Effects of nanoparticle feed location during nanocomposite compounding.
Polymer composites with a second phase, whether inorganic or organic, represent the most common type of reinforced polymers. Distinct from conventional composites, polymer nanocomposites (PNC) arc defined as polymers that include fillers having at least one dimension less than 100 nm (1). Polymer composites filled with micrometer-sized fillers show improvements in their mechanical properties (e.g., modulus and yield strength). Incorporating nanoparticles, however, in a polymer matrix can give rise to remarkable increases in mechanical and thermal properties at low filler concentrations (typically < 10%). In a polypropylene system examined by Sumita et al. (2), (3) nanoparticle-filled polypropylene exhibited considerable improvement in the yield stress (30%) and Young's modulus (170%) compared with the polypropylene filled with micrometer-sized particles.
The enhancement of properties in PNC depends on the (1) type of nanoparticle, (2) the dispersion of the nanoparticles in the polymer matrix, and (3) the adhesion of nanoparticles to the polymer. Nanoparticles are characterized by their shape, size, and aspect ratio. They can be classified as plate-like, rod-like, and particulate (4), with much of the more recent research focused on plate-like and rod-like nanoparticles which can provide greater improvements in nanocomposite properties. Dispersion of various nanoparticles has been obtained using solution, insitu polymerization, and melt compounding methods (5). The former two methods provide easier dispersion, but are limited to selected polymer systems and are less compatible with current manufacturing practices. Although melt compounding has been used for wide variety of polymer systems (6-17), complete nanoparticle dispersion is highly dependent on polymer-nanoparticle compatibility and on melt compounding parameters. The nanoparticles and polymer matrices have been compatibilized by treating the nanoparticles and/or the polymers (18). For melt compounding, twin screw mixers have produced better dispersion than single screw extruders, batch mixers, and other equipment (19), (20). The level of shearing, mixing (residence) time, and degree of mixer fill have emerged as the critical parameters in melt mixing (19), (21-24). Anderson (24) also noted that feed location was the most important processing parameter when compounding nanoclay into polypropylene using a corotating twin screw extruder. The greatest improvements in dispersion and mechanical properties occurred when the polymer and nanoparticles were fed together (24). Additionally, adding nanoclay downstream produced same results as feeding nanoclay by itself downstream (24). This feeding practice currently is commonly used for twin screw compounding of nanocomposites. These results are noteworthy because they are contrary to the results obtained by Chavarria et al, (25). They determined shorter particle lengths and lowest matrix reinforcement when polymer and nanoclay was fed together (25).
Working with nanoparticles can possibly result in occupational exposure to airborne particle concentrations (aerosols) that pose a potential hazard to human health (26-28). The small particle size and shape--along with corresponding changes surface area, chemical reactivity, porosity, and roughness--impact the biological implications following inhalation, ingestion, or transdermal delivery of airborne particles (26-35). Nanoparticles have been shown to accumulate in lungs (29), (30), (34); penetrate the skin (33); and travel along the nerves from the eyes and nose to the brain (29). Because surface area increases with decreasing particle size, nanoscale particles, including dust, are usually more reactive than comparable microparticles (28). Additionally, high aspect ratio nanoparticles, such as carbon nanotubes, have the potential to behave like asbestos, a fibrous particle that caused severe health effects in exposed workers (29), (31), (32). Overall, the impacts of nanoparticles on worker health and the environment are not well understood. The National Institute for Occupational Safety and Health has suggested preliminary guidelines for working safely with nanomate-rials (36) and has identified areas requiring additional research (37), whereas the U.S. Environmental Protection Agency intends to develop a roadmap for environmental heath and safety research needs (38).
This work was part of a larger study examining the impact of twin screw extrusion compounding practices on dispersion, mechanical properties, and the level of aerosolized nanoparticles. For this work, model poly(methyl-methacrylate)-nano aluminum oxide nanocomposites were compounded using three locations and methods for feeding nanoaiumina into twin screw compounding equipment. Transmission electron microscopy (TEM) was used to assess the dispersion of the nanoparticles in the polymer matrix, whereas mechanical properties were measured using ISO test methods. A fast mobility particle analyzer was employed to measure airborne material loss from the feeders and feed ports. The airborne particle levels were correlated with material loss. An indirect relationship was established among material loss, measured dispersion levels, and mechanical properties.
The base resin used in this work was poly(methylmethacrylate) (PMMA), grade H12-003 from Degussa Corporation. The measured glass transition temperature, [T.sub.g], was 114[degrees]C; the reported melt index is 7 g/10 min at 230[degrees]C and 3.8 kg; and density is 1.19 g/[cm.sup.3] with no prior added additives as mentioned in Degussa's technical information sheet. The resin was compounded with nanoa-luminum oxide ([Al.sub.2][O.sub.3]), grade Al-015-003-025, obtained from Nanophase Technologies Corporation. This nanoalumina is spherical in shape with average particle size ranging from 27 to 56 nm, density is 3.6 g/[cm.sup.3] and it exists as delta and gamma phase (70:30 ratio) as reported in Nanophase Corporation's technical information sheet.
