Adaptable Porous Materials

多孔材料在存储、分离和催化技术中有着广泛的应用。刚性多孔材料（如沸石或活性炭）对客体分子的吸附是由孔的固定大小和形状控制的。大自然用蛋白质催化化学过程和操纵分子。蛋白质的特点是对环境的适应性反应，是通过从能量低和动力学可达的状态的大集合中选择适当的功能结构（例如酶催化，或离子通道中小电导的机械敏感通道的孔隙打开）产生的。这是由于多肽链具有多种扭转作用，从而可以折叠成所需的结构。该项目团队最近（Rabone etal Science 2010）使用一种简单的二肽连接物通过金属结合组装了一个晶体多孔框架。所得到的材料结合了预先形成的孔隙和构象选择所需的肽连接体的自由度。这种基于肽的开放框架显示出适应性强的孔隙度，可根据客体内容从开放结构不断演变为部分无序的封闭结构。功能多孔行为是非常规的，当孔隙形貌随占用宿主孔隙体积的客体数量而变化时，表现出合作相互作用的合作反馈特征。多肽基材料显示适应性孔隙度。由于存在许多可达的吸附态，它们之间的能垒较低，因此在客体载荷作用下会发生动态结构变化。这种能量格局的出现是因为肽连接体结构的扭转变化所需的能量较低。这开启了设计适应性强的多孔材料的可能性，这种材料以一种类似于生物分子进行构象选择的方式响应客人，由几个金属中心周围的多个氨基酸残基的模块化组装产生，以进入大型和功能多样化的单元细胞。这一愿景目前还无法实现，因为这些材料中含有大量潜在的化学成分，而且缺乏计算工具来理解和预测它们对客人的反应。这种材料将是前所未有的，并提供新的和潜在有用的吸附和催化功能。拟议的研究旨在开发工具，通过专注于分别含有两个和三个氨基酸残基的化学简单二肽和三肽连接剂衍生的适应性多孔材料来分离这些材料。我们将确定个体肽连接剂和金属基单位的特征，它们具有适应性强的多孔行为。这将允许开发第二代系统，其中使用多个肽和金属单元，在功能多孔固体中演示模块化氨基酸残基组装方面取得关键进展。刚性连接体将与肽一起引入，以产生刚性亚基被柔性肽基单元重新定位的结构，类似于蛋白质折叠中刚性螺旋和片状单元的重新定位。考虑到可能的肽成分的多样性，包括机器学习在内的基于描述符的计算方法将被开发为第二代和第三代家族中合成靶标选择的补充方法。适应性多孔材料对客体的响应不遵循经典模型，将从实验吸附、动力学和结构数据以及适用于客体周围适应性多孔主体动态重构的计算模型进行评估，以超越目前合成多孔材料中主客体相互作用的静态观点。

Porous materials find widespread application in storage, separation and catalytic technologies. The sorption of a guest molecule by a rigid porous material such as a zeolite or active carbon is controlled by the fixed size and shape of the pores.Nature catalyses chemical processes and manipulates molecules using proteins. Proteins are characterised by an adaptable response to their environment, produced by conformational selection of an appropriate functional structure (e.g. for enzyme catalysis, or pore opening of the mechanosensitive channel of small conductance in an ion channel) from a large ensemble of energetically low-lying and kinetically accessible states. This is enabled by the manifold torsions available to polypeptide chains, which allow folding into the required structures.The project team have recently (Rabone et al Science 2010) used a simple dipeptide linker to assemble a crystalline porous framework through metal-binding. The resulting material combines pre-formed pores with the degrees of freedom from a peptide linker required for conformational selection. This peptide-based open framework displays adaptable porosity that evolves continuously from an open to a partially disordered closed structure in response to the guest content. The functional porous behavior is unconventional, displaying cooperative feedback characteristic of cooperative interactions as the pore topography changes in response to the number of guests occupying the pore volume of the host. The peptide-based material displays adaptable porosity. It undergoes dynamical structural changes on guest loading because there are many accessible sorption states with low energy barriers between them. This energy landscape arises because of the low energies required for torsional changes to the structure of the peptide linker.This opens up the possibility of designed adaptable porous materials which respond to guests in a manner analogous to that of a biomolecule undergoing conformational selection, produced by the modular assembly of multiple amino acid residues around several metal centres to access large and functionally diverse unit cells. This vision cannot be presently realized due to the large numbers of potential chemical constituents of such materials and the absence of computational tools to understand and predict how they would respond to guests. Such materials would though be unprecedented and offer new and potentially useful sorption and catalysis functionality.The proposed research aims to develop the tools to allow the isolation of such materials by focussing on adaptable porous materials derived from chemically simple di- and tripeptide linkers containing two and three amino acid residues respectively. We will identify the characteristics of both individual peptide linkers and metal-based units which give adaptable porous behaviour. This will allow the development of second-generation systems in which multiple peptide and metal units are used, making the key advance of demonstrating modular amino acid residue assembly in a functional porous solid. Rigid linkers will be introduced together with the peptides to produce structures where rigid sub-units are repositioned by the flexible peptide-based units, in a manner analogous to the repositioning of rigid helix and sheet units in protein folding.Given the diversity of possible peptide components, descriptor-based computational methods including machine-learning will be developed as a complementary approach to the selection of synthetic targets in the second- and third-generation families.The response of adaptable porous materials to guests does not follow classical models, and will be evaluated from experimental sorption, dynamics and structural data coupled with computational models appropriate to the dynamical restructuring of the adaptable porous host around a guest, to move beyond the current static view of host-guest interactions in synthetic porous materials.