Combined Experimental and Molecular Simulation Investigation of the

Jul 11, 2016 - User Resources. About Us · ACS Members · Librarians · Authors & Reviewers · Website Demos · Privacy Policy · Mobile Site ...
0 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/JPCA

Combined Experimental and Molecular Simulation Investigation of the Individual Effects of Corexit Surfactants on the Aerosolization of Oil Spill Matter Zenghui Zhang,†,⊥ Paria Avij,†,⊥ Matt J. Perkins,‡ Thilanga P. Liyana-Arachchi,§ Jennifer A. Field,‡ Kalliat T. Valsaraj,† and Francisco R. Hung*,†,∥,# †

Cain Department of Chemical Engineering, Louisiana State University, Baton Rouge, Louisiana 70803, United States Department of Environmental & Molecular Toxicology, Oregon State University, Corvallis, Oregon 97331, United States § Department of Chemistry, University of Florida, Gainesville, Florida 32611, United States ∥ Center for Computation & Technology, Louisiana State University, Baton Rouge, Louisiana 70803, United States ‡

S Supporting Information *

ABSTRACT: We report laboratory aerosolization experiments and classical molecular dynamics (MD) simulations, with the objective of investigating the individual effects of the two Corexit surfactants Span 80 (nonionic) and dioctyl sodium sulfosuccinate (DOSS, ionic), on the aerosolization of oil spill matter to the atmosphere. Our simulation results show that Span 80, DOSS, and the oil alkanes n-pentadecane (C15) and n-triacontane (C30) exhibit deep free energy minima at the air/seawater interface. C15 and C30 exhibit deeper free energy minima at the interface when Span 80 is present, as compared to the situation when DOSS or no surfactants are at the interface. These results suggest that Span 80 makes these oil hydrocarbons more likely to be adsorbed at the surface of seawater droplets and carried out to the atmosphere, relative to DOSS or to the situation where no surfactants are present. These simulation trends are in qualitative agreement with our experimental observations in a bubble-column setup, where larger amounts of oil hydrocarbons are ejected when Span 80 is mixed with oil and injected into the column, as compared to when DOSS is used. Our simulations also indicate that Span 80 has a larger thermodynamic incentive than DOSS to move from the seawater phase and into the air/seawater interface. This observation is also in qualitative agreement with our experimental measurements, which indicate that Span 80 is ejected in larger quantities than DOSS. Our simulations also suggest that DOSS predominantly adopts a perpendicular orientation with respect to the air/seawater interface at a dispersant to oil ratio (DOR) of 1:20, but has a slight preference to lie parallel to the interfaces at a DOR = 1:5; in both cases, DOSS molecules have their tails wide open and stretched. In contrast, Span 80 has a slight preference to align parallel to the interfaces with a coiled conformation at both DOR values. mass, respectively.10 The main difference between Corexit 9500A and 9527A is that in the latter the solvent 2butoxyethanol is present; however, this compound raised toxicity concerns and was removed from the 9500A formulation.7,13 The toxicity and environmental fate of Corexit surfactants in water10,11,14,15 and in sediments16 have been the subjects of intensive research. In contrast, few studies have addressed the ejection of Corexit surfactants to the atmosphere,17−20 where most of these reports focused on the evaporation of volatile organic compounds.17−19 Evaporation of volatile and intermediate volatile organic compounds (VOCs and IVOCs, i.e., species having 16 carbon atoms) is also another important transport mechanism of oil spill matter into the atmosphere. Results from laboratory experiments using a bubble column reactor,20 and from classical molecular dynamics (MD) simulations,21 suggest that heavier IVOCs and SVOCs, as well as Corexit components with limited volatility, can be carried into the atmosphere adsorbed on the surface of water droplets and sea sprays. Sea surface phenomena such as bubble bursting and whitecaps generate major quantities of sea sprays, which represent a major source of production of particulate matter flowing into the atmosphere.22−27 The aerosols detected downwind from the DWH oil spill site were attributed to a mechanism involving evaporation of hydrocarbons from the sea surface, oxidation in the atmosphere, and condensation to form secondary organic aerosol (SOA) because of the following reasons: (1) evidence of new particle formation was observed downwind from the DWH site,28 (2) sub-μm sized aerosols increased in abundance in the plume downwind from the DWH site,17,18 and (3) oxygenated organic species were detected in the aerosols in the plume.29 However, the ratio of hydrocarbonlike organic aerosol (HOA) to oxygenated organic aerosol (OOA) downwind from the spill site exhibited higher levels (up to 2-fold) compared to those measured at the spill site.17,20 This observation is in contradiction to the hypothesis of attributing the formation of aerosol plume to purely oxidative SOA formation. Furthermore, the bubble column reactor used here and in the study of Ehrenhauser et al.20 could produce aerosols that are slightly larger than 1 μm in size30 (between 0.63 and 2.41 μm, as reported in our previous study20). Moreover, the instruments used to observe organic aerosols downwind from DWH would not have detected super-μm aerosols. All these observations suggest that both the formation of SOAs and the sea spray mechanism studied here and in our previous studies20,21 could have contributed to the formation of oil spill matter aerosols during the DWH accident. Building on our previous studies,20,21 here we present a combination of experiments and molecular simulations that aim at establishing the individual effects of two of the main components of Corexit (DOSS and Span 80), on the aerosolization of oil spill matter to the atmosphere. Experimental ejection rates of IVOC and SVOC alkanes, as well as DOSS and Span 80, were determined using an aerosolization reactor, whereas our simulations consisted of potential of mean force (PMF) calculations as well as conventional MD simulations. Only a few simulation studies have focused on studying the properties of oil and surfactants at the air/water interface.21,31 We note that in our previous simulation paper21 we considered several model surfactants that aimed at representing the tail of the nonionic surfactants found in Corexit dispersants; however, neither the ionic surfactant DOSS nor specific Corexit nonionic surfactants were considered in our previous study.21 Here we used realistic united atom models for DOSS and Span 80, and for the linear oil alkanes n-pentadecane (C15) and n-triacontane (C30), which are representatives of heavier IVOCs and SVOCs found in Louisiana sweet crude oil. For simplicity we did not model the nonionic Corexit surfactants Tween 80 and Tween 85, which are complex mixtures of a homologous series of

compounds that vary in the degree of polyethoxylation.32 In the experimental part of this paper, we premixed pure DOSS and pure Span 80 with surrogate oil, injected the mixtures into the bubble column reactor, and quantified the ejection rates of the oil alkanes, DOSS, and Span 80. In a separate paper,33 we report experiments where we premix the dispersants Corexit 9500A and 9527A with surrogate oil, inject the mixtures in our bubble column reactor filled with actual seawater, and determine the ejection rates of the species involved. With our combined experimental−simulation study, we aim at fundamentally understanding the properties of oil alkanes and surfactants at the air/seawater interface, which is relevant to the possible ejection of oil spill matter into the atmosphere through sea sprays and water droplets. The rest of this article is arranged as follows. Details of our experimental and computational methods are described in Section 2. Section 3 presents the main results and discussions, and concluding remarks are summarized in Section 4.

