Plastic recycling technologies are being actively developed and implemented to cope with increasing volume of
plastic. Such technologies require new analytical tools able to control the quality of the recycled polymers to be
further integrated in production processes. Here, we propose a rapid and selective quality assessment method for
polymer materials made of high-density polyethylene using electronic nose with aluminum doped zinc oxide
sensing material in combination with the RandomForestClassifier machine learning tool. We test total content of
volatile organic compounds both odor-active responsible for the smell and odorless of primary and secondary
plastics, and evaluate corresponding organic vapors emitted by the plastics by headspace gas chromatography
and mass-spectrometry at optimized conditions like sample temperature, sensor signal recovery time. The
electronic nose demonstrated the good correlation of vector signal with the emitted volatile compounds with an
accuracy more than 98.5% when discriminating between primary and secondary plastics. Addition of zeolites to
the recycled plastic is shown to decrease the appearance of off-odors.
