Cu2O@CeO2 heterostructure nanozyme with synergistic amplification effect-enhanced immunoassay for sensitive detection of Salmonella in food Authors: Ruya Guo, Zhuorui Li, Jingjiang Qiu, Zhongwei Guo, Wuzhen Qi, Li Xue, Xiangyang Gao, Ronghan Wei, Jianhan Lin Published: 2026 ABSTRACT AbstractColorimetric immunoassays based on nanozyme labelling offer promising solution to rapid detect pathogenic bacteria for ensuring food safety. However, sluggish electron transfer kinetics and finite active sites of nanozyme result in unsatisfactory catalytic activity, achieving sensitive detection remains a challenge. Here, we report a Cu2O@CeO2 heterostructure nanozyme with synergistic amplification effect integrating rapid charge transfer kinetics and abundant active sites to enhance the colorimetric immunoassay for sensitive detection of pathogenic bacteria. Highly active CeO2 nanoparticles (NPs) are uniformly and firmly anchored onto stable Cu2O nanocubes to form Cu2O@CeO2 core-shell heterostructure, which effectively alleviates the aggregation of NPs and achieves extremely exposure of catalytic active sites. Density functional theory (DFT) calculations reveal that electron transfer at the heterointerface induces a strong built-in electric field, along with abundant density and high affinity for H2O2, thus enhancing the reaction kinetics and catalytic activity of the nanozyme. Benefiting from these advantages, the Cu2O@CeO2 heterostructure nanozyme exhibits excellent peroxidase-like activity, which could effectively catalyze the substrate 3,3′,5,5′-tetramethylbenzidine (TMB) to generate blue oxidized TMB. More importantly, the colorimetric immunoassay utilizing Cu2O@CeO2 nanozyme as labeling probe achieves sensitive detection of Salmonella with a low limit of detection (LOD) of 26 CFU/mL and presents high specificity and stable applicability in spiked food samples. This work provides scientific insights for developing high-activity heterostructure nanozyme and advancing the exploitation of fast bacteria detection technologies to ensure food safety.Keywords: Cu2O@CeO2 heterostructure, Synergistic amplification, Peroxidase-like nanozyme, Salmonella detection, Food safety Introduction Foodborne diseases have become a growing health issue worldwide with severe impact on global economy and public health. Pathogenic bacteria (such as Salmonella) are recognized as the primary cause of foodborne diseases according to the World Health Organization (WHO, 2026). Thus, early detection of pathogenic bacteria is pivotal in preventing and controlling the outbreaks of foodborne diseases. For bacteria detection, culture method (Srikanth et al., 2022) is gold standard, which is accurate but time-consuming (2-4 d). Polymerase chain reaction (PCR) (Cui et al., 2024) is a fast detection method, which significantly shorten the detection time to several hours. Nonetheless, PCR rely on sophisticated operations, cumbersome equipment and experienced technicians (Gao et al., 2024). Alternatively, various colorimetric immunoassays have achieved increasing attention for bacteria detection in recent years owing to their easy operations and naked-eye observation (Nasrollahpour and Khalilzadeh, 2024). However, colorimetric immunoassays based on enzyme labelling still suffer from insufficient sensitivity because of the unsatisfactory activity of employed enzymes. For instance, horseradish peroxidase (HRP) as a natural enzyme is widely employed as the labeling probe in colorimetric immunoassay to catalyze the substrate to produce the color change (Park et al., 2025). However, the catalytic activity of HRP is prone to affect by the environmental factors (Liang and Yan, 2019), which could lead to the insufficient sensitivity for bacteria detection. Hence, it is an urgent need to explore superior labeling probe to enhance the performance of colorimetric immunoassays for sensitive detection of pathogenic bacteria. Nanozymes are catalytic nanomaterials with enzyme-like characteristics that have attracted extensive interest over the past decade owing to their tunable catalytic activity, excellent stability even under harsh conditions and low cost, positioning them as promising alternatives to natural enzymes (Liang and Yan, 2019; Lin et al., 2024). Especially, peroxidase (POD)-like nanozymes, including noble metal nanoparticles (Au, Pt, etc.) (Huang et al., 2026; Zhang et al., 2025a; Lin et al., 2026), metal-organic frameworks (MOFs) (Bian et al., 2026; Hong et al., 2023; Shafique et al., 2025), and metallic oxides (Fe3O4, Co3O4, Cu2O, MnO2, CeO2, etc.) (Yuan et al., 2023; Dong et al., 2025; , H. Wang et al., 2026b; Xue et al., 2021; Nguyen et al., 2022), have shown great potential to replace HRP in the development of colorimetric immunoassays. Among them, transition metal-based nanozymes (Zhang et al., 2024a) exhibit outstanding catalytic activity due to their unique physical and chemical properties, diverse crystal structures, and remarkable electron transfer capabilities. CeO2 nanoparticles (NPs) feature oxygen vacancy and uniquely reversible Ce3+/Ce4+ redox pair (Liao et al., 2024), which facilitates high redox reaction and tunable activity, receiving particular interest. Despite CeO2 NPs have shown greater performance due to more oxygen vacancies, the activity still faces the challenge of deactivation due to the drawback of easy aggregation, which has seriously limited their further application. Searching the suitable scaffolds to disperse CeO2 NPs and constructing the core-shell structures could be an efficient approach to improve the stability and enhance the catalytic activity of CeO2-based nanozymes. At present, various scaffold materials have been employed to disperse CeO2 NPs and construct heterojunctions, including carbon-based materials (Graphene (Adegoke et al., 2021), N-C (Gai et al., 2023)), metal oxides (TiO2 (Wang