Voltammetric Studies of Nitrate Reduction at Boron‐Doped Diamond Thin‐Film Electrodes of Different Microstructure
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Electroanalysis |
| Published | 2026-05-01 |
| Authors | Margaret R. Conway, Greg M. Swain |
| DOI | 10.1002/elan.70153 |
A nanocrystalline boron-doped diamond electrode reduced nitrate to ammonia. It achieved an 85 ± 5% Faradaic efficiency and a production rate of 42 ± 5 μmol cm⁻² h⁻¹. The measurement came from a 1 h electrolysis at −1.8 V versus Ag/AgCl in 0.1 M NaNO₃ and 1 M NaOH. Across three highly boron-doped polycrystalline films, microstructure changed peak current but did not shift nitrate onset or reduction peak potentials. The authors identify nanocrystalline BDD as a nonmetal electrode for ammonia production, but the product claim remains preliminary.
Electrode growth and material control
Section titled “Electrode growth and material control”Three BDD films were grown by microwave-assisted CVD on polished, conducting p-type Si(111) substrates. Every film received 0.1% B₂H₆ diluted in H₂. The estimated boron concentration was approximately 10²¹ B cm⁻³, or about 0.5% B/C.
Each CH₄/H₂ deposition ended with H₂-plasma cooling. Power and pressure fell to 300 W and 5 Torr over 30 min. This maintained hydrogen termination and reduced adventitious surface sp² carbon.
Table 1. Film growth conditions. Every film used the 0.1% B₂H₆/H₂ dopant.
| Film | Source gas | Power and pressure | Growth time | Nominal thickness |
|---|---|---|---|---|
| BMD | 2% CH₄/H₂ | 1000 W, 40 Torr | 5.5 h | 3 μm |
| BND | 1% CH₄/H₂ | 1000 W, 40 Torr | 4 h | 1 μm |
| BUNCD | 1% CH₄/5% H₂/94% Ar | 800 W, 140 Torr | 4 h | 0.5 μm |
Microstructure and redox activity
Section titled “Microstructure and redox activity”SEM measured average crystallite sizes of 0.43 μm for BMD, 0.28 μm for BND, and 0.16 μm for BUNCD. BMD included facets as long as 1.6 μm. Each BUNCD cluster contained smaller grains about 10 nm across.
The first-order diamond Raman peaks were 1312 cm⁻¹ for BMD and BUNCD and 1308 cm⁻¹ for BND. The reference position for single-crystal diamond was 1332 cm⁻¹. A broad band near 1500 cm⁻¹ indicated some nondiamond sp² carbon in every film.
All three electrodes gave reversible to quasi-reversible responses for Fe(CN)₆³⁻/⁴⁻, Ru(NH₃)₆³⁺/²⁺, IrCl₆²⁻/³⁻, and methyl viologen. Ferri/ferrocyanide peak separations were 81 to 96 mV. For Fe(CN)₆, Ru(NH₃)₆, and IrCl₆, oxidation peak currents increased with the square root of scan rate from 50 to 250 mV/s. Across formal potentials from −1.0 to 1.0 V versus Ag/AgCl, the authors found semi-metallic electronic behavior in all films.
Nitrate reduction potentials
Section titled “Nitrate reduction potentials”Cyclic voltammetry used deaerated 0.1 M Na₂SO₄ at pH 6 to 7. Solutions were N₂-sparged for 20 min, and an N₂ blanket prevented oxygen redissolution. Scans ran from 0.0 to −2.0 V at 25 to 100 mV/s. The exposed area was 0.28 cm², and currents were normalized to it.
Each BDD electrode showed two cathodic peaks. The first near −1.2 V was assigned to nitrate-to-nitrite reduction with 2H⁺/2e⁻ stoichiometry. The second near −1.6 V was assigned to nitrite-to-ammonia or ammonium reduction with 8H⁺/6e⁻ stoichiometry. Nitrite-only measurements began near −1.5 V and supported that assignment.
Table 2. Background-subtracted fifth-scan potentials at 50 mV/s in 16 mM NaNO₃ and 0.1 M Na₂SO₄. Potentials are versus Ag/AgCl. The paper reports n = 3 for each BDD type.
| Electrode | Onset (V) | First peak (V) | Second peak (V) |
|---|---|---|---|
| BMD | −0.852 ± 0.011 | −1.049 ± 0.023 | −1.582 ± 0.031 |
| BND | −0.871 ± 0.029 | −1.126 ± 0.043 | −1.654 ± 0.014 |
| BUNCD | −0.875 ± 0.019 | −1.136 ± 0.079 | −1.595 ± 0.090 |
| Cu | −0.796 ± 0.009 | −1.010 ± 0.009 | −1.641 ± 0.096 |
BND had the largest nominal second-peak current at −781 ± 122 μA cm⁻². Copper reached −3580 ± 728 μA cm⁻² at its second peak, and its onset was about 50 mV less negative. Copper voltammograms were less stable across repeated scans. On BDD, hydrogen evolution began near −1.6 V. On copper, it began near −1.3 V.
Reaction assignment and kinetic tests
Section titled “Reaction assignment and kinetic tests”Detailed tests used only BND, made with 1% CH₄/H₂. Repeated scans kept Eₚ,₁ and Eₚ,₂ stable. Current at Eₚ,₂ declined during each scan, but solution mixing restored its original value. The authors attributed the loss to a depletion layer rather than fouling.
