1

The oxidation of water and the reduction of CO2 to fuels Investigators Thomas F. Jaramillo, Assistant Professor; Jens K. Nørskov, Professor; Joel Varley, Postdoctoral Researcher; Lars Grabow, Postdoctoral Researcher; Kendra Kuhl, Graduate Student; Etosha Cave, Graduate Student; David Abram, Graduate Student; Desmond Ng, Graduate Student; Toru Hatsukade, Graduate Student; Jakob Kibsgaard, Postdoctoral Researcher; Christopher Hahn, Postdoctoral Researcher; Samuel Fleischman, Postdoctoral Researcher. Abstract The goal of this project is to combine experimental and theoretical methods in investigating and developing three types of catalysts: metal surfaces for CO2 reduction, metal sulfides for CO2 reduction, and metal oxides for the oxygen evolution reaction (OER). In this project, we have: (1) Developed new instrumentation and methodology to conduct CO2-reduction studies. In particular, we have developed a custom CO2 electrolysis reactor with on-stream head-space analysis by gas chromatography. We have also developed 1H-NMR and 13C-NMR methods to detect liquid products in the liquid electrolyte post-reaction. (2) Studied CO2 electroreduction on 7 different metals: Cu, Ni, Pt, Fe, Au, Ag, and Zn, establishing their activity and selectivity for different gaseous and liquid products as a function of applied potential. Due to the unprecedented sensitivity of our detection methods, several new and/or previously unseen products were identified, leading to further insights into the mechanisms of CO2 reduction on metal surfaces. (3) Examined metal sulfide catalysts such as [4Fe-4S] cubanes, FeS2, FeS, MoS2, and VS2 for CO2 electroreduction. The metal sulfide catalysts evolved H2 as a major product and showed little selectivity to CO2 reduction. (4) Studied the effect of surface chemical modification on the activity and selectivity of CO2 reduction catalysts by modifying Pt electrodes with polyaniline. Changes in Faradaic efficiencies indicate that polyaniline can influence the selectivity of a metal surface for the CO2 reduction reaction. (5) Synthesized and investigated MnOx and Co-MnOx catalysts for the OER and the oxygen reduction reaction (ORR). The OER activities are comparable to precious metals, suggesting that these catalysts could be a cost-effective replacement. Introduction The objective of this project is to develop catalysts for two key energy conversion reactions: CO2 electroreduction and the oxygen evolution reaction (OER). Developing catalysts for these reactions will enable technologies that can produce carbon-neutral fuels and/or commodity chemicals when coupled to renewable energy sources, e.g. wind and solar (Figure 1). These molecules are the foundation of our energy infrastructure and chemical industry. As a result, the global demand for these compounds is massive, with annual production rates on the order of 1013 kg/yr.1 Such technologies could be modular to fit both portable (10W-100kW) and stationary (1kW-100MW) production scales. Table I shows hypothetical production rates (kg/hr) of various CO2 reduction products for both portable and stationary production scales. All values assume that both CO2 reduction and OER catalysts operate at 0.3V overpotential

2

Figure 1. Electro-reduction of CO2 coupled to renewable electricity sources, such as wind or solar, can enable a CO2-neutral energy cycle.

each. These production rates suggest that further development of CO2 electroreduction and OER catalysts will allow for the scalable production of fuels and commodity chemicals at rates which could meet global demands. Currently, there are no efficient catalysts for CO2 reduction, and the best OER catalysts are precious metals. Therefore, significant breakthroughs in catalyst performance will be necessary to clear the path shown in Figure 1 to produce renewable, sustainable chemical fuels. In order to achieve our research goals of electrocatalyst development, we are combining experimental and theoretical methods in investigating three types of catalysts: (1) metal surfaces for CO2 reduction, (2) metal sulfides for CO2 reduction, and (3) metal oxides for the OER.

Table I. Production rates (kg/hr) of various CO2 reduction products as a function of input power.

3

Background In the past several years, there have been a number of developments in the field of CO2 reduction and water oxidation. Particularly relevant results in the field of CO2 reduction come from laboratories at UC Berkeley, the Technical University of Denmark (DTU), Stanford University, Leiden University, the University of Illinois, Urbana-Champaign (UIUC), Princeton University, and UC Irvine. At UC Berkeley, Prof. John Newman provided a detailed kinetic analysis of CO2 electroreduction on Au and Ag surfaces.2,3 However, they focused entirely on gas phase products, without examining any liquid phase products, though any liquid products would be small relative to the major gas phase products of H2 and CO. Theoretical analysis by Prof. Jens Nørskov's group, at Stanford University and DTU, has explored the effect of different surface orientations on activity and selectivity, and developed activity descriptors for CO2 reduction metal surfaces to guide the development of better catalysts.4,5 At DTU, Prof. Ib Chorkendorff's group has collaborated with Prof. Jens Nørskov's group to experimentally confirm their prediction that a rougher surface would favor multi-carbon products.6 The design of a new CO2 electrolysis reactor by Prof. Thomas Jaramillo's group, at Stanford University, allowed for the detection of previously unobserved products and new insights into reaction mechanisms on Cu surfaces.7 At Leiden University in the Netherlands, researchers in Prof. Marc Koper's group used an on-line mass spectrometer to understand the mechanism of CO2 reduction on Cu electrodes.8 At UIUC, Prof. Paul Kenis and Richard Masel used ionic liquids as catalysts to electrochemically convert CO2 to CO at extremely low overpotential.9 At Princeton University, Prof. Andy Bocarsly has continued in his efforts to understand the mechanistic pathway for methanol production involving pyridine and a metal electrode (e.g. Pd).10,11 Prof. Markus Ribbe’s group at UC-Irvine has reported the discovery that vanadium nitrogenase from Azotobacter vinelandii can remarkably catalyze the (electro-) reduction of CO under ambient pressure and temperature.12

In the field of water oxidation, efforts are continuing in Prof. Daniel Nocera's group at MIT in understanding their Co-phosphate based electrode, though this catalyst operates at > 0.4 V overpotential, leaving much room for improvement.13 Also at MIT, Prof. Yang Shao-Horn's group has identified 3d orbital occupancy as an activity descriptor for perovskite OER catalysts.14 At UC Berkeley, Prof. Alex Bell has been studying the OER on Au electrodes, revealing the presence of the –OOH intermediate for the first time – a breakthrough in understanding the reaction mechanism.15

