Novel 3S-shaped biophotonic sensor utilizing MoS₂–NSs/ZnO–NWs/AuCu–NCs for rapid detection of Shigella flexneri bacteria Open Access

Food-borne diseases with symptoms such as diarrhea and dysentery cause 685 × 10⁶ cases of infection and 2.12 × 10⁶ deaths each year, while the major pathogens of the most severe infectious food-borne disease are E. coli, Salmonella, and Shigella.^1, Shigella can be transmitted in food and water, with four serogroups: Shigella flexneri (S. flexneri), Shigella dysenteriae, Shigella boydii, and Shigella sonnei.² Shigellosis is primarily a disease present in crowded communities suffering from poor resources, a lack of adequate sanitation equipment, and a lack of safe drinking water.³ Despite the prevalence of Shigella in different regions, the disease and death caused by Shigella are more severe in countries with lower economic levels and less developed countries.^4, S. flexneri has more acid tolerance and a higher mortality rate.⁵ It is estimated that 60% and 16% of shigellosis cases in developing and developed countries, respectively, are caused by this bacterium. In some developing countries, treatment for S. flexneri is often ineffective. Cases of multiple drug resistance due to inappropriate testing facilities or misuse of antibiotics have been documented worldwide.⁶ Therefore, it is necessary to study S. flexneri for the convenience of diagnosis, food safety surveillance, and the prevention and control of infectious diseases.

Traditional methods for the detection of S. flexneri include culture and biochemical identification, which are time-consuming, laborious, and subject to some error.⁷ With the development of biotechnology, modern detection methods for S. flexneri have been greatly improved and innovative. These methods of S. flexneri mainly include molecular biological techniques and immunological techniques. Molecular biological techniques, such as polymerase chain reaction (PCR) and real-time PCR, can detect the presence of S. flexneri rapidly and accurately.⁸ Immunological techniques, such as enzyme-linked immunosorbent assay (ELISA) and immunochromatography, can confirm infection by detecting antigens produced by S. flexneri.⁵ Mass spectrometry, PCR, and gene sequencing are precision techniques for S. flexneri detection.⁹ However, their application is limited due to the high cost, need for special tools, and complicated sample preparation.¹⁰ Therefore, it is necessary to develop a rapid, sensitive, and specific S. flexneri detection method. To overcome these problems, fiber optic biosensors have emerged, which are simple to prepare, easy to operate, and highly sensitive. It mainly works by receiving more evanescent waves (EWs) to stimulate (local) surface plasmonic resonance (SPR/LSPR) phenomena. Fiber optic biosensors based on LSPR are widely used in various fields of chemical and biological sensing.¹¹ Recently, Mendes et al. proposed a method of SPR induced by gold nanoparticles.¹² Compared to the typical SPR, the sensing performance is greatly improved. Moslemi et al. proposed an LSPR-based fiber optic sensor that is able to excite the LSPR mode of a single AuNPs, as well as the aggregation state of nanoparticles or multiple layers, further improving the sensitivity of the sensor.¹³ 

In recent years, the development of fiber-based biophotonic has rapidly grown in different fields, such as medicine,¹⁴ food safety,¹⁵ and environmental detection.¹⁶ In 2014, Xiao et al. proposed a portable evanescent-wave fiber-optic biophotonic sensor for Shigella detection and compared it to real-time fluorescent quantitative PCR, indicating its potential substitutability.¹⁷ Furthermore, in 2019, Kaushik et al. used SPR-based MoS₂ nanosheet (NSs)-modified optical fibers to detect E. coli with a sensitivity of 0.6 nm/(CFU/ml).^18,19 In 2021, Kumar et al. achieved detection of Shigella using multi-core optical fibers based on LSPR.²⁰ In 2023, Qiu et al. used tapered optical fibers to detect Salmonella typhimurium by coating antibodies directly on the fibers.²¹ Based on the high electronic communication and good biocompatibility properties of ZnO nanomaterials,²² Xia et al. developed a biophotonic sensor for glucose detection.²³ AuCu nanoclusters (NCs) have excellent biocompatibility as well as unique optical properties and can be used for signal transduction in biosensors. The tunable LSPR of AuCu–NCs enables precise control of the detection wavelength and signal enhancement, helping improve the sensitivity of biosensing applications.²⁴ In recent years, many optical fiber immunosensors based on nanomaterial surface modification have been reported to improve the sensitivity and biocompatibility of the sensing platform.²⁵ Among them, MoS₂ has large surface volume ratio, high light absorption efficiency, thermal stability, and the existence of free hydrophobic sulfur groups. These two-dimensional nanosheets provide a higher binding site for antibody immobilization and are, therefore, widely used in fiber optic biosensors. ZnO nanowires (NWs) and precious metal nanoparticles form a 3-D array that increases the sensing region and reduces light loss by capturing light signals from the sensing region.^26,27 Compared to aptamers, antibodies have a higher affinity for binding to target molecules, allowing for sensitive detection and analysis. Therefore, we chose anti-S. flexneri as the specific binding of S. flexneri to achieve better sensing performance.

