All field emission experiments are noisy, … are any meaningful?

Abstract

Representing field emission data on a Fowler–Nordheim plot is both very common and strongly not recommended. It leads to a spurious estimation of the emitter parameters despite a very good data fit. There is a lack of a reliable method of analysis and a proper estimation of the uncertainty in the extracted parameters. In this article, we show that the uncertainty in the estimation of the field enhancement factor or the emission area can be as high as ±50% even for a tungsten single emitter in good ultrahigh vacuum conditions analyzed by the Murphy–Good model. Moreover, the choice of the exact Murphy–Good method can have a noticeable impact. We found that advanced analysis methods, based on the measurement of the differential conductance of the emitter, are so demanding in terms of emitter stability that up to now its requirements are probably out of reach in any field emission laboratory.