Mercury Capture from Petroleum Using Deep Eutectic Solvents

Subscriber access provided by Kaohsiung Medical University

Separations

Mercury Capture from Petroleum Using Deep Eutectic Solvents Samah Warrag, Evgenii O. Fetisov, Dannie van Osch, David B. Harwood, Maaike C. Kroon, J. Ilja Siepmann, and Cor J. Peters Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b00967 • Publication Date (Web): 06 Jun 2018 Downloaded from http://pubs.acs.org on June 6, 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 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 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.

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 26 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

Mercury Capture from Petroleum Using Deep Eutectic Solvents Samah E. E. Warrag,†,‡ Evgenii O. Fetisov,¶ Dannie J.G.P. van Osch,‡ David B. Harwood,¶ Maaike C. Kroon,†,‡ J. Ilja Siepmann,∗,¶,§ and Cor J. Peters∗,† Department of Chemical Engineering, The Petroleum Institute, Khalifa University of Science and Technology, P.O. Box 2533, Abu Dhabi, United Arab Emirates, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology, P. O. Box 513, 5600 MB Eindhoven, The Netherlands, Department of Chemistry and Chemical Theory Center, University of Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455-0431, United States, and Department of Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Avenue SE, Minneapolis, Minnesota 55455-0132, United States E-mail: [email protected]; [email protected]

Abstract Mercury capture is a major challenge in petroleum and natural gas processing. Recently, ionic liquids (ILs) have been introduced as mercury extractants from oil and gas. ILs yield very high mercury extraction efficiencies (> 95%) from hydrocarbons, but their drawbacks include complex synthesis, toxicity, and difficult regeneration after mercury capture. In this ∗ To

whom correspondence should be addressed Petroleum Institute ‡ Eindhoven University of Technology ¶ Chemistry, University of Minnesota § Chemical Engineering and Materials Science, University of Minnesota † The

1

ACS Paragon Plus Environment

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

work, a new technology using deep eutectic solvents (DESs) for elemental mercury (Hg0 ) extraction from hydrocarbons is demonstrated. DESs are an innovative class of designer solvents exhibiting similar properties as ILs, such as low vapor pressure and low flammability, but DESs are formed from inexpensive hydrogen-bond donor and acceptor compounds that are often biodegradable. In this work, four DESs were investigated including choline chloride:urea, choline chloride:ethylene glycol, choline chloride:levulinic acid, and betaine:levulinic acid, where the molar ratio is 1:2 in all cases. The DESs were tested for their thermal stability, density, and viscosity. Their performance for mercury extraction was assessed using saturated solutions in n-dodecane as the model oil. It was found that solvent to feed ratios of 1:1 and 2:1 at temperatures of 303.15 and 333.15 K and atmospheric pressure yield extraction efficiencies greater than 80% for all four DESs. First principles molecular dynamics simulations probing the solvation in choline chloride:urea indicate a tight first coordination shell for mercury. Calculation of the Hg–Hg potential of mean force supports formation of a mercury–mercury polycation for a pair of Hg1+ ions, but not for pairs of Hg0 and Hg2+ species. Geometric analysis of the speciation and Mulliken population analysis support a redox reaction involving Hg2+ + 2 Cl− .

Introduction Energy efficient, economically feasible and environmentally friendly processes for the production of fuels and chemical feedstocks are highly desirable. In particular, carbon dioxide capture and natural gas sweetening technologies have garnered significant research efforts over the past two decades. 1–3 On the other hand, comparatively little effort has been directed towards mercury (Hg) capture due to its very small concentration that varies between 0.01 ppb and 10 ppm, depending on the geological location of the oil/gas reservoir. 4 Mercury in crude oil/natural gas is present in different toxic species: elemental mercury (Hg0 ) is most prevalent, but mercuric halides (mostly, HgCl2 ), organic mercury compounds (RHgR’ and RHgCl) and mercury-sulfur complexes can also be found. 4 Beyond mercury’s health and safety risks for the biosphere, mercury is also a major

2


Page 2 of 26

Page 3 of 26 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


problem in oil and gas processing units as it deposits in the cryogenic units forming amalgams with different metals (e.g., aluminum) that lead to equipment degradation, and it poisons many catalysts. 4 Additionally, Hg emissions are a major environmental concern and are classified as hazardous air pollutants. 5 Owing to mercury’s adverse environmental effects, as well as the operational issues in the oil and gas processing industry, it is rendered mandatory to develop an efficient removal process. Several technologies are commercially available for mercury capture from liquid/gaseous hydrocarbon streams based on either amalgamation, 6 physical adsorption, 7 chemical adsorption or reactive absorption. 8 The most mature technologies are adsorption on activated carbon and on sulfur/transition metal sulfides impregnated on a solid support, such as activated carbon, alumina, zeolite or silica. 8–10 Due to the sensitivity of sulfur to moisture in organic systems, the latter is less suitable for application in liquid streams. 9 Some other technologies employ regenerative molecular sieves impregnated with silver, but this is an expensive option compared to activated carbon beds. 9 Recently, Clariant and Petronas announced the commercialization of a new solid-supported ionic liquid (SSIL) mercury removal technology. 11 SSILs containing chlorocuprate(II) effectively remove elemental, organic and inorganic mercury from natural gas. 12 Mancini et al. 13 studied the mechanism of the removal of mercury ions from aqueous solutions using hydrophobic/Clcontaining ILs in the absence of chelating agent and suggested that Hg2+ ions are transferred to the IL phase through the formation of polyanion species HgCl−n+2 (where n is 1 to 4) and then n extraction by the IL. Cheng et al. 14 evaluated the extraction of Hg0 from flue gas using 1-alkyl3-methylimidazolium chloride ILs and identified the formation of a complex [IL cation]HgCl3 on a solid adsorbent by means of Raman and UV-Vis spectroscopies. Other ILs also show promise as mercury capture technology, 12,15–17 but ILs are a relatively expensive class of solvents, and economical regeneration approaches have not yet been reported. In 2003, Abbott et al. 18 introduced an innovative class of solvents, the so-called deep eutectic solvents (DESs), which could be potential alternatives to ILs. DESs consist of at least one compound acting predominantly as hydrogen-bond donor (HBD) and another compound acting pre-

