Recent Progress in Heteroatom-Containing Metalloporphyrin-Based Catalysts for CO₂ Reduction

Abstract

Metalloporphyrins, owing to their structural resemblance to natural enzyme active sites and highly tunable coordination environments, have emerged as promising catalysts for converting CO₂ into value-added chemicals and fuels. Considerable efforts have been made to modify metalloporphyrins to improve their catalytic capability for CO₂ reduction. One approach involves modifying the metal coordination environment (known as the first coordination sphere) to generate heteroatom-containing metalloporphyrins, particularly N-confused and O/S-substituted variants. While heteroatom-containing metalloporphyrins were first synthesized in 1989, their use in CO₂ reduction catalysis was not reported until after 2020. Herein, we review the recent progress in the design, catalytic performance, and mechanistic studies of N-confused and O/S-substituted metalloporphyrins towards CO₂ reduction. This review encompasses both experimental and theoretical computational work, as well as the use of porphyrins as catalysts in photocatalysis and electrocatalysis. Finally, based on the current research advances, we present critical recommendations and future research directions, with a focus on theoretical studies, in the hope of facilitating the rational design of novel catalysts for sustainable energy conversion and environmental remediation.

1. Introduction

The increasing concerns about climate change and the depletion of fossil fuels have raised the need for sustainable energy solutions [1,2]. The electrochemical reduction of carbon dioxide (CO₂) to value-added hydrocarbons and chemical feedstocks, such as alcohols, aldehydes, ketones, etc., represents a promising approach for dealing with CO₂ emissions and alleviating energy and environmental crisis [3,4,5]. However, the inherent chemical inertness of CO₂ and the existence of a competitive hydrogen evolution reaction poses a challenge for its efficient conversion [6,7,8], thereby driving the need for catalyst development [9,10].

Metalloporphyrins are a class of coordination complexes in which a metal ion (e.g., Fe, Mg, Co, Zn, Cu, Ni) is bound to the nitrogen atoms of a porphyrin ligand. They are ubiquitous in nature and serve essential functions in various critical life processes. Typical examples include iron porphyrins (e.g., heme and cytochrome P450), magnesium porphyrins (e.g., chlorophyll), and cobalt porphyrins (e.g., vitamin B12) [11,12]. These metalloporphyrins share a common mechanistic hallmark: enabling electron transfer via redox transitions of the central metal ion (e.g., Fe²⁺/Fe³⁺) or activating small molecules (e.g., O₂, CO₂) through metal–ligand coordination [13,14]. Their unparalleled catalytic versatility has driven breakthroughs in artificial enzyme design, with transformative applications in precision medicine like targeted drug metabolism and renewable energy, particularly CO₂-to-fuel conversion [15,16,17]. Regarding the latter, much empirical evidence has been reported in the literature. For example, iron porphyrin compounds can act as homogeneous catalysts for the reduction of CO₂ to CO in N, N-dimethylformamide (DMF) solution [18]. Nickel porphyrin compounds featuring exocyclic 1,3-ditert-butylimidazolin-2-ylidenamino (NItBu) substituents also exhibit catalytic activity for CO₂-to-CO conversion in DMF [19]. Additionally, cobalt porphyrin compounds can be incorporated into metal–organic frameworks (MOFs) or covalent–organic frameworks (COFs) through organic linkers, serving as efficient catalysts for CO₂ reduction to CO [20,21,22,23,24,25].

How to optimize the catalytic activity and selectivity of the metalloporphyrin-based catalysts is the key to these studies. To reach this end, modifying the structures of the metalloporphyrins has been shown to be an effective strategy [26,27]. The distinctive structure of porphyrin molecules provides a highly modifiable platform for functionalization. As shown in Figure 1a, porphyrins are macrocycles that are composed of four pyrrole subunits interconnected via methine bridges, forming a planar, aromatic, and highly stable framework [28]. It readily anchors metal ions at its core, forming a stable M–N₄ coordination environment (Figure 1b) [29]. In addition to changing the type of the central metal, common strategies for modifying metal porphyrins include introducing substituents at the meso or β positions of the porphyrin, coordinating axial ligands to the central metal, and modifying the first coordination or the outer coordination sphere [27,30,31,32,33,34]. Among these strategies, modifying the first coordination sphere of metal porphyrin compounds has received relatively less attention, likely due to the greater synthetic challenges in constructing asymmetrically coordinated metal porphyrin systems. Nevertheless, recent experimental and theoretical studies have made significant progress in this direction, demonstrating it as an effective strategy for enhancing catalytic efficiency [27,35,36,37]. Of note, this strategy has been widely used in graphene-based M–N–C type single-atom catalysts [38,39,40,41,42].

A number of recent review articles have provided comprehensive coverage of utilizing metal porphyrin-based catalysts for CO₂ conversion [14,26,43,44,45]. These reviews all focus on traditional N₄-metalloporphyrin systems. For example, Philipp Gotico et al. [44] (2020) and Domingo-Tafalla et al. [26] (2023) reviewed research advances in outer coordination sphere modifications through functional group effects (e.g., hydrogen bonding, electrostatic interactions, and electronic effects) and hierarchical material design (e.g., MOFs, COFs, and carbon composites) of metalloporphyrins. To our knowledge, no review has addressed modifications of the first coordination sphere in metal porphyrins for CO₂ reduction, motivating our summary of recent advances in this field.

Modifying the first coordination sphere in metal porphyrins enables access to diverse derivatives with potentially improved catalytic performance. For example, N-confused porphyrins, which are isomers of porphyrins, are formed by inverting the N and C atoms of one or multiple pyrrole rings (Figure 1c). Additionally, replacing one or two pyrrole N atoms with other heteroatoms such as O, S, Se, B, P, and C results in a class of porphyrinoids known as heteroatom-containing porphyrins or core-modified porphyrins (Figure 1d) [46,47,48]. Among the latter, metalloporphyrins incorporating oxygen or sulfur heteroatoms demonstrate relatively facile synthesis. These substitutions induce significant modifications in both the geometric and electronic structures of the metalloporphyrins, and they fundamentally change the reaction coordinate and introduce additional mechanistic complexity in CO₂ reduction catalytic pathways.

Herein, we present a review of the advancements of first coordination sphere modification on metalloporphyrins for CO₂ reduction, both experimentally and theoretically. We particularly focus on the N-confused and O- or S-substituted species. The contents of this review are organized as follows: Session 2 briefly reviews the application of metalloporphyrins in CO₂ reduction and introduces the commonly used computational methods in modeling this process. Session 3 reviews the synthesis and catalytic progress of N-confused metalloporphyrin for CO₂RR. Session 4 reviews the synthesis and catalytic progress of O/S-substituted metalloporphyrin for CO₂RR. Session 5 presents a summary and outlook. By reviewing the current state of research on first coordination modification on metalloporphyrins, we hope to provide insights into the design principles and strategies that can guide the development of next-generation catalysts for sustainable energy conversion.

