Understanding Continuous Lithium mediated Electrochemical Nitrogen Reduction

Highlights

• •

  Lithium-metal-mediated approach for nitrogen electroreduction to ammonia is studied

• •

  Ammonia production rate of (7.9 ± 1.6) × 10⁻⁹ mol cm⁻² s⁻¹ is achieved

• •

  Faradaic efficiency of 18.5% ± 2.9% is obtained

• •

  A kinetic-transport model incorporating observed diffusion limitations is developed

Context & Scale

Ammonia is an important commodity chemical that is produced predominantly via the Haber-Bosch process, which is CO₂-emitting and highly centralized. Electrochemical nitrogen reduction is proposed as a renewable and distributed alternative to Haber-Bosch, which could allow access to ammonia in remote regions of the world. Herein, a lithium-mediated nitrogen reduction approach is presented. A coupled kinetic-transport model is developed to determine the rate-limiting factors in the system. The low rate of nitrogen diffusion was found to be an important feature of the system that limits theoretical and practical yields. Lithium metal was found to undergo an activation process, only after which it can reduce nitrogen to ammonia. These results suggest that future studies in lithium metal activation and architectural changes to alleviate nitrogen transport limitations are imperative to making this process viable in practice.

Summary

Ammonia is a large-scale commodity chemical that is crucial for producing nitrogen-containing fertilizers. Electrochemical methods have been proposed as renewable and distributed alternatives to the incumbent Haber-Bosch process, which utilizes fossil fuels for ammonia production. Herein, we report a mechanistic study of lithium-mediated electrochemical nitrogen reduction to ammonia in a non-aqueous system. The rate laws of the main reactions in the system were determined. At high current densities, nitrogen transport limitations begin to affect the nitrogen reduction process. Based on these observations, we developed a coupled kinetic-transport model of the process, which we used to optimize operating conditions for ammonia production. The highest Faradaic efficiency observed was 18.5% ± 2.9%, while the highest production rate obtained was (7.9 ± 1.6) × 10⁻⁹ mol cm⁻² s⁻¹. Our understanding of the reaction network and the influence of transport provides foundational knowledge for future improvements in continuous lithium-mediated ammonia synthesis.

Graphical Abstract

Keywords

• electrochemical ammonia production
• nitrogen reduction
• ammonia
• electrochemistry
• lithium-mediated
• mechanistic studies
• kinetic modeling

Introduction

Ammonia is one of the most widely produced industrial chemicals, with a worldwide production of 170 million tons per annum.

¹

• Food and Agriculture Organization of United Nations

World fertilizer trends and outlook to 2020.

• Google Scholar

At least 80% of ammonia is used in fertilizer production, making ammonia production critical for agriculture.

Today, ammonia is produced predominantly via the Haber-Bosch process, which involves reacting nitrogen and hydrogen over heterogeneous catalysts at high temperatures of 450°C–500°C and pressures up to 200 atm.

Hydrogen gas is typically generated on site from fossil fuels via steam reforming; this leads to significant carbon dioxide emissions of 1.9–9.3 tons CO₂ per ton of NH₃.

^,

In total, the Haber-Bosch process consumes approximately 1% of the world's energy supply and emits over 450 million tons of CO₂ per year.

^,

⁶

• Institute for Industrial Productivity

Ammonia. Industrial Efficiency Technology Database.

• Google Scholar

Due to the harsh operating conditions and on-site hydrogen production, the process is economically viable only at large scales.

⁷

• Foster S.L.
• Bakovic S.I.P.
• Duda R.D.
• Maheshwari S.
• Milton R.D.
• Minteer S.D.
• Janik M.J.
• Renner J.N.
• Greenlee L.F.

Catalysts for nitrogen reduction to ammonia.

• Crossref
• Scopus (599)
• Google Scholar

This makes it difficult to produce ammonia in a modular and distributed manner through Haber-Bosch. Distributed ammonia production, if achieved, could reduce inefficiencies and costs associated with fertilizer distribution.

In order to overcome these challenges, methods to produce ammonia electrochemically from nitrogen and water at ambient conditions have been proposed, which could help to reduce the CO₂ footprint of the process. Operating at ambient conditions without the use of steam reforming enables smaller scale processes with lower capital costs.

⁹

• Chen J.G.
• Crooks R.M.
• Seefeldt L.C.
• Bren K.L.
• Bullock R.M.
• Darensbourg M.Y.
• Holland P.L.
• Hoffman B.
• Janik M.J.
• Jones A.K.
• et al.

Beyond fossil fuel–driven nitrogen transformations.

• Crossref
• PubMed
• Scopus (743)
• Google Scholar

The reduced CO₂ footprint and capital costs, as well as the economics associated with distributed production, can make electrochemical reduction competitive with industrial Haber-Bosch in certain contexts. Many catalyst chemistries have been proposed for electrochemical nitrogen reduction, including structured noble metals,

¹⁰

• Bao D.
• Zhang Q.
• Meng F.-L.
• Zhong H.-X.
• Shi M.-M.
• Zhang Y.
• Yan J.
• Jiang Q.
• Zhang X.

Electrochemical reduction of N₂ under ambient conditions for artificial N₂ fixation and renewable energy storage using N₂/NH₃ cycle.

• Crossref
• Scopus (796)
• Google Scholar

metal oxides,

¹¹

• Kong J.
• Lim A.
• Yoon C.
• Jang J.H.
• Ham H.C.
• Han J.
• Nam S.
• Kim D.
• Sung Y.
• Choi J.
• et al.

Electrochemical synthesis of NH₃ at low temperature and atmospheric pressure using a γ-Fe₂O₃ catalyst.

• Crossref
• Scopus (184)
• Google Scholar

metal nitrides,

¹²

• Yang X.
• Nash J.
• Anibal J.
• Dunwell M.
• Kattel S.
• Stavitski E.
• Attenkofer K.
• Chen J.G.
• Yan Y.
• Xu B.

Mechanistic insights into electrochemical nitrogen reduction reaction on vanadium nitride nanoparticles.

• Crossref
• PubMed
• Scopus (292)
• Google Scholar

metal sulfides,

nitrogen- and boron-doped carbon,

¹⁴

• Lv C.
• Qian Y.
• Yan C.
• Ding Y.
• Liu Y.
• Chen G.
• Yu G.

Defect engineering metal-free polymeric carbon nitride electrocatalyst for effective nitrogen fixation under ambient conditions.

• Crossref
• PubMed
• Scopus (460)
• Google Scholar

and lithium metal.

^,

¹⁶

• McEnaney J.M.
• Singh A.R.
• Schwalbe J.A.
• Kibsgaard J.
• Lin J.C.
• Cargnello M.
• Jaramillo T.F.
• Nørskov J.K.

Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

• Crossref
• Google Scholar

^,

^,

¹⁸

• Kim K.
• Lee S.J.
• Kim D.-Y.
• Yoo C.-Y.
• Choi J.W.
• Kim J.-N.
• Woo Y.
• Yoon H.C.
• Han J.