In this paper, a three-S-tapered (3S) fiber-optic biophotonic sensor is developed for label-free, characteristic, and fast detection of S. flexneri. First, the optical field distribution of the 3S structure and single S-tapered structure is simulated by RSoft software, and the superiority of the 3S structure is verified theoretically. Second, in order to improve the performance of the biophotonic sensor, MoS₂ nanosheets (NSs), ZnO NWs, and AuCu NCs are used. MoS₂ NSs and ZnO NWs form a 3D spatial structure, which provides more adhesion points for AuCu. At the same time, their optical and electronic properties promote the LSPR of AuCu. Then, the performance of the sensor and the possibility of detecting S. flexneri biofilm in different environments are discussed, as shown in Fig. 1. The experimental results show that the sensor has a wide range of S. flexneri detection capabilities.

Single-mode fiber (SMF, 8.2 μm, 125 μm, EB-Link Technologies Co., Shenzhen) is used to fabricate sensor probes with a 3S structure. N-methyl-2-pyrrolidone (NMP) is the solvent used to dissolve MoS₂. S. flexneri was purchased from Shanghai, and anti-shigella flexneri type II was purchased from Jinan Osemer. The FITC-labelled aptamer (5′ CCGGACTAGGGCTGGTTAGCTTCAATACTGCTGGGCGAGG-3′) for marking Shigella flexneri was purchased from Shanghai Generay Biotechnology Company, Shanghai. Tryptic soy broth soybean-casein digest medium (TSB) and tryptic soy agar soybean-casein digest agar (nTSA) were purchased from Shanghai BD Medical Device Co., Ltd. for bacterial culture. Acetone, H₂O₂, sulfuric acid solution (H₂SO₄), ultrapure water, (3-mercaptopropyl)trimethoxysilane (MPTMS), 3-(trimethoxysilyl)propyl methacrylate solution (silane agent), ethanol, and nitrogen gas are used for coating (immobilization) of MoS₂-NSs, ZnO-NWs, and AuCu NCs over the 3S fiber structure. In addition, the necessary reagents are used in the process of attaching special modifiers to the surface of the probe, such as phosphate-buffered saline (PBS), 11-mercaptoundecanoic acid (MUA), N-hydroxysuccinimide (NHS), and ethyl (dimethylaminopropyl) carbodiimide (EDC). Other commonly used reagents, such as TMB (3,3′,5,5′-tetramethylbenzidine) are obtained directly from local chemical suppliers. All the aqueous solutions used in the experiments are prepared with deionized water (DI water) with a specific resistance of 18 MΩ cm obtained from the Milli-Q system.

In this study, the following high-precision testing instruments are used to obtain more accurate experimental results. First, a 3SAE combiner manufacturing system (CMS, USA) was used to fabricate and analyze a well-structured optical fiber. A UV–visible (UV–vis) spectrophotometer (U-3310, Hitachi, Japan) was used to determine the photometric value of S. flexneri solution. A high-resolution transmission electron microscope (HR-TEM, Talos L120C, Thermo Fisher Scientific) was used to detect the morphology of the different nanomaterials. The morphology of the nanomaterial coating on the surface of the sensing probe was observed by using a scanning electron microscope (SEM, Carl Zeiss Microscopy, Germany), and the elemental composition of the probe surface was determined by SEM-energy dispersive spectroscopy (EDS). Fluorescence spectra of synthetic AuCu–NCs were detected using the steady-state/transient + fluorescence test system (UC920, Edinburgh Instruments Ltd., UK), and the results of S. flexneri counting and aptamer experiments were validated using a fluorescence microscope. In addition, the detection device utilizes a tungsten–halogen light source (HL-2000, Ocean Optics Inc., USA), a USB2000+ spectrometer (Ocean Optics Inc., USA), and a computer.

