Filler exfoliation and dispersion in polypropylene/as-received graphite nanocomposites via cryogenic milling.
Polymer nanocomposites are an emerging class of specialty materials that has gained significant distinction in recent years [1-59], In contrast to conventional macro- and micro-composites, the filler particles of the nanocomposites have at least one dimension on the nanometer length scale, and offer a much higher surface area to volume ratio, as well as a potential molecular interaction with the matrix polymer (1-3). As a result, a nanocomposite can achieve significantly enhanced physical properties at low filler loadings, typically 1-10 wt%. Because less filler is required, there is minimal sacrifice to the original performance of the polymer matrix, and the overall cost of nanocomposites can remain low. Within a variety of possible nanofiller choices, single- and multi-walled carbon nanotubes (4-18) have been widely studied because of their high tensile strength and electrical conductivity, whereas nanoclays (19-36), due to the layered morphology, are extensively applied in composite materials where enhanced mechanical, thermal, and gas barrier properties are desired.
More recently, grapnne, in tne rorm or inaiviauat graphene sheets or graphite nanoplatelets, has been pursued as a nanofiller of choice due to similar, or sometimes even more significant, property enhancements that the filler can provide when incorporated into a polymer matrix (15), (37-59). Often more cost effective than other nanofillers, graphite is chemically similar to carbon nanotubes and structurally analogous to clay; the repeating graphene sheet structure of graphite resembles that of the layered silicates in montmorillonite whereas the [sp.sup.2] -hybridized six-carbon ring network structure within each individual graphene layer is identical to that of carbon nanotubes. A graphene layer is known to possess excellent thermal conductivity on the order of 5000 W/(m-K) (60), mechanical stiffness of ~1 TPa (61), and electron mobility of above 15,000 [cm.sup.2] /(V*s) (62). Many of the previous reports have focused on the processing aspect of the graphite- and graphene-based nanocomposites, as achieving optimum filler structure and composite morphology has been practically very challenging. Graphite, in its natural state, exists as flakes of microns in size, which are composed of repeated layers of graphene sheets 0only 3.35 A apart (61). Complete exfoliation transforms the filler into individual graphene layers of -----1 nm thick whereas partial exfoliation leads to thin nanoplatelets ranging in thickness from 500 nm down to several nm.
Different conventional processing methods, such as solution-mixing (40), (42), (44-46), (48), (49), (51-54), in situ polymerization (37-39), (41), and melt extrusion (15), (43), (50), (55-57), have been applied to disperse graphite and graphene fillers in a polymer matrix, with varying levels of success in terms of property improvements. Almost all the cases, however, involve fillers that have been treated or modified as expanded/expandable graphite, exfoliated graphite/graphite nanoplatelets, graphite/graphene oxides, and functionalized graphite/graphene. Solid-state shear pulverization (SSSP) (63-65) is a different, robust technique that can achieve in situ dispersion of unmodified, as-received nanofillers in a polymer matrix (11), (12), (58), (66); a polypropylene/as-received graphite nanocomposite fabricated via this method was reported to have a number of improved physical properties, including a 100% increase in Young's modulus (58). The major reason for stable nanocomposite morphology and enhanced properties is the low-temperature operation, where the mechanical shear and compression of the nanofiller with the hard and brittle polymer pellets lead to mechanical exfoliation and dispersion of the nanotiller, without being influenced by the thermodynamic and transport phenomena applicable in the solution or melt state. In contrast, batch-scale solid-state processing at or below ambient temperature has also been studied for the fabrication of polymer nanocomposites, with several variations of the technique differentiated by the milling medium and temperature; these methods are termed as solid-state shear compounding/milling and high-energy/attrition/impact/cryogenic ball milling (17), (18), (30), (31), (33-36), (59), (67-69). This article reports a systematic processing-structure-property characterization on polymer/as-received graphite nanocomposites fabricated via batch-scale cryogenic impact milling (cryomilling). Materials prepared in a simple milling setup with a short, facile procedure show appreciable levels of property improvements due to exfoliation of graphite into nanoplatelets. The effects of cryomilling and subsequent process parameters are explored and compared with conventional melt mixing, and the effect of graphite loading (1-10 wt%) is investigated to determine the percolation threshold and optimum filler content.
Polypropylene (PP) from Total Petrochemicals, PP3276 (density = 0.905 g/ [cm.sup.3], MFI = 2.0 g/10 min at 230[degrees]C/ 2.2 kg), was the model polymer matrix in the study. As-received graphite (ARG) from Asbury Carbon, Grade 4827 (average flake diameter = 2, um, surface area =113 [m.sup.2]/g), was employed without any pretreatment. No solvents, compatibilizers, or additives were used during the processing of nanocomposites.
