Scalable Quantum Chemistry with Flexible Embedding

我们建议开发一种新的软件，用于复杂系统中反应性的分子模拟，特别是在溶剂（如水）存在下的表面化学和多相催化。我们目标科学的例子是 * 在水存在下工作的金属氧化物催化 * 污染物物种的结合（重金属或有毒有机物）转化为环境中的天然矿物质，或特殊设计的无机材料，这些材料可能会将它们从环境中去除 * 更复杂的矿物质的催化作用，现有的处理方法不能很好地处理，但具有重要的工业意义 * 通过无机表面上的有机层设计传感器和光电材料我们将结合量子化学来实现这一目标。（也称为第一原理）技术，其可以用经典的化学反应中心来处理化学反应中心。（或经验）模型的其余部分的系统（板的材料是建模的一个martefae，以及其上的溶剂层。我们方法的新奇在于，我们在许多重要方面扩展了经典建模环境与量子力学计算之间相互作用的质量。我们正在结合我们自己的一些想法，如调整现有的模型，包括自旋极化效应与想法（水分子的冻结电子密度模型）已经由其他人开发和测试。我们选择开源量子化学程序NWChem作为新功能的基础，因为它是从头开始设计的，用于高性能计算机，并且是模块化设计，这将有助于支持我们需要进行的更改。MM代码（GULP）已经在我们现有的工作中使用，并在并行机上运行良好。并行效率是设计接口代码的关键驱动力，并将为我们实现新方法的方式提供信息。从长远来看，这里开发的代码将形成更广泛系统（包括生物系统）建模框架的核心，建立在我们在过去软件项目中的经验基础上（www.chemshell.org）。

We propose to develop and a new piece of software for molecular modelling of reactivity in complex systems, in particular, surface chemistry and heterogeneous catalysis in the presence of solvent such as water.Examples of the science we are targetting are * metal oxide catalysis that work in the presence or water * the binding of pollutant species (heavy metals or toxic organics) to natural minerals in the environment, or specially designed inorganic materials that could potentially remove them from the environment * catalysis by more complex minerals, not well treated by existing treatments but of great industrial importance * design of sensors and photoelectric materials by organic layers on inorganic surfacesWe will do this by combining quantum chemistry (also known as first principles) techniques, which can treat the chemically reacting centres with classical (or empirical) models for the rest of the system (the slab of material which is modelling a surgfae, and the solvent layers on top of it.The novelty in our approach is that we are extending the quality of the interaction between the classically modelled environment and the quantum mechanical calculation in a number of significant respects. We are combining some of our own ideas, such as adapting existing models to include spin polarisation effects with ideas (frozen electron density models for water molecules) that have been developed and tested by others.We are choosing an Open Source quantum chemistry program, NWChem as the basis for the new features because it has been designed from scratch for use on high-performance computers, and is modular in design which will help support the changes we need to make. The MM code (GULP) is already used in our existing work, and works well on parallel machines. Parallel efficiency is a key driver for the design of the interfaced code and will inform the way we implement the new methods.In the long run, it is intended the code developed here will form the nucleus of a modelling framework for a wider range of systems (including biological ones) building on experience we have had in past software projects (www.chemshell.org).