2. METHODS 2.1. Experimental Methods. 2.1.1. List of Chemical and Reagents. See Supporting Information for a detailed list of chemicals and reagents used in our experiments. 2.1.2. Laboratory Aerosolization Experiments. The laboratory aerosolization experiments were performed in a bubble column reactor at 298 K, developed after Smith et al.34 A complete description of the bubble column reactor and aerosol generator is provided in our previous paper.20 The model seawater phase in the bubble column consisted (for simplicity) of a solution of sodium chloride (3.5% w/w) in deionized water. The individual Corexit surfactants DOSS and Span 80 (Table 1) were thoroughly mixed with the surrogate oil at a Table 1. Chemical, Physical, and Structural Characteristics of DOSS and Span 80

a

HLB: hydrophilic−lipophilic balance.

dispersant to oil ratio (DOR) of 1:20 (w/w) at room temperature. This mixture was injected into the aerosolization reactor using a syringe pump at an injection rate of 50 μL/min in separate experiments. When the system reached steady state, the effluent of the reactor was collected by two sampling methods as in our previous study:20 (1) constant mass-flow airsampling nozzle, which allows collection of both vapors and particles; and (2) electrostatic precipitator (ESP), which only allows particle collection. Duplicate samples from each method were collected over a time period of 15 min. 2.1.3. GC−FID and GC−MS Analyses To Determine Ejection Rates of Alkanes. The alkane contents of the reactor effluent were quantified using gas chromatography equipped with a flame ionization detector (GC−FID) or a single B

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A quadrupole mass spectrometer (GC−MS) as described by Ehrenhauser et al.20 with minor modifications (see Supporting Information). The ejection rates of alkanes were calculated as described in the Supporting Information. The ejection rates of volatile and intermediate-volatile alkanes (C10−C19) are reported as the mean of duplicate measurements collected by nozzle. The ejection rates of semivolatile alkanes (C20−C29) are reported as the mean of quadruplicate measurements collected by nozzle (n = 2) and ESP (n = 2). The pooling of ESP and nozzle data for C20−C29 is justified as these alkanes are ejected in the form of particles and are not expected to be present in the vapor phase. More details are available in the Supporting Information. 2.1.4. LC−MS/MS Analysis to Determine Ejection Rates of Span 80 and DOSS. Splits of the same reactor effluent collected by nozzle or ESP were shipped by ground to Oregon State University (OSU) in isopropanol in glass scintillation vials at ambient temperature. Upon arrival at OSU, the samples were stored at −20 °C until analysis. The reactor effluent was diluted 0 to 100-fold in an Instant Ocean (IO)−isopropanol (IPA) solution (final ratio 75:25 IO/IPA) and analyzed by LC−− MS/MS as described by Place et al.,10 with minor modifications (see Supporting Information). The ejection rates of DOSS and Span 80 were calculated as described for alkanes (see Supporting Information). The ejection rates of DOSS and Span 80 were reported as the mean of quintuplet (DOSS) or quadruplicate (Span 80) measurements collected by nozzle sampling (n = 3 for DOSS; n = 2 for Span 80) and ESP (n = 2 for both DOSS and Span 80). Pooling of the data for DOSS and Span 80 is justified because neither surfactant is expected to be in the vapor phase. More details are available in the Supporting Information. 2.2. Computational Models and Methods. Classical MD simulations were performed using GROMACS,35 to investigate systems of oil alkanes and Corexit surfactants at the air/ seawater interface at 298 K. We used united atom models to represent the linear alkanes n-pentadecane (C15) and ntriacontane (C30) as well as the surfactants DOSS and Span 80. As in our previous work,20 for simplicity, seawater was modeled as a solution of NaCl in water, where the nonpolarizable SPC/ E model36 was used for water and the nonpolarizable parameters of Vácha et al.37 were used to model NaCl. DOSS was represented using the model proposed by Chowdhary et al.,38 whereas Span 80 and the linear alkanes C15 and C30 were represented using the TraPPE-UA force field.39−43 Partial charges for Span 80 were determined from ab initio calculations at the HF/6-31G* level of theory, followed by a restrained electrostatic potential charge fitting procedure.44,45 These partial charges are reported in Table S1 and Figure S1 (Supporting Information). Parameters from GAFF46 were used to account for all bonded interactions in our simulations. As in our previous study,21 we performed all our simulations in the canonical ensemble, where we used an elongated orthorhombic simulation box (6 × 6 × 30 nm3) with a seawater slab (5805 water molecules and 63 NaCl ion pairs) placed in the center of the box, creating two air/seawater interfaces (Figure 1). The potential of mean force (PMF) values of the surfactants Span 80 and DOSS and of the linear alkanes C15 and C30 were determined in our simulations, by moving a molecule of these species between the liquid and gas phases and across the air/seawater interface. We determined the PMF profiles of the linear alkanes when the air/seawater interface is

Figure 1. Side view of our simulated setup, showing the elongated orthorhombic simulation box (6 × 6 × 30 nm3) with a seawater slab (5805 water molecules and 63 NaCl ion pairs) placed in the center of the box, creating two air/seawater interfaces. Black = C15, green = Span 80, red = water.

bare or coated with 12 molecules of either Span 80 or DOSS. For the surfactants, we computed their PMF profiles for the cases where the air/seawater interface is bare or coated with 16 molecules of either C15 or C30. These molecules of surfactants or alkanes coating the interface were placed randomly, ensuring a monolayer-like coverage. All of the simulations in our PMF calculations were run for at least 20 ns using the constraint force method. In this method, the z-distance between the center of mass of the molecule of interest (C15, C30, DOSS, or Span 80) and the center of mass of the water slab was constrained, and we monitor the force that we need to apply to keep the molecule of interest at this constrained z-distance. Finally, we also performed conventional MD simulations of systems of n-alkanes placed at the air/seawater interfaces containing varying amounts of either Span 80 or DOSS (DOR = 1:20, 1:10, and 1:5 in each interface). The rest of the simulation details are exactly the same as in our previous study.21

3. RESULTS AND DISCUSSION 3.1. Individual Effects of DOSS and Span 80 on Experimental Ejection Rates. In Figure 2 we show experimental measurements of the ejection rates of oil alkanes [namely, VOCs (C10−C14), IVOCs (C15−C19), and SVOCs (C20−C29)] and the Corexit surfactants DOSS and Span 80. The results for alkane ejection rates indicate that mixing DOSS with the surrogate oil leads in general to lower ejection rates for all alkane groups considered, relative to the rates determined

