1. Introduction
Liquid hydrogen carriers (LHC), such as liquid organic hydrogen carriers (LOHC) and ammonia (NH
3), are a promising and environmentally friendly alternative to liquid fuels [
1], the usage of which increases the CO
2 content in the atmosphere. Traditionally, catalytic methods have been used to produce LHC. Recently, however, the attention of researchers has increasingly turned to electrochemical methods. A comparison of these methods [
2,
3] shows the advantage of the latter, since the synthesis can be carried out at room temperature and hydrogen can be obtained
in situ most often from water.
Promising methods for ammonia synthesis are electrochemical, electrocatalytic, photocatalytic, photoelectrocatalytic, and bio-catalytic [
4]. Electrocatalytic methods are the most attractive because they combine the advantages of catalytic and electrochemical methods. The advantages of the latter are the use of an optimum potential or current and the ability to replace known industrial processes [
5]. Moreover, the advantage of electrochemical methods of ammonia synthesis over the traditional Haber-Bosch process in terms of greenhouse gas formation was shown [
4]. Electrochemical synthesis of ammonia can be carried out not only in aqueous electrolytes, but also in ionic liquids and solid electrolytes [
6].
Various nitrogen-containing compounds can be used for electrocatalytic production of ammonia. The electroreduction reactions of nitrogen (NRR), nitrogen oxides (NOxRR), nitrite (NO
2RR) and nitrate (NO
3RR) are actively investigated [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17]. Two fundamentally different NRR mechanisms are reviewed [
8]: dissociative and associative. The first mechanism requires the use of a highly efficient catalyst to break the inert triple bond in the nitrogen molecule. Free energy diagrams for the electrochemical synthesis of ammonia obtained by DFT calculations are presented [
10]. The limiting stage can be reductive adsorption of nitrogen to form a *N
2H particle or simultaneous transfer of protons and electrons to form a *NH
2 particle. Based on thermodynamic data [
16], the most promising way is the nitrate reduction reaction compared to other substances. The mechanism of NO
3RR in acidic, neutral and alkaline media has been described in detail [
7,
12]. Moreover, NO
3RR is significant from the point of view of environmental problems as a means of nitrate removal from wastewater [
15,
16]. The application of NO
3RR in water treatment of nitrate has some serious limitations that do not yet allow this process to be implemented in the industrial scale. These are, firstly, the multi-step electron and proton transfer, and secondly, the low concentration of ions to be removed. While producing ammonia by electrocatalytic reduction, a side reaction of hydrogen evolution (HER) is possible, since the reaction potential of ammonia synthesis in an alkaline medium is E
o = -0.132 V (vs. SHE, pH=14) [
17]:
The hydrogen release is also preferable from the kinetic point of view, because fewer electrons are required [
8]. Based on the above, one of the criteria for the NO
3RR selectivity [
18] is the evaluation of the NO
3RR selectivity compared to HER. Other important characteristics of electrocatalytic reduction are the Faradaic efficiency (FE, %), ammonia yield rate (mol s
-1), turnover frequencies (TOF), energy efficiency (EE, %), and NO
3- conversion. To determine these characteristics, it is necessary to know the concentration of ammonia in the solution obtained after reduction. Spectrophotometry, ion chromatography,
1H NMR spectroscopy, enzymatic, fluorometric methods, the use of ion-selective electrodes, conductivity, and titrimetric methods are used for this purpose [
9]. Linear voltammetry (LsV) in a solution containing nitrate ions is used to determine the intermediate stages of reduction, evaluate the reaction efficiency and catalyst stability [
19,
20].
Electrode-catalysts for electroreduction reactions should provide high activity, selectivity, and stability [
20]. Several reviews [
7,
20] outline the strategy for preparation of efficient catalysts, including porous materials, using bimetals and alloys, single atom catalysts (SAC), modification by hetero- and nanostructures.
Noble metals (Pt, Pd, Rh, Ru) as catalyst-electrodes have been mainly investigated to establish the mechanism [
7], since they show weak adsorption and low rate of nitrate reduction and significant contribution of HER. In terms of efficiency and cost, Cu- and Co-based catalysts are of interest [
19,
21]. Copper, its alloys with Fe, Co, Ni and noble metals, copper oxides, Cu single atom catalysts were reviewed [
19]. Cu SAC achieved a maximum conversion of NH
3.
