Characterization of Pt species on Tiv-Ti3-xC2Ty MXenesThe Pt atomic loading level of Pt species (Pt SA, and NPs) synthesized via an atomic layer deposition (ALD) strategy was detected by inductively coupled plasma mass spectrometry (ICP-MS) (Supplementary Table 1). The products were named by the structure of Pt and their atomic loading level. For example, Pt SA-0.06 wt%-Ti3-xC2Ty MXenes means that Pt SA was loaded in the MXenes catalysts at a content of 0.06 wt%. With an increase in the number of ALD cycles, the Pt NPs formed. Similarly, the products of Pt NPs-0.46 wt%-Ti3-xC2Ty MXenes represented that Pt NPs were loaded in the MXenes catalysts with a content of 0.46 wt%. Figure 1a illustrates that Pt SA formed at a single ALD cycle on the Ti3-xC2Ty MXenes. Figure 1b illustrates the distribution and structure of Pt NPs on the Ti3-xC2Ty MXenes as the number of ALD cycles increased (Fig. 1b). To further explore the nature of neighboring Pt atomic species, and their oxidation states in Ti3–xC2Ty, XANES, X-ray photoelectron spectroscopy (XPS), and EXAFS spectroscopy were characterized. Figure 1c shows the XANES spectra of Pt L3-edge of Pt NPs-Ti3–xC2Ty MXenes (curve i), Pt foil (curve ii), PtO2 (curve iii), and Pt SA-Ti3–xC2Ty MXenes (curve iv). the white-line intensity of the Pt SA-Ti3-xC2Ty MXenes is higher than that of Pt NPs-Ti3-xC2Ty MXenes and Pt foil, but between those of Pt foil and PtO2. This suggests that the Ptδ+ single-atom carries a more positive charge rather than Pt NPs anchored on Ti3-xC2Ty MXenes and the oxidation state is between 0 and +4. The results are consistent with the XPS results shown in Fig. 1d. The Pt 4f XPS spectra of the Pt species-Ti3-xC2Ty MXenes exhibit three pairs of peaks. The most intense pair represents metallic Pt (Pt0) with binding energies of 70.9 and 74.2 eV. The second pair corresponds to Pt2+ in PtO (72.1 eV) or Pt(OH)2 (75.5 eV), and the third pair corresponds to Pt4+ in species such as PtO228. Compared with that of Pt NPs-Ti3-xC2Ty MXenes (curve ii), the Pt 4f peak position of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes (curve i) was shifted to higher energy, indicating a higher oxidation state of Pt in Pt SA-0.06 wt%-Ti3-xC2Ty MXenes due to the strong EMSI effect. The Fourier transform (FT) k²χ(k)-R space curves and EXAFS fitting analysis of the Pt L3-edge for PtO2 (curve i), Pt SA -Tiv-Ti3-xC2Ty MXenes (curve ii), Pt foil (curve iii), and Pt NPs-Tiv-Ti3-xC2Ty MXenes (curve iv) in Fig. 1e reveal distinct fine structure features. The FT curves of Pt SA-Tiv-Ti3-xC2Ty MXenes display two main peaks at approximately 1.9 Å and 2.3 Å, which correspond to Pt–C and Pt–Ti scattering, respectively. Notably, there is no Pt–Pt peak at 2.8 Å, which indicates that the Pt in Pt SA-Tiv-Ti3-xC2Ty MXenes is present in an atomically dispersed state. The results of the wavelet transform (WT) analysis are in accordance with those of the other techniques, indicating the presence of a local intensity maximum at k = 5.0 Å−1 for Pt–C in Pt SA-Tiv-Ti3-xC2Ty MXenes (Fig. 1f), with no Pt–Pt path at 8.5 Å−1 in Pt NPs-Tiv-Ti3-xC2Ty MXenes (Fig. 1g) and Pt foil (Supplementary Fig. 1). The EXAFS fitting results in Supplementary Table 2 provide further support for this conclusion, with coordination numbers (CN) of Pt–N at 2 and Pt–Ti at 2.5, which demonstrate the successful formation of atomically dispersed Pt on the MXenes support.Fig. 1: Morphological characterization and valence analysis.TEM images of (a) Pt SA-0.06 wt%-Ti3-xC2Ty MXenes, (b) Pt NPs-0.46 wt%-Ti3-xC2Ty MXenes. c Pt L3-edge XANES for (i) Pt NPs-Ti3-xC2Ty MXenes, (ii) Pt foil, (iii) PtO2, and (iv) Pt SA-Ti3-xC2Ty MXenes. d XPS of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes (i), Pt NPs-0.46 wt%-Ti3-xC2Ty MXenes (ii). Insert shows the Bader charge transfer between TiV-Ti3-xC2Ty MXenes support and central Pt species. In the model schematic, the color blue ball represents Pt atoms, yellow ball represents Ti atoms, brown ball represents C atoms, and the green ball represents the capped functional group, Ty. e The k2-weighted R-space FT spectra from EXAFS for (i) PtO2, (ii) Pt SA-Ti3-xC2Ty MXenes, (iii) Pt foil, and (iv) Pt NPs-Ti3-xC2Ty MXenes. WT for the k2-weighted Pt L3-edge EXAFS signals in the (f) Pt SA-Ti3-xC2Ty MXenes, and the (g) Pt NPs-Ti3-xC2Ty