Abstract
Noroviruses are a leading cause of acute gastroenteritis worldwide, and their rapid and reliable detection remains a significant challenge for current diagnostic technologies. In this work, we present a plasmonic metasurface platform based on Y-shaped nanocavity arrays designed for label-free viral detection using surface-enhanced Raman spectroscopy (SERS). The nanostructured metasurface was fabricated using electron-beam lithography and engineered to support plasmonic resonances near the 785 nm excitation wavelength, enabling strong electromagnetic field localization and Raman signal enhancement. SERS measurements were performed on human norovirus (HNoV) and murine norovirus (MNV), producing distinct vibrational fingerprints associated with viral capsid biomolecules. Despite the structural similarity between both viruses, multivariate statistical analysis based on principal component analysis (PCA) enabled clear discrimination between their spectral signatures, with the first two principal components explaining more than 98% of the total variance captured. These results demonstrate that plasmonic metasurfaces combined with SERS fingerprinting and statistical analysis provide a powerful strategy for the label-free identification of closely related viral pathogens. The proposed platform highlights the potential of nanophotonic biosensors for rapid pathogen detection in biomedical diagnostics and environmental monitoring.
1 Introduction
Plasmonic and photonic crystals have become key components in modern nanophotonics due to their unique ability to manipulate light at the nanoscale (Yablonovitch, 1987; ; ). Metal–dielectric nanostructures, in particular, enable strong electromagnetic field confinement and localized surface plasmon resonances (LSPR), providing powerful signal enhancement mechanisms for label-free molecular detection (Vitiello, 2014; ; Zito et al., 2021; ). Combining Surface Plasmon Resonance (SPR) with Surface-Enhanced Raman Scattering (SERS) enables the simultaneous monitoring of binding kinetics and molecular fingerprints, achieving superior sensitivity and selectivity in biosensing applications (; ; ; ; ).
Recent advances in nanolithography have enabled the fabrication of highly reproducible and tunable plasmonic substrates capable of supporting multiple resonance modes (Yu et al., 2011; Yu and Capasso, 2014; ; ; ). Among these platforms, nanocavity-based meta-crystals have emerged as particularly promising architectures due to their ability to support super-radiant modes and strong plasmonic coupling. In this context, iso-Y-shaped geometries provide a distinct advantage over conventional nanohole arrays by generating intense localized electromagnetic hotspots at the cavity vertices, which are essential for amplifying the Raman scattering cross-section (; ; ). The Y-shaped nanostructure represents a particularly interesting configuration. It consists of a symmetric cavity with three identical arms arranged at 120°, capable of supporting highly localized plasmon modes along its edges. When implemented as a periodic array of nanocavities, the Y-shaped architecture exhibits strong plasmonic activity and enhanced nonlinear optical responses, making it a promising platform for highly sensitive SERS detection (; ; ; ; ; ).
Noroviruses (NoV), first identified in 1972, belong to the Caliciviridae family and are single-stranded RNA viruses responsible for a large fraction of non-bacterial acute gastroenteritis cases worldwide. Conventional diagnostic methods, including reverse transcription quantitative polymerase chain reaction (RT-qPCR), enzyme-linked immunosorbent assays (ELISA), and immunochromatographic tests, provide reliable detection but are often limited by high costs, processing time, and the need for labeling or specialized reagents. These limitations highlight the need for rapid and label-free detection approaches, such as SERS-based sensing (; ; ; Vinjé, 2015; ; Wang et al., 2021). Recent efforts have therefore focused on improving the sensitivity, portability, and response time of NoV detection systems. Emerging molecular techniques, including loop-mediated isothermal amplification (LAMP) and recombinase polymerase amplification (RPA), enable nucleic acid amplification under isothermal conditions, thereby eliminating the need for thermocycling and reducing assay time. In parallel, biosensor-based technologies—particularly electrochemical and optical sensors—have been increasingly investigated to provide real-time, label-free detection with improved analytical performance. These advances aim to bridge the gap between high-sensitivity laboratory diagnostics and rapid point-of-care testing (Wang et al., 2021).
Surface-Enhanced Raman Scattering has recently been explored as a powerful tool for NoV detection. demonstrated the feasibility of detecting norovirus using a dual SERS nanotag immunoassay based on magnetic functionalized MoO3 nanocubes assembled on a graphene oxide substrate. In this approach, the sensing performance relied strongly on the specificity of antigen–antibody interactions forming a sandwich immunocomplex.
Building upon these advances, the present work investigates the use of engineered plasmonic nanostructures for label-free viral discrimination. Specifically, we evaluate the physical and sensing performance of dual Y-shaped nanostructured SERS substrates designed to enhance electromagnetic field confinement while maintaining high signal reproducibility. To demonstrate their analytical capability, the substrates were used to detect and discriminate between two closely related caliciviruses—human norovirus (HNoV) and murine norovirus (MNV-1). Due to their structural and biochemical similarities, these viruses represent an ideal model system to evaluate the discriminative capability of the plasmonic platform through SERS spectral fingerprinting combined with Principal Component Analysis (PCA). To the best of our knowledge, this work represents one of the first demonstrations of label-free discrimination between Human Norovirus (HNoV) and Murine Norovirus (MNV) using engineered plasmonic metasurfaces combined with SERS fingerprinting and multivariate analysis. Unlike most previously reported SERS-based viral sensing strategies, which rely on antibodies, aptamers, Raman reporters, molecular amplification, or target-specific functionalization, the proposed platform exploits exclusively the intrinsic vibrational fingerprints of the viral particles. This capability enables the discrimination of two closely related caliciviruses through a simplified, label-free nanophotonic sensing approach, highlighting the potential of plasmonic metasurfaces for rapid pathogen classification.
