Tribology International 135 (2019) 189–199

Contents lists available at ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

Properties of Polyetheretherketone (PEEK) transferred materials in a PEEK- T

steel contact
Debashis Puhan, Janet S.S. Wong∗
The Tribology Group, Department of Mechanical Engineering, Imperial College London, Exhibition Road, London, SW7 2AZ, United Kingdom

ARTICLE INFO ABSTRACT

Keywords: Polyetheretherketone (PEEK) is a high performance polymer that can be an alternative to metal for some moving
Polyetheretherketone components in unlubricated conditions. During rubbing, PEEK is transferred to the counterface. The formation
Transfer film formation and properties of PEEK transfer films on steel and sapphire are studied by in-situ observations of PEEK wear
In situ triboplasma process, contact temperatures and triboemission, as well as FTIR and Raman spectroscopies ex-situ. Our results
In situ contact temperature
suggest that frictional heating alone may not be sufficient to generate PEEK degradation observed in the transfer
materials. Triboplasma observed during rubbing, together with mechanical shear, may promote generations of
radicals and degradation of PEEK, which subsequently influence the properties of PEEK transfer film and per-
formance of polymer-metal tribopair.

1. Introduction The formation of PEEK polymeric transfer films on metallic coun-

terfaces has been attributed to its adhesion properties [9]. Few studies
Polyetheretherketone (PEEK, see Fig. 1), a member of the poly- have characterised these films in details. Laux et al. [8], using a pin-on-
aryletherketone family, is a high performance polymer (HPP) widely disc geometry, concluded that the mean transfer film thickness and the
used in applications involving relative motion. Its good chemical re- amount of PEEK wear were not correlated. In addition, the transfer film
sistance, high strength and high wear resistance make it an ideal ma- was the thickest when the sliding motion was perpendicular to the
terial for tribological applications [1,2] and hence PEEK has been used counterface roughness direction. The wear rates increase with in-
as alternatives to various metallic components such as seal rings, creasing contact pressure. Zhang et al. conducted a FTIR analysis of
transmission gears, bushings, sliders, bearings and valves. In these ap- worn PEEK surfaces and suggested that changes in chemical bondings
plications, it is common that PEEK is rubbed against a metal counter- and crosslinking improved the adhesion of plasma treated PEEK to steel
face. The properties of the PEEK-metal interface thus govern the tri- countersurface [10]. The steel surface however was not characterised.
bological performance of PEEK-metal tribopairs, including contact Properties and the formation of transfer films are commonly be-
temperature, friction and wear [3–5]. lieved to be contact temperature driven. Other factors, including tri-
A polymeric transfer film may form on the counterface when a boemission, may also play a role. It has been found that plasma can
polymer is rubbed against metal. It can be formed by physical forces cause polymer degradation [11] and affect friction and wear [12]. The
such as adhesion or by chemical reactions [6,7]. PEEK films have been aim of this paper is to examine factors that contribute to the properties
described as continuous, uniform, and thick [8]. Yet, there is no stan- of PEEK transfer films in a PEEK-steel contact in non-lubricated con-
dard definition of a good transfer film. Usually a smooth and con- ditions. This understanding is achieved by observing in-situ PEEK ma-
tinuous transfer film is considered beneficial for protecting the polymer terials transfer process to counterfaces. A rotating PEEK disc was
surface from the hard-metallic asperities, reducing wear. The material rubbed against a stationary steel or sapphire ball. Chemical and topo-
transfer process and properties of transferred materials are thus crucial graphical analysis were conducted on transferred materials. Results
for the development of better materials with increased durability and were correlated with contact temperature and plasma (UV emission)
lower friction. However in-depth understanding of this process is generated during rubbing.
lacking. Only qualitative and often subjective descriptions are made.
Hence, determining the correlations among transfer film properties,
friction and wear have been challenging.

∗
Corresponding author.
E-mail address: [email protected] (J.S.S. Wong).

https://doi.org/10.1016/j.triboint.2019.02.028
Received 11 December 2018; Received in revised form 15 February 2019; Accepted 18 February 2019
Available online 26 February 2019
0301-679X/ © 2019 Elsevier Ltd. All rights reserved.
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 1. Chemical structure of Polyetheretherketone (PEEK).

2. Materials and methodologies

2.1. Materials system

Fig. 2. Schematic of the friction test experimental setup.

A grade of commercially available Victrex PEEK was cut from an

injection moulded sheet stock of 10 mm thickness. Its glass transition Table 2
temperature (Tg) and a melting temperature (Tm) are 143 °C and 343 °C Operating conditions.
respectively. The crystallinity of the sample, determined using differ- PEEK-Sapphire PEEK-Steel
ential scanning calorimetry, was less than 30%. The 6 mm diameter
steel balls (AISI52100) were purchased from PCS Instruments Ltd and Ball diameter mm 5 6
Load N 10
sapphire hemispheres of 5 mm diameter were purchased from Swiss
Speed m/s 2
Jewel, USA. The roughness of the disc (Ra) was 1–1.2 μm while that of Time s 1800
the balls were less than 0.02 μm. The material properties of the balls Sliding distance m 3600
and discs are listed in Table 1. Initial average pressure, Pmean GPa 0.12 0.11
Initial maximum pressure, Pmax GPa 0.18 0.16
New disc and ball specimens were used for each test and no lu-
Semi contact width (a) m 0.16 × 10−3 0.17 × 10−3
bricant was used. Prior to tests, steel and sapphire balls were cleaned
with toluene (analytical reagent grade, 99.8%, Fischer Chemicals) in an
ultrasonic bath and washed with isopropanol (analytical reagent grade, PEEK-sapphire contact in-situ [17]. A FLIR X6540 SC camera, with a
99.8%, Fischer Chemicals) while the discs were wiped with lint-free high resolution 640 × 512 pixel detector sensitive to IR wavelengths
wipes soaked in isopropanol. Both ball and disc samples were thor- from 3 to 5 μm, was positioned above the stationary sapphire hemi-
oughly dried using a drier before each test. All solvents were used as sphere. Thermal radiation from the contact was transmitted through the
received. All test were done at room temperature (22–27 °C) with re- IR transmissive sapphire specimen and was collected through a
lative humidity (40–55%). 5 × objective lens. IR images were captured at 4 Hz for 30 min of
sliding. Note that the temperature rise at the asperity level cannot be
2.2. Friction test measured since such heat source is very short lived (10−3–10−5 s) [18].
Hence only the steady state nominal temperature at the contact is ob-
Coefficient of friction was measured with a CETR UMT tribometer in tained.
a ball on disc configuration. A point contact was created by a stationary The main contribution to the radiation detected by the IR detector
steel ball loaded against a unidirectional rotating PEEK disc (see Fig. 2). are from (1) PEEK disc, (2) sapphire hemisphere, and (3) reflected
Test conditions are listed in Table 2. Average coefficient of friction is ambient radiation. In this work, the emissivity of PEEK and sapphire are
reported from the average of three tests. An electrometer (Keithley taken as 0.95 [19] and 0.45 [20] respectively, while the contributions
6517B) connected to a homebuilt charge sensor (details in Ref. [16]) from secondary reflection and reflected radiation were neglected as in
was used to record the charge accumulation on the disc surface. The Ref. [21]. Having said that, the contact temperature obtained may be an
sensor was placed diametrically opposite to the contact, with its tip overestimation. The bulk temperature of the sapphire hemisphere was
positioned 2 mm above the wear track. monitored by a K type contacting thermocouple to ascertain the lower
A K type thermocouple was attached to the side of the ball with high bound of the contact temperature.
temperature superglue to monitor the temperature of the ball. RS Pro
1384 Data Logger recorded the temperature at 1 Hz during the test. The 2.4. Plasma observations
disc surface temperature was measured using an Omega IR-USB in-
frared sensor, suitable for non-contact temperature monitoring of Based on the setup shown in Fig. 2, the PEEK-steel sliding contact
polymers. The temperature was recorded at 1 Hz by a PC using a was viewed from the side to observe the plasma generated using a UV
dedicated software. sensitive camera (Hamamatsu ImagEM X2). Images were captured with
a 5 × objective (Mitotoyo) at 8 frames/second under no background
2.3. In situ IR thermography light conditions.
The optical spectrum of the generated plasma was obtained using a
Infrared IR thermography was used to obtain temperature of a spectrometer (Ocean Optics QE Pro) coupled with a fibre optic through
a stationary sapphire hemisphere when it was rubbed against a rotating
Table 1
PEEK disc.
Properties of the ball and disc material.
2.5. Surface analysis using SEM-EDS, FTIR and Raman
Properties PEEK Disc [13] Sapphire Ball Steel Ball
[14]
Optical images were captured using a HIROX RH 2000 microscope.
Density kg/m3 1320 3980 7700 SEM images were taken with a Hitachi S3400 instrument in secondary
Thermal Conductivity W/ 0.29 46 21 [15] electron mode and back scattered electron mode at 20 kV. The energy
m-K
dispersive X-ray (EDX) mode on the SEM was used for elemental ana-
Thermal Diffusivity m2/s 1.47 × 10−7 1.51 × 10−6 1.29 × 10−5
Specific heat J/kg-K 1340 7610 461 lysis of the test samples.
FTIR and Raman spectroscopies were used to characterise PEEK