Electrochemical synthesis of ammonia from water and nitrogen: a lithium-mediated approach using lithium-ion conducting glass ceramics.

• Crossref
• PubMed
• Scopus (43)
• Google Scholar

^,

Despite the large variety of reported catalysts, many of them exhibit low Faradaic efficiencies (FEs) and rates for producing ammonia, often below 10% and 10⁻¹⁰ mol cm⁻² s⁻¹, respectively (Table S1).

^,

Among these chemistries, lithium-metal-based methods report some of the highest FEs (Table S1).

^,

¹⁶

• McEnaney J.M.
• Singh A.R.
• Schwalbe J.A.
• Kibsgaard J.
• Lin J.C.
• Cargnello M.
• Jaramillo T.F.
• Nørskov J.K.

Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

• Crossref
• Google Scholar

^,

^,

¹⁸

• Kim K.
• Lee S.J.
• Kim D.-Y.
• Yoo C.-Y.
• Choi J.W.
• Kim J.-N.
• Woo Y.
• Yoon H.C.
• Han J.

Electrochemical synthesis of ammonia from water and nitrogen: a lithium-mediated approach using lithium-ion conducting glass ceramics.

• Crossref
• PubMed
• Scopus (43)
• Google Scholar

Lithium metal is unique in that it can undergo a bulk reaction in which it spontaneously splits the nitrogen triple bond at ambient conditions.

In lithium-mediated nitrogen reduction, lithium ions are reduced to lithium metal, which spontaneously reacts with nitrogen to form lithium nitride (Li₃N). The Li₃N then reacts with a proton source to form ammonia. High FEs in lithium-mediated nitrogen reduction were often achieved by running the process in a batchwise fashion,

¹⁶

• McEnaney J.M.
• Singh A.R.
• Schwalbe J.A.
• Kibsgaard J.
• Lin J.C.
• Cargnello M.
• Jaramillo T.F.
• Nørskov J.K.

Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

• Crossref
• Google Scholar

^,

^,

¹⁸

• Kim K.
• Lee S.J.
• Kim D.-Y.
• Yoo C.-Y.
• Choi J.W.
• Kim J.-N.
• Woo Y.
• Yoon H.C.
• Han J.

Electrochemical synthesis of ammonia from water and nitrogen: a lithium-mediated approach using lithium-ion conducting glass ceramics.

• Crossref
• PubMed
• Scopus (43)
• Google Scholar

at elevated pressures,

or at elevated temperatures at certain steps.

¹⁶

• McEnaney J.M.
• Singh A.R.
• Schwalbe J.A.
• Kibsgaard J.
• Lin J.C.
• Cargnello M.
• Jaramillo T.F.
• Nørskov J.K.

Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

• Crossref
• Google Scholar

^,

¹⁸

• Kim K.
• Lee S.J.
• Kim D.-Y.
• Yoo C.-Y.
• Choi J.W.
• Kim J.-N.
• Woo Y.
• Yoon H.C.
• Han J.

Electrochemical synthesis of ammonia from water and nitrogen: a lithium-mediated approach using lithium-ion conducting glass ceramics.

• Crossref
• PubMed
• Scopus (43)
• Google Scholar

It would be attractive to produce ammonia continuously at ambient temperatures and pressures. In this vein, a continuous method has been reported in which lithium metal is plated onto metal electrodes from a 0.2 M lithium perchlorate, 1% ethanol in tetrahydrofuran (THF) solution in the presence of nitrogen to produce ammonia.

However, mechanistic understanding of this process is lacking, which makes it difficult to rationally improve the process yields.

Herein, we report a mechanistic study of continuous lithium-mediated ammonia production at room temperature and pressure. We determined the rate laws of the two main rate-limiting reactions in the system: the reaction between lithium and ethanol to form hydrogen (lithium protolysis) and the reaction between lithium and nitrogen to form Li₃N (lithium nitridation), from which ammonia is evolved. These rate laws were integrated into a coupled kinetic-transport model that incorporated the competing reactions in a simple reaction network. The model is able to explain experimental results with great fidelity. The model also suggests that transport limitations can be observed in the system for nitrogen reduction, demonstrating the fast intrinsic kinetics of the lithium-mediated approach. Through a series of improvements in experimental setup, including utilization of a separator, drying of solvents, purification of reagents, and galvanostatic operation, and through mechanistic understanding of the process, we obtained high ammonia FEs and rates at ambient conditions. The maximum FE and production rate obtained were 18.5% ± 2.9% and (7.9 ± 1.6) × 10⁻⁹ mol cm⁻² s⁻¹, respectively.

Results

Electrochemical Approach

The lithium-mediated nitrogen reduction cycle was run continuously in a 2-compartment electrochemical cell (Figures 1, S1, and S2). In this setup, lithium metal is plated onto an inert metal substrate from solution in a polar, aprotic solvent. Nitrogen gas dissolved in the electrolyte reacts with the plated lithium to form Li₃N. The nitride reacts with the proton carrier, here ethanol, to form ammonia that remains dissolved in the electrolyte.

Control Experiments

It is imperative to verify that detected ammonia is a result of N₂ reduction and does not originate from either adventitious sources or contaminants.

^,

²²

• Greenlee L.F.
• Renner J.N.
• Foster S.L.

The use of controls for consistent and accurate measurements of electrocatalytic ammonia synthesis from dinitrogen.

• Crossref
• Scopus (168)
• Google Scholar

We performed a series of control experiments in order to confirm N₂ reduction (Figure 2). The concentration of ammonia in the cell was found to increase with time when applying a constant current, as determined by taking periodic aliquots from a single operating cell (Figure 2A), which supports the assertion that ammonia is generated during electrolysis and is not present as a contaminant before the experiment. As HNO₃ is used in preparing the cathode (see Supplemental Experimental Procedures), we sought to rule out nitrate reduction as the source of ammonia. When HCl-etched and unetched copper foils were used as the cathode, ammonia was still detected, confirming that contamination from HNO₃ etching is not responsible for ammonia formation (Figure S6). When argon was used as the feed gas, little ammonia was detected (Figure 2B); the ammonia that was detected in this argon control was likely produced because of residual nitrogen in the cell (see Supplemental Experimental Procedures). This confirms that the presence of dinitrogen is required for ammonia generation.

Figure 2 Control Experiments Confirming Ammonia Generation via Nitrogen Reduction

Show full caption

(A) Evolution of ammonia concentration over time in a cell containing 0.2 M EtOH at an applied current density of 3 mA cm⁻².

(B) Concentrations of ammonia in the electrolyte solutions following experiments in which ¹⁵N₂, ¹⁴N₂, and Ar were used as feed gases as measured by the colorimetric assay. A 15 mA cm⁻² current was applied for 15 min in experiments with current; no current was applied for the same amount of time in the other experiments. Error bars depict the standard deviation in the quantification assay based on n = 2.

