Systematic Study of Various Functionalization Steps for Ultrasensitive Detection of SARS-CoV-2 with Direct Laser-Functionalized Au-LIG Electrochemical Sensors

Abstract

The 2019 coronavirus (COVID-19) pandemic impaired global health, disrupted society, and slowed the economy. Early detection of the infection using highly sensitive diagnostics is crucial in preventing the disease's spread. In this paper, we demonstrate electrochemical sensors based on laser induced graphene (LIG) functionalized directly with gold (Au) nanostructures for the detection of SARS-CoV-2 with an outstanding limit of detection (LOD) of ∼1.2 ag·mL^-1. To achieve the optimum performance, we explored various functionalization parameters to elucidate their impact on the LOD, sensitivity, and linearity. Specifically, we investigated the effect of (i) gold precursor concentration, (ii) cross-linker chemistry, (iii) cross-linker and antibody incubation conditions, and (iv) antigen-sensor interaction (diffusion-dominated incubation vs pipette-mixing), as there is a lack of a systematic study of these parameters. Our benchmarking analysis highlights the critical role of the antigen-sensor interaction and cross-linker chemistry. We showed that pipette-mixing enhances sensitivity and LOD by more than 1.6- and 5.5-fold, respectively, and also enables multimodal readout compared to diffusion-dominated incubation. Moreover, the PBA/Sulfo-NHS: EDC cross-linker improves the sensitivity and LOD compared to PBASE. The sensors demonstrate excellent selectivity against other viruses, including HCoV-229E, HCoV-OC43, HCoV-NL63, and influenza H5N1. Beyond the ability to detect antigen fragments, our sensors enable the detection of antigen-coated virion mimics (which are a better representative of the real infection) down to an ultralow concentration of ∼5 particles·mL^-1.

Keywords: SARS-COV-2; antigen; biosensor; electrochemical; laser-induced graphene.

Figure 1.. Fabrication of L-Au/LIG.

First, a CO₂ laser carbonizes polyimide into laser induced graphene (LIG), patterning the working electrode (WE). Au is directly doped into LIG through a second lasing step to form L-Au/LIG. Effect of different Au concentrations is studied. Carboxylic groups are activated on the WE surface with PBASE or PBA/Sulfo-NHS: EDC as crosslinker. Effect of incubation humidity is studied. SARS-CoV Spike S2 antibody covalently bonds to the carboxylic groups. The effect of incubation temperature for this step is studied. Bovine serum albumin (BSA) is then used to block unbound sites, followed by testing with the virus samples. Specifically, two different antigen mixing techniques are investigated in this study.

Figure 2.. Electrochemical characterization of L-Au/LIG.

(a) Electrochemically active surface area (ECSA) of LIG and L-Au/LIG functionalized with different concentrations of HAuCl₄ (the gold precursor solution). Data are presented as mean (n=9). Error bars are standard error of the mean. (b) $k^0$ of LIG and L-Au/LIG functionalized with different concentrations of HAuCl₄. Data are presented as mean (n=9). Error bars are S.E.M.

Figure 3.. Material characterization of LIG and L-Au/LIG sensor.

(a) SEM image of the LIG confirms the 3D porous structure of the material (with 6,500× magnification). Scale bar is 20 μm. (b) Average Raman spectra of the 1-pass (n=24) and 2-pass LIG sensor (n=8). Error bands are standard deviation (STD). (c) XPS results of 1-pass LIG (n=4), 2-pass LIG (n=4), and L-Au/LIG sensors (n=4) showing their chemical components. Error bands are STD.

Figure 4.. L-Au/LIG electrochemical sensor performance.

(a) Differential pulse voltammetry (DPV) curves after each step of functionalization process with PBASE crosslinker. Antigen concentration is 30 ng.$m{L}^{-1}$. (b) Representative differential pulse voltammetry (DPV) curves after each step of functionalization process with PBA/Sulfo-NHS: EDC crosslinker. Antigen concentration is 30 ng.$m{L}^{-1}$. (c) Representative DPV curves in response to varying concentrations of SARS-CoV-2 antigen. (d) The charge-transfer resistance ($R_{ct}$) after each step of functionalization process for PBASE crosslinker. Antigen concentration is 30 ng.$m{L}^{-1}$. Data are presented as mean with n = 3. Error bars are standard error of the mean. (e) The charge-transfer resistance ($R_{ct}$) after each step of functionalization process for PBA/Sulfo-NHS: EDC crosslinker. Antigen concentration is 30 ng.$m{L}^{-1}$. Data are presented as mean with n = 3. Error bars are standard error of the mean.

Figure 5.. Studying the effect of crosslinker chemistry and antigen-sensor interaction (mixing method).

(a) Charge transfer resistance ($R_{ct}$) of L-Au/LIG sensors (normalized w.r.t. $R_{ct}$ of samples after BSA treatment) in response to various concentrations of SARS-CoV-2 antigen in PBS using PBASE vs. PBA/Sulfo-NHS: EDC crosslinkers. Sensors are incubated with antigen using the diffusion method. Data are presented as mean (n=12). Error bars are standard error of the mean (S.E.M.). (b) Normalized $R_{ct}$ of 1-pass LIG and 2-pass LIG sensors (with PBA linker chemistry) in response to various concentrations of SARS-CoV-2 antigen in PBS using diffusion-dominated incubation vs. pipette-mixing methods. Data are presented as mean (n=12). Error bars are S.E.M. (c) Normalized $R_{ct}$ of L-Au/LIG sensors in response to various concentrations of SARS-CoV-2 antigen in PBS using diffusion-dominated incubation vs. pipette-mixing methods. Data are presented as mean (n=12). Error bars are S.E.M.

Figure 6.. Investigation of the sensor performance in saliva and with virion-like mimics.

(a) Charge transfer resistance ($R_{ct}$) derived from EIS data (normalized w.r.t. $R_{ct}$ of samples after BSA treatment) after incubation with various concentrations of SARS-CoV-2 antigen in artificial saliva (3—300 ag.$m{L}^{-1}$). Data are presented as mean (n=9). Error bars are standard error of the mean (S.E.M.). (b) Selective response of L-Au/LIG sensors against 30 ag.$m{L}^{-1}$ concentration of SARS-CoV-2 and non-target antigens. Data are presented as mean (n=9). Error bars are S.E.M. (c) $R_{ct}$ after incubation with various concentrations of SARS-CoV-2 virus mimics in PBS (2 × 10² — 2 × 10⁴ $particles.m{L}^{-1}$) using 2-pass WE LIG sensors. Data are presented as mean (n=12). Error bars are S.E.M. (d) $R_{ct}$ after incubation with various concentrations of SARS-CoV-2 virus mimics in PBS (2 × 10²—2 × 10⁴ $particles.m{L}^{-1}$) using L-Au/LIG sensors. Data are presented as mean (n=12). Error bars are S.E.M.