Exploiting dopant characteristics to enhance the performance of lithium garnet solid electrolytes

Stockham, Mark ORCID: 0000-0001-7254-0266 (2022). Exploiting dopant characteristics to enhance the performance of lithium garnet solid electrolytes. University of Birmingham. Ph.D.

[img]
Preview
Stockham2022PhD_Redacted.pdf
Text - Redacted Version
Available under License All rights reserved.

Download (46MB) | Preview

Abstract

Lithium garnet materials are promising candidates as inorganic solid state electrolytes for use in all solid state batteries, owing to their high ionic conductivity, wide electrochemical window and high safety. However, garnets suffer from energy demanding synthesis, rapid proton exchange and high interfacial resistance. Furthermore, even the most promising lithium garnet material (Ga\(_x\)Li\(_{7-3x}\)La\(_3\)Zr\(_2\)O\(_{12}\)) requires specialised handling and has not reached the conductivity of current solvent based electrolytes. In this work alternative dopant strategies into the lithium garnet system are investigated to unlock new approaches to solid state batteries.

Firstly, the flexibility of the Pr dopant was explored, with primary aims to increase the Li conductivity in Li\(_5\)La\(_3\)Nb\(_2\)O\(_{12}\) based systems. Here, Pr was doped onto the Nb and La site with Li\(_{5+x}\)La\(_3\)Nb\(_{2-x}\)Pr\(_x\)O\({12}\) and Li\(_5\)La\(_{3-x}\)Pr\(_x\)Nb\(_2\)O\(_{12}\) prepared respectively. The Li\(_5.8\)La\(_3\)Nb\(_{1.2}\)Pr0.8O12 system had a room temperature conductivity of 0.41 mS cm\(^{-1}\), the highest reported for a lithium garnet with < 6 Li per formula unit.

Secondly, Hf based garnets have been reported to potentially offer increased electrochemical stability compared to the Zr analogue. Therefore, Ga\(_x\)Li\(_{7-3x}\)La\(_3\)Hf\(_2\)O\(_{12}\) was synthesised for the first time and its electrochemical properties assessed, with results indicating increased critical current density compared to Ga\(_x\)Li\(_{7-3x}\)La\(_3\)Zr\(_2\)O\(_{12}\). This work also tackled the energy demanding synthesis of garnet materials and proposed a new lower temperature water-based route.

Amongst the best performing lithium garnet materials are the Li substituted M\(_x\)Li\(_{7-3x}\)La\(_3\)Zr\(_2\)O\(_{12}\) (M = Ga, Al) systems, however both dopants exsolve to the grain boundary when heated to the high temperatures required for densification. This phenomenon, although posing some potential problems, has not been explored as a method to reduce garnet/Li metal resistance via Ga/Al alloying. In this work, the Ga dopant was deliberately destabilised by substitution of the smaller Nd for La to form Ga\(_x\)Li\(_{7-3x}\)Nd\(_2\)Zr\(_2\)O\(_{12}\). The Li/garnet interfacial resistance in these systems was determined to be 67 Ω cm\(^2\), compared to 949 Ω cm\(^2\) for the undoped Li\(_7\)Nd\(_3\)Zr\(_2\)O\(_{12}\), with results indicating significant Ga exsolution after Li metal contact. A range of Nd garnets were also studied for the first time.

Much work has been conducted on the cubic lithium garnets, but comparatively little on the tetragonal systems due to their poorer ionic mobility. These systems, however, transition to a highly conductive cubic phase at high temperature, thought to be driven by entropic factors and increased unit cell size. It would, therefore, be ideal if this transition was reduced to room temperature, but this first requires a greater understanding of the tetragonal-cubic transition. To that end, a series of nine tetragonal garnets were made; A\(_3\)B\(_2\)Li\(_7\)O\(_12\) (A=La, Pr, Nd) (B=Zr, Hf), La\(_3\)Zr\(_{1.75}\)Ce\(_{0.25}\)Li\(_7\)O\(_{12}\) and LaSr\(_2\)B\(_2\)Li\(_7\)O\(_{12}\) (B=Nb, Ta). It was shown that a dependence on lattice parameters alone is inaccurate, however the B site dopant plays a critical role.

Next, the compositional flexibility of the lithium garnet materials was tested with two high entropy garnet systems; Ga\(_{0.2}\)Li\(_{5.75}\)La\(_{2.5}\)Nd\(_{0.5}\)Nb\(_{0.65}\)Ce\(_{0.1}\)Zr\(_1\)Ti\(_{0.25}\)O\(_{12}\) and Ga\(_{0.2}\)Li\(_{5.75}\)La\(_{2.5}\)Nd\(_{0.5}\)Nb\(_{0.35}\)Ta\(_{0.3}\)Ce\(_{0.1}\)Zr\(_{0.75}\)Hf\(_{0.25}\)Ti\(_{0.25}\)O\(_{12}\). Here, not only were high performing membranes obtained, especially with such low Li content, numerous dopant properties were exploited simultaneously and both systems (outside of stoichiometric weighing) proved very simple to form.

