Development of Poly ( butylene adipate-co-terephthalate ) Filled with Montmorillonite-Polypyrrole for Pressure Sensor Applications

A novel pressure sensing material composed of poly (butylene adipate-co-terephthalate) (PBAT) and montmorillonite-polypyrrole (Mt-PPy) was prepared using melt mixing and injection molding. The structure and properties of the PBAT/Mt-PPy composites were evaluated and compared with those of the PBAT/PPy blends. The PBAT/Mt-PPy displays a very sharp insulator-conductor transition and its percolation threshold was reached at 6.5 wt% of Mt-PPy, which was lower than that for PBAT/ PPy (11.0 wt% of PPy). The transmission electron microscopy analysis shows that Mt-PPy displays a high aspect ratio and was better distributed and dispersed into PBAT compared to PPy. PBAT/Mt-PPy exhibits a decrease in the electrical resistivity with an applied compressive stress due to the formation of new conducting pathways. The electromechanical response was dependent on Mt-PPy and the maximum sensitivity was observed for the composite containing 10 wt% of Mt-PPy. In this system, the electrical resistivity drops from 9 x 106 to 2 x 106 Ω cm when a compressive stress of 0.25 MPa was applied. The electrical resistivity changes with applied compressive stress, the reproducibility and the reversibility makes PBAT/Mt-PPy a suitable material for the development of pressure sensors.


Introduction
In recent years, the development of electrically conducting polymer composites (ECPC) for pressure sensor applications has increased significantly.This increase is due to their flexibility, low cost, easy to manufacture and the electrical conductivity can be changed by applying a compressive force [1][2][3][4][5][6][7][8][9] .These materials are composed of an insulating polymer matrix and a dispersed phase of intrinsically conducting polymer (ICP) [10][11][12][13][14] , such as polypyrrole (PPy) or polyaniline (PAni), or carbonaceous fillers, such as carbon black [15][16][17] and graphite 18 .Among available conducting fillers, several works in the literature have reported the use of conducting polymers, especially PPy, for sensing applications 1,5,8,9,[19][20][21][22][23][24][25][26][27][28] .Merlini et al. described the development of poly(vinylidene fluoride) (PVDF) / polypyrrole (PPy) blends, and the maximum sensitivity was obtained for a blend with 9 wt% PPy.The electrical resistivity of this blend dropped by two orders of magnitude, e.g. from 10 8 to 10 6 Ω cm 24 .Souza, Michel and Soares reported that the electrical resistivity of poly(styrene-butadiene-styrene)/polyaniline blends drops one order of magnitude when a compressive stress from 0 to 4 MPa is applied 11 .The changes of the electrical conductivity in ECPC are explained by the percolation theory 10 .Under compressive stress, the conductive particles come into a closer contact forming a new conducting network.On the other hand, by releasing the compression stress, the conducting network is disrupted, and the electrical conductivity returns to the initial value to that without pressure 24 .The electrical sensitivity and reproducibility of the response of these composites depend on various factors including the structure and properties of the dispersed phase (such as, intrinsically electrical conductivity, aspect ratio and surface area), the properties of the matrix, ICP content, interaction of both polymer phases and dispersion of the ICP into the insulating polymer matrix.The main challenge in the development conductive polymer composites based on ICPs is to reach an adequate pressure sensitivity at low conducting filler concentration in order to minimize processing problems and a reduction of the mechanical properties.In general, when PPy is used as a filler to develop pressure-sensitive composites, large amounts are required to reach significant increases in the electrical conductivity, and there are difficulties to disperse the filler.An appropriate strategy that has been proposed is the development of conducting composites composed of a nanostructured filler, such as montmorillonite/polypyrrole (Mt-PPy) or montmorillonite/ polyaniline (Mt-PAni), dispersed in an insulating polymer matrix.This nanostructured conductive filler can be obtained through chemical in situ polymerization of pyrrole (Py) in the presence of Mt suspension.The montmorillonite can be exfoliated by the Py and then the polymerization that occurs in the Mt layers results in a conductive filler with a layered lamellar structure, with a high aspect ratio and surface area, compared to the neat PPy, which allows a strong interaction with the insulating polymer.According to Ramoa et al. 29 and Boukerma et al. 30 , the exfoliation of Mt-PPy promotes conductive pathways into the insulating polymer matrix with a lower Mt-PPy content than that observed for conducting polymer blends containing neat PPy as the conductive filler.
In addition, interesting results concerning the preparation of conducting polymer composites with Mt-PPy and an insulating polymer matrix have also been reported by Moučka et al. 31 and Peighambardoust and Pourabbas 32 .
The manufacturing method and test conditions also influence the electrical properties of the conducting polymer composites.Often these materials are developed by incorporating ICP into an insulating polymer matrix, such as thermoplastic polymers or rubbers through in situ pyrrole or aniline oxidative polymerization 5,[33][34] , solution casting 12,23 or the melt mixing method 3 .Alternatively, the electrospinning technique can be also used in order to produce highly pressure-sensitivity fibrous membranes 9,36 .Among these, the melt mixing method offers the advantages of large-scale production and reduced costs.However, studies in the literature have demonstrated that ECPC with the same chemical composition prepared through solution casting showed higher pressure sensitivity than that produced by melt mixing 12,[36][37][38] .This behavior can be attributed to the degradation of the conducting polymer due to the high melting process temperature and the difficulty to disperse the conductive phase in the insulating polymer matrix.
In this context, poly(butylene adipate-co-terephthalate) (PBAT), which is a flexible aliphatic-aromatic copolyester, is an interesting insulating polymer matrix to produce flexible conductive Mt-PPy nanocomposites with high electrical conductivity with a low percolation threshold.PBAT is among the most versatile thermoplastics since it has a low processing temperature (160 ºC), and therefore does not reach the degradation temperature of PPy during the melt mixing method, low modulus, stiffness, high flexibility and toughness.Moreover, the ester groups of PBAT can interact with the -NH groups of PPy leading to better distribution and dispersion of the conductive filler into polymer matrix.Furthermore, due to the efforts of academic and industrial researchers to develop ecofriendly material, PBAT is an excellent candidate to develop pressure-sensitive composites due to its biodegradability and natural origin 39 .
Based on the above, the aim of this work is to investigate PBAT/Mt-PPy composites produced using a twin screw extruder followed by injection molding.Particular attention is focused on producing sensitive composites with low filler content.For comparative purposes, neat polypyrrole (PPy) was added separately in the PBAT under the same processing conditions.To the best of our knowledge, there are no studies dealing with the production of composites based on biodegradable PBAT with Mt-PPy using the melt mixing method for pressure sensor applications.

