Signal drift of oxygen optical sensors. Part I: Rationalization of the drift nature and its experimental check with a light intensity detection based sensor

A way to quantify the signal drift causes of an oxygen optical sensor is reported. Theoretical formalization were experimentally confirmed by using a polysulfone-based thin layer membrane embedding platinum meso-tetra-(pentafluorophenyl)-porphyrine working in light emission detection mode. Photochemical, thermal and oxidative degradation of both luminophore and polymeric matrix were rationalized and tested. Experimentally determinable light intensity drift, D I , and Stern–Volmer constant ( K ′ S V ) drift, D K , were related to matrix and luminophore modifications. The mathematical modeling of the drift evidenced that D I , is constant by increasing the %O2 only in the absence of D K . It has been possible to quantify all the contributions independently through a very good match between theory and experiments. The sensor drift analysis allowed understanding whether luminophore and/or polymer degradations were operative. In the studied caseoxidative degradation was absent; thermal degradation of the sole luminophore, was independent of %O2 and caused a relative light intensity drift of − 0.028 ( 0.002 )  day − 1 ; photochemical degradation was present both on luminophore and polymeric matrix. When D K ≠ 0 , either light intensity and phase-shift and the life-time based detection modes lead to incorrect oxygen concentration values. The results reported in this paper suggest the design of a new, simple, cheap and robust light intensity-based sensor working with high accuracy and precision. This will be done in the Part II of the workby setting up an algorithm able to correct all the drift effects operating during the life-time of the sensor.