This research is purely directed toward studying how feeding affects dispersion, mechanical properties, and airborne particles and is not interested in final applications of the nanocomposites so much as defining a model system that is reasonably representative of other nanocomposites and that can readily work with for the purposes of this study. First, this aluminum oxide was selected because it exists as small spherical particles and the skew-ness/Quadrat method cannot be readily applied to high aspect ratio particles. Second, nanoaiumina was used because it is readily available and has a high density (3.6 g/[cm.sup.3], vs. Silica, 2.2 g/[cm.sup.3], which will enhance the contrast between polymer and filler in TEM. Third, because of its shape, it can be easily measured using the Fast Mobility Particle Sizer (FMPS); (particles with an aspect ratio greater than one complicate particle size measurements in this instrument). The particles usually exist as four particle aggregates which flow easily in hoppers and feeders (39). As with all nanoparticles, its surface chemistry influences the degree of agglomeration and the ease with which the particle can be dispersed (40). Previous research with this nanoalumina has been limited to (1) tri-biological properties in poly(ethylene terephthalate) (PET) and poly(tetrafluroethylene) (41-46) and (2) thermal behavior in PET and poly(methylmethacrylate) (47-49), and dispersion studies in PET (39). Although the nanoaiumina particles exhibit better adhesion to PET than to PMMA, the highly reactive particle surface causes degradation of PET and necessitates compounding PET under a nitrogen atmosphere (39). Because such compounding did not facilitate measurement of aerosolized particles, PMMA, which does degrade when compounded in an oxygen-containing atmosphere, was chosen as the matrix resin for this study.
Prior to compounding, the PMMA was oven dried at 80[degrees]C for 8 h and the nanoalumina at 100[degrees]C for 24 h under vacuum. Both, neat PMMA and PMMA-nanoalu-mina based nanocomposites was compounded using a corotating twin screw extruder (Werner & Pfeilderer ZSK-30 mm, L/D = 32) using screw configuration shown in Fig. 1. This screw configuration has downstream feeding options and is illustrated in Fig. 2. Figure 2a-c shows a schematic diagram of three configurations used for feeding nanoalumina. The PMMA was always fed into extruder's primary feed port (hopper) through the single screw volumetric feeder for all formulations. As shown in Fig. 2 and Table 1, nanoalumina filler loadings of 2% and 5% were controlled using a twin screw volumetric feeder (K-Tron Soder K2-MV-T60, 55 to 5200 [dm.sup.3]/h) or pre-mixed with the PMMA. This volumetric feeder consists of long pitch twin concave screws with a filling capacity of 90%. For premixed formulations (nanoalumina and PMMA), distributive mixing was carried out in a rotary mill (Thomas-Wiley Rotary Mill, Model 4). The location of nanoalumina feeding was varied (Table 1). Strands emerging from the extrusion die were cooled in a water bath and then strand pelletized. The temperature zones of the extruder from the hopper to the die were maintained at 250[degrees]C, whereas the screw speed was set at 80 rpm Extruded PMMA-nanoalumina pellets produced via the above the blending process were subsequently used in injection molding
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
TABLE 1. Feeding configurations for twin screw extruder. Concentration of nanoalumina Feeding method for nanoalumina 0% 2% 5% Single-screw feeder into primary feed port Neat Premix of PMMA and PMMA nanoalumina Twin-screw feeder into primary feed port -- Nanoalumina Twin-screw feeder into secondary feed port -- Nanoalumina
The nanocomposite pellets were oven dried at 80[degrees]C for a period of 8 h prior to injection molding. Using a 50-ton machine (Ferromatik Cincinnati Milacron LUSS 5.8), tensile bar specimens were molded according to ISO 3167 (ASTM D 5936) specifications. The temperatures used, from zone 1 to nozzle, were 210[degrees]C, 230[degrees]C. 250[degrees]C, and 250[degrees]C for all formulations. The mold temperature (27[degrees]C) and other processing conditions were held constant for all samples.
Tensile testing was performed on molded specimens in accordance with a modified ISO 527-1 (ASTM D5937) technique. For each formulation, live tensile specimens were tested on a universal testing machine (Instron 6205) with a standard gage length of 115 mm and a testing speed of 1 mm/min. General purpose strain gages (350 ohms grid resistance) from Vishay micromeasurements were used to determine strain values. Young's modulus, tensile stress at yield, and tensile strain at break were calculated from these measurements.
Dynamic Mechanical Analysis
Investigations of the effects of processing on the visco-elastic behavior of the various batches were studied through dynamic mechanical analysis (DMA). A TA Instrument's Q800-0603 dynamic mechanical analyzer was used with a multifrequency strain (temperature ramp/ frequency sweep) program to determine the elastic and viscous moduli (G' and G", respectively). Each of the molded specimens was conditioned according to ISO 6721-5 (ASTM D 5023) and oven dried for 4 h prior to testing. Flexural (three-point bending) tests were carried out using a temperature range of 10-150[degrees]C, heating rate of 3[degrees]C/min, and frequency of 1 Hz (6 radians/sec).
Differential Scanning Calorimetry
A TA Instruments DSC Q200 differential scanning calorimeter (DSC) was used to determine the glass transition temperature of the specimens in accordance with ISO 11357-2 (ASTM D 3418). Specimens in the form of oven-dried pellets with constant mass (18 mg) were heated at a constant rate of 10[degrees]C/min for a range of temperatures of 30-300[degrees]C and the heat flow was recorded as a function of temperature.
