The impacts of surface polarity on the solubility of nanoparticle

, ,
Jianzhuo Zhu, Xinwen Ou, Jiguo Su , and Jingyuan Li

Citation: J. Chem. Phys. 145, 044504 (2016); doi: 10.1063/1.4959805

View online: http://dx.doi.org/10.1063/1.4959805
View Table of Contents: http://aip.scitation.org/toc/jcp/145/4
Published by the American Institute of Physics
 THE JOURNAL OF CHEMICAL PHYSICS 145, 044504 (2016)

The impacts of surface polarity on the solubility of nanoparticle

Jianzhuo Zhu,1 Xinwen Ou,2,3 Jiguo Su,1,a) and Jingyuan Li2,b)
1
College of Science, Yanshan University, Qinhuangdao 066004, China
2
CAS Key Laboratory for Biomedical Effects of Nanomaterials and Nanosafety,
Institute of High Energy Physics, Chinese Academy of Sciences, Yuquan Road 19B, Beijing 100049, China
3
Key Laboratory of Design and Assembly of Functional Nanostructures, Fujian Institute
of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002, China
(Received 1 April 2016; accepted 12 July 2016; published online 27 July 2016)

In order to study the dependence of water solubility and hydration behavior of nanoparticles on
their surface polarity, we designed polar nanoparticles with varying surface polarity by assigning
atomic partial charge to the surface of C60. The water solubility of the nanoparticle is enhanced by
several orders of magnitude after the introduction of surface polarity. Nevertheless, when the atomic
partial charge grows beyond a certain value (qM), the solubility continuously decreases to the level of
nonpolar nanoparticle. It should be noted that such qM is comparable with atomic partial charge of a
variety of functional groups. The hydration behaviors of nanoparticles were then studied to investigate
the non-monotonic dependence of solubility on the surface polarity. The interaction between the
polar nanoparticle and the hydration water is stronger than the nonpolar counterpart, which should
facilitate the dissolution of the nanoparticles. On the other hand, the surface polarity also reduces the
interaction of hydration water with the other water molecules and enhances the interaction between
the nanoparticles which may hinder their dispersion. Besides, the introduction of surface polarity
disturbs and even rearranges the hydration structure of nonpolar nanoparticle. Interestingly, the polar
nanoparticle with less ordered hydration structure tends to have higher water solubility. Published by
AIP Publishing. [http://dx.doi.org/10.1063/1.4959805]

I. INTRODUCTION
Nanomaterials exhibit great potential for applications
in many fields.1–6 The water solubilities of a variety of
pristine nanoparticles are very low, which often limit their
applications in physiological environment.7–9 There have
been enormous attempts to improve the solubility and
colloidal stability of nanoparticles, many of which are through
surface functionalization.8,9 For example, the hydroxylation of
fullerene gives rise to fullerenol. As the number of hydroxyl
groups grows, the solubility of fullerenol can increase up
to several mM. Enhanced water solubility of fullerenol is
largely attributed to the introduction of surface polarity.9
While the surface functionalization does not solely regulate
the surface polarity, it also changes the size, shape, and
surface topology of the nanoparticle, imposing difficulties
to the direct investigation about the dependence of water
solubility on surface polarity of nanoparticle.
There have been intensive works to study the hydration
of nonpolar (hydrophobic) solute.10–19 As indicated by
Lum-Chandler-Weeks theory, the hydration of hydrophobic
solute is size-dependent.20,21 Small solutes can fit into the
hydrogen-bond network of water molecules, while large
solutes cause reorganization of hydrogen-bond network and
depletion of surrounding waters.14 And the crossover length
scale is around 1 nm.20,22 On the other hand, there are limited
studies about the hydration behavior of polar solute,23,24
a)Electronic mail: jiguosu@ysu.edu.cn
b)Electronic mail: lijingyuan@ihep.ac.cn
especially the ones with the size of ∼1 nm. The impacts of
surface polarity, e.g., the distribution of functional groups, on
the hydration behavior as well as the solubility of nanoparticle
remain largely elusive. On top of this, the size of many soluble
proteins is also around several nanometers. And the protein
surface is mainly composed of polar and charged groups. The
studies about the hydration behavior of polar nanoparticle
and the impacts of surface polarity will also facilitate the
understanding about the solubility and hydration of proteins.
In this work, we constructed the nanoparticles with
varying surface polarity by assigning atomic partial charge
to C60s and studied the impact of surface polarity on their
water solubility. The solubility increases by several orders
of magnitude after the addition of polarity; however, when
the polarity grows beyond a certain level, the solubility
decreases to the level of nonpolar nanoparticle. Interestingly,
the assigned atomic partial charge corresponding to the highest
solubility is comparable to that of a variety of functional
groups. Such a non-monotonic relationship between water
solubility and surface polarity of nanoparticle may be related
to the interplay of different aspects of nanoparticle’s hydration
behavior: while the introduction of surface polarity enhances
the interaction between the nanoparticle and the hydration
water which should facilitate the dispersion of nanoparticles,
polar nanoparticle also reduces the interaction of hydration
waters with the other water molecules and enhances the
interaction between the nanoparticles, which may hinder
their dispersion. Besides, the introduction of surface polarity
disturbs and rearranges the original hydration structure of
nonpolar nanoparticle, such as the radial distribution and

