Composites Part B 184 (2020) 107625

Contents lists available at ScienceDirect

Composites Part B
journal homepage: www.elsevier.com/locate/compositesb

Multifunctional performance of carbon nanotubes and graphene

nanoplatelets reinforced PEEK composites enabled via FFF
additive manufacturing
M.F. Arif a, f, H. Alhashmi a, K.M. Varadarajan b, c, Joseph H. Koo d, A.J. Hart e, S. Kumar a, *
a
Department of Mechanical Engineering, Khalifa University of Science and Technology, Masdar Campus, Masdar City, P.O. Box 54224, Abu Dhabi, United Arab
Emirates
b
Department of Orthopaedic Surgery, Harris Orthopaedics Laboratory, Massachusetts General Hospital, 55 Fruit St, Boston, USA
c
Department of Orthopaedic Surgery, Harvard Medical School, A-111, 25 Shattuck Street, Boston, USA
d
Department of Mechanical Engineering, University of Texas at Austin, Austin, TX, 78712, USA
e
Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA, 02139, USA
f
Department of Materials Engineering, Institut Teknologi Sumatera, South Lampung, 35365, Indonesia

A R T I C L E I N F O A B S T R A C T

Keywords: The study is focused on multifunctional performance of carbon nanotubes (CNT) and Graphene nanoplatelets
PEEK nanocomposites (GNP) reinforced PEEK composites enabled via fused filament fabrication (FFF) additive manufacturing (AM)
Fused filament fabrication (FFF) utilizing in-house nanoengineered filaments. Thermo-physical, mechanical and wear characteristics of electro-
Carbon nanostructures
conductive PEEK nanocomposites are reported. The coefficient of thermal expansion (CTE) is found to
Coefficient of thermal expansion (CTE)
Wear
decrease by 26% and 18% with the incorporation of 5 wt% GNP and 3 wt% CNT into PEEK polymer, respectively.
The decrease in CTE provides better dimensional stability to resulting nanocomposite structures. Due to uniform
dispersion of CNT and GNP in the PEEK matrix, the crystallization temperature and degree of crystallinity are
both increased. The 3D printed PEEK nanocomposites reveal interfacial voids between the beads and intra-bead
pores and thus exhibit lower density compared to that of the 3D printed neat PEEK. Young’s and storage moduli
are found to increase by 20% and 66% for 3 wt% CNT loading and by 23% and 72% for 5 wt% GNP loading
respectively. However, the PEEK nanocomposites exhibit similar tensile strength to that of neat PEEK. The co­
efficient of friction obtained from fretting wear tests is found to decrease by 67% and 56% for 1 wt% CNT and 3
wt% GNP loaded PEEK nanocomposites, respectively and the decrease is attributed to reduced hardness and
increased porosity. Multifunctional performance of carbon nanostructures reinforced AM-enabled PEEK com­
posites demonstrated here makes them suitable for a range of applications such as orthopedics, oil and gas,
automotive, electronics and space.

1. Introduction on to a heated plate to form a 3D part, following a predefined 3D slicing

pattern.
Additive manufacturing (AM) enables fabrication of complex ge­ Despite abundant availability of FFF machines, only a few are suit­
ometries through layer-by-layer formation and is rapidly advancing to­ able for 3D printing high performance polymers, such as polyether ether
wards new materials and applications [1,2]. AM offers user friendly and ketone (PEEK), polyetherimide (PEI), etc., as they process at high tem­
low-cost fabrication route that provides greater flexibility to locally perature. These polymers exhibit superior mechanical properties as well
design the material architecture in three-dimensions [3–6]. Fused fila­ as high thermal and chemical resistance [7]. Among the high perfor­
ment fabrication (FFF), also known as fused deposition modeling (FDM) mance polymers, PEEK, a semi-crystalline polymer is suitable for many
is the most popular AM technique for thermoplastics due to its relative engineering applications, such as biomedical implants, due to its
simplicity and low-cost. In FFF process, the filament feedstock is fed into biocompatibility, corrosion resistance, radiopacity, excellent fatigue
a heated nozzle where the polymer is fused and subsequently deposited and wear resistance [8]. PEEK is also utilized for space applications as

* Corresponding author.
E-mail address: [email protected] (S. Kumar).

https://doi.org/10.1016/j.compositesb.2019.107625
Received 10 June 2019; Received in revised form 21 August 2019; Accepted 25 November 2019
Available online 27 November 2019
1359-8368/© 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
M.F. Arif et al. Composites Part B 184 (2020) 107625

