Efficient and Durable Photochemical CO₂ Reduction by TiO₂‐Immobilized Metal Porphyrin Catalysts

Abstract

The advancement of molecular catalysts featuring 3d metals provides a sustainable alternative to noble‐metal‐based systems for CO₂‐to‐C1 fuel conversion; however, their widespread application remains limited by modest activity and selectivity. Herein, we report a dye‐sensitized ternary hybrid photocatalyst composed of a visible‐light‐absorbing Ir(III) complex, a TiO₂ electron mediator, and a covalently anchored porphyrin catalyst incorporating 3d metals (Fe, Co, Ni, and Cu). Immobilization of the porphyrin via carboxylate groups enables strong electronic coupling with the TiO₂ scaffold, facilitating efficient multielectron accumulation and directional charge transfer. Under visible‐light irradiation, the TiO₂‐immobilized porphyrins (M‐TCPP; M = Fe, Co, and Ni) exhibit markedly enhanced CO₂‐to‐CO conversion performance with turnover numbers exceeding 3000, surpassing their homogeneous analogs. Notably, Cu‐TCPP shows distinct reactivity, favoring H₂ evolution over CO formation. Electrochemical analysis reveals a metal‐dependent mechanistic divergence, in which Cu‐TCPP undergoes protonation to form a Cu–H intermediate, whereas Fe‐, Co‐, and Ni‐based porphyrins proceed through conventional M–COOH pathways. These findings highlight the synergistic role of TiO₂ in tuning interfacial electron transfer and promoting efficient CO₂ reduction using earth‐abundant molecular catalysts.

A dye‐sensitized TiO₂ scaffold bearing IrP and 3d‐metalloporphyrin catalysts (M‐TCPP; M = Fe, Co, Ni, Cu) drives efficient CO₂‐to‐CO conversion and H₂ evolution. Robust TiO₂ immobilization and rapid interfacial electron transfer within the ternary architecture underpin the markedly enhanced activities relative to the mixed homogeneous systems.

1. Introduction

The catalytic reduction of carbondioxide (CO₂) to energy‐rich fuels and chemical feedstocks remains a central goal in the development of carbon‐neutral technologies [1, 2, 3]. Molecular approaches to CO₂ reduction have advanced markedly, enabled by improved understanding of electron transfer pathways, ligand field effects, and catalyst stability [4, 5, 6, 7, 8, 9]. However, many of the most active systems continue to rely on precious metal centers such as rhenium and ruthenium [4, 5, 6]. While these complexes often exhibit high turnover frequencies and excellent selectivity, their widespread application is fundamentally constrained by limited natural abundance and high cost [4, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]. In light of these limitations, a significant research effort has shifted toward the development of molecular catalysts based on earth‐abundant 3d transition metals [15, 16, 17, 18, 19, 20, 21]. Although these first‐row metal complexes offer the promise of scalability and economic viability, their catalytic performance—particularly in terms of CO₂ binding affinity, reduction kinetics, and product selectivity—often falls short of their noble metal analogs. Within this context, porphyrin‐based ligand scaffolds have emerged as particularly attractive platforms due to their inherent structural and electronic features [22, 23, 24, 25, 26, 27, 28, 29, 30]. The rigid, planar macrocyclic architecture of porphyrins provides unhindered access to the metal center while enabling substrate coordination and modulation of the product selectivity through metal‐centered redox chemistry [22]. In addition, the extended π‐conjugation framework promotes efficient multielectron transfer processes, which are essential for overcoming the thermodynamic and kinetic barriers associated with CO₂ activation [22, 26]. Moreover, the synthetic modularity of porphyrins allows for systematic tuning of electronic properties via peripheral substitution [23, 24, 25, 26, 27, 28]. For instance, introduction of hydroxyl functionalities into iron porphyrins has been shown to enhance CO₂ electroreduction activity, presumably through proton‐coupled electron transfer (PCET) pathways, as originally demonstrated by Savéant and coworkers [23, 24]. Other studies have established that electron‐donating substituents (e.g., methoxy) increase electron density at the metal center, thereby facilitating CO₂ binding and accelerating overall turnover [22, 23, 24, 25, 26, 27, 28]. In contrast, electron‐withdrawing groups (e.g., fluorine) lower overpotentials by stabilizing the reduced state of the metal, although at the expense of CO₂ binding affinity—underscoring the importance of balancing electronic effects to achieve optimal performance [26, 28].

Despite these advances, homogeneous porphyrin‐based systems generally exhibit limited durability under photochemical or electrochemical conditions, largely due to catalyst decomposition, aggregation, or loss from solution [31, 32, 33]. One promising route to address this issue is heterogenization, wherein molecular catalysts are immobilized on conductive supports to create hybrid systems that combine the precision of molecular design with the operational robustness of heterogeneous catalysis [33]. In contrast to solution‐phase catalysis, where multielectron processes require high‐energy collisional electron transfer, surface‐anchored catalysts can access electron reservoirs directly, enabling successive single‐electron reductions at lower driving force (Figure 1) [15, 26, 33, 34, 35] This strategy avoids the need for high‐energy low‐valent intermediates (e.g., M⁰ or M^I) and enables more efficient electronic coupling between the donor species and the catalytic center. Among the common heterogenization strategies, drop‐casting of molecular complexes onto electrodes remains simple but suffers from poor dispersion, catalyst aggregation, and weak interfacial contact—all of which impair performance and longevity [30]. Alternatively, porphyrin‐functionalized metal–organic frameworks and covalent organic frameworks offer structurally defined architectures, but face persistent limitations in electrical conductivity, mass transport, mechanical stability, and scalability [27, 29, 30, 35]. Consequently, the development of hybrid systems that integrate molecular catalysts into electronically conductive, chemically robust, and optically active scaffolds has become an area of growing interest.

FIGURE 1.

Proposed CO₂ reduction mechanisms for (a) homogeneous (soluble porphyrins) and (b) heterogeneous (surface‐immobilized porphyrins) systems.

