Accelerating the Catalyst Development Cycle: Integrating Predictive Computational Modeling, Tailored Materials Synthesis, and in situ Characterization Capabilities through the ChemCatBio Consortium—Video Text Version

This is the text version for the Accelerating the Catalyst Development Cycle: Integrating Predictive Computational Modeling, Tailored Materials Synthesis, and in situ Characterization Capabilities through the ChemCatBio Consortium webinar.

[Music]

>>Susan Habas: I'm Susan Habas.

>>Carrie Farberow: I'm Carrie Farberow.

>>Jesse Hensley: I'm Jesse Hensley.

>>Matthew Yung: I'm Matthew Yung.

>>Lu Wang: I'm Lu Wang.

>>Josh Schaidle: My name is Josh Schaidle.

>>Emily Roberts: I'm Emily Roberts.

>>Jae-Soon Choi: I am Jae-Soon Choi.

>>Dan Ruddy: I'm Dan Ruddy.

>>Ted Krause: My name is Ted Krause.

>>Fred Baddour: I'm Fred Baddour.

>>Kinga Unocic: My name is Kinga Unocic.

>>Kurt Van Allsburg: I'm Kurt Van Allsburg.

>>Josh Schaidle: Recent reports out of the Bioenergy Technologies Office within the Department of Energy suggest that by the year 2030, there could be over a billion tons of biomass that could be sustainably sourced every year in the United States. The challenge becomes how do we mobilize that domestic resource and convert it into fuels and chemicals that can be used in today's society? And the biggest challenge associated with that is that conversion step, and if you look at existing processes in today's economy, 85 percent of those processes rely on a catalytic step. And the catalysis market across the globe is about $20 billion per year.

And so we don't expect this to change moving into using biomass as a feedstock for producing fuels and chemicals, but the challenges are likely going to change for those catalytic steps. My name is Josh Schaidle. I'm the director of the Chemical Catalysis for Bioenergy Consortium. I also lead the Thermochemical Biomass Conversion Platform within the National Bioenergy Center at NREL.

The Chemical Catalysis for Bioenergy Consortium, or ChemCatBio for short, was started back in the fall of 2016, and the focus of the consortium is to overcome catalysis challenges for biomass conversion applications. The goals at ChemCatBio are really two-part. One is specifically to bridge that gap that currently exists between foundational science and applied engineering. The second major component of ChemCatBio is accelerating the catalyst and process development cycle.

In existing literature and knowledge, it takes about 15 to 20 years to get a chemical process from initial ideation and development out into the market. One of the other major problems that ChemCatBio hopes to address is in regards to the complex nature of the feedstocks. You're often dealing with multifunctional molecules made up in a larger mixture of sometimes tens, even hundreds of molecules in a process stream. And you have to design a catalyst to address all of those multifunctionalities within those molecules across that mixture in order to control selectivity towards your desired products.

>>Susan Habas: Today, we'll highlight transition metal carbides as a class of materials that has the potential to meet the critical goals of low cost and superior performance under severe conditions. We'll use this promising but complex material system to demonstrate how the integration of predictive computational modeling, tailored material synthesis, and in situ characterization capabilities within ChemCatBio, is accelerating catalyst development.

When we started to think about designing catalysts for different biomass conversion processes, taking into account things like cost, stability in harsh chemical environments, and then also the ability to selectively perform transformations that are of interest to us, a class of materials that came to mind were metal carbides, and this is something that our team has a lot of expertise in.

Well, metal carbides are essentially metal lattices in which carbon is incorporated into the interstitial sites within the parent lattice. And as a result of this carbon in the lattice, you get very different properties from the parent metal. Particular properties that we're interested in are the fact that by putting the right amount of carbon in the right structure, you can start to get behavior that is a lot like a platinum group metal.

At the same time, you get a multifunctional catalyst as a result of incorporating carbon in the lattice, and this allows you to have hydrogenation sites and acidic sites in close proximity with one another.

