ADVANO’s Approach to Manufacturing: Envisioning a Factory for the Future
In a previous blog post, we covered a set of design rules to transform silicon into an anode material for rechargeable Lithium-ion batteries. The blog discussed a diverse range of solutions to tackle a similar class of problems— structuring silicon in unique ways to extract optimal performance as an anode material within the electrochemical environment of a battery cell. We also described in some detail ADVANO’s silicon anode solution — REALSi. This blog aims to demonstrate our approach to the manufacturing of REALSi, and how we envision it to be scaled for mass manufacturing while retaining quality and consistency. We will first walk through our manufacturing technology and associated manufacturing philosophy, including a technical vision for how we intend to upscale our manufacturing to meet domestic and global demand for silicon anode materials.
Design for Manufacturing
The manufacturing of REALSi consists of three major stages: nanostructuring, granulation, and pyroprocessing. Let’s visit the design rationales behind these manufacturing processes.
Nanostructuring: Silicon is a brittle material, prone to fracture under stress along certain crystallographic planes. Using high-purity metallurgical silicon as the feedstock, we exploit this tendency to repeatedly fracture the silicon with millions of collisions induced by the stochastic motion of an inert high-strength media in a reactive liquid phase. The choice of the reactive liquid phase is made using a blend of low-cost, stable organic solvents containing certain additives to control the surface chemistry of the silicon during the nanostructuring process. Fracturing the silicon in this manner derives silicon nanoparticles that maintain a tight particle size distribution with a certain characteristic shape and functionalized surface. The resulting silicon nanostructure possesses several key characteristics: excellent dispersion stability in the reactive phase and surface chemistry that protects the silicon against oxidation during subsequent stages of manufacturing.
Granulation: In this phase, a blend of carbon precursors and certain additives are incorporated into the silicon nanoparticle suspension. The suspension is derived in such a manner that the carbon precursors can form interfacial bonds with the functionalized silicon nanoparticles, without disturbing the state of the particle dispersion achieved in the previous step. This suspension then undergoes a granulation process where it is atomized into microscale droplets. By the end of the process, we derive an intermediate product — a nanocomposite particle composed of uniformly distributed silicon nanoparticles immobilized within a solidified carbon matrix. The motivation behind granulation can be inferred from our previous post. To recap, one of the important reasons is to generate a powder material with the characteristics required for it to be compatible with industry-standard electrode coating processes (both wet slurry and dry electrode).
Pyroprocessing: In this stage, we stabilize the granulated particles and transform the carbon phase into a well-defined porous microstructure. The state of dispersion of the silicon nanoparticles is preserved, with the particles being connected by a shared conductive carbon matrix. Our choice of carbon precursors plays a fundamental role in obtaining the appropriate type of carbon matrix that offers optimal compatibility for both battery electrode processing and electrochemical activity. The carbon itself is also capable of participating in electrochemical activity in a cell, instead of only functioning as a protective barrier for the entrapped silicon.
We incorporate the following aspects in our manufacturing philosophy:
◾ Location-agnostic manufacturing: Considering our choice of raw materials, we do not necessarily have to commission a production facility in a geographic location close to the source of the feedstock. Rapid advances in the purification and processing of metallurgical silicon material have broadened the availability of battery-grade silicon raw material. Our partnership with Ferroglobe PLC provides us with long-term supply chain security assurance.
◾ Safe process design: Each stage of ADVANO’s manufacturing process is designed to incorporate ingredients that are safe for storage, handling, and transportation. Each ingredient used to manufacture REALSi and their usage in each step of manufacturing is carefully vetted employing sound chemical engineering principles. The design of each stage of our process also incorporates strict process safety guidelines.
◾ Process output efficiency: Our process is designed to maximize the yield of intermediate products passing through each stage of manufacturing. We intentionally select ingredients that are feasible to be recycled and purified for reuse in our manufacturing process.
◾ Equipment selection: Our manufacturing process is designed to incorporate engineered solutions already in wide use in typical materials processing industries including minerals, battery materials, pharmaceuticals, and food products. This approach alleviates the level of effort required to re-engineer a solution that can be scaled for mass production. In other words, it helps us to focus more on material design and process engineering rather than custom equipment design.
Our Approach to Scaling Beyond the Laboratory
Scaling the manufacturing of a product to industrially relevant volumes is easier said than done. The stakes are even higher when the product is a specialty material incorporating nanotechnology and advanced material engineering. Considering the class of problem, the following basic questions need to be answered to address the scaleup problem:
◾ Are the physics of the different stages of the manufacturing process transferable to larger batch sizes in larger equipment?
◾ What factors are responsible for manufacturing REALSi with a high degree of consistency in bulk physical attributes, material properties, and quality?
