You are on page 1of 62

Role of out of plane deformations on fracture toughness of

2D graphene

B. N. K. Goutham

A Dissertation Submitted to
Indian Institute of Technology Hyderabad
In Partial Fulfillment of the Requirements for
The Degree of Master of Technology

Department of Mechanical and Aerospace Engineering

Month, 2021
Declaration

I declare that this written submission represents my ideas in my own words, and where
others’ ideas or words have been included, I have adequately cited and referenced the
original sources. I also declare that I have adhered to all principles of academic honesty and
integrity and have not misrepresented or fabricated or falsified any idea/data/fact/source
in my submission. I understand that any violation of the above will be a cause for disciplinary
action by the Institute and can also evoke penal action from the sources that have thus not
been properly cited, or from whom proper permission has not been taken when needed.

(Signature)

_________________________
(B.N.K. Goutham)

_________________________
(Roll No: ME16B20M000005)

ii
Approval Sheet
This thesis entitled Role of out of plane deformations on fracture toughness of 2D
graphene by B.N.K. Goutham is approved for the degree of Master of Technology from IIT
Hyderabad.

Prof. M. Ramji
Department of Mechanical and Aerospace Engineering
Indian Institute of Technology Hyderabad
Examiner

Dr. Syed Nizamuddin Khaderi


Associate Professor
Department of Mechanical and Aerospace Engineering
Indian Institute of Technology Hyderabad
Examiner

Dr. Viswanath Chintapenta


Assistant Professor
Department of Mechanical and Aerospace Engineering
Indian Institute of Technology Hyderabad
Adviser

Prof. S. Suriya Prakash


Professor
Department of Civil Engineering
Indian Institute of Technology Hyderabad
Chairman

iii
Acknowledgements

I would like to thank my advisor, Dr. Viswanath Chinthapenta, for his unwavering support,
motivation, and direction in helping me finish this thesis. I would sincerely thank Chaitanya
Sagar for all the help with LAMMPS, Linux and postprocessing of simulations. His experience
helped me learn things very fast. I would like to show my thankfulness to IIT Hyderabad's
Mechanical and Aerospace Department for providing me with the incredible chance to
pursue my MTech at this prestigious school. I would also like to thank my lab mates and
friends for their constant support. Finally, I would like to thank my Amma, Nanna for
everything.

iv
Dedicated to

Amamma and Mavayya.

v
Abstract

In this thesis, the fracture toughness of graphene is investigated. In-spite of several previous
studies available on the topic, there is a need to relook into the fracture properties of
graphene due to large variability reported in literature. This variability is due to several
factors. Few of them are crack angle, temperature, and out-of-plane deformations of 2D
materials. In this study, the role of crack orientation, crack angle, initial temperature and
out-of-plane deformations on the fracture toughness is investigated methodically. It was
observed that up to 20% variation in fracture toughness is seen in graphene due to out-of-
plane deformations. It was also observed that there is 14% drop in tensile strength with
change in temperature from 300K to 1000K.

vi
Nomenclature
Abbreviation Full form
AC Armchair
AFM Atomic Force Microscopy
AIREBO Adaptive Intermolecular Reactive
Empirical Bond Order
CNT Carbon Nanotubes
CVD Chemical Vapor Deposition
GEM Global Energy Method
LAMMPS Large-scale Atomic/Molecular
Massively Parallel Simulator
LEFM Linear Elastic Fracture Mechanics
LFM Local Force Method
MEMS Micro-electromechanical systems
SIF Stress Intensity Factor
SWD Stone-Wales Defect
UTS Ultimate Tensile Strength
ZZ Zigzag

vii
Symbol Parameter
m Mass
F Force
r Position
U Potential Energy
𝞇 State of System
θ Temperature
v Velocity
n Degrees of Freedom
ĸb Boltzmann’s Constant
bo Bond Order
q Charge
V Total Volume
Gc Energy Release Rate
φ Potential Energy
a Crack length
z Thickness
KIC Stress Intensity Factor
𝝈 Stress
Y Young’s Modulus

viii
Contents

Declaration ......................................................................................................................... ii
Approval Sheet .................................................................................................................. iii
Acknowledgements ........................................................................................................... iv
Abstract.............................................................................................................................. vi

Nomenclature ...................................................................................................................... viii

1 Introduction ......................................................................................................................... 1
1.1 2D Materials........................................................................................................... 1
1.2 Properties and Applications................................................................................... 2
1.3 Manufacturing of 2D Materials ............................................................................. 3
1.4 Graphene ............................................................................................................... 5
1.5 Defects in Graphene .............................................................................................. 6

2 Literature Survey ................................................................................................................. 9


2.1 Preliminary Work ................................................................................................... 9
2.2 Research Gaps ..................................................................................................... 13

3 MD Simulations and Formulation ..................................................................................... 14


3.1 Introduction to Molecular Dynamics ................................................................... 14
3.2 Characteristics of Molecular Dynamics ............................................................... 15
3.3 Interatomic Forces ............................................................................................... 20
3.4 Fracture Parameters ............................................................................................ 24

4 Results and Discussion ...................................................................................................... 28


4.1 Effect of initial crack length, orientation, and angle on tensile strength ............ 28
4.2 Effect of out-of-plane deformations on fracture toughness ............................... 34
4.3 Effect of Initial crack on out of plane deformations of the graphene sheet ....... 36
4.4 Effect of temperature on fracture parameters and out of plane deformations . 38

5 Conclusions and Future Work ........................................................................................... 43


5.1 Concluding marks and key contributions ............................................................ 43
5.2 Future Scope ........................................................................................................ 44

References ............................................................................................................................ 45

ix
Chapter 1

Introduction

1.
1.1 2D Materials
Initially, two-dimensional materials and the many sorts of 2D materials are covered in
this chapter. In Section 1.2, it was expanded to include the study of the characteristics and
uses of 2D materials. In Section 1.3, different approaches for fabricating two-dimensional
materials are discussed. Section 1.4 and 1.5, are specifically about graphene and defects
observed in graphene.
When one dimension of material is nano-sized and other dimensions are much larger,
it is called a 2D material. If two dimensions of material are restricted to nano size, then it is
1D material. The first two-dimensional material found was Graphene [1]. It is single atom
thick graphite. Graphene was discovered in 2004, after rubbing a highly oriented pyrolytic
graphite against another surface. Later, different kinds of novel two-dimensional materials
were discovered. Few examples are Molybdenum Disulphide (MoS2), Boron Nitride (BN), and
Phosphorene. There are many families of 2D materials discovered. 2D transition metal
dichalcogenides (TMDCs), 2D layered transition metal oxide (LTMOs), 2D layered halide
perovskites, 2D carbides, and nitrides (MXenes), 2D layered double hydroxides (LDHs), etc.
are few families which belong to 2D materials. TMDCs are MX2 kind of two-dimensional
materials. Here M represents a metal element, and X represents a sixth group element.
Nickel telluride (NiTe2), Molybdenum telluride (MoTe2), Tungsten Disulphide (WS2), Nickel
Disulphide (NiS2), and Molybdenum Diselenide (MoSe2) are few examples for TMDCs. TMDCs
are comprised of 3 atomic layers, a metal layer between the layers of chalcogens. TMDCs
have an exceptional electron mobility of 200 cm2V-1s-1. LTMOs are MO, MO2, MO3 type 2D
materials. Titanium dioxide (TiO2), Tungsten Trioxide (WO3), and Manganese dioxide (MnO2)
are few examples of LTMOs. O2- polarizes LTMOs, which leads to nonlinear and non-uniform
distribution of charges in their lattice. Mn+1BYn is the typical chemical formula for 2D

1
carbides. Here, M represents a transition metal atom, B represents a group–A element, Y
represents carbon or nitrogen. Ti3AlC2, Ti4N3 are few examples of 2D carbides and nitrides.

1.2 Properties and Applications of 2D materials


Because of their unique features and applicability in diverse sectors, two-dimensional
materials have piqued the curiosity of scientists. Properties like high surface area per mass,
high mechanical strength, and high electron mobility. In 2D materials, atoms are bonded
covalently in-plane and through van der Waal forces between different layers. When
compared to a covalent connection, van der Waal forces are exceedingly weak, allowing
layers to slide over each other readily even when just a little amount of shear stress is
applied. This gives rise to very good lubrication properties for 2D materials. Few single-layer
materials displayed exceptional optical characteristics due to their intrinsic features. Young's
modulus of 2D materials can reach till 1TPa, breaking stress of 130GPa. It is mainly because
of the low disorder and impurities. 2D materials have surface area per mass as high as 2600
m2/gm (3gms of a 2D material can cover a whole football ground). Surface forces are thought
to have a crucial influence in deciding the applications of 2D materials because of their huge
surface area per volume. Atomic Force Microscopy (AFM) is utilized to investigate and
examine the two-dimensional material's mechanical and tensile characteristics.
Two-dimensional materials can be used in various disciplines. Polymer composites
tend to have increased mechanical strength when graphene was added to the matrix. New
techniques were introduced to create polymer matrices blended with 2D materials for
improving tensile strength. Even with the addition of a very little portion of two-dimensional
materials (0.2 – 0.3% wt.), a drastic improvement is observed in tensile strength, breaking
strength, and Young’s modulus [2], [3]. Composites hybridized with 2D materials tend to
have higher thermal diffusivity [4]. 3D printing remains a promising application of reinforced
polymers. In the case of bone tissue engineering, polypropylene fumerate polymer, and
functionalized PEG (polyethylene glycol), are mixed, which resulted in an improvement of
mechanical strength [5]. Because of their great strength, low thickness, and great flexibility,
two-dimensional materials are suitable for electronic applications. MoS2 based transistors
exhibits electron mobility of 30cm2V-1s-1 [6]. The electron mobility of most two-dimensional
materials (Molybdenum Disulphide, Graphene, Boron Nitride, tungsten Di) was greater than
45 cm2V-1s-1. They also sustain the strain of 1.5 to 3%, without much effect on the electronic
properties [7]. Another promising application of 2D material is due to its property of

2
piezoelectricity. The trait of piezoelectricity is predicted to occur in most 2D materials [8],
[9]. It is observed due to the broken symmetry of 2D materials. Graphene is intrinsically non-
piezoelectric material due to its symmetry. Symmetry can be broken through various surface
doping techniques [10]. This property of piezoelectricity in 2D materials is utilized for sensors
and nanogenerators. Through performing biaxial loading on suspended graphene, a
nanogenerator, a piezoresistive sensor sensitive for low values of pressure can be fabricated.
Strain tensors were also manufactured from MoS2 [11]. Piezoelectricity of MoS2 helps
improve the efficiency of solar cells prepared from MoS2 [12].
Because of their low thickness, high strength, and low density, 2D materials were
easily adaptable to MEMS (Micro-electromechanical systems) devices [13]. Resistance of
graphene is sensitive to the strain applied, which can be utilized for strain and force sensors.
2D materials-based nano resonators have potential in applications, such as sensors, mass
sensors, and optoelectronic devices [14]–[16]. Graphene squeeze-film pressure sensors
show 45 times more responsivity with a surface area of at least 25 times less than previous
MEMS sensors.
In addition to these, 2D materials also have applications in gas sensors, magnetic
resonance imaging, light-emitting diodes, and biosensing.

