Carbon capture by physical adsorption: Materials ... · A review on carbon capture by physical adsorption is provided. The review covers carbon capture materials, experimental and

Applied Energy 161 (2016) 225–255

Contents lists available at ScienceDirect

Applied Energy

journal homepage: www.elsevier .com/locate /apenergy

Review

Carbon capture by physical adsorption: Materials, experimentalinvestigations and numerical modeling and simulations – A review

http://dx.doi.org/10.1016/j.apenergy.2015.10.0110306-2619/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +966 138604467; fax: +966 138602949.E-mail address: [email protected] (M.A. Habib).

R. Ben-Mansour, M.A. Habib ⇑, O.E. Bamidele, M. Basha, N.A.A. Qasem, A. Peedikakkal, T. Laoui, M. AliMechanical Engineering Department and KACST-TIC on CCS, KFUPM, Dhahran 31261, Saudi Arabia

h i g h l i g h t s

� A review on carbon capture by physical adsorption is provided.� The review covers carbon capture materials, experimental and numerical research.� Challenges for the post combustion adsorption materials are presented.� Gaps are found in the research of carbon dioxide adsorption of post-combustion.� Materials of high selectivity, CO2 uptake with water vapor stability are needed.

a r t i c l e i n f o

Article history:Received 4 May 2015Received in revised form 13 September2015Accepted 2 October 2015Available online 22 October 2015

Keywords:Carbon captureAdsorption techniquesPost-combustionExperimental studiesNumerical investigations

a b s t r a c t

This review focuses on the separation of carbon dioxide from typical power plant exhaust gases using theadsorption process. This method is believed to be one of the economic and least interfering ways for post-combustion carbon capture as it can accomplish the objective with small energy penalty and very fewmodifications to power plants. The review is divided into three main sections. These are (1) the candidatematerials that can be used to adsorb carbon dioxide, (2) the experimental investigations that have beencarried out to study the process of separation using adsorption and (3) the numerical models developedto simulate this separation process and serve as a tool to optimize systems to be built for the purpose ofCO2 adsorption. The review pointed the challenges for the post combustion and the experiments utilizingthe different adsorption materials. The review indicates that many gaps are found in the research of CO2

adsorption of post-combustion processes. These gaps in experimental investigations need a lot ofresearch work. In particular, new materials of high selectivity, uptake for carbon dioxide with stabilityfor water vapor needs significant investigations. The major prerequisites for these potential new materi-als are good thermal stability, distinct selectivity and high adsorption capacity for CO2 as well as suffi-cient mechanical strength to endure repeated cycling.

� 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2272. Post-Combustion carbon capture technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

2.1. Adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
2.1.1. CO2 capture using chemical sorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2292.1.2. CO2 capture using physical sorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
2.2. Adsorption process types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
3. Materials for adsorption carbon capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2303.2. Porous materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2313.3. Carbon based adsorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323.4. Solid materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

http://crossmark.crossref.org/dialog/?doi=10.1016/j.apenergy.2015.10.011&domain=pdf

http://dx.doi.org/10.1016/j.apenergy.2015.10.011

mailto:[email protected]


http://www.sciencedirect.com/science/journal/03062619

http://www.elsevier.com/locate/apenergy

Nomenclature

CF,j feed concentration of component j (mol m3)Cj gas phase concentration of component j (mol m3)Cv ;g specific heat at constant volume for gas mixture

(J kg�1 K�1)Cp;g specific heat at constant pressure for gas mixture

(J kg�1 K�1)Cs specific heat capacity of solid adsorbent (J kg�1 K�1)Cp;w specific heat capacity of adsorption column wall

(J kg�1 K�1)Dax axial dispersion coefficient (m2/s)dp adsorbent particle diameter (m)dint adsorption bed diameter (m)e adsorption bed void fraction�DHj enthalpy of component j in gas mixture (kJ/mol)hf film heat transfer coefficient between the gas and solid

adsorbent (Wm�2 K�1)hw internal convective heat transfer coefficient between

the gas and the column wall (Wm�2 K�1)KL;j overall mass transfer coefficient of component j (s�1)Ko;j adsorption constant of component j at infinite dilution

(Pa�1)l column wall thickness (m)n polytropic indexP total pressure of gas mixture (Pa)Pj partial pressure of component j in gas mixture (Pa)QF feed volumetric flow rate of gas mixture (m3/s)�qj average amount of adsorbed of component j (mol/kg)q�j the amount of component j adsorbed at equilibrium

(mol/kg)

qm;j Toth model parameter for amount of component jadsorbed in activated carbon at equilibrium (mol/kg)

t time of adsorption/desorption (s)tst stoichiometric time (s)Ts temperature of solid adsorbent (K)Tw temperature of column wall (K)Tg gas mixture temperature (K)U superficial velocity of the gas mixture (m/s)u x-component of the superficial velocity of the gas mix-

ture (m/s)v y-component of the superficial velocity of the gas mix-

ture (m/s)w z-component of the superficial velocity of the gas mix-

ture (m/s)V adsorption bed volume (m3)yj mole fraction of component j in gas mixture

Greek Lettersaw the ratio of the internal surface area to the volume of

adsorption column wall density (m�1)awl the ratio of the algorithm mean surface area of the

column shell to the volume of the column wall (m�1)e adsorption bed void fractionkL thermal conductivity of gas of the gas mixture in axial

direction (Wm�1 K�1)lg dynamic viscosity of gas mixture (Pa s�1)qg gas mixture density (kg/m3)qp adsorbent particle density (kg/m3)qw adsorption column wall density (kg/m3)

226 R. Ben-Mansour et al. / Applied Energy 161 (2016) 225–255

3.5. Adsorption of CO2 by carbon nanotubes (CNTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323.6. Metal organic frameworks (MOFs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333.7. Comparison of different CO2 adsorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

4. Experimental studies on adsorption carbon capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2344.2. Experimental studies on adsorption by MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
4.2.1. Adsorption desorption regeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2364.2.2. Adsorption and kinetic studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2364.2.3. Temperature swing adsorption methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2374.2.4. Performance in presence of water vapor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

4.3. Experimental studies on adsorption by zeolites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
4.3.1. Pressure swing adsorption process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2374.3.2. Vacuum swing adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2384.3.3. Zeolite testing under humid conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
4.4. Experimental studies on adsorption by carbon-based materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
4.4.1. Activated carbon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2394.4.2. Carbon fibre composites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
4.5. Other experimental studies on adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
4.5.1. Regeneration process techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2404.5.2. Adsorbent packing processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
4.6. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
5. Numerical investigations and mathematical models for fixed bed column adsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2415.2. Some existing mathematical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

5.2.1. CO2 in a binary mixture (with CH4, N2, H2 or He) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2425.2.2. CO2 mixture (with CH4 and H2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2445.2.3. CO2 (with Air) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2445.2.4. CO2 mixture (CO2, CO, H2, and CH4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2445.2.5. CO2 mixture (with N2 and O2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244

5.3. Modeling of adsorption of CO2 for carbon capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
5.3.1. Fixed bed adsorption model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2485.3.2. Governing equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248

R. Ben-Mansour et al. / Applied Energy 161 (2016) 225–255 227

5.4. Overview of results of numerical simulations of adsorptive carbon capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
5.4.1. A comparison of breakthrough simulation results using Linear Driving ForceModel (LDF) with breakthrough experimental result 2495.4.2. Simulated results of the breakthrough behaviour of Mg-MOF-74. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2495.4.3. Simulated results for adsorptive storage of CO2 on MOF-5 & MOF-177 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2505.4.4. Simulated results of PSA of CO2 on Mg-MOF-74 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252

1. Introduction

In May 2013, most world environmental organizations havedeclared that a critical level of carbon dioxide concentration of400 ppm was reached. This event has forced all countries, includ-ing those who were reluctant to take serious action about carbonemissions, to take unprecedented measures to reduce carbon diox-ide emissions. Fossil fuels are the dominant source of the globalprimary energy demand, and will likely remain so for the next sev-eral decades. Carbon dioxide (CO2) is regarded as one of the mainpromoters for climate change. Carbon capture (CC) is essential toenable the use of fossil fuels while reducing the emissions of CO2

into the atmosphere, and thereby mitigating global climate change.Research is needed to address technical challenges to CC such asimproved efficiency and reduced cost of CO2 capture [1]. Amongthe main sources of CO2 emissions, the road transport fieldaccounts for about 25% of CO2 emissions, while energy electricitygeneration involves 26% of the total emissions. Therefore, CO2

emissions from fixed and mobile sources should be drasticallyreduced in the forthcoming decades. Reducing CO2 emissions fromfixed and mobile sources are equally important though the mobilesources may pose more difficult challenges to be addressed. Globalpursuit of sustainable and healthy environment has been the sub-ject of the day in recent years and it cannot be overemphasized.Global warming/greenhouse effect results in increase in tempera-ture of the earth’s surface beyond the normal, leading to gross dis-comfort for inhabitants of earth. Greenhouse effect is caused bygreenhouse gases such as; carbon dioxide, nitrogen oxide, methaneand water vapor. The most predominant of these greenhouse gasesis carbon dioxide [2].

Due to the necessity of energy resources for man’s continualcomfortable living, development of energy efficient, fossil fueloperated power plants is a major task that can be used to minimizethe level of greenhouse gases emissions [1]. In addition to this,reduction of greenhouse gas emissions due to combustion of fossilfuels to the atmosphere can be further achieved through [3,4]: (i)reducing fossil fuels burning (ii) improving coal fired plant effi-ciency (iii) capture and storage of carbon dioxide and (iv) enhance-ment of CO2 partial pressure in exhaust gas. The first step might bedifficult because it entails reduction in electricity production andfinding a replacement for fossil fuels. The second step suggestedmay have insufficient effect when compared to the target of reduc-ing CO2 emission to near-zero. Hence, Herzog et al. [4] suggestedthe third step (Carbon Capture and Storage, CCS) to be a matchlessmethod that could permit continuous use and reduction of emis-sions associated with fossil fuels combustion and it would alsobuy time for the development of a new alternative to fossil fuels.The fourth step has been suggested as a means to achieve betterelectrical energy efficiency in the third step [5–7]. Carbon capturecould be executed using three methods: (i) Pre-Combustion CarbonCapture, (ii) Oxy-Combustion Carbon Capture, and (iii) Post-Combustion Carbon Capture. Oxy-Combustion Carbon Capture,instead of air, makes use of highly pure Oxygen (P95%) for fuelcombustion. Pre-Combustion Carbon Capture implies the removal

of Carbon before combustion. This method has an advantage overPost Combustion Carbon capture because it is cheaper [4]. PostCombustion capture involves the separation of CO2 fromnitrogen–carbon dioxide mixture as the main constituent of fluegas generated in power plants is nitrogen.

Research in the Carbon Capture and Sequestration (CCS) is fastgrowing. A broad variety of technologies is investigated and devel-oped by the day [8,9]. Some technologies have been developed,however most researched technology need further improvementsin terms of efficiency and associated cost reduction. The majorchallenges for CO2 capture methods are stated briefly as follows.In oxy-fuel combustion capture we are faced with (a) high energyconsumption for supply of pure oxygen and (b) the lack of fullreadiness for this technology with very little experience on a com-mercial scale. In pre-combustion capture, the challenges include(a) high cost (b) insufficient technical know-how for good operabil-ity (c) absence of single concise process for overall operationalperformance; and (d) lack of development work for industrialapplication. For post-combustion capture case, the difficultiesinclude: (a) additional energy requirement for compression of cap-tured carbon dioxide, (b) need for treatment of high gas volumes,because CO2 has low partial pressure and concentration in fluegas and (c) large energy requirement for regeneration of sorbente.g. amine solution.

A wide variety of potential methods and materials for CarbonCapture and Sequestration (CCS) applications that could beemployed in post-combustion processes are being suggested assubstitutes for the traditional chemical absorption process. Thesuggested processes comprise: the use of membranes, physicalabsorbents, adsorption of the gases on solids with the use of Tem-perature Swing or Pressure Swing (PSA/TSA) processes, hydrateformation, cryogenic distillation, and the use of metal oxides forchemical-looping combustion, and adsorption. A popular technol-ogy of post-combustion carbon capture involves the absorptionof carbon dioxide in amine solution. This method has been in useon industrial scale for quite a long time. At the same time, varietiesof some other of materials are available for other similar technolo-gies (e.g. adsorption), some of which are old while some are newlydeveloped.

Post-Combustion Carbon Capture is advantageous because ofthe following reasons:

(a) It is easier to integrate into existing plant without needing tosubstantially change the configuration/combustion technol-ogy of the plant.

(b) It is more suitable for gas plants than the Oxy-Combustionor the Pre-Combustion plants.

(c) It is flexible as its maintenance does not stop the operationof the power plant and it can be regulated or controlled.

The post-combustion CO2 capture technology is widelydeployed in chemical processing. However, the application of thistechnology to CC specific applications needs further investigationespecially in the area of optimizing CO2 capture systems for fixed

Absorption vessel Desorption

vessel

Separated CO2 (to storage and

sequestration)


and mobile sources. The priority activities in this task are: (1)development of better materials for post-combustion CO2 capture;(2) identifying optimal capture process designs and ways of inte-grating the capture systems with emitting sources to reduceenergy loss and environmental impact; (3) identifying advantagesand limitations of precipitating systems (e.g., carbonates) and (4)carrying out a detailed assessment of the environmental impactof various CO2 capture technologies.

(From plant)

Fig. 2. Schematics of absorption carbon capture process using amine.

2. Post-Combustion carbon capture technologies

A few Post-Combustion separation technologies have beenreported, some of which are; (a) absorption CO2 separation [10](b) membrane CO2 separation [11,12] (c) cryogenic CO2 separation[13] (d) Micro algal bio-fixation (e) Condensed CentrifugalSeparation [14] and (f) adsorption. Fig. 1 and the followingparagraphs briefly describe these methods [1].

Absorption of carbon dioxide (Fig. 2) is a process whereby Car-bon dioxide is taken in or embedded (absorbed) from flue gas intoan absorbent solution (e.g. amine) by chemical action, leaving theremaining gas stream to pass through the absorption column freely[15]. The dilute absorbent is re-concentrated (regenerated) forreuse in CO2 capture. CO2 absorption using amine based solventspresents a great deal of disadvantages. Some of these disadvan-tages are: (i) high heat/power requirement for solvent regenera-tion, (ii) need for corrosion control measures and (iii) thesensitivity of the solvents to losses in chemical purity/qualitydue to infiltrations from other by-products (e.g. SOx, NOx, etc.) inthe flue gas streams, which leads to reduction in efficiencies andincrement in costs of power supply [16].

Membrane separation of carbon dioxide (Fig. 3) involves the useof polymer/ceramic made membranes to sieve out the CO2 gas

Fig. 1. Post combustion car

from the flue gas. The membranes are made from polymer or cera-mic materials and their configurations are specially designed forCO2 selectivity. Challenges are still being faced in the applicationof this technique on a large scale, and in the design of membranesthat would operate efficiently for the desired purpose at relativelyhigh temperatures.

Cryogenic CO2 separation technique, Fig. 4, uses the principle ofliquid state temperature and pressure difference in constituentgases of flue gas. In this technique, CO2 is cooled and condensed,thereby removed from stream of flue gases [13].

Micro-Algae bio fixation is a potential technique for removal ofCO2 from flue gases. This technique entails the use of photosyn-thetic organisms (microalgae) for anthropogenic CO2 capture inCCS. Aquatic microalgae have been suggested to be of greaterpotential because they have higher carbon fixation rates than land

bon capture processes.

Fig. 3. Schematics of membrane carbon capture process.

Fig. 4. Schematics of cryogenic carbon capture process.


plants. Micro-algal culturing is quite expensive but the processproduces other compounds of high value that can be used forrevenue generation. Micro-algal photosynthesis also leads toprecipitation of calcium carbonate that can serve as long lastingsink for Carbon [17].

The present review is focused on carbon capture by physicaladsorption and considers materials and experimental investiga-tions. The remaining sections provide critical reviews of carboncapture methods, materials for adsorption carbon capture,experimental studies on adsorption carbon capture and numericalinvestigations and mathematical models for fixed bed columnadsorption.

2.1. Adsorption

Adsorptive separation, Fig. 5, is a mixture separating processwhich works on the principle of differences in adsorption/desorp-tion properties of the constituent of the mixture [5–7]. The wordadsorption is defined as the adhesion of ions, atoms or moleculesfrom a liquid, gas or dissolved solid to a surface. The adhered ions,

atoms or molecules form film on the surface of the materials towhich they are attached and are called adsorbate while the mate-rial on which they are attached is called the adsorbent. Adsorptionis different from absorption because in absorption, the fluid (absor-bate) is dissolved by a solid or liquid (absorbent). Adsorptionoccurs on the surface while absorption entails the whole materialvolume. Sorption is related to the two processes while desorptionis the counter reaction or reverse process of adsorption. In adsorp-tion, superficial atoms of the adsorbents are not completelyencompassed by the remaining adsorbent atoms. Adsorptionresults in surface energy due to the filling of these bonding require-ments of the adsorbent by the adsorbate atoms. The particular typeof bonding involved is a function of the involved species. Adsorp-tion may take place physically; this will involve weak van derWaals forces (physi-sorption). It may take place chemically, whichwill involve covalent bonding (chemi-sorption) and it may occurdue to electrostatic attraction.

Adsorption has a major advantage with regard to the ease ofadsorbent regeneration by thermal or pressure modulation [18],reducing the energy of Post-Combustion Carbon Capture. Son-golzadeh et al. [18] in their review of adsorbents defined adsorp-tion to be; a physical process that involves attachment of fluid tosolid surface. Important factors in adsorption include; (i) ease ofregeneration of adsorbed CO2, (ii) durability of adsorbent, (iii)selectivity of adsorbent for CO2, (iv) adsorption capacity and, (v)stability of adsorbent after several adsorption/desorption cycle[18].

Several challenges are being faced by scientists and engineersalike with respect to commercialization of these materials. This isso because the researched materials require further work toimprove their performance and stability. Suitable materials for car-bon capture must account for size of gas molecules and electronicbehavior of such molecules. There is no much difference in thekinematic diameters of gas molecules; this makes it difficult tobase CO2 separation solely on gas molecule size (CH4: 3.76 Å,CO2: 3.30 Å, N2: 3.64 Å) [8,9]. However, electronic properties likequadru-polar moment and polarization have been of great help,as bases of separation as they are significantly different for eachgas.

2.1.1. CO2 capture using chemical sorbentsIn order to overcome these challenges, a lot of research has been

carried out on advanced materials. However, despite the extent ofinvestigations, it has been difficult to find a single technology thatis able to meet the requirements set by the Department of Energy(DOE) and National Energy Technology Laboratory (NETL): i.e.below 35% increment in cost of electricity for 90% CO2 capture[8,9]. Most chemical adsorption and absorption processes, in car-bon capture/separation procedures involve the interactionbetween chemicals which lead to the formation of molecular struc-tures that are CO2-based, after which regeneration of the capturedCO2 is done through sufficient increase in temperature by heating.This procedure (i.e. regeneration) consumes most of the powerrequirement in CCS. Hence, there is a need to develop efficientmaterials and processes for CO2 capture that can greatly decreaseoperation cost through reduction in regeneration cost.

2.1.2. CO2 capture using physical sorbentsCO2 capture using physical sorbents and inorganic porous mate-

rials (e.g. carbonaceous materials and zeolites respectively) con-sumes lesser energy when compared with CCS with chemicalsorbents. This is because no new bond is formed between the sor-bate and sorbent, therefore much lesser energy is required for CO2

regeneration. Nevertheless, some well-known materials (e.g. acti-vated carbon), have the disadvantage of poor CO2/N2 selectivity.If the challenges of selectivity in physical sorbents and membranes

(a) Adsorption bed

(b)Typical experimental breakthrough curve

Fig. 5. Schematics of adsorption carbon capture process in a cylindrical bed (a) and typical breakthrough curve (b).


are successfully overcome, their use for CO2 capture could be agood potential for energy saving by the dominant amine-basedabsorption systems. Zeolites show much higher selectivity, but,they also have a disadvantage of lower CO2 loading and their effi-ciency is reduced in the presence of water [8,9]. Furthermore,molecular sieve membranes have great potentials, however, tradi-tional molecular sieves (e.g. zeolites) have restricted use in CO2/N2

separation because of similar kinetic diameters of N2 (3.64 Å) andCO2 (3.3 Å). In all, development of advanced physical adsorbentswith high CO2 capacity and selectivity is crucial. Good stability,CO2 affinity, scalability and additional required energy are majorconcerns in carbon capture research. This is crucial to the researchand development of potential carbon capture materials that willchallenge the available technologies that have been discussedabove. More attention should be paid to better understandingmolecular level gas-sorbent synergy.

2.2. Adsorption process types

It has been reported that the incurred cost in CO2 capture andits associated procedures, with the use of liquid solvent absorption,can be cut down by a great deal if adsorption separation techniqueis used [19]. Numerous technological successes have been reportedrecently in the research of adsorption carbon capture processes.Out of the researched technologies for adsorption carbon capture,two potential technologies have been considered feasible forindustrial scale CCS:

(a) Pressure/Vacuum Swing Adsorption (PSA/VSA) [20,21] Car-bon capture capacity in a PSA system is affected by two mainfactors: Adsorption selectivity and carbon dioxide workingcapacity [22]. In PSA, adsorption step is done at elevatedpressure than atmospheric pressure while in VSA adsorptionis performed at atmospheric pressure or lower.

(b) Temperature Swing Adsorption (TSA) [23,24]. In tempera-ture swing system, the adsorption bed heating is done usinga feed of hot gas or steam. Following the regeneration step isthe cooling of the adsorption bed by a feed of cold gasstream before the next adsorption step.

Of these two processes, it has been demonstrated that PSA is abetter option [23] because of (i) simplicity in application with widerange of temperature and pressure application, (ii) low energydemand and (iii) lower investment cost.

In adsorption carbon capture process, material selection pre-cedes process design. Before an adsorption process is designed,selection of suitable adsorbent, with desired properties for therequired purpose must be done. In doing this, properties such as:adsorbent selectivity, adsorption capacity, ease of and energyrequired in desorption are of great importance. In view of this, alot of research has been carried on broad species of materials suchas: synthetic zeolite, metal oxides, silica’s, carbon molecular sieves,and activated carbon.

3. Materials for adsorption carbon capture

3.1. Introduction

Different classes of Carbon capture materials have been identi-fied over the years e.g. Songolzadeh et al. [18] discussed twoclasses of CO2 adsorbents: (i) physical and (ii) chemical adsorbents.Physical adsorbents have substantial benefits for energy efficiencyin comparison with chemical and physical absorption routes. Theadsorption involves either physisorption (van der Waals) orchemisorption (covalent bonding) interaction between the gasmolecules and the surface of the material. An important factor inthe case of physical adsorbent is balancing a solid affinity forremoving the undesired component from a gas mixture with the


energy consumption required for their regeneration. Selectivity isanother factor in addition to the adsorption capacity, which is rel-evant to the adsorptive gas separation. The following mechanismare proposed for adsorptive separation: (a) the molecular sievingeffect, based on size/shape exclusion of the components in thegas mixture; (b) the thermodynamic equilibrium effect, thatdepends on the surface–adsorbate interactions; (c) the kineticeffect, due the diffusion rate differences in the gas mixture compo-nents [25].