A TA Instruments TGA Q50 thermogravimetric analyzer (TGA) was employed for determining the compositional analysis of the specimens in accordance with ISO 11358 (ASTM E 1131-03). A 12-mg specimen (oven- dried compounded pellets) was heated at a constant rate of 10[degrees]C/min. for a temperature range of 30-650[degrees]C, and the mass was recorded as a function of temperature. Constant flow of nitrogen gas (40 ml/min) was maintained throughout the analysis. Mass loss over specific temperature ranges provided a compositional analysis for each specimen.
Transmission Electron Microscopy
TEM imaging was used to assess nanoalumina dispersion in the PMMA polymer matrix. Samples of each nanocomposite were embedded in epoxy in individual test-tube molds and sectioned using a cryo ultra microtome in preparation for TEM. Initially, a glass knife was used to section the epoxy mold, but to section nanocomposite specimens, a diamond knife was used. The sectioned specimens were collected on a TEM grid [Electron Microscopy Sciences, 200 mesh copper carbon (50t)]. TEM was carried out using Philips EM--400 with an accelerating voltage of 100 kV.
Fast Mobility Particle Sizer
During compounding, the concentrations of nanoparticles in the atmosphere were measured by the FMPS. The FMPS spectrometer (Model 3091, TSI) in the range from 5.6 to 560 nm, offering a total of 32 channels of resolution (16 channels per decade). The FMPS spectrometer performs particle size classification based on differential electrical mobility classification.
The FMPS spectrometer draws an aerosol sample into the inlet continuously. Using a corona charger, particles are charged to a precisely predictable distribution. The charged particles then enter the measurement region along the surface of a high-voltage electrode. They move down the length of the electrode alongside a sheath of HEPA-filtered air which separates them onto a series of annular electrometer electrodes. This produces particle-size-distribution measurements with one-second resolution, providing the ability to visualize particle events and changes in particle size distribution in real time. When particles strike the electrometers, they transfer the charge they carry to the electrometer where it is measured in real time. A particle with a high electrical mobility strikes an electrometer near the top of the measurement region, whereas a particle with lower electrical mobility travels farther and strikes an electrometer lower down the stack. The particle number concentration is determined by measurement of the electrical current collected on a series of electrodes.
During compounding trials, data were collected near the feed ports as the source location (75 mm) and at the operator's feeding zone (at the feed ports) (50). Typically, measurements started during extruder warm up and continued until the compounding trials were finished. The filter and cyclone in the FMPS were cleaned prior to each trial.
RESULTS AND DISCUSSION
TGA analysis was performed on the PMMA-nanoalumina nanocomposites to determine the actual filler concentrations. As shown Table 2, the change in mass for single and twin screw feeders, [DELTA]m, varied from 90% to 93%. Change of mass vs. temperature curves of TGA was divided into three stages. The first stage, denoting the temperature range between 30[degrees]C and 315[degrees]C with a weight loss of about 5%, is attributed to the loss of water. The second stage showed a significant weight loss of about 93% (2% loading) and 90% (5% loading) in the temperature range of 315-475[degrees]C. The third stage, featuring no weight loss between 475[degrees]C and 650[degrees]C, is related to the presence of 2% and 5% nanoalumina particles, respectively. In the case of neat PMMA there is a 100% weight loss in the range 315-475[degrees]C and is attributed to the decomposition of PMMA. Ash et al. (47) reported that physisorbed water in PMMA-nanoalumina nanocomposites desorbed at temperatures of 100-400[degrees]C, whereas water bonded to hydroxyl and other polar function groups was lost at temperatures greater than 400[degrees]C. Addition of the 5% water loss to the measured changes in mass provided the corrected change in mass, [DELTA][m.sub.c], values listed in Table 2. The measured changes in mass for both single and twin screw feeders were consistent with the filler concentrations of 98% and 95% at nanoalumina loadings of 2% and 5%, respectively. These results, which suggest that nanoalumina was fed consistently through the hopper (single screw volumetric feeder), supporting Anderson's observations for feeding of nanoclays with polypropylene pellets (24) and Kim's results for nanoalumina and PET (39).
TABLE 2. Change in mass and degradation temperatures obtained from TGA. [DELTA]m (%) al [DELTA] [T.sub.d] loading of [m.sub.c] ([degrees]C) (%) at at loading of loading of Feeding method for 0% 2% 5% 0% 2% 5% 0% 2% 5% nanoalumina Single-screw feeder into 95 92.78 89.98 100 98 95 419 388 387 primary feed port Twin-screw feeder into -- 92.99 89.85 -- 98 95 -- 386 386 primary feed port Twin-screw feeder into -- 93.46 90.77 -- 98 96 -- 385 385 secondary feed port
Thermal stability of neat PMMA and PMMA compounded with nanoalumina was also evaluated using TGA analysis. The maximum degradation temperatures obtained from TGA analysis are given in Table 2. Degradation temperatures varied from 419[degrees]C for neat PMMA to 385-388[degrees]C for PMMA-nanoalumina nanocomposites, indicating that the feeding methods and nanoalumina concentration did not substantial impact the stability of the PMMA-nanoalumina nanocomposites. However, it was seen that this thermal stability decreased on addition of nanoalumina irrespective of the feeding methods. It is also important to note that maximum degradation temperature for neat PMMA was 419[degrees]C which was considerably higher than the temperature at which PMMA decarboxylizes. This illustrates thermal stability of neat PMMA resin toward degradation. This thermal stability can be correlated to the additive (thermal stabilizers) which might be present within PMMA (originally by the resin manufacturer). The nanocomposites also exhibit thermal stability at melt processing temperatures; these results were consistent with those reported by Ash et al. (47).