In this context, dye‐sensitized photocatalytic systems featuring TiO₂ have emerged as versatile platforms for visible light‐driven CO₂ reduction, owing to their modular design and favorable photophysical properties [15, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50]. Covalent attachment of molecular components through phosphonate or carboxylate anchors ensures robust electronic coupling, suppresses charge recombination, and supports multielectron accumulation at the catalytic interface. TiO₂ serves not only as a chemically stable scaffold and electron mediator but also as an electron reservoir capable of modulating interfacial charge transfer through chemical and electrostatic means [51, 52]. Importantly, the conduction band edge of TiO₂ can be tuned by adjusting the identity and concentration of proton donors (e.g., TEOA, TFE, or H₂O), thereby modulating the thermodynamic driving force for directional electron transfer—from the photoexcited dye to the TiO₂ conduction band, and subsequently to the immobilized molecular catalyst [53, 54, 55]. The inherent n‐type character of TiO₂ introduces a subtle electron‐withdrawing effect, which can reduce electron density at the metal center and facilitate catalytic turnover at lower overpotentials [41, 44]. Prior studies have shown that TiO₂‐based hybrid photocatalysts incorporating various polypyridyl‐type molecular catalysts outperform their homogeneous counterparts in terms of activity, selectivity, and long‐term operational stability [15, 42]. Extending this strategy to porphyrin‐based catalysts can offer a promising route to overcome the persistent limitations in activity and durability associated with molecular porphyrin systems while simultaneously harnessing their inherent multielectron redox capability and structurally tunable coordination environment to promote efficient CO₂ reduction.

Herein, we report a dye‐sensitized TiO₂‐based ternary hybrid photocatalyst consisting of an organometallic Ir(III) photosensitizer (IrP), TiO₂ nanoparticles, and a 3d transition metal (Fe, Co, Ni, Cu)‐based porphyrin catalyst. As illustrated in Figure 2, this system integrates three functionally distinct components: a visible‐light‐absorbing Ir(III) complex, a semiconductor electron relay (TiO₂), and a surface‐anchored molecular porphyrin catalyst. IrPE refers to [(C )₂Ir(III)(4,4′‐Y₂‐bpy)]⁺, where C  = 1‐phenylisoquinoline and Y = CH₂PO(OEt)₂; IrP is the phosphonic acid analog (Y = CH₂PO(OH)₂), chemically modified for robust immobilization on TiO₂. The porphyrin catalysts include M‐TCPP (M = Fe(III), Co(III), Ni(II), or Cu(II)), bearing four carboxylate groups for covalent grafting to the oxide surface, and M‐TPPOMe, a methyl‐protected analog used in control experiments under homogeneous conditions. The sacrificial electron donor (dimethylphenylbenzimidazoline, BIH) facilitates reductive quenching of the photoexcited Ir(III) complex under visible light irradiation (λ >500 nm). Under visible light irradiation, the ternary hybrid system achieves turnover numbers (TONs) of up to 3000 for CO and 400 for H₂ over 18 h, significantly surpassing the performance of the corresponding homogeneous systems. Interestingly, the Cu‐TCPP favors H₂ evolution (TONs of >3500), consistent with electrochemical data indicating facile hydride formation and a mechanistic divergence from the Fe, Co, and Ni congeners. These findings position TiO₂ immobilized porphyrin catalysts as efficient, durable, and tunable platforms for solar‐driven CO₂ reduction, and highlight the broader potential of dye‐sensitized hybrid architectures in solar fuel production.

FIGURE 2.

Schematic representation of the ternary hybrid photocatalytic system composed of Ir(III) photosensitizer (IrP), TiO₂ electron relay, and M‐TCPP catalyst (M = Fe, Co, Ni, and Cu) co‐immobilized on TiO₂ particles.

2. Results and Discussion

2.1. Preparation and Characterization of the Porphyrin‐Based Photocatalytic System

All metal–porphyrin complexes, including M‐TPPOMe and M‐TCPP (M = Fe, Co, Ni, and Cu), were synthesized following literature procedures with minor modifications [56], as summarized in Scheme S1 of the Supporting Information (SI). The heteroleptic Ir(III) photosensitizers (IrPE and IrP) were prepared according to our previous report [39]. To evaluate the photocatalytic behavior of M‐TCPP under homogeneous and heterogeneous conditions, two systems were constructed: (i) a mixed homogeneous solution containing IrP and M‐TCPP and (ii) a heterogeneous ternary hybrid formed by coanchoring IrP and M‐TCPP onto TiO₂ via carboxylate‐mediated adsorption (Figure 2). The electronic interaction between M‐TCPP and the TiO₂ scaffold was examined by comparing the solution‐phase UV–vis spectra of M‐TCPP with the diffuse‐reflectance spectra (DRS) of M‐TCPP/TiO₂ composites. Upon immobilization, Co‐TCPP exhibited distinct red shifts in both the Soret (˜420 nm) and Q‐band (˜540 nm) regions (Figure 3), indicative of significant electronic coupling between the porphyrin π‐system and the electron‐deficient n‐type TiO₂ surface. Similar trends were observed for other M‐TCPP derivatives (Figure S1). These spectral perturbations, together with the pronounced depletion of porphyrin absorbance in the post‐adsorption supernatants, confirm quantitative binding of M‐TCPP to TiO₂ through carboxylate anchoring (Figures S2 and S3 of the Supporting Information). This interfacial binding enables through‐structure electronic communication between the lowest unoccupied molecular orbitals of porphyrin molecule and the conduction band of TiO₂, facilitating efficient electron transfer from photoreduced TiO₂ (TiO₂(e^‐)) to the surface‐bound porphyrin catalyst under illumination. Notably, anchoring M‐TCPP on the TiO₂ surface reduces the local electron density at the metal center relative to the homogeneous phase, thereby lowering the energetic barrier for sequential single‐electron reduction events. This modulation of the electronic environment is directly correlated with the observed enhancement in photocatalytic CO₂‐to‐CO conversion activity (vide infra).

FIGURE 3.

UV–vis spectrum of Co‐TCPP (1.0 μM in DMF) and DRS of Co‐TCPP/TiO₂ powders and bare TiO₂ particles. Right: Schematic illustration of the Co‐TCPP/TiO₂ interface, depicting the electronic coupling between surface‐anchored Co‐TCPP complex and the adjacent n‐type TiO₂ semiconductor. The proximity to TiO₂ is proposed to lower the electron density of the Co center through interfacial charge delocalization.

