“…The beamline allows a micro-beam size of about 2*2 μm 2 , which makes it possible to select glass surfaces that are devoid of excessive alteration and are not covered by grisaille. However, as the use of a microbeam causes photoreduction or photooxidation effects (e.g., Chalmin et al, 2009;Ferrand et al, 2015;Yamashita et al, 2004), the beam was defocused to cover a larger area (15x15 μm 2 ), reducing the exposure per unit area of the sample. The stability of the measurements suggests that possible photoinduced effects have a limited impact on the experimental redox values.…”