Facial Preparation of Cyclometalated Iridium (III) Nanowires as Highly Efficient Electrochemiluminescence Luminophores for Biosensing

In this study, highly efficient ECL luminophores composed of iridium complex-based nanowires (Ir–NCDs) were synthesized via covalently linking bis(2-phenylpyridine)-(4-carboxypropyl-2,2′-bipyridyl) iridium(III) hexafluorophosphate with nitrogen-doped carbon quantum dots (NCDs). The ECL intensity of the nanowires showed a five-fold increase in ECL intensity compared with the iridium complex monomer under the same experimental conditions. A label-free ECL biosensing platform based on Ir–NCDs was established for Salmonella enteritidis (SE) detection. The ECL signal was quenched linearly in the range of 102–108 CFU/mL for SE with a detection limit of 102 CFU/mL. Moreover, the relative standard deviations (RSD) of the stability within and between batches were 0.98% and 3.9%, respectively. In addition, the proposed sensor showed high sensitivity, selectivity and stability towards SE in sheep feces samples with satisfactory results. In summary, the excellent ECL efficiency of Ir–NCDs demonstrates the prospects for Ir(III) complexes in bioanalytical applications.


Introduction
Electrochemiluminescence (ECL) is the luminescent emission resulting from highly exergonic electron transfer reactions between electrochemically generated oxidised and reduced species in the vicinity of the electrode surface [1]. ECL has some significant advantages; for example, no external light source is required, so the optical background noise is extremely low, and the sensitivity is very high. ECL has become a very powerful detection technique for a wide range of bio-related species via the ability to use ECL luminophores as labels in antibody/antigen or nucleic acid assays [2,3]. As a result, the technique has been widely used in the areas of clinical diagnostics, and life science research [4]. Recently, iridium(III) complexes have been investigated as alternative ECL reagents to the commonly employed ruthenium(II) complexes, due to various advantageous properties such as high photoluminescence efficiency, easily tuned emission colour, and large stokes shifts [5,6]. It was reported that (pq) 2 Ir(acac) showed greater ECL intensity than Ru(bpy) 3 2+ with tri-n-propylamine (TPA) co-reactant under optimum conditions (pq = 2-phenylquinoline anion, acac = acetylacetonate anion) [7]. However, their application in bio-assays has been limited due to their poor aqueous solubility and the limited availability of derivatives with bio-conjugatable ligands. The in-depth studies of cyclometalated iridium (III) complexes and their application in bioassays are still a challenge.
To address these challenges, the research focus has been devoted to synthesizing cyclometalated iridium(III) complex-based materials with higher ECL efficiency and water the SE antibody to the film surface. In the presence of SE, the antibody specifically recognizes the antigen and forms an antigen-antibody complex, insulating the GCE surface and hindering the electron transport process, thereby quenching the ECL signal. This labelfree sensing strategy enabled the simple detection of SE with excellent selectivity and high stability.

Instruments
Transmission electron microscopy, used to image the morphologies of the synthesized nanomaterials, was performed using a Talos F200X (Thermo Fisher Scientific, Waltham, MA, USA) microscope. Ultraviolet-visible spectrometry was performed using a UVvis spectrometer (UV-2700, Shimadzu, Kyoto, Japan). A fluorescence spectrophotometer (Shanghai Cold Light Technology, Shanghai, China, F97pro) was used to obtain the fluorescence spectrum of raw materials and products. Scanning electron microscopy, used to take topography and mapping of materials, was performed using a Hitachi S-4800 SEM (Hitachi, Tokyo, Japan). The constructed sensor was tested by electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) using an electrochemical workstation (Xian Remai). ECL measurements were performed on an MPI-E analyzer (Xi'an Ruimai Analytical Instruments Co., Ltd., Xi'an, China) using a traditional three-electrode system.

