(483d) Development of Adsorption Processes for the Separation of Close-Boiling Gas Mixtures

Distillation is the most widely used process for the separation of homogeneous mixtures in industry. However, it is very expensive when the components to be separated have similar volatilities. This issue is even more pronounced for highly volatile gas mixtures which do not condense at ambient conditions. For such mixtures, very high pressures or low temperatures, or a combination of both, are required in distillation systems, thus significantly increasing the cost of equipment as well as operating costs.

Efforts to increase the energy efficiency and reduce costs associated with these types of processes have primarily focused on improving or optimising existing technologies, leading to highly heat-integrated and complex flowsheets. This approach, although leading to improvements in process performance, is inherently limited in its potential to bring about energy savings, as the core technology is the same. Novel approaches are required to truly revolutionise the performance of such processes.

One possible alternative is the use of pressure or temperature swing adsorption to achieve the required separation. As separation by adsorption is based on differences in chemical interactions between the components in the mixture and a solid adsorbent, rather than volatility differences, adsorption-based flow sheets could potentially bypass bottlenecks seen in distillation processes.

In this work, we investigate the potential of adsorption processes as alternatives to cryogenic distillation-based processes for the separation of close boiling gas mixtures using a systematic two-stage approach. In the first stage, simple shortcut adsorption process models are used for quick screening and ranking of candidate adsorbents, and to identify the most appropriate cyclic process (i.e. PSA or TSA) for a given separation. Detailed dynamic adsorption column models are then used for more in-depth evaluation of promising adsorbents and separation schemes identified in the first stage. This approach allows for suitable adsorbents and separation schemes to be identified with minimal computational cost, and without recourse to cyclic adsorption experiments, which could be very costly if novel adsorbents are being investigated, and are time-consuming in any case.

The separation of CO – a major feedstock in the production of acetic acid – from syngas containing N₂ is the case study in this work. We investigate the potential of metal-organic frameworks (MOFs) – a class of highly-tuneable adsorbent materials – for high purity CO recovery from N₂ containing mixtures. Conventional cryogenic distillation-based processes for CO production are restricted to low N₂ feeds as energy costs are prohibitive for high N₂ feeds: it is difficult to separate CO and N₂ by distillation as they have very similar physical properties. Adsorption-based flowsheets could potentially result in lower energy costs for traditional feedstock, as well as potentially allowing the use of cheaper feedstock with higher N₂ content, such as the off-gas from steel production, for the production of CO.

Results from the benchmarking of existing cryogenic distillation-based technology for CO production, systematic screening of several MOFs for CO/N₂ separation, and dynamic modelling of adsorption processes for CO/N₂ separation will be presented. The effect of the selected process scheme, choice of adsorbent material and operating conditions on the energy demand and capital cost of the adsorption-based flowsheets are investigated. The performance of the adsorption-based flowsheets is compared with conventional technology.