DNA‐Regulated Catalytic‐Plasmonic Colocalization Enables Synergistic Signal Amplification in SERS and Nanozyme‐Based Sensors

ABSTRACT

The critical need for accessible disease monitoring underscores the urgency of developing advanced point‐of‐care testing (POCT). We present a DNA‐regulated catalytic‐plasmonic colocalization‐based synergy coupling mechanism that resolved the spatiotemporal disjunction commonly present in conventional SERS‐catalysis systems through programmed spatial confinement. Rolling circle amplification‐derived DNA nanocages were employed as programmable spatial regulators to precisely position Au@Pt nanozymes within plasmonic hotspots via base pair encoded hybridization, thereby enforcing spatial and temporal consistency between catalytic generation of Raman‐active species and electromagnetic field enhancement. This strategy integrated coordinated interface, pore, and interlayer confinement, enabling cross‐scale signal amplification from molecular to microscale levels. As a result, it yielded exceptional SERS enhancement (an approximate 41‐fold versus controls), sensitivity (10² exosomes µL⁻¹), and reproducibility (6.5% RSD). To translate this mechanism into practical application, a portable dual‐modal detection platform with potential point‐of‐care applicability was developed that preserved catalytic‐plasmonic colocalization during both colorimetric screening and SERS quantification. The device achieved radical miniaturization (95% volume, 91% weight reduction) and cost‐efficiency (90% reduction vs. commercial systems). By coupling programmable nanomaterial design with customizable device engineering, we established a robust paradigm for next‐generation POCT, providing a promising platform for biomedical detection, environmental surveillance, and food safety monitoring.

We developed a DNA‐regulated catalytic‐plasmonic colocalization‐based synergy coupling mechanism that resolved the spatiotemporal disjunction in conventional SERS‐catalysis systems. The proposed strategy is a hierarchical SERS active substrate that employs tripartite spatial confinement to enable tripartite signal amplification. It was further translated into a portable dual‐modal platform for colorimetric screening and SERS quantification in point‐of‐care applications.

1. Introduction

The evolution of point‐of‐care testing (POCT) was propelled by paradigm shifts in medical frameworks, escalating demands for decentralized healthcare solutions, and multifaceted socioeconomic drivers. As a streamlined alternative to conventional laboratory diagnostics, POCT emerged as a pivotal tool across clinical diagnostics [1, 2, 3, 4, 5, 6], environmental surveillance [7, 8, 9, 10], and food safety monitoring [11, 12, 13, 14, 15, 16, 17]. As a POCT‐related methodology, colorimetric analysis stands out for its operational simplicity, rapidity, cost‐effectiveness, and intuitive interpretability. Yet, its inherent limitations, including moderate sensitivity, compromised specificity, and restricted quantitation capacity, hinder reliable detection of low‐abundance analytes [18, 19, 20]. To address these challenges, dual‐modal detection strategies gained prominence as a viable pathway to enhance analytical performance while advancing POCT toward precision medicine. Recent advancements in biosensing witnessed substantial progress in dual‐modal detection technologies, with prominent examples including colorimetric‐fluorescence [21, 22, 23, 24], colorimetric‐electrochemistry [25, 26, 27], and colorimetric‐Raman spectroscopy [28, 29, 30, 31]. Extensive studies confirmed that dual‐modal approaches enhanced analytical reliability through cross‐validation and extended applicability across diverse detection scenarios. Among them, colorimetric‐Raman detection exhibits several distinctive advantages, including universal compatibility across heterogeneous samples, the elimination of complex pretreatment protocols, and a non‐invasive detection capability that preserves sample integrity [32, 33, 34]. These attributes collectively position colorimetric‐Raman systems as a promising solution for POCT, particularly in resource‐limited settings where operational simplicity and analytical robustness were paramount. However, it remains a technical hurdle that achieve synergistic integration of dual detection modalities with analytical robustness.

Dual‐modal detection technologies leverage mutually independent signal transduction mechanisms, enabling cross‐validation of analytical results to significantly improve interference resistance and operational reliability [35, 36, 37, 38]. The key point of this paradigm hinges on developing integrated probes that generate dual‐modal signals. Recent years, remarkable advancements were studied in multifunction probes, particularly metal‐organic frameworks (MOFs) [39, 40, 41, 42, 43] and metal oxides [44, 45, 46, 47, 48], which exploited their inherent advantages of high surface area, tunable porosity, and customizable nanoscale structure. These materials predominantly served as host matrices for immobilizing functional components such as luminescent moieties or catalytic nanoparticles (NPs), thereby synergistically enhancing optical, electronic, and catalytic functionalities. Notable developments included that Niu et al. [18] developed a dual‐modal signal probe, integrating MOF scaffolds with aggregation‐induced emission luminogens and PtNPs to achieve high fluorescence quantum yield and peroxidase‐like activity. Duan et al. [49] demonstrated Fe/Co‐MOF@Pt nanocomposites with triple fluorescence‐colorimetric‐surface‐enhanced Raman scattering (SERS) signal responsiveness. Xu et al. [50] pioneered Fe₃O₄@MoS₂@Pt nanotags for dual‐modal lateral flow immunoassays, enabling simultaneous detection of SARS‐CoV‐2 nucleocapsid and H1N1 antigens through catalytic‐photothermal coupling effect. Despite these present innovations, fundamental limitations still persist, such as MOF/metal oxide synthesis often involving complex protocols, low catalytic site accessibility, uneven growth, and the intrinsic activity ceilings of monometallic nanozymes (e.g., PtNPs). Fortunately, bimetallic nanozymes opened new avenues for dual‐modal detection, and offered potential solutions to these challenges.

The programmable synthesis of bimetallic nanozymes enabled precise modulation of catalytic properties through strategic compositional tuning of host–guest metal pairs and rational engineering of electronic/geometric configurations [51]. These advantages driven the development of diverse bimetallic architectures for biosensing applications, such as Ag@Au [52, 53, 54, 55], Pd@Ru [56, 57], Cu₂O@Au [58, 59], Fe₃O₄@Ag [60], Fe₃O₄@Au [61], Pd@Pt [62, 63, 64]. Among of them, the Au@Pt nanozymes [19, 65, 66] emerged as a paradigm for colorimetric‐Raman dual‐modal detection, synergistically combining the exceptional peroxidase‐like activity of Pt with the stable SERS enhancement of Au. Its dominance arises from three engineered advantages, including facile one‐pot synthesis requiring only reducing agent control, catalytic activity enhancement over monometallic counterparts due to synergistic and enhanced effects, particles synthesized at the nanoscale with great stability and biocompatibility. Recent applications demonstrated the potential of this system. Hong et al. [66] achieved 30 min Salmonella typhimurium detection using phage‐assisted magnetic separation coupled with AuPt‐decorated nanoflowers and smartphone‐based analysis. However, AuPt NPs reliance on nonspecific electrostatic conjugation to nanoflowers led to inconsistent nanozyme loading in different batches and suboptimal catalytic site utilization. To address this, Zhang et al. [67] and Wang et al. [19] implemented target‐specific functionalization through aptamer/antibody modifications, and further improved detection accuracy and probe efficiency. Jiang et al. [68] integrated loop‐mediated isothermal amplification‐Cas9 cascades for two‐stage signal amplification, and attained attomolar sensitivity. Nevertheless, their sequential workflow risked probe loss during Cas9‐mediated strand displacement, causing uneven Au@Pt distribution that compromised sensitive analysis reliability. In summary, the programmable synthesis of bimetallic nanozymes offers significant advantages for biosensing, particularly in developing colorimetric‐Raman dual‐modal detection systems. Au@Pt nanozymes emerge as a promising candidate due to their catalytic activity, stability, and biocompatibility. However, inadequate signal enhancement and non‐uniform catalytic efficiency are still the limitations that resulted from a lack of spatiotemporal coordination between catalytic and optical amplification.

