The chemicals and preparation methodologies have been thoroughly outlined in the previous study [25] and illustrated in Fig 1. We used the chemicals without any further purification as follows: Ammonium Heptamolybdate Tetrahydrate ((NH₄)₆Mo₇O₂₄.4H₂O, 99.0%, from Tianjin Chemical Reagent Factory, Tianjin, China), Thioacetamide (C₂H₅NS, 99.0%, from Shanghai Zhanyun Chemical Co., Ltd, Shanghai, China), Ethanol (C₂H₅OH, 99.5%, from Xilong Scientific Co., Ltd., Guangdong, China), and deionized (DI) water. Briefly, we used the hydrothermal method to prepare hybrid MoS₂ nanosheets. We dissolved and mixed two precursor chemicals of (NH₄)₆Mo₇O₂₄.4H₂O and C₂H₅NS in 20 mL of deionized water. After that, 20 mL ethanol was gradually added and stirred for 30 minutes. The solid product was transferred to a 200 mL Teflon-lined stainless-steel autoclave. The hydrothermal temperature was set at 180° C for 5 hours. After this process was done. The precipitation was collected by centrifugation at 5000 rpm, washed with DI water, and dried in a vacuum at 60°C for 3 hours.

The B. subtilis strain was obtained from the Microbiology and Genetics Lab at the Hanoi University of Science and Technology in Hanoi, Vietnam. Starting with 1.5 mL of an overnight B. subtilis culture grown in Luria Broth (LB) medium, the cells are first pelleted by centrifuging at 8,000×g for 5 minutes using a Hettich Mikro 200R centrifuge (Tuttlingen, Germany). The supernatant is discarded, and the resulting cell pellet is resuspended in 740 μL of TE buffer. Subsequently, 20 μL of 100 mg/mL Lysozyme is added to break down the cell wall, then an incubation occurred at 37°C for 30 minutes. Next, 40 μL of 10% SDS and 8 μL of Proteinase K (10 mg/mL) (all from Biobasic, Canada) are introduced, assisting in protein digestion and membrane disruption. After a further incubation at 56°C for 3 hours, 100 μL of 5 M NaCl and heated CTAB/NaCl (from Merck, Germany) at 65°C are added sequentially to promote DNA precipitation. After an incubation at 65°C for 10 minutes, the sample undergoes a chloroform: isoamyl alcohol extraction (from Sigma Aldrich) to separate the DNA from impurities. After centrifuging at 12,000×g for 10 minutes at room temperature, the aqueous phase containing the DNA is transferred to a new tube. This extraction step is repeated until no white protein layer is visible. The DNA is then precipitated using cold 100% ethanol (Merck, Germany) and incubated at -20°C for 2 hours overnight. After additional centrifugation at 12,000×g for 15 minutes at 4°C, the DNA pellet is washed with 50 μL of 70% ethanol to remove salts and other impurities. Once the pellet dries, it’s resuspended in the TE buffer for storage. Ideally, the isolated DNA should be stored at −20°C for future use. All the DNA utilized in this study was evaluated using the OD260/280 ratios using a DeNovix UV-Visible spectrometer (Model: DS-11 FX+), yielding results around 2.0, indicating the high purity of the DNA samples.

This study utilized an oligonucleotide probe with the sequence amine—5’-CCTACGGGAGGCAGCAGTAG-3’, complementary to B. subtilis DNA [8]. The probe was diluted to 30 nM in TE buffer in all measurements. DNA solutions were prepared by dissolving and diluting them in 1×TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). B. subtilis DNA was pretreated using the heating method, which involved heating the samples at 95°C for 30 minutes. For a specific test, designated concentrations of the probe and hybrid MoS₂ nanosheets were utilized to determine the sensors’ absorbance and photoluminescence (PL). We incorporated 900 μL of hybrid MoS₂ into a 10 mm cuvette, with TE buffer as the solvent. We then added 100 μL of the probe and gradually added 100 μL DNA to the cuvette to create concentrations ranging from 23.6 to 130 aM. At each step, PL measurements were performed. The fluorescence intensities at an excitation wavelength of white light, using a slit width of 300 μm and an exposure duration of 1 second, were recorded. To investigate the effects of the sensing materials, the experiment is repeated with varying concentrations of hybrid MoS₂ (10 mg/L, 20 mg/L, 30 mg/L, 40 mg/L, and 50 mg/L).