We commonly make metal carbide catalysts by starting with a metal oxide precursor and heating it up in the presence of a reactive carbon source. We most often use a methane hydrogen mixture and then heat it up to temperatures as high as 590 degrees for an extended time.

Recently, we've been interested in looking at lower severity solution phase synthesis routes to make metal carbides, and these are of interest to us because it affords you a lot of control over the resulting metal carbide structure. And at the same time, you can access these metal carbides at significantly lower temperatures, as low as even 250 degrees.

We're really excited about some of the solution synthesis routes that are enabling us to make nanoscale metal carbide catalysts. Metal carbides are promising catalysts for many biomass-related conversion processes. But at the same time, the degree of tune-ability that enables this promise also makes it difficult to identify specific catalyst structures that promote targeted chemical transformations.

One way that we handle this complexity is to engage our computational modeling team early in the process to help identify key phases, compositions, and surface structures to increase catalyst selectivity.

>>Carrie Farberow: The challenges associated with modeling carbide catalysts include things like the fact that the choice of surface to model isn't straightforward. There could be many possible exposed surface facets or terminations on a carbide particle, and our understanding of the relative stability of these isn't as comprehensive as it is for transition metals.

Hi. I'm Carrie Farberow. I'm a research scientist at the National Renewable Energy Lab, and I use computational modeling to understand how chemical reactions occur on catalyst surfaces.

NREL currently has an HPC system called Peregrine to support research simulations and visualization. Peregrine is comprised of about 2,500 nodes, or more than about 60,000 cores, and has a peak performance of 2.3 petaflops.

We use atomic scale modeling combined with experimental kinetics and catalyst characterization to understand how reactions occur on catalyst surfaces. We use that information to understand what's limiting the catalyst activity or selectivity, and then propose modifications to the catalyst formulation to try and improve performance.

Once we've identified computationally interesting candidates, we need to make these a reality, using innovative synthetic routes that we've developed within ChemCatBio.

>>Dan Ruddy: I'm Dan Ruddy, and I'm a senior scientist at NREL, specializing in the design and synthesis of catalyst materials. In addition to standard bulk synthetic routes where the metal carbide is synthesized at very high temperatures, we're also interested in low-temperature solution synthetic routes to metal carbide nanoparticles. The high level of control afforded by the mild solution synthesis conditions enables us to create model catalyst systems with key features that have been predicted by our computational team.

We are particularly interested in using tailored molecular precursors that have built-in functionality to help us achieve targeted catalyst structures. These targeted functionalities can include preformed metal-carbon bonds with specific metal-to-carbon ratios, and ligands that promote decomposition of the precursor at temperatures relevant for metal carbide nanoparticle formation. The ligands also facilitate solubility in a solvent that can accommodate the desired decomposition temperature. We also introduce additional organic ligands during synthesis that can interact with the growing catalyst nanoparticle to help control attributes such as size, shape, and [inaudible].

Some of the metal carbide nanoparticles that we make are in the single nanometer range, such as two to four nanometers. We can support these particles on different metal oxides or carbons to give us catalysts with very high active surface areas.

>>Susan Habas: Characterization plays a critical role in all aspects of catalyst development. During the early stages of the cycle, we seek feedback on how effectively we've been able to translate computationally predicted attributes into a model catalyst system, so that we can rapidly iterate on catalyst synthesis.

As the development cycle progresses, our focus shifts to the evolution of the catalyst structures under realistic pre-treatment, reaction, and regeneration conditions, so that we can accurately describe and improve upon the real working catalyst system.

>>Ted Krause: Welcome to Argonne National Laboratory. My name is Ted Krause. I'm the principal investigator for the X-ray characterization studies being conducted on catalysts as part of the advanced catalyst synthesis and characterization team, which is an activity under ChemCatBio.

Here at Argonne National Lab, I'm sitting at the Advanced Photon Source, also known as the APS, which is the brightest X-ray source in the Western hemisphere. We use X-rays to understand the chemical and physical properties of the catalysts. In X-ray spectroscopy, we use two techniques, X-ray near edge spectroscopy, and X-ray absorption fine-edge structural analysis. This provides insight into chemical and physical properties of the catalyst, such as the catalyst oxidation state, the particle size, and the nearest neighbors around the atom of interest.