Several physical adsorbents have been studied for CO2 captureincluding metal oxides, hydrotalcite-like compounds, microporousand mesoporous materials (including activated carbon and carbonmolecular sieves, zeolites, chemically modified mesoporous mate-rials) [26–29]. Physical adsorbents (physisorbents) are barely dis-turbed during adsorption. Pore sizes are of great importance inphysical adsorption. When pores are of size 2 nm, they are termedmicro-pores, pores of sizes between 2 and 50 nm are termed meso-pores, and when pores are of size 50 nm, they are termedmacro-pores. Materials with micro pores have better adsorptionselectivity for CO2 over CH4. Some examples of physical adsorbentsinclude activated carbon, zeolite, hydrotalcites, carbon nanotubes(CNTs), coal, etc. Activated carbon has high adsorption capacityfor CO2, high hydrophobicity, low cost, little regeneration energyrequirement and is insensitive to moisture. Zeolite on theother hand has better selectivity for CO2/N2 than carbonaceousmaterials.

Some examples of metal oxides that have been studied for car-bon capture include: calcium oxide (CaO), magnesium oxide (MgO)and lithium oxides (e.g. Li2ZrO3, Li4SiO4) [30,31]. Some examples ofmetal salts are lithium silicate and lithium zirconate, both of whichare alkali metal compounds. Magnesium oxide and calcium oxideare examples of alkali earth metal compounds. Some other exam-ples of chemical adsorbents are the hydrotalcites and double salts.During CO2 adsorption, solid compounds react with CO2 to formnew compounds e.g. Metal Carbonates. These reactions can bereversed in regenerators to harvest CO2 for storage. Metal oxidesare promising capture materials with high adsorption capacitiesat above 300 �C [32]. Lithium based oxides found recent attractionfor their high CO2 adsorption capacities [33]. Calcium oxide is ofspecial interest to researchers because it is cheap and it has highadsorption capacity for CO2 compared to lithium salts which aremore expensive especially in production. Hydrotalcites are anionicand basic clays and their derivatives are also found suitable for CO2

adsorbents at temperatures as high as 400 �C [34]. Most naturallyoccurring and well-studied hydrotalcite is Mg–Al–CO3. Hydrotal-cites have the disadvantage of high loss in adsorption capacity aftercycles of operation. During CO2 adsorption, solid compounds reactwith CO2 to form new compounds e.g. metal carbonates. Materialswith at least one dimension less than 100 nm (nanomaterials) havealso been investigated [35]. These materials have improvedstability and they maintain CO2 capturing capacity for longeradsorption/desorption cycles. However, nanomaterials have disad-vantage of high cost and complicated process of synthesis. Webb[36] stated that CO2 capture efficiency, rate of absorption, requiredregeneration energy and volume of absorber are some of the majorchallenges of CO2 absorption method. They reviewed adsorbentsand some meso-porous solid adsorbents with polyamines embed-ded in them. They stated that some factors for adsorbent selectionare rate of adsorbent, cost, and capacity of the adsorbent to adsorbCO2 and thermal stability. They identified of the following types ofadsorbents;

Chemical adsorbents e.g. amine based adsorbent. Amines weresaid to have low heat of regeneration due to low heat capacity ofsolid support. They are costly and they have low CO2 adsorptioncapacity, therefore, they are difficult to commercialize. CO2 adsorp-tion properties of amines can be improved by preparation of

support with high Amine loading, by increasing the nitrogen con-tent in amines and by improving methods of Amine introduction.Two special cases are amine impregnated adsorbents and amine-grafted adsorbents. In amine impregnated adsorbents, increasedpolyethyleneimine loading would lead to improved CO2 adsorptioncapacity, reduced surface area for adsorption, pore size and vol-ume. Therefore, it was suggested [37] that amine impregnatedadsorbents do not have thermal stability in desorption. In amine-grafted adsorbents and in order to overcome the limitations ofamine impregnated adsorbents it is suggested that CO2 adsorptioncapacity for this group of materials can be improved through silyli-sation. They can be grafted covalently to the intra-channel surfaceof meso-porous Silica. It is indicated that improvement of Amineloaded adsorbent could be improved by infusing amines intomeso-porous support with the use of effective solvents. This wastermed supercritical fluid approach. However, this group of mate-rials has disadvantages of high toxicity, low diffusivity and highviscosity. These features can lead to lower adsorption capacityand high pressure drop. Due to large volume of flue gases are tobe treated, and low partial pressure of CO2 in flue gas, chemicaladsorption would be more feasible for CO2 capture than physicaladsorption. However, it has the disadvantage of being an energyintensive process. It was indicated that that physical adsorptionis good for CO2 adsorption at high pressure and low temperature.In this light, they might not me practically applicable for post com-bustion carbon capture.

Physical adsorbents. These include activated carbon with advan-tage of enormous availability, zeolites with advantage of highlycrystalline structure, high surface area, ability to alter their compo-sition structure and ratio. They also include meso-porous silicawith advantage of high volume, surface area and tunable pore size,thermal and mechanical stability and Metal Organic Frameworks(MOFs) with advantages of very high surface area, adjustable porespaces, pore surface properties, and exceptional adsorption capac-ity for CO2. They however stated that activated carbon has disad-vantage of application to only high pressure gases, at hightemperature they have high sensitivity and low selectivity. Theyalso stated that Zeolites have very low selectivity, zeolites arehydrophilic and their CO2 adsorption capacity drops with the pres-ence of moisture in gas. The authors further mentioned that theadsorption capacity of meso-porous silica is not sufficient mostespecially at atmospheric pressure. They stated that MOFs havethe disadvantages of reduction in adsorption capacity on exposureto gas mixture and insufficient research on them, however, theyare prospective materials. Generally, CO2 capture by physical pro-cess requires less energy when compared to typical procedureusing chemical sorbents. As mentioned earlier, this is because ofthe absence of newly formed chemical bonds between the sorbateand sorbent, which reduce the energy requirement for regenera-tion [9].

3.2. Porous materials

Zeolites are the most commonly used physical adsorbents forcommercial hydrogen production using pressure swing adsorptionwith most popular zeolites 13X [26,38]. They are used at high pres-sures (above 2 bars) and their capacity is greatly reduced by thepresence of moisture in the gas; resulting in very high regenerationtemperatures [39,40]. Experimental and computational studies ofCO2 removal from flue gas using naturally occurring zeolites andother synthetic zeolites 5A and 13X indicate that synthetic zeolitesare most promising adsorbents for CO2 capture from flue gas mix-ture [39,41]. However, they experience weak adsorbent–adsorbateinteractions which are not well-suited with a high CO2/N2 selectiv-ity. The low SiO2/Al2O3 ratio and presence of cations in the zeolitestructure can enhance the adsorption. The presence of cations


leads to strong electrostatic interactions of the zeolites with CO2

[39]. Although these adsorbents are satisfactory for pressure swingadsorption, significant energy is needed for their regeneration andthat possibly leads to the disadvantages of these materials.

In the meantime, it is possible to modify these porous solidmaterials by impregnating active alkyl amines into their internalsurfaces leading to an enhancement in their gas adsorption proper-ties at low pressures. Several amine modified silica have beeninvestigated [26,42]. Carbamate species are formed throughadsorption of CO2 in the surface modified silica with primaryamines. Removal of CO2 can be performed at lower temperaturesthan those required for the regeneration of amine solvents[43,44]. A significant enhancement in the CO2 adsorption capacityis obtained through pressure swing adsorption using MCM-41 withimpregnated polyethyleneimine [45]. Amine immobilized supportsuch as poly(methyl methacrylate) has exhibited increased adsorp-tion capacities [46]. However, after impregnation, the materialssuffer from a lack of stability over repeated cycles. To increasethe stability of the materials in repeated cycles, alkylamines havebeen covalently tethered to the surface of the mesoporous support.For example, polymerization of aziridine on the surface of meso-porous silica generates a hyperbranched material which showsreversible CO2 binding and multi-cycle stability under simulatedflue gas conditions using temperature swing adsorption [42]. Thegrafted monoamino, diamino, triamino ethoxysilanes SBA-15 havebeen used to study the effect of amine and the presence of mois-ture on CO2 adsorption performance [47]. The capacity slightlydecreased for primary amine, but increased for secondary and ter-tiary amines. Although amine grafting materials show significantimprovement over non-grafted materials, it is very important thatthe amount of grafted amine be optimal for the particular CO2 cap-ture process. It is also important to study the influence of the quan-tity of grafting reagent added to the actual amount of amine that iscovalently attached to the surface.

3.3. Carbon based adsorbents

Carbon based materials such as activated carbon, charcoal andcoal have been reported for high pressure CO2 capture applications[26,48]. The key advantages of these materials are their low cost,their insensitivity to moisture and the possibility of their produc-tion/synthesis from numerous carbon based naturally existing orspent materials [49]. The activated carbon materials are amor-phous porous forms of carbon that can be prepared by pyrolysisof various carbon containing resins, fly ash, or biomass [26]. Thesematerials have lower capacities for CO2 compared with zeolites atlower pressures due to relatively uniform electric potential on thesurfaces of activated carbons leading to a lower enthalpy ofadsorption for CO2. However, their significantly high surface arealead to greater adsorption capacities at high pressures, which hasresulted in activated carbons being considered for a variety ofhigh-pressure gas separation applications. The major target appli-cation for these materials is the high-pressure flue gas produced inpre-combustion CO2 capture. Indeed, one study has shown that theupper limit for the CO2 adsorption capacity within activated car-bon materials is approximately 10–11 wt.% under post-combustion CO2 capture conditions, while it reaches 60–70 wt.%under pre-combustion CO2 capture conditions [50]. The volumetricCO2 adsorption capacity of carbon-based adsorbents is greater thansome of the highest surface area MOFs at high pressure throughthe careful selection of the material precursors and the reactionconditions employed [51].

One additional advantage of activated carbons over zeolites isthat their hydrophobic nature results in a reduced effect of thepresence of water, and they subsequently do not suffer frombreakdown or decreased capacities under hydrated conditions.

Moreover, consistent with the lower heat of adsorption for CO2,activated carbon requires a lower temperature for regenerationcompared with zeolites. Activated carbon and charcoal wereshownmoderate adsorption selectivity for CO2 over N2 at low pres-sure (below 1 atm.) and increasing the pressure reduces the selec-tivity [52]. Moreover, consistent with the lower heat of adsorptionfor CO2, activated carbons require a lower temperature for regener-ation compared with zeolites [52]. Activated carbon and charcoalwere shown moderate adsorption selectivity for CO2 over N2 atlow pressure below 1 atm. and increasing the pressure reducesthe selectivity [53]. The CO2 capture using physical adsorbentsincluding carbon based and zeolites is much more energy efficientcompared to the metal oxides and others. This is due to theabsence of the formation of new chemical bonds between the sor-bate and sorbent, thereby requiring significantly less energy forregeneration. However, the selectivity of carbon based materialsis very low, whereas zeolites exhibit significantly higher selectivitywhile they suffer from lower CO2 loading and their performance isreduced in the presence of moisture.

3.4. Solid materials

Organic calixarene compounds, for example non-porous self-assembled p-tert-butylcalix[4]arene organic solids have been con-sidered for CO2 capture [54,55]. Their structure involves cone-shaped calixarene molecules and the molecules are stabilized byintramolecular hydrogen bonds and the presence of hydrophobicnanodimensional channels [54]. The material may be suitable forhigh pressure CO2/H2 syngas separations. Other potential solidsreported for CO2 capture are covalent organic frameworks (COFs)[56]. They are microporous materials similar to MOFs but withframeworks with light weight organic components instead of themetal connectors. For example, COF-102 (C25H24B4O8) is con-structed with tetra(4-(dihydroxy)borylphenyl)methane unit andshows the highest CO2 uptake in this class (27 mmol g�1 at55 bar and 298 K) [56]. Molecular simulation studies performedon these materials predict also their exceptional high uptake[57,58].

3.5. Adsorption of CO2 by carbon nanotubes (CNTs)

The adsorption of CO2 on various carbonaceous materials suchas activated carbon [59–63] and CNTs [64–70] attracted the atten-tion of many researchers in recent years. AC, derived from differentsources of carbon materials, was the first carbon adsorbent agentused for CO2 capture [71–74]. Currently, CNTs are being consideredin this field due to their promising physical and chemical proper-ties, high thermal and electrical conductivity, along with the possi-bility to modify their surfaces chemically by adding a chemicalfunction group, using fisher esterification method, yielding highadsorption storage capacity [75–84]. These CNTs have proven tohave good potential as highly adsorbent materials for removingdifferent kinds of inorganic and organic pollutants and microor-ganisms [85–91]. The large adsorption capacity of pollutants byCNTs is mainly attributed to their surface charge densities, andwide spectrum of surface functional groups, achieved by chemicalmodification or thermal treatment to make CNTs possess optimumperformance for a particular purpose. Therefore, it is believed thata chemical modification of CNTs would also be expected to have agood potential for CO2 capture from a flue gas. However, suchstudies are still very limited in the literature. Functionalized CNTswith amino-functional groups [92–95] have been considered. Suet al. [96] investigated the effect of functionalized CNTs with3-aminopropyltriethoxysilane (APTES) at different adsorptiontemperatures. They found that by increasing the temperature ofthe system, the adsorption storage capacity decreased, while


increasing the water content increased the adsorption capacity,which reflected the exothermic process of adsorption. Their exper-imental CO2 adsorption capacity of �2.59 mmol/g at 293 K forAPTES–CNT is the evidence for the potential of CNTs as low-temperature adsorbents. Hsu et al. [66] combined vacuum andthermal adsorption system in order to trim down the regenerationtime. They were able to sustain adsorption/regeneration of CNT–APTES for twenty cycles at 493 K while maintaining the CNTs’physiochemical properties and adsorption capacity. Dillon et al.[97] functionalized the surfaces of single-walled CNTs with poly-ethylene Imine (PEI) functional group and reached a maximumadsorption capacity of 2.1 mmol/g at 300 K. The reported goodCO2 capture capacities suggest that the amine-functionalized CNTsare promising CO2 adsorbents, given that the adsorption mainlydepends on physical effects, thus relatively low energy is requiredfor the regeneration. Very few works are reported on the use ofCNTs as membrane for CO2 capture. Mixed matrix membranes ofpolyvinylalcohol containing amines with MWCNTs dispersed asmechanical reinforcing fillers demonstrated high stability for gasseparation at high pressures up to 1.5 MPa. Selectivity and perme-ability of 43 and 836 Barrers have been achieved even at such highpressures [98].

3.6. Metal organic frameworks (MOFs)

About two decades ago, a new class of materials was discov-ered; they are made of MOFs and are simply called MOFs [5–7].They are organic–inorganic hybrid, porous, solid materials. Out ofall knownmaterials to date, MOFs have the highest adsorption sur-face area per gram. They have great potentials for CO2 capture,flexible design-ability in terms of structure and function. This hasmade these materials highly used in research works of Carbon Cap-ture and Sequestration. MOFs has emerged and first synthesized byHoskins and Robson in 1989. MOFs, also known as coordinationpolymers [99] have been described as porous hybrid nano-cubesthat harness bi-properties; they establish properties of organicand inorganic porous materials. The descriptive termMOF was firstintroduced by Yaghi and co-workers in 1995. MOFs are a class ofporous crystalline materials constructed from metal-containingnodes that bonded or linked through organic ligands [9,7]. Thelinked metal and organic ligands bridges and assembled to form1D, 2D and 3D coordination network., The metal containing unitwhich is referred as secondary building units (SBUs) linked withorganic ligands using strong bonds [7]. MOFs have shown extraor-dinary porosity and can be used for wide application such as gasstorage, gas separation and catalysis. One of the most advantagesof MOFs shows its possibility of tuning the pore size from severalangstroms to nanometres by controlling the length and functional-ity of the ligands. These properties are not achievable in the case ofzeolites and porous carbon materials. The most prominent and dis-tinctive property of MOFs are its large surface area. The surfacearea, pore size and framework topology can be tuned by using dif-ferent organic building blocks and metal ions.

The metals ions can vary from transition metals to lanthanidesand even some p-block metals to form wide range of networktopologies. There are wide range of network topologies are knownand they are constructed with different combination of metal ionand the ligands. The organic linkers and metal SBUs can be variedand that leads to variety of thousands of MOFs and that numberincreasing year and year [100]. The layered zinc terephthalatewas the first proof of permanent porosity of MOF observed by mea-suring nitrogen and carbon dioxide isotherms. Later the thrust waslooking for ultrahigh porosity MOFs that can be achieved by usinglonger linkers which eventually increase the storage space andumber of adsorption sites. The longer hurdles were using thelonger linkers that always prone to form the network to undergo

interpenetration. The interpenetration can be avoided by targetingthe topology which are not prone to interpenetrate. Since theemergence of MOFs as potential material for carbon capture, alot of research has been done on MOFs.

Since MOFs provide reversible carbon dioxide adsorption, theyare excellent materials for the carbon capture. Carbon dioxideadsorption first reported using MOF-2 in 1998. The systematic car-bon dioxide adsorption study of MOF-177 with an uptake of1470 mg/g at 35 bar which exceeded that of any known porousmaterial in similar conditions. Li et al. [8], worked on carbon cap-ture using MOFs as adsorbent. CO2 adsorption in MOFs depends onpore size or volume and nature of pore surface. MOFs have higheradsorption capacity than Zeolite and activated Carbon becausethey have more surface area and larger pore size in contrast tothem. The volume and nature of pore to a great extent determinethe shape of adsorption isotherms; due to interaction betweenmolecules of CO2 leading to large condensation. Typically, MOFsare synthesized in a hydro/solvothermal reaction which involvescombination of organic ligands and metal salts in dilute solutionof polar solvents such as water, alcohol, alkyl formamides (suchas DMF, DEF) or DMSO and heated at comparatively low tempera-tures usually below 50–300 �C. The solvent utilized in the synthe-sis itself act as a template and the solvent can provide theframework intact and accessible porosity. It is important to gethigh quality single crystals to characterize the MOF crystals.Although solvothermal technique used extensively other tech-niques also known for example slow evaporation of the solutionprecursors, layering or slow diffusion. Hydro/solvothermal tech-niques have advantage over other former techniques since theyreduce the synthesis time. The ligand properties such as ligandlength, bulkiness, bond angles, and chirality act as major factorsto determine the frame work topology of the resultant compound[101]. The synthesis of MOF also depends on the concentration,solvent polarity, pH and temperature. A minor change in the for-mer parameters can leads to poor quality crystals, lower yields oreven the formation of new structures. To improve the crystalgrowth mixed solvent are often used which also provide to tunethe polarity of the solution. Besides this standard method, someother methods have been described by researchers. These methodsinclude: The mixture of non-miscible solvents [102], spray dryingtechnic [103], an electrochemical approach [104,105], and a high-throughput approach [106] and microwave irradiation. Micro waveirradiation enables access to increased range of temperatures, itcan be used to reduce crystallization time and for controlling dis-tribution of particle size and face morphology [107,108]. Micro-wave irradiation however has a disadvantage of small crystal sizeformation, therefore difficult to get enough size crystal for singlecrystal X-ray diffraction.

Over time, several MOFs have been prepared by different groupof researchers with the aim of arriving at a suitable formulation forefficient capture of CO2. As at August 2012, a total of about 37,241MOF structures were available in the Cambridge Structure Database [109]. A typical example is MOF-177 [110] synthesized usingZn(NO3)2�6H2O and of 4,40,400-benzene-1,3,5-triyl-tri-benzoic acid(H3BTB) were dissolved in 10 mL of DEF inside a 20 mL vial. Itwas subjected to heat at temperature of 100 �C for 20 h. The solu-tion drained; the resulting clear crystals were washed in DMF andreplaced with CHCl3 three times in three days. Evacuated of thematerial was carried out at 125 �C for 6 h prior to further analysis.For proper selection of appropriate building blocks for any desiredapplication, a proper understanding of the influence of characteris-tics of the building blocks and resulting material on the adsorptionbehavior is important. Hydrothermal stability of MOFs could beestimated by exposing MOFs to steam at concentration and tem-perature more than anticipated in practical operating conditionof flue gas. A throughput apparatus could be employed for the


steaming. After which, sample materials are exposed to X-raydiffraction (XRD) examination to ascertain their structural stability[110].

MOFs could be rigid or flexible, depending on whether there isrelative movement within their frameworks or not [9]. Severalresearches have been carried out on this topic: [111–114]. Usually,rigid MOFs; MOFs that do not display movement within frame-works show adsorption isotherms that are I-shaped. However,some MOFs have bi-porous structures that have channels andcages existing together within them. This makes them havingstepwise adsorption isotherms [115] e.g. at low temperature,NiII2NiIII(_3-OH)(pba)3(2,6-ndc)1.5 (MCF-19; pba = 4-(pyridin-4-yl)benzoate, 2,6-ndc = 2,6-naphthalenedicarboxylate). Some otherMOFs with ultrahigh pores have sigmoidal isotherms at low temper-ature (close to room temperature) and high pressure e.g. Zn4O(btb)2(MOF-177, btb = benzene-1,3,5-tribenzoate), Zn4O(bdc)3 (MOF-5 orIRMOF-1, bdc = 1,4-benzenedicarboxylate), and Zn4O(bte)14/9(-bpdc)6/9 (MOF-210, bte = 4,40,400-(benzene-1,3,5-triyltris(ethyne-2,1-diyl))tribenzoate, bpdc = biphenyl-4,40-dicarboxylate). On the otherhand, flexible MOFs; MOFs that show flexible behavior due to move-ment within frameworks; display stepwise or hysteretic desorptionfor CO2 and other gases [8]. Such MOFs are said to ‘breath’ duringadsorption/desorption e.g. M(OH)(bdc) (MIL-53) series, Sc2(bdc)3etc. Flexible MOFs show great potential for selectivity and they havethe advantage of smooth increment in volume with increase in CO2

loading. The flexibility of such MOFs can be improved as it is relatedto the post added group alkyl chain length. Gate phenomenon inMOFs has been given quite attention over the years [116,117].Kitagawa et al. observed a phenomenon which was termed ‘‘gate”effect in some flexible MOFs e.g. Cu(pyrdc)(bpp) (pyrdc = pyridine-2,3 dicarboxylate, bpp = 1,3-bis(4-pyridyl)propane). This wasdescribed as an abrupt rise in adsorption isotherm at relativelylow pressure. This pressure was termed ‘‘gate” opening pressure.Saturation of the materials occurred at a different pressure. How-ever, the isotherms for desorption, did not follow reverse trace ofthe adsorption isotherm, rather, it showed a sudden drop at anotherpressure (third pressure). Gate phenomenon also noticed in[Cu(4,40-bipy)(H2O)2(BF4)2](4,40-bipy) (4,40-bipy = 4,40-bipyridine),when bared to water. Similarly, Rosseinsky et al. reported that Zn(Gly-Ala)2; a peptide base MOF; exhibited ‘‘gate” behavior atpressure of about 2 bar.

