Surface & Coatings Technology 235 (2013) 603610

Contents lists available at ScienceDirect

Surface & Coatings Technology

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

Investigation of surface properties and mechanical and tribological

behaviors of polyimide based composite coatings
Camelia Demian a,, Hanlin Liao a, Remy Lachat b, Sophie Costil a
a
b

Laboratoire IRTES-LERMPS, Universit de Technologie de Belfort-Montbliard, 90010 Belfort Cedex, France

Laboratoire IRTES-M3M, Universit de Technologie de Belfort-Montbliard, 90010 Belfort Cedex, France

a r t i c l e

i n f o

Article history:
Received 17 June 2012
Accepted in revised form 19 August 2013
Available online 29 August 2013
Keywords:
Polyimide based composite coatings
Total surface energy
Friction coefcient
Wear rate
Thermal stability

a b s t r a c t
This article focuses on the development and characterization of thermoset polyimide (PI) based composite coatings on aluminum substrates. In order to improve the tribological behavior, PTFE and SiC llers were added into
pure PI to develop composite coatings. A thermal study to validate the condition of the pure PI coating after the
elaboration process was performed using DSC analyses and the Tg evolution with the temperature of a pure
polyimide sample was investigated.
Then, the inuence of the llers (PTFE and SiC) on surface properties, and mechanical and tribological behaviors
of the PI composite coatings is considered. Results showed that, by adding PTFE particles into the pure PI, lower
surface energies and lower and stable friction coefcients can be obtained. Besides, the addition of SiC particles
improved the mechanical behavior such as hardness and wear resistance of the composite PIPTFESiC coatings.
Following the obtained results, correlations between wear rate depending on total surface energy and
microhardness were established. Thermogravimetric analysis (TGA) of the PI and PI composite coatings was
carried out. The results revealed that the addition of llers into PI pure matrix improved the thermal stability
of the composite coatings.
2013 Elsevier B.V. All rights reserved.

1. Introduction
Advanced composites exhibit appropriate physical and chemical
properties that include light weight characteristics coupled with high
stiffness and strength, dimensional and thermal stability, chemical
resistance, exural performance and relatively easy processing. The
resin systems used to manufacture advanced composites concern
two basic types: thermosetting and thermoplastic polymers. Thermosetting resins such as epoxy, polyurethane, phenolic and amino, and
bismaleimide (BMI, polyimide) are widely used in advanced composite
manufacturing. As a thermoset polymer, polyimide (PI) possesses
outstanding properties such as excellent mechanical and electrical
insulating properties, good thermal stability and chemical inertness,
high wear resistance and resistance to radiation which make it suitable
for a wide range of applications [1]. In spite of the good characteristics of
polyimide, studies have been carried out to improve their surface properties and tribological behavior by dispersing different ller materials
throughout the polymer matrix. Thus, the PI based composite becomes
a promising material with controlled mechanical or tribological properties. Besides, llers contribute to optimize the operational properties in
different applications, such as microelectronics or biomedical devices,
components for electrical, aerospace or automotive industries [24].
Corresponding author at: Laboratoire SPCTS, UMR CNRS 7315, Universit de Limoges,
12 rue Atlantis, 87068 Limoges Cedex, France. Tel.: +33 0 5 87 50 24 03.
E-mail address: camelia.demian@unilim.fr (C. Demian).
0257-8972/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.surfcoat.2013.08.032

Polyimide is also known as a thermo-stable polymer due to its excellent properties at elevated temperatures (250 and 350 C). The thermal
properties of different synthesized polyimides were already studied by
many researchers. For example, high glass transition temperature of
polyimide up to 310315 C [5,6] or 340 C [7] and no melting point
[58] have been observed on different DSC curves. The excellent thermal stability of polyimide from ambient temperature up to 350 C [9]
or 420 C [6] has been also reported. Moreover, the polyimide has a
degradation temperature above 500 C [5,6].
Polymer based composites reinforced at micro or nanoscale with different llers such as bers or solid lubricates have gained development
and become promising bulk or coating materials for tribological applications. The mechanical properties of the glassepoxy composites can be
improved with the addition of SiC ller and this composite presents a
lower wear rate [10]. The addition of the SiC particles in the polymer
matrix as a secondary ller along with graphite into glassepoxy matrix
increases the wear resistance of the composite material [11]. Like many
kinds of polymer materials, the PI based composite coatings were also
studied. For example, the PI resins lled with solid lubricants such as
graphite, MoS2 and PTFE particles and reinforced with carbon bers
(CF) have shown better friction and anti-wear behavior under waterlubrication rather than dry sliding [1]. The addition of appropriate content
of Al2O3 nanoparticles into PI improved signicantly the tribological behavior [12] of the specimens prepared by compression molding method.
Besides, the tribological properties of polyimide based composites
demonstrated good friction behavior and improved wear resistance

