View Article Online

View Journal

PCCP
Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: B. S. Dwadasi, R.
Gupta and B. Rai, Phys. Chem. Chem. Phys., 2018, DOI: 10.1039/C8CP04044C.

Volume 18 Number 1 7 January 2016 Pages 1–636 This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
PCCP
Physical Chemistry Chemical Physics Accepted Manuscripts are published online shortly after
www.rsc.org/pccp

acceptance, before technical editing, formatting and proof reading.

Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.

You can find more information about Accepted Manuscripts in the

author guidelines.

Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s
ISSN 1463-9076 standard Terms & Conditions and the ethical guidelines, outlined
PERSPECTIVE
in our author and reviewer resource centre, still apply. In no
Darya Radziuk and Helmuth Möhwald
Ultrasonically treated liquid interfaces for progress in cleaning and
separation processes
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.

rsc.li/pccp
 Page 1 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

Effect of Surface Coverage and Chemistry

on Self-Assembly of Monolayer Protected

Physical Chemistry Chemical Physics Accepted Manuscript

Gold Nanoparticles: A Molecular Dynamics
Simulation Study
Dwadasi Balarama Sridhar, Rakesh Gupta* and Beena Rai
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Physical Science Research Area, TCS Research

Tata Research Development and Design Centre, Tata Consultancy Services,
54B, Hadapsar Industrial Estate, Pune – 411013, India
*Corresponding author: gupta.rakesh2@tcs.com
Fax: 91-20-66086399
Tel: 91-20-66086203

Abstract

Gold Nanoparticles (AuNP) are being used in variety of applications ranging from drug
delivery systems to nano-coolants. The structure and surface properties of AuNP can be
manipulated using coatings of thiols, carrying different charges. Different densities of surface
coverages can result in formation of different structures. A molecular basis to quantify the
interactions between AuNP covered with different densities (20, 60 and 100%) of surface
coverage and various thiol chains (neutral, cationic and anionic) is obtained using Potential of
Mean Force (PMF) calculations. Further self-assembly simulations were performed to study
the formation of aggregates/dispersed solutions with different densities of surface coverage
(20, 40, 60, 80 and 100%). It is found that neutral coatings in all surface coverage densities
studied, and charged coatings (anionic and cationic) of less than 60% surface coverage
density results in formation of aggregates. The aggregation and dispersion of AuNPs is
explained in terms of an interplay between van der Waals and electrostatic forces. The results
obtained are in good agreement with the data available in literature and suggest that
aggregation behaviour can be controlled by modifying the surface coverage and chemistry.
 Physical Chemistry Chemical Physics Page 2 of 21
View Article Online
DOI: 10.1039/C8CP04044C

1. Introduction

Nanoparticles are currently being used in various applications such as medical1,2, cosmetics3,

Physical Chemistry Chemical Physics Accepted Manuscript

energy,4,5 and electronics,6 to name a few. The manipulation of material properties at nano-
scale has led to the development of various kind of nanomaterials for use in various
biomedical applications since last few decades.7-9 Numerous kind of nanoparticles such as
metal, sugars, polymers, biopolymers, lipids, organic and inorganic nanoparticles have been
synthesized and used in various applications.10,11 Gold nanoparticles (AuNPs) attracted a lot
of attention from researchers since past few decades owing to their ease of preparation and
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

surface modification by various chemical or bioactive molecules.12,13

Researchers have used various surface coated AuNPs for applications such as drug delivery,
gene/protein delivery and imaging.10-16 In the past, it has been shown through extensive
experiments17-21 and simulations22-25 that the gold nanoparticles, having various surface
morphology and size, were able to deliver drugs in cell membrane and in the deeper layer of
the skin as well. A firm control over the interaction between the nanoparticle and biological
system has led to the successful application of these nanoparticle technology. These
interactions arrive mainly due to the surface properties of nanoparticles. The permeation of
AuNPs in experiments17-21 and in simulations22-25 has been found to be highly dependent on
surface morphology, size and shape of gold nanoparticles.

Engineered AuNPs are generally modified by the self-assembled monolayers (SAMs). The
SAMs are organic assemblies formed by adsorption of ligand molecules over solid surfaces.14
These monolayers form tightly packed and highly ordered structure over the AuNP and have
ability to elucidate fundamental properties of biological interactions at molecular scales.14
The surface morphology of SAMs are determined by the kind of ligand molecules employed
to coat the AuNPs. It has been shown that, for a binary mixture of ligands, various 3D pattern
(such as patched, striped or Janus domains) on gold nanoparticles may form based on the size
and shape of the ligands.14,16,26 The situation becomes more complex and more number of
patterns are possible if multiple ligands are present in SAMs.

