Additive manufacturing of ceramic materials for energy applications: Road map and opportunities

Abstract

Among engineering materials, ceramics are indispensable in energy applications such as batteries, capacitors, solar cells, smart glass, fuel cells and electrolyzers, nuclear power plants, thermoelectrics, thermoionics, carbon capture and storage, control of harmful emission from combustion engines, piezoelectrics, turbines and heat exchangers, among others. Advances in additive manufacturing (AM) offer new opportunities to fabricate these devices in geometries unachievable previously and may provide higher efficiencies and performance, all at lower costs. This article reviews the state of the art in ceramic materials for various energy applications. The focus of the review is on material selections, processing, and opportunities for AM technologies in energy related ceramic materials manufacturing. The aim of the article is to provide a roadmap for stakeholders such as industry, academia and funding agencies on research and development in additive manufacturing of ceramic materials toward more efficient, cost-effective, and reliable energy systems.

Introduction

Among engineering materials, ceramics are indispensable in energy applications such as batteries, capacitors, solar cells, smart glass, fuel cells and electrolyzers, nuclear power plants, thermoelectrics, thermionics, carbon capture and storage, piezoelectrics, turbines and heat exchangers, among others. They are of particular interest because of their inherent high temperature capabilities and corrosion resistance, but they possess many other useful properties at ambient and elevated temperatures and pressures. Fig. 1 shows the US advanced ceramic market size, it was valued at USD 97.0 billion in 2019 and is expected to grow at a compound annual growth rate (CAGR) of 3.7% by 2027 [1]. Rising demand from industry, including renewable energy and medical sectors, are expected to propel market growth over the forecast period. Rising product demand from the clean technology industry will also support market growth. Many of these applications are directly for energy usage and storage and rely on the advancements of ceramic processing and materials technology.

Ceramics are used in many energy applications, and some of them are specifically introduced in section. Ceramics are used in emission reduction, for example through control of emissions from combustion engines, and CO₂ (or carbon) capture. For emission control in combustion engines, ceramic honeycombs (more than 90% of honeycombs currently used worldwide in combustion engines) function as a filtering device and building foundation hosting the catalytic coating. For these applications, the ceramic of choice is often cordierite, an extremely low-cost refractory oxide ceramic capable of withstanding substantial temperature variations downstream of an engine.

Significant advances in battery energy storage technologies have occurred over the past decade with solid state batteries; in these systems, the materials used for the electrolyte and cathode are monolithic ceramic oxides. Much of this work has led to battery pack price decreases of over 90% since 2010, with the lowest reported price under $100/kWh in 2020 [2]. Li-ion battery performance has increased, and the cost has decreased, which is especially good for the current high demand for Li-ion batteries for electric vehicles (EVs). The annual deployment of Li-ion batteries is projected to increase roughly eightfold over the next 10 years, reaching nearly 2 TWh of capacity globally [3]. However, there is a strong demand for increased performance, system safety, lower manufacturing costs and lower environmental footprint. For example, transportation applications will require additional advances including cost reductions below $60/kWh of usable energy at the pack level, fast charging capabilities of less than 15 min, and energy densities that will allow increased EV range between charges [4].

Nuclear reactors of both the fission and fusion types impose high temperatures, aggressive corrosion requirements, and high radiation fluxes on materials. Silicon carbide (SiC) fiber reinforced SiC matrix composite (SiC_f/SiC), among other ceramics, have seen significant worldwide interest in the development of cladding materials that are more resilient to high temperature steam oxidation as would be experienced in many design-basis and beyond design-basis nuclear accidents. Specifically in nuclear applications, SiC has resistance to swelling with high dose of neutron fluences of up to 100 displacements per atom (dpa) [5]. Because of their high temperature severe environment capabilities and the unique functionality, ceramic materials, in monolithic or composite forms, are essential for enabling fusion energy, in components such as the flow channel inserts in liquid metal blankets, tritium breeders, the radio frequency plasma heating window, diagnostic mirrors, the blanket and first wall structures. Additionally, ceramics with high thermal conductivity (to transfer heat effectively) and high strength (to afford thin tubular shapes or channels) are considered the best materials for high temperature heat exchangers. These materials include metal oxides, nitrides, and carbides such as WC, SiC, Si₃N₄, and Al₂O₃.

