A Molecular Simulation Study of CO Adsorption in Metal-Organic Frameworks

A molecular simulation study of CO2 adsorption in
metal-organic frameworks
Pan YANG
Master of Philosophy (Engineering)
Supervisor: Dr Qinghua Zeng

Co-Supervisor: Dr Kejun Dong
School of Computing, Engineering and Mathematics

Western Sydney University
Sydney, Australia
May 2018
ACKNOWLEDGEMENTS
I would like to thank my principal supervisor Dr Qinghua Zeng for offering me the
opportunity to do the molecular dynamics research on metal-organic frameworks. I really
appreciate his patience, enthusiasm, encouragement and broad knowledge. I was
continuously advised and supported by his guidance in all the time of research timeframe and
writing of this thesis.
I am also grateful for the comments and feedbacks given by my co-supervisor Dr Kejun
Dong during my 2-year research period and my thesis writing.
Also, I appreciate many friends and visiting scholars for their insightful views and valuable
suggestions.
Lastly, thanks to the research service team of SCEM and HDR, Western Sydney University
for organising conferences and providing administrative and technical support.
ii
ABSTRACT
The increasing carbon dioxide density in the atmosphere has led to the global warming and
other environmental issues. Such increase in carbon dioxide comes mainly from the
combustion of thousands of tons of fossil fuels (coal, oil and natural gas). Thus, the
development of novel materials for CO2 capture, separation and sequestration is becoming
critically essential. Many materials including aqueous amine solvent, micro and mesoporous
solid substances have been extensively investigated for CO2 absorption/adsorption. It is found
that quite a lot of distinct metal-organic frameworks (MOFs) have remarkable CO2
adsorption capacity in room temperature, particularly HKUST-1 and MIL-68(In). HKUST-1
(Hong Kong University of Science and Technology) is a MOF made up of copper nodes with
1,3,5-benzenetricarboxylic acid struts between them. MIL-68(In) is an indium-based MOF
made up of the chains of InO4(OH)2 octahedral units which are interconnected with
terephthalate ligands to form central triangle and hexagon channels. To understand the micro-
mechanisms and hence to further improve their adsorption capacities, grand canonical Monte
Carlo and molecular dynamics simulations have been employed in this study to explore CO2
adsorption uptake. The results show that both fluorinated HKUST-1 and MIL-68(In) have a
superior CO2 adsorption performance than their unmodified structures. In addition, radial
distribution function and mean square displacement analysis methods have been carried out
to explore the CO2 adsorption sites and the self-diffusion within MOFs. The results indicate
that CO2 molecules in HKUST-1 are more likely to attach to the metal sites with the pressure
increase, however, the adsorbed CO2 in HKUST-1F and MIL-68F (In) are bound firmly to
fluoride atoms, reducing their mobility.
Keywords: HKUST-1, MIL-68(In), fluorination, CO2 adsorption, diffusion, molecular
dynamics simulation
iii
LIST OF FIGURES
Figure 1. Brief concept of carbon dioxide capture and storage/sequestration (CCS) (Lee &
Park 2015). ................................................................................................................................. 5
Figure 2. General reaction mechanisms for the chemical absorption of CO2 in amine-based
solvent (Bosoaga, Masek & Oakey 2009; Rochelle 2009). ....................................................... 7
Figure 3. (a) Solid adsorption mechanisms, (b) Adsorption process in packed beds, and (c)
Adsorption process in fluidized beds (Yu, Huang & Tan 2012). .............................................. 8
Figure 4. (a) Cu-BTC (dry) (b) 4wt. % hydrated Cu-BTC. Orange, red, and grey spheres
represent Cu, O, and C atoms, respectively. The oxygen atom of the coordinated water
molecule is omitted in blue (Yazaydın et al. 2009). ................................................................ 15
Figure 5. Periodic boundary condition..................................................................................... 19
Figure 6. Schematic diagram of a basic MD code(Zeng, Yu & Lu 2010). .............................. 21
Figure 7. Mean square displacement. ...................................................................................... 22
Figure 8. Radial distribution function. ..................................................................................... 23
Figure 9. The snapshot of a CO2 adsorbed equilibrium MOF structure: CO2 adsorbed MOF
(upper), isolated MOF (lower left) and isolated adsorbed CO2 molecules (lower right). ....... 24
Figure 10. Unit cells of HKUST-1 (a) and HKUST-1F (b), Colour: Copper in orange,
oxygens in red, carbons in grey, hydrogen in white and fluorides are light blue. ................... 27
Figure 11. Adsorption isotherm of CO2 in HKUST-1 and HKUST-1F at 298K under the
pressure of from 0.1 to 5bar. .................................................................................................... 29
Figure 12. The CO2 (red spots) adsorption snapshot in HKUST-1, 1 bar (left) and 5 bar
(right). ...................................................................................................................................... 30
Figure 13. The CO2 (red spots) adsorption snapshot in HKUST-1F, 0.1 bar (left) and 5 bar
(right). ...................................................................................................................................... 31
Figure 14. Radial distribution function of atomic pairs from HKUST-1 and CO2 at 298 K and
different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar, respectively. ...... 34
Figure 15. Radial distribution function of atomic pairs from HKUST-1F and CO2 at 298 K
and different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar, respectively. 37
Figure 16. Mean square displacement over time of CO2 in HKUST-1and HKUST-1F.......... 38
iv
Figure 17. Unit cells of MIL-68 (a) and MIL-68F (In) (b), Colour: Indium in dark red,
oxygens in red, carbons in grey, hydrogen in white and fluorides are light blue. ................... 41
Figure 18. Adsorption isotherm of CO2 in MIL-68(In) and MIL-68F (In) at 298K under the
pressure of from 0.1 to 5 bar. ................................................................................................... 43
Figure 19. Comparison of CO2 adsorption isotherms of CO2 in HKUST-1 and MIL-68(In). 44
Figure 20. The CO2 (red spots) adsorption snapshot of MIL-68(In), 0.1 bar (left) and 5 bar
(right). ...................................................................................................................................... 45
Figure 21. The CO2 (red spots) adsorption snapshot of MIL-68F (In), 0.1 bar (left) and 5 bar
(right). ...................................................................................................................................... 45
Figure 22. The radial distribution function for atomic pairs of MIL-68 and CO2 at 298 K and
Figure 23. Radial distribution function of atomic pairs from MIL-68F and CO2 at 298 K and
Figure 24. Mean squared displacement over time of CO2 in MIL-68(In) and MIL-68F (In).. 52
v
LIST OF TABLES
Table 1 The Atomic partial charge q (e) and force field of HKUST-1 and HKUST-1F ........ 28
Table 2 CO2 adsorption capacity on HKUST-1(Cu-BTC) under different conditions ............ 29
Table 3 Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1 .................. 39
Table 4 Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1F................ 39
Table 5 The Atomic partial charge q (e) and force field of MIL-68(In) and MIL-68F (In) ... 41
Table 6 Interaction potential energies (Kcal/mol) of CO2 adsorbed in MIL-68 (In) ............... 53
Table 7 Interaction potential energies (Kcal/mol) of CO2 adsorbed in MIL-68F (In)............. 53
vi
LIST OF PUBLICATIONS
Conference paper: A molecular dynamics of CO2 adsorption in fluoridised cations and non-
fluoridised HKUST-1. 1st International Conference on Mineral Engineering and Materials
Science (iCMEMS2017), 20-22 Nov 2017, Sydney, Australia.
vii
TABLE OF CONTENTS
DECLARATION ........................................................................................................................ i
ACKNOWLEDGEMENTS .......................................................................................................ii
ABSTRACT ............................................................................................................................. iii
LIST OF FIGURES .................................................................................................................. iv
LIST OF TABLES .................................................................................................................... vi
LIST OF PUBLICATIONS .....................................................................................................vii
CHAPTER 1 INTRODUCTION ............................................................................................... 1
1.1 Research background and objectives ................................................................................... 1
1.2 Thesis outline ....................................................................................................................... 2
CHAPTER 2 LITERATURE SURVEY .................................................................................... 3
2.1 Global warming and carbon emissions ................................................................................ 3
2.2 Carbon dioxide capture and storage ..................................................................................... 4
2.3 Amine absorption for CO2 emissions .................................................................................. 5
2.4 Metal-organic frameworks for CO2 adsorption and separation ........................................... 7
2.4.1 Experimental work for CO2 capture on MOFs ................................................................. 9
2.4.2 Molecular simulations for CO2 capture on MOFs .......................................................... 11
2.5 Current research summary and challenges ........................................................................ 14
CHAPTER 3 METHODOLOGY ............................................................................................ 16
3.1 Input parameters................................................................................................................. 16
3.1.1 Atomic partial charge ...................................................................................................... 16
3.1.2 Force field ....................................................................................................................... 17
3.2 Ensemble ............................................................................................................................ 18
3.3 Periodic boundary condition and cut-off distance ............................................................. 18
3.4 Principle of Grand Canonical Monte Carlo simulation ..................................................... 19
3.5 Principle of molecular dynamics simulation...................................................................... 20
viii
3.6 Molecular dynamics analysis methods .............................................................................. 21
3.6.1 Mean square displacement .............................................................................................. 22
3.6.2 Radial distribution function ............................................................................................ 22
3.6.3 Adsorption strength ......................................................................................................... 23
3.7 Summary ............................................................................................................................ 25
CHAPTER 4 CARBON DIOXIDE ADSORPTION IN HKUST-1 ........................................ 26
4.1 Model configuration........................................................................................................... 26
4.2 Grand Canonical Monte Carlo Simulation details ............................................................. 27
4.2.1 Adsorption isotherms of CO2 in HKUST-1 and HKUST-1F ......................................... 28
4.2.2 Comparison of CO2 adsorption isotherms of CO2 in HKUST-1 .................................... 29
4.3 Molecular dynamics simulation results ............................................................................. 30
4.3.1 Adsorption sites of CO2 in HKUST-1 and HKUST-1F .................................................. 30
4.3.2 Self-diffusivity of CO2 in HKUST-1 and HKUST-1F ................................................... 37
4.3.3 Interaction strength between CO2 and HKUST-1 ........................................................... 38
4.4 Summary ............................................................................................................................ 39
CHAPTER 5 CARBON DIOXIDE ADSORPTION IN MIL-68(In) ...................................... 40
5.1 Model configuration........................................................................................................... 40
5.2 Grand Canonical Monte Carlo Simulation details ............................................................. 41
5.2.1 Adsorption isotherms of CO2 in MIL-68(In) and MIL-68F (In) .................................... 42
5.2.2 Comparison of CO2 adsorption isotherms of CO2 in HKUST-1 and MIL-68(In) .......... 43
5.3 Molecular dynamics simulation results ............................................................................. 44
5.3.1 Adsorption sites of CO2 in MIL-68(In) and MIL-68F (In) ............................................. 44
5.3.2 Self-diffusivity of CO2 in MIL-68(In) and MIL-68F (In) .............................................. 51
5.3.3 Interaction strength between CO2 and MIL-68(In) ......................................................... 52
5.4 Summary ............................................................................................................................ 53
CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS ............................................ 54
6.1 Conclusions ........................................................................................................................ 54
ix
6.2 Recommendations on future work ..................................................................................... 55
REFERENCES ........................................................................................................................ 56
x
CHAPTER 1 INTRODUCTION
1.1 Research background and objectives
The global energy demand is increasingly being driven by the population boom and
economic growth (Krausmann et al. 2009). Fossil fuels (e.g. coal, natural gas and oil) are
playing a significant role in meeting such enormous demand since industrialization
revolution; on the other hand, the combustion of fossil fuels is contributing to tonnes of
greenhouse gases emissions (Quadrelli & Peterson 2007). The rising level of CO2
concentration in the atmosphere is one of the greatest environmental concerns facing our
civilization today. This presents an urgent need for carbon dioxide emissions being mitigated
from the thermoelectric power plants and other industrial manufactures (Cormos 2014).
Aqueous amine solvents are currently used in some power generation plants for CO2 capture
and scrubbing, which could reduce the carbon emissions from the exhausted streams
(Rochelle, GT 2009). However, this technique is recognised to be energy intensive and will
lead to a sharp decrease in plant efficiency (Rao & Rubin 2002). In order to lower the energy
penalty, various porous solid materials (e.g. zeolites, metal oxides, activated carbons as well
as metal-organic frameworks) have been synthesized and probed (Yu, Huang, & Tan 2012).
Among them, MOFs, emerging as a novel type of porous solid adsorbents, have attracted
many research interests in the past two decades because of their outstanding physical and
chemical properties for gas adsorption and separation (Yazaydin et al. 2009), For instance,
the highest CO2 uptake of MOF-177 is reported to be 147 wt. % (Millward & Yaghi 2005).To
date, there are more than 20,000 distinct MOFs being reported for the control of CO2
emissions (Banerjee 2012; Saha et al. 2012).
Great efforts have been made by the conventional experiments on CO2 capture and separation
in MOFs such as MOF-5, MIL-101 and MOF-177 (Banerjee 2012; Liu, Wang & Zhou 2012;
Samanta et al. 2012). However, it is impractical to synthesize and screen hundreds of MOFs
1
in laboratory, especially the MOFs having specific pore size/shape and functional groups
(Torrisi, Bell & Mellot-Draznieks 2010). Moreover, the direct observations in laboratory of
CO2 diffusion and other dynamic properties are extremely difficult to achieve (Salles et al.
2008). In contrast, with the ever-growing computational technique, molecular simulations
(e.g. Monte Carlo and molecular dynamic simulation) have become a more favorable method
to understand the mechanisms of MOFs for gas capture or separation, particularly for CO2
adsorption performance and diffusion behavior (Sabouni, Kazemian & Rohani 2014).
The present work aims to investigate CO2 adsorption capacities in HKUST-1 and MIL-68(In)
as well as their fluorinated structures (HKUST-1F and MIL-68F (In)) , identify CO2 binding
sites and self-diffusion in these four MOFs and calculate adsorption energy between MOFs
and CO2 via a combined grand canonical Monte Carlo (GCMC) & molecular dynamics (MD)
simulations.
1.2 Thesis outline
The layout of this thesis is described as follows: Chapter 1 briefly introduces the background
information and objectives; Chapter 2 reviews the development of techniques and materials
for CO2 capture and separation, particularly in MOFs; Chapter 3 consists of theories of
computational simulations and analysis methods; Chapter 4 presents the grand canonical
Monte Carlo and molecular dynamics simulations results of CO2 adsorption in HKUST-1 and
its fluorinated structure (HKUST-1F); In Chapter 5 discusses grand canonical Monte Carlo
and molecular dynamics simulations results of CO2 adsorption in MIL-68(In) and MIL-68F
(In); Chapter 6 summarises the present outcomes, and some recommendations for future
work are also proposed.
2
CHAPTER 2 LITERATURE SURVEY
2.1 Global warming and carbon emissions
The global energy demand is predicted to surge significantly in line with world population
growth and the urbanization development. The emissions of greenhouse gases from the
combustion streams of fossil fuels are playing a critical role in leading to global warming
which makes our human beings endangered and the entire planet impaired (Achten, Almeida
& Muys 2013; Lee et al. 2013). Unfortunately, due to the huge energy need for economic
growth and industrialisation in more and more countries, these emissions are expected to
have a non-stop increase in the short-to medium-term (Bains, Psarras & Wilcox 2017; Manan
et al. 2017). What’s more, it seems unfeasible in the near future relying on some new
resources or renewable energies, e.g. wind, solar, tidal, wave and offshore powers to meet the
enormous energy demand. Therefore, fossil fuels will remain to be an overwhelming energy
provider to fulfil the ever-growing global energy needs (Cherp et al. 2016).
It is estimated that about 86% of greenhouse gas emissions are from the combustion of fossil
fuels. In Australia, over 70% of the energy demand is contributed by the primary resources,
particularly by coal and natural gas (Quadrelli & Peterson 2007). Moreover, among the
exhausted combustion streams (e.g.,CO2, CO, NOx, SOX, etc.), CO2 contributes to over 60%
of global warming and greenhouse gas effect (Bains, Psarras & Wilcox 2017). In 2012,
statistics from World Metrological Organisation (WMO) alarmed that the concentration of
CO2 in the atmosphere had already crept up to 393.1 ppm (Fang et al. 2015; Fang et al. 2014).
In April 2015, the United States National Oceanic and Atmospheric Administration (NOAA)
announced that carbon dioxide concentration exceeded 400 ppm in the atmosphere and would
reach 550 ppm by the year of 2050 (Schlör, Fischer & Hake 2015; Vujic & Lyubartsev
2017). Besides, the increased CO2 concentration in the atmosphere plays a negative effect on
the radiative balance of the Earth leading to the destruction of the ecosystems (Barcza 2017;
Jackson et al. 2017). The global warming leads to floods, droughts, heat waves as well as sea
level rise (AghaKouchak et al. 2014; Levin et al. 2012). As a result, the economic loss due to
climate change is predicted up to be 20% of the world’s gross domestic product (GDP)
(Mohor & Mendiondo 2017; Takakura et al. 2017).
3
2.2 Carbon dioxide capture and storage
Owing to the concerns of the atmospheric carbon density and urgent needs for the carbon
dioxide emissions mitigation, a wide range of technologies and strategies have been
developed and promoted to combat the anthropogenic CO2 contributions from power plants
and other industries (Boot-Handford et al. 2014; Pietzcker et al. 2014). Among them, carbon
dioxide capture and storage/sequestration (CCS) is believed to be prospective for carbon
emissions reduction, it was reported that the CCS technology could help to control
approximately one third of the global CO2 emissions. The Intergovernmental Panel on
Climate Change (IPCC) has established a number of academic and practical collaborations
and programmes promoting the greenhouse emission reduction. IPCC estimates that the CCS
applications can reduce up to 80% of CO2 emissions in the coal-fired thermoelectric power
plants. Furthermore, the European Union (EU) predicts that there will be 30 billion tons of
CO2 captured and stored by CCS to the year of 2050-2060 (Azar et al. 2010; Shackley &
Verma 2008).
In general, CCS technique embodies a group of techniques, as shown in Figure 1: (I)
Compression of CO2, e.g., absorbing or adsorbing the carbon dioxide gases from the
thermoelectric plants or other chemical industrials by aimed-based, alkaline solvents or
advanced porous solid materials (Cormos 2012; de Coninck, Stephens & Metz 2009). (II)
Transportation of CO2, utilising the pipeline or ship to freight the captured or compressed
CO2 gases to the facilities for storage (Chrysostomidis et al. 2009; Jung et al. 2013). (III)
CO2 permanent storage, e.g. marine reserve in the deep sea bed, carbonates solidification
underground deposit (oilfield and basins) and the storage post-management such as CO2
storage monitoring and observation (Eldevik et al. 2009; Randolph & Saar 2011). In addition,
CCS also embraces other strategies to reduce CO2 emissions, such as improving energy
combustion efficiency and developing alternative powers (Hussain et al. 2013; Kowalczyk et
al. 2013).
In summary, the development of CCS schemes is of great importance in removing
anthropogenic CO2 emissions and stabilizing carbon density in the atmosphere, CCS is a
direct and feasible manner to ease the global warming and environmental degradation.
Although efforts have been made, challenges still exist due to the different policies in
developed and developing countries and difficulties of global collaborations on CO2
emissions reduction(Benson et al. 2012; Gibbins & Chalmers 2008) .
4
Figure 1. Brief concept of carbon dioxide capture and storage/sequestration (CCS)
(Lee & Park 2015).
2.3 Amine absorption for CO2 emissions
Among the schemes of CCS, carbon dioxide capture technology is the most critical aspect
which accounts for approximately 75% of the total cost. Furthermore, the regeneration energy
requirement in the capture phase is regarded as the most key factor for the CCS technology
commercialization. The improvement of CO2 capture will play a significant role in cutting
down the total cost of CCS in near-term (Chu 2009).
Since the thermoelectric power plants contribute to the largest proportion (nearly 45%) of the
global carbon emissions, the CO2 capture from the post-combustion streams will have