The optical27, electronic28 and biocompatibility29 properties of monolayer protected AuNPs

are highly affected by their agglomeration. Because of it, the assemblies of monolayer
protected AuNPs into the super structure are of great interest for researchers. The study of
agglomeration behaviour of SAMs protected AuNPs in aqueous solution will also give
fundamental insight for their use as delivery or therapeutic agents.29-31 The surface coverage
 Page 3 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

of AuNPs by SAMs plays an important role in their agglomeration/dispersion. Kim et al.32

performed experiments on various degrees of surface coverage of gold nanoparticles by poly-
(styrenesulfonate)-doped poly(3,4-ethylenedioxythiophene) (PEDOT:PSS). It was shown that

Physical Chemistry Chemical Physics Accepted Manuscript

fully covered AuNP surface led to particle size of 10 nm while partially covered AuNPs led
to particle size of 50 nm.32 Yang et al.33 reported that line-assembly mechanism of gold nano-
chain was due to the preferential binding of cetyltrimethylammonium bromide (CTAB)
molecules on a certain facet of AuNP and a balance of AuNP electrostatic interactions. It was
also shown that the 11-mercaptoundecanoic acid was effective to prevent the aggregation of
CTAB-modified gold colloids.33
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

However, molecular level detailed information of the aggregation behaviour of monolayer-

protected AuNPs is still lacking. Molecular dynamics (MD) simulations provide a convenient
way to gain critical insights happening at molecular levels that are difficult to obtain from
experiments because of associated time and length scale. Potential of Mean Force (PMF)
calculated from MD simulations gives effective interaction between particles in a given
media.

In this study, we report the molecular mechanism of agglomeration of monolayer coated

AuNPs of various surface coverage. Three AuNPs, a) neutral, b) cationic and c) anionic with
varying percentage of surface coverage are studied by means of coarse grained (CG) MD
simulations. We first show how the interactions between AuNP of different surface coverage
changes by means of potential of mean force (PMF) calculations. Later on, the effect of
surface coverage on agglomeration behaviour of each AuNP, is reported. We also report the
effect of concentration of AuNP on the aggregation behaviour.

2 System, Models and Parameters

MD simulations are faster than first principle methods but still achieving time scale of
microseconds is highly computationally demanding using atomistic simulations. Earlier
simulation studies,34,35 performed using atomistic model of AuNP, were limited to use only
two AuNP. In order to model the long time scale phenomena happening at micro second scale,
the CG model (based on the MARTINI force field )36,37 is used. In the MARTINI model, the
atomistic atoms are mapped to CG beads using various mapping rules37. Generally, on an
average four or three (for ring structures) heavy atoms are represented by a single bead. Four
different kind of beads namely polar (P), nonpolar (N), apolar (C), and charged (Q), are
considered. Further, each bead is divided into several subtypes, which facilitate more accurate
 Physical Chemistry Chemical Physics Page 4 of 21
View Article Online
DOI: 10.1039/C8CP04044C

representation of the chemical nature of the underlying atomic structure. For a given bead
type, subtypes are distinguished by two parameters: (i) based on hydrogen-bonding
capabilities (d=donor, a=acceptor, da=both and 0 =none) and (ii) based on degree of polarity

Physical Chemistry Chemical Physics Accepted Manuscript

(from 1, low polarity, to 5, high polarity). There are in total 18 subtypes present in MARTINI
force field. These mapping scheme can be found in original publication of MARTINI force
field.36,37
The CG parameters of thiol coated AuNPs were taken from previous simulation
studies.22,23,38,39 Using these parameters, the effect of size and surface charge on permeation of
AuNP (2-5 nm in size) through a model skin bilayer was previously studied in our group23.
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Although in principle, AuNP of any given size can be studied using CGMD, owing to
computational restraints, we have used an AuNP of size 3nm in the present study for all the
three cases. The gold (Au) atoms were mapped to one C5 bead in MARTINI
Forcefield22,23,38,39. In order to obtain the structure of a bare AuNP, we first created an FCC
lattice of Au beads of size 3nm. The AuNP of 3nm is then carved out of this lattice with the
distance between the centre of the AUNP and surface beads being 3nm. Each bare AuNP has
887 beads in it. The forcefield parameters used are given in supporting information (Fig S6
and Table S2). For modeling the charged AuNPs, the terminal beads of thiol chain of coated
AuNP were assigned a charge of +1 or -1 for cationic and anionic AuNP, respectively.
Adequate numbers of counter ions were added to the system to maintain the neutrality. More
details on the parameters used in our simulation can be found in Supporting Information (Fig
S6 and Table S2). In order to prepare systems of different surface coverages, a system of bare
gold nanoparticle was first equilibrated with excess of the dodecane thiol (DDT) molecules
(neutral, cationic and anionic) and the molecules were let to form a mono layer on the
particle. The excess molecules were then removed and a system of maximum surface
coverage was obtained. Going further, this will be referred to as 100% surface coverage.
Correspondingly, nanoparticles with 20, 40, 60 and 80 % coverages are prepared, by
randomly removing the excess chains. The details of the number of thiol chains used under
different surface coverage is given in Table S1 of Supporting Information. The CG model of
various NPs used in this study are shown in Fig. 1 (a-e).
To start with, simulations were performed to calculate potential of mean force (PMF) between
AuNP in aqueous solution. PMF between the nanoparticles is a measure of the effective
interaction between AuNP in the given medium. We have employed Umbrella Sampling
simulations40 along with Weighted Histogram Analysis Method (WHAM)41 to obtain the
PMF. As these calculations are expensive and time consuming, nanoparticles with 20, 60 and
 Page 5 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

100 % surface coverage of neutral, cationic and anionic thiols were only taken. The AuNP
were placed 9 nm apart and were pulled till they are in contact to generate initial
configurations. Windows of 0.2 nm were taken for the Umbrella Sampling. Each window was