2 Materials and methods
2.1 Materials
Indium Tin Oxide (ITO)-coated BK7 glass substrates, ZEP520A electron-beam resist, and gold pellets (99.99% purity) were purchased from commercial suppliers (ZEONREX® and Sigma-Aldrich). Human Norovirus (HNoV) and Murine Norovirus (MNV-1) samples were employed as model analytes for the SERS measurements. All chemicals and reagents were of analytical grade and used without further purification.
The viral samples were provided by the Istituto Zooprofilattico Sperimentale del Mezzogiorno (IZSM, Italy). The preparations employed in this study consisted of non-infectious, inactivated viral samples obtained from previously available laboratory stocks and used exclusively for spectroscopic characterization. No viral propagation, cell culture experiments, or handling of infectious viral material were performed during the present investigation.
2.2 Sample cleaning and preparation
ITO glass substrates were cleaned by ultrasonic treatment for 2 h in a detergent/deionized water solution (1:1). The substrates were subsequently rinsed with ultrapure water, acetone, and isopropanol, and then dried under a nitrogen (N2) flow.
A 5.2% w/w ZEP520A resist solution was spin-coated at 7,000 rpm for 60 s, producing an approximately 80 nm thick polymer film suitable for Electron Beam Lithography (EBL).
2.3 Electron beam lithography
The nanostructured patterns were fabricated using a Raith 150 electron beam lithography system with an exposure dose of 22 μC/cm2 and a beam current of 12.8 pA. Y-shaped nanocavity units were arranged in a hexagonal lattice with periodicities Λ1 = 350 nm and Λ2 = 347.3 nm (Figure 1b,c).
FIGURE 1
After exposure, the resist was developed in n-amyl acetate and subsequently rinsed in a 1:3 Methyl Isobutyl Ketone (MIBK) and Isopropyl Alcohol (IPA) solution for 60 s, followed by a final rinse in IPA for 30 s.
2.4 Gold deposition and morphological characterization
A 50 nm thick gold layer was deposited by electron-beam evaporation using a SISTEC CL-400C system. The unit cell geometry and material composition of the plasmonic structure are illustrated in Figure 1a. Scanning electron microscopy (SEM), performed using the Raith 150 system, confirmed the accurate reproduction of the designed geometry and the uniformity of the nanocavity structures (Figure 1b).
Surface plasmon resonance (SPR) measurements were performed in transmission mode using a white halogen light source, a ×40 objective (NA = 0.65), and an Ocean Optics USB4000 spectrometer. The resonance wavelength (λSPR) was monitored as a function of the surrounding refractive index. The optical setup used for the measurements is shown in Figure 1e.
2.5 SERS measurements
SERS measurements were performed using a custom-built automated Raman platform integrating a high-sensitivity Raman detector with an Olympus B51 optical microscope. The system employed a collimated 785 nm excitation laser (Laser-785 nm-LAB-FC, Ocean Insight) with a beam diameter of approximately 1 mm to illuminate the samples. The scattered Raman signal was collected through the microscope optical path and analyzed using an Enhanced QEPro Raman spectrometer (Ocean Insight) equipped with a 10 μm slit, operating in the spectral range of 200–2,500 cm-1.
Automated sample scanning was enabled by a motorized stage driven by two precision linear actuators (M-403, PI Instruments) controlled through C-863 controllers (PI Instruments), enabling accurate XY positioning during mapping measurements as shown in Figure 2a. Data acquisition and stage control were managed through custom MATLAB scripts. The laser power was maintained at 20 mW with an integration time of 10 s per spectrum and a single spectral average. The sampling interval was set to 1 s ().
FIGURE 2
2.6 Numerical simulation
Numerical simulations of the Y-shaped nanocavity transmittance spectra (Figure 1f) were performed using a full three-dimensional Rigorous Coupled-Wave Analysis (RCWA) based on Fourier modal expansion (41 k-vectors), implemented in the Lumerical 2024 R1.2 software environment. In the RCWA model, the periodic structure was divided into a stack of uniform layers along the propagation direction (z-axis). (; ).
To investigate the electromagnetic mechanism responsible for the SERS enhancement, the electromagnetic enhancement factor (EF) and the corresponding hotspot distribution (Figure 1d) were calculated using the Finite-Difference Time-Domain (FDTD) method implemented in the Lumerical FDTD solver. FDTD simulations allow accurate numerical solutions of Maxwell’s equations in complex plasmonic nanostructures, providing the spatial distribution of the local electric field (Eloc). The SERS enhancement factor was estimated using the commonly adopted fourth-power approximation: EF ≈ |Eloc/E0|4 which accounts for the plasmonic amplification of both the excitation field and the Raman-scattered emission.
2.7 Virus fingerprint measurements
For SERS fingerprint acquisition, droplets of virus solution with a concentration of 105 Tissue Culture Infectious Dose 50% (TCID50)/mL were directly deposited onto the plasmonic sensors. After deposition, the substrates were incubated for 1 h in a humid chamber at room temperature.
The samples were then rinsed seven times with 1 mL of bidistilled water and dried under nitrogen flow before SERS spectral acquisition. For reproducibility assessment, four independently fabricated plasmonic metasurfaces were employed in the study. Two sensors were functionalized with Human Norovirus (HNoV) samples and two sensors with Murine Norovirus (MNV) samples. From each sensor, 25 SERS spectra were acquired at different spatial locations across the active sensing area, resulting in a total dataset of 100 spectra (50 HNoV and 50 MNV spectra). Measurements were performed during separate experimental sessions and by different operators to evaluate the robustness of the sensing platform against temporal and operator-dependent variability.
