- Email: [email protected]
Solvation of alanine and histidine functionalized carbon nanotubes in aqueous media: A Monte Carlo simulation study
Journal of Molecular Liquids 208 (2015) 191–195
Contents lists available at ScienceDirect
Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq
Solvation of alanine and histidine functionalized carbon nanotubes in aqueous media: A Monte Carlo simulation study Leila Rahmani a, Sepideh Ketabi b,⁎ a b
Department of Biochemistry, Falavarjan Branch, Islamic Azad University, Falavarjan, Iran Department of Chemistry, East Tehran Branch, Islamic Azad University, Tehran, Iran
a r t i c l e
i n f o
Available online xxxx Keywords: CNT Alanine Histidine DFT MC simulation Solvation free energy
a b s t r a c t In this research, the effect of adsorption of two amino acids on the solvation of carbon nanotubes is investigated. The complexes of β-Alanine and Histidine with armchair single wall carbon nanotube are studied by density functional theory. Then, the outcomes of Monte Carlo simulations of these nanotubes immersed in water to look at the effects of nanotube type on the solvation in water are reported. It is found that the binding energy of the interaction of β-Alanine with nanotube is larger than Histidine. The results of computer simulation in aqueous solution indicate that amino acid functionalization increases the intermolecular interactions of carbon nanotube and water. Computed solvation free energies are in this order: CNTAlanine N CNTHistidine N pure CNT. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Carbon nanotubes (CNTs), due to their one dimensional structure and remarkable mechanical, thermal, optical, and electronic properties, may find many applications in materials and life sciences [1]. The welldefined shape and size of carbon nanotubes (CNTs) make them attractive candidates for theoretical and experimental studies of various nanoscopic phenomena such as protection and confinement of molecular species as well as transport of molecules through their interior pores. Nanotubes are believed to open the road toward different modern fields, either technological or biological. CNTs have recently become promising materials in various biological applications such as drug delivery [2], tumor therapy [3], biosensors [4], and templates for biomolecule assembly [5]. However, the applications of nanotubes have been badly impeded for the poor solubility in water which is especially essential for studies in the presence of living cells. Therefore, water soluble samples are in demand candidates for theoretical studies of various nanoscopic phenomena such as protection and confinement of molecular species as well as transport of molecules. However, there are still challenges facing the carbon nanotube industry, such as how to effectively disperse CNTs in solution, and how to assemble CNTs and other molecules into useful nanostructures [6–8]. An important technique to increase the solubility and reactivity of CNT is through functionalization, which increases the electrical dipole moments [9]. A theoretical study of the functionalization of SWCNTs with some organic molecules has shown the changes in the NT properties due to the functionalization [10]. Among numerous functional ⁎ Corresponding author. E-mail address: [email protected] (S. Ketabi).
http://dx.doi.org/10.1016/j.molliq.2015.04.021 0167-7322/© 2015 Elsevier B.V. All rights reserved.
species, functionalization of NTs with the assistance of biological molecules (such as nucleic acids and proteins) improves the solubility in aqueous, thus facilitating the application of NTs in biotechnology, biomedicine, and bioengineering. Theoretical studies on DNA bases [11] and functionalized CNTs with DNA bases [12] indicate the enhancement of solubility of NTs in water through favorable changes in the solvation energy. Some specific proteins/peptides have been proven to bind to CNTs in experiments, thus can be used to disperse CNTs through their interior pores [13–15], At the same time the wide biomedical applications raise the biosafety concerns of CNTs due to their unintended interactions with proteins and other biological molecules [16–18]. Despite their technological and biological interest there is not enough detailed theoretical analysis of the interactions of CNTs and biological molecules such as amino acids. As a starting point in understanding interactions with much more complex biological systems, we carried out the interaction of CNTs and two amino acids β-Alanine and Histidine (two amino acids in the structure of carnosine dipeptide). Carnosine has protective functions additional to anti-oxidant and free radical scavenging roles. It extends cultured human life span, kills transformed cells, protects cells against aldehydes and an amyloid peptide fragment and inhibits in vitro, protein glycation and DNA/protein cross linking. Carnosine is an aldehyde scavenger, a likely lipofuscin (age pigment) precursor and possible modulator of diabetic complications, atherosclerosis and Alzheimer's disease [19–25]. It has been shown that CNTs can be used as liquid filled nanoparticles for drug delivery tool improves the bioavailability of carnosine. This is an example of the potential applications of carbon nanotubes in biomedicine as drug or vaccine carriers and biomolecular recognition [26].
192
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195
In this study we investigate the interaction of armchair (5, 5) carbon nanotube and the amino acids β-Alanine and Histidine in gas phase and aqueous solution within density functional theory calculations and Monte Carlo simulation. Since carnosine dipeptide structure consists of β-Alanine and L-Histidine, we are interested in these amino acids. Details on the model as well as on the computational methods employed are explained more thoroughly in the proceeding section.
2. Computational method This study comprised two sections: quantum mechanics (QM) and Monte Carlo (MC) simulation. In the quantum mechanical part, isolated molecules (CNT, β-Alanine and L-Histidine functionalized CNTs and isolated β-Alanine and L-Histidine) were optimized. Then it has been applied Monte Carlo simulation for dilute solutions of nanotubes in water. Therefore the study was divided into four parts: (i) Geometry optimization of the amino acids, and (5, 5) CNT; (ii) interactions of amino acids with substrate (CNT); (iii) stability analysis structure of the CNTAlanine (CNTAla) and CNTHistidine (CNTHis) complexes in gas phase: and (iv) stability analysis structure of the CNTAla and CNTHis complexes in water medium. In the first step of our investigation it was important to find the most stable isomers of β-Alanine and L-Histidine. As mentioned before, we were interested to study the interaction of alanine and histidine with CNT, because these two amino acids participate in carnosine dipeptide structure. Therefore, the sites of amino acids that take part in peptide bonding were blocked to prevent interacting with CNT. The sections that would form part of the peptide bonding have all been replaced by terminating hydrogen. This approach has been utilized in other researches before [27]. Hence, among the possible isomers, the protonated form of these two amino acids (Fig. 1) has been used. It should be noticed that these protonated forms of amino acids are stable under their isoelectric pH [28]. The protonated forms of histidine and alanine were fully optimized to confirm the stability of related structures. Then separately optimized geometries were used in the combined system. To perform an accurate QM calculation to a nanosized system without ending up in a prohibitively large computation, an approach is the cluster model. Most of the previous attempts to study this phenomenon are using a cylindrical part of the tube while dangling bonds at the ends of the tube were saturated with hydrogen atoms to minimize the edge effect [29]. Similar methodology was used successfully to study the gas adsorption to CNTs [30,31]. The ab initio study of the interaction of glycine amino acid and CNT has used the same approach [32]. A large part of an armchair (4, 4) and a zigzag (8, 0) CNT containing 64 carbon atoms (or 3 hexagon layers) were separated and treated as an individual system. In the study of DNA-base Li doped SiC nanotubes [12], 5 hexagon layers of SiC nanotubes have been used.
