Numerical simulation of photothermal deflection measurement: Efficiency limits of the quadrant photodetector.

Significant discrepancies are observed between the values of thermal conductivity and thermal diffusivity obtained from amplitude data and the corresponding values obtained from phase data in one and the same photothermal deflection measurement. Detailed investigation by means of numerical simulation, invoking the probe beam deflection as well as the signal generation in the detector, revealed that the quadrant detector suffers from a lack of efficiency with regard to measuring the deflection amplitude under the condition of relatively strong deflection. As a consequence, the measured amplitude significantly underestimates the actual deflection amplitude in the region of low excitation frequencies. It is demonstrated that this nonlinear behavior of the quadrant detector, never reported before, may lead to large systematic errors in the values for thermal conductivity and thermal diffusivity obtained by means of multiparameter fitting of the amplitude vs square root frequency curve. The origin of this detector inefficiency is discussed on the basis of comparing the time-resolved detector signal with the theoretically calculated deflection. It is found that phase data are much less affected by the reported detector inefficiency and can, therefore, be considered more reliable than the amplitude data. It is concluded that special care should be taken not to rely on the amplitude data unless compatibility with the corresponding phase data has been extensively proved.