Near-infrared photoluminescence of orange color standards – then and now

Abstract

The presence of near-infrared (NIR) photoluminescence has been recently reported in some of the second series of Ceramic Color Standards (CCSII) that are widely used in the calibration and performance evaluation of color measuring instruments. The impact of this photoluminescence effect can cause significant colorimetric errors particularly for broadband measurements using a detector with high spectral responsivity in the NIR region. The magnitude of this effect has been demonstrated for specific color standards and specific instrument systems but has not been unambiguously quantified to allow general predictions or absolute comparisons of different instrument designs or different ceramic tiles. Here we present absolute NIR photoluminescence measurements on three different formulations of the CCSII orange ceramic color standard using the National Research Council of Canada (NRC) Reference Spectrofluorimeter whose spectral range has been recently extended to 1000 nm. The validation for this extended spectral range is shown by comparison of an independent method of instrument calibration using a different combination of physical standards. It is convincingly shown that the two different leaded formulations of this ceramic orange standard issued in 2000 and 2011 have no significant photoluminescence and thus can be used for calibration with any type of spectrophotometer design whereas the unleaded formulation issued in 2011 has significant NIR photoluminescence and should not be used for instrument calibration and validation over an extended range into the NIR for certain spectrophotometers with relatively high throughput in the NIR region, such as a spectrophotometer with polychromatic illumination mode using a xenon source or with monochromatic illumination mode using a Si detector. It is shown that for colorimetric applications, the impact of this NIR fluorescence is only significant for the latter spectrophotometer design with broadband detection with a Si or spectrally flat detector and is negligible with a narrowband PMT detector. These calculated colorimetric results are also consistent with previously estimated colorimetric errors for this type of orange CCSII ceramic tile used to transfer calibration between these two types of detector systems.