Exploring artificial neural networks to model interatomic and intermolecular potential energy surfaces

We demonstrate how a back-propagation artificial neural network can be trained to represent a potential energy surface (PES) in a formless manner with limited data points and exploited to predict interaction energies for configurations not included in the training set. A similar exercise is undertaken for predicting the eigenvalues and eigenvectors of a model Hamiltonian matrix that delicately depends on parameters and exhibits crossing of eigen values.     

Electron and positron pair emission by low energy positron impact on surfaces

The emission of electron pairs from surfaces has the power to reveal details about the electron–electron interaction in condensed matter. This process, stimulated by a primary electron or photon beam, has been studied both in experiment and theory over the last two decades. An additional pathway, namely positron–electron pair emission, holds the promise to provide additional information. It is based on the notion that the Pauli exclusion principle does not need to be considered for this process.We have commissioned a laboratory based positron source and performed a systematic study on a variety of solid surfaces. In a symmetric emission geometry we can explore the fact that positron and electron are distinguishable particles. Following fundamental symmetry arguments we have to expect that the available energy is shared unequally among positron and electron. Experimentally we observe such a behavior for all materials studied. We find an universal feature for all materials in the sense that on average the positron carries a larger fraction of the available energy. This is qualitatively accounted for by a simplified scattering model. Numerical results, which we obtained by a microscopic theory of positron–electron emission from surfaces, reveal however that there are also cases in which the electron carries more energy. Whether the positron or the electron is more energetic depends on details of the bound electron state and of the emission geometry. The coincidence intensity is strongly material dependent and there exists an almost monotonic relation between the singles and coincidence intensity. These results resemble the findings obtained in electron and photon stimulated electron pair emission. An additional reaction channel is the emission of an electron pair upon positron impact. We will discuss the energy distributions and the material dependence of the coincidence signal which shows similar features as those for positron–electron pairs.     

Geometric analysis of anharmonic downward distortion following paths

A mathematical aspect of the anharmonic downward distortion following (ADDF) path is discussed. The ADDF method is utilized as an automated reaction path search method, which can explore transition state geometries on a potential energy surface from a potential minimum. We show that the maximum number of the ADD stationary paths intersecting the potential minimum is 2^{f + 1} ? 2 , where f denotes the degree of freedom of the system. We also show that the bifurcation of the ADD stationary path is essential to detect all the transition states connected from a given minimum. The ADDF computation is demonstrated for a H₂O molecule in which pitchfork bifurcation is observed.     

二维材料用于渗透能转换的研究进展

在河水与海水的交界处实现渗透能提取与捕获是解决未来能源危机的重要方式之一. 渗透能因为储量大, 容易获取以及绿色可持续的优势受到广泛关注. 反向电渗析技术是一种能够有效捕获渗透能的方法之一, 目前已经得到了深入的研究与发展. 离子交换膜是反向电渗析技术转换渗透能的关键组件, 其性能的优异程度决定能量转换效率的高低. 常见的膜材料主要是高分子聚合物及其改性化合物, 最近一些二维材料如石墨烯、 氧化石墨烯、 二硫化钼、 各种框架材料及其改性复合物因优异的选择性离子传输、 纳米级通道、 丰富的表面功能基团以及可修饰性成为捕获渗透能的重要膜材料. 本文综合评述了二维材料作为离子传输通道的类型以及相应的传输机理; 例举了二维材料及其复合物的设计方案和在渗透能转换方面的具体应用; 最后提出了目前二维材料在渗透能转换领域中面临的挑战以及未来的发展方向.     

高通量虚拟筛选CDK2/Cyclin A2靶点抑制剂

CDK2/Cyclin A2复合蛋白的异常表达与乳腺癌、 口腔癌、 食管鳞状细胞癌的发生密切相关. CDK2/ Cyclin A2复合蛋白的活性位点不同于CDK2单体. 至今临床上尚无靶向此复合蛋白的药物分子. 针对CDK2/Cyclin A2复合蛋白, 以实验报道的10个抑制剂分子构建药效团模型, 通过药物体外药代动力学（ADME）、 Docking、 聚类分析、 毒性预测, 从DrugBank, ChEMBL和TCM@Taiwan 3个数据库约90万组数据中进行高通量虚拟筛选, 进一步进行MD模拟、 MM/PBSA结合自由能计算、 能量分解和平均非共价作用(aNCI)分析, 筛选出3个抑制效果优于阳性实验药Roscovitine的先导分子: DrugBank-2004, DrugBank-583和ChEMBL-7122. 与CDK2蛋白相比, CDK2/Cyclin A2复合蛋白结合位点空间变大, 先导分子与Lys33, Asp86, Lys129和Asp145残基之间的排斥作用有所降低, 导致结合自由能更大.     

