Electrical and Mechanical Behavior of Carbon Black--Filled Poly( Vinyl Acetate) Latex--Based Composites.
Balancing electrical conductivity with desirable mechanical behavior is one of the largest challenges facing the use of filled polymer composites in applications such as anti-static layers (1, 2), chemical vapor sensors (3-5), and thermal resistors (6). Typically. polymer matrices require large concentrations of carbon black (CB), or other conductive filler, to achieve sufficient electrical conductivity for various applications. These high filler concentrations often come at the expense of mechanical performance and ease of processing (7). It is precisely this ease of processing and durability of the final product that make conductive polymer composites attractive replacements for applications traditionally reserved for metals. As the interplay between electrical and mechanical behavior of these composites becomes better understood, further applications will likely present themselves.
Several methods of improving mechanical properties, while maintaining electrical conductivity, of polymer composites have been proposed. For example, Huang et al. (8) added plasticizer to a CB-filled elastomer and found that flexural modulus, hardness and secant modulus could all be maintained near that of the unfilled elastomer. Furthermore, elongation at break was actually greater than that of the unfilled elastomer. Electrical conductivity was slightly improved with the addition of plasticizer when compared to an equivalent unplasticized composite. Despite these improvements, plasticizer tends to degrade overall composite strength by disturbing the interaction between the matrix and filler. The percolation threshold seems unaffected by the addition of plasticizer. Another approach to solve this issue of property balancing is to lower the percolation threshold and thereby reduce the amount of filler needed to achieve electrical conductivity within the composite. It has been shown that polymer composites wi th lower filler concentration tend to have higher yield strength (9) and lower modulus (10, 11) than do comparable composites with higher filler loading. Several researchers have improved yield strength by surface treating filler, thereby improving interfacial strength between polymer and filler (12, 13). The silane coupling agents used to treat the filler, however, add a new variable that may alter other properties (e.g., modulus, electrical conductivity, etc.) and increase the cost of the overall composite system.
More recently, latex has been used as the composite matrix starting material and the percolation threshold has been reduced by almost an order of magnitude relative to comparable solution-(or melt-) based composites (14-16). In this report, the electrical and mechanical properties of carbon black filled latex composites were investigated in an effort to achieve an optimal balance of electrical and mechanical performance. Two different commercial poly(vinyl acetate) (PVAc) latices were used to produce conductive composites: one is polydisperse while the other is nearly monodisperse. Poly(vinyl acetate) latex has the unique ability to form glassy, coherent films at room temperature due to the large disparity between its glass transition temperature ([T.sub.g] [approximately equal to] 35[degrees]C) and minimum film formation temperature (MFFT [approximately equal to] 15[degrees]C). Properties of these latex-based composites are compared to those of a solution-(or melt) based composite wherever possible. The effe ct of latex coalescence on electrical and mechanical behavior of these composites is also investigated. Finally, a figure of merit has been proposed to determine the filler loading at which the most advantageous combination of electrical conductivity, break strength and modulus is achieved.
Three aqueous-based, polymer-matrix, starting materials were used in this study. Vinac XX210 (Air Products, Inc.) is a polydisperse PVAc homopolymer latex with 49.8 vol% solids and a number average ([D.sub.n]) and volume average ([D.sub.v]) particle sizes of 108 nm and 2.5 p.m. respectively. PD 0202 (H. B. Fuller Co.) is a nearly monodisperse PVAc homopolymer latex with 40.9 vol% solids and a [D.sub.n] of 116 nm. Particle size analysis was performed using a Coulter LS Particle Size Analyzer 3.00.40. Both of these latices have glass transitions temperatures ([T.sub.g]'s) of approximately 35[degrees]C and minimum film formation temperatures (MFFTs) near 15[degrees]C. Poly(N-vinylpyrrolidone) (PVP) powder was purchased from Aldrich and had a molecular weight ([M.sub.w]) of 55,000 g/mol and a [T.sub.g] of 175[degrees]C. PVP is a water-soluble polymer that acted as the solution-based composite matrix starting material. When creating aqueous composite mixures, all of the polymers were first prepared as suspensions or solutions with 10 vol% solids. The carbon black (trade name Conductex 975 Ultra) used as the conductive filler was supplied by Columbian Chemicals Company. Conductex carbon black has a 21 nm primary particle size, 1.89 g/[cm.sup.3] density and a [N.sub.2] surface area of 242 [m.sup.2]/g.