1.3 Manufacturing of 2D Materials


Two distinct techniques may be used to create 2D materials. The first is a top-down
strategy, and the second is a bottom-up one. The top-down approach is through controlled
removal of material from 3D solids, whereas in bottom-up, the material is allowed to react
and grow from the atomic level.
In the top-down approach, thin sheets can be fabricated either through physical force
or through chemical separation. Physical force separation comprises of different techniques,
through use of mechanical force, using ultrasonic waves to separate 2D layers from bulk
solids.
Graphene can be fabricated from graphite through micromechanical exfoliation. It
was introduced by Novoselov and Geim [17] in the year 2004. The process follows a
repetitive exfoliation of 3D van der Waal solid using adhesive tape. It is a simple way of
synthesizing one atom layer nanosheets without disrupting crystal structure and properties.
Only van der Waals solids can be manufactured through this method. Because of its

3
inexpensive cost, this technique has become well-liked for manufacturing two-dimensional
materials.
One type of top-down technique for creating 2D material is ultrasonic exfoliation. It is
more effective and productive as compared to mechanical exfoliation. A bulk solid of desired
material is dispersed in an aqueous solution. It is bubbled with an inert gas to avoid
oxidation. The mixture is then sonicated. These bubbles are formed in the negative pressure
region. Ultrasonic waves travel longitudinally and form a cavitation effect, which eventually
increases pressure. This high pressure causes small explosions, which results in constant
exfoliation. After ultrasonic treatment, the resultant is centrifuged to remove the
unexfoliated component. Recently, Hui et.al [18] fabricated ultrathin black phosphorous
using ultrasonic exfoliation. Despite the various benefits of ultrasonic exfoliation, it is
difficult to obtain pure single layer material.
Lithium-intercalated and exfoliation are efficient in synthesizing ultrathin sheets than
ultrasonic exfoliation. The bulk of desired material is used as the cathode, lithium as an
anode. Electrochemical exfoliation is performed using galvanostatic discharge. During this
galvanostatic discharge, the Li+ ions intercalate in between the layers of bulk material. This
intercalation weakens the van der Waals interactions and eventually results in exfoliation.
This method is very effective in preparing single-layer metal dichalcogenide. This method is
limited to van der Waals solids and not applicable in the case of layered ionic solids. Ion-
change exfoliation is used for the preparation of layered ionic solids.
Even though the top-down approach is very useful in fabricating ultrathin nanosheets,
it is limited to materials where bulk crystals are layered [19]. Also, the sheet fabricated
through a top-down approach is limited to small size. Top-down strategies are more
applicable for initial research in a lab than in mass production. In the case of bottom-up
strategies, 2D sheets are synthesized from atomic and molecular substrates, which are
suitable to react and grow.
Wet chemical synthesis is through chemical reactions in the solution phase using
precursors. Hydrothermal synthesis and template synthesis are two different types of wet
chemical synthesis strategies. These tactics are used to manufacture 2D materials that
cannot be created via top-down methods. It is a promising method in preparing 2D materials
with high yield, mass production, and low cost.
Hydro/solvothermal synthesis is one kind of wet chemical synthesis, where desired
solid is dissolved in solvent (water, benzene methanol, methanethiol), and mixed

4
strenuously. This mixture is later transferred to a steel autoclave. The advantage of this
method over others is, it is carried out mostly at low temperatures. Template synthesis
comprises desired morphology of the template, where the crystal is specified to grow in a
specified dimension. Through this method, all kinds of nanostructures can be fabricated (0D,
1D, 2D). Many non-layered 2D nanomaterials can be fabricated using template synthesis
[20].
The most well-known method for manufacturing two-dimensional materials is
chemical vapor deposition (CVD). It is extensively utilized to mass-fabricate ultrathin
materials in large quantities. Thin films are formed on the heated substrate through a
chemical reaction of gaseous precursors. It is useful in generating high-quality 2D materials
with controllable thickness and scalable size. CVD is so far the best method to fabricate large
graphene sheets at a macroscale. Recently, good progress was made in fabricating large
graphene sheets using copper and nickel as substrates through CVD [21], [22]. Carbon was
supplied in gaseous form and allowed to grow on a metal sheet. Metal sheet acted as both
substrate and catalyst.

1.4 Graphene
Graphene is a honeycomb structure made up of one layer of sp2 hybridized carbon
atoms. The Discovery of graphene has revolutionized the research in nanomaterials. It
serves as the basic component for making many carbon-based nanomaterials. Carbon
nanotubes (CNT) are obtained by rolling graphene. With 0.34 Å as the least possible
thickness, graphene remains as the thinnest material discovered so far. With the breaking
stress of more than 130 GPa, pristine graphene remains as strongest material to date.
Graphene displays extraordinary properties like high carrier mobility (electron mobility),
high stiffness, and high tensile strength. Despite all the above-mentioned magnificent
characteristics, the 2D graphene sheet fabricated in laboratories has intrinsic imperfections
(discussed in Section 1.5).
Graphene has Young’s Modulus of 1.01 ± 0.03 TPa [23]. Graphene can sustain up to
20% strain (in the Uniaxial direction), before breaking. Young's Modulus of graphene was
calculated through nano-indentation. Indentation was performed on a graphene sheet
which was suspended on a silicon substrate. Another notable property of graphene is its
Young's modulus does not alter appreciably till it reaches around 1200K. This widens the
applications of graphene to higher temperatures. Graphene has a thermal conductivity of

5
5000 W/mK (at least 20 times higher than copper) [24], and a thermal expansion coefficient
of -7 x 10-6 K-1. These tremendous characteristics of graphene widen its applications in the
fields of thermal management, energy, and electronics. Despite these, graphene also
exhibits excellent electronic properties. With electron mobility of 250,000 cm2/Vs (for
comparison, Silicon has mobility of 1500 cm2/Vs), graphene stands as a material with one of
the highest electron mobility. It is very high as compared to other conventional materials.
Also, these properties were not easily affected by temperature change (which is not the case
for many conventional materials) [25]. Electrons and holes act as charge carriers when it
comes to normal electronic substances. However, in graphene, the charge carriers behave
like a phonon (whose speed is just 300 times less than speed of light).
Graphene has applications in many diverse fields like Energy, biosensors, electronics
composites, membranes. Recently, graphene is receiving more attention in chemical
sensors. Chemical gas sensors prepared of graphene can spot even one molecule of gas [26]
(the adsorbed molecule have change the local charge density, which leads to a difference in
electric resistance). Along with these, even biosensors were developed. A sensor that can
identify glucose was developed [27]. The glucose concentration is monitored throughout the
procedure of reduction of oxygen. These findings were in good accord compared to those
obtained from previously used sensors.

1.5 Defects in Graphene


The second law of thermodynamics says that perfect crystalline materials exist only
at absolute zero, and at any amount of temperature, there are imperfections in crystal.
Imperfections in the fabrication process might lead to defects in materials. These
imperfections have a significant impact on the material's mechanical, thermal, electrical, and
optical characteristics.
Intrinsic defects occur when crystalline structure is broken without any existence of
different atoms, whereas extrinsic defects occur when the crystalline structure is disrupted
by an external atom. Intrinsic imperfections can be of any dimension. For example, point
imperfections and vacancies are zero-dimensional defects. Dislocations, stacking faults, and
voids are one-, two- and three-dimensional defects, respectively. Since graphene itself is 2D
material, the dimensionality of defects cannot go beyond 1D. Point defects are quite familiar
in the case of graphene.

6
Stone-Wales defect
Stone-Wales (SW) defect is when four hexagonal rings are transferred into two
heptagon rings and two pentagon rings, with 900 rotation of one C-C bond. It does not have
any added or removed atoms. The formation energy of SW defect is 5 eV [28], [29]. The
kinetic barrier when the transformation is through 900 rotation of C-C bond is 10 eV. The
reverse transformation’s barrier is 5 eV. Once the defect was formed, the high reverse
formation energy makes the SW defect more stable at room temperature. For higher
temperatures, the energy barrier is crossed and there is a high chance of the generation of
Stone-Wales defect. The most common process of fabricating graphene is CVD, which takes
place at high temperatures. Stone-Wales defects are also observed during the failure of a
graphene sheet.

Figure 1: Formation of SW defect (5-7-7-5 ring) with the 900 rotation of C-C bond. (This picture is
taken from Tserpes [30])

Single Vacancy defect


It is the most basic imperfection that occurs in most of the material. It is associated
with the absence of just one carbon atom. Single vacancy defects are experimentally
observed in graphene [31], [32]. When a single carbon atom is removed, Jahn-Teller
distortion occurs, causing the other two dangling bonds to be saturated towards the absent
atom. It is possible that five and nine-membered rings are developed as a result of this.
Single vacancy defect have a formation energy of 7.5 eV [33], [34]. It has been observed that
the mechanical and electronic characteristics and applications of graphene sheets are
significantly impacted by both Stone-Wales and vacancy [35], [36]. If there is a loss of
multiple atoms in graphene, it leads to multiple vacancy defects. Even these defects might

7
lead to distortions and form two pentagon and an octagon ring, in place of four hexagon
rings.
The properties and applications of graphene, various kinds of imperfections that are
formed during fabrication are discussed in this chapter. In the next Chapter, several works
done by different studies on graphene and the effects of these defects (primarily focusing
on point defects like Stone-Wales, vacancy defects) on mechanical and tensile parameters
is discussed. Later, the simulation techniques and formulation are discussed. In Chapter 4,
the results obtained by our simulations are discussed.

Figure 2: Formation of Jahn-Teller distortion due to Single vaccancy defect in a graphene


sheet. (This picture is taken from Valencia [37])

8
Chapter 2

Literature Survey

2.
2.1 Preliminary Work
In this section, the preliminary work that was done by other researchers is discussed.
Initially, the importance of studying defects is discussed. Then, how one topological defect
affects the mechanical properties of graphene is discussed. Later this is extended, to
studying the effect of two topological defects, and their interaction of graphene sheet.
Theoretical background on fracture parameters like strain energy release rate is discussed.
Properties of a crack in non-inert environments (oxygenated or hydrogenated) is discussed.
Finally, the importance of the selection of proper interatomic potential and the relevance of
understanding how out-of-plane deformations impact graphene's mechanical and fracture
characteristics is also discussed.
Even though the fabrication techniques for large pristine sheets have seen significant
changes [38], [39], for macroscale fabrication, the majority depends on the chemical vapor
deposition (CVD) technique[21], [22]. The likelihood of topological defect generation
increases as the size of the intended graphene sheet grows larger. In CVD, the material is
allowed to grow on substrate. There is a very high chance of vacancy defects in this process.
Also, the whole process of CVD occurs at high temperatures, where the graphene sheet
crosses the kinetic barrier of Stone-Wales defect (10eV). Due to the substrate graphene
interactions, there is also a high chance of the formation of extrinsic defects like
substitutional defects. This shows that the formation of topological defects like vacancies,
cracks, and Stone-Wales defects becomes unavoidable while fabricating graphene sheets.
These topological imperfections affect the mechanical and electronic characteristics to a
significant extent.[40], [41]. A recent study showed that the graphene’s fracture toughness
can be very low as 15 Jm-2 [42]. Hence, to investigate the mechanical and electrical
properties of a graphene sheet, it is more feasible and acceptable to simulate a graphene
sheet with imperfections and defects rather than a pristine one. Macroscope properties like

9
ultimate tensile strength (UTS), energy release rate, and stress intensity factor (SIF), which
are used in parameterizing the fracture strength of graphene sheets, are also affected by
these defects.
Dewapriya et al. [43] in the year 2013, modeled graphene sheets of size 27 x 27 nm
with different center crack lengths ranging from 0.4 nm to 2.5 nm at two different
temperatures, 1K and 300K. The interatomic interactions were defined using the AIREBO
potential [44]. The fracture strength of graphene sheets was calculated using MD
simulations, Griffith [45] formulation, and Inglis [46] formulation. He also normalized
Griffith's formulation to the molecular level, by subtracting a constant from failure stress
which is discussed in Section 3.4. It was shown that fracture strength calculated from MD
simulations and Continuum approaches are in line with very little difference. It was also
shown, a graphene sheet with 2.5 nm has around 55% of the strength of a pristine graphene
sheet. Later in 2014, Dewapriya and Rajapakse [47] extended this work by studying the
effect of temperature and strain rate. It was shown that at a reduced strain rate (that is of
order 0.001 s-1) and higher temperatures, the decrease in strength was very rapid. The
decrease in UTS of graphene sheets at 1500K is five times greater than at 300K. The strain
rate in MD simulations is of order of 109 s-1. At this strain rate, the effect of strength
reduction at high temperatures observed is not as high as in 10-3 s-1. It was concluded that
in MD, strength reduction is over predicted in case of higher temperatures. Georgios et al.
[48] examined how the length of the center crack and the temperature affected the tensile
strength of graphene sheets. MD simulations were used in modeling graphene sheets with
different crack lengths at different temperatures. AIREBO potential was used in defining the
interatomic forces. It was shown that graphene sheets break at lower stresses and strains
when exposed to high temperatures. Ansari et al. [49] observed the effect of a vacancy
defect and a Stone-Wales defect on Young's modulus. A graphene sheet of size 61 x 48 nm
was modeled to study. The Tersoff-Brenner potential was used in defining interatomic
forces. It was concluded that vacancy defect has a larger effect on graphene sheet than
Stone-Wales defect.
While the studies on the effects of a single topological defect, on strength of
graphene, were going, Dewapriya and Meguid [50] extended this to the effect of two
different imperfections and their interaction on a graphene sheet. A 60 x 60 nm graphene
sheet with 10 nm edge crack and a vacancy, was modeled for MD simulations.
Intermolecular forces were defined by using AIREBO potential. For measuring fracture

