不同形貌和结构的纳米颗粒具有不同的物理化学性质，使其在光学、催化、电磁、生物等多个领域有着广阔的应用前景。实现纳米材料工业应用的关键在于如何可控地合成特定形貌和结构的颗粒并进行规模化生产，所以形貌和结构调控是纳米材料科学领域的重要课题。经过几十年的努力，科学家们开发出了气相沉积、沉淀法、溶胶-凝胶法、溶剂热法等多种纳米颗粒合成方法，并通过模板法和表面包覆等技术实现对颗粒形貌的有效控制。然而，传统结晶学理论大多基于热力学原理，而实际的纳米颗粒合成大都是在远离热力学平衡态下进行的，受动力学因素的影响更大，这就导致传统结晶理论无法解释和预测很多新的结晶现象，以及缺少调控纳米材料形貌的通用理论和方法。 人们逐渐认识到动力学过程对颗粒形貌演化的重要性，发现物质的反应和扩散相互作用是导致一些复杂的时间和空间结构的关键因素，并使用反应-扩散来解释某些特殊斑图的形成过程，发展了扩散限制聚集和反应限制聚集的基本动力学模型。本课题组从化学工程的角度出发，提出了反应-扩散调控颗粒形貌的新方法，实现了对以碳酸钙和贵金属为主的多种颗粒材料的有效调控，并初步获得了反应-扩散对颗粒形貌的调控规律。然而，我们前期的工作只是停留在定性研究的阶段，缺乏关键的动力学模型参数，还未形成普适性的调控机制。 所以，本文在组内前期工作的基础上，以银纳米颗粒为主要研究对象，开展反应-扩散对颗粒形貌的定性和定量调控，并进一步探索对流传质（混合）的调控作用，建立基于反应-传质（包括扩散和对流）的普适性和通用性颗粒生长机制。本文主要有以下三点发现： （1）反应-扩散对银颗粒形貌的定性调控规律。运用传统的溶液还原法制备了形貌均一的枝状银颗粒，改变反应温度和反应物浓度，发现形貌变化的规律性很差，并分析了常规实验参数对形貌调控作用的局限性。通过改变溶液酸度和粘度分别对反应和扩散进行定性调控，发现银枝晶生成于快反应条件下，随着反应下降，侧枝的生长受到抑制，结构变得紧凑乃至最终侧枝消失；而随着扩散降低，颗粒尺寸明显减小；晶向分析表明扩散也影响枝晶的生长方向，快扩散提供充足的单体，满足高能晶面和低能晶面同时生长，扩散减慢使得单体供应速率下降，仅能维持高能晶面优先生长。良好的演变规律，证明了反应-扩散对颗粒形貌调控的有效性。 （2）反应-扩散对银颗粒形貌的定量调控规律和颗粒生长的浓度场机制。利用电化学沉积方法制备了银颗粒，分别通过改变外加电位和电解液溶剂粘度来调控反应和扩散，并采取循环伏安法和计时电量法测定了Ag+/Ag电对的标准反应速率常数和Ag+的扩散系数，发现随着粘度增加，除了扩散系数降低以外，标准速率常数也会因介电弛豫效应而有所减小，只不过变化幅度比扩散系数小；根据Butler-Volmer关系，计算了每种溶剂中不同电位下的表观反应速率常数，随着电位降低，反应呈指数增加，银颗粒均呈现出“多面体-枝晶-纳米晶”的三段式演化规律；对比不同溶剂体系，随着粘度增加，扩散降低，枝晶起点电位处的颗粒侧枝生长更充分，而枝晶终点电位有所提前。接着，引入达姆科勒准数（Damko?hler number，Da），量化了反应和扩散的相对变化，发现几乎所有枝晶都存在于一个固定的Da区间。根据Da与颗粒形貌的对应关系，提出了基于微观浓度场的生长机制，较小的Da代表扩散控制，在晶核附近形成的浓度梯度诱导生长界面Mullins-Sekerka（MS）失稳，从而演化成为枝晶；而较大和极小的Da所对应的反应受限和极度扩散受限，都不能产生枝晶生长的必要浓度梯度。设计脉冲电沉积实验证实了微观浓度场的存在，验证了该机制的合理性，并在其他溶剂体系以及电沉积铜和金颗粒中发现同样的形貌演化规律，证明了反应-扩散调控方法的普适性。 （3）对流传质（混合）对银颗粒形貌的调控规律。在间歇体系中用溶液还原法制备银颗粒，改变前驱体和还原剂的混合比例与混合顺序，发现银枝晶结构发生了改变甚至是消失，说明对流传质对形貌结构也有重要影响。在T型混合器中进行连续反应，通过控制前驱体和还原剂的流率来精确调控混合强度，以Villermaux-Dushman方法测得的混合时间来定量表征混合条件，发现缓慢混合下生成片状银颗粒，快速混合产生枝状银颗粒；结晶性分析表明片状颗粒的裸露晶面为 (111) 面；从由分析离心技术得到的产物分布特性看，颗粒的当量尺寸及其分布宽度都随着混合强度增加而减小。得出的结论是：低混合强度形成接触不充分的反应混合物，低的有效过饱和度导致低的银源转化速率，高能晶面优先生长，最稳定的 (111) 面保留下来形成片状结构；而高混合强度形成接触充分的反应混合物，高的有效过饱和度导致快速成核以及银在晶核上的快速沉积，形成微观浓度梯度，诱导枝晶演化；混合强度增加还使得反应混合物分布更均匀，更均匀的生长环境使得产物尺寸差异变小、分布变窄。由此，通过对流传质效应拓展了反应-扩散生长机制，建立了适用于连续生产的反应-传质调控颗粒形貌的普适性方法。;The diversity of morphologies and structures equips nanoparticles with a lot of unique physical and chemical properties, which enable nanomaterials to show promising applications in various fields, like optics, catalysis, electromagnetics, biology and so forth. Controllable synthesis and large-scale production of particles with specific shapes are the key to turn those potential applications into reality, so shape control is an absolutely essential issue in the nanoscience. Vapour deposition, precipitation, sol-gel synthesis and solvothermal methods have been developed for preparing nanoparticles in the past decades. Also, the particle shapes have already been effectively regulated by templates and surface capping. However, those conventional methods of shape control are mainly confined on the thermodynamic aspect, while an overwhelming majority of particle growth occurs far from the thermodynamic equilibrium and is dominated by kinetic factors, which result in the severe shortage of accesses to the shape evolution. Thus, a general theory and a universal methodology have not yet been built up for the manipulation of particle morphologies and structures. Based on the existed