Fluid Phase Equilibria 379 (2014) 175179

Contents lists available at ScienceDirect

Fluid Phase Equilibria

journal homepage: www.elsevier.com/locate/fluid

Adsorption of He gas on the Agn nanoclusters: A molecular dynamic

study
Hamed Akbarzadeh , Mahmood Mohammadzadeh
Department of Chemistry, Faculty of Basic Sciences, Hakim Sabzevari University, Touhidshahr, 96179-76487 Sabzevar, Iran

a r t i c l e

i n f o

Article history:
Received 26 October 2013
Received in revised form 19 July 2014
Accepted 21 July 2014
Available online 28 July 2014
Keywords:
Molecular dynamics simulation
Physisorption
Ag nanocluster
Gas adsorption
Enthalpy of adsorption

a b s t r a c t
Using molecular dynamics simulations, the adsorption of He gas on the Ag nanoclusters were investigated
as a function of pressure, temperature and diameter of nanoclusters. We have calculated the average
interaction energy between the gas atoms and the Ag nanoclusters, adsorption constant and enthalpy of
adsorption for all of the Ag nanoclusters. Results show that the adsorption constant, the value of average
interaction energy and the enthalpy of adsorption decrease when cluster size increases.
2014 Elsevier B.V. All rights reserved.

1. Introduction
In recent years, metallic nanoclusters have attracted a large portion of research interest due to their different physical and chemical
properties in comparison with macroscopic metals [1,2]. These differences in most cases are related to the large surface/volume ratio
in nanoscale. Also, another important feature of nanoclusters is that
their properties are usually size and shape-dependent [3]. As it has
been known before, the intrinsic properties of metallic nanoclusters depend on their size, shape, composition, crystallinity, and
structure, considerably [46]. Therefore, study on the effect of mentioned parameters on the chemical activity of metallic nanoclusters
is a running and challenging area of computational chemistry. Ag
nanoclusters are one of the important metallic clusters which have
several applications such as in catalysis [6], electrocatalysis [7],
nano optics [3,8], solar cells [9], molecular sieve membranes [10],
and antibacterial activities [1115] which can be produced by various methods [2,1620].
One of the important factors that can affect the shape of metallic
nanoclusters, is the inuence of gas atmosphere around the cluster
which usually leads to adsorption of gases on their surface. These
adsorption phenomena can be chemical or physical which depends
on the chemical activity of gas particles and clusters. Hansen et al.

Corresponding author. Tel.: +0098 571 4003323/+98 915 3008670;

fax: +0098 571 4003323.
E-mail address: akbarzadehhamed@yahoo.com (H. Akbarzadeh).
http://dx.doi.org/10.1016/j.uid.2014.07.028
0378-3812/ 2014 Elsevier B.V. All rights reserved.

studied the effect of gas atmosphere on the shape of Cu nanocrystals

dispersed on ZnO and silica supports using transmission electron
microscopy (TEM) in H2 O/H2 and CO/H2 gas mixtures [21]. They
observed considerable distortions in the cluster shape by changing
the gas atmosphere. Lamas et al. performed a MD simulation to
investigate the effect of inert gas adsorbates (Ar, Xe, and He) on
the shape and structure of Pt nanoclusters supported on a graphite
substrate. They found that the gas phase substantially alters the
vacuum cluster structure, and the changes are mostly irreversible in
the time frame of the simulations (2 ns), especially at temperatures
well below the cluster melting point [22].
Also, some MD simulations have been performed for Ag clusters. Baletto et al. used MD simulations to study the growth of
free silver nanoclusters [23]. They found that the nal shape of
the cluster depends on deposition ux and temperature. Also, they
reported a reentrant morphology transition at intermediate values
of temperature and deposition ux in which a decahedral window is formed and around it icosahedrals preferentially grow. In
another work, Baletto et al. studied freezing of the Ag nanodroplets
in time scales by MD according to experimental conditions [24].
Their calculations show that at small sizes (23 nm), both crystalline and non-crystalline clusters (icosahedral and decahedral)
are created, while at large sizes silver droplets solidify preferentially as non-icosahedral clusters. Ao et al. used MD simulations
based on SuttonChen many-body potentials to study the size
dependency of melting point (Tm ) and Kauzmann temperature (TK )
of Ag nanoclusters [25]. Their simulations indicate that as the size of
Ag particles decreases, TK and Tm functions reduce. However, the TK

