Molecular Modeling Analysis of CO2 Absorption by Glymes - The

1 hour ago - The properties of diglyme + CO2 systems were analysed through density functional theory and molecular dynamics methods with the objective...
2 downloads 17 Views 1MB Size
Subscriber access provided by READING UNIV

Article 2

Molecular Modeling Analysis of CO Absorption by Glymes Alberto Gutiérrez, Mert Atilhan, and Santiago Aparicio J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.7b10276 • Publication Date (Web): 29 Jan 2018 Downloaded from http://pubs.acs.org on January 31, 2018

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.

The Journal of Physical Chemistry B 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 33 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

The Journal of Physical Chemistry

Molecular Modeling Analysis of CO2 Absorption by Glymes Alberto Gutiérrez,a Mert Atilhan,*b,c and Santiago Aparicio*a

a b

Department of Chemistry, University of Burgos, 09001 Burgos, Spain

Department of Chemical Engineering, Texas A&M University at Qatar, Doha, Qatar c

Gas and Fuels Research Center, Texas A&M University, College Station, TX, USA

*

Corresponding authors: [email protected] (M. A.) and [email protected] (S. A.)

ABSTRACT: The properties of diglyme + CO2 systems were analysed through density functional theory and molecular dynamics methods with the objective of inferring the microscopic properties of CO2 capturing by glyme-based solvents and the effect of ether group regarding solvents affinity toward CO2. Calculations of diglyme + CO2 molecular clusters using density functional theory allowed to accurately quantify and characterize short range intermolecular forces between these molecules, whereas molecular dynamics simulation of diglyme + CO2 liquid mixtures, for different CO2 contents, allowed to infer the properties and dynamics of bulk liquid phases upon CO2 absorption. Likewise, liquid diglyme + CO2 gas interfaces were also studied using molecular dynamics methods in order to study the kinetics of CO2 capture, adsorption at the gas-liquid interface and the mechanism of interface crossing, which is of pivotal importance for the design of CO2 capturing units.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Introduction The problem of increasing atmospheric levels of CO2

1,2

and its effects on climate change3

due to excessive utilization of fossil based fuels,4 requires the development of new, sustainable and effective carbon capture technologies in chemical process industries.5,6 Emissions from fossil-fuelled power plants are among the largest CO2 sources,7,8 and thus, developing suitable sorption technologies would contribute to a decrease and controlling of CO2 atmospheric levels and it can be considered as one of the pivotal challenge that chemical sciences facing for a while.9 Therefore, the development of economically and technologically viable carbon capture technologies is required in the transition toward a fully decarbonized electricity production framework,10,11 which may last for decades. The development a sustainable and economically viable CO2 capture technologies at industrial scale for power plants must consider several engineering challenges due to the large scale of the issue.12 There is a growing technical consensus that the main drawback stands on the absence of suitable high-performance sorbents (either physical or chemical),13 which can perform the absorption and desorption steps favourable enough in comparison to current state of the art method(s) both technically and economically. From the technological viewpoint, the main challenges of developing such materials for CO2 capture in fossil-fuel based power plants stand on the proper characteristics of flue gases, low CO2 partial pressures,14 which require highly efficient sorbents to avoid increasing costs of electricity production to unfeasible levels.15,16 Therefore, many materials have been studied this last decade considering very different chemical and physical characteristics. Traditional aminebased CO2 capturing methods17-20 have been employed in the natural gas industry with success for many decades, but the characteristics of natural gas treatment and those of CO2 capturing from flue gases are remarkably different, different pressure ranges and CO2 concentrations, which are also combined with high corrosion effect,21-23 high solvent degradation,24 and high regeneration costs25-29 lead to huge increase in cost of electricity production when using amine-based carbon capture units. Hence, the use of materials such as ionic liquids,30 deep eutectic solvents,31,32 membranes,33 metal organic frameworks,34 covalent organic frameworks,35 and many other types of sorbents have been studied and combined with different technological approaches.36-39

