Effect of Water on the CO2 Adsorption Capacity of Amine

Publication Date (Web): May 2, 2017. Copyright © 2017 American ... Pore Modified FDU-12 as a Novel Container for Dendron Growth. Junxing Han , Yueyun...
0 downloads 0 Views 2MB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Effect of water on the CO2 adsorption capacity of amine-functionalized carbon sorbents Peter Psarras, Jiajun He, and Jennifer Wilcox Ind. Eng. Chem. Res., Just Accepted Manuscript • Publication Date (Web): 02 May 2017 Downloaded from http://pubs.acs.org on May 3, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Effect of water on the CO2 adsorption capacity of amine-functionalized carbon sorbents Peter Psarras1, Jiajun He2, and Jennifer Wilcox1,* 1

Department of Chemical and Biological Engineering, Golden, CO 80401; 2Department of

Energy Resources Engineering, Stanford University, Stanford, CA 94305. KEYWORDS. adsorption, carbon dioxide capture, nitrogen functionalization, solid sorbent

ABSTRACT Molecular simulation is used as a tool to improve understanding of CO2 adsorption in nitrogenfunctionalized carbon sorbents in the presence of water vapor, which is crucial to the advancement of adsorption approaches to CO2 separation from exhaust streams of coal- and natural gas-fired power plants. Molecular simulations were carried out for binary mixtures of CO2 and H2O over 4 N-functionalized surfaces and 3 variations of the quaternary group with increasing wt. % N. The quaternary group was found to be most stable, with a 13% loss in CO2 capacity observed, followed by the pyrrolic and pyridonic groups, which lost 25 and 28% CO2 loading capacity, respectively. The oxidized pyridinic group demonstrated a dramatic loss in capacity, i.e., 58% when compared to ideal loading. The quaternary group was the only functionality to display loading in excess of 2.0 mmol CO2 g-1 sorbent under ambient temperature

ACS Paragon Plus Environment

1

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

and 1% humidity (2.40 mmol CO2 g-1 sorbent). Further, the two functional groups without oxygen were shown to be more resistant to competitive H2O adsorption at low humidity. In general, increasing nitrogen content appears to buffer the CO2 capacity loss under low humidity, yet such systems appear to be incompatible with CO2 separation at ambient temperature and 10% humidity. Further, the results of this work suggest that materials modified with pyrrolic and pyridonic groups and pore size weighted in the supermicroporous region are most resistant to compromised CO2 loading under 10% humidity.

Introduction Over the past several years, carbon dioxide adsorption technologies have been investigated as an alternative CO2 mitigation strategy, driven in part by concerns over the water and energyintensity associated with more mature and conventional amine solvent-based processes. Solid sorbents can offer several benefits over their solvent-based counterparts. First, elimination of solvent is critical to lowering the heat of regeneration, which is influenced greatly by the heat capacity of solution (for example, 3.496 J g-1K-1 for 30 wt % monoethanolamine (MEA) solutions at 298.15 K, 0.40 mol CO2/mol MEA loading)1. It has been estimated that 80% of the energy required to release CO2 goes toward heating of the solution.2 Naturally, solid materials have lower heat capacities than liquids (compare, 0.7 J g-1K-1 for graphite at 298.15 K)3; further, the gas-solid interaction can be weaker depending upon the physical or chemical nature of the adsorption process (. . ,  < 40 kJ mol-1).4 Moreover, solid sorbents offer tunability over pore geometries and pore dimensions, as well as flexibility for heteroatom doping or surface functionalization, which are essential for optimized gas diffusivity and/or sorbent-CO2 interactions.19, 20 Solid sorbents possess additional advantages over amine solvents, such as a relatively wide range of operating temperatures, and less waste produced for disposal.21 However, many of the factors that lead to a lower theoretical energy of separation can also compromise loading. For instance, Gray et al. estimate 3–4 mmol CO2 g-1 sorbent (298 K, 1 atm)

ACS Paragon Plus Environment

2

Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

as the minimum CO2 working capacity to be considered cost competitive with conventional amine scrubbing.5

Solid sorbents may also be advantageous when paired with intermittent renewable energy sources such as solar and wind. Renewable energy is increasingly identified as a necessary component in climate change mitigation, with the potential to achieve carbon neutrality and negativity when paired with carbon capture utilization and sequestration (CCUS). Unfortunately, several technical barriers exist that stall significant deployment of renewable energy development and integration, namely land management and energy intermittency. In traditional solvent-based separation processes, much of the energy penalty associated with solvent heating is circumvented through heat exchange via the reboiler.6,7 This exchange is contingent on the uninterrupted processing of flue gas, which is in turn contingent on a non-intermittent power supply; thus, pairing of solvent-based processes with renewable energy – even if these solvents show a higher working capacity for CO2 – may be energetically wasteful and unpractical. Conversely, solid sorbents operating under pressure swing adsorption (PSA) may integrate well with intermittent sources. Unlike temperature swing adsorption (TSA) which typically experience longer loading cycles and where thermal integration with power plants is critical to efficient performance,8 PSA features relatively fast cycles with the absence of thermal integration and thus it is capable of being operated in an intermittent way powered by solar or wind. In summary, the advances of sorbent technologies create opportunities for the deployment of renewable energy-assisted carbon capture.

