Reaction Environment Engineering for Selective C₃₊ Formation in CO₂ Electroreduction: Progress and Perspectives

ABSTRACT

Electrocatalytic CO₂ reduction (ECRR) offers a sustainable way to produce high‐value chemicals. While C₁ and C₂ electrosynthesis have achieved notable success, ECRR‐to‐C₃₊ conversion remains limited by two key challenges: low overall C₃₊ selectivity and poor selectivity for specific C₃₊ products. To address these issues, this review summarizes three main formation mechanisms: *CO–*CO–*CO coupling, *CO₂–*OCHCH₂ coupling (*CO₂ insertion), and *CO–*OCHCH₂ coupling (*CO insertion). We employ thermodynamic analysis on Cu(100) to build a comprehensive reaction network. This practice identifies key branching points and the competition between hydrocarbon and oxygenate pathways. Building on these insights, we highlight a couple of new reaction environment engineering strategies, which include catalyst architecture and electrolyte interface engineering. This integrated approach holds promise for simultaneously promoting C–C couplings and tuning post‐coupling selectivity. Finally, we propose challenges and perspectives for further development of ECRR‐to‐C₃₊ conversion, aiming to guide future research in this field.

This work focuses on how to improve the selectivity and activity of electrocatalytic CO₂ reduction to C₃₊ products, by the integration of electrocatalyst and electrolyte co‐design. We summarize key C₃₊ formation mechanisms and provide a comprehensive reaction network through thermodynamic analysis. By reviewing recent publications on reaction environments, we highlight design strategies by integrating catalyst architecture and electrolyte interface engineering.

1. Introduction

Atmospheric CO₂ levels have risen from 270 ppm in the early 19th century to over 410 ppm today. This increase drives global challenges, including extreme weather, ocean acidification, and biodiversity loss [1, 2]. Natural processes like photosynthesis help regulate CO₂ but cannot maintain a stable atmospheric balance [3]. Therefore, CO₂ conversion technologies are needed to establish a sustainable artificial carbon cycle. Researchers have developed various methods, such as biocatalysis, photocatalysis, and electrocatalysis [4, 5, 6, 7, 8, 9, 10]. Each method offers a pathway to convert CO₂ into valuable products and reduce emissions.

Among these, the electrocatalytic CO₂ reduction reaction (ECRR) attracts significant attention. It converts CO₂ into chemicals and fuels using renewable energy sources like wind and solar. ECRR offers high conversion efficiency, operates under mild conditions, and provides catalytic flexibility [11, 12]. These features make it highly compatible with industrial applications. Current ECRR technologies achieve high efficiency for simple molecules. Reports show Faradaic efficiencies (FE) exceeding 95% for single‐carbon (C₁) products like CO, formic acid (HCOOH), methane (CH₄), and methanol (CH₃OH) [11, 12, 13]. Recent breakthroughs also allow the production of double‐carbon (C₂) hydrocarbons and oxygenates, such as ethanol (CH₃CH₂OH) and ethylene (CH₂CH₂), with activity and selectivity approaching industrial viability [14, 15, 16, 17, 18, 19, 20].

Building on this progress, ECRR shows promise for producing multi‐carbon (C₃₊) compounds. These products offer higher economic value due to their greater energy density and compatibility with existing fuel infrastructure.

As summarized in Table 1, C₃₊ products cover a wide range of industrial uses. They include fuel molecules such as propane and butane, key chemical feedstocks such as propylene, and oxygenated compounds such as acetone, butanol, and propanol. These products differ substantially in market size, economic value, and downstream processing requirements. Propylene stands out as a major chemical feedstock. It is currently the fourth most‐produced chemical in the world. With a market value over 120 billion USD each year, it is also the most consumed C₃₊ product. In contrast, oxygenates such as propanol, butanol, and acetone are produced at smaller volumes but attract higher prices. From a processing point of view, they also offer a practical advantage. Because they are liquids at room temperature, they can be collected directly from the electrolyte, which avoids the additional separation steps needed for gaseous products like propane. Propanol, for example, is growing in demand by 7.0% because more solvents are needed for medicine and cleaning (Table 1).

TABLE 1.

Market size and carbon emission of key C₃₊ products in 2023.

Products		Market			Emission [21]	Source
Name	Usage	Size /USD billion p.a.	Volume /million‐ton p.a.	Expected growth rate	Million‐ton CO2‐eq p.a.
Butane	Liquefied petroleum gas	112	200	4.1	237.4	a
Propane	Fuel	98	189	12.5	90.6	b
Propylene	Chemical feedstock	121	100	5.4	86.6	c
Acetone	solvent	6.5	8.1	3.8	11.9	d
Butanol	Chemical feedstock	9	5.5	6.2	12.1	e
Propanol	solvent	4.2	3.2	7.0	4.8	f

^a The data are obtained from: https://www.mordorintelligence.com/industry‐reports/propane‐market.

^b https://www.mordorintelligence.com/industry‐reports/propane‐market.

^c https://www.mordorintelligence.com/industry‐reports/propylene‐market‐633.

^d https://www.mordorintelligence.com/industry‐reports/propane‐market.

^e https:// www.statista.com/statistics/1245205/acetone‐market‐volume‐worldwide/.

^f https://www.statista.com/statistics/1245211/n‐butanol‐market‐volume‐worldwide/.

These diverse products offer great potential, but their current production presents significant sustainability bottlenecks. Although traditional fossil‐based manufacturing methods are well‐established, they remain highly carbon‐intensive and energy‐consuming, creating an imperative demand for cleaner alternatives [22, 23]. For instance, conventional propylene production relies heavily on fossil fuels. This process consumes 1.2 exajoules of energy annually and emits over 80 million tons of CO₂ (Figure 1). Production will likely reach 190 million tons by 2030, which will intensify these environmental and energy challenges [24]. Developing ECRR for efficient C₃₊ production helps address these issues and maintains a technological edge over other conversion methods [25].

FIGURE 1.

Environmental impact of the top four industrial reactions. Evaluated by production volumes, energy consumption, and greenhouse gas (GHG) emissions. Data obtained from reference [26].

However, ECRR must work effectively under harsh industrial conditions to replace traditional methods. This requirement reveals several technical gaps. For instance, it is still hard to reach high Faradaic efficiency at industrial‐level current densities for C₃₊ products (Table 2). This challenge stems from the complexity of ECRR‐to‐C₃₊ conversion. The process involves multi‐step C–C coupling and multiple proton‐coupled electron transfer (PCET) steps [27, 28, 29, 30]. This complexity causes two major selectivity issues. The first is the low overall C₃₊ selectivity. Forming long carbon chains is initiated by *C₁–*C₁ coupling. Subsequent elongation (*C₁–*C₂ coupling) extends the backbone [28, 29, 31, 32, 33]. Theoretical [33, 34, 35, 36, 37, 38, 39, 40] and experimental [19, 28, 41, 42, 43, 44, 45] studies reveal sluggish kinetics for both stages. Strong *CO adsorption is necessary to drive hydrogenation. Conversely, the resulting *C₁ species requires weaker binding to facilitate coupling. However, inherent scaling relationships between intermediate adsorption energies prevent this simultaneous optimization [33, 46, 47]. Similarly, *C₁–*C₂ coupling demands a precise balance. *C₂ intermediates must adsorb weakly enough to couple with *C₁ species. However, excessively weak adsorption triggers premature desorption. This results in the loss of *C₂ intermediates as by‐products [31, 48, 49]. Consequently, the FE for CO₂‐to‐C₃₊ conversion remains below 40% at commercially relevant current densities (>100 mA cm^{– 2}) [21, 50, 51].

TABLE 2.

Summary of state‐of‐the‐art ECRR‐to‐C₃₊ performance.

Product	Catalyst	Electrolyte	Electrode	Electrolyzer	E /VRHE	FE /%	j /mA cm−2
Butane	Cu2O‐Cu [52]	0.1 M KCl	GDE	H‐cell	−1.6	0.9	7
Propane	ImF‐Mo3P with anion‐exchange ionomer [53]	1 M KOH	GDE	Flow cell	−0.8	91	395
Propylene	CuCl‐derived nanocubes [54]	1 M KOH+0.2 M CsI	GDE	Flow cell	−0.6	1.5	273.7
	Ni3(PO4)2‐derived Ni catalyst [55]	0.1 M KHCO3	GDE	Flow cell	−1.2	2.2	13
Acetone	Ni2P‐Ho2O3 [56]	0.1 M KHCO3	GDE	H‐cell	—	0.95	25.4
Butanol	a‐CuTi@Cu [57]	0.1 M KHCO3	GDE	H‐cell	−0.8	6.85	2
2‐Propanol	Cu94Ag6 [58]	1 M CsHCO3	GDE	H‐cell	−0.7	56.7	100
1‐Propanol	Pd‐doped Cu [50]	1 M KOH	GDE	Flow cell	−0.68	47	38
	CuAg5%N20h [59]	1 M CsOH	GDE	Flow cell	−0.75	45	150
	Ag‐Ru‐Cu nanoparticles [45]	1 M KOH	GDE	MEA	—	36	300
	Ag‐doped Cu [27]	1 M KOH	GDE	Flow cell	−0.46	33	14
	Cu2O‐derived Cu [60]	1 M KOH	GDE	Flow cell	−0.44	22	26
	Cu2O‐derived nanocavities [28]	1 M KOH	GDE	Flow cell	−0.56	21	37
	Commercial Cu nanoparticles [61]	1 M KOH	GDE	Flow cell	—	20	155
	Fragmented Cu nanoparticles [48]	1 M KOH	GDE	Flow cell	−0.45	20	43
	CuS x with double‐sulfur vacancy [43]	1 M KHCO3	GDE	Flow cell	−0.85	15.4	64

