目前在单分子电学特性的高效表征、海量数据的智能解析和分子体系的合成设计等方面的研究模式依旧是以科研人员的人为操作为主。这一劳动密集型研究模式费时费力、效率低下。近年来,由于计算能力、大数据和机器学习算法的不断进步,人工智能技术以其强大的学习能力和迭代能力、为传统研究模式的革新带来了新的机遇。综合以上挑战和机遇，本方向致力于建立基于AI驱动的单分子电子学闭环研发模式。面向分子电子学，通过研究智能表征仪器建立AI驱动的高质量数据产生方法；通过研究智能分析算法建立AI驱动的数据挖掘方法；通过研究智能解析模型探索AI驱动的智能决策方法。
At present, research models in the efficient characterization of single-molecule electrical properties, intelligent interpretation of massive data, and synthetic design of molecular systems, are still mainly based on manual operations by researchers. This labor-intensive research model is time-consuming, laborious, and inefficient. In recent years, with the continuous advancement in computing capabilities, big data, and machine learning algorithms, artificial intelligence technology offers new opportunities for innovation in traditional research models due to its strong learning and iterative capabilities. Combining these challenges and opportunities, this direction is committed to establishing an AI-driven closed-loop research model for single-molecule electronics. In the field of molecular electronics, we aim to establish AI-driven methods for high-quality data generation by studying intelligent characterization instruments; establish AI-driven data mining methods by studying intelligent analysis algorithms; and explore AI-driven intelligent decision-making methods by studying intelligent interpretation models.
单分子器件微纳制造与阵列集成
微/纳米制造技术极大地推动了分子电子学的发展，基于无噪声超精密加工与表征实验室，我们发展了可达亚3纳米精度的基于电子和离子等高能粒子束的微纳加工方法，实现了可达亚皮米级电极操纵空间分辨率的可控裂结技术，从而显著提升分子结制备的成功率和稳定性，基于此我们致力于实现具有复杂结构和电路的分子电子器件乃至阵列集成。
我们旨在探索一系列可用于单分子尺度光、电、热等测量实验的微纳芯片。我们首创性地引入弹性金属基底结合无机氧化物绝缘/牺牲层的机械可控裂结芯片工艺路线、实现了石墨烯/分子/石墨烯场效应晶体管器件芯片制备、首创基于MEMS工艺的超高微量热分辨率基底与针尖芯片、开发了将驱动部件集成至芯片上的片上可控裂结芯片技术，利用上述芯片的稳定悬停性能，结合电学、光谱学、噪声谱学、热导联用表征技术，可以对分子结结构、反应始终态和中间体进行表征，构建分子结结构与电输运和声子性质之间的关系，进一步基于上述分子电子器件性质开发具有逻辑运算、类脑计算功能的分子电子器件阵列集成工艺。
Micro/nano manufacturing technologies have greatly promoted the development of molecular electronics. Based on the noise free ultra-precision manufacturing and characterization laboratory, we have developed micro/nano fabrication methods based on high-energy particle beams such as electrons and ions with sub-3 nanometer accuracy, and realized controllable single-molecule break junction technologies with spatial resolution of subpicmeter-level electrode manipulation. Thus, the success rate and stability of molecular junction are significantly improved, based on which we are committed to realizing the integration of molecular electronic devices and even arrays with complex structures and circuits.
We aim to explore a series of micro-chips that can be used for optical, electrical and thermal measurement experiments on a single-molecule scale. We first introduced the process route of mechanically controllable break junction chips with elastic metal substrate combined with inorganic oxide insulation/sacrifice layer, realized the chip preparation of graphene/molecular/graphene field effect transistor devices, pioneered the ultra-micro thermal resolution substrate and tip chip based on MEMS process, and developed the on-chip break junction chip technology that integrates the actuating components onto the chip. By using the stable molecular junction hover performance of the above chip, combined with the characterization techniques of electricity, spectroscopy, noise spectroscopy and thermal conductance, the molecular junction structure, reaction state and intermediate can be characterized, and the relationship between the molecular junction structure and electrical transport and phonon properties can be constructed. Based on the above properties of molecular electronic devices, a molecular electronic device array integration process with logic operation and brain-like computing functions is further developed.
