Integration of Graphene in Poly(Lactic) Acid by 3D

Printing to Develop Creep and Wear-Resistant

Hierarchical Nanocomposites

Jenniffer Bustillos, Daniela Montero, Pranjal Nautiyal, Archana Loganathan, Benjamin Boesl,
Arvind Agarwal
Plasma Forming Laboratory, Department of Mechanical and Materials Engineering, Florida International
University, Miami, 33174 Florida

VC 2017 Society of Plastics Engineers

Polylactic acid (PLA) and graphene reinforced
polylactic acid (PLA-graphene) composites have been
fabricated by three-dimensional (3D) fused deposition
modeling (FDM) printing. Indentation creep resistance
was analyzed in terms of the strain-rate sensitivity
index of PLA (0.11) and PLA-graphene (0.21).
Enhanced creep resistance in PLA- graphene is
attributed to the restriction of the polymeric chains by
graphene, caused by low strain rates identified during
secondary creep. The tribological properties of PLA
and PLA-graphene composites were evaluated by
ball-on-disk wear tests. Wear resistance was increased
by a 14% in PLA-graphene as compared to PLA. A
two- stage coefficient of friction (COF) behavior has
been observed for PLA-graphene. Initially, PLA-
graphene exhibits a 65% decrease in COF as
compared to PLA. During the second stage, PLA-
graphene approached similar COF behavior and
value of PLA (rv0.58). PLA- graphene composites
have shown significant improve- ment in creep and
wear resistance demonstrating 3D printing to be a
novel manufacturing route. POLYM. COM- POS., 00:000–
000, 2017. VC 2017 Society of Plastics Engineers

INTRODUCTION
The recent developments in the additive manufacturing
process such as three-dimensional (3D) printing, where
CAD-based functional structures are generated in a layer-
by-layer fashion [1], present advantages in the design and
optimization of models with complex features and tailored
mechanical properties for several applications including
orthopedic scaffolds and patient-specific implants [2].
Three-dimensional printing has proven to overcome many
of the limitations and complexities that conventional proce-
dures pose in the design of complex geometries, such as the

Correspondence to: A. Agarwal; e-mail: agarwala@fiu.edu

Contract grant sponsor: Program Manager of Low Density Material Pro-
gram (at the Air Force Office of Scientific Research); contract grant
number: W911NF-15-1-0458.
The manuscript was written through contributions of all authors. All
authors have given approval to the final version of the manuscript.
DOI 10.1002/pc.24422
Published online in Wiley Online Library (wileyonlinelibrary.com).
POLYMER COMPOSITES—2017
limited control in the customization of porosity size
distri- bution [3, 4]. A few studies [2, 5, 6] have
demonstrated the capabilities of 3D printing in the
manufacturing of bone- like structure, combining good
mechanical properties and low density. An
interconnected porosity in the printed scaf- folds will
affect the mechanical properties of the structure but will
provide the optimal environment to promote cell
adhesion and proliferation [7]. However, the
functionality of 3D printed scaffolds and implants is
highly material dependent. Hence, the selection of
material for 3D printed scaffolds and implants has
become a prominent area of research, as their
properties and behavior are affected by the type of
manufacturing technique used [5].
Polylactic acid (PLA) has been extensively studied
as a biopolymer for making orthopedic scaffolds due
to its biodegradable properties [8, 9]. Moreover, ease
of proc- essing and biocompatibility of PLA under
physiological environment makes it a suitable material for
the afore- mentioned application. PLA has also been
reinforced by nanomaterials to improve its strength and
functionality [10–13]. Graphene’s excellent mechanical,
electrical and thermal properties make it an attractive
candidate for the reinforcement of several polymers,
including PLA [14]. Due to its two-dimensional
2
arrangement of sp bonded carbon atoms, graphene has
shown to improve the wear resistance and reduction of
friction [11, 12]. Graphene’s
addition to polymer matrices has resulted in
composites exhibiting superior mechanical strength while
retaining its flexibility, as well as tailorable thermal and
electrical conductivity as a consequence of the
generated graphene network in the matrix [15, 16].
Recently, few studies have reported the successful
development of graphene-based reinforced polymer
(acry- lonitrile-butadine-styrene (ABS) and PLA)
filaments for 3D printing [17–19]. For instance, ABS-
graphene compo- sites filaments were synthesized and
extruded followed by a 3D printing process. An ideal
graphene loading of 5.6 wt% for printing is reported to
avoid clogging during extrusion due to aggregation of
graphene sheets [18].