The 3S structure designed in this paper is fabricated using a combiner manufacturing system (CMS) machine: a CMS machine with a semi-vacuum working environment and a unique three-electrode plasma heating technology, equipped with a dedicated computer operating system for control. Inside, there are two independent stretching platforms and a high-precision data ruler control shaft. There are two high-definition cameras at the front and bottom to observe the working process in real time. CMS has unique advantages in optical fiber processing; in addition to fiber taper, on-line precision cutting, and fusion functions, there is a convenient visual interface and high reproducibility. The S-tapered structure is obtained by applying opposite non-axial forces to the left and right sides of the fiber after the fiber is heated and softened. Therefore, there is a need to set the computer’s CMS machine’s left and right platforms at the y-axis dislocation distance. We also need to set the lifting speed, electrode heating power, platform lifting speed, length of the taper, and diameter. Then, remove the coated fiber from the CMS machine and click on the computer to run. Many experiments are required to fabricate the required optical fiber structure. The 3S structure in this paper is three S-tapered cascades of the same length and different diameters. So, it is necessary to get the parameters of S-tapered with different diameters to make a 3S structure. First, an S-tapered with a diameter of 80 μm is drawn in the optical fiber, and then, the electrodes are translated to a certain distance. Just input the value of the S-tapered with a diameter of 60 μm in the program, click and run, and then, we can draw the double S-tapered cascade with a diameter of 80 and 60 μm. By this operation, the S-tapered with a diameter of 40 μm can be drawn to fabricate a 3S structure.

The fiber probe was immersed in acetone for 20 min to remove most of the organic impurities and dust on the probe. Due to the need to thoroughly clean the probe surface, the probe was then immersed in a “Piranha solution” (H₂SO₄:H₂O₂ = 7:3) for 30 min. The probe is then cleaned with DI water and dried in an oven at 60° for 30 min. After cleaning, the probe was immersed in a 10 ml MoS₂-NSs solution for 20 s, dried in the oven for 2 min, and repeated eight times. The MoS₂ coated probe was immersed in a solution of ZnO-NWs for 10 min and then in a 70° oven for 30 min, and the procedure was repeated three times. MoS₂ and ZnO coated probes were placed in a 1% ethanol MPTMS solution (0.1 ml of MPTMS stock mixed with 9.9 ml of ethanol solution) for 12 h. Then, wash with alcohol and blow dry with nitrogen. The probe was immersed in an AuCu NCs solution for 48 h. AuCu NCs immobilized probes were washed with ethanol and dried with nitrogen gas.

The antibody was fixed as follows: the probe was immersed in MUA ethanol solution (10 ml, 0.5 mM) for 5 h of carboxylation. Then, 10 ml of EDC aqueous solution (200 mM) and NHS aqueous solution (50 mM) were immersed for 30 min to activate the carboxyl group. The probe was washed with DI water and immediately placed in the newly prepared antibody solution for 2 h, keeping the liquid at a low temperature. Remove the solution and soak in 1% milk powder for 10 min before washing to block the non-specific binding to the fiber. Clean with a 1X PBS solution and dry at room temperature. Store in a −20 °C refrigerator and soak in PBS for 15 min before use. The process of sensor probe functionalization is shown in Fig. 2.

In this process, the bacteria-related instruments were sterilized with an autoclave for 15 min before use, and the operation was also carried out in the sterile fume hood. TSB nutrient solution: weighed 7.5 g TSB dissolved in 250 ml DI water (500 ml flask), heated to be completely dissolved. TSA Petri dish: Say 10 g of TSA dissolved in 250 DI water, heated to boiling after 1 min. After high temperature sterilization, pour 50 ml of TSA solution into each dish, cool, solidify, seal, and then put into the refrigerator at 4 °C. The TSB nutrient solution (1 ml) in ampoules was transferred by using a pipette gun to the vials containing bacteria, and the bacteria were distributed evenly in the TSB solution by repeated blowing and sucking. A total of 1 ml of S. flexneri solution in a penicillin flask was added to two TSA dishes, and 100 μl of bacterial solution was added to each dish. The ring is then smeared with the bacterial solution until slightly dry.

