Poly (ethylene 2,6-naphthalene dicarboxylate) (PEN), a high-performance thermoplastic polyester, has attracted considerable attention for many years due to its superior thermal, mechanical, barrier and chemical resistance properties, which make PEN useful in a wide range applications, such as films, magnetic tapes and packaging (Ford, 1996). This aromatic polyester used recently in electrical engineering presents praiseworthy physical performances which make it an interesting material to investigate and may be considered as a serious competitor of polyethylene terephthalate (PET) for manufacture of capacitors. PEN is a polymer obtained by dicarboxylic polycondensation of the naphthalene-2,6 and ethylene glycol acid. The process used for the manufacture of films calls upon the dicarboxylic cross esterification of the dimethyl ester of the naphthalene-2, 6 acid with the ethylene glycol (Buchner et al., 1989).
This material is often subjected to higher electric fields which can induce mechanical stress within the insulator structure surface producing a local mechanical strain. In previous experimental investigations (Mamy et al., 2004; Chavez-Lara and Martinez-Vega, 2004) an attempt was made to assess the mechanical response of PET under gradually increasing electrical fields. It was suggested that these electrical stresses were responsible for localized mechanical strains.
The aim of this study is the quantification of the mechanical strain induced
by a high dc electric field in Poly (Ethylene-2, 6-Naphthalene dicarboxylate)
PEN films by using an optical measurement method (tracking). The tracking method
allowed measuring induced mechanical strains on the flat gold-metallized surfaces
of amorphous and semi-crystalline PEN thin films. Several PEN samples with different
crystallinity rates and thickness have been considered. The kinetics
of crystallization and melting of PEN have been investigated by using Differential
Scanning Calorimeter (DSC) to ensure the morphology status of the samples provided.
DSC was performed using a TA Instruments device of type DSC 2010. The DSC cell
was calibrated using pure indium. This technique allows us to determine the
glass transition (Tg) and the melting point temperature (Tm)
of the sample. The knowledge of crystallization (ΔHc) and melting
(ΔHm) enthalpies of the samples allows for the calculation of
the crystallinity rate (χ(%)), as mentioned in (Martinez-Vega et al.,
2001). Based on experimental considerations by using DSC, a thermal cycle of
crystallization was carried out. Different specimens were obtained in this way,
starting with as-received amorphous polymers. However, in order to improve the
comprehension of phenomena that might lead to the electrical breakdown and ageing
mechanism, a quantitative study by a non destructive optical technique without
involving any physical contact, based on the follow up of four computerized
markers on the sample surface was used to investigate the effect of the thermal
treatment on the induced mechanical strain of partially crystallized PEN films.
Both amorphous and semi-crystalline PEN samples were subjected to dc electric
fields for different durations up to their electrical breakdown, with or without
depolarization. In the present work we compare the level and the evolution of
electrical field induced strain in amorphous and partially crystallized PEN
samples; emphasizing the quantification of field induced strain occurring before
the electrical breakdown, as function of crystallinity and the electrical field.
Interpretation of the morphological changes and the levels of the field-induced
mechanical deformation corresponding to various measurement protocols are discussed
MATERIALS AND METHODS
The experimental measurement method is based on the follow-up of four markers
on the sample surface. This technique makes it possible to follow in real time
the mechanical behavior of PEN thin films under high electrical fields.
Material: Commercial PEN (Teonex TM) provided by Teijin DuPont Films (Luxembourg), in A4 sheet format were employed for experiments. The as-received samples were PEN films, partially crystallized Teonex Q51 12 and 25 μm thick, as well as amorphous 14, 25 and 70 μm thick.
Experimental device and samples preparation: Figure 1
shows the schematic diagram of the experimental apparatus developed to conduct
the electrical field induced strain measurement by using an optical technique
based on the tracking of successive position displacement of four computerized
markers by DEFTAC software (associated with an optical system formed by an optical
microscope connected to a Charge Coupled Device (CCD) camera (resolution = 768
pixels. 576 pixels with 256 gray levels) provided by MATROX -METEOR II, image
processing card permitted the observation of the sample surface in reflected
light (Dupré et al., 2005). To guarantee a better electrode/polymer
contact, the test samples were metalized by gold coating using a S150B plasma
sputter coater. To determine the morphology of the samples under study and the
impact of gold metallization on the initial PEN morphology, we analyzed two
samples with Differential Scanning Calorimetry (DSC): one was as-received (for
the morphological study) and the other was gold-metalized (for the study of
the effect of the metallization on the thermal properties of the material).
||The schematic diagram of the experimental setup
The analyses were carried out under liquid nitrogen with specimens of about
11 mg; each sample was subjected to a rise in temperature from ambient temperature
to 290°C at 10°C min¯1.The DSC thermograms show practically
no difference between the as-received and metalized samples indicating no appreciable
effect of gold metallization on the morphological properties of the PEN.
