Mechanochemistry for ammonia synthesis under mild conditions

Ammonia, one of the most important synthetic feedstocks, is mainly produced by the Haber–Bosch process at 400–500 °C and above 100 bar. The process cannot be performed under ambient conditions for kinetic reasons. Here, we demonstrate that ammonia can be synthesized at 45 °C and 1 bar via a mechanochemical method using an iron-based catalyst. With this process the ammonia final concentration reached 82.5 vol%, which is higher than state-of-the-art ammonia synthesis under high temperature and pressure (25 vol%, 450 °C, 200 bar). The mechanochemically induced high defect density and violent impact on the iron catalyst were responsible for the mild synthesis conditions.

1. Weekly Group Presentation and Discussion. DaTE : 25/12/2020 fRIDAY

2. ABSTRACT : Ammonia, one of the most important synthetic feedstocks, is mainly produced by the Haber–Bosch process at 400–500 °C and above 100 bar (~99 atm ). This work demonstrates that ammonia can be synthesized at 45 °C and 1 bar via a mechanochemical method using an iron-based catalyst. With this process the ammonia final concentration reached 82.5 vol% , which is higher than state-of-the-art ammonia synthesis under high temperature and pressure ( 25 vol%, 450 °C, 200 bar ). The mechanochemically induced high defect density and violent impact on the iron catalyst were responsible for the mild synthesis conditions.

3. Why The Chemical Ammonia is Important ? Ammonia (NH 3 ) is one of the biggest feedstocks for fertilizers, explosives, plastics and other chemicals. In 2019, approximately 140 million tonnes were produced worldwide according to the US Geological Survey, making it one of the top ten chemicals produced.

4. How we are making Ammonia in 21’s Haber–Bosch process. The synthesis is conducted with molecular nitrogen (N 2 ) and hydrogen (H 2 ) gases at temperatures in the range of 400–500 °C and pressures above 100 bar. The most common catalyst of the Haber-Bosch process based on K 2 O or Al 2 O 3 -promoted iron catalyst. The mechanism that involves the heterogeneous catalyst is proposed to have the following steps… N 2(g)  → N 2(adsorbed) (1) N 2(adsorbed)  → 2N (adsorbed) (2) H 2(g)  → H 2(adsorbed) (3) H 2(adsorbed)  → 2H (adsorbed) (4) N (adsorbed)  + 3H (adsorbed)  → NH 3(adsorbed) (5) NH 3(adsorbed)  → NH 3(g) (6)

5. Why this Process is a novel one ? Recently, chemical looping method for ammonia synthesis under moderate conditions. The typical reaction media are alkali metals and their imides as well as nitrides, but the popularity feds because of high cost and high-temperature needed. An alternative approach has been proposed, based on mechanochemistry , which is a green , solvent-free and scalable method. Compared with traditional methods, the mechanochemical method has several unique properties, including metastable non-equilibrium states , high impact force and high defect densities , which allow a new opportunity for avoiding the need for harsh reaction conditions (comes with high expenses) Continued….

6. Overview of the Process. This novel Mechanochemical method for synthesizing ammonia under mild conditions, as low as 45 °C and 1 bar. Achieving a final ammonia concentration of up to 82.5 vol% , which is much higher than the state-of-the-art Haber–Bosch process ( 25 vol% at 450 °C, 200 bar ) and the electrochemical method (10–2,900 ppm, that is 0.6–170 μM in electrolyte). The Mechanochemical ball-milling process generates a series of unique properties in the iron catalyst. High defect densities, which are generated in situ by mechanical collisions, accelerate nitrogen dissociation, and the transferred energy produced from the dynamic relaxation during violent impacts facilitates the desorption of strongly adsorbed intermediates.

7. Inside the Mechanochemical Process… Ammonia is synthesized via ball milling using iron powder as the catalyst. The entire synthesis process can be divided into two steps: nitrogen dissociation and subsequent hydrogenation. In the first step, stable N 2 is adsorbed and dissociated into atomic nitrogen on the defects of the iron particles [Fe(N*)], where Fe(N*) denotes nitrogenated iron particles. The low-coordinated defects , which are created by repeated collisions during ball milling, are considered highly active for nitrogen dissociation. In the next step, N* is hydrogenated into NH x * species ( x = 1–3). With the help of additionally transferred energy, the strongly adsorbed NH x * species are subsequently detached from the iron surface.

8. Ammonia synthesis kinetics… Before the experiments, a controlled experiment was conducted to rule out nitrogen contamination by Ar gas. Experiments showed that the amount of adsorbed N 2 ( N N ) exhibited a typical volcano plot with respect to the rotation speed. The kinetic energy generates defects and provides additional energy to drive the nitrogen dissociation. By contrast, the generated mechanical heat, which is observed as the container temperature has a negative effect on the adsorption of N 2 due to high entropy at high temperature.