The application of pure cobalt electrocatalysts encounters problems related to low conductivity and low stability in acidic media. The use of functionalized multi-walled carbon nanotubes (MWCNTs) as carriers for active cobalt catalysts demonstrated an FE of 84.72% in 0.1 M KOH with 0.1 M NO
3- at -0.16 V vs. RHE. Surface composition analysis by XPS showed the presence of Co
3O
4 [
22]. A comparison of 12 common transition metal oxide catalysts for NO
3RR at a high cathodic current density was carried out [
23]. The Co
3O
4 catalyst shows the highest FE of 85.15% at -0.25 V vs. RHE.
The study and application of Fe containing catalysts for the electrochemical reduction of nitrate to ammonia has so far received insufficient attention compared to other base metals. Iron as a catalyst is employed in industry (Haber-Bosch process) and is present in various forms as a catalyst in the environment (nitrogenase enzyme) [
24]. The disposal of such catalysts and release into the environment is not a factor that seriously limits the development and application of these electrocatalystsThe review [
25] reported the use of pure iron, nanocomposites, bimetallic catalysts and iron oxides in the NO
3RR reaction. The FE was in the range of 74 - 98 %. Pure iron has low stability in aqueous media due to corrosion. Nanocomposites and bimetallic catalysts, as well as reasonable design of iron single-atom catalysts (Fe SAC) [
26] can be a solution to this problem. For Fe SAC, the maximum FE was ~75% at -0.66 V with the highest NH
3 production rate of 0.46 mmol/h/cm
2 at -0.85 V. The rate of ammonia production on Fe SAC is not high enough, so the rational choice of a carbon substrate, increase of the Fe SAC content on the substrate, introduction of components preventing aggregation of Fe SAC particles are important. Iron is used in alloys or bimetallic catalysts to increase the efficiency of NO
3RR.
According to [
24], the proximity of the energies of the d-orbitals of Cu with the lowest unoccupied molecular π* orbitals of NO
3- facilitates electron transfer, which makes Cu an attractive electrode-catalyst. A dual-atom catalyst Fe/Cu due to the improved electron transfer on Cu demonstrates FE of 92.51% for NRR under alkaline conditions.
In conclusion, the turn to non-platinum group metals, especially catalysts based on Fe, Co, Cu, and Ni, has opened more economical approaches to NO
3RR since adsorption of *NO
3 and its conversion to *NO
2 occurs readily on the 3d-transition metal centers, and these metal centers play a crucial role mainly in the adsorption pathways at the O- and N-ends leading to the formation of NH
3 [
27].
The aim of this work was to obtain (one-step synthesis) and characterize a model bicatalyst for NO3RR, which is in demand in the composition and, first of all, serves as a starting point for future work on NO3RR, NO2RR and NRR. As shown above, Fe- and Fe-Co-containing electrocatalysts are not yet as well studied as, for example, systems containing Cu and Cu-Co and even more so noble metals. Nanoparticles of Co, Fe, and bimetallic Fe/Co catalysts were obtained by the method of electrodeposition from aqueous solutions on a graphite substrate. Mono- and bicatalysts were tested in the reaction of electro-reduction of nitrate into ammonia and experimental data were obtained, which allowed us to suggest the nature of active metal nanoparticles (NPs).
3. Materials and Methods
3.1. Materials
The commercial cobalt (II) sulfate heptahydrate (CoSO4⋅7H2O, chemical purity), sodium nitrate (NaNO3, chemical purity), sodium sulfate (Na2SO4, chemical purity), boric acid (H3BO3, chemical purity), and graphite plates (mark V2L12) were used. All reagents were used without additional purification. Distilled water was employed for all experiments.
