1.Centre for Soft Condensed Matter Physics and Interdisciplinary Research, Soochow University, Suzhou 215006, China 2.College of Chemistry, Chemical Engineering and Materials Science, Soochow University, Suzhou 215123, China
Fund Project:Project supported by the National Natural Science Foundation of China (Grant Nos. 11574222, 11704269, 21522404) and the Natural Science Foundation of the Jiangsu Higher Education Institutions of China (Grant No.17KJB140020).
Received Date:05 March 2019
Accepted Date:18 March 2019
Available Online:01 May 2019
Published Online:20 May 2019
Abstract:The nature of crystallization is considered to be one of the most fundamental research problems in condensed matter physics. With single particle resolution offered by video microscopy, colloidal suspensions provide a novel model system for studying crystallization, melting and other phase transitions, where the structures and dynamics of the particles during the transitions can be quantitatively probed. Traditional systems for studying the crystallization typically focus on the infinitely large systems in order to obtain the equilibrium state. However, studies of the crystallization in finite-sized systems such as crystallization in thin films and porous media, are rare despite the fact that they are the common phenomena in natural world. In this paper, we experimentally investigate the crystallization in a finite-sized colloidal system with attractive interactions. The colloidal suspension is composed of polystyrene microspheres dispersed in a mixture of water and 2, 6-lutidine, in which the interaction between the particles can be tuned by adjusting the temperature. We increase the temperature to 34 °C to induce attractions between the particles and thus producing a cluster, and then reduce the temperature to 33 °C to tune off the attractions. Thus we obtain a finite-sized liquid cluster of the colloidal particles. Crystallization is triggered by increasing the temperature to 34 °C. The crystallization process is recorded by video microscopy and the video data are analyzed by a standard particle tracking algorithm. Through the analysis of radial distribution function, Voronoi diagram, and local order parameter, we find that the crystallization of the finite colloidal system starts from the central dense region of the liquid cluster. This leads to a crystalline phase in the center and a liquid phase on the edge of the cluster. As time elapses, the central crystalline region grows while outer liquid region shrinks. The crystallization process exhibits a two-step scenario: a fast crystallization initially and a slow crystallization at the later stage. At the initial stage, the center of the system forms a dense metastable liquid phase, which lowers the free-energy barrier of crystallization and results in a fast crystallization. As the crystalline region grows, the metastable phase disappears, and thus the crystallization rate decreases. Moreover, a bimodal distribution of the orientational order parameter is observed during the crystallization in our finite-sized colloidal system, which is consistent with that in a large system. This indicates that the bimodal distribution is a common feature of the two-dimensional crystallization. Keywords:colloids/ crystallization/ structure/ phase transition
其中, n表示粒子数密度, $\rho {\rm{ = }}\sum {_{i = 1}^{N\left( t \right)}\delta \left( {r - {r_i}\left( t \right)} \right)} $表示区域内的粒子分布, $\left\langle {} \right\rangle $表示对时间和空间上的平均. 径向分布函数表征了在距离粒子r处找到其他粒子的几率[23]. 图2给出了体系在不同结晶时刻径向分布函数的变化规律. 从图2可以看出, 随着结晶时间的增长, 径向分布函数g(r)的峰变高并且峰的数目增多, 说明体系粒子位置关联性增强, 表明体系结构从无序向有序转变. 图 2 不同结晶时刻的径向分布函数g(r) Figure2. Pair correlation function g(r) at different times of crystallization.
