(223a) Water Vapour Capture on MIL-125(Ti)_NH2 By (P)TSA Process | AIChE

(223a) Water Vapour Capture on MIL-125(Ti)_NH2 By (P)TSA Process

Authors 

Ferreira, A. - Presenter, LSRE - Laboratory of Separation and Reaction Engineering - Associate Laboratory LSRE/LCM
Silva, M., LA / LSRE - LCM
Ribeiro, A. M., LSRE - Laboratory of Separation and Reaction Engineering - Associate Laboratory LSRE/LCM
Silva, C. G., LA / LSRE - LCM
Narin, G., University of Usak
Nogueira, I., LA / LSRE - LCM
Lee, U. H., KRICT
Faria, J. L., LA / LSRE - LCM
Chang, J. S., Korea Research Institute of Chemical Technology
Rodrigues, A., LSRE - Laboratory of Separation and Reaction Engineering - Faculty of Engineering - University of Porto
Introduction

Water (H2O) scarcity is a global problem and it might worsen in the future, especially in arid and semi-arid areas. To cover this lack of potable water, several techniques are being developed. One of them is the extraction of H2O from atmospheric air by adsorption based technologies using Metal Organic Frameworks (MOFs). Some MOFs present great potential and stability under H2O adsorption/desorption cycles. In the present work, the potential of the MIL-125(Ti)_NH2 for water capture from air is assessed.

Experimental

MIL-125(Ti)_NH2 was synthesized at the Korea Research Institute of Chemical Technology (KRICT). The preparation of granulated materials was performed by wet granulation method. The adsorption equilibrium isotherms were measured in a Rubotherm magnetic suspension microbalance (precision of 0.01 mg), in batch mode. Prior to adsorption equilibrium isotherms measurements, the material was activated at 423 K during 12h under vacuum. CO2, N2, O2, and H2O vapour adsorption equilibrium isotherms were performed at 283/293 K, 303 K, and 323 K. Adsorption breakthrough experiments were conducted on a bench-scale fixed-bed unit at 298 K and 1 bar. The stainless-steel column was packed with the MOF and both ends of the column were capped with glass wood. MOF was activated by heating up to 423 K under helium for 12 h. Diffuse reflectance infrared Fourier transform (DRIFT) spectra measurements were performed using Nicolet 510-P FTIR spectrometer were performed to assess adsorption kinetics. The samples were analysed without any dilution. In a typical experiment, the sample was placed inside the chamber, which was then sealed and continuously purged with helium. MIL-125(Ti)_NH2 was activated under He flow at 423 K for 3h. Then, the sample was cooled to 298 K and a He flow saturated with water was passed over it. Infrared spectra were recorded periodically. The spectrum of pure He at ambient temperature (298 K) was used as background. Finally, a phenomenological model was developed to describe the adsorption dynamic behavior in the adsorber, packed with the shaped MIL-125(Ti)_NH2 and was validated against the bench-scale breakthrough experimental results.

Results and discussion

CO2, N2, O2, and H2O vapour adsorption equilibrium isotherms on MIL-125(Ti)_NH2 were attained. H2O vapour isotherms show type V behaviour, according IUPAC classification and is not visible any hysteresis loop, though it is difficult to confirm this without higher resolution at low partial pressures. The O2 and N2 adsorption equilibrium isotherms present an almost linear shape while the CO2 adsorption equilibrium isotherms are of Type I according to the IUPAC classification.

In the experimental H2O breakthrough curve on MIL-125(Ti)_NH2, H2O vapour was carried with pure He, at about 50% RH. Adsorption breakthrough curve of water vapor presented a dispersive front up to P/Po=0.21 and then a compressive front up to P/Po=0.5. A similar pattern was also observed during the desorption with a dry flow of helium through the bed: dispersive from P/Po=0.5 to 0.23 and then compressive, as expected from the shape of the water adsorption isotherms. This behaviour was well predicted by the developed fixed bed model,

The water breakthrough curve in co-adsorption with CO2 on MIL-125(Ti)_NH2 was also assessed. The water vapor is transported by a stream of CO2, at 40% RH, and fed into column, mixed with a dry helium (50%) stream. It starts with a dispersive front up to P/Po=0.18 followed by front a compressive up to P/Po=0.40. During desorption step, a dispersive front from P/Po=0.4 to 0.24 is presented and then compressive front, similar to the experiment without CO2. The fixed bed model can predict well the dynamic behaviour of water adsorption in the presence of CO2, without taking in consideration any competition between both. This can be explained by the strong water adsorption.

The DRIFT spectrum of dry MIL-125(Ti)_NH2 exhibited three typical vibrational bands in the 3800–3200 cm-1 region, peaking at 3500 cm−1, 3380 cm−1, and 3680 cm−1. After several cycles of adsorption/regeneration, the DRIFT spectra of the neat MIL-125(Ti)_NH2 remained mostly unchanged, which indicates that the material didn’t suffer any major structural modification upon cyclic water adsorption. This conclusion is corroborated by the textural characterization performed to the sample used in the microbalance to access the adsorption equilibrium.

To evaluate the potential of the shaped MIL-125(Ti)_NH2 as adsorbent in a type 2 Atmospheric Water Vapor Processing (AWVP) technology, a column with about 1 m3 of volume (L – 1.25 m, Ø – 1 m) was considered to operate under adiabatic conditions, packed with 345 kg of MOF, and a bed porosity of 0.4. With the aim of designing a technology of easy implementation, only two steps were considered, feed and co-current purge. The regeneration temperature was considered to be 333 K, 353 K, and 373 K, while the condensation of the purge outlet stream was proposed to be 283 K, 288 K, and 298 K, for the TSA processes. The different designs were compared in terms of their productivity (l/day). The feed was considered to be at 298 K, 1.1 bar, and a RH of 50 %. As expected the highest productivity (320 l·day-1·ton-1) was obtained for the regeneration at the highest considered value (373 K), and condensation at the lowest considered value (283 K).

Conclusions

H2O isotherms present a Type V isotherm shape, not being visible any hysteresis. The amount adsorbed measured was 20.2 mol·kg-1 at P/Po=0.83. The most adsorbed component in this adsorbent was CO2, followed by O2, and N2. H2O breakthrough history has the expected behaviour according with H2O adsorption equilibrium isotherms. First, presents a dispersive front at low P/P0, and then a compressive front. The dynamic adsorption behaviour of water adsorption in the fixed bed was well predicted by the developed model. DRIFT analysis revealed that MIL-125(Ti)_NH2 structure remained stable after several adsorption/regeneration cycles. Additionally, water adsorption kinetics seems to be fast enough, and equilibrium is achieved in about 100 min. The (P)TSA proposed process shows that MIL-125(Ti)_NH2 is a promising material for water harvesting from atmosphere, with productivity between 40 l·day-1·ton-1 and 300 l·day-1·ton-1, depending on the regeneration conditions.