Another property, for gas adsorption, which can affect CO2

uptake capacity of MOFs, is heat of adsorption [9]. Heat of adsorp-tion can be estimated with the use of adsorption isotherms of a CCSprocess at various temperatures. This property is an important fac-tor in desorption. High of heat of adsorption brings about highenergy requirement for regeneration/desorption. Heat of adsorp-tion reduces with increase in loading. The tenability of pores inMOFs is one of the important properties that distinguish themfrom other porous materials. Often, the length of organic linkersis the major determinant of the pores size in MOFs [118]. An anal-ysis of the sorbate/framework interactions by Düren [99] showedthat one dimensional pores with sharp edges are good for gas sep-aration and gas storage at low pressure. However, this is less feasi-ble at higher pressure because of the small volume of thesepreferred energetic corner regions. This was illustrated with theinvestigation of the adsorption of pure methane and ethane in ZnMOFs of different pore morphologies (e.g. 3D cubic, 1D Rhombic,1D triangular). It was shown that at lower pressure, as the porevolume is designed smaller, the selectivity becomes better whilethe adsorption rate per unit volume becomes higher. However, sat-uration is quicker due to smaller pore volume. However, at higherpressure, there is much lower uptake because of the small porevolumes. It was concluded that adsorption in MOFs with one

dimensional pore is as a result of presence of sharp corners whichbrings about more framework atoms in the sharp corners.

Some of the ways by which CO2 uptake of MOFs have to beimproved include the following. (1) Capacity of MOFs at pressurecan be improved by introduction of metal ions like Magnesium,Cobalt, Vanadium, Titanium etc. [110,119]. (2) After-synthesis-exchange of extra framework cations inside anionic MOFs. (3)Introduction CNTs into MOFs, which could be ameliorated by addi-tion of lithium and (4) Functionalizing the pores with alkyl aminogroup.

3.7. Comparison of different CO2 adsorbents

The data of the different materials are summarized in Table 1.The table provides the different properties of CO2 uptake, surfacearea, CO2/N2 selectivity and stability in humid conditions. The dataare provided for materials of the different groups including carbon-based adsorbents, Zeolites and MOFs. The table indicates thedependence of the properties on the application pressure. It alsoindicates that some new materials are well stable in humid condi-tions. However, many materials require more development forconsideration for carbon capture of flue gases of the industrialapplications. As well, the CO2 uptake in some materials needsimprovement.

Another table (Table 2) provides a comparison of the differentmaterials group of zeolites, MOFs and activated carbon basedmaterials. It is shown that MOFS have much priority on othermaterials regarding the capacity but it is very expensive. As wellMOFs in general are not stable in humid conditions. The threegroups discussed in the table differ in terms of conductivity, ther-mal and chemical stability and possibility of tuning. The selectivityof CO2/N2 changes form low in zeolites to moderate in carbon-based absorbents and becomes high in MOFs.

4. Experimental studies on adsorption carbon capture

4.1. Introduction

Generally speaking, post-combustion carbon capture is a costlyprocess due to process challenges including many parameters.These include design of capture CO2 process and materials, struc-turing of carbon capture materials, dealing with impurities withCO2 that can cause adverse effect on capture materials. They alsoinclude CO2 storage and thermodynamics of power plants, integra-tion of heat dissipation during carbon capture with heat dissipatedin power plants, optimization of carbon capture materials withrespect to ease of recycling, rate of carbon capture, CO2 selectivityand capacity etc. [140]. Many types of MOFs and zeolites as adsor-bents for carbon capture by adsorption in post combustion werestudied in terms of CO2/N2 selectivity, adsorption capacity andbreakthrough time [22]. Furthermore, many types of MOFs studiedin literature for post combustion CO2 capture were tabulated [141]regarding to CO2 and N2 uptake and selectivity for conditionsclosed to the ambient conditions which generally mimicked thepost combustion exhaust conditions. This section presents theexperimental studies that are available for CO2 adsorption. Theseare provided in two sub-sections including adsorption by MOFsand adsorption by zeolites and other materials.

4.2. Experimental studies on adsorption by MOFs

A large number of literature investigations related to carboncapture is focused on methods and procedures for synthesis andtesting of materials for post combustion capture. MOF type

Table 1Adsorbent materials utilized for CO2 capture.

Sorbent Temp.(�C)

Pressure(kPa)

CO2 molar fraction (%) Uptake CO2 (mol/kg)

Surface area (m2/g) BET Selectivity CO2/N2 Stability in humid conditions Reference

Activated carbon basedNCLK3 25 120 – 3.5 – 30 (at 130 kPa, 323 K) – González et al. [120]NCHA29 25 120 – 2.3 – 20 (at 130 kPa, 323 K) – González et al. [120]NaSB31 25 4000 100 27 3024 – – Marco-Lozar et al. [121]KL31 25 4000 100 22 2540 – – Marco-Lozar et al. [121]KA21 25 4000 100 17.5 2156 – – Marco-Lozar et al. [121]NORIT R2030CO2 30 120 17 2.4 942 7 Plaza et al. [122]Carbon fiber

composites25 101.3 13 3.1 490.6 – – Thiruvenkatachari et al.

[123]Olive stones 50 120 14 0.61 1113 18 Hydrophobic and high

stabilityGonzález et al. [124]

Almond shells 50 120 14 0.58 822 20 Hydrophobic and highstability

González et al. [124]

No1KCla-600 25 120 50 2.03 1091 2.54 over CH4 – Gil et al. [125]No1KClb-1000 25 120 50 1.91 804 2.69 over CH4 – Gil et al. [125]No2OS-1000 25 120 50 1.83 1233 2.26 over CH4 – Gil et al. [125]Cu/Zn–16% AC 30 100 15 1.98 730.53 – – Hosseini et al. [126]Cu/Zn–20% AC 30 100 15 2.26 599.41 – – Hosseini et al. [126]Cu–20% AC 30 100 15 1.99 645.21 – – Hosseini et al. [126]

ZeoliteZeolite 13X 50 100 15 3 585.5 – – Dantas et al. [127–129]Zeolite 13X-APG 30 100 15.9 4.3 – – – Wang et al. [130]Zeolite A5 30 100 16 3 499 – – Wang et al. [130]LEZ -13X 50 101.3 – 4.6 12.7 – Stable Cho et al. [131]LEZ -A5 50 101.3 – 5.2 16.8 Stable Cho et al. [131]ZSM�5 25 120 25 0.7 – 4.6 – Hefti et al. [132]Zeolite 13X 25 120 25 4.5 – 28 – Hefti et al. [132]

MOFSHKUST-1 30 1000 20 8.07 1326 – Stable Ye et al. [133]MIL-101(Cr) 30 1000 20 7.19 2549 – Stable Ye et al. [133]Zn2(hfipbb)2(ted) 25 101.3 – 0.4545 – 40 – Xu et al. [134]CPM-5 0–25–40 105 15 3–2.3–1 – 14.2 (273 K)–16.1

(298 K)Stable for few weeks Sabouni et al. [135]

MOF-177 40 100 15 0.65 4690 3 – Mason et al. [24]Mg2-MOF-74 40 100 15 7.5 1800 63 – Mason et al. [24]IRMOF-1 25 3500 100 11.1 2833 – – Millward and Yaghi [136]IRMOF-3 25 3500 100 10.3 2160 – – Millward and Yaghi [136]IRMOF-6 25 3500 100 10.5 2516 – – Millward and Yaghi [136]IRMOF-11 25 3500 100 8.9 2096 – – Millward and Yaghi [136]HKUST-1 25 3500 100 7.3 1781 – – Millward and Yaghi [136]Zn-MOF-74 25 3500 100 7.1 816 – – Millward and Yaghi [136]MOF-505 25 3500 100 0.70 1547 – – Millward and Yaghi [136]Cu-TDPAT 25 100 10 0.59 1938 79 – Li et al. [5–7]Na-rhoZMOF 25 100 20 6.2 – 440 – Nalaparaju et al. [137]Mg-rhoZMOF 25 100 20 8 – 680 – Nalaparaju et al. [137]Al-rhoZMOF 25 100 20 8 – 590 – Nalaparaju et al. [137]MIL-53(Al) 30 1000 100 5 – 5.5 – Camacho et al. [138]MIL-100(Fe) 30 101.3 15 0.67 1894 4.6 Stable Xian et al. [139]MIL-101(Cr) 30 101.3 15 1.05 3360 5.5 Stable Xian et al. [139]

R.Ben-M

ansouret

al./Applied

Energy161

(2016)225–

255235

Table 2Comparison of different adsorbents.

Specifications Zeolites Carbon-based adsorbents MOFs

Major application H2 production High pressure CO2 adsorption flue gas CO2 separationCO2/N2 selectivity Low Moderate selectivity for CO2 over N2 HighEnergy for regeneration Significant Lower temperature for regeneration compared to

zeolites. Better energy efficiency compared to metaloxides

Limited by low temperatures forgeneration, but still low economicefficiency

Capacity Moderate Lower than zeolites at low pressures and gets high athigh pressures

High

Stability under moistureconditions

Reduced capacity Do not suffer from breakthrough or decreasedcapacity under moist conditions

Mainly unstable: improvement underresearch

Cost Low production cost Reasonable cost ExpensiveAdvantages � Large micropores/mesopores

� Medium CO2 adsorption atambient conditions

� High conductivity� High thermal and chemical stabilities� Light weight with high surface areas as well aslarge pore volumes

� Energy consumption is low

� Possibility of tuning the pore size� Large surface area

Disadvantages � Adsorb moisture, so CO2

adsorption is poor with mois-ture existence

� High energy consumption� Difficult readiness

� Low adsorption and desorption temperatures� Low CO2 uptake compared to some types of Zeo-lites and MOFs

� Has low performance at partial pres-sure of CO2

� Low economic efficiency� Synthesis is tedious and complicated� So sensitive to moisture� It is difficult to use at high tempera-tures due to destroying the MOFconstruction


UiO-66 was synthesized and evaluated by Andersen et al. [142] asadsorbent for post combustion CO2 capture using vacuum swingadsorption (VSA) process. The study focused on equilibrium iso-therm, breakthrough curves, purity, and recovery of CO2 (for 15%dry CO2 and for 15% of CO2 associated with 9% of water vapor;the remaining fraction was N2). Single adsorber column of 1.1 cmdiameter and 10.5 cm of length was used in experimental work.The gases were directed by solenoid valves while the mass flowcontrollers determined the need amounts of CO2 and N2 to mixand to purge into the adsorbent. Six steps represented the VSAcycle. These are feed pressurization, counter-current blow-down(adsorption), concurrent rinse with CO2, counter-current evacua-tion (desorption), and counter-current evacuation with nitrogenpurge (completing desorption). Equilibrium isotherms of CO2 andN2 were obtained at 303 K and 328 K for pressure increased upto 100 kPa. The results showed that the best CO2 adsorbedamounts were obtained at high pressures and low temperatures.Breakthrough curves were evaluated for three different conditionsof pressure (2 bar, 3 bar and 4 bar) and the obtained values showedthe longer time was for the higher pressure which exhibited thebetter adsorption process. Increasing the times for adsorptionand rinse processes (up to 61% and 13% of CO2 breakthrough timefor adsorption and rinse time, respectively) enhanced the recoveryand purity of CO2 up to 70% and 60%, respectively. The effect ofwater vapor was also studied through 50 consecutive cycles; itshowed that the CO2 capacity of adsorbent is reduced 25% withoutany deterioration of MOF compared to dry cases.

4.2.1. Adsorption desorption regenerationAdsorption, desorption and regeneration of CO2 in two types of

MOFs (HKUST-1 and MIL-101(Cr)) were experimentally investi-gated by Ye et al. [133]. The experimental set-up was built fromone adsorbent bed connected to two cylinders; one had mixtureof CO2 (20% by volume) and N2 and the other was filled by pureN2 (for supporting desorption process). The concentrations of efflu-ent gases from adsorbent bed were measured by a dual channel gaschromatograph fitted with a thermal conducted detector using H2

as the carrier gas. The study started focusing on the CO2 adsorptioncapacity of both HKUST-1 and MIL-101(Cr) at temperaturevaried between 30 and 200 �C and pressure up to 10 bar. Thecorresponding results showed that the maximum CO2 adsorptioncapacities were 8.07 and 7.19 mmol/g for HKUST-1 and MIL-101

(Cr), respectively, at 30 �C and 10 bar. This is attributed to the factthat the pore volume of HKUST-1 (0.58 cm3/g) is smaller than thatin MIL-101(Cr) (1.3 cm3/g), even though, the surface area of MIL-101(Cr) (2549 m2/g) was over that of HKUST-1 (1326 m2/g). Thecomparison between both MOFs was done by TSA at 25 �C foradsorption and 100 �C for desorption (with purging N2). It wasnoticed that HKUST-1 had a higher CO2 adsorption capacity(1.82 mmol/g) than MIL-101(Cr) (1.17 mmol/g) at this condition.Furthermore, HKUST-1 was exploited to compare the sorptioncapacity for TSA and VSA processes. The CO2 regeneration showedobviously that the TSA is better than VSA. The amount of CO2 des-orbed by VSA was about 1.05 mmol/g for 16 min while the desorp-tion of CO2 by TSA process was up to 1.85 mmol/g for 100 �C after6 min only. These behaviors were interpreted by the MOFs con-taining co-ordinately unsaturated metal sites (CUMs) that mightnot be efficient desorption by VSA. Xu et al. [134] synthesizedtwo types of MOFs (Zn2(hfipbb)2(ted) and Co2(hfipbb)2(ted))and only investigated the CO2 adsorption in one of them(Zn2(hfipbb)2(ted)). The study reported microporous MOFs synthe-sis, crystal structure analysis, porosity characterization and CO2

adsorption selectivity and capacity as well. For 298 K and 1 atmcondition, the equilibrium isotherms showed the maximum CO2

adsorption was about 2% (by wt.) and the selectivity rangedbetween 208 and 40 for low vacuum pressure and up to 1 atm.These values of selectivity were claimed to be higher than zeolitematerials and some MOFs as Cu-TPBTM, CuBTTri and PCN-61. Itwas observed that the adsorption heat was close to be constant(27 kJ/mol). The other results concerned with H2 adsorption andpure CO2 adsorption.

4.2.2. Adsorption and kinetic studiesAnother MOF called CPM-5 was synthesized and undergone to

CO2 adsorption equilibrium and kinetic study by Sabouni et al.[135]. Adsorption studies of carbon dioxide started by investigat-ing the adsorption equilibriums of CO2 and N2 for pressure up to105 kPa and for three different temperatures (0, 25 and 40 �C).BET instruments were used for measuring the adsorption equilibri-ums volumetrically and ASAP 2010 system equipped with software(Rate of Adsorption program) to measure CO2 adsorption rates. Theexperiments commenced with degassed process at 423 K andvacuum pressure (10–6 kPa) previous to adsorption process.Unlike many of MOFs, CPM-5 showed stable structure under Lab


conditions with relevant humidity of 62% for several weeks.Regarding to experimental isotherms at several conditions, CO2

adsorption rate was about 3 mmol/g (13.2 wt.%), 2.3 mmol/g(10.1 wt.%) and 1 mmol/g (4.3 wt.%) at 105 kPa for 273 K, 298 Kand 318 K, respectively. Moreover, the selectivity factor of CPM-5was evaluated as 14.2 for 273 K and 16.1 for 298 K. CO2 diffusivityin CPM-5 at 273 K, 289 K and 318 K for the same pressure(105 kPa) was estimated as 1.86 ⁄ 10�12 m2/s, 7.04 ⁄ 10�12 m2/sand 7.87 ⁄ 10�12 m2/s, respectively, while the maximumadsorption heat was about 36 kJ/mol. Comparison to other MOFsin the literature in terms of adsorption capacity performance, theCPM-5 showed a better CO2 adsorption performance than somekinds of MOFs as MOF-5 and MOF-177 and in the same adsorptioncapacity performance of MIL-53(Al), UMCM-150 and Ni-STA-12.However, the adsorption capacity of CPM-5 is lower than function-alized and open metal sites MOFs such as HKUST-1, Mg-MOF-74and NH2MIL-53(Al).

Fourteen different types of MOFs were investigated for captur-ing CO2 from the flue gas by Yazaydın et al. [143]. Seven types ofMOFS were synthesized, characterized and measured regardingto the adsorption properties while the other 7 types were takenfrom the literature to study their CO2 capture capability. Someexperimental and simulation work was done for this purpose;the simulation study was performed by use Grand CanonicalMonte Carlo (GCMC) at the ambient conditions (room temperatureand 0.1 bar, the normal partial pressure of CO2 in flue gas). Theexperimental work demonstrated that the best types could be usedfor CO2 adsorption were Mg/DOBDC (above 250 mg/g) followed byNi/DONDC (180 mg/g) and CO/DOBDC (140 mg/g). On the otherhand, the worst types were ZIF-8, IRMOF-3, IRMOF-1, UMCM-1and MOF-177 (all of them less than 10 mg/g). Another point wasthe reversal effect of the metal–organic (M–O) bond length, itshowed that the good captured CO2 was for lower M–O bondlength (Mg–O (1.069 Å) is better than Ni–O (2.003 Å)). The simula-tion study proved only some agreements with experimental datain the cases of the best MOFs types for CO2 pressure about 0.5and 1 bar.

4.2.3. Temperature swing adsorption methodsTwo types of MOFs (MOF-177 and Mg2-dobdc (Mg/DOBDC))

were compared to capture CO2 for post-combustion by using tem-perature swing adsorption method (TSA) [24]. Effect of tempera-ture range between 20 �C and 200 �C on CO2 caption wasinvestigated at low pressure (0.15 bar for CO2 in flue gas) to studythe equilibrium isotherms of both MOFs as well as of zeolite NaX(well known in the literature). The results showed thatMg2-dobdc exhibited the best capture performance: in term ofamount of adsorbed CO2, Mg2-dobdc adsorbed 189 mg/g at 40 �Cwhereas Zeolite NaX and MOF-177 captured about 81 and4.3 mg/g, respectively. Furthermore, the selectivity of Mg2-dobdcis the highest (148.1 at 50 �C, while 87.4 and unity for zeoliteNaX and MOF-177, respectively). In addition, the working capacityby means of desorbing amount of CO2 at higher temperaturesindicated a superior amount for Mg2-dobdc over the others. Thus,0–176 mg/g could be desorbed by Mg2-dobdc for temperaturebetween 90–120 �C and about 0–75 mg/g could be desorbed byzeolite NaX while MOF-177 did not express any positive valuesof desorbed CO2 at the same range of temperature.

4.2.4. Performance in presence of water vaporThe most issue faces the use of MOFs as the adsorbents in sep-

aration processes is the decomposition under exposure to humidair. A few researches deal with this issue because the majoritydealt with flue gas as a dry mixture gas only consists of CO2 andN2. Han and his co-workers [144] studied the stability of seventypes of MOFs (CdZrSr, Ni–Nic, La–Cu, Eu–Cu, Zn–NDC, ZnPO3

and Cu–HF) under exposure to moist air, liquid water, SO2 andNO2. They significantly emphasized on three types: Cu–HF,Zn–NDC and Ni–Nic as they had larger adsorption capacity andselectivity than the other four types. Exposing Cu–HF, Zn–NDCand Ni–Nic to liquid water and NO2 during 5 days decreased theCO2 adsorption capacity of Zn–NDC by about 30% due to partialdecomposition of organic structure, whereas, Cu–HF and Ni–Nicdid not suffer from decompositions. Oppositely, Cu–HF andNi–Nic showed decreases in CO2 adsorption capacity under expos-ing to humidity (3 days) and SO2 (2 days) while Zn–NDC expressedsome increasing in adsorption in the same exposed gases.

The best MOF type (Mg-MOF-74) also has some CO2 adsorptiondeficiency with existing of moister, unlike HKUST-1 type. The studyinvestigated by Yu and Balbuena [145] showed the decreasing ofCO2 adsorption at several conditions. For 1 bar and 298 K, the dryMg-MOF-74 could adsorb about 8.4 mmol/g of CO2 while withhydration 6.5% and 13% the CO2 adsorbed amounts were6.7 mmol/g and 5.4 mmol/g, respectively. Meanwhile, the CO2/N2

selectivity increased significantly due to drop in N2 adsorption inhydrated gas. The interpretation of CO2 decreases with existinghumidity was the strong binding energy between CO2 and co-ordinately unsaturated metal sites in MOF more than the bindingenergy between CO2 and coordination water interacting. Thereverse action (the binding energy between CO2 and coordinationwater interacting is stronger) made the HKUST-1 adsorbing moreCO2 under increasing of hydration level. IRMOF-74-III as a MOFwas covalently functionalized by anime [146] to study impact ofhumidity on the MOF construction and CO2 adsorption capacity.The anime compounds added to IRMOF-74-III were –CH3, –NH2,–CH2NHBoc, –CH2NMeBoc, –CH2NH2, and –CH2NHMe. IRMOF-74-III-CH2NH2 showed high adsorption capacity of CO2 (3.2 mmol/gat 106 kPa and 298 K) and was not affected by water vapor. Com-paring dry and wet (RH = 65%) cases of flue gas (16% CO2, and thebalance was N2), the breakthrough curves ware identical for bothcases (dry and wet by using IRMOF-74-III-CH2NH2).

4.3. Experimental studies on adsorption by zeolites

4.3.1. Pressure swing adsorption processFlue gas separation by zeolite 13X through pressure swing

adsorption process (PSA) was investigated by experimental andmathematical model at two different temperatures (50, 100 �C),Dantas et al. [129]. The experimental set-up relayed on fixed bedfilled with zeolite 13X which was undergone to four steps to rep-resent separation process namely: pressurization, flue gas feed(15% CO2, 85% N2 by volume), blowdown (depressurization), purg-ing. The gas chromatograph unit was used to measure the outletconcentrations of CO2 and N2 and mass flow controllers were usedto control the flow amount of gases during working. Pressurizationprocess was used to rise the pressure of the bed up to 1.3 bar withpurging nitrogen, and then, the mixture of CO2 and N2 was fed tothe bed at constant pressure (1.3 bar) to represent the adsorptionprocess. After CO2 saturation observed, the inlet gases was closedwith depressurization the bed down to 0.1 bar for remove adsor-bent amount of CO2. For enhancing the desorption process, someamount of nitrogen was purging to the bed under low pressure(0.1 bar), this process called purging process. The experimentaland theoretical equilibrium isotherms showed that zeolite 13Xcould adsorb 3 mmol/g of CO2 at 1 bar and 50 �C and about1 mmol/g of CO2 for 100 �C at the same pressure while the notice-able adsorbed amount of N2 was less than 0.25 mmol/g for thesame conditions. The results also showed good percentages ofCO2 recovery reached about 91.8% and 90% for temperatures 50and 100 �C (P = 1.3 bar), respectively, while the CO2 purity exhib-ited low percentages about 33.3% for 50 �C and about 36.8% for100 �C. The decrease in purity of CO2 can be solved by adding rinse


process after adsorption by purging pure amount of CO2 into theadsorbent to remove N2 and replaced by CO2. This processincreases the cycle cost, but it is a solution when the pure CO2

(above 90%) is needed. Fig. 6 and Table 3 show a schematic ofPSA and valve sequencing for different steps in the cyclerespectively.

In the PSA set up (Fig. 6), the first column (M1) is fed with fluegas at a pressure above atmospheric pressure, the packed bedselectively remove CO2 from the gas stream leaving nitrogen richeffluent to flow out from valve 7(V7). After a set time e.g. break-through, the adsorbent packed in M1 is saturated hence, it nolonger adsorbs CO2. The feed is then directed to the second column(M2). In order to regenerate the saturated bed (M1), valve 3(V3) isopened to initiate pressure drop within the bed. The induced pres-sure causes desorption of the adsorbed CO2 making the gas exitingV3 rich in CO2. A purge step is then initiated to facilitate additionalremoval of CO2 from the column. After purging, the bed pressure isrestored by pressurizing with the less adsorbed gas. These are thefour steps that make up a typical PSA cycle. At the end of a com-plete cycle additional cycles can be conducted to ensure furtherpurity of the desorbed stream.