The second, and perhaps harder, challenge is the unclear selectivity trends among specific C₃₊ products [62, 63]. The thermodynamic redox potentials for products like ethylene, ethanol, ethane, 1‐propanol, 2‐propanol, propylene, and propane are very similar. Their half‐reaction potentials overlap in the range of 0.06–0.14 V (vs. the standard hydrogen electrode, SHE) [64, 65]. This suggests we should see comparable selectivity among products. In practice, however, only 1‐propanol appears in appreciable yields [27, 28, 30, 41, 45, 46, 48, 50, 66, 67, 68, 69, 70, 71, 72]. It reaches a maximum selectivity of ∼38% in membrane electrode assemblies at 1 A cm⁻² [50]. Conversely, 2‐propanol is rarely reported, even though it is the most thermodynamically stable C₃ alcohol [58, 67]. Similarly, propylene (a C₃ olefin) typically appears only at trace levels. This contrasts sharply with n‐propanol (the corresponding alcohol) [54, 55, 67, 73, 74, 75, 76, 77, 78, 79]. It also contrasts with C₂ products, where ethylene (olefin) production significantly exceeds ethanol (alcohol) production [80, 81]. Overall, C₃₊ hydrocarbon selectivity is much lower than oxygenate selectivity [51, 65, 74]. These discrepancies highlight a gap in our understanding of the C₃₊ reaction mechanism and structure‐activity relationships. We must address these gaps to design advanced catalytic systems and evaluate ECRR scalability.

In this review, we summarize the reported mechanisms for ECRR‐to‐C₃₊ conversion. We propose thermodynamically feasible pathways to complement established mechanisms. This provides a more complete picture for reducing CO₂ to both C₃₊ hydrocarbons and oxygenates (Figure 2a). Using this roadmap, we highlight reaction environment engineering as a critical strategy to solve the two selectivity issues mentioned above. This approach integrates both catalyst architecture and electrolyte interface engineering. For catalyst architecture engineering, we discuss sublayer modifications [82, 83, 84, 85], heterostructure/interface design [86], and single‐atom alloys and derivatives [87, 88] (Figure 2b). These strategies help retain key intermediates and overcome intrinsic scaling relationships to boost C₃₊ selectivity. For electrolyte interface engineering, we focus on cation effects optimization [89], coordinative solvent control [90], and polymer‐induced interfacial effect [20, 91]. These measures create a local environment that works with the electrocatalyst to control the selectivity‐determining step after *C₁–*C₂ coupling (Figure 2c). Finally, we outline challenges and future opportunities for ECRR‐to‐C₃₊ conversion to guide research in this field.

FIGURE 2.

Overview of this review. (a) Road map for controlling selectivity from CO₂ to C₃₊ hydrocarbons and oxygenates. Proposed (b) catalyst architecture and (c) electrolyte interface strategies, and how they might contribute to C₃₊ production.

2. C₃₊ Formation Mechanism Through ECRR

ECRR‐to‐C₃₊ formation is a multi‐step process with diverse surface species and pathways. The upstream *C₁–*C₁ coupling mechanism is well established [80, 81, 92, 93, 94, 95]. Adsorbed CO₂ first converts to a *CO intermediate [96]. This species undergoes direct *CO–*CO coupling or forms deeply reduced *C₁ intermediates, such as *CHO, *COH, and *CH _x [97, 98]. Besides *CO–*CO coupling, six other pathways exist. These include *CHO–*CO, *CHO–*CHO, *COH–*CO, *CH _x –*CO₂, *CH _x –*CO, and *CH _x –*CH _x [28, 99]. In contrast, identifying *C₁–*C₂ coupling and post‐coupling pathways is difficult (Figure S1). This complexity arises from the large number of intermediates, involving over 10³ reaction steps [74]. Therefore, we do not cover the extensively studied *C₁–*C₁ coupling stage here. Instead, we focus on three main C₃₊ mechanisms starting from *C₁–*C₂ coupling (Figure 3). The *CO–*CO–*CO coupling mechanism involves the stepwise or concerted coupling of three *CO intermediates, followed by hydrogenation. The *CO insertion and *CO₂ insertion mechanisms share a common pathway and the same precursor, *OCHCH₂. Primary by‐products in all mechanisms include H₂, CO, acetate, ethanol, and ethylene. Finally, post‐*C₁–*C₂ coupling steps often determine selectivity toward specific C₃₊ products more than the coupling kinetics. We summarize these critical steps in Figures S2–S4.

FIGURE 3.

Overview of C₃₊ production pathways, highlighting three possible coupling mechanisms (in coloured boxes) and main by‐products (blue colour; CO and C₂H₄). The dashed arrow indicates multi‐PCET steps. Detailed mechanisms with alternative pathways can be found in the Supporting Information.

2.1. *CO–*CO–*CO Coupling Mechanism

The pathway based on successive *CO coupling, often described as *CO–*CO–*CO, is widely considered a key route toward C₃₊ products (Figure 3, red box; details in Figure S2). It builds on earlier experimental observations suggesting that *OCCO species can form as transient precursors. For example, Holmes et al. utilized neutralization–reionization mass spectrometry to detect signals corresponding to ethylenedione. The reported microsecond‐scale lifetime, inferred from the kinetic energy release of dissociating neutral species, suggests that this intermediate can persist long enough to participate in downstream reactions [100]. The further evolution into the *OCCOCO trimer depends on *CO coverage and the coupling barrier. In this context, the nanoconfinement strategy reported by Sargent and co‐workers provides critical insight. By creating open Cu nanocavities, they showed that restricting mass transport effectively retains locally formed *OCCO species within confined regions [28]. This spatial confinement increases the probability of subsequent coupling with *CO, leading to measurable C₃ formation. Experimentally, this is reflected in a clear shift in selectivity, where the C₃‐to‐C₂ ratio increases substantially compared with conventional nanoparticles, along with a reported propanol FE of 21%. This experimental observation supports the view that nanoconfinement can increase the local availability of C₂ or CO‐derived intermediates and thereby promote *CO–*CO–*CO coupling pathway and facilitate carbon chain elongation to C₃₊ products. Complementarily, density functional theory (DFT) calculations show that interfacial electric fields and solvation effects significantly enhance *CO coverage and stabilize the transition state [19, 27, 28, 31].

Recent studies by Tang et al. [49] and Pablo‐García et al. [74] reveal a high activation barrier for stepwise *CO–*OCCO coupling (ΔG_a > 1 eV) on Cu(100). This kinetic hurdle prompted the exploration of alternative pathways. Crucially, it suggests that chain growth need not rely solely on standard *CO and *OCCO precursors.

For example, it has been demonstrated that moderate *CO binding is essential for activation and coupling [34, 81], and the co‐adsorption energy of three *CO intermediates has been further identified as a key activity descriptor [31]. A constructed volcano plot predicts the optimal adsorption strength. In this optimal range, *CO undergoes a concerted trimerization mechanism. This *CO‐trimerization pathway directly produces the *OCCOCO trimer, avoiding the high energy barrier of *CO–*OCCO coupling and bypassing the competitive C₂ desorption route (Figure S2, yellow pathway) [29, 48].

Subsequent hydrogenation of the *OCCOCO trimer targets three oxygen sites. Two dehydration steps follow to yield *CCCOH. Saturation of the carbon backbone leads to 1‐propanol desorption (Figure S2, cyan pathway). This process involves 12 PCET steps. Regarding the *CH₂CH = COH intermediate, hydrogenation at the α‐carbon shifts the C = C bond to the β‐ and γ‐carbons. Further hydrogenation would release allyl alcohol. However, kinetic barriers likely suppress this step as it is experimentally undetected. Additionally, the reaction pathways branch during the dehydroxylation of the adsorbed *OCHOCCOH intermediate. Although thermodynamic analysis on pristine Cu surfaces suggests that dehydroxylation preferentially cleaves the middle OH group over the side group (ΔG = −0.45 eV vs. −0.29 eV), Voiry and co‐workers recently demonstrated a strategy to stabilize the middle OH group and steer the pathway toward 2‐propanol [58]. Under CO₂ supersaturation condition combined with a highly dispersed CuAg alloy catalyst, they observed a different reaction pathway. The presence of nearby Ag atoms changes how the intermediate interacts with the surface. Specifically, the binding at the middle carbon is weakened while the C–O bonds are strengthened. This effectively inhibits the dehydroxylation of the middle oxygen, thereby favoring the formation of 2‐propanol production, a traditionally rare product.

