372
J. Eur. Opt. Society-Rapid Publ. 22, 37( 2026)
Fig
. 2.( a) PAS signal amplitude of samples C, C-Au and the substrate( without Au), excited at normal incidence and modulated with 54Hz, 408 Hz and 625 Hz chopper frequency; the excitation is at normal incidence with linear polarization. All the signals are normalized to the incident laser power at the corresponding wavelength.( b) Normalized PAS signal for both nanostructured samples is obtained by subtracting the substrate signal from the NW signal, and normalizing each spectra to its maximum. The resonant modes appear at higher frequencies, and their bandwidth shrinks.( c) PAS difference between the NWs and the substrate for Sample C and Sample C-Au agree with the NW modes predicted by simulations.
wavelength dependence of this effect in ensembles of thinner NWs by reflection spectroscopy; the optical response was measured by placing a photo-diode at only one angle in the space to observe the specular reflection [ 11 ]. This scattering-influenced signal characterizes reflection dissymmetry under LCP and RCP excitations, but it should not be confused with the real, absorption-defined CD.
Our PAS set-up allows for the broadband and sensitive characterization of chirality from absorption response. We next fix the chopper frequency to 1225 Hz, and investigate Sample C-Au under oblique incidence, Figure 3. In the schematic, we show the sample orientation which obeys the condition of extrinsic chirality: the non-planar triad of vectors between the light wave-vector ~ k, the surface normal ~ n, and the direction perpendicular to the Au sidewalls ~ s. In Figure 3a we plot the PAS signal difference between the NW and the substrate, when Sample C-Au is excited with LCP or RCP, at ± 20 °. Inawellalignedset-up, substrate absorption is not dependent on the incident polarization, so the difference comes from the nanostructured part. At �20 ° NWs better absorb RCP, while this behavior inverts at + 20 ° across the whole wavelength range. Figure 3b plots PAS signal differences between LCP and RCP excitation for Sample C-Au: as expected, A LCP – A RCP are of the opposite sign between + 20 ° and �20 ° across the whole wavelength range, while at normal incidence, difference tends to zero. Next, the normalized absorption difference is calculated as 100( A LCP � A RCP)/( A LCP + A RCP) [%]; Figure 3c plots this difference at two wavelengths, as a function of angle. CD dependence on h is in line with our previous results, while we note that the signal can be enhanced by increasing the NW surface density, as shown with single wavelength CD measurements in Ref. [ 23 ]. Figure 3d plots the simulated absorption density when the single C-Au NW is excited at 720 nm, under + 20 ° or �20 °, with LCP and RCP. The Au layer confines and enhances the absorption; its asymmetry leads to stronger coupling with LCP, thus revealing the nanoscale origin of the extrinsic chirality. Therefore, PAS with a widely tunable laser and frequency is a scattering-free method able to detect polarization response stemming from asymmetric structuring of a thin Au layer; even though it reveals the overall absorption signal of the NW ensemble, the possibility of high modulation frequency allows to access the upper parts of the samples and detect low chirality.
Besides precise microscope positioning and beam-waist tuning on the nanostructured areas [ 26 ], the novelty of our set-up lies in parallel chopper frequency and incident power tuning. Namely, with nanostructures on thermally insulating substrates( such as commercial soda-lime glass), PAS is highly sensitive to the nanomaterial on the top, as the nanostructure absorption creates the heat wave which stays confined to the top. In this case, even with powers on the order of mW and at lower frequencies( f = 81 Hz),