(C) NMR spectra of the solutions from aforementioned experiments. The ¹⁵N₂ plot has been scaled down relative to the other spectra for readability. The different proton-splitting patterns in spectra resulting from ¹⁴N₂ and ¹⁵N₂ gas feeds support the claim that ammonia comes from reduction of the feed gas.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

Isotopic labeling of nitrogen allowed us to directly trace conversion of dinitrogen to ammonia.

^,

²³

• Zhang L.
• Ji X.
• Ren X.
• Luo Y.
• Shi X.
• Asiri A.M.
• et al.

Efficient electrochemical N₂ reduction to NH₃ on MoN nanosheets array under ambient conditions.

• Crossref
• Scopus (165)
• Google Scholar

^,

²⁴

• Zhang R.
• Zhang Y.
• Ren X.
• Cui G.
• Asiri A.M.
• Zheng B.
• Sun X.

High-efficiency electrosynthesis of ammonia with high selectivity under ambient conditions enabled by VN nanosheet array.

• Crossref
• Scopus (131)
• Google Scholar

^,

^,

²⁶

• Zhou F.
• Azofra L.M.
• Ali M.
• Kar M.
• Simonov A.N.
• McDonnell-Worth C.
• Sun C.
• Zhang X.
• MacFarlane D.R.

Electro-synthesis of ammonia from nitrogen at ambient temperature and pressure in ionic liquids.

• Crossref
• Google Scholar

When isotopically labeled nitrogen gas, ¹⁵N₂, was used as the feed gas, ammonia was first detected via the colorimetric assay (Figures 2B and S3–S5), then analyzed by NMR. When ¹⁴N₂ was used as the feed gas, a triplet with equal peak intensities was observed at 7.12 ppm with J = 52.2 Hz, corresponding to ¹⁴N-¹H coupling in NH₄ ⁺ (Figure 2C).

²⁷

• Berger S.
• Braun S.
• Kalinowski H.-O.

NMR Spectroscopy of the Non-Metallic Elements.

• Google Scholar

The NMR spectrum of the electrolyte solution obtained following the electrolysis of ¹⁵N₂ contains a doublet at 7.12 ppm with J = 73.2 Hz, corresponding to ¹⁵N-¹H coupling in NH₄ ⁺ (Figure 2C),

²⁷

• Berger S.
• Braun S.
• Kalinowski H.-O.

NMR Spectroscopy of the Non-Metallic Elements.

• Google Scholar

which confirms that ammonia was made from the feed ¹⁵N₂. As ammonia contamination of isotopically labeled nitrogen feedstocks is possible,

²⁸

• Dabundo R.
• Lehmann M.F.
• Treibergs L.
• Tobias C.R.
• Altabet M.A.
• Moisander P.H.
• Granger J.

The contamination of commercial ¹⁵N₂ gas stocks with ¹⁵N–labeled nitrate and ammonium and consequences for nitrogen fixation measurements.

• Crossref
• PubMed
• Scopus (167)
• Google Scholar

^,

we performed additional control experiments in which both ¹⁵N₂ and ¹⁴N₂ were fed to the cell in the absence of polarization. No ammonia was detected in either case by NMR or colorimetric assay, confirming the purity of the gas feeds, and further validating that ammonia was formed via nitrogen reduction.

Effect of Nitrogen Concentration on Nitrogen Reduction

As nitrogen is the species that is being reduced, we sought to understand the impact of nitrogen partial pressure on the observed FE toward ammonia. The partial pressure of nitrogen was varied between 0 and 1 bar by diluting the nitrogen stream with argon; the concentration of ethanol in the system was kept constant at 0.2 M and a constant current of 3 mA cm⁻² was applied to produce ammonia. The FE toward ammonia increased linearly with nitrogen pressure (Figure 3). As nitrogen concentration in solution is proportional to the partial pressure of nitrogen because of Henry's law, we can conclude that the reaction between lithium and nitrogen is first order in nitrogen. The partial current density toward ammonia in these experiments is significantly lower than the transport-limited current density (see Supplemental Information), which implies that the first-order behavior is a manifestation of the kinetics of the process not the transport. The fact that ammonia yields decrease at lower partial pressures of nitrogen helps to further validate that ammonia comes from nitrogen reduction and not from adventitious sources.

Figure 3 Effect of Nitrogen Partial Pressure on Ammonia Production

Show full caption

Variation in (A) Faradaic efficiency and (B) production rate toward ammonia with varying partial pressures of N₂. The partial pressure was controlled by diluting N₂ with Ar. When no N₂ was present, no ammonia was detected. Error bars represent the standard deviation between replicates of the same experiment with n = 3.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

Effect of Proton Carrier Concentration

We elucidated the effect of the proton carrier, ethanol, by varying its concentration while holding the nitrogen pressure and current density constant at 1 bar and 3 mA cm⁻², respectively. No ammonia or hydrogen was generated in the absence of ethanol, and a significant layer of lithium accumulated on the surface of the cathode (Figure S7); this implies that the applied current goes toward plating lithium, which remains unreactive in the absence of ethanol. The ammonia FE initially increases with increasing ethanol concentration and reaches a peak value at 0.1 M, with an average FE toward ammonia of 7.8% ± 3.0% (Figure 4A; Table S3). At ethanol concentrations > 0.1 M, the FE for ammonia decreases monotonically with ethanol concentration; the main side product under these conditions is hydrogen gas (Figure S15). The rate of ammonia generation exhibits an approximately −1.5 order with respect to ethanol concentration in this regime (Figure 4B). We believe this effect can be explained by a decrease in the amount of lithium available for nitridation at higher ethanol concentrations due to increased competition from direct lithium protolysis by ethanol; this is reflected in the lower coverage of lithium at higher ethanol concentrations (Figures S7A–S7C).

Figure 4 Effect of Ethanol Concentration on Ammonia Production

Show full caption

The effect of varying ethanol concentration on the (A) Faradaic efficiency and (B) production rate of ammonia. There is an optimal ethanol concentration at which ammonia yields are maximized. At higher concentrations, ethanol competes with nitrogen to react with lithium, which leads to a negative order with respect to ethanol, as seen in (B). Error bars represent the standard deviation between replicates of the same experiment with n = 3.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

It is important to address the economic and practical aspects of using ethanol as a proton carrier for ammonia synthesis. We define the proton carrier to be the species that protonates lithium and Li₃N in the reactions at the cathode, while the hydrogen source is the species that loses hydrogen atoms in the overall reaction at both electrodes. In this study, ethanol is used as a model proton carrier that can lead to high ammonia yields while forming lithium ethoxide; it is also the hydrogen source. While ethanol is a renewable hydrogen source, as it is produced from biomass, it is not economically viable to use ethanol as the hydrogen source for ammonia synthesis at industrial scales. It would be desirable to use water instead of ethanol as the hydrogen source because of its lower cost. Adding water to the catholyte was found to negatively affect ammonia yields, which may hinder its use as a direct proton source and carrier (Figure S9). In future work, one way to utilize water as the hydrogen source could be by oxidizing it at the anode to produce oxygen gas and protons; the protons could react with the free proton carrier and be transported to the cathode (Figure S17). Alternatively, hydrogen gas sourced from water splitting could be oxidized at the anode. In both of these setups, ethanol, or any other chosen carrier molecule, would only act as the proton carrier to the cathode and would not be consumed.

