Research Article www.acsami.org
Strain-Induced Lithium Losses in the Solid Electrolyte Interphase on Silicon Electrodes Ravi Kumar,† Peng Lu,‡ Xingcheng Xiao,‡ Zhuangqun Huang,§ and Brian W. Sheldon*,† †
School of Engineering, Brown University, 182 Hope Street, Box D, Providence, Rhode Island 02912, United States General Motors Global R&D Center, 30500 Mound Road, Warren, Michigan 48090, United States § Bruker Nano Surfaces, 112 Robin Hill Road, Goleta, California 93117, United States ‡
ABSTRACT: The chemical and mechanical stability of SEI layers are particularly important for high capacity anode materials such as silicon, which undergoes large volume changes (∼300%) during cycling. In this work, we present a novel approach for applying controlled strains to SEI films with patterned Si electrodes to systematically investigate the impact of large volume changes on SEI formation and evolution. Comparisons between patterned silicon islands and continuous silicon thin films make it possible to correlate the irreversible capacity losses due to expansion and contraction of underlying silicon. The current work demonstrates that strain in the SEI layer leads to more lithium consumption. The combination of in situ AFM and electrochemical lithium loss measurements provides further information on SEI layer growth. These experiments indicate that inplane strains in the SEI layer lead to substantial increases in the amount of inorganic phase formation, without significantly affecting the overall SEI thickness. These observations are further supported with EIS and TOF-SIMS results. A map of irreversible capacity evolution with strain in the SEI is obtained from the experimental results. KEYWORDS: solid electrolyte interphase, capacity loss, silicon anode, in situ atomic force microscope, lithium-ion battery
■
INTRODUCTION Passivation films play a critical role in the performance of Li-ion batteries. This film, usually referred to as the solid electrolyte interphase (SEI), forms due to reductive decomposition of electrolyte on the anode surface. In an ideal case, it should be electronically insulating, allow fast Li+ ion transport, and not undergo mechanical degradation. However, in silicon anodes the huge volume changes during cycling (>300%) can cause corresponding large strains on the SEI layer that lead to serious concerns about its mechanical integrity. This has led to widespread recognition that the stability of SEI layers is a major roadblock in the commercialization of silicon-based anodes. Although substantial efforts have been made to improve SEI stability and performance, the evolution and degradation of these films during cycling are still not well understood. Our interest is largely focused on how Li-induced expansion and contraction of silicon affects SEI. Most of the existing improvements in passivation have been achieved using electrolyte additives,1−3 encapsulation of silicon particles,4−6 and artificial SEI layers.7−9 A wide variety of characterization © 2017 American Chemical Society
techniques including X-ray photoelectron spectroscopy (XPS), 10−12 electrochemical impedance spectroscopy (EIS),10,13 Raman spectroscopy,14 and atomic force microscopy (AFM)15−17 have been used to characterize the formation, evolution, and functionality of SEI layers. However, the existing knowledge obtained from these investigations is largely focused on SEI chemistry, while very little is known about the mechanical integrity of the SEI. Previous research on mechanical degradation in silicon anodes has largely focused on understanding and mitigating active material damage such as cracking18−20 and delamination.21 This understanding has led to the design of nanostructured anodes, such as silicon nanowires, silicon nanoparticles, and hollow silicon nanostructures, which are more resistant to fracture.22 These electrodes still undergo huge volume expansion and contraction during cycling, leading to stresses in the SEI layer. However, there is Received: May 11, 2017 Accepted: August 3, 2017 Published: August 3, 2017 28406
DOI: 10.1021/acsami.7b06647 ACS Appl. Mater. Interfaces 2017, 9, 28406−28417
Research Article
ACS Applied Materials & Interfaces
Figure 1. Comparison of deformation behavior of SEI on (a) continuous silicon film and (b) patterned silicon island during lithiation. Due to lateral sliding of silicon islands during lithiation, SEI is subjected to tensile strains which could lead to cracking/delamination of SEI and hence further capacity loss. (c) Cycling recipe to capture the irreversible capacity loss or growth of SEI.
the first time the correlation between strain in the SEI layer and capacity loss due to mechanical damage of the SEI.