et al., 2025), Co3O4 (Zhang et al., 2025c), NiO (Khyrun et al., 2023)) and MOFs (Su et al., 2024). Despite the encouraging progress in mitigating CeO2 aggregation by the aforementioned scaffolds, they generally suffer from limitations on weak interfacial interactions, difficulty in morphological control and poor biocompatibility, making it difficult to meet the complex biosensing requirements of foodborne pathogens. In this regard, Cu2O nanocubes (NCs) feature complete three-dimensional rigid framework, abundant oxygen vacancies and unsaturated surface Cu sites, which are considered excellent multifunctional building species (Zhang et al., 2025d). Meanwhile, the CeO2 shell tightly encapsulates the Cu2O core, forming a complete and continuous Cu2O-CeO2 interface. This continuous interfacial channel allows the rapid redox cycles of Ce4+/Ce3+ and Cu+/Cu2+ to continuously generate new active sites, significantly enhancing the catalytic activity of nanozymes (Liu et al., 2025; Ansari et al., 2025). Furthermore, heterostructure could spontaneously form a built-in electric field (BIEF) at the interface, which accelerates electron reconstruction and transport (Shi et al., 2024; Chen et al., 2025), while improving the reaction kinetics between nanozymes and substrates, further enhancing the catalytic activity of nanozymes. Notably, the intact core-shell heterostructure provides a more stable interfacial environment and an excellent biomolecule immobilization platform, which is beneficial for constructing biosensing systems with good reproducibility and high reliability (Wang et al., 2026a). Therefore, the Cu2O@CeO2 core-shell heterostructure nanozymes have the great potential to be employed as labeling probe for enhancing the performance of colorimetric immunoassays and achieving sensitive detection of pathogenic bacteria. Herein, we report Cu2O@CeO2 heterostructure nanozymes with synergistic amplification effect as signal probes to enhance the colorimetric immunoassay, aiming at sensitive detection of Salmonella (Fig. 1). The Cu2O@CeO2 nanozyme modified with antibody and immune magnetic bead (IMNB) are employed to label the target bacteria for obtaining MNB-Salmonella-Cu2O@CeO2 immunocomplex. Cu2O@CeO2 nanozyme as signal probe on the immunocomplex possess highly POD-like activity, catalyzing the substrate TMB to generate blue oxidized TMB (oxTMB), which is analyzed for Salmonella determination. The ingeniously designed heterostructure nanozymes have the following advantages: (i) The stable Cu2O NCs core prevents the aggregation of highly active CeO2 NPs shells, which achieves the maximum exposure of active sites and enhances the overall stability of the nanozyme structure. (ii) The CeO2 shell effectively alleviates the Cu2O core from dissolving, oxidizing, and collapsing in complex food environments, while retaining the POD-like activity of Cu2O-CeO2 heterointerfaces without blocking mass transfer of H2O2 and chromogenic substrates. (iii) Electron transfer from Cu2O to CeO2 induces a strong BIEF at the heterointerface, accelerating interfacial electron transport and nanozyme catalytic reaction kinetics. (iv) Cu2O@CeO2 heterostructure nanozyme integrates the strong synergistic effect between rapid electron transfer and abundant active sites, thus exhibiting superior catalytic activity and significantly improving the sensitivity and stability of nanozyme-based immunoassay. Moreover, the effect mechanism of interfacial electron transfer on the catalytic activity of heterostructure nanozymes is explored by density functional theory (DFT) calculations. The analytical performance of this Cu2O@CeO2-based immunoassay is investigated for Salmonella detection by evaluating the sensitivity, specificity, and applicability in spiked food samples. This work has shown promising potential of the heterostructure nanozyme for application of sensitive Salmonella detection, advancing the exploitation of fast detection technologies for ensuring food safety. Schematic illustration for (a) the fabrication of immune Cu2O@CeO2 heterostructure nanozyme, (b) the application of Cu2O@CeO2 nanozyme for colorimetric immunoassay of Salmonella. Experimental section Materials Cupric sulfate (CuSO4), sodium hydroxide (NaOH), ascorbic acid (AA), ammonium ceric nitrate ((NH4)2Ce(NO3)6), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC·HCl), N-hydroxysulfosuccinimide sodium (NHSS), 3,3′,5,5′-tetramethylbenzidine (TMB), hydrogen peroxide (H2O2), isopropanol (IPA), p-benzoquinone (BQ), sodium azide (NaN3), 2,2,6,6-tetramethyl-4-piperidone hydrochloride (TEMP), 5,5-dimethyl-1-pyrroline N-oxide (DMPO), phosphate buffered saline (PBS), and bovine serum albumin (BSA) were purchased from Aladdin (Shanghai, China). Carboxylated magnetic nanobeads (MNBs) were procured from Allrun Nano (Shanghai, China). The monoclonal antibodies (Catalog No. C86309M) against Salmonella typhimurium were purchased from Meridian (Cincinnati, OH, USA). The anti-Salmonella polyclonal antibodies (Catalog No. 20C-CR7100RP) were obtained from Biosynth (Compton, Berkshire, UK). Luria-Bertani (LB) broth and LB agar were obtained from Aoboxing (Beijing, China). All other reagents were of analytical grade and supplied by Solarbio (Beijing, China). Synthesis of Cu2O@CeO2 First, 8 mL of CuSO4 (0.01 M) and 72 mL of polyethylene glycol 600 (PEG 600, 0.1 M) were mixed in a beaker. Then 6 mL of NaOH (0.5 M) and 1 mL of AA (0.1 M) were added in sequence and stirred for 10 min, followed by incubating for 30 min undisturbed. Subsequently, the precipitates were collected by centrifuging at 6800g for 10 min and washing three times with deionized water and ethanol to remove residual reagents. Finally, the Cu2O NCs were obtained after drying in a vacuum oven at 60°C overnight. Cu2O@CeO2 nanozymes were synthesized via template-assisted method. 