石英玻璃分子动力学模拟中的原子电荷转移与系综选择

介绍了SiO2体系分子动力学模拟中的Si、O原子电荷转移问题;采用Morse势函数研究了原子电荷转移对石英玻璃模拟的影响,发现原子电荷转移在影响模型密度的同时,还直接影响着原子的最近邻距离.NPT和NVT系综下的模拟结果对比显示,系综对模型中原子最近邻情况影响不大,但在NVT系综下模拟结果表明实际玻璃中存在的较大的空隙结构,找到了以往模拟中密度结果偏高的原因,提出了一种较好的石英玻璃分子动力学建模的方法.该方法不但解决了在调整电荷时维持原子最近邻距离与保证模型密度之间的矛盾,而且可以很好地描述石英玻璃在远程结构上密度不均、存在较大空隙的无序结构.此外,原子自扩散系数的计算结果展示了空隙结构在石英玻璃扩散性质研究中的作用.     

分子间相互作用的Ab Initio计算研究——氢键复合物体系(HF)_2中的非键相互作用

本文以6-311 G~(**)为基组对氢键复合物(HF)_2进行了Ab Initio计算,给出了2个HF在距离及取向平衡点附近的势能面。用势诱导最小二乘拟合法计算了体系中各原子的原子电荷。用(exp-6-1)势函数分析法分解了2个HF间的氢键相互作用能,表明其中主要的相互作用是电荷转移而不是静电性的。2个HF分子间的势能曲线呈Morse函数性质。     

两体扰动势和H2O通道反应的三原子体系解析势

提出了两体扰动势的概念.通过两体扰动势构造多原子相互作用体系解析势的理论方法,获得了非线性三原子H2O相互作用体系的解析势函数.用势能面正确描述了O（1D）＋H2→H2O→OH＋H通道反应的性质.理论分析和实验结果非常符合.该方法比较容易推广构造多原子相互作用体系解析势.     

两体扰动势和H2O通道反应的三原子体系解析势

提出了两体扰动势的概念.通过两体扰动势构造多原子相互作用体系解析势的理论方法,获得了非线性三原子H2O相互作用体系的解析势函数.用势能面正确描述了O(1D)+H2→H2O→OH+H通道反应的性质.理论分析和实验结果非常符合.该方法比较容易推广构造多原子相互作用体系解析势.     

疏水性微孔中水的结构和扩散性质的分子模拟

用分子动力学（MD）方法模拟了受限在疏水性微孔中的水的结构与动力学行为.分别考察了孔径、温度和压力对水在孔道方向的密度分布和自扩散系数的影响,计算了不同温度下水的径向分布函数.发现在小孔径的微孔中,随着温度的降低,水分子沿孔道的分布逐渐变得不均匀,最终导致气-液相分离,微孔孔道内有明显的分段现象.受限在小孔径微孔中水的自扩散系数大约为体相流体水的20％～30％,并且随着孔径的减小,自扩散系数也减小.同时还发现沿孔道方向的自扩散系数分量大约为孔径方向的4～5倍.提出了微孔中水自扩散系数的关联模型.     

TbSin(n=2-13)团簇的结构、电子及磁学性质

利用相对论密度泛函理论在广义梯度近似下研究TbSi_n (n=2-13)团簇的结构、稳定性、电子和磁学性质. 对团簇的平均结合能、离解能、电荷转移、最高占据分子轨道(HOMO)和最低未占据分子轨道(LUMO)的能级差、Mulliken 电荷分析和磁学性质进行了计算和讨论. TbSi_n团簇并没有像实验推测的那样在n=10形成嵌入式的结构. 我们推断电子亲和势的急剧变化不仅与嵌入式的结构有关, 而且与电子的固有稳定性相关.Mulliken 电荷分析表明电荷总是从Tb 原子转向Si 原子. 团簇的磁矩主要局域在Tb 原子的周围, 并且主要由f电子贡献, f 电子表现出局域性并且不参与化学成键. 以TbSi₁₀为例的分波态密度分析表明Tb与Si 原子间存在很强的sp轨道杂化.     

