Vol.9 No.3 2017
61/74

Research paper : A super-growth method for single-walled carbon nanotube synthesis (K. Hata)−174−Synthesiology - English edition Vol.9 No.3 (2017) However, from the perspective of industrial mass production, the sputtering method had low productivity and high facility costs, and that meant high overall costs. Therefore, it was necessary to shift to wet catalysts that were low cost, needed small facility investment, and had high productivity. In the sputtering process when the article was published in Science, the iron catalyst was sputtered immediately after sputtering the alumina catalyst support with RF. However the RF sputtering of alumina had extremely slow lm forming rate, and the productivity of the catalyst was significantly low. Therefore, we developed the technology in which the aluminum was DC sputtered in an atmosphere containing oxygen thereby oxidizing the aluminum on the spot to form the alumina. Through the series of research on catalyst development using sputtering, it was found that various factors such as the composition of alumina catalyst support and surface smoothness greatly affected the growth of single-walled CNTs.Thinking that the thickness restriction for the iron catalyst was too strict and that it was difcult to apply the iron thin film evenly within the allowed range, we developed iron colloid nanoparticles as wet catalysts synthesized in the iron carboxyl solution (Fig. 11).[13] By thinly coating the iron colloid nanoparticles onto the silicon substrate using spin coating, we were able to grow single-walled CNTs similar to the ones by sputtering thin films. However, when it was investigated closely, the single-walled CNTs did not grow from individual iron colloid nanoparticles. It was found that the iron colloid nanoparticles fused when the catalyst was reduced by hydrogen before synthesis, and then turned back into ne particles again. Therefore, when wet catalysts were used, application of a thin coat to the iron catalyst lm was also sufcient.We first looked at the method called capillary coating as a method of coating the ultra-thin iron lm. This was a method for coating the substrate material with a solution containing iron salt that was sucked into the ultra-fine tubes by the capillary effect, and then used for evenly coating substances such as liquid crystals. Using this method, we succeeded in coating ultra-thin iron catalysts onto a at substrate material, but due to its principle, we were unable to apply the solution evenly to a substrate with distortion or deformation. From the experiments for large-area synthesis, we found that the distortion and deformation of substrates increased as the substrate material increased in surface area and as the substrate material was reused repeatedly. Therefore, we were forced to develop a totally different method. After many twists and turns, we nally succeeded in coating the ultra-thin iron lm onto large-area deformed substrate material by the wet method, and this was a very inexpensive and highly productive method for growing single-walled CNTs.Next, like the iron catalyst, we conducted technological development for coating with alumina catalyst supports. Since alumina had extremely high carburizing resistant property, the substrate material consisting of Ni-Fe-Cr alloys coated with alumina catalyst supports showed excellent durability in the highly concentrated hydrocarbon environment used in the super-growth method. The catalyst we developed became a system that powerfully inhibited the deformation and carburization that were problems in reusing the large-area substrate material.4.1.3 Development of the synthesis technology optimal for mass productionIt was necessary to conduct much technological development for the mass production process in synthesis technology. When the article was published in Science, the CNT was synthesized by placing the substrate horizontally in a horizontal synthesis furnace of 1 inch diameter and supplying ethylene and water vapor gas from the side. This small, horizontal synthesis furnace had an optimal structure for creating a laminar flow and preventing gas turbulence. Any gas turbulence significantly decreased the synthesis efficiency of CNTs, and such equipment configuration was optimal in lab scale synthesis. However, this lab scale synthesis furnace or substrate material could not be upscaled to an industrial scale. There was a major problem that most of the supplied gas did not hit the substrate, passed over the catalyst without reaction, and only about 1 % of the supplied ethylene gas was converted to CNTs. It was necessary to supply the gas from the top to enable upscaling of the synthesis furnace and substrate material, and to greatly improve the conversion rate of the carbon source to CNTs. Therefore, we developed a showerhead (Fig. 12).[14] Various changes were made to the showerhead to evenly supply extremely minute amount of water to the catalysts on the substrate material. Next, the synthesis furnace was made vertical and the diameter was increased from 1 inch, to 2 inches, and then 4 inches. When the furnace was turned vertical and the diameter increased, turbulence occurred immediately. We conducted fluid simulation and Journal of Physical Chemistry C. 111, 17961 (2007)CVDSpin coatingFe colloidFig. 11 Development of coated wet catalyst[13]

元のページ  ../index.html#61

このブックを見る