Rubber World — January 2012
Evaluation of micronized rubber powders with cost/performance benefits
The objectives of this article are twofold. The first objective involves characterizing the micronized rubber powders (MRPs) using various physical and chemical techniques that are essential to understanding the physical and chemical structure and morphology. This enables their cost effective application and further optimization in various markets such as high performance tires and rubber goods, paints/coatings, asphalts and plastics. The MRPs are produced using a cryogenic turbo mill technology from 400 ìm down to 50 ìm size range. A majority of earlier investigations (refs. 1-3) used large size (180-400 ìm) ambient, ambient-wet or thermally regenerated ground tire rubber (GTR) powders in plastics. The commercial ambient processes can routinely produce 400 ìm particles, however, at particle sizes less than 100 ìm, throughputs are low. To understand the effects of the grinding process on the rubber particles, an ambient sample with 400 ìm nominal size is also included in the study. Previous research suggested that ambient grind particles have a rougher texture and higher surface area compared to cryogenic grind particles at the same nominal size (refs. 4 and 5). The present study shows deeper insight into the similarities and differences between the cryogenic and ambient grind particles.
The second objective focuses on an attempt to correlate the results of the characterization work to the processing and properties of rubber compounds made with MRP. Previous work (refs 6-8) in MRP-containing rubber composites indicated an effect on both processing and properties; in this work, we attempt to relate those properties back to a fundamental analysis of the MRP surface.
MicroDyne is a cryogenically ground micronized rubber powder produced by Lehigh Technologies. It is a free flowing, black rubber powder produced from end-of-life tires that easily disperses into a multitude of systems and applications (e.g., roof coatings, adhesives, asphalt, plastic resins, sealants, etc.). Several products were used for characterization in this study, as shown in table 1. These products are typically less expensive than the rubber compounds into which they are substituted. For example, as of this writing, the current cost for tire rubber is approximately $1.50 /lb.
For comparison, ambient ground tire derived rubber powders were also included in the study (table 1).
Thermogravimetric analysis (TGA)
TGA on all the rubber powder samples was performed to evaluate the compositions of the constituents in the ground rubber. Rubber powders typically contain process oils, rubber hydrocarbons, carbon black and inorganic fillers. The experiments were conducted using a Perkin Elmer Pyris-1 TGA analyzer. A typical cycle consists of heating from 30°C to 530°C under a nitrogen blanket at 10°C/min. The cycle is held for 10 minutes at 530°C. The gas is switched from nitrogen to oxygen. The sample is then further heated to 850°C at 10°C/min. This method enables quantification of all the constituents in a single TGA cycle.
Scanning electron microscopy (SEM)
The morphology of the samples was observed using a Zeiss scanning electron microscope. The samples were gold coated to avoid charging.
The sieve analysis of all samples was done using Tyler equivalent meshes with a portable Rotap sieve shaker per ASTM D- 5603-01. Typically about 100 gm of the sample were placed in the Rotap and shaken for 10 minutes. The weight of the particles on each screen were measured and percentage of weight retained on the screen was determined to obtain the particle size distribution.
Surface area measurement
The surface area of the samples was determined using the multipoint Krypton ‘BET’ adsorption isotherm method. The experiments were conducted on Quantachrome’s Autosorb-1. The test was run three times and an average value was taken for interpretation.
Electron spectroscopy for chemical analysis (XPS/ESCA) XPS spectra were obtained by using a Thermo-K-Alpha spectrometer. The x-ray was the monochromated Al K-alpha line (1486 eV). The analyzer pass energy was typically 50 eV with 200 watts x-ray power (spot size 200 µm). The vacuum was recorded as 1 x 10-7 mbar. The charge neutralization was active during the acquisition of data. The samples were used as produced without any pre-treatment. All the experiments were done at room temperature. The binding energy assignments for particular peaks were made using known literature.
Characterization of rubber powders
The thermogravimetric charts representing percent weight loss with temperature are shown in figure 1. Table 2 represents the breakdown of the constituents in the samples. The weight loss around 300°C is considered to be due to low molecular weight additives and process oil in the rubber sample. The polymer undergoes thermal degradation, resulting in weight loss of approximately 55% up to 530°C. The step loss seen on the graph at around 600°C was attributed to the carbon black in the rubber sample. From the table, it was consistently in the range of 30%.