Aqueous composite mixtures were prepared using a high-speed impeller. Carbon black powder was added, over the course of 5 minutes, to polymer solutions and suspensions containing 10 vol% solids while the impeller stirred at a rate of 880 rpm. In an effort to reduce the formation of foam during mixing. 5-10 drops of DrewPlus[R] L-483 foam control agent (Ashland, Inc.) was used. Following carbon black incorporation, the impeller speed was increased to 3600 rpm for 15 minutes. Composite mixtures with lower carbon black concentrations were prepared by further diluting the mixtures with more polymer solution or suspension followed by another 15 minutes of impelling at 3600 rpm. Final composite films were prepared by pouring the aqueous mixture into one inch square plastic molds and drying for 24 hours under ambient conditions followed by another 24 hours in a vacuum dessicator to completely remove residual water. The resulting composite films were 150-350 [micro]m thick.
Electrical conductivity measurements were made on freestanding composite films using a Veeco FPP-5000 four-point probe (Veeco Instruments, Inc.). Electrical conductivity below [10.sup.-5]S/cm could not be measured with this device. Cross-sectional images of the final composite films were obtained using a Hitachi S-800 field emission-gun, scanning electron microscope (FEGSEM). Surface images of the composite films were created using both the Hitachi microscope and a Nanoscope[R] III Multimode Scanning Probe Microscope (Digital Instruments) operated in tapping mode (TM). This scanning probe microscopy is often referred to as atomic force microscopy (AFM).
All of the composite mechanical properties were evaluated using a Perkin-Elmer DMA 7e dynamic mechanical analyzer. Storage modulus (E') was determined using a tensile fixture operated at a frequency of i Hz with 2200 and 2000 mN of static and dynamic load respectively to obtain oscillation amplitudes above 5 mm. For PVP-based composites, E' was measured using a three-point-bend fixture because of sample rigidity. Ultimate tensile strength (UTS) values were obtained by applying a static force ramp in tension from 3 to 6400 mN, at a rate of 500 mN/minute, using the extension fixture.
RESULTS AND DISCUSSION
The microstructural differences between the three composite systems are shown in amplitude images of composite surfaces obtained with tapping mode atomic force microscopy (TMAFM) (Fig. 1). Amplitude images are often referred to as error signal images due to the AFM's need to maintain a constant level of oscillation during imaging. Although true height information is lost in an amplitude image, fine features (such as the distribution of 21 nm carbon black particles) are better resolved (16). Figure 1 shows TMAFM surface images of the three composite systems under consideration at carbon black concentrations of 4 and 15 vol%. Composites made with the solution-based PVP (Fig. la-b) show no organized microstructure with regard to carbon black distribution, while a nicely segregated, carbon black network can be observed in the composites made with polydisperse latex (Fig. lc-d). As water evaporates from the composite system containing latex, polymer particles push the smaller carbon black into the interstitial sp aces between them. The excluded volume created by the latex particles causes the first networked carbon black structure to be formed at concentrations below 4 vol%. Another contributor to this segregation is the inability of the individual latex particles to wet the conductive filler, which would create a more homogeneous microstructure. For composites prepared with PVP solution, the carbon black is free to situate itself anywhere within the composite film and may even be stabilized by a thin coating of solvated polymer. This added stabilization and absence of excluded volume creates more evenly distributed carbon black and forces network formation to occur at much higher filler concentrations relative to the latex.
In the case of the monodisperse latex (Fig. 1e-f). the small latex particles (shown at a higher magnification relative to the previous samples) are only partially coalesced. It is difficult to discern the segregated carbon black network in these images because the latex particles are of a similar magnitude to carbon black aggregates. It is also possible that some small quantity of commercial additive may have deposited itself at the surface during ifim formation, which further disguises the carbon black. If these images were on the same scale as Fig. 1a-d). they would look much like Fig. 1 a and b but somewhat grainier because of the lack of full latex coalescence. Despite the inability to fully visualize carbon black in the monodisperse latex-based lattice, its presence is evident in the cross-sectional images obtained with SEM (see Fig. 2) and in electrical conductivity results (see Fig. 3).
