Encapsulation mechanism of N 2 molecules into the central hollow of carbon nitride multiwalled nanofibers

Transcription

1 Carbon 44 (2006) Encapsulation mechanism of N 2 molecules into the central hollow of carbon nitride multiwalled nanofibers Ji Hoon Yang a, Dae Ho Lee a, Min Hyung Yum a, Yong Sook Shin a, Eun Jung Kim a, Chong-Yun Park a, *, Myeung Hoi Kwon a,b, Cheol Woong Yang c, Ji-Beom Yoo c, Ha-Jin Song d, Hyun-Joon Shin d, Yong-Wan Jin e, Jong-Min Kim e a Center for Nanotubes and Nanostructured Composites (CNNC), Institute of Basic Science, Sungkyunkwan University, 300 Chunchun-dong, Suwon , Republic of Korea b Physics Department, University of Incheon, 177 Dowha-dong, Incheon , Republic of Korea c CNNC, School of Metallurgy and Materials Engineering, Sungkyunkwan University, 300 Chunchun-dong, Suwon , Republic of Korea d Beamline Research Division, Pohang Accelerator Laboratory, POSTECH, Pohang, Kyunbuk , Republic of Korea e Materials Lab., Samsung Advanced Institute of Technology, P.O. Box 111, Suwon , Republic of Korea Received 5 October 2005; accepted 28 February 2006 Available online 17 April 2006 Abstract Nitrogen molecules have been encapsulated into the central hollows of vertically aligned carbon nitride (CN) multiwalled nanofibers by dc plasma-enhanced chemical vapor deposition with C 2 H 2,NH 3, and N 2 gases on a Ni/TiN/Si(100) substrate at 650 C. X-ray photoelectron spectroscopy and near-edge X-ray absorption fine structure spectra showed the existence of nitrogen molecules in CN nanofibers. Elemental mapping images with electron energy loss spectroscopy of the CN nanofiber and catalyst metal, and optical emission spectroscopy spectra of the plasma showed the distribution of nitrogen atoms and molecules in the CN nanofiber, catalyst metal, and gaseous precursor, respectively. These studies showed that atomic nitrogen diffused into the catalytic metal particle because of the concentration gradient and then saturated at the bottom of the particle. Saturated nitrogen atom participated in the formation of the CN nanofiber wall but most of nitrogen was trapped in the central hollow of the nanofiber as molecules. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Carbon nanofibers; Chemical vapor deposition; Scanning electron microcopy; Electron energy loss spectroscopy; Gas storage 1. Introduction Ever since the discovery of carbon nanotubes (CNTs), scientists have been investigating the specific properties of CNTs [1 5] and the possible roles CNTs might play in various kinds of nanotechnologies and molecular electronics [6,7]. In particular, many researchers have investigated the possibility of CNTs being used as energy storage devices such as in hydrogen fuel cells by taking advantage of the high-surface-area of single-walled and multiwalled CNTs [8]. Almost all of their efforts were focused on the * Corresponding author. Fax: address: cypark@skku.edu (C.-Y. Park). interaction between hydrogen and CNTs through physisorption or chemisorption. Recent studies have revealed that Ar and N 2 molecules can also be stored in CNTs [9,10]. The experiments involving Ar introduced Ar gas into CNTs for 48 h at 650 C under an Ar pressure of 170 MPa. However for those involving N 2 storage, nitrogen molecules were trapped directly in the hollow core of the CNTs during the growth procedure. These observations may provide the way for the fabrication of gas storage components using CNTs. However, the encapsulation mechanism of gaseous molecules is, as yet, unclear. Here, we report an alternative method for storing nitrogen molecules in carbon nitride (CN) multiwalled nanofibers /$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi: /j.carbon