3.2. Catalyst Preparation
For the preparation of the electrocatalyst, graphite plates with dimensions of 50×7×1 mm were used as a support. Graphite plates were mechanically abraded and pre-cleaned, washed thoroughly with distilled water and air dried. Fe catalysts were prepared by electrodeposition on the graphite substrate from a 0.1 M FeSO
4 aqueous solution under galvanostatic conditions at a current density of -1 mA cm
2 for 5 min or 30 min. The catalysts with Co-NPs were prepared by electrodeposition from an aqueous solution of 0.1 M CoSO
4 with 1 M Na
2SO
4 as a background electrolyte and 0.5 M H
3BO
3 as a pH buffer [
38]. Deposition of cobalt was carried out under potentiostatic conditions at E = - 0.75 V (vs Ag/AgCl) for 5 min and 30 min. A conventional cylindrical single-compartment (30 mL) electrochemical cell was used, where a platinum wire and a chlorosilver electrode (Ag/AgCl) were used as the counter electrode and reference electrode, respectively. The Autolab PGSTAT 302N potentiostat-galvanostat equipped with Nova 2.1.5 (Netherlands-Switzerland) software was used for electrodeposition.
3.3. Electrochemical Measurements
Electrochemical measurements were carried out at room temperature using an Autolab PGSTAT 302N potentiostat with a three-electrode cell and an Ag/AgCl electrode as a reference electrode. Electrocatalysts on the graphite substrate were used as the working electrode, and platinum wire was used as the counter-electrode.
Linear voltammetry (LsV) over the range between (-0.25 V) and ≈ (-1.185 V) vs. RHE at a potential scan rate of 50 mV s-1 was performed in a cathode-anode space-separated cell with a total volume of 60 mL.
Electrochemical reactions of NO3RR were carried out during 1 hour in a cell with a separated cathode-anode space and a total volume of 60 mL. The electrolyte was a solution of 100 ppm (1.2 mmol/L) NaNO3 in 0.05 M Na2SO4, degassed by an Ar flow before the tests. All potential values were recalculated vs. the reversible hydrogen electrode (RHE) according to the formula ERHE = E applied Ag/AgCl + 0.202 + 0.059 × pH, unless otherwise noted.
Chronoamperometry tests were carried out in the potential range (-0.385 V) to (-1.185 V) vs. RHE for 1 h to determine the ammonia yield rates and Faradaic efficiencies.
3.4. Detection of Ammonia
The detection of the ammonia content after NO
3RR was carried out using the indophenol method, according to the methodology given elsewhere [
39]. UV-vis absorption spectra (
Figure 10a) were recorded using a Shimadzu 3600 Plus spectrophotometer (Shimadzu, Japan) in a standard 1 cm quartz cuvette. Two milliliters of 5 wt% sodium salicylate in 1.0 M NaOH were added to 2 ml of the tested solution, then 1 ml of 0.05 M NaClO and Na
2[Fe(NO)(CN)
5] (0.2 ml, 1 wt%) were added. The solutions were kept at 40 °C for 1 hour. The absorption maximum is observed at λ=652 nm, for which a calibration graph was plotted. The resulting calibration graph is described by the equation (y = 0.4217 x + 0.0641; R² = 0.9993) and shows a good linear relationship between the absorbance value and NH
3 concentration in the range from 0.25 mg/mL to 10 mg/L (Figure 10b).
The Faradaic efficiency was determined by the formula:
where n (NH
3) denotes the amount (mol) of NH
3;
F is the Faradaic constant (96,485 C mol
-1);
Q is the total charge passed through the electrode, 8 is the number of electron (
n) transfers required to form 1 mol of ammonia.
The ammonia yield rate (yield) was defined as:
where C
NH3 denotes the mass concentration (μg ml
−1) of NH
3 calculated from the UV-Vis spectra,
t is the electrolysis time;
S is the geometric area of the working electrode (1 cm
2);
V is the volume of the electrolyte.
3.5. ECSA Evaluation
The ECSA value was calculated from the value of the double layer electrochemical capacitance (C
dl) obtained by measuring CV (cyclic voltammogram) in the double layer potential range, i.e., the non-Faradaic area. All catalysts were scanned in the potential range from 0.165 V to 0.265 V vs. RHE in NaNO
3 (1.2 mM) in 0.05 M Na
2SO
4 at different scan rates (10 to 100 mV s
-1). The values of the current density at 0.215 V vs. RHE at different scan rates were calculated and the curves of the dependence of scan rates for each catalyst were plotted. The dependences of the current densities on scan rates were obtained, and the Cdl values were obtained accordingly. ECSA was calculated as
where
Cs (= 0,4 F∙m
-2) is the total specific capacitance for an atomically smooth planar surface under homogeneous electrolytic conditions.