Several classes of sorbents for CO2 capture have been characterized in the literature.9-11 Generally, zeolites are associated with moderate surface area (typically under 2000 m2/g), moderate production costs ($0.20/mol CO2 captured vs. $0.25 and $0.544/mol CO2 captured for activated carbon and MEA, respectively, assuming 2 mol CO2 loading per kg activated carbon, 5 mol CO2 loading per kg zeolite)12 and typically show sensitivity to humid operating conditions. More recently, zeolitic imidazole frameworks (ZIF) exhibit enhanced chemical and thermal stability.13 Further, hydrophobic ZIFs exhibit essentially no loss in CO2 capacity when co-fed with wet N2 (80% humidity);14 however, they remain plagued by the lack of pore tunability.

ACS Paragon Plus Environment

3

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

Metal organic frameworks (MOFs) are known to exhibit ultra-high surface areas (in excess of 6000 m2/g BET specific surface area )15 and excellent high-pressure CO2 capacities (as high as 54.5 mmol g-1 at 50 bar);23 yet, typical MOFs suffer from competitive water adsorption or chemical instability to moisture.24 For example, one family of MOF materials with superb CO2 capacity (upwards of 23.6 wt %) show a 67% reduction in CO2 capacity in the presence of 9% humidity.16 Recently, McDonald et al. reported a chemisorption strategy to improve CO2 adsorption in the presence of moisture.25 Nevertheless, the material cost can be high, taking into account the scalability, time for preparation and post-synthesis functionalization cost. Carbonaceous sorbents are earth-abundant materials that are generally low cost alternatives to MOFs and display reasonable chemical stability and cyclability. Additionally, the non-polar nature of carbon materials lends to excellent stability under humid conditions. Unfortunately, the unmodified carbon surface has relatively low working capacity for CO2 (< 2 mmol g-1 sorbent)17 making it unsuitable for deployment at scale. It has been shown that chemical surface modification of carbon surfaces can lead to enhanced CO2 uptake at low CO2 partial pressure and, in some cases, enhanced CO2/N2 selectivity.18,

19, 22

Modifications via oxygen-based

functionalities20 and nitrogen-based functionalities21 have been explored via molecular simulations, allowing for a direct examination of the connection between individual functionalities and pore size on CO2 performance, i.e., working capacity and CO2/N2 selectivity.

It is important to consider that a typical flue gas derived from fossil-fuel fired power plants will contain 8–10% H2O.22 The presence of water vapor has the potential to compromise the sorbent CO2 adsorption capacity by virtue of competitive dispersion-repulsion interactions with the surface and, more importantly, competing electrostatic interactions. The former is defined by the Lennard-Jones (LJ) potential:    

= 4   −   

(1)

ACS Paragon Plus Environment

4

Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

where  denotes the potential well depth and  represents the finite distance at which the interparticle potential reaches zero. Electrostatic interactions are described by a combination of energy contributions:2 . = −

1  1 %! ! + −#! + $ 2 2 %

(2)

where terms on the right-hand side correspond to (from left to right) polarization, field-dipole, and the field-gradient quadrupole. An additional contribution from sorbate-sorbate interactions on the surface may be included at high coverage. Competitive adsorption arises from the fact that the species present in a flue gas mixture have different LJ and electrostatic parameters as shown in Table 1. TABLE 1. Potential parameters used in calculation of force field &&⁄'( K

&& Å

q(e)a

Ref.

C(CO2)

27.0

2.8

+ 0.70

23

O(CO2)

79.0

3.05

– 0.35

23

+ 0.417

24

H(H2O) O(H2O)

76.58

3.15

– 0.834

24

C(surface)

28.0

3.4

calc.

25

N(surface)

60.39

3.296

calc.

26, 27

O(surface)

79.0

3.1

calc.

25

H(surface)

30.0

1.31

calc.

25

a

Bader charge assignments for surface atoms are calculated separately for each system.

Further, differences in physical properties for CO2 and H2O lend to different contributions to the electrostatic potential (Eq. 2). A comparison of these properties is presented in Table 2. TABLE 2. Comparison of select physical properties of CO2 and H2O. Adapted from Ref. [2]. Kinetic Diameter (Å)

Dipole Moment (Debye)

Quadrupole moment (10–40 Coulomb m2)

Polarizability (10–24 cm3)

ACS Paragon Plus Environment

5

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

CO2

0.330

0

–13.71, –10.0

2.64, 2.91, 3.02

H 2O

0.280

1.85

6.67

1.45, 1.48

Several studies have successfully modeled the effects of water adsorption using the parameters outlined in Table 1.28-30 This study examines the effect of water vapor presence in low (1%) and typical (10%) amounts on the loading enhancement of several nitrogen-modified carbon surfaces. Though it has been demonstrated that N-functionalization can improve CO2 uptake through the introduction of surface charge and interaction of the local field gradient with the CO2 quadrupole moment, the dipole interaction of water is notably stronger and is expected to compromise CO2 loading capacity through the preferential occupation of reactive surface sites. This effect will be simulated through the investigation of CO2/H2O binary mixtures within pores of various widths functionalized by four unique surfaces: pyrrolic nitrogen (N5), oxidized pyridinic nitrogen (NO), pyridonic nitrogen (NP), and quaternary nitrogen (NQ) (Figure 1). These functional groups have been confirmed to exist within N-doped carbons and have been frequently reported in the literature.17, 31-33 Further, previous experimental analysis on similar functionalities revealed no hysteresis on desorption,34 and excellent cyclic stability35 making these materials suitable candidates for use in PSA. Additionally, the presence of water vapor will be tested against four surfaces of increasing wt. % N (i.e., 3.6 – 28.0 %), henceforth abbreviated NQ, Q2, Q4, and Q8. These results will provide further knowledge of the relative stability of Nfunctionalities toward CO2 capacity in varying humidity.