604

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

Fig. 1. DSC curve of the PI coatings.

if they are lled with carbon nanotubes (CNT) and Al2O3 [12,13]. The
graphite additives can control the interfacial sliding conditions, either
acting abrasively or adhesively for sintered PI composites [14]. By the
addition of graphite as internal lubricant into pure polyimide, lower
friction coefcients can be obtained under high loads and high sliding
velocities [15]. Besides, increasing SiC whisker content in a SiCAlN
polyimide matrix composite increased the toughness and decreased
the exural modulus [16].
The use of polyimide based composites as surface protection can
improve the tribological behavior and surface properties of different
substrate materials. For example, aluminum is widely used in industrial
applications for diverse parts and devices, but it possesses poor friction
properties. Moreover, polyimide based composite coatings also offer
high surface thermal stability when used in tribological applications,
even at elevated temperatures. It is already demonstrated that PTFE
signicantly improved the performances of other polymers such as
PEEK [17] and PI [18] because it controls the friction and wear behaviors.
Besides, SiC particles as a solid ller also contribute to enhance the
mechanical and tribological properties of PI based composites. However, few information about the mechanical and tribological behaviors of
polyimide composite coatings lled with PTFE and SiC was reported in
the literature. The aim of this study is to investigate the mechanical
and tribological behaviors of PIPTFE and PIPTFESiC composite

coatings under dry condition. For this purpose different PTFE weight
concentrations as well as a x amount of SiC particles were incorporated
into a fully imidized polyimide resin matrix to fabricate composite
coatings. The DSC and TGA analyses were performed on the polyimide
solution and pure polyimide and its composite coatings in order to
determine the structure after the heat treatment as well as the phenomena which occur during the polymer heating such as glass transition,
desolventation and thermal characteristics. Microscopic observations
of the coating surface and cross-section were performed in order to
reveal the surface morphology. The results of microhardness and tribological tests were correlated with the total surface energy determined
by sessile drop method.
2. Coating elaboration and characterization methods
2.1. Composite coating elaboration
PTFE particles with an average diameter of 30 m were added in 20,
30 and 40 wt.% into a polyimide fully imidized P84 solution (provided
by Evonik Industries, Evonik Fibres GmbH, Austria) and well mixed.
Polyimide P84 solution is presented in the NEP (N-ethyl-pyrrolidone)

Glass transition temperature

Tg (OC)

350
300
250
200
150
100
150

200

250

300

350

400

450

Isothermal holding (OC)

Fig. 2. Evolution of Tg with temperature.

Fig. 3. Surface morphology of PI20% PTFE5% SiC composite coating before polishing.

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

Fig. 4. PI based composite coating microstructures: A) PI20% PTFE; B) PI20% PTFE5%

SiC.

solvent, which represents 75% of the total mass. 5 wt.% SiC particles
having an average diameter of 1.7 m were added into the PIPTFE
mixture as the second solid ller. Because of the high viscosity of the
PIPTFE mixture, the chosen amount of SiC particles is almost the
limit in order to favor the application on the Al substrate. The resulting
composite mixtures were then applied on Al plates of 50 50 3 mm
to nally obtain a coating thickness between 60 and 120 m. Then,
the coatings were dried in a forced-air furnace at 80 C for a few
hours to evaporate the solvent, and then heated up to 315 C
with a heating rate of 1.5 C/min to remove the residual solvent and
to stabilize the properties of the composite material.

