| 학회 |
한국화학공학회 |
| 학술대회 |
2019년 가을 (10/23 ~ 10/25, 대전컨벤션센터) |
| 권호 |
25권 2호, p.1695 |
| 발표분야 |
열역학 (Thermodynamics) |
| 제목 |
Construction of MOF Structures from Topology and Building Blocks |
| 초록 |
We have developed a python library for the construction of metal-organic framework (MOF) structures from the topology and the building blocks. The library includes a parser for the topology information file of “cgd” format from the RCSR database. The topology structures are relaxed according to the given building blocks. For the relaxation of the topology, we have implemented the objective function using the automatic differentiation of Tensorflow 2 to use the gradient information. The gradients obtained from the automatic differentiation significantly improve the speed of optimization of the topology. For that reason, this library can handle topologies of a large number of vertices and edges without limitations. The orientation and permutation searching algorithm developed in this work can locate a building block to the assigned node in any kind of topologies while the previous works on MOF construction only offer limited kind of topologies. From the results, this library generated almost infinite number of MOF structures from the finite number of building blocks and topologies. We expect that the result significantly influences screening works on MOFs and MOF applications. |
| 저자 |
이상원1, 김지한2
|
| 소속 |
1한국과학기술원, 2한국과학기술원 생명화학공학과 |
| 키워드 |
열역학; 분자모델링 및 전산모사
|
| E-Mail |
|
| 원문파일 |
초록 보기 |
