The catalytic reduction of carbon dioxide to carbon onion particles by platinum catalysts

Letters to the Editor / Carbon 43 (2005) 2618–2641

2621

Fig. 5. Scheme of elongating a-CT.

Acknowledgement We thank the National Science Council of Taiwan, the Republic of China (Projects NSC-91-2113-M-009020 and NSC-91-2218-E-007-002) for financial support. References [1] Angus JC, Koidl P, Domitz S. Plasma deposited thin films. In: Mort J, Jansen F, editors. Boca Raton Florida: CRC Press; 1986. p. 89. [2] Khan RUA, Silva SRP. A review of the effects of carbon selfimplantation into amorphous carbon. Diamond Relat Mater 2001;10:224–9.

[3] Scharff P. New carbon materials for research and technology. Carbon 1998;36:481–6. [4] Gerstner EG, Lukins PB, McKenzie DR. Substrate bias effects on the structural and electronic properties of tetrahedral amorphous carbon. Phys Rev B 1996;54:14504–10. [5] Kyotani T, Tsai L, Tomita A. Preparation of ultrafine carbon tubes in nanochannels of an anodic aluminium oxide film. Chem Mater 1996;8(8):2109–13. [6] Chang Y-H, Wang L-S, Chiu H-T, Lee C-Y. SiCl3CCl3 as a novel precursor for chemical vapor deposition of amorphous carbon film. Carbon 2003;41:1169–74. [7] Wang L-S, Lee C-Y, Chiu H-T. New nanotube synthesis strategy—application of sodium nanotubes formed inside anodic aluminium oxide as a reactive template. Chem Commun 2003:1964–5.

The catalytic reduction of carbon dioxide to carbon onion particles by platinum catalysts Marco G. Crestani a, Ivan Puente-Lee a, Luis Rendo´n-Vazquez b, Patricia Santiago b, Federico del Rio c, David Morales-Morales c, Juventino J. Garcı´a a,* a

Facultad de Quı´mica, Universidad Nacional Auto´noma de Me´xico, Circuito Interior. Cd. Universitaria, Me´xico, D.F. 04510, Mexico b Instituto de Fı´sica, Universidad Nacional Auto´noma de Me´xico, Me´xico, D.F. 04510, Mexico c Instituto de Quı´mica, Universidad Nacional Auto´noma de Me´xico, Me´xico, D.F. 04510, Mexico Received 15 February 2005; accepted 28 April 2005 Available online 1 July 2005

Keywords: Carbon onions; Catalytically grown carbon; Activation; Catalyst; Nuclear magnetic resonance

Since their discovery, carbon nanomaterials [1] have attracted a great deal of interest from many research groups worldwide, since they exhibit unique chemical and physical properties [2]. Both carbon nanotubes (CNT) and carbon onion particles (COP) have been prepared by a variety of methods that can be classified into

*

Corresponding author. Tel.: +52 55 56223514; fax: +52 55 56162010. E-mail address: [email protected] (J.J. Garcı´a). 0008-6223/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2005.04.029

two broad categories: physical and chemical. These techniques have been reviewed recently [3], the physical methods include electric arc discharge synthesis, laser ablation, resistivity vaporization, electron or ion beam vaporization and sunlight vaporization. The chemical methods are, for example, the catalytic pyrolysis reactions of hydrocarbons, catalytic disproportion of carbon monoxide (CO) [4,5], the reduction of perfluorinated hydrocarbons by an alkali metal amalgam, hydrothermal growth from amorphous carbon, catalytic reduction of CO and the thermal decomposition of Fe(CO)5 in a

2622

Letters to the Editor / Carbon 43 (2005) 2618–2641

Fig. 1. HRTEM pictures of the product after workup. (A) Nanoonions, (B) twin spiral cores, (C) tilted multilayers.