605

tribological behaviors of the composite coatings. Thermal investigations

of the PI coating, PI solution and PI based composite coatings were carried out using differential scanning calorimetry (DSC) and thermal
gravimetric (TGA) measurements. These analyses were conducted in
air atmosphere using Q10 and Q50 TA Instrument apparatus, respectively. The data were recorded using dedicated Universal AnalysisV4 acquisition software. The sample was embedded in an aluminum pan in
the case of DSC analysis and then heated until 500 C at a heating rate
of 10 C/min. Following the same experimental protocol, the PI based
composite coatings were also analyzed by a series of TGA tests. The samples were heated until 800 C at a heating rate of 10 C/min.
Coating morphology was investigated using a JSM 5800 LV, JEOL,
Japan scanning electron microscope. The microscopic observations in
cross-section view have been performed using a Leica DMRM optical
microscope. The grinding and polishing processes of metallographic
samples, which were embedded in epoxy resin under vacuum, were
performed according to an adapted procedure using a SiO2 solution
for polishing.
The surface properties were evaluated using sessile drop method to
determine the wettability and surface energies according to Owen
Wendt theory [19].
The sample surfaces used to evaluate the tribological properties of
the coatings and surface energy were polished to reach an average surface roughness between 1 and 3 m.
The friction coefcients of the composite coatings were determined
by tribological tests using the pin-on-disk method at ambient temperature. Friction tests were accomplished using a ball-on-disk CSM
tribometer (CSM Instruments, Swiss). The counterpart consisted of a
6 mm diameter 100C6 steel ball with a mirror nished surface (Ra =
0.02 m) and a hardness of 62 HRC. The friction force was measured
with a linear variable differential transformer sensor and dynamically
recorded into a computer. The friction coefcients were obtained when
the measured forces were divided by the applied load.
To determine the wear rate, prolometry traces were taken across
the wear tracks generated on the coating surface after tribological test,
using a 3D AltiSurf 500 prolometer (Altimet, France). The volume
loss of the coating was determined by multiplying the cross-section
area by the perimeter of the wear track. The wear rate was calculated as:
V
A  l mm3

w
d F d F Nm

!
1

where,
V
D
F
A
L

the volume loss (mm3);

the sliding distance (m);
the applied load (N);
the cross-section area of the wear track (mm2);
the perimeter of the circular track (mm): l = 2 3.142 r
(mm), where r (mm) represents the radius of the wear track.

2.2. Characterization methods

Several investigation methods were used to observe the inuence
of ller particles on the surface characteristics and mechanical and
Table 1
Coating thicknesses and surface roughness measurements.
Coating

Thickness (m) SD

Roughness Ra (m) SD

PI
PI20% PTFE
PI20% PTFE5% SiC
PI30% PTFE
PI30% PTFE5% SiC
PI40% PTFE
PI40% PTFE5% SiC

78
69
63
109
95
120
81

0.03
5.2
4.8
9.1
7.5
13.2
11.6

3
7
2
8
3
7
7

0.01
0.3
0.6
0.3
0.9
1.1
1.0

Finally, mechanical properties in terms of hardness were determined. At least 8 indentations using a load of 10 g were performed on
the metallographic samples in cross-section using a Leica VMHT 30A
Vickers tester. Then, to have a clear interpretation of the measures the
average and the standard deviation of the results were calculated.
3. Results and discussion
3.1. Thermal study of the polyimide matrix
The PI matrix is a fully imidized solution containing 25% of polyimide
and 75% of NEP solvent, provided by Evonik Industries GmbH, Austria.
The thermal study of the pure PI coating aims to validate the condition
of the coating after the elaboration process. From DSC tests conducted

606

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

Table 2
Contact angles and total surface energies of PI and PI composite coatings.
Material coatings

PI
PI20% PTFE
PI20% PTFE5% SiC
PI30% PTFE
PI30% PTFE5% SiC
PI40% PTFE
PI40% PTFE5% SiC

Contact angle, ()

Total surface energy (mN/m)

H2O (water)

C2H6O2 (ethylene-glycol)

C10H7Br (-bromonaphthalene)

