Patel H C, Tokarski J S, Hopfinger A J
Laboratory of Molecular Modeling and Design, College of Pharmacy, University of Illinois at Chicago 60612-7231, USA.
Pharm Res. 1997 Oct;14(10):1349-54. doi: 10.1023/a:1012156318612.
The purpose of this study was to identify the key physicochemical molecular properties of polymeric materials responsible for gaseous diffusion in the polymers.
Quantitative structure-property relationships, QSPRs were constructed using a genetic algorithm on a training set of 16 polymers for which CO2, N2, O2 diffusion constants were measured. Nine physicochemical properties of each of the polymers were used in the trial basis set for QSPR model construction. The linear cross-correlation matrices were constructed and investigated for colinearity among the members of the training sets. Common water diffusion measures for a limited training set of six polymers was used to construct a "semi-QSPR" model.
The bulk modulus of the polymer was overwhelmingly found to be the dominant physicochemical polymer property that governs CO2, N2 and O2 diffusion. Some secondary physicochemical properties controlling diffusion, including conformational entropy, were also identified as correlation descriptors. Very significant QSPR diffusion models were constructed for all three gases. Cohesive energy was identified as the main correlation physicochemical property with aqueous diffusion measures.
The dominant role of polymer bulk modulus on gaseous diffusion makes it difficult to develop criteria for selective transport of gases through polymers. Moreover, high bulk moduli are predicted to be necessary for effective gas barrier materials. This property requirement may limit the processing and packaging features of the material. Aqueous diffusion in polymers may occur by a different mechanism than gaseous diffusion since bulk modulus does not correlate with aqueous diffusion, but rather cohesive energy of the polymer.
本研究的目的是确定聚合物中负责气体扩散的聚合物材料的关键物理化学分子性质。
使用遗传算法对16种聚合物的训练集构建定量结构-性质关系（QSPR），这些聚合物的二氧化碳、氮气、氧气扩散常数已被测量。在构建QSPR模型的试验基础集中使用了每种聚合物的九种物理化学性质。构建线性互相关矩阵并研究训练集成员之间的共线性。使用六种聚合物的有限训练集的常见水扩散测量值来构建“半QSPR”模型。
绝大多数情况下，发现聚合物的体积模量是控制二氧化碳、氮气和氧气扩散的主要物理化学聚合物性质。还确定了一些控制扩散的次要物理化学性质，包括构象熵，作为相关描述符。为所有三种气体构建了非常显著的QSPR扩散模型。内聚能被确定为与水扩散测量值相关的主要物理化学性质。
聚合物体积模量对气体扩散的主导作用使得难以制定通过聚合物选择性传输气体的标准。此外，预计有效的气体阻隔材料需要高体积模量。这种性能要求可能会限制材料的加工和包装特性。聚合物中的水扩散可能通过与气体扩散不同的机制发生，因为体积模量与水扩散无关，而与聚合物的内聚能有关。
