Effect of temperature on the fracture strength of perfect and

defective mono layered graphene

A Synopsis Submitted in
Partial Fulfillment of the Requirements for the Degree of
Master of Technology

Kritesh Kumar Gupta

Registration No. 17-22-306

Under the Supervision of

Dr. Sudip Dey

Department of Mechanical Engineering

National Institute of Technology Silchar
2018
1. Introduction
The extraordinary characteristics of graphene have been used to develop nanoresonators [1],
biosensors [2], nanocomposites [3] and many more nanoelectromechanical systems (NEMS).
So far research has been conducted in order to determine the exact mechanical properties and
strength of monolayer graphene sheet. Graphene is found to be one of the lightest (0.77
mgm-2) and still stiffest material known. Its fracture strength and Young modulus are
recorded as 130 GPa and 1 TPa, respectively. Finding the mechanical properties of graphene
through experimental procedures are limited due to the fact that setting up the
experimentation for the nanoscale material has its own complications and limitations.
Experimentally one of the common methods to extract graphene from graphite is CVD
(Chemical Vapor Decomposition), which is performed at elevated temperatures, it is obvious
that during extraction some defects may be introduced to the graphene structure by vacancy,
chemical reactions with foreign elements or by a change in the bond structure at elevated
temperatures. This leads to getting incorrect desired results from the experimentation. The
defect such as vacancies remarkably reduces the strength of graphene, which in return affects
the performance of the NEMS.
In past, researchers studied on properties of graphene. Tsai et. al. [4] investigated the elastic
properties of the graphene along with graphite flakes through molecular dynamics simulation
and concluded that the single graphene layer has the higher value of Young’s modulus of
elasticity and shear modulus in comparison with graphite flakes.
Scarpa et al. [5] calculated the elastic moduli of SLGS using analytical models of truss and
the theory of cellular material mechanics while Thomas et al. [6] performed the large
atomistic simulation using MD simulation and Tersoff Interatomic potential and determined
the mechanical, thermal and structural properties of the single-layered graphene. In contrast,
Ansari et al. [7] performed the atomistic simulation using MD and Tersoff-Brenner potential
energy function to determine the mechanical properties of perfect and defective single-
layered graphene while Dewapriya and Rajapakse [8] investigated the temperature and strain
rate dependent fracture strength of defective graphene using molecular dynamics and an
atomistic model. The strength of graphene highly depends on vacancy concentration,
temperature, and strain rate. Molecular dynamics simulations, which are generally performed
under high strain rates, exceedingly over-predict the strength of graphene at elevated
temperatures. Molecular dynamics simulations on the crack propagation reveal that the
energy dissipation rate indicates proportionality with the strength. These findings provide a
remarkable insight into the fracture strength of defective graphene, which is critical in
designing experimental and instrumental applications. Fenley et al. [9] developed the
software to calculate the atomistic mechanical stress to be used in the MD simulations of
biomolecules and nanomaterials using AMBER interatomic potential function. Clavier et
al.[10] computed the elastic stiffness tensor at 1 bar and 60K of FCC argon through
molecular dynamics simulation and compared the result obtained by using NPT and NVT
ensembles of MD. Papageorgio et al.[11] reviewed the mechanical properties of graphene
and described the strategies to prepare the bulk graphene-based nanocomposites. It was
concluded from the study that graphene as reinforcement in composites may improve the
strength of the composite up to the next level. Carpio et al. [12] studied the stability and
evolution of various elastic defects in a flat graphene sheet and the electronic properties of
the most stable configurations. Unlike the most studied case of carbon nanotubes, Stone-

2
 Wales defects are unstable in the planar graphene sheet. Similar defects in which one of the
pentagon-heptagon pairs is displaced vertically with respect to the other one are found to be
dynamically stable. Singh et al.[13] predicted the elastic properties of the graphene sheet by
combining MM3 inter-atomic potential with the Cauchy-Born rule at infinitesimal and finite
strains. The elastic properties calculated using the given approach were somewhat better as
compared to those predicted through Tersoff-Brenner and second generation REBO
potentials. Shao et al.[14] analyzed the temperature dependent ultimate strength, thermal
expansion coefficients, lattice geometries, and elastic constants of graphyne and graphene by
combining first-principles calculation with a quasi-harmonic approximation. Zaeri et al. [15]
carried out molecular dynamics simulations to predict the elastic constants, i.e. the elements
of the stiffness tensor, and the elastic moduli, namely the Young’s and shear moduli, of
various single-walled carbon nanotubes. Poisson’s ratios were also calculated.

