This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.
Article Cite This: ACS Omega 2018, 3, 11497−11503
http://pubs.acs.org/journal/acsodf
Effects of Size and Functionalization on the Structure and Properties of Graphene Oxide Nanoflakes: An in Silico Investigation Enxi Peng, Nevena Todorova, and Irene Yarovsky* School of Engineering, RMIT University, GPO Box 2476V, 3001 Melbourne, Victoria, Australia
Downloaded via 193.93.194.135 on September 20, 2018 at 17:25:31 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
S Supporting Information *
ABSTRACT: Graphitic nanoparticles, specifically, graphene oxide (GO) nanoflakes, are of major interest in the field of nanotechnology, with potential applications ranging from drug delivery systems to energy storage devices. These applications are possible largely because of the properties imparted by various functional groups attached to the GO surface by relatively simple production methods compared to pristine graphene. We investigated how varying the size and oxidation of GO flakes can affect their structural and dynamic properties in an aqueous solution. The all-atom modeling of the GO nanoflakes of different sizes suggested that the curvature and roughness of relatively small (3 × 3 nm) GO flakes are not affected by their degree of oxidation. However, the larger (7 × 7 nm) flakes exhibited an increase in surface roughness as their oxidation increased. The analysis of water structure around the graphitic nanoparticles revealed that the degree of oxidation does not affect the water dipole orientations past the first hydration layer. Nevertheless, oxygen functionalization induced a well-structured first hydration layer, which manifested in identifiable hydrophobic and hydrophilic patches on GO. The detailed all-atom models of GO nanoflakes will guide a rational design of functional graphitic nanoparticles for biomedical and industrial applications. ity,2,6 enzyme inhibitors, and drug delivery vehicles.2,6−9 More recently, several studies identified the potential of these nanoparticles to mediate the formation of amyloid protein fibrils, known for their implications in debilitating neurodegenerative diseases.10−12 In this paper, we explore the structural modeling of the GO nanoflakes for subsequent applications in controlling the amyloid fibril formation. Computational modeling provides direct access to the structural characterization of nanomaterials complementary to experiments.13,14 Several computational studies focused on the structure of GO nanoparticles, mostly with the aim of improving the mechanical properties of this nanomaterial.15−17 In particular, Khoei et al. demonstrated that increasing the proportion of the oxygen-containing functional groups also increased the C−C bond length of the hexagonal lattices of graphene, causing a distortion in the hexagonal lattice.15 This resulted in the alteration of the mechanical properties of the material such as shear modulus and elastic constants. Wei et al. modeled the wetting properties of GO relative to PG and found that the oxidized regions contributed to the spreading of water droplets across the surface.16 Controlling the proportion of the oxygen functionality can thus be used for modulating the water behavior on the GO surfaces. These studies suggest that
1. INTRODUCTION Graphitic nanomaterials have been a highly topical area of research in the past decade, with an array of possible applications ranging from drug delivery to biosensing and prevention of biofouling.1,2 In recent years, graphene oxide (GO), the surface-functionalized cousin of pristine graphene (PG), has become widely recognized as the more accessible nanomaterial relative to PG. GO is chemically synthesized using Hummer’s method,3 which was refined by Lerf and Klinowski et al.,4 where a mixture of KMnO4 and concentrated H2SO4 is reacted with graphite to produce GO with the chemical composition of C10O1(OH)1(COOH)0.5. The distribution of functional groups comprises two epoxy groups, two hydroxyl groups on both sides of the basal plane, and one carboxyl group on the edge for every 20 carbon atoms. This chemical structure allows GO flakes to be easily synthesized, transported, and handled, relative to native PG flakes. Graphene can also be functionalized with various types of functional groups other than oxygen through both covalent and noncovalent modifications.5 Because of the presence of the large and highly functionalized surface, GO allows for noncovalent interactions with biological molecules via electrostatics, π−π stacking, and hydrogen bonding. This functionalization also contributes to the desirable optical, mechanical, and electrical properties2 of GO, which are exploited in biomedical applications, for example, biosensors for the detection of biomolecules with high sensitivity and selectiv© 2018 American Chemical Society
Received: May 1, 2018 Accepted: September 7, 2018 Published: September 20, 2018 11497
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
crucial to reveal how these properties influence the GO interactions with the environment, primarily the behavior of the surrounding solvent and biomolecules. This paper reports a theoretical modeling investigation of the effects of GO nanoflake size and degree of functionalization on the GO surface structure, dynamics, and interactions in the aqueous solution relevant to biological environments. The outcomes of this study provide the atomistic insights needed for a rational design of novel graphitic nanomaterials with a desired behavior and an optimized and controlled function for biomedical applications.