钪凝固过程局域原子结构演化的第一性原理研究

采用第一性原理分子动力学模拟方法,研究了不同温度下钪熔体的局域原子结构演化和动力学行为。通过均方位移和速度自相关函数的分析,研究了钪熔体的原子动力学。当从高温降到接近熔点时,增强的背散射效应将直接影响钪熔体的动力学特性,如自扩散系数。进一步利用对关联函数、静态结构因子和键角分布分析了液态钪的局域原子结构。体系的第一配位间距随着温度降低而增大,有序度增加,扭曲二十面体结构增加。通过与其他常见金属比较,发现稀土金属钪在凝固过程的演化行为与其存在相似性的同时还保留独特的性质。本研究可以为从量子力学视角解释稀土金属熔体中的物理现象提供理论基础。     

铋系氧化物超导体晶体结构的模拟

自Michel等人首次在Bi-Sr-Cu-O体系中获得7—22 K的Bi_2Sr_2CuO_6超导相以来, 一类新型的双层铋系超导体引起人们的注意, 研究结果表明超导相的转变温度T_c随CuO_2层数的增加而增高. 因此, 晶体结构和超导性之间的关系是人们注意的一个问题. 铋系超导体的晶体完整性很差, 文献仅给出晶体平均结构的晶胞参数. 本文采用原子经验势方法, 对Bi_2Sr_2Ca_(n-1)Cu_nO_(2n+4)(n=1,2,3,4)的理想晶体结构作了模拟计算. 计算了单胞的结构参数和原子的位置参数. 计算表明: Cu-O(3)的原子间距和CuO-BiO的层间距离随n增加而减小. 根据电荷转移模型, 载流子的数目和电导层(CuO_2)及电荷储蓄层(BiO)之间的电荷转移有关. 层间距离的缩短似乎有利于载流子的形成.     

用渗流理论计算高分子溶液中溶剂的自扩散系数

基于微观渗流理论建立了溶剂小分子在高分子溶液中的自扩散模型,并据此模型对不同温度和浓度下的聚苯乙烯(PS)-苯、PS-甲苯、PS-乙苯和PS-四氢呋喃4个体系中小分子的自扩散系数进行了关联,计算出在不同温度下溶剂分子扩散所需的临界浓度。结果表明,在PS玻璃化温度以下,本模型对于温度和浓度具有很好的适用性和关联精度。     

二十面体准晶对非晶形成影响的模拟

采用分子动力学模拟技术,以液态金属Ni为例,研究了在不同冷却条件下形成晶体及非晶的过程.模拟采用镶嵌原子法(EAM)作用势,得到了不同温度、不同冷却速度下Ni的径向分布函数以及原子组态变化的重要信息,利用键对分析技术探讨了二十面体准晶对非晶形成的影响.     

First principles molecular dynamics of molten NaI: structure, self-diffusion, polarization effects, and charge transfer

The structure and self-diffusion of NaI and NaCl at temperatures close to their melting points are studied by first principles Hellmann-Feynman molecular dynamics (HFMD). The results are compared with classical MD using rigid-ion (RI) and shell-model (ShM) interionic potentials. HFMD for NaCl was reported before at a higher temperature [N. Galamba and B. J. Costa Cabral, J. Chem. Phys. 126, 124502 (2007)]. The main differences between the structures predicted by HFMD and RI MD for NaI concern the cation-cation and the anion-cation pair correlation functions. A ShM which allows only for the polarization of I- reproduces the main features of the HFMD structure of NaI. The inclusion of polarization effects for both ionic species leads to a more structured ionic liquid, although a good agreement with HFMD is also observed. HFMD Green-Kubo self-diffusion coefficients are larger than those obtained from RI and ShM simulations. A qualitative study of charge transfer in molten NaI and NaCl was also carried out with the Hirshfeld charge partitioning method. Charge transfer in molten NaI is comparable to that in NaCl, and results for NaCl at two temperatures support the view that the magnitude of charge transfer is weakly state dependent for ionic systems. Finally, Hirshfeld charge distributions indicate that differences between RI and HFMD results are mainly related to polarization effects, while the influence of charge transfer fluctuations is minimal for these systems.     

