【作者】 张哲；

【导师】 施展；

【作者基本信息】 吉林大学 ， 无机化学， 2025， 博士

【摘要】 全球可持续发展面临的核心挑战之一是能源问题。寻找和开发清洁、高效的可再生能源以替代传统的化石能源已经成为了一项迫在眉睫的任务。水是地球上最重要、最丰富的自然资源之一,它不含任何污染元素和重金属元素,因此,利用水进行能源转换,即开发水系催化,被视为一种具有极大潜力的清洁能源替代方案。然而,在实际的能源转换过程中,通常需要投入较高的外部能量才能进行有效的反应,这种能量损失极大地限制了能源转换效率。为了解决这个问题,需要开发新型的高性能催化剂来降低这些反应的能量门槛。共价有机框架（COFs）是一种具有巨大潜力的催化剂,它具有结构明确、可调性高、比表面积大以及稳定性好等优点。然而,目前催化剂的研发过程主要依赖于实验试错的方式,这种方法既耗费时间和精力,又难以全面揭示材料的全部潜力。为了克服这一挑战,我们从两个关键领域着手:首先,我们希望创新催化剂的开发策略,结合理论计算和人工智能等新方法,以提高研发效率;其次,我们深入研究并归纳出具有普适性的催化反应机制,这将为催化剂的设计和优化提供清晰的指导。这两个方向是推动这类材料从基础研究走向实际应用的关键。具体研究内容如下:首先,针对现有材料开发方式的局限性,提出了一种基于催化单元的材料开发新方法。该方法将传统的实验方法与理论计算、机器学习等新技术有机结合,将材料开发的过程简化为类似组装积木的过程,这不仅可以快速地设计出适合特定工况的催化剂,还可以对大量的筛选数据进行数据挖掘,提取出具有普适性的描述符和化学见解,最大程度地发挥了数据的价值。为了验证这一方法,我们对一系列双原子Salphen基催化剂进行计算筛选,随后使用机器学习方法对数据进行挖掘。我们将筛选出的优秀催化单元组装到共价有机框架的骨架中,并通过大量的表征技术证实了其结构的准确性。在电化学测试中,合成的催化剂具有很高的氧析出效率和稳定性。作为对比,我们还合成了理论预测效果不佳的催化剂,实验结果表明,这些催化剂的性能与理论预测高度一致。这一结果进一步证明了基于双原子单元筛选方法的有效性。此外,我们还探讨了基于配位单元的催化剂选择方法可能存在的问题。通过这种方式,能够更深入地理解这些催化剂的工作原理,从而为未来的催化剂设计提供更多的启示。其次,我们从计算出发,对氧还原反应的潜在机制进行了探索。通过基于密度泛函理论的第一性原理计算,首次揭示了随着配位电子结构对称性的降低,d-π相互作用的强度会增大,从而提升d_xz/d_yz轨道的能级,增强对中间体的吸附。相比于传统的卟啉或酞菁基和石墨基热解型催化剂,具有较低对称性的过渡金属-共价有机框架（TM-COFs）在表现出了最强d-π的相互作用,并且显示出了最高的氧还原反应活性。这种设计具有高度的可调性,可以通过改变d电子和π电子修饰进行双重调节。此外,我们还发现,当在苯环上引入氧取代基时,可以有助于过渡金属中d轨道的重排,从而提升了它们的自旋,增强了对^*OOH中间体的吸附,进一步提升了催化性能。为了验证这一理论预测,我们在实验室中合成了一系列的TM-COF结构,并对其进行了氧还原性能的测试。值得一提的是,在碳纳米管上原位生长的Co-COF-O表现出了优异的氧还原性能。它具有高半波电位（0.83 V vs.RHE）、低Tafel斜率(38.8 mV dec^-1)以及优异的4电子转移选择性（95%）,显示出了极高的商业应用潜力。这一工作增强了对d-π相互作用对过渡金属基催化剂性能影响的理解,并为设计高效的ORR催化剂提供了新的见解和启示。最后,我们从实验出发,揭示了堆积工程对共价有机框架光能转换能力的影响。我们设计并制备了四种具有相同主链但堆积方式不同的共价有机框架,以阐明共价有机框架堆叠工程对光子能量转换影响的重要性。这些共价有机框架的主链是通过四（4-氨基苯基）芘和2-羟基间苯二甲醛的席夫碱缩合反应形成的,展示出sql拓扑链的特性。通过对这些一维sql拓扑链边缘顶点的修饰和连接,成功地调节了链间距离和堆叠强度,从而得到了一系列不同链间距离的共价有机框架材料,包括1D-COF,1D-Me COF和1D-tBuCOF。我们还通过共价键连接相邻链的顶点,制备了2D-COF。这些COFs在化学组成、连接和孔隙环境上具有相似性,只是堆叠方式不同,因此它们为研究共价有机框架堆叠工程对光子能量转换影响提供了优秀的模型。实验结果表明,这些主链相同但堆叠方式不同的共价有机框架在光催化析氢反应中表现出显著的性能差异。我们发现,堆叠方式对共价有机框架的光物理性质有显著影响。其中,平面内1D堆叠的共价有机框架通过光生内建电场促进载流子在不同链间迁移,实现了“自催化”效应。此外,在堆叠中引入适当的榫卯结构可以增强π晶格的连续性,从而提高光吸收能力和载流子输运能力。这一研究结果增强了对堆叠结构-性能关系的理解,并为共价有机框架基光能转换材料的开发提供了有效的设计策略。更多还原

