Abstract Oxide nanoparticle growth from vapor phase precursors occurs in high temperature aerosol reactors first via the formation of nanoclusters, which are nanometer-scale condensed-phase species composed of 101-102 atoms. The binding rate for nanoclusters, defined as the rate at which two nanoclusters collide with and stick to one another, is hence of interest in predicting nanoparticle size distribution functions in an aerosol. We have utilized molecular dynamics (MD) trajectory calculations to determine the homogeneous (equal-sized) and heterogeneous (disparate-sized) binding rate coefficients of SiO2 (silica) nanoclusters composed of 18, 144, and 333 atoms. MD trajectory calculations incorporated all-atom models of nanoclusters using a combined Born–Huggins–Mayer-Lennard-Jones potential model, which accounts for both short range interactions and electrostatic interactions. MD calculations were utilized to determine the binding probability as a function of both initial relative velocity and impact parameter; integration of the binding probability across impact parameter and relative velocity space yields the binding rate coefficient. MD trajectory calculations reveal that the most common type of encounters between nanoclusters leading to binding are grazing collisions, i.e. instances where collision would not occur without induced dipole potential influences. The resulting binding rate coefficients are found to be extremely weakly dependent on system temperature, which is in contrast to the use of rate coefficient models which are the product of a hard-sphere collision rate coefficient and a constant enhancement factor (leading to a rate coefficient proportional to the square root of temperature). Enhancement factors defined from MD trajectory calculations fall in the range of 3–9 as system temperature decreases from 1500 K to 300 K. Such large values suggest that potential interactions need to be considered when modeling oxide nanocluster growth in the gas phase.

中文翻译：

---

来自分子动力学轨迹计算的二氧化硅纳米团簇结合率系数

摘要 在高温气溶胶反应器中，气相前驱体的氧化物纳米颗粒生长首先通过纳米团簇的形成发生，纳米团簇是由 101-102 个原子组成的纳米级凝聚相物质。纳米团簇的结合率，定义为两个纳米团簇相互碰撞和粘附的速率，因此在预测气溶胶中的纳米颗粒尺寸分布函数中很重要。我们利用分子动力学 (MD) 轨迹计算来确定由 18、144 和 333 个原子组成的 SiO2（二氧化硅）纳米团簇的均质（等尺寸）和异质（不同尺寸）结合率系数。MD 轨迹计算使用组合的 Born-Huggins-Mayer-Lennard-Jones 势模型结合了纳米团簇的全原子模型，这说明了短程相互作用和静电相互作用。MD 计算用于确定结合概率作为初始相对速度和冲击参数的函数；跨冲击参数和相对速度空间的结合概率的积分产生结合率系数。MD 轨迹计算表明，导致结合的纳米团簇之间最常见的相遇类型是掠碰撞，即在没有诱导偶极电位影响的情况下不会发生碰撞的情况。发现所得结合率系数极弱地依赖于系统温度，这与使用速率系数模型形成对比，后者是硬球碰撞速率系数和恒定增强因子的乘积（导致速率系数与温度的平方根成正比）。随着系统温度从 1500 K 降低到 300 K，由 MD 轨迹计算定义的增强因子在 3-9 的范围内。如此大的值表明在模拟气相中的氧化物纳米团簇生长时需要考虑潜在的相互作用。