Adsorption and Desorption of Carbon Dioxide on Sodium OxideImpregnated AluminaMingheng Li,* Tong Li, Kenneth Kelley, and Cindy Pan

Department of Chemical and Materials Engineering, California State Polytechnic University, Pomona, California 91768, UnitedStates

ABSTRACT: Adsorption and desorption characteristics of carbon dioxide onsodium oxide impregnated alumina particles are investigated using a small-scalepacked-bed reactor (PBR) at different feed concentrations (5−14%), differentbed temperatures (25−300 °C), and in the presence or absence of steam. Theadsorption capacities are calculated from the breakthrough curves measured inthe PBR. The adsorption capacities at different temperatures suggest that aphysisorption mechanism may dominate at low temperatures, while achemisorption mechanism dominates at high temperatures. The presence ofsteam enhances CO2 adsorption. Implications for adsorption-enhancedprocesses are discussed.

■ INTRODUCTION

Fossil fuel combustion supplies around 80% of energy andcontributes to the majority of CO2 emissions worldwide.1 Withrapid economic growth and industrialization in developingcountries, there has been an ever-increasing demand forenergy, which poses tremendous challenges in reducing CO2emissions. This has motivated significant efforts in developingtechniques for CO2 capture in power plants, such aspostcombustion capture, precombustion capture, and oxygen-fuel combustion.2

Preferential CO2 adsorption on solid adsorbents anddesorption at a different condition is an important methodfor CO2 capture.

3 It is fundamentally different from the liquid-based absorption approach.4−7 CO2 adsorption can be foundin both postcombustion and precombustion capture applica-tions. In the precombustion capture scheme, the fuel is firstconverted to H2 and CO2 via steam reforming/partialoxidation/gasification and water gas shift (WGS) steps.8 TheCO2 may be captured from the mixture using pressure swingadsorption (PSA), leaving pure H2 for clean energy conversion.Recently, adsorption-enhanced reforming (AER) has emergedas a novel clean fuel processing technique that enablesreforming, WGS, and H2 purification in a single unit.9

Adsorption-enhanced WGS (AEWGS) has also been inves-tigated by various groups.10−13 The adsorbents used in suchapplications should have decent CO2 capacities at elevatedtemperatures (500 °C or above for AER14 and about 200−250°C for low-temperature AEWGS15). Since adsorption occurssimultaneously with the reaction, the CO2 partial pressure seenby the adsorbent may be lower than what is suggested by thereaction stoichiometry.14 Reported adsorbents in AERapplications are mainly alkali metal carbonates and oxidessuch as calcium oxide,16,17 hydrotalcite,14 lithium zirconate,18

and alumina.19 Interested readers are referred to a reviewpaper20 and references therein for CO2 adsorbents at elevatedtemperatures.In the postcombustion capture scheme,21 the combustion

product mixture leaves the stack as flue gas after heat recovery.A typical mole fraction of CO2 in flue gas is 7.4−7.7% and12.5−12.8% for gas and coal fired furnaces, respectively.22

Since the flue gas has a total pressure of about 1 atm, thepartial pressure of CO2 is fairly small. The flue gas may passthrough emission control units (to remove particulates, SOx,NOx, etc.) and then enter the CO2 capture unit before ventingto the atmosphere. A list of common adsorbents that may beused for postcombustion CO2 capture is available in theliterature.23

In oxygen-fuel combustion, the oxygen may be supplied bycryogenic air separation, where solid adsorbents are useful forremoving CO2 and other impurities before cryogenicdistillation.24 If not removed sufficiently, these componentsmay freeze and lead to blocking and plugging issues.This work aims to investigate the characteristics of CO2

adsorption and desorption on sodium oxide impregnatedalumina under different operating conditions. The processvariables include temperature (25−300 °C), CO2 partialpressure (0.05−0.14 atm), and the presence (or lack thereof)of steam. This CO2 sorbent has a relatively lower workingtemperature as compared to hydrotalcite,14 and it has potentialapplications in adsorption-enhanced WGS and methanol

Special Issue: Christos Georgakis Festschrift

Received: August 21, 2019Revised: November 26, 2019Accepted: December 3, 2019Published: December 3, 2019

Article

pubs.acs.org/IECRCite This: Ind. Eng. Chem. Res. 2020, 59, 2642−2647

© 2019 American Chemical Society 2642 DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

reforming15 as well as flue gas CO2 cleanup. Lee et al. studiedthe same adsorbent in the temperature range of 250−450 °Cand concluded that the CO2 adsorption mechanism at suchhigh temperatures was chemisorption.19 In processes such asAEWGS, the adsorption of CO2 always occurs in the presenceof steam. Therefore, quantifying the effect of water vapor onCO2 breakthrough is deemed to be useful for process design.The studies are carried out using a small-scale packed-bedreactor (PBR).