Electrodes of 20 mm diameter and 30 nm thickness were thus obtained on both sides. The prepared sample was placed between two external brass electrodes. The upper one, the negative, was constructed as a hollow cylinder of 12 mm diameter. This permitted the use of a flexible light source to illuminate the upper face of the sample. The lower electrode, the positive, was connected to a high dc voltage source (HCN 35-20000; 20 kV and 1.5 mA limited current) with controllable output. The samples were placed between electrodes in the measuring cell; the electrodes were short-circuited a few hours before the testing in order to eliminate the initial charges existing on the sample faces before applying the electric field.
Figure 2 shows the obtained image of the gold-metalized surface
of a sample using. This image reveals small, contrasting spots, which represent
the microscopic light contrast of traditional metallization. To manipulate the
test sample as little as possible, we identified four of these spots with computerized
marking. The four markers thus obtained allowed us to quantify the electric
field induced strain in real time.
Principle of strain measurement: The principle of strain measurement
consists of four markers (Fig. 3), forming a cross, positioned
on the sides of a parallelogram.
Knowing the coordinates of these spots, at each state of loading, we can use
the calculation of the lengths (a1 and a2) and orientations
(α1 and α2) to determine, into a large deformation
formulation, the Green-Lagrange strain tensor ()
where, the gradient tensor function of the following components:
The experimental technique makes it possible to measure the induced mechanical
deformation components associated with the plane surface of a film. We label
these components εx, εy and γxy,
which are associated with the directions parallel to the x-, y- and xy-shearing
||Disposition of four markers on gold-metalized amorphous PEN sample
||Diagram of the follow-up of four markers
The ε1 and ε2 components are associated with
the principal a direction at which γxy is null. We have assumed
the homogeneity of the deformation on the measurement base and as a result,
the reported values are averages.
The principle of mark tracking is to calculate the geometric center of the
spot from a rectangular zone (the research zone), defined at the beginning of
the computation, around each spot by its upper left coordinates (xz,
yz) and size (Nx, Ny). For large deformations
and movements, this research zone is automatically moved at each step of time,
with consideration given to the measured displacement, to keep the spot inside.
Then, the coordinates (xg and yg) of each marker are derived
with the following equations:
I(x, y) is the light intensity of the pixel with coordinates x and y.
where, Is is the lower limit of the light intensity (beyond this limit, the pixel is not discriminated) and I(x,y) is the light intensity of the pixel with coordinates x and y.
Procedure: Measurements were carried out at the room temperature at atmospheric pressure and for very short durations (less than 30 min), in order to minimize the influence of the environmental conditions. Several protocols of dc electric stress application were adopted (Zegnini et al., 2006); the differences consisted in the duration of dc voltage application and/or the existence of a depolarization period. To assess the level of deformation resulting from high dc voltage application, PEN samples were stressed for periods of 5, 10, 15 and 30 min followed by a similar depolarization period at every voltage level with different CCD camera sampling rates (one photo per three seconds and six seconds, respectively). After the test, we removed the electric stress to study the return of the material to its original state. To obtain measurement reproducibility, we carried out measurements of the noise levels before the application of the electric field for a certain period.
RESULTS AND DISCUSSION
Differential scanning calorimetry: The thermal analysis DSC was successfully
performed in order to obtain significant degrees of crystallinity. The as-received
amorphous PEN samples were first maintained at 170°C for different annealing
times 5, 10, 15, 30, 45, 60, 90, 120, 180 min; the protocol of isothermal crystallization
of as-received amorphous PEN is given in Fig. 4.
