A novel approach to reduction of CO₂ into methanol by water splitting with aluminum over a copper catalyst

Abstract

A novel method of water splitting for CO₂ reduction into methanol was proposed by directly using commercially available and non-precious Al and Cu powder as a reductant and catalyst, respectively. A 22% yield of methanol was obtained from CO₂. This technique is simple and environmentally friendly, and also provides an example of a multitude of possible chemical reactions for CO₂ conversion.

---

The concentration of CO₂ in the atmosphere is increasing at an accelerating rate from 312 ppm in 1955 to 400 ppm in 2014, and the high CO₂ level is the primary source of the greenhouse effect.¹ Thus, developing a sustainable approach to CO₂ utilization as a cheap and abundant carbon source to produce organic chemicals would be an especially fascinating technology, which can lead to a carbon-neutral fuel cycle in theory.² Many promising methods for CO₂ reduction have been proposed. Among these methods, water splitting for CO₂ reduction can be considered as a promising approach because water can be used as an in situ liquid hydrogen source and then hydrogen storage, transportation and purity are not required.

Methanol is one of chemical fuels from CO₂ reduction, which has been used as a transportation fuel in either modified internal combustion engines or direct methanol fuel cells, and also as a raw material for the synthesis of important platform chemicals for methyl tert-butyl ether and chloromethane, etc.³

In recent decades, various processes have exemplified the concept that methanol can be produced from CO₂ and hydrogen base on Cu-based catalysts.⁴ However, the hydrogen sources for CO₂ hydrogenation are generally from fossil fuel, and also the hydrogen storage, transportation and purity requirement are the other technical obstacles. In addition, a complex or precious catalyst is generally needed.⁵ Hydrothermal chemistry has received much attention in organic chemical synthesis and biomass conversion due to the unique inherent properties of high temperature water, including a high ion product, a low dielectric constant, fast reaction rate and environmental benign property. We have conducted many researches involving biomass conversion under hydrothermal reactions.⁶ On the other hand, hydrothermal reactions have played an important role in the formation of fossil fuel in the earth's crust and deep-sea hydrothermal vent such as the abiotic synthesis of hydrocarbons from the dissolved CO₂.⁷ Thus, CO₂ reduction in the simulated hydrothermal vent system should be a very desirable research. With this in mind, recently, we turned our attention from biomass conversion into the much more challenging CO₂ reduction under hydrothermal conditions with metal Fe as a reductant. Fe was selected because hydrothermal fluid enriches reduced ferrous-bearing mineral, which is considered to generate reduction conditions by the reaction with hot water.⁸ Our previous results with Fe as a reductant have demonstrated the formation of formic acid from CO₂.⁹ These findings inspired us to further investigated CO₂ reduction with other metals such as Zn, Mn and Al, and it was found that the relative efficiency of CO₂ reduction into formic acid was Al > Zn > Mn > Fe.¹⁰ In these reactions, methanol formation was also confirmed when Zn was used with Cu as a catalyst, however, the yield of methanol was very low,¹¹ and therefore the reactions require further investigation for CO₂ utilization. Since Al has a better performance than Zn for CO₂ reduction into formic acid, an enhancement of methanol production from CO₂ can be expected by using Al instead of Zn. Moreover, a redox process of ZnO/Zn, FeO/Fe₃O₄ and Mn(IV)/Mn(II) with solar energy have been reported, even for MgO/Mg, its redox can be achieved by solar power concentration using laser technology.¹² The redox of MgO/Mg suggests that the reduction of Al_xO_y would be much easier than MgO when using solar energy because Mg is more active than Al. Thus, a circulation of Al would be achieved by combining solar energy.

To the best of our knowledge, the hydrothermal reduction of CO₂ into methanol with Al as a reductant was not yet reported previously. In the present article, we report the discovery of hydrothermal conversion of CO₂ to methanol with two commercially available and non-precious metal powders of Al and Cu directly working as a reductant and a catalyst, respectively. In this process, water not only acts as an excellent greener solvent, but also as a greener hydrogen sources. Further, neither elaborately prepared catalysts nor expensive reagents were used.

