Development of a new catalyst system that promotes the cleavage of a weak metal-metal bond
The catalytic cleavage of the boron-boron (B-B) bond and its functionalization is an important subject in organic chemistry because the resulting boron-containing molecules can be further manipulated in several ways. So far, late transition metal catalysts (Group 8-11 elements) have been used for this purpose. In order to validate the tendency of the cleavage of B-B bond by computational analysis, B2(eg)2 was set as a model diboron compound. The AFIR calculation was conducted using Pt(0), MePd(II)OAc, Pt(0), Cu(I)F, Cu(I)Ot-Bu, Cu(I)OMe, Ag(I)OMe, and Au(I)OMe in combination with PH3, a model phosphine ligand. Additional calculation was performed using Cu(I)OMe, Ag(I)OMe, and Au(I)OMe with NHC ligand. As a result, all calculated metal complexes except Pd(0)-PH3 and Au(I)OMe-NHC smoothly promoted the cleavage of B-B bond (the calculated activation energy is less than 100 kJ/mol). Subsequently, an early transition-metal complex such as Zr(IV)Cp2Cl2 was used in the calculation. However, in this case, the activation energy was very high (155.8 kJ/mol), which suggested the B-B bond cleavage was difficult.
Next, we designed a chemical reaction using Pt(0) and B2(pin)2 without a ligand (the calculated activation energy is 1.2 kJ/mol: almost barrierless!), which is a relatively rare combination for the catalytic transformation of diboron compounds. To develop a new functionalization method, we turned our attention to silylation of diborons. Iridium-catalyzed silylation of diborons with a silylhydride to prepare silylboranes was actively studied by Hartwig (Organometallics 2008, 27, 6013–6019). However, synthesizable silylboranes are limited. In particular, the synthesis of trialkylsilylboranes bearing steric hindrance or functional groups has not been reported. Therefore, if a general synthetic method of trialkylsilylboranes is developed, such a method would be effective due to its novelty and good performance. Furthermore, silylboranes thus obtained are useful precursors for C-Si and Si-Si bond formation reactions, serving as silyl anion equivalents in the presence of a transition-metal catalyst or an appropriate base like MeLi.
We thus began the chemical experiment using a catalytic amount of Pt(0)/C in the presence of B2(pin)2 and R3Si-H. Fortunately, various types of trialkylsiliylborans were successfully obtained in synthetically useful yields. Moreover, we conducted the AFIR study to reveal each step of the catalytic cycle, proposing a reasonable Pt(0)-mediated catalytic cycle under ligand-free conditions.
This research was done in a collaborative way with experimental chemists (Ito group) and computational chemists (Maeda group). The interdisciplinary research grant was mainly used for the setup of the mixed lab to enhance the mutual collaboration between experimental and computational chemists. Expensive experimental devices including a high-vacuum manifold (300,000 yen), a crushed ice maker (450,000 yen), and a diaphragm pump (350,000 yen) were purchased in order to accelerate the chemical experiments.
Synthesis of luminescent complexes using a theoretical calculation approach
Lanthanide complexes show characteristic emission with high color purity and prominent chiroptical properties; thus, they have garnered considerable attention as luminescent materials. However, the lack of understanding of their detailed electronic structure has suppressed the development of design strategies for targeted photo-function of lanthanide complexes. Therefore, I have undertaken joint research with computational chemists at WPI-ICReDD and set up the workstation in our group (experimental chemists) to simulate the electronic structures of lanthanide complexes.
Development of new solid-state organic reactions through collaborative researches
In this year, in order to launch the fusion research on the development of new solid-state organic reactions at ICReDD, experimental investigations that would become the seeds of future collaborative research with calculation and information scientists were conducted. The start-up budget was primarily used to purchase ball milling equipment, as well as glassware and chemical reagents needed for chemical experiments.