In our research a large adequate part of an armchair (5, 5) CNT containing 100 carbon atoms was separated and treated as an individual system. Five layers of carbon ring hexagons along the tube axis were modeled where the two edge dangling bonds are saturated by hydrogen atoms to emulate the bulk properties. Then interactions of amino acids with the central ring were studied. Each species (amino acid, CNT and their complexes) was optimized by the DFT/B3LYP method using the 6–31 G* [33,34] basis set. The length of the tube has been chosen constant (equal to 11 Å). In all calculations, nanotubes were capped with hydrogen atoms. All structural optimization for the nanotube-amino acid systems was performed by using the GAMESS-US quantum chemistry package [35]. Then investigations of the structural properties of water surrounding nanotubes (containing CNT, CNTAla and CNTHis) were studied by performing fully atomistic Monte Carlo simulation in water. The FORTRAN code developed by the corresponding author [36] was used in simulation. In all separate MC simulations performed here, throughout a standard manner the Metropolis sampling technique [37] in canonical (T, V, N) ensemble was used. Each setup includes two sections: a solute fragment and water molecules. All calculations were performed in a cubic box at the experimental density of water, 1 g/cm3. The optimum edges of the box were 50 × 50 × 50 Å, which corresponds to almost 4000 H2O molecules of pure solvent. The interactions between the solute and the water molecules were defined by a site–site interaction potential consisting of a LJ potential to represent the short range interactions and a long range Coulomb potential with parameters, εi, σi, and qi for each atoms in nanotube and amino acids: " E
AB
¼ 4εi j
σij ri j
!12
σij − ri j
!6 # þ
qi q j e2 ri j
ð1Þ
EAB is the interaction energy between two molecules, A and B, which are expressed by the pair wise sum of their interaction contributions. Appropriate Lennard–Jones parameters for atoms in CNT [38] and amino acids [39,40] are given in Table 1. The Transferable intermolecular potential function (TIP3) [41] was applied for modeling water molecules. Periodic boundary conditions were employed in computation. The system was thoroughly equilibrated using 107 configurations. To calculate the solvation free energies of the nanotubes, the thermodynamic perturbation method was applied in these computations. The free energy difference between two states A and B of a system may be derived from classical statistical mechanics [42] allowing this difference to be expressed as Eq. (2) as the free energy perturbation (FEP) master equation. ΔG ¼ GB −GA ¼ −RT ln h exp−ðEB −EA Þ=RT i
ð2Þ
(EB − EA) is the potential energy difference (ΔE) between states A and B of the system. R is the molar gas constant, T is the absolute temperature,
Fig. 1. Amino acids (a) β-Alanine and (b) Histidine.
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195 Table 1 Lennard–Jones parameters for the atoms in CNT and amino acids. Site
ε, kcal/mol
σ, Å
C O N C in C_O Other C H on N H on C
0.08 0.21 0.17 0.105 0.08 0 0.05
3.5 2.96 3.25 3.75 3.5 0 2.5
and the symbol 〈〉 indicates an ensemble average taken using the potential for state A. In practice, one introduces a coupling parameter (λ) into the potential function which allows us to calculate the free energy difference over a series of more closely related states. Using this parameter, the free energy difference between states A and B is as following: D h i E N N ΔG ¼ ∑i¼1 ΔGi ¼ ∑i¼1 −RT ln exp− Eλðiþ1Þ −EλðiÞ =RT
λi
ð3Þ
The solvation free energy of the species A (CNT, CNTAla and CNTHis) can be written in terms of perturbations where the species disappear to nothing in the gas phase and in solution: ΔGsol ðAÞ ¼ ΔGgas ðA→0Þ−ΔGsol ðA→0Þ:
ð4Þ
3. Results and discussion To prove potential application of CNT as a drug delivery, we investigated the possibility of β-Alanine and Histidine adsorption onto CNT in water. In current research, we studied the interaction of the possible sites of β-Alanine and Histidine with (5, 5) armchair CNT. While the carboxyl group of β-Alanine and amine group of histidine take part in the peptide bonding, the other desirable sites for interacting with CNT are the amine group of alanine and imidazole ring of histidine (see Fig. 1). The binding of imidazole ring of histidine to CNT can occur via π–π stacking and NH–π interaction. The peptide bond that chemically links two amino acid residues along a poly peptide chain is a C–N bond of an amide linkage and shows partial double-bond character. Therefore, the molecule cannot freely rotate around the peptide bond [43]. So the interaction of carboxyl group of histidine with CNT was not considered. Orientations and positions have been specifically
193
chosen so that amine group of β-Alanine (Fig. 2a) and imidazole ring of Histidine (Fig. 2b & c) approach directly on top of CNT atoms. The optimized structures are shown in Fig. 2. The binding energies of an amino acid, Eb, adsorbed on the exterior surface of a CNT were calculated according to the expression Eb ¼ EðCNT−amino acidÞ−½EðCNTÞ þ Eðamino acidÞ
ð5Þ
where E(CNT) denotes the total energy of the optimized pure CNT, E(amino acid) is the energy of amino acid, and E(CNT-amino acid) is the energy of the CNT with the adsorbed amino acid. Eb b 0 corresponds to a stable optimized configuration and indicates bonding. The binding energies calculated by Eq. (5) are only a rough estimation because the basis set superposition error (BSSE) should be considered. The calculations were performed with and without the counterpoise corrections for BSSE [44] for all complexes. Of course it has been shown that DFT is much less affected by BSSE than other methods [45] and it appears as a practical and useful method, with a small BSSE [46]. The binding energy with BSSE correction can be defined as: Eb ðBSSE correctedÞ ¼ Eb þ BSSE:
ð6Þ
BSSE which is always positive in this formulation is added to the binding energy of the complex yielding the corrected Eb. Binding energy values were changed after considering the BSSE correction. However, the sign of the binding energies are not changed upon the BSSE correction. The results are given in the Table 2. It demonstrates that π–π interaction via aromatic ring of Histidine with CNT has the positive binding energy. Therefore, between two interactions which are considered with CNT, Histidine makes NH–π interaction with the binding energy of − 0.9872 eV. As can be seen, the binding energy of the interaction of β-Alanine and CNT is − 2.1365 eV. So the interaction of β-Alanine with CNT is stronger than Histidine. In the Monte Carlo simulations, a very dilute solution of CNT and its complexes with amino acids were used. So one molecule of solute has merged in water and the average energies were then calculated from Monte Carlo simulations. The resultant configuration of the Monte Carlo simulation of CNT-Ala and CNT-His in water is shown in Fig. 3. This gives a qualitative idea of the formation of the solvation shell around the solute. The total energy of the compounds in water was calculated. Etot can be explained as the sum of the energy contributions from solute-solvent
Fig. 2. Interaction of (a) β-Alanine and CNT, (b) L-Histidine and CNT via NH–π, (c) L-Histidine and CNT via π–π.