Figure 2. Experimental ejection rates of alkanes and surfactants (Span 80 and DOSS). Results were determined by averaging over samples obtained from the sampling nozzle and the electrostatic precipitator (except for C10−C14 and C15−C19; see text), at an injection rate of 50 μL/min of the mixture of surrogate oil with surfactants into the reactor. The left, dark red bar in each group is the measured ejection rate for the given species when oil is premixed with Span 80; the right, orange bar represents the ejection rate of the species when oil is premixed with DOSS. C

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

injected into the bubble column. This enhancement in ejection rates seems to increase with the molecular weight of the oil alkanes. Furthermore, our results also suggest that Span 80 has a larger ejection rate than DOSS. These results can be combined with those presented in our previous experimental study,20 where larger alkane ejection rates were observed when oil was premixed with Corexit and injected into the column, relative to the situation where oil was directly injected into the bubble column without premixing with dispersant. The results shown in Figure 5 of ref 20 indicate that adding Corexit increased the ejection rate of all alkane groups: C10−C14 (47.7%), C15−C19 (24.4%), C20−C24 (168.2%), and C25− C29 (126.1%), as measured at an injection rate of 100 μL min−1. We also note that at the injection rate of 50 μL min−1 used here, the results reported in Figure 5 of ref 20 also exhibit significant increases in ejection rate for all alkane groups when Corexit is added, as compared to the situation where only oil is injected into the column. 3.2. Potentials of Mean Force (PMF) of Surfactants and n-Alkanes. In Figure 4 we show the PMF obtained by moving a molecule of the n-alkanes C15 and C30 between the seawater and gas phases and across an interface that are either bare or coated with 12 molecules of DOSS or Span 80 at 298 K. In all PMF results presented in this article, we arbitrarily assumed the PMF of the molecule of interest in the gas phase is equal to zero. In all cases the n-alkanes show deep free energy minima at the air/seawater interface, indicating that they have a thermodynamic preference to remain at these interfaces regardless of the absence or presence of surfactants. As the molecule of n-alkane moves deeper into the seawater phase, the value of PMF increases to large positive values. C30 exhibits a deeper PMF minimum at the interface than C15 (−10.1 kJ/ mol for C15 and −16.8 kJ/mol for C30), as well as a larger PMF value inside the seawater phase than C15. Coating the air/seawater interface with the ionic surfactant DOSS leads to deeper PMF minima at the interface (−16.5 kJ/mol for C15 and −27.7 kJ/mol for C30), as compared to the case of bare interfaces. These numbers represent an increase in the depth of the free energy minima of 63.4% for C15 and 64.9% for C30, when compared to the case of bare interfaces. Adding the nonionic surfactant Span 80 leads to an even deeper minima for the n-alkanes at the air/seawater interface (−17.7 kJ/mol for

when Span 80 is mixed with oil. For the heavier SVOCs, the difference in ejection rates when oil is mixed with Span 80 can be larger than an order of magnitude with respect to the rates observed when DOSS is mixed with oil. The results shown in Figure 2 also suggest that Span 80 is ejected at a higher rate than DOSS. It has been shown that DOSS has a higher partitioning rate from crude oil into seawater, as compared to other Corexit components.9 In contrast, Span 80, which is the most hydrophobic surfactant within Corexit, strongly attaches to the oil droplets due to its lower affinity for the aqueous phase, and partitions very slowly from the oil into seawater phase.9 This difference in partitioning kinetics could influence the ejection rates determined in our experiments. Visual observations of the reactor during these sets of experiments (Figure 3) suggest that

Figure 3. Visual comparison of the effect of DOSS Span 80 (left) and DOSS (right) with the same DOR (1:20) in oil injected into the aerosolization reactor.

application of DOSS allows most of the oil compounds to be dispersed within the water column, as suggested by the more brownish color of the salt water within the reactor relative to the color of the seawater when Span 80 is present. In any case, in our lab column the rising bubbles within the reactor may carry a significant part of the oil up to the surface; part of this can then transfer into the atmosphere, either by direct evaporation or adsorbed on the surface of bursting bubbles and jet sprays at the sea surface.47 Overall, our results suggest that larger ejection rates of oil alkanes are observed when Span 80 is premixed with oil, as compared to the situation where DOSS is mixed with oil and

Figure 4. Potentials of mean force (PMF) associated with moving one molecule of the n-alkanes C15 (left) and C30 (right) between the gas phase (left) and the seawater phase (right), and across an air/seawater interface that is either bare (blue solid curve) or coated with 12 molecules of Span 80 (red dashed curve) or DOSS (green dotted curve), at 298 K. The PMF of the n-alkanes in the gas phase was arbitrarily assumed to be zero. The gray dashed line represents the location of the air/seawater interface (arbitrarily defined as the point where the density of seawater in the simulation box reaches 500 kg/m3). D

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A C15 and −31.4 kJ/mol for C30; an increase in the depth of the PMF minima of 75.2% and 86.9% with respect to the case of bare interfaces). These results suggest that the presence of Span 80 at the air/seawater interface makes it more stable thermodynamically for the adsorption of n-alkanes relative to DOSS; in turn, DOSS makes the interface more stable for nalkane adsorption when compared to a bare air/seawater interface. These results are in agreement with the trends observed in our previous study,20 where deeper PMF minima were also observed for the n-alkanes when the interfaces were coated with model surfactant molecules resembling the tail of the nonionic surfactants present in Corexit. In Figure 5 we present the PMF obtained by moving one molecule of DOSS or Span 80 between the gas and the

Table 2. Summary of Results Obtained in Potential of Mean Force (PMF) Calculations system C15, bare interface C15, DOSScoated interface C15, Span 80coated interface C30, bare interface C30, DOSScoated interface C30, Span 80coated interface Span 80, bare interface Span 80, C15coated interface Span 80, C30coated interface DOSS, bare interface DOSS, C15coated interface DOSS, C30coated interface

PMF minimum (kJ/mol)

[PMF minimum] − [PMF value inside seawater phase], (kJ/mol)