“自下而上”的原子制造
利用“自下而上”的化学和生物组装技术来制造复杂器件，已成为当前纳米科学和技术领域的重要研究方向。针对目前单分子器件传统制备工艺存在的稳定性差、成品率低和难以集成等一系列问题，我们拟引入“自下而上”组装技术。该技术有望提高器件可靠性及集成度，并探索如何精准操控原子或分子以实现纳米级别的可控制造。同时，与传统的“自上而下“的微纳加工技术进行结合，开展基于原子与超限制造的系统集成与应用研究，以此打破全尺度芯片制造的关键技术壁垒，为单分子器件的制备开辟一条新道路，进而推动其在半导体领域的广泛应用。
Manufacturing complex devices using "bottom-up" chemical and biological assembly techniques has become a significant research direction in the field of nanoscience and technology. In order to address the issues of poor stability, low yield, and difficult integration faced by traditional fabrication processes for single-molecule devices, we propose to introduce "bottom-up" assembly methods. This strategy has the potential to enhance device reliability and integration, while exploring how to precisely manipulate atoms or molecules for nanomanufacturing. At the same time, by integrating these methods with the traditional "top-down" micro-nano-processing technology, we can conduct research on system integration and applications based on atom-scale and ultra-precision manufacturing. This approach aims to overcome the key technological barriers in manufacturing chips at all scales, opening up a new path for the preparation of single-molecule devices, and promoting their application in the semiconductor field.
单分子自旋量子器件
自旋电子学是在传统微电子学电荷自由度的基础上考虑自旋自由度进行传感、信息存储、传输和处理，可大幅提高数据处理速度、降低电力消耗和提高集成密度。我们以单分子为研究载体，聚焦于单分子器件中电子自旋自由度的测量和操纵，包括自旋量子输运、自旋注入和探测、自旋弛豫、相干性操纵等诸多科学问题，以此来为单分子自旋量子器件的应用和制备提供理论依据。
Spintronics involves the consideration of spin degrees of freedom in addition to the charge degrees of freedom in traditional microelectronics for sensing, information storage, transmission, and processing. Spintronics can greatly improve data processing speed, reduce power consumption, and increase integration density. We focus on the measurement and manipulation of electron spin degrees of freedom in single molecule devices, using single molecules as the research carrier. This involves a broad range of scientific issues such as spin quantum transport, spin injection and detection, spin relaxation, and coherence manipulation. Our goal is to provide theoretical foundation for the application and preparation of single molecule spin quantum devices.
类脑计算
随着人工智能时代的到来，信息量爆炸式增长对计算机算力提出了巨大的挑战，传统冯·诺依曼计算架构将存储与运算单元分离，导致处理大规模数据时消耗巨大的能量并且耗时更长。近年来，神经形态计算作为一种新技术崭露头角，有望替代传统计算机，执行人工智能任务。生物突触是大脑学习和记忆的基本单元。我们致力于开发基于单个分子、单分子层和分子薄膜的可调谐多电导态忆阻器，以模拟突触的结构和功能。这有助于实现纳米尺度下高运算速度和低能耗的非易失性存储器开发。进一步地，我们将通过微纳加工技术集成高性能存算一体的分子芯片，实现基于神经网络算法的类脑计算。
With the advent of the artificial intelligence era, the explosive growth of information has posed a huge challenge to computing power. The traditional von Neumann computing architecture separates storage and computing units, leading to significant energy consumption and longer processing times for large-scale data. In recent years, neuromorphic computing has emerged as a new technology with the potential to replace traditional computers and perform artificial intelligence tasks. Biological synapses are fundamental to learning and memory in the brain. We are committed to developing tunable multi-conductivity memristors based on single molecules, single molecular layers, and molecular thin films to mimic synapse structure and function. This will aid in developing non-volatile memory with high computational speed and low energy consumption at the nanoscale. Additionally, we will integrate high-performance memory and computing into molecular chips using micro/nano processing technology to achieve neuromorphic computing based on neural network algorithms.
