Characterization of graphene-gold spr biosensor based on statistical analysis of reflection spectrum

A new approach using the full width by autocorrelation method (FWAM) is presented to estimate the real width of a surface plasmon resonance biosensor spectrum. The novelty of statistical analysis is the expansion an arbitrary angular distribution in terms of Gaussian functions and the use of Gaussian parameters inside FWAM. A new formalism is presented for the sensitivity which the amplification strategy of FWAM is appeared at the evaluation of sensitivity. It is found that the FWAM causes the spectrum width of an angular distribution to be between two limit values calculated from the full width at half maximum (FWHM) and the full width by standard deviation. It also leads to that the mean value of the relative sensitivity is dependent on incident wavelength and 5, 6 and 2 orders of magnitude is better than the mean value achieved by the regular intensity-amplitude relative sensitivity, for wavelengths of λ = 660 nm, λ = 785 nm, and λ = 850 nm, respectively. (paper)