Both peak currents increased linearly with nitrate concentration over 16, 50, and 100 mM. The second peak remained about twice the first. The first-peak current also rose linearly with the square root of scan rate. The second peak showed the same trend with a weaker correlation. The authors considered the first reaction diffusion controlled and the second consistent with diffusion control.
The proposed reaction sequence was NO₃⁻ to NO₂⁻ to possibly NO to NH₃. Nitric oxide was not measured, so the proposed pathway remains partly interpretive. From pH 7 to 11 in phosphate buffer, both peaks shifted about −20 mV per pH unit. This supported proton transfer in the reductions. Phosphate may compete for adsorption sites, but that effect was not resolved.
Batch electrolysis and product analysis
Section titled “Batch electrolysis and product analysis”The divided cell used two chambers of about 6 mL each, separated by a glass frit. A 1.13 cm² BDD cathode faced an equal-area graphite foil anode. Catholyte was stirred at about 300 rpm, and both solutions were Ar-sparged for 20 min.
Constant potential was −1.8 V for 1 h in 0.1 M NaNO₃. The supporting electrolyte was 0.1, 0.5, or 1 M NaOH. Current density increased as hydroxide concentration increased. The authors associated this trend with less charge going to hydrogen evolution.
At 1 M NaOH, ammonia had an 85 ± 5% Faradaic efficiency and a production rate of 42 ± 5 μmol cm⁻² h⁻¹. Nitrite had a 7 ± 0.6% efficiency and a rate near 14 μmol cm⁻² h⁻¹. Griess and salicylic-hypochlorite assays measured nitrite at 540 nm and ammonia at 655 nm. About 7% of the charge was unaccounted for, possibly through hydrogen evolution or products such as nitrogen. Nitrogen gas was not measured.
Limits and engineering implications
Section titled “Limits and engineering implications”Three films differed in gas composition, thickness, and sp² carbon as well as grain size, so the experiment did not isolate grain size. Detailed concentration, scan-rate, pH, and batch tests used only the 1% CH₄/H₂ BND electrode.
Some films had been stored for years in laboratory air, leaving them mainly hydrogen terminated with some oxygen. The authors did not test termination independently. They also reported variable peak currents and suggested differences in electrochemically active area as one cause.
Earlier work cited by the authors found favorable grains about 2 μm and linked performance to fewer grain boundaries and less sp² carbon. Here, grain size did not change onset or peak potentials among the three films. The authors call the sp² carbon role unclear and propose testing more CH₄/H₂ compositions.
The neutral electrolyte voltammetry and alkaline batch electrolysis were separate tests. Only initial 1 h divided-cell runs were reported, with no flow-cell, long-duration, or nitrogen-gas balance data.
For material selection, the measured data support highly boron-doped BDD when low background current and stable repeated scans matter. The 1% CH₄/H₂ nanocrystalline film gave the highest nominal current and the demonstrated ammonia selectivity. Grain class alone should not set selection because measured pH and electrolyte conditions changed the response. Prior work identifies termination as a control, and the authors call for reactor optimization.
Materials and support
Section titled “Materials and support”The paper requires a conductive, highly boron-doped diamond thin-film cathode on a Si substrate and a Cu current collector. Its electrodes were mainly hydrogen terminated, which is the surface condition used for the reported performance. 6CCVD’s diamond materials catalog includes boron-doped diamond grown by MPCVD and Cu metallization, while engineering support is the contact for defining a custom film and electrical connection.
View Original Abstract
The electrochemical reduction of nitrate at polycrystalline boron‐doped diamond (BDD) electrodes possessing three different microstructures was studied. The BDD thin films were grown on conducting Si substrates using 2 and 1% CH 4 /H 2 and 1% CH 4 /Ar source gas mixtures. All films were highly boron‐doped using 0.1% B 2 H 6 /H 2 in the source gas to produce a doping level of ca. 10 21 cm −3 or ~5000 ppm. BDD films with a hydrogen surface termination were investigated. All BDD films microstructures exhibited a high degree of electrochemical activity for a suite of redox systems: (IrCl 6 −2/−3 , Fe(CN) 6 −3/−4 , Ru(NH 3 ) +3/ + 2 , and methyl viologen +2/+/0 ). The reduction of nitrate proceeded similarly on the BDD electrodes in terms of onset and peak potentials in 0.1 M Na 2 SO 4 . In other words, the BDD electrode microstructure, at least for these three film types, had no differential influence on the nitrate reduction polarization curves. However, the largest peak current densities were observed for the 1% CH 4 /H 2 nanocrystalline BDD film. The reaction proceeds through a two‐step process involving the initial reduction of nitrate to nitrite at less negative potentials (ca. −1.2 V vs. Ag/AgCl), followed by further reduction to ammonia or other products at more negative potentials (ca. −1.6 V). Initial batch electrolysis experiments confirmed ammonia as the primary electrolysis product at the nanocrystalline BDD thin films in 1 M NaOH (−1.8 V for 1 h), achieving a production rate of up to 42 µmol/cm 2 h at a Faradaic efficiency of 85%. The results align with earlier reports in the literature and contribute meaningfully to the advancement of nitrate reduction research using BDD electrodes.
Boron Doped Diamond On Titanium Bdd Electrode
View Specs →
Boron Doped Diamond Electrode On Niobium Bdd Nb Electrode
View Specs →