Physical Chemistry Chemical Physics Accepted Manuscript

equilibrated for 100 ns at 310 K and 1 atmosphere and was further run for 200 ns to collect
data. The resulting force files were analyzed using the g_wham42 module of GROMACS43-45
to obtain the potential of mean force. In case of self-assembly simulations, for each system,
total 4 AuNPs were packed in a box of 20nm x 20 nm x 20 nm and filled with 30000 CG
water beads (1 CG water bead = 4 water molecules). The nanoparticles are kept apart from
each other as shown in Fig. 1 (f). Each system was energy minimized using steepest decent
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

method. The systems were further equilibrated for 100 ns each in NVT and NPT ensemble
conditions by keeping AuNPs fixed to ensure the proper solvation. Properly solvated systems
were then finally run for 500 ns in NPT ensemble without restraints, to study the dynamic
process of self-assembly. However, one should note that the effective times of simulations are
much higher (up to four times36,37) owing to the smoothened energy barriers in a coarse
grained simulation.
All simulations were carried out in NVT and NPT ensemble using the GROMACS MD
package.43-45 The temperature was controlled at a temperature of ~ 310 K, using the
Berendsen (equilibration run) and Nose-Hoover (production run) thermostat with a time
constant of 2 ps. Pressure was controlled by Berendsen (equilibration run) and Parrinello-
Rahman (production run) barostat with a time constant of 6 and 12 ps, respectively and
compressibility of 4.0 x 10-5 bar-1 with isotropic coupling. The LJ potentials were smoothly
shifted to zero between a distance rshift = 0.9 nm and the cutoff distance of 1.2 nm. The pair
list was updated at every 20 steps. The timestep used in all the simulations was 20 fs. The
configuration was sampled at every 200 ps in the production run.
 Physical Chemistry Chemical Physics Page 6 of 21
View Article Online
DOI: 10.1039/C8CP04044C

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Figure 1. Nanoparticles with 20, 40, 60, 80 and 100% surface coverage used in the system
(a-e) respectively. The nanoparticle is represented in yellow in Vdw style, Sulphur atom is in
pink and the Carbon chain is shown in cyan. f. Initial structure of the simulation cell taken
(water is omitted for the sake of clarity), where the AuNP are kept apart to avoid any bias.
AuNP 1,2,3 and 4 are coloured in red, gray, orange and yellow respectively. Images are
obtained using VMD Software58

3. Results and Discussions

3.1 Potential of Mean Force Calculations
Typically, Zeta Potential (ZP) is used as a measure to quantify the stability of a suspension. It
is believed that a ZP lower than -30 mV and higher than +30 mV results in a highly stable
suspension46. The ZP can be captured using Non Equilibrium Molecular Dynamics (NEMD)
simulations47-48. However, ZP captures only electro static repulsions leaving the van der
Waals attraction forces46. The DLVO theory for colloids46,49 considers electro static repulsion
and van der Waals attraction between particles, where the suspension is said to be stable if
the electro static repulsion dominates the van der Waals attraction. PMF captures all these
interactions and gives a good measure to comment upon the stability of a suspension. The
PMF gives the effective interaction between the particles, including the effects of solvent. As
the particles are pulled closer, a highly negative PMF indicates attraction between the
particles, thereby indicating aggregation. Constrained MD simulations (Umbrella Sampling)
were performed in order to obtain the PMF between nanoparticles of varying surface
coverage (20, 60 and 100%) for the three cases of neutral, anionic and cationic coating. PMF
 Page 7 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

profile was obtained by employing the Weighted Histogram Analysis Method (WHAM).
Figure 2 shows the PMF profiles for all the systems. The reaction coordinate represented by ξ
is the distance between the centre of mass of the two nanoparticles taken. For the sake of

Physical Chemistry Chemical Physics Accepted Manuscript

clarity, error bars are not shown here; the individual PMF profiles with error bars for each
system are plotted in Fig. S2 (Please see supporting information). In case of charged coatings
(fig. 2b and 2c), the PMF indicates that interaction between particles with 100% coverage
turned repulsive as compared to those with 20 and 60% charged coverage. This could be due
to the increased effective charge on the particles, which leads to an increased electrostatic
repulsion, dominating the van der Waals attraction between the particles. This indicates that,
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

a suspension of AuNP coated with less than 60% of charged thiols (cationic and anionic) will
result in aggregates, while 100% coverage will remain stable. This was also observed in
experiments, where fully covered similar charged coatings resulted in stable dispersions50
while a modification in the charge coverage, resulted in controlled formation of aggregates33.
However, in case of neutral thiol coated AuNP, the interaction between the nanoparticles
turned more attractive as the surface coverage increased (fig. 2a). This could be attributed to
the fact that although the core interactions of the AuNP get shielded with the increased
surface coverage, the hydrophobic chains in the aqueous medium tend to form entanglements.
The non-DLVO hydrophobic force comes into picture, as can be seen from the entanglements
formed between the hydrophobic chains in water (fig. S1 of Supporting Information). Such
entanglements were also observed experimentally using TGA and SAX measurements51. This
indicates that, neutral thiol coating leads to the formation of aggregates in case of any surface
coverage.