10
parameters, Griffith’s formulation of SIF was used. The crack-vacancy interaction was
studied by placing them at different orientations and angles. It was demonstrated that with
the proper placement of crack-vacancy, graphene’s fracture toughness can be improved.
Additionally, it was shown that crack-vacancy interaction can be used effectively for crack
healing. With a designed distribution of crack-vacancy, a 67% increase in fracture strength
was observed. Similarly, Rajasekharan and Parashar [51] investigated the impact of Stone-
Wales defects and crack interaction on fracture toughness of graphene sheets. AIREBO
potential was used for defining interatomic forces. SIF was calculated using Griffith’s
formulation. Initially, it was shown that the tensile strength was reduced when the
concentration of Stone-Wales defects in graphene sheet was increased. Later, when the
defects were introduced in a designed distribution around the crack, the fracture was
improved by 16% and 8% in ZZ and AC directions, respectively. Due to the crack, Stone-Wales
defect interaction, the material property changed from brittle to semi ductile configuration.
Strain energy release rate is the energy released per additional surface area
generated, during the crack propagation. At the nanoscale, most of the macroscale fracture
parameters are not completely applicable. But Energy release rate remains a crucial
parameter to characterize fracture properties of a material. Jin and Yuan [52] introduced
two different methods for the calculation of energy release rate and SIF. Global Energy
Method (GEM) and Local Force Method (LFM). LFM is centered on the concept of virtual
work required for stopping further fracture propagation, while GEM is measured from the
energy release rate during crack propagation. Further formulation is discussed in detail, in
Section 3.4. To the accuracy of GEM and LFM, a graphene sheet of size 15 x 17 nm (9840
atoms) with a center crack ranging, 1 nm to 6.5 nm. All the simulations were run at 0K. The
Tersoff-Brenner potential is used in defining interatomic forces. The fracture parameters
were calculated from GEM, LFM, and LEFM methods, and compared. All three methods gave
similar results with very little difference. Using this GEM formulation, Zhang et al. [42]
calculated the energy release rate of three different sample sheets (with an edge crack) of
different lengths, at three different temperatures (1K, 300K, and 1000K) and strain rates.
AIREBO potential was used for defining interatomic forces. Sheets with different initial crack
angles were also considered for simulations. It was shown that with an increase in
temperature, the energy release rate tends to reduce. The influence of external loading on
the crack path is more than crack angle, orientation, and temperature.

11
Most of the literature so far was more focused on graphene under non-inert
conditions. Few researchers studied how the fracture properties of graphene are affected in
a non-inert environment. Huang et al. [53] modeled a 60 Å x 60 Å graphene sheet, and it has
a preexisting notch. MD simulations of tearing of graphene sheets were modeled at three
different environments. Vacuumed, hydrogenated, and oxygenated environments. In the
case of hydrogenated and vacuum environments, notch tends to grow in the AC direction,
whereas in an oxygenated environment case, crack tends to propagate in the ZZ direction.
Akarsh and Parashar [54] studied how fracture properties of graphene are affected by
functional groups like oxide, hydroxy, ether, and epoxide. ReaxFF potential was used in
defining interatomic forces. It was showed that the graphene’s fracture strength reduces as
oxygen functionalization increases.
In most of the research done on graphene, the interatomic forces were defined by
AIREBO potential. Even though AIREBO potential replicates physical forces accurately to an
extent, it has some shortcomings. In AIREBO formulation, the influence of an atom on
another is made zero if the distance between them crosses a cutoff distance. Even though
the effects of torsion are considered in AIREBO (which in many non-reactive potentials was
not considered), the effects of some multibody forces like angle interactions were not
considered. Even the van der Waals forces were not included in AIREBO. Which is not the
case for ReaxFFC-2013. In ReaxFFC-2013 instead of taking cutoff distance, the effect of one atom
over the other takes transition very smoothly. It also considers the effects of lone pair
valence electrons, which were mostly neglected by other potential functions. ReaxFFC-2013 is
discussed in detail, in Section 3.3. Chemical reactions and bond cleavages are depicted
accurately with ReaxFFC-2013 [55]. Jensen et al. [56] in 2015, performed MD simulations on
graphene sheet of dimensions 3.5 nm x 3.54 nm, using ReaxFFC-2013, to confirm the
potential’s accuracy. Simulation-derived Young's Modulus and breaking stress values are
evaluated with experimental results. The results were very accurate. Later, he also
performed MD simulations on a 5.24 nm x 5.19 nm graphene sheet, using different values
of time steps, strain rates, and coupling coefficients to find the condition for convergence of
energy [57]. Chaitanya et al. [58] modeled graphene sheets (size of 10 nm x 10 nm) with
different concentrations of Stone-Wales defects. ReaxFFC-2013 potential was used for defining
interatomic forces. He also studied the effect of defect density on out-of-plane deformations
(rippling and wrinkling) of graphene sheets during uniaxial tension.

12
As quoted by the Mermin-Wagner theorem, except at absolute zero, perfect 2D
material does not exist [59]. There are experimental observations of ripples in suspended
graphene sheets [60], [61]. Thermal fluctuations are the primary reason for these ripples
and wrinkles in graphene sheets [62]. These ripples have a major impact on graphene's
physical characteristics, such as tensile strength [63], [64].

2.2 Research Gaps


For the past decade, there has been a lot of research on the tensile and fracture
characteristics of graphene under different conditions. Especially on how fracture
parameters were affected due to different kinds of imperfections and temperature. But not
much was explored on studying the out-of-plane deformations (rippling and wrinkling) and
their effects on crack propagation, mechanical and fracture properties.
Most of the research on graphene (especially simulations) was done at ideal
conditions. Most were done in a vacuum. Chemical reactions were not considered, which
might not be the case in the physical world.
Graphene has potential applications in space crafts (low earth space orbits), because
of its tremendous properties like high thermal conductivity, high tensile strength. During the
atmospheric reentry, the temperature of orbit reaches 2500 K. Even though there has been
significant work on graphene at high temperatures, it was limited to at max 1000 K.
The impact of initial fracture length, crack angle, and orientation on UTS and SIF of a
2D graphene sheet is studied in our work. Later the effect of out-of-plane deformations was
studied by taking two different cases. In one case, out-of-plane deformations are restricted,
by making the out-of-plane velocity zero. In this case, the sheet remains 2D. In the other
case, there was no restriction on out-of-plane velocity. The tensile strength and fracture
parameters of the graphene sheet in both cases were compared. Along with the effect of
out-of-plane deformations fracture strength, the effect of crack length on out-of-plane
deformations is also studied. Sheets with different crack lengths were taken and the values
of out-of-plane deformations were calculated at different points of loading. Effects of initial
conditions like temperature, on fracture parameters, and out-of-plane deformations were
also studied.

13
Chapter 3

MD Simulations and Formulation

3.
3.1 Introduction to Molecular Dynamics
Several techniques in both experiments and computation were developed to
investigate properties of graphene. Most of the experimental techniques are pre-existing
ones. There are many limitations to these techniques. Phenomenon like rippling, folding and
striations might lead to inaccuracies in the experimental results. This shows that material
properties cannot be determined completely accurately with just experiments. This widened
the scope for computational approaches. To investigate the characteristics of graphene,
scientists used a variety of computational approaches. [65], [66]. MD simulations were used
to study how the out-of-plane deformations impact tensile, fracture parameters of a
graphene sheet.
MD simulations determine the position and velocity of a given model using
approximate finite difference equations from classical mechanics. The origin of MD
simulations dates to the mid-20th century (1950s). Alder and Wainwright [67] introduced the
MD methods, to study the phase transition of the hard-sphere system. The important
feature of MD simulations is their ability to solve multi-body problems of classical
mechanics. MD simulations derive the macroscale properties of a material, by solving the
atomic level equations. Molecular Dynamics comprises of two parts mainly, Molecular
mechanics and Quantum mechanics. Classical Newtonian equations are used in solving
molecular mechanics. By using a semi-empirical method, quantum mechanical calculations
can be solved. MD simulations can be used for different scales. MD serves as a connecting
link between theoretical research and experimental work. MD can also be used for extreme
cases, where experimentation becomes impossible.

14
3.2 Characteristics of Molecular Dynamics
Central concept of Molecular Dynamics is to solve Newton's second law of motion
iteratively in order to determine every atom’s position and velocity. Initial positions of atoms
were provided and later were calculated using the below-given Equation (1) for every step.
To ensure the model's correctness, the entire procedure is separated into a huge number of
time steps.
𝑑2 𝑟𝑗 𝐹𝑗
2
= (1)
𝑑𝑡 𝑚𝑗

Here 𝑟𝑗 and 𝑚𝑗 are position and mass of given specific atom 𝑗. The total force exerted
on the atom 𝑗 is 𝐹𝑗 . Position and velocity are obtained by integrating the above equation for
every time step. For calculating 𝑟(𝑡0 + Δ𝑡) and 𝑣(𝑡0 + Δ𝑡) from 𝑟(𝑡0 ) and 𝑣(𝑡0 ), the force
acting on each atom 𝐹(𝑡0 + Δ𝑡) should be calculated. Force is calculated by differentiating
Potential Energy with respect to position, as given in Equation (2). This process of integrating
at every interval of time is called as finite difference method. Velocity Verlet and leap
algorithm are few examples of finite difference methods. Velocity Verlet algorithm is used
in our simulations. The error in this algorithm is of order O(𝝙t2) [68].
𝜕𝑈𝑖
𝐹𝑖 = − (2)
𝜕𝑟𝑖

Where, 𝐹𝑖 is total acting on 𝑖 atom. Position and Potential Energy are represented by
𝑈𝑖 and 𝑟𝑖 respectively. Total PE calculation is explained in Section 3.3.
While performing MD simulations, the first step is, establishing the starting
coordinates of all atoms which belong to the system considered. Initial positions can be
obtained from experiments. If the material is perfect crystalline, lattice parameters can be
used in defining positions. It can also be given randomly for amorphous material. It should
be made sure that atoms are not overlapping. The positions should be defined in a way that
the behavior of the whole system is close to an actual physical system.
There are many different approaches in integrating the given Equation (1). The
detailed procedure is as given follows,
1. Consider the initial state of the system as 𝞇(t0). The position of an atom is ⃗𝑟𝑖 .
Whereas velocity is ⃗⃗⃗
𝑣𝑖 respectively at time t0.

15
2. Using these positions and velocity, the system’s total energy is computed from
the interatomic potential formulation (which is discussed in detail in Section 3.3).

3. Using this energy, Force ⃗⃗𝐹𝑖 (𝑡) is calculated using Equation (2).
4. The state of the system is 𝞇(t0 + 𝝙t) in sequential time step.

5. Integrating ⃗⃗𝐹𝑖 (𝑡) to find the velocity at sequential step ⃗⃗⃗


𝑣𝑖 (𝑡 + Δ𝑡).
6. Integrating once again to find the position ⃗𝑟𝑖 (𝑡 + Δ𝑡).
This loop continues, till the time load is applied.

Figure 3: Flow chart of velocity verlet algorithm.

In velocity Verlet, two Taylor series are combined to give the final equation of
position.
⃗𝑟𝑖 (𝑡 + 𝛥𝑡) + ⃗𝑟𝑖 (𝑡 − Δ𝑡) − 2𝑟⃗𝑖 (𝑡)
= ⃗𝑟̈𝑖 (𝑡) (3)
2Δt

The truncation error is of order O(𝝙t4). In this method, there is no need of storing the
values of velocity at each step. Velocity can be determined from the Equation (4)
⃗𝑟𝑖 (𝑡 + Δ𝑡) − ⃗𝑟𝑖 (𝑡 − Δ𝑡)
⃗𝑟̇𝑖 (𝑡) = (4)
2Δt

The main disadvantage of this method is, there is a high probability of producing
errors [69]. As a result, the leapfrog method was introduced used to correct this. The velocity

16
is determined at a half-time step in the leapfrog method, while the position is determined
Δ𝑡
at a full-time step. As shown in Equation (5), the velocity at 𝑡 + is calculated using force
2

at previous time step ⃗⃗𝐹𝑖 (𝑡).

⃗𝑟̇𝑖 (𝑡 + Δ𝑡) − ⃗𝑟̇𝑖 (𝑡 − Δ𝑡)


2 2 = 𝑚−1 ⃗⃗𝐹 (𝑡) (5)
𝑖 𝑖
Δ𝑡
.
Δ𝑡
Using ⃗⃗⃗
𝑣𝑖 (𝑡 + 2
), ⃗𝑟𝑖 (𝑡 + Δ𝑡) is calculated as shown in Equation (6).