−10.1 −16.5

−50.2 −59.4

−17.7

−60.6

−16.8 −27.7

−87.5 −98.4

−31.4

−102

−67.5

−60.9

−57.7

−67.6

−56.2

−76.7

−119 −98.3 −102

−57.3 −49.6 −60.3

Span 80 but causes a slightly deeper PMF minimum for DOSS (Figure 5). 3.2.1. Relation between Simulated PMF Profiles and Experimental Ejection Rates. We first note here that both thermodynamics and transport phenomena play a role in the ejection rates measured in our experiments. In contrast, our PMF calculations solely probe the thermodynamics (free energies) involved when a single molecule of n-alkane (C15, C30) or surfactant (DOSS, Span 80) is dragged between the air and seawater phases, and through an interface that is bare, coated with 12 molecules of DOSS or Span 80 (when dragging an n-alkane, Figure 3) or coated with 16 molecules of C15 or C30 (when moving a surfactant, Figure 4). These PMF results are summarized in Table 2, which contains the values of PMF minima for each system, as well as the differences between the PMF minima and the PMF values determined when the molecule of interest is in the seawater phase (rightmost plateau values from Figures 3 and 4). This last column in Table 2 provides a measurement of the “thermodynamic incentive” that a given molecule has to move to the air/seawater interface, as compared to staying dissolved in the seawater phase. The results shown in Figure 4 suggest that the ionic DOSS has more affinity with seawater than the nonionic Span 80, as indicated by the lower values of PMF observed for DOSS than for Span 80 when these surfactants are in the seawater phase. These results are consistent with the reported hydrophilic−lipophilic balance (HLB) values of DOSS (10.5) and Span 80 (4.3),48 indicating that the former is more hydrophilic and less volatile than the latter.48−50 HLB values are important to design mixtures of surfactants for oil dispersion purposes,51,52 as surfactants with HLB ≈ 9−13 typically succeed in dispersing large portions of oil. Figure 4 and Table 2 indicate that adding Span 80 to the air/ seawater interface makes it thermodynamically more stable for the adsorption of n-alkanes, as signaled by (1) the increase in the PMF depth and (2) the larger differences between the PMF minimum and PMF inside seawater (Table 2), when compared

Figure 5. Potential of mean force (PMF) associated with moving one molecule of the Corexit surfactants Span 80 (red curves) and DOSS (green curves) between the gas phase (left) and the seawater phase (right) and across an air/seawater interface that is either bare (solid curves) or coated with 16 molecules of C15 (dotted curves) or C30 (dashed curves), at 298 K. The PMF of the surfactants in the gas phase was arbitrarily assumed to be zero. The gray dashed line represents the location of the air/salt water interface (arbitrarily defined as the point where the density of seawater in the simulation box reaches 500 kg/ m3).

seawater phases and across an air/seawater interface that is either bare or coated with 16 molecules of C15 or C30. In analogy to the PMF plots shown in Figure 4 for n-alkanes, the PMF profiles of both Span 80 and DOSS also exhibit a deep minimum at the bare air/seawater interface (−67.5 kJ/mol for Span 80 and −119 kJ/mol for DOSS). Therefore, thermodynamics indicate that both surfactants prefer to stay at the air/ seawater interface, rather than being in the seawater or in the gas phases; DOSS has a much deeper minima at the interface than Span 80. When n-alkanes are present in the air/seawater interface (dashed and dotted curves in Figure 5), the PMF minima for both DOSS and Span 80 become shallower (i.e., less negative). For DOSS the observed PMF minima are −98.3 and −102 kJ/mol when C15 and C30 are at the interfaces (a reduction of 17.5% and 14.5% with respect to the case of bare interfaces). Similarly, for Span 80, the PMF minima become −57.7 kJ/mol and −56.2 kJ/mol when C15 and C30 are at the interfaces, representing a reduction of 14.5% and 16.7% with respect to the case of bare interfaces. The results shown in Figure 5 suggest that the presence of n-alkanes in the air/ seawater interface makes it thermodynamically less stable for the adsorption of the Corexit surfactants Span 80 and DOSS resulting in a decreased PMF deph of 15−18% (Table 2; see also discussion below). Increasing the size of the n-alkane from C15 to C30 leads to a slightly shallower PMF minimum for E

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 6. Density profiles of water, the n-alkanes C15 (left column) and C30 (right column), and the head and tail groups of the Corexit surfactants Span 80 (top row) and DOSS (bottom row), from conventional molecular dynamics simulations of air/seawater systems at 298 K (see simulation setup in Figure 1). All systems contain 40 molecules of C15 or C30 and two molecules of Span 80 or DOSS per air/seawater interface, giving a DOR = 1:20. The density profile of each species is normalized by dividing by the maximum value of the local density of each species in the simulation box.

“extended” configurations are also present in our PMF simulations, when the n-alkanes or the surfactants are in the seawater phase but close to the air/seawater interface coated with surfactants or n-alkanes; the molecules stretch to try to touch the species coating the interface (see Figure S2, Supporting Info). Huston and Larson53 argue that such an energy barrier can lead to hysteresis effects in the PMF. In our systems, these effects impact the accuracy of the PMF values determined when the molecule of interest is in the seawater phase (see, e.g., the differences in PMF values at the rightmost side of Figure 5), as well as the accuracy of the values reported in the last column of Table 2. For molecules simpler than Tween 80, Huston and Larson53 could run 2D-biased umbrella sampling simulations to overcome hysteresis effects, but for their model Tween 80 molecule they argued that running these simulations was too expensive computationally. Similarly, in our study we did not attempt to overcome these hysteresis effects (our PMF simulations are already computationally very expensive). Therefore, we could argue that the PMF results shown in Figures 3 and 4 and Table 2 have much larger error bars than PMF calculations performed by us in the past for much smaller molecules.54−56 The largest error bars in our PMF results are on the order of 20 kJ/mol, which correspond to the discrepancies observed in the PMF values at the rightmost side of Figure 5, when the surfactant molecules are in the seawater phase but close to the n-alkanes at the interface. However, we emphasize that these large error bars do not affect the trends observed in our results and discussed here, and thus, our conclusions still hold. 3.3. Structural Properties of Alkanes and Dispersants at the Air/Seawater Interface. In Figure 6 we depict the density profiles of n-alkanes (C15 or C30), surfactants (DOSS or Span 80), and water molecules in air/seawater systems at 298 K, as obtained from conventional MD simulations. In this figure, the local densities of the surfactants were separated into contributions from the head and tail groups. For Span 80, we considered the five-membered ring and the ester group (Table 1) as part of the headgroup and the carbon chain as the tail group. For DOSS, we considered the sulfonate moiety as the