Freeze-fractured cross sections of these composite films emphasize other differences between the latex-and solution-based composites. In Fig. 2 the two latex-based systems under consideration are shown loaded with 9 and 15 vol% carbon black and compared to PVP solution-based composites with the same filler levels. The segregated carbon black is very apparent in the 9 vol% images of the two latex systems (Figs. 2c and e). In these images, chainlike structures of carbon black are seen separated by relatively large areas of polymer that does not contain any noticeable amount of carbon black. At 9 vol% the image of the PVP composite (Fig. 2a) shows significant charging (seen as a white wavy pattern throughout the image) due to lack of composite electrical conductivity. No charging is observed in either the monodisperse (Fig. 2c) or polydisperse (Fig. 2e) latex composites because these systems have attained electrical conductivity with 9 vol% carbon black (see Fig. 3). With 15 vol% carbon black the PVP composite no longer exhibits the intricate charging pattern (Fig. 2b) observed at the lower filler concentration. It could be assumed from these images that the percolation threshold for the PVP composite system is between 9 and 15 vol% carbon black and for the two latex-based systems it falls below 9 vol%. A plot of electrical conductivity as a function of carbon black concentration, known as a loading curve, provides the precise location of the percolation threshold for these composites.
Electrically conductive polymer composites have been shown to obey the empirical, power-law conductivity relationship with volume concentration filler (V), once the percolation threshold is exceeded. This classical percolation power-law has the form (17):
[sigma] = [[sigma].sub.0][(V-[V.sub.c]).sup.s] (1)
where [sigma] is the composite conductivity (S/cm), [[sigma].sub.0] is a proportionality constant which often resembles the intrinsic conductivity of the filler (10-100 S/cm for carbon black), s is the power-law exponent (typically between 1.6 and 2 for a three-dimensional conductor (18)) and [V.sup.c] is the volume fraction of filler at the percolation threshold (near 0.15 for random three-dimensional systems). Loading curves for the three composite systems are shown in Fig. 3 with the solid curves shown representing the classical percolation power-law (Eq 1) fit to the experimental data. Table 1 summarizes the important fitting parameters obtained from the data in Fig. 3.
Both of the latex-based composite systems have percolation thresholds near 2 vol% carbon black, which is almost an order of magnitude lower than that of the solution-based composite system. The two most important factors contributing to this dramatic reduction in percolation threshold are lack of filler stabilization and excluded volume effects. In the solution-based system, the polymer is dissolved in water and may adsorb on the filler particles prior to film formation. This thin layer of polymer serves to stabilize the filler suspension and prevent aggregation, or intimate particle-particle contacts, during drying. Furthermore, the solution-based composite has no excluded volume effects during drying because the polymer matrix is fully solvated. In the case of the latex-based composites, the polymer matrix is suspended, not solvated, and pushes the smaller carbon black into its interstices. The onset of electrical conductivity in filled polymer composites is marked by the formation of pathways of conductiv e particles and the concentration of conductive filler needed to create the first pathway is taken to be the percolation threshold. By forcing the carbon black into interstitial space during drying, the latex is lowering the percolation threshold by incorporating filler particles in a non-random manner, thus creating a connected network at very low filler concentration (16). Despite. having lower percolation thresholds, the latex-based composites are not able to achieve as high a maximum conductivity as the solution-based composites at high filler concentration (Fig. 3). The same excluded volume that serves to lower the percolation threshold also creates porosity within the composite (Fig. 2d), which serves to destroy conductive networks at higher filler concentration. In the next section, the effect of percolation and porosity on the mechanical behavior of these composites is examined.
Storage modulus (E') is one measure of a material's elastic response to deformation; it has nearly the same value as Young's modulus (E) at room temperature (19). Figure 4 shows the results of dynamic mechanical analysis performed on the three composite systems. The lower E' values for the monodisperse latex-based system are due to flaws that result when film formation occurs in a highly ordered manner (20, 21). These flaws are made worse by a lack of complete coalescence at room temperature. Despite having [T.sub.g]'s above 30[degrees]C, the two PVAc latices have minimum film formation temperatures (MFFT) near 15[degrees]C. Above its [T.sub.g], unfilled latex will become a transparent film able to sustain load, indicating a high level of coalescence of the polymer particles. Below the MFFT, most unfilled latices would appear white (or opaque) owing to light scattered by uncoalesced polymer particles and there would be no load bearing capability (22). Between MFFT and [T.sub.g], the latter of which is almost always greater, a coherent film is formed but lack of full coalescence can be observed using an electron microscope (16). The mechanical behavior of these latex systems can ultimately be improved by drying the composite films at temperatures above the [T.sub.g] of the matrix as will be shown in the next section.