◾ How amenable is the manufacturing process for tweaking REALSi to provide customized solutions to meet unique customer requirements?
ADVANO has conducted intensive process engineering at our current facility in New Orleans, Louisiana. This has allowed the team to exhaustively understand the influence of manufacturing process conditions, process parameters, and equipment selection on REALSi production yield, material quality, and material characteristics. We have conducted detailed characterization of almost every conceivable aspect of REALSi at multiple scales: individual particle level, bulk material level, and cell level. 90% of these multiscale characterizations were accomplished by our capable in-house teams and our own equipment, and the rest in collaboration with external partners. By now we have amassed a significantly large dataset that gives us insight into the engineering required for scaling up REALSi manufacturing beyond the lab scale. It does not stop there — we have also intimately collaborated with internationally renowned equipment vendors to qualify their large-scale production equipment for compatibility with our REALSi manufacturing technology. This equipment and related plant engineering know-how will be part of our upcoming 10 TPY specialty production plant (SPP), awaiting construction in Louisiana.
Mapping the Regimes of Manufacturing
For mass-scale manufacturing of a specialty material, it is necessary to map out the process-structure-property (P-S-P) interrelationships to truly understand the regimes within which scaling up is feasible. Scaling up these processes is also a function of understanding the volume dependence of the physics involved in each of the unit operations. The high-level process flow hierarchy shown below demonstrates the controllable input parameters and derived output parameters in each stage of manufacturing. Having already done thousands of physical experiments, we now know to a high degree of detail the windows of each parameter and the effect of their interrelationships on the final product. This intimate knowledge of our processes is what also provides us the capability to select and qualify production-scale equipment. This level of granularity essentially enables us to not only achieve high consistency in product manufacturing but also provides us the broad ability to tune the REALSi anode material, targeting specific characteristics as requested by our partners.
Intelligent Manufacturing — A Path to a Factory for the Future
To truly scale beyond several tons per year to hundreds or thousands of tons per year, we envision enhancing our engineering prowess beyond physical parameter mapping. Data-driven intelligent manufacturing will play an increasingly influential role in increasing manufacturing scalability by leaps and bounds. To this end, the initial stage requires three real-time data sources: logging of operational parameters, logging of machine conditions, and monitoring the process itself. Production equipment will be interconnected by distributed control systems for providing continuous real-time operational parameter monitoring, which will generate the datasets necessary for P-S-P mapping. Using sensor fusion based on high-performance IoT platforms, we will be also able to incorporate real-time machine condition monitoring (MCM) in our production equipment to generate datasets that will be necessary to map the dependence of equipment uptime on process parameter regimes. Fast-throughput in-line material analysis systems will also need to be integrated to enable real-time monitoring of process conditions.
With our current lab line dataset including the dataset that will be derived from the SPP, it would be possible to harness the power of artificial intelligence toolsets to optimize our unit processes to an extreme degree. The motivation behind this thinking is multifold. As shown in the process flow diagram above, there are many individual input and output parameters that at first sight (not all are shown here), seem to be flowing unidirectionally. While the manipulable bounds of these parameters are tied to the capabilities of each process, I envision that there exists a quantifiable influence of “parameter sets” — unique groupings of parameters that exercise unique influence on the step-wise outputs and the ultimate outputs. This approach is different from using standard statistical process control (SPC) measures, which are more relevant to product quality optimization if only a limited variety of products were to be manufactured. However, as the usage of silicon anodes in commercial rechargeable lithium-ion batteries continues to expand in the coming years, this will require customized offerings of silicon anode material demanding increasingly higher performance capabilities. Since the individual parameters are already tagged and known in advance, supervised learning models employing hierarchical artificial neural networks (ANNs) will be possible to be used to design predictive models for advanced process mapping.
The combination of all of these models will also enable us to understand what specific component in a piece of equipment we can upgrade or retrofit to expand its operational window. Additionally, this will also enable us to get a broader understanding of our process sensitivity and tolerance. This lessens the need for the obsessive development of complicated computational models simulating the granular multiscale physics of each of the unit operations, which is often a non-trivial endeavor on its own.
Reaching Cell Maturity with REALSi
ADVANO has dedicated significant scientific and engineering efforts to designing every aspect of REALSi. We have been able to exercise great granular control on its characteristics by studying the performance of REALSi in simulated electrochemical environments (coin cells and pouch cells). The team is now well-prepared in our ability to consistently manufacture our product in multiple customized variants. It is now time to validate the performance regimes of REALSi in practical cell formats. To that end, in CY2023 we have successfully completed cell builds in multi-layer pouch and cylindrical cell form factors in close collaboration with both domestic and overseas cell partners, with additional builds scheduled throughout this year. Instead of blindly relying only on the expertise and experience of these cell partners, ADVANO’s team has worked in close collaboration with the partners to select the appropriate cell design, and provided detailed data on REALSi integration, including the selection of the cell ecosystem. Our intention with these cell builds is to generate a structured dataset, i.e. a dataset derived from systematically conducted investigations. These cells are going to be subjected to an exhaustive test matrix at our facility, with our partners, and with independent validators. This endeavor will provide us with valuable insight regarding the electrochemical performance regimes of REALSi in such cells, and the influence of cell design and cell ecosystems. The ultimate objective of generating this dataset and upcoming datasets is to continually upgrade our technology readiness level (TRL), via multiple stages of cell maturity qualification of REALSi.