4.3.2. Vacuum swing adsorptionThe problems associated with use vacuum swing adsorption

were investigated by Chaffee et al. [147] by improving the cycledesign with good temperature control. The adsorbent was zeolite13X to capture CO2 from flue gas (simulated by adding pure CO2

to the air). This adsorbent material was insensitive to moisture.Furthermore according to the results, the CO2 adsorption mightbe increased in the presence of H2O; N-containing hybrid materialadsorbed higher amount of CO2 than N2 (contained in feed flue

M1 M2

V5

V1 V2

V6

V7 V8

V3 V4

Fig. 6. Schematics design of two-column PSA unit.

Table 3Valve sequencing for different steps in PSA cycle.

M1 Feed Blow down Purge PressurizationV1, V7 V3 V5, V3 V1

M2 Purge Pressurization Feed Blow downV4, V6 V2 V2, V8 V4

gas). The study also claimed that the fully filled pores adsorbedby N-contains had lower CO2 caption at low temperature (roomtemperature) while significant amount of CO2 was adsorbed forhigher temperature (as 70 �C), and the vice versa for partially filledpores (open pores) by N-containing. Generally, for open poresadsorbent, increasing the gas feed temperature decreased theamount of adsorbed CO2 while increasing the feed pressureimproved the captured CO2; the optimum vacuum pressure tominimize the power used for adsorption process was 0.04 bar.

4.3.3. Zeolite testing under humid conditionsExperimental investigation of CO2 capture from wet (humid)

flue gas was studied by Li et al. [148]. Zeolite X13 was used andthe vacuum swing adsorption method was applied to study theimpact of moist flue gas (PH = 95%) on the adsorption and desorp-tion processes at 30 �C. The investigation demonstrated that theCO2 recovery reduced by 22% with existence of H2O. Furthermore,high concentration of H2O appeared during vacuum process andabout 27% of the condensed H2O was accumulated in the vacuumpump itself. A comparative experimental study between twoadsorbents (13X and A5 Zeolites) for CO2 capture by indirect ther-mal swing adsorption (indirect heating/cooling by internal heatexchanger) was studied by Mérel et al. [149]. 90% of N2 and 10%of CO2 were modeled the flue gas to pursue CO2 capturing. The Zeo-lite A5 showed the better performance than Zeolite 13X for captur-ing CO2 such as the capture rate of CO2, volumetric productivityand specific heat consumption were (+14.5%), (+22%) and (�19%),respectively, for Zeolite A5 over than 13X.

The experimental work for CO2 capture from flue gas of coalfired power plant is studied by Wang et al. [150] using zeolite13XAPG by vacuum pressure swing adsorption technique VPSA).The capture plant consisted of two units: dehumidification unitand CO2 capture unit. The dehumidification unit consisted of twocylinders filled with 156 kg of alumia for removing water vaporand the contaminants amount of SOx and NOx via temperatureswing process. The output gases of this unit were CO2 (15.5–16.5% by volume) and N2 and less than 0.5% of relative humidity.The other unit formed of three column cylinders (adsorbers) occu-pied by 261 kg of zeolite 13XAPG for representing CO2 capture unitby VPSA process. The cycle of the VPSA was quite complicated toconsist of eight steps for each adsorber such as pressurization, feed,depressurization, rinse, provided pressure equalization, blowdown,purge, and received equalization. All processes were done auto-matically by programmable logic controller and software. Theresults showed the beds reached steady state after 100 operatingcycles and the adsorption temperature raised to 323 K. The adsorp-tion isotherms announced the maximum CO2 adsorption wasabout 4.3 mmol/g comparing with 3 mmol/g with using A5 molec-ular sieve in their previous work at the same conditions (T = 303 K,P = 100 kPa). For inlet flow rate of flue gas about 32.9–45.9 Nm3/h,the CO2 recovery and purity were about 85–95% and 37–82%,respectively, with power consumed for blower and vacuum pumpabout 1.79–2.14 MJ/kgCO2 (two third of the consumed power wasby vacuum pump). The maximum CO2 productivity of the unit was0.207 molCO2/m3 adsorbent.

Zeolite 13X-APGwas utilized byWang and his co-workers [130]as the adsorbent for post combustion CO2 capture by VTSA process.Experimental and simulation investigation focused mainly on thetype of process such as TSA, VSA and VTSA that was more efficientin terms of CO2 recovery and purity. The setup consisted substan-tially of one bed heated and cooled indirectly by oil passing aroundthe adsorber. The studied flue gas had 15% of carbon dioxide byvolume while the complement percentage was nitrogen. The max-imum isotherm adsorption was about 4.3 mmol/g of CO2 at 303 Kand 100 kPa. The comparison of results among the three genera-tion methods (TSA, VSA and VTSA) illustrated that the best CO2


recovery and purity for VSA process were 78.6% and 78.4%, respec-tively, at P = 3 kPa for 5 min of evacuation and 0.15 SLPM of N2

purging while alone. TSA process without evacuation could achieve78.1% of CO2 recovery and 91.6% of CO2 purity for 443 K of desorp-tion. The cooling was at close to ambient conditions during 10 minto maximize adsorption capability. In the other hand, the com-bined processes in one process (VTSA) at 403 K of desorption tem-perature and 3 kPa of vacuum pressure could reach 98.2% and 94%of CO2 recovery and purity, respectively. Furthermore, researchersconducted with the Zeolite 13X as adsorbent for CO2 capture andthe generation processes correspondingly are shown in Table 4 toshow the ability of this material (Zeolite 13X) of adsorption CO2

at several conditions.

4.4. Experimental studies on adsorption by carbon-based materials

4.4.1. Activated carbonGonzález et al. [124,120] prepared a cheap activated carbon

from spent coffee grounds to study the potential of CO2 captureby adsorption of flue gas mimicking the post combustion CO2/N2

percentages. Two types of activated carbon obtained from spentcoffee ground were investigated in this study such as NCLK3 andNCHA29 at pressure between 0 and 120 kPa and temperature var-ied between 0, 25 and 50 �C by volumetric apparatus. The isother-mal adsorption showed NCLK3 had about 3.5 mmol/g of CO2 as amaximum adsorption at 120 kPa and 25 �C with average heat ofadsorption about 27.19 kJ/mol while NCHA29 was less efficientwith CO2 adsorption with about 2.3 mmol/g at the same conditionsand 36.42 kJ/mol of isosetric adsorption heat. The selectivity andadsorption working capacity also showed some advantages forNCLK3 over NCHA29 in which the authors claimed that NCLK3was competitive with zeolite 13X.

The main properties of the adsorbent affecting CO2 capture byadsorption was experimentally investigated by Marco-Lozar et al.[121] through comparing the adsorption performance of 17 typesof activated carbon. The different pore size distribution and densityof the adsorbent were found to play main roles of selection ofadsorbent type at proper pressure. For pressure between 0.1 and1.2 MPa and ambient temperature (post combustion case), it wasobserved that the adsorption capacity did not change much byincreasing microspore volume and it was appropriate to considerthe volume of the microspore less than 0.7 nm. However, in appli-cation that have higher operation pressure (>1.2 MPa: pre combus-tion and oxy combustion cases), the microspores volume should belarger to adsorb more amount of carbon dioxide. Regarding to den-sity, the adsorbent bed has a specific volume, so the less adsorbentdensity means a little amount of the solid material would occupythe size and that significantly reduces the overall amount ofadsorbate material. Therefore, the larger density with highadsorption capacity was preferable. Plaza et al. [122] focused on

Table 4CO2 capture by zeolite 13X.

Process Cycle steps CO2%(by vol.)

Ads./des.pressure (kPa

PSA FP,FD,DP,PUR 15 130/10PSA FP,FD, DP,PUR 8.3 303/101.3VPSA (2-stages) 1st-stage: EQ,FP,FD,EQ,DP,PUR 10.5 6.67

2st-stage: EQ,FP,FD,EQ,DP 15 13.34VSA FP,FD,DP 11.2 118/3VSA FD,PR1,PR2,EQ,RIN, DP,EQ,PR3,PR4 8–15 130/5–6VSA FP,FD,EQ,RIN1,RIN2,DP,EQ 13 172/5.07TSA 10 101VTSA FP,FD,H,DP,PUR,C 15 101/3

FP, pressurization with feed; FD, feed; RIN, rinse; EQ, pressure equalization; DP, depress

the commercial activated carbon NORIT R2030CO2 to study itsCO2 adsorption capability from flue gas (17% CO2, 83% N2 by vol-ume) and comparing some regeneration methods. The set-up ofexperimental work consisted of one adsorbent bed receiving a mix-ture CO2/N2 from two cylinders, each for one gas controlled bymass flow controller and then mixing by a helical distributor.The bed was heating by a coil around it and the outlet of the bedwas connected by pressure regular and then by dual channel chro-matograph fitted with thermal conductive detector to calibrate andmeasure the output concentrations of effluent gases (CO2 and N2).The study addressed the comparison between TSA, VSA and VTSAfor flow rate of 34 cm3/min and adsorption pressure of 130 kPaas well as the adsorption temperature was 303 K. The isothermsshowed the maximum CO2 adsorption was about 2.4 at 120 kPaand 303 K and the CO2/N2 selectivity was 7 at the same conditions.TSA announced the smallest values of the CO2 recovery and pro-ductivity by about 40% and 0.8 mmol/g hr, respectively, at thementioned adsorption conditions (T = 303 K, P = 303 kPa) since N2

purging for desorption process (at 373 K and 2.7 cm3/min). How-ever, VSA adsorption performed under the vacuum (P = 5 Pa) andtemperature about 303 K produced about 1.7 mmol/g hr of CO2

with 87% of recovery. For enhancing the performance, VTSA wasapplied to produce about 1.9 mmol/g hr and to increase the CO2

productivity up to 97% under the vacuum conditions and increas-ing temperature to 323 K.

4.4.2. Carbon fibre compositesCarbon fibre composites also promised a better CO2 capture

compared to other types of activated carbon, Thiruvenkatachariet al. [123]. It was synthesized by consolidation as a one brick.There were some small tubes put inside the material for air andwater heating and cooling during desorption and adsorption pro-cesses, respectively. Two large beds (2 m) were filled with adsor-bent for investigation the CO2 capture at ambient conditions(298 K and 1 bar) from flue gases which contained 13% CO2, 5.5%O2 and the remaining was N2. The setup controls and monitorsincluded flow mass meter, CO2 analyzer, O2 analyzer and volumemeter. The study relied firstly on temperature swing adsorptionmethod for adsorbent regeneration at T = 383 K and ambient pres-sure (1 bar) without purging any gas and then on vacuum swingadsorption for ambient temperature and 30 kPa of pressure. How-ever, the results showed the two methods were not sufficient forefficient recovery CO2 and then suggested vacuum temperatureswing adsorption for efficient regeneration process.

The maximum adsorbed CO2 showed by adsorption isothermswas 2.51–3.1 mmol/g at ambient condition which added someadvances to activated carbon CO2 capture research. Regarding todesorption techniques, TSA at 398 K and 1 bar had 100% of CO2

concentration and the CO2 recovery was less than 20% while VSAat 298 K and 30 kPa presented lower than 5% of CO2 recovery with

)Ads./des. temperature(�C)

RecoveryCO2 (%)

PurityCO2 (%)

Ref.

50–100 91.8–90 33.3–36.8 Dantas et al. [129]25 50 78 Gomes and Yee [151]30 80 99 Cho et al. [152]30 78.8 99.730 78.5 69 Li et al. [148]40 60–70 90–95 Zhang et al. [153]30 69 99.5 Choi et al. [154]15/110 56 �100 Merel et al. [155]30/90 98.5 94.4 Wang et al. [130]

urization; PUR, purge; PR, re-pressurization; H, heating; C, cooling.


higher energy consumption by vacuum pump. On the other hand,utilizing two methods simultaneously (VTSA: T = 398 K andP = 75 kPa) enhanced the performance significantly. Besides VTSA,flushing some amount of pure CO2 soon after adsorption process(for remove the amounts of adsorbed N2 and O2 from the bed)improving the CO2 recovery up to 97% with 100% of the purity.Two cheap activated carbon adsorbents were made from olivestones and almond shells with single step activation for investigat-ing CO2 adsorption separation of flue gas [124]. The study consid-ered the equilibrium isotherms at different conditions (0, 25 and50 373 K and 2.7 cm3/min of for pressure reached 120 kPa). Forolives stone carbon, the maximum adsorption of CO2 was about3.2 mmol/g for 100 kPa and 25 �C while almond shell carbonshowed about 2.5 mmol/g at the same conditions. Simulating theflue gas by 14% CO2 and 86% N2 and passing it through adsorbents,the obtained breakthrough curves determined that the break-through time of olive stone-based carbon had lower time thanalmond shell-based carbon (by 1 min out of 8 min). But, the CO2

adsorption capacity of olive stone-based carbon expressed a littlehigher value than that in the almond shell-based carbon(0.61 mmol/g for olives type and 0.58 for almond shell one at120 kPa and 50 �C). The desorption process in this study was doneby passing helium gas, because it only focused on adsorption pro-cess regardless the complete cycle methods.

4.5. Other experimental studies on adsorption

4.5.1. Regeneration process techniquesThe regeneration process (desorption) refers to the rejection of

the adsorbed amount and the best measures for its performancethat is CO2 recovery and CO2 purity. The performance of regenera-tion process techniques for purity and recovery of flue gas wassummarized as shown in Table 5, Clausse et al. [156]. It is clear fromthis table and as mentioned above [122,123,120,156] that the bestpercentages of CO2 recovery and purity above 90% were obtainedby combined processes such as pressure temperature swingadsorption (PTSA) and vacuum temperature swing adsorption(VTSA). Also, the CO2 recovery and purity reasonable percentagescan be obtained from vacuum pressure swing adsorption process.

A hybrid adsorbent consisted of monolithic activated carbonand zeolite was investigated for CO2 capturing performance usingelectrical swing adsorption technique (ESA) [157]. The holes inconsolidated activated carbon were filled by Zeolite 13X to occupyabout 82% of the volume of the bed. ESA was designed to desorbthe adsorbed amount of CO2 inside the adsorbent by electrother-mal regeneration (Joule effect) with temperature reached about460 �C. Furthermore, ESA was represented by two cases: first casewas performed by four steps such as feeding, electrothermal

Table 5Comparison among different regeneration processes in terms of CO2 purity andrecovery [156].

Process CO2 purity (%) CO2 recovery (%)

ESA 23.33 29.57VPSA 99 53–70PTSA 99 902-bed-2-step PSA 18 90VPSA 99.5–99.8 34–69PSA 99.5 69PSA/VSA 58–63 70–75VSA 90 90PSA/VSA 58 87PSA/VSA 82.7 17.43-bed VSA 90–95 60–70TSA 95 81ESA 89.7 79PSA 16 89

desorption, purging with electrothermal desorption, and coolingwhile the other case study expanded the capture cycle to six stepssuch as feeding, rinsing, electrothermal desorption, purging withelectrothermal desorption, purging and cooling. The flue gas in the-ses cycles was about 8.1% (by vol.) of CO2 and the balance is N2. Forthe same cycle time of the two cases study, the results showed thesix steps cycle had higher CO2 purity about 46.6% compared to44.8% of four step cycle due to rinse process, whereas the CO2

recovery had high percentage for four step cycle with 92.4% andthe lower was 81.4% for the six steps cycle. However, the cost ofboth was considered high compared ESA to other process tech-niques with about 44.8 GJ/tonCO2 and 33.3 GJ/tonCO2 for fourand sex steps cycles, respectively. Moreover, the hybrid adsorbentaddressed some drawbacks as enlarging the mass transfer zonedue to non-homogeneously. Some increasing in adsorbed amountof dioxide carbon and elongating the breakthrough curve weredue to existing of zeolite 13X itself with good percentage (82%).

The adsorption behavior of zeolite 13X to Methane, Nitrogenand Carbon Dioxide were investigated experimentally by Cavenatiet al. [158]. Activation of Zeolite 13X samples was done withHelium, under vacuum through the night at temperature of593 K. The samples were heated at a rate of 2 K/min while Iso-therms were measured at 293, 308 and 323 K at pressure rangeof 0–5 MPa. All of the Isotherms were made completely reversible.A Magnetic Suspension Microbalance (Rubotherm) was employedto perform adsorption equilibrium of the pure gases. The authors’data fitted with the Toth and Multisite Langmuir Model. A strongCO2 adsorption was recorded, which make them recommend Zeo-lite 13X as potential material for CO2 sequestration from flue gas.Casas et al. [159] performed breakthrough experiment, describingpre-combustion CO2 capture using MOFs (e.g. USO-2-Ni MOF)and UiO-67/MCM-41 hybrid adsorbents by Pressure Swing Adsorp-tion (PSA). MOF UiO-67/MCM-41 hybrid was designed jointly withmeso-porous silica, (i.e. MCM-41), of average sized particles: say1 mm. MCM-41 has a very good adsorption capacity, stabilizingeffect, and lower Henry’s constant. These are favorable characteris-tics for desorption at high pressure. Furthermore, the 1 mm parti-cles qualify for use at industrial level, for feasible range of resultingpressure drop. On the other hand, formulation of USO2-Ni MOFparticles is yet to be up scaled; therefore, only particles of size0.2–0.5 mm were produced. Material and particle densities werecharacterized by Helium pycnometery and Hg-pycnometeryrespectively. The material heat capacity of the two materials wasestimated with the use of a Differential Scanning Calorimeter(DSC). The authors [159] performed process scale up by first con-ducting a fixed bed experiment, during which the adsorbent waspacked into column after which three grades of CO2/H2 mixtureswere feed through them at temperature of 25 �C and pressurerange of 1–25 bar to determine the transfer parameters. In thebreakthrough experiment, it was found that the feed flow ratehad negligible impact on the mass adsorbed and heat transferredunder the considered span of conditions. Adsorption Isothermwas measured for pure components with the aid of a Magnetic Sus-pension Balance (Rubotherm), Langmuir isotherm and correspond-ing isotherm parameters were reported. A comparison was madebetween their results obtained at another result of the same pro-cess using activated carbon as adsorbent. The comparison showedthat the selectivity and productivity of the PSA (Pressure SwingAdsorption) process was increased by the introduction of USO-2-Ni MOF, compared to activated carbon. UiO-67/MCM-41 hybridshowed faintly lower selectivity, but higher specific adsorbent pro-ductivity compared to activated carbon.

4.5.2. Adsorbent packing processesFormation of particle is very important; it has a huge effect on

the adsorbent packing properties, hence, on the process


performance by Casas et al. [159]. They concluded that existingresearch on formulated MOFs with average particles size greaterthan 1 mm (that permit scaling up) is adequate to enable theirexploration for industrial scale usage. In addition, bed densityand particles, are of great importance in process design. This isbecause they are responsible for the quantity of adsorbent materi-als that can be packed in enclosed column volume. In this light, theUiO-67/MCM-41 hybrid showed good packing properties, notwithstanding, further research and improvement is required intheir mechanical stability in order to make them useable on indus-trial scale. Dantas et al. [128] worked on fixed bed CO2 adsorptionfrom a gas mixture of 20%CO2/N2. The adsorption medium usedwas activated carbon. Helium was used for pre-treatment of thebed. Break through curves were obtained by varying temperatures,while Linear Driving Force approximation (LDF) was used for themass balance, the momentum and energy balance were alsoaccounted for in order to reproduce the break through curves.Investigation of changes in the surface of the activated carbon useddue to CO2 accumulation was carried out with Fourier TransportInfrared Spectroscopy (FTIR) and X-ray Photoelectron Spectroscopy(XPS) analysis. Gas mixture was subjected to different tempera-tures of 301 K, 323 K, 373 K, and 423 K at a total pressure of1.02 bar. The adsorption column was located inside a furnace foreasy control of the process temperature, the column and furnace;which was the adsorption system; were therefore assumedadiabatic as they were isolated using a fiberglass layer and anon-convective refractory material. However, the breakthroughexperiment was treated adiabatically. Siriwardane et al. [38] alsoobserved similar behavior while using 13X zeolite for CO2/N2 gasmixture adsorption. Dantas et al. [127] suggested that resistancesto internal mass transfer are negligible in the adsorption system.It was suggested that for turbulent system, mass spread is due toaxial dispersion [129].

4.6. Concluding remarks

Some concluding remarks can be summarized in the following.It is indicated that the MOFs types had higher pore volume andsurface area than zeolite types. The most issue faces the use ofMOFs as the adsorbents in separation processes is the decomposi-tion under exposure to humid air. The best MOF type (Mg-MOF-74)also has some CO2 adsorption deficiency with existing of moisture,unlike HKUST-1 type. The experimental work demonstrated thatthe best types could be used for CO2 adsorption were Mg/DOBDC(above 250 mg/g) followed by Ni/DONDC (180 mg/g) and CO/DOBDC (140 mg/g). On the other hand, the worst types were ZIF-8, IRMOF-3, IRMOF-1, UMCM-1 and MOF-177 (all of them less than10 mg/g). The effect of water vapor was also studied through 50consecutive cycles; it showed that the CO2 capacity of adsorbentis reduced 25% without any deterioration of MOF compared todry cases. The CO2 regeneration showed obviously that the TSA isbetter than VSA. Comparing to other MOFs in the literature interms of adsorption capacity performance, the CPM-5 showed abetter CO2 adsorption performance than some kinds of MOFs asMOF-5 and MOF-177 and in the same adsorption capacityperformance of MIL-53(Al), UMCM-150 and Ni-STA-12. ExposingCu–HF, Zn–NDC and Ni–Nic to liquid water and NO2 decreasedthe CO2 adsorption capacity of Zn–NDC by about 30% due to partialdecomposition of organic structure, whereas, Cu–HF and Ni–Nicdid not suffer from decompositions. Oppositely, Cu–HF andNi–Nic showed decreases in CO2 adsorption capacity under expos-ing to humidity and SO2 (2 days) while Zn–NDC expressed someincreasing in adsorption in the same exposed gases. Generally,for open pores adsorbent, increasing the gas feed temperaturedecreased the amount of adsorbed CO2 while increasing the feedpressure improved the captured CO2. The isotherms showed the

maximum CO2 adsorption was about 2.4 at 120 kPa and 303 Kand the CO2/N2 selectivity was 7 at the same conditions. The bestpercentages of CO2 recovery and purity above 90% were obtainedby combined processes such as pressure temperature swingadsorption (PTSA) and vacuum temperature swing adsorption(VTSA). Also, the CO2 recovery and purity reasonable percentagescan be obtained from vacuum pressure swing adsorption process.In the breakthrough experiment, it was found that the feed flowrate had negligible impact on the mass adsorbed and heattransferred under the considered span of conditions. The resultsshowed that the selectivity and productivity of the PSA (PressureSwing Adsorption) process was increased by the introduction ofUSO-2-Ni MOF, compared to activated carbon. UiO-67/MCM-41hybrid showed faintly lower selectivity, but higher specific adsor-bent productivity compared to activated carbon. The resultsshowed that the adsorption selectivity was high (>100) in sometypes of MOFs and Zeolites. However, the adsorption capacity val-ues highlighted the type MgMOF-74 as the best (3.3 mmol/g for300 K and 100 kPa). In additions, MgMOF-74 had the longestbreakthrough time which got advantages to increase amount ofCO2 adsorbed.

