Combined Experimental and Theoretical Approach to Grain Boundary Engineering in Organic Semiconducting Crystals

有机半导体材料（OSC）具有颠覆性地改变许多行业的潜力，例如IT部门或光伏领域，并且确实已经在某些应用中得到了应用。与传统的无机半导体相比，它们有许多改进，比如低能量生产成本、新型材料特性和仅由地球上丰富的元素组成。长期以来，OSCs新材料的实验和理论设计工作主要集中在已知材料的增量改进上，而现代数据驱动的工作直到最近才发现它们的用途。然而，迄今为止，所有这些努力都只考虑了块状材料的特性，通常是在理想化的情况下。另一方面，现实材料总是会表现出缺陷，例如不同颗粒之间的边界，最近发现的电荷载流子的障碍和陷阱表明，任何全面的设计努力都必须包括障碍的影响。因此，在提出的工作中，我们将采用实验和理论相结合的方法来优化有机分子材料，包括由于晶界导致的传输效率低下。在理论方面，我们将明确地计算晶界电荷输运的能垒。实验上，通过扫描探针显微镜、电荷载流子输运和光电流光谱的实空间成像将实验地确定晶界高度，并将其与基于密度泛函理论和有效建模相结合的理论计算进行比较。虽然建立了扫描探针显微镜方法和输运测量来确定晶界高度，但绘制晶界是一个连续且相对缓慢的技术。相比之下，光电流光谱方法目前只用于无机量子阱，将首次用于有机材料，可以快速筛选各种材料的晶界高度。最初，我们将局限于对苝基分子进行化学修饰，因为它们已经证明可以产生可重复的高质量薄膜。因此，理论不仅将阐明电荷在界面上传输的原子细节，而且还将表明势垒与所使用的分子和界面的各种性质之间的相关性。它还将为进一步扩大对初始苝结构以外的其他分子的研究奠定基础。最后，我们将调整我们其中一人开发的机器学习方法，以包括跨颗粒传输参数。有了这个方案，它也将被修改以纳入实验数据，我们将能够有效地搜索分子设计空间，以识别用于有机半导体的有前途的分子。

Organic semiconductor materials (OSC) have the potential to disruptively transform a number of industries—such as the IT sector or the field of photovoltaics—and are indeed already being used in some applications. They offer a number of improvements over traditional inorganic semiconductors such as low energetic production costs, novel materials properties and composition of only earth-abundant elements. Experimental and theoretical design efforts of new materials for OSCs for a long time mostly focused on incremental improvements of known materials, with modern data-driven efforts only very recently finding their use. Yet, all of these efforts to date only considered the properties of the bulk material, often in an idealized setting. Realistic materials, on the other hand, will always show imperfections such as boundaries between different grains and the recent discovery of barriers and traps for charge carriers demonstrated that any comprehensive design effort will have to include the influence of barriers. In the proposed work, we will therefore employ a combined experimental and theoretical approach for the optimization of organic molecular materials including the transport inefficiencies due to grain boundaries. On the theory side, we will explicitly compute the energetic barriers to charge transport at grain boundaries. Experimentally, real-space imaging by scanning probe microscopy, charge carrier transport as well as (photocurrent-)spectroscopy will experimentally identify grain boundary heights and compare this to theory computations based on a combination of density functional theory and effective modelling. While the scanning probe microscopy method and transport measurements are established to identify grain boundary heights, it is a serial and comparably slow technique to map grain boundaries. In contrast, the photocurrent spectroscopy method has up to now only been used for inorganic quantum wells, and will be her firstly adopted for organic materials and allow a fast screening of grain boundary heights of various materials. Initially, we will restrict ourselves to chemical modifications of perylene-based molecules as they have shown to generate thin films of reproducibly high quality. Theory will thereby not only elucidate the atomistic details of charge transport across the interfaces but also show the correlation of barriers with various properties of the employed molecules and the interfaces. It will also lay the foundation for a further, expanded investigation of other molecules beyond the initial perylene structures. Finally, we will adapt the machine learning approach developed by one of us to include cross-grain transport parameters. With this scheme, which will also be modified to incorporate experimental data, we will be able to efficiently search the molecular design space to identify promising molecules for use in organic semiconductors.