190
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

transferred materials on counterfaces. A Fourier transform infrared coloured streaks (see Tables S2–1, Figs. S2–1; and Tables S2–2, Figs.
(FTIR) Spectrometer (Spectrum 100, Perkin Elmer, USA) was used to S2–2 for EDX results from sapphire and steel wear scars respectively).
collect IR spectra. Each spectrum consists of 100 scans performed be- Micro-cracks are observed on the thin streaks, suggesting that they are
tween 700 cm−1 and 3600 cm−1 at a wavenumber resolution of brittle (see Fig. 4g and i). The cracks may form due to repetitive loading
4 cm−1. IR spectra collected were distorted due to the low reflectance [22]. These may also be associated with the deformation of the polymer
of PEEK disc. Kramers-Kronig conversion was applied to obtain cor- below its Tg or a change in flexibility of polymer chains arising from
rected absorption spectra. chain scission [23] or UV irradiation [24].
A HORIBA Labram HR Evolution Raman spectroscope, with a Wear tracks on PEEK discs formed by PEEK-sapphire and PEEK-steel
785 nm laser, was used to collect Raman spectra. To avoid sample de- contacts are shown in Fig. 5. They show grooves, and fish-scale-like
gradation by laser heating, the laser power was kept as low as possible. features. Grooves occur when hard asperities of the ball surface plough
Results were obtained from 20 scans of 1 s each between 700 cm−1 and through the softer polymer. It is more severe in the PEEK-steel contact.
1800 cm−1 using a 20 × objective. Information of peak assignments for Fish-scale-like patterns could be formed due to tensile drawing followed
FTIR and Raman spectra can be found in section S-1. Note that SI stnads by compressive ironing of the polymer surface [25] in the presence of
for Supplementary Information. Sections, figures and table numbers flowing and non-flowing regions of the polymer, giving rise to stick-slip
starting with 'S' are in SI-1. motion [26]. The parabolic folds or ridges suggests that the temperature
may have been below the Tg [10]. These results suggest that in these
3. Results and discussions contacts both abrasive and adhesive wear occurred.
Factors that govern polymer wear processes include materials
3.1. Summary of friction test results properties, test conditions and contact conditions. The similarity of the
wear scar and wear tracks obtained from PEEK-sapphire and PEEK-steel
Fluctuations in frictional forces, and hence friction coefficients, are contacts is likely due to the severity of the chosen test condition. As a
observed during friction tests (Fig. 3). Friction coefficient of the PEEK- result, PEEK in both contacts undergoes similar deformation and de-
steel contact shows larger variation than that of the PEEK-sapphire gradation. This allows us to apply results obtained from in-situ wear
contact. Since friction tests were conducted in unlubricated conditions, and contact temperature measurements from a PEEK-sapphire contact,
this level of variations among tests are common and deemed accep- presented in sections 3.2.1 and 3.3.1 respectively, to a PEEK-steel
table. Average steady state friction coefficients of the PEEK-steel and contact.
PEEK-sapphire contacts are around 0.26 ± 0.05 and 0.17 ± 0.01 re-
spectively. 3.2. PEEK wear process
After friction tests, rubbing surfaces were examined. While more
debris are found with the PEEK-steel contact (Fig. 4a), features on 3.2.1. In-situ wear observations
rubbed steel and sapphire surfaces are similar. Loose debris, including A PEEK-sapphire contact was monitored in-situ through the sap-
thick accumulations at the inlet and outlet of the contact (labelled as phire hemisphere during rubbing. This allowed the PEEK wear process
sheared film, see Fig. 4b and d), can be removed easily by wiping the to be observed in real-time (Video SI 2). Snapshots of in-situ wear
surface with cleaning tissue paper. Even more debris are removed after process in PEEK-sapphire contacts (Fig. 6) show that debris are formed
sonication in acetone (compare Fig. 4b vs. 4c and Fig. 4d vs. 4e), after by two processes (i) primary debris are formed by the detachment of the
which strongly adhered PEEK transferred materials is exposed. Transfer ploughed out or deformed materials; and (ii) secondary debris formed
layers on wear scars are heterogeneous and consist of two distinct re- by shredding (tearing) or shearing of the primary debris.
gions: patches with a thickness of about 2–4 μm, referred as transfer Supplementary video related to this article can be found at https://
films; and rainbow coloured regions which are very thin films (about doi.org/10.1016/j.triboint.2019.02.028.
20–200 nm, see Fig. 4b and c for definitions). Elongated patches along During rubbing, long strands and plate like debris are formed. Long
the sliding direction are termed as streaks. strands (see Fig. 6c) are more common at the beginning of sliding when
The strongly adhered materials in wear scars were examined with sharp asperities of the countersurface plough through the soft polymer
backscattered SEM electron images (Fig. 4g–i). Heavy and light atoms surface, resulting in groves on the polymer surface. As sliding con-
appear brighter and darker respectively in these images. The darker tinues, these asperities become blunt either by fracture due to the shear
regions are polymeric, consisting of carbon and oxygen. Both counter- stress acting on them or are covered with the softer transferred mate-
face materials and PEEK-based materials are detected in the lighter- rials. Thermal effects and fatigue may also become more prominent and
polymer is removed by a combination of ploughing and micro-cutting.
The resulting plastic flow may give rise to plate-like debris as shown in
Fig. 6d–f.
As a result of repeated cyclic loading, the deformed, ploughed out
polymer may be removed from the contact and the wear track, forming
loose debris (see Fig. 6h–i). They may accumulate at the inlet (Fig. 6f).
Some of them re-enter the contact and may be shredded into smaller
pieces (Fig. 6d and g). They may re-adhere to the polymer surface or
become trapped between asperities of the counterface by compression
[6,27], forming strongly adhered thin film or transfer film.

