A precision freezing point apparatus with platinum resistance thern~on~eter was used to investigate the system hydrogen peroxide -water over the whole concentration range. The freezing point of the purest sarnple of hydrogen peroxide obtained by repeated fractional crystallizations of a large quantity of 99.6% pure solrrtion was found to be -0.461°C.; that of the dihydrate was -52.1OoC. T h e two eutectlcs occur a t concentrations of 45.2% and d1.2yO Hz02 and a t temperatures of -52.4" and -5G.5"C. respectively. Contrary to what has been reported previously, water and hydrogen peroxide do not form solid solutions together. This was proved conclusively by applying the technique of radioactive tracers to the 'wet residr~e' method of Schreinemakers.The purpose of this investigation was threefold. First, to redetermine the freezing point of pure hydrogen peroxide, with greater accuracy than heretofore (3). Second, to scttle the question of solid solutions in the systenl hydrogen peroxide -water (6). Third, to use freezing point measurements for evaluating the degree of purity of very concent~atcd hydrogen peroxide. Indeed, the usual chemical nlethods of analysis are of little use for that purpose since the remaining traces of water cannot be determined directly (4). Other physical metl~ocls investigated so far are either not sensitive enough or subject to errors due to spontaneous decomposition. These three objectives arc closely interconnected; for instance, the existence of solid solutions would render determinations of freezing point rather problematic. At the same time it could rule out the possibility of using this method for accurate analysis.
Experimental Methods and Results
Freezing PointsAn apparatus similar to that originated by Mair (12) and perfected subsequently by Rossini and his collaborators (7,13,20) a t the National Bureau of Standards was constructed for this purpose. T h e double-walled, silvered freezing tube had an internal diameter of 3 cm. and the rate of heat transfer was controlled by adjusting the pressure inside the annular space. The cold head was provided by a large Dewar flask filled with dry ice and acetone. The problem of stirring the sample presented some experimental difficulties sincc most materials catalyze the clecornposition of hydrogen peroxide. Stirrers made of glass rods were too fragile. Aluminum or highly polished stainless steel has been recommended a s suitable for this purpose (18); the latter was selected because of its greater strength. T h e stirrer was in the form of a double helix fashioned iliunz~script receiued October 4, 1950.