Yes, our website features an image of a robot, but what we do here is not creating or designing robots!
We recently redesigned our website to illustrate our aspiration for creating molecular assemblies that work like robots in the future.

Molecular assemblies abound in nature. Proteins and nucleic acid molecules that make up our body also form molecular assemblies, which then aggregate into further assemblies, moving up in scale to form life.

In our lab, we synthesize specially designed new organic molecules and assemble them hierarchically to generate unique molecular assemblies. We are currently working on two major types of molecular designs, as introduced below.

Supramolecular polymers are polymer-like molecular assemblies that are composed of monomer molecules connected by non-covalent bonds. Shown below is an atomic force microscopy (AFM) image of one-dimensionally elongated supramolecular polymers.

Being formed by non-covalent bonds, supramolecular polymers are not as strong as conventional polymers. However, they can be easily disassembled and regenerated, making them increasingly attractive materials around the world. In the human body, different supramolecular polymers undergo disassembly and regeneration to play a part in biological functions.

In our lab, we have discovered a molecular design that imparts curvature to elongated supramolecular polymers that conventionally are one-dimensional (Acc. Chem. Res. 2019 ; Acc. Mater. Res. 2022). By exploiting this molecular design, we have successfully produced supramolecular polymers with unique structures.

　

・Rings and Coils
One of our representative supramolecular polymer structures is the ring (second from the left in the figure below). Many ring-shaped tools and structures can be seen around us, and they must be shaped as such for a reason. We therefore aimed to create rings with supramolecular polymers. As shown in the AFM images below, we started with one-dimensional fibers and gradually changed their molecular structures, eventually succeeding in imparting curvature to the fibers and creating the world’s first uniform rings (J. Am. Chem. Soc. 2009; Angew. Chem. Int. Ed. 2012; Angew. Chem. Int. Ed. 2016).

When we first obtained these rings, we wondered what they could be used for. However, we then discovered that some of these ring-shaped molecules further extended while maintaining curvature to form random structures. By further controlling the rate of extension and improving the molecular design, we were able to obtain exquisite helical and wavy structures (Chem. Commun. 2021; bottom images in the figure above).

Moreover, these coils exhibit a very unusual property: they are adulterated with random structures when initially formed, but eventually fold into beautiful helical coils after about a week in solution (Science Adv. 2018), just like protein folding.

・Structure Control by Supramolecular Copolymerization
The molecule that formed these beautiful helical coils had a naphthalene core. So next, we synthesized compounds with anthracene, hoping that its superior fluorescence would yield fluorescent helical coils.

Unfortunately, the molecules containing anthracenes never formed helical coils, producing only one-dimensionally elongated fibers. Such attempts often do not yield the expected results, but such challenges are actually opportunities for expanding our research. We explored a way to use these anthracene molecules that did not fold helically, and tried mixing the naphthalene molecules (red) with the anthracene molecules (blue) to generate new co-polymers. Lo and behold, we obtained chimeric structures consisting of one-dimensionally elongated fibers with helically coiled ends (Nature Commun. 2019), suggesting the formation of supramolecular block copolymers comprising anthracene and naphthalene connected in a sequence such as AAAAAABBBBBB.

　

There are other ways to generate helical coils. The ring-forming molecules that we discovered in 2012 contained ether groups so we wondered what would happen if we replaced these with ester groups, and tried synthesizing molecules with ester groups. The resulting ester molecules did not form rings, but produced short fibers. It is intriguing how such a change can be caused by a difference of a single carbonyl group. Perhaps the carbonyl groups repel each other.

Next, we tried mixing the ether molecules (red) with ester molecules (green) as before. Initially, we observed no rings or fibers; they seemed to be working against each other. But to our surprise, after one week in solution, helical coils started to grow. It seems that the ether molecules alone form rings and the ester molecules alone form fibers, but their mixture forms helical coils. Such a phenomenon is referred to as a synergistic effect.

