This paper examines the impact of the sampling error caused by the small size of the focused laser spot when using surface-enhanced Raman scattering (SERS) as a quantitative readout tool to analyze a sandwich immunoassay. The assay consists of a thin-film gold substrate that is modified with a layer of capture monoclonal antibodies (mAbs) and extrinsic Raman labels (ERLs) that consist of gold nanoparticle cores (60 nm diameter) coated with a monolayer of a Raman reporter molecule and a layer of human IgG mAbs to tag the captured antigen. The contribution of sampling error to the measurement is delineated first by constructing and analyzing an antigenic random accumulation model; this is followed by an experimental study of the analysis of an assay substrate using two different laser spot sizes. Both sets of findings indicate that the analysis with a small laser spot can lead to a sampling error (i.e., undersampling) much like that found when the size of a measured soil sample fails to accurately match that of a larger, more representative sample. That is, the smaller the laser spot size, the larger probable deviation in the accuracy of the measurement and the greater the imprecision of the measurement. Possible implications of these results with respect to the general application of SERS for quantitative measurements are also briefly discussed.

A method for rapid, highly portable immunoassays at low limits of detection has been developed by combining solid-phase microextraction membranes, gold nanoparticle labels, and surface-enhanced Raman scattering.

In this paper, we describe a novel method for analyte quantitation that does not rely on calibrants, internal standards, or calibration curves but, rather, leverages the relationship between disparate and predictable surface-directed analyte flux to an array of sensing addresses and a measured resultant signal. To reduce this concept to practice, we fabricated two flow cells such that the mean linear fluid velocity, U, was varied systematically over an array of electrodes positioned along the flow axis. This resulted in a predictable variation of the address-directed flux of a redox analyte, ferrocenedimethanol (FDM). The resultant limiting currents measured at a series of these electrodes, and accurately described by a convective-diffusive transport model, provided a means to calculate an “unknown” concentration without the use of calibrants, internal standards, or a calibration curve. Furthermore, the experiment and concentration calculation only takes minutes to perform. Deviation in calculated FDM concentrations from true values was minimized to less than 0.5% when empirically derived values of U were employed.