Welcome to

Lancashire Online Knowledge

Image Credit Header image: Artwork by Professor Lubaina Himid, CBE. Photo: @Denise Swanson


Surrogate-assisted cyclic performance optimisation of direct air capture using amine-functionalised metal–organic frameworks

Nasiri-ghiri, Maryam, Nasriani, Hamid Reza orcid iconORCID: 0000-0001-9556-7218, Khajenoori, Leila orcid iconORCID: 0000-0002-1632-2296, Rasmussen, Samira Khani and Williams, Karl S orcid iconORCID: 0000-0003-2250-3488 (2026) Surrogate-assisted cyclic performance optimisation of direct air capture using amine-functionalised metal–organic frameworks. Separation and Purification Technology, 383 . p. 136177. ISSN 1383-5866

[thumbnail of VOR]
Preview
PDF (VOR)
Available under License Creative Commons Attribution.

5MB

Official URL: https://doi.org/10.1016/j.seppur.2025.136177

Abstract

Direct Air Capture (DAC) using solid sorbents has emerged as a promising technology for achieving net-negative CO₂ emissions and meeting global climate targets. Among the available sorbent materials, amine-functionalised metal–organic frameworks (MOFs) have gained significant attention due to their tuneable structures and strong affinity for CO₂ under ambient conditions. In particular, mmen-Mg2(dobpdc) has demonstrated exceptional CO₂ uptake capacity, making it a strong candidate for DAC applications. However, its process-level performance optimisation under realistic operating conditions remains insufficiently explored. This study introduces the first comprehensive multi-objective optimisation of a temperature–vacuum swing adsorption (TVSA) process employing the amine-functionalised metal–organic framework mmen-Mg₂(dobpdc) as the sorbent for direct air capture (DAC) of CO₂. The optimisation simultaneously targets minimisation of energy consumption and maximisation of CO₂ recovery and productivity, while ensuring high product purity, thereby providing new insights into the process–material interactions governing DAC performance. To achieve this, a validated dynamic temperature vacuum swing adsorption (TVSA) model was developed in Aspen Adsorption, integrated with a surrogate artificial neural network (ANN) and optimised using the Non-dominated Sorting Genetic Algorithm (NSGA-II). This approach facilitates efficient multi-objective optimisation of key process variables, significantly reducing computational time from approximately 350 days to two hours. The resulting Pareto fronts reveal clear trade-offs between specific energy consumption (SEC), recovery, and productivity at purities above 95 %. The optimised design achieved a 37 % increase in recovery, a threefold improvement in productivity, and a 14.9 % reduction in SEC, at the cost of a modest 3 % decrease in CO₂ purity (from 98 % to 95 %) compared to the base case. Moreover, the study highlights the strong influence of ambient temperature on process performance, showing that mmen-Mg2(dobpdc) exhibits enhanced CO₂ uptake below 8 °C, demonstrating its suitability for DAC operation in cool climates.


Repository Staff Only: item control page