2 2220 J.H. Yang et al. / Carbon 44 (2006) (CN-MWNFs) with central hollow using dc plasmaenhanced chemical vapor deposition (PECVD) methods during the growth with a Ni catalytic material and gas mixture of acetylene, ammonia and nitrogen. In addition, we propose the encapsulation mechanism of nitrogen molecules based on the results obtained from the near-edge X-ray absorption fine structure (NEXAFS) and the mapping of the elements using transmission electron microscope (TEM). 2. Experimental CN-MWNFs with a central hollow were grown using a dc PECVD technique on a 100 nm thick Ni film which was used as a catalytic layer on the base TiN/Si(1 0 0) substrate. In order to enhance the adhesion of the catalytic layer, a layer of TiN at a thickness of 100 nm was deposited on the Si(100) substrate. The substrate was loaded into the dc PECVD system which was evacuated to a pressure of Torr. A gas mixture of C 2 H 2 (10 sccm), NH 3 (40 sccm), and N 2 (60 sccm) was used as precursor. The working pressure was 5 Torr. The substrate was heated to 650 C using a tungsten filament and the temperature was monitored using a thermocouple. The growth of CN- MWNFs was carried out for 15 min. The structure of the CN-MWNFs and the distribution of the elements in the CN-MWNFs were measured by scanning electron microscope (SEM) (JEOL, JSM6700F) and TEM (JEOL, JEM2100F)/electron energy loss spectrometer (EELS) (Gatan Image Filter 2001). The electronic structure of the CN-MWNFs was analyzed from NEXAFS and X-ray photoelectron spectroscope (XPS) (VSW 125) data taken at 8A1 beamline of Pohang Accelerator Laboratory. The NEXAFS at N 1s absorption edge was measured with an energy resolution of approximately 80 mev. The composition of the plasma in the dc PECVD chamber was monitored by optical emission spectrometer (OES) (SC Technology, PCM402). 3. Results and discussion Fig. 1(a) shows a typical SEM image of an array of vertically aligned CN-MWNF with an average diameter of 50 nm and length of 10 lm. As can be seen from the TEM images in Fig. 1(b) and (c), the CN-MWNF has a stacked cone arrangement with small angle between the graphite basal planes and the tube axis [11] and a compartmented central hollow, which is similar to that observed in CN x nanotube [12] with Ni catalysts at the tip of the tube. Fig. 2(a) shows wide scan XPS spectra of as-grown CN- MWNFs and annealed CN-MWNFs at 900 C. N 1s peaks at 399 ev and 401 ev, a C 1s peak at ev and an O 1s peak at 528 ev were observed. The peak at 399 ev reveals the presence of pyridinic nitrogen, while the peak at 401 ev comes from the trivalent nitrogen replacing the carbon in the hexagonal structure [13]. The XPS spectra clearly show that the CN-MWNFs contain nitrogen at temperatures up to 900 C. Fig. 1. (a) SEM image. (b) TEM image of nitrogen-incorporated CN- MWNFs. (c) Magnification of TEM image. Fig. 2(b) shows NEXAFS spectra at the N 1s absorption edge of the CN-MWNFs taken at several elevated temperatures. Here, N 1s! p * excitation peaks marked with arrows a, b and c are at 399.1, 399.8, and ev, respectively and the N 1s! r * excitation peak, labeled d, is at ev. These peak positions and shapes are close to those of CN x thin films [14] except peak c. As seen in the NEXAFS spectrum, it is noticeable that peak c consists of several peaks, which was not visible on the CN x thin film spectra, and very similar to the N 1s! p * excitation peaks of molecular N 2 in both photon energy and profile. This strongly suggests that molecular nitrogen is present within the CN-MWNFs [15]. These peaks did not change, even after being annealed at temperatures up to 530 C. If the nitrogen molecules were adsorbed on the surface of the CN-MWNFs, annealing at high temperatures such as 530 C would desorb the molecules from the surface [16]. For a more detailed investigation of the profile of peak c, highly resolved spectra were acquired with high-energy resolution as shown in Fig. 2(c). Peak c appears to consist of more than six peaks and did not change in either intensity or shape with annealing at temperatures up to 530 C as mentioned earlier. After annealing at 900 C, the intensity of peak c decreased and broadened out, because of a reaction between the nanofiber wall and molecular nitrogen at such high temperatures.