POLYMER COMPOSITES—2017
Three-dimensional printed composites demonstrated printed dense rectangular structures is characterized by a
enhanced stiffness, improved electrical conductivity and a ball-on-disk tribometer in dry sliding condition. Micro-
slight increase in its glass transition temperature by rv4% scopic examination was conducted to obtain a deeper
[18]. In a separate study, Dul et al. [19] observed the understanding of the deformation and wear mechanism
effects on the mechanical and thermal properties of 3D experienced by 3D printed microstructure of PLA-
printed ABS-graphene nanoplatelets (GNP) as a result of graphene.
the deposition orientation. The horizontal deposition was
found to exhibit the greatest enhancement in elastic mod- EXPERIMENTAL
ulus (32%), storage modulus (23%) and an increased ther-
mal stability by 26% compared to pure ABS, as a result Materials and Processing
of the alignment of GNPs along the tensile loading direc-
tion during deposition [19]. Zhang et al. [17] reported Polylactic acid (PLA) filament was procured from
enhanced electrical properties in 3D printed flexible cir- MakerBot Industries (Brooklyn, New York). It had a
cuits of reduced graphene oxide (r-GO) reinforced PLA diameter of 1.75 mm and density of 1.25 g/cm3. PLA-
composites prepared by melt blending with compositions based graphene (PLA-graphene) filament was acquired
of 6 wt%. A 1.75 mm diameter filament for fused deposi- from BlackMagic3D (Calverton, NY) with a diameter of
tion modeling (FDM) was obtained by melt extrusion of 1.75 mm, a volume resistance of 0.6 X=cm and a density
the composite. Three-dimensional printed samples demon- of 1.77 g/cm3 [22]. Two geometries were printed: (i)
strated an increase by 37% in elastic modulus, 75% in dense rectangular strips of 85 mm in length, 10 mm
21
tensile strength and a conductivity of 4.76 S cm [17]. width, 2.5 mm thickness and (ii) scaffold structures with
The enhanced mechanical and electrical properties are dimensions of 20 mm length, 20 mm width and 3 mm
attributed to the orientation process graphene undergoes thickness. Scaffold structure had an increasing pore size
during the extrusion of the filament. In addition to the gradient of 400–850 lm across all surfaces, using FDM
solid filament, ink based feedstock has also been investi- technique on a MakerBot Replicator 2 (Brooklyn, NY).
gated for 3D printing of graphene-reinforced polylactide- All tests were performed on dense 3D printed structures.
co-glycolide (PLG) for tissue engineering applications The purpose of printed scaffold structures is to provide
[20]. Higher electrical conductivity was found to be evidence of the potential of PLA-graphene. Samples of
achieved in samples extruded with tips of diameters of both geometries were printed at a temperature of 2208C
<400 lm, as a response to the shear forces allowing and an extrusion speed of 110 mm/s. Samples were
realignment of graphene sheets during extrusion [20]. The printed with a layer height of 0.2 mm and 100% infill
biocompatibility of 3D printed scaffold was evaluated in aiming to obtain the highest printing resolution.
vivo using a mouse model, where no inflammatory Extrusion nozzles with diameters of 0.4 and 0.6 mm
response was observed during a period of 30 days [20]. were used for PLA and PLA-graphene, respectively. A
The unique combination of rapid developments in 3D nozzle with a larger diameter was used for the
printing as a processing technique and feedstock materials extrusion of PLA- graphene to prevent clogging of the
as filament or ink has led to new challenges in the funda- nozzle resulting from the larger volume of molten
mental understanding of “processing-structure-properties” material being extruded caused by the higher thermal
of 3D printed structures. The challenge becomes more conduction of graphene in the filament.
complex with varying type, and composition of reinforce-
ment (e.g., Gr, rGO, GO and graphene) added to polymer
filament due to changing nature of interfacial interactions, Thermal Properties Characterization
which in turn influence the printing, curing and properties Differential scanning calorimetry (DSC) was carried
of the polymer composite. Although biocompatibility of out for as-received PLA, and PLA-graphene filaments
graphene-reinforced PLA has been proven in a previous and 3D printed samples, to investigate the effect of gra-
study [21], its potential application as load-bearing phene on the crystallization and melting behavior of the
implants, the resultant performance under frictional loads 3D printed samples. Studies were performed by a TA
and viscoelastic behavior needs to be evaluated. Instruments SDT Q600 (New Castle, DE). Samples were
In the present study, commercially available polylactic heated from 25 to 2008C at a heating rate of 58C/min.
acid (PLA) and graphene reinforced polylactic acid (PLA- Argon gas was passed through the cell at a rate of 50 ml/
graphene) composite filaments are used to manufacture min during the test. The degree of crystallization was
scaffold structure with a graded porosity and dense rectan- determined by the following expression:
gular samples by FDM based 3D printing. The mechanical
properties of the 3D printed PLA-graphene composite and DHm2DHc
control PLA sample v e ic in
are evaluated by i l nature,
instrumented indenta- s a
tion technique. Since c s
polymers are o t

2 POLYMER COMPOSITES—2017 DOI 10.1002/pc

Xcð%Þ5 H 3100;
D o
m

time-dependent plastic where DHm represents the

deformation is probed by melting enthalpy, DHc is m
depth- the
sensing micro-scale creep cold is the
testing. The wear behavior crystallization meltin
of 3D enthalpy and g
o
DH