为了直观地表示体系的有序/无序的结构, 图3给出了体系的泰森多边形(Voronoi diagram)分布图随结晶时间的变化[5], 不同颜色的多边形代表粒子周围的近邻粒子数不同. 当体系中的粒子都具有相同的近邻粒子数时, 泰森多边形分布图为单一颜色, 表明体系结构具有周期性, 代表有序结构; 而当粒子近邻粒子有很大差异时, 分布图表现为多种颜色, 说明体系处于无序结构. 因此, 泰森多边形可以直观地表征体系的结构特征, 并且可以有效地分析体系内缺陷产生、聚集和消失等行为. 对于二维单分散的圆球体系, 当粒子的近邻粒子数都为6时, 表明体系形成了六角晶格, 是一种完美的二维有序结构. 从图3可以看出, 随着时间增加, 体系中央的高密度区域含有6个近邻粒子数的粒子首先增多, 然后向外层扩散. 这说明体系的结构越来越有序, 体系发生了结晶行为, 而且是从中间高密度区域向外变化. 此外, 统计近邻粒子数为6的粒子占总粒子数的比例(图4). 结果显示, 体系在结晶过程中, 结构有序性在400 s之前先快速增加, 而后缓慢增加. 这种统计变化规律与图3中直观图的结果一致. 这些结果说明, 有限尺寸胶体体系发生结晶时存在一个特征尺寸: 当晶相小于特征尺寸时, 结晶快速生长; 而当晶相达到特征尺寸后, 结晶速率降低. 对于结晶, 其晶相生长速率取决于促使体系结晶的驱动力和结晶自由能能垒[24,25]. 驱动力越大、能垒越低, 结晶速率会越快. 对于特定的体系, 其自由能能垒在一般情况下是保持不变的. 因此, 驱动力的大小决定了体系的结晶速率. 在本文的胶体体系中, 这种驱动力是粒子间的相互吸引力. 由于实验过程中吸引力大小恒定, 因此体系结晶速率应该保持不变. 然而, 在实验中观察到了两个结晶速率不同的阶段. 从而可以推断, 在结晶初期, 中央区域是高密度的亚稳态液体, 会降低体系结晶自由能能垒, 导致体系在初始阶段快速结晶. 这种由亚稳态导致体系结晶自由能能垒降低从而加快体系结晶速率的行为已经得到了理论计算和实验研究的证实[16,24,26]. 随着晶相成长到一定的特征尺寸后, 中央高密度亚稳态消失, 于是结晶速率变慢. 这两步结晶行为与文献报道的小尺寸晶体的两步熔化过程类似: 晶体熔化时晶相首先缓慢减小, 当减小到一定尺寸后体系进入高密度亚稳态, 然后迅速熔化[15]. 图 3 不同结晶时刻的泰森多边形分布图, 图中的色标表示粒子的最近邻粒子的个数(参见图标) Figure3. Snapshots of Voronoi diagram at different times. The color code indicates the number of nearest neighbors for the particles (see the legend).
图 4 近邻粒子数为6的粒子数占总粒子数的比例随结晶时间的变化, 两段不同斜率的红色线条表明, 体系结构变化大约分为两个阶段: 先快后慢 Figure4. The fraction of particles with 6 nearest neighbors versus crystallization time. The red line is a guide to the eye.
为了阐明这一小尺寸胶体体系结晶规律, 我们计算了粒子的局部取向序参量[27], ${\psi _{6i}} = \dfrac{1}{{n{}_i}}\displaystyle\sum\limits_{j = 1}^{{n_i}} {{{\rm{e}}^{{\rm i}{\theta _{ij}}}}} $, 其中ni是粒子i的最近邻粒子数, ${\theta _{ij}}$是粒子i与它最近邻的第j个粒子的连线同水平坐标轴的夹角. 当${\psi _{6i}}=1$时, 说明粒子呈现完美的六边形排列; ${\psi _{6i}}=0$表示粒子呈现随机排列的无序结构. 图5给出不同结晶时刻体系中粒子的局部取向序参量分布图. 在起始阶段, 局部取向序参量值总体较低, 体系处于液态. 随着结晶的开始, 体系中间高密度的区域局部取向序参量值迅速增加, 中间区域首先变得有序, 呈现中间有序周围无序的结构, 即固液共存的状态. 随着结晶时间的增长, 这种有序区域不断增大, 说明晶相不断增大而液相不断减小, 最后体系形成晶态. 另外, 这种结晶过程还可以考察体系局部取向序参量平均值的变化规律. 图6给出了局部取向序参量的平均值随结晶时间的变化: 400 s之前迅速增长, 而后缓慢增加. 这同样说明了体系结晶过程中结构的变化分为先快后慢两个阶段, 与粒子最近邻粒子数的结果一致(图3和图4). 图 5 体系取向序参量随时间分布图, 颜色代表粒子取向序参量${\psi _{6}}$值(参见色带图) Figure5. Snapshots of orientational order parameter ${\psi _{6}}$ at different times of crystallization. The color of particles represents the value of ${\psi _{6}}$ (see the color bar).
图 6 粒子取向序参量的平均值随结晶时间的变化 Figure6. The time-dependent averaged orientational order parameter, ${\psi _{6}}$. The red line is a guide to the eye.
对于大尺度二维体系, 已有的研究表明在结晶过程中序参量会出现特征的分布规律[14]. 为了探究有限尺寸的结晶过程是否也有类似的特征分布, 考察了取向序参量在结晶过程中的分布变化(图7). 结果显示: 结晶开始时, 体系序参量的分布出现单一的峰, 峰位对应的${\psi _{6}}$的值很小, 说明体系呈现无序结构, 是液态; 随着时间增加, 序参量的分布出现双峰, 分别对应液态和固态结构; 随着时间进一步增加, 序参量的分布只出现一个峰, 峰位对应的${\psi _{6}}$的值接近于1, 说明体系转变为固态结构. 这种取向序参量分布的变化规律与大尺度二维体系的结晶行为一致[14], 说明它是二维结晶行为的普遍规律, 可以作为判断二维结晶相变发生的标志. 图 7 取向序参量的值在不同结晶时刻的分布 Figure7. The distribution of orientational order parameter at different crystallization times.