisotropic PDI alignment on their direction and speed ofself-propelled motion of the disk is currently under study inour groups.ConclusionsIn summary, we have shown an autonomous, centimeter-scale oscillatory motion induced by the generation ofsurface-active dianion species of perylenediimide (PDI)through an interfacial reduction reaction. We demonstratedhow the solubility, and hence the supply rate, of surface-active species, as well as the surface tension of the buffersolution, could be modulated by variations in the side chainsof the PDI π-scaffold and in the reductant concentration.Further modulation of the self-propelled motion on waterwas achieved by changing the molecular alignment of PDIfilms at the solid-water interface through an annealingtreatment. The design principle of self-propelled objectsbased on π-conjugated molecules and their self-assembliespresented here will pave the way to create supramolecularsystems under non-equilibrium conditions that achievedesired motion using chemical energy.AcknowledgementsWe acknowledge Ms. Izumi Matsunaga (NIMS) for herassistance with the experiments, especially with the synthe-ses of the PDI compounds, Dr. Takanobu Hiroto (NIMS)for his assistance with the XRD measurements, and Dr.Takeshi Yasuda (NIMS) for his assistance to prepare thinfilm samples. A.T. is grateful for financial support fromJSPS (a Grant-in-Aid for Scientific Research (C);JP23K04725), the Izumi Science and Technology Founda-tion, and the Inamori Foundation. N.J.S. is grateful forfinancial support from JSPS (a Grant-in-Aid for ScientificResearch (C); JP23K03347 and (B); JP21H01004,JP20H02712, and JP20H01871). We also thank support froma JSPS Grant-in-Aid for Transformative Research Areas(A) “Materials Science of Meso-Hierarchy” (JP24H01734 toA.T. and JP23H04874 to K.H.) and ARIM of MEXT(JPMXP1223NM5141 and JPMXP1224NM5109).Conflict of InterestThere are no conflicts to declare.Data Availability StatementThe data that support the findings of this study are availablein the supplementary material of this article.Keywords: Perylene dyes · Nonequilibrium processes ·Reduction · Marangoni flow · Oscillatory motion[1] a) A. Goldbeter, Biochemical Oscillations and CellularRhythms: The Molecular Bases of Periodic and ChaoticBehaviour, Cambridge University Press, Cambridge, UK, 1996;b) E. Karsenti, Nat. Rev. Mol. Cell Biol. 2008, 9, 255; c) B. A.Grzybowski, Chemistry in Motion, Reaction-Diffusion Systemsfor Micro- and Nanotechnology, John Wiley & Sons, Chi-chester, UK, 2009; d) J.-M. Lehn, Angew. Chem. Int. Ed. 2013,52, 2836; e) P. K. Hashim, J. Bergueiro, E. W. Meijer, T. Aida,Prog. Polym. Sci. 2020, 105.[2] a) I. Lagzi, B. Kowalczyk, D. W. Wang, B. A. Grzybowski,Angew. Chem. Int. Ed. 2010, 49, 8616; b) E. Mattia, S. Otto,Nat. Nanotechnol. 