Ligand engineering tailors hydrophobic microenvironments for efficient electrocatalytic oxidation of fatty alcohol

Abstract

The intrinsically low solubility of fatty alcohols in aqueous electrolytes imposes critical mass transport limitations for electrocatalytic oxidation. Conventional anodic electrocatalysts typically suffer from poor activity along with severe oxygen evolution reaction due to their hydrophilicity. Here, we present a ligand engineering strategy utilizing nickel-based metal-organic frameworks (Ni-MOFs) to construct hydrophobic microenvironments. Precise modulation of the aromatic structures of organic ligands promotes local enrichment of fatty alcohols and enhances structural robustness through π-π stacking interactions. Consequently, the optimized Ni-MOF exhibits enhanced activity for octanol oxidation compared to Ni(OH)₂, achieving a threefold higher production rate and significantly improved Faradaic efficiency (84.7% versus 30.8%). The octanoic acid production rate is competitive with that of state-of-the-art thermocatalytic aerobic oxidation. This work highlights a general design principle for hydrophobic microenvironment to overcome mass transport barriers in organic electrocatalysis, demonstrating its potential for practical applications.

Electrocatalytic oxidation of fatty alcohols in aqueous media is limited by mass transport due to their low solubility. Here, the authors propose ligand engineering to create hydrophobic microenvironments in Nickel MOFs, enabling efficient oxidation to fatty acids by enriching reactants.

Introduction

Organic electrocatalysis presents a promising solution for the mild synthesis of value-added fine chemicals in aqueous media, powered by renewable electricity^1–4. However, compared to well-established aqueous-phase processes such as water electrolysis and the chlor-alkali process, organic electrocatalysis faces distinct challenges that have hindered its industrial translation. One of the principal obstacles is the inherently low aqueous solubility of many organic substrates, which limits their concentration at the electrode–electrolyte interface and consequently results in sluggish mass transport^5–7. The electrocatalytic oxidation of fatty alcohols exemplifies this challenge. Electrocatalytic oxidation of fatty alcohols circumvents the use of heavy metal oxidants and toxic organic solvents commonly employed in traditional processes^8–10, thus providing a green and sustainable alternative for the production of high-value fatty acids¹¹. While Ni(OH)₂-based electrocatalysts have demonstrated high activity for the oxidation of water-miscible alcohols such as methanol, ethanol and glycerol^12–15, replicating such performance for long-chain fatty alcohols (C6+) remains difficult¹⁶. The intrinsic hydrophobicity of fatty alcohols stands in stark contrast to the hydrophilicity of most inorganic electrocatalysts, including Ni(OH)₂^17,18. This mismatch severely restricts reactant accessibility, limits oxidation current densities, and aggravates the competing oxygen evolution reaction (OER), particularly at high anodic potentials^19,20.

To overcome these challenges, precise engineering of the electrocatalytic microenvironment is crucial to enhance reactant accessibility and reaction efficiency^21–23. In nature, enzymes finely tune the microenvironment surrounding catalytic active sites via amino acid residues, thereby achieving both high activity and selectivity^24–26. As depicted in Fig. 1, chymotrypsin, a hydrolytic enzyme, utilizes aromatic residues (tryptophan [Trp], tyrosine [Tyr], and phenylalanine [Phe]) to construct “hydrophobic pockets” that facilitate selective binding of hydrophobic reactants at the active site^27,28. Furthermore, π–π stacking interactions among aromatic amino acids have been shown to play a pivotal role in maintaining enzymatic stability^29,30. Inspired by these biological principles, we propose to employ metal organic frameworks (MOFs) to offer a versatile platform for constructing enzyme-mimetic hydrophobic microenvironments owing to their highly tunable organic ligands^31–33. Nevertheless, MOF-based anodic electrocatalysts often undergo significant structural reconstruction in alkaline electrolytes^34,35, resulting in ligand leaching and the collapse of the well-designed coordination architectures and microenvironments^36–38. Thus, the development of MOF anode electrocatalyst that combine high electrocatalytic activity with robust structural stability remains a challenge.

Fig. 1. Schematic illustration of the proposed ligand engineering strategy for efficient electrocatalytic oxidation of fatty alcohol.

Gray, red, white, and blue spheres represent carbon, oxygen, hydrogen, and nickel atoms, respectively. Source data for Fig. 1 are provided as a Source Data file.

In this work, we present a ligand engineering strategy utilizing nickel MOFs (Ni-MOFs) to tailor hydrophobic microenvironments for efficient electrocatalytic oxidation of fatty alcohols. By systematically varying the number of benzene rings in aromatic carboxylate ligands, three distinct Ni-MOFs were rationally designed and synthesized. Comprehensive material characterizations combined with molecular dynamics simulations reveal that the Ni-MOF incorporating the terphenyl dicarboxylate ligand exhibits enhanced affinity for the local enrichment of fatty alcohols, as well as reinforced structural robustness against reconstruction (Fig. 1). These advantages are attributed to the pronounced hydrophobicity and strong interligand π–π stacking interactions imparted by its aromatic ligand³⁹. Consequently, the optimized Ni-MOF delivers enhanced electrocatalytic performance, achieving a threefold higher octanoic acid production rate from octanol oxidation at 1.6 V vs. RHE compared to Ni(OH)₂. Moreover, complete conversion of octanol can be achieved within only 3.5 h at 1.55 V vs. RHE, with a production rate competitive with traditional thermocatalytic aerobic methods. Additionally, we demonstrate a coupled system integrating octanol oxidation with hydrogen evolution in a flow cell, which significantly lowers the cell voltage required for hydrogen production and maintains stable operation for 48 h at 40 mA cm⁻², highlighting its practical applicability. In summary, this work establishes an enzyme-inspired paradigm for MOF ligand engineering, enabling precise control over electrocatalytic microenvironment, and provides a broadly applicable strategy to overcome longstanding mass transport bottlenecks in organic electrocatalysis.