Figure 3 presents the TEM images of nanoalumina dispersed in the PMMA matrix. At the 2% nanoalumina loading, feeding of nanoalumina through twin screw volumetric feeder (secondary feeder) into the primary feed port produced visibly better dispersion than the other two feeding methods. With filler concentration of 5%, twin-screw feeding of nanoalumina into the primary feed port again provided the best dispersion, but twin-screw feeding into the primary feed port was better than feeding the nanoalumina as a premix with PMMA. Chavarria et al. (25) found similar results with lower degree of dispersion when polymer was premixed with nanoclay for polyamide (PA-6) nanocomposites system. These results contradict earlier results obtained by Anderson (24) predicting an increase in degree of dispersion when a polymer was premixed with nanoclay. This difference in results is more complicated than just feed location, depending on the materials system being studied. Magnification of the nanocomposites with 5% nanoalumina (Fig. 3g-i) again shows more primary particles and aggregates with twin-screw feeding than with single-screw feeding. This increase in primary particles and aggregates with twin-screw feeding at 5% nanoalumina is mainly because of its twin-screw design thus allowing more nanoalumina particles to be fed into the extruder resulting in higher density of particles. This high density of nanoalumina particles and agglomerates are easy to detect at higher magnifications when compared with particles with low density. As shown in Fig. 3h, twin-screw feeding into the primary feed port produced the largest number of primary particles and the average particle size was about 50-60 nm.
[FIGURE 3 OMITTED]
Because many of the TEM images were relatively similar in appearance, the difference in dispersion was quantified by the degree of asymmetry of a statistical distribution around its mean and can be quantified by its skewness [beta], which is defined by Eq. 1
[beta] = ([q/[(q - 1)(q - 2)]]) [SIGMA] [([[[N.sub.qi] - [N.sub.q.sup.mean]]/[sigma]]).sup.3] (1)
where q is the total number of quadrats studied, [N.sub.qi] is the number of particles in the ith quadrat (i = 1, 2, ..., q), [N.sub.q.sup.mean] is the mean number of particles per quadrat, and [sigma] is the standard deviation of the [N.sub.q] distribution. In this method, variation in quadrat size can result in different skewness values. Too small a quadrat size will show aggregation, because there will be many empty quadrats. However, too large a quadrat size may not reflect true aggregative effects, because the quadrats will tend to have the same number of particles in each (48), (49). So, an optimum grid of 720 square cells was placed on each of the TEM images shown in Fig. 3a-f and the number of particles in each cell was counted (49). These particle counts are used to calculate skewness which is inversely related to dispersion of particles in a polymer matrix. Therefore, lower skewness values correlate to good dispersion. Karnezis et al. (51) correlated increase in [beta] values to agglomeration of particles. Smaller values of [beta] were correlated to less agglomeration and uniform distribution of particles. Kim et al. (49) demonstrated similar results using PET with nanoalumina. The skewness values for the PMMA-nanoalumina nanocomposites are listed in Table 3.
TABLE 3. Skewness values. [beta] at loading of Feeding method for nanoalumina 2% 5% Single-screw feeder into primary feed port 224 13.7 Twin-screw feeder into primary feed port 10.5 1.1 Twin-screw feeder into secondary feed port 42.5 4.4
At a nanoalumina loading of 5%, twin-screw feeding of nanoalumina produced skewness values of 1.1 and 4.4 for the primary and secondary feed ports, respectively. This result was not unexpected because nanoalumina fed into the primary feed port had a longer extruder residence time than the material fed into the secondary feed port. Feeding of premixed PMMA and nanoalumina into the primary feed port via a single-screw feeder, however, produced a significantly greater [beta] value (13.7). The same trend occurred for the nanocomposites with 2% nanoalumina although the skewness values were much greater than for 5% nanoalumina materials. As seen in Fig. 3a, feeding of premixed PMMA with 2% nanoalumina using single-screw feeder have lower density of particles when compared with premixed PMMA with 5% nanoalumina (Fig. 3d). At the optimum size of each quadrat, lower density of particles in case of premixed PMMA with 2% nanoalumina has large number of empty quadrat resulting in higher values of skewness ([beta]). This high [beta] value correlates to poor dispersion. The [beta] values for the nanocomposites with 5% nanoalumina prepared using twin-screw feeders are significantly lower than single-screw feeder. This difference is again related to the particle size and density of particles. By visually comparing TEM images of single-screw feeding (Fig. 3c) with twin-screw feeding into secondary feed port (Fig. 3f) for 5% nanoalumina, we find relatively same density of particles. However, with close observation, a large clump of agglomerated nanoalumina is noticed in single-screw feeding (Fig. 3c). This clump is regarded as a single particle and, thus, increases the [beta] value significantly. Switching the twin-screw feeder from the primary to secondary feed port provided a four-fold increase in skewness, regardless of the filler concentration. Using the single-screw feeder at the primary feed port gave 21- and 12-fold increases in skewness for nanoalumina loadings of 2% and 5%, respectively. The lack of consistency with the single-screw feeder indicates nonuniform feeding of the nanoalumina, but comparison of feeder types suggests that the twin-screws feeding mechanism is helping to break up the nanoalumina aggregates.
