A Fairly Common Error in Processing the Results of Knudsen Effusion Mass Spectrometry: Sublimation Enthalpy of Vanadium as an Example

Abstract

It has been shown in this paper, that a fairly common approach of Knudsen effusion mass spectrometry based on the so called “self-calibration” for the instrument sensitivity (combination of single measurement for absolute vapor pressure by Knudsen method with the mass spectrometry for its temperature dependence) leads to a very large uncertainty for the enthalpy of sublimation for the substance under investigation. The uncertainty may not be less than the uncertainty of initial single measurement for vapor pressure, and so, the array of data obtained by its combination with the mass spectral data is not the array of independent measurements. The difference between the uncertainties for two approaches (considered and criticized) may be near one order of magnitude for vapor pressure. There are presented the results of 13 measurements of vapor pressure for vanadium in the paper (1). Formal statistical treating of these results leads to sublimation enthalpy with the uncertainty less than ±1 kJ·mol-1, although the uncertainty of the calibration experiments (2 experiments) amounts near ±13 kJ·mol-1.