⃗𝑟𝑖 (𝑡 + Δ𝑡) − ⃗𝑟𝑖 (𝑡 + Δ𝑡) 𝛥𝑡


= ⃗𝑟̇𝑖 (𝑡 + ) (6)
𝛥𝑡 2

There are other algorithms like gear’s predictor-corrector algorithm. This algorithm's
fundamental idea is to estimate the velocity, position of each atom in system at 𝑡 + Δ𝑡 step,
utilizing the data available at 𝑡. These values are corrected using observed values.
To ensure that total unknowns are not greater than equations and also to imitate the
real physical system accurately, various macroscopic characteristics like temperature,
energy, pressure, and volume should be controlled in an MD simulation. The instantaneous
temperature in an MD simulation of a multibody system can be obtained by Equation (7).
𝑀 𝑛
𝜃= 〈∑ ⃗⃗⃗ 𝑣𝑖 〉
𝑣𝑖 . ⃗⃗⃗ (7)
3𝑛𝜅𝑏 𝑖

Where, 𝜃 is instantaneous temperature, ⃗⃗⃗


𝑣𝑖 is the velocity of atom 𝑖. Boltzmann
constant and total degrees of freedom’s count are represented by 𝜅𝑏 and n. It is measured
by minimum coordinates required to describe/specify the atom. An atom has six degrees of
freedom in total, three of which are in translation and three of which are in rotation.
The temperature of the system can be controlled in many ways. One simple way is by
adjusting the temperature as desired value at each step during simulation. This process can
cause perturbations, which have a significant impact on the material's characteristics, which
makes it an undesirable process. To maintain temperature, without much effect on the
properties and behavior of materials, thermostats were introduced. The basic principle on
which the thermostat works is, each atom is given a certain velocity with a probability
distribution, whose average velocity value gives us the desired temperature. Thermostat
developed by Anderson designates a fresh velocity from a Maxwellian distribution to the

17
velocity of a random particle. Later, in thermostats developed by Berendsen and Nose-
Hoover, a damping term was introduced in the equation of motion, which controls the
temperature change. A group of systems having the same macroscopic state, but distinct
microscopic states are referred to as an ensemble. Replicating real-world physical
circumstances is the fundamental goal of an ensemble. The micro-canonical ensemble is the
most frequently used. This ensemble has two different approaches. The first one is where
the total number of particles (N), Total Volume of the system, and Temperature of the
System (T) are maintained as constant. This is called an NVT ensemble. Total Energy (E) is
kept constant instead of temperature in the other case. This is called an NVE ensemble. But
in practical applications, keeping energy constant is not feasible.
In MD simulations, the system can have three types of boundary conditions, free,
fixed, and periodical. The free boundary condition is mainly applicable to the cases where
the influence of boundary is negligible. It is applicable to systems like atoms with very high
speeds and a single cluster of particles in a vacuum. Fixed boundary conditions are most
unphysical models. It is used as a complimentary one with other BCs for a more accurate
model. Periodic boundary helps us model macroscopic properties from fewer particles. Also,
in periodic boundary, the boundary effects caused are minimum (due to repetition of the
unit cell). A periodic boundary is more suitable for our case.
In our situation, LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator)
software was used for MD simulations. This is open-source free software. It is mainly used
for solving classical MD problems. It was developed by Steve Plimpton. LAMMPS input script
comprises four parts. Initializing, defining the type of atoms and position of atoms, defining
potential and output parameters. OVITO has been used for the visualization of simulation
output.
A graphene sheet of size 20 x 20 nm is considered for simulations. The graphene sheet
was positioned on an XY plane, with the Z-axis designated as the graphene sheet's normal.
Initially, uniaxial loading was performed in Mode-1 (opening mode) fracture, in which crack
is perpendicular to external loading. Three different initial crack lengths of size varying from
0.7 to 2.4 nm, in case of a crack in AC direction and 0.7 to 2.6 nm, when the crack is in ZZ
direction. Initial crack length size is kept less than 1/10th of characteristic dimensions of
graphene sheet to avoid finite-size effects. A graphene sheet with crack oriented in AC
direction is demonstrated in Figure 4.

18
Figure 4: Computational model of the graphene with dimensions, 20 nm x 20 nm. The
initial center cracks length in the AC direction, and load is applied in a direction
perpendicular to the initial crack. Illustration of AC crack.

Later, the load was applied in mixed mode. Graphene sheets with initial crack inclined
at angles 600 and 1200 were considered for mixed-mode of loading. Graphene sheets with
four different crack lengths ranging from 0.7 nm to 2.4 nm, were considered for simulations.
Loading in both the AC and ZZ directions was examined. Figure 5 shows a graphene sheet
with an inclined fracture.

19
Figure 5: Computational model of the graphene with dimensions, 20 nm x 20 nm. The
initial center cracks length at 600 to the AC direction, and load is applied in either of the
direction (AC or ZZ). Illustration of inclined crack.

Loading is applied at 1 ns-1. Here, timestep is critical in determining the simulation's


stability and accuracy. The vibrations of atoms are at the order of 1015 Hz. To define the
position and velocity of an atom at a certain point of point accurately, the time step should
be less than the order of 10-15 sec. Jensen et al. [57] checked the stability of simulations with
different time steps, thermostats, and strain rates. He also verified the accuracy of ReaxFFC-
2013 potential using these simulations. He showed that, for timestep above 1 fs, the potential
energy tends to diverge. From Figure 6, it can be observed that for all time steps below 1 fs,
the potential energy is not diverging. The maximum possible time step for convergence
condition is 0.5 fs.

Figure 6: Potential Energy w.r.t strain of graphene sheet, at different time steps. The
image is taken from Jensen [57].

3.3 Interatomic Forces


ReaxFFC-2013 is used in defining the interatomic forces for our simulations. For the past
two decades, it has been widely recognized and developed. The computational cost of
quantum chemistry limits its applications to smaller systems. Traditional force fields, which
can describe mostly physical forces are limited for very large systems. Interatomic force
fields (potentials) acts as a medium between these two extremes. ReaxFFC-2013 is bond-order
based potential. ReaxFF potential does not have any discontinuity in forces and energies. It

20
reports the formation and breaking of chemical bonds accurately. To define an interatomic
force field, a detailed parametric study on atomic bonding, bending, torsion properties of
every atom and their interaction with the whole system, using both experimental and
quantum data. This helps in creating a dynamic model. ReaxFF also comprises physical
interactions like van der Waals and columbic interactions. It serves as a perfect link
connecting quantum chemistry and interatomic potential force field. ReaxFF can simulate
systems having lakhs of molecules, which is highly impossible using quantum chemistry.
ReaxFF was used in modeling, many types of chemical systems like in fuel cells [70],
nanotubes [71], combustion [72]. Van Duin et al. [73] created ReaxFF for the first time in
2001.
Initially, ReaxFF parameters were developed to model organic systems [73]. Later
parameters were developed for other metal and non-metal elements [74], [75]. Eventually,
ReaxFF became more diverse. Van Duin et al. [74] in 2003, modeled interatomic forces for
Silicon oxide systems, which are in line with the quantum mechanics calculations. Later it
was developed for other elements like Aluminum, Zinc, Magnesium, etc.. [76]–[78]. In the
ReaxFF potential, parameters of nearly all elements were generated.
ReaxFF has a wide range of applications. Chemical reactions like oxidation, pyrolysis (which
takes place at extreme conditions of temperature and pressure) are tough to study in detail,
by experiments. ReaxFF helps in studying all the organic reactions which take place in
extreme conditions. ReaxFF has applications in defining high energy materials, hydro-carbon
elements, coal, and nanomaterials.
ReaxFF for hydrocarbons (ReaxFFC) was first developed in the year 2001. Since then,
it has been frequently employed to investigate the processes of hydrocarbon-based
chemical processes. Later, in 2008, it was further optimized and developed to investigate
gas oxidation at higher temperatures. The interatomic potential, used in defining the
interatomic forces is ReaxFFC-2013, which is an optimized version of ReaxFFC-2008.

21
Figure 7: Calculation of Energy from atomic positions through interatomic potential.

Equation (8)Error! Reference source not found. gives the equation for the system's total
energy.
𝑈𝑠𝑦𝑠𝑡𝑒𝑚
= 𝑈𝑏𝑜𝑛𝑑 + 𝑈𝑜𝑣𝑒𝑟 + 𝑈𝑢𝑛𝑑𝑒𝑟 + 𝑈𝑣𝑎𝑙 + 𝑈𝑣𝑎𝑛𝑑𝑒𝑟 𝑊𝑎𝑎𝑙 + 𝑈𝑡𝑜𝑟𝑠𝑖𝑜𝑛 (8)
+ 𝑈𝑐𝑜𝑙𝑜𝑢𝑚𝑏 + 𝑈𝑙𝑝

Where, 𝑈𝑏𝑜𝑛𝑑 is energy related to chemical bonds. It is determined by bond order. It


describes the formation and dissociation of bonds. 𝑈𝑜𝑣𝑒𝑟 is over-coordination penalty.
𝑈𝑢𝑛𝑑𝑒𝑟 describes the stability of under-coordination. 𝑈𝑣𝑎𝑛𝑑𝑒𝑟 𝑊𝑎𝑎𝑙 and 𝑈𝑐𝑜𝑙𝑜𝑢𝑚𝑏 describe
van der Waals energy and columbic interactions, respectively. These are non-bonded forces.
𝑈𝑣𝑎𝑙 and 𝑈𝑙𝑝 represents energy related to valence and lone-pair interactions, respectively.
𝑈𝑡𝑜𝑟𝑠𝑖𝑜𝑛 and 𝑈𝑎𝑛𝑔𝑙𝑒 describes the energy associated with four-body torsion and valence
angle strain.
All the forces in ReaxFFC-2013 can be divided into two types. Bonded and non-bonded
forces. Non-bonded comprise Columbic and van der Waals interactions. Rest all are bonded
interactions. The majority of non-bonded interactions are determined by the distance
among the atoms. Initially, bond-order for each atom pair should be determined to calculate
bonded interactions.
Bond order is given in Equation (9).

22
𝜎 𝜋 𝜋𝜋
𝑏𝑜𝑖𝑗 = 𝑏𝑜𝑖𝑗 + 𝑏𝑜𝑖𝑗 + 𝑏𝑜𝑖𝑗 (9)

𝜎 𝜋
𝑏𝑜𝑖𝑗 is total bond order, 𝑏𝑜𝑖𝑗 is bond order due to single bond, 𝑏𝑜𝑖𝑗 is bond order
𝜋𝜋
contribution of the double bond, whereas 𝑏𝑜𝑖𝑗 is the contribution of the triple bond. The

contribution of single bond is shown in Equation (10). The equations for double and triple
bonds are similar to a single bond.
𝑃𝑏𝑜2
𝜎
𝑟𝑖𝑗
𝑏𝑜𝑖𝑗 = exp (𝑃𝑏𝑜1 ( 𝜎 ) ) (10)
𝑟0

𝑟0𝜎 represents the C - C (𝝈 bond) equilibrium distance. 𝑟𝑖𝑗 is the actual distance
between two atoms. 𝑃 is bonding term, it is parameterized in a way such that it yields a
value that agrees with the values predicted through quantum mechanics.
It can be observed that total bond order is a smooth and continuous function. In the
case of non-reactive force fields, the effect of one atom on another is made zero after a
certain cut-off distance, which is not the case in ReaxFF.
Determining the atomic positions is the first step in estimating the overall energy of
the whole system. With these atomic positions, energy contributed by nonbonded
interactions like van der Waals force (𝑈𝑣𝑎𝑛𝑑𝑒𝑟 𝑊𝑎𝑎𝑙 ) and columbic forces (𝑈𝑐𝑜𝑙𝑜𝑢𝑚𝑏 ) can be
determined. For calculating energy contributed by bonded interactions, bond order is
calculated from atomic positions. With this bond energy (𝑈𝑏𝑜𝑛𝑑 ) can be calculated. Further,
bond order is adjusted for over-coordination and 𝑈𝑜𝑣𝑒𝑟 is calculated. Eventually, with this
corrected bond order, energies related to three-body, and four-body interactions like
torsion (𝑈𝑡𝑜𝑟𝑠𝑖𝑜𝑛 ) and angle (𝑈𝑎𝑛𝑔𝑙𝑒 ) interactions are calculated.