The absence of direct spectroscopic evidence for *OCCOCO or *HOCCOCO suggests the involvement of alternative C₂ precursors. Conversely, Fourier Transform Infrared Spectroscopy (FTIR) studies have detected *OCCOH [101]. This observation implies that the *OCCO dimer may undergo protonation to *OCCOH rather than coupling with *CO. Figure S2 (green pathway) illustrates this route: *OCCO hydrogenates to *OCCOH, dehydrates to *OCC, and hydrogenates further to *OCCH. Finally, the *OCCH species couples with *CO to form a *OCCHCO precursor. Asadi et al. proposed that following this coupling, the pathway toward propane is driven by sequential dehydroxylation and hydrogenation, a mechanism supported by their controlled experiments [53]. During electrochemical CO₂ reduction at −0.8 V (vs. the reversible hydrogen electrode, RHE), in situ Raman spectroscopy showed peaks at 275, 334, and 2,294 cm⁻¹. These peaks are assigned to the frustrated rotation of adsorbed *CO, the Mo–CO stretching mode, and the *C–O stretching mode. Crucially, the intensity of these peaks was significantly enhanced on functionalized Mo₃P compared with pristine nanoparticles, indicating a stabilized and high concentration of adsorbed *CO intermediates on the surface. The enriched *CO environment increases the probability for *OCCH to be coupled with another *CO molecule, forming a *OCCH+*CO precursor. This step is consistent with the observation that the system predominantly yields hydrocarbons when the *CO coverage remains high. Following the formation of *OCCHCO, subsequent reduction steps proceed sequentially to yield propane.

A third possible pathway involves deeply reduced C₁ precursor. Instead of direct coupling with *CO, *OCCH may couple with *CH (*CO reduction product) to form *HCCOCH (Figure S2, pink pathway). This mechanism resembles Fischer–Tropsch synthesis. Zhou et al. proposed this route on polarized Ni and validated it by adding CO or formaldehyde feedstocks [55]. Their experiments show that adding either CO or formaldehyde to a CO₂ feed increases the formation rate of multicarbon products compared with using CO₂ alone. Adding formaldehyde was a key part of the experiment. It provided a steady supply of *CH _x pieces, which are necessary for the carbon chain to grow through an insertion process. This was further supported by the product distribution, which followed the Anderson–Schulz–Flory model, highlighting a polymerization‐like behavior of adsorbed C₁ intermediates (*CO and *CH _x ) facilitated by polarized Ni sites. These findings indicate that while Cu favors CO dimerization, polarized Ni surfaces enable a distinct pathway involving the coupling of *COOH with deeply reduced *CH _x units to initiate and propagate long carbon chains.

Overall, *CO–*CO–*CO coupling is a common mechanism for C₃₊ production on a number of catalysts, but it cannot explain propylene formation on Cu. Cu catalysts do not produce propylene from pure CO feedstock. Cu only yields trace amounts of propylene when CO₂ is the reactant. This suggests the existence of other *C₁–*C₂ coupling mechanisms, such as the *CO₂–*OCHCH₂ pathway.

2.2. *CO₂–*OCHCH₂ Coupling (*CO₂ Insertion) Mechanism

The *CO₂ insertion mechanism involves the simultaneous adsorption of *CO₂ and the *OCHCH₂ intermediate. Subsequently, *OCHCH₂ cleaves its C–O bond. The carbon fragment couples with *CO₂, leaving *O adsorbed on the surface (Figure 3, green box; details in Figure S3).

Gao et al. investigated the *C₁ precursor in this mechanism [54]. They observed negligible propylene production during pure CO reduction. The process yielded a cathodic partial current density below 0.10 mA cm⁻² and a production rate under 90 pmol s⁻¹ cm⁻² at −0.50 to −0.75 V (vs. RHE). These metrics suggest that *C₁ precursor forms prior to *CO reduction. Consequently, potential intermediates include *CO₂ and its mono‐hydrogenated derivatives (e.g., *OCOH, *OCHO).

The same group also identified the corresponding *C₂ intermediate [54]. They analyzed the linear correlation between the partial current density (j) of propylene and other C₂ products. A strong correlation exists with both ethylene and ethanol. This suggests that propylene production shares a deep‐reduced *C₂ intermediate with these pathways. Operando technologies cannot yet directly probe these key selectivity‐determining intermediates (SDIs). To bridge this gap, Kortlever et al. combined DFT studies with controlled experiments [102]. They identified *OCHCH₂ as the SDI under weak *OCCO binding conditions. Hydrogenation of *OCHCH₂ forms acetaldehyde (an ethanol precursor) or leads to ethylene desorption. Therefore, *OCHCH₂ serves as the active *C₂ intermediate for coupling with *CO₂.

Further hydrogenation of the resulting trimer *OCOCHCH₂ presents four sub‐pathways. Across these routes, dehydroxylation consistently proves more energetically unfavorable than carbon hydrogenation (Figure S3). This thermodynamic landscape confirms 1‐propanol as the dominant product, while propylene formation remains rare. The first pathway involves hydrogenation on one oxygen atom. Sequential dehydroxylation initiates the hydrocarbon route. Retaining β‐carbon unsaturation requires three consecutive PCET steps. This sequence generates propylene (Figure S3, pink pathway). When the intermediates become fully hydrogenated, propane is generated as the terminal product. This result agrees with experimental findings in CO₂ electroreduction on Cu‐based catalysts, where propane is usually observed as a deeply reduced C₃ product under conditions that favor strong adsorption and retention of CO‐derived species [103]. Operando studies have shown that at relatively high overpotentials, Cu surfaces can sustain high *CO coverage, which promotes C–C coupling and enables successive PCET steps. Under these conditions, C₃ intermediates can undergo extensive hydrogenation, ultimately leading to the formation of fully saturated hydrocarbons such as propane.

The second pathway targets the α‐carbon. Further oxygen hydrogenation forms *HOCHOCHCH₂. Subsequent dehydroxylation leads to *OCHCHCH₂. However, this step faces a high energy barrier (ΔG = 0.86 eV vs. 0.16 eV). Pablo–García et al. attribute scarce propylene production to this unfavorable formation of allyl alkoxy (*OCH₂CHCH₂) [74]. Once this intermediate forms, subsequent steps become thermodynamically and kinetically more accessible. Experiments reducing allyl alcohol on OD‐Cu support this mechanism [74]. Specifically, electrochemical reduction of allyl alcohol yields a mixture of C₃ products, including propylene, propane, and 1‐propanol, with measurable formation rates of ∼1.44 µmol cm⁻² h⁻¹ for propylene and ∼1.97 µmol cm⁻² h⁻¹ for allyl alcohol‐derived products under reducing conditions. These results demonstrate that once the allyl alkoxy intermediate is formed, it serves as a key branching point: C–O bond cleavage leads to propylene, whereas further hydrogenation of the same intermediate produces more saturated products such as propane and 1‐propanol. The third pathway involves hydrogenation at the β‐carbon. Early loss of unsaturation eliminates the possibility of propylene formation. This route predominantly yields 1‐propanol. Similarly, the fourth pathway (hydrogenation at the γ‐carbon) converges with the second sub‐pathway. This also leads primarily to 1‐propanol.

In summary, the thermodynamic landscape of these sub‐pathways strongly favors C₃₊ oxygenates over hydrocarbons. While this aligns with the experimental dominance of 1‐propanol, the high energy barriers associated with propylene formation in this route suggest that the *CO₂ insertion mechanism alone cannot fully capture the complex C₃₊ reaction network. Therefore, to comprehensively map the alternative routes for C₃₊ production, we must consider a third possibility: the *CO–*OCHCH₂ coupling pathway.

2.3. *CO–*OCHCH₂ Coupling (*CO Insertion) Mechanism

The *CO insertion mechanism involves the simultaneous adsorption of *CO and the *OCHCH₂ intermediate. *CO inserts into *OCHCH₂ via one of three sub‐pathways (Figure 3, blue box; details in Figure S4).

The first sub‐pathway mirrors the *CO₂ insertion mechanism. Cleavage of the C–O bond in *OCHCH₂ allows the CHCH₂ group to couple with *CO. This step leaves adsorbed *O on the catalyst surface. The resulting *OCCHCH₂ undergoes sequential hydrogenation at the α‐carbon until saturation. Subsequent hydrogenation at the α‐carbon forms adsorbed propenal (*OCHCHCH₂). Desorbing long‐chain liquid molecules proves difficult (ΔG = 0.37 eV) [74]. Consequently, further dehydroxylation leads to the *CH₂CHCH₂ precursor. This step ultimately yields propylene. This oxygen‐cleavage route is theoretically selective for propylene. However, experimental evidence remains absent. High energy input for the *C₁–*C₂ coupling step likely suppresses this pathway. Our calculations on Cu(100) support this view (Figure S1). The free energy change (ΔG) for C–O bond cleavage and coupling reaches 0.12 eV. This value represents the highest among *CO–*OCHCH₂ coupling mechanisms (Figure S4). It significantly exceeds that of the *CO₂–*OCHCH₂ mechanism (−1.0 eV).

In the second sub‐pathway, the C–O bond in *OCHCH₂ remains intact. Coupling occurs directly at the β‐carbon with *CO. This forms *OCHCH₂CO, with a ΔG of −0.42 eV. Sequential hydrogenation and dehydroxylation theoretically allow hydrocarbon formation. However, this pathway strongly favors 1‐propanol. Each step in this route is energetically downhill.

In the third sub‐pathway, the C–O bond also remains intact. Coupling occurs directly at the α‐carbon with *CO to form *OCCHOCH₂. Qi et al. reported an alternative route (Figure S4, blue highlight) [58]. Here, *OCHCH₂ first undergoes hydrogenation at the β‐carbon. Subsequent coupling at the α‐carbon with *CO forms the *OCHCHOCH₃ intermediate. This species is a bidentate adsorbate, binding to the surface through both carbon and oxygen atoms. The interaction between the oxygen on the β‐carbon and surface Cu may stabilize the C–O bond and suppress dehydroxylation. Consequently, the reaction yields 2‐propanol. This pathway demands multifunctional active sites. These sites must possess both carbonophilic and oxygenophilic properties, which can be achieved on alloy catalysts [58].

