UC Berkeley

Electrochemical reduction of CO2 (CO2R) to chemical fuels and feedstock via the use of renewable electrical energy is a potential way to mitigate rising atmospheric CO2 emissions. The ultimate goal of this research field is the ability to develop catalysts capable of converting CO2 selectively and efficiently into specific products. Of all the possible products from CO2R, products with 2 or more carbons (C2+ products) such as ethylene and oxygenates (e.g. ethanol) are highly desirable as they are already used as fuels in the existing energy infrastructure or as chemical feedstock in industry. Therefore in this thesis, I explore various catalytic materials to facilitate CO2R to generate these products.

Initially, I was excited by reports in the literature on the use of heteroatom doped carbon materials as catalytic materials. For example, it was claimed that these materials could convert CO2 selectively to CO, formate or even to methane. These materials typically consist of a carbon structure doped with ~1-5% of heteroatoms such as nitrogen, sulfur, phosphorous or even boron. Synthesis of these materials is simple; they usually involve high temperature annealing of a carbon material (e.g. graphene) with a source containing the heteroatom of choice. However, I discovered that these materials can carry various levels of metallic impurities depending on their preparation method. Therefore, the true catalytic activity of these carbon materials can be obscured by these catalytically active impurities. In Chapter 2, I therefore investigate the role of metallic impurities present in different carbon materials during CO2R.

Next, I investigated nanostructured catalysts for CO2R. These catalysts are prepared by electrochemical oxidation and reduction cycling of Cu foil assisted by halide anions in the solution. These halide anions serve to promote the oxidation process of Cu to create Cu2O and subsequent reduction back to metallic Cu0 results in a nanostructured surface. These catalysts have been shown to be more selective towards C2+ products (especially ethylene) compared to flat, polycrystalline Cu. In Chapter 3, I thus examine the effect of different halide anions on the nanostructuring process and its resultant catalytic activity for CO2R.

In the literature, oxide-derived Cu catalysts have been shown to be promising catalysts to facilitate the reduction of CO2 to C2+ products. Typically, these catalysts are synthesized by oxidation of Cu to Cu2O and/or CuO and subsequent reduction back to Cu0. Also, the cation in the electrolyte used for CO2 reduction used in the has been shown to be very important in determining the C2+ selectivity of Cu catalysts. In particular, a larger cation such as Cs+ promotes a higher C2+ selectivity compared to K+. An interesting question is whether oxide-derived Cu catalysts could be combined with Cs+ to yield very high C2+ selectivity. In Chapter 4, I prepare oxide-derived catalysts via 4 different means according to previous reports and I study their catalytic activity for CO2R in aqueous electrolyte containing either K+ or Cs+. I show that a very high faradaic efficiency for formation of C2+ products (~70%) can be achieved by optimizing the roughness factor of these catalysts with Cs+ containing electrolyte.

It has been proposed by various researchers that oxides in oxide-derived Cu catalysts are stable despite the highly reducing conditions applied during CO2 reduction. Therefore, a large amount of residual oxides remain in the catalyst and these have been proposed to be the reason for high C2+ selectivity for oxide-derived Cu catalysts. In chapter 5, I investigated whether oxides are truly stable during CO2 reduction. I fabricated 18O labeled catalysts and used secondary ion mass spectrometry to determine the 18O content before and after CO2 reduction. I found that <1% of the original 18O content remains after CO2 reduction, showing that residual oxides are unstable, as thermodynamically predicted.

The previous chapters all describe catalysts which have ill-defined structures. I therefore sought to obtain greater control over the dimensions and composition of my catalysts and I turned to photolithography to fabricate model catalysts. In Chapter 6, I describe a device which I named the “AuCu device” to investigate a process termed as “CO crossover” in CO2R. CO spillover occurs when a CO producing catalyst produces CO from CO2 and this can then crossover onto Cu sites. This result in an increased production of C2+ oxygenates at the expense of ethylene. The AuCu device which I fabricated consists of interdigitated lines of Au and Cu on an insulating SiO2 substrate. In this device, Au serves to produce CO and the Cu serves to reduce this CO as well as externally supplied CO2 to further reduced products. The lines are designed to be electronically isolated and as such can be independently actuated. Such a system is therefore ideal to study the CO crossover effect.

Additionally, in chapter 6, I describe another catalyst system, which was also fabricated via photolithography. This system, consisting of Cu dots or lines patterned onto a Ag substrate is designed to take full advantage of the CO crossover effect to achieve a high selectivity to oxygenates. Exposed Ag sites serve to produce CO and thus by varying the areal coverage of Cu dots/lines, systematic control of CO spillover can be exercised. This allows me to tune the ratio of oxygenates to ethylene from 0.59 to 2.39. In the best system, an oxygenate selectivity of >40% can be achieved.

In conclusion, I investigated a variety of catalyst systems for CO2R for my graduate studies. My thoughts and future outlook are summarized in Chapter 7. I outline the main challenges in this field and my thoughts on how selectivity and catalyst activity might be improved. I also emphasize the importance of efforts to understand the mechanism of CO2 reduction.