Elsevier

Polymer

Volume 49, Issue 16, 28 July 2008, Pages 3500-3509
Polymer

Influence of twin-screw extrusion conditions on the dispersion of multi-walled carbon nanotubes in a poly(lactic acid) matrix

https://doi.org/10.1016/j.polymer.2008.06.010Get rights and content

Abstract

Twin-screw extrusion using a co-rotating Berstorff ZE25 extruder was applied to disperse multi-walled carbon nanotubes (MWNT) in poly(lactic acid) (PLA). The masterbatch dilution technique was used whereas four different masterbatches were produced under variation of MWNT content, screw profile, temperature profile, and rotation speed which then were diluted to composites with 0.75 wt% MWNT under varied process conditions. The state of dispersion was investigated by light microscopy from which a dispersion index was quantified. Transmission electron microscopy was performed to observe the MWNT dispersion and network formation in the sub-micron scale.

The state of MWNT dispersion within the diluted composites was predominated by the state of filler dispersion in the masterbatches. High rotation speed (500 rpm) that still ensures a certain residence time of the melt combined with a screw profile containing mainly mixing elements were found to be highly convenient to disperse and distribute the MWNT in the PLA matrix as well during masterbatch production as the dilution step. The temperature profile showed less influence, however, an increasing profile resulted in slightly better nanotube dispersions. By means of these processing conditions a percolation set was performed indicating an electrical percolation threshold below 0.5 wt% MWNT content as measured on compression molded samples.

Introduction

The outstanding performance of carbon nanotubes (CNT) [1], [2], [3], which combines unique electrical [4], [5], [6], [7], [8], [9] and thermal properties [10], [11], [12], [13] with high mechanical strength [14], [15], [16], [17], [18] predestines them as filler for polymer matrices. These new composite materials can be used in a wide range of industrial applications in the fields of electrostatic dissipation [19], [20], electromagnetic interference- shielding [21], [22], [23], and electrically conductive materials achieving at the same time enhanced stiffness, strength, impact properties, thermal stability, tribological properties, and reduced thermal expansion [24].

To pave the way for these applications of polymer/CNT composites, it is important to develop processing technologies, which are compatible to already available industrial technologies. For this reason it is essential to reveal the relations between processing conditions and properties of melt mixed composites, wherefore this work contributes. Most of the previous studies on melt mixing of CNT into polymers were done using small-scale mixing devices in the gram scale and results obtained may not be transferable directly towards large-scale processes [25], [26]. Even if large-scale melt mixing of nanotubes into thermoplastic polymers is reported in literature [27], [28], [29] no systematic investigations on the influence of melt mixing conditions on nanotube dispersion could be found. Therefore, the influence of extrusion conditions using a twin-screw extruder on the dispersion of multi-walled carbon nanotubes (MWNT) in a poly(lactic acid) (PLA) matrix is investigated in this study with the aim to develop a guideline for plastic fabricators.

The key challenge for the successful implementation of applications based on polymer/CNT composites is a suitable distribution and dispersion of the filler inside the polymer matrix to ensure low percolation thresholds combined with high mechanical performance. Remaining primary agglomerates of the nanotubes not only reduce the nanotube amount available for electrical percolation, but also act as imperfections in mechanical tests. Therefore, the CNT material has to be individualised during the melt mixing process and for electrical conductivity they finally have to form a network, where the distance of neighbouring tubes is smaller than the maximum tunnelling distance of electrons, which is reported to be about 1.8 nm [30]. This objective is hindered by the intrinsic properties of the raw MWNT materials, which are characterised by the presence of compact primary agglomerates (up to diameters of several millimetres) often combined with physical entanglements of the spaghetti like flexible MWNT, which result from the synthesis process as well as the high van-der-Waals forces between the tubes.