Preparation of Nanocomposites
Solid-state fabrication of PP/ARG nanocomposites was carried out using a Spex SamplePrep 6870 Freezer/Mill, a cryogenic impact mill (cryomill). A 12 g batch of polymer-graphite mixture of known graphite loading (1, 3, 5, 8, and 10 wt%, indicated hereafter as %) was added to the 50-mL polycarbonate sample vial with stainless steel end plugs and magnetized impacting bar, 19 mm in diameter and 70 mm in length. The sample vial was loaded into the mill chamber and completely submerged within a liquid nitrogen bath. Following a 15 min precooling period, samples were processed in 8 min cycles, each consisting of a 4 min milling period followed by a 4 min sample recooling time. Controlled oscillation of the impacting bar between the end plugs at the rate of 10 Hz creates the milling and colliding action that mechanically fragments and intimately mixes the polymer pellets and the graphite particles; the output of the cryomilling process was a homogeneous, uniformly colored powder. The number of cycles was varied at 5, 15, and 50 cycles. Cryo-milled samples are identified by CM designation followed by the number of cycles. Table 1 shows the cycle to grinding time conversion. A subset of the cryomill-pulver-ized nanocomposite powder samples was subsequently melt-mixed for evaluation of the postcryomill melt process (specimens identified with the suffix - MM). Cryo-milled PP/ARG composite powder was added to an Atlas Laboratory cup-and-rotor mixer maintained at 200[degrees]C, in which three 4 mm steel mixing balls were inserted to promote intimate and chaotic mixing ; the mixing was carried out at 170 rpm for 5 min. Preparation of a melt-extruded PP/ARG sample (denoted EXT), for comparison, was performed using a Davis-Standard Killion KLB075 single-screw extruder (D = 19 mm and LID = 24), operating at the speed of 30 rpm and the average barrel temperature of 200[degrees]C. Test specimens were compression-molded in a Carver Press held at 200[degrees]C, and subsequently quenched to room temperature.
TABLE 1. Cryomilling and melt mixing processing parameters. Method Cryomilling lime Subsequent melt Total processing (min) mixing time (min) time (min) CM 5 20 0 40 CM15 60 o 120 CM50 200 0 400 CM5-MM 20 5 45 CM15-MM 60 5 125 CM50-MM 200 5 405
Direct imaging of the composite morphology by scanning electron microscopy (SEM) was performed on a compression-molded specimen, prepared in the following fashion: cryofractured surface was plasma-etched for selective removal of the surface PP matrix using Harrick PDC-32G plasma etcher at 13.6 MHz, and subsequently sputter-coated with gold using the Denton Vacuum Desk IV. JEOL JSM-6390LV SEM was operated at an accelerating voltage of 15 kV and a sample-to-detector distance of 10 mm. For X-ray diffraction (XRD) measurements, compression-molded specimens were examined in a PANalytical X'Pert Pro Multi Purpose Diffractometer system. CuK [ALPHA] monochromatic rays ([lambda] = 1.542 A) at 45 kV and 40 mA were used, and the detector was set at 0.04[degrees] acquisition steps. The linear scale of the XRD spectra was normalized by known PP crystallinity (as determined by calorimetry, described below). Thermogravimet-ric analyzer (TGA) was used to investigate the thermal degradation characteristics of the nanocomposites. TA Instruments Q600 SDT was programmed to heat the sample and reference pans at the constant rate of 10[degrees]C/min in a nitrogen environment. Further thermal characterization was conducted using a TA Instruments Q1000 differential scanning calorimeter (DSC). Isothermal crystallization run was conducted by heating the specimen sealed in a her-mitic pan to 200[degrees]C at 10cC/min, and subsequently quenching the specimen to 140[degrees]C and holding at the temperature for 2 h. Heat flow was recorded as polymer crystals developed over time. Tensile test coupons were prepared from 0.5-mm-thick compression molded sheets, using a Dewes Gumbs 1.5T DGD Expulsion Press with an American Society for Testing and Materials (ASTM) standard D1708 die. Uniaxial crosshead speed of 1.27 cm/ min (strain rate of 0.01/s) was applied in Tinius Olsen H5K-S, operating in conjunction with in-house data acquisition and analysis software. Electrical conductivity was measured by way of impedance spectroscopy using a Gamry Series 6750 Impedance Spectrometer and Gamry Framework software. Resistance was measured across the width on melt-molded specimens with dimensions 3.18 mm (h) X 12.7 mm (w) x 31.8 mm (l), with an applied voltage of 100 mV over a frequency range of 300,000-0.01 Hz, A Mocon OX-TRAN Model 2/21 oxygen permeation analyzer was used for barrier characterization; measurements were conducted on 50 [cm.sup.2] areas of 0.5-mm-thick compression-molded sheets, at room temperature and zero relative humidity.
[FIGURE 1 OMITTED]
RESULTS AND DISCUSSION
SEM micrographs in Fig. 1 shows the treated surfaces of cryomilled PP/1%ARG samples, as well as those of extruded PP/1%ARG and neat PP as references. In cryomilled nanocomposite samples, the graphite nanoplatelets are clearly seen on the surface as randomly oriented layered particles of varying dimensions. The SEM image of the extruded (EXT) PP/1%ARG sample closely resembles that of the neat PP in Fig. 1 with very sparsely embedded graphite particles 1-3 [micro] m in size. This considerable difference in filler morphology between cryomilled and extruded samples of identical filler loading suggests a lack of exfoliation of the graphite nanoplatelets in the latter. Figure 2 shows XRD spectra of neat PP and PP/1%ARG that were extruded and cryomilled at different cycles, in the region around the characteristic peak (2 [theta] ~ 26.6[degrees]; d ~ 3.35 A) of graphite. As the height of a nanofiller XRD peak is a qualitative manifestation of stacked, repeated layers present in the sample, the XRD spectrum of extruded PP/ARG sample suggests that the graphite particles remained in unexfoliated slacks upon extrusion. Thus, extrusion simply mixed, but did not exfoliate or disperse, the graphite fillers throughout the polymer matrix.
