Devulcanisation of truck tyre tread vulcanisates in supercritical carbon dioxide using diphenyl disulphide and 2,2- dithiobis(benzothiazole)

A lot of work has been done in the recycling industry in an effort to increase the amount of reclaimed rubber used in new tyre formulations. The major drawback has been inferior physical and mechanical properties of reclaimed/virgin rubber blends in comparison to the virgin rubber material. Deterioration in these properties has been identified to be a result of chain degradation during reclamation processes as well as presence of crosslinks in the final reclaim product. Devulcanisation techniques have gained precedence due to the relatively improved properties of devulcanised/virgin rubber blends. The concept of devulcanisation is to reverse vulcanisation, resulting in total or partial cleavage of crosslinks. In this way, chain degradation is minimised while crosslink scission is maximised, thereby resulting in good quality devulcanised rubber. However, due to the persistence of chain degradation and crosslinks during devulcanisation processes, a very limited number of reports have claimed success in achieving this goal. Therefore there is still the need to develop a devulcanisation method that ensures improved quality and productivity of devulcanised rubber. Typical truck tyre tread vulcanisates were used for optimisation of time, temperature, heating rate, pressure and amount of devulcanising agent while monitoring percentage devulcanisation in supercritical carbon dioxide medium. Optimisation of the devulcanisation conditions was done by employing a twolevel central composite design in the isothermal and non-isothermal heating stages. This was followed by a single factor analysis of devulcanisation conditions in the non-isothermal stage. The effect of the presence of carbon black was investigated by comparing the percentage devulcanisation of carbon black filled and unfilled samples. The results show that supercritical carbon dioxide is an effective medium of devulcanisation using diphenyl disulphide (DD) and 2,2-dithiobis(benzothiazole) (MBTS). The relatively higher degree of devulcanisation observed during the non-isothermal stage compared to the isothermal stage, led to a shift of focus to non-isothermal devulcanisation. Temperature and time were found to have a significant antagonistic effect on the percentage devulcanisation, while changes in pressure above critical point and mass of devulcanising agent showed no effect on percentage devulcanisation. The heating rate was determined by the set-point, of which 180 ℃ set-point temperature resulted in desirable degree of devulcanisation for both DD and MBTS. 76.18 ± 5.50 % devulcanisation in 5 minutes at 102 ℃ was observed for DD whilst 70.92 ± 4.10 % devulcanisation in 4 minutes at 97 ℃ was observed for MBTS. Changes in pressure above critical point and mass of devulcanising agent used in devulcanisation showed no significant effect in the percentage devulcanisation and so they were kept constant at 80 bars and 1.00 % v (of weight of rubber sample) devulcanisation agent, respectively. The presence of carbon black was found to have an effect on the degree of devulcanisation; 87.95 % and 81.33 % devulcanisation was observed for unfilled samples devulcanised using DD and MBTS respectively. Thermogravimetric analysis of the natural rubber/styrene butadiene rubber (NR/SBR respectively) relative composition of devulcanisates indicated uneven devulcanisation when using DD, whereas MBTS did not show any form of preference. DD showed preference for NR devulcanisation over SBR. Further analysis of the sol and gel fractions were performed using; Differential Scanning Calorimetry, Fourier Transform Infrared Spectroscopy, Gel Permeation Chromatography and Gas Chromatography coupled with Mass Spectroscopy. Application of the optimised conditions to devulcanise ground tyre rubber (GTR) resulted in relatively lower degrees of devulcanisation for both DD and MBTS; 41.22 ± 4.22 and 22.41 ± 1.97 respectively. The differences in the degree of devulcanisation of the laboratory prepared vulcanisates and the GTR was determined to be due to sample differences; i.e. sample constituents, particle dimensions and crosslink network (crosslink distribution in particular)