2

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33 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

The Journal of Physical Chemistry

Among the studied liquid state CO2 absorbents that work with physisorption,37,40 glymes (glycol dieters) showed large affinity for CO2 molecules leading to high absorption performances.41 Although the enhancement of CO2 absorption by oxygenation of alkyl chains is a known effect,42 the mechanism of interaction between glymes and CO2 molecules is still not understood, especially at the microscopic level. Therefore, in order to obtain a microscopic knowledge of the mechanism of CO2 capture by glyme– based physical absorbents, a computational chemistry study was carried out in this work. Diglyme, bis(2methoxyethyl) ether (2G), was considered as a model system of glymes and their properties regarding CO2 absorption were studied. In the first stage, Density Functional Theory (DFT) calculations on 2G + CO2 molecular clusters were carried out to obtain an accurate characterization of glymes – CO2 intermolecular interactions. However, the reported DFT calculations in gas phase lack to infer additional effects in bulk liquid phases, such as larger molecular clusters, development of solvation shells or molecular packing effects. Thus, molecular dynamics (MD) simulations were carried out. The study of liquid phase properties via MD was done considering 2G + CO2 liquid mixtures for different mixture compositions, i.e. increasing CO2 mole fraction, according to experimental compositions reported by Kodama et al.41 The analysis of intermolecular forces, molecular arrangements and distributions, and dynamic properties allow characterizing the effect of CO2 on 2G properties and thus the CO2 capture mechanism as well. Likewise, to infer the kinetics of CO2 capturing at microscopic level, MD simulations on liquid 2G in contact with CO2 gas phases was carried out, inferring the properties at liquid-gas interface (i.e. CO2 adsorption), diffusion across the interface and mass transfer form gas to liquid phases. The reported computational study herein focuses on the microscopic characterization of glymes – CO2 systems, analyzing the effect of ether-functionalization on CO2 capture and contributes to the development of suitable solvents for effective CO2 capture.

Methods The modelling of 2G-CO2 intermolecular interactions was carried using molecular clusters with one 2G molecule and n (=1,2 or 3) CO2 molecules, which were optimized using B3LYP43,44,45 functional, with Grime’s method (DFT-D3) for treating dispersion interactions,46 and 6-311++G(d,p) basis set. For the calculation of interaction energies, ΔE, counterpoise 3

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

procedure was applied for correcting Basis Set Superposition Error (BSSE).47 All DFT calculations were carried out by using ORCA program.48 DFT calculations were carried out only at B3LYP-D3/6-311++G(d,p) theoretical level and no comparison was carried out with higher level approaches such as MP2, nevertheless, previous studies have showed that differences for ΔE between B3LYP with dispersion corrections and MP2 methods are lower than 0.5 kcal mol-1,49 and thus, results reported in this work are a reasonably accurate approach to ΔE50,51,52 for the studied systems. Intermolecular interactions were also analyzed based on Bader’s theory (Atoms in a Molecule, AIM, method) using the Multiwfn code.53 MD simulations were carried out using the MDynaMix v.5.2 program.54 The forcefield parameterization for 2G was previously reported by our group55 whereas for CO2 was obtained from the literature.56 Two different types of systems were considered for studying the behaviour of 2G + CO2 by using MD: i) liquid mixtures of 2G + CO2 with different compositions, to study the mechanism of interaction, intermolecular forces and other effects controlling 2G affinity for CO2 molecules; ii) liquid 2G in contact with a CO2 gas phase, to study the kinetics of CO2 absorption, adsorption of CO2 molecules at 2G liquid interface and other interfacial phenomena controlling CO2 capturing by 2G. 2G + CO2 mixtures were modelled using the systems reported in Table 1, considering a fixed number of 2G molecules and different contents of CO2 to mimic experimental absorption data by Kodama et al.41 Simulation boxes were built using Packmol57 program. MD simulations were developed in two stages: i) initial equilibration (10 ns); followed by ii) production (10 ns) runs. The MD simulations were carried out in a NPT ensemble at 313 K and pressures reported in Table 1. Nose–Hoover method was used for pressure and temperature controlling, Ewald method58 (15 Å for cut-off radius) for treating coulombic interactions, Lorentz-Berthelot mixing rules for Lennard-Jones cross terms Tuckerman–Berne double time step algorithm59 (1 and 0.1 fs for long and short time steps) for solving equations of motion. Periodic boundary conditions were applied. Regarding simulations of 2G – CO2 systems, a 2G liquid phase was put in contact (i.e. developing an interface) with a gas phase containing 2000 CO2 molecules, both phases previously equilibrated at 313 K. Simulations as function of time (10 ns long) were carried out in the NVT ensemble, with the same methods as for the mentioned simulations for 2G + CO2 mixtures.

