From Greenhouse Gas to Green Fuel
Zooming in on a scalable carbon cycle
Anna Wuttig, NSF Graduate Research Fellow, MIT Chemistry
While nature’s processes of photosynthesis and respiration allow for CO2 to be utilized in a cyclic and sustainable way, society has yet to find a scalable carbon cycle to do the same. I decided to pursue a PhD in chemistry because, in my undergraduate studies, I learned that both a molecular-level understanding of known catalytic systems and innovative synthetic strategies to design more efficient and selective catalysts are necessary to achieve chemical recycling of carbon. 
I remember my first conversations over Skype with Professor Yogesh Surendranath. He had just accepted his assistant professorship at the time, and I was inspired by his emphasis on the importance of gaining mechanistic insight into known catalytic systems for various fuel-forming half reactions in order to rationally design improved systems. As a graduate student in the Surendranath Laboratory, my primary aim has been to understand how electrified metals convert CO2 to fuels. In principle, this process would utilize waste CO2 while simultaneously storing electricity from intermittent renewable sources such as solar and wind in a high-energy-density form. However, I quickly learned that there are several competing pathways!
CO2 reduction in aqueous electrolytes suffers efficiency losses due to the wide array of products accessible over a narrow potential range in addition to the simultaneous reduction of water to hydrogen gas (H2). A key problem is that we do not know how the electrons and protons interact to selectively produce one fuel over another. Are there differential proton coupling requirements for CO2 reduction versus the competing H2 evolution? To answer this question, I utilized a gold (Au) model electrocatalyst, which is known to produce only carbon monoxide (CO) and H2 under CO2 reduction conditions. I measured the rate of CO2conversion to CO as a function of various experimental parameters using electrokinetic studies and in-line gas chromatography. I found that the rate-limiting steps are not coupled with proton transfer. The same analysis for the simultaneously occurring H2 production revealed that it involves both electron and proton transfer in its rate-limiting step. I assembled the data to construct a mechanistic model that predicted that impeding proton transfer to the surface is an effective strategy for controlling and improving CO2-to-fuels catalyst selectivity, especially since the slow hydration (~23 s half life) of CO2 impedes the replenishment of viable proton donors, bicarbonate and hydronium, at an appreciable rate. Using this knowledge, we postulated that slowing proton transfer would limit H2 production and not affect CO production. A postdoc, Shoji Hall, and a graduate student, Youngmin Yoon, engineered an elegant porous Au electrode and illustrated that proton transfer dynamics can be used to control selectivity in fuel formation, which supports our mechanistic model.
In addition to preventing parasitic H2, selective multi-electron, multi-proton reduction of CO2 processes requires electrocatalysts capable of binding reaction intermediates and promoting complex bond rearrangements. In a pioneering study by Yoshio Hori of Chiba University in Japan and coworkers around 30 years ago, CO2 reduction electrocatalysts were classified based on their selectivities for various CO2 reduction products. While the reduction of CO2 on these surfaces demands a high overpotential or energy penalty in all cases, the study provides insight into how catalyst selectivity may be achieved. 
This study led to the contemporary hypothesis, supported by Density Functional Theory calculations and ultra-high vacuum thermochemical studies, that catalyst selectivity is correlated with the adsorption enthalpy of the two-electron-reduced CO intermediate. For example: (1) main group elements display low CO binding strengths (adsorption enthalpy of ~0 eV for tin (Sn)) and high selectivities for formate (88% for Sn); (2) group 11 elements exhibit moderate CO adsorption enthalpies (ex. −0.40 eV for Au) and display good Faradaic yields for CO (87% for Au); and (3) group 10 elements display high CO adsorption enthalpies (ex. −1.30 eV for platinum (Pt)) and are readily poisoned by CO, leading to parasitic H2 production(ex. 96% for Pt). Curiously, copper (Cu), which displays a comparable CO adsorption enthalpy to other group 11 elements (−0.43 to −0.49 eV), is unique as the only known metal catalyst that generates higher order carbonaceous products (33 % for methane and 25% for ethylene) from CO2. Clearly, CO is a key intermediate in CO2 reduction and its adsorptive behavior changes from metal to metal. This raises another key question: why does copper yield hydrocarbons and alcohols from CO2 while other metals stop at CO? I believed that the question could be tackled through spectroscopic resolution of the key surface-bound intermediates.
Despite the key role that these surface binding interactions may play, there are few tools capable of probing them in real time. In our studies, we relied on the technique of in situ surface-enhanced infrared absorption spectroscopy (SEIRAS) in an attenuated reflectance mode developed by Professor Masatoshi Osawa at Hokkaido University in Japan. This method allows for the quantitative characterization of surface-bound intermediates on an electrode surface by taking advantage of the enhanced local electromagnetic field and changes in dipole moments of a nanostructured catalyst material. I initiated a collaboration with the Osawa group to learn how to set up the instrumentation and execute the synthetic steps necessary to perform SEIRAS. Through the MIT International Science and Technology Initiatives (MISTI) program, I was fortunate to be able to travel to Japan for a few months to learn this technique and perform initial measurements with graduate student Momo Yaguchi and with Kenta Motobayashi, an Assistant Professor in the Osawa group. As I observed the incoming spectroscopic data for the first time after many attempts, I remember being excited about the power of this technique; a complex heterogeneous surface suddenly became less of a black box and more of a well-defined, determinable structure. Now, we also have this capability in the Surendranath group!
Through these experiments, we have learned that the intermediate CO binding processes are far more complex than we initially thought. For example, there are two binding modes of CO on Au. One is bonded to one Au atom, linearly bonded CO, and another is bonded to two Au atoms, bridge-bonded CO. By tracking these spectroscopic signatures for CO under CO2 reduction conditions, we discovered that the linearly bonded CO is reversibly bonded to the surface and exists in low surface population, so it can be readily liberated from the surface to form the CO product. Conversely, bridge-bonded CO is irreversibly bonded to the surface to about 20% of the Au sites and acts as a kinetically inert spectator. In contrast to Au, where kinetically competent linearly bonded CO dissociates from the surface, labile CO species accumulate on the Cu surface providing a pool of reactant primed for further reduction to higher order products. Taken together with copper’s unique ability to mediate higher order fuel synthesis beyond CO, these spectroscopic observations provide insight into the dynamic nature and population landscape of this common intermediate that underpins CO2-to-fuels conversion.
Using these strategies, we were able to formulate new mechanistic models of CO2 electroreduction catalyzed by transition metal surfaces, and I have begun to disentangle the effects of competing reactions. While, in the short-term, my studies motivate me to further identify the extent of these bifurcating processes, I aspire to expand my knowledge and, more importantly, to build the process of how I think about scientific problems in the context of addressing societal challenges and advocating for solutions. At MIT, I have been continuously motivated and invigorated by the deep, careful, and relentless approach to understanding systems on a molecular level. I have been grateful to have opportunities to engage in interdisciplinary energy-focused discussions such as attending the MIT Energy Initiative (MITEI) Fall Research Conferences and MIT Energy Night and also learn from peers in the Department of Chemistry by participating in and organizing the chemistry student seminars with fellow graduate students Kathleen White and Timothy Barnum. I am thankful for the many opportunities that Professor Surendranath and the Chemistry Department provide, and I cannot wait to see what we uncover next.
Please contact Elizabeth Chadis if you are considering a gift to the School of Science:

Elizabeth Chadis

Assistant Dean for Development
t: 617-253-8903
Edit People Page