Measurement of NO2 at low concentrations (tens of ppts) is non-trivial. A variety of techniques exist, with the conversion of NO2 into NO followed by chemiluminescent detection of NO being prevalent. Historically this conversion has used a catalytic approach (molybdenum); however, this has been plagued with interferences. More recently, photolytic conversion based on UV-LED irradiation of a reaction cell has been used. Although this appears to be robust there have been a range of observations in low-NOx environments which have measured higher NO2 concentrations than might be expected from steady-state analysis of simultaneously measured NO, O-3, jNO(2), etc. A range of explanations exist in the literature, most of which focus on an unknown and unmeasured "compound X" that is able to convert NO to NO2 selectively. Here we explore in the laboratory the interference on the photolytic NO2 measurements from the thermal decomposition of peroxyacetyl nitrate (PAN) within the photolysis cell. We find that approximately 5aEuro-% of the PAN decomposes within the instrument, providing a potentially significant interference. We parameterize the decomposition in terms of the temperature of the light source, the ambient temperature, and a mixing timescale (aEuro parts per thousand 0.4aEuro-s for our instrument) and expand the parametric analysis to other atmospheric compounds that decompose readily to NO2 (HO2NO2, N2O5, CH3O2NO2, IONO2, BrONO2, higher PANs). We apply these parameters to the output of a global atmospheric model (GEOS-Chem) to investigate the global impact of this interference on (1) the NO2 measurements and (2) the NO(2)aEuro-:aEuro-NO ratio, i.e. the Leighton relationship. We find that there are significant interferences in cold regions with low NOx concentrations such as the Antarctic, the remote Southern Hemisphere, and the upper troposphere. Although this interference is likely instrument-specific, the thermal decomposition to NO2 within the instrument's photolysis cell could give an at least partial explanation for the anomalously high NO2 that has been reported in remote regions. The interference can be minimized by better instrument characterization, coupled to instrumental designs which reduce the heating within the cell, thus simplifying interpretation of data from remote locations.