the degree of oxidation has a significant effect on the mechanical and wetting properties of the graphitic flakes. In addition, several recent theoretical studies investigated the behavior of GO sheets and GO nanoflakes in the presence of biomolecules and solvents.7,18−25 These studies indicated that the highly functionalized GO interacted most favorably with the biomolecules via hydrogen bonding, whereas the GO with low oxygen functional group densities prevalently interacted through π−π stacking.16,17 The theoretical studies utilized various modeling approaches to simulate the bonded and nonbonded interatomic interactions of GO. The most common approach for modeling GO7,16,19,22,24 is to use the LJ sp2 carbon potentials from AMBER96, the bond angle parameters from CHARMM27,26 and all other interaction parameters from the OPLS-AA force field.27 Although this approach may be adequate to model simple homogeneous systems, it has not been validated for modeling GO interactions with biomolecules. In addition, the CHARMM force field was used more recently by Bansal et al. for the modeling of aqueous GO dispersion.28 This study concluded that the functional groups on the GO surfaces prevented the GO aggregation in water through strong repulsive forces, which were shown to increase with an increasing concentration of functional groups. The analysis of hydrogen bonding data also indicated that both epoxy- and hydroxyl-functionalized GO surfaces were stabilized via the intra- and interlayer watermediated H-bonds.28 Similarly, Tang et al.18 highlighted that the type of functional group, the oxygen content, and the initial distance between the GO flakes all play a significant role in the aggregation kinetics of GO nanoparticles. Alternatively, the GO−biomolecular complexes have been simulated using material-orientated force fields such as COMPASS,17,29,30 originally developed for modeling organic (polymer) interfaces with metal oxides. For example, Rahmani et al. utilized the COMPASS II force field to model the adsorption of various amino acids on GO surfaces.29 Their study concluded that the surface functionalization of GO disrupted the π−π stacking between the surface and the aromatic residues of the amino acids. Class II force fields are generally computationally expensive, as the functional form of the Hamiltonian includes anharmonic terms, such as quartic stretching and quartic angle bending as well as a variety of other important intramolecular coupling interactions.31 However, COMPASS has undergone limited development and testing for biomolecular systems, and, as a result, the computational cost and biomolecular force field deficiencies do not generally justify its application to sampling complex biological phenomena. Unlike PG, GO is highly soluble in water; thus, its influence on the surrounding water structure is one of the important factors that will ultimately control its interactions with biomolecules.32−35 As a result, it is important to understand the mechanisms of interactions that govern the structure and dynamic behaviors of the GO nanoflakes in an aqueous environment. Moreover, as demonstrated by previous studies10,11,20,36 the nanoparticle size, functionalization, and curvature play an important role in the biomolecular recognition and protein adsorption onto the surface of the nanoparticle. Despite the noteworthy research efforts on GO, there is a gap in understanding the interconnection between the size and the degree of oxidation of the GO nanoflakes and their physical and (bio)chemical properties. To exploit the unique properties of the GO particles in biomedicine, it is
2. RESULTS AND DISCUSSION 2.1. Surface Roughness and Flexibility of Graphitic Nanoflakes. The shape and curvature of nanoparticles were highlighted to be a major factor implicated in protein binding.36 The average surface roughness (or curvature) of all the nine graphitic models is shown in Figure 1, indicating a
Figure 1. Average roughness of graphitic nanoflakes as a function of size and degree of oxidation.
general trend of the average roughness increase with the increased size and degree of functionalization of the graphitic flakes. For the 3 nm flakes, there was a small increase (∼0.1 nm) in the surface roughness with increasing oxidation, suggesting that the nanoparticles are too small for the functional groups to induce the significant response in the curvature. By contrast, the presence of oxygen-containing groups in the larger (5 and 7 nm) rGO and GO flakes increased the curvature of the nanoparticle approximately twofold, compared to PG of the same size. Moreover, the 5 nm flakes of rGO and GO had a negligible difference (within the standard deviation) in their surface roughness, indicating that the curvature in rGO is largely induced by the edge functional groups. Overall, the 7 nm flakes exhibited the most significant increase in surface roughness as the degree of oxidation increased, with the PG measuring roughness of 0.9 ± 0.05 nm, and rGO and GO of 1.2 ± 0.07 nm and 1.6 ± 0.13 nm, respectively. To understand the relationship between the measured surface roughness and the flexibility of the particle, the rootmean square fluctuation (RMSF) of individual atoms in each graphitic flake was calculated for the final 50 ns of the simulated trajectories. The average RMSF values are presented in Figure 2, using the color scale ranging from 1 to 10. The blue-colored areas indicate a high RMSF, whereas the redcolored areas indicate low fluctuations. The ratio of the stiff (red) to mobile (blue) areas is presented below each plot. For all the 3 nm flakes, high fluctuations were observed at the 11498
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
dynamic differences of GO could affect protein folding and adsorption. Several studies have investigated the properties that contributed toward biomolecule adsorption to nanoparticles. For example, works by Hughes and Walsh had emphasized that exploiting interfacial water structuring was key for understanding selective behavior in peptides binding to multifaceted gold nanoparticles and graphene.33,37 It was also highlighted that a highly structured water layer provides favorable conditions for peptide adsorption at the nanostructured surfaces. Another study uncovered that intermittent interactions with water can mediate protein adsorption on surfaces, whereas longer lasting interactions control the diffusion of water around the adsorption sites.34,38 Another study highlighted that the hydration shells of unfolded proteins are more compressible than those on the folded ones contributing to protein denaturation because of pressure.39 Thus, by understanding the hydration patterns around the graphitic nanoparticles, we can help elucidate how they could influence the binding motifs of proteins on the nanoparticle surface.1,32,40−42 To better understand the water molecule behavior around the graphitic nanoparticles, GO−water radial distribution function (RDF) and water dipole analyses were performed. Figure 3 shows the solute (O)−water (H) RDFs for the rGO and GO flakes. For the PG flakes, the RDF was calculated between the carbon atoms and the water molecules. In Figure 3A, a formation of hydration shells around the rGO and GO flakes can be clearly identified by two characteristic peaks. The
Figure 2. Average fluctuations of individual atoms in each graphitic nanoflake. The color scale represents the normalized RMSF values. The numbers below each plot indicate the ratio of stiff to mobile surface areas.