But we also do in situ studies, and this gives the real true picture of how the catalyst is performing. We also not only look at those properties that control catalyst activity and selectivity, but we look at the changes in the catalyst properties as the catalyst deactivates. We then use this information to understand what is driving catalyst deactivation. We then can pass this information along to NREL, which helps them to develop synthesis techniques that then not only optimize the catalyst's performance, but also increases catalyst lifetime and durability.

Operando studies involve evaluating the catalyst under working conditions. In this case, we roll in a microreactor. We actually insert it into the beam line, and the catalyst is operating under the real-world reaction processes that you would find in a commercial biofuels production. By studying the catalyst under the real working conditions, we have a better picture of those chemical and physical properties that are influencing catalyst performance and activity.

In situ and operando techniques allow us to monitor how the catalyst changes with time, and to understand which factors are governing the deactivation. Then using this information, we can develop regeneration processes, such as steam stripping to remove carbon deposition, a major cause of catalyst deactivation in biofuel processing.

So by understanding what the effect of the regeneration process is, again, we can help tailor synthesis processes to address any issues that the regeneration process might create.

>>Susan Habas: Now that we have a better idea of what the working catalyst looks like, we can combine this information with detailed catalytic evaluation to understand how specific catalyst features influence the catalytic performance. For metal carbides, model compounds can be used to quantify the relative rates of hydrogenation and deoxygenation. This information can be used to design a second generation of catalysts, with a tailored ratio of active sites, to shift the product slate to desired chemicals or fuel molecules.

>>Jesse Hensley: My name is Jesse Hensley. I'm the group manager for the Heterogeneous Research and Development Group in the National Bioenergy Center at NREL. The FSCL allows real time analysis of reactions under controlled conditions. Heterogeneous catalysts are tested in any of our six custom-engineered reactors, each designed to answer specific questions about catalyst performance.

Support equipment, like gas chromatographs, streamline experiment setup and product analysis, and a fully integrated control system allows around the clock operation and continuous collection of process data. We can perform hydrogenation, oxidation, carburization, deoxygenation, dehydration, dehydrogenation, cyclization, aromatization, cracking, carbonylation, and isomerization reactions of gases and liquids using solid catalyst materials at pressures up to 100 atmospheres and temperatures up to 1,000 C.

>>Susan Habas: Catalyst regeneration requirements can constitute a significant portion of the final cost of fuels or chemicals. Consequently, we've learned that it's important to begin investigating effective regeneration procedures early in the catalyst development cycle. We've used a number of operando methods to investigate the deactivation modes of metal carbide catalysts during reaction, and treatments to regenerate the active carbide structure as quickly as possible under conditions as close to those used during reaction as possible.

>>Matthew Yung: Hi. I'm Matthew Yung. I'm a research scientist specializing in materials characterization and catalysis. Catalysts are commonly regenerated through a high-temperature oxidation, which can remove carbon deposits that can block catalysts' active sites.

One of the challenges in regenerating metal carbide catalysts is that they oxidize under very mild conditions and at low oxygen concentrations, and we see this using in situ X-ray diffraction. In situ X-ray diffraction is a technique that allows us to look at the crystalline structure of a catalyst while it's undergoing chemical reactions. We're using operando laser Raman spectroscopy to study regeneration protocols for carbide catalysts.

Operando laser Raman spectroscopy is a really cool technique that enables us to probe the catalysts during the reaction while we have a combination of gas chromatography and mass spectrometry to enable us to measure the catalyst's activity. This allows very good correlation between certain physical or molecular attributes of the catalyst and the activity of the catalyst for a chemical reaction, and to simultaneously see the formation of carbon species during the reaction and the removal of carbon species during the regeneration.

This allows us to select the proper operating temperatures and conditions to enable long-term activity of a catalyst.