2 POLYMER COMPOSITES—2017 DOI 10.1002/pc

 FIG. 1. Top and cross-sectional microstructure of as-received (a) PLA and (b) PLA-graphene filaments.
[Color figure can be viewed at wileyonlinelibrary.com]

enthalpy of pure crystalline PLA, the following being D tests were carried out at the normal loads of 5, 10 and 20
Hom5 93.6 J/g [21, 23]. N using a linear speed of 16 mm/s, making a 6 mm diam-
eter track for a total period of 30 min. A non-contact
Mechanical Characterization three-dimensional optical profilometer (Nanovea, Irvine,
CA) was employed to obtain line profiles of the wear
Quasi-static nanoindentation tests were conducted to tracks. Wear depth and width were obtained by analyzing
obtain the nanomechanical properties of 3D printed PLA the wear track’s 3D profile using Scanning Probe Image
and PLA-graphene composite using a Hysitron TI900 Tri- Processor (SPIP) software (Hørsholm, Denmark). Volume
boindenter (Hysitron Inc., Minneapolis, MN). A Berko- loss during the wear test was computed by multiplying
vich diamond indenter, with tip radius of 100 nm is used. the cross-sectional area of the wear track and circumfer-
A loading and unloading period of 10 s was kept, charac- ence of the track. A minimum of three worn surfaces for
terized by elastoplastic behavior to obtain measurements each load (5, 10 and 20 N) of PLA and PLA-graphene
of reduced elastic modulus and hardness. Since polymers were obtained, respectively. Analysis to determine aver-
are viscoelastic in nature, their creep behavior is very age volume loss and depth included obtaining 5 sections
important. Microscale creep tests of 3D printed PLA and corresponding to the cross section of the worn surface.
PLA-graphene samples were carried out using a Universal Further investigation of the worn surfaces was analyzed
Surface Tester (UST) (INNOWEP, W€urzburg, by using a JEOL JSM 6330F field emission scanning
Germany). Creep test performed with a 0.8 mm electron microscopy (SEM) (Tokyo, Japan) to obtain an
diameter spherical indenter, consisted on a continuous understanding of the mechanisms controlling the tribolog-
deformation under static loadings ranging between 10 and ical behavior.
90 mN over a period of 10 s. Elastic deformation
experienced by the 3D printed specimens during creep RESULTS
test is eliminated from the curve allowing further
evaluation of creep induced deformation. Filament and 3D Printed Macro/Microstructure
The microstructure of the as-received PLA and PLA-
Tribological Characterization graphene filaments on top surface and cross-section are
shown in Fig. 1. Pure PLA shows a dense microstructure,
Wear tests were performed by using a Nanovea ball-
while PLA-graphene demonstrates a highly porous surface
on-disk tribometer (Irvine, CA) to evaluate the wear resis-
with the presence of dispersed graphene flakes embedded
tance and coefficient of friction of 3D printed PLA and
in the polymeric matrix.
PLA-graphene, using a 3 mm diameter stainless steel ball
The morphology of the 3D printed scaffolds shown in
as the counter material. Parameters of the wear studies
Fig. 2 demonstrates interconnected porous structures with
were chosen by taking into consideration the conditions
a well-defined pore size gradient as described earlier. A
the material must endure in its intended application as
gradient pore size was chosen to demonstrate the potential
orthopedic implants. In the case of load-bearing implants,
application as the orthopedic scaffold where cells can dif-
resistance to high contact pressure will be required. Wear ferentiate and proliferate as per the pore size distribution

DOI 10.1002/pc POLYMER COMPOSITES—2017 3

 FIG. 2. Graded porous structure in 3D printed (a) PLA and (b) PLA-graphene scaffold structures. [Color
figure can be viewed at wileyonlinelibrary.com]

