Controlling Metal Organic Framework (MOF) Crystallization Using High Throughput Synthesis and Computer Vision

Abstract

Co2(dobdc) is a metal organic framework (MOF) that is essentially used in gasseparation. Problems with current synthetic solvothermal routes involve a lack of creativity, the obtaining of elongated needle-shaped crystals along the c-axis with a high aspect ratio (AR), and slow synthetic approaches. Herein, we propose automating the synthesis of Co2(dobdc) MOFs using the OT-2 liquid handling robot (Mara) to improve experimental design for higher throughput, improved accuracy and increased efficiency. By doing so, we introduce strategies to create Co2(dobdc) MOFs with lower ARs in accelerated high throughput manner and characterize crystals with novel computer vision algorithm, Bok Choy. AR is a unit-less measurement for the fraction of the width by length of a crystal. Lower ARs are significantly important for MOFs to grant more control over porosity, augment surface-to-volume ratio, and enhance the characteristics of MOFs such as colloidal dispersion, diffusion, catalysis, mechanical strength, and anisotropic properties. With these strategies, we solve a set of drawbacks in synthesizing MOFs in lab such as low throughput, safety risks, human error and time consumption.