All of the systems in Fig. 4 show an initial increase in E' with carbon black concentration. Eventually, as filler loading increases, a given composite system exceeds its critical pigment volume concentration (CPVC) and E' decreases. At the CPVC, a polymer matrix can no longer contact all of the filler particles and voids form at the interstices between them (23). These voids behave as zero modulus filler and are the ultimate source of modulus degradation in these composites (24). The concentration of filler at which voids will form depends on the nature of both polymer and filler. Generally speaking, the smaller particle size of fillers have lower CPVC's due to their greater surface area for a given volume fraction. Using latex as the matrix starting material rather than polymer solution further reduces the CPVC due the inability of the polymer particles to effectively coat the filler during film formation. Fig. 4 illustrates this point nicely in that the CPVC of the polydisperse latex composite system is ar ound 10 vol% carbon black as compared with the PVP solution-based system with a CPVC near 25 vol% (25). Another property that exhibits a sharp transition at the CPVC is break strength, as shown in the next section.
Ultimate Tensile Strength & Elongation at Break
Ultimate tensile strength is a measure of a material's ability to sustain load prior to failure. Figure 5 a shows UTS as a function of carbon black content for the two latex-based composite systems under consideration and a comparable carbon black-filled poly(ethylene-vinyl acetate) prepared from the melt (7). The polydisperse latex has comparable UTS values, up to 10 vol% carbon black, to those exhibited by the melt-based composites at similar and larger carbon black concentrations. At carbon black concentrations greater than 10 vol%, the UTS of the polydisperse latex composites drops sharply due to porosity that is characteristic of materials that exceed their CPVC. This reduction in ultimate tensile strength as a function of carbon black concentration has been observed in other filled polymer systems (26-28) and is due to a weak filler-matrix interface (29), which is another contributing factor for the composites shown here. Despite maintaining its break strength beyond a loading of 20 vol% carbon black, the melt-based composite system's electrical conductivity is only around [10.sup.-15] S/cm at this concentration of filler (7). The polydisperse latex composite has an electrical conductivity greater than 0.1 S/cm at a carbon black loading of 10 vol%, making it a better choice from both a properties and processing standpoint, although melt processing allows easier formation of complex shapes. The UTS of the monodisperse latex system is low at all carbon black concentrations due to its susceptibility to processing defects as discussed previously.
Elongation at break was measured at the same time as ultimate tensile strength; the results for the two latex-based systems are shown in Fig. 5b. These materials are glassy at room temperature and therefore exhibit a fairly low elongation at break. The melt-based carbon black-filled poly(ethylene-vinyl acetate) composites used for comparison in Fig. 3 have elongations of 200% or more (7) due to greater plasticity imparted by the addition of polyethylene. There is no dramatic change in elongation at carbon black loadings greater than 10 vol% as was seen for storage modulus and break strength. The overall trend shows elongation decreasing as the carbon black content increases for a given composite. As polymer is replaced by the far more rigid carbon black, the composite film's ability to stretch becomes steadily diminished. Increasing carbon black content will also diminish the viscoelastic character of a composite film as evidenced by the creep behavior of these systems.
Effects of Latex Coalescence
Composite Microstructure Alteration
During the process of latex coalescence, the latex particles deform and individual polymer molecules gradually interdiffuse between them (22). Coalescence is improved at elevated temperatures because latex particles deform more readily and polymer molecules diffuse at faster rates. The result of improved coalescence is a reduction in residual porosity due to the disappearance of interstitial space and boundaries between latex particles. Figure 6 reveals the change in composite microstructure with temperature for the poly-disperse (Fig. 6a-c) and monodisperse (Fig. 6d-f) latex-based composite systems. The images in Fig. 6 are freeze-fractured crosssections of composite films that had been dried at 20[degrees]C, 60[degrees]C and 110[degrees]C prior to fracture. Large microstructural differences can be seen between films dried at 20[degrees]C and 60[degrees]C. The films dried at 20[degrees]C show a significant level of porosity created by a carbon black aggregation and lack of polymer to fill interstitial space . Both of the latices under consideration have glass transition temperatures around 35[degrees]C and will not deform or fully coalesce at room temperature (~ 20[degrees]C) as a result. Much of the porosity seen in Fig. 6a and d was eliminated when these films were dried at 60[degrees]C (Fig. 6b and e) because of the ability of the latex particles to deform and fill interstitial space. Drying these films at 110[degrees]C (Fig. 6c and [Florin]) does not appear to show significant improvement over drying at 60[degrees]C. Reduced porosity with drying temperature has a notable effect on the percolation threshold (Fig. 7 and Table 2), storage modulus (Fig. 8) and ultimate tensile strength (Fig. 9) for films loaded with greater than 10 vol% carbon black.
