 Thanks, Richard, for the nice introduction. Hi, everyone. So we are interested in making clean transportation fuels from natural gas, biomass, or even CO2 from the atmosphere. And one promising pathway is through syngas conversion. And we are looking into the catalyst that can convert syngas into higher oxygenates, for example, ethanol. So these higher oxygenates can be used as very high performance and also clean transportation fuels. And also, they can be used as the basic chemical feedstocks for manufacturing consumer products. And rodent-based catalyst is one of the major candidates. One very interesting and important factor is that the support material or the promoters can actually largely determine the catalytic performance. So in heterogeneous catalysis, promoter refers to a substance added into the catalyst to improve its activity, selectivity, or lifetime. And for rodent catalysts, by using different metal oxides as either the support material or the promoters, the activity and selectivity can be dramatically changed. Therefore, our research is focused on understanding how these metal oxide species influence the catalytic performance. We know that in heterogeneous catalysis, the surface properties are crucial. And also, the interface between rhodium and these metal oxide species could play a critical role. Therefore, we want to directly interact. Therefore, we want to directly create such interface layers and control the surface properties. And we do that by using atomic layer deposition. So atomic layer deposition is based on self-saturating layer-by-layer growth model. It can achieve very uniform coding with angstrom level control on high surface area substrates. So we designed two models in order to study the effect of manganese oxide in this case. In the first case, we deposited an ultra-thin layer of manganese oxide by ALD on the silicon oxide substrate. And then we deposited the rhodium nanoparticles. So in this case, the manganese oxide is a support modification layer. In the other case, we deposited rhodium nanoparticle first and then deposited the ALD manganese oxide as an overcoat layer. And I should point out that in this catalyst, the rhodium nanoparticles are deposited by a wire chemistry technique called incipient-vanus impregnation. And besides using ALD to modify the catalyst structure, we also use a conventional method, coimpregnation, to incorporate these manganese promoters. The coimpregnation method is simply done by mixing the rhodium and manganese precursor together into the impregnation solution. This method is simple and effective. However, it doesn't have good control over how much manganese is mixed into the rhodium nanoparticles and how much manganese ends up on the nanoparticle surface. Also, the nanoparticle size can usually be smaller from this coimpregnation method. Therefore, by using ALD, we're hoping to have less variables so that we can focus on the role of these metal oxides without changing other parameters, for example, the size of the nanoparticles or the porosity of the support. Then we took all these different types of catalyst to test in syngas conversion reaction. So this diagram shows the selectivity going into different products. Our desirable higher-oxygenate product is shown in orange. And the method shown in gray is the most competing byproduct, which we want to minimize. With manganese oxide ALD as a support layer, it actually achieved the highest production of higher oxygenates, especially the most important parameter selectivity was significantly improved. In comparison, using manganese oxide as an overcoat layer was not as effective as the support layer. The selectivity improvement was much less. We also observed that actually the coimpregnation catalyst gave also pretty good activity and selectivity. Then we used different types of characterizations to understand how manganese oxide functions in these different types of catalyst. First of all, we used a TEM to characterize the nanoparticle size of rhodium. This is because by changing the nanoparticle size, we can change what types of surface size are exposed. And different types of surface sites can have different intrinsic activity and selectivity. But it turns out that the manganese oxide support layer didn't change the size of rhodium nanoparticles compared to the n-promoted. So this is indicating that the manganese oxide mainly functions by forming the interface sites around the parameter where rhodium nanoparticles are anchored. In comparison, the coimpregnation catalyst results in much smaller nanoparticle size. And this is not surprising because the two precursors were mixed together during the synthesis. So it's very likely that the two species are highly intermixed. From the reaction results, we know that the manganese oxide overcoat layer was not as effective as the support layer. However, it's usually hypothesized that the manganese species present on the rhodium