Correlating mechanical and electrical properties of filler-loaded polyurethane fluoroelastomers: the influence of carbon black.
Perfluoropolyethers (PFPEs) are high-performance lubricant fluids characterized by excellent thermal, chemical, and photochemical stability (1). A variety of bifunctional and polyfunctional PFPE oligomeric derivatives (2), (3) have been developed in the past decades as specialty lubricants, additives, and macromers for the synthesis of segmented copolymers. Among them, various structures such as polyamides, polytriazines, polyesters, polyurethanes, and polyacetals have been described (4-10). In particular, PFPE-segmented polyurethanes have found interesting applications because of their peculiar properties. In fact, the very low surface tension and high photochemical stability of low-molecular-weight PFPE macromers were exploited for the design of easy-cleaning and durable surfaces to be applied both as solventborne as well as waterborne coatings (11-13). Higher molecular weight dihydroxy-functionalized PFPEs were instead used for the preparation of segmented thermoplastic and thermosetting elastomers (14), (15). PFPE-based polyurethane rubbers are characterized by high chemical resistance to both hydrocarbon and polar chemicals and excellent low temperature elasticity. In the very specialized field of fluoroelastomers for seals and gaskets (16), the use of PFPE-based rubbers (17) may offer the advantage of a better low-temperature sealing ability when compared with conventional polyfluoroolefines characterized by much higher [T.sub.g] values.
Most PFPE polyurethanes described so far are unfilled cast elastomers, not designed for processing and compounding with standard rubber mixing machines and technology. Only recently a few examples of millable PFPE polyurethanes fluoroelastomers have been described in the literature, with particular emphasis on their microstructure and viscoelastic behavior (18).
To further explore the characteristics of this type of materials, we present in this work a study on the dynamic mechanical and electrical properties of PFPE-based polyurethanes filled with different amounts and types of carbon black. Carbon black is normally used in fluoroelastomeric compounds both for cost reduction and as a property modifier (19). In the latter case, carbon black can dramatically change both the mechanical behavior (hardness, abrasion resistance, and damping) and the electrical conductivity of the rubber compound, targeting special uses such as antistatic or electromagnetic interference (EMI) shielding products. With the results obtained from our work, a more detailed knowledge of the basic relations between carbon black-filled compound composition and properties can be achieved, thus providing useful information for the development of a new family of high-performance sealing materials.
Materials and Compound Preparation
The preparation of the PFPE-polyurethane matrix used in this study as starting material is based on a two-step process described elsewhere (18). Polymers were filled with different amounts (10-50 phr) and types of carbon black [fast extrusion furnace (FEF) N550 (finely structured) or medium thermal (MT) N990 (coarsely structured) supplied by Carbocrom S.R.L.; see Table 1] as well as with dicumylperoxide in a Brabender-type mixer (20 min, room temperature). The mixing chamber volume was 50 [cm.sup.3]. All compounds were then casted and press cured at 160[degrees]C for 20 min to obtain 2-mm-thick samples. Postcuring at 120[degrees]C for 24 h allowed complete vulcanization. All reagents and monomers were purchased from Sigma-Aldrich and used as received.
TABLE 1. Main features of finely structured N550 (FEF) and coarsely structured N990 (MT) carbon blacks considered in this study as provided by supplier. Methods (units) FEF MT Surface Nitrogen adsorption 42 9 area ([m.sup.2]/g) Iodine adsorption 43 8 (mg/g) Structure DBP absorption 121 43 (ml/100 g) Particle 40-18 201-500 size (nm) pH value 6-9 9-11 Apparent 360 -- (bulk) density (g/l) Hydrogen 0.30-0.39 0.27-0.34 (%) Oxygen 0.15-0.65 0.00-0.10 (%) Abbreviation: DBP, dibutylphthalate.
Dynamic mechanical analysis (DMA) was performed with a DMA/SDTA[861.sup.e] Mettler Toledo dynamic mechanical analyzer in the linear viscoelastic range with 2-mm-thick, 5-mm-wide specimens in shear-sandwich mode in temperature scans from -140 to +150[degrees]C and oscillation frequencies of 1000, 100, 10, and 1 Hz. The linear visco-elastic range was separately determined with isothermal strain amplitude sweeps at 1 Hz.