3



dominantly as hydrogen-bond acceptor (HBA) that form a mixture exhibiting a significantly lower freezing point than both of the pure compounds. DESs have similar properties as ILs (including low vapor pressure, wide liquid range, and low flammability), but are generally much easier to prepare from low-cost HBD and HBA ingredients that that are mostly biodegradable. 18–22 DESs have already been utilized in a number of separations relevant to oil and gas industries, such as desulfurization, dearomatization, and sweetening, 23 and also for gas separation 24 and water reclamation. 25 Additionally, the application of DES-functionalized carbon nanotube adsorbents for mercury removal from water has been studied. 26 Abai et al. also investigated one chlorocuprate(II) DES, choline chloride – ethylene glycol – CuCl2 ·H2 O, and found similar extraction properties to those of chlorocuprate(II) containing ILs used for the SSIL technology. 12 Based on the observed color change, however, it is likely that the Cu-Hg redox pair also plays an important role in the dissolution with the chlorocuprate(II) DES. 12 In this work, the use of DESs as extracting agents for the removal of mercury from Hgcontaining liquid hydrocarbons is reported. The selection of an extraction solvent for mercury from hydrocarbon liquids depends on four factors: (i) its strong affinity for solvating various mercury species, (ii) its low mutual solubility with hydrocarbons, (iii) its thermal stability, and (iv) its regenerability. Here, only the first three factors are addressed. Given the low vapor pressure of the DESs, a (partial) regeneration via vaporization of elemental mercury may be feasible. DESs are very polar and, hence, their mutual solubility with aliphatic hydrocarbons is expected to be very low. 27 n-Dodecane was selected to represent the aliphatic hydrocarbons in petroleum. It is known that halogen- (particularly, Cl) and nitrogen-containing ILs exhibit excellent extraction efficiency for mercury 17 and, hence, salt based/polar DESs are good candidates for this application. The selected DESs were (i) choline chloride:urea (DES-1), (ii) choline chloride:ethylene glycol (DES-2), (iii) choline chloride:levulinic acid (DES-3), and (iv) betaine:levulinic acid (DES-4), where the molar ratio is 1:2 in all four cases. DES-4 was chosen to test the influence of replacing a salt-based HBA with a zwitter-ionic HBA. These DESs were tested for their thermal stability by observing their degradation, and for their transport properties by determining glass transition tem-

4


Page 4 of 26

Page 5 of 26 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


Table 1: Chemicals Used in This Work Chemical Choline chloride Urea Ethylene glycol Levulinic acid Betaine Dodecane Mercury

Purity (wt %) ≥ 98 ≥ 98 ≥ 99.8 ≥ 98 ≥ 98 ≥ 99 Extra pure

Source Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich Merck Merck

peratures and measuring viscosities. DES-1 and DES-2 are widely used and their densities (1.25 g cm−3 and 1.12 g cm−3 at T = 298 K, respectively) and viscosities (750 mPa s and 37 mPa s at T = 298 K, respectively) have been reported previously. 28–31 The densities and viscosities of DES-3 and DES-4 were determined in this work at temperatures from 298.15 to 333.15 K and atmospheric pressure. The extraction efficiency for the system [n-dodecane + Hg0 + DES] was determined by direct solvent-feed extraction in ratios of 1:1 and 2:1 at T = 303.15 and 333.15 K and atmospheric pressure. These experimental operating conditions were chosen based on the conventional processing temperatures for the Hg adsorbent beds. 9 First principles molecular dynamic simulations were used to obtain molecular-level information on the mercury solvation and to probe charge transfer and potential redox reactions.

Experimental Section Materials and DES Preparation The chemical compounds used in this work, along with their sources and purities, are reported in Table 1. The choline chloride was dried under vacuum prior to use. The other chemicals were used as obtained. The molecular structures of the constituents for the four DESs are provided in Scheme 1. Here, the DESs were prepared in 50 g batches using a 1:2 molar ratio for HBA:HBD. The constituents were weighed using a Sartorius ED 224S analytical balance with a precision of ±0.1 mg, then 5



Page 6 of 26

added together in a closed 100 mL glass bottles and mixed thoroughly using a Vortex mixer (VWR). The mixtures were stirred at 323.15 K in a temperature controlled oil bath with a temperature controller (IKA ETS-D5, uncertainty = ±0.1 K), until a homogeneous clear liquid was formed. Three DESs were prepared with choline chloride as the HBA and the HBDs were urea, ethylene glycol, and levulinic acid. A fourth DES was prepared from betaine and levulinic acid. HBA

HBD O

N+

HO

H 2N NH 2 urea

Cl-

choline chloride

OH HO ethylene glycol

O N+ betaine

O O-

CH 3

HO

O levulinic acid

Scheme 1: Molecular Structures of the Constituents for the DESs Investigated in This Work

DES Characterization The water content, w, of each DES was determined using the Karl Fischer titration method (899 Coulometer, Metrohm Karl Fischer). Three measurements were carried out for each DES to obtain its water content. The specific density, ρ, and the dynamic viscosity, η, of DES-3 and DES-4 were measured over a temperature range from 298.15 to 333.15 K in steps of 5 K at atmospheric pressure using an Anton Paar SVM 3000 Stabinger Viscometer, with an instrumental uncertainty of ±0.0005 g cm−3 for the specific density, relative uncertainty of ±0.35% for the dynamic viscosity, and ±0.01 K for the temperature. The degradation temperatures, Tdeg , for each DES were obtained using a thermogravimetric analyzer (TGA 4000, PerkinElmer). The weighing precision 6


Page 7 of 26 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


and sensitivity of the balance are ±0.01% and 1 µg, respectively. The instrumental uncertainty in temperature is ±1 K. The thermographs of the DESs (9-10 mg each) were obtained using the following heating protocol: First, the sample was heated from ambient temperature to 313 K at heating rate of 5 K min−1 in a ceramic crucible under a continuous nitrogen flow of 20 cm3 min−1 and a gas pressure of 0.2 MPa. Then, the sample was held for 120 min at 313 K. Finally, the change in mass was scanned by heating from 313 K to 673 K at a rate of 5 K min−1 . Two thermographs were obtained for each DES. Differential scanning calorimetry (TA Instruments, DSC Q100) was used to determine the glass transition temperature, Tg , of the DES with a scanning rate of 5 K min−1 and temperature range from 193.15 K to 303.15 K. The instrumental uncertainty in T is ±0.1 K. The calorimeter precision and sensitivity are ±0.1% and 0.2 µW, respectively. The structure of the DESs was verified by 400 Bruker nuclear magnetic resonance (NMR) spectrometer for proton and carbon NMR (Figures S1 to S8).