194
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195
Table 2 Binding energy (Eb) and dipole moment of CNT, amino acids and their complexes.
Stoichiometry Dipole moment (Debye) Energy (Hartree) Energy (eV) Binding energy (eV) Binding energy (eV) (BSSE corrected)
CNT
β-Alanine
Histidine
CNTAla
CNTHis (NH–π)
CNTHis (π–π)
C100H20 0.0449 −3801.7088 −103,448.2976 – –
C3H8O2N 1.9299 −322.4477 −8774.1239 – –
C6H10O2N3 4.0926 −546.2478 −14,863.9478 – –
C103H26O2N3 18.355 −4124.2381 −112,224.6439 −2.2224 −2.1365
C106H30O2N3 14.6193 −4347.9979 −118,313.3705 −1.1251 −0.9872
C106H30O2N3 28.1245 −4347.5302 −118,302.3834 9.8620 10.1442
(Esoln), solvent–solvent (Esolv), and intramolecular (Eint) interactions. The average energy (bEtotN) calculated from Monte Carlo simulations is given in Table 3. This table also includes the number of solvent molecules (NH2O) in the cubic box, local density of water inside and outside of the nanotube, Electrostatic and van der Waals contribution in solute– solvent interaction energy. The results indicated that the absolute total energies of these compounds in water appear in the following order: Etotal ðCNT−AlaÞNEtotal ðCNT−HisÞNEtotal ðCNTÞ: In the polar solvent (water), the electrostatic terms play an important role in the total energy. Point charges on the material's surface can improve total energy because they increase the interaction energy of H2O molecules and the solute. The atomic charges in the solutes affect the electrostatic terms of intermolecular energies directly. Quantum mechanical calculations indicated that the atoms in the functionalized CNTs with amino acids carry partial charges due to the interaction of Alanine and Histidine with CNT. In fact the electrical dipole moments increased with the functionalization, and this augment the intermolecular interaction of nanotube with water. Therefore, the absolute total energy of CNT-amino acid in water is larger than that of pure CNT. Quantum mechanical calculations indicated that the dipole moment of CNTAla is larger than that of CNTHis (See Table 2). So it is predicted that the interaction of CNTAla and water (polar solvent) is larger than that of CNTHis. Computer simulation results are compatible with these conclusions. As can be seen in Table 3, total energy of CNTAla in water medium is greater than CNTHis. As can be seen in Table 3, in functionalized CNTs electrostatic contribution of the CNT–water interaction energy is larger than the van der Waals contribution due to the partial atomic charges in the surface of functionalized CNTs. Pure CNT has no atomic charges and it is predicted that the total energy of CNT in water is lesser than the functionalized CNTs. therefore electrostatic contribution in CNT–water interaction
energy is negligible. The results in Table 3 are compatible with these consequences. The amount of water molecules in the NTs has represented as local densities. Calculated values for CNT, CNTAla and CNTHis samples are given in Table 2. As can be seen, the local density of water inside CNTamino acid is larger than non-functionalized CNT which corresponds to the larger atomic charges on the tube wall and less hydrophobic effect of CNTAla and CNTHis samples. The outside local densities for all three NTs are equal to 0.032 that corresponds to the limit of bulk water in all samples. Solvation free energy is the change in Gibbs energy when a molecule is transferred from a vacuum (or the gas phase) to a solvent. Solvation free energies of nanotubes are inaccessible experimentally due to problems of low volatility. In this case, the application of theoretical methods may be the best approach to gain physical insights into the effects of solvation. Solvation free energy calculations for each of the solutes (CNT, CNTAla and CNTHis) were carried out. To model the CNTs, a simulation was performed on the system with the solute fully represented and then its electrostatic and van der Waals parameters decreased to zero. The computed solvation free energies, ΔGsol, are also presented in Table 3. It could be seen in Table 3 that CNT had the positive value of the solvation free energy. Whereas CNTAla and CNTHis had the negative value of ΔGsol. Computed solvation free energies indicated that amino acid functionalization increases the solvation of CNT in water. As it is mentioned before, CNT has only van der Waals intermolecular interactions with water molecules (Table 3) and so it has poor solubility in water. The atomic charges in the solutes affect on electrostatic terms of intermolecular energies directly. The electrostatic contribution of intermolecular interaction between solute–solvent in CNTAla and CNTHis solutions is −1.346 and −1.161 respectively (Table 3). Thus as expected, amino acid functionalized CNTs have higher solvation free energies than pure CNT. So it is concluded that amino acid functionalization increases the solubility of CNT in water. These consequences were compatible to those outcomes from other functionalization of CNTs [10,
Fig. 3. Snapshot that issued from the simulation of (a) CNT-Ala and (b) CNT-His in water.
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195 Table 3 MC simulation results. NT
CNT
CNT-Ala
CNT-His
NH2O bEN (kcal mol−1) inside local density Outside local density Electrostatic contribution in Esoln van der Waals contribution in Esoln ΔGsol (kcal mol−1)
4034 −8.456 0.002 0.032 0.000 −0.287 0.2550
4024 −10.042 0.006 0.032 −1.346 −0.178 −12.335
4020 −9.921 0.008 0.032 −1.161 −0.182 −10.953
Local densities are in molecule/Å3.
47]. Solvation free energy of CNTAla (− 12.335) is larger than CNTHis (−10.953). Consequently, the solubility of CNTAla is more than CNTHis. 4. Conclusion We have performed theoretical calculations on the structural properties of CNTs upon adsorption of β-Alanine and Histidine in water. Between these complexes, the β-Alanine has binding energies of − 2.1365 eV that is larger than Histidine with the binding energy of −0.9872 eV. Since Aqueous nanotubes are believed to open the road toward different modern fields, either technologically or biologically, we also studied the role of structural composition on the behavior of NTs in water by computer simulation method. The results indicated that adsorption of β-Alanine and Histidine can increase the total energy of CNT in water. Between two complexes, CNTAla has a larger interaction energy with water. In order to understand the effect of functionalization on the solubility of CNTs, solvation free energies were also computed. Computations indicated that solubility of the species is in the order: CNTAlanine N CNTHistidine N pure CNT. References [1] P. Singh, J. Kumar, F.M. Toma, J. Raya, M. Prato, B. Fabre, S. Verma, A. Bianco, J. Am. Chem. Soc. 131 (2009) 13555. [2] A.A. Bhirde, V. Patel, J. Gavard, G. Zhang, A.A. Sousa, A. Masedunskas, R.D. Leapman, R. Weigert, J.S. Gutkind, J.F. Rusling, ACS Nano 3 (2009) 307. [3] S. Jain, V.S. Thakare, M. Das, A.K. Jain, S. Patil, Nanomedicine 5 (2010) 1277. [4] R.J. Chen, S. Bangsaruntip, K.A. Drouvalakis, N.W.S. Kam, M. Shim, Y.M. Li, W. Kim, P.J. Utz, H. Dai, J. Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 4984. [5] W.F. DeGrado, G. Grigoryan, Y.H. Kim, R. Acharya, K. Axelrod, R.M. Jain, L. Willis, M. Drndic, J.M. Kikkawa, Science 332 (2011) 1071. [6] D. Nepal, K.E. Geckeler, Small 3 (2006) 1259. [7] S.Q. Wang, E.S. Humphreys, S.Y. Chung, D.F. Delduco, S.R. Lustig, H. Wang, K.N. Parker, N.W. Rizzo, S. Subramoney, Y.M. Chiang, A. Jagota, Nat. Mater. 2 (2003) 196. [8] G.R. Dieckmann, A.B. Dalton, P.A. Johnson, J. Razal, J. Chen, G.M. Giordano, E. Munoz, I.H. Musselman, R.H. Baughman, R.K. Draper, J. Am. Chem. Soc. 125 (2003) 1770.