corners of each flake, while the inner area of the flakes remained stable or less dynamic. This indicates that the minor variances in roughness between the 3 nm graphitic flakes are mostly attributed to the edge fluctuations. The 3 nm rGO sheet has significantly more stiff areas within the sheet as its stiff/mobile area ratio is 1.1, compared to 0.53 and 0.43 in the PG and GO flakes, respectively. This is likely due to the lack of edge functional groups to induce higher fluctuations around the edges as observed in the GO flakes, combined with the smaller size of the flake where oxidation stabilized the inner areas. Increases in fluctuations were also observed around the edges of the 5 nm graphitic flakes, with reduced fluctuations around the center of the flakes. Figure 2 shows that the introduction of functional groups reduces the flexibility of the rings, evidenced by less fluctuations exhibited by the densely functionalized areas. This oxidation-induced rigidity is the main contributor to the variances in roughness seen between different flake sizes. This effect was also observed by Khoei and Khorrami15 as discussed previously, where an increase in oxidation agents in graphene sheets leads to decreased elastic constants. The RMSF results for the 7 nm flakes revealed relatively larger areas of low fluctuations manifesting in the overall reduced dynamics compared to the smaller flakes. Therefore, it can be suggested that, for larger flakes, it is the increased concentration of oxygen functional groups on the graphitic surface that induced the localized structural rigidity. This also played an important role in the increased surface roughness, whereas the edge effects play a dominant role in the roughness and dynamics of the small size flakes. 2.2. Water Structure and Dynamics. Investigating the behaviors of water around the graphitic flakes can shed light on how water structuring resulting from chemical, structural, and
Figure 3. RDFs between water and graphene functional groups on differently sized GO flakes: (A) 3, (B) 5, and (C) 7 nm. 11499
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
first water shell for all the three flake sizes (3, 5, and 7 nm) can be found around 0.15−0.25 nm from the nanoparticle surface, with a second water shell forming between 0.25 and 0.35 nm. The RDF results for the PG flakes identified the first hydration shell at 0.25−0.3 nm from the nanoparticle because of graphene’s hydrophobic nature of hydration.43 For both rGO and GO flakes, the amplitude of each RDF peak is consistent between the two flake sizes for both water shells. This suggests that because of the low roughness and smaller hydrophobic areas on the 3 nm rGO and GO flakes, there is higher density of water molecules in the first water shells around these nanoparticles. A small peak indicating a possible third water layer between 0.5 and 0.7 nm can be seen for the 3 nm rGO and GO flakes. However, this characteristic third peak is not visible for the 5 and 7 nm flakes, as after 0.6 nm, the RDF approaches that of bulk water. It is important to note that the positions of the first two hydration shell peaks do not change with the flake size; however, the magnitude does. In Figure 3B,C, the RDF for GO is significantly higher than for rGO, which suggests that as the flake size and the degree of functionalization increase, more water molecules are packed into the first hydration shell. These results demonstrate that the graphitic nanoflakes can be tailored to have different curvatures and provide varying degrees of affinity toward water, which can be exploited to mediate protein binding. Furthermore, water dipole analysis was performed for all simulation systems in order to determine how surface functional groups affect water orientation in each of the observed hydration layers (see figures in the Supporting Information). The dipole orientation of each water molecule within each hydration layer and bulk solution was calculated (measured relative to Z-axis). For the 3 and 5 nm particles, the dipole angle distributions of the first and second hydration layers as well as the bulk water all have a characteristic bell shape. The peak of this bell curve rests around a 90° angle, indicating that the majority of water molecules in these hydration layers are not orientationally influenced by the surface functional groups of the smaller (3 and 5 nm) graphitic flakes. By contrast, the water dipole angle distribution for the 7 nm GO flake exhibited a rougher profile compared to the smaller nanoflakes with notable peaks between 40°−75° and 120°−160°. This suggests that the highly oxidized larger flakes induce water dipole orientation because of the significant number of water interactions with hydrophilic functional groups on the flake surfaces and edges. However, this induced dipole orientation does not extend past the first water layer as observed in Figure S3. Although no induced dipole orientation was seen across any of the simulated flakes below the 7 nm size, we observed an increase in bound water with an increasing degree of oxidation across all nanoparticle sizes, in line with the RDF results shown in Figure 3. Overall, it can be suggested that for the GO nanoflakes studied here, the surface functionalization does provide interaction sites for water; however, these interactions are not sufficient to induce dipole orientation effects past the first hydration layer and into bulk water. 2.3. Solvent-Accessible Surface Area. The solventaccessible surface area (SASA) of each graphitic model was also calculated and expressed in terms of water-covered surface fractions (%), as illustrated in Figure 4. It can be clearly seen that as the degree of functionalization increased, the percentage of water coverage also increased. For GO flakes, the average covered fraction of the surface by the first water
