Design of Organic Electronic Materials using Predictive Modelling

我的总体项目目标是用响应性自组装有机分子取代智能设备中使用的金属。在我们的日常技术中使用金属是有问题的。这些快速消耗的金属的获取、开采和处置在经济和环境方面存在许多问题。使用有机材料作为金属的可能替代品是一种解决方案，因为它们更丰富，更便宜，并且使用的加工方法能耗更低。最近，有机基材料的例子越来越多，它们成功地用于LED等设备，并被用作光催化剂，这些材料的性能超过了它们的金属竞争对手。这表明有机物可以是一个真实的选择。然而，有机基材料的一个问题是，不仅在分子结构上，而且在分子的组装以产生不同的聚集体、组装后处理以及然后干燥成薄膜以最终制备器件方面存在如此多的可能性。这些高性能的有机物通常是偶然发现的，或者是经过多年的研究。有这么多不同的迭代相同的分子更不用说不同的分子，它可以压倒性地知道从哪里开始寻找。这是我们可以做出真实的改变的地方。为了克服在哪里开始测试新分子的这种模糊性，我将开发一个预测模型，研究人员将能够根据所需的功能缩小制备分子的范围，以及分子所需的聚合和加工。我将开发这通过定量结构-性质关系（QSPR）为基础的预测模型，使用通过高通量方法收集的数据。这将使研究人员能够快速收集有关不同分子、组装方法和条件、组装后添加剂和对齐的数据。我将从机械反应装置开始。这些设备是使用可以通过弯曲时改变其电阻率来感测运动的材料制备的。它们被用于分娩病房的分娩力计和截肢者的智能假肢等设备。我的实验室目前有材料显示在这一领域的承诺，但我想了解是什么使这些分子的工作，而其他人没有。这将通过探索这些材料的分子结构和自组装如何影响超分子结构，以及这种形态如何影响导电性和柔韧性等特性来实现。简单地观察化学结构和形态与材料性能之间的关系将非常有用，但我的目标是通过谈论所有这些信息并使用它来构建预测模型来更进一步。然后将通过探索化学空间来测试这些模型，并告诉我们哪些分子和哪些自组装方法最有可能为我们提供具有所需特性的材料，哪些不会。这些材料将被合成和测试，并再次将信息反馈到这些模型中，继续改进它们。预测模型对有机电子学领域将是无价的，因为能够预测分子的性质有可能大幅扩展该领域。目前，材料通常是先制成，然后再应用于应用程序，而在这里，我将首先从应用程序开始，并定制设计和制造，而不会浪费时间和资源在材料发现上。这将使更多的时间用于测试和进一步开发，使其与金属替代品竞争。

My overall project aim is to replace the metals used in smart devices with responsive self-assembled organic molecules. The use of metal in our everyday technologies is problematic. The acquisition, mining and disposal of these rapidly depleting metals have many issues financially and environmentally. The use of organic materials as a possible alternative to metals is a solution, as they are more abundant, less expensive and use processing methods that are less energy intensive. Recently there is an ever-expanding list of examples of organic based materials being successfully used for devices such as LEDs and being used as photocatalysts, that are outperforming their metal competitors. This shows that organics can be a real option. However, a problem with organic-based materials is that there is so much possibility not only in the molecular structure, but also in the assembly of the molecules to produce different aggregates, post-assembly processing and then drying into thin films to eventually prepare a device. These high-performing organics are often found serendipitously or after years of research. With so many different iterations of the same molecule let alone different molecules, it can be overwhelming knowing where to start to look. This is where we can make a real difference. To overcome this ambiguity in where to start testing new molecules, I will develop a prediction model where researchers will be able to narrow down both which molecules to prepare based on the desired functionality, and also the aggregation and processing required of the molecule. I will develop this through quantitative structure-property relationship (QSPR) based prediction models by using data collected through a high-throughput approach. This will enable researchers to quickly collect data on different molecules, assembly methods and conditions, post-assembly additives and alignment. I will start by focusing on mechanoresponsive devices. These devices are prepared using materials that can sense movement by changing their resistivity upon being bent. They are used in devices such as tocodynamometers on labour wards and in smart prosthetics for amputees. My lab currently has materials that show promise in this area, but I want to understand what makes these molecules work whilst others do not. This will be achieved by exploring how molecular structure and the self-assembly of these materials affect the supramolecular structure, and then how this morphology influences properties such as conductivity and flexibility. Simply observing a relationship between chemical structure and morphology to material properties will be really useful, but I aim to go a step further by talking all this information and using it to build predictive models. The models will then be tested by exploring chemical space and informing us which molecules and which self-assembly methods are most likely to give us materials with the desired properties and which will not. The suggested materials will then be synthesised and tested, and again the information fed back into these models, continuing improving them.The prediction models will be invaluable to the field of organic electronics, as being able to predict what molecules to make for their properties has the potential to expand the field drastically. Currently, materials are often made and then applied to an application afterwards, whereas here I will start with the application first and tailor the design and fabrication without wasting time and resources on material discovery. This will allow more time to be used on testing and further development to make them competitive with metal-based alternatives.