Results CO2 Electrolysis Reactor Design One difficulty in studying CO2 electroreduction is the need to measure both gas and liquid phase products. There is no 'standard' method in the field for conducting these kinds of studies, thus our first goal was to develop a method that would allow for high sensitivity in identifying and quantifying gaseous and liquid-phase products of reaction. To achieve this goal we ran electrolysis experiments for 1 hr at a range of potentials using a custom cell designed to have a large electrode area and small electrolyte volume to maximize product formation and concentration. CO2 was constantly flowed through the cell during electrolysis to achieve good mass transport and gas phase products were measured by a gas chromatograph (GC) downstream of the electrolysis cell. Liquid

4

Figure 2. Electrolysis cell used for CO2 reduction with representative gas chromatograph and NMR data used to quantify products.

phase products were measured by nuclear magnetic resonance (NMR) after electrolysis. Figure 2 shows a schematic of the cell and examples of the data used to quantify the products that were formed.

CO2 Electrolysis on Cu Metal The CO2 reduction activity of Cu metal is well known, so we began by benchmarking Cu foils in our custom electrolysis reactor to compare with literature reports. CO2 reduction on Cu is known to produce methane, ethylene, formate, CO, ethanol, acetaldehyde, propanol, propionaldehyde, and allyl alcohol along with hydrogen as a side product.16 In addition to these products, methanol and acetate have been detected in a few studies.17,18 Using the methods described in the previous section, we detected and quantified all of the previously detected CO2 reduction products on Cu (from all previous studies combined in one single set of experiments). Due to the enhanced sensitivity of our methods, we were also able to detect a number of previously unreported products such as: glyoxal, ethylene glycol, glycolaldehyde, acetone, and hydroxyacetone. Current efficiencies (fraction of current towards each product) and partial current densities (scales linearly with turnover frequency) are plotted as a function of voltage to examine the effect of overpotential on CO2 reduction products (Figure 3). At lower overpotentials, CO2 is reduced only to CO and formate, which are believed to have smaller kinetic barriers to form than the more reduced products. At higher overpotentials where the kinetic barriers can be overcome, hydrocarbons such as ethylene and methane are formed as major products. While lower overpotentials produce more ethylene than methane, this trend is reversed as the overpotential is increased. The pathways to multi-carbon products, such as ethylene, must contain a C-C coupling step in addition to the necessary proton and electron transfers. Therefore, one theory for this decrease in ethylene production is that higher overpotentials cause the proton and electron transfers to become more favorable than C-C coupling, leading to the production of less multi-carbon products and more methane and hydrogen.

5

Figure 3. Current efficiency and Tafel plots of partial current going to products over the voltage range measured.

An examination of the partial currents of products with multiple carbons shows that they tend to track each other across the voltage range, suggesting that these products share common mechanistic pathways (Figure 3). However, the observed multi-carbon products display a wide range of functional groups such as: alcohols, carboxylates, ketones, and double bonds. As the first lab to have identified and quantified this more comprehensive view of the reaction products and rates of production, we were able to provide further insights into reaction mechanisms that could potentially lead to such diverse products. One important insight was the realization that the products containing a carbonyl group can have other possible forms due to tautomerization. By examining these tautomers, we realized that a plausible mechanism could proceed through the dehydroxylation of the enol or diol form of the various products. Figure 4 shows this hypothesized pathway in which the enol or diol form of the products in present as an intermediate on the electrode surface, and either desorbs or is dehydroxylated leading to the formation of more reduced products.

6

Figure 4. Proposed pathway for multi-carbon product formation.

Examining CO2 reduction on Cu demonstrates the utility of our new methodology for product detection. More importantly, the increased product sensitivity of our method has led to several new insights into the reaction products and pathways of CO2 reduction on Cu. A greater understanding of the mechanism of CO2 reduction could lead to catalysts which are engineered to selectively form the desired product for the specific application. Studies of electrochemical reduction of CO2 on other transition metals

In order to further our understanding of CO2 reduction, we explored the activities of Au, Ag, and Zn, which bind CO weakly, and Ni, Pt, and Fe, which bind CO strongly. Figure 5a shows the average current density drawn across a range of potentials on each metal and Figure 5b shows the percentage of that current that went towards CO2RR instead of making hydrogen. The plot confirms what has been reported previously, Au, Ag, and Zn are good CO2 reduction catalysts, displaying a high current efficiency for CO2 reduction, whereas Ni, Pt, and Fe are poor CO2 reduction catalysts and mostly produce hydrogen. Cu shows good current efficiency for CO2RR, but less than Au, Ag, or Zn.

7

Figure 5. (a) Partial current density during CO2RR electrolysis on metal electrodes. (b) Current efficiency for CO2RR. (c) Partial current density of methane. (d) Partial current density of methanol.

The unprecedented sensitivity of our experimental methods for minor product

quantification allowed us to detect methane and/or methanol formation on all metals studied. Figures 5c and 5d show the partial current density of methane and methanol across the potential range explored. Among the metals that make both products, methane and methanol appear at similar onset potentials and the partial current densities of each product track each other across the potential range. This suggests that the mechanistic pathway to methane and methanol share many common intermediates and that the two pathways are only differentiated late in the reaction. Conventional wisdom suggests that only Cu metal catalyzes the formation of hydrocarbons and alcohols, but our more sensitive methods show that production of these compounds is a universal property of transition metals.

Figure 6 contains volcano plots of CO2RR activity vs. the binding energy of CO to

each metal, as suggested by theory.5 The plots demonstrate the Sabatier principle,5,19,20 which states that there is an optimal binding energy for reaction intermediates and that intermediates bound more strongly or weakly than the optimum value will lead to lower catalyst activity. Two metrics of catalyst activity, the partial current density (6a) and the onset potential (6b), for CO2RR are shown. By either metric, Au is the closest to the top of the volcano, suggesting that it has a CO binding energy nearer to the ideal value than

8

Figure 6. Volcano plot of CO2RR activity vs. CO binding strength. (a) Activity measured by partial current density of CO2RR at -0.8 V. (b) Activity measured by onset potential.

any other pure metal. Ag and Zn form too weak of bonds to CO and the other metals bind CO too tightly, so they show lower activity for CO2RR. Other trends are shown by the shaded regions. The red region spans the onset potential of HER on the different metals and shows that there is a weak correlation between earlier onset voltage of HER and stronger CO binding energy. The blue region spans the onset voltages where hydrocarbons and alcohols appear. There is a trend toward earlier onset for more methane and methanol on metals that bind CO strongly.