In summary, we combine reported mechanisms with our own thermodynamic analysis to outline a comprehensive roadmap for converting CO₂ into C₃₊ hydrocarbons and oxygenates. This roadmap, constructed using DFT within the computational hydrogen electrode (CHE) framework [104], serves as a “diagnostic filter” to bridge fundamental theory and practical regulation strategies. One key role of this scheme is to pinpoint the critical energetic bottlenecks. It separates pathways that are thermodynamically unfavorable from those that are energetically accessible but rarely observed in experiments. The former usually points to limits in the catalyst itself, where changes in structure may be needed to adjust intrinsic relationships. The latter suggests kinetic limits, where both catalyst design and the local reaction environment may need to work together. The scheme also helps guide how to improve selectivity. It points out selectivity‐determining steps (SDSs). If this step is the coupling between *C₁ and *C₂ species, then tuning the catalyst structure becomes more important as it affects the active site geometry and charge distribution. If the key step appears later, such as hydrogenation or removal of hydroxyl groups from coupled intermediates, then adjusting the electrolyte environment is more effective in directing the reaction selectivity.

Notably, these mechanistic insights are obtained using Cu(100) as the reference surface. We selected this facet because its specific four‐fold hollow sites provide an optimal electronic environment to stabilize CO at high surface coverage. In surface science, electron back‐donation from metal d‐orbitals to the 2π* antibonding orbitals of adsorbed *CO plays a crucial role in activating *CO. Cu(100) serves as a baseline for this mechanism. For this reason, Cu(100) is often used as a model to study multicarbon formation [105]. In moving from idealized single‐crystal surfaces to complex practical catalysts, such as those with Cu(100)/Cu(111) interfaces, the fundamental physics remains the same. These advanced interfaces work by creating local regions with a reduced work function, which effectively lowers the energy barrier for electron injection [106]. In other words, the interface does not change the fundamental mechanism identified on Cu(100); rather, it provides a more energetically efficient pathway for the same electronic process. Therefore, the conclusions drawn from Cu(100) provide a general framework for understanding how different catalytic environments, whether they are idealized surfaces or complex industrial catalysts, optimize intermediate binding and steer product pathways through electronic modulation. These baseline insights allow us to rationally evaluate how the diverse regulation strategies discussed in the subsequent sections, such as sublayer modification and interface design, influence key mechanistic steps.

3. Catalyst Architecture Engineering

ECRR for C₃₊ formation is highly complex, involving multi‐step C–C coupling and multiple PCET steps. While debates on specific pathways continue, the primary challenge remains the inherently low overall selectivity due to the high kinetic barrier for *C₁–*C₂ coupling. Theoretical insights offer a critical solution. Our previous studies suggest that efficient coupling favors intermediates with high surface coverage and distinct charge distributions. This configuration minimizes repulsive dipole–dipole interactions between adsorbates [107]. Consequently, achieving high C₃₊ selectivity necessitates designing asymmetrical catalytic sites with diverse electronic configurations. For instance, in the concerted *CO trimerization mechanism, tuning active sites to induce different electronegativities in adsorbed *CO molecules creates charge asymmetry. This facilitates the formation of the *HOCCOCO intermediate while suppressing C₁ and C₂ pathways [31]. This section examines three catalyst architecture strategies to construct such surfaces: sublayer modification, single‐atom alloys and derivatives, and heterostructure and interface design. We highlight the potential of tailored catalyst architectures to enhance overall C₃₊ selectivity and address this critical challenge.

3.1. Sublayer Modification

Sublayer modification introduces a secondary material beneath the catalyst surface. This strategy modulates surface electronic structures, influences local charge distribution, and stabilizes key intermediates [82]. We categorize these strategies into three types based on the sublayer material: oxygen species, foreign metals, and other elements such as sulfur or nitrogen‐doped supports.

First, oxygen sublayers. These alter the catalyst surface's electronic structure, improving the binding and activation of reaction intermediates and facilitating key coupling steps in the reaction pathway [84]. Long et al. synthesized a catalyst with metallic Cu on the surface and Cu oxides (Cu₂O or CuO) underneath (Figure 4a). Under reductive conditions, the surface reconstructs into metallic Cu. However, the sublayer oxides persist due to dynamic oxidation‐reduction interactions. This creates a “CuOD‐Cu” structure with highly undercoordinated surface atoms. These features regulate the interaction between *CO and *HOCCOH precursors, promoting *C₁–*C₂ coupling. This catalyst achieves an n‐propanol FE of ∼17.9% [108].

FIGURE 4.

Sublayer modification. a) Top: Simulated reconstruction pathways from CuO and Cu₂O to CuOD‐Cu and Cu₂OD‐Cu. Bottom: FE of CO₂RR products on CuOD‐Cu (left) and Cu₂OD‐Cu (right) in H‐cell. Reproduced with permission [108]. Copyright 2023, The American Association for the Advancement of Science. b) Top: Scanning electron microscopy (SEM) image of Au/Cu catalyst showing 3–5 nm Au nanoparticles on polycrystalline Cu. Bottom: Rate of CO₂ reduction to >2e⁻ products. Reproduced with permission [109]. Copyright 2018, Springer Nature. c) Atomic model of Cu with a vacancy and subsurface sulfur. Bottom: FE of alkenes and alcohols on the core‐shell‐vacancy (CSVE) catalyst under various potentials. Reproduced with permission [68]. Copyright 2018, Springer Nature.

Second, foreign metal sublayers. These introduce synergistic catalytic behavior by spatially separating reaction domains [85]. Morales‐Guio et al. developed a gold‐on‐copper (Au/Cu) catalyst using electron‐beam physical vapor deposition (PVD). By precisely controlling the deposition rate, they formed a discontinuous layer of Au nanoparticles rather than a complete film on the polycrystalline Cu foil (Figure 4b) [109]. This architecture is critical because it leaves the underlying copper sublayer exposed and accessible. The top Au nanoparticles locally increase CO concentration and pH, effectively activating the copper substrate beneath. This proximity allows the exposed copper sublayer to efficiently reduce the locally generated CO to alcohols. This tandem effect enables the system to outperform pure gold, pure copper, or uniform alloys.

Third, other elements such as sulfur or nitrogen‐doped carbon. Stabilizing *C₂ intermediates for efficient *C₁–*C₂ coupling is challenging. Integrating sulfur into the sublayer and engineering controlled vacancies creates geometrically favorable sites to address this. Zhuang et al. synthesized a Cu₂S–Cu‐V (vacancy) nanoparticle structure (Figure 4c). This core–shell–vacancy engineering (CSVE) strategy uses subsurface sulfur and surface copper vacancies to stabilize active sites. This promotes efficient coupling, achieving a C₂₊ alcohol partial current density of 126 ± 5 mA cm⁻² with a FE of 32 ± 1% [68]. Similarly, nitrogen‐doped carbon supports can also induce sublayer‐like electronic effects [83]. Zhi et al. dispersed Cu/Cu₂O nanoparticles on nitrogen‐doped graphene (NG) [97]. Strong electronic interactions between the NG support and Cu species modify the surface electronic structure. This reduces the activation energy of adsorbates on the copper surface, provides additional active sites and improves nanoparticle stability. Consequently, the catalyst achieved a remarkable C₂–C₃ FE of 56%.

3.2. Single‐Atom Alloys and Derivatives

ECRR‐to‐C₃₊ conversion relies on successive *C₁–*C₁ and *C₁–*C₂ coupling. The catalyst must balance moderate *CO binding with high *CO coverage. Single‐atom alloys (SAAs) and their derivatives offer a robust solution. SAAs merge the strengths of alloys and single‐site catalysts [87]. This structure allows precise electronic tuning to control reactivity [88]. Such control is critical for C₃₊ production, as asymmetric precursors demand multi‐functional sites for activation. We classify these strategies into three categories: standard single‐atom alloys, multiple‐atom alloys, and vacancy‐engineered structures.

First, standard single‐atom alloys. These offer precise control over electronic properties at atomic sites, enhancing coupling efficiency by stabilizing non‐symmetric intermediates. Copper usually serves as the host metal with dopants such as Zn, Ag, Au, or Pd. Chen et al. theoretically embedded Ag single atoms into copper nanopyramids (Cu‐DANs, Figure 5a) [31]. This structure promotes the co‐adsorption of three *CO molecules with optimized binding energy. Consequently, *CO molecules form a long carbon chain via a one‐step concerted trimerization mechanism. This generates the *HOCCOCO intermediate while avoiding the competitive C₂ desorption pathway. Experimentally, Wang et al. identified that Ag doping induces strain and ligand effects [27]. These asymmetric sites consist of neighboring Cu atoms with distinct electronic structures, lowering activation barriers for coupling. Their Ag‐doped Cu catalyst achieved a record n‐propanol FE of 33% at −0.46 V vs. RHE. Jeong et al. introduced the facet‐defined dilute CuAu alloy nanorods (NRs) with precisely controlled Au atomic content from ca. 1% to 16% [72]. All these NR alloy catalysts expose preferential (100) surfaces, among which the NR2 catalyst with a nominal composition of Cu₉₈Au₂ exhibited the highest FE of 18.2 ± 0.3% for 1‐propanol at −0.41 V vs. RHE.