3.2.2. Analysis based on FTIR spectra

FTIR spectra of PEEK transferred materials on steel balls, PEEK wear
track and pristine PEEK disc are shown in Fig. 7. Important observa-
tions are summarised in Table 3. Details of the IR spectra can be found
in Figs. S3–1 and Tables S3–1.
To facilitate discussion, seven zones are highlighted in Fig. 7. The
spectrum of thin rainbow coloured film on steel ball wear scar (black
Fig. 3. Coefficient of friction of PEEK-steel and PEEK-sapphire contacts re- line, no symbol) is mostly featureless, with a broad peak between 1200
corded during rubbing. and 700 cm−1 (zone 1). This suggests that the ketone and ether groups

191
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 4. (a) PEEK debris (appeared white) on a

steel ball after friction test; (b) PEEK transferred
materials on a steel ball right after the test; (c)
PEEK transfer film on a steel ball after cleaning;
(d) PEEK transferred materials on a sapphire ball
right after the test; (e) PEEK transferred mate-
rials on a sapphire ball after cleaning; (f) SEM
image of PEEK transferred materials on a steel
ball right after the test; (g) Magnified view
showing cracks in the square region on image (f)
in backscattered electron mode (BSE); (h)
Backscattered electron scanning micrograph of
PEEK transferred materials on a sapphire ball
right after the test; (i) Magnified view showing
cracks and fracture in the square region on
image (h). Cleaning refers to wiping and soni-
cation in acetone.

Fig. 5. Wear tracks on PEEK discs formed by (a) PEEK-steel and (b) PEEK-sapphire contacts.

present on the PEEK chain have been severely oxidized and the remains Changes in FTIR spectra from transferred materials, other than thin
are mainly amorphous carbonaceous material. The low thickness (in films, show peak shifts compared to that of pristine PEEK, and hence
sub-micron range) of thin films may make their identification difficult suggest that rubbing promotes chemical changes of PEEK (see zone
by FTIR in reflection mode. 2–4). The relative intensities and shoulders of many peaks are altered,

192
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 6. Snapshots of a PEEK-sapphire contact imaging at 50 Hz, showing the wear process of PEEK in situ.

In case of PEEK transferred materials, the bands are broader, and

new peaks and shoulders appear. Generally, the bands related to car-
bonyl stretching and ring vibration (zone 3) shift towards amorphous
(higher) wavenumbers while bands related to out of plane bending
modes from the aromatic hydrogens (1000–700 cm−1) towards crys-
talline (lower) wavenumbers. The peak shifts towards higher wave-
numbers suggests a decrease in bond length. This may occur due to a
change in electronegativity of the neighbouring atom caused by cross-
linking after bond scission, or loss of hydrogen or electrons. The
broadening of peaks is due to increased intermolecular interactions
such as hydrogen bonding. Chain scission and crosslinking may occur
simultaneously which leads to a decrease and an increase in crystal-
linity respectively. Chalmers et al. found a relationship between the
degree of crystallinity and the ratio of the absorbance at band around
1306 to that around 1280 cm−1 [28]. Based on this relationship, our
results suggest that the degree of crystallinity in descending order:
PEEK disc > PEEK wear track > transferred film > sheared film >
debris.
Fig. 7. FTIR spectra of pristine PEEK disc, PEEK wear track and various PEEK Peaks not usually associated with PEEK are found in spectra of PEEK
transferred materials on steel. transferred materials. A peak around 1790 cm−1 (between zone 4 and
5, Fig. 7), which is linked to the formation of aryl and α,β-unsaturated
suggesting that the transfer films, sheared films and debris underwent acid peroxides [29], is observed. Bands of overtones due to the aromatic
some degree of degradation such as oxidation, crosslinking, changes in ring vibrations appear in the region 2000–1800 cm−1 (zone 5, Fig. 7).
molecular orientation and substitution, as summarised in Table 3 (see They are indicative of the distribution, and positions of substituents in
also Section S-4). the aromatic rings [29]. The peaks 1884 and 1922 cm−1 arise from the
The FTIR spectrum of the PEEK wear track (green triangle) have 1,4 (para) di-substituted benzene ring. Peaks at 2072 cm−1 and at
bands that are narrower and shift towards slightly higher frequencies as 2028 cm−1 are characteristics of symmetric and asymmetric stretching
compared to pristine PEEK (navy pentagon), which suggests a slight of CeO bond respectively [30] which may suggest the formation of gym
increase in crystallinity on the wear track. dicarbonyl species (CO-Ph-CO or CO-Metal-CO). Their intensities vary

193
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Table 3
Observed changes in spectra of PEEK transferred materials on steel as compared to the spectrum of pristine PEEK and possible mechanisms of such changes.
Wavenumber (cm−1) Peak assignment Observation Inference Driving factors

3000–3500 Polymeric hydrogen-bonded OH stretching New Phenolics or Alcohols bands Oxidation UV irradiation
2700–2800 HeC]O New Aldehyde bands Oxidation UV irradiation
1730-1740,1780 OeC]O Shoulder of Ester bands Oxidation Thermal treatment
Ion irradiation
1200–1100 CeOeC (bending and stretching) Shifting, broadening and splitting of Ether bands Crosslinking & Chain scission Thermal treatment
Ion irradiation
1500 C=C phenyl ring vibration Decreased intensity Oxidation UV irradiation
Thermal treatment
Ion irradiation
1000–800 CeH of phenyl rings Decreased intensity Reduced crystallinity Mechanical shearing
Thermal treatment
Ion irradiation

