J Infrared Milli Terahz Waves (2011) 32:1457–1463 DOI 10.1007/s10762-011-9832-3

Nondestructive Evaluation of Rubber Compounds by Terahertz Time-Domain Spectroscopy Yasuyuki Hirakawa & Yoshitomo Ohno & Toyohiko Gondoh & Tetsuo Mori & Kei Takeya & Masayoshi Tonouchi & Hideyuki Ohtake & Tomoya Hirosumi

Received: 19 April 2011 / Accepted: 13 September 2011 / Published online: 22 September 2011 # Springer Science+Business Media, LLC 2011

Abstract Rubber compounds were investigated by terahertz time-domain spectroscopy. Terahertz absorption spectra of crude rubbers and additives were measured as well as those of acrylonitrile-butadiene rubber compounds, which included the additives. It was found that carbon black, which is one of the additives and serves as a filler, dominates the terahertz absorption owing to its metallic characteristics. Thus, terahertz spectroscopy is a useful method for rapid nondestructive inspection during the rubber production. Keywords THz . THz-TDS . Rubber . Nondestructive

1 Introduction Terahertz (THz) applications are an active research area now that suitable THz sources and detectors have become available. Consequently, a variety of THz-related investigations have been performed [1]. THz time-domain spectroscopy (THz-TDS) [2, 3] is especially attractive for nondestructive evaluation of materials such as semiconductors, biological materials, and polymers [4–12]. Rubber is a complex but important industrial product. Many unknown properties arise not only in the rubber itself but also in the manufacturing processes of mixing and vulcanization. In particular, rubber products are made using mixing and chemical reactions with various additives. The production process is as follows: a masticated crude rubber is mixed with several additives in a mixing machine. The resulting materials are called rubber compounds. Subsequently, these rubber compounds are heated and pressed in a mold to make vulcanized rubber products. In a manufacturing Y. Hirakawa (*) : Y. Ohno : T. Gondoh : T. Mori Kurume National College of Technology, Kurume 830-8555, Japan e-mail: [email protected] K. Takeya : M. Tonouchi Institute of Laser Engineering, Osaka University, Suita 565-0871, Japan H. Ohtake : T. Hirosumi AISIN SEIKI Co., Ltd., Kariya 448-8650, Japan

1458

J Infrared Milli Terahz Waves (2011) 32:1457–1463

setting, the decision about various parameters such as the selection of the additives, the mixing time, and the vulcanization time and temperature are usually based on past experience, and it is not always optimal because detailed information about most rubbers is still lacking. In order to improve the manufacturing processes, it is necessary to develop a quantitative, nondestructive, and rapid method for evaluating rubber compounds. In the present study, as a first step in developing such analytical methods, rubber compounds are evaluated using THz-TDS. Although the THz optical parameters of polydimethylsiloxane were investigated [13], few studies about rubber compounds were so far reported using THzTDS. To the authors’ knowledge, this is the first report on systematic THz measurements of rubber compounds. In the experiments, THz absorption spectra of crude rubbers and additives are measured as well as those of acrylonitrile-butadiene rubber compounds. Technical difficulties in the proposed methods are also discussed.

2 Experimental The experimental setup used in this study was the compact THz-TDS system [14] shown in Fig. 1. The source for the THz emission and detection was a femtosecond fiber laser (IMRA Femtolite C-20, 780 nm, 100 fs, 20 mW). The THz radiation was emitted from a dipole-type low-temperature-grown GaAs photoconductive (PC) switch pumped by a portion of the laser beam. The generated THz radiation was incident on a rubber sample.

Fig. 1 THz-TDS system used in this study.

J Infrared Milli Terahz Waves (2011) 32:1457–1463

1459

The ensuing transmission was detected using the same type of PC switch as the emitter. The whole THz-TDS system was placed in a special box purged with dry nitrogen gas in order to keep the ambient humidity of 5% during the experiments. The inset of Fig. 1 shows the typical THz frequency spectrum of the reference. The THz beam diameter was about 1 mm at the focal point (at the sample surface) measured by the knife-edge method. Three kinds of crude rubbers were used as samples: natural rubber (NR), styrenebutadiene rubber (SBR), and acrylonitrile-butadiene rubber (NBR). A compound sample was also measured, namely an unvulcanized rubber containing several additives including carbon black (CB), zinc oxide (ZnO), stearic acid (SA), sulfur (S), N-cyclohexyl-2benzothiazyl sulfenamide (CBS), tetramethyl thiuram disulfide (TMTD), and dibenzothiazyl disulfide (MBTS). When dealing with rubbers, a special unit of measure indicating the composition is used in the industry: “phr” or “per hundred rubber” meaning the relative weight when the crude rubber weight is 100 (in any units). The NBR compound contained 100 phr NBR, 40 phr CB, 5 phr ZnO, 1 phr SA, 1 phr S, 1 phr CBS, 1 phr TMTD, and 1 phr MBTS. The size of the rubber samples were 10 mm×10 mm with a thickness of about 0.5 mm. Since the additives used in these experiments are powdered, each of them was pressed using a tablet machine (Ichihashi-Seiki HANDTAB-Jr) under a pressure of 30 MPa. Because CB has a very low THz transmission due to its metallic character, the CB powder was carefully mixed with polyethylene powder (PE) at a mixing weight ratio of CB:PE= 1:30. The tablets had a diameter of 10 mm and a thickness of about 1.5 mm. The thickness of the samples was measured carefully by a caliper (Mitutoyo, NK15) and a thickness gauge (Peacock, dial thickness gauge G). As for the rubber samples, the measured thickness values were reconfirmed by the numerical simulations [15–17] to calculate the sample thickness. All of the THz transmissions were evaluated based on an absorbance per unit length. Since the surface of the rubber samples did not always indicate perfect flatness such as tablet samples, it was experimentally confirmed in advance that the influence of the surface roughness to the THz measurements was negligible for this rubber estimation. The THz absorption spectra of the rubber samples with different surface conditions evaluated by a laser microscope (Olympus OLS3100) were compared with each other, and it was confirmed that their THz spectra had the same profiles and absorption levels.