2.8 Spectral data analysis
Spectral data were analyzed using R version 2024.04.2 + 764. Prior to multivariate analysis, all spectra were subjected to baseline correction and spectral normalization to minimize background contributions and enable reliable comparison between measurements acquired from different sensors and experimental sessions. Peak assignment, spectral processing, and multivariate analysis were performed within the R environment. Principal Component Analysis (PCA) was used to reduce the dimensionality of the spectral dataset while preserving the main sources of variance. PCA was applied to discriminate between human and murine norovirus spectra. The analysis was carried out using the ChemoSpec package in R (; ; ).
3 Results
3.1 Optical characterization
Scanning electron microscopy (SEM) analysis confirmed the high structural fidelity of the fabricated Y-shaped nanocavity arrays. The patterned area (150 μm × 150 µm) ensures uniform signal collection across the entire sensing region. The nanocavities exhibit a consistent geometry with uniform orientation and periodicity, as shown in Figure 1b.
The plasmonic resonances of the metasurface were experimentally observed near 670 nm, which can be attributed to the coupling between localized surface plasmon modes and propagating plasmon modes within the periodic nanocavity array, as shown in Figure 1f (dark green line). Numerical simulations also predict plasmonic resonances near 670 nm and 742 nm, as illustrated in Figure 1f (orange line).
As shown in Figure 1f, the experimentally observed plasmonic resonance at 670 nm is in good agreement with the simulated spectrum. The numerical model predicts two resonance peaks located at approximately 670 nm and 742 nm. The slight differences between experimental and simulated spectra can be attributed to nanofabrication tolerances. In particular, minor rounding of the Y-shaped cavity vertices during the resist development process can occur compared to the perfectly sharp corners assumed in the RCWA model.
Such geometric rounding is known to modify the effective plasmonic mode volume and the interaction length of localized surface plasmons, which can lead to the appearance of additional resonance features. Furthermore, small variations in the refractive index of the polymeric adhesion layer (ZEP520A) between the theoretical model and the experimentally fabricated structure may also contribute to the observed spectral shifts.
Although the dominant extinction maximum is experimentally observed near 670 nm, the optical response of the Y-shaped metasurface exhibits a broad multimodal behavior extending toward the near-infrared region. In particular, numerical simulations predict an additional plasmonic mode around 742 nm, providing substantial spectral overlap with the 785 nm excitation wavelength employed for SERS measurements. It is important to note that the wavelength associated with maximum far-field extinction does not necessarily coincide with the condition of maximum local electromagnetic enhancement. Since SERS performance is primarily governed by near-field hotspot distributions rather than by the absolute position of the extinction maximum, efficient Raman amplification can still be achieved through excitation of the broad plasmonic modes extending into the near-infrared spectral region. Moreover, the use of a 785 nm excitation source significantly reduces fluorescence background commonly encountered in biological samples, thereby improving spectral quality and signal-to-noise ratio during viral detection.
Overall, the good agreement between experimental and simulated optical responses confirms the successful fabrication of the designed plasmonic metasurface and supports its suitability for SERS-based label-free viral discrimination.
3.2 Electromagnetic enhancement and hotspot distribution
To quantify the plasmonic enhancement capability of the Y-shaped nanocavities, near-field electromagnetic simulations were performed using the Lumerical FDTD platform to determine the spatial distribution of the local electric field (Eloc) at the SERS excitation wavelength of λ = 785 nm. The electromagnetic Surface-Enhanced Raman Scattering (SERS) enhancement factor (EF) was estimated using the widely adopted fourth-power approximation:where E0 represents the amplitude of the incident electric field. This fourth-power dependence accounts for the double electromagnetic enhancement process characteristic of SERS, which involves the amplification of both the excitation field and the Raman-scattered emission. Figure 1d shows the two-dimensional map of the electromagnetic enhancement factor (|E|4) calculated across the top surface of the gold nanocavity within a unit cell of the hexagonal lattice (Λ1 = 350 nm, Λ2 = 347.3 nm) illustrated in Figure 1c.
The map clearly reveals highly localized electric field confinement, with maximum intensities concentrated at the sharp inner vertices of the Y-shaped geometry. These regions, commonly referred to as plasmonic hotspots, represent the locations where the strongest electromagnetic enhancement occurs. Two main physical mechanisms contribute to this enhancement. The first is the lightning rod effect, in which the high curvature at the tips of the Y-shaped arms leads to charge accumulation and strong local electric field amplification at the nanoscale. The second mechanism arises from inter-cavity plasmonic coupling, where electromagnetic interactions between neighboring cavities within the periodic array further intensify the local field distribution.
The simulations indicate a maximum electromagnetic enhancement of approximately EFmax≈ 6.4 × 103. Although this value is not among the highest reported for SERS hotspots, its relevance lies in the uniform spatial distribution of the enhanced electromagnetic field across the periodic metasurface. Higher enhancement factors have been reported for alternative plasmonic architectures, including pyramidal nanohole metasurfaces previously developed by our group (EF ∼1.3 × 106) (). This homogeneous distribution of hotspots across the sensor area significantly reduces spectral variability during measurements, ensuring that the SERS signal is generated consistently across the entire patterned region. Such electromagnetic uniformity plays a critical role in enabling the high spectral reproducibility observed in the experimental measurements.
In particular, this structural and electromagnetic homogeneity supports the tight clustering observed in the subsequent multivariate analysis, where the first two principal components explain more than 98% of the total spectral variance captured.