2015, 10, 111; c) S. N. Semenov, A. S. Y.Wong, R. M. van der Made, S. G. J. Postma, J. Groen,H. W. H. van Roekel, T. F. A. de Greef, W. T. S. Huck, Nat.Chem. 2015, 7, 160; d) S. N. Semenov, L. J. Kraft, A. Ainla, M.Zhao, M. Baghbanzadeh, V. E. Campbell, K. Kang, J. M. Fox,G. M. Whitesides, Nature 2016, 537, 656; e) G. Ashkenasy,T. M. Hermans, S. Otto, A. F. Taylor, Chem. Soc. Rev. 2017,46, 2543; f) S. A. P. van Rossum, M. Tena-Solsona, J. H.van Esch, R. Eelkema, J. Boekhoven, Chem. Soc. Rev. 2017,46, 5519; g) J. Leira-Iglesias, A. Tassoni, T. Adachi, M. Stich,T. M. Hermans, Nat. Nanotechnol. 2018, 13, 1021; h) S. Dhi-man, S. J. George, Bull. Chem. Soc. Jpn. 2018, 91, 687; i) S. N.Semenov, in Out-of-Equilibrium (Supra)molecular Systems andMaterials (Eds.: N. Giuseppone, A. Walther), Wiley-VCH,Weinheim, Germany, 2021, pp. 91; j) F. Sheehan, D. Sementa,A. Jain, M. Kumar, M. Tayarani-Najjaran, D. Kroiss, R. V.Ulijn, Chem. Rev. 2021, 121, 13869; k) M. G. Howlett, A. H. J.Engwerda, R. J. H. Scanes, S. P. Fletcher, Nat. Chem. 2022, 14,805; l) M. ter Harmsel, O. R. Maguire, S. A. Runikhina,A. S. Y. Wong, W. T. S. Huck, S. R. Harutyunyan, Nature 2023,621, 87; m) B. Dúzs, I. Lagzi, I. Szalai, ChemSystemsChem2023, 5, e202200032.[3] a) S. Ramaswamy, Annu. Rev. Condens. Matter Phys. 2010, 1,323; b) W. Wang, W. Duan, S. Ahmed, T. E. Mallouk, A. Sen,Nano Today 2013, 8, 531; c) S. Sánchez, L. Soler, J. Katuri,Angew. Chem. Int. Ed. 2015, 54, 1414; d) Y. Tu, F. Peng, A.Adawy, Y. Men, L. K. E. A. Abdelmohsen, D. A. Wilson,Chem. Rev. 2016, 116, 2023; e) P. Illien, R. Golestanian, A.Sen, Chem. Soc. Rev. 2017, 46, 5508; f) Self-organized Motion:Physicochemical Design based on Nonlinear Dynamics (Eds.:S. Nakata, V. Pimienta, I. Lagzi, H. Kitahata, N. J. Suematsu),The Royal Society of Chemistry, Cambridge, UK, 2018; g) Z.-Z. Nie, M. Wang, H. Yang, Commun. Chem. 2024, 7, 58.[4] a) S. Nakata, M. Murakami, Langmuir 2010, 26, 2414; b) T.Ikegami, Y. Kageyama, K. Obara, S. Takeda, Angew. Chem.Int. Ed. 2016, 55, 8239; c) S. Fujii, S. Yusa, Y. Nakamura, Adv.Funct. Mater. 2016, 26, 7206; d) Y. Tu, F. Peng, X. Sui, Y. Men,P. B. White, J. C. M. van Hest, D. A. Wilson, Nat. Chem. 2017,9, 480; e) A. Altemose, M. A. Sánchez-Farrán, W. Duan, S.Schulz, A. Borhan, V. H. Crespi, A. Sen, Angew. Chem. Int.Ed. 2017, 56, 7817; f) R. Tenno, Y. Gunjima, M. Yoshii, H.Kitahata, J. Gorecki, N. J. Suematsu, S. Nakata, J. Phys. Chem.B 2018, 122, 2610; g) Y. Matsuda, K. Ikeda, Y. Ikura, H.Nishimori, N. J. Suematsu, J. Phys. Soc. Jpn. 2019, 88, 093002;h) C. Zhou, N. J. Suematsu, Y. Peng, Q. Wang, X. Chen, Y.Gao, W. Wang, ACS Nano 2020, 14, 5360; i) S. Arnaboldi, G.Salinas, A. Karajic, P. Garrigue, T. Benincori, G. Bonetti, R.Cirilli, S. Bichon, S. Gounel, N. Mano, A. Kuhn, Nat. Chem.2021, 13, 1241; j) G. Vantomme, L. C. M. Elands, A. H.Gelebart, E. W. Meijer, A. Y. Pogromsky, H. Nijmeijer, D. J.Broer, Nat. Mater. 