176

H. Akbarzadeh, M. Mohammadzadeh / Fluid Phase Equilibria 379 (2014) 175179

and Tm ratio of Ag nanoparticles are size independent. Kuzmin et al.

studied the behavior of Ag clusters with magic numbers of atoms
in the 01300 K temperature range with embedded atom model by
MD method [26]. Their work exhibits structural transitions of cubic
octahedral clusters to the stable icosahedral structures. Luo et al.
calculated the Gibbs free energy (G) of Ag nanoparticles from the
bulk free energy and surface free energy of both solid and liquid
phases using MD simulations based on the embedded atom model
[27]. Their calculations indicate that thermodynamic properties
of mentioned nanoparticles can be divided into two parts: bulk
quantity and surface quantity; and thermodynamic properties of
nanoparticles depends on the number of surface atoms. However,
in despite of mentioned investigations, size dependence of adsorption properties of inert gaseous atmosphere on Ag nanoclusters has
not been investigated clearly.
In this work, we have investigated adsorption of He atoms on the
Ag nanoclusters of different sizes at some temperatures and pressures using MD simulations. The inert gas atmosphere was selected
due to the fact that reactive gas atmosphere leads to chemisorption
process. This process includes charge transfer between adsorbates
and the cluster and it inuences the structure and chemical activity
of nanoparticles. It would be more complex if we had used reactive
gases.
2. Calculation methods
Ag is a metal with a FCC structure. Spherical Ag nanoclusters were constructed from this large FCC structure, using various
spherical cutoff radii [28]. The cluster cutoff radius is dened as

Rc = Rg

5
+ Ri
3

(1)

and
where Ri is the interatomic radius for the Ag cluster (Ri = 1.44 A)
the radius of gyration Rg is given by
Rg2 =

2
1 
Rj Rc.m.
N

(2)

where (Rj Rc.m. ) is the distance between j atom and the cluster
center of mass. We constructed Ag nanoclusters ranging from 138
to 1057 atoms to take into account a variety of size ranges in the
mesosclae region [28,29].
The spherical Ag nanoclusters were initially subjected to mild
annealing in the 0300 K intervals. The shape of Ag nanoclusters
remained almost spherical after the annealing process.
The interactions between metal atoms were modeled by quantum SuttonChen potential (QSC) [30,31]:

Utot

1 
=
2
i=
/ j

 n
a
rij

1/2
i

(3)

The rst term is pair repulsion potential and the second term
describes the cohesion associated with the local density i :
i =

  a m
j=
/ i

rij

(4)

In Eq. (3), rij is the separation distance between atoms i and j, is

the potential well-depthwith dimensions of energy, a is the lattice
constant with the dimensions of length, c is a dimensionless parameter scaling the attractive terms, and m and n are positive integer
values with n > m. For AgAg interactions, the QSC parameters are
n = 7, m = 6 and c = 16.399.
= 0.033147 eV, a = 4.05 A,
In these simulations, we considered two different force eld
models for the He gas:

Table 1
The LennardJones 126 parameters for HeHe and AgHe interactions.
interaction

(eV)

 ()