as PEEK debris > sheared film > transferred film. intensity at 1226 cm−1 and the relative intensity increase of peaks at
Peak at 1635 cm−1, which is linked to hydrogen bonded carbonyl 1653, 1599, and 1493 cm−1 observed in this work (peaks between
(ν(CO …..HeOeH)) or ν(CO)α-hydroxyl diaryl ketone groups and is zones 3 and 4, Fig. 6) have been attributed to a change in chemical
usually associated with PEEK [31], is not observed in spectra of PEEK bonding due to plasma treatment [9] and melting [35,39]. The band
transferred materials. This suggests that the carbonyl bonds are not 3044 cm−1 is known to increase with an increase in ion irradiation [40]
hydrogen bonded. Peaks at 1750, 1745, 1715 cm−1, which are linked to and decrease with UV exposure [38,41,42]. The roles of ion/UV irra-
production of new carbonyl species such as esters and fluorenone type diation and temperature in the chemistry of the transferred materials
structures, are also absent. Therefore, ester and fluorenone type struc- are discussed in section 3.3. Note that severe chain scission and mas-
tures are unlikely to have formed in our PEEK-steel rubbing contacts. tication of polymer chains can also lead to very thin carbonaceous films
Very weak peaks characteristic of aldehydes (O]CeH) stretch are bound to the counterface by strong force of adhesion.
observed in the region 2830-2695 cm−1 (zone 6, Fig. 7) for the debris
and sheared films (see also Figs. S3–1 and Tables S3–1). The peaks 3.2.3. Analysis based on Raman spectra
between 3000 and 3100 cm−1 (zone 7, Fig. 7) such as 3100, 3088, 3044 The FTIR spectra suggest that thin films present in the steel wear
and 3008 cm−1 are generally related to the stretching vibrations scar underwent the most severe degradation with complete oxidation of
of = CeH stretch from aromatic rings. In particular, the 3044 cm−1 ketone and ether moieties, leading to the formation of amorphous
absorption band has been assigned to phenyl ring stretching vibration carbonaceous film. Rebelo de Figueiredo et al. observed similar PEEK
ν(C3eH) and ν(C5eH). The intensities of these bands vary in the order degradation after rubbing using Raman spectroscopy (with 532 nm
sheared film > debris > transfer film > PEEK wear track. excitation). D and G bands of amorphous carbon around 1350 and
A broad peak with a maximum around 3330 cm−1 and another 1580 cm−1 were found on the transfer films on steel surfaces [43] al-
band around 3270 cm−1 are observed only in the case of the central though the spectrum of PEEK wear track did not contain peaks related
transfer film and is linked to the hydroxyl absorption (see Figs. S3–1 for to PEEK degradation. Instead peaks related to iron oxides hematite
detailed spectra). The former lies in the range of polymeric hydrogen- (Fe2O3) and magnetite (Fe3O4) were present. This is inconsistent with
bonded OH stretch while the latter is in the range linked to OH vibra- our FTIR results. It is unclear if the authors have taken into account the
tion of acid groups [32]. background fluorescence of PEEK, which may hinder PEEK degradation
Our results suggest that rubbing promotes chain scission, opening of detection. To reduce PEEK background fluorescence, Raman spectro-
the aromatic rings, substitution, crosslinking, along with loss of crys- scopy using a 785 nm laser for excitation was employed in this study
tallinity, and coplanarity of the rings. Degradation associated with ring (see Fig. 8, see also Figs. S4–1 for Raman spectra obtained with 532 nm
opening, crosslinking and substitution is possible only by radical gen- excitation). Conventionally, the relative intensities of spectral modes is
eration which requires dehydrogenation/hydrogen abstraction. obtained by keeping the intensity of the most intense band, the
Random chain scission at the ether and ketone linkages must occur first 1146 cm−1 peak in this case, constant. However, this peak has been
to generate stable radicals intermediates [33,34]. Dehydrogenation has reported to change [44,45]. Consequently, the second most intense
been shown to be affected by thermal treatment [35,36], UV treatment band at 1595 cm−1 is fixed, as shown in Fig. 8.
[32,37,38], and ion irradiation [31]. The relative decrease in the peak Under 785 nm excitation, Raman peaks of PEEK are observed at

Fig. 8. Raman spectra (with 785 nm excitation) of clean steel ball, pristine PEEK disc, PEEK wear track, and transfer film and thin film on steel wear scar.

194
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Table 4
Relative intensities of the Raman modes of PEEK.
Wavenumber Pristine Sheared Materials Change
PEEK Disc relative to
PEEK Transfer film Pristine PEEK
Wear
track

1596.41 (CeC) fixed 0.549 0.535 0.721

1644.55 (C]O) 0.235 0.258 0.379
I1644/1596 0.427 0.482 0.525 positive
1201.76 (CeOeC)assym 0.212 0.190 0.241
I1201/1596 0.385 0.356 0.335 negative
1146.83 (CeOeC)sym 0.994 1 0.975
I1146/1596 1.808 1.866 1.352 negative
808.35 (CeH) 0.471 0.468 0.482
I808/1596 0.857 0.874 0.669 negative

808, 1146, 1202, 1596, 1608 and 1644 cm−1 (see Fig. 8a) [45,46]. The
peak for the C]O stretching mode (1644 cm−1) have the same wave- Fig. 9. Bulk temperature of balls and discs during rubbing (one data point is
number for the pristine PEEK disc, the PEEK wear track and the transfer displayed for every 50 points collected for clarity).
film. This suggests that they have similar crystallinity. This contradicts
our observations with FTIR studies and may be due to the relatively low the ball is stationary and is thermally more conductive than the PEEK
signal-to-noise ratio in Raman signals. A broader phenyl ring stretching disc. The bulk temperatures of steel (stars, Fig. 9) and sapphire balls
vibration band (1595 cm− 1) is observed in the Raman spectrum of the (triangles, Fig. 9) are about 100 and 70 °C respectively, and bulk tem-
transfer film. New broad bands also appear around 1400 cm−1, be- peratures of their corresponding PEEK disc are about 40 (squares,
tween 1200 and 1300 cm−1, and between 900 and 1000 cm−1 (see Fig. 9) and 30 °C (circles, Fig. 9) at the end of the tests.
arrows, Fig. 8a). These new bands may arise from the formation of During rubbing, a sharp temperature rise, called the flash tem-
other aromatic rings [47] or ring opening reactions by oxidative attack perature, occurs at the rubbing interface due to frictional heating
on phenyl moieties [41]. [51,52]. For the PEEK-sapphire contact, the contact temperature was
The relative peak heights of stretching mode of ether (CeOeC) obtained by IR thermography. Note that the contact temperature de-
linkage (1201 and 1146 cm−1), stretching mode of CeH (808 cm−1) pends on pressure, which changes with positions in the contact. Local
and C]O stretching mode (1644 cm−1) to CeC stretching temperature can be raised at a local asperity level. This is however a
(1596 cm−1), denoted as I1201/1596, I1146/1596, I808/1596 and I1644/1596 transient phenomenon, lasts only as long as 10 μs, and is difficult to
respectively, obtained from various spectra are shown in Table 4. I1201/ measure. The local temperature can also increase as debris accumulate
1596, I1146/1596, I808/1596 are lower, while I1644/1596 is higher for the around the contact and debris entrain into the contact, causing the local
transfer film than those of the pristine PEEK. Previous studies have pressure to change. This is confirmed by IR thermographic snapshots
shown that the relative intensity of 1146 cm−1 peak (CeOeC stretch for a PEEK-sapphire contact (Fig. 10, see also Video SI 3). Part of the
mode) in PEEK increases with temperature until about 100 °C and de- contact is above 100 °C almost as soon as rubbing starts (see Fig. 10b,
creases until about 120 °C and then increases again until melting measured temperature can be found in Fig. S5–1). The temperature at
[45,48]. The opposite trend is observed in C]O stretch in the same the inlet, where the ploughed material accumulates and is sheared
temperature range. These changes were linked to the thermally induced (Fig. 10c–h), increases. Some of the sheared materials are drawn into
movement of the ether linkages of PEEK below Tg [45,48]. The results the contact where they are shredded. The formation and breaking of
here thus suggest that the temperature experienced by PEEK transferred sheared films is a chaotic process. Some are swiped around the contact,
materials are likely to be below Tg, and the highest temperature that the while other are eventually entrained into the contact, forming local hot
transfer film is subjected to is between 100 and 120 °C. spots. The high local temperature the sheared films are subjected to
The Raman spectrum of a clean steel ball shows a broad asymmetric may prompt their degradation. The longer they remain in the contact
band between 600 and 900 cm−1 (Fig. 7b) as the A1g mode of magnetite the more heat, and hence more degradation they experience. For the
(Fe3O4) at 669 cm−1 superimposes with the broad maghemite band (γ- entrained debris, the high temperature makes them more susceptible to
Fe2O3) at 720 cm−1 [49]. The peak at 1380 cm−1 is attributed to he- shear deformation and degradation.
matite (α -Fe2O3). A weak peak at 1060 cm−1 corresponds to α-FeOOH It is anticipated that similar phenomena applies to the PEEK-steel
is also observed. Apart from these oxide related peaks, the spectrum of contact. Since the bulk temperatures of the PEEK-steel contact are
the thin film on the steel ball wear scar also shows bands around 1350 higher than those of the PEEK-sapphire contact (see Fig. 9), the contact
and 1580 cm−1. They are linked to the D and G bands of amorphous temperature for the former is likely to be higher.
carbon, although the former can also be assigned to γ -Fe2O3 [50]. A Heating PEEK at 400 °C in air has been shown to cause decom-
new broad band between 400 and 600 cm−1 (dotted rectangle, Fig. 7b) position, chain branching, and crosslinking [53]. Zhang et al. estimated
is observed on the spectrum of the thin film. This can be linked to iron interfacial temperatures in the range 300–345 °C at which PEEK plastic
oxides such as δ-FeOOH, Fe3O4 and FeO [50]. It should be noted that no flow occurs based on the chemistry of PEEK wear debris [39]. It was
signature of PEEK is found. This supports the FTIR results that thin films suggested that thermo-oxidative cross-linking reactions occurred in
on steel wear scar are highly degraded. PEEK which reduced its crystallinity [39]. However, our results support
that the contact temperature is only near the Tg of PEEK. These contact
3.3. Factors influencing PEEK degradation temperatures match the temperature suggested by results in the Raman
study (section 3.2.3) and are alone insufficient to cause thermal de-
3.3.1. Frictional heating gradation of PEEK. While our results cannot eliminate the existence of
Bulk temperatures of the ball and the disc were measured during high flash temperature due to the lack of temporal resolutions, they
rubbing. The steel ball (stars, Fig. 9) always has higher temperature support that factors other than temperature, such as UV/ion irradiation
than the PEEK disc (squares, Fig. 9). This is because in our experiments due to triboemission, assisted by mechanical shear, may contribute to

