A direct ammonium carbonate fuel cell with an anion exchange membrane

Abstract

Direct ammonium carbonate- and ammonia-based fuel cells with an anion exchange membrane (AEM) and Pt/C catalyst have been constructed and their performance has been evaluated. Ammonium carbonate, which has never been used as a fuel, can become massively available when CO₂ is captured with ammonia as a non-recyclable, single-use adsorbent. It is a solid at ambient conditions and thus as a fuel option, has the advantages of volumetric energy density, ease of transportation, and storage safety. In this paper, the oxidative activities of the Pt/C catalyst on ammonium carbonate, along with ammonia, were evaluated by means of a half-cell test. Ammonium carbonate exhibited approximately 75% of the oxidation activity of ammonia; this reduced activity was likely attributable to the existence of the carbonate. This was also observed in a single cell test: the direct ammonium carbonate fuel cell generated approximately 50% lower maximum power density than that of an ammonia-based counterpart. Although its performance is seemingly inferior, it turns out that the energy power is comparable to pure ammonia or at least in the same order of magnitude. We showed that ammonium carbonate has enough potential as a fuel in low temperature polymer fuel cells.

---

1. Introduction

Ammonia is one of the future fuel candidates owing to its low cost (U.S. $200 per ton) and high energy density compared with hydrogen gas (11. 5 MJ L⁻¹ (aq. 30%) vs. 0.01 MJ L⁻¹). Ammonia is primarily produced from natural gases, and in the long run, from renewable sources. Accordingly, ammonia has received considerable attention in the research sector. Its potential as a hydrogen carrier was suggested¹ and various oxidation catalysts have also been reported.² Unfortunately, however, the aforementioned beneficial properties of energy density and price, which are superior to the hydrogen fuel, are not sufficient to become the ultimate fuel of choice. Additionally, ammonia is gaseous at ambient conditions and thus a large amount of energy (roughly 10 000 hPa) is required for liquefaction. These problems can be solved by way of using similar yet different ammonia-based compounds: one great candidate is ammonium carbonate. Ammonium carbonate still has a high energy density (ammonium carbonate: 11.4 MJ L⁻¹) and is a solid below 58 °C, and thus safe storage and easy transportation are possible. Despite its high potential, any information in relation to fuel cells has never been collected thus far.

Ammonium carbonate can be produced when ammonia is used as a single-use absorbent for capturing carbon dioxide^3–5

aNH₃ + bCO₂ + cH₂O → NH₄HCO₃; (NH₄)₂CO₃; NH₄COONH₂; (1)

(NH₄)₂CO₃ and NH₄HCO₃ are easily produced in the form of a white solid at low temperature and atmospheric pressure.⁶ These reactions are advantageous in that not much energy and/or noble metal catalysts are needed, because of the low kinetic barrier of the precipitation mechanism.

Ammonia has recently been spotlighted, in particular as a future fuel for a hydrogen-based society. It is regarded as a potential candidate fuel for the solid oxide fuel cell (SOFC). Using this fuel cell system, Fuerte et al. reported that ammonia exhibited approximately 75% of the performance of hydrogen.⁷ In the case of polymer electrolyte membrane fuel cells (PEFCs), employing an acidic electrolyte, such as Nafion, ammonia has been treated as a poisoning impurity.^8,9 Serious performance degradation happened with even a small amount of it (1–10 ppm) because of the decreased conductivity caused by the replacement of the H⁺ by NH₄⁺ on the ionomer in both electrode catalysts and on the membrane. In the case of alkaline fuel cells (AFCs) using KOH electrolyte, on the other hand, ammonia has been studied as a fuel since the early 1960s.¹⁰ Although ammonia can be electrochemically oxidized in an alkaline electrolyte, the formation of a carbonate salt causes a serious performance drop in conventional alkaline fuel cells.¹¹ The use of anion exchange membranes (AEMs) allows us to overcome this issue,^12,13 as the anion exchange membranes, comprising organic quaternary ammonium hydroxide branched to polymers, are free from cations like K⁺. Unfortunately, however, the current version of the fuel cell suffers from low performance (i.e., under 1 mW cm⁻²), requiring dramatic improvement through either active catalysts, new anion exchange membranes or both. De Vooys et al. reported ammonia oxidation activities of single metals related to the adsorption of N_ads.¹⁴ Electro-catalytic ammonia oxidation activities varied in the order of Ru < Rh < Pd < Ir < Pt. Pt–Ru/C,¹⁵ Pt–Ir/C,¹⁶ and Pt–Ni/C¹⁷ catalysts have been introduced as bimetallic catalysts, displaying improved activity for ammonia oxidation. Substantial further improvement is needed to show its commercialization potential.