[24]. Three-dimensional printed scaffold structures exhibit Thermal Properties

an approximate porosity of 13.6 6 3.2% (computed
DSC heating curves of as-received PLA and PLA-
using ImageJ) with a full connectivity throughout the
graphene filaments and corresponding 3D printed struc-
structure. Three-dimensional printed PLA-graphene (Fig.
tures are shown in Fig. 4. Table 1 showed a lower glass
2b) shows a rough and irregular surface compared to the
transition (Tg) temperature of both filament and printed
smooth sur- face in pure PLA (Fig. 2a), which is
PLA-graphene as compared to PLA which differs with
directly related to the porous nature of the PLA-
the literature reporting an increase in Tg with graphene
graphene observed in Fig.
addition to polymer composites [18, 19].
1. Moreover, PLA-graphene filament experiences higher
This could be attributed to the presence of unknown
cooling and nucleation rate due to graphene flakes’ higher
chemistry of the binder material in commercial PLA-
thermal conductivity which will add to a rougher surface.
graphene filament. However, 3D printed PLA-graphene
The rough surface presented by PLA-graphene in its scaf-
demonstrates an increase by 12% in its Tg compared to
fold could signify an advantage for its potential applica-
that found in the filament of the same composition. The
tion in tissue engineering by promoting greater cell
increase in Tg observed after printing can be understood
adhesion as shown in the previous study [20].
as a function of the stronger interaction between graphene
Three-dimensional printed PLA, and PLA-graphene’s
and molten PLA polymer chains during the 3D printing
top surfaces of dense strips are shown in Fig. 3a and b.
The top surface of the printed PLA reveals a layer thick- [21]. A greater nucleation rate is inferenced in samples
containing graphene from the decrease in its cold crystal-
ness of rv0.36 6 0.01 mm (Fig. 3a) with a greater
lization temperature (Tc) which directly affects the crys-
inter- layer adhesion. PLA-graphene has a higher layer
tallinity of the printed samples. A lower Tc (Table 1)
thickness of rv0.48 6 0.05 mm (Fig. 3b) PLA-
indicates that a higher crystallization will be experienced
graphene’s incomplete interlayer attachment and irregular
by PLA-graphene as a response to the graphene flakes
microstruc- ture with the presence of voids are attributed
to a combi- nation of graphene’s thermal conduction and acting as a nucleating agent. However, a decrease in the
the processing parameters chosen. As material is being crystallinity of 3D printed PLA-graphene by rv8% is
extruded through a heated nozzle and deposited in a semi- observed. A primary contributor to the decline in crystal-
solid state, a higher cooling rate will be experienced by linity is ascribed to the rapid cooling experienced by
PLA-graphene, due to graphene’s higher thermal conduc- PLA-graphene upon impinging the unheated substrate
tivity (2,000–4,000 W/m K) [25]. It results in a thermal during solidification process of 3D printing, leading to a
strain mismatch between matrix and reinforcement, lead- higher degree of polymer amorphization. During the pro-
ing to poor intralayer and interlayer bonding [26]. Cross- cess of 3D printing, the melted PLA-graphene filament
sectional fracture surfaces of PLA and PLA-graphene are experiences a rapid cooling as a response to graphene’s
shown in Fig. 3c and d. Layered structures demonstrate higher thermal conductivity. Melting of PLA and PLA-
the presence of interlayer voids. In contrast, PLA- graphene is shown in Fig. 4 for which the area bounded
graphene shows a very porous cross-section within each by its characteristic peak represents the heat absorbed
layer and the greater interlayer gap which is consistent during melting. A unimodal endothermic peak is observed
with as-received filament (Fig. 1b) and the top surface for all samples. The enthalpy of melting for 3D printed
(Fig. 3c).

4 POLYMER COMPOSITES—2017 DOI 10.1002/pc

 FIG. 3. Top surface microstructure of 3D printed (a) PLA and (b) PLA-graphene. Fractured cross-sectional
microstructure of 3D printed (c) PLA and (d) PLA-graphene. [Color figure can be viewed at
wileyonlinelibrary.com]

PLA was 58.97 J/g while that found in PLA-graphene reduces the crystallinity. The rapid cooling experienced
was 55.48 J/g. A decrease in the evaluated percent crys- by PLA-graphene restrict the fluidity and mobility for the
tallinity was observed for 3D printed PLA-graphene, con- rearrangement of polymer chains to take place resulting
sistent with the literature where the addition of graphene in reduced crystal structure formation [27].

FIG. 4. DSC heating curves of PLA and PLA-graphene (a) as-received filament and (b) 3D printed. [Color
figure can be viewed at wileyonlinelibrary.com]

DOI 10.1002/pc POLYMER COMPOSITES—2017 5

TABLE 1. Glass transition (Tg), crystallization (Tc) and melting tem-
peratures (Tm) of PLA and PLA-graphene filaments and 3D printed
samples.

%
Sample Tg (8C) Tc (8C) Tm (8C) Crystallinity

PLA filament 59.3 88.01 168.7 32.8

PLA-graphene filament 45.9 67.8 164.05 39.6
3D printed PLA 60.7 87.3 168.5 34.2
3D printed PLA-graphene 51.2 72.1 165.5 31.4