PEEK has low outgassing property and lightweight relative to metals [9]. 2. Materials and methods
PEEK and PEEK composites are also used in oil and gas industries for
gaskets and seals. These components interact with potentially harsh 2.1. Materials
environments, such as extreme temperatures and high pressures, cor­
rosive fluids and gases, lubricating oils and solvents [10]. The draw­ The KetaSpire® KT-880 high flow, injection molding grade PEEK
backs of PEEK are higher cost (compared to other commodity (Solvay Specialty Polymers, www.solvay.com) was used as the matrix.
thermoplastics), high processing temperature (above 350 � C) and the This is a medical grade PEEK supplied in a lubricated pellet form, with
processing difficulties associated with its semi-crystallinity. FFF tech­ 0.01% calcium stearate lubricant to aid the pellet flow through plasti­
nique has been successfully utilized to fabricate PEEK [11–17]. The cisation screws. Two carbon nanostructures were used as nano-
expensive selective laser sintering (SLS) AM technique has also been reinforcements: Graphistrength™ C100 multiwall CNT (Arkema, Lacq,
utilized to fabricate PEEK [18]. France; www.graphistrength.com) and Grafmax HC 11 GNP (Nacional
Precursor materials (e.g., filament, mixed photo-resin) containing de Grafite, San Paulo, Brazil; www.grafite.com). Graphistrength® C100
micro and/or nano-fillers, such as carbon fiber [19,20], carbon black MWCNT is a black powder with an apparent density of 50–150 kg/m3,
[21], GNP [22,23], CNT [24], montmorillonite [25], stainless steel, mean agglomerate size of 200–500 μm, C content greater than 90 wt%,
bronze and copper powders [26,27], glass microspheres and nanoclay mean number of walls of 5–15, an outer mean diameter of 10–15 nm,
[28], and wood flour [27] are increasingly being developed to realize and length about 0.1–10 μm [43]. Grafmax® HC 11 GNP is a black
AM-enabled composites. The main idea is to harness the synergistic ef­ powder with a C content great than 99.5 wt%. It has a D32 ¼ 6.99 μm,
fects of their constituents and their tailorable microstructure to realize and a platelet thickness from about 0.34 to 100 nm [43].
lightweight composites with superior mechanical, electrical, thermal
and coefficient of thermal expansion (CTE) properties. Strong parts with 2.2. Filament fabrication
low CTE are desirable to ensure dimensional stability. However, typical
polymers possess high CTE. It was found that the PEEK CTE can be To manufacture filament of an appropriate size for use in FFF ma­
reduced by the incorporation of low CTE and high thermal conductivity chines, a filament extrusion system consisting of a twin-screw extruder,
Al2O3 particles [29]. Additionally, polymers reinforced with electrically water bath, a customized filament puller, and a winding device was
conductive nanofibers, such as GNP and/or CNT have also been explored setup as shown in Fig. S1. To achieve precise speed control, the filament
for their functional performance, such as for temperature, strain and puller consists of a potentiometer, an Arduino controller, and a stepper
damage-sensing [30–34]. Stepashkin et al. [19]. reported that Carbon motor. For extrusion, a Thermo Scientific Process 11 parallel co-rotating
fiber-reinforced PEEK composites fabricated by FFF process exhibit twin-screw extruder was used to melt compound the CNT/PEEK (1 and
25–30% lower thermal conductivity than cast CF/PEEK composites. 3 wt%) and GNP/PEEK (3 and 5 wt%). To ensure a homogenous
Anisotropy of thermal properties was observed, as the short carbon fi­ dispersion, formulations with varying concentrations were stir mixed
bers get oriented along the FFF deposition path. Berretta et al. [35]. prior to melt compounding. It is interesting to note that the GNP/PEEK
fabricated PEEK nanocomposites reinforced with 1 wt% and 5 wt% CNT compound with 3–5 wt% GNP and CNT/PEEK compound with 1 wt%
by FFF process and found that the incorporation of CNT into the PEEK CNT showed good melt processability, whereas CNT/PEEK compound
matrix, does not influence the mechanical performance of the PEEK with 3 wt% CNT was very viscous. Viscous compound, due to its un­
nanocomposites. Gonçalves et al. [36]. fabricated PEEK nanocomposite stable melt flow, makes the filament diameter non-uniform. This in turn
filaments reinforced with CNT/GNP. The resulting nanocomposite fila­ influences the printability of filaments and the properties of printed
ments exhibit electrical conductivity between 1.5 and 13.1 S/m. parts. The weight fraction of carbon nanostructures is chosen such that
The reinforcement of PEEK with CNT and GNP also influences the they form electrically percolating network in the resulting nano­
tribological characteristics of resulting PEEK composites. For example, composite (electrical conductivity σ ¼ 9.1 � 10 7 S/cm for 1 wt% CNT)
addition of GNP into PEEK was found to reduce the coefficient of friction but still small enough for the nanocomposite to remain biocompatible.
by 38% of the nanocomposite due to the lubrication action of GNP [37].
In another study, incorporation of CNT into PEEK matrix was found to 2.3. FFF fabrication
result in a ~24% reduction in the friction compared to the neat PEEK
[38]. The microscale wear (known as fretting wear) test is usually per­ Indmatec (Apium) HPP 155 3D printer was used to fabricate the
formed to assess the tribological performance of PEEK and PEEK nano­ specimens. Neat and nanoengineered PEEK filaments with an average
composites used especially for artificial joints in orthopedics and diameter of 1.75 mm were used as feedstock. The deposition path and
mechanical bearings [38]. Artificial joints, experience microscale FFF parameters were defined using Simplify3D (a 3D printing slicer
oscillatory movements between the mating surfaces [39–41]. Fretting software). Nozzle movement speed of 1000 mm/min, nozzle tempera­
wear analysis can predict the possibility of loosening of the implant due ture of 390 � C, bed temperature of 100 � C, layer height of 0.1 mm,
to wear at the joints [42]. extrusion width of 0.48 mm, and infill density of 100% were selected as
Either most of the aforementioned studies focused on conventional the FFF process parameters. The specimens were fabricated in two
fabrication of PEEK nanocomposites or 3D printed composites utilizing different print orientations, namely, horizontal and vertical, as
filaments produced by third party. In this study, we developed carbon described in Fig. 1. One specimen of horizontal print orientation, and
nanostructures (CNT and GNP) engineered PEEK filaments and fabri­ four specimens of vertical print orientation were produced per fabrica­
cated electro-conductive PEEK nanocomposites enabled via FFF AM. We tion cycle. Different number of specimens were selected to ensure a
investigated the effect of GNP (3 and 5 wt%) and CNT (1 and 3 wt%) on successful fabrication, as discussed in our recent work [17].
multifunctional properties of PEEK filaments and their corresponding
FFF-PEEK nanocomposites. The melting temperature, degree of crys­ 2.4. Characterization
tallinity, and density of the filament feedstocks are analyzed. The den­
sity, CTE, tensile properties, dynamic mechanical properties, and Differential scanning calorimetry (DSC) measurements (Netzsch)
surface morphology of the PEEK nanocomposites are then investigated. were performed under nitrogen atmosphere to study the melting and
The Fretting wear performance of PEEK, CNT/PEEK and GNP/PEEK crystallization behavior of neat PEEK and in-house nanoengineered
composites is assessed against stainless steel counter body. The multi­ PEEK filaments. The samples were heated from 30 � C to 430 � C at a
functional performance of 3D printed GNP/PEEK and CNT/PEEK heating rate of 10 C/min and then maintained isothermally at 430 � C
�

electro-conductive composites demonstrated in this study signify their for 5 min before cooling the samples to 30 � C at the same rate of 10 C/
�

potential for several structural and functional applications. min. Subsequently after soaking isothermally for 5 min at 30 C, the
�

2
M.F. Arif et al. Composites Part B 184 (2020) 107625

Fig. 1. Additive manufacturing of PEEK and

PEEK nanocomposites by FFF. The CTE
specimens were fabricated either a) hori­
zontally with 0� raster angle or b) vertically
with 90� raster angle. Insert (a,b) shows the
magnified view of the deposition path or
raster angle. Configuration of the specimen
for DMA characterization is similar to that of
CTE specimen. c) Horizontally fabricated
dogbone specimens were used for tensile
testing (also see Fig. S2). Complex 3D prin­
ted scaffold geometries of d) neat PEEK and
e) CNT/PEEK.