Effect of Current Density

In order to study the effect of the rate of lithium plating on the production of ammonia, the applied current density was varied between 0.3 mA cm⁻² and 25 mA cm⁻². The concentration of ethanol in these experiments was held constant at 0.2 M, a concentration higher than the optimal one. This value was chosen in order to maintain a low surface concentration of the plated lithium, which is useful for mechanistic analysis (see Discussion) as well as to obtain results with less variation in FE between runs since the standard deviation in FE is smaller at ethanol concentrations higher than the optimal one. The FE and rate of ammonia production change significantly with applied current density. The rate for ammonia production increases monotonically with increasing current density (Figure 5C). This implies that the nitride formation rate increases with current density; this is consistent with previously reported observations of faster nitrogen fixation in lithium-ion batteries at increased plating current densities.

The FE toward ammonia, however, increases with current density only at low current densities (Figure 5A). In this region, the FE increases proportionately to the square root of current density. At approximately 15 mA cm⁻², the FE reaches a maximum value with an average FE of 8.6% ± 2.2%. At higher current densities, the FE toward ammonia begins to decrease. We believe that the peak in FE is caused by nitrogen transport limitations, as discussed below.

Figure 5 Effect of Applied Current Density on Ammonia Production

Show full caption

The effect of current density on (A and B) Faradaic efficiency (FE) and (C and D) production rate toward ammonia. (B) and (D) show that the order of FE and production rate with respect to current density is approximately 1/2 and 3/2, respectively. Thus, the FE initially increases proportionally to the square root of the current density. At approximately 15 mA cm⁻², the FE peaks and begins to decrease. The production rate monotonically increases with current density but begins to stabilize above 15 mA cm⁻². Error bars represent the standard deviation between replicates of the same experiment with n = 3.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

Discussion

The existence of an optimal ethanol concentration at which ammonia production is maximized reveals many mechanistic details of ammonia synthesis (Figure 4). The negative order with respect to ethanol at higher ethanol concentrations can be explained by competition between nitrogen and ethanol for reacting with lithium metal and will be discussed in more detail below through our mechanistic model. The reasons for the positive order in ethanol at low concentrations are less obvious, however. One potential explanation could be that at low concentrations of ethanol there is insufficient ethanol to protonate Li₃N that forms when nitrogen reacts with lithium, which leads to accumulation of Li₃N. The accumulation of material on the cathode surface (Figure S7A) and low hydrogen FEs (Figure S15) at low ethanol concentrations support this hypothesis. However, after dissolving the residual solids in 0.1 M HCl following electrolysis, insignificant amounts of ammonia were detected in the resulting solution (Figure S12). If Li₃N were present, rapid protolysis by HCl would produce large amounts of ammonia. This suggests that the solids do not contain Li₃N; instead, they most likely contain unreacted lithium.

Alternatively, at low ethanol concentrations, increasing the ethanol concentration may assist nitride formation. This could occur if higher ethanol concentrations disrupt the solid-electrolyte interphase (SEI) on the lithium.

This would expose more fresh lithium for reactions. Another mechanism by which ethanol could assist nitride formation is by causing defects in the structure of lithium. A similar effect has been observed in solid-state lithium nitridation, wherein the reaction is accelerated in the presence of small amounts of water vapor.

In this case, water vapor acts as a proton source and creates defects at which nitrogen can react.

Regardless of the mechanism, increases in ethanol concentration promote reactions between lithium and both ethanol and nitrogen, leading to significantly higher ammonia and hydrogen FEs (Figures 4 and S15) than those observed in the absence of ethanol.

At high ethanol concentrations, a great deal of mechanistic insight can be extracted from the order of the ammonia production rate with respect to nitrogen and ethanol, as well as from the dependence of FE on current density. In particular, the observed orders with respect to current density and ethanol are non-integer, which implies interactions among elementary reactions. In order to explain the observed phenomena, we developed a model that describes the reactions of plated lithium with nitrogen or ethanol to form ammonia or hydrogen, respectively. We assume that there is a small amount of lithium on the surface that is constant in time but can change with operating conditions. This amount is assumed to be significantly lower than the equilibrium amount of lithium defined by the applied potential; the amount is instead defined by the competing lithium formation and consumption reactions after an initial plating time and can be found by invoking a quasi-steady-state approximation (QSSA) for lithium.

The amount of lithium per geometric surface area, which is similar to a surface concentration, is denoted by

$\left[\mathit{Li}\right]$

. The assumption of low surface concentration is justified by the associated assumption of high ethanol concentration (Figure S7). The reactions with lithium have rates proportional to the lithium concentration to some power:

$\begin{array}{c}{\mathit{r}}_{{\mathit{H}}_{\mathit{2}}}\mathit{=}{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}\end{array}$

(Equation 1)

$\begin{array}{c}{\mathit{r}}_{\mathit{N}{\mathit{H}}_{\mathrm{3}}}\mathit{=}{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]\end{array}\text{.}$

(Equation 2)

Here, the rate constants

${\mathit{k}}_1,{\mathit{k}}_2$

are defined to result in rates in units of current density, i.e., they contain Faraday's constant F and any proportionality constants derived from stoichiometry. The rate constants are assumed to be independent of applied current or potential as they describe thermochemical reactions. [N ₂] and [EtOH] are the concentrations of nitrogen and ethanol at the electrode surface, respectively. We assume the nitridation reaction is first order in nitrogen from the measured nitrogen partial pressure data (Figure 3). While this reaction formally generates Li₃N, we assume that the Li₃N protolysis reaction is fast, such that the reaction between lithium and nitrogen is rate limiting for ammonia production. This assumption is supported by a lack of Li₃N observed in the solids on the cathode after electrolysis (Figures S7 and S12). In addition, to simplify the analysis, we assume that the rate of the ammonia-forming reaction is significantly lower than the rate of the hydrogen-forming reaction:

$\begin{array}{c}{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]\mathit{\ll }{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}\end{array}\text{.}$

(Equation 3)

This assumption is valid as the partial current toward ammonia is approximately an order of magnitude lower than the partial current toward hydrogen at ethanol concentrations above the optimal one (Figures 4 and S15).

