Ni-decorated carbon nanotubes (CNTs) derived from ethanol for electrooxidation of furan derivatives featuring H₂ production

Abstract

This study explores advancements in concurrent biorefinery and hydrogen production using Ni-decorated CNTs derived from bioethanol as electrocatalysts, showing the complete conversion of furan derivatives to the desired products with 95–97% selectivity and a hydrogen production rate of 75 μmol h⁻¹. These findings highlight the simultaneous upgrading of biomass-derived compounds and hydrogen production.

---

The production of green hydrogen is crucial for addressing global energy requirements and mitigating the impact of greenhouse gas emissions. Electrocatalytic water splitting, well-known as water electrolysis, is one of the promising processes for hydrogen production without any harmful environmental impacts.¹ Although this technology is widely recognized owing to its effectiveness, the efficiency of the process is often constrained by the kinetic barriers associated with the oxygen evolution reaction (OER), which requires higher energy compared to the hydrogen evolution reaction (HER).² Recent advancements in catalyst design, electrode engineering, and renewable energy integration have significantly enhanced the viability of sustainable hydrogen production through water electrolysis. The shift towards biologically relevant materials and coupled organic oxidation systems improves energy efficiency,³ simultaneously facilitates the conversion of biomass derivatives into valuable products and enhances the scalability and commercial viability of processes like electrocatalytic furfural oxidation.^4–8 Importantly, furanic compounds such as furfural (FF) and 5-hydroxymethylfurfural (5-HMF) can be transformed towards furoic acid (FA) and 2,5-furandicarboxylic acid (FDCA), respectively.⁹ In addition, the electrooxidation of 5-HMF to FDCA is an essential process using green solvents and ambient conditions. Importantly, FDCA is a crucial intermediate for creating sustainable materials like bio-based poly(ethylene 2,5-furandicarboxylate) (PEF), replacing the traditional polyethylene terephthalate (PET).¹⁰

Furthermore, the trend to develop electrocatalysts is gradually shifting from noble metal catalysts to non-noble metals such as Ni, Co, Fe, and Mn.^11–13 Several metal nanoparticles supported on conductive materials have been used as efficient electrocatalysts, for example, Ni nanoparticles deposited on nickel foam. However, it often suffers from the leaching of metal nanoparticles from solid supports. Therefore, it is still challenging to develop stable electrocatalysts with the simultaneous presence of catalytic active sites and electrically conductive parts such as carbon materials.

Indeed, Ni-decorated carbon materials such as Ni/CNTs dramatically enhance catalytic performances. Although CNTs exhibit superior electron transfer capabilities,^14,15 controlling the conductivity of commercial CNTs is challenging due to variations in their diameters, qualities, and impurities. To address these limitations, a controlled synthesis approach using well-defined microporous zeolites as templates has been illustrated.¹⁶ Typically, the metallic catalysts, especially Fe, Co, and Ni supported on porous materials like zeolites, significantly impact the quality and yield of the synthesized CNTs. These catalysts provide highly dispersed metal active sites, eventually enhancing CNTs yield and quality. To date, using renewable carbon sources like bioethanol, derived from renewable biomass such as sugar fermentation, has also been one of the environmentally friendly approaches for green CNTs synthesis. Recently, we have successfully developed a CNTs synthesis approach using bioethanol.¹⁷ Although the controlled diameters of multilayered CNTs can be observed, achieving the high efficiency of these materials as electrocatalysts for HMF electrooxidation remains challenging.

As the electrochemical cell contains anode and cathode compartments, the electrooxidation of HMF to FDCA can be combined simultaneously with the HER at the anode and cathode, respectively. For example, bifunctional electrocatalysts such as MoO₂-FeP@C¹⁸ and NiSe@NiO_x¹⁹ exhibit as promising electrocatalysts for featuring HMF electrooxidation and hydrogen production. However, the simultaneous electrooxidation of biomass-derived compounds and hydrogen production is still in an early stage of development. To the best of our knowledge, concurrent biorefinery via electrooxidation of biomass-derived compounds and hydrogen production on non-noble metal nanoparticle-decorated CNTs as electrocatalysts has not yet been reported so far. In particular, there is a lack of knowledge for developing non-noble metals supported on carbon materials as both the anode and cathode for electrooxidation of furan derivatives and hydrogen production, respectively.