In this study, ammonium carbonate, a potential ammonia compound that can be widely available in the future, was investigated as a fuel for the alkaline fuel cell employing an anion exchange membrane and Pt/C. Electrochemical studies based on half cells and single cells were carried out. Ammonia was also tested for comparison.

2. Experimental

Commercial Pt/C (J&M 20 wt%) was used as a catalyst for the oxidation of ammonia and ammonium carbonate. Electrochemical analysis was carried out on a three-electrode cell with an electrochemical analyzer (CH instruments 700a). 158.7 μg cm⁻² catalyst was loaded on a glassy carbon electrode (ALS, Q1073) which was used as the working electrode. Platinum wire and a Ag/AgCl electrode (BAS Co., MF2052 RE-5B) were adopted as a counter electrode and a reference electrode, respectively. 1 M H₂SO₄, 1 M KOH, 0.2 M NH₄OH + 1 M KOH, and 0.2 M (NH₄)₂CO₃ + 1 M KOH were prepared as electrolytes. Cyclic voltammetry (CV) was conducted with a scan rate of 15 mV s⁻¹ at room temperature. All electrolytes were purged by nitrogen gas prior to any measurement. All chemicals (reagent grade) were purchased from Aldrich. All potentials are quoted with respect to the Ag/AgCl scale.

Membrane electrode assemblies (MEAs) were fabricated for the single cell test. The mixture of Pt/C (20 wt%), DI water, 1-propanol, and anion exchange ionomer solution (AS-4, Tokuyama Corp.) was prepared as a catalyst ink. The prepared ink was loaded on a carbon cloth (SCCG-5N, CNL energy) using a blushing method. The loading amount of Pt on both the anode and cathode electrodes was 0.6 mg_Pt cm⁻². The properties of the anion exchange membrane (AHA, Astom. Japan) are listed in Table A1.† Ammonia solution (25 wt%) and ammonium carbonate solution (25 wt%) were fed into the anode compartment at flow rates of 20 cm³ min⁻¹. Pure oxygen was fed into the cathode at a flow rate of 50 cm³ min⁻¹. The single cells with active areas of 4 cm² were operated at 80 °C and their performance was analyzed by an electrochemical analyzer (CH instruments 700a) with a scan rate of 15 mV s⁻¹. A schematic diagram of the single cell is shown in Fig A1.† Data from the single cell were collected for 1 h after each fuel was injected.

3. Results and discussion

3.1 Electrocatalytic oxidation of ammonia and ammonium carbonate in alkaline aqueous solutions

Electrochemical analysis for the oxidation reactions of both ammonia and ammonium carbonate was conducted in alkaline solutions by means of cyclic voltammetry (CV). Fig. 1 shows CV results obtained with the Pt/C catalyst. In addition, the proton adsorption–desorption area of the Pt/C catalyst was calculated from the CV result obtained in 1 M H₂SO₄ solution, from which the electrochemically active surface area (EAS) was estimated for the catalyst (Fig. A2†). Furthermore, OH⁻ adsorption phenomena were also observed at potentials between −0.9 V and 0.6 V (Fig. 1). Ammonia was found to be oxidized at around −0.31 V, at which current densities of ammonia oxidation were 0.92 mA m⁻² in 0.2 M NH₄OH + 1 M KOH electrolyte, and 0.69 mA m⁻² in 0.2 M (NH₄)₂CO₃ + 1 M KOH electrolyte. Specific activities of both ammonia electrolytes, calculated from the EAS and current density, were found to be 337.0 mA m⁻² for the 0.2 M NH₄OH + 1 M KOH electrolyte and 252.7 mA m⁻² for the 0.2 M (NH₄)₂CO₃ + 1 M KOH electrolyte. The Pt/C catalyst exhibited up to 25% lower oxidation activity in the ammonium carbonate electrolyte than that in the ammonia electrolyte. Electrochemical results are summarized and listed in Table 1. A possible mechanism of ammonia oxidation on noble metal catalysts was proposed by Bunce and Bejan¹⁸ as follows (where ‘M’ in equation 5 denotes catalysts (noble metal, e.g. Pt):

NH₃ + HO⁻ → NH₂(ads) + H₂O + e⁻ (2)

NH₂(ads) + HO⁻ → NH(ads) + H₂O + e⁻ (3)

NH(ads) + HO⁻ → N(ads) + H₂O + e⁻ (4)

2N(ads) → 2M + N₂ (5)

Fig. 1 Cyclic voltammogram of Pt/C in 1 M KOH (black line), 1 M KOH + 0.2 M NH₄OH (red line), 1 M KOH + 0.2 M (NH₄)₂CO₃ (blue line) solutions at a scan rate of 15 mV s⁻¹ from −0.9 V to 0.7 V.