Figure 4. Average SASAs for the GO models expressed in terms of water coverage fraction (%). A representative frame was selected from the equilibrium portion of the molecular dynamics (MD) trajectories to visualize the first hydration layer.
layer is approximately ∼88 ± 1−2% across all the three flake sizes. As the degree of functionalization decreased to rGO and PG, a significant drop in water coverage was observed. Specifically, the water coverage range was ∼60−70 ± 3−4% for rGO and ∼42−53 ± 3−6% for PG. Interestingly, the increased size did not increase the coverage. In fact, at a lower functionalization, the degree of water coverage decreased with an increase in the flake size. The 3 nm rGO flake had an average coverage of 71 ± 4%, which reduced to 59 ± 3% for the 7 nm flake. A similar trend was also observed for the PG flakes, with 53 ± 3% and 42 ± 6% coverage fractions for 3 and 7 nm flakes, respectively. This decrease in water coverage can be attributed to the significant increases in the surface area of hydrophobic patches on either side of the basal plane. This observation explains the aggregation behavior illustrated by Bansal et al.,28 where repulsive interactions between the flakes increased with increasing functionalization. As a result, more water was displaced from the surface of less functionalized flakes allowing for smaller interlayer spacing to occur, driven by hydrophobic surface association. As highlighted by Tang et al.,18 the initial distance between the graphitic flakes plays an important role in their aggregation velocity. Therefore, the increased fraction of hydrophobic patches observed here for less functionalized GO flakes can explain the experimentally observed aggregation behaviors.18,28
3. CONCLUSIONS The MD simulation characterized the surface roughness of a series of nanoscale PG flakes with various degrees of functionalization, as well as the characterized water behavior around these graphitic nanoparticles. Our results showed that the roughness of the GO flakes increases with their size and degree of functionalization. In addition, increasing the surface 11500
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
functionalization greatly increased the GO affinity to water, increasing the ratio of hydrophilic to hydrophobic surface areas as the degree of oxidation increased. However, the affinity for water did not cause any water structuring past the first water layer, with the effects of functional groups on water dipole alignment only evident for the 7 nm GO flake. The findings of this study contribute to understand the structure dynamics and hydration of graphitic nanoparticles. The ability to control the shape and hydration level of the graphitic nanoparticles can lead to the design and optimization of nanomaterials to modulate the protein binding in biomedical applications.
4. METHODOLOGY 4.1. Graphitic Nanoflake Models. The GO flakes of different sizes and degrees of oxidation were built using the BIOVIA Materials Studio software suite.44 In order to model GO with varying functionalities, a script was generated to randomly assign functional groups to a PG surface of [3 × 3] nm, [5 × 5] nm, and [7 × 7] nm in size (herein simply denoted as 3, 5, and 7 nm particles for ease of discussion) in accordance with the chemical composition described in the literature.4 These GO flake sizes are consistent with the applications of graphitic particles in a biologically relevant environment, where size compatibility is required to study the peptide/protein adsorption on the particle surface. We ensured that the induced randomness in the functional group distribution did not violate any physicochemical principles.45,46 Initial geometry optimization was then performed using the COMPASS II47 force field to ensure no steric clashes existed. A total of three oxidation states of GO were created for each nanoflake size: PG, reduced GO (rGO with the C/O ratio of 10:1), and GO (GO with the C/O ratio of 5:1). The degree of oxidation used in this study to model the GO and rGO nanoflakes are in line with the experimentally determined ratios of 4:1 for GO48 and 10:1 for rGO.49 This generated nine different graphene nanoflake models for subsequent simulations in an explicit solvent (Figure 5). The bonded interactions for GO were taken from the CGenFF force field,50,51 which utilizes an atom-typing approach whereby the parameters are assigned through analogies with an existing database of molecules. This approach was applied in recent studies investigating the accuracies of force fields for modeling the chemical and physical properties of GO.52 Such a method of parameterization avoids the need to combine the parameters from various other force fields.7,19,22,24 It is important to recognize that when the interactions between a nanoparticle and the biological molecules are of interest, a consistent set of interatomic potentials needs to be used to model each individual system component as well as the entire nano−bio complex.14,53−56 With this in mind, the choice of CGenFF, which is compatible with the CHARMM group of biological force fields, is appropriate for applications to nano-biosystems aimed as a follow-up of this work.57 In this study, the bonded GO parameters from CGenFF were plugged into CHARMM22* for all simulations. The QEq algorithm58 was used to determine the partial atomic charges based on the experimental zeta-potential of GO as measured by Konkena and Vasudevan for GO at varying pH values.59 Specifically, the zeta-potential of GO at neutral pH (−44 mV) was used, which resulted in a total surface charge of −0.04 e/nm2, and neutralized by counterions. The partial charges for PG and
Figure 5. Model parameters investigated in this study.
