Ed Wolfrum, Feedstock-Conversion Interface Consortium Principal Investigator: Welcome to the FCIC webinar, “Unveiling Signatures of Feedstock Variability.” My name is Ed Wolfrum, I serve as the principal investigator. We'll get started in just one minute. Thank you all for attending. We still see a number of people still logging on, so we'll give it one more minute and then we'll get started.
Okay, that looks like it's one minute after the hour and again, thanks for everyone for joining. And we'll get started. I suspect we'll have a number more people joining us as we go. But I do want to respect everybody's time, so let's get going.
Again, welcome to the third in a series of webinars about the Feedstock-Conversion Interface, the FCIC.
My name is Ed Wolfrum and I'm fortunate enough to serve as a principal investigator at the FCIC, and I'm glad you all were able to take an hour out of your day to join us. So the FCIC is led by the Department of Energy's Bioenergy Technologies Office, BETO. It's a collaborative effort among researchers from nine different national laboratories to address the challenges posed by feedstock and process variability on buyer refineries.
Last December, I presented an overview of the FCIC, and in February, a couple of researchers talked about their computational modeling work. Those recordings are available on the FCIC website, and we'll get you a link, but you can just Google "FCIC BETO" to get that. Today, and for the folks who missed the first talk, I'm going to spend about a minute on the background of the FCIC before the main event, which is presentations by two FCIC researchers: Dr. Allison Ray of Idaho National Lab and Dr. Bryon Donohoe of the National Renewable Energy Laboratory will be talking about unveiling signatures of feedstock variability.
Because of the world we live in and the uncertainties associated with virtual presentations, we have prerecorded both of those talks, so there will be a bit of a pause as I queue them up. Each talk is about 20 minutes or so.
I do have to read a disclaimer, so please bear with me for a second. This webinar, including all audio and images of participants and presentation materials, may be recorded, saved, edited, distributed, used internally, posted on DOE's website, or otherwise made publicly available. If you continue to access this webinar and provide such audio or image content, you consent to such use by or on behalf of DOE and the government for government purposes and acknowledge that you will not inspect or approve, or be compensated for, such use.
So with that, let's move on to the quick overview. Here's the one-slide guide to the FCIC. Here's my one-minute overview. Again, we're a consortium led by the Department of Energy, a collaborative effort among researchers at nine different national laboratories, and we're focused on two key ideas. First, biomass feedstock properties are variable and different from other commodities. They're variable in that they change over time and space. So depending on how the biomass material—the picture on the right shows corn stover bales at the top and wood chips at the bottom—and how those materials are harvested, how they're stored, how they're transported, and how they're preprocessed all will change their properties. And they're different in the sense that they're different from other agricultural commodities, like grains, for example, that have about a 100-year head start on bioenergy. And we know how to handle those, obviously.
The empirical approaches that have been used to address this variability and these differences haven't been entirely successful. We can't just use hoppers and bins and conversion equipment designed for grains and things like that for biomass. So the FCIC is developing first principles-based knowledge and tools first to understand, and then to mitigate the impacts of feedstock and process variability on buyer refineries.
And there's the link. You can find out all you want to know about the FCIC by visiting us at www.energy.gov/FCIC.
Today's speakers we have Dr. Allison Ray of Idaho National Lab, and Dr. Bryon Donohoe at NREL. Allison is a senior scientist and the Research Excellence Lead for Science and Technology at Idaho. She has extensive experience in biomass science and feedstock logistics and in managing feedstock variability and quality for effective conversion into fuels and chemicals. She has a Ph.D. in environmental microbiology from Idaho State University.
Bryon is a senior scientist at NREL within the bioenergy science and technology directorate. He leads research at NREL's Biomass Surface Characterization Laboratory, the BSCL, and uses a variety of advanced imaging technologies to probe the structural changes in biomass during conversion processes. His Ph.D. is in molecular and cellular biology, MCB, from the University of Colorado.
The feedstock variability task within the FCIC works to understand the intrinsic variability of biomass resources, like corn stover and woody materials, what Mother Nature gives us, and then works with other FCIC tasks to understand how this variability impacts downstream processes.
So Allison will first discuss this intrinsic variability and quality impacts in general. Bryon will then talk about the specific tools the team is using to prove that impact. And downstream you can get operations like material handling, flowability.
