Nanoparticles as SERS substrates: recent progress in hormone sensing

1 Introduction

The endocrine system regulates a variety of physiological activities, including growth, metabolism, reproduction, and homeostasis, by secreting hormones into the circulation.

Being key biochemical messengers, hormones are active at extremely low concentrations and serve crucial roles in maintaining the physiological system in an equilibrium [1], [2], [3], [4]. However, hormonal imbalances or abnormal secretions are the characteristics of various clinical illnesses (Table 1), including diabetes mellitus [5], thyroid dysfunction [6], infertility, and polycystic ovarian syndrome (PCOS) [7] etc. Thus, the sensitive, specific, and real-time detection of hormones is thus key for early diagnosis [1], 8], 9] treatment monitoring [10], 11], and personalized therapy [12]. Together with their ultra-low concentrations, the structural diversity of hormones ranging from hydrophilic peptides/proteins [Insulin, procalcitonin (PCT), somatostatin etc.] [8], 13], 14] to lipophilic steroids (cortisol, P4, testosterone) [15], 16] and amino acid derived hormones (DA, epinephrine (EPI), norepinephrine(NE) etc.) [3], 4] poses significant challenges in their sensing. Differences in solubility, stability, and molecular size demand varied detection platforms. This complexity hampers the development of a universal strategy for sensing hormones. Traditional approaches (Table 1) have long been considered the gold standard for hormone detection and quantification. While these procedures are reliable and sensitive, they frequently need tedious sample preparation, costly reagents, and trained people, and are unsuitable for POC settings [17], [18], [19], [20].

In 1974, Fleischmann and coworkers noticed an enhanced Raman intensity of pyridine adsorbed on electrochemically roughened Ag electrodes [58]. This finding was separately confirmed in 1977 by Van Duyne’s group, who ascribed it to increased static electromagnetic (EM) fields, and by Creighton’s group, who explained it in terms of resonant Raman scattering caused by the widening of molecule electronic energy levels [59], 60]. This effect was later termed SERS and is characterized by exceptionally high sensitivity, extending down to the single-molecule detection limit. In recent years SERS has emerged as a powerful biosensing analytical platform, offering ultra-high sensitivity, specificity, and label-free, non-destructive analysis in complicated biological systems [61], 62]. SERS enhances the fundamentally weak Raman scattering signatures by several orders of magnitude by electromagnetic (EM) and chemical (CM) enhancement mechanisms [8], 13], 14]. SERS has gained popularity for diagnostic applications due to its unique ability to detect molecules at trace levels [62], [63], [64].

SERS has rapidly progressed as a powerful tool for hormone sensing by leveraging advances in plasmonic and non-plasmonic nanomaterials and molecular recognition strategies. Early studies primarily focused on AgNPs and AuNPs for highly sensitive detection and structural characterization of peptide hormones such as insulin at nM to Pm concentrations [48], 65], 66]. More recent developments emphasize selectivity and real-world applicability through Ab, aptamer, and MIP functionalized SERS substrates, allowing reliable detection of small molecule hormones in complex biofluids including saliva, sweat, and serum [2], 15], 67], 68]. The integration of flexible, wearable, and microfluidic SERS platforms further highlights the transition of hormone sensing from laboratory demonstrations toward point-of-care (POC) and continuous health monitoring applications [44], 45]. The use of SERS shows great promise hormone sensing, where target analytes are frequently present at femtomolar (fM) to pM levels [2], 63], 64]. Insulin, cortisol, estrogen, testosterone, thyroid hormones (T3, T4), and catecholamines like dopamine (DA), epinephrine (EPI) and norepinephrine (NE) all have natural Raman vibrational modes that can be used to identify and quantify them. The schematic (Figure 1) illustrates the overview of SERS for hormone sensing. Furthermore, SERS’s quick detection capabilities (typically within minutes) and interoperability with portable Raman instruments make it a feasible POC option for hormone sensing [69], [70], [71], [72]. Emerging systems also combine microfluidics [44], 73] and machine learning (ML) [48], 74] to provide automated, high-throughput, and smart hormone profiling, which is beneficial in clinical endocrinology.

