Comment on "Freezing point mixtures of H216O with H217O and those of aqueous CD3CH2OH and CH313CH2OH solutions"

In a recent article, Kiyosawa [J. Solution Chem. 33, 323 (2004)] reports that the freezing points of isotopic mixtures of ordinary water and O-17 enriched water show an unexpectedly large linear dependence on the concentration of (H2O)-O-17. Surprisingly, the constant of proportionality to the (H2O)-O-17 concentration is nearly five times larger than that of (H2O)-O-18 found in earlier studies by Kiyosawa [J. Solution Chem. 20, 583 (1991)]. We show that the (H2O)-O-17 result is not consistent with other data or models. For example, a recent determination of the triple point temperature dependence on isotopic composition in naturally and artificially depleted waters [White et al. in Temperature, Its Measurement and Control in Science and Industry, Vol. 7, D. C. Ripple, Ed., AIP CP 684, 221-226 (2003)] is consistent with the (H2O)-O-18 and D2O results from Kiyosawa (1991) [White and Tew in Report of the 22nd Meeting of the Consultative Committee for Thermometry, Document CCT/03-21, BIPM, Severes, France, 2003] but is inconsistent with the (H2O)-O-17 results from Kiyosawa (2004). Additionally, the results from Kiyosawa. (1991) are close to what would be found in ideal solutions for those isotopic forms, whereas the (H2O)-O-17 proportionality from Kiyosawa (2004) is about 10 times larger than similarly predicted. One possible explanation is that the original O-17 enriched water sample contained a small amount of D2O, and the sample, if available, should be subject to isotopic analysis to help resolve these inconsistencies.