5. Numerical investigations and mathematical models for fixedbed column adsorption

5.1. Introduction

In order to achieve a suitable and effective design of adsorptionprocess, there is need for an appropriate model to describe thedynamics of the adsorption system [128,38,127]. Most of sug-gested models are mathematical models and more recently, Artifi-cial Neural Network models (ANN) [160] amongst others. Thecomputer simulation tool requires experimental validation forthe development of new system. Since experimental setups arequite costly and time consuming, a mixed design approach usinga well validate simulation tool with reasonable experimental vali-dation seems to give the best design results. The simulation tool iscomposed of a descriptive mathematical model to predict theadsorption system (fixed bed/column) behavior [161]. Suchmathematical models are experimentally verified and make useof independent parameters to estimate the required dynamicproperties of the adsorption system with no extra time and costas compared to the experimental procedures. The models alsoenable break through curve estimation, temperature profile of con-stituent gases at different time and point within the adsorptioncolumn. Varieties of materials and their properties could be quicklyand easily tested using the mathematical models. In addition,variations in compositions and temperatures within the adsorbentcolumn, with respect to time and space, and their effect on theoverall performance of the adsorbent system; can be modeledand simulated [162].

Mathematical models capable of predicting the dynamics ofadsorption systems are made of coupled partial differential equa-tions representing the flow field, mass and energy transfer withinthe field (mass, species, momentum and energy balances) [128].The flow field is usually modeled as a fixed bed (with suitableboundary condition) in which adsorption takes place. A simultane-ous solution is required for the system of PDE’s, making the solu-tion to the system involved and complex, hence the need for asimplified model with good assumptions for easier computationand optimization. The study of modeling and optimization of CO2

adsorption on fixed bed has grown over the years and is still ofimportant interest in the field of Carbon Capture and Sequestration(CCS). The dynamic behavior of an adsorption chamber system canbe categorized based on the nature of the relationship between the


constituent gas species and the solid at equilibrium and the com-plexity of the mathematical needed for describing the adsorptionmass transfer process [163]. The complexity of the mathematicalmodel for describing adsorption process behavior depends on thelevel of concentration, the choice of rate equation and the choiceof flow model [163].

The fixed bed mathematical models are used to temporarilyforecast the performance of an adsorption system in terms ofdynamic property variation of the gas and the adsorption bed dur-ing adsorption e.g. flow rate, temperature, concentration, etc. Thedescription of the pattern of flow within the adsorption columnis usually done using the plug flow model or axially dispersed plugflow model. Some assumptions are usually made but, they differfrom one model to another. E.g. some models account for theeffects of heat generation and heat transfer in the adsorbent bed,based reasons that it may affect the adsorption rates etc. Some ofthese assumptions include (a) Ideal gas behavior, (b) Negligibleradial gradient of concentration (and temperature and pressurewhere applicable), (c) Negligible heat transfer between gas andsolid phase for non-isothermal operation i.e. instantaneous ther-mal equilibrium and (d) Negligible pressure drop across bed. Theassumption of negligible radial gradient has been made by a num-ber of researchers [164,165]. A lot of existing models are based onthe effects of finite mass transfer rate with mathematical modelsclosely representing real process. Most of the popular existingmodels use a linear driving force approximation for the descriptionof mass transfer mechanism in CO2 adsorption process. After sev-eral years of research it has been discovered that it is equallyimportant to consider the effect of momentum balance and heatgeneration and heat transfer in the adsorbent bed. This isimportant because the concentration profile has a dependence ontemperature variations, may be eminent for high-concentrationfeeds, because the heat of adsorption in high concentration feedgenerates thermal waves which travel in axial and radial directions[166].

Adsorption equilibrium has been mostly represented with non-linear isotherms such as the Langmuir isotherm/hybrid Langmuir–Freundlich isotherm. Linear isotherms have been used but only fewcases. The Langmuir model works on the assumption of ideal local-ized molecular interaction between adsorbate and adsorbent withno further interaction on other groups of identical sites. Adsorptionsystem hardly adhere strictly to Langmuir model assumptions,most times, their equilibrium isotherms deviate from the Langmuirmodel form. This may be due to the variation in heat of adsorptionwhich is required to be constant based on Langmuir. From this, itcan be stated that: Since the heat of adsorption changes with con-centration, at lower concentration, the Langmuir model can give anappropriate representation of the system, however, as the concen-tration of the gas to be tested increases, the accuracy of the modelwould drop [163]. Due to the limitations of the Langmuir model,several authors e.g. Freudlich have modified the model e.g. byintroduction of power law expression (Langmuir–Freundlich equa-tions), and a host of other authors. The gas phase material balanceincludes an axial dispersion term, convective term, fluid phaseaccumulation, and the source term due to adsorption of the gasmolecules (adsorbate) on the solid surface (adsorbent). The equa-tion accounts for: The variation in adsorbate velocity and concen-tration in fluid phase with distance along the bed, the averageconcentration of adsorbate components in the solid adsorbent par-ticles, while the axial dispersion coefficient represents the effect ofaxial mixing and the contributing mechanisms. This equation isused to find the transportation of gas composition along the bed,with an assumption of negligible radial variation in gas concentra-tion and solid loading [127,128]. Danchkwert’s boundary condi-tions are applied here [162,167].

5.2. Some existing mathematical models

Mathematical modeling of CO2 adsorption and separationdepends mainly on the mixture from which CO2 is to be separated.It also depends on the type of adsorption process and the adsor-bent media. The following are examples of CO2 separation from dif-ferent mixtures such as CO2/CH4, CO2/N2, CO2/H2, CO2/He, CO2/Air,CO2/CO and flue gas mixtures as well as pressure swing or vacuumswing adsorption.

5.2.1. CO2 in a binary mixture (with CH4, N2, H2 or He)Kumar [168] obtained a mathematical model to describe

adsorption separation of CO2 from binary gas mixtures of Carbondioxide (CO2) and Nitrogen (N2), Carbon dioxide (CO2) andmethane (CH4), and carbon dioxide and Hydrogen. The modelwas made up of a system of coupled partial differential equations.The adsorption media (adsorbents) used was 5A zeolite and BPLcarbon. The flow pattern was described using plug flow model,while the mass transfer pattern was described using local equilib-rium model. The mathematical model was solved numericallyusing finite difference method after which adiabatic simulationwas carried out. The following assumptions were made: Negligibleradial variation in temperature, concentration, negligible pressuredrop within bed, thermal equilibrium between the gas and solidparticles, and non-isothermal heat effects. A Langmuir–Freundlichequilibrium isotherm was assumed. It was concluded that isother-mal assumption was improper for the process design, but it couldbe useful for semi-quantitative forecast of adsorption columnbehavior.

Delgado et al. [169,170] described a mathematical model todescribe the adsorption separation of CO2 from binary gas mixtures(CO2–N2, CO2–He and CO2–CH4) on sepiolite, silicate pellets and aresin. The flow pattern was described using axial dispersed plugflow model, while the mass transfer pattern was described usingthe LDF approximation model. The mass transfer coefficient wasdetermined by fitting the experimental data (i.e. lumped). Ergun’sequation was employed to describe the momentum balance of thesystem. The PDE’s in the mathematical model were solved numer-ically using method of orthogonal collection on finite elementusing PDECOL software. The following assumptions were made:Negligible radial variation in temperature and concentration, neg-ligible pressure drop within bed, thermal equilibrium between thegas and the solid particles, and non-isothermal heat effects. AnExtended Langmuir equilibrium isotherm was assumed. The math-ematical model gave a good description of the breakthrough exper-iment with lower CO2 concentration. However, for the experimentswith high concentration of CO2 were predicted with higher per-centage of error. It was suggested that, introduction of interactionfactor into the model boosted the accuracy of the model based onthe interaction between adsorbed molecules of CO2. Shafeeyanet al. [162] reviewed different existing mathematical modelingmethods of the fixed-bed adsorption of carbon dioxide. Shendal-man and Mitchell [171] obtained a linear mathematical modelusing characteristic method while working on a mathematicalmodel to describe Pressure Swing Adsorption separation of CO2

from a binary gas mixture of Carbon dioxide and Helium (CO2–He). Their adsorption medium (adsorbent) was Silica gel. The flowpattern was described using plug flow model, while the masstransfer pattern was described using local equilibrium model.The mathematical model was solved analytically, by assuming:Negligible radial variation in concentration, negligible pressuredrop, trace system and isothermal heat effects. A linear equilibriumisotherm was assumed. Their model had a limitation of neglectingthe mass transfer resistance effect which made their results differfrom experimental results. Cen and Yang [172] obtained a mathe-


matical model to describe Pressure Swing Adsorption separation ofCO2 and other gaseous products of coal gasification. Their adsorp-tion medium (adsorbent) was activated carbon. The flow patternwas described using plug flow model while the mass transfer pat-tern was described using local equilibrium model and Linear Driv-ing Force approximation model (LDF). An empirical relation wasused for the determination of the mass transfer coefficient forCO2. The mathematical model was solved using the implicit finitedifference method, by assuming: Negligible radial variation in tem-perature and concentration, thermal equilibrium between gas andsolid phase, and non-isothermal heat effect. A Langmuir–Freundlich isotherm was assumed. Their model differs from exper-imental data. This was more pronounced in the CO2 concentration.However, the LDF was closer to the experimental data. A mathe-matical model was developed by Raghavan et al. [173] to describePressure Swing Adsorption separation of CO2 from a binary gasmixture of Carbon dioxide and Helium (CO2–He). Their adsorptionmedium (adsorbent) was Silica gel. The flow pattern was describedusing axial dispersed plug flowmodel, while the mass transfer pat-tern was described using Linear Driving Force approximationmodel. The mathematical model was solved by orthogonal collec-tion and by using finite difference method and by assuming: Neg-ligible radial variation in concentration, negligible pressure drop,traces system inverse dependence of the mass transfer coefficientwith pressure, and isothermal heat effects. A linear equilibriumisotherm was assumed. Their model succeeded in making a goodrepresentation of experimental results.

A mathematical model that describes Pressure Swing Adsorp-tion separation of CO2 from a gas mixture of Carbon dioxide(CO2) and methane (CH4) was developed by Kapoor [174]. BothCO2 and CH4 have equal proportion by volume. The adsorptionmedium (adsorbent) was carbon molecular sieve. The flow patternwas described using plug flow model, while the mass transfer pat-tern was described using LDF approximation model, with a coeffi-cient of mass transfer that is cycle time dependent. Themathematical model was solved using implicit backward finite dif-ference method and, by assuming: Negligible radial variation inconcentration, negligible pressure drop within bed, and isothermalheat effects. A Langmuir equilibrium isotherm was assumed. Theresults provided by the model is reportedly said to be very closeto the experimental data used within about 3% margin of error[162]. Cavenati et al. [175] worked on a mathematical model todescribe the adsorption separation of a gas mixture of carbon diox-ide (CO2) and methane (CH4) on Tekada carbon molecular sieve byVacuum Swing and Pressure Swing (VSA–PSA). The flow patternwas described using axial dispersed plug flow model, while themass transfer pattern was described using double LDF approxima-tion model. Pressure variation in the system was described usingErgun equation. The PDE’s in the mathematical model were solvednumerically using method of orthogonal collection for twenty-five(25) finite elements, with two collection point per element, afterwhich the evolving ODE’s were solved using gPROMS. The follow-ing assumptions were made: Negligible transfer of mass, momen-tum and heat in radial direction, adiabatic and non-isothermal heateffects. A multisite Langmuir equilibrium isotherm was assumed.The mathematical model gave a qualitative description of thebreakthrough experiment and temperature curves. The modelhad a limitation of how to determine new values of mass transfercoefficient for new runs.

Similarly, Ahn and Brandani [176] predicted the dynamics ofCO2 breakthrough on carbon monolith, with different set ofassumptions. The flow pattern was also described using axial dis-persed plug flow model, while the mass transfer pattern wasdescribed using the LDF approximation model. The PDAE’s in themathematical model were solved numerically using gPROMS.The following assumptions were made: Negligible change in

concentration in radial direction. A Langmuir equilibrium isothermwas assumed. The mathematical model which accounted for adetailed structure of the adsorbent gave a qualitative descriptionof the breakthrough experiment. It gave results of very close matchto the experimental data used. However, another model based onthe equivalent channel approach produced wrong results that fore-cast higher separation efficiency for the system. Hwang et al. [177]described a mathematical model to describe the adsorption sepa-ration of CO2 on activated carbon using helium as the carrier gas.The flow pattern was described using plug flow model, while themass transfer pattern was described using LDF approximationmodel. The mass transfer coefficient was lumped i.e. it was deter-mined by fitting the experimental data. The PDE’s in the mathe-matical model were solved numerically using method of lines,after which the evolving ODE’s were solved using DIVPAG. Theremaining algebraic equations were solved using DNEQNF. The fol-lowing assumptions were made: Negligible radial velocity, negligi-ble radial variation in temperature and concentration, negligiblepressure drop within bed, non-adiabatic, and isothermal heateffects. A Langmuir equilibrium isotherm was assumed. The math-ematical model gave a qualitative description of the breakthroughexperiment and temperature curves. The model had a limitation ofhow to determine new values of mass transfer coefficient for newruns.

The mathematical modeling of the adsorption separation of CO2

from flue gas (20% CO2, 80% N2) on zeolite 13X by Vacuum Swing(VSA) was provided by Chou and Chen [178]. The mixture presentstypical dry conditions of flue gas on industrial applications. Theflow pattern was described using axial dispersed plug flow model,while the mass transfer pattern was described using local equilib-rium model. The PDE’s in the mathematical model were solvednumerically using method of lines with adaptive grid points, afterwhich an estimate of the flow rate was done using the cubic splineapproximation. The evolving ODE’s were solved by integrationwith respect to time of flow in adsorption bed using LSODE fromODEPACK software. The remaining algebraic equations were solvedusing DNEQNF. The following assumptions were made: Negligibleradial variation in temperature and concentration, negligible pres-sure drop within bed, thermal equilibrium between the gas and thesolid particles, and non-isothermal heat effects. An Extended Lang-muir equilibrium isotherm was assumed. The mathematical modelgave results similar to the experimental data used but with lowervalues than those of the experiment. This discrepancy was sug-gested to be due to the use of non-specific isotherm.

Recently, Dantas et al. [127,128] worked on a mathematicalmodel to describe the adsorption separation of binary gas mixturesof carbon dioxide and Hydrogen (CO2–H2), carbon dioxide andHelium (CO2–He) on activated carbon and zeolite 13X. The flowpattern was described using axial dispersed plug flowmodel, whilethe mass transfer pattern was described using LDF approximationmodel. The mass transfer coefficient was determined by fitting theexperimental data (i.e. lumped). The momentum balance in thesystem was described using Ergun equation. The PDE’s in themathematical model were solved numerically using method oforthogonal collection for six (6) finite elements, with three (3) col-lection point per element, after which the evolving ODE’s weresolved using gPROMS. The following assumptions were made: Neg-ligible change in temperature and concentration in radial direction,adiabatic and non-isothermal heat effects. Adiabatic and non-adiabatic systems were considered. Toth equilibrium isothermwas assumed. The mathematical model gave a qualitative descrip-tion of the breakthrough experiment for different feed concentra-tion and temperatures. The Toth model gave adequate results forsingle components but deviations were noticed for multicompo-nent gas mixture. Mulgundmath et al. [179] worked on a mathe-matical model to describe the adsorption separation of binary


gas mixtures of carbon dioxide and Nitrogen (90% N2–10%CO2),carbon dioxide and Helium (CO2–He) on zeolite 13X. The flow pat-tern was described using axial dispersed plug flow model, whilethe mass transfer pattern was described using LDF approximationmodel. The PDE’s in the mathematical model were solved numer-ically using method of orthogonal collection for six (6) finite ele-ments, with three (3) collection point per element, after whichthe evolving ODE’s were solved using gPROMS. The followingassumptions were made: Negligible change in temperature andconcentration in radial direction, negligible pressure drop andnon-isothermal heat effects. Adiabatic and non-adiabatic systemswere considered. Langmuir equilibrium isotherm was assumed.The mathematical model gave a qualitative description of thebreakthrough experiment for with good accuracy at the tempera-ture break through point. However, the model gave results of loweraccuracy for the energy balance in the system.

5.2.2. CO2 mixture (with CH4 and H2)Doong and Yang [180] described a mathematical model to

describe Pressure Swing adsorption separation of CO2 from a gasmixture of Carbon dioxide (CO2), methane (CH4) and Hydrogen(H2); all of equal proportion by volume. Their adsorption medium(adsorbent) was activated carbon. The flow pattern was describedusing plug flow model, while the mass transfer pattern wasdescribed using local equilibrium model and pore diffusion model.The mathematical model was solved numerically using finite dif-ference method and, by assuming: Negligible radial variation inconcentration, negligible pressure drop within bed, and non-isothermal heat effects. A Langmuir–Freundlich equilibrium iso-therm was assumed. It was concluded that Knudsen and surfacetension model produced results close to the experimental dataused, while the ILE model produce results with lower CO2 concen-tration with longer break through. They suggested that the latterresult may be due to the assumption of infinite rate of porediffusion.

5.2.3. CO2 (with Air)Diagne et al. [181] worked on a mathematical model to describe

Pressure Swing Adsorption separation of CO2 from air using molec-ular sieves zeolite (13X, 5X, and 4A). The flow pattern wasdescribed using plug flow model, while the mass transfer patternwas described using LDF approximation model. The set of equa-tions in the mathematical model was solved by Euler’s method.The following assumptions were made: Negligible radial variationin concentration, negligible pressure drop within bed, trace sys-tem, and isothermal heat effects. A Langmuir equilibrium isothermwas assumed. The mathematical model gave a qualitative descrip-tion of the breakthrough experiment and temperature curves. Themodel showed good agreement with experimental data except forpoints at which ratio of feed/lean flow rate was less than 2.

5.2.4. CO2 mixture (CO2, CO, H2, and CH4)

Lee et al. [182] obtained a mathematical model to predict thePressure Swing Adsorption separation of coke oven gas mixture(i.e. CO2, CO, N2, and CH4) on two different adsorbents (Zeolite5A and activated carbon). The adsorption bed was made in layers.The flow pattern was described using axial disperse plug flowmodel, while the mass transfer pattern was described using LDFapproximation model. The mass transfer coefficient was lumped.The PDE’s in the mathematical model were solved numericallyusing second order finite difference method (for second orderspace derivatives) and second order backward difference method(for first order space derivatives). The following assumptions weremade: Negligible radial variation in temperature and concentra-tion, thermal equilibrium between gas and solid phase. Effect ofpressure drop along bed was taken into account using the Ergun

equation. A Langmuir–Freundlich isotherm equilibrium isothermwas assumed. The LDF model successfully predicted the adsorptionand desorption steps and gave good simulation results that agreedwith experimental data. It has been reported that the experimentaldata gave higher gas recovery with error range of 4% [162].

The pressure swing adsorption separation of cracked gas mix-ture (i.e. CO2, CO, H2, and CH4) on two different adsorbents (Zeolite5A and activated carbon) was predicted by Park et al. [183]. In theirmathematical model, the adsorption bed was made in layers. Theflow pattern was described using axial disperse plug flow model,while the mass transfer pattern was described using LDF approxi-mation model. The mass transfer coefficient was lumped. ThePDE’s in the mathematical model were solved numerically usingbackward difference method, after which the evolving ODE’s weresolved using GEAR method. The following assumptions weremade: Negligible radial variation in temperature and concentra-tion, thermal equilibrium between gas and solid phase, negligiblepressure drop in axial direction within bed and non-isothermalheat effects. A Langmuir equilibrium isotherm was assumed. Theresults predicted by the LDF model for a single component systemwas close to experimental results of adsorption and desorptioncurves. The model gave a good prediction of the experimental data;however, the model had a limitation of lower residual gas temper-ature than the one gotten from the experiment. This is due to theneglecting of heat loss to the column end.

5.2.5. CO2 mixture (with N2 and O2)Choi et al. [154] worked on a mathematical model to describe

Pressure Swing Adsorption separation of CO2 from flue gas (83%N2, 13% CO2 and 4% O2) using zeolite 13X. The flow pattern wasdescribed using plug flow model, while the mass transfer patternwas described using LDF approximation model. The set of equa-tions in the mathematical model was solved by Euler’s method.The following assumptions were made: Negligible radial variationin temperature and concentration, negligible pressure drop withinbed, and non-isothermal heat effects. An extended Langmuir equi-librium isotherm was assumed. The mathematical model wassolved using MATLAB function which was operated on the princi-ple of Sequential Quadratic Programming (SQP). The model gavea close agreement with experimental data, with little differencesin the temperature data. Kaguei and Wakao [184] described amathematical model while working on the theoretical and experi-mental research on CCS. The adsorption system was a columnpacked with activated carbon. The flow pattern was describedusing axial dispersed plug flowmodel, while the mass transfer pat-tern was described using pore diffusion model. The mathematicalmodel was solved analytically using Laplace domain, by assuming:semi-infinite column Negligible radial variation in temperatureand concentration within column, uniform temperature over col-umn cross section, negligible pressure drop in the axial direction,fixed column wall temperature, and non-isothermal heat effects.A linear equilibrium isotherm was assumed. Their model gave agood prediction of thermal waves at different axial locations.

In order to predict the adsorption separation of CO2 and CO onactivated carbon, Hwang and Lee [185] obtained a mathematicalmodel in which, the flow pattern was described using axial dis-perse plug flow model. The mass transfer pattern was describedusing LDF approximation model. The mass transfer coefficientwas made pressure dependent. The PDE’s in the mathematicalmodel was solved numerically using the method of orthogonal col-lection, after which the evolving ODE’s were solved using DGEARthrough a Gear’s stiff method in different orders and step size.The following assumptions were made: Negligible radial variationin concentration, negligible pressure drop within bed and isother-mal heat effects. A Langmuir equilibrium isotherm was assumed.The results predicted by the LDF model for a single component

Table 6Detailed review of adsorption numerical models including mass isotherm type and mass transfer models.