Nanocomposite materials based on electrically conducting percolative networks, particularly those consisting of carbon-based secondary phases dispersed within a ceramic matrix (as in polymer-derived ceramics) possess high piezoresistive response and are attractive for applications operating in harsh conditions, e.g., high temperatures, corrosive environments, etc., because of the intrinsic robustness of the ceramic matrix materials. Piezoelectric ceramics such as lead zirconate titanate, barium titanate, lead titanate, and potassium sodium niobate are excellent candidates for sensors and energy harvesting devices, such as in wind energy harvesters.

Yttria-stabilized zirconia (YSZ) for the electrolyte and YSZ-based composites (for example Ni-YSZ cermet) as the fuel electrode are the state-of-the materials for solid oxide fuel cells (SOFCs) and solid oxide electrolysis cells (SOECs). In these devices, the electrolyte must be sufficiently densified to avoid leakage of the fuel/oxidant gases through the electrolyte to the electrodes, while both cathode and anode are required to be porous, electrically conductive and should possess high activities for fuel oxidation and oxygen reduction.

Semiconductor materials alloyed to provide high ZT values are used in thermoelectric solid-state energy conversion devices that convert heat to electricity and vice versa. For conversion of very high temperature heat (from sources such as hydrocarbon combustion, concentrated sunlight, or nuclear generation processes) directly into electricity, ceramic materials with high power density outputs are needed for heat-to-electricity conversion efficiencies of interest in thermionic energy converters (TECs).

Ceramic turbine blades and vanes are needed to further improve turbine efficiency owing to the refractive ceramic properties including higher melting temperature, potentially high creep resistance, potential for high corrosion resistance at elevated temperatures, and low density. Ceramic cores and molds are also essential to the casting of the highest performance conventional, superalloy turbine components. High precision and careful control of thermal and mechanical properties of the ceramic cores and molds during superalloy casting is critical to obtain the complex, internal cooling passageways needed in high performance turbine components.

Smart glass and smart glass devices are glass-based materials or devices utilizing constructions for active or adaptive response to environmental or interventional stimulation to tailor functional properties. Most prominently, these can be smart windows or other components of a building with adaptive energy (photovoltaic or heat) harvesting ability, cooling functions or emissivity control and shading techniques.

Ceramics have inherently high melting temperatures, and this makes them difficult to process by melting. Oxide ceramics have some degree of ionic bonding, which allows for self-diffusion, so they can be sintered in powder form. Because of high degree of covalent bonding in carbide, nitride, and boride ceramics, their sintering is very difficult. Traditionally, casting and molding, sintering, and machining is carried out to reach high precision monolithic ceramic parts. Additionally, for many carbides and ultra-high temperature ceramics, hot pressing and spark plasma sintering with subsequent machining is done. These methods achieve highly dense material with good properties and microstructure, but the geometries formed by these methods are simple and axis symmetrical, as shown in Fig. 2A and B. Also, these methods require tooling to shape the preform (as in injection molding) or tooling to keep the powder compact under constant mechanical load during heating and sintering (as in hot pressing). After densifying into objects with simple shapes, they may need to be machined into complex shapes, which requires machining with expensive tooling.

Several methods have been developed to form SiC preforms with more complex geometries such as die extrusion, slip casting, tape casting, gel casting, and injection molding [6], [7], [8]. These methods utilize solvent-based slurries with dispersed powders to shape parts, typically with dies. These methods can achieve high solids loadings, which translates to high green density and less shrinkage during sintering. These methods also provide high volume production. In die extrusion, axisymmetric dies are used in extrusion producing geometries like the ones in Fig. 2B. The geometries achievable with the other slurry-based casting methods are shown in Fig. 2C. Although the forming provides more freedom compared to compacts and pressure-assisted methods with dies (not including hot isostatic pressing (HIP)), there are still certain designs such as ones with internal features that are not achievable. Additionally, there is a cost associated with die tooling, so there is motivation to use methods without the associated dies.