2.2. Photocatalytic CO₂ Conversion

The photo‐induced CO₂ reduction activities of the M‐TCPP complex were examined on two different platforms: the mixed homogeneous system (IrP + M‐TCPP) and the M‐TCPP catalyst‐heterogenized TiO₂ hybrid system (IrP/TiO₂/M‐TCPP). All samples, consisting of CO₂‐saturated DMF solution with proton sources (16.7 vol% TEOA, 3 or 10 vol% TFE, 10 vol% phenol, 3 or 10 vol% H₂O) and 0.1 M BIH (sacrificial electron donor), were evaluated for the CO₂‐to‐CO conversion activities of the porphyrin catalyst. It is important to note that TEOA does not function as a proton source under the reaction conditions because its intrinsic acidity is too weak (pKa ≈ 14–16). Instead, as reported for related systems [57], BIH serves as the primary sacrificial electron and proton donor, and its oxidized form (BIH^•+) engages in hydrogen bonding with TEOA, subsequently transferring a proton to generate protonated TEOA (TEOA(H⁺)). The formation of TEOA(H⁺) effectively establishes a secondary proton reservoir with a pKa of approximately 8 [58]. For this evaluation, all photoreactions were performed under LED light of wavelengths longer than 500 nm obtained by passing the light through a K₂CrO₄ solution layer of 1.0 cm path length (see the comparative UV−vis absorption spectra of the Ir(III) photosensitizer in Figure S4 of the Supporting Information). Table 1 lists the turnover numbers (TONs, mol of products/mol of catalyst) and the net formation amounts (μmol) of the two reduction products (CO and hydrogen) as a function of irradiation time for both the homogeneous and ternary hybrid systems. Figure 4 and Figures S5–S7 depict the photocatalysis results (TONs and μmol for CO and H₂ production) of homogeneous and heterogeneous systems performed in CO₂‐saturated DMF solutions containing 0.1 M BIH and various proton sources (TEOA, TFE, H₂O, or phenol). The 17 h photoreaction of the homogeneous system (IrP + M‐TCPP) with addition of 16.7 vol% TEOA produced small but detectable amounts of CO (2.6–7.4 μmol) and H₂ (6.0–12.0 μmol) (entries 9–12 in Table 1 and Figure S5).

TABLE 1.

Results of the Visible‐Light‐Driven CO₂ Reduction with Porphyrin Catalyst‐Based Photocatalytic Systems in Different Conditions.

Entrya	System	Proton sources	CO			H2
			TON / t irr	µmol	ΦCO b (%)	TON / t irr	µmol	ΦH2 b (%)
1	IrP (2 µmol)/TiO2/Fe‐TCPP (0.01 µmol)	16.7 vol% TEOA	2932 ± 212 / 17	29.32 ± 2.12	0.49	490 ± 36 / 17	4.9 ± 0.36	0.042
2	IrP (2 µmol)/TiO2/Co‐TCPP (0.01 µmol)	16.7 vol% TEOA	3020 ± 193 / 17	30.2 ± 1.93	0.47	390 ± 16 / 17	3.9 ± 0.16	0.039
3	IrP (2 µmol)/TiO2/Ni‐TCPP (0.01 µmol)	16.7 vol% TEOA	3030 ± 236 / 17	30.3 ± 2.36	0.53	320 ± 22 / 17	3.2 ± 0.22	0.045
4	IrP (2 µmol)/TiO2/Cu‐TCPP (0.01 µmol)	16.7 vol% TEOA	1830 ± 144 /17	18.3 ± 1.44	0.13	3600 ± 219 / 17	36.0 ± 2.19	0.37
5	IrP (2 µmol)/TiO2/Fe‐TCPP (0.1 µmol)	16.7 vol% TEOA	306 ± 13.6 / 35	30.6 ± 1.36		121 ± 9.2/ 35	12.1 ± 0.92
6	IrP (2 µmol)/TiO2/Co‐TCPP (0.1 µmol)	16.7 vol% TEOA	284 ± 16.0 / 35	28.4 ± 1.60		91 ± 5.2 / 35	9.1 ± 0.52
7	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	16.7 vol% TEOA	349 ± 15.2 / 35	34.9 ± 1.52		85 ± 5.7 / 35	8.5 ± 0.57
8	IrP (2 µmol)/TiO2/Cu‐TCPP (0.1 µmol)	16.7 vol% TEOA	116 ± 7.6 / 35	11.6 ± 0.76		331 ± 21.2 / 35	33.1 ± 2.12
9	IrP (0.3 mM) + Fe‐TCPP (0.1 mM)	16.7 vol% TEOA	740 ± 85 / 17	7.4 ± 0.85		600 ± 65 / 17	6.0 ± 0.65
10	IrP (0.3 mM) + Co‐ TCPP (0.1 mM)	16.7 vol% TEOA	390 ± 69 / 17	3.9 ± 0.69		420 ± 54 / 17	4.2 ± 0.54
11	IrP (0.3 mM) + Ni‐ TCPP (0.1 mM)	16.7 vol% TEOA	380 ± 38 / 17	3.8 ± 0.38		600 ± 91 / 17	6.0 ± 0.91
12	IrP (0.3 mM) + Cu‐ TCPP (0.1 mM)	16.7 vol% TEOA	260 ± 29 / 17	2.6 ± 0.29		1200 ± 152 / 17	12.0 ± 1.52
13	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	3 vol% TFE	144 ± 11.6 / 30	14.4 ± 1.16		n.d. g	n.d. g
14	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	10 vol% TFE	49 ± 1.3 / 22	4.9 ± 0.13		16 ± 0.7 / 22	1.6 ± 0.07
15	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	3 vol% H2O	221 ± 12.3 / 30	22.1 ± 1.23		15 ± 1.0 / 30	1.5 ± 0.10
16	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	10 vol% H2O	33 ± 1.0 / 22	3.3 ± 0.10		52 ± 3.5 / 22	5.2 ± 0.35
17	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)	10 vol% Phenol	46 ± 2.8 / 22	4.6 ± 0.28		n.d. g	n.d. g
18	IrP (2 µmol)/TiO2 c	16.7 vol% TEOA	n.d. g	n.d. g		n.d. g	n.d. g
19	TiO2/Ni‐TCPP (0.1 µmol)d	16.7 vol% TEOA	n.d. g	n.d. g		n.d. g	n.d. g
20	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)e	16.7 vol% TEOA	n.d. g	n.d. g		n.d. g	n.d. g
21	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)e	None	n.d. g	n.d. g		n.d. g	n.d. g
22	IrP (2 µmol)/TiO2/Ni‐TCPP (0.1 µmol)f	16.7 vol% TEOA	n.d. g	n.d. g		n.d. g	n.d. g
23	IrP (2 µmol)/ZrO2/Ni‐TCPP (0.1 µmol)	16.7 vol% TEOA	20	2.0		15	1.5

^a Standard photocatalytic conditions: 0.1 M BIH was employed as a sacrificial electron donor in a total solution volume of 3 mL. Reactions were carried out in CO₂‐saturated DMF/TEOA (16.7 vol% TEOA) under irradiation with λ >500 nm from a 60 W LED lamp (Cree Inc.).

^b See the detailed method for AQY measurements in the Experimental Section.

^c In the absence of Ni‐TCPP.

^d In the absence of Ir(III) photosensitizer.

^e Without electron donor (BIH).

