We have measured the density of liquid sulfur as a function of temperature near the polymerization temperature. We measure between 424 and 445 K, with a precision in density of 2×10−5 and an accuracy of 5×10−4. We see upward shifts of the density within 3 K of the polymerization temperature of about 8×10−4 with each cycle in temperature; these shifts could be due to the reaction of the sulfur with the quartz cell, or to the persistence of metastable polymeric forms of sulfur. Thus we consider the data from the first heating run to be the best, since the sample was then the purest. Our results differ from those in the literature in that, while we see a change in slope at the transition, we do not see a sharp minimum or a singularity. We compare our data to two models of the sulfur polymerization as a second order phase transition: the mean field model [A. V. Tobolsky and A. Eisenberg, J. Am. Chem. Soc. 81, 780 (1959)] and the n→0 magnet model [S. J. Kennedy and J. C. Wheeler, J. Chem. Phys. 78, 1523 (1983)]. Both models assume that the sulfur forms linear polymers which are in an ideal solution in monomeric sulfur, and that the thermal expansions of monomeric and polymeric sulfur are linear within the temperature range under consideration. For the most physically reasonable choice of thermodynamic parameters for the models, the mean field model describes the data better than does the n→0 model. The n→0 model is superior if the change of specific volume on polymerization is adjusted. One possible explanation is that sulfur belongs in the n=1 universality class, not in the n→0 class.

Zheng, K. M., & Greer, S. C. (1992). The density of liquid sulfur near the polymerization temperature. The Journal of Chemical Physics, 96(3), 2175–2182.