3.6. Impedance Response Testing
Impedance spectra were measured in a three-electrode undivided cell (60 mL) at room temperature in a solution 1.2 mM of NaNO3 in 0.05 M Na2SO4. An Ag/AgCl reference electrode was used. The auxiliary electrode was a platinum wire. Measurements were carried out with a P-40X potentiostat with an electrochemical impedance measurement module FRA-24M (Electrochemical instruments, Russia) in the frequency range from 50 kHz to 0.01 Hz at an AC voltage amplitude of 20 mV. The time of immersion of the sample corresponded to the time of impedance measurement without preliminary exposure in the medium.
3.7. Material Characterization
Scanning electron microscopy (SEM) analysis was made with a LEO EVO 50 xvp electron microscope (Carl Zeiss, Germany) equipped with an X-ray energy-dispersive spectrometer (EDS).
The powder X-ray diffractograms (XRD) were obtained with a STOE STADI P diffractometer (STOE & Cie GmbH: Darmstadt, Germany) equipped with a Ge-monochromator, CuKα1 emission, λ = 1.54056 Ǻ, linear PSD in the transmittance geometry. The samples were examined in the region 2θ = 10-90° with a scanning step of 0.01° and an exposure time of 15 seconds per point. The samples were identified by comparing theoretical and experimentally obtained X-ray diffraction patterns using WinXPOW version 2.24.
XPS measurements were performed using a PREVAC EA15 spectrometer. In the current work, AlKα radiation (hν = 1486.6 eV, 150 W) was used as a primary radiation source. The pressure in the analytical chamber did not exceed 5×10−9 mbar during spectra acquisition. The binding energy scale was pre-calibrated using the positions of Ag 3d5/2 (368.3 eV) and Au 4f7/2 (84.0 eV) from silver and gold foils, respectively. The powdered samples were supported onto a double-sided conducting scotch tape. To take into account the effect of surface charging, the C1s line at (Eb = 284.8 eV) from the carbon contamination was used as an internal standard.
Figure 1.
SEM images: (a) Co-NPs deposited for 5 min on a graphite substrate; (b) Co-NPs deposited for 30 min on a graphite substrate; (c) Fe(5min)/C; (d) Fe(30min)/C; (e) Fe(5min)@Co(5min)/C; (f) Fe(5min)@Co(30min)/C; (g) elemental mapping for Co, Fe, and O elements on a carbon substrate for Fe(5min)@Co(30min)/C.
Figure 1.
SEM images: (a) Co-NPs deposited for 5 min on a graphite substrate; (b) Co-NPs deposited for 30 min on a graphite substrate; (c) Fe(5min)/C; (d) Fe(30min)/C; (e) Fe(5min)@Co(5min)/C; (f) Fe(5min)@Co(30min)/C; (g) elemental mapping for Co, Fe, and O elements on a carbon substrate for Fe(5min)@Co(30min)/C.
Figure 2.
XP spectra for the sample Fe(5min)/C: (a) survey spectrum, and high-resolution spectra: (b) Fe 2p, (c) C 1s, (d) O 1s.
Figure 2.
XP spectra for the sample Fe(5min)/C: (a) survey spectrum, and high-resolution spectra: (b) Fe 2p, (c) C 1s, (d) O 1s.
Figure 3.
XP spectra for the sample Fe(5min)@Co(30min)/C: (a) survey spectrum, and high-resolution spectra: (b) Fe 2p3/2 , (c) Co 2p, (d) C 1s, (e) O 1s.
Figure 3.
XP spectra for the sample Fe(5min)@Co(30min)/C: (a) survey spectrum, and high-resolution spectra: (b) Fe 2p3/2 , (c) Co 2p, (d) C 1s, (e) O 1s.
Figure 4.
XRD pattern for electrocatalyst samples: (a) Fe(30min)C; (b) Fe(30min)@Co(30min)/C; (c) Co(30min)/C; (d) Graphite substrate and all samples in the insert.
Figure 4.
XRD pattern for electrocatalyst samples: (a) Fe(30min)C; (b) Fe(30min)@Co(30min)/C; (c) Co(30min)/C; (d) Graphite substrate and all samples in the insert.