^f In the dark.

^g Not detected.

FIGURE 4.

Time‐dependent formation profiles of H₂ and CO under visible‐light irradiation (λ >500 nm, 298 K) for the heterogeneous powder sample (10 mg, IrP (2 μmol)/TiO₂/M‐TCPP (0.01 μmol)) in CO₂‐saturated DMF/TEOA (16.7 vol% TEOA) containing BIH (0.1 M) as electron donor. M = (a) Fe, (b) Co, (c) Ni, and (d) Cu.

Under identical conditions, the heterogeneous hybrid systems (IrP/TiO₂/M‐TCPP) exhibited markedly enhanced photocatalytic performance. CO evolution proceeded with significantly elevated turnover numbers (TONs: 3030 for Fe‐TCPP, 3090 for Co‐TCPP, 3180 for Ni‐TCPP, and 1840 for Cu‐TCPP), accompanied by concomitant hydrogen generation (TONs: 490 for Fe‐TCPP, 390 for Co‐TCPP, 320 for Ni‐TCPP, and 3600 for Cu‐TCPP) (Figure 4). The apparent quantum yields (AQYs) for CO and H₂ production measured at 500 nm across the heterogeneous samples were found to range from 0.13 to 0.53% for CO and from 0.039 to 0.37% for H₂ (entries 1–4 in Table 1 and Figure S8). It should be emphasized that the measured AQY does not correspond to the true quantum yield in the strict scientific definition but rather provides a relative estimate of how efficiently photogenerated electrons are channeled into CO formation with respect to the incident light intensity, as the exact number of photons absorbed by the sensitizer cannot be accurately determined under these conditions. The relatively low AQY values likely arise, at least in part, from limited light harvesting by the dye due to significant light scattering within the particle dispersion system. To investigate the effect of proton donors on photocatalytic CO₂ reduction of heterogeneous system, we conducted a series of photocatalysis experiments with variation of proton source (TEOA, TFE, H₂O, and phenol) (see entries 13–17 of the Table 1). When 3 vol% TFE and H₂O were introduced, the heterogeneous systems generated the relatively lower CO₂‐to‐CO conversion activities (TONs, 152 for 3 vol% TFE and 221 for 3 vol% H₂O) with no hydrogen evolution. When the volume amount of TFE and H₂O was increased to 10 vol%, the TiO₂‐immobilized Ni‐TCPP produced the sharply dropped CO₂‐to‐CO conversion activities (TONs, 49 for 10 vol% TFE and 53 for 10 vol% H₂O). The pronounced decrease in CO production is likely attributable to shifts in the conduction band (CB) edge of TiO₂, which is known to be highly sensitive to the concentration of protonic sources. As established in our previous reports [53, 54, 55], increasing the concentration of TFE or H₂O (from 3 to 10 vol%) induced a pronounced positive shift in the conduction band (CB) edge potential of TiO₂ (Table 2, –1.70 to –1.18 V vs standard hydrogen electrode (SHE)), thereby attenuating the thermodynamic driving force for electron injection from the photoexcited Ir(III) photosensitizer (E _red = –1.12 V for IrPE) into the TiO₂ CB, as well as for subsequent electron transfer from TiO₂ to the adsorbed M‐TCPP (Table 2, E _red1(M^II/I) = –0.49 to –0.84 V vs SHE). Compared to the homogeneous system, the heterogeneous system exhibited higher and more durable CO₂ conversion activities (Table 1 and Figure 4). The enhanced performance of the ternary hybrid system is attributed to the delayed degradation of the photosensitizing unit, enabled by TiO₂ heterogenization, as demonstrated in our previous studies [15, 59]. In control experiments, systems lacking either functional component (porphyrin catalyst or Ir(III) photosensitizer), such as IrP/TiO₂ or TiO₂/Ni‐TCPP, showed negligible CO₂ conversion activities (entries 18 and 19 in Table 1). These results highlight that the co‐immobilization of the molecular porphyrin catalyst and Ir(III) photosensitizer is essential for the stable CO₂ reduction performance of the heterogenized hybrid system. Furthermore, in the absence of either BIH or light (entries 20 and 21 in Table 1), the IrP/TiO₂/Ni‐TCPP system exhibited no detectable catalytic activity. This result unambiguously confirms that CO₂ reduction by the TiO₂‐anchored porphyrin catalyst is initiated through a photochemical electron transfer, wherein the Ir(III) photosensitizer absorbs visible light (IrP + hν → IrP*) and subsequently undergoes reductive quenching by BIH (IrP* + BIH → IrP^•– + BIH^•+). It should be noted that unlike the other metal porphyrin catalysts, Cu‐TCPP generates mainly hydrogen (TONs, 1840 for CO; 3600 for H₂). These results imply that copper may exhibit a higher propensity for metal hydride formation compared to other metals.

TABLE 2.

Electrochemical data of porphyrin catalyst and Ir(III) photosensitizer.

Complexes	E 1/2 ox2 (V)a	E 1/2 ox1 (V)	E 1/2 red1 (V)	E 1/2 red2 (V)a	E 1/2 red3 (V)a	E 1/2 red4 (V)a
Fe‐TPPOMe	—	1.12b	0.07	–0.75	–1.31	–1.69b
Co‐TPPOMe	1.42	0.76b	–0.49	–1.59	–1.72	—
Ni‐TPPOMe	—	1.31	–0.84	–1.21	–1.49	–1.69
Cu‐TPPOMe	—	1.37	–0.84	–1.29	–1.57	–1.80
IrPE	—	1.49	–1.12	–1.49	–1.72	—

^a E _pa = anodic peak potential, E _pc = cathodic peak potential, and E _1/2 = (E _pc + E _pa)/2 vs SHE.

^b Irreversible peak.

To evaluate whether intermolecular electron transfer between co‐adsorbed M‐TCPP and IrP contributes to the observed photocatalysis, we carried out control experiments using ZrO₂, a wide band‐gap semiconductor that effectively suppresses electron injection from both chromophores [60]. If direct lateral electron transfer between M‐TCPP and IrP were operative, measurable activity should persist even on ZrO₂. Instead, the ZrO₂‐supported systems displayed only minimal reactivity, indicating that such molecule‐to‐molecule electron transfer is negligible (entry 23 in Table 1) [61]. This behavior mirrors our prior findings using various dye–catalyst combinations, where replacing TiO₂ with ZrO₂ resulted in a near‐complete loss of photocatalytic activity [38, 39, 41, 44, 59]. Additional evidence comes from the large disparity in surface loading: IrP is immobilized on TiO₂ at levels 20–200 times higher than those of M‐TCPP. Such a substantial population difference strongly biases initial photoexcitation and electron‐injection events toward IrP, making IrP → TiO₂ charge transfer overwhelmingly dominant. The injected electrons stored in TiO₂ (TiO₂(e^‐)) can then be delivered to the sparsely distributed M‐TCPP molecules as needed, consistent with well‐established models for semiconductor‐mediated electron relays in dye‐sensitized photocatalysis [15, 51, 62, 63, 64].

