Design Of Catalysts Research Articles

B, S co-doped mesoporous carbon as an efficient catalyst for H2O2 electrosynthesis in natural seawater

The small-scale electrochemical synthesis of hydrogen peroxide (H2O2) in marine environments is of great significance for environmental wastewater treatment, and the design of electrocatalysts for complicated marine conditions is the key for this technology. Although B-doped carbon materials have been demonstrated to be resistant to Cl− poisoning, the deposition of Ca2+ on the surface of catalyst in seawater and the sluggish reaction kinetics of non-metallic catalysts remain to be solved. Herein, B, S co-doped mesoporous carbon materials were synthesized by a simple NaCl template method. The H2O2 yield of the as-prepared sample was up to 3.66 mol/(gcat h) with a Faradaic Efficiency (FE) of 97.98 % at 100 mA/cm2 in natural seawater, and achieved 100 % inactivation of bacteria and Chlorella within 5 min at a low current density. Based on the structural characterization and electrochemical results, the S-containing structure on the carbon matrix suppresses Ca2+ deposition due to its excellent proton adsorption activity. Furthermore, the electron transfer process between the reactant and BC3 was enhanced by the abundant mesoporous structure, so as to optimize the sluggish oxygen reduction reaction (ORR) kinetics on BC3. This work offers a strategy for inactivation of marine microorganisms and the degradation of organic pollutants, which promotes the advancement of marine wastewater treatment technology, and also provides reference for the design of metal free catalysts used in marine environments.

Novel Alkane Dehydrogenation Routes via Tailored Catalysts

Alkane dehydrogenation to alkene represents a promising alternative route for conventional petroleum cracking processes heavily reliant on fossil resources. This Concept outlines the latest advancements of novel alkane dehydrogenation routes that offer potential solutions to intrinsic problems existing in routine direct dehydrogenation or oxidative dehydrogenation reactions, including CO2 assisted alkane dehydrogenation, chemical looping alkane oxidative dehydrogenation, dual‐path dehydrogenation and auto‐accelerated multiple alkane dehydrogenation route. These novel reaction routes exhibit unique advantages in enhancing the activity, selectivity and stability of the alkane dehydrogenation reaction system. The rational design of novel catalysts and proper choice of the reaction condition that well‐fitted these reaction routes is the key for achieving highly efficient and stable alkane dehydrogenation process. These present examples in this manuscript also highlight the importance of fundamental in‐depth understandings on the mechanism and structure‐function relations, which is the foundation for developing novel alkane dehydrogenation routes for practical applications.

Tandem catalysis of furfural to γ-valerolactone over polyoxometalate-based metal-organic frameworks: Exploring the role of confinement in the catalytic process

Polyoxometalate-based metal-organic frameworks (POMOFs) with bifunctional Lewis-Brønsted acid were designed by confining polyoxometalates (POMs: phosphotungstic acid (PW), phosphomolybdic acid (PMo) or silicotungstic acid (SiW)) within the Zr-MOFs. The confined POMs not only have their own Brønsted acid, but can also regulate the Lewis acid strength of the POMOFs via charge regulation, ultimately allowing the POMOFs to display superior catalytic performance in the tandem catalysis of furfural (FAL) to γ-valerolactone (GVL). Among the three catalysts (POMOF-PW, POMOF-PMo and POMOF-SiW), a complete conversion of FAL was achieved, while POMOF-PW exhibited the highest GVL yield (58.1 %) at 160 °C within 23 h. The leaching experiment and structural characterization of the catalyst before and after recycling confirmed the stability of the catalyst structure. The humin originates from the polymerization of furan ring-containing compounds driven by Brønsted acid. The adsorption configuration of the substrate molecules on the catalyst offers a new explanation for the tandem reaction process, i.e. that perpendicular adsorption performs only the first step of the tandem reaction, while the coexistence of perpendicular and horizontal adsorption permits the entire reaction process. This work gives unique insights into the tandem reaction of FAL and can guide the design of efficient tandem reaction catalysts.