MXenes. Source data are provided as a Source Data file.The ECL behaviors of luminol on different Pt species on the TiV-Ti3-xC2Ty MXenesIt is noted that distinguished cathodic ECL behaviors of luminol were observed on the different Pt species anchored to TiV-Ti3-xC2Ty MXenes modified electrodes (Fig. 2a). The Pt SA-0.06 wt%-Ti3-xC2Ty MXenes modified luminol cathodic ECL curve exhibited an obvious ECL peak at −0.54 V. In comparison, the cathodic ECL peak at around −0.9 V was observed for the Pt nanoparticles in MXenes. Additionally, the ECL intensity at −0.54 V was found to be dependent on the Pt SA content. To identify the ECL behavior, ECL spectrums of the different Pt species on Ti3-xC2Ty MXenes support were collected (Fig. 2b). The spectra presented the same peak wavelength at approximately 460 nm, which corresponds to the luminescence of 3-aminophthalate29,30. It was demonstrated that the ECL signal intensity of luminol on various Pt species on the TiV-Ti3-xC2Ty MXenes was highly dependent on the luminol concentrations (Supplementary Fig. 2) and the resolved oxygen in the solution (Supplementary Fig. 3). Consequently, the electrochemical oxygen reduction of the different Pt species anchored on the TiV-Ti3-xC2Ty MXenes modified electrode was examined (Supplementary Fig. 3). During the ORR process in Supplementary Fig. 4, the Pt species on the TiV-Ti3-xC2Ty MXenes support exhibited significant differences in the 2e−-4e− selectivity and average electron transfer number (n). Pt SA-0.06 wt%-Ti3-xC2Ty MXenes showed a boosted selectivity for H2O2 production in ORR, while Pt NPs exhibited poor selectivity for H2O2 due to the cleavage of the O-O bond by the adjacent atom sites31. To investigate the mechanism of luminol cathodic ECL in the Pt species loaded MXenes catalysts, radical scavengers were introduced to demonstrate reactive oxygen species (ROS) species during Pt SA- NPs-TiV-Ti3-xC2Ty MXenes catalyzed ORR. The luminol solution was supplemented with 1-butanol (OH• scavengers), BQ (O2•− scavengers) NaN3 (1O2 scavengers), Na2SO3 (O2 scavengers) (Supplementary Fig. 5). When using Pt SA-0.06 wt%-Ti3-xC2Ty MXenes as the correction accelerator, the ECL emission was largely quenched with the addition of BQ, indicating the critical role of O2•− in the ECL reaction. However, the ECL signal of Pt NPs-Tiv-Ti3-xC2Ty MXenes can be inhibited by both 1-butanol and BQ, suggesting that both O2•− and OH• play critical roles in the ECL process. In contrast, only a slight decrease in the ECL intensity was observed with NaN3, Na2SO3, and catalase, indicating that 1O2, O2, and H2O2 were not the primary reactive ROS species in the ECL reaction.Fig. 2: The ECL and ORR behaviors of luminol on different Pt species on the TiV-Ti3-xC2Ty MXenes.a ECL-potential curves, and (b) ECL spectra of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes (curve i), and Pt NPs-0.46 wt%-Ti3-xC2Ty MXenes (curve ii). c Potential dependent in situ Raman spectra during the negative scan process of the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes in 0.1 M KOH solution with a 785 nm laser source. d The potential dependence of the integrated band intensity for the O–O stretching mode of O2•− on Pt SA-0.06 wt%-Ti3-xC2Ty MXenes. Curve i represents the trend of ECL intensity with potential, while curve ii represents the trend of O2•− Raman peak signal intensity with potential. e The potential dependence of the integrated band intensity for the O–O stretching mode of H2O2 on Pt SA-0.06 wt%-Ti3-xC2Ty MXenes. Curve i represents the trend of ECL intensity with potential, while curve ii represents the trend of H2O2 Raman peak signal intensity with potential. Source data are provided as a Source Data file.Furthermore, in situ Raman spectra were used to characterize the potential-dependent ROS formation on the Pt species on the TiV-Ti3-xC2Ty MXenes support during the negative scanning process (Fig. 2c). The stretching mode of adsorbed O2•− was observed at 1200 cm−1 and the band at 900 cm−1 was assigned to the bridge-bonded peroxide or superoxide (Supplementary Fig. 6)32. No cathodic polarization current or Raman signal of reactive oxygen species was detected within the potential range of 0 to −0.1 V. At −0.2 V, a cathodic polarization current began to accumulate, and vibration (ν) at 1200 cm−1 was