A ¹³C isotope tracer experiment was performed on the Ni‐TCPP‐heterogenized ternary hybrid system in a ¹³CO₂‐saturated mixed DMF‐d ₇/TEOA solvent (16.7 vol% TEOA, 0.1 M BIH) to identify the carbon source of the CO product. After 18 h of photocatalysis, the ¹³C NMR spectrum exhibited a distinct resonance at 184.5 ppm, attributable to ¹³CO, thereby confirming that CO originates from the introduced ¹³CO₂ gas (Figure 5). Complementary gas chromatography–mass spectrometry (GC–MS) analysis further revealed that the isotopic abundance of ¹³CO exceeded 97% (Figure S9 of the Supporting Information). These results unambiguously establish CO₂ as the carbon source of the observed CO.

FIGURE 5.

¹³C NMR spectra of the ternary hybrid system (IrP/TiO₂/Ni‐TCPP; 10 mg) in DMF‐d ₇ containing BIH (0.1 M) and TEOA (16.7 vol%, 2 mL total volume), recorded (a) before and (b) after 18 h of visible‐light irradiation (λ >500 nm). The signals marked with ▪, ▾, and • correspond to TEOA, DMF, and BIH, respectively.

2.3. Electrochemical Characterization

To elucidate the electrochemical properties of porphyrin catalysts and their relevance to CO₂ reduction, we examined the corresponding M‐TPPOMe analogs by cyclic voltammetry (CV). The CV profile of Fe‐TPPOMe (Figure 6a) displays multiple redox events over the range of 1.5 to –2.1 V (vs SHE). Notably, three successive reduction waves are observed at 0.07, –0.75, and –1.31 V, consistent with literature‐reported values [65], and can be attributed to stepwise one‐electron reductions from Fe(III) to Fe(0), accompanied by chloride ligand dissociation. An irreversible oxidation peak near + 1.12 V corresponds to the Fe(III)/Fe(IV) redox couple. The CV of Co‐TPPOMe (Figure 6b) reveals one irreversible oxidation event at + 1.42 V and two reversible reduction waves at –0.49, –1.59, and –1.72 V, assignable to Co^II/III, Co^II/I, and [Co^IL]^0/− processes, respectively [66]. For Ni‐TPPOMe (Figure 6c), a reversible oxidation wave is detected near + 1.31 V, along with four reversible reduction peaks between –0.84 and –1.69 V (vs SHE). Among them, the oxidation and initial two reduction waves (at –0.84 and –1.21 V) are ascribed to the Ni^II/III, Ni^II/I, and Ni^I/0 couples, respectively, while the more negative reductions are attributed to electron uptake by the porphyrin ligand framework [67]. Similar electrochemical features are observed for Cu‐TPPOMe (Figure 6d), which exhibits an oxidation event at + 1.37 V and two reversible reduction waves at –0.84 and –1.29 V, corresponding to Cu^II/III, Cu^II/I, and Cu^I/0 redox processes [68]. Additional reductions at –1.57 and –1.80 V (vs SHE) are assigned to stepwise reductions of the porphyrin macrocycle ([Cu⁰L]^0/– and [Cu⁰L]^–/2–).

FIGURE 6.

Schematic illustration of stepwise reduction pathways for M‐TPPOMe complexes (M = Fe, Co, Ni, Cu), followed by cyclic voltammograms of (a) Fe‐TPPOMe, (b) Co‐TPPOMe, (c) Ni‐TPPOMe, and (d) Cu‐TPPOMe (1 mM) recorded in Ar‐saturated DMF containing 0.1 M TBAP at a scan rate of 100 mV/s.

To investigate the electrochemical properties of porphyrin catalysts under conditions relevant to photocatalytic reaction, CV measurements were carried out in the presence of TEOA under either an Ar or CO₂ atmosphere. As shown in Figure 7, the addition of TEOA under an Ar atmosphere rendered initially reversible reduction peaks irreversible. Under CO₂‐saturated conditions with 16.7 vol% TEOA, a new reduction peak emerged alongside a notable increase in catalytic current, which is attributed to the formation of key intermediates via CO₂ coordination to the reduced metal center.

FIGURE 7.

Cyclic voltammograms of (a) Fe‐TPPOMe, (b) Co‐TPPOMe, (c) Ni‐TPPOMe, and (d) Cu‐TPPOMe (1 mM) recorded in DMF containing 0.1 M TBAP at a scan rate of 100 mV/s under three different conditions: Ar‐saturated DMF (black line), Ar‐saturated DMF with 16.7 vol% TEOA (blue line), and CO₂‐saturated DMF with 16.7 vol% TEOA (red line).

Notably, Cu‐TPPOMe exhibited a substantial increase in catalytic current upon addition of H₂SO₄, in stark contrast to the Fe‐, Co‐, and Ni analogs, which showed negligible changes under identical conditions (Figure 8). This acid‐dependent current enhancement centered near –1.0 V is consistent with the involvement of a Cu–hydride intermediate. Although the Cu–H species is not directly observable under our electrochemical or photocatalytic conditions, such behavior aligns closely with previously reported mechanistic studies on Cu‐porphyrin systems, in which protonation of a two‐electron‐reduced Cu state generates a short‐lived but catalytically competent Cu–H intermediate that drives H₂ evolution [69]. Likewise, Peng et al.. demonstrated that electronically tuned Cu–porphyrins undergo a reduction–protonation sequence in which protonation of reduced Cu(I) or Cu(0) species to form Cu–H is obligatory for H₂ formation, even though the hydride itself remains experimentally elusive [70].

FIGURE 8.

Cyclic voltammograms of (a) M‐TPPOMe (M = Fe, Co, Ni, Cu; 1 mM) in Ar‐saturated DMF solvent containing 140 mM TFA, highlighting the catalytic current enhancement specific to Cu‐TPPOMe. (b) Cu‐TPPOMe in Ar‐saturated DMF with increasing concentrations of TFA (0 to 140 mM), showing a monotonic increase in cathodic peak current, indicative of the stepwise formation of Cu–H intermediates. (c) Comparative CVs of Cu‐TPPOMe in the presence of different proton donors (TEOA, H₂SO₄, and TFA). (d) Proposed electron transfer and mechanistic pathways for Cu‐TPPOMe, involving sequential reduction and protonation steps to generate Cu–H intermediates and facilitate CO₂‐to‐CO conversion under photocatalytic conditions.