20 mg of Cu2O NCs were first added in 80 mL of ethanol and mixed with 2 mL of NaCl aqueous solution (0.855 M). Then the mixture was fully dispersed by ultrasonication (Frequency: 40 kHz, Power: 600 w) for 30 min at 20°C using ultrasonic cleaner (Fuyang F-100SD, Shenzhen, China), followed by transferring to an oil bath maintained at 40°C. After that, 20 mL of (NH4)2Ce(NO3)6 ethanolic solution (0.1 M) was added drop by drop under stirring and incubated for 60 min. Finally, the Cu2O@CeO2 nanozymes were obtained by washing three times (6800g, 10 min) with deionized water and ethanol and dried in a vacuum oven overnight. Preparation of immune MNBs and immune Cu2O@CeO2 Carboxylated MNBs were modified with antibodies via covalent binding (Jiang et al., 2024). First, 1 mg of MNBs were washed three times by magnetic separation (2 min) with phosphate buffer (PB, 0.01 M, pH 6.0) and resuspended in 3 mL of PB. Subsequently, the freshly prepared solutions of EDC·HCl (1 mg) and NHSS (1 mg) were added in sequence and reacted for 1 h with shaking, followed by washing twice with 3 mL of PB, 0.1 mg of anti-Salmonella polyclonal antibodies were added to incubate with the activated MNBs for 2 h. Then, the unbound antibodies were removed by magnetic separation for 2 min and washed twice with PB. After that, 0.3 mL of BSA (10%, w/v) was employed to block excessive sites on the surface of MNBs for 1 h. Finally, the immune MNBs were collected after washing twice with PBS and stored in 1 mL of PBS containing 1% BSA and 25% sucrose at 4°C for future use. Cu2O@CeO2 nanozymes were coated with antibodies through electrostatic adsorption (Duan et al., 2024). Briefly, 0.2 mg of Cu2O@CeO2 nanozymes were dispersed in 1 mL of 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES) buffer (pH 8.0), followed by adding 0.1 mL of anti-Salmonella monoclonal antibodies solution (0.2 mg/mL) dropwise under gentle stirring for 2 h. After that, the unbound antibodies were removed by centrifugation at 6500g for 10 min and washed twice with HEPES buffer. Then 0.1 mL of BSA (10%, w/v) was added and incubated for 1 h to prevent nonspecific adsorption. The immune Cu2O@CeO2 nanozymes were obtained by centrifugation at 6500g for 10 min and resuspended with 0.4 mL of HEPES buffer (pH 7.4) containing 1% BSA and 25% sucrose, followed by storing at 4°C until used. Materials characterization Scanning electron microscopy (SEM) was performed by JSM-7610FPlus (JEOL, Japan). For SEM characterization, 10 mg of powdered samples were adhered to conductive carbon tape and cleaned with pressurised air. SEM imaging was performed at an accelerating voltage as high as 15 kV and a working distance of 8 mm. Transmission electron microscopy (TEM), high-resolution TEM (HRTEM), High-angle annular dark-field scanning TEM (HAADF-STEM) elemental mapping and energy dispersive spectrometer (EDS) were implemented in Talos F200s (FEI, USA), the specimens were prepared by dispersing the powdered samples in ethanol solution and dropped on copper mesh (10 μL), then dried at 60°C in vacuum dryer. TEM related imaging was carried out at an accelerating voltage up to 200 kV. X-ray photoelectron spectroscopy (XPS, AXIS Supra, Shimadzu, Japan) was performed to analyze the element composition, valence state and interfacial electron transfer, the specimens were prepared using 10 mg of powdered samples. The crystal structure and phase conformation were measured by X-ray diffraction (XRD, Empyrean, Malvern Panalytical, UK), 20 mg of powdered samples were used for testing. Dynamic light scattering (DLS) and zeta potential measurement were studied by Zetasizer Nano ZS90 analyzer (Malvern Panalytical, UK), the specimen was prepared by adding 10 mL (2 mg/mL) of each sample dispersed in deionized water to a special cuvette. Preparation and culture of bacteria The pathogenic bacteria employed in this study were listed in Table S1. Salmonella typhimurium was used as target bacteria. Escherichia coli (E. coli) O157:H7, Listeria monocytogenes, Staphylococcus aureus and Bacillus cereus were used as non-target bacteria. All these bacterial strains were stored at −20°C with 50% (v/v) glycerol until used. These strains were first cultured in LB broth at 37°C with shaking for 16-18 h, and then 10-fold diluted in sterile PBS to obtain a serial dilution from 101 to 105 CFU/mL of bacterial suspensions. The enumeration of bacteria was used the standard plate counting method. Briefly, the bacterial cultures were 10-fold serially diluted with sterile PBS. Then 100 μL of the diluents were surface plated on the LB agar plates. After incubation at 37°C for 20-22 h, the visible colonies were counted to determine the bacteria concentration. For calibration curve establishment, the Salmonella concentrations were determined from 4.2 × 101 to 4.2 × 105 CFU/mL. Immunoassay procedure for Salmonella detection Prior to test, the centrifuge tubes were blocked by 1%BSA for 1 h and then washed with sterile PBS. First, 20 μg of immune MNBs and 15 μg of immune Cu2O@CeO2 nanozymes were added into 1 mL of PBS containing Salmonella cultures at different concentrations from 101 CFU/mL to 105 CFU/mL, respectively. The mixture was vortexed homogeneously and incubated for 40 min to form MNB-Salmonella-Cu2O@CeO2 sandwich complexes. Then, the complexes were collected by magnetic separation (2 min) and washing twice to remove the unbound Cu2O@CeO2 nanozymes. Finally, 200 μL of H2O2-TMB substrates (10 mM of H2O2, 1 mM of TMB) in acetate buffer (0.2 M, pH 4.0) were added and catalyzed at 40°C for 10 min by the Cu2O@CeO2 nanozymes on the sandwich complexes, leading to the color change from colorless to blue, followed by magnetic separation (2 min) to remove MNB-Salmonella-Cu2O@CeO2 sandwich complexes (stop the catalytic reaction) and collect the supernatant containing the reacted substrate solution, the absorbance at 652 nm of the solution was measured by microplate reader (Multiskan Sky, Thermo, USA) to determine Salmonella concentration. Detection of Salmonella in spiked food samples To evaluate the applicability of this proposed immunoassay, low fat milk (Yili, Inner Mongolia, China) and fresh chilled chicken breast meat of white-feathered chicken (Charoen Pokphand Food, Beijing, China) were purchased from a local supermarket for preparation of spiked samples. 