AgHe
HeHe

0.017613
0.000900

2.597
2.550

(1) The LennardJones (LJ) 126 potential was employed for the
HeHe.
(2) The FeynmanHibbs (FH) potential was applied for HeHe
interactions. The HF potential for He can be found in study of
Tchouar et al. [32]
We compared the results obtained from the two models. There
was no signicant difference in results between the models. Therefore, the quantum effects are not important for He in these
simulations. In this paper, we have reported the results obtained
from the rst model.
Also, the LennardJones (LJ) 126 potential was employed for
the AgHe interactions [33,34] (see Table 1). The LJ parameters
for AgHe atoms were obtained by the LorentzBerthelot mixing
rules (the LennardJones parameters for AgAg are = 0.3447 eV
and  = 2.644 A [33]).
Constant temperature and constant volume MD simulations
were carried out by using DLPOLY 2.20 [35,36]. To keep temperature constant, the Berendsen thermostat [37] with a relaxation
time of 0.1 ps was applied. In these simulations Verlets leapfrog
algorithm [38] was used in conjunction with the multiple time
step method to integrate Newtons equation of motion over time.
In each simulation we considered a simulation box, which had periodic boundaries in three directions. The size of the box was chosen
such that the interaction between the particle and its own image in
the neighboring box is negligible. For all simulations a time step of
0.5 fs was used and statistical data were collected every 1000 time
step. After an equilibration period of 500 ps, data production runs of
1 ns were conducted for all the simulations. A cutoff radius of 10.0 A
was applied to the nonbonding interactions. Simulations were performed under different isothermal conditions (150350 K) and for
pressures up to 500 kbar for different sizes of Ag nanoclusters.
MD simulations were carried out in a cubic box for a system
including the annealed spherical Ag nanoclusters and He gas (He
atoms were located randomly in the cell). Various gas pressures at
constant volume and temperature were created by changing the
number of gas atoms. So the pressure was reduced in several steps.
At each step, gas atoms were removed, and after the end of each
simulation, its output was used as the input for the next step. In
these simulations, pressure was calculated by van der Waals (VDW)
equation of state for He. Fig. 1 gives snapshots of a Ag457 nanocluster
surrounded by He atoms at 150 K. This gure shows that He gas
adsorption does not change the shape of the Ag nanocluster surface.
Also, He atoms diffuse freely on the Ag nanocluster surface and do
not form clusters on the surface. Also, He atoms do not prefer to
be adsorbed on certain sites of the surface. The distance between
Helium atoms is almost much. Therefore, He atoms are gas under
the conditions of study.
3. Results and discussions
The adsorption constant represents the ratio of the number of
gas atoms within the rst layer and the surface area of the Ag nanoclusters. These gas atoms are located in the rst layer and are
directly adsorbed by the surface of Ag nanoclusters. For counting
the number of adsorbed gas atoms, we used the AgHe radial distribution function (RDF). The position of the rst peak in RDF was
applied as a criterion to locate the adsorbed gas atoms position.

H. Akbarzadeh, M. Mohammadzadeh / Fluid Phase Equilibria 379 (2014) 175179

177

14

0.07

12

0.06

10

0.05

Adsorption constant

g (r)

Fig. 1. Snapshots of a Ag457 nanocluster in He gas at 150 K (Ag, blue; He, gray) (For interpretation of the references to color in this gure legend, the reader is referred to the
web version of this article.).

8
6
4

0.04
0.03
0.02
0.01

2
0

T= 150 K
T= 250 K
T= 350 K

0.00

200

The rst peak in AgHe RDF represents the rst layer of adsorbed
gases that were located around the Ag naonocluster. Therefore, the
number of adsorbed gas atoms was calculated by:

r1
r 2 g(r)dr

n1 (r) = 4

(5)

Here n1 (r) is the number of adsorbed gas atoms in the rst layer,
g(r) is the radial distribution function, r1 is the rst minimum in the
RDF (the position of the rst peak in RDF) and  is the gas phase
density. Fig. 2 shows the radial distribution function (RDF) of Ag138
nanoclusterHe at T = 150 K and P = 1 kbar. To determine the area
of the nanocluster, the procedure used is similar to that introduced
in Ref. [39] (Because the Ag nanoclusters are almost spherical).
In Fig. 3, we have plotted adsorption constant as a function of
cluster size for Ag nanoclusters for different isotherms at P = 1 kbar.
The dimension of adsorption constant is He/2 . This gure shows
a decrease in the adsorption constant when cluster size increases.
As cluster size decreases, a larger fraction of atoms are on the surface of the cluster. Because surface atoms have less binding energy,
compared to the bulk atoms, therefore with decrease in number
of particles, cohesive energy is reduced and a more unstable structure results. So, these surface atoms have a greater tendency to
adsorb gas atoms to increase their cohesive energy. Consequently,
a decrease in the adsorption constant with increase of nanocluster size should be expected which is in accordance with Fig. 3.
Also this gure shows that the adsorption constant decreases with
temperature for each size of Ag nanoclusters. Because the average number of He atoms on the Ag nanocluster and in the gas
phase are the same, there is a dynamic equilibrium between them.