Figure 2. Potential of Mean Force between two nanoparticles in aqueous medium. The
alphabet N, C and A indicate Neutral (a), Cationic (b) and Anionic (c) coating respectively.
The numbers 20,60 and 100 indicate the percentage of Surface coverage
 Physical Chemistry Chemical Physics Page 8 of 21
View Article Online
DOI: 10.1039/C8CP04044C

3.2 Self Assembly Simulations

The results of PMF showed that charged thiol coatings of 20 and 60% may result in

Physical Chemistry Chemical Physics Accepted Manuscript

formation of aggregates, while, in case of neutral coating, all the cases studied may lead to
the formation of aggregates. To check this observation obtained from the PMF profiles,
unconstrained simulations of each nanoparticle were performed. In each system, four
nanoparticles were kept apart, and solvated in water, as explained in the section on system
taken. The size of the nanoparticle, as measured by the radius of gyration increased as the
surface coverage increased. Figure 3 shows the final structures of the system (solvent is not
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

shown for clarity) at the end of 500 ns simulation in case of neutral thiol coated AuNP. As
suggested by the PMF calculations, the nanoparticles aggregate in all cases to form different
shapes, starting with a rod like shape at low surface coverage to a more spherical form as the
surface coverage increases. At low surface coverage, the AuNP, which are hydrophobic in
nature tend to form aggregates in the aqueous medium. Given the surface coverage is less, the
cores of the AuNP are close to each other. On the other hand, as the surface coverage
increases the hydrophobicity of the chains covering the AuNP comes into play, and the chains
interdigitate to form entanglements. This is verified by the fact that Pradeep et al.52 observed
interdigitation of chains in case of AuNP coated with DDT using Quasielastic Neutron
Scattering experiments. Terrill et. al.51 suggested interdigitation of chains in case of a fully
covered system using TGA analysis and SAX measurements. In this case, the AuNP cores
seem to be relatively far from each other. This is quantitatively seen in Fig. 4, where the
distances between all the AuNP pairs (AuNP-AuNP COM distance) are plotted. It can be seen
that the distance between the AuNP forming an aggregate is increasing as the surface
coverage is increasing. In the simulated time, all the systems formed aggregates which
remained stable once formed. Terrill et. al.51 have estimated the center-center distance
between AuNP cores in fully covered system to be around 3.896 nm. The distances between
centre of mass of AuNP, from Fig. 4 (e) falls in this range. Unlike PEG coverage53 where the
increased surface coverage showed a stable suspension, in case of neutral DDT increasing the
surface coverage led to interdigitation of chains, leading to formation of aggregates. Lin et
al.39 have also shown formation of similar aggregates in case of fully covered neutral DDT
coated AuNP using CG MD simulation. Meena Devi54 has performed united atom MD
simulations of thiol coated 0.78 nm AuNP and found similar formation of aggregates at
higher surface coverage density. At lower surface coverage density, they observed formation
of small size aggregates (couple of nano particles) over 100 ns. However, we observed that
 Page 9 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

over a time of 500 ns, which is effectively 2 µs as pointed out earlier (4 times higher effective
time), the AuNP formed stable aggregates in all the cases of neutral thiol coating.

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Figure 3. Structures of the AuNP formed at the end of 500 ns of simulation for neutral thiol
coated AuNP with a. 20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage
respectively. The nanoparticle is represented in Vdw style, Sulphur atom is in pink and the
Carbon chain is shown in cyan. AuNP 1,2,3 and 4 are coloured in red, gray, orange and
yellow respectively. Images are obtained using VMD Software58
 Physical Chemistry Chemical Physics Page 10 of 21
View Article Online
DOI: 10.1039/C8CP04044C

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Figure 4. Distances between centres of mass of pairs of neutral thiol coated AuNP with a.
20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage, respectively. Four AuNP are
taken. The numbers indicate the AuNP under consideration. Distance between AuNP in direct
contact is shown by a line and between AuNP not in contact is shown by a dotted line.

Further, systems of AuNP coated with cationic and anionic thiol coating were simulated.
Figures 5 and 6 show the final structures at the end of 500 ns simulation. One should note
that, as the surface coverage increases, the effective charge on each of the AuNP increases. In
both cases, at low surface coverages, the AuNP core interaction dominates the charge effect
and tend to form aggregates. However, at higher coverage, the electrostatic repulsion comes
into picture, and the AuNP settle away from each other. Similar phenomenon has been
observed by Liu et. Al.50 using Transmission Electron Microscopy (TEM), where they found
AuNP coated with citrate ions (negatively charged) dispersed steadily in a solution. Yang et.
Al.33 have observed that nanoparticles stabilized by citrates remain stable and aggregation
behaviour of the AuNP can be controlled by modifying the monolayer.
 Page 11 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Figure 5. Structures of the AuNP formed at the end of 500 ns of simulation for cationic thiol
coated AuNP with a. 20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage
respectively. The nanoparticle is represented in Vdw style, Sulphur atom is in pink and the
Carbon chain is shown in cyan. AuNP 1,2,3 and 4 are coloured in red, gray, orange and
yellow respectively. Images are obtained using VMD Software58

Figure 6. Structures of the AuNP formed at the end of 500 ns of simulation for anionic thiol
coated AuNP with a. 20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage
respectively. The nanoparticle is represented in Vdw style, Sulphur atom is in pink and the
Carbon chain is shown in cyan. AuNP 1,2,3 and 4 are coloured in red, gray, orange and
yellow respectively. Images are obtained using VMD Software58
 Physical Chemistry Chemical Physics Page 12 of 21
View Article Online
DOI: 10.1039/C8CP04044C