Synthesis of conducting nanocomposites (Mt-PPy)
The preparation procedure of nanostructured conductive filler Mt-PPy with the surfactant DBSA was based on the method described by Ramoa et al. 29 2.5 g of Mt was dispersed into 250 mL of water, containing the DBSA, and stirred for 2h at room temperature.The dispersion was sonicated with 35% power (263 W) for 20 min using a Sonics VCX 750 ultrasonic processor (Sonics & Materials, Inc., USA).The molar ratio of the surfactant/pyrrole (Py) used in the polymerization was 1/5.Then FeCl 3 •6H 2 O (0.2542 mol) dissolved in 125 mL of distilled water was added to the aqueous Mt dispersion under stirring at room temperature.After which, 50 mL of a 0.26 mol•L -1 aqueous dispersion of Py (0.1105 mol) was added drop-wise over a 15 min period.The polymerization was carried out for 1 h under stirring at room temperature.After 24 h, the conductive filler (Mt-PPy) was filtered washed with distilled water and dried at 60 ºC.The same procedure was used to prepare neat polypyrrole (PPy).

Preparation of PBAT/Mt-PPy composites
Composites based on PBAT/Mt-PPy and PBAT/PPy were prepared using a 15g-capacity DSM microextruder (Midi 2000 Heerlen, Holland) with co-rotating screws.The mixture was sheared under nitrogen atmosphere for 3 min with 100 rpm speed at 160 ºC and then injected in a mold at 40 ºC to obtain discs with 40 mm diameter, according to the method described by Livi et al. 40 .