Mercury Extraction Experiments 25 mL of n-dodecane (> 99% purity) was saturated with elemental mercury (extra pure) at ambient conditions to a concentration of approximately 4000 µg kg−1 . The saturated n-dodecane solution was added to the DESs using both a 1:1 and 2:1 solvent-to-feed mass ratios. The two-phase systems were initially mixed for a short time using a Vortex mixer followed by shaking the solutions for 2 h using an incubating shaker (IKA KS 4000i) at temperatures of 303.15 and 333.15 K. The mixtures were left to settle for 30 min until liquid–liquid coexistence was visually observed with the ndodecane and DES being the upper and lower phases, respectively. A sample from the n-dodecane phase was taken using a syringe without disturbing the equilibrium interface. The n-dodecane sample was then analyzed for its mercury content using a Milestone Direct Mercury Analyzer DMA-80 pyrolysis/AA analyzer. A sample of the n-dodecane phase (20-30 mg) was introduced in the DMA-80, in which the sample was initially dried at T = 573 K and then thermally decomposed at T = 1123 K in an oxygen flow (200 mL min−1 ) and a gas pressure of 4 bar. The combustion products were carried off and further decomposed in a hot catalyst bed at T = 873 K. The mercury 7



vapors were trapped on a gold amalgamator and subsequently desorbed at T = 1173 K. Finally, the mercury content is determined using atomic absorption spectrophotometry at 254 nm.

Computational Methods First principles molecular dynamics (FPMD) simulations were used to provide molecular-level information on the mercury solvation. Due to the significant expense of FPMD, only DES-1 was investigated at an elevated temperature of 363.15 K to improve the sampling. Three different systems containing 8 choline chloride formula units and 16 urea molecules as the solvent and either 2 Hg atoms, 2 HgCl formula units, or 2 HgCl2 formula units were studied to probe the effects of Hg speciation. The systems were initialized using Monte Carlo simulations with the modified AMBER 32 force field for the solvent and the Universal force field for mercury 33 to obtain initial configurations for the FPMD simulations. Two initial configurations were generated for each system: “short" and “long" with the initial Hg–Hg distances being 2.6 and 5.0 Å, respectively. The Monte Carlo simulations consisted of approximately 3.6 million Monte Carlo moves using translations of the atomic species and translations and rotations for rigid molecular species (choline and urea). The simulations were carried out in the canonical ensemble with the box volumes determined from the molecular volume of DES-1 34 and the van der Waals or ionic volumes of the mercury species. The resulting box volumes for the Hg, HgCl, and HgCl2 systems were 3018 Å3 , 3047 Å3 , and 3094 Å3 , respectively. All FPMD simulations were performed with the CP2K simulation package, 35 which solves the Kohn-Sham formulation of density functional theory with the Gaussian plane wave method. 36 The BLYP functional 37,38 with the third-generation dispersion correction (D3) of Grimme 39 was used along with a triple-zeta, double polarization basis set 40 for non-metal atoms, double-zeta MOLOPT basis set for mercury atoms, 40 and GTH pseudopotentials. 41,42 The plane wave cut-off was set to 400 Ry. A time step of 0.5 fs was used, and the temperature was controlled using massive Nosé–Hoover chain thermostats. 43,44 All systems were equilibrated for 40 ps in the canonical ensemble at the box volumes mentioned above. Thereafter, at least 60 ps of production trajecto8


Page 8 of 26

Page 9 of 26 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


ries were generated for subsequent analysis. Configurations for structural analysis were collected during production runs every ten time steps. Additional simulations in the canonical ensemble with umbrella sampling of Hg–Hg distances were performed for the three systems to calculate the potential of mean force (PMF). The Hg–Hg distances ranged from 2.6 to 7.1 Å with windows equally spaced by 0.3 Å. A harmonic potential with a force constant of 400 kJ mol−1 Å−2 was used to constrain the Hg–Hg distance. In each window, the system was equilibrated for 5 ps, and statistics for the PMF were collected over at least 30 ps. To extract the Helmholtz free energy, the weighted histogram analysis method was used. 45 To provide additional information on charge transfer and redox reactions, a Mulliken population analysis was carried out for the umbrella region corresponding to the minimum in the PMF for each of the three systems, but caution is needed in interpretation of the partial atomic charges obtained from periodic calculations and for liquid phases with large fluctuations in local structure.

Results and Discussion DES Characterization Figure 1 illustrates the temperature dependencies (298.15 K ≤ T ≤ 333.15 K) of the specific density, ρ, and of the dynamic viscosity, µ, for DES-3 and DES-4 (the numerical data are provided in the Supporting Information, Table S1). Over this 35 K temperature range, the specific densities are well described by linear fits as follows: DES − 3 : ρ = (−0.652 ± 0.002) [kg m−3 K−1 ] T + (1331.2 ± 0.5) [kg m−3 ]

(1)

DES − 4 : ρ = (−0.723 ± 0.004) [kg m−3 K−1 ] T + (1373.4 ± 1.1) [kg m−3 ]

(2)

The standard deviations of the experimental data from the corresponding linear fits are 0.124 and 0.116 kg m−3 for DES-3 and DES-4, respectively. The viscosity is a key property for selecting a proper absorbent candidate because it is related 9



1.16

-3

ρ [g cm ]

