Rheological behavior of polypropylene nanocomposites at low concentration of surface modified carbon nanotubes.
Ever since their discovery in 1990s, carbon nanotubes (CNTs) have generated great interest particularly as fillers for preparation of a variety of high-performance polymer nanocomposites due to their ability to transfer some of their outstanding properties into the polymer matrices (1),(2). However, it is difficult to disperse CNT in solvents or polymer melts because of their stable structure and strong resistance to wetting (3). Recent reviews on viscoelastic behavior of CNT-filled polymer nanocomposites reported a transition from melt to solid-like response at low oscillation frequencies, which is attributed to the formation of a filler network in the nanocomposites (4-6). However, CNTs have a tendency to form a network of agglomerates during their synthesis because of the existence of strong van der Waal's attraction between the particles, which rendered them extremely difficult to disperse uniformly in the polymer matrix 121. Thus, a significant challenge in developing a high performance polymer/CNT composite is to disperse and ensure efficient load transfer across the CNT-polymer matrix interface.
Recent studies reveal that functionalization of the CNT surface leads to improved dispersion of the nanofillers and consequent improvement in composite fracture strength (7-11). In a previous communication, it has been reported that the CNT surface can be functionalized by phenol resulting in increase in mechanical properties (12). Functionalization of CNT by aliphatic alcohols has been reported earlier(13-161).
Rheological test is known to be a sensitive macro-scopic tool for probing the interactions and structures of polymer nanocomposites, apart from being used as a quality control for providing valuable information on the melt processing conditions. However, theological studies on isotactic polypropylene (iPP) filled with functionalized CNT (f-CNT) are scanty. This article reports the result of studies on the melt rheological behavior of iPP/CNT nanocomposites, with special reference to the effect of CNT loading, both in the case of pristine CNT and phenol and 1-octadecanol (C18) f-CNT. The direct current (DC) volume resistivity of the nanocomposites was studied to supplement some of the findings in rheological studies.
A commercially available iPP (P510P, SABIC, Saudi Arabia) with density 905 kg [M.sup.-3] and MFR 12 g/10 min (ASTM D4101-10) was used as the matrix for this investigation. Multiwall CNTs with nanotube diameter of 8-15 urn, length 10-50 [micro]m, specific area 230 [m.sup.2][g.sup.-1], and purity of more than 95% was procured from Nanostructure and Amorphous Material.
Preparation and characterization of phenol f-CNT have been published in a previous communication (12). 1-Octadecanol (C 18) f-CNT was prepared from the acid-treated CNT. First, 1-octadecanol was heated to its melting point in a reaction flask and was maintained slightly below its boiling point, while acid-treated CNT was added into the reaction flask containing the molten surfactant, as the CNT:surfactant ratio is maintained at 1:10 parts by weight. The mixture was stirred for a few minutes, and a few drops of sulfuric acid were added to initiate the reaction. The reaction was allowed to continue for 2 h beyond which the resulting f-CNT was washed with toluene several times to remove any unreacted 1-octadecanol, followed by washing with deionized water to remove acid traces (Scheme 1).
Both CNT and C18-CNT were characterized using Fourier Transform Infrared Spectroscopy (FTIR) PERKIN ELMER 16F PC instrument. FTIR samples were prepared by grinding dried C18-CNT together with potassium bromide (KBr) to make a pellet. Both phenol functionalized and C18 functionalized CNTs are designated as f-CNT.
iPP/CNT nanocomposites were prepared by melt blending of both unmodified CNT and f-CNT into iPP at different loadings (i.e., 0.1, 0.25, 1.0, and 5.0 wt%) each, using a Haake twin-screw mini extruder at 190[degrees]C. Preparation and characterization of phenol f-CNT and composites of IPP and f-CNT have been reported in a previous communication (12).
Rheological measurements were performed using an advanced rheometrics expansion system. The measurements were performed using cone and plate geometry (25 mm diameter 0.1 rad cone angle and 0.4554 mm gap) at 190[degrees]C in nitrogen atmosphere. Frequency sweeps with an angular velocity of 0.1-100 rad [s.sup.-1] were performed in the linear viscoelastic regime at low strain of 5%. The samples were left to equilibrate for 5 min before each measurement.
The DC volume resistivity of the composite materials having wide range of resistivity was measured using instruments Agilent 4339B (High Resistance Meter attached with Agilent 16008B Resistivity Cell) for measurement range covering from [10.sup.6] ohm to [10.sup.16] ohm and GOM-802 (GW Instek DC milli Ohm Meter) for low resistance measurement. GOM-802 is attached with a home-made electrode, covering range from [10.sup.6] ohm to [10.sup.-3] ohm.