Figures 7 and 8 show the AuNP-AuNP distance profiles along the trajectory of the cationic
and anionic thiol coated AuNP, respectively. It can be seen that, at low coverages, the AuNP
settle at closer distance, indicating the formation of aggregates. As the coverage increases, the

Physical Chemistry Chemical Physics Accepted Manuscript

distance profile shows disorder and the distance being relatively higher. This can be
explained by the PMF result, which showed repulsive interactions at 100 % coverages. Thus
by visual inspection of structure, distance analysis, and PMF calculation we can comment
that at coverages over 80%, the AuNPs do not form aggregates. Similar observation was
made by Lin et al.39 where they have simulated 2.2 nm sized AuNP coated with charged thiol
ligands at full surface coverage. However, their claim that AuNP coated with charged thiol
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

chains is valid only for AuNP with high surface coverage. As seen from our results, even
charged thiol coated AuNP aggregate at low surface coverages. Thus, surface coverage along
with the charge on the ligands is important in determining the stability of the suspension.

Figure 7. Distances between centres of mass of pairs of cationic thiol coated AuNP with a.
20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage respectively. Four AuNP are
taken. The numbers indicate the AuNP under consideration. Distance between AuNP in direct
contact is shown by a line and between AuNP not in contact is shown by a dotted line.
 Page 13 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Figure 8. Distances between centres of mass of pairs of anionic thiol coated AuNP with a.
20%, b. 40%, c. 60%, d. 80% and e. 100% surface coverage respectively. Four AuNP are
taken. The numbers indicate the AuNP under consideration. Distance between AuNP in direct
contact is shown by a line and between AuNP not in contact is shown by a dotted line.

To quantify the aggregation, aggregation propensity (AP) is determined based on the

definition of Frederix et. al.55, which is defined as the ratio of solvent accessible surface area
(SASA) of the AuNP in the initial configuration to the final configuration obtained at the end
of the simulation. AP value greater than 1 indicates formation of aggregates, while anything
less than or equal to 1 indicates otherwise. However, one should note that in case of AuNP
with charged coating, the SASA increases as the AuNP gets closer due to the effect of charge
interactions among the chains on AuNP. This results in low values of AP as the charge on
AuNP increases, even if they form aggregates. SASA was calculated using the gmx sasa
module of GROMACS43-45 using a probe radius of 0.25 nm, which is closer to the van der
Waals radius of the MARTINI CG bead56. Table 1 shows the values of AP calculated for
each system. The AP closely follows the result of PMF, as neutral or low charged AuNP tend
to form aggregates in water, while, AuNP with increased charged surface coverage resulted
in stable dispersion. The cluster density, as calculated using the module, for 100% neutral
thiol coated AuNP was 3.354 g/cc, while, Terrill et. al.51 have reported a value of 3.06 g/cc.
 Physical Chemistry Chemical Physics Page 14 of 21
View Article Online
DOI: 10.1039/C8CP04044C

This shows that the CG model is successful in capturing the dynamics of self-assembly of
AuNP.

Physical Chemistry Chemical Physics Accepted Manuscript

Systems with 8 AuNP were also simulated to find the effect of concentration on the
formation of aggregates. The number of water molecules were kept equal to the previous
case. The final configurations are shown in Fig. S3-S5 in Supporting Information. Table 1
shows the value of AP when 8 nanoparticles are taken. The AP value reduces in case of
charged coating (cationic and anionic) as the percentage of coverage increases and is least for
80 and 100% surface coverages. This is in concurrence with the results obtained from PMF
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

calculations, that charged coatings with high surface coverage result in a stable dispersion. In
case of neutral thiol coating, the AP continued to increase and was higher than 1 in all the
cases, indicating formation of aggregates. This shows that, irrespective of the concentration,
surface coverage and the charges on the ligands play a major role in the stability of AuNP
suspensions.

Table 1. Aggregation Propensities Calculated for Different systems. The alphabet A, N and C
indicates Anionic, Neutral and Cationic respectively. The number indicates percentage of
surface coverage

System AP AP Final Structure

4 Nano particles 8 Nano particles
A_20 0.96 1.04 Aggregates
A_40 0.78 0.84 Aggregates
A_60 0.77 0.79 Small Aggregates
A_80 0.77 0.64 Dispersed
A_100 0.77 0.62 Dispersed
N_100 1.24 1.57 Aggregates
N_80 1.16 1.62 Aggregates
N_60 1.49 1.98 Aggregates
N_40 1.44 2.03 Aggregates
N_20 1.49 2.01 Aggregates
C_20 0.91 1.08 Aggregates
C_40 0.72 0.81 Aggregates
C_60 0.79 0.88 Small aggregates
C_80 0.76 0.76 Dispersed
 Page 15 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

C_100 0.78 0.72 Dispersed

Although we have used CGMD to study the interactions between AuNP and were able to
relate it to the self-assembly of AuNP in aqueous solution, we have to remind the readers that

Physical Chemistry Chemical Physics Accepted Manuscript

these simulations are limited to the time scale of few microseconds and length scale of few
nanometers. Large numbers of nanoparticles of various sizes have to be simulated to study
formation of phases57 or the morphology of structure33,50-52,57. Higher order computational
methods might be helpful in studying such systems. Also, a detailed atomic level study will
help in capturing hydrogen bonding phenomenon between water and the functional groups on
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

the thiol chain, but capturing such events using atomistic simulation is still computationally
expensive.