锂嵌入型化合物-氢气电池

可充电氢气电池作为一种新兴的电池体系在大规模能源储存领域显示出富有前景的电化学性能. 锂嵌入型化合物作为一大类的锂离子电池正极材料能够很好地用作可充电氢气电池的正极. 本文开发了 2种新型锂嵌入型化合物-氢气电池. 通过使用钴酸锂与磷酸铁锂2种正极材料分别与氢气负极在硫酸锂 水系电解液中进行匹配, 得到了钴酸锂-氢气电池与磷酸铁锂-氢气电池. 钴酸锂-氢气电池展现出约1.27 V 的放电电位, 约97 mA·h·g^-1的比容量及10C的高倍率; 磷酸铁锂-氢气电池展现出约0.66 V的放电电位, 约125 mA·h·g^-1的比容量以及10C的高倍率. 虽然, 钴酸锂-氢气电池和磷酸铁锂-氢气电池因为使用了未经优化的、 不稳定的锂嵌入型化合物正极材料而导致全电池容量衰减, 但这2种电池经过氢气负极的再循环利用均表现出优异的恢复能力. 本文结果证明了氢气电池的化学稳定性及其在未来长寿命电池中具有的大规模能源储存潜力.     

优化金属磷化物电催化析氢性能的策略

日益严重的能源危机和环境污染问题使得探索清洁的可再生能源载体及减少对传统化石燃料的过度依赖成为人们面临的一项重要任务.因此,各种可持续能源如太阳能、风能、海洋能和生物质能等得到了广泛研究并取得了一定的进展.然而,这些能源因存在间歇性和不稳定性等缺点阻碍了其实际应用.近年,氢气作为一种能源载体,以其高能量密度和无碳排放的优点引起了人们的广泛关注,被认为是缓解日益严重的污染问题的最有前途的环保能源.对比目前采用的天然气热解和煤炭气化等传统制氢策略,电催化水裂解由于催化效率高,制氢纯度高和不产生温室气体,被认为是高效、环保、可持续的制氢策略.电催化水裂解由两个独立的半反应组成,分别是析氢反应和析氧反应.析氢反应作为水裂解的一个半反应,在降低制氢成本及提高产氢催化效率方面起着关键作用.然而,目前的核心问题之一是要开发高效的析氢电催化剂,以加快反应速度.目前,铂和铂基纳米材料被认为是高效的析氢电催化剂,但是其稀缺性和高成本阻碍了大规模实际应用.金属磷化物由于具有较高的本征活性并且在不同的电解质中都具有良好的电催化析氢性能,被证明是一种优良的析氢电催化剂.此外,与普通催化剂相比,金属磷化电催化剂还具有合成简便、效率高、成本低、省时等优点.本文详细介绍了近年人们在金属磷化物用于电催化析氢研究中取得的进展.首先,介绍了电催化析氢反应机理,金属磷化物的结构及作用,并对其优缺点进行了总结;随后,综述了金属磷化物的合成方法,包括后处理、原位生成和电沉积策略,并对不同方法进行了比较和讨论.此外,从元素掺杂、界面工程、空穴工程、修饰特定载体、构建特定纳米结构、设计双或多金属磷化物和其他发展的新方法等七个方面详细总结了促进金属磷化物电催化活性的多种策略,并进行了对比和讨论.最后,归纳了金属磷化物在电催化析氢应用中存在的问题和面临的挑战,并对未来的研究发展提出了展望.     

碱性介质中氢氧化反应电催化剂的开发:从机理认识到材料设计

阴离子交换膜(AEM)燃料电池因具有使用非贵金属作为催化剂的优点而受到广泛关注.然而,在碱性体系中,AEM燃料电池中氢氧化反应(HOR)的反应动力学比在酸性介质中的慢两个数量级.针对HOR在碱中动力学缓慢的问题,有两种主要的理论来解释,(1)pH相关的氢结合能作为主要影响因素来控制HOR动力学的理论;(2)质子和氢氧根离子的吸附共同作为影响因子来控制HOR在碱性条件下的动力学的双功能理论.本文首先讨论了在碱性电解质中可能的HOR反应机理及其Tafel性能变化.除了传统的Tafel-Volmer和Heyrovsky-Volmer-HOR机理外,还讨论了最新提出的氢氧根离子吸附参与的HOR机理来说明在酸性和碱性介质中HOR机理的差异.然后,总结了具有代表性的碱性HOR催化剂(如贵金属、合金、金属间化合物、镍基合金、碳化物、氮化物等),简要介绍了它们相应的HOR反应机理,从而进一步理解在碱性介质中不同基元反应步骤给HOR性能带来的差异.最后,提出了一种未来设计HOR碱性催化剂的可行性方案,为今后碱性环境下的HOR催化剂设计提供参考.     