1.15

1.14

1.13

1.12

1.11

1200

µ [MPa s]

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 26

900

600

300

0

300

310

320

330

T [K]

Figure 1: Experimental data for the specific density (top) and the dynamic viscosity (bottom) for DES-3 (triangles) and DES-4 (circles) as function of temperature. The dashed lines show linear fits for the specific density and VFT fits for the dynamic viscosity. The statistical uncertainties are smaller than the symbol size. to the mass transfer rate of the solute from the feed to the solvent. As illustrated in Figure 1, the viscosities of DES-3 and DES-4 decrease sharply over the temperature range investigated here. Although the two DESs share a common HBD compound, DES-4 with the zwitter-ionic HBA yields µ values exceeding those of DES-3 with the ionic HBA by factors ranging from 2.5 at 333.15 K to 4.8 at 298.15 K. The Vogel–Fulcher–Tammann (VFT) equation is widely used to describe the temperature dependence of the dynamic viscosity for DESs. 46,47 Here we also observe a good description of the µ values by the VFT equation as follows:

(857 ± 24) [K] DES − 3 : µ = (0.126 ± 0.012) [mPa s] exp T − (186 ± 2) [K]

10


(3)

Page 11 of 26 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


Table 2: Water Content (w), Degradation Temperature (Tdeg ), and glass transition temperature (Tg ) of the four DESs DES w [ppm] Tdeg [K] Tg [K] DES-1 1510 ± 20 455.4 ± 0.5 205.15 DES-2 1870 ± 20 386.0 ± 1.5 245.15 DES-3 7150 ± 40 445.1 ± 1.3 243.15 DES-4 6690 ± 40 443.4 ± 0.1 214.15

(1078 ± 20) [K] DES − 4 : µ = (0.059 ± 0.006) [mPa s] exp T − (190 ± 1) [K]

(4)

The standard deviations of the experimental data from the corresponding VFT fits are 0.042 and 0.282 mPa s for DES-3 and DES-4, respectively. Knowledge of the thermal stability range is important to assess the viability of an extraction solvent. The thermographs for the four DESs are depicted in Figure 2. As can be seen, the thermal degradation proceeds in two major steps for DES-2, DES-3, and DES-4, but only one broad step for DES-1. Conventionally, the degradation temperature, Tdeg , is determined from the intersection of the extrapolated tangent where the weight loss starts (the first derivative equals to zero) and the slope of the weight loss at the inflection point (the peak of the first derivative). 48 The same method was applied here using the inflection point for the first step in the thermographs, and the results are presented in Table 2. Since any absorbed water is likely to evaporate first as the heating protocol is started, the water content was also determined and found to range from 1510 ppm for DES-1 to 7150 ppm for DES-3. The thermograph for DES-2 showed a 3% decrease in weight after keeping the sample at T = 313.15 K for 2 h, which can only be attributed to the evaporation of volatile impurities because the moisture content was low. DES-2 also is thermally least stable with Tdeg = 386 K, whereas the Tdeg values of the other three DESs are above 440 K. Thus, all four DESs are thermally stable at the operating temperature of the extraction experiments. The glass transition temperatures were also determined for the four DESs using differential scanning calorimetry. The half-step glass transition temperature method 47 was adopted here, and Tg values range from 205 K for DES-1 to 245 K for DES-2 (see Table 2).

11



DES-1 DES-2 DES-3 DES-4

100

80

m / m0 [%]

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 26

60

40

20

0 300

400

500 T [K]

600

700

Figure 2: Thermographs of the four DESs obtained with a heating rate of 5 K min−1 .

Mercury Extraction Efficiencies The extraction performance for the system [n-dodecane + Hg0 + DES] was evaluated for solvent to feed ratios of 1:1 and 2:1 at T = 303.15 and 333.15 K and atmospheric pressure. No color change was noticed when the DESs were mixed with the n-dodecane solution and the mercury was transferred from the non-polar alkane phase to the polar DES phase. The initial and final mercury concentrations in the n-dodecane solution, Ci and Cf , respectively, were measured in triplicate for each sample, and each extraction experiment was done in duplicate. The extraction efficiencies, E, were calculated as follows: E = (Ci −Cf )/Ci

(5)

The results are listed in Table 3. It should be noted that the statistical uncertainties are fairly large, as one should expect for extraction experiments using very dilute solute concentrations. At 303.15 K, the two extraction experiments yield consistent C f values when the same feed solution was used. At 333.15 K, however, the C f values for DES-1 show significant scatter despite use of the same feed solution. Increasing the solvent-to-feed ratio is found to increase the extraction

12


Page 13 of 26 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


Table 3: Extraction Performance Data for the Four DESsa DES 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4

T feed CiA CfA CiB CfB E ∆G −1 −1 −1 −1 [K] [µg kg ] [µg kg ] [µg kg ] [µg kg ] [%] [kJ mol−1 ] 303.15 1:1 3670 ± 90 260 ± 10 3670 ± 90 280 ± 10 93 ± 3 −7.7 ± 0.2 303.15 2:1 3420 ± 30 200 ± 10 4120 ± 140 600 ± 10 90 ± 7 −5.3 ± 1.8 333.15 1:1 3950 ± 80 830 ± 20 3950 ± 80 520 ± 20 83 ± 6 −5.8 ± 1.1 333.15 2:1 3950 ± 80 460 ± 20 3950 ± 80 790 ± 10 84 ± 7 −4.2 ± 1.3 303.15 1:1 3670 ± 90 570 ± 20 3670 ± 90 590 ± 10 84 ± 3 −5.5 ± 0.1 303.15 2:1 3420 ± 30 240 ± 10 4120 ± 140 480 ± 10 91 ± 5 −5.4 ± 1.0 333.15 1:1 3990 ± 150 370 ± 10 3990 ± 150 330 ± 10 91 ± 4 −7.9 ± 0.3 333.15 2:1 3990 ± 150 210 ± 10 3990 ± 150 170 ± 10 95 ± 5 −7.8 ± 0.5 303.15 1:1 3680 ± 90 380 ± 10 3670 ± 90 460 ± 20 88 ± 4 −6.5 ± 0.4 303.15 2:1 3420 ± 30 270 ± 10 4120 ± 140 270 ± 10 93 ± 4 −6.0 ± 0.4 333.15 1:1 4970 ± 40 140 ± 10 4970 ± 40 110 ± 10 97 ± 6 −11.5 ± 0.5 333.15 2:1 4970 ± 40 30 ± 2 4970 ± 40 24 ± 2 99 ± 6 −14.0 ± 0.5 303.15 1:1 4160 ± 60 550 ±10 4160 ± 60 480 ± 20 88 ± 3 −6.2 ± 0.3 303.15 2:1 4160 ± 60 460 ± 20 4120 ± 140 220 ± 10 92 ± 6 −5.8 ± 1.4 333.15 1:1 3710 ± 110 400 ± 10 3710 ± 110 450 ± 10 88 ± 3 −7.1 ± 0.3 333.15 2:1 3710 ± 110 490 ± 20 3710 ± 110 490 ± 20 87 ± 4 −4.7 ± 0.2 a The superscripts A and B denote repeat experiments for a given feed to solvent ratio.