Results

Structural characterization of Ni-MOFs

Three types of Ni-MOFs were synthesized via a solvothermal method, employing terephthalic acid (BDC), biphenyl-4,4′-dicarboxylate (BPDC), and p-terphenyl-4,4′′-dicarboxylate (TPDC) as aromatic dicarboxylate ligands (Fig. 2a). The resulting materials are denoted as Ni-BDC, Ni-BPDC, and Ni-TPDC, respectively. Scanning electron microscopy (SEM, Supplementary Fig. 1) and transmission electron microscopy (TEM, Supplementary Fig. 2) images reveal that all Ni-MOFs form nanosheet morphologies uniformly anchored on nickel foam. Energy-dispersive X-ray (EDX) mapping (Supplementary Figs. 3–5) indicates that all materials predominantly consist of Ni, O, and C, with the relative carbon content increasing with ligand length.

Fig. 2. Material characterization of Ni-MOFs.

a Molecular structures of BDC, BPDC and TPDC. The gray, red and white spheres represent carbon, oxygen and hydrogen atoms, respectively. HR-TEM images of b Ni-BDC, c Ni-BPDC, and d Ni-TPDC. e XRD patterns of Ni-BDC, Ni-BPDC, and Ni-TPDC. f Ni K-edge XANES and g Fourier-transformed EXAFS spectra of Ni(OH)₂, Ni-BDC, Ni-BPDC, and Ni-TPDC. Ni K-edge wavelet-transformed EXAFS contour plots for h Ni(OH)₂, i Ni-BDC, j Ni-BPDC, and k Ni-TPDC. Source data for Fig. 2 are provided as a Source Data file.

Structural differences among these Ni-MOFs were elucidated via high-resolution TEM (HR-TEM) and powder X-ray diffraction (XRD). HR-TEM images (Fig. 2b–d) exhibit well-defined lattice fringes corresponding to the (001) crystallographic planes; the interplanar spacing increases with ligand length, consistent with selected area electron diffraction results in Supplementary Fig. 6. Specifically, Ni-BDC exhibits a spacing of 1.06 nm, whereas Ni-TPDC shows the largest value of 1.86 nm, both aligning closely with ligand lengths. These results confirm that all three rigid ligands result in layered MOF structures, consistent with previous literature³⁹. XRD patterns (Fig. 2e) show the strongest diffraction peak at the (001) plane, which systematically shifts to lower angles as the ligand length increases, in accordance with Bragg’s law⁴⁰. These findings demonstrate that layered Ni-MOFs were prepared successfully through coordination between organic ligands and nickel centers.

To further probe the local coordination environments, Ni K-edge X-ray absorption spectroscopy (XAS) was performed. As shown in Fig. 2f, the Ni K-edge X-ray absorption near-edge structure (XANES) spectra reveal that the Ni centers in all Ni-MOFs possess similar oxidation states, as corroborated by the Ni 2p X-ray photoelectron spectroscopy (XPS) spectra in Supplementary Fig. 7. Fourier-transformed extended X-ray absorption fine structure (EXAFS) in Fig. 2g and wavelet-transformed EXAFS spectra in Fig. 2i–k show a prominent peak at 1.6 Å for Ni-MOFs, attributable to Ni–O bonds in the first coordination shell^41,42, with bond lengths close to those in Ni(OH)₂ (Fig. 2h). These findings indicate that aromatic ligands coordinate to Ni via carboxylate groups and that the ligand type does not significantly alter the local nickel coordination environment.

Ligand incorporation was further verified by C K-edge XANES. As presented in Supplementary Fig. 8, all Ni-MOFs exhibit two characteristic peaks at 285.6 eV and 289.0 eV, corresponding to π* (C=C) from benzene rings and σ* (C=O) from carboxyl groups, respectively^43,44. Ni-BPDC and Ni-TPDC exhibit higher intensity of π* (C=C) signals than Ni-BDC, owing to enhanced conjugation between benzene rings. Additionally, attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectra in Supplementary Fig. 9 also reveal organic ligand features. The peaks at 1376 and 1582 cm⁻¹ are assigned to symmetric and asymmetric carboxylate (–COO–) stretching, and the peaks at 823 and 754 cm⁻¹ to aromatic C–H out-of-plane bending modes^45,46. Collectively, the Ni-MOFs synthesized via the solvothermal route possess well-defined MOF structures as designed.

Effect of organic ligands in enriching fatty alcohol

The inclusion of various organic ligands enables tunable hydrophobicity within the MOF materials⁴⁷. To illustrate ligand-induced hydrophobicity differences, underwater contact angle measurements were performed to assess the wettability of octanol on Ni-MOFs in 1 M KOH electrolyte, using hydrothermally prepared Ni(OH)₂ as a blank control (see Supplementary Fig. 10 for detailed characterizations). As shown in Fig. 3a, Ni(OH)₂ exhibits a high contact angle of octanol (139.8°), indicative of poor wettability due to its intrinsic hydrophilicity. In contrast, the contact angle with octanol decreases as the ligand length increases for the Ni-MOFs, with Ni-TPDC demonstrating the lowest value (83.1°), suggestive of high wettability. This trend evidences that increasing the benzene ring in ligand enhances hydrophobicity of Ni-MOFs, as confirmed by water contact angle measurements (Supplementary Fig. 11). Furthermore, fluorescence adsorption experiments were carried out to quantify the octanol adsorption capacities of Ni-MOFs, using oil-soluble Nile Red dye as a probe⁴⁸. As shown by fluorescence emission spectra in Fig. 3b, the resulting fluorescence intensity at 628 nm increases in the sequence: Ni(OH)₂ < Ni-BDC <Ni-BPDC <Ni-TPDC, indicating enhanced wettability, which is consistent with the underwater contact angle results. Quantitatively, Ni-TPDC displays nearly twice the octanol adsorption per unit mass compared to Ni(OH)₂ (Supplementary Fig. 12).