rGO were equilibrated to produce a neutral particle, in line with a standard simulation practice.24 4.2. Simulation Details. All simulations were performed using the GROMACS 5.0.5 software package,60−62 using classical MD with interatomic interactions parameterized as described in the above section. The van der Waals and electrostatic interactions were truncated at 1 nm with longrange electrostatics calculated by the particle mesh Ewald summation method.63 Each graphitic particle was suspended in a cubic simulation box sufficiently large (allowing at least 1.5 nm separation between the particle and the edge of the box) to prevent mirror image interactions. In vacuo energy minimization was performed to eliminate any steric clashes using the steepest descent algorithm. The simulation box was then solvated with TIP3P water64 and another energy minimization was performed. An MD simulation at a constant number of particles, constant volume, and temperature (NVT ensemble) was conducted for 1 ns to equilibrate the water molecules around the graphitic nanoparticle. Following this, the MD simulations were performed for 200 ns for each system under the isothermal−isobaric conditions (NPT ensemble). The constant pressure and temperatures were maintained via the Berendsen barostat65 and v-rescale thermostat,66 respectively. The LINCS algorithm61 was applied to constrain the bond lengths to their equilibrium values, which allowed for a simulation time step of 2 fs to be used. 4.3. Data Analysis. All analyses were carried out on the final 50 ns of the simulated trajectory, with the convergence being confirmed by the RMSD (within 0.3 nm) of the graphitic structure and total energy reaching a plateau. The average surface roughness for each graphitic system was measured using the 10-point roughness method,67 where Z-axis coordinates of every carbon atom relative to the starting structure are used, ignoring the “out-of-plane” and diffusive motion. The five maximum and five minimum values of the Zaxis coordinates for each frame were then determined, and the 11501
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
phospholipids from Escherichia coli membranes by graphene nanosheets. Nat. Nanotechnol. 2013, 8, 594−601. (9) Gao, X.; Jiang, D.-e.; Zhao, Y.; Nagase, S.; Zhang, S.; Chen, Z. Theoretical Insights into the Structures of Graphene Oxide and Its Chemical Conversions Between Graphene. J. Comput. Theor. Nanosci. 2011, 8, 2406−2422. (10) Li, Q.; Liu, L.; Zhang, S.; Xu, M.; Wang, X.; Wang, C.; Besenbacher, F.; Dong, M. Modulating Aβ33-42Peptide Assembly by Graphene Oxide. Chemistry 2014, 20, 7236−7240. (11) Wang, J.; Cao, Y.; Li, Q.; Liu, L.; Dong, M. Size Effect of Graphene Oxide on Modulating Amyloid Peptide Assembly. Chemistry 2015, 21, 9632−9637. (12) Mesarič, T.; Baweja, L.; Drašler, B.; Drobne, D.; Makovec, D.; Dušak, P.; Dhawan, A.; Sepčić, K. Effects of surface curvature and surface characteristics of carbon-based nanomaterials on the adsorption and activity of acetylcholinesterase. Carbon 2013, 62, 222−232. (13) Makarucha, A. J.; Todorova, N.; Yarovsky, I. Nanomaterials in biological environment: A review of computer modelling studies. Eur. Biophys. J. 2011, 40, 103−115. (14) Charchar, P.; Christofferson, A. J.; Todorova, N.; Yarovsky, I. Understanding and Designing the Gold-Bio Interface: Insights from Simulations. Small 2016, 12, 2395−2418. (15) Khoei, A. R.; Khorrami, M. S. Mechanical properties of graphene oxide: A molecular dynamics study. Fuller. Nanotub. Car. N. 2016, 24, 594−603. (16) Wei, N.; Lv, C.; Xu, Z. Wetting of graphene oxide: a molecular dynamics study. Langmuir 2014, 30, 3572−3578. (17) Zhang, J.; Jiang, D. Molecular dynamics simulation of mechanical