Extended Data Fig. 4: Single-molecule sensitivity by BiASERS.
From: Digital colloid-enhanced Raman spectroscopy by single-molecule counting
The concentration-dependent heatmaps of each analyte involved in the BiASERS measurements (left and middle) and the histograms of the contribution of (a, c, d) crystal violet (CV) or (b) 4-NBT in each spectrum (right). The pure events are defined as the probability of one of the analytes approaching 0 (or 1) and the mixed events are defined as the probability in between. (a) Mixtures of CV and Nile blue (NB) measured using Hya-Ag colloids. At 10−8 M of CV/NB, both analytes show clear signal in nearly all voxels, exhibiting a Gaussian-like histogram of the probability of CV. On the contrary, at 10−9 M CV, some spectra exhibit neither the CV nor NB signal but within all signal spectra, there is a growing proportion of pure-signaled ones, leading to an increase of 1/0 measurements than those in the middle. At 10−10 M CV, most signal spectra exhibit pure signals with extremely rare ones showing both signals. Since these pure-signaled spectra should be statistically dominated by single-molecule events, the single-molecule regime for CV should be below 10−9 M. (b) Mixture of 4-NBT and 4-chlorothiophenol (4-CBT) measured using Hya-Ag colloids, exhibiting the expected single-molecule behavior for 4-NBT below 10−8 M. (c) Mixtures of CV and NB measured by citrate-Ag colloids and (d) citrate-Au colloids. The same trend can be observed in the other types of SERS colloids though at different analyte concentrations due to differences in hotspot density/intensity, surface binding affinity, among other properties.