The discovered trends in catalyst activity can guide the search for catalysts with

higher CO2RR activity. The data suggest that the hunt for new catalysts (ie. transition metal alloys) should focus on materials with CO binding energies between Au and Cu, with the hope that they would be capable of forming more reduced products, but with an earlier onset and higher partial current density for CO2RR than Cu. Better CO2RR catalysts would open up more alternatives to fossil fuels for electricity generation and as a source of commodity chemicals. CO2 Reduction with Sulfides Taking inspiration from the enzyme carbon monoxide dehydrogenase (CODH),21 we explored the CO2 reduction activity of Fe4S4 cubanes supported on highly oriented pyrolytic graphite (HOPG). While the STM image in Figure 7a demonstrates that we were successful in attaching the cubanes to the surface, XPS spectra taken before and after water exposure show that the Fe4S4 complexes are unstable. Therefore, further

9

Figure 7. (a) STM of cubanes attached to an HOPG surface with XPS showing changes to the Fe and S signal after water exposure. (b) Pyrite film before and after electrochemistry. (c) Current efficiency (%) of hydrogen production on metal sulfides.

work will be necessary to improve the stability of Fe4S4 cubanes for use in aqueous systems.

Next, we explored the CO2 reduction activity of thin films of FeS2, FeS, MoS2, and

VS2. Similar to the case of Fe4S4 cubanes, both FeS2 and FeS were unstable during electrolysis (Figure 7b). A plot of the current efficiency (%) of hydrogen production shows that hydrogen was the major product detected for all sulfides during the CO2 reduction experiment (Figure 7c). Previous literature reports show that sulfides such as MoS2 are active catalysts for hydrogen evolution.22,23 Therefore, our experiments suggest that the tested sulfides have poor activity for CO2 reduction, and instead will selectively reduce protons into hydrogen. Studies of Electrochemical Reduction of CO2 on Pt metal and Polyaniline-Pt

In an effort to understand how to impact catalyst activity and selectivity on metal surfaces, we have begun exploring metal alloys as well as chemically modified surfaces. In this study, we investigate the impact of polyaniline (PANi) on Pt. As an aqueous CO2 reduction catalyst, Pt has been reported to produce primarily H2 with 0.1% faradaic efficiency for formate24 at 5mA/cm2. One study demonstrated that methanol, formaldehyde, and trace amounts of methane are produced as well.25 Using our recently

10

Figure 8. Comparison of Faradaic efficiciencies for carbon-containing products of CO2 electrolysis for PANi-Pt and Pt catalysts at various potentials. Electrolysis measurements were done in 0.1 M KHCO3 electrolyte for current densities up to 10 mA/cm2.

developed detection methods, we first quantified Faradaic efficiencies for H2, formate, CO, methanol, and methane production on Pt metal (Figure 8).

To investigate effects of PANi at the Pt-liquid interface, approximately 10 nm of

PANi was electrochemically deposited on Pt metal foils using cyclic voltammetry. The PANi-Pt electrodes were subsequently tested using the same CO2 reduction procedures as for bare Pt. While H2 was still the dominant species produced, the product distribution of among the minor species was significantly altered (Figure 8). The most significant changes, caused by the addition of the PANi film, are an increase in formate Faradaic efficiency from 0.3% to 1.5%, and an increase in CO Faradaic efficiency from 0.1% to 0.8%. The CO production also appears to increase at higher overpotentials. 13CO2 experiments confirmed that the increased formate was not attributable to degradation of the PANi film. Since formate and CO are 2e- products with the same proposed rate-limiting step while methanol has a different proposed rate-limiting step, it is possible that the amine groups of PANi are stabilizing the intermediate for the 2e- products and destabilizing methanol intermediates. Understanding these effects are critical to influence CO2 reduction selectivity, and ongoing efforts are aimed at determining how and why PANi can produce such differences.

Theory Theoretical studies of CO2 Reduction with Dehydrogenase enzymes To supplement and guide the experimental efforts on bioinspired systems, we have explored the relevant reaction mechanisms with first-principles calculations. To this end we have used the CO dehydrogenase enzyme (CODH) as a model system to study the

11

redox reaction CO + H2O ↔ CO2 + 2H+ + 2e–. The active site of this metalloenzyme consists of a Ni and Fe4S4 cluster akin to a cubane system, which we have also investigated. Figure 9 shows a schematic of the studied structures, each of which provides insight into different mechanisms underlying the catalytic activity of the enzyme. These include the preferred adsorption sites on the cluster(s) and the role of the surrounding ligand environment in adsorbate stabilization and selectivity, both of which can ultimately offer routes to engineering superior catalysts.

A CB

In Figure 10 we show the relaxed atomic geometries of the adsorbed COOH* and CO* species and their energetics on one type of model the dehydrogenase species from the Methanosarcina barkeri bacteria (Mb-CODH).26 We have included the first-order ligand environment surrounding the active site to explicitly account for stabilization provided by hydrogen bonding, which we have found has a significant effect in the observed activity. The free energy diagram describing the CO2 + 2H+ + 2e– ↔ CO + H2O reaction is shown in Figure 10c, where we find a modest overpotential of -0.47 eV necessary to drive the reaction electrochemically, in excellent agreement with experiments on other CODH enzymes.27 We find that the effect of the different cluster atoms also contribute to the binding energies, as can be see from the COOH* where the C binds to Ni and an O forms a secondary bond to Fe (Figure 10a). Our future plans include exploiting this as a design principle for breaking scaling relations between adsorbates that bind via C or O, as recently detailed in Ref 28.

Figure 9. A) Model of a CODH dehydrogenase enzyme with the immediate ligand environment as determined by high-resolution X-ray crystallography. B) Model of a CODH enzyme without the ligand environment. C) Model cubane structures of different Ni-Fe-S stoichiometry. Light green atoms represent Ni, orange represents Fe, and yellow represents S in all clusters shown.