9. Dependence of adsorbed N 2 with rotation speed and pressure of N 2 The adsorbed N 2 reached a vertex at 400 r.p.m . However, a rotation speed of 250 r.p.m . gives the highest N* fixation yield in terms of energy consumption . The Fe powders were ball-milled in N 2 (9 bar) at a rotation speed of 500 rpm. The N N value is also linear with the natural logarithm of the N 2 pressure [that is N N ∝ ln( p )], but p has little influence on N N over the moderate pressure range. Nitrogen dissociation can occur at a pressure as low as 1 bar.

10. The ammonia yield ( N A ) and concentration depended on the initial charge pressure… The final ammonia concentration reached as high as 82.5 vol% when the initial charge pressure was 2 bar. After the reaction, the remnant pressure was only 1.4 bar, which indicates that hydrogenation can proceed at near atmospheric pressure (1 bar). Gas chromatography (GC) results showed that there was only a trace of N 2 byproduct (0.02 vol%) after hydrogenation.

11. Optimization of rotation Speed... Unlike nitrogen dissociation, the ammonia yield from hydrogenation exhibited a monotonously increasing relation with rotation speed and container temperature. The hydrogenation step is an endothermic process , whereas the dissociative nitrogen adsorption step is an exothermic process . Hydrogenation could proceed at a rotation speed as low as 350 r.p.m . when the container temperature was as low as 45 °C. Energy consumption analysis demonstrated that the rotation speed of 450 r.p.m . resulted in the highest ammonia yield per unit of energy consumed (91.8 mmol kWh −1 ).

12. Cost efficiency of the process… The N* fixation and ammonia yield rates per US dollar were calculated to be 204 and 937 millimoles per hour per US dollar. Except for yielding slightly less than the Haber–Bosch process (at 1,022 millimoles per hour per US dollar), mechanochemical method( at 937 milimoles / hr /US dollar ) is superior to the others. In this method, the bottleneck in the weighted ammonia yield rate is the N* fixation rate, which is only one-fifth of the ammonia yield rate. Thus, improving the N* fixation rate could greatly enhance the weighted ammonia yield rate.

13. Hydrogenation and Stability analysis… The hydrogenation process proceeds very quickly at the start and then becomes sluggish. A mathematical analysis indicates that the natural logarithm of hydrogenation rate decays linearly with time ln(( d N A /d t ) ∝ − λt ). The decay constant λ was calculated to be 0.22, according to the absolute value of the slope. The total N A reached 41.2 mmol for a nitrogenation time of 30 h. Stability was also checked. Except for a drop after the first cycle, the N A value remained stable during subsequent cycles. The drop after the first cycle may be related to the initial macro size of the iron particles, which can generate additional fresh defects during cracking.

14. Catalyst characterization… Since N 2 is mainly dissociatively adsorbed on the surface of the iron powder, the nitrogen content can be classified into three types: bulk, surface and average concentration. Here all types of nitrogen content used here are normalized to iron by atomic ratio. The bulk nitrogen concentration of Fe(N*) was calculated from the X-ray diffraction (XRD) pattern (3.7%)which is much higher than the theoretical thermodynamic equilibrium (2.05 × 10 −6 ). This indicates that the bulk nitrogen is in thermodynamic non-equilibrium, which is closely related to the high number of defect densities at low temperature. The surface concentration was determined by X-ray photoelectron spectroscopy (XPS) (16.0%). We can see that the surface nitrogen concentration is 18.7-fold higher than commercial Fe x N , shows ultrahigh concentration of surface nitrogen of Fe(N*). The average nitrogen concentration was obtained by measuring the hydrogenated nitrogen (9.4%).

15. Contaminatory cleansheet of the Catalyst… After hydrogenation, the Fe(N*) powder is transformed to activated iron powder again. There is no detectable nitrogen in the activated iron, only the α-Fe phase was observed using Mössbauer spectroscopy. The absence of nitrogen was further verified by the zero shift relative to standard iron powder. The result demonstrates that the adsorbed N* can be thoroughly removed.

16. Characterisation with HRTEM… HRTEM indicates that the activated iron (which is the iron powder after hydrogenation) and Fe(N*) particles have high defect densities, even showing amorphous phases. In the case of Fe(N*), the phase with good lattice fringes is composed of Fe 16 N 2 . Using Mössbauer spectroscopy, we speculate that the highly defective phase (nearly amorphous) features the short-range Fe 4 N structure.

