Atomic layer deposition for lithium-ion batteries
- Author
- Felix Mattelaer
- Year
- 2017
- Abstract & Cover
- Source of Information
- CoCooN website, www.cocoon.ugent.be
- University
- Ghent University, Department of Solid State Sciences, CoCooN research group (Ghent, Belgium)
- External Link
- Read Thesis
introduction and research objectives Lithium-ion batteries (LIB) power most electronic devices today, and are applied in an extremely wide range of applications: large scale off-grid and grid-levelling solutions, electric and hybrid electric vehicles, modern power tools and portable electronics such as laptops and smartphones. Recently, even smaller batteries are coming into focus, such as small-scale microbatteries for on-chip storage, autonomous sensors and implantables. These emerging markets drive the lithium-ion battery towards new concepts and architectures, such as nano-structured electrodes and all-solidstate devices. Furthermore, evermore stringent demands are placed on existing and novel battery chemistries, such as improved cycle life, fast charging, high safety and biocompatibility. Atomic layer deposition (ALD) is a deposition technique where thin films are deposited from sequential gas-surface reactions. Thanks to the self-limiting nature of these reactions, ALD provides interesting deposition characteristics such as sub-nanometre level thickness control, low deposition temperatures and excellent conformality. These unique features enabled ALD to become a well-established tool in semiconductor manufacturing. Other domains, such as catalysis or energy storage, can also benefit from these interesting coating characteristics. However, for these applications ALD is yet to break the bonds of academia into industry. In this work, atomic layer deposition is investigated for uses in the field of lithium-ion batteries. Two application areas are proposed: • Novel battery concepts such as the 3D thin-film lithium-ion battery are very promising in applications such as on-chip energy storage, as a result of their potential high power, high energy and good safety. To obtain these 3D thin-films, conformal deposition technique are required. Atomic layer deposition is one of the few techniques that can achieve conformal thin films in very high aspect ratio structures. • As battery ageing for traditional lithium-ion batteries can already be problematic, novel architectures, battery concepts and material chemistries pose increasingly greater challenges in this direction. Surface coating has always been one of the concepts to overcome ageing and improve battery performance. To obtain a high quality surface coating on the complex structures of battery electrodes, even in traditional batteries, a conformal deposition technique such as atomic layer deposition is paramount. atomic layer deposition for thin film electrodes Atomic layer deposition and phase control of manganese oxides An ALD process for the deposition of manganese dioxide was reported using the Mn(thd)3 precursor, i.e. manganese tris(2,2,6,6-tetramethyl-3,5- heptanedionato), by Nilsen et al before the start of this PhD. As the growth rate of this particular process chemistry was reported to be very low (0.1 to 0.2 Å/cycle), three new plasma-enhanced ALD processes were developed. These processes obtained a higher, temperature independent growth rate of 0.2 Å/cycle. Interestingly, depending on the nature of oxidative or reductive species in the plasma (H2 , NH3 , H2O) or ozone, films with different oxidations states were obtained: MnO2 (ozone), Mn3O4 (H2O plasma), Mn3O4/MnO (H2 plasma) or MnO (NH3 plasma). The oxidation and reduction behaviour of these films in inert, reducing and oxidative atmospheres were studied using in-situ X-ray diffraction. In this way, all manganese oxide phases could be obtained, including the Mn2O3 phase which would not be obtained directly from (PE-)ALD growth with the Mn(thd)3 precursor. Applications of ALD manganese oxides The four as-deposited films were characterised as thin-film electrodes. Their different oxidation states resulted in different electrochemical behaviour: MnO2 behaved as a positive electrode, while the MnO-containing films could be used as negative electrodes. However, the slow growth rate limits the thickness of these films to ≤30 nm, restricting their use as thinfilm electrodes. Furthermore, the conformality of these ALD processes was subpar, so 3D thin-film mangnanese oxide electrodes could not be demonstrated. Two alternative applications were envisioned where only ultra-thin films were required. MnO2 and Mn2O3 were examined as thin-film catalysts for water splitting, and performed comparably to platinum/carbon black, a benchmark catalyst. Ultra-thin MnO2 films were also examined as a protective seed layer for the electrodeposition on oxidizable current collectors, which enabled the growth of thicker, electro-active MnO2 films Atomic layer deposition of vanadium oxides ALD of vanadium oxides was performed using TEMAV, i.e. V(NEtMe)4 , as a precursor for vanadium, while water, ozone and oxygen plasma were used as