800

1000

1200

Fig. 3. Size dependence of the adsorption constant (He/2 ) for different isotherms
at P = 1 kbar.

Of course, individual He atoms may detach from or adsorb to the

Ag nanocluster, but the average number remains the same. Also,
because the gas phase is very large, the pressure reduction is very
small as a result of adsorption of some He atoms onto the Ag
nanocluster.
Fig. 4 shows the constant adsorption of Ag nanoclusters of various sizes at 150 K versus pressure. This gure shows that Fig. 3 at the
pressure of 1 kbar is far from monolayer coverage. This gure illustrates that the adsorption increases with pressure and gradually
approaches to a constant value, due to the nanocluster surface saturated by a layer of gas atoms. Also, it can be known that the pressure
0.20
0.18

Adsorption constant

Fig. 2. Radial distribution function (RDF) of Ag138 nanoclusterHe at T = 150 K and

P = 1 kbar.

600

Number of Ag nanocluster atoms

r(A)

400

0.16

Ag 138

0.14

Ag 457

Ag 263
Ag 735

0.12

Ag 1057

0.10
0.08
0.06
0.04
0.02
0.00

100

200

300

400

500

Pressure (kbar)
Fig. 4. Pressure dependence of the constant adsorption for given Ag nanoclusters
at 150 K.

178

H. Akbarzadeh, M. Mohammadzadeh / Fluid Phase Equilibria 379 (2014) 175179

Ag1057
Ag735
Ag457
Ag263
Ag138

13.0

-0.2

12.0
11.0

-0.3

ln (P/P 0)

Average interaction energy ( 10-3 kJ/mol)

-0.1

-0.4

10.0
9.0
8.0
7.0

T= 150 K
T= 250 K
T= 350 K

-0.5

6.0
0.003

-0.6
-0.7

0.004

0.005

0.006

0.007

1/T

200

400

600

800

1000

Fig. 6. Logarithm of the P/P0 as function of the inverse of temperature.

1200

Number of atoms nanocluster (Ag)

Fig. 5. Size dependence of the average interaction energy for different isotherms at
P = 1 kbar.

dependence of the adsorption constant for the smaller nanoclusters

is greater. Consequently, an adsorbent containing smaller granules
can adsorb signicantly greater amounts of adsorbates.
Fig. 5 shows the variation in the average interaction energy
with the number of Ag nanocluster atoms at P = 1 kbar. The average interaction energy (Eaverage ) is the ratio of the intermolecular
energy between the adsorbed He atoms and the surface atoms of
the Ag nanoclusters (EHeAg ) and the total number of Ag nanocluster
surface atoms (Nsurf ):
Eaverage =

EHeAg
Nsurf

Enthalpy of adsorption (kJ/mol)

0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0

(6)

(7)

where Esystem is the calculated energy of the given geometry containing the Ag nanocluster and the He atoms; Enanocluster is the
energy of the Ag nanocluster; and EHe is the energy of the He atoms.
We have performed several separate MD simulations in order to
calculate the energy of the isolated Ag nanocluster (Enanocluster ) and
the energy of the He atoms (EHe ).
There are two types of atoms in the Ag nanoclusters:
(1) Bulk atoms (interior atoms): their coordination number is
approximately 12.
(2) Surface atoms: their coordination number is less than bulk
atoms.
Therefore, the total number of Ag nanocluster surface atoms
(Nsurf ) is calculated from counting the number of atoms that have
a lower coordination number than bulk atoms. Consequently, the
value of the average interaction energy represents the strength
of the interaction energy between surface of the Ag nanocluster
and the He gases around it. As this gure shows, the value of the
average interaction energy decreases with increasing nanocluster
size. As the results show, increasing the temperature decreases
the average interaction energy. This is fundamentally consistent
with the fact that higher temperatures give the adsorbates more
kinetic energy and this, in turn, reduces the adsorbateadsorbent
attraction energy. When the gas atoms are physisorbed on the nanocluster surface, increasing the temperature makes the adsorbed
system unstable and therefore decreases the average interaction
energy.