To address this, we proposed a DNA‐regulated catalytic‐plasmonic colocalization‐based synergy coupling (CPCSC) mechanism. In this design, rolling circle amplification (RCA)‐derived DNA nanocages functioned as programmable spatial regulators that precisely positioned Au@Pt nanozymes into plasmonic hotspots via base pair encoded hybridization. This structure enforced spatial and temporal consistency between the catalytic generation of Raman‐active species and electromagnetic field enhancement. Within this strategy, three confinement modes were deliberately coordinated. Interface confinement (Figure 1A): Au@Pt bimetallic nanozymes promoted electron transfer and molecular adsorption, enabled high‐efficiency catalysis while generating dual‐modal (Raman‐colorimetric) signals for reliability. Pore confinement (Figure 1B): 3D flower‐shaped nanocages, which were synthesized by RCA technology, provided enhanced mass transfer, high‐density 3D hotspots, and abundant specific molecular binding sites. It can improve the probe loading capacity and target affinity. Through base‐pair‐directed assembly, Au@Pt nanozymes were precisely and stably integrated into these nanocages, forming a multifunctional catalytic‐plasmonic colocalization‐based synergy coupling SERS substrate (CPCSCS). Interlayer confinement (Figure 1C): the connection of CPCSCS, target exosomes, and the nanocone array substrate (NAS), which forms a 3D micro‐nano hierarchical structure. This design effectively enhanced photon confinement and light absorption in a spatially restricted area, further amplifying the localized electromagnetic field for Raman signal enhancements, and significantly improved detection sensitivity and anti‐interference capability.

FIGURE 1.

Schematic illustration of the core research concept and the catalytic‐plasmonic colocalization‐based synergy coupling mechanism. Fabrication workflow encompassing (A) programmable synthesis of bimetallic Au@Pt nanozyme, (B) RCA‐driven self‐assembly of 3D flower‐shaped nanocage, and (C) NAS engineering. (D,E) A portable Raman‐colorimetric dual‐modal detection device integrating the active substrate, serving as an analytical platform for actual samples in POCT application scenarios.

Overall, this tripartite spatial confinement (interface, pore, interlayer confinement) enabled cross‐scale signal amplification (molecular, nanoscale, microscale level amplification): (1) electron/mass transfer: Au@Pt nanozymes boosted electron transfer from AuNPs to PtNPs while nanocages enabled efficient mass transfer for TMB catalysis within a confined environment, accelerating molecular level catalytic kinetics; (2) electromagnetic enhancement: nanocages and the NAS created high‐density 3D hotspots, significantly amplifying the local electromagnetic field at the nanoscale level; (3) light trapping: the multiscale micro‐nano hierarchical structure of nanocages and nanocones enhanced light absorption in a confined space, increasing the interaction of matters and photons at the microscale level. To better understand that, we carefully summarized the contributions of each spatial confinement to Raman signal amplification mechanisms, as shown in Figure 2. To the best of our knowledge, this catalytic‐plasmonic colocalization strategy was not been explicitly reported previously. To translate this mechanism into a practical application, we developed a dual‐modal point‐of‐care testing platform that integrated colorimetric screening and SERS quantification on a single solid‐phase substrate (Figure 1D,E). This platform was specifically designed for preserving the spatial colocalization to achieve the CPCSC mechanism. As a result, the proposed method provided a mechanistically well‐defined and sensitive exosome detection strategy with potential applicability in point‐of‐care analytical scenarios.

FIGURE 2.

Schematic illustration of the core research concept and the catalytic‐plasmonic colocalization‐based synergy coupling mechanism.

2. Results and Discussion

2.1. Fabrication and Characterization of CPCSCS

The overall design strategy aimed to enforce spatial colocalization of catalytic centers and plasmonic hotspots, which was essential for validating the proposed CPCSC mechanism. RCA was employed to generate 3D flower‐shaped nanocages with programmable anchor sites, enabling base pair encoded positioning of Au@Pt nanozymes within a confined 3D nanostructure to form CPCSCS (see Sections S1.2–S1.4 for detailed experimental procedures and characterization analyses). As illustrated in Figure 3A, the linear template was annealed with a ligation primer, and the template circularization was achieved through T4 DNA ligase‐mediated phosphodiester bond formation between the 5′‐phosphate and 3′‐hydroxyl termini. With the assistance of phi29 DNA polymerase and deoxyribonucleotide triphosphates (dNTPs), the circular templates were amplified via RCA to obtain long single‐stranded DNA (RCA products). To verify the circularization and amplification ability of the circular template, polyacrylamide gel electrophoresis (PAGE) was performed on the corresponding products obtained from each process (Figure 3D). Section S1.5 described the experimental procedures of PAGE. The reactants added to each lane were detailed in Table S1. Compared with the linear template (lane I), the electrophoretic mobility of the ligation primer was increased (lane II) due to its small molecular weight. A clear band appeared after template‐primer dsDNA anneal (lane III) or circularization of the template by T4 ligase (lane IV), indicating successful synthesis of the circular template. Moreover, a distinct band in lane V was stuck in the gel well and could not migrated into the gel, which indicated that an ultralong DNA with a large molecular weight was formed during the RCA process. These results indicated the successful execution of the RCA reaction. The stability of CPCSCS during 4°C storage (2‐10 days) was also verified by PAGE. As depicted in Figure S1, no discernible band diffusion in the gel well or new bands in the gel were observed, revealing that the RCA products maintained the original structure.

FIGURE 3.

Fabrication and characterization of CPCSCS. (A) Schematic workflow of circular DNA template design, ligation, and RCA reaction mechanism. (B) Assembly mechanism of nanocage self‐assembly during RCA. (C) Schematic illustration of Au@Pt nanozymes precisely organized within the RCA products through Watson‐Crick base pairing. (D) PAGE analysis of the RCA process. Lane I: linear template (including CD63 aptamer), lane II: ligation primer, lane III: annealed template‐primer dsDNA, lane IV: circular template, lane V: RCA products. (E) SEM images of nanocages prepared via RCA with different reaction times. 0–1 h: irregular sheet‐like assemblies, 2–3 h: spherical intermediates, 4 h: flower‐shaped nanocages, 8 h: flower‐shaped nanocages with interparticle coalescence and uncontrolled overgrowth. (F) Energy dispersive spectroscopy (EDS) elemental mapping of CPCSCS prepared by a 4 h RCA reaction. TEM images of nanocages (G) and CPCSCS (H).