FIGURE 5.

Single‐atom alloy & derivatives. a) Structural model of Ag‐doped Cu‐DAN (top) and corresponding activity volcano plot between ΔG_3*CO and ΔG_{*CO–CO–COH} (bottom). Reproduced with permission [31]. Copyright 2021, Royal Society of Chemistry. b) Left: DFT‐calculated reaction barriers (E_a ) for C₁–C₁ and C₁–C₂ coupling on M‐doped Cu systems (M  =  Ag, Au, Ru, Rh, and Pd). Right: FEs of major C₂/C₃ products on Ag‐doped vs. pristine Cu catalysts under different potentials. Reproduced with permission [27]. Copyright 2019, Springer Nature. c) Left: Mechanism of n‐propanol formation on CuS _x with double sulphur vacancies (DSV), involving *CO–*OCCO coupling. Right: CO₂RR product distribution using CuS _x ‐DSV catalysts in H‐cells. Reproduced with permission [43]. Copyright 2021, Springer Nature.

Second, multiple‐atom alloys. These derivatives combine the strengths of dopant elements. This synergy increases *C₁ surface coverage and stabilizes *C₂ intermediates, promoting ECRR‐to‐C₃₊ conversion. Wang et al. synthesized an Ag‐Ru‐Cu catalyst using a two‐step galvanic replacement method [45]. They sequentially doped Cu nanoparticles with RuCl₃ and AgNO₃. Theoretical calculations revealed that isolated Ru and Ag dopants induce higher *CO coverage and lower activation barriers for both *C₁–*C₁ and *C₁–*C₂ coupling. Consequently, this Ag‐Ru‐Cu catalyst achieved an n‐propanol FE of 37% at a current density of 300 mA cm⁻² for over 100 h. Duong et al. extended this concept to metal‐doped Cu nitrides. They investigated a series of catalysts, among which Cu_Ag5%N_20h showed an impressive 1‐propanol FE of 45% at 150 mA cm⁻² [59]. They attributed the higher C₃₊ selectivity to increased roughness and porosity. Additionally, the Ag dopant lowers the *C₁–*C₂ coupling barriers.

Third, vacancy‐engineered structures. Vacancies function similarly to metal dopants. They modulate the electronic structure and coordination environment of the host metal. This fine‐tunes the adsorption of key coupling intermediates and facilitates C–C coupling steps. In this sense, vacancies act as a “pseudo‐dopant” within the SAA family. Peng et al. introduced lithium atoms into CuS materials via electrochemical intercalation [43]. This process triggers reactions with sulfur atoms, leading to sulfur loss and the formation of double sulfur vacancies (DSVs) in CuS. As illustrated in Figure 5c, these DSVs significantly enhance CO₂ electroreduction to n‐propanol. Mechanistically, they stabilize key intermediates (*CO and *OCCO) and promote efficient *CO–*OCCO coupling to form the *C₃ intermediate *OCCOCO. Furthermore, the vacancies optimize Cu–Cu distances, creating favorable geometries for *C1–*C₂ coupling. The DSV‐rich CuS _x catalyst achieved an n‐propanol FE of 15.4% at −1.05 V vs. RHE in an H‐cell. It retained 78% of its efficiency after 10 h, demonstrating high effectiveness and durability.

3.3. Heterostructure/Interface Design

Heterostructures enhance catalytic performance by creating unsymmetrical surface environments [71]. These structures introduce local electronic asymmetry and multifunctional active sites that decouple CO formation from subsequent coupling. Monometallic systems leverage different crystal facets or oxidation states to generate diverse reactive sites. In alloy systems, secondary metals such as Ni, Ag, Pd, Pt, Au, Rh, Ir, Pb, and Ga provide tailored electronic and geometric properties to optimize activity and selectivity. We classify these heterostructure and interface designs into three primary forms: grain boundaries, phase boundaries, and bimetallic interfaces.

First, grain boundaries between different crystalline facets. These boundaries generate strain and defect‐rich zones that spatially separate reaction domains [34]. This facilitates stepwise intermediate formation. Pang et al. synthesized highly fragmented copper (HF‐Cu) catalysts with abundant grain boundaries between Cu(111) and Cu(100) facets [48]. The spatial arrangement of these boundaries lowers the energy barriers for both *CO–*CO and *CO–*OCCO coupling steps compared to single facets (Figure 6a). This catalyst achieved an n‐propanol FE of 20% at −0.45 V vs. RHE with a partial current density of 8.5 mA cm⁻². Similarly, Gao et al. utilized copper nanocrystals (CuNCs) with coexisting Cu(100) and Cu(111) facets [54]. Mechanistically, Cu(111) facets increase the surface coverage of the *C₁ precursor, while Cu(100) facets promote the selective formation of the *C₂ precursor. The grain boundaries between these facets then act as highly reactive sites for the subsequent *C₁–*C₂ coupling (Figure 6b). Additionally, these boundaries suppress the competitive hydrogen evolution reaction (HER). Consequently, the CuNCs achieved a propylene FE of 1.4% at −0.60 V with a peak partial current density of 5.46 mA cm^{– 2}.

FIGURE 6.

Heterostructure and interface. a) Left: Interface between Cu(111) and Cu(200) facets enables spatial proximity of optimal *C₁ and *C₂ sites for C₃₊ formation. Right: Energy profiles for *CO–*CO and *CO–*OCCO coupling on Cu(111), Cu(100) and the interface. Reproduced with permission [48]. Copyright 2019, Springer Nature. b) Left: Transmission electron microscopy (TEM) image of CuNCs with mixed facets. Right: FE of allyl alcohol, propylene and n‐propanol on CuNCs at varying potential. Reproduced with permission [54]. Copyright 2023, Springer Nature. c) Initial C–C coupling steps on polarized Ni sites and schematic of the insertion mechanism for C₃₊ hydrocarbon formation. Reproduced with permission [55]. Copyright 2019, Springer Nature. d) Top: Synthesis scheme for Pb–Cu and Cu catalysts. Bottom: High‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) (left and middle) and energy dispersive spectroscopy (EDS) mapping images (right) showing structural and compositional features of Pb–Cu catalysts. Reproduced with permission [50]. Copyright 2023, Springer Nature.

Second, phase boundaries between distinct oxidation states. Despite identical chemical compositions, phase boundaries leverage differences in valence and d‐orbital occupancy. High‐valence metal phases, with partially unfilled d‐orbitals, stabilize key *C₂ intermediates. Zero‐valent phases, in contrast, possess high electron density to promote C–C coupling [110]. Zhou et al. reported nickel oxygenate‐derived catalysts with stable phase boundaries between Ni⁰ and Ni^δ+ [55]. Unlike pure Ni, which is largely inactive for CO₂ reduction, these phase boundaries stabilize *CO intermediates and enable sequential CH _x insertions via a Fischer–Tropsch‐like mechanism. This system achieved a C₃–C₆ hydrocarbon FE of 6.5% at −1.2 V (Figure 6c), surpassing state‐of‐the‐art copper‐based catalysts in selectivity.

Third, bimetallic interfaces between different elements. These interfaces utilize disparities in Fermi levels and lattice strain to tune intermediate adsorption and enhance coupling kinetics. For instance, Niu et al. doped Pb onto Cu surfaces [50]. This process forms numerous atomic Pb‐concentrated grain boundaries and induces under‐coordinated Cu sites via structural distortion (Figure 6d). These interfaces facilitate a spatial separation of catalytic roles: Pb stabilizes *C₁ intermediates, while Cu promotes C–C coupling. This synergy is critical for C₃₊ formation. The Pb‐Cu catalyst achieves a remarkable n‐propanol FE of 47% with exceptional stability over 100 h.

4. Electrolyte Interface Engineering

Improving catalyst architecture alone is insufficient to achieve product‐specific control in ECRR. Targeting uncommon C₃₊ backbones such as propylene, propane, or 2‐propanol requires precise modulation of the local catalyst–electrolyte microenvironment. Experimental evidence underscores this necessity. For example, in an H‐cell with 0.05 M CsHCO₃ and atmospheric CO₂ pressure, the selectivity for 2‐propanol was 39.6% at 32 mA cm⁻². However, increasing the concentration to 1 M CsHCO₃ and the pressure to 10 bar raised the FE to 56.7% and the total current density to 100 mA cm⁻² under otherwise identical conditions [58]. Similarly, the local microenvironment dictates product distribution in flow cells. A typical catholyte creates a thick hydrated layer that favors acetic acid. Conversely, combining anion exchange membranes with a CsOH electrolyte creates a locally hydrophobic microenvironment. This shift significantly enhances ethyl acetate activity, achieving a FE of 23% and a current density of 250 mA cm⁻² [111]. Therefore, the electrolyte components, including water, ions, and solvents, collectively shape the reaction interface. Fine‐tuning these components stabilizes key SDIs in the post‐*C₁–*C₂ coupling pathway. This section examines three electrolyte strategies to fine‐tune the reaction interface: cation effects optimization, coordinative solvent control, and polymer‐induced interfacial engineering. We highlight the potential of tailored electrolyte interfaces to improve product specificity and advance C₃₊ electroreduction.