Mechanical Properties
Nano-scale hardness and reduced elastic modulus (Er)
values were obtained from nanoindentation of 3D printed
samples and are listed in Table 2. The representative
load-displacement curves from individual indents of 3D FIG. 5. Representative load–displacement curves obtained from nano-
printed PLA and PLA-graphene are shown in Fig. 5. The indentation of 3D printed PLA and PLA-graphene. [Color figure can be
viewed at wileyonlinelibrary.com]
evaluation of the nanomechanical properties through
nanoindentation, represent an idealized and small volume
Upon application of an instantaneous constant load,
of the 3D printed structures where the presence of poros-
primary creep takes place, where the material subjected
ity will have a minimal effect on the measurements.
to stress will experience a rapid deformation with a con-
Assessing the localized properties of the composite, an
18% improvement in nanohardness is obtained in PLA- tinuously decreasing strain rate [28]. It is evident by the
graphene. An increase in the elastic modulus of PLA- curve that this is achieved at a lower creep displacement
by PLA-graphene as compared to PLA. This observation
graphene by rv11% and resistance to displacement
suggests an increase in creep resistance attributed to gra-
(rv25%), suggest there is a higher resistance to plastic
phene’s impeding effect in strain hardening. Secondary
deformation in the reinforced samples. This behavior can
creep is characterized by an almost constant strain rate,
be attributed to the effective transfer of stress to the gra-
for which a combination of strain hardening and recovery
phene reinforcement. During the unloading period, PLA-
mechanisms take place [28, 29]. Figure 6b shows the
graphene can recover 43% of the elastic deformation
maximum displacements achieved during creep tests dur-
experienced, while PLA was able to recover 25%, indicat-
ing a hold period of 10 s with loads of 10, 25, 50, 75 and
ing that the presence of graphene in the filament
90 mN. It is observed that even at higher loads the addi-
decreases the permanent deformation. The improvement
tion of graphene in the filament decreases the creep dis-
in the nanomechanical properties of 3D printed PLA-
placement by around 20.5%.
graphene composite indicates that the refinement of 3D
printing processing parameters could result in further
improvement of the overall mechanical properties. Tribological Behavior
Wear behavior of 3D printed PLA and PLA-graphene
Time-Dependent Viscoelasticity was evaluated using a ball-on-disk tribometer under dry
sliding conditions. The coefficient of friction (COF) dur-
Since polymers are visco-elastic in nature and exhibit ing wear test as a function of time for loads of 5, 10 and
creep deformation, the time-dependent viscoelastic behav- 20 N are shown in Fig. 7. Pure PLA exhibited a higher
ior of the 3D printed PLA and PLA-graphene were evalu- COF in the range of 0.5–0.6, after a stabilization period
ated at the microscale length using a depth-sensing (run-in period) The initial instability in the COF curve of
universal surface tester. Figure 6a shows the creep dis- PLA is attributed to the unbalanced contact between the
placement curves of 3D printed PLA and PLA-graphene steel counter surface and 3D printed layers [26]. PLA-
under a load of 25 mN. graphene (Fig. 7b and c) experiences a significantly lower
COF (0.1–0.35) during the first 10 min of the test. After
a transition period between 10 and 15 min of the wear
TABLE 2. Reduced elastic modulus and nanohardnesses of 3D printed
PLA and PLA-graphene. tests, the curve representing the COF of PLA-graphene

starts approximating values (0.45–0.55) and approaches

Sample Er PLA 3.5 6
(GPa) 0.7 123
6 0.1
PLA-graphene 3.9 6
Nanohardness (MPa)
0.8 146

6 POLYMER COMPOSITES—2017 DOI 10.1002/pc

6 0.1 behavior similar to that
encountered in pure PLA
(Fig. 7a).
The wear resistance of
3D printed samples was
evalu- ated by obtaining
the 3D wear track profiles
of PLA (Fig.

6 POLYMER COMPOSITES—2017 DOI 10.1002/pc

 FIG. 6. (a) Creep displacement–time curves of 3D printed PLA and PLA-graphene. (b) Maximum creep
displacement at 10, 25, 50, 75 and 90 mN. [Color figure can be viewed at wileyonlinelibrary.com]

8a) and PLA-graphene (Fig. 8b) as well their respective under normal loads of 20 N revealed a wear track with
linear wear patterns (Fig. 8c). Superior wear resistance decreased average width by rv21% as opposed to PLA
was observed in PLA-graphene, where tests performed which also was characterized with a higher wear depth of

FIG. 7. COF’s of PLA (a) and PLA-graphene (b) at loads of 5, 10, and 20 N for sliding periods of 30 min.
(c) Average COF during the initial 15 min of the wear test. [Color figure can be viewed at
wileyonlinelibrary.com]

DOI 10.1002/pc POLYMER COMPOSITES—2017 7

 FIG. 8. Wear tracks profiles of 3D printed: (a) PLA (20 N, 30 min) and (b) PLA-graphene (20 N, 30 min). (c)
2D line profiles showing width and depth of the wear track obtained in Figure 8a and b. (d) Wear volume loss
comparison of 3D printed PLA and PLA-graphene. [Color figure can be viewed at wileyonlinelibrary.com]

rv0.18 mm. Wear volume loss computed from the track is allowing the rearrangement of polymer chains to take place
shown in Fig. 8d for tests conducted under a normal load [29]. A restriction of PLA’s polymer network during strain
of 20 N for PLA and PLA-graphene. The incorporation of hardening is responsible for the increase in creep resistance
graphene in the PLA based filament resulted in a 14% experienced by PLA-graphene. In the case of PLA-
enhancement in wear resistance. graphene, the presence of graphene in the matrix restrains
the movement of segments, resulting in the effective distri-
DISCUSSION bution of load to the graphene reinforcement preventing
further deformation of the polymer network. Over a small
Time-Dependent Viscoelasticity period, steady-state creep is achieved, and strain rate
Creep mechanism in 3D printed PLA and PLA- becomes almost constant achieving its minimum value. Lit-
graphene. The behavior of microscale creep-induced defor- erature has proven this region to be strain-rate dependent,
mation of PLA and PLA-graphene is highly dependent on where the response of the material to stresses will be a
its molecular structure. Due to its noncrystalline structure com- bination of that of an elastic solid and a viscous
with long entangled chains and covalent bonds, higher fluid [29, 31]. To prove the creep mechanism is occurring
chain mobility is experienced in PLA under the application in the com- posite, a stress vs. strain rate curve in the
of a load. A significant rotation and translation of chains logarithmic scale is plotted to analyze the strain-rate
are observed during the exertion of a constant load, as that sensitivity index [32]. Contact stress is measured by the
experienced during a creep test, for which deformation following formula [13]:
presents a visco-elastic behavior [30, 31]. It was noted that F
a higher creep displacement was acquired by PLA during r5 ; (2)
its transient region as compared to PLA-graphene. During Ap
primary creep, strain rate tends to decrease with time where Ap represents the projected contact area using the
known geometry of the indenter at any depth displaced.
8 POLYMER COMPOSITES—2017 DOI 10.1002/pc
 index value allows for higher level of stresses to achieve
equivalent deformations [36]. That is, the enhanced
response to creep-induced deformation exhibited by PLA-
graphene is due to the stresses becoming dependent on
the constricted chain mobility experienced. Similar behav-
ior has been observed in tensile creep testing of polymer/
graphene reinforced composites [10, 13, 37], confirming
the concluded creep mechanism in the restriction of vis-
cous flow of polymers.