Enhancing hydrogen evolution and oxidation kinetics through oxygen insertion into nickel lattice

Nickel-based materials, including metallic Ni and Ni oxide, have been widely studied in the exploration of non-precious-metal hydrogen electrocatalysts, but neither pure Ni nor NiO is ideal for the hydrogen evolution reaction (HER) and hydrogen oxidation reaction (HOR). In this paper, an oxygen insertion strategy was applied on nickel to regulate its hydrogen electrocatalytic performance, and the oxygen-inserted nickel catalyst was successfully obtained with the assistance of tungsten dioxide support (denoted as O-Ni/WO2). The partial insertion of oxygen in Ni maintains the face-centered cubic arrangement of Ni atoms, simultaneously expanding the lattice and increasing the lattice spacing. Consequently, the adsorption strength of *H and *OH on Ni is optimized, thus resulting in superior electrocatalytic performance of O-Ni/WO2 in alkaline HER/HOR. The Tafel slope of O-Ni/WO2@NF for HER is 56 mV dec−1, and the kinetic current density of O-Ni/WO2 for HOR reaches 4.85 mA cm−2, which is ahead of most currently reported catalysts. Our proposed strategy of inserting an appropriate amount of anions into the metal lattice could provide more possibilities for the design of high-performance catalysts.

Design and optimization of CuCo-LDO catalysts via ZIF-67 sacrificial templates for enhanced toluene oxidation: A comprehensive study on morphology, structure, and catalytic activity

The investigation into the development of an active transition-metal oxide catalyst with controllable morphology for the degradation of volatile organic compounds (VOCs) is a subject of considerable importance and warrants further exploration. In this study, ZIF-67 was prepared and employed as a sacrificial template, with varying quantities of Cu ions incorporated to synthesize layered copper-cobalt double oxide (CuCo-LDO) featuring a modifiable surface state. A comprehensive characterization of the CuCo-LDO catalysts was conducted, encompassing structural properties, morphology, surface chemical state, and redox properties, utilizing a range of analytical techniques. Subsequent evaluations included testing the catalytic activity for toluene oxidation and assessing stability performance. The results revealed that differing Cu contents played a pivotal role in inducing changes in the morphology and structure of CuCo-LDO, thereby significantly influencing its oxidation activity towards toluene. Notably, the flower-like structure of 3CuCo-LDO (Cu% = 20.14 wt.%) exhibited outstanding toluene oxidation activity attributable to its unique structure and composition. This research contributes novel insights to the design of highs-performance catalysts targeting volatile organic compounds.

Electrochemical CO2 Activation and Valorization on Metallic Copper and Carbon‐Embedded N‐Coordinated Single Metal MNC Catalysts

AbstractThe electrochemical reductive valorization of CO2, referred to as the CO2RR, is an emerging approach for the conversion of CO2‐containing feeds into valuable carbonaceous fuels and chemicals, with potential contributions to carbon capture and use (CCU) for reducing greenhouse gas emissions. Copper surfaces and graphene‐embedded, N‐coordinated single metal atom (MNC) catalysts exhibit distinctive reactivity, attracting attention as efficient electrocatalysts for CO2RR. This review offers a comparative analysis of CO2RR on copper surfaces and MNC catalysts, highlighting their unique characteristics in terms of CO2 activation, C1/C2(+) product formation, and the competing hydrogen evolution pathway. The assessment underscores the significance of understanding structure–activity relationships to optimize catalyst design for efficient and selective CO2RR. Examining detailed reaction mechanisms and structure‐selectivity patterns, the analysis explores recent insights into changes in the chemical catalyst states, atomic motif rearrangements, and fractal agglomeration, providing essential kinetic information from advanced in/ex situ microscopy/spectroscopy techniques. At the end, this review addresses future challenges and solutions related to today's disconnect between our current molecular understanding of structure–activity‐selectivity relations in CO2RR and the relevant factors controlling the performance of CO2 electrolyzers over longer times, with larger electrode sizes, and at higher current densities.