Subsequently, as shown in Figure 3B, dNTPs hydrolysis liberated substantial pyrophosphate ions (PPi⁴⁻) during RCA reactions, which immediately complexed with counterions (Mg²⁺) to form magnesium pyrophosphate (Mg₂PPi) precipitates. This precipitation initiated an anisotropic DNA liquid crystallization process, where ultralong ssDNA amplicons self‐assemble into spherical particles through Mg₂PPi‐mediated interfacial interactions [69, 70, 71]. Subsequently, these nanocages undergo progressive structural evolution into well‐defined nanocage structures. By regulating the factors such as reaction time and template sequence, the particle size and functionality of the nanocages could be controlled. To validate the structural evolution of the 3D flower‐shaped nanocages during RCA processing, scanning electron microscope (SEM) analysis was performed at distinct time intervals. Consistent with theoretical predictions, the SEM images revealed a distinct morphological evolution. As shown in Figure 3E, at the early stage of RCA (0‐1 hours), irregular sheet‐like assemblies without distinct 3D architectures were observed, indicating limited accumulation of RCA products. With prolonged amplification (2–3 h), these structures gradually evolved into larger spherical intermediates through continuous self‐assembly of RCA‐generated DNA strands. After 4 h of RCA, uniformly distributed flower‐shaped nanocages (≈1 µm) with well‐defined porous architectures and high monodisperses were formed, suggesting optimized DNA strand growth and structural self‐assembly. In contrast, excessive RCA amplification (8 hours) induced interparticle coalescence and uncontrolled overgrowth, resulting in aggregated structures exceeding 3 µm with poor structural uniformity. These results demonstrate that RCA reaction time critically affects nanocage morphology, with 4 hours identified as the optimal amplification duration for generating monodisperse porous nanocages. Complementary SEM‐based energy dispersive spectroscopy (EDS) elemental mapping confirmed homogeneous distribution of AuNPs within the nanocages, alongside characteristic P, O, C, N, and Mg element signatures from RCA‐derived DNA‐inorganic hybrids (Figure 3F; Figure S2). The aforementioned results demonstrated that the synthesized nanocages exhibit pronounced 3D porous nanostructures, structural stability, and tunable dimensions/functionalities, serving as a robust scaffold for CPCSCS fabrication.

The circular template is an essential component in nanocage fabrication, encoding the sequences to determine its properties. For that, a rationally designed linear template (Figure 3C) served as the foundational blueprint, containing two functional domains: (1) target recognition: the CD63 antisense aptamer sequence (blue) for specific CD63 protein binding; (2) probe anchoring: the complementary sequence (green) for precise nanozymes immobilization. This sequence encoded design ensured that catalytic centers were positioned at predetermined locations rather than randomly adsorbed. As shown in Figure S3A, Au@Pt nanozymes were functionalized via Au─S bonding with an anchor probe (AP), which was complementary to the probes binding region of RCA products (Figure S4). During the RCA‐driven nanocage assembly, these probes became precisely and stably organized within the nanocages through complementary base pairing to form CPCSCS. Transmission electron microscopy (TEM) analysis demonstrated the morphological characteristics of nanocages and CPCSCS, as displayed in Figure 3G,H, where the resulting nanostructures exhibit a well‐defined spherical structure, and the combination of Au@Pt nanozyme exhibited a relatively small influence on the morphology of the nanocage. To verify that the observed colocalization originated from specific base complementary interactions rather than nonspecific adsorption, control experiments employing substitutive AP and probe free conditions were performed (Figure S5). The corresponding TEM images and EDS elemental mappings in Figure S6 revealed a clear spatial overlap of Au and Pt with P/Mg/O/N signals from the 3D flower‐shaped nanocage only when the complementary AP was used. In contract, substitutive or absent probes failed to achieve such nanocages loaded with Au@Pt nanozyme. These results demonstrated that the spatial structure of Au@Pt nanozymes was encoded by nucleic acid hybridization and provided direct structural validation of the proposed colocalization strategy.

2.2. SERS Enhancement and Catalytic Performance Optimization of Au@Pt Nanozymes in CPCSCS

To systematically investigate the plasmonic enhancement mechanism of CPCSCS, comparative studies were conducted using AuNP‐hybridized flower‐shaped nanocages (AHFN) as simplified electromagnetic field enhancement models. Section S1.4 described the experimental procedures and characterization analyses of AHFN. As previously established, Au@Pt nanozymes were immobilized within nanocages via complementary base pairing between AP sequences and the probes binding region of RCA products. As shown in Figure 4A, the 20‐base probe region (6.8 nm) facilitated the generation of 1D hotspots. Transitioning to AHFN architectures, it may induce 3D hotspots formation through two synergistic mechanisms: (1) plasmonic coupling between adjacent AuNPs and (2) electric field intensity within the nanocage. To optimize hotspot density and minimize steric hindrance, four AuNP variants (5/20/40/60 nm diameters) were synthesized and characterized (see SectionsS1.2 and S1.3 for detailed experimental procedures and characterization analyses). TEM analysis (Figure S7) confirmed monodisperse AuNPs with well‐defined morphologies across various sizes. Corresponding UV–vis absorption spectra (Figure S8) exhibited characteristic red shifts from 513 to 531 nm with increasing particle diameters. Figure 4B shows the morphology of the distribution of gold particles of different sizes on the AHFN. SEM characterization of AHFN confirmed preserved structural integrity with NPs spatially ordered within the nanocages. Smaller AuNPs (5/20 nm) displayed interparticle distances of 10–20 nm, whereas larger variants (40/60 nm) achieved sub‐5 nm proximity through geometric confinement effects. To elucidate size‐dependent electromagnetic responses, finite‐difference time‐domain (FDTD) simulations modeled electric field distributions for four AuNP sizes (5/20/40/60 nm) within nanocages. As shown in Figure 4C, simulation parameters replicated experimental configurations: interparticle distances were systematically varied as (i) 10 nm, (ii) 5 nm, (iii) 1 nm, and (iv) overlapping (−5 nm). Figure 4D presented y‐z plane simulation results, revealing maximum local electric field enhancement factors (Max |E _loc/E ₀|) of 2.8, 18.0, 49.7 and 90.4 for 5, 20, 40 nm, and 60 nm AuNPs, respectively. Larger NPs (40–60 nm) exhibited extended electric field over a wider range and higher intensity due to intensified plasmonic coupling between adjacent NPs and the pore‐confinement effect of the nanocage. Notably, despite enhanced coupling in larger NPs, excessive interparticle proximity (<1 nm) induced charge‐exchange phenomena, causing them to behave as equipotential bodies and resulting in electrostatic shielding [72]. It should be noted that the conventional FDTD simulations employed here are based on classical dielectric electrodynamics and therefore cannot accurately describe quantum tunneling and charge‐exchange phenomena at sub‐nanometer interparticle gaps. Consequently, the experimentally observed signal attenuation from 60 nm AuNPs assembled nanocages contradicted the simulation results.

FIGURE 4.

