Within days of arriving at Oxford University in 1981, South African electrochemist Michael Thackeray made a remarkable discovery that was eventually to have deep implications for lithium battery development.

Putting lithium in its place

The recent pioneers of the lithium battery are diverse — an academic piece of research here, a minor breakthrough there — but Michael Thackeray has been an essential figure in putting lithium in its place at the heart of the computing revolution that has transformed the world this past 20 years.

Thackeray’s early post-graduate research work was spent at the Council for Scientific and Industrial Research (CSIR), where he joined the Crystallography Division of the National Physical Research Laboratory in Pretoria, his home town. He was to remain at CSIR, off and on, for two decades.

Michael Thackeray

Michael Thackeray

Here lithium electrochemistry fascinated him. Moreover, the first generation of room temperature, primary (non-rechargeable) lithium cells was being produced by industry — it was one of the hottest research topics.

In 1980, Thackeray applied to do research with John Goodenough, an Oxford University professor and then the world expert on lithium as an energy storage source. The plan was to spend a post-doctoral period with him to learn the art of room temperature lithium electrochemistry. Goodenough had recently pioneered the discovery of LiCoO2 as a lithium insertion electrode.

Spinel samples 

Because his earlier work at CSIR on the electrochemical behaviour of iron oxide electrodes in high-temperature lithium cells had shown that, in the charged state, iron oxide spinel structures were formed, Thackeray had arrived in Oxford with several spinel samples, including magnetite, Fe3O4, and Hausmannite, Mn3O4.

“At my first meeting with Goodenough, I suggested my research plan to investigate the electrochemical behaviour of spinels at room temperature. Goodenough responded gently with the comment, ‘Well, you do know that spinels are like gems, stable line-phases… so where is the space in the structure to accommodate the inserted lithium ions during discharge? By all means try, but I suggest that you look around the laboratory to see what other projects are going on.”

The mineral ‘spinel’ (MgAl2O4) is a semi-precious gem. Goodenough then left for a visit to India. He returned two weeks later to a very different situation.

Thackeray immediately set to work, first conducting a chemical reaction of lithium with Fe3O4 at room temperature to mimic the electrochemical reaction.

Thackeray observed that the magnetic Fe3O4 particles, which clung to the magnetic stirrer in the reaction vessel, gradually fell away from the stirrer on the addition of the lithium reagent, n-butyl-lithium — evidence of iron reduction and reaction between lithium and the spinel structure.

On obtaining the powder X-ray diffraction pattern and a compositional analysis of the lithiated product, Thackeray determined that, on lithiation, the cubic unit cell of the Fe3O4 host structure had expanded by approximately 3% and that there were changes in the relative intensities of the diffraction peaks, confirming the lithium insertion process.

On Goodenough’s return from India, Thackeray recalls: “I told him that lithium was indeed going into the spinel, Fe3O4. Goodenough immediately led me into his office, saying ‘Sit down, tell me all.’”

Refinement of the LixFe3O4 data showed that during the lithiation of magnetite, Fetet[Fe2]octO4, the [Fe2] octO4 spinel framework (with iron in octahedral sites) had remained intact, whereas the iron that resided in tetrahedral sites outside the framework were displaced into neighboring empty octahedral sites to make place for the uptake of one lithium ion to generate a rock salt structure (LiFe)oct[Fe2] octO4; ie, without iron extrusion as in high-temperature cells.

Thackeray reported the results of the refinement to Goodenough with similar findings for Hausmannite (Mn3O4) that, on lithiation, formed the corresponding ordered rock salt structure (LiMn)oct[Mn2]octO4.

The results had immediate scientific and technological implications.

Goodenough, familiar with the spinel structure from his earlier research on their magnetic properties, suggested an investigation of the lithium spinel Litet[Mn2]octO4, which accommodated lithium by the same principle as Fe3O4 and Mn3O4 to form the ordered rock salt configuration (Li2) oct[Mn2]octO4; in this case, the interstitial space of the [Mn2]octO4 spinel framework contained only lithium ions that could migrate, unimpeded, through a three-dimensional network of face-sharing tetrahedra and octahedra, providing fast kinetics and a high power electrode.