The produced materials were initially assessed for their structure and morphology using XRD and SEM imaging techniques. As illustrated in Fig 2A, these materials reveal a nanosheet structure. The XRD pattern, displayed in Fig 2B, presents five unique peaks at positions (101), (012), (015), (110), and (113), suggestive of the MoS₂-3R structure (PDF#17–0744, as analyzed with JADE software by MDI Materials Data). The resulting composite was also detected alongside MoS₂-3R, (NH₄)₆Mo₇O₂₄. However, the SEM image, shown in Fig 2A, emphasizes the existence of multilayer nanosheets within the hybrid material. This observation suggests that (NH₄)₆Mo₇O₂₄ plays a critical role in creating the lamellar MoS₂, with NH₄⁺ ions occupying the layers in between and indicates that (NH₄)₆Mo₇O₂₄ either functionalizes the MoS₂ surface or decomposes into molecules. Furthermore, we explored the optical properties of the manufactured materials. An absorbance peak at 235 nm and a photoluminescence peak at 558 nm can be observed in Fig 2C.

This study aims to design a simple optical sensing platform to detect a range of B. subtilis DNA concentrated from 23.6 to 130 aM, reflected by the number of copies of the testing sample from 16×10⁶ to 16×10⁷. In our experiment, the probe concentration was 30 nM (using 100 μL, which contains 1.83×10¹² copies, a much more considerable amount than the number of ssDNA copies) and the concentration of sensing materials was varied. The sensors’ fluorescence was examined at hybrid MoS₂ concentrations of 10 mg/L, 20 mg/L, 30 mg/L, 40 mg/L, and 50 mg/L, and was presented in Fig 3.

Fig 3A shows an example of PL spectra of sensors based on 30 mg/L hybrid MoS₂ nanosheets in contact with B. subtilis DNA. The other MoS₂ concentration sensors had similar shapes and quenching effects with DNA concentrations increased. They all had fluorescent peaks at 558 nm. The fluorescence spectra of all sensors based on 10 mg/L to 50 mg/L hybrid MoS₂ before adding DNA are presented in Fig 3B. Based on the intensity I and the ratio I/I₀ at the wavelength 558 nm, we established the calibration lines of I vs C and I/I₀ vs. C in Fig 3C & 3D, where I is the intensity of sensors with specific concentration of DNA; I₀ is the initial fluorescent intensity of sensors; C is the concentration of B. subtilis (aM). The operating functions were estimated and shown in Table 1. All sensors can operate linearly with high precision (all values of R² were about 0.98). Among them, the sensors based on 50 mg/L have the highest sensitivity with the highest slope of -325. However, the 30 mg/L MoS₂-based sensors have the best detection limit (LOD) of 18.7 aM and high sensitivity (slope of -233.8). In Table 1, the equation determined the limit of detection is LOD = blank signal + 3 standard derivations. Overall, the detection limits are slightly different, about 19 aM. That might be because the sensing material’s concentration is not much different (10 mg/L to 50 mg/L, only 40 mg/L difference).

In all experiments, the quenching effects have occurred and are reflected in the negative slopes of all operating functions. The quenching effect can be explained by the fact that in this study, we employed the NH₂-5′-CCTACGGGAGGCAGCAGTAG-3′ probe to identify the complementary target B. subtilis DNA, adhering to the Watson-Crick base-pairing principles [8]. The probe was chemically modified with an amine group (NH₂) to enhance bonding to the MoS₂ surface. In the prepared nanomaterials, NH₄⁺ ions occupying the layers in between either functionalize the MoS₂ surface or help to bind with NH₂- of the amine probe. Fig 4 provides a proposed schematic of the adsorbed single-stranded DNA (ssDNA) on the hybrid MoS₂ surface. The ssDNA can adhere to this surface, altering the dielectric properties of MoS₂. However, when the ssDNA hybridizes with its complementary DNA, the resultant double-stranded DNA (dsDNA) establishes poor contact with the hybrid MoS₂, distancing itself from the MoS₂ surface and changing the dielectric environment from DNA to water and reducing the photoluminescence [25]. The more complementary ssDNA added to the sensor, the more dsDNA was formed. Consequently, the photoluminescence was reduced more intensively. The intensity I at 558 nm, and the ratio of I/I₀ at 558 nm can be used to estimate the unknown B. subtilis DNA concentrations.

To validate the operational performance of this sensor, we prepared various DNA concentrations exceeding the limit of detection (LOD) and within the range defined by the calibration line. The concentrations examined were 74.3 aM, 97.5 aM, and 115.6 aM. The corresponding fluorescence measurements are depicted in Fig 5A. Utilizing the fluorescence values at 558 nm into the calibration line in Fig 5B, we determined the measured concentrations listed in Table 2. The calibrated concentrations were in good agreement with the experimental concentrations with small percentages of difference (smaller than 10%). When the test sample had a high concentration, the measured intensity was higher, and the precision improved. For example, with the sample of 115.6 aM, the difference between the actual concentration and the calibrated one was only 1.52%. The result confirms the reliability and repeatability of our proposed sensor.