flow of CO [6]. Among the chemical methods there are two reports for the use of carbon dioxide (CO2) as a carbon source to produce carbon nanomaterials, both of which use supercritical CO2, one with melted metallic magnesium at 1000 C and a pressure of 10,000 atm [7] and the other one with melted metallic lithium at 550 C and a pressure of 700 atm [8]. Because carbon dioxide is the final product of many industrial and biological processes, being the former a major source of global warming and since only plants can effectively regenerate useful products from it, the aim of the present report is to develop a low-temperature process for the preparation of carbon nanomaterials using CO2 as the carbon source, based on the use of platinum thiaplatinacycles as catalyst precursors. For a number of years we have been studying the synthesis of thiaplatinacycles of the type [Pt(g2-C,S-thiophene)(PEt3)2] and their role in homogeneous hydrodesulfurization (HDS) reactions [9,10] the former being useful intermediates for organic transformations [11] and active precursors in homogeneous HDS catalytic reactions [12]. In the present study, we investigated the reaction of CO2 gas with thiaplatinacycles, such as [Pt(g2C,S-C12H8)(PEt3)2], 1, in a stainless-steel autoclave (300 mL), charged also with hydrogen and pressurized with argon up to 1380 psi at 40 C (approx. 94 atm) to ensure the CO2 being in a supercritical state (critical point, 31 C, 73 atm), all used gases were in a high purity grade 99.998%. The main product identified was carbon, particularly in the form of carbon nanomaterials. Our interpretation of the chemical reaction that occurs is proposed as:

ing structures turned out to be carbon nanomaterials, the main products (80%) being the well-known COP [13], Fig. 1A. Also, a few closely related particles with two spiral cores were observed [14], Fig. 1B, along with some tilted layers of carbon, characteristic of the ones present in multilayer CNT, Fig. 1C. The X ray diffraction (XRD) pattern of the product shows a very small peak at 2h = 26.6, regularly associated with the presence of graphite; the intensity of such peak is very small (Fig. 2), which is consistent with the low temperature used in the process [8]. Solid state 13C nuclear magnetic resonance (NMR) studies of the product obtained, were carried out by several groups to further characterize carbon nanomaterials (Fig. 3). The signal in a typical sample, located at d 110.0, is considerably narrower (half-height line width is only about 10 ppm) than those observed previously [15], which may be attributed to uniform composition, packing or relatively few defects [16]. The absence of signals around 35 ppm and 179 ± 10 ppm implies that there are neither hydrocarbon nor graphite impurities, respectively. The reaction, which was performed with thiaplatinacycles derived from dibenzothiophene [Pt(g2-C,S-

CO2 þ 2H2 ! C þ 2H2 O The HRTEM (high resolution transmission electron microscopy) micrographs depicted in Fig. 1 show the major structures obtained after workup. All the result-

Fig. 2. XRD pattern for the sample.

Letters to the Editor / Carbon 43 (2005) 2618–2641

2623

This research demonstrates that useful products like carbon nanomaterials, particularly COP, can be prepared by a rather simple chemical method, under catalytic conditions and very mild pressure and temperature. Acknowledgements We thank DGAPA-UNAM (IN-205603) and CONACYT (C02-42467) for support and Dr. Alma Are´valo for technical assistance. Appendix A. Supplementary data

Fig. 3. Solid state

13

C NMR.