Figure 7 shows the effect of increasing drying temperature on the electrical conductivity of the polydisperse (Fig. 7a) and monodisperse (Fig. 7b) composite systems. The data shown in Fig. 7 are summarized in Table 2. Both composite systems show an increasing percolation threshold ([V.sub.c] in Table 2) with drying temperature, but the monodisperse latex system appears to be affected to a much greater extent than its polydisperse counterpart. Maximum electrical conductivity ([[sigma].sub.max], measured at a loading of 20 vol% carbon black) also appears to increase with temperature. The polydisperse latex composites dried at 110[degrees]C had a significant amount of air bubbles at the higher carbon black loadings, which may have contributed to a lower value of [[sigma].sub.max] than that at 60[degrees]C. There does not appear to be any trend concerning the critical percolation exponent (s) or the proportionality constant ([[sigma].sub.0]), which is close to the intrinsic conductivity of carbon black (on the o rder of 10 S/cm) at all temperatures. This increase in [V.sub.c] with temperature is due to the improved coalescence that serves to cutoff conductive pathways that would otherwise be part of the "infinite" network in a more weakly coalesced composite.
Figure 8 shows storage modulus as a function of carbon black concentration, at various drying temperatures, for both the polydisperse (Fig. 8a) and monodisperse (Fig. 8b) PVAc latex-based composites. For the polydisperse system, the most dramatic difference between composites dried at 20[degrees]C and those dried at 60[degrees]C is the greater modulus achieved in the latter at loadings above 10 vol% carbon black. By reducing the porosity in these films, which is the overriding factor for modulus loss above concentrations of 10 vol%, films dried at 60[degrees]C (and 110[degrees]C for that matter) have storage moduli 18%-33% greater than those dried at 20[degrees]C. The data points for composites dried at 110[degrees]C are shown in Fig. 8a but are superimposed on the 60[degrees]C data at carbon black loadings above 10 vol%. Assuming that full PVAc latex coalescence can occur at 60[degrees]C, composites dried at both 60 and 110[degrees]C would be expected to have very similar moduli. This agreement in modulus a ppears to be true for higher carbon black loading ([greater than or equal to]10 vol%); however, the disparity at smaller carbon black concentrations is most likely due to air bubbles and surface abnormalities present in these films.
An improvement in modulus is also observed for the monodisperse latex-based composites when dried at 60[degrees]C (Fig. 8b). The storage modulus appears to be larger at every comparable carbon black concentration, which implies elimination or minimization of the defects attributed to incomplete latex coalescence. A few data points were included for samples dried at 110[degrees]C confirm agreement with 60[degrees]C composites. Based on the trends in Fig. 8 and the microstructural evidence provided in Fig. 6, 60[degrees]C is a good drying temperature for these types of composites. Higher drying temperatures only serve to increase the percolation threshold (see Fig. 7). The lower moduli for the monodisperse system relative to the polydisperse is proposed to be due to statistical fluctuations in porosity that are greater in the monodisperse system. Verification of this is found in the large difference in UTS (see Fig. 9) found when comparing monodisperse and polydisperse composites. A statistically larger collec tion of porosity in the monodisperse latex-based composites acts like a reduced area as well as an increased stress concentration, which lowers both modulus and break strength. The mechanical property disparity remains between these two composite systems at higher drying temperatures due to statistical weak spots in the monodisperse system, which have an effect similar to porosity. Ultimate tensile strength shows a similar improvement with drying temperature.
