The Freezing Points of High Purity Metals as Precision Temperature Standards:v. Thermal Analyses on 10 Samples of Tin With Purities Greater Than 99.99+%

Abstract

Extensive thermal analyses have been made on 10 samples (suppliers' analyzed impurity contents <0.2 to <I00 p.p.m.) of high purity tin, including zone-refined metal; liquidus points have been intercompared with a precision of about 0.0002" C and alloy melting ranges have been examined following different types of freezing with and without overnight anneals near the solidus temperature. Samples of nominal 99.9999% purity tin were found t o have such narrow alloy melting ranges that any ambiguity, arising from unlcnown impurity conceutrations, in specifying the liquidus point of pure tin is well inside 0.001" C ; a value of 231.913" C (Int. 1918) was found for the standard liquidus point of puretin. An ~C C O L I I I~ is give11 of the supercooling that was observecl on the bull; samples and of anon~alous structures that were found 011 melting curves. f i n appendix gives the results of long-term intercomparisons of the temperatures realized i r~ four water triple point cells.'Manuscript

Mechanism to Diminish the Supercooling of the Tin Freezing Point by using Graphite Powder

“…Fine graphite powder was observed adhering to the wall surface of the graphite crucible to form a porous substrate. The supercooling in the cells treated with this technique is observed to decrease to 2-5 K depending on the thermal history of the molten tin, as indicated in the literature [4]; the lower the temperature of the molten tin, the smaller the supercooling. The small supercooling makes possible a natural cooling operation, whereby a tin fixed-point cell is cooled down naturally within a furnace to realize the tin freezing point.…”

“…The final report of the CCT-K3 comparison [7] indicated that the freezing point of the open cell is 1.3 mK lower than the CCT-K3 reference value. The literature [4,8,9] indicated that the impurities with an equilibrium distribution coefficient k < 1 depress the freezing point. This open cell was fabricated in the 1970s.…”

“…The depth of the supercool appears to depend on many variables. McLaren and Murdock [4] surmised that the purity of the tin samples, the addition of a graphite powder topping, and the thermal history of the previous tin melt influence the extent of the supercool. Topping the tin melts with graphite powder was observed by them to decrease significantly the supercooling of extremely pure tin ingots, where graphite powder was applied as an oxide inhibitor.…”

“…High-purity tin should be kept in an inert atmosphere and not be topped with graphite powder: when graphite powder becomes distributed throughout a sample, inadequate supercooling (<4 • C) may result." The literature [2,4] emphasized the importance of adequate supercooling for the realization of freezing curves of tin but did not relate the degree of supercooling to the equilibrium temperature of the tin phase change.…”

Standard Reference Materials

“…If the freeze were initiated by lowering the furnace temperature with the tin-point cell in the furnace, the furnace block temperature could not be returned to the freezing-point temperature fast enough to avoid excessive freezing of the tin. Hence, the tin freeze is initiated by employing the "outside nucleated" freeze technique described earlier by McLaren and Murdock [26].…”