In this section, the selectivity and stability of the proposed sensors are investigated. First, we recorded the fluorescence of B. subtilis DNA of concentration from 23.6 to 130 aM. The fluorescences when the proposed sensors were in contact with various TE buffer concentrations, were also measured. Fig 6A showed that the tested B. subtilis DNA concentrations had low fluorescence and were almost the same intensity of 1.1×10⁴. Hence, our proposed sensors’ fluorescence changes in contact with B. subtilis DNA were not due to the changes in B. subtilis DNA fluorescence. To confirm the feasibility and selectivity of the proposed sensors, we prepared sensors based on 30 mg/L hybrid MoS₂ nanosheets. These sensors were exposed to different analytes, including TE buffer, Vibrio proteolyticus DNA, and Escherichia coli DNA. The fluorescence spectra in Fig 6B–6D reveal that the spectra and the changes in intensity for TE buffer and mismatched DNAs (V. proteolyticus and E. coli DNA) were nearly identical and slightly different. The intensity was reduced from 4.7×10⁴ to 4.0×10⁴ for three analytes. In contrast, the intensity decreased significantly when the target was B. subtilis DNA from 4.7×10⁴ to 2.0×10⁴ (see Fig 3A). This suggests that the E. coli and V. proteolyticus DNA did not induce any changes and merely diluted the hybrid MoS₂ suspension, mirroring the effect of added TE buffer. This observation aligns with Fig 6E & 6F, where the intensity I or ratio I/I₀ at 558 nm for TE, V. proteolyticus, and E. coli overlapped and changed slightly. B. subtilis DNA concentration can be determined by the linear function, as shown in Fig 6, within a range of 23.6–130 aM with high precision (R² = 0.98).

For better visualization, the quenching effects were quantified by: Quenching(%)=I0−II0×100% (1)

Fig 7 illustrates the quenching effect of the sensors when 130 aM analytes compared to the sensors’ photoluminescence before adding analytes. In the case of "ONLY B. subtilis DNA" the reference photoluminescence was the TE buffer. The quenching percentage of B. subtilis DNA was -1.12%, which means when adding DNA, the fluorescence enhanced. This enhancement is reasonable because the more DNA added, the higher the fluorescence was. However, with the maximum concentration of DNA, the increased intensity was deficient. The non-complementary DNA and TE buffer induced almost the same quenching percentage of 13%, representing the same diluting effect of adding these analytes. Only B. subtilis DNA, as we discussed above, had a significant quenching (%) of 55%. These findings validate that the proposed fluorescence sensors are functional and offer high sensitivity, reliability, and selectivity.

Furthermore, we explored the stability of the proposed sensors by repeating the tests of PL measurements of 30 mg/L hybrid MoS₂-based sensors in contact with B. subtilis DNA every week over 4 weeks. The results of intensities at 558 nm depending on the DNA concentrations and quenching percentage were shown in Fig 8A and 8B, respectively. As observed from Fig 8A and 8B, the intensity of proposed sensors at 558 nm wavelength and quenching percentage were stable over four weeks. These results confirmed the high stability of our proposed sensors to B. subtilis DNA over time.

Our findings revealed that the relatively straightforward arrangement of an amine probe-hybrid MoS₂ nanosheets demonstrated significant promise for the detection of B. subtilis DNA. The proposed sensors offered impressive sensitivity, stability, and specificity, with a detection range of 23.6 to 130 aM and a detection limit of 18.7 aM for 30 mg/L hybrid MoS₂ nanosheet-based sensors. Notably, there’s been a lack of literature about Bacillus DNA sensors. The existing biosensors either have a higher detection limit, are built on costlier materials, and involve more intricate procedures. For instance, Fei Chen et al. designed optical biosensors to identify B. subtilis DNA with a detection limit of 10⁵ CFU/mL, utilizing a combination of alkaline phosphatase/graphene oxide nanoconjugates and D-glucose-6-phosphate-functionalized gold nanoparticles [26]. Ivan Magnrina’s team presented a novel dual electrochemical genosensor for simultaneously amplifying and detecting Bacillus anthracis DNA, with a detection limit of 0.8 fM [27]. Zahra Izadi formulated an electrochemical DNA-based biosensor for Bacillus cereus detection employing an Au-nanoparticle-modified pencil graphite electrode with a detection limit of 9.4×10⁻¹² M [28]. Mukhil Raveendran produced an electrochemical DNA biosensor to identify Bacillus anthracis, which leveraged a thiol probe anchored on gold-modified screen-printed electrodes and had a detection limit of 10 pM [29]. Furthermore, the utilization of hybrid MoS₂ nanosheets is still in its infancy and remains largely untapped. Based on the authors’ understanding, there’s no documented evidence of using the innovative hybrid-MoS₂ nanosheets and (NH₄)₆Mo₇O₂₄ materials to detect B. subtilis DNA. Our findings will likely pave the way for further research in pathogen detection applications, which are still in the nascent stages of development.