to the situations when DOSS is present at the interface, or when no surfactant is present. Therefore, thermodynamics suggest that the oil alkanes are more likely to be carried out to the atmosphere adsorbed at the surface of bursting bubbles when Span 80 is added, as compared to when DOSS is added or when no surfactants are present. This observation agrees qualitatively with the experimental findings in Figure 2 that more oil hydrocarbons are ejected when Span 80 is present in the system, as compared to when DOSS is added. The presence of n-alkanes at the air/seawater interface makes it slightly less stable for the adsorption of DOSS and Span 80, as suggested by PMF minima that are about 15−18% shallower relative to the case of a bare interface (Figure 5 and Table 2). Span 80 is predicted to have a shallower PMF minimum at the air/seawater interface as compared to DOSS, according to Figure 5. However, the differences between the PMF minima and the PMF inside the seawater phase in these systems (last column of Table 2) suggest that Span 80 has a larger thermodynamic incentive to move from the seawater phase and into the interface (difference ranges from −61 to −77 kJ/mol), as compared to DOSS (difference ranges from −50 to −60 kJ/ mol). This last observation agrees qualitatively with the experimental results that Span 80 is ejected in slightly larger quantities than DOSS in our bubble column (Figure 2). Nevertheless, again we emphasize that both thermodynamics and transport processes play roles in the aerosolization of oil spill matter, while our molecular simulations only probe thermodynamics in these systems. We also have to mention that the results shown in Figures 3 and 4 (and Table 2) might be subject to hysteresis effects, similar to those discussed in detail by Huston and Larson53 for a model of Tween 80 in water/oil systems. In this study, the authors determined PMFs for several molecules, including their model Tween 80, using an umbrella sampling method. They discuss that as their model Tween 80 molecule is in the water phase but close to the interface with oil, there is an energy barrier between configurations where the tail of the molecule is either “retracted” or “extended” to reach out and touch the oil phase (see Figure S3 in ref 53). These “retracted” and F

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A headgroup and the two hydrocarbon chains connected to the sulfonate headgroup as the tail group. All systems contained 40 molecules of C15 or C30 and two molecules of DOSS or Span 80 per air/seawater interface, giving an overall DOR = 1:20, which is the same value used in our experiments, as well as during response operations of the DWH oil spill. However, as we discussed before,55,56 this value of overall DOR (which could be considered as a measure of overall concentrations of oil and dispersant) can be significantly different from the surface or local value of DOR, which is more difficult to determine from experiments and in the field. A given system can have a macroscopic, overall DOR = 1:20, but the surface/local/ molecular value of DOR could be much larger. Side view snapshots of systems containing surfactants and nalkanes at the air/seawater interfaces are shown in Figure 7.

Figure 8. Definition of the angular parameters θ used to describe the molecular orientation of the molecules of the surfactants DOSS (left) and Span 80 (right) with respect to the air/seawater interface. Blue = water; black = C15; yellow = surfactant DOSS or Span 80. The red beads in the surfactant molecules represent the atoms used to define the angle θ.

sulfonate group (Table 1). In Span 80 (Figure 8), the angle θ is formed between the vector normal to the interface (zdirection) and the vector that joins the end carbon atom on the tail with the carbon atom connecting the five-membered ring with the linear carbon chain (Table 1). A value of cos(θ) = 1 indicates that the surfactant molecule aligns parallel to the zaxis and thus is perpendicular to the air/seawater interface (see insets in Figure 9), whereas a value of cos(θ) = 0 means that the surfactant molecule aligns perpendicular to the z-axis and lies parallel to the air/seawater interface (insets, Figure 9). The probability distributions of cos(θ) for DOSS and Span 80 at DOR = 1:20 and 1:5, in the presence of either C15 or C30, are shown in Figure 9. At DOR = 1:5 both surfactants tend to predominantly align parallel to the interface, although both surfactants exhibit a wide distribution of orientations. However, at DOR = 1:20, DOSS aligns mostly perpendicular to the air/ seawater interface, whereas the distribution of orientations of Span 80 becomes wider than that observed for the same surfactant at DOR = 1:5. Varying the chain length of the nalkane seems to have little influence on the orientation of DOSS and Span 80, although at a DOR = 1:20, DOSS seems to have a wider distribution of orientations in the presence of C30, as compared to C15. The trends observed at DOR = 1:10 fall in between the angle distributions presented in Figure 9 and are not shown for brevity. An intramolecular angle α for both DOSS and Span 80 (Figure 10) was monitored in our simulations, to gain insights on the effects of DOR values within the surfactant molecules. In DOSS, α is the angle formed by the two tails with the vertex given by the carbon atom connected to the sulfonate group (Figure 10); for Span 80, α is the angle formed between the headgroup and the tail group (Figure 10). A value of cos(α) = −1 indicates that the two tails of DOSS are fully opened and that the Span 80 molecule is fully stretched out (see insets in Figure 11); in contrast, cos(α) = 1 indicates that the two tails of DOSS are parallel, and the Span 80 molecule is coiled (inset, Figure 11). Figure 11 depicts the probability distribution of angle α at two DOR values (1:5 and 1:20), in the presence of either C15 or C30. Although wide distributions for cos(α) are observed in our simulations, the results suggest that the DOSS molecules predominantly have their two tails wide open and stretched, whereas Span 80 prefers to mostly adopt a coiled conformation (Figure 11). For Span 80, changing the DOR from 1:20 to 1:5 cause the distributions of cos(α) to become flatter; for DOR = 1:5, the angle distributions seem to become

Figure 7. Side view of representative snapshots from molecular dynamics simulations of oil n-alkanes and surfactants in air/seawater systems at 298 K. All systems contain 40 molecules of C15 (left column) or C30 (right column) and two molecules of Span 80 (top row) or DOSS (bottom row) per air/seawater interface, giving a DOR = 1:20. Black = C15 or C30; red = seawater; cyan = headgroup of surfactant molecules; yellow = tail group of surfactant molecules.

MD simulations with different overall values of DOR (1:10 and 1:5) were also performed; the results are very similar to those reported in Figures 6 and 7 and thus will not be shown for brevity. Both the density profiles and the snapshots indicate that the n-alkanes and the Corexit surfactants prefer to stay at the air/seawater interfaces, as suggested by the PMF results discussed in Section 3.2. The density profiles indicate that the n-alkanes tend to remain at the air side of the interface, while the surfactants prefer to stay between the n-alkanes and the seawater side of the interface. Not surprisingly, the head groups of Span 80 and DOSS tend to remain closer to the seawater side of the interface, while the tail groups overlap with the nalkanes. These observations are consistent with the snapshots shown in Figure 7. In our simulations we monitored the orientation of the surfactant molecules DOSS and Span 80 at the air/seawater interface, as given by the angle θ. For DOSS (Figure 8) this angle is formed between the vector normal to the interface (zdirection) and the plane defined by the two carbon atoms at both ends of the tail and by the carbon atom connected to the G

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 9. Angle distribution of the surfactant molecules DOSS (left) and Span 80 (right) at the air/seawater interface at 298 K at two DOR values (1:5 and 1:20). All systems contain 40 molecules of C15 or C30 per interface.