>>Susan Habas: In addition to the catalyst characterization techniques we've described that assess changes to the overall catalyst structure, it's important to understand what's happening to the catalyst locally. Electron microscopy techniques in particular help to provide a clear picture of how the active metal carbide surfaces evolve during regeneration.

>>Kinga Unocic: At Oak Ridge National Lab, we have a number of aberration-corrected microscopes, in conjunction with EDX or EELS allow us to characterize the catalyst, and it's very crucial in catalyst development.

Using spectroscopy analysis, it allows us to identify the distribution of different elements throughout the catalyst particle. It allows us to identify the clustering of the elements, what are the precipitates made of in the catalyst, and identify the sub-shell formation. STEM and higher STEM imaging allow us to capture the morphology, size, and the structure of the catalyst. Highest STEM imaging also allows us to identify and look at atomic level of the different interfaces, and also localize the coke formation, and identify it using EELS. Therefore, STEM imaging and spectroscopy analysis provide important insights into synthesis and computational modeling team at NREL for future catalyst development and development of next generation of the catalyst.

In situ closed gas cell reaction allow us to study a range of different materials, and their response to the temperature and the reactive gases. In situ STEM allow the reactions to be carried at temperatures up to 1,000 degrees C, and atmospheric pressures. In situ STEM at ORNL uses proto-chip system that includes computer control, gas delivered in a manifold, and in situ gas reaction TEM, atmosphere holder. Recently, we performed additional upgrades to the in situ system, and we add a residual gas analyzer that is connected to the stainless steel capillary at the end of the atmosphere TEM holder.

This allows us the measurement of the gas on the exit side, and quantitative measurements. Recently, it's been used for studying ORNL developed molybdenum carbides, where we try to regenerate the spent catalyst. It was tested bio-oil upgrading reactor at PNNL. With capability of studying the reaction of the catalyst at different magnification, we can not only observe the types and the location of the coke formation, but also, we can study the reactivity.

For example, in molybdenum carbides, we found that the surface coke was very reactive, and was removed somewhere between 200 degrees C and 370 degrees C. This is very important, because cokes deactivate the catalyst and decreases the life of the material.

What is really important about this in situ STEM microscopy is that molybdenum carbide can be regnerable, and also, that this type of reaction can be missed by other techniques, since it is performed at such a small-scale level.

>>Susan Habas: As the catalyst development cycle continues, key catalysts are selected to move forward for larger scale evaluation and process developments, while continued advancements are sought for lab scale formulations through the iterative process of computational prediction, tailored synthesis, advanced characterization, and catalytic evaluation.

>>Fred Baddour: I'm Fred Baddour, and I'm a scientist at the National Renewable Energy Lab, and I specialize in inorganic synthesis and nanomaterials. When we start to think about scaling up a promising catalyst material, the most common approach that comes to mind is simple volumetric scaling. This works well with more traditional catalyst synthesis techniques, such as incipient wetness impregnation, for which engineering scale equipment and processes already exist.

The solution phase synthesis of nanoparticle catalysts, on the other hand, pose a unique set of challenges that make volumetric scaling difficult. During the synthesis of metal carbide nanoparticles, the molecular precursors can undergo sublimation. At the same time, volatile byproducts are being produced that can explosively vaporize. Additionally, the key features of the nanoparticles that we are trying to control are highly sensitive to thermal and chemical inhomogeneities. These inhomogeneities can arise due to the increased solvent volume and changes in heating rate when transitioning to larger scales.

Addressing these challenges really requires us to look at alternative approaches to producing large quantities of nanoparticles with the critical properties that have been demonstrated and refined during the catalyst development cycle. This is why we're excited about our partnership with the researchers at the University of Southern California who have been scaling up the synthesis of nanoparticles using continuous flow approaches.

>>Lu Wang: Hi. I'm Lu Wang. I'm working with Noah Malmstadt in chemical engineering in University of Southern California. We have been trying to fabricate high-throughput, high-temperature reactors for nano-fabrication in continuous flow.