First, various transition-metals such as Fe, Zn, Mn and Al were tested as a reductant for the reduction of NaHCO₃, as a source of CO₂, with Cu-based catalysts. The results are summarized in Table 1. Among all metals investigated, the reaction did not give the desired methanol in the absence of Fe or Mn as a reductant (entry 1–7). Interestingly, methanol was formed when using Zn or Al as a reductant (entry 8 and 10) with Cu as a catalyst. Evidently, the methanol yield with Al (1.5%) was higher than that with Zn (0.6%), which was in accordance with their activities. Moreover, no formation of methanol was observed when using Al with CuO or using only Cu without Al (entry 11 and 12). Thus, only for the case of using both Al (as a reductant) and Cu (as a catalyst), CO₂ can be reduced into methanol. Additionally, in order to test whether Cu²⁺ can act as a catalyst in the conversion of CO₂ into methanol, a reaction with CuSO₄ instead of Cu powder was conducted. The results showed that no methanol was produced (entry 13), indicating that metallic Cu rather than Cu²⁺ acted as a catalyst in the reduction of CO₂ into methanol. Analysis for solid residues after reactions by XRD showed that Al was oxidized to AlO(OH) and most of Cu still existed in Cu(0) (see Fig. SI-1†), confirming the reductant role of Al and catalytic role of Cu in the reduction of CO₂ into methanol. Thus, the overall reaction of the reduction of CO₂ with Al in water can be explained as eqn (1):

Table 1 Effect of reductants and catalysts on the yield of methanol^a,b

Entry	Reductant	Catalyst	Yield (%)
a Reaction conditions: 350 °C, 2 h, 35% of water filling, 1 mmol NaHCO3, 6 mmol metal, 0.5 g Cu, or 40 mmol NaHCO3, 60 mmol Zn, 40 mmol Cu (entries 8 and 9), or 40 mmol NaHCO3, 40 mmol Al, 50 mmol Cu (entries 10 and 11), or 20 mmol NaHCO3, 60 mmol Al, 50 mmol Cu, 0.08 mol L^−1 CuSO4 (entries 12 and 13).b The yield of methanol is defined as the percentage of methanol to the initial NaHCO3 on a carbon basis (see eqn S1 in ESI).
1	—	—	0
2	Fe	Cu	0
3	Fe	CuO	0
4	Fe	Cu2O	0
5	Mn	Cu	0
6	Mn	Cu2O	0
7	Mn	CuO	0
8	Zn	Cu	0.6
9	Zn	—	0
10	Al	Cu	1.5
11	—	Cu	0
12	Al	CuO	0
13	Al	CuSO4	0

---

Subsequently, the liquid samples and gaseous samples after the reactions were analyzed by GC/FID, GC/MS, HPLC and GC/TCD, respectively. Results showed that the major product was methanol with a few amount of formic acid as well as trace quantity of acetic acid and methane (see Fig. S3–S5†).

Although methanol could be produced from CO₂ with Al on the Cu catalyst under hydrothermal conditions, formic acid was formed mainly as a competitive product of methanol. To enhance the selectivity of methanol, the effect of H⁺ concentration on the yield of methanol and formic acid was investigated by adding hydrochloric acid from 0 to 2.0 mol L⁻¹, because some research has reported that a weak acidic system was adverse to formic acid formation in CO₂ hydrogenation.^9b,10b As shown in Fig. 1, the yield of formic acid decreased sharply with an increase in the initial concentration of hydrochloric acid, while the yield of methanol increased greatly from 1.5% to 5.4%, suggesting that an acidic system was favorable for the production of methanol. However, when the concentration of hydrochloric acid exceeded 1.2 mol L⁻¹, the yield of methanol decreased. In order to further confirm the effect of H⁺ concentration on the production of methanol, reactions with sulfuric acid instead of hydrochloric acid with the concentration of 0.6 mol L⁻¹ (H⁺ concentration of 1.2 mol L⁻¹) was conducted. As a result, the yield of methanol was 4.8% (see the red point in Fig. 1), which was close to the one obtained with hydrochloric acid in the same H⁺ concentration of 1.2 mol L⁻¹. The fact that a low pH is favorable for the production of methanol suggests that methanol comes from CO₂ rather than HCO₃⁻ because a low pH is favorable for the formation CO₂ in the equilibrium of CO₂, HCO₃⁻ and CO₃²⁻ in water solution.

Fig. 1 Effect of hydrochloric acid concentration on the yield of methanol (temperature: 350 °C, time: 2 h, NaHCO₃: 40 mmol, Al: 40 mmol, Cu: 50 mmol, water filling: 50%).