# Authors’ names Application type ModelDimension

Mass transfer model Isotherm Type Energy model Pressure and velocity model Solution type

1 Carter andHusain [186]

Modelling of adsorption of Carbon dioxideand water vapour on molecular sieve

1-D,transient

From experimentaldata

Langmuirisotherm

Isothermal Negligible pressure drop Numerical solution onFortran

2 Kumar [168] Modelling of blow down of adsorption ofCO2 from gaseous mixture of; CO2/H2 CO2/CH4 CO2/N2 on Zeolite 5A and BPL carbon byPSA

1-D,transient

Local equilibriummodel

Langmuirisotherm

Non-Isothermal Negligible pressure gradientacross adsorption bed

Numerical solutionAdiabatic systemNegligible radialtemperature gradient

Flow behaviour: Plug flow Finite difference methodwith the use of IBM 370/165

3 Hwang and Lee[185]

Modelling of adsorption and desorption ofgaseous mixture of CO2 and CO on activatedcarbon by breakthrough experiment

1-D,transient

LDF approximationmodel

Langmuirisotherm

Isothermal Negligible pressure gradientacross adsorption bed

Numerical solution with theuse of DGEAR commercialcodeTemperature of column

wall, adsorbent and gaswere all accounted for

Flow behaviour: Axialdispersed plug flow

4 Chue et al. [40] Modelling of the adsorption of CO2 fromCO2/N2 mixture on Zeolite 13X and activatedcarbon by PSA

1-D,transient

Adsorbedconcentration by IASmodel

Langmuirisotherm

Non-isothermal Negligible pressure drop inbed

Adiabatic Flow behaviour: Axialdispersed plug flowThermal equilibrium

between gas and solidphase

5 Hwang et al.[177]

Modelling of adsorption of gaseous mixtureof CO2 and CO on activated carbon bybreakthrough experiment

1-D,transient


ExtendedLangmuirisotherm

Isothermal Negligible pressure gradientacross adsorption bed

Numerical solution

Lumped mass transfercoefficient

Non-adiabatic and adiabaticsystems

Flow behaviour: Plug flow Linear algebras were solvedusing DIVPAG commercialcode while non-linearalgebra equations weresolved using DNEQNFcommercial code

Temperature of columnwall, adsorbent and gaswere all accounted for

Negligible radial velocity

Negligible radialtemperature gradient

6 Diagne et al.[181]

Modelling of adsorption of CO2 from air byPSA on Zeolite (5A, 13X and 4A)

1-D,transient


Langmuirisotherm

Isothermal Negligible pressure drop Euler’s methodFlow behaviour: Ideal plugflow

7 Ding and Alpay[187]

Modelling of adsorption and desorption ofCO2 on hydrotalcite at high temperature

1-D,transient

LDF model based onpore diffusion

Langmuirisotherm

Non-isothermal. Negligibleradial temperature gradient

Pressure distribution byErgun’s equation

Numerical solution with theuse of gPROMS commercialcodeThermal equilibrium

between fluid and particlesFlow behaviour: Axialdispersed plug flow

8 Takamura et al.[188]

Modelling of CO2 adsorption from gaseousmixture of CO2 and N2 on Zeolites (Na–X andNa–A)

1-D,transient


Langmuirisotherm

Isothermal Negligible pressure drop Discretisation of coupledPDEA equations in spaceand time. Final solution ofODE with variable time step

Flow behaviour: Plug flowFlow behaviour: Plug flow

9 Choi et al. [154] Modelling of CO2 adsorption from flue gasmixture containing 13% CO2, 83% N2 and 4%O2 on zeolite 13X by break throughexperiment and PSA operation

1-D,transient.



Non-isothermal Negligible pressure drop inradial direction

Numerical solution with theuse of MATLAB function

Adiabatic system Flow behaviour: Plug flowNegligible temperaturegradient in radial direction

Gas flow rate in bed ismainly affected by bedheight

10 Chou and Chen[178]

Modelling of CO2 adsorption from flue gasmixture containing 20% CO2 and 80% N2 onzeolite 13X by VSA.

1-D,transient

Local equilibriummodel



Negligible pressure gradient Analytical + numericalsolution

Thermal equilibriumbetween fluid and particles


Solution of spatialderivatives by upwinddifferenceSolution of flow rates bycubic splineSolution of temperature,concentration and adsorbedmass by integration withthe use of LSODE fromODEPACK commercial code

(continued on next page)

R.Ben-M

ansouret

al./Applied

Energy161

(2016)225–

255245

Table 6 (continued)



11 Cavenati et al.[158]

Modelling of fixed bed adsorption of CO2,CH4 and N2 on Zeolite 13X at high pressureby breakthrough experiment

Experimentalmeasurement

Toth Isothermand

Isothermal Experimental measurement Numerical solution to solvefor mass deposited inadsorbent using MATLABcommercial code

MultisiteLangmuirisotherm


Modelling of fixed bed adsorption of CO2

from a gaseous mixture of 45% CO2 and 55%CH4on carbon molecular sieve 3 K by PSA

1-D,transient

A double LDFapproximation model

MultisiteLangmuirisotherm



Numerical solution with theuse of gPROMS commercialcodeFlow behaviour: Axial

dispersed plug flow13 Ahn and

Brandani [176]Modelling of fixed bed adsorption anddesorption of CO2 on Carbon Monoliths bybreak through experiment

1-D,transient


Langmuirisotherm

Isothermal Relationship betweenaverage velocity andaverage pressure drop wasestimated with the use ofequation by Cornish 1928

Numerical solution with theuse of gPROMS commercialcode



Modelling of fixed bed adsorption of CO2

from a gaseous mixture of 20% CO2/60% CH4/and 20% N2 on zeolite 13X by LayeredPressure Swing Adsorption (LPS)

1-D,transient

Bi-LDF model Multicomponentextension ofmultisiteLangmuir

Non-Isothermal Pressure distribution byErgun’s equation

Numerical solution with theuse of gPROMS

Temperature of columnwall, adsorbent and gaswere all accounted for


Negligible radialtemperature gradient

15 Moreira et al.[190]

Modelling of fixed bed adsorption of Heliumdiluted CO2 on hydrotalcite (Al–Mg)

1-Dtransient


Langmuirisotherm

Isothermal Negligible pressure drop Numerical with the use ofPDECOL in FORTRANcommercial codeCalculation of mass

transfer coefficient bytheoretical correlations


16 Delgado et al.[169,170]

Modelling of fixed bed adsorption of CO2from gaseous mixture of; CO2/He CO2/CH4

CO2/N2 on Silicalite pellets, sepiolite, andresin using break through experiment

1-Dtransient





Numerical solution byPDECOL commercial code


Pressure variation in timeand spaceFlow behaviour: Axialdispersed plug flow

17 Dantas et al.[127–129]

Fixed bed adsorption of gaseous mixture of;CO2/N2 and CO2/He on zeolites 13X andactivated carbon

1-D,transient


Toth Isotherm Non-Isothermal Pressure distribution byErgun’s equation

Numerical solution usinggPROMS commercial code

by break through experiment and PSA Lumped mass transfercoefficient

Adiabatic and non-adiabaticsystem

Axial dispersed plug flow

Model accounted for Heattransfer in gas, solid andwall

18 Biswas et al.[191]

Modelling of adsorption separation ofgaseous mixture of CO, CH4, H2, CO2 onZeolite 5A and activated carbon

1-D,transient

LDF model MultisiteLangmuir model

Isothermal Pressure distribution byErgun’s equation

Discretisation by Newtonbased approach


Assuming temperature ofwall, gas phase andadsorbent are equal


Algebraic solution

19 Agarwal [192] Fixed bed adsorption of CO2 from gaseousmixture of CO2/N2, 45% CO2/55% H2 by PSA

1-D,transient


Dual siteLangmuirisotherm

Temperature equilibriumbetween gas phaseadsorbent


Numerical solution with theuse of interior point NPLsolver


Constant column walltemperature


20 Krishna andvan Baten [22]

Modelling of PSA performance and breakthrough characteristics of zeolites (MFI,JBW, AFX, NaX) and MOFs (MgMOF-74,MOF-177, CuBTTri-mmen) for gaseousmixture of CO2/N2

1-D,transient

Isotherm Negligible pressure drop Molecular simulation withthe use of Configuration-Bias Monte Carlo (CBMS)

Assumed flow behaviour:Plug flow

246R.Ben-M

ansouret

al./Applied

Energy161

(2016)225–

255

Table 6 (continued)



21 Casas et al.[193]

Fixed bed adsorption of CO2 from gaseousmixture of CO2/H2 on activated carbon bybreak through experiment

1-D,transient

LDF model Langmuir and Sipisotherms

Thermal equilibriumbetween gas stream andadsorbent


Finite volume method andtime integration on IMSLDIVPAG commercialpackage using Gear’smethod


Column wall temperature isaccounted for separately

Flow behaviour: Plug flow

22 Mulgundmathet al. [179]

Fixed bed adsorption of CO2 from gaseousmixture of 10% CO2/90% N2 on Ceca 13X bybreak through experiment

1-D,transient

LDF approximationmodel for external fluidfilm mass transfer

Langmuirisotherm

Non-Isothermal Negligible pressure dropTemperature of columnwall, adsorbent and gaswere all accounted for


23 Casas et al.[159]

Mathematical modelling of CO2 adsorptionfrom CO2/H2 mixture in MOF and UiO-67/MCM-41 by PSA and break throughexperiment

1-D,transient

Mass transfercoefficient determinedby fitting ofexperimental datameasured in the rangeof interest

Langmuirisotherm

Non-Isothermal Pressure distribution byErgun’s equation

Integration via Gear’smethod with the use ofIMSL DIVPAG (Fortran)commercial code

AdiabaticModel accounted for Heattransfer in gas, solid andwallIsosteric heat of adsorptionand heat capacities of thefluid and the solid phase

24 Sabouni [194] Modelling of adsorption of CO2 From inCPM-5 by breakthrough experiment

1-D,transient

Mass transfercoefficient determinedby fitting ofexperimental data

Langmuir–Freundlichisotherm

Isothermal Negligible pressure dropthrough column

Numerical solution with theuse of COMSOL

Constant gas velocitythrough column

25 Ribeiro et al.[157]

Modelling of CO2 adsorption from flue gasby a mixture of Activated carbonhoneycomb monolith and Zeolite 13Xhybrid system by Electrical SwingAdsorption (ESA)

1-D,transient

Two different LDFmodels; one for micropores and the other formacro pores

MultisiteLangmuir model

Temperature equilibriumbetween the solid phases


Numerical solution with theuse of gPROMS commercialcode

Lumped mass transferparameter for mesopores and micro pores;obtained fromBosanquet equation

Negligible temperaturegradient in adsorbent

Assumed flow behaviour:Axial plug flow

26 Krishnamurthyet al. [195]

Modelling of CO2 adsorption from dry fluegas in Zeochem zeolite 13X by breakthrough experiment and VSA

1-D,transient


Extended dualsite Langmuirmodel

Non-Isothermal Non Isobaric Numerical solution by stiffODE solver; ode23s inMATLAB commercial code

Pressure distribution byDarcy’s equation

R.Ben-M

ansouret

al./Applied

Energy161

(2016)225–

255247


system was close to experimental results of adsorption and des-orption curves. The mass transfer coefficient and the assumptionsgave good results, close the experimental data for adsorption anddesorption for multi-component sorption system. Table 6 providesdetailed review of Adsorption Numerical models including massisotherm type and mass transfer models. The table gives a detailedaccount of the mathematical models used in previous studies. Thetable presents the type of the two most important properties usedin the models; namely the adsorption isotherm model and themass transfer model. Other important consideration such as theheat/energy transfer as well as the pressure drop models are alsoreviewed up to 2014.

Most of the modelling studies indicated that the gases flowthrough the bed are treated as one dimensional flow (1D) andthe effect of radial direction or 3D simulation still need modellingand performance optimizations investigations. Moreover, the idealgas behaviour assumption for gases dominated most of the CO2

separation numerical investigations. Another point is that theavailable data obtained by experimental work as adsorption andthermal properties of adsorbent and adsorbate materials couldonly be used in the modelling to validate the simulation and inves-tigate the adsorption process behaviour and its performance opti-mization. Therefore, the modelling is restricted by what has beenperformed by experimentation.

5.3. Modeling of adsorption of CO2 for carbon capture

In this section we present the mathematical model and somesample simulations results of the adsorption fixed bed (seeFig. 5) for CO2 separation. The fixed bed represents high aspectratio column or cylinder often used industry for separating gasesusing the PSA or TSA process. A similar geometry to that used inexperimental adsorption studies to measure the capacity of testedmaterials is used in the break-through setup. Because of the highaspect ratio of such system, the concentration, temperature, pres-sure and velocity gradients are mainly along the axis of the cylin-drical bed. These axial gradients are much larger than the radialgradients; hence the one-dimensional (1D) assumption is madeas done in all the studies listed in Table 6 above. After validationof the numerical model with experimental data, we use it to sim-ulate different materials including the commonly used ones suchas activated carbon and the novel material such as MOF-5, MOF-74 and MOF-177 which have been recently developed and havevery good potential to become the adsorption materials of thefuture. In addition, we have simulated three different operationmodes of the fixed bed namely (i) the break-through test simula-tion, (ii) the storage simulation and (iii) the PSA simulation. Thethermo-physical properties of the bed materials, the geometricdetails and the operating parameters including temperatures, pres-sures, gas mixture inlet compositions are given for each case.

5.3.1. Fixed bed adsorption modelThe conservation of mass, species, momentum and energy

equations are developed to describe the fixed bed adsorption sys-tem (Fig. 5). Since the bed has a large aspect ratio, the gradient inradial direction are ignored and hence the 1D approximation. Theflow behavior is characterized with axially dispersed plug flowmodel and the mass transfer rate is assumed to follow the LinearDriving Force (LDF) model. The LDF model for mass transfer wasinitially developed by Glueckauf and Coates [196]. They suggestedthat the uptake rate of a species into adsorbent solid particles isproportional to the difference between the concentration of thatspecies at the outer surface of the particle, denoted as q�

j (equilib-rium adsorption amount) and its average concentration within theparticle (volume-averaged adsorption amount) denoted as �qj. Thismodel is expressed in Eq. (2) below. In addition, the equilibrium

maximum adsorption rates are determined from proper isothermswhich describe the adsorbed amount as a function of pressuregiven a certain temperature. In the fixed bed adsorption models,the ideal gas behavior is normally considered with constant fluidproperties and constant bed porosity. It is also assumed that theprocess is adiabatic. The fixed bed adsorption model is describedby the following equations which are derived from, mass, momen-tum and energy equations:

5.3.2. Governing equationsSpecies (mass balance) conservation for CO2 and N2, Dantas

et al. [128,129]

e@Cj@t

þ @uCj@z

� �¼ eDax

@2Cj@z2

!� ð1� eÞqp

@�qj

@tð1Þ

As discussed above, the LDF mass transfer model for each compo-nent is given by:

@�qj

@t¼ KL;j q�

j � �qj

� �ð2Þ

where the maximum adsorption is determined from the followingisotherm equation:

q�j ¼ qm;jKeq;jPj

�1þ Keq;jPj

� �nh i1=nð2aÞ

and the equilibrium adsorption coefficient is given by:

Keq;j ¼ Ko;jeð�DHj=RTg Þ ð2bÞMass Conservation

qg@u@z

¼Xj

ð1� eÞe

qp@�qj

@t

� �ð3Þ

The momentum equation is simplified for the porous media caseunder very slow flow rate to the Darcy model equation:

� @p@z

¼ 150lg 1� eð Þ2

e3d2p

uþ 1:75ð1� eÞe3dp

qgu2 ð4Þ

The energy equation for the gases can be written as:

eqgCv ;g@Tg

@tþ qgCp;g

u@Tg

@z

� �¼ ekL

@2Tg

@z2

!� Csð1� eÞqp

@Ts

@t

þ ð1� eÞqp

Xj

�DHj@�qj

@t

� �

� 4hw

dintðTg � TwÞ ð5Þ

While the energy conservation for the solid part of the bed can bewritten as:

qpCs@Ts

@t¼ qp

Xj

�DHj@�qj

@t

� �þ 6hf

dpðTg � TsÞ ð6Þ

The bed wall temperature is solved from the following equation:

qwCp;w@Tw

@t¼ awhwðTg � TwÞ and ð7Þ

aw ¼ dint=lðdint þ lÞ ð7aÞThe properties of the flue gas within the adsorption operating win-dowwere modelled through the gas mixture concepts and are givenas follows:

Density of mixture

qg ¼PP

iyiMi� �RuT

ð8Þ

Fig. 7. Validation breakthrough curve for CO2 & N2 on activated carbon forTgfeed = 373 K. Experimental data [128].

Fig. 8. Validation breakthrough curve for CO2 & N2 on activated carbon forTgfeed = 423 K. Experimental data [128].


where M is the molecular weight.Ru is the gas constant (8.314 J/mol K).Thermal conductivity can be obtained using Wilke’s approach

kg ¼Xni¼1

yikiPnj¼1yiUij

ð9Þ

and

Uij ¼1þ li

lj

� �0:5MiMj

� �0:25 ffiffiffi8

p1þ Mi

Mj

� �h i0:5 ð10Þ

The boundary conditions for the above equations are written asfollows:

z ¼ 0 : eDax@Cj@z

� ��zþ

¼ �u Cj

��z� � Cj

��zþ

� �

z ¼ L :@Cj@z

� ��z�

¼ 0

z ¼ 0 : ekL@Tg

@z

� ��zþ

¼ �ucCs Tg

��z� � Tg

��zþ

� �

z ¼ L :@Tg

@z

� ��z�

¼ 0

z ¼ 0 : uCj

��z� ¼ uCj

��zþ

The initial conditions

t ¼ 0 : P ¼ Pinlet; Tw ¼ Tg ¼ Ts ¼ Tg;inlet and Cjðz;0Þ ¼ �qjðz;0Þ ¼ 0

The system properties that describe the heat and mass parametersmodelled as follows:

kL ¼ kg � ð10þ 0:5 � Pr � ReÞhw ¼ kg � ð12:5þ 0:048 � ReÞhf ;j ¼ Nu � kg=dp

Dax ¼ Uinlet � LPe

Dimensionless Numbers:

Re ¼ ðqg � Uinlet � dpÞlg

Pr ¼ ðCp;g � lgÞkg

Nu ¼ 2:0þ 1:1 � ðRe0:6 � Pr1=3Þ

Pe ¼ 0:508 � ðRe0:020 � LÞdd

5.4. Overview of results of numerical simulations of adsorptive carboncapture

5.4.1. A comparison of breakthrough simulation results using LinearDriving Force Model (LDF) with breakthrough experimental result

The work of Dantas et al. [128] presents breakthrough experi-ments for a temperature range of 28–150 �C (301–423 K) on acti-vated carbon. The adsorption bed used was 0.171 m � 0.02 mØin size and feed flow rate was 30 mL/min. The data provides vari-ation of CO2 and N2 concentrations with time at the exit section.Figs. 7 and 8 show a comparison of experimental data and LDFmodel simulation for the break through curves for the adsorptionof CO2 from binary gas mixture of 20% CO2wt N2 for 100 �C &150 �C (273 K & 423 K) respectively. These figures show the ratioof species concentrations at bed exit to the feed concentration.The total feed gas flow rate in each case is 30 mL/min. Theroll-up behaviour of N2 remains as explained before i.e. the

concentration of N2 at the outlet becomes greater than the feedconcentration [8,9], which is a common behaviour in multi compo-nent gaseous mixture adsorption. The quick breakthrough of theLDF model compared to the experimental model may be due thesome differences in binary mixture properties used, constant fluidproperties (e.g. density, viscosity etc.) and the ideal gas lawassumption.

As shown from Figs. 7 and 8, the present model captures thechanges in CO2 and N2 concentrations with time quite well fromthe qualitative point of view. The peak in the nitrogen concentra-tion is well captured by the model. The model shows under-prediction of the saturation time. The figures also indicate thatthe model provides better agreement as the temperature becomeshigh. The breakthrough time gets shorter as the temperatureincreases. This is attributed to the fact that nitrogen has higher dif-fusion at higher temperatures, thus, nitrogen adsorption becomesfaster. After implementing the LDF model for different situationsincluding breakthrough experiment or pressure swing operation;and different materials (AC, Zeolite X13, MOF5 and MOF74) wecan conclude that this model gave good agreement between theexperimental measurement and model predictions.

5.4.2. Simulated results of the breakthrough behaviour of Mg-MOF-74The adsorption breakthrough curves CO2 and N2 on Mg-MOF-74

for the separation of CO2 from a binary gas mixture of 15% CO2 wtN2 which is as shown in Fig. 9. The adsorption bed used was0.171 m � 0.02 mØ in size and feed flow rate was 30 mL/min.Fig 9 portrays that Mg-MOF-74 has very high selectivity for CO2

which conforms to existing reports. The plots also show that thebreak through time for CO2 in the described mixture decreasedwith temperature. At a feed gas temperature of 301 K, CO2 adsorp-tion took well above 500 min before breakthrough which conforms

0 500 1000 1500-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Time (min)

C/C

inle

t

Concentration at exit 1500 min

conc of CO2 at exitconc of N2 at exit

0 500 1000 1500-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Time (min)

C/C

inle

t



0 100 200 300 400 500-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Time (min)

C/C

inle

t



0 100 200 300 400 500-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Time (min)

C/C

inle

t



Tgfeed = 301K

Pfeed = 1.2bar

Tgfeed = 323K

Pfeed = 1.2bar

Tgfeed = 373K

Pfeed = 1.2bar

Tgfeed = 423K

Pfeed = 1.2bar

Fig. 9. Breakthrough curves for CO2 & N2 adsorption on Mg-MOF-74 at various feed temperatures.

Pfeed = 50bar

5bar 4bar 3bar

10bar

40bar

30bar

20bar

2bar 1bar

Fig. 10. Profile of amount of CO2 stored on MOF-5 for 50 min for various feedpressures. Pfeed = 50bar

4bar3bar2bar

5bar

10bar

20bar

30bar

40bar

1bar

Fig. 11. Profile of amount of CO2 stored on MOF-177 for 30 min for varying feedpressures.


quite closely to existing reports for similar conditions [24,197].This breakthrough time decreases as the feed gas temperatureincreases, which may be due to reduction in the value of the Lang-muir adsorption equilibrium parameter with temperature, whichturn decreases the retention time and leads to longer break-through. The continuous adsorption of some quantity of CO2 atbed exit after breakthrough might be due to existing suggestionthat the single component single site Langmuir model inade-quately predicts CO2 adsorption in Mg-MOF-74 even at loadingbelow 8 mol/kg [24]. The roll-up exhibited by Nitrogen in all fourcases conforms to existing reports for multicomponent adsorption[8,9]. This phenomenon is due to the displacement effect of CO2 onNitrogen which happens during initial continuous adsorption of

CO2 by the material which leads to a steep rise in the concentrationof Nitrogen at bed exit.

5.4.3. Simulated results for adsorptive storage of CO2 on MOF-5 &MOF-177

Simulation of the adsorbed mass of CO2 on MOF-5 & MOF-177show increase in the amount of CO2 adsorbed on bed with theincrease in gas feed pressure which is as shown in Figs. 10 and11 respectively for pressure values from 1 to 50 bar, respectively.The adsorption bed used was 0.171 m � 0.02 mØ in size and feedflow rate was 30 mL/min. As the pressure increases, CO2 moleculesare pressed against the surface of the solid. This increases the avail-


able surface area for CO2 adsorption within existing adsorptionsites. The increase in the available surface area for adsorption inturn leads to an increase in the maximum possible adsorption ofCO2, hence, the amount of adsorbed CO2 increases. The highestadsorbed amount for MOF-5 after 50 min of adsorption is achievedwith 50 bar feed pressure which is about 7.4 g. This matched withthe work of Zhao et al. [111]. The highest adsorbed amount forMOF-177 after 30 min of adsorption is achieved with 50 bar feedpressure which is about 11.3 g. This matched with the work of Sahaand Bao [198].