Fast Mobility Particle Sizer
The raw results from the FMPS are presented as particle normalized number concentration vs. average particle diameter at source location for 2% and 5% nanoalumina (Fig. 4a-d). These measurements were obtained by sampling at source location (75 mm) and at the feed location of the appropriate feed ports. Although these data were later corrected for other particles in the laboratory using a previously established method (49), the raw data suggest mechanisms that support the TEM results. For premixed PMMA and nanoalumina (2% and 5%, respectively), the particle normalized number concentration vs. average particle diameter curves at the source location showed peaks at about 45 and 190 nm. These two peaks are also present in the curves when nanoalumina was fed into the secondary feed port, but the larger diameter peak was less intense. Feeding nanoalumina through the twin-screw feeder into the primary feed port produced curves with single peak at about 60 nm. For premixed PMMA and nanoalumina (2% and 5%, respectively), the particle normalized number concentration vs. average particle diameter curve at the feed location showed a two sharp peaks at about 60 and 190 nm. These two peaks were also present when 2% nanoalumina was fed into the secondary feed port with the larger diameter peak less intense. However, for 5% nanoalumina, a sharp single peak is present at about 60 nm when fed using any of the feeding methods. The peak at 45-60 nm on Fig. 4a,b is because of background polymer fume that is always present, independent of feeding mechanism. Premix feeding was a more energetic method to release aggregate nanoalumina particles into the room because of the dumping and transportation of a mixture of polymer pellets and nanoalumina particles. The particle concentration peaking at 190 nm was the aggregate nanoalumina particles. Overall, these results suggest that the twin screws in the volumetric feeder are assisting in break up of the nanoalumina aggregates.
[FIGURE 4 OMITTED]
Glass Transition Temperature
The glass transition temperature of this PMMA-nanoa-lumina nanocomposite was determined by DMA and differential scanning calorimetry (DSC). As shown in Table 4, both measurement techniques gave similar glass transition temperatures. These glass transition temperatures were 96-97C regardless of amount of or method used in feeding the nanoalumina, but they were substantially lower than the [T.sub.g] of the neat PMMA (114[degrees]C). Although the glass transition temperature should increase, or at least remain constant, upon introduction of a high modulus filler into a polymer (41), (52-54), some nanopar-ticulate fillers have produced reductions in [T.sub.g] (47), (52), (54-56). For a 5% loading of this specific nanoalumina, Kim (39) reported a 2.5[degrees]C decrease in the [T.sub.g] of PET. Using the same nanoalumina. Ash et al. (54) found that, at an optimum weight fraction, the glass transition temperature of the PMMA-nanoalumina nanocomposites produced via in-situ polymerization decreased by more than 20[degrees]C and that loadings as low as 0.5% could induce a 25[degrees]C reduction in the [T.sub.g] of PMMA (13). The current 17[degrees]C reduction in [T.sub.g] is more consistent with the latter results, which may be because of increased mobility near the surface of noninteracting nanoparticles (52), (55).
TABLE 4. Glass transition temperatures determined by (a) dynamic mechanical analysis and (b) differential scanning calorimetry at varying nanoalumina loadings. [T.sub.g] [T.sub.g] ([degrees]C) ([degrees]C) from DMA at form DSC at loading of loading of Feeding method for nanoalumina 0% 2% 5% 0% 2% 5% Single-screw feeder into primary feed port 114 96 97 114 96 96 Twin-screw feeder into primary feed port -- 97 96 97 96 Twin-screw feeder into secondary feed port -- 97 97 -- 96 96
Table 5 presents the average values for tensile properties of the PMMA-nanoalumina nanocomposites. The Young's modulus, E, and tensile yield strength, [sigma], for the neat PMMA with the same heat history were 3130 and 56 MPa, respectively; the elongation at break, [[epsilon].sub.b] was 3.32%. As shown in Fig. 5a, the modulus of the nanocomposites compounded using the premixed PMMA and nanoalumina fed via a single-screw feeder increased linearly at a rate of 46 MPa per percent nanoalumina added to the PMMA. For these nanocomposites, the mixing closely followed the rule of mixtures; the coefficient of determination, [r.sup.2], was 0.99. A similar trend was observed for tensile yield strength, with the rate of increase being about 1 MPa per percent nanoalumina and the [r.sup.2] being 0.90 (Fig. 5b). When the nanoalumina was fed from the twin-screw feeder, however, both modulus and yield strength increased sharply with the addition of even 2% alumina. Longer mixing time (i.e., feeding through the primary feed port) further increased the modulus and yield strength. At a nanoalumina loading of 5%, feeding of the premix into the primary feed port produced a 7% increase in modulus and an 11% improvement in yield strength. With twin-screw feeding of the nanoalumina through the secondary and primary feed ports, the moduli increased by 13% and 20%, respectively. The yield strength exhibited 28% and 48% increases for similar feeding configurations. These results were consistent with the dispersion of the nanoalumina in the PMMA.