0021-9606/2016/145(4)/044504/6/$30.00 145, 044504-1 Published by AIP Publishing.

044504-2 Zhu et al.
FIG. 1. Polar C60s with alternative (left) and patchy (right) charge distribu-
tion. The green and red sites represent the negatively and positively charged
carbon atoms, respectively.
the orientation of hydration waters. These findings should
be helpful for better understanding of the solubility and
hydration of polar nanomaterials as well as the biological
molecules.
II. METHODOLOGY
Two series of polar C60s with different charge
distributions were studied in this work (Figure 1). In the first
series of polar C60, the positive and negative charge sites are
alternatively distributed (denoted as alternative distribution).
As for the second series, the positively and negatively charged
atoms are distributed into patches composed of 3–5 atoms, i.e.,
patchy distribution. In both cases, the numbers of positive and
negative charge sites are equal, and the nanoparticle remains
charge neutral. Five types of polar C60 with alternative charge
distribution were studied, with the atomic partial charge
q = 0.2, 0.4, 0.6, 0.8, and 1.0 e. As for the second series, three
types of polar C60 with q = 0.2, 0.4, and 0.6 e were studied.
Besides, pristine C60 was studied for comparison. In each
system, there are three C60s solvated in the water box with
the initial size of 10 × 10 × 10 nm3, and the corresponding
concentration is around 5 mM.
All molecular dynamics (MD) simulations were per-
formed with GROMACS program.25 The atom type of
aromatic carbon (CA) of OPLSAA force field was assigned to
the carbon atom, and the TIP3P26 water model was adopted.
Each system was first minimized using the steepest descent
energy minimization method, and a 500 ns NPT simulation
FIG. 2. (a) Solvent accessible surface area (SASA) of nonpolar C60, polar C60 with alternative charge distribution (q = 0.2 , 0.6, 1.0 e) and patchy charge
distribution (q = 0.4, 0.6 e). Smoothed data (red) are also shown for reference. (b) The probabilities of fully dissolved state.
J. Chem. Phys. 145, 044504 (2016)
(P = 1 bar and T = 300 K) was then carried out. The last
450 ns trajectory was used to estimate the solubility of C60.
A time step of 2 fs was used, unless otherwise denoted. The
cutoff distances for both the short-ranged electrostatic and
van der Waals interactions were set to 10 Å. The particle
mesh Ewald (PME) method was adopted for the calculation
of long-range electrostatic interactions.27
III. RESULTS AND DISCUSSION
We first studied the water solubility of polar C60s. The
total solvent accessible surface area (SASA) was used to
characterize the dispersion state of C60s. The SASA of
the three C60s was calculated from the trajectories, and
the results are shown in Figure 2(a). The SASA of ∼12
and ∼13 nm2 corresponds to the C60 trimer in compact
and string arrangement, respectively; the SASA of around
14 nm2 corresponds to the case of a dimer with a separated
C60; and the SASA of ∼15 nm2 represents fully dissolved
state. As shown in Figure 2(a), the designed C60s become
more dispersed after the introduction of polarity, e.g., the
alternative distribution with q = 0.2, 0.6 e, as well as the
patchy distribution with q = 0.4 e. When the polarity further
increases (e.g., q = 1.0 e for the alternative distribution and
q = 0.6 e for the patchy distribution), the solubility decreases.
We then calculated the probabilities of fully dissolved state
from the data of SASA to estimate the dependence of
solubility on surface polarity (Figure 2(b)). Pristine C60s
mainly aggregate as a trimer, and the probability of fully
dissolved state is only 0.6%. (The concentration of C60 in this
system is 5 mM, far beyond the water solubility of C60, i.e.,
1.8 × 10−11 mM.28) The probability of fully dissolved state
for polar C60 (alternative charge distribution) increases to
53% when the atomic partial charge is 0.6 e. In addition,
a separate MD simulation with a time step of 1 fs for
polar C60 of alternative charge distribution with q = 0.6 e
was also conducted. The SASA and the probability of fully
dissolved state are shown in Figure S1 of the supplementary
material.29 The results are similar to that obtained through
the simulation with the time step of 2 fs. As for polar C60
with patchy charge distribution, the probability is 75% when
q = 0.4 e. In other words, the solubility of C60 increases by