What we talk about when we talk about breaking scaling relations

For more than a dozen years, “breaking scaling relations” has been regarded by many as a top priority in computational and experimental electrocatalysis. Numerous works claim to have done it, and literature reviews are available that summarize the strategies to do it. However, in most cases, it seems more like a slogan than a clear, experimentally realizable catalyst design paradigm. It is worth asking whether we know of any materials that unmistakably break scaling relations and, more importantly, if the breaking of a problematic scaling relation is the end of all troubles in electrocatalysis. On the basis of some quantitative examples related to CO2 electroreduction and the concept of “ghost scaling relation,” I will try to address these and other fundamental questions related to scaling relations.

Electrochemical CO2 Activation and Valorization on Metallic Copper and Carbon-Embedded N-Coordinated Single Metal MNC Catalysts.

The electrochemical reductive valorization of CO2, referred to as the CO2RR, is an emerging approach for the conversion of CO2-containing feeds into valuable carbonaceous fuels and chemicals, with potential contributions to carbon capture and use (CCU) for reducing greenhouse gas emissions. Copper surfaces and graphene-embedded, N-coordinated single metal atom (MNC) catalysts exhibit distinctive reactivity, attracting attention as efficient electrocatalysts for CO2RR. This review offers a comparative analysis of CO2RR on copper surfaces and MNC catalysts, highlighting their unique characteristics in terms of CO2 activation, C1/C2(+) product formation, and the competing hydrogen evolution pathway. The assessment underscores the significance of understanding structure-activity relationships to optimize catalyst design for efficient and selective CO2RR. Examining detailed reaction mechanisms and structure-selectivity patterns, the analysis explores recent insights into changes in the chemical catalyst states, atomic motif rearrangements, and fractal agglomeration, providing essential kinetic information from advanced in/ex situ microscopy/spectroscopy techniques. At the end, this review addresses future challenges and solutions related to today's disconnect between our current molecular understanding of structure-activity-selectivity relations in CO2RR and the relevant factors controlling the performance of CO2 electrolyzers over longer times, with larger electrode sizes, and at higher current densities.

High‐Efficiency Iridium‐Yttrium Alloy Catalyst for Acidic Water Electrolysis

AbstractProton exchange membrane (PEM) water electrolysis holds great promise in revolutionizing clean energy production by enabling the efficient generation of hydrogen. Nevertheless, a formidable challenge persists in the realm of designing electrocatalysts that are both highly active and acid‐resistant during the oxygen evolution reaction (OER), thereby mitigating the substantial kinetic barrier. In this study, the facile synthesis of iridium‐yttrium (IrY) alloy nanocatalysts via a thermal shock method is introduced, which exhibits exceptional activity in the context of acidic water oxidation. Through the strategic incorporation of dispersed Y into the lattice of Ir metal, the IrY catalyst demonstrates a notably low overpotential of 255 mV at a current density of 10 mA cm−2 and showcases remarkable catalytic stability in acidic electrolytes, enduring for over 500 h with a high current density of 100 mA cm−2. Through a comprehensive set of in situ characterizations and analytical methods, the formation of a surface Ir‐based oxide layer, induced by deprotonation and electrochemical oxidation is unveiled, which is notably stabilized by the presence of Y dopants. This stabilization of the active site imparts enhanced resistance to over‐oxidation and dissolution, underpinning the exceptional stability of the catalyst. Theoretical calculations suggest that the incorporation of Y into the catalyst structure has a significant impact on enhancing the reactivity of the oxygen intermediate (O*) at adjacent Ir sites, thus lowering the overpotential and promoting OER activity. The alloying approach presents a straightforward method for achieving atomic‐level modifications in catalyst design and can pave the way for the development of more effective and economically viable OER catalysts and beyond.