SEM image of the PP/1%ARG CM5 sample in Fig. 1 exhibits significantly delaminated graphite particles in many regions of the specimen, on the order of microns in lateral dimensions and tens to hundreds of nanometers in thickness. These graphite nanoplatelets are far from individual graphene layers reported in the literature [45, 60, 62], but still serve as effective nanofillers in PP. As the cryomilling time increases and the material is subjected to additional mechanical milling, the graphite nanoplatelets are fragmented into particles with smaller lateral dimensions, as seen in the comparison of the PP/1%ARG CM50 sample to the CM5 sample in Fig. 1. It is noteworthy that the filler content in these PP/1%ARG cryomilled specimens converts to only 0.4 vol%, and the considerable apparent volume that the nanoplatelets occupy in these SEM images qualitatively confirm the significant levels of exfoliation that has occurred in the cryomilled samples. Normalized XRD spectra in Fig. 2 confirm that cryomilling indeed leads to delamination of graphite particles along the thickness of the nanoplatelets. The fact that the height of the characteristic graphite peak in the CM5 sample is slightly lower than that of the EXT sample, along with the trend of progressively decreasing peak height with CM 15 and CM50, suggests that cryomilling leads to in situ partial exfoliation of the graphite layers, with the exfoliation level increasing with longer cryomilling cycles or time. It is rather intriguing that 50 cycles of cryomilling (corresponding to ~3.3 h of processing) can lead to a near absence of the graphite peak in XRD.
[FIGURE 2 OMITTED]
Some of the cryomilled samples were subsequently melt-mixed for an investigation of whether the secondary, postcryomill melt processing has any effect on the original graphite morphology. Melt mixing of powder samples from solid-state fabrication is relevant to potential industrial applications, where processing and molding usually takes place in the melt. Masuda and Torkelson performed a similar study of PP/carbon nanotube nanocomposites by employing solid-state shear pulverization followed by melt mixing, and reported that the two-step process promoted further nanotube dispersion compared to the solid-state shear pulverization alone (11). In this study, the layered nature of the filler resulted in two phenomena. The SEM micrograph of PP/1%ARG processed by 50-cycle cryomilling plus melt mixing (CM50-MM) in Fig. 1 illustrates similar-sized fragmented graphite nanoplatelet particles, but significantly more distributed in the polymer matrix when compared with that of CM50 alone. This type of further dispersion is also seen in the CM5 and CM 15 cases. However, when the XRD spectra of CM5-MM, CM15-MM, and CM50-MM (dashed curves in Fig. 2) are compared with those of corresponding CM5, CM 15, and CM50 (solid curves), there is a systematic increase in the graphite peak height, which is an evidence for nanoplatelet reagglomeration (restacking). Therefore, a postcryomill melt mixing step is detrimental to the exfoliation (separation) state of the graphite nanoplalelets, but advantageous in the dispersion (distribution) of the nano-filler in the PP matrix. In addition, peak shifts to lower 2 [theta] values are observed in all CM-MM nanocomposites. which suggest that the postcryomill melt mixing leads to an intercalation of PP chains in the interlayer spacing of ARG. However, because the extent of the XRD peak shift is small (~0.2 [degrees] ), we refrain from attributing this shift to major morphological phenomena.
[FIGURE 3 OMITTED]
Figure 3 shows the graphite peak region of the XRD spectra for the PP/ARG series of varying filler loadings, processed via the CM5 method. A monotonic trend of nanocomposites with higher graphite content displaying more intense graphite peaks simply corresponds to the increased presence of graphite particles, hence graphene layers, in the sample. A qualitative evaluation that the peak intensities are similar at higher graphite content samples, specifically 5%, 8% (not shown), and 10%, suggests that the extent of exfoliation of graphite nanoplate-lets are significantly affected by the amount of filler present, and in turn the amount of graphite that the cryomill-ing medium is exposed to. There is a saturation limit above which cryomilling cannot further exfoliate the graphite nanoplatelets. Figure 4 compares the SEM images of four PP/ARG samples of different graphite content. The overall size of the graphite nanoplatelets, both in the lateral and layer thickness dimensions, remains progressively larger with higher graphite loading. In these overloaded nanocomposites, the filler particles have less free space in the polymer matrix to exfoliate and fragment, and thus most likely remain in its stacked state. Such ineffective nanocomposites are expected to exhibit physical properties that are more similar to extruded analogs of corresponding filler content.