SERS enhancement performance verification and optimization of CPCSCS. (A) Schematic diagram of the transition from 1D to 3D hotspot. (B) SEM images, (C) 2D modal, and (D) FDTD simulations of the electric field distribution for (i) 5 nm, (ii) 20 nm, (iii) 40 nm and (iv) 60 nm AuNPs internalized into the AHFN. (E) Comparative SERS intensity profiles of AuNPs and corresponding AHFN across nanoparticle sizes.

To investigate size‐dependent electromagnetic enhancement mechanisms, AuNPs were functionalized with malachite green isothiocyanate (MGITC) via Au–S bonding (Figure S1B). As shown in Figure S9 and Table S2, MGITC exhibits distinct vibrational fingerprints, with the 1614 cm⁻¹ peak serving as the quantification metric. Figure 4E revealed two critical trends: (1) SERS intensity increased with the increase of AuNP diameter, aligning with established principles of size‐dependent plasmon coupling effect [73, 74, 75]. (2) AHFN incorporating sub‐40 nm AuNPs demonstrated 4‐ to 10‐fold signal amplification versus individual NPs, attributed to nanocage‐induced probe enrichment and nanoconfined plasmonic hotspots generation. However, 60 nm AuNP‐decorated AHFN caused 30% signal attenuation, consistent with the theoretical predictions of interparticle charge‐exchange quenching. These findings collectively established 40 nm NPs as the optimal size for Au@Pt nanozymes fabrication.

Based on the optimized AuNP size (40 nm) obtained from plasmonic enhancement optimization, Au@Pt nanozymes with corresponding nanoscale dimensions were subsequently synthesized to further optimize catalytic activity and catalytic‐plasmonic coupling performance. The catalytic activity originated from the Au@Pt nanozymes, which exhibited peroxidase‐like activity due to interface‐confinement, as schematically illustrated in Figure 5A. Electron transfer from Au to Pt facilitated H₂O₂ decomposition into hydroxyl radicals (·OH), which subsequently oxidize 3,3′,5,5′‐tetramethylbenzidine (TMB) from a reducing state to an oxidizing state (oxTMB), inducing a chromogenic transition from colorless to blue. Under acidic conditions (pH<4.0), oxTMB undergoes further oxidation to a stable quinone derivative (λ _max = 451 nm), yielding a yellow chromogenic product. This cascade was essentially driven by ·OH generation via metal‐catalyzed decomposition of H₂O₂ on Au@Pt nanozymes surfaces, followed by TMB oxidation at adsorbed sites. Therefore, the process involved the adsorption of both H₂O₂ and TMB onto the Au@Pt nanozymes surfaces, followed by a reaction between TMB and ·OH. When TMB is in great excess with respect to the H₂O₂, [TMB] remains constant at its initial value, [TMB]₀. Therefore, this catalytic reaction follows pseudo first‐order kinetic as described by the following rate equations:

$$\left[\mathrm{TMB}\right]={\left[\mathrm{TMB}\right]}_0{\mathrm{e}}^{-k_{\mathrm{obs}}\mathrm{t}}$$ (1)

$$-\ln \frac{\left[\mathrm{TMB}\right]}{{\left[\mathrm{TMB}\right]}_0}=k_{\mathrm{obs}}\mathrm{t}$$ (2)

where k _obs is the apparent pseudo‐first‐order rate constant. Furthermore, according to the Beer‐Lambert law, the absorbance intensity is proportional to the concentration of the chromophore:

$$A=\epsilon cl$$ (3)

where ε is the molar extinction coefficient, c is the concentration, and l is the optical path length. Since the concentration variation of TMB is linearly correlated with the absorbance intensity change during the reaction process, Equation (2) can be further transformed into:

$$-\ln \left(\frac{I}{{\mathrm{I}}_0}\right)=k_{\mathrm{obs}}\mathrm{t}$$ (4)

where I ₀ and I are the absorption intensity of oxTMB at 451 nm before the reaction and at reaction time t, respectively. The observed rate constant k _obs serves as the primary metric for catalytic efficiency comparison [76, 77].

FIGURE 5.

The catalytic performance characterization and optimization of CPCSCS. (A) Schematic of the peroxidase‐like activity mechanism of Au@Pt nanozymes. TEM images (B) and EDS elemental mapping (C) of AuNPs, Au@Pt nanozyme variants (5:1/1:1/1:5 Au:Pt ratios), and PtNPs. (D) UV–vis absorption spectra of five nanoparticle variants under distinct reaction conditions. (TMB concentration: 0.001%–0.25%, H₂O₂ concentration: 0.01%‐3%, buffer pH: 1–11). (E) Absorbance of oxTMB (451 nm) and Raman intensity of MGITC (1614 cm⁻¹) of five nanoparticle variants under optimized parameters.

To prepare the Au@Pt nanozymes that simultaneously exhibited catalytic activity and programmable DNA hybridization capacity, a precise optimization of the Au:Pt molar ratio was essential. Figure S1A illustrated the synthesis scheme involving controlled co‐reduction of HAuCl₄ and H₂PtCl₆ precursors. Three Au@Pt nanozyme variants with three Au:Pt ratios (5:1/1:1/1:5) were synthesized through systematic precursor ratio modulation. TEM characterization (Figure 5B) revealed uniform spherical morphologies (≈40 nm diameter), with Pt deposition density correlating to precursor concentration‐surfaces transitioned from smooth to granular as H₂PtCl₆ proportion increased. EDS elemental mapping (Figure 5C) also confirmed that an increasing amount of platinum elements were localized on the surface of AuNPs. UV–vis absorption spectra (Figure S10) demonstrated progressive absorption peak attenuation around 520 nm with increasing Pt content, accompanied by colloidal color transitions from red to black. This optical evolution originated from the alloy transformed the optical response from selective absorption to broadband absorption by altering the electronic structure, dielectric properties, and particle morphology. This was consistent with the conclusions drawn from the TEM images and UV–vis absorption spectra.