200

400

600

800

1000

1200

Number of atoms nanocluster (Ag)

The intermolecular energy between the adsorbed He atoms and

the surface atoms of the Ag nanoclusters (EHeAg ) is dened by the
following equation:
EHeAg = Esystem (Enanocluster + EHe )

Fig. 7. Enthalpy of adsorption as a function of nanocluster size.

Eq. (8) represents the relation of the logarithm of p/p0 with the
1/T.
ln

p
p0

=
coverage

ads H
 S
ads
RT
R

(8)

We have calculated the logarithm of p/p0 versus the 1/T for

adsorption isotherms of gas. So, at each isotherm, the pressure was
dened and plotted in Fig. 6 [40,41]. Then, the enthalpy of adsorption was calculated from the slope of the curves. Coverage is the
ratio of number of adsorbed gas atoms in the rst layer and the
total number of Ag nanocluster surface atoms whose coordination
is lower than that of the bulk FCC structure.
Fig. 7 shows the enthalpy of adsorption of He gas for all nanoclusters in 0.3 coverage. It shows a trend toward higher enthalpy
of adsorption at the lower cluster size limit. The more the enthalpy
adsorption, the more favorable the adsorption process, thus, we
assume that the highest enthalpy of adsorption represents the thermodynamically most favorable gas adsorption. As the size of the
nanocluster decreases, a larger fraction of atoms are on the surface of the nanocluster. Because surface atoms have less cohesive
energy, therefore with decrease in number of particles, cohesive
energy is reduced and a more unstable structure results. Therefore,
the more number of He atoms will get the chance to be adsorbed
and so the enthalpy of adsorption will get to higher values.
4. Conclusion
This paper investigated He adsorption on Ag nanoclusters ranging from 138 to 1057 Ag atoms by molecular dynamics simulations.

H. Akbarzadeh, M. Mohammadzadeh / Fluid Phase Equilibria 379 (2014) 175179

The many body quantum SuttonChen (QSC) potential was used to

describe the AgAg interactions, also, the LennardJones (LJ) 126
potential was applied for remainder interactions (such as AgHe,
HeHe). It was found that the adsorption constant decreases with
Ag nanocluster size. As cluster size decreases, a larger fraction of
atoms are on the surface of the cluster. Because surface atoms
have less binding energy, cohesive energy is reduced with decrease
in number of particles. Consequently, these surface atoms have a
greater tendency to adsorb gas atoms to increase their cohesive
energy. The average interaction energy was calculated for each
size of Ag nanoclusters. Results show that the value of the average
interaction energy decreases with increasing nanocluster size. Also,
the enthalpy of adsorption was computed for all of the nanoclusters in 0.3 coverage. Results show that the enthalpy of adsorption
increases with decreasing cluster size. Consequently, Ag138 have
the thermodynamically most favorable gas adsorption.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]

F. Baletto, F. Riccardo, Rev. Mod. Phys. 77 (2005) 371.

Q. Zhang, J. Xie, Y. Yu, J.Y. Lee, Nanoscale 2 (2010) 1962.
P.C. Ray, Chem. Rev. 110 (2010) 5332.
Z.L. Wang, J. Phys. Chem. B 104 (2000) 1153.
Y. Sun, Y. Xia, Science 298 (2002) 2176.
R. Xu, D. Wang, J. Zhang, Y. Li, Chem. Asian J. 1 (2006) 888.
V. Bansel, V. Li, A.P. OMullane, S.K. Bhargava, Cryst. Eng. Commun. 12 (2010)
4280.
S. Lal, S. Link, N.J. Halas, Nat. Photonics 1 (2007) 641.
B.P. Rand, P. Peumans, S.R. Forrest, J. Chem. Phys. 96 (2004) 7519.
J.N. Barsema, J. Balster, V. Jordan, N.F.A. Van der Vegt, M. Wessling, J. Membr.
Sci. 219 (2003) 47.
J.S. Kim, E. Kuk, K.N. Yu, J.H. Kim, S.J. Park, H.J. Lee, S.H. Kim, Y.K. Park, Y.H. Park,
C.Y. Hwang, Y.K. Kim, Y.S. Lee, D.H. Jeong, M.H. Cho, Nanomed.: Nanotechnol.
Biol. Med. 3 (2007) 95.