2. Metalloporphyrins for CO₂RR

The use of metalloporphyrins as catalysts for CO₂ reduction can be traced back to 1979. Toshima et al. found that cobalt porphyrins with carboxyl (–COOH) and sulfonic acid groups (–SO₃H) showed electrocatalytic activity in reducing CO₂ to formic acid, HCOOH [49]. Since then, metalloporphyrins have garnered increasing attention as catalysts for CO₂ reduction, especially those with low-cost transition metal centers like iron, cobalt, and nickel [18,19,43,50,51]. For example, Savéant et al. (2016) reported electrochemical CO₂-to-CO conversion by Fe tetraphenylporphyrins with four positively charged trimethylanilinium groups (Fe-p-TMA) [33]. The Omar Yaghi group (2015) synthesized co-porphyrin covalent organic frameworks (COF-366-Co) for the electrocatalytic reduction of CO₂ to CO [20]. The Koper group (2015) immobilized cobalt porphyrins on pyrolyzed graphene as electrocatalysts to reduce CO₂ to CO and CH₄ [52]. Dey’s group (2021) achieved the conversion of CO₂ to HCOOH via electroreduction by introducing a pendant amine group onto the iron porphyrin (Fe-2,3,7,8-tetracetyl-15-phenylporphyrin, FeTEsP) [53]. Liu et al. (2013) reported copper porphyrin-based metal–organic frameworks (Cu-5,10,15,20-tetrakis(4-carboxyphenyl) porphyrin, Cu-TCPP-MOFs) as photocatalysts for converting CO₂ to CH₃OH [54].

The above studies indicate that metal porphyrin compounds predominantly catalyze the reduction of CO₂ to C1 products, including carbon monoxide (CO) [18,20,31,33,50,55], formic acid (HCOOH) [49,53], methanol (CH₃OH) [54], and methane (CH₄) [52,56,57]. C2 products are scarcely observed on porphyrin-based catalysts because of the lack of suitable adsorption sites for C–C coupling to C2 intermediates. Hence, modifications to the coordination sphere of metal porphyrins retain this selectivity, with C1 species—particularly CO—remaining the dominant reduction products.

Elucidating the precise mechanism of CO₂ reduction by metalloporphyrin compounds faces several fundamental challenges: the complexity of possible reaction pathways sensitive to experimental conditions, the transient nature of key intermediates limiting experimental detection, and ambiguous electronic configurations arising from redox-active metal centers and porphyrin ligands [58,59]. Hence, a computational study plays an important role. In theoretical calculations, density functional theory (DFT) is almost exclusively employed. Researchers have adopted two distinct modeling approaches, molecular cluster-based DFT calculations and periodic DFT calculations. Molecular cluster models are used for molecular metalloporphyrin complexes as homogeneous catalysts. Electron transfer (ET) and proton transfer (PT) processes are typically treated separately [56,60,61,62,63,64,65,66], enabling the computation of reduction potentials for different species and identifying the order and/or synchronicity of ET and PT. The periodic DFT calculations, which are widely used for heterogeneous catalysts, are primarily applied to covalent organic frameworks (COFs) and metal–organic frameworks (MOFs) constructed with metalloporphyrin units [37,67,68]. Periodic calculations generally consider concerted proton-electron transfer (CPET) as a unified process, as shown in Figure 2. In this case, for example, the reduction of CO₂ to CO follows the mechanism of the adsorption of CO₂, the formation of *COOH and *CO, and the desorption of CO. The computational hydrogen electrode (CHE) model [69] was usually used to account for the energy change in adding a proton and an electron, i.e., a hydrogen atom. The effects of pH and applied potential can be incorporated through an energy correction term [70].

3. N-Confused Metalloporphyrin

N-confused porphyrins (NCPs) are isomers of porphyrin. The first NCP, 2-aza-21-carbaporphyrin, was synthesized by Latos-Grażyński and Furuta et al. in 1994 [72]. In principle, inverting one, two, or three pyrrole rings of porphyrins could generate a first coordination sphere with N₃C₁, N₂C₂, or N₁C₃ configurations. To date, synthetic efforts reported in the experimental literature have exclusively produced configurations featuring either single or double pyrrole ring inversion, while no successful construction of triply inverted configurations has been achieved. For this reason, theoretical studies mainly focused on N₃C₁ and N₂C₂ configurations.

The formal oxidation state of the central metal in a M–N₄-type metalloporphyrin is +2. Anchoring metal elements into NCPs gives N-confused metalloporphyrins, and the central metal remains a formal oxidation state of +2. The first example of N₃C₁-type metalloporphyrins was A Ni(II) complex of 21-(CH₃)–NCTPP and was reported by Latos-Grażyński et al. in 1995 [73]. Later, in 2000, the same group reported the synthesis of Cu^II–N(–CH₃)CTPP [74]. Fe^II(–Br)–NCTPP (2001) and Mn^II(–Br)–NCTPP (2002) were synthesized by Hung et al. [75,76]. Cu^II/Ag^III/Pd^II–NCP–(C₆F₅)₄ was synthesized by Furuta et al. in 2003 [77], and Co^IINCTPP was synthesized by Bettelheim et al. in 2004 [78]. Other than these, N₃C₁-type metalloporphyrins with central metals of Mo (2005) [79], Rh (2008) [80], Re (2007) [81], Pt (2000) [82], Au (2005) [83], and Hg (2008) [84] have also been synthesized experimentally. Shao and coworkers (2025) recently reported the synthesis of NiN₃C- and CoN₃C-coordinated porphyrin-based COFs [85]. For N₂C₂-type metalloporphyrins, the syntheses of complexes with two inverted carbon atoms both in the ortho-(cis-N₂CP) and para-positions (trans-N₂CP) have been reported. Cu^III/Ag^III–cis-N₂CP–(C₆F₅)₄ and Cu^III–trans-N₂CP–(C₆F₅)₄ were synthesized by Furuta et al. in 2000 and 2003 [86,87], respectively. These N-confused metalloporphyrins are illustrated in Figure 3. Computational studies on these systems with other central metals, such as Co, have been conducted [71].

The application of N-confused metalloporphyrins as catalysts for CO₂ reduction has only been demonstrated recently. In 2023, Ren and coworkers reported theoretical studies of N-confused metalloporphyrin-based covalent organic frameworks (Por-COFs) for enhanced electrocatalytic CO₂RR, where periodic DFT with PBE-D3 functional was used [71]. The study employed a monolayer COF model constructed through imine-linked 5,10,15,20-tetrakis(4-benzaldehyde)porphyrin (Por-CHO) and p-phenylenediamine (PPDA), forming a 2D periodic structure (Figure 4b). Through comprehensive screening of ten 3d transition metals (Sc-Zn) based on stability and catalytic selectivity and activity, CoN₄-Por-COF was identified as a promising candidate for catalyzing CO₂ reduction to CO (limiting potential U_L = −0.89 V), while CrN₄-Por-COF catalyzes CO₂ reduction to HCOOH (U_L = −0.78 V). The authors further modified the coordination environment of CoN₄-Por-COF by inverting one or two pyrrole units to design mono-pyrrole or dipyrrole-inverted N-confused Por-COFs, named CoN₃C₁- and CoN₂C₂-Por-COFs (Figure 4a). A free energy analysis (ΔG) of key intermediates (*COOH vs. *H) revealed the designed N-confused materials effectively suppressed hydrogen evolution (HER) competition. Detailed calculation of the CO₂ reduction pathway (CO₂ → COOH → CO → further reduction) showed that the N-confused CoN₃C₁- and CoN₂C₂-Por-COFs display improved performance in catalyzing CO₂RR than parent CoN₄-Por-COF. The limiting potential of reducing CO₂ to CO decreases from −0.89 V (CoN₄) to −0.76 V (CoN₃C₁) and −0.60 V (CoN₂C₂), and applying high potential can yield deep-reduction degree C₁ products such as CH₃OH and CH₄ (Figure 4c–e). Electronic structure analysis revealed that substituting CoN₄ to CoN₃C₁/CoN₂C₂ increases the electron density on the Co atom and raises the d-band center, thus stabilizing the key intermediates *COOH and lowering the ΔG of respective PDSs (PDS refers to potential-determining step). These findings underscore the promise of coordination engineering in COFs, positioning CoN₂C₂ and CrN₃C₁ as superior candidates for selective CO₂RR. This theoretical framework not only rationalizes experimental trends but also sets the stage for exploring asymmetric coordination motifs in other transition-metal systems.