■ EXPERIMENTAL SECTIONBoth in-house and commercially made sodium oxideimpregnated alumina particles are used in the experiments.To prepare the sorbent particles in the lab, Na2CO3 is firstdissolved in deionized water. Alumina particles from SASOLare then submerged in the solution. After soaking for 50 h, theparticles are dried and heated before being loaded into thePBR. The commercial adsorbent is ActiSorb CL2 from SudChemie, which contains 85−95% alumina and 5−15% sodiumoxide. Its Brunauer−Emmett−Teller (BET) N2 surface area is165 m2/kg.25 The experimental setup is shown in Figure 1.

Industrial grade N2 and CO2 (purity >99%) from gas cylinderssupplied by Praxair and Airgas are fed separately throughrotameters followed by two Aalborg mass flow controllers(MFCs) to maintain the desired total flow rates and CO2concentration. Deionized water is supplied into the systemusing a GE peristaltic pump and is vaporized in an evaporator.The characteristic curve of the peristaltic pump (or rpm settingvs volumetric flow) is determined offline (Figure 2). On thebasis of this calibration curve, the water feed rate is calculatedto achieve a desired steam flow into the PBR. After beingmixed, the water vapor, N2, and CO2 streams pass through astainless steel preheater powered by a Variac variable powersupply to minimize the thermal transiency. The PBR iscomprised of two stainless steel tubes (2.3 cm inner diameterby 59 cm height and 0.3 cm inner diameter by 52.5 cm height,respectively). The annular space of the reactor is packed withsodium oxide promoted alumina (the weight of which is 192 gfor lab made particles or 150 g for commercial particles).Stainless steel meshes are used at both ends of the reactor toprevent the leaking of particles into the pipe.

The reactor is wrapped with an external heater controlled bya Holfman PID temperature controller to maintain the desiredbed temperature. Two three-way valves are used to control theflow direction. At the outlet of the reactor, the mixture passesthrough a condenser and a desiccant dryer to remove waterbefore reaching the Vaisala CO2 analyzer.During the adsorption step, the gases enter the smaller tube

in the center, travel through the bed, and then reach the CO2concentration analyzer before venting. The adsorption stepflow direction is represented by the solid lines in Figure 1.During the desorption step, CO2 is turned off and pure N2enters the annular space and exits through the center tube. Theflow direction is opposite to the one in the adsorption step andis represented by the red dashed lines in Figure 1.The experimental system is monitored and controlled using

LabVIEW. Process data such as feed rates of N2 and CO2 andCO2 outlet concentration are logged in real time. Beforeconducting any adsorption experiments, the Vaisala CO2concentration analyzer is calibrated by correlating the steadystate analyzer readouts with known CO2 concentrations (basedon the set points in MFCs).Figure 3 shows the mole fraction of CO2 in the exhaust gas

recorded by the analyzer in one run at room temperature (i.e.,25 °C), which included six adsorption and desorption cycles.From the real-time data recorded by LabVIEW, the

following equation is used to calculate the capacity of thesorbate in the packed-bed reactor during the adsorption step:

∫* = −

− −

∞ i

kjjjjjj

y

{zzzzzzN F

y

yF

y ty t

t1

( )1 ( )

dCO0

N0

0N2 2 2

(1)

where FN2is the molar flow rate of the carrier N2 gas, y0 is the

inlet CO2 mole fraction, and y(t) is the exit CO2 mole fraction.Similarly, the sorbate capacity during desorption is

∫* = −

∞N F

y ty t

t( )

1 ( )dCO

0N2 2 (2)

The reported capacity in this work was calculated as theaverage of values determined by eqs 1 and 2. For each of thetest conditions, the bed was saturated and flushed at least threetimes. The averaged capacity and standard deviation at each setof operating conditions were calculated and are presented in

Figure 1. Experimental setup of the packed-bed reactor for CO2adsorption and desorption.

Figure 2. Calibration curve of the GE peristaltic water pump.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

2643

the appropriate graphs. Several factors that may contribute tothe errors are the manual operation of the mode switchbetween adsorption and desorption as well as the measurementerrors of the mass flow controllers and the CO2 analyzer.