As the experiments were performed galvanostatically, we introduce an applied current density I into the model. We have experimentally observed that almost all the current applied goes toward lithium plating; only a small fraction of the current, typically ∼1%, goes toward other reactions (Figures S13 and S14), likely EtOH or water reduction. For simplicity, we assume all the current goes toward lithium plating in the model. The lithium can then react only via one of the two aforementioned reactions (Scheme 1); we assume that no other side reactions, such as SEI formation, occur. The rate of change of lithium concentration is given by:

$\begin{array}{c}\begin{array}{c}\mathit{F}\frac{\mathit{d}\left[\mathit{Li}\right]}{\mathit{dt}}=\mathit{I}-{\mathit{k}}_1{\left[\mathit{Li}\right]}^{\alpha }{\left[\mathit{EtOH}\right]}^{\mathit{x}}-{\mathit{k}}_2{\left[\mathit{Li}\right]}^{\beta }\left[{\mathit{N}}_2\right]\text{.}\end{array}\end{array}$

(Equation 4)

Scheme 1 A Visual Representation of the Reactions Assumed to Occur in the Model

Show full caption

Lithium metal is plated from lithium ions in solution by the application of a current, yielding solid lithium. The lithium metal can react with the proton carrier (HA) to form hydrogen or with nitrogen to form lithium nitride. Lithium nitride is protonated quickly to form ammonia and does not contribute to measured kinetics; this reaction is kinetically facile and is not probed by our kinetic analyses.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

From the QSSA assumption made for lithium and the relative rates of ammonia and hydrogen evolution (Equation 3), the lithium concentration can be solved for in terms of applied current density:

$\begin{array}{c}\mathit{I}-{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}-{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]\approx \mathit{I}-{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}\approx 0\end{array}$

(Equation 5)

$\begin{array}{c}\left[\mathit{Li}\right]\mathit{=}{\left(\frac{1}{{\mathit{k}}_1{\left[\mathit{EtOH}\right]}^{\mathit{x}}}\right)}^{\frac{1}{\alpha }}{\mathit{I}}^{\frac{1}{\alpha }}\text{.}\end{array}$

(Equation 6)

The FE for ammonia can be found to be:

$\begin{array}{c}\mathit{F}{\mathit{E}}_{\mathit{N}{\mathit{H}}_{\mathrm{3}}}=\frac{{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]}{\mathit{I}}\approx \frac{{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]}{{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}}=\left(\frac{{\mathit{k}}_{\mathrm{2}}}{{\mathit{k}}_{\mathrm{1}}^{\mathrm{1}\mathit{+}\frac{\mathit{\beta }\mathrm{-}\mathit{\alpha }}{\mathit{\alpha }}}}\right)\frac{\left[{\mathit{N}}_{\mathrm{2}}\right]}{{\left[\mathit{EtOH}\right]}^{\mathit{x}\left(\mathrm{1}\mathit{+}\frac{\mathit{\beta }\mathrm{-}\mathit{\alpha }}{\mathit{\alpha }}\right)}}{\mathit{I}}^{\frac{\mathit{\beta }\mathrm{-}\mathit{\alpha }}{\mathit{\alpha }}}\text{.}\end{array}$

(Equation 7)

According to this model, named the kinetic-only model, the ammonia FE should change with current density as

${\mathit{I}}^{\frac{\beta -\alpha }{\alpha }}$

. Experimental data (Figure 5) suggest that

$\frac{\beta -\alpha }{\alpha }=\frac{1}{2}$

. One possible interpretation of this result is that

$\alpha =2$

and

$\beta =3$

, which corresponds to the number of lithium atoms required to produce one molecule of H₂ or NH₃, respectively.

The model predicts that the ammonia FE should decrease with increasing ethanol concentration, with an order equal to

$-\mathit{x}\left(1+\frac{\beta -\alpha }{\alpha }\right)=-\mathit{x}\left(1+\frac{1}{2}\right)$

. As the observed order for ethanol from experimental data is −1.5, we can deduce that

$\mathit{x}=1$

, from which we infer that the reaction between lithium and ethanol is first order in ethanol. The observed non-integer order for ammonia FE as a function of ethanol concentration is thus explained by a combination of increased competition from ethanol and decreased availability of lithium to form ammonia.

The FE toward ammonia exhibits a peak as applied current density is varied, which is unexplained by the kinetic model. One explanation for the FE peaking behavior could be a change in the morphology in the plated lithium at higher current densities and thus a change in its reactivity. Lithium morphology has been observed to change at higher current densities, often turning more dendritic.

³⁴

• An S.J.
• Li J.
• Daniel C.
• Mohanty D.
• Nagpure S.
• Wood D.L.

The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling.

• Crossref
• Scopus (917)
• Google Scholar

However, if changes in morphology led to changes in reactivity, the production rate of ammonia could be expected to decrease alongside the FE, which is not observed (Figure 5C).

Alternatively, the lithium concentration could be reaching a maximum possible value at high current densities; this may be reflected in the analytical model as a constant, maximum value of

$\left[\mathit{Li}\right]$

. One of the assumptions made in the model is that the rates of reactions are proportional to the amount of lithium on the surface. This assumption breaks down when lithium is plated rapidly enough to begin plating on top of itself. This could result in a fraction of lithium that cannot contribute to ammonia or hydrogen formation. Visually, there is not a clear difference in the amount of lithium plated at different current densities after electrolysis (Figure S7). However, this does not necessarily disprove this hypothesis, as the difference between thinner and thicker lithium layers may be difficult to observe by visual inspection. In order to reject this hypothesis, we performed an experiment where the concentration of ethanol was decreased to 0.1 M while maintaining the applied current density of 15 mA cm⁻², thus increasing the amount of lithium present (Equation 6). If the lithium concentration was reaching a finite maximum value, then the FE for ammonia would be unaffected. This was not observed; the FE increased to 15.2% ± 3.1% (Figure 6), which is higher than the previously observed value of 8.6% ± 2.2%. Thus, a maximum lithium concentration is not the cause of a peak in ammonia FE.

Figure 6 A Comparison between Proposed Kinetic Models

Show full caption

The models seek to explain the appearance of an optimum current density to maximize FE. The expressions for the finite lithium concentration model described in the text are given in the Supplemental Information. All models use the 0.2 M EtOH, 15 mA cm⁻² point to fit parameters. Only the coupled kinetic-transport model matches experimental results well under all conditions. Error bars represent the standard deviation between replicates of the same experiment with n = 3.

• View Large Image
• Figure Viewer
• Download Hi-res image
• Download (PPT)

Another way to explain the observed optimal current density is that the peak is caused by a local decrease of dissolved nitrogen concentration at the electrode surface due to the limited rate of diffusion of nitrogen. In this case, the partial current density for ammonia would be approaching the mass transfer limited current density.

³⁵

• Bard A.J.
• Faulkner L.R.

Electrochemical Methods. Fundamentals and Applications.

• Google Scholar

While the limiting current (see Supplemental Information) is significantly higher than the peak partial current toward ammonia (11 ± 8 mA cm⁻² versus 1.3 ± 0.3 mA cm⁻²), transport limitations can still be used to explain the peak behavior, as shown by modifying the model below. The rate of diffusion of nitrogen in units of current density to the electrode surface is given by:

³⁵

• Bard A.J.
• Faulkner L.R.

Electrochemical Methods. Fundamentals and Applications.