Filler Exfoliation/Dispersion and Polymer Crystallization
Figure 5 illustrates the rate of polymer crystallite development in neat PP and PP/1%ARG nanocomposites, determined from the DSC isothermal crystallization analysis at 140 [degrees] C. The curves were constructed based on the total polymer crystallinity that each sample can attain, and plotted against time that the materials took to reach percentages of the total crystallinity. Crystallization half times ([tau]1/2), a conventional quantitative measure of the crystallization rate, are tabulated in Table 2. According to crystallization kinetics theory, well-exfoliated filler particles act as heterogeneous nucleation sites and lead to a higher polymer crystallization rate of the nanocomposite. All of the nanocomposites under investigation indeed benefit from faster isothermal crystallization due to the presence of graphite filler particles in PP, but cryomilled samples exhibit especially higher rates of crystallization compared to the EXT equivalent. Longer cryomilling time results in lower [tau]1/2 values and thus faster crystallization rate, because a higher number of graphite nanoplatelet particles are created from exfoliation and lateral fragmentation. The inferred degrees of exfoliation are consistent with XRD analysis, which demonstrated that CM50 promoted the most nanofiller exfoliation, followed by CM15 and CM5. It is interesting to note that postcryomill melt mixing further enhances polymer crystallization rates; CM5-MM, CM15-MM, and CM50-MM all exhibited lower [tau]1/2 than the corresponding CM samples. This phenomenon was not consistent with what the XRD data suggested, where postcryomill melt mixing lead to restacking of the filler layers. From the earlier discussion of enhanced filler dispersion observed in postcryomill melt mixing, there is a strong indication that the dispersion also plays a separate, significant role in raising the polymer crystallization rate in nanocomposites. A closer inspection of the Avrami sigmoidal curves in Fig. 5 reveals that cryomilled-plus-melt-mixed samples have systematically higher slopes than simple cryomilled analogues, during the intermediate (crystal growth) stage. Table 2 quantifies this observation and lists the isothermal crystallization start time, defined as the time at which 5% of total crystallinity is completed, and the slope at crystallization half time for each sample. For a given CM cycle set, subsequent melt mixing does not shorten the crystallization start time (nucleation), but it does raise the slope, i.e. the rate at which the crystals undergo growth. We therefore interpret that the postcryomill melt mixing process disperses the numerous heterogeneous nucleation sites, which had already been created via exfoliation, throughout the sample and in turn generates more extensive spaces for polymer crystallization. While further study regarding the mechanism behind separate nucleation and growth steps in exfoliated nanocomposites is warranted, this analysis ultimately validates that ARG dispersion in cryomilled materials can be further advanced by following the solid-state process with melt mixing.
TABLE 2. Properties of PP/1%ARG nanocomposites. Sample Isothermal Isothermal Isothermal Onset crystal, crystal, crystal, at140 degradation at140 at140 [degrees] C, temperature, [degrees] [degrees] C, slope at half at 5 wt% loss C, half start [time.sup.c] ([degrees] [time [time.sup.b] (% C) .sup.a] min) crystals/min) (min) Neat PP 54.4 23.2 0.02 420 EXT 11.6 6.0 0.11 415 CM5 93 2.8 0.08 420 CM 15 7.9 2.5 0.10 428 CM50 5.2 1.6 0.15 426 CM5-MM 7.0 3.0 0.15 428 CM15-MM 6.3 2.6 0.16 431 CM50-MM 4.4 1.6 0.22 429 Sample Young's Yield Elongation Electrical modulus(MPa) stress(MPa) at break conductivity(S/cm) Neat PP 900 [+ or -] 27 [+ or -] 8.0 [+ or No 50 1 -] 0.2 EXT 1070 [+ or -] 32 [+ or -] 0.4 [+ or No 40 1 -] 0.2 CM5 1100 [+ or -] 30 [+ or -] 7.1 [+ or 1.0 * [10.sup.-6] 80 0 -] 0.3 CM 15 1310 [+ or -] 33 [+ or -] 6.1 [+ or 7.0 * [10.sup.-5] 70 1 -] 0.7 CM50 1300 [+ or -] 35 [+ or -] 4.4 [+ or 1.4 * [10.sup.-4] 140 3 -] 0.2 CM5-MM 1230 [+ or -] 32 [+ or -] 6.3 [+ or No 50 1 -] 0.3 CM15-MM 1440 [+ or -] 33 [+ or -] 6.4 [+ or No 60 1 -] 0.1 CM50-MM 1380 [+ or -] 35 [+ or -] 6.3 [+ or No 120 1 -] 0.8 Sample [o.sup.2] permeability(cc mil/([m.sup.2] day)) Neat PP 3090 EXT 3180 CM5 2220 CM 15 2540 CM50 2600 CM5-MM 2710 CM15-MM 2660 CM50-MM 2700 a Time at which 50% of the total crystallinity is achieved, b Time at which 5% of the total crystallinity is achieved. c Slope of the "% of total crystallinity vs. time" curve at 50% of the total crystallinity.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
Results from the isothermal DSC experiment performed on PP/ARG series of varying filler contents are shown in Table 3. Increase in PP crystallization rate is expected in samples containing higher amounts of graphite, simply from the heterogeneous nucleation site argument; indeed, a monotonic increase in [tau] 1/2 values is observed as graphite loading increased. The increase in polymer crystallization rate for CM5 samples of higher filler loadings is on the same order of magnitude as that caused by increasing the cryomilling cycles in the PP/ 1 %ARG sample.
TABLE 3. Properties of PP/ARG nanocomposites of varying filler content, ail processed as CM5. Sample isothermal Onset Young's Yield Elongation crystallization degradation at 140 [degrees] temperature, modulus stress at at C, half 5 wt% loss (MPa) (MPa) break [time.sup.a] (min) 54.4 ([degrees] C) Neat 54.4 420 900 [+ 27 [+ 8.0 [+ or PP or -] or -] -] 0.2 50 1 1 % 9.3 420 1100 [+ 30 [+ 7.1 [+ or ARG or -] or -] -] 0.3 80 0 3% ARG 7.8 425 1280 [+ 33 [+ 4.4 [+ or or -] or -] -] 0.2 110 1 5% ARG 6.7 431 1350 [+ 32 [+ 0.3 [+ or or -] or -] -] 0.2 70 1 8% ARG 6.4 432 1350 [+ 32 [+ 0.1 [+ or or -] or -] -] 0.0 120 1 10% 5.7 428 1330 [+ 30 [+ 0.1 [+ or ARG or -] or -] -] 0.0 160 3 Sample [o.sup.2] permeability (cc mil/ [m.sup.2] day))y Neat 3090 PP 1 % 2220 ARG 3% ARG 2150 5% ARG 2000 8% ARG 1940 10% 1840 ARG a Time at which 50% of the total crystallinity is achieved.