12

We have previously shown for a variety of transition metal surfaces that the CO* and COOH* binding energies both determine the activity in CO2 reduction and follow scaling relations.29 We have combined our current results on CODH and cubane systems with ongoing work in a project studying metal surfaces for the same reaction into a microkinetic model as seen in Figure 11, where we find that both the hydrogen bonding due to ligands and the “alloyed” surfaces of the CODH and cubanes can lead to deviations from the transition metal scaling relations. Such deviations can lead to enhanced activity, as seen by some of the cubane and CODH points that fall closer to the top of the Volcano Plot in Figure 11. The enhanced activity exhibited by the Mb-CODH and the Ni2Fe2S4 cubane cluster were then analyzed further to characterize the relative importance of the geometry (stoichiometry) and stabilization effects of surrounding ligands.

Figure 10. A) Most stable configuration of COOH* and B) CO* adsorbed on the model CODH dehydrogenase enzyme. C) Gibbs free energy diagram for CO2 reduction to CO* and H2O. The black curve shows no applied potential, while the red curve shows the minimum potential necessary to make the reaction exergonic.

13

To further assess the structure-property relationships metalloenzymes exhibit as catalysts for CO2 and CO reduction, we performed a systematic study on cubanes of varying composition. We studied the range of compositions of the FexNi(4-x)S4 clusters as seen in Figure 12, a range which includes the building blocks of both the nitrogenase cofactors and the active sites of the CODH enzymes. a b dc e

Ni

Fe

Cl

S

In Figure 13 we show the predicted activities for the clusters in the kinetic model of Reference 30, where we find clusters with 2 and 3 Ni to be most active. We find the scaling relations exhibited by the clusters deviate significantly from that of elemental metals, and a number of clusters are predicted to be more active for CO evolution than

Figure 11. Plot of the Sabatier Volcano for the CO2 reduction activity using the binding energies of COOH* and CO* as descriptors. The trend line reflects the scaling relations exhibited by transition metal surfaces, of which only the noble metals are in the vicinity of the active region. The pink squares represent different Ni-Fe-S cubanes and the green triangles represent different dehydrogenase enzymes, both of which show favorable deviations from the scaling relation.

Figure 12. Compositions of the model FexNi(4-x)S4 cubanes studied. They range from an all Fe cluster (Fe4S4) shown in (a) to an all Ni cluster (Ni4S4) shown in (e).

14

the best known transition metal catalysts.30 Unfortunately, we find that clusters containing more than 1 Ni are unstable in conditions necessary for CO2 reduction. This offers insight into engineering more stable synthetic cubanes and may offer an explanation to why nature selected cubane compositions that we find to be less active than similar alternatives.

Fischer-Tropsch Synthesis with Nitrogenase enzymes Nitrogenase enzymes, known for their ability to synthesize ammonia from atmospheric N2, have recently been shown to also produce a variety of hydrocarbons from CO and CN-.31-33 These studies suggest nitrogenase enzymes may make extremely promising catalysts for Fischer-Tropsch synthesis under ambient conditions, yet offer a limiting understanding of how the reactions take place. Using first-principles calculations, we studied model systems for the Mo- and V- containing nitrogenase variants (FeMoco and FeVco). We performed extensive calculations on different chemical intermediates in the reduction pathways of CO and CN- to a number of observed hydrocarbon products. We include a pathway for CO reduction by FeMoco and FeVco in Figure 14. Our results indicate that the active sites for CO and CN- reduction are the under-coordinated sulfur sites found in both variants of nitrogenase cofactors, and that these processes are in direct competition with H2 evolution. Our calculated onset potentials are consistent with experiment,31,32 and give insight into why CN- is more easily reduced than CO and which (electro)chemical steps limit the production of hydrocarbons by nitrogenase cofactors. Furthermore, our results can also explain the low yields of hydrocarbons relative to H2 production, suggesting a poor selectivity of nitrogenase for Fischer-Tropsch synthesis. Our results were recently published in Ref.32.

Figure 13. Kinetic volcano for CO evolution at an overpotential of 0.35 V from the model of Hansen et. al. (2013), shown with the clusters in charge states ranging from -2 (a) to 0 (c). The solid (black) lines represent the scaling relation of transition metals (Hansen et. al. 2013), and the dotted line represents a trend line for scaling exhibited by the cubanes. The cubane scaling deviates significantly from that of the transition metals and falls much closer to the peak of the activity volcano.

15

Manganese Oxide Water Oxidation Catalysts A working device for the reduction of CO2 requires a source of protons and electrons; employing water oxidation, i.e. the oxygen evolution reaction (OER), could serve as the other half-reaction for this purpose. Through this GCEP-funded effort, we aimed to develop MnOx catalysts for OER catalysis using two approaches. The first project involved developing a new and efficient method for synthesizing active supported MnOx catalysts. This work was based off of work done by Gorlin et al.34 who synthesized a highly active thin film of MnOx via anodic electrodeposition and subsequent calcination on a glassy carbon (GC) disk. The ORR and OER activity of this thin film MnOx is superior to that of the best MnOx catalysts reported in literature, and is comparable to that of the best known precious metal catalysts. However, the synthesis process is slow, small-scale, and substrate-selective. Therefore, this procedure is not amenable for commercial use. To overcome these limitations while retaining the excellent ORR activity, we designed a new and efficient synthesis procedure which incorporated MnOx particles onto GC via a classic impregnation technique. This was followed by a high-temperature calcination process that generated a nanostructured morphology for the

Figure 14. Plot of the CO reduction pathways to produce CH4 for the FeMoco (a) and FeVco (b). The black lines represent the free energies for each intermediate, while the red lines indicate the free energies at the specified potentials. These potentials indicate the theoretical overpotentials necessary to make the pathways exergonic and compare well with experiment. Both (a) and (b) are shown for the case of a single unprotonated bridging S site, which we find to be the most active sites for CO and CN reduction.