195
[9] M. Mananghaya, E. Rodulfo, G.S. Santos, A. Villagracia, J. Nanotechnol. 1–6 (2012). [10] J. Robles, M.J. López, J. Alonso, Modeling of the functionalization of single-wall carbon nanotubes towards its solubilization in an aqueous medium, Eur. Phys. J. D Atom. Mol. Opt. Phys. 61 (2) (2011) 381–388. [11] M. Monajjemi, S. Ketabi, M. Hashemian Zadeh, A. Amiri, Biochem. Mosc. 71 (1) (2006) S1–S8. [12] S. Ketabi, S.M. Hashemianzadeh, M. MoghimiWaskasi, J. Mol. Model. 19 (2013) 1605–1615. [13] S.S. Karajanagi, A.A. Vertegel, R.S. Kane, J.S. Dordick, Langmuir 20 (2004) 11594. [14] P. Goldberg-Oppenheimer, O. Regev, Small 3 (2007) 1894. [15] J. Zhong, L. Song, J. Meng, B. Gao, W. Chu, H. Xu, Y. Luo, J. Guo, A. Marcelli, S. Xie, Z. Wu, Carbon 47 (2009) 967. [16] Y.L. Zhao, G.M. Xing, Z.F. Chai, Nat. Nanotechnol. 3 (2008) 191. [17] G. Zuo, W. Gu, H. Fang, R. Zhou, J. Phys. Chem. C 115 (2011) 12322. [18] G. Zuo, Q. Huang, G. Wei, R. Zhou, H. Fang, ACS Nano 4 (2010) 7508. [19] A.R. Hipkiss, Int. J. Biochem. Cell Biol. 30 (1998) 836–868. [20] A.R. Hipkiss, C. Brownson, Cell. Mol. Life Sci. 1 (2000) 217–223. [21] A.R. Hipkiss, Mechanisms of aging and development, 126 (2005) 1034–1039. [22] A.R. Hipkiss, J. Alzheimers Dis. 11 (2007) 229–240. [23] A.R. Hipkiss, Food Nutr. Res. 57 (2009) 87–154. [24] A.R. Hipkiss, Mol. Asp. Med. 32 (2011) 267–278. [25] Y. Horii, et al., Neurosci. Lett. 501 (2012) 1–5. [26] M. Panthuis, Chem. Biol. 10 (2003) 897. [27] T. Roman, W.A. Dino, H. Nakanishi, H. Kasai, Eur. Phys. J. D 38 (2006) 117–120. [28] A.L. Lehninger, D.L. Nelson, M.M. Cox, Principles of Biochemistry, 5th edition, 2008. [29] C. Kim, Y.S. Choi, J.T. Park, B. Kim, Y.H. Lee, J. Am. Chem. Soc. 124 (2002) 9906. [30] G.E. Froudakis, Nano Lett. 1 (2001) 179–182. [31] G.E. Froudakis, Nano Lett. 1 (2001) 531–533. [32] A. Mavrandonakis, S.C. Farantos, G.E. Froudakis, J. Phys. Chem. B 110 (2006) 6048–6050. [33] G.A. Petersson, M.A. Al-Laham, J. Chem. Phys. 94 (1991) 6081(1991). [34] G.A. Petersson, A. Bennett, T.G. Tensfeldt, M.A. Al-Laham, W.A. Shirley, J.J. Mantzaris, Chem. Phys. 89 (1988) 2193. [35] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jense, S. Koseki, N. Matsunaga, K.A. Nguyen, S.J. Su, T.L. Windus, M. Dupuis, J.A. Montgomery, J. Comput. Chem. 14 (1993) 1347–1363. [36] H.H. Haeri, S. Ketabi, S.M. Hashemianzadeh, The solvation study of carbon, silicon and their mixed nanotubes in water solution, J. Mol. Model. 18 (2012) 3379–3388. [37] N. Metropolis, A.W. Rosenbulth, M.N. Rosenbulth, A.H. Teller, E. Telle, J. Chem. Phys. 21 (1953) 1087. [38] Y. Jia, M. Wang, L. Wu, Sep. Sci. Technol. 42 (2007) 3681–3695. [39] W.L. Jorgensen, C.J. Swenson, Optimized intermolecular potential functions for amides and peptides. Hydration of amides, J. Am. Chem. Soc. 107 (1985) 1489–1496. [40] R.C. Rizzo, W.L. Jorgensen, OPLS all-atom model for amines: resolution of the amine hydration problem, J. Am. Chem. Soc. 121 (1999) 4827–4836. [41] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Lein, Comparison of simple potential functions for simulating liquid water, J. Chem. Phys. 79 (1983) 926–935. [42] D.L. Beveridge, F.M. Di Capua, Free Energy Via Molecular Simulation: Applications to Chemical and Biomolecular Systems, Annu. Rev. Biophys. Biophys. Chem. 18 (1989) 431–492. [43] K.E. Van Holde, W.C. Johnson, P.S. Ho, Principles of physical Biochemistry, Par 1: Macromolecular structure and dynamics, 2nd edition, 2006. [44] S. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553. [45] G.Th. de Jong, M. Solà, L. Visscher, F.M. Bickelhaupt, J. Chem. Phys. 121 (2004) 9982. [46] K.H. Kim, J.C.H. Kim, Y.K. Han, Bull. Korean Chem. Soc. 30 (4) (2009) 794–796. [47] S. Ketabi, M. Aarabmaghsoudi, Effect of metal doping on the solvation of carbon nanotubes: a Monte Carlo simulation study, Phys. Chem. Liq. 52 (6) (2014) 763–776.