samples underwent a 2nd heating cycle. The melting temperature and strains over the gauge length zone of the specimen. We tested only
heat of fusion data were obtained from the 2nd heating cycle to elimi­ horizontally fabricated samples, with three repeats for statistical pur­
nate the influence of any thermal history and for a better surface contact pose. Vertically fabricated samples were not considered as this config­
between the solid sample and the DSC crucible. The degree of crystal­ uration leads to poor mechanical performance [17]. Prior to testing,
linity Xc of the samples was calculated using Xc ¼ ΔHc � 100= ðΔH0c wÞ, samples were annealed at 200 � C for 3 h to remove thermal history.
where ΔHc is the heat of crystallization and ΔH0c is the heat of crystal­ Fractured surfaces of the tensile specimens were examined using Nova
lization of 100% crystalline PEEK (130 J/g [44]), and w is the mass NanoSEM to investigate the morphology of the samples and the
fraction of PEEK in the nanocomposites. dispersion quality of CNT and GNP in the PEEK matrix. The samples
Density of the filament feedstocks and the FFF-samples were deter­ were mounted on a metal stub and gold sputtered to reduce the charging
mined using a gas (Helium) displacement pycnometry system - AccuPyc effects.
II 1340, Micromeritics. Density of FFF-samples made for CTE measure­ Glass transition temperatures (Tg), storage modulus, and tan δ of the
ments was measured. Ten repeated measurements were obtained for PEEK nanocomposites were measured using a dynamic mechanical
each sample type. The CTE of the samples was determined using Dila­ analyzer (DMA NETZSCH-404 F1 Pegasus®). A bending mode testing
tometer DIL 802 from Bahr Thermoanalyse. The temperature was varied was applied during the DMA scans, and the scanning range was from 30
to 250 � C at a heating rate of 3 C/min and at a frequency of 1 Hz under
�

from 30 to 200 � C with a scan rate of 5 C/min. Cylindrical samples with

�

a diameter of 5 mm and length of 50 mm were analyzed (see Fig. S3). nitrogen atmosphere. Each sample was tested at least twice to ensure
The CTE was estimated from the relationship: ΔL=L0 ¼ CTE* ΔT, where repeatability.
ΔL=L0 is the average thermal strain along the length of the sample and Bulk hardness of the processed samples was analyzed using Vickers’s
ΔT is the change in temperature. Prior to testing, samples were annealed hardness tester (BAREISS-VTEST; Bareiss Prüfgera €tebau, Germany) by
at 200 � C for 3 h to remove thermal history. applying 200-g load (n ¼ 3) with a dwell time of 10 s. The diagonals of
Room temperature tensile tests were performed on a (2.5 kN load the Vickers indents were then measured using an optical microscope to
cell) Zwick-Roell Z005 machine, following ISO 527 standard with a find out the hardness number. To investigate wear performance of the
crosshead speed of 1 mm/min. Digital image correlation (DIC) tech­ samples, the Fretting wear tests were performed using a ball on a flat
nique was used to obtain the distribution of axial and lateral engineering configuration (Reciprocating Friction and Wear Monitor TR 281 M

3
M.F. Arif et al. Composites Part B 184 (2020) 107625

fretting wear testing machine; Bangalore, India). AISI E52100 stainless Table 1
steel ball with a diameter of 6 mm was used as counter body against The DSC melting and crystallization data of neat and CNT/GNP reinforced PEEK
fretting of PEEK and PEEK nanocomposites. All the samples were tested composites.
at a frequency of 5 Hz and an applied load of 10 N with an amplitude of Tm (oC) Tc (oC) ΔHm (J/g) ΔHc (J/g) Xc (%)
100 μm for 10,000 cycles. The wear scar was observed using an optical
PEEK 342 298 29.2 40.8 31
profilometer (Bruker, contour G T, GTKOX-14-150) to quantify wear GNP/PEEK (3 wt%) 343 302 28.1 41.5 33
volume and wear rate. The wear rate, W (mm3/Nm) was then calculated GNP/PEEK (5 wt%) 343 301 27.4 39.7 32
V
using the relationship: W ¼ PλS, where, V is the wear volume in mm3, P is CNT/PEEK (1 wt%) 344 303 28.9 42.4 33
the load in N, λ is the amplitude of displacement in m and S is the CNT/PEEK (3 wt%) 344 304 27.6 39.0 31

number of cycles.
CNT/PEEK samples at low (contain several beads) and high magnifica­
3. Results and discussion tion, respectively. Fig. 3a shows smooth, brittle-like fracture surface of
the base PEEK samples. The interfaces between beads of the base PEEK
3.1. DSC are not well identifiable, signifying a relatively good adhesion between
beads. CNT/PEEK specimen shows surface morphology similar to that of
DSC was performed on filament feedstocks (see, Fig. 2). The melting the base PEEK, but with more visible interfaces between beads (Fig. 3c).
temperature (Tm), crystallization temperature (Tc), heat of fusion ðΔHm Þ, This indicates that CNT/PEEK sample has more interfacial voids be­
heat of crystallization ðΔHc Þ; and degree of crystallinity ðXc Þ of neat tween the beads compared with base PEEK and thus lower interfacial
PEEK and the PEEK nanocomposite filaments were analyzed (see, strength. GNP/PEEK sample (see, Figs. 3b and 4b) exhibits a
Table 1). The CNT and GNP reinforced PEEK exhibit a shift in melting morphology different to those of base PEEK and CNT/PEEK, indicating
and crystallization peaks to a higher temperature compared to the un­ that the addition of GNP promotes ductility of the nanocomposite. Fig. 4
reinforced PEEK. The Tm is increased by 1 � C and 2 � C with the addition shows high magnification SEM images of the base PEEK, GNP/PEEK,
of GNP and CNT, respectively. The Tc shows higher temperature shift and CNT/PEEK. It can be seen that CNT and GNP are uniformly
than Tm as the Tc increases by 3–4 � C and 5–6 � C with addition of GNP dispersed in the PEEK matrix. Moreover, GNP/PEEK composite has large
and CNT, respectively. It indicates that the presence of CNT and GNP in extent of micro and nano-pores. It is worth noting that in all samples,
PEEK facilitates PEEK crystallization. This is further confirmed by the voids with diameters ranging from 50 to 150 μm are observed (see SI,
increase in degree of crystallinity of GNP/PEEK with 3 wt% GNP and Fig. S4). This is commonly observed in FFF process, arising due to issues,
CNT/PEEK with 1 wt% CNT. The nano-reinforcements serve as nucle­ such as non-uniformity of the filament diameter and non-uniform flow
ating agents permitting molecular chains of PEEK to pack into a closer of polymer melt.
arrangement [45,46]. Higher loading of nano-reinforcements
(GNP/PEEK with 5 wt% GNP and CNT/PEEK with 3 wt% CNT) exhibit
3.3. Density
negligible increase in degree of crystallinity. This could be due to the
agglomeration of CNT and GNP. Agglomeration reduces the effective
It can be seen from Table 2 that the density of the filament feedstock
contact surface area between the nano-reinforcements and the matrix.
slightly increases with increasing CNT and GNP content. This is due to
higher density of CNT (1.6 g/cm3) and GNP (2.267 g/cm3) than that of
3.2. SEM morphology the PEEK. All FFF-fabricated nanocomposites exhibit lower density than
those of the filament feedstocks. The reduced density of FFF-samples is
Fracture surfaces of tensile specimens were examined to investigate due to the formation of multiscale voids both within the beads and at the
the morphology and dispersion state of the GNP and CNT in PEEK ma­ interfaces between the beads during layer-by-layer build-up of the
trix. Figs. 3 and 4 show the SEM images of base PEEK, GNP/PEEK, and samples. It is interesting to note that horizontally fabricated samples