4

ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33 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

The Journal of Physical Chemistry

Results and discussion The characterization of 2G-CO2 intermolecular forces carried out using DFT was done for 2G + n (=1,2,3) CO2 molecular clusters, at different interaction sites. The presence of ether group is considered to enhance the affinity toward CO2 molecules, and thus several configurations were considered with CO2 molecules placed close to oxygen atoms in 2G, Figure 1. Two different types of oxygen atoms may be considered in glymes: those placed at the end of the chains (O1), between a methylene and a terminal methyl group, and those in the chain between two methylene groups (O2). Therefore, the interaction of CO2 molecules with these two types of oxygen atoms was studied. In the case of a single CO2 molecules interacting with 2G, results that are presented in Figure 1 show that larger binding energies for interactions through O1 than through O2 sites, although the low values for both sites are indicative of van der Waals interactions of moderate strength. Interaction through O2 site is characterized by CO2 molecules adopting a parallel arrangement to 2G chains with carbon atom on top of oxygen atom (at 3.07 Å) and oxygen atoms on top of the two-methylene carbons neighbouring O2 site. In the case of O1 site, the distance between carbon atom in CO2 and the O1 site is 2.75 Å, which is lower than for the interactions through O2 site, which would justify the larger binding energies through O1 than through O2. Likewise, the interaction through the O1 site is characterized by the interaction between the hydrogen atom in the terminal methyl group and oxygen atom in CO2 which is stronger than the interaction with the methylene hydrogens when CO2 interacts through the O2 site. This is confirmed by the bond paths reported in the AIM analysis (Figure S1, Supporting Information). Additional details of AIM analysis for intermolecular interactions are reported in the Supporting Information. Larger clusters were considered to infer the effect of several CO2 molecules competing and interacting with a single 2G molecule, Figure 1. The interaction of two CO2 molecules with one 2G molecule leads to binding energies as the sum of those corresponding to single interactions, but when one of the molecules is out of the main interaction sites (O1 and O2) binding energy decreases remarkably, as reported in Figure 1. Regarding larger number of CO2 molecules, results in Figure 1 confirm additive binding energies and no perturbation on interactions between neighbour sites. Therefore, each 2G molecule can fit one CO2 molecule per O1 or O2 site, confirming the sorption capacity experimentally reported by Kodama et al.41

