Influences of Detection Pinhole and Sample Flow on Thermal Lens Detection in Microfluidic Systems

Abstract

Thermal lens microscopy (TLM), due to its high temporal ( $${\sim }\mathrm{ms}$$ ) and spatial resolution ( $${\sim }\upmu \mathrm{m}$$ ), has been coupled to lab-on-chip chemistry for detection of a variety of compounds in chemical or biological fields. Due to the very short optical path length (usually below 100 $$\upmu \mathrm{m}$$ ) in a microchip, the sensitivity of TLM is unfortunately still 10 to 100 times lower than conventional TLS with 1 cm sample length. Optimization of the TLM optical configuration was made with respect to different pinhole aperture-to-beam size ratios for the best signal-to-noise ratio. In the static mode, the instrumental noise comes mainly from the shot noise of the probe beam when the chopper frequency is over 1 kHz or from the flicker noise of the probe beam at low frequencies. In the flowing mode, the flow-induced noise becomes dominant when the flow rate is high. At a given flow rate, fluids with a higher density and/or a higher viscosity will cause larger flow-induced noise. As an application, a combined microfluidic flow injection analysis ( $$\upmu \mathrm{FIA}$$ )–TLM device was developed for rapid determination of pollutants by colorimetric reactions. Hexavalent chromium [Cr(VI)] was measured as a model analyte. Analytical signals for 12 sample injections in 1 min have been recorded by the $${\upmu }$$ FIA–TLM. For injections of sub- $$\upmu $$ L samples into the microfluidic stream in a $$50\,\upmu \mathrm{m}$$ deep microchannel, a limit of detection of $$4\,\mathrm{ng}{\cdot }\mathrm{mL}^{-1}$$ was achieved for Cr(VI) in water at 60 mW excitation power.