Over Coordination of bond order


Bonded forces begin at a considerably greater distance than those reported in many
reactive force fields [79], [80]. ReaxFF is capable of correctly modeling long-range, partially
bonded interactions. But due to these long-range forces, the bond order due to the second
nearest neighbor is overestimated (sometimes carbon might get bond order more than 4,
due to this overestimation).

Three-body interactions

23
In non-reactive force fields, torsion, and angle (three-body) interactions were defined
very rigidly. Irrespective of the bond order (weak or strong), the same function is used in
defining these interactions. But in the case of ReaxFF, torsion and angle interactions are
dependent on bond order. The transition in these forces is smooth when an atom breaks
from a molecule. The force imposed on it as a result of three-body contact gradually
decreases.
For a three-atom body, the expression for energy-related to angle interaction is given
in Equation (11).
𝑈𝑎𝑛𝑔𝑙𝑒
(11)
= [1 − 𝑒𝑥𝑝(𝜆. 𝑏𝑜13 )]. [1 − 𝑒𝑥𝑝(𝜆. 𝑏𝑜23 )]. {𝜅𝑎 − 𝜅𝑏 . 𝑒𝑥𝑝(−𝜅𝑏 . (𝜙 − 𝜙0 )2 )}

Bond ordering of two bonds linking three atoms is 𝑏𝑜1 and 𝑏𝑜2 . 𝜙0 and 𝜙 are the angle at
equilibrium and actual angle between the two bonds. 𝜆 is the parameter that makes sure
that the values are in line with quantum mechanics calculations. The depth and width of
angular potential are determined by 𝜅𝑎 and 𝜅𝑏 .

Non-bonded Interactions
ReaxFF potential uses the hardness and electronegativity parameters of every atom
in the system to measure how atoms are polarized within molecules. Irrespective of the
presence of a chemical bond, forces like coulomb and van der Waal interactions are
computed in the ReaxFF force field, which is not the case with many non-reactive force fields
[73], [81], [82]. Energy associated with columbic interaction is given in Equation (12).

𝑞𝑖 𝑞𝑗
𝑈𝑐𝑜𝑢𝑙𝑜𝑚𝑏 = 𝐶. 1 (12)
3 3
1
{𝑟𝑖𝑗3 + (𝛾 ) }
[ 𝑖𝑗 ]

𝑞𝑖 and 𝑞𝑗 are amount of polarization of atoms. 𝑟𝑖𝑗 is separation of distance among two
atoms. 𝛾𝑖𝑗 is shielding factor (for avoiding excessive columbic and van der Waals
interactions).
The total energy calculated from this section is used in calculating forces, through which
velocity and position of atoms are calculated eventually.
3.4 Fracture Parameters

24
The fracture parameters which are discussed in this section are ultimate tensile
strength, stress intensity factor, and energy release rate. The fundamental quantity that
must be assessed to determine additional fracture characteristics is stress. So, initially stress
formulation at an atomic scale is discussed. Formulation for energy release rate and stress
intensity factor are discussed eventually.

Virial Stress Formulation


The macroscopic definition of Cauchy stress does not have much significance at the
atomic scale (since it does not come under continuum scale). At nanoscale, an equivalent
definition for stress is necessary. To measure and calculate local atomic stress, several
definitions were introduced [83]–[86]. Virial stress was derived from the virial theorem given
by Clausius and Maxwell in 1870.
The expression of Virial stress tensor is given in (13).
𝑁
1 1 𝛽 𝛼𝛽
𝜎𝑖𝑗 = ∑ [ ∑ (𝑅𝑖 − 𝑅𝑖𝛼 ) 𝐹𝑗 − 𝑚𝛼 𝑣𝑖𝛼 𝑣𝑗𝛼 ] (13)
𝑉 2
𝛼 𝛽=1

Here, (i, j) are defined by the x, y, and z directions. 𝛽 vary from 1 to N and the total
neighboring atoms count is described by N of 𝛼. 𝑉 is the total volume. 𝑚𝛼 represents the
mass, whereas 𝑣𝑖𝛼 represents velocity of atom 𝛼, respectively. 𝑅 𝛼 defines the position of the
𝛼𝛽
atom 𝛼. 𝐹𝑗 represents the force applied on atom 𝛼 by atom 𝛽 in direction 𝑗.

Virial stress (which is calculated from MD simulations) should be averaged over a


certain amount of time, to make it identical to Cauchy stress. It is calculated using the
volumetric derivative of the system's total energy. In Equation (13), the first term describes
the system's potential energy, whereas the second term describes the system's kinetic
energy.

Energy Release Rate Formulation


Strain Energy release rate was introduced by Griffith in 1921 to evaluate material
property [87]. It may be thought of as the total energy released per unit new surface area
created during the crack propagation process [42], [52]. Since the process of loading and
crack propagation is a pseudo-static process, the energy release rate should be calculated
for a small length of crack propagation.

25
𝑑𝜙 Δ𝜙
𝐺𝑐 = − = − (14)
𝑧(2𝑑𝑎) 𝑧(2Δ𝑎)

2𝑎 is the initial center crack length. 2Δ𝑎 and Δ𝜙 are increase/change in center crack length
and increase of system’s total energy, respectively. 𝑧 is the thickness of the graphene sheet.

Stress Intensity Factor Formulation


The stress Intensity Factor is a measurement of a material's fracture strength. It was
introduced by George R Irwin, in the year 1957.
Stress Intensity Factor is formulated using two different methods, Linear Elastic
Fracture Mechanics (LEFM) and Global Energy Method (GEM). LEFM was developed by
Griffith [87] in the year 1921. Later this was extended and modified by Irwin [88] and Rice
[89] in the years 1957 and 1968, respectively.
𝐾𝐼𝐶 = 𝜎𝑓 √2𝜋𝑎 (15)

𝐾𝐼𝐶 is Stress Intensity Factor, 𝜎𝑓 is failure stress and 2𝑎 is center crack length. Later, this was
normalized by Dewapriya [43] for graphene sheet with crack as given in Equation (16).
𝐴
𝜎𝑓 = +𝑐 (16)
√𝑎

𝜎𝑓 is the stress where sheet fails, 𝑎 is initial crack length, A and c are constant for a given
temperature. Accordingly, the expression for SIF is modified as given in Equation (17).

𝐾𝐼𝐶 = (𝜎𝑓 − 𝑐)√𝜋(2𝑎) (17)

𝑐 is constant for all given crack lengths. It is used for normalizing the solution for SIF.
Griffith’s formulation for SIF (LEFM) is modeled mainly for fracture at the continuum scale.
Its applicability at the nano scale is not yet explored completely. Apart from LEFM, Jin and
Yuan [52] modeled two different methods for the calculation of SIF, the Global Energy
Method (GEM) and the Local Force Method (LFM). GEM is through calculation of total
energy dissipated during the crack propagation (For a small increase in crack length). In
GEM, first, strain energy release rate (Mentioned in Equation (14)) is calculated.
After that, SIF should be calculated from the formula given in Equation (18).

26
𝐾1𝐶 = √𝐺1𝐶 𝑌 (18)

𝐾1𝐶 is Stress Intensity Factor of Mode – 1 fracture, 𝐺1𝐶 is Strain Energy release rate
(Calculated from Equation (14)). 𝑌 is Young’s Modulus of graphene sheet. Initial 10% of
uniaxial loading is considered for calculation of Young’s Modulus.
LEFM assumes that fracture development is self-similar, whereas GEM does not. The
basic principle of LFM is the amount of virtual work needed to stop the growth of the
fracture.

27
Chapter 4

Results and Discussion

4.
In this chapter, initially, the effect of crack length, crack orientation, and initial crack
angle on Ultimate tensile strength (UTS) is discussed. This is extended in studying the
effect of out-of-plane deformations at different crack lengths, on the fracture parameters
like Ultimate tensile strength (UTS), stress intensity factor (SIF), and Energy release rate. It
was also studied how different initial crack parameters like length, orientation, and angle
affect the out-of-plane deformation. Finally, the effect of the initial condition like
temperature on out-of-plane deformations and fracture toughness is discussed.

4.1 Effect of initial crack length, orientation, and angle on tensile strength
Load was applied in Mode-1 or opening mode (perpendicular to the crack). It was
observed that pristine graphene has tensile strength of 137.5 GPa in AC direction and 123.5
GPa in ZZ direction. The graphene has high ultimate tensile stress and strain in armchair
orientation. With the insertion of 0.7 nm center crack in 20 x 20 nm graphene, the UTS has
reduced by 20.7% (from 137.5 GPa to 108.9 GPa) in AC direction (load is in AC direction,
crack in ZZ direction) and 5% (from 123.5 GPa to 119.1 GPa) in ZZ direction. This shows, for
the same crack length, the graphene sheet is more affected by the crack in ZZ direction than
AC direction. Graphene sheets with cracks in AC direction, tend to have higher ultimate
tensile stress than those with cracks in ZZ direction, but they have lower tensile strain. A
graphene sheet with a crack length of 0.7 nm in AC orientation has a tensile strain of 0.26,
whereas a sheet with the same crack length but oriented in ZZ has a tensile strain of 0.44.
Even though the tensile stress is low when the crack is oriented in the ZZ direction, it can
stand for a higher amount of strain than sheets with cracks oriented in the AC direction. Till
the point of failure, the trend remained similar for the given orientation of crack.

28
Figure 8: Stress vs Strain of graphene sheet at different crack lengths and orientations.

Along with the effect of fracture orientation and length, the effect of initial crack angle
on tensile strength was also studied. Two different initial crack angles apart from opening
mode, (600 and 1200) were considered. From Figure 9a, it can be said that UTS is almost same
for the case where the initial angle of crack is 600 and 1200. For most of the crack lengths,
sheets with an initial angle 600 have UTS slightly higher than the sheet with an initial angle
1200. Sheets tend to have a higher tensile strength in an armchair than zigzag irrespective of
crack length and crack angle (similar behavior is observed in the opening mode of the crack).

29
(a)

(b)

Figure 9: (a) Stress vs Strain of graphene sheet with initial crack oriented in two different
crack angles (θ = 600 and 1200), and at 4 different crack lengths, with load acting in AC
direction. (b) Stress vs Strain of graphene sheet with initial crack oriented in two
different crack angles (θ = 600 and 1200), and at 4 different crack lengths, with load
acting in ZZ direction.

To study the effect of crack angle more effectively, stress vs strain at the same crack
length and orientation but the different crack angle was plotted. Crack length of 1.7 nm and
loading in ZZ direction is kept constant throughout for 3 different initial crack angles (00, 600,
and 1200).
From Figure 10, it can be observed that graphene sheet tends to have higher
strength in opening mode, than in the case of mixed mode. For initial crack angle 60 0 and
1200, the tensile strength is almost equal. It can be concluded that, by decreasing the crack
angle, the ability to resist the load increases for a graphene sheet.

30
Figure 10: Stress vs Strain of graphene sheet with crack length 1.7 nm, at three different
initial crack angles.

The process of crack propagation in the opening mode of the crack is depicted in
Figure 11. The crack propagation is in the direction perpendicular to loading. This
phenomenon is observed at all crack lengths. In the case of pristine, failure occurs in diagonal
direction.

(a) (b)

31
(c) (d)

Figure 11: Crack propagation in the opening mode of fracture.

Crack propagation in mixed mode of crack (θ = 600) is depicted in Figure 12. Crack
propagated in a direction perpendicular to loading (similar to opening mode). Irrespective
of crack angle and crack length, the propagation is in the direction perpendicular to external
loading. This shows us that the direction of crack propagation is heavily influenced by the
direction of external loading than the initial crack angle.

(a) (b)

32
(c) (d)

Figure 12: Crack propagation in a mixed mode of fracture.

The initial angle of the crack also affects the energy release rate. From Figure 13, It
can be noted, the energy release rate takes values of 15 ± 3 J/m2. Whereas it is 30 ± 2 J/m2
in case of the opening mode of crack. So, with a different initial crack angle, there is a 50%
drop observed in the energy release rate.

Figure 13: Energy release rate vs Crack length of graphene sheets with various initial
crack angles at 300K.

33
4.2 Effect of out-of-plane deformations on fracture toughness
In this section, the effect of out-of-plane deformations on fracture parameters like
UTS, SIF, and energy release rate is discussed.
The effect of out-of-plane deformations is studied by taking two different cases. One
case, where the out of plane velocity (velocity in Z-direction) is constrained to zero. Since
the sheet remains in a plane, it is referred to as 2D case. In another case, where out of plane
velocity remains unconstrained (3D case).