In our system, the pronounced activation of this Cu–H pathway provides a mechanistic basis for the uniquely high H₂ evolution activity of TiO₂‐anchored Cu porphyrin catalyst, distinguishing it from the Fe, Co, and Ni analogs that preferentially undergo CO₂‐to‐CO conversion. Further support for this interpretation is obtained from CV experiments employing trifluoroacetic acid (TFA) as an alternative proton source (Figure 8). Among all metal–porphyrin complexes examined, only Cu‐TPPOMe displayed a clear, concentration‐dependent increase in current at –1.30 V (vs SHE), indicative of accelerated Cu–H formation. Taken together, these findings indicate that Cu porphyrins access hydride‐mediated H₂ evolution more readily than their Fe, Co, or Ni counterparts, thereby favoring proton reduction via a Cu–H intermediate rather than CO₂ reduction via a Cu–COOH pathway (Figure 8d) [26, 69, 70].

2.4. Discussions

The pronounced discrepancy in catalytic performance between the homogeneous and heterogeneous systems highlights the critical role of TiO₂ immobilization in modulating the reactivity and stability of porphyrin‐based catalysts for CO₂ photoreduction. Under identical photocatalysis conditions, the mixed homogeneous systems comprising IrP and M‐TCPP yielded only modest amounts of CO and H₂, likely due to catalyst deactivation via photodegradation of molecular components and unwanted reaction between reactive species in solution. In sharp contrast, the heterogeneous ternary system (IrP/TiO₂/M‐TCPP) exhibited dramatic improvements in both catalytic turnover and durability, achieving TON_CO values exceeding 3000 for TiO₂‐anchored M‐TCPP (M = Fe, Co, and Ni) over 18 h of irradiation. The superior performance of the heterogenized system arises from multiple synergetic effects facilitated by the TiO₂ scaffold. First, covalent anchoring of the porphyrin catalysts via peripheral carboxylate groups enables intimate electronic coupling between the porphyrin π*‐orbitals and the TiO₂ conduction band, as evidenced by the red‐shifted absorption features observed in DRS spectra (Figure 3). This through‐structure electronic communication promotes efficient sequential electron transfer from reduced TiO₂(e^‐) to the immobilized porphyrin catalysts, bypassing the need for high‐energy M⁰ intermediates typical of homogeneous systems [26]. Second, the n‐type character of TiO₂ exerts a mild electron‐withdrawing effect, lowering the electron density at the metal center and thereby facilitating redox transitions at reduced overpotentials—a key advantage for catalysis under low driving force conditions [41, 44]. Moreover, TiO₂ functions as a dynamic electron reservoir: electrons injected from the reductively quenched Ir(III) photosensitizer are stored in shallow trap states near the conduction band edge and subsequently relayed to the immobilized catalyst as needed [15]. This temporal electron storage facilitates multielectron accumulation at the catalyst interface, which is indispensable for multistep PCET sequences, such as the stepwise reduction of CO₂ via M–COOH and M–CO intermediates. The resulting unidirectional charge transfer cascade (IrP → TiO₂ → M‐TCPP) minimizes back‐electron transfer and enhances the overall efficiency of photochemical energy conversion.

Additionally, heterogenization imparts structural and chemical stability to the molecular components. Immobilization of the porphyrin catalysts on TiO₂ mitigates undesired aggregation and photodecomposition, while spatial confinement at the oxide surface ensures sustained access to reagents (CO₂ and H⁺) for CO₂ reduction and H₂ production [15, 38, 40]. Note that the identity and concentration of the proton source further illustrate the tunability of the system: variations in the proton source's acidity and solvation properties shift the TiO₂ conduction band edge, thereby modulating the thermodynamic driving force for electron injection and subsequent catalyst reduction (entries 13–17 in Table 1) [53, 54, 55].

Mechanistic insights from electrochemical analysis reveal that the identity of the metal center plays a decisive role in dictating the product distribution in the ternary photocatalytic system. Fe‐, Co‐, and Ni‐porphyrins preferentially promote CO₂‐to‐CO conversion through the classical sequence involving M–CO₂ and M–COOH intermediates, consistent with well‐established pathways for first‐row metalloporphyrins [26]. In contrast, TiO₂‐anchored Cu‐TCPP exhibits a distinct reactivity pattern, selectively facilitating H₂ evolution. This divergence is strongly correlated with the acid‐dependent current enhancement observed exclusively for Cu‐TCPP upon addition of TFA or H₂SO₄ (Figure 8), behavior that is consistent with the formation of a transient Cu–hydride intermediate [69, 70]. Figure 9 summarizes the proposed photo‐induced electron transfer and catalytic pathways operative in the TiO₂‐immobilized porphyrin system. Reductive quenching of photoexcited IrP* by BIH initiates electron injection into the TiO₂ semiconductor [39], generating TiO₂(e^‐) species that subsequently reduce the surface‐bound metalloporphyrin [15, 50, 51, 62, 63, 64]. For Fe–, Co–, and Ni–porphyrins, the one‐electron‐reduced M(I) state binds CO₂ to yield M(I)–CO₂ (step a ), followed by protonation to M–COOH (step b ) and a second reduction/protonation sequence that affords M–CO and releases H₂O (step c ). CO dissociation regenerates the M(I) species (step d ), completing the CO‐forming cycle [26]. For Cu‐TCPP, an alternative protonation‐first pathway may dominate. The singly reduced [Cu(I)–porphyrin]⁻ species can undergo protonation, followed by TiO₂(e⁻)‐driven one‐electron reduction, to generate a catalytically competent hydride species [Cu(II)–H]^‐ (step b′ ). This intermediate can subsequently react with an additional proton to produce H₂, thereby enabling a hydride‐mediated H₂ evolution pathway that competes effectively with CO₂ binding and reduction [71, 72].

FIGURE 9.

Schematic of the electron transfer processes in the present ternary hybrid system and proposed mechanistic pathways of the TiO₂‐immobilized M‐TCPP complex for photochemical CO₂‐to‐CO conversion in the heterogeneous ternary system.

