Elsevier

Chemical Engineering Science

Volume 64, Issue 20, 15 October 2009, Pages 4210-4221
Chemical Engineering Science

Thermal ignition and self-heating of carbon nanotubes: From thermokinetic study to process safety

https://doi.org/10.1016/j.ces.2009.06.072Get rights and content

Abstract

A comprehensive investigation on the kinetics of combustion of multiwall carbon nanotubes (MWCNTs) produced by vapour chemical deposition has been undertaken. The kinetics parameters were determined from isothermal and non-isothermal combustion tests i.e. by both thermal gravimetric analysis (TGA) and differential scanning calorimetry (DSC). The activation energy of CNT combustion was found to be about 150 kJ mol−1. The oxidation of carbon nanotubes (CNTs) appears to be a single step reaction well represented by a cylindrical shrinking core model. The consistency of this model was assessed both by comparing the activation energy extracted from TGA and DSC, and by theoretical considerations on the geometrical development of CNT oxidation. The influence of oxygen concentration on CNT combustion was also studied. Finally, these results combined with those obtained by self-ignition tests in baskets lead to recommendations for a safe handling and storage of CNTs.

Introduction

Carbon nanotubes (CNT) market is expected to increase in the coming years and will probably reach over $800 millions by 2011 (Oliver, 2007). In fact, CNTs are considered to be one of the most promising nanomaterials in the future. Consequently, the occurrence of risks related to such materials will certainly increase, which imposes the evaluation of potential environmental and occupational hazards, especially flammability ones. So far, literature studies concerning the evaluation of CNT flammability essentially aimed to determine their reactivity and the related oxidative phenomena by electron microscopic analysis (SEM & TEM) and qualitative thermogravimetric analysis (Andrews et al., 2001; Bom and Andrews, 2002; Chiang et al., 2001; Landi et al., 2005; Morishita and Takarada, 1997; Pang et al., 1993; Shimada et al., 2004; Yang et al., 2002); only two TGA quantitative studies on CNTs were reported in 2007 (Brukh and Mitra, 2007; Illekova and Csomorova, 2005). In the process safety field, the need of a global risk assessment of CNT storages and production sites has recently been expressed (notably during the Integrated European Project Nanosafe 2). This need is emphasised by the application of European directive 1999/92/EC (EU, 1999), which defines the minimum requirements to protect the workers from potentially explosive atmospheres within the meaning of directive 89/391/EEC on safety and health of workers at work. Reliable and quantitative information on kinetics data are thus needed, which is the aim of our present work.

Solid state kinetics have extensively been studied by thermal methods (Brown et al., 2000). However, kinetics analysis has been criticised a lot. In fact, kinetics measurements and onset temperatures obtained from reactive materials may not be reliable because of heat and mass transfer effects which corrupt the extraction of fundamental chemical kinetics rate parameters (Park et al., 2004). For instance, it has been shown that the calcium nitrate nanoparticles reactivity can be underestimated by as much as three orders of magnitude by using thermogravimetric analysis because of heat and mass transfer effects (Mahadevan et al., 2002). Despite this fact, solid state kinetics has continued to develop together with improvements in experimental and computational methods. Thus, thermal gravimetric analysis (TGA) and differential scanning calorimetry (DSC) are both used in this study, with the cautions expressed by the previous remarks, in order to provide isothermal and non-isothermal kinetics data which allows the determination of accurate kinetics parameters (i.e. conversion model, Arrhenius factor and activation energy) of bulk CNT combustion.

The results of these thermal analyses will be combined with the data collected by means of self ignition tests in baskets. In addition, some ignition and explosivity data will be presented. Finally, recommendations for safe handling and storage of CNTs will be put forward.

Section snippets

Main characteristics of the carbon nanotubes

CNT samples are raw multiwalled carbon nanotubes (MWCNTs), provided by an industrial partner. Some specifications were unknown and were determined during this work. The particle-size distribution, determined in isopropanol by using a laser diffraction analyzer (Mastersizer 2000S, Malvern Instrument) was characterized by the d10, d50 and d90, with respectively 20, 200 and 560μm. CNTs seems to form large agglomerates, which is confirmed by the transmission electron microscopy (TEM). TEM

Thermogravimetric analysis—isothermal analysis

The oxidation of CNTs materials was investigated by means of isothermal thermogravimetry using a Setsys 12 thermobalance (Setaram) under atmospheric pressure and temperatures ranging from 450 to 600 °C. The mass loss and the temperature were systematically recorded. The studies were performed under isothermal conditions and an oxidant flow of dry air (330 mL min−1). For each run (450, 500, 550, 600 °C), approximately 2 mg were used as it is difficult to place more nanotubes into the crucibles due to

Thermogravimetric analysis—selection of reliable conversion models

In a first stage, there is a weight increase which can be attributed to catalyst oxidation (Fig. 2). In the case of raw SWCNTs containing metal catalyst, this specific pre-oxidation was also reported several times (Chiang et al., 2001; Landi et al., 2005). Then, there is a single step weight loss due to carbon oxidation over a 2500–5000 s period. Finally there is a stabilisation when the maximum conversion rate is reached.