RESULTS AND DISCUSSION
Figure 1 shows the IR spectra of CNT and octadecanol f-CNT (CNT-C18). The IR spectrum of CNT shows 1698 [cm.sup.-1] assigned to carboxylic C=0 stretching and 1097 [cm.sup.-1] corresponding to C--0 stretch in alcohols. The presence of these functional groups on the surface of pristine CNT indicates their introduction during removal of metal catalysts in nanotubes purification processes (17). Carboxylic C=0 stretching peak observed at 1693 [cm.sup.-1] can be attributed to acid treatment of CNT 1181. Treatment of CNT with 1-octadecanol gives indicative peaks at 2920 [cm.sup.-1] corresponding to C[H.sub.2] stretching of long alkyl molecule of 1-octadecanol and 1473, 1458 and 1068 [cm.sup.-1] corresponding to the ether formation of 1-octadecanol with the carboxylic groups in CNT (191.
Figures 2 and 3 show the frequency dependence of storage modulus (G') and loss modulus (C') at different loadings of CNT and f-CNT. It is interesting to note that at the lowest nanofiller loading of 0.1 wt%, neat PP shows the highest modulus, which decreases on addition of CNT and f-CNT and the effect, is pronounced in the low frequency region. In contrast to this observation, at the next higher nanofiller loading of 0.25 wt% the effect of unmodified CNT on modulus has been marginal, but the f-CNT causes a drop in the modulus, and the effect is more pronounced in the C18 f-CNT, when compared with phenol f-CNT. When the nanofiller loading is further increased to 1 wt%, the effects of both unmodified CNT and f-CNT on the modulus become inconsequential. At 5 wt% of nanofiller loading, the effects are reversed, as both unmodified CNT and f-CNT show higher modulus, when compared with neat PP, the f-CNT registering lower modulus than the pristine CNT. It is apparent that CNT, at loadings of 0.1 wt% act as a processing aid for iPP and the effect is most pronounced in the case of C18 1-CNT. The processing aid effect is observable even at 0.25 wt% of f-CNT, but it is less pronounced in the case of unmodified CNT at this loading. At the nanofillers loading of 1 wt%, however, the processing aid effect of both unmodified and f-CNT is no longer observable and all the fillers (Fig. 2c) show similar modulus in the frequency range studied. The processing aid effect altogether disap-pears at 5 wt% loading of the nanofillers (Fig. 2d), when the neat PP registers the minimum modulus.
Figure 4 shows the plots of complex viscosity of iPP and PP composites containing different loadings of unmodified CNT and f-CNT. The complex viscosity of PP decreases on addition of the nanofillers at 0.1 wt[degrees]) and the effect is most pronounced in the case of C18 modified CNT. At the next higher nanofiller loading of 0.25 wt%, the processing aid effect is still observable, although the effect is less pronounced than that in 0.1 wt%. As was observed in the case of storage modulus and loss modulus, the effects of nanofiller on the complex viscosity disappears at the loading of 1 wt% for both modi fled and unmodified CNT (Fig. 4c). At 1 wt% loading, unmodified CNT and phenol modified CNT register only a marginal increase in complex viscosity, while C18 modified CNT registers a marginal decrease in complex viscosity, and this effect becomes less prominent at higher frequencies. At the higher loading of 5 wt% of the nanofillers, however, the neat PP show the minimum complex viscosity, which increases on incorporation of both unmodified CNT and f-CNT.
It is also evident that at low loadings of the nanofiller, the PP composites behaves more as Newtonian fluids compared with the unfilled PP sample, and the effect is most pronounced in the case of 0.1 and 0.25 wt% of C18 modified CNT. However, at 5 wt% of the nanofiller loading, the PP composites tend to show non-Newtonian behavior and in retrospect it is found that 1 wt% of the nanofiller may be termed as the cross over loading when the nanofillers shifts from behaving like a processing aid to a reinforcing filler.
In addition, our results indicate that there is no network formation of f-CNT inside the PP matrix for up to the maximum loading studied (5 wt%). This conclusion is derived from the observation that there is no clear plateau in G' values observed at low frequency range for the f-CNT-filled samples (Fig. 2), which is usually used in literature as a signature for a pseudo solid-like behavior and the formation of filler network for different nanocomposite systems. However, by carefully examining Fig. 2d, the slope of G' vs. w in the low frequency range for 5 wt% loading of unmodified CNT-filled sample seems to be considerably lower than that of the f-CNT-filled samples, indicating that network formation just started in the case of the unmodified CNT but not completed.