During the presentation we do have the chat box active, so please pop a question in there. If it's something simple, Allie and Bryon will be able to answer it. If it's longer, my colleague Amy Slider, who serves as the FCIC project manager, is watching that chat box very closely. She'll watch for all the questions, and at the end of the presentation she'll queue those up for Allison and Bryon to answer those questions. If we run out of time, we have too many questions for the time that is allotted, I promise we will follow up with people individually.
So with that, please bear with me while I queue up the presentations.
Allison Ray, Idaho National Laboratory: Good morning, or good afternoon, depending on your time zone. Thank you for joining us today for this third installment of the FCIC webinar series. I am Allie Ray with INL, and I am joined by my NREL colleague, Bryon Donohoe, and we will be presenting “Unveiling Signatures of Feedstock Variability.” I'll be presenting part one: “Sources of Variability and Impacts on Quality.”
I wanted to begin today by taking a moment to acknowledge all of my feedstock variability colleagues, including our current and former team members. We are fortunate to have a superb team of diverse and talented PIs representing six national labs, from National Renewable Energy Laboratory, Los Alamos, Berkeley, Sandia, Oak Ridge, and Idaho. And this task offers a great example of how collaboration accelerates progress.
Feedstock variability has been cited as a major operational challenge, with ash, moisture, and soil identified as key factors that impact biomass quality, process uptime, and throughput. Herbaceous feedstocks like corn stover typically require storage due to a once-yearly harvest window. Biological degradation often occurs during field-side storage, resulting in dry matter loss and changes to bulk composition and the structural integrity of the bale. Annual variations in rainfall represent another compounding source of variability that can exacerbate the extent of degradation during field-side storage, while distinct plant tissues and anatomical fractions behave and respond differently in mechanical preprocessing and thermochemical conversion.
A 2016 biorefinery optimization workshop report indicated that fundamental characterization of non-pristine biomass was needed to understand the quality of available resources for integrated preprocessing and conversion pathways.
Historically, feedstock variability has been assessed largely on the basis of bulk compositional measures. However, it's clear that physical and mechanical properties also impact biomass behavior, and this is different in different unit operations, and it varies with the type of biomass, as well as with operational parameters. Interactions among these properties reflect the complexity of lignocellulosic biomass and highlights the challenge of really understanding the quality metrics and the logistics that supplying feedstocks with consistent quality.
The goal of the feedstock variability task is to develop the tools that can quantify and understand sources of variability in the biomass resource with the aim of bridging the gap in our explanation and understanding of feedstock variability.
The goal of the FCIC is to develop first principles-based knowledge and tools to understand and mitigate the effects of biomass feedstock and process variability on biorefineries. So I will highlight some of the knowledge and tools that we've developed through the feedstock variability task. Now, the motivation for our task is that characterization at multiple scales offers insights to the sources of variability, compositional and structural attributes that impacts the biomass value chain. And as part of our research we begin to identify these emergent properties of biomass that are [audio skips] underlying physical, chemical, and structural attributes that exist and interact at spatial scales that impact the behavior of biomass and biorefinery operations, which have really struggled with the challenge of accommodating feedstock variability. In our task, we take a fresh approach to characterizing biomass by exploring these emergent properties through the lens of quality by design.
Our task combines multiscale characterization with data analytics to extract features that explain variability. In one study we performed a retrospective analysis of 24 bales across four counties in Iowa, representative of a realistic biorefinery supply shed. Using more than 200 core samples collected from bales, clustering analysis was performed with a combination of 16 organic and inorganic features that were measured from bale cores. Results showed a partitioning of corn stover bales that was driven by significant variations in inorganic speciation, revealing a connection back to county of origin. Though the root cause of such variations could not be determined from these data, differences in growing conditions, tillage, nutrient inputs, or harvesting methods likely played a role.
Clustering analyses indicated that moisture and ash alone on a bulk compositional basis did not differentiate the bales of corn stover from the fields that were examined here. This highlights that county-level differences, even within a supply shed in a high-yielding region, are an important consideration, not only for siting future biorefinery locations, but also for determining optimal process configurations and operational parameters for managing variations in biomass quality.