• Google Scholar

$\begin{array}{c}\begin{array}{c}{\mathit{N}}_{{\mathit{N}}_2}=\frac{6\mathit{FD}}{\delta }\left({\mathit{C}}_{\mathit{bulk}}-\left[{\mathit{N}}_2\right]\right)\text{.}\end{array}\end{array}$

(Equation 8)

Nitrogen diffusing to the surface will react with lithium according to Equation 2. Over time, the system will reach a steady state, in which the concentration profile of nitrogen in the transport boundary layer does not change with time. At this point, the reaction rate of the nitrogen will be equal to its diffusion rate:

$\begin{array}{c}\frac{6\mathit{FD}}{\delta }\left({\mathit{C}}_{\mathit{bulk}}-\left[{\mathit{N}}_2\right]\right)={\mathit{k}}_2{\left[\mathit{Li}\right]}^{\beta }\left[{\mathit{N}}_2\right]\text{.}\end{array}$

(Equation 9)

Solving for the concentration of nitrogen at the surface, we obtain:

$\begin{array}{c}\left[{\mathit{N}}_2\right]=\frac{{\mathit{C}}_{\mathit{bulk}}}{\mathrm{1}\mathit{+}\frac{\mathit{\delta }}{\mathit{6FD}}{\mathit{k}}_2{\left[\mathit{Li}\right]}^{\mathit{\beta }}}\end{array}\text{.}$

(Equation 10)

At this point, we can derive an expression for the FE toward ammonia. We once again assume that the partial current for hydrogen is significantly higher than the partial current for ammonia (Equation 3). The same conclusions are obtained regardless of whether this assumption is made; we make it in order to simplify the resulting expressions. The FE toward ammonia is given by:

$\begin{array}{c}\begin{array}{c}\mathit{F}{\mathit{E}}_{\mathit{N}{\mathit{H}}_{\mathrm{3}}}\mathit{=}\frac{{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]}{{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\left[{\mathit{N}}_{\mathrm{2}}\right]\mathit{+}{\mathit{k}}_{\mathrm{1}}{\left[\mathit{Li}\right]}^{\mathit{\alpha }}{\left[\mathit{EtOH}\right]}^{\mathit{x}}}\approx \frac{{\left[\mathit{Li}\right]}^{\mathit{\beta }\mathrm{-}\mathit{\alpha }}\left[{\mathit{N}}_{\mathrm{2}}\right]}{\left(\frac{{\mathit{k}}_{\mathrm{1}}}{{\mathit{k}}_{\mathrm{2}}}{\left[\mathit{EtOH}\right]}^{\mathit{x}}\right)}\mathit{=}\frac{{\mathit{C}}_{\mathit{bulk}}{\left[\mathit{Li}\right]}^{\mathit{\beta }\mathrm{-}\mathit{\alpha }}}{\left(\frac{{\mathit{k}}_{\mathrm{1}}}{{\mathit{k}}_{\mathrm{2}}}{\left[\mathit{EtOH}\right]}^{\mathit{x}}\right)\left(\mathrm{1}\mathit{+}\frac{\mathit{\delta }}{\mathrm{6}\mathit{FD}}{\mathit{k}}_{\mathrm{2}}{\left[\mathit{Li}\right]}^{\mathit{\beta }}\right)}\end{array}\end{array}\text{.}$

(Equation 11)

Substituting the expression for lithium concentration (Equation 6), we obtain:

$\begin{array}{c}F{E}_{N{H}_{\mathrm{3}}}=\frac{C_{\mathit{bulk}}{I}^{\frac{\beta \mathrm{-}\alpha }{\alpha }}}{\left(\frac{k_{\mathrm{1}}^{\frac{\beta }{\alpha }}}{k_{\mathrm{2}}}{\left[\mathit{EtOH}\right]}^{\frac{\beta }{\alpha }x}\right)\left(\mathrm{1}+\frac{\delta k_2}{\mathrm{6}\mathit{FD}{k}_{\mathrm{1}}^{\frac{\beta }{\alpha }}}\frac{I^{\frac{\beta }{\alpha }}}{{\left[\mathit{EtOH}\right]}^{\frac{\beta }{\alpha }x}}\right)}\end{array}\text{.}$

(Equation 12)

This model is named the kinetic-transport model. It can be shown that when ammonia FE is maximized by varying the applied current density, the partial current density toward ammonia is one-third of the transport limited current density (see Figure S8; Supplemental Information):

$\begin{array}{c}{\mathit{I}}_{\mathit{peak}}=\frac{{\mathit{I}}_{\mathit{lim}}}{3}\end{array}\text{.}$

(Equation 13)

From this result, we can conclude that the FE toward ammonia can exhibit a peak with respect to applied current density at any fixed ethanol concentration due to transport limitations with a partial current that is only one-third of the predicted transport-limited current. The partial current toward ammonia at peak FE is 1.3 ± 0.3 mA cm⁻² (Figure 5), which implies that the limiting current density in the system is 3.9 ± 0.9 mA cm⁻². This value lies within the estimated range of possible limiting current densities (11 ± 8 mA cm⁻²; see Supplemental Information); hence, it is possible that some transport effects are observed at higher current densities.

The above model predicts that decreasing the ethanol concentration from 0.2 M to 0.1 M while applying 15 mA cm⁻² increases both the production rate and FE toward ammonia, which was experimentally observed (Figure 6). According to the model, if transport limitations are present, the obtained value of FE is not optimal as the partial current density toward ammonia is above one-third of the transport-limited current, which was measured at 0.2 M ethanol (2.3 ± 0.5 mA cm⁻² versus 1.3 ± 0.3 mA cm⁻²). Therefore, decreasing the applied current density should increase the FE. This is, in fact, observed: at 8 mA cm⁻² and 0.1 M, the FE increases to 18.5% ± 2.9% (Figure 6), which further supports the hypothesis that nitrogen transport limitations are observed in this system. As lowering the concentration of ethanol further would decrease ammonia production (Figure 4), this FE is close to optimal in this system. Under these conditions, ammonia is produced at an energy efficiency of 1.45%, with most of the energy lost to resistive dissipation (Figure S16). While we did not explicitly optimize for energy efficiency, it is an important metric to consider when designing electrochemical nitrogen reduction systems.

Conclusion

In summary, this work demonstrates that lithium-metal chemistry can be used to fix nitrogen to ammonia continuously in a non-aqueous solvent at high rates at room temperature and ambient pressure. We found that the reaction between nitrogen and lithium is promoted by a small amount of proton carrier, in this case ethanol. We measured rate laws for the reactions between lithium, ethanol, and nitrogen by varying the applied current density and concentrations of relevant species. We found that the lithium nitridation and lithium protonation reactions are first order in nitrogen and ethanol, respectively. The orders with respect to lithium differ for the two reactions, with the nitridation reaction being more sensitive to the amount of lithium present. As a result of this difference, the FE toward ammonia was found to increase with increasing applied current density. This finding could potentially allow for the development of a process that can produce ammonia at higher FEs. In the present system, the FE for ammonia reaches a maximum value when varying current density while maintaining a high production rate, likely due to nitrogen transport limitations. We hope that mechanistic understanding of this system will serve as a starting point and inspiration for the development of more efficient continuous lithium-mediated methods to produce ammonia.