Figure 14: Stress vs Strain of 2D and 3D graphene sheet at various crack lengths.

The 3D graphene sheet has a higher tensile strength than the 2D graphene sheet, as
can be shown in Figure 14. 11.8% difference in tensile strength is observed between the 2D
and 3D case of graphene sheet with 2.4 nm crack length in AC direction. This shows that
ripples stabilize the graphene sheet and helps it to stand for a high load.
Work done by external loading is utilized for an increase in both kinetic and potential
energy, in the case of 3D. Whereas in the case of 2D, it is utilized only for increasing potential
energy, which might lead to failure at low stress.
Another important term to consider while studying fracture propagation at the
nanoscale is the energy release rate. It can be defined as the total energy released per new
surface generated. Since the deformation is a pseudo-static phenomenon, the energy
release rate is calculated for an infinitesimal increase in crack length (𝝙a = 0.256 nm). The

34
total new surface area generated is the product of thickness of the sheet and increase in
crack length.

Figure 15: Energy release rate vs Crack length at 300K of 2D and 3D sheets.

Figure 15 shows the trend of energy release rate with respect to crack length for the
case of 3D and 2D graphene sheets. The energy release rate remained almost constant
throughout different crack lengths for a given specific case. For a sheet with a crack in the
AC direction, the energy release rate varies from 24 to 29 J/m2, in the case of a crack in the
ZZ direction, it is 25 to 32 J/m2. The obtained values of energy release rate are consistent
with Zhang's findings [42]. There is no significant difference between 2D and 3D cases, but
the average energy release rate is slightly higher for a 3D sheet than a 2D sheet.
Stress Intensity factor is calculated using two different methods, Linear elastic fracture
mechanics and the Global Energy method.
SIF calculated from LEFM has low precision and high discrepancy as compared to
GEM. SIF of 3D sheets is slightly more than 2D sheets when calculated through LEFM. 3D
sheets with ZZ crack (loading in AC), have values ranging from 5.91 ± 0.8 MPa√m and 5.08 ±
0.7 MPa√m, when calculated from LEFM and GEM, respectively. Similarly, for a 3D sheet
with AC crack, the values are 5.11 ± 0.2 MPa√m and 5.06 ± 0.27 MPa√m, when calculated
from LEFM and GEM, respectively.

35
The crack considered here is a center crack. While calculating the SIF, the
formulation in linear elastic fracture mechanics considers self-similarity. This may not be the
case in crack propagation at the nanoscale. Whereas the Global energy method does not
consider self-similarity, it is directly calculated from the energy release rate. This shows that
GEM is a better method to measure SIF.

Figure 16: Stress Intensity Factor vs Crack length for both 2D and 3D sheets at 300K.

4.3 Effect of Initial crack on out of plane deformations of the graphene sheet
In this section, the effect the initial fracture conditions like initial crack length, angle,
and orientation affect out-of-plane deformations. Initially, the effect of initial fracture length
on out-of-plane deformations is studied. Out-of-plane deformations are caused mainly due
to thermal fluctuations. However, the amplitude of out-of-plane deformations is influenced
by a number of different variables.
Figure 17 shows the out-of-plane deformation amplitudes of graphene sheets with
different crack lengths in AC and ZZ directions. It can be observed, with small enlargement
in crack length, there is a drastic increase in the out-of-plane deformation’s amplitude.
Graphene sheets with crack lengths of 2.6 nm in the AC direction have 112 percent higher
amplitude than those sheets with a crack length of 0.7 nm. Similarly, the out-of-plane
deformation’s amplitude of graphene sheet with initial crack length 2.4 nm in ZZ direction is
97% higher than the sheet with crack length 0.7nm. Also, sheets with crack oriented in the

36
ZZ direction have a higher amplitude of out-of-plane deformations than sheets with crack
oriented in AC direction. Both follow a similar trend (almost parallel).

Figure 17: Out-of-plane deformations of graphene sheet at various crack lengths and
orientations.

Initial crack angle’s influence (at different crack lengths) on out-of-plane


deformations can be observed in Figure 18. Graphene sheets with initial crack oriented in
1200 have a slightly higher amplitude of out-of-plane deformations than a sheet with initial
crack oriented in 600. The difference is more at higher crack lengths. The influence of the
initial crack angle is significantly smaller than the effect of the initial crack length.

Out-of-plane deformations have been found to have a substantial impact on


fracture characteristics and crack propagation (discussed in 4.2). In this section, it was
observed that out-of-plane deformations are significantly influenced by crack length, which
shows that there is a coupling effect between the two.

37
Figure 18: Out-of-plane deformations of graphene sheet with different initial crack
angles and crack lengths.

4.4 Effect of temperature on fracture parameters and out of plane deformations


In this section, the effect of temperature on fracture parameters and out-of-plane
deformations is discussed. Figure 19 shows a stress versus strain curve for graphene sheets
with various fracture lengths, orientations (AC and ZZ), and temperatures (300K and 1000K).
It can be noted from the figure, with raise in temperature, there is a drop in the ultimate
tensile strength of the sheet. There is almost a 14% drop in UTS for a given specific sheet
when the temperature is increased from 300K to 1000K. Till the point of failure, the trend in
the plot is similar for both temperatures.

38
Figure 19: Stress vs Strain of graphene sheet at different crack lengths and two different
temperatures (300K and 1000K).

Figure 20 shows the energy release of graphene sheets with different crack lengths,
loaded in opening mode of crack at 1000K. The energy release rate of graphene sheets with
crack oriented in ZZ and AC have values varying from 17.4 to 19.1 J/m2 and 16.7 to 20.7 J/m2
respectively. There is a 33% difference in energy release rate as compared to sheets at 300K.

39
Figure 20: Energy release rate vs Crack length of graphene sheets with various crack
lengths oriented in AC and ZZ directions, at 1000K.

At 1000K, graphene sheets have an average energy release rate of 18 J/m2, whereas,
for graphene sheets at 300K, it is 29 J/m2. The energy release rate has seen a noticeable drop
with an increase in temperature. Thermal fluctuations are the main cause of this drop.
The Stress Intensity Factor of graphene sheets with different crack lengths, at 1000K
is shown in Figure 21. Graphene sheets are loaded in opening mode of crack.

Figure 21: SIF vs crack length of graphene sheets at 1000K.

SIF of graphene sheet at 1000K have values ranging in 4.0 to 4.5 MPa√m. Whereas,
for graphene sheet at 300K, the values are ranging from 4.9 to 5.3 MPa√m. Similar to the
trend in energy release rate and UTS, there is a drop in SIF.
The effect of temperature on out-of-plane deformations is studied in two different
cases. In the first case, amplitudes were taken just before the loading (after the
equilibration). Its purpose is to investigate the effect of temperature in the absence of any
loading. In the second example, the average amplitude at various stresses is calculated
during loading. Figure 22 depicts the 1st case. It can be observed, before loading, the sheet
with high temperature (1000K) has higher amplitudes than the sheet with low temperature

40
(300K), which shows that, as temperature increases, the out-of-plane deformations increase
due to thermal fluctuations.

Figure 22: Out of plane deformations of graphene sheet at various initial crack lengths,
at two different temperatures (room temperature and 1000K) just before loading.

Figure 23 shows the amplitude of graphene sheets at different crack lengths,


orientations, and two different temperatures during loading. During the loading, the
amplitude of out-of-plane deformations is calculated at four different strains and averaged
out (out of plane deformation in previous sections was also calculated the same way). The
trend was very similar to the results in previous section 4.3. The amplitude tends to increase
drastically at higher crack lengths. It is also observed that, during loading, the amplitude
increases initially, reaches a peak, and reduces before reaching the failure point.

41
Figure 23: Out of plane deformations of graphene sheet with different initial crack
lengths, at two different temperatures (room temperature and 1000K) during loading.

42
Chapter 5

Conclusions and Future Work

5.
In this chapter, the concluding marks of our results are discussed, in Section 5.1. Later
the future scope of the work is discussed.

5.1 Concluding marks and key contributions

1. Cracks in graphene sheets affect the tensile strength of the sheet to a greater extent.
With the insertion of 0.7 nm crack, the graphene sheet’s ultimate tensile stress was
reduced by almost 20%.
2. Graphene sheets are more affected by a crack in ZZ direction than in AC direction.
This shows that sheets have higher UTS when the crack is oriented in AC direction
than ZZ direction.
3. Even though graphene sheets have higher UTS when the crack is oriented in the AC
direction, sheets with a crack oriented in the ZZ direction don’t break for greater
strains. For graphene sheet with crack length 0.7 nm oriented in ZZ direction, have
70% higher tensile strain than a sheet with the same crack length but oriented in AC
direction.
4. The ultimate tensile strength of graphene sheets with the same start fracture length
but various initial crack angles is almost the same, but with the decrease in initial
crack angle, the strength of the graphene sheet increases. Graphene sheets tend to
have higher strength in opening mode than mixed mode.
5. The external load’s direction has a greater impact on the direction of crack
development than the initial direction of crack. The crack tends to propagate in the
direction perpendicular to external loading, irrespective of the initial crack angle.

43
6. Similar to the trend of tensile strength, graphene’s energy release rate is higher in
the case of opening mode than mixed mode. In the mixed-mode scenario, the
energy release rate is 50% lower than in the opening mode case.
7. 3D graphene sheets have higher tensile strength as compared to 2D graphene
sheets. This demonstrates that out-of-plane deformations improve the tensile
strength of graphene sheets.
8. The energy release rate remains almost constant for graphene sheets with the same
initial crack angle and temperature but different crack lengths. The average energy
release rate of 3D sheets is slightly higher than 2D sheets.
9. The of out-of-plane deformation’s amplitude tend to grow with fracture length. A
graphene sheet with an initial crack length of 2.6 nm in the ZZ direction, has 112%
higher amplitude than a graphene sheet with an initial crack length of 0.7 nm.
10. Graphene sheets with crack oriented in AC direction tend to have higher amplitudes
than sheets with crack oriented in ZZ direction.
11. The initial crack length has a higher influence on out-of-plane deformations than the
initial crack angle.
12. Initial crack length, orientation, and angle impact out-of-plane deformations, and
out-of-plane deformations alter graphene fracture characteristics. This suggests
that there is a coupling effect.
13. The tensile strength of the material decreases as the initial temperature rises. A 38%
difference is observed in the energy release rate of graphene sheets at 300K and
1000K.
14. Graphene sheets show greater out-of-plane deformation’s amplitude at high
temperatures, regardless of the presence of load.

5.2 Future Scope


This work can be extended further to other two-dimensional materials like
Molybdenum Disulphide (MoS2) and Boron Nitride (HBN). It can also be extended to study
the effect of some other factors like oxidized and hydrogenated atmosphere on fracture and
tensile properties of 2D materials.