For experimental results obtained at 500, 550 and 600 °C, respectively, the

Conclusions

The combustion kinetics of raw MWCNTs was studied by means of isothermal and non isothermal thermogravimetric experiments. The activation energy was determined from both Kissinger method (without assumption of a model) and from isothermal experimental data. It was shown that the activation energy of MWCNT oxidation was 150 kJ mol−1, which is slightly greater than the SWCNTs’ one (Illekova and Csomorova, 2005) as expected. This study has also shown that the combustion of CNTs seems to be a

Notation

AArrhenius’ preexponential factor
CO2molar concentration of oxygen, mol m−3
CNTscarbon nanotubes
Cpspecific heat capacity, J kg−1 K−1
dxquantiles of the volumetric distribution. Size at which x% of the particles are smaller
DSCdifferential scanning calorimetry
EAactivation energy, J mol−1
f(X)reaction model
FO2consumption rate of oxygen, mol s−1
Henthalpy of reaction, J g−1
ΔHtotal heat released during oxidation, J g−1
kreaction rate constant—k(T) is function of the temperature
m0mass of the sample at the

Acknowledgements

This study was carried out with the financial support of the European Commission through the Sixth Framework program for Research and Technological Development NMP2-CT-2005-515843 contract “NANOSAFE2”.

References (37)

  • Boumahmaza, M., 1991. Etude cinétique de la gazéification d’un charbon du maroc (anthracite de Jerada) par...
  • M.E. Brown

    Introduction to Thermal Analysis

    (2001)
  • R. Brukh et al.

    Kinetics of carbon nanotube oxidation

    Journal of Materials Chemistry

    (2007)
  • CEN, 2006. Determination of the spontaneous ignition behaviour of dust accumulations. European project of standard PrEN...
  • I.W. Chiang et al.

    Purification of single-wall carbon nanotubes (SWNTs) obtained from the gas-phase decomposition of CO

    Journal of Physical Chemistry B

    (2001)
  • R.K. Eckhoff

    Dust Explosions in the Process Industries

    (2003)
  • EU, European directive 1999/92/EC 1999. Directive on minimum requirements for improving the safety and health...
  • J.H. Flynn

    The ‘temperature integral’—Its use and abuse

    Thermochimica Acta

    (2007)
  • Cited by (34)

    • Manganese-based metal-organic-framework derived hydrophilic cathode with carbon nanotubes introduced for long-life and high-performance aqueous zinc-ion battery

      2022, Journal of Alloys and Compounds
      Citation Excerpt :

      Thermal gravimetric analyze (TGA) of Mn-MOF and Mn-MOF@CNT in air atmosphere were showed at Fig. S5 and Fig. S6, which can analyze that the generation temperature of manganese oxide is 230 °C and the stable temperature is 300 °C. It can be determined that the pyrolysis temperature of CNTs in air is 520 °C through the TGA curve of Mn-MOF@CNT materials [47]. Therefore, 400 °C was chosen as the calcination temperature to ensure the formation and stable existence of the product.

    • Influence of carbon black nanoparticles on the front flame velocity of methane/air explosions

      2017, Journal of Loss Prevention in the Process Industries
      Citation Excerpt :

      Nano-titanium is a highly sensitive powder as the dust self-ignition is easily achieved and the explosion severity seems to increase when the particle size decreases (Boilard et al., 2013). Although the explosion overpressure has comparable values for micro and nano-sized powders, similar studies of nano-dust have shown that the explosion sensitivity is considerably higher the smaller the particle size (Vignes et al., 2009). In fact, the similarities on the explosion severity between microparticle and nanoparticle powders could be explained because of the limited homogeneous dispersibility and the high coagulation phenomena on nanoparticle powders (Eckhoff, 2012, 2013; Worsfold et al., 2012).

    View all citing articles on Scopus
    View full text