Binary mixtures dataset acquired with a temperature-modulated redundant MOX gas sensor array

This dataset contains measurements of binary mixtures of volatile compounds (ethanol, acetone, and butanone) acquired with a temperature-modulated MOX gas sensor array. The sensing platform combines: multiple sensor types (8), sensor replication (12), temperature modulation, and multiple load resistances (16), resulting in a highly redundant representation of the chemical space. The dataset is designed to capture continuous transitions between pairs of analytes at a fixed total concentration (120 ppm), enabling the study of compositional effects in chemical sensing. Three binary systems are included: ethanol–acetone acetone–butanone butanone–ethanol For each system, the mixture composition varies in seven steps from one pure analyte to the other. The experimental sequence is repeated 10 times with randomized exposure order, and baseline measurements using synthetic dry air are acquired before each exposure. The combination of sensor diversity, redundancy, temperature modulation, and controlled mixture design results in a high-dimensional dataset suitable for: mixture deconvolution regression and calibration models transfer learning and domain adaptation feature extraction in multivariate chemical sensing robustness analysis under complex stimuli Data are provided in MATLAB .mat format, with accompanying scripts for Python and R. This dataset complements the dataset 'Pure substances dataset acquired with a temperature-modulated redundant MOX gas sensor array': https://doi.org/10.5281/zenodo.19557093