2021, 20, 1702; k) R. Yoshida, Polym. J.2022, 54, 827.AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (7 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1038/nrm2357https://doi.org/10.1002/anie.201208397https://doi.org/10.1002/anie.201208397https://doi.org/10.1002/anie.201004231https://doi.org/10.1038/nnano.2014.337https://doi.org/10.1038/nchem.2142https://doi.org/10.1038/nchem.2142https://doi.org/10.1038/nature19776https://doi.org/10.1039/C7CS00117Ghttps://doi.org/10.1039/C7CS00117Ghttps://doi.org/10.1039/C7CS00246Ghttps://doi.org/10.1039/C7CS00246Ghttps://doi.org/10.1038/s41565-018-0270-4https://doi.org/10.1246/bcsj.20170433https://doi.org/10.1021/acs.chemrev.1c00089https://doi.org/10.1038/s41557-022-00949-6https://doi.org/10.1038/s41557-022-00949-6https://doi.org/10.1038/s41586-023-06310-2https://doi.org/10.1038/s41586-023-06310-2https://doi.org/10.1146/annurev-conmatphys-070909-104101https://doi.org/10.1146/annurev-conmatphys-070909-104101https://doi.org/10.1016/j.nantod.2013.08.009https://doi.org/10.1002/anie.201406096https://doi.org/10.1021/acs.chemrev.5b00344https://doi.org/10.1039/C7CS00087Ahttps://doi.org/10.1021/la903509zhttps://doi.org/10.1002/anie.201600218https://doi.org/10.1002/anie.201600218https://doi.org/10.1002/adfm.201603223https://doi.org/10.1002/adfm.201603223https://doi.org/10.1038/nchem.2674https://doi.org/10.1038/nchem.2674https://doi.org/10.1002/anie.201703239https://doi.org/10.1002/anie.201703239https://doi.org/10.1021/acs.jpcb.7b11903https://doi.org/10.1021/acs.jpcb.7b11903https://doi.org/10.7566/JPSJ.88.093002https://doi.org/10.1021/acsnano.9b08421https://doi.org/10.1038/s41557-021-00798-9https://doi.org/10.1038/s41557-021-00798-9https://doi.org/10.1038/s41563-021-00931-6https://doi.org/10.1038/s41428-022-00638-8https://doi.org/10.1038/s41428-022-00638-8https://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode=[5] a) C. Tomlinson, Proc. R. Soc. London 1862, 11, 575; b) S.Nakata, Y. Iguchi, S. Ose, M. Kuboyama, T. Ishii, K.Yoshikawa, Langmuir 1997, 13, 4454; c) Y. Ikezoe, G. Wash-ino, T. Uemura, S. Kitagawa, H. Matsui, Nat. Mater. 2012, 11,1081; d) T. Bánsági, M. M. Wrobel, S. K. Scott, A. F. Taylor, J.Phys. Chem. B 2013, 117, 13572; e) J. M. P. Gutierrez, T.Hinkley, J. W. Taylor, K. Yanev, L. Cronin, Nat. Commun.2014, 5, 5571; f) J. Cejková, T. Banno, M. M. Hanczyc, F.Stepánek, Artif. Life 2017, 23, 528; g) M. J. Cheng, G. Q. Zhu,L. Li, S. Zhang, D. Q. Zhang, A. J. C. Kuehne, F. Shi, Angew.Chem. Int. Ed. 2018, 57, 14106; h) C. H. Meredith, P. G.Moerman, J. Groenewold, Y. J. Chiu, W. K. Kegel, A.van Blaaderen, L. D. Zarzar, Nat. Chem. 2020, 12, 1136; i) Y. F.Yano, H. Tada, E. Arakawa, W. Voegeli, T. Ina, T. Uruga, T.Matsushita, J. Phys. Chem. Lett. 2020, 11, 6330; j) A.van der Weijden, M. Winkens, S. M. C. Schoenmakers, W. T. S.Huck, P. A. Korevaar, Nat. Commun. 2020, 11, 4800; k) D.Babu, N. Katsonis, F. Lancia, R. Plamont, A. Ryabchun, Nat.Chem. Rev. 2022, 6, 377; l) T. Kojima, H. Kitahata, K. Asakura,T. Banno, Cell Rep. Phys. Sci. 2023, 4, 101222; m) Y. Sumino,R. Yamashita, K. Miyaji, H. Ishikawa, M. Otani, D. Yamamo-to, E. Okita, Y. Okamoto, M. P. Krafft, K. Yoshikawa, A.Shioi, Sci. Rep. 2023, 13, 12377; n) S. W. Song, S. Lee, J. K.Choe, A. C. Lee, K. Shin, J. W. Kang, G. Kim, H. Yeom, Y.Choi, S. Kwon, J. Kim, Nat. Commun. 