Electrochemical impedance spectroscopy (EIS) was carried out with a frequency-response analyzer Solartron 1260, applying a sinusoidal voltage (2 V maximum) in the frequency range between [10.sup.-2]and [10.sup.6] Hz. All EIS measurements were performed under dry atmosphere and constant temperature so as to ensure test reproducibility and accuracy. Each compound was tested at 40, 50, 60, 70, and 80[degrees]C.
Differential scanning calorimetry (DSC) was performed with a DSC/[823.sup.e] Mettler Toledo differential scanning calorimeter at a scan rate of 20 K/min.
RESULTS AND DISCUSSION
Dynamic Mechanical Analysis
DMA allowed the determination of the complex shear modulus G*, the storage shear modulus G', and the loss shear modulus G" for each compound. Figure 1 shows DMA curves (G' as a function of temperature T) at 1 Hz for compounds filled with increasing amount of finely structured carbon black (N550, from here on referred to as FEF) or coarsely structured carbon black (N990, from here on referred to as MT).
As a comparison, DMA curves for the nonvulcanized (M) and the vulcanized (VM) fluoroelastomer matrix are also shown. Above the PFPE glass transition temperature (Tg) occurring at approximately -110[degrees]C (see Fig. 2), the uncured matrix M shows a value for G' (about 10 MPa) higher than both 10 FEF (fluoroelastomeric matrix loaded with 10 phr of FEF) and all MT compounds. This behavior may be attributed to the presence of urethane crystalline domains in the elastomeric matrix M, which are largely suppressed when vulcanization occurs. The vulcanized matrix (VM) shows the lowest value of G' when compared with all filled compounds, likely because the crosslinking process hinders conformational regularity needed for crystallization. In accordance with DSC measurements (see Supporting Information), M shows a sharp decrease in G' at around 60[degrees]C where melting of the crystalline domains occurs.
The mechanical behavior of filled fluoroelastomers appears to be highly dependent on the type of carbon black, as expected. As shown in Fig. 1, the introduction of FEF carbon black enhances G' both below and above the low-temperature [T.sub.g], resulting in a two-order-of-magnitude increase in the case of maximum carbon black loading (50 FEF) in comparison with the VM. Similar trends were found at higher oscillation frequency, with a shift of thermal transition phenomena to higher temperatures (see Supporting Information). Conversely, G' for MT compounds does not show any improvement with respect to M even for high MT loading. This behavior may be explained by considering the different nature and features of carbon black types used in this work (see Table 1) (20), (21). Actually MT consists of large size, low surface area, crystalline particles, and it does not normally show any reinforcing effect.
The loss factor tan [delta] (= G"/G') versus temperature at 1 Hz for FEF compounds at increasing FEF loading is shown in Fig. 2, where tan [delta] curves for M and VM are also included for reference. Three regions where relaxation phenomena occur can be clearly distinguished in accordance to the polyphasic nature of the fluoroelastomers. A first transition occurs at around -100[degrees]C, where a sharp tan [delta] peak is observed. This peak can be assigned to the glass transition of PFPE segments in the macromolecular chain: the very low-solubility parameter (around 10.5 [MPa.sup.0.5]) of the fluorophase in PFPE block copolymers makes them nearly immiscible in hydrogenated chain polymers, giving rise to a distinct glass transition. This behavior was already observed for a variety of PFPE copolymer structures in comparison with nonfluorinated materials (7), (22), and it seems unaffected by the presence of filler particles. A second broad transition is found in the temperature range 0-20[degrees]C, where an asymmetric tan 6 peak is observed ascribable to a transition associated to the urethane group. Finally, a third transition can be seen at temperatures close to the melting temperature of the elastomeric matrix M (60[degrees]C/100[degrees]C), which can be related to the presence of residual crystalline or paracrystalline interactions not completely suppressed by crosslinking. Similar trends were observed on all MT-based compounds, as shown in the Supporting Information. When considering the transition occurring at -100[degrees]C (Fig. 2 and Supporting Information), the VM shows the highest intensity of viscoelastic dissipation when compared with both FEF and MT compounds. On the contrary, the nonvulcanized matrix M shows a considerably lower tan [delta] peak intensity when compared with VM (about 40% lower). This may be attributed to the presence of residual crystalline domains in the nonvulcanized matrix, with reduction of the relative amorphous phase to which the dissipation phenomena can be ascribed.