2. Problem Statement
In practice, it is possible that significant uncertainties may arise from the defects in the
monolayer graphene during the production phase. To accommodate these uncertainties in the
modeling of the monolayered graphene sheet and observe the effects of temperature variation
on the fracture strength of the same. In order to exploit the full potential of graphene, it is
essential to explore the thorough understanding of the mechanics of graphene.

3. Objectives
To gain an understanding that how graphene performs under varying temperature and how
the vacancy defects influence the strength of the graphene.

4. Materials and Methods

4.1. Modeling and Simulation

The graphene sheet of 5nm X 5nm with 1008 carbon atoms is modeled using VMD (Visual
Molecular Dynamics), the output file received from VMD is fed to Avogadro (an open
source atom manipulator) in order to introduce a 2.5% vacancy in the perfect graphene sheet.
Both the sheets (perfect and defected) were uniaxially deformed in x-direction using
LAMMPS which is an acronym of Large-scale Atomic/Molecular Massively Parallel
Simulator. It is an open source package by Sandia National Laboratories USA. To avoid free
edges, all the three directions was applied with periodic boundary conditions. The sheets
were firstly equilibrated to minimize the energy of the system, over 30 ps before applying
strain using time-step 0.5fs. In the relaxation period, the pressure component of the x and y
directions was maintained zero using NPT ensemble. The NPT ensemble holds the no. of
atoms, pressure, and temperature of the system constant, it uses Nose-Hoover thermostat to
control the temperature of the system. Furthermore, to uniaxially deform the sheets a strain
was imposed by elongation of the simulation box in both (armchair and zigzag) dimension

3
and the applied strain rate was 109s-1. The stress components in y and z directions were kept
at zero during simulating the tensile deformation of graphene by keeping in check the
pressure component in the said directions using NPT ensemble. Figure 1 shows the topology
of the perfect graphene sheet and graphene introduced to 2.5% vacancy concentration.

(a) (b)
Fig. 1. A graphene sheet (a) perfect, (b) with 2.5% vacancy concentration

4.2. Potential Field

In molecular dynamics simulation, potential field is an element which signifies the

energy interactions between the atoms. The adaptive intermolecular reactive empirical
bond order (AIREBO) potential is employed in the present study. The AIREBO potential
is made by combining three different potential fields. First, the original REBO potential,
second, the Lenard-Jones (LJ) and third, the torsion potential. The original REBO term
imparts the inter-atomic attractive and repulsive energy, the Lenard-Jones to account for
the intermolecular interaction that includes dispersion and short-range repulsion and
torsion component to signify the torsional interactions between single bonds. The general
form to represent the AIREBO potential can be expressed as [17],

1
E [ EijREBO  EijLJ  
2 i 1 i  j

k i l i , j , k
TORSION
Eijkl ] (1)

where E refers the total potential energy with atoms i, j, k, and l. The REBO term which
accounts for interatomic attraction and repulsion is represented as:

4
 EijREBO  VijR (rij )  bijVijA (rij ) (2)

where VA and VR refer the attractive and repulsive parts of the potential respectively; rij
signifies the distance between i and j atom up to the cutoff distance and bij refers the bond
order which allows for a change in the strength of the bond depending on the
environment.

The LJ term of the potential can be described as:

EijLJ  f (rij )VijLJ (3)

12 6
  ij    ij 
VijLJ
 4 ij [    (4)
 r   r 
 ij   ij 

where f(rij) represents the set of cutoff functions and  ij and  ij are the distance at which
interparticle potential is zero and the potential barrier respectively.

Finally, the torsion term can be represented as:

TORSION
Eijkl  f (rij , rjk , rkl )V TORSION (wijkl ) (5)

 256  w 1 
V TORSION ( wijkl )    cos10     (6)
 405  2  10 

where rij, rjk, and rkl are the bond lengths, wijkl is the dihedral angle and  is the barrier
height of the potential.