performance of graphene/graphene oxide paper based polymer composites. Carbon 2014, 67, 784−791. (18) Tang, H.; Liu, D.; Zhao, Y.; Yang, X.; Lu, J.; Cui, F. Molecular Dynamics Study of the Aggregation Process of Graphene Oxide in Water. J. Phys. Chem. C 2015, 119, 26712−26718. (19) Chen, J.; Chen, L.; Wang, Y.; Chen, S. Molecular dynamics simulations of the adsorption of DNA segments onto graphene oxide. J. Phys. D Appl. Phys. 2014, 47, 505401. (20) Baweja, L.; Balamurugan, K.; Subramanian, V.; Dhawan, A. Hydration patterns of graphene-based nanomaterials (GBNMs) play a major role in the stability of a helical protein: a molecular dynamics simulation study. Langmuir 2013, 29, 14230−14238. (21) Sun, X.; Feng, Z.; Hou, T.; Li, Y. Mechanism of graphene oxide as an enzyme inhibitor from molecular dynamics simulations. ACS Appl. Mater. Interfaces 2014, 6, 7153−7163. (22) Chen, J.; Zhou, G.; Chen, L.; Wang, Y.; Wang, X.; Zeng, S. Interaction of Graphene and its Oxide with Lipid Membrane: A Molecular Dynamics Simulation Study. J. Phys. Chem. C 2016, 120, 6225−6231. (23) Borthakur, P.; Boruah, P. K.; Hussain, N.; Sharma, B.; Das, M. R.; Matić, S.; Ř eha, D.; Minofar, B. Experimental and Molecular Dynamics Simulation Study of Specific Ion Effect on the Graphene Oxide Surface and Investigation of the Influence on Reactive Extraction of Model Dye Molecule at Water-Organic Interface. J. Phys. Chem. C 2016, 120, 14088−14100. (24) Shih, C.-J.; Lin, S.; Sharma, R.; Strano, M. S.; Blankschtein, D. Understanding the pH-Dependent Behavior of Graphene Oxide Aqueous Solutions: A Comparative Experimental and Molecular Dynamics Simulation Study. Langmuir 2011, 28, 235−241. (25) Guo, J.; Li, J.; Zhang, Y.; Jin, X.; Liu, H.; Yao, X. Exploring the influence of carbon nanoparticles on the formation of beta-sheet-rich oligomers of IAPP(2)(2)(-)(2)(8) peptide by molecular dynamics simulation. PLoS One 2013, 8, No. e65579. (26) Patra, N.; Wang, B.; Král, P. Nanodroplet activated and guided folding of graphene nanostructures. Nano Lett. 2009, 9, 3766−3771. (27) Tummala, N. R.; Striolo, A. Role of Counterion Condensation in the Self-Assembly of SDS Surfactants at the Water−Graphite Interface. J. Phys. Chem. B 2008, 112, 1987−2000.
difference between the average maximum and minimum values was taken as the roughness measured of the flake at that specific frame. Atomic RMSF analysis was performed on the aligned trajectories (as described above) to determine the effects of size and functionalization on the flake dynamics. In addition, RDFs between water molecules and the functional groups and carbon atoms, as well as water dipole analyses, were utilized to better understand the water structuring around the graphitic particles and its role in the observed curvature patterns and dynamics of the GO flakes. The SASA of each flake was measured using the VMD SASA analysis tool.68
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.8b00866. Water dipole orientation analysis within the first two hydration shells and bulk water for each flake size and functionalization degree (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: +61 3 9925 2571 (I.Y.). ORCID
Irene Yarovsky: 0000-0002-4033-5150 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS I.Y. acknowledges the Australian Research Council for financial support under the Discovery Project scheme (DP140101888 and DP170100511). This research was undertaken with the assistance of resources from the National Computational Infrastructure (NCI), grant number (e87), the Pawsey Supercomputing Centre, and Melbourne Bioinformatics, Australia. Dr. Adam J. Makarucha and Dr. M. Harunur Rashid are acknowledged for their guidance at the early stage of this research.
