第574回のスポットライトリサーチは、北海道大学 触媒科学研究所 触媒構造研究部門 高草木研究室のCan Liu（劉燦）博士にお願いしました。
高草木研では、走査型トンネル顕微鏡（STM）やX線吸収微細構造（XAFS）などの先端的ナノ構造解析手法を、原子レベルでのその場/オペランド表面計測法として高度化し、これらを用いて触媒・電極反応プロセスの可視化と機構解明、制御指針の獲得を行っています。また、従来の熱による活性化のみに頼る触媒反応だけではなく、プラズマや電場による活性化も併用した非在来型触媒プロセスの開拓と機構解明にも挑戦しています。
本プレスリリースの研究内容は触媒表面に吸着した分子の動きについてで、触媒表面に吸着した反応分子（中間体吸着種）がどのように表面上を拡散し、活性サイトに到達して生成物に変換されるのかを、本研究グループは原子レベルで可視化することに成功しました。
Figure 1. (a) STM image of methoxy intermediates adsorbed on the five-fold coordinated Ti4+ sites (Ti5c) of Pt/TiO2(110) surface. (b) Scheme showing how the methoxy intermediates on the TiO2(110) surface diffuse to the Pt active sites and decompose to CO and CH4.
この研究成果は、「Journal of the American Chemical Society」誌に掲載され、またプレスリリースにも成果の概要が公開されています。
Can Liu, Bang Lu, Hiroko Ariga-Miwa, Shohei Ogura, Takahiro Ozawa, Katsuyuki Fukutani, Min Gao, Jun-ya Hasegawa, Ken-ichi Shimizu, Kiyotaka Asakura, and Satoru Takakusagi
J. Am. Chem. Soc. 2023, 145, 36, 19953-19960
研究室を主宰されている高草木 達教授よりLiu博士についてコメントを
この研究は私にとってとても思い入れが強いものです。研究を開始したのは20数年前で、私が博士課程在籍時でした。担持金属触媒において、金属ナノ粒子上に吸着した反応ガスが酸化物担体表面へ溢れ出すスピルオーバー現象が古くから知られていましたが、STMによる直接可視化に初めて成功しました。この成果は2010年にアメリカ化学会のLangmuir誌に発表済みです。その後はSTM装置を自身で保有することができず、研究を中断していましたが、STM装置を運よく中古で手に入れ、さらに大型プロジェクトに参画する機会を得ました。そこで博士研究員としてやってきたのが、劉燦さんです。彼女の専門は材料合成や電気化学で、固体触媒の研究やSTM等の超高真空装置の扱いは未経験でした。しなしながら、持ち前の器用さと深い洞察力で1年を待たずにこの分野で通じる優秀な研究者に成長しました。今回の成果は私が長年抱いていた、“スピルオーバーした反応中間体は、最終的にどこで反応してどのような生成物を与えるのか？酸化物上それとも金属上？また、そこに至るまでの中間体の動きが触媒活性・選択性にどのように影響するのか？”、を研究テーマとして彼女にアサインし、それを見事に解明してくれたものです。今後は彼女自身のアイディアをプラスして、触媒反応の機構解明や新しい反応制御法の提案などをしていって欲しいと期待しています。
Q1. 今回プレスリリースとなったのはどんな研究ですか？簡単にご説明ください。
The mechanisms of surface reactions are at the core of understanding heterogeneous catalysis. The dynamic behavior of intermediate adsorbates, such as their diffusion, spillover, and reverse spillover, determines the activity and product selectivity of oxide-supported metal catalysts. Understanding these surface processes at the atomic level is critical, but challenging due to the limitations of conventional characterization techniques.
Figure 2. TPD spectra of the methoxy intermediates on the Pt/TiO2(110) surfaces with Pt particle densities of 5.4 ×1016 /m2 (H-Pt/TiO2(110), black curve) and 2.4 ×1016 /m2 (L-Pt/TiO2(110), red curves), respectively: (a) CO (m/z = 28), (b) CH4 (m/z = 16). (c) Estimated selectivity toward carbonaceous products (CO, CH4, H2CO, and CH3OH) for the H-Pt/TiO2(110) and L-Pt/TiO2(110) samples.
We have previously observed the dissociation of methanol molecules on Pt nanoparticles into methoxy intermediates, which then spilled over to the five-fold coordinated Ti4+ sites (Ti5c) of the TiO2(110) surface (see Fig. 1a). In this work, through temperature programmed desorption (TPD) experiments, we found that most of the methoxy intermediates were thermally decomposed to CO and CH4 above 350 K (see Fig. 2). The activity and product selectivity for methoxy decomposition depended on the Pt nanoparticle density. It suggests that the Pt nanoparticles acted as active sites and the decomposition property was controlled by the diffusion and reverse-spillover of the methoxy intermediates (see Fig. 1b).
Figure 3. (a)-(d) Diffusion of methoxy intermediates on the Pt/TiO2(110) surface at room temperature. The STM images were acquired 12 s per frame (5.0×5.0 nm2).