In this contribution, we illustrate the preparation and application of Ni-decorated carbon nanotubes (NiCNTs) derived from 99% purity bioethanol obtained through sugar fermentation (Fig. S1, ESI†). These NiCNTs were utilized as both the anode and cathode for the concurrent electrooxidation of furan derivatives (e.g., HMF and FF) and hydrogen production. As shown in Scheme 1 and Schemes S1 and S2 (ESI†), at first, the Ni-decorated CNTs were prepared by CVD on highly dispersed Ni supported on ZSM-5 using ethylene, generated by bioethanol dehydration on H-ZSM-5 as an acid catalyst. Subsequently, the synthesized NiCNTs after zeolite removal were deposited on nickel foam (NiCNTs/NF) and used as both the anode and cathode for the simultaneous electrooxidation of furan derivatives and hydrogen production, respectively.

Scheme 1 Schematic illustration of the reaction pathways regarding electrooxidation of furan derivatives and hydrogen production at the anode and cathode, respectively.

The crystalline patterns of all the prepared zeolites are consistent with those of the MFI topology (Fig. S2, ESI†). After loading Ni with 5 wt% on conventional ZSM-5 (5 wt% NiCONZSM-5) and hierarchical ZSM-5 (5 wt%NiHieZSM-5), the combined characteristic peaks of the MFI topologies at 2θ of 8.0°, 8.9°, 14.9°, 23.3°, 24.1°, and 30.2°, corresponding to the crystal plane indices of (101), (200), (301), (501), (303) and (503), respectively, and nickel oxides at 2θ of 37.4° and 43.5°, assigned to the plane indices of (111) and (200), respectively, were observed (Fig. S2, ESI†).²⁰ Additionally, the chemical composition of Ni on 5 wt%NiCONZSM-5 and 5 wt%NiHieZSM-5 is approximately 5 wt% (Table S1, ESI†). Scanning electron microscopy (SEM) images (Fig. S3a and b, ESI†) reveal that 5 wt%NiCONZSM-5 possesses uniform hexagonal prisms of conventional ZSM-5 particles with an average particle size of 1.6 ± 0.28 μm. In contrast, 5 wt%NiHieZSM-5 exhibits a self-assembly morphology of hierarchical ZSM-5 zeolite (Fig. S3c and d, ESI†). Notably, the particle size of 5 wt%NiHieZSM-5 is significantly smaller than that of 5 wt%NiCONZSM-5, with an average size of 134 ± 18 nm. Transmission electron microscopy (TEM) images reveal the presence of nickel oxide particles on zeolite crystals with average particle sizes of 14.1 ± 5.3 nm and 5.2 ± 1.1 nm for 5 wt%NiCONZSM-5 and 5 wt%NiHieZSM-5, respectively (Fig. S4, ESI†). The textural properties of the as-synthesized catalyst were determined by N₂ sorption isotherms, as presented in Fig. S5 and Table S2 (ESI†). It should be noted that both CONZSM-5 and 5 wt% NiCONZSM-5 follow the type I isotherm, indicating the presence of microporous structures. Conversely, HieZSM-5 and 5 wt%NiHieZSM-5 exhibited the combined types I and IV, confirming hierarchical structures containing micropores and mesopores.²¹ Indeed, nickel incorporation on zeolites led to reduced surface area and porosity due to the pore blockage from nickel nanoparticles (Table S2, ESI†). After the deposition of CNTs on 5 wt%NiCONZSM-5 and 5 wt.%NiHieZSM-5, TEM revealed nickel metals supported on CONZSM-5 and HieZSM-5 with average particle sizes of 28.0 ± 11.9 nm and 15.0 ± 3.8 nm, respectively (Fig. S6, ESI†). In addition, SEM and TEM analyses confirmed CNT formation on the external surfaces of the zeolites (Fig. S6 and S7, ESI†). The inner and outer diameters of CNTs obtained using 5 wt%NiCONZSM-5 were 26.2 ± 2.51 nm and 14.2 ± 3.4 nm, respectively. In contrast, smaller CNTs with inner diameters of 9.4 ± 3.4 nm and outer diameters of 13.0 ± 5.2 nm were formed using 5 wt%NiHieZSM-5 (Table S3, ESI†). The large CNT diameters observed with the 5 wt%NiCONZSM-5 catalyst can be attributed to nickel aggregation due to metal sintering during the CNT growth process at high temperatures. In strong contrast, the good-dispersion of Ni active sites on HieZSM-5 leads to the formation of smaller carbon nanotubes (CNTs).