Ultimate Tensile Strength & Elongation at Break
Figure 9 shows ultimate tensile strength as a function of carbon black content for both polydisperse (Fig. 9a) and monodisperse (Fig. 9b) PVAc latex-based composites. For the polydisperse system, UTS for composites dried at 60[degrees]C is smaller at low carbon black loadings but a crossover occurs around 10 vol% carbon black, at which point the composites dried at 20[degrees]C have much lower strengths. The 10%-15% disparity between these films at small carbon black concentrations may be due to a slight degradation in filler reinforcement that occurred as a result of the surface energy or modulus differences between carbon black and poly(vinyl acetate). Unlike modulus. UTS is a property that depends directly on the local stress concentration and ductility of a material, which creates a situation in which E' increases with filler loading while UTS decreases (27), as shown here. It is improved coalescence, with its reduced porosity, which accounts for the greater UTS observed for the 60[degrees]C composites a t carbon black concentrations greater than 10 vol%. Elongation at break for these composites seems to be slightly smaller for composites dried at 60[degrees]C relative to those dried at 20[degrees]C (see Table 3). This diminishing elongation at break with increasing drying temperature may be due to a weak filler-polymer interface.
For the monodisperse system, it is difficult to adequately compare the composites dried at 20[degrees]C to those dried at 60[degrees]C because of the small number of samples that could be tested from the former set (see Fig. 9 and Table 3). Despite a lack of data for the composites dried at 20[degrees]C, the data show the same general trend noted for the polydisperse system (Fig. 9a). Unlike the polydisperse system, these composites actually show improved ultimate tensile strength with increased carbon black concentration. This improvement is not dramatic but it is evidence of reduced porosity and/or a stronger interface between polymer and filler than was suggested for the polydisperse composites. An improved interface may be the result of a different stabilizing agent that is more compatible with carbon black or a different starting pH of the latex system that would serve to alter the surface charges on the latex particles in a more favorable manner. Whatever the reason, it is apparent that mechanical perf ormance is enhanced overall when composites are dried at higher temperatures, especially at carbon black loadings above 10 vol%. Elongation at break for the monodisperse composite systems seems to follow the same trend as for the polydisperse system, although comparison between the 20[degrees]C and 60[degrees]C data is tenuous because of a lack of sufficient data at 20[degrees]C. All the storage modulus, ultimate tensile strength and elongation at break data for composites dried at 20[degrees]C and 60[degrees]C is summarized in Table 3.
Figure of Merit
Figures of merit can provide a gauge for material performance when there are multiple properties that need to be balanced. These types of parameters are especially common among materials combining transport properties with mechanical or optical properties. Transparent conductive oxides (29), polymer-based light-emitting diodes (30) and piezoelectric sensors (31) all use some type of figure of merit as a property balancing assessment. In this study, two primary mechanical properties and electrical conductivity have been evaluated as a function of carbon black concentration. These three properties have been combined into a figure of merit:
[gamma] = [sigma]E'UTS (2)
where [gamma] is the figure of merit and all other terms have been previously defined. Other combinations of these three properties have been examined as alternate figures of merit (32). Figure 10 shows [gamma] as a function of carbon black concentration for the monodisperse latex-based composite system dried at 60[degrees]C and the polydisperse latex-based composite system dried at 20[degrees]C and 60[degrees]C. The polydisperse system dried at 20[degrees]C shows a peak in [gamma] at 10 vol%, which was expected based upon trends in E' (Fig. 4) and UTS (Fig. 5). Both composite systems dried at 60[degrees]C show a peak in [gamma] at 17 vol% carbon black. None of these composites is able to display a [gamma] value greater than 20, which provides a performance target for the development of new composite systems. From a property balancing perspective, it is best to dry these composites at 60[degrees]C if the goal is large [sigma] with minimal porosity and reasonable mechanical performance. It should also be noted that in systems where the filler is lower cost than the polymer, drying composite films at 60[degrees]C will provide the largest figure of merit with a filler concentration greater than 14 vol%, thus lowering the overall cost of the composite. If filler were the more costly component, it would be better to dry the composite film at room temperature and achieve a good [gamma] value at 10 vol% or less.
Low percolation thresholds, near 2 vol% carbon black, were achieved with poly(viny1 acetate) latices as the composite matrix starting material. The storage modulus and ultimate tensile strength of these materials, examined as a function of carbon black concentration, show a decrease near 10 vol% carbon black that is correlated to the critical pigment volume concentration (CPVC). Porosity develops in these composites when the concentration of filler exceeds the CPVC, which leads to a degradation in mechanical performance. Drying these latex-based composites at temperatures above the [T.sub.g] of the latex was shown to improve their mechanical performance, but increase their percolation thresholds. A figure of merit ([gamma]) was used to assess the optimal balance of electrical conductivity, storage modulus and ultimate tensile strength as a function of carbon black content. The optimal level of carbon black for the polydisperse latex-based composite, dried at 20[degrees]C. was determined to be 10 vol% based up on the peak value of [gamma]. Composite systems dried at 60[degrees]C showed a tendency to have a maximum value of [gamma] at a larger concentration of carbon black due to an apparent shift in the CPVC.