Then, how do the methoxy intermediates reach the Pt active sites? We performed sequential scanning tunneling microscopy (STM) imaging to monitor the movement of individual adsorbates. As is shown in Fig. 3, the methoxy intermediates diffuse along the [001] or 
direction and can thus move throughout the entire surface. Based on the visual evidence, we further investigated the atomistic details of the diffusion paths using density functional theory (DFT) calculations. Figure 4 shows the atomic model of diffusion in the [001] direction, as well as the energy diagram showing a reasonable energy barrier. The temporal formation of molecularly adsorbed methanol is the key to the easy diffusion along the [001] direction. As for the diffusion along the direction, apart from the formation of molecularly adsorbed methanol, the presence of an additional proton on the adjacent bridging oxygen is also necessary to reduce the energy barrier to a reasonable value. Therefore, we have resolved the diffusion pathways and proposed the model in which the methoxy intermediates diffuse and reach the Pt active sites for their decomposition (see the scheme in Fig. 1b).
Figure 4. (a)-(e) Selected configurations in the potential energy profile ((f)) illustrating the diffusion of methoxy species along [001] direction via a molecularly adsorbed methanol (“methanol-mediated model”).
This study successfully elucidates the complex elementary steps in heterogeneous catalysis at the atomic level. It shows how the intermediate adsorbates diffuse on the catalyst surface and reach the active sites to yield the corresponding products. It also provides new inspiration for catalyst design that activity and selectivity can be modulated by considering the surface diffusion.
Q2. 本研究テーマについて、自分なりに工夫したところ、思い入れがあるところを教えてください。
The direct STM visualization of the movement of the intermediate adsorbates on the catalyst surface is certainly one of the most attractive features of this study. Our group has a long history of research on methanol adsorption on the Pt/TiO2(110) surface, which can be traced back to our first paper published in Langmuir in 2010 (S. Takakusagi et al., Langmuir 26 (2010) 16392). Since then, we have devoted much effort to the systematic study of the preparation and observation of the methoxy intermediates. This study focused on the investigation of their dynamic behavior and decomposition property, which was realized by precisely controlling the experimental conditions and by combining various surface science techniques, such as TPD and XPS.
STM imaging of the behavior of the molecules at the atomic level naturally bridges the gap between experiment and theory by pinpointing the rational diffusion pathways for DFT calculation. In addition, the comparison between high and low densities of Pt nanoparticles effectively proves that the decomposition is controlled by diffusion and reverse-spillover, and the main active sites are the Pt nanoparticles. Imaging the diffusion of methanol on Pt/TiO2 catalyst helps to reveal more accurate decomposition mechanism. In this study, fundamental surface science approach showed its power and beauty by resolving a surface catalytic reaction at atomic level.
Q3. 研究テーマの難しかったところはどこですか？またそれをどのように乗り越えましたか？
Although STM is a well-established technique, obtaining high-resolution images on a routine basis is still a difficult and time-consuming task. First, high-resolution STM images require sharp and robust tungsten tips. However, the shapes of the tips prepared by electrochemical etching are very sensitive to the preparation conditions, resulting in a low yield. To make matters worse, the only effective way to check the quality of the tips is to use them in a condition similar to a practical experiment. Therefore, ensuring a perfect tip is time-consuming. Second, the scanning condition is also critical. Any surface contamination could alter the dissociative adsorption of methanol and the movement of the adsorbates. In addition, scanning parameters such as area, speed, and bias voltage were optimized to follow the movement of the methoxy intermediates. A large number of sequential images had to be taken to capture the moment of hopping. The TPD measurements of the methoxy intermediates on the Pt/TiO2(110) were also not easy because their coverage was rather small and high sensitivity detection of the desorption products was required. I performed them in collaboration with the group of Prof. Katsuyuki Fukutani (Univ. of Tokyo), an expert in TPD measurements, and then succeeded in obtaining the reliable results.
Q4. 将来は化学とどう関わっていきたいですか？
Chemistry is closely related to the chemical industry and has great value, but is ultimately complex. The theory of physical chemistry in homogeneous phase has been well developed over years. However, it is difficult to establish mechanism and theory for the complex surface reactions such as heterogeneous catalysis and electrocatalysis due to limited experimental evidence. Surface science techniques have the potential to solve this problem. The surface chemical processes can be studied at the atomic/molecular level and accurately compared with theoretical analysis. I will try my best to connect experiment and theory of surface chemistry in my future research career.
Q5. 最後に、読者の皆さんにメッセージをお願いします。
Repeated experimentation is tedious. Scientific discovery, on the other hand, is exciting. Hard work deserves success. In particular, I would like to thank my supervisor Professor Takakusagi for his dedicated support and guidance of the project, my associate supervisor Professor Asakura for his helpful discussions and suggestions, and all the collaborators for their efforts and contributions with professional skills and knowledge.
研究者の略歴
Can Liu 劉燦
Institute for Catalysis, Hokkaido University
2009.6, BEng in Materials Physics at University of Science and Technology Beijing
2014.7, PhD in Materials Science and Engineering at Tsinghua University; Supervisor Prof. Zhengjun Zhang; Research theme: Preparation and electrochemical property adjustment of molybdenum oxide film
2014.9 -2017.3 , Postdoctoral researcher at Institute for Catalysis, Hokkaido University; Supervisor Prof. Shen Ye; Research theme: In-situ studies on the electrode reactions of lithium-oxygen battery
2018.4 -present, Postdoctoral researcher at Institute for Catalysis, Hokkaido University; Supervisor Prof. Satoru Takakusagi; Research theme: Molecular adsorption and reaction on the model catalyst surfaces