Later, Huang and coworkers (2024) [88] synthesized a cobalt porphyrin complex with N-confused coordination, and Shao and coworkers (2025) [85] synthesized a CoN₃C porphyrin-based COF. Both works demonstrated that the CoN₃C systems enhance electrocatalytic activity for the reduction of O₂ to H₂O. Although the experiments were not for CO₂RR, the authors confirmed the N₄ to N₃C coordination change reduced electron transfer from cobalt to the porphyrin ligand, which is the key for catalytic performance improvements, consistent with the theoretical prediction by Ren’s work [71]. While the periodic DFT approach effectively modeled the electronic structure of monolayer COFs, the study [71] acknowledged limitations including unaccounted interlayer interactions, solvent effects, and kinetic barriers that might lead to overestimated catalytic performance in practical applications. By assuming a CPET process, the kinetic barriers were not estimated in this periodic DFT work.

In 2024, Hua and coworkers reported a N-confused copper(II) tetraphenylporphyrin (CuNCP) complex for efficient electrochemical CO₂ reduction to methane (CH₄) under pulsed potential electrolysis (PPE) [89]. Two copper porphyrin complexes were synthesized via a modified Geier method: symmetric copper tetraphenylporphyrin (CuTPP, Cu-N₄ coordination) and asymmetric N-confused copper porphyrin (CuNCP, Cu-N₃C coordination). Experimental results revealed that CuTPP maintained structural integrity under constant potential electrolysis (CPE) but primarily produced hydrogen (H₂) (Figure 4b). CuNCP exhibited higher CO₂ reduction activity (Figure 5a), which is attributed to the asymmetric N₃C coordination but suffers from structural instability during electrolysis. In situ X-ray absorption spectroscopy (XAS) and transmission electron microscopy (TEM) confirmed the irreversible reduction in CuNCP to Cu⁰ nanoparticles under CPE, leading to diverse products (CO, C₂H₄, H₂, etc.). However, applying PPE (anodic potential: +1.3 V, cathodic potential: −1.6 V, pulse cycle: 5 s/10 s) significantly enhanced CH₄ selectivity, achieving 60% Faradaic efficiency (FE) with a partial current density of 170 mA cm⁻² (Figure 5c,d). Combined in situ infrared (IR), Raman spectroscopy, and scanning electrochemical microscopy (SECM) analyses suggested that PPE promotes water dissociation at high anodic bias, consuming OH⁻ and lowering local pH, thereby enriching proton availability for the successive hydrogenation of *CO intermediates and ultimately boosting methane production. Although initial hypotheses proposed dynamic Cu(I)/Cu(II) cycling to restore molecular structure, experimental data ruled out catalyst regeneration, instead highlighting the role of electrolysis protocols in optimizing interfacial chemical environments. This study underscores the trade-off between enhanced catalytic activity and compromised stability in asymmetric coordination (CuN₃C), and it proposes a novel strategy of tailoring electrolytic conditions to catalyst properties for efficient CO₂ reduction.

In 2024, Peng and coworkers reported a nickel(II) N-confused tetraphenylporphyrin complex (NiNCP) (Figure 6a) that efficiently photocatalyzes CO₂ to CO reduction in acetonitrile solution under irradiation from a blue LED (420 nm), using [Ru(bpy)₃]²⁺ as a photosensitizer and triisopropanolamine (TIPA) as a sacrificial electron donor [36]. The complex achieved a turnover number (TON) of up to 217,000 with 98% selectivity (Figure 6b). In their experiments, the authors synthesized NiNCP by improving upon the Lindsey and Choi method, confirming its structure through characterization techniques such as MALDI-TOF, XPS, and ¹H NMR. They discovered that the acidic N–H group in NiNCP is deprotonated by TIPA during the catalytic process, forming an anionic active species, NiNCP⁻. Photocatalytic experiments revealed that NiNCP produced 0.32 mmol of CO within the first hour of the reaction, and the TON increased significantly as the catalyst concentration decreased (Figure 5b). Through steady-state fluorescence and time-resolved spectroscopy analyses, the authors found that NiNCP efficiently extracts electrons from the photosensitizer, with its LUMO energy level (3.40 eV) well matched to those of the photosensitizer (3.19 eV) and the CO₂/CO couple (3.91 eV) (Figure 6c). Furthermore, DFT calculations indicated that NiNCP⁻ attacks the carbon atom of CO₂ via a ligand-centered Lewis basic site, forming a stable *COOH intermediate stabilized by a strong C–C bond (1.5 Å), rather than following the traditional metal-centered activation pathway, thereby significantly reducing the reaction energy barrier (Figure 6d,e). The work redefines the role of coordination asymmetry, demonstrating that ligand participation—not just metal tuning—can unlock unprecedented catalytic efficiency.

To our knowledge, only the aforementioned studies have been reported on N-confused porphyrins for CO₂ research. However, additional works exist regarding metallated NCP complexes applied to the oxygen reduction reaction [85,88]. In the next section, we will discuss the O/S-substituted metalloporphyrins.

4. O/S-Substituted Metalloporphyrin

Since one oxygen or sulfur atom is equivalent to one NH group, directly substituting MN₄ to MN₃O or MN₃S of metalloporphyrin leads to the formal oxidation state of the central metal to +1. If one wants to maintain a +2 oxidation state of M, an additional axial ligand, typically a Cl atom, is often introduced. Similarly, when two NH groups are replaced to form N₂S₂ or N₂O₂-coordinated metalloporphyrins, the central metal is coordinated by two Cl ligands. Experimentally, a few O/S-coordinated metalloporphyrins have been synthesized, including Fe/Co/Ni/Cu/Zn/Mn/Re-21-oxaporphyrin (N₃O) [90,91,92,93,94], Li/Fe/Ni/Cu/Ru/Rh/Pd/Re/Hg-21-thiaporphyrin (N₃S) [93,95,96,97,98,99,100], Ni-21,23-dioxaporphyrin (N₂O₂) [90], and Ru/Re-21,23-dithiaporphyrin (N₂S₂) [101,102]. To date, no experimental studies have reported the successful synthesis of metalloporphyrin complexes incorporating three heteroatoms, such as N₁O₃ or N₁S₃. Moreover, MOFs or COFs that incorporate metalloporphyrins with oxygen or sulfur have not yet been reported, although their synthesis looks feasible. In principle, oxygen and sulfur can be interchanged and combined to form metalloporphyrin complexes, such as MN₂OS porphyrins. Such structures have not yet been synthesized experimentally, but they have been studied by computational works [67].

