Expansion cloud chamber measurements of the gas phase self- initiated thermal polymerization of styrene

Abstract

The precision Wilson expansion cloud chamber [I] at the GCCPR has been used to detect polymers formed in the vapor phase by the thermal self-initiation of styrene. This expansion chamber has a large volume that is uniform in temperature and vapor content. Polymerization in the vapor phase involves initiation, propagation by addition of monomers and diffusion and loss to the walls but not chain transfer or termination (as is found in the liquid). The polymers that are formed are detected by expanding the chamber to a supersaturation sufficient to nucleate the polymer (cluster) yet not start homogeneous nucleation. This type of detection allows one to detect, for example, I to 1000 polymers/cm 3 in a vapor of 5 x 1017 monomers/cm 3. Previous work on styrene was done in a diffusion chamber by El-Shall, Bahta, Rabeony and Reiss [2]. Initial measurements were made with pure styrene vapor in an argon carrier gas. It was difficult to prevent the liquid styrene from polymerizing in this case and subsequently oxygen was added to the carrier gas to enhance the action of the inhibitor in the liquid. Figure I shows data for pure styrene vapor at 15 C. This preliminary data may show some influence from polymerization of the liquid. Rabeony and Reiss [3] have modeled the case where polymers form in a gas and diffuse to the walls of a closed cylinder. They predict that the number of drops/cm 3 in the vapor, N, can be calculated from the expression: N(I,D,K,J*) = I • F where I is the initiation rate and F is a factor that can be calculated from a lengthy expression where D is the diffusion constant, K the propagation constant and J the critical cluster size. Figure 2 shows the factor, F, as a function of the critical cluster size for various values of K. Using the theory of Reiss, Rabeony, El-Shall and Bahta [4], one can calcuate the critical cluster size as a function of supersaturation in styrene. When the results of this calcuation are applied to the data in Fig. I, one finds the slope as shown in Fig. 2.