3.3 SERS performance - fingerprint of murine and human norovirus
The plasmonic metasurface was evaluated for its capability to detect and characterize the SERS fingerprint spectra of the two noroviruses under investigation. Human norovirus (HNoV) and murine norovirus (MNV) were deposited on the same plasmonic substrate at a concentration of 105 TCID50/mL in phosphate-buffered saline (PBS). A total of 100 spectra were collected from four independently fabricated plasmonic metasurfaces. Two sensors were used for Human Norovirus measurements and two sensors for Murine Norovirus measurements. For each sensor, 25 spectra were acquired from different locations across the sensing area. Measurements were performed during separate acquisition sessions and by different operators, providing an initial assessment of substrate-to-substrate, spatial, temporal, and operator-dependent reproducibility. The resulting spectra exhibited high consistency and well-defined vibrational features across the investigated datasets. Figure 2c shows the average SERS spectra obtained for the two viruses in PBS. The dark gray curve corresponds to human norovirus, whereas the red curve represents murine norovirus.
A comparative analysis of peak positions and relative intensities reveals several similarities as well as distinctive spectral features between the two viruses. Both spectra exhibit several spectral features commonly reported in Raman and SERS studies of viral particles and protein-rich biological systems. These bands have been previously associated with vibrational contributions originating from disulfide bonds, aromatic amino acids, protein backbone modes, and nucleic-acid-related structures. However, it should be noted that multiple biomolecular components may contribute simultaneously to the same spectral region, and therefore the assignments reported in Table 1 should be regarded as putative rather than definitive molecular identifications. The presence of several common spectral features suggests that both viral species share broadly similar biochemical characteristics, which is consistent with their close taxonomic relationship and structural similarity as members of the Caliciviridae family.
TABLE 1
| Wavenumber [cm-1] | Assignment | Human NoV | Murine NoV | Ref. |
|---|---|---|---|---|
| 524 | Disulfide stretching | X | | |
| 541 | S-S stretch | X | X | |
| 593 | Disulfide bridges | X | X | |
| 605 | Phenylalanine | | X | |
| 663 | CH2 rocking, Phenylalanine | X | X | |
| 687 | Glycine | X | X | Yadav et al. (2021) |
| 730 | Adenine | X | | |
| 746 | Adenine | | X | |
| 779 | Histidine | | X | |
| 799 | Tryptophan | X | X | |
| 841 | Tyrosine | | X | |
| 876 | Tryptophan | X | X | |
| 895 | DNA/RNA, phosphodiester, deoxyribose, ν(C-OO−) | X | X | Zhou et al. (2023) |
| 939 | C-COO− | X | | |
| 958 | Tryptophan | | X | |
| 970 | Valine | X | | |
| 1,000 | Phenylalanine (symmetrical ring breathing) | X | X | |
| 1,017 | Phenylalanine (in-plane C-H bending) | X | | |
| 1,026 | Phenylalanine (in-plane C-H bending) | | X | Tripathi et al. (2023) |
| 1,088 | Histidine | X | X | |
| 1,160 | C-C/C-N stretching in proteins | X | | Zhou et al. (2023) |
| 1,178 | δ (C–H) and tyrosine | X | X | Xu et al. (2020) |
| 1,231 | Amide III (β-sheet) | X | X | |
| 1,312 | Amide III (α-helix structure) | | X | |
| 1,378 | Alanine | X | X | |
| 1,454 | CH2 deformation | X | X | |
| 1,473 | Adenine C-N stretching | X | | |
| 1,534 | Amide II in anti-parallel β-sheet | X | X | |
| 1,601 | Tyrosine and phenylalanine | X | X | Tripathi et al. (2022) |
Putative Raman band assignments associated with the principal spectral features observed in the SERS fingerprints of human norovirus and murine norovirus.
Despite this general similarity, some notable differences were observed. The Human NoV spectrum exhibited several spectral features that were either absent or less pronounced in the Murine NoV dataset, whereas the Murine NoV spectra displayed additional features in other spectral regions. Based on previously reported Raman assignments, these differences may reflect variations in protein composition, capsid organization, amino-acid exposure, and nucleic-acid-related vibrational contributions. Nevertheless, due to the complexity of viral Raman spectra, individual bands should not be interpreted as unique biomarkers of a specific molecular component.
The observed spectral differences likely reflect subtle variations in the overall biochemical composition and structural organization of the viral particles. However, given the overlapping nature of Raman signatures in complex biological systems, it is not possible to attribute these differences unambiguously to specific molecular species or isolated biochemical mechanisms. These reproducible spectral differences provide sufficient information for multivariate statistical analysis, enabling reliable discrimination between the two viral species through their overall Raman fingerprints.
3.4 Multivariate statistical analysis (PCA)
Given the significant spectral overlap between the two noroviruses, a multivariate statistical approach was required to identify subtle but systematic differences in their SERS fingerprints. Principal Component Analysis (PCA) was therefore applied to reduce the dimensionality of the spectral dataset while highlighting the most relevant sources of variance. This approach provides a robust method for evaluating the analytical performance of the nanostructured platform and for verifying that the observed spectral differences are statistically meaningful and reproducible.
Figure 2b presents the PCA score plot in a two-dimensional space defined by the first two principal components (PC1 and PC2). The data points correspond to human norovirus (HNoV), represented by dark gray squares, and murine norovirus (MNV), represented by red circles, based on the collected SERS spectra. The score plot reveals two clearly separated clusters corresponding to the two viral strains, demonstrating the strong discrimination capability of the plasmonic metasurface platform. The tight grouping of the data points within each cluster indicates high spectral reproducibility across the measurements, while the clear separation between clusters confirms the sensitivity of the SERS fingerprints to subtle structural differences between the human and murine capsids. Quantitatively, the first principal component (PC1) accounts for 87.1% of the total spectral variance, while the second principal component (PC2) explains an additional 11.6%, resulting in a cumulative explained variance exceeding 98% the discrimination is evidenced by the cluster separation.