Each flask containing 50 ml of TSB was filled with 400 μl of S. flexneri solution. The culture dish and small flask with bacteria is put in the incubator for 24 h, with the incubator temperature set at 36 °C and 140 rpm. Here is the first generation of bacteria. In order to better the experimental results, we need to re-culture with the following: take two previously bacterial-free TSA dishes and drop 0.1 ml of the first generation of bacteria into each dish; fill two new flasks with 25 ml of TSB solution and drop 1 ml of bacterial solution into each flask; put in the incubator one night at a temperature of 37 °C, 140 rpm; and here is the second generation of bacteria, used for experiments.

According to the theory of electromagnetic wave continuity, even though total internal reflection (TIR) occurs,²⁸ the energy of the incident wave does not return to the optical dense medium completely, and some light waves continue to propagate in the optical sparse medium, and the amplitude of the optical signal in this direction decreases exponentially with the transmission depth. Such transmitted waves produced by TIR in optical sparse media are called evanescent waves (EWs).²⁹ When incident light irradiates the metal nanoparticles (MNPs) surface, free electrons on the MNPs surface interact with photons to produce collective oscillations.³⁰ This collective oscillation reaches its maximum amplitude at a specific frequency called the “resonant frequency,” resulting in LSPR. Therefore, the electromagnetic field in the nano-scale region near the surface of the nanoparticles (NPs) is significantly enhanced, and the absorption and scattering of light by the NPs are the strongest.¹¹ When the RI changes in the environment around the MNPs, the resonance wavelength changes.³¹ 

Fabrication of an S-tapered with a gradual change in diameter structure on SMF causes the light in the fiber core to leak into the cladding due to fiber bending and diameter reduction.³² The light in the cladding excites the LSPR at the interface between the noble MNPs and the cladding. When the amount of analytes (bacteria in our case) to be measured on the surface of MNPs changes, the resonance wavelength will change.³³ The light that escapes into the cladding via the S-tapered excites the higher-order modes, and after propagation for some distance, part of the cladding light is coupled back into the core at the second bend of the S-tapered fiber structure. Therefore, the change in concentration of bacteria can be detected by using a spectrometer, so that analysis and monitoring of the measured substance can be realized. A schematic of the 3S structure is shown in Fig. 3(a).

We simulate and study the performance of the 3S fiber structure using the beam propagation method within RSoft software. Each S-tapered length is set to 4 mm in simulation, which corresponds to the actual S-tapered length. The RI of the fiber cladding is set to 1.4447, and the RI of the fiber core is set to 1.4504. Figure 3(b) shows an S-tapered with a diameter of 40 μm, and panel (c) shows a 3S with a diameter of 80, 60, and 40 μm cascades. When the light wave propagates in the S-tapered, the light in the core at the first bend of the S-tapered leaks into the cladding and excites the higher-order mode, and then, the core is coupled back at the second bend of the S-tapered. “Mode” is the field distribution curve of the first four modes representing the cladding, which is consistent with the theory. “Launch” represents the field in the core, and only a fraction of the light that leaks into the cladding can be coupled back into the core. Compared to the energy curve of the single S-tapered structure, the curve of the 3S structure at the third S-tapered fluctuates more, which indicates that the light in the core and cladding has a stronger interaction. In other words, 3S is more sensitive to changes in the environment. The 3S structure optimizes the defect of insufficient light energy in the front cladding of single S-tapered. With stronger evanescent field, it is more suitable for an optical fiber biophotonic sensor. Therefore, the 3S fiber-based WaveFlex structure designed in this experiment has better sensing performance than the single S-tapered structure.

AuCu NCs was synthesized by using the method described by Chen et al.³⁴ Penicillamine solution (30 mM, 1.8 ml) was mixed with HAuCl₄ (10 mM, 0.1 ml) in a magnetic stirrer (1000 rpm), shaken vigorously. The nitric acid solution (0.1M, 0.1 ml) containing 5 mM Cu(NO₃)₂ was then added, and after 1 min of vigorous shaking, the color of the solution changed from brown to milky white rapidly, indicating the formation of AuCu NCs. To obtain repeatable results, the solution was then shaken in the dark for 2 h at room temperature (25 °C).