On the other hand, the BI^• species, generated via deprotonation of the one‐electron‐oxidized form of sacrificial donor (BIH^•+), exhibit a high reduction potential (−1.44 V vs SHE), enabling electron transfer to various acceptors such as neutral Ir(III) photosensitizer (E _red1 = −1.12 V vs SHE) and dispersed TiO₂ particles in mixed DMF/TEOA (16.7 vol% TEOA) solvent (E _fb = −1.34 V vs SHE) (Figure 9) [73]. Taken together, these findings demonstrate that TiO₂ not only stabilizes and activates molecular porphyrin catalysts but also serves as a multifunctional platform for enhancing charge separation, electron delivery, and product selectivity in visible light‐driven CO₂ reduction. The superior performance of the heterogenized system relative to its homogeneous counterpart clearly establishes dye‐sensitized TiO₂/porphyrin hybrid systems as an efficient and tunable strategy for solar‐to‐chemical energy conversion using earth‐abundant organometallic catalyst.

3. Conclusions

This study establishes dye‐sensitized TiO₂/porphyrin hybrid systems as effective and robust platforms for visible light‐driven CO₂ reduction using earth‐abundant 3d transition metals. Integration of Ir(III) photosensitizers, TiO₂ electron relays, and carboxylate‐functionalized porphyrin catalysts (M‐TCPP; M = Fe, Co, Ni, and Cu) into a ternary architecture leads to substantial improvements in catalytic turnover. The TiO₂ scaffold functions not only as a directional electron mediator and multielectron reservoir but also as a stabilizing matrix that mitigates porphyrin decomposition and suppresses deleterious side reactions. Under optimized photocatalysis conditions (λ >500 nm, CO₂‐saturated DMF containing 16.7 vol% TEOA), TiO₂‐immobilized Fe–, Co–, and Ni–porphyrin catalysts (Fe‐TCPP, Co‐TCPP, and Ni‐TCPP) selectively promote CO evolution with turnover numbers exceeding 3000. In contrast, the Cu‐TCPP system favors H₂ production, consistent with electrochemical evidence for a mechanistically distinct, protonation‐first pathway involving Cu–hydride intermediates. These findings highlight the multifaceted role of TiO₂ not only as an electron mediator and reservoir but also as a structural modulator of catalytic reactivity. Taken together, this work showcases the potential of dye‐sensitized TiO₂/porphyrin assemblies as a modular, tunable, and scalable platform for solar‐to‐chemical energy conversion using non‐precious metal catalysts.

4. Experimental Section

4.1. General Procedure

All reagents were purchased from Aldrich and were used without further purification. All manipulations were performed under a dry nitrogen or argon atmosphere by using standard Schlenk techniques. N,N‐Dimethylformamide (DMF) was distilled from calcium hydride, and stored over molecular sieves. The ¹H and ¹³C NMR spectra were recorded by Bruker Ascend 400 spectrometry in the KBSI Ochang Center (operating at (700.1 and 176.1) MHz), respectively. The absorption and photoluminescence spectra were recorded by Agilent Technologies Cary 5000 UV–Vis–NIR spectrophotometry and VARIAN Cary Eclipse fluorescence spectrophotometry, respectively. ESI‐MS (Electrospray Ionization Mass Spectrometry) was performed by the linear ion trap of the LTQ XL ETD mass spectrometer (Thermo Fisher Scientific, LCQ Fleet Hyperbolic Ion Trap MS/MSⁿ Spectrometer). Gas chromatography–mass spectrometry (GC–MS) analyses were performed using an Agilent 7890A GC system coupled with a 5975C inert mass selective detector with Triple‐Axis capability, equipped with a SUPELCO Carboxen 1010 PLOT fused silica capillary column.

4.2. Synthesis

[Ir(C )₂(4,4′‐Y₂‐bpy)]⁺ (IrPE, C  = 1‐phenylisoquinoline; Y = CH₂PO(OEt)₂) [39], [Ir(C )₂(4,4′‐Y₂‐bpy)]⁺ (IrP, Y = CH₂PO(OH)₂) [39], and 1,3‐dimethyl‐2‐phenyl‐1,3‐dihydro benzimidazole (BIH) [74] were prepared by using previous methods. 5,10,15,20‐Tetrakis(4‐methoxycarbonylphenyl)porphyrin (TPPOMe), 5,10,15,20‐Tetrakis(4‐methoxycarbonylphenyl)porphyrin‐M (M‐TPPOMe), 5,10,15,20‐Tetrakis(4‐carboxyphenyl)porphyrin (TCPP), and [5,10,15,20‐Tetrakis(4‐carboxyphenyl)porphyrinato]‐M (M‐TCPP) were prepared by using previous methods [56].

4.3. Preparation of TiO₂

Commercially available Hombikat UV‐100 (TiO₂) was thoroughly washed with distilled water, ultrasonically treated in water, separated by centrifugation, and then dried in an oven under N₂.

4.4. Photocatalytic Reaction

A wavelength‐selective illumination system was constructed using a cylindrical quartz vessel filled with an aqueous K₂CrO₄ solution (inner optical path length: ca. 1 cm). Three high‐intensity LED lamps (Cree, Inc., 60 W) were arranged symmetrically around the cylinder. Photocatalytic samples were placed inside the cylinder on a custom‐built merry‐go‐round rotating platform, enabling uniform exposure across all vials. A K₂CrO₄ solution (0.1 M) was used as a liquid optical filter. A portion of the solution was extracted and analyzed by UV–vis spectroscopy, which confirmed complete absorption of wavelengths below 500 nm and thereby ensured that only λ >500 nm light was transmitted to the photocatalytic samples (Figure S10a of the Supporting Information). The light intensity within the center of the cylinder was measured using a calibrated laser/optical power meter (FieldMate, Coherent Inc.). The measured irradiance under typical photocatalytic operation was 59.8 mW (Figure S10b). Dispersions of IrP/TiO₂/M‐TCPP particles (10 mg containing 2 μmol of IrP and 0.01 or 0.1 μmol of M‐TCPP) in 3 mL of DMF with various proton sources (16.7 vol% TEOA, 3 or 10 vol% TFE, 10 vol% phenol, or 3 or 10 vol% H₂O) and 0.1 M BIH were placed in a Pyrex cell (1 cm path length; total volume 12.8 mL), bubbled with CO₂ for 30 min, sealed with a septum, and irradiated with a 60 W LED lamp (Cree) under continuous stirring. A series of samples was set on a homemade merry‐go‐round apparatus and then irradiated under magnetic stirring with an LED lamp (60 W, model Fc‐6051, Cree Inc.). Homogeneous photoreactions were conducted in 3 mL of a DMF/TEOA mixture (16.7 vol% TEOA) containing 0.3 mM IrP, 0.1 mM M‐TCPP, and 0.1 M BIH. The amounts of CO evolved in the headspace were quantified by gas chromatography (7890B GC equipped with a TCD detector) using a SUPELCO Carboxen 1010 PLOT Fused Silica Capillary column.