greatest potential for reducing the greenhouse gas emissions (Drage et al. 2012).
In practical, aqueous amine absorbents have been employed industrially to scrub the post-
combustion gaseous streams for over 50 years. This technique has been currently equipped in
the coal or gas-fired electricity plants for the removal of CO2 and other acid gases such as
CO, NOX and SOx. This process involves the chemical reaction between aqueous amines and
CO2 to form bicarbonate, carbamate or carbonate under different conditions (e.g.
temperature, H2O and PH), as shown in Figure 2 (Bosoaga, Masek & Oakey 2009; Rochelle
2009). And the regeneration of the amine-based adsorbents normally occurs at elevated
5
temperatures (approximately 373-403 K) to release the CO2 gases (D'Alessandro, Smit &
Long 2010).
The primary alkanol-amine, monoethanolamine (MEA) is one of the most used amine
solvents utilized for CO2 capture industrially. A typical commercial process allow MEA
absorption solution (25-30 wt.%) to react with exhausted gas streams to form the amine-CO2
liquid, then this amine-CO2 liquid will be pumped to stripper tower and heated at high
temperature to regenerate the MEA and desorb the CO2 (Mangalapally et al. 2009). More
recently, the secondary amines such as diethanolamine(DEA) and triethanolamine (TEA) ,
the tertiary amine N-mythyl-diethanolamine (MDEA) as well as other amine mixtures are
also successfully implemented in the industrial manufactures to reduce the post-combustion
CO2 emissions (D'Alessandro, Smit & Long 2010; Garcia-Abuin, Gomez-Diaz & Navaza
2014; Pinto et al. 2014).
The apparent benefit of aqueous amines is that the technology has been maturely and
commercially equipped at a large number of power plants and other industrials for a couple
of years, the extensive experiences of this technique have been obtained. However, there still
exists a great deal of drawbacks and limitations using this technique such as relatively low
CO2 absorption capacity, the low mass transfer and reaction kinetics due to slow reaction
between CO2 and solvent, chemical degradation caused by other by-products (e.g. NOx and
SOx), corrosion and oxidative issues and, most importantly, the high energy demand of the
regeneration process to recycle the amines and liberate CO2 gases at elevated temperature
will sharply increase the total energy requirement by 25-40% of a power generation plant
which is equivalent to an additional CO2 capture cost by $50-60/tonne (Abanades, Rubin &
Anthony 2004; Le Moullec & Kanniche 2011; Lucquiaud & Gibbins 2011; Rochelle, G et al.
2011). Although advances and efforts have been made to overcome the limitations such as
the low absorption capacity, for example, the MDEA possess a higher CO 2 loading capacity
compared with MEA, the significant energy penalties still present a great challenge for amine
absorption technique.
6
Figure 2. General reaction mechanisms for the chemical absorption of CO2 in amine-
based solvent (Bosoaga, Masek & Oakey 2009; Rochelle 2009).
2.4 Metal-organic frameworks for CO2 adsorption and separation
Given the obvious disadvantages of the amine-based solvent for post-combustion CO2
capture, especially the intensive energy penalties for regeneration process, research attempts
have been turned on the diverse range of solid adsorbents for CO2 adsorption and separation.
A great many of studies and papers have highlighted the advances and advantages of solids
adsorbents for gas adsorption and separation, particularly the solid adsorbents have striking
benefits such as low energy requirement in the regeneration process and chemical stability
compared with aqueous amines (Abanades, Rubin & Anthony 2004; Samanta et al. 2012; Yu,
CH, Huang, CH & Tan, CS 2012).
The adsorption mechanism of solid adsorbents typically involves weak van der Waals (vdW)
forces for physisorption, stronger covalent bonding for chemisorption or, in some cases, a
hybrid sorption in packed or fluidized beds as shown in Figure 3 (Yu, C-H, Huang, C-H &
Tan, C-S 2012). The adsorption and removal (desorption) of CO2 is achieved by three main
cycles, temperature swing adsorption (TSA), pressure swing adsorption (PSA) and vacuum
swing adsorption (VSA). In practical, TSA and PAS approaches are often combined to
accomplish the adsorption process by heating/increasing the temperature and reducing/lowing
the pressure to recycle the adsorbents and liberate CO2 (Li et al. 2013; Liu, Wang & Zhou
2012; Nugent et al. 2013; Yu, Huang & Tan 2012).
7
Figure 3. (a) Solid adsorption mechanisms, (b) Adsorption process in packed beds,
and (c) Adsorption process in fluidized beds (Yu, Huang & Tan 2012).
In general, the adsorption capacity is a one of the most significant factors evaluating the
performance of solid adsorbents which describes the number or the amount of the gases can
be captured in adsorbent. In addition to adsorption capacity, the affinity strength (bonding or
packing interactions) and kinetic effect (diffusion or transport behaviors) are simultaneously
playing a critical role in characterizing and assessing the adsorbents (Li, Kuppler & Zhou
2009).To date, tones of thousands of solid adsorbents which are microporous or mesoporous
materials such as zeolites (e.g. zeolite 5A and 13X), silica materials, metal oxides (e.g. CaO
and MgO), hydrotalcite-like compounds (HTLCs), carbon or carbonaceous adsorbents (e.g.
MAXSORB) as well as metal-organic frameworks (MOFs)have been extensively considered
and investigated for CO2 capture and separation under different conditions. However, there
are some distinct limitations for these solid physical adsorbents. Firstly, the weak affinity and
relatively low adsorption capacity such as activated carbons and carbon fibers. Secondly, the
degradation phenomenon of the adsorbents, for example, the CaO is easily to degrade after
several carbonation-calcination cycles at high temperature. Thirdly, zeolites and silica
materials are moisture sensitive which is unavailable for wet streams. Lastly, due to the
complex topology of HTLCs compounds and silica materials as well as the poor thermal
stability, these adsorbents confront great challenges for recycling and renewal in the
regeneration step. Thus, recent research attentions have been drawn to overcome these
shortcomings (Alhamami, Doan & Cheng 2014; Chałupnik et al. 2013; Miller, Akbar &
Morris 2014; Rossi, Campos & Souza 2016; Seema et al. 2014; Thommes & Cychosz 2014).
8
In the past two decades, metal-organic frameworks (MOFs), emerging as a new type of
progressive materials, have attracted a great deal of research interests for gas adsorption and
separation (Li, Y-W et al. 2017). A large variety of MOFs have been studied and reported in
the journals and literatures on the basis of their outstanding structural diversities (e.g. pore
size/shape and geometries), high surface area, prominent thermal stability, moisture
insensitivity and distinct functionalization (Furukawa et al. 2013; Hu et al. 2014; Li, Kuppler
& Zhou 2009). In general, MOFs are microporous crystalline substances with uniform pore
size in the range of 3-20 Å which can be controlled and tailored. They are typically three-
dimensional coordination networks consisting of inorganic nodes (metal-based single node or
clusters) and surrounded by specific functional organic ligands or linkers (e.g. carboxylate,
tribenzoate, imidazolate and pyridyl) (Maurin et al. 2017; Tudisco et al. 2016; Zhang, S et al.
2016). Today, MOFs materials have been widely investigated and explored for the gas
adsorption and separation (e.g. H2, CH4, N2 and CO2), which presents both opportunities and
challenges (Hu et al. 2014; Zhang, Z et al. 2013).
2.4.1 Experimental work for CO2 capture on MOFs
Due to the distinct advantages and favourable properties such as high void volume,
remarkable thermal stability and low crystal density, MOFs have become one of the most
prospective research candidates for gas capture and adsorption, particularly for small gas
molecules. Intensive systematic research programmes have been implemented for the CO2 on
MOFs both experimentally and numerically (Li et al. 2013; Li et al. 2015; Shi et al. 2014).
The great progresses have been achieved in the last two decades for MOFs’ design, synthesis
and characterization. Basically, MOFs includes two main categories: rigid and flexible, rigid
MOFs often have stable pores such as MOF-177 and MOF-5, while the structure of flexible
MOFs like MIL-53 and MIL-47 will have a great change during the adsorption process.
Experimental test is recognized as the most direct approach to evaluate the adsorption
performance of the MOFs by the observation of equilibrium isotherms. To date, several
synthetic approaches have been significantly developed and a large scale of MOFs have been
synthesized and explored for CO2 adsorption and separation in laboratories. (Canivet et al.
2014; Ghosh, Colon & Snurr 2014; Xian et al. 2015)
MOFs usually have exceptional CO2 adsorption capacity compared with zeolites and
activated carbons due to the outstanding pore/surface properties, for example, the surface area
of activated carbon is 400-1000 m2/g and zeolite of 1500 m2/g maximum, whereas MOFs has
9
a typical surface area of 1500-4500 m2/g. As a result, a superior CO2 adsorption of MOF-177
with a surface area of 4746 m2/g was 147 wt. % which was reported by Yaghi’s team
(Millward & Yaghi 2005). Moreover, It was demonstrated that MOF-210 (surface area 6240
m2/g) and MOF-200 (surface area 4530 m2/g) which are both zinc-based MOFs illustrated
carbon dioxide uptake of 74.2 wt. % and 73.9 wt. %, respectively at 298K under a high
pressure (50 bar) (Furukawa et al. 2010).
In addition to the optimal pore/surface properties, the effects of open metal site in many
MOFs have been extensively exploited. Wu and co-workers studied Mg-MOF-74 and
HKUST-1, they reported that both MOFs had a striking CO2 adsorption capacity which may
probably attribute by the strong electrostatic interaction due to the correlatively unsaturated
Mg and Cu cations (Wu et al. 2010). Millward and co-workers demonstrated that MOF-177
contains open Zn4+ metal sites could reach an excellent CO2 uptake which was much larger
other adsorbents such as zeolite 13X and MAXSORB (activated carbon) (Hao, Li & Lu
2011).Furthermore, MIL-100 and MIL-101, containing unsaturated Cr3+ sites, revealed a
superior uptake of CO2 over CH4 by the difference of (quadruple) polarity/ non-polarity, the
mechanism involves a coordination of O=C=O and Cr3+ metal centre (Llewellyn et al. 2008).
Thus, the open metal site in MOFs is of great importance for CO2 adsorption exploration.
More interestingly, some flexible MOFs possess a unique “gate-opening” or “breathing”
phenomenon during the adsorption process; these MOFs usually have a small amount of
adsorption capacity at the low pressure but show a sudden increase after a threshold pressure ,
temperature or gas loading (Alhamami, Doan & Cheng 2014; Millange et al. 2008). MIL-53
series are favourable candidates for the study of CO2 adsorption due to their phase
transformation. For example, MIL-53(Cr) had an inferior CO2 uptake under 5bar, however, a
remarkable adsorption increase in CO2 uptake occurred as the narrow-pore (NP) transformed
to be large-pore (LP) (Hamon, L. et al. 2009). Similarly, the “gate” effect Kitagawa and co-
workers illustrated that Cu (pyrdc) (bpp) (pyrdc=pyridine-2, 3-dicarboxylate, bpp=1, 3-bis (4-
pyridyI) propane) had a distinct CO2 uptake increase at a comparatively low pressure(Maji,
Ohba & Kitagawa 2005).
Additionally, the chemical modification and functionalization on MOFs have simultaneously
been investigated and probed to understand the CO2 adsorption capacity enhancement in
recent years.(Feijani, Tavasoli & Mahdavi 2015; Luo et al. 2015). A great many of research
attentions have been drawn on shafting the functional groups on the organic ligands or metal
sites of MOFs (Ma, Abney & Lin 2009; Wang & Cohen 2009). In 2009, Botas et.al observed
an outstanding 58.0 wt. % CO2 uptake at 273K and 10 bars on MOF-5 (Zn4OBDC,
10
BDC=1,4-benzenedicarboxylate) when doped with Co (Botas et al. 2010). Moreover, Yan
et.al examined and compared the adsorption performances on the simple chemically modified
and parent Cu-BTC (BTC=1,3,5-benezenetricaboxylate), it showed that the functionalized
MOF had a superior CO2 uptake than pristine Cu-BTC (Xie et al. 2012). Later, It was also
reported that CO2 adsorption capacity increased significantly from 79.2 to 176 wt. % on MIL-
101 by the ethanol and NH4F treatment (Berdonosova et al. 2013).
Tremendous achievements and endeavours have been made for the CO2 adsorption on MOFs
experimentally (Li, J-R et al. 2011). However, challenges still exist. Firstly, as metal-organic
frameworks usually consist of various atoms, the array of metal clusters and organic ligands
could be infinite, thus, it may be impractical to synthesize and characterize immense
hypothesized and perceived MOFs in laboratory. Additionally, the experiment is time-
consuming and cost-inefficient for MOFs sampling and testing, and the experiments under
extreme high temperature or pressures may be quite difficult to implement. More importantly,
the insight observations of kinetic and transport properties for adsorbate and adsorbent could
not be directly viewed by experimental method. In contrast, with rapid-growing
computational resources, computational simulations could provide anther optimal research
option to understand CO2 adsorption and separation within MOFs (Duren, Bae & Snurr 2009;
Park, Lively & Sholl 2017; Walton et al. 2008).
2.4.2 Molecular simulations for CO2 capture on MOFs
Computational simulations can provide the great help for researchers to screen and
characterize not only current microporous materials but also construct the conceived
substances’ architectures which are difficult to synthesize experimentally. In addition, owing
to the limited experiment explorations on CO2 dynamics (diffusion) within MOFs,
computational simulations are also able to supplement the shortcomings in this field(de Pablo
et al. 2014; Yu et al. 2016). Conventional computational techniques mainly consist of
quantum-based calculations (ab initio and density function theory) and molecular simulations
(Grand Canonical Monte Carlo and Molecular Dynamics). These methods are widely utilized
to understand structure-property relationships for porous materials (Evins 2013; Hafner
2007).
The ab initio and density functional theory (DFT) are often employed to simulate the detailed
chemical or physical interactions between solid adsorbents and small gases. In many porous
materials such as activated carbons, the CO2 molecules are physically bonded within the
11
pores via van der Waals forces. Ab initio could accurately calculate such weakly bonded
energy. One of the high level ab initio methods, second-ordered Moller-Plesset (MP2) shows
quite reliable and accurate results in describing the noncovalent interactions, however, the
major drawback is that MP2 method could only compute for small systems containing tens of
atoms which may be deficient to characterize many state-of-art microporous material
(Vogiatzis et al. 2009).
In terms of DFT technique, it is more practical and feasible to investigate relatively “big”
systems, and the B3LYP (a typical DFT method) has a high computational efficiency
compared with MP2. Although B3LYP can provide a finer approximation of electron
correlation effects, it has some apparent deficiencies when calculating the long-range
electrostatic interactions and the short-term vdW potentials regarding in simulation of CO2
adsorption in MOFs (Dreizler & Gross 1990; Handy 1996; Orio, Pantazis & Neese 2009).
Since molecular simulation force field is a more sensible approach to calculate the vdW
interactions, it has been implemented as an alternative to compute the CO2 adsorption
performance within MOFs(Tian, Dai & Jiang 2016).
To date, molecular simulations method is regarded as the one of the most prospective
technology to predict the gas adsorption properties within porous materials such as MOFs
and zeolites. Molecular simulations mainly comprise of grand canonical Monte Carlo
(GCMC) and molecular dynamics (MD). To date, GCMC simulation is extensively employed
to computationally predict the gas uptake in porous materials, whereas MD simulation is
more appropriate for the investigation of structural and dynamic properties of adsorbate
molecules in a wide range of materials. In recent years, molecular simulations have become
an increasingly critical instrument to explore and study not only the adsorption capacity but
also dynamic properties between various MOFs and CO2 (Mosher et al. 2013; Torrisi, Bell &
Mellot-Draznieks 2010).
GCMC simulation is recognized as one of the most direct and efficient computational
approach to examine the gas adsorption capacity in MOFs. In addition, GCMC can also
provide a fine benchmark or verification for experimental results (Chen et al. 2014; Lawler,
Hulvey & Forster 2015). Apart from the adsorption capacity, the diffusion explorations are
simultaneously essential to understand the kinetic properties. In this respect, MD simulation
plays a significant role to address the research gaps which is hard to achieve experimentally
(Ciccotti, Ferrario & Schuette 2014).
Yazaydin et al. compared the adsorption performance of dry Cu-BTC and hydrated Cu-BTC
structures, as shown in Figure 4. The dry Cu-BTC (Figure 4a) illustrated a CO2 adsorption
12
capacity of 18.4 wt. % under 1 bar at 298K. Interestingly, when the Cu-BTC was hydrated
with 4% of water molecular (Figure 4b), the capacity surged remarkably to 27 wt. %. The
results of computer simulations were subsequently verified by experiments. They identified
that the electric field created by water molecules could coordinate with the copper metal site
to significantly enhance CO2 uptake (Yazaydın et al. 2009). In 2007, Yang’s group conducted
a combined GCMC simulation on the subject of separation performance of CO2-containing
gas mixtures on Cu-BTC, the simulation results demonstrated that the intense interactions
between quadrupole moment of CO2 and Cu2+ metals could improve the CO2 selectivity at
low pressure (Yang et al. 2007). Yang and Zhong carried out a hybrid GCMC simulation and
Ideal adsorbed solution theory (IAST) approach to examine the adsorption performance on a
number of porous materials, revealing that MOFs exhibited advantageous CO2 adsorption
capacity than zeolite and carbon adsorbents (Yang, Zhong & Chen 2008). In 2012, Vaesen
and co-workers conducted a joint modelling and experiment approach to study the various
gas uptake and separation on a series of MIL-68. They reported that the all of the CO2
uptakes in MIL-68s were higher than CH4 (Yang et al. 2012).
In addition to the adsorption isotherms, the dynamic properties such as diffusion behaviour
and permeability are of great significance characterizing and screening MOFs, which is
favourable for pre-design and post-synthesis of the current and new MOFs (Jelfs & Cooper
2013; Meier, Laesecke & Kabelac 2001; Parkes et al. 2014).
In 2008, Babarao and Jiang studied the CO2 and methane diffusion in IRMOF-1, silicate and
C168 Schwarzite, and they reported decreased self-diffusivities of CO2 in these three
adsorbents as the loading increased, which is a common tendency in many porous materials
due to the steric hindrance(Babarao & Jiang 2008). Salles et al. observed the self and
transport diffusivity of CO2 in MIL-47(V) by a combined MD simulations and quasi-elastic
neutron scattering experiments which revealed that CO2 molecules in MIL-47(V) did not
perform like an idea gas due to their intermolecular interactions(Salles, Fabrice et al. 2013).
Sholl and co-workers employed a combined GCMC and MD simulation approach to probe
the separation of CO2 and methane mixture in a membrane-based MOF. They proved that the
diffusion of CO2 was less likely to be influenced by the slowly moving methane, more
importantly, this results corrected the previous view that the membrane-based separation
selectivity would be lower than adsorption-based separation(Keskin & Sholl 2007). Similarly,
Yang’s group demonstrated the adsorption and diffusion of CO2 would be at the same time
and in the same direction during the adsorption process(Yang & Zhong 2006). Later, another
study on bio-MOF-11 for CO2 and H2 adsorption (CO2/H2=15:85) illustrated that CO2
13
molecules were less close to the metal sites but preferentially binding to the organic linkers
as reflected from the radial distribution function (RDF) results after the MD simulation(Atci,
Erucar & Keskin 2011). In summary, molecular simulations can overcome some of
experimental limitations, particularly the microscopic insight observations and
characterizations at the aspects of dynamic and transport properties.
2.5 Current research summary and challenges
In this section, the great research efforts have been summarized and highlighted in the field of
CO2 capture within MOFs both experimentally and computationally. Researchers in
laboratory have devoted to synthesize and screen hundreds of MOFs for gas adsorption and
separation(Li, Kuppler & Zhou 2009), however, it may be quite impossible to compound
such numerous MOFs’ models for experimental testing (Li et al. 2011). On the other hand,
with ever-growing computational technology, GCMC and MD simulations are becoming
more rational and efficient to evaluate the adsorption properties effectively and economically
on MOFs. Moreover, the structural, kinetic and transport analysis methods could provide a
microscopic guide to refine and design of the current and novel MOFs (Ciccotti, Ferrario &
Schuette 2014; Parkes et al. 2014; Zhou, Long & Yaghi 2012).
However, great research challenges and gaps still exist. The MOFs’ performance is usually
determined by interplay factors such as porosity, thermodynamic equilibrium and kinetic
effect. In practice, it is more often a combination of these effects caused by open metal site,
structural flexibility and functional groups in MOFs (Luo et al. 2015; Lyndon et al. 2015;
Xian et al. 2015; Yan et al. 2015; Zhou et al. 2015). The effect of the unsaturated metal sites
in MOFs has been widely examined, for example, the enhanced CO2 adsorption uptake was
observed in the Cu-BTC (HKUST-1) by a coordinative interaction of CO2 and Cu2+ (Bordiga et al.
2007)
. In addition to the open metal site effect, the “gate-opening” or “breathing” mechanism
often occurs in flexible MOFs, especially in MIL-53 serious. MIL-53(Cr) possesses an
insignificant CO2 uptake below 5 bars, whereas a remarkable adsorption increase is observed
as the structure restore after CO2 molecules entering into the specific pores (Bourrelly et al.
2005). Moreover, the surface functionalization on metal sites or organic ligands could play a
critical role to enhance the affinity between CO2 and MOFs, but the synthesis and
investigations on fluoride treated MOFs were quite rare. Few attentions have been drawn on
the research of dual effects, for example, the coordination effect between structural
flexibility/open metal site and chemical functionalization. To our best acknowledge, few
14
molecular simulations have been implemented for CO2 binding locations and self-diffusion
on fluorinated MOFs, especially on HKUST-1 and MIL-68(In). Therefore, this present work,
we constructed fluoride modified HKUST-1 and MIL-68(In) on their organic linkers, by
using fluoride to replace all of the hydrogen atoms attached on the organic ligands, and then
we examined CO2 adsorption capacities on these MOFs by GCMC simulation, we examined
CO2 binding sites after MD simulation by RDF and observed the CO2 self-diffusion by MSD
method. Lastly, the energy interaction strength between MOFs and CO2 was calculated to
verify the GCMC results.
Figure 4. (a) Cu-BTC (dry) (b) 4wt. % hydrated Cu-BTC. Orange, red, and grey
spheres represent Cu, O, and C atoms, respectively. The oxygen atom of the
coordinated water molecule is omitted in blue (Yazaydın et al. 2009).
15
CHAPTER 3 METHODOLOGY
In present research work, a combined GCMC and MD simulation approach is adopted to