observed simultaneously due to the ν(O2•−). However, no ν(H2O2*) signal was detected at this potential. This indicates that the adsorbed O2 is first reduced to O2•− by a single electron transfer on the surface of the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes. When the potential was scanned negatively to −0.3 V, a weak signal at 900 cm−1 appeared, which originated from the H2O2*. It was worth noting that the variation trend of the luminol cathodic ECL of the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes was completely consistent with the variation trend of O2•− signal with potential (Fig. 2d). However, the peak potential of H2O2* was 0.2 V, which was more negative than the peak potential of ECL. The concentration variation trend of H2O2* with peak potential was not consistent with that of the luminol cathodic ECL trend (Fig. 2e). This result suggests that O2•− is the main coreactant responsible for the luminol cathodic ECL reaction on the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes modified electrode. This finding is consistent with the results of the above ECL quenching experiments. The OH• variation trend for the Pt NPs-0.22 wt%-Ti3-xC2Ty MXenes was found to be consistent with the variation trend of the luminol cathodic ECL (Supplementary Fig. 7). The different ROS products catalyzed by Pt species were found to be responsible for the distinguished cathodic ECL behaviors of luminol. In the case of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes, the ORR is predominantly governed by a 2e− pathway, resulting in the feasible formation of H2O2. This occurrence can be attributed to the absence of adjacent Pt active sites and the hindrance of O–O bond cleavage. During the ORR process, the O2•− species generated is preferentially accommodated on the Pt SA due to the absence of continuous Pt–Pt bonds. In contrast, Pt NPs-Ti3-xC2Ty MXenes tend to follow a 4e− pathway, resulting in the production of H2O. The OH• generated during this process is the primary active species responsible for the cathodic ECL of luminol. The unique cathodic ECL characteristics, which include varying peak potentials, can be attributed to the specific ROS species produced during the ORR process on the Pt species attached to Ti3-xC2Ty MXenes.Furthermore, DFT calculations were used to calculate Bader charge analysis and reveal the electronic transfer between the TiV-Ti3-xC2Ty MXenes support and the central metal Pt. It is crucial to understand the electronic metal-support interaction (EMSI) effect in metal-supported electrocatalysts, as it significantly influences their catalytic performance33,34. The transfer of electrons between TiV-Ti3-xC2Ty MXenes support and Pt active centers caused changes in the electron cloud density of the Pt atoms, affecting the electrocatalyst’s catalytic performance. DFT calculations revealed the charge transfer in various Pt species-TiV-Ti3-xC2Ty MXenes. The Pt SA-0.06 wt%-Ti3-xC2Ty MXenes showed a charge transfer of about 1.25e- from the Pt atom to the TiV- Ti3-xC2Ty MXenes support, while the electron transfer of about 0.30 e− decreased between the Pt NPs and the TiV- Ti3-xC2Ty MXenes support. The results suggest that various dispersed forms of Pt active centers impact the electron transfer between the TiV-Ti3-xC2Ty MXenes support and central metal Pt. To validate this experimentally, the chemical composition and surface electronic states of different Pt active centers were characterized by XPS and XANES. Compared with Pt NPs-Ti3-xC2Ty MXenes, the Ptδ+ single-atom of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes had more positive charge than Pt NPs on Ti3-xC2Ty MXenes due to the strong EMSI effect. The distinct downshift levels of the d-band centers of the supported Pt species towards the Fermi level, caused by differences in oxidation states, alter the redox potential of metal ions and consequently influence the ROS products during the ORR process20. This ultimately leads to distinguished luminol cathodic ECL peaks, allowing for qualitative analysis of Pt SA.The quantitative analysis of the Pt SA loading level on the TiV-Ti3-xC2Ty MXenes by the luminol cathodic ECL behaviorsThe ECL stability of the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes has been thoroughly investigated by modifying it on a glassy