Latos-Grażyński and co-workers made pioneering efforts in the field of heteroatom-containing porphyrins synthesis. They reported the first synthesis of Cu(II), Fe(II), and Ni(II) complexes of 21-thiaporphyrin (N₃S–TPP) in 1989 [95]. X-ray diffraction showed that the large radius of the S atom disrupts the planar structure in MN₄ porphyrins. All these three metalloporphyrins contain an axial chloride ligand, so as to maintain the the metal center in the +II oxidation state. For Ni^II(–Cl)N₃S–TPP, Latos-Grażyński and co-workers carried out chemical and electrochemical reduction to prepare its one-electron reduced Ni^IN₃S-TPP (1989) [103]. The same group synthesized Rh^III(–Cl₂)N₃S–TPP in 1989 [96] and Pd^II(–Cl)N₃S–TPP(–CH₃)₂ in 1994 [97]. In 2000, LiN₃S–TPP(–CH₃)₂ was synthesized by Arnold et al. [98]. Hg^II(–Cl)N₃S–TPP was prepared by Hwang et al. in 2002 [99]. Later, Ru^II(–Cl)(–CO)N₃S–TPP(–CH₃)₄ (2011) [100] and Re^I(–(CO)₃)N₃S–TPP (2012) [93] were reported by Hung et al. and Ghosh et al., respectively. For N₂S₂-porphyrins, the large atomic radius of S atom disrupts the planar stability, so only N₂S₂-porphyrins containing the heavy metal Ru and Re have been synthesized so far, i.e., Ru^II(–Cl₂)N₂S₂–TPP(–CH₃)₄ (2001) [101] and Re^I(–(CO)₃)N₂S₂–TPP(–CH₃)₄ (2014) [102]. These S-substituted metalloporphyrins are illustrated in Figure 7.

The first example of O-coordinated metalloporphyrins was synthesized in 1997. Latos-Grażyński and co-workers synthesized a five-coordinated Ni(II)-21-oxaporphyrin with an axial chloride ligand, Ni(–Cl)N₃O–TPP, in which the N₃O moiety adopts a planar configuration [90]. Subsequently, they conducted an electroreduction experiment to convert it into a four-coordinated Ni^IN₃O-TPP. In 2002, the same group also reported the formation of Fe^III-21-oxaporphyrin [91]. In 2012, Ghosh et al. synthesized Re^I(–(CO)₃)N₃O–TPP [93]. In 2013, Gloe and co-workers reported the Mn(II), Co(II), Zn(II) and Cu(II) complexes of N₃O porphyrins [94]. For the N₂O₂ system, Ni^II-21,23-dioxaporphyrin was prepared in 1997 [90]. There are two chloride ligands on Ni, and the NiN₂O₂ motif attains a planar structure. These O-substituted metalloporphyrins are illustrated in Figure 8. The chemical properties of heteroatom-containing porphyrins and metalloporphyrins have been summarized in several review articles [46,47,48,104,105].

Studies concerning the catalytic performances of heteroatom-coordinated metalloporphyrins were only available recently. In 2021, Choi and co-workers investigated the effects of a mono-oxygen substitution in nickel tetraphenylporphyrin (NiN₄–TPP) on electrocatalytic CO₂ reduction [27]. The modified nickel-21-oxatetraphenylporphyrin, the Ni(–Cl)N₃O–TPP catalyst (Figure 9a), was synthesized via a tailored Lindsey method using 2,5-bis (phenylhydroxymethyl) furan as the precursor, followed by metalation and characterization. Electrodes were prepared by mixing the porphyrins with carbon black, and the CO₂RR was conducted in a 0.5 M KHCO₃ solution. Ni(–Cl)N₃O–TPP displays much better electrocatalytic CO₂RR activity and selectivity than NiN₄–TPP. NiN₄–TPP exhibits predominant H₂ formation over the whole potential range, but minor CO formation is detected at a high overpotential region below −0.75 V_RHE with a maximum FE_CO of only ca. 2%. (Figure 9c). In sharp contrast, Ni(–Cl)N₃O–TPP shows an onset potential of CO formation at approximately −0.55 V_RHE, and its maximum FE_CO reaches ca. 80% at −0.65 V_RHE.

The improved performance was attributed to the broken D_4h symmetry in Ni(−Cl)N₃O−TPP, which results in an increase in the Ni redox potential yielding Ni^I. NiN₄-TPP shows no reduction between −0.4 and −0.8 V_SHE, while Ni(–Cl)N₃O–TPP is reduced at −0.62 V_SHE. This is also supported by DFT calculations, where their respective calculated reduction potential is −1.04 and −0.54 V_SHE. As a result, the catalytically active species are different under the applied potential. Ni^II in NiN₄–TPP and Ni^I in NiN₃O–TPP are the actual chemical species responsible for the CO₂RR. In DFT calculations using the M06 functional, the authors assumed a CPET mechanism under an applied potential of –0.6 V_RHE, and the reaction follows the steps of CO₂ adsorption, the formation of *COOH and *CO, and the desorption of CO (Figure 9b). The [NiN₃O–TPP]^– greatly stabilizes the critical species *COOH (Figure 9d), thus leading to a much lower overpotential. In comparison, the calculated COOH binding energy for NiN₄-TPP is −0.77 eV, and it is −1.73 eV for [Ni–N₃O–TPP]^–. The authors pointed out that the weak ligand field formed by N₃O coordination reduced the energy required for Ni to redistribute electrons to the ligand, thus increasing its Lewis acidity for forming a strong Ni–C bond and stabilizing the *COOH intermediate. At the same time, the lowering of the empty Ni d orbital energy due to this electron redistribution results in a positive shift in its reduction potential. Meanwhile, the author also found that the Ni(–Cl)N₃O–TPP complex suffers from instability over a potential of <−0.6 V_RHE, posing a challenge for its practical usage. This combined spectroscopic and computational study revealed that the broken ligand-field symmetry is the key for active CO₂ electrolysis, and controlling the ligand-field strength would be a challenge for future synthesis.

In 2024, He and coworkers [35] reported similar nickel porphyrin catalyst systems in photocatalytic CO₂ reduction. They incorporated O and S into tetracarboxyphenyl ([–TPP–(COOH)₄]) nickel porphyrin (NiN₄-Por) and synthesized Ni(Cl)ON₃-Por and Ni(Cl)SN₃-Por. Cyclic voltammetry indicated that the reduction potentials of O/S-substituted species are more positive than the original N₄-Por. The values are −1.52 V vs. Fc/Fc⁺ for NiN₄-Por, −1.17 V for Ni(Cl)ON₃-Por, and −0.82 V for Ni(Cl)SN₃-Por. DFT calculations at the B3LYP-D3 level showed that the lowest unoccupied molecular orbitals (LUMOs) are lowered from NiN₄-Por (−2.41 eV) to Ni(Cl)ON₃Por (−2.87 eV) and Ni(Cl)SN₃Por (−3.06 eV). The lowering of the LUMO suggests higher electron-accepting capability, and it was used to support the positive shift in reduction potential observed in experiment. Although it is possible to give the calculated reduction potential values, the work did not report them.