assess the CO2 adsorption performance on HKUST-1 and MIL-68(In) by Materials Studio
software. The combined GCMC and MD simulation is a method to predict not only the gas
adsorption capacity in the materials by grand canonical Monte Carlo simulation but also
explore the adsorption dynamics of the adsorption process by molecular dynamics simulation.
This method is, firstly carried out by GCMC simulation to numerically calculate the quantity
of the adsorbed gas, and MD simulation will be implemented to study adsorption, mechanism
of adsorbed gases such as diffusion properties of gases in MOFs. In present study, a
combined GCMC and MD simulation allows a more precise and practical approach to
achieve the research objectives.
Materials Studio is a comprehensive modelling and simulation tool for materials and science
engineering. It allows the researchers to better understand the materials’ properties at a
microscopic level. The GCMC simulation is utilized to predict the CO2 capacity under a
pressure swing adsorption, and the MD simulation is employed to observe the structural and
dynamic properties of CO2 molecules within MOFs.
3.1 Input parameters
In all of the GCMC and MD simulations, the input parameters including atomic partial
charge and force field are of great importance which should be determined and assigned prior
to the simulations.
3.1.1 Atomic partial charge
In some early computational studies for CO2 adsorption within MOFs, the intramolecular
interactions may be neglected in some cases. However, with the development of
computational method, the potentials of chemical bonds in MOFs have become significant
and may contribute to the computational accuracy. Therefore, it becomes essential to assign
the atomic charges for the precise calculation of electrostatic potentials within MOFs. And
due to the quadrupole moment of CO2, it is important to assign the atomic charges of O and C
atoms to calculate the electrostatic interactions. Please note that these charges have never
16
been observed experimentally or quantum mechanically (Collins & Woo 2017; Hamad et al.
2015; Li, W et al. 2017; Wilmer & Snurr 2011).
In general, there are a few approximations can characterize these charges. The most
extensively adopted method is electrostatic potential (ESP) which is the quantum-based
approach on the basis of DFT. However, this technique is often applied on non-periodic
models. The Mulliken Population Analysis method which is another DFT-based calculation
approach usually overestimates the charges. Very recently, a Molecular Charge Equilibration
method, the molecular charge equilibration (QEq) technique is usually performed for
molecular systems. Many studies have been carried out this method to obtain the atomic
charges in MOFs because it takes into account the geometry and the electronegativities of the
various atoms which is more precise for MOFs models with periodic boundary (Rappe &
Goddard 1991).
3.1.2 Force field
Apart from the atomic partial charge, the force field is another critical parameter for the
molecular simulations. In recent computational simulations, it is essential to describe all of
the related interactions between the atoms in the systems (Rappe & Goddard 1991). Force
field (FF) is known as a set of potential energies describing the bond (valence) and non-bond
interactions. The valence term includes the intramolecular interactions (e.g., bond, angle,
torsion and inversion). And the non-bond term generally consists of Lennard-Jones (LJ) and
long-rang electrostatic potentials (Columbic interactions). To date, a number of force fields
have been developed and adopted for the molecular simulations including DREIDING, UFF
(universal force field), COMPASS (Condensed-phase Optimised Molecular Potentials for
Atomistic Simulation Studies) and other specific force fields (Martin 2006; Pongsajanukul et
al. 2017). Among them, the UFF is the most extensively utilized force field for molecular
simulations as UFF has been rigorously tested by the Materials Studio researchers and results
are in agreement with published work. UFF is a purely diagonal, harmonic force field. In
UFF, the potential energy E is expressed as a sum of valence and nonbonded interactions in
Eq 1,
E= ER + Eθ + Eɸ + Eω + Evdw + ECoulombic (1)
Where bond stretching (ER) is described by a harmonic term, angle bending (Eθ) by a three-
term Fourier cosine expansion, and torsions (Eɸ) and inversions (Eω) by cosine-Fourier
expansion terms, the van der Waals (Evdw) interactions are described by the Lennard-Jones
17
potential, and electrostatic interactions (ECoulombic) are described by atomic monopoles and a
screened (distance-dependent) Coulombic term (Casewit, Colwell & Rappe 1992; Rappe et
al. 1992). UFF is adopted for present study since UFF presents a good compromise in the
molecular simulation in regarding to computational effectiveness and accuracy. In addition,
UFF is developed in conjunction with QEq method, thus universal force field usually has a
fine collaboration with the QEq charge equilibration method(Rappe & Goddard 1991).
3.2 Ensemble
An ensemble is a collection of all possible systems which have different microscopic states
but have an identical macroscopic or thermodynamic state. In other words, an ensemble is a
set of points in phase space satisfying the conditions of a particular thermodynamic
state(Nosé 1984). Different macroscopic environment constraints lead to different types of
ensembles with particular statistical characteristics (Adams 1975). Four ensembles are often
used in molecular simulations: Grand canonical ensemble (µVT), which has fixed chemical
potential (µ), volume (V) and temperature (T); Micro canonical ensemble (NVE), which has
fixed number of atoms (N), volume (V) and energy (E);Canonical ensemble (NVT), which has
fixed number of atoms (N), volumes (V) and temperature (T); Isothermal-isobaric ensemble
(NPT), which has fixed number of atoms (N), a fixed pressure (P) and temperature (T) (Din &
Michaelides 1997; Lamb & Jorgensen 1997). Besides, a numerical heat bath technique will
be adopted to add or eliminate heat in the course of time in NVT or NPT ensembles. The
Berendsen thermostat and Nosé-Hoover thermostat are two common used techniques for
molecular simulations (Berendsen et al. 1984; Hoover 1985; Nosé 1984).
3.3 Periodic boundary condition and cut-off distance
The term of periodic boundary condition refers to the simulation of structures consisting of
periodic lattices of identical motifs. This is crucial as it allows atoms or molecules in
simulation box to maintain a continuous trajectory while still experiencing the same potential
energy field (Makov & Payne 1995). For example, as shown in Figure 5, the central square
(Yellow) is the simulated zone and the surrounding squares simulation boxes are virtual
imagines. The particle (grey) in the central square moving out will reflect and emerge from
18
the neighbouring virtual simulation boxes to guarantee the force or potential energy can be
assigned on the particle.
All periodic boundary conditions imply a cut-off criterion to make a balance of computing
accuracy and efficiency. For instance, when the distance of two atoms is beyond a certain
range, the interactions between them can be ignored. Basically, the minimum image
convention advises a maximum distance for this cut-off will be no more than half the cell
dimensions (Pivkin & Karniadakis 2005).
Figure 5. Periodic boundary condition.
3.4 Principle of Grand Canonical Monte Carlo simulation
With well-defined force field to calculate the interactions, the grand canonical Monte Carlo
(GCMC) simulation, one of the most common methods predicting the gas adsorption in
porous materials, is adopted to examine the CO2 uptake in MOFs. In this method, the
temperature T, volume V and the chemical potentials µ are kept fixed, and the particles or
molecules are allowed to have trial moves (e.g., insertion and deletion) before the system
reaches to the equilibrium phase (pressure or temperature), these movements have equal
19
probability to randomly translate or rotate an existing particles or molecules, then the
adsorbed amount N of the particles or molecules will be calculated by a statistically approach
after the equilibrium stage (Duren, Bae & Snurr 2009; Frenkel & Smit 2002). In most cases,
the Metropolis is the most common used technique in GCMC simulation for CO2 adsorption
in MOFs (Metropolis et al. 1953). The adsorption isotherm which is the most significant
thermodynamic information could be obtained from the GCMC simulation. Such adsorption
isotherm can provide a comparison for the experimental results.
At the moment, GCMC simulation is a practical and sensible computational technique to
investigate the adsorption properties of single component CO2 or mixtures in MOFs
(Akkermans, Spenley & Robertson 2013; Satoh 2010).
3.5 Principle of molecular dynamics simulation
Up to date, the studies on diffusion properties of gust molecules in MOFs are quite limited
compared with the adsorption achievements. Among various simulation methods, MD
simulation, which is based on the Newton’s second law of motion, is feasible and effective to
investigate the structural and dynamic properties of adsorbate molecules in microporous
metal-organic frameworks (Parkes et al. 2014; Satoh 2010).
MD simulation is solving the equations of motion or Newton’s second law. After force
determination on each atom, the acceleration of each atom in the system is possible to
calculate. And then, integration of the equations of motion could generate a trajectory of the
atom positions, velocities and other information with time (Nose & Klein 1983). See Eq 2.
𝜕𝑈(𝒓1 ,𝒓2 ,⋯,𝒓𝑛 )
𝑚𝑖 𝒓̈ 𝑖 = − = 𝑭𝑖 (2)
𝜕𝑟𝑖
Where 𝑭𝑖 is the force on atom i, 𝑚𝑖 is the mass of atom i, and U is the potential energy of the
system.
Therefore, given the initial coordinates and velocities and other dynamic information at time
t, the updated coordinates and velocities at a time 𝑡 + 𝛥𝑡 can be calculated and determined. In
this respect, the Verlet algorithm uses position 𝒓𝒊 and force 𝑭𝒊 at time 𝑡 and the position from
time 𝑡 − 𝛥𝑡 to calculate new position 𝒓𝒊 at time 𝑡 + 𝛥𝑡 and obtain the
velocity 𝑽𝒊 (Grubmuller et al. 1991). See Eq 3 and 4,
𝑭𝒊 (𝛥𝑡)
𝒓𝒊 (𝑡 + 𝛥𝑡) = −𝒓𝒊 (𝑡 − 𝛥𝑡) + 2𝒓𝒊 (𝑡) + (𝛥𝑡)2 + 𝑂((𝛥𝑡)3 ) (3)
𝑚𝑖
𝑑𝒓𝒊 𝒓𝒊 (𝑡+𝛥𝑡)−𝒓𝒊 (𝑡−𝛥𝑡)
𝑽𝒊 (𝑡) = = + 𝑂((𝛥𝑡)2 ) (4)
𝑑𝑡 2𝛥𝑡
20
Note that the time step 𝛥𝑡 depends on the integration method as well as the system itself.
In summary, MD simulation is a continuous trajectory describing displacement of all
molecules and atoms. The principle of MD is simple, in each MD simulation step, the initial
position r and velocity v of particle are defined, and then all the molecules’ equations of
motion are solved numerically after the force field F is calculated. Consequently, the position
and velocity are updated to generate a dynamic view of the system evolution (Andersen 1980;
Ciccotti, Ferrario & Schuette 2014; Erpenbeck & Wood 1982; Haile 1992; Zeng, Yu & Lu
2010), as shown in Figure 6.
.
Figure 6. Schematic diagram of a basic MD code(Zeng, Yu & Lu 2010).
3.6 Molecular dynamics analysis methods
Very recently, MD simulation has been extensively deployed to explore and determine the
diffusion and dynamic properties including self-diffusivity, transport diffusivity and corrected
diffusivity of CO2 in MOFs. The majority of the studies of molecules diffusion are obtained
21
by single MD or an experimental combined MD method (Atci, Erucar & Keskin 2011;
Krishna & Van Baten 2006; Li, J-R, Kuppler & Zhou 2009; Parkes et al. 2014; Salles,
Fabrice et al. 2013; Skoulidas & Sholl 2005).
3.6.1 Mean square displacement
Mean square displacement (MSD) analysis is a technique which is critical in characterizing