efficiency only for six of the eight cases, but even in those cases the increase in E falls within the statistical uncertainties. All four solvents yield extraction efficiencies in excess of 80% for both temperatures and both solvent:feed ratios. Overall, 9 out of 16 E values fall into the range from 87 to 93%, and E values tend to be somewhat higher at T = 333.15 K with 3 out of 8 satisfying E ≥ 95%. Particularly, DES-3 extracts more than 97% of the mercury at 333.15 K, but here a feed solution with an unusually high Ci value was used. The good extraction performance for DES-1, DES-2, and DES-3 can likely be attributed to the coordination of mercury by Cl anions, but coordination by electronegative oxygen atoms of the solvent may also play a role. For the zwitter-ionic DES-4, coordination by the carboxylate group may provide a favorable environment for mercury. An affinity of Hg atoms to solvate in acidic mediums has been described previously. 49 A molecular view of the solvation in DES-1 is provided in the next section. Gibbs free energies of transfer were also estimated from the concentration data (see Table 3) and are found to range from −4 to −14 kJ mol−1 (with uncertainties being less than 2 kJ mol−1 ); thus, solvation in the DESs is

13



clearly much preferred over solvation in a petroleum oil, but the uncertainties preclude a detailed ranking of the DESs investigated here. Recently, Abbott et al. 50 investigated the solvation of sulfur-containing and of oxygenated compounds in a variety of DESs. Using extraction data at two temperatures (differing by 10-15 K), the van’t Hoff relation was used to estimate enthalpic and entropic contributions to the Gibbs free energy of transfer. Although two temperatures are also used in the present work, the temperature difference is 30 K and heat capacity effects 51 may play a significant role. With caution (also considering the significant uncertainties in the present measurements), it appears that the change in the transfer free energies indicates that partitioning from alkane into DES is exothermic for DES-1 and endothermic for DES-2 and DES-3.

First Principles Molecular Dynamics Simulations Out of necessity, the FPMD simulations contain Hg in a very high concentration (about four orders of magnitude higher than the experimental systems investigated here). Thus, the first question that needs to be addressed is whether the presence of Hg in these concentrations significantly perturbs the structure of DES-1. Previously, we have already investigated the structure and dynamics of anhydrous and hydrous DES-1 using FPMD simulations. 52 Figure 3 shows the radial distribution functions (RDFs) and the corresponding number integrals (NIs) involving the trans- and cis-hydrogen atoms of urea as HBD and either Cl anion or the oxygen atom of urea as the HBA for neat (anhydrous) DES-1 and for the DES-1 system with two additional Hg atoms. In agreement with previous force-field based simulation studies, 32 the FPMD simulations demonstrate a preferential solvation of the Cl anion by the trans-hydrogen atoms of urea and double peaks for urea–urea pairs that indicate that the carbonyl oxygen either is involved in two hydrogen bonds with cis-hydrogen atoms from two different urea molecules or in two hydrogen bonds with the two trans-hydrogen atoms of the same urea molecule. The small differences between the two systems (neat DES-1 and DES-1 with two Hg atoms) observed for these RDFs and the corresponding NIs demonstrate that the solvent structure is not significantly perturbed by the presence of Hg in high 14


Page 14 of 26

Page 15 of 26

12

N (r)

2 9

g (r)

1 6

0

2

3

4

H(urea) - O(urea)

3

0 0.8 N (r)

4

0.4 3 g (r)

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


0

2

3

4

2 H(urea) - Cl 1

0

2

3

4

5

6

7

r [Å]

Figure 3: Effect of mercury on the structure of DES-1: radial distribution functions and corresponding number integrals for H(urea)–O(urea) and H(urea)–Cl are shown in the top and bottom parts. The solid lines show RDFs for neat DES-1 taken from previous work. 52 The dashed lines RDFs for DES-1 with two additional Hg atoms. The blue and green lines represent RDFs involving the trans- and cis-hydrogen atoms of urea. concentrations. This may be an indication that the loading capacity in DES-1 may actually be quite high.

Mercury–chlorine RDFs are shown in Figure 4 for the systems including pairs of Hg atoms added in three different formal oxidation states and two different initial Hg–Hg distances. For the Hg(0) system, the simulations for the two different initial Hg–Hg distances both yield a peak at r ≈ 3.7 Å for one Hg atom that is coordinated on average by 1.5 to 2 chlorine atoms as indicated by the NI (see Figure 4), whereas the other Hg atom is only weakly coordinated by chlorine atoms. In

