Considerations for Nanomaterials in Lithium Rechargeable Batteries

W. F. Howard, Ph.D. Edmond, Oklahoma

Introduction

$8.6 billion: that’s the estimated worldwide nanotechnology research funding for 2004 (Lux Research report, 16 August 2004). Here in the U.S., the 21st Century Nanotech Research Act will pump $3.7 billion into the field over the next four years. In 2005, private support is projected to exceed government-sponsored efforts for the first time. But nanotechnology is not an industry unto itself; rather, it has applications across many sectors, and has become THE buzzword for active inorganic materials.

Potential applications for nanomaterials are legion. This is especially true in lithium rechargeable batteries (LiRB), and teams of scientists and engineers are leaping at the opportunities therein. Nanometric particles allow the circumvention of certain limitations common to micron-sized powders (improved rate capability is most frequently cited), but as with any novel technology, new hurdles must be overcome, both in powder and electrode production. This paper focuses on nanoparticulate electrode materials, discusses the challenges facing this fledgling industry, and suggests promising LiRB applications for such materials.

Background

What makes nanopowders behave differently from conventionally-sized particles (0.1-100µm) with an identical chemical formula? (I define nano- as referring to atomic clusters or molecules with a maximum dimension of 25nm.) It is well known that the properties of bulk catalysts are distinct from those of dispersed atomic or molecular clusters with the same elemental composition. Intercalation materials require the insertion (or extraction) of Li+ into (or out of) the crystal lattice, thus we have a bulk process. But when we consider nanoscale proportions, particles approach two dimensions, and intercalation becomes a surface reaction, a major distinction.

The ramifications of this crystallinity loss (i.e., z ~0 in the molecular lattice) are several. The first, as mentioned above, is an accelerated Li+ uptake (or release), manifested by improved rate capability. Simply put, the lithium ion has no transport restrictions within the electrode material: the electrochemical reaction takes place on the particle surface, without wending its way through oxygen-bounded channels or planes. Olivines (for example, LiFePO4) and lamellar Li(Mn/Ni)O2 compounds, with indifferent performance above 2C charge/discharge rates, could benefit greatly from nanotechnology.

The second advantage to nanomaterials is the likelihood of capacity enhancement. There are few, if any, dead spots within the “crystal” since there are no grain or phase boundaries truncating Li+ movement. With all the active sites working, capacities should approach their theoretical limits. Top candidates for capacity improvement are members of the Li(Co,Ni)O2 family.

Finally, it is reasonable to expect reduced capacity fade from structural issues. Again, the lack of a three-dimensional crystal structure is beneficial: there are no channels or planes to distort during cycling, therefore Li+ transport remains unfettered. In this case, metal oxides such as MnO2 or TiO2, normally considered as poor rechargeable species with large first-cycle capacity losses, have exhibited surprisingly good performance in lab tests.

Before declaring victory and constructing huge manufacturing facilities, remember that batteries have numerous components. What effect will nanosizing have on the interaction between electrolytes and cathode materials? Most metals with multiple oxidation states, typically found in Li intercalation compounds, are active catalysts. High voltage cathodes become even more reactive as the particle size shrinks, and organic solvents may not be robust enough to withstand the electrochemical pressure. A possible alternative, still very much in the development stage, is quaternary amine ionic liquids (MacFarlane), with voltage windows exceeding 5V. Polymer electrolytes (so far) don’t support pulse operations, which would negate one of the nanoparticles’ attributes.

Other pertinent considerations are the high surface area and low density common to nanomaterials. The first will result in significantly more solvent (or polymer) than is usually necessary for cathode production to prepare slurries suitable for coating. This leads to the costly question of solvent removal and subsequent recovery or disposal. Secondly, low density nanoparticles will provide a loading challenge. Even with rigorous wetting and compression techniques, it will be difficult to construct batteries with enough active material to achieve energy output objectives.

Candidate Materials

Not all intercalation compounds will benefit from a reduction in particle size. Following is an overview of existing and potential LiRB electrode materials, and the pros and cons of utilizing nanoparticles of such species. In this context, it is assumed that the intercalation materials are uncoated (nanoparticle coating is very expensive), but may be modified for stability with any of several dopants. The caveat, as always, is that technology marches on, and today’s lemons become tomorrow’s lemonade.