So as we saw from the previous study, data analytics can be used to reveal important insights about the origins of biomass variability. Our task has employed things like correlation matrices and the hierarchical clustering to the compositional components of corn stover, and this has provided insights to some of the key sources of variability that impact quality. For instance, when we had looked at amino acid profiles in storage-degraded corn stover, it reveals that samples cluster as a function of the extent of biological degradation or the severity of biological degradation. However, when you look at the inorganic species and inorganic components in storage-degraded corn stover, it reveals a connection back to harvest, with samples mapping back to the original bales from which they were derived.
So these findings illustrate how data analytics can be used to glean key insights about the sources of variability that impact biomass quality to better inform management strategies that enable biomass utilization and conversion.
We also developed a molecular characterization approach to look at cell wall modification in corn stover as a function of storage and biological degradation. We used a low-temperature analytical pyrolysis combined with multidimensional GCMS in order to detect changes in biopolymer structure and chemistry. We found that biological degradation disrupted the cell wall structure, fragmenting the hemicellulose and cellulose change. And analytical pyrolysis with GCMS can be a beneficial strategy to improve opportunities for cell wall characterization and provide insights for understanding variability, as well as informing a storage best management practices. Furthermore, these types of data can contribute to techno-economic analysis as a loss of carbohydrates during storage, which are valuable to downstream conversion, have to be balanced by potential opportunities to utilize the residence time of storage for the reduction of recalcitrants through this fragmentation of hemicellulose that we see here.
As I mentioned in an earlier slide, biological degradation that occurs during storage can often be observed by the toppled over appearance of a bale, as well as color change. However, there is a lack of fundamental understanding and characterization of structural modifications that occur in response to biological degradation. Our task has employed FTIR and NMR spectroscopy in order to understand the structural properties of lignin and hemicellulose, and how they're changed in response to biological degradation that occurs during storage. Our findings suggest that the oxidation of lignin, ether cleavage of lignin, and hydrolysis of hemicellulose occurred in degraded corn stover, and this is consistent with our pyrolysis GC/GCMS studies that were presented in the earlier slide.
These findings provide insights that help understand the mechanisms of biological self-heating and degradation processes, and also provide important information about critical structural changes that can impact downstream pretreatment and conversion processes.
Variability is inherent to biomass and can arise from the distinct anatomical fractions, tissue, and cell types within a plant. The traditional approaches for the bulk processing of biomass have ignored the natural inherent variability and the unique properties at anatomical and tissue scales. For instance, if we consider something like corn stover, which is made up of different anatomical fractions in different proportions, we know that all of these fractions appear differently; they have obvious morphological differences at the macro scale, they have different physical and chemical attributes. And then if we look at them under a high-resolution microscope with an SEM, we can see very different structural features. For instance, you can see that leaves of corn stover have long fibers with tightly packed cellular structures. If you look at cobs, they have a dense structure with a woody ring that's stiff and lignified. When we look at the stalks, we can see well-organized and intact sclerenchyma and vascular bundles, which are stiff and very recalcitrant.
And it's obvious and evident that all of these fractions are going to respond and behave quite differently in response to mechanical and chemical processing. Therefore, it's quite important to consider inherent sources of variability when you're considering configurations for the processing of biomass.
We also examined the inherent and introduced chemical compositional variability and anatomical fractions of corn stover. Anatomical fractions all have different chemical compositions, and if we consider the inherent within-plant variability, we can see that leaves are high in ash but much lower in glucan and xylan content when compared to stalks and cobs.
When we further considered the introduced variability that results from storage degradation, we also saw substantial differences. We performed a study using a bale that had been stored for approximately 2 years, and we collected samples that represented a range of degradation from mild to severe. You can see the mild sample on the lower left-hand picture in this slide with the severe sample that shows corn stover that's nearly black. If we look at the chemical composition of the bulk material, it shows a variable response of cell wall composition to degradation. With increasing extents of degradation, we see reductions in ash, reductions in glucan and xylan content, with a substantial increase in extractives and enrichment in lignin. In addition, if we look at the ash and inorganic distribution in the different fractions of stover and as a function of degradation, we see differences across the fractions, particularly in the components of potassium and silica.
