|COMPENSATED REFERENCING FOR SURFACE PLASMON RESONANCE SENSORS|
One difficulty for any sensor is environmental drift such as temperature, which in many cases can cause apparent sensor response not associated with the quantity being measured. For SPR sensors, the high sensitivity to the refractive index of the sample means that small changes in the temperature of the sample, with the resulting bulk refractive index change, appear as a sensor signal. Unless compensated for, this environmental signal may be confused with the desired signal resulting from a change in the surface chemistry. Other effects may require compensation, such as non-specific binding or surface fouling.
One method of compensation which has been tested at the University of Washington is the use of a reference channel, used in parallel with the sensor channel. In the graph above, both channels change with time, and the change in the signal will consist of the measured quantity, environmental drift, and other effects such as non-specific binding or surface fouling. If the surface chemistry of the reference channel is chosen carefully, then the specific response will occur only in the signal channel, while drift and non-specific effects should affect both channels to about the same degree. The reference channel response may therefore be subtracted from the signal channel response to provide a compensated measure of the instrument response.
In the graph above, the sensor channel was functionalized using anti-alkaline phosphatase (AP) antibodies, while the parallel reference channel was functionalized with anti-fluorescein antibodies. The reference channel neatly cancels out environmental drift and bulk refractive index changes as the solutions are changed.
For the experiment shown above, the reference and sensor channels were swapped, so that the anti-AP channel was used to compensate for the anti-F channel during detection of E. coli bacteria treated to contain fluorescein in their outer protein coats. Since bacteria are relatively large, diffusion is not significant and the kinetics of binding are slow. The reference channel helps to compensate for slow environmental drift over the hour required in this experiment.