Figure 1:

Schematic illustration of SERS-based hormone detection. Hormones secreted from various glands are detected using SERS active NPs functionalized with specific capture elements and RRM, enabling enhanced Raman signal acquisition for sensitive and selective hormone detection.

Despite these benefits, practical use of SERS in hormone detection remains challenging. The repeatability of SERS signals, substrate batch-to-batch variability, signal instability induced by hotspot heterogeneity, and interference from non-specific binding in complex matrices must all be carefully considered [61], 75], 76]. 3D-ordered SERS substrates [77], 78], the use of, functionalized nanostructures, SERS tags, internal references [26], 79] are all potential solutions to these limitations [62]. In addition, hybrid SERS substrates [2], 68] that combine EM and CM enhancement mechanisms are being developed to improve sensitivity and stability. Thus, by combining nanotechnology and surface chemistry SERS offers a powerful platform for precise, rapid, and sensitive hormone detection, paving the way for improved endocrine diagnostics and therapy monitoring. The current review explores the emerging role of various NPs in SERS-based hormone sensing. It covers recent advances in direct and indirect detection methods, highlighting functionalization for selectivity, biological matrices, and quantitative performance. Key clinical challenges and future direction like integration with wearables, real-time monitoring, and multiplexed hormone detection are also discussed.

2 Detection principal hormones by SERS

SERS signal amplification arises from two complementary mechanisms: EM enhancement [80] and CM enhancement [81]. The EM mechanism originates from LSPR in metallic nanostructures. When these nanostructures are illuminated by laser light, collective oscillations of free electrons generate an intense localized EM field, particularly at sharp features and nanoscale junctions known as hotspots (Figure 2a and b). Hormone molecules positioned within these hotspots experience amplified incidents and scattered fields, leading to a strong increase in Raman intensity without involving any electronic interaction between the molecule and the substrate [4], 61], 82]. Beyond this, an additional CM enhancement contribution arises from charge transfer (CT) interactions between the substrate and the hormone molecules (Figure 2c). Upon adsorption, the electronic energy levels of the hormone couple with the Fermi level of metals or the valence and conduction bands of semiconductors. Under laser excitation, electrons can transfer either from the substrate to the hormone or from the hormone to the substrate, transiently modifying the molecular polarizability. The relaxation of these excited charge carriers back to the ground state produces enhanced Raman scattering that carries the vibrational fingerprint of the hormone [61], 83].

Figure 2:

Fundamental mechanisms of SERS. (a) Illustration of SERS from a molecule adsorbed on AuNPs, showing Stokes and anti-Stokes Raman processes. (b) Schematic depiction of localized surface plasmon resonance generating enhanced EM fields that drive the SERS effect [61], 83]. c) Illustration of SERS from CM mechanism [84].