In general, according to Einstein rule, addition of solid hard fillers into melt of polymers is expected to result in an increase in viscosity (due to hydrodynamic effect), which will be proportional to the percentage by volume of the filler particles present (20). However, a
considerable reduction in complex viscosity, storage modulus, and loss modulus of the melt composites at a loading of less than 1.0 wt% of f-CNT was observed on application of shear stress during melt rheology. It is believed that before shear application, there are entanglements of polymer chains in high proportions. However, on application of shear stress, f-CNT alters polymer chain dynamics by the disruption of polymer chains entanglements and subsequently giving rise to increase in polymer chains fluidity. At lower concentration of f-CNT concentration (less than 1 wt%), this disruption of polymer entanglements effect is stronger than the hydrodynamic effect. A further increase in f-CNT loading (above 1.0 wt%) causes a sharp increase in the viscosity, and it supersedes the viscosity of increased neat iPP matrix when the hydrodynamic effect becomes the dominating factor over the disruption of polymer entanglements effect. The effect is more pronounced in the case of C 18 functionalization than the phenol functionalization. It is likely that the relatively long linear carbon chain on the surface of CNT in the case of C18, as shown in Scheme 1, would facilitate the disruption of polymer chain entanglements more than the short phenolic group in the case of phenol f-CNT and certainly much more than the unmodified CNT.
Although our data for the viscosity reduction at low filler loading is unusual, especially for C18 modified CNT, it is not unique for nanocomposites systems, and it has been observed by different investigators for different systems. For example, Zhang et al. (21) produced nanocomposites of ultra high molecular weight polyethylene and single-walled carbon nanotubes (SWNT). They observed an initial drop in viscosity ([eta]') up to 0.25% of SWNT loading. Also, Jain (22) reported a decrease in viscosity when silica nanoparticles were used as filler. They attributed the decrease in viscosity due to the selective physico-adsorption of polymer chains onto the surface of the filler. However, physico-adsorption is difficult to be quantified and justified. Another reason which is frequently mentioned in the composite literature and industry is due to the plasticization effect. The plasticizer usually increases chain mobility and the free volume; hence, it increases the fluidity of the polymer composite. However, CNT is not expected to act as a true plasticizer for our PP system, because in an independent study we measured experimentally the stiffness and glass transition of the CNT/PP system and found them to be higher at 0.1 wt% and 0.25 wt% loading compared with unfilled samples (12),(18). Therefore, the proposed hypothesis of the disruption of polymer chains entanglements at low filler loading is the most plausible explanation for the unusual rheological behavior. Also, by referring back to Figs. 2a, 3a, and 4a, it is interesting to notice that adding 0.1 wt% loading of 08 f-CNT caused a considerable drop in the viscous properties of the neat polymer not only in the low frequency range but also at the highest frequency used (100 rad [s.sup.-1]), which is close to the typical window of process ability. Therefore, as a practical application, the 0.1 wt% loading of CI8 f-CNT could be exploited as a processing aid for improving polypropylene process ability.
As a complement to rheological measurement, electrical property was also investigated by measuring the DC volume resistivity. It is evident from Fig. 5 that the volume resistively decreases with increase in loading of the nanotubes. But, the pattern of decrement is different in the case of f-CNT, when compared with unmodified CNT. At 0.1 wt%, there is a sharp drop in the resistivity in the case of f-CNTs compared with unmodified CNT. Among phenol and C18 functionalizations, C 18 shows slightly more decrement than phenol at 0,1 and 0.25 wt% loading, which can be better understood from the inset figure in Fig. 5. However, at the loading of 1 wt%, it is interesting to note that the volume resistivity is same for both unmodified CNT and f-CNTs. Interestingly, the results of the rheological measurements of process parameters also reveal that at 1 wt% loading of both CNT and f-CNT, the composites display similar modulus and complex viscosity as that of the neat iPP (Figs. 2c, 3c, and 4c). However, at the loading of 5 wt%, the reduction in volume resistivity is less than that of f-CNT. This is because functionalization of CNT by C 18 and phenol results in better dispersion of CNT inside the hydrophobic PP matrix compared with the unmodified CNT. Therefore, at 5 wt% loading, most likely unmodified CNT inside the matrix start to form aggregates and start to build network, whereas in the case of f-CNT, the network formation has not yet started. This is in agreement with the storage modulus data by comparing the slope of G' vs. co in the low frequency range (Fig. 2d), as discussed earlier.
The effects of multiwall CNT loading, and C 18 and phenol functionalization of CNTs on the rheological parameters of polypropylene/CNT nanocomposites have been investigated. Storage modulus, loss modulus, and complex viscosity decrease in the low concentration region (<1%) of nanofillers in iPP/CNT composites, especially C18 f-CNT sample. Rheological properties, however, register higher modulus and viscosity in the high concentration region (>1%). It is believed that the processing aid behavior due to the disruption of polymer entanglements effect dominates in the low concentration region especially for C18 modified CNT, while the hydrodynamic effect overshadows it in the high concentration region for all types of CNT studied. The interesting observation that the viscosity of the polypropylene melt decreases significantly on the addition of trace amount of C 18 f-CNT (0.1 wt%) even at the highest frequency used (100 rad [s.sup.-1]), which is close to the typical window of processability, could have a practical application for using C18 f-CNT as an effective processing aid for improving the polypropylene processability.
