CARBON
4 7 ( 2 0 0 9 ) 2 –2 2
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/carbon
A review of vapor grown carbon nanofiber/polymer conductive composites Mohammed H. Al-Saleh, Uttandaraman Sundararaj* Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2G6
A R T I C L E I N F O
A B S T R A C T
Article history:
Vapor grown carbon nanofiber (VGCNF)/polymer conductive composites are elegant materi-
Received 9 April 2008
als that exhibit superior electrical, electromagnetic interference (EMI) shielding effective-
Accepted 17 September 2008
ness (SE) and thermal properties compared to conventional conductive polymer
Available online 23 September 2008
composites. This article reviews recent developments in VGCNF/polymer conductive composites. The article starts with a concise and general background about VGCNF production, applications, structure, dimension, and electrical, thermal and mechanical properties. Next composites of VGCNF/polymer are discussed. Composite electrical, EMI SE and thermal properties are elaborated in terms of nanofibers dispersion, distribution and aspect ratio. Special emphasis is paid to dispersion of nanofibers by melt mixing. Influence of other processing methods such as in-situ polymerization, spinning, and solution processing on final properties of VGCNF/polymer composite is also reviewed. We present properties of CNTs and CFs, which are competitive fillers to VGCNFs, and the most significant properties of their composites compared to those of VGCNF/polymer composites. At the conclusion of the article, we summarize the most significant achievements and address the future challenges and tasks in the area related to characterizing VGCNF aspect ratio and dispersion, determining the influence of processing methods and conditions on VGCNF/polymer composites and understanding the structure/property relationship in VGCNF/polymer composites. 2008 Elsevier Ltd. All rights reserved.
Contents 1.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1. Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.2. VGCNF properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Abbreviations: ABS, acrylonitrile–butadiene–styrene; ASI, Applied Sciences Inc.; CB, carbon black; CF, carbon fiber; CNT, carbon nanotubes; CNF, carbon nanofiber; CPC, conductive polymer composites; CTE, coefficient of thermal expansion; CuNW, copper nanowires; EM, electromagnetic; EMI, electromagnetic interference; ESD, electrostatic discharge; GF, glass fiber; HDPE, high density polyethylene; HIPS, high impact polystyrene; HRTEM, high resolution transmission electron microscopy; I–V, current–voltage; VGCF, vapor grown carbon fiber; VGCNF, vapor grown carbon nanofibers; LCP, liquid crystal polymer; MCP, metal coated polymer; MNW, metal nanowire; MWNT, multi-walled carbon nanotubes; PE, polyethylene; PES, poly(ether sulfone); PMMA, poly(methyl methacrylate); PP, polypropylene; PS, polystyrene; PVA, poly(vinyl alcohol); PVDF, poly(vinylidene fluoride); SE, shielding effectiveness; SEM, scanning electron microscope; SMP, shape memory polymer; SS, stainless steel; s-VGCNF, short-vapor grown carbon nanofiber; SWNT, single wall carbon nanotube; TEM, transmission electron microscopy; VE, vinyl ester. * Corresponding author: Fax: +1 780 492 2881. E-mail address:
[email protected] (U. Sundararaj). 0008-6223/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2008.09.039
CARBON
2.
3. 4. 5.
1.