16

Figure 15. (a) Rotating disk voltammograms displaying oxygen electrode activities of MnOx-GC particles and MnOx thin film. The inset shows an SEM image of MnOx-GC particles which display a nanostructured morphology. (b) Cyclic voltammograms showing oxygen evolution activities of Mn2O3, Co3O4, and Co-MnOx thin films. The inset on the left displays the overpotential required to reach 10 mA/cm2 for the various catalysts, while the inset on the right shows an SEM image of the Co-MnOx catalyst (Co/Mn = 2.2) which has a hollow rod-like morphology.

catalyst. The MnOx-GC particles displayed comparable ORR and OER activity with the MnOx thin films (Figure 15a), and can be easily loaded onto carbon paper substrate for use in fuel cells and electrolyzers.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

-6

-4

-2

0

2

4

6

8

10

12

14(a)

0.1M KOH5 mV/s, 1600 rpm

Cu

rren

t D

ensi

ty (

mA

/cm

2 ge

o)

Potential (V versus RHE)

OER

ORR

MnOx-GC particles MnOx thin film (Gorlin et al) GC particles, calcined

Eo H2O/O2

0.0 0.2 0.4 0.6 0.8

0

5

10

15

20

200 nm

Mn2O

3

Co/Mn 0.95 Co/Mn 2.2 Co/Mn 2.8 Co/Mn 7 Co

3O

4

Cu

rren

t D

ensi

ty (

mA

/cm

2

ge

o)

Potential (V vs Ag/AgCl)

Eo H2O/O2

OER0 1 2 3 4 5 6 7

0.44

0.48

0.52

0.56

0.60

0.64

Co/Mn Mol Ratio in Film

Ove

rpo

ten

tial

fo

r 10

mA

/cm

2

Mn2O3

(b)

To further improve the OER activity, our second project involved adding a Co precursor to synthesize Mn-Co mixed oxides. An active MnOx OER catalyst can be synthesized via cathodic electrodeposition and a subsequent high temperature calcination step. The Mn-Co oxides were generated by adding a Co precursor into the deposition solution. SEM revealed that the Mn-Co mixed oxides have a hollow rod-like morphology which is different from that of the pure Mn oxides and Co oxides. Electrochemical testing showed that the Mn-Co mixed oxides are more active than the pure Mn and Co oxides as seen in (Figure 15b). The most active mixed oxide is that with a Co/Mn ratio of 2.2, which corresponds to MnCo2O4 based on XPS and XRD data. MnCo2O4 requires 470 mV of

17

overpotential to reach 10 mA/cm2, and stability testing showed that it was still active after 1000 cycles. Conclusions In conclusion, we have successfully integrated experimental and theoretical methods to provide insight on the design of better CO2 reduction and OER catalysts. Initially, we designed a custom CO2 electrolysis reactor which increased the sensitivity of product detection by NMR and GC. Using these methods, we conducted a detailed study of CO2 reduction products on Cu, Ni, Pt, Fe, Au, Ag, and Zn. Due to the increased sensitivity of our methods, we were able to detect several previously unreported products, leading to new insights into the reaction mechanism on metal surfaces. Next, we examined sulfide materials such as [4Fe-4S] cubanes, FeS2, FeS, MoS2, and VS2 for their CO2 reduction activity. However, we detected H2 as the only major product and found that some materials were unstable at the negative potential required to reduce CO2. Theoretical efforts have been instrumental in guiding us toward more promising avenues with sulfides. By understanding how biological catalysts can effectively catalyze CO2 reduction or ammonia synthesis, we believe that we can utilize those principles to design solid-state surfaces with similar functionality. In addition to bare surfaces, we studied a PANi modified Pt surface to examine the effect of a conducting polymer on CO2 reduction selectivity and selectivity. We discovered that the modification changed Faradaic efficiencies for CO and formate production. Finally, we have developed new synthesis methods for scalable MnOx and Co-MnOx OER catalysts. The activity of these catalysts suggests that they could be useful as replacements for precious metal OER catalysts. Based on the results obtained through this project, we have established new insights into the reaction chemistry of CO2 reduction with new and exciting avenues ahead. One area of future interest is furthering our mechanistic understanding of CO2 reduction on Cu and examining the catalytic activity and selectivity of more pure metals and alloys. Our ultimate goal is to understand the factors that govern catalytic activity and selectivity for this reaction, and to design better catalysts by engineering the surface chemistry. Bimetallic surfaces and polymer/metal interactions are promising means to achieve these goals. In addition to pursuing metallic systems, continued development of mechanistic understanding of CO2 reduction on the biological catalysts such as nitrogenase and CODH enzymes will be of value, with an aim to apply this knowledge to their cubane counterparts. Characterizing the activity and selectivity in terms of adsorbate binding energies and geometries is essential, seeking favorable deviations from the scaling relations exhibited by transition metal surfaces. When coupled with microkinetic modeling, this will give insight into the design of optimal catalysts. There is much work to be done in this field to realize commercial-scale, renewable electrochemical conversion of CO2 to fuels and chemicals; the knowledge disseminated in this project has helped shaped the field and primed it for advances ahead. We look forward to finalizing the dissemination of these results and to moving forward with new directions ahead.

18

Publications & Presentations 1. Hansen, H. A.; Varley, J. B.; Peterson, A. A.; Norskov, J. K. Journal of Physical Chemistry

Letters 2013, 4, 388. 2. Varley, J. B.; Norskov, J. K. Chemcatchem 2013, 5, 732. 3. T.F. Jaramillo, "Catalyzing chemical transformations in renewable energy: Tailoring

Electrocatalyst Materials for Activity, Selectivity, and Stability," Center for Catalytic Science and Technology at the University of Delaware, Newark, DE, May 2013.

4. T.F. Jaramillo, "Catalyzing key chemical transformations for renewable, sustainable energy," Harvard University Center for the Environment (HUCE), Materials for Energy Seminar Series, Cambridge, MA, May 2013.

5. Kuhl, K.P.; Cave, E.R.; Abram, D.N.; Hatsukade, T.; and Jaramillo, T.F. “Trends in Transition Metal Catalysts for Electrochemical CO2 Reduction,” Materials Research Society (MRS) National Meeting, San Francisco, CA, April 2013.

6. Cave, E.R.; Kuhl, K.P.; Abram, D.N.; Hatsukade, T.; and Jaramillo, T.F. “Electrochemical Reduction of CO2 on Metal Surfaces with a Focus on Formate Production,” Invited talk at the NASA-Ames Research Center, Moffet Field, CA., March 2013.

7. T.F. Jaramillo, "Developing Electrocatalysts for the Synthesis of Renewable Fuels," Lawrence Berkeley National Laboratory, Berkeley, CA, December 2012.

8. T.F. Jaramillo, "Tailoring electrocatalyst materials to enhance activity, stability, and selectivity for key energy conversion reactions," University of New Mexico, Dept. of Chemical and Nuclear Engineering, Albuquerque, NM, December 2012.

