The Effect of Surface Functionalization on the Electrochemical Lithiation of Silicon for Li Ion Batteries

Lithium-ion batteries are widely used in consumer electronics and are commercially available for transportation applications. However, the limited energy capacities of current electrode materials mean that widespread applications in transportation are potentially limited. Therefore, there are substantial research efforts on new materials with higher capacities than those of current electrode materials. Among them, silicon has attracted widespread attention as an alternative anode material to graphite due to its large gravimetric and volumetric capacities. The lithiation of crystalline Si, however, results in significant structural changes as compared to transition metal oxides or graphitic material. The large volume expansion increases the interaction with electrolyte leading to detrimental effects on the capacity.

Si atom has four valence electrons, requiring four bonds to saturate its valence shell. In the presence of H and O in the vicinity of Si surface, surface Si atoms are saturated with these passivating agents. It is therefore often observed that Si surfaces are hydrogenated or oxidized. It has also been suggested that Mn can be present on the anode, coming from dissolution of cathode materials such as lithium manganese oxide. Given these considerations, the motivations of this study are thus twofold.  The first such motivation is to build an atomistic-level understanding of the effects of the realistic surface terminations on the lithiation process. This analysis complements and extends previous work on the lithiation of clean Si surfaces.¹ The second motivation is that, because of the generally recognized importance of SEI layer, which forms on the electrode surface as a result of decomposition of thermodynamically unstable electrolyte molecules, we develop strategies to predict which types of functionalized surfaces may impact SEI formation. Thus, we studied here how the functionalization impacts the lithiation process itself.

In this talk, we will present the results of an extensive computational study for the lithiation of functionalized Si surfaces. First principles calculations, with a history-dependent lithium insertion algorithm, are used to model the mechanism of lithiation of the functionalized crystalline Si. The lithiated configurations for the Mn/Si(100), (110), and (111) surfaces at different lithiation stages show the progression of amorphous crystalline boundary with increasing Li content. For all surfaces, we find that Si is amorphized by breakdown of Si-Si chains upon Li insertion. For Mn/Si(111), the sheets of connected Si atoms are formed and maintained up to the highest Li content. At high Li concentration (x>3), some isolated Si atoms are formed, in particular for Mn/Si(110) and Mn/Si(111) surfaces. For all surfaces, Mn is found trapped below the surface, underneath the Si dimers. The analysis of the voltage curves reveals both similar and different characteristics with those of the clean Si surfaces. (110) surface has the highest voltage plateau as compared to other surfaces. We also find the maximum Li that can be inserted to vary from one surface to another, and the trends differ from those of the clean Si surfaces. For hydroxilated Si surface, on the other hand, we find the lithiated configurations resemble to those of clean Si surfaces; formation of Si dimers are observed as well as few isolated Si atoms at high Li content. The voltage plot suggests an additional voltage drop at low Li content originating from the presence of OH/H species on the surface, and a slight reduction in the maximum Li that can be inserted as compared to the clean Si surfaces. The atomistic and electronic structure origins of the differences will be discussed.

[1] M. K. Y. Chan, C. Wolverton, and J. Greeley, Journal of the American Chemical Society 134 (2012) 14362