DSC measurements have been carried out from 30 to 300°C, in order to characterize
the glass transition, the melting point and the crystallization degree of the
material. The results obtained with as-received material as well as with the
amorphous one. The glass transition, clearly observable in the scan corresponding
to the amorphous sample, lies at approximately Tg = 123°C: obviously
the change of baseline becomes less pronounced when the isothermal annealing
time increases. This amorphous sample crystallizes between 160 and 220°C,
preceding the fusion of the material at 268°C. The scan corresponding peak,
which indicates a high degree of crystallinity in the material. The main endothermic
peak, located at 268°C, corresponds to the fusion of the crystalline phase
and is preceded by a small endothermic peak probably associated with a solid
state transition in the material (Fig. 5).
||Protocol of isothermal crystallization
||DSC thermograms of PEN samples annealed at 170°C
According to the extrapolated heat of fusion for a pure PEN crystal (103.3
J g¯1) and the total heat flow involved in the endothermic peaks,
a crystallinity degree of about 4.17% is estimated for the as-received amorphous
Table 1 shows an augmentation in the amount of crystalline
phase as isothermal annealing duration increases. Three stages of evolution
are observed, the first one corresponds to annealing times lower than 15 min,
where the material remains almost, amorphous. Subsequently, a fast crystallization
rate is observed, between 15 and 45 min. As result the peak of the cold crystallization
disappeared then a small pre-melting peak is observed instead. Finally, a constant
percentage of crystalinity about 42% is obtained for the longest periods. In
addition it was observed at higher temperatures, the melting peak is around
268°C and it does not be affected by this thermal treatment. The obtained
results are explored to analyze the dependence of the rate of crystallinity
on the mechanical response of PEN thin films.
||Annealed as-received amorphous PEN 70 μm samples
Performance of the strain measurement: The accuracy of the strain measurement
is a function of the marker position and the distance between the spots. The
measurements of the noise level were carried out for an as-received PEN sample
70 μm thick on a mechanical deformation without an electric field. The
measurements of the noise level were carried out for an as-received PEN sample
on a mechanical deformation without an electric field. Figures
6a and b show the measured components of the noise levels
corresponding to the distant markers.
Figure 6a and b show that in the case the
noise levels were very small and corresponded to the predicted error (2x10¯4).
For This result we could neglect the effect of the weight of the higher electrode.
Moreover the measured mechanical strain deformations in this study corresponded
only to the strain induced by the electric field application.
Field induced strain measurements: During the present study, we highlighted
the optical observation of the surface of PEN thin films for the analysis field
induced strain. To perform this investigation PEN films were subjected for periods
of 300 sec at gradually increasing 0.5 kV constant step of high applied voltage.
It is important to mention that the maximum applied voltage that we have imposed
was 7 kV; the deformation of the film was recorded with constant CDD camera
using a sampling rate (one image per three seconds). An attempt was made to
compare the level and the evolution of field induced strain in the as-received
amorphous and partially crystallized PEN samples (Zegnini et al., 2007).
Two different samples with different degrees of crystalinity were chosen to
study the breakdown phenomena. Figure 7 shows the sample with
a crystallinity of (χ = 17, 41%), which broke down at electric field of
6.21 kV. The figure shows the moment of spark due to its electrical breakdown.
The geometry of the arc due to breakdown is very well sparking defined in Fig.
Figure 8 shows the induced strain produced during the test.
Figure 8a shows the principal strain components of the film
deformation as function of time. It is interesting to note that ε1
is positive and ε2 is negative (even through it is close to
zero). This anisotropy is probably to the biaxial orientation of the film micro-morphology
produced during its manufacturing.
||Strain components of the noise levels (electric field = 0)
||Final captured image after breakdown the PEN sample 25 μm thick with
a crystallinity of (χ = 17, 41%)
Figure 8b shows the contributions of the average fitted
values ε1 and ε2 to evaluate of the electric
field induced strain ε as a function of the electrical field, until the
electrical breakdown (Ford, 1963).