The TMB/H₂O₂ chromogenic system was governed by three critical parameters: the concentration of TMB and H₂O₂, as well as buffer pH. Insufficient reagent concentrations impeded complete oxidation, while excessive levels promoted nonspecific coloration or nanozyme inhibition. Therefore, systematic optimization of TMB concentration (0.001%–0.25%), H₂O₂ concentration (0.01%–3%), and buffer pH (1–11) were conducted. The reactions were uniformly quenched with 2 m H₂SO₄ after 5 min to ensure temporal consistency. The results in Figure 5D were derived from the absorption peak of oxTMB at 451 nm. A gradient of TMB concentrations (0.001%–0.25%) was evaluated and the maximum absorbance was at the concentration of 0.01%. Above this threshold, signal attenuation likely arose from TMB oversaturation and subsequent precipitate formation, as evidenced by colloidal darkening in Figure S11. Suboptimal absorbance below 0.01% resulted from incomplete oxidation due to limited TMB availability. Subsequent optimization of H₂O₂ concentration (0.01%–3%) under fixed 0.01% TMB revealed non‐obvious chromogenic behavior (Figure S12), with maximum absorbance achieved at 0.3% H₂O₂. The pH‐dependent studies using optimized reagents (0.01% TMB, 0.3% H₂O₂) demonstrated spontaneous oxTMB oxidation under strong acidic conditions (pH 1–3), generating false‐positive yellow coloration without the addition of H₂SO₄ (Figure S13). Under physiological to mildly alkaline conditions (pH 5–9), stable oxTMB formation was maintained, and the maximal catalytic efficiency was observed at pH 5. It validated the robust peroxidase‐like activity of Au@Pt nanozymes in slightly acidic environments. Collectively, optimized parameters for the TMB/H₂O₂ system were established as 0.01% TMB, 0.3% H₂O₂ in a pH 5.0 acetate‐citrate buffer. Leveraging these optimized parameters, comparative profiling of five nanoparticle variants was subsequently performed to elucidate the catalytic and plasmonic roles of different compositions. As shown in Figures S14–S16, Au@Pt nanozymes demonstrated significantly enhanced peroxidase‐like activity compared to monometallic Au or Pt counterparts, especially with a 1:5 Au–Pt bimetallic ratio, followed by 5:1. Furthermore, their kinetic analysis with the oxidation of the TMB system was investigated to compare the catalytic efficiency between five nanoparticle variants. The relationship between absorbance ratio and reaction time was shown in Figure S17, where the slope corresponded to the catalytic reaction rate constant (k _obs). The Au@Pt nanozymes (1:5 Au:Pt ratio) exhibited superior catalytic performance with a k _obs was 0.1428 min⁻¹, which was 59.5‐fold and 3.38‐fold enhancement over AuNPs and PtNPs, respectively. Comparatively, the 5:1 Au:Pt variant achieved a k _obs of 0.1244 min⁻¹, corresponding to 51.8‐fold and 2.94‐fold improvements over the monometallic counterparts. These results confirmed that alloying Au with Pt significantly enhanced peroxidase‐like catalytic kinetics. Complementarily, DNA anchoring capacity was validated using MGITC as a DNA substitute. The SERS intensities exhibited a positive correlation with MGITC levels, confirming Au‐S bond‐dependent probe immobilization (Figure S18). As shown in Figure 5E, the dual‐factor comparison histogram (catalytic efficiency vs. DNA loading) identified the Au@Pt nanozymes with an Au–Pt bimetallic ratio (5:1) as the optimal condition. Although the 1:5 variant exhibited the best catalytic activity, the 5:1 composition provided a balance of sufficient catalytic efficiency and enhanced probe anchoring, which was critical for subsequent spatial colocalization and SERS signal generation.

2.3. Fabrication and Characterization of Nanocone Array Substrate and Exosome Recognition

The performance of SERS active substrates mainly depends on NAS. Therefore, we first verified the performance of NAS. Figure S19a–d schematically illustrated the fabrication of NAS through inductively coupled plasma (ICP) anisotropic etching, with corresponding SEM images documenting the structural evolution in Figure 6A‐a–d. Section S1.6 described the experimental procedures for fabrication of NAS. Initially, polystyrene (PS) microsphere monolayers were assembled on polyethylene terephthalate (PET) as an etch mask. Progressive ICP etching induced PET degradation while sculpting microscale conical architectures, ultimately yielding triangular nanocones with nanoscale wrinkle structures (31.8 ± 3.5 nm, Table S3). The height of the nanocone morphology (Figure 6A‐c) was measured via Image J as 885.7 ± 18.9 nm. Subsequent electron‐beam evaporation deposited a 50 nm Au film onto NAS for DNA probe immobilization, as confirmed by homogeneous Au distribution in EDS elemental mapping (Figure S20). Systematic optimization of NAS fabrication was achieved through controlling ICP etching duration (40–250 s), and their corresponding morphological evolution was documented in Figure S21A. Progressive oxygen plasma exposure induced sequential transformations: (1) PS microsphere diameter reduction (40–190 s) through ICP etching, concurrent PET substrate sculpting into nanocone structures; (2) structural transition to sharp nanocone morphology with constricted midsection (190–250 s). It revealed a 6‐fold decrease in nanocone midsection diameter from 411 ± 15.4 nm (40 s) to 69 ± 12.7 nm (250 s). It demonstrated the precise nanofeature control capability through temporal modulation of plasma etching duration (Table S4). Systematic investigation of etching duration effects on SERS performance revealed critical structure‐activity relationships (Figure S21B). NAS substrates with 160 s ICP etching demonstrated maximum Raman enhancement, attributed to synergistic effects of nanocones and nanoscale wrinkle structures. Prolonged etching (>160 seconds) induced progressive signal attenuation (≈60% intensity loss at 250 s), attributed to morphological transitions and concomitant wrinkle degradation. These findings suggested 160 s as the optimal etching duration. To comprehensively assess NAS performance, MGITC‐functionalized substrates were subjected to SERS mapping across a concentration gradient (10⁻⁶–10⁻¹⁰ m, Figure S22). The corresponding averaged spectra analysis (Figure S23) confirmed detectable MGITC signatures at ultralow concentrations (10⁻¹⁰ m), which the characteristic peak intensities increased with the increase of MGITC concentrations. Take the mapping image of 10⁻⁶ as an example, high‐resolution mapping (324 pixels, 1 pixel = 1 µm × 1 µm) demonstrated substrate reproducibility (RSD = 8.0%, Figure S24). This combination of nanomolar sensitivity and spatial uniformity validated NAS as a robust SERS substrate. Further exploration into the reasons for the uniformity of the nanocone array revealed that the detection volume is an approximate cylinder of the laser focus, which has a high enough power density. Utilizing a 633 nm laser with a 0.5 NA objective, the airy disk diameter (D = 1.22 λ/NA) was calculated to be 1.54 µm. Given the midsection diameter of the nanocone with 160 seconds ICP etching was 171 ± 17 nm (Table S4), each laser focal volume statistically encompasses multiple nanocones (Figure S25). This ensemble averaging effect across ≈9 nanocones per sampling area inherently compensated for individual nanostructural variations, yielding reproducible SERS signals.

FIGURE 6.

Integrated characterization and validation of nanocone array substrate and exosome recognition. (A) SEM images of the SERS active substrate fabrication and exosome recognition processes. (a) PET film, (b) self‐assembly PS microspheres on PET, (c) ICP etching, (d) deposition of AuNPs by E‐beam, (e) CP modified NAS, (f) exosomes captured on NAS, (g) CPCSCS@EXs@NAS, (h) CPCSCS@EXs@NAS after TMB catalytic reaction. (B) Molecular binding mechanism between CPCSCS and exosomes. (C) EDS elemental mapping of Au/Pt/P distribution on active substrate. (D) UV–vis absorption spectra tracking optical property evolution during fabrication. (a) PET film, (b) self‐assembly PS microspheres on PET, (c) ICP etching, (d) deposition of AuNPs by E‐beam, (e) CP modified NAS, (f) exosomes captured on NAS, (g) CPCSCS@EXs@NAS, (h) CPCSCS@EXs@NAS after TMB catalytic reaction, (i) Au deposited‐PET, (j) CPCSCS. (E) Control substrate modifications and feasibility assessment schematics with corresponding Raman spectra. (i) Without gold film, (ii) without exosomes, (iii) substitutive CP, (iv) substitutive CPCSCS, (v) Au@Pt NP@EXs@NAS SIC structure, (vi) CPCSCS@EXs@NAS SIC structure. (F) The Raman spectra of the catalytic performance of Au@Pt NP@EXs@NAS SIC (v) vs CPCSCS@EXs@NAS SIC (vi) structure. The asterisk represents the characteristic peak at 1607 cm⁻¹. Comparative SEM images (G), morphology models (H) and FDTD simulations of electromagnetic field enhancement (I) of Au@Pt NP@EXs@NAS SIC vs CPCSCS@EXs@NAS SIC structure.