Ps

D
s

Ts

73.1
80.5
79.1
83.9
87.5
97.9
98.1

39.7
57.7
60.3
62.9
69.4
76.9
72.4

8.8
42.1
41.0
44.7
54.6
54.9
49.0

5.1
4.6
5.0
3.8
3.5
0.9
0.6

41.2
31.0
30.4
29.5
25.1
25.5
29.1

46.4
35.6
35.4
33.2
28.6
26.5
29.7

on a PI solution sample, the Tg evolution with the temperature is observed. It serves as an indicator of mastery of the coating.
Several thermal events were observed on the DSC curve of the PI
coating presented in Fig. 1. The rst part of the curve corresponds to
the PI heat capacity. Then, two exothermic events occur at temperatures
of 128 C and 410 C, respectively. The second exothermal event which
could represent the expulsion of the residual solvent enclosed in the
polymer molecules is overlapped on the beginning of the degradation
process of the polymer. This process is possible due to the relaxation
of the molecular chains during the glass transition phenomenon. Subsequently, an acceleration of heat ow can be observed. A similar thermal
event was observed on the DSC curve of PI ber [6].
In order to detect the position of the Tg on the DSC curve of the PI
coating, a complementary DSC analysis was performed on a PI solution
sample. Fig. 2 shows the Tg evolution with the temperature. During the
heating several isothermal holdings were performed as follows:
150 min at 80 C, 80 min at 250 C and 300 C, and 20 min at 350 C
and 400 C. It can be observed that the Tg points follow a logarithmic
evolution with the temperature increase, but a logarithmic trendline
cannot be drawn due to point displacements. The increase of Tg with
temperature is due to the solvent expulsion which, at high temperatures, acts as a plasticizer.
Therefore, associating these results with the DSC curve presented in
Fig. 1, the position of the Tg point of the pure PI coating can be dened
around the temperature of 263 C. Besides, using a fully imidized PI
solution no melting point was observed on the DSC curve as shown in
Fig. 1.
3.2. Composite coating morphology
The composite coatings with different ller concentrations presented
similar morphology. For example, Fig. 3 presents the surface morphology
of PI20% PTFE5% SiC coating. PTFE and SiC particles appear quite
homogeneously dispersed into the PI matrix with respect to the
different particle diameters: about 30 and 1.7 m (particle dimensions
between 0.4 and 6.5 m) for the PTFE and SiC particles, respectively.

The cross-section microstructures of the PI based composite coatings

are shown in Fig. 4. A relative homogenous distribution of llers into PI
matrix can be distinguished. PTFE content does not affect the composite
coating morphology (Fig. 4A). When SiC particles are added, a homogeneous distribution of the reinforcing agents can be observed (Fig. 4B).
The composite coating thickness and roughness are listed in Table 1.
Both thickness and roughness increase with the ller content. The PI
coating has a very smooth surface with an average roughness Ra of
0.03 m compared with the PI based composites. The surface roughness
considerably increases with the increase of the PTFE particle content in
PI matrix. The addition of SiC particles has a positive effect on the surface
roughness for all composite coatings. When a mass amount of PTFE
above 20% is added, the composite coating thickness increases from
69 to 109 and 120 m corresponding to PI30% PTFE and PI40% PTFE,
respectively. Moreover, the roughness also increases from 5 m corresponding to the PI20% PTFE to 9 and 13 m corresponding to PI30%
PTFE and PI40% PTFE, respectively. The addition of SiC particles slightly
reduces the thickness and roughness values of the composite coatings
lled with 20, 30 and 40 wt.% PTFE.

3.3. Evolution of the surface properties of coatings

The measured contact angles and surface energies of the PI and PI
based composite coatings are listed in Table 2. The wetting occurs if
the contact angle between the water droplet and the surface is less
than 90. Consequently the surface is considered hydrophilic. Conversely, non-wetting occurs if the contact angle is higher than 90 and the
surface is considered hydrophobic. For the pure PI coating the water
contact angle has a value of 73. As a result, the surface is considered either wetting or hydrophilic. Adding PTFE and SiC reinforcing agents the
angle increases up to 98 and the surface becomes hydrophobic for PI
40% PTFE and PI40% PTFE5% SiC coatings. Besides, the total surface energy is directly related to the wettability and represents an important
parameter which can strongly affect the surface properties. The total

30
25

HV0.01

20
15
10
5

Friction coefficient

0.4

F = 5 N; v = 0.6 m/s

0.35
0.3
0.25

PI coating

0.2
0.15
0.1

PI based composite coatings

0.05
0
0

200

0
PI
PI-40%PTFE-5%SiC
PI-40%PTFE

Fig. 5. Vickers microhardness of PI and PI based composite coatings.

400

600

800

1000

Sliding distance d (m)

PI-20%PTFE-5%SiC
PI-20%PTFE

PI-30%PTFE-5%SiC
PI-30%PTFE

Fig. 6. Friction coefcient of PI and PI based composite coatings depending on sliding

distance.