We further examined elemental distribution and compositional variability introduced by storage degradation. SEM-EDS mapping of elemental distribution for quantitation of inorganic species was performed, and SEM images on the bottom row reveal that severe degradation of corn stover stems resulted in the translocation of silica from the pith to the outer epidermal tissues, as well as a reduction of potassium shown in green on the right. The translocation of silica and reduction of potassium were further confirmed by quantitation in each tissue type, with translocation from the parenchyma and vascular bundle to the epidermis in the severe condition. Chemical composition shows an increase in degradation leads to dramatic decreases in glucan and xylan contents with an increase in extractives. This is likely due to disruption of cell wall structure and the release of extractable inorganic species and structural carbohydrates that's caused by severe biological degradation.
In summary, we have characterized inorganic species variations and distinct plant fractions from corn stover, where inorganics with localization represents a new aspect of the old attribute of total ash content from standard approaches for bulk materials. We conducted a first-of-a-kind study on the dynamic elemental variability and distributions observed in corn stover fractions as functions of a source of introduced variability storage degradation. This provides fundamental understanding that can be used to inform strategies for harvest and collection, wear and abrasion, selective biomass preprocessing, and equipment design for enhanced valorization. Furthermore, this research also highlights key considerations for storage beyond just mitigating for dry matter loss.
Water chemistry is important to pretreatment and conversion, and our task is exploring variations in free and bound water across anatomical fractions and for degradation state and biomass. Water is essential to biomass hydrolysis, offering a medium by which catalysts can access substrates and products can diffuse away from the reaction site. Water pools in biomass environments include bound water, which interacts closely with cell wall polymers, and free water, which resides in cell lumen. Water distribution and interactions with biomass microstructure influence physical and chemical changes during storage and preprocessing.
This research is led by Dr. Ling Ding at INL, and she gave a really fantastic presentation on this topic yesterday at the SBFC virtual conference, in case you can catch a replay.
Low-field time-demand NMR relaxometry was employed to understand the influence of local environments and elucidate free versus bound water in biomass. Variations in spin spin or transverse relaxation times, or T₂ relaxation times, reflect differences in water status, where higher values of T₂, or greater than 10 milliseconds, reflect pools of water with enhanced mobility, while lower values of T₂, or less than 1 millisecond, reflect water that's more tightly bound to the biomass matrix.
Anatomical fractions of pine have differences in their physical and chemical attributes like microstructure, composition, and water contents, and the results shown here shows significant differences in distributions of free and bound water across anatomical fractions of pine, and as a function of age at harvest, with needles having a substantial pool of free water, while the bark in juvenile pine has more substantial pools of free water relative to pulpwood-age trees. This may indicate a loosening or fragmentation of lumen structure that enables enhanced mobility in the younger tree. This offers a tool that provides fundamental understanding of biomass environments that dictate diffusion, enzyme access, and recalcitrants.
Characterizing the inherent variability of biomass informs opportunities for advanced fractionation and valorization. As part of the FCIC, the feedstock variability task is exploring how sources of variability, like biological degradation and distinct anatomical fractions of the plant, confound standard approaches to bioprocessing. For instance, in our task, we've measured a lignin modification that occurs in response to biological degradation during storage, and that can have effects on the potential for lignin utilization in biomass. In addition, anatomical fractions of the plant have variable responses in mechanical and chemical processing. Leaves are pulverized upon impact, while things like husks and stalk might need sheer-based size reduction approaches. Leaves are more hydrophilic than cob, which also suggests they would have variable responses to pretreatment and fermentation. Understanding a feedstock variability allows for variations of the anatomical and tissue scales to be exploited and to derive value from the overlooked fractions of biomass.
If you have any questions, please feel free to reach out to send me an email; I would be glad to hear from you. Thanks so much for listening and I hope we meet again soon.
Ed: Great. That's the first one. Let's see if we can't get the–
Bryon Donohoe, National Renewable Energy Laboratory: Bryon Donohoe from the National Renewable Energy Laboratory, and I'm going to continue unveiling signatures of feedstock variability with a focus on attributes that have an impact on downstream materials handling and flowability of the biomass. For my part of this talk, I'm going to revisit some of the problems and perspectives that the feedstock variability task has been tackling. I'll start with a couple of stories about some attributes that we've got a pretty good handle on regarding particle morphology and the surface properties of feedstock particles, then get into a couple of topics that we're still working on, dealing with lignin structure and how pine residue particles change during heating. Then we'll wrap up with some final thoughts and ideas about the remaining challenges of understanding feedstock variability.