properties of oil alkanes and surfactants at the air/seawater interface. This understanding is relevant to the possible transport of oil spill matter into the atmosphere, adsorbed on the surface of sea sprays and water droplets. Our potential of mean force (PMF) calculations suggest that the presence of Span 80 makes the air/seawater interface more stable thermodynamically for the adsorption of the n-alkanes C15 and C30, as compared to the case where DOSS is present or no surfactant is present in the system. These trends are signaled by the depth of the PMF minimum when the molecule of interest is at the air/seawater interface, as well as the difference between such minimum and the PMF value observed when the molecule is in the seawater phase (which can be seen as a measure of the thermodynamic incentive for a molecule to move from the seawater phase and into the interface). Therefore, thermodynamics suggests that Span 80 makes these oil hydrocarbons to be more likely to be carried out to the atmosphere adsorbed at the surface of water droplets, followed by DOSS and last by the situation where no surfactants are present. These simulation trends are in qualitative agreement with our experimental results, which indicate that larger amounts of oil hydrocarbons were ejected when Span 80 is present in the system as compared to when DOSS is added. Our PMF results also indicate that Span 80 has a larger thermodynamic incentive to move from the seawater phase and into the air/seawater interface, as compared to DOSS. This observation is in qualitative agreement with the experimental measurements, which indicate that Span 80 is ejected in larger quantities than DOSS. Nevertheless, we emphasize that both

Figure 10. Definition of the angular parameter α, representing the intramolecular angle between the two tails in DOSS (left) and between the headgroup and the tail group in Span 80 (right). Gray = carbon; red = oxygen; yellow = sulfur; hydrogen atoms not shown for clarity.

slightly flatter when the n-alkane length decreases from C30 to C15, but at DOR = 1:20, the angle distributions are not significantly affected by changes in the length of the n-alkanes. The angle distribution in DOSS seems to be more sensitive to these variables: if the oil phase is pure C15, changing the DOR from 1:20 to 1:5 causes the distribution to become slightly flatter, but the opposite trend is observed if the oil phase is pure C30 (i.e., the angle distribution becomes less wide at DOR = 1:5)

4. CONCLUDING REMARKS We performed experiments using a laboratory scale aerosolization reactor, as well as computer simulations, with the objective of determining the individual effects of two of the main components of Corexit, namely, the surfactants DOSS and Span 80, on the aerosolization of oil spill matter to the atmosphere. With our combined experimental−simulation study, we also aimed at fundamentally understanding the

Figure 11. Distribution of the intramolecular angle α of the surfactant molecules DOSS (left) and Span 80 (right) at the air/seawater interface at 298 K at two DOR values (1:5 and 1:20). All systems contain 40 molecules of C15 or C30 per interface. H

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

■

thermodynamics and transport processes play roles in the aerosolization of oil spill matter, while our molecular simulations only probe the thermodynamics in these systems. Classical MD simulations of systems containing oil and surfactants at the air/seawater interface suggest that Span 80 has a slight preference to align parallel to the interfaces with a coiled conformation at both DOR values. In contrast, DOSS predominantly adopts a perpendicular orientation with respect to the interface at a dispersant to oil ratio (DOR) of 1:20 but has a slight preference to lie parallel to the interfaces at a DOR = 1:5; in both cases, DOSS molecules have their tails wide open and stretched. Finally, the united-atom models used here are computationally less expensive than the all-atom models used in our previous study,21 providing us with a convenient set of models for future research ideas, such as studying the synergistic effects of two or more dispersant components at air/seawater and oil/ seawater interfaces, and the partitioning of dispersant components in oil/seawater systems.

■

REFERENCES

(1) The Economic and Environmental Impact of the BP Oil Spill in the Gulf States. http://www.thefederation.org/documents/24. BP%20Oil%20Spill-Currie.pdf (accessed May 2, 2016). (2) U.S. Chemical Safety and Hazard Investigation Board. Explosion and Fire at the Macondo Well. http://www.csb.gov/assets/1/7/Vol_ 2_Final_Version.pdf (accessed May 2, 2016). (3) Oil Budget Calculator Deepwater Horizon. http://www. restorethegulf.gov/sites/default/files/documents/pdf/OilBudgetCalc_ Full_HQ-Print_111110.pdf (accessed May 2, 2016). (4) Gray, J. L.; Kanagy, L. K.; Furlong, E. T.; Kanagy, C. J.; McCoy, J. W.; Mason, A.; Lauenstein, G. Presence of the Corexit Component Dioctyl Sodium Sulfosuccinate in Gulf of Mexico Waters After the 2010 Deepwater Horizon Oil Spill. Chemosphere 2014, 95, 124−130. (5) Prince, R. C.; Butler, J. D. A Protocol for Assessing the Effectiveness of Oil Spill Dispersants in Stimulating the Biodegradation of Oil. Environ. Sci. Pollut. Res. 2014, 21, 9506−9510. (6) Lessard, R. R.; DeMarco, G. The Significance of Oil Spill Dispersants. Spill Sci. Technol. Bull. 2000, 6, 59−68. (7) Oil Spill Dispersants: Efficacy and Effects; Ocean Studies Board, National Academies Press, 2005. (8) Kleindienst, S.; Seidel, M.; Ziervogel, K.; Grim, S.; Loftis, K.; Harrison, S.; Malkin, S. Y.; Perkins, M. J.; Field, J.; Sogin, M. L.; Dittmar, T.; Passow, U.; Medeiros, P. M.; Joye, S. B. Chemical Dispersants Can Suppress the Activity of Natural Oil-Degrading Microorganisms. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 14900− 14905. (9) Riehm, D. A.; McCormick, A. V. The Role of Dispersants’ Dynamic Interfacial Tension in Effective Crude Oil Spill Dispersion. Mar. Pollut. Bull. 2014, 84, 155−163. (10) Place, B. J.; Perkins, M. J.; Sinclair, E.; Barsamian, A. L.; Blakemore, P. R.; Field, J. A. Trace Analysis of Surfactants in Corexit Oil Dispersant Formulations and Seawater. Deep Sea Res., Part II 2016, 129, 273−281. (11) Kujawinski, E. B.; Kido Soule, M. C.; Valentine, D. L.; Boysen, A. K.; Longnecker, K.; Redmond, M. C. Fate of Dispersants Associated with the Deepwater Horizon Oil Spill. Environ. Sci. Technol. 2011, 45, 1298−1306. (12) Ramirez, C. E.; Batchu, S. R.; Gardinali, P. R. High Sensitivity Liquid Chromatography Tandem Mass Spectrometric Methods for the Analysis of Dioctyl Sulfosuccinate in Different Stages of an Oil Spill Response Monitoring Effort. Anal. Bioanal. Chem. 2013, 405, 4167− 4175. (13) Gray, J. L.; Kanagy, L. K.; Furlong, E. T.; McCoy, J. W.; Kanagy, C. J. Determination of the Anionic Surfactant Di(Ethylhexyl) Sodium Sulfosuccinate in Water Samples Collected from Gulf of Mexico Coastal Waters Before and After Landfall of Oil from the Deepwater Horizon Oil Spill, May to October, 2010. U.S. Geological Survey Open-File Report 2010-1318. http://pubs.usgs.gov/of/2010/1318/ (accessed May 2, 2016). (14) Gray, J. L.; Kanagy, L. K.; Furlong, E. T.; Kanagy, C. J.; McCoy, J. W.; Mason, A.; Lauenstein, G. Presence of the Corexit Component Dioctyl Sodium Sulfosuccinate in Gulf of Mexico Waters After the 2010 Deepwater Horizon Oil Spill. Chemosphere 2014, 95, 124−130. (15) Campo, P.; Venosa, A. D.; Suidan, M. T. Biodegradability of Corexit 9500 and Dispersed South Louisiana Crude Oil at 5 and 25 C. Environ. Sci. Technol. 2013, 47, 1960−1967. (16) White, H. K.; Lyons, S. L.; Harrison, S. J.; Findley, D. M.; Liu, Y.; Kujawinski, E. B. Long-Term Persistence of Dispersants Following the Deepwater Horizon Oil Spill. Environ. Sci. Technol. Lett. 2014, 1, 295−299. (17) de Gouw, J. A.; Middlebrook, A. M.; Warneke, C.; Ahmadov, R.; Atlas, E. L.; Bahreini, R.; Blake, D. R.; Brock, C. A.; Brioude, J.; Fahey, D. W.; et al. Organic Aerosol Formation Downwind from the Deepwater Horizon Oil Spill. Science 2011, 331, 1295−1299. (18) Middlebrook, A. M.; Murphy, D. M.; Ahmadov, R.; Atlas, E. L.; Bahreini, R.; Blake, D. R.; Brioude, J.; de Gouw, J. A.; Fehsenfeld, F. C.; Frost, G. J.; et al. Air Quality Implications of the Deepwater