4 7 ( 2 0 0 9 ) 2 –2 2
Electrical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1. Effect of VGCNF surface chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2. Effect of polymer type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3. Effect of processing method and processing conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3.1. Composites by melt mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3.2. Composites by solution processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.3.3. Composites by in-situ polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3.4. Composite by heterocoagulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.4. Effect of VGCNF alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.5. Effect of HDPE adsorption on VGCNF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.6. Effect of volume exclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.7. Effect of using immiscible multiphase matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Shielding effectiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Thermal properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Summary and conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Introduction
In the last few years, conductive nanofiller/polymer composites have been widely investigated in academia and industry because of their outstanding multifunctional properties compared to conventional conductive polymer composites (CPCs) [1–3]. Those composites have been mainly formulated using high aspect ratio 1D conductive nanofillers including carbon nanotubes (CNTs) [4–18], vapor grown carbon nanofibers (VGCNFs) [19–28] and more recently metal nanowires (MNWs) [29–31]. This article is intended to review in depth the progress and accomplishments achieved to date in the VGCNF/ polymer conductive composites field in terms of nanofiber characterization and composite compounding and properties. VGCNF/polymer composites were recently concisely reviewed by Tibbetts and coworker [20]. The review covered a portion of the contributions published to date and did not go into detail about thermal, electrical and electromagnetic interference (EMI) shielding effectiveness (SE) properties of the VGCNF/ polymer composites. It is not the purpose of this work to review progress in the CNT/polymer composites field. However, we have summarized some of the most significant contributions attained using CNTs for purposes of comparison or to add more knowledge that might be transferable to VGCNFs. Interested readers in CNT/polymer composites are advised to refer to a general review by Thostenson and coworkers [32], CNT/polymer composites reviews Breuer and Sundararaj [1] and by Moniruzzaman and Winey [4], reviews of mechanical reinforcement of CNTs by Coleman and coworkers [33,34] and by Miyagawa and coworkers [5], review of dispersion and alignment of CNTs in polymer by Xie and coworkers [35], CNT/polymer fibers by Ciselli and coworkers [36] and reviews of polymer grafted-carbon nanofillers by Tsubokawa [37] and Homenick and coworkers [38]. Compared to CNTs, VGCNFs have received less research attention as nanofillers because CNTs have better mechanical properties (due to less microstructural defects [39]), smaller diameter and lower density than VGCNFs. However, because of their availability and relatively low price, VGCNFs are an excellent alternative for, CNTs and in addition, VGCNFs could 1
3
be used for research purposes to build knowledge that might be transferable to the more expensive CNTs. Multi-walled carbon nanotubes (MWNTs) are 2–3 times more expensive than VGCNFs and single wall carbon nanotubes (SWNTs) are even more expensive. In 2007, the prices of VGCNFs, MWNTs, and 90% pure SWNTs were $125/lb, $350/lb and $30,000/lb, respectively1 [40]. It is forecasted that due to increase in production capacities, VGCNF price might drop significantly [41]. For electrical applications, VGCNFs are also competitive fillers with carbon fibers (CFs) and high structure carbon black (CB), owing to the lower loading of VGCNFs compared to CFs and CB required to achieve certain electrical conductivities. VGCNFs are produced by catalytic chemical vapor deposition of a hydrocarbon (such as natural gas, propane, acetylene, benzene, ethylene, etc.) or carbon monoxide over a surface of a metal (Fe, Ni, Au, Co) or metal alloy (such as Ni–Cu, Fe–Ni) catalyst [25,42–45]. The catalyst can be deposited on a substrate or directly fed with the gas phase [19,46]. The reaction is usually carried out in reactor operated at a temperature of 500–1500 C [19]. Pyrograf III nanofibers (Applied Sciences Inc. (ASI), Ohio, USA) have been the most widely investigated VGCNFs. They are produced in a gas phase reactor operated at 1100 C. The feedstocks for the reactor are natural gas, metal catalyst (produced by decomposition of (Fe(CO)5), hydrogen sulfide (used to disperse and activate the catalyst) and ammonia. The nanofibers are produced by decomposing of hydrocarbon on the metal catalyst. The decomposition both nucleates and grows the nanofibers [47]. The residence time of carbon in the reactor is only few milliseconds. Nanofibers produced by this method have a volume resistivity of about 4 · 103 X cm. However, this resistivity can be further decreased by graphitization [20]. Depending on the feedstock, catalyst and operating conditions different morphologies and characteristics of VGCNFs were obtained. For example, Lee and coworkers [25] synthesized VGCNF using different types of catalysts (nickel–copper and pure nickel) and different feedstocks (propane, ethylene and acetylene). VGCNF produced using propane was linear; whereas, VGCNF produced using ethylene resulted in a twisted conformation. For VGCNF produced using acetylene,
[cited 2007 December 31]; Available from: http://www.cheaptubesinc.com
4
CARBON
4 7 ( 2 0 0 9 ) 2 –2 2
both twisted and helical confirmations were observed. Surface area and electrical conductivity of VGCNF were also found to depend on the catalyst and feedstock types. VGCNF produced using propane feedstock and nickel–copper as catalyst had the highest surface area (348 m2/cm). However, the highest conductivity was observed for VGCNF produced using ethylene as a feedstock and pure nickel as a catalyst (28.7 S/ cm).