Wear Behavior
To understand the wear behavior of 3D printed poly-
mer and composite, the wear tracks were examined using
SEM. Figure 10a shows a SEM image of the wear track
of 3D printed PLA. The wear track shows large sized
FIG. 9. Linear fitting of logarithmic stress–strain rate of PLA and PLA- delamination regions characteristic of adhesive wear (Fig.
graphene. [Color figure can be viewed at wileyonlinelibrary.com] 10a). Degraded polymer features illustrated in Fig. 10b
are attributed to the frictional heat generated during
For a spherical indenter, the projected area of contact is shearing of the surface. During this period, a shear-
given by [33] induced softening of the polymer takes place giving rise
to the detachment of polymer particles [38], resulting in
Ap52pRihc: (3) the transfer of clusters of debris throughout the wear track
responsible for the higher COF. In a similar manner, Fig.
Where Ri represents the radius of the spherical indenter
10c reveals the presence of finer debris on the wear track
and hc represents the contact depth. Strain rate is com-
of 3D printed PLA-graphene. The reduced size of debris
puted given the following formula [34]:
could be attributed to the easy separation of graphene
1 dh flakes from the polymer matrix aided by the low crystal-
e_ 5 : linity of PLA- Graphene
and the weak interfacial
h
bonding between PLA
d
and graphene. Figure
t
10d show the pres-
For which h represents the ence of a graphene flakes
instantaneous depth of the along the wear track,
indenter and dh=dt which contributes to the
represents the rate of enhanced tribological
change of inden- tation performance exhibited by
depth. Computation of the PLA-graphene.
strain-rate sensitivity index A contributing factor
(m) for viscoelastic in the wear volume loss
materials obeys the power differ- ence can be
law assuming a steady- attributed to the material’s
state process has been hardness. Improvement in
approximated [32, 35]: hardness of the material
has proven to increase
r5 the wear resistance as per
Cð Archard’s relation shown
e_ in Eq. 6, where dV refers
m
Þ ; to the wear volume loss,
k is wear resistance factor
where r represents the and dx is the wear distance
stress experienced by the traveled,
constant load applied over
the projected area of the
indenter, e_ is
the strain-rate, m is the strain-rate sensitivity index

DOI 10.1002/pc POLYMER COMPOSITES—2017 9

and C represents a k
dynamic modulus. Index P
dV
values obtained rep- 5
resent the capability of (6)
the material to resist H
plastic defor- mation, Taking into consideration
while denoting its the hardness values
dependence in the obtained from
molecular structure of the nanoindentation, values
specimens [32]. It can be for wear volume loss
assumed from the nature were calculates using
of the displacement-time Archard’s equation. A
curve (Fig. 6a) and the 34.2% improve- ment in
provided holding time of wear resistance is
10 s, that secondary estimated. However,
creep region achieves experi- mental results
steady-state. Linear demonstrated only a 14%
regression was car- ried improvement. High
out in stress-strain rate interlayer porosity
plots as shown in Fig. 9. present in the PLA-
It is demonstrated that graphene composite
PLA-graphene yields a contributes toward the
higher strain rate difference in wear resis-
sensitivity index tance measured. Also,
equivalent to 0.21 as com- the theoretical prediction
pared to PLA with 0.11. of wear loss as a
Higher index values function of hardness
obtained for PLA- depends on the localized
graphene are indicative nanomechanical
of the reduced plastic properties of the material
flow and stability attained obtained through
during the visco-elastic nanoindentation as
region. In analyzing the opposed to the
time-dependent properties experimental frictional
of 3D printed PLA and test, where a larger area
PLA-graphene, a larger of the sample is sub-
strain-rate sensitivity jected to wear, and
porosity will have a
higher impact in

DOI 10.1002/pc POLYMER COMPOSITES—2017 9

 FIG. 10. SEM images of (a) PLA wear track showing adhesive wear with large size debris, (b) PLA enlarged
view of delamination showing shearing features, (c) PLA-graphene worn surface showing fine debris and (d)
presence of graphene along the wear track. [Color figure can be viewed at wileyonlinelibrary.com]

the wear response. To understand the reduced COF exhib- Figure 11a shows the formation of a smooth tribofilm
ited by PLA-graphene during the first 10–15 min of the composed of graphene flakes acting as a lubricant during
wear test, the subsurface of the wear track was analyzed. the wear test. Figure 11b shows a high magnification

