Ceramic Processing Research

Transcription

1 Journal of Ceramic Processing Research. Vol. 8, No. 2, pp. 114~118 (2007) J O U R N A L O F Ceramic Processing Research Preparation of copper nanoparticles with an organic coating by a pulse wire ischarge metho K. Murai a, Y. Watanabe a, Y. Saito a, T. Nakayama a, H. Suematsu a, W. Jiang a, K. Yatsui a, K.B. Shim b an K. Niihara a a Extreme Energy-ensity Research Institute, Nagaoka University of Technology, Nagaoka, Niigata , Japan b epartment of Avance Materials Engineering, Hanyang University, Seoul , Korea In the pulse wire ischarge metho, a coating technique to prepare metal nanoparticles covere with organic matter has been investigate. Copper nanoparticles covere with organic matter have been successfully prepare by evaporation of a copper wire in an oleic aci vapor/mist. The thickness of the coating layer was a few nanometres. The meian iameter of the power was 25 nm, an became 10 nm smaller than that without the coating because of the inhibition of particle growth by the formation of the coating. From phase ientification by transmission electron microscopy an X-ray iffraction analysis, it was foun that the copper nanoparticles have been passivate an have not been oxiize in over 2 months. Key wors: Nanoparticles, Pulse wire ischarge, Copper, Passivation. Introuction Soli particles with an average grain size in the range of 1 to 100 nm are calle nanoparticles. Various unique characteristics can be exhibite by the reuction of the particle size [1, 2]. Since some of these characteristics are attribute to the high surface activity of the nanoparticles, maintaining the activity is important as on effective use of these characteristics. One of the major eteriorations of the activity is cause by oxiation in the case of base metals. The base metal nanoparticles react easily with oxygen in air even at room temperature an form an oxie phase on the surface, which has an unesirable effect on usage. Consequently, special isolation an passivation techniques of the nanoparticles prepare are require. Along with a recent remarkable growth of nanotechnology, nanoparticles as a raw material with superior or aitional properties are in strong eman. A hybriization of the base metal nanoparticles is the most realistic way to obtain these properties. There is a possibility of a useful effect evelopment at the hybriizing interface, an increasing the egree of mixing of the nanoscale mixture seems to result in a better composite effect. To passivate an hybriize nanoparticles, a coating appears to be a promising metho. Physical vapor eposition (PV) is a major synthesis metho of nanoparticles in gas phases. In this metho, nanoparticles are prepare by a homogeneous nucleation from the evaporate raw material, namely by conensation an coagulation. It is possible to prepare high *Corresponing author: Tel : ext Fax: kk11@etigo.nagaokaut.ac.jp purity power by PV. However, the particle size istribution is too broa compare with monoisperse particles [3]. In PV, some ways to overcome the problems mentione above have been reporte by many researchers over the past years. An oler metho of particle passivation is the formation of a thin an ense oxie layer on the surface of the particle immeiately after its preparation. Though the oxie layer plays a role as an oxiation-resistant coating, in some cases, it is ifficult to remove the oxie layer later on for usage. Yatsuya et al. successfully prepare isolate aluminum an silver particles on the surface of substrate coate with oil by vacuum evaporation on a running oil substrate (VEROS) metho [4]. Kwon et al. ae a stearic aci solution to prepare tungsten power before contact with air, an passivate it [5]. Nakayama et al. evaporate metal an organic matter simultaneously, an mae special isolate eposits of several kins of metal particles with organic matter on a substrate [6]. These techniques are more like immersion than coating. A limitation of the organic matter species is not so severe in the PV metho, compare with the solution phase synthesis metho. The pulse wire ischarge (PW) metho, where the exploing wire technique is applie, is one of the methos for preparing nanoparticles [7-9]. In PW, a soli wire is evaporate by a pulse current to prouce a vapor, which is then coole by an ambient gas to form nanoparticles. Preparations of metal, oxie an nitrie nanoparticles by PW have been reporte [7-11]. This metho has potentially a high prouction rate an a high energy efficiency. Furthermore, the nanoparticles are prepare in a simple apparatus consisting of a vacuum chamber, a power collection filter an a ischarging circuit. Therefore, PW is consiere to 114