To further confirm the CNT yield and quality, TGA analysis and Raman spectroscopy were applied (Table S3, ESI†). In addition, the relative intensity ratio (I_D/I_G) was observed to determine the quality of the produced CNTs. The characteristic D and G bands were observed at 1336 and 1589 cm⁻¹, respectively.²² Theoretically, a low I_D/I_G ratio relates to high graphitic behavior, relating to the higher quality of CNTs. The CNT yield over the catalyst surface was determined by %weight loss in the temperature range of 350 and 650 °C obtained from TGA curves. The CNT yield of approximately 10.4 and 21.8% is observed when using 5 wt%NiCONZSM-5 and 5 wt%NiHieZSM-5 as catalysts, respectively (Table S3, ESI†). These observations confirm that when using the CNTs (NiHieZSM-5-CNTs) as a catalyst, uniform, small tube diameter and high yield CNTs can be observed. It is, therefore, reasonable to choose the NiHieZSM-5-CNTs catalyst for further investigation.

After removing the zeolite template, the synthesized NiCNTs (NiHieZSM-5-CNTs) were deposited onto nickel foam (NiCNTs/NF), and applied as the electrode for both the anode and cathode compartments for the electrochemical oxidation of HMF and HER, respectively. At first, LSV curves over the NiCNTs/NF electrode in the presence and absence of 5 mM 5-HMF using 1.0 M KOH as a supporting electrolyte are illustrated in Fig. 1a. These curves were used to assess the redox properties over the NiCNTs/NF electrode surface. Notably, a typical oxidation peak of nickel species occurs at 1.55 V vs. RHE, and water oxidation begins after 1.60 V (Fig. 1a, black curve). However, when adding 5 mM 5-HMF into the supporting electrolyte, the oxidation current at 1.32 V is dramatically increased (Fig. 1a, red curve), indicating HMF oxidation over the oxidized Ni³⁺ surfaces.^23,24

Fig. 1 Linear sweep voltammograms (LSV) of the synthesized NiCNTs/NF in 1 M KOH at a scan rate of 20 mV s⁻¹ in the presence and absence of 5 mM 5-HMF; (b) LSV curves of the HER in 1 M KOH at a scan rate of 20 mV s⁻¹ with and without 5 mM 5-HMF (see inset the enlarged view of LSV curves); (c) and (d) kinetic curves of (c) 5-HMF oxidation and (d) the HER at E_cell of 1.45 V, using NiCNTs/NF as both the anode and cathode, respectively. (e) and (f) Recyclability test of (e) 5-HMF oxidation at the anode coupled with (f) HER at the cathode at 1.45 V over NiCNTs/NF for five consecutive cycles; HMF conversion (yellow line and bar), product selectivity of FDCA (grey), FFCA (light blue), DFF (pink), HMFCA (orange), FDCA yield (black bar), faradaic efficiency for FDCA production (FE_FDCA) (dark blue bar), hydrogen production (green bar) and faradaic efficiency for hydrogen production (FE_H2) (purple bar).