The SERS enhancement factor (EF) provides a quantitative measure of how effectively a plasmonic substrate amplifies the Raman signal of an analyte relative to conventional Raman conditions [2], 4], 62]. In general, EF is calculated by comparing the Raman intensity of a characteristic vibrational band obtained under SERS conditions with that measured under normal Raman conditions, while properly normalizing for the corresponding analyte concentrations (or number of probed molecules) [2]. A representative, well-resolved Raman peak typically associated with a specific bond vibration of the target molecule is selected as the marker band. The EF is then expressed as EF = (I _SERS/C _SERS) (I _Raman/R _aman). where I _SERSand I _Ramanare the measured intensities of the chosen Raman band under SERS and normal Raman conditions, respectively, and C _SERSand C _Ramanare the corresponding analyte concentrations. This approach assumes identical optical acquisition conditions and enables direct comparison of enhancement performance between different substrates. A higher EF indicates stronger EM and/or CM enhancement provided by the SERS-active material, thereby reflecting its superior capability for sensitive molecular detection. Further, a variety of SERS substrates (Figures 3 and 4) have been reported in the literature for hormone sensing. However, the choice of an appropriate SERS substrate is not arbitrary; it is intrinsically linked to the physicochemical characteristics of the target hormone. Factors such as molecular size, polarity, available functional groups, and the inherent binding affinity toward plasmonic surfaces collectively govern the adsorption behavior and, consequently, the detection performance [4], 61], 82]. Peptide and protein hormones such as insulin require biocompatible substrates that preserve native conformation while providing strong EM enhancement. Adaptive Ag films, citrate-reduced AgNPs, and functionalized AuNPs have been shown to enable soft adsorption and conformationally sensitive detection of insulin at sub monolayer and even aM levels [48], 65], 85]. In contrast, though small steroid hormones such as cortisol, estradiol (E2), progesterone exhibit weak intrinsic affinity for bare metal surfaces especially for AgNPs [82] through polar functional groups such as carbonyl (C=O) and hydroxyl (–OH) moieties, the interaction is reversible and orientation dependent necessitating substrates with molecular recognition elements [15], 85]. Catecholamine hormones such as DA, EPI and NE, benefit from Ag/Au nanostructures with high-density nanogaps, where strong chemisorption through phenolic groups and controlled aggregation dramatically enhance sensitivity and enable detection at nM to pM concentrations [4], 77], 86]. Overall, rationally matching hormone chemistry with substrate composition, morphology, and surface functionalization is key to achieve sensitivity, selectivity, and reliability in SERS-based hormone sensing. Hormone specificity is achieved through surface functionalization using Ab, aptamers, or MIPs, which selectively bind the target hormone and position it within the SERS-active region [53], 74], 87], 88]. This strategy enables selective sensing even in complex biofluids. Recent SERS biosensing studies show that capture strategies control selectivity, interference tolerance, reproducibility, and translational potential. Boronic-acid–functionalized Ag substrates enable interaction-driven capture via cis-diol binding, offering intrinsic selectivity, serum tolerance, reduced hotspot variability through frequency-shift readout, good batch consistency, and low cost [89]. Ab-based immunocapture provides excellent specificity in complex matrices but involves multistep functionalization, higher cost, and variability linked to antibody and nanotag quality [1], 9], 62], 74]. Aptamer-mediated systems offer tunable selectivity and improved interference control through conformational design, albeit with greater fabrication complexity [15], 87]. Magnetic Ab–AuNP platforms can exclude washing steps and enhance reproducibility but demand stricter batch control and higher material costs [49], 54].

Figure 3:

A comprehensive overview of different nanostructures with tailored compositions that have been used in SERS based hormone sensing.

Figure 4:

Classification of the SERS substrate types used for hormone detection.

3 Direct sensing of hormones by SERS

Direct SERS sensing of hormones involves the absorption of hormones directly onto the surface of SERS-active substrates, enabling the acquisition of detailed vibrational spectra without additional labeling [4], 45], 48], 86]. Studies have extensively demonstrated direct detection of insulin, cortisol, DA and other hormones including neurotransmitters (EPI, NE) through direct adsorption on Ag, Au and composite NPs (Table 2), exploring the molecule-surface interactions to achieve both fingerprint identification [14], 48], 65] and quantification [6], 77]. For instance, adaptive silver films (ASF) distinguished insulin isomers at fM levels [65], while DA was detected down to pM concentrations using flexible or nanostructured substrates [63], 64], 78]. Direct sensing enables ultrasensitive, label-free detection of hormones while preserving the molecular specificity of Raman scattering. In direct sensing, hormones are identified through their intrinsic vibrational fingerprint unique bands. Enhancement of these peaks provides insights into hormone structure, substrate interactions, and biomarker relevance in disease diagnosis. Summarizing these spectral signatures aids cross-study comparison, guides selective sensor design, and support clinical translation. Table 3 lists the SERS peak positions of various hormones with their chemical attributions for use in direct clinical hormone analysis.