Fig. 3. Mechanistic investigation of reactant enrichment.

a Contact angles of octanol on Ni(OH)₂ and Ni-MOFs in 1 M KOH. Insets show optical photographs of octanol droplets on the corresponding material surfaces captured during the measurements. b Fluorescence emission spectra of adsorbed Nile Red (excitation at 510 nm). MD simulation snapshots of the c Ni(OH)₂ and d Ni-TPDC systems. Gray, red, white, and blue spheres represent carbon, oxygen, hydrogen, and nickel atoms, respectively. e Mass density distribution of octanol perpendicular to the solid–liquid interface. f Van der Waals interaction energies between octanol and Ni(OH)₂ or Ni-TPDC. Source data for Fig. 3 are provided as a Source Data file.

Molecular dynamics (MD) simulations were employed to gain insight into the diffusion and adsorption behavior of octanol at the solid-liquid interface. Simulation systems comprising water and octanol were constructed, with Ni(OH)₂ and Ni-TPDC crystalline models serving as the solid phases to form the solid-liquid interface (see Supplementary Information for details). Figure 3c, d show simulation snapshots after 2000 ps for Ni-TPDC and Ni(OH)₂, respectively. These results indicate local enrichment of octanol molecules on Ni-TPDC, whereas such enrichment is absent for Ni(OH)₂. The octanol mass concentration profiles shown in Fig. 3e further corroborate these findings, displaying significantly higher local concentrations proximal to the Ni-TPDC surface, in good agreement with experimental observations. As shown in Fig. 3f, the calculated van der Waals (vdW) interaction energies between octanol and the respective materials revealed that Ni-TPDC (−56.2 kcal mol⁻¹) provides much stronger vdW interactions than Ni(OH)₂ (−2.5 kcal mol⁻¹). Thus, octanol enrichment near Ni-TPDC is primarily driven by vdW interactions stemming from enhanced hydrophobicity^49,50. These results suggest that rational ligand design can construct hydrophobic microenvironment in Ni-MOFs, facilitating reactant enrichment and improving mass transport. It should be noted that the MD simulations are based on simplified models and may not fully capture the dynamic structural evolution and surface heterogeneity of the catalyst.

Structural evolution and reconstruction of Ni-MOFs

Previous studies have established that MOFs utilized as anodic electrocatalysts, generally necessitate an activation pretreatment under electrochemical conditions, during which the MOF structure undergoes reconstruction^51–53. Hence, the insights into reconstruction behavior are essential for clarifying the structure-activity relationship in MOF-based electrocatalysts. Here, Ni-MOFs were activated via 50 cyclic voltammetry (CV) cycles at 50 mV s⁻¹ in 1 M KOH. As illustrated in Supplementary Fig. 13, the current density increases and then stabilizes as activation progresses, exhibiting Ni(II)/Ni(III) redox peaks. This indicates the formation of NiOOH as the active oxygen species via reconstruction according to previous work⁴¹. Notably, as ligand length increases, the peak current and area decrease significantly. To quantify NiOOH surface coverage, we performed pulsed chronoamperometry experiments for three Ni-MOFs (Supplementary Fig. 14). By fully oxidizing Ni(II) to Ni(III) at 1.5 V vs. RHE and then reducing back at 1.1 V vs. RHE, the integral reduction charge reflects the number of electrochemically accessible NiOOH sites. As shown in Supplementary Fig. 14c, the reduction charge follows the order: Ni-BDC-act > Ni-BPDC-Act > Ni-TPDC-Act, suggesting a reduced surface coverage of NiOOH and thus a lower degree of reconstruction. The activated Ni-MOFs after 50 cycles are denoted as Ni-BDC-Act, Ni-BPDC-Act, and Ni-TPDC-Act for subsequent characterizations.

To investigate the structural changes upon activation, both surface and bulk analytic techniques were employed. In situ Raman spectroscopy was utilized to monitor surface reconstruction at different CV cycles (Fig. 4a–c). Before activation, all three Ni-MOFs show ligand-related peaks in Raman spectra, with the main peak at 1610 cm⁻¹ assigned to aromatic C=C stretching (Supplementary Fig. 15)^54,55. However, for Ni-BDC and Ni-BPDC, these peaks vanish after only a few CV cycles. Simultaneously, two new peaks emerge at 472 and 553 cm⁻¹, which can be attributed to δ(Ni³⁺-O) and ν(Ni³⁺-O) vibrations of NiOOH, respectively. Due to the relatively weak coordination between Ni and carboxylate groups in Ni-MOFs, the ligands are exchanged with OH^- from the electrolyte at high potentials, resulting in ligand leaching. In contrast, Ni-TPDC (Fig. 4c) retains ligand-related signals throughout the activation process, even after 50 CV cycles (Supplementary Fig. 16). ATR-FTIR spectra further confirm that no ligand-related signals can be detected on the surface of Ni-BDC-Act and Ni-BPDC-Act due to complete ligand leaching, whereas the carboxylate and aromatic signals are clearly detected on the surface of Ni-TPDC-Act (Supplementary Fig. 17).

Fig. 4. Mechanistic investigation of structural evolution and reconstruction.

In situ Raman spectra of a Ni-BDC, b Ni-BPDC, and c Ni-TPDC after CV activation for different cycle numbers in 1 M KOH. d XRD patterns of Ni-BDC-Act, Ni-BPDC-Act, and Ni-TPDC-Act. e O K-edge and f C K-edge XANES spectra of Ni-TPDC and Ni-TPDC-Act. Source data for Fig. 4 are provided as a Source Data file.

XRD was used to probe the bulk structural changes of Ni-MOFs after activation. As shown in Fig. 4d, both Ni-BDC-Act and Ni-BPDC-Act show no crystalline diffraction peak, indicating amorphous structures and thus complete collapse of the frameworks during activation. Ni-TPDC-Act retains its pre-activation crystalline structure (Fig. 2e), indicating high bulk structural stability. HR-TEM image in Supplementary Fig. 18 shows that Ni-TPDC-Act preserves its pristine lattice fringes corresponding to the (001) plane of Ni-TPDC after the activation. EDX mapping results (Supplementary Fig. 19) reveal that a high content of carbon is homogeneously distributed in Ni-TPDC-Act, confirming the retention of TPDC ligands within the MOF structure. Additionally, time of flight secondary ion mass spectrometry (TOF-SIMS) was utilized to analyze the depth profiles of carbon elements in Ni-TPDC-Act (Supplementary Fig. 20). As sputtering time increases, the signal of carbon species gradually increases and then stabilizes, confirming the integrity of the MOF structure in the bulk and that reconstruction occurs only on the surface. Furthermore, the high structural stability of Ni-TPDC also ensures the stability of its hydrophobicity. As shown by the fluorescence adsorption results in Supplementary Fig. 21a, the adsorption capacity of Ni-TPDC-Act for Nile Red is only slightly reduced compared to Ni-TPDC, a marked improvement over Ni-BPDC (Supplementary Fig. 21b).