Dimethyl Carbonate Synthesis from CO2 over CeO2 with Electron-Enriched Lattice Oxygen Species.

Direct synthesis of dimethyl carbonate (DMC) from CO2 plays an important role in carbon neutrality, but its efficiency is still far from the practical application, due to the limited understanding of the reaction mechanism and rational design of efficient catalyst. Herein, abundant electron-enriched lattice oxygen species were introduced into CeO2 catalyst by constructing the point defects and crystal-terminated phases in the crystal reconstruction process. Benefitting from the acid-base properties modulated by the electron-enriched lattice oxygen, the optimized CeO2 catalyst exhibited a much higher DMC yield of 22.2 mmol g-1 than the reported metal-oxide-based catalysts at the similar conditions. Mechanistic investigations illustrated that the electron-enriched lattice oxygen can provide abundant sites for CO2 adsorption and activation, and was advantageous of the formation of the weakly adsorbed active methoxy species. These were facilitating to the coupling of methoxy and CO2 for the key *CH3OCOO intermediate formation. More importantly, the weakened adsorption of *CH3OCOO on the electron-enriched lattice oxygen can switch the rate-determining-step (RDS) of DMC synthesis from *CH3OCOO formation to *CH3OCOO dissociation, and lower the corresponding activation barriers, thus giving rise to a high performance. This work provides insights into the underlying reaction mechanism for DMC synthesis from CO2 and methanol and the design of highly efficient catalysts.

Dimethyl Carbonate Synthesis from CO2 over CeO2 with Electron‐Enriched Lattice Oxygen Species

AbstractDirect synthesis of dimethyl carbonate (DMC) from CO2 plays an important role in carbon neutrality, but its efficiency is still far from the practical application, due to the limited understanding of the reaction mechanism and rational design of efficient catalyst. Herein, abundant electron‐enriched lattice oxygen species were introduced into CeO2 catalyst by constructing the point defects and crystal‐terminated phases in the crystal reconstruction process. Benefitting from the acid‐base properties modulated by the electron‐enriched lattice oxygen, the optimized CeO2 catalyst exhibited a much higher DMC yield of 22.2 mmol g‐1 than the reported metal‐oxide‐based catalysts at the similar conditions. Mechanistic investigations illustrated that the electron‐enriched lattice oxygen can provide abundant sites for CO2 adsorption and activation, and was advantageous of the formation of the weakly adsorbed active methoxy species. These were facilitating to the coupling of methoxy and CO2 for the key *CH3OCOO intermediate formation. More importantly, the weakened adsorption of *CH3OCOO on the electron‐enriched lattice oxygen can switch the rate‐determining‐step (RDS) of DMC synthesis from *CH3OCOO formation to *CH3OCOO dissociation, and lower the corresponding activation barriers, thus giving rise to a high performance. This work provides insights into the underlying reaction mechanism for DMC synthesis from CO2 and methanol and the design of highly efficient catalysts.

Mechanistic exploration in controlling the product selectivity via metals in TiO2 for photocatalytic carbon dioxide reduction

