Calibration of fiber optic based surface plasmon resonance sensors in aqueous systems

Abstract This manuscript addresses two issues with calibration of surface plasmon resonance sensors which can track refractive index changes to measure correlated bulk properties. First, employing non-air references do not return the traditional parabolic dip in the SPR spectra; instead the returned SPR spectra are more ‘derivative’ in shape. Investigated are five different ways to calibrate SPR spectra when non-air references are employed. The minimum hunt method (MHM) calculates the position of this minimum by fitting a parabola to the curve. MHM is shown to consistently achieve prediction errors of 3×10 −4 RI units (RIU) using an air reference and a RI calibration set of aqueous sucrose samples measured with an Abbe refractometer accurate to 1×10 −4 RIU. Use of principal component regression (PCR) with air or water references generates prediction errors at best the same as MHM but worse as the concentration range of samples increases. Second, a method for calibrating SPR spectra across a wide temperature range is presented. It is shown that this method is capable of successfully mitigating the effect of temperature drifts up to 20 °C. MHM was subsequently used to predict the concentration of 0.00–6.99 wt.% KCl (aq) samples between 6 and 29 °C, a range found in surface ocean waters. Prediction error as small as 0.073 wt.% correlates to 2×10 −4 RIU shows MHM holds over a wide dynamic range.