Contents lists available at ScienceDirect
Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq
Solvation of alanine and histidine functionalized carbon nanotubes in aqueous media: A Monte Carlo simulation study Leila Rahmani a, Sepideh Ketabi b,⁎ a b
Department of Biochemistry, Falavarjan Branch, Islamic Azad University, Falavarjan, Iran Department of Chemistry, East Tehran Branch, Islamic Azad University, Tehran, Iran
a r t i c l e
i n f o
Available online xxxx Keywords: CNT Alanine Histidine DFT MC simulation Solvation free energy
a b s t r a c t In this research, the effect of adsorption of two amino acids on the solvation of carbon nanotubes is investigated. The complexes of β-Alanine and Histidine with armchair single wall carbon nanotube are studied by density functional theory. Then, the outcomes of Monte Carlo simulations of these nanotubes immersed in water to look at the effects of nanotube type on the solvation in water are reported. It is found that the binding energy of the interaction of β-Alanine with nanotube is larger than Histidine. The results of computer simulation in aqueous solution indicate that amino acid functionalization increases the intermolecular interactions of carbon nanotube and water. Computed solvation free energies are in this order: CNTAlanine N CNTHistidine N pure CNT. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Carbon nanotubes (CNTs), due to their one dimensional structure and remarkable mechanical, thermal, optical, and electronic properties, may find many applications in materials and life sciences [1]. The welldefined shape and size of carbon nanotubes (CNTs) make them attractive candidates for theoretical and experimental studies of various nanoscopic phenomena such as protection and confinement of molecular species as well as transport of molecules through their interior pores. Nanotubes are believed to open the road toward different modern fields, either technological or biological. CNTs have recently become promising materials in various biological applications such as drug delivery [2], tumor therapy [3], biosensors [4], and templates for biomolecule assembly [5]. However, the applications of nanotubes have been badly impeded for the poor solubility in water which is especially essential for studies in the presence of living cells. Therefore, water soluble samples are in demand candidates for theoretical studies of various nanoscopic phenomena such as protection and confinement of molecular species as well as transport of molecules. However, there are still challenges facing the carbon nanotube industry, such as how to effectively disperse CNTs in solution, and how to assemble CNTs and other molecules into useful nanostructures [6–8]. An important technique to increase the solubility and reactivity of CNT is through functionalization, which increases the electrical dipole moments [9]. A theoretical study of the functionalization of SWCNTs with some organic molecules has shown the changes in the NT properties due to the functionalization [10]. Among numerous functional ⁎ Corresponding author. E-mail address: [email protected] (S. Ketabi).
http://dx.doi.org/10.1016/j.molliq.2015.04.021 0167-7322/© 2015 Elsevier B.V. All rights reserved.
species, functionalization of NTs with the assistance of biological molecules (such as nucleic acids and proteins) improves the solubility in aqueous, thus facilitating the application of NTs in biotechnology, biomedicine, and bioengineering. Theoretical studies on DNA bases [11] and functionalized CNTs with DNA bases [12] indicate the enhancement of solubility of NTs in water through favorable changes in the solvation energy. Some specific proteins/peptides have been proven to bind to CNTs in experiments, thus can be used to disperse CNTs through their interior pores [13–15], At the same time the wide biomedical applications raise the biosafety concerns of CNTs due to their unintended interactions with proteins and other biological molecules [16–18]. Despite their technological and biological interest there is not enough detailed theoretical analysis of the interactions of CNTs and biological molecules such as amino acids. As a starting point in understanding interactions with much more complex biological systems, we carried out the interaction of CNTs and two amino acids β-Alanine and Histidine (two amino acids in the structure of carnosine dipeptide). Carnosine has protective functions additional to anti-oxidant and free radical scavenging roles. It extends cultured human life span, kills transformed cells, protects cells against aldehydes and an amyloid peptide fragment and inhibits in vitro, protein glycation and DNA/protein cross linking. Carnosine is an aldehyde scavenger, a likely lipofuscin (age pigment) precursor and possible modulator of diabetic complications, atherosclerosis and Alzheimer's disease [19–25]. It has been shown that CNTs can be used as liquid filled nanoparticles for drug delivery tool improves the bioavailability of carnosine. This is an example of the potential applications of carbon nanotubes in biomedicine as drug or vaccine carriers and biomolecular recognition [26].
192
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195
In this study we investigate the interaction of armchair (5, 5) carbon nanotube and the amino acids β-Alanine and Histidine in gas phase and aqueous solution within density functional theory calculations and Monte Carlo simulation. Since carnosine dipeptide structure consists of β-Alanine and L-Histidine, we are interested in these amino acids. Details on the model as well as on the computational methods employed are explained more thoroughly in the proceeding section.
2. Computational method This study comprised two sections: quantum mechanics (QM) and Monte Carlo (MC) simulation. In the quantum mechanical part, isolated molecules (CNT, β-Alanine and L-Histidine functionalized CNTs and isolated β-Alanine and L-Histidine) were optimized. Then it has been applied Monte Carlo simulation for dilute solutions of nanotubes in water. Therefore the study was divided into four parts: (i) Geometry optimization of the amino acids, and (5, 5) CNT; (ii) interactions of amino acids with substrate (CNT); (iii) stability analysis structure of the CNTAlanine (CNTAla) and CNTHistidine (CNTHis) complexes in gas phase: and (iv) stability analysis structure of the CNTAla and CNTHis complexes in water medium. In the first step of our investigation it was important to find the most stable isomers of β-Alanine and L-Histidine. As mentioned before, we were interested to study the interaction of alanine and histidine with CNT, because these two amino acids participate in carnosine dipeptide structure. Therefore, the sites of amino acids that take part in peptide bonding were blocked to prevent interacting with CNT. The sections that would form part of the peptide bonding have all been replaced by terminating hydrogen. This approach has been utilized in other researches before [27]. Hence, among the possible isomers, the protonated form of these two amino acids (Fig. 1) has been used. It should be noticed that these protonated forms of amino acids are stable under their isoelectric pH [28]. The protonated forms of histidine and alanine were fully optimized to confirm the stability of related structures. Then separately optimized geometries were used in the combined system. To perform an accurate QM calculation to a nanosized system without ending up in a prohibitively large computation, an approach is the cluster model. Most of the previous attempts to study this phenomenon are using a cylindrical part of the tube while dangling bonds at the ends of the tube were saturated with hydrogen atoms to minimize the edge effect [29]. Similar methodology was used successfully to study the gas adsorption to CNTs [30,31]. The ab initio study of the interaction of glycine amino acid and CNT has used the same approach [32]. A large part of an armchair (4, 4) and a zigzag (8, 0) CNT containing 64 carbon atoms (or 3 hexagon layers) were separated and treated as an individual system. In the study of DNA-base Li doped SiC nanotubes [12], 5 hexagon layers of SiC nanotubes have been used.
