An objective, principal-component-analysis (PCA) based, method which improves the quartz-crystal-microbalance (QCM) sensing performance

The quartz-crystal-microbalance (QCM) is a widespread analytical tool used in various fields such as chemistry, life-sciences, and soft-matter research. The QCM can detect mass and viscoelastic changes over its sensors surface. The sensor signal stems from the variation of the quartz resonant acoustic frequencies, called harmonics. To date, the operator selects and analyses one or few of these harmonics, therefore introducing an operator bias. This standard method also causes a massive loss of information owing to the discarding of the unanalyzed signals. Here, we develop a novel method for QCM data-analysis based on principal-component-analysis (PCA) that takes advantage from the whole harmonic dataset and avoids any operator bias. Such method is applied to two different molecular adsorption experiments. One is a standard benchmark experiment based on the streptavidin-biotin binding, widely used in the biosensing field. The other experiment is a case-study where an antigen-antibody binding is tested in presence of biological noise. The antigen of choice is the glial-fibrillary-acidic-protein (GFAP), recently discovered as a brain-damage biomarker. In both cases, we observed an improved sensitivity and limit-of-detection compared to the standard method. Our analysis further reveals a decreased, or nearly unvaried, noise-level. The proposed PCA-based approach standardizes the QCM data-analysis procedure and improves its performance. This method, besides being a useful tool for the QCM community, can be customized to other sensing technologies whose response is made of several variables.