FIG. 11. (a) PLA-graphene subsurface with tribofilm. (b) Presence of graphene flakes in the subsurface of
PLA-graphene wear track. [Color figure can be viewed at wileyonlinelibrary.com]

10 POLYMER COMPOSITES—2017 DOI 10.1002/pc

 technique. The microstructure of 3D printed PLA-
graphene composite reveals a higher degree of intrinsic
porosity as compared to PLA due to the porous structure
of as-received filament. Three-dimensional printed PLA-
graphene resulted in a decrease in crystallinity by
8% which is attributed to the higher cooling experienced
after extrusion. PLA-graphene showed higher hardness
(146 MPa) as compared to PLA (123 MPa) in spite of
porous printed structure. Three-dimensional printed PLA-
graphene exhibited 20.5% improvement in microscale
creep resistance as compared to PLA at loads of 90 mN.
A strain-rate dependence is observed during the steady-
state region, characterizing PLA and PLA-graphene with
strain-rate sensitivity index (m) values of 0.11 and 0.21,
respectively. A higher index (m) value for PLA-graphene
denotes its superior stress-bearing capabilities while
achieving smaller deformations as compared to PLA. Dry
FIG. 12. Normal pressure against time during the wear test for PLA- sliding wear tests revealed a two-stage COF behavior as a
graphene under normal loads of 20 N. [Color figure can be viewed at
result of the self-lubricating mechanism of PLA-graphene.
wileyonlinelibrary.com]
It is demonstrated that under high normal stresses during
wear test, graphene flakes will dissociate from the com-
image of graphene flakes present in the subsurface. A
posite’s matrix resulting in lower COF of rv0.1–0.35 com-
combination of graphene’s mechanical strength along the
pared to PLA with rv0.45. The higher hardness and the
planar direction and easy shear capability among layers
contributed to the improved wear resistance and reduction presence of graphene in the planar direction contribute to
in COF. the improved wear resistance of PLA-graphene by a 14%
as compared to pure PLA. The promising results in the
The improved mechanical and tribological properties
present study prove the feasibility of 3D printed PLA-
observed in graphene reinforced polymer composites has
graphene for potential use as scaffolds for orthopedic
been attributed to two primary observations, limited poly-
applications.
mer chain mobility and efficient load transfer to the rein-
forcement [14, 19]. To further understand the two-stage
ACKNOWLEDGMENTS
COF behavior of PLA-graphene, the normal stress exerted
during the wear tests under 20 N load is plotted as a The authors would like to acknowledge Dr. Joycelyn C.
function of time (Fig. 12). Harrison, Program Manager of Low Density Material Pro-
The normal stress applied to the surface in contact is gram at the Air Force Office of Scientific Research (grant
computed by dividing the normal load by the effective con- W911NF-15–1-0458). Pranjal Nautiyal thanks Florida
tact area between the 3 mm stainless steel ball and the com- Inter- national University Graduate School for the
posite at any given point in time. It was noted that with Presidential Fellowship Award. Advanced Materials
greater periods of time, the stress exerted in the composite Research Institute (AMERI) at FIU is also acknowledged
lowered due to having a higher contact area. Moreover, a for the research facil- ities used in this work.
larger amount of stress was encountered during the initial
10–15 min of the sliding test characterized by the decreased REFERENCES
COF. It is proposed that during this period, the rise in tem-
perature in the sliding interface along with the normal 1. T. Galeta, P. Raos, J. Stoj�sic´, and I. Pak�si, Proc.
Eng., 149, 100 (2016).
stress exerted by the counter material in the composite
was high enough to dissociate graphene flakes from the 2. S. Barui, S. Chatterjee, S. Mandal, A. Kumar, and B. Basu,
Mater. Sci. Eng. C, 70, 812 (2016).
polymer matrix [38, 39], resulting in the formation of a
graphene tri- bofilm, acting as a protective layer to the 3. A. Kishore, R. Begum, and G. Arumaikkannu, IOSR J.
Mech. Civ. Eng., 1, 2278 (2014).
worn surface and as lubrication. After 15 min of the
4. R. Begum and G. Arumaikkannu, Biomed. Res., 26, 29 (2015).
duration of the test, the stress exerted was not sufficient to
continue with the lubri- cation mechanism causing the 5. X. Wang, S. Xu, S. Zhou, W. Xu, M. Leary, P. Choong, M.
Qian, M. Brandt, and Y. Min Xie, Biomaterials, 83, 127
sudden increase in COF dominated by pure PLA wear.
(2016).
6. F. Libonati, G.X. Gu, Z. Qin, L. Vergani, and M.J. Buehler,
CONCLUSIONS
Adv. Eng. Mater., 18, 1354 (2016).
PLA-graphene composite and PLA scaffold structures 7. B. Dhandayuthapani, Y. Yoshida, T. Maekawa, and D.
were fabricated using 3D fused filament printing Sakthi Kumar, Int. J. Polym. Sci., 2011, 1 (2011).