Then, the authors conducted photocatalytic experiments with a 465 nm LED lamp in CH₃CN/H₂O mixture (v/v, 4:1) solution and with [Ru(bpy)₃]Cl₂ as the photosensitizer and triethanolamine (TEOA) as the electron sacrificial agent. Ni(Cl)ON₃-Por and Ni(Cl)SN₃-Por have remarkable CO₂RR activity, with a CO production rate of 24.7 and 38.8 mmol g⁻¹ h⁻¹ with a selectivity of 94.0 and 96.4%, respectively (with H₂ as the by-product), outweighing that of NiN₄–Por (a CO production rate of 6.6 mmol g⁻¹ h⁻¹ and selectivity of 82.8%). At the same time, the authors also examined the CO production when using H₂N₄–Por, HON₃–Por, and HSN₃P–Por as catalysts. They are far less than the CO production from the Ni-porphyrin counterparts, suggesting the Ni center to be the active site. Since the reduction potential of the photosensitizer, [Ru(bpy)₃]Cl₂ (PS), PS⁺/PS* is at −1.30 V vs. Fc/Fc⁺, the active catalytic species were proposed to be Ni^II in [NiN₄–Por]⁰, Ni^I in [NiON₃–Por]⁰ (after one-electron reduction), and Ni^I in [NiSN₃Por]⁻ (after two-electron reduction) (Figure 10a–c). Then, these active species followed a CPET mechanism, the PDS was the formation of *COOH, and the energy barrier decreased from 1.30 to 0.67 and 0.58 eV for N₄ to N₃O and N₃S systems (Figure 10d). Similarly, the authors attributed such changes to the disruption of D_4h symmetry in O/S-substituted Ni-porphyrin. The calculated energy level of Ni d_z² orbital is elevated in the NiON₃–Por and NiSN₃–Por (where the writer believed the comparisons should be made between the active species), thus enhancing the adsorption of CO₂ (Figure 10e). At the same time, the increased electron density on Ni (supposedly, of the active species) stabilizes the *COOH intermediate (Figure 10f–h) and finally leads to faster CO₂-to-CO conversion. This work demonstrates how O/S-substitution can alter the electronic states of Ni-porphyrin catalysts and improve their catalytic performance.

To our knowledge, experimental studies on O- or S-substituted metalloporphyrins for CO₂ reduction have been limited to Ni metal centers. Studies that involve other metal centers like Co or Fe are theoretical works.

In 2021, Lu et al. reported a DFT study on Fe-N₂S₂ tetraphenylporphyrin (TPP) and predicted that it can improve CO₂ reduction performance to multiple C1 products, including CO, HCOOH, CH₃OH, and CH₄ [37]. The systems are molecular complexes. For the computations, the VASP 6.1.1 software was used with the PBE-D3 functional and implicit solvent model (water as solvent). The computational hydrogen electrode (CHE) model was used. Fe–N₂S₂ porphyrin forms a non-coplanar framework with elongated Fe–N bonds (2.14 Å vs. 2.04 Å in Fe–N₄). The band gap of the Fe–N₂S₂ porphyrin (0.34 eV) is much lower than that of the Fe–N₄ porphyrin (1.62 eV), due to the raised orbital level of d-orbitals of Fe and p-orbitals of N. The electronic density of states (DOS) reveals additional Fe d_z² orbital contributions in Fe–N₂S₂, promoting intermediate adsorption in the axial direction. The Fe atom in Fe–N₂S₂ was found to have a higher electron density than the Fe atom in Fe–N₄ porphyrin. Fe–N₂S₂ exhibits slightly stronger CO₂ adsorption (E_ads = –0.20 eV) compared to Fe–N₄ (E_ads = –0.16 eV), attributed to increased electron transfer from Fe to CO₂. Note that the adsorption energy difference is about 0.04 eV or 0.9 kcal/mol, which is within the accuracy of DFT calculations. For the CO₂-to-CO pathway, the PDS of both Fe–N₄ and Fe–N₂S₂ systems is the desorption of CO, with respective ΔG of 1.45 eV and 1.11 eV (Figure 11a). For the HCOOH pathway, the PDS is the desorption of HCOOH for both FeN₄ and FeN₂S₂ porphyrins and the respective ΔG of 0.70 eV and 0.38 eV (Figure 11b). For the CH₃OH pathway, the PDS is CH₃OH desorption for the Fe–N₄ porphyrin and *CO→*CHO step for the Fe–N₂S₂ porphyrin, with the respective ΔG of 0.90 eV and 0.40 eV. For the CH₄ pathway, the PDS is the *CO→*CHO step for the Fe–N₄ porphyrin, and *CH₃OH→*CH₃ step for FeN₂S₂ porphyrin, with the respective ΔG of 0.73 eV and 0.56 eV. As a result, the calculated limiting potentials U_L for all pathways were lowered in FeN₂S₂ in comparison to the Fe–N₄ case. The respective values are listed in

Table 1. Calculated limiting potential (U_L/V) of the products for the H₂, CO, HCOOH, CH₃OH and CH₄ production on metalloporphyrin systems and DFT methods.

	H2	CO	HCOOH	CH3OH	CH4	DFT Method	Ref.
FeN4-TPP-Por	−0.59	−1.45	−0.70	−0.90	−0.73	Cluster PBE-D3	[37]
FeN2S2-TPP-Por	−0.26	−1.11	−0.38	−0.40	−0.56
FeN4-Por-COF	−1.25	−1.25	/	/	/
FeN3S-Por-COF	−1.25	−1.18	/	/	/	Periodic PBE-D3
FeN2S2-Por-COF	−1.03	−0.91	/	/	/		[67]
FeN3O-Por-COF	−1.50	−1.24	/	/	/
FeN2O2-Por-COF	−1.26	−0.66	/	/	/
CoN4-Por	/	−1.27	/	−0.80	−0.80		[106]
CoN3S-Por	/	−0.58	/	−0.91	−0.91	Cluster ωB97XD-D3BJ
CoN2S2-Por	/	−1.17	/	−0.66	−1.20
CoN3O-Por	/	−1.11	/	−1.04	−1.04
CoN2O2-Por	/	−2.12	/	−0.96	−1.40
CoN4-Por-COF	−0.91	−0.89	−1.36	−1.31	−1.63		[71]
CoN3C1-Por-COF	−0.79	−0.76	/	−0.76	−0.99	Periodic PBE-D3
CoN2C2-Por-COF	−0.63	−0.60	/	−0.60	−1.44
CoN3S-Por-COF	−0.75	−0.74	/	/	/	Periodic PBE-D3
CoN2S2-Por-COF	−0.84	−0.88	/	/	/		[67]
CoN3O-Por-COF	−0.64	−0.67	/	/	/
CoN2O2-Por-COF	−0.79	−0.58	/	/	/
NiN4-Por-COF	−1.52	−1.52	/	/	/
NiN3S-Por-COF	−1.16	−1.08	/	/	/
NiN2S2-Por-COF	−1.28	−1.25	/	/	/	Periodic PBE-D3	[67]
NiN3O-Por-COF	−0.97	−1.35	/	/	/
NiN2O2-Por-COF	−1.23	−0.98	/	/	/

.