||Induced strain versus electrical field of PEN 25 μm thick sample
partially crystallized (χ = 17.41%)
The evolution of the electric field induced shows three zones. Indeed, initially
there is a region of very low deformation until an electrical field threshold
is attained after which an increasing induced strain level is observed. Finally,
there is a diminution of the strain that could be produced by a local densification
of the material. This behaviour suggests the presence of a thermally dominated
mechanism attributed to the propagation of local breakdowns caused by large
local fields in micro-voids. It is rather a mechanism of electrical aging having
a mechanical origin. Thus, on the basis of a previous image-acquisition program
used to explore the influence of crystallinity on the mechanical response of
the same sample, two distinct domains resulting from dielectric breakdown were
investigated (Fig. 9). The first is with dielectric breakdown
zone more vulnerable to the field-induced mechanical deformation as compared
to the second which is situated outside this region. However, the level of the
induced mechanical deformation seems to have similar evolution, in particular
when the values of the electric stress are relatively small (Fig.
||Definition of two zones the PEN 25 μm thick sample with a crystallinity
of (χ = 5.14 %); the first one (zone 1) inside the damage area and
the second one (zone 2) outside this area
||Comparison of the induced strain at each zone
This observation links in a straightforward way the local strain with the
electrical breakdown phenomena. The possible explanation of this behavior is
probably the diminution of the trapped charges mobility of responsible for the
reorganization of macromolecular chains in zone I and the conformational changes
probably vary the molecular inter-space. It is possible that the deformation
at the breakdown area could be set up by pre-existent defects from the chemical
or physical manufacturing processes, where the strain concentrated until the
final electrical breakdown.
Figure 11a and b shows the influence
of the crystallinity on the strain behaviour and the electrical breakdown field.
||Influence of the crystallinity on the strain behaviour and electrical
breakdown (a) behaviour at zone I and (b) behaviour at zone II
It is observed that the lower crystallinity sample (χ = 5, 14%) presents
a larger strain that the other one (χ = 17, 41%).
The previous observations are very interesting because it show that the crystallinity produces an electrical weakness of the tested sample. It is important to indicate that the conductivity of the amorphous PEN is almost larger than the semi-crystalline PEN (Chavez-Lara et al., 2004). Therefore, the higher crystalline sample is capable of accumulating a larger electrostatic energy, which seems to be the main factor that the electrical breakdown.
In addition the influence of crystallinity on the induced strain in treated
PEN 70 μm thick samples, the results given in Fig. 12
reveal higher deformation levels in case of amorphous samples. Furthermore;
for annealing time 60 and 120 min, the crystalinity degree is constant respectively
43.31 and 44.34%. But the local strain observed for ta = 120 min
is lower compared to ta = 60 min, at the high electric field range.
||Influence of crystalinity percentage on the field induced strain
This can be explained by a more ordered microstructure that has higher mechanical
strength. This morphology increases the stiffness of the polymer, hence its
A non destructive optical technique has been applied to the measurement of
field induced strain in PEN thin films in order to give a clear concept of ageing
phenomenon and breakdown mechanism in a polymeric material. The resulting induced
strain levels and patterns were dependent on the level, measurement protocols
(ramp test, with depolarization test) and duration of electric stress application,
thickness and morphologies of the samples. A thermal protocol of isothermal
crystallization has been conducted to obtained PEN film crystallized samples
with different crystallinity degrees, starting with as-received amorphous. The
tracking method was employed to evaluate the local induced strain on the basis
of a previous image-acquisition program at two different zones of the tested
sample; breakdown zone (zone 1) inside the damage area and zone II outside this
area. The results revealed higher deformations levels in zone I compared to
the distant unaffected zones (ie zone 2). This difference in induced strain
was observed even at very weak electrical fields. Moreover plots of field induced
strain to analyze the influence of crystallization on the level of mechanical
deformation and threshold field. It was observed that the field induced strain
was dependent on the crystallinity of the treated samples dependent on the crystallinity
and the electrical breakdown of tested samples. The filed induced strain diminished
when the crystallinity increased, as well as the maximum voltage supported by
the film before electrical breakdown. The obtained results implies further work
to validate a clear concept of ageing phenomenon and breakdown mechanism in
a polymeric material under operating conditions, in which several stresses are
The authors are indebted to DUPONT TEIJIN FILMS (Luxembourg) for providing PEN samples. Also they wish to thank J.C. Dupré, Directeur de recherches, Laboratoire de mécanique des solides (UMR-CNRS 6610) Bd Marie et Pierre Curie, Téléport 2, 86962 Futuroscope Chasseneuil, cedex, France for his help, support and collaboration in this research work.