the diffusion behavior of adsorbed molecules in the system. It determines the particles
average displacement in a molecular dynamics simulation as shown in Figure 7 (Kärger
1992). In addition, self-diffusion constant (Ds) can be obtained by using the Einstein relation
which is take the slope of MSD over time (Chandler 1975; Michalet 2010).
300
250
Mean Square Displacement (Ds)
Pressure 0.1bar
200
150
100
50
0
0 100 200 300 400 500
Time(Ps)
Figure 7. Mean square displacement.
3.6.2 Radial distribution function
Besides, MD simulation provides a direct image to examine particles’ packing or ordering

probability in the system by the radial distribution function (RDF) analysis. RDF is a pair
correlation function which describes how atoms or particles in a system are radically packed
around each other(Nijboer & Van Hove 1952). It could provide an effective way to describe
22
the overall structure properties of disordered molecular systems as shown in Figure
8(Waisman 1973). RDF can be expressed by the following equation (Eq 5)
𝑔(𝑟) = 𝑛(𝑟)/(𝜌4𝜋𝑟2d𝑟) (5)
Where 𝑔(𝑟) is the RDF; 𝑛(𝑟) is the mean number of atoms in a shell of width d𝑟 at
distance 𝑟, 𝜌 is the mean atom density.
Moreover, the RDF results could provide a fine supplement with X-ray or neutron diffraction
(Bentley et al. 1990; Filipponi 1994; Li, F & Lannin 1990). Thus, the binding sites of CO2
molecules can be explored by RDF method after MD simulation.
Figure 8. Radial distribution function.
3.6.3 Adsorption strength
In addition, the adsorption energy (Ep) between MOFs and CO2 is quite difficult to assess
both experimentally and numerically. However, it may be achieved in MD simulation by
calculating the average value of CO2 adsorbed MOFs equilibrium frames, i.e. 6 frames from
relatively stable 400 to 500 ps. Thus, the adsorption energy (Ep) could be calculated by the
Eq 6,
23
Ep=Etotal-EMOF-ECO2 (6)
Where Etotal, EMOF and ECO2 are the energy of CO2 adsorbed MOF extracted MOF and
extracted CO2 molecules, respectively, as shown in Figure 9. By exacting the CO2 adsorbed
MOFs equilibrium frames, the total potential energy Etotal can be calculated by the Forcite in
Materials Studio. After removing the all of the adsorbed CO2 in the structure, we are able to
achieve the potential energy of the MOF at the equilibrium stage. We can also accomplish the
potential energy calculation of the CO2 molecules by deleting the MOF structure in the
frame, thus, the interaction strength (adsorption energy) could be measured by the Eq 6 to
understand the adsorption strength between CO2 and MOF under specific pressure.
Figure 9. The snapshot of a CO2 adsorbed equilibrium MOF structure: CO2 adsorbed
MOF (upper), isolated MOF (lower left) and isolated adsorbed CO2 molecules (lower
right).
24
3.7 Summary
To summarize, this section highlights the main research approaches including the GCMC and
MD simulations. Prior to the simulations, the input parameters such as atomic partial charges
and force field are important which should be defined appropriately. With the intermolecular
and intramolecular interactions are appropriately described by the force fields, the grand
canonical Monte Carlo (GCMC) simulation is often implemented to predict the gas
adsorption capacity in a periodic simulation box ,in which the temperature T, the volume V
and chemical potentials are kept constant, only the number of adsorbed molecules N are
allowed to fluctuate by randomly creating, rotating , deleting as well as translating/displacing
in the system, and then the average amount of adsorbed molecules is statistically calculated
when the system reached equilibrium stage. At the moment, GCMC simulation is a practical
and popular computational technique to investigate the adsorption properties of single
component CO2 or mixtures in MOFs. Moreover, MD simulation, based on the Newton’s
equitation of motion, is becoming increasingly popular to investigate the kinetic and transport
properties of the adsorbate in microporous materials, especially in MOFs. The RDF and
MSD analysis method will be adopted to explore the CO2 binding locations and its self-
diffusion within MOFs. Lastly, the energy strength between CO2 and MOFs is calculated to
reveal the related relationship.
25
CHAPTER 4 CARBON DIOXIDE ADSORPTION IN HKUST-1
4.1 Model configuration
HKUST-1, also known as Cu-BTC, is one of the most extensively investigated MOFs for gas
adsorption and separation after it was first synthesized by Chui and co-workers in 1999(Chui
et al. 1999).
The model of HKSUT-1 is employed from the Crystallography Open Database
(No.2300380). This structure of HKUST-1 was reported by Yakevenko and co-workers in
2013 (Yakovenko et al. 2013).
HKUST-1 consists of paddlewheel-coordinated copper clusters and surrounded by 1, 3, 5-
benzenetricarboxylate (BTC) organic linkers. Each metal node is composed of two copper
atoms connected to the oxygens of four BTC ligands (see Figure 10a). In order to construct
the fluorinated model (HKUST-1F), all of the hydrogen atoms on BTC organic linkers are
manually replaced by fluoride elements (Figure 10b) by using the module of Materials Studio
Visualizer.
(a)
26
(b)
Figure 10 Unit cells of HKUST-1 (a) and HKUST-1F (b), Colour: Copper in orange,
oxygens in red, carbons in grey, hydrogen in white and fluorides are light blue.
4.2 Grand Canonical Monte Carlo Simulation details
As described above, the atomic partial charges and force field are critical for molecular
simulations. For this reason, all of the charges of HKSUT-1 and HKUST-1F were initially
reset to be zero, and then the atomic partial charges in frameworks were obtained by QEq
method (see the details in Table 1), and maintained in the following simulations.
Prior to the GCMC simulation, the energy calculation and geometry optimization were
applied on MOFs in order to obtain an optimal configuration. The force field of UFF was
performed on HKUST-1 and HKUST-1F, Ewald technique was utilized to calculate the
electrostatic interactions. Atom based method with a cut-off distance of 15.5 Å was used in
the calculations of van der Waals (vdW) interactions.
Given the quadrupole movement of carbon dioxide, CO2 was modelled as a linear molecule
with double C=O, and the atomic partial charge of O and C atom was calculated by QEq
method which were -0.446e and 0.892e, respectively, when implemented the energy and
geometry optimization on CO2, the UFF was determined and Atom based method was used
for both vdW interactions electrostatic interactions calculation.
27
Table 1. The Atomic partial charge q (e) and force field of HKUST-1 and HKUST-1F.
Element Atomic partial Force field type Hybridization
charge
Cu 2.388 Cu3+1 Trigonal-bipyramid
O -0.702 O_3 Trigonal
H 0.052 H_ No hybridization
Cu* 3.001 Cu3+1 Trigonal-bipyramid
O* -0.715 O_3 Trigonal
F* -0.8 F_ No hybridization
* is the element in HKUST-1F
Unlike the most simulation studies on HKUST-1 published before to keep the MOFs and CO2
as rigid, in present work, the MOFs and CO2 are allowed to be flexible during the GCMC
simulations. The periodic boundary condition was applied. For HKUST-1, since the cell
parameters were a=b=c=26.3 Å. we employed the unit cell as the simulation box, GCMC
method was defined as normal Metropolis technique, the adsorption isotherms of HKUST-1
and the HKUST-1F were performed at 298 K by Sorption modules in Materials Studio with
the pressure from 0.1 to 5 bars, and equilibration process of 1 × 106 steps followed by 1 × 107
production process steps with a cut-off radius of 18.5 Å. The Ewald technique was used to
handle the electrostatic interactions.
4.2.1 Adsorption isotherms of CO2 in HKUST-1 and HKUST-1F
The GCMC adsorption isotherm is a significant feature to access MOF’s property and
performance. Adsorption isotherms of CO2 in HKUST-1 and HKUST-1F are shown in Figure
11, which demonstrate that the adsorption capacity of HKUST-1F is much higher than that of
non-fluorinated HKUST-1. For example, the adsorbed CO2 in HKUST-1F unit cell is 176 at
0.1 bar compared with 62 in HKUST-1unit cell. At 1 bar, HKUST-1F adsorbs 194 carbon
dioxide molecules, which is almost the double of the HKUST-1. In addition, the uptake of
CO2 in HKUST-1 increases significantly with the pressure from 62 at 0.1 bar to 136 at 5 bars.
28
Yet, HKUST-1F reaches the saturated adsorption equilibrium at a relatively low pressure
(approximately between 0.5-1bar).
Figure 11. Adsorption isotherm of CO2 in HKUST-1 and HKUST-1F at 298K under
the pressure of from 0.1 to 5bar.
4.2.2 Comparison of CO2 adsorption isotherms of CO2 in

HKUST-1
The comparison of CO2 adsorption isotherms in HKUST-1 under different conditions

(experiment and simulation) are shown in Table 2. The results show that both HKUST-1 and
HKUST-1F in this work have a superior adsorption capacity. In particular, fluorinated
HKUST-1 has a much higher uptake than that of non-fluorinated structure, which may be
attributed to the strong electrostatic interactions between fluoride and CO2.
Table 2. CO2 adsorption capacity on HKUST-1(Cu-BTC) under different conditions.

MOFs Pressure (Bar) Temperature (K) Adsorption Adsorption
capacity (mmol/g) capacity (wt %)
HKUST-1 1 298 10.5 46.3
HKUST-1 5 298 14 61.7
HKUST-1F 1 298 17.2 75.5
29
MOFs Pressure (Bar) Temperature (K) Adsorption Adsorption
capacity (mmol/g) capacity (wt %)
HKUST-1F 5 298 17.3 76.3
HKUST-1a 10 327 7.19 31.6
HKUST-1b 15 298 12.7 -
HKUST-1c 1 273 11.6 51

(Ethanol+NH4Cl)
Cu-BTCd 10 298 11.9 52.4
a (Ye et al. 2013),b (Liang, Marshall & Chaffee 2009),c (Yan, X et al. 2014), (Zhao et al.
2011).
4.3 Molecular dynamics simulation results
4.3.1 Adsorption sites of CO2 in HKUST-1 and HKUST-1F
The MD simulation not only gives an adsorption prediction of CO2 in MOFs but also provide
an image of CO2 probability in MOFs, as shown in Figure 12 and 13 for HKUST-1 and
HKUST-1F. The results indicate that adsorbed CO2 in HKUST-1 at low pressure (1bar) are
more close to paddle-wheel-coordinated Cu sites, and the additional adsorbed CO2 are more
likely to appear at the central channel at high pressure (5 bar). For HKUST-1F, there is no
remarkable change for CO2 adsorbed in MOF from low to high pressure.
Figure 12. The CO2 (red spots) adsorption snapshot in HKUST-1, 1 bar (left) and 5
bar (right).
30
Figure 13. The CO2 (red spots) adsorption snapshot in HKUST-1F, 0.1 bar (left) and 5
bar (right).
In order to verify the more accurate and precise binding sites of CO2 in MOFs, radial
distribution function is implemented for analysis after MD simulation. The peaks indicate the
most possible locations of an atom from the reference atom at a specific radius. In present
study, RDF results of HKUST-1 and CO2 under different pressures are shown in Figure 14.
Due to the distinct orientations of oxygen atoms in CO2 molecule, the central carbon atom
was selected to represent the entire CO2, thus the atomic pairs are determined to be Cu
(MOF)-C (CO2), O (MOF) - C (CO2), and H (MOF) - C (CO2).
Figure 14a displays the shortest distance (first peak) is around 2.91 Å between H of HKUST-
1 and C of CO2, revealing the orientation of CO2 with C atoms toward H atoms of HKUST-1.
The next peak is located at 3.49 Å for O atoms of HKUST-1 and C atoms of CO2, followed
by a sharp and narrow peak at 3.93 Å for Cu atoms of HKUST-1 and C atoms of CO2. We
could predict that the CO2 molecules are more close to H atom (BTC linker) of HKSUT-1 at
the pressure of 0.1 bar.
With the pressure rise (Figure 14b, 3c, 3d and 3e), we found that the first peak of Cu- C pair
is still situated at around 3.9 Å, and the peak of the O-C pair is at about 3.49 Å. However, the
peak of H-C pair becomes flat and broad, indicating majority of additional CO2 adsorbed
move toward Cu sites, which may be attributed by the open metal site effect.
31
32
33
Figure 14. Radial distribution function of atomic pairs from HKUST-1 and CO2 at
298 K and different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar,
respectively.
Similar RDF analyses are shown for HKUST-1F in Figure 15. The atomic pairs are Cu
(MOF)-C (CO2), O (MOF)-C (CO2) and F-C (CO2). RDF results of HKUST-1F indicate that
there is a stronger interaction is between F atoms in MOF and the C atoms of CO2, reflected
from the drastic and sharp first peak at 2.75 Å. The peaks are located at the same distance
with the pressure rising to 5 bar, but peaks are becoming slightly wide for F-C atomic pair,
demonstrating CO2 molecules are bound tightly to F atoms. The peak of Cu-C atomic pair has
a slight decrease due to the amount of the carbon dioxide adsorbed increases and CO2 are
more close to F sites.
34
35
36
Figure 15. Radial distribution function of atomic pairs from HKUST-1F and CO2 at
298 K and different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar,
respectively.
4.3.2 Self-diffusivity of CO2 in HKUST-1 and HKUST-1F
To further probe the diffusion property of adsorbed CO2 in MOFs, the mean square
displacement (MSD) method has been performed to probe the diffusivity of CO2 in MOFs.
We take advantage of the carbon atom representing the CO2 as an approximation. Figure 16
shows the self-diffusivity of CO2 in HKUST-1 and HKUST-1F by taking the slope of the
MSD curves over time from 0 to 400 ps. The results indicate the self-diffusion of CO2
molecules in HKUST-1 and HKUST-1F declines with pressure increase. This is an expected
result and can be explained by the increase in adsorbed CO2 and then the steric hindrance of
movement for adsorbed CO2 molecules within MOFs’ structures. Note that the diffusivity
(MSD slope) of CO2 in HKUST-1F is much lower than that of non-fluorinated HKUST-1,
suggesting adsorbed CO2 is attached firmly.
37
Figure 16. Mean square displacement over time of CO2 in HKUST-1and HKUST-1F.
4.3.3 Interaction strength between CO2 and HKUST-1
In order to obtain insight of the interaction strength between CO2 and MOFs, we extracted the
structures of MOFs and adsorbed CO2 molecules, respectively, and then calculated the
interaction energy (Ep). The results as shown in Table 3 and Table 4 indicate that interaction
strength between MOFs and CO2 increases with pressure (i.e. CO2 loading). Moreover, the
interaction is attractive in nature of CO2 due to the negative potential energy. The Ep of
HKUST-1F is higher than HKUST-1, reflected to the increased loading number of CO2
(adsorption capacity). The potential energy of extracted CO2 in HKUST-1 reduced
significantly from positive (repulsion) to negative (attraction), suggesting the distance
between CO2 increases (i.e. loosely adsorbed). For HKUST-1F, potential energy of CO2 is
relatively low, which indicates CO2 molecules are stable and could hardly move.
38
Table 3. Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1.
Pressure(bar) Loading(number Total Energy of MOF Energy of CO2 Ep
per cell) Energy
0.1 62 7974.9 8505.1 49.1 -579.3
0.5 83 7863.9 8527.5 36.4 -700
1 99 7739.7 8520.3 13.8 -794.3
3 126 7557.9 8519.9 -29.1 -932.9
5 136 7488 8520.9 -54.2 -978.7
Table 4. Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1F.