195
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 10. Snapshots of in-situ temperature measurement using IR imaging. The colour bars are temperature scales which change among images. (For interpretation of
the references to colour in this figure legend, the reader is referred to the Web version of this article.)

the change in chemistry of transferred materials evidenced by FTIR charged wear track. Occasionally, a glow at the edge of the steel ball is
studies. observed as well as a localized corona discharge (Fig. 11f and h). The
overall intensity of the transient plasma luminescence reduces as sliding
3.3.2. Triboemission progresses. A detailed explanation of plasma luminescence from
Plasma, consisting of electrons, ions and photons, is known to polymer contact can be found in Ref. [58].
promote surface modification through processes such as etching, chain
scission, cross-linking and chemical functionalization [54]. Similarly 3.3.2.2. Potential links between plasma generations and chemistry of PEEK
triboplasma may modify rubbing surfaces by producing high energy transferred materials. Although some of the chemical changes reported
active sites which are extremely reactive [55]. Indeed triboelectrifica- in Table 3 have been linked to ion irradiation on PEEK surface
tion and triboemission (emission of electrons) are known to occur [31,40,59,60], PEEK in general has very good resistance to ion
during rubbing and have been linked with polymer chain scission [56]. irradiation [60]. Its UV resistance however is weak and it adsorbs UV
Note that some features in FTIR spectra of PEEK transferred materials in the range, λ = 300–380 nm [41]. The damage from UV irradiation is
are similar to those of plasma treated PEEK in literature [9,31,38,57]. supported by the fracture of transfer films shown in Fig. 4g and i [24].
Hence, the effect of triboplasma on the rubbing surfaces and transferred Based on the plasma emission spectrum in Fig. 12, the energy of the
materials must be considered. UV photons is in the range 3.2–4.43 eV. This is sufficient to break
bonds, including CeC (3.48 eV), CeH (4.13 eV), OeH (3.66 eV), CeO
3.3.2.1. Observations of triboplasma. The PEEK wear track was (3.6 eV). The resulting radicals can participate in various reactions. A
positively charged, (see Fig. S6–1). Consequently, an electric field is competition between chain scission and crosslinking in PEEK under UV
created between the steel ball and the charged PEEK wear track. When irradiation may occur [38,41,42,61]. This could be a reason for IR
the strength of the electric field surpasses the dielectric strength of the spectra of transferred materials to have some peaks (1457–1453, 1415,
air (3 kV/mm for air), electric breakdown of air occurs by a corona 1310, 1281, 1226, 1159, 1187, 1115, 1101, 864, 847, 838, 870, 964,
discharge process [70]. Plasma streamers are generated as electrons of 951, 928, 768 cm−1) shifted towards crystalline bands while other
the neutral steel ball (earthed conductor) are pulled towards the peaks (1647, 1593, 1500, 1493, 1485 and 1410 cm−1) towards amor-
charged wear track. This is observed at the gap between the phous bands (see Table S3–1).
stationary ball and a rotating PEEK disc, and along the wear track The presence of bands corresponding to H*, OH*, CH* and C2* in
(see Fig. 11, also see Video SI 4). The spectrum of the emitted plasma the plasma emission spectrum confirm the formation of radicals R* in
resembled the N2 breakdown spectrum as shown in Fig. 12. the polymer. These radicals may promote PEEK degradation. For ex-
The intensity of plasma varies with time. Initially, plasma is ob- ample, CH* may react with unsaturated carbon-carbon bonds by ad-
served in the gap between the ball and the disc at the outlet. After a dition across the pi-bonds of neighbour radicals. The free hydrogen
while, it is also seen at the inlet (Fig. 11f). The variation is linked to the radicals, which can diffuse very easily through the polymer matrix, may
gap geometry and the electric field generated, which change due to the cause hydrogen abstraction of polymer chains, or recombine with a
generation of wear debris, materials transfer and the re-entering polymer radical (crosslinking) or with hydrogen radicals. H* may also

196
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 11. Plasma generated when a steel ball was rubbed against a PEEK disc. Note that the brightness and contrast has been adjusted in (f) and (h) to show the glow
along ball edge and localized corona discharge events.

induce oxidation and cleavage of the rings of intermediate species to

form chain aliphatic acids in the presence of reactive oxygen species
(e.g. ReOO* and OH*). ReOO* are known to form when alkenyl or
phenacyl radicals react with molecular oxygen, which is available in
PEEK [67]. Peroxy-radicals can form at temperatures < 100 °C in pre-
sence of UV light from the plasma [68,69].
The presence of a transition metal such as iron (Fe2+) at the in-
terface may promote the reduction of the radicals and increased che-
lation. Radicals are known to react with metal ions via the hydro-
genation of the organic radical [70–73]. Moreover, metal carbonates
(ester) may be generated by the linkage of CO* radicals with the co-
ordinatively unsaturated Lewis acid-base pair sites (Mn+-O2−) on the
counterface. These reactions may be responsible for the strong adhesion
observed with the transfer materials on steel surfaces.
Note that the scission of polymer chains and formation of primary
radicals may occur due to mechanical shear [7,74–77]. The primary
Fig. 12. Emission spectrum of the observed triboplasma. Strong emission is
(end) radicals are very reactive and react readily with adjacent polymer
observed from the first negative system of the nitrogen molecular ion (N2+
molecules to form secondary (internal) radicals, producing low mole-
(B2Σu-X2Σg), 329.3 nm–586.5 nm) and the second positive system of the ni-
trogen molecule (N2 (C3Πu-B3Πg), 281.4 nm–497.6 nm). Weak emissions from cular weight fragments through auto oxidation [69]. For PEEK, Zhang
the first positive system of the nitrogen molecule (N2 (B3Πg-A3Πu), et al. have suggested that friction and wear can lead to the diphenyl
503.1 nm–800 nm) (about 100 times weaker) [62–64] and nitric oxide γ-system ether segment undergoing chain scission, and the formation of oxida-
(NO (A2Σ - X2Π), 250–300.8 nm) [65] are also detected. Non-nitrogen radicals, tive crosslinks by combining with adjacent radicals [39]. Hence the
such as CH (A2Δ-X2Π) and C2 (A3Πg -X2Πu), including bands at 400, 425 and existence of both triboplasma and mechanical shear are likely to ex-
475 nm, may be generated from polymer bond scission and triboemission of acerbate PEEK degradation.
volatile products [66].