逻辑运算
单分子晶体管器件是未来分子逻辑电路的核心单元，理解单分子晶体管器件的运行机制，发展高性能单分子晶体管器件对实现分子逻辑运算乃至分子计算机具有重要的意义。如何系统表征单分子晶体管的电学特性仍存在较大挑战。针对上述问题，我们首先通过单分子裂结测试技术与电化学门控技术相结合，构筑单分子电化学晶体管，实现单分子沟道的能级分布和化学结构的高效调控，通过解析单分子沟道电学性质对电极电势的响应规律，评估单分子沟道的电子传输机制与场效应特性，以此筛选性能优异的单分子半导体功能材料。在获得性能优异的单分子材料后，我们将进一步构筑固态单分子器件，探索器件在直流电压运行过程的性能参数，并结合单分子阻抗与矢量网络分析仪等器件高频电学表征技术，揭示单分子电阻、晶体管等基本电子元件的高频电学特性，以此推动单分子器件的实用化进程。
Single-molecule transistors are the core units of future molecular logic circuits. Understanding the operating mechanisms of single-molecule crystal transistor devices and developing high-performance single-molecule transistor devices have significant implications for realizing molecular logic operations and even molecular computers. However, how to systematically characterize the electrical properties of single-molecule transistors remains a significant challenge.
In response to these issues, we fabricate the single-molecule electrochemical transistors by combining single-molecule break junction technique with electrochemical gate, which allows the efficient manipulation of the chemical structures and energy level alignment of the single-molecule channels. By analyzing the response of the single-molecule conducting channel to electrode potential, we evaluate the field-effect performance and electron transport mechanism of the single-molecule channel, so as to select the high-performance single-molecule semiconductor materials.
After obtaining high-performance single-molecule materials, we will further construct solid-state single-molecule devices, and explore the key parameters of device performance under the DC voltage. Moreover, we will combine high-frequency electrical characterization techniques, such as single-molecule impedance and vector network analysis, to reveal the high-frequency electrical properties of basic electronic components like single-molecule resistors and transistors. Through the above approaches, we aim to advance the practical application of single-molecule devices.
单分子发光谱学与技术
单分子发光谱学与成像技术旨在发展单分子尺度下超快、高灵敏且高时空分辨的光谱技术。面向电化学与能源化工领域，在工况条件下研究单分子与表界面作用的动态行为、分子间相互作用与反应动力学、分子输运和能量传递等单分子尺度下的“三传一反”过程与原理。目标发展与建立纳微尺度下的传递过程新机制与模型，为智能与高效化工新过程开发提供理论基础与方向。
另一方面，发展电致单分子发光技术以实现在纳米尺度下光子与物质相互作用的机制及应用研究。将纳米光子学与量子信息领域相结合，为量子态的制备与量子信息器件的设计与集成提供变革性技术基础。
Single molecular spectroscopy and imaging techniques aim to develop ultrafast, highly sensitive and spatiotemporal-resolved spectroscopy techniques at the individual molecules level. Toward the field of electrochemistry and energy chemical industry, operando single molecular spectroscopy is applied to investigate the dynamic behaviors between single molecule and interface, the interaction and reaction kinetics of molecules, molecular transport and energy transfer under working conditions. We aim to develop innovative mechanisms and models of transport phenomena at the nano-scale, and to provide theoretical basis and direction for the development of novel intelligent and efficient chemical processes.
On the other hand, the electro-induced single molecule luminescence technology is developed to understand the interaction between photon and matter toward regulation and control of the behavior of single photon. The combination of nano-photonics and quantum information provides a transformative technical basis for the preparation of quantum states and the fabrication of quantum information devices.
超快单分子电子学
单分子尺度金属纳米腔中的光与物质相互作用研究能为单分子光电子学、量子光学、分子光化学等前沿领域带来新的认知。我们聚焦于飞秒脉冲研究金属-分子-金属结这一复杂体系中的光与单分子结相互作用以及动力学机制。自主搭建了锁相检测技术的超快裂结系统（包括扫描隧道裂结技术以及机械可控裂结技术）使我们能够通过检测瞬时光电流的方式获得光与单分子结相互作用的信息。结合光学泵浦-探测技术能够将如今单分子电子学的研究时间尺度扩展至飞秒时间尺度。目前，我们的研究重点是理解界面光电输运过程，单分子与等离激元耦合的动力学机制，以及光如何操纵分子的电输运。
Investigating the interaction between light and matter in metal nanocavities at single-molecule-scale holds the potential to advance our understanding in the forefront areas of single-molecule optoelectronics, quantum optics, and molecular photochemistry. We focus on probing the interaction and dynamics between light and single-molecule junctions within the complex system of metal-molecule-metal structures using femtosecond pulse techniques. Through the customer-designed lock-in detection-based break junction techniques (including scanning tunneling microscope break junction and mechanically controllable break junction), we are able to acquire information about the interaction between light and single-molecule junctions by detecting transient photocurrents. Combined with optical pump-probe techniques, the research extends the time scale of single-molecule electronics studies to the femtosecond regime. Currently, our research is primarily aimed at understanding the processes of interface optoelectronic transport, the dynamic mechanisms of coupling between single molecules and plasmons, and how light manipulates the transport through single-molecule junctions.