The effect of frequency (1-1000 Hz) on the low-temperature tan [delta] peak for FEF compounds is shown in Fig. 3. A 70% decrease in tan [delta] peak intensity is observed when passing from 10 to 50 phr compounds for all frequencies. Similar trends were found for MT compounds, although a less marked decrease (about 40%) was observed from 10 to 50 phr (see Supporting Information). The reduction of tan [delta] peak intensity at increasing carbon black loading may be explained in terms of polymer-filler and filler-filler interactions (23-26). In the carbon black-filled compound, the filler aggregates are surrounded by a thin polymer film tightly bound to the filler surface that acts as a separating layer between the aggregates. This layer possesses a polymer chain mobility lower than the matrix bulk as a result of the strong polymer-filler interactions. In addition, part of the polymer matrix may get trapped in the filler aggregates leading to an increase of the effective volume fraction of the filler. By increasing filler loading, the filler-to-filler distance decreases and results in the formation of a three-dimensional network of filler aggregates containing trapped low-mobility polymer chains. The existence of this network substantially increases the effective volume fraction of the filler because of the polymer matrix trapped in the agglomerates and leads to an increase in the elastic modulus and a decrease in the hysteresis. The interactions between rubber and filler particles as well as the reinforcing effect of the bound rubber at a given filler content strongly depend on the features of the filler. In particular, this effect is found to be less evident when surface area, structure, and surface activity of the filler are lower, as confirmed by the trends observed in MT compounds when compared with FEF compounds (see Supporting Information). A linear dependence of the transition temperature on the applied frequency was found for both FEF and MT compounds, confirming the second-order nature of this thermal transition. Accordingly, linear interpolation of transition temperature data plotted against frequency on logarithmic scale results in a constant slope of about 5 [degrees], which is typical for an [alpha]-transition (27) such as the glass transition (Table 2).
TABLE 2. Slope dT/d(log f) [+ or -] standard error (SE) estimated for linear regression of the low-temperature tan [delta] peak temperature as a function of the logarithm of frequency f and obtained for each type of carbon black (FEF, MT) at increasing loading (10 to 50 phr). Carbon black dT/d(log f) [+ or -] SE X [10.sup.-2] (Adj [R.sup.2]) loading (phr) FEF MT 10 5.8 [+ or -] 0.4 (0.985) 6.3 [+ or -] 0.3 (0.993) 20 5.2 [+ or -] 0.5 (0.974) 5.6 [+ or -] 0.4 (0.985) 30 5.5 [+ or -] 0.2 (0.996) 5.9 [+ or -] 0.6 (0.971) 40 6.1 [+ or -] 1.1 (0.911) 6.0 [+ or -] 0.5 (0.981) 50 4.8 [+ or -] 0.4 (0.975) 6.4 [+ or -] 0.7 (0.967) The adjusted determination coefficient (Adj [R.sup.2]) is also shown for each estimated dT/d(log f) value.
The width of tan [delta] curve in the -125[degrees]C to -75[degrees]C temperature range (see Fig. 2) for both FEF and MT compounds at varying filler content was examined as a qualitative estimation of the distribution of relaxation times. Each tan (5 curve was first normalized to the maximum height of the peak, then the rising and descending parts of each curve were linearly interpolated, and the magnitude of the slope of the interpolating line was taken. As reported in Table 3, a decrease of the slope in the interpolating line (indicating a widening of the peak) is observed at increasing carbon black loading for both FEF and MT compounds. This effect is more evident for FEF carbon black likely because of its more porous and less graphitized nature when compared with MT carbon black, which determines a more efficient interaction with the polymer chains. This stronger interaction may hamper the velocity of the thermal transition, thus causing a more pronounced widening of the relaxation time distribution.