The authors thank King Fahd University of Petroleum and Minerals (KFUPM) for using its research and laboratory facilities.
(1.) S. Iijima, Nature, 56, 354 (1991).
(2.) Z. Spitalsky, D. Tasis, K. Papagelis, and C. Galiotis, Prog. Polym. Sci., 35, 357 (2010).
(3.) S.H. Lee, M.W. Kim, S.H. Kim, and J.R. Youn, Eur. Polym. J., 44, 1620 (2008).
(4.) Q.H. Zhang, S. Rastogi, and D.J. Chen, Carbon, 44, 778 (2006).
(5.) K.Q. Xiao, L.C. Zhang, and I. Zarudi, Compos. Sci. Technol., 67, 177 (2007).
(6.) M. GanB, B.K. Satapathy, and M. Thunga, Acta Mater., 56, 2247 (2008).
(7.) A. Hirsch and 0. Vostrowsky, Top Curr. Chem., 245, 193 (2005).
(8.) D. Tasis, N. Tagmatarchis, A. Bianco, and M. Prato. Chem. Rev., 106, 1105 (2006).
(9.) B.X. Yang, K.P. Pramoda, G.Q. Xu, and S.H. Goh, Adv. Funct. Mater. 17, 2062 (2007).
(10.) M. Yang, Y. Goa, H. Li, and A. Andronov, Carbon, 45, 2327 (2007).
(11.) H.S. Goh, Y. Bing-Xing, H.J. Shi, and K.P. Pramoda, Cornpos. Sci. Technol., 68, 2490 (2008).
(12.) S.P. Thomas, S.A. Girei, A.A. Al-Juhani, K. Mezghani, S.K. De, and M.A. Atieh, Polym. Eng. Set, 52, 525 (2012).
(13.) S. Banerjee, M.G.C. Kahn, and S.S. Wong, Chem. Eur. J., 9, 1898 (2003).
(14.) L. Zhang, V.U. Kiny, H. Peng, J. Zhu, R.F.M. Lobo, J.L. Margrave, and V.N. Khabashesku, Chem. Mater., 16, 2055 (2004).
(15.) N. Zhang, J. Xie, and V.K. Varadan, Smart Mater. Struct., 15, 123 (2006).
(16.) R. Tian, X. Wang, M. Li, H. Hu, R. Chen, F. Liu, H. Zheng, and L. Wan, Appl. Surf. Sci., 255, 3294 (2008).
(17.) L. Vaisman, H.D. Wagner, and G. Marom, Adv. Colloid Interface Sci., 128-130, 37 (2006).
(18.) S.A. Girei, S.P. Thomas, M.A. Atieh, K. Mezghani, S.K. De, S. Bandyopadhyay, and A.A. Al-Juhani, J. Thermoplast. Compos. Mater., in press. DOI: 10.1177/0892705711406159.
(19.) G. Socrates, Infrared Characteristic Group Frequencies, Wiley, Bristol (1980).
(20.) W. Guth, J. Appl. Phys., 16, 20 (1945).
(21.) Q. Zhang, D.R. Lippits, and S. Rastogi, Macromolecules, 39, 658 (2006).
(22.) S.H. Jain, "Nano-scale Events with Macroscopic Effects in Polypropylene/Silica Nanocomposites; Effect of Polymer Adsorption on Processability and Properties," Ph.D. Thesis, Technische Universiteit Eindhoven (2005).
Selvin P. Thomas, (1) Salihu Adamu Girei, (1) Muataz Ali Atieh, (1), (2) S.K. De, (1) Abdulhadi Al-Juhanil (1)
Correspondence nr. Abdulhadi Al-Juhani; e-mail: firstname.lastname@example.org Contract grant sponsor: King Abdulaziz City for Science and Technology (KACST) through the Science and Technology Unit, King Fahd University of Petroleum and Minerals (KFUPM); contract grant number: AR-28-118.
(1) Chemical Engineering Department, King Fand University of Petroleum and Minerals, 31261 Dhahran, Saudi Arabia
(2) Center of Excellence in Nanotechnology (CENT), King Fand University of Petroleum and Minerals, 31261 Dhahran, Saudi Arabia
|Printer friendly Cite/link Email Feedback|
|Author:||Thomas, Selvin P.; Girei, Salihu Adamu; Atieh, Muataz Ali; De, S.K.; Al-Juhani, Abdulhadi|
|Publication:||Polymer Engineering and Science|
|Date:||Sep 1, 2012|
|Previous Article:||Surface texture effect on scratch behavior of injection molded plastics.|
|Next Article:||Viscoelastic constitutive model for uniaxial time-dependent ratcheting of polyetherimide polymer.|