High shear forces during the melt mixing process and relatively long processing times are found to be suitable for the successful individualisation of MWNT, as reported for different systems [25], [26], [27], [28], [29], [31], [32], [33], [34], [35], [36]. Pötschke et al. [36] discussed for small-scale mixing the influence of rotation speed on the state of MWNT dispersion and electrical and dielectric properties for concentrations below and above the percolation threshold. At MWNT contents below the percolation threshold, increased screw speed resulted in better dispersion, whereas above the percolation threshold a decrease of DC conductivity was observed, which could be explained only with an enhanced breakage of the MWNT. Furthermore it is reported that an increase of mixing time improved the state of MWNT dispersion significantly for all MWNT concentrations. Similar results were reported by Takase [37] who showed that increase in rotation speed of a twin-screw extruder decreased the agglomerate size.

These practical results are comparable to findings on the dispersion kinetics of carbon black (CB) in non-Newtonian fluids. It is reported that dependent on the shear stress level either CB agglomerates are ruptured into smaller agglomerates or erosion from their surfaces takes place. The erosion process is more gradual and takes place at lower shear stresses [38], [39], however, both the processes occur simultaneously. The work of Li et al. [40] showed that the dispersibility of CB agglomerates decreases with increasing aggregate structuring, which may be assumed also for CNT agglomerates, even if CNT and CB are quite different filler materials and may not be directly comparable.

In case of CNT a certain decrease of the aspect ratio during melt mixing may improve the dispersibility due to the decreasing number of physical entanglements as indicated by investigations of Andrews et al. [35]. The authors reported that during melt mixing of MWNT with polystyrene in a laboratory mixer a shortening of MWNT occurred with mixing time down to a third of the origin length, despite the high elasticity and flexibility of the nanotubes. At the same time a better MWNT dispersion was generated with increasing mixing energy input.

For industrial applications of melt mixing extrusion technique the two-step masterbatch dilution seems to be favourable as compared to the direct nanotube incorporation. Here, in the first step a concentrate of polymer containing high CNT contents is produced which is diluted in the second step with the same polymer in order to get well dispersed low amounts of nanotubes within the matrix. Besides the fact that high nanotube contents in the masterbatches resulting in high melt viscosity lead to high shear stresses during melt mixing, the handling of polymer bonded CNT is much easier for fabricators. In addition, the accurate dosage of small CNT amounts is easily achievable. Such masterbatch dilution processes are described in literature to be an appropriate technique to disperse and distribute CNT in polymer melts [20], [31], [32], [33], [41], [42] whereas in these examples industrially available masterbatches were used. Therefore, the masterbatch dilution technique was applied in our work. In contrast to microcompounders that have a fixed assembly, a twin-screw extruder has a modular assembly, which can be varied in the diameter/length ratio of the screw, the screw profile itself, and the feeding positions. In this contribution, especially the screw profile, the temperature profile, and the rotation speed were varied systematically as well during masterbatch production as during the dilution process to determine their impact on the MWNT dispersion and distribution within the matrix.

PLA was selected as matrix polymer exemplarily. This type of polymer has generated great interest as one of the most innovative materials being developed for a wide range of applications. This polymer is thermoplastic and biodegradable, which makes it highly attractive for biological and medical applications. Especially in the field of tissue engineering PLA was found to be one of the most favourable matrix materials to create high performance fibres. Some reports were found on PLA/CNT composites. Moon et al. [43] reported about mechanical, thermal, and electrical properties of composites containing MWNT, which were processed using solution technique. Besides an increase of Young's modulus an improvement in electromagnetic wave shielding effectiveness was observed by adding CNT. Wu and Liao [44] presented a study on PLA with MWNT where melt mixing was applied on acrylic-grafted PLA and chemically modified MWNT. In this paper, thermal and mechanical characterisation of the composites was in focus next to the proof of the expected chemical reaction whereas the state of nanotube dispersion and electrical conductivity were not regarded. Furthermore, McCullen et al. [45] reported about PLA/MWNT fibres, which were produced by electro spinning from a PLA solution to develop a scaffold for tissue engineering. Electrical percolation of fibre mats was reached at about 0.3 wt% CNT. For good spinning ability in electro spinning or melt spinning a good dispersion and distribution of the MWNT are needed to ensure a high mechanical level of the fibres and a certain melt strength required for the formation of the fibres.