■ RESULTS AND DISCUSSIONBreakthrough Curves and Capacities at Dry Con-

ditions. In-house made particles are used to measure theadsorption/desorption characteristics in the absence of steam.The flow rate to the reactor is 5 SLPM. Figure 4 shows that at

the same temperature and CO2 partial pressure, the adsorptioncapacity does not change much when the total flow rate variesbetween 2 and 5 SLPM. This is reasonable since the flow ratewould affect the kinetics but not the equilibrium of CO2adsorption. However, a general trend indicates that higher totalflow rates are associated with smaller variations in experimentaldata. One possible reason is that the residence time of thegases in the reactor is smaller when the flow rate is larger, thus

reducing measurement error. Therefore, 5 SLPM is chosen forthe total flow rate in all of the other experiments.Representative curves of the CO2 mole fraction measured at

the outlet of the PBR during both adsorption and desorptionin three separate runs using conditions T = 250 °C, FN2

+ FCO2

= 5 SLPM, and y0 = 10% for adsorption are shown in Figure 5.

The adsorption breakthrough and desorption elution curves inall cycles were plotted using the same initial time to examinethe repeatability of the results. The transport phenomena asrevealed by dynamic modeling of the convection-dispersion-adsorption process in PBRs26,27 are as follows. When themixture is fed to the reactor, CO2 is adsorbed on the surface ofthe adsorbate and no CO2 is detected by the analyzer initially.As the loading of CO2 on the adsorbent increases, the wave ofCO2 concentration in the gas propagates along the reactor.The breakthrough occurs when the wavefront reaches theoutlet of the reactor. Subsequently, the CO2 concentrationdetected by the analyzer gradually increases as the bedcontinues to approach saturation. When the bed reaches its fullcapacity, the process operates at a steady state in which theinlet concentration is equal to the outlet concentration. Whenthe flow is reversed with pure N2, CO2 is gradually desorbed

Figure 3. Six cycles of CO2 adsorption/desorption at T = 25 °C andy0 = 5%.

Figure 4. Adsorption capacity of CO2 on sodium oxide promotedalumina as a function of total flow rate. y0 = 5%.

Figure 5. Three (a) adsorption and (b) desorption cycles at T = 250°C and y0 = 10%.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

2644

from the adsorbate. In each cycle, the difference between thecalculated adsorption and desorption capacities is less than 5%,implying that the adsorption/desorption processes arereversible.19 For this particular case, the variation of adsorptionor desorption capacity in different cycles is also less than 5%,suggesting the cycle is repeatable as well. The breakthroughcapacity, calculated based on the time for the exhaust to reach5% of its inlet concentration, is about two-thirds of the exhaustcapacity.As previously mentioned, the adsorption/desorption dy-

namics are investigated under different temperatures and CO2partial pressures. The results are summarized in Figure 6. It

appears that the dependence of the capacity on thetemperature is not monotonic. Instead, it has a valley atabout 50 °C and a local peak at about 200 °C. In all caseswithin the scope of this study, initially the capacity is high, butas temperature increases, the capacity drops as if the systemwas following a physical adsorption mechanism. Somewherebetween 50 and 200 °C, the system appears to transition fromphysisorption to chemisorption.Thermogravimetric analysis (TGA) was used to investigate

CO2 adsorption on a different batch of homemade sodiumoxide impregnated alumina particles in the temperature rangeof 150−250 °C. The capacity was calculated as the weight gaindivided by the initial weight. The results are shown in Figure 7,which confirm the local peak of the adsorption capacityoccurring around 200 °C.The physisorption mechanism results in an inverse relation-

ship between the temperature and capacity. As the internalenergy of the adsorbate increases, the weak interactionsbetween the adsorbent surface and the adsorbate are notstrong enough for the two to remain bonded. Chemisorptionfollows a trend better represented by a quadratic function.Initially the increasing temperature helps supply the activationenergy necessary for the adsorbate and adsorbent to undergochemical bonding. These bonds are much stronger than theweak physisorption bonds, and they can withstand higheradsorbate kinetic energies. Eventually, when the temperaturebecomes too high for the adsorbent and adsorbate to remainbonded, the capacity begins to drop with increasing temper-

ature similarly to the physisorption model. This system followsa physisorption mechanism at low temperatures and reversiblechemisorption mechanism at high temperatures.19 Further-more, the superposition of the physisorption curve and thechemisorption curve is believed to result in the capacity−temperature relationship.

