(223d) An Equilibrium-Based Shortcut Temperature Swing Adsorption (TSA) Model for Screening Adsorbents for Heavy Component Recovery

1. Introduction

The chemical industries are seeking efficient separation technologies that can improve resource- and energy-efficiency, reduce greenhouse gas emissions, maximise process yield and increase profit. Metal organic frameworks (MOFs) are attracting attention as adsorbents that could outperform conventional technologies such as distillation. However, with over 80,000 possible MOFs to choose from, the development of novel adsorption processes applying MOFs is challenging. The performance of adsorption processes depends significantly on the chosen adsorbent, and in turn on the process configuration and operating conditions. To unlock the full potential of adsorptive processes for separation, an integrated approach for identifying appropriate adsorbents, separation configurations and their operating conditions is required.

The literature on adsorbent screening focuses on pressure (or vacuum) swing adsorption (P/VSA) systems, rather than temperature swing adsorption (TSA) which present opportunities for heat integration but generally have lower productivity than PSA processes because of the time required for heating and cooling the bed. The lack of screening approaches for TSA adsorbents means that novel TSA processes cannot easily be identified. This work addresses this gap in capabilities by developing a shortcut TSA model for quick screening of adsorbents for recovering the more strongly-adsorbing component (A) of a binary (A/B) mixture. The model may also be used for parametric analysis to identify suitable operating conditions. The model is applied to screen adsorbents for the recovery of CO – a major feedstock in the production of acetic acid – from syngas mixtures primarily containing CO and N₂.

2. Methods

The new shortcut model considers a four-step TSA cycle with heating, cooling, pressurisation and adsorption steps. The model mirrors the model of Maring and Webley (2013) for screening adsorbents for application in pressure swing adsorption processes.

The initial condition is a bed saturated with gas at the feed composition (after breakthrough). This initial condition is selected because it is the point at which most information is known about the bed (i.e. temperature and gas-phase composition). The heating step follows, where one end of the column is open and the bed is heated (indirectly) to the desorption temperature. This step is modelled as a ‘frozen’ front: the step is divided into a discrete number of temperature steps without considering the rate of change of temperature. Material and energy balances are solved simultaneously for each temperature step to determine the amount of A and B recovered in each step, and the gas-phase composition. These values provide the initial guess for the next temperature step, until the final desorption temperature is reached. The cooling step which follows is modelled similarly to the heating step. Both ends of the bed are closed while the bed is cooled to the adsorption temperature. As no material flows into or out of the bed, the bed pressure reduces as cooling progresses. A pressurisation step follows, where the feed mixture enters the bed at one end, restoring it to the adsorption pressure. This step is also solved as a frozen front; the material and energy balances are solved simultaneously to determine the bed temperature, gas-phase composition and amount fed at each intermediate step. Finally, in the feed step, both ends of the bed are opened and the feed is introduced to the bed; the weakly adsorbed component (B) is recovered until breakthrough occurs. As the bed conditions at the start and end of the feed step are known, the amount of fresh feed and of component B recovered during the feed step can be calculated from a material balance. The solution procedure for the shortcut model is implemented in MATLAB.

3. Results and discussion

To validate the shortcut model, post-combustion capture of CO₂ from a binary CO₂/ N₂ mixture, earlier investigated by Joss et al., (2017), is modelled using a detailed column model. The CO₂ purity and recovery predicted by the shortcut and detailed models are within 5%, suggesting that the shortcut model provides a meaningful representation of the actual process performance.

The shortcut model is also used to screen several MOFs for CO/N₂ (A/B) separation. Several combinations of adsorption and desorption temperatures were simulated in the range 300–550 K to recover CO from a feed at atmospheric pressure comprising 60 mol% CO. The performance of each candidate material was compared to that of a commercial MOF, HKUST-1, which achieved a maximum CO purity and recovery of 78% and 68% respectively in the investigated range. Ni-MOF-74, a 1-D hexagonal-structured MOF, performed best, recovering up to 95% of CO at purities of up to 99 mol%. The better performance of Ni-MOF-74 correlates with the higher density of open-metal sites to which CO binds preferentially.

4. Conclusions

The new shortcut TSA model enables candidate adsorbents to be screened and ranked for recovering a strongly adsorbing component from a binary mixture, considering process conditions. The shortcut model is shown to predict process performance reasonably well, with predictions within 5% of those of a detailed model. This work, for the first time, provides a tool for quick, systematic screening of adsorbents for application in TSA processes, which can offer benefits over PSA in some cases, but also still present significant engineering challenges. The model could be coupled with high-throughput molecular simulation and materials engineering for large-scale screening and to aid the discovery of new materials with optimal properties.

References

• J. Maring and P.A. Webley, 2013, A new simplified pressure/vacuum swing adsorption model for rapid adsorbent screening for CO2 capture applications, Int. J. Greenh. Gas Control, 15: 16–31.
• Joss, L., Gazzani, M., Mazzotti, M., 2017. Rational design of temperature swing adsorption cycles for post-combustion CO₂ Chemical Engineering Science 158 (2017) 381–394