Other variables such as dispersing agents, latex [T.sub.g] and filler particle size could potentially alter the balance of these properties and will be examined in future studies.
The authors would like to thank the Eastman Kodak Company, the University of Minnesota Graduate School, through its Doctoral Dissertation Fellowship, and the Industrial Partnership for Research in Interfacial and Materials Engineering (IPRIME), through its Coating Process Fundamentals Program, for financial support of this research.
(*.) Corresponding author.
(1.) D. M. Bigg and D. E. Stutz, Polym. Compos., 4, 40 (1983).
(2.) J. P. Reboul. in Carbon Black Polymer Composites, Chapter 3. K. Sichel, ed., Marcel Dekker: New York, 1982.
(3.) B. Lundberg and B. Sundqvist, J. Appl. Phys., 60, 1074 (1986).
(4.) N. S. Lewis, M. C. Lonergan, E. J. Severin, B. J. Doleman, and R. H. Grubbs, SPIE, 3079, 660 (1996).
(5.) P. Talik, M. Zabkowska-Waclawek, W. Waclawek, J. Mater. Sci., 27, 6807 (1992).
(6.) D. M. Moffat, J. P. Runt, A. Halliyal, and R. E. Newnham, J. Mater. Sci., 24, 609 (1989).
(7.) X.-S. Yi, G. Wu, and D. Ma, J. Appl. Polym. Sci., 67, 131 (1998).
(8.) J. C. Huang, S. F. Chuang, T. Y. Su, and S. J. Grossman, J. Polym. Eng., 19, 39 (1999).
(9.) J. Z. Liang and R. K. Y. Li, Polym. Compos., 19, 698 (1998).
(10.) C. P. Wong and R. S. Bollompally, J. Appl. Polym. Sci., 74, 3396 (1999).
(11.) J. C. Smith, Polym. Eng. Sci., 16, 394 (1976).
(12.) K. Nakane, T. Yamashita, K. Iwakura, and F. Suzuki, J. Appl. Polym. Sci., 74. 133 (1999).
(13.) A. C. Roulin-Moloney, W. J. Cantwell, and H. H. Kautsch, Polym. Compos., 8, 314(1987).
(14.) J. C. Grunlan, W. W. Gerberich, and L. F. Francis, Mater. Res. Soc. Symp. Proc., 576, 383 (1999).
(15.) Y. Wang and C. Anderson, Macromolecules, 32, 6172 (1999).
(16.) J. C. Grunlan, W. W. Gerberich, and L. F. Francis, J. Appl. Polym. Sci., 80, 692 (2001).
(17.) S. Kirkpatrick, Rev. Mod. Phys., 45, 574 (1973).
(18.) G. Deutscher, A. Kapitulnik, and M. Rappaport, in Percolation Structures and Processes,, Chapter 10, G. Deutscher, R. Zallen, and J. Adler, Eds., Annals of the Israel Physical Society, v. 5. American Institute of Physics; New York, (1983).
(19.) N. Agarwal and R. J. Farris, Polym. Eng. Sci., 40, 376 (2000).
(20.) D. Y. Lee, H. Y. Choi, Y. J. Park, M. C. Khew, C. C. Ho, and J. H. Kim, Langmuir, 15, 8252 (1999).
(21.) F. Dobler, T. Pith, Y. Holl, and M. Lambia, J. Appl. Polym. Sci., 44, 1075 (1992).
(22.) J. L. Keddie, Mater. Sci. Eng. R, 21, 101 (1997).
(23.) W. K. Asbeck and M. VanLoo, Ind. Eng. Chem., 41, 1470 (1949).
(24.) W. J. Culhane, D. T. Smith, and C. P. Chiang, J. Coat Technol., 55, 53 (1983).
(25.) J. C. Grunlan, W. W. Gerberich, and L. F. Francis, J. Mater. Res., 14, 4132 (1999).
(26.) A. Lekatou, S. E. Faidi, S. B. Lyon and R. C. Newman, J. Mater, Res., 11, 1293 (1996).
(27.) Z. Gao and A. H. Tsou, J. Polym. Sci. B: Polym. Phys., 37, 155 (1999).