To further understand the reconstruction process of Ni-MOFs from the perspective of coordination structure, ex situ O K-edge XAS experiments were performed. As shown in Supplementary Fig. 22, the O K-edge XANES spectra of pristine Ni-MOFs display three characteristic peaks. The peak at 532.6 eV is attributed to the O 1s→ Ni 3d-O 2p hybrid orbitals⁵⁶, corresponding to the Ni–O–C coordination bond formed between Ni and carboxylates. The peak at 535.1 eV corresponds to the O 1s → π*(C = O) transitions of the carboxylate ligands, and a feature near 540.0 eV is assigned to O 1s → Ni 4sp–O 2p hybrid orbitals^57,58. After activation, the characteristic peak corresponding to the Ni 3d-O 2p hybrid orbitals in Ni-BDC-Act and Ni-BPDC-Act shifts to 534.0 eV (Supplementary Fig. 23), coinciding with the Ni–O–H coordination peak in Ni(OH)₂ (Supplementary Fig. 24)⁵⁹, indicating that the Ni-carboxylate coordination structure is completely destroyed during activation and fully converted to NiOOH and finally reduced to Ni(OH)₂. In other words, both Ni-BDC and Ni-BPDC undergo “complete reconstruction”⁶⁰. In contrast, Ni-TPDC-Act exhibits “partial reconstruction”, with the Ni 3d-O 2p hybrid signal appearing as a merger of Ni–O–C and Ni–O–H coordination bonds, as illustrated in Fig. 4e. Moreover, C K-edge XANES spectra provide more information on the coordination structure related to the organic ligands. As shown in Fig. 4f, Ni-TPDC-Act displays two characteristic peaks like Ni-TPDC, corresponding to the π*(C=C) bonds of aromatic rings (285.6 eV) and the σ*(C=O) bonds of carboxylates (289.0 eV). Notably, the relative intensity of the σ*(C=O) bond decreases after activation, indicating partial destruction of the metal-ligand coordination due to reconstruction. However, the π*(C=C) signal increases, possibly due to preferential orientation of aromatic rings during electrochemical activation, which enhances conjugation and facilitates π–π stacking interactions.

To strengthen the comparison between “partial reconstruction” and “complete reconstruction”, operando Ni K-edge XAS characterizations for both Ni-BDC and Ni-TPDC were also performed. As shown in Supplementary Fig. 25a, the Fourier-transformed EXAFS spectrum of activated Ni-BDC (Ni-BDC-Act) exhibits a prominent peak at 2.77 Å, which can be assigned to Ni-Ni scattering paths⁶¹. This feature is characteristic of NiOOH species, confirming that the original MOF framework of Ni-BDC is completely transformed into a metal (oxy)hydroxide phase upon activation (“complete reconstruction”)⁶². In contrast, Ni-TPDC shows negligible changes in coordination structure both before and after activation (Supplementary Fig. 25b). This indicates that the bulk coordination structure of Ni-TPDC is largely preserved during CV activation, consistent with our ex situ XRD results. The reconstruction is thus confined to the surface layer (“partial reconstruction”), leaving the hydrophobic framework intact.

The structural stability of Ni-TPDC is fundamentally attributed to the strong π–π stacking among multiple aromatic rings. Previous studies have demonstrated that π–π stacking between aromatic ligands in MOFs provides additional stabilization energy compared to metal-ligand coordination alone^39,40 To quantitatively support this conclusion, we calculated the interligand π–π stacking interaction energies within the crystal structures of the three Ni-MOFs. As shown in Supplementary Fig. 26, the π–π stacking energy increases substantially with the number of benzene rings in the ligand (Ni-BDC < Ni-BPDC < Ni-TPDC). The TPDC ligand, with its extended terphenyl moiety, facilitates much stronger π–π stacking compared to BDC and BPDC. The non-covalent interactions provide supplemental stabilization force that reinforces the MOF lattice against ligand leaching in the harsh alkaline electrolyte, explaining why only Ni-TPDC exhibits significant stabilization.

Overall, inspired by the structure–activity paradigms of aromatic residues in chymotrypsin, the systematic design and characterization of three Ni-MOFs with aromatic carboxylate ligands were undertaken. The TPDC ligand, containing triple benzene rings, creates an optimal hydrophobic microenvironment for enriching fatty alcohol and confers high bulk stability via π–π stacking, positioning Ni-TPDC as a promising candidate for electrocatalytic oxidation of fatty alcohol.

Electrocatalytic performance for octanol oxidation

To assess electrocatalytic performance, octanol oxidation was selected as the model reaction for subsequent studies. As shown in Supplementary Fig. 27, linear sweep voltammetry (LSV) in 1 M KOH demonstrates that both Ni-TPDC and Ni(OH)₂ exhibit a pre-OER anodic peak at 1.45–1.5 V vs. RHE, corresponding to Ni(II)/Ni(III) redox transitions, with oxygen evolution observed above 1.6 V vs. RHE. Upon the addition of 0.1 M octanol (Fig. 5a), Ni(OH)₂ still displays the anodic peak, suggesting that the formed NiOOH species accumulate due to sluggish kinetics of octanol oxidation. In contrast, Ni-TPDC shows a linear increase in oxidation current between 1.4–1.6 V vs. RHE, indicating a significantly accelerated octanol oxidation. This observation is further corroborated by electrochemical impedance spectroscopy (EIS). As presented in Supplementary Fig. 28, Nyquist plots at 1.5 V vs. RHE show the charge-transfer resistance of Ni-TPDC (2.1 Ω) is much lower than that of Ni(OH)₂ (30.1 Ω), highlighting substantially improved interfacial reaction kinetics.