10 mL of milk was added to 90 mL of sterile PBS to prepare the samples containing different concentrations of Salmonella cultures (5.7 × 101-5.7 × 104 CFU/mL). 25 g of chicken meat was added into 225 mL of sterile PBS containing different concentrations of Salmonella cultures (6.1 × 101-6.1 × 104 CFU/mL) and homogenized for 3 min to obtain the spiked chicken samples. The spiked food samples were then detected by the proposed immunoassay and the recovery rates (%) were analyzed to assess the applicability. Statistical analysis All error bars in the figures represent standard deviation (SD) calculated from three repeated experimental data (n = 3). Statistical comparison was performed using one-way analysis of variance (ANOVA) with Dunnett's post-test, P < 0.05 was considered statistically significant. The limit of detection (LOD) was determined based on three times the SD (3 × SD) of negative control samples (Moulahoum and Ghorbanizamani, 2024) and the calibration curve. Results and discussion Characterization The morphology is observed visually by SEM. Fig. 2a shows that Cu2O exhibits a highly uniform cubical nanostructure, which is suitable for acting as scaffolds to support and disperse the CeO2 NPs. Meanwhile, the diameter of Cu2O NCs is approximately 180-200 nm. After growing CeO2 NPs on the surface of Cu2O NCs, the cubical nanostructure is still well maintained (Fig. 2b). To obtain more detailed nanostructure and interface relationship, TEM and HRTEM are further performed. From Fig. 2c, the morphology of Cu2O shows consistent morphology with SEM with good dispersity. Fig. 2d and e reveal that the CeO2 NPs are uniformly anchored onto stable Cu2O NCs to form Cu2O@CeO2 core-shell structure, which is conducive to extremely expose the active sites for improving the catalytic activity. From HRTEM image, the obvious phase boundary between Cu2O and CeO2 is identified (Fig. 2f), indicating the successful formation of heterostructure interface that will spontaneously induce the generation of BIEF to accelerate electron transport and improve the reaction kinetics. Additionally, the lattice spacing of 0.246 nm and 0.312 nm correspond to Cu2O (111) and CeO2 (111) planes, respectively. EDS image demonstrates the only existence of three elements including Cu, O and Ce in Cu2O@CeO2 (Fig. S1). HAADF-STEM and corresponding elemental mapping images illustrate uniform distribution of Cu, O and Ce elements in Cu2O@CeO2 nanozymes (Fig. 2g). Notably, Ce element is mainly located at the edges while the Cu element is mainly located in the center, further revealing the formation of Cu2O@CeO2 core-shell structure. SEM images of (a) Cu2O and (b) Cu2O@CeO2. TEM images of (c) Cu2O and (d) Cu2O@CeO2. (e, f) HRTEM images of Cu2O@CeO2. (g) HAADF-STEM and corresponding elemental mapping images. XPS is conducted to gain insight into the surface chemical states and interface electronic interactions. The high-resolution Ce 3d spectrum of Cu2O@CeO2 contains two spin orbitals of 3d5/2 and 3d3/2, which could be deconvolved into eight well-resolved peaks (Fig. 3a) (Xu et al., 2024; Zhang et al., 2025b). The peaks at 881.85, 888.17, 897.69, 900.39, 906.89 and 916.80 eV corresponds to the surface Ce4+ species, while others two peaks at 884.24 and 902.39 eV are assigned to Ce3+ species, indicating the presence of surface oxygen vacancies (Zhang et al., 2025b). Notably, the binding energy of Ce 3d in Cu2O@CeO2 shifts negatively by 0.6 eV in contrast with CeO2, indicating that the formation of the heterojunction induces electron transfer from Cu2O and CeO2 and strong electronic interaction, thereby increasing the electron density and lowering oxidation state of Ce (Zhang et al., 2025b). In Cu 2p spectrum of Cu2O@CeO2 (Fig. 3b), the peak at 932.51 eV (Cu 2p3/2) and 952.39 eV (Cu 2p1/2) reveal strong characteristic signal for Cu+ species, with satellite peaks (Zhang et al., 2024b; Zheng et al., 2025). Similarly, the 2p spectrum of Cu2O@CeO2 shifts to a higher binding energy by 0.4 eV compared to Cu2O, which further confirms the strong interfacial electronic interaction and stable heterojunction between Cu2O and CeO2 (Zhang et al., 2024b; Zheng et al., 2025). Furthermore, the positive shift in binding energy indicates that the Cu+ center can serve as an abundant electron donor (Zhang et al., 2025e), which can provide electrons for the efficient conversion of Ce4+/Ce3+, thereby enhancing the kinetics of the nanozyme catalytic reaction. The O 1s spectrum (Fig. 3c) reveals the coexistence of surface oxygen vacancies (Os) and lattice oxygen (Ol) in Cu2O@CeO2 (Zhang et al., 2025b; Zhang et al., 2024b). Electrophilic Os features high mobility and strong adsorption capacity for substrates and reaction intermediates in redox reaction, promoting electron transfer and nanozyme catalytic activity (Ding et al., 2023). Therefore, quantitative Os/Ol ratios are calculated in different nanozyme (Fig. S2). As expected, the Os/Ol ratio in Cu2O@CeO2 is as high as 2.35, exceeding that of CeO2 (1.91) and Cu2O (0.12), indicating higher catalytic activity of Cu2O@CeO2 nanozyme (Zhang et al., 2025b). The crystallographic structures are analyzed by XRD pattern, Fig. 3d presents that diffraction peaks are matched well with cubic Cu2O phase (space group: Pn-3m(224), PDF#99-0041 (Zhang et al., 2024b)) and CeO2 phase (space group: Fm-3m(225), PDF#34-0394 (Xu et al., 2024)), further confirming the successfully construction of