TABLE 3. Slope k [+ or -] standard error (SE) estimated by numerically fitting the rising and descending parts of the low- temperature tan [delta] peak at 1 Hz obtained for each type of carbon black (FEF, MT) at increasing loading (0-50 phr). Carbon Rising part, Descending part, black k [+ or -] SE X k [+ or -] SE X [10.sup.-2] [10.sup.-2] (Adj [R.sup.2]) (Adj [R.sup.2]) loading FEF MT FEF MT (phr) 0 8.14 [+ or 6.68 [+ or -] -] 0.24 0.17(0.98) (0.96) 10 8.15 [+ or 8.33 [+ or 6.54 [+ or 6.57 [+ or -] 0.09 -] -] -] 0.20 (0.99) 0.14(0.99) 0.19(0.96) (0.94) 20 7.86 [+ or 7.94 [+ or 6.47 [+ or 6.63 [+ or -] 0.16 -] 0.15 -] 0.13 -] 0.13 (0.98) (0.98) (0.98) (0.98) 30 7.88 [+ or 7.78 [+ or 5.81 [+ or 6.51 [+ or -] 0.08 -] -] -] (0.99) 0.16(0.98) 0.07(0.98) 0.13(0.98) 40 7.89 [+ or 7.91 [+ or 5.86 [+ or 6.39 [+ or -] 0.15 -] -] -] 0.09 (0.98) 0.09(0.99) 0.10(0.98) (0.99) 50 7.60 [+ or 8.20 [+ or 5.52 [+ or 6.36 [+ or -] 0.17 -] 0.15 -] -] (0.98) (0.99) 0.10(0.99) 0.12(0.98) The adjusted determination coefficient (Adj [R.sup.2]) is also shown for each estimated k value.
Hydrodynamic Effect of Carbon Black
The increase of G' achievable by compounding elastomers with carbon black particles can be partly related to the hydrodynamic effect of the dispersed solid particles in the elastomeric matrix. The extent of such effect depends on the volume fraction of carbon black present in the compound. To predict the mechanical properties of carbon black-loaded elastomers compounds as a function of the loading fraction, Guth's theory can be used (28). Based on the assumptions that the medium wets the filler particles but does not chemically react with the filler surface, Guth's theory gives an estimate of G' for a filler-loaded elastomer as follows:
G' = [[G.sub.u].sup.'](1 + 2.5c + 14.1[c.sup.2]) (1)
where [[G.sub.u].sup.'] is the storage modulus for the neat elastomer and c is the volume concentration of the filler. To take into account the contribution to G' given by the fraction of polymer trapped in the filler's pores, Medalia's correction to the effective volume of filler ce is applied (29):
[c.sub.e]/c = 0.5[1 + (l + 0.02139 x DBP)/1.46] (2)
where DBP is the volume of dibutylphthalate absorbed by the filler particles, taken as a measure of the volume fraction of elastomeric matrix trapped in the filler's pores. For nonspherical asymmetrical filler particles or anisotropic aggregates, the following equation was proposed (3O]:
G' = [[G.sub.u].sup.'](1 + 0.67sc + 1.62[s.sup.2][c.sup.2]) (3)
where s is a shape factor (ratio of diameter to width of particle).
In this work, the values of G' in the rubbery plateau obtained from DMA analysis at varying carbon black phr were compared with the values of G' calculated by means of four different models: Guth's model (Eq. 1), Guth's model including Medalia's correction to the effective volume of filler [c.sub.e] (Eq. 2), Guth's model including a shape factor s (Eq. 3), and Guth's model including both Medalia's correction and a shape factor. The numerical results are summarized in Table 4. When included, the value of shape factor was tuned so that the best fit to the experimental G' values could be attained.
In the case of FEF compounds at low carbon black content (see Supporting Information), all models give a G' value higher than that experimentally measured by DMA analysis, suggesting a negligible influence on calculated G' of both particle asymmetry and polymer fraction trapped in the filler's pores. By progressively increasing the amount of carbon black in the compound, the effect of particle asymmetry becomes increasingly important, as shown by the progressively lower value of G' calculated by the models corrected with the shape factor. In particular, for 50 FEF compound, all models predict a G' value lower than what found experimentally. As shown in Table 4, the shape factor value for FEF compounds increases with carbon black phr, suggesting that simple Guth's model cannot give an accurate estimate of G' for high carbon black loading. This effect may be explained by considering the formation of a three-dimensional network of particle aggregates and agglomerates forming at high FEF carbon black loading (above a threshold value around 30 phr) that would result in an actual reinforcing effect higher than what predictable by simple Guth's model. When Guth's model including Medalia's correction is considered, the shape factor appears negligible up to very high phr (50 FEF). In addition, G' calculated by this model attains higher values than the experimental up to 40 phr. This may be related to a negligible contribution of the matrix trapped within the filler particle pores to the mechanical reinforcement of the compound. Accordingly, Guth's model corrected only by the shape factor appears to give a more accurate prediction of G'.