per cell) Energy
0.1 176 -1469.3 919 7.94 -2396.6
0.5 189 -1526.5 956.3 3.1 -2486
1 194 -1538.6 975.2 2.13 -2515.9
3 195 -1558.6 961.9 2.08 -2522.9
5 198 -1549.2 994.9 2.09 -2546.3
4.4 Summary
In summary, results of the GCMC simulation demonstrate that HKUST-1 and HKUST-1F in
this study have an exceptional CO2 adsorption capacity up to 5bar. And the fluoride modified
HKUST-1 has a higher uptake than that of non-fluorinated structure. The powerful
electrostatic interactions between fluoride and CO2 contribute to the superior gas uptake. MD
simulation has been performed to investigate the CO2 adsorption location and binding
strength in HKUST-1 and HKUST-1F. The results demonstrate that HKUST-1F has a
superior adsorption capacity for CO2, particularly at low pressure. Moreover, adsorbed CO2
molecules are bound closely to fluoride atoms of BTC organic linkers in HKUST-1F,
however, may hinder their diffusion to copper metal sites.
39
CHAPTER 5 CARBON DIOXIDE ADSORPTION IN MIL-68(In)
5.1 Model configuration
MIL-n(M) materials, Materials of Institute Lavoisier, are mainly connected with trivalent
metal cations M3+ (e.g. Al3+, Cr3+, Fe3+, V3+, Ga3+ or In3+) and organic linkers of benzene
dicarboxylic acid which could form three-dimensional networks such as MIL-53(M), MIL-
68(M), and MIL-101(M), etc. Interestingly, MIL-53(M) and MIL-68(M) have the same
formula, same inorganic and organic units but show different framework features and
properties. MIL-53 series have been widely investigated for gas adsorption due to its unique
structure flexibility (Bourrelly et al. 2005; Hamon, Lomig et al. 2009; Lebedev et al. 2005;
Millange et al. 2008). The model of MIL-68(In) is adopted from the Crystallography Open
Database (No.4306785). MIL-68(In) is an Indium based metal-organic frameworks. The
MIL-68(In) is constructed with the chains of InO4 (OH) 2 octahedral units and interconnected
with the terephthalate ligands to form central triangle and hexagon channels (see Figure 17a).
In order to build up the fluorinated model which is MIL-68F (In), we utilized the Materials
Visualizer to do the modification, i.e. hydrogen atoms on terephthalate linkers are all
manually replaced by fluoride elements (Figure 17b) .
(a)
(b)
40
Figure 17. Unit cells of MIL-68 (a) and MIL-68F (In) (b), Colour: Indium in dark red,
oxygens in red, carbons in grey, hydrogen in white and fluorides are light blue.
5.2 Grand Canonical Monte Carlo Simulation details
Similarly, the atomic partial charges of MIL-68(In) and MIL-68F (In) were obtained by QEq
method and the UFF force filed was assigned for both MOFs as shown in in Table 5, these
parameters remain constant in the following simulations.
Table 5. The Atomic partial charge q (e) and force field of MIL-68(In) and MIL-68F (In).
Element Atomic partial Force field type Hybridization
charge
In 2.289 In3+3 Octahedral
O -0.519 O_3 Tetrahedral
H 0.1 H_ No hybridization
In* 2.697 In3+3 Octahedral
O* -0.498 O_3 Tetrahedral
F* -0.554 F_ No hybridization
* is the element in MIL-68F (In)
41
The energy calculation and geometry optimization were implemented on MIL-68(In) and
MIL-68F (In) in order to seek an optative configuration for the following simulations. In
these calculations, Ewald technique was used to calculate the electrostatic interactions. Atom
based method with a cut-off distance of 15.5 Å was used in the calculations of van der Waals
(vdW) interactions.
Likewise, the CO2 molecule was modelled as a linear molecule with double C=O, and the
atomic partial charge of O and C atom were -0.446e and 0.892e which are the same as the
HKUST-1 project. UFF was selected and Atom based method was used for both vdW
interactions and electrostatic interactions calculation for energy and geometry optimization
on CO2.
In present work, the MIL-68(In), MIL-68F (In) and CO2 are allowed to be flexible during the
GCMC simulations. For HKUST-1, we employed a single unit cell as the simulation box;
however, in this case, the simulation box amplified to 3 (1×1×3) of unit cells because the
original cell parameters were a=21.73 Å, b=37.68 Å and c=7.23 Å, respectively. Likewise,
GCMC method was defined as normal Metropolis technique; the adsorption isotherms of
MIL-68(In) and the Mil-68F (In) were performed at 298 K by Sorption modules in Materials
Studio by the pressure from 0.1 to 5 bars, and equilibration process of 1 × 106 steps followed
by 1 × 107 production process steps with a cut-off distance of 15.5 Å which is smaller than of
HKUST-1(18.5 Å). Atom based method was used to calculate vdW interactions and the
Ewald technique was used to compute the Columbic interactions.
5.2.1 Adsorption isotherms of CO2 in MIL-68(In) and MIL-68F

(In)
Figure 18 shows the adsorption isotherms of CO2 in MIL-68(In) and MIL-68F (In), indicating
that the fluorinated MIL-68F (In) performs well than MIL-68(In) for CO2 uptake. For
instance, at pressure of 0.1 bar, only 6 CO2 molecules are adsorbed in the MIL-68(In) unit
cell, compared with 67 in the MIL-68 F (In). With the pressure increases to 1 bar, the number
of CO2 in MIL-68(In) grows to 23, while the number of CO2 adsorbed in MIL-68F (In) is 92.
After the pressure rise to 5 bars, the MIL-68(In) adsorption capacity escalates to 57 while the
MIL-68(In) F elevates to 117. Given the tendency of adsorption curves, both MIL-68(In) and
MIL-68 F (In) may not have reached to the adsorption equilibrium at 5 bars. It is expected
42
that MIL-68 and its fluorinated structures would have an advantageous CO2 adsorption
capacity at higher pressures.
Figure 18. Adsorption isotherm of CO2 in MIL-68(In) and MIL-68F (In) at 298K
under the pressure of from 0.1 to 5 bar.
5.2.2 Comparison of CO2 adsorption isotherms of CO2 in

HKUST-1 and MIL-68(In)
The previous experimental studies reported the CO2 adsorption capacities in MIL-68(In) and
MIL-68(In)-NH2 were 4 and 8 wt. % at 1 atm and 298 K, respectively(Arstad et al. 2008),
compared with present simulation results were 2.9 and 7.81 wt.% for MIL-68(In) and MIL-
68F (In) at 1 bar and 298 K, suggesting that the simulation result of MIL-68 is
underestimated, but the results for amine and fluoride functionalized MIL-68 are more close.
The results of the adsorption isotherms in MOFs, as shown is Figure 19, show that both
HKUST-1 and HKUST-1F have a relatively high adsorption capacity. In addition, fluorinated
MOFs, HKUST-1F and MIL-68 F (In) perform well for CO2 adsorption than that of non-
fluorinated structures, which may be attributed to the strong electrostatic interactions between
fluoride atoms and CO2. Note that the unit cell of MIL-68 is triple of original cells.
43
Figure 19. Comparison of CO2 adsorption isotherms of CO2 in HKUST-1 and MIL-
68(In).
5.3 Molecular dynamics simulation results
5.3.1 Adsorption sites of CO2 in MIL-68(In) and MIL-68F (In)
The adsorption sites images of CO2 in MIL-68 (In) and MIL-68F (In) of MD simulations are
displayed in Figures 20 and 21. It shows that adsorbed CO2 molecules in MIL-68(In) at 0.1
bar have great possibility arising in the central triangle pore, however, CO2 are more likely to
emerge at hexagon channel at 5 bar, while, the appearance probability of CO2 in MIL-68F
(In) varies insignificantly.
44
Figure 20. The CO2 (red spots) adsorption snapshot of MIL-68(In), 0.1 bar (left) and 5
bar (right).
Figure 21. The CO2 (red spots) adsorption snapshot of MIL-68F (In), 0.1 bar (left)
and 5 bar (right).
A series of RDF results of MIL-68 and CO2 under different pressures are shown in Figure 22.
The atomic pairs are In (MOF)-C (CO2), O (MOF) - C (CO2), and H (MOF) - C (CO2).
At pressure of 0.1 bar, as shown in Figure 22a, the relatively strongest interaction (first peak)
is between O in MOF and the C of CO2, as located at 3.41 Å, and closely followed by the
peak of H (MOF) - C (CO2), which is located at 3.61 Å. The peak of In–C pair is located 4.41
Å, revealed as a relatively intense and sharp peak. Thus, the adsorbed CO2 molecules are
bound closely to O and H atoms in MIL-68(In), which is the triangle pore of the structure.
45
With pressure increasing to 0.5 bar, as shows in Figure 22b, the first peak appears at 3.15 Å,
which is O-C pair, the next peak, H-C pair, is located at 3.85 Å, followed by a strong peak
of In-C pair, which is located at 4.79 Å, indicating that additional adsorbed CO2 at 0.5 bar
are more close to O atoms. When the pressure rises to 1 bar (see Figure 22c), the H-C and O-
C atomic pairs are becoming wide and fewer intense, while, the peak of In-C remains sharp,
probably suggesting that additional adsorbed CO2 molecules become to emerge in hexagon
void of MIL-68(In) structure. With the pressure continues to creep up to 3 and 5 bars, the
CO2 molecules are more close to the H atoms, however, they reside in a more widespread
area, and majority of molecules may probably be adsorbed in the hexagon area, revealed as
three broad peaks in the Figures 22d and 22e.
46
47
Figure 22. The radial distribution function for atomic pairs of MIL-68 and CO2 at 298
K and different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar,
respectively.
48
In terms of MIL-68F (In), Figure 23 displays the RDF results of CO2 in this structure. The
atomic pairs are In (MOF)-C (CO2), O (MOF)-C (CO2) and F-C (CO2). The trend of these
curves is quite similar to that of HKUST-1F. The results demonstrate a forceful interaction
between F atoms of MOF and the C atoms of CO2, which corresponds to a drastic first peak
at 2.85 Å, and with the pressure rising to 5 bars; the radius of first peak which is F-C atomic
pair has no change. In addition, the G(r) of F-C peak reduces slightly because more CO2
molecules are binding closely to F atoms with the pressure increasing. The peaks of O-C and
In-C pairs are extensive and wide, indicating that CO2 are less likely to be bound with them.
49
50
Figure 23. Radial distribution function of atomic pairs from MIL-68F and CO2 at 298
K and different pressure: (a) 0.1 bar, (b) 0.5 bar, (c) 1 bar (d) 3 bar and (e) 5 bar,
respectively.
5.3.2 Self-diffusivity of CO2 in MIL-68(In) and MIL-68F (In)
Likewise, the MSD analysis technique was carried out to observe CO2 self-diffusivity in
MIL-68 (In) and MIL-68F (In), as shown in Figure 24.
Interestingly, MSD results unfold that self-diffusivity of CO2 in MIL-68F (In) is much lower,
demonstrating that CO2 are adsorbed strongly within the structure, which is quite similar to
HKUST-1F. For MIL-68(In), the diffusion of CO2 is slow at low pressure (0.1bar), with the
pressure increase to 1 bar, the diffusion of CO2 has a remarkable surge, and however, when
the pressure rise to 5 bar, the movement of CO2 molecules becomes difficult. Since the MIL-
n (M) materials have been extensively studied on their structure flexibility such as “gate-
opening” or “breathing” phenomenon, we may infer that the pore diameter in MIL-68 may
have a relatively significant increase at 1 bar, resulting in the high self-diffusivity coefficient
of CO2 molecules. However, we did not observe a noticeable rise in the GCMC adsorption
isotherm at 1 bar. Further analysis must be implemented to examine and probe on pore
diameters variation.
51
Figure 24. Mean squared displacement over time of CO2 in MIL-68(In) and MIL-68F
(In).
5.3.3 Interaction strength between CO2 and MIL-68(In)
Likewise, Tables 6 and 7 illustrate that interaction strength between MOFs and CO2 increases
with pressure/CO2 loading. The strength of fluorinated structure is much higher than non-
fluorinated one, corresponding to the uptake amount of CO2. Besides, the potential energy of
CO2 molecules in MIL-68 and MIL-68 (In) F increases notably from 0.1 to 3 bars,
demonstrating the average distance of adsorbed CO2 reduces, however, the CO2 potentials
decline with the pressure increasing to 5 bars, suggesting CO2 molecules are becoming loose
in the MOFs.
52
Table 6. Interaction potential energies (Kcal/mol) of CO2 adsorbed in MIL-68 (In).
per cell) Energy
0.1 6 -22305.1 -22258 5.72 -52.84
0.5 16 -22384.5 -22272.3 20.23 -132.41
1 23 -22442.4 -22230 34.85 -177.24
3 43 -22447.5 -22191.4 48.65 -304.77
5 57 -22378.8 -22022.5 32.89 -385.23
Table 7. Interaction potential energies (Kcal/mol) of CO2 adsorbed in MIL-68F (In).

per cell) Energy
0.1 67 -33261.6 -32664.4 -1.35 -595.86
0.5 84 -33256.3 -32623.7 34.95 -667.56
1 92 -33224.6 -32584 35.1 -695.78
3 107 -33074.7 -32394.8 67.31 -747.13
5 117 -2824.8 -32103.7 53.77 -774.85
5.4 Summary
In conclusion, the MD simulation results of MIL-68 (In) and its fluorinated structure reveal
that adsorption strength of MIL-68F (In) is much higher than the untreated MOF, which is in
line with the adsorption loading of GCMC results. In addition, the tight binding of CO2
molecules to fluoride functional organic ligands not only restricts the movement of the
molecules but also may have no effect on the structure flexibility of MIL-68 (In).
53
CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusions
In this thesis, a hybrid GCMC and MD simulation method is adopted to explore the CO2
adsorption performances on two typical and distinct MOFs, HKUST-1 and MIL-68(In).
Previous studies have paid numerous attentions on the single predominant factor or effect
such as open metal site, structure flexibility and group functionalization. In addition, the
kinetic and transport properties of CO2 in MOFs are less studied either.
In present work, the intense literature surveys have been done to understand the experimental
and computational accomplishments of CO2 adsorption behaviour on HKUST-1 and MIL-
68(In). It is evident that there is few work on the fluorinated MOFs for CO2 capture and
adsorption, especially on HKUST-1 and MIL-68(In). Moreover, the study on CO2 diffusion
properties is also scant compared with the tremendous programmes on the adsorption
capacity. In order to investigate the performances of fluorinated HKUST-1 and MIL-68(In)
for CO2 capture. The fluoride modification of structures of HKUST-1 and MIL-68(In) are
carried out by Material Studio Visualizer.
With the atomic partial charges and the force fields (UFF) being assigned, the GCMC
simulation is employed to predict CO2 adsorption isotherms in MOFs. The MD simulation is
implemented to explore the CO2 adsorption sites in MOFs and the CO2 self-diffusion during
the adsorption process. Some major findings have been achieved: (I) the GCMC results
exhibit that the both fluorinated MOFs have a significant improved CO2 adsorption capacity
than that of non-fluorinated structures, indicating that the fluoride modification treatment is
favourable to enhance CO2 adsorption, particular the HKUST-1. (II) MD simulation results
including RDF and MSD measurement of HKUST-1F demonstrate the interaction between
fluoride and CO2 is dominant at the low pressure which will restrict the diffusion of CO2
molecules, however, may be negative for the CO2 adsorption at the high pressure or CO2
desorption for adsorbent regeneration. (III) In the case of MIL-68, although MIL-68 and its
fluorinated structure do not perform as well as HKUST-1 for CO2 adsorption under the
pressure of 5 bars, it is expected to increase at higher pressures according to the adsorption
isotherm tendency. Besides, the functional group (fluoride modification) may not be effective
for the structural “open-up” of MIL-68 as there is no obviously significant CO2 adsorption
increase from 0.1 to 5 bar.
54
6.2 Recommendations on future work
The further studies can be considered from this work; firstly, GCMC simulations can be
implemented up to 20-30 bars to examine the CO2 adsorption on MIL-68(In) and its
fluorinated structure to explore the adsorption increase at a specific threshold pressure.
Secondly, the various fluoride modification sites on the organic linkers could be exploited, in
this case, all of the H atoms have been replaced by fluoride, however, there are still many
replacement options to compromise and enhance the CO2 adsorption capacity and self-
diffusivity in MOFs. For example, the number of the fluoride modification can be one or two
on the organic ligands and operational location can be 1.3, 1.5 or 3.5. Thirdly, with regard to
the refinement of force filed, although UFF is a good approximation to describe the
interactions in many MOFs and in this work, it may not be accurate for fluorinated HKUST-
1and MIL-68F (In) owing to the intense electrostatic field generated by the fluoride atoms.
55
REFERENCES
Abanades, JC, Rubin, ES & Anthony, EJ 2004, 'Sorbent cost and performance in CO2 capture systems',
Industrial & Engineering Chemistry Research, vol. 43, no. 13, pp. 3462-6.
Achten, WMJ, Almeida, J & Muys, B 2013, 'Carbon footprint of science: More than flying', Ecological
Indicators, vol. 34, pp. 352-5.
Adams, DJ 1975, 'GRAND CANONICAL ENSEMBLE MONTE-CARLO FOR A LENNARD-JONES FLUID',