15



contrast, for the Hg(+1) system, a very sharp peak at r ≈ 2.6 Å is the most prominent feature that is followed by deep and broad minimum. The NI for this peak is 2.0 and there is a broad plateau due to the minimum in the RDF. This structural feature clearly corresponds to a strongly bound HgCl2 species and is present for both Hg atoms in the simulation started with these atoms in close proximity, but only for one of the Hg atoms in the simulation started with the two Hg atoms at a large distance. Thus, there may be a free energy barrier preventing the two simulations to converge to the same answer. For the Hg(+2) system, both the sharp peak at r ≈ 2.6 Å and the broader peak at r ≈ 3.7 Å are present. To overcome the sampling bottlenecks and to provide more detailed insight on the speciation, potentials of mean force (PMFs) were calculated as function of Hg–Hg separation (see Figure 5) for the three systems containing different numbers of Cl atoms that reflect different oxidation states of the mercury species that are extracted from liquid petroleum or natural gas. These PMFs show a global minimum at r ≈ 3.9 Å for the Hg(0) and Hg(+2) systems and lend support to the similarity of the Hg–Cl solvation environments for these two systems observed in the RDFs. For the Hg(+1) system, only a weak local minimum is present at r ≈ 3.8 Å. For the Hg(+2) system, a weak minimum is also found at r ≈ 2.6 Å. The Hg–Hg species at r ≈ 2.6 Å correspond to the well-known mercury polycation, whereas the minimum at r ≈ 3.9 Å corresponds to a pair of Hg species separated by their van der Waals diameter. 53 To characterize the speciation present at these minima, the Hg2 Cln aggregation was determined using configurations from the two umbrella sampling windows from 3.5 to 4.1 Å for the Hg(0) and Hg(+2) systems, and using only the window from 2.6 to 2.9 Å for the Hg(+1) system. For the Hg(0) system, Hg2 Cl3 is the most prevalent speciation with about 40% of all aggregates. Among these Hg2 Cl3 aggregates, about two thirds have a structure with two Cl atoms coordinating to only one of the Hg atoms, and the other Cl atom either coordinates to both Hg atoms or to the second Hg atom. For these aggregates in the Hg(0) system, the average Mulliken partial charges on the Hg atoms and chlorine anions are −0.05 and −0.6 |e|, respectively. For the Hg(+2) system, Hg2 Cl3 is also the most prevalent speciation with about 60% of all aggregates. Here, more than half of

16


Page 16 of 26

Page 17 of 26

6

Hg(0) N (r)

3

4

2

g (r)

1 0

2

3

4

5

3

4

5

3

4

5

2

24

Hg(+1) N (r)

3

16

2

g (r)

1 0

2

8

9

Hg(+2) N (r)

3

6

2 1

g (r)

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


0

2

3

0

2

3

4

5

6

7

r [Å]

Figure 4: Hg–Cl radial distribution functions and the corresponding number integrals for DES-1 systems with the addition of 2 Hg (top), 2 HgCl (middle), and 2 Hg2 Cl2 compounds (bottom). The solid and dashed lines indicate RDFs for simulations started with the Hg atoms placed either in close proximity or far apart. The purple and green colors distinguish the Hg atoms with the instantaneously tighter and looser coordination by Cl anions. these aggregates have a structure with one shared Cl atom and both of the other two Cl atoms coordinating to one of the Hg atoms. Thus, it appears that one of the Hg atoms in the Hg(+2) system is reduced to a lower oxidation state. This conjecture is also supported by the observation that the system contains one Cl2 species where the two nuclei are separated by less than 2.1 Å, i.e., a distance that would be extremely unlikely for two Cl anions. The Mulliken population analysis 17



15

0

2 Hg Hg2Cl2 2 HgCl2

12

∆F [kcal/mol]

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

9

6

3

0

2.9

3.5

4.1

4.7

5.3

5.9

6.5

r [Å]

Figure 5: Hg–Hg potentials of mean force for DES-1 containing 2 Hg0 , Hg2 Cl2 , and 2 HgCl2 yields distinctly different average values of +0.64 and +0.01 |e| for the two Hg atoms, two Cl atoms with an average charge of −0.06 |e| and ten Cl anions with an average charge of −0.5 |e|. That is, solvation of Hg(+2) compounds in DES-1 can involve a redox reaction. Based on the uneven coordination of the Hg contact pair and the Mulliken charges, the redox reaction appears to be 2 Hg2+ + 2 Cl− → Hg2+ + Hg0 + Cl2 . Given the small changes in the structure of the DES, we speculate that such a redox reaction would also occur at lower mercury concentrations. For the Hg(+1) system, about 40% of the polycation aggregates are found with the Cl2 HgHgCl2 structure, 30% with the Cl2 HgClHgCl structure (where the Cl in the middle coordinates to both Hg atoms), 10% with the ClHgCl2 HgCl structure, and 20% as Hg2 Cl5 aggregates with various structures. Here, the Mulliken population analysis yields average partial charges of +0.23 |e| and −0.65 |e| for the Hg (poly)cations and Cl anions.

Conclusion The present work describes the first application of DESs as extraction solvent for the removal of elemental mercury (Hg0 ) from liquid petroleum. The DESs were composed using either choline 18


Page 18 of 26

Page 19 of 26 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


chloride or betaine as the hydrogen-bond acceptor, and either urea, ethylene glycol, or levulinic acid as the hydrogen-bond donor. These compounds are widely available or can be synthesized from bio-renewable resources. For example, choline chloride is an additive in chicken feed and levulinic acid can be derived from degradation of cellulose. All four DESs were found to be very good extraction solvents with extraction efficiencies exceeding 80% for 1:1 and 2:1 solvent:feed ratios and Gibbs free energies of transfer from n-dodecane to DES being negative (i.e., favorable) with a magnitude from about 4 to 14 kJ mol−1 . First principles molecular dynamics simulations for the choline chloride:urea system indicate that the coordination sphere of mercury is populated predominantly by chloride anions with remarkably similar solvation structures for Hg atom pairs with initially different formal oxidation states. For the system initially containing two Hg2+ cations, the simulations point to a redox reaction resulting in the formation of one Hg(0) atom and one Cl2 molecule.

Acknowledgement Financial support for this collaborative project at the Petroleum Institute (PI) and at the University of Minnesota from the Petroleum Institute Research Center (PIRC) through a grant entitled “Advanced PVT-Properties and Molecular Modeling of Complex Fluids in Support of Safe and Green Hydrocarbon Production" (Project Code LTR14009) is gratefully acknowledged. The authors would also like to thank Eindhoven University of Technology (TU/e) for the kind hospitality in the Separation Technology Group (SEP). Sincere gratitude goes to SEP team members and W. Weggemans for their help in the experimental activates. Computer resources were provided by the Minnesota Supercomputing Institute.