Moreover, we also found that these helical coils are formed by alternating supramolecular copolymerization, in which ether and ester molecules are connected in a sequence such as ABABABABABAB through perfect interaction. Such regular arrangement of two molecular species is unfavorable in terms of entropy, which is a measure of disorder. However, perfect interaction between the two molecules is favorable in terms of enthalpy, likely resulting in the gradual helicoidal growth.

Since entropy is affected by temperature, increasing the temperature will result in entropy exceeding enthalpy and the coils collapsing catastrophically, a phenomenon referred to as phase transition. Supramolecular polymers usually dissolve gradually with increasing temperature, so the present finding is an intriguing temperature response (Nature Commun. 2020)。

　

As shown above, we successfully generated novel structures using block and alternating copolymerization in supramolecular polymers. We then tried random copolymerization, in which two molecules are arranged randomly in a sequence such as AABABBABBAAAB. Creating a novel structure by random copolymerization will require two molecules that associate with similar forces and at the same time assemble into distinct structures.

　

One researcher was able to discover such a molecule which associates with comparable force with the ring-forming molecule and at the same time elongates one-dimensionally. We then tried mixing the one-dimensionally elongating molecule (yellow) with the ring-forming molecule (red). The mixture initially formed a linearly elongating polymer, into which ring-forming molecules gradually integrated to impart curvature, eventually forming a spirally folded random copolymer. Moreover, the yellow molecules have UV-reactive substituents, which allows us to dissociate the red and yellow molecules by UV-illumination and increasing the temperature (J. Am. Chem. Soc. 2022).

　

These studies enabled us to achieve block, alternating, and random copolymerization, the three major approaches to polymer synthesis, in supramolecular polymers, thereby generating novel structures.

・Supramolecular Polymers that Spontaneously Degrade and Crystallize
In addition to the ester molecule mentioned earlier , recently we observed some interesting phenomena by introducing esters. For example, an ester molecule forms a random structure and emits orange light, but when left in solution, the molecules rearrange themselves and undergo a structural transition to form bright green-emitting crystalline nanosheets (Chem. Commun. 2020; Chem. Lett. 2020). Further improving the structure of the aromatic moiety resulted in even brighter luminescent nanosheets (Chem. Sci. 2021).

　

This phenomenon of a single molecule species taking on multiple distinct assemblies is referred to as polymorphism. Polymorphism can be exploited to isolate supramolecular polymers with a specific structure. For example, suppose we have a mixture of rings without ends and random structures with ends. When this mixture in solution is heated, only the unstable fibers with exposed ends exhibit polymorphism, in which the molecules rearrange themselves into crystalline fibers and precipitate. The rings, on the other hand, are stable because they have no exposed ends, and remain unchanged upon heating. Thus, filtering the precipitated crystals will produce a solution with only the rings (J. Am. Chem. Soc. 2012). Though we know in proteins that an ends-free circular structure is chemically stable, it is interesting to see similar effect in the world of molecular assemblies.

More recently, we have been able to induce polymorphism without esters by selecting appropriate solvents. We can now isolate supramolecular polymers with a specific structure by using appropriate solvents (J. Am. Chem. Soc. 2019). Though we know in proteins that an ends-free circular structure is chemically stable, it is interesting to see similar effect in the world of molecular assemblies.

More recently, we have been able to induce polymorphism without esters by selecting appropriate solvents. We can now isolate supramolecular polymers with a specific structure by using appropriate solvents (Chem. Eur. J. 2020).

・Photoresponsive Supramolecular Polymers
Our research group has pioneered the study of photoresponsive molecular assemblies (Chem. Eur. J. 2005; Chem. Soc. Rev. 2008), and has been working on imparting photoresponsivity to supramolecular polymers with unique structures.