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.6b04988. Additional methods details for experimental determination of ejection rates of n-alkanes and Corexit surfactants DOSS and Span 80; additional model details for Span 80; and additional PMF results (PDF)

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 617-373-2989. Present Address #

Department of Chemical Engineering, Northeastern University, Boston, Massachusetts 02115, United States. Author Contributions ⊥

These authors contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Gulf of Mexico Research Initiative (GoMRI), as part of C-MEDS (Consortium for the Molecular Engineering of Dispersant Systems, http:// dispersant.tulane.edu/). Additionally, this research was made possible in part by a grant from The Gulf of Mexico Research Initiative supporting the ECOGIG (Ecosystem Impacts of Oil & Gas Inputs to the Gulf, https://ecogig.org/) consortium, and in part by the National Institute of Environmental Health Sciences of the National Institutes of Health under Award Number T32ES007060. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This is ECOGIG manuscript number 434. Data are publicly available through the Gulf of Mexico Research Initiative Information & Data Cooperative (GRIIDC) at DOI: 10.7266/N73R0QW8. All molecular simulations were performed on the clusters from High Performance Computing at Louisiana State University (http://www.hpc.lsu.edu) and the Louisiana Optical Network Initiative (http://www.loni.org). I

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Horizon Oil Spill. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 20280− 20285. (19) Coe, H. Aerosol Chemistry and the Deepwater Horizon Spill. Science 2011, 331, 1273−1274. (20) Ehrenhauser, F. S.; Avij, P.; Shu, X.; Dugas, V.; Woodson, I.; Liyana-Arachchi, T.; Zhang, Z.; Hung, F. R.; Valsaraj, K. T. Bubble Bursting as an Aerosol Generation Mechanism During an Oil Spill in the Deep-Sea Environment: Laboratory Experimental Demonstration of the Transport Pathway. Environ. Sci.: Processes Impacts 2014, 16, 65−73. (21) Liyana-Arachchi, T. P.; Zhang, Z. H.; Ehrenhauser, F. S.; Avij, P.; Valsaraj, K. T.; Hung, F. R. Bubble Bursting as an Aerosol Generation Mechanism During an Oil Spill in the Deep-Sea Environment: Molecular Dynamics Simulations of Oil Alkanes and Dispersants in Atmospheric Air/Salt Water Interfaces. Environ. Sci.: Processes Impacts 2014, 16, 53−64. (22) Topics in Current Aerosol Research Vol. 2; Hidy, G. M.; Brock, J. R., Eds.; Pergamon Press: Oxford, 1971. (23) Spillane, M. C.; Monahan, E. C.; Bowyer, P. A.; Doyle, D. M.; Stabeno, P. J.: Whitecaps and Global Fluxes. In Oceanic Whitecaps and Their Role in Air-Sea Exchange Processes; Monahan, E. C., Niocaill, G. M., Eds.; D. Reidel Publishing Company: Dordrecht, Holland, 1986; pp 209−218. (24) Blanchard, D. C.; Liss, P. S.; Slinn, W. G. N. The Production, Distribution and Bacterial Enrichment of the Sea-Salt Aerosol. Air-Sea Exchange of Gases and Particles 1983, 108, 407−454. (25) Saint-Louis, R.; Pelletier, E. Sea-to-Air Flux of Contaminants Via Bubbles Bursting. An Experimental Approach for Tributyltin. Mar. Chem. 2003, 84, 211−224. (26) Andreas, E. L.; Edson, J. B.; Monahan, E. C.; Rouault, M. P.; Smith, S. D. The Spray Contribution to Net Evaporation from the Sea: A Review of Recent Progress. Boundary-Layer Meteorology 1995, 72, 3−52. (27) MacIntyre, F. Flow Patterns in Breaking Bubbles. J. Geophys. Res. 1972, 77, 5211−5228. (28) Brock, C. A.; Murphy, D. M.; Bahreini, R.; Middlebrook, A. M. Formation and Growth of Organic Aerosols Downwind of the Deepwater Horizon Oil Spill. Geophys. Res. Lett. 2011, 38, L17805. (29) Bahreini, R.; Middlebrook, A. M.; Brock, C. A.; de Gouw, J. A.; McKeen, S. A.; Williams, L. R.; Daumit, K. E.; Lambe, A. T.; Massoli, P.; Canagaratna, M. R.; et al. Mass Spectral Analysis of Organic Aerosol Formed Downwind of the Deepwater Horizon Oil Spill: Field Studies and Laboratory Confirmations. Environ. Sci. Technol. 2012, 46, 8025−8034. (30) Keene, W. C.; Maring, H.; Maben, J. R.; Kieber, D. J.; Pszenny, A. A. P.; Dahl, E. E.; Izaguirre, M. A.; Davis, A. J.; Long, M. S.; Zhou, X.; Smoydzin, L.; Sander, R. Chemical and Physical Characteristics of Nascent Aerosols Produced by Bursting Bubbles at a Model Air-Sea Interface. J. Geophys. Res. 2007, 112, D21202. (31) Tang, X.; Huston, K. J.; Larson, R. G. Molecular Dynamics Simulations of Structure-Property Relationships of Tween® 80 Surfactants in Water and at Interfaces. J. Phys. Chem. B 2014, 118, 12907−12918. (32) Ayorinde, F. O.; Gelain, S. V.; Johnson, J. H.; Wan, L. W. Analysis of Some Commercial Polysorbate Formulations Using Matrix-Assisted Laser Desorption/Ionization Time-Of-Flight Mass Spectrometry. Rapid Commun. Mass Spectrom. 2000, 14, 2116−2124. (33) Avij, P.; Perkins, M. J.; Huang, H.; Hung, F. R.; Field, J. A.; Valsaraj, K. T. Ejection of surfactants from Corexit 9500A/9527A and alkanes from oil into the atmosphere via aerosolization from the sea surface. Environ. Sci.: Processes Impacts 2016, submitted. (34) Smith, J. S.; Burns, L. F.; Valsaraj, K. T.; Thibodeaux, L. J. Bubble Column Reactors for Wastewater Treatment. 2. The Effect of Sparger Design on Sublation Column Hydrodynamics in the Homogeneous Flow Regime. Ind. Eng. Chem. Res. 1996, 35, 1700− 1710. (35) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E. GROMACS 4: Algorithms for highly efficient, load-balanced, and scalable molecular simulation. J. Chem. Theory Comput. 2008, 4, 435−447.