4.3. Calculation of Stress

In the molecular dynamic simulation, stress may be interpreted as Cauchy stress or virial
stress. When it comes to the computational efficiency the Cauchy stress is proved to be
more considerable than virial stress. Although, some amount of initial stress (when strain

5
 is zero) is induced by the Cauchy stress at higher temperatures. In this study, the
computation of virial stress is carried out through atomistic simulation. The virial stress
can be defined as [16]:

1 
  R  R  F  m v v 
N
1
 ij    
i

i
  
j i j (7)
V  2 1 
where  and  refer to the number assigned to an atom and the number assigned to the
neighboring atom, respectively; while, i and j signify the directional indices, m and vi
are the mass and velocity of the atom  , respectively. Ri is the location of an atom  in
the i direction; Fj is the force due to an atom  on atom  along the j direction, V
refers the total volume of the material system. The virial stress represents a product of
stress and volume i.e., the stress value computed by LAMMPS would be in the units of
pressure*volume, in order to extract the stress from the computed value it is essential to
divide the value calculated by the LAMMPS with the total volume. Furthermore, in order
to calculate the virial stress, the instantaneous volume is used i.e. when the imposed
strain is ‘  ’, the value of volume at the given instant would be V0(1+  ), where V0
represents the initial volume of the system.

5. Results and Discussion

5.1. Effect of temperature variation
The stress-strain behavior of perfect graphene and defected graphene at different
temperatures i.e. 1K, 300K, 900K and 1200K is recorded in both armchair and zigzag
directions and is shown in Fig. 2 and Fig. 3 respectively. It is evident by the recorded
values that increasing temperature drastically reduces the fracture strength of the
graphene in both the directions and failure occurs at fewer strain values. However, it is
observed that the graphene shows more resistance to failure in the zigzag direction as
compared to its armchair direction. It can also be seen that on uniaxially loading
graphene in both the directions (armchair and zigzag) at 1K, it shows the sharp yielding
point before fracture.
In both the cases i.e., loading in an armchair and zigzag direction, the defected graphene
does not show the instant fracture or it can be said that on introducing the 2.5% vacancy
to the graphene sheet it shows the ductile fracture instead of brittle fracture. However,
after some strain, it starts showing the yielding.

6
 (a) (b)

(c) (d)
Fig. 2. A stress-strain behavior of perfect and defective graphene loaded in an armchair direction (a) at 1K, (b) at
300K, (c) at 900K, (d) at 1200K. Defective sheets have 2.5% of randomly distributed vacancies.

(a) (b)

(c) (d)

Fig.3. A stress-strain behavior of perfect and defective graphene loaded in a zigzag direction (a) at 1K, (b) at 300K,
(c) at 900K, (d) at 1200K. Defective sheets have 2.5% of randomly distributed vacancies.

7
 The given Table 1 and Table 2 represent the recorded values of Young’s modulus (E),
fracture strength and failure strain of perfect graphene sheet at various temperatures in an
armchair and zigzag directions respectively.

Table 1. Young’s modulus (E), fracture strength and failure strain of perfect graphene sheet
loaded in an armchair direction due to variation in temperature.

E (in Fracture strength (in Failure

Temperature
TPa) GPa) strain
1K 0.814 146.9 0.293
300K 0.812 145.4 0.236
900K 0.809 135.7 0.217
1200K 0.799 126.6 0.201

Table 2. Young’s modulus (E), fracture strength and failure strain of perfect graphene sheet
loaded in a zigzag direction due to variation in temperature.

E (in Fracture strength (in Failure

Temperature
TPa) GPa) strain
1K 0.745 246.3 0.407
300K 0.756 261.9 0.411
900K 0.764 211.4 0.370
1200K 0.760 200.9 0.357

6. Conclusions
In summary, an MD simulation of uniaxial loading the perfect graphene sheet and
defective graphene sheet (2.5% vacancy concentration) is carried out at various
temperatures (1K, 300K, 900K,1200) in both armchair and zigzag directions. The
AIREBO potential field is used to incorporate the interatomic interactions between the
carbon atoms and the Nose-Hoover thermostat is used to control the temperature of the
system by using the NPT ensemble. The fracture strength of perfect and defective
graphene sheet is observed at various temperatures and it is concluded that the
temperature has a profound impact over the fracture strength of the graphene sheet. It’s
been pointed above that how sometimes it is obvious that the graphene sheet produced
may have some defects, inherited by the way it is produced. The stress-strain behavior of
perfect and defective graphene sheet is recorded at various temperatures by loading the
graphene in both the directions, and it is concluded that graphene sheet posses higher
fracture strength in the zigzag direction as compared to its armchair direction.
Furthermore, it is also been observed that the defective graphene sheet does not depict the

8
 brittle fracture. Although, it shows yielding at very low strain values than perfect
graphene sheet. The findings of the study provide the insights of variability in the fracture
strength of graphene with respect to increasing temperature and with vacancy defects,
using which the usability of graphene could be decided for the applications at elevated
temperatures and where defects are unavoidable.