Such a high cumulative variance indicates that the vast majority of spectral information is captured by the first two principal components, enabling reliable discrimination between the two norovirus variants. The reproducibility of the observed clustering is further supported by the use of multiple independently fabricated sensors, measurements acquired from different regions of the substrates, and separate acquisition sessions conducted by different operators. The persistence of the discrimination pattern across these experimental conditions indicates that the observed separation is driven primarily by intrinsic spectral differences between the viral species rather than by measurement artifacts or substrate variability.
To further investigate the spectral origin of the observed clustering, PCA loading vectors corresponding to the first two principal components were analyzed and are provided in the Supplementary Information (Supplementary Figure S1). The loading profiles indicate that the discrimination arises from distributed contributions across multiple Raman regions rather than from a single dominant spectral feature. The loading profiles support the conclusion that discrimination originates from distributed spectral contributions across multiple Raman regions rather than from isolated spectral bands. Additional information regarding the spectral variability of the complete dataset is provided in Supplementary Figure S2.
3.5 Discussion
The proposed Y-shaped plasmonic metasurface enabled the label-free discrimination of Human Norovirus (HNoV) and Murine Norovirus (MNV) through their intrinsic SERS fingerprints, despite the close structural similarity between both viral species. The clear separation observed in the PCA space demonstrates that the platform captures reproducible spectral differences associated with the overall biochemical composition of the viral particles, highlighting the potential of combining nanophotonic metasurfaces, Raman fingerprinting, and multivariate analysis for pathogen classification without antibodies, molecular labels, or amplification strategies.
Although the present work was not designed to determine the analytical limit of detection, the investigated concentration (105 TCID50/mL) lies within the range commonly associated with active norovirus infections and therefore represents a clinically relevant benchmark for evaluating the discrimination capability of the platform (; TEUNIS et al., 2015). Within this context, the study should be regarded as a proof-of-concept demonstration of label-free viral fingerprint recognition rather than a fully optimized quantitative diagnostic assay.
A key feature of the proposed architecture is the balance between electromagnetic enhancement and spectral reproducibility. While the estimated enhancement factor (EF ≈ 6.4 × 103) is lower than values reported for highly localized plasmonic hotspots, the metasurface was intentionally engineered to provide a homogeneous electromagnetic field distribution over a large sensing area. Such uniformity is particularly advantageous for multivariate fingerprint analysis, where reproducibility and spectral consistency are often more critical than maximizing local enhancement.
The robustness of the sensing strategy is supported by measurements acquired from four independently fabricated plasmonic metasurfaces, multiple spatial locations, separate acquisition sessions, and different operators. Under these conditions, the spectral fingerprints consistently clustered according to viral type, indicating that the observed separation is driven primarily by intrinsic spectral differences rather than measurement variability.
To better contextualize the present results, Table 2 compares representative SERS-based viral sensing approaches reported in the literature. Unlike many highly sensitive assays that rely on antibodies, aptamers, Raman reporters, or molecular amplification, the proposed platform operates in a fully label-free configuration and enables the discrimination of closely related norovirus species through their intrinsic Raman signatures. These characteristics highlight the potential of engineered plasmonic metasurfaces as simplified sensing platforms for rapid pathogen classification.
TABLE 2
| Study | Target virus | Sensing strategy | Label-free | Sample preparation complexity | Reported detection capability | Main advantage |
|---|---|---|---|---|---|---|
| Norovirus | Dual SERS nanotag immunoassay | No | High (antibodies + nanotags) | Highly sensitive detection | Specific immunorecognition | |
| SARS-CoV-2 | SERS detection of S and N proteins | Yes | Low | Viral protein identification | Rapid label-free analysis | |
| Zhou et al. (2023) | SARS-CoV-2 | SERS + PCA | Yes | Low | Virus identification | PCA-assisted discrimination |
| Tripathi et al. (2023) | Zika virus | Ag nanoisland SERS platform | Yes | Low | High-sensitivity viral detection | Portable sensing approach |
| Yadav et al. (2021) | HIV-1 | Ag nanorod SERS substrate | Yes | Moderate | HIV detection and tropism determination | Rapid viral characterization |
| Live SARS-CoV-2 | SERS-active substrates | Yes | Low | Variant differentiation | Live-virus analysis | |
| Shiga toxins | Plasmonic metasurface + PCA | Yes | Low | Toxin discrimination | High classification capability | |
| Shiga toxin variants in serum | SERS + PCA | Yes | Moderate | Variant discrimination in biological matrix | Clinically relevant samples | |
| This work | Human and murine norovirus | Y-shaped plasmonic metasurface + PCA | Yes | Low | Discrimination of closely related viral species | Label-free viral fingerprinting without antibodies or molecular labels |
Representative examples of previously reported SERS-based viral sensing platforms reported in the literature, highlighting sensing strategy, label-free capability, analytical complexity, detection capability, and the principal advantages of each approach relative to the proposed Y-shaped plasmonic metasurface.
Future studies will focus on evaluating sensor performance in clinically and environmentally relevant matrices, assessing potential interference effects, and expanding the statistical framework through larger spectral datasets and supervised machine-learning approaches. Previous studies from our group have demonstrated that algorithms such as LDA, SVM, and Random Forest can significantly enhance classification performance when combined with SERS fingerprinting (; ; ; ; ), suggesting a promising pathway toward more advanced diagnostic implementations.
4 Conclusion
In this work, we demonstrated a plasmonic metasurface platform based on Y-shaped nanocavity arrays for the label-free detection of norovirus using surface-enhanced Raman spectroscopy (SERS). The engineered nanostructure supports plasmonic resonances that generate localized electromagnetic hotspots, enabling efficient Raman signal amplification under 785 nm excitation.