To remove the excess penicillamine ligand, the AuCu NCS solution was centrifuged at a relative centrifugal force of 4000 rpm. The collected particles were re-dispersed in water (20 ml). For simplicity, we express the concentration of the prepared AuCu NCs as 1X. AuCu–NCs remained stable for two months at room temperature. Similarly, MoS₂-NSs were extracted by ultrasonic-assisted liquid stripping with NMP as an organic solvent. MoS₂-NSs (30 mg) were added to NMP solution (10 ml) in a tube, and an ultrasonic machine was used in an ultrasonic bath (power 500 W) for 1 h. Then, the MoS₂ suspension was centrifuged at 5000 rpm/min for 1 h. The supernatant is collected and stored in a glass jar for later use. The suspension of ZnO-NWs was prepared, and 10 mg of solid ZnO-NWs was put into a 10 ml anhydrous ethanol solution.

The schematic representation of the experimental setup is shown in Fig. 1. The optical signal from the light source is transmitted to the spectrometer by using the sensing probe, which is propagated in the optical fiber, as shown in Fig. 1. The spectrometer transmits the signal to the computer through a data line, and the computer converts the signal into data for processing.

Synthesized nanomaterials were characterized by the SPR absorption peak wavelength of the nanomaterials solution and photographed for microscopic particle to identify the shape and size of different NPs. Figures 4(a) and 4(b) show the MoS₂-NSs and ZnO-NWs sample solutions captured using HR-TEM, respectively. The morphology of MoS₂ shown in Fig. 4(a) is thin and uniformly distributed in solution, while that of ZnO-NWs shown in Fig. 4(b) is distributed like cylinder in solution. The result showed that MNPs in the graphs have uniform distribution and good morphology, which meet the experimental requirements. The HR-TEM images in Fig. 4(c) show that synthesized AuCu NCs are spherical in shape with uniform distribution thus indicating the formation of well-dispersed AuCu NCs. Figure 4(d) shows the fluorescence spectra of AuCu NCs recorded using a fluorescence test system with a peak wavelength of 614 nm³⁴ at an excitation wavelength of 270 nm. The mean size of the synthesized AuCu NCs was determined to be 3.23 (±0.73) nm, as shown in Fig. 4(e). Thus, the above-mentioned result indicates the synthesis of well-dispersed and uniform shaped MNPs.

CMS is an important equipment for a fabricated 3S structure, and its three-electrode plasma heating plays an important role in fiber fabrication by setting the parameters of CMS, such as electrode power, motor speed, taper length, and diameter; the structure of optical fiber is optimized. The whole tapering process can be seen through the computer screen and through the scanning function to determine whether the structure of the manufactured fiber morphology is qualified for sensing applications. Figure 5(a) shows a 3S diameter scan of five optical fibers. As can be seen from the diagram, the diameters of the three cascade S-tapered are 80, 60, and 40 μm, respectively. Five fibers have the same taper position and smooth diameter curve, which shows that the fiber structure prepared by using this method has good repeatability and good morphology. Even with a special fiber structure with the same appearance, its transmission intensity is not necessarily the same. Therefore, we have tested the transmission intensity of five 3S structured fibers, and the results are shown in Fig. 5(b). The transmission intensity of all fibers is highly coincident, and their peak wavelength is almost same, which indicates that the fabricated fiber structure has high repeatability.

SEM has been used to confirm the immobilization of nanomaterials over sensor probe. Figure 6(a) shows the structure of a fabricated sensor probe, and Fig. 6(b) shows the uniform distribution of MoS₂-NSs over the fiber surface. Thereafter, it is cleared from Fig. 6(c) that ZnO-NWs has been successfully immobilized on the surface of the optical fiber with an intact morphological structure. Figure 6(d) shows the coating of AuCu NCs along with MoS₂-NSs and ZnO-NWs. AuCu NCs is not visible due to its small size. However, the presence of AuCu NCs was confirmed by EDS, as shown in Fig. 6(e). Among them, Au, Cu, Mo, and Zn are the main elements of the three nanomaterials, while Si and O are the elements of the optical fiber, and C is the main component of the organic material used in the fixed nanomaterials. Therefore, from the above-mentioned results, it can be proved that all the nanomaterials are well immobilized on the fiber surface. These nanomaterials provide an enhanced surface area for antibody attachment and enhance the sensing ability of optical fiber sensors.

Stability testing is usually used to investigate the accuracy of a sensor. Continuous measurements were carried out to determine whether the performance of the sensor remains stable over time. The sensor was immersed in a 1X PBS solution, and after the response was completed, the peak wavelength of the measured signal was recorded. This was repeated fifteen times, and the experimental results are shown in Fig. 7(d).