In our research a large adequate part of an armchair (5, 5) CNT containing 100 carbon atoms was separated and treated as an individual system. Five layers of carbon ring hexagons along the tube axis were modeled where the two edge dangling bonds are saturated by hydrogen atoms to emulate the bulk properties. Then interactions of amino acids with the central ring were studied. Each species (amino acid, CNT and their complexes) was optimized by the DFT/B3LYP method using the 6–31 G* [33,34] basis set. The length of the tube has been chosen constant (equal to 11 Å). In all calculations, nanotubes were capped with hydrogen atoms. All structural optimization for the nanotube-amino acid systems was performed by using the GAMESS-US quantum chemistry package [35]. Then investigations of the structural properties of water surrounding nanotubes (containing CNT, CNTAla and CNTHis) were studied by performing fully atomistic Monte Carlo simulation in water. The FORTRAN code developed by the corresponding author [36] was used in simulation. In all separate MC simulations performed here, throughout a standard manner the Metropolis sampling technique [37] in canonical (T, V, N) ensemble was used. Each setup includes two sections: a solute fragment and water molecules. All calculations were performed in a cubic box at the experimental density of water, 1 g/cm3. The optimum edges of the box were 50 × 50 × 50 Å, which corresponds to almost 4000 H2O molecules of pure solvent. The interactions between the solute and the water molecules were defined by a site–site interaction potential consisting of a LJ potential to represent the short range interactions and a long range Coulomb potential with parameters, εi, σi, and qi for each atoms in nanotube and amino acids: " E
AB
¼ 4εi j
σij ri j
!12
σij − ri j
!6 # þ
qi q j e2 ri j
ð1Þ
EAB is the interaction energy between two molecules, A and B, which are expressed by the pair wise sum of their interaction contributions. Appropriate Lennard–Jones parameters for atoms in CNT [38] and amino acids [39,40] are given in Table 1. The Transferable intermolecular potential function (TIP3) [41] was applied for modeling water molecules. Periodic boundary conditions were employed in computation. The system was thoroughly equilibrated using 107 configurations. To calculate the solvation free energies of the nanotubes, the thermodynamic perturbation method was applied in these computations. The free energy difference between two states A and B of a system may be derived from classical statistical mechanics [42] allowing this difference to be expressed as Eq. (2) as the free energy perturbation (FEP) master equation. ΔG ¼ GB −GA ¼ −RT ln h exp−ðEB −EA Þ=RT i
ð2Þ
(EB − EA) is the potential energy difference (ΔE) between states A and B of the system. R is the molar gas constant, T is the absolute temperature,
Fig. 1. Amino acids (a) β-Alanine and (b) Histidine.
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195 Table 1 Lennard–Jones parameters for the atoms in CNT and amino acids. Site
ε, kcal/mol
σ, Å
C O N C in C_O Other C H on N H on C
0.08 0.21 0.17 0.105 0.08 0 0.05
3.5 2.96 3.25 3.75 3.5 0 2.5
and the symbol 〈〉 indicates an ensemble average taken using the potential for state A. In practice, one introduces a coupling parameter (λ) into the potential function which allows us to calculate the free energy difference over a series of more closely related states. Using this parameter, the free energy difference between states A and B is as following: D h i E N N ΔG ¼ ∑i¼1 ΔGi ¼ ∑i¼1 −RT ln exp− Eλðiþ1Þ −EλðiÞ =RT
λi
ð3Þ
The solvation free energy of the species A (CNT, CNTAla and CNTHis) can be written in terms of perturbations where the species disappear to nothing in the gas phase and in solution: ΔGsol ðAÞ ¼ ΔGgas ðA→0Þ−ΔGsol ðA→0Þ:
ð4Þ
3. Results and discussion To prove potential application of CNT as a drug delivery, we investigated the possibility of β-Alanine and Histidine adsorption onto CNT in water. In current research, we studied the interaction of the possible sites of β-Alanine and Histidine with (5, 5) armchair CNT. While the carboxyl group of β-Alanine and amine group of histidine take part in the peptide bonding, the other desirable sites for interacting with CNT are the amine group of alanine and imidazole ring of histidine (see Fig. 1). The binding of imidazole ring of histidine to CNT can occur via π–π stacking and NH–π interaction. The peptide bond that chemically links two amino acid residues along a poly peptide chain is a C–N bond of an amide linkage and shows partial double-bond character. Therefore, the molecule cannot freely rotate around the peptide bond [43]. So the interaction of carboxyl group of histidine with CNT was not considered. Orientations and positions have been specifically
193
chosen so that amine group of β-Alanine (Fig. 2a) and imidazole ring of Histidine (Fig. 2b & c) approach directly on top of CNT atoms. The optimized structures are shown in Fig. 2. The binding energies of an amino acid, Eb, adsorbed on the exterior surface of a CNT were calculated according to the expression Eb ¼ EðCNT−amino acidÞ−½EðCNTÞ þ Eðamino acidÞ
ð5Þ
where E(CNT) denotes the total energy of the optimized pure CNT, E(amino acid) is the energy of amino acid, and E(CNT-amino acid) is the energy of the CNT with the adsorbed amino acid. Eb b 0 corresponds to a stable optimized configuration and indicates bonding. The binding energies calculated by Eq. (5) are only a rough estimation because the basis set superposition error (BSSE) should be considered. The calculations were performed with and without the counterpoise corrections for BSSE [44] for all complexes. Of course it has been shown that DFT is much less affected by BSSE than other methods [45] and it appears as a practical and useful method, with a small BSSE [46]. The binding energy with BSSE correction can be defined as: Eb ðBSSE correctedÞ ¼ Eb þ BSSE:
ð6Þ
BSSE which is always positive in this formulation is added to the binding energy of the complex yielding the corrected Eb. Binding energy values were changed after considering the BSSE correction. However, the sign of the binding energies are not changed upon the BSSE correction. The results are given in the Table 2. It demonstrates that π–π interaction via aromatic ring of Histidine with CNT has the positive binding energy. Therefore, between two interactions which are considered with CNT, Histidine makes NH–π interaction with the binding energy of − 0.9872 eV. As can be seen, the binding energy of the interaction of β-Alanine and CNT is − 2.1365 eV. So the interaction of β-Alanine with CNT is stronger than Histidine. In the Monte Carlo simulations, a very dilute solution of CNT and its complexes with amino acids were used. So one molecule of solute has merged in water and the average energies were then calculated from Monte Carlo simulations. The resultant configuration of the Monte Carlo simulation of CNT-Ala and CNT-His in water is shown in Fig. 3. This gives a qualitative idea of the formation of the solvation shell around the solute. The total energy of the compounds in water was calculated. Etot can be explained as the sum of the energy contributions from solute-solvent
Fig. 2. Interaction of (a) β-Alanine and CNT, (b) L-Histidine and CNT via NH–π, (c) L-Histidine and CNT via π–π.