C12H8)(PEt3)2], 1, and thiophene [Pt(g2-C,SC4H4)(PEt3)2], 2, as catalytic precursors, did not proceed when the analog derived from benzothiophene [Pt(g2C,S-C8H6)(PEt3)2], 3, was used. This difference in reactivity can be explained on account of the higher stability toward the reductive elimination reaction of complex 3 to release the benzothiophene ligand, which has been demonstrated by our group [9]. Consequently, compounds 1 and 2 are better precursors of the ‘‘Pt(PEt3)2’’ moiety, which may be involved in the catalytic cycle. Experiments adding metallic mercury were performed to rule against a heterogeneous system [17]. Although none of the experiments using 1 or 2 was inhibited by mercury, the reaction was probably quenched by the H2O produced, which typically occurs approximately after 40 turnovers for 1 and 22 turnovers for 2. The use of a diphosphine as 1,2-bis(diphenylphosphino)ethane (dppe) as an ancillary ligand, in the thiaplatinacycle [Pt(g2-C,S-C12H8)(dppe)], 4, gave an improved catalysis up to 60 turnovers (calculated as moles of product per mole of catalyst). Detailed studies are currently underway to establish the nature of some of the intermediaries involved in the catalytic cycle; however, we can anticipate the reductive elimination of the thiophenic moiety, followed by CO2 coordination and its stepwise reduction to CO and finally to carbon and H2O as products. The total release of dibenzothiophene from 1 and 4 and thiophene from 2 was monitored and quantified by gas chromatography coupled to mass spectrometry (GC–MS) determinations at the end of the reaction. The final fate of platinum was metallic platinum, which was also detected by HRTEM (mean particle size 13 nm). None of the metallic particles was nested or encapsulated within the carbon nanomaterials; most of it can be eliminated with acid washings without affecting the COP obtained, also the PEt3 or dppe were totally recovered and quantified by GC–MS. Yield of product after workup was 0.2%.

Experimental details for the catalytic experiments, DfG0 calculation and additional HRTEM and SEM micrographs. Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbon.2005.04.029. References [1] Iijima S. Helical microtubules of graphitic carbon. Nature 1991;354(6348):56–8. [2] Harris PFJ. Carbon nanotubes and related structures. New York: Cambridge University Press; 1999. [3] Rakov EG. Methods for preparation of carbon nanotubes. Russ Chem Rev 2000;69(1):35–52. [4] Nolan PE, Schabel MJ, Lynch DC, Cutler AH. Hydrogen control of carbon deposit morphology. Carbon 1995;33(1):79–85. [5] Nasibulin AG, Moisala A, Brown DP, Kauppinen EI. Carbon nanotubes and onions from carbon monoxide using Ni(acac)(2) and Cu(acac)(2) as catalyst precursors. Carbon 2003;41(14): 2711–24. [6] Nikolaev P, Bronikowski MJ, Bradley RK, Rohmund F, Colbert DT, Smith KA, et al. Gas-phase catalytic growth of single-walled carbon nanotubes from carbon monoxide. Chem Phys Lett 1999;313(1–2):91–7. [7] Motiei M, Hacohen YR, Calderon-Moreno J, Gedanken A. Preparing carbon nanotubes and nested fullerenes from supercritical CO2 by a chemical reaction. J Am Chem Soc 2001;123(35):8624–5. [8] Lou ZS, Chen QW, Wang W, Zhang YF. Synthesis of carbon nanotubes by reduction of carbon dioxide with metallic lithium. Carbon 2003;41(15):3063–7. [9] Garcia JJ, Mann BE, Adams H, Bailey NA, Maitlis PM. Equilibria of the thiametallacycles with tris(triethylphosphine)platinum(0) and dibenzothiophene, benzothiophene, or thiophene—the hydrodesulfurization reaction. J Am Chem Soc 1995;117(8):2179–86. [10] Arevalo A, Bernes S, Garcia JJ, Maitlis PM. Ring opening of methylbenzothiophenes and methyldibenzothiophenes by tris(triethylphosphine)platinum(0). Organometallics 1999;18(9):1680–5. [11] Garcia JJ, Arevalo A, Montiel V, DelRio F, Quiroz B, Adams H, et al. Analysis of a hydrodesulfurization process. 3. Acid cleavage of thiaplatinacycles. Organometallics 1997;16(14):3216–20. [12] Hernandez M, Miralrio G, Arevalo A, Bernes S, Garcia JJ, Lopez C, et al. Reactivity of substituted thiophenes toward tris(triethylphosphine)platinum(0), -palladium(0), and -nickel(0). Organometallics 2001;20(19):4061–71. [13] Ugarte D. Curling and closure of graphitic networks under electron-beam irradiation. Nature 1992;359(6397):707–9.