The development of new solid-state reactions using mechanical energy as a driving force was targeted as a seed project for future fusion research. As a result, we succeeded for the first time in the development of the mechanoredox system, a new redox reaction driven by the mechanical impact generated by a ball mill (Kubota, K.; Ito, H. et al., Science 2019, 366, 1500). In this catalytic system, the agitation of piezoelectric materials via ball milling generates temporarily highly polarized particles that can act as strong reductants to transfer electrons to small organic molecules, followed by oxidative quenching of a donor, thus inducing the selective formation of bonds in a manner analogous to photoredox catalysis. By utilizing this new solid-state redox reaction using ball milling, C–H arylation and borylation of hardly soluble compounds, which could not be applied by a conventional photoredox catalysis, with aryldiazonium salts were achieved. The present mechanoredox reactions can be carried out on gram scale without organic solvents in air, and do not require special operating conditions. This operational simplicity suggests that the present approach may complement existing photoredox transformations in a practical and environmentally friendly manner. In the next year, we will aim to clarify the detailed mechanism and improve the efficiency through collaborative research with computational and information scientists in the ICReDD.
We have also discovered selective mechanochemical monoarylation reactions of unbiased dibromoarenes using in situ crystallization (Kubota, K.; Ito, H. et al., J. Am. Chem. Soc. 2020, 142, 9884). Suzuki–Miyaura cross-coupling reactions of unbiased dibromoarenes in solution tend to provide a mixture of mono- and diarylated products. The selective trend toward monoarylation is mostly likely derived from the fact that the liquid substrate is more reactive than the solid substrate under solvent-free mechanochemical conditions. This work has been selected as a supplementary cover art in the Journal of the American Chemical Society. To probe the mechanism, we are currently working on the construction of a mathematical model in collaboration with Associate Professor Dr. Tetsuya Yamamoto at ICReDD.
Understanding of physical properties at the atomic level of a supramolecular bonding network
Recently, we synthesized novel polyampholyte hydrogels (PA gels) forming a hierarchical structure, which show high toughness, high fatigue resistance, and self-healing. PA gels consist of reversible ionic bonds at the 1-nm scale, a cross-linked polymer network at the 10-nm scale, and bicontinuous hard/soft phase networks at the 100-nm scale. Such hierarchical structures are found to strongly depend on the chemical structure of the monomers, which is essential for the high toughness and fatigue resistance of PA gels. Here we studied how the hierarchical structure forms and how it influences the mechanical properties of PA gels. With real-time X-ray scattering, we revealed the structure formation process of PA gels. By using an osmotic stress method, we established the correlation between the relative strength of the soft and hard phases, the viscoelastic properties, the toughness and the self-recovery behavior of PA gels. In addition, we also studied the effect of the phase-separated structure on the fatigue behavior of PA gels and highlight the role of the multiscale structure on fatigue resistance.
The phase separation data may lead to future fusion research of “phase separation behavior of tough and self-healing polyampholyte hydrogels”. Furthermore, my experimental conditions have been largely improved by the start-up funding. I purchased some equipment such as a temperature-controlled oven, an electronic balance and so on. My office condition was also improved by the start-up funding, such as through the purchase of a monitor for my paperwork.
Development of a new reaction using the reaction path search
The current trial-and-error approach to the development of new reactions is time-consuming and inefficient. A core technology of ICReDD uses “reaction path search methods” using the AFIR (Artificial Force Induced Reaction) method in combination with machine learning, proposing possible experimental parameters such as starting materials and catalysts, etc. To develop a bioisostere of amino acids, we planned to synthesize an α,α-difluoroglycine derivative, which is the most simple α-fluorinated α-amino acid. We first retro-synthesized α,α-difluoroglycine by AFIR, proposing three basic and simple starting materials such as NR3 (amine), :CF2 (difluorocarbene), and CO2 (carbon dioxide). Subsequently, when NMe3 (R = Me) and –CF2Br, which is a reasonable difluorocarbene precursor, were used as reactants in further calculation, Me3N+-CF2-CO2– was obtained via three-component assembly in 99.99% yield. Encouraged by this promising result, we conducted the chemical synthesis of a difluoroglycine derivative. As a result, Me3N+-CF2-CO2Me∙BF4– was obtained in 80% yield after methyl esterification. It took only two months to complete the synthesis, which has emphasized the power of AFIR.
I set up our new mix lab as a chief by the support of this interdisciplinary research fund, which is mostly used for experimental tools, glassware, and chemical reagents. I also advertised our research activities at the 14th International Conference on Cutting-Edge Organic Chemistry in Asia (ICCEOCA-14) held in Niseko, Hokkaido, Japan, September 26-29th, 2019.