crystallization theories, material scientists were aware of the significance of kinetic processes for shape evolution. It was found that the interaction between chemical reaction and diffusion often leads to the emergence of intricate spatial and/or temporal structures. Reaction and diffusion were applied to explain the formation of some special patterns, and subsequently were developed two basic kinetic models, called diffusion-limited aggregation and reaction-limited aggregation. Therefore, by combining the former achievements and principles of chemical engineering, our research team proposed a new reaction-diffusion-based strategy to control the particle shapes, which preliminarily showed its power in shaping calcium carbonate and noble metal particles. In spite of nice regularities, our previous work was restrict to qualitative investigations for the parameters of kinetic models were not achieved, and thus general mechanisms of shape manipulation could not be set up. Accordingly, taking silver particles as the technically relevant model system, both qualitative and quantitative researches of shape regulations by the reaction-diffusion protocol are reported in this thesis. The effects of convection (mixing) are included as well, and a general reaction-mass transfer based mechanism of particle growth is ultimately founded. The main findings are shown in more detail as follows: (1) Regulation of silver particle shapes by qualitatively manipulating reaction and diffusion. A traditional solution-based reduction was applied to synthesize monodispersed silver dendrites. Reaction temperatures and chemical concentrations were initially adjusted, but no regularity was obtained for the shape variations, which indicated the limitation of shape manipulation by simply altering experimental parameters. Then, the acidity and solvent viscosity were adopted to regulate reaction and diffusion, respectively. It was found that dendritic structures were produced under fast reaction conditions. When reaction was reduced, the branches were suppressed, became compact and finally vanished. Diffusion depression brought about the decrease of particle sizes and also changed the orientations of branched structures. Sufficient monomer supply by quick diffusion favoured the simultaneous growth of both high-energy and low-energy facets, while reduced diffusion could only support the preferential growth of high-energy facets. Above discoveries implied the importance and necessity of the reaction-diffusion protocol for shaping particles. (2) Regulation of silver particle shapes by quantitatively manipulating reaction and diffusion. The electrochemical method was employed to deposit silver particles, in which reaction and diffusion were modulated by varying the applied potential and the solvent viscosity of the electrolyte, respectively. Reaction and diffusion were separately depicted by the standard rate constant of Ag+/Ag couple and the diffusion coefficient of Ag+, which were derived from cyclic voltammetry and chronocoulometry, respectively. These two parameters were coupled and both of them decreased with increasing solvent viscosity, merely a smaller declining amplitude for standard rate constant than diffusion coefficient. With