The FCIC's work is motivated by the unsolved challenges that remain for the efficient operation of integrated biorefineries. Several of these challenges were highlighted and ranked by the participants of a BETO workshop on biorefinery optimization. In this workshop, they listed feedstock flowability and feedstock variability as two of their top concerns. When thinking about feedstock variability, we really had to consider: What's the most critical physical scale to be interrogating all these variable attributes at? For example, the barriers to conversion tend to be down at that molecular or at the nanoscale, whereas the challenges to flowability and conveyance are more particle-scale issues, and both conversion and conveyance are impacted by tissue-scale variability that's inherent in plants.
Our feedstock variability team has worked to both broaden and deepen our fundamental understanding of biomass attributes. In addition to scale, we've been considering the different sources of the variability. Some of them are inherent to the feedstock materials—they come from the biological sources like the plant genotype or the growth conditions—but other sources of variability are introduced to the feedstock materials through the practices and choices that are made during harvest, or storage, or preprocessing, for example. One of the critical material attributes that conversion folks will request for their reactors is a specific particle size distribution. They're often going to describe it with a single value like 2 millimeters or 2 inch. One thing to realize is biomass particles are never isotropic; they are basically always longer than they are wide and wider than they are tall. Particle size distribution is clearly one of the feedstock variability attributes that has induced the choice of type of mill and screen and how they run and dictate the particle size and morphology, and it is of course influenced by some other attributes, like moisture content.
These images are just to show some of the morphologies that can be seen in a handful of hammer milled corn stover. You can imagine that some of these look like they're more likely to just completely interlock with their neighbors than they are to flow peaceably beside them. Luckily there's an entire team in FCIC working specifically on preprocessing, and they're looking more into the knife mills and rotary shear mills that can generate more uniform particle morphologies.
Allie had discussed some of the new understandings we've come to about biomass ash content in terms of its specific speciation and localization and potentially the migration during biological degradation. Another aspect of high ash content is how it impacts the formation of fines. In these images you can visually see the increased fines content in the samples on the left relative to the ones on the right. The higher extrinsic ash or dirt content of these materials is both physically part of the fines and also likely contributing to the formation of additional fines by acting as a grinding agent during milling.
I also wanted to draw your attention to some of the work we published last year with Dave Sievers, working with a large data set of images captured during several pilot-scale biomass conversion runs.
I'll refer you to our paper for the details of how we applied machine learning to classify the images in the context of motor torque values from a cross-feeder component of the conveyer system. What I wanted to address with you today is how we're developing image analysis tools to measure surface texture as a potential proxy for particle size and particle shape analysis. With this kind of tool, we can analyze piles of particles instead of having to isolate individual particles to image separately.
In the case of these way belt images, it worked pretty well. In other words, the images of piles of biomass that corresponded with the low and high loads on the cross-feeder could also be distinguished based on surface textural features. In the example images on the right, you can see how surface texture is distinguishing more typical-looking biomass on the left from the biomass on the right that has a stringy texture to it and has become dominated by these high-aspect-ratio particles. We've since applied this method to additional image data sets of more finely milled material and are continuing to develop this tool going forward.
As Allie had already mentioned, we spent quite a bit of time and effort with these degraded bales and analyzing the impact of storage as a source of induced feedstock variability. I like this picture of the corn stover loaded truck because I can really empathize with the biorefinery operator watching this load rolling in and thinking, you know, what exactly is the value of all those different bales on that load and what's the plan for handling those bales at the back of the load? The picture on the left reminds us that even within a bale there can be a lot of variability. And the graph on the right shows us that the dominant anatomical fractions within these bales were the leaf and stem tissues, which is what we focused on with a lot of our characterization. But we also tried to spend some time with the cobs because they're quite a unique fraction.
Because feedstock flowability is such a major concern and a challenge, we decided to turn our attention for quite a while on surface attributes, and in this case looking at the effect of degradation and the variability among the different anatomical fractions. We measured surface texture at the millimeter scale using optical microscopy and at the micrometer scale using electron microscopy. And here I also want to point out this is one of the scenarios where figuring out the best scale to work at wasn't obvious. We started at the millimeter scale with optical microscopy, we thought that would probably be the easiest and probably be the most meaningful for thinking about how particles interact and the frictional forces between them.