■
REFERENCES
(1) Linklater, D. P.; Baulin, V. A.; Juodkazis, S.; Ivanova, E. P. Mechano-bactericidal mechanism of graphene nanomaterials. Interface Focus 2018, 8, 20170060. (2) Zhao, J.; Liu, L.; Li, F. Graphene OxidePhysics and Applications; Springer: London, 2015; Vol. 1, p 161. (3) Hummers, W. S.; Offeman, R. E. Preparation of Graphitic Oxide. J. Am. Chem. Soc. 1958, 80, 1339. (4) Lerf, A.; He, H.; Forster, M.; Klinowski, J. Structure of Graphite Oxide Revisited. J. Phys. Chem. B 1998, 102, 4477−4482. (5) Kuila, T.; Bose, S.; Mishra, A. K.; Khanra, P.; Kim, N. H.; Lee, J. H. Chemical functionalization of graphene and its applications. Prog. Mater Sci. 2012, 57, 1061−1105. (6) Celik, N.; Manivannan, N.; Balachandran, W. Graphene-based biosensors: methods, analysis and future perspectives. IET Circ. Dev. Syst. 2015, 9, 434−445. (7) Chen, J.; Wang, X.; Dai, C.; Chen, S.; Tu, Y. Adsorption of GA module onto graphene and graphene oxide: A molecular dynamics simulation study. Physica E Low Dimens. Syst. Nanostruct. 2014, 62, 59−63. (8) Tu, Y.; Lv, M.; Xiu, P.; Huynh, T.; Zhang, M.; Castelli, M.; Liu, Z.; Huang, Q.; Fan, C.; Fang, H.; Zhou, R. Destructive extraction of 11502
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503
ACS Omega
Article
(28) Bansal, P.; Panwar, A. S.; Bahadur, D. Molecular-Level Insights Into the Stability of Aqueous Graphene Oxide Dispersions. J. Phys. Chem. C 2017, 121, 9847−9859. (29) Rahmani, F.; Nouranian, S.; Mahdavi, M.; Al-Ostaz, A. Molecular simulation insights on the in vacuo adsorption of amino acids on graphene oxide surfaces with varying surface oxygen densities. J. Nanopart. Res. 2016, 18, 320. (30) Shao, G.; Lu, Y.; Wu, F.; Yang, C.; Zeng, F.; Wu, Q. Graphene oxide: the mechanisms of oxidation and exfoliation. J. Mater. Sci. 2012, 47, 4400−4409. (31) Hwang, M. J.; Stockfisch, T. P.; Hagler, A. T. Derivation of Class II Force Fields. 2. Derivation and Characterization of a Class II Force Field, CFF93, for the Alkyl Functional Group and Alkane Molecules. J. Am. Chem. Soc. 1994, 116, 2515−2525. (32) Penna, M.; Ley, K.; Maclaughlin, S.; Yarovsky, I. Surface heterogeneity: a friend or foe of protein adsorption - insights from theoretical simulations. Faraday Discuss. 2016, 191, 435−464. (33) Hughes, Z. E.; Walsh, T. R. What makes a good graphenebinding peptide? Adsorption of amino acids and peptides at aqueous graphene interfaces. J. Mater. Chem. B 2015, 3, 3211−3221. (34) Ley, K.; Christofferson, A.; Penna, M.; Winkler, D.; Maclaughlin, S.; Yarovsky, I. Surface-water Interface Induces Conformational Changes Critical for Protein Adsorption: Implications for Monolayer Formation of EAS Hydrophobin. Front. Mol. Biosci. 2015, 2, 64. (35) Penna, M. J.; Mijajlovic, M.; Biggs, M. J. Molecular-level understanding of protein adsorption at the interface between water and a strongly interacting uncharged solid surface. J. Am. Chem. Soc. 2014, 136, 5323−5331. (36) Todorova, N.; Makarucha, A. J.; Hine, N. D. M.; Mostofi, A. A.; Yarovsky, I. Dimensionality of carbon nanomaterials determines the binding and dynamics of amyloidogenic peptides: multiscale theoretical simulations. PLOS Comput. Biol. 2013, 9, No. e1003360. (37) Wright, L. B.; Palafox-Hernandez, J. P.; Rodger, P. M.; Corni, S.; Walsh, T. R. Facet selectivity in gold binding peptides: exploiting interfacial water structure. Chem. Sci. 2015, 6, 5204−5214. (38) Penna, M. J.; Mijajlovic, M.; Tamerler, C.; Biggs, M. J. Molecular-level understanding of the adsorption mechanism of a graphite-binding peptide at the water/graphite interface. Soft Matter 2015, 11, 5192−5203. (39) Sarupria, S.; Garde, S. Quantifying water density fluctuations and compressibility of hydration shells of hydrophobic solutes and proteins. Phys. Rev. Lett. 2009, 103, 037803. (40) Godawat, R.; Jamadagni, S. N.; Garde, S. Characterizing hydrophobicity of interfaces by using cavity formation, solute binding, and water correlations. