Study of quantitative influence of sample defects on measurements of resistivity of thin films using van der Pauw method

Abstract The paper is devoted to the methodology of measurement of resistivity of thin films using van der Pauw (VDP) technique – in particular to the consequences of typical “defects” that can appear in the samples. To a certain extent, it can be actually deduced from the measurement whether the defect is fatal, and the sample has to be excluded from the experimental results (incorrigible defect), or whether the error can be subsequently corrected numerically (corrigible defect). This is important especially for ultrathin films close to the limit of their connectivity, where the consequences of these defects can be easily mistaken for actual anomalous behaviour. Therefore, we decided to simulate the consequences of such defects experimentally. Using standard photolithographic techniques and subsequent selective etching, typical sample defects were simulated on geometrically symmetric samples of thin Cu foil: these defects ranged from non-symmetric contacting, isolated and non-isolated cracks and holes, to the influence of their size and position within the sample. The impact of these defects on the sequence of values measured in one measurement cycle of the VDP method is demonstrated, and a comparison with characteristic values of perfect samples is carried out. Then, we summarise the manifestations and consequences of the defects and demonstrate practical possibilities of their numerical correction. The goal of this study is to show experimentally the types of defects that can lead to electrical asymmetry of the originally symmetric samples, which of these are theoretically amenable to correction, and which cannot be corrected in this way. It was experimentally confirmed that values measured on “simply connected” samples are generally corrigible but in other cases (interior cracks and holes) are incorrigible.