Tuan A. Ho

Simulation Insights for Graphene-Based Water Desalination Membranes

Molecular dynamics simulations were employed to study the transport of water and ions through pores created on the basal plane of one graphene sheet (GS). Graphene pore diameters ranged from 7.5 to 14.5 Å. Different pore functionalities were considered, obtained by tethering various functional groups to the terminal carbon atoms. The ease of ion and water translocation across the pores was monitored by calculating the potential of mean force along the direction perpendicular to the GS pore. The results indicate that effective ion exclusion can be achieved only using nonfunctionalized (pristine) pores of diameter ~7.5 Å, whereas the ions can easily penetrate pristine pores of diameters ~10.5 and 14.5 Å. Carboxyl functional groups can enhance ion exclusion for all pores considered, but the effect becomes less pronounced as both the ion concentration and the pore diameter increase. When compared to a carbon nanotube of similar pore diameter, our results suggest that GS pores functionalized with COO(-) groups are more effective in excluding Cl(-) ions from passing through the membrane. Our results suggest that narrow graphene pores functionalized with hydroxyl groups remain effective at excluding Cl(-) ions even at moderate solution ionic strength. The results presented could be useful for the design of water desalination membranes.

Interfacial water on crystalline silica: a comparative molecular dynamics simulation study

Understanding the properties of interfacial water at solid–liquid interfaces is important in a wide range of applications. Molecular dynamics is becoming a widespread tool for this purpose. Unfortunately, however, the results of such studies are known to strongly depend on the selection of force fields. It is, therefore, of interest to assess the extent by which the implemented force fields can affect the predicted properties of interfacial water. Two silica surfaces, with low and high surface hydroxyl density, respectively, were simulated implementing four force fields. These force fields yield different orientation and flexibility of surface hydrogen atoms, and also different interaction potentials with water molecules. The properties for interfacial water were quantified by calculating contact angles, atomic density profiles, surface density distributions, hydrogen bond density profiles and residence times for water near the solid substrates. We found that at low surface density of hydroxyl groups, the force field strongly affects the predicted contact angle, while at high density of hydroxyl groups, water wets all surfaces considered. From a molecular-level point of view, our results show that the position and intensity of peaks observed from oxygen and hydrogen atomic density profiles are quite different when different force fields are implemented, even when the simulated contact angles are similar. Particularly, the surfaces simulated by the CLAYFF force field appear to attract water more strongly than those simulated by the Bródka and Zerda force field. It was found that the surface density distributions for water strongly depend on the orientation of surface hydrogen atoms. In all cases, we found an elevated number of hydrogen bonds formed between interfacial water molecules. The hydrogen bond density profile does not depend strongly on the force field implemented to simulate the substrate, suggesting that interfacial water assumes the necessary orientation to maximise the number of water–water hydrogen bonds irrespectively of surface properties. Conversely, the residence time for water molecules near the interface strongly depends on the force field and on the flexibility of surface hydroxyl groups. Specifically, water molecules reside for longer times at contact with rigid substrates with high density of hydroxyl groups. These results should be considered when comparisons between simulated and experimental data are attempted.

Polarizability effects in molecular dynamics simulations of the graphene-water interface.

The importance of including the polarizability of both water and graphene in molecular dynamics simulations of the water/graphene system was quantified. A thin film of either rigid single point charge extended (SPC/E) water or polarizable simple 4-site water model with Drude polarizability (SWM4_DP) water on non-polarizable and polarizable graphene surfaces was simulated. The graphene surface was either maintained neutral or charged, positively and negatively. The results suggest that SPC∕E and SWM4_DP water models yield very similar predictions for the water structural properties on neutral non-polarizable graphene, although they yield slightly different dynamical properties of interfacial water on neutral non-polarizable graphene. More pronounced were the differences obtained when graphene was modeled with a polarizable force field. In particular, the polarizability of graphene was found to enhance the number of interfacial SWM4_DP water molecules pointing one of their OH bonds towards the neutral surface. Despite this structural difference, the dynamical properties predicted for the interfacial SWM4_DP water were found to be independent on polarizability as long as the polarizability of a carbon atom is smaller than α = 0.878 A. On charged graphene surfaces, the effect of polarizability of graphene on structural properties and some dynamical properties of SWM4_DP water is negligible because electrostatic forces due to surface charge dominate polarization forces, as expected. For all cases, our results suggest that the hydrogen bond network is insensitive to the polarizability of both water and graphene. Understanding how these effects will determine the accumulation of ions near neutral or charged graphene could have important implications for applications in the fields of energy storage and water desalination.

Molecular dynamics simulation of the graphene–water interface: comparing water models

In this work, different water models (i.e. SPC/E, TIP3P, TIP4P/2005, TIP5P, SPC/Fw, TIP4P/2005f and SWM4_DP) are implemented to simulate water on neutral, negatively charged and positively charged graphene. In all cases ambient conditions are considered. Structural and dynamical properties for water are calculated to quantify the differences among various water models. The results show that SPC/E, TIP4P/2005, SPC/Fw, TIP4P/2005f and SWM4_DP water models yield a similar structure for interfacial water on graphene, whether it is neutral, negatively charged or positively charged. TIP5P is the model whose predictions for the structure of the interface deviate the most from those of the other models. Although qualitatively the results are for the most part similar, a large quantitative variation is observed among the dynamical properties predicted when various water models are implemented. Although experimental data are not available to discriminate the most/least accurate of the model predictions, our results could be useful for comparing results for interfacial water obtained implementing different models. Such critical comparison will benefit practical applications such as the development of energy-storage and water-desalination devices (e.g. electric double-layer capacitors), among others.