4.1. Cation Effect Optimization

During ECRR, net negative charges on the catalyst surface electrostatically attract cations. Consequently, cations accumulate at the electrochemical interface to interact with the catalyst surface. Such interactions modify both adsorption energies and reaction pathways [89]. Common cations include Li⁺, Na⁺, K⁺, Cs⁺, Mg²⁺, Ca²⁺, and NH₄ ⁺. Selecting the right cation helps stabilize key SDIs. Stable SDIs are vital for directing the reaction toward C₃₊ products. We group these cation effects into three mechanisms: electric field modulation, non‐electric field (NEF) effects, and steric regulation.

First, electric field control via hydrated cation. The Gouy–Chapman–Stern (GCS) model explains this effect. It shows that electric field strength depends on the size of the hydrated cation. Larger naked cations, such as Cs⁺, have smaller hydrated radii than smaller cations like Li⁺. A compact radius allows Cs⁺ to approach the electrode surface more closely (Figure 7a). Closer proximity generates a stronger interfacial electric field and electrostatic interaction. Xu et al. recently proved that this intense field selectively stabilizes the highly polar transition state of the rate‐determining step (RDS), thereby boosting reaction kinetics [112]. Experimental results support this enhancement. For instance, Duong et al. observed that switching the electrolyte from 1 M KOH to 1 M CsOH increased the n‐propanol FE from 39% to 45% [59]. While originally attributed to *CO binding, the distinct stabilization of both the C–C coupling transition state and O‐containing group by Cs⁺ provides a more rigorous explanation for this selectivity boost.

FIGURE 7.

Cation effect optimization. a) Left: Schematic of Li⁺ and Cs⁺ interactions with *CO and corresponding ensemble average distances between cations and O atoms. Right: Influence of cation identity on *CO adsorption free energies and rate‐determining step (RDS) energetics. Reproduced with permission [112]. Copyright 2024, Springer Nature. b) Left: Product distribution at –0.7 V in 0.1 M KOH, with and without crown ether, and in 0.1 M LiOH. Right: A larger effective cation size, resulting from either chelation (C–K⁺) or a large hydration shell (Li⁺), expands the outer Helmholtz layer. Reproduced with permission [113]. Copyright 2020, Science. c) Left: Schematic illustration of *CO₂ adsorption‐assisted alloy formation. Right: Potential‐dependent CO₂RR FE on Cu‐Ag alloy under CO₂‐supersaturated conditions. Reproduced with permission [58]. Copyright 2023, Springer Nature.

Second, non‐electric field (NEF) effects. Malkani et al. showed that electric field strength alone fails to explain the reactivity trends [113]. They proposed a NEF component. This effect relies on the cation's hydration shell structure and charge distribution. These physical properties control where *CO adsorbs (terrace vs. step sites) and stabilize the reaction intermediates (Figure 7b). NEF effects also change the water structure at the interface. Singh et al. reported that the strong electrostatic field near the cathode lowers the pKa of cation hydrolysis [114]. Crucially, this pKa reduction is more pronounced for larger cations. Consequently, hydrated K⁺, Rb⁺, and Cs⁺ act as buffers at the interface. This buffering lowers the local pH, increasing the concentration of dissolved CO₂. This suppresses HER and enhances CO₂ reduction currents, even on Ag cathodes (favoring CO) and Cu cathodes (favoring C₂₊ products).

Third, steric regulation. Beyond field effects, Qi et al. revealed a critical geometric requirement for C₃₊ formation [58]. They found that C₃₊ products form only in the presence of Cs⁺. Smaller cations failed to trigger this pathway, even under high CO₂ pressure. This stems from the specific coordination geometry of Cs⁺. Its large ionic radius allows a single cation to physically bridge two *CO intermediates. This steric interaction promotes the challenging C–C coupling step. After the C₃ backbone forms, control shifts to the Ag atoms. These sites electronically weaken the surface–oxygen binding to prevent C–O bond cleavage. This sequential synergy achieved a high 2‐propanol FE of 56.7% at −0.70 V vs. RHE (Figure 7c).

4.2. Co‐ordinative Solvent Control

Co‐ordinative solvent species modulate the interfacial water structure and chemically stabilize key SDIs. These dual functions guide reaction pathways specifically toward C₃₊ products along the post‐*C₁–*C₂ coupling pathway. Common coordinative agents range from nitriles, such as acetonitrile (ACN), and carbonates, like propylene carbonate (PC), to amides, including dimethylformamide (DMF) and N‐methyl‐2‐pyrrolidone (NMP). Other effective oxygenated solvents include ethanol (EtOH), 1,2‐dimethoxyethane (DME), dimethyl sulfoxide (DMSO), and tetrahydrofuran (THF). We classify these solvent‐induced mechanisms into three representative roles: stabilizing SDIs via transient coordination, enabling synergistic cation‐solvent interactions, and modulating interfacial hydrogen bonding.

First, stabilizing SDIs via functionalized ionomers in acid. Acidic electrolytes are essential to eliminate the carbonate formation issue (CO₂ loss) inherent to alkaline systems. However, acidity typically compromises C₃₊ selectivity. While alkali cations can improve this, they suffer from salt precipitation, limiting stability to less than 15 h. To resolve this trade‐off, Fan et al. developed a ‘cation‐free’ strategy using a coordinative ionomer layer in strong acid (0.2 M H₂SO₄, pH 0.4, Figure 8a) [115]. By coating the Cu surface with functionalized polymers, they effectively replaced bulk cations. Acting as a cation mimic, this layer generates a strong surface electric field and increases the Outer Helmholtz Plane (OHP) potential. Beyond field effects, the polymer matrix buffers local acidity and restricts proton diffusion, thereby suppressing the competing HER. Most critically, the immobilized functional groups form transient bonds with intermediates. This direct coordination stabilizes the precursors for C–C coupling. Consequently, this engineered microenvironment enables efficient multi‐carbon formation.

FIGURE 8.

Coordinative solvent control. a) Top: MD‐simulated atomic configurations of Cu/CG systems after NVE simulation (ΔU  =  2.5 V). Bottom: Full‐cell voltage performance in a slim flow cell under current densities of 25−125 mA cm⁻². Reproduced with permission [115]. Copyright 2023, Springer Nature. b) Top: Schematic of interfacial structures in Cu(100)/water/Na⁺/DMSO systems. Bottom: In situ attenuated total reflection infrared (ATR‐IR) spectra at −1.1 V vs. RHE in CO₂‐saturated NaHCO₃ and NaHCO₃ + DMSO electrolytes. Reproduced with permission [90]. Copyright 2024, Wiley‐VCH. c) Top: Effect of methyl4N⁺ (or ethyl4N⁺) on interfacial water structure and CO dimer stabilization via hydrogen bonding. Bottom: Corresponding electrochemical current densities based on geometric electrode area. Reproduced with permission [116]. Copyright 2019, PNAS.

Second, synergistic cation‐solvent interactions. In acidic media, solvents and cations often achieve higher reactivity when working in tandem. Bai et al. demonstrated this synergy by pairing Na⁺ cations with DMSO on Cu electrodes (Figure 8b) [90]. Constant potential ab initio molecular dynamics (AIMD) simulations clarify the cooperative mechanism: while Na⁺ cations initiate the process by reducing water network connectivity to inhibit C₁ pathways (CH₃OH/CH₄), adding DMSO intensifies this disruption. The solvent molecules aggressively displace interfacial water, creating a water‐poor environment. This specific local environment drastically enhances the adsorption of key SDIs like *OCCO. Thus, the hybrid electrolyte directs the reaction toward multi‐carbon products more effectively than either component alone.

Third, modulating interfacial hydrogen bonding. The steric structure of the solvent molecule dictates the delicate balance of interfacial hydration. A comparative study on the DMF family reveals a clear structure‐function relationship: Methyl₄N⁺ and Ethyl₄N⁺ promote multi‐carbon products, whereas Butyl₄N⁺ and Propyl₄N⁺ inhibit them (Figure 8c) [116]. The difference stems from water displacement behavior. Bulky hydrophobic groups (Butyl/Propyl) strip away the water layer sitting directly on top of the adsorbed *CO. This removal severs the critical hydrogen bonds needed to stabilize the terminal oxygen atoms of *OCCO. Destabilizing this intermediate hinders the C–C coupling pathway. This finding establishes a key design principle: solvent engineering must displace enough water to suppress HER, yet retain sufficient hydration to stabilize oxygenated intermediates via hydrogen bonding.

4.3. Polymer‐Induced Interfacial Effect

Polymers and functionalized copolymers in the electrolyte modulate local pH, ionic strength, and surface *CO coverage [91]. These macromolecules create a specialized microenvironment. They reduce the involvement of unwanted species and stabilize reaction intermediates. Common examples include polyvinyl alcohol (PVA), polymers containing N, O, and F elements, and polymers with amino functional groups. We classify their roles into three key mechanisms: stabilization of SDIs via functional group interactions, enhancement of interfacial integration and catalyst stability, and creation of a hydrophobic microenvironment.