carbon electrode (GCE). Supplementary Fig. 8a shows that the ECL stability of the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes modified GCE was assessed in the potential range from 0 to −1.0 V with continuous potential scanning. Remarkably, the ECL emission exhibited a uniform intensity for each cycle, with a low relative standard deviation (RSD) of 2.88%. The stability of Pt SA-0.06 wt%-Ti3-xC2Ty MXenes demonstrates their potential as a promising candidate for constructing ECL sensors. Additionally, evaluating the reproducibility of the ECL sensors is a crucial performance metric. To evaluate the reproducibility, five ECL sensors prepared under identical conditions were tested. Supplementary Fig. 8b presents the RSD as 2.37%, indicating a commendable level of reproducibility for the Pt SA-0.06 wt%-Ti3-xC2Ty MXenes modified GCE-based ECL sensors. The exceptional stability and reproducibility make it a highly attractive choice for the development of advanced ECL sensors.In addition to investigating the luminol cathodic ECL behaviors influenced by different Pt species on Ti3-xC2Ty MXenes, we also examined the cathodic ECL behaviors of luminol on Pt SAs anchored on Ti3-xC2Ty MXenes with varying concentrations. The Pt SA loading levels were further confirmed using ICP-MS in Supplementary Table 3. The ECL intensity of Ti3-xC2Ty MXenes in the Pt SA was increased gradually with the increasing loading level of Pt SA, as shown in Fig. 3a. A linear relationship was found between the intensity of the ECL peak and the Pt SA loading level within the range of 0.02–0.12 wt%. The detection limit of 0.014 wt% was impressively low (Fig. 3b). It is worth noting that the loading level of Pt SA detected by this luminol cathodic ECL strategy was comparable to that of current mainstream Pt SA quantification techniques like DRIFTS14,21,35. This observation underscores the exceptional sensitivity and accuracy of our ECL-based method, thereby confirming its suitability for the precise detection of trace amounts of Pt SA on Ti3-xC2Ty MXenes.Fig. 3: The quantitative analysis of the Pt SA loading level on the TiV-Ti3-xC2Ty MXenes by the luminol cathodic ECL behaviors.a ECL intensities of assay versus different loading of Pt obtained by ALD (from i to v was 0.02, 0.04, 0.06, 0.10, 0.12 wt%, respectively). b The plot of ECL intensity versus the corresponding linear value of the Pt SA loading. Error bar shows the standard deviation (n = 3). c The ECL peak intensity trend of Pt SA (curve i) and Pt NPs (curve ii) with the number of cycles turns, d The corresponding ECL curve before (curve i) and after (curve ii) the stability test. Source data are provided as a Source Data file.To validate the accuracy and reliability of detecting trace Pt SAs on Ti3-xC2Ty MXenes using the ECL strategy, Ti3-xC2Ty MXenes with different Pt SA loading were synthesized using different methods, including one-pot synthesis reported in the literature36. The samples were then analyzed to assess the presence and concentration of Pt SA using the developed ECL strategy. The Pt SA loading levels were further confirmed using ICP-MS, as shown in Supplementary Table 4. The results demonstrated a consistent relationship between the ECL peak intensity and the Pt SA content. As illustrated in Supplementary Fig. 9a, the ECL intensity progressively with the escalating loading level of Pt SA. This correlation was further validated by obtaining a linear relationship between the intensity of the ECL peak and the Pt SA loading level (Supplementary Fig. 9b).To extend the generalizability of the developed ECL method for analyzing various MXene substrate-loaded Pt SAs, a series of MXenes-loaded Pt SA were synthesized and characterized. Generally, stabilizable-loaded Pt SAs of MXene and unstabilizable-loaded Pt SAs can be classified according to the formation energies of Pt SAs on the MXenes37. Among the stabilizable-loaded Pt SAs, Pt SAs are stable on Ti2C, V2C, Nb2C, Mo2C, Ti3C2, Zr3C2, Nb4C3, and Ta4C3 monometal carbide MXenes, and Ti3C2, Nb2C, and V2C substrates loaded with Pt SAs exhibited excellent ORR performance. Additionally, Pt SA catalysis on Zr3C2, Ta4C3, and Mo2C MXene supports resulted in less effective ORR, attributed to the strong adsorption of ROS produced during Pt SA catalysis. For a while, Pt SA