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

Table 3
Friction coefcients at different sliding velocities and an applied load of 5 N as well as a
calculated wear rate.
Sample

Sliding velocity v
(m/s)

Average friction
coefcient SD

Wear rate w
(mm3/Nm)

PI

0.4
0.6
0.4
0.6
0.8
0.4
0.6
0.8
0.4
0.6
0.8
0.4
0.6
0.8
0.4
0.6
0.8
0.4
0.6
0.8

0.38
0.34
0.16
0.16
0.16
0.18
0.17
0.17
0.16
0.15
0.15
0.18
0.17
0.17
0.17
0.16
0.16
0.18
0.17
0.17

0.42
0.71
0.67
0.58
0.56
0.62
0.57
0.52
13.03
2.99
1.71
0.74
0.61
0.60
14.02
4.19
2.34
1.95
1.47
1.34

PI20% PTFE

PI20% PTFE5% SiC

PI30% PTFE

PI30% PTFE5% SiC

PI40% PTFE

PI40% PTFE5% SiC

0.08
0.03
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.01
0.01
0.01
0.01
0.01

surface energy decreases with the PTFE content. The SiC content has less
inuence on the total surface energy of the composite coatings. A minor
reduction of the total surface energy can be observed in the case of PI
20% PTFE5% SiC and PI30% PTFE5% SiC, respectively. An exception
can be noticed in the case of PI40% PTFE, where adding the SiC
particles slightly increases the total surface energy.
3.4. Mechanical behavior
The mechanical behavior of the PI and PI based composite coatings
was determined by microhardness measurements. The HV0.01 experimental results are listed in Fig. 5.
PI coating microhardness reached a value of about 27 HV0.01 which is
higher than that of the PI based composite coatings (Fig. 5). If different
wt.% of PTFE particles are added into the polyimide matrix a signicant
decrease in microhardness of the composite coatings was observed.
However, adding SiC particles into the PIPTFE matrix the microhardness
of the PIPTFESiC composite coatings slightly increases.
3.5. Tribological behavior
The tribological tests were conducted at room temperature and lab
air environment. For an applied load of 5 N the sliding speeds of 0.4,
0.6 and 0.8 m/s were selected to investigate their inuences on the friction coefcients and wear rates. For every test, the ball slides 1000 m on
the coating surface. Fig. 6 illustrates the friction coefcient curves of the
PI and PI based composite coatings versus sliding distance. Generally,
the friction coefcient of composite coatings has a relatively stable
value from the start of the sliding track.
Table 3 illustrates the data input for the friction tests as well as the
average friction coefcients obtained for all composite coatings and
the calculated wear rates.
PI coating has a friction coefcient of 0.34 which is approximately
two times higher than that of the PI based composite coatings. For
each PI composite coating, the friction coefcient decreases slightly
along with sliding velocity increase but its value varies over a small
range. For example, in the case of PI20% PTFE5% SiC, the friction coefcients have the values of 0.18 and 0.17 for the sliding velocities of 0.4
and 0.6 and 0.8 m/s, respectively. Besides, depending on the ller content, the composite coatings have different friction behaviors. From
the results listed in Table 3, it can be noticed that the addition of PTFE
into the PI matrix leads to a friction coefcient decrease in comparison