It can be seen from the graph that the peak wavelength fluctuates only once in the process of many measurements that proves that the sensor has good stability.

The viability of S. flexneri varies across different pH environments, which can impact the performance of the sensor differently depending on the pH conditions. The sensor was then used to test S. flexneri in various pH environments and identify which pH environment is most suited for the bacterium to live in. The pH 3 and 6 solvents were diluted with strong hydrochloric acid, the pH 7.4 solvents with 1X PBS, and the pH 10 and 13 solvents with potassium hydroxide. The peak wavelengths of S. flexneri at the lowest (1 CFU/ml) and maximum (10⁸ CFU/ml) concentrations were determined under different pH environments. Subtracting the peak wavelength of a low concentration from the peak wavelength of a high concentration of S. flexneri gives the peak wavelength shift at various pH values, as shown in Fig. 7(e). The results demonstrated that the sample's most significant change occurred at pH = 7.4. At a pH of 7.4, bacteria are more alive and bind more strongly to the sensor’s antibodies. As a result, the sensor performs better when detecting S. flexneri in a solution with a pH of 7.4.

Reproducibility relates to whether different sensor probes can conduct the same test, whereas reusability refers to whether the same sensor probe can take many readings. Both serve as the foundation for evaluating sensor accuracy and dependability, as well as crucial variables in determining the realistic nature of sensor applications. To assess the repeatability of the produced sensor probes, the same concentration of S. flexneri (10⁸ CFU/ml) was analyzed with two separate sensor probes. As shown in Fig. 7(f), the transmission intensities of the two sensor probes are of the same wavelength and the constructed sensor is very reproducible. At the same time, the same concentration of S. flexneri solution was tested using the same probe at different times, and the fiber was thoroughly washed with 1X PBS solution between each test and then the antibody functionalization process was repeated. As shown in Fig. 7(g), the transmission intensity of the same probe at 10⁴ CFU/ml concentration at different times was highly consistent. It shows that the developed sensor can be reused many times without affecting the results.

Selectivity means that the sensor only responds to specific objects in a complex environment, and the selectivity directly affects the accuracy and reliability of the sensor. In order to test whether the developed sensor only responds to S. flexneri, four other bacteria, S. aureus, E. coli, V. anguillarum, and S. sonnei, have also been tested. First, the peak wavelength of each bacterium at a concentration of 1 × 10⁰ CFU/ml was measured. Then, the peak wavelength of the bacterial solution was measured at a concentration of 1 × 10⁸ CFU/ml. Then, the corresponding peak wavelength shift for each bacterium has been recorded, as shown in Fig. 8(a). It is possible that the cross-reaction and non-specific binding of the S. flexneri antibody to S. aureus resulted in the large wavelength shift of S. aureus. However, the peak wavelength shift of S. flexneri was much larger than that of other bacteria. The presence of S. flexneri antibodies on the sensor probe is responsible for the selectivity of the sensor and, therefore, the sensor has the highest recognition of S. flexneri compared to other analytes. The results show that the developed sensor has high specificity and recognition ability for the S. flexneri analyte.

S. flexneri is transmitted through food or water, so consumption of contaminated food may lead to infection. After infection, it may cause diarrhea, with symptoms including diarrhea, abdominal pain, fever, nausea, and vomiting. Therefore, it is very important to detect whether the food is infected by S. flexneri. To assess the versatility of our developed sensor in S. flexneri detection, we conducted experiments with various food samples, including milk, fish, juice, and chocolate. Commercially available milk, juice, and chocolate were highly concentrated and required dilution with 1X PBS for testing. The fish solution is obtained by mashing raw fish and diluting it with 1X PBS. Then, five tubes of 1 × 10⁸ CFU/ml bacterial solution were centrifuged; the top layer of liquid was removed; and equal volumes of 1X PBS, diluted milk, fish, fruit juice, and chocolate solution were added, respectively. First, the peak wavelength of the probe in 1X PBS solution was measured, and then, the peak wavelength of each solution containing bacteria was measured. In real samples, the proposed sensor exhibited a recovery rate of ∼99% with a relative standard deviation (RSD) value of around 25%. Therefore, the sensor still has good detection ability in different environments.