As a proof‐of‐concept test, the isolated exosomes (EXs) were primarily characterized (see Section S1.7 for detailed experimental procedures for cell culture and exosome isolation). Nanoparticle tracking analysis (NTA) quantified a monodisperse population (70–200 nm diameter) with a mean particle size of 101 nm and a concentration of 8.3 × 10⁸ particles µL⁻¹ (Figure S26A), aligning with established exosome biomarkers [78, 79, 80, 81]. Cryo‐electron microscopy imaging (Figure S26B) confirmed the preservation of the canonical cup‐shaped morphology and membrane integrity. These findings validated the suitability of isolated exosomes for subsequent analytical workflows. The construction process of CPCSCS‐based SERS active substrate for exosome recognition was schematically illustrated in Figure S19e–h, and their corresponding SEM verification was shown in Figure 6A‐e–h. Exosomes were captured through NAS‐immobilized capture probe (CP, EpCAM aptamer). Concurrently, the antisense CD63 aptamer sequence encoded within the linear template enabled programmable CPCSCS‐exosome conjugation via multivalent DNA hybridization, forming octopus‐like stable complexes through multiple binding “tentacles” (Figure 6B). This CPCSCS@EXs@NAS sandwich immunocomplex (SIC) structure retained the advantage that the cone array has a large specific surface area to provide abundant binding sites for exosomes, significantly enhanced binding stability and detection specificity. Morphological analysis revealed that each CPCSCS could cover multiple nanocones (Figure 6A‐g), suggesting their capacity for multivalent exosome binding. EDS elemental mapping (Figure 6C) confirmed these spherical structures exhibited characteristic Au/Pt/P signatures, verifying successful capture of CPCSCS on the NAS.

The integration of CPCSCS with nanocones significantly enhanced photon confinement within a 3D NAS substrates. This multiscale micro‐nanohierarchical structure demonstrated greater signal packing density compared to planar counterparts, while enhanced light absorption in a confined space and increasing matter‐photons interaction at the microscale level. UV–vis absorption analysis (Figure 6D) tracked the optical evolution during the stepwise assembly of the SERS‐active substrate. Sequential conjugation of the CP (Figure 6D‐e), exosomes (Figure 6D‐f), and CPCSCS (Figure 6D‐g) progressively enhanced the broadband absorption across 350–800 nm, confirming successful layer‐by‐layer integration. The final SERS active substrate (Figure 6D‐g) exhibited a threefold absorption enhancement compared to the Au deposited‐PET (Figure 6D‐i), while CPCSCS particles (Figure 6D‐j) showed a sixfold enhancement versus PET (Figure 6D‐a), underscoring their intrinsic light trapping. Subsequent introduction of the TMB chromogenic system (pH 5.0) induced an absorbance reduction, attributable to CPCSCS structural reconfiguration under acidic conditions‐a phenomenon corroborated by SEM images. This pH‐dependent structural instability originated from Mg₂PPi crosslinker dissolution in acidic environments (pH < 5.5), consistent with prior studies of nanoflower‐to‐linear structural transitions in acetate buffer systems [71].

Systematic control experiments validated the specificity and functionality of the active substrate. As demonstrated in Figure 6E, condition (i) exhibited intrinsic PET Raman signatures (1614 cm⁻¹, Table S5) but lacked oxTMB chromogenic responses (1607 cm⁻¹, Table S6) because the Au‐free substrate prevented CP immobilization via Au─S bonding, precluding CP‐mediated exosome anchoring. In contrast, conditions (ii–iv) yielded negligible response due to the failure to form complete immunocomplexes, including exosome‐negative controls, substitution with non‐EpCAM aptamer‐encoded CP, and substitution with non‐CD63 aptamer‐encoded CPCSCS, all of which disabled target recognition. In all cases, the absence of anchored exosomes precluded subsequent CPCSCS localization required for TMB catalysis. Comparative analysis of catalytic performance revealed CPCSCS with superior signal amplification capacity. The CPCSCS‐based system (vi) demonstrated an approximate 41‐fold SERS intensity enhancement at 1607 cm⁻¹ versus Au@Pt nanozymes controls (v) (Figure 6F), attributable to synergistic effects of nanocage‐enrichment‐enhanced catalytic activity and plasmonic coupling. FDTD simulations validated this enhancement that morphology‐accurate models (Figure 6H) reconstructed from SEM image (Figure 6G). Under wavelength‐specific excitation at 633 nm, the CPCSCS@EXs@NAS SIC structure demonstrated superior surface plasmon resonance (SPR) characteristics compared to the Au@Pt NP@EXs@NAS SIC structure, generating a higher density of electromagnetic hotspots localized at interparticle interfaces (Figure 6I). This enhancement arose from synergistic photon confinement enabled by three spatially coordinated factors: (1) multi‐scale micro‐nano hierarchical structure‐mediated light trapping; (2) plasmonic coupling between Au@Pt NPs; (3) the alloyed Au‐Pt composition further amplified hotspots generation through interfacial electron density modulation. Importantly, these confinement effects did not act independently. Instead, their spatial overlap created a confined catalytic‐plasmonic microreactor in which catalytic generation of Raman‐active oxTMB and electromagnetic field enhancement occurred within the same nanoscale area. The coordinated spatial confinement transformed the incremental enhancements into a coupled amplification process, which fully adapted to the CPCSC mechanism.

