Essential Amino Acids - Detection on a GaAs Chip and via Machine Learning

We demonsrate that it is possible to detect all essential amino acids in a water solution on a GaAs chip, at ambient conditions. The chip is made of AlGaAs/GaAs heterostructure in a simple microfluidfic configuration, and the detection method is based on differential conductance. We identify between 3 and 7 characteristic peaks for each amino acid that can be used for idenitification putposes in machine learning algorithms. Our experiments show the changes in Debye length as we alter the pH factor of the solution, pointing towards optimum detectability for various amino acids. In parallel, we find correlations between the X-ray absorptivity of amino acids, their detectability on our chips and some of their physico-chemical properties.