We've since migrated this surface roughness analysis to focus even more at the micrometer scale because we think it's actually giving us more fundamental information that informs how the biomass particles interact, and in particular how they trap water between adjacent particles.
These black-and-white SEM micrographs in the lower left-hand portion of the slide highlight another feature of particles at this scale, is that they can have wildly different surfaces on the different sides of the same particle. A piece of stalk, for example, depending on the way it fractures during milling, can have an extremely smooth surface that's the original exterior of the stalk, while on the other side of the particle the original interior of the stalk is extremely rough—not to mention that these two surfaces have very different chemistries as well.
Another set of surface-oriented characterization has been being developed by Troy Semelsberger's group, at LANL. These techniques actually interrogate multiple surfaces and multiple scales, so it's not just the visible external surfaces of milled particles like we were looking at with the imaging techniques, but in addition, more of the internal surfaces that need to be accessed by pretreatment chemistries and enzymes. Some of the measurements here being uniquely applied to biomass feedstocks are surface energy, surface area, hydrophilicity, and pore volume. And here you see even more evidence for the variability caused by degradation and how it's impacting different tissue types very differently. The lead fraction in particular was super susceptible to changes induced by degradation, at least in terms of how it increased its surface area, surface energy, and pore volume, whereas the cobs were relatively impervious and really didn't change at all.
When I think the cross-sectional anatomy of a leaf and all the mesophyll tissue and the high pectin content, I mean this might make some sense. But it's really not all that intuitive, because if we were to take these same anatomical fractions into the biorefinery and subject them to milling, pretreatment, and enzymatic hydrolysis, we'd find that the cob is actually far and away the most digestible material. So we heard about the losses of hemicelluloses and we saw changes in surface attributes, but another feature of degraded corn stover is that it becomes more brittle. This has an impact not only on the integrity of the bales and stacks, but also on how the material grinds or mills. So degraded materials generate a lot more fine particles, which is generally undesirable.
Amber tested the breaking strength of these corn stover stalks and quantified the reduction in strength with degradation. It was really this brittleness attribute that got us thinking more about how lignin may be changed, because it wouldn't normally just implicate loss of hemicellulose in changing strength of the materials. At least when the plants are alive, they normally strengthen themselves either by beefing up their cellulose content to resist tensile forces, or by increasing lignin content to resist compressive forces.
Our first evidence that something had actually changed about the lignin came from Yining's work with lifetime fluorescence lifetime imaging, where the progressively degraded materials were showing a progressively shortened fluorescence lifetime. So what we're seeing in these images is a trend in multiple tissue types—leaf, stalk, and both in cross-section and longitudinal sections—from an orange and yellow color to the blue and greens that indicate a shorter lifetime. The shortened lifetime of lignin fluorescence is an indication that something fundamental has changed about either the lignin's molecular environment or the structure of the lignin itself.
Some further analysis of the fluorescence signal showed that the severely degraded materials, there was a reduction by half in the number of fluorescence emitters that can be measured. Also, there is an attenuation in the pattern of fluorescence emission intensity using a polarization of the excitation laser rotated across 180 degrees. So if you look in the control, the fluorescence is brighter when the excitation laser is oriented parallel with the long axis of the fiber cell wall, and that pattern is lessened in the degraded materials. This combined with the lifetime suggests that what's happening with the lignin is that it's going from a more extended, more linear morphology to a more compacted and folded one that allows energy funneling, creating the reduced number of emitters, and reducing the influence of the polarized excitation.
To answer the question of whether lignin is just partially collapsed or whether it's actually condensed and has formed new carbon-carbon bonds, we analyze the samples by NMR. And Renee was able to confirm that there was no dramatic change in either lignin content or composition. In fact, while there's some shifting of peaks, the β-O-4 content has actually dropped in the severely degraded samples, suggesting that, if anything, there's some slight depolymerization of the lignin polymers going on.
When Xiaowen took these same materials and applied his deacetylation and mechanical refining pretreatment to them, he found that only about half as much lignin was extracted during the pretreatment and more of the lignin was left behind in the pretreated feedstock. Then when those materials were subsequently subjected to enzymatic hydrolysis, there was a 20%–30% reduction in the yield of both glucose and xylan.