Figure 5.
Linear voltammetric curves in Na2SO4 electrolyte containing and not containing nitrate ions at a potential scan rate of 50 mV s-1 for electrocatalyst samples: (a) Fe(5min)C; (b) Fe(30min)/C; (c) Fe(5min)@Co(5min)/C; (d) Fe(5min)@Co(30min)/C.
Figure 5.
Linear voltammetric curves in Na2SO4 electrolyte containing and not containing nitrate ions at a potential scan rate of 50 mV s-1 for electrocatalyst samples: (a) Fe(5min)C; (b) Fe(30min)/C; (c) Fe(5min)@Co(5min)/C; (d) Fe(5min)@Co(30min)/C.
Figure 6.
The process of ammonia synthesis by NO3RR for Fe(5min)@Co(30min)/C; at different potentials: (a) chronoamperometric curves in the 0.05 M Na2SO4 with 1.2 mM NaNO3 electrolyte; (b) UV-vis spectrum corresponding to the concentrations of the resulting product at λ=652 nm.
Figure 6.
The process of ammonia synthesis by NO3RR for Fe(5min)@Co(30min)/C; at different potentials: (a) chronoamperometric curves in the 0.05 M Na2SO4 with 1.2 mM NaNO3 electrolyte; (b) UV-vis spectrum corresponding to the concentrations of the resulting product at λ=652 nm.
Figure 7.
The values of FE and yield rate in NO3RR: (a) the resulting graph; and for single electrocatalysts: (b) Fe(5min)@Co(30min)/C, (c) Fe(5min)@Co(5min)/C, (d) Fe(5min)C and (e) Fe(30min)/C. (f) NH3 yield rate of NO3RR at the potential -0.785 V.
Figure 7.
The values of FE and yield rate in NO3RR: (a) the resulting graph; and for single electrocatalysts: (b) Fe(5min)@Co(30min)/C, (c) Fe(5min)@Co(5min)/C, (d) Fe(5min)C and (e) Fe(30min)/C. (f) NH3 yield rate of NO3RR at the potential -0.785 V.
Figure 8.
(a) Cyclic voltammograms for the Fe(5min)@Co(30min)/C electrocatalyst sample for a series at scan rates of 10, 20, 40, 60, 80 and 100 mV s−1 from 0.165 to 0.265 V (RHE). (b) The electrochemically active surface of the electrocatalyst samples presented as a double layer capacity.
Figure 8.
(a) Cyclic voltammograms for the Fe(5min)@Co(30min)/C electrocatalyst sample for a series at scan rates of 10, 20, 40, 60, 80 and 100 mV s−1 from 0.165 to 0.265 V (RHE). (b) The electrochemically active surface of the electrocatalyst samples presented as a double layer capacity.
Figure 9.
(a) Nyquist curves for electrocatalyst samples Fe(5min)@Co(30min)/C, Fe(5min)/C and a graphite substrate in 1.2 mM NaNO3 with 0.05 М Na2SO4; (b) enlarged fragment.
Figure 9.
(a) Nyquist curves for electrocatalyst samples Fe(5min)@Co(30min)/C, Fe(5min)/C and a graphite substrate in 1.2 mM NaNO3 with 0.05 М Na2SO4; (b) enlarged fragment.
Figure 10.
(a) UV-Vis spectra, and (b) calibration line for testing NH3.
Figure 10.
(a) UV-Vis spectra, and (b) calibration line for testing NH3.
Table 1.
The Samples of electrocatalysts and features of their preparation.
Table 1.
The Samples of electrocatalysts and features of their preparation.
Designation |
General brief description |
C |
Graphite is the initial substrate |
Fe(30min)/C |
Fe-NPs (deposited at 30 min) on the substrate С |
Fe(5min)/C |
Fe-NPs (deposited at 5 min) on the substrate С |
Fe(5min)@Co(5min)/C |
Fe-NPs (deposited at 5 min) on the Сo-NPs layer (deposited at 5 min) on the substrate C |
Fe(5min)@Co(30min)/C |
Fe-NPs (deposited at 5 min) on the Сo-NPs layer (deposited at 30 min) on the substrate C |