The purification of commercially available 1-buty1-3-methyl-imidazolium hexafluorophosphate (BMIPF(6)) is described, which is still required in addition to the usual vacuum drying for detailed double-layer studies. By using a molecular sieve of 0.3 nm pore size, the level of impurities originating either from an omnipresent humidity or from IL synthesis, is reduced to a point, where cyclic voltammograms in essence represent only double-layer behaviour of a Au(1 1 1)/BMIPF(6) interface. It is demonstrated that pore sizes larger than 0.3 nm are counterproductive: the situation is worse than without any sieve. (C) 2010 Elsevier B.V. All rights reserved.