Cu2O@CeO2 heterostructure. XPS spectra of (a) Ce 3d, (b) Cu 2p, (c) O 1s. (d) XRD patterns of Cu2O and Cu2O@CeO2. (e) TEM image of immune Cu2O@CeO2-Salmonella complex. (f) DLS results of Cu2O@CeO2 and immune Cu2O@CeO2. (g) Zeta potentials of Cu2O@CeO2 and immune Cu2O@CeO2. (h) TEM image of immune MNBs-Salmonella complex. (i) DLS results of MNBs and immune MNBs. (j) Zeta potentials of MNBs and immune MNBs. To confirm the modification of antibody on the surface of Cu2O@CeO2, TEM, DLS and zeta potential measurement are performed. As shown in Fig. 3e, immune Cu2O@CeO2 nanozymes are successfully bonded with target bacteria (Salmonella) to form immune Cu2O@CeO2-Salmonella complex due to the antibody-antigen immunobinding, indicating the successful antibody conjugation to the nanozymes. DLS is employed to measure the size of nanozymes. Fig. 3f shows that the size of Cu2O@CeO2 is 181.4 nm, which closely match the size results of SEM and TEM images. The polydispersity index (PDI) is as low as 0.062, indicating the good dispersity of Cu2O@CeO2 in solution. The size of immune Cu2O@CeO2 is 193.5 nm with uniform dispersity (PDI: 0.169). Fig. 3g presents that the zeta potentials of Cu2O@CeO2 and immune Cu2O@CeO2 are 13.8 mV and −20.6 mV, respectively. The significant changes in the size and zeta potential further confirm the successful preparation of immune Cu2O@CeO2. Similarly, TEM, DLS and zeta potential measurement are performed to prove the modification of antibody on the surface of MNBs. TEM image shows that immune MNBs are successfully bonded with Salmonella to form immune MNBs-Salmonella complex (Fig. 3h), verifying the successful antibody conjugation to the MNBs. The sizes of MNBs and immune MNBs are measured by DLS to be 168.9 nm (PDI: 0.022) and 182.3 nm (PDI: 0.102) with good dispersity, respectively (Fig. 3i). Fig. 3j presents that the zeta potentials of MNBs and immune MNBs are −31.7 mV and −19.5 mV, respectively. The obvious changes in the size and zeta potential both further verify the successful modification of antibody on the surface of MNBs. To further quantitative analysis of antibody loading on Cu2O@CeO2 nanozymes, bicinchoninic acid (BCA) protein assay is performed. The antibody loading capacity is determined by subtracting the unbound antibody concentration in the supernatant from the antibody concentration (Dong et al., 2025). The antibody concentration is calculated through the antibody calibration curve. Thus, the different concentrations of antibodies are first measured using BCA assay (Fig. S3a) and the calibration curve of BCA assay of antibodies is established according to the absorbance at 562 nm (Fig. S3b). The antibody loading capacity is then determined as 38.99 μg for 1 mL (0.5 mg/mL) Cu2O@CeO2 nanozyme based on the BCA results of antibody concentration in supernatant (Fig. S3c). The high antibody capacity of this nanozyme contributes to confirm sufficient antigen binding capacity and achieve the high sensitivity in the proposed immunoassay. To evaluate the stability of immune signal probes, we incubate the immune Cu2O@CeO2 nanozymes in PBS buffers with varying ionic strengths (0-100 mM NaCl) for 2 h. After separation, the supernatant is analyzed by the BCA protein assay. As shown in Fig. S3d, no significant antibody leaching is detected across all tested ionic strengths, indicating the high stability of immune Cu2O@CeO2 nanozymes. Catalytic performance The POD-like catalytic activity of Cu2O@CeO2 heterostructure nanozyme is assessed using TMB as substrate. As shown in Fig. 4a, the TMB, H2O2, Cu2O@CeO2 and TMB + H2O2 solutions present colorless and almost no absorption peak while only TMB + H2O2 + Cu2O@CeO2 solution turns blue. Meanwhile, a strong absorption peak at 652 nm can be observed in the case of TMB + H2O2 + Cu2O@CeO2. These results indicate that Cu2O@CeO2 nanozymes indeed possess the high POD-like catalytic activity, which greatly accelerates the oxidation reaction of TMB to blue oxTMB in the presence of H2O2. In addition, a tiny absorption peak at 652 nm in the case of TMB + Cu2O@CeO2 is observed (Fig. S4), revealing the ignorable oxidase-like activity of Cu2O@CeO2 nanozyme (Bian et al., 2026; Lin et al., 2025). Given the catalytic reaction conditions including pH, temperature, and reaction time have a crucial impact on the nanozyme, the catalytic activity of Cu2O@CeO2 is investigated under different conditions. The solutions with different pH from 3.6 to 6 are prepared using acetate buffer (pH 3.6-5.6) and PB (pH 6). From Fig. 4b and c, the catalytic activity of Cu2O@CeO2 exhibits superior performance under the conditions of pH 4 and 40°C, which are chosen the optimal parameters for subsequent experiments. From Fig. 4d, the value of absorbance at 652 nm increases with reaction time increase from 2 min to 12 min. The difference between the signal for 8 min and 10 min is significant (P < 0.05), while there is no significant difference (P > 0.05) between the signal for 10 min and 12 min. Thus, 10 min is chosen the optimal reaction time. To further evaluate the catalytic behavior of Cu2O@CeO2 nanozyme, different concentrations of Cu2O@CeO2 from 0.78 to 50 μg/mL are used to catalyze the TMB substrate under optimal conditions. From Fig. 4e, the blue color changes from light to dark and the value of absorption peak at 652 nm increase with the concentrations of Cu2O@CeO2 increase. A linear relationship is found between concentrations of Cu2O@CeO2 and absorbance at 652 nm, which is described by the equation as y=0.0715x+0.2666(R2=0.9803), indicating the good catalytic performance of Cu2O@CeO2 nanozyme at a concentration range from 0.78 to 50 μg/mL. Moreover, the stability of Cu2O@CeO2 nanozyme is investigated after incubating at various pH and environmental temperatures for 2 h. Fig. S5a and S5b display that there are no obvious changes on the catalytic performance