(36) Berendsen, H. J. C.; Grigera, J. R.; Straatsma, T. P. The Missing Term in Effective Pair Potentials. J. Phys. Chem. 1987, 91, 6269−6271. (37) Vacha, R.; Siu, S. W. I.; Petrov, M.; Bockmann, R. A.; BaruchaKraszewska, J.; Jurkiewicz, P.; Hof, M.; Berkowitz, M. L.; Jungwirth, P. Effects of Alkali Cations and Halide Anions on the DOPC Lipid Membrane. J. Phys. Chem. A 2009, 113, 7235−7243. (38) Chowdhary, J.; Ladanyi, B. M. Molecular Dynamics Simulation of Aerosol-OT Reverse Micelles. J. Phys. Chem. B 2009, 113, 15029− 15039. (39) Wick, C. D.; Martin, M. G.; Siepmann, J. I. Transferable Potentials for Phase Equilibria. 4. United-Atom Description of Linear and Branched Alkenes and Alkylbenzenes. J. Phys. Chem. B 2000, 104, 8008−8016. (40) Chen, B.; Potoff, J. J.; Siepmann, J. I. Monte Carlo Calculations for Alcohols and their Mixtures with Alkanes. Transferable Potentials for Phase Equilibria. 5. United-Atom Description of Primary, Secondary, and Tertiary Alcohols. J. Phys. Chem. B 2001, 105, 3093−3104. (41) Stubbs, J. M.; Potoff, J. J.; Siepmann, J. I. Transferable Potentials for Phase Equilibria. 6. United-Atom Description for Ethers, Glycols, Ketones, and Aldehydes. J. Phys. Chem. B 2004, 108, 17596−17605. (42) Maerzke, K. A.; Schultz, N. E.; Ross, R. B.; Siepmann, J. I. TraPPE-UA Force Field for Acrylates and Monte Carlo Simulations for their Mixtures with Alkanes and Alcohols. J. Phys. Chem. B 2009, 113, 6415−6425. (43) Keasler, S. J.; Charan, S. M.; Wick, C. D.; Economou, I. G.; Siepmann, J. I. Transferable Potentials for Phase Equilibria−United Atom Description of Five-and Six-Membered Cyclic Alkanes and Ethers. J. Phys. Chem. B 2012, 116, 11234−11246. (44) Bayly, C. I.; Cieplak, P.; Cornell, W.; Kollman, P. A. A WellBehaved Electrostatic Potential Based Method Using Charge Restraints for Deriving Atomic Charges: The RESP Model. J. Phys. Chem. 1993, 97, 10269−10280. (45) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A Second Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 1995, 117, 5179−5197. (46) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development and Testing of a General Amber Force Field. J. Comput. Chem. 2004, 25, 1157−1174. (47) Scott, J. C.: The Effect of Organic Films on Water Surface Motions. In Oceanic Whitecaps; Monahan, E. C., Niocaill, G. M., Eds.; Springer: Netherlands, 1986; Vol. 2; pp 159−165. (48) The HLB System: A Time-saving Guide to Emulsifier Selection. ICI Americas, Incorporated, 1984. (49) Liu, H.; Li, S.; Wang, Y.; Han, F.; Dong, Y. Bicontinuous WaterAOT/Tween85-Isopropyl Myristate Microemulsion: A New Vehicle for Transdermal Delivery of Cyclosporin A. Drug Dev. Ind. Pharm. 2006, 32, 549−557. (50) Lewandowski, H.; Schwuger, M. J. α-Sulfomonocarboxylic Ester. In Novel Surfactants: Preparation, Applications, and Biodegradability, 2nd ed.; Holmberg, K., Ed.; CRC Press, 2003; pp 440. (51) Boyd, J. v.; Parkinson, C.; Sherman, P. Factors Affecting Emulsion Stability, and the HLB Concept. J. Colloid Interface Sci. 1972, 41, 359−370. (52) Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. What is So Special About Aerosol-OT? 2. Microemulsion Systems. Langmuir 2000, 16, 8741−8748. (53) Huston, K. J.; Larson, R. G. Reversible and Irreversible Adsorption Energetics of Poly(ethylene glycol) and Sorbitan Poly(ethoxylate) at a Water/Alkane Interface. Langmuir 2015, 31, 7503− 7511. (54) Liyana-Arachchi, T. P.; Hansel, A. K.; Stevens, C.; Ehrenhauser, F. S.; Valsaraj, K. T.; Hung, F. R. Molecular Modeling of the Green Leaf Volatile Methyl Salicylate on Atmospheric Air/Water Interfaces. J. Phys. Chem. A 2013, 117, 4436−4443. (55) Liyana-Arachchi, T. P.; Valsaraj, K. T.; Hung, F. R. Molecular Simulation Study of the Adsorption of Naphthalene and Ozone on J

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Atmospheric Air/Ice Interfaces. J. Phys. Chem. A 2011, 115, 9226− 9236. (56) Liyana-Arachchi, T. P.; Zhang, Z.; Vempati, H.; Hansel, A. K.; Stevens, C.; Pham, A. T.; Ehrenhauser, F. S.; Valsaraj, K. T.; Hung, F. R. Green Leaf Volatiles on Atmospheric Air/Water Interfaces: A Combined Experimental and Molecular Simulation Study. J. Chem. Eng. Data 2014, 59, 3025−3035.

K

DOI: 10.1021/acs.jpca.6b04988 J. Phys. Chem. A XXXX, XXX, XXX−XXX