In the field of bacterial sensors, temperature is an important factor affecting the sensitivity of the sensors. At different temperatures, we tested the S. flexneri at different temperatures. S. flexneri solution of 1 × 10⁸ CFU/ml at 4, 15, 37, 50, and 100 °C was prepared and measured from low to high. In the course of scientific experiments, the sensing probe’s surface underwent cleaning with 1X PBS before each S. flexneri temperature measurement. These experimental outcomes are shown in Fig. 8(b). At the same concentration, the wavelength shift of the sensor is the largest at 37 °C, that is, the sensor is the most sensitive to detect S. flexneri at 37 °C. The reason is that at 37 °C, S. flexneri and its antibody binding affinity is strong; S. flexneri is more easily recognized by the antibody.

In order to develop the alternative detection system using the developed sensor, the authors developed different bacterial concentration detection systems based on TMB and H₂O₂ to explore whether they can be used as an important tool to detect microbes such as S. flexneri. It has been well reported that in the presence of peroxidase mimetic nanozyme, colorless TMB can be converted to blue color oxidized product, which in turn can alter the RI of developed sensor. Thus, it can be hypothesized here that AuCu particles on the probe surface can oxidize TMB in the presence of H₂O₂, causing RI changes in the sensing region and thus wavelength changes. Higher concentration of bacteria present in the system can form a dense layer over the developed sensor, which in turn is not able to interact with TMB and hence caused reduction to the RI of developed sensor. The probe was coated with different concentrations of S. flexneri, and the same amount of TMB and hydrogen peroxide mixture was measured. Three probes were immersed in different concentrations of 10⁸, 10⁶ CFU/ml bacterial solution, and 1X PBS for 30 min. The probe was then washed with PBS and immersed in a mixed solution of TMB (1 mM) and H₂O₂ (2M). The peak wavelength of each probe was recorded after 5 min. After 30 min, the peak wavelength of each sensor probe was measured again. Figure 8(c) shows a wavelength shift plot of the peak wavelength recorded at 30 min per probe minus 5 min. When the concentration of bacteria was 10⁰ CFU/ml, many AuCu participated in the redox, and the experimental results showed that the peak wavelength shifted to the long wavelength. When the bacterial concentration was between 10⁶ and 10⁸ CFU/ml, AuCu covered by bacteria increased and AuCu involved in redox decreased, which showed a shift from peak wavelength to short wavelength. In addition, the greater the number of bacteria, the greater the wavelength shift. The results showed that sensor probes coated with more S. flexneri were more sensitive to the redox. Therefore, the detection system can become an important tool for detection of microbes, such as S. flexneri.

Bacterial biofilm is a sticky substance formed on the surface by a large number of bacteria that secrete polysaccharides and proteins. Because the bacteria are protected by biofilms, their resistance to antibiotics has increased. Moreover, bacterial biofilms can adhere to biological tissues or metal surfaces, which can lead to metal corrosion or the infection of organisms. Therefore, it is very important for human health and environmental protection to study and detect the formation of bacterial biofilm. The S. flexneri solution was diluted to the target optical density (OD) of 0.1, 0.5, and kept at a constant temperature of 37 °C. The peak wavelengths of the two sensors in the air were measured, and then, the probes were immersed in S. flexneri solution with photometric values of 0.1 and 0.5, respectively. The S. flexneri were allowed to form a biofilm on the surface of the fiber for 3 h. The probe was then removed, immersed in sterile water for 1 s and dried for 3 min, and the peak wavelength of the sensing probe was measured again. The probe was then immersed in 99% methanol for three minutes. Finally, the probe was immersed in a 0.2% crystal violet solution for 30 min, washed and dried with PBS, and the peak wavelength was measured. “After water” is the peak wavelength of the second measurement minus the peak wavelength of the first measurement, and “after pbs” is the peak wavelength of the third measurement minus the peak wavelength of the first measurement, as shown in Fig. 8(d). It is clear from the figure that the peak wavelength shifts of the probe with a dip photometric value of 0.5 is greater than that with a dip photometric value of 0.1. When methyl violet binds to the bacteria, it changes the refractive index of the probe’s surface, making the shift in peak wavelength more pronounced. Therefore, the sensor has the potential to detect biofilm.