197
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

Fig. 13. Summary of factors influencing properties of polymeric transfer film on rubbing counterface.

4. Conclusions Appendix A. Supplementary data

When PEEK is rubbed against sapphire and steel, it is transferred to Supplementary data to this article can be found online at https://
the counterfaces under our test conditions. The formation of PEEK doi.org/10.1016/j.triboint.2019.02.028.
transfer layers was examined by in-situ monitoring of the wear process,
contact temperature, and triboplasma generation. As rubbing starts, the References
PEEK surface is initially ploughed by the asperities of the steel ball.
Some of these materials are entrained and sheared in the contact. [1] Lu ZP, Friedrich K. On sliding friction and wear of PEEK and its composites. Wear
Debris form, as well as materials transfer occurs. 1995;181–183:624–31. https://doi.org/10.1016/0043-1648(95)90178-7.
[2] Bijwe J, Kadiyala AK, Kumar K, Puhan D, Parida T, Trivedi P. Development of high
The chemistry of PEEK transferred materials on wear scars differ performance poly (ether-ketone) composites based on novel processing technique.
from that of pristine PEEK. The thin film, which are formed between the Mater Des 2015;73:50–9. https://doi.org/10.1016/j.matdes.2015.02.007.
thicker transfer films and the counterface, is mainly amorphous car- [3] Freidrich K. Wear of reinforced polymers by abrasive counterparts vol. 1. 1986.
https://doi.org/10.1016/B978-0-444-42524-9.50012-0.
bonaceous materials. FTIR results of other PEEK transferred materials [4] Briscoe BJ, Lin Heng Yao, Stolarski TA. The friction and wear of poly(tetra-
suggest scission of PEEK chains occurs at various positions in the ether fluoroethylene)-poly (etheretherketone) composites: an initial appraisal of the op-
and ketone groups. In addition, opening of the aromatic rings, sub- timum composition. Wear 1986;108:357–74. https://doi.org/10.1016/0043-
1648(86)90013-X.
stitution, crosslinking, along with loss of crystallinity, and co-planarity
[5] Puhan D, Bijwe J, Parida T, Trivedi P. Investigations on performance properties of
of the rings are observed. Carbonate and carboxylic acid may form and nano-micro composites based on polyetherketone, short carbon fibers and hexa-
react with steel or sapphire surface through an acid-base reaction, boron nitride. Sci Adv Mater 2015;7. https://doi.org/10.1166/sam.2015.2142.
[6] Bahadur S. The development of transfer layers and their role in polymer tribology.
forming the thin and robust transfer films.
Wear 2000;245:92–9. https://doi.org/10.1016/S0043-1648(00)00469-5.
In-situ IR thermography shows that the nominal contact tempera- [7] Jintang G. Tribochemical effects in formation of polymer transfer film. Wear
ture is below PEEK Tg even though local temperature is raised by the 2000;245:100–6. https://doi.org/10.1016/S0043-1648(00)00470-1.
entrainment of debris. Results from Raman studies support that the [8] Laux KA, Schwartz CJ. Effects of contact pressure, molecular weight, and supplier
on the wear behavior and transfer film of polyetheretherketone (PEEK). Wear
contact temperature (100–120 °C) is below the Tg of PEEK. Hence 2013;297:919–25. https://doi.org/10.1016/j.wear.2012.11.013.
contact temperature alone may not be sufficient to generate the PEEK [9] Zhang R, Häger A, Friedrich K, Song Q, Dong Q. Study on tribological behaviour of
degradations observed. The presence of brittle cracks on the thin film plasma-treated PEEK and its composites. Wear 1995;181–183:613–23. https://doi.
org/10.1016/0043-1648(95)90177-9.
on the steel wear scar also suggests that the deformation temperature [10] Zhang G, Yu H, Zhang C, Liao H, Coddet C. Temperature dependence of the tri-
may be relatively low and the film may have exposed to UV irradiation. bological mechanisms of amorphous PEEK (polyetheretherketone) under dry sliding
The shear experienced by the rubbing surfaces leads to their tri- conditions. Acta Mater 2008;56:2182–90. https://doi.org/10.1016/J.ACTAMAT.
2008.01.018.
boelectrification. As a result, triboplasma is generated during rubbing. [11] Nakayama K, Mirza SM. Verification of the decomposition of perfluoropolyether
This triboplasma has sufficient energy, which together with the me- fluid due to tribomicroplasma. Tribol Trans 2006;49:17–25. https://doi.org/10.
chanical shear, can cause chain scission and generate radicals. This 1080/05698190500414524.
[12] Matta C, Eryilmaz OL, De Barros Bouchet MI, Erdemir A, Martin JM, Nakayama K.
promotes transfer film formation and leads to crosslinking and de- On the possible role of triboplasma in friction and wear of diamond-like carbon
gradation of PEEK. Our results show that mechanical shear, as well as films in hydrogen-containing environments. J Phys D Appl Phys
frictional heating and triboplasma all contribute to the formation and 2009;42:075307https://doi.org/10.1088/0022-3727/42/7/075307.
[13] VICTREX ® HIGH PERFORMANCE PAEK POLYMERS ©Victrex plc Revision
properties of the PEEK transferred materials on the rubbing counterface
VICTREX ® PEEK 450CA20 2014:1–2. https://www.victrex.com/∼/media/
(Fig. 13). Keeping the possibility of UV plasma generation in mind, the datasheets/victrex_tds_650g.pdf (accessed May 14, 2018).
design of future polymer and polymer composites should take the [14] Properties of Synthetic Sapphire n.d. https://www.swissjewel.com/materials/
possibility of surface charging and the potential effect it may have on sapphire/(accessed June 10, 2018).
[15] Reddyhoff T, Schmidt A, Spikes H. Thermal conductivity and flash temperature.
transfer film formation and degradation into considerations. Tribol Lett 2019;67:22. https://doi.org/10.1007/s11249-018-1133-8.
[16] Singh SV, Kusano Y, Morgen P, Michelsen PK. Surface charging, discharging and
chemical modification at a sliding contact. J Appl Phys 2012;111:83501. https://
doi.org/10.1063/1.3698311.
Acknowledgement [17] Vollmer M, Möllmann K-P. Wiley InterScience (Online service). Infrared thermal
imaging : fundamentals, research and applications. Wiley-VCH; 2010.
This work is partly funded by EPSRC (Grant number: EP/L023202/ [18] Blok H. Theoretical study of temperature rise at surfaces of actual contact under
oiliness lubricating conditions. Proc. Gen. Discuss. Lubr. Lubr. 1937;2:222–35.
1). The authors gratefully acknowledge Dr Philippa Cann for her help in [19] Hirschen C, Gülhan A. Infrared thermography and pitot pressure measurements of a
the FTIR studies and Dr Tom Reddyhoff for his advice on IR thermo- scramjet nozzle flowfield. J Propul Power 2009;25:1108–20. https://doi.org/10.
graphy measurements and observations of triboplasma. Test samples 2514/1.41787.
[20] Wittenberg AW. Total hemispherical emissivity of sapphire. J Am Ceram Soc
were kindly donated by Victrex and Hoerbiger. 1965;55:432–5. https://doi.org/10.1364/JOSA.55.000432.
[21] Rowe KG, Bennett AI, Krick BA, Gregory Sawyer W. In situ thermal measurements