Fig. 5. Electrocatalytic performance for octanol oxidation.

a LSV curves of Ni(OH)₂ and Ni-TPDC in 1 M KOH with 0.1 M octanol at scan rate of 5 mV s⁻¹. b Steady-state current density normalized to NiOOH at various potentials for Ni(OH)₂ and Ni-TPDC. Octanoic acid c production rates and d FEs at various potentials on Ni(OH)₂ and Ni-TPDC in 1 M KOH with 0.1 M octanol. e FEs of octanoic acid on Ni(OH)₂ and Ni-TPDC via constant potential and intermittent electrolysis in 1 M KOH with 0.1 M octanol. f Octanoic acid production rates and FEs at 1.55 V vs. RHE on Ni-BDC, Ni-BPDC, and Ni-TPDC in 1 M KOH with 0.1 M octanol. Error bars represent the standard deviation derived from three independent experiments. g Cycling stability tests of Ni-TPDC at 1.5 V vs. RHE. h Time-dependent octanoic acid concentrations and production rates over 3.5 h electrolysis at 1.55 V vs. RHE on Ni-TPDC. All electrolysis experiments were conducted in three-electrode cell without iR compensation at 25 °C and atmospheric pressure. The stirring rate was 600 rpm for all electrochemical tests. Source data for Fig. 5 are provided as a Source Data file.

To confirm that NiOOH acts as the catalytic site for octanol oxidation, multi-potential experiments were performed. As shown in Supplementary Fig. 29, in 1 M KOH, Ni-TPDC undergoes oxidation and reduction at 1.5 V and 1.1 V, respectively, corresponding to the reversible redox transition between Ni(OH)₂ and NiOOH. Upon introducing octanol at open-circuit potential (OCP) and then switching to 1.1 V, the reduction current significantly decreases, indicating that NiOOH spontaneously oxidizes octanol as the active oxygen species. Therefore, a pulsed chronoamperometry method was employed to measure NiOOH active sites. As shown in Supplementary Fig. 30, NiOOH is first generated at 1.5 V vs. RHE and subsequently reduced at 1.1 V vs. RHE. The charge passed during reduction quantitatively reflects the number of NiOOH active sites, showing that the number of active sites on Ni-TPDC is lower than on Ni(OH)₂. This rules out the possibility that the enhanced performance of Ni-TPDC originates from increased active site. In addition, the double-layer capacitance (C_dl), determined by CV at various scan rates (Supplementary Fig. 31), was used to estimate the electrochemically active surface area (ECSA). Linear fitting results (Supplementary Fig. 31c) reveal that Ni-TPDC exhibits a slightly higher C_dl (1.36 mF cm⁻²) compared to Ni(OH)₂ (1.11 mF cm⁻²), further ruling out the contribution of surface area to the enhanced catalytic activity of Ni-TPDC. Furthermore, multistep potential chronoamperometry was used to evaluate steady-state current density, eliminating the current contribution from self-oxidation of Ni(OH)₂ (Supplementary Fig. 32). Current densities were normalized through the number of NiOOH active sites and ECSA. As shown in Figs. 5b and S33, Ni-TPDC exhibits a much higher normalized current density than Ni(OH)₂ across a wide potential range. As shown in Supplementary Fig. 34, the Tafel plot for Ni(OH)₂ shows no clear linear region, indicative of severe mass transport limitations dominating the reaction kinetics. In contrast, Ni-TPDC exhibits a distinct linear Tafel region (1.42–1.47 V vs. RHE), allowing for the estimation of a Tafel slope (166 mV dec⁻¹). The emergence of this kinetically controlled region for Ni-TPDC directly supports the argument that reactant enrichment improves mass transport.

High-performance liquid chromatography (HPLC) was employed to quantify product during constant-potential electrolysis. Figure 5c shows the production rate of octanoic acid at 1.45, 1.5, and 1.55 V vs. RHE. Clearly, the production rate on Ni-TPDC increases with potential, reaching four times that of Ni(OH)₂ at 1.6 V vs. RHE. The Faradaic efficiency (FE) results in Fig. 5d further show that the FE for octanol oxidation on Ni(OH)₂ drops sharply with increasing potential, reaching only 30.8 ± 3.5% at 1.6 V vs. RHE. This is due to severe OER competition at high potentials, which also explains the decrease in octanoic acid yield with increasing potential on Ni(OH)₂. In contrast, Ni-TPDC maintains FEs above 80% even at high potentials, indicating effective suppression of OER.

To demonstrate that enhanced mass transport is the key to the enhanced performance of Ni-TPDC, an intermittent electrolysis protocol was implemented based on previous studies²⁰. Specifically, the protocol consists of alternating 5 s of electrolysis at 1.6 V and 1 s rest at OCP, allowing reactants to diffuse and replenish at the electrode-electrolyte interface during the rest period. As shown in Fig. 5e, compared to constant-potential electrolysis (1.6 V vs. RHE), intermittent electrolysis greatly improves the FE for octanoic acid from 30.8 ± 3.5% to 78.7 ± 6.5 % on Ni(OH)₂, while the FE on Ni-TPDC increases from 86.4 ± 1.6% to 99.9 ± 5.4%. This strongly supports that the suppression of OER on Ni-TPDC is mainly due to its hydrophobic microenvironment, which enriches reactants at the electrode–electrolyte interface. Moreover, the comparative analysis of octanol oxidation across the three Ni-MOFs reveals a progressive enhancement in both production rate and FE from Ni-BDC to Ni-BPDC and ultimately to Ni-TPDC (Fig. 5f), thereby substantiating the effectiveness of the rational ligand engineering.

To assess the durability of Ni-TPDC, repeated cycling tests were carried out at 1.5 V vs. RHE. As shown in Fig. 5g, Ni-TPDC exhibits stable performance over seven cycles, with octanoic acid FE consistently above 85%. Additionally, long-term electrolysis at 1.55 V vs. RHE was conducted, with product concentration and production rate shown in Fig. 5h. Complete octanol conversion is achieved within only 3.5 h on Ni-TPDC. Importantly, the product selectivity for octanoic acid is 100%, with no byproducts detected (Supplementary Figs. 35 and 36), highlighting the advantages of electrocatalytic oxidation.