Introducing azobenzene into helically assembling molecules and exposing the helices to ultraviolet light results in the unfolding and stretching of the helical coils (Nature Commun. 2017), just like watching a protein denature. These helically assembling azobenzene molecules have a flexible structure, and thus the helical coils loosen and unfold from anywhere and everywhere.

Interestingly, when we modified the molecular design to make the azobenzene molecules more rigid, the resulting helical coils, when illuminated with UV light, unfolded from the ends (Angew. Chem. Int. Ed. 2021). The AFM image capturing this unfolding from the ends is really fascinating.

As mentioned above, rings are very stable, but what would happen if azobenzene is introduced into the ring-shaped supramolecular polymers? The rings in a less polar solvent remained stable and unchanged when illuminated with UV light. However, when we added a small amount of polar solvent, which induces molecular exchange, the rings opened to form elongated supramolecular polymers upon illumination (Angew. Chem. Int. Ed. 2019). We call this photo-induced ring-opening supramolecular polymerization.

The azobenzene helical coils that unfolded upon UV-illumination never reassembled back into helical coils once illuminated. However, by using diarylethene molecules, we were able to achieve reversible switching between helical coils and random structures using UV and visible light (J. Am. Chem. Soc. 2021). If such polymers can be synthesized in large quantities, it will be possible to develop gel-like materials whose viscosity and other properties can be modified instantly by using light.

・The Ultimate Molecular Assemblies: Opening the Door to the Mesoscopic Regime
Finally, we obtained molecular assemblies that are a breakthrough for our group, and indeed, the world!(Nature 2020, Nature NEWS AND VIEWS, C&EN, YouTube). We naturally believed that the ring was the dead-end structure and therefore allowed no further organization. However, we discovered that the generation of rings facilitates the generation of new rings from the existing rings, a phenomenon referred to as secondary nucleation. By identifying the conditions that successfully trigger this phenomenon, we succeeded in generating self-assembled polycatenanes.

The image below shows a catenane composed of five interlocking rings that resemble the Olympic rings ([5]catenane). In terms of catenanes synthesized by organic reactions, a [5]catenane was synthesized by Nobel laureate Prof. Stoddart in 1994 and named olympiadane. Our catenane is as big as 80 nm in size, so we named it nanolympiadane. By further modifying the preparation procedure, we also obtained a nanopolycatenane with an estimated 22 interlocking rings including branches. This research makes extensive use of a number of techniques developed in our group, and we encourage you to check out our paper published in Nature . An animation illustrating the process of generating a polycatenane is also available on YouTube.

As shown above, our group has been focusing on studying uniquely structured curved supramolecular polymers. These studies are conducted at the scale of 5 to 500 nm, bridging the range from a few nm to μm. This scale range is referred to as the mesoscopic regime, and is considered the most challenging regime in which to control material's structures. Conventional supramolecular polymers spontaneously extend linearly, making it difficult to generate shapes in the mesoscopic regime. By exploiting curved supramolecular polymers and hierarchical organization techniques, we have successfully generated a variety of meaningful mesoscopic structures.

Another of our proud achievements is the beautiful nano world of molecules that we call scissors-shaped molecules. Shown at the top of the figure below are these scissors-shaped molecules. They open and close just like a pair of scissors. When the solution temperature is high, the scissors are open. When the solution temperature is lowered, the scissors close and become wedge-shaped, eventually assembling themselves into a ring. “Another ring?” you might ask, but these rings stack themselves and form a tube when the solution is further cooled. We cannot exactly tell if the structure is a tube (hollow) from the AFM image, but observation by transmission electron microscopy (TEM) confirmed that it is in fact a hollow cylinder. The wall thickness is 4 nm, which corresponds to the size of one pair of scissors.

The rings we saw earlier with the curved supramolecular polymers did not stack, which suggests that these scissors-shaped molecules form a ring with entirely different properties. Furthermore, the scissors-shaped molecules contain azobenzene, so when exposed to UV light, the rings and tubes in solution dissociate into monomers (J. Am. Chem. Soc. 2012; Eur. J. Org. Chem. 2020; Org. Biomol. Chem. 2020). By creating a special environment inside the rings or tubes, we might be able to devise mechanisms such as photoresponsive release of guest molecules in the future.