On the Fe–N₄ porphyrin, U_L follows the sequence HCOOH > CH₄ > CH₃OH > CO. On the Fe-N₂S₂ porphyrin, U_L follows the sequence HCOOH > CH₃OH > CH₄ > CO, with the value of −0.38 V for HCOOH and −0.40 V for CH₃OH, surpassing most reported catalysts. At the same time, since their values of U_L are very close, the HCOOH and CH₃OH product channels are in great competition. Compared to conventional Fe–N₄, Fe–N₂S₂ porphyrin shifts product selectivity toward HCOOH and CH₃OH, with reduced energy barriers for key steps such as *CO → *CHO. An important drawback of the Fe–N₂S₂-system is that the hydrogen evolution also becomes easier. The calculated free energy change of H⁺ + e⁻ → *H → H is 0.59 eV on the Fe–N₄ porphyrin and 0.26 on the Fe–N₂S₂ porphyrin. This work underscores the potential of double-heteroatom coordination strategies in designing high-efficiency electrocatalysts.

In 2023, Qi and co-workers [106] reported a theoretical study investigating a series of O- or S-substituted cobalt porphyrin molecular complexes (Figure 12a, where the peripheral ligands are H atoms) for electrochemical CO₂RR. The authors performed DFT calculations using Gaussian 09 with a ωB97XD-D3BJ functional and SMD implicit solvent model with water as the solvent. The computational hydrogen electrode model was used. Based on the Co–N₄ porphyrin (Co1), they modified the ring structure and designed three structures, including cobalt corrole (Co2), cobalt octahydroporphyrin (Co3), and cobalt 1,5,9,13-tetraazacyclohexadecane (Co4). In terms of the core modification, they introduced O and S heteroatoms and formed six structures, including cobalt 21-oxaporphyrin (Co5, N₃O), cobalt 21,23-dioxaporphyrin (Co6, N₂O₂), 21,22-dioxaporphyrin (Co7, N₂O₂), cobalt 21-thiaporphyrin (Co8, N₃S), cobalt 21,23-dithiaporphyrin (Co9, N₂S₂), and cobalt 21,22-dithiaporphyrin (Co10, N₂S₂). Here, we focus on the comparison between Co5–Co10 and Co1. The authors considered different products of CO₂ reduction, including CO, CH₃OH, and CH₄. The product channel of HCOOH and the HER was not discussed in this work.

Similarly, the core-modification approach decreased the band gap as compared to the CoN₄ porphyrin. This serves as a hint that the intramolecular electron transfer of these cobalt catalysts would be improved during the CO₂RR process. The PDS of CO₂-to-CO reduction on Co1 (N₄), Co6 (N₂O₂), Co7 (N₂O₂), Co9 (N₂S₂), and Co10 (N₂S₂) is the formation of *COOH, where their respective ΔG values are 1.27, 2.12, 1.89, 1.17, and 1.41 eV. The PDS of Co5 (N₃O) is CO desorption with ΔG of 1.11 eV, indicating a promising catalyst for the further hydrogenation reactions of CO into CH₃OH (U_L = −1.04 V) and CH₄ (U_L = −1.04 V) (Figure 12b,c). The PDS of Co8 (N₃S) is CO₂ adsorption, with a minimum ΔG of 0.58 eV among the studied systems, indicating the best catalytic activity for CO₂-to-CO reduction (Figure 12d). Co8 also has the smallest band gaps (3.40 eV), highlighting its strongest reduction capability. The authors concluded that single O/S-substituted cobalt porphyrin exhibits higher activity than double O/S-substituted cobalt porphyrin, and CoN₃S is better than CoN₃O. The adsorption energies of CO in the designed complexes were analyzed. Change analyses showed that Co5–Co10 transferred more electrons (−0.046 |e| ~ −0.182 |e|) to CO, in comparison to Co1 (0.149 |e|, indicating electron transfer from CO to Co1), hence, a stronger CO binding, and this facilitates the formation of deeper reduction products like CH₃OH and CH₄. They also claimed that the broken D₄h symmetry altered the electron density of the Co porphyrin and improved CO₂RR performance.

In 2024, Ren et al. [67] performed periodic DFT calculations with PBE-D3 functional on O/S-substituted metalloporphyrin-based COFs (Por-COFs) for electrocatalytic CO₂ reduction. This work designed 15 types of heteroatomic Por-COFs, featuring M–N_xO_yS_z (M = Fe, Co, Ni; x + y + z = 4) centers, including single-substituted models: N₃O, N₃S, and double-substituted models: N₂O₂, N₂S₂, and N₂OS (Figure 13a). It focused on the electrocatalytic reduction of CO₂ to CO and assumed a CPET mechanism. The PDS for all the target M–N_xO_yS_z-Por-COFs is the formation of the *COOH intermediate. Figure 13b summarizes the corresponding theoretical limiting potentials U_L. In general, changing the coordination from N₄ to N_xO_yS_z shifts the U_L^CO₂RR towards the positive direction, meaning an easier reduction and thus higher CO₂-to-CO catalytic activity. The majority of the designed M–N_xO_yS_z-Por-COFs display a distinct preference for CO₂RR. Exceptions are Ni–N₃O, Co–N₃O, Co–N₃S, and Co–N₂S₂ systems, which show a slight preference for HER. Based on calculated U_L values, the four CO₂-to-CO catalyst candidates ranked at the top are Co–N₂O₂, Fe–N₂OS, Fe–N₂O₂, and Co–N₃O-Por-COFs. The best designed system, Co–N₂O₂-Por-COF, achieved a limiting potential (U_L^CO₂RR) of −0.58 V, outperforming the parent Co–N₄-Por-COF (U_L^CO₂RR = −0.89 V) (Figure 13c). Electronic structure analysis attributed this improvement to enhanced orbital overlap between Co 3d orbitals and *COOH 2p orbitals, alongside increased charge transfer (0.32 e⁻ in Co–N₂O₂ vs. 0.15 e⁻ in Co–N₄), stabilizing the critical *COOH intermediate (Figure 13d). In comparison, the U_L values of the Ni-based systems are more negative than the Fe- and Co-based catalysts. Ren et al. predicted that the U_L follows the order of Ni–N₄ (−1.52 V) < Ni–N₃O (−1.35 V) < Ni–N₃S (−1.08 V). Although the Ni–N₃S and Ni–N₃O systems were not the best, this prediction aligns with the experimental superiority of Ni(Cl)–SN₃Por and Ni(Cl)ON₃Por observed in the photocatalytic CO₂ reduction [35], affirming the accuracy of this computational work. Furthermore, the machine learning model (Gradient Boosting Regression, GBR) further identified central metal mass (m) and metal-ligand bond lengths (d_M-X) as pivotal descriptors, underscoring the synergy between metal identity and coordination environment. This work not only predicts high-performance Por-COF candidates but also establishes coordination engineering as a universal strategy for tailoring COF-based electrocatalysts.