Fig. 2. DSC curves for the neat PEEK and PEEK nanocomposites.

4
M.F. Arif et al. Composites Part B 184 (2020) 107625

Fig. 3. SEM fracture surface morphology of a) base PEEK, b) GNP/PEEK specimen with 5 wt% GNP and c) CNT/PEEK specimen with 3 wt% CNT. The raster angle
and loading direction are normal to the surface.

exhibit lower density than that of vertically fabricated samples. This the specimen. Significant difference in density is particularly observed
difference is attributed to the difference in thermal histories of the two for reinforced nanocomposites. For similar FFF process parameters such
configurations, as discussed in our recent work [17]. as the nozzle movement speed, nozzle temperature and layer height, the
The polymer melt in vertically fabricated sample is deposited onto a mass flow rate of material dispensed through the nozzle for nano-
small cross-sectional area, as seen in Fig. 1. Therefore, the deposition engineered filaments is lower than that for neat PEEK filament. This is
time in the x y plane is very short prior to moving in the z -direction. due to the increased viscosity and change in flow behavior of fused
This allows a good degree of compaction of the sample, resulting in nanocomposite. Therefore, the extent of interfacial voids between beads
higher density, since the polymer is deposited onto a layer that is still in of the nanocomposites is higher than that of the base PEEK. This leads to
a molten or partially molten state. The deposition time in the x y plane reduced density of the nano-reinforced PEEK composites compared to
of the horizontally fabricated sample is more than that of the vertically base PEEK. Moreover, presence of micro and nano-pores within beads
fabricated sample. Therefore, it is likely that the bead is deposited on to are prevalent especially in GNP/PEEK composites, contributing to
a solidified preceding layer, resulting in lower degree of compaction of reduced density of the nano-reinforced PEEK composites.

5
M.F. Arif et al. Composites Part B 184 (2020) 107625

Fig. 4. High magnification micrographs of tensile fracture images of 3D printed a) base PEEK, b) GNP/PEEK composite with 5 wt% GNP, and c) CNT/PEEK
composite with 3 wt% CNT. The raster angle and loading direction are normal to the surface.

3.4. CTE of PEEK nanocomposites fabrication configurations. One can see that for temperature below Tg,
the CTE of PEEK nanocomposite decreases by 16% and 26% with the
The thermal expansion behavior of the samples is influenced by addition of 3 wt% GNP and 5 wt% GNP, respectively. Incorporation of
CNT/GNP content, as well as the fabrication configuration. In this study, CNT into the PEEK matrix reduces the CTE as well but to a lesser extent.
we measure an average CTE considering thermal expansion only along The addition of 1 wt% and 3 wt% CNT reduces the CTE of PEEK nano­
the longitudinal direction. As seen from Fig. 5, the specimen elongates as composites by 9% and 18%, respectively. The CTE of the nanocomposite
the temperature is constantly increased. Two distinct regions, separated above Tg is also evaluated. It should be noted that the mechanical per­
by change in slope of strain ΔL=L0 vs. temperature increase ΔT are formance of the reinforced PEEK nanocomposites becomes poor in this
observed. This corresponds to the temperature below and above Tg. The temperature regime, with a storage modulus of ~10 � lower than that
CTE was estimated by evaluating the slope of ΔL=L0 vs. ΔT, below and below Tg (see, Fig. 6). The extent of decrease in CTE of the samples with
above Tg. Table 3 summarizes CTE of neat PEEK and PEEK composites GNP remains the same relative to that below Tg. However, CNT incor­
with CNT and GNP reinforcements for both horizontal and vertical porated PEEK specimens above Tg do not exhibit significant CTE changes

6
M.F. Arif et al. Composites Part B 184 (2020) 107625

Table 2 nanocomposites increases with increase in nano-reinforcement content.

