Combination of covalent-organic frameworks and metal-organic frameworks to explore the overall picture of the structure-property relationship for dynamic adsorption cooling performance

With the development of nanoporous adsorbents, adsorption-driven cooling (AC) systems have been expected to achieve better performance by utilizing outstanding adsorbent-adsorbate working pairs. Metal-organic frameworks (MOFs) and Covalent-organic frameworks (COFs) are two classes of well-established structures that allow atomically precise integration of frameworks to build promising nanopores for AC. The adsorption cooling performance of COFs at the material level based on the thermodynamic cycle has been investigated, indicating that large capacity is the dominant factor in exhibiting a high coefficient of performance (COP). Meanwhile, the dynamic performance of COF-related working pairs at the system level awaits evaluation, especially for the specific cooling power (SCP) that is quite important during operation. Hence, we evaluated the AC performance of COF/ethanol working pairs based on the lumped parameter model, in which the COP and SCP were calculated by performing molecular simulation and mathematical simulation. It was revealed that combining the COF and MOF results presents an overall picture of the structure-property relationship, in which relatively small pore size (8–12 Å), moderate void fraction (0.6–0.8), suitable density (0.3–0.6 kg/m3), and sufficient pore volume (1–3 cm3/g) leading to high performance. Meanwhile, all the adsorption cooling performance evaluated by the lumped-parameter model is lower than the value obtained from the thermodynamic model, ascribed to the mass transfer resistance between the adsorbent and the environment was taken into consideration. From the dynamic perspective, medium transport diffusion (10−9-10−7 m2/s) with a large working capacity (0.4 g/g) and appropriate location (0.2–0.4) of the stepwise adsorption isotherm leads to better performance. It was also demonstrated that free energy can provide insightful information about the adsorption site in COF structures. With the assistance of free energy, the successful implementation of transfer learning algorithms paves the way for speeding up the assessment of the AC performance of limited nanoporous structures.