Effect of Steam on CO2 Adsorption and Desorption.The experimental testing of the in-house made adsorbents atvarious temperatures and inlet concentrations lasted over ayear, and some decay in performance was observed. Toinvestigate the influence of steam on CO2 adsorption/desorption, the reactor was reloaded with a new batch ofadsorbent particles from Sud Chemie. The dry run consists ofN2 gas and CO2 gas, and the wet run is the same, exceptdeionized water is fed to the system. The pump was operatedin a setting that maintained the water feed rate at about 0.8mL/min, which is equivalent to 1 SLPM steam into the PBR.In both cases, the mass of the absorbent, flow rate,temperature, and percent of CO2 (on a dry basis) were keptconstant. The experiments were performed at the followingconditions: total N2 and CO2 flow rate of 5 SLPM, bedtemperature of 250 °C, and CO2 inlet concentration of 8%(similar to typical flue gas concentrations). Each run consistsof three cycles as denoted by the different colored sections(blue, red, and green) in Figure 8. Both dry and wet runs wereplotted in the same figure with the same initial time for acomparison. In the adsorption mode, it took 114 s on averagefor the exhaust to reach 50% of the inlet concentration in thedry run, as compared to 128 s in the wet run. Moreover, thebreakthrough curves in the wet run are under those in the dryrun. These indicate that the adsorption capacity is enhanced inthe presence of steam. In the desorption mode, it took 63−64 sin both dry and wet runs for the exhaust gas to reach 50% ofthe bed saturation concentration. However, the desorptionelution curves in the wet run are above those in the dry run.The saturation capacity is calculated to be 0.42 ± 0.05 mol/kgin the dry run and 0.51 ± 0.05 mol/kg in the wet run.Despite the fact that H2O may compete with CO2 for sites

on the adsorbent (i.e., multicomponent adsorption), itspresence promotes CO2 adsorption. FTIR spectroscopystudies indicate that adsorbed H2O may react with CO2 to

Figure 6. Adsorption capacity of in-house made sodium oxidepromoted alumina as a function of temperature and CO2 partialpressure in the packed-bed reactor.

Figure 7. TGA measurement of CO2 adsorption capacity on anotherbatch of in-house made Na promoted alumina particles. Freundlichisotherms are used for the correlation.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

2645

yield adsorbed carbonic acid, which undergoes deprotonationto yield adsorbed carbonate and protonated hydroxyl groups.28

The formation of carbonic acid may result from water beingcoadsorded simultaneously with CO2 or adsorbed before CO2.This mechanism might explain the apparent enhancement ofCO2 adsorption on sodium oxide promoted alumina particlesin the presence of H2O.Implications to Adsorption-Enhanced Processes. In

industrial applications, low-temperature WGS occurs at 190−250 °C with copper/zinc oxide/alumina catalysts, while high-temperature WGS is performed at 350−450 °C with promotediron oxide catalysts.29 The local peak in the CO2 adsorptioncapacity versus temperature curve occurring around 200 °C iscompatible with the working temperature of the low-temperature shift catalysts. As temperature goes beyond 250°C, the CO2 adsorption capacity on sodium oxide promotedalumina drops significantly.19 Therefore, it may not be ideal foruse in high-temperature WGS.This work shows that the adsorption capacity of sodium

oxide promoted alumina at 200 °C is comparable to the one atroom temperature. It might be a promising adsorbent toremove CO2 from a hot waste stream without cooling to roomtemperature.

For adsorption-enhanced process applications, the break-through capacity is more meaningful than the saturationcapacity calculated from eq 1 or 2. In other words, theswitching from adsorption mode to desorption mode occursmuch earlier before the bed is saturated with CO2. Otherwise,the product stream may be contaminated with CO2. Underexperimental conditions in this work, the breakthroughcapacity for sodium oxide promoted alumina particles isroughly two-thirds of the saturation capacity (see, e.g., Figures5 and 8). The required adsorbent amount in the adsorption-enhanced processes is better estimated using the breakthroughcapacity.