First, stabilization of SDIs via functional group interactions. Specific chemical groups within the polymer matrix actively participate in the reaction pathway. Cheng et al. used polyaniline (PANI) to modify the Cu/electrolyte interface (Figure 9a) [20]. This polymer concentrates and activates CO₂, enabling high *CO coverage. Acting as a proton relay, PANI shuttles protonic H⁺ through its benzenoid ─NH─ groups. This process protonates the oxygen of *CO to form *COH. Subsequently, this intermediate asymmetrically couples with nearby *CO to form *OCCOH. This pathway offers favorable energetics. By stabilizing this specific SDI, PANI enhances the selectivity for multi‐carbon products. Additionally, the polymeric backbone interacts with the catalyst surface. This interaction improves stability and catalytic efficiency, facilitating the retention of SDIs for C₃₊ formation.

FIGURE 9.

Polymer‐induced interfacial effect. a) Left: Oxygen density profiles along the z‐direction under varying negative bias. Middle: Free energy profiles for the Volmer step on PPy/Cu and PANI/Cu under −0.8 V_RHE and pH 14. Right: Energy profile for *CO+*COH→*OCCOH on PANI/Cu. Reproduced with permission [20]. Copyright 2023, Wiley‐VCH. b) Left: Copolymer‐modified Cu electrodes for tuning CO₂RR product distribution. Right: FE and current densities toward various products. Reproduced with permission [117]. Copyright 2021, ACS. c) Left: Schematic of Mo₃P nanoparticles coated with imidazolium and anion‐exchange ionomer on carbon. Right: FE for the ImF‐Mo₃P catalyst at different potentials. Reproduced with permission [53]. Copyright 2023, Springer Nature.

Second, enhancement of interfacial integration and catalyst stability. Wang et al. developed a tricomponent copolymer to enhance C₃₊ production through a robust, integrated design. The copolymer structure integrates multiple functions into a single interface (Figure 9b) [117]. The random copolymer architecture enhances porosity, facilitating efficient reactant transport to active sites. Concurrently, the polymer induces a localized electric field to promote C–C coupling. Functional groups like urea and imidazolium are integrated into the chain to stabilize intermediates (*CO, *OCCO) via hydrogen bonding and electrostatic interactions. Crucially, the incorporation of phenylpyridinium significantly improves film robustness. This structural reinforcement prevents delamination, ensuring sustained catalytic performance. These synergistic effects position the copolymer as a highly effective, stable modifier.

Third, creation of a hydrophobic microenvironment. Polymer effects extend beyond Cu, enabling selective propane production on alternative materials. Asadi et al. achieved this by coating Mo₃P nanoparticles with an imidazolium‐functionalized ionomer (Figure 9c) [53]. Both the functionalization and the coating create a relatively hydrophobic environment. This modification increases the CO₂/water ratio near the catalytic surface, effectively suppressing the competing HER. The positively charged imidazolium further modifies the interfacial electric field and adsorption energies on Mo atoms. These electronic changes direct the reaction toward a propane‐formation pathway, initiated by the co‐adsorption of *OCCH + *CO. Owing to the synergy between Mo₃P and the hydrophobic ionomer coating, the system achieved a remarkable propane FE of 91% at –0.8 V vs. RHE.

In summary, reaction environment engineering serves as a systematic approach to improving ECRR‐to‐C₃₊ efficiency, focusing either on enhancing overall C₃₊ selectivity through catalyst architecture engineering, or increasing product‐specific selectivity via electrolyte interface engineering. However, we highlight that the success of these engineering strategies arises from the interplay between both components, rather than either one acting in isolation.

For instance, Duong et al. coupled an Ag‐doped Cu₃N catalyst with a CsOH electrolyte [59]. The introduction of Ag optimizes the surface electronic structure, weakening C‐binding while strengthening the C–O bond to suppress hydrogenation on O atoms. The nearby Cs⁺ cations induce a strong local electric field that further stabilizes O atoms against early protonation. This structural and interfacial synergy significantly retains oxygen‐containing groups and promote *C₁–*C₂ coupling, yielding a record n‐propanol FE of 45%.

The second example revisits the C₃ hydrocarbon formation discussed above. Asadi et al. employed imidazolium‐functionalized Mo₃P nanoparticles encapsulated within an anion‐exchange ionomer coating [53]. The Mo₃P nanoparticles provide highly active sites with a strong affinity for carbon‐based intermediates, while the chemisorbed imidazolium ligand electronically modulates the surface Mo atoms, shifting the Mo d‐band center upward to reinforce intermediate adsorption. Concurrently, the outer ionomer coating stabilizes this functionalized interface and optimizes the local water availability to suppress HER. This structural and interfacial coordination effectively lowers the barrier for oxygen protonation relative to carbon hydrogenation. As a result, oxygen species are progressively removed as water, enabling the formation of deeply reduced C₃ hydrocarbon species and achieving propane production with a record FE of 91%.

5. Conclusion and Perspective

The push for sustainability has increased the need for low‐footprint industrial production methods. Renewable electricity‐powered ECRR technology offers a promising alternative to producing C₃₊ chemicals, which are traditionally produced through energy‐ and carbon‐intensive processes. Advances in coupling mechanism, characterization technologies, and catalyst/electrolytic system design have improved ECRR‐to‐C₁/C₂ conversion. However, ECRR‐to‐C₃₊ is still in its early stages, hindered by low conversion rates and poor selectivity.

To address these challenges, this work discusses mechanistic insights, highlights opportunities and key aspects of electrocatalyst and electrolyte engineering. We summarize three primary mechanisms for ECRR‐to‐C₃₊ conversion: *CO–*CO–*CO coupling, *CO₂–*OCHCH₂ coupling (*CO₂ insertion), and *CO–*OCHCH₂ coupling (*CO insertion), aiming to provide researchers with a comprehensive understanding of the process. Furthermore, we perform extensive thermodynamic calculations on the Cu(100) surface to construct a comprehensive C₃₊ reaction network and incorporate alternative pathways as potential mechanisms, offering a more complete roadmap to inspire researchers in exploring practical strategies for commercialization.

Based on this roadmap, we highlight reaction environment engineering. This serves as a systematic approach to improving ECRR‐to‐C₃₊ efficiency. We focus on two main goals. First, to enhance overall C₃₊ selectivity through catalyst architecture engineering. Second, to increase product‐specific selectivity via electrolyte interface engineering. This strategy holds significant potential for advancing more sustainable and efficient production processes, as it represents a holistic approach that simultaneously tunes the electrocatalyst and electrolyte. However, capturing these coupled effects requires intensive computational efforts. Accurately simulating the dynamic solid‐liquid interface presents a formidable challenge, primarily due to the lack of robust computational models capable of describing realistic, complex electrolyte environments. To address this, we offer practical suggestions for integrating the currently underexplored electrolyte engineering into computational workflows, alongside further insights into the intricate reaction pathway networks and future research directions in this field.

5.1. First, Operando Computational Modeling Techniques

Exploring the reaction mechanisms for ECRR‐to‐C₃₊ conversion remains a significant challenge due to the highly dynamic nature of the catalyst‐electrolyte interface under operando conditions. Current state‐of‐the‐art mechanistic analyses primarily rely on a combination of in situ spectroscopic characterization and static theoretical simulations. Unfortunately, these conventional methods are still unable to directly detect intrinsic active sites as they evolve during the reaction. They also fail to monitor the dynamic transformation of adsorbed intermediates along reaction pathways. These limitations create significant barriers to advancing mechanistic studies and establishing robust structure‐activity relationships for catalysts. To overcome the inherent limitations of traditional static simulations conducted under zero Kelvin (0 K) and ultra‐high‐vacuum (UHV) conditions, the development of advanced operando computational methods is indispensable. Techniques such as constrained AIMD, slow‐growth thermodynamic integration, constant electro‐potential simulations, and microkinetic reaction network models hold great promise. By considering critical variables such as applied potential, explicit solvents, and ions, these advanced simulations can faithfully replicate realistic reaction environments. Insights derived from such models can accurately identify and verify key intermediates in this highly dynamic system, providing robust theoretical corroboration for experimental observations.

5.2. Second, Machine Learning‐Based Potentials (MLP)

Challenges in extending the reaction mechanisms for ECRR‐to‐C₃₊ conversion also stem from the reaction network complexity. This network involves thousands of reaction branches. It is formed by numerous *C₁, *C₂, and *C₃ intermediates. Operando modeling techniques are highly advanced. However, their high computational cost limits feasibility. It is difficult to independently construct a comprehensive roadmap. Recently, a promising simulation approach has emerged: machine learning‐based potentials (MLP). These techniques use deep learning, and they could model and predict atomic interactions by training on extensive datasets obtained from operando computational modeling or first‐principles calculations. MLPs can predict atomic forces and potential energies in new scenarios with high accuracy. Integrating MLP with operando modeling enables new capabilities. This approach makes the identification of most likely *C₁–*C₂ coupling steps more feasible. It could also identify bifurcation points leading to different C₃₊ products in a more efficient way. Thus, it offers a more feasible pathway toward constructing a complete reaction roadmap. For example, Sun et al. used first‐principle machine learning (FPML) [118] to explore ECRR‐to‐C₃₊ mechanisms on graphdiyne (GDY)‐based catalysts. They focused on C–C–C coupling processes. Without additional DFT calculations, FPML revealed key competitions among C₁, C₂, and C₃₊ pathways, providing a novel approach to investigate complex C₃₊ mechanisms. It further identified six promising electrocatalysts, including GDY‐Cu and GDY‐Mn. Propanol emerged as the favored product. This guides the design of advanced electrocatalysts with improved selectivity.