Thermal, Mechanical, Electrical, and Barrier Properties
Table 2 contains thermogravimetric analysis data of neat PP and PP/1%ARG nanocomposites processed by different methods. The reported onset thermal degradation temperature ([[TAU].sub.deg]) is defined here as temperature at 5 wt% of total sample loss, which we interpret as a general measure of thermal stability. Pure ARG is extremely thermally stable; negligible weight loss was observed over the temperature range of measurement (-----1 wt% at 550 [degrees] C). Enhancement in thermal stability in PP/ARG nanocomposites arises from filler particles acting as thermal and transport barriers in the polymer matrix; previously studied PP/graphite systems have reported increases in rdeg as large as 35 [degrees] C . The EXT sample exhibits a lower thermal stability than the neat polymer, possibly due to the slight polymer chain degradation that had occurred in the melt extrusion process. All of the CM and CM-MM specimens show at least equal, otherwise appreciably enhanced thermal stability compared with the neat polymer--as much as 11[degrees]C increase in Tdeg. As discussed earlier, longer cryomilling time results in a higher level of exfoliation whereas postcryomill melt mixing contributes to an enhanced dispersion of the exfoliated graphite nanoplatelets. Both these morphological changes evidently lead to higher thermal stability, as shown in Table 2. The concept of graphite filler acting as a thermal barrier is also consistent with the trend shown in Table 3, where for a consistent cryomill processing method of CM5, higher loading of ARG in PP contributes to higher thermal stability, up to 8 wt%. A combination of high ARG loading (around 8 wt%) and higher cryomilling time with subsequent melt mixing (15CM-MM) is expected to yield samples with highest thermal stability.
Tensile test data for FF/AKU samples of diflerent processing methods and varying filler contents are presented in Tables 2 and 3, respectively. Young's modulus and yield stress data show that all graphite nanocomposite samples benefit from modest increases in stiffness and strength, which arise from a filler effect of high-modulus and high-strength material in a less stiff and strong polymer matrix, based on a simple composite theory. This is not necessarily an indication of well-exfoliated nanocom-posites, as noted in the 18% modulus increase of the extruded sample. However, elongation at break is at least 10-fold lower for the extruded sample compared with any cryomiil-processed samples. Often identified as the major disadvantage of conventional composites, this significant increase in brittleness arises from micro- and macro-scale filler particles acting as physical defect themselves, or introducing similar-sized voids and imperfections surrounding them. In contrast, cryomilled and cryomilled-plus-melt-mixed samples show appreciable levels (up to ~45% for CM and -60% for CM-MM series in Young's modulus compared to neat PP) of increase in stiffness and strength with minimal sacrifice (as little as 11% reduction) in elongation at break. The level of mechanical enhancement depends on several factors. First, longer cryomilling time, as indicated by the change in stiffness and strength among CM5, CM15, and CM50 in Table 2, contributes to higher level of exfoliation of graphite nano-platelets, although excessive cryomilling evidently lead to brittleness, often an undesired outcome. Second, post-cryomill melt mixing, as exemplified by the values of CM15-MM sample in Table 2, enhances the dispersion of the partially exfoliated fillers and, therefore, raises the stiffness and strength whereas only lightly influencing the elongation at break. Finally, Table 3 shows that an increase in filler content corresponds to an increase in nanocomposite stiffness and strength in general accordance with the composite theory, up to a certain point; in the case of CM5, 3 wt% is likely an optimum ARG content, which leads to 42% increase in stiffness with 45% reduction in elongation at break; samples with higher filler content have minimal additional increase in stiffness and become impractically brittle.
Gas barrier properties were characterized by way of oxygen permeation. Oxygen permeability values for PP/ 1%ARG nanocomposites in Table 2 indicates graphite fillers generally act as a physical barrier to the oxygen molecules and provide a more tortuous pathway for the gas within the PP matrix, though the EXT sample is an exception. A slight increase in oxygen permeability in the extruded sample may be attributed to voids and other defects surrounding the micro- and macro-scale fillers in the specimen. As discussed earlier, CM5 process leads to exfoliation of graphite nanoplatelets, which reduces oxygen transport within the matrix (28% reduction compared with neat PP). Surprisingly, further cryomilling beyond CM5 results in an increase in oxygen permeability. This negative effect on the barrier property is likely because excessive cryomilling significantly cuts down large lateral dimensions (and thus high aspect ratio) of graphite nanoplatelets. Exfoliated but simultaneously fragmented graphite nanoplatelet particles are not as effective in providing the tortuous path for gases through the matrix. Similarly, it is rather counterintuitive to observe that postcryomill melt-mixed samples have higher oxygen permeability than corresponding cryomill-only analogs. These results suggest that oxygen barrier behavior is more effective when the exfoliated graphite nanoplatelets are less dispersed but instead exist in clumps. Table 3 indicates that the [0.sub.2] permeability decreases monotonically with increasing filler content, which is consistent with the concept of graphite nanofillers acting as physical hindrance to oxygen transport. Therefore, desired filler morphology for barrier functionality is very different from, or even opposite of, that for mechanical and thermal property enhancements.