Harnessing the power of the sun, the photocatalytic CO2 reduction process emerges as a pivotal sustainable solution, utilizing solar energy a premier, clean energy source known for its efficiency and economic viability in the current energy paradigm. This study investigates the impact of metals (Cu, Ag, Au) on enhancing the photocatalytic CO2 reduction capabilities of TiO2. In-depth mechanistic analysis conducted using in-situ DRIFTS for identifying the key reaction intermediates, highlights that the catalytic performance, both in terms of activity and product selectivity, is tightly linked to the underlying reaction mechanisms route, optoelectronic properties, and substrate’s (CO2 and H2O) affinities to the catalyst surface. Enhanced light absorption and reduced charge recombination (lifetime enhancement from 0.21 ns to ∼1.2 ns) in case of metal incorporated resulted in enhanced overall photocatalytic performance. With Au as dopant, demonstrated highest electron selectivity for H2 (>97%) compared to Ag and Cu, which showed relatively higher CO2 reduction electron selectivity (∼20% for Cu). The study reveals the crucial interplay of intermediate stabilization, demonstrating its critical role in governing the reaction pathways i.e., carbene or formaldehyde routes as observed in case of Ag and Cu respectively and thus the specific final products. For enhancing the selectivity of CO, engineering the catalyst surface to favour weak CO adsorption is vital. Observations in the CO-TPD measurements demonstrated that Ag sites were effective in promoting the weaker CO* stabilization thus displayed more selectivity towards CO as compared to Cu. While higher CH3OH formation rates in case of Cu, were found to be related to the augmented reactive M=O species stabilization which promotes the oxidation of CH4 to CH3OH. Additionally, in case of Cu, the CH3O* intermediates stabilized effectively, which contributed to more Methanol selectivity through favoured formaldehyde route. While in case of Ag, the CH2* stabilization suggests more probability of the carbene pathway in comparison to the formaldehyde route (as in case of Cu) for producing the CH4 as the final hydrogenated C1 product. These insights guide the strategic design of catalysts for controlling the reaction pathways and thus selective C1 product production in photocatalytic CO2 reduction, offering insights for advanced catalyst design.

Transition metals of Pt and Pd on the surface of topological insulator Bi2Se3.

Transition metal catalysts supported on topological insulators are predicted to show improved catalytic properties due to the presence of topological surface states, which may float up to the catalysts and provide robust electron transfer. However, experimental studies of surface structures and corresponding catalytic properties of transition metal/topological insulator heterostructures have not been demonstrated so far. Here, we report the structures, chemical states, and adsorption behaviors of two conventional transition metal catalysts, Pt and Pd, on the surface of Bi2Se3, a common topological insulator material. We reveal that Pt forms nanoparticles on the Bi2Se3 surface. Moreover, the interaction between Pt and surface Se is observed. Furthermore, thermal dosing of O2 onto the Pt/Bi2Se3 heterostructure leads to no oxygen adsorption. Detailed scanning tunneling microscopy study indicates that Pt transforms into PtSe2 after the thermal process, thus preventing O2 from adsorption. For another transition metal Pd, it exhibits approximate layer-island growth on Bi2Se3, and Pd-Se interaction is also observed. Our work provides significant insights into the behaviors of transition metals on top of a common topological insulator material and will assist in the future design of catalysts built with topological materials.

Advances and Challenges in Oxygen Carriers for Chemical Looping Partial Oxidation of Methane

To cope with global warming and increasing carbon emissions, the chemical looping process has attracted attention due to its excellent ability to convert fossil fuel and capture CO2. In this case, chemical looping partial oxidation technology has become the focus of attention due to its advantages in the production of syngas and hydrogen, especially with respect to the design and selection of oxygen carriers, which directly affect the efficiency of the production of syngas and hydrogen. In particular, the conversion of methane can reach 95% in the chemical looping partial oxidation of methane, and the selectivity of syngas, in the range of 700 °C to 900 °C at atmospheric pressure, can reach 99% for twenty or more cycles. In this review, from the perspective of metal oxide selection and structure regarding the chemical looping partial oxidation process, we discuss the role of oxygen carriers in the chemical looping partial oxidation cycle, in which the specific surface area, the lattice oxygen mobility, and the thermal stability are understood as the important factors affecting reactivity. We hope to summarize the design and development of efficient oxygen carriers with high oxygen-carrying capacity and syngas selectivity, as well as contribute to the selection, design, optimization, and redox reaction mechanism of redox catalysts.

Electrifying Energy and Chemical Transformations with Single-Atom Alloy Nanoparticle Catalysts.