Molecular physics, vol. 29, no. 1, pp. 307-11.
AghaKouchak, A, Cheng, LY, Mazdiyasni, O & Farahmand, A 2014, 'Global warming and changes in
risk of concurrent climate extremes: Insights from the 2014 California drought', Geophysical
Research Letters, vol. 41, no. 24, pp. 8847-52.
Akkermans, RL, Spenley, NA & Robertson, SH 2013, 'Monte Carlo methods in materials studio',
Molecular Simulation, vol. 39, no. 14-15, pp. 1153-64.
Alhamami, M, Doan, H & Cheng, C-H 2014, 'A Review on Breathing Behaviors of Metal-Organic-
Frameworks (MOFs) for Gas Adsorption', Materials, vol. 7, no. 4, p. 3198.
Andersen, HC 1980, 'Molecular dynamics simulations at constant pressure and/or temperature', The
Journal of Chemical Physics, vol. 72, no. 4, pp. 2384-93.
Arstad, B, Fjellvåg, H, Kongshaug, KO, Swang, O & Blom, R 2008, 'Amine functionalised metal organic
frameworks (MOFs) as adsorbents for carbon dioxide', Adsorption, vol. 14, no. 6, pp. 755-62.
Atci, E, Erucar, I & Keskin, S 2011, 'Adsorption and Transport of CH4, CO2, H-2 Mixtures in a Bio-MOF
Material from Molecular Simulations', Journal of Physical Chemistry C, vol. 115, no. 14, pp. 6833-40.
Azar, C, Lindgren, K, Obersteiner, M, Riahi, K, van Vuuren, DP, den Elzen, KMGJ, Möllersten, K &
Larson, ED 2010, 'The feasibility of low CO2 concentration targets and the role of bio-energy with
carbon capture and storage (BECCS)', Climatic Change, vol. 100, no. 1, pp. 195-202.
Babarao, R & Jiang, J 2008, 'Molecular screening of metal− organic frameworks for CO2 storage',
Langmuir, vol. 24, no. 12, pp. 6270-8.
Bains, P, Psarras, P & Wilcox, J 2017, 'CO2 capture from the industry sector', Progress in energy and
combustion science, vol. 63, pp. 146-72.
Banerjee, R 2012, 'Functionalized Metal Organic Frameworks (MOFs) for reversible gas storage and
sequestration applications', Journal of the Indian Chemical Society, vol. 89, no. 9, pp. 1197-202.
Barcza, S 2017, 'Greenhouse effect from the point of view of radiative transfer', Acta Geodaetica Et
Geophysica, vol. 52, no. 4, pp. 581-92.
Benson, SM, Bennaceur, K, Cook, P, Davison, J, Coninck, Hd, Farhat, K, Ramirez, CA, Simbeck, D,
Surles, T, Verma, P & Wright, I 2012, 'Carbon Capture and Storage', in Global Energy Assessment -
Toward a Sustainable Future, Cambridge University Press, pp. 993 - null.
56
Bentley, J, Angelini, P, Gove, AP, Sklad, PS & Fisher, AT 1990, 'RADIAL-DISTRIBUTION FUNCTIONS OF
AMORPHOUS MATERIALS FROM ONLINE MEASUREMENTS OF DIFFRACTED ELECTRON INTENSITIES',
Institute of Physics Conference Series, no. 98, pp. 107-10.
Berdonosova, EA, Maletskaya, NV, Kogan, EV, Kovalenko, KA, Klyamkin, SN, Dybtsev, DN & Fedin, VP
2013, 'Synthesis and gas sorption properties of halogen-doped mesoporous chromium(iii)
terephthalate', Russian Chemical Bulletin, vol. 62, no. 1, pp. 157-62.
Berendsen, HJC, Postma, JPM, Vangunsteren, WF, Dinola, A & Haak, JR 1984, 'MOLECULAR-
DYNAMICS WITH COUPLING TO AN EXTERNAL BATH', Journal of Chemical Physics, vol. 81, no. 8, pp.
3684-90.
Boot-Handford, ME, Abanades, JC, Anthony, EJ, Blunt, MJ, Brandani, S, Mac Dowell, N, Fernandez, JR,
Ferrari, M-C, Gross, R, Hallett, JP, Haszeldine, RS, Heptonstall, P, Lyngfelt, A, Makuch, Z, Mangano, E,
Porter, RTJ, Pourkashanian, M, Rochelle, GT, Shah, N, Yao, JG & Fennell, PS 2014, 'Carbon capture
and storage update', Energy & Environmental Science, vol. 7, no. 1, pp. 130-89.
Bordiga, S, Regli, L, Bonino, F, Groppo, E, Lamberti, C, Xiao, B, Wheatley, PS, Morris, RE & Zecchina, A
2007, 'Adsorption properties of HKUST-1 toward hydrogen and other small molecules monitored by
IR', Physical Chemistry Chemical Physics, vol. 9, no. 21, pp. 2676-85.
Bosoaga, A, Masek, O & Oakey, JE 2009, 'CO2 Capture Technologies for Cement Industry', in J Gale, H
Herzog & J Braitsch (eds), Greenhouse Gas Control Technologies 9, vol. 1, pp. 133-40.
Botas, JA, Calleja, G, Sanchez-Sanchez, M & Orcajo, MG 2010, 'Cobalt Doping of the MOF-5
Framework and Its Effect on Gas-Adsorption Properties', Langmuir, vol. 26, no. 8, pp. 5300-3.
Bourrelly, S, Llewellyn, PL, Serre, C, Millange, F, Loiseau, T & Férey, G 2005, 'Different adsorption
behaviors of methane and carbon dioxide in the isotypic nanoporous metal terephthalates MIL-53
and MIL-47', Journal of the American Chemical Society, vol. 127, no. 39, pp. 13519-21.
Canivet, J, Fateeva, A, Guo, Y, Coasne, B & Farrusseng, D 2014, 'Water adsorption in MOFs:
fundamentals and applications', Chemical Society Reviews, vol. 43, no. 16, pp. 5594-617.
Casewit, CJ, Colwell, KS & Rappe, AK 1992, 'APPLICATION OF A UNIVERSAL FORCE-FIELD TO

ORGANIC-MOLECULES', Journal of the American Chemical Society, vol. 114, no. 25, pp. 10035-46.
Chałupnik, S, Franus, W, Wysocka, M & Gzyl, G 2013, 'Application of zeolites for radium removal
from mine water', Environmental Science and Pollution Research, vol. 20, no. 11, pp. 7900-6.
Chandler, D 1975, 'Rough hard sphere theory of the self‐diffusion constant for molecular liquids',
The Journal of Chemical Physics, vol. 62, no. 4, pp. 1358-63.
Chen, L, Reiss, PS, Chong, SY, Holden, D, Jelfs, KE, Hasell, T, Little, MA, Kewley, A, Briggs, ME,
Stephenson, A, Thomas, KM, Armstrong, JA, Bell, J, Busto, J, Noel, R, Liu, J, Strachan, DM, Thallapally,
PK & Cooper, AI 2014, 'Separation of rare gases and chiral molecules by selective binding in porous
organic cages', Nature Materials, vol. 13, p. 954.
Cherp, A, Jewell, J, Vinichenko, V, Bauer, N & De Cian, E 2016, 'Global energy security under different
climate policies, GDP growth rates and fossil resource availabilities', Climatic Change, vol. 136, no. 1,
pp. 83-94.
57
Chrysostomidis, I, Zakkour, P, Bohm, M, Beynon, E, de Filippo, R & Lee, A 2009, 'Assessing issues of
financing a CO2 transportation pipeline infrastructure', Energy Procedia, vol. 1, no. 1, pp. 1625-32.
Chu, S 2009, 'Carbon Capture and Sequestration', Science, vol. 325, no. 5948, pp. 1599-.
Chui, SSY, Lo, SMF, Charmant, JPH, Orpen, AG & Williams, ID 1999, 'A chemically functionalizable
nanoporous material Cu-3(TMA)(2)(H2O)(3) (n)', Science, vol. 283, no. 5405, pp. 1148-50.
Ciccotti, G, Ferrario, M & Schuette, C 2014, 'Molecular Dynamics Simulation'.
Collins, SP & Woo, TK 2017, 'Split-Charge Equilibration Parameters for Generating Rapid Partial
Atomic Charges in Metal-Organic Frameworks and Porous Polymer Networks for High-Throughput
Screening', Journal of Physical Chemistry C, vol. 121, no. 1, pp. 903-10.
Cormos, C-C 2012, 'Integrated assessment of IGCC power generation technology with carbon
capture and storage (CCS)', Energy, vol. 42, no. 1, pp. 434-45.
Cormos, CC 2014, 'Economic evaluations of coal-based combustion and gasification power plants
with post-combustion CO2 capture using calcium looping cycle', Energy, vol. 78, pp. 665-73.
D'Alessandro, DM, Smit, B & Long, JR 2010, 'Carbon dioxide capture: prospects for new materials',
Angewandte Chemie International Edition, vol. 49, no. 35, pp. 6058-82.
de Coninck, H, Stephens, JC & Metz, B 2009, 'Global learning on carbon capture and storage: A call
for strong international cooperation on CCS demonstration', Energy Policy, vol. 37, no. 6, pp. 2161-5.
de Pablo, JJ, Jones, B, Kovacs, CL, Ozolins, V & Ramirez, AP 2014, 'The Materials Genome Initiative,
the interplay of experiment, theory and computation', Current Opinion in Solid State and Materials
Science, vol. 18, no. 2, pp. 99-117.
Din, XD & Michaelides, EE 1997, 'Kinetic theory and molecular dynamics simulations of microscopic
flows', Physics of Fluids, vol. 9, no. 12, pp. 3915-25.
Drage, TC, Snape, CE, Stevens, LA, Wood, J, Wang, J, Cooper, AI, Dawson, R, Guo, X, Satterley, C &
Irons, R 2012, 'Materials challenges for the development of solid sorbents for post-combustion
carbon capture', Journal of Materials Chemistry, vol. 22, no. 7, pp. 2815-23.
Dreizler, RM & Gross, EKU 1990, 'Introduction', in Density Functional Theory: An Approach to the
Quantum Many-Body Problem, Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 1-3.
Duren, T, Bae, Y-S & Snurr, RQ 2009, 'Using molecular simulation to characterise metal-organic
frameworks for adsorption applications', Chemical Society Reviews, vol. 38, no. 5, pp. 1237-47.
Eldevik, F, Graver, B, Torbergsen, LE & Saugerud, OT 2009, 'Development of a Guideline for Safe,
Reliable and Cost Efficient Transmission of CO2 in Pipelines', Energy Procedia, vol. 1, no. 1, pp. 1579-
85.
Erpenbeck, JJ & Wood, WW 1982, 'Molecular-dynamics calculations of the velocity-autocorrelation

function. Methods, hard-disk results', Physical Review A, vol. 26, no. 3, p. 1648.
58
Evins, R 2013, 'A review of computational optimisation methods applied to sustainable building
design', Renewable and Sustainable Energy Reviews, vol. 22, no. Supplement C, pp. 230-45.
Fang, SX, Tans, PP, Steinbacher, M, Zhou, LX & Luan, T 2015, 'Comparison of the regional CO2 mole
fraction filtering approaches at a WMO/GAW regional station in China', Atmospheric Measurement
Techniques, vol. 8, no. 12, pp. 5301-13.
Fang, SX, Zhou, LX, Tans, PP, Ciais, P, Steinbacher, M, Xu, L & Luan, T 2014, 'In situ measurement of
atmospheric CO2 at the four WMO/GAW stations in China', Atmospheric Chemistry and Physics, vol.
14, no. 5, pp. 2541-54.
Feijani, EA, Tavasoli, A & Mahdavi, H 2015, 'Improving Gas Separation Performance of
Poly(vinylidene fluoride) Based Mixed Matrix Membranes Containing Metal-Organic Frameworks by
Chemical Modification', Industrial & Engineering Chemistry Research, vol. 54, no. 48, pp. 12124-34.
Filipponi, A 1994, 'The radial distribution function probed by X-ray absorption spectroscopy', Journal
of Physics: Condensed Matter, vol. 6, no. 41, p. 8415.
Frenkel, D & Smit, B 2002, 'Understanding Molecular Simulation. Computational Science Series',
Academic Press, San Diego Adcock SA, McCammon JA (2006) Molecular dynamics: survey of methods
for simulating the activity of proteins. Chem Rev, vol. 106, pp. 1589-615.
Furukawa, H, Cordova, KE, O’Keeffe, M & Yaghi, OM 2013, 'The Chemistry and Applications of Metal-
Organic Frameworks', Science, vol. 341, no. 6149.
Furukawa, H, Ko, N, Go, YB, Aratani, N, Choi, SB, Choi, E, Yazaydin, AÖ, Snurr, RQ, O’Keeffe, M, Kim, J
& Yaghi, OM 2010, 'Ultrahigh Porosity in Metal-Organic Frameworks', Science, vol. 329, no. 5990, pp.
424-8.
Garcia-Abuin, A, Gomez-Diaz, D & Navaza, JM 2014, 'New processes for amine regeneration', Fuel,
vol. 135, pp. 191-7.
Ghosh, P, Colon, YJ & Snurr, RQ 2014, 'Water adsorption in UiO-66: the importance of defects',
Chemical Communications, vol. 50, no. 77, pp. 11329-31.
Gibbins, J & Chalmers, H 2008, 'Carbon capture and storage', Energy Policy, vol. 36, no. 12, pp. 4317-
22.
Grubmuller, H, Heller, H, Windemuth, A & Schulten, K 1991, 'GENERALIZED VERLET ALGORITHM FOR
EFFICIENT MOLECULAR DYNAMICS SIMULATIONS WITH LONG-RANGE INTERACTIONS', Molecular
Simulation, vol. 6, no. 1-3, pp. 121-42.
Hafner, J 2007, 'Materials simulations using VASP—a quantum perspective to materials science',
Computer Physics Communications, vol. 177, no. 1, pp. 6-13.
Haile, J 1992, Molecular dynamics simulation, vol. 18, Wiley, New York.
Hamad, S, Balestra, SRG, Bueno-Perez, R, Calero, S & Ruiz-Salvador, AR 2015, 'Atomic charges for
modeling metal-organic frameworks: Why and how', Journal of Solid State Chemistry, vol. 223, pp.
144-51.
59
Hamon, L, Llewellyn, PL, Devic, T, Ghoufi, A, Clet, G, Guillerm, V, Pirngruber, GD, Maurin, G, Serre, C,
Driver, G, van Beek, W, Jolimaitre, E, Vimont, A, Daturi, M & Ferey, G 2009, 'Co-adsorption and
Separation of CO2-CH4 Mixtures in the Highly Flexible MIL-53(Cr) MOF', Journal of the American
Chemical Society, vol. 131, no. 47, pp. 17490-9.
Hamon, L, Serre, C, Devic, T, Loiseau, T, Millange, F, Férey, G & Weireld, GD 2009, 'Comparative
study of hydrogen sulfide adsorption in the MIL-53 (Al, Cr, Fe), MIL-47 (V), MIL-100 (Cr), and MIL-101
(Cr) metal− organic frameworks at room temperature', Journal of the American Chemical Society,
vol. 131, no. 25, pp. 8775-7.
Handy, NC 1996, 'Density Functional Theory', in D Bicout & M Field (eds), Quantum Mechanical
Simulation Methods for Studying Biological Systems: Les Houches Workshop, May 2–7, 1995,
Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 1-35.
Hao, G-P, Li, W-C & Lu, A-H 2011, 'Novel porous solids for carbon dioxide capture', Journal of
Materials Chemistry, vol. 21, no. 18, pp. 6447-51.
Hoover, WG 1985, 'CANONICAL DYNAMICS - EQUILIBRIUM PHASE-SPACE DISTRIBUTIONS', Physical