3.4 Nanocomposites as SERS-active substrates

Traditional noble metal (Ag, Au) SERS substrates have achieved ultrasensitive detection of hormones [106]. However, their high cost severely limits their development for practical applications, especially for clinical applications [106]. Nanocomposites enable synergistic EM and CM enhancement, improved dielectric coupling, and controlled hotspot formation [107]. The composite architecture enhances analyte adsorption, suppresses nanoparticle aggregation, reduces oxidation of Ag, and improves signal reproducibility in complex biological matrices, which is crucial for reliable hormone detection [107]. The excellent selectivity enhancement of nanocomposite SERS substrates enables them to accurately identify target molecules in complex systems, and their outstanding spectral stability and repeatability make them emerging SERS-active nanomaterials [1], 8], 47], 108]. nanocomposite substrates have significantly advanced the capabilities of SERS based hormone sensing by combining the powerful EM field enhancement of noble metals with the chemical stability and functional tunability of semiconductors. This fusion offers superior sensitivity, specificity, and applicability across a range of bioanalytical contexts [47], 67], 108].The performance of these hybrid substrates is primarily driven by EM and CM enhancement. Noble metal NPs, particularly Ag and Au exhibit LSPR, generating intense EM fields or “hot spots” that amplify Raman signals by several orders of magnitude. Semiconductors complement this by facilitating CT between the analyte [83] and the metal substrate [50], [51], [52]. Together, they produce hybrid nanostructures with highly active SERS responses [47], 107], 109] enhancing signal clarity, improving selectivity, and sensitivity. Structurally, several strategies have been employed to maximize these effects. Vertically aligned SiNWs have been used to anchor AgNPs (SiNWs@Ag) in 3D porous architecture (Figure 5), increasing surface area and improving analyte adsorption. This approach proved effective for neurotransmitter (DA) quantification in plasma and serum with detection limit as low as 0.1 nM and the quantification range between 0.1 and 0.5 nM, enabling significant differentiation between healthy and clinical populations in depression diagnostics [77]. A similar configuration (SiNWs@Ag) was used to detect DA and serotonin with aM sensitivity and mixture discrimination, showcasing their real biological fluid applicability [78].

Figure 5:

SiNWs@Ag enabled SERS platform for quantification of dopamine in biological matrices. (A) Fabrication of silver-enriched silicon nanowires (SiNWs@Ag) and their application for SERS-based detection of plasma dopamine (DA) in human blood samples. (B) Raman spectra of DA at concentrations ranging from 10⁻⁷ to 10⁻¹¹ M in ethanol. (C) Calibration curve showing the linear relationship between −log C (DA) and the peak intensity at 1,148 cm⁻¹. (D) Raman spectra of DA (10⁻⁷ to 10⁻¹¹ M) in human serum. (d) Calibration curve for −log C (DA) versus peak intensity at 880 cm⁻¹. (e) Raman spectra of DA (10⁻⁷ M), norepinephrine (NE, 10⁻⁷ M), and a DA–NE mixture (both 10⁻⁷ M) in ethanol; black numbers indicate Raman peaks common to both DA and NE [77].

Further, graphene (Gr) based nanocomposites have also been extensively incorporated due to their excellent electrical, chemical and biocompatible properties [8], 67], 108], large surface area and [108] capacity for uniform nanoparticle dispersion. Graphene is a one-atom-thick carbon material with a hexagonal crystal structure ideal for sensing applications. It has been functionalized with AgNPs and utilized as a SERS substrate [8]. A two-dimensional (2D) graphene (single layer graphene) based configuration combined with Ab functionalized AgNPs, achieving sensitive detection of PCT for early sepsis diagnosis [8]. Further enhancement of SERS performance has been achieved through functional modifications such as the use of mesoporous silica. This material offers controlled pore structures that regulate nanoparticle dispersion and interparticle gaps, improving signal-to-background ratios. A mesoporous silica modified rGO platform demonstrated excellent reproducibility and sensitivity (<0.2 ng/mL) in detecting parathyroid hormone (PTH) relevant to chronic kidney disease [67]. Similarly, a flower-like hybrid composed of graphene oxide and MoS₂, functionalized with a cortisol-specific DNA aptamer, achieved rapid and selective detection in serum samples, reflecting its potential for stress biomarker monitoring [47]. 3D Gr-based substrates, when combined with Gr quantum dots (GQDs) and Ag NPs, have also demonstrated exceptional SERS activity for detecting DA with a LOD of 0.1 nM [8]. Across these studies, hybrid platforms consistently achieved detection limits ranging from fM to low nM concentrations, often with strong reproducibility and minimal sample preparation. Their adaptability to various biomarkers including neurotransmitters (DA, NE) and other hormones, inflammatory proteins (PCT), and metabolic peptides (Insulin) highlights their versatility in biomedical sensing. Importantly, many of these platforms demonstrated compatibility with complex biological fluids such as blood serum or plasma [47], 77], 108] without compromising sensitivity or selectivity.