ACS Paragon Plus Environment

6

Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 1. Surface view of the 7 unique 3-layered modified graphitic models examined in this study. Model abbreviation included in parentheses.

2. Method 2.1 Density Functional Theory (DFT): Model structure and optimization Hierarchical, carbon-based pore structures were approximated as a collection of independent, functionalized graphitic slit-pores. This approximation has been used with success in several studies

20, 36-38

. All structures were modeled as 3-layered graphitic slabs in 4 - 4 carbon ring unit

cells (96 atoms for pure graphene). Functional groups were substituted into the top layer at a rate of one nitrogen atom per unit cell. In some cases, it was necessary to remove additional top layer carbon atoms to simulate a graphene edge; thus, nitrogen coverage varied slightly according to functionalization. All cells were subject to geometric optimization within the Vienna ab initio simulation package (VASP), 39 with a van der Waals correction40 applied for proper optimization of graphitic layer spacing. A plane-wave basis set was employed and truncated at 750 eV to achieve a balance of computational efficiency within the desired force threshold of 1 meV/Å.

ACS Paragon Plus Environment

7

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

Projector augmented wave (PAW)41 potentials were used to describe core electronic behavior. The exchange-correlation functional of Perdew, Burke, and Ernzerhof (PBE)42 was employed for non-local corrections, and a 6 - 6 - 1 Monkhorst-Pack43 grid was used to sample the first Brillouin zone. All optimizations were carried out using the conjugate-gradient algorithm. Charge partitioning was achieved through the Bader method,44 and core-level charge was included in the integrations. The fine-grid was adjusted until the total system charge was preserved to within ± 0.001 e.

The molecular simulations result in pore density profiles; however, if a target material pore size distribution (PSD) is made available, adsorption isotherms may be predicted. Here we use a PSD based on the work of He et al.,35 which served as the basis for adsorption isotherm predictions in greater detail in previous work and is demonstrated in Figure 2 as an example.21 The material (named SU-MAC 500, for Stanford University mesoporous activated carbon, with an activation temperature of 500

) demonstrated excellent CO2 working capacity, CO2/N2 selectivity, and

reasonable cyclic stability under humid conditions. The detailed methodology for the experimental PSD is provided in our previous study.35 The experimentally determined PSD was partitioned such that the full structure could be approximated as a weighted sum of pores with the following internal diameters: 3.5, 4, 5, 6, 7, 8, 9, 10, 12, 14, 18, 22, 28, 34, 40, 50, 60, 70, 80, 90, 100, 120, 140, 160, 180, and 200 Å, where the pore width was assumed as the distance between carbon atoms at opposing pore walls less the collision diameter of a surface carbon atom (3.35 Å). Pores were created by first transposing the functional group from the center of the cell (where optimized) to off-center, then mirroring

ACS Paragon Plus Environment

8

Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

the VASP-optimized cells about a point of inversion at the pore center. This method prevented unnatural overlap of opposing wall functionalities in the smaller pore sizes.

Figure 2. Pore size distribution used as the basis for adsorption isotherm predictions. Modeled after the material SUMAC 500 from Ref. [35]

2.2 Molecular simulations Grand canonical Monte Carlo (GCMC) simulations45 were carried out to describe CO2 and H2O adsorption on the idealized functional pores previously discussed. Carbon dioxide was described according to TraPPE parameters23, which have been shown to quantitatively reproduce CO2 vapor-liquid equilibria. This three-site rigid model treats the intrinsic quadrupole moment of CO2 by assignment of partial charge at each site (qC = 0.70 e and qO = -0.35 e). The C=O bond lengths were fixed at 1.16 Å and the O=C=O bond angle fixed at 180.0 degrees. Water was modeled using TIP3P parameters,24 which effectively captures the intrinsic dipole moment. This three-site rigid model assigns partial charges of +0.417 e to two hydrogen sites and –0.834 e to the oxygen site, with H–O bond lengths of 0.957 Å and an H–O–H bond angle of 104.52 degrees. Potential parameters for surface carbon, oxygen and hydrogen were modeled after the studies of Steele46 and Tenney and Lastoskie25, respectively, while the parameters for surface

ACS Paragon Plus Environment

9

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

nitrogen were modeled after the work of Gubbins et al.27 All potentials associated with different LJ sites were calculated using standard Lorentz-Berthelot mixing rules.47 The surface charge was assigned according to those calculated in the Bader method described previously.

Simulations were carried out in the #VT ensemble at 298 K and 333 K and for a pressure range of 0.001 to 1 bar. The Peng-Robinson equation of state was used to calculate fugacity and to relate bulk experimental pressure to chemical potential. Accepted move types included energybiased insertions, deletions, rotations and translations. A total of 10 million GCMC moves were attempted during each simulation to ensure adequate system equilibration. To reduce system load, a rigid framework was assumed and sorbate-surface interactions were interpolated to a pretabulated potential map with grid spacing of 0.1 Å (sorbate-sorbate interactions were calculated on-the-fly).