2 Preparation of copper nanoparticles with an organic coating by a pulse wire ischarge metho 115 be suitable for mass-prouction of nanoparticles at low cost. In this stuy, a coating technique to prepare metal nanoparticles covere with organic matter has been examine. Specifically, the influence of the coating on the particle size an the proceeing oxiation have been investigate. Experimental Setup The experimental setup is shown in Figure 1. The system consiste of a chamber (2.5 l in volume), an oil-free ry vacuum pump, a power collection filter (which inclues a col trap an a filter), an a simple ischarge circuit (which inclue a high-voltage power supply, a capacitor, an a spark gap switch). A copper wire (99.9% in purity, Nilaco) with length of 25 mm an iameter of 0.25 mm was place in the chamber which was fille with nitrogen gas. The ambient gas pressure was kept at 100 kpa. Before the ischarge, 15 mg of oleic aci (Tech. 90%, SIGMA-ALRICH) was heate an evaporate into the chamber by an electrical heater at a temperature of 473 K. This temperature is the boiling point of the oleic aci use in this experiment, which was measure by a thermobalance. A pulse current was generate from a 10 µf capacitor which was charge up to 5.2 kv. In orer to evaporate the copper wire, an energy of 68 J is require theoretically [12]. The store energy in the capacitor was therefore enough to evaporate the wire. The current was riven through the wire by a triggering pulse input to the spark gap switch, i.e. closing the switch. The amount of energy eposite in the wire was calculate from the current an voltage waveforms. After the ischarge, the chamber was evacuate through a membrane filter (pore size 0.1 µm, Millipore) for power collection. The col trap, which was coole by liqui nitrogen, in front of the filter plays a role as an organic matter vapor/mist collector. The efficiency of trapping the organic vapor was estimate by the weight change of the filter in an experiment without PW. With the col trap at 77 K, only 0.1 mg of oleic aci existe on the filter. Aitionally, the copper power without the coating was prepare using almost the same conitions but without the evaporation of oleic aci for comparison. The power morphology was observe by a transmission electron microscope (TEM; JEOL JEM2010, at an accelerating voltage of 200 kv). The particle size istribution was analyze by measuring the iameter of more than 1000 nanoparticles on bright fiel images (BFIs). The crystal structure of the powers was ientifie with a power X-ray iffractometer (XR; Rigaku RINT2000, Cu-Kα raiation of nm, 50 kv, 300 ma). The ientification of the organic matter was carrie out by a Fourier transform infrare spectroscope (FT-IR; Nicolet IMPACT410) using the infrare reflection metho. Experimental Results Shown in Fig. 2 are the waveforms from the current an voltage measurements. The energy eposite in the Fig. 2. Current an voltage waveforms, an energy eposite in the wire. Here, 0 µs is a point in time at which the spark gap switch was close. Fig. 1. Experimental setup.

3 116 K. Murai, Y. Watanabe, Y. Saito, T. Nakayama, H. Suematsu, W. Jiang, K. Yatsui, K.B. Shim an K. Niihara Bright fiel TEM images of copper particles prepare in an oleic aci vapor/mist (a) with/(b) without an acetone ispersion treatment. The arrow in Fig. 3(b) shows a coating layer. Fig. 3. wire is plotte in the same figure. These results are almost the same as those previously which were measure with the same electric experimental conitions but without the evaporation of oleic aci [13]. It can be seen that the evaporation of oleic aci oes not influence the ischarge uner these conitions. Since several investigations concerning the ischarge in the exploing wire phenomenon have been reporte [13, 14], this will not be iscusse in etail here. Bright fiel TEM images of particles prepare in the oleic aci vapor/mist are shown in Fig. 3. Here, Fig. 3(a) is an image of particles after the ispersion treatment to measure particle iameters. Figure 3(b) is an image of prepare particles without the ispersion treatment, to observe the surface conition. Most nanoparticles o not have iameters above 100 nm. The shape of nanoparticles is spherical, an the nanoparticles are aggregate. Thin coating layers were observe on all particles, an the thickness was a few nm. Figure 4 shows a typical bright fiel TEM image of particles prepare in nitrogen gas without