of Cu2O@CeO2 under different pH (2-10) and temperature (10-85°C). After one month storage, the catalytic activity remains significantly unchanged (Fig. S5c), suggesting the high stability of Cu2O@CeO2 nanozyme. The catalytic performance of Cu2O@CeO2 heterostructure nanozyme. (a) Visible absorption spectra and photograph of different solutions. The effect on the catalytic activity of Cu2O@CeO2 under different conditions: (b) pH, (c) temperature and (d) catalytic reaction time. (e) Visible absorption spectra and photograph under different concentrations of Cu2O@CeO2. (f) The linear relationship between concentration of Cu2O@CeO2 and absorbance at 652 nm. (g) Michaelis-Menten kinetic curves of Cu2O, CeO2 and Cu2O@CeO2 nanozyme under varying concentrations of TMB. (h) Michaelis-Menten kinetic curves of Cu2O, CeO2 and Cu2O@CeO2 nanozyme under varying concentrations of H2O2. (i) Optimization of the amount of immune Cu2O@CeO2 nanozymes. To explore the POD-like catalytic mechanism of Cu2O@CeO2, reactive oxygen species (ROS) scavenging experiments and electron spin resonance (ESR) are performed to identify the specific species of ROS generated during the catalytic process. IPA, BQ and NaN3 are used as scavengers for hydroxyl radicals (•OH), superoxide anion radicals (O2•-) and singlet oxygen (1O2), respectively (Bian et al., 2026; Dang et al., 2022). As shown in Fig. S6a, the POD-like activity of Cu2O@CeO2 nanozyme decreases significantly when IPA is used as the scavenger, while no significant change on the activity upon addition of BQ and NaN3. The results indicate that •OH is the primary ROS responsible for the POD-like activity of Cu2O@CeO2 nanozyme. Furthermore, ESR are measured by using DMPO and TEMP as the spin trapping agents. As shown in Fig. S6b, the ESR spectra present typical quartet signal peak with 1:2:2:1 intensity corresponding to the characteristic of DMPO-•OH, demonstrating the generation of •OH during the catalytic process (Bian et al., 2026; Zhu et al., 2025). However, the ESR spectra of H2O2 show no signal peaks (Fig. S6c) and no signal peaks corresponding to O2•- or 1O2 are observed (Fig. S6d and e). Collectively, the results of ROS scavenging experiments and ESR measurements fully verified that the dominant specific species of ROS is •OH, which could be derived from the Cu2O@CeO2 nanozyme catalyzing the decomposition of H2O2 via Fenton-like reaction (Kang et al., 2024). To in-depth evaluate the catalytic activity of Cu2O@CeO2 nanozyme, the catalytic kinetics analysis is conducted and the details can be found in supplementary material. As displayed in Fig. 4g, h and Fig. S7a and b, the Michaelis-Menten kinetic curves and Lineweaver-Burk plots reveal that Cu2O@CeO2 exhibit superior catalytic efficiency than pure Cu2O or CeO2 for the substrates including TMB and H2O2, suggesting that the fabrication of Cu2O@CeO2 heterostructure show a strong synergistic effect than single Cu2O or CeO2. The kinetic parameters of Cu2O, CeO2, and Cu2O@CeO2 for TMB and H2O2 are determined and shown in Table S2. Compared with HRP, Cu2O, CeO2 and previously reported nanozymes, the Km value of Cu2O@CeO2 for both TMB and H2O2 are much lower and the Vmax value of Cu2O@CeO2 for both TMB and H2O2 are significantly higher. Furthermore, the Vmax/Km value of Cu2O@CeO2 for both TMB and H2O2 are significantly higher than that of other enzymes. These results indicate Cu2O@CeO2 nanozymes exhibit higher affinity for TMB and H2O2 and faster catalytic efficiency. The Cu2O@CeO2 nanozymes possess the superior catalytic performance than HRP, pure Cu2O or CeO2, and nanozymes summarized in Table S2, which underpins the signal amplification and high sensitivity achieved in the proposed immunoassay. Furthermore, the amount of immune Cu2O@CeO2 nanozymes in this immunoassay is crucial for Salmonella detection, thus different amounts of immune Cu2O@CeO2 nanozymes from 2.5 to 20 μg are used for detecting Salmonella at 4.2 × 105 CFU/mL. From Fig. 4i, the value of absorbance at 652 nm increases with the amounts of immune Cu2O@CeO2 nanozymes increase from 2.5 to 20 μg. The difference between the signal for 10 μg and 15 μg is significant (P < 0.05), while there is no significant difference (P > 0.05) between the signal for 15 μg and 20 μg, proving 15 μg of immune Cu2O@CeO2 nanozymes is sufficient to label Salmonella. Thus, 15 μg is selected the optimal dosage for Salmonella detection. Density functional theory analysis Density functional theory (DFT) calculations are performed to explore the effect mechanism of interfacial electron transfer on the catalytic activity of heterostructure nanozymes, selecting CeO2 (111) and Cu2O (111) planes to construct the heterostructure. The corresponding optimized atomic structure configurations are shown in Fig. 5a and Fig. S8. The density of states (DOS) in Fig. 5b reveals the differences in band structure among the various nanozymes. It can be found that CeO2 and Cu2O@CeO2 exhibit fully occupied states at the Fermi level, indicating typical metallic properties (Li et al., 2024). Notably, the DOS of Cu2O@CeO2 mainly originates from Ce 3d orbitals (Fig. 5c), while the present of local cross-orbital and energy level mixing phenomena, proving robust heterostructure between Cu2O and CeO2. To further investigate electron distribution and transfer in heterostructures, differential charge density calculations and analyses are performed. As a result, the CeO2 side exhibits electron accumulation (yellow region), while the Cu2O side shows electron depletion (cyan region), indicating a charge redistribution at the interface (Fig. 5d, Fig. S9). The transfer of charge carriers alters the Fermi level and band alignment, thereby generating a BIEF at the interface pointing from CeO2 to Cu2O, which could produce additional electromotive force, enhancing electron transfer and nanozyme catalytic