Experimental Procedures

A polished copper foil served as the cathode, onto which lithium was plated electrochemically; a platinum foil served as the anode, and a polyporous polyethylene separator (Daramic 175) was used to separate the chambers of a polyether ether ketone (PEEK) electrochemical cell (Figures S1 and S2). The use of a separator prevented bulk convection of ammonia from the cathode to the anode, at which ammonia may be oxidized (Figure S11). The electrolyte solution consisted of 1 M lithium tetrafluoroborate (LiBF₄) dissolved in molecular sieve-dried THF with varying concentrations of ethanol (EtOH). A detailed description of the cell geometry, electrolyte preparation procedure, and other experimental details can be found in the Supplemental Information. The choice of solvent and proton carrier was based on previous work,

while the electrolyte and lithium source were chosen for safety and economic reasons. Many factors are essential to obtain high ammonia yields, including dryness of the solvent (Figure S9), purity of the reagents (Figure S10), and the use of a thick separator (Figure S11).

The cathode compartment of the cell was continuously purged with 10 sccm of THF-saturated N₂ gas both before polarization and during the experiment. Lithium metal was galvanostatically plated onto the copper cathode. The lithium metal reacted continuously with nitrogen dissolved in the electrolyte solution to form Li₃N, which was subsequently protonated by the ethanol in solution to form ammonia (Figure 1A). The electrolyte solution containing ammonia was diluted with water; the ammonia content of the resulting solution was quantified using the salicylate method.

¹⁶

• McEnaney J.M.
• Singh A.R.
• Schwalbe J.A.
• Kibsgaard J.
• Lin J.C.
• Cargnello M.
• Jaramillo T.F.
• Nørskov J.K.

Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

• Crossref
• Google Scholar

^,

^,

Acknowledgments

Funding for this research was provided by the Abdul Latif Jameel World Water and Food Systems Lab (J-WAFS) at MIT . This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under grant no. 1122374 . We thank Matt Wolski of Daramic for providing us with polyporous separator samples. We thank Yuzhang Li and Nathan Corbin for productive and helpful discussions.

Author Contributions

Conceptualization, N.L. and K.M.; Methodology, N.L.; Investigation, N.L.; Validation, Z.J.S.; Writing – Original Draft, N.L.; Writing – Review & Editing, N.L., K.W., and K.M.; Supervision. K.M.

Declaration of Interests

The authors declare no competing interests.

Supplemental Information

References

1. • Food and Agriculture Organization of United Nations

   World fertilizer trends and outlook to 2020.

2. • Erisman J.W.
   • Sutton M.A.
   • Galloway J.
   • Klimont Z.
   • Winiwarter W.

   How a century of ammonia synthesis changed the world.

   Nat. Geosci. 2008; 1 : 636-639
3. • Giddey S.
   • Badwal S.P.S.
   • Munnings C.
   • Dolan M.

   Ammonia as a renewable energy transportation media.

   ACS Sustain. Chem. Eng. 2017; 5 : 10231-10239

4. Nørskov, J., and Chen, J. (2016). Sustainable ammonia synthesis: exploring the scientific challenges associated with discovering alternative, sustainable processes for ammonia production. US Department of Energy, Office of Science. Round Table Report, February 18, 2016. https://science.energy.gov/∼/media/bes/pdf/reports/2016/SustainableAmmoniaReport.pdf.

5. • Kahrl F.
   • Li Y.
   • Su Y.
   • Tennigkeit T.
   • Wilkes A.
   • Xu J.

   Greenhouse gas emissions from nitrogen fertilizer use in China.

   Environ. Sci. Policy. 2010; 13 : 688-694
6. • Institute for Industrial Productivity

   Ammonia. Industrial Efficiency Technology Database.

7. • Foster S.L.
   • Bakovic S.I.P.
   • Duda R.D.
   • Maheshwari S.
   • Milton R.D.
   • Minteer S.D.
   • Janik M.J.
   • Renner J.N.
   • Greenlee L.F.

   Catalysts for nitrogen reduction to ammonia.

   Nat. Catal. 2018; 1 : 490-500
8. • Schiffer Z.J.
   • Manthiram K.

   Electrification and decarbonization of the chemical industry.

   Joule. 2017; 1 : 10-14
9. • Chen J.G.
   • Crooks R.M.
   • Seefeldt L.C.
   • Bren K.L.
   • Bullock R.M.
   • Darensbourg M.Y.
   • Holland P.L.
   • Hoffman B.
   • Janik M.J.
   • Jones A.K.
   • et al.

   Beyond fossil fuel–driven nitrogen transformations.

   Science. 2018; 360 : eaar6611
10. • Bao D.
    • Zhang Q.
    • Meng F.-L.
    • Zhong H.-X.
    • Shi M.-M.
    • Zhang Y.
    • Yan J.
    • Jiang Q.
    • Zhang X.

    Electrochemical reduction of N₂ under ambient conditions for artificial N₂ fixation and renewable energy storage using N₂/NH₃ cycle.

    Adv. Mater. 2017; 29 : 1604799
11. • Kong J.
    • Lim A.
    • Yoon C.
    • Jang J.H.
    • Ham H.C.
    • Han J.
    • Nam S.
    • Kim D.
    • Sung Y.
    • Choi J.
    • et al.

    Electrochemical synthesis of NH₃ at low temperature and atmospheric pressure using a γ-Fe₂O₃ catalyst.

    ACS Sustain. Chem. Eng. 2017; 5 : 10986-10995
12. • Yang X.
    • Nash J.
    • Anibal J.
    • Dunwell M.
    • Kattel S.
    • Stavitski E.
    • Attenkofer K.
    • Chen J.G.
    • Yan Y.
    • Xu B.

    Mechanistic insights into electrochemical nitrogen reduction reaction on vanadium nitride nanoparticles.

    J. Am. Chem. Soc. 2018; 140 : 13387-13391
13. • Furuya N.
    • Yoshiba H.

    Electroreduction of nitrogen to ammonia on gas-diffusion electrodes loaded with inorganic catalyst.

    J. Electroanal. Chem. Interfacial Electrochem. 1990; 291 : 269-272
14. • Lv C.
    • Qian Y.
    • Yan C.
    • Ding Y.
    • Liu Y.
    • Chen G.
    • Yu G.

    Defect engineering metal-free polymeric carbon nitride electrocatalyst for effective nitrogen fixation under ambient conditions.

    Angew. Chem. Int. Ed. 2018; 57 : 10246-10250
15. • Tsuneto A.
    • Kudo A.
    • Sakata T.

    Lithium-mediated electrochemical reduction of high pressure N₂ to NH₃.

    J. Electroanal. Chem. 1994; 367 : 183-188
16. • McEnaney J.M.
    • Singh A.R.
    • Schwalbe J.A.
    • Kibsgaard J.
    • Lin J.C.
    • Cargnello M.
    • Jaramillo T.F.
    • Nørskov J.K.

    Ammonia synthesis from N₂ and H₂O using a lithium cycling electrification strategy at atmospheric pressure.