3 J.H. Yang et al. / Carbon 44 (2006) Fig. 2. (a) XPS spectra of as-grown CN-MWNFs and annealed CN-MWNFs at 900 C. (b) NEXAFS spectra from as-grown and the annealed CN- MWNFs at 50, 350, 530, and 900 C. (c) Fine-scanned NEXAFS spectra of as-grown and annealed CN-MWNFs at 50, 350, 530, and 900 C. (d) Fitted NEXAFS spectrum by Voigt function of as-grown CN-MWNFs. The as-grown spectrum showed a good fit to the eight components (dotted curves) at 400.9, 401.1, 401.3, 401.5, 401.8, 402.0, and ev using Gaussian functions, asymmetric Gaussian function and an error function. For N 2 gas, eight peaks appeared corresponding to the eight vibration levels of N 2 gas in N 1s! p * absorption [15]. A specific difference between spectra from gas and the CN-MWNFs is that all peaks of CN-MWNFs are broader and the intensity of the second peak is relatively stronger than that of the gas. It appears that these phenomena are attributed to the existence of small quantities of nitrogen bonded to carbon within the CN-MWNFs. These analyzed results reveal clearly that free nitrogen molecules exist in the CN-MWNFs. The distribution of nitrogen in the CN-MWNFs was examined by mapping the elements using TEM with a Gatan image filter (GIF). Fig. 3(a) (d) shows the zero-loss, carbon-mapping and nitrogen-mapping images, and a line profile, taken from the CN-MWNFs. The zero-loss image of the nanofiber, as shown in Fig. 3(a), shows a bamboolike structure with hollows at center. Fig. 3(b) shows that the density of carbon in the hollow of the CN-MWNFs is significantly lower than that found in the wall. In contrast to the distribution of carbon, Fig. 3(c) shows that most of the nitrogen is present in the innermost hollow of the CN-MWNFs. As mentioned earlier in NEXAFS analysis, it strongly suggests that most of the nitrogen is in molecular form within the hollows of the tube. However, a small amount of nitrogen still exists in the nanofiber wall. These results are confirmed again by the line profiles of the carbon- and nitrogen-mapping images, shown in Fig. 3(d). In Fig. 3(c), it is shown that the nitrogen signal is enhanced at the outermost surface of the CN-MWNF. A thermal drift and vibration of CN-MWNFs during TEM analysis is inevitable because of long and thin structure of CN-MWNFs. Those thermal drift and vibration of CN-MWNFs are reasons of a blur or overlap images in TEM analysis. In order to investigate the mechanism for the encapsulation of the N 2 molecules, the carbon and nitrogen distributions in the Ni catalyst tip were measured using GIF analysis. Fig. 3(e) (h) shows zero-loss, nickel-mapping, carbon-mapping, and nitrogen-mapping images of the Ni catalyst tip. In Fig. 3(e), the zero-loss image of the Ni tip indicates that the nanofiber showed tip-growth (or topgrowth). Fig. 3(f) shows the nickel-mapping image has many areas where the contrast is non-uniform. This indicates that the shape of the Ni particles inside the tip is not symmetric or uniform. The carbon- and nitrogen-mapping images in Fig. 3(g) and (h) shows clearly that carbon and nitrogen are distributed over both the surface and inside of the Ni tip. It seems that the contrast visible in the images is related to the shape of the Ni particles in both cases. These results strongly suggest that carbon and nitrogen diffused into the inside via top surface of Ni and then became segregated during the growth. According to previous reports, there are three primary supply routes for the carbon source required for the growth of the CNT: surface diffusion [17], bulk diffusion [18],anda combination of both [19]. The catalyst particle melted at

