Complex Liquids Laboratory, Department of Physics, National Central University, Chungli, 320, Taiwan.
J Comput Chem. 2021 Feb 15;42(5):310-325. doi: 10.1002/jcc.26457. Epub 2020 Dec 18.
The DFTB theory was combined with the isothermal Brownian-type molecular dynamics (MD) and metadynamics molecular dynamics (MMD) algorithms to perform simulation studies for Au clusters. Two representative DFTB parametrizations were investigated. In one parametrization, the DFTB-A, the Slater-Koster parameters in the DFTB energy function were determined focusing on the ionic repulsive energy part, E and the other, the DFTB-B, due attention was paid to the electronic band-structure energy part, E . Minimized structures of these two parametrizations were separately applied in MD and MMD simulations to generate unbiased and biased trajectories in collective variable (CV) space, respectively. Here, we found the MD simulations monitored at 300 K manifest fluxional characteristics in planar cluster Au /DFTB-A, but give no discernible tracts of fluxionality for planar Au /DFTB-A and Au /DFTB-B, for nonplanar Au /DFTB-A and, to some extent, for nonplanar Au /DFTB-B; they are plausibly being hindered by higher-than k T energy barriers. Very recent FIR-MPD spectroscopy measurements, however, were reported to have detected at 300 K both the planar and nonplanar neutral Au clusters in the size range 5 ≤ n ≤ 13. The failure of MD simulations has prompted us to apply the MMD simulation and construct the free energy landscape (FEL) in CV space. Through scrutinizing the FELs of these clusters and their associated structures, we examine the relative importance of E /DFTB-A and E /DFTB-B in unraveling the covalent-like behavior of valence electrons in Au . Most important of all, we shall evaluate the DFTB parametrization in MMD strategy through comparing extensively the simulation data recorded with the gas-phase experimental data.
DFTB 理论与等温布朗型分子动力学(MD)和元动力学分子动力学(MMD)算法相结合,对 Au 团簇进行了模拟研究。研究了两种有代表性的 DFTB 参数化方法。在一种参数化方法(DFTB-A)中,DFTB 能量函数中的 Slater-Koster 参数是针对离子排斥能部分(E)来确定的;另一种参数化方法(DFTB-B)则特别关注电子能带结构能量部分(E)。分别用这两种参数化方法的最小化结构对 MD 和 MMD 模拟进行了应用,以分别在集体变量(CV)空间中生成无偏和有偏轨迹。在这里,我们发现 300K 下的 MD 模拟在平面 Au/DFTB-A 簇中表现出了流变性特征,但在平面 Au/DFTB-A 和 Au/DFTB-B 中没有明显的流变性轨迹,对于非平面 Au/DFTB-A,并且在某种程度上对于非平面 Au/DFTB-B,由于高于 kT 的能量障碍而受到阻碍。然而,非常新的 FIR-MPD 光谱测量报告在 300K 下检测到了尺寸范围为 5≤n≤13 的平面和非平面中性 Au 团簇。MD 模拟的失败促使我们应用 MMD 模拟并在 CV 空间中构建自由能景观(FEL)。通过仔细研究这些团簇及其相关结构的 FEL,我们检验了 DFTB-A 和 DFTB-B 在揭示 Au 中价电子的共价行为方面的相对重要性。最重要的是,我们将通过广泛比较与气相实验数据记录的模拟数据,来评估在 MMD 策略中的 DFTB 参数化。