7. References
1. Dai, M.D., Kim, C.W., Eom, K., Nonlinear vibration behavior of graphene resonators and
their applications in sensitive mass detection, Nanoscale Research Letters: a Springer
open journal, 7(1), (2012).
2. Kuila, T., Bose, S., Khanra, P., Mishra, A.K., Kim, N.H., Lee, J.H., Recent advances in
graphene based biosensors, Biosensors and Bioelectronics, 26, 4637– 4648, (2011).
3. Wang, G., Wang, B., Wang, X., Park, J., Dou, S., Ahn, H. and Kim, K., Sn/graphene
nanocomposite with 3D architecture for enhanced reversible lithium storage in lithium
ion batteries, Journal of Materials Chemistry, 19, 8378-8384, (2009).
4. Tsai, J.L., Tu, J.F., Characterizing mechanical properties of graphite using molecular
dynamics simulation, Materials and Design, 31, 194-199, (2010).
5. Scarpa, F., Adhikari, S., Phani, A.S., Effective elastic mechanical properties of single
layer graphene sheets, Nanotechnology, 20, (2009).
6. Thomas, S., Ajith, K.M., Molecular dynamics simulation of the thermo mechanical
properties of monolayer graphene sheet, Procedia Materials Science, 5, 489-498, (2014).
7. Ansari, R., Ajori, S., Motevalli, B., Mechanical properties of defective single layered
graphene sheets via molecular dynamics simulation, Superlattices and Microstructures,
51, 274-289, (2012).
8. Dewapriya, M.A.N., Rajapakse, R.K.N.D., Molecular dynamics simulations and
continuum modeling of temperature and strain rate dependent fracture strength of
graphene with vacancy defects, Journal of Applied Mechanics, 81, (2014).
9. Fenley, A.T., Muddana, H.S., Gilson, M.K., Calculation and visualization of atomistic
mechanical stresses in nanomaterials and biomolecules, PLoS ONE, 9(12), (2014).
10. Clavier, G., Desbiens, N., Bourasseau, E., Lachet, V., Brussel-Dupend, N., Rousseau, B.,
Computation of elastic constants of solids using molecular simulation: comparison of
constant volume and constant pressure ensemble methods, Molecular Simulation, 43,
1413-1422, (2017).
11. Papageorgio, D.G., Kinloch, I.A., Young, R.J., Mechanical properties of graphene and
graphene based nanocomposites, Progress in Material Science, 90, 75-127 (2017).
12. Carpio, A., Bonilla, L.L., di Juan, F., Vozmediano, M.A.H., Dislocations in graphene,
New Journal of Physics, 10(5), (2008).
13. Singh, S., Patel, B.P., Nonlinear elastic properties of graphene sheet using MM3 potential
under finite deformation, Composites Part B, 136, 81-91, (2018).
14. Shao, T., Wen, B., Melnik, R., Yao, S., Kawazoe, Y., Temperature dependent elastic
constants and ultimate strength of graphene and graphyne, The Journal of Chemical
Physics, 137 (2012).
15. Zaeri, M.M., Ziaei-Rad, S., Molecular dynamics investigation of the elastic constants and
moduli of single walled carbon nanotubes, J. Nanoanalysis, 4(1), 65-75, (2017).

9
16. Thompson, A. P., Plimpton, S.J., Mattson, W., General formulation of pressure and stress
tensor for arbitrary many body interaction potentials under periodic boundary conditions,
The Journal of Chemical Physics, 131, (2009).
17. Stuart, S.J., Tutein, A.B., Harrison, J.A., A reactive potential for hydrocarbons with
intermolecular interactions, Journal of Chemical Physics, 112(14), (2000).
18. Humphrey, W., Dalke, A. and Schulten, K., VMD - Visual Molecular Dynamics, J.
Molec. Graphics, vol. 14, pp. 33-38, (1996).
19. Hanwell, M.D., Curtis, D.E., Lonie, D.C., Vandermeersch, T., Zurek, E., and Hutchison,
G.R., Avogadro: An advanced semantic chemical editor, visualization, and analysis
platform, Journal of Cheminformatics, 4:17, (2012).

10