TABLE 4. Best-fitting values of shape factor s [+ or -] standard error (SE) estimated for the storage modulus G'according to Guth model (Eq. 3) without and with Medalia's correction (Eq. 2) at increasing FEF loading. Carbon Guth model without Guth model without black Medalia's correction Medalia's correction loading k [+ or s Adj k [+ or s Adj (phr) -] SE [R.sup.2] -] SE R.sup.2] 0 N/A N/A N/A N/A N/A N/A 10 N/A N/A N/A N/A N/A N/A 20 N/A N/A N/A N/A N/A N/A 30 9.362 [+ 3.622 0.989 N/A N/A N/A or -] 0.046 40 12.696 [+ 3.880 0.994 N/A N/A N/A or -] 0.046 50 26.456 [+ 5.447 0.989 26.456 [+ 3.150 0.989 or -] or -] 0.127 0.127 Linear regressions of [[G.sub.FEF].sup.'] versus [[G.sub.VM].sup.'] plots at varying FEF loading were performed in the rubbery plateau, and s was obtained from the slope k of the regression curve according to Eq. 3. [[G.sub.FEF].sup.'] is the storage modulus of FEF-loaded compounds and [[G.sub.VM].sup.'] is the storage modulus of the vulcanized matrix VM. The adjusted determination coefficient (Adj [R.sup.2]) is also shown for each estimated k value.
In the case of MT compounds (see Supporting Information), all models predict a value of G' higher than measured for all carbon black phr explored in this work, suggesting a negligible contribution of aggregate asymmetry and trapped elastomer to compound reinforcement. The best fit with experimental data is observed at low carbon black loading (10 MT), likely because the carbon black volumetric fraction is still too low to affect significantly G'.
Shear Displacement Tests
Shear displacement tests were carried out on all compounds to examine the effect of different carbon black loading on compound dynamic behavior at relatively large displacement (up to 10%). An increase of G' is observed at increasing phr both in FEF compounds (Fig. 4a), where this effect is more noticeable, and in MT compounds (see Supporting Information). In the case of FEF compounds, the neat elastomer M shows higher G' than both VM and 10 FEF, in accordance with results obtained from DMA as discussed above. As expected, the linearity region where G' maintains a constant value decreases at increasing carbon black loading.
The so-called Payne effect (31), (32) was reproducibly observed in FEF compounds (Fig. 4), where a decrease of storage modulus G' was found at increasing displacements together with the occurrence of a maximum of the loss modulus G" in the same region where G' decreases. This effect is attributed to the rapid breaking and rebuilding of filler aggregates due to changes in the compound microstructure induced by the increasing deformation. Conversely, MT compounds do not show any peak in G", even at high phr (see Supporting Information). Because MT particles possess a more graphitic-like structure than FEF particles, together with a lower surface area and a lower structure, they can only weakly interact with the polyurethane chains, without generating any further dissipative effect.
Electrochemical Impedance Spectroscopy
EIS was used to evaluate the electrical behavior of all FEF and MT compounds. Nyquist plot (33) for the vulcanized unfilled elastomer at increasing temperature (Fig. 5a) shows the semicircular shape typical of conductive materials, which confirms the slightly conductive character of the unfilled polymer matrix. This behavior is not unusual and may be attributed to the presence of ionic impurities (typically the catalyst) in the polyurethane. In contrast, two semicircles are observed in the case of carbon black--filled compounds (10 FEF was chosen as an example in Fig. 5b), suggesting the presence of two distinct relaxation phenomena. In the high-frequency region, a larger semicircle is observed similar to what found for the elastomeric matrix, ascribable to charge motion in the bulk of the material (ionic conductivity). On the other hand, a smaller semicircle is observed at low frequencies, likely clue to charge accumulation giving rise to polarization phenomena at the interface between carbon black particles and elastomeric matrix. This effect is observed at low frequencies where ions are given a longer time to move toward accumulation regions.