3. Results and Discussion 3.1 Effect of functionality on CO2 loading capacity Ideal CO2 adsorption isotherms are presented here against binary CO2:H2O mixtures for humidity conditions of 1% H2O and 10% H2O (Figure 3). Adsorption isotherms presented on the left represent CO2 loading over NQ, N5, NO, and NP in mild humidity (1% H2O). These results are presented against ideal adsorption isotherms for the same materials obtained in an earlier study21 to help directly visualize the loss in loading capacity as function of water vapor presence. The quaternary group (NQ) was found to be most stable, with a 13% loss in CO2 capacity

ACS Paragon Plus Environment

10

Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

observed, followed by the pyrrolic and pyridonic groups which lost 25 and 28% CO2 loading capacity, respectively. The oxidized pyridinic group saw a dramatic loss in capacity – 58% – when compared to ideal loading; however, due to exceptional ideal loading performance, the relative capacity for CO2 uptake at 1 bar (1.81 mmol CO2 g-1 sorbent) is similar to the N5 and NP groups (1.76 and 1.95 mmol CO2 g-1 sorbent, respectively). The quaternary group was the only functionality to display loading in excess of 2.0 mmol CO2 g-1 sorbent under these conditions (2.40 mmol CO2 g-1 sorbent). Further, the two functional groups without oxygen were shown to be more resistant to competitive H2O adsorption at low humidity.

Figure 3. Comparison of CO2 loading on 4 unique N-functionalized surfaces at 298 K. Ideal loading presented as a reference and reproduced from Ref. [21] with permission from the PCCP Owner Societies. Binary CO2 loading indicated by point-down triangles. The lower plots are zoomed-in regions of the full isotherm, representing low

ACS Paragon Plus Environment

11

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

pressure (i.e., 0 – 10 mbar) loading performance. Isotherms on the right-hand side show greatly compromised CO2 loading in typical flue conditions (10% H2O).

Generally, CO2 loading capacity was greatly compromised for all modified sorbents at 10% humidity, with the greatest percentage loss in loading corresponding to the quaternary (98%) and oxidized pyridinic groups (97%). As indicated in the adsorption isotherm (Figure 3, top right), CO2 loading is dramatically reduced (even at low pressures) and marginal loading is observed for p/po > 0.1. Interestingly, the pyrrolic group is considerably more stable over 0 < p/po < 0.25, with a sharp decline in loading observed at p/po = 0.3. A comparison of the zoomed-in low-pressure isotherms reveals that CO2 remains competitive with H2O for high-energy surface sites (functional substitutions) at low pressures, as indicated by the relative isotherm slopes for the low and moderate humidity conditions. An exception to this trend is observed for the pyrrolic group, whereby initial competition for pyrrolic sites appears to be dominated by H2O. Considering that the pyrrolic group is indeed the most stable under the moderate humidity of the four groups studied here, it is worth a moment to comment on the pyrrolic surface chemistry.

ACS Paragon Plus Environment

12

Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 4. Bader partial charge assignments for select surface atoms. Bonds and select atoms omitted for clarity. Smaller magnitude charges (|| 0 0.10 ) not shown.

Aside from the absence of oxygen, the pyrrolic nitrogen has a smaller Bader partial charge than the quaternary and pyridonic groups, a consequence of bonding to the less electronegative hydrogen (Figure 4). Additionally, the strained 5-membered pyrrolic ring creates more vacancy space within the immediate vicinity of the functional group. The combination of these latter two effects may leave the pyrrolic nitrogen sterically unhindered – in spite of early (low pressure) H2O bonding – allowing the pyrrolic nitrogen to continue to participate in CO2 loading at higher partial pressures. These simulations were repeated at 333.15 K. Resulting isotherms are presented together for 1% and 10% humidity as shown in Figure 5. For the surfaces NQ, N5, and NP at 1% humidity, CO2 loading declined by 38, 19, and 19%, respectively relative to the 1% humidity capacity at 298.15 K. From the kinetic theory of gases, it is intuitive to anticipate a drop in single-component working capacity at elevated temperature. For gas mixtures, a balance exists between the

ACS Paragon Plus Environment

13

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

increasing kinetic energy of each gas-phase component, the relative heats of adsorption, and relative partial pressures; thus, an increase in loading for one component in a binary mixture may be observed if, for instance, a competitive sorbate is more susceptible to temperature change. This effect is manifested here, albeit slightly, by a 1% increase to CO2 loading over the NO surface, which also demonstrated the strongest CO2 capacity of the four surfaces (1.71 mmol CO2 g-1 sorbent). The relative CO2 capacity at 333.15 K and 1% humidity decreased on the order of NO > NP > NQ > N5, which does not align with the trend observed at 298.15 K (NQ > NP > NO > N5), indicating that specific functional performance is nonlinear with temperature.

Figure 5. Comparison of CO2 loading at 1% and 10% humidity at 333.15 K. Carbon dioxide is shown to be more competitive with H2O over N-modified surfaces at elevated temperatures (when compared to results at 298.15 K, Figure 3). Loading for the pyrrolic group is relatively unaffected when the humidity is increased from 1% to 10%, an extension of the trend observed at 298.15 K.