Section snippets

Materials

PLA Biomer® L9000 (Biomer, Germany) is based on poly-l-lactic acid produced by polymerisation of naturally occurring l-lactate and represents a semi-crystalline and biodegradable polymer. Chemically, PLA is an alpha hydroxyl ester, which tends to hydrolyse (degrade). This hydrolysis process is fast in humid conditions at temperatures above the glass transition temperature (Tg  55 °C). Due to this fact, the material was dried at least for 8 h at 40 °C in a vacuum oven before each processing step.

PLA

Influence of extrusion conditions on the dispersion of MWNT in PLA based masterbatches

As MWNT contents in the masterbatches 15.0 wt% and 7.5 wt% were selected. The content of 15.0 wt% follows the amount, which is typically incorporated in industrial available masterbatches, i.e. by Hyperion Catalysis International Inc. (Cambridge, USA). The amount was then reduced to 7.5 wt% since the torque during extrusion was near the limit using the 15.0 wt% and a further increase in temperature was not possible due to expected PLA degradation. After comparing both masterbatches under otherwise

Summary and conclusions

The work demonstrates how processing conditions during the twin-screw extrusion process as an industrially relevant melt mixing process influence the distribution and dispersion of multi-walled carbon nanotubes in a poly(lactic acid) matrix. A masterbatch dilution technique was applied.

For the four masterbatches produced under variation of MWNT loading, temperature profile, screw profile, and rotation speed dispersion indices ranging between 13.3% and 94.3% and agglomerate numbers between 4.1

Acknowledgement

We gratefully acknowledge financial support from the project INTELTEX (Intelligent multi-reactive textiles integrating nanofiller based CPC-fibres) – a European Integrated Project supported through the Sixth Framework Programme for Research and Technological development. In addition, we thank Nanocyl S.A. for providing the nanotubes within the INTELTEX project.

References (52)

  • S. Iijima

    Materials Science and Engineering B – Solid State Materials for Advanced Technology

    (1993)
  • R.S. Ruoff et al.

    Carbon

    (1995)
  • B.G. Demczyk et al.

    Materials Science and Engineering A – Structural Materials Properties Microstructure and Processing

    (2002)
  • T. Villmow et al.

    Composites Science and Technology

    (2008)
  • O. Meincke et al.

    Polymer

    (2004)
  • B. Schartel et al.

    European Polymer Journal

    (2005)
  • P. Pötschke et al.

    Polymer

    (2002)
  • P. Pötschke et al.

    European Polymer Journal

    (2004)
  • P. Pötschke et al.

    Polymer

    (2003)
  • P. Pötschke et al.

    Polymer

    (2005)
  • C.-S. Wu et al.

    Polymer

    (2007)
  • Y. Wang et al.

    Polymer

    (2008)
  • S. Pegel et al.

    Polymer

    (2008)
  • B. Kalb et al.

    Polymer

    (1980)
  • A.R. Bhattacharyya et al.

    Polymer

    (2003)
  • K.P. Ryan et al.

    Composites Science and Technology

    (2007)
  • S. Iijima

    Nature

    (1991)
  • S. Iijima et al.

    Nature

    (1992)
  • J.C. Charlier et al.

    Physical Review Letters

    (1993)
  • R. Saito et al.

    Journal of Applied Physics

    (1993)
  • C.T. White et al.

    Physical Review B

    (1993)
  • M.S. Dresselhaus et al.

    European Physical Journal D

    (1999)
  • T.W. Ebbesen et al.

    Nature

    (1996)
  • H.J. Dai et al.

    Science

    (1996)
  • J. Hone et al.

    Applied Physics A – Materials Science and Processing

    (2002)
  • S. Berber et al.

    Physical Review Letters

    (2000)
  • Cited by (389)

    • Uranium adsorption property of carboxylated tubular carbon nanofibers enhanced chitosan microspheres

      2023, Fabrication and Functionalization of Advanced Tubular Nanofibers and their Applications
    View all citing articles on Scopus
    View full text