A decrease of resistivity [rho] is reported at increasing temperature both in the elastomeric matrix and in the carbon black--filled compounds. In particular, an Arrhenius-type dependence of [rho] on temperature is observed for both FEF (Fig. 6a) and MT (Fig. 6b) compounds. The value of resistance R (and [rho] thereby) was obtained from the Z' = Re(Z) value extrapolated from the Nyquist plot at [omega] [right arrow] 0 (where [omega] is the angular oscillation frequency). Interpolation of data of [rho] versus 1/T allows the calculation of the activation energy [E.sub.A] of the electrical conduction process for all compounds (Table 5).
TABLE 5. Activation energy [E.sub.[LAMBDA] [+ or -] standard error (SE) estimated for the Arrhenius-type curves of resistivity [rho] and obtained for each type of carbon black (FEF, MT) at increasing loading (0-50 phr). FEF MT Carbon [E.sub.[LAMBDA] Adj [E.sub.[LAMBDA]] Adj black [+ or -] SE [R.sup.2] [+ or -] SE (eV) [R.sup.2] loading (eV) (phr) 0 0.672 [+ or -] 0.993 0.672 [+ or -] 0.993 0.032 0.032 10 0.743 [+ or -] 0.998 0.741 [+ or -] 0.998 0.015 0.015 20 0.730 [+ or -] 0.988 0.807 [+ or -] 0.992 0.041 0.036 30 0.511 [+ or -] 0.997 0.646 [+ or -] 0.996 0.013 0.019 40 <0 -- 0.671 [+ or -] 0.999 0.009 50 <0 -- 0.656 [+ or -] 0.995 0.022 The adjusted determination coefficient (Adj [R.sup.2]) is also shown for each estimated [E.sub.[LAMBDA] value.
The values of resistivity [rho] calculated at 40[degrees]C as a function of phr for both MT and FEF compounds are shown in Fig. 7. In the case of MT compounds, no significant variation of the activation energy and of the resistivity is observed at increasing carbon black loading. Only small fluctuations in the [E.sub.A] value are observed, which can be attributed to anisotropic conductivity within the graphitized carbon black particles. In addition, the presence of MT carbon black does not appear to modify significantly the activation energy of the elastomeric matrix, suggesting that the electrical conductivity observed in MT compounds can be attributed to the intrinsic conductivity of the pure elastomeric matrix. After the application of an electric field, ionic charges present in the sample and resulting from the presence of impurities from the polymerization reaction are allowed to move through the sample and accumulate at the interface with the electrodes where they can form a charged layer.
In the case of FEF compounds, the activation energy of the conduction process appears to be affected by carbon black loading. For low phr values (up to 20 phr), a behavior similar to that reported for MT compounds is observed, with a plateau value for resistivity at 40[degrees]C comparable with that measured for the elastomeric matrix. At 30 phr, a decrease of resistivity is observed together with a 30% drop in the EA value. This suggests that by increasing carbon black loading, the influence of thermal activation on the conduction process progressively weakens, leading to [E.sub.A] values that are typical of electronic conduction. Concurrently, a sharp decrease in [rho] at 40[degrees]C is observed (Fig. 7). For higher carbon black loading (40 and 50 phr), the dependence of [rho] on temperature is reversed, and the compounds show the typical behavior of electronic conductors. The value of resistivity at 40[degrees]C undergoes a nine-orders-of-magnitude decrease when the carbon black content is increased up to 50 phr. Considering the rather low temperature at which EIS tests are carried out ([T.sub.test] = 40-80[degress]C), it is likely that thermal expansion of the sample under test is negligible, and therefore, any increase in resistivity due to separation of carbon black particles from each other is prevented. Furthermore, [T.sub.test] is high enough to exclude the tunnel effect from being the main conduction mechanism occurring in the material. As a result, hopping of electrons between carbon black particles is believed to be the main conduction mechanism occurring in FEF compounds.
The different electrical behavior observed for MT when compared with FEF compounds may be related to the different dimensions and surface characteristics of MT with respect to FEF carbon black particles. MT particles are larger than FEF and show a highly crystalline surface. Conversely, FEF particles possess a very fine structure, prone to formation of aggregates within the compound that tend to become larger at increasing phr. Therefore, when the carbon black loading is high enough, electrical percolation may occur, resulting in a sharp decrease of compound resistivity.