active sites (Zhu et al., 2025). The adsorption behavior of H2O2 on the nanozyme is further investigated, and the optimized atomic structure model is shown in Fig. S10. Differential charge density analysis reveal that the electrons are transferred from the nanozyme side to the H2O2 side (Fig. 5e–g, Fig. S11). Bader charge analysis quantified the charge transfer behavior, while revealing a transfer of 1.301 e− in the Cu2O@CeO2 heterostructure significantly higher than that in Cu2O (0.298 e−) and CeO2 (0.837 e−), indicating that the robust electron transfer capability enhances the catalytic activity of nanozyme (Qiao et al., 2025; Feng et al., 2023). Meanwhile, among the optimized models of Cu2O, CeO2 and Cu2O@CeO2 after H2O2 adsorption (Fig. 5h), the O-O bond length in Cu2O@CeO2 is the longest (1.586 Å), substantially longer than those in Cu2O (1.457 Å) and CeO2 (1.479 Å), which indicates stronger O-O bond activation and promotes the generation of •OH (Dang et al., 2022; Hong et al., 2022). Significantly, the adsorption energies of Cu2O, CeO2, and Cu2O@CeO2 for H2O2 are −0.295 eV, −1.409 eV, and −1.860 eV, respectively (Fig. 5i), indicating that Cu2O@CeO2 possesses the strongest adsorption capacity for H2O2, consistent with the catalytic kinetic analysis. The strong affinity for H2O2 facilitates the transport of H2O2, thereby accelerating the kinetics of nanozyme catalytic reactions (Wang et al., 2023). The strong BIEF generated by the construction of heterostructure nanozymes promotes interfacial electron reconstruction and transfer, enhancing the chemisorption of H2O2, electron transfer capacity and nanozyme catalytic activity. DFT calculations. (a) Front view of CeO2, Cu2O and Cu2O@CeO2 models. (b) Total DOS of CeO2, Cu2O and Cu2O@CeO2. (c) DOS of Cu2O@CeO2 with elements. (d) Front and 2D views of charge density difference at Cu2O@CeO2 heterostructure. Charge density difference and Bader charge transfer of H2O2 on the surface of (e) Cu2O, (f) CeO2 and (g) Cu2O@CeO2. (h) O-O bonds on the optimized models. (i) the corresponding O-O band length and H2O2 adsorption energies for CeO2, Cu2O and Cu2O@CeO2. Cyan and yellow regions represent charge depletion and aggregation, respectively. Analytical performance of Cu2O@CeO2-enhanced immunoassay Encouraged by the excellent POD-like catalytic performance of Cu2O@CeO2 heterostructure nanozyme, we next employ Cu2O@CeO2 as signal probe to develop a colorimetric immunoassay. Analytical performance of this Cu2O@CeO2-enhanced immunoassay is investigated by evaluating the sensitivity, specificity, and applicability for Salmonella detection. To evaluate the sensitivity, the positive samples containing different concentrations of Salmonella from 4.2 × 101 to 4.2 × 105 CFU/mL are detected under optimal conditions. The same volume of PBS without Salmonella is used as negative control for detection. From Fig. 6a and b, the signal of negative control is lower than that of positive samples containing different concentrations of Salmonella and the absorbance at 652 nm gradually increases as the concentration of Salmonella increases to 4.2 × 105 CFU/mL. The calibration curve of this immunoassay is established by plotting the value of absorbance at 652 nm as y-axis and the logarithmic concentration of Salmonella as x-axis. Fig. 6c presents a good linear correlation between the absorbance at 652 nm and the logarithmic concentration of Salmonella, which is expressed as y=0.5269x−0.3690(R2=0.9794). From the calibration curve, the working range is 4.2 × 101-4.2 × 105 CFU/mL. The LOD is determined as low as 26 CFU/mL, demonstrating the superior sensitivity of this immunoassay. The specificity of this immunoassay is evaluated by detecting the target bacteria (Salmonella) at 4.2 × 105 CFU/mL, four non-target bacteria (E. coli, Listeria, Bacillus and Staphylococcus) at 4.2 × 107 CFU/mL, and their mixture. As displayed in Fig. 6d, the signals for Salmonella and mixture are notably higher than (P < 0.001) those for non-target bacteria and negative control even the concentration of non-target bacteria is 100 times higher than that of Salmonella, proving the high specificity of this immunoassay. Analytical performance of this proposed immunoassay for Salmonella detection. (a) Visible absorption spectra of Salmonella detection at different concentrations and negative control. (b) The absorbance at 652 nm of Salmonella detection at different concentrations and negative control. (c) The calibration curve of this immunoassay for Salmonella detection. (d) The specificity evaluation of this immunoassay. ***P < 0.001. (e) Salmonella detection in spiked chicken samples. (f) Comparison of this proposed immunoassay with recently developed nanozyme-based methods for pathogenic bacteria detection. The applicability of this immunoassay is demonstrated by detecting spiked food (chicken and milk) samples containing different concentrations of Salmonella. The spiked Salmonella concentrations (6.1 × 101 to 6.1 × 104 CFU/mL for chicken samples, 5.7 × 101 to 5.7 × 104 CFU/mL for milk samples) fall within the linear range of the calibration curve. The PBS containing food samples without Salmonella is used as negative control. From Fig. 6e, the signal of negative control is lower than that of positive samples containing different concentrations of Salmonella and the signal increases with the spiked concentration of Salmonella increases. To further evaluate the analytical reliability of this immunoassay, both intra-assay and inter-assay on spiked food samples containing different concentrations of Salmonella are conducted. Intra-assay is performed within one day with six replicates at each spiked concentration while inter-assay is performed over three continuous days with six replicates at each spiked concentration. The recovery rates (%) are expressed as the ratio of the detected Salmonella