Distinct SERS fingerprints were obtained for human norovirus (HNoV) and murine norovirus (MNV), despite their close structural similarity. Multivariate statistical analysis using principal component analysis (PCA) revealed clear spectral separation between the two viral strains, confirming the capability of the platform to discriminate closely related viral capsids without the need for antibodies or molecular labels.
These results highlight the potential of nanostructured plasmonic metasurfaces combined with SERS fingerprinting and statistical analysis as a rapid and reliable approach for viral identification. The proposed strategy represents a promising proof-of-concept toward next-generation plasmonic biosensors for pathogen identification. Future studies will focus on analytical sensitivity, validation in clinically relevant biological matrices, and implementation of advanced classification algorithms to further evaluate the translational potential of the platform.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
BG: Validation, Software, Formal Analysis, Writing – review and editing, Methodology, Writing – original draft, Data curation, Visualization, Conceptualization, Supervision, Investigation. AM: Investigation, Data curation, Methodology, Software, Writing – review and editing, Validation, Formal Analysis, Writing – original draft, Conceptualization, AD’: Investigation, Methodology, Writing – review and editing. DS: Writing – review and editing, Investigation. MR: Investigation, Writing – review and editing. VM: Investigation, Writing – review and editing. GF: Writing – review and editing, Validation, Investigation. LP: Validation, Formal Analysis, Supervision, Project administration, Writing – review and editing, Conceptualization, Investigation, Funding acquisition, Resources.
Funding
The author(s) declared that financial support was received for this work and/or its publication. The authors gratefully acknowledge the support for this work from European Union-NextGenerationEU Call PRIN2022 Development of a plasmonic nanobiosensor for the rapid diagnosis of Shiga toxin producing E. coli human infections at the point of care-SENSOSTEC (CUP B53D23019990006).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fphot.2026.1864447/full#supplementary-material
References
1
AchaduO. J.AbeF.SuzukiT.ParkE. Y. (2020). Molybdenum trioxide nanocubes aligned on a graphene oxide substrate for the detection of norovirus by surface-enhanced raman scattering. ACS Appl. Mater Interfaces12, 43522–43534. 10.1021/acsami.0c14729
2
AoY.XieX.DongX.JinM.DuanZ. (2019). Genetic analysis of an emerging GII.P2–GII.2 norovirus associated with a 2016 outbreak of acute gastroenteritis in China. Virol. Sin.34, 111–114. 10.1007/s12250-019-00084-6
3
AtmarR. L.OpekunA. R.GilgerM. A.EstesM. K.CrawfordS. E.NeillF. H.et al (2008). Norwalk virus shedding after experimental human infection. Emerg. Infect. Dis.14, 1553–1557. 10.3201/eid1410.080117
4
CaligiuriV.TedeschiG.PaleiM.MiscuglioM.Martin-GarciaB.Guzman-PuyolS.et al (2020). Biodegradable and insoluble cellulose photonic crystals and metasurfaces. ACS Nano14, 9502–9511. 10.1021/acsnano.0c03224
5
CetinA. E.EtezadiD.GalarretaB. C.BussonM. P.EksiogluY.AltugH. (2015). Plasmonic nanohole arrays on a robust hybrid substrate for highly sensitive label-free biosensing. ACS Photonics2, 1167–1174. 10.1021/acsphotonics.5b00242
6
ChenS.LiG.ZeunerF.WongW. H.PunE. Y. B.ZentgrafT.et al (2014). Symmetry-selective third-harmonic generation from plasmonic metacrystals. Phys. Rev. Lett.113, 033901. 10.1103/PhysRevLett.113.033901
7
ChenD.ZhouJ.RippaM.PettiL. (2015). Structure-dependent localized surface plasmon resonance characteristics and surface enhanced raman scattering performances of quasi-periodic nanoarrays: measurements and analysis. J. Appl. Phys.118, 163101. 10.1063/1.4934490
8
D’AvinoA.MilanoA.MarchesanoV.GuilcapiB.SagnelliD.RippaM.et al (2025). Flexible gold nanoparticle SERS tape for rapid, label-free and ultrasensitive detection and differentiation of Shiga toxin variants. Biosens. Bioelectron.27, 100696. 10.1016/j.biosx.2025.100696
9
D’AvinoA.MilanoA.MarchesanoV.GuilcapiB.SagnelliD.RippaM.et al (2026). “Stick-and-Sense”: a flexible paper SERS tape for rapid, selective and high – sensitivity food and water pathogen detection. Food Chem., 149599. 10.1016/j.foodchem.2026.149599
10
FangY.LinT.ZhengD.ZhuY.WangL.FuY.et al (2021). Rapid and label‐free identification of different cancer types based on surface‐enhanced raman scattering profiles and multivariate statistical analysis. J. Cell Biochem.122, 277–289. 10.1002/jcb.29857
11