Methods for reinforced ceramics also have more specialized processing and are of large interest in energy applications. As such, ceramic matrix composites (CMCs) are processed by different methods because they rely on continuous fiber reinforcement, interface coating, and a matrix material that can bear and transfer mechanical load giving the CMC damage tolerance through interfacial debonding, crack deflection and propagation along the interface leading to fiber sliding and pullout [10]. To process CMCs, the fibers must be spun into tows, pyrolyzed, and sintered or slightly crystallized. Next, fiber tows are woven into 2-D plies or fabrics and coated with an interface coating. Then, the cloths are prepregged with slurries to help hold the fabrics together and improve the fiber packing and density and laid up onto tools where they are subjected to some pressing with or without heating. Before deciding on a densification method, the pre-impregnated preform is pyrolyzed. Preforms are infiltrated with methods such as reactive melt infiltration (RMI), chemical vapor infiltration (CVI), or polymer impregnation and pyrolysis (PIP).

The market for ceramics made via additive manufacturing is also growing, as many industries realize the benefits provided by advanced techniques for design, cost savings and, in some cases, time saving. A recent report analyzed pros and cons and generated revenues of the dominant AM technologies for processing ceramics, both technical (advanced) and traditional (clay-like) [11]. Trends can be seen in Fig. 3. The advanced methods in the study included material extrusion, photopolymerization and binder jetting 3D printing technologies. As the aggressive design demands call for low-cost, high temperature ceramics in complex shapes, the need for improved solutions is significant. Compared to traditional ceramic processing techniques and even casting techniques, solid free-form fabrication (SFF), additive manufacturing (AM), or 3D printing will be used in energy industries where it is necessary and makes sense for the design, properties, and cost. In this article, the three terms will be called AM to limit confusion. For ceramics AM, there are several leading technologies for shaping preforms [12], [13]. One method is called Vat photopolymerization, where a laser (traditionally called stereolithography) or light projection (digital light processing (DLP) with LEDs or LCD screens) is used to cure the resin. Another Vat photopolymerization method is called Lithography-based Ceramic Manufacturing (LCM) and can use a laser, LED, or LCD system. Some newer light-based curing techniques use machines without vats but rather use reservoirs of slurry or paste that is spread onto build platforms and subsequently cured with a light source (such as companies Admatec and 3DCeram). Another method is nozzle extrusion and is called direct ink write or robocasting if a liquid material (gel/slurry/paste/thermosets) is used and FDM if a solid, thermoplastic material is used. A method of forming dry powder bed with liquid binder is typically called binder jet 3D printing (BJ3DP). A dry powder bed method with laser source to sinter layers is called selective laser sintering or melting (SLS/SLM). Low viscosity slurry forming is called material jetting/inkjet printing.

Vat photopolymerization techniques such as stereolithography (SLA) and digital light processing (DLP) cure 2D layers of photosensitive monomer resins loaded with powder or preceramic polymers by curing with light. High resolution and low surface roughness are achievable with lithography, and it is a proven technology for powders with low light absorbance such as oxide ceramics. It is also a proven technique for preceramic polymer forming. The drawbacks of lithography are the material limitation and use of support material. As pointed out in [14], debinding of printed parts is achieved when the crosslinked resin decomposes, so pyrolysis must be done. High solids loading is difficult in resins, and limitation on solids loading in relation to viscosity and spreading must be monitored. Between crosslinked binder and more resin present in the green bodies, debinding can be difficult and lead to more defects like the ones seen in [15], [16]. Generally, across all 3D printing methods, there are more defects and deviations in porosity and microstructure at layer interfaces.

