Introduction

Lithium-silicon batteries are a subclass of the well-known lithium-ion batteries. The difference is that they replace the typical anode material, graphite, with lithium-based composites or materials such as elemental silicon, silicon oxides or graphite combined with different silicon structures. Most of these have yet to see any commercial usage but have highly desirable qualities for example higher specific capacity. Also, considering the cost and the scarcity of typical battery electrode materials, silicon is a prime candidate as one of the most abundant materials in earths crust while also being non-toxic.

In regards to battery capacity, choosing a higher capacity anode is paramount, as the main factor in defining the overall capacity of the battery is the capacity of negative anode^[1] Total cell capacity is however calculated from both cathode and anode capacities which might present situations where choosing a lower capacity option is preferable if improvements in battery lifetime may be had.

Composition

Cathode

As is typical for lithium-ion batteries, the cathode material consists of some lithium containing metal oxides (LiMO_x) such as lithium cobolt oxide (LiCoO₂), lithium manganese oxide (LiMn₂O₄) or Lithium iron phosphate (LiFePO₄). These cathode materials have mostly replaced lithium-free cathodes as highly reactive lithium anode is not needed. ^[3]

Anode

When lithiated, silicon can form different compounds with lithium but theoretically, each silicon atom can "hold" up to 4,4  lithium atoms per silicon compared to 1 lithium per 6 carbon atoms, thus, the theoretical maximum capacity of the batteries using silicon containing anodes should be higher compared to typical anode materials.. This theoretical maximum is achieved with a silicon-lithium alloy Li₂₂Si₅. Many other silicon anode materials have been proposed such as utilizing understoichiometric silicon oxide or graphite combined with silicon nano structures.

Si-Li alloys

Silicon-lithium alloys have been already studied since the discovery of Li₂₂Si₅ in 1966. Multiple Si-Li phases such as LiSi, Li₁₂Si₇, Li₁₅Si₄ and Li₂₂Si₅ exist simultaneosly during alloying process, however, amorphous phases for Li_xSi alloys (x=0-3,75) are favoured during lithiation. Amorphous phases on the other hand are more prone to larger volume changes compared to their crystalline counterparts.^[1] Lithium silicide (15/4) forms in cubic crystal system belonging to the I-43d (220) space group. During lithiation, lithium atoms are trapped inside the silicon structures due to Li-Li bond being shorter than Li-Si. ^[4]

Figure 2. Crystal structure of Li₁₅Si₄. Figure by Joona Vilander, Crystal Structure from ICSD

The main advantage in using silicon in battery lithium-ion battery anodes is it's larger theoretical specific capacity compared to the typical anode material, graphite. Li₁₅Si₄ for example, can reach specific capacity of 4200 mAhg^-1 compared graphite's lower 372 mAhg^-1. ^[5] Even though elemental silicon and some silicon-lithium alloys have the highest theoretical maximum capacity of any silicon based anode, they have seen little use however due to their inherent problems when used as battery anodes. Electrode reaction for Li15Si4 in room temperature is as follows. 

\( 4 Si + 15 Li \rightarrow Li_15Si_4 \ \ (3578 \ mAhg^{-1}) \)

The theoretical capacity of the alloy product doesn't quite reach the maximum of 4200 mAhg^-1 of Li₂₂Si₅ but is still quite high at 3578 mAhg^-1.  

Electrolyte

Commercial lithium batteries typically use organic solvents capable of dissolving lithium salts. In the electrolyte solution, lithium-ions act as charge carriers between the cathode and anode. A good electrolyte should balance the viscosity, ionic conductivity and the ability to form a stable enough passivation layer. ^[2] The formation of the passivation layer called solid electrolyte interphase layer (SEI) begins during the lithiation as a result of degrading electrolyte. SEI's role is crucial in battery operation as it prevents further electrolyte degradation during consequent lithiation and dilithiation cycles. This, however, presents a problem for silicon anode batteries as the cracks in SEI films caused by large volume changes lead to formation of new SEI layers. Continuous formation of SEI degrades the electrolyte even further causing issues for battery cycling properties. ^[1]