reagents. Amorphous VO2 films were deposited with the first two. Oxygen plasma was reported to deposit crystalline V2O5 , but the process was modified to lower temperatures (<100 ◦C) and lower precursor dosage to obtain amorphous V2O5 . The crystallisation and phase change behaviour of these two flavours of amorphous VO2 were investigated using in-situ XRD in atmospheres ranging from oxidizing to inert. The influence of the substrate, temperature, ambient and initial flavour of VO2 were unravelled. All crystalline phases in the Wadsley series (VO2 -V2O5 ) could be obtained. Vanadium oxides as thin-film electrodes The ALD-derived crystalline films in the Wadsley series, i.e. VO2 (B), V6O13, V4O9 , V3O7 and V2O5 , showed good electrochemical activity as positive lithium-ion electrodes, with very high storage capacities up to 1.4 A h cm−3 . V4O9 demonstrated the highest capacity of the crystalline films, while VO2 (B) had the best trade-off between capacity, rate capability and cycle life. Amorphous VO2 and V2O5 were benchmarked as electrodes against their crystalline counterparts. The amorphous films generally demonstrate higher capacities and exhibit better rate capability thanks to lower densities and superior lithium diffusion coefficients. The cycle life for these amorphous films was not perfect, and capacity fading was related to vanadium dissolution at high state-of-charge. 3D thin-film electrodes Contrary to the manganese oxide ALD, the conformality of the thermal vanadium oxides processes was excellent. Vanadium oxide films were deposited on micro- and nano-structured scaffolds, i.e. silicon micropillars and carbon nanotubes. Footprint capacities up to ∼100 µA h cm−2 were demonstrated on both structures, with excellent rate capabilities inherited from the thin-film nature of the electrodes. atomic layer deposition for interface modifications Atomic layer deposition is considered as an excellent candidate to modify and stabilize interfaces of lithium-ion battery particles or electrodes. ALDdeposited Al2O3 is the most studied surface modifier, but many other ALD films are also explored. In most published work, these coatings were investigated to extend the cycle life of various electrodes, but the effect on the rate capability is often omitted. Here, this influence on the electrode kinetics is studied in detail, and the best coating is evaluated on an ALD V2O5 3D nano-structured thin-film electrode. ALD interface modifications and the rate capability Two thin-film electrodes were used as simple model systems to unravel the role of ALD coatings on lithium-ion battery electrodes. Anatase TiO2 was considered a near-ideal model system as it does not suffer from SEI formation, metal dissolution of large volume changes. 0.5 nm to 5.0 nm films of amophous ALD Al2O3 and TiO2 were deposited on the surface of this electrode. ALD Al2O3 appeared to be very resistive towards lithium, and significantly deteriorated rate capabilities were found. ALD TiO2 on the other hand retained the rate capability of the model electrode. Overcharged thin-film LiMn2O4 was used as a model system for solvent decomposition. A 1 nm ALD Al2O3 film already posed a large impedance, and could not alleviate the solvent decomposition, resulting in severely degraded rate performance. 5 nm ALD TiO2 on the other hand could prevent the solvent decomposition, which appreciably improved the rate capability compared to the uncoated electrode. ALD-modified vanadium oxide 3D thin-film electrodes A 3D thin-film electrode was constructed from an ALD-derived crystalline V2O5 film deposited on carbon nanotubes (CNTs). This V2O5/CNTs electrode suffered from capacity fading when charged to 2 V vs Li+/Li, which was shown to be related to vanadium dissolution. The interface was modified with 5 and 25 cycles of ALD TiO2 , which maintained the excellent thin-film kinetics of the V2O5/CNTs electrodes. Vanadium dissolution was completely supressed with 25 cycles of ALD TiO2 , which resulted in a stabilized electrode capacity conclusions Atomic layer deposition was investigated for applications in lithium-ion batteries. First, electrodes grown from existing and novel ALD processes were evaluated. If the conformality of the deposition allows it, as was the case for ALD vanadium oxides, these electrodes can be nano-structured, which increases their footprint capacity and at the same time maintains the excellent rate capability of the thin-film electrodes, yielding high-energy and high-power electrodes. Next, atomic layer deposition was evaluated for electrode surface modifications. The lithium diffusion through the deposited coatings needs to be considered when choosing a coating, as a poor lithium transfer can significantly degrade battery performance, as was shown for ALD Al2O3 . Finally, 3D nano-structured electrodes and interface modifications were combined in an ALD TiO2/ALD V2O5/CNTs electrode, which exhibited excellent rate capability, high footprint capacity and an excellent cycle life, demonstrating both applications of atomic layer deposition for lithium-ion batteries.