9. T.F. Jaramillo, "Insights into the electrochemical conversion of CO2 to fuels and chemicals on transition metal surfaces," 2012 Fall Meeting of the Materials Research Society (MRS), Boston, MA, November 2012.

10. T.F. Jaramillo, "Catalyzing chemical transformations in renewable energy: Tailoring electrocatalyst materials for activity, selectivity, and stability," Princeton University, Dept. of Chemical and Biological Engineering Colloquium, Princeton, NJ, November 2012.

11. T.F. Jaramillo, "Electrocatalysis 101," Global Climate Energy Project Annual Research Symposium: Creating a Bright Energy Future, Stanford University, Stanford, CA, October 2012.

12. T.F. Jaramillo, “The electrocatalytic conversion of CO2 to fuels and chemicals,” 222nd Meeting of the Electrochemical Society, PRiME Pacific Rim Meeting on Electrochemical and Solid-State Science, Honolulu, HI, October 2012.

13. T.F. Jaramillo, "Tailoring electrocatalyst materials to enhance activity, stability, and selectivity for key energy conversion reactions," University of California, Berkeley, Department of Chemical and Biomolecular Engineering, Berkeley, CA, October 2012.

14. Kuhl, K.P.; Cave, E.R.; Abram, D.N.; Hatsukade, T.; and Jaramillo, T.F. “Insights into CO2 Electroreduction to Multi-Carbon Products,” 2012 Annual Meeting of the American Institute of Chemical Engineers (AIChE), Pittsburg, PA., October 2012.

15. Cave, E.R.; Kuhl, K.P.; Abram, D.N.; Hatsukade, T.; and Jaramillo, T.F. “Electrochemical Reduction of CO2 on Modified Gold Surfaces,” 2012 Annual Meeting of the American Institute of Chemical Engineers (AIChE), Pittsburg, PA., October 2012.

16. Kuhl, K.P.; Cave, E.R.; Abram, D.N.; Hatsukade, T.; and Jaramillo, T.F. “Insights into the electrochemical reduction of CO2 on metal surfaces,” Global Climate Energy Project (GCEP) Distinguished Student Lecture, GCEP Symposium, Stanford, CA., October 2012.

17. T.F. Jaramillo, "Electrocatalyst development for the synthesis of renewable fuels from water and CO2,"2012 PIRE-CCI Summer School, Dalian, China, September 2012.

18. T.F. Jaramillo, "Solar Fuels by Photocatalysis and Photoelectrochemistry," CASE-CINF Summer School, Reactivity of Nanoparticles for More Efficient and Sustainable Energy Conversion II, Kobaek Strand, DK, August 2012.

19. T.F. Jaramillo, "Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity, Selectivity, and Stability for Energy Conversion Reactions," United Technologies Research Center, Hartford, CT, April 2012.

20. Kuhl, K.P.; Cave, E.R.; Abram, D.N.; Jaramillo, T.F. Energy Environ. Sci. 2012, 5, 7050. 21. Jaramillo, T.F. "Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity,

Selectivity, and Stability for Energy Conversion Reactions," University of California, Santa Barbara, Dept. of Chemistry and Dept. of Chemical Engineering, March 2012.

19

22. Jaramillo, T.F. "Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity, Selectivity, and Stability for Energy Conversion Reactions," Stanford University, Dept. of Mechanical Engineering, High Temperature Gas Dynamics Laboratory Seminar Series, Stanford, CA., March 2012.

23. Jaramillo, T.F. "Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity, Selectivity, and Stability for Energy Conversion Reactions," Massachusetts Institute of Technology (MIT), Energy Initiative Seminar Series, Cambridge, MA., December 2011.

24. Jaramillo, T.F. "Catalyzing the production of clean fuels from renewable energy resources," University of California, Berkeley, Applied Science & Technology, Berkeley, CA., December 2011.

25. Jaramillo, T.F. “Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity, Selectivity, and Stability for Energy Conversion Reactions,” California Institute of Technology, Chemical Engineering Department Seminar, Pasadena, CA., October 2011.

26. Jaramillo, T.F. “Tailoring Electrocatalyst Materials at the Nano-Scale: Controlling Activity, Selectivity, and Stability for Energy Conversion Reactions,” University of New Mexico, Chemical and Nuclear Engineering Department Seminar, Albuquerque, NM., October 2011.

27. Jaramillo, T.F.; Kuhl, K.P.; Cave, E.R.; Abram, D.N. “Electrocatalytic Conversion of CO2 to Fuels on Metal Surfaces,” 2011 Annual Meeting of the American Institute of Chemical Engineers (AIChE), Minneapolis, MN., October 2011.

28. Jaramillo, T.F. “Tailoring electrocatalyst materials at the nano-scale: Controlling activity, selectivity, and stability for energy conversion reactions,” Catalysis for Sustainable Energy (CASE) Seminar Series, Technical University of Denmark, Lyngby, DK., August 2011.

29. Jaramillo, T.F. “Semiconductors and catalysts for the production of solar fuels,” Haldor Topsøe Catalysis Forum: Catalysis & Future Energy, Munkerupgaard, DK., August 2011.

30. Jaramillo, T.F. “Tailoring electrocatalyst materials at the nano-scale: Controlling activity and selectivity for energy conversion reactions,” Lawrence Livermore National Laboratory, Livermore, CA., July 2011.

31. Jaramillo, T.F. “Tailoring electrocatalyst materials at the nano-scale: Controlling activity and selectivity for energy conversion reactions,” Global Climate Energy Project (GCEP) Distinguished Lectureship Series, Exxon Mobil Corporation, Clinton, NJ., June 2011.

32. Jaramillo, T.F. “Tailoring electrocatalyst materials at the nano-scale: Controlling activity and selectivity for energy conversion reactions,” Global Climate Energy Project (GCEP) Distinguished Lectureship Series, General Electric Corporation, Niskayuna, NY., June 2011.

33. Jaramillo, T.F. “Tailoring electrocatalyst materials at the nano-scale: Controlling activity and selectivity for energy conversion reactions,” Global Climate Energy Project (GCEP) Distinguished Lectureship Series, Schlumberger, Inc. Cambridge, MA., June 2011.

34. Jaramillo, T.F. “Nanomaterials for efficient chemical transformations in energy conversion reactions,” Santa Clara University, Department of Chemistry, Santa Clara, CA., May 2011.