To further confirm the electrooxidation characteristics of 5-HMF, the anodic current density increases as a function of enhanced 5-HMF concentration (Fig. S8, ESI†). In addition, the impact of 5-HMF on the HER activity over NiCNTs/NF at the cathodic compartment was studied. LSV curves for the HER, before and after the addition of 5 mM 5-HMF in the cathode, exhibit a slight cathodic shift of 9 mV at −10 mA cm⁻² (Fig. 1b). The HER-HMF oxidation system shows a higher current density compared to the HER-OER system. For example, achieving 10 mA cm⁻² requires only 1.32 V vs. RHE onset cell voltage for the HER-HMF compared to 1.55 V vs. RHE for the HER-OER.

Therefore, in this work, NiCNTs/NF was used as both the anode and cathode, facilitating the simultaneous generation of H₂ and FDCA, respectively. The NiCNTs catalyst achieved nearly 100% 5-HMF conversion at various 5-HMF concentrations (5 mM, 10 mM, and 20 mM), with the FDCA yield decreased with increasing 5-HMF concentrations (Fig. S9, ESI†) due to side-product formation at high 5-HMF concentrations. Additionally, for two-electrode systems with the oxidation of 5-HMF and the HER on NiCNTs/NF electrodes, potentials ranging from 1.30 to 1.50 V between the two working electrodes were applied using 5 mM 5-HMF in 1 M KOH (Fig. S10, ESI†). The optimal potential of 1.45 V can achieve 98.9 ± 1.3% HMF conversion and approximately 94.6 ± 1.0% FDCA selectivity with the highest FE_FDCA of 97.4 ± 2.5%. At this potential, the amount of produced hydrogen was approximately 161.8 μmol cm⁻², with a FE_H2 of 93.7 ± 2.2%. As the potential increased to 1.50 V, the FDCA selectivity dropped to 88.9 ± 4.4% due to the competitive oxygen evolution reaction (OER).¹¹ These observations highlight the importance of optimizing the cell potential at 1.45 V to balance 5-HMF oxidation and hydrogen production.

The reaction kinetics behaviors for HER and 5-HMF oxidation over NiCNTs/NF surfaces were studied by maintaining a constant E_cell (ESI†) potential of 1.45 V in 1 M KOH with 5 mM 5-HMF. As shown in Fig. 1c, the reaction mixtures were collected every 20 minutes and analyzed using HPLC for the solution obtained from the cathodic compartment. It was found that the FDCA selectivity dramatically increased up to 95.2 ± 1.5% at the complete HMF conversion level (100%). Notably, the reaction preferably takes place via the HMFCA intermediate. The oxidation likely occurred at the –OH group when operated at high pH (pH = 14), leading to further oxidation to –COOH. This process typically involves sequential conversions from HMF to HMFCA, which can be further converted to FFCA and FDCA.²⁵ Concurrently, the hydrogen production was quantified using gas chromatography; the content significantly increased to 142.6 ± 1.5 μmol cm⁻² over 2 hours at a constant cell potential of 1.45 V, with FE_FDCA and FE_H2 generation of approximately 97.6 ± 0.6% and 98.7 ± 0.6%, respectively. These results underscore the high activity over NiCNTs/NF for simultaneous HMF oxidation and H₂ production.

Furthermore, the catalytic stability of the designed catalyst in a combined cell setup was studied by performing five catalytic cycles at a constant cell potential of 1.45 V using the identical NiCNTs/NF as electrodes for HMF oxidation coupled with the HER. The FE_FDCA and FE_H2 are in the range of 97.4 ± 2.9% to 90.5 ± 0.76% and 98.1 ± 2.2% to 94.3 ± 1.5%, respectively (Fig. 1e and f). This demonstrates the remarkably high stability of NiCNTs/NF for both HMF oxidation and the HER. The designed NiCNTs/NF electrodes always preserved good catalytic performances with HMF conversion ranging from 98.6 ± 1.7% to 93.2 ± 1.7%, FDCA yield ranging from 94.5 ± 1.5% to 85.6 ± 2.9%, and hydrogen production ability ranging from 164.4 ± 18.3 μmol cm⁻² to 148.0 ± 10.9 μmol cm⁻² when reusing the electrodes.