several orders of magnitude after the introduction of surface
044504-3 Zhu et al.
FIG. 3. The interaction potential between the designed C60 and the hydration
water. Exponential model fits the data very well.
polarity. It is interesting to note that the atomic partial charges
corresponding to the highest water solubility (i.e., qM = 0.6 e
for the alternative charge distribution and qM = 0.4 e for the
patchy charge distribution) are comparable with atomic partial
charge of a variety of functional groups, e.g., the C atom of
benzoic acid carboxyl group (0.635 e)30 and the N atom of
amino group (−0.4 e).31
The introduction of surface polarity does significantly
increase the solutes’ solubility, while full dispersion cannot
be achieved by any studied polar C60s. This suggests the
solubility of designed polar C60s should be less than 5 mM.
The potential of mean force between two polar C60s of patchy
charge distribution with q = 0.4 e was computed (Figure S2
in the supplementary material29). The minimum (at 0.98 nm)
corresponds to the direct contact of the two C60s,32 which
implies that the state of direct contact is still rather stable. It
should be noted that the solubility of fullerene derivatives can
be up to 11.9 mM (C60(OH)36) and 37.7 mM (C60(OH)40).9
The surface topology of nanoparticles may also affect their
dispersion.
The impact of surface polarity on the hydration behavior
of nanoparticle was then studied. The calculations were
restricted to the fully dissolved state. Additional 22-ns
simulation was performed for each system where three C60s
were restraint to their initial well separated positions. The
interaction potential (the sum of electrostatic and van der
FIG. 4. Average number of hydrogen bonds (a) and the interaction potentials (b) of hydration waters with other water molecules. Exponential model fits the
data very well.
J. Chem. Phys. 145, 044504 (2016)
Waals interaction potential) is usually used to depict the
interactions of solute and solvent.33,34 First, we calculated
the interaction potential between the hydration water and
the solute, denoted as PWS (Fig. 3). The water molecule is
considered as the hydration water if the water oxygen has a
distance of less than 0.8 nm from the center of mass (COM)
of C60 (the radius of C60 is 0.34 nm).35 PWS decreases from
−2.56 kJ/mol (pristine C60) to −14.21 kJ/mol (alternative
charge distribution with q = 1.0 e) and −19.01 kJ/mol (patchy
charge distribution with q = 0.6 e), as the surface polarity
increases. In fact, water molecules as well as other molecules
are likely to be adsorbed on the sites of charged atoms.36,37
Interestingly, for each series of polar C60, PWS can be fit to
an exponential function of q very well.
In addition, we studied the interaction of hydration
water with other water molecules by calculating the number
of hydrogen bonds formed with other water molecules
(Figure 4(a)). A geometric definition of hydrogen bond was
adopted. Namely, two water molecules are considered to form
hydrogen bond if their oxygen–oxygen distance is less than
3.5 Å, and the angle H–O· · · O is smaller than 30◦. In the case
of pristine C60, the average number of hydrogen bond is 3.24,
slightly (2.7%) less than bulk value, which is consistent with
previous work.38 The introduction of polarity considerably
suppresses the formation of hydrogen bond of hydration
waters. In the case of alternative charge distribution, the
hydrogen bond number continuously decreases to 3.18 (0.6 e)
and 3.09 (1.0 e). The disruption of hydrogen bond is more
profound in the case of patchy charge distribution, where the
average number of hydrogen bond decreases to 3.09 (0.4 e)
and further to 2.90 (0.6 e). Similar tendency is also exhibited
in the interaction potential of hydration water with other
water molecules (PWW) (Figure 4(b)): PWW keeps reducing
as the polarity increases, and the reduction is more profound
in the cases of patchy charge distribution. For pristine C60,
PWW (per water molecule) is −76.0 kJ/mol, slightly (2.8%)
weaker than the bulk value. After the introduction of polarity,
PWW continuously reduces to −69.4 kJ/mol (alternative charge
distribution with q = 1.0 e) and −65.4 kJ/mol (patchy charge
distribution with q = 0.6 e), respectively. Interestingly, PWW
can also be fitted to an exponential function of q very
well.
044504-4 Zhu et al. J. Chem. Phys. 145, 044504 (2016)