Although the direct sensing approach provides high sensitivity and molecular specificity, the low abundance of hormones, and the complexity of biological fluids contribute to low signal intensities, making identification challenging and intricate. Furthermore, uneven aggregation of SERS active nanostructures may cause low reproducibility [62], 106]. Therefore, choosing more advanced SERS substrates such as SERS tags with a tailored composition is the key for achieving high sensitivity and reproducibility.

4 Indirect hormone sensing by SERS tags

One of the most significant challenges in hormone sensing is the complexity of biological matrices, where co-existing molecules such as proteins, lipids, nucleic acids and other metabolites can mask or interfere with the Raman signal of the target analyte [62], 76], 110]. But SERS could overcome this issue by employing a unique architecture of nanoprobes commonly known as SERS tag. SERS tags are often constructed with a plasmonic core (Ag or Au NPs) that serves as the signal-enhancing element, a RRM for generating distinct spectral fingerprints, and a specific capture element such as Ab, aptamers, or molecularly imprinted polymers (MIPs) to ensure selective target binding. has broadened the field of hormone sensing by SERS [3], 44], 93]. These customized SERS tags function as ‘three-in-one’ probes, allowing for selective binding, signal production, and multiplexed analysis [79] in a single step. SERS tags overcome the intrinsic limitations of low Raman cross-sections and matrix interferences by generating stable, amplified, and highly specific Raman signals upon interaction with the target hormone molecule even in the complex biological matrix. The RRMs such as 4-mercaptobenzoic acid (4-MBA) [44], 4-mercaptophenylboronic acid (4-MPBA) [3], aminothiophenol (4-ATP) [54], or methylene blue (MB) produces intense fingerprint peaks (Table 4) when positioned at plasmonic “hotspots” of nanostructures like nanostars, nanorods, core–shell dimers, or binary nanosphere arrays. An innovative approach of wearable technology came through the development of a silk fibroin-based opal structure (SFOS) skin patch. The SFOS skin patch was prepared by first forming a monolayer polystyrene (PS) opal structure (∼1 µm spheres) on a silk fibroin film via interfacial self-assembly. A 10 nm chromium adhesion layer and 90 nm gold film were then deposited by magnetron sputtering. Polyvinylpyrrolidone (PVP)-capped silver nanospheres (∼58 nm) were self-assembled onto the Au-coated opal structure to create a binary nanosphere array. This flexible SFOS/Au/Ag NSs substrate was integrated into a multi-layer wearable microfluidic patch for SERS detection invasive cortisol monitoring in sweat (Figure 6), which utilizes a pseudoknot-assisted aptamer that undergoes conformational switching upon cortisol binding, bringing MB closer to the nanosphere array and significantly enhancing the Raman signal of the 4-MBA attached to the binary Ag-NPs array [44]. Another approach involves Ag nanostructures functionalized with 4-MPBA, where DA binding induces a constrained vibrational shift in the Raman reporter’s peak position from 1,075.6 cm⁻¹–1,084.7 cm⁻¹ providing a ratiometric signal readout directly proportional to DA concentration in serum [3]. Zipper-like Ag nanodimers with uniform 1 nm gaps functionalized with DA-specific aptamers and Raman reporter DTNB, the sensor achieved a 10 aM LOD in buffer and 10 fM in serum. Integrated into a microfluidic chip with a 3D neuronal culture unit, it enabled real-time monitoring of drug-stimulated DA release from living neurons highlighting potential for in situ clinical diagnostics and neurological research [93]. DA is the most explored hormone for its detection and quantification by innovative SERS strategies, Lu et al. 2023 proposed a nanoarray of n-type semiconductors with tungsten trioxide (WO₃) and tin oxide (SnO₂) flake-based SERS tag for detecting DA in SH-SY5Y (neuroblastoma) cell lysates [72]. The study used single stranded DNA as aptamer and to which, the MB was attached as RRM. The study demonstrated a significant SERS activity due to the combined effect of charge transfer from WO₃ ⁻SnO₂ and the molecular resonance effect of MB. Further, Ab-conjugated AuNPs based SERS immunoassays have received a lot of interest in recent years due to their high sensitivity. In a more classical immunoassay format, studies have demonstrated the effectiveness of plasmonic NPs as SERS substrates for the sensitive detection of peptide hormones such as recombinant EPO (rEPO) and PCT. Zhang et al., 2023 demonstrated a double-antibody sandwich format. Raman-labeled (XP013) core–shell BNPs (Au@XP013@AgNPs) and streptavidin-coated magnetic beads enable magnetic enrichment and detection of PCT within 15 min. The assay shows excellent linearity across 0–20 ng/mL PCT and a low detection limit of 0.012 ng/mL [9]. Whereas the other study Au-nanorods functionalized with Raman reporters (DTNB) and anti-rEPO antibodies enabled two SERS-based immunoassays, achieving detection limits below 0.1 pg/mL in urine through nanoextraction and sandwich assays [53]. In the former case, XP013 was used as Raman reporter. Further, studies have also demonstrated SERS-based lateral flow immunoassay (LFIA) strategies for ultrasensitive TSH detection using AuNPs with malachite green isothiocyanate (MGITC) as reporter molecule. Because LFIA strips offer notable advantages over conventional SERS-based immunoassays, including rapid analysis, operational simplicity, minimal interference from chromatographic separation, and enhanced long-term stability [111], 112]. Choi et al. (2017) employed a conventional single-flow LFIA format with ∼37 nm AuNPs functionalized with MGITC and anti-TSH antibodies, achieving a limit of detection (LOD) of 0.025 μIU/mL two orders of magnitude more sensitive than naked-eye LFIA – enabling detection in both hyper- and hypothyroidism ranges with strong reproducibility and minimal cross-reactivity, demonstrating clear potential for POC translation due to its 10 min assay time and compatibility with serum samples [113]. In the dual-flow LFIA developed by Kim et al. (2021), two distinct sizes of Raman reporter–labeled AuNPs; 25 nm biotin-BSA–AuNPs and 45 nm avidin–AuNPs – were sequentially delivered to the test zone (Figure 6). The high-affinity biotin–avidin interaction brought the nanoparticles into proximity, while the size disparity promoted dense interparticle coupling. This arrangement generated abundant plasmonic “hot spots,” significantly amplifying the electromagnetic enhancement effect and thereby boosting the SERS signal. Such dual-size, biorthogonal nanoparticle pairing can overcome the sensitivity limitations of single-size systems, enabling reliable detection of very low TSH concentrations critical for early hyperthyroidism diagnosis [114].