>>Emily Roberts: Continuous flow microfluidic reactors are powerful tools for synthesizing chemicals and materials. In particular, for manufacturing colloidal nanoparticle catalysts, there are benefits of doing syntheses in microflow, such as superior heat and mass transport, reproducibility, automatability, and reduced environmental and safety risk.

Hi. I'm Emily Roberts, and I'm a graduate student at the University of Southern California. I work for Professor Richard Brutchey in the chemistry department, and in collaboration with Professor Noah Malmstadt and graduate student Lu Wang, we work on developing high-temperature reactors for the synthesis of nanoparticle catalysts.

In a typical molybdenum carbide synthesis, we prepare our precursor solutions and then put them into syringes. The syringes are then placed into syringe pumps and flowed at different flow rates, depending on the desired residence time. Upon entering the reactor, it is heated to a desired temperature, in this case, upwards of 300 degrees, and this is where the reaction takes place.

>>Lu Wang: With this synthesis, we have two major challenges. One is the temperature. This temperature is well above the traditional microfluidic reactions, which is around 300 degrees, and the traditional materials, such as PET and Teflon, is no longer suitable for this reaction.

>>Emily Roberts: Therefore, we fabricated a helical millifluidic reactor out of borosilicate glass that can withstand the temperatures necessary for carbide nanoparticle synthesis. Using this reactor, we can safely intensify the process by a factor of eight, which would not be possible in the corresponding lab batch reaction. This allowed us to make over 18 grams of molybdenum carbide nanoparticles in continuous flow in a day, compared to 100 milligrams in a single-batch reaction. This equates to more than 370 grams of 5 weight percent catalyst in a 24-hour period.

>>Kurt Van Allsburg: Non-traditional scaling approaches, like the continuous flow synthesis of carbide nanoparticles being pioneered at USC, are helping to bridge the gap between the lab-scale batch syntheses and engineering scale application of these materials. But how do we know when we've made a breakthrough in cost?

I'm Kurt Van Allsburg. I'm a scientist at NREL, and I study catalytic materials and their electrochemical properties, their synthesis, and especially their cost.

Bringing any technology to market requires considering both performance and cost, but unlike performance for catalysts, which can easily be interrogated via various methods in the lab, cost is not something that most scientists have the tools to answer.

Understanding the cost of a material depends on a lot of information that researchers don't have good access to. We've developed a tool to help researchers easily get cost insight on their catalytic materials based only on the lab-scale synthesis characteristics.

Starting a cost estimate for a new catalyst using the tool is very easy. You simply select a process template that most closely approximates your synthesis method, and populate it with your reagents and associated lab -\\-scale quantities. We've incorporated powerful visualizations that make it easy to understand and to use your data to make informed decisions about catalyst materials and synthesis processes.

Because the tool is easy to use, it can provide insights starting at the earliest stages of the catalyst development cycle. The Excel spreadsheet and web tool will be free and publicly available in fall of 2018.

>>Susan Habas: Another important factor in the successful scaling of a metal carbide catalyst is the creation of a technical catalyst that has the appropriate properties, such as particle size, porosity, and attrition resistance, for integration into an engineering scale process. This is something we're investigating through a number of unique approaches at Oak Ridge National Laboratory.

>>Jae-Soon Choi: The challenge that we are facing in developing biomass conversion catalysts is difficulty in transforming novel catalytic materials, usually powder, into engineer the catalyst, such as pellets, beads, [inaudible] required for reactor testing or implementation at relevant scales.

Because biomass conversion conditions are very different from material processing, for example, there is a large amount of liquid water, so conventional catalysts [inaudible] method are often not suitable for bioenergy catalysts, such as molybdenum carbide. So CCB is trying to fill this gap by providing catalyst [inaudible] solutions tailored to bioenergy applications.