Table 1 A summary of the electrochemical analysis results

Half cell test				Single cell test
Electrolyte	Current density (mA cm^−2)	Mass activity (mA g^−1Catal)	Specific activity (mA m^−2)	Fuel	OCV (V)	Max power density (mW cm^−2)
0.2 M NH4OH	0.92	5796	337.0	NH4OH	0.36	0.22
0.2 M (NH4)2CO3	0.69	4347	252.7	(NH4)2CO3	0.32	0.11

---

The final destination of this tentative pathway is N₂ and H₂O. According to Gerischer and Mauerer's notion, N(ads) is stabilized by chemisorption, which is a cause of poisoning of the Pt surface.¹⁹ Based on their study about the relations between adsorbates (i.e., NH_ads and N_ads) and the activity of ammonia oxidation,¹⁴ De Vooys et al. claimed that the N_ads species, which are oxidation intermediates, must be easily and rapidly removed to maintain high catalytic activity. We speculate that this mechanism can be applied to ammonium carbonate electrolyte without a great deal of impairment to its highly acceptable reasoning. This can be synergically backed up by reported electrochemical and theoretical studies of carbonate adsorption on a Pt(111) surface.^20,21 One sure reason for the lower activity in ammonium carbonate can be found in the adsorption of carbonate on the catalyst surface. The theoretically calculated desorption energy of carbonate from a Pt(111) surface is 30.2 kJ mol⁻¹.²⁰ The desorption energy of NH₃ on a Pt surface, measured by temperature programmed desorption (TPD) in an ultra-high vacuum condition (UHV) is 96 kJ mol⁻¹,²² strongly suggests that adsorption of ammonia on the Pt surface is indeed interrupted by carbonate ions. Furthermore, it is also likely that carbonate acts as a N_ads-like species, as in the case of the ammonia oxidation process. Therefore to ensure acceptably high catalytic activity of our fuel of choice, ammonium carbonate, different types of catalysts, particularly with a low adsorption energy of N_ads and/or a powerful carbonate-blocking method, must be developed.

Simultaneous oxidation of ammonia and H₂O was detected above 0.5 V, where ammonia was reported to be oxidized to NO₂⁻ and NO₃⁻.²³

NH₃ + 3HO⁻ → 1/2N₂ + 3H₂O + 3e⁻ (6)

NH₃ + 7HO⁻ → NO₂⁻ + 5H₂O + 6e⁻ (7)

NH₃ + 9HO⁻ → NO₃⁻ + 6H₂O + 8e⁻ (8)

N₂H_y and NH₂OH could also be formed as a result of the simultaneous oxidation.²⁴ Pt/C catalysts exhibited a somewhat degraded activity in the ammonium carbonate solution, in comparison to that in the ammonia solution, which must also be something to do with the aforementioned carbonate effect on the Pt surface.

3.2 Single cell test of ammonia and ammonium carbonate

The single cell was operated with either ammonia or ammonium carbonate as fuels and their results are shown in Fig. 2. Open-circuit voltages (OCVs) of ammonia and ammonium carbonate were not very different: 0.36 V and 0.32 V, respectively. The theoretical OCV of the ammonia fuel cell is 1.17 V at 25 °C, in accordance with the following half reactions.

Anode: 2NH₃ + 6OH⁻ → N₂ + 6H₂O + 6e⁻, E_o = −0.77 V (9)

Cathode: O₂ + 2H₂O + 4e⁻ → 4OH⁻, E_o = + 0.40 V (10)

Overall: 4NH₃ + 3O₂ → 2N₂ + 6H₂O, E_o = + 1.17 V (11)

Fig. 2 I–V test results of the AEM fuel cells using Pt/C as the anode catalyst using (1) ammonium carbonate (red line) and (2) ammonia (blue line) as fuels.

This unreasonably low OCV in the ammonia fuel cell, which was actually only 30% of the theoretical value, was not unusual. Suzuki et al. reported 0.37 V in their ammonia fuel cell with an AEM and Pt/C catalyst.¹² One reasonable cause for this suboptimal performance can be found from Pt catalysts. Some previous studies proved that Pt is not a good catalyst towards ammonia oxidation due to the high adsorption energy of N_ads species on the catalyst surface.^14,18 Another cause is ammonia leakage through the AEM, termed fuel-crossover. This crossover becomes worsened in wet conditions.¹² Ammonia-based fuel cells, especially those using an AEM, have been barely studied and thus not much is known about intrinsic issues such as this. Hence, systematic and in-depth investigation must be conducted. The maximum power densities of ammonia and ammonium carbonate fuel cells were 0.22 mW cm⁻² and 0.11 mW cm⁻², respectively. These results clearly support ammonia as a better fuel than ammonium carbonate, which is expected and reasonable, considering the aforementioned negative effects of carbonate. When taking into consideration ammonia alone and not the entire molecule, very different and interesting results are brought about. The ammonia portion in ammonium carbonate is only 35 wt%, thus the recalculated power output of ammonium carbonate, done by normalizing the power density values, is roughly 40% greater than that of ammonia. Based on all these findings, we tend to believe that ammonium carbonate is a good fuel option, if not better than ammonia. Of course, the performance is not even close to the commercial competitors like direct methanol fuel cells and thus, it is in great need of further improvement. We are currently undertaking this path in a more systematic way.