no clear way to directly study the mechanical deformation of SEI in these structures. This difficulty motivated our use of thin-film electrodes to investigate the impact of mechanical deformation on SEI layer formation and evolution processes. In another recent paper, we conducted in situ and operando AFM imaging of patterned silicon islands and reported extensive cracking of SEI layers in the shear lag zone of patterned silicon islands.15 However, new SEI formation and evolution at the bottom of these cracks was not clearly evident due to AFM scanning limitations. The experiments reported here are designed to obtain additional information that is not readily obtained from AFM. By comparing continuous and patterned silicon thin film configurations, we have directly obtained quantitative data on “extra” capacity losses resulting from mechanical degradation of the SEI layer. Detailed studies1,23,24 of SEI composition have employed silicon thin films that are essentially identical to our reference configuration, where expansion occurs only in the out-of-plane direction and there is minimal lateral strain in the SEI. Nadimapalli et al.25 estimated the capacity losses associated with initial SEI formation on Si thin film electrodes. However, these estimates differ significantly from practical electrodes where SEI is subjected to repeated expansion and contraction. The primary aim of the current paper is to examine the implications of these strains on the formation and evolution of SEI layers. We previously demonstrated that thin patterned silicon islands can undergo lateral expansion and contraction due to a shear lag effect.26,27 However, continuous silicon films only expand in the out-of-plane direction due to the constraint from the substrate. Figure 1 shows a schematic of Li-induced deformation of these two configurations. This indicates that SEI on continuous silicon films is free of any strains whereas the expansion of patterned islands parallel to the substrate results in significant strains in the SEI layer. By comparing the lithium losses in these two configurations, it is possible to investigate the SEI layer response to mechanical strains that are induced by the underlying silicon. The experiments reported here show for
■
EXPERIMENTAL SECTION
Sample Preparation. Two electrode morphologies were used in this study: 1) continuous silicon films (50 nm thick) and 2) patterned silicon films (25 μm × 25 μm × 50 nm). A thickness of 50 nm was chosen to avoid cracking and delamination of the active material during cycling. These samples were prepared by first depositing a 500 nm thick copper thin film on top of copper foil by electron beam evaporation at a rate of 2.0 Å/s to provide a defect free surface. Silicon was then deposited by electron beam evaporation at a rate of 1.5 Å/s. The patterned samples were made by through-mask e-beam evaporation on copper foil, using custom brass masks (0.005 in. thick with 25 μm × 25 μm holes and 10 μm spacing between holes). Capacity Losses and SEI Formation and Evolution. Figure 1a,b shows schematic representations of deformation behavior of continuous films and patterned islands during lithiation. As described above, the continuous silicon films only expand in the out-of-plane direction due to the constraint from the substrate. However, in the patterned islands interfacial sliding near the edges (shear lag zones) leads to lateral extension parallel to the substrate.26,27 Prior investigations of patterned islands indicate that the size of this shear lag zone is in the micrometer range, which is smaller than the size of the islands used here.15 Therefore, only silicon in the shear lag zone undergoes lateral extension, while the center of these islands are similar to continuous films where expansion only occurs normal to the substrate. In the shear lag zone, the lateral expansion and contraction of the underlying silicon results in tensile and compressive strains in the SEI layer, respectively. It is important to note that the concurrent out-of-plane expansion of silicon leads to strains in the shear lag zone that are not simply in-plane. Even with a more complex deformation field, comparisons between these samples provides a direct way to probe the impact of large volume changes in Si−Li alloys on SEI formation and evolution. Electrochemical Measurements. The thin film silicon specimens were used as working electrodes with pure lithium metal foil as a counter electrode in CR-2032 coin cells. These cells were assembled with a Celgard separator and a standard liquid electrolyte (1 M LiPF6 with a 1:1 volume ratio of ethylene carbonate (EC) and diethyl carbonate (DEC)). They were cycled with a Bio-Logic VSP potentiostat at C/20 currents. 28407
DOI: 10.1021/acsami.7b06647 ACS Appl. Mater. Interfaces 2017, 9, 28406−28417
Research Article
ACS Applied Materials & Interfaces
Figure 2. Comparison of irreversible capacity evolution of copper foil and 50 nm thick continuous silicon film during (a) first cycle and (b) second cycle. In order to study the dynamic behavior of the initial SEI layer formation, the first few cycles were studied in more detail. To apply controlled strains to the SEI layer, the modified galvanostatic cycling methodology in Figure 1c was employed. The electrodes were cycled by gradually lowering the cutoff voltage, and capacity losses were recorded at different states-of-charge. This procedure allowed us to control the lithiation of silicon and to correlate the capacity losses with both initial SEI formation and its deformation. Since the silicon thin films used here are very thin (50 nm), cracking and delamination of the active material are unlikely. Also, there are no inactive components such as carbon black and binder. Therefore, the capacity losses are primarily associated with SEI formation. The in situ AFM measurements were conducted with a Dimension ICON electrochemical AFM inside of an argon-filled glovebox (Nano Surfaces Division, Bruker), where both H2O and O2 were below 1 ppm. The unique PeakForce tapping mode was used with MLCT tips (Bruker AFM Probes), composed of a silicon nitride cantilever with a sharp silicon nitride tip (spring constant: 0.6 N/m; nominal tip radius: 20 nm). For SEI composition analysis, the samples were run for 3 cycles. After cycling, the coin cells were opened inside of a glovebox, and the electrodes were rinsed with DMC to remove electrolyte residue and dried in Ar. These samples were then transferred in a sealed apparatus (no exposure to air) to the time-of-flight secondary ion mass spectrometer (TOF-SIMS) system. The time-of-flight secondary ion mass spectroscopy (TOF-SIMS) analyses were conducted with a PHI TRIFT V nanoTOF (Physical Electronics, Chanhassen, MN). A 30 kV Au+ source was employed for analysis and sputtering. Depth profiles were collected from a 50 × 50 μm2 area inside a 200 × 200 μm2 sputtering area. For patterned silicon, the depth profiles were reconstructed from only the center area of the island Si. TOF SIMS images of different ion fragments were also collected from a 50 × 50 μm2 area at various depths in the SEI. The base pressure of the analysis chamber was maintained at