Characterization of capillary inner surface conditions with streaming potential

Streaming potential is created when an electrolyte solution is forced to flow pass a charged surface. For an uncoated fused silica capillary, the streaming potential is measured between the inlet and outlet vials while applying a pressure across the capillary. The changes in streaming potential can be used to characterize the properties of the capillary inner surface. In this work, HCl, NaCl, and NaOH solutions ranging from 0.4 to 6 mM were used as the background electrolyte (BGE) at temperatures of 15 to 35 °C for the mesurements. The streaming potential decreases with the increase in BGE concentration, and the trend is amplified at higher temperatures. When buffer solutions in the pH range of 1.5 to 12.7 were used as the BGE, streaming potential was shown to be sensitive to changes in pH but reaches a maximum at around 9.5. At pH < 3.3, no streaming potentials were observed. The pH of zero surface charge (streaming potential equals 0) changes with temperature, and is measured to be 3.3 to 3.1 when the temperature is changed from 15 to 35°C. Zeta potentials can be calculated from the measured streaming potential, conductivity, and the solution viscosity. Surface charge densities were calculated in this work using the zeta potentials obtained. We demonstrated that capillary surface conditions can significantly change the streaming potential, and with three different solutions, we showed that analyte-dependent adsorption can be monitored and mitigated to improve the peak symmetry and migration times reproducibility.     

Isoelectric point of fluorite by direct force measurements using atomic force microscopy

Interaction forces between a fluorite (CaF2) surface and colloidal silica were measured by atomic force microscopy (AFM) in 1 x 10(-3) M NaNO3 at different pH values. Forces between the silica colloid and fluorite flat were measured at a range of pH values above the isoelectric point (IEP) of silica so that the forces were mainly controlled by the fluorite surface charge. In this way, the IEP of the fluorite surface was deduced from AFM force curves at pH approximately 9.2. Experimental force versus separation distance curves were in good agreement with theoretical predictions based on long-range electrostatic interactions, allowing the potential of the fluorite surface to be estimated from the experimental force curves. AFM-deduced surface potentials were generally lower than the published zeta potentials obtained from electrokinetic methods for powdered samples. Differences in methodology, orientation of the fluorite, surface carbonation, and equilibration time all could have contributed to this difference.     

Insights into electrocatalysis

When standard reversible potentials for bulk solution reactions, U(0), are known, the reversible potentials when the reactant and product are adsorbed on an electrocatalyst surface, U(surf)(rev), are given in terms of these potentials and the adsorption Gibbs energy bond strengths: U(surf)(rev) = U0 + D(ads)G (Ox)/F-Δ(ads)G (R)/F (i). When the Δ(ads)G (Ox) and Δ(ads)G (Red) values are known at potential U, this equation is exact. When the overpotential for a multi-electron transfer reaction is minimal, each electron transfer takes place at the standard reversible potential for the overall reaction. In the case of O(2) reduction to water via the intermediate step OOH(aq) → O(aq) + OH(aq), or via O(2)(g) → 2O(aq), the respective endergonic O-O dissociation Gibbs energies are shown to be 2.52 eV and 4.76 eV. When the oxygen product and water reactant adsorb weakly, as on platinum, the adsorption Gibbs energies, Δ(ads)G, for O, OH, and OOH intermediates can be uniquely predicted using these data. All of the above depend exclusively on experimentally determined data. Reversible potentials have been calculated for oxygen reduction steps on the platinum electrocatalyst surface using Interface 1.0, a comprehensive computational code for the potential dependence of the electrochemical interface. Using these results as benchmarks, is found to be accurate to around 0.1 V when the Δ(ads)G are values calculated for the potentials of zero charge, instead of 1.229 V, which is a significant simplification. The variation in Δ(ads)G values between the calculated potentials of zero charge and 1.229 V are found to be 0.2 eV V(-1) or less. Prior work, using internal adsorption energies calculated at the potential of zero charge in place of Gibbs energies in was found to be accurate to within about 0.2 V. On platinum Δ(ads)G of the reaction OOH(ads) → O(ads) + OH(ads) is calculated at the potential of zero charge for the reactant and product to be about 1.2 eV exergonic under Langmuir conditions, and this Gibbs energy loss reduces the 1.229 V four-electron reversible potential on the platinum surface to an effective reversible potential of about 0.93 V for this mechanism on platinum. The effective reversible potential is a consequence of efficiency loss, not kinetics. Based on these values, the onset potential for four-electron oxygen reduction will be less than or equal to the effective reversible potential and on pure Pt(111) it appears to be equal to it.     