negatively shifting the potential, the apparent rate constant calculated by Butler-Volmer formula increased exponentially and silver particles underwent a “polyhedron-dendrite-nanocrystal” shape variation in each solvent system. A transverse comparison of results from different solvent systems indicated that diffusion reduction induced by viscosity rise promoted better structural development at the starting potential of dendrites while made the ending potential of dendrites arise in advance. Then, the Damko?hler number (Da) was introduced to quantify the diffusion and reaction processes during particle growth and to relate reaction-diffusion conditions and particle shapes, and it was found that almost all dendrites were produced in a fixed Da range. A Da-based concentration field mechanism of particle growth was put forward: diffusion control with small Da values generated local concentration gradients, which triggered Mullins-Sekerka (MS) instability on the growing interfaces and in turn favoured dendritic evolution; howbeit, large or extremely small Da and the accompanying reaction control or severe diffusion limitation could not provide the necessary concentration gradients for dendritic growth. Afterwards, the existence of local concentration fields and the rationality of the growth mechanism were verified by carefully designed pulsed electrodepostion. In addition, the same tendency of shape variation was achieved in more solvent systems and in other material systems (copper and gold), which implied the good generality of the reaction-diffusion method. (3) Regulation of silver particle shapes by manipulating convection (mixing). Batch experiments with solution-based reduction were initially employed to synthesize silver particles, and the dendritic structures were changed and even faded away via applying different mixing ratios and injection procedures of precursor and reducing agent solutions, indicating convention has significant effects on particle shapes. Then, continuous reactions were carried out in a T-mixer, and the mixing intensity quantified by previously determined mixing time using the Villermaux-Dushman protocol. It was found that slow mixing brought about plate-like particles, whose planar surfaces were (111) facets according to the crystallinity analysis, while fast mixing led to quite dendritic structures. Both the equivalent sizes and the widths of equivalent size distributions derived from analytical centrifugation decreased with improving the mixing quality. It was concluded: poor mixing led to low effective supersaturation, which gave rise to a preferential growth of high-energy facets of face-centered cubic crystals, and thus plate-like particles were generated; on the contrary, good mixing led to fast nucleation rates, and accordingly microscopic concentration gradients formed around the nuclei resulting in dendritic products; and better mixing quality led to smaller and more uniform particles. Therefore, a more general growth mechanism was achieved by incorporating mixing effects into the reaction-diffusion mechanism, and a kinetic embryo of the reaction-mass transfer methodology was established for continuous production of particles with controlled shapes.