■ CONCLUSIONS

This work was devoted to studying the relationships betweencapacity and temperature, flow rate, and concentration of CO2in a packed-bed reactor. The variation of the flow rate did notaffect the capacity much. At 25 °C, the physisorptionmechanism was believed to be dominant, which was reflectedin the relatively high adsorption capacity versus inlet CO2concentration relationship observed. Finally, the capacityversus temperature graph suggested that the adsorptionmechanism of CO2 on sodium oxide promoted alumina isinitially physisorption at low temperature and chemisorption athigh temperature. The capacity versus temperature curveappears to be the superposition of the individual physisorptionand chemisorption curves. The local peak of adsorptioncapacity occurring around 200 °C is confirmed by TGAmeasurement. The presence of steam promotes CO2adsorption, probably attributed to the reaction of adsorbedH2O with CO2.

■ AUTHOR INFORMATION

Corresponding Author*Tel.: +1-909-869-3668. Fax: +1-909-869-6920. E-mail:minghengli@cpp.edu.

ORCIDMingheng Li: 0000-0002-6611-425XNotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This work is partly supported by ACS-PRF and a DOEsubcontract. Undergraduate students Sevana Baghdasarian,Tito Echeverria, Sergio Lainez, Oscar Lin, Joshua Obidah,Jason Rodriguez, and Joshua Wilson are acknowledged forassistance in experimental setup and testing. Dr. Tong Liwould like to thank Professor Christos Georgakis for hisimmense support and guidance.

■ REFERENCES(1) Fossil fuel energy consumption (% of total), OECD/IEA.https://data.worldbank.org/indicator/eg.use.comm.fo.zs, 2014.(2) Damen, K.; Troost, M. v.; Faaij, A.; Turkenburg, W. AComparison of Electricity and Hydrogen Production Systems withCO2 Capture and Storage. Part A: Review and Selection of PromisingConversion and Capture Technologies. Prog. Energy Combust. Sci.2006, 32, 215−246.(3) Liang, Y.; Harrison, D. P.; Gupta, R. P.; Green, D. A.;McMichael, W. J. Carbon Dioxide Capture Using Dry Sodium-BasedSorbents. Energy Fuels 2004, 18, 569−575.

Figure 8. CO2 outlet mole fraction in three (a) adsorption and (b)desorption cycles at T = 250 °C and y0 = 8% in the presence orabsence of steam. The dotted lines are thicker than the solid lines.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