A correlation may be found between electrical and dynamic mechanical properties of FEF and MT compounds. As shown by electrical measurements, an electrical percolation threshold is observed for FEF compounds at 30 phr, followed by a sharp decrease in compound resistivity. Starting from the same filler fraction, DMA measurements and calculations showed that filler shape and asymmetry increasingly affected mechanical reinforcement in the case of FEF compounds. This shape effect was explained by considering the possible formation of carbon black agglomerates and three-dimensional anisotropic structures at high FEF carbon black loading. Therefore, both filler loading and filler shape appear to influence electric and mechanical properties in FEF compounds. On the other hand, filler shape and asymmetry for MT carbon black structures did not appear to give a significant contribution to mechanical reinforcement of MT compounds. Concurrently, no electrical percolation threshold is reached for MT compounds up to 50 phr.
The mechanical and electrical properties of a family of carbon black--filled polyurethane-based fluoroelastomeric compounds were investigated in this work as a function of filler type and loading. DMA on the unfilled polymer revealed the presence of three distinct relaxation phenomena, confirming the phase-segregated nature of the copolymer. In particular, the [alpha]-transition at low temperature appeared to be highly dependent on both the amount and the type of filler, suggesting the presence of strong polymer--filler interactions influencing the distribution of relaxation times.
Well-known models of mechanical reinforcement were used to fit the experimental DMA data. It was shown that for finely structured carbon black (FEF) compounds, the effect of filler particle asymmetry on compound reinforcement becomes increasingly important as the filler content increases above a threshold value. This behavior may be explained by considering the formation of a three-dimensional network of particle aggregates and agglomerates within the fluoroelastomeric matrix, which participates in the reinforcement of the elastomeric compound. In addition, the effective filler volume fraction may increase due to part of polymer matrix trapped in the filler aggregates. This effect was not observed in coarsely structured carbon black (MT)--filled compounds, suggesting no formation of higher level structures.
Impedance spectroscopy measurements highlighted some differences in the electrical behavior of filled fluoroelastomers. Although no significant variation was reported in the case of MT compounds, an electrical percolation threshold was observed for FEF compounds, accompanied by a sharp decrease in resistivity. This percolation threshold is observed for the same FEF fraction as that found in DMA, supporting the hypothesis of the presence of carbon black aggregates and three-dimensional anisotropic structures forming at high FEF loading.
The results presented in this work provide a greater understanding of structure-to-property relationships and polymer--filler interactions in carbon black--filled polyurethane fluoroelastomeric compounds, giving useful guidelines for the design of high-performance carbon black--filled polyurethane fluoroelastomers.
The authors thank Gigliola Clerici for carrying out DMA measurements. They also thank Prof. Giovanni Dotelli and Prof. Marinella Levi for helpful discussions.
(1.) G. Marchionni, G. Ajroldi, and G. Pezzin, "Structure--Property Relationships in Perfluoropolyethers: A Family of Polymeric Oils," in Comprehensive Polymer Science, S. Russo, L.S. Aggarwal, and G. Allen, Eds., Pergamon Press, Oxford, 347 (1996).
(2.) S. Turn, M. Scicchitano, and C. Tonelli, J. Polym. Sci. Part A. Polym. Chem., 34, 3263 (1996).
(3.) C. Tonelli, A. Di Meo, S. Fontana, and A. Russo, J. Fluorine Chem., 118, 107 (2002).
(4.) R.W. Rosser and R.A. Korus, J. Polym. Sci. Polym. Lett. Ed., 18, 135 (1980).
(5.) J.F. Ding, A.R. Khan, M. Proudmore, R.H. Mobbs, F. Heatley, C. Price, and C. Booth, Macrornol. Chem. Phys., 195, 3137 (1994).
(6.) G. Caporiccio, E. Strepparola, G. Bargigia, G. Novaira, and G. Peveri, Makromol. Chem., 184, 935 (1983).
(7.) D.M. Bigg, K.J. Heater, T.L. Skidmore, and M. Arai, Polym. Eng. Sci., 45, 1622 (2005).
(8.) M. Levi and S. Turn, J. Polym. Sci. Part A: Polym. Chem., 36, 939 (1998).