In addition, XPS analysis confirmed the presence of Ni, C, and O in both fresh and spent NiCNTs/NF electrodes, with consistent elemental composition before and after the electrochemical processes (Fig. S11 and Table S4, ESI†). The chemical states of the fresh NiCNTs/NF, the spent anodic catalyst (after five cycles of 5-HMF oxidation), and the spent cathodic catalyst (after five cycles of HER) were investigated by XPS analysis. Theoretically, the Ni peaks at 852.7, 854.1, and 856.0 eV are attributed to Ni⁽⁰⁾, Ni²⁺, and Ni³⁺, respectively.²⁶ After HMF electrooxidation, the peak at 856.0 eV is more pronounced, corresponding to the formation of Ni³⁺ species.²⁷ It should be noted that the Ni³⁺/(Ni⁰+Ni²⁺+Ni³⁺) ratio of the spent NiCNTs/NF electrode increased to 0.31, whereas no Ni³⁺ was observed for the fresh one. For the fresh and the spent cathodic catalysts, no apparent differences in Ni species were observed before and after the reaction, suggesting that the Ni valence state does not change under electroreduction conditions. In addition, the leaching percentage of Ni species was calculated relative to the initial Ni concentration in the fresh catalyst, with only 0.47% for the anode compartment and 1.56% for the HER. These findings confirm that nickel remains stably adhered to the CNTs and nickel foam throughout the reactions (Table S5, ESI†). Furthermore, the electrochemical characterization of the NiCNTs/NF and NF electrodes provides key insights into the superior performance of NiCNTs/NF. The Tafel slopes (Fig. S12, ESI†) demonstrate significantly enhanced reaction kinetics for the NiCNTs/NF (89.2 mV dec⁻¹) compared to those of the bare NF (162.5 mV dec⁻¹). This improvement is further supported by the Nyquist plots (Fig. S13, ESI†), where the charge transfer resistance (R_ct) for NiCNTs/NF is drastically reduced compared to NF, indicating more efficient electron transfer. The enhanced electron conductivity can be attributed to the high conductivity of the NiCNTs composite and the excellent electron transport properties of the CNTs structure.

Additionally, the electrochemically active surface area (ECSA), which correlates to the number of active sites for the electrochemical reaction, is notably higher for the NiCNTs/NF compared to the bare NF. This is evidenced by the double-layer capacitance (C_dl) values obtained through cyclic voltammetry (CV) measurements (Fig. S14, ESI†). The Cdl for NiCNTs/NF is 0.72 mF cm⁻², compared to 0.19 mF cm⁻² for the NF, confirming the increased ECSA of the prepared catalyst.

Apart from 5-HMF electrooxidation, the electrocatalytic reaction of other furan derivatives, such as FF, over the designed NiCNTs/NF electrode was further studied. All LSV curves for FF oxidation and the HER were measured in 1 M KOH as a supporting electrolyte in the typical three-electrode system with RHE as a reference electrode. According to Fig. S15a (ESI†), NiCNTs/NF demonstrated a high catalytic current density, with an onset potential of 1.36 V and 1.60 V vs. RHE for FF oxidation and the OER, indicating that FF oxidation was more advantageous than the OER. As shown in Fig. S15b (ESI†), NiCNTs/NF exhibited a comparable current density for the HER in the presence and absence of FF, requiring −0.11 V vs. RHE to attain a current density of 10 mA cm⁻². Moreover, for the concurrent FF electrooxidation and HER, a two-electrode electrolyzer employing NiCNTs/NF electrocatalysts for both the anode and cathode was used to quantify the production of 2-FA and hydrogen. At a constant cell potential of 1.45 V, approximately ∼11C of charge was passed with nearly 100% selectivity and FE_2-FA of 92.2 ± 2.8% of 2-FA production. Notably, the produced hydrogen is close to the theoretical value, with a production rate of 47.8 μmol cm⁻² and FE_H2 of 87.2 ± 1.60% (Fig. 2a and b).