Instruments
Transmission electron microscopy, used to image the morphologies of the synthesized nanomaterials, was performed using a Talos F200X (Thermo Fisher Scientific, Waltham, MA, USA) microscope. Ultraviolet-visible spectrometry was performed using a UV-vis spectrometer (UV-2700, Shimadzu, Kyoto, Japan). A fluorescence spectrophotometer (Shanghai Cold Light Technology, Shanghai, China, F97pro) was used to obtain the fluorescence spectrum of raw materials and products. Scanning electron microscopy, used to take topography and mapping of materials, was performed using a Hitachi S-4800 SEM (Hitachi, Tokyo, Japan). The constructed sensor was tested by electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) using an electrochemical workstation (Xian Remai). ECL measurements were performed on an MPI-E analyzer (Xi'an Ruimai Analytical Instruments Co., Ltd., Xi'an, China) using a traditional three-electrode system. In the three-electrode system, a glassy carbon electrode (GCE) was used for the working electrode. The sensor was constructed with a platinum wire counter electrode, and Ag/AgCl formed the reference electrode (saturated KCl). Solutions were adjusted to the correct pH with a 920 Precision pH meter.

Synthesis of NCDs and Ir-NCDs Nanowires
The amine-rich carbon dots (NCDs) were synthesized via the hydrothermal method according to previous work [30]. Briefly, the L-arginine and ethylenediamine were succes- sively dissolved in ultrapure water and mixed. This mixed solution was then poured into a sealed Teflon lined reaction chamber, and the solution was then heated for 12 h at the programmed temperature of 180 • C. During the heating process, the color of the solution changes from clear to yellow due to the formation of NCDs. The solution was then filtered using a filter membrane, and the product was further purified by dialysis (membrane cutoff Mw = 2 kDa) to remove the raw material and other products. The NCDs product was dried by freeze-drying and stored in a refrigerator.
Ir-NCDs products are obtained by amidation of the carboxyl group of the iridium complex with the amine group of NCDs. The iridium complex is added to the synthetic NCDs solution, and the solution is mixed via ultrasonication until it becomes clear and transparent. EDC/NHS was added to activate the carboxyl group, and then stirred at room temperature for 16 h to obtain an orange-red product. The product is called Ir-NCDs. As a control experiment, Ir complex and NCDs are simply mixed without the addition of EDC/NHS, negating the formation of the Ir-NCDs. This Ir/NCDs mixture is used to compare ECL signals with Ir-NCDs and confirm the ECL enhancement from the nanowire formation. The mass concentration of iridium atoms in the Ir-NCDs was determined by an inductively coupled plasma emission spectrometer (ICP), and the molar concentration of the iridium complex was calculated by a linear regression equation ( Figure S1 in Supplementary Materials).

Fabrication of ECL Sensor
The glassy carbon electrode was polished in turn with 0.3 µM and 0.05 µM alumina slurries, and then rinsed sequentially with ultrapure water, ethanol and acetone followed by ultrasonication before use. Then GCE electrodes were activated and further cleaned in 0.5 mol/L H 2 SO 4 solution by CV (scanning from 1.0~−1.0 V) until an ideal voltammogram was obtained (no further change in background current). To construct the sensor, Ir-NCDs and PDDA (1:1) were mixed and sonicated for 30 min, and then 10 µL of the mixed solution was drop-coated on the glassy carbon electrode. PDDA is a polyelectrolyte and acts to increase the stability of the Ir-NCDs, while its conductive property enhances the electron transport to the modified electrode. The electrode was then drop-coated with 10 µL of GA, which acts as an intermediate linker between the Ir-NCDs and the antibodies. Subsequentially, 5 µL of antibody solution was drop-coated on the GCE surface and reacted overnight in a refrigerator at 4 • C. Finally, 10 µL of BSA was dropped on the electrode surface to react for 2 h, blocking the unbound sites. Before analysis, different concentrations of salmonella were incubated with the modified electrode for 1 h.