Single-atom alloys (SAAs) have attracted considerable attention as promising electrocatalysts in reactions central to energy conversion and chemical transformation. In contrast to monometallic nanocrystals and metal alloys, SAAs possess unique and intriguing physicochemical properties, positioning them as ideal model systems for studying structure-property relationships. However, the field is still in its early stages. In this Perspective, we first review and summarize rational synthesis methods and advanced characterization techniques for SAA nanoparticle catalysts. We then emphasize the extensive applications of SAAs in a range of electrocatalytic reactions, including fuel cell reactions, water splitting, and carbon dioxide and nitrate reductions. Finally, we provide insights into existing challenges and prospects associated with the controlled synthesis, characterization, and design of SAA catalysts.

Boron-doped Cu-Co catalyst boosting charge transfer in photothermal carbon dioxide hydrogenation

Photothermal catalysis of CO2 hydrogenation holds great potential in achieving carbon neutrality. However, the rational design of cost-effective photothermal catalysts with high performance for CO2 hydrogenation is an ongoing challenge. Herein, we employed CuCoyBOₓ composite catalysts by incorporating both CuBOₓ and amorphous CoBOₓ to efficiently catalyze the CO2 hydrogenation. Besides the high selectivity (98.0%), the optimized CuCo5BOx demonstrated a remarkable CO yield of 124.7 mmol g⁻¹ h⁻¹ under illumination at 300°C, which was 3 times higher than the traditional thermocatalysis. Experimental results and theoretical calculations demonstrated CuBOₓ acted as a photocatalyst, donating electrons to CoBOₓ active sites. Coupling boron-oxygen species with CuCo bimetallic-oxides synergically enhanced the CO2 adsorption capacity and accelerated charge transfer, thereby enhancing the photothermal synergistic effect and consequently boosting the catalytic activity. This study establishes an approach for the rational design of non-precious metal-boride catalysts, enabling efficient photothermal conversion of CO2 into CO under mild conditions.

Ab initio calculations of the chemisorption of atomic H and O on Pt and Ir metal and on bimetallic Pt x Ir y surfaces

Abstract Understanding the chemisorption of atoms on precious metal surfaces is of substantial interest for the rational design of heterogeneous and electrochemical catalysts. In this study, we report density functional theory (DFT) investigations of the chemisorption of atomic H and O on bimetallic Pt x Ir y (111) surfaces for bifunctional anode catalyst materials in polymer electrolyte membrane (PEM) fuel cells. We found that for both adsorbates, the adsorption on the Pt(111) surface is in general less exothermic than on the Ir(111) surface. Our study has revealed that chemisorption on the bimetallic surfaces becomes more stable with increasing number of Ir surface atoms at the adsorption site. While for hydrogen atoms the ONTOP sites yield the most negative adsorption energies, the chemisorption of oxygen atoms appears to be most stable on the FCC sites for both the mono- and bimetallic surfaces. Using the ab initio thermodynamics approach, we calculated phase diagrams for the chemisorption of H and O atoms on these metal surfaces in order to transfer our findings to finite temperature and pressure conditions. Our theoretical results may provide an improved understanding of the hydrogen oxidation reaction (HOR) and oxygen evolution reaction (OER) on intermetallic Pt x Ir y (111) surfaces and may be helpful for the rational design of new bifunctional PEM fuel cell anode catalyst materials.

Tandem CatalystsEnabling Efficient C-N Coupling toward the Electrosynthesis of Urea.