194
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195
Table 2 Binding energy (Eb) and dipole moment of CNT, amino acids and their complexes.
Stoichiometry Dipole moment (Debye) Energy (Hartree) Energy (eV) Binding energy (eV) Binding energy (eV) (BSSE corrected)
CNT
β-Alanine
Histidine
CNTAla
CNTHis (NH–π)
CNTHis (π–π)
C100H20 0.0449 −3801.7088 −103,448.2976 – –
C3H8O2N 1.9299 −322.4477 −8774.1239 – –
C6H10O2N3 4.0926 −546.2478 −14,863.9478 – –
C103H26O2N3 18.355 −4124.2381 −112,224.6439 −2.2224 −2.1365
C106H30O2N3 14.6193 −4347.9979 −118,313.3705 −1.1251 −0.9872
C106H30O2N3 28.1245 −4347.5302 −118,302.3834 9.8620 10.1442
(Esoln), solvent–solvent (Esolv), and intramolecular (Eint) interactions. The average energy (bEtotN) calculated from Monte Carlo simulations is given in Table 3. This table also includes the number of solvent molecules (NH2O) in the cubic box, local density of water inside and outside of the nanotube, Electrostatic and van der Waals contribution in solute– solvent interaction energy. The results indicated that the absolute total energies of these compounds in water appear in the following order: Etotal ðCNT−AlaÞNEtotal ðCNT−HisÞNEtotal ðCNTÞ: In the polar solvent (water), the electrostatic terms play an important role in the total energy. Point charges on the material's surface can improve total energy because they increase the interaction energy of H2O molecules and the solute. The atomic charges in the solutes affect the electrostatic terms of intermolecular energies directly. Quantum mechanical calculations indicated that the atoms in the functionalized CNTs with amino acids carry partial charges due to the interaction of Alanine and Histidine with CNT. In fact the electrical dipole moments increased with the functionalization, and this augment the intermolecular interaction of nanotube with water. Therefore, the absolute total energy of CNT-amino acid in water is larger than that of pure CNT. Quantum mechanical calculations indicated that the dipole moment of CNTAla is larger than that of CNTHis (See Table 2). So it is predicted that the interaction of CNTAla and water (polar solvent) is larger than that of CNTHis. Computer simulation results are compatible with these conclusions. As can be seen in Table 3, total energy of CNTAla in water medium is greater than CNTHis. As can be seen in Table 3, in functionalized CNTs electrostatic contribution of the CNT–water interaction energy is larger than the van der Waals contribution due to the partial atomic charges in the surface of functionalized CNTs. Pure CNT has no atomic charges and it is predicted that the total energy of CNT in water is lesser than the functionalized CNTs. therefore electrostatic contribution in CNT–water interaction
energy is negligible. The results in Table 3 are compatible with these consequences. The amount of water molecules in the NTs has represented as local densities. Calculated values for CNT, CNTAla and CNTHis samples are given in Table 2. As can be seen, the local density of water inside CNTamino acid is larger than non-functionalized CNT which corresponds to the larger atomic charges on the tube wall and less hydrophobic effect of CNTAla and CNTHis samples. The outside local densities for all three NTs are equal to 0.032 that corresponds to the limit of bulk water in all samples. Solvation free energy is the change in Gibbs energy when a molecule is transferred from a vacuum (or the gas phase) to a solvent. Solvation free energies of nanotubes are inaccessible experimentally due to problems of low volatility. In this case, the application of theoretical methods may be the best approach to gain physical insights into the effects of solvation. Solvation free energy calculations for each of the solutes (CNT, CNTAla and CNTHis) were carried out. To model the CNTs, a simulation was performed on the system with the solute fully represented and then its electrostatic and van der Waals parameters decreased to zero. The computed solvation free energies, ΔGsol, are also presented in Table 3. It could be seen in Table 3 that CNT had the positive value of the solvation free energy. Whereas CNTAla and CNTHis had the negative value of ΔGsol. Computed solvation free energies indicated that amino acid functionalization increases the solvation of CNT in water. As it is mentioned before, CNT has only van der Waals intermolecular interactions with water molecules (Table 3) and so it has poor solubility in water. The atomic charges in the solutes affect on electrostatic terms of intermolecular energies directly. The electrostatic contribution of intermolecular interaction between solute–solvent in CNTAla and CNTHis solutions is −1.346 and −1.161 respectively (Table 3). Thus as expected, amino acid functionalized CNTs have higher solvation free energies than pure CNT. So it is concluded that amino acid functionalization increases the solubility of CNT in water. These consequences were compatible to those outcomes from other functionalization of CNTs [10,
Fig. 3. Snapshot that issued from the simulation of (a) CNT-Ala and (b) CNT-His in water.
L. Rahmani, S. Ketabi / Journal of Molecular Liquids 208 (2015) 191–195 Table 3 MC simulation results. NT
CNT
CNT-Ala
CNT-His
NH2O bEN (kcal mol−1) inside local density Outside local density Electrostatic contribution in Esoln van der Waals contribution in Esoln ΔGsol (kcal mol−1)
4034 −8.456 0.002 0.032 0.000 −0.287 0.2550
4024 −10.042 0.006 0.032 −1.346 −0.178 −12.335
4020 −9.921 0.008 0.032 −1.161 −0.182 −10.953
Local densities are in molecule/Å3.
47]. Solvation free energy of CNTAla (− 12.335) is larger than CNTHis (−10.953). Consequently, the solubility of CNTAla is more than CNTHis. 4. Conclusion We have performed theoretical calculations on the structural properties of CNTs upon adsorption of β-Alanine and Histidine in water. Between these complexes, the β-Alanine has binding energies of − 2.1365 eV that is larger than Histidine with the binding energy of −0.9872 eV. Since Aqueous nanotubes are believed to open the road toward different modern fields, either technologically or biologically, we also studied the role of structural composition on the behavior of NTs in water by computer simulation method. The results indicated that adsorption of β-Alanine and Histidine can increase the total energy of CNT in water. Between two complexes, CNTAla has a larger interaction energy with water. In order to understand the effect of functionalization on the solubility of CNTs, solvation free energies were also computed. Computations indicated that solubility of the species is in the order: CNTAlanine N CNTHistidine N pure CNT. References [1] P. Singh, J. Kumar, F.M. Toma, J. Raya, M. Prato, B. Fabre, S. Verma, A. Bianco, J. Am. Chem. Soc. 131 (2009) 13555. [2] A.A. Bhirde, V. Patel, J. Gavard, G. Zhang, A.A. Sousa, A. Masedunskas, R.D. Leapman, R. Weigert, J.S. Gutkind, J.F. Rusling, ACS Nano 3 (2009) 307. [3] S. Jain, V.S. Thakare, M. Das, A.K. Jain, S. Patil, Nanomedicine 5 (2010) 1277. [4] R.J. Chen, S. Bangsaruntip, K.A. Drouvalakis, N.W.S. Kam, M. Shim, Y.M. Li, W. Kim, P.J. Utz, H. Dai, J. Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 4984. [5] W.F. DeGrado, G. Grigoryan, Y.H. Kim, R. Acharya, K. Axelrod, R.M. Jain, L. Willis, M. Drndic, J.M. Kikkawa, Science 332 (2011) 1071. [6] D. Nepal, K.E. Geckeler, Small 3 (2006) 1259. [7] S.Q. Wang, E.S. Humphreys, S.Y. Chung, D.F. Delduco, S.R. Lustig, H. Wang, K.N. Parker, N.W. Rizzo, S. Subramoney, Y.M. Chiang, A. Jagota, Nat. Mater. 2 (2003) 196. [8] G.R. Dieckmann, A.B. Dalton, P.A. Johnson, J. Razal, J. Chen, G.M. Giordano, E. Munoz, I.H. Musselman, R.H. Baughman, R.K. Draper, J. Am. Chem. Soc. 125 (2003) 1770.