[FIGURE 5 OMITTED]
TABLE 5. Tensile properties of nanocomposiles. E (GPa) at loading of Feeding method 0% 2% 5% for nanoalumina Single-screw 3.1 [++or--] 2.8 3.2 [++or--] 2.0 3.3 [++or--] 1.9 feeder into primary feed port Twin-screw -- 3.5 [++or--] 1.2 3.7 [++or--] 1.7 feeder into primary feed port Twin-screw -- 3.4 [++or--] 6.7 3.5 [++or--] 2.1 feeder into secondary feed port [delta] (MPa) at loading of Feeding method for 0% 2% 5% nanoalumina Single-screw 56 [++or--] 1.1 60 [++or--] 0.7 62 [++or--] 0.9 feeder into primary feed port Twin-screw feeder -- 71 [++or--] 0.5 83 [++or--] 1.2 into primary feed port Twin-screw feeder -- 65 [++or--] 1.1 71 [++or--] 0.6 into secondary feed port [[epsilon].sub.b] at loading of Feeding method 0% 2% 5% for nanoalumina Single-screw 3.3 [++or--] 0.1 2.6 [++or--] 0.2 2.5 [++or--] 0.1 feeder into primary feed port Twin-screw -- 2.5 [++or--] 0.1 2.1 [++or--] 0.2 feeder into primary feed port Twin-screw -- 2.6 [++or--] 0.1 2.4 [++or--] 0.1 feeder into secondary feed port
Elongation at break, however, decreased with higher nanoalumina loadings and better dispersion (see Fig. 6). At a nanoalumina loading of 5%, feeding of the premix into the primary feed port produced a 23%; decrease in elongation, whereas twin-screw feeding of the nanoalumina through the secondary and primary feed ports provided reductions of 26% and 35%, respectively. In previous work with this nanoalumina and PMMA, this behavior and the reductions in the glass transition temperature indicate poor adhesion and a weak interface between the nanoalumina particles at low nanoalumina loadings resulting in an increase in ductility (48), (53), (55).
[FIGURE 6 OMITTED]
Table 6 lists the storage and loss moduli, G' and G", for the PMMA-nanoalumina nanocomposites. The G' and G" values of the neat PMMA were 2237 and 1987 MPa, respectively. As shown in Fig. 7a,b, the storage modulus and loss modulus exhibited trends similar to tensile modulus. Feeding of premixed PMMA and nanoalumina produced nanocomposites with storage and loss moduli that obeyed the rule of mixtures. These moduli increased at rates of 50 and 74 MPa per percent added to the PMMA, respectively; [r.sup.2] was 0.98 for both moduli. Feeding with the twin-screw feeder produced greater increases in moduli and these increases greater for better dispersed systems. With the best dispersion of the nanoalumina in the PMMA, storage modulus increased by 27% whereas loss modulus was increased by 36%;. Lines shown in Figs. 5-7 are only included for purpose of clarity and there is no analytical model attached to it.
[FIGURE 7 OMITTED]
TABLE 6. Storage and loss moduli of nanocomposites at a frequency of 1 Hz. G' (GPa) at loading of Feeding method 0% 2% 5% for nanoalumina Single-screw 2.2 [++or--] 1.0 2.3 [++or--] 1.2 2.4 [++or--] 1.1 feeder into primary feed port Twin-screw -- 2.6 [++or--] 0.3 2.8 [++or--] 0.2 feeder into primary feed port Twin-screw -- 2.5 [++or--] 0.7 2.6 [++or--] 0.4 feeder into secondary feed port G" (GPa) at loading of Feeding method 0% 2% 5% for nanoalumina Single-screw 1.9 [++or--] 1.3 2.1 [++or--] 1.5 2.3 [++or--] 1.4 feeder into primary feed port Twin-screw -- 2.4 [++or--] 0.7 2.7 [++or--] 0.6 feeder into primary feed port Twin-screw -- 2.3 [++or--] 1.1 2.4 [++or--] 0.9 feeder into secondary feed port
PMMA-nanoalumina nanocomposites were compounded in a twin screw extruder using three different feeding configurations to feed the nanoalumina; PMMA was always fed into extruder's primary feed port. Results from thermogravimetric analysis indicated that feeding of premixed PMMA and nanoalumina using a single-screw volumetric feeder produced inconsistent feeding, but that use of a twin-screw volumetric feeder produced uniform feeding of the nanoalumina. Transmission electron micrographs, statistical analysis of those images (which produced skewness, [beta], values), and measurements of aerosolized nanoalumina particles showed that the twin-screw feeder was helping to break up nanoalumina aggregates. As a result, twin-screw feeding of nanoalumina into the primary feed port produced the best dispersion of the nanoparticles, whereas single-screw feeding a premixed PMMA and nanoalumina into the same port provided the least dispersion.