As shown above, the introduction of polarity considerably

enhances the interaction between the solute and the hydration
water. In addition, the PWS between the solute and the
water molecules with the water oxygen has a distance
between 0.8 nm and 1.1 nm from the COM of the solute
(0.8 < d < 1.1 nm) which was also calculated. The results
are shown in Figure S3 of the supplementary material.29 It
is shown that the change of PWS for water molecules of
0.8 < d < 1.1 nm is negligible as compared to the hydration
water. Token together, the interaction between the solute
and the water molecules is monotonically enhanced as the
atomic partial charge increases. However, it does not result
in continuous increment of their solubility. And the non-
monotonic dependence of the solubility of nanoparticles on
their surface polarity should be related to the fact that the polar
nanoparticles considerably disturb the interaction of hydration
water with other water molecules, which should hinder their
dissolution.
To further study the impacts of polarity on the hydration
structure of designed C60s, the radial distribution functions
(RDFs) gSO(r) of water oxygen atoms around C60 were
calculated (Figure 5). RDF profiles exhibit two distinct peaks
FIG. 6. (a) A schematic diagram about the definition of the vectors µ̂, σ̂, and
located at r = ∼0.65 and ∼0.95 nm (the radius of C60 is
γ̂, as well as the angles θ and ω. ((b)–(d)) Representative configurations of
0.34 nm), respectively. For both series of polar C60, these two hydration water with orientations A, B, and C, respectively.
peaks continuously shift to the left. In the case of alternative
charge distribution, the height of two peaks decreases as
the atomic partial charge increases to 0.8 e, suggesting 0.8 e, patchy charge distribution with q = 0.2, 0.4 e) tend to
that the structure of hydration layer becomes less ordered. have higher water solubility. And the separation between the
Interestingly, when the atomic partial charge further increases two peaks also becomes larger as the atomic partial charge
to 1.0 e, the height of the first peak begins to increase: increases.
the hydration layer becomes ordered again. In addition, the The impact of surface polarity on the orientation of
separation between the two peaks becomes larger as the atomic hydration water was further investigated. The water orientation
partial charge increases. The larger separation may result from is characterized by the combination of angles θ and ω
the weaker interaction between these water molecules. (Figure 6(a)). θ is defined as the angle between the water
Similar phenomena can be observed in the patchy charge dipole ( µ̂) and the vector from C60 COM to water oxygen
distribution. The height of both peaks decreases as the (σ̂), and ω is the angle between the vectors σ̂ and γ̂ (the
atomic partial charge increases to 0.2 e; however, when q normal of H–O–H plane, pointing away from C60).
further increases, the height of first peak begins to increase. In the case of pristine C60, there are two preferred
More strikingly, the polar C60s with less ordered hydration orientations for hydration water: (i) both hydroxyl groups
structure (i.e., alternative charge distribution with q = 0.6, are tangential to the surface of C60 (i.e., θ = 115◦/ω = 25◦,