44
References

[1] K. S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V. Morozov, et


al., “Two-dimensional atomic crystals,” Proc. Natl. Acad. Sci. U. S. A., vol. 102, no.
30, pp. 10451–10453, 2005, doi: 10.1073/pnas.0502848102.
[2] O. Eksik, J. Gao, S. A. Shojaee, A. Thomas, P. Chow, S. F. Bartolucci, et al., “Epoxy
nanocomposites with two-dimensional transition metal dichalcogenide additives,”
ACS Nano, vol. 8, no. 5, pp. 5282–5289, 2014, doi: 10.1021/nn5014098.
[3] C. Zhi, Y. Bando, C. Tang, H. Kuwahara, and D. Golberg, “Large-scale fabrication of
boron nitride nanosheets and their utilization in polymeric composites with
improved thermal and mechanical properties,” Adv. Mater., vol. 21, no. 28, pp.
2889–2893, 2009, doi: 10.1002/adma.200900323.
[4] W.-L. Song, P. Wang, L. Cao, A. Anderson, M. J. Meziani, A. J. Farr, et al.,
“Polymer/Boron Nitride Nanocomposite Materials for Superior Thermal Transport
Performance,” Angew. Chemie, vol. 124, no. 26, pp. 6604–6607, 2012, doi:
10.1002/ange.201201689.
[5] A. M. Díez-Pascual and A. L. Díez-Vicente, “Poly(propylene fumarate)/Polyethylene
Glycol-Modified Graphene Oxide Nanocomposites for Tissue Engineering,” ACS
Appl. Mater. Interfaces, vol. 8, no. 28, pp. 17902–17914, 2016, doi:
10.1021/acsami.6b05635.
[6] H. Y. Chang, S. Yang, J. Lee, L. Tao, W. S. Hwang, D. Jena, et al., “High-performance,
highly bendable MoS2 transistors with high-K dielectrics for flexible low-power
systems,” ACS Nano, vol. 7, no. 6, pp. 5446–5452, 2013, doi: 10.1021/nn401429w.
[7] T. Georgiou, R. Jalil, B. D. Belle, L. Britnell, R. V. Gorbachev, S. V. Morozov, et al.,
“Vertical field-effect transistor based on graphene-WS 2 heterostructures for
flexible and transparent electronics,” Nat. Nanotechnol., vol. 8, no. 2, pp. 100–103,
2013, doi: 10.1038/nnano.2012.224.
[8] C. Sevik, D. Çaklr, O. Gülseren, and F. M. Peeters, “Peculiar Piezoelectric Properties
of Soft Two-Dimensional Materials,” J. Phys. Chem. C, vol. 120, no. 26, pp. 13948–
13953, 2016, doi: 10.1021/acs.jpcc.6b03543.
[9] K. A. N. Duerloo, M. T. Ong, and E. J. Reed, “Intrinsic piezoelectricity in two-

45
dimensional materials,” J. Phys. Chem. Lett., vol. 3, no. 19, pp. 2871–2876, 2012,
doi: 10.1021/jz3012436.
[10] M. T. Ong and E. J. Reed, “Engineered piezoelectricity in graphene by chemical
doping,” Tech. Proc. 2012 NSTI Nanotechnol. Conf. Expo, NSTI-Nanotech 2012, no. 2,
pp. 161–164, 2012.
[11] J. Qi, Y. W. Lan, A. Z. Stieg, J. H. Chen, Y. L. Zhong, L. J. Li, et al., “Piezoelectric effect
in chemical vapour deposition-grown atomic-monolayer triangular molybdenum
disulfide piezotronics,” Nat. Commun., vol. 6, no. May, pp. 1–8, 2015, doi:
10.1038/ncomms8430.
[12] D. Q. Zheng, Z. Zhao, R. Huang, J. Nie, L. Li, and Y. Zhang, “High-performance piezo-
phototronic solar cell based on two-dimensional materials,” Nano Energy, vol. 32,
no. December 2016, pp. 448–453, 2017, doi: 10.1016/j.nanoen.2017.01.005.
[13] C. Chen, S. Lee, V. V. Deshpande, G. H. Lee, M. Lekas, K. Shepard, et al., “Graphene
mechanical oscillators with tunable frequency,” Nat. Nanotechnol., vol. 8, no. 12,
pp. 923–927, 2013, doi: 10.1038/nnano.2013.232.
[14] R. J. Dolleman, D. Davidovikj, S. J. Cartamil-Bueno, H. S. J. Van Der Zant, and P. G.
Steeneken, “Graphene Squeeze-Film Pressure Sensors,” Nano Lett., vol. 16, no. 1,
pp. 568–571, 2016, doi: 10.1021/acs.nanolett.5b04251.
[15] R. A. Barton, I. R. Storch, V. P. Adiga, R. Sakakibara, B. R. Cipriany, B. Ilic, et al.,
“Photothermal self-oscillation and laser cooling of graphene optomechanical
systems,” Nano Lett., vol. 12, no. 9, pp. 4681–4686, 2012, doi: 10.1021/nl302036x.
[16] I. E. Berinskii, D. I. Indeitsev, N. F. Morozov, D. Y. Skubov, and L. V. Shtukin,
“Differential graphene resonator as a mass detector,” Mech. Solids, vol. 50, no. 2,
pp. 127–134, 2015, doi: 10.3103/S0025654415020028.
[17] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson, I. V.
Grigorieva, et al., “Two-dimensional gas of massless Dirac fermions in graphene,”
Nature, vol. 438, no. 7065, pp. 197–200, 2005, doi: 10.1038/nature04233.
[18] H. Wang, X. Yang, W. Shao, S. Chen, J. Xie, X. Zhang, et al., “Ultrathin Black
Phosphorus Nanosheets for Efficient Singlet Oxygen Generation,” J. Am. Chem. Soc.,
vol. 137, no. 35, pp. 11376–11382, 2015, doi: 10.1021/jacs.5b06025.
[19] C. Tan and H. Zhang, “Wet-chemical synthesis and applications of non-layer
structured two-dimensional nanomaterials,” Nat. Commun., vol. 6, 2015, doi:
10.1038/ncomms8873.

46
[20] W. Cheng, J. He, T. Yao, Z. Sun, Y. Jiang, Q. Liu, et al., “Half-unit-cell α-Fe2O3
semiconductor nanosheets with intrinsic and robust ferromagnetism,” Journal of
the American Chemical Society, vol. 136, no. 29. pp. 10393–10398, 2014, doi:
10.1021/ja504088n.
[21] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, et al., “Large-area synthesis of high-
quality and uniform graphene films on copper foils,” Science (80-. )., vol. 324, no.
5932, pp. 1312–1314, 2009, doi: 10.1126/science.1171245.
[22] A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, et al., “Large area, few-layer
graphene films on arbitrary substrates by chemical vapor deposition,” Nano Lett.,
vol. 9, no. 1, pp. 30–35, 2009, doi: 10.1021/nl801827v.
[23] Y. Jing, Y. Sun, H. Niu, and J. Shen, “Chirality and size dependent elastic properties
of silicene nanoribbons under uniaxial tension,” 13th Int. Conf. Fract. 2013, ICF
2013, vol. 7, no. Md, pp. 5663–5668, 2013.
[24] A. A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan, F. Miao, et al.,
“Superior thermal conductivity of single-layer graphene,” Nano Lett., vol. 8, no. 3,
pp. 902–907, 2008, doi: 10.1021/nl0731872.
[25] R. F. Service, “Carbon sheets an atom thick give rise to graphene dreams,” Science
(80-. )., vol. 324, no. 5929, pp. 875–877, 2009, doi: 10.1126/science.324_875.
[26] F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson, et al.,
“Detection of individual gas molecules adsorbed on graphene,” Nat. Mater., vol. 6,
no. 9, pp. 652–655, 2007, doi: 10.1038/nmat1967.
[27] P. Wu, Q. Shao, Y. Hu, J. Jin, Y. Yin, H. Zhang, et al., “Direct electrochemistry of
glucose oxidase assembled on graphene and application to glucose detection,”
Electrochim. Acta, vol. 55, no. 28, pp. 8606–8614, 2010, doi:
10.1016/j.electacta.2010.07.079.
[28] J. Ma, D. Alfè, A. Michaelides, and E. Wang, “Stone-Wales defects in graphene and
other planar s p2 -bonded materials,” Phys. Rev. B - Condens. Matter Mater. Phys.,
vol. 80, no. 3, pp. 1–4, 2009, doi: 10.1103/PhysRevB.80.033407.
[29] L. Li, S. Reich, and J. Robertson, “Defect energies of graphite: Density-functional
calculations,” Phys. Rev. B - Condens. Matter Mater. Phys., vol. 72, no. 18, pp. 1–10,
2005, doi: 10.1103/PhysRevB.72.184109.
[30] K. I. Tserpes and P. Papanikos, “The effect of Stone-Wales defect on the tensile
behavior and fracture of single-walled carbon nanotubes,” Compos. Struct., vol. 79,

47
no. 4, pp. 581–589, 2007, doi: 10.1016/j.compstruct.2006.02.020.
[31] M. H. Gass, U. Bangert, A. L. Bleloch, P. Wang, R. R. Nair, and A. K. Geim, “Free-
standing graphene at atomic resolution,” Nat. Nanotechnol., vol. 3, no. 11, pp. 676–
681, 2008, doi: 10.1038/nnano.2008.280.
[32] J. C. Meyer, C. Kisielowski, R. Erni, M. D. Rossell, M. F. Crommie, and A. Zettl, “Direct
imaging of lattice atoms and topological defects in graphene membranes,” Nano
Lett., vol. 8, no. 11, pp. 3582–3586, 2008, doi: 10.1021/nl801386m.
[33] A. V. Krasheninnikov, P. O. Lehtinen, A. S. Foster, and R. M. Nieminen, “Bending the
rules: Contrasting vacancy energetics and migration in graphite and carbon
nanotubes,” Chem. Phys. Lett., vol. 418, no. 1–3, pp. 132–136, 2006, doi:
10.1016/j.cplett.2005.10.106.
[34] A. El-Barbary, H. Telling, P. Ewels, I. Heggie, and R. Briddon, “Structure and
energetics of the vacancy in graphite,” Phys. Rev. B - Condens. Matter Mater. Phys.,
vol. 68, no. 14, pp. 1–7, 2003, doi: 10.1103/PhysRevB.68.144107.
[35] F. Banhart, J. Kotakoski, and A. V. Krasheninnikov, “Structural defects in graphene,”
ACS Nano, vol. 5, no. 1, pp. 26–41, 2011, doi: 10.1021/nn102598m.
[36] Ç. Ö. Girit, J. C. Meyer, R. Erni, M. D. Rossell, C. Kisielowski, L. Yang, et al.,
“Graphene at the edge: Stability and dynamics,” Science (80-. )., vol. 323, no. 5922,
pp. 1705–1708, 2009, doi: 10.1126/science.1166999.
[37] A. M. Valencia and M. J. Caldas, “Single vacancy defect in graphene: Insights into its
magnetic properties from theoretical modeling,” Phys. Rev. B, vol. 96, no. 12, pp. 1–
9, 2017, doi: 10.1103/PhysRevB.96.125431.
[38] Y. Hao, M. S. Bharathi, L. Wang, Y. Liu, H. Chen, S. Nie, et al., “The role of surface
oxygen in the growth of large single-crystal graphene on copper,” Science (80-. ).,
vol. 342, no. 6159, pp. 720–723, 2013, doi: 10.1126/science.1243879.
[39] J. H. Lee, E. K. Lee, W. J. Joo, Y. Jang, B. S. Kim, J. Y. Lim, et al., “Wafer-scale growth
of single-crystal monolayer graphene on reusable hydrogen-terminated
germanium,” Science (80-. )., vol. 344, no. 6181, pp. 286–289, 2014, doi:
10.1126/science.1252268.
[40] R. Grantab, V. B. Shenoy, and R. S. Ruoff, “Anomalous strength characteristics of tilt
grain boundaries in graphene,” Science (80-. )., vol. 330, no. 6006, pp. 946–948,
2010, doi: 10.1126/science.1196893.
[41] Y. Wei, J. Wu, H. Yin, X. Shi, R. Yang, and M. Dresselhaus, “The nature of strength

48
enhancement and weakening by pentagong-heptagon defects in graphene,” Nat.
Mater., vol. 11, no. 9, pp. 759–763, 2012, doi: 10.1038/nmat3370.
[42] Z. Zhang, X. Wang, and J. D. Lee, “An atomistic methodology of energy release rate
for graphene at nanoscale,” J. Appl. Phys., vol. 115, no. 11, pp. 0–8, 2014, doi:
10.1063/1.4869207.
[43] M. A. N. Dewapriya, R. K. N. D. Rajapakse, and A. Srikantha Phani, “Molecular
dynamics simulation of fracture of Graphene,” 13th Int. Conf. Fract. 2013, ICF 2013,
vol. 1, no. June, pp. 847–852, 2013.
[44] S. J. Stuart, A. B. Tutein, and J. A. Harrison, “A reactive potential for hydrocarbons
with intermolecular interactions,” J. Chem. Phys., vol. 112, no. 14, pp. 6472–6486,
2000, doi: 10.1063/1.481208.
[45] A. . A. . Griffith, “The Phenomena of Rupture and Flow in Solids Author ( s ): A . A .
Griffith Source : Philosophical Transactions of the Royal Society of London . Series A
, Containing Papers of a Mathematical or Physical Character , Vol . 221 ( 1921 ), pp .
163-198 Publish,” Philos. Trans. R. Soc. london. Ser. A, Contain. Pap. a Math. or
Phys. character, vol. 221, no. 1921, pp. 163–198, 1920.
[46] C. E. Inglis, “Stress in a plate due to the presence of cracks and sharp corners,”
Spring Meetings of the Fifty-fourth Session of the Institution of Naval Architects. pp.
219–241, 1913.
[47] M. A. N. Dewapriya, R. K. N. D. Rajapakse, and A. S. Phani, “Atomistic and
continuum modelling of temperature-dependent fracture of graphene,” Int. J.
Fract., vol. 187, no. 2, pp. 199–212, 2014, doi: 10.1007/s10704-014-9931-y.
[48] G. I. Giannopoulos and G. S. Avntoulla, “Tensile strength of graphene versus
temperature and crack size: Analytical expressions from molecular dynamics
simulation data,” Proc. Inst. Mech. Eng. Part N J. Nanomater. Nanoeng. Nanosyst.,
vol. 231, no. 2, pp. 67–73, 2017, doi: 10.1177/2397791417712845.
[49] R. Ansari, S. Ajori, and B. Motevalli, “Mechanical properties of defective single-
layered graphene sheets via molecular dynamics simulation,” Superlattices
Microstruct., vol. 51, no. 2, pp. 274–289, 2012, doi: 10.1016/j.spmi.2011.11.019.
[50] M. A. N. Dewapriya and S. A. Meguid, “Tailoring fracture strength of graphene,”
Comput. Mater. Sci., vol. 141, pp. 114–121, 2018, doi:
10.1016/j.commatsci.2017.09.005.
[51] G. Rajasekaran and A. Parashar, “Enhancement of fracture toughness of graphene