Figure 6:

Representative examples of SERS tag–based platforms for hormone detection. (Top left) wearable silk fibroin–based microfluidic SERS patch for noninvasive cortisol monitoring, integrating Ag nanostructures, Raman reporter molecules, and aptamer functionalization [44]. (Top right) schematic illustration of the SERS-based lateral flow immunoassay platform for the highly sensitive detection of TSH [113]. (Bottom) dual-flow lateral flow immunoassay (LFIA) incorporating AuNP-based SERS tags for TSH detection [114].

4.2 Internal/self-referencing standards

Internal/self-referencing standards in SERS are employed to improve quantitative reliability by correcting signal fluctuations arising from uneven EM or chemical enhancement, hotspot heterogeneity, and instrumental variations [117], 118]. An ideal internal standard (IS) should be chemically and spectrally stable, uniformly distributed within or on the SERS substrate, with at least one characteristic band free from analyte interference [117], 119], 120]. Common organic IS molecules include thiol or pyridine-based compounds such as 4-MPBA, 4-Mercaptobenzonitrile (4-MBN) 4,4-DP, and β-mercaptoethylamine [117], 119], which bind strongly to noble metal surfaces and provide reproducible signals. The distinct Raman bands of the embedded IS exhibit minimal spectral overlaps with SERS fingerprint of hormones, enabling straightforward ratiometric calibration without complex multivariate analysis [79]. Because the internal reference and analyte signals arise from spatially distinct plasmonic hotspots, straightforward intensity normalization or peak‐ratio analysis can be employed without the need for complex spectral deconvolution. This strategy markedly enhances the linearity and reproducibility of quantitative SERS measurements, as reflected by higher R² values in linear fitting for hormone detection [79]. Feng et al. 2022 demonstrated the detection and quantification of multiple neurotransmitters (DA, EPI, NE, ST) using colloidal stellate NPs (Figure 7) having 4,4′-dipyridyl (4,4′-DP) as an IS embedded inside the hollow gap of stellate AuNPs, which could quantify multiple hormones in a concentration range (1 µM–10 mM) [79]. The Raman peaks for 4,4′-DP were distinct for all the four neurotransmitters enabling their unambiguous detection and quantification. Furthermore, normalization of the neurotransmitter Raman signals using the 4,4′-DP reference markedly enhanced the linearity of the calibration curves for all four neurotransmitters as shown in Figure 7d.