Conflict of Interest Statement. Peters, Kroon, and Warrag are named as inventors on a patent application relating the subject matter of this work.

19



Supporting Information Available NMR spectra of the DESs, the experimental density and viscosity data, and the experimental extraction data. This material is available free of charge via the Internet at http://pubs.acs. org/.

20


Page 20 of 26

Page 21 of 26 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


References 1. Shah, M. S.; Tsapatsis, M.; Siepmann, J. I. Hydrogen sulfide capture: from absorption in polar liquids to oxide, zeolite, and metal-organic framework adsorbents and membranes. Chem. Rev. 2017, 117, 9755–9803. 2. Sanz-Pérez, E. S.; Murdock, C. R.; Didas, S. A.; Jones, C. W. Direct capture of CO2 from ambient air. Chem. Rev. 2016, 116, 11840–11876. 3. Abanades, J.; Arias, B.; Lyngfelt, A.; Mattisson, T.; Wiley, D.; Li, H.; Ho, M.; Mangano, E.; Brandani, S. Emerging CO2 capture system. Int. J. Greenh. Gas Con. 2015, 40, 126–166. 4. Wilhelm, S.; Bloom, N. Mercury in petroleum. Fuel Process. Technol. 2000, 63, 1–27. 5. Portney, P. R. Policy watch: economics and the clean air act. J. Econ. Perspect. 1990, 4, 173– 181. 6. Markovs, J.; Heights, Y.; Cintins, P. E.; Plaines, D. Purification of fluid streams containing mercury US Patent, 4874525A, 1989. 7. Li, L.; Li, X.; Lee, J.-Y.; Keener, T. C.; Liu, Z.; Yao, X. The effect of surface properties in activated carbon on mercury adsorption. Ind. Eng. Chem. Res. 2012, 51, 9136–9144. 8. Granite, E. J.; Pennline, H. W.; Hargis, R. A. Novel sorbents for mercury removal from flue gas. Ind. Eng. Chem. Res. 2000, 39, 1020–1029. 9. Eckersley, N. Advanced mercury removal technologies. Hydrocabon Process. 2010, 29–35. 10. Hiroshi, N.; Toshio, A.; Katsuya, N. Process for removal of mercury vapor and adsorbant therefor US Patent, 4500327A, 1985. 11. Rogers, R. D.; Holbrey, J.; Rodriguez, H. Process for removing metals from hydrocarbons WO Patent, 116165A2, 2010.

21



12. Abai, M.; Atkins, M. P.; Hassan, A.; Holbrey, J. D.; Kuah, Y.; Nockemann, P.; Oliferenko, A. A.; Plechkova, N. V.; Rafeen, S.; Rahman, A. A.; Ramli, R.; Shariff, S. M.; Seddon, K. R.; Srinivasan, G.; Zou, Y. An ionic liquid process for mercury removal from natural gas. Dalton Trans. 2015, 44, 8617–8624. 13. Mancini, M. V.; Spreti, N.; Di Profio, P.; Germani, R. Understanding mercury extraction mechanism in ionic liquids. Sep. Purif. Technol. 2013, 116, 294–299. 14. Cheng, G.; Bai, B.; Zhang, Q.; Cai, M. Hg0 removal from flue gas by ionic liquid/H2 O2 . J. ˘ S773. Hazard. Mater. 2014, 280, 767âA ¸ 15. Ji, L.; Thiel, S. W.; Pinto, N. G. Room temperature ionic liquids for mercury capture from flue gas. Ind. Eng. Chem. Res. 2008, 47, 8396–8400. 16. Ji, L.; Thiel, S. W.; Pinto, N. G. Pyrrolidinium imides: promising ionic liquids for direct capture of elemental mercury from flue gas. Water, Air, Soil Pollut. Focus 2007, 8, 349–358. 17. Rogers, R. D.; Holbrey, J. Ionic liquid solvents of perhlide type for metals and metal compounds US Patent, 20120090430A1, 2012. 18. Abbott, A. P.; Capper, G.; Davies, D. L.; Rasheed, R. K.; Tambyrajah, V. Novel solvent properties of choline chloride/urea mixtures. Chem. Commun. 2003, 70–71. 19. Zhang, Q.; Vigier, K. D. O.; Royer, S.; Jérôme, F. Deep eutectic solvents: syntheses, properties, applications. Chem. Soc. Rev. 2012, 41, 7108–7146. 20. Francisco, M.; van den Bruinhorst, A.; Kroon, M. C. Low-transition-temperature mixtures (LTTMs): a new generation of designer solvents. Angew. Chem. Int. Ed. 2013, 52, 3074–3085. 21. Dai, Y.; van Spronsen, J.; Witkamp, G.-J.; Verpoorte, R.; Choi, Y. H. Natural deep eutectic solvents as new potential media for green technology. Anal. Chim. Acta 2013, 766, 61–68. 22. Hayyan, A.; Mjalli, F. S.; AlNashef, I. M.; Al-Wahaibi, Y. M.; Al-Wahaibi, T.; Hashim, M. A. Glucose-based deep eutectic solvents: physical properties. J. Mol. Liq. 2013, 178, 137–141. 22