FIG. 5. Radial distribution functions of water oxygen atoms around designed C60 with alternative (a) and patchy charge distribution (b). Here r is the distance
between water oxygen atom and the COM of C60 (the radius of C60 is 0.34 nm). RDFs of water oxygen atoms around designed C60 were also calculated from
the MD simulations with the time step of 1 fs, the corresponding results are shown in Figure S4 of the supplementary material.29
044504-5 Zhu et al. J. Chem. Phys. 145, 044504 (2016)

FIG. 7. Probability distributions of the angles θ and ω of hydration water.

denoted as orientation A, Figure 6(b)); (ii) one hydroxyl C60s in direct contact, i.e., the COM distance of two C60s
group is tangential to C60 surface and the other points away is less than 1.05 nm (the radius of C60 is about 0.34 nm),
from C60 (i.e., θ = 65◦/ω = 85◦, denoted as orientation B, was calculated, denoted as PSS (Figure 8). The interaction
Figure 6(c)). It should be noted that the tangential arrangement potential between pristine C60s is −37.5 kJ/mol. After
of hydration waters around small hydrophobic solute has been the introduction of polarity, PSS exponentially decreases to
reported in both experimental and simulation works.35,39 The
introduction of polarity considerably disturbs both orientations
of hydration water (Fig. 7). As the polarity further increases,
such as alternative charge distribution with q = 0.6 e and
patchy charge distribution with q = 0.4 e, hydration water
begins to adopt new orientation, i.e., one hydroxyl group
points towards C60, and the other hydroxyl is tangential
to the surface of C60 (i.e., θ = 135◦/ω = 85◦, denoted as
orientation C, Figure 6(d)). Such an orientation is similar
to that of water molecules on some mineral surfaces, e.g.,
Mg(OH)2 (001)40 and α-Al2O3 (0010).41 It is interesting to
note that the emergence of the induced orientation C (i.e.,
alternative charge distribution with q = 0.6 e and patchy
charge distribution with q = 0.4 e) corresponds to the polar
C60 with the highest water solubility.
Besides the impacts on the hydration waters, the
introduction of surface polarity also affects the interaction FIG. 8. The interaction potential between two contacted nanoparticles.
between nanoparticles. The interaction potential between two Exponential model fits the data very well.
044504-6 Zhu et al.
−76.2 kJ/mol (alternative charge distribution with q = 1.0 e)
and −192.5 kJ/mol (patchy charge distribution with q = 0.6 e),
respectively. The enhanced interaction between nanoparticles
may also hinder the dissolution of polar C60. And in some
cases, the strong interaction between nanoparticles may even
be the main origin that hinders the dissolution of nanoparticles.
Taken together, non-monotonic dependence of solubility on
the surface polarity is largely attributed to the interplay
of different aspects of nanoparticle’s hydration behavior,
especially the interplay of PWW, PWS, and PSS.