The development of a methodology for synthesizing value-added urea (CO(NH2)2) via a renewable electricity-driven C-N coupling reaction under mild conditions is highly anticipated. However, the complex catalytic active sites that act on the carbon and nitrogen species make the reaction mechanism unclear, resulting in a low efficiency of C-N coupling from the co-reduction of carbon dioxide (CO2) and nitrate (NO3-). Herein,we propose a novel tandem catalyst of Mo-PCN-222(Co), in which the Mo sites serve to facilitate nitrate reduction to the *NH2 intermediate, while the Co sites enhance CO2 reduction to carbonic oxide (CO), thus synergistically promoting C-N coupling. The synthesized Mo-PCN-222(Co) catalyst exhibited a noteworthy urea yield rate of 844.11mg h-1 g-1, alongside a corresponding Faradaic efficiency of 33.90% at -0.4 V vs. reversible hydrogen electrode (RHE). By combining in situ spectroscopic techniques with density functional theory calculations, we demonstrate that efficient C-N coupling is attributed to a tandem system in which the *NH2 and *CO intermediates produced by the Mo and Co active sites of Mo-PCN-222(Co) stabilize the formation of the *CONH2 intermediate. This study provides an effective avenue for the design and synthesis of tandem catalysts for electrocatalytic urea synthesis.

Tandem Catalysts Enabling Efficient C‐N Coupling toward the Electrosynthesis of Urea

The development of a methodology for synthesizing value‐added urea (CO(NH2)2) via a renewable electricity‐driven C‐N coupling reaction under mild conditions is highly anticipated. However, the complex catalytic active sites that act on the carbon and nitrogen species make the reaction mechanism unclear, resulting in a low efficiency of C‐N coupling from the co‐reduction of carbon dioxide (CO2) and nitrate (NO3‐). Herein,we propose a novel tandem catalyst of Mo‐PCN‐222(Co), in which the Mo sites serve to facilitate nitrate reduction to the *NH2 intermediate, while the Co sites enhance CO2 reduction to carbonic oxide (CO), thus synergistically promoting C‐N coupling. The synthesized Mo‐PCN‐222(Co) catalyst exhibited a noteworthy urea yield rate of 844.11mg h−1 g−1, alongside a corresponding Faradaic efficiency of 33.90% at −0.4 V vs. reversible hydrogen electrode (RHE). By combining in situ spectroscopic techniques with density functional theory calculations, we demonstrate that efficient C‐N coupling is attributed to a tandem system in which the *NH2 and *CO intermediates produced by the Mo and Co active sites of Mo‐PCN‐222(Co) stabilize the formation of the *CONH2 intermediate. This study provides an effective avenue for the design and synthesis of tandem catalysts for electrocatalytic urea synthesis.

Photocatalytic Efficiency for CO2 Reduction of Co and Cluster Co2O2 Supported on g-C3N4: A Density Functional Theory and Machine Learning Study.

It is well known from experimental results that a single atom of cobalt supported on g-C3N4 is an efficient photocatalyst for the reduction of CO2 to CO, with a better photocatalytic activity than g-C3N4. In this work, we investigate the performance as catalysts for the CO2 reduction of single atoms of cobalt and Co2O2 clusters supported on graphitic carbon nitride (g-C3N4). Employing density functional theory plus Hubbard (DFT + U) calculations, we investigate in detail the reduction mechanisms to CO and HCOOH for the first time. We find that deposition of cobalt on g-C3N4 decreases the work function of g-C3N4 to 6.6 eV and provides a better candidate for the reduction reaction. In addition, we find that the preferred product of CO2 reduction on Co@g-C3N4 is CO, with a rate-determining barrier of 0.97 eV, while on Co2O2@g-C3N4, CO2 reduces to formate with a rate-determining barrier of 0.44 eV. We determine the creation of CO2 from COOH to only take place on Co2O2@g-C3N4, with a (relatively high) barrier of 2.27 eV. In order to obtain more easily the transition state energies of the reactions mentioned above, we applied machine learning methods to search for high-importance descriptors for these quantities, in the case of single transition metal atoms supported on C3N4. Interestingly, our results show that our quantities of interest are closely related to the adsorption energies of products and normalized valence electrons of the products of the elementary reactions as well as those of the metal atoms. The former of these two sets of features can be straightforwardly obtained via DFT, while the latter energies are extensively tabulated. Our results offer guidance for the design of catalysts and photocatalysts for CO2 reduction on single-metal atoms supported on C3N4.