Characterization of the NCDs and Ir-NCDs
The morphology of the NCDs and Ir-NCDs were characterized by TEM. The TEM image and the dark field image of the as-prepared NCDs showed spherical morphology with monodispersing properties (Figures 1A and S2). Dynamic light scattering analysis showed that the average diameter ranged from 2 to 5 nm with an average diameter of 3.0 ± 0.4 nm ( Figure S2). The high-resolution TEM image indicated a lattice spacing of 0.21 nm ( Figure S2), which was consistent with the in-plane lattice spacing of graphene (100 facet). UV-visible absorption of NCDs showed two characteristic peaks centered at 280 nm and 358 nm, which were attributed to the π-π* transition of C=C and the n-π* transition f C=O ( Figure S5).
The TEM image of the Ir-NCDs showed nanowires morphology due to the crosslinking between one iridium complex and multiple amine-rich Ir-NCDs ( Figure 1B). The enlarged images of Figure 1C,D further demonstrate that the NCDs can be successfully encapsulated into the nanofibers. The EDS mapping analyses were carried out to analyze the surface elements of the nanowires ( Figure 1E-I). As anticipated, the characteristic elements of Ir, C, N, and O were visible. The specific position and element distribution at spot 1 and spot 2 are shown in Figure S3. Additionally, the morphology of the iridium monomer in its solid state and solution was characterized by TEM ( Figure S4). The iridium complexes formed a crystal structure in the solid state and were spherical in shape in solution, which further confirmed that the iridium complex can form nanowires due to the existent of NCDs. In addition, the UV-visible and fluorescence spectra were recorded to further confirm the successful synthesis of Ir-NCDs ( Figure S5). The fluorescence spectrum of the Ir-NCDs showed two peaks at 430 and 580 nm, which was consistent with the fluorescence maximum emission wavelengths of NCDs and the Ir complex recorded separately. encapsulated into the nanofibers. The EDS mapping analyses were carried out to analy the surface elements of the nanowires ( Figure 1E-I). As anticipated, the characteristic e ments of Ir, C, N, and O were visible. The specific position and element distribution spot 1 and spot 2 are shown in Figure S3. Additionally, the morphology of the iridi monomer in its solid state and solution was characterized by TEM ( Figure S4). The iridi complexes formed a crystal structure in the solid state and were spherical in shape in lution, which further confirmed that the iridium complex can form nanowires due to existent of NCDs. In addition, the UV-visible and fluorescence spectra were recorded further confirm the successful synthesis of Ir−NCDs ( Figure S5). The fluorescence sp trum of the Ir−NCDs showed two peaks at 430 and 580 nm, which was consistent with fluorescence maximum emission wavelengths of NCDs and the Ir complex recorded s arately.

ECL Performance of Ir−NCDs and Feasibility Study
The ECL properties of the iridium monomer, Ir−NCDs (covalent bond) and Ir/NC (simple mixture) were compared. As shown in Figure 2A, the iridium complex alone hibits weak ECL emission. After the addition of NCDs to the iridium complex soluti the ECL intensity of Ir/NCDs is nearly two times higher than that of the monomer, a the ECL intensity of Ir−NCDs is further increased to five times that of the iridium compl These results suggest that the higher ECL efficiency of Ir−NCDs may be attributed to contribution of the covalently bound NCDs. NCDs are a new type of quantum dots, wh are environmentally friendly and contain a large number of amine groups that can be us as a co-reactant reagent with the iridium complex, thereby enhancing its ECL emissi After the amidation reaction, the electron transport distance between NCDs and the ir ium complex was shortened, and the ECL intensity of the nanowires was improved s nificantly. The mechanism of ECL emission is shown in Figure S6.
The effect on the ECL of the stepwise construction of the sensor was studied by tak SE as an example. The results are shown in Figure 2B. First, the mixed solution of Ir−NC and PDDA was modified on the electrode to obtain relatively high ECL signal, and th glutaraldehyde (GA) was modified on the electrode, which resulted in a decrease in