At 10% humidity, CO2 capacity increased dramatically for NQ, N5, and NO (with a concurrent decline in capacity of ca. 10% for NP) relative to the 298.15 K 10% humidity results. Adsorption

ACS Paragon Plus Environment

14

Page 15 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

isotherms for NQ and NP pass through a maximum at low pressure, then loading steadily declines to 1 bar. Since H2O is likely adsorbed more strongly than CO2 (see Eq.2 and Table 2), CO2 is gradually displaced at higher partial pressures of H2O. Interestingly, this trend is absent from the N5 isotherm, further supporting how low incorporation of pyrrolic groups yield a surface with moderate hydrophobicity. The observed relative increase in CO2 loading can be attributed to the effect of temperature on the interaction between H2O and the surface. This is discussed further in Section 3.3.

3.2 Effect of increased N wt % on CO2 loading capacity The models examined in Section 3.1 had similar nitrogen content, ca. 3.6 wt. %. Here we examine the effect of increased nitrogen content via systematic substitution of nitrogen for graphitic surface carbon to obtain models of 7.22, 14.4, and 28.0 wt. % N (Q2, Q4 and Q8, respectively from Figure 1). Results are presented for 298.15 K and 333.15 K in Figure 6.

Figure 6. Carbon dioxide adsorption isotherms for three surfaces with increasing wt. % N at 298.15 K (left) and 333.15 K (right) in 1% (circles) and 10 % humidity. Only the Q8 model (28.0 wt. % N) demonstrated +2.0 mmol CO2 g-1 sorbent loading at both temperatures. Increase in temperature had a negligible effect on loading on all models at 1% humidity, while higher temperatures appear to stabilize CO2 on the N-modified surface at 10%

ACS Paragon Plus Environment

15

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

humidity.

Loading at ambient temperature was found to be similar to loading at elevated temperature for all models, with variations measured at +/– 1 to 3 %. Likewise, the loss in 1 bar loading relative to ideal CO2 capacity is reported here as a decrease in 52, 52, and 45 % for the Q2, Q4, and Q8 systems, respectively. Only the heavily modified Q8 system demonstrated a working capacity in excess of 2.0 mmol CO2 g-1 sorbent. At ambient temperature, all systems suffered great losses in CO2 loading capacity at 10% humidity, with decreases ranging from 97 to 99% of the ideal CO2 capacity. Naturally, such a loss in CO2 loading is accompanied by increased water loading. For example, at 1 bar and 333.15 K, the H2O loading jumps from the marginal 0.02 and 0.01 mmol H2O g-1 sorbent for Q4 and Q8, respectively to 8.58 and 7.76 mmol H2O g-1 sorbent, respectively, when the humidity increases from 1 to 10%. At a lower temperature of 298.15 K and 10% humidity, loadings of 13.7 and 18.1 mmol H2O g-1 sorbent were achieved for the same sorbents at 1 bar. In summary, increasing nitrogen content appears to buffer the CO2 capacity loss under low humidity, yet such systems appear to be incompatible with CO2 separation at ambient temperature and 10% humidity.

3.3 Pore densities at select pore widths To gain a better understanding of the influence of material design on H2O stability, pore densities are reported for select pore widths within the ultramicroporous (d = 5.0 Å), supermicroporous (d = 10.0 Å), and near mesoporous (d = 22.0 Å) regions. Results are presented in Figure 7.

ACS Paragon Plus Environment

16

Page 17 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 7. a) Pore densities for pores of width 5, 10, and 22 Å at 1 bar. Notable is the relative stability of the pyrrolic and pyridonic groups at ambient temperature and 10% humidity, likewise the absence of CO2 loading for the 5 Å pore at the same conditions. b) Probability density function distributions for CO2 and H2O loading taken along the z cell axis at ambient temperature and low humidity. The lack of CO2 loading observed in the 5 Å pore may be a result of complete pore filling by H2O at higher humidity, as indicated by the preference of H2O near the pore wall. Note changes in the y-axis scaling.

Generally, loading decreased with pore widening at low (1%) humidity, likewise with an increase in temperature. Carbon dioxide loading in the 5 Å pore was similar at both temperatures in the oxidized pyridinic and pyridonic models, with greater loss in capacity observed in the wider pores. This effect is commonly observed in narrow pores, whereby opposing pore wall potentials can overlap to enhance sorbate uptake. It has been demonstrated elsewhere that loading

enhancement

induced

by

surface

functionalization

becomes

marginal

for

ACS Paragon Plus Environment

17

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

supermicropores and converges with the unmodified surface for mesopores.21 This trend has inspired materials with large ultramicroporous volumes – such as the SU-MAC 500 – designed to maximize benefits in loading and selective CO2 uptake. However, pore density results in the presence of 10% humidity indicate that very narrow pores are in fact non-ideal for CO2 uptake (Figure 7a). Further, these results suggest that materials modified with pyrrolic and pyridonic groups and pore size weighted in the supermicroporous region are most resistant to compromised CO2 loading under 10% humidity. An examination of the probability density functions (PDFs) for CO2 and H2O in 5 Å pore widths reveal that H2O preferentially adsorbs at the pore wall (represented as the left and right vertical axes in Figure 7b), whereas CO2 exists primarily in the pore center; thus, as the humidity is increased to 10%, ultramicropores become saturated with H2O and are consequently inaccessible to CO2. Here, results for the NQ and NO groups are presented, where multi-layering appears to be slightly enhanced for the NO model. This observation is reflected in higher pore densities as reported in Figure 7a. Two distinct trends can be observed when considering the effect of temperature on loading capacity. The first involves loadings at 1% humidity in the 5 Å pore, where loadings for NP and NO are marginally reduced, and N5 and NQ undergo large reductions. Here, the NP and NO pores are clearly saturated with CO2 in both cases, indicating that the strength of interaction between these functionalities and CO2 outweighs the decrease in surface residence time onset by elevated temperature. This is not the case, however, for the N5 and NQ functionalized pores. As these latter functionalities can be considered less polar than NO and NP (i.e., they lack oxygen) it is unlikely that H2O is preferentially adsorbing at elevated temperature, an effect not seen for NP and NO by virtue of unchanged loading at 333.15 K; rather, the decrease in loading is likely a