Characterization
The electrical conductivity of the conductive fillers, PPy and Mt-PPy, and low-resistivity composites of PBAT/Mt-PPy and PBAT/PPy blends were measured using the four probe standard method using a Keithley 6220 (USA) current source to apply the current and a Keithley Model 6517A (USA) electrometer to measure the potential difference.For pure PBAT and high-resistivity composites the measurements were made using a Keithley 6517A (USA) electrometer connected to a Keithley 8009 (USA) test fixture.Five measurements were taken for each sample at room temperature.
Transmission electron microscopy (TEM) was carried out at the Technical Center of Microstructures (University of Lyon) on a Philips CM 120 (Germany) microscope operating at 120 kV.The samples were cut using a Leica Ultracut UCT ultramicrotome (Leica, Germany) equipped with a diamond knife to obtain 60 nm thick ultrathin sections and deposited on a 200 mesh copper grid for subsequent observations.Attenuated Total Reflectance-Fourier Transform Infrared spectroscopy (ATR-FTIR) was performed with a Bruker (USA) spectrometer model Tensor 27, in the range of 2000 to 600 cm -1 by accumulation of 32 scans at 4 cm -1 resolution.
The X-ray diffraction (XRD) patterns of all samples were obtained on a Phillips X'PERT (Philips, Germany) X-ray diffractometer with CuK (λ= 0,154 nm) radiation source operating at a voltage of 40 kV and a 30 mA current.The samples were evaluated on a 2θº range from 2 to 50º at steps of 0.05º and time step of 1 s.The Mt, PPy and Mt-PPy samples were analyzed in powder form while neat PBAT, PBAT/Mt-PPy composites and PBAT/PPy blends were analyzed in the form of discs.
Dynamic mechanical properties were measured by a dynamic mechanical analyzer (DMA 242 E Artemis, Germany) under tensile mode on rectangular specimens with a width of 5 mm and a length of 12 mm.The analyses were performed at a frequency of 1 Hz, and amplitude of 50 µm using a 3 ºC min -1 temperature ramp from -80 to 70 ºC.
Thermogravimetric analysis (TG) was carried out using a STA 449 F1 Jupiter ® (Netzch, Germany) thermogravimetric analyzer.The analyses were performed for the conductive additives Mt-PPy and PPy, PBAT/Mt-PPy and PBAT/PPy at 10 ºC min -1 from 30 ºC to 650 ºC under air atmosphere with a flow rate of 50 mL min -1 .
The electromechanical tests were performed according to the methodology reported by Merlini et al. 8 .To evaluate the electrical-mechanical response of PBAT/Mt-PPy and PBAT/PPy composites the setup used consisted in a universal testing machine MTS Acumen (USA) with a 0.5 kN load cell to apply a controlled stress and an electrometer (Keithley 6517A)(USA) to acquire the resistivity data, using software developed exclusively for this purpose (Figure 1).The sample with a diameter of 22.5 mm was placed between two electrodes which were confined in a cylinder made of poly(tetrafluoethylene) (PTFE) in order to electrically isolate the samples.The device was then placed between the testing plates of the universal testing machine.The electrodes were connected to the electrometer for measuring the volume resistivity of the composites during loading and unloading.The samples were loaded up to 0.25 MPa at a loading rate of 0.00018 MPa/s, the compressive stress was then released at the same rate (Figure 1).For each sample, loading-unloading sequences of 5 and 25 cycles were performed on different specimens.The electrical resistivity (ρ) values (in Ω cm) were calculated using Eq.1, where R is the measured resistance (Ω), d is the sample diameter (cm) and w is the specimen thickness (cm).The curves of electrical resistivity and compressive stress were plotted as a function of time (Figure 1).
(1) The percolation threshold (f p ) and critical exponent (t) of both polymer systems were calculated according to the model proposed by Kirkpatrick 41 .The f p and t parameters of PBAT/Mt-PPy were ~6.5 wt% and 2.7, respectively, while  those parameters for PBAT/PPy were ~11 wt% and 2.5.A critical exponent in the range from 2 to 4 is in agreement with the classical theory for tridimensional systems.