Figure 7:

SERS-based quantification of neurotransmitters. (a) Schematic illustration of the strategy for SERS detection. (b) SERS spectra of four neurotransmitters and a blank incubated with stellate AuNPs with internal reference 4,4′-DP. (c) SERS spectra of DA with different concentrations (from 1 μM to 10 mM). (d–g) plots of Raman intensity of neurotransmitters versus their concentrations: (d) DA, (e) epinephrine, (f) serotonin, (g) norepinephrine. Bottom plots are extracted directly from the raw data, whereas top plots are results after calibration by reference signals. The error bars show the standard deviation from 10 measurements with different concentrations. Laser wavelength: 785 nm [79].

5 Clinical significance of SERS based hormone sensing

Hormone detection should be quick and precise to diagnose and manage a variety of health issues, including endocrine diseases, mental illnesses and metabolic syndromes. The advent of noble metal NPs, nanocomposites and their combinations as SERS active substrates has opened a new avenue in hormone detection and quantification by enabling label-free, ultrasensitive, POC and real-time molecular analysis in physiologically relevant fluids. Though SERS is not yet clinically approved diagnostic tool, looking at the current advancements and the anticipated future efforts would further strengthen the technique for hormone sensing. For luteinizing hormone (LH), a passive microfluidic immunoassay employing 4-ATP labeled AuNPs achieved rapid detection in serum with a detection limit of 0.036 IU/L, can potentially eliminate the need for a specialized laboratory [54]. In the case of insulin, a series of SERS studies demonstrated its detection at ultralow concentrations in pancreatic islet secretions, serum, and aqueous solutions. Label-free 3D AuNPs sensors quantified insulin secretion dynamics under glucose stimulation with detection reaching up to 35 pM [25], while AgNPs based microdisk electrodes achieved serum insulin detection with limits of 70 pg/mL [1]. The reported detection limits are sufficient to monitor the physiological concentrations of Insulin [121]. Besides, SERS also help in identifying insulin isomer conformational differences [65] and monitored time- and pH-dependent structural changes associated with amyloid aggregation [85], 94] highlighting the application of SERS in drug safety and functional efficacy monitoring. Clinically, this aids in quality control and preventing aggregation-related complications. Cortisol, a key biomarker for stress, has been detected across multiple platforms, including flexible and wearable formats. Skin-mounted sensors using silver triangle nanoplates on PVC/SEBS films demonstrated high sensitivity with detection limits around 5.5 × 10⁻⁸ M [45]. A silk fibroin-based SERS patch combined aptamer-based specificity with pH calibration for dynamic sweat cortisol monitoring [44], while a lateral-flow paper-based sensor enabled detection in saliva with minimal matrix interference. Additionally, molecularly imprinted polymers and Ag@Au nanostructures provided robust cortisol detection in saliva down to 10 pM [2]. Such ultrasensitive detection could enable non-invasive, real-time cortisol monitoring in sweat and saliva, crucial for assessing stress and hypothalamic-pituitary-adrenal (HPA) axis disorders [70]. Wearable, portable and paper-based platforms offer POC convenience and minimal discomfort, enhancing accessibility. Their high sensitivity supports early detection of conditions like Cushing’s, Addison’s disease, and chronic stress [122]. However, flexible/wearable substrates often suffer from mechanical deformation, sample induced corrosion, and long-term signal drift, while nanofabrication inconsistencies lead to hotspot heterogeneity and poor inter-device reproducibility [123]. Their field deployment is further complicated by the lack of robust on-site calibration strategies to correct laser power fluctuations, substrate aging, and matrix effects [44], 45], 73]. Addressing these problems requires well-defined internal IS materials [62], 79], batch-level quality control workflows, and unified data formats with metadata on substrates, acquisition parameters, and signal preprocessing to ensure cross-study comparability and translational reliability [123]. Further, the matrix effects remain key challenge in SERS-based hormone sensing because proteins, salts, metabolites, and pH variations regulate nonspecific adsorption, hotspot accessibility, and background noise across biofluids [49], 54]. In plasma and serum, high protein content leads to signal suppression and cross-reactivity, which has been effectively mitigated using immunomagnetic capture and microfluidic compartmentalization [49], 54]. Saliva presents lower protein interference but variable viscosity and contaminants, necessitating simple pretreatment and standard-addition calibration to ensure reliable recovery, as reported in [2]. Sweat-based wearable SERS sensing is further complicated by low analyte concentrations, high ionic strength, fluctuating pH, and mechanical deformation; microfluidic sweat collection, aptamer or competitive immunoassays, and matrix-matched calibration improve selectivity and recovery [44], 45], 73], 88].