the oleic aci vapor/mist. The shape of these particles is clearly ifferent from that shown in Fig. 3(a). Specifically, interparticle neck growth ue to particle collision/coalescence was observe frequently. Figure 5 shows size istributions of particles prepare with/without the coating. These istributions were Typical bright fiel TEM image of particles prepare in nitrogen gas without the oleic aci vapor/mist. Fig. 4. Fig. 5. Size istributions of particles prepare with/without coating. obtaine from multiple images incluing those shown in Figs. 3(a) an 4. The frequency was fitte to soli or ashe curves base on the following log-normal istribution function, (log log 50)2 1 -, ( )= exp πlogσg 2log2σg (1) f where Σ i (log i log 50)2 logσg= , log n Σ ni Σ i log i -. = n 50 Σ ni Here, ( ) enotes the log-normal istribution function, is the particle iameter, i is the number of the nanoparticle with iameter of i, 50 is the meian iameter, an σg is the geometric stanar eviation. It was foun that the 50 an σg of the particles with the coating were 25.0 nm an 1.79, while those of the particles without the coating were 35.0 nm an 1.49, respectively. As a tren, the particle iameter with the coating was smaller an the ivergence from the log-normal istribution function was larger than those of particles without the coating. Figure 6 shows FT-IR spectra of (a) the oleic aci, (b) the evaporate oleic aci an (c) the prepare particles with the coating. In orer to obtain the FT-IR spectrum of the coating matter, a solution was prepare by issolving the matter from the power in acetone. A test sample for the FT-IR measurement was prepare by conensing the matter from this solution. It was confirme that the molecular structure of the oleic aci ha been retaine after the evaporation. However, there are some ifferences on comparing Fig. 6(a) with Fig. 6(c). Specifically, the isappearance of a broa spectrum ue to the O-H stretching vibration ( cm 1), the isappearance of some spectral characteristics ue to the C-O stretching an O-H eformation vibration ( cm 1), an an intensity reuction of the spectral characteristics ue to the C=O stretching vibration (1700 cm 1) were observe after the coating. As can be seen from these ifferences, the coate matter i not retain its original molecular structure of the oleic aci. Since f n

4 Preparation of copper nanoparticles with an organic coating by a pulse wire ischarge metho FT-IR spectra of (a) oleic aci, (b) evaporate oleic aci an (c) the coating matter on the prepare particles. Fig. 6. some peaks ue to C-H stretching an eformation vibrations ( cm 1 an cm 1, respectively) remain in the spectrum in Fig. 6(c), the coate matter consiste of some types of organic matter, (i.e. it i not consist mainly of soot). Figures 7 an 8 show XR patterns of particles prepare with/without the coating, respectively. In the first case, since the measurements were carrie out for particles prepare an ahere to the membrane filter, there are several iffraction lines from the filter in Fig. 7. It was seen that the crystallize particles immeiately after the preparation consiste only of copper. It was foun that the copper nanoparticles with the coating have been 117 Change of XR patterns of particles prepare without the coating over time ((a) as-prepare, (b) after 5 ays, (c) after 10 ays an () after 3 months). Fig. 8. passivate an have not been oxiize in over 2 months. Since it is clear from Fig. 8 an an other stuy [8] that the copper nanoparticles without the coating oxiize graually, it was emonstrate that the organic coating on the metal nanoparticles is effective in preventing the oxiation of them. High resolution electron microscope (HR-TEM) images of a particle without a coating 5 ays after the preparation an its closeup are shown in Figs. 9(a) an 9(b). Many grains with the size of a few nm are observe on the particle. Lattice fringes of 2.3 Å in Fig. 9(b) are seen on the surface, which is in goo agreement with the interplanar spacings of CuO (111). iscussion First, the place an time of the coating are investigate. As mentione above, oleic aci vapor which has not ahere to the particles is suppose to be conense completely by the col trap. From Fig. 3(b), the coating must not be forme on the filter uring the collection but must be generate in the chamber uring PW. Change of XR patterns of particles prepare with the coating over time ((a) as-prepare, (b) after 2 months). Fig. 7. HR-TEM images of a particle without the coating 5 ays after the preparation ((a) overall, (b) close-up view). Fig. 9.