Advanced Visualization of the Reaction Route Network
Visualization of automated reaction pathway search results as graphs can provide an intuitive picture to help understand the entirety of chemical reaction routes. Following our previous work to reduce a dimension of a set of reference structures along the IRC by a classical multidimensional scaling (CMDS) approach (J. Chem. Theory Comput. 2018, 14, 4263), we proposed a method to project on-the-fly trajectories into a reduced-dimension subspace determined by the global reaction route map, using the out-of-sample extension of CMDS. As a demonstration, the method was applied to the SN2 reaction, OH– + CH3F, and to the structural transformation of Au5 cluster, which was accepted by JCTC.
Although the method of mapping the reaction path network for a few-atom system in two dimensions makes sense, it is difficult to employ this approach directly to a reaction type such as A + B → C + D or a dissociation into fragments, which appears commonly in catalysis and organic reactions. To extend the methodology to such reaction types, we exchanged thoughts and ideas on the role of the reaction route maps in reaction design with experimentalists and information scientists within ICReDD. There is a need to develop a notation that extracts important pathway information from complex reaction route maps and allows the experimentalists to search for similar reaction paths as for different compounds. Based on the knowledge and methods obtained in this project, we have just started a fusion study to speed up reaction development by using reaction pathway maps as knowledge sharing notes for reaction design.
To facilitate the creation of models that accurately represent the results of the calculations, a 3D printer was introduced in this project. I developed the program and set up environments to create the models. In combination with the reaction route maps, discussions using the models from the 3D printer could contribute to the development of new catalytic reactions.
Development of an automatic reaction path search method for macromolecular systems
In a collaboration among organic, medical, theoretical and information researchers, we tried to extend the multistructral microiteration method. In previous MSM(-ME) methods, an electrostatic interaction between reaction center and surrounding atoms has been treated classically. To describe the interaction quantum mechanically, we have extended and developed it into the MSM(-EE) method. The resulting MSM-EE method gave results in reasonably good agreement with experimental ones, while the MSM-ME method underestimated the barrier height by more than 5 kcal mol-1. Based on these results, we expect that our method will become an important tools for reaction design using enzymes and proposed new research.
With the start-up budget, we bought a software and arranged the research environment. Especially our development and test calculations improved due to the new software.
Elucidation of the reprogramming phenomenon of cancer stem cells using a polymer hydrogel and establishment of a new cancer stem cell-targeting therapy
We have previously found polymer hydrogels to induce cancer cell reprogramming in a very short time to create cancer stem cells (CSCs) (Japanese Patent Application 2017-028833, PCT/JP2018/ 005884, US 16/487,247). To develop a clinical application, we here analyzed the reprogramming induction phenomenon by hydrogels by combining computational science, information science, and experimental science at ICReDD.
(i) Creation of reprogramming-guided hydrogel substrate (collaboration with the Gong group)
In order to efficiently induce the reprogramming of cancer cells, we optimized the hydrogels by adjusting the elastic modulus and charge state. In addition, for the rapid and efficient large-scale drug screening targeting CSCs induced by the hydrogel, we succeeded in constructing a system consisting of a sheet-shaped gel that can be easily cut and installed into a 24-well plate. This is an extremely important technological innovation for “development of a CSC diagnostic kit using hydrogels”.
(ii) Creation of a reprogramming analysis platform based on hydrogels (Tanaka group)
To quantify spatiotemporal variables in hydrogel-induced CSCs formation, we performed scRNAseq of brain tumor cells cultured on hydrogels, and succeeded in identifying multiple markers specific to CSCs. This will be valuable information for the selection and creation of CSCs-targeting drugs, and also will be the basic knowledge for the creation of a new academic field called “Material Genomics” which is the ultimate goal of this project within ICReDD.
(iii) Creation of the technology and theory for predicting and manipulating characteristics and dynamics of cell populations (collaboration with the Komatsuzaki group)
We are now developing a method for predicting CSC transformation from spatiotemporal cell dynamics information based on information science methods, which makes it possible to predict the induction process of CSCs. In addition, we are constructing rapid drug screening technology with accuracy guarantee using a Multi-armed Bandit algorithm. This enables the rapid selection of CSC-targeting therapeutic agents for clinical application.
Taken together, the results of these fusion research projects will lead to the development of CSC diagnostic methods based on highly functional hydrogels and the creation of CSC-targeting therapeutic agents as medical applications of the ICReDD project.