Page 22 of 26

Page 23 of 26 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


23. Warrag, S. E.; Peters, C. J.; Kroon, M. C. Deep eutectic solvents for highly efficient separations in oil and gas industries. Curr. Opin. Green Sustain. Chem. 2017, 5, 55–60. 24. Garcia, G.; Aparicio, S.; Ullah, R.; Atilhan, M. Deep eutectic solvents: physicochemical properties and has separation applications. Energy Fuels 2015, 29, 2616–2644. 25. Mahto, A.; Mondal, D.; Polisetti, V.; Bhatt, J.; M. R, N.; Prasad, K.; Nataraj, S. K. Sustainable water reclamation from different feed streams by forward osmosis process using deep eutectic solvents as reusable draw solution. Ind. Eng. Chem. Res. 2017, 56, 14623–14632. 26. AlOmar, M. K.; Alsaadi, M. A.; Jassam, T. M.; Akib, S.; Hashim, M. A. Novel deep eutectic solvent-functionalized carbon nanotubes adsorbent for mercury removal from water. J. Colloid Interface Sci. 2017, 497, 413–421. 27. Smith, J. G. Organic Chemistry, 2nd ed.; McGraw-Hill: New York, 2008; pp 95–99. 28. Abbott, A. P.; Capper, G.; Gray, S. Design of improved deep eutectic solvents using hole theory. ChemPhysChem 2006, 7, 803–806. 29. D’Agostino, C.; Harris, R. C.; Abbott, A. P.; Gladden, L. F.; Mantle, M. D. Molecular motion and ion diffusion in choline chloride based deep eutectic solvents studied by 1H pulsed field gradient NMR spectroscopy. Phys. Chem. Chem. Phys. 2011, 13, 21383–21391. 30. Abbott, A. P.; Harris, R. C.; Ryder, K. S. Application of hole theory to define ionic liquids by their transport properties. J. Phys. Chem. B 2007, 111, 4910–4913. 31. Shahbaz, K.; Mjalli, F. S.; Hashim, M. A.; Al Nashef, I. M. Using deep eutectic solvents for the removal of glycerol from palm oil-based biodiesel. J. Appl. Sci. 2010, 10, 3349–3354. 32. Perkins, S. L.; Painter, P.; Colina, C. M. Molecular dynamic simulations and vibrational analysis of an ionic liquid analogue. J. Phys. Chem. B 2013, 117, 10250–10260.

23



33. Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard III, W. A. and Skiff, W. M. UFF, a full periodic table force field for molecular mechanics and molecular dynamics. J. Am. Chem. Soc. 1992, 114, 10024–10035. 34. Yadav, A.; Pandey, S. Densities and viscosities of (choline chloride + urea) deep eutectic solvent and its aqueous mixtures in the temperature range 293.15 K to 363.15 K. J. Chem. Eng. Data 2014, 59, 2221–2229. 35. Hutter, J.; Ianuzzi, M.; Schiffmann, F.; VandeVondele, J. CP2K: atomistic simulations of condensed matter systems. WIREs: Comput. Mol. Sci. 2014, 4, 15–25. 36. VandeVondele, J.; Krack, M.; Mohamed, F.; Parrinello, M.; Chassaing, T.; Hutter, J. QUICKSTEP: fast and accurate density functional calculations using a mixed Gaussian and plane waves approach. Comput. Phys. Comm. 2005, 167, 103–128. 37. Becke, A. D. Density-functional exchange-energy approximation with correct asymptotic behavior. Phys. Rev. A 1988, 38, 3098–3100. 38. Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B 1988, 37, 785–789. 39. Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and accurate Ab Initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010, 132, 154104. 40. VandeVondele, J.; Hutter, J. Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases. J. Chem. Phys. 2007, 127, 114105. 41. Goedecker, S.; Teter, M.; Hutter, J. Separable dual-space Gaussian pseudopotentials. Phys. Rev. B 1996, 54, 1703–1710. 42. Hartwingsen, C.; Goedecker, S.; Hutter, J. Relativistic separable dual-space Gaussian pseudopotentials from H to Rn. Phys. Rev. B 1998, 58, 3641–3662. 24


Page 24 of 26

Page 25 of 26 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


43. Hoover, W. G. Canonical dynamics: equilibrium phase-space distributions. Phys. Rev. A 1985, 31, 1695–1697. 44. Martyna, G. J.; Klein, M. L.; Tuckerman, M. Nosé–Hoover chains: the canonical ensemble via continuous dynamics. J. Chem. Phys. 1992, 97, 2635–2643. 45. Grossfield, A. WHAM: the weighted histogram analysis method, version 2.0.9, 2016. 46. Rodriguez, N. R.; Requejo, P. F.; Kroon, M. C. Aliphatic–aromatic separation using deep eutectic solvents as extracting agents. Ind. Eng. Chem. Res. 2015, 54, 11404–11412. 47. Zubeir, L. F.; Lacroix, M. H. M.; Kroon, M. C. Low transition temperature mixtures as innovative and sustainable CO2 capture solvents. J. Phys. Chem. B 2014, 118, 14429–14441. 48. Vyazovkin, S. Thermogravimetric analysis. In Characterization of Materials; Kaufmann, E. N., Ed., 2nd ed.; John Wiley & Sons: New York, 2012; pp 1 – 12. 49. ShamsiJazeyi, H.; Kaghazchi, T. Investigation of nitric acid treatment of activated carbon for enhanced aqueous mercury removal. J. Ind. Eng. Chem. 2010, 16, 852–858. 50. Abbott, A. P.; Al-Murshedi, A. Y.; Alshammari, O. A.; Harris, R. C.; Kareem, J. H.; Qader, I. B.; Ryder, K. Thermodynamics of phase transfer for polar molecules from alkanes to deep eutectic solvents. Fluid Phase Equil. 2017, 448, 99–104. 51. Wick, C. D.; Siepmann, J. I.; Schure, M. R. Temperature dependence of transfer properties: Importance of heat capacity effects. J. Phys. Chem. B 2003, 107, 10623–10627. 52. Fetisov, E. O.; Harwood, D. B.; Kuo, I.-F. W.; Warrag, S. E. E.; Kroon, M. C.; Peters, C. J.; Siepmann, J. I. First-principles molecular dynamics study of a deep eutectic solvent: choline chloride/urea and its mixture with water. J. Phys. Chem. B 2018, 122, 1245–1254. 53. Earnwood, N. N.; Greenshaw, A. Chemistry of the Elements, 2nd ed.; Elsevier B.V.: Oxford, 1997. 25



Graphical TOC Entry

26


Page 26 of 26

Mercury Capture from Petroleum Using Deep Eutectic Solvents

Recommend Documents