49
via crack bridging with stone-thrower-wales defects,” Diam. Relat. Mater., vol. 74,
pp. 90–99, 2017, doi: 10.1016/j.diamond.2017.02.015.
[52] Y. Jin and F. G. Yuan, “Nanoscopic modeling of fracture of 2D graphene systems,” J.
Nanosci. Nanotechnol., vol. 5, no. 4, pp. 601–608, 2005, doi: 10.1166/jnn.2005.071.
[53] X. Huang, H. Yang, A. C. T. Van Duin, K. J. Hsia, and S. Zhang, “Chemomechanics
control of tearing paths in graphene,” Phys. Rev. B - Condens. Matter Mater. Phys.,
vol. 85, no. 19, pp. 1–6, 2012, doi: 10.1103/PhysRevB.85.195453.
[54] Akarsh Verma and Avinash Parashar, “Molecular dynamics based simulations to
study the fracture strength of monolayer graphene oxide,” pp. 0–31, 2019.
[55] K. Li, H. Zhang, G. Li, J. Zhang, M. Bouhadja, Z. Liu, et al., “ReaxFF Molecular
Dynamics Simulation for the Graphitization of Amorphous Carbon: A Parametric
Study,” J. Chem. Theory Comput., vol. 14, no. 5, pp. 2322–2331, 2018, doi:
10.1021/acs.jctc.7b01296.
[56] B. D. Jensen, K. E. Wise, and G. M. Odegard, “Simulation of the Elastic and Ultimate
Tensile Properties of Diamond, Graphene, Carbon Nanotubes, and Amorphous
Carbon Using a Revised ReaxFF Parametrization,” J. Phys. Chem. A, vol. 119, no. 37,
pp. 9710–9721, 2015, doi: 10.1021/acs.jpca.5b05889.
[57] B. D. Jensen, K. E. Wise, and G. M. Odegard, “The effect of time step, thermostat,
and strain rate on ReaxFF simulations of mechanical failure in diamond, graphene,
and carbon nanotube,” J. Comput. Chem., vol. 36, no. 21, pp. 1587–1596, 2015, doi:
10.1002/jcc.23970.
[58] T. C. Sagar, V. Chinthapenta, and M. F. Horstemeyer, “Effect of defect guided out-
of-plane deformations on the mechanical properties of graphene,” Fullerenes
Nanotub. Carbon Nanostructures, vol. 0, no. 0, pp. 1–17, 2020, doi:
10.1080/1536383X.2020.1813720.
[59] N. D. Mermin and H. Wagner, “Absence of ferromagnetism or antiferromagnetism
in one- or two-dimensional isotropic Heisenberg models,” Phys. Rev. Lett., vol. 17,
no. 22, pp. 1133–1136, 1966, doi: 10.1103/PhysRevLett.17.1133.
[60] Y. Mao, W. L. Wang, D. Wei, E. Kaxiras, and J. G. Sodroski, “Graphene structures at
an extreme degree of buckling,” ACS Nano, vol. 5, no. 2, pp. 1395–1400, 2011, doi:
10.1021/nn103153x.
[61] J. C. Meyer, A. K. Geim, M. I. Katsnelson, K. S. Novoselov, T. J. Booth, and S. Roth,
“The structure of suspended graphene sheets,” Nature, vol. 446, no. 7131, pp. 60–

50
63, 2007, doi: 10.1038/nature05545.
[62] A. Fasolino, J. H. Los, and M. I. Katsnelson, “Intrinsic ripples in graphene,” Nat.
Mater., vol. 6, no. 11, pp. 858–861, 2007, doi: 10.1038/nmat2011.
[63] V. Geringer, M. Liebmann, T. Echtermeyer, S. Runte, M. Schmidt, R. Rückamp, et al.,
“Intrinsic and extrinsic corrugation of monolayer graphene deposited on SiO2,”
Phys. Rev. Lett., vol. 102, no. 7, pp. 1–4, 2009, doi:
10.1103/PhysRevLett.102.076102.
[64] I. F. Herbut, V. Juričić, and O. Vafek, “Coulomb interaction, ripples, and the minimal
conductivity of graphene,” Phys. Rev. Lett., vol. 100, no. 4, pp. 1–4, 2008, doi:
10.1103/PhysRevLett.100.046403.
[65] F. Hao, D. Fang, and Z. Xu, “Mechanical and thermal transport properties of
graphene with defects,” Appl. Phys. Lett., vol. 99, no. 4, 2011, doi:
10.1063/1.3615290.
[66] H. Bu, Y. Chen, M. Zou, H. Yi, K. Bi, and Z. Ni, “Atomistic simulations of mechanical
properties of graphene nanoribbons,” Phys. Lett. Sect. A Gen. At. Solid State Phys.,
vol. 373, no. 37, pp. 3359–3362, 2009, doi: 10.1016/j.physleta.2009.07.048.
[67] B. J. Alder and T. E. Wainwright, “Phase transition for a hard sphere system,” J.
Chem. Phys., vol. 27, no. 5, pp. 1208–1209, 1957, doi: 10.1063/1.1743957.
[68] K. N. Dzhumagulova and T. S. Ramazanov, “Sustainable numerical scheme for
molecular dynamics simulation of the dusty plasmas in an external magnetic field,”
J. Phys. Conf. Ser., vol. 905, no. 1, 2017, doi: 10.1088/1742-6596/905/1/012022.
[69] S. Toxvaerd, Algorithms for canonical molecular dynamics simulations, vol. 72, no.
1. 1991.
[70] A. C. T. Van Duin, B. V. Merinov, S. S. Jang, and W. A. Goddard, “ReaxFF reactive
force field for solid oxide fuel cell systems with application to oxygen ion transport
in yttria-stabilized zirconia,” J. Phys. Chem. A, vol. 112, no. 14, pp. 3133–3140, 2008,
doi: 10.1021/jp076775c.
[71] K. D. Nielson, A. C. T. Van Duin, J. Oxgaard, W. Q. Deng, and W. A. Goddard,
“Development of the ReaxFF reactive force field for describing transition metal
catalyzed reactions, with application to the initial stages of the catalytic formation
of carbon nanotubes,” J. Phys. Chem. A, vol. 109, no. 3, pp. 493–499, 2005, doi:
10.1021/jp046244d.
[72] C. Ashraf and A. C. T. Van Duin, “Extension of the ReaxFF Combustion Force Field

51
toward Syngas Combustion and Initial Oxidation Kinetics,” J. Phys. Chem. A, vol.
121, no. 5, pp. 1051–1068, 2017, doi: 10.1021/acs.jpca.6b12429.
[73] A. C. T. Van Duin, S. Dasgupta, F. Lorant, and W. A. Goddard, “ReaxFF: A reactive
force field for hydrocarbons,” J. Phys. Chem. A, vol. 105, no. 41, pp. 9396–9409,
2001, doi: 10.1021/jp004368u.
[74] A. C. T. Van Duin, A. Strachan, S. Stewman, Q. Zhang, X. Xu, and W. A. Goddard,
“ReaxFFSiO reactive force field for silicon and silicon oxide systems,” J. Phys. Chem.
A, vol. 107, no. 19, pp. 3803–3811, 2003, doi: 10.1021/jp0276303.
[75] S. S. Han, J. K. Kang, H. M. Lee, A. C. T. Van Duin, and W. A. Goddard, “The
theoretical study on interaction of hydrogen with single-walled boron nitride
nanotubes. I. The reactive force field ReaxFF HBN development,” J. Chem. Phys.,
vol. 123, no. 11, 2005, doi: 10.1063/1.1999628.
[76] Q. Zhang, T. Çağln, A. Van Duin, W. A. Goddard, Y. Qi, and L. G. Hector, “Adhesion
and nonwetting-wetting transition in the Al/α−Al2O3 interface,” Phys. Rev. B -
Condens. Matter Mater. Phys., vol. 69, no. 4, pp. 1–11, 2004, doi:
10.1103/PhysRevB.69.045423.
[77] C. J. Tainter and G. C. Schatz, “Reactive Force Field Modeling of Zinc Oxide
Nanoparticle Formation,” J. Phys. Chem. C, vol. 120, no. 5, pp. 2950–2961, 2016,
doi: 10.1021/acs.jpcc.5b09511.
[78] S. Cheung, W. Q. Deng, A. C. T. Van Duin, and W. A. Goddard, “ReaxFF MgH reactive
force field for magnesium hydride systems,” J. Phys. Chem. A, vol. 109, no. 5, pp.
851–859, 2005, doi: 10.1021/jp0460184.
[79] K. F. Herzfeld and V. Griffing, “Activation energies,” Fundam. Phys. Gases, vol. 279,
no. 1931, pp. 55–68, 2015.
[80] J. Che, T. Çaǧin, and W. A. Goddard, “Studies of fullerenes and carbon nanotubes by
an extended bond order potential,” Nanotechnology, vol. 10, no. 3, pp. 263–268,
1999, doi: 10.1088/0957-4484/10/3/307.
[81] A. C. T. Van Duin and J. S. Sinninghe Damsté, “Computational chemical investigation
into isorenieratene cyclisation,” Org. Geochem., vol. 34, no. 4, pp. 515–526, 2003,
doi: 10.1016/S0146-6380(02)00247-4.
[82] N. Chen, M. T. Lusk, A. C. T. Van Duin, and W. A. Goddard, “Mechanical properties
of connected carbon nanorings via molecular dynamics simulation,” Phys. Rev. B -
Condens. Matter Mater. Phys., vol. 72, no. 8, pp. 1–9, 2005, doi:

52
10.1103/PhysRevB.72.085416.
[83] J. H. Irving and J. G. Kirkwood, “The statistical mechanical theory of transport
processes. IV. The equations of hydrodynamics,” J. Chem. Phys., vol. 18, no. 6, pp.
817–829, 1950, doi: 10.1063/1.1747782.
[84] D. H. Tsai, “The virial theorem and stress calculation in molecular dynamics,” J.
Chem. Phys., vol. 70, no. 3, pp. 1375–1382, 1979, doi: 10.1063/1.437577.
[85] K. S. Cheung and S. Yip, “Atomic-level stress in an inhomogeneous system,” J. Appl.
Phys., vol. 70, no. 10, pp. 5688–5690, 1991, doi: 10.1063/1.350186.
[86] J. Cormier, J. M. Rickman, and T. J. Delph, “Stress calculation in atomistic
simulations of perfect and imperfect solids,” J. Appl. Phys., vol. 89, no. 1, pp. 99–
104, 2001, doi: 10.1063/1.1328406.
[87] A. a a Griffith, P. Character, B. a a Griffith, M. Eng, and A. Estcblishment, “The
Phenomena of Rupture and Flow in Solids VI . The Phenomena of Rupture and Flow
in Solids . larger question of the nature of intermolecular cohesion . The original
object of the work , which was carried out at the Royal Aircraft Estab- lishment ,
was t,” Philos. Trans. R. Soc. London. Ser. A, Contain. Pap. a Math. or Phys.
Character, vol. 221, no. 1921, pp. 163–198, 1921.
[88] G. R. Irwin, “Linear fracture mechanics, fracture transition, and fracture control,”
Eng. Fract. Mech., vol. 1, no. 2, pp. 241–257, 1968, doi: 10.1016/0013-
7944(68)90001-5.
[89] J. R. Rice, “A path independent integral and the approximate analysis of strain
concentration by notches and cracks,” J. Appl. Mech. Trans. ASME, vol. 35, no. 2, pp.
379–388, 1964, doi: 10.1115/1.3601206.

53

You might also like