Because the graphite filler is based on sp2-hybridized carbon structure, the resulting nanocomposites have the potential to conduct electricity. Electrical conductivity in the PP/ARG system, where the PP matrix is nonconduc-tive., relies on the physical connectivity of the filler particles across the sample, and therefore exfoliation and dispersion into a connective network becomes an important factor, especially in specimens of low filler loading. Electrical conductivity data of PP/1%ARG nanocomposites, measured via impedance spectroscopy, are presented in Table 2. The extruded sample did not record measurable conductivity, which is because the graphite fillers are not exfoliated to provide a connective path. In contrast, PP/ 1%ARG samples processed via different extents of cryo-milling exhibit electrical conductivity values on the order of [10.sup-6]--[10.sup.-4] S/cm, graphite nanoplatelets are exfoliated to sufficient degrees such that, even with a random orientation in a compression-molded specimen, a continuous filler network existed. Table 2 shows that the conductive network is more prominent following higher cycle cryo-milling, despite the SEM results showing significant lateral scission of high-aspect ratio layers. Interestingly, subsequent melt mixing of cryomilled PP/1%ARG samples eradicates the graphite network and causes loss of electrical conductivity. This phenomenon can be attributed to the partial reagglomeration and excessive dispersion of the limited amount of graphite nanoplatelets promoted by the postcryomill melt mixing process, as corroborated by the XRD and SEM results.
Electrical conductivity values of FP/ARG nanocompo-sites processed via CM5 are plotted as a function of filler loading in Fig. 6. A general rise in electrical conductivity with increasing filler content results from a larger amount of high-aspect ratio graphite nanoplatelets present in the matrix, leading to a more robust connective pathway through the sample. A significant rise up to 3 wt% filler content and subsequent plateau of conductivity with additional filler loading seen in Fig. 6 is a manifestation that percolation threshold is achieved in the series of samples between 1 and 3 wt% (0.4 and 1.2 vol%) graphite content. This threshold value is in close proximity to that found in PP/ARG nanocomposites processed via solid-state shear pulverization , reported to be the lowest percolation threshold achieved in polymer nanocomposites with unmodified graphite. The highest level of electrical conductivity achieved in this study is on the order of [10.suo.-3] S/cm, which is relatively modest because the ARG used has a reported neat-state electrical conductivity of only 6-7 S/cm . Although a more highly conductive graphite filler can be used to improve the nanocomposite conductivity, modest levels of electrical conductivity similar to what is observed here are often sufficient for such basic applications as electrostatic dissipation.
The solid-state cryomilling of PP with 1-10 wt% of ARG leads to varying and considerable levels of graphite nano-platelet exfoliation. Higher cycles of cryomilling raise the extent of exfoliation, while also fragmenting further the graphite nanoplatelets in the lateral dimensions. Melt mixing was explored as a postcryomill processing step, and was shown to cause nanofiller reagglomeration to an extent, but at the same time further enhance the dispersion of the nanofiller.
Physical properties of the cryomilled nanocomposites exhibited notable enhancements, in different degrees depending on the experimental parameters. Increasing the cryomilling cycles, mainly from 5 to 15, leads to improvements in such properties as thermal degradation temperature, Young's modulus and yield strength and electrical conductivity, while reducing the oxygen barrier capacity. Applying a postcryomill melt mixing enhances thermal stability and Young's modulus, but reduces oxygen barrier properties and eliminates electrical conductivity due to excessive dispersion. An increase in filler content generally improves properties like thermal stability, Young's modulus, and oxygen permeation, but there is an optimum filler content for most properties at 3-5 wt%.
The cryomilling technique was demonstrated to be a simple, robust small-scale fabrication technique for PP/ ARG nanocomposites, with a range of achievable properties to tailor to desired performance and applications. As solid-state processes become a viable alternative for fabricating polymer nanocomposites, studies that relates batch-scale cryomilling to continuous, industrially scalable solid-state shear pulverization is warranted. Such studies are currently underway.
Test materials were graciously provided by David L. Turner at Total Petrochemicals and Albert V. Tama-shausky at Asbury Carbons. Sample characterization was facilitated by Michael D. Gross, Erin L. Jablonski, Timothy M. Raymond, Brad C. Jordan, and Diane S. Hall at Bucknell University. PJH acknowledges the support from Bucknell Graduate Research Fellowship.
(1.) P. Rittigstein and J.M. Torkelson,./. Polym. Sci. B: Polym. Phys., 44, 2935 (2006).
(2.) D. Ciprari, K. Jacob, and R. Tannenbaum, Macroniolecules, 39, 6565 (2006).
(3.) L.S. Schadler, S.K. Kumar, B.C. Benicewicz, S.L. Lewis, and S.E. Harton, MRS Bull., 32, 335 (2007).
(4.) R. Haggenmueller, H.H. Gommans, A.G. Rinzler, J.E. Fischer, and K.I. Winey, Chem, Phys. Lett., 330, 219 (2000).