FazioB.D’AndreaC.FotiA.MessinaE.IrreraA.DonatoM. G.et al (2016). SERS detection of biomolecules at physiological pH via aggregation of gold nanorods mediated by optical forces and plasmonic heating. Sci. Rep.6, 26952. 10.1038/srep26952
12
GuilcapiB.MilanoA.AvinoA. D.SagnelliD.RippaM.MarchesanoV.et al (2026). Machine learning-enhanced SERS on silicon-gold sensors for ultra-high-throughput pathogen detection. Medinformatics3, 162–177. 10.47852/bonviewMEDIN62028939
13
JohnS. (1987). Strong localization of photons in certain disordered dielectric superlattices. Phys. Rev. Lett.58, 2486–2489. 10.1103/PhysRevLett.58.2486
14
KannanP.LosM.LosJ. M.Niedziolka-JonssonJ. (2014). T7 bacteriophage induced changes of gold nanoparticle morphology: biopolymer capped gold nanoparticles as versatile probes for sensitive plasmonic biosensors. Analyst139, 3563–3571. 10.1039/C3AN02272B
15
KolkowskiR.PettiL.RippaM.LafargueC.ZyssJ. (2015). Octupolar plasmonic meta-molecules for nonlinear chiral watermarking at subwavelength scale. ACS Photonics2, 899–906. 10.1021/acsphotonics.5b00090
16
LeeJ.-H.KimB.-C.OhB.-K.ChoiJ.-W. (2013). Highly sensitive localized surface plasmon resonance immunosensor for label-free detection of HIV-1. Nanomedicine9, 1018–1026. 10.1016/j.nano.2013.03.005
17
LinC.LiuZ.FangF.ZhaoS.LiY.XuM.et al (2023). Next-generation rapid and ultrasensitive lateral flow immunoassay for detection of SARS-CoV-2 variants. ACS Sens.8, 3733–3743. 10.1021/acssensors.3c01019
18
LiuL.MooreM. D. (2020). A survey of analytical techniques for noroviruses. Foods9, 318. 10.3390/foods9030318
19
MatarazzoV.De NicolaS.ZitoG.MormileP.RippaM.AbbateG.et al (2011). Spectral characterization of two-dimensional thue–morse quasicrystals realized with high resolution lithography. J. Opt.13, 015602. 10.1088/2040-8978/13/1/015602
20
MilanoA.D’AvinoA.MarchesanoV.SagnelliD.RippaM.GuilcapiB.et al (2025). Advancing medical diagnostics: rapid, label-free detection and differentiation of Shiga toxin variants in human serum using a cost-effective PCA-assisted SERS platform. ACS Appl. Mater Interfaces17, 63237–63252. 10.1021/acsami.5c18171
21
MilanoA.D’AvinoA.MarchesanoV.PerrottaI. S.GuilcapiB.SagnelliD.et al (2026). SERS and supervised learning for proof-of-concept rotavirus detection in wastewater matrices. Biosens. Bioelectron.311, 118880. 10.1016/j.bios.2026.118880
22
NairR. V.VijayaR. (2010). Photonic crystal sensors: an overview. Prog. Quantum Electron34, 89–134. 10.1016/j.pquantelec.2010.01.001
23
NucaraL.GrecoF.MattoliV. (2015). Electrically responsive photonic crystals: a review. J. Mater Chem. C Mater3, 8449–8467. 10.1039/C5TC00773A
24
PalermoG.RippaM.ContiY.VestriA.CastagnaR.FuscoG.et al (2021). Plasmonic metasurfaces based on pyramidal nanoholes for high-efficiency SERS biosensing. ACS Appl. Mater Interfaces13, 43715–43725. 10.1021/acsami.1c12525
25
PalermoG.GuilcapiB.KolkowskiR.GuglielmelliA.AcetiD. M.ValenteL.et al (2026). Plasmonic metasurfaces with structural chirality and pseudo-chirality for enhanced circular dichroism and enantiomeric recognition. Adv. Opt. Mater14, e02488. 10.1002/adom.202502488
26
PatelM. M.HallA. J.VinjéJ.ParasharU. D. (2009). Noroviruses: a comprehensive review. J. Clin. Virology44, 1–8. 10.1016/j.jcv.2008.10.009
27
PettiL.RippaM.CapassoR.NennaG.MauroA. D. G. D.MaglioneM. G.et al (2013). Novel organic LED structures based on a highly conductive polymeric photonic crystal electrode. Nanotechnology24, 315206. 10.1088/0957-4484/24/31/315206
28
RippaM.CastagnaR.PannicoM.MustoP.BobeicoE.ZhouJ.et al (2017). Plasmonic Nanocavities-based aperiodic crystal for protein-protein recognition SERS sensors. Opt. Data Process. Storage3, 3. 10.1515/odps-2017-0007
29
RippaM.MilanoA.MarchesanoV.SagnelliD.GuilcapiB.D’AvinoA.et al (2025). Diagnostic oriented discrimination of different Shiga toxins via PCA-Assisted SERS-Based plasmonic metasurface. Nanophotonics14, 4005–4018. 10.1515/nanoph-2024-0696
30
SalomonA.PriorY.FedorukM.FeldmannJ.KolkowskiR.ZyssJ. (2014). Plasmonic coupling between metallic nanocavities. J. Opt.16, 114012. 10.1088/2040-8978/16/11/114012
31
SanchezJ. E.JaramilloS. A.SettlesE.Velazquez SalazarJ. J.LehrA.GonzalezJ.et al (2021). Detection of SARS-CoV-2 and its S and N proteins using surface enhanced raman spectroscopy. RSC Adv.11, 25788–25794. 10.1039/D1RA03481B
32
SchiattarellaC.SanitàG.Guilcapi AlulemaB.LanzioV.CabriniS.LambertiA.et al (2022). High-Q photonic aptasensor based on avoided crossing bound states in the continuum and trace detection of ochratoxin A. Biosens. Bioelectron.12, 12. 10.1016/j.biosx.2022.100262
33
ShechtmanD.BlechI.GratiasD.CahnJ. W. (1984). Metallic phase with long-range orientational order and no translational symmetry. Phys. Rev. Lett.53, 1951–1953. 10.1103/PhysRevLett.53.1951