5.4.4. Simulated results of PSA of CO2 on Mg-MOF-74The simulated pressure curve and total amount of CO2 adsorbed

in bed exit for PSA with 10s of counter-current pressurization, 50sof feed, 25s of counter-current depressurization and 50s ofcounter-current purge are as shown in Figs. 12 and 13 respectively.The adsorption bed used was 0.2 m � 0.02 mØ in size, feed andpurge pressures are 1.3 bar & 0.5 bar respectively while feed andpurge flow rates are 5e�5 m3/s & 8.3e�6 m3/s respectively. Forthese cycles, the feed gas temperature and concentration for eachcycle were maintained at 373 K and 15% CO2 wt 85% N2 respec-tively. Twenty-five (25) PSA cycles were simulated. From the curve,it can be seen that increase in the duration of purge has effect onthe adsorption capacity of the bed. For the same bed size and feedtemperature and similar operating condition, increase in the dura-tion of purge from 25s to 50s helped to slow down the rate of beddeterioration during the first nine cycles with about 0.61 g, 0.51 g,0.42 g and 0.36 g of CO2 adsorbed in the 2nd, 3rd and 4th cyclesrespectively as compared to 0.58 g, 0.45 g, 0.36 g and 0.32 g

Fig. 12. Pressure curve at bed exit for 25 cycles of four-step PSA of CO2 on Mg-MOF-74 from gas mixture of 15%CO2, 85%N2 at 373 K for PSA run with tpurge = 50s.

Fig. 13. Total amount of CO2 adsorbed in bed for 25 cycles of four-step PSA of CO2

on Mg-MOF-74 from gas mixture of 15%CO2, 85%N2 at 373 K for PSA run withtpurge = 50s.

adsorbed in similar cycles of PSA run when the purge time wasset to 25s. About 0.26 g of CO2 adsorbed during the ninth cycle,after which the rate of deterioration of the material reduced.Between the ninth and twenty fifth cycles, the rate of deteriorationof the material slowed down as the material achieved cyclic steadystate. The amount CO2 adsorbed in the fifteenth cycle is about0.24 g.

6. Conclusions

A review on the separation of carbon dioxide from typicalpower plant exhaust gases using the adsorption process is pre-sented. This method is believed to be one of the most economicand least interfering ways for post-combustion carbon capture asit can accomplish the objective with small energy penalty and veryfew modifications to existing power plants. The review focused onthe candidate materials that can be used to adsorb carbon dioxide,the experimental investigations that have been carried out to studythe process of separation using adsorption and the numerical mod-els developed to simulate this separation process and serve as atool to optimize systems to be built for the purpose of CO2 adsorp-tion. The review pointed out some of the remaining challenges forpost combustion carbon-capture. In particular, to handle typicalCO2 mixtures in exhaust gases (78 N2, 13% CO2, 9% H2O), newmaterials of high selectivity and high adsorption of carbon dioxidehigh stability with water vapor, good thermal stability, good ther-mal conductivity, high specific heat, good corrosion resistance aswell as sufficient mechanical strength to endure repeated cyclingare required. It is indicated that there is a need for evolution of acontemporary class of more effective, comparatively cheap, andindustrially applicable materials for carbon capture and storageapplications in order to minimize the uncontrolled emissions ofgreenhouse gases into the atmosphere, which is necessary on anational and international scale. In terms of experimental investi-gations, the present work done on physical adsorption experimentsrelied on small amount of adsorbents (in few grams). Adsorptionbeds with larger mass should be studied to reflect the capabilityfor utilizing such systems in the actual applications. In addition,the number of adsorption/desorption cycles that the adsorbentcan handle without deterioration lacks a long time operationsand recordings. As well, more configurations of the bed (other thanthe tubular beds) are also important in this field. The knowledgegap related to the modelling is that the present simulation consid-ered only one dimensional flow and ignored the radial or 3D ther-mal and adsorption behaviours. The mass transfer rate in themajority of simulations is currently represented by a linear drivingforce (LDF) model. More physically realistic approaches should beimplemented and comparison between them is needed to achievesignificant accuracy and more importantly good agreement withexperimental results. Finally we suggest that more investigationsare carried out on the thermodynamic analyses, using the firstand the second law efficiencies; of physical adsorption of CO2

capture.

Acknowledgments

The authors wish to acknowledge the support received fromKing Abdulaziz City for Science and Technology (KACST) CarbonCapture and Sequestration Technology Innovation Center (CCS-TIC #32-753) at King Fahd University of Petroleum and Minerals(KFUPM) for funding this work through Project No. CCS10. The sup-port of KFUPM through the Research Institute and the Deanship ofScientific Research is greatly appreciated.


References

[1] IPCC. Special report on carbon dioxide capture and storage. In: Metz B,Davidson O, Coninct H, Loos M, Meyer L, editor. Cambridge (University Press);2005. <http://ipcc.ch/report/srccs/>.

[2] D’Alessandro DM, McDonald T. Toward carbon dioxide capture usingnanoporous materials. Pure Appl Chem 2011;83(1):57–66.

[3] GCEP Energy Assessment Analysis Project. Global climate & energy project anassessment of carbon capture technology and research opportunities. In:2005, Global Climate & Energy Project, Stanford University.

[4] Herzog H, Meldon J, Hatton A. Advanced post-combustion CO2 capture. CleanAir Task Force; 2009. April.

[5] Li B et al. Enhanced binding affinity, remarkable selectivity, and high capacityof CO2 by dual functionalization of a rht-type metal–organic framework.Angew Chem Int Ed 2012;51:1412–5.

[6] Li H, Ditaranto M, Yan J. Carbon capture with low energypenalty: supplementary fired natural gas combined cycles. Appl Energy2012;97:164–9. http://dx.doi.org/10.1016/j.apenergy.2011.12.034.

[7] Li JR, Sculley J, Zhou HC. Metal-organic frameworks for separations. Chem Rev2012;112(2):869–932.

[8] Li G, Xiao P, Xu D, Webley PA. Dual mode roll-up effect in multicomponentnon-isothermal adsorption processes with multilayered bed packing. ChemEng Sci 2011;66(9):1825–34.

[9] Li JR, Ma Y, McCarthy MC, Sculley J, Yu J, Jeong HK, et al. Carbon dioxidecapture-related gas adsorption and separation in metal-organic frameworks.Coord Chem Rev 2011;255(15–16):1791–823.

[10] Dong R, Lu H, Yu Y, Zhang Z. A feasible process for simultaneous removal ofCO2, SO2 and NOx in the cement industry by NH3 scrubbing. Appl Energy2012;97(x):185–91. http://dx.doi.org/10.1016/j.apenergy.2011.12.039.

[11] Lv Y, Yu X, Jia J, Tu S-T, Yan J, Dahlquist E. Fabrication and characterization ofsuperhydrophobic polypropylene hollow fiber membranes for carbon dioxideabsorption. Appl Energy 2012;90(1):167–74. http://dx.doi.org/10.1016/j.apenergy.2010.12.038.

[12] Lv Y, Yu X, Tu S-T, Yan J, Dahlquist E. Experimental studies on simultaneousremoval of CO2 and SO2 in a polypropylene hollow fiber membranecontactor. Appl Energy 2012;97:283–8. http://dx.doi.org/10.1016/j.apenergy.2012.01.034.

[13] Song CF, Kitamura Y, Li SH. Evaluation of stirling cooler system for cryogenicCO2 capture. Appl Energy 2012;98:491–501. http://dx.doi.org/10.1016/j.apenergy.2012.04.013.

[14] Van Benthum RJ, van Kemenade HP, Brouwers JJH, Golombok M. Condensedrotational separation of CO2. Appl Energy 2012;93:457–65. http://dx.doi.org/10.1016/j.apenergy.2011.12.025.

[15] Zhang M, Guo Y. Rate based modeling of absorption and regeneration for CO2capture by aqueous ammonia solution. Appl Energy 2013;111(x):142–52.http://dx.doi.org/10.1016/j.apenergy.2013.04.074.

[16] Yan S, Fang M, Wang Z, Luo Z. Regeneration performance of CO2-rich solventsby using membrane vacuum regeneration technology: Relationshipsbetween absorbent structure and regeneration efficiency. Appl Energy2012;98:357–67. http://dx.doi.org/10.1016/j.apenergy.2012.03.055.

[17] Nakamura T. Recovery and sequestration of CO2 from stationary combustionsystems by photosynthesis of microalgae quarterly technical progress report# 9 reporting period start date: 1 October 2002 Reporting Period End Date :31 December 2002 Prepared by National Energy; 2003.

[18] Songolzadeh M, Ravanchi MT, Soleimani M. Carbon dioxide capture andstorage: a general review on adsorbents. World Acad Sci Eng Technol2012:225–32.

[19] Ho MT, Allinson GW, Wiley DE. Reducing the cost of CO2 capture from fluegases using pressure swing adsorption. Ind Eng Chem Res 2008;47(14):4883–90.

[20] Yoo J, Cho SH, Yang T. Comparison of activated carbon and zeolite 13X for COarecovery from flue gas by pressure swing adsorption. Ind Eng Chem Res1995;34:591–8.

[21] Krishna R. Adsorptive separation of CO2/CH4/CO gas mixtures at highpressures. Microporous Mesoporous Mater 2012;156(July):217–23.

[22] Krishna R, van Baten JM. A comparison of the CO2 capture characteristics ofzeolites and metal–organic frameworks. Sep Purif Technol 2012;87:120–6.

[23] Clausse M, Bonjour J, Meunier F. Adsorption of gas mixtures in TSA adsorbersunder various heat removal conditions. Chem Eng Sci 2004;59(17):3657–70.

[24] Mason J, Sumida K, Herm ZR, Krishna R, Long JR. Evaluating metal–organicframeworks for post-combustion carbon dioxide capture via temperatureswing adsorption. Energy Environ Sci 2011;4(8):3030.

[25] Li JR, Kuppler RJ, Zhou HC. Selective gas adsorption and separation in metal–organic frameworks. Chem Soc Rev 2009;38:1477.

[26] Choi S, Drese JH, Jones CW. Adsorbent materials for carbon dioxide capturefrom large anthropogenic point sources. Chem Sus Chem 2009;2:796.

[27] Zhao D, Yuan D, Yakovenko A, Zhou H-C. A NbO-type metal-organicframework derived from a polyyne-coupled di-isophthalate linker formedin situ. Chem Commun (Camb) 2010;46(23):4196–8.

[28] Zhao D, Yuan D, Krishna R, van Baten JM, Zhou H-C. Thermosensitive gatingeffect and selective gas adsorption in a porous coordination nanocage. ChemCommun (Camb) 2010;46(39):7352–4.

[29] Akhtar F, Andersson L, Keshavarzi N, Bergström L. Colloidal processing andCO2 capture performance of sacrificially templated zeolite monoliths. ApplEnergy 2012;97:289–96. http://dx.doi.org/10.1016/j.apenergy.2011.12.064.

[30] Wang J, Manovic V, Wu Y, Anthony EJ. A study on the activity of CaO-basedsorbents for capturing CO2 in clean energy processes. Appl Energy 2010;87(4):1453–8. http://dx.doi.org/10.1016/j.apenergy.2009.08.010.

[31] Sun R, Li Y, Liu H, Wu S, Lu C. CO2 capture performance of calcium-basedsorbent doped with manganese salts during calcium looping cycle.Appl Energy 2012;89(1):368–73. http://dx.doi.org/10.1016/j.apenergy.2011.07.051.

[32] Feng B, An H, Tan E. Screening of CO2 adsorbing materials for zero emissionpower generation systems. Energy Fuels 2007;21:426.

[33] Yamaguchi T, Niitsuma T, Nair BN, Nakagawa K. Lithium silicate basedmembranes for high temperature CO2 separation. J Membr Sci 2007;294:16.

[34] Ochoa-Fernandez E, Rusten HK, Jakobsen HA, Ronning M, Holmen A, Chen D.Sorption enhanced hydrogen production by steam methane reforming usingLi2ZrO3 as sorbent: sorption kinetics and reactor simulation. Catal Today2005;106:41.

[35] Ma S, Simmons JM, Yuan D, Li J-R, Weng W, Liu D-J, et al. A nanotubularmetal-organic framework with permanent porosity: structure analysis andgas sorption studies. Chem Commun (Camb) 2009;1(27):4049–51.

[36] Webb PA. Introduction to chemical adsorption analytical techniques andtheir applications to catalysis. MIC Technical Publications; 2003.

[37] Yu C, Hung C, Tan C. A review of CO2 capture by absorption and adsorption.Aerosol Air Qual Res 2012:745–69.

[38] Siriwardane RV, Shen M, Fisher EP, Poston JA, Box PO, Virginia W.Adsorption of CO2 on molecular sieves and activated carbon. Energy Fuels2001;15:279–84.

[39] Harlick PJE, Sayari A. Applications of pore-expanded mesoporous silicas,triamine silane grafting for enhanced CO2 adsorption”. Ind Eng Chem Res2006;45:3248.

[40] Chue KT, Kim JN, Yoo YU, Cho SH, Yang RT. Comparison of activated carbonand zeolite 13X for CO2 recovery from flue gas by pressure swing adsorption.Ind Eng Chem Res 1995;34:591.

[41] Maurin G, Llewellyn PL, Bell RG. Adsorption mechanism of carbon dioxide infaujasites: grand canonical monte carlo simulations and microcalorimetrymeasurements. J Phys Chem B 2005;109:16084.

[42] Hicks JC, Drese JH, Fauth DJ, Gray ML, Qi G, Jones CW. Designing adsorbentsfor CO2 capture from flue gas-hyperbranched aminosilicas capable ofcapturing CO2 reversibly. J Am Chem Soc 2008;130:2902.

[43] Leal O, Bolivar C, Ovalles C, Garcia J, Espidel Y. Reversible adsorption of carbondioxide on amine surface-bonded silica gel. Inorg Chim Acta 1995;240:183.

[44] Sirwardane RV. Solid sorbents for removal of carbon dioxide from gas streamsat low temperatures. U.S. Patent 6,908,497 B1, 1, 2005.

[45] Xu X, Song C, Miller BG, Scaroni AW. Adsorption separation of carbon dioxidefrom flue gas of natural gas-fired boiler by a novel nanoporous. Fuel ProcessTechnol 2005;86:1457.

[46] Lee S, Filburn TP, Gray M, Park J-W, Song H-J. Screening test of solid aminesorbents for CO2 capture. Ind Eng Chem Res 2008;47:7419.

[47] Hiyoshi N, Yogo DK, Yashima T. Adsorption of carbon dioxide on aminemodified SBA-15 in the presence of water vapor. Chem Lett 2004;33:510.

[48] Mazumder S, van Hemert P, Busch A, Wolf K-H, Tejera- Cuesta P. Flue gas andpure CO2 sorption properties of coal: a comparative study. Int J Coal Geol2006;67:267.

[49] Plaza MG, González AS, Pevida C, Pis JJ, Rubiera F. Valorisation of spent coffeegrounds as CO2 adsorbents for postcombustion capture applications. ApplEnergy 2012;99:272–9. http://dx.doi.org/10.1016/j.apenergy.2012.05.028.

[50] Martin CF, Plaza MG, Pis JJ, Rubiera F, Pevida C, Centeno TA. On the limits ofCO2 capture capacity of carbons. Sep Purif Technol 2010;74:225.

[51] Silvestre-Albero J, Wahby A, Sepǔlveda E, Martínez-Escandell M, Kaneko K,Rodríguez-Reinoso F. Ultrahigh CO2 adsorption capacity on carbon molecularsieves at room temperature. Chem Commun 2011;47:6840.

[52] Plaza MG, Pevida C, Arias B, Fermoso J, Arenillas A, Rubiera F, et al. ThermAnal Calorimeters 2008;92:601–6.

[53] Radosz M, Hu X, Krutkramelis K, Shen Y. Flue-gas carbon capture oncarbonaceous sorbents: toward a low-cost multifunctional carbon filter for‘‘green” energy producers. Ind Eng Chem Res 2008;47:3783.

[54] Thallapally PK, McGrail PB, Atwood JL, Gaeta C, Tedesco C, Neri P. Carbondioxide capture in a self-assembled organic nanochannels. Chem Mater2007;19:3355.

[55] Thallapally PK, McGrail PB, Dalgarno SJ, Schaef HT, Tian J, Atwood JL. Gas-induced transformation and expansion of a non-porous organic solid. NatMater 2008;7:146.

[56] Furukawa H, Yaghi OM. Storage of hydrogen, methane, and carbon dioxide inhighly porous covalent organic frameworks for clean energy applications. JAm Chem Soc 2009;131:8876.

[57] Babarao R, Jiang JW. Molecular screening of metal�organic frameworks forCO2 storage. Langmuir 2008;24:6270.

[58] Babarao R, Jiang JW. Unprecedentedly high selective adsorption of gasmixtures in rho zeolite-like metal�organic framework: a molecularsimulation study. J Am Chem Soc 2009;131:11417.

[59] Deng S, Wei H, Chen T, Wang B, Huang J, Yu G. Superior CO2 adsorption onpine nut shell-derived activated carbons and the effective micropores atdifferent temperatures. Chem Eng J 2014;253:46–54. 1 October.

[60] Heidari A, Younesi H, Rashidi A, Ghoreyshi AA. Evaluation of CO2adsorption with eucalyptus wood based activated carbon modified byammonia solution through heat treatment. Chem Eng J 2014;254:503–13.15 October.

http://ipcc.ch/report/srccs/

http://refhub.elsevier.com/S0306-2619(15)01238-6/h0205























































































































































[61] Balsamo M, Budinova T, Erto A, Lancia A, Petrova B, Petrov N, et al. CO2adsorption onto synthetic activated carbon: kinetic, thermodynamic andregeneration studies. Sep Purif Technol 2013;116:214–21. 15 September.

[62] Yin G, Liu Z, Liu Q, Wu W. The role of different properties of activated carbonin CO2 adsorption. Chem Eng J 2013;230(15):133–40.

[63] Caglayan BS, Erhan Aksoylu A. CO2 adsorption on chemically modifiedactivated carbon. J Hazard Mater 2013;252–253:19–28. 15 May.

[64] Babu DJ, Lange M, Cherkashinin G, Issanin A, Staudt R, Schneider JJ. Gasadsorption studies of CO2 and N2 in spatially aligned double-walled carbonnanotube arrays. Carbon 2013;61(September):616–23.

[65] Sawant SY, Somani RS, Bajaj HC, Sharma SS. A dechlorination pathway forsynthesis of horn shaped carbon nanotubes and its adsorption properties forCO2, CH4, CO and N2. J Hazard Mater 2012;227–228:317–26. 15 August.

[66] Hsu SC, Lu C, Su F, Zeng W, Chen W. Thermodynamics and regenerationstudies of CO2 adsorption on multi-walled carbon nanotubes. Chem Eng Sci2010;65(4):1354–61.

[67] Lithoxoos GP, Labropoulos A, Peristeras LD, Kanellopoulos N, Samios J,Economou IG. Adsorption of N2, CH4, CO and CO2 gases in single walledcarbon nanotubes: a combined experimental and Monte Carlo molecularsimulation study. J Supercrit Fluids 2010;55(2):510–23.

[68] Su F, Lu C, Cnen W, Bai H, Hwang JF. Capture of CO2 from flue gas viamultiwalled carbon nanotubes. Sci Total Environ 2009;407(8):3017–23.

[69] Cinke M, Li J, Bauschlicher Jr CW, Ricca A, Meyyappan M. CO2 adsorption insingle-walled carbon nanotubes. Chem Phys Lett 2003;376(5–6):761–6.

[70] Zhou X, Yi H, Tang X, Deng H, Liu H. Thermodynamics for the adsorption ofSO2, NO and CO2 from flue gas on activated carbon fibre. Chem Eng J2012;200–202(15):399–404.

[71] Yang H, Gong M, Chen Y. Preparation of activated carbons and theiradsorption properties for greenhouse gases: CH4 and CO2. J Nat Gas Chem2011;20(5):460–4.

[72] Zhang Z, Xu M, Wang H, Li Z. Enhancement of CO2 adsorption on high surfacearea activated carbon modified by N2, H2 and ammonia. Chem Eng J2010;160(2):571–7.

[73] Shen C, Grande CA, Li P, Yu J, Rodrigues AE. Adsorption equilibria and kineticsof CO2 and N2 on activated carbon beads. Chem Eng J 2010;160(2):398–407.

[74] Moon SH, Shim JW. A novel process for CO2/CH4 gas separation on activatedcarbon fibers-electric swing adsorption. J Colloid Interface Sci 2006;298(2):523–8.

[75] Iijima S. Helical microtubules of graphitic carbon. Nature 1991;354:56.[76] Abuilaiwi FA, Laoui T, Al-HArthi M, Mautaz AA. Modification and

functionalization of multiwalled carbon nanotube (MWCNT) via Fischeresterification. Arabian J Sci Eng 2010;35(1C):37.

[77] Ebbesen TW, Ajayan PM. Large-scale synthesis of carbon nanotubes. Nature1992;358(6383):220–2.

[78] Thess A, Lee RP, Nikolaev HJ, Dai P, Petit J, Robert CH, et al. Crystalline ropes ofmetallic carbon nanotubes. Science 1996;273(52):483–7.

[79] Deepak S, Chenyu W, Kyeongjae C. Nanomechanics of carbon nanotubes andcomposites. ASME 2003;56(2):215–29.

[80] Harris P. Carbon nanotubes and related structures: new materials for thetwenty first century. United Kingdom: Cambridge University Press; 2001.

[81] Ching-Hwa K, William G, Robert B, Donald S. Carbon nanotubes with single-layer walls. Carbon 1995;33(7):903–14.

[82] Jung SH, Kim MR, Jeong SH, Kim SU, Lee OJ, Lee KH, et al. High-yield synthesisof multi-walled carbon nanotubes by arc discharge in liquid nitrogen. ApplPhys A – Mater Sci Process 2003;76(2):285–6.

[83] Journet C, Bernier P. Production of carbon nanotubes. Appl Phys A – Mater SciProcess 1998;67(1):1–9.

[84] Andreas T, Roland L, Pavel N, Hongjie D, Pierre P, Jerome R, et al. Crystallineropes of metallic carbon nanotubes. Science 1996;273:483–7.

[85] Kabbashi NA, Muataz AA, Al-Mamun A, Mohamed ES, Alam MZ, Yahya N.Kinetic adsorption of application of carbon nanotubes for Pb(II) removal fromaqueous solution. J Environ Sci 2009;21:539–44.

[86] Tawabini, Al-Khaldi S, Muataz AA, Khaled M. Removal of mercuryfrom water by multi-walled carbon nanotubes. Water Sci Technol 2010;61(3):591.

[87] Muataz AA, Bakather OY, Tawabini BS, Bukhari AA, Khaled M, Al-Harthi M,et al. Removal of chromium (III) from water by using modified andnonmodified carbon nanotubes. J Nanomater 2010. 9 pages Article ID 232378.

[88] Muataz AA, Omer YB, Alaadin AB, Faraj AA, Mohamed F. Effect of carboxylicfunctional group functionalize on carbon nanotubes surface on the removal oflead from water. J Bioinorgan Chem Appl 2010:9 Article ID 603978.

[89] Tawabini S, AL-Khaldi SF, Khaled MM, Muataz AA. Removal of arsenic fromwater by iron oxide nanoparticles impregnated on carbon nanotubes. JEnviron Sci Health Part A 2011;46:215–23.

[90] Al-Hakami SM, Khalil AB, Laoui T, Atieh MA. Fast disinfection of escherichiacoli bacteria using carbon nanotubes interaction with microwave radiation.Bioinorg Chem Appl 2013;458943:1–9.