The following reversible electrochemical reaction occurs during charging and discharging of Si-Li batteries. ^[6]

Charging process:

Discharging process:

Challenges with silicon

The main difficulty in creating lithium-silicon batteries is silicon swelling i.e. the tendency for silicon to undergo large volume changes (up to 300%) during lithiation process.^[6]  This is due to silicon's ability to accommodate a high amount of lithium atoms during lithiation process (charge/recharge) which can be seen in the enormous difference in unit cell volume between silicon (40,88 ų) and Li₂₂Si₅ (1617 ų). ^[5] The swelling leads to high mechanical stress inside the electrode causing cracks and delamination which decreases overall capacity of the battery and thus its cycle lifetime. Coupled with the formation of unstable SEI layer on the anode, pulverization may occur which causes permanent capacity losses. Uncontrolled SEI layer formation may also lead to nucleation and dendrite formation which ultimately can damage the separator leading to short circuiting of the battery cell. ^[6]  In addition to the previously mentioned problems, silicon suffers from slow reaction kinetics due to low electronic conductivity of only 10^-5 to 10^-3 S cm^-1 and Li-ion diffusion rate from 10^-14 to 10^-12 CM² s^-1 causing silicon batteries to have low cycling efficiency which can limit the utilization of the battery's full capacity. ^[6]

Alternative silicon containing anodes

Understoichiometric silicon oxide (SiO_x, 0<x<2)

Utilizing understhoichimetric silicon dioxides provides some benefits over silicon-lithium alloys as it is easier to handle in non-regulated atmosphere compared to elemental silicon due to lower reactivity. The resulting Li₄SiO₄ forms in monoclinic crystal system and belongs to the space group P121/m1 (11).^[7]

Figure 3. Crystal structure of Li₄SiO_4. Figure by Joona Vilander, crystal structure from ICSD

The precise stchoiciometry of SiO_x is still somewhat under debate but evidence points to it being a mixture of SiO₂, intermediate phase with less oxygen and elemental silicon. The following reaction mechanism presents the alloying of silicon with lithium.

\( 4 SiO + 17.2 Li \rightarrow 3 Si + Li4_4SiO_4 \ (608 \ mAhg^{-1}) \rightarrow 3Li_{3.75}Si + Li_4SiO_4 \ (1708 \ mAhg^{-1}) \)

The initial reaction is irreversible as only elemental silicon can reversibly alloy with lithium, thus the formed lithium silicate remains inactive during the second reaction. The low capacity of the irreversible reaction is due to the higher oxygen content.^[8] However, this Li4SiO4 matrix can mitigate the volume changes in the anode structure also enhancing the lithium diffusion. ^[8]

Silicon nanomaterials

Most recent interest in silicon batteries has been with the utilization of silicon nanostructures in lieu of a typical bulk material such Li-Si alloys. Compared to previously mentioned bulk materials, anodes containing nanostructures can mitigate massive volume changes better as more free space is provided into the anode structure when utilizing nanoparticles such as 0D nanoparticles, porous silicon or nanotubes.^[9] One of the more promising options of these nanostructures is silicon 0D nanoparticles as their synthesis methods are already commercially available. These can further be improved with novel binder materials such as carboxymethyl cellusole (NaCMC) instead of typical poly(vinylidene fluoride) (PVDF). Using this binder material, a capacity of 1200 mAhg-1 was maintained through 70 cycles due to the materials or polyacrylic acid (PAA) with capacity of 2500 mAhg-1 through 100 cycles. Binder materials help in creating a deformable and more stable SEI layer while also giving access to lithium atoms to the surface of silicon.^[9]

Another approach would be using silicon nanotubes (1D) which would essentially lower the lithium diffusion distance as both inner and outer walls of the nanotubes are exposed to the electrolyte solution. The additional void inside the nanotubes also allows for more expansion without compromising internal structure of the anode, preventing pulverization.^[9] 

References