34
ShuL.ZhouJ.YuanX.PettiL.ChenJ.JiaZ.et al (2014). Highly sensitive immunoassay based on SERS using nano-Au immune probes and a nano-Ag immune substrate. Talanta123, 161–168. 10.1016/j.talanta.2014.02.015
35
SitjarJ.LiaoJ.-D.LeeH.TsaiH.-P.WangJ.-R.LiuP.-Y. (2021). Challenges of SERS technology as a non-nucleic acid or -antigen detection method for SARS-CoV-2 virus and its variants. Biosens. Bioelectron.181, 113153. 10.1016/j.bios.2021.113153
36
SitjarJ.LiaoJ.-D.LeeH.TsaiH.-P.WangJ.-R.ChenC.-H.et al (2023). Detection of live SARS-CoV-2 virus and its variants by specially designed SERS-Active substrates and spectroscopic analyses. Anal. Chim. Acta1256, 341151. 10.1016/j.aca.2023.341151
37
TawilN.SacherE.MandevilleR.MeunierM. (2012). Surface plasmon resonance detection of E. coli and methicillin-resistant S. aureus using bacteriophages. Biosens. Bioelectron.37, 24–29. 10.1016/j.bios.2012.04.048
38
TeunisP. F. M.SukhrieF. H. A.VennemaH.BogermanJ.BeersmaM. F. C.KoopmansM. P. G. (2015). Shedding of norovirus in symptomatic and asymptomatic infections. Epidemiol. Infect.143, 1710–1717. 10.1017/S095026881400274X
39
TripathiM. N.SinghK.YadavU.SrivastavaR. R.GangwarM.NathG.et al (2022). SERS based rapid and ultrasensitive detection of Japanese encephalitis virus. Antivir. Res.205, 105382. 10.1016/j.antiviral.2022.105382
40
TripathiM. N.JangirP.AakritiR. S.GangwarM.NathG.SaxenaP. S.et al (2023). A novel approach for rapid and sensitive detection of zika virus utilizing silver nanoislands as SERS platform. Spectrochim. Acta A Mol. Biomol. Spectrosc.302, 123045. 10.1016/j.saa.2023.123045
41
VinjéJ. (2015). Advances in laboratory methods for detection and typing of norovirus. J. Clin. Microbiol.53, 373–381. 10.1128/JCM.01535-14
42
VitielloM. S. (2014). Terahertz Photonic Devices, 91–111. 10.1007/978-94-017-8828-1_5
43
WangN.PanG.LiuP.RongS.GaoZ.LiQ. (2021). Advances and future perspective on detection technology of human norovirus. Pathogens10, 1383. 10.3390/pathogens10111383
44
XuY.WangY.LinH.LiuX.ZhengZ.WangT.et al (2020). Serum analysis method combining cellulose acetate membrane purification with surface‐enhanced raman spectroscopy for non‐invasive HBV screening. IET Nanobiotechnol14, 98–104. 10.1049/iet-nbt.2019.0274
45
YablonovitchE. (1987). Inhibited spontaneous emission in solid-state physics and electronics. Phys. Rev. Lett.58, 2059–2062. 10.1103/PhysRevLett.58.2059
46
YadavS.SenapatiS.DesaiD.GahlautS.KulkarniS.SinghJ. P. (2021). Portable and sensitive Ag nanorods based SERS platform for rapid HIV-1 detection and tropism determination. Colloids Surf. B Biointerfaces198, 111477. 10.1016/j.colsurfb.2020.111477
47
YuN.CapassoF. (2014). Flat optics with designer metasurfaces. Nat. Mater13, 139–150. 10.1038/nmat3839
48
YuN.GenevetP.KatsM. A.AietaF.TetienneJ.-P.CapassoF.et al (2011). Light propagation with phase discontinuities: generalized laws of reflection and refraction. Sci.334, 333–337. 10.1126/science.1210713
49
ZhouL.VestriA.MarchesanoV.RippaM.SagnelliD.PicazioG.et al (2023). The label-free detection and identification of SARS-CoV-2 using surface-enhanced raman spectroscopy and principal component analysis. Biosens. (Basel)13, 1014. 10.3390/bios13121014
50
ZitoG.SanitàG.Guilcapi AlulemaB.Lara YépezS. N.LanzioV.RiminucciF.et al (2021). Label-free DNA biosensing by topological light confinement. Nanophotonics10, 4279–4287. 10.1515/nanoph-2021-0396
Summary
Keywords
label-free biosensing, nanophotonics, norovirus detection, plasmonic metasurfaces, principal component analysis (PCA), surface-enhanced Raman spectroscopy (SERS)
Citation
Guilcapi B, Milano A, D’Avino A, Sagnelli D, Rippa M, Marchesano V, Fusco G and Petti L (2026) Plasmonic metasurface-enhanced Raman spectroscopy for label-free discrimination of human and murine noroviruses. Front. Photonics 7:1864447. doi: 10.3389/fphot.2026.1864447
Received
24 April 2026
Revised
10 June 2026
Accepted
17 June 2026
Published
08 July 2026
Volume
7 - 2026
Edited by
Eden Morales-Narváez, Universidad Nacional Autónoma de México, Mexico
Reviewed by
Hulya Yilmaz, Leibniz Institute of Photonic Technology (IPHT), Germany
Harshala Naik, The Institute of Science Mumbai, India
Updates
Copyright
© 2026 Guilcapi, Milano, D’Avino, Sagnelli, Rippa, Marchesano, Fusco and Petti.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Bryan Guilcapi, bryan.guilcapi@isasi.cnr.it; Alessia Milano, alessia.milano@isasi.cnr.it; Lucia Petti, lucia.petti@isasi.cnr.it
† These authors have contributed equally to this work and share first authorship
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.