The observed mechanism of the inhibition by boron of the catalytic oxidation of carbon composite differs from that of phosphorus compounds. In addition, to the reduction in the Fermi level of carbon, the protective effect of boron has been proposed to occur in two ways: (1) by reduction of the number of carbon active sites and (2) by blockage of carbon by formation of boron oxide which melts at 450C. The effect of boron doping and of boron oxide deposition was observed by Wu and Radovic to completely suppress the calcium-catalysed reaction. At with phosphorus oxide compounds, boron oxide may disrupt the initial interfacial contact between calcium and carbon resulting in its deactivation. A combination of SEM and XRD data that were presented by Wu and Radovic has demonstrated the effect of blockage of the carbon fibres and matrix by B2O3. According to McKee, the site-blocking effect occurs because the O-O bond distance in the –BO3 group is close to the distance between the zigzag carbon atoms in the grapheme layer. On the other hand, results of X-ray photoelectron spectroscopy presented by Cermignani et al. identified the presence of boron oxycarbides rather than B2O3 following a four-hour oxidation at 600C.
The need for oxygen to diffuse through the encapsulating B2O3 or boron oxycarbides may be another factor for the reduction in the rate of the catalytic oxidation as reported by Wu and Radovic. However, boron-doping and boron oxide deposition had only a minimal inhibiting effect on the potassium-catalysed reaction. It was suggested that the reason for this was the high mobility rate of potassium salts and their ability to readily re-disperse and maintain good interfacial contact with carbon. A sharp decrease in the oxidation rate after a maximum had been reached was pondered to be due to the reaction of boron oxide with potassium catalysts to form a glass-like layer, but the authors provided no data to substantiate this. However, if correct, this approach would seem to be potentially the most effective to prevent the catalytic oxidation of C-C composites. The effect of this prevention approach would be to convert the catalyst to another stable compound which would inhibit the catalysis reaction.
The most likely mechanism of the catalytic oxidation of C-C composites by sodium and potassium salts is probably the one involving the redox reaction between the monoxides and the peroxides and carbon as proposed by McKee and Chartterji; thus scavenging the catalyst by prevention of the reaction of alkali metal oxides to their respective peroxides would seem to be the most sensible approach to eliminate catalysis. Support for this idea is provided by the results reported by Tricot et al. who brushed mono-aluminium phosphate solution onto the surface of C-C samples. Following drying and a heat-treatment at 650C in an inert atmosphere, some samples were brushed with various acetate solutions before undergoing isothermal oxidation at 650C. After 3 h of oxidation, the unprotected samples exhibited significant levels of oxidation; for example, the sample that had been exposed to potassium acetate lost 62% carbon, while the samples with sodium and calcium acetates lost 60% and 30% respectively. However, in the presence of these acetates, the samples that were protected with the mono-aluminium phosphate lost less than 0.5% carbon over the same period. According to the authors, the presence of the Al(PO3)3 coating prevented the catalysts from establishing good contact with the carbon. In addition, XRD analysis of the protected sample that had been contaminated with sodium acetate, showed evidence of small amounts of NaAlP2O7. This demonstrated the ability of the Al(PO3)3 to convert the catalytic species into an inactive compound. A similar approach was attempted by Devecerski et al. who added 4% by weight of boron in the glassy matrix of the composite.
They showed evidence of oxidation of the boron which then formed B2O3; subsequent tests revealed inhibition against catalytic oxidation with sodium acetate. It was proposed that the inhibition to catalysis was due to reaction between B2O3 and Na2O to form Na2B4O7 thus preventing the formation of sodium peroxide. They tried to verify the formation of Na2B4O7 by XRD and FTIR analyses, but the results were inconclusive perhaps due to the low amounts of the borate in the samples. The authors then soaked a composite sample in a saturated solution of Na2B4O7.10H2O for 30 min and subsequent oxidation runs showed results that were similar to those that had exhibited catalytic inhibition. This result provided some basis to their assumption that Na2B4O7 could have formed and was probably the cause of the catalytic inhibition, but further work is required to confirm the mechanism.