Many scientific instruments utilise multiple element detectors, e.g. CCD's or photodiode arrays, to monitor the change in a position of an optical pattern. For example, instruments for affinity biosensing based on surface plasmon resonance (SPR) or resonant mirror are equipped with such detectors. An important and desired property of these bioanalytical instruments is that the calculation of the movement or change in shape follows the true change. This is often not the case and it may lead to linearity errors, and to sensitivity errors. The sensitivity is normally defined as the slope of the calibration curve. A new parameter is introduced to account for the linearity errors, the sensitivity deviation, defined as the deviation from the undistorted slope of the calibration curve. The linearity error and the sensitivity deviation are intimately related and the sensitivity deviation may lead to misinterpretation of kinetic data, mass transport limitations and concentration analyses. Because the linearity errors are small (e.g. 10 pg/mm2 of biomolecules on the sensor surface) with regard to the dynamic range (e.g. 30 000 pg/mm2), they can be difficult to discover. However, the linearity errors are often not negligible with regard to a typical response (e.g. 0-100 pg/mm2), and may therefore cause serious problems. A method for detecting linearity errors is outlined. Further on, this paper demonstrates how integral linearity errors of less than 1% can result in a sensitivity deviation of 10%, a value that in our opinion cannot be ignored in biospecific interaction analysis (BIA). It should also be stressed out that this phenomenon also occurs in other instruments using array detectors. (C) 2000 Elsevier Science S.A.Many scientific instruments utilize multiple element detectors, e.g. CCD's or photodiode arrays, to monitor the change in a position of an optical pattern. For example, instruments for affinity biosensing based on surface plasmon resonance (SPR) or resonant mirror are equipped with such detectors. An important and desired property of these bioanalytical instruments is that the calculation of the movement or change in shape follows the true change. This is often not the case and it may lead to linearity errors, and to sensitivity errors. The sensitivity is normally defined as the slope of the calibration curve. A new parameter is introduced to account for the linearity errors, the sensitivity deviation, defined as the deviation from the undistorted slope of the calibration curve. The linearity error and the sensitivity deviation are intimately related and the sensitivity deviation may lead to misinterpretation of kinetic data, mass transport limitations and concentration analyses. Because the linearity errors are small (e.g. 10 pg/mm2 of biomolecules on the sensor surface) with regard to the dynamic range (e.g. 30 000 pg/mm2), they can be difficult to discover. However, the linearity errors are often not negligible with regard to a typical response (e.g. 0-100 pg/mm2), and may therefore cause serious problems. A method for detecting linearity errors is outlined. Further on, this paper demonstrates how integral linearity errors of less than 1% can result in a sensitivity deviation of 10%, a value that in our opinion cannot be ignored in biospecific interaction analysis (BIA). It should also be stressed out that this phenomenon also occurs in other instruments using array detectors.

A new, multiple wavelength surface plasmon resonance apparatus for imaging applications is presented. It can be used for biosensing, e.g., for monitoring of chemical and biological reactions in real time with label-free molecules. A setup with a fixed incident angle in the Kretschmann configuration with gold as the supporting metal is described, both theoretically and experimentally. Simulations of the sensor response based on independently recorded optical (ellipsometric) data of gold show that the sensitivity for three-dimensional recognition layers (bulk) increases with increasing wavelength. For two-dimensional recognition layers (adlayer) maximum sensitivity is obtained within a limited wavelength range. In this situation, the rejection of bulk disturbances, e.g., emanating from temperature variations, decreases, with increasing wavelength. For imaging surface plasmon resonance the spatial resolution decreases with increasing wavelength. Hence, there is always a compromise between spatial resolution, bulk disturbance rejection, and sensitivity. Most importantly, by simultaneously using multiple wavelengths, it is possible to maintain a high sensitivity and accuracy over a large dynamic range. Furthermore, our simulations show that the sensitivity is independent of the refractive index of the prism. (C) 2000 American Institute of Physics. [S0034-6748(00)02909-9].