WO1995003438A1

WO1995003438A1 - Electrolytic carbon dioxide reduction method

Info

Publication number: WO1995003438A1
Application number: PCT/IB1994/000223
Authority: WO
Inventors: Jan Augustynski; Beat Jermann; Robert Kostecki
Original assignee: Universite De Geneve
Priority date: 1993-07-23
Filing date: 1994-07-21
Publication date: 1995-02-02
Also published as: AU7082394A; CH686788A5

Abstract

A method for electrolytically reducing carbon dioxide in a CO2-saturated aqueous saline solution with a neutral to slightly alkaline pH and in the presence of a copper or silver cathode, wherein the cathode is electrochemically activated by at least one cathode potential scanning cycle including a linear increase in potential up to a positive potential relative to the starting electrolytic potential, followed by a linear return to said starting potential.

Description

PROCESS FOR ELECTROLYTIOUS REDUCTION OF CARBON BIOXIDE

The present invention relates to an improved process for electrolytic reduction of CO2 •

Carbon dioxide (CO2) is, in the long term, an essential raw material for the chemical industry and the manufacture of liquid (methanol, ethanol) and gaseous (methane) fuels. The conversion of CO2 into fuel or intermediate chemicals can be carried out by reduction with hydrogen in the gas phase (high temperature and pressure) or directly by electrolytic reduction in an aqueous medium (environmental conditions). In both cases, such an operation requires the supply of electrical energy either to manufacture the hydrogen intended for the reduction process in the gas phase, or to conduct a direct electrolysis of CO2 in an aqueous environment. The direct electrolytic process, on condition of being effective, on the other hand has an obvious advantage of simplicity and of considerably lower operating cost.

In the shorter term, the process of reducing CO2 by electrolysis can also constitute a good means of storing electrical energy, whatever its origin. Indeed, methanol or methane are easier to store than hydrogen, the use of which for this purpose has often been envisaged. This is why, a great deal of research has been undertaken to try to improve this more electrolytic reduction process. especially in order to increase its yield.

On the other hand, we already know in such a process the activity of copper electrodes for the reduction of CO2 which was first described in 1985 (Hori et al., Chemistry Letters, 1695, 1985) and which, since that date, has been the subject of numerous publications. This interest is explained by the fact that copper currently remains the only electrode material making it possible to obtain the most advantageous products for reducing CO2, namely methane, ethylene and ethanol. However, it soon became apparent that the yield (relative to current) of the high CO2 reduction products in the initial period (30 min. To 1 hour) of electrolysis then dropped rapidly due to a poisoning of the copper electrode, hydrogen becoming the predominant product of electrolysis and hydrocarbons tending to disappear completely. Despite various attempts, no effective remedy against blockage of the copper electrode has been described so far.

Thus, the aim of the present invention is to provide a process for the electrolytic reduction of CO2 in aqueous saline solution saturated with CO2 and in the presence of a copper or silver cathode, which allows the above-mentioned drawbacks of known processes to be overcome. , and which more particularly provides an improved yield and which prevents poisoning and thereby blockage of the cathode.

The above object is achieved by the method according to the invention, in which the copper or silver cathode is subjected to an electrochemical activation consisting of the application of at least one scan cycle of the poten¬ tial of the cathode comprising a linear increase in this potential up to a positive potential compared to the initial electrolysis potential, then the linear return to it .

According to a first variant of the invention, the cathode is made of copper and the activation can then comprise from 1 to 5 cycles, preferably 3 cycles, succeeding each other every 3 to 15 minutes, preferably every 5 minutes approx. . The actual scanning consists in a linear increase in the potential of the cathode up to 0.5 to 1.5 V, preferably up to around 1V, compared to a normal calomel reference electrode (ENC), at a speed of 0.5 to 20 V / sec, or better still of 1 to 10 V / sec, and preferably around 5V / sec, the potential then returning to its value of usual origin at a similar speed.

The reduction of CO 2 is in principle carried out at ambient temperature and at atmospheric pressure, in aqueous saline solution saturated with CO 2 (for example acidic carates, chlorides, perchlorates or sulphates of alkali metals), at a neutral to slightly alkaline pH. Electrolysis is generally carried out at a potential of -2V (compared to an ENC reference electrode).