4 2222 J.H. Yang et al. / Carbon 44 (2006) Fig. 4. (a) OES spectra of the plasma created by C 2 H 2,NH 3, and N 2. (b) The spectra of the plasma created by C 2 H 2 and NH 3. Fig. 3. Elemental mapping images with GIF of CN-MWNFs. (a) Zeroloss image. (b) Carbon-mapping image. (c) Nitrogen-mapping image. (d) Line profile of carbon-mapping and nitrogen-mapping images. Elemental mapping images at the tip of CN-MWNFs. (e) Zero-loss image. (f) Nickelmapping image. (g) Carbon-mapping image. (h) Nitrogen-mapping image. relatively low temperatures due to the effects of the plasma in the CVD process [20] and the effect of the nanoparticle size [21,22]. Thus, carbon diffusion into the melted Ni was enhanced. However, in this experiment, it was also observed that nitrogen, as well as carbon, initially diffused into the bulk Ni during the growth phase. This strongly suggests that the carbon became saturated at the bottom of the Ni tip and formed a wall in the CN-MWNF and that the nitrogen also diffused to the bottom of the catalyst tip through the bulk, became saturated and then trapped in the CN-MWNF in molecule form. However, some of the nitrogen atoms bonded with the carbon atoms and formed a wall in the CN-MWNF. During nitrogen diffusion into the catalyst tip, nitrogen must be in the atomic phase than the molecular phase. To investigate the source of the atomic nitrogen, the constituents of the plasma in the chamber were measured using OES, as shown in Fig. 4 and analyzed. Fig. 4(b) shows primarily CN radicals present in the plasmas created with C 2 H 2 and NH 3. When N 2 gas was then introduced additionally into the chamber, the following various constituents were observed in the plasmas as shown in Fig. 4(a): CN, N 2,N þ 2, and NH, all of which, are potential sources of atomic nitrogen. It was reported that CN dissociates to form atomic carbon and nitrogen by annealing about 580 C [16]. Therefore, it appears that from among these components, CN radicals are the most important source of atomic nitrogen during the growth of CN-MWNFs [23]. We confirmed that CN-MWNFs grown with C 2 H 2 and NH 3 precursors contained nitrogen molecules in the hollow and the amount of encapsulated nitrogen molecules increased as further nitrogen molecules were added (not shown in here). Analyzing the results of the elemental mapping of the tip of the nanofiber and the results of the OES analysis, the mechanism of encapsulation of nitrogen in nanofibers was deduced as follows: (1) Precursors dissociated into various radicals including the CN radicals in the plasma. (2) CN radicals became embedded on the top surface of the melted catalyst tip. (3) CN radicals dissociated into atomic carbon and nitrogen, and then diffused into the melted tip. (4) Carbon and nitrogen atoms diffused through to the bottom of the tip via concentration gradient mechanisms and became saturated at the bottom. (5) Saturated carbon and nitrogen atoms participated in the formation of the CN- MWNF wall and a lot of released nitrogen atoms formed molecular nitrogen which then became trapped in the hollow of the CN-MWNF. Fig. 5 shows the suggested schematic of the nitrogen-encapsulation-mechanism during the growth of CN-MWNFs.

5 J.H. Yang et al. / Carbon 44 (2006) Fig. 5. Schematic of the mechanism of nitrogen encapsulation and growth of CN-MWNFs. 4. Conclusion We have investigated the growth mechanism of CN- MWNFs incorporating nitrogen and the mechanism for the encapsulation of the nitrogen molecules inside the hollow of the CN-MWNFs by using TEM/GIF, NEXAFS, SEM, OES and XPS. Through analysis of these experimental results, it was found that CN radicals in the plasma became embedded on the top surface of the melted Ni catalyst tip and dissociated into atomic phases of C and N, following which the atomic C and N then diffused into the tip. This diffusion mechanism plays a key role in the formation of CN-MWNF walls incorporating nitrogen and in the encapsulation of nitrogen molecules in the hollow of the nanofiber. To the best of our knowledge, the confinement mechanism of nitrogen molecules in the hollow of CN-MWNF has been demonstrated. It is hoped that this research will be able to supply significant information for the growth mechanism of CN x nanofibers and the