3 Results and discussion Industrial crude rubber samples after mastication were cut into pieces with a thickness of about 0.5 mm and their THz transmission was measured. Figure 2 plots the resulting absorbance for the NBR, SBR, and NR samples. All the data in the experiments were averaged for 10 measurements, and all of the spectra presented in this paper were smoothed using three-point averaging after calculating the effective dynamic range [18] to clarify the THz frequency range with a good signal to noise ratio (SNR) for each sample. No narrow peaks are visible and weak Fabry-Pérot oscillations were observed in the THz spectra for these crude samples. The NBR sample displays the strongest absorption, while the crude NR and SBR samples have weak spectral profiles. These broad THz spectra result from the fact that (crude) rubbers are amorphous aggregates of polymer molecules. The broad NR spectrum is similar to that previously reported by Rungsawang et al. [10]. Polar molecules give rise to significant THz absorption [19, 20]. The strong NBR absorption is probably due to the polar nitride group (CN). Carbon black, the most important additive in rubber products, is a form of amorphous carbon. It is expected to reflect THz radiation due to its metallic characteristics and thus

1460

J Infrared Milli Terahz Waves (2011) 32:1457–1463

Fig. 2 Absorption spectra of NBR, SBR, and NR crude samples.

control the THz characteristics of rubber compounds. In order to clarify the effects of the CB and other additives in the THz spectral region, the absorption spectra of various additives were measured by THz-TDS. The results are graphed in Fig. 3. For the CB data, the absorbance was estimated from measurements of mixed CB/PE and pure PE tablets Fig. 3 Absorption spectra of additives: (a) CB, ZnO and SA; and (b) other additives. The bulk densities were 0.72, 3.4, 0.99, 1.8, 1.2, 1.3 and 1.5 g/cm3 for CB, ZnO, SA, S, CBS, TMTD and MBTS, respectively.

J Infrared Milli Terahz Waves (2011) 32:1457–1463

1461

Fig. 4 Calculated THz absorption spectrum based on data shown in Figs. 2 and 3 according to their compounding ratios, together with the measured THz absorption spectrum for the NBR compound containing 40 phr CB.

using the mixing ratio of CB/PE and the bulk densities. As Fig. 3(a) indicates, the CB absorbance is much larger than that of any other additives. However, the THz reflection was not separately measured in this study and thus contributes to the apparent absorption. As seen in Fig. 3(b), the vulcanization accelerators exhibit several absorption peaks in contrast to the CB, the vulcanization activator SA and the crude rubbers. Although the peak intensities of these accelerators are large, the quantity of the compounding ingredients was only 1 phr except for the crude rubber (100 phr), CB (40 phr), and ZnO (5 phr). Therefore, these THz spectral components should be negligible in the overall THz spectra of the rubber compounds. To confirm this assumption, the total spectrum was calculated using the individual THz spectra in Fig. 3, the compounding ratio, and the bulk density of the materials. The crude NBR was selected as the base rubber because the ideal crude rubber in this simulation must exhibit stable THz spectra. The crude NBR always has similar THz absorptivity and stable characteristics while the SBR samples exhibit some spectral variations and the NR samples lack physical uniformity. In the simulations, the CB quantity was fixed at 40 phr. Figure 4 shows the calculated results together with a measured THz absorption spectrum for a NBR compound containing a 40 phr CB. In the case of the NBR with the CB 40 phr sample, the good SNR region was Fig. 5 Measured THz absorption spectra for crude NBR and for NBR compounds with and without CB.