2023, 14, 3597.[6] a) S. Nakata, M. Yoshii, S. Suzuki, R. Yoshida, Langmuir 2014,30, 517; b) N. J. Suematsu, Y. Mori, T. Amemiya, S. Nakata, J.Phys. Chem. Lett. 2016, 7, 3424; c) N. J. Suematsu, S. Nakata,Chem. Eur. J. 2018, 24, 6308; d) T. Fujino, M. Matsuo, V.Pimienta, S. Nakata, J. Phys. Chem. Lett. 2023, 14, 9279; e) M.Matsuo, K. Ejima, S. Nakata, J. Colloid Interface Sci. 2023, 639,324; f) G. Zhu, S. Zhang, G. Lu, B. Peng, C. Lin, L. Zhang, F.Shi, Q. Zhang, M. Cheng, Angew. Chem. Int. Ed. 2024, 63,e202405287.[7] M. M. Sun, K. Müllen, M. Z. Yin, Chem. Soc. Rev. 2016, 45,1513.[8] a) E. Shirman, A. Ustinov, N. Ben-Shitrit, H. Weissman, M. A.Iron, R. Cohen, B. Rybtchinski, J. Phys. Chem. B 2008, 112,8855; b) Y. Jiao, K. Liu, G. Wang, Y. Wang, X. Zhang, Chem.Sci. 2015, 6, 3975; c) J. Leira-Iglesias, A. Sorrenti, A. Sato,P. A. Dunne, T. M. Hermans, Chem. Commun. 2016, 52, 9009;d) C. Chen, Z. Guan, Chem. Eur. J. 2023, 29, e202300347;e) C. F. Chen, J. S. Valera, T. B. M. Adachi, T. M. Hermans,Chem. Eur. J. 2023, 29, e202202849.[9] a) M. R. Wasielewski, Acc. Chem. Res. 2009, 42, 1910; b) F.Würthner, C. R. Saha-Möller, B. Fimmel, S. Ogi, P. Leowana-wat, D. Schmidt, Chem. Rev. 2016, 116, 962.[10] a) H. Langhals, Heterocycles 1995, 40, 477; b) C. Huang, S.Barlow, S. R. Marder, J. Org. Chem. 2011, 76, 2386; c) Y. He,S. Savagatrup, L. D. Zarzar, T. M. Swager, ACS Appl. Mater.Interfaces 2017, 9, 7804.[11] The corresponding radical anions, [PDI]*� should be formedtransiently, but we could not observe them under theexperimental conditions owing to the immediate transforma-tion into [PDI]2� with the excess reductant.[12] After about one day of standing, the color of the solutionbecame clear and a red solid was observed precipitating on thesurface of the solution. The recovered solid was dissolved inCDCl3 for1H NMR measurement, and it was determined to bePDI 1 that had returned to the neutral state.[13] We chose a CHES buffer solution at pH 9.4�0.2 because theNa2S2O4 aqueous solution and [PDI]2� were stable under thesebasic conditions. We confirmed that the PDI disks also swamin HEPES buffer solution at lower pH 8.2�0.2, but theexperimental results varied due to the low stability of [PDI]2�as well as the reductant.[14] a) J. P. Gong, S. Matsumoto, M. Uchida, N. Isogai, Y. Osada,J. Phys. Chem. 1996, 100, 11092; b) H. Kitahata, K. Iida, M.Nagayama, Phys. Rev. E 2013, 87, 010901; c) H. Kitahata, Y.Koyano, J. Phys. Soc. Jpn. 2020, 89, 094001; d) Y. Koyano, H.Kitahata, Phys. Rev. E 2021, 103, 012202.[15] Note that although the other PDI disks showed less visiblecolor change of the solute than the PDI 4 disk, when thesolutions around the disks were monitored from the side or leftstanding even after the disks stopped, a dark purple solutecorresponding to dissolved [PDI]2� was clearly observeddiffusing from underneath the disk.