Then, the reaction characteristics of the reduction of CO₂ into methanol with Al and Cu, and also the parameters design for getting a high yield of methanol were investigated in the presence of hydrochloric acid concentration of 1.2 mol L⁻¹. First, the effect of the initial amounts of Cu and Al was studied. As shown in Fig. 2, the methanol yield increased evidently with the increase in the initial amount of Cu, when the ratio of Cu/NaHCO₃ increased to 0.5, the yield of methanol verged to a constant. For the effect of Al amount, the yield of methanol increased clearly with the increase in the amount of Al, which is probably because a stronger reduction condition or a large amount of hydrogen can improve the CO₂ reduction. Considering the cost saving, although the yield of methanol still increased with the increase in the initial amount of Al, Al/NaHCO₃ = 1 : 1 was chosen in further investigations.

Fig. 2 Effect of the amount of Al and Cu on the methanol yield (350 °C, 2 h, 40 mmol NaHCO₃, 1.2 mol L⁻¹ HCl, 50% of water filling).

In the following experiments, the effects of some important parameters such as reaction temperature, reaction time, water filling and NaHCO₃ amount on the formation of methanol were investigated. As shown in Fig. 3, the increase in the reaction temperature, time and water filling (pressure) were favorable for the conversion of CO₂ into methanol, and also an increase in the initial amount of NaHCO₃ led to an increase in the yield of methanol. The highest yield of methanol reached to 18.5%, which occurred at 350 °C for 7 h with water filling 50% and NaHCO₃ 40 mmol. Further, the result obtained with gaseous CO₂ at 350 °C for 2 h showed that the yield of methanol was 22%, which was higher than that with NaHCO₃. These results confirmed again the methanol formation was from CO₂.

Fig. 3 Effect of reaction temperature, time, water filling and the initial amount of NaHCO₃ on methanol production (40 mmol Al, 50 mmol Cu, 1.2 mol L⁻¹ HCl; (A) 2 h, 50% of water filling, 40 mmol NaHCO₃; (B) 350 °C, 50% of water filling, 40 mmol NaHCO₃; (C) 350 °C, 2 h, 40 mmol NaHCO₃; (D) 350 °C, 2 h, 50% of water filling).

Additionally, a reaction with directly using gaseous CO₂ and H₂ with Cu catalyst was conducted. As a result, a methanol yield of 20.3% was obtained at 350 °C for 2 h, which demonstrates the suitability of using CO₂, H₂ with Cu catalyst to obtain methanol under hydrothermal conditions.

Generally, a complex catalyst is needed in the traditional hydrogenation of CO₂ into methanol. The observed highly catalytic activity of a general Cu powder under hydrothermal conditions suggests that some intermediates might act as a synergetic catalyst with Cu in the formation of methanol from CO₂. To examine this assumption, the solid residues after reactions were further analyzed by XRD. Unexpectedly, a small amount of Cu₂O was detected (Fig. SI-2(a)†). The formation of Cu₂O was probably because of the oxidation of Cu in high temperature water, which was supported by the fact that a few of Cu₂O was detected by XRD (see Fig. SI-2(b)†), and the grain diameter of Cu₂O was less than 100 nm in a further experiment with Cu directly reacted with NaHCO₃ and HCl at 350 °C. Although further evidence is needed, these results suggest that the high methanol yield with Cu might be attributed to a synergistically catalytic role of the interface between Cu and Cu₂O because Cu is a conductive material and Cu₂O is a good semiconductor material with narrow band, the interface between Cu and Cu₂O would be the active site for the deeper six-electron-reduction product methanol from CO₂. Work along these lines is now in progress.

Conclusions

In summary, we have developed an efficient method for reducing CO₂ into six-electron-reduction product methanol in water with commercially available and non-precious metals of Al and Cu, which directly worked as a reductant and a catalyst, respectively. A high methanol yield of 22% was obtained from CO₂. The proposed method is simple, green, and environmentally friendly because an environmental benign solvent of water and the hydrogen from water are used in the proposed process. Furthermore, a carbon cycle could be achieved by coupling geochemistry, specifically the metal-based water splitting for CO₂ reduction, with solar driven thermochemistry for reducing metal oxide into metal.

Acknowledgements

The authors thank the financial support of the National Natural Science Foundation of China (no. 21277091), the State Key Program of National Natural Science Foundation of China (no. 21436007), Key Basic Research Projects of Science and Technology Commission of Shanghai (no. 14JC1403100).