17. How it behaves like a smart material… The grain size of activated iron (around 9 nm) is larger than that of Fe(N*) (2–4 nm). The results agree well with XRD [8.5 and 3.7 nm, respectively, according to the Scherrer equation. It was interesting to find that the crystal grain sizes were different. This means that one iron grain can be comminuted into many small Fe(N*) grains. The comminution of big iron grains generates more defects via cracking, and the small Fe(N*) grains are capable of adsorbing more N*. The iron catalyst can self-regulate automatically. The underlying switch is that N* is able to suppress the self-diffusion of iron atoms.

18. EXAFS spectroscopy was applied to get insight of Fe–N. In the EXAFS-derived radial distribution function (RDF), the Fe–Fe in Fe(N*) is expanded relative to activated iron. The result agrees well with the XRD patterns. XPS and soft X-ray absorption near edge structure (XANES) spectroscopy were used. Because the iron on the surface can be easily oxidized during transfer, and since the nitrides have an improved anti-oxidation ability, nitrogen will be intensively characterized here.

19. Fig. 3d shows, the commercial Fe x N only displays a peak at 397.15 eV (Fe 4 N). The Fe(N*) has two additional peaks at higher energy. It is known that the Fe–N shift to higher energy (398.10 eV) corresponds to a high nitrogen concentration and weak bonding. The peak at 399.30 eV is assigned to oxynitrides and/or physisorbed α-N 2 . This result indicates that the nitrogen in the Fe(N*) presents mainly in a weakly bonded state due to the repulsion effect at such high coverage.

20. Thermal Stability… The thermal stability of Fe(N*) was measured using thermal desorption spectrometry. The results demonstrate that nitrogen atoms begin to be desorbed at about 300 °C, reaching a vertex at 730 °C and are completely released at 780 °C.

21. Theoretical analysis… During ball milling, the mechanically generated nanoparticles store a large amount of excess energy in surface defects and disordered structure in a non-equilibrium form. DFT calculations showed that the iron atoms with the lower coordination number exhibited higher activity towards nitrogen dissociation, and the two adjacent iron atoms can facilitate nitrogen dissociation relative to one atom.

22. The adopted low temperature is inherently able to reduce the energy barrier of nitrogen dissociation due to its exothermic reaction nature. Unlike a classic iron-based catalyst, which needs to be activated by a reducing agent (usually H 2 ) at high temperature, the activation of the iron catalyst in our reaction can be accomplished directly at low temperature. The low-temperature condition makes it easier to preserve the high defect density, and their high surface energies accelerate mass diffusion on the surface. The low-coordinated iron atoms and the adsorbed N* can readily diffuse away. Subsequently, the surfaces are always dynamically reconstructed, due to aggregation by Ostwald ripening or coalescence and comminution by mechanical force.

23. Fast mass diffusion offers the possibility that the active sites for nitrogen dissociation and hydrogenation can be different. Although defect sites are highly active for nitrogen dissociation, they do not work for hydrogenation because the NH x ( x = 0–2) intermediates are too strongly adsorbed on the defects. Since the reduction of surface energy is a spontaneous process, the nitrogen atoms can easily move from the high-energy defect sites to the low-energy extended surface. Thus, the hydrogenation process on the extended facet (110) of iron was studied in the present work. It is known that strongly adsorbed hydrogen blocks the iron surface at low temperature, because the low-mass atom (H) is not easily affected by entropy. That is to say, the rate-determining step at low temperature is hydrogenation. This is different from ammonia synthesis at high temperature, where nitrogen dissociation is the rate-determining step. That is why elevated temperature is inevitably adopted in the traditional Haber–Bosch process, to overcome the desorption energy barrier of the hydrogen-related intermediates. Here, we rely on violent impacts to avoid the need for high temperature.

24. Conclusive Remarks… The major advantage of ball milling that is often quoted is its easy scale-up potential. In addition, the mild reaction conditions involved enable a simple manufacturing device configuration. Unlike traditional ammonia synthesis, which requires centralized reactors on a very large scale, this method allows on-site manufacturing on a flexible scale, and near the point of consumption. A decentralized infrastructure can save storage and transportation costs, as well as mitigate reactive nitrogen contamination through the risk of leakage. The energy consumption needed to produce one tonne of ammonia is 4.5 × 1012 J, which is competitive in comparison with the Haber–Bosch process at the laboratory scale. However, it is substantially higher than the state-of-the-art Haber–Bosch process at the industrial scale. The reason is that the production scale has a predominant contribution to energy consumption per tonne . Since the mechanochemical method for ammonia synthesis is still in its infancy, there should be large scope for improving these metrics by further. Enhancing the catalytic performance as well as optimizing the reactor architecture.