With the addition of 2% and 5% nanoalumina, the glass transition temperature decreased from 114[degrees]C to 97[degrees]C. Tensile modulus and yield strength as well as storage moduli measured via three-point bending increased with the addition of nanoalumina. When the PMMA and nanoalumina was fed as a mixture using a single-screw volumetric feeder, these properties of the nanocomposite followed the rule of mixtures. Nanocomposites for which the nanoalumina was fed using twin-screw feeder, however, has properties that increased more rapidly than predicted by the rule of mixtures. For all the PMMA-nanoalumina nanocomposites, the relative changes in tensile modulus and yield strength as well as storage moduli exhibited a power law relationship with dispersion as quantified using skewness. In contrast to the other properties, (tensile) elongation at break decreased with nanoalumina loading and better dispersion.
Results from FMPS indicate that single screw feeding into the primary feed port resulted in higher release of aggregate nanoalumina particles. Results from Fig. 4a-d shows that the twin screw volumetric feeder into the primary feed port resulted in high background concentration caused by polymer fume which is present at all times and is independent of the feeding method.
This work was supported by the National Science Foundation as a Nanoscale Science and Engineering Centers Program (Award # NSF-0425826).
(1.) Y. Kojima, A. Usuki, M. Kawasumi, A. Okada, T. Kurau-chi, and O. Kamigaito, J. Polym. Sci. Parr A: Polym. Chem., 31, 983 (1993).
(2.) M. Sumita, Y. Tsukumo, K. Miyasaka, and K. Ishikawa, Mater. Sci., 18, 1758 (1983).
(3.) M. Hussain, A. Nakahara, S. Nishijima, and K. Niihara, Mater. Lett., 27, 21 (1996).
(4.) J. Edenbaum, Plastics Additives and Modifiers Handbook, Van Nostrand Reinhold, New York, NY (1992).
(5.) S.S. Ray and M. Okamoto, Prog. Polym. Sci, 28, 1548 (2003).
(6.) L. Liu, Z. Qi, and X. Zhu, J. Appl. Polym. Sci., 71, 1133 (1999).
(7.) T.D. Fornes, P.J. Yoon, D.L. Hunter, H. Keskkula, and D.R. Paul. Polymer, 43, 5915 (2002).
(8.) N. Hasegawa, H. Okamoto, M. Kato, A. Usuki, and N. Sato, Polymer, 44, 2933 (2003).
(9.) A. Usuki, M. Kato, A. Okada, and T. Kurauchi, J. Appl. Polym. Sci., 63, 137 (1997).
(10.) M. Kawasumi, N. Hasegawa, M. Kato, A. Usuki, and A. Okada, Macromolecules, 30, 6333 (1997).
(11.) N. Hasegawa, M. Kawasumi, M. Kato, A. Usuki, and A. Okada, J. Appl. Polym. Sci., 67, 87 (1998).
(12.) P.H. Nam, P. Maiti, M. Okamoto, T. Kotaka, N. Hasegawa, and A. Usuki, Polymer, 42, 9633 (2001).
(13.) X. Liu and Q. Wu, Polymer, 42. 10013 (2001).
(14.) D. Kaempfer, R. Thomann, and R. Mulhaupt, Polymer, 43, 2909 (2002).
(15.) C.H. Davis, L.J. Mathias, J.W. Gilman, D.A. Schiraldi, J.R. Shields, P. Trulove, T.E. Sutto, and H.C. Delong, J. Polym. Sci. Part B: Polym. Phys., 40, 2661 (2002).
(16.) B.J. Chisholm, R.B. Moore, G. Barber, F. Khouri, A. Hempstead, M. Larsen, E. Olson, J. Kelley, G. Balch, and J. Car-aher, Macromolecules, 35, 5508 (2002).
(17.) X. Huang, S. Lewis, W.J. Brittain, and R.A. Vaia, Macromolecules, 33, 2000 (2000).
(18.) J.W. Cho and D.R. Paul, Polymer, 42, 1083 (2001).
(19.) H.R. Dennis, D.L. Hunter, D. Chang, S. Kim, J.L. White, J.W. Cho, and D.R. Paul. Polymer, 42, 9513 (2001).
(20.) M.K. Dolgovskij, P.D. Fasulo, F. Lortie, C.W. Macosko, R.A. Ottaviani, and W.R. Rodgers, SPE ANTEC, 49, 2255 (2003).
(21.) P.D. Fasulo, W.R. Rodgers, R.A. Ottaviani, and D.L. Hunter, Polym. Eng. Sci., 44, 1036 (2004).
(22.) W. Lertwimolnun and B. Vergnes, Polymer, 46, 3462 (2005).
(23.) D. Kim, J.S. Lee, C.F. Barry, and J.L. Mead, Polym. Eng. Sci., 45, 1031 (2005).
(24.) P.G. Anderson, Proc. Ann. Conf. Soc. Plast. Eng., 48 (2002).
(25.) F. Chavarria, R.K. Shah, D.L. Hunter, and D.R. Paul, Polym. Eng. Sci., 47(11), 1847 (2007).
