Neural Network Temperature Predictions based on the Optical Properties of Quantum Dots

Cadmium telluride quantum dots, as well as other types of quantum dots, have potential applications as localized temperature sensors in microfluidic devices due to the temperature-dependent qualities of their photoluminescence (PL) lifetime lengths. To take advantage of these properties, PL spectral data and time-resolved PL data of various samples were collected at a range of temperatures, and a machine learning algorithm was trained to output a temperature prediction based on the input data. Two different cadmium telluride quantum dot samples have been tested already and show promising results, with a mean absolute error (MAE) as low as 0.1 K for those emitting PL at 620 nanometers. More samples, such as perovskite quantum dots will be investigated in the future.