Following the development in recent years of progressively more accurate approximations to the exchange–correlation functional, the use of density functional theory (DFT) methods to examine increasingly large and complex systems has grown, in particular for solids and other condensed matter systems. However the cost of these calculations is high, often requiring the use of specialist HPC facilities. As such, for the purpose of large-scale high-throughput screening of material properties, a hierarchy of simplified DFT methods has been proposed that allows rapid electronic structure calculation of large systems, and we have recently extended this scheme to the solid state (sol-3c). Here, we analyze the applicability and scaling of the new sol-3c DFT methods to molecules and crystals composed of light-elements, such as small proteins and model DNA-helices. Furthermore, the calculation of the electronic structure of large to very large porous systems, such as metal–organic frameworks and inorganic nanoparticles, is discussed. The new composite methods have been implemented in the CRYSTAL17 code, which efficiently implements hybrid functionals and enables routine application of the new methods to large-scale calculations of such materials with excellent performance, even with small-scale computing resources.

中文翻译：

---

用于大规模固态计算的经济高效的复合方法。

随着近年来对交换相关泛函的更精确近似的发展，密度泛函理论 (DFT) 方法在检查日益庞大和复杂的系统中的使用有所增加，特别是对于固体和其他凝聚态系统。然而，这些计算的成本很高，通常需要使用专业的 HPC 设施。因此，为了对材料特性进行大规模高通量筛选，已经提出了一种简化的 DFT 方法层次结构，允许对大型系统进行快速电子结构计算，并且我们最近将该方案扩展到固态（溶胶） -3c)。在这里，我们分析了新的 sol-3c DFT 方法对由轻元素组成的分子和晶体的适用性和缩放比例，例如小蛋白质和模型 DNA 螺旋。此外，还讨论了大到非常大的多孔系统的电子结构计算，例如金属有机骨架和无机纳米粒子。新的复合方法已在 CRYSTAL17 代码中实现，它有效地实现了混合泛函，并使新方法能够以优异的性能日常应用于此类材料的大规模计算，即使是小规模的计算资源。