2624

Letters to the Editor / Carbon 43 (2005) 2618–2641

[14] Ozawa M, Goto H, Kusunoki M, Osawa E. Continuously growing spiral carbon nanoparticles as the intermediates in the formation of fullerenes and nanoonions. J Phys Chem B 2002;106(29):7135–8. [15] Selvan R, Unnikrishnan R, Ganapathy S, Pradeep T. Macroscopic synthesis and characterization of giant fullerenes. Chem Phys Lett 2000;316(3–4):205–10.

[16] Peng HQ, Alemany LB, Margrave JL, Khabashesku VN. Sidewall carboxylic acid functionalization of single-walled carbon nanotubes. J Am Chem Soc 2003;125(49):15174–82. [17] Widegren JA, Finke RG. A review of the problem of distinguishing true homogeneous catalysis from soluble or other metalparticle heterogeneous catalysis under reducing conditions. J Mol Catal A 2003;198:317–41.

An easy method for the synthesis of ordered microporous carbons by the template technique Peng-Xiang Hou, Toshiaki Yamazaki, Hironori Orikasa, Takashi Kyotani

*

Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Aoba-ku, Sendai 980-8577, Japan Received 23 February 2005; accepted 3 May 2005 Available online 23 June 2005

Keywords: Porous carbon; Chemical vapor deposition; BET surface area; Microporosity

Recently we prepared a long-range ordered microporous carbon with a structural regularity of zeolite Y [1– 3]. This microporous carbon possesses a high BET specific surface area more than 3000 m2/g and almost no mesoporosity. Furthermore, its pore size distribution is very narrow in comparison with commercial high surface area carbons and most of the pore sizes are in the range of 1.0–1.5 nm [4]. Such monodispersed pore size distribution is suitable for the application to the storage media for methane gas and the electrodes of electric double-layer capacitor. However, the carbon preparation method is rather complicated, because it always requires many different types of steps as follows. Zeolite Y was impregnated with furfuryl alcohol (FA), which was polymerized inside the zeolite channels. The resulting polyfurfuryl alcohol (PFA)/zeolite composite was heated to 700 C and then propylene chemical vapor deposition (CVD) was performed at this temperature for 1 h. After the CVD, the composite was further heat-treated at 900 C and the resultant carbon was liberated from the zeolite framework by HF washing. If the use of FA could be avoided, the preparation method would become much simpler. Moreover, many laborious operations such as stirring, filtering and drying for the wet impregnation step could be omitted. Here we try to synthesize similar ordered microporous carbons only by a dry CVD process. *

Corresponding author. Tel.: +81 22 217 5625; fax: +81 22 217 5626. E-mail address: [email protected] (T. Kyotani).

0008-6223/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2005.05.001

In the previous report [3], we concluded that both of the two processes (the PFA carbonization and the subsequent propylene CVD) were indispensable to obtain the ordered microporous carbon with a large surface area and large micropore volume, because each single process did not allow us to prepare such porous carbon. For the PFA carbonization, the amount of the carbon derived from PFA was not large enough to preserve the regularity of the zeolite structure. On the other hand, the single CVD process deposited large amount of carbon (when the CVD temperature and/or period are increased), but carbon deposition on the external surface of zeolite particles was inevitable and the regularity of the resultant carbons was found to be quite low. For diminishing such unnecessary carbon deposit, the CVD temperature should be as low as possible. Moreover, if the molecular size of CVD gas is small in comparison with the inner diameter of the zeolite channels, the diffusion through the channels would be easy and therefore the gas molecules could go inside without serious pyrolytic decomposition on the external surface. With this assumption in mind, we use acetylene as CVD gas and perform acetylene CVD at as a low temperature as 600 C. Briefly, acetylene CVD (5% in N2) was performed over powdered zeolite Y (Na-form, SiO2/Al2O3 = 5.6) at 600 C for 4 h and then a portion of the zeolite/carbon composite powder was subjected to further CVD using acetylene (5% in N2) or propylene (2% in N2) at 700 C for 1 h. Finally, the heat-treatment at 900 C in