Conclusion

The effect of surface coverage and chemistry on the aggregation of thiol coated AuNP was
studied using Coarse Grained Molecular Dynamics Simulations. Neutral, Cationic and
Anionic thiol coatings were considered. Molecular basis for the interaction between coated
AuNPs was established using PMF calculations. Neutral thiol coated AuNP at all surface
coverage densities and charged thiol coated AuNP at low surface coverage densities showed
an attractive interaction, indicating a propensity to form aggregates, while dense charged thiol
coated AuNP showed the opposite phenomenon. Self-assembly simulations with four AuNP
in water supported the results of PMF. Aggregation propensity of the AuNP was calculated
and gives a measure of the tendency of AuNP to form aggregates. Effect of concentration of
AuNP on aggregation behaviour was studied. These results suggest that the aggregation
behaviour of AuNP could be controlled by changing the surface coverage density on the
AuNP and choosing the charge on the ligand.

Funding & Resources

This research was funded by Tata Consultancy Services (TCS), CTO organization.
Acknowledgement
Authors would like to thank High Performance Computing at Tata Consultancy Services
(TCS) for providing access to EKA Super computer and Mr. K Ananth Krishnan, CTO and
EVP, Tata Consultancy Services for his constant encouragement and support during this
project
 Physical Chemistry Chemical Physics Page 16 of 21
View Article Online
DOI: 10.1039/C8CP04044C

Supporting Information
S1. Entanglement between fully covered neutral thiol coated AuNP; S2. PMF profiles
with error bars; S3. Final configurations of 8 AuNP with different percentage of surface

Physical Chemistry Chemical Physics Accepted Manuscript

coverage of neutral thiol; S4. Final configurations of 8 AuNP with different percentage
of surface coverage of cationic thiol and S5. Final configurations of 8 AuNP with
different percentage of surface coverage of anionic thiol

References
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

1. Salata OV. Applications of nanoparticles in biology and medicine. Journal of

nanobiotechnology. 2004 Dec;2(1):3.

2. Decuzzi P, Mitragotri S. Introduction to special issue on “Nanoparticles in medicine:

Targeting, optimization and clinical applications”. Bioengineering & translational medicine.
2016 Mar;1(1):8-9.

3. Raj S, Jose S, Sumod US, Sabitha M. Nanotechnology in cosmetics: Opportunities

and challenges. Journal of pharmacy & bioallied sciences. 2012 Jul;4(3):186.

4. Wang ZL, Wu W. Nanotechnology enabled energy harvesting for self powered

micro /nanosystems. Angewandte Chemie International Edition. 2012 Nov
19;51(47):11700-21.

5. Guo KW. Green nanotechnology of trends in future energy: a review. International

journal of energy research. 2012 Jan;36(1):1-7.

6. Cao Q, Rogers JA. Ultrathin films of single walled carbon nanotubes for electronics
and sensors: a review of fundamental and applied aspects. Advanced Materials. 2009 Jan
5;21(1):29-53.

7. West JL, Halas NJ. Engineered nanomaterials for biophotonics applications:

improving sensing, imaging, and therapeutics. Annual Review of Biomedical Engineering.
2003 Aug;5(1):285-92.

8. Nalwa HS. A special issue on reviews in biomedical applications of nanomaterials,

tissue engineering, stem cells, bioimaging, and toxicity. Journal of biomedical
nanotechnology. 2014 Oct 1;10(10):2421-3.
 Page 17 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

9. Lavik E, Von Recum H. The role of nanomaterials in translational medicine. ACS

nano. 2011 May 24;5(5):3419-24.

Physical Chemistry Chemical Physics Accepted Manuscript

10. DeLouise LA. Applications of nanotechnology in dermatology. Journal of
Investigative Dermatology. 2012 Mar 1;132(3):964-75.

11. Heddle JG. Gold nanoparticle-biological molecule interactions and catalysis.

Catalysts. 2013 Sep 3;3(3):683-708.

12. Tiwari PM, Vig K, Dennis VA, Singh SR. Functionalized gold nanoparticles and their
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

biomedical applications. Nanomaterials. 2011 Jun 14;1(1):31-63.

13. Daniel MC, Astruc D. Gold nanoparticles: assembly, supramolecular chemistry,

quantum-size-related properties, and applications toward biology, catalysis, and
nanotechnology. Chemical reviews. 2004 Jan 14;104(1):293-346.

14. Pengo P, Şologan M, Pasquato L, Guida F, Pacor S, Tossi A, Stellacci F, Marson D,

Boccardo S, Pricl S, Posocco P. Gold nanoparticles with patterned surface monolayers for
nanomedicine: current perspectives. European Biophysics Journal. 2017 Dec 1;46(8):749-71.

15. Kobayashi K, Wei J, Iida R, Ijiro K, Niikura K. Surface engineering of nanoparticles

for therapeutic applications. Polymer journal. 2014 Aug;46(8):460.