As regards the temperature and pressure of the electrolysis, this can also be carried out between 0 ° and 60 ° C and at a pressure higher than atmospheric pressure, for example hitherto around 5 atm.

As an electrolyte, it is advantageous to use, for example, bicarbonate, sulfate or perchlorate. Na or K, in a concentration of 0.1 to 1 M, preferably 0.1 to 0.5 M.

According to a second variant of the method, the copper cathode can be replaced by a silver cathode; in this case the solution will contain, in addition to a perchlorate, sulfate or alkali chloride, a small amount of a soluble copper salt, for example CUSO, in a concentration of 10 ^~ 6 to 5 x 10- * - ^* mole / 1, preferably between 10 - * - * and 5 x 1 0 ~ 4 mole / 1. The electrolytic reduction of CO2 dissolved in the solution will then be accompanied by a deposit of metallic copper on the silver cathode. With such a silver electrode, the scanning consists of a linear increase in the potential of the cathode up to about 0.4 V, preferably about 0.3 V, relative to the ENC, at a speed from 0.1 to 1 V / sec, preferably at about 0.2 V / sec, the potential then returning to its initial value at a similar speed.

The method according to the invention will now be described in more detail and with reference to the following two nonlimiting examples.

Figures 1, 2 and 3 are graphs representing the faradaic yields (in%) relative to the duration of the electrolysis for three experiments carried out with the method according to the invention.

FIGS. 1A and 2A are graphs corresponding to those of FIGS. 1 and 2, but showing the results of corresponding experiments carried out in the usual manner, that is to say without activation. Example 1

The electrolysis is carried out in a glass cell, with two compartments separated by a Nafion membrane. The catholyte (30ml) is continuously traversed by a CO2 flow of approximately 15ml / min. The working electrode here is a copper rod 6 mm in diameter (surface = 0.28 cm ^* -), polished with 0.3 μm alumina, then pickled for 15 seconds in 10% HC1, just before beginning of electrolysis. Instead of this chemical etching, one can use an electrochemical preparation method of the electrode, which has the advantage of giving better reproducible surfaces, and which consists in applying an anodic current of 100 mA for 5 minutes and in an 85% H3PO4 solution.

The three solutions used are 0.1M sodium bicarbonate, 0.5M potassium bicarbonate and 0.5M potassium sul¬ fate. The reference electrode is the ECN, the potentials mentioned below relating to it. The electrolyses were carried out at a potential of -2V. Activation, which prolongs the life of the cathode, consisted of a succession of three cycles of scanning of the potential up to an anode limit lying between +1.5 and +0.5 V and at the speed of about 5 V / sec. These cycles were repeated every five minutes. Corresponding control electrolyses were carried out under the same conditions, but without the activation cycles.

Analysis of reduction products (here methane and ethylene), both in the gas phase, during of an electrolysis, that in the solution, at the end of an experiment, was carried out using a gas chromatograph provided with an FID detector.

The three experiments which have been carried out are therefore distinguished by the use of different electrolytes.

The first was carried out in 0.5M KHCO3, at 25 ° C, and the results obtained are presented in Figures 1 and 1A. With activation according to the invention (fig. 1), the faradic yields are around 40 to 50% for methane and 10% for ethylene, and remain stable, even over an electrolysis period of 24 hours. . During this experiment, a few percent of ethanol was detected, as well as only fractions of a percent of CO and formates. On the contrary, the results obtained without activation (fig. 1A) confirm that very quickly (less than 2 hours for ethylene and approximately 3 hours for methane) the faradaic yields fell to negligible values close to zero . The cathodic current progresses during an electrolysis, in this solution, of about 40 to 60 mA / cm ² . By way of comparison, the current densities used during the electrolysis of water on an industrial scale are of the order of 200 mA / cm ² .

With 0.1 M NaHCO3 as electrolyte, the second experiment was carried out under conditions identical to those of the first, but the faradaic yields obtained are slightly lower, namely approximately 40% of CH4 and 10-20%. of C2H (fig. 2). The much weaker currents (10 to 20 mA / cm ² ) make it possible to extend the reduction over more than 50 hours while maintaining the constant things. On the other hand, as before, it can be seen that without activation, the yield drops to insignificant values very quickly, after 1 h 30 to approx. 2 hours (FIG. 2A).