Density of the filament feedstocks and FFF-fabricated base PEEK and PEEK However, the reinforced nanocomposites exhibit similar tensile strength
nanocomposites. to that of the base PEEK. The GNP/PEEK nanocomposites exhibit ductile
Density (g/cm3) fracture behavior while CNT/PEEK nanocomposites exhibit brittle
Filament PEEK 1.30 � 0.0009
fracture behavior, as discussed earlier in SEM morphology section. This
GNP/PEEK (3 wt%) 1.31 � 0.0003 is reflected in the strain at break and tensile toughness properties of the
GNP/PEEK (5 wt%) 1.33 � 0.0014 nanocomposites. Compared to base PEEK, the GNP/PEEK nano­
CNT/PEEK (1 wt%) 1.30 � 0.0014 composites exhibit higher strain at break and toughness properties,
CNT/PEEK (3 wt%) 1.31 � 0.0017
while the CNT/PEEK nanocomposites show decreased performance. The
PEEK Fabricated Horizontally 1.28 � 0.0016
Fabricated Vertically 1.30 � 0.0006 Poisson’s ratio of the base PEEK and CNT/PEEK nanocomposite samples
GNP/PEEK (3 wt%) Fabricated Horizontally 1.15 � 0.0006 are similar. However, GNP/PEEK nanocomposite exhibits lower Pois­
Fabricated Vertically 1.27 � 0.0009 son’s ratio. This could be attributed to the greater extent of micro-voids
GNP/PEEK (5 wt%) Fabricated Horizontally 1.17 � 0.0005 in GNP/PEEK samples, as indicated by SEM image presented in Fig. 4b.
Fabricated Vertically 1.24 � 0.0007
CNT/PEEK (1 wt%) Fabricated Horizontally 1.21 � 0.0005
It should be noted that the mechanical properties of FFF-parts are
Fabricated Vertically 1.29 � 0.0004 influenced by factors such as the bead-bead interfacial strength,
CNT/PEEK (3 wt%) Fabricated Horizontally 1.23 � 0.0008 nanostructure-PEEK interfacial strength, dispersion state of the nano-
Fabricated Vertically 1.25 � 0.0006 reinforcement, interfacial voids between beads and the voids within
the beads. The dominating factors that govern the mechanical properties
of nano-reinforced FFF samples are the dispersion state of the nano-
compared to that of base PEEK. In summary, the horizontally printed
reinforcement and the interfacial strength between CNT/GNP and
nanocomposite samples exhibit lower CTE than the base PEEK, except
PEEK. The macroscopic density measurement presented in Table 2 is
for CNT reinforced PEEK above Tg. The in-plane CTE of GNP and CNT
influenced by the density of the individual constituents (fiber and ma­
reinforced PEEK nanocomposites is lower than that of the out-of-plane
trix), the intra-bead pores, the interfacial voids between beads and the
direction. This is optimally manifested by the horizontally fabricated
compaction degree during processing. One cannot see a linear rela­
samples since GNP and CNT tend to orient along the deposition path
tionship between the density (see, Table 2) and the mechanical prop­
direction. Insignificant changes in CTE of CNT reinforced PEEK nano­
erties of the horizontally fabricated parts because of these multiple
composites above Tg could be due to different intrinsic thermal expan­
influencing factors.
sion property of CNT above Tg. The decrease in CTE with incorporation
of GNP and CNT indicates better dimensional stability of the nano­
composites as compared with base PEEK.
The vertically fabricated samples exhibit different trend as compared
Table 3
to horizontally fabricated samples. GNP/PEEK samples show increase in CTE and standard error of base, and CNT and GNP reinforced PEEK nano­
CTE, while CNT/PEEK samples do not exhibit significant changes in composites. Statistically significant increases/decreases vs. neat PEEK are shown
CTE, both below and above Tg. The CNT and GNP in vertically fabricated in bracket.
samples are oriented transversely to the specimen’s longitudinal direc­
CTE of Horizontally Fabricated CTE of Vertically Fabricated
tion and thus intrinsically low CTE of CNT and GNP cannot be man­ Samples [10 6/K] Samples [10 6/K]
ifested by the composites. Increased CTE in vertically fabricated GNP/
Below Tg Above Tg Below Tg Above Tg
PEEK samples could be due to constraints offered by carbon nano­
structures for lateral expansion (as reinforcements are oriented trans­ PEEK 52.17 � 0.40 128.0 � 1.25 56.15 � 0.58 132.4 � 2.31
GNP/PEEK 44.04 � 0.53 105.9 � 1.29 64.18 � 0.78 153.4 � 1.50
versely to longitudinal direction) and the resulting Poisson’s effect. (3 wt%) ( 16%) ( 17%) (þ14%) (þ16%)
GNP/PEEK 38.63 � 0.64 92.2 � 1.91 63.88 � 0.77 161.8 � 2.10
3.5. Tensile properties (5 wt%) ( 26%) ( 28%) (þ14%) (þ22%)
CNT/PEEK 47.30 � 0.60 135.6 � 1.64 59.69 � 0.56 128.0 � 1.64
(1 wt%) ( 9%) (þ6%) (þ6%) ( 3%)
The tensile performance of base and PEEK nanocomposites is sum­ CNT/PEEK 42.70 � 0.19 124.3 � 1.47 59.19 � 0.21 124.9 � 2.12
marized in Table 4. The stress-strain response of the samples are shown (3 wt%) ( 18%) ( 3%) (þ5%) ( 6%)
in Fig. S5. The Young’s modulus of GNP and CNT reinforced PEEK

Fig. 5. Thermal expansion behavior of the base and reinforced PEEK composites for a) horizontally fabricated samples and b) vertically fabricated samples.

7
M.F. Arif et al. Composites Part B 184 (2020) 107625

Fig. 6. Storage modulus and tan δ of the base and reinforced PEEK nanocomposites for a) horizontally fabricated samples and b) vertically fabricated samples.

the storage modulus onset and peak of tan δ.