Capacitance enhancement via electrode patterning

The necessity of increasing the energy density in electric double layer capacitors to meet current demand is fueling fundamental and applied research alike. We report here molecular dynamics simulation results for aqueous electrolytes near model electrodes. Particular focus is on the effect of electrode patterning on the structure of interfacial electrolytes, and on the potential drop between the solid electrodes and the bulk electrolytes. The latter is estimated by numerically integrating the Poisson equation using the charge densities due to water and ions accumulated near the interface as input. We considered uniform and patterned electrodes, both positively and negatively charged. The uniformly charged electrodes are modeled as graphite. The patterned ones are obtained by removing carbon atoms from the top-most graphene layer, yielding nanoscopic squares and stripes patterns. For simplicity, the patterned electrodes are effectively simulated as insulators (the charge remains localized on the top-most layer of carbon atoms). Our simulations show that the patterns alter the structure of water and the accumulation of ions at the liquid-solid interfaces. Using aqueous NaCl solutions, we found that while the capacitance calculated for three positively charged electrodes did not change much, that calculated for the negatively charged electrodes significantly increased upon patterning. We find that both water structure and orientation, as well as ion accumulation affect the capacitance. As electrode patterning affects differently water structure and ion accumulation, it might be possible to observe ion-specific effects. These results could be useful for advancing our understanding of electric double layer capacitors, capacitive desalination processes, as well as of fundamental interfacial electrolytes properties.

A pore scale study describing the dynamics of slickwater distribution in shale gas formations following hydraulic fracturing

Hydraulic fracturing with slickwater to stimulate shale gas wells is routinely employed to enable increased contact with larger reservoir volumes and has the advantages of lower cost, the ability to create larger and more complex fractures, less formation damage and easier cleanup. However, a common observation is that during flow back only 10 to 20% of the frac water is recovered, even though a typical stimulation job requires several million gallons of water. Although there have been some attempts to address this phenomenon, the associated theories are lacking in scientific rigor. Due to the nanoporous nature of shales where pore proximity effects and strong inter-molecular interactions may dominate, a fundamental pore-level analysis is employed in this work to better understand and leverage the dynamics of the physiochemical processes during and after fracturing. By varying pore size in organic and inorganic pores in shales, we study the dynamics of water and gas molecules, as well as that of ions. The results of our study demonstrate that the mechanics of water entrapment and the water and ions distribution are strongly linked to the pore-surface mineralogy. Understanding the placement and distribution of frac water in both organic and inorganic pores in shales will potentially help in improved forecasting of well performance and address concerns related to the contamination of groundwater resources. Copyright 2013, Society of Petroleum Engineers.

A Pore Scale Study of Slickwater Systems in Shale Reservoirs: Implications for Frac-Water Distribution and Produced Water Salinity

Pore-level molecular dynamics simulation studies are conducted towards an understanding of poor recovery of fracwater, progressive increase in produced water salinity, and identification of potential trapping mechanisms for fracwater and its influence on long-term well productivity in shale gas and oil reservoirs. The kerogen pores of shales are respesented by two organic pore models. The first model containing only carbon is intended to mimic the nature of highly mature kerogen. The second model helps understanding of the fluid behavior in partially mature shales containining oxygenated functional groups with non-zero oxygen to carbon ratio. The maturation processes of these kerogen models are described by means of a molecular dynamics simulation. These models are shown to describe effectively the essential structural features obseved in SEM images which indicate surface roughness, tortuous paths, material disorders, and imperfect pore openings of kerogen pores, and are therefore superior to the frequently assumed graphene slit pore systems. The effect of maturation, pore surface mineralogy, and pore roughness on the wettability characteristics of organic kerogen pores is delineated. Distribution of saline water in organic and inorganic pores is described as a function of pore size and morphology. These pore-scale studies reveal important insights about the distribution of dissolved ions and water in organic pores, and the frac-water distribution and produced water salinity following hydraulic fracturing. Introduction Shale gas and oil development activities have continually undergone several stages of refinement and continues to be driven by our ability to create extensive multi-stage hydraulic fracture treatments along several thousands of feet of horizontal laterals. Although these efforts have largely progressed successfully, unfortunately, our current understanding of the complex interplay of hydrocarbons and water in organic and inorganic shale nanopores is rather limited. Among the key questions remaining unanswered are related to the explanation of the poor recovery of frac-water, the progressive increase in produced water salinity, and the potential trapping mechanisms for frac-water and its influence on long-term well productivity. Shale formations are characterized by extremely low permeabilities in the orders of nanodarcies. Therefore, hydraulic fracturing treatments in combination with horizontal well completions are necessary to contact large volumes of the reservoir and to enhance well productivity. There are several field observations linked to the recovery of frac-fluids such as slickwater and the salinity of produced water. Generally, the recovery of slickwater is very low, even though several million gallons of fluid is initially injected to hydraulically fracture the well. This recovery may range from as little as 5% in the Haynesville shale to as much as 50% in the Barnett and Marcellus shales (King 2012). Additionally, we generally observe an increasing salinity of produced water over time. Even though the original injected slickwater is fresh, the salt concentration can be as high as 80,000 to 100,000 ppm at later times (King 2010). A better understanding of these issues may enable us to design better fracture treatments, to URTeC 2013 Page 1284