Adaptation of electrodes and printable gel polymer electrolytes for optimized fully organic batteries

Despite intensive scientific efforts on the development of organic batteries, their full potential is still not being realized. The individual components, such as electrode materials and electrolytes, are in most cases developed independently and are not adjusted to each other. In this context, we report on the performance optimization of a full-organic solid-state battery system by the mutual adaptation of the electrode materials and an ionic liquid (IL)-based gel polymer electrolyte (GPE). The formulation of the latter was designed for a one-step manufacturing approach and can be applied directly to the electrode surface, where it is UV-cured to yield the GPE without further post-treatment steps. Herein, a special focus was placed on the applicability in industrial processes. A first significant capacity increase was achieved by the incorporation of the IL into the electrode composite. Furthermore, the GPE composition was adapted applying acrylate- and methacrylate-based monomers and combinations thereof with the premise of a fast curing step. Furthermore, the amount of IL was varied, and all combinations were evaluated for their final performance in cells. The latter variation revealed that a high ionic conductivity is not the only determining factor for a good cell performance. Next to a sufficient conductivity, the interaction between electrode and electrolyte plays a key role for the cell performance as it enhances the accessibility of the counter ions to the redox-active sites.     

水分子在均三嗪基g-C3N4上的吸附

气体分子与光催化剂之间的相互作用对于光催化反应的触发非常重要.对于TiO2,ZnO和WO3等传统金属氧化物光催化剂上的水分解反应而言,已有许多报道研究了水分子在它们表面的吸附行为.结果表明,水分子与催化剂表面的原子形成了O-H…O氢键.石墨相氮化碳(g-C3N4)是一种具有可见光响应且化学性质稳定的光催化剂,对其进行修饰以增强其分解水产氢性能的研究非常多.本文通过密度泛函理论计算,全面研究了水分子在均三嗪(s-triazine)基g-C3N4上的吸附情况.首先构建了一系列初始吸附模型,考察了各种吸附位和水分子的朝向.通过比较分析计算得到的吸附能,确定了一种最优的吸附构型,即水分子以竖直的朝向吸附于褶皱的单层g-C3N4表面.水分子中的一个极性O-H键与g-C3N4中一个二配位富电子的氮原子结合形成了分子间的O-H…N氢键.其中,H原子与N原子的间距为1.92?,O-H键的键长由0.976?增至0.994?.进一步通过计算Mulliken电荷,态密度和静电势曲线分析了该吸附体系的电子性质.结果发现在分子间氢键的桥接作用下,g-C3N4上的电子转移至水分子,由此导致g-C3N4的费米能级降低,功函数由4.21 eV增至5.30 eV.在该吸附模型的基础上,考查了不同的吸附距离.当水分子与g-C3N4的间距设为1至4?时,几何优化后总是能得到相同的吸附构型,吸附能和氢键长度也十分相近.随后,通过改变吸附基底g-C3N4的大小和形状,验证了这种吸附构型具有很强的重复性.将2′2单层g-C3N4吸附基底替换为2′2多层g-C3N4(2至5层),3′3和4′4单层g-C3N4,以及具有不同管径的单壁g-C3N4纳米管后,水分子的吸附能随着体系原子数的增多而增大,但吸附模型的几何结构和电子性质基本不变,包括O-H…N氢键的形成和键长,以及电子转移和增大的功函数.另外还研究了非金属元素(P,O,S,Se,F,Cl和Br)掺杂对吸附能的影响.构建模型时,杂质原子以取代二配位氮原子的方式进行掺杂,水分子放置于杂质原子上方.结果显示,引入杂质原子后水分子的吸附能增大,在理论上从吸附的角度解释了元素掺杂增强g-C3N4分解水活性.总之,本文揭示了一种在分子间氢键的作用下,具有高取向性的水分子吸附的g-C3N4构型,这有助于g-C3N4基光催化剂上水分解过程的理解和优化设计.