To investigate the structural stability of Ni-TPDC during long-term electrolysis, operando Ni K-edge XAS was performed. As shown in Fig. S37, the bulk coordination structure of Ni-TPDC exhibits no significant changes at 1.55 V vs. RHE over a period of 3.5 h. In addition, both TEM (Supplementary Fig. 38) and XRD (Supplementary Fig. 39) analyses confirm the robust crystalline structure of Ni-TPDC even after prolonged electrolysis. Furthermore, the Raman spectrum (Supplementary Fig. 40) shows that the characteristic signals of the ligand in Ni-TPDC are well preserved, confirming the stability of its hydrophobic microenvironment.

To benchmark the practical potential of the electrocatalytic oxidation method, the average octanoic acid production rate achieved by Ni-TPDC at 100% conversion was normalized to catalyst mass and compared against state-of-the-art thermocatalytic aerobic oxidation methods (Supplementary Table 1). As shown in Fig. 5i, Ni-TPDC delivers a significantly higher mass-specific production rate of 6.3 mmol h⁻¹ g⁻¹, which is competitive with the performance of reported aerobic oxidation systems. It is noteworthy that current aerobic oxidation methods usually depend on precious metal catalysts (e.g., Au, Pd), which increase costs and hinder large-scale deployment. In contrast, the Ni-based electrocatalyst developed here combines low cost and high stability, offering attractive application potential.

Flow cell test

Given the high catalytic activity of Ni-TPDC for fatty alcohol oxidation, we assembled a flow cell to test the feasibility of continuous-flow octanol oxidation coupled with hydrogen evolution. Figure 6a presents a schematic diagram of the flow cell, featuring Ni-TPDC as the anode and Ni-Mo foam as the cathode, separated by an anion exchange membrane (AEM) to prevent gas crossover. Catholyte (1 M KOH) and anolyte (1 M KOH with 0.1 M octanol) were circulated separately by peristaltic pumps (see Supplementary Fig. 41). LSV curves before and after octanol addition are shown in Fig. 6b. After adding octanol, a significant increase in current density is observed, confirming that octanol oxidation has more favorable kinetics than OER. Figure 6c displays the steady-state cell voltages under constant current. After adding octanol, the cell voltage decreases by ~190 mV at 20–50 mA cm⁻², confirming that anodic octanol oxidation can reduce the energy input for hydrogen production. Figure 6d shows that the octanoic acid production rate increases with current density, reaching 0.41 mmol cm⁻² h^-1 and a FE of 88.6% at 50 mA cm⁻², demonstrating high electrocatalytic performance of Ni-TPDC under continuous flow. Meanwhile, the cathodic hydrogen product was quantified by gas-collecting method (Fig. 6e). The experimentally measured hydrogen volume over 60 min matches the theoretical value, demonstrating the FE of hydrogen evolution reaction (HER) reaches 100%. Moreover, operational stability of the flow cell system is demonstrated by maintaining a stable cell voltage over 48 h at 40 mA cm⁻² (Fig. 6f). The above results underscore the practical potential of Ni-TPDC as the anodic electrocatalyst for fatty alcohol oxidation in flow cells, achieving energy-efficient hydrogen production and high-value fatty acid synthesis simultaneously.

Fig. 6. Flow cell tests.

a Schematic illustration of the flow cell used in this work. b LSV curves of the flow cell before and after adding 0.1 M octanol in the anolyte (1 M KOH). c Cell voltage before and after adding 0.1 M octanol in the anolyte. d Octanoic acid production rates and FEs of the flow cell at various current densities. e Experimentally measured and theoretically calculated H₂ volumes produced at the cathode at 50 mA cm⁻². f Cell voltage at 40 mA cm⁻² during long-term stability test, with electrolyte refreshed every 12 h. The Schematic illustration was created using SolidWorks. Source data for Fig. 6 are provided as a Source Data file.

Discussion

In summary, inspired by the “hydrophobic pocket” of chymotrypsin constructed by aromatic amino acid residues, we have proposed a ligand engineering strategy based on Ni-MOFs to unlock efficient fatty alcohol electrocatalytic oxidation by constructing hydrophobic interfacial microenvironment. By tuning the number of benzene rings in aromatic dicarboxylate ligands, we successfully synthesized three Ni-MOFs and confirmed their well-defined metal-ligand coordination structures by techniques such as Ni K-edge XAS. Combined experimental and computational studies show that increasing benzene ring content enhances the hydrophobicity of Ni-MOFs, enabling enrichment of reactant at the surface and improved mass transport. In situ Raman and ex situ XAS studies of structural evolution reveal that Ni-TPDC exhibits the best bulk structural stability among Ni-MOFs, undergoing only “partial reconstruction” during activation. This is mainly due to strong π–π stacking among aromatic ligands, which maintains lattice stability. With the synergistic optimizations of hydrophobic microenvironment and structural stability, Ni-TPDC exhibits enhanced electrocatalytic activity and stability for octanol oxidation. Compared with conventional Ni(OH)₂, Ni-TPDC delivers a threefold increase in octanoic acid yield at 1.6 V vs. RHE. By enhancing local reactant concentration and mass transport, Ni-TPDC significantly suppresses competing OER at high potentials, raising the FE for octanoic acid from 30.8% (Ni(OH)₂) to 84.7%. Notably, Ni-TPDC achieves 100% octanol conversion within only 3.5 h at 1.55 V vs. RHE, and the production rate (0.40 mmol h⁻¹ cm⁻²) is competitive with that of current thermocatalytic aerobic oxidation methods, offering advantages in both activity and cost. Finally, integration of Ni-TPDC anode with a Ni-Mo cathode in a flow cell demonstrates the feasibility of coupling octanol oxidation with hydrogen evolution, achieving stable operation at 40 mA cm⁻² for up to 48 h. This work represents a rational ligand design strategy for optimizing electrocatalytic microenvironments, which is applicable to other material systems (e.g., two-dimensional conjugated metal organic frameworks, covalent organic frameworks) and paves the way for efficient aqueous electrocatalytic transformation of a wider range of hydrophobic organic substrates.