607

with the pure PI coating. For example, adding 20 wt.% of PTFE in the PI
matrix the friction coefcient decreases from 0.34 to 0.16 for a sliding
velocity of 0.6 m/s. Along with the increase of PTFE content into the PI
matrix, the friction coefcients vary in a small range for PI30% PTFE
and PI40% PTFE. By the addition of SiC particles into the PIPTFE matrix
a slight increase of the friction coefcient can be observed. Similar
friction coefcient of 0.15 was reported for PI15% PTFE composite
coating, but under harder conditions (sliding velocity of 1.4 m/s and
an applied load of 200 N) [18].
The wear dependence of PI and PI based composite coatings on sliding velocity is illustrated in Fig. 7. The wear rate of PI coating is lower
than that of PIPTFE and PIPTFESiC composite coatings. Besides, for
all composite coatings the wear rate decreases with the sliding velocity
increase. An exception can be noticed in the case of PI coating, whose
wear rate increases with the sliding velocity increase, e.g. from a wear
rate of 0.42 to 0.71 mm3/Nm for sliding velocities of 0.4 and 0.6 m/s,
respectively. For the PI based composite coatings, the wear rate increases along with the PTFE content. For example, at a sliding velocity
of 0.6 m/s, the wear rate of PI20% PTFE coating has a value of
0.58 mm3/Nm which increases at 2.99 and 4.19 mm3/Nm for the
PI30% PTFE and PI40% PTFE, respectively. The PI40% PTFE composite
coating reaches a maximum wear rate of 14 mm3/Nm for a sliding velocity of 0.4 m/s. In general, for all composite coatings, the addition of
SiC ller leads to a wear rate decrease. For example, in the case of PI
20% PTFE matrix, the addition of SiC particles led to a slight decrease
of the wear rate from 0.56 to 0.52 mm3/Nm. This difference is
highlighted with the increase of PTFE content.
When developing wear protection coatings for tribological applications, the total surface energy also plays an important role. Fig. 8a and
b illustrates the linear regression of wear rate depending on total surface energy as well as wear rate depending on microhardness. The linear
regression curve of wear rate depending on total surface energy is illustrated in Fig. 8a. The addition of PTFE llers results in a wear rate
increase and a surface total energy decrease. Besides, with the addition
of SiC particles, both wear rate and total surface energy decrease. The
linear regression presented in Fig. 8b shows that except the PI coating,
the composite coatings with a relatively high hardness exhibit less
wear, especially for the composite coatings that contain SiC particles.
3.6. Thermogravimetric analysis of the pure PI and PI based composite
coatings
Thermal stability of PI and PI based composite coatings were evaluated by TGA (Fig. 9). The pure PI coating did not lose any signicant
mass, other than due to the loss of moisture or the residual solvent
expulsion, until about 400 C. For example, an amount of 1.4% moisture
evaporation occurred at a temperature of 93 C and then, a weight loss
of 4% at a temperature of 378 C. Above 400 C, the degradation process
started slowly, and an offset point can be observed at a temperature of
569 C. After that, the mass loss was much faster and by 700 C the coating was completely burnt.
The TGA curves show that the composite coatings are more stable
than the pure PI coating. This may be due to the PTFE ller melting
point (about 330 C) which occurs in the same temperature range as
the expulsion of solvent. On the other side, PTFE has a lower degradation temperature than the PI, resulting in the degradation of PI based
composite coatings. SiC ller was found as a residue but it does not
alter the degradation process of the composite coatings.
The real coating compositions were determined by TGA tests. Accordingly with TGA curves shown in Fig. 9, the proportion of different
components in the solidied coatings should be noted which are
displayed in Table 4.
The proportions found in the coating residues contain a SiC ller. Besides, it is not possible to distinguish the proportion of PI and PTFE
which are two organic polymers that degrade in the same temperature
range. The only distinction that is possible to realize is that due to the

608

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

14

Wear w (mm3/N3m)

12
10
8
6
4
2
0

0.4 0.6

0.4 0.6 0.8

0.4 0.6 0.8

0.4 0.6 0.8

0.4 0.6 0.8

0.4 0.6 0.8

0.4 0.6 0.8

Sliding velocity v (m)

Fig. 7. Wear rate of PI and PI composite coatings depending on sliding velocity.

high concentration of PTFE in the composite coatings the thermal events

associated to PI are masked by those associated to PTFE. This justies the
main studies concerning the thermal behavior of the pure PI coating.
3.7. Discussion
DSC analyses performed on the pure PI coating revealed a high Tg
point which along with the TGA results, indicate that the PI matrix is
thermally stable above 400 C. Moreover, the analysis performed on a

Wear rate w (mm3/Nm)

5
4
3
2
1
0
20

Wear rate w (mm3/Nm)

30

40

50

20

30

Total surface energy sT (mN/m2)

5
4
3
2
1
0

10

HV0.01
Fig. 8. Linear regression curves of PI and PI based composite coatings: a) wear rate
depending on total surface energy; b) wear rate depending on microhardness.