The fluorescent aptamer of S. flexneri was used to verify that the number of bacteria bound to the probe surface was different in different concentrations of S. flexneri solution. Three concentrations of S. flexneri solution (10³, 10⁶, and 10⁸ CFU/ml) were prepared, and the probe was immersed in the solution for 30 min. The probe was then removed, washed with 1X PBS, and placed in an aptamer (1 μM) solution for 1 h. Finally, the probe was removed, the surface of the probe was cleaned with 1X PBS, and the results were observed with a fluorescence microscope. Figure 8(f) shows the SEM image of the probe at 10⁸ CFU/ml and Figs. 8(g), 8(h), and 8(i) show the fluorescence image of the probe at 10³, 10⁶, and 10⁸ CFU/ml, respectively. The sensor probe immersed in a high concentration of bacterial solution has more bacteria attached to the surface and more fluorescent aptamers bound to it, so the fluorescence intensity is stronger. In order to show the above-mentioned experimental results more clearly, the fluorescence intensities shown in Figs. 8(g), 8(h), and 8(i) were processed by using software. As shown in Fig. 8(e), the fluorescence intensity of the probe surface increased with the increase in bacterial concentration.

At present, an optical fiber-based biophotonic sensor has not been developed for the detection of S. flexneri. Therefore, the sensor is compared to the various principles of detection of Shigella reports. As presented in Table I, this sensor is compared in detail to other available reports in terms of materials used, detection range, and detection limits. In the aspect of material selection, the sensor adopts excellent materials to improve the detection effect of S. flexneri. In terms of detection range, the sensor shows a wide range of adaptability and can accurately detect different types of Shigella. The results show that the sensor has potential for the detection of S. flexneri and provides strong support for related research and application.

We have introduced a novel 3S fiber-based biophotonic sensor to identify S. flexneri, subjecting it to a comprehensive assessment of sensitivity, repeatability, reproducibility, and selectivity. Simulation techniques have illustrated the advantages of the 3S structure compared to the single S-tapered structure. Beyond the customary MoS₂-NSs and ZnO-NWs, we have employed novel AuCu NCs to improve the sensor’s sensitivity, a significant stride in optical fiber biosensing. Furthermore, we have evaluated the sensor’s performance in diverse environments, confirming its consistency across varied conditions, and assessed its capacity to detect S. flexneri biofilms. Employing an aptamer experiment and leveraging AuCu’s oxidation–reduction properties, we have augmented the TMB experiment, leading to improved S. flexneri detection. To summarize, our research marks the development of a novel fiber-optic sensor for S. flexneri detection, offering substantial academic contributions to this field.

This work was supported by the Double-Hundred Talent Plan of Shandong Province, China; Special Construction Project Fund for Shandong Province Taishan Mountain Scholars; Liaocheng University (Grant No. 318052341); the Science and Technology Support Plan for Youth Innovation of Colleges and Universities of Shandong Province of China (Grant No. 2022KJ107); and the Science and Technology Plan of Youth Innovation Team for Universities of Shandong Province (Grant No. 2019KJJ019). This work was also developed within the scope of the projects CICECO (Grant Nos. LA/P/0006/2020, UIDB/50011/2020, and UIDP/50011/2020) and DigiAqua (Grant No. PTDC/EEI-EEE/0415/2021), financed by national funds through the [Portuguese Science and Technology Foundation/MCTES (FCT I.P.)]. The research was co-funded by financial support from the European Union under the REFRESH – Research Excellence for Region Sustainability and High-tech Industries Project No. CZ.10.03.01/00/22_003/0000048 via the Operational Programme Just Transition. This work was also supported by the Ministry of Education, Youth, and Sports of the Czech Republic conducted by the VSB-Technical University of Ostrava, under Grant Nos. SP2024/081 and SP2024/059.

The authors have no conflicts to disclose.

Lucan Xiao: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Writing – original draft (lead). Ragini Singh: Conceptualization (equal); Methodology (equal); Validation (equal); Writing – original draft (equal); Writing – review & editing (equal). Jan Nedoma: Methodology (equal); Resources (equal); Validation (equal); Writing – review & editing (equal). Qinglin Wang: Methodology (equal); Writing – review & editing (equal). Feng-Zhen Liu: Methodology (equal); Writing – review & editing (equal). Daniele Tosi: Methodology (equal); Writing – review & editing (equal). Carlos Marques: Methodology (equal); Resources (equal); Validation (equal); Writing – review & editing (equal). Bingyuan Zhang: Resources (equal); Supervision (equal); Validation (equal); Writing – review & editing (equal). Santosh Kumar: Conceptualization (equal); Funding acquisition (equal); Resources (equal); Supervision (equal); Validation (equal); Writing – review & editing (equal).

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.