(28.) J. R. M. D'Almeida and L. H. De Carvalho, J. Mater. Sci., 33, 2215 (1998).
(29.) R. G. Gordon, MRS Bull., 25, 52 (2000).
(30.) P. Broms, J. Birgerson and W. R. Salaneck, Synth. Met., 88, 255 (1997).
(31.) L. Wang, J. Zhu, X. Zou and F. Zhang, Sens. Actuators, B66, 266 (2000).
(32.) J. C. Grunlan, W. W. Gerberich, and L. F. Francis, Mater. Res. Soc. Symp. Proc., 661, in press.
Table 1 Percolation Parameters at Room Temperature. Matrix [[sigma].sub.0](S/cm) 100[V.sub.c] Solution 252 [+ or -] 42.3 14.5 [+ or -] 0.1 Polydisperse latex 13.5 [+ or -] 9.0 2.39 [+ or -] 0.08 Monodisperse latex 15.8 [+ or -] 29.4 2.20 [+ or -] 0.64 Matrix S [R.sup.2] (*) Solution 2.3 [+ or -] 0.6 0.94 Polydisperse latex 1.6 [+ or -] 0.2 0.99 Monodisperse latex 2.2 [+ or -] 0.7 0.98 (*)The [R.sup.2] value is a measure of the quality of fit between the classical percolation power law (Eq 1) and experimental data. The relatively low value for the solution system is due to a small experimental data set. Table 2 Percolation Parameters and Maximum Conductivity at Various Drying Temperatures. Matrix T([degrees]C) [[sigma].sub.0](S/cm) Polydisperse latex 20 13.5 [+ or -] 9.0 60 4.5 [+ or -] 5.0 110 31.0 [+ or -] 43.5 Monodisperse latex 20 15.8 [+ or -] 29.4 60 76.4 [+ or -] 54.2 110 12.8 [+ or -] 8.91 Matrix 100[V.sup.a.sub.c] s [R.sup.2] Polydisperse latex 2.39 [+ or -] 0.08 1.6 [+ or -] 0.2 0.99 2.76 [+ or -] 0.47 1.4 [+ or -] 0.4 0.99 2.85 [+ or -] 2.00 2.2 [+ or -] 0.8 0.99 Monodisperse latex 2.20 [+ or -] 0.64 2.2 [+ or -] 0.7 0.98 4.37 [+ or -] 0.74 2.6 [+ or -] 0.4 0.99 5.60 [+ or -] 0.75 1.7 [+ or -] 0.4 1.00 Matrix [[sigma].sub.max] (S/cm) Polydisperse latex 0.798 1.270 0.594 Monodisperse latex 0.338 0.460 0.498 Table 3 Mechanical Property Data for Latex-Based Composites. 20[degrees]C E' VTS E' Matrix Vol% CB (GPa) (MPa) %[epsilon] (GPa) Polydiesperse latex 0 3.14 28.3 1.96 3.33 2 3.25 22.7 3.42 3.30 3 3.26 4 3.25 25.0 1.76 5 3.36 22.2 1.43 3.40 6 3.34 7 3.43 21.6 1.54 8 3.46 3.28 10 3.07 18.5 2.27 2.97 12 14 2.42 6.3 1.31 2.89 17 2.9 1.20 20 1.73 3.3 0.52 2.49 Monodisperse latex 0 1.26 4.3 1.63 1.49 2 3 2.06 11.2 2.21 2.36 5 2.62 7 2.66 10 2.33 8.0 0.34 2.91 12 2.85 14 2.70 60[degrees]C VTS Matrix (MPa) %[epsilon] Polydiesperse latex 25.2 1.77 21.5 1.08 18.5 1.31 19.2 1.03 18.6 1.55 18.6 1.24 15.0 1.20 17.3 1.32 11.8 0.86 Monodisperse latex 8.7 1.10 9.6 0.77 9.7 0.73 13.4 0.84 12.9 1.05 12.9 1.00
|Printer friendly Cite/link Email Feedback|
|Author:||Grunlan, Jaime C.; Gerberich, William W.; Francis, Lorraine F.|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 2001|
|Previous Article:||Effects of [CO.sub.2] on Crystallization Kinetics of Polypropylene.|
|Next Article:||Effect of Interlayer Structure, Matrix Viscosity and Composition of a Functionalized Polymer on the Phase Structure of Polypropylene-Montmorillonite...|