So these results alone don't tell us that it's actually a structural change in the lignin that's the cause of the reduced conversion. In fact, there's quite a complicated series of things going on, and ferreting this out is the topic of work that's continuing this year in the FCIC.
Just like Allie had showed, while we did start with corn stover, we have managed to look at some other feedstocks. So we’ve begun looking into pine forestry residues as well. With the pine, here the emphasis has been on understanding how anatomical fractions, and particularly the needles and the bark, behave in the context of high-temperature conversion. One of the tools we've been using is in situ microscopy to replicate the heat ramp that pine particles experience in the feed screw that transfers the milled pine from a hopper into the fast pyrolysis reactor.
In the upper left you can see some movies playing, and it gives you an idea of how these materials heat up, and when that happens their surfaces get soft and the particles begin sort of changing shape. In the lower panels we have some still frames from those in situ heating movies from five different pine tissues. So the original size of the particle are outlined in red, and what you're actually seeing is an image of the final particle after it reaches 375°C. All the particle types are still at a maximum volume at 250°C, and then reach a maximum rate of shrinkage at 375°C. And while we let the in situ experiments go all the way out to pyrolysis temperatures, it's really around that 375°C that's mimicking the temperatures that would be experienced inside of the screw feeder. The two tissue types that exhibited the most change were the cambium and the whitewood.
A part of the hypothesis with this whole study is that there might be particular tissue types that are the bad actors and the ones responsible for maybe evolving volatile materials that can condense back on the surfaces of nearby particles and become a source for nucleating blockages that can plug up feeders. We did actually see evidence for the kind of thing we were envisioning, but only on one tissue type, and that was actually the cambium. At 375°C the cambium tissue surfaces showed these small drop-like deposits. We're not sure yet if that's enough to actually cause particles to stick together, and we're going to repeat some of these experiments with a higher density of particles and mixtures of anatomical fractions.
We also performed confocal Raman microscopy to study the surface chemistry of these particles. And while all the anatomical fractions start out relatively similar, likely because the spectra dominated by the main polymers of the tissues in the cell walls, by the time they hit 375°C, there's some distinct differences among the different fractions. We're still working with this data, and one of our projects for this year is trying to determine whether these changes are significant enough to implicate them in creating a sticky surface that could cause plugging and blockages.
So to wrap up, we talked about how all biomass particles are anisotropic and how some tissues are especially prone to generate high-aspect-ratio particles, which are a problem for flowability. We looked at the way biological degradation affects the surface properties of corn stover and how that impact is different on the different anatomical fractions. It looks like even lignin is being modified in degraded materials, with potential negative impacts on the performance of pretreatment and enzymatic saccharification.
And finally, the different anatomical fractions in pine residues respond differently to the heating regime that they experience in a screw feeder, and some of the tissue types may be developing a sticky surface, which could contribute to plugging.
In addition to thanking the entire feedstock variability team that Allie acknowledged in one of her initial slides, I wanted to call out some of the key individuals from both feedstock variability and the other teams within FCIC whose data or ideas I presented here today. And again, a big thanks to the Bioenergy Technologies Office for its leadership and support of this research.
And with that I'll leave you with our contact information. If any of these topics we've been discussing resonate with you, we'd be happy to chat about it anytime. And with the remaining time we have we'd be happy to answer any questions you might have.
Ed: Great. Thanks to Bryon and Allie. I think I'm just showing the—go to the questions slide—you know, a fair comment. This picture was taken, and the one that Bryon showed was clearly pre-COVID, so just for the record there. I see Bryon's picture and I don't have the chat handy. I see—okay, now I see the chat.
Rohit asked a question to Allie, "Great presentation. From an anaerobic digestion perspective it feels as if one could design to cope with variations. So is there an angle—is there a way to design around feedstock variability? Not only all—" Rohit's asking specifically about anaerobic digestion, but maybe I'll tee that one up for other processes. And then, Allie, if you want to take a crack at that, and then Bryon as well? I think I'll pass that to you guys.
Allison: So I guess I would agree with you. Other than that, from a design perspective, to cope with variations, I think I'm going to hand that one off to Bryon. [Laughs]
Bryon: Sure. I don't know I have a great answer for this, but I mean I love the sentiment about coping with variations. I mean this just hearkens back to this idea of the sort of intrinsic or inherent variability versus the induced. There's so much inherent variability that I think part of our charter is to come up with the technologies that can cope with it. I mean, the technologies that would eliminate it are kind of pretty tough to conceive in some cases, and likely impossible in others.