单分子生物电子学
单分子技术的出现使得监测单个蛋白质的动态变化过程成为可能，并且为探索和阐明生命活动中新现象和新机理提供了一个有力的工具。在这些单分子技术中，STM-BJ技术能够在短时间内快速构建成千上万个单分子结来捕获单个分子的动态变化过程，并以电信号的形式呈现，具有重复性好、时间精度高的特点。因此，我们小组主要采用STM-BJ的技术去分析生命过程中分子内和分子间的相互作用；捕捉动态过程中的中间态和过渡态；检测他们的动态行为；捕获微秒尺度间的构象变化，阐明单分子尺度下酶催化反应的机理。在不久的未来，试图通过单分子手段解析机制并指导酶的理性设计和改造。
Single-molecule technique makes monitoring the dynamical changes of protein at the single-molecule level a reality, which also provides a powerful tool for the further exploration of new phenomena and new mechanisms of life activities. Among them, the STM-BJ technique is achieved by repeatedly and rapidly forming thousands of molecular junctions, by which to capture the changes of single-molecule incisively. Thus, our group focuses on employing the STM-BJ technique to analyze intramolecular and intermolecular interactions, capture intermediate/transition states, observe the moving behaviors, and conformational changes in a microsecond, and elucidate the mechanism of enzyme-catalyzed reaction in the nanoscale. In the future, the enzyme catalytic mechanism analysis from single-molecule conductance would guide the rational design and modification of the enzyme.
单分子传感器
发展结构小型化、功能智能化的高密度传感器已成为学术与工业界的共同追求，同时也对极限灵敏度传感技术的研发提出了新的挑战。基于原子制造的单分子电学表征技术，以隧穿电流为检测信号，能够在单分子水平上直接对目标分析物进行精确识别和定量检测。这一方法具有高选择性、极致灵敏度、免标记和快速响应的原理性优势，在传感领域具有巨大的应用潜力。通过前期研究，我们已经验证该方法在不同物质检测中具有广泛的适用性，可以实现对酶（辅酶NADH）、环境污染物（有机磷杀虫剂）、爆炸物（TNT等）和金属离子等在超宽线性范围的定量检测，检测限低至aM（10-18 mol/L）水平。未来，针对环境污染物监测、生命健康诊断以及电子化学品质量评估等领域所面临的复杂性与动态监测难题，我们将从实际检测需求出发，进一步研发单分子电学智能化传感表征平台，以期在单分子层面上实现更为精准、快速的检测，从而推动相关领域的技术革新和应用拓展。
The development of high-density sensors with miniaturized structures and intelligent functions has become a common pursuit of academia and industry, while posing new challenges for the research and development of ultra-sensitive sensing technology. The single-molecule electrical characterization technique based on atomic manufacturing, which leverages tunneling current as a detection signal, enables precise identification and quantitative detection of target analytes at the single-molecule level. This method has theoretical advantages of high selectivity, ultimate sensitivity, label-free detection, and rapid response, promising significant application potential in the realm of sensing. Our preliminary research has substantiated the broad applicability of this method in detecting various substances. It has achieved quantitative detection of enzymes (specifically coenzyme NADH), environmental pollutants (such as organophosphorus pesticides), explosives (e.g., TNT), and metal ions within an ultra-wide linear range, reaching a detection limit as low as aM (10-18 mol/L). In the future, aiming at the complexity and dynamic monitoring challenges faced by environmental pollutant monitoring, life health diagnosis, and electronic chemical quality assessment, we will further develop a single-molecule intelligent sensing and characterization platform based on actual detection requirements. We hope to achieve more precise and rapid detection at the single-molecule level, thereby promoting technological innovation and application expansion in related fields.