IV. CONCLUSIONS
With the help of model system, we studied the influence
of surface polarity on the water solubility and the hydration
behavior of nanoparticles. The introduction of surface polarity
enhances the water solubility of nanoparticle by several orders
of magnitude. Nevertheless, the solubility of nanoparticle
considerably decreases when the polarity grows beyond a
certain level. The introduction of surface polarity enhances the
interaction between the nanoparticle and the hydration water
(PWS) which should facilitate the dissolution of nanoparticle.
On the other hand, the surface polarity also reduces the
interaction of hydration water with other water molecule
(PWW) and enhances the interaction between the nanoparticles
(PSS) which may hinder their dispersion. Interestingly, the
interaction potentials between nanoparticle and hydration
water, between hydration water and other water molecules,
and between solutes change exponentially with the assigned
atomic partial charge. Besides, the surface polarity disturbs
and rearranges the hydration structure. Interestingly, the polar
nanoparticles with less ordered hydration structure tend to
have higher water solubility. Taken together, out results
illustrate the non-monotonic dependency of the solubility of
nanoparticles on their surface polarity and the corresponding
hydration behaviors. It should be noted that the assigned
atomic partial charge corresponding to the highest solubility
is comparable to that of a variety of functional groups; further
works are highly demanded to study whether the findings
about the impact of surface polarity can be applied to other
systems, such as functional fullerene and proteins. Moreover,
there are a variety of nanoparticles such as (MgO)n (n ≤ 90)
and (ZnO)n (n ≤ 360)42,43 which have similar surface charge
distribution as our model systems. Our study should be also
helpful for the understanding of the solubility and hydration
of these nanoparticles.
ACKNOWLEDGMENTS
This work is supported in part by the Ministry of Science
and Technology (MOST) 973 Program No. 2013CB933704,
the 100-Talent Project from the Chinese Academy of Science,
and the National Natural Science Foundation of China (NSFC)
Grant Nos. 21273240 and 21473207.
J. Chem. Phys. 145, 044504 (2016)
1E. Mathiowitz, J. S. Jacob, Y. S. Jong, G. P. Carino, D. E. Chickering, P.
Chaturvedi, C. A. Santos, K. Vijayaraghavan, S. Montgomery, M. Bassett,
and C. Morrell, Nature 386, 410 (1997).
2S. M. Marinakos, J. P. Novak, L. C. Brousseau, A. B. House, E. M. Edeki,
J. C. Feldhaus, and D. L. Feldheim, J. Am. Chem. Soc. 121, 8518 (1999).
3K. Lu, L. Lu, and S. Suresh, Science 324, 349 (2009).
4B. Nowack, Science 330, 1054 (2010).
5P. D. Howes, R. Chandrawati, and M. M. Stevens, Science 346, 1247390
(2014).
6H. Vach, Phys. Rev. Lett. 112, 197401 (2014).
7A. Firooz, S. Nafisi, and H. I. Maibach, Int. J. Pharm. 495, 599 (2015).
8C. Maciel, E. E. Fileti, and R. Rivelino, Chem. Phys. Lett. 507, 244