Further, the LOD and linear dynamic ranges reported for SERS-based hormones sensing are not directly comparable across studies because of multiple interdependent factors as demonstrated by Moody et al., who evaluated the detection of seven neurotransmitters on Au and Ag nanoparticles using 532, 633, and 785 nm excitation [4] Aromatic and indole-containing molecules such as melatonin, serotonin, dopamine, norepinephrine, and epinephrine achieved lower LODs (sub-nM to low µM) due to strong π-interactions with metal surfaces, particularly on AuNPs at 785 nm where optimal overlap with Au LSPR and reduced fluorescence background maximized enhancement [124]. In contrast, amino-acid neurotransmitters lacking extended aromatic systems (glutamate and GABA) showed weak enhancement on AuNPs but significantly improved sensitivity on AgNPs at 633 nm, reaching sub-µM to nM LODs through stronger carboxylate and amine coordination with silver [125]. Despite similar average enhancement factors (∼10⁵–10⁶) for Au and Ag colloids, molecule-specific adsorption and orientation led to large practical differences in LODs, underscoring that sensitivity arises from the interplay of surface chemistry, hotspot accessibility, and excitation conditions rather than EF alone [4]. These issues are amplified in complex biofluids, where proteins and salts further reduce effective hotspot occupation and elevate background, resulting in higher practical LODs [77], 82], 126]. Consequently, intensity-based methods become highly sensitive to hotspot heterogeneity. In such case reference-assisted or peak-shift-based approaches improve detection robustness but may report higher nominal LODs and linear dynamic rages [3], 79].

6 Conclusions

The rapid advancements in the field of SERS-active NPs have notably changed the way of hormone sensing. SERS-active NPs have enabled detection of several clinically important hormones at extremely low concentrations, sometimes as low as attomole which in turn facilitated the faster and accurate detection of hormones even in smaller volumes of complex biological matrices such as sweat, saliva, blood, cells and cell lysate. Also, the advancements in the design of plasmonic NPs have not only facilitated developing the sensitive hormone sensors, but also portable, wearable and ready-for real world applications. Wearable patches made of silk or flexible polymers have been used to check cortisol levels during exercise or in stressful situations, providing a look into totally individualized, real-time health monitoring. Most importantly, label-free SERS based sensors not only help in just quantitative detection of the hormones but also help in assessing the structural dynamics of the peptide hormones like insulin offering important insights into stability and therapeutic effectiveness. In neurology and psychiatry, SERS systems have demonstrated the sensitive and selective detection of neurotransmitters (DA, EPI, and NE) with a high specificity, indicating their diagnostic capabilities. The integration of SERS active substrates into POC devices and smart wearables foreshadows the future of in situ hormone monitoring. Further, studies have primarily focused on the SERS-based detection of insulin, cortisol, and various catecholamines; however, substantial opportunities remain to expand SERS investigations to a broader range of hormones, which is essential for enhancing its clinical relevance.