4.5. Apparent Quantum Yields Measurement

The quantum yields for CO and H₂ production (Φ) were determined for the heterogeneous system (IrP/TiO₂/M‐TCPP). A 495–505 nm band‐pass filter was used to isolate 500 nm excitation light from the emission of a 300 W Xe lamp (Asahi Spectra, MAX‐350). The quantum yields were measured using an Asahi Spectra PRA‐201 apparatus equipped with a Xe lamp (Asahi Spectra, MAX‐350) and calculated according to Equation (1) (Figure S11 of the Supporting Information). The reported AQY values may be lower than instrument records due to diffuse scattering and multiple scattering losses arising from suspended TiO₂ particles; nonetheless, these data remain reliable for relative comparison.

$$\mathrm{QY}\left(\mathrm{\%}\right)=\frac{\text{amount of product generated per unit time}}{\text{number of incident photons per unit time}}\times 100$$ (1)

4.6. Electrochemical Measurements

CV measurements were carried out in distilled DMF solutions, degassed with Ar or CO₂, containing 1 mM M‐TPPOMe and 0.1 M tetrabutylammonium perchlorate (TBAP), in the presence of various proton sources (none, 16.7 vol% TEOA, or TFA). Experiments were performed at room temperature using a CHI660e electrochemical analyzer (CH Instruments). All experiments were recorded using a glassy carbon working electrode (d = 2.0 mm), a platinum wire counter electrode, and a nonaqueous Ag/AgNO₃ (with inner electrolyte: 0.01 M AgNO₃ and 0.1 M Bu₄NPF₆ in DMF) reference electrode. Ferrocene/ferrocenium (Fc/Fc⁺) was used as an internal potential marker. The potential values of the compounds were calibrated against the ferrocene/ferrocenium (Fc/Fc⁺) redox couple and subsequently converted to the SHE scale, using a Fc/Fc⁺ potential of 0.624 V versus SHE, to enable direct comparison with literature‐reported values [75].

Supporting Information

Additional supporting information can be found online in the Supporting Information section. Supporting Scheme S1: Synthetic routes of the porphyrin catalysts (M‐ TPPOMe and M‐TCPP, M = Fe, Co, Ni ,Cu)^[S1] Supporting Fig. S 1: UV–vis spectrum of M ‐TCPP (1.0 μM in DMF) and diffuse‐reflectance spectra (DRS) of M ‐TCPP/TiO₂ powders and bare TiO₂ particles. M = (b) Fe, (b) Co, (c) Ni, and (d) Cu. Supporting Fig. S 2: Comparable absorption spectra of DMF solution of M‐TCPP before and after adsorption process of M‐TCPP with TiO₂ particles. Supporting Fig. S 3: Photographs of bare TiO₂ powders and TiO₂/M‐TCPP hybrid samples prepared with different metalloporphyrins (Fe‐TCPP, Ni‐TCPP, Co‐TCPP, and Cu‐TCPP). The progressive color variation from white (bare TiO₂) to brown‐pink or reddish‐brown color reflects the incorporation of M‐TCPP catalysts onto the TiO₂ surface. Supporting Fig. S 4: (a) UV–vis spectrum of Ir(III) photosensitizers (10 μM in DMF) and (b) diffuse‐reflectance spectra (DRS) of IrP/TiO₂ powders and bare TiO₂ particles. Supporting Fig. S 5: Time‐dependent formation profiles of H₂ and CO under visible‐light irradiation (λ >500 nm, 298 K) for the homogeneous solution containing IrP (0.3 mM) and M‐TCPP (0.1 mM) in CO₂‐saturated DMF/TEOA (16.7 vol% TEOA) containing BIH (0.1 M) as electron donor. M = (a) Fe, (b) Co, (c) Ni, and (d) Cu. Supporting Fig. S 6: Time‐dependent formation profiles of (a) CO and (b) H₂ under visible‐light irradiation (λ >500 nm, 298 K) for the heterogeneous powder sample (10 mg, IrP (2 μmol)/TiO₂/M‐TCPP (0.1 μmol)) in CO₂‐saturated DMF/TEOA (16.7 vol% TEOA) containing BIH (0.1 M) as electron donor. Supporting Fig. S 7: Plots of CO and H₂ formation vs irradiation time for the heterogeneous powder sample (10 mg, IrP (2 μmol)/TiO₂/M‐TCPP (0.1 μmol)) in CO₂‐saturated DMF/additive solution containing BIH (0.1 M) as electron donor (additive = (a) 3 vol% H₂O, (b) 10 vol% H₂O, (c) 3 vol% TFE, (d) 10 vol% TFE, and (e) 10 vol% phenol). Supporting Fig. S 8: (a) Photon‐dependent CO evolution for the heterogeneous sample IrP (2 μmol)/TiO₂/M‐TCPP (0.01 μmol) (M = Fe, Co, Ni, Cu) with the corresponding apparent quantum yields (AQYs), reported as relative estimates due to light scattering in the TiO₂ dispersion. (b) Photon‐dependent H₂ evolution under identical conditions, showing the distinct apparent Φ_hydrogen values for each metal–porphyrin system. Supporting Fig. S 9: Gas chromatography mass spectrometry (GC‐MS) of gas in the reaction vessel before (black line) and after (red line) photoreaction and mass spectra of each retention time ((a) 3.2 min, (b) 3.9 min, and (c) 24.1 min): GC‐MS spectra was measured by Agilent Technologies 7890A GC equipped with 5975C inert MSD with Triple‐Axis detector using a SUPELCO CarboxenTM 1010 PLOT Fused Silica Capillary column. Supporting Fig. S 10: (a) Photograph of the cylindrical K₂CrO₄ solution filter (inner thickness ≈1 cm) used to achieve selective λ >500 nm irradiation, together with the UV–vis spectrum of the extracted solution confirming complete absorption below 500 nm. (b) Photocatalytic irradiation setup showing the merry‐go‐round sample holder placed inside the filter cylinder. The internal irradiance measured at the sample position is 59.8 mW, as determined using a calibrated laser power meter. Supporting Fig. S 11: Photographs and instrument settings for the quantum yield (QY) measurement using the Asahi Spectra PRA‐201 analysis system equipped with a MAX‐350 monochromatic light source. The excitation wavelength was set to 500 nm using the internal filter wheel (Filter position F2: 500 nm), and the calibrated excitation light intensity at the sample position was 15.0 mW, as shown in the instrument settings panel.

Funding

This study was supported by National Research Foundation of Korea (NRF) (RS‐2025‐00562897) and Korea University Grant.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Material

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.