1.1.
Applications
EMI shielding and electrostatic discharge (ESD) protection are the major applications for conductive polymer composites [48–51]. Surface/volume resistivity of the filled polymer determines its application range. Polymer composites used for EMI shielding applications typically have a surface resistivity lower than 10 X/sq, whereas for the ESD applications, the optimum surface resistivity range is 106–109 X/sq.2 Fig. 1 shows surface resistivity ranges for plastic, antistatic, static dissipative, conductive and metal materials. Currently, metal coated polymers (MCP) and carbon black (CB)-filled polymers are used for EMI shielding and ESD protection, respectively [52–55]. These materials have many disadvantages and limitations. For example, MCP can delaminate and are difficult to recycle. In some cases, there are hidden costs in coating applications that increase the production cost [54,55]. Likewise, for CB-filled polymers, the high concentration of the CB required to achieve good electrical properties reduces certain composite mechanical properties and ease of processing, while increasing the cost [56]. In addition, for CB powders, sloughing is an environmental concern and could damage the packaged electronics [57–59]. Conductive VGCNF/polymer composites can be used as sensors for organic vapors [60]. The mechanism is based on changing the composite conductivity when it is exposed to an organic vapor because of swelling of the polymer matrix. Organic vapors can be distinguished based on the composite conductivity after certain time of exposure. A good sensor is the one that can give different electrical conductivity for different vapors [60]. VGCNFs have potential applications in automotive industry that could lead to better quality, lower cost, less fuel consumption and lower environmental emissions. Those applications include: electrostatic painting of exterior panels, shielding of automotive electronics and addition of VGCNFs to tires to improve stiffness [61]. VGCNFs are promising materials for batteries, where multifunctional carbon materials are used as electrodes or as support materials [62]. For this application, carbon materials have many advantages over other materials such as metal oxides and sulfides in terms of cost, thermal and chemical stability, ease of formulation in various shapes and environmental impact [63]. In the addition to the conductivity related applications, polymers filled with VGCNFs have potential biological applications. VGCNFs are more attractive than SWNTs and MWNTs for applications require incorporation of biological components such as proteins and DNAs in the hollow core of the fiber, because they have much larger hollow core diam2
[cited 2005 July 25]; Available from: wwwesd.org
Fig. 1 – Classification of materials according to their surface resistivity and application ranges. (Note: [cited 2007 July]; Available from: http://www.rtpcompany.com/products/ conductive/index.htm.)
eter [64]. In addition to the above mentioned applications, VGCNFs could be used in many applications that previously required CB or conventional CFs.
1.2.
VGCNF properties
Dimension, structure, electrical, mechanical and thermal properties of the VGCNF depends on the production technique and post-treatment methods [65]. Table 1 summarizes some typical properties of VGCNFs, CFs, MWNTs and SWNTs. VGCNFs are high aspect ratio nanofibers. They have larger diameters than CNTs, but smaller than CFs. The length of VGCNFs is comparable to that of CNTs, but shorter than that of CFs [1,20]. VGCNFs are hollow core nanofibers comprise of a single layer, as shown in Fig. 2, or a double layer, as shown in Fig. 3, of graphite planes stacked parallel or at a certain angle from the fiber axis [70,71]. The stacked planes are nested with each other and have different structures including bamboolike, parallel and cup-stacked [32,70,72,73]. Fig. 4 is a high resolution transmission electron microscopy (HRTEM) image of a side-wall of a VGCNF (the inset is a schematic illustrating the structure of cup-stacked VGCNF). The nanofiber is clearly seen to have a hollow core surrounded by concentric cupstacked planes. This type of structure has a large number of reactive edges both inside and outside the nanofiber [74]. Parallel layers in single layer VGCNFs were also observed using HRTEM [70]. The d-spacing of the graphene sheets was reported as 0.34 nm (the same as that in MWNTs and graphite platelets). Uchida and coworkers [71] observed the two different morphologies of VGCNF graphite layers, namely: single layer and double layer. Fig. 5 depicts schematic sketches demonstrating the structure of a single layer VGCNF and a double layer VGCNF. The inset is a HRTEM showing the presence of loops of about 3–5 graphite sheets at the inner and outer
CARBON
5
4 7 ( 2 0 0 9 ) 2 –2 2
Table 1 – Typical properties of VGCNF, SWNT, MWNT and CF Property
VGCNFa
Diameter (nm) Length (lm) Aspect ratio Density (g/cm3) Thermal conductivity (W/m K) Electrical resistivity (X cm) Tensile strength (GPa) Tensile modulus (GPa)
50–200 50–100 250–2000 2 1950 1 · 104 2.92 240
SWNTb
MWNTb
0.6–1.8
5–50
100–10,000 1.3d 3000–6000f 1 · 103–1 · 104 50–500g 1500
100–10000 1.75e 3000–6000f 2 · 103–1 · 104 10–60g 1000
CFc 7300 3200 440 1.74 20 1.7 · 103 3.8 227
a From Refs. [1,20,62,66]. b From Ref. [2]. c Properties of Akzo Nobel Fortafil 243 PAN-based fibers [67]. d From Ref. [33]. e From Ref. [68]. f From Ref. [69]. g From Ref. [35].