Long-Term Ambitions
But even after going through this expansive pursuit, is it guaranteed to get us to a point where we accomplish offtake agreements with cell and EV partners to continually scale up mass production? It is indeed a circular problem — go through the qualification cycles with each partner, wait to sign an offtake or supply agreement, then ramp up production by expanding manufacturing throughput which is not a trivial task to accomplish rapidly. We also think it is not sufficient to rest our laurels on simply being a battery materials manufacturer subject to the mercy of the external environment in the long run.
At the end of the day, silicon anode materials are intended to be used as a minor component of lithium-ion battery cells, with their proportion in the cell increasing over time (We briefly covered what it will take to gradually increase its proportion in a previous blog). The crux is that we are the ones who know most intimately the characteristics of our material. In that effort, we found it necessary to take the initiative to continuously upgrade our team’s capabilities to gain in-house expertise in cell design, cell fabrication, and cell assembly — albeit on small scales. Bidirectional knowledge sharing with our cell partners is continually increasing our understanding of the intricacies of cell manufacturing. We find ourselves more and more in the role of a silicon anode solution provider, not just a material manufacturer. Not every partner is well-versed in silicon anode integration in their baseline cells, and with the availability of multiple types of silicon anode materials, this situation is becoming more complicated over time. It often befalls us to provide expert consultation for REALSi integration. We also see that compared to the impressive ubiquity of overseas cell manufacturing, domestic cell manufacturing capability has a significant amount of room to grow. Therefore, in the long term, our ambitions also include laying the groundwork for a transformation into a domestic cell manufacturer. These ambitions should not necessarily be viewed as us emerging as a competitor to established cell manufacturers, but rather as an endeavor to carve our niche in a rapidly growing industry to sustain the momentum required for the transition to clean energy. We do not have to stretch ourselves too thin trying to tackle so many technical solutions all at the same time. With appropriate fundraising and resource allocation, we also envision upscaling our talent acquisition and cultivating the culture necessary for forward-thinking with highly disciplined, systematic, and granular planning. This talent will serve as the necessary springboard to expand our expertise and capabilities in both material manufacturing and cell manufacturing.
The technical groundwork required for these ambitions can also be viewed through the lens of systems engineering. A very large system can always be subdivided into smaller subsystems, and prioritizing the order of the development of smaller subsystems would gradually take us toward tackling the larger system as a whole. To demonstrate this point on a simple basis, firstly consider that the sole point of integration of REALSi in a battery cell is in the anode electrode. Upscaling our electrode process design capability from manual or lab-scale techniques to pilot-scale processes using semi-automated equipment at one of our domestic cell partner’s sites is already in progress. The simplified process flow diagram exhibits how we are currently investigating the process parameter windows to achieve high-quality electrodes integrating REALSi as a drop-in with graphite in the anode. In the short-term such investigations on our end are intended to shorten the development cycle time of our cell partners for integrating REALSi materials in their cell designs.
Aside from expanding our electrode manufacturing knowledge base, we have also already taken initial steps to lay the groundwork for complete large-format cell assembly, and over time our capabilities will gradually improve as we keep chipping at the problem. In the distant future, as our database of parameters keeps growing, we can apply similar manufacturing scale-up considerations as explained above for material manufacturing.
Takeaways
ADVANO is on a focused mission to consistently keep advancing forward in playing an integral role in accelerating the clean energy transition by harnessing our silicon anode solutions, targeting step changes in energy storage performance in current and future generations of rechargeable lithium-ion batteries. We are continually upscaling our material manufacturing capabilities and we have already made significant advances in our short-term targets of qualification of REALSi in commercial cells. Simultaneously, we have started the gradual but diligent pursuit of laying the groundwork for transitioning beyond a battery materials manufacturer in the long term. We envision these efforts to ultimately enable us to establish ADVANO’s bold ambitions in realizing a “factory of the future” by offering solutions that bring value to our partners, customers, and the world as a whole.
The blog is written by Saheem Absar.
Share this blog with your friends, especially if they are interested in the abovementioned areas. You can also connect with ADVANO on Medium, Linkedin, and Twitter.