The three computational studies mentioned above, although they involve different peripheral ligands, in part share the same first coordination spheres. All calculations assumed a CPET mechanism, and the computational results showed both similarities and differences. The similarities include that, after O/S coordination, the band gap of the metal porphyrin decreases, the electron density at the metal center increases, and the adsorption energies of key intermediates (such as CO₂, CO, COOH, and others) change, leading to different catalytic activities for CO₂ reduction.

As for the differences, we highlight the following for instance. For the FeN₂S₂ system, both Lu et al. and Ren et al. predicted a lower U_L for CO₂-to-CO conversion compared to the FeN₄ system (Table 1). However, the former work (employed cluster model) identified CO desorption as the PDS, whereas the latter work (employed periodic model) identified the COOH formation as the PDS. For the Co–N_xO_y system, Qi et al. (employed cluster model) suggested that the single-substituted Co–N₃O-Por performs better than the double-substituted Co–N₂O₂-Por system, and, in comparison to Co–N_xS_y systems, S-substitution is better than O-substitution. However, Ren et al.’s calculations (employed periodic model) yielded the opposite conclusion, favoring the double-substituted Co–N₂O₂-Por-COF system, and O-substitution is better than S-substitution for Co–N₄-system. These discrepancies may arise from ligand effects and, of course, the differences in computational models and methods.

Theoretical studies by Ren et al. in 2023 [71] and 2024 [67] employed the same Por-COF model and DFT method, allowing us to compare the calculated CO₂-to-CO performance of N-confused and O/S-substituted Co-Por-COF systems. These systems shared the same PDS, i.e., the formation of the *COOH intermediate. As shown in Table 1, for singly substituted Co-Por-COF, the order of calculated U_L (thus the order of catalytic performance) is CoN₃O (−0.67 V) > CoN₃S (−0.74 V) > CoN₃C (−0.76 V) > CoN₄ (−1.27 V); for double-substitution, the order of U_L is CoN₂O₂ (−0.58 V) > CoN₂C₂ (−0.60 V) > CoN₂S₂ (−0.88 V). Based on these U_L values, all these systems improved the catalytic performance as compared to the original CoN₄ system, because O, S, and C substitutions increased the electron density of the Co center, thereby strengthening the binding with *COOH. CoN₂O₂ is the best, and double substitution is better than single substitution for C and O, but the trend reverses for S. This latter decline was attributed to the fact that the large atomic radius of S distorts the catalyst structure, reducing orbital overlap and weakening binding to *COOH. This underscores the necessity of balancing electronic modulation and structural stability in heteroatom-substitution strategies, providing crucial guidance for the rational design of high-performance CO₂ reduction catalysts.

5. Summary and Outlook

In this review, we have summarized the recent progress in the application of N-confused and O/S-substituted metalloporphyrins-based systems for CO₂ reduction, both experimentally and theoretically. In general, the first coordination sphere modification by C, O, and S atoms represents an effective method to improve the catalytic activity and selectivity of CO₂ reduction. By breaking the symmetry of traditional M–N₄ coordination, these heteroatoms induce distinct electronic perturbations that optimize intermediate adsorption and lower reaction energy barriers. Furthermore, these heteroatom coordination structures significantly influence product selectivity. The catalytic performances of these systems are summarized in

Table 2. Summary of different catalyst systems and catalytic performance of metalloporphyrin complexes in CO₂RR (E_red: reduction potential; U_L: limiting potential; EC: electrocatalysis; PC: photocatalysis; Expt.: experimental; TPP: tetraphenylporphyrin; and Por: porphyrin).

Catalysts	Cores	Product, FE	Potential	Method	Cat. Type	Solvent	Ref.
Iron-TPP	FeN4	HCOOH	−0.70 VRHE (UL)	DFT	EC	H2O	[37]
Iron-21,23-dithia-TPP	FeN2S2	HCOOH CH3OH	−0.38 VRHE (UL) −0.40 VRHE (UL)
Cobalt-TPP	CoN4	CO, 95%	−0.60 VRHE (Ered)	Expt.	EC	0.5 M NaHCO3	[34]
Cobalt 21-thia-Por	CoN3S	CO	−0.58 VRHE (UL)	DFT	EC	H2O	[106]
Cobalt 21,23-dithia-Por	CoN2S2	CH3OH	−0.66 VRHE (UL)
Cobalt 21-oxa-Por	CoN3O	CH3OH, CH4	−1.04 VRHE (UL)
Cobalt 21,23-dioxa-Por	CoN2O2	CH3OH	−0.96 VRHE (UL)
CoN4-Por-COFs	CoN4	CO	−0.89 VRHE (UL)	DFT	EC	H2O	[71]
CoN3C-Por-COFs	CoN3C	CO, CH3OH	−0.76 VRHE (UL)
CoN2C2-Por-COFs	CoN2C2	CO, CH3OH	−0.60 VRHE (UL)
Nickel-TPP	NiN4	CO, 2% CO, 29%	−0.75 VRHE (Ered) −1.10 VRHE (Ered)	Expt.+DFT Expt.	EC	0.5 M KHCO3 DMF	[27] [19]
Nickel-21-oxa-TPP	NiN3O	CO, 80%	−0.65 VRHE (Ered)	Expt.+DFT	EC	0.5 M KHCO3	[27]
Nickel-TPP(-COOH)4	NiN4	CO, 82.8%	−1.52 VFc+/Fc(Ered)/−1.29 VRHE	Expt.+DFT	PC	CH3CN/H2O	[35]
Nickel-21-oxa-TPP(-COOH)4	NiN3O	CO, 94.0%	−1.41 VFc+/Fc(Ered)/−1.18 VRHE
Nickel-21-thia-TPP(-COOH)4	NiN3S	CO, 96.4%	−0.82 VFc+/Fc(Ered)/−0.60 VRHE
Nickel-TPP	NiN4	CO, <5%	/	Expt.+DFT	PC	CH3CN	[36]
Nickel-NCTPP	NiN3C	CO, 98%	−1.41 VAg/Ag+(Ered)/−1.26 VRHE
Copper-TPP	CuN4	CO, <15%	−1.8 VRHE(Ered)	Expt.	EC	1 M KOH	[89]
Copper-NCTPP	CuN3C	CH4, >60%	−1.6 VRHE(Ered)

, where properties including reduction potential, products and yields, and Faradaic efficiencies are present. It shows that the reactivity changes depend on the type of substituted atoms, the metal center, and experimental conditions. In addition, computational models and methods also affect the results.