Review A, vol. 31, no. 3, pp. 1695-7.
Hu, X-L, Liu, F-H, Wang, H-N, Qin, C, Sun, C-Y, Su, Z-M & Liu, F-C 2014, 'Controllable synthesis of
isoreticular pillared-layer MOFs: gas adsorption, iodine sorption and sensing small molecules',
Journal of Materials Chemistry A, vol. 2, no. 36, pp. 14827-34.
Hussain, D, Dzombak, DA, Jaramillo, P & Lowry, GV 2013, 'Comparative lifecycle inventory (LCI) of
greenhouse gas (GHG) emissions of enhanced oil recovery (EOR) methods using different CO2
sources', International journal of greenhouse gas control, vol. 16, no. Supplement C, pp. 129-44.
Jackson, RB, Le Quere, C, Andrew, RM, Canadell, JG, Peters, GP, Roy, J & Wu, L 2017, 'Warning signs
for stabilizing global CO2 emissions', Environmental Research Letters, vol. 12, no. 11.
Jelfs, KE & Cooper, AI 2013, 'Molecular simulations to understand and to design porous organic
molecules', Current Opinion in Solid State and Materials Science, vol. 17, no. 1, pp. 19-30.
Jung, J-Y, Huh, C, Kang, S-G, Seo, Y & Chang, D 2013, 'CO2 transport strategy and its cost estimation
for the offshore CCS in Korea', Applied Energy, vol. 111, no. Supplement C, pp. 1054-60.
Kärger, J 1992, 'Straightforward derivation of the long-time limit of the mean-square displacement in
one-dimensional diffusion', Physical Review A, vol. 45, no. 6, p. 4173.
Keskin, S & Sholl, DS 2007, 'Screening metal-organic framework materials for membrane-based
methane/carbon dioxide separations', The Journal of Physical Chemistry C, vol. 111, no. 38, pp.
14055-9.
Kowalczyk, A, Harnisch, E, Schwede, S, Gerber, M & Span, R 2013, 'Different mixing modes for biogas
plants using energy crops', Applied Energy, vol. 112, no. Supplement C, pp. 465-72.
Krausmann, F, Gingrich, S, Eisenmenger, N, Erb, K-H, Haberl, H & Fischer-Kowalski, M 2009, 'Growth
in global materials use, GDP and population during the 20th century', Ecological Economics, vol. 68,
no. 10, pp. 2696-705.
60
Krishna, R & Van Baten, J 2006, 'Describing binary mixture diffusion in carbon nanotubes with the
Maxwell-Stefan equations. An investigation using molecular dynamics simulations', Industrial &
Engineering Chemistry Research, vol. 45, no. 6, pp. 2084-93.
Lamb, ML & Jorgensen, WL 1997, 'Computational approaches to molecular recognition', Current

Opinion in Chemical Biology, vol. 1, no. 4, pp. 449-57.
Lawler, KV, Hulvey, Z & Forster, PM 2015, 'On the importance of a precise crystal structure for
simulating gas adsorption in nanoporous materials', Physical Chemistry Chemical Physics, vol. 17, no.
29, pp. 18904-7.
Le Moullec, Y & Kanniche, M 2011, 'Screening of flowsheet modifications for an efficient

monoethanolamine (MEA) based post-combustion CO2 capture', International journal of greenhouse
gas control, vol. 5, no. 4, pp. 727-40.
Lebedev, O, Millange, F, Serre, C, Van Tendeloo, G & Férey, G 2005, 'First direct imaging of giant
pores of the metal-organic framework MIL-101', Chemistry of Materials, vol. 17, no. 26, pp. 6525-7.
Lee, S-Y & Park, S-J 2015, 'A review on solid adsorbents for carbon dioxide capture', Journal of
Industrial and Engineering Chemistry, vol. 23, no. Supplement C, pp. 1-11.
Lee, ZH, Sethupathi, S, Lee, KT, Bhatia, S & Mohamed, AR 2013, 'An overview on global warming in
Southeast Asia: CO2 emission status, efforts done, and barriers', Renewable & Sustainable Energy
Reviews, vol. 28, pp. 71-81.
Levin, K, Cashore, B, Bernstein, S & Auld, G 2012, 'Overcoming the tragedy of super wicked
problems: constraining our future selves to ameliorate global climate change', Policy Sciences, vol.
45, no. 2, pp. 123-52.
Li, F & Lannin, JS 1990, 'RADIAL-DISTRIBUTION FUNCTION OF AMORPHOUS-CARBON', Physical

review letters, vol. 65, no. 15, pp. 1905-8.
Li, J-R, Kuppler, RJ & Zhou, H-C 2009, 'Selective gas adsorption and separation in metal-organic
frameworks', Chemical Society Reviews, vol. 38, no. 5, pp. 1477-504.
Li, J-R, Ma, Y, McCarthy, MC, Sculley, J, Yu, J, Jeong, H-K, Balbuena, PB & Zhou, H-C 2011, 'Carbon
dioxide capture-related gas adsorption and separation in metal-organic frameworks', Coordination
Chemistry Reviews, vol. 255, no. 15, pp. 1791-823.
Li, W, Rao, ZZ, Chung, YG & Li, S 2017, 'The Role of Partial Atomic Charge Assignment Methods on
the Computational Screening of Metal-Organic Frameworks for CO2 Capture under Humid
Conditions', Chemistryselect, vol. 2, no. 29, pp. 9458-65.
Li, Y-W, Li, J-R, Wang, L-F, Zhou, B-Y, Chen, Q & Bu, X-H 2013, 'Microporous metal-organic
frameworks with open metal sites as sorbents for selective gas adsorption and fluorescence sensors
for metal ions', Journal of Materials Chemistry A, vol. 1, no. 3, pp. 495-9.
Li, Y-W, Xu, J, Li, D-C, Dou, J-M, Yan, H, Hu, T-L & Bu, X-H 2015, 'Two microporous MOFs constructed
from different metal cluster SBUs for selective gas adsorption', Chemical Communications, vol. 51,
no. 75, pp. 14211-4.
61
Li, Y-W, Yan, H, Hu, T-L, Ma, H-Y, Li, D-C, Wang, S-N, Yao, Q-X, Dou, J-M, Xu, J & Bu, X-H 2017, 'Two
microporous Fe-based MOFs with multiple active sites for selective gas adsorption', Chemical
Communications, vol. 53, no. 15, pp. 2394-7.
Liang, ZJ, Marshall, M & Chaffee, AL 2009, 'CO2 Adsorption-Based Separation by Metal Organic
Framework (Cu-BTC) versus Zeolite (13X)', Energy & fuels, vol. 23, no. 5-6, pp. 2785-9.
Liu, YY, Wang, ZYU & Zhou, HC 2012, 'Recent advances in carbon dioxide capture with metal-organic
frameworks', Greenhouse Gases-Science and Technology, vol. 2, no. 4, pp. 239-59.
Llewellyn, PL, Bourrelly, S, Serre, C, Vimont, A, Daturi, M, Hamon, L, De Weireld, G, Chang, JS, Hong,
DY, Hwang, YK, Jhung, SH & Ferey, G 2008, 'High uptakes of CO2 and CH4 in mesoporous metal-
organic frameworks MIL-100 and MIL-101', Langmuir, vol. 24, no. 14, pp. 7245-50.
Lucquiaud, M & Gibbins, J 2011, 'On the integration of CO2 capture with coal-fired power plants: A
methodology to assess and optimise solvent-based post-combustion capture systems', Chemical
Engineering Research & Design, vol. 89, no. 9, pp. 1553-71.
Luo, QX, An, BW, Ji, M, Park, SE, Hao, C & Li, YQ 2015, 'Metal-organic frameworks HKUST-1 as porous
matrix for encapsulation of basic ionic liquid catalyst: effect of chemical behaviour of ionic liquid in
solvent', Journal of Porous Materials, vol. 22, no. 1, pp. 247-59.
Lyndon, R, Konstas, K, Thornton, AW, Seeber, AJ, Ladewig, BP & Hill, MR 2015, 'Visible Light-
Triggered Capture and Release of CO2 from Stable Metal Organic Frameworks', Chemistry of
Materials, vol. 27, no. 23, pp. 7882-8.
Ma, L, Abney, C & Lin, W 2009, 'Enantioselective catalysis with homochiral metal-organic
frameworks', Chemical Society Reviews, vol. 38, no. 5, pp. 1248-56.
Maji, TK, Ohba, M & Kitagawa, S 2005, 'Transformation from a 2D stacked layer to 3D
interpenetrated framework by changing the spacer functionality: Synthesis, structure, adsorption,
and magnetic properties', Inorganic chemistry, vol. 44, no. 25, pp. 9225-31.
Makov, G & Payne, MC 1995, 'PERIODIC BOUNDARY-CONDITIONS IN AB-INITIO CALCULATIONS',

Physical Review B, vol. 51, no. 7, pp. 4014-22.
Manan, ZA, Nawi, W, Alwi, SRW & Klemes, JJ 2017, 'Advances in Process Integration research for CO2
emission reduction - A review', Journal of Cleaner Production, vol. 167, pp. 1-13.
Mangalapally, HP, Notz, R, Hoch, S, Asprion, N, Sieder, G, Garcia, H & Hasse, H 2009, 'Pilot plant
experimental studies of post combustion CO2 capture by reactive absorption with MEA and new
solvents', Energy Procedia, vol. 1, no. 1, pp. 963-70.
Martin, MG 2006, 'Comparison of the AMBER, CHARMM, COMPASS, GROMOS, OPLS, TraPPE and
UFF force fields for prediction of vapor–liquid coexistence curves and liquid densities', Fluid phase
equilibria, vol. 248, no. 1, pp. 50-5.
Maurin, G, Serre, C, Cooper, A & Ferey, G 2017, 'The new age of MOFs and of their porous-related
solids', Chemical Society Reviews, vol. 46, no. 11, pp. 3104-7.
62
Meier, K, Laesecke, A & Kabelac, S 2001, 'A Molecular Dynamics Simulation Study of the Self-
Diffusion Coefficient and Viscosity of the Lennard–Jones Fluid', International Journal of
Thermophysics, vol. 22, no. 1, pp. 161-73.
Metropolis, N, Rosenbluth, AW, Rosenbluth, MN, Teller, AH & Teller, E 1953, 'EQUATION OF STATE
CALCULATIONS BY FAST COMPUTING MACHINES', Journal of Chemical Physics, vol. 21, no. 6, pp.
1087-92.
Michalet, X 2010, 'Mean square displacement analysis of single-particle trajectories with localization
error: Brownian motion in an isotropic medium', Physical Review E, vol. 82, no. 4, p. 041914.
Millange, F, Guillou, N, Walton, RI, Greneche, J-M, Margiolaki, I & Ferey, G 2008, 'Effect of the nature
of the metal on the breathing steps in MOFs with dynamic frameworks', Chemical Communications,
no. 39, pp. 4732-4.
Miller, DR, Akbar, SA & Morris, PA 2014, 'Nanoscale metal oxide-based heterojunctions for gas
sensing: A review', Sensors and Actuators B: Chemical, vol. 204, no. Supplement C, pp. 250-72.
Millward, AR & Yaghi, OM 2005, 'Metal-organic frameworks with exceptionally high capacity for
storage of carbon dioxide at room temperature', Journal of the American Chemical Society, vol. 127,
no. 51, pp. 17998-9.
Mohor, GS & Mendiondo, EM 2017, 'Economic indicators of hydrologic drought insurance under
water demand and climate change scenarios in a Brazilian context', Ecological Economics, vol. 140,
pp. 66-78.
Mosher, K, He, J, Liu, Y, Rupp, E & Wilcox, J 2013, 'Molecular simulation of methane adsorption in
micro- and mesoporous carbons with applications to coal and gas shale systems', International
Journal of Coal Geology, vol. 109-110, no. Supplement C, pp. 36-44.
Nijboer, B & Van Hove, L 1952, 'Radial distribution function of a gas of hard spheres and the
superposition approximation', Physical Review, vol. 85, no. 5, p. 777.
Nosé, S 1984, 'A molecular dynamics method for simulations in the canonical ensemble', Molecular
physics, vol. 52, no. 2, pp. 255-68.
Nose, S & Klein, ML 1983, 'CONSTANT PRESSURE MOLECULAR-DYNAMICS FOR MOLECULAR-

SYSTEMS', Molecular physics, vol. 50, no. 5, pp. 1055-76.
Nugent, P, Belmabkhout, Y, Burd, SD, Cairns, AJ, Luebke, R, Forrest, K, Pham, T, Ma, S, Space, B &
Wojtas, L 2013, 'Porous materials with optimal adsorption thermodynamics and kinetics for CO2
separation', Nature, vol. 495, no. 7439, pp. 80-4.
Orio, M, Pantazis, DA & Neese, F 2009, 'Density functional theory', Photosynthesis Research, vol.
102, no. 2, pp. 443-53.
Park, J, Lively, RP & Sholl, DS 2017, 'Establishing upper bounds on CO2 swing capacity in sub-ambient
pressure swing adsorption via molecular simulation of metal-organic frameworks', Journal of
Materials Chemistry A, vol. 5, no. 24, pp. 12258-65.
63
Parkes, MV, Demir, H, Teich-McGoldrick, SL, Sholl, DS, Greathouse, JA & Allendorf, MD 2014,
'Molecular dynamics simulation of framework flexibility effects on noble gas diffusion in HKUST-1
and ZIF-8', Microporous and Mesoporous Materials, vol. 194, no. Supplement C, pp. 190-9.
Pietzcker, RC, Longden, T, Chen, W, Fu, S, Kriegler, E, Kyle, P & Luderer, G 2014, 'Long-term transport
energy demand and climate policy: Alternative visions on transport decarbonization in energy-
economy models', Energy, vol. 64, no. Supplement C, pp. 95-108.
Pinto, DDD, Knuutila, H, Fytianos, G, Haugen, G, Mejdell, T & Svendsen, HF 2014, 'CO2 post
combustion capture with a phase change solvent. Pilot plant campaign', International journal of
greenhouse gas control, vol. 31, pp. 153-64.
Pivkin, IV & Karniadakis, GE 2005, 'A new method to impose no-slip boundary conditions in
dissipative particle dynamics', Journal of Computational Physics, vol. 207, no. 1, pp. 114-28.
Pongsajanukul, P, Parasuk, V, Fritzsche, S, Assabumrungrat, S, Wongsakulphasatch, S,

Bovornratanaraks, T & Chokbunpiam, T 2017, 'Theoretical study of carbon dioxide adsorption and
diffusion in MIL-127 (Fe) metal organic framework', Chemical Physics, vol. 491, pp. 118-25.
Quadrelli, R & Peterson, S 2007, 'The energy–climate challenge: Recent trends in CO2 emissions from
fuel combustion', Energy Policy, vol. 35, no. 11, pp. 5938-52.
Randolph, JB & Saar, MO 2011, 'Coupling carbon dioxide sequestration with geothermal energy
capture in naturally permeable, porous geologic formations: Implications for CO2 sequestration',
Energy Procedia, vol. 4, no. Supplement C, pp. 2206-13.
Rao, AB & Rubin, ES 2002, 'A technical, economic, and environmental assessment of amine-based
CO2 capture technology for power plant greenhouse gas control', Environmental science &
technology, vol. 36, no. 20, pp. 4467-75.
Rappe, AK, Casewit, CJ, Colwell, KS, Goddard, WA & Skiff, WM 1992, 'UFF, A FULL PERIODIC-TABLE
FORCE-FIELD FOR MOLECULAR MECHANICS AND MOLECULAR-DYNAMICS SIMULATIONS', Journal of
the American Chemical Society, vol. 114, no. 25, pp. 10024-35.
Rappe, AK & Goddard, WA 1991, 'CHARGE EQUILIBRATION FOR MOLECULAR-DYNAMICS