198
D. Puhan and J.S.S. Wong Tribology International 135 (2019) 189–199

of sliding contacts. Tribol Int 2013;62:208–14. https://doi.org/10.1016/J. Corrosion behaviour of mild steel in 1-alkyl-3-methylimidazolium tricyanometha-
TRIBOINT.2013.02.028. nide ionic liquids for CO2 capture applications. RSC Adv 2014;4:5300. https://doi.
[22] Zum Gahr K-H. Microstructure and wear of materials. Elsevier; 1987. org/10.1039/c3ra45872e.
[23] Ebewele R. Polymer science and technology. New York: CRC Press LLC; 2000. [50] de Faria DLA, Venâncio Silva S, de Oliveira MT. Raman microspectroscopy of some
https://doi.org/10.1201/9781420057805. iron oxides and oxyhydroxides. J Raman Spectrosc 1997;28:873–8. https://doi.org/
[24] Nikafshar S, Zabihi O, Ahmadi M, Mirmohseni A, Taseidifar M, Naebe M. The effects 10.1002/(SICI)1097-4555(199711)28:11<873::AID-JRS177>3.0.CO;2-B.
of UV light on the chemical and mechanical properties of a transparent epoxy- [51] Kok JF, Renno NO. Electrostatics in wind-blown sand. Phys Rev Lett
diamine system in the presence of an organic UV absorber. Mater (Basel, 2008;100:014501https://doi.org/10.1103/PhysRevLett.100.014501.
Switzerland) 2017;10. https://doi.org/10.3390/ma10020180. [52] Blok H. The flash temperature concept. Wear 1963;6:483–94. https://doi.org/10.
[25] Jiang H, Browning R, Sue HJ. Understanding of scratch-induced damage mechan- 1016/0043-1648(63)90283-7.
isms in polymers. Polymer (Guildf) 2009;50:4056–65. https://doi.org/10.1016/j. [53] Day M, Sally D, Wiles DM. Thermal degradation of poly(aryl‐ether‐ether‐ketone):
polymer.2009.06.061. experimental evaluation of crosslinking reactions. J Appl Polym Sci
[26] Zhang S, Li JC. Slip process of stick–slip motion in the scratching of a polymer. 1990;40:1615–25. https://doi.org/10.1002/app.1990.070400917.
Mater Sci Eng, A 2003;344:182–9. https://doi.org/10.1016/S0921-5093(02) [54] Kusano Y. Atmospheric pressure plasma processing for polymer adhesion: a review.
00409-4. J Adhes 2014;90:755–77. https://doi.org/10.1080/00218464.2013.804407.
[27] Pooley CM, Tabor D. Friction and molecular structure: the behaviour of some [55] Gengenbach TR, Griesser HJ. Post-deposition ageing reactions differ markedly be-
thermoplastics. Proc Roy Soc A Math Phys Eng Sci 1972;329:251–74. https://doi. tween plasma polymers deposited from siloxane and silazane monomers. Polymer
org/10.1098/rspa.1972.0112. (Guildf) 1999;40:5079–94. https://doi.org/10.1016/S0032-3861(98)00727-7.
[28] Chalmers JM, Gasldn WF, Mackenzie MW. Crystallinity in poly(AryI-ether-ketone) [56] Lacks DJ, Mohan Sankaran R. Contact electrification of insulating materials. J Phys
plaques studied by multiple internal reflection spectroscopy. Polym Bull D Appl Phys 2011;44. https://doi.org/10.1088/0022-3727/44/45/453001.
1984;11:433–5. [57] Mylläri V, Ruoko T-P, Järvelä P. The effects of UV irradiation to poly-
[29] Socrates G. Infrared and Raman characteristic group frequencies. John Wiley & etheretherketone fibers – characterization by different techniques n.d.
Sons; 2004. https://doi.org/10.1002/jrs.1238. [58] Puhan D, Nevshupa R, Wong JSS, Reddyhoff T. Transient aspects of plasma lumi-
[30] El-Bahy ZM. Adsorption of CO and NO on ceria-and Pt-supported TiO 2 : in situ FTIR nescence induced by triboelectrification of polymers. Tribol Int 2019;130:366–77.
study. Mod Res Catal 2013;2:136–47. https://doi.org/10.4236/mrc.2013.24019. https://doi.org/10.1016/j.triboint.2018.09.026.
[31] Al Lafi AG. FTIR spectroscopic analysis of ion irradiated poly (ether ether ketone). [59] Hnatowicz V, Havránek V, Bočan J, Macková A, Vacík J. Modification of poly(ether
Polym Degrad Stabil 2014;105:122–33. https://doi.org/10.1016/j. ether ketone) by ion irradiation n.d. doi:10.1016/j.nimb.2007.11.031.
polymdegradstab.2014.04.005. [60] Ferain E, Legras R. Modification of PEEK model compounds and PEEK film by en-
[32] Lin-Vien D, Colthup N, Fateley W, Grasselli J. The handbook of infrared and Raman ergetic heavy ion and ultraviolet irradiations. Nucl Instrum Methods Phys Res Sect
characteristic frequencies of organic molecules. 1991. B Beam Interact Mater Atoms 1993;83:163–6. https://doi.org/10.1016/0168-583X
[33] Perng L, Tsai C, Ling Y. Mechanism and kinetic modelling of PEEK pyrolysis by TG/ (93)95921-Q.
MS. Polymer (Guildf) 1999;40:7321–9. https://doi.org/10.1016/S0032-3861(99) [61] Mylläri V, Ruoko TP, Järvelä P. The effects of UV irradiation to poly-
00006-3. etheretherketone fibres - characterization by different techniques. Polym Degrad
[34] Patel P, Hull TR, Mccabe RW, Flath D, Grasmeder J, Percy M. Mechanism of thermal Stabil 2014;109:278–84. https://doi.org/10.1016/j.polymdegradstab.2014.08.003.
decomposition of poly(ether ether ketone) (PEEK) from a review of decomposition [62] Deng XL, Nikiforov AY, Vanraes P, Leys C. Direct current plasma jet at atmospheric
studies. Polym Degrad Stabil 2010;95:709–18. https://doi.org/10.1016/j. pressure operating in nitrogen and air. J Appl Phys 2013;113:023305https://doi.
polymdegradstab.2010.01.024. org/10.1063/1.4774328.
[35] Nguyen HX, Ishida H. Molecular analysis of the melting behaviour of poly(aryl- [63] Lofthus A, Krupenie PH. The spectrum of molecular nitrogen. J Phys Chem Ref Data
ether-ether-ketone). Polymer (Guildf) 1986;27:1400–5. https://doi.org/10.1016/ 1977;6:113–307. https://doi.org/10.1063/1.555546.
0032-3861(86)90041-8. [64] Camacho JJ, Poyato JML, Díaz L, Santos M. Optical emission studies of nitrogen
