Optimization of immersion direction and time of covalently self-assembled monolayer gold nanourchins on glass as SERS substrate

A glass-based surface-enhanced Raman scattering (SERS) substrate using 90 nm gold nanourchin (GNU) was fabricated as a localized surface plasmon resonance (LSPR) sensor. Self-assembly of GNU monolayer is achieved by silanizing the glass surface with amino-terminated (3-aminopropyl) triethoxysilane (APTES) and covalently grafting with thiol-terminated mercaptoundecanoic (MUA). Six substrates are used in vertical direction for immobilization of GNU at two concentrations of 5 pM (S1, S2, S3) and 15 pM (S4, S5, S6) for three sets of time 1, 3, and 5 h in each case, respectively, and one substrate (S7) is immersed horizontally at 15 pM for 5 h. The substrates are assessed using reflective surface-enhanced Fourier transform near-infrared (SEFT-NIR) and SERS. The results show that S3 (5 pM, 3 h) ≈ 96%, S4 (15 pM, 1 h) ≈ 97%, and S7 ≈ 97.8% have the highest GNU coverage, respectively. However, higher GNU coverage does not necessarily produce the maximum SERS signal as in our case S2 > S3 > S1 and S4 > S5 > S6. Principal component analysis (PCA) results can be interpreted as showing significant variance between the groups with less variation between the trials in the SEFT-NIR data, while the PC plots of the Raman data show that the spectra of all groups look very similar. Both techniques showed S3 and S4, and S7 as optimized substrates. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) indicated a dispersed distribution of GNU with peak heights of ≤ 10 nm. This work is fundamentally important for designing and modeling LSPR-based nanosensors.