    Energy Environ. Sci. 2017; 10 : 1621-1630
17. • Ma J.-L.
    • Bao D.
    • Shi M.-M.
    • Yan J.-M.
    • Zhang X.-B.

    Reversible nitrogen fixation based on a rechargeable lithium-nitrogen battery for energy storage.

    Chem. 2017; 2 : 525-532
18. • Kim K.
    • Lee S.J.
    • Kim D.-Y.
    • Yoo C.-Y.
    • Choi J.W.
    • Kim J.-N.
    • Woo Y.
    • Yoon H.C.
    • Han J.

    Electrochemical synthesis of ammonia from water and nitrogen: a lithium-mediated approach using lithium-ion conducting glass ceramics.

    ChemSusChem. 2018; 11 : 120-124
19. • Cui X.
    • Tang C.
    • Zhang Q.

    A review of electrocatalytic reduction of dinitrogen to ammonia under ambient conditions.

    Adv. Energy Mater. 2018; 8 : 1800369
20. • Kyriakou V.
    • Garagounis I.
    • Vasileiou E.
    • Vourros A.
    • Stoukides M.

    Progress in the electrochemical synthesis of ammonia.

    Catal. Today. 2017; 286 : 2-13
21. • Greenwood N.N.
    • Earnshaw A.

    Chemistry of the Elements.

    Second Edition. Elsevier, 1997
22. • Greenlee L.F.
    • Renner J.N.
    • Foster S.L.

    The use of controls for consistent and accurate measurements of electrocatalytic ammonia synthesis from dinitrogen.

    ACS Catal. 2018; 8 : 7820-7827
23. • Zhang L.
    • Ji X.
    • Ren X.
    • Luo Y.
    • Shi X.
    • Asiri A.M.
    • et al.

    Efficient electrochemical N₂ reduction to NH₃ on MoN nanosheets array under ambient conditions.

    ACS Sustain. Chem. Eng. 2018; 6 : 9550-9554
24. • Zhang R.
    • Zhang Y.
    • Ren X.
    • Cui G.
    • Asiri A.M.
    • Zheng B.
    • Sun X.

    High-efficiency electrosynthesis of ammonia with high selectivity under ambient conditions enabled by VN nanosheet array.

    ACS Sustain. Chem. Eng. 2018; 6 : 9545-9549
25. • Chen G.-F.
    • Cao X.
    • Wu S.
    • Zeng X.
    • Ding L.-X.
    • Zhu M.
    • Wang H.

    Ammonia electrosynthesis with high selectivity under ambient conditions via a Li⁺ incorporation strategy.

    J. Am. Chem. Soc. 2017; 139 : 9771-9774
26. • Zhou F.
    • Azofra L.M.
    • Ali M.
    • Kar M.
    • Simonov A.N.
    • McDonnell-Worth C.
    • Sun C.
    • Zhang X.
    • MacFarlane D.R.

    Electro-synthesis of ammonia from nitrogen at ambient temperature and pressure in ionic liquids.

    Energy Environ. Sci. 2017; 10 : 2516-2520
27. • Berger S.
    • Braun S.
    • Kalinowski H.-O.

    NMR Spectroscopy of the Non-Metallic Elements.

    First Edition. John Wiley & Sons, Inc, 1997
28. • Dabundo R.
    • Lehmann M.F.
    • Treibergs L.
    • Tobias C.R.
    • Altabet M.A.
    • Moisander P.H.
    • Granger J.

    The contamination of commercial ¹⁵N₂ gas stocks with ¹⁵N–labeled nitrate and ammonium and consequences for nitrogen fixation measurements.

    PLoS One. 2014; 9 : e110335
29. • Minteer S.D.
    • Christopher P.
    • Linic S.

    Recent developments in nitrogen reduction catalysts: A virtual issue.

    ACS Energy Lett. 2019; 4 : 163-166
30. • Wang H.
    • Zhang W.-D.
    • Deng Z.-Q.
    • Chen M.-C.

    Interaction of nitrogen with lithium in lithium ion batteries.

    Solid State Ionics. 2009; 180 : 212-215
31. • Peled E.
    • Menkin S.

    Review—SEI: past, present and future.

    J. Electrochem. Soc. 2017; 164 : A1703-A1719
32. • McFarlane E.F.
    • Tompkins F.C.

    Nitridation of lithium.

    Trans. Faraday Soc. 1962; 58 : 997
33. • Atkins P.
    • de Paula J.

    Physical Chemistry.

    Eighth Edition. Oxford University Press, 2006
34. • An S.J.
    • Li J.
    • Daniel C.
    • Mohanty D.
    • Nagpure S.
    • Wood D.L.

    The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling.

    Carbon. 2016; 105 : 52-76
35. • Bard A.J.
    • Faulkner L.R.

    Electrochemical Methods. Fundamentals and Applications.

    Second Edition. John Wiley & Sons, Inc, 2001
36. • Verdouw H.
    • Van Echteld C.J.A.
    • Dekkers E.M.J.

    Ammonia determination based on indophenol formation with sodium salicylate.

    Water Res. 1978; 12 : 399-402
37. • Yu X.
    • Han P.
    • Wei Z.
    • Huang L.
    • Gu Z.
    • Peng S.
    • Ma J.
    • Zheng G.

    Boron-doped graphene for electrocatalytic N₂ reduction.

    Joule. 2018; 2 : 1610-1622

Article Info

Publication History

Published: March 7, 2019

Accepted: February 4, 2019

Received in revised form: January 11, 2019

Received: December 13, 2018

Identification

DOI: https://doi.org/10.1016/j.joule.2019.02.003

Copyright

© 2019 Elsevier Inc.

User License

Elsevier user license |

How you can reuse

Permitted

For non-commercial purposes:

• Read, print & download
• Text & data mine
• Translate the article

Not Permitted

• Reuse portions or extracts from the article in other works
• Redistribute or republish the final article
• Sell or re-use for commercial purposes

Elsevier's open access license policy

ScienceDirect

Access this article on ScienceDirect

• View Large Image
• Download Hi-res image
• Download .PPT

Linked Article

• Kinetics of Li-Mediated N2 Electroreduction

  Joule April 9, 2019

  • In Brief

    Electrochemical ammonia production is a possible sustainable alternative to the energy-intense Haber-Bosch process. An active and selective catalyst for electrochemical nitrogen reduction is thus highly desired. Developing a more active and selective catalyst requires detailed kinetic studies; however, poor selectivity and low rates of nitrogen electroreduction on many candidate materials challenge kinetic analysis. Li-mediated nitrogen electroreduction provides sufficient nitrogen reduction reaction (NRR) rates for kinetic analysis. In this issue of Joule, Manthiram and coworkers perform a detailed examination of NRR kinetics for this system. They further use the obtained information to optimize the reaction conditions for electrochemical ammonia formation.

  • Full-Text
  • PDF
  Open Archive

Related Articles

jacobsondows1957.blogspot.com

Source: https://www.cell.com/joule/fulltext/S2542-4351(19)30052-2

Share this post