ECL Performance of Ir-NCDs and Feasibility Study
The ECL properties of the iridium monomer, Ir-NCDs (covalent bond) and Ir/NCDs (simple mixture) were compared. As shown in Figure 2A, the iridium complex alone exhibits weak ECL emission. After the addition of NCDs to the iridium complex solution, the ECL intensity of Ir/NCDs is nearly two times higher than that of the monomer, and the ECL intensity of Ir-NCDs is further increased to five times that of the iridium complex. These results suggest that the higher ECL efficiency of Ir-NCDs may be attributed to the contribution of the covalently bound NCDs. NCDs are a new type of quantum dots, which are environmentally friendly and contain a large number of amine groups that can be used as a co-reactant reagent with the iridium complex, thereby enhancing its ECL emission. After the amidation reaction, the electron transport distance between NCDs and the iridium complex was shortened, and the ECL intensity of the nanowires was improved significantly. The mechanism of ECL emission is shown in Figure S6.
The effect on the ECL of the stepwise construction of the sensor was studied by taking SE as an example. The results are shown in Figure 2B. First, the mixed solution of Ir-NCDs and PDDA was modified on the electrode to obtain relatively high ECL signal, and then glutaraldehyde (GA) was modified on the electrode, which resulted in a decrease in the ECL signal measured. Further modification of the modified layer with antibodies and BSA resulted in further reductions in the ECL signal due to the quenching effect of protein molecules. Finally, the specific binding of SE with the antibody further quenches the ECL, depending on the concentration of SE, indicating that the constructed ECL quenching sensor can realize the sensitive detection of SE.
Biosensors 2023, 13, 459 6 of 11 ECL signal measured. Further modification of the modified layer with antibodies and BSA resulted in further reductions in the ECL signal due to the quenching effect of protein molecules. Finally, the specific binding of SE with the antibody further quenches the ECL, depending on the concentration of SE, indicating that the constructed ECL quenching sensor can realize the sensitive detection of SE.

Electrochemical Characterization of the Immunosensor
To investigate the assembly process of the immunosensor electrochemically, cyclic voltammetry was performed in 0.  Figure 3A, a decreased current response was found for Ir−NCDs/PDDA, indicating that the current was suppressed due to the electron-blocking effect of the Ir−NCDs/PDDA layer. After modification with GA, the current signal was further decreased, which is mainly due to the decreased electron transfer of the redox probe on the electrode surface. Weaker peak currents were observed after modifying with Ab. This is due to the successful assembly of the proteins on the electrode, which results in an additional electron-blocking layer of protein. After BSA modification on the electrode, the current value decreases further because BSA blocks the unreacted active sites. After the capture of SE at the electrode surface, a further subsequent decrease in current was observed due to insulting blocking of protein molecules by SE, which illustrates that the immunosensor for detection of SE was successfully constructed.
To further confirm the successful assembly of stepwise surface modification on the electrode surface ( Figure 3B [31], the equivalent circuit (as shown in the inset), was applied as of model to analyze the EIS spectra. The equivalent circuit included the electrolyte solution resistance Rs, the surface electron transfer resistance Ret, the Warburg impedance Zw and the double layer capacitance Cdl. The diameter of the semicircle of the Nyquist curve reflects the electron transfer resistance (Ret) at the electrode interface. Compared with the bare glassy carbon electrode (GCE, curve a), a small semicircle portion of the curve was observed for Ir−NCDs/PDDA. After modification with GA, the Ret increased due to the layer of GA. The assembly of Ab on the electrode results in a barrier to the interfacial charge, which further increases Ret. Further stepwise increases in the semicircle portion of the curve were observed after modification with BSA, likely originating from the partial insulation of the electrode by the BSA protein. After the incubation of the electrode with SE and subsequent binding of SE to Ab, the Ret is further increased, due to

Electrochemical Characterization of the Immunosensor
To investigate the assembly process of the immunosensor electrochemically, cyclic voltammetry was performed in 0.1 M PBS containing 5.0 mM K 3 [Fe(CN) 6 ]/K 4 [Fe(CN) 6 ] and 0.1 M KCl. As shown in Figure 3A, a decreased current response was found for Ir-NCDs/PDDA, indicating that the current was suppressed due to the electron-blocking effect of the Ir-NCDs/PDDA layer. After modification with GA, the current signal was further decreased, which is mainly due to the decreased electron transfer of the redox probe on the electrode surface. Weaker peak currents were observed after modifying with Ab. This is due to the successful assembly of the proteins on the electrode, which results in an additional electron-blocking layer of protein. After BSA modification on the electrode, the current value decreases further because BSA blocks the unreacted active sites. After the capture of SE at the electrode surface, a further subsequent decrease in current was observed due to insulting blocking of protein molecules by SE, which illustrates that the immunosensor for detection of SE was successfully constructed.
Biosensors 2023, 13, x FOR PEER REVIEW 7 of 12 the relatively large particle size of Salmonella enteritidis. Thus, the transmission of electrons is easily hindered, and the current value also decreases significantly. The EIS spectrum was consistent with the CV results, which demonstrates the successful immobilization of each modification step.