ACS Paragon Plus Environment

18

Page 19 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

function of lower isosteric hearts of adsorption for N5 and NQ and decreased surface residence time of CO2. Second, we consider the sharp increase of loading for N5 (22 Å) at 333.15 K and 10% humidity. It should be pointed out that all functionalities observed an increase in CO2 loading within the 22 Å pore at elevated temperature. This observation is counter-intuitive from a kinetic view, leaving only H2O presence as responsible for the enhanced uptake in larger pores, possibly through new acid-base interactions of the co-adsorbed species. Interestingly, the greatest enhancement was observed for N5 and NQ, the two functionalities most comprised in 5 Å loading. Collectively, these results underlie how the mechanism of loading as a function of temperature is extremely sensitive to both surface chemistry and pore size, where subtle changes to either property can yield large changes in loading behavior. This phenomenon has been observed elsewhere.48 Conclusions Molecular simulations were carried out for CO2 and H2O binary mixtures over 4 Nfunctionalized surfaces and 3 variations of the quaternary group at increasing wt. % N. Generally, these materials were found to be largely unsuitable for CO2 capture under 10% humidity at ambient temperature, with severe loss to CO2 loading observed, with a decrease of up to 99% of the ideal capacity. Reductions in loading were partially offset by increased temperature (333 K); however, loading results for individual functionalities were found to be non-linear with increasing temperature. Of the functionalities tested, the pyrrolic group was found to be most stable under 1 and 10% humidity, a result that is likely a consequence of additional surface vacancy space created by the strained 5-memberd pyrrolic ring, and the relative “softness” of the nitrogen partial charge due to bonding with the less-electronegative hydrogen. The latter condition creates a less polar environment and, due to vacancy spacing, an

ACS Paragon Plus Environment

19

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

environment that is likely less sterically hindered to accommodate CO2 adsorption in spite of water presence.

Increasing nitrogen content upwards of 28.0 wt. % yielded adequate CO2 loading (2.4 mmol g-1 sorbent) despite low humidity. This could be one tool to offset anticipated reductions to CO2 capacity in practice. However, the aforementioned reductions in CO2 loading capacity at 10% humidity further underscore the need for moderate stream dehydration. Further studies will focus on optimizing the material-humidity level balance. Additionally, ultramicroporous volumes were shown to be especially susceptible to water uptake. This challenges the viewpoint that materials should be designed with high ultramicroporous volumes to maximize pore wall-induced benefits in CO2 uptake and selectivity. Instead, materials weighted toward the supermicroporous volumes may represent a suitable compromise between functionality enhancement and stability to moisture. Future work will focus on the effects of other common flue constituents. For example, coalderived flue gas contains fly ash particles, oxygen, moisture, carbon monoxide and many acid gases. A typical untreated flue gas derived from the combustion of a US Low Sulfur Eastern bituminous coal can contain: 5 - 7% H2O, 3 - 4% O2, 15-16% CO2, 1 ppb total Hg, 20 ppm CO, 10 ppm hydrocarbons, 100 ppm HCl, 800 ppm SO2, 10 ppm SO3, 500 ppm NOx, and balance N2.49-51 There can be potential impacts of acid gases such as SO2, SO3, NO, NO2, and HCl upon the CO2 capture through competing acid-base surface chemistries. The possibility of oxidative degradation of the sorbent by oxygen and these same acid gas species also merits future study.52 Corresponding Author

ACS Paragon Plus Environment

20

Page 21 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

*Email Address: [email protected] Funding Sources This work is partially funded by the Global Climate and Energy Project (GCEP) and Precourt Institute for Energy.