Table 4
Tensile properties of horizontally fabricated samples of base and reinforced
PEEK nanocomposites. Statistically significant increases/decreases vs. neat 3.7. Wear and hardness
PEEK are shown in bracket.
PEEK GNP/ GNP/ CNT/ CNT/ Fig. 7a shows the coefficient of friction (COF) as a function of number
PEEK (3 PEEK (5 PEEK (1 PEEK (3 of cycles for base PEEK and PEEK nanocomposites. COF of the PEEK
wt%) wt%) wt%) wt%)
decreased with the reinforcement of carbon nanostructures from ~0.25
Young’s 3.15 3.68 � 3.89 � 3.37 � 3.77 � (PEEK) to 0.08 for CNT/PEEK composite at 1 wt % CNT loading and 0.1
Modulus, Et � 0.11 0.05 0.15 0.04 (7%) 0.04 for GNP/PEEK composite at 3 wt % GNP loading. The wear volume and
(GPa) (17%) (23%) (20%)
wear rate were calculated by scanning the worn surface with the help of
Tensile 66.2 66.1 � 58.5 � 62.5 � 63.4 �
strength, σm � 2.61 0.96 3.06 2.13 1.02
an optical profilometer and the results are presented in Table 5. The
(MPa) ( 12%) ( 6%) ( 4%) surface profiles of the scar generated due to wear are shown in Fig. 7b, c
Tensile 3.10 3.46 � 3.55 � 2.66 � 2.41 � and 7d for PEEK, CNT/PEEK and GNP/PEEK respectively. Bulk hardness
elongation, � 0.32 0.15 0.17 0.15 0.21 measured through Vickers’s indentation test is presented in Table 5.
εb (%) (12%) (15%) ( 14%) ( 22%) The decrease in the COF for PEEK with the addition of carbon
Toughness 1.25 1.54 � 1.50 � 0.98 � 0.90 � nanostructure can be correlated with the hardness. The hardness of
(MJ.m 3) � 0.19 0.10 0.16 0.14 0.12
CNT/PEEK and GNP/PEEK nanocomposites decreases from ~302 (HV)
(23%) (20%) ( 22%) ( 28%)
Poisson’s Ratio 0.43 0.37 � 0.32 � 0.43 � 0.43 �
to ~ 238 and ~279, respectively. The contact areas of the counter body
ν, � 0.01 0.01 0.01 0.01 0.01 (6 mm dia SS ball) on the samples were calculated using image of the
( 14%) ( 26%) wear scar obtained. Wear rate was minimum in neat PEEK samples (1.23
mm3/Nm). Carbon nanostructure reinforced PEEK samples, exhibit
higher wear rate compared to neat PEEK, i.e., 2.97 and 2.72 mm3/Nm
3.6. Dynamic mechanical analysis
for CNT/PEEK and GNP/PEEK respectively.
Neat PEEK shows lower wear rate because of higher bulk hardness
DMA analysis for base and nano-reinforced PEEK composites were
and higher density, whereas CNT/PEEK and GNP/PEEK nanocomposites
performed to examine the effect of nano-reinforcement and fabrication
showed increased wear rate but lower COF due to their lower hardness
configuration on the thermomechanical properties of the PEEK nano­
and higher multiscale porosity. The increase in surface roughness caused
composites. The storage modulus ðE’Þ and tan δ of the PEEK nano­
by the CNT/GNP in the FFF printing process and larger extent of micro
composites are shown in Fig. 6. In horizontally fabricated samples,
and nano-pores within the beads and between beads lead to reduced
incorporation of GNP and CNT into the PEEK matrix increases the
hardness.
storage modulus in the glassy region (below Tg). The higher content of
nano-reinforcement results in higher storage modulus. The storage
4. Summary and conclusions
modulus of the samples above Tg reduces drastically, with no significant
difference between the base and reinforced PEEK nanocomposites. In
Carbon nanostructures reinforced PEEK filaments were developed
contrast, GNP and CNT reinforcements do not significantly influence the
for FFF additive manufacturing and the multifunctional performance of
storage modulus of vertically fabricated samples in both regimes (below
resulting FFF-PEEK nanocomposites were thoroughly investigated. In­
and above Tg). The improved storage modulus of horizontally fabricated
fluence of GNP (3 and 5 wt%) and CNT (1 and 3 wt%) content on the
samples is due to uniform dispersion of GNP/CNT in the PEEK matrix
density, CTE, tensile properties, dynamic mechanical properties, wear,
(see, Fig. 4b and c) which enables efficient stress transfer between the
friction and hardness characteristics of 3D printed PEEK nano­
GNP/CNT and PEEK and the improved bead-bead interfacial strength.
composites were investigated. The crystallization temperature and the
The vertically fabricated samples do not show significant improvement
degree of crystallinity of FFF-PEEK increase with the addition of carbon
with GNP/CNT incorporation because the load (in bending mode) in this
nanostructures. The print direction influences the extent and distribu­
sample is parallel to multiple weak interfaces between beads. For all the
tion of pores and thus the density. For horizontally fabricated samples,
samples, the storage modulus shows a sharp drop at a temperature of
the CTE of PEEK nanocomposites significantly decreases by up to 26%
150 � C which indicates the Tg. No significant changes in Tg of both
and 18% with the addition of 5 wt% GNP and 3 wt% CNT, respectively,
horizontally and vertically fabricated samples are noticed, as seen from
indicating a better dimensional stability of the resulting nanocomposite

8
M.F. Arif et al. Composites Part B 184 (2020) 107625

Fig. 7. The COF and 2D images of the wear scar obtained from fretting wear test. (a) COF vs number of cycles, (b) wear scar on PEEK, (c) wear scar on CNT/PEEK
(1.0 wt% CNT) nanocomposite and (d) wear scar on GNP/PEEK (3.0 wt% GNP) nanocomposite.

evince higher wear rate but lower COF due to their lower hardness and
Table 5
higher multiscale porosity. All nanocomposite samples examined here
The wear characteristics and hardness obtained for the PEEK and PEEK
formed electrically percolating network and exhibited a minimum
nanocomposites.
electrical conductivity σ ¼ 9.1 � 10 7 S/cm (for 1 wt% CNT loading).
Sample COF Wear Wear rate Contact Hardness The multifunctional performance and multiscale characteristics of FFF
volume (10 4) area (HV)
(μm3) (mm3/Nm) (mm2)
AM GNP/PEEK and CNT/PEEK nanocomposites demonstrated in this
study exhibit their promise for several load bearing and functional ap­
PEEK 0.263 123,876 1.23 0.126 302.40 �
plications such as orthopedics, space, oil and gas and automotive.
� 0.038 22.69
CNT/ 0.087 297,411 2.97 0.404 238.25 �
PEEK � 0.003 69.65 Acknowledgement
(1.0 wt
%) This publication is based upon work supported by the Khalifa Uni­
GNP/ 0.116 272,877 2.72 0.103 279.63 �
PEEK � 0.011 45.52
versity of Science and Technology under Award No. CIRA-2018-128.
(3.0 wt
%) Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.

structures as compared with neat PEEK.
org/10.1016/j.compositesb.2019.107625.
The tensile modulus of GNP/PEEK and CNT/PEEK nanocomposites
increases with increasing nano-reinforcement content. The GNP/PEEK
References
nanocomposites exhibit ductile fracture with concomitant increase in
strain at break and toughness relative to neat PEEK. CNT/PEEK nano­ [1] Go J, Schiffres SN, Stevens AG, Hart AJ. Rate limits of additive manufacturing by
composites exhibit brittle fracture with reduced toughness. Higher fused filament fabrication and guidelines for high-throughput system design. Addit
micro-voids in GNP/PEEK results in lower Poisson’s ratio. Incorporation Manuf 2017;16:1–11. https://doi.org/10.1016/j.addma.2017.03.007.
[2] Hart AJ, Rao A. How to print a 3D object all at once. Science (80-) 2019;363:
of GNP and CNT increases the storage modulus in the glassy region due 1042–3. https://doi.org/10.1126/science.aaw7062.
to uniform dispersion of GNP and CNT in the PEEK matrix for hori­
zontally fabricated samples. CNT/PEEK and GNP/PEEK nanocomposites