That said, you know, we definitely, in a lot of the things that we've done, we're kind of advocating for sort of amelioration techniques that involve like fractionation. I mean, there seems to be no question that if you can afford to parse some of these things that are widely different apart ahead of your convergence strategy, you're probably going to have an easier time of it.
Ed: Great. Thanks. I will add one minor thing to that, is one of the points that I think Allie and Bryon made that maybe I should've emphasized in my introduction as well, is that within the FCIC we're moving away from names of materials to the attributes of those materials. And if you've got two different materials, whether they be the tissue anatomical fractions of an herbaceous material like corn stover or of a wooded material like pine, those will behave differently in all of the various unit operations that comprise our refinery.
And I can—you know, people have heard me say this a bunch of times—but Richard Feynman, the physicist, said, "I can tell you the name of something, but I haven't told you anything about it." So by focusing on the attributes, the physical, the mechanical, the chemical attributes of the materials, we have more levers to understand that variability, but then also to mitigate it.
Allison: Yeah, I was going to—I was sort of thinking about the same comment, Ed, that the nice thing about the quality-by-design framework that's been established through the FCIC was that it really allows us to focus on the attributes rather than the feedstock so much. So understanding this relationship between attributes and impacts allows you to move toward this feedstock-agnostic approach, to open up more potential feedstocks and potential pathways so that you can really have a good understanding of attributes that are tailored to pathways rather than the other way around.
Ed: Thanks, Allie. Rohit followed up by using, "If you could use novel enzymes as opposed to EH (enzymatic hydrolysis), perhaps these microbes could be designed to cover a broader range of substrates." I think you're preaching to the choir there, Rohit. By understanding—again, we're, if you will, agnostic to potential downstream conversion processes, if you will. But if you were to understand that variability better on a chemical, molecular, mechanical, and physical levels of the different scales that Bryon alluded to, and the chemistry versus the mechanical aspects that Allie alluded to, then you can design your conversion process or processes, multiple unit operations to take advantage of that.
And so I wouldn't want to pass judgment on AD versus low-temperature conversion versus high-temperature conversion, that's outside my area of expertise, but I agree, understanding the variability of the attributes and that what we're trying to do, what Allie's and Bryon's team are trying to do I think provides tools and knowledge to the folks like you, who are trying to optimize a given unit operation. So right on.
I don't think I'm seeing any other questions in the chat, so I'll go—we'll do a going once, going twice. But I would like to remind everybody a couple things. First of all, thank you all for joining us today. There are a bunch of ways—there are a number of ways to get some more information about the FCIC. You can check out our website; again, just Google "BETO FCIC" and you can get the website. On that website you'll see a complete list of publications, many of them with Bryon and Allie's name on them. They've done a really great job in publishing their findings. You also, if you want a higher-level perspective on it, there's a brief description of our structure and the various tasks, as well as a downloadable 2020 year-in-review report, real high-level discussion of all the impacts that we had.
And then finally, please reach out to us. All of our contact information is available on the website as well.
It will take a couple of weeks typically to get this video up to the website. We actually close-caption it, so that takes a little bit of time. But that will be available along with the other webinars that we've given. And then in about a month we'll be doing another one of these webinars in our series, this time on signatures of where on processes across the bar refinery, and that will be led by our—presented by some colleagues at the Oak Ridge National Laboratory and the Argonne National Laboratory.
So with that end material, I'll thank everybody, and I'll give everybody one last chance. If you do have questions, put them in the chat, but if not, we will give everybody a couple minutes of their morning back. Again, thank you very much. And I would be remiss if I didn't say thanks again to BETO for the support. And thanks to all of you again for taking time out of your day.
So thanks very much and I think we are—I guess we'll end it here. So thank you all very much. I see a hand up, but I can't seem to see a question.
Bryon: Oh, yeah, the hand was from Tim Sheets and he asked if we got his email. So he must have sent his questions in an email to someone. So we'll definitely follow up with him.
Ed: Okay. Well then, thanks, everybody. Have a pleasant morning or afternoon, as the case may be.
[End of Audio]