Peter Kalisvaart

Origin of non-SEI related coulombic efficiency loss in carbons tested against Na and Li

Partially ordered but not graphitized carbons are widely employed for sodium and lithium ion battery (NIB and LIB) anodes, either in their pure form or as a secondary supporting phase for oxides, sulfides and insertion electrodes. These “pseudographitic” materials ubiquitously display a poor initial coulombic efficiency (CE), which has been historically attributed to solid electrolyte interface (SEI) formation on their large surface areas (up to ∼2500 m2 g−1). Here we identify the other sources CE loss by examining a pseudographitic carbon with a state-of-the-art capacity (>350 mA h g−1 for NIB, >800 mA h g−1 for LIB), but with a purposely designed low surface area (14.5 m2 g−1) that disqualifies SEI from having a substantial role. During the initial several (<5) cycles both Na and Li are irreversibly trapped in the bulk, with the associated CE loss occurring at higher desodiation/delithiation voltages. We measure a progressively increasing graphene interlayer spacing and a progressively increasing Raman G band intensity, indicating that the charge carriers become trapped not only at the graphene defects but also between the graphene planes hence causing them to both dilate and order. For the case of Li, we also unambiguously detected irreversible metal underpotential deposition (“nanoplating”) within the nanopores at roughly below 0.2 V. It is expected that in conventional high surface area carbons these mechanisms will be a major contributor to CE loss in parallel to classic SEI formation. Key implications to emerge from these findings are that improvements in early cycling CE may be achieved by synthesizing pseudographitic carbons with lower levels of trapping defects, but that for LIBs the large cycle 1 CE loss may be unavoidable if highly porous structures are utilized.

Si nanotubes ALD coated with TiO2, TiN or Al2O3 as high performance lithium ion battery anodes

Silicon based hollow nanostructures are receiving significant scientific attention as potential high energy density anodes for lithium ion batteries. However their cycling performance still requires further improvement. Here we explore the use of atomic layer deposition (ALD) of TiO2, TiN and Al2O3 on the inner, the outer, or both surfaces of hollow Si nanotubes (SiNTs) for improving their cycling performance. We demonstrate that all three materials enhance the cycling performance, with optimum performance being achieved for SiNTs conformally coated on both sides with 1.5 nm of Li active TiO2. Substantial improvements are achieved in the cycling capacity retention (1700 mA h g−1vs. 1287 mA h g−1 for the uncoated baseline, after 200 cycles at 0.2 C), steady-state coulombic efficiency (∼100% vs. 97–98%), and high rate capability (capacity retention of 50% vs. 20%, going from 0.2 C to 5 C). TEM and other analytical techniques are employed to provide new insight into the lithiation cycling-induced failure mechanisms that turn out to be intimately linked to the microstructure and the location of these layers.

Silicon nanowire lithium-ion battery anodes with ALD deposited TiN coatings demonstrate a major improvement in cycling performance

We demonstrate that nanometer-scale TiN coatings deposited by atomic layer deposition (ALD), and to a lesser extent by magnetron sputtering, will significantly improve the electrochemical cycling performance of silicon nanowire lithium-ion battery (LIB) anodes. A 5 nm thick ALD coating resulted in optimum cycling capacity retention (55% vs. 30% for the bare nanowire baseline, after 100 cycles) and coulombic efficiency (98% vs. 95%, at 50 cycles), also more than doubling the high rate capacity retention (e.g. 740 mA h g−1vs. 330 mA h g−1, at 5 C). We employed a variety of advanced analytical techniques such as electron energy loss spectroscopy (EELS), focused ion beam analysis (FIB) and X-ray photoelectron spectroscopy (XPS) to elucidate the origin of these effects. The conformal 5 nm TiN remains sufficiently intact to limit the growth of the solid electrolyte interphase (SEI), which in turn both improves the overall coulombic efficiency and reduces the life-ending delamination of the nanowire assemblies from the underlying current collector. Our findings should provide a broadly applicable coating design methodology that will improve the performance of any nanostructured LIB anodes where SEI growth is detrimental.

Magnesium and magnesium-silicide coated silicon nanowire composite anodes for lithium-ion batteries

We synthesized composites consisting of silicon nanowires (SiNWs) coated with magnesium (Mg) and magnesium silicide (Mg2Si) for lithium-ion battery anodes and studied their electrochemical cycling stability and degradation mechanisms. Compared to bare SiNWs, both Mg- and Mg2Si-coated materials show significant improvement in coulombic efficiency during cycling, with pure Mg coating being slightly superior by ∼1% in each cycle. XPS measurements on cycled nanowire forests gave quantitative information on the composition of the SEI layer and showed lower Li2CO3 and higher polyethylene oxide content for coated nanowires, thus revealing a passivating effect towards electrolyte decomposition. Extensive characterization of the microstructure before and after cycling was carried out by scanning- and transmission electron microscopy aided by focused ion beam cross-sectioning. The formation of large voids between the nanowire assembly and the substrate during cycling, causing the nanowires to lose electrical contact with the substrate, is identified as an important degradation mechanism.

Nanometer-scale Sn coatings improve the performance of silicon nanowire LIB anodes

We demonstrate that a thin partially dewetted coating of Sn will improve the cycling performance of silicon nanowire (SiNWs) lithium ion battery (LIB) anodes. The optimum architecture 3Sn/SiNWs (i.e. a Sn layer with an average film thickness of a 3 nm covering the nanowire) maintained a reversible capacity of 1865 mA h g−1 after 100 cycles at a rate of 0.1 C. This is almost double of the baseline uncoated SiNWs, where the reversible capacity after 100 cycles was 1046 mA h g−1 (∼78% improvement). The 1Sn/SiNWs and 3Sn/SiNWs electrodes demonstrated much improved cycling coulombic efficiency, with >99% vs. 94–98% for the baseline. At a high current density of 5 C, these nanocomposite offered 2× the capacity retention of bare SiNWs (∼20 vs. ∼10% of 0.1 C capacity). It is demonstrated that the Sn coating both lithiates and delithiates at a higher voltage than Si and thus imparts a compressive stress around the nanowires. This confines their radial expansion in favor of longitudinal, and reduces the well-known failure mode by lithiation-induced nanowire stranding and fracture. TOF-SIMS analysis on the post-cycled delithiated specimens shows enhanced Li signal near the current collector due to accelerated SEI formation at the interface. FIB demonstrates concurrent en-masse delamination of SEI agglomerated sections of the nanowires from the current collector. Both of these deleterious effects are lessened by the presence of the Sn coatings.

Effect of alloying magnesium with chromium and vanadium on hydrogenation kinetics studied with neutron reflectometry

The effect of chromium and vanadium alloying on the hydrogenation of a magnesium thin film is studied by neutron reflectometry. Immediate formation of a blocking MgD(2) layer is observed in pure Mg, however in the alloyed film deuteration is rapid and almost completely homogeneous.