The tips of the blades of the scissors-shaped molecules shown above have chiral side chains. When we synthesized these molecules with the chirality of the side chains reversed (the R configuration in blue and S configuration in red), the resulting rings and tubes were completely indistinguishable by their appearance. However, observation by circular dichroism (CD) spectroscopy revealed that these molecules become twisted when the scissors close, giving right-handed or left-handed scissors depending on the chirality of the side chains, thereby imparting chirality to the resulting rings and tubes.

Furthermore, when we tried mixing the R and S enantiomers at different ratios, we observed that all molecules assume the configuration of whichever enantiomer had a higher concentration, exhibiting the majority-rules effect. In other words, the R configuration is supposed to give right-handed scissors under normal conditions, but when R and S enantiomers are mixed in the ratio of 4:6, the R enantiomers, which are the minority here, will follow the S enantiomers, the majority, and become left-handed scissors. As a result, a 4:6 mixture of R and S enantiomers will only yield left-handed rings and tubes. This is the first time that the majority-rules effect has been observed in hierarchical assemblies.

The scissors-shaped molecule has many other interesting aspects, one of which is that it has other modes of assembly besides the ring. For example, when we introduced stilbene in place of azobenzene into the scissors-shaped molecules, they exhibited stronger intermolecular interactions and failed to form rings. Instead, they stacked in a helical fashion and assembled themselves into helical fibers. Illumination with UV light induced photoreaction among stilbene molecules and resulted in the synthesis of a trace amount of foreign products, which then triggered the formation of a new assembly with reversed chirality (Nature Commnun. 2015).

Furthermore, when we introduced a perfluoroalkyl group exhibiting strong intermolecular interaction into the side chains, the molecules again assembled themselves into fibers instead of rings, and the solution gelatinized (Chem. Comm. 2020). It seems clear that stronger interaction among the scissors-shaped molecules leads to failure to form rings, guiding more divergent self-assembly.

We also obtained some unexpected results. When we introduced a very bulky substituent into the side chains of the scissors-shaped molecules, the resulting rings never assembled into tubes even when the solution temperature was lowered. The AFM image of these rings shows that their centers are raised, suggesting that the reduced flatness of the rings prevents them from stacking. In the future, it may be possible to modify the properties of the hole in the ring by introducing side chains with different properties, thereby generating mesoscopic host compounds.

Finally, what if we introduced even more bulky side chains exhibiting strong interaction? Would they assemble into rings, or fibers? To find out, we introduced bulky yet cohesive cholesterol. Interestingly, the cholesterol-modified scissors-shaped molecules formed both rings and fibers. Moreover, we were able to control the yield of these assemblies to some extent by adjusting the intensity of light. This means that the ring-closing and the elongation pathways are kinetically competitive (Chem. Sci. 2022).

Described above are two research projects being pursued in our group. The key to both projects is curvature, and we successfully generated curvature by two entirely different molecular designs. In both cases, the rings guided the discovery of curvature. The process of opening the rings and connecting them to form helical coils, or stacking them to form tubes, led us to this new world.

We have yet to figure out how these materials can be used. After all, molecular assemblies with uniform curvature have never existed before, meaning that we have discovered unprecedented materials, phenomena, and underlying principles.

Whenever we find something interesting, we scrutinize and discuss it as a team, asking questions such as “What caused this structure to form?” and “Why are the results not what we expected?” Through this process, we discover hidden new science. Our instinctive urge to find something interesting is crucial.

Now that we fully understand the fascination of curvature, we feel that this is the beginning of a new endeavor. By exploiting the science of curvature, we may be able to develop supramolecular polymers that work like robots in the mesoscopic domain. Our true challenge is just beginning.