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

The characterization of bulk 2G liquid phases upon CO2 absorption is also analysed in this work. The absorption of CO2 may be accompanied by solvent swelling, i.e. volume expansion upon gas sorption, %Vexp, which was quantified according to method propsed by Gallager et al.60 and it is reported in Figure 2. The large %Vexp reported in Figure 2 should be analysed considering the fact that the high CO2 concentrations in 2G, which was also confirmed by Kodama et al.41 This expansion show disruption of 2G structuring upon gas capturing. CO2 concentrations lower than 0.6 yielded percentage volume expansion of 40 %, which could be considered as reasonable regarding to the solvents’ possible application for industrial scale applications. The 2G + CO2 mixtures showed a non-ideal behaviour as the evolution of density with composition, as shown in Figure 3. When the density profile is examined, there is a region at which sudden increase take place between 0 – 0.2 CO2 more fraction. For CO2 concentrations between 0.2 and 0.6, there is a transitional density region observed and there is a linear evolution in the density for densities that are higher mole fractions than that of 0.6; which are in agreement with the behaviour of %Vexp. Therefore, three regions in the behaviour of 2G + CO2 mixtures are inferred, which is confirmed by the evolution dielectric constant, ε. The used force field parameterizations are able to predict ε accurately for both pure 2G and CO2; 6.886 for 2G at 318.15 K was reported experimentally by Lago et al.61, which was obtained as 6.1 as a result of MD simulations at 313 K (Figure 3). Yet, experimental ε values were reported close to 1 at 313.15 K were for pure CO262, which is also in agreement with values for CO2-rich mixtures as presented in Figure 3. One of the main origins on CO2 capturing by liquid sorbents stands on the rearrangement of available empty molecular space or cavities to fit the absorbed molecules. The spontaneous rearrangement of fluids cavities has been considered as a pivotal factor for CO2 capture in fluids such as ionic liquids,63,64 but there are not enough scientific studies that handles 2G for the same purpose in molecular level. In order to analyse the distribution of cavities in pure 2G and in 2G + CO2 mixtures, the method studied by Margulis65 was considered; and to do that, a network of random points was distributed in simulation boxes and the distance between each point and the edge of the nearest atom (defined by its van der Waals radius) is considered as the size of the cavity defined as spheres. Probabilities distributions of cavities are reported in Figure 4 for pure 2G and for mixtures with different CO2 mole fractions. These results show that a rearrangement of 2G spherical cavities is 6

ACS Paragon Plus Environment

Page 6 of 33

Page 7 of 33 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

The Journal of Physical Chemistry

produced upon CO2 absorption, with a decrease of cavities sizes, in agreement with the increase of density reported in Figure 3. The molecular arrangements at the molecular level were analysed using radial distribution functions (RDFs), Figure 5. Pure 2G is a largely structured fluid and its center-ofmass RDF reported in Figure 5a shows a first and narrow solvation peak (with maximum at 4.6 Å) which is followed by three additional peaks that are wider and less intense than the first one and separated roughly 4 Å. This structuring is maintained upon CO2 absorption, with a slight decrease in the intensity of RDFs peaks because of dilution of 2G molecules, and thus showing the strong trend of 2G molecules to self-associate. Regarding the behaviour of CO2 molecules when absorbed in 2G, results in Figure 5b also shows trend to self-associate as the presence of at least two well defined RDFs peaks shows. RDFs for 2G-CO2 pairs, Figure 5c, show complex behaviour, five RDFs peaks are obtained for distances lower than 20 Å, with their maxima remaining constant upon increase of CO2 concentration. Nevertheless, the intensity of RDFs for 2G-CO2 pairs is remarkably lower than those for 2G-2G or CO2-CO2 pairs, which confirm the large trend of both 2G and CO2 to self-associate. The behaviour of RDFs shows the existence of microdomains of 2G and CO2 molecules, with some regions allowing the development of 2G-CO2 interactions but with the trend to self-associate being the pivotal role in the structuring of these mixed fluids. The formation of microdomains have also been reported for other types of CO2 absorbents, such as ionic liquids,66,67 although in this case the origin of these microdomains use to be produced by the presence of large apolar chains with self-aggregation trends leading to apolar microdomains68 whereas in the case of 2G microdomains are produced even with the presence of polar ether group because of the trend of 2G molecules to develop parallel arrangements for improving van der Waals interactions. The number of molecules in the solvation shells obtained from RDFs is reported in Figure 6, regarding CO2 molecules around 2G (Figures 6a and 6b) a non-linear evolution with mixture composition is inferred, with three well-defined regions corresponding to those for macroscopic properties in Figures 2 and 4. The number of CO2 molecules in the first solvation shell (with radius of 4.2 Å, Figure 5c) is almost constant for 2G-rich mixtures (x(CO2)