The remaining residue was due to the inorganic fillers, such as silica, in the sample. All the samples showed approximately 5.5 to 6% of ash content except the MD50 sample. The weight loss for MD50 is pushed upward by the amount of additional ash content compared to the rest of the samples. This was attributed to the processing aid added during the Rotap particle size analysis for avoiding agglomeration of the rubber particles. The size range of the processing aid is below 10 µm, which allowed it to pass through all the larger size screens.
TGA results of MD400 and Amb400 samples were about the same.
Particle size distributions
The particle size distribution of the powders prepared by ambient and cryo grinding was examined in their “as produced” form. Figure 2a compares the size distributions of rubber samples having nominal particle size of 400 microns (40 mesh).
Clearly, the distributions depend on the production plant’s mode of grinding and the sieving technique. It was observed that the cryogenically ground powder had a larger amount of fine particles than the corresponding size ambient grind powder. In the cryogenic method, freezing the rubber below its embrittlement point followed by fracturing in a high speed turbo mill evidently produces a larger amount of fines. This suggested that the cryogenic distributions should have higher surface area as compared to the ambient distributions of the same nominal mesh size.
Figure 2b compares GTR samples produced by the cryogenic method. The cryogenic turbo mill technology produces fine rubber particles below 50 ìm. The chart compares particle size distributions for nominal 400 ìm (40 mesh), 180 ìm (80 mesh), 105 ìm (140 mesh) and 50 ìm (300 mesh) size particles. For all the powder samples, large amounts of fines were observed relative to their nominal mesh sizes. The distribution curve was found to depend on many processing variables that included nitrogen ratio, milling rpm, mill gap and sieving technique. It was seen that by adjusting the parameters, the particle size distribution can be controlled as needed. The size distribution data essentially suggested very high surface area particles at 50 ìm as compared to 400 ìm GTR.
Surface area measurements
This study was focused on understanding the effects of the grinding process and of the surface areas of the particles on GTR/PP composites. A systematic investigation has been made of the effect of the grinding process on the surface area of the particles produced. All the samples were as produced commercially, the distribution curves of which are shown in figure 2. The surface area of all the GTR samples was determined using a multipoint Krypton BET adsorption isotherm. The experiments were repeated to obtain consistent results. Previously, it has been reported that the ambient-ground rubber had somewhat higher surface area than the cryo-ground rubber for the same nominal particle size (ref. 9). This was thought to be a reflection of the porous or rough morphology of ambient ground materials. In some instances, the higher surface area was believed to be due to higher fiber content in the ambient samples.
Figure 3 shows a comparison of cryogenic grind and ambient grind particles. For reference, a model based on spherical particles was also included. Both ambient and cryogenic particles showed high surface areas relative to the spherical particle model. The cryogenic particles showed two orders of magnitude increase in surface area from MD400 to MD50. It was observed that the cryogenic particles had higher surface area than the ambient particles at all the nominal particle sizes. This was contrary to previous observations and arguments about cryogenic rubber powders (refs. 1 and 9). Comparing ambient and cryo-grind powders, the surface area of cryogenic powders increased by a factor of 1.2x at 400 microns to about 2 to 2.2x for 180 and 200 micron particles, respectively (table 3). The observations correlate with higher content of fines in the cryogenic powders.
Morphology of rubber particles
The surface morphology of the rubber particles is expected to depend on the method of production. The size, shape and morphological features on the particle surface can affect the properties of the polymer/GTR composites. The differences in surfaces of cryogenic and ambient grind rubber particles were studied by scanning electron microscopy.
The effect of the grinding process on particle morphology is compared in figure 4. For simplicity, as-produced rubber samples of 400 µm nominal size are shown. Figures 4a to c show surfaces at low and high magnifications for MD400; while d to f represent corresponding surfaces for Amb400 rubber. Figure 4a showed distribution of cryo particles with size scales in the range of 200 to 400 µm. Larger ambient particles in the size scale of 400 µm were observed in figure 4d. The cryogenic particles largely demonstrated smooth surface morphology. The ambient particles showed two distinct morphologies; in the first, the surface was covered in porous, rough nodules, and in another, the surface was smooth. Previously, it was reported that the ambient particles exclusively form porous, rough texture (refs. 4 and 10). However, our observations of Amb400 and other ambient and wet-ambient grind samples indicated that it consisted of both types of morphology, along with the presence of fibers. From figure 4d, the composition of the two textures appeared to be in equal proportions. Figure 4e shows the texture of one of the smooth ambient particles, it appears very similar to that of the corresponding cryogenic particle image in figure 4b. Further magnification of about 50KX showed that the details on the surface of both ambient and cryogenic smooth particles look about the same. The morphology of the rough ambient particles was comparable with that reported in the literature (refs. 4 and 10).
