Scalable route to well-dispersed polyolefin/carbon nanotube composites

21 April 2010
Edward Davis, Vinod Radhakrishnan, and Virginia Davis
The initial step in blending polypropylene with unmodified and surface-functionalized single-walled carbon nanotubes has a significant effect on final dispersion and thermal properties.

Single-wall carbon nanotubes (SWNTs) are seamless carbon cylinders with diameters on the order of 1nm and lengths of a few hundred nanometers to a few microns. Their Young's modulus (a measure of elasticity) is close to 1.25TPa (terapascals), tensile strength over 37GPa (gigapascals), and they have a low density of ~1.5g/cm3. In other words, they have over 60 times the strength of steel at just one-sixth the weight.1 Translating the remarkable properties of pure carbon nanotubes into improved properties in thermoplastic polymer nanocomposites requires that the nanotubes be well dispersed. However, as they exist in tightly packed bundles held together by strong attractive interactions, melt processing alone does not result in sufficient dispersion.

Significant effort has gone into developing methods to fully disperse SWNTs in a polymer matrix. Techniques that have been evaluated include hot coagulation (a form of solution mixing), spray-coating polymer powder or pellets, and dry mixing prior to melt processing. Dry mixing is the simplest and most economical method but has failed to produce high levels of dispersion. Both hot coagulation and spray coating have shown promising results, but it has been difficult to compare them since previous studies have all used different polymers, nanotubes, and processing conditions.

We have directly compared the efficacy of three preblending processes followed by melt compounding for the dispersion of carbon nanotubes in polypropylene (PP). The three preblending methods examined were dry blending, hot coagulation,2,3 and a modification of the spray-drying process4 that we call rotary evaporation. In addition, we have compared the effects of these processes on both pristine SWNTs and SWNTs that have been functionalized by attaching dodecyl (12-carbon-atom-long) groups to their sidewalls (C12SWNTs). Functionalization was expected to improve dispersion. The use of a single polymer, standardized processing conditions, and SWNTs from a single production batch have allowed us to directly compare the effects of the preblending process on the degree of dispersion.

Dry blending was accomplished by shaking the nanomaterial and polymer together. Hot coagulation was performed with 1, 2-dichlorobenzene as the solvent. A rotary evaporator was used to mimic the spraying additives into tumbling dryers such as those already in commercial polymer-manufacturing plants. The main advantage of this process over previous studies (looking at the spraying of SWNT dispersions onto polymer prior to melt compounding) is that in this case the polymer was continually agitated. Figure 1 shows the distinct differences in the product of these methods prior to melt compounding, which was conducted on all materials to produce the final nanocomposite.

(a) Dry-mixed, (b) rotary-evaporation, and (c) hot-coagulation samples prior to melt extrusion. (Adapted with permission.5)

Optical microscopy images of melt-extruded nanocomposites: (a, d) PP/SWNTs and PP/C12SWNTs by dry mixing, (b, e) PP/SWNTs and PP/C12SWNTs by rotary evaporation, and (c, f) PP/SWNTs and PP/C12SWNTs by hot coagulation. (Adapted with permission.5)

The microstructure of the nanocomposites was evaluated by scanning electron microscopy, optical microscopy (see Figure 2), Raman spectroscopy mapping, and rheology (flow). While microscopy methods provide qualitative information, Raman spectroscopy and rheology provide quantitative information on microstructure and are not affected by sample opacity. All of these methods indicated that hot coagulation resulted in the best dispersion and dry mixing in the worst. However, the rheological results also indicated that hot coagulation resulted in some polymer degradation. Rotary evaporation, on the other hand, did not result in polymer degradation, and the dispersion was significantly better than for the dry-mixed samples. Surprisingly, for each process, the C12SWNTs were more poorly dispersed than equivalently processed SWNTs. This result is contrary to the expectation that surface modification should produce better dispersion.

Ultimately, the motivation for adding nanomaterials to polymers is to improve properties such as thermal stability. The thermal decomposition temperature of the composites was generally found to trend with the dispersion uniformity. The dry-mixed samples had decomposition temperatures similar to pure PP. However, both the hot-coagulation and rotary-evaporation nanocomposites showed significant increases of ~10°C.

This work represents the first direct comparison of three potential dispersion techniques—dry blending, hot coagulation, and rotary evaporation (a model process that mimics commercial additive spray drying)—each followed by melt extrusion. Morphological evaluation shows that the best dispersion is achieved from the hot-coagulation process, the poorest from dry blending, with the rotary evaporation method falling somewhere in between. However, the hot-coagulation process entails polymer degradation. In addition, scale-up of hot coagulation is challenging due to the large quantity of solvent required. Rotary evaporation on the other hand, could likely make use of existing commercial processes. We plan further work to characterize the effects of distributive and dispersive blending elements on an industrial-scale compounding line.


Edward Davis
Department of Polymer and Fiber Engineering, Auburn University

Edward W. Davis received his PhD from the University of Akron (1996). He worked in the commercial plastics industry for 11 years, including positions with Shell Chemicals in Louvain-la-Neuve (Belgium) and Evalca in Houston, TX. He joined the faculty at Auburn University in the fall of 2007.

Vinod Radhakrishnan
Department of Chemical Engineering, Auburn University

Vinod K. Radhakrishnan is a graduate student in Virginia A. Davis's group at Auburn University. He received his BE first class with distinction in chemical engineering from Bangalore University (India) in 2001, and his MS in chemical engineering from the Texas A & M University, Kingsville (2005).

Virginia Davis
Department of Chemical Engineering, Auburn University

Virginia A. Davis is an assistant professor. She has 11 years of experience in the polymers industry. She received her PhD in Chemical Engineering from Rice University (2006).


  1. R. H. Baughman, A. A. Zakhidov and W. A. de Heer, Carbon nanotubes—the route toward applications, Science 297 (5582), pp. 787-792, 2002.

  2. R. Haggenmueller, J. E. Fischer and K. I. Winey, Single wall carbon nanotube/polyethylene nanocomposites: nucleating and templating polyethylene crystallites, Macromolecules 39 (8), pp. 2964-2971, 2006.

  3. F. Du, J. E. Fischer and K. I. Winey, Coagulation method for preparing single-walled carbon nanotube/poly(methyl methacrylate) composites and their modulus, electrical conductivity, and thermal stability, J. Polym. Sci., Part B: Polym. Phys. 41 (24), pp. 3333-3338, 2003.

  4. Q. H. Zhang, S. Rastogi, D. J. Chen, D. Lippits and P. J. Lemstra, Low percolation threshold in single-walled carbon nanotube/high density polyethylene composites prepared by melt processing technique, Carbon 44 (4), pp. 778-785, 2006.

  5. V. K. Radhakrishnan, E. W. Davis and V. A. Davis, Influence of initial mixing methods on melt-extruded single-walled carbon nanotube-polypropylene nanocomposites, Poly. Eng. Sci., in press.

DOI:  10.2417/spepro.002910