16. Verma A, Stellacci F. Effect of surface properties on nanoparticle–cell interactions.

Small. 2010 Jan 4;6(1):12-21.

17. Dean HJ, Haynes J, Schmaljohn C. The role of particle-mediated DNA vaccines in
biodefense preparedness. Advanced drug delivery reviews. 2005 Jun 17;57(9):1315-42.

18. Nativo P, Prior IA, Brust M. Uptake and intracellular fate of surface-modified gold
nanoparticles. ACS nano. 2008 Jul 22;2(8):1639-44.

19. Sonavane G, Tomoda K, Sano A, Ohshima H, Terada H, Makino K. In vitro

permeation of gold nanoparticles through rat skin and rat intestine: effect of particle size.
Colloids and Surfaces B: Biointerfaces. 2008 Aug 1;65(1):1-0.

20. Labouta HI, El-Khordagui LK, Kraus T, Schneider M. Mechanism and determinants
of nanoparticle penetration through human skin. Nanoscale. 2011;3(12):4989-99.
 Physical Chemistry Chemical Physics Page 18 of 21
View Article Online
DOI: 10.1039/C8CP04044C

21. Huang Y, Yu F, Park YS, Wang J, Shin MC, Chung HS, Yang VC. Co-administration
of protein drugs with gold nanoparticles to enable percutaneous delivery. Biomaterials. 2010
Dec 1;31(34):9086-91.

Physical Chemistry Chemical Physics Accepted Manuscript

22. Gupta R, Rai B. Penetration of gold nanoparticles through human skin: unraveling its
mechanisms at the molecular scale. The Journal of Physical Chemistry B. 2016 Jul
15;120(29):7133-42.

23. Gupta R, Rai B. Effect of size and surface charge of gold nanoparticles on their skin
permeability: a molecular dynamics study. Scientific Reports. 2017 Mar 28;7:45292.
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

24. Gupta R, Kashyap N, Rai B. Molecular mechanism of transdermal co-delivery of

interferon-alpha protein with gold nanoparticle–a molecular dynamics study. Molecular
Simulation. 2018 Mar 4;44(4):274-84.

25. Gupta R, Kashyap N, Rai B. Transdermal cellular membrane penetration of proteins

with gold nanoparticles: a molecular dynamics study. Physical Chemistry Chemical Physics.
2017;19(11):7537-45.

26. Ghorai PK, Glotzer SC. Atomistic simulation study of striped phase separation in
mixed-ligand self-assembled monolayer coated nanoparticles. The Journal of Physical
Chemistry C. 2010 Oct 25;114(45):19182-7.

27. Barnes WL, Dereux A, Ebbesen TW. Surface plasmon subwavelength optics. nature.
2003 Aug 14;424(6950):824.

28. Zhang Y, Gu C, Schwartzberg AM, Chen S, Zhang JZ. Optical trapping and light-
induced agglomeration of gold nanoparticle aggregates. Physical Review B. 2006 Apr
6;73(16):165405.

29. Albanese A, Chan WC. Effect of gold nanoparticle aggregation on cell uptake and
toxicity. ACS nano. 2011 Jun 21;5(7):5478-89.

30. Shukla R, Bansal V, Chaudhary M, Basu A, Bhonde RR, Sastry M. Biocompatibility

of gold nanoparticles and their endocytotic fate inside the cellular compartment: a
microscopic overview. Langmuir. 2005 Nov 8;21(23):10644-54.

31. Peer D, Karp JM, Hong S, Farokhzad OC, Margalit R, Langer R. Nanocarriers as an
emerging platform for cancer therapy. Nature nanotechnology. 2007 Dec;2(12):751.
 Page 19 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

32. Kim SH, Bae TS, Heo W, Joo T, Song KD, Park HG, Ryu SY. Effects of gold-
nanoparticle surface and vertical coverage by conducting polymer between indium tin oxide
and the hole transport layer on organic light-emitting diodes. ACS applied materials &

Physical Chemistry Chemical Physics Accepted Manuscript

interfaces. 2015 Jun 29;7(27):15031-41.

33. Yang Y, Matsubara S, Nogami M, Shi J. Controlling the aggregation behavior of gold
nanoparticles. Materials Science and Engineering: B. 2007 Jun 15;140(3):172-6.

34. Schapotschnikow P, Pool R, Vlugt TJ. Selective adsorption of alkyl thiols on gold in
different geometries. Computer physics communications. 2007 Jul 1;177(1-2):154-7.
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

35. Schapotschnikow P, Pool R, Vlugt TJ. Molecular simulations of interacting

nanocrystals. Nano letters. 2008 Aug 13;8(9):2930-4.

36. Marrink SJ, De Vries AH, Mark AE. Coarse grained model for semiquantitative lipid
simulations. The Journal of Physical Chemistry B. 2004 Jan 15;108(2):750-60.

37. Marrink SJ, Risselada HJ, Yefimov S, Tieleman DP, De Vries AH. The MARTINI
force field: coarse grained model for biomolecular simulations. The Journal of Physical
Chemistry B. 2007 Jul 12;111(27):7812-24.

38. Lin J, Zhang H, Chen Z, Zheng Y. Penetration of lipid membranes by gold

nanoparticles: insights into cellular uptake, cytotoxicity, and their relationship. ACS nano.
2010 Aug 27;4(9):5421-9.

