Single-atom catalysts (SACs), where metal atoms are dispersed individually on a support, represent a promising approach to enhance catalytic activity and reduce costs compared to traditional bulk metal catalysts. SACs bridge the gap between homogeneous and heterogeneous catalysis, ideally offering the uniform active site distribution of homogeneous systems within the heterogeneous realm. However, a growing body of evidence indicates significant heterogeneity in SAC active sites, complicating the identification of the true active sites and precise quantification of intrinsic activity. This heterogeneity hinders rational catalyst design. This study focuses on platinum (Pt) SACs, widely used in various catalytic reactions. Pt SACs generally consist of Pt ions in the +2 oxidation state (PtII), which, with its d8 electronic configuration, typically favors square planar geometry (D4h) due to ligand field theory. Extended X-ray absorption fine structure (EXAFS) analyses often suggest porphyrin-like geometries (e.g., Pt-N4, Pt-S4) for PtII in carbon-supported Pt SACs. However, theoretical models of these symmetric structures predict poor catalytic activity, while broken or unsaturated coordination geometries are expected to exhibit enhanced activity. This discrepancy between experimental performance and theoretical predictions underscores the need for further investigation into the actual active sites in Pt SACs. Previous studies have suggested asymmetric coordination geometries or in situ modifications of symmetric geometries as possible explanations, but the role of broken geometric symmetry in electrocatalysis has remained unclear. This uncertainty stems from the heterogeneity of Pt sites, making it difficult to distinguish between different coexisting Pt moieties even using powerful techniques like EXAFS. The current study aims to address this challenge by investigating low-coordinated PtII species in carbon-based Pt SACs and their contribution to electrocatalysis using the chlorine evolution reaction (CER) as a model reaction.

The literature extensively explores single-atom catalysis, highlighting its potential advantages over traditional heterogeneous catalysts. Studies such as Wang et al. (Nat. Rev. Chem. 2018) and Ji et al. (Chem. Rev. 2020) have reviewed the synthesis and applications of SACs. The unique characteristics of SACs, bridging homogeneous and heterogeneous catalysis, have been discussed in several works (e.g., Cui et al., Nat. Catal. 2018; Kim et al., Acc. Chem. Res. 2022), emphasizing the importance of understanding active site heterogeneity. The challenges associated with identifying and quantifying the intrinsic activity of SACs have also been highlighted (e.g., Christopher, ACS Energy Lett. 2019; Mitchell et al., Angew. Chem. Int. Ed. 2018; Zitolo et al., Nat. Mater. 2015). Specifically regarding Pt SACs, research has demonstrated their versatility in electro-, photo-, and heterogeneous catalysis (Cheng et al., Nat. Commun. 2016; Gao et al., J. Am. Chem. Soc. 2016; Qiao et al., Nat. Chem. 2011; Choi et al., Nat. Commun. 2016). Theoretical studies have explored the influence of coordination geometry on Pt SAC activity (e.g., Hossain et al., Nat. Commun. 2020; Xiao et al., Nano Res. 2022; Li & Wang, Nano Res. 2022), predicting that low-coordinated sites are more active. However, experimental evidence supporting these theoretical predictions has been limited. Previous work by the authors and others has explored the catalytic behavior of Pt SACs (Lim et al., Nat. Commun. 2020; Lim et al., ACS Catal. 2021; Zhao et al., Nat. Commun. 2022; Fang et al., Nat. Commun. 2020; Kwon et al., J. Am. Chem. Soc. 2018) providing some understanding of the active sites, but the importance of broken geometric symmetry was not thoroughly investigated.

The researchers synthesized a model Pt SAC catalyst, Pt1(3)/CNT, by heat-treating a mixture of PtII meso-tetraphenylporphine (PtTPP) and acid-treated carbon nanotubes (CNTs) at 700 °C under N2. The catalyst was characterized using various techniques including high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), X-ray diffraction (XRD), X-ray absorption spectroscopy (XAS), and X-ray photoelectron spectroscopy (XPS). The Pt content was determined using inductively coupled plasma-optical emission spectroscopy (ICP-OES). The catalytic performance of Pt1(3)/CNT for the chlorine evolution reaction (CER) was evaluated using electrochemical techniques such as polarization curves, online differential electrochemical mass spectrometry (DEMS), and rotating ring disk electrode (RRDE) measurements. The durability of the catalyst was assessed by performing 500 iterative cyclic voltammograms (CVs). Real-time Pt dissolution was analyzed using online inductively coupled plasma-mass spectrometry coupled with an electrochemical flow cell (EFC/ICP-MS). To investigate the role of surface functional groups, additional catalysts with abundant oxygen (O-Pt1(3)/CNT) and chlorine (Cl-Pt1(3)/CNT) functionalities were prepared via post-treatment. Control experiments were conducted using Pt SAC catalysts with different Pt loadings (Pt1(1)/CNT and Pt1(0.15)/CNT). In situ XAS measurements were also performed to monitor changes in the catalyst structure under electrochemical conditions. Density functional theory (DFT) calculations were used to model different Pt coordination environments (Pt-N4, Pt-N3, Pt-N3V) and investigate the reaction mechanisms and stability of these sites under CER conditions. The DFT calculations provided free-energy diagrams for CER and the competing oxygen evolution reaction (OER). Finally, to further analyze the chemical composition of the aged catalyst, XPS was used to analyze the samples before and after degradation.