Fig. 2 (a) Electrochemical oxidation of FF at the anode coupled with (b) HER at the cathode, with a cell potential (E_cell) of 1.45 V.

In summary, we present a novel strategy using Ni-decorated CNTs derived from bioethanol as electrocatalysts for the simultaneous conversion of furan derivatives, such as 5-HMF and FF, alongside hydrogen production. The system achieves nearly 100% faradaic efficiency for FDCA/2-FA and hydrogen production at the optimized potential. Our study highlights the economic and technical advantages of NiCNTs, which provide a cost-effective material alternative to noble metal-based catalysts. The NiCNTs, synthesized via a bioethanol-based CVD process, demonstrate high faradaic efficiency, selectivity, and reduced energy consumption. This work opens up new possibilities for electrocatalytic applications in energy production and biorefinery, combining sustainability and industrial viability in the future.

This project is funded by the National Research Council of Thailand and Vidyasirimedhi Institute of Science and Technology (N42A660307), the National Science and Technology Development Agency (NSTDA), the Thailand Science Research and Innovation (FRB670026/0457), and the Program Management Unit for Human Resources & Institutional Development, Research and Innovation (B42G670029).

Data availability

The data supporting the findings of this study, including experimental results, are available within the ESI.† Additional data can be requested from the corresponding author.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303.
2. M. T. M. Koper, Nat. Chem., 2013, 5, 255–256.
3. Q. Wang, et al. , Chem. Eng. Sci., 2022, 446, 136878.
4. H. G. Cha and K.-S. Choi, Nat. Chem., 2015, 7, 328–333.
5. Y. Yang and T. Mu, Green Chem., 2021, 23, 4228–4254.
6. T. Harhues, et al. , ACS Sustainable Chem. Eng., 2023, 11, 8413–8419.
7. S. Liu, et al. , Nat. Commun., 2024, 15, 1141.
8. S. R. Kubota and K.-S. Choi, ChemSusChem, 2018, 11, 2138–2145.
9. L. Gong, et al. , J. Catal., 2019, 373, 322–335.
10. H. Xia, et al. , RSC Adv., 2018, 8, 30875–30886.
11. W.-J. Liu, et al. , ACS Catal., 2018, 8, 5533–5541.
12. Y. Xie, et al. , Adv. Funct. Mater., 2021, 31, 2102886.
13. J. Jin, et al. , Sci. China Mater., 2022, 65, 2704–2710.
14. I. Dumitrescu, et al. , Chem. Commun., 2009, 6886–6901.
15. X. Geng, et al. , Carbon, 2023, 214, 118347.
16. A. Thivasasith, et al. , Carbon Trends, 2022, 7, 100158.
17. S. Tantisriyanurak, et al. , ACS Appl. Nano Mater., 2023, 6, 8784–8794.
18. G. Yang, et al. , Adv. Mater., 2020, 32, 2000455.
19. L. Gao, et al. , Appl. Catal., B, 2020, 261, 118235.
20. Z. Shams-Ghahfarokhi and A. Nezamzadeh-Ejhieh, Mater. Sci. Semicond. Process., 2015, 39, 265–275.
21. A. Chakraborty and B. Sun, Appl. Therm. Eng., 2014, 72, 190–199.
22. X. Zhang, et al. , Analyst, 2016, 141, 5527–5534.
23. H. Jin, et al. , Chem. Rev., 2018, 118, 6337–6408.
24. K. Xu, et al. , J. Am. Chem., 2015, 137, 4119–4125.
25. C. Chen, et al. , J. Energy Chem., 2021, 54, 528–554.
26. T. Kahlstorf, et al. , Green Chem., 2023, 25, 8679–8686.
27. N. Heidary and N. Kornienko, Chem. Commun., 2020, 56, 8726–8734.

---

Footnote

† Electronic Supplementary Information (ESI) available. See DOI: https://doi.org/10.1039/d4cc03356f

---