Experimentally, a Cu–NCP complex was found to improve the electroreduction of CO₂-to-CH₄ products under pulsed potential electrolysis (PPE), mainly because PPE lowered the local pH. A Ni–NCP complex was reported to improve the photoreduction of CO₂-to-CO, and it was attributed to a ligand-involved mechanism. In addition, Ni–N₃(Cl)O- and/or Ni–N₃(Cl)S-porphyrin systems were found to improve the activity and selectivity of CO₂-to-CO electrochemically and photochemically. The detailed underlying mechanisms of enhanced catalytic reactivity are different for C and O/S substitution, as well as for Cu and Ni. For the Ni-porphyrin system, common findings include that the C/O/S substitutions shift the reduction potential to positive, the Ni^I species are the active catalytic species, and S substitution is better than O substitution.

Both electrocatalytic works mentioned that the modified porphyrin systems experienced critical instability issues, indicating that the asymmetric coordination may weaken macrocyclic stability under redox cycling. Under high potential, Cu-NCP electrolysis to Cu clusters/nanoparticles, applying PPE only delayed and cannot avoid the decomplexation of Cu-NCP [89]. The Ni ions were dissolved from the Ni(–Cl)N₃O-porphyrin system to form metallic Ni or oxidized NiO and Ni(OH)₂ precipitations [27]. In periodic DFT calculations, the stability of original catalysts is often predicted by computing the formation energy and dissolution potential (U_diss) of the metalloporphyrin materials. Predicting the stability of catalysts during the catalytic process, i.e., under applied potential, is normally beyond the calculation capability, due to the difficulty in modeling the real working conditions. To account for the role of solvent in affecting the stability of the catalysts, one may include explicit solvent molecules, but the computational cost is much higher. Moreover, including the effect of potential during the simulations is computationally challenging.

In contrast, both photocatalytic works reported that the catalysts are stable under the experimental conditions. It seems to indicate that, in terms of stability, the first coordination modified porphyrins are suitable for photocatalysis reactions, rather than electrocatalytic reactions.

Computational works extended the systems from single substitution to double and mixed substitutions. Most works focus on Fe, Co, and Ni metal centers. Although the computational models and methods are not the same in those works, the calculation results revealed a number of universal observations, no matter whether the substituted atoms are C, O, or S. The heteroatom substitution changes (mainly increases) the electron density of the center metals, which may enhance the electron transfer from M to other incoming species. It decreases the HOMO-LUMO gap (or band gap) of the porphyrin systems, either by lowering the LUMO or raising the HOMO. The energy level of the d_z² orbital was raised; this can enhance the interaction with incoming species with a proper orbital orientation. These electron redistributions and the change in orbital levels can lead to a shift in the reduction potential and stabilization of key intermediates (such as CO₂, COOH, CO, etc.) during CO₂RR processes and then change the reaction barrier. For different metals, heteroatom substitution affects the intermediates differently, so the potential determining steps, and thus the selectivity, may differ. The major CO₂ reduction product of metalloporphyrin complexes is CO and H₂. After introducing heteroatoms in the first coordination sphere, calculations suggest that Ni systems still prefer CO, Fe systems (like Fe–N₂S₂) may prefer HCOOH and CH₃OH, and Co systems (like Co–N₃C₁, Co–N₂C₂) have increased preference for CH₃OH and CH₄.

Theoretical calculations predicted improved CO₂RR activity for Fe- and Co-centered porphyrin with C or O or S coordination, and these predictions have not been validated by experiments yet, which suggests a potential avenue for further experimental exploration. We would like to point out that the calculations mentioned above are based on the thermodynamics results as computed from the DFT method by assuming CPET, and the CHE model was usually used. We believe many important issues are worth further exploring. For example, treat the protonation and electron-transfer steps separately [60,61,62,66,107], examine the transition states (as the system may be kinetically controlled) [64], track the flow of electrons to check the oxidation state and clarify the active catalytic species, consider the mechanism that ligands serve as active sites, account for the solvent effect by including explicit solvent molecules, etc. It is worth mentioning that these calculations did not consider factors like catalyst degradation or mass transport in real-world electrolysis. A direct comparison between the overpotential measured in the experiment and the limiting potentials (U_L) calculated by DFT is not anticipated. Nevertheless, comparing the calculated parameters, including U_L values and activation energies, across different catalysts provides valuable information on the change in catalytic activity and selectivity.

One may be interested in comparing the influence of the first coordination and outer-sphere modification on the catalytic performance of CO₂ reduction. Here, we take the comparison between NiN₄–TPP, NiN₃O-TPP, and Ni–TPP–NItBu (Ni-1,3-ditert-butylimidazolin-2-ylidenamino-tetraphenylporphyrin) for illustration. Experimentally, the unmodified Ni–N₄–TPP exhibits extremely low catalytic activity in CO₂ electroreduction. Through the first coordination sphere modification, the O-coordinated Ni–N₃O–TPP achieved an FE_CO of 80% at −0.65 V_RHE and experienced a positive shift in reduction potential [27]. Through outer-sphere modification, adding the NItBu group to peripheral benzene ring forms Ni–TPP–NItBu. It increased the FE_CO to 62% at −1.10 V_RHE but had no obvious effect on the reduction potential [64]. Both strategies improved the CO₂-to-CO reduction performance, but the underlying mechanisms are not the same. For the former, O-substitution disrupts the D_4h symmetry of NiN₄ and redistributes the electron density on Ni. This promotes the formation of the active species Ni^I and stabilizes the *COOH intermediate, thus enhancing the CO₂ to CO reduction. For the latter, appending the NItBu group on the outer-sphere increases the electron density at the Ni center, for NItBu is a strong electron-donating group. This facilitates the binding of CO₂ and improves catalytic performance. We can see that the first coordination sphere affects the electron density of the central metal through ligand-field regulation, whereas the outer-coordination sphere affects it via longer-distance electronic interactions.

In comparison, O-substitution is more effective than appending peripheral NItBu groups in increasing the catalytic reactivity towards CO₂-to-CO reduction. However, based on this single comparison, we cannot draw a general conclusion that first coordination modification is more effective than outer-sphere modification. Electron-withdrawing or cationic groups were found to have distinct effects as compared to electron-donating groups. The effects of different substituents were studied for Fe and Co porphyrin systems but not for Ni porphyrin systems [33,34]. We anticipate that combining these two strategies may achieve the maximum enhancement of catalytic performance. In summary, heteroatom-containing metalloporphyrins represent a versatile platform for catalysis for CO₂ and other small molecules. This approach not only provides a deeper understanding of the structure and activity relationships governing reduction reactions but also offers a clear pathway for the rational design of next-generation electrocatalysts. We propose the following promising directions for future investigations: the synergistic effect between the modification of first- and peripheral coordination sphere modification [108], the incorporation of multi-metal centers within expanded porphyrin analogs [109,110,111], and the integration with composite materials such as carbon nanostructures [112,113], covalent organic frameworks [114], metal–organic frameworks [115,116], or conductive polymers [117,118]. Furthermore, to establish catalyst design principles, mechanistic investigations require not only theoretical and computational approaches but also the experimental monitoring of dynamic processes. Operando spectroscopy (e.g., XAS, SEIRAS, DEMS) has been applied to study the catalytic mechanism of CO₂ reduction [118,119]. These techniques can capture the reaction intermediates and obtain critical information such as reaction rates, product formation kinetics, and the oxidation states of active sites on catalysts, thereby bridging the gap with theoretical calculations and facilitating better catalyst design.