(5.) P.M. Ajayan, L.S. Schadler, C. Giannaris, and A. Rubio, Adv. Mater., 12, 750 (2000).
(6.) P. Potschke, T.D. Fornes, and D.R. Paul, Polymer, 43, 3247 (2002).
(7) O. Breuer and U. Sundararaj, Polym. Compos., 25, 630 (2004).
(8.) T. McNally, P. Potschke, P. Halley, M. Murphy, D. Martin, S.E.J. Bell, G.P. Brennan, D. Bein, P. Lemoine, and J.P. Quinn, Polymer, 46, 8222 (2005).
(9.) T. Ramanathan, H. Liu, and L.C. Brinson, J. Polym. Sci. B: Polym. Phys., 43, 2269 (2005).
(10.) M. Moniruzzaman and K.I. Winey, Macromoiecules, 39, 5194 (2006).
(11.) J. Masuda and J.M. Torkelson, Macromoiecules, 41, 5974 (2008).
(12.) M. Mu, A.M. Walker, J.M. Torkelson, and K.I. Winey, Polymer, 5, 1332 (2008).
(13.) S. Pujari, T. Ramanathan, K. Kasimatis, J. Masuda, R. Andrews, J.M. Torkelson, L.C. Brinson, and W.R. Bur-ghardt,./. Polym. Sci. B: Polym. Phys., 47, 1426 (2009).
(14.) A.I. Isayev, R. Kumar, and T.M. Lewis, Polymer, 50, 250 (2009).
(15.) P. Potschke, M. Abdel-Goad, S. Pegei, D. Jehnichen, J.E. Mark, D. Zhou, and G. Heinrich,./. Macromol. Sci. A, 47, 12 (2010).
(16.) B.P. Grady, Macromol. Rapid. Comm., 31, 247 (2010).
(17.) G. Gorrasi, M. Sarno, A. Di Bartolomeo, D. Sannino, P. Ciambeili, and V. Viuoria,./. Polym. Sci. B: Polym. Phys., 45, 597 (2007).
(18) G. Terife and K.A. Narh, Soc. Plast. Eng. ANTEC, 67, 349 (2009).
(19.) A. Usuki, Y. Kojima, M. Kawasumi, A. Okada, Y. Fukush-ima, T. Kurauchi, and O. Kamigaito, J. Mater. Res., 8, 1179 (1993).
(20.) E.P. Giannelis, Adv. Mater., 8, 29 (1996).
(21.) P.C. LeBaron, Z. Wang, and T.J. Pinnavaia, Appl. Clay. Sci., 15, 11 (1999).
(22.) M. Alexandre and P. Dubois, Mater. Sci. Em, R, 28, 1 (2000).
(23.) A. Okada and A. Usuki, Macromol. Mater. Eng., 291, 1449 (2006).
(24.) D.R. Paul and L.M. Robeson, Polymer, 49, 3187 (2008).
(25.) E. Manias, A. Touny, L. Wu, K. Strawhecker, B. Lu, and T. Chung, Chem. Mater., 13, 3516 (2001).
(26.) S.S. Ray and M. Okamoto, Prog, Polym. Sci., 28, 1539 (2003).
(27.) N. Sheng, M.C. Boyce, D.M. Parks, G.C. Rutledge, J.I. Abes, and R.E. Cohen, Polymer, 45, 487 (2004).
(28.) T. Gopakumar, J. Lee, M. Kontopoulou, and J. Parent, Polymer, 43, 5483 (2002).
(29.) S. Pavlidou and CD. Papaspyrides, Prog. Polym. Sci., 33, 1119(2008).
(30.) H. Koerner, D. Misra, A. Tan, L. Drummy, P. Mirau, and R. Vaia, Polymer, 47, 3426 (2006).
(31.) W. Shao, Q. Wang, and H. Ma, Polym. Int., 54, 336 (2005).
(32.) R.A. Vaia and E.P. Giannelis, MRS Bull., 26, 394 (2001).
(33.) M. Abareshi, S.M. Zebarjad, and E.K. Goharshadi, J. Compos. Mater., 43, 2821 (2009).
(34.) C. Koo, H. Ham, M. Choi, S. Kim, and I. Chung, Polymer, 44, 681 (2003).
(35.) G. Wang, Y. Chen, and Q. Wang,./. Polym. Sci. B: Polym. Phys., 46, 807 (2008).
(36.) G. Zhang, A.K. Schlarb, S. Tria, and O. Elkedim, Compos. Sci. Technoi, 68, 3073 (2008).
(37.) W.P. Wang, Y. Liu, X.X. Li, and Y.Z. You, J. Appl. Polym. Sci., 100, 1427 (2006).
(38.) G.H. Chen, D.J. Wu, W.G. Weng, and W.L. Yan,./. Appl. Polym. Sci., 82, 2506(2001).
(39.) P. Xiao, M. Xiao, and K.C. Gong, Polymer, 42, 4813 (2001).
(40.) W.G. Zheng, S.C. Wong, and H.J. Sue, Polymer, 43, 6767 (2002).
(41.) X.S. Du, M. Xiao, and Y.Z. Meng, J. Polym. Sci. B: Polym. Phys., 42, 1972 (2004).
(42.) J.J. Mack, L.M. Viculis, A. Ali, R. Luoh, G.L. Yang, H.T. Hahn, F.K. Ko, and R.B. Kaner, Adv. Mater., 17, 77 (2005).
