In situ electrochemical regeneration of nanogap hotspots for continuously reusable ultrathin SERS sensors

Abstract

Surface-enhanced Raman spectroscopy (SERS) harnesses the confinement of light into metallic nanoscale hotspots to achieve highly sensitive label-free molecular detection that can be applied for a broad range of sensing applications. However, challenges related to irreversible analyte binding, substrate reproducibility, fouling, and degradation hinder its widespread adoption. Here we show how in-situ electrochemical regeneration can rapidly and precisely reform the nanogap hotspots to enable the continuous reuse of gold nanoparticle monolayers for SERS. Applying an oxidising potential of +1.5 V (vs Ag/AgCl) for 10 s strips a broad range of adsorbates from the nanogaps and forms a metastable oxide layer of few-monolayer thickness. Subsequent application of a reducing potential of -0.80 V for 5 s in the presence of a nanogap-stabilising molecular scaffold, cucurbit[5]uril, reproducibly regenerates the optimal plasmonic properties with SERS enhancement factors ≈10⁶. The regeneration of the nanogap hotspots allows these SERS substrates to be reused over multiple cycles, demonstrating ≈5% relative standard deviation over at least 30 cycles of analyte detection and regeneration. Such continuous and reliable SERS-based flow analysis accesses diverse applications from environmental monitoring to medical diagnostics.

Fig. 1. Preparation and in-flow EC-SERS analyte detection, cleaning, and regeneration with MLagg-CB[5].

a Preparation of MLagg-CB[5] SERS aggregate from self-assembly of AuNPs with CB[5], followed by deposition onto solid support. Photo and scanning electron micrograph (SEM) show MLagg-CB[5] deposited on FTO-coated glass. b Schematic illustraties the integration of an MLagg-CB[5] into an EC-SERS flow system. c Cross-section of the EC-SERS flow cell (CE = counter electrode, RE = reference electrode, and WE = working electrode). d Schematic of in situ electrochemical SERS analyte detection and cleaning/regeneration protocol. Potentials are vs Ag/AgCl. e SERS spectra from: initial MLagg-CB[5] (grey), after detection of 10 µM adenine (ADN) (red), after oxidative cleaning step (blue), and after regeneration step (black). ADN peak at 732 cm⁻¹ is marked by asterisk. SERS spectra are collected with 1 s integration time and 1 mW 785 nm laser.

Fig. 2. Analyte detection, cleaning, and regeneration cycles with and without CB[5].

a SERS spectra from 30 cycles of 10 µM ADN detection (red) and regeneration with CB[5] (black). Spectra are offset for clarity. b ADN peak areas (${\nu }_{\mathrm{ADN}}$ = 732 cm⁻¹) from the SERS spectra of each ADN detection and cleaning/CB[5]-regeneration cycle. Dotted horizontal line represents the average ADN peak area of all analyte detection cycles. c CB[5] peak areas (${\nu }_{\mathrm{CB}\left[5\right]}$ = 830 cm⁻¹, black circles) and integrated SERS background (purple squares) for the CB[5]-regenerated MLagg. Dotted horizontal line represents the average CB[5] peak area of all regeneration cycles. d Overlaid SERS spectra from 15 cycles of 10 µM ADN detection and e after regeneration without CB[5]. A constant background was subtracted from all spectra to facilitate comparison across 15 cycles. f ADN peak areas (${\nu }_{\mathrm{ADN}}$ = 732 cm⁻¹) from the SERS spectra of each ADN detection and cleaning/regeneration cycle without CB[5]. g CB[5] peak areas (${\nu }_{\mathrm{CB}\left[5\right]}$ = 830 cm⁻¹, black circles) and integrated SERS background (purple squares) for the MLagg regenerated without CB[5].