It was claimed before that ambient grinding gives higher surface area, attributing to the rough texture as compared to cryogenic grinding (ref. 4). Additionally, it was suggested that the rougher texture may nullify the effect of more fines in cryogenic powders (ref. 9). As observed previously in table 3, the surface areas of cryogenic particles were consistently higher than the ambient grind particles. Also, we observed that not all particles are rough in ambient grind as claimed; but it consisted of a significant amount of smooth particles as well.
Chemical analysis of the micronized rubber powders
It is known that polymer-filler interfacial adhesion is a key factor in improving the composites’ mechanical properties. Very little is known about surface chemistry and morphology of GTR particles. X-ray photoelectron spectroscopy (XPS) of the rubber particles was performed for determining the surface elemental compositions. The study investigated for the presence of any oxidized species and/or unsaturation that can potentially react with the thermoplastics and/or compatibilizers. Figure 5 compares the XPS spectra of MD400 and Amb400 rubber samples. The major peaks observed in both the spectra were attributed to the following elements:
• Si peaks - at 103 and 153 eV;
• C1s peak - 285 eV;
• O1s peak - 532 eV; and
• Zn peaks - 1,022 and 1,045 eV.
The XPS of other cryogenic particles revealed similar spectra. The relative intensities of different atoms present on the rubber surface are summarized in table 4. Ambient particles showed the presence of somewhat more oxygen and less silica as compared to cryogenic particles.
Interestingly, no sulfur was observed on the surface. Previous researchers noted the presence of about 0.34-0.66 atomic wt. % sulfur on ambient GTR particles (ref. 4).
Processing of rubber compounds with MRP
A designed experiment was conducted on a standard SBR/ BR tread compound to understand the effects of MRP addition. In the design, MRP content was varied from 2-14%, accelerator (N-tert-butyl-2 benzothioazolesulphenamide, BBTS) was varied from 0.2-1.4 phr, and sulfur content was varied from 1.5-4 phr. The processing was a standard two pass mix, with the MRP added in the masterbatch along with the carbon black. The MRP was Lehigh PD80, an -80 mesh micronized rubber powder made via cryogenic turbo-milling of rubber from end-of-life radial truck tires. Tensile plaques were cured at 160ºC for 20 minutes. Cure times were measured @ 160ºC on an Alpha Technologies Rheometer MDR 2000 per ASTM D5289. Tensile properties were measured per ASTM D412.
A sample of the process dependence is shown in figure 6. MRP content is on the x-axis and BBTS accelerator content on the y-axis. The iso-lines shown are the cure times (T10 – time to 10% cure) in minutes. It is clear that increasing the MRP content at constant accelerator levels results in reduced cure times. Alternatively, constant cure rates can be achieved with increasing MRP content by reducing the amount of accelerator added. This effect is consistent with the XPS results that indicate the presence of accelerator compounds (N) on the MRP surface.
The property dependence is shown in figure 7. MRP content is on the x-axis and sulfur content on the y-axis. In this case, the iso-lines shown are for the 300% modulus in MPa. In this case, increasing the MRP content at constant sulfur has the effect of depressing modulus. Alternatively, increasing the S content with increasing MRP content can be done to maintain modulus. These results are consistent with the XPS results that did not indicate the presence of sulfur on the MRP surface.
Cost/performance benefits in rubber compounds
As stated previously, the cost of tire tread compound is significantly higher than MRP such as MD105. Replacing part of the virgin rubber in the rubber compound without sacrificing the necessary performance is the key to developing new MRP formulations with sustainability. Figure 8 shows the effect of MD105 substitution at 3, 6 and 10 weight percent on tensile strength against tensile elongation normalized to the control rubber compound without MRP. It was observed that the fine rubber particles improved the strength and elongation of the control compound at 3% loading. With increasing MRP composition, the strength was within 95% of the control compound and the elongation was about the same. This is well within the acceptable standards for a typical tire tread compound.
Micronized rubber powder is a valuable resource for improving environmental quality, reducing dependence on imported raw materials and hedging volatile commodities costs. Its successful usage depends on a proper understanding of the surface science of MRP and how that affects the processing and properties of the materials in which it is used. The MRP materials in this study have a cost of approximately half that of prevailing virgin rubber compound prices. The tensile strength and elongation of 3% MD105 exceeded the control tread compound and at higher loading of up to 10%, it was comparable to that of the control. Thus, proper understanding and utilization of micronized rubber powders can result in significant economic and environmental benefits.