This reaction occurred electrochemically at 3V in a lithium cell. Because the concept of spinel electrodes had originated at CSIR, Goodenough graciously agreed to cede title of the spinel patent that was filed to SAIDCOR, one of the South African sponsors of Thackeray’s visit to Oxford, later to be licensed to industry.

Over recent years, Li[Mn2]O4 has been widely exploited. It is often used as a blend in the cathodes of lithium-ion batteries, notably for portable electronics and electric vehicles, such as the all-electric Nissan Leaf and the hybrid-electric Chevy Volt.

Return to CSIR 

Thackeray returned to South Africa at the end of 1982. He remained at CSIR, where he established a new battery team to build on the lithium battery materials research he had initiated, while providing R&D support to the South African battery industry, including among others, Zebra Power Systems, Willard Batteries, a producer of lead acid batteries, and Delta EMD, a manufacturer of electrolytic manganese dioxide (EMD) for the alkaline (Zn/MnO2) battery market.

Over the next 10 years, Thackeray and his CSIR team continued to innovate the design of new lithium battery electrode materials, structures and compositions, first focusing, on his return from Oxford, on the family of lithium spinels containing manganese, vanadium or iron.

In 1994, Ernst Ferg, Gummow, de Kock and Thackeray demonstrated that safe 2.5V lithium-ion cells could be fabricated by coupling a lithium titanate spinel anode (Li4Ti5O12), which operates 1.5V above the potential of metallic lithium, with high voltage cathodes, notably a stabilized Li1+xMn2-xO4 spinel — a system that is being exploited for devices requiring high power batteries, such as hybrid electric vehicles.

The CSIR team adopted strategies to engineer and patent new lithium insertion materials, which included manganese oxide electrodes with one-dimensional, two-dimensional and three-dimensional pathways for lithium-ion transport; they were successful on all three counts, designing and evaluating lithia-stabilized alpha- MnO2, lithia-stabilized layered-MnO2 and lithia-stabilized spinel-MnO2 electrode materials, respectively.

Anhydrous-layered MnO2 structures were unknown at the time.

Here chance, or good fortune, intervened. At an Electrochemical Society Meeting in Toronto in October 1992, Thackeray met Don Vissers, head of the Battery Department at Argonne National Laboratory.

Vissers invited Thackeray to lead a materials R&D effort at Argonne for a new lithium-polymer battery project to be sponsored by the US Department of Energy and the United States Advanced Battery Consortium (USABC). Thackeray, sensing a bright future for lithium battery technology and recognizing an excellent opportunity, accepted. He and his family left in January 1994.

As with CSIR and Oxford, Thackeray got off to a quick start — within two months of his arrival, while working on a high energy, all solid state, lithium-polymer battery project for electric vehicles supported by the US Department of Energy, 3M Corporation and Hydro-Quebec, he had identified a lithia-stabilized vanadium oxide cathode material (based on his earlier research at CSIR) that provided the 3M/HQ cells with 30% more energy and superior power relative to the original material being used.

This materials technology was later transferred to, and scaled up by, 3M and implemented by Hydro-Quebec/ Avestor in commercial battery products for stationary energy storage. However, subsequent problems related to lithium dendrite formation and short circuiting on long-term cycling ultimately led to the withdrawal of the batteries from the market.

Towards the end of the 1990s, noticing that other research groups were initiating studies on Li2MnO3, Thackeray, Johnson, Khalil Amine, Jaekook Kim and an expanding Argonne team intensified their efforts to exploit composite ‘layered-layered’ xLi2MnO3•(1-x)LiMO2 (M=Mn, Ni, Co) and ‘layered-spinel’ xLi2MnO3•(1-x)LiM2O4 (M-Mn, Li, Ni, Co) electrode structures, which formed the basis of a broad patent portfolio.

Argonne’s intellectual property was subsequently licensed worldwide to major lithium battery materials and cell manufacturers. The first generation Chevy Volt, in particular, uses a lithium-ion battery with a blend of the Argonne composite materials and stabilized LiMn2O4 spinel in the cathode.