REFERENCES 1. Weiland, R. H.; Dingman, J. C.; Cronin, D. B., Heat capacity of aqueous monoethanolamine, diethanolamine, N-methyldiethanolamine, and N-methyldiethanolaminebased blends with carbon dioxide. Journal of Chemical & Engineering Data 1997, 42, (5), 10041006. 2. Wilcox, J., Carbon capture. Springer: 2012. 3. Pop, E.; Varshney, V.; Roy, A. K., Thermal properties of graphene: Fundamentals and applications. MRS bulletin 2012, 37, (12), 1273-1281. 4. Atkins, P. W., Physical Chemistry. 4 ed.; Oxford University Press: 1990. 5. Gray, M. L.; Champagne, K. J.; Fauth, D.; Baltrus, J. P.; Pennline, H., International Journal of Greenhouse Gas Control 2008, 2, (3). 6. Rubin, E. S.; Chen, C.; Rao, A. B., Cost and performance of fossil fuel power plants with CO2 capture and storage. Energy policy 2007, 35, (9), 4444-4454. 7. Freguia, S.; Rochelle, G. T., Modeling of CO2 capture by aqueous monoethanolamine. AIChE Journal 2003, 49, (7), 1676-1686. 8. Hedin, N.; Andersson, L.; Bergström, L.; Yan, J., Adsorbents for the post-combustion capture of CO2 using rapid temperature swing or vacuum swing adsorption. Applied Energy 2013, 104, 418-433. 9. Choi, S.; Drese, J. H.; Jones, C. W., Adsorbent materials for carbon dioxide capture from large anthropogenic point sources. ChemSusChem 2009, 2, (9), 796-854. 10. Drage, T. C.; Snape, C. E.; Stevens, L. A.; Wood, J.; Wang, J.; Cooper, A. I.; Dawson, R.; Guo, X.; Satterley, C.; Irons, R., Materials challenges for the development of solid sorbents for post-combustion carbon capture. Journal of Materials Chemistry 2012, 22, (7), 2815-2823. 11. Samanta, A.; Zhao, A.; Shimizu, G. K. H.; Sarkar, P.; Gupta, R., Post-combustion CO2 capture using solid sorbents: a review. Industrial & Engineering Chemistry Research 2011, 51, (4), 1438-1463. 12. Abanades, J. C.; Rubin, E. S.; Anthony, E. J., Sorbent Cost and Performance in CO2 Capture Systems. Industrial & Engineering Chemistry Research 2004, 43, (13), 3462-3466. 13. Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.; Yaghi, O. M., Control of Pore Size and Functionality in Isoreticular Zeolitic Imidazolate Frameworks and their Carbon Dioxide Selective Capture Properties. Journal of the American Chemical Society 2009, 131, (11), 3875-3877.

ACS Paragon Plus Environment

21

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

14. Nguyen, N. T. T.; Furukawa, H.; Gándara, F.; Nguyen, H. T.; Cordova, K. E.; Yaghi, O. M., Selective Capture of Carbon Dioxide under Humid Conditions by Hydrophobic Chabazite-Type Zeolitic Imidazolate Frameworks. Angewandte Chemie International Edition 2014, 53, (40), 10645-10648. 15. Wang, T. C.; Bury, W.; Gómez-Gualdrón, D. A.; Vermeulen, N. A.; Mondloch, J. E.; Deria, P.; Zhang, K.; Moghadam, P. Z.; Sarjeant, A. A.; Snurr, R. Q.; Stoddart, J. F.; Hupp, J. T.; Farha, O. K., Ultrahigh Surface Area Zirconium MOFs and Insights into the Applicability of the BET Theory. Journal of the American Chemical Society 2015, 137, (10), 3585-3591. 16. Kizzie, A. C.; Wong-Foy, A. G.; Matzger, A. J., Effect of Humidity on the Performance of Microporous Coordination Polymers as Adsorbents for CO2 Capture. Langmuir 2011, 27, (10), 6368-6373. 17. Sevilla, M.; Valle-Vigón, P.; Fuertes, A. B., N-Doped Polypyrrole-Based Porous Carbons for CO2 Capture. Advanced Functional Materials 2011, 21, (14), 2781-2787. 18. Gotzias, A.; Tylianakis, E.; Froudakis, G.; Steriotis, T., Effect of surface functionalities on gas adsorption in microporous carbons: a grand canonical Monte Carlo study. Adsorption 2013, 19, (2-4), 745-756. 19. Dasgupta, T.; Punnathanam, S. N.; Ayappa, K. G., Effect of functional groups on separating carbon dioxide from CO2/N2 gas mixtures using edge functionalized graphene nanoribbons. Chemical Engineering Science 2015, 121, (0), 279-291. 20. Liu, Y.; Wilcox, J., Effects of surface heterogeneity on the adsorption of CO2 in microporous carbons. Environmental science & technology 2012, 46, (3), 1940-1947. 21. Psarras, P.; He, J.; Wilcox, J., Molecular simulations of nitrogen-doped hierarchical carbon adsorbents for post-combustion CO 2 capture. Physical Chemistry Chemical Physics 2016, 18, (41), 28747-28758. 22. Integrated Environmental Control Model (IECM), C. M. U., http://www.iecm-online.com/ Accessed December 1, 2009. 23. Potoff, J. J.; Siepmann, J. I., Vapor–liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen. AIChE journal 2001, 47, (7), 1676-1682. 24. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L., Comparison of simple potential functions for simulating liquid water. The Journal of chemical physics 1983, 79, (2), 926-935. 25. Tenney, C. M.; Lastoskie, C. M., Molecular simulation of carbon dioxide adsorption in chemically and structurally heterogeneous porous carbons. Environmental progress 2006, 25, (4), 343-354. 26. Jorgensen, W. L.; Tirado-Rives, J., The OPLS [optimized potentials for liquid simulations] potential functions for proteins, energy minimizations for crystals of cyclic peptides and crambin. Journal of the American Chemical Society 1988, 110, (6), 1657-1666. 27. Travis, K. P.; Gubbins, K. E., Transport diffusion of oxygen-nitrogen mixtures in graphite pores: A nonequilibrium molecular dynamics (NEMD) study. Langmuir 1999, 15, (18), 60506059. 28. Mark, P.; Nilsson, L., Structure and dynamics of the TIP3P, SPC, and SPC/E water models at 298 K. The Journal of Physical Chemistry A 2001, 105, (43), 9954-9960.