On the use of big-bang method to generate low-energy structures of atomic clusters modeled with pair potentials of different ranges

The efficiency of the so-called big-bang method for the optimization of atomic clusters is analysed in detail for Morse pair potentials with different ranges; here, we have used Morse potentials with four different ranges, from long- ρ = 3) to short-ranged ρ = 14) interactions. Specifically, we study the efficacy of the method in discovering low-energy structures, including the putative global minimum, as a function of the potential range and the cluster size. A new global minimum structure for long-ranged ρ = 3) Morse potential at the cluster size of n= 240 is reported. The present results are useful to assess the maximum cluster size for each type of interaction where the global minimum can be discovered with a limited number of big-bang trials.     

Comparison of spectroscopic potentials and an a priori analytical function. The potential energy curve of the ground state of the sodium dimer, X1Sigmag(+) Na2

The results of a "universal" potential energy function, one that incorporates electronegativity and Slater's effective nuclear charge into a Morse-type function, are compared to spectroscopically derived potential energy curves of the X1Sigmag(+) state of Na2. The function is a priori in that it does not require prior knowledge of the actual potential and has no adjustable parameters. Criteria used to evaluate the performance of the function are comparisons of predicted versus spectroscopic energies at Rydberg-Klein-Rees (RKR) procedure turning points, predicted distances at measured energies versus RKR distances, and eigenvalues derived from the a priori potential versus spectroscopically deduced energy levels. The a priori function describes the Na2 potential with deviations approaching the magnitude of those found among some spectroscopic potentials from different sources. By examining the behavior of the "spectroscopic" parameter of the Morse function, irregularities are found in five of the seven spectroscopic potentials examined. A new procedure is demonstrated for correcting irregularities on the inner branch of spectroscopic potentials at high extents of dissociation and for extending reliably the potential in this region beyond the domain of the measurements.     

The calculation of vapor-liquid coexistence curve of Morse fluid: application to iron

The vapor-liquid coexistence curve of Morse fluid was calculated within the integral equations approach. The critical point coordinates were estimated. The parameters of Morse potential, fitted for elastic constants in solid phase, were used here to apply the results of present calculations to the determination of iron binodal. The properties of copper and sodium were considered in an analogous way. The calculations of pair correlation functions and isobars at liquid phase have shown that only for sodium these potential parameters allow one to obtain agreement with the measurements data. For iron another parameters are necessary to get this agreement in liquid phase. However, they give rise to very low critical temperature and pressure with respect to the estimates of other authors. Consequently, one can suppose that Morse potential is possibly inapplicable to the calculation of high temperature properties of non-alkali metals in disordered phases.     

非晶态Co-B的局域电子结构的Xα原子簇计算

In the research on metallic glass, there are arguments against the "rigid band charge transfer", model which assumes a charge transfer from metalloid atoms to transition metal atoms to explain the experimental evidence of linear reduction of average atomic magnetic moment with the increasing concentration of metalloid atoms, but they could not explain the experimental relation of the reduction. In the present work, spin-polarized SCC-DV-X_a calculation for atomic clusters for metallic glass Co-B has been empoloyed to investigate the local electronic structure and magnetic property of the metallic glass. As opposed to the "rigid band charge transfer" model, calculation in the present work indicates that charge transfers from Co4s to both B and Co3d. It is found that there is Co3d4s-B2p hybird bonding in Co-B, which leads to the linear reduction of average atomic magnetic moment. Thus the explanation removes the above controversy.