본문 바로가기
  • Home

Open-atmosphere spinning of carbon nanotube fibers sans hydrogen flow by floating catalyst chemical vapor deposition: an insight into the mechanism

  • Carbon Letters
  • Abbr : Carbon Lett.
  • 2025, 35(3), pp.1125~1138
  • DOI : 10.1007/s42823-024-00843-w
  • Publisher : Korean Carbon Society
  • Research Area : Natural Science > Natural Science General > Other Natural Sciences General
  • Received : July 8, 2024
  • Accepted : January 20, 2025
  • Published : June 5, 2025

Alexander Rajath 1 Kaushal Amit 1 Singh Jaspreet 2 Dasgupta Kinshuk 1

1Materials Group, Bhabha Atomic Research Centre
2Technical Physics Division, Bhabha Atomic Research Centre

Accredited

ABSTRACT

This study introduces a novel method for synthesizing carbon nanotube (CNT) fibers using floating catalyst chemical vapor deposition (FC-CVD) in an open-atmosphere without the need for hydrogen as a carrier gas. Traditional FC-CVD techniques depend on hydrogen gas and require a harvest box with inert gas purging, which restricts scalability. Our approach utilizes nitrogen gas as the sole carrier, allowing for CNT fiber production without a harvest box. To understand the spinning process mechanism in an open-atmosphere, we conducted thermodynamic and computational fluid dynamics (CFD) analyses. Methanol was selected as the carbon source based on thermodynamic calculations, which revealed that at high temperatures, methanol forms CO and H2 as thermodynamically stable species instead of carbon (C), thereby preventing soot formation. Moreover, methanol undergoes catalytic cracking exclusively in the presence of catalysts, further preventing soot formation. This approach allows operation at high partial pressure, even above the upper explosive limit (UEL), effectively preventing combustion. A 600 mm cooling zone was incorporated into the reactor to lower the outlet gas temperature below methanol's auto-ignition point, mitigating combustion risks. CFD calculations were employed to determine the necessary cooling zone length. Additionally, we developed a predictive model using the XGBoost machine learning method to efficiently map the parameter space for CNT fiber spinning, achieving an accuracy of 95.24%. The resulting CNT fibers demonstrate high electrical conductivity (240 ± 24 S/cm) and a low ID/IG ratio, indicating a high degree of crystallinity.

Citation status

* References for papers published after 2024 are currently being built.