was not stable on Ta2C, W2C, and Sc2C MXenes, and it was difficult to get pure Pt SA rather than the mixture of Pt SA and Pt nanoparticles37. To further elucidate usability, Nb2CTy and V2CTy MXenes (with good ORR performance) and Mo2CTy MXenes (with poor ORR performance), as well as Ta2CTy MXenes (unable to stably load Pt SA) were chosen as the model for detailed analysis.The cathodic ECL behaviors of luminol on Pt SAs anchored on V2CTy (Supplementary Fig. 10a), and Nb2CTy (Supplementary Fig. 10c) MXenes with varying concentrations were characterized (Supplementary Tables 5 and 6). These MXenes substrates loaded with Pt SA showed a similar ECL peak at approximately −0.5 V that is similar to that of Pt SAs on Ti3-xC2Ty MXenes. Additionally, the ECL peak intensity of Pt SAs anchored on V2CTy (Supplementary Fig. 10b), and Nb2CTy (Supplementary Fig. 10d) MXenes was also linear to the loading level of Pt SA. The facts demonstrate that the as-designed ECL method can be applied for the Pt SA on the MXene qualitatively and quantitatively. In addition, the cathodic ECL behaviors of luminol on Pt SAs anchored on Mo2CTy MXenes (Supplementary Fig. 11) were also characterized. A similar ECL peak at approximately −0.5 V was observed, while the ECL peak intensity is weak to those above, which might be attributed to the strong ROS adsorption on Mo2CTy MXenes that hinder the ECL reaction. The facts demonstrated that the ECL method can be applied to Pt SA quantitative analysis. The cathodic ECL behaviors of luminol on Pt SAs anchored on Ta2CTy MXenes (Supplementary Fig. 12) were characterized. It was observed that a wide cathodic ECL peak was presented, which combined the ECL response of Pt SA at −0.5 V and the response of Pt nanoparticles at −0.8 V, suggesting the characterization of Pt SA on Ta2CTy MXenes. These facts demonstrated that the developed ECL methods are suitable to the qualitative characterization of Pt SAs on carbide MXenes, and can also be used for quantitative characterization of Pt SA stabilized carbide MXene.The ECL method developed in this study was used to evaluate Pt species during Pt SA-catalyzed ORR processes in situ. A prolonged catalytic test was performed to assess the catalytic conversion of Pt SA. As depicted in Fig. 3c, the ECL peak intensity of Pt SA (curve i) gradually decreases with an increase in the number of cycles. After 1000 cycles, the ECL peak intensity of Pt SA decreased by 13%, and an approximately 16% aggregation of Pt SA was observed based on the linear relationship obtained above, as shown in Fig. 3d. Additionally, the characteristic peak of Pt NPs emerged (Supplementary Fig. 13), and the ECL peak intensity of Pt NPs (curve ii) gradually increased with the number of cycles. This observation can be attributed to the partial aggregation of Pt SA after the extended catalytic test. The results show that the ECL method developed can be used for both qualitative and quantitative analysis of Pt SA and its aggregation during the electrocatalytic process. This innovative approach creates opportunities for investigating the catalytic mechanisms involved in various electrocatalytic reactions.In summary, a feasible luminol cathodic ECL characterization technique was established to be applied for the qualitative and quantitative analysis of the Pt SA loading on Ti3-xC2Ty MXenes. The qualitative analysis relies on the distinctive ORR catalytic pathways, resulting in the generation of specific ROS that drive luminol cathodic ECL. Furthermore, within the same catalytic mechanism, the quantitative analysis depends on the variation in Pt SA loading levels on the TiV-Ti3-xC2Ty MXenes support. The developed ECL method exhibits a detection limit for Pt SA, reaching as low as 0.014 wt%, comparable to that of current mainstream Pt SA quantification techniques like DRIFTS. Importantly, this developed ECL method also enables successful in situ analysis of Pt SA during electrocatalysis, showing great promise in qualitative and quantitative evaluations of Pt atom states, as well as facilitating SA catalysts development and discovery of catalytic mechanisms. The well-designed ECL platform presented here represents a significant advancement in the effective control and rational tuning of single atoms, propelling future research efforts in this field.