(26.) A.D. Maynard, R.J. Aitken, T. Butz, V. Colvin, K. Donaldson, G. Oberdorster, M.A. Philbert, J. Ryan, A. Seaton, V. Stone, S.S. Tinkle, L. Tran, N.J. Walker, and D.B. Warheit, Nature, 444(7717), 267 (2006).
(27.) K. Donaldson, L. Tran, L.A. Jimenez, R. Duffin, D.E. Newby, N. Mills, W. MacNee, and V. Stone, Part. Fibre Toxicol, 2, 10 (2005).
(28.) World Health Organization Publication No. 80. Evaluation of Exposure to Airborne Particles in the Work Environment, World Health Organization, Geneva (1984).
(29.) M.R. Gwinn and V. Vallyathan, Environ. Health Perspect., 114(12), 1818 (2006).
(30.) P.H.M. Hoet, 1. Bruske-Hohlfeld, and O.V. Salata, J. Nano-biotechnol., 2, 12 (2004).
(31.) K. Donaldson, R. Aitken, L. Tran, V. Stone, R. Duffin, G. Forrest, and A. Alexander, Toxicol. Sci., 92(1), 5 (2006).
(32.) C.-W. Lam, J.T. James. R. McCluskey, S. Arepalli, and R.L. Hunter, Crit. Rev. Toxicol., 36(3), 189 (2006).
(33.) N.A. Monteiro-Riviere and A.O. Inman, Carbon, 44(6), 1070 (2006).
(34.) P.J.A. Borm, Inhal. Toxicol., 14(3), 311 (2002).
(35.) L.E. Murr, E.V. Esquivel, and J.J. Bang, J. Mater. Sci.: Mater. Med., 15, 237 (2004).
(36.) National Institute for Occupational Safety and Health (NIOSH). Approaches to Safe Nanotechnology: An Information Exchange with NIOSH, Version 1.1, 51, (July, 2006).
(37.) National Institute for Occupational Safety and Health (NIOSH), Strategic Plan for NIOSH Nanotechnology Research: Filling the Knowledge Gaps, Draft Document, 69. (September 28, 2005).
(38.) EPA Nanotechnology Workgroup, Nanotechnology White Paper, U.S. Environmental Protection Agency 100/B-07/001, EPA Science Policy Council, 132, (February 15, 2007).
(39.) D. Kim, Investigation of Melt Process for Preparation of Alumina Nanocomposites, Doctoral Dissertation, University of Massachusetts Lowell (2006)
(40.) R.A. Vaia and H.D. Wagner, Mater. Today, 7, 32 (2004).
(41.) P. Cousin and P. Smith, J. Polym. Sci. Part B: Polym. Phys., 32, 459 (1994).
(42.) W.G. Sawyer, K.D. Freudenberg, P. Bhimaraj, and L.S. Schadler, Wear, 254, 573 (2003).
(43.) B. Wetzel, F. Haupert, and M. Qui Zhang, Composite Sci. Technol., 63, 2055 (2003).
(44.) P.Y. Lee and T. Yano, J. Eur. Ceramic Soc, 24, 3359 (2004).
(45.) G. Shi, M.Q. Zhang, M.Z. Rong, B. Wetzel, and K. Frie-drieh, Wear, 256, 1072 (2004).
(46.) P. Bhimaraj, D.L. Burris, J. Action, W.G. Sawyer, C.G. Toney, R.W. Siegel, and L.S. Schadler, Wear, 258, 1437 (2005).
(47.) B.J. Ash, R.W. Siegel, and L.S. Schadler, J. Polym. Sci.: Part B: Polym. Phys., 42, 4371 (2004).
(48.) A. Rogers, Statistical Analysis of Spatial Dispersion: The Quadrat Method, Pion, London (1974).
(49.) D. Kim, J.S. Lee, C.F. Barry, and J.L. Mead, Polym. Eng. Sci., 47, 2049 (2007).
(50.) S.-J. Tsai, A. Ashter, E. Ada, J.L. Mead, C.F. Barry, and M.J. Ellenbecker, J. Sci. Technol. Adv. Mater., 8(2), 160 (2008).
(51.) P.A. Karnezis, G. Durrant, and B. Cantor, Mater. Char., 40, 97 (1998).
(52.) B.J. Ash, D.F. Rogers, C.J. Wiegand, L.S. Schadler, R.W. Siegel, B.C. Benicecwicz, and T. Apple, Polym. Composites, 23, 1014 (2002).
(53.) L. Nielsen and R. Landel, Mechanical Properties of Polymers and Composites, Marcel Dekker Inc., New York (1994).
(54.) B.J. Ash, R.W. Siegel, and L.S. Schadler, Macromolecules, 37, 1358 (2004).
(55.) B.J. Ash, L.S. Schadler, and R.W. Siegel, Mater. Lett.. 55, 83 (2002).
(56.) C. Becker, H. Krug, and H. Schmidt, Mater. Res. Soc. Symp. Proc, 435, 237 (1996).
Ali Ashter, Su-Jung Tsai, Jun S. Lee, Michael J. Ellenbecker, Joey L. Mead, Carol F. Barry
Center for High-Rate Nanomanufacturing, University of Massachusetts Lowell, One University Avenue, Lowell, Massachusetts
Correspondence to: Ali Ashter; e-mail: email@example.com