AVS 70 Session TF1-TuA: Thin Films for Energy Applications II: Batteries

Sulfide solid state electrolytes (SSEs) are considered one of the most promising materials to replace liquid electrolytes in next-generation batteries due SSEs astonishing ionic conductivities. However, the high reactivity of sulfide SSE materials precludes direct contact with Li metal anodes. One approach to stabilize the sulfide SSEs is passivating the sulfide interface with thin films of stable chemical species, forming an artificial solid electrolyte interphase (ASEI) layer. Ideal passivation materials for the anode/SSE interface are generally comprised of Li₃N or LiF species, which are stable at Li metal potentials. Here, we investigate the impact of plasma treatment on sulfide SSEs, as well as deposition of plasma-based ASEI materials using atomic layer deposition (ALD).

I will discuss the implications of Ar, N₂, and NH₃ plasma exposures on sulfide solid electrolytes (LGPS, LPSCl) and the resulting surface chemical changes as determined by in-vacuo x-ray photoelectron spectroscopy (XPS) measurements. N₂ and NH₃ plasma exposure can cleave Li₂S surface bonds and replace them with Li₃N surface species, however due to a lack of Li available at the surface this coverage is limited to ~30% and cannot form complete ASEI layers. Additionally, I will report the changes in electrochemical behavior of these pellets in Li/SSE/Li symmetric cells. Due to the incomplete surface coverage of Li₃N species, the surface nitride passivation is reduced and still allows deleterious reactions to occur after contact with Li metal. These results indicate that for complete Li₃N surface coverage, additional Li species must be available on the sulfide SSE surface before plasma treatment. I will also discuss the initial stages of ALD ASEI layer deposition on the sulfide SSE materials and evaluate their protection efficacy.

Fast Li⁺ ion conducting thin film solid state electrolytes (SSEs) by atomic layer deposition (ALD) enable high power density, fast time constants, and high operating frequency regimes in solid state ionic devices. However, the ionic conductivity of thin film SSEs fabricated by ALD has been limited by material development challenges, phase purity concerns when dealing with tertiary and quaternary systems, and challenges related to crystallinity. These complexities require careful consideration of the material system, precursor selection, process parameters, and post-annealing conditions to realize ideal ALD thin-film SSE materials with ionic conductivities >10^-6 S/cm. Bulk oxide SSEs, such as Li_1+xAl_xTi_2-x(PO₄)₃ (LATP) have a NASICON-type crystal structure, show high ionic conductivity when crystalline (10^-3 S/cm), demonstrate air & water stability, and a high voltage stability window. ALD synthesis of LATP can be broken down into well-known constituent processes (e.g., Li₃PO₄, Al₂O₃, TiP₂O₇) which can be combined with respect to stoichiometric ratios.

LATP can be considered Al-doped LiTi₂(PO₄)₃ (LTP), in which Al⁺³ ions partially replace Ti⁺⁴ ions in the NASICON-type structure. This facilitates fast Li⁺ ion conduction through the 3D network. With that in mind, an ALD process for LTP was first developed by alternating between Li₂O and TiP₂O₇ sub-processes. The Li₂O sub-process uses lithium tert-butoxide (LiO^tBu) as the lithium source and water to complete the surface reaction. The TiP₂O₇ sub-process uses titanium (IV) isopropoxide (TTIP) as the titanium source, trimethyl phosphate (TMP) as the phosphorous source, and water. The growth rate of LTP was measured to be 0.4 Å/cycle at 300 ^oC. By adjusting the ratio between Li₂O and TiP₂O₇ cycles, the Li concentration in LTP can be tuned between 8.4-34.3 at % Li.

A NASICON-type crystalline structure is achieved by post-annealing the LTP films between 650 ^oC – 850 ^oC. The ideal annealing temperature was found to be a function of Li-ion concentration. The resulting measured ionic conductivity of partially crystalline LTP thin films is 3.6x10^-5 S/cm at 80 ^oC. In this talk, we will discuss the complex issues related to process parameters, stochiometric tunability, and the resulting ionic conductivity when exploring ternary and quaternary LTP/LATP metal phosphates phase space.

Trimethylaluminum (TMA) stands as a keystone in atomic layer deposition (ALD), widely esteemed for its role in crafting aluminum-based coatings across diverse applications. Notably, TMA is used to produce Al₂O₃ with H₂O, AlF₃ with HF-pyridine, and Al₂S₃ with H₂S, adhering to well-defined ALD mechanisms. These processes have permeated various substrates, leading to ultra-thin Al-based coating layers, particularly within the realm of energy storage materials such as battery cathodes, anodes, and solid electrolytes.

Our research group has embarked on a profound exploration of TMA's ALD chemistry, unveiling captivating interactions with reactive metal substrates, solid-state electrolytes, and battery cathode materials. Interestingly, when subjected to Li-metal anodes, TMA deviates from conventional ALD pathways, engendering a carbon composite layer atop the Li metal. This unconventional behavior arises from the inherent reactivity of both TMA and Li metal, diverging from established ALD paradigms. Moreover, our investigations reveal TMA's propensity to engage with Li-based impurities on solid electrolyte and cathode surfaces, yielding protective layers with beneficial attributes. Astonishingly, despite TMA's extensive utilization, this phenomenon has eluded prior scrutiny.

In this presentation, we unveil our findings on the intricate chemistry between TMA and battery materials (Li-metal, cathode, solid electrolytes) elucidated through in-situ quartz crystal microbalance (QCM), high-resolution X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), and Raman spectroscopy studies. Furthermore, we demonstrate the transformative potential of these discoveries in the realm of battery materials surface treatment, offering tantalizing prospects for streamlined industrial applications.

Understanding lateral diffusion of lithium in thin-film solid-state battery (SSB) materials is critical to improving SSB performance, stability, lifespan, and architecture. Using model test structures of sputtered V₂O₅/W/LiPON, we use Raman spectroscopy peak shifts to map Lithium concentration in V₂O₅. We evaluate the diffusion of lithium from LiPON layers into thin-film V₂O₅, measure the impact of different sputter deposition process conditions on lateral lithium diffusion, and investigate the effects of post-annealing the fabricated test structures. We will discuss the benefits of front versus back-side Raman spectroscopy for analyzing masked devices and correlate Raman spectroscopy mapping to XPS data to confirm compositional accuracy. Additionally, we will discuss the impact of electrochemical cycling at C/10 on lateral Lithium diffusion for symmetric V₂O₅/LiPON/V₂O₅ devices. Lastly, we will discuss the important implications of the lateral spacing limitations of clustered SSB devices due to lateral diffusion and propose architectural design rules based on this diffusion behavior for optimized device performance.