2.4. Construction and Performance Verification of the Dual‐Modal Testing Platform

Based on the validated CPCSC mechanism, a dual‐modal point‐of‐care testing platform was constructed to translate synergistic amplification into a practical sensing format. As we validated the practical application potential of this substrate, an interesting phenomenon was observed: laser irradiation induced rapid interfacial fluid evaporation under atmospheric exposure, leading to >90% SERS signal attenuation within 60 seconds (Figure 7A,B). This evaporation‐driven instability disrupted the spatial and temporal consistency between catalytic generation of Raman‐active species and optical readout, thereby fundamentally compromising analytical reproducibility and masking synergistic amplification effects. To address this, a portable microfluidic chip integrating dual functionality was engineered: (1) hermetic encapsulation to minimize liquid evaporation and (2) spatially partitioned zones for dual‐modal detection. Unlike conventional liquid‐phase dual‐modal systems, this platform operated in a solid‐phase configuration, ensuring that colorimetric and SERS readouts interrogated the same catalytic‐plasmonic microenvironment. This design was critical for the spatial colocalization required by CPCSC mechanism. Magnetic actuation enabled stable transfer of the SERS‐active substrate between the colorimetric screening and Raman detection zones, which kept spatial colocalization from disrupting. As a result, spatial and temporal consistency between catalytic signal generation and optical readout was maintained, which was essential for maintaining synergistic amplification. As illustrated in Figure 7C and Video S1, the chip fabrication and operational protocol involved: (1) stereolithography 3D printing of microfluidic chip mold; (2) polydimethylsiloxane (PDMS) curing and structure formation through thermal crosslinking; (3) stripping and dimensional refinement of the polymeric architecture; (4) immobilization of the SERS active substrate on a round iron sheet; (5) conducting TMB chromogenic reactions in a deep reservoir containing the substrate; (6) magnetically transferring the substrate to the Raman analysis zone while retaining the TMB solution for colorimetric analysis. Section S1.8 described the experimental procedures for fabrication and dual‐modal detection of a portable chip. A nanoscale aqueous interlayer, stabilized by capillary forces between the substrate and glass slide, simultaneously optimized laser focusing and mechanical adhesion. SERS mapping was utilized to evaluate substrate stability under POCT conditions, with 90 pixels (1 pixel = 1 µm × 1 µm) analyzed at the 1607 cm⁻¹ peak (Figure 7D). The resultant intensity distribution exhibited enhanced reproducibility (RSD = 6.5%, Figure S27) compared to NAS substrates (RSD = 8.0%), confirming improved spatial uniformity. This portable microfluidic chip design enabled innovative dual‐modal detection through magnetically actuated solid‐phase substrate translocation, departing from conventional liquid‐phase handling methods. By cost‐effective 3D printing (material cost: $2.55 per unit, Table S7), the printing model and the photograph of the mold were shown in Figure S28. The fabrication cost of one portable microfluidic chip for dual‐modal detection is $2.71 including the chip materials and the SERS‐active substrate. A companion portable device integrating colorimetric imaging and SERS measurement was developed with assembly schematics and operational demonstrations detailed in Figure S29 and Video S2, respectively. Optical configuration of the portable dual‐modal POCT device was shown in Figure S29A.

FIGURE 7.

Performance optimization and evaluation of the portable dual‐modal POCT device. Schematic of evaporation‐induced Raman signal attenuation (A) and the Raman spectra (B) of interfacial fluid evaporation during Raman measurements. (C) Portable chip fabrication and operational protocol for dual‐modal (Raman‐colorimetric) detection. (D) Substrate homogeneity assessment: SERS intensity distribution at 1607 cm⁻¹ across the portable chip. (E) Au@Pt nanozymes concentration optimization: comparative SERS response curves (0–2 nm) for commercial Raman system vs. this work. (F) Catalysis time optimization: temporal evolution of TMB chromogenic signals (0–120 min). (G) SERS mapping of exosome concentrations (a‐f: 10⁵‐10⁰ particles µL⁻¹) acquired using a commercial Raman system (18 × 18 µm²). (H) The photographs and representative Raman spectra of various exosome concentrations acquired from a commercial Raman system and this work. Curves a–f represent 10⁵–10⁰ particles µL⁻¹ exosome concentrations. (I) The calibration curve at 1607 cm⁻¹ with various exosome concentrations was performed on a commercial Raman system and this work. (J) Selectivity analysis against interferent proteins (point i‐vi represents exosome, BSA, AFP, CEA, PSA, thrombin).

Furthermore, to further validate the stability and robustness of the colorimetric readout, we performed additional control experiments focusing on the influence of detection chamber geometry on the TMB chromogenic response. Specifically, as shown in Figure S30A, three different detection chamber geometries, including hexagonal (a), rectangular (b), and triangular (c) chambers, were newly fabricated and systematically compared with the round detection chamber of portable chip. Two independent TMB chromogenic conditions, denoted as groups (i) and (ii), were then tested across all geometries. As the same reaction conditions, the corresponding photographs demonstrated that no obvious difference presented among four chambers. To evaluate the robustness of the colorimetric readout, RGB heatmap analysis was performed for all chambers (Figure S30B). The RGB profiles showed consistent color distributions among different chamber geometries within the same experimental group, while maintaining obvious difference between groups (i) and (ii). In addition, grayscale intensity analysis (Figure S30C) further confirmed the reproducibility of the colorimetric signals, with relative standard deviation (RSD) values of 3.4% and 6.2% for groups (i) and (ii), respectively. These low signal variations demonstrated that the proposed colorimetric strategy possessed good reproducibility and resistance to different detection chamber geometries. It should be noted that, despite the current dependence on laboratory‐assisted exosome isolation, the proposed platform substantially reduces instrumentation complexity, space utilization, and operational cost compared with conventional systems, thereby demonstrating the potential for future decentralized and point‐of‐care analytical applications.

To validate the performance of the portable Raman‐colorimetric dual‐modal detection device, we performed a comparative assessment against the commercial Raman system. Based on the previous optimization of these critical parameters including NP dimensions (40 nm), Au–Pt bimetallic ratio (5:1), TMB chromogenic system (0.01% TMB, 0.3% H₂O₂, pH 5.0), and ICP etching duration of NAS (160 s), subsequent optimization experiments focused on CPCSCS‐doped Au@Pt nanozymes concentration gradients (0–2 nm) and TMB catalytic time optimization (0–120 min). As shown in Figure 7E, both detection strategies demonstrated concentration‐dependent SERS signal amplification at 1607 cm⁻¹, reaching maximum intensity at 2 nm. Beyond this threshold, signal saturation occurred and a similar tendency was found from colorimetric analysis. The TMB chromogenic strategy achieved optimal SERS response within 5 min of catalytic reaction initiation (Figure 7F). However, the inherent chemical instability and photoinduced degradation of oxTMB caused progressive chromophore fading over 30 min. These findings established 2 nm Au@Pt nanozymes concentration and 5 minutes of catalysis as optimal parameters for reliable detection. This also demonstrated the performance of the portable dual‐modal POCT device: although its precision was not as high as that of the commercial Raman system due to insufficient focusing effect, the overall trend was consistent and could meet the clinical examination threshold. On the other hand, the portable device has significant advantages in terms of assembly cost, size, and weight, etc., in the POCT field. The comparative data were shown in Table S8, compared to a commercial Raman system, the portable device significantly reduced equipment expenses (price only 10%, volume only 5%, weight only 9%).

Under optimized conditions, the analytical performance of the exosome detection platform was systematically evaluated. Figure 7G presented SERS mapping results for exosome concentrations spanning 10⁵‐10⁰ particles µL⁻¹, acquired using a commercial Raman system across 324 pixels (1 pixel = 1 µm × 1 µm). As an example, the uniformity of Figure 7G‐a was calculated to be 6.8% (Figure S31). Corresponding SERS spectra (Figure 7H) exhibited concentration‐dependent intensity attenuation at the 1607 cm⁻¹ peak, demonstrating a logarithmic correlation between signal intensity and exosome concentration. Comparative analysis revealed weaker sensitivity for the portable device (LOD = 10² particles µL⁻¹) versus the commercial Raman system (LOD = 10 particles µL⁻¹), with both platforms exhibiting concentration‐dependent signal attenuation. This sensitivity range (10²–10⁵ particles µL⁻¹) was within the clinically relevant exosome concentrations in human blood (10²–10⁹ particles µL⁻¹) [82, 83], confirming diagnostic utility. The 1607 cm⁻¹ peak, prominently observed in SERS spectra, was designated as the quantification marker to establish the exosome concentration‐response relationship. As shown in Figure 7I, the SERS intensity increased with the increase of exosome concentration from 0 to 10⁵ particles µL⁻¹. Selectivity, a critical analytical parameter, was evaluated against common interferents, including bovine serum albumin (BSA), alpha‐fetal protein (AFP), carcinoembryonic antigen (CEA), prostate‐specific antigen (PSA), and thrombin. Negligible SERS signals were observed for non‐target analytes (Figure 7J), confirming high specificity for exosome recognition.