PI solution sample demonstrated that the Tg increases with the temperature increase and results in a thermally stable PI at an elevated temperature. Still, a mass loss of PI coating was detected on the TGA curve
beginning with the temperature of 310 C (Fig. 9) which could correspond to the residual solvent evaporation. A complete desolventation
of the polyimide is hard to accomplish, because it is directly related to
the temperature and time and with a temperature increase, the polymer
degrades. If the heat treatment of the PI coating is performed at a temperature above 350 C the polymer could be damaged, but it contains
a less solvent amount, which could inuence the mechanical properties
of the composite coating matrix.
Concerning the stability of pure PI matrix in the absence of solvent,
an incorrect heat treatment which ended up with a lower Tg could affect
the properties of the nal coatings, e.g. a decrease of the mechanical
properties of the nal composite coatings. The effect of degassing process results in an effective process control for the drying procedure of
the PIPTFESiC composite coatings.
Analyzing the results presented in Tables 2 and 3, it can be noticed
that low total surface energies lead to low friction coefcients. This is
due to the addition of the PTFE ller which is known for its low total surface energy. The total surface energy decreases along with the PTFE content increase from 46 to 26 mN/m2 for PI and PI40% PTFE coatings,
respectively. Moreover, along with total surface energy decrease the
friction coefcient decreases from 0.34 to 0.16 which corresponds to a
sliding speed of 0.6 m/s for PI and PI40% PTFE coatings. However
when increasing the amount of PTFE ller into the PI matrix, the wear
rate increases quickly (PTFE presents a low hardness). For example, in
the case of a sliding velocity of 0.4 m/s, the wear rate of PI is
0.42 mm3/Nm against 14.02 mm3/Nm for PI40% PTFE. The addition
of SiC ller reduced the wear rate caused by the action of strong cyclical
solicitations and kept in the same time the low friction coefcients. Consequently, the addition of 5 wt.% SiC particles into the fully imidized PI
resin demonstrated an improved mechanical behavior by a slight increase in hardness. Moreover, the wear rate signicantly decreases in
the case of PI30% PTFE5% SiC with about 94% in comparison with
PI30% PTFE or with 86% for PI40% PTFE5% SiC in comparison with
PI40% PTFE. For PI20% PTFE5% SiC coating, the wear rate has a low
reduction of 8% in comparison with PI20% PTFE. The PI coating demonstrated a low wear rate because of its high hardness. From the point of
view of both friction and wear behaviors, it can be observed that the
addition of 20% PTFE and 20% PTFE + 5% SiC into PI matrix has the
best behavior in terms of a moderate friction coefcient, a low wear

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

609

Fig. 9. TGA curves of PI and PI composite coatings.

rate and an acceptable mechanical behavior. Under harder dry sliding

conditions, the addition of 15 wt.% PTFE and 10 wt.% CNT into a PI
matrix highly decreased the friction coefcient to 0.12 [18].
A correlation between the mechanical and tribological behaviors can
be noticed. The PIPTFESiC composite coatings which demonstrated a
good hardness exhibit less wear. Another trend suggested by this study
is the possibility to use the PIPTFESiC composite coatings for high
temperature applications, due to the thermal stability of the matrix.
Moreover, this is also underlined by the thermal stability of SiC and
PTFE particles like llers and their positive inuence on hardness, total
surface energy and friction coefcient of the composite coatings.

4. Conclusions
This study proposed to investigate the surface properties as well as
the tribological and mechanical behaviors of the PIPTFESiC composite
coatings, especially the effect of PTFE and SiC reinforcing agents. In addition, DSC and TGA investigations of the pure polyimide solution samples and PI and PI based composite coatings were performed. Based on
the results obtained with respect to PI composite coating investigations,
several conclusions can be made. PTFE addition has a positive inuence
on the surface properties of the PI composite coatings but negative with
respect to the mechanical and tribological behaviors in terms of hardness and wear rate of the PI composite coatings. The addition of the
SiC particles has no inuence on the surface properties of the PI based
composite coatings but signicantly improves the mechanical and tribological behaviors of the PIPTFESiC composite coatings.

Table 4
Composition of the PI composite coatings, according to TGA results.
Initial composition of the
composite coatings
(Before the furnace drying)

Final composition of the

composite coatings
(After the furnace drying)