9
M.F. Arif et al. Composites Part B 184 (2020) 107625

[3] Kumar S, Wardle BL, Arif MF. Strength and performance enhancement of bonded [24] Yu WW, Zhang J, Wu JR, Wang XZ, Deng YH. Incorporation of graphitic nano-filler
joints by spatial tailoring of adhesive compliance via 3D printing. ACS Appl Mater and poly(lactic acid) in fused deposition modeling. J Appl Polym Sci 2017;134:
Interfaces 2017;9:884–91. https://doi.org/10.1021/acsami.6b13038. 1–11. https://doi.org/10.1002/app.44703.
[4] Kumar S, Wardle BL, Arif MF, Ubaid J. Stress reduction of 3D printed compliance- [25] Weng Z, Wang J, Senthil T, Wu L. Mechanical and thermal properties of ABS/
tailored multilayers. Adv Eng Mater 2018;20. https://doi.org/10.1002/ montmorillonite nanocomposites for fused deposition modeling 3D printing. Mater
adem.201700883. 1700883. Des 2016;102:276–83. https://doi.org/10.1016/j.matdes.2016.04.045.
[5] Liljenhjerte J, Upadhyaya P, Kumar S. Hyperelastic strain measurements and [26] Ryder MA, Lados DA, Iannacchione GS, Peterson AM. Fabrication and properties of
constitutive parameters identification of 3D printed soft polymers by image novel polymer-metal composites using fused deposition modeling. Compos Sci
processing. Addit Manuf 2016;11:40–8. https://doi.org/10.1016/j. Technol 2018;158:43–50. https://doi.org/10.1016/j.compscitech.2018.01.049.
addma.2016.03.005. [27] Lebedev SM, Gefle OS, Amitov ET, Zhuravlev DV, Berchuk DY, Mikutskiy EA.
[6] Liljenhjerte J, Kumar S. Pull-out performance of 3D printed composites with Mechanical properties of PLA-based composites for fused deposition modeling
embedded fins on the fiber. MRS Proc 2015;1800:mrss15–2135597. https://doi. technology. Int J Adv Manuf Technol 2018;97:511–8. https://doi.org/10.1007/
org/10.1557/opl.2015.645. s00170-018-1953-6.
[7] de Leon AC, Chen Q, Palaganas NB, Palaganas JO, Manapat J, Advincula RC. High [28] Wu H, Sulkis M, Driver J, Saade-Castillo A, Thompson A, Koo JH. Multi-functional
performance polymer nanocomposites for additive manufacturing applications. ULTEMTM1010 composite filaments for additive manufacturing using fused
React Funct Polym 2016;103:141–55. https://doi.org/10.1016/j. filament fabrication (FFF). Addit Manuf 2018. https://doi.org/10.1016/j.
reactfunctpolym.2016.04.010. addma.2018.10.014.
[8] Panayotov IV, Orti V, Cuisinier F, Yachouh J. Polyetheretherketone (PEEK) for [29] Goyal R, Tiwari AN, Mulik UP, Negi YS. Novel high performance Al2O3/poly(ether
medical applications. J Mater Sci Mater Med 2016;27:118. https://doi.org/ ether ketone) nanocomposites for electronics applications. Compos Sci Technol
10.1007/s10856-016-5731-4. 2007;67:1802–12. https://doi.org/10.1016/j.compscitech.2006.10.020.
[9] Rinaldi M, Ghidini T, Cecchini F, Brandao A, Nanni F. Additive layer [30] Reddy SK, Kumar S, Varadarajan KM, Marpu PR, Gupta TK, Choosri M. Strain and
manufacturing of poly (ether ether ketone) via FDM. Compos B Eng 2018;145: damage-sensing performance of biocompatible smart CNT/UHMWPE
162–72. https://doi.org/10.1016/j.compositesb.2018.03.029. nanocomposites. Mater Sci Eng C 2018;92:957–68. https://doi.org/10.1016/j.
[10] Yasin S, Shakeel A, Ahmad M, Ahmad A, Iqbal T. Physico-chemical analysis of msec.2018.07.029.
semi-crystalline PEEK in aliphatic and aromatic solvents. Soft Mater 2019;00:1–7. [31] Gupta TK, Choosri M, Varadarajan KM, Kumar S. Self-sensing and mechanical
https://doi.org/10.1080/1539445X.2019.1572622. performance of CNT/GNP/UHMWPE biocompatible nanocomposites. J Mater Sci
[11] Valentan B, Kadivnik Z, � Brajlih T, Anderson A, Drstven�sek I. Processing poly(ether 2018;53:7939–52. https://doi.org/10.1007/s10853-018-2072-3.
etherketone) on a 3d printer for thermoplastic modelling. Mater Tehnol 2013;47: [32] Gupta TK, Kumar S, Khan AZ, Varadarajan KM, Cantwell WJ. Self-sensing
715–21. performance of MWCNT-low density polyethylene nanocomposites. Mater Res
[12] Vaezi M, Yang S. Extrusion-based additive manufacturing of PEEK for biomedical Express 2018;5:015703. https://doi.org/10.1088/2053-1591/aa9f9e.
applications. Virtual Phys Prototyp 2015;10:123–35. https://doi.org/10.1080/ [33] Arif MF, Kumar S, Gupta TK, Varadarajan KM. Strong linear-piezoresistive-
17452759.2015.1097053. response of carbon nanostructures reinforced hyperelastic polymer
[13] Wu WZ, Geng P, Zhao J, Zhang Y, Rosen DW, Zhang HB. Manufacture and thermal nanocomposites. Compos Part A Appl Sci Manuf 2018;113:141–9. https://doi.org/
deformation analysis of semicrystalline polymer polyether ether ketone by 3D 10.1016/j.compositesa.2018.07.021.
printing. Mater Res Innov 2014;18. https://doi.org/10.1179/ [34] Marasso SL, Cocuzza M, Bertana V, Perrucci F, Tommasi A, Ferrero S, et al. PLA
1432891714Z.000000000898. S5-12-S5-16. conductive filament for 3D printed smart sensing applications. Rapid Prototyp J
[14] Wu W, Geng P, Li G, Zhao D, Zhang H, Zhao J. Influence of layer thickness and 2018;24. https://doi.org/10.1108/RPJ-09-2016-0150. 00–00.