[36] Nguyen HX, Ishida H. Molecular analysis of the crystallization behavior of poly plasma generated by IR CO 2 laser pulses. J Phys B At Mol Opt Phys J J Camacho Al
(aryl‐ether‐ether‐ketone). J Polym Sci, Part B: Polym Phys 1986;24:1079–91. J Phys B At Mol Opt Phys J Phys B At Mol Opt Phys 2007;40. https://doi.org/10.
https://doi.org/10.1002/polb.1986.090240510. 1088/0953-4075/40/24/003. 4573–4573.
[37] Munro HS, Clark DT, Recca A. Surface photo-oxidation of phenoxy resin and [65] Laux CO, Spence TG, Kruger CH, Zare RN. Optical diagnostics of atmospheric
polyetheretherketone. Polym Degrad Stabil 1987;19:353–63. https://doi.org/10. pressure air plasmas. Plasma Sources Sci Technol Plasma Sources Sci Technol
1016/0141-3910(87)90036-X. 2003;12:125–38.
[38] Nakamura H, Nakamura T, Noguchi T, Imagawa K. Photodegradation of PEEK [66] Rusanov A, Nevshupa R, Martin JM, Garrido MÁ, Roman E. Tribochemistry of
sheets under tensile stress. Polym Degrad Stabil 2006;91:740–6. https://doi.org/10. hydrogenated amorphous carbon through analysis of mechanically stimulated gas
1016/J.POLYMDEGRADSTAB.2005.06.003. emission. Diam Relat Mater 2015;55:32–40. https://doi.org/10.1016/j.diamond.
[39] Zhang MQ, Lu ZP, Friedrich K. Thermal analysis of the wear debris of poly- 2015.02.007.
etheretherketone. Tribol Int 1997;30:103–11. https://doi.org/10.1016/0301-679X [67] Schweitzer-Chaput B, Boess E, Klussmann M. Acid-catalyzed activation of perox-
(96)00028-X. yketals: tunable radical initiation at ambient temperature and below. Org Lett
[40] Švorčík V, Prošková K, Rybka V, Vacík J, Hnatowicz V, Kobayashi Y. Changes of 2016;18:4944–7. https://doi.org/10.1021/acs.orglett.6b02419.
PEEK surface chemistry by ion irradiation. Mater Lett 1998;36:128–31. https://doi. [68] Clark BDE. Peroxides and peroxide- forming compounds. vol. 9098. Div Chem Heal
org/10.1016/S0167-577X(98)00030-5. Saf Am Chem Soc Publ by Elsevier Sci; 2001. p. 1–22. https://doi.org/10.1016/
[41] Giancaterina S, Rossi A, Rivaton A, Gardette J. Photochemical evolution of poly S1074-9098(01)00247-7.
(ether ether ketone). Polym Degrad Stabil 2000;68:133–44. https://doi.org/10. [69] Smith MB, JM. Advanced organic chemistry: reactions, mechanisms, and structure.
1016/S0141-3910(99)00181-0. 2007.
[42] Shi H, Sinke J, Benedictus R. Surface modification of PEEK by UV irradiation for [70] Khudyakov IV, Kuz ’min VA. Oxidation-reduction reactions of free radicals. Uspekhi
direct co-curing with carbon fibre reinforced epoxy prepregs. Int J Adhesion Adhes Khimii Russ Chem Rev 1978;47144:39–82.
2017;73:51–7. https://doi.org/10.1016/J.IJADHADH.2016.07.017. [71] Collinson E, Dainton FS, Smith DR, Trudel GJ, Tazuké S. The oxidation and re-
[43] Rebelo de Figueiredo M, Bergmann C, Ganser C, Teichert C, Kukla C, Mitterer C. duction of free radicals by metal ions in aqueous solution. Discuss Faraday Soc
Adhesion tendency of polymers to hard coatings. Int Polym Process 1960;29:188–204. https://doi.org/10.1039/DF9602900188.
2013;28:415–20. https://doi.org/10.3139/217.2767. [72] La Mare D, Kochi JK, Rust FF, De La Mare E, Kochi K. Organic peroxides vol. 83.
[44] Cole KC, Casella IG. Fourier transform infra-red spectroscopic study of thermal Interscience Publishers, Inc; 1961.
degradation in poly(ether ether ketone)-carbon composites. Polymer (Guildf) [73] Behar D, Samuni A, Fessenden RW. Reactions of iron(II) and titanium(III) with
1993;34:740–5. https://doi.org/10.1016/0032-3861(93)90357-G. organic radicals. J Phys Chem 1973;77:2055–9. https://doi.org/10.1021/
[45] Briscoe BJ, Stuart BH, Thomas PS, Williams DR. A comparison of thermal- and j100636a005.
solvent-induced relaxation of poly(ether ether ketone) using Fourier transform [74] Radtsig VA, Butyagin PY. The free radicals forming during the mechanical de-
Raman spectroscopy. Spectrochim Acta Part A Mol Spectrosc 1991;47:1299–303. struction of polyethylene and polypropylene. Polym Sci USSR 1967;9:2883–7.
https://doi.org/10.1016/0584-8539(91)80219-9. https://doi.org/10.1016/0032-3950(67)90153-0.
[46] Agbenyega JK, Ellis G, Hendra PJ, Maddams WF, Passingham C, Willis HA, et al. [75] Balestrin LBS, Duque DD, Silva DS, Galembeck F, Beraldo da Silveira Balestrin L,
Applications of Fourier Transform Raman spectroscopy in the synthetic polymer Del Duque D, et al. Triboelectricity in Insulating Polymers: Evidence For A
field. Spectrochim Acta Part A Mol Spectrosc 1990;46:197–216. https://doi.org/10. Mechanochemical Mechanism vol. 170. 2014. p. 369–83. https://doi.org/10.1039/
1016/0584-8539(90)80090-L. c3fd00118k.
[47] Larkin P, Peter J. Infrared and Raman spectroscopy : principles and spectral inter- [76] Burgo TAL, Ducati TRD, Francisco KR, Clinckspoor KJ, Galembeck F, Galembeck SE.
pretation. Elsevier; 2011. Triboelectricity: macroscopic charge patterns formed by self-arraying ions on
[48] Briscoe BJ, Stuart BH, Sebastian S, Tweedale PJ. The failure of poly (ether ether polymer surfaces. Langmuir 2012;28:7407–16. https://doi.org/10.1021/la301228j.
ketone) in high speed contacts. Wear 1993. https://doi.org/10.1016/0043- [77] Onodera T, Park M, Souma K, Ozawa N, Kubo M. Transfer-film formation me-
1648(93)90524-P. chanism of polytetrafluoroethylene: a computational chemistry approach. J Phys
[49] Molchan IS, Thompson GE, Lindsay R, Skeldon P, Likodimos V, Romanos GE, et al. Chem C 2013;117:10464–72. https://doi.org/10.1021/jp400515j.

199