Lithium manganese oxide (spinel or LMO) is severely disadvantaged by the instability of Mn+3, the electrochemically-active ion, and experience has shown that reducing LMO particle size hastens capacity fade. Without a major breakthrough in electrolyte chemistry (i.e., acid-free solutions), spinel is a poor candidate for nanotechnology. Further, spinel is one of the least expensive cathode materials, an advantage sure to be lost with nano-LMO production.

The layered lithium metal oxides, including LiCoO2 (LCO), LiCo1-xNixO2, and the LiMxMnyNizO2 (LMN) family, have mediocre high rate or pulse capabilities, and smaller particle size should be beneficial. LCO and derivatives tend to deoxygenate at high voltage, a destructive and potentially dangerous property exacerbated by nanoscale particles (more reactive crystal edges). A further concern is the catalytic activity of Co and Ni, which would be greater with nanometric clusters compared to micron crystallites. The high capacity (250-325mAh/g) found with LMN compounds, however, suggests that these intercalants warrant further exploration. LMN syntheses already require a solution (or gel) phase to ensure homogeneous products, thus pointing to spray pyrolysis or related nanotech processes.

Lithium iron phosphate, LiFePO4, is an emerging cathode material that may not incur an advantage from nanotechnology. This compound, prepared by anaerobic solid state calcination, already exhibits near-theoretical capacity (170mAh/g) and extremely low fade. There are also indications that high rate performance is enhanced by (proprietary) modifications to the reaction process. Other olivines, with their complex structures that slow Li+ transport, may be better nanotech candidates.

Vanadium oxides, especially Li1+xV3O8, are strong oxygen coordinators at high potentials and have a tendency to dissolve into the electrolyte. Rate capabilities of nanoparticulate vanadium intercalants would likely be enhanced, however. As with spinel, new electrolytes are required to get the full value from vanadium-based cathode materials.

Binary transition metal oxides, such as TiO2, MnO2, Fe2O3, CrO3, etc., typically undergo irreversible structural damage with Li+ intercalation, with debilitating capacity fade. Exploratory efforts with Ti and Mn oxide cathodes show MOx nanoparticles have very reversible behavior and high capacity in Li cells. So far, the process costs outweigh the performance benefits, but this may be a product looking for an application. The encouraging results dictate additional work with these compounds.

The real action for nanomaterials in rechargeable batteries will likely reside with the anode, which falls into two categories: metals (alloys) and intercalants. The unfortunate truth about nano-sized metal powders is that they are pyrophoric, and safety-conscious manufacturers will look elsewhere. Non- or semi-metallics, such as group IVb elements Si or Ge, alloy with Li and may have enough oxidation resistance to be viable anode candidates. Yazami suggested 10nm Si is the optimum anode material, although it is not commercially available. Minimizing particle size ameliorates the problem of large volume changes during Li alloying or extraction, but raises formidable economic barriers.

Intercalating metal oxide anodes are bit players in this opera: only Li4Ti5O12 (LTO) has received much press. LTO is very stable, has excellent high rate performance, and is relatively easy to make. The barrier to LTO acceptance is its high operating potential (~1.5V), which substantially reduces cell energy output. There are numerous metal oxides, lithiated or not, that will intercalate Li+ but share a common weakness: during the first charge, their crystal structures are irreversibly damaged by Li retention. This capture causes up to 50% loss of theoretical capacity in the first charge-discharge cycle, even though capacity may stabilize during subsequent battery operation. Here is an opportunity for anode nanomaterials, which would restrict Li+ to surface absorption and minimize structural degradation.

Possibilities in this category include LiMOx species (M is W, Mo, or multi-metallic), complex Nb-, Zr-, or Hf-based oxides, cermets (metal borides, carbides, etc.), and inverse spinels (LiMVO4). First-cycle capacity hysteresis with these species as micron-sized powders is formidable, typically 20-50%. The primary objection to the group VIa compounds is the relatively high reduction potential (1.2-1.6V). Group IVa and Va oxides have lower potentials, but require process temperatures above 1100°C for phase purity, beyond the reach of conventional furnaces. Cermets also need specialized process equipment capable of holding non-oxidizing atmospheres above 1000°C. Both types of anode materials are synthesized by processes compatible with nanotechnology, however. Finally, inverse spinels intercalate Li below 1V, but until recently were considered only as (poor) high-voltage cathode materials. Preferred reaction routes include a solution step, again pointing to the feasibility of a spray-type process. Although speculative, the intercalating materials in this paragraph are good candidates for exploratory nanotech research.