Note on the thermal degradation of polytetrafluoroethylene as a first-order reaction

Abstract

Addit ion a l experim e nts 011 the rates of t hermal degrad ation of polytetraflu oroethylene in a vacuum confirm a n earlier conclu sion t hat a first-order rate law is involved in the degradation reactio n. In ~, sLudy made by Madorsky, Hart, Straus, and Sedlak Io n tbe rates and activation energy of thermal degradution of polytetrafiuoroethylene in a vacuum, two method were employed: a gravimetric method, using a very sensitive tungsten spring balance in a vacuum system to measure the rate of loss of weight of the degrading sample, and a pressure method, using a multiplying manometer to m easure the pressure of the C2F4 formed in the reaction. The material that was used was in the form of a tape 0.07 mill thiclc. Weight of tbe sample was about 7 mg in the gravimetric experiments and 5 to 306 mg in the pressure experiments. Th e rates obtained by the gravimetric method are l'Cproduced in figure 1, plotted as a function of percentage loss of weight of the sample for 480, 490, 500, and 510°C. The initial rates were obtained by extrapolating th e rate curves to t he ordinate. The rate curves beyond the initial 5 to 18 percent loss of weight of the sample are straight lines, and wh en extended to the right they approach ncar the zero rate at 100 percent volatilization. The rates obtained by the pressure method were studied at 10 diff erent temperatures ranging from 423.5 to 513.0 dc. Logarithms of the initial rates obtained by both methods are shown in figure 2 plotted against the inverse of absolute temperature. 2 From the Arrhenius equation the slope of the resulting straight line indicates an activation energy of 80 .5 kcal/mole. From the appearan ce of the curves in figure 1 it seemed logical to conclude that the reaction involved in the thermal degradation of polytetrafluoroethylene in a vacuum follows a fu'st-order law. At a later date ,Vall and Michaelson 3 studied the rate of thermal degradation of polytetrafluoroethy-lene at 460 DC in a stream of nitrogen. They used a gmvimetric method by heating I-g samples of a powdered material and weighing the r esidues at IS. L. ]VJ:ac!orsky, V. E. H a rt, S. Straus, a nd V. A. Sedlak, J. Research N BS 51, 327 (l953) RP2461. 2 In th e earlier pa pe r (see footnote \) , the abscissa of a simil ar graph (Ag. 5) was incorrectly labelled log (I/T)XLO' , whereas it should have bee n (l/1')X IO' .