How Much Data Are Enough? Toward Statistically Robust Adhesion Experiments by Atomic Force Microscopy

Direct force measurements by atomic force microscopy (AFM) have become an indispensable analytical tool in the last decades. Force measurements have been widely used for adhesion measurements, often in combination with the colloidal probe technique. For the latter technique, a colloidal particle is attached to the end of an AFM cantilever, proving great flexibility in terms of colloid/surface interaction to be studied. Interestingly, the question of how much data is necessary to obtain statistically reliable results has been addressed in this context only sparsely. By contrast, the value and necessity of determining and reporting distributions of adhesion forces has been widely accepted. Here, simple statistical methods of the experimental design have been applied to address this question. It has been demonstrated that it is possible to determine an optimal number of force curves even during data acquisition. This approach would be essential to prevent oversampling of data. Moreover, it allows to address questions like heterogeneity of the sample in a more reliable or less time-consuming way. In AFM measurements with colloidal probes, most statistical variation results from the surface roughness of the probe particle. In this case, the use of different colloidal particles is important, which can be achieved by techniques such as fluidic force microscopy (FluidFM). The latter enables to combine a real-time determination of the required data size and high-throughput techniques of unattended measurements, which will open new fields of analysis.