Calculated melting point of the catalyst particles observed at the tip of carbon fibres are not far from their temperature of growth. In order to determine whether this melting favours or poisons lengthening of precursor filaments, it is necessary to discuss the experimental conditions at which lengthening occurs. With that purpose, a previously described procedure for making carbon fibres from a CH4-H-2 mixture has been modified by heating first the catalyst precursor under hydrogen flow up to 1028-degrees-C. H-2 was then replaced by CH4-H-2 MiXtUre (30 moles CH4 percent) and, after 10 min (or longer time) of this flow at the constant temperature of 1028-degrees-C, carbon filaments had grown up to some micrometers. Fibres (length greater-than-or-equal-to 1 mm) only appear when the procedure is pursued by heating up these filaments to 1150-degrees-C without any carbonaceous gas and then reintroducing the 30% CH4-H-2 mixture at 1150-degrees-C. N2 seeMS to be the most convenient gas to be flowed during heating from 1028 to 1150-degrees-C. Fibre yield is less important when N2 is replaced by H-2 or helium. It is less important, too, when the temperature plateau at 1028-degrees-C under the 30% CH4 mixture is continued under a 60% CH4 mixture (Fig. 5). Thus 60% CH4 mixture at 1028-degrees-C prevents most fibres to lengthen. The main part of lengthening occurs therefore, when no CH4-rich mixture is introduced, at higher temperature than 1028-degrees-C. As heating from 1028 to 1150-degrees-C is performed without any carbonaceous gas, this lengthening stage can only occur at 1150-degrees-C. Shortening of the 1150-degrees-C stage to the minimum time allowed by our apparatus (30 seconds) does not prevent fibre growth. Lengthening seems therefore to occur during the first seconds when 30% CH4 mixture is reintroduced in the apparatus at 1150-degrees-C. This result depends upon the history of the catalyst. Catalyst particle diameters measured at 1028-degrees-C before methane introduction are much larger than at 950-degrees-C. When the CH4-H-2 Mixture comes into contact with our catalyst at a temperature close to 1030-degrees-C, carbon germination fixes particle diameters to sizes which are unfavourable to an important lengthening of carbon filaments at that temperature. If the carbonaceous atmosphere is replaced by an inert ps after this germination and during heating of the catalyst up to 1150-degrees-C, then main part of filament lengthening occurs when the CH4-H-2 mixture is reintroduced. The important fibre yield obtained by this process does not agree with the idea of a poisoning of catalyst particles by their melting. On the contrary, lengthening occurs at the temperature at which carbonaceous ps comes into contact with catalyst particles in the liquid state. The harmful role, observed by other authors, of heating on fibre growth may be issued from the sintering of catalytic particles. Heating only favours lengthening when the sizes of particles are fixed by carbon germination.