Fig. 3 – TEM micrograph of a double layer VGCNF [70]. Fig. 2 – TEM micrograph of VGCNF [20].
surfaces of the single layer VGCNF. The inner and outer diameters of the single layer VGCNF are 25 and 60 nm, respectively, while for the double layer VGCNF, the inner and outer diameters are 20 and 83 nm, respectively. In addition of having double layers and larger diameter, the graphene planes in the outer layer of the double layer VGCNF were parallel to the fiber axis; whereas, both the single layer VGCNF and the inside layer of the double layer have a truncated cone morphology. Because of their high electrical conductivity, VGCNFs are favorable fillers to formulate CPCs. As-produced VGCNFs are usually covered with layers of amorphous carbon that degraded the conductivity of the nanofibers. Post treatment is required in order to remove those layers of less conductive carbon (increasing crystallinity). Endo and coworkers [62] 3 4
measured the volume electrical conductivity of short-VGCNF (s-VGCNF); 100–200 nm in diameter and 10–20 lm in length; using a four-point method. They found that the intrinsic volume resistivity of nanofibers decreased to 103 and 104 X cm by heat treating the fibers at 1200 C (carbonization) and 2800 C (graphitization), respectively. Compared to normal vapor grown carbon fiber (VGCF), which is several microns in diameter, VGCNFs have higher volume resistivity due to their lower crystallinity [62]. Volume resistivity of normal VGCF after graphitization is 6 · 105 X cm [66]. With a thermal conductivity value of 1950 W/(m K), VGCNFs have the highest thermal conductivity among all commercial CFs3,4 [41]. Experimentally measured thermal conductivity of MWNT is around 3000 W/(m K) [69]. This value is much lower than the 6600 W/(m K) theoretically estimated for SWNT using molecular dynamics simulations [69,75].
[cited 2008 April 4]; Available from: http://www.apsci.com/ngm-pyro1.html [cited 2008 April 04]; Available from: http://www.electrovac.com
6
CARBON
4 7 ( 2 0 0 9 ) 2 –2 2
No direct measurements of Pyrograf III nanofibers mechanical properties have been conducted. The 2.92 GPa tensile strength and 237 GPa tensile modulus usually used in literature are for 7.5 lm in diameter VGCF. Patton and coworkers [77] estimated the lower limit of the tensile modulus and tensile strength of Pyrograf III nanofibers based on the rule of mixtures and experimentally measured mechanical properties of 15.5 vol% VGCNF reinforced epoxy. Their calculations revealed that the lower limit of tensile modulus and tensile strength are in the range of 88–166 GPa and 1.7– 3.38 GPa, respectively.
2.
Fig. 4 – HRTEM image shows a side-wall of a VGCNF having a cup-stacked structure. The inset is a schematic illustrates the cup-stacked structure [74].
One of the major drawbacks of VGCNFs is their poor tensile properties compared to those of CNTs. Before talking about stress transfer, adhesion, etc., knowing intrinsic mechanical properties of VGCNFs is very crucial since it determines the ultimate properties that can be achieved when using them as mechanical reinforcing fillers. VGCNF and VGCF tensile properties depends on the fiber diameter [62,76]. For example, tensile strength of graphitized s-VGCNF having a diameter of 100 and 300 nm are 2.2 GPa and 1.77 GPa, respectively. The dependence of tensile properties on nanofiber diameter may reveal a change in nanofiber morphology (crystallinity, graphene planes orientation) with increasing diameter (for example, changing from single layer to double layer) and/or increase in defects with increasing fiber diameter.