1462

J Infrared Milli Terahz Waves (2011) 32:1457–1463

limited below 1.8 THz. In this THz region both spectra agree well, although the THz transmission was disturbed by reflection and scattering losses due to the CB. Another simulated spectrum, calculated by adding together only the crude NBR and the 40 phr CB absorptivity curves without any other additives, had almost the same profile as that estimated using all the additives, suggesting that CB dominates the THz transmissivity compared to all other additives in the rubber. In order to clarify the effect of the CB, the absorption spectra of the NBR samples with and without the CB were measured. All other conditions, such as the additive compounding ratios and the mixing processes, remained unchanged. Figure 5 plots the THz absorbance for the crude NBR, and for NBR compounds with and without CB. At frequencies less than 1.8 THz, the slope of the spectrum increases in the presence of CB and other additives. This observation qualitatively agrees with Rungsawang’s results [10] in which the absorption coefficient of carbon nanotubes (CNTs) in vulcanized NR was measured. The absorption spectra in both Rungsawang’s and our experiments increased with the carbon content rate of the samples. The comparison between Rungsawang’s and our spectra suggest that the absorption of vulcanized NR with CNTs is stronger than that of our (unvulcanized) NBR compound considering their carbon content rates. Rungsawang’s composite contained only 5 wt% carbon at a maximum and showed an absorption coefficient of more than 200 cm-1, whereas our result indicated an absorbance of 100 cm-1 for the sample having 40 phr CB (25 wt%). This was probably due to the difference of the carbon shape and rubber internal cross-linkages caused by the vulcanization. These aspects should be investigated in future experiments along with THz reflection measurements. The absorption below 1.5 THz contains information not only about the CB content but also about other additives. However, although the slope increment can specifically be used to detect the existence of CB, the variation in the slope seems much smaller or practically naught for other additives. Further experiments are necessary using other rubber materials and different experimental conditions to clarify the effect of such additives. Therefore, from the present experimental results, the slope variation makes it possible to use THz absorption as a nondestructive means of evaluating the CB existence. It is easy to extend this evaluation method to a two-dimensional imaging of the CB distribution, which has a direct impact on the properties of the manufactured rubber products. In conclusion, in this study rubber compounds were evaluated by THz-TDS. The absorption spectra of crude rubbers, additives, and NBR compounds were measured. The CB in samples is easily detectable based on THz absorption below 1.5 THz. This is thus a useful method for rapid and nondestructive inspection during rubber production. Acknowledgement This study was supported by the 2009 Japan Science and Technology (JST) project to develop innovative seeds.

References 1. M. Tonouchi, Nature Photonics, 1, 97 (2007) 2. S. Nishizawa, K. Sakai, M. Hangyo, T. Nagashima, M. W. Takeda, K. Tominaga, A. Oka, in Terahertz Optoelectronics, ed. by K. Sakai, (Springer, Berlin, 2005), p. 203 3. J. B. Baxter and C. A. Schmuttenmaer, in Terahertz Spectroscopy PRINCIPLES AND APPLICATIONS, ed. by S. L. Dexheimer, (CRC Press, Boca Raton, 2007), p. 73 4. C. A. Schmuttenmaer, Chem. Rev., 104, (2004) p. 1759 5. R. A. Kaindl, R. D. Averitt, in Terahertz Spectroscopy PRINCIPLES AND APPLICATIONS, ed. by S. L. Dexheimer, (CRC Press, Boca Raton, 2007), p. 119

J Infrared Milli Terahz Waves (2011) 32:1457–1463

1463

6. N. Nagai, R. Fukasawa, Chem. Phys. Lett., 388, 479 (2004) 7. E. J. Heilweil and D. F. Plusquellic, in Terahertz Spectroscopy PRINCIPLES AND APPLICATIONS, ed. by S. L. Dexheimer, (CRC Press, Boca Raton, 2007), p. 269 8. J. Son, J. Appl. Phys., 105, 102033 (2009). 9. H. Hoshina, A. Hayashi, N. Miyoshi, F. Miyamaru, C. Otani, Appl. Phys. Lett., 94, 123901 (2009) 10. R. Rungsawang, V. G. Geethamma, E. P. J.Parrot, D. A. Ritchie, E. M. Terentjev, J. Appl. Phys., 103, 123503 (2008) 11. R. Piesiewicz, C. Jansen, S. Wietzke, D. Mittleman, M. Koch, and T. Kürner, J. Infrared Milli. Terahz Waves, 28, 363 (2007) 12. Y.-S. Jin, G.-J. Kim, and S.-G. Jeon, J. Korean Phys. Soc., 49, 513 (2006) 13. A. Podzorov, G. Gallot, Chem. Phys. Lett., 495,46 (2010) 14. Y. Hirakawa, Y. Ohno, T. Gondoh, T. Mori, K.Takeya, M. Tonouchi, H. Ohtake, T. Hirosumi, Extended Abstracts of International Workshop on Terahertz Technology 2009 (TeraTech’09), 1p–43 (2009) 15. M. Scheller, C. Jansen, M. Koch, Opt. Commun., 282, 1304 (2009) 16. T. D. Dorney, R. G. Richard, G. Baraniuk, D. M. Mittleman, J. Opt. Soc. Am. A, 18, 1562 (2001) 17. I. Pupeza, R. Wilk, M. Koch, Opt. Express, 15, 4335 (2007) 18. P. U. Jepsen, B. M. Fischer, Opt. Lett., 30, 29 (2005) 19. B. N. Flanders, R. A. Cheville, D. Grischkowsky, N. F. Scherer, J. Phys. Chem., 100, 11824 (1996) 20. T. Ohkubo, M. Onuma, J. Kitagawa, Y. Kadoya, Appl. Phys. Lett., 88, 212511 (2006)