Results and Discussion
The electrical conductivity of the filler Mt-PPy and PPy are similar 7.2 ± 4.4 S cm -1 and 6.2 ± 5.0 S.cm -1 , respectively.Thus, the intrinsic conductivity of the fillers is not a factor that significantly influences the final electrical conductivity of the composites.The lower percolation threshold and higher electrical conductivity of PBAT/Mt-PPy composites may be associated to the layered structure of the Mt-PPy.This layered morphology results in a higher aspect ratio and surface area (3.52 m 2 g -1 ) 42 than found for the spherical particles of PPy (2.06 m 2 g -1 ) 42 , and consequently increases the interaction of Mt-PPy with the PBAT matrix and thus improves its ability to form a conductive network in the matrix.Similar results have been reported by Ramoa et al. 35 for thermoplastic polyurethane (TPU) filled with Mt-PPy and PPy.Also, Contri et al. 43 and Vargas et al. 41 have found similar results for epoxy and thermoset polyurethane, respectively, with the same nanoparticles.
Figure 3 illustrates the TEM micrographs from which it is possible to evaluate the influence of the conductive filler on the microstructure of the composites.The PBAT/Mt-PPy nanocomposite micrographs show that the layered conductive fillers are better dispersed and distributed in the PBAT matrix compared to the PBAT/PPy blends.Moreover, the layered Mt-PPy fillers display an intercalated/partially exfoliated structure that facilitates the contact between the particles, allowing the formation of a conductive network throughout the matrix.On the other hand, the microstructure of the PBAT/PPy blends revealed a phase separated morphology with the presence of PPy agglomerates into PBAT matrix.The difference in the microstructure and aspect ratio of the fillers indicates that Mt-PPy was able to induce a conductive network formation with a low content of conductive filler in the PBAT matrix, and consequently, increases the electrical conductivity of the nanocomposites more significantly.
Figure 4 presents the XRD diffractogram of Mt, conductive fillers and nanocomposites.The (001) reflection and d-value for neat Mt were 6.11° (2θ) corresponding to an interlayer distance of 1.44 nm while for Mt-PPy were 2.12 o (2θ) which corresponds to a basal spacing of 4.17 nm, respectively.The shift of the peak to lower angles in the XRD and higher d-value for Mt-PPy can be attributed to the PPy chain intercalation between the Mt layers 30,[44][45] .Moreover, the low-intensity reflection at 6.11° for the Mt-PPy filler suggests that most of the Mt layers were exfoliated [44][45][46] during in situ polymerization.On the other hand, PPy shows (001) reflection at 2.78 o (2θ) and d-value of 3.20 nm.The X-ray diffraction pattern of PBAT is characterized by 4 peaks at 16.11°, 17.45°, 20.21°, 23.19° which are related to the basal reflection (011), (010), ( 111) and (100), respectively, indicating a crystalline structure of PBAT 47 .All composites showed a lower peak intensity than those observed for neat PBAT.Interestingly, the peak at 6.11º belonging to Mt, disappears in the PBAT/Mt-PPy composite, which suggests that the Mt layers were partially exfoliated in the PBAT matrix [48][49][50] .
The FTIR spectra of the Mt-PPy, PPy, PBAT and PBAT/ Mt-PPy composites and PBAT/PPy blends from 2000 to 600 cm -1 are shown in Fig. 5 (a) and (b).Ramoa et al. gave a detailed study of the FTIR spectra of the fillers 29 .The PBAT spectrum exhibits absorption bands at 1710 cm -1 , which are assigned to the carbonyl groups in the ester linkage, while bands at 1263 cm -1 correspond to the C-O in the ester linkage and 726 cm -1 represents four or more adjacent methylene (-CH 2 -) groups [50][51][52] .The spectra of PBAT/Mt-PPy and PBAT/ PPy exhibited overlapped absorption bands of PPy and PBAT.
An in-depth analysis of the infrared spectra in the range from 1780 to 1680 cm -1 for PBAT and its composites was carried out.The absorption bands centered at 1730 and 1710 cm -1 are attributed to the free and site-specific interactions of the carbonyl group, respectively.The interaction between PBAT and PPy can be determined from the ratio between free carbonyl absorption bands indicates that a site-specific interaction between N-H and C=O functional groups is operative in both polymer systems 35 .
Figure 6 (a, b) shows the variation in the storage modulus (E') and loss tangent (tan δ) as a function of temperature for neat PBAT, PBAT/Mt-PPy and PBAT/PPy.Over the entire temperature range, E' values of both polymer systems, PBAT/ Mt-PPy and PBAT/PPy, are higher compared to neat PBAT, which is due to the reinforcement effect of the conductive fillers.Moreover, the chemical interaction between the conductive fillers and PBAT also contribute to an increase of the mechanical properties.In addition, the storage modulus of the composites increases when the filler content increases.
A sudden fall in the storage modulus is observed for neat PBAT at -20 o C, which corresponds to the glass-rubber transition of the PBAT.This transition can also be related to the peak of tan δ, as shown in the Fig. 8(c) and (d).The peak intensity, which corresponds to the relaxation temperature (T α ), related to the glass transition temperature of PBAT reduces as the amount of conductive filler increases.These variations in the magnitude of the loss tangent are related to the incorporation of Mt-PPy or PPy, which reduces the  the peak areas of the absorption bands at 1710 (H-C=O) and 1730 (free C=O) (I 1710 /I 1730 ) 5 .Table 1 shows the area under the peak related to the C=O groups (hydrogen bonded C=O (A 1710 ) and free C=O (A 1730 )) and the ratio between them.Moreover, the absorption area ratio between the bonded and free carbonyl groups (A 1710/1730 ) increases significantly with an increase of Mt-PPy and PPy content.A reduction of the flexibility of the PBAT 53 .Moreover, the decrease in the peak intensity of the loss tangent is higher for PBAT/Mt-PPy because the incorporation of Mt increases the stiffness of the composites.