[16] H. Takeuchi, in Powder Technology Handbook, 4th ed. (Eds.:K. Higashitani, H. Makino, S. Matsusaka), CRC Press, BocaRaton, US, 2019, pp. 171.[17] The oxidation of [PDI]2� by dioxygen in air at the air-waterinterface and the formation of aggregates, which is not solublein water, may also be one of the reasons for the cessation ofmovement.[18] We employed a polyimide film owing to its thermal stability upto 400 °C or higher. The diameter of the disks was increasedfor improved ease of handling, and the amount of PDI (20 μgeach) to drop-cast was unified.[19] a) S. Tatemichi, M. Ichikawa, T. Koyama, Y. Taniguchi, Appl.Phys. Lett. 2006, 89, 112108; b) T. N. Krauss, E. Barrena, X. N.Zhang, D. G. de Oteyza, J. Major, V. Dehm, F. Würthner,L. P. Cavalcanti, H. Dosch, Langmuir 2008, 24, 12742; c) T.Yasuda, K. Sakamoto, Org. Electron. 2018, 62, 429.Manuscript received: June 6, 2024Accepted manuscript online: July 31, 2024Version of record online: October 2, 2024AngewandteChemieResearch ArticleAngew. Chem. Int. Ed. 2024, 63, e202410671 (8 of 8) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2024, 46, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202410671 by Cochrane Japan, Wiley Online Library on [04/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1021/la970196phttps://doi.org/10.1038/nmat3461https://doi.org/10.1038/nmat3461https://doi.org/10.1021/jp405364chttps://doi.org/10.1021/jp405364chttps://doi.org/10.1002/anie.201808294https://doi.org/10.1002/anie.201808294https://doi.org/10.1038/s41557-020-00575-0https://doi.org/10.1021/acs.jpclett.0c01205https://doi.org/10.1038/s41570-022-00392-8https://doi.org/10.1038/s41570-022-00392-8https://doi.org/10.1016/j.xcrp.2022.101222https://doi.org/10.1021/la403675zhttps://doi.org/10.1021/la403675zhttps://doi.org/10.1021/acs.jpclett.6b01539https://doi.org/10.1021/acs.jpclett.6b01539https://doi.org/10.1002/chem.201705171https://doi.org/10.1021/acs.jpclett.3c02130https://doi.org/10.1016/j.jcis.2023.02.039https://doi.org/10.1016/j.jcis.2023.02.039https://doi.org/10.1039/C5CS00754Bhttps://doi.org/10.1039/C5CS00754Bhttps://doi.org/10.1021/jp8029743https://doi.org/10.1021/jp8029743https://doi.org/10.1039/C5SC01167Ahttps://doi.org/10.1039/C5SC01167Ahttps://doi.org/10.1039/C6CC01192Fhttps://doi.org/10.1021/ar9001735https://doi.org/10.1021/acs.chemrev.5b00188https://doi.org/10.3987/REV-94-SR2https://doi.org/10.1021/jo2001963https://doi.org/10.1021/acsami.6b15791https://doi.org/10.1021/acsami.6b15791https://doi.org/10.1021/jp960398ehttps://doi.org/10.1021/la8030182https://doi.org/10.1016/j.orgel.2018.08.041https://onlinelibrary.wiley.com/action/rightsLink?doi=10.1002%2Fanie.202410671&mode= Reduction-Induced Self-Propelled Oscillatory Motion of Perylenediimides on Water Introduction Results and Discussion Conclusions Acknowledgements Conflict of Interest Data Availability Statement