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 15119− 15124. (41) Granick, S.; Bae, S. C. CHEMISTRY: A Curious Antipathy for Water. Science 2008, 322, 1477−1478. (42) Zhang, X.; Zhu, Y. X.; Granick, S. Hydrophobicity at a Janus interface. Science 2002, 295, 663−666. (43) Patra, N.; Wang, B.; Král, P. Nanodroplet activated and guided folding of graphene nanostructures. Nano Lett. 2009, 9, 3766−3771. (44) Materials Studio 6.1; Accelrys Software Inc: San Diego, CA, USA, 2012. (45) Hu, R.; Fan, Z. Q.; Fu, C. H.; Nie, L. Y.; Huang, W. R.; Zhang, Z. H. Structural stability, magneto-electronics and spin transport properties of triangular graphene nanoflake chains with edge oxidation. Carbon 2018, 126, 93−104. (46) Yuan, P. F.; Zhang, Z. H.; Fan, Z. Q.; Qiu, M. Electronic structure and magnetic properties of penta-graphene nanoribbons. Phys. Chem. Chem. Phys. 2017, 19, 9528−9536. (47) Sun, H. COMPASS: An ab Initio Force-Field Optimized for Condensed-Phase Applications Overview with Details on Alkane and Benzene Compounds. J. Phys. Chem. B 1998, 102, 7338−7364. (48) Pei, S.; Cheng, H.-M. The reduction of graphene oxide. Carbon 2012, 50, 3210−3228. (49) Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.; Kleinhammes, A.; Jia, Y.; Wu, Y.; Nguyen, S. T.; Ruoff, R. S. Synthesis
of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide. Carbon 2007, 45, 1558−1565. (50) Vanommeslaeghe, K.; MacKerell, A. D., Jr. Automation of the CHARMM General Force Field (CGenFF) I: bond perception and atom typing. J. Chem. Inf. Model. 2012, 52, 3144−3154. (51) Vanommeslaeghe, K.; Raman, E. P.; MacKerell, A. D., Jr. Automation of the CHARMM General Force Field (CGenFF) II: assignment of bonded parameters and partial atomic charges. J. Chem. Inf. Model. 2012, 52, 3155−3168. (52) Fonseca, A. F.; Liang, T.; Zhang, D.; Choudhary, K.; Sinnott, S. B. Probing the accuracy of reactive and non-reactive force fields to describe physical and chemical properties of graphene-oxide. Comput. Mater. Sci. 2016, 114, 236−243. (53) Walsh, T. R.; Knecht, M. R. Biointerface Structural Effects on the Properties and Applications of Bioinspired Peptide-Based Nanomaterials. Chem. Rev. 2017, 117, 12641−12704. (54) Walsh, T. R. Pathways to Structure-Property Relationships of Peptide-Materials Interfaces: Challenges in Predicting Molecular Structures. Acc. Chem. Res. 2017, 50, 1617−1624. (55) Heinz, H.; Ramezani-Dakhel, H. Simulations of inorganicbioorganic interfaces to discover new materials: insights, comparisons to experiment, challenges, and opportunities. Chem. Soc. Rev. 2016, 45, 412−448. (56) Ozboyaci, M.; Kokh, D. B.; Corni, S.; Wade, R. C. Modeling and simulation of protein-surface interactions: achievements and challenges. Q. Rev. Biophys. 2016, 49, No. e4. (57) Peng, E.; Todorova, N.; Yarovsky, I. Effects of forcefield and sampling method in all-atom simulations of inherently disordered proteins: Application to conformational preferences of human amylin. PLoS One 2017, 12, No. e0186219. (58) Rappe, A. K.; Goddard, W. A. Charge equilibration for molecular dynamics simulations. J. Phys. Chem. 1991, 95, 3358−3363. (59) Konkena, B.; Vasudevan, S. Understanding Aqueous Dispersibility of Graphene Oxide and Reduced Graphene Oxide through pKa Measurements. J. Phys. Chem. Lett. 2012, 3, 867−872. (60) Berendsen, H. J. C.; van der Spoel, D.; van Drunen, R. GROMACS: A message-passing parallel molecular dynamics implementation. Comput. Phys. Commun. 1995, 91, 43−56. (61) 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. (62) Lundborg, M.; Lindahl, E. Automatic GROMACS topology generation and comparisons of force fields for solvation free energy calculations. J. Phys. Chem. B 2015, 119, 810−823. (63) Darden, T.; York, D.; Pedersen, L. Particle mesh Ewald: An N· log(N) method for Ewald sums in large systems. J. Chem. Phys. 1993, 98, 10089−10092. (64) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 1983, 79, 926−935. (65) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R. Molecular dynamics with coupling to an external bath. J. Chem. Phys. 1984, 81, 3684−3690. (66) Bussi, G.; Donadio, D.; Parrinello, M. Canonical sampling through velocity rescaling. J. Chem. Phys. 2007, 126, 014101. (67) Crawford, R. J.; Webb, H. K.; Truong, V. K.; Hasan, J.; Ivanova, E. P. Surface topographical factors influencing bacterial attachment. Adv. Colloid Interface Sci. 2012, 179−182, 142−149. (68) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 1996, 14, 33−38.
11503
DOI: 10.1021/acsomega.8b00866 ACS Omega 2018, 3, 11497−11503