Methods

Chemicals

Potassium hydroxide (KOH, 90%) and nickel nitrate hexahydrate (Ni(NO₃)₂·6H₂O, 99%) were obtained from Aladdin Reagent Co., Ltd (Shanghai, China). N-octanol (99.5%) and n-octanoic acid (99.5%) were purchased from Meryer (Shanghai) Biochemical Technology Co., Ltd. Analytical-grade terephthalic acid (BDC, 99%), biphenyl-4, 4′-dicarboxylate (BPDC, 98%), and p-terphenyl-4,4′′-dicarboxylate (TPDC, 98%) were acquired from Shanghai Macklin Biochemical Technology Co., Ltd. HPLC-grade acetonitrile (99.9%) was supplied by Shanghai Titan Scientific Co., Ltd. All chemicals were used as received without further purification.

Material synthesis

Ni-BDC, Ni-BPDC, and Ni-TPDC were prepared using a one-step solvothermal method as follows⁴⁵. A mixture of 0.5 mmol Ni(NO₃)₂·6H₂O and 0.5 mmol organic ligand (BDC, BPDC, and TPDC) was first dispersed in 30 mL of DMF containing 1.5 mL ultrapure water and 1.5 mL ethanol. This solution was then transferred into a Teflon-lined stainless-steel autoclave with a 1.5 mm thick nickel foam (1 × 5 cm²) as substrate, and heated at 120 °C for 12 h. Prior to use, the nickel foam was cleaned sequentially with 3 mol L⁻¹ hydrochloric acid, water, and ethanol. After the heating process, the samples were rinsed with ethanol and ultrapure water and vacuum dried at 60 °C. The mass loading of Ni-TPDC was determined to be 68.2 mg per cm². Ni(OH)₂ was fabricated on nickel foam via a hydrothermal method based on previous studies¹⁹. Initially, 1.5 mmol Ni(NO₃)₂·6H₂O, 4.16 mmol NH₄F, and 16.6 mmol urea were dissolved in 40 mL ultrapure water. The solution was then transferred into a Teflon-lined stainless-steel autoclave, with nickel foam (1 × 5 cm²) as substrate, and heated at 100 °C for 10 h. After heating, the samples were cleaned and dried using the same method as above.

Material characterization

Scanning electron microscope (SEM) images were acquired using a JEOL JSM-7900F instrument. Transmission electron microscope (TEM) images were obtained with a JEOL JEM-2010 transmission electron microscope. X-ray photoelectron spectroscopy (XPS) measurements were performed on an AXIS Supra+ spectrometer (Kratos, England). Contact angle measurements were conducted using an OCA15EC contact angle measuring instrument. Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectra were recorded on a Bruker INVENIO S spectrometer. The spectra were collected in the range of 400–4000 cm⁻¹ with a resolution of 4 cm⁻¹, and each spectrum represents an average of 32 scans to ensure a satisfactory signal-to-noise ratio. Raman spectra were collected using a HORIBA HR Evolution high-resolution Raman spectrometer equipped with a 532 nm laser, with an acquisition time of 30 s. The spectra were acquired with a 600 grooves/mm grating, providing a spectral resolution of approximately 2 cm⁻¹. A 50× objective lens was used to focus the laser beam and collect the scattered light. Time-of-flight secondary ion mass spectrometry (ToF-SIMS) was performed on a TOF.SIMS5 instrument (ION-TOF GmbH, Germany). The depth profiling and high-resolution mapping were conducted with the Bi source in the “high current bunched” mode. For static SIMS experiments the primary ion beam was rastered across the area of 200 × 200 μm² field of view. For depth profiling the sputtering was conducted by using 2 keV Cs⁺ erosion to erode a 500 × 500 μm² crater and to analyze the center 200 × 200 μm². XAS experiments were conducted at TableXAFS-500 (Anhui Chuangpu Instrument Technology Co., LTD.). The data were collected at 25 °C in transmission mode using N₂-Ar-filled ionization chamber. All samples were pelletized into disks with a 13 mm diameter. The XAFS and EXAFS data were processed using ATHENA and ARTEMIS modules from the IFEFFIT software package. EXAFS data and wavelet transforms were obtained by Fourier transforming χ(k) data in k-space to real ® space, using Hanning windows to isolate contributions from varied coordination environments. We extend our appreciation to the staff of the BL12b-a beamline station (https://cstr.cn/31131.02.HLS.XMCD.a) at the National Synchrotron Radiation Laboratory, Hefei, for their support and assistance in soft X-ray absorption spectroscopy data acquisition and analysis.

Fluorescence adsorption experiment

Fluorescence adsorption experiments quantified the octanol adsorption using oil-soluble Nile Red dye as a probe. Specifically, 2 mg of dried powder samples of Ni(OH)₂, Ni-BDC, Ni-BPDC, and Ni-TPDC were dispersed in an n-octanol solution containing 1 mg mL⁻¹ Nile Red and subjected to ultrasonication for 20 min. The powder was then separated by centrifugation, and 2 mL of ethanol was added to fully dissolve the adsorbed Nile Red. Fluorescence emission spectra were obtained using an Edinburgh FS5 spectrofluorometer with an excitation wavelength of 510 nm. Calibration curves were created by a linear fit of the fluorescence intensity at the maximum emission wavelength (628 nm) against Nile Red concentration for accurate quantification of Nile Red adsorption (Supplementary Fig. 12a).