SIMULATIONS', Journal of Physical Chemistry, vol. 95, no. 8, pp. 3358-63.
Rochelle, G, Chen, E, Freeman, S, Van Wagener, D, Xu, Q & Voice, A 2011, 'Aqueous piperazine as the
new standard for CO2 capture technology', Chemical engineering journal, vol. 171, no. 3, pp. 725-33.
Rochelle, GT 2009, 'Amine Scrubbing for CO2 Capture', Science, vol. 325, no. 5948, pp. 1652-4.
Rossi, TM, Campos, JC & Souza, MMVM 2016, 'CO2 capture by Mg–Al and Zn–Al hydrotalcite-like
compounds', Adsorption, vol. 22, no. 2, pp. 151-8.
Sabouni, R, Kazemian, H & Rohani, S 2014, 'Carbon dioxide capturing technologies: a review focusing
on metal organic framework materials (MOFs)', Environmental Science and Pollution Research, vol.
21, no. 8, pp. 5427-49.
64
Saha, S, Chandra, S, Garai, B & Banerjee, R 2012, 'Carbon dioxide capture by metal organic
frameworks', Indian Journal of Chemistry Section a-Inorganic Bio-Inorganic Physical Theoretical &
Analytical Chemistry, vol. 51, no. 9-10, pp. 1223-30.
Salles, F, Ghoufi, A, Maurin, G, Bell, RG, Mellot-Draznieks, C & Ferey, G 2008, 'Molecular Dynamics
Simulations of Breathing MOFs: Structural Transformations of MIL-53(Cr) upon Thermal Activation
and CO2 Adsorption', Angewandte Chemie-International Edition, vol. 47, no. 44, pp. 8487-91.
Salles, F, Jobic, H, Devic, T, Guillerm, V, Serre, C, Koza, MM, Ferey, Gr & Maurin, G 2013, 'Diffusion of
Binary CO2/CH4 Mixtures in the MIL-47 (V) and MIL-53 (Cr) Metal–Organic Framework Type Solids: A
Combination of Neutron Scattering Measurements and Molecular Dynamics Simulations', The
Journal of Physical Chemistry C, vol. 117, no. 21, pp. 11275-84.
Samanta, A, Zhao, A, Shimizu, GKH, Sarkar, P & Gupta, R 2012, 'Post-Combustion CO2 Capture Using
Solid Sorbents: A Review', Industrial & Engineering Chemistry Research, vol. 51, no. 4, pp. 1438-63.
Satoh, A 2010, Introduction to practice of molecular simulation: molecular dynamics, Monte Carlo,
Brownian dynamics, Lattice Boltzmann and dissipative particle dynamics, Elsevier.
Schlör, H, Fischer, W & Hake, J-F 2015, 'The system boundaries of sustainability', Journal of Cleaner
Production, vol. 88, no. Supplement C, pp. 52-60.
Seema, H, Kemp, KC, Le, NH, Park, S-W, Chandra, V, Lee, JW & Kim, KS 2014, 'Highly selective CO2
capture by S-doped microporous carbon materials', Carbon, vol. 66, no. Supplement C, pp. 320-6.
Shackley, S & Verma, P 2008, 'Tackling CO2 reduction in India through use of CO2 capture and
storage (CCS): Prospects and challenges', Energy Policy, vol. 36, no. 9, pp. 3554-61.
Shi, J-T, Yue, K-F, Liu, B, Zhou, C-s, Liu, Y-L, Fang, Z-G & Wang, Y-Y 2014, 'Two porous metal-organic
frameworks (MOFs) based on mixed ligands: synthesis, structure and selective gas adsorption',
CrystEngComm, vol. 16, no. 15, pp. 3097-102.
Skoulidas, AI & Sholl, DS 2005, 'Self-diffusion and transport diffusion of light gases in metal-organic
framework materials assessed using molecular dynamics simulations', The Journal of Physical
Chemistry B, vol. 109, no. 33, pp. 15760-8.
Takakura, J, Fujimori, S, Takahashi, K, Hijioka, Y, Hasegawa, T, Honda, Y & Masui, T 2017, 'Cost of
preventing workplace heat-related illness through worker breaks and the benefit of climate-change
mitigation', Environmental Research Letters, vol. 12, no. 6.
Thommes, M & Cychosz, KA 2014, 'Physical adsorption characterization of nanoporous materials:

progress and challenges', Adsorption, vol. 20, no. 2, pp. 233-50.
Tian, Z, Dai, S & Jiang, D-e 2016, 'What can molecular simulation do for global warming?', Wiley
Interdisciplinary Reviews: Computational Molecular Science, vol. 6, no. 2, pp. 173-97.
Torrisi, A, Bell, RG & Mellot-Draznieks, C 2010, 'Functionalized MOFs for Enhanced CO2 Capture',
Crystal Growth & Design, vol. 10, no. 7, pp. 2839-41.
65
Tudisco, C, Zolubas, G, Seoane, B, Zafarani, HR, Kazemzad, M, Gascon, J, Hagedoorn, PL & Rassaei, L
2016, 'Covalent immobilization of glucose oxidase on amino MOFs via post-synthetic modification',
RSC Advances, vol. 6, no. 109, pp. 108051-5.
Vogiatzis, KD, Mavrandonakis, A, Klopper, W & Froudakis, GE 2009, 'Ab initio Study of the
Interactions between CO2 and N-Containing Organic Heterocycles', ChemPhysChem, vol. 10, no. 2,
pp. 374-83.
Vujic, B & Lyubartsev, AP 2017, 'Computationally based analysis of the energy efficiency of a CO2
capture process', Chemical Engineering Science, vol. 174, pp. 174-88.
Waisman, E 1973, 'The radial distribution function for a fluid of hard spheres at high densities: mean
spherical integral equation approach', Molecular physics, vol. 25, no. 1, pp. 45-8.
Walton, KS, Millward, AR, Dubbeldam, D, Frost, H, Low, JJ, Yaghi, OM & Snurr, RQ 2008,
'Understanding inflections and steps in carbon dioxide adsorption isotherms in metal-organic
frameworks', Journal of the American Chemical Society, vol. 130, no. 2, pp. 406-+.
Wang, Z & Cohen, SM 2009, 'Postsynthetic modification of metal-organic frameworks', Chemical

Society Reviews, vol. 38, no. 5, pp. 1315-29.
Wilmer, CE & Snurr, RQ 2011, 'Towards rapid computational screening of metal-organic frameworks
for carbon dioxide capture: Calculation of framework charges via charge equilibration', Chemical
engineering journal, vol. 171, no. 3, pp. 775-81.
Wu, H, Simmons, JM, Srinivas, G, Zhou, W & Yildirim, T 2010, 'Adsorption Sites and Binding Nature of
CO2 in Prototypical Metal-Organic Frameworks: A Combined Neutron Diffraction and First-Principles
Study', Journal of Physical Chemistry Letters, vol. 1, no. 13, pp. 1946-51.
Xian, S, Peng, J, Zhang, Z, Xia, Q, Wang, H & Li, Z 2015, 'Highly enhanced and weakened adsorption
properties of two MOFs by water vapor for separation of CO2/CH4 and CO2/N2 binary mixtures',
Chemical engineering journal, vol. 270, no. Supplement C, pp. 385-92.
Xie, J, Yan, N, Qu, Z & Yang, S 2012, 'Synthesis, characterization and experimental investigation of
Cu-BTC as CO2 adsorbent from flue gas', Journal of Environmental Sciences, vol. 24, no. 4, pp. 640-4.
Yakovenko, AA, Reibenspies, JH, Bhuvanesh, N & Zhou, HC 2013, 'Generation and applications of
structure envelopes for porous metal-organic frameworks', Journal of Applied Crystallography, vol.
46, pp. 346-53.
Yan, X, Komarneni, S, Zhang, Z & Yan, Z 2014, 'Extremely enhanced CO2 uptake by HKUST-1 metal–
organic framework via a simple chemical treatment', Microporous and Mesoporous Materials, vol.
183, no. Supplement C, pp. 69-73.
Yan, XL, Li, SN, Jiang, YC, Hu, MC & Zhai, QG 2015, 'Synthesis, crystal structures and gas adsorption of
two porous pillar-layered MOFs decorated with different functional groups', Inorganic Chemistry
Communications, vol. 62, pp. 107-10.
Yang, Q, Vaesen, S, Vishnuvarthan, M, Ragon, F, Serre, C, Vimont, A, Daturi, M, De Weireld, G &

Maurin, G 2012, 'Probing the adsorption performance of the hybrid porous MIL-68(Al): a synergic
66
combination of experimental and modelling tools', Journal of Materials Chemistry, vol. 22, no. 20,
pp. 10210-20.
Yang, Q, Xue, C, Zhong, C & Chen, JF 2007, 'Molecular simulation of separation of CO2 from flue
gases in CU‐BTC metal‐organic framework', AIChE Journal, vol. 53, no. 11, pp. 2832-40.
Yang, Q & Zhong, C 2006, 'Molecular simulation of carbon dioxide/methane/hydrogen mixture

adsorption in metal-organic frameworks', The Journal of Physical Chemistry B, vol. 110, no. 36, pp.
17776-83.
Yang, QY, Zhong, CL & Chen, JF 2008, 'Computational study of CO2 storage in metal-organic
frameworks', Journal of Physical Chemistry C, vol. 112, no. 5, pp. 1562-9.
Yazaydın, AO, Benin, AI, Faheem, SA, Jakubczak, P, Low, JJ, Willis, RR & Snurr, RQ 2009, 'Enhanced
CO2 adsorption in metal-organic frameworks via occupation of open-metal sites by coordinated
water molecules', Chemistry of Materials, vol. 21, no. 8, pp. 1425-30.
Yazaydin, AO, Snurr, RQ, Park, TH, Koh, K, Liu, J, LeVan, MD, Benin, AI, Jakubczak, P, Lanuza, M,
Galloway, DB, Low, JJ & Willis, RR 2009, 'Screening of Metal-Organic Frameworks for Carbon Dioxide
Capture from Flue Gas Using a Combined Experimental and Modeling Approach', Journal of the
American Chemical Society, vol. 131, no. 51, pp. 18198-+.
Ye, S, Jiang, X, Ruan, L-W, Liu, B, Wang, Y-M, Zhu, J-F & Qiu, L-G 2013, 'Post-combustion CO2 capture
with the HKUST-1 and MIL-101(Cr) metal–organic frameworks: Adsorption, separation and
regeneration investigations', Microporous and Mesoporous Materials, vol. 179, no. Supplement C,
pp. 191-7.
Yu, C-H, Huang, C-H & Tan, C-S 2012, 'A review of CO2 capture by absorption and adsorption',
Aerosol Air Qual. Res, vol. 12, no. 5, pp. 745-69.
Yu, CH, Huang, CH & Tan, CS 2012, 'A Review of CO2 Capture by Absorption and Adsorption', Aerosol
and Air Quality Research, vol. 12, no. 5, pp. 745-69.
Yu, M, Yankovich, AB, Kaczmarowski, A, Morgan, D & Voyles, PM 2016, 'Integrated Computational
and Experimental Structure Refinement for Nanoparticles', Acs Nano, vol. 10, no. 4, pp. 4031-8.
Zeng, Q, Yu, A & Lu, G 2010, 'Evaluation of Interaction Forces between Nanoparticles by Molecular
Dynamics Simulation', Industrial & Engineering Chemistry Research, vol. 49, no. 24, pp. 12793-7.
Zhang, S, Yang, Q, Liu, X, Qu, X, Wei, Q, Xie, G, Chen, S & Gao, S 2016, 'High-energy metal–organic
frameworks (HE-MOFs): Synthesis, structure and energetic performance', Coordination Chemistry
Reviews, vol. 307, no. Part 2, pp. 292-312.
Zhang, Z, Zhao, Y, Gong, Q, Li, Z & Li, J 2013, 'MOFs for CO2 capture and separation from flue gas
mixtures: the effect of multifunctional sites on their adsorption capacity and selectivity', Chemical
Communications, vol. 49, no. 7, pp. 653-61.
Zhao, L, Yang, Q, Ma, Q, Zhong, C, Mi, J & Liu, D 2011, 'A force field for dynamic Cu-BTC metal-
organic framework', Journal of molecular modeling, vol. 17, no. 2, pp. 227-34.
67
Zhou, H-C, Long, JR & Yaghi, OM 2012, 'Introduction to metal–organic frameworks', Chemical
Reviews, vol. 112, no. 2, pp. 673-4.
Zhou, Z, He, C, Xiu, JH, Yang, L & Duan, CY 2015, 'Metal-Organic Polymers Containing Discrete Single-
Walled Nanotube as a Heterogeneous Catalyst for the Cycloaddition of Carbon Dioxide to Epoxides',
Journal of the American Chemical Society, vol. 137, no. 48, pp. 15066-9.
68

A Molecular Simulation Study of CO Adsorption in Metal-Organic Frameworks

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Molecular Simulation Study of CO Adsorption in Metal-Organic Frameworks

Uploaded by

Copyright:

Available Formats

A molecular simulation study of CO2 adsorption in

Supervisor: Dr Qinghua Zeng

School of Computing, Engineering and Mathematics

Figure 5. Periodic boundary condition..................................................................................... 19

Figure 6. Schematic diagram of a basic MD code(Zeng, Yu & Lu 2010). .............................. 21

Figure 7. Mean square displacement. ...................................................................................... 22

Figure 8. Radial distribution function. ..................................................................................... 23

Table 2 CO2 adsorption capacity on HKUST-1(Cu-BTC) under different conditions ............ 29

Table 3 Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1 .................. 39

Table 4 Interaction potential energies (Kcal/mol) of CO2 adsorbed in HKUST-1F................ 39

Table 7 Interaction potential energies (Kcal/mol) of CO2 adsorbed in MIL-68F (In)............. 53

ABSTRACT ............................................................................................................................. iii

LIST OF FIGURES .................................................................................................................. iv

LIST OF TABLES .................................................................................................................... vi

LIST OF PUBLICATIONS .....................................................................................................vii

CHAPTER 1 INTRODUCTION ............................................................................................... 1

1.1 Research background and objectives ................................................................................... 1

1.2 Thesis outline ....................................................................................................................... 2

CHAPTER 2 LITERATURE SURVEY .................................................................................... 3

2.1 Global warming and carbon emissions ................................................................................ 3

2.2 Carbon dioxide capture and storage ..................................................................................... 4

2.3 Amine absorption for CO2 emissions .................................................................................. 5

2.4 Metal-organic frameworks for CO2 adsorption and separation ........................................... 7

2.4.1 Experimental work for CO2 capture on MOFs ................................................................. 9

2.4.2 Molecular simulations for CO2 capture on MOFs .......................................................... 11

2.5 Current research summary and challenges ........................................................................ 14

CHAPTER 3 METHODOLOGY ............................................................................................ 16

3.1 Input parameters................................................................................................................. 16

3.1.1 Atomic partial charge ...................................................................................................... 16

3.1.2 Force field ....................................................................................................................... 17

3.2 Ensemble ............................................................................................................................ 18

3.3 Periodic boundary condition and cut-off distance ............................................................. 18

3.4 Principle of Grand Canonical Monte Carlo simulation ..................................................... 19

3.5 Principle of molecular dynamics simulation...................................................................... 20

3.6.1 Mean square displacement .............................................................................................. 22

3.6.2 Radial distribution function ............................................................................................ 22

3.6.3 Adsorption strength ......................................................................................................... 23

3.7 Summary ............................................................................................................................ 25

CHAPTER 4 CARBON DIOXIDE ADSORPTION IN HKUST-1 ........................................ 26

4.1 Model configuration........................................................................................................... 26

4.2 Grand Canonical Monte Carlo Simulation details ............................................................. 27

4.2.1 Adsorption isotherms of CO2 in HKUST-1 and HKUST-1F ......................................... 28

4.2.2 Comparison of CO2 adsorption isotherms of CO2 in HKUST-1 .................................... 29

4.3 Molecular dynamics simulation results ............................................................................. 30

4.3.1 Adsorption sites of CO2 in HKUST-1 and HKUST-1F .................................................. 30

4.3.2 Self-diffusivity of CO2 in HKUST-1 and HKUST-1F ................................................... 37

4.3.3 Interaction strength between CO2 and HKUST-1 ........................................................... 38

4.4 Summary ............................................................................................................................ 39

CHAPTER 5 CARBON DIOXIDE ADSORPTION IN MIL-68(In) ...................................... 40

5.1 Model configuration........................................................................................................... 40

5.2 Grand Canonical Monte Carlo Simulation details ............................................................. 41

5.2.1 Adsorption isotherms of CO2 in MIL-68(In) and MIL-68F (In) .................................... 42

5.3 Molecular dynamics simulation results ............................................................................. 44

5.3.1 Adsorption sites of CO2 in MIL-68(In) and MIL-68F (In) ............................................. 44

5.3.2 Self-diffusivity of CO2 in MIL-68(In) and MIL-68F (In) .............................................. 51

5.3.3 Interaction strength between CO2 and MIL-68(In) ......................................................... 52

5.4 Summary ............................................................................................................................ 53

CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS ............................................ 54

6.1 Conclusions ........................................................................................................................ 54

1.1 Research background and objectives

playing a significant role in meeting such enormous demand since industrialization

emissions (Banerjee 2012; Saha et al. 2012).