Highly Porous Silicon Embedded in a Ceramic Matrix: A Stable High-Capacity Electrode for LiIon Batteries Dragoljub Vrankovic,† Magdalena Graczyk-Zajac,*,† Constanze Kalcher,‡ Jochen Rohrer,*,‡ Malin Becker,† Christina Stabler,† Grzegorz Trykowski,§ Karsten Albe,‡ and Ralf Riedel† †
Disperse Feststoffe and ‡Materialmodellierung, Technische Universität Darmstadt, 64289 Darmstadt, Germany Faculty of Chemistry, Nicolaus Copernicus University in Torun, 87-100 Torun, Poland
§
S Supporting Information *
ABSTRACT: We demonstrate a cost-effective synthesis route that provides Si-based anode materials with capacities between 2000 and 3000 mAh·gSi−1 (400 and 600 mAh· gcomposite−1), Coulombic efficiencies above 99.5%, and almost 100% capacity retention over more than 100 cycles. The Si-based composite is prepared from highly porous silicon (obtained by reduction of silica) by encapsulation in an organic carbon and polymer-derived silicon oxycarbide (C/SiOC) matrix. Molecular dynamics simulations show that the highly porous silicon morphology delivers free volume for the accommodation of strain leading to no macroscopic changes during initial Li−Si alloying. In addition, a carbon layer provides an electrical contact, whereas the SiOC matrix significantly diminishes the interface between the electrolyte and the electrode material and thus suppresses the formation of a solid−electrolyte interphase on Si. Electrochemical tests of the micrometer-sized, glass-fiber-derived silicon demonstrate the up-scaling potential of the presented approach. KEYWORDS: Li-ion battery, porous silicon, molecular dynamics simulations, silicon oxycarbide, nanocomposite anode material binder or conductive additive,22−27 and (v) preparation of nanowires, nanotubes, and nanostructured particles.28−32 These approaches indeed lead to a significant improvement of cycle life and specific capacities. However, complex preparation procedures and expensive raw materials together with a relatively low areal mass loading still limit their practical application. In the present work, we propose a cost-effective and easy to implement preparation technique that provides Si-based anode materials with capacities between 2000 and 3000 mAh·gSi−1, high CE above 99.5%, and almost 100% capacity retention over 100 cycles and more. The materials consist of porous, carboncoated Si that is embedded in a ceramic SiOC matrix (Si/C/ SiOC). Key to the stable electrochemical behavior is (i) the synthesis of highly porous Si (50 to 70% of the relative porosity) by alumino- or magnesiothermic reduction of silica in molten AlCl3 described by Lin et al.33 to supply sufficient free volume for accommodation of lithium ions, (ii) carbon
S
ilicon and silicon oxide (SiOx)-based systems have long been considered as desirable anode materials for the next generation of Li-ion batteries, which is attributed to the abundance of Si, the low working potential (280%) during alloying/dealloying with lithium that can lead to disintegration of the electrode particle itself and to rupture and repetitive growth and thickening of a solid−electrolyte interphase (SEI). In consequence, this leads to low Coulombic efficiencies (CE) and capacity fading.5 Numerous advanced nanomaterial design strategies have been developed in the past to mitigate volume expansion and to stabilize SEI on the silicon electrodes. A few examples include (but are not limited to) (i) nanosizing (keeping the size of silicon nanoparticles (NPs) below a threshold value of about 150 nm),6 (ii) embedding of nanosilicon particles into active or inactive matrix materials,7−15 (iii) synthesis of nanosilicon− carbon composites with free volume around the silicon particles,1,16−21 (iv) chemical bonding of silicon NPs to a © 2017 American Chemical Society
Received: August 24, 2017 Accepted: October 23, 2017 Published: October 23, 2017 11409
DOI: 10.1021/acsnano.7b06031 ACS Nano 2017, 11, 11409−11416
Article
www.acsnano.org
Cite This: ACS Nano 2017, 11, 11409-11416
Article
ACS Nano
Figure 1. Schematics of composite preparation. Silica is first reduced by alumino/magnesiothermic reduction and then washed to obtain clean porous Si. The porous Si is further dispersed in a solution of fructose which forms a carbon coating after carbonization at 1100 °C. Finally, the C-coated porous Si (Si/C) is mixed with the preceramic commercial polymer SPR 684a (Starfire Systems) and pyrolyzed at 1100 °C, leading to embedding of Si/C particles in a silicon oxycarbide (SiOC) matrix.
coating17 of porous Si to ensure good electrical contact, and (iii) encapsulation of coated Si in carbon-rich SiOC matrix17,31,32,34−36 to minimize SEI formation at Si surfaces. In the following we give a brief description of the preparation route of Si/C/SiOC, including structural and microstructural analyses of intermediate and final products, thereby focusing on Si in the form of NPs. We then present the results of electrochemical testing and elucidate the highly reversible and stable performance with the aid of atomistic simulations. After this proof-of-principle, we demonstrate the upscaling potential of the preparation technique by replacing the Si NPs by micron-sized fibers that can easily be obtained from commercially available silica glass fibers (GFs) and whose electrochemical performance is only marginally below that of the NPs.
resulting Si samples are denoted as SiNP‑Al, SiNP‑Mg, SiGF‑Al, and SiGF‑Mg. Our X-ray diffraction (XRD) and Raman Spectroscopy measurements confirm that the prepared SiNP samples consist of phase-pure silicon (Figure 2c,d). Rietveld refinement shows that the samples are mainly amorphous with