(2011).
9K. Kokubo, K. Matsubayashi, H. Tategaki, H. Takada, and T. Oshima,
ACS Nano 2, 327 (2008).
10R. Zangi, M. Hagen, and B. J. Berne, J. Am. Chem. Soc. 129, 4678 (2007).
11G. Graziano, Chem. Phys. Lett. 499, 79 (2010).
12S. Sarupria and S. Garde, Phys. Rev. Lett. 103, 037803 (2009).
13R. Godawat, S. N. Jamadagni, and S. Garde, Proc. Natl. Acad. Sci. U. S. A.
106, 15119 (2009).
14D. Chandler, Nature 437, 640 (2005).
15A. L. Ferguson, N. Giovambattista, P. J. Rossky, A. Z. Panagiotopoulos, and
P. G. Debenedetti, J. Chem. Phys. 137, 144501 (2012).
16H. Xu and K. A. Dill, J. Phys. Chem. B 109, 23611 (2005).
17G. Hummer, S. Garde, A. E. Garcia, A. Pohorille, and L. R. Pratt, Proc. Natl.
Acad. Sci. U. S. A. 93, 8951 (1996).
18R. C. Remsing and J. D. Weeks, J. Phys. Chem. B 117, 15479 (2013).
19T. M. Raschke and M. Levitt, Proc. Natl. Acad. Sci. U. S. A. 102, 6777
(2005).
20K. Lum, D. Chandler, and J. D. Weeks, J. Phys. Chem. B 103, 4570 (1999).
21S. Rajamani, T. M. Truskett, and S. Garde, Proc. Natl. Acad. Sci. U. S. A.
102, 9475 (2005).
22X. Huang, C. J. Margulis, and B. J. Berne, Proc. Natl. Acad. Sci. U. S. A.
100, 11953 (2003).
23N. Giovambattista, P. G. Debenedetti, and P. J. Rossky, J. Phys. Chem. B
111, 9581 (2007).
24N. Giovambattista, P. G. Debenedetti, and P. J. Rossky, J. Phys. Chem. C
111, 1323 (2007).
25B. Hess, C. Kutzner, D. van der Spoel, and E. Lindahl, J. Chem. Theory
Comput. 4, 435 (2008).
26W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey, and M. L.
Klein, J. Chem. Phys. 79, 926 (1983).
27T. Darden, D. York, and L. Pedersen, J. Chem. Phys. 98, 10089 (1993).
28K. N. Semenov, N. A. Charykov, and V. N. Keskinov, J. Chem. Eng. Data
56, 230 (2011).
29See supplementary material at http://dx.doi.org/10.1063/1.4959805 for a
variety of additional figures mentioned in the text.
30W. L. Jorgensen, D. S. Maxwell, and J. Tirado-Rives, J. Am. Chem. Soc.
118, 11225 (1996).
31W. L. Jorgensen and J. Gao, J. Phys. Chem. 90, 2174 (1986).
32W. Li and Y. Mu, Nanoscale 4, 1154 (2012).
33L. Hua, R. Zhou, D. Thirumalai, and B. J. Berne, Proc. Natl. Acad. Sci.
U. S. A. 105, 16928 (2008).
34P. Das and R. Zhou, J. Phys. Chem. B 114, 5427 (2010).
35D. R. Weiss, T. M. Raschke, and M. Levitt, J. Phys. Chem. B 112, 2981
(2008).
36J. Gebhardt, F. Vines, P. Bleiziffer, W. Hieringer, and A. Gorling,
Phys. Chem. Chem. Phys. 16, 5382 (2014).
37C. Wang, B. Zhou, Y. Tu, M. Duan, P. Xiu, J. Li, and H. Fang, Sci. Rep. 2,
358 (2012).
38N. Choudhury, J. Chem. Phys. 125, 34502 (2006).
39J. L. Finney and A. K. Soper, Chem. Soc. Rev. 23, 1 (1994).
40X. Ou, J. Li, and Z. Lin, J. Phys. Chem. C 118, 29887 (2014).
41D. Argyris, T. Ho, D. R. Cole, and A. Striolo, J. Phys. Chem. C 115, 2038
(2011).
42C. R. Catlow, S. T. Bromley, S. Hamad, M. Mora-Fonz, A. A. Sokol, and
S. M. Woodley, Phys. Chem. Chem. Phys. 12, 786 (2010).
43P. J. Ziemann and A. W. Castleman, J. Chem. Phys. 94, 718 (1991).