【Abstract】 One of the core challenges facing global sustainable development is the energy issue.The search for and development of clean and efficient renewable energy sources to replace traditional fossil fuels has become an urgent task.Water is one of the most important and abundant natural resources on Earth.It contains no pollutants or heavy metals,and thus,the utilization of water for energy conversion,namely the development of aqueous catalysis,is regarded as a highly promising alternative for clean energy.However,in the actual process of energy conversion,we often need to input a significant amount of external energy to achieve effective reactions.This energy loss greatly limits the efficiency of energy conversion.To address this problem,we need to develop new high-performance catalysts to reduce the energy barriers of these reactions.Covalent organic frameworks（COFs）are a type of catalyst with great potential,characterized by their well-defined structure,high tunability,large specific surface area,and good stability.However,the current development of catalysts mainly relies on experimental trial-and-error methods,which are not only time-consuming and labor-intensive but also fail to fully reveal the potential of the materials.To overcome this challenge,we focus on two key areas:First,we aim to innovate the development strategies for catalysts by integrating new methods such as theoretical calculations and artificial intelligence to improve the efficiency of research and development.Second,we conduct in-depth studies to identify universal catalytic reaction mechanisms,which will provide clear guidance for the design and optimization of catalysts.These two directions are crucial for advancing these materials from fundamental research to practical applications.First,in response to the limitations of existing material development methods,we propose a new material development method based on catalytic units.This method organically combines traditional experimental methods with new technologies such as theoretical calculation and machine learning,simplifying the material development process to a process similar to assembling building blocks.This not only allows us to quickly design catalysts suitable for specific conditions,but also allows us to mine data from a large amount of screening data,extracting universal descriptors and chemical insights,maximizing the value of data.To verify this method,we calculated and screened a series of bimetallic Salphen-based catalysts,and then used machine learning methods to mine the data.We assembled the excellent catalytic units screened out into the skeleton of the covalent organic framework and confirmed its structural accuracy through a large number of characterization techniques.In electrochemical tests,the synthesized catalysts showed high oxygen evolution efficiency and stability.For comparison,we also synthesized catalysts that were predicted to perform poorly in theory.The experimental results showed that the performance of these catalysts was highly consistent with the theoretical predictions,further proving the effectiveness of our method based on bimetallic unit screening.In addition,we also explored the potential problems of catalyst selection methods based on coordination units.Through this approach,we can gain a deeper understanding of the working principles of these catalysts,providing more insights for future catalyst design.Second,we started from computation and explored the potential mechanism of the oxygen reduction reaction.Through first-principles calculations based on density functional theory,we revealed for the first time that as the symmetry of the coordination electron structure decreases,the strength of the d-πinteraction increases,which raises the energy level of the dxz/dyz orbitals and enhances the adsorption of intermediates.Compared with traditional porphyrin or phthalocyanine-based and graphite-based heat-bonded catalysts,we found that transition metal-covalent organic frameworks（TM-COFs）with lower symmetry showed the strongest d-πinteraction and displayed the highest oxygen reduction reaction activity.This design has high adjustability and can be doubly regulated by changing d-electron andπ-electron modifications.In addition,we found that introducing oxygen substituents on the benzene ring can contribute to the rearrangement of d-orbitals in transition metals,thereby enhancing their spin,strengthening the adsorption of the^*OOH intermediate,and further enhancing catalytic performance.To verify our theoretical predictions,we synthesized a series of TM-COF structures in the laboratory and tested their oxygen reduction performance.Notably,the Co-COF-O grown in situ on carbon nanotubes exhibited excellent oxygen reduction performance.It has a high half-wave potential（0.83 V vs.RHE）,a low Tafel slope(38.8 mV dec^-1),and excellent 4-electron transfer selectivity（95%）,showing great commercial application potential.This work enhances our understanding of the impact of d-πinteractions on the performance of transition metal-based catalysts and provides new insights and inspirations for the design of efficient ORR catalysts.Finally,from an experimental perspective,we revealed the impact of stacking engineering on the photon energy conversion ability of covalent organic frameworks.We designed and prepared four types of covalent organic frameworks with the same main chain but different stacking methods,to clarify the importance of covalent organic framework stacking engineering on photon energy conversion.The main chains of these covalent organic frameworks are formed by the Schiff base condensation reaction of tetra（4-aminophenyl）naphthalene and 2-hydroxyterephthalaldehyde,showing the characteristics of the sql topological chain.We successfully adjusted the chain spacing and stacking strength by modifying and connecting the edge vertices of these one-dimensional sql topological chains,thereby obtaining a series of covalent organic framework materials with different chain spacings,including 1D-COF,1D-Me COF,and 1D-tBuCOF.We also prepared 2D-COF by connecting the vertices of adjacent chains with covalent bonds.These COFs have similar chemical composition,connections,and pore environments,only the stacking methods are different,so they provide excellent models for us to study the impact of covalent organic framework stacking engineering on photon energy conversion.Experimental results show that these covalent organic frameworks with the same main chain but different stacking methods show significant performance differences in photocatalytic hydrogen evolution reactions.We found that the stacking method has a significant impact on the photophysical properties of covalent organic frameworks.Among them,the covalent organic framework with in-plane 1D stacking promotes the migration of carriers between different chains through the photogenerated built-in electric field,realizing the"self-catalysis"effect.In addition,we found that introducing appropriate mortise and tenon structures in the stacking can enhance the continuity of theπlattice,thereby improving the light absorption ability and carrier transport ability.This research result enhances our understanding of the relationship between stacking structure and performance,and provides effective design strategies for the development of covalent organic framework-based photon energy conversion materials.更多还原