Optimization of Experimental Conditions
To obtain the best performance of the immunosensor, the pH of PBS, scan rate, amount of PDDA and ratio of the nanowires and PDDA, reaction time, and temperature According to previous reports [31], the equivalent circuit (as shown in the inset), was applied as of model to analyze the EIS spectra. The equivalent circuit included the electrolyte solution resistance Rs, the surface electron transfer resistance Ret, the Warburg impedance Zw and the double layer capacitance Cdl. The diameter of the semicircle of the Nyquist curve reflects the electron transfer resistance (Ret) at the electrode interface. Compared with the bare glassy carbon electrode (GCE, curve a), a small semicircle portion of the curve was observed for Ir-NCDs/PDDA. After modification with GA, the Ret increased due to the layer of GA. The assembly of Ab on the electrode results in a barrier to the interfacial charge, which further increases Ret. Further stepwise increases in the semicircle portion of the curve were observed after modification with BSA, likely originating from the partial insulation of the electrode by the BSA protein. After the incubation of the electrode with SE and subsequent binding of SE to Ab, the Ret is further increased, due to the relatively large particle size of Salmonella enteritidis. Thus, the transmission of electrons is easily hindered, and the current value also decreases significantly. The EIS spectrum was consistent with the CV results, which demonstrates the successful immobilization of each modification step.

Optimization of Experimental Conditions
To obtain the best performance of the immunosensor, the pH of PBS, scan rate, amount of PDDA and ratio of the nanowires and PDDA, reaction time, and temperature were optimized. The corresponding results are shown in Figure 4. Firstly, the pH of PBS (0.10 M) was optimized. An increasing ECL signal from the Ir-NCDs nanowires was observed with the pH increasing from 5.0-7.5. When the pH surpassed 7.5, the ECL signal dropped gradually. Therefore, pH 7.5 was chosen as the optimal pH.

Performance of Proposed Biosensor
Based on the optimum conditions, the analytical performance of the immunosensor for detecting SE with various concentrations was investigated. As shown in Figure 5A, with the increase in SE concentration, the ECL from the Ir−NCDs at the electrode surface decreased, resulting in a quenching of the ECL signal. Figure 5B shows the linear relationship between the concentration of SE and the ECL intensity. The linear fitting equation was y = 14,760-1423 lgC, with a detection limit of 1 × 10 2 CFU/mL (S/N = 3). Additionally, the LOD and linear range of detecting SE in this work was compared with various methods reported in the literature (Table S1), confirming that the proposed method showed The effect of PDDA with different mass fractions (0.5%, 1%, 2%, 3%, 4%) on the ECL signals was studied. As shown in Figure 4B, the ECL signals decreased obviously with the increasing PDDA mass fraction. Therefore, 0.5% was chosen as the optimal mass fraction. As shown in Figure 4C, the ratio of PDDA and Ir-NCDs was also optimized. The responses of ECL signals increased with the ratio of PDDA and Ir-NCDs from 1:3 to 3:1, At the ratio of 1:1 for PDDA and nanowires, the immunosensor achieved the maximum ECL intensity. Therefore, Ir-NCDs:PDDA = 1:1 was determined as the optimum ratio.
As shown in Figure 4D, the ECL signal decreased significantly with increasing SE binding time. These results indicate that more SE was specifically bound to Ab as the incubation time increased. When the reaction time reached 60 min, the ECL signal remained stable, indicating that the reaction was basically saturated at 60 min. Therefore, 60 min was selected as the best incubation time. In addition, it can be seen from Figure 4E that the ECL signal also changes with the incubation temperature. Considering the activity of the antigen at a certain temperature, the incubation temperature was finally selected to be 37 • C.