encapsulation mechanism of gas molecules in nanofibers. Acknowledgements The authors acknowledge the support of the Science Research Center program (Center for Nanotubes and Nanostructured Composites at Sungkyunkwan University) of Ministry of Science & Technology and the Korea Science and Engineering Foundation, and Samsung Advanced Institute of Technology. The experiment at Pohang Light Source was partially supported by POSCO. References [1] Hamada N, Sawada S, Oshiyama A. New one-dimensional conductors: graphitic microtubules. Phys Rev Lett 1992;68(10): [2] Wilder JWG, Venema LC, Rinzler AG, Smalley RE, Dekker C. Electronic structure of atomically resolved carbon nanotubes. Nature 1998;391(6662): [3] White CT, Todorov TN. Carbon nanotubes as long ballistic conductors. Nature 1998;393(6682): [4] Kataura H, Kumazawa Y, Maniwa Y, Umezu I, Suzuki S, Ohtsuka Y, et al. Optical properties of single-wall carbon nanotubes. Synth Met 1999;103(1 3): [5] de Jonge N, Allioux M, Doytcheva M, Kaiser M, Teo KBK, Lacerda RG, et al. Characterization of the field emission properties of individual thin carbon nanotubes. Appl Phys Lett 2004;85(9): [6] Collins PG, Arnold MS, Avouris P. Engineering carbon nanotubes and nanotube circuits using electrical breakdown. Science 2001;292(5517): [7] Kong J, Franklin NR, Zhou C, Chapline MG, Peng S, Cho K, et al. Nanotube molecular wires as chemical sensors. Science 2000;287 (5453): [8] Dillon AC, Jones KM, Bekkedahl TA, Kiang CH, Bethune DS, Heben MJ. Storage of hydrogen in single-walled carbon nanotubes. Nature 1997;386(6623): [9] Gadd GE, Blackford M, Moricca S, Webb N, Evans PJ, Smith AM, et al. The world s smallest gas cylinders? Science 1997;277 (5328): [10] Terrones M, Kamalakaran R, Seeger T, Rühle M. Novel nanoscale gas containers: encapsulation of N 2 in CN x nanotubes. Chem Commun 2000(23): [11] Meyyappan M, Delzeit L, Cassell A, Hash D. Carbon nanotube growth by PECVD: a review. Plasma Sources Sci Technol 2003;12(2): [12] Terrones M, Terrones H, Grobert N, Hsu WK, Zhu YQ, Hare JP, et al. Efficient route to large arrays of CN x nanofibers by pyrolysis of ferrocene/melamine mixtures. Appl Phys Lett 1999;75(25): [13] Nath M, Satishkumar BC, Govindaraj A, Vinod CP, Rao CNR. Production of bundles of aligned carbon and carbon nitrogen nanotubes by the pyrolysis of precursors on silica-supported iron and cobalt catalysts. Chem Phys Lett 2000;322(5): [14] Shimoyama I, Wu G, Sekiguchi T, Baba Y. Study of electronic structure of graphite-like carbon nitride. J Electron Spectrosc Relat Phenom 2001; : [15] Chen CT, Ma Y, Sette F. K-shell photoabsorption of the N 2 molecule. Phys Rev A 1989;40(11): [16] van Hardeveld RM, van Santen RA, Niemantsverdriet JW. C N coupling in reactions between atomic nitrogen and ethylene on Rh(111). J Phys Chem B 1997;101(40): [17] Helveg S, López-Cartes C, Sehested J, Hansen PL, Clausen BS, Rostrup-Nielsen JR, et al. Atomic-scale imaging of carbon nanofibre growth. Nature 2004;427(6973): [18] Shiratori Y, Hiraoka H, Yamamoto M. Vertically aligned carbon nanotubes produced by radio-frequency plasma-enhanced chemical vapor deposition at low temperature and their growth mechanism. Mat Chem Phys 2004;87(1):31 8. [19] Grobert N, Terrones M, Trasobares S, Kordatos K, Terrones H, Olivares J, et al. A novel route to aligned nanotubes and nanofibres using laser-patterned catalytic substrates. Appl Phys A 2000;70 (2): [20] Okai M, Muneyoshi T, Yaguchi T, Sasaki S. Structure of carbon nanotubes grown by microwave-plasma-enhanced chemical vapor deposition. Appl Phys Lett 2000;77(21): [21] Castro T, Reifenberger R, Choi E, Andres RP. Size-dependent melting temperature of individual nanometer-sized metallic clusters. Phys Rev B 1990;42(13): [22] Kuang MH, Wang ZL, Bai XD, Guo JD, Wang EG. Catalytically active nickel {110} surfaces in growth of carbon tubular structures. Appl Phys Lett 2000;76(10): [23] Hudis M. Study of ion-nitriding. J Appl Phys 1973;44(4):