[91] Zeino A, Abulkibash A, Khaled M, Atieh M. Bromate removal from water usingdoped iron nanoparticles on multi-walled carbon nanotubes (CNTs). JNanomater 2014;2014:1–9 ID 561920.

[92] Fatemi S, Vesali-Naseh M, Cyrus M, Hashemi J. Improving CO2/CH4adsorptive selectivity of carbon nanotubes by functionalization withnitrogen-containing groups. Chem Eng Res Des 2011;89(9):1669–75.

[93] Gui MM, Yap YX, Chai SP, Mohamed A. Multi-walled carbon nanotubesmodified with (3-aminopropyl)triethoxysilane for effective carbon dioxideadsorption. Int J Greenhouse Gas Control 2013;14(May):65–73.

[94] Su F, Lu C, Chung AJ, Liao CH. CO2 capture with amine-loaded carbonnanotubes via a dual-column temperature/vacuum swing adsorption. ApplEnergy 2014;113(January):706–12.

[95] Liu Q, Shi Y, Zheng S, Ning L, Ye Q, Tao M, et al. Amine-functionalized low-costindustrial grade multi-walled carbon nanotubes for the capture of carbondioxide. J Energy Chem 2014;23(1):111–8.

[96] Su F, Lu C, Chen H-S. Adsorption, desorption, and thermodynamic studies ofCO(2) with high-amine-loaded multiwalled carbon nanotubes. Langmuir2011;27(13):8090–8.

[97] Dillon EP, Crouse CA, Barron AR. Synthesis, characterization, and carbondioxide adsorption of covalently attached polyethyleneimine-functionalizedsingle-wall carbon nanotubes. ACS Nano 2008;2(1):156–64.

[98] Zhao YY, T Jung B, Ansaloni L, Ho WSW. Multiwalled carbon nanotube mixedmatrix membranes containing amines for high pressure CO2/H2 separation. JMembr Sci 2014;459:233–43.

[99] Düren T. How does the pore morphology influence the adsorptionperformance of metal-organic frameworks: a molecular simulation study ofmethane and ethane adsorption in Zn-MOFs. In: International zeoliteconference, vol. i; 2007.p. 2042–2047.

[100] Furukawa H, Cordova KE, O’Keeffe M, Yaghi OM. The chemistry andapplications of metal-organic frameworks. Science 2013;341:1230444.

[101] Kuppler RJ, Timmons DJ, Fang Q-R, Li J-R, Makal Ta, Young MD, et al. Potentialapplications of metal-organic frameworks. Coord Chem Rev 2009;253(23–24):3042–66.

[102] Lu H. Interfacial synthesis of metal-organic frameworks. McMasterUniversity; 2012.

[103] Carné-Sánchez A, Imaz I, Cano-Sarabia M, Maspoch D. A spray-dryingstrategy for synthesis of nanoscale metal-organic frameworks and theirassembly into hollow superstructures. Nat Chem 2013;5(3):203–11.

[104] Campagnol N, Van Assche T, Boudewijns T, Denayer J, Binnemans K, De Vos D,et al. High pressure, high temperature electrochemical synthesis of metal–organic frameworks: films of MIL-100 (Fe) and HKUST-1 in differentmorphologies. J Mater Chem A 2013;1(19):5827.

[105] Li M, Dinc M. Reductive electrosynthesis of crystalline metal-organicframeworks. American Chemical Society (MIT Open Access Articles); 2014.

[106] Monika P. High throughput assisted investigation on lanthanide (III)tetrakisphosphonates. Ludwig-Maximilans-Universitat Munchen; 2009.

[107] Lu CM, Liu J, Xiao K, Harris AT. Microwave enhanced synthesis of MOF-5 andits CO2 capture ability at moderate temperatures across multiple capture andrelease cycles. Chem Eng J 2010;156(2):465–70.

[108] Kharissova OV, Kharisov BI, Méndez UO. Microwave-assisted synthesis ofcoordination and organometallic compounds. INTECH; 2009.

[109] CCDC. Support Solution @ www.ccdc.cam.ac.uk; 2014.[110] Plaines D, Levan D, Brandani S, Snurr R, Matzger A, Arbor A. Carbon dioxide

removal from flue gas using microporous metal organic frameworks finaltechnical report reporting start date : reporting end date : principal author :report issued : DOE Award Number : Submitting Organization : 1 April 2007Richard Willis; 2010.

[111] Zhao Z, Li Z, Lin YS. Adsorption and diffusion of carbon dioxide on metal–organic framework (MOF-5). Ind Eng Chem Res 2009;48:10015–20.

[112] Xie LH, Suh MP. Flexible metal-organic framework with hydrophobic pores.Chemistry 2011;17(49):13653–6.

[113] Yuan D, Zhao D, Zhou H-C. Pressure-responsive curvature change of a ‘rigid’geodesic ligand in a (3,24)-connected mesoporous metal-organic framework.Inorg Chem 2011;50(21):10528–30.

[114] Coudert FX. The osmotic framework adsorbed solution theory: predictingmixture coadsorption in flexible nanoporous materials. Phys Chem ChemPhys 2010;12(36):10904–13.

[115] Yuan D, Getman RB, Wei Z, Snurr RQ, Zhou H-C. Stepwise adsorption in amesoporous metal-organic framework: experimental and computationalanalysis. Chem Commun (Camb) 2012;48(27):3297–9.

[116] Zhao G, Aziz B, Hedin N. Carbon dioxide adsorption on mesoporous silicasurfaces containing amine-like motifs. Appl Energy 2010;87(9):2907–13.http://dx.doi.org/10.1016/j.apenergy.2009.06.008.

[117] Mellot-draznieks C, Fuchs AH, Boutin A. Prediction of breathing and gate-opening transitions upon binary mixture adsorption in metal–organicframeworks franc. JACS Commun 2009:11329–31.

[118] Nakhla J. Metal-organic-frameworks @ www.sigmaaldrich.com. AldrichChem files; 2009.

[119] Leus K. The development of vanadium and titanium containing metal organicframeworks for applications in adsorption and heterogeneous catalysis karenleus department of inorganic and physical chemistry. Universiteit Ganet;2012.

[120] González A, Plaza M, Pis J, Rubiera F, Pevida C. Post-combustion CO2 captureadsorbents from spent coffee. Energy Procedia 2013;37:134–41.

[121] Marco-Lozar JP, Kunowsky M, Suarez-Garcıa F, Linares-Solano A. Sorbentdesign for CO2 capture under different flue gas conditions. Carbon2014;72:125–34.

[122] Plaza M, García S, Rubiera F, Pis J, Pevida C. Post-combustion CO2 capturewith a commercial activated carbon: comparison of different regenerationstrategies. Chem Eng J 2010;163:41–7.

[123] Thiruvenkatachari R, Su S, Yu XX, Bae J-S. Application of carbon fibrecomposites to CO2 capture from flue gas. Int J Greenhouse Gas Control2013;13:191–200.

[124] González A, Plaza M, Rubiera F, Pevida C. Sustainable biomass-based carbonadsorbents for post-combustion capture. Chem Eng J 2013;230:456–65.






































































































































http://www.ccdc.cam.ac.uk


















http://www.sigmaaldrich.com

















[125] Gil MV, Alvarez-Gutierrez N, Martinez M, Rubiera F, Pevida C, Moran A.Carbon adsorbents for CO2 capture from bio-hydrogen and biogas streams:breakthrough adsorption study. Chem Eng J 2015;269:148–58.

[126] Hosseini Soraya, Bayesti Iman, Marahel Ehsan, Babadi Farahnaz Eghbali,Abdullah Luqman Chuah, Choong Thomas SY. Adsorption of carbon dioxideusing activated carbon impregnated with Cu promoted by zinc. J TaiwanInstit Chem Eng 2015;52:109–17.

[127] Dantas TLP, Luna FMT, Silva Jr IJ, Torres AEB, De Azevedo DCS, Rodrigues AE,et al. Modeling of the fixed-bed adsorption of carbon dioxide and a carbondioxide–nitrogen mixture on zeolite 13X. Brazilian J Chem Eng 2011;28(03):533–44.

[128] Dantas TLP, Luna FMT, Silva IJ, de Azevedo DCS, Grande CA, Rodrigues AE,et al. Carbon dioxide–nitrogen separation through adsorption on activatedcarbon in a fixed bed. Chem Eng J 2011;169(1–3):11–9.

[129] Dantas TL, Luna FMT, Silva Jr IJ, Torres AEB, Azevedo DCd, Rodriguesc AE, et al.Carbon dioxide–nitrogen separation through pressure swing adsorption.Chem Eng J 2011;172:698–704.

[130] Wang L, Liu Z, Li P, Yu J, Rodrigues AE. Experimental and modelinginvestigation on post-combustion carbon dioxide capture using zeolite13X-APG by hybrid VTSA process. Chem Eng J 2012;197:151–61.

[131] Cho Youngmin et al. LiOH-embedded zeolite for carbon dioxide captureunder ambient conditions. J Ind Eng Chem 2015;22:350–6.

[132] Hefti Max, Marx Dorian, Joss Lisa, Mazzotti Marco. Adsorption equilibrium ofbinary mixtures of carbon dioxide and nitrogen on zeolites ZSM-5 and 13X.Microporous Mesoporous Mater 2015;215:215–28.

[133] Ye Sheng et al. Post-combustion CO2 capture with the HKUST-1 and MIL-101(Cr) metal–organic frameworks: Adsorption, separation and regenerationinvestigations. Microporous Mesoporous Mater 2013;179:191–7.

[134] Xu WW, Pramanik S, Zhang Z, Emge TJ, Li J. Microporousmetalorganicframework[M2(hfipbb)2(ted)] (M¼Zn, Co; H2hfipbb¼4,4-(hexafluoroisopropylidene)-bis(benzoicacid); ted¼triethylenediamine):Synthesis, structure analysis, pore characterization, smallgasad-sorptionandCO2/N2 separation properties. J Solid State Chem 2013;200:1–6.

[135] Sabouni R, Kazemian H, Rohani S. Carbon dioxide adsorption in microwave-synthesized metal organic framework CPM-5: equilibrium and kineticsstudy. Microporous Mesoporous Mater 2013;175:85–91.

[136] Millward AR, Yaghi OM. Metal�organic frameworks with exceptionally highcapacity for storage of carbon dioxide at room temperature. J Am Chem Soc2005;127:17998–9.

[137] Nalaparaju A, Khurana M, Farooq S, Karimi IA, Jiang JW. CO2 capture incation-exchanged metal-organic frameworks: holistic modeling frommolecular simulation to process optimization. Chem Eng Sci 2015;124:70–8.

[138] Camacho Barbara CR, Ribeiro Rui PPL, Esteves Isabel AAC, Mota Jose PB.Adsorption equilibrium of carbon dioxide and nitrogen on the MIL-53 (Al)metal organic framework. Sep Purif Technol 2015;141:150–9.

[139] Xian Shikai, Peng Junjie, Zhang Zhijuan, Xia Qibin, Wang Haihui, Li Zhong.Highly enhanced and weakened adsorption properties of two MOFs by watervapor for separation of CO2/CH4 and CO2/N2 binary mixtures. Chem Eng J2015;270:385–92.

[140] Hedin N, Andersson L, Bergström L, Yan J. Adsorbents for the post-combustion capture of CO2 using rapid temperature swing or vacuumswing adsorption. Appl Energy 2013;104:418–33. http://dx.doi.org/10.1016/j.apenergy.2012.11.034.

[141] Sumida K, Rogow DL, Mason JA, McDonald TM, Bloch ED, Herm ZR, et al.Carbon dioxide capture in metal-organic frameworks. Chem Rev 2012.

[142] Andersen A, Divekar S, Dasgupta S, Cavka JH, Nanoti Aarti, Spjelkavik A, et al.On the development of Vacuum Swing adsorption (VSA) technology for post-combustion CO2 capture. Energy Procedia 2013;37:33–9.

[143] Yazaydın AAO, Snurr QR, Park T-H, Koh K, Liu J, LeVan MD, et al. Screening ofmetal-organic frameworks for carbon dioxide capture from flue gas using acombined experimental and modeling approach. J Am Chem Soc2009;131:18198–9.

[144] Han S, Huang Y, Watanabe T, Nair S, Walton KS, Sholl DS, et al. MOF stabilityand gas adsorption as a function of exposure to water, humid air, SO2, andNO2. Microporous Mesoporous Mater 2013;173:86–91.

[145] Yu J, Balbuena PB. Water effects on postcombustion CO2 capture in Mg-MOF-74. J Phys Chem 2013;117:3383–8.

[146] Fracaroli AM, Furukawa H, Suzuki M, Dodd M, Okajima S, Gándara F, et al.Metal�organic frameworks with precisely designed interior for carbondioxide capture in the presence of water. J Am Chem Soc 2014.

[147] Chaffee AL, Knowles GP, Liang Z, Zhang J, Xiao P, Webley PA. CO2 capture byadsorption: materials and process development. Int J Greenhouse Gas Control2007;1:11–8.

[148] Li G, Xiao P, Webley P, Zhang J, Singh R, Marshall M. Capture of CO2 from highhumidity flue gas by vacuum swing adsorption with zeolite 13X. Adsorption2008;14:415–22.

[149] Mérel J, Clausse M, Meunier F. Carbon dioxide capture by indirect thermalswing adsorption using 13X zeolite. Environ Prog 2006;25:327–33.

[150] Wang L, Yang Y, Shen W, Kong X, Li P, Yu J, et al. Experimental evaluation ofadsorption technology for CO2 capture from flue gas in an existing coal-firedpower plant. Chem Eng Sci 2013;101:615–9.

[151] Gomes VG, Yee KW. Pressure swing adsorption for carbon dioxidesequestration from exhaust gases. Sep Purif Technol 2002;28:161–71.

[152] Cho S, Park J, Beum H, Han S, Kim J. A 2-stage PSA process for the recovery ofCO2 from flue gas and its power consumption. Stud Surf Sci Catal2004;153:405–10.

[153] Zhang J, Webley P, Xiao P. Effect of process parameters on powerrequirements of vacuum swing adsorption technology for CO2 capturefrom flue gas”. Energy Convers Manage 2008;49:346–56.

[154] Choi W, Kwon T, Yeo Y, Lee H, Song H. Optimal operation of the pressureswing adsorption (PSA) process for CO2 recovery. Korean J Chem Eng2003;20:617–23.

[155] Merel J, Clausse M, Meunier F. Experimental investigation on CO2 post-combustion capture by indirect thermal swing adsorption using 13X and 5Azeolites. Ind Eng Chem Res 2008;47(1):209–15.

[156] Clausse M, Merel J, Meunier F. Numerical parametric study on CO2 capture byindirect thermal swing adsorption. Int J Greenhouse Gas Control2011;5:1206–13.

[157] Ribeiro RPPL, Grande CA, Rodrigues AE. Activated carbon honeycombmonolith – Zeolite 13X hybrid system to capture CO2 from flue gasesemploying Electric Swing Adsorption. Chem Eng Sci 2013;104:304–18.

[158] Cavenati S, Grande CA, Rodrigues AE. Adsorption equilibrium of methane,carbon dioxide, and nitrogen on zeolite 13X at high pressures. J Chem EngData 2004;49(4):1095–101.

[159] Casas N, Schell J, Blom R, Mazzotti M. MOF and UiO-67/MCM-41 adsorbentsfor pre-combustion CO2 capture by PSA: breakthrough experiments andprocess design. Sep Purif Technol 2013;112(July):34–48.

[160] Sipöcz N, Tobiesen FA, Assadi M. The use of Artificial Neural Network modelsfor CO2 capture plants. Appl Energy 2011;88(7):2368–76. http://dx.doi.org/10.1016/j.apenergy.2011.01.013.

[161] Mohsen S. Mathematical modeling of single and multi-componentadsorption fixed beds to rigorously predict the mass transfer zone andbreakthrough curves. Iran J Chem Chem Eng 2009;28(3):25–44.

[162] Shafeeyan MS, Wan Daud WMA, Shamiri A. A review of mathematicalmodeling of fixed-bed columns for carbon dioxide adsorption. Chem Eng ResDes 2013(August).

[163] Ruthven DM. Principles of adsorption and adsorption processes.pdf. 10thed. New Brunswick, Fredericton: Wiley & Sons; 1984.

[164] Kim MB, Bae Y-S, Choi D-K, Lee C-H. Kinetic separation of landfill gas by atwo-bed pressure swing adsorption process packed with carbon molecularsieve: nonisothermal operation. Ind Eng Chem Res 2006;45(14):5050–8.

[165] Jee J, Park H, Haam S, Lee C. Effects of nonisobaric and isobaric stepson O2 pressure swing adsorption for an aerator. Ind Eng Chem Res2002;41:4383–92.

[166] Rezaei F, Grahn M. Thermal management of structured adsorbents in CO2capture processes. Ind Eng Chem Res 2012.

[167] Khalighi M, Farooq S, Karimi IA. Nonisothermal pore diffusion model for akinetically controlled pressure swing adsorption process. Ind Eng Chem Res2012.

[168] Kumar R. Adsorption column blowdown: adiabatic equilibrium model forbulk binary gas mixtures. Ind Eng Chem Res 1989;28(11):1677–83.

[169] Delgado JA, Uguina MA, Sotelo JL, Ruíz B. Fixed-bed adsorption of carbondioxide–helium, nitrogen–helium and carbon dioxide–nitrogen mixturesonto silicalite pellets. Sep Purif Technol 2006;49(1):91–100.

[170] Delgado JA, Uguina MA, Sotelo JL, Ruíz B, Rosário M. Separation of carbondioxide/methane mixtures by adsorption on a basic resin. Adsorption2007;13(3–4):373–83.

[171] Shendalman LH, Mitchell JE. A study of heatless adsorption on the systemCO2 in the He(l). Chem Eng Sci 1972;27:1449–58.

[172] Cen P, Yang RT. Separation of a five-component gas mixture by PSA. Sep SciTechnol 1985;20:725.

[173] Raghavan NS, Hassan MM, Ruthven DM, Joboken NJ. Numerical simulation ofa PSA system. Part I: isothermal trace component system with linearequilibrium and finite mass transfer resistance (1985). AIChE J 1985;31(1985):385–92.

[174] Kapoor A. Kinetic separation of methane-carbon dioxide mixture byadsorption on molecular sieve carbon. Chem Eng Sci 1989;44(8).

[175] Cavenati S, Grande CA, Rodrigues AE. Upgrade of methane from landfill gas bypressure swing adsorption. Energy Fuels 2005;19(6):2545–55.

[176] Ahn H, Brandani S. Dynamics of carbon dioxide breakthrough in a carbonmonolith over a wide concentration range. Adsorption 2005;11(S1):473–7.

[177] Hwang KYES, Jun JAEHO, Lee WONK. Fixed-bed adsorption for bulkcomponent. Chem Eng Sci 1995;50(5).

[178] Chou CT, Chen C-Y. Carbon dioxide recovery by vacuum swing adsorption.Sep Purif Technol 2004;39(1–2):51–65.

[179] Mulgundmath VP, Jones RA, Tezel FH, Thibault J. Fixed bed adsorption for theremoval of carbon dioxide from nitrogen: breakthrough behaviour andmodelling for heat and mass transfer. Sep Purif Technol 2012;85:17–27.

[180] Doong SJ, Yang RT. Bulk separation of multi-component gas mixtures bypressure swing adsorption: pore/surface diffusion and equilibrium models.AIChE J 1986;32(3):397–410.

[181] Diagne D, Gotob M, Hiroseb T. Numerical analysis of a dual refluxed PSAprocess during simultaneous removal and concentration of carbon dioxidedilute gas from air. J Chem Technol Biotechnol 1996;65:29–38.

[182] Lee CH, Yang J, Ahn H. Effects of carbon-to-zeolite ratio on layered bed H2 PSAfor coke oven gas. AIChE J 1999;45(3):535–45.

[183] Park JH, Kim J-N, Cho S-H. Performance analysis of four-bed H2 PSA processusing layered beds. AIChE J 2000;46(4):790–802.

[184] Kaguei S, Wakao N. Validity of infinite bed assumption in the estimation ofparameters from thermal waves measured in a non-isothermal adsorptioncolumn. Chem Eng Sci 1985;40(10):1851–3.


















































































































































































[185] Hwang KSH, Lee WK. The adsorption and desorption breakthrough behaviorof carbon dioxide on activated carbon. effect of total pressure and pressure-dependent mass transfer coefficients. Sep Purif Technol 1994;29(14):1857–91.

[186] Carter JW, Husain H. The simultaneous adsorption of carbon dioxide andwater vapour by fixed beds of molecular seives. Chem Eng Sci 1974;29(1):267–73.

[187] Ding Y, Alpay E. Equilibria and kinetics of CO2 adsorption on hydrotalciteadsorbent. Chem Eng Sci 2000;55(17):3461–74.

[188] Takamura SUY, Narita S, Aoik J, Hironaka S. Evaluation of double bed pressureswing adsorption for CO2 recovery from boiler gas exhaust. Elsevier Sep PurifTachnol 2001;24:519–28.

[189] Cavenati S, Grande CA, Rodrigues AE. Separation of mixtures by layeredpressure swing adsorption for upgrade of natural gas. Chem Eng Sci 2006;61(12):3893–906.

[190] Moreira RFPM, Soares JL, Casarin GL, Rodrigues AE. Adsorption of CO2 onhydrotalcite-like compounds in a fixed bed. Sep Sci Technol 2006;41(2):341–57.

[191] Biswas P, Agrawal S, Sinha S. Modeling and simulation for pressure swingadsorption system for hydrogen purification. Chem Biochem Eng Q 2010;24(4):409–14.

[192] Agarwal A. Advanced strategies for optimal design and operation of pressureswing adsorption processes; 2010.

[193] Casas N, Schell J, Pini R, Mazzotti M. Fixed bed adsorption of CO2/H2 mixtureson activated carbon: experiments and modeling. Adsorption 2012;18(2):143–61.

[194] Sabouni R. Carbon dioxide adsorption by metal-organic frameworks(synthesis, testing and modelling). University of Western Ontario –Electronic Thesis and Dissertation Repository. Paper 1472; 2013 <http://ir.lib.uwo.ca/etd/1472>.

[195] Krishnamurthy S, Rao VR. Separations: materials, devices and processes CO2

capture from dry flue gas by vacuum swing adsorption: a pilot plant study.AIChE J (Separations Mater Dev Process) 2014;60(5).

[196] Glueckauf E, Coates JI. Theory of chromatography, Part IV, The influence ofincomplete equilibrium on the front boundary of chromatograms and on theeffectiveness of separation. J Chem Soc (Resumed) 1947:1315–21.

[197] Bao Z, Yu L, Ren Q, Lu X, Deng S. Adsorption of CO2 and CH4 on amagnesium-based metal organic framework. J Colloid Interface Sci 2011;353(2):549–56.

[198] Saha D, Bao Z. Adsorption of CO2, CH4, N2O and N2 on MOF-5, MOF-177, andZeolite 5A. Environ Sci Technol 2010;44(5):1820–6.




























http://ir.lib.uwo.ca/etd/1472

http://ir.lib.uwo.ca/etd/1472












Documents

Carbon capture by physical adsorption: Materials ... · A review on carbon capture by physical adsorption is provided. The review covers carbon capture materials, experimental and