179

[12] S. Pal, Y.K. Tak, J.M. Song, Appl. Environ. Microbiol. 73 (2007) 1712.
[13] O. Choi, K.K. Deng, N.J. Kim, L. Ross, R.Y. Surampalli, Z. Hu, Water Res. 42 (2008)
3074.
[14] M. Rai, A. Yadav, A. Gade, Biotechnol. Adv. 27 (2009) 76.
[15] Z.M. Xiu, Q.B. Zhang, H.L. Puppala, V.L. Colvin, P.J.J. Alvarez, Nano Lett. 12 (2012)
4271.
[16] Q. Zhang, J. Xie, J. Yang, J.Y. Lee, ACS Nano 3 (2009) 139.
[17] H. Xu, K.S. Suslick, Adv. Mater. 22 (2010) 1078.
[18] H. Xu, K.S. Suslick, ACS Nano 4 (2010) 3209.
[19] J. Zeng, Y. Zheng, M. Rycenga, J. Tao, Z.Y. Li, Q. Zhang, Y. Zhu, Y. Xia, J. Am. Chem.
Soc. 132 (2010) 8552.
[20] Q. Zhang, N. Li, J. Goebl, Z. Lu, Y. Yin, J. Am. Chem. Soc. 133 (2011) 18931.
[21] P.L. Hansen, J.B. Wagner, S. Helveg, J.R. Rostrup-Nielsen, B.S. Clausen, H. Topse,
Science 295 (2002) 2053.
[22] E.J. Lamas, P.B. Balbuena, J. Phys. Chem. B 107 (2003) 11682.
[23] F. Baletto, C. Mottet, R. Ferrando, Phys. Rev. Lett. 84 (2000) 5544.
[24] F. Baletto, C. Mottet, R. Ferrando, Chem. Phys. Lett. 354 (2002) 82.
[25] Z.M. Ao, W.T. Zheng, Q. Jiang, Nanotechnology 18 (2007) 255706.
[26] V.I. Kuzmin, D.L. Tytik, D.K. Belashchenko, A.N. Sirenko, Colloid J. 70 (2008) 284.
[27] W. Luo, W. Hu, S. Xiao, J. Phys. Chem. C 112 (2008) 2359.
[28] K.R.S.S. Sankaranarayanan, V.R. Bhethanabotla, B. Joseph, Phys. Rev. B: Condens.
Matter 71 (2005) 195415.
[29] Y. Qi, T. Cagin, W.L. Johnson, W.A. Goddard III, J. Chem. Phys. 115 (2001) 385.
[30] A.P. Sutton, J. Chen, Philos. Mag. Lett. 61 (1990) 139.
[31] Y. Qi, T. Cagin, Y. Kimura, W.A. Goddard III, Phys. Rev. B: Condens. Matter 59
(1999) 3527.
[32] N. Tchouar, F. Ould-Kaddour, D. Levesque, J. Chem. Phys. 121 (2004) 7326.
[33] M. Neek-Amal, R. Asgari, M.R. Rahimi Tabar, Nanotechnology 20 (2009)
135602.
[34] R.C. Reid, J.M. Prausnitz, B.E. Poling, The Properties of Gases and Liquids, fourth
ed., McGraw-Hill, Boston, MA, 1987.
[35] W. Smith, T.R. Forester, J. Mol. Graphics 14 (1996) 136.
[36] W. Smith, I.T. Todorov, Mol. Simul. 32 (2006) 935.
[37] H.J.C. Berendsen, J.P.M. Postma, W.F. Gunsteren, A. Dinola, J.R. Haak, J. Chem.
Phys. 81 (1984) 3684.
[38] M.P. Allen, D.J. Tildesley, Computer Simulation of Liquid, Clarendon, Oxford,
1997.
[39] W.H. Qi, M.P. Wang, J. Mater. Sci. Lett. 21 (2002) 1743.
[40] J.M. Simon, O.E. Haas, S. Kjelstrup, J. Phys. Chem. C 114 (2010) 10212.
[41] D.C. Meier, D.W. Goodman, J. Am. Chem. Soc. 126 (2004) 1892.