5 118 K. Murai, Y. Watanabe, Y. Saito, T. Nakayama, H. Suematsu, W. Jiang, K. Yatsui, K.B. Shim an K. Niihara Fig. 10. Comparison of the theoretical enthalpy to heat a copper particle from the boiling point of oleic aci to the boiling point of copper with the heat of vaporization of the oleic aci layer. Seconly, the reason for the particle size istribution observe is iscusse. There are significant ifferences of particle size istributions between the particles prepare with an without the coating. These ifferences can be explaine by the influence of the oleic aci vapor/mist on the particle growth process. The formation of the coating ue to the conensation on the particle surface occurre uring the particle growth process. It is expecte that the coating acts as an impeiment to collision/coalescence, so that a ecrease of the neck formation an grain growth processes occurs. At the time of collision/coalescence, re-vaporization of the coating ue to the heat of the particle probably has also occurre. Figure 10 shows a comparison of the theoretical enthalpy to heat a copper particle from the boiling point of oleic aci to the boiling point of copper [12] with the heat of vaporization of the oleic aci layer [15]. These were calculate by multiplication of the enthalpy an volume as a function of particle iameter. Here, the temperature of a copper particle is 2843 K. It is assume that the thickness of the oleic aci layer is 1 nm. From Fig. 10, with copper particles in the range of 1 to 20 nm in iameter, it is shown that the heat content of a copper particle is not enough to vaporize the oleic aci layer, i.e. the coating remains. On the other han, with copper particles of 20 nm-iameter or more, it is possible to re-vaporize the oleic aci coating layer, consiering the heat available. From these computational results, it appears that the coating prevents grain growth in particular on smaller particles, an this tenency can explain the ifference in the sizes shown in Fig. 5. In copper particles of 20 nm-iameter or more, it seems that the excessive heat results in ecomposition of the oleic aci coating, as shown in Fig. 6(c). Conclusions Preparation of copper nanoparticles covere with organic matter was carrie out an analyses of the morphology, the molecular structure an the crystal structure were conucte to investigate such a nanoparticle synthesis process. The following conclusions have been obtaine. (1) Copper nanoparticles covere with organic matter have been successfully prepare by evaporation of copper wire in an oleic aci vapor/mist. The thickness of the coating layer was a few nanometres. (2) The meian iameter of the power was 25 nm, an became 10 nm smaller than that without the coating. (3) The coating matter has not kept its original molecular structure, because of the ecomposition by the heat from a growing particle. (4) The copper nanoparticles covere with organic matter have not oxiize several months after the preparation. Acknowlegments This work was partly supporte by the 21 st Century COE Project, Ministry of Eucation, Culture, Sport, Science an Technology, Japan. The authors acknowlege Prof. H. Saito an Mr. Ohshio in Nagaoka University of Technology for their help in FT-IR analysis. References 1. R. Kubo, J. Phy. Soc. Jpn. 17[6] (1962) M. Blackman an J.R. Sambles, Nature 226 (1970) F.E. Kruis, H. Fissan, an A. Pele, J. Aerosol Sci. 29[5/6] (1998) S. Yatsuya, Y. Tsukasaki, K. Mihama an R. Uyea, J. Crystal Growth 45 (1978) Y.S. Kwon, A.A. Gromov, A.P. Ilyin, A.A. itts, J.S. Kim, S.H. Park an M.H. Hong, International J. Refractory Metals an Har Mater. 22 (2004) T. Nakayama, Y. Yamamoto, T. Sekino, T.A. Yamamoto, T. Kusunose, Y. Hayashi an K. Niihara, Proc. of 6 th Seminar on Core Univ. Prog. between Japan an Korea, (2002) F.G. Karioris an B.R. Fish, J. Colloi Sci. 17 (1962) W. Jiang an K. Yatsui, IEEE Trans. Plasma Sci. 26[5] (1998) Y.A. Kotov, J. Nanoparticle Res. 5 (2003) T. Suzuki, K. Keawchai, W. Jiang, an K. Yatsui, Jpn. J. Appl. Phys. 40 (2001) Y. Kinemuchi, K. Murai, C. Sangurai, C. Cho, H. Suematsu, W. Jiang, an K. Yatsui, J. Am. Ceram. Soc. 83[3] (2003) Stanar formation enthalpy of gas from NIST Chemistry Webbook, Enthalpy change from room temperature to the boiling point was referre from Chase. M.W. Jr., NIST-JANAF Themochemical Tables, Fourth Eition, J. Phys. Chem. Ref. ata, Monograph 9 (1998) K. Murai, C. Cho, H. Suematsu, W. Jiang an K. Yatsui, IEEJ Trans. FM 125[1] (2005) C. Cho, K. Murai, T. Suzuki, H. Suematsu, W. Jiang an K. Yatsui, IEEE Trans. Plasma Sci 32[5] (2004) Eit committee of Chemical ictionary: Encyclopeia Chimica, (Kyoritsu publisher, Japan, 1977), 2, p.206.