DOI 10.1002/pc POLYMER COMPOSITES—2017 11

 8. J.C. Middleton and A.J. Tipton, Biomaterials, 21, 2335 25. E. Pop, V. Varshney, and A.K. Roy, MRS Bull., 37, 1273
(2000). (2012).
9. C.M. Agrawal, G.G. Niederauer, and K.A. Athanasiou, Tis- 26. A. Tsouknidas, Adv. Tribol., 2011, 1 (2011).
sue Eng., 1, 241 (1995). 27. E. Tarani, A. Wurm, C. Schick, D.N. Bikiaris, K.
10. A. Bhattacharyya, S. Chen, and M. Zhu, eXPRESS Polym. Chrissafis, and G. Vourlias, Thermochim. Acta, 643, 94
Lett., 8, 74 (2014). (2016).
11. D. Lahiri, F. Hec, M. Thiesse, A. Durygin, C. Zhang, and 28. J. Vincent, “Basic elasticity and viscoelasticity,” in Struc-
A. Agarwal, Tribol. Int., 70, 165 (2014). tural Biomaterials, V. Julian, Ed.,. Princeton University
12. H.S. Dongn and S.J. Qi, Biosurf. Biotribol., 1, 229 Press, Princeton, NJ, 1 (2012).
(2015). 29. M.A. Meyers and K.K. Chawla, “Creep and superplasticity,”
13. M. Tehrani, M. Safdari, and M.S. Al-Haik, Int. J. Plast., 27, in Mechanical Behavior of Materials, Cambridge University
887 (2011). Press, New York, 653 (2008).
14. I.A. Ovidko, Rev. Adv. Mater. Sci., 34, 1 (2013). 30. T.A.P. Engels, S.H.M. S€ontjens, T.H. Smit, and L.E.
15. K. Hu, D.D. Kulkarni, I. Choi, and V.V. Tsukruk, Prog. Govaert, J. Mater. Sci. Mater. Med., 21, 89 (2009).
Polym. Sci., 39, 1934 (2014). 31. E. Jacobus, “Intrinsic deformation behaviour: Stress acti-
16. G. Mittal, V. Dhanda, K. Yop Rhee, S.J. Park, and W.R. vated mobility,” in Mechanical Properties of Solid Poly-
Lee, J. Ind. Eng. Chem., 21, 11 (2015). mers: Constitutive Modelling of Long and Short Term
17. D. Zhang, B. Chi, B. Li, Z. Gao, Y. Du, J. Guo, and J. Wei, Behaviour, Technische Universiteit Eindhoven, Eindhoven,
Synth. Met., 217, 79 (2016). The Netherlands, 4 (2005).
18. X. Wei, D. Li, W. Jiang, Z. Gu, X. Wang, Z. Zhang, and Z. 32. D.L. Goble and E.G. Wolff, J. Mater. Sci. Mater. Med., 28,
Sun, Sci. Rep., 5, 1 (2015). 5986 (1993).
19. S. Dul, L. Fambri, and A. Pegoretti, Composites Part A, 85, 33. A. Fischer-Cripps, “Nanoindentation testing,” in Nanoinden-
181 (2016). tation, Springer, New York, NY, 21 (2011).
20. A.E. Jakus, E.B. Secor, A.L. Rutz, S.W. Jordan, M.C. 34. P. Nautiyal, J. Jain, and A. Agarwal, Mater. Sci. Eng. A,
Hersam, and R.N. Shah, ACS Nano, 9, 4636 (2015). 630, 131 (2015).
21. A.M. Pinto, C. Gonc¸alves, I.C. Gonc¸alves, and F.D. 35. R.A. Vasin, F.U. Enikeev, and M.I. Mazurski, J. Mater. Sci.
Magalh~aes, Eur. Polym. J., 85, 431 (2016). Eng. A, 224, 131 (1997).
22. Graphene 3D Lab Inc. (Ed), 3D, BlackMagic. Conductive 36. M.A. Nazzal, M.K. Khraisheh, and F.K. Abu-Farha, J.
Graphene Filament (Msds), Graphene 3D Lab Inc. Calver- Mater. Process. Technol., 191, 189 (2007).
ton, New York, USA (2015). 37. L.C. Tang, X. Wang, L.X. Gong, K. Peng, L. Zhao, Q.
23. P. Nanthananon, M. Seadan, S. Pivsa-Art, and S. Chena, L.B. Wu, J.X. Jiang, and G.Q. Lai, Compos. Sci.
Suttiruengwong, IOP Conf. Ser. Mater. Sci. Eng., 87, 1108 Technol., 91, 63 (2014).
(2015). 38. B.J. Briscoe and S.K. Sinha, Proc. Inst. Mech. Eng., Part J,
24. A. Di Luca, B. Ostrowska, I. Lorenzo-Moldero, A. 216, 401 (2001).
Lepedda, W. Swieszkowski, C. Van Blitterswijk, and L. 39. C. Zhang, A. Nieto, and A. Agarwal, Nanomater. Energy, 5,
Moroni, Sci. Rep., 6, 1 (2016). 1 (2016).

12 POLYMER COMPOSITES—2017 DOI 10.1002/pc