Electrical properties
A critical filler loading must be incorporated to transfer the composite from the insulative state into the conductive state. At this critical concentration, which is known as the percolation threshold, the electrical conductivity of the composite suddenly increases by several orders of magnitude. Often, at the percolation threshold, the filler forms a continuous network inside the polymer matrix, and increasing in the filler loading further usually has little effect on the composite electrical resistivity, as shown in Fig. 6. However, if a remarkable decrease in the composite’s electrical resistivity is noticed with increasing the filler loading above the percolation threshold, this means that the three dimensional conductive network has not yet been formed at the percolation concentration, and thus the composite conductivity is due to tunneling in addition to direct contact between the particles. In some cases, tunneling could be the dominant mechanism. Tunneling conduction occurs when the distance between the filler particles are close enough, roughly less than 10 nm [78]. Investigating the current–voltage (I–V) relationship gives an indication whether the composite conductivity is due to tunneling or direct contact between the particles [79,80]. Linear I–V relationship, Ohm’s law, indicates that direct contact between the filler particles is the dominant conduction mechanism. However, tunneling mechanism is the dominant mechanism for composites characterized by power law I–V relation [80,81].
Fig. 5 – Schematic illustrate the structure of (a) a single layer VGCNF and (b) a double layer VGCNF (c) a HRTEM showing loops (two loops have been enclosed in white ellipse as a guide for the eye) on the side-wall of a single layer VGCNF [70].
CARBON
4 7 ( 2 0 0 9 ) 2 –2 2
7
Fig. 6 – Schematic sketch shows typical electrical resistivity as a function of filler loading of high aspect ratio filler/polymer system.
Fig. 7 – Schematic sketches show the effect of 1D filler on the conductivity of polymer composite.
A power law fit [82], as given in Eq. (1), derived from the percolation theory is often used to fit resistivity data at filler loading above the percolation threshold: q ¼ qo ðv vc Þt ð1Þ Here q is the composite electrical resistivity (X cm), qo is a scaling factor (X cm), v is the filler volume fraction, vc is the filler critical volume fraction at the percolation threshold
and t is the critical exponent related to the lattice dimensionality. For spherical particles in an insulating matrix, the t value is around 1.3 for a 2D system and about 1.9 for a 3D system [82]. Composites filled with fibers typically have a t value higher than 2 because of the higher aspect ratio of fibers compared to spheres. For VGCNF/polymer composites, a critical exponent (t) value of 3 were reported for several composites using different types of VGCNFs [66,83]. Similarly, t values
8
CARBON
4 7 ( 2 0 0 9 ) 2 –2 2
in the range of 3 were also reported for SWNT/polymer composites [84,85]. This indicates that VGCNF/polymer composites have similar dimensionality to that of CNT/polymer composites. More information about the percolation theory and other models developed to predict electrical resistivity and percolation threshold in composites of conductive filler in insulative matrices can be found in reviews by Weber and Kamal [82] and Lux [86]. When it comes to the electrical properties of VGCNFs or CNTs filled polymer composites, reducing the electrical percolation threshold is critical to make those composites economically feasible since VGCNFs and CNTs are much more expensive than the polymer matrix. Building a conductive network within an insulative matrix at lower concentration does not necessarily require well distributed filler. However, it does need well dispersed filler, as illustrated in Fig. 7. The figure shows the ability of 1D fibers in percolating a 2D plane based on different dispersion and distribution scenarios. The number of fibers shown is not enough to perfectly percolate the 2D structure. Sketches (a) and (b) show that the poor dispersion of the fibers prohibits network formation, while sketch (d) shows that perfect distribution of well dispersed fibers increases the gap between the fibers. Only the preferential distribution, case (c), of well dispersed fibers forms a conductive 2D network. It should be noted that formation a conductive network by preferential distribution of the conductive filler might not be suitable for applications such as ESD, because the poor homogeneity of the surface, depending on the degree of distribution and the filler dimensions, might allow for static charge build up. Percolation threshold and electrical conductivity of VGCNF/polymer composites depends on many factors including VGCNF aspect ratio, nanofiber dispersion, nanofiber distribution, nanofiber conductivity, polymer matrix crystallinity and polymer matrix surface tension. VGCNFs have a very high aspect ratio enabling them to percolate systems at low volume fraction [87], as shown in Fig. 8. Several techniques have been utilized to reduce the percolation threshold in VGCNF/ polymer composites. In the following sections, all the significant contributions reported to date toward reducing VGCNF percolation threshold, enhancing the electrical conductivity and understanding the conduction mechanism in VGCNF/ polymer composites are discussed. Table 2 summarizes the percolation threshold concentrations that have been reported so far in literature for VGCNF/polymer composites and some other competitive composites. The table includes the VGCNF type, processing conditions and polymer type.