Extrusion printing methods utilize solvent, gel, paste, thermoplastic, or thermoset slurries that are extruded through a nozzle onto a build platform. Extrusion methods can print particulate with orientation as well as many different materials, but their drawbacks are resolution, speed, throughput, and filament adherence. One other drawback is that viscosity must also be monitored in order to get enough shear and flow through the nozzle. BJ3DP utilizes a dry powder bed where a solvent and polymer mix is sprayed onto the powder bed in 2D layers [17], [18]. BJ3DP can achieve high throughput and high resolution of parts because of the printing nozzles and accuracy of the jets. The drawbacks of printing with BJ3DP are the large particle sizes needed, random orientation of printed fibers and particulate, depowdering difficulty and time, and low green (and final) density. Material jetting is the expulsion of a low viscosity slurry onto a platform with subsequent drying of the layers. This method provides a thin layer of material, but the solids loading is typically low and drying effects can cause cracking.

Selective laser sintering or melting utilizes a dry powder bed and a laser to sinter or melt patterns into 2D layers. It can achieve some solid-state or liquid-phase sintering of ceramics, but it is not as commonly used for ceramics because of cracking of parts due to large thermal gradient during the process. It should be noted that SLS is the only method that can provide some direct consolidation of the ceramic powder during printing because of the heat applied. For all the other ceramic AM techniques, post processing steps such as solid-state sintering, liquid-phase sintering, reaction synthesis sintering, reactive melt infiltration (RMI), chemical vapor infiltration (CVI), or polymer impregnation and pyrolysis (PIP) are needed to consolidate the printed preforms into dense, usable parts. A schematic of these AM methods with images of the processes used for ceramics is shown in Fig. 4.

Multi-material ceramic AM is still a considerable challenge both in terms of the actual printing as well as the subsequent processing (debinding and sintering), although some recent advancements have occurred, with the development of multi-vat lithographic printers and multi-nozzle extrusion and in-line mixing devices [19].

Additive manufacturing opens the design space and allows for improved design in terms of material placement, architecture, microstructures, and properties. Designs that were not thought possible have become a reality. As an example, there are many architectures to consider when manufacturing a battery, and Fig. 5 presents a variety of 3D printing techniques suitable for electrode fabrication and the pattern and its mechanism for the optimization of electrodes. These architectures are achievable with various AM methods, so these materials and architectures can be made and have a large impact on the performance and properties of batteries. Another example is small channels and tortuosity that can be achieved with AM of SiC and other materials for high temperature heat exchangers. It is the same parallel for many other ceramic materials used in energy.

Each energy application in Fig. 6 has certain design criteria for success and improvements that are dictated by the properties. Based on the properties outlined, the most appropriate material can be selected. For given materials, the manufacturing process can be selected to provide the best selected material with the best properties to meet the application design needs. This process and idea are outlined in Fig. 6. These factors may result in several potential AM technologies to be appropriate for each application. More specifically, energy applications outlined above may require high fracture toughness and high strength, high thermal conductivity, high corrosion resistance, high temperature stability, radiation resistance, high electromechanical coupling coefficients, ion conductivity, high ZT values, high power density outputs, among others. Added to these properties may be the requirement for low-cost and availability (in the cases of high-volume productions). In certain applications, two or more properties may be required. As an example, high temperature and high-pressure heat exchangers require high thermal conductivity (as usual), with the added requirement for high fracture toughness to withstand high pressures. In these scenarios, a good approach is using multi-physics topology optimization (TO) to obtain the optimal designs. The next step would be identifying AM technologies that are able to process and manufacture those certain ceramics in identified geometries from TO results. In the latter process, the design for additive manufacturing (DFAM) principle should be considered.

This review article is divided into sections focused on specific energy applications. For each application, its requirements are discussed, followed by why ceramic materials are the best choice for that application. Traditional fabrication and manufacturing processes for each application is outlined, and the case is made for potential improvements AM technologies may offer for each application. There are several reports in the literature on covering various AM processes for specific applications [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]. The focus on this article is on ceramic materials and processing, and hence the readers are referred to those reports for in depth discussion on various AM processes.