Silicon anodes could have a lithium storage capacity of up to almost 10 times that of commercial graphite anodes (3,579 mAh/g vs. 372 mAh/g) in lithium ion batteries (LIB). However, silicon’s use is hampered by a dramatic volume change of ∼300% during lithiation, which gives rise to a range of electrochemical and mechanical effects including fracture of the particle when the particle diameter is larger than a critical value of ∼150 nm. These effects result in fast capacity decay upon cycling.

Additionally, the researchers noted, the intrinsic existence of silicon oxide on the silicon surface induces undesirable interfacial interactions that inevitably consume lithium, leading to a chain of effects that include large first-cycle capacity loss, low specific capacity, high impedance, and incomplete lithiation.

One approach to stabilizing the cycling performance of silicon is surface modification of the electrode materials. Researchers have shown that the use of aluminum oxide layers via atomic layer deposition (ALD) can stabilize highly reactive interfaces of Li-ion electrodes, effectively protecting the surface from electrolyte attack. ALD has shown itself itself to be the best method to deposit continuous, conformal, and pinhole-free films at low temperatures.

Earlier, National Renewable Energy Laboratory (NREL) materials scientist Chunmei Ban—one of the corresponding authors of this new study—and her colleagues found that they could use molecular-layer deposition (MLD) of a flexible coating onto Si electrodes to produce high-capacity Si nanocomposite anodes. (Piper 2014) The coating was an aluminum alkoxide hybrid organic
inorganic film fabricated using sequential, self-limited reactions between inorganic trimethylaluminum and organic glycerol precursors: “alucone”. The resulting nano-Si electrodes could be cycled more than 100 times with capacities of nearly 900 mAh g⁻¹ and Coulombic efficiencies in excess of 99%.

MLD-coated silicon anodes show sustainable cycling performance with a high Coulombic efficiency above 99.9%. The MLD approaches also enable independent manipulation of the mechanical properties of the composite electrode and the potential modification of the electronic and ionic conductivity. This manipulation permits optimization of the mechanical integrity without sacrificing rate capability.

However, researchers did not know how this coating improved the performance of the silicon nanoparticles. The nanoparticles naturally grow a hard shell of silicon oxide on their surface, much like stainless steel forms a protective layer of chromium oxide on its surface. No one understood if the oxide layer interfered with electrode performance, and if so, how the rubbery coating improved it.

To better understand how the coating worked, Pacific Northwest National Laboratory (PNNL) materials scientist Chongmin Wang and colleagues, including Ban, turned to expertise and a unique instrument at DOE’s Environmental Molecular Sciences Laboratory (EMSL), a DOE Office of Science User Facility at PNNL.

Ban’s group—which developed the alucone coating for silicon electrodes, called alucone, and is currently the only group that can create alucone-coated silicon particles—took high magnification images of the particles in an electron microscope. But Wang’s team has a microscope that can view the particles in action, while they are being charged and discharged. So, Yang He from the University of Pittsburgh probed both uncoated and coated silicon nanoparticles upon cyclic lithiation/delithiation using in situ transmission electron microscopy (TEM) at EMSL.

We discovered that upon initial lithiation, the native oxide layer (∼2 nm in thickness) reacts with Li to form crystalline Li₂O (c-Li₂O), which partially insulates the particle during subsequent lithiation/delithiation cycles. Porous structures form upon delithiation, and a high fraction of particles investigated appear to become difficult to delithiate following the first lithiation.

On the other hand, the alucone MLD coating process almost fully removed the native oxide layer. The alucone coating does not incur Li₂O formation and appears to possess great flexibility, showing compatible stretching or shrinking in accordance with the expansion or shrinkage of the silicon nanoparticles upon lithiation and delithiation. Conductivity measurements also indicate a remarkable improvement of conductivity of the lithiated alucone coating. Both reaction rate and reversible volume expansion are larger than that of native oxide-coated particles. The electrical, ionic, and mechanical characteristics of the surface-modifying alucone layers are thought to give rise to the different behaviors observed.

—He et al.

The team discovered that, without the alucone coating, the oxide shell prevents silicon from expanding and limits how much lithium the particle can take in when a battery charges. At the same time, they found that the alucone coating softens the particles, making it easier for them to expand and shrink with lithium.

Silicon nanoparticles coated in alucone (yellow spheres outlined in orange) expand and contract easily on charging and use. But left to their native silicon oxide covering (yellow spheres in blue), they break down fast on recharging. Source: PNNL. Click to enlarge.

The images also revealed that the rubbery alucone replaces the hard oxide, allowing the silicon to expand and contract during charging and discharging, preventing fracturing.

We were amazed that the oxide was removed. Normally it’s hard to remove an oxide. You have to use acid to do that. But this molecular deposition method that coats the particles completely changed the protective layer.

—Chongmin Wang

In addition, the particles with the oxide shells tend to merge together during charging, increasing their size and preventing lithium from permeating the silicon. The rubbery coating kept the particles separated, allowing them to function optimally.

In the future, the researchers would like to develop an easier method of coating the silicon nanoparticles.

This work was supported by the DOE Office of Energy Efficiency and Renewable Energy and PNNL.

Resources

• Yang He, Daniela Molina Piper, MengGu, Jonathan J. Travis, Steven M. George, Se-Hee Lee, Arda Genc, Lee Pullan, Jun Liu, Scott X. Mao, Ji-Guang Zhang, Chunmei Ban, and Chongmin Wang (2014) “In Situ Transmission Electron Microscopy Probing of Native Oxide and Artificial Layers on Silicon Nanoparticles for Lithium Ion Batteries,” ACS Nano doi: 10.1021/nn505523c

• Piper, D. M., Travis, J. J., Young, M., Son, S.-B., Kim, S. C., Oh, K. H., George, S. M., Ban, C. and Lee, S.-H. (2014) “Reversible High-Capacity Si Nanocomposite Anodes for Lithium-ion Batteries Enabled by Molecular Layer Deposition,” Adv. Mater., 26: 1596–1601 doi: 10.1002/adma.201304714