39. Lin JQ, Zhang HW, Chen Z, Zheng YG, Zhang ZQ, Ye HF. Simulation study of
aggregations of monolayer-protected gold nanoparticles in solvents. The Journal of Physical
Chemistry C. 2011 Sep 14;115(39):18991-8.

40. Torrie GM, Valleau JP. Nonphysical sampling distributions in Monte Carlo free-
energy estimation: Umbrella sampling. Journal of Computational Physics. 1977 Feb
1;23(2):187-99.

41. Kumar S, Rosenberg JM, Bouzida D, Swendsen RH, Kollman PA. The weighted
histogram analysis method for free energy calculations on biomolecules. I. The method.
Journal of computational chemistry. 1992 Oct;13(8):1011-21.

42. Hub JS, De Groot BL, Van Der Spoel D. g_wham: A Free Weighted Histogram
Analysis Implementation Including Robust Error and Autocorrelation Estimates. Journal of
Chemical Theory and Computation. 2010 Nov 16;6(12):3713-20.
 Physical Chemistry Chemical Physics Page 20 of 21
View Article Online
DOI: 10.1039/C8CP04044C

43. Hess B, Kutzner C, Van Der Spoel D, Lindahl E. GROMACS 4: algorithms for highly
efficient, load-balanced, and scalable molecular simulation. Journal of chemical theory and
computation. 2008 Mar 11;4(3):435-47.

Physical Chemistry Chemical Physics Accepted Manuscript

44. Kutzner C, Van Der Spoel D, Fechner M, Lindahl E, Schmitt UW, De Groot BL,
Grubmüller H. Speeding up parallel GROMACS on high latency networks. Journal of
computational chemistry. 2007 Sep;28(12):2075-84.

45. Pronk S, Páll S, Schulz R, Larsson P, Bjelkmar P, Apostolov R, Shirts MR, Smith JC,
Kasson PM, Van Der Spoel D, Hess B. GROMACS 4.5: a high-throughput and highly
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

parallel open source molecular simulation toolkit. Bioinformatics. 2013 Feb 13;29(7):845-54.

46. Bhattacharjee S. DLS and zeta potential–What they are and what they are not?.
Journal of Controlled Release. 2016 Aug 10;235:337-51.

47. Predota M, Machesky ML, Wesolowski DJ. Molecular origins of the zeta potential.
Langmuir. 2016 Sep 29;32(40):10189-98.

48. English NJ, Long WF. Estimation of zeta potentials of titania nanoparticles by
molecular simulation. Physica A: Statistical Mechanics and its Applications. 2009 Oct
1;388(19):4091-6.

49. Chow MK, Zukoski CF. Gold sol formation mechanisms: role of colloidal stability.
Journal of colloid and interface science. 1994 Jun 1;165(1):97-109.

50. Liu Y, Lin XM, Sun Y, Rajh T. In situ visualization of self-assembly of charged gold
nanoparticles. Journal of the American Chemical Society. 2013 Feb 27;135(10):3764-7.

51. Terrill RH, Postlethwaite TA, Chen CH, Poon CD, Terzis A, Chen A, Hutchison JE,
Clark MR, Wignall G. Monolayers in three dimensions: NMR, SAXS, thermal, and electron
hopping studies of alkanethiol stabilized gold clusters. Journal of the American Chemical
Society. 1995 Dec;117(50):12537-48.

52. Pradeep T, Mitra S, Nair AS, Mukhopadhyay R. Dynamics of alkyl chains in

monolayer-protected Au and Ag clusters and silver thiolates: A comprehensive quasielastic
neutron scattering investigation. The Journal of Physical Chemistry B. 2004 Jun
3;108(22):7012-20.
 Page 21 of 21 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C8CP04044C

53. Lin J, Zhang H, Morovati V, Dargazany R. PEGylation on mixed monolayer gold

nanoparticles: Effect of grafting density, chain length, and surface curvature. Journal of
colloid and interface science. 2017 Oct 15;504:325-33.

Physical Chemistry Chemical Physics Accepted Manuscript

54. Devi JM. Aggregation of thiol coated gold nanoparticles: A simulation study on the effect
of polymer coverage density and solvent. Computational Materials Science. 2014 Apr
15;86:174-9.

55. Frederix PW, Ulijn RV, Hunt NT, Tuttle T. Virtual screening for dipeptide
aggregation: toward predictive tools for peptide self-assembly. The Journal of Physical
Published on 17 September 2018. Downloaded on 9/21/2018 7:11:06 AM.

Chemistry Letters. 2011 Sep 6;2(19):2380-4.

56. de Jong DH, Singh G, Bennett WD, Arnarez C, Wassenaar TA, Schäfer LV, Periole
X, Tieleman DP, Marrink SJ. Improved parameters for the martini coarse-grained protein
force field. Journal of Chemical Theory and Computation. 2012 Nov 28;9(1):687-97.

57. Ong Q, Nianias N, Stellacci F. A review of molecular phase separation in binary self-
assembled monolayers of thiols on gold surfaces. EPL (Europhysics Letters). 2017 Nov
29;119(6):66001.

58. Humphrey W, Dalke A, Schulten K. VMD: visual molecular dynamics. Journal of

molecular graphics. 1996 Feb 1;14(1):33-8.