So we have extensive capabilities in [inaudible] chemistry, either [inaudible]. This chemistry is highly scalable and also tunable, and works well with many materials that are of interest to us. For example, [inaudible] with titanium. Then we can also make all the traditional metal carbides, one good example is nickel [inaudible] carbide that we recently applied to bio-oil [inaudible].

Our approach includes inexpensive precursors, which we [inaudible] transition [inaudible] microscopic catalytic structures. So basically, we can create [inaudible] the catalysts using water stable catalytic materials only without binders, which can greatly help enhance catalyst stability and activity.

>>Susan Habas: Once we have a technical catalyst formulation that meets the specifications required for engineering-scale evaluation, we can take advantage of a variety of industry-relevant reactor capabilities available within ChemCatBio to better understand the interplay between catalysts and catalytic process at a larger scale.

>>Kim Magrini: My name's Kim Magrini, and I manage the thermochemical process development group at NREL, and my background is material science and catalyst development. And one of the components of our research program is called the Davison Circulating Riser system, fondly nicknamed the DCR. And what that is, is an industry accepted fluid cat cracking unit that allows us to evaluate catalyst performance for taking biomass vapors and liquids to hydrocarbon fuels.

The DCR unit uses two kilos of catalyst, and we can feed to the pyrolizer that's coupled with our DCR anywhere from 1 to 5 kilograms an hour of biomass that gets vaporized during fast pyrolysis. A slip stream of that vapor is then fed to the DCR unit for upgrading to hydrocarbon fuels.

The DCR we primarily use to take industrial catalysts, or we use modified variants of those industrial catalysts, and we test the performance at a process relevant scale. And what we do is understand how biomass impacts fluid cat cracking, and also the changes that are induced on the catalyst pre- and post-use. And so what we do is directly transferable to industry for their use, and we as well can provide testing for industry, if they have catalysts to test, or if they want to evaluate their materials with various biomass feedstocks.

>>Susan Habas: I hope we've been able to convey how ChemCatBio works together to predict key catalyst features, translate these active features to laboratory-scale catalyst systems, and understand how these features evolve under real working conditions. This enables us to design the next generation of catalysts that can then be transitioned to engineering-scale catalyst formulations for evaluation under industry relevant conditions. We think that this approach will help to bridge the gap between fundamental catalytic research and industrial implementation to accelerate the adoption and promising but complex catalyst materials, such as the metal carbides.

This is just one of the many examples of how ChemCatBio is solving catalyst and process development challenges associated with emerging bioenergy applications, and we're always looking for new opportunities to collaborate.

>>Josh Schaidle: There are numerous ways that academia and industry partners can work with ChemCatBio. These include engaging on existing projects and even supplying materials, expertise, knowledge, guidance, even in many cases just playing an advisory role to helping give an industrial context to the ongoing R&D, and helping us to assess our research progress.

In addition, there can be opportunities to pursue funding through the Department of Energy and the Bioenergy Technologies Offices, through funding opportunity announcements or proposal calls that can be held both by the Department of Energy or by ChemCatBio to address specific industry challenges.

Further, there's a number of mechanisms that exist, whether these be cooperative research and development agreements, or technical service agreements, that already are in place within the national lab system, and are available to academic and industry partners to help work with ChemCatBio.

One of the things that we wanted to try to tackle with ChemCatBio is to enable easier interactions for industry and academia with the national lab complex. And so ChemCatBio is set up such that there's a single point of contact. You go directly through our website. It comes directly to myself as the director of ChemCatBio. And I can ascertain what the challenge is being faced, what your needs are as a potential partner, and can direct you to the right person within the national lab complex, within ChemCatBio Consortium, that can help address your challenge.

[Music]

>>Susan Habas: This might be the best thing you've ever filmed, [inaudible].

[Crosstalk]

[Music]

>>Kurt Van Allsburg: Represents a dramatic improvement in the something something something something something. But how do we know when we've made a breakthrough in catalyst cost?

[Music]

>>Matthew Yung: Hi. I'm Matthew Yung, research engineer specializing in catalysis and materials characterization. And welcome to the ChemCatBio Consortium.

[End of Audio]