195
[9] M. Mananghaya, E. Rodulfo, G.S. Santos, A. Villagracia, J. Nanotechnol. 1–6 (2012). [10] J. Robles, M.J. López, J. Alonso, Modeling of the functionalization of single-wall carbon nanotubes towards its solubilization in an aqueous medium, Eur. Phys. J. D Atom. Mol. Opt. Phys. 61 (2) (2011) 381–388. [11] M. Monajjemi, S. Ketabi, M. Hashemian Zadeh, A. Amiri, Biochem. Mosc. 71 (1) (2006) S1–S8. [12] S. Ketabi, S.M. Hashemianzadeh, M. MoghimiWaskasi, J. Mol. Model. 19 (2013) 1605–1615. [13] S.S. Karajanagi, A.A. Vertegel, R.S. Kane, J.S. Dordick, Langmuir 20 (2004) 11594. [14] P. Goldberg-Oppenheimer, O. Regev, Small 3 (2007) 1894. [15] J. Zhong, L. Song, J. Meng, B. Gao, W. Chu, H. Xu, Y. Luo, J. Guo, A. Marcelli, S. Xie, Z. Wu, Carbon 47 (2009) 967. [16] Y.L. Zhao, G.M. Xing, Z.F. Chai, Nat. Nanotechnol. 3 (2008) 191. [17] G. Zuo, W. Gu, H. Fang, R. Zhou, J. Phys. Chem. C 115 (2011) 12322. [18] G. Zuo, Q. Huang, G. Wei, R. Zhou, H. Fang, ACS Nano 4 (2010) 7508. [19] A.R. Hipkiss, Int. J. Biochem. Cell Biol. 30 (1998) 836–868. [20] A.R. Hipkiss, C. Brownson, Cell. Mol. Life Sci. 1 (2000) 217–223. [21] A.R. Hipkiss, Mechanisms of aging and development, 126 (2005) 1034–1039. [22] A.R. Hipkiss, J. Alzheimers Dis. 11 (2007) 229–240. [23] A.R. Hipkiss, Food Nutr. Res. 57 (2009) 87–154. [24] A.R. Hipkiss, Mol. Asp. Med. 32 (2011) 267–278. [25] Y. Horii, et al., Neurosci. Lett. 501 (2012) 1–5. [26] M. Panthuis, Chem. Biol. 10 (2003) 897. [27] T. Roman, W.A. Dino, H. Nakanishi, H. Kasai, Eur. Phys. J. D 38 (2006) 117–120. [28] A.L. Lehninger, D.L. Nelson, M.M. Cox, Principles of Biochemistry, 5th edition, 2008. [29] C. Kim, Y.S. Choi, J.T. Park, B. Kim, Y.H. Lee, J. Am. Chem. Soc. 124 (2002) 9906. [30] G.E. Froudakis, Nano Lett. 1 (2001) 179–182. [31] G.E. Froudakis, Nano Lett. 1 (2001) 531–533. [32] A. Mavrandonakis, S.C. Farantos, G.E. Froudakis, J. Phys. Chem. B 110 (2006) 6048–6050. [33] G.A. Petersson, M.A. Al-Laham, J. Chem. Phys. 94 (1991) 6081(1991). [34] G.A. Petersson, A. Bennett, T.G. Tensfeldt, M.A. Al-Laham, W.A. Shirley, J.J. Mantzaris, Chem. Phys. 89 (1988) 2193. [35] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jense, S. Koseki, N. Matsunaga, K.A. Nguyen, S.J. Su, T.L. Windus, M. Dupuis, J.A. Montgomery, J. Comput. Chem. 14 (1993) 1347–1363. [36] H.H. Haeri, S. Ketabi, S.M. Hashemianzadeh, The solvation study of carbon, silicon and their mixed nanotubes in water solution, J. Mol. Model. 18 (2012) 3379–3388. [37] N. Metropolis, A.W. Rosenbulth, M.N. Rosenbulth, A.H. Teller, E. Telle, J. Chem. Phys. 21 (1953) 1087. [38] Y. Jia, M. Wang, L. Wu, Sep. Sci. Technol. 42 (2007) 3681–3695. [39] W.L. Jorgensen, C.J. Swenson, Optimized intermolecular potential functions for amides and peptides. Hydration of amides, J. Am. Chem. Soc. 107 (1985) 1489–1496. [40] R.C. Rizzo, W.L. Jorgensen, OPLS all-atom model for amines: resolution of the amine hydration problem, J. Am. Chem. Soc. 121 (1999) 4827–4836. [41] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Lein, Comparison of simple potential functions for simulating liquid water, J. Chem. Phys. 79 (1983) 926–935. [42] D.L. Beveridge, F.M. Di Capua, Free Energy Via Molecular Simulation: Applications to Chemical and Biomolecular Systems, Annu. Rev. Biophys. Biophys. Chem. 18 (1989) 431–492. [43] K.E. Van Holde, W.C. Johnson, P.S. Ho, Principles of physical Biochemistry, Par 1: Macromolecular structure and dynamics, 2nd edition, 2006. [44] S. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553. [45] G.Th. de Jong, M. Solà, L. Visscher, F.M. Bickelhaupt, J. Chem. Phys. 121 (2004) 9982. [46] K.H. Kim, J.C.H. Kim, Y.K. Han, Bull. Korean Chem. Soc. 30 (4) (2009) 794–796. [47] S. Ketabi, M. Aarabmaghsoudi, Effect of metal doping on the solvation of carbon nanotubes: a Monte Carlo simulation study, Phys. Chem. Liq. 52 (6) (2014) 763–776.