Performance of Proposed Biosensor
Based on the optimum conditions, the analytical performance of the immunosensor for detecting SE with various concentrations was investigated. As shown in Figure 5A, with the increase in SE concentration, the ECL from the Ir-NCDs at the electrode surface decreased, resulting in a quenching of the ECL signal. Figure 5B shows the linear relationship between the concentration of SE and the ECL intensity. The linear fitting equation was y = 14,760-1423 lgC, with a detection limit of 1 × 10 2 CFU/mL (S/N = 3). Additionally, the LOD and linear range of detecting SE in this work was compared with various methods reported in the literature (Table S1), confirming that the proposed method showed higher sensitivity than other methods, with an excellent detection range.

Application of ECL Bacterial Sensor in Real Samples
To verify that the prepared bacterial sensor has good reliability in practical application, the SE concentration in sheep feces was quantified by the proposed immunosensor. The recovery test was performed by a standard addition method. A 1 g sample of sheep excrement was ultrasonically dispersed into 5 mL of 0.01 M PBS, followed by centrifugation. The supernatant was then collected and filtered. SE was spiked into the sheep feces sample to final concentrations of 1 × 10 8 , 1 × 10 7 , and 1 × 10 6 CFU/mL before analysis. As shown in Table 1, the recovery rate was between 94.99% and 110.0%, indicating that the sensor yields accurate results from a complex sample matrix, demonstrating the promising practicality for determination of bacterial concentrations in real samples.  The reproducibility of the immunosensor was evaluated via the intra-and inter-assays ( Figure 5C). After the three modified electrodes were incubated with the same concentration of target, the relative standard deviations for the intra-batches and inter-batches were 0.98% and 3.9% respectively. Additionally, the ECL intensity was scanned six times using repetitive cyclic voltammetry. The relative standard deviation (RSD) was 1.68%. These investigations show the excellent stability and reproducibility of the proposed immunosensor.
The selectivity of the immunosensor for SE detection was demonstrated using five representative bacteria with similar structures, using Vibrio alginolyticus (VA), Vibrio parahaemolyticus (VP), Escherichia coli (E. coli) and Vibrio sinairobi (VS) as interferences. The concentrations were all 1 × 10 9 CFU/mL, 100-fold larger than the SE sample. As shown in Figure 5D, The ECL signals of the interferences were found to be indistinguishable from that of the blank; however, the ECL intensity for SE showed a large decrease, highlighting the selectivity of this sensing approach. The results illustrate that the ECL signal was clearly decreased only in the presence of the target SE, showing the excellent specificity of the immunosensor.

Application of ECL Bacterial Sensor in Real Samples
To verify that the prepared bacterial sensor has good reliability in practical application, the SE concentration in sheep feces was quantified by the proposed immunosensor. The recovery test was performed by a standard addition method. A 1 g sample of sheep excrement was ultrasonically dispersed into 5 mL of 0.01 M PBS, followed by centrifugation. The supernatant was then collected and filtered. SE was spiked into the sheep feces sample to final concentrations of 1 × 10 8 , 1 × 10 7 , and 1 × 10 6 CFU/mL before analysis. As shown in Table 1, the recovery rate was between 94.99% and 110.0%, indicating that the sensor yields accurate results from a complex sample matrix, demonstrating the promising practicality for determination of bacterial concentrations in real samples.

Conclusions
In this work, Ir-NCDs nanowires were synthesized through a simple cross-linking between amidation of the carboxyl groups of cyclometallic iridium complexes with the amino groups of nitrogen-doped carbon quantum dots. A reliable, sensitive and selective label-free ECL immunosensor based on the Ir-NCDs was proposed for detecting Salmonella enteritidis for the first time. Carbon dots provided an enhanced contribution to the ECL performance of Ir-NCDs due to efficient intramolecular electron transfer processes. The fabricated ECL immunosensor was constructed through a specific assembly between antigens and antibodies. The immunosensor has sufficient stability and selectivity for determination of SE with a detection limit as low as 10 2 CFU/mL. Therefore, the immunosensor is expected to provide a new method for preparing a novel ECL luminophore and expanding the application of iridium complexes in bio-related analysis.