2.1.
Effect of VGCNF surface chemistry
As-produced VGCNFs are typically covered with layers of amorphous polycyclic aromatic hydrocarbons [66]. The thickness of the amorphous layers depends on the production conditions. These layers have much lower conductivity than the crystalline VGCNFs (VGCNF composed of graphene planes only). The presence of amorphous layers increases the electrical percolation threshold, decreases the composite conductivity and affects the conduction mechanism because they increase the contact resistance between nanofibers, decrease
Fig. 8 – Effect of filler aspect ratio on the critical filler concentration needed to induce bulk conductivity in a filled polymer [88].
the nanofibers conductivity and increase the gap between the conductive cores. For example, tunneling was found to be the dominant conduction mechanism in as-grown VGCNF/PP composites even at high filler loadings [83]. Whereas in our lab, we found that direct contact between nanofibers is the dominant conduction mechanism of graphitized-VGCNF/ polymer composite. The negative influence of the polycyclic aromatics on the percolation threshold and electrical resistivity of polymer composites can be decreased by decreasing the thickness of the polycyclic layers [66]. From the discussion above, it becomes clear that post treatment of as-produced VGCNFs is a must to enhance their electrical properties. Several techniques can be used including carbonization, graphitization, etching in air at about 400–600 C and soaking in nitric acid, sulfuric acid or sulfuric/nitric acid mixture [20,23,101]. Among those techniques, graphitization has been proven to be the most effective in enhancing nanofibers conductivity, reducing percolation threshold and enhancing composite conductivity. Undoubtedly, increasing filler conductivity can reduce the percolation threshold by decreasing the filler/filler contact resistance. However, after surface treatment, the nanofibers often have better interaction with the polymer and this will lead to different dispersion, distribution, wetting and final aspect ratios of the nanofibers after compounding. Therefore, the decrease in percolation threshold due to the enhancement of VGCNF conductivity may not be the complete story, and characterizations of microstructure and nanofibers aspect ratio should also be performed. For example, Lee and coworkers [25] found that the surface area and/ or the morphology of VGCNF are more important than the conductivity of VGCNF in determining the conductivity of the composite. Similarly, Wu and Chung [102] reported that surface area of CF has a significant influence on the EMI SE of CF/polymer composites. But to date, a full understanding of these factors is lacking. Thus, there should be more work to fully understand how surface area and morphology affect conductivity.
CARBON
9
4 7 ( 2 0 0 9 ) 2 –2 2
Table 2 – Percolation threshold of different polymer composites Polymer PP
PC PC PC VEf VE PE PA-6 P(S-co-BuA)g PS PS PMMA PMMA PMMA Epoxy Epoxy Epoxy Epoxy Epoxy
Fibers type Percolation threshold Compounding/MoldingA VGCNF VGCNF VGCNF VGCNF VGCNF VGCNF VGCNF VGCNF VGCNF VGCNF MWNTs MWNTs CB CB VGCNF VGCNF MWNT
VGCNF MWNT MWNT VGCNF VGCNF VGCNF
VGCNF SWNT SWNT MWNT
5.6 vol% 3 vol% 2.47 vol% 5.2 vol% 4.97 vol% 0.5 vol% 9–18 wt% 18 wt% 7.5 wt 4–5 vol% 0.05 vol% 8 wt% 9.5 wt% 3 phr 6.3 wt% 7.5 1–1.5 wt% 2 wt% 4 wt% 2.5 phr 4–6 wt% 1.5 vol%