(9.) P.G. Edelman, D.G. Castner, and B.D. Ratner, ACS Polym. Prepr., 31, 314 (1990).
(10.) S. Turn, A. Sanguineti, and M. Levi, Macroniol. Chem. Phys., 198, 3215 (1996).
(11.) G. Simeone, S. Turn, M. Scicchitano, and C. Tonelli, Angew, Makromol. Chem., 236, 111 (1996).
(12.) S. Turn, S. Radice, R. Canteri, G. Speranza, and M. Anderle, Swf. Interface Anal., 29, 873 (2000).
(13.) R. Canteri, G. Speranza, M. Anderle, S. Turn, and S. Radice, Swf. Interface Anal., 35, 318 (2003).
(14.) C. Tonelli, T. Trombetta, M. Scicchitano, G. Simeone, and G. Ajroldi, J. Appl. Polym. Sci., 59, 311 (1996).
(15.) M. Proudmore, J.F. Ding, R.H. Mobbs, S.G. Yeates, C. Price, and C. Booth, J. Polym. Sci. Part A: Polym. Chem., 33, 1615 (1995).
(16.) B. Ameduri and B. Boutevin, J. Fluorine Chem., 126, 221 (2005).
(17.) S. Turri, M. Levi, M. Cristini, and A. Sanguineti, J. Polym. Res., 14, 141 (2007).
(18.) S. Turri, R. Valsecchi, M. Levi, M. Cristini, and A. Sanguineti, Eur. Polym. J., 44, 2951 (2008).
(19.) V. Arcella and R. Ferro, "Fluorocarbon Elastomers," in Modern Fluoropolymers: High Performance Polymers for Diverse Applications, J. Scheirs, Ed., John Wiley & Sons, 81 (1997).
(20.) M.J. Wang, S. Wolff, and J.B. Donnet, Rubber Chem. Technol., 64, 714 (1991).
(21.) S. Wolff, and M.J. Wang, Rubber Chem. Technol., 65, 329 (1992).
(22.) G. Consolati, M. Levi, and S. Turn, Polymer, 42, 9723 (2001).
(23.) J. Frohlich, W. Niedermeier, and H.D. Luginsland, Compos. A, 36, 449 (2005).
(24.) J.L. Leblanc, Prog. Polym. Sci., 27, 627 (2002).
(25.) M.J. Wang, S. Wolff, and E.H. Tan, Rubber Chem. Technol., 66, 178 (1993).
(26.) S. Wolff, M.J. Wang, and E.H. Tan, Rubber Chem. Technol., 66, 163 (1993).
(27.) L.W. Nielsen, Mechanical Properties of Polymers and Composites, Marcel Dekker, New York (1974).
(28.) A.R. Payne, "Dynamic Properties of Filler-Loaded Rubbers," in Reinforcement of Elastomers, G. Kraus, Ed., Inter-science Publishers, New York, 69 (1965).
(29.) A.I. Medalia, Rubber Chem. Technol., 45, 1171 (1972).
(30.) S. Wolff and J.B. Donnet, Rubber Chem. Technol., 63, 32 (1990).
(31.) A.R. Payne, J. Appl. Polym. Sci., 6, 57 (1962).
(32.) A.R. Payne, J. Appl. Polym. Sci., 6, 368 (1962).
(33.) J. Ross Macdonald, Ann. Biomed. Eng., 20, 289 (1992).
Gianmarco Griffini, Raffaella Suriano, Stefano Turri
Department of Chemistry, Materials and Chemical Engineering 'Giulio Natta' Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy
Additional Supporting Information may be found in the online version of this article.
Correspondence to: Gianmarco Griffini; e-mail: gianmarco.griffini@chem. polimi.it
Published online in Wiley Online Library (wileyonlinelibrary.com).
[C] 2012 Society of Plastics Engineers
|Printer friendly Cite/link Email Feedback|
|Author:||Griffini, Gianmarco; Suriano, Raffaella; Turri, Stefano|
|Publication:||Polymer Engineering and Science|
|Date:||Dec 1, 2012|
|Previous Article:||Biodegradation of plastified starch obtained by corotation twin-screw extrusion.|
|Next Article:||Polyvinylidene fluoride/poly(ethylene terephthalate) conductive composites for proton exchange membrane fuel cell bipolar plates: crystallization,...|