Fig. 3. Morphology changes with or without CB[5] during regeneration.

a DF scattering spectra and (b) representative SEM images of the MLagg-CB[5] before and after undergoing 10 and 30 cycles of analyte detection, cleaning, and regeneration with CB[5]. c DF scattering spectra and (d) representative SEM images of the MLagg-CB[5] throughout different cycles of analyte detection, cleaning, and regeneration without CB[5]. Yellow arrows in (d) point to bridges formed between neighbouring AuNPs. For the DF spectra, solid lines and shaded area represent mean and ±1 s.d. of n = 150 spectra obtained across the area of a MLagg-CB[5]. Grey line highlights initial chain mode peak wavelength. Spectra are offset for clarity.

Fig. 4. In situ analyte switching.

a Sequential SERS spectra from alternating detection of 100 µM ADN (red) and 100 µM CYT (blue) with the corresponding regenerated MLagg-CB[5] (black). A 50:50% mixture of 100 µM ADN and 100 µM CYT (purple) is also tested. b Normalized peak areas of ADN (${\nu }_{\mathrm{ADN}}$ = 732 cm⁻¹, red) and CYT (${\nu }_{\mathrm{CYT}}$ = 792 cm⁻¹, blue) per detection-regeneration cycle. c Sequential spatially-averaged SERS spectra from alternating detection of 100 µM ADN (red) and 100 µM 4-biphenylthiol (BPT), 4,4-biphenyldithiol (BPDT), 1,4-benzenedithiol (BDT), 2-naphthalenethiol (2-NT), and 4-mercaptopyridine (4-MPY) with the corresponding regenerated MLagg-CB[5] (black). d Normalized peak area of ${\nu }_{\mathrm{ADN}}$ per ADN/thiol detection-regeneration cycle. Peak areas are plotted as the mean with error bars representing ±1 s.d. of n = 10 spectra obtained from different points across the MLagg-CB[5] area. For cycle 6 (marked by asterisk), alternative ADN peak at 970 cm⁻¹ is used due to overlap with BDT peak at 732 cm⁻¹. e Sequential SERS spectra from detection of a series of biological compounds: ADN, CYT, hypoxanthine (HYP), creatinine (CR), nicotinamide (NAM), paracetamol (PAR), norepinephrine (NEPI), tryptophan (TRP), nicotinic acid (NIA), methylene blue (MB), and ADN (red) again with the corresponding regenerated MLagg-CB[5] (black). Time (t) axis marks progress of both analyte detection and ReSERS (30 s between each detection cycle). All SERS spectra collected a1 s integration time and with 1 mW 785 nm excitation laser.

Fig. 5. Quantitative analysis of ADN.

a ADN peak areas at ${\nu }_{\mathrm{ADN}}$ from sequentially measured calibration standards (0.1 to 100 µM) in 50 mM potassium phosphate buffer (pH 7.0) and regenerated MLagg-CB[5] (black circles). b Sequential SERS spectra from ADN calibration standards measured from the same MLagg-CB[5] SERS substrate. SERS spectra collected at 1 s integration time and 1 mW 785 nm laser. c Relative peak areas of CB[5] (${\nu }_{\mathrm{CB}\left[5\right]}$ = 830 cm⁻¹) and ADN (${\nu }_{\mathrm{ADN}}$.  = 732 cm⁻¹) vs [ADN] (log scale). Peak areas are plotted as the mean with error bars representing ±1 s.d. from n = 3 measurements of an ADN calibration standard using the same MLagg-CB[5] SERS substrate after multiple ReSERS cycles. Solid lines are fits. d Schematic of the standard addition analysis of ADN in urine. e ADN peak area vs added [ADN] standard. Points are plotted as mean ±1 s.d. from n = 3 measurements of an ADN calibration standard using the same MLagg-CB[5] SERS substrate after multiple ReSERS cycles. Grey solid line is linear fit, dashed grey line is linear extrapolation. Black arrow points to the extrapolated x-intercept. Red dashed line highlights the trend towards non-linearity at higher concentration. Inset shows close-up of ADN peak from SERS spectra of calibration standards.