The TG curves of Mt, Mt-PPy, PPy, PBAT, PBAT/Mt-PPy and PBAT/PPy are shown in Figure 7.The MMt display  thermal stability until 600 ºC.PPy and Mt-PPy present a small weight loss below 100 ºC due to the elimination of water and a weight loss starting from 317.9 ºC and 321.7 ºC respectively, which corresponds to the degradation of the polymer chain.At 650 °C, the thermal decomposition of PPy is completed.On the other hand, the Mt-PPy sample displays a residual mass that can be related to the Mt that is thermally stable at this temperature.PBAT shows a main weight loss starting at 383.7 ºC which is attributed to the decomposition of the polymer chain.The derivate thermogram (DTG) (Table 2) clearly shows that PBAT/Mt-PPy and PBAT/PPy have two weight loss steps.The first step is related to the polymer chain decomposition of both components and the extra derivate peak is observed around 500 ºC, which corresponds to the oxidation of char residue under an atmosphere of air 54 .All the composites start to decompose at lower temperatures than neat PBAT as shown in Table 2.This behavior can be related to the presence of the PPy, which presents lower thermal stability 55 .
Figures 8 a-h show the variation of the electrical resistivity of PBAT/Mt-PPy and PBAT/PPy composites that have been submitted to 5 consecutive compressive loading-unloading stress cycles, until a compressive stress of 0.25 MPa.Compressive stress has an influence on the electrical resistivity of PBAT/Mt-PPy composites due to the presence of conductive Mt-PPy.This behavior can be explained by the fact that when the system is submitted to a compressive stress, the contact between the Mt-PPy particles improves and forms new conducting pathways, resulting in a reduction in the electrical resistivity of the composite 8 .When a composite with 5 wt% Mt-PPy is submitted to a compressive stress, the contact between the conductive fillers is not enough to form new conductive pathways, therefore there is no variation in the electrical resistivity of the sample.On the other hand, a compressive stress has a remarkable effect on the electrical resistivity of a composite with 10 wt% Mt-PPy.This composite displays a better electromechanical response under compressive stress.When the system is subjected to a compressive stress (loading cycle), the resistivity value decreases by one order of magnitude as the pressure increases.Moreover, when the applied stress is removed, the resistivity returns to near initial value.This behavior indicates that new conductive pathways are created on applying a compressive stress and a reversible reorganization of Mt-PPy particles occurs in the PBAT matrix.This system presents a reproducible response after 25 consecutive compressive loading-unloading stress cycles (Figure 9).For the composite with 15 wt% Mt-PPy the change in resistivity is less pronounced, and when the compressive stress is applied the electrical resistivity decreases from 8 x 10 6 to 3 x 10 6 Ω cm.However, when the conductive filler content is increased to 20 wt%, the response is markedly different from that presented for the other samples.It is possible to note that this composite does not display a significant change of the electrical resistivity when the compressive stress is applied.This behavior can be attributed to the large amount of conductive additive of Mt-PPy in the PBAT matrix, which already forms a conducting network, even without a compression stress, as observed in TEM micrographs.It suggests that the Mt-PPy particles are too close to each other to be influenced by the compressive stress applied 8 .The eletromechanical tests was performed on the all the PBAT/PPy blend samples; however the response was completely different in the PBAT/PPy blends with 5,  10 and 15 wt% of filler compared with the Mt-PPy samples, there was a very small variation in the electrical resistivity under compressive stress and the response was not linear and was irreversible.However, the electrical resistivity of the PBAT/PPy blend with 20 wt% PPy was influenced slightly by the loading and displayed a nonlinear response.The electromechanical response of the PBAT/Mt-PPy composites was superior to that of the composite with neat PPy due to the higher aspect ratio of the conductive additive that facilitates the contact between the particles under loading.Among the studied compositions, the composite with 10 wt% of Mt-PPy presented the highest sensitivity (2.5 MPa -1 ).The sensitivity of this composite prepared in this study was quite similar to the values found for the polymer system with 30 wt% of conductive contents prepared by Muller et al. 23 , Souza Jr., Michel and Soares 11 and Barra et al. 12 .It is important to note that the compressive stress range applied in this work (0.25 MPa) is significantly lower than those reported in literature (from 0 to 10 MPa) 8,[15][16][23][24] , as shown in Table 3. On th other hand, the sensitivity of PBAT/Mt-PPy is lower when compared to porous materials studied by Xue et al. 28 and Wu et al. 17 .The great advantages of this c are its low cost, good sensitivity, signal reproducibility and the biodegradability of PBAT.

