Modeling of Label-Free Optical Waveguide Biosensors with Surfaces Covered Partially by Vertically Homogeneous and Inhomogeneous Films.

Optical Waveguide Lightmode Spectroscopy (OWLS) is widely applied to monitor protein adsorption, polymer self-assembly, and living cells on the surface of the sensor in a label-free manner. Typically, to determine the optogeometrical parameters of the analyte layer (adlayer), the homogeneous and isotropic thin adlayer model is used to analyze the recorded OWLS data. However, in most practical situations, the analyte layer is neither homogeneous nor isotropic. Therefore, the measurement with two waveguide modes and the applied model cannot supply enough information about the parameters of the possible adlayer inhomogeneity and anisotropy. Only the so-called quasihomogeneous adlayer refractive index, layer thickness, and surface mass can be determined. In the present work, we construct an inhomogeneous adlayer model. In our model, the adlayer covers the waveguide surface only partially and it has a given refractive index profile perpendicular to the surface of the sensor. Using analytical and numerical model calculations, the step-index and exponential refractive index profiles are investigated with varying surface coverages from 0 to 100%. The relevant equations are summarized and three typically employed waveguide sensor structures are studied in detail. We predict the errors in the calculated optogeometrical parameters of the adlayer by simulating the OWLS measurement on an assumed inhomogeneous adlayer. We found that the surface coverage has negligible influence on the calculated refractive index below film thicknesses of 5nm; the calculated refractive index is close to the refractive index of the adlayer islands. But the determined quasihomogeneous adlayer refractive index and surface mass are always underrated; the calculated quasihomogeneous thickness is heavily influenced by the surface coverage. Depending on the refractive index profile, waveguide geometry, and surface coverage, the thickness obtained from the homogeneous and isotropic modeling can even take negative and largely overestimated values, too. Therefore, experimentally obtained unrealistic adlayer values, which were dismissed previously, might be important indicators of layer structure.