raster angle on the mechanical properties of 3D-printed PEEK and a comparative [35] Berretta S, Davies R, Shyng YT, Wang Y, Ghita O. Fused Deposition Modelling of
mechanical study between PEEK and ABS. Materials 2015;8:5834–46. https://doi. high temperature polymers: exploring CNT PEEK composites. Polym Test 2017;63:
org/10.3390/ma8095271. 251–62. https://doi.org/10.1016/j.polymertesting.2017.08.024.
[15] Cicala G, Latteri A, Del Curto B, Lo Russo A, Recca G, Far� e S. Engineering [36] Gonçalves J, Lima P, Krause B, P€ otschke P, Lafont U, Gomes J, et al. Electrically
thermoplastics for additive manufacturing: a critical perspective with experimental conductive polyetheretherketone nanocomposite filaments: from production to
evidence to support functional applications. J Appl Biomater Funct Mater 2017;15. fused deposition modeling. Polymers 2018;10:925. https://doi.org/10.3390/
https://doi.org/10.5301/jabfm.5000343. 0. polym10080925.
[16] Yang C, Tian X, Li D, Cao Y, Zhao F, Shi C. Influence of thermal processing [37] Pu�ertolas JA, Castro M, Morris JA, Ríos R, Ans� on-Casaos A. Tribological and
conditions in 3D printing on the crystallinity and mechanical properties of PEEK mechanical properties of graphene nanoplatelet/PEEK composites. Carbon N Y
material. J Mater Process Technol 2017;248:1–7. https://doi.org/10.1016/j. 2019;141:107–22. https://doi.org/10.1016/j.carbon.2018.09.036.
jmatprotec.2017.04.027. [38] Kalin M, Zalaznik M, Novak S. Wear and friction behaviour of poly-ether-ether-
[17] Arif MF, Kumar S, Varadarajan KM, Cantwell WJ. Performance of biocompatible ketone (PEEK) filled with graphene, WS 2 and CNT nanoparticles. Wear 2015;
PEEK processed by fused deposition additive manufacturing. Mater Des 2018;146: 332–333:855–62. https://doi.org/10.1016/j.wear.2014.12.036.
249–59. https://doi.org/10.1016/j.matdes.2018.03.015. [39] Malucelli G, Marino F. Abrasion resistance of polymer nanocomposites - a review.
[18] Berretta S, Evans K, Ghita O. Additive manufacture of PEEK cranial implants: In: Adamiak M, editor. Abrasion resist. Mater., IntechOpen; 2012.
manufacturing considerations versus accuracy and mechanical performance. Mater [40] Dasari A, Yu Z-Z, Mai Y-W. Wear/scratch damage. 2016. p. 207–26. https://doi.
Des 2018;139:141–52. https://doi.org/10.1016/j.matdes.2017.10.078. org/10.1007/978-1-4471-6809-6_9.
[19] Stepashkin АА, Chukov DI, Senatov FS, Salimon AI, Korsunsky AM, Kaloshkin SD. [41] Zalaznik M, Kalin M, Novak S, Jak�sa G. Effect of the type, size and concentration of
3D-printed PEEK-carbon fiber (CF) composites: structure and thermal properties. solid lubricants on the tribological properties of the polymer PEEK. Wear 2016;
Compos Sci Technol 2018;164:319–26. https://doi.org/10.1016/j. 364–365:31–9. https://doi.org/10.1016/j.wear.2016.06.013.
compscitech.2018.05.032. [42] Chao J. Fretting-fatigue induced failure of a connecting rod. Eng Fail Anal 2019;96:
[20] Liao G, Li Z, Cheng Y, Xu D, Zhu D, Jiang S, et al. Properties of oriented carbon 186–201. https://doi.org/10.1016/j.engfailanal.2018.10.006.
fiber/polyamide 12 composite parts fabricated by fused deposition modeling. [43] Koo JH. Polymer nanocomposites: processing, characterization, and applications.
Mater Des 2018;139:283–92. https://doi.org/10.1016/j.matdes.2017.11.027. second ed. New York: McGraw-Hill; 2019.
[21] Schmitz DP, Ecco LG, Dul S, Pereira ECL, Soares BG, Barra GMO, et al. [44] Blundell DJ, Osborn BN. The morphology of poly(aryl-ether-ether-ketone).
Electromagnetic interference shielding effectiveness of ABS carbon-based Polymer 1983;24:953–8. https://doi.org/10.1016/0032-3861(83)90144-1.
composites manufactured via fused deposition modelling. Mater Today Commun [45] Kazakova MA, Selyutin AG, Semikolenova NV, Ishchenko AV, Moseenkov SI,
2018;15:70–80. https://doi.org/10.1016/j.mtcomm.2018.02.034. Matsko MA, et al. Structure of the in situ produced polyethylene based composites
[22] Prashantha K, Roger F. Multifunctional properties of 3D printed poly(lactic acid)/ modified with multi-walled carbon nanotubes: in situ synchrotron X-ray diffraction
graphene nanocomposites by fused deposition modeling. J Macromol Sci Part A and differential scanning calorimetry study. Compos Sci Technol 2018;167:
2017;54:24–9. https://doi.org/10.1080/10601325.2017.1250311. 148–54. https://doi.org/10.1016/j.compscitech.2018.07.046.
[23] Dul S, Fambri L, Pegoretti A. Fused deposition modelling with ABS–graphene [46] Yang B, Wang D, Chen F, Su L-F, Miao J-B, Chen P, et al. Melting and crystallization
nanocomposites. Compos Part A Appl Sci Manuf 2016;85:181–91. https://doi.org/ behaviors of poly(lactic acid) modified with graphene acting as a nucleating agent.
10.1016/j.compositesa.2016.03.013. J Macromol Sci Part B 2019;58:290–304. https://doi.org/10.1080/
00222348.2018.1564222.

10