Finally, additional work not yet published, concentrates upon prior reports which indicate doping tetragonal LLZO with Ce reduces Li metal interfacial resistance. Here, a new co-doped Li\(_{6.5}\)La\(_3\)Nb\(_{0.5}\)Zr\(_1\)Ti\(_{0.25}\)Ce\(_{0.25}\)O\(_{12}\) system was made to assess if the favourable interfacial properties are maintained, but also if the increased lattice parameters (in a cubic cell) enable better performance. Curiously, not only is this found true, but this system also shows extremely rapid, and simultaneous, sintering and densification (<1h at 1100°C at 100 °C min\(^{-1}\)). This garnet was also shown to be synthetically scalable (>15g) and to react, and remove, short circuits caused by lithium dendrites after short rest periods.

Secondly, Hf based garnets have been reported to potentially offer increased electrochemical stability compared to the Zr analogue. Therefore, Ga\(_x\)Li\(_{7-3x}\)La\(_3\)Hf\(_2\)O\(_{12}\) was synthesised for the first time and its electrochemical properties assessed, with results indicating increased critical current density compared to Ga\(_x\)Li\(_{7-3x}\)La\(_3\)Zr\(_2\)O\(_{12}\). This work also tackled the energy demanding synthesis of garnet materials and proposed a new lower temperature water-based route.

Amongst the best performing lithium garnet materials are the Li substituted M\(_x\)Li\(_{7-3x}\)La\(_3\)Zr\(_2\)O\(_{12}\) (M = Ga, Al) systems, however both dopants exsolve to the grain boundary when heated to the high temperatures required for densification. This phenomenon, although posing some potential problems, has not been explored as a method to reduce garnet/Li metal resistance via Ga/Al alloying. In this work, the Ga dopant was deliberately destabilised by substitution of the smaller Nd for La to form Ga\(_x\)Li\(_{7-3x}\)Nd\(_2\)Zr\(_2\)O\(_{12}\). The Li/garnet interfacial resistance in these systems was determined to be 67 Ω cm\(^2\), compared to 949 Ω cm\(^2\) for the undoped Li\(_7\)Nd\(_3\)Zr\(_2\)O\(_{12}\), with results indicating significant Ga exsolution after Li metal contact. A range of Nd garnets were also studied for the first time.

Much work has been conducted on the cubic lithium garnets, but comparatively little on the tetragonal systems due to their poorer ionic mobility. These systems, however, transition to a highly conductive cubic phase at high temperature, thought to be driven by entropic factors and increased unit cell size. It would, therefore, be ideal if this transition was reduced to room temperature, but this first requires a greater understanding of the tetragonal-cubic transition. To that end, a series of nine tetragonal garnets were made; A\(_3\)B\(_2\)Li\(_7\)O\(_12\) (A=La, Pr, Nd) (B=Zr, Hf), La\(_3\)Zr\(_{1.75}\)Ce\(_{0.25}|)Li\(_7\)O\(_{12}\) and LaSr\(_2\)B\(_2\)Li\(_7\)O\(_{12}\) (B=Nb, Ta). It was shown that a dependence on lattice parameters alone is inaccurate, however the B site dopant plays a critical role.

Next, the compositional flexibility of the lithium garnet materials was tested with two high entropy garnet systems; Ga\(_{0.2}\)Li\(_{5.75}\)La\(_{2.5}\)Nd\(_{0.5}\)Nb\(_{0.65}\)Ce\(_{0.1}\)Zr\(_1\)Ti\(_{0.25}\)O\(_{12}\) and Ga\(_{0.2}\)Li\(_{5.75}\)La\(_{2.5}\)Nd\(_{0.5}\)Nb\(_{0.35}\)Ta\(_{0.3}\)Ce\(_{0.1}\)Zr\(_{0.75}\)Hf\(_{0.25}\)Ti\(_{0.25}\)O\(_{12}\). Here, not only were high performing membranes obtained, especially with such low Li content, numerous dopant properties were exploited simultaneously and both systems (outside of stoichiometric weighing) proved very simple to form.

Finally, additional work not yet published, concentrates upon prior reports which indicate doping tetragonal LLZO with Ce reduces Li metal interfacial resistance. Here, a new co-doped Li\(_{6.5}\)La\(_3\)Nb\(_{0.5}\)Zr\(_1\)Ti\(_{0.25}\)Ce\(_{0.25}\)O\(_{12}\) system was made to assess if the favourable interfacial properties are maintained, but also if the increased lattice parameters (in a cubic cell) enable better performance. Curiously, not only is this found true, but this system also shows extremely rapid, and simultaneous, sintering and densification (<1h at 1100°C at 100 °C min\(^{-1}\)). This garnet was also shown to be synthetically scalable (>15g) and to react, and remove, short circuits caused by lithium dendrites after short rest periods.

Type of Work: Thesis (Doctorates > Ph.D.)
Award Type: Doctorates > Ph.D.
Supervisor(s):
Supervisor(s)EmailORCID
Slater, PeterUNSPECIFIEDorcid.org/0000-0002-6280-7673
Ding, YulongUNSPECIFIEDorcid.org/0000-0001-8490-5349
Li, YongliangUNSPECIFIEDorcid.org/0000-0001-6231-015X
Licence: All rights reserved
College/Faculty: Colleges (2008 onwards) > College of Engineering & Physical Sciences
School or Department: School of Chemistry
Funders: Other
Other Funders: University of Birmingham
Subjects: Q Science > QD Chemistry
URI: http://etheses.bham.ac.uk/id/eprint/13169

Actions

Request a Correction Request a Correction
View Item View Item

Downloads

Downloads per month over past year