Notes and references

1. (a) D. M. Xiang, D. Magana and R. B. Dyer, J. Am. Chem. Soc., 2014, 136, 14004; (b) J. Hansen, M. Sato, G. Russell and P. Kharecha, Philos. Trans. R. Soc., A, 2013, 371, 20120294.
2. (a) V. S. Thoi, N. Kornienko, C. G. Margarit, P. Yang and C. J. Chang, J. Am. Chem. Soc., 2013, 135, 14413; (b) N. J. Brown, J. Weiner, K. Hellgardt, M. S. Shaffer and C. K. Williams, Chem. Commun., 2013, 49, 11074.
3. (a) G. A. Olah, Catal. Lett., 2004, 93, 1; (b) X. M. Liu, G. Q. Lu and Z. F. Yan, Ind. Eng. Chem. Res., 2003, 42, 6518; (c) E. E. Ortelli, J. Wambach and A. Wokaun, Appl. Catal., A, 2001, 216, 227; (d) G. A. Olah, A. Goeppert, M. Czaun and G. K. S. Prakash, J. Am. Chem. Soc., 2013, 135, 648.
4. (a) F. Güder, E. Frei, U. M. Kücükbayrak, A. Menzel, R. Thomann, R. Luptak and M. Zacharias, ACS Appl. Mater. Interfaces, 2014, 6, 1576; (b) L. C. Grabow and M. Mavrikakis, ACS Catal., 2011, 1, 365; (c) M. Lim, S. Kang, H. Chae, J. Bae and K. Jun, Ind. Eng. Chem. Res., 2009, 48, 10448; (d) F. Pontzen, W. Liebner, V. Gronemann, M. Rothaemel and B. Ahlers, Catal. Today, 2011, 171, 242.
5. (a) M. S. Spencer, Top. Catal., 1999, 8, 259; (b) J. F. Deng, Q. Sun, Y. L. Zhang, S. Y. Chen and D. Wu, Appl. Catal., A, 1996, 139, 75; (c) J. H. Lee, S. W. Kim, B. S. Ahn and D. J. Moon, J. Nanosci. Nanotechnol., 2015, 15, 400.
6. (a) J. F. Dubreuil and M. Poliakoff, Pure Appl. Chem., 2006, 78, 1971; (b) N. Akiya and P. E. Savage, Chem. Rev., 2002, 102, 2725; (c) N. Akiya and P. E. Savage, Ind. Eng. Chem. Res., 2001, 40, 1822.
7. (a) F. Jin, X. Zeng, Z. Jing and H. Enomoto, Ind. Eng. Chem. Res., 2012, 51, 9921; (b) E. L. Shock and M. D. Schulte, J. Geophys. Res., 1998, 103, 28513; (c) A. Yamaguchi, M. Yamamoto, K. Takai, T. Ishii, K. Hashimoto and R. Nakamura, Electrochim. Acta, 2014, 141, 311.
8. (a) F. Ji, H. Zhou and Q. Yang, Origins Life Evol. Biospheres, 2008, 38, 117; (b) M. T. McCollom and J. S. Seewald, Chem. Rev., 2007, 107, 382.
9. (a) B. Wu, Y. Gao, F. Jin, J. Cao, Y. Du and Y. Zhang, Catal. Today, 2009, 148, 405; (b) H. Takahashi, L. H. Liu, Y. Yashiro, K. Ioku, G. Bignall and N. Yamasaki, J. Mater. Sci., 2006, 41, 1585.
10. (a) F. Jin, Y. Gao, Y. Jin, Y. Zhang, J. Cao, Z. Wei and R. L. Smith Jr, Energy Environ. Sci., 2011, 4, 881; (b) L. Lyu, X. Zeng, J. Yun, F. Wei and F. Jin, Environ. Sci. Technol., 2014, 48, 6003.
11. Z. Huo, M. Hu, X. Zeng, J. Yun and F. Jin, Catal. Today, 2012, 194, 25.
12. (a) P. Haueter, S. Moeller, R. Palumbo and A. Steinfeld, Sol. Energy, 1999, 67, 161; (b) S. Uchida, SolarPACES, Sept 11–14, 2012, I-11; (c) M. Galvez, G. Loutzenhiser, I. Hischier and A. Steinfeld, Energy Fuels, 2008, 22, 3544; (d) E. Alonso, C. Hutter, M. Romero, A. Steinfeld and J. Aguilar, Energy Fuels, 2013, 27, 4884.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, characterization data. See DOI: 10.1039/c5ra02872h

---