ACS Paragon Plus Environment

22

Page 23 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

29. Huang, H.; Zhang, W.; Liu, D.; Zhong, C., Understanding the effect of trace amount of water on CO2 capture in natural gas upgrading in metal–organic frameworks: a molecular simulation study. Industrial & Engineering Chemistry Research 2012, 51, (30), 10031-10038. 30. Liu, Y.; Wilcox, J., Molecular Simulation Studies of CO2 Adsorption by Carbon Model Compounds for Carbon Capture and Sequestration Applications. Environmental Science & Technology 2013, 47, (1), 95-101. 31. Pels, J. R.; Kapteijn, F.; Moulijn, J. A.; Zhu, Q.; Thomas, K. M., Evolution of nitrogen functionalities in carbonaceous materials during pyrolysis. Carbon 1995, 33, (11), 1641-1653. 32. Gu, J.-M.; Kim, W.-S.; Hwang, Y.-K.; Huh, S., Template-free synthesis of N-doped porous carbons and their gas sorption properties. Carbon 2013, 56, 208-217. 33. Wang, H.; Maiyalagan, T.; Wang, X., Review on Recent Progress in Nitrogen-Doped Graphene: Synthesis, Characterization, and Its Potential Applications. ACS Catalysis 2012, 2, (5), 781-794. 34. He, J.; To, J. W. F.; Psarras, P. C.; Yan, H.; Atkinson, T.; Holmes, R. T.; Nordlund, D.; Bao, Z.; Wilcox, J., Tunable Polyaniline-Based Porous Carbon with Ultrahigh Surface Area for CO2 Capture at Elevated Pressure. Advanced Energy Materials 2016, 6, (14), n/a-n/a. 35. To, J. W. F.; He, J.; Mei, J.; Haghpanah, R.; Chen, Z.; Kurosawa, T.; Chen, S.; Bae, W.-G.; Pan, L.; Tok, J. B. H.; Wilcox, J.; Bao, Z., Hierarchical N-Doped Carbon as CO2 Adsorbent with High CO2 Selectivity from Rationally Designed Polypyrrole Precursor. Journal of the American Chemical Society 2016, 138, (3), 1001-1009. 36. Jin, D.; Lu, X.; Zhang, M.; Wei, S.; Zhu, Q.; Shi, X.; Shao, Y.; Wang, W.; Guo, W., The adsorption behaviour of CH 4 on microporous carbons: effects of surface heterogeneity. Physical Chemistry Chemical Physics 2014, 16, (22), 11037-11046. 37. Liu, Y.; Wilcox, J., Molecular simulation of CO2 adsorption in micro- and mesoporous carbons with surface heterogeneity. International Journal of Coal Geology 2012, 104, (0), 83-95. 38. Do, D. D.; Do, H. D., Evaluation of 1-Site and 5-Site Models of Methane on Its Adsorption on Graphite and in Graphitic Slit Pores. The Journal of Physical Chemistry B 2005, 109, (41), 19288-19295. 39. Kresse, G.; Furthmüller, J., Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Physical Review B 1996, 54, (16), 11169. 40. Grimme, S., Semiempirical GGA-type density functional constructed with a long-range dispersion correction. Journal of computational chemistry 2006, 27, (15), 1787-1799. 41. Kresse, G.; Joubert, D., From ultrasoft pseudopotentials to the projector augmentedwave method. Physical Review B 1999, 59, (3), 1758. 42. Perdew, J.; Burke, K.; Ernzerhof, M., Generalized Gradient Approximation Made Simple. Physical Review Letters 1996, 77, (18), 3865-3868. 43. Monkhorst, H.; Pack, J., Special points for Brillouin-zone integrations. Physical Review B 1976, 13, (12), 5188-5192. 44. Bader, R., Atoms in Molecules - A Quantum Theory. Oxford University Press: Oxford, 1990. 45. Gupta, A.; Chempath, S.; Sanborn, M. J.; Clark, L. A.; Snurr, R. Q., Object-oriented programming paradigms for molecular modeling. Molecular Simulation 2003, 29, (1), 29-46. 46. Steele, W. A., The physical interaction of gases with crystalline solids: I. Gas-solid energies and properties of isolated adsorbed atoms. Surface Science 1973, 36, (1), 317-352.

ACS Paragon Plus Environment

23

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

47. Allen, M. P.; Tildesley, D. J., Computer simulation of liquids. Oxford university press: 1989. 48. Young, P. D.; Notestein, J. M., The role of amine surface density in carbon dioxide adsorption on functionalized mixed oxide surfaces. ChemSusChem 2011, 4, (11), 1671-1678. 49. Presto, A. A.; Granite, E. J.; Karash, A., Further investigation of the impact of sulfur oxides on mercury capture by activated carbon. Industrial & Engineering Chemistry Research 2007, 46, (24), 8273-8276. 50. Presto, A. A.; Granite, E. J., Impact of sulfur oxides on mercury capture by activated carbon. Environmental science & technology 2007, 41, (18), 6579-6584. 51. Granite, E. J.; Pennline, H. W., Photochemical removal of mercury from flue gas. Industrial & Engineering Chemistry Research 2002, 41, (22), 5470-5476. 52. White, C. M.; Strazisar, B. R.; Granite, E. J.; Hoffman, J. S.; Pennline, H. W., Separation and capture of CO2 from large stationary sources and sequestration in geological formations— coalbeds and deep saline aquifers. Journal of the Air & Waste Management Association 2003, 53, (6), 645-715. -

ACS Paragon Plus Environment

24

Page 25 of 25

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

25