40TH ANNIVERSARY material science involved in the processes governing this behaviour and the use of this knowledge to produce fibres CH with greatly enhanced properties have been the preoccu pation of a number of laboratories across the world for the last quarter of a century. The result has been the develop ment of three generations of fibres, the latest of which has produced fibres with properties approaching the limits of what is physically possible with silicon carbide C CH3 Figure I Repeat unit of polycarbosilane(PCS). The numbers allow easy 2. First generation fine Sic fibres The interest in silicon carbide as a reinforcement has prompted the development of several types of fibres in luding fibres with diameters usually greater than 100 um made by CVD onto a core filament and also monocrys- talline short filaments known as whiskers with diameters of the order of l um [2, 3]. The Cvd fibre is finding inter est as a reinforcement for titanium but its large diameter handling processes, which are commonly used with finer exploited in the production of SiC fibres. The numbers refer to the positions fibres used in the majority of composite materials. The Sic whiskers also present serious handling difficulties to- gether with worries about health related problems The announcement of the production of Sic based fi- by the cong diameters in the range of 10 to 20 um,made of polycarbosilane precursors, excited considerable interest amongst those people looking for reinforcements capable of operating in an oxidising atmo- sphere at over 1000oC, above the limits of nickel based alloys[4, 5]. Yajima and his colleagues explored a number of routes to produce polycarbosilane which could be used as a precursor for a ceramic fibre [5, 6]. The difficulty lay in the production of a form of polycarbosilane which Figure 3 The first generation of fine SiC fibres were made by cross-linking could be spun and converted into ceramic filaments. The the precursor PCS with oxygen. decomposition of polydimethylsilane(PDs) which was heated in an autoclave at 470oC for 14 h, was eventu- cross-linked. The PCs precursor was made infusible, in ally chosen as the route for the production of PCS as it the first generation of fibres, by crosslinking in air, in the gave a precursor which, although difficult to spin, could temperature range from 145 to 200oC, which introduced be spun from the melt and converted into a ceramic fibre. oxygen into the polymer, as shown in Fig 3. The con- The repeat element in the chemical structure of polycar- version of the cross-linked precursor fibre into a ceramic bosilane is given in Fig. 1. A steric view of this molecule fibre was explored again by Yajima et al. [7-9] and would show that the cycle of carbon and silicon atoms, others [10]. The cross-linked PCS fibres were insoluble with some bonds removed, is arranged in the form of a in all solvents. Heating the cross-linked precursor up to chair configuration, as illustrated in Fig. 2. This reflects 550C induced the evaporation of low molecular weight the arrangement seen in B-SiC. The numbering of the ar- components in the carbosilane which led to a consider- rangement of the atoms is so that a comparison can be able weight loss but resulted in an increase in molecular made to those in Fig. 1. The groups, shown in Fig. l, weight. Above this temperature and up to around 800oC, numbered 2, 3, 5 and 6 are in the same plane whilst 1 hydrogen and methane were lost from the side groups and 4 are out of this plane "Me"represents the methyl in the PCS leaving behind free carbon and cross-linking group(CH3). Unlike most polymers which are spun into was enhanced. Further heating to 1200oC showed that fibres, it was found that very high molecular weight was gas evolution was almost complete at 1000oC At 1050C not necessarily best and polymers with molecular weights hydrogen was again given off and the XRD patterns be- of around 1500 were eventually used for commercial pro- came sharper indicating greater regularity in the struc- duction. The precursor filaments were then spun from the ture. At and above 1300C the free carbon created by the melt in a nitrogen atmosphere at around 300C and then destruction of the methyl groups reacted with the Si-o 82440TH ANNIVERSARY material science involved in the processes governing this behaviour and the use of this knowledge to produce fibres with greatly enhanced properties have been the preoccu￾pation of a number of laboratories across the world for the last quarter of a century. The result has been the develop￾ment of three generations of fibres, the latest of which has produced fibres with properties approaching the limits of what is physically possible with silicon carbide. 2. First generation fine SiC fibres The interest in silicon carbide as a reinforcement has prompted the development of several types of fibres in￾cluding fibres with diameters usually greater than 100 µm made by CVD onto a core filament and also monocrys￾talline short filaments, known as whiskers, with diameters of the order of 1 µm [2, 3]. The CVD fibre is finding inter￾est as a reinforcement for titanium but its large diameter makes it unsuitable for weaving and other types of fibre handling processes, which are commonly used with finer fibres used in the majority of composite materials. The SiC whiskers also present serious handling difficulties to￾gether with worries about health related problems. The announcement of the production of SiC based fi- bres having diameters in the range of 10 to 20 µm, made by the conversion of polycarbosilane precursors, excited considerable interest amongst those people looking for reinforcements capable of operating in an oxidising atmo￾sphere at over 1000◦C, above the limits of nickel based alloys [4, 5]. Yajima and his colleagues explored a number of routes to produce polycarbosilane which could be used as a precursor for a ceramic fibre [5, 6]. The difficulty lay in the production of a form of polycarbosilane which could be spun and converted into ceramic filaments. The decomposition of polydimethylsilane (PDS) which was heated in an autoclave at 470◦C for 14 h, was eventu￾ally chosen as the route for the production of PCS as it gave a precursor which, although difficult to spin, could be spun from the melt and converted into a ceramic fibre. The repeat element in the chemical structure of polycar￾bosilane is given in Fig. 1. A steric view of this molecule would show that the cycle of carbon and silicon atoms, with some bonds removed, is arranged in the form of a chair configuration, as illustrated in Fig. 2. This reflects the arrangement seen in β-SiC. The numbering of the ar￾rangement of the atoms is so that a comparison can be made to those in Fig. 1. The groups, shown in Fig. 1, numbered 2, 3, 5 and 6 are in the same plane whilst 1 and 4 are out of this plane. “Me” represents the methyl group (CH3). Unlike most polymers which are spun into fibres, it was found that very high molecular weight was not necessarily best and polymers with molecular weights of around 1500 were eventually used for commercial pro￾duction. The precursor filaments were then spun from the melt in a nitrogen atmosphere at around 300◦C and then Figure 1 Repeat unit of polycarbosilane (PCS). The numbers allow easy reference to Fig. 2. CH Si CH2 SiMe 1 2 3 6 H 4 5 CH2 SiMe 1 Figure 2 The steric conformation of the repeat unit of polycarbosilane exploited in the production of SiC fibres. The numbers refer to the positions shown in Fig. 1. Figure 3 The first generation of fine SiC fibres were made by cross-linking the precursor PCS with oxygen. cross-linked. The PCS precursor was made infusible, in the first generation of fibres, by crosslinking in air, in the temperature range from 145 to 200◦C, which introduced oxygen into the polymer, as shown in Fig. 3. The con￾version of the cross-linked precursor fibre into a ceramic fibre was explored again by Yajima et al. [7–9] and others [10]. The cross-linked PCS fibres were insoluble in all solvents. Heating the cross-linked precursor up to 550◦C induced the evaporation of low molecular weight components in the carbosilanes which led to a consider￾able weight loss but resulted in an increase in molecular weight. Above this temperature and up to around 800◦C, hydrogen and methane were lost from the side groups in the PCS leaving behind free carbon and cross-linking was enhanced. Further heating to 1200◦C showed that gas evolution was almost complete at 1000◦C. At 1050◦C hydrogen was again given off and the XRD patterns be￾came sharper indicating greater regularity in the struc￾ture. At and above 1300◦C the free carbon created by the destruction of the methyl groups reacted with the Si–O 824