The study's key findings demonstrate the significant role of low-coordinated PtII species with broken D4h symmetry in the high CER activity of Pt SACs. Initially, Pt1(3)/CNT exhibited excellent CER activity, but significant activity loss (61%) was observed after 500 CV cycles, while Pt loss was only 22%, suggesting a change in the turnover frequency (TOF) rather than just a decrease in the number of active sites. The analysis revealed the formation of oxygen and chlorine functional groups on the carbon support after the durability test, but control experiments indicated that these groups do not significantly affect the TOF. The synthesis of Pt SACs with lower Pt loadings (Pt1(1)/CNT and Pt1(0.15)/CNT) revealed a surprising increase in TOF with decreasing Pt content. EXAFS analysis of these catalysts showed a decrease in the Pt-N coordination number (CN) as the Pt loading decreased, with Pt1(0.15)/CNT exhibiting a CN of 3.0 indicating low-coordinated species, either trigonal-planar-like PtII-N3 or T-shaped Pt-N3V. The high CER activity of Pt1(0.15)/CNT demonstrates that Pt-N3(V) sites are more active than the Pt-N4 sites. DFT calculations supported this finding, showing that the T-shaped Pt-N3V site is the most active for CER, with a significantly lower free energy barrier compared to the Pt-N4 site. DFT calculations also showed that the oxidative demetallation potential (Udiss) was significantly lower for Pt-N3(V) sites compared to Pt-N4 sites, which is consistent with the observed higher lability of Pt-N3(V) sites. In situ XAS showed a higher coverage of CER intermediates on Pt1(0.15)/CNT and that the CN of Pt-N/O increases to 4 due to formation of a Pt-O bond.

The findings resolve the long-standing discrepancy between experimental observations and theoretical predictions regarding the activity of Pt SACs in electrocatalysis. The study's key contribution is the identification of the low-coordinated PtII species, particularly the T-shaped Pt-N3V site, as the primary contributor to the high CER activity. The observed increase in TOF with decreasing Pt loading suggests a synergistic effect between the low-coordinated sites and the carbon support. The fact that the decrease in CER activity correlates better with Pt loss when lower Pt loadings are used implies that the labile, highly active Pt-N3(V) sites are responsible for this degradation. The different reaction pathways for CER on Pt-N4 and Pt-N3(V) sites further highlight the importance of understanding the active site structure-activity relationships. This work changes the focus from maximizing total Pt content to maximizing the number of three-coordinated sites with a broken symmetric geometry to optimize performance. The study provides valuable insights for the design of next-generation SACs using other dn metal ions.

This study demonstrates that the high CER activity of Pt1/CNT catalysts is primarily attributed to the presence of low-coordinated PtII species, particularly the T-shaped Pt-N3V sites, which exhibit superior activity but lower stability than their Pt-N4 counterparts. Maximizing the density of these low-coordinated sites is a critical factor for achieving high electrocatalytic performance. Future research should focus on developing synthetic strategies to control the coordination geometry of Pt sites and enhance the stability of the active Pt-N3(V) sites under electrochemical conditions. Understanding the stabilization of these rarely observed species on the supporting substrate is a high priority, and these findings may be generalizable to other dn metal ions in SACs.

The study mainly focuses on the CER. While the findings are expected to be relevant to other electrocatalytic reactions, further investigation is needed to confirm the generality of these observations. The DFT calculations involved approximations and simplifications, and the precise nature of the interaction between low-coordinated Pt sites and the carbon support may require further investigation. The linear extrapolation used to estimate Pt-N3(V) and Pt-N4 content may not be fully accurate since it assumes the TOF of Pt-N3(V) sites is constant across all catalysts. The study mainly focused on a specific type of carbon support (CNTs); further studies are needed to investigate the impact of different support materials.