Conclusion
A pressure sensitivity material based on PBAT/Mt-PPy was successfully prepared by melt mixing with a microextruder and then injection molding.The structure and properties of these materials are significantly influenced by the composition of the conductive filler (Mt-PPy and PPy) incorporated in the PBAT.The PBAT/Mt-PPy composite exhibits a higher electrical conductivity and lower percolation threshold than the PBAT/PPy blend due to the layered morphology of the filler that displays a high aspect ratio and surface area, which improves the formation of conductive paths.Moreover, the intercalated/partially exfoliated structure results in better dispersion of the Mt-PPy in the matrix, which also contributes to increase the conductivity of the composite.The electromechanical response is dependent on the amount and type of conductive filler in the insulating polymer matrix.Under a compressive stress, among the studied compositions, PBAT containing 10 wt% of Mt-PPy, showed the highest sensitivity, with a variation of electrical resistivity of about one order of magnitude, when a compressive stress of 0.25 MPa was applied.For this composite, the electrical resistivity response is highly reproducible after 25 repeated loading-unloading cycles, which makes this composition a promising material for the development of pressure sensors.

Figure 2 Figure 1 .
Figure2shows the curves of the electrical conductivity of PBAT/Mt-PPy and PBAT/PPy as a function of the filler content.PBAT/Mt-PPy composites show a very sharp insulator-conductor transition and the electrical conductivity enhanced significantly as the Mt-PPy content increased due to the formation of a conductive network in the insulating polymer, reaching a maximum value of 10 -6 S cm -1 .On the other hand, for the PBAT/PPy blends the increase in electrical conductivity as a function of the PPy content is less pronounced.The composite containing 10 wt% of Mt-PPy shows an electrical conductivity of 1.3 x 10 -6 S cm -1 , which was higher compared to the PBAT/PPy blend (4.3 x 10 -12 S cm -1 ) for the same amount of conductive filler.

Figure 2 .
Figure 2. Effect of the conducting filler contents (Mt-PPy and PPy) on electrical conductivity.

Table 1 .
Area of the peaks and ratio of the area under the peaks of the C=O groups (hydrogen bonded C=O (A 1710 ) and free C=O (A 1730) ).

Table 3 .
Electroactive materials used in pressure sensors.
*The results of this work.