nanoparticle surface should be responsible for the promotion effect. So we were wondering why when we directly deposited manganese oxide on the catalyst surface, the results weren't quite good. Performing XPS, we noticed that after the catalyst absorbs CO, the surface manganese signal largely decreased. Therefore, this is telling us that the manganese oxide overcoat layer is not stable upon CO absorption. Therefore, the promotion effect was not effective. Then we looked into how the carbon monoxide interacts with the catalyst surface using infrared spectroscopy. The most intense peak is CO binds to one rhodium atom. And another major peak occurring around 1900 wave number is the bridge peak, where CO binds to two or more rhodium atoms. So this bridge peak indicates the presence of extended rhodium surface. With manganese oxide as a support layer, besides this regular bridge peak, we also saw another peak showing up at much lower wave number, around 1700 wave number in this case. So this peak is commonly assigned to CO binds to the interface between rhodium and manganese. Since the manganese cation is oxafilic, so it helps pull apart the CO bond and therefore increase the activity. So this feature suggests that we effectively formed interface sites between rhodium and the manganese oxide support without disrupting the extended rhodium surface. Well, on the catalyst with manganese oxide as an overcoat layer, the formation of such interface sites was not obvious. Again, due to the instability of manganese oxide overcoat layer upon CO absorption. For deco impregnation catalysts, actually, we saw a lot of formation of such interface sites. However, the signal from the extended rhodium surface largely diminished. And this picture actually agrees pretty well with the results from TEM showing much smaller nanoparticle sites due to the highly intermixing between rhodium and the manganese. And besides the improved activity from this manganese oxide support layer, we also saw enhanced selectivity towards the higher oxygenates. And there could be two main reasons. One is that during impregnation of the rhodium nanoparticles, some trace amount of manganese species could be transferred onto the nanoparticle surface. And such species should preferentially stay on the stepped or defect sites of the nanoparticles, since these kind of sites are highly active to make methane. So this site blockage or site blockage and deactivation can bring down the methane production. Another reason would be the manganese oxide could change the reaction energetics. Through a collaboration with Professor Jens Norskopf's group on density functional theory calculation, we found that manganese oxide present on the rhodium surface could stabilize the key transition state towards higher oxygenate synthesis. But it doesn't really affect the methane synthesis. Therefore, around the interface between manganese oxide and rhodium, the selectivity towards higher oxygenates could be improved. So in conclusion, by controlling the rhodium catalyst structure using atomic layer deposition, we identified that the rhodium manganese oxide interface sites is responsible for improved higher oxygenate production. And we are also very excited to apply similar strategy to study more heterogeneous catalysis system that are very important for the chemical industry and energy conversion. With that, I'd like to thank my advisor, Professor Stacy Bant, all the Bant group members, and also the fundamental insights provided by Professor Norskopf's group. And also, we want to thank the financial support provided by GSAP. Thank you all for listening. I'd be happy to take questions. Two or three minutes of questions, if anyone has any questions? Over here, yes. Hello. Thank you for that very interesting talk. Manganese and oxygen form at least 30 known binary compounds from manganese 2 up to manganese 7. And some of the common ones you might be forming are manganese oxide with manganese 2, or Bixbyite MN2O3, or Hausmanite MN3O4, and others, which are commonly accessible. They have very different reactivity and also transfer reactions. And so do you actually know what compound you have formed, and is that relevant to the catalysis you're seeing? Yeah, that's a very good point. So from atomic layer deposition, the manganese oxide comes out to be 2 plus. And so we actually haven't performed any operando characterization to look at the manganese oxidation state under reaction condition. But from the literature, there's enough evidence showing that manganese should stay at 2 plus under reaction condition. Although some group did use some calculation to show that it could be possible to form rhodium and manganese metallic alloy, but we don't believe that's the case here. So it will mostly stay as oxidized, probably 2 plus, but we haven't done any in-situ characterization. Actually, the time is up. So let's thank Naya one more time. Thank you, Naya. Thank you.