concentration divided by the spiked Salmonella concentration. The spiked Salmonella concentration is precisely determined using the standard plate counting method. As shown in Table 1, for spiked chicken samples, the recovery for intra-assay ranges from 96.72% to 113.11% with the CV ranges from 3.83% to 10.62% and the recovery for inter-assay ranges from 95.03% to 114.81% with the CV ranges from 4.96% to 12.99%. For spiked milk samples (Table S3), the recovery for intra-assay ranges from 95.09% to 107.93% with the CV ranges from 3.62% to 9.62% and the recovery for inter-assay ranges from 99.65% to 114.07% with the CV ranges from 5.12% to 11.28%. These results verify that this immunoassay has good applicability in spiked food samples. The CVs of both intra-assay and inter-assay are all below 13%, demonstrating the good reproducibility of this proposed immunoassay. Table 1Spiked chicken samples analysis using this immunoassay for Salmonella detection.Table 1Spiked Salmonella (CFU/mL)Intra-assayInter-assayDetectedRecoveryCVDetectedRecoveryCV(CFU/mL)(%)(%)(CFU/mL)(%)(%)615996.7210.6265106.5612.99610690113.117.59670109.846.606100598698.133.83579795.034.966100067827111.195.9470036114.817.56 Spiked chicken samples analysis using this immunoassay for Salmonella detection. To further assess the analytical performance of this immunoassay, other rapid methods for pathogenic bacteria detection (including ELISA, PCR, and nanozyme-based colorimetry) are summarized for comparison. As shown in Table S4, the LOD of this proposed colorimetric method is two orders of magnitude lower than that of conventional PCR and ELISA, which clearly highlights the superior sensitivity of this immunoassay. Moreover, compared with other nanozyme-based colorimetric methods (Fig. 6f and Table S4), this proposed immunoassay also exhibits lower LOD and a shorter detection time, indicating the superior performance of this approach. This could attribute to the remarkable catalytic activity of Cu2O@CeO2 heterostructure nanozyme, which derives from the unique heterostructure interface with strong BIEF that accelerates electron transport and enhance the reaction kinetics. Conclusions In summary, we have successfully constructed a Cu2O@CeO2 heterostructure nanozyme with synergistic amplification effect integrating fast charge transfer kinetics and abundant active sites as signal probe and developed enhanced colorimetric immunoassay, achieving sensitive detection of Salmonella. Cu2O NCs are employed as stable scaffolds to support and disperse CeO2 nanoparticles, which efficiently improves the exposure of the active sites. Meanwhile, the CeO2 shell acts as a protective barrier, improving the structural stability and anti-interference ability of the nanozyme. Remarkably, Cu2O@CeO2 heterostructure nanozyme is demonstrated with superior POD-like catalytic activity and high stability based on the experimental studies and catalytic kinetics analysis. DFT calculations verify the existence of a strong BIEF within the Cu2O@CeO2 heterostructure interface, resulting in faster interfacial electron transfer and higher H2O2 adsorption capacity, thereby exhibiting the excellent catalytic performance of Cu2O@CeO2 heterostructure nanozyme. More importantly, the colorimetric immunoassay using Cu2O@CeO2 nanozyme as labeling probe realizes sensitive detection of Salmonella with a LOD of 26 CFU/mL within 1 h. Additionally, this proposed colorimetric immunoassay exhibits great specificity and stable detection performance in spiked food samples. This work offers a promising approach on the advancement of heterostructure nanozyme for improving the detection of pathogenic bacteria. CRediT authorship contribution statement Ruya Guo: Conceptualization, Methodology, Software, Investigation, Data curation, Writing-original draft. Zhuorui Li: Software, Investigation. Jingjiang Qiu: Investigation, Methodology. Zhongwei Guo: Investigation. Wuzhen Qi: Software. Li Xue: Methodology. Xiangyang Gao: Supervision, Resources, Funding acquisition, Conceptualization, Writing-review & editing. Ronghan Wei: Supervision. Jianhan Lin: Supervision. Data availability Data will be made available on request. Submission declaration The authors declare that: ·The work described has not been published previously; The work described has not been published previously; ·The article is not under consideration for publication elsewhere; The article is not under consideration for publication elsewhere; ·The article's publication is approved by all authors and, where required, by the responsible authorities where the work was carried out; and The article's publication is approved by all authors and, where required, by the responsible authorities where the work was carried out; and ·If accepted, the article will not be published elsewhere in the same form, in English or in any other language, including electronically, without the written consent of the copyright holder. If accepted, the article will not be published elsewhere in the same form, in English or in any other language, including electronically, without the written consent of the copyright holder. Declaration of generative Al use The authors did not use generative Al or Al-assisted technologies in the writing process of this manuscript. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgment This work was supported by the China Postdoctoral Science Foundation (Grant No. 2024M763000, 2024M762990 and GZC20241537), Natural Science Foundation of Henan (Grant No. 252300420404) and Key Research and Development Program of Henan Province (Grant No. 231111312400). Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.crfs.2026.101544. Supplementary data The following is the Supplementary data to this article: Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1 Data Availability Statement Data will be made available on request.