136
10 Conclusions and Future Works
This material, Li 1.2 Ni 0.2 Mn 0.6 O 2 , is in the “bump” region in the layered material on
the Li–Mn–Ni–O face in Fig. 10.1. Materials in this part of the phase diagram were
shown to contain metal site vacancies in Chap. 8. The presence of some vacancies
could very well account for the lowering of the overall magnetic moment of the samples. X-ray absorption near-edge structure (XANES) and redox titrations confirmed
that manganese was in fact in the 4+ state in these materials. This work also demonstrated that there is a solid solution line from Li 2 MnO 3 to Li[Ni 1/6 P 1/6 Mn 2/3 ]O 2 .
Along this line, all the transition metal layers have 2/3 manganese occupation allowing for ordering on two
√
3 x
√
3 superlattices. This gave rise to sharp superlattice
peaks in the XRD patterns and the Monte Carlo simulation demonstrated that these
structures do not phase separate into layered–layered composites.
VIII Some layered materials in the Li–Mn–Ni–O system lose a significant amount
of lithium during synthesis. Typically, researchers counter this by adding a small
amount of excess lithium and assume it is lost during synthesis (e.g. Refs. [51, 52]).
However, the phase diagrams here show that most lithium-rich structures in the Li–
Mn–Ni–O system can accommodate more lithium without phase separating. It can
no longer be assumed that all excess lithium is lost during synthesis. Elemental analysis is therefore required to determine the composition of samples after heating. The
contour plots generated in Chap. 7 can also be used to obtain approximate compositions. These contours show that as one adds excess lithium the c axis decreases
rapidly, while the a axis is reduced more gradually. As such, it is vital to use both
parameters in identifying final materials. Recognizing the actual compositions of the
sample also helps explain differences in electrochemistry found in the literature. For
example, Lin et al. [53] made a LiNi 0.5 Mn 0.5 O 2 sample without excess lithium that
showed a lower capacity than materials made by Lu et al. [2]. The samples made
by Lu used LiOH to cause coprecipitation of the mixed transition metal hydroxide
precursor such that a small amount of excess lithium was inevitable. As such, the Lu
sample was just within the layered region while the Lin sample was slightly outside.
In the previous chapter, it was shown that this results in the single-phase material
having a significantly higher capacity, consistent with the published results from Lu
and Lin. A more extreme example of making LiNi 0.5 Mn 0.5 O 2 with excess lithium
was published by Ohzuku et al. [52]. The contour plots were used to demonstrate that
this sample lost little of the excess lithium during synthesis and the electrochemical
performance was therefore attributed to being a lithium-rich layered material.
Two articles show very strange XRD patterns for LiNi 0.5 Mn 0.5 O 2 .
IX The first was published by Jo et al. [54] and showed extreme peak broadening
when regular cooled in oxygen. This was attributed to crystallites on the order of
14 nm. Here, it was shown that samples phase separate into layered-layered composites under such conditions. Therefore, the peak broadening seen by Jo was due to
phase separation and not small crystallites since 14 nm is extremely small for having
been heated for 3 h at 800
◦ C.
X The second strange result was obtained by Hinuma et al. [55] who found
that a single-phase LiNi 0.5 Mn 0.5 O 2 underwent a complex phase separation upon
annealing at 600
◦ C and a return to single-phase upon reheating to 1,000
◦ C. The
new phases appearing upon annealing were not identified in the original paper. The
transformations can now be understood. The layered boundary is high in the Gibbs
10 Conclusions and Future Works
This material, Li 1.2 Ni 0.2 Mn 0.6 O 2 , is in the “bump” region in the layered material on
the Li–Mn–Ni–O face in Fig. 10.1. Materials in this part of the phase diagram were
shown to contain metal site vacancies in Chap. 8. The presence of some vacancies
could very well account for the lowering of the overall magnetic moment of the samples. X-ray absorption near-edge structure (XANES) and redox titrations confirmed
that manganese was in fact in the 4+ state in these materials. This work also demonstrated that there is a solid solution line from Li 2 MnO 3 to Li[Ni 1/6 P 1/6 Mn 2/3 ]O 2 .
Along this line, all the transition metal layers have 2/3 manganese occupation allowing for ordering on two
√
3 x
√
3 superlattices. This gave rise to sharp superlattice
peaks in the XRD patterns and the Monte Carlo simulation demonstrated that these
structures do not phase separate into layered–layered composites.
VIII Some layered materials in the Li–Mn–Ni–