Toward nanofabrication of SERS substrates with two-photon polymerization

Abstract

Surface-enhanced Raman spectroscopy (SERS) has shown its ability to characterize biological substances down to a single-molecule level without a specific biorecognition mechanism. Various nanofabrication technologies enable SERS substrate prototyping and mass manufacturing. This study reports a complete cycle of design, fabrication, prototyping, and metrology of SERS substrates based on two-photon polymerization (2PP). Highly controllable direct laser writing allows the fabrication of individual nanopillars with up to an aspect ratio of 4. The developed SERS substrates show up to 10⁶ Raman signal enhancement, comparable to commercial substrates. Moreover, the rapid prototyping of the 2PP-printed SERS substrates takes from a minute to less than 2 hours, depending upon the nano-printing approach and aspect ratio requirements. The process is well-controlled and reproducible for achieving a uniform distribution of nanostructure arrays, allowing the SERS substrates to be used for a broad range of applications and the characterization of different molecules.

Fig. 1. Parameters of the voxel array. (a) Monolayer or so-called single-voxel array; (b) accumulative or so-called multivoxel array.

Fig. 2. AFM (a and c) and SEM (b and d) images of nanostructures fabricated with the single-voxel-based approach. Parameters here are the designed values given in Table 1 in ESI.

Fig. 3. AFM images of nanostructures fabricated with the multivoxel-based approach. (a and b) Correspond to the same lowest laser power with different height and pitch values. (c and d) Differ only in power values, and (e) corresponds to the highest applied laser power. Parameters here are the designed values given in Table 2 in ESI.

Fig. 4. (a) AFM image of Hamamatsu gold-coated dot array. (b and d) AFM and (c and e) SEM images of 2PP-fabricated nanostructures. The height of the structures is designed to be (b and c) 1 μm and (d and e) 1.6 μm. The pitch in XZ and YZ directions is respectively 360 and 380 nm, with a 300 nm voxel diameter.

Fig. 5. AFM image (a) and FDTD simulation (b) of the single-voxel-based array with structures of 600 nm pitch, 490 nm diameter, and 850 nm height. Color bars represent (a) the height of the structures and (b) the electromagnetic field intensity.

Fig. 6. (a) Spectrum of BPE using above mentioned SERS substrates, compared to a Hamamatsu substrate. The spectra are averaged over a map of 256 spectra, and the baseline is corrected using the intelligent polynomial baseline correction of Renishaw's WiRE software. (b) 30 μm × 30 μm experimental enhancement factor heatmap (2 μm resolution) max-normalized to examine collapsed vs. uniform nanostructures for the 1606 cm⁻¹ peak of 2.5 μM BPE. The red square indicates the surface area corresponding to the AFM-measured region. On the color bar, 0 corresponds to 1.02 × 10⁵ and 1 to 7.36 × 10⁵, respectively.

Fig. 7. The mean (a) and maximum (b) enhancement factor over the measured map of the SERS substrates for 1.5 μM BPE's peaks of interest. The error bars show the map-to-map variation of the EF values for the selected peak, while the mean and maximum values are calculated over each map.

Fig. 8. The mean (a) and minimum (b) LOD over the measured map of the SERS substrates for BPE's peaks of interest in spectra averaged over each map. The bars, i.e. mean values, are calculated over multiple measured maps. The error bars show the map-to-map variation of the LOD.