Electrochemical measurement

One molar KOH electrolyte was prepared by dissolving KOH powders in ultrapure water and stored in polypropylene containers. The pH of the freshly prepared electrolyte was determined to be 13.84 ± 0.01 at 25 °C. All electrochemical measurements were conducted in a single-compartment three-electrode cell (25 mL) using a CHI760E electrochemical workstation. The as-prepared catalyst electrode (1 × 1 cm²), a Hg/HgO reference electrode, and a platinum foil counter electrode (1.5 × 1.5 cm²) were employed. The Hg/HgO reference electrode was calibrated against a reversible hydrogen electrode (RHE), which consisted of a Pt wire immersed in H₂-saturated 1.0 M KOH. The stable open circuit potential between the two electrodes was measured to be 0.930 V. Accordingly, all potentials were converted to the RHE scale by adding 0.930 V to the measured potentials versus Hg/HgO. Electrochemical impedance spectroscopy (EIS) tests were conducted between 0.1 and 100000 Hz at 5 mV amplitude. Cyclic voltammetry (CV) activation pretreatments were performed at 50 mV s⁻¹ in 1 M KOH between open circuit potential and 1.8 V vs RHE. Product analysis of constant voltage electrolysis was performed after passing 100 °C. For Tafel slope analysis, 90% iR compensation was applied based on the solution resistance obtained from EIS tests. The C_dl was determined from the slope of the charging current versus scan rate plot, obtained via CV in a non-Faradaic potential window. The electrochemically active surface area (ECSA) was then estimated by dividing the measured C_dl by a specific capacitance value of 0.040 mF cm^−237,38.

Product analysis

Product concentrations were determined using high-performance liquid chromatography (HPLC, Shimadzu LC-20ADXR) with an ultraviolet (UV) detector at a wavelength of 210 nm. A BioRad Aminex 87H column was employed, with an eluent of 5 mM aqueous H₂SO₄ mixed with 20% acetonitrile at a flow rate of 0.6 mL min⁻¹. Standard calibration curves for octanoic acid are provided in Supplementary Fig. 42. Faradaic efficiency (FE) for octanoic acid (OAc) was calculated using the following equations, with n = 4 representing the number of electrons transferred for OAc formation, and 96,485 C mol⁻¹ as the Faraday constant.

$$\text{OAc}\mathrm{FE}\left(\%\right)=\frac{\mathrm{mole}\mathrm{of}\mathrm{produced}\mathrm{OAc}}{\mathrm{total}\mathrm{charge}\mathrm{passed}/\left(n\times 96485\mathrm{C}{\mathrm{mol}}^{-1}\right)}\times 100\%$$ 1

The conversion of octanol (OA) was calculated using the following equations.

$$\text{OA}\mathrm{Conversion}\left(\%\right)=\frac{\mathrm{Concentrations}\mathrm{of}\mathrm{produced}\mathrm{OAc}}{\mathrm{Initial}\mathrm{concentration}\mathrm{of}\mathrm{OA}}\times 100\%$$ 2

The generated hydrogen was quantified using a gas-collecting method. The theoretical H₂ volume was calculated as follows, with 0.0244 mL mol⁻¹ representing the molar volume of H₂ at 25 °C and 1 bar.

$$\text{Theoretical}{\mathrm{H}}_2\mathrm{volume}\left(\mathrm{mL}\right)=0.0244\mathrm{mL}{\mathrm{mol}}^{-1}\times \mathrm{total}\mathrm{charge}\mathrm{passed}/\left(2\times 96485\mathrm{C}{\mathrm{mol}}^{-1}\right)$$ 3

Gas chromatography (GC) was performed on Shimadzu GC-2014 equipped with a flame ionization detector (FID). An DB-FFAP column was used, with nitrogen as the carrier gas. ¹H nuclear magnetic resonance (NMR) spectra were recorded on a Bruker AVANCE III spectrometer (400 MHz). The liquid product after electrolysis was extracted and dissolved in deuterated chloroform for analysis.

Theoretical calculation

Molecular dynamics simulations were executed using the Forcite module within Materials Studio software, leveraging the Universal force field for all simulations^63,64. Crystal models for Ni(OH)₂ and Ni-MOFs referenced previous research literature⁴⁰. The solution box was initially set to dimensions of 3.7 × 7.4 × 10.0 nm³, filled with water and octanol molecules. Appropriately, slabs of the (010) crystal face of Ni(OH)₂ and Ni-TPDC (3.7 × 7.4 nm²) were placed at opposite ends of the solution box. Simulations were performed in the NVT ensemble, maintaining system temperature at 298 K. Long-range electrostatic interactions were calculated using the Ewald method, while van der Waals interactions were measured with the cut-off method, using a cut-off distance of 1.55 nm. The time step was set to 1 fs, with each system enduring 2000 ps of MD simulation. Frame structures were recorded at intervals of 5 ps. The initial and final configurations of MD simulations in this work are provided in Supplementary Data 1. DFT calculations of the π-stacking energy were performed with the DMol3 program of Materials Studio using the B3LYP functional and DND 3.5 basis. The atomic coordinates of the optimized computational models in this work are also provided in Supplementary Data 1.

Flow cell test

The flow cell was primarily assembled with a nickel metal housing, electrolyte inlets and outlets, along with fluororubber gaskets and M6 stainless steel screws (Supplementary Fig. 31). The anode was composed of Ni-TPDC, while the cathode consisted of a Ni-Mo foam alloy (1.5 mm thickness), with an anion exchange membrane (Selemion DSV, 95 μm thickness) separating the cathode and anode chambers. The membrane (3 × 3 cm²) was pretreated with 1 M KOH for 24 h. The catholyte was 1 M KOH solution, and the anolyte was 1 M KOH with 0.1 M octanol. The active electrode area was fixed at 1 cm². A peristaltic pump (L100-1s-1) was used to circulate the electrolytes through the cathode and anode chambers at a flow rate of 50 mL min⁻¹.

Author contributions

R.D. conducted most of the experiments and analysis and wrote the initial manuscript. Z.C. supported catalyst synthesis and testing. B.Z. contributed XAS measurement and analysis. S.W. carried out contact angle measurement and analysis. K.N. supported XPS and Raman spectra. B.Y. and J.B. contributed experimental equipment and funding. Y.C. contributed the project concept, funding and reviewed & edited the manuscript.

Peer review

Peer review information

Nature Communications thanks Arafat Khan, Peng Li and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

All data that support the findings of this study are present in the paper and the Supplementary Information files. Source data are provided in this paper. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70501-4.