(43.) P.M. Uhl, Q. Yao, H. Nakajima, E. Manias, and C.A. Wilkic, Polym. Degrad. Stab., 89, 70 (2005).
(44.) A. Yasmin, J.J. Luo, and l.M. Daniel, Compos. Sci. Techno!., 66, 1182 (2006).
(45.) S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohl-haas, E.J. Zimncy, E.A. Stach, R.D. Piner, S.T. Nguyen, and R.S. Ruoff, Nature, 442, 282 (2006).
(46.) B. Debelak and K. Lafdi, Carbon, 45, 1727 (2007).
(47.) J. Li and J.K. Kim, Compos. Sci. Techno!., 67, 2114 (2007).
(48.) T. Ramanathan, A.A. Abdala, S. Stankovich, D.A. Dikin, M. Herrera-Alonso, R.D. Piner, D.H. Adamson, H.C. Schniepp, X. Chen, R.S. Ruoff, S.T. Nguyen, I.A. Aksay, R.K. Prud'homme, and L.C. Brinson, Nature Nanotech., 3, 327 (2008).
(49.) T. Ramanathan, S. Stankovich, D.A. Dikin, H. Liu, H. Shen, S.T. Nguyen, and L.C. Brinson, J. Polym. Sci. B: Polym. Phys., 45, 2097 (2007).
(50.) P. Steurer, R. Wissert, R. Thomann, and R. Muelhaupt, Macromol. Rapid Comm., 30, 316 (2009).
(51.) J.L. Vickery, A.J. Patil, and S. Mann, Adv. Mater., 21, 2180 (2009).
(52.) S. Ansari and E.P. Giannelis, J, Polym. Sci. B: Polym. Phys., 47, 888 (2009).
(53.) S. Bourdo, Z.R. Li, A.S. Biris, F. Watanabe, T. Viswana-than, and I. Pavel, Adv. Fund. Mater., 18, 432 (2008).
(54.) X.M. Chen, J.W. Shen, and W.Y. Huang, J. Mater. Sci. Lett., 21, 213 (2002).
(55.) T.G. Gopakumar and DJ.Y.S. Page, Polym. Eng. Sci., 44, 1162(2004).
(56.) K. Kalaitzidou, H. Fukushima, P. Askeland, and L.T. Drzal,./. Mater. Sci., 43, 2895 (2008).
(57.) H. Kim and C.W. Macosko, Macromolecules, 41, 3317 (2008).
(58.) K. Wakabayashi, C Pierre, D.A. Dikin, R.S. Ruoff, T. Ram-anathan, L.C. Brinson, and J.M. Torkelson, Macromolecules, 41, 1905 (2008).
(59.) S.C. Wong, E.M. Sutherland, and F.M. Uhl, Mater. Manuf. Process., 21, 159 (2006).
(60.) A.A. Balandin, S. Ghosh, W.Z. Bao, I. Calizo, D. Tewel-debrhan, F. Miao, and C.N. Lau, Nano. Lett., 8, 902 (2008).
(61.) B.T. Kelly, Physics of Graphite, Applied Science, London (1981).
(62.) A.K. Geim and K.S. Novoselov, Nat. Mater., 6, 183 (2007).
(63.) N. Furgiucle, A.H. Lebovitz, K. Khait, and J.M. Torkelson, Macromolecules, 33, 225 (2000).
(64.) A.H. Lebovitz, K. Khait, and J.M. Torkelson, Macromolecules, 35, 8672 (2002).
(65.) K. Khait and S.H. Carr, Solid-State Shear Pulverization: A New Polymer Processing and Powder Technology, Tech-nomic, Lancaster, PA (2001).
(66.) K.G. Kasimatis and J.M. Torkelson, PMSE Prepr., 91, 173 (2004).
(67.) G. Sui, W.H. Zhong, X.P. Yang, Y.H. Yu, and S.H. Zhao, Polym. Advan. Techno!., 19, 1543 (2008).
(68.) J. Gonzalez-Benito and G. Gonzalez-Gaitano, Macromolecules, 41, 4777 (2008).
(69.) Y.G. Zhu, Z.Q. Li, D. Zhang, and T. Tanimoto, J. Polym. Sci. B: Polym. Phys., 44, 1161 (2006).
(70.) M. Marie and C.W. Macosko, Polym. Eng. Sci., 41, 118 (2001).
(71.) K. Wakabayashi, P.J. Brunner, J. Masuda, S.A. Hewlett, and J.M. Torkelson, Polymer, 51, 5525 (2010).
Paul J. Hubert, Krishna Kathiresan, Katsuyuki Wakabayashi Department of Chemical Engineering, Bucknell University, Lewisburg, Pennsylvania 17837
Correspondence to: Katsuyuki Wakabayashi; e-mail: email@example.com Contract grant sponsor: NSF Major Research Instrumentation; contract grant number: CMMI #0820993.
|Printer friendly Cite/link Email Feedback|
|Author:||Hubert, Paul J.; Kathiresan, Krishna; Wakabayashi, Katsuyuki|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 2011|
|Previous Article:||Kinetics of the anionic polymerization of [epsilon] - caprolactam from an isocyanate bearing polystyrene.|
|Next Article:||Effects of acrylonitrile content and molecular weight on the scratch behavior of styrene-acrylonitrile random copolymers.|