4. Conclusions

A direct ammonium carbonate fuel cell and an ammonia fuel cell with an AEM and Pt/C catalyst have been evaluated. Ammonium carbonate shows oxidation activity 25% lower than ammonia due to the effect of carbonate. In a single cell test, OCVs of ammonia and ammonium carbonate were not very different: 0.36 V and 0.32 V, respectively. Additionally, the direct ammonium carbonate fuel cell has half the performance of the direct ammonia fuel cell, based on maximum power density. Considering the ammonia ratio between ammonia and ammonium carbonate fuels, however, ammonium carbonate fuel exhibits cell performance that is 40% higher than ammonia fuel. Ammonium carbonate can be a potential fuel option in the AEM-based fuel cell system, provided that substantial performance improvement is made.

Acknowledgements

This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government Ministry of Education, Science and Technology (NRF-2012M1A2A2026587). This work was also supported by the New & Renewable Energy of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korean government Ministry of Knowledge Economy (2012T100201665).

References

1. A. Klerke, C. H. Chirstensen, J. K. Norskov and T. Vegge, J. Mater. Chem., 2008, 18, 2304–2310.
2. F. Schuth, R. Palkovits, R. Schlogl and D. S. Su, Energy Environ. Sci., 2012, 5, 6278–6289.
3. A. C. Yeh and H. Bai, Sci. Total Environ., 1999, 228, 212.
4. G. Valenti, D. Bonalumi and E. Macchi, Energy Proc., 2009, 1, 1059–1066.
5. S. Choi, J. H. Drese, P. M. Eisenberger and C. W. Jones, Environ. Sci. Technol., 2011, 45, 2420–2427.
6. A. Bandyopadhyay, Clean Technol. Environ. Policy, 2011, 13, 269–294.
7. A. Fuerte, R. X. Valenzuela, M. J. Escudero and L. Daza, J. Power Sources, 2009, 192, 170–174.
8. F. A. Uribe, S. Gottesfeld and T. A. Zawodzinski, J. Electrochem. Soc., 2002, 149, A293–A296.
9. R. Halseid, P. J. S. Vie and R. Tunold, J. Power Sources, 2006, 154, 343–350.
10. E. J. Cairns, L. Simons and A. D. Tevebaugh, Nature, 1968, 217, 780–781.
11. G. F. Mclean, T. Niet, S. Prince-Richard and N. Dijilali, Int. J. Hydrogen Energy, 2002, 27, 507–526.
12. S. Suzuki, H. Muroyama, T. Matsui and K. Eguchi, J. Power Sources, 2012, 208, 257–262.
13. R. Lan and S. Tao, Electrochem. Solid-State Lett., 2010, 13(8), B83–B86.
14. A. C. A. de Vooys, M. T. M. Koper, R. A. van Santen and J. A. R. van Veen, J. Electroanal. Chem., 2001, 506, 127–137.
15. K. Endo, K. Nakamur, Y. Katayama and T. Miura, Electrochim. Acta, 2004, 49, 2503–2509.
16. T. L. Lomocso and E. A. Baranova, Electrochim. Acta, 2011, 56, 8551–8558.
17. K. Yao and Y. F. Cheng, J. Power Sources, 2007, 173, 96–101.
18. N. J. Bunce and D. Bejan, Electrochim. Acta, 2011, 56, 8085–8093.
19. H. Gerischer and A. Mauerer, J. Electroanal. Chem., 1970, 25, 421–433.
20. A. Markovits, M. G. Herna, J. M. Ricart and F. Illas, J. Phys. Chem. B, 1999, 103, 509–518.
21. A. Berna, A. Rodes, J. M. Feliu, F. Illas, A. Gil, A. Clotet and J. M. Ricart, J. Phys. Chem. B, 2004, 108, 17928–17939.
22. J. M. Bradley, A. Hopkinson and D. A. King, Surf. Sci., 1997, 371, 255–257.
23. F. Fichter, Electrochem. Z. Physik. Chem., 1913, 18, 647.
24. K. Endo, Y. Katayama and T. Miura, Electrochim. Acta, 2005, 50, 2181–2185.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra44057e

---