Finally, the third experiment, carried out in a solution of 0.5 M K2SO4 and at 5 ° C., made it possible to maintain high faradaic yields, that is to say around 60% for methane and around 10% for ethylene, over a period of at least 75 hours and with cathode currents of the order of 80 mA / cm ² (fig. 3).

Example 2

For this second example, the copper cathode has been replaced by a silver cathode. As in the case of the copper cathode alone, the proper functioning of the CO 2 electrolysis required periodic activation of the silver cathode. Such an activation here consisted of a succession of three cycles of a linear scan of the potential of the cathode between the usual potential of electrolysis, for example -2V, and + 0.3V. This series of 3 scanning cycles was carried out at a speed of 0.2 V / sec. and repeated every 5 min. during electrolysis. During these activation cycles, a greater or lesser part of the copper deposited at the cathode during the electrolysis of CO 2 is oxidized and passes into the solution in the form of Cu ²⁺ ions.

This method makes it possible to obtain spectacular yields of the main products which are comparable, or even superior to the copper electrode, while at the same time also avoiding deactivation of the cathode. For example, during an electrolysis carried out in 0.1M aCl04, saturated with CO2, containing 3.3 x 10 ^-4 mole / 1 of CUSO4, with the silver cathode polarized at -2V (versus ENC) and activated every 5 min., the faradaic yields after 5 h. electrolysis were about 35-40% CH ₄ , 30% C ₂ H ₄ , 9% C ₂ H ₅ OH and 3% CO.

It is therefore clear from the above experiments illustrating the invention by way of examples that the activation which is the subject of this invention makes it possible to remove the blocking of the cathode, and thus to obtain high faradaic yields, even on electrolysis times longer than 24 hours. Such an improvement in yields and in the possible duration of the process therefore makes it possible to envisage the use of electrolytic reduction of CO2 on an industrial scale.

Claims

1. Process for the electrolytic reduction of carbon dioxide in an aqueous saline solution at neutral pH with weak alkali saturated with CO2 and in the presence of a method of copper or silver, in which this cathode is subjected to an electrochemical activation consisting in the application of at least one cycle of a scanning of the poten¬ tial of the cathode comprising a linear increase in this potential up to a positive potential compared to the initial electrolysis potential, then the return linear to this one.

2. Method according to claim 1, in which the ca¬ thode is made of copper and it is subjected to a succession of 2 to 5 cycles of a linear scanning of the potential of the cathode, which consists in an increase up to 0 , 5 to 1.5 V of said potential relative to a calomel reference electrode, at a speed of 0.5 to 20 v / sec, then on returning to the initial electrolysis potential at a similar speed.

3. Method according to claim 2, wherein the linear ba¬ layering of the cathode potential is carried out up to about 1V, and the speed of this scanning from 1 to 10 V / sec, preferably about 5 V / sec.

4. The method of claim 1, wherein the method is silver and it is subjected to a succession of 2 to 5 cycles of a linear scan of the potential of the cathode, which consists of an increase to about 0.4 V of said potential relative to a calomel reference electrode, at a speed of 0.1 to 1 V / sec, then to the return to electrolysis potential initial at a similar rate, and in which the electrolyte contains a soluble copper salt in concentration of 10 "^ to 5-10 - * - * 'mole / lt.

5. Method according to claim 4, in which the linear ba¬ layering of the cathode potential is carried out up to approximately 0.3V, and the speed of this scanning is approximately 0.2 v / sec.

6. Method according to one of claims 2 to 5, in which the frequency of the scanning cycles is 3 to 15 minutes, preferably about 5 minutes.

7. Method according to one of claims 1 to 6, in which the electrolyte chosen from acid carbonates, chlorides, perchlorates and sulphates of alkali metals, in a concentration of 0.1 to 1M.

8. Method according to claims 2 and 7, wherein the electrolyte is chosen from potassium bicarbonate, sodium bicarbonate and potassium sulfate, in a concentration of 0.1 to 0.5 M.

9. Method according to claims 4 and 7, wherein the electrolyte is sodium perchlorate 0.1 to 0.5 M.