2646

(4) Taheri, M.; Dai, C.; Lei, Z. CO2 Capture by Methanol, IonicLiquid, and Their Binary Mixtures: Experiments, Modeling, andProcess Simulation. AIChE J. 2018, 64, 2168−2180.(5) Wang, C.; Luo, H.; Jiang, D.; Li, H.; Dai, S. Carbon DioxideCapture by Superbase-Derived Protic Ionic Liquids. Angew. Chem.,Int. Ed. 2010, 49, 5978−5981.(6) Rao, A. B.; Rubin, E. S. A Technical, Economic, andEnvironmental Assessment of Amine-Based CO2 Capture Technologyfor Power Plant Greenhouse Gas Control. Environ. Sci. Technol. 2002,36, 4467−4475.(7) Lei, Z.; Han, J.; Li, Q.; Chen, B. Process Intensification on theSupercritical Carbon Dioxide Extraction of Low-ConcentrationEthanol from Aqueous Solutions. Ind. Eng. Chem. Res. 2012, 51,2730−2737.(8) Kolb, G. Fuel Processing: for Fuel Cells; Wiley-VCH: Weinheim,Germany, 2008.(9) Hufton, J. R.; Mayorga, S.; Sircar, S. Sorption-EnhancedReaction Process for Hydrogen Production. AIChE J. 1999, 45,248−256.(10) Najmi, B.; Bolland, O.; Colombo, K. E. A systematic approachto the modeling and simulation of a Sorption Enhanced Water GasShift (SEWGS) process for CO2 capture. Sep. Purif. Technol. 2016,157, 80−92.(11) Pichardo, P. A.; Karagoz, S.; Manousiouthakis, V. I.; Tsotsis, T.;Ciora, R. Techno-Economic Analysis of an Intensified IntegratedGasification Combined Cycle (IGCC) Power Plant Featuring aCombined Membrane Reactor - Adsorptive Reactor (MR-AR)System. Ind. Eng. Chem. Res. 2019, DOI: 10.1021/acs.iecr.9b02027.(12) Duyar, M. S.; Farrauto, R. J.; Castaldi, M. J.; Yegulalp, T. M. Insitu CO2 Capture Using CaO/γ-Al2O3 Washcoated Monoliths forSorption Enhanced Water Gas Shift Reaction. Ind. Eng. Chem. Res.2014, 53, 1064−1072.(13) Steen, W.; Richardson, C.; Lu, Y.; Lu, H.; Rostam-Abadi, M.Sorbent Enhanced Water Gas Shift-Rethinking Carbon Capture inIGCC. Energy Procedia 2014, 63, 1686−1702.(14) Duraiswamy, K.; Chellappa, A.; Smith, G.; Liu, Y.; Li, M.Development of a High-Efficiency Hydrogen Generator for Fuel Cellsfor Distributed Power Generation. Int. J. Hydrogen Energy 2010, 35,8962−8969.(15) Li, M.; Duraiswamy, K.; Knobbe, M. Adsorption EnhancedSteam Reforming of Methanol for Hydrogen Generation inConjunction with Fuel Cell: Process Design and Reactor Dynamics.Chem. Eng. Sci. 2012, 67, 26−33.(16) Florin, N. H.; Harris, A. T. Enhanced Hydrogen Productionfrom Biomass with in situ Carbon Dioxide Capture Using CalciumOxide Sorbents. Chem. Eng. Sci. 2008, 63, 287−316.(17) Lysikov, A. I.; Trukhan, S. N.; Okunev, A. G. SorptionEnhanced Hydrocarbons Reforming for Fuel Cell Powered Gen-erators. Int. J. Hydrogen Energy 2008, 33, 3061−3066.(18) Ochoa-Fernandez, E.; Rusten, H. K.; Jakobsen, H. A.; Ronning,M.; Holmen, A.; Chen, D. Sorption Enhanced Hydrogen Productionby Steam Methane Reforming Using Li2ZrO3 as Sorbent: SorptionKinetics and Reactor Simulation. Catal. Today 2005, 106, 41−46.(19) Lee, K. B.; Beaver, M. G.; Caram, H. S.; Sircar, S.Chemisorption of Carbon Dioxide on Sodium Oxide PromotedAlumina. AIChE J. 2007, 53, 2824−2831.(20) Yong, Z.; Mata, V.; Rodrigues, A. E. Adsorption of CarbonDioxide at High Temperature - A Review. Sep. Purif. Technol. 2002,26, 195−205.(21) Aaron, D.; Tsouris, C. Separation of CO2 from Flue Gas: AReview. Sep. Sci. Technol. 2005, 40, 321−348.(22) Xu, X.; Song, C.; Miller, B. G.; Scaroni, A. W. AdsorptionSeparation of Carbon Dioxide from Flue Gas of Natural Gas-FiredBoiler by a Novel Nanoporous “Molecular Basket” Adsorbent. FuelProcess. Technol. 2005, 86, 1457−1472.(23) Samanta, A.; Zhao, A.; Shimizu, G. K. H.; Sarkar, P.; Gupta, R.Post-Combustion CO2 Capture Using Solid Sorbents: A Review. Ind.Eng. Chem. Res. 2012, 51, 1438−1463.

(24) Schmidt, W. P.; Kovak, K. W.; Licht, W. R.; Feldman, S. L.Managing trace contaminants in cryogenic air separation. AIChESpring National Meeting, Atlanta, GA, 2000.(25) Slimane, R.; Wagner, J.; Pratapas, J.; Kurek, H.; Reguraman, S.;Saveliev, A. V. Waste Heat Recovery from Corrosive IndustrialExhaust Gases. CEC-500-2014-098; California Energy CommissionReport: 2013.(26) Li, M. Dynamics of CO2 Adsorption on Sodium OxidePromoted Alumina in a Packed-Bed Reactor. Chem. Eng. Sci. 2011, 66,5938−5944.(27) Babaei Pourkargar, D.; Kamali Shahri, S. M.; Rioux, R. M.;Armaou, A. Spatiotemporal Modeling and Parametric Estimation ofIsothermal CO2 Adsorption Columns. Ind. Eng. Chem. Res. 2016, 55,6443−6453.(28) Baltrusaitis, J.; Schuttlefield, J. D.; Zeitler, E.; Jensen, J. H.;Grassian, V. H. Surface Reactions of Carbon Dioxide at the AdsorbedWater-Oxide Interface. J. Phys. Chem. C 2007, 111, 14870−14880.(29) Ratnasamy, C.; Wagner, J. P. Water gas shift catalysis. Catal.Rev.: Sci. Eng. 2009, 51, 325−440.

Industrial & Engineering Chemistry Research Article

DOI: 10.1021/acs.iecr.9b04588Ind. Eng. Chem. Res. 2020, 59, 2642−2647

2647