5.3. Third, Catalyst Materials With Multi‐Function Active Sites

ECRR‐to‐C₃₊ conversion inherently requires multiple C–C coupling steps that could include *C₁–*C₁ and *C₁–*C₂ coupling. Our previous studies on small‐molecule coupling suggest that efficient coupling is favored when intermediates exhibit high surface coverage [119]. Furthermore, the two carbonaceous molecules involved are ideally non‐identical. They should possess distinct charge distributions to minimize repulsive dipole–dipole interactions. Here, achieving facile coupling requires asymmetrical catalytic sites. These sites need diverse electronic configurations. In addition, *C₁–*C₁ coupling may involve two *C₁ intermediates at different reduction stages. Meanwhile, *C₁–*C₂ coupling involves *C₁ and *C₂ intermediates with significantly different chemical properties. These scenarios are difficult to achieve efficiently using single materials with homogeneous active sites. Therefore, developing catalyst materials with multifunctional active sites is essential. Ideally, such multi‐site catalysts should provide diverse active sites, which could synergistically regulate precursor adsorption. This facilitates carbon chain elongation to C₃₊ products. Moreover, increasing active site diversity may enable novel mechanisms. For instance, doping Cu nanopyramids with Ag single atoms provides diverse *CO binding environments across planar and under‐coordinated edge sites [31]. Through the spatial and electronic synergy of these distinct sites, adsorbed *CO undergoes a unique one‐step concerted trimerization, circumventing conventional two‐step dimerization to selectively yield 1,2‐propanediol.

5.4. Fourth, Alternative Mechanisms on Copper‐Based Materials

Current research predominantly focuses on Cu catalysts where 1‐propanol serves as the principal C₃₊ product, with *CO acting as the primary intermediate. However, unlocking non‐traditional pathways on engineered Cu surfaces is critical because the conventional routes inherently favor 1‐propanol, severely limiting the diversity and economic viability of the C₃₊ product portfolio. For example, the synthesis of propylene via ECRR on Cu remains highly challenging; the state‐of‐the‐art FE remains below 2% [54], significantly lower than that of 1‐propanol. Achieving this much‐needed alternative pathway to propylene (illustrated in Figure S4) requires overcoming substantial thermodynamic and kinetic barriers. The route initiates with the cleavage of the C–O bond in *OCHCH₂, enabling the remaining CHCH₂ group to couple with *CO. To facilitate this bond‐breaking requirement, catalyst architecture engineering can be employed to tailor the Cu surface, enhancing its oxophilicity for C–O bond activation while simultaneously maintaining high carbon affinity to ensure sufficient *CO coverage for the subsequent *C₁–*C₂ coupling. Furthermore, regulating the post‐coupling stages is equally vital for propylene selectivity. Here, electrolyte interface engineering provides effective solutions by preventing hydrogenation at the β‐carbon, thereby preserving the unsaturated C = C double bond in the final product. This propylene route is just one example within our roadmap, which also encompasses thermodynamically feasible pathways toward other valuable products like propane and 2‐propanol. Ultimately, the judicious application of reaction environment engineering to realize these alternative mechanisms on Cu represents a critical direction for addressing both diverse market demands and environmental challenges.

5.5. Fifth, Promising Catalyst Materials Beyond Copper

Beyond copper, other catalyst systems are gaining attention for C₃₊ electrosynthesis. Copper remains the main benchmark yet it often requires large overpotentials and offers limited control over product selectivity. This limitation has led to interest in non‐Cu materials such as Mo and Ni compounds. The key idea in these systems is to tune the metal d‐states. Pure Mo and Ni bind CO too strongly and can suffer from CO poisoning. When combined with non‐metals such as sulfur, phosphorus, or nitrogen, the binding strength becomes more moderate and more suitable for reaction.

Non‐Cu catalysts show both strengths and weaknesses when compared with Cu. In terms of activity, non‐Cu catalysts have demonstrated remarkably low overpotentials (< 100 mV) for C₃/C₄ oxygenates, simulating enzymatic‐like efficiency that Cu cannot achieve. For instance, P‐rich nickel phosphides (NixPy) can convert CO₂ to C₃ methylglyoxal at exceptionally low overpotentials (< 50 mV), where the rate‐determining step is identified as a non‐electrochemical surface hydride transfer rather than a standard electron transfer process [120]. Selectivity also shows clear differences. Non‐Cu materials can target rare hydrocarbons like propane with FEs exceeding 90% via reaction environment engineering [53], whereas Cu predominantly favors 1‐propanol. In addition, N‐based catalysts can follow a Fischer–Tropsch like pathway. In this route, *COOH reacts with *CH _x species and then grows through repeated insertion steps, which allows the formation of a wider range of hydrocarbons from C₃ to C₆ [55]. These pathways are not easily accessible on Cu. At the same time, non‐Cu systems face a major challenge. The HER competes strongly at high current density, which limits their performance under practical conditions. To address this issue, researchers have combined catalyst design with control of the electrolyte interface. Using imidazolium functionalization together with ionomer coatings on Mo₃P can create a more hydrophobic local environment. This environment suppresses hydrogen evolution and significantly enhances the selectivity toward C₃ hydrocarbons [53]. Overall, non‐Cu catalysts provide a viable way to expand the range of C₃₊ products. Their reaction pathways often involve different intermediates, such as formate or hydrocarbon species, rather than only *CO. This difference opens new possibilities beyond the product distribution commonly observed on Cu.

5.6. Sixth, the Engineering of Electrolyzer Configurations and Reactor Topologies

Choosing an appropriate reactor platform is fundamental, as the cell geometry directly dictates the interfacial microenvironment across three distinct aspects: (1) Mass transport and coverage optimization: Transitioning from transport‐limited H‐cells to gas diffusion electrode (GDE)‐based flow cells or MEAs overcomes CO₂ solubility limits, successfully enhancing reactant availability and stabilizing high *CO coverage at the triple‐phase boundary to enable industrially relevant current densities [121]. (2) Water activity and local pH regulation: Utilizing MEA architectures combined with anion‐exchange membranes can precisely restrict localized water activity near the catalyst surface, suppressing the competing HER to favor the preservation of oxygenated intermediates for selective C₃₊ synthesis [111]. (3) Intermediate residence time prolongation: Tailoring the physical parameters of the catalyst layer, such as increasing thickness or specific surface area, prolongs the retention time of transient C₂ intermediates, maximizing their kinetic probability for subsequent *C₁–*C₂ coupling.

These advances show strong performance at the laboratory level, yet scaling up brings new challenges. One major issue is the formation of carbonates from the reaction between OH– and CO₂. This side reaction consumes carbon without producing useful products and can reduce system stability. In addition, practical systems must consider energy use, carbon efficiency, and production rate at the same time. To address these issues, tandem reactor strategies, categorized as E_cat+E_cat [122], E_cat+T_cat [123], or E_cat+Bio_cat [124], have emerged as high‐technology‐readiness‐level (TRL) solutions. For instance, a two‐step E_cat+E_cat system utilizing a high‐temperature solid oxide electrolysis cell (SOEC) to convert CO₂ to CO, followed by a specialized CO reduction electrolyzer, can circumvent carbonate formation entirely and optimize the energy efficiency of 1‐propanol production by independently tuning each stage's overpotential [122]. Furthermore, for products that are traditionally energy‐intensive to separate, such as propanol azeotropes, industrial viability can be strengthened by steering the selectivity toward gaseous C₃₊ hydrocarbons like propane or butane through synergistic E_cat+T_cat dimerization [125]. A complete evaluation also requires economic analysis. Tools such as techno economic assessment (TEA) help define the operating conditions that are needed for large scale use.

Overall, by integrating mechanistic understanding with rational electrocatalyst–electrolyte‐reactor co‐design, the selective growth of carbon chains toward C₃₊ chemicals can transition from fundamental discovery to a robust technological reality. Continued progress in this direction should support the transition from laboratory studies to integrated systems and eventually to industrial use.

Author Contributions

Ling Chen: writing – review and editing, writing – original draft, data curation, formal analysis, methodology, software, investigation, validation, visualization. Damien Voiry: conceptualization, writing – review and editing, funding acquisition. Yan Jiao: conceptualization, writing – review and editing, supervision, funding acquisition, resources, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: The authors have cited additional references within the Supporting Information [126, 127, 128, 129, 130].

Biographies

Dr. Ling Chen received his PhD degree in 2024 from Adelaide University, Australia. He is currently a postdoctoral researcher under the guidance of Prof. Yan Jiao at School of Chemical Engineering of Adelaide University. His current research focuses on the computational design and discovery of efficient, low‐cost catalysts for sustainable energy applications.

Prof. Damien Voiry earned his PhD from the University of Bordeaux in 2010. After postdoctoral research at Rutgers University (2011–2016), he joined CNRS as a staff scientist at the Institut Européen des Membranes in Montpellier. His research focuses on low‐dimensional materials for energy conversion, including CO₂ electro‐conversion and separation membranes. Awarded an ERC Starting Grant in 2018, he has also received the CNRS Bronze Medal, SCF Young Researcher Awards, and the 2024 Young Scientist SDG Award. He joined the Young Academy of Europe in 2020.

Prof. Yan Jiao is a professor working at the School of Chemical Engineering at Adelaide University. She obtained her Ph.D. from the University of Queensland, Australia in 2012. Her research focuses on the design of catalyst materials for clean energy conversion reactions by molecular modelling and development of computational electrochemistry theory.

Data Availability Statement

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