35. Jaramillo, T.F. “Nanostructured Catalysts for Chemical Transformations in Energy,” Stanford University Energy & Environment Affiliates Program, Stanford, CA., May 2011.

36. Jaramillo, T.F. “Electrocatalytic conversion of CO2 to fuels on metal surfaces,” Materials Research Society (MRS) National Meeting, San Francisco, CA., April 2011.

37. Cave, E.R.; Kuhl, K.P.; and Jaramillo, T.F. “Surface Modification of Gold for CO2 Electrochemical Reduction,” 2011 Electrochemical Society Fall Meeting, Boston, MA., October 2011.

38. Kuhl, K.P.; Cave, E.R.; and Jaramillo, T.F. “Electrocatalytic Conversion of CO2 to Fuels on Metal Surfaces,” 2011 Electrochemical Society Fall Meeting, Boston, MA., October 2011.

39. Abram, D.N.; Vezie, M.; and Jaramillo, T.F. “Conductive Polymer - Metal Nanoparticle Composites for Electrocatalytic Reduction Reactions,” 2011 Electrochemical Society Fall Meeting, Boston, MA., October 2011.

40. Abram, D.N.; Kuhl, K.P.; Cave, E.R.; and Jaramillo, T.F. “Carbon Dioxide Electrochemical Reduction on Metals,” Fall 2011 GCEP Symposium, Stanford, CA.

41. Kuhl, K.P.; Cave, E.R.; Abram, D.N.; and Jaramillo, T.F. “Carbon Dioxide Electroreduction on Copper,” Fall 2011 GCEP Symposium, Stanford, CA.

42. Ng, J; Baeck, S; and Jaramillo, T.F. “MnOx Electrocatalyst Development for ORR/OER,” 2011 GCEP Seventh Annual Research Symposium, Stanford, CA.

20

43. Jens Nørskov: Nanostructured Catalysts for Conversion of Chemical Energy, Gordon Research Conference: Nanomaterials for Applications in Energy Technology, Ventura, USA, February 2013

44. Jens Nørskov: Tailoring Surface Chemical Properties Using Electronic Structure Theory, APS March Meeting, Boston, USA, February 2012, 243rd ACS National Meeting, San Diego, USA, March 2012

45. Jens Nørskov: Activity descriptors for CO2 electroreduction to methane on transition-metal catalysts, 243rd ACS National Meeting, San Diego, USA, March 2012

46. Jens Nørskov: Catalysis for sustainable energy, 243rd ACS National Meeting, San Diego, USA, March 2012

47. Jens Nørskov: Transition metal surface catalysis, 2012 Summer School on Reactivity of Nanoparticles for More Efficient and Sustainable Energy Conversion – II, Skaelskor, Denmark, August 2012

48. Jens Nørskov: Concepts and trends in surface reactivity, CAMD Summer School: Electronic Structure Theory and Materials Design, Kongens Lyngby, Denmark, August 2012

49. Jens Nørskov: (Photo-)electro-catalytic fuel production, The 63rd Annual Meeting of the International Society of Electrochemistry: Electrochemistry for Advanced Materials, Technologies and Instrumentation, Prague, Czech Republic, August 2012

50. Jens Nørskov: Understanding catalytic and electro-catalytic CO2 reduction, Gordon Research Conference on Chemical Reactions at Surfaces, Ventura, USA, February 2011

51. Jens Nørskov: Catalysis for Sustainable Energy, 22nd North American Catalysis Society Meeting, Detroit, USA, June 2011

52. Jens Nørskov: Catalysis for Sustainable Energy, DOE SCGF 2011 Annual Research Meeting, Oak Ridge, TN, USA, July 2011

53. Jens Nørskov: From descriptive to predictive models of surface reactions, 100 year Anniversary Symposium for the Fritz-Haber Institute, Berlin, Germany, October 2011

54. Jens Nørskov: (Photo-)electro-catalytic CO2 Reduction, MRS 2011 Fall Meeting, Boston, USA, Nov 2011

55. Jens Nørskov: Electrochemical CO2 Reduction: Theoretical Investigations, MURI Annual Review Meeting, University of California San Diego, USA, December 2011

56. Jens Nørskov: Adsorbate-surface interactions and Concepts and trends in surface reactivity, CAMD Summer School, Lyngby, Denmark, August 2010 Jens Nørskov: Catalysis for sustainable energy, ACS 240th National Meeting, Boston, USA, August 2010

57. Jens Nørskov: Fuels from sunlight – a challenge to electronic structure theory, Psi-k 2010 Conference, Berlin, Germany, September 2010

58. Jens Nørskov: Fuels from Sunlight: The Role of Theory, Symposium honoring Juergen Hafner, University of Vienna, Austria, September 2010

59. Jens Nørskov: Catalysis for sustainable Energy, Welch Foundation 2010 Conference on Chemical Research: Green Chemistry and Sustainable Energy, Houston, USA October 2010

60. Jens Nørskov: Understanding CO2 electrochemical CO2 reduction, Fall Symposium on Future Directions in CO2 Conversion Chemistry, Princeton University, USA, October 2010

Manuscripts in preparation, to be submitted shortly after end of project

1. Kuhl, K.P.; Hatsukade, T.; Cave, E.R.; Abram, D.N.; Kibsgaard, J.; Jaramillo, T.F. "Trends in transition metals for the electro-reduction of CO2 to methane and methanol," In Preparation.

2. Ng, J; Gorlin, Y; and Jaramillo, T.F. “Nanostructured particles of manganese oxide for ORR in alkaline-based fuel cells,” In Preparation.

3. Cave, E.R; Kuhl, K.P; Abram, D.N.; and Jaramillo T. F. “Electrochemical Reduction of Carbon Dioxide on Gold,” In Preparation.

4. Abram, D.N.; Kuhl, K.P.; Cave, E.R.; and Jaramillo, T.F. “Electrochemical Reduction of CO2 on Platinum and Polyaniline-Coated Platinum and Modification of Product Distribution,” In Preparation.

21

5. Hatsukade, T.; Kuhl, K.P.; Cave, E.R.; Abram, D.N.; and Jaramillo, T.F. “Insights into the Electrocatalytic Reduction of CO2 on Metallic Silver Surfaces through Potential-Dependent Investigation,” In Preparation.