PI20% PTFE
PI20% PTFE5% SiC
PI30% PTFE
PI30% PTFE5% SiC
PI40% PTFE
PI40% PTFE5% SiC

PI50% PTFE
PI44% PTFE12% SiC
PI63% PTFE
PI57% PTFE10% SiC
PI73% PTFE
PI66% PTFE9% SiC

DSC analysis revealed a high Tg point of the pure PI coating. Besides,

the Tg of the pure polyimide solution increases with the temperature
increase.
Sessile drop method investigations revealed that increasing the PTFE
amount up to 30 wt.% the PI composite coatings have a hydrophilic
character then a hydrophobic character is noticed for the PI40% PTFE
and PI40% PTFE5% SiC. The calculated total surface energy decreases
with the increase of the PTFE wt.%. Thus, PI20% PTFE5% SiC composite
coating demonstrated a lower total surface energy than the pure PI coating. Increasing the amount of PTFE to 30 and 40 wt.% and adding SiC
particles the total surface energy still decreases.
Moreover, the PTFE ller considerably decreases the friction coefcient of the PIPTFESiC composite coatings. Thereby, the addition of
PTFE signicantly contributes to the total surface energy decrease and
consequently to the reduction of the friction coefcient of the PI based
composite coatings. The hardness as well as the wear rate increases
with the increase of PTFE amount into the PIPTFE composite coatings.
Adding 20 wt.% of PTFE ller into the pure PI matrix is sufcient to
have interesting results with respect to the tribological and mechanical behaviors of the PI based composite coatings. Thus the PI20%
PTFE composite coating revealed a low friction coefcient. By the addition of 5 wt.% SiC particles, the PI20% PTFE5% SiC composite
coating still presents a low friction coefcient, a slightly decreased
wear rate and a satisfactory hardness in comparison with the other
PIPTFESiC composite coatings.
The TGA results showed that the PI coating is thermally stable at an
elevated temperature (above 400 C). Moreover, the addition of PTFE
and SiC llers into pure PI matrix thermally stabilizes the PIPTFESiC
composite coatings.
Acknowledgments
The authors are grateful to the company Evonik Industries-Evonik
Fibres GmbH, Austria for the polyimide P84 solution supply.
References
[1] J.H. Jia, H.D. Zhou, S.Q. Gao, J.M. Chen, Mater. Sci. Eng. A 356 (2003) 4853.
[2] A. Tanaka, K. Umeda, S. Takatsu, Wear 257 (2004) 10961102.
[3] C.-K. Min, T.-B. Wu, W.-T. Yang, C.-L. Chen, Compos. Sci. Technol. 68 (2008)
15701578.

610

C. Demian et al. / Surface & Coatings Technology 235 (2013) 603610

[4] G. Zhao, I. Hussainova, M. Antonov, Q. Wang, T. Wang, Friction and wear of ber
reinforced polyimide composites, Wear 301 (2013) 122129, http://dx.doi.org/10.10
16/j.wear.2012.12.019.
[5] S. Zhang, Y. Li, D. Yin, X. Wang, X. Zhao, Y. Shao, S. Yang, Eur. Polym. J. 41 (2005)
10971107.
[6] G.S. Bhat, R. Schwanke, J. Therm. Anal. 49 (1997) 399405.
[7] Q. Jin, T. Yamashita, K. Horie, R. Yokota, I. Mita, J.PolymSci. A 31 (1993) 23452351.
[8] M. Ghaemy, R. Alizadeh, H. Behmadi, Eur. Polym. J. 45 (2009) 31083115.
[9] Y.J. Song, S.H. Meng, F.D. Wang, C.X. Sun, Z.C. Tan, Thermochim. Acta 389 (2002) 1924.
[10] B. Suresha, G. Chandramohan, P.R. Sadananda Rao, P. Sampathkumaran, S.
Seetharamu, J. Reinf. Plast. Compos. 26 (6) (2007) 565578.

[11] S. Basavarajappa, K.V. Arun, J. Paulo Davim, J. Miner. Mater. Charact. Eng. 8 (2009)
379391.
[12] H. Cai, F. Yan, Q. Xue, W. Liu, Polym. Test. 22 (2003) 875882.
[13] H. Cai, F.Y. Yan, Q.J. Xue, Mater. Sci. Eng. A 364 (2004) 94100.
[14] P. Samyn, G. Schoukens, P. De Baets, Wear 270 (2010) 5772.
[15] P. Samyn, P. De Baets, G. Schoukens, B. Hendrickx, Polym. Eng. Sci. 43 (8) (2003)
14771487.
[16] L. Li, D.D.L. Chung, J. Electron. Mater. 23 (6) (1994) 557564.
[17] J. Bijwe, S. Sen, A. Ghosh, Wear 258 (2005) 15361542.
[18] S. Yijun, M. Liwen, F. Xin, L. Xiaohua, J. Appl. Polym. Sci. 121 (2011) 15741578.
[19] D.K. Owens, R.C. Wendt, J. Appl. Polym. Sci. 13 (1969) 17411747.