2.5. Clinical Sample Analysis

To rigorously validate the clinical diagnostic utility of our dual‐modal platform, we performed comprehensive analyses using exosomes isolated from both cell culture media and human blood samples. The system achieved intrinsic synergy through the combination of rapid colorimetric screening and confirmatory SERS quantification, enabling a streamlined diagnostic workflow suitable for point‐of‐care settings. As illustrated in Figure 8A, exosome detection was carried out in parallel using our portable Raman‐colorimetric dual‐modal detection device and a commercial enzyme‐linked immunosorbent assay (ELISA) kit, enabling direct methodological comparison. Section S1.10 described the experimental procedures for ELISA analysis. While both platforms employ colorimetric readouts, our approach uniquely integrated SERS‐based verification to enhance analytical reliability.

FIGURE 8.

Clinical sample analysis. (A) Schematic of exosome detection in blood samples and cell culture medium using this method versus commercial ELISA. (B) Linear relationship between absorption intensities (450 nm) and logarithmic exosome concentrations. (C) Linear correlation of SERS intensity (1607 cm⁻¹) versus logarithmic exosome concentrations. (D) Recovery rates of exosomes at different concentrations across three detection methods. (E) RGB heatmap analysis of exosomes from prostate cancer patients (n = 9) versus healthy controls (n = 9). (F, G) SERS/absorbance intensities and grayscale intensities for patient/control cohorts. (H) Exosome concentrations quantified by this work versus ELISA.

The highly sensitive analysis results demonstrated concentration‐dependent signal variations against logarithmic exosome concentrations (10⁰–10⁵ particles µL⁻¹), with R ² values of 0.99 (ELISA), 0.98 (commercial Raman system), and 0.95 (this work) (Figure 8B,C). The corresponding fitting equations were provided as Equations (5)‐(7). Notably, recovery rate tests demonstrated significantly improved accuracy for our platform (79.0‐127.0%) compared to ELISA (113.4‐199.0%) and the commercial Raman system (94.6–156.0%) (Figure 8D), underscoring the enhanced quantification reliability afforded by dual‐modal cross‐validation.

$$\mathrm{ELISA}:\mathrm{y}=0.19-0.03\mathrm{x}+0.03{\mathrm{x}}^2$$ (5)

$$\mathrm{Commercial}\mathrm{Raman}:y=3170-3839.27x+4188.54x^2$$ (6)

$$\mathrm{This}\mathrm{work}:y=949-1146.25x+1488.7x^2$$ (7)

Colorimetric analysis further confirmed the practicality of system, exhibiting a progressive color shift toward white (RGB 255, 255, 255) with decreasing exosome concentration (Figure S32). A linear relationship between grayscale intensity and logarithmic exosome concentration was observed (R ² = 0.96 for this work vs 0.97 for ELISA, Figure S33), supporting the consistency and visual interpretability of the colorimetric screening step.

In clinical validation, the blood samples were obtained from prostate cancer patients (n = 9) and healthy controls (n = 9), and the detection platform demonstrated preliminary applicability in complex biological matrices. As summarized in Figure 8E, distinct RGB profiles clearly differentiated the two cohorts. The corresponding SERS/absorbance intensities and grayscale intensities further supported the feasibility of distinguishing patient and control clinical samples under the current experimental conditions (Figure 8F,G), and a strong correlation between our method and ELISA quantification validated its detection accuracy (Figure 8H).

Critically, the dual‐modal design addressed key limitations of conventional single‐mode assays: the colorimetric channel enabled rapid preliminary screening, while the SERS channel provided molecularly specific quantification, effectively minimizing false positives/negatives. This synergistic operation met the demand of clinical applications, that initial screening should be rapid and reliable. By integrating two detection modalities into one portable platform, the present work demonstrated both high‐throughput screening and precision diagnostics, highlighting the potential of the proposed strategy for future POCT‐related applications. Future development of simplified or direct exosome enrichment strategies would be important to achieve fully integrated sample‐to‐answer POCT systems. In addition, future studies may focus on expanding clinical cohort sizes and conducting systematic diagnostic evaluations, including ROC analysis, blinded testing, and sensitivity/specificity assessments.

3. Conclusion

This study resolved the critical spatiotemporal disjunction in catalytic‐optical amplification by establishing a DNA‐regulated catalytic‐plasmonic colocalization‐based synergy coupling mechanism. By employing rolling circle amplification‐derived DNA nanocages as programmable spatial regulators, Au@Pt nanozymes were precisely confined within plasmonic hotspots through base‐pair‐encoded hybridization. This strategy enforced spatial and temporal consistency between catalysis and optical readout, enabling Raman‐active species to be generated directly within regions of maximum electromagnetic field enhancement. Through coordinated interface, pore, and interlayer confinement, the result architectures achieved cross‐scale signal amplification spanning molecular, nanoscale, and microscale levels. This cross‐scale confinement strategy achieved exceptional performance: approximate 41‐fold SERS enhancement, ultralow detection limit (10² exosomes µL⁻¹), and high reproducibility (6.5% RSD). Importantly, this mechanistic advance was translated into a dual‐modal point‐of‐care testing platform specifically designed to preserve catalytic–plasmonic colocalization during operation. The device synergizes rapid colorimetric screening with confirmatory SERS quantification, achieving 95% volume reduction and 90% cost savings versus commercial systems as well as maintaining laboratory‐level accuracy. Beyond the present exosome detection model, the CPCSC mechanism may provide typical design considerations for future catalytic‐plasmonic sensing systems. Future studies may explore the extension of this strategy to other nanozyme systems, alternative plasmonic materials, and dynamic biological microenvironments, as well as its integration with multiplexed or in situ analytical platforms. Collectively, this work advances both the mechanistic understanding and practical implementation of synergistic catalytic‐plasmonic sensing.

4. Experimental Section

Detailed experimental materials and methods can be found in the Supporting Information.

Author Contributions

Yeru Wang: writing – original draft, methodology, investigation, and data curation. Rongke Gao: writing – review & editing, validation, supervision, methodology, and funding acquisition. Zihao Wang: methodology and data curation. Changbiao Zhan: writing – review & editing. Hancheng Liu: writing – review & editing. Wei Peng: writing – review & editing. Haiyang Wei: writing – review & editing. Long Li: writing – review & editing. Yiyue Yu: writing – review & editing. Wenbo Zhou: writing – review & editing. Yujie Feng: supervision and methodology. Yang Lu: writing – review & editing. Liandong Yu: writing – review & editing, validation, supervision, methodology, and funding acquisition.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: adma74043‐sup‐0001‐SuppMat.docx.

Supporting File 2: adma74043‐sup‐0002‐VideoS1‐S2.zip.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.