6. Kuhl, K.P.; Hansen, H.; Cave, E.R.; Abram, D.N.; Hatsukade, T.; Nørskov, J.K., Jaramillo, T. F. “Trends in electrochemical CO2 reduction to CO activity”, In Preparation.

7. Hansen, H.; Kuhl, K.P.; Montoya, J.; Chi, C.; Cave, E.R.; Abram, D.N.; Hatsukade, T.; Jaramillo, T.F.; Nørskov, J.K. “Understanding methane/methanol selectivity on CO2 electroreduction”, In Preparation.

References (1) Vesborg, P. C. K.; Jaramillo, T. F. RSC Advances 2012, 2, 7933. (2) Delacourt, C.; Ridgway, P. L.; Newman, J. J. Electrochem. Soc. 2010, 157, B1902. (3) Delacourt, C.; Newman, J. J. Electrochem. Soc. 2010, 157, B1911. (4) Durand, W. J.; Peterson, A. A.; Studt, F.; Abild-Pedersen, F.; Norskov, J. K. Surf. Sci. 2011, 605,

1354. (5) Peterson, A. A.; Norskov, J. K. Journal of Physical Chemistry Letters 2012, 3, 251. (6) Tang, W.; Peterson, A. A.; Varela, A. S.; Jovanov, Z. P.; Bech, L.; Durand, W. J.; Dahl, S.;

Norskov, J. K.; Chorkendorff, I. Phys. Chem. Chem. Phys. 2012, 14, 76. (7) Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo, T. F. Energy Environ. Sci. 2012, 5, 7050. (8) Schouten, K. J. P.; Kwon, Y.; van der Ham, C. J. M.; Qin, Z.; Koper, M. T. M. Chem. Sci. 2011, 2,

1902. (9) Rosen, B. A.; Salehi-Khojin, A.; Thorson, M. R.; Zhu, W.; Whipple, D. T.; Kenis, P. J. A.; Masel,

R. I. Science 2011, 334, 643. (10) Morris, A. J.; McGibbon, R. T.; Bocarsly, A. B. ChemSusChem 2011, 4, 191. (11) Cole, E. B.; Lakkaraju, P. S.; Rampulla, D. M.; Morris, A. J.; Abelev, E.; Bocarsly, A. B. J. Am.

Chem. Soc. 2010, 132, 11539. (12) Lee, C. C.; Hu, Y. L.; Ribbe, M. W. Science 2010, 329, 642. (13) Surendranath, Y.; Kanan, M. W.; Nocera, D. G. J. Am. Chem. Soc. 2010, 132, 16501. (14) Suntivich, J.; May, K. J.; Gasteiger, H. A.; Goodenough, J. B.; Shao-Horn, Y. Science 2011, 334,

1383. (15) Yeo, B. S.; Klaus, S. L.; Ross, P. N.; Mathies, R. A.; Bell, A. T. ChemPhysChem 2010, 11, 1854. (16) Hori, Y.; Murata, A.; Takahashi, R. J. Chem. Soc. Far. Trans. I 1989, 85, 2309. (17) Hori, Y.; Takahashi, I.; Koga, O.; Hoshi, N. Journal of Physical Chemistry B 2002, 106, 15. (18) Le, M.; Ren, M.; Zhang, Z.; Sprunger, P. T.; Kurtz, R. L.; Flake, J. C. J. Electrochem. Soc. 2011,

158, E45. (19) Deutschmann, O.; Knözinger, H.; Kochloefl, K.; Turek, T. In Ullmann's Encyclopedia of

Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: 2000. (20) Peterson, A. A.; Abild-Pedersen, F.; Studt, F.; Rossmeisl, J.; Norskov, J. K. Energy Env.Sci. 2010,

3, 1311. (21) Parkin, A.; Seravalli, J.; Vincent, K. A.; Ragsdale, S. W.; Armstrong, F. A. J. Am. Chem. Soc.

2007, 129, 10328. (22) Chen, Z. B.; Cummins, D.; Reinecke, B. N.; Clark, E.; Sunkara, M. K.; Jaramillo, T. F. Nano

Letters 2011, 11, 4168. (23) Jaramillo, T. F.; Jorgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I. Science

2007, 317, 100. (24) Hori, Y. W., H.; Tsukamoto, T.; Koga, O. Electrochim. Acta 1994, 39, 1833. (25) Brisard, G. M. C., A.P.M.; Nart, F.C.; Iwasita, T. Electrochem. Comm. 2001, 3, 603. (26) Kung, Y.; Doukov, T. I.; Seravalli, J.; Ragsdale, S. W.; Drennan, C. L. Biochemistry 2009, 48,

7432. (27) Parkin, A.; Seravalli, J.; Vincent, K. A.; Ragsdale, S. W.; Armstrong, F. A. J. Am. Chem. Soc.

2007, 129, 10328. (28) Peterson, A. A.; Norskov, J. K. J. Phys. Chem. Lett. 2012, 3, 251. (29) Abild-Pedersen, F.; Greeley, J.; Studt, F.; Rossmeisl, J.; Munter, T. R.; Moses, P. G.; Skulason, E.;

Bligaard, T.; Norskov, J. K. Physical Review Letters 2007, 99. (30) Hansen, H. A.; Varley, J. B.; Peterson, A. A.; Norskov, J. K. Journal of Physical Chemistry

Letters 2013, 4, 388.

22

(31) Lee, C. C.; Hu, Y.; Ribbe, M. W. Angewandte Chemie-International Edition 2012, 51, 1947. (32) Varley, J. B.; Norskov, J. K. Chemcatchem 2013, 5, 732. (33) Hu, Y.; Lee, C. C.; Ribbe, M. W. Science 2011, 333, 753. (34) Gorlin, Y.; Jaramillo, T. F. J. Am. Chem. Soc. 2010, 132, 13612. Contacts Thomas F. Jaramillo: jaramillo@stanford.edu Jens K. Nørskov: norskov@stanford.edu Etosha R. Cave: cave@stanford.edu Kendra Kuhl: kpkuhl@stanford.edu David Abram: dabram@stanford.edu Desmond Ng: desmond1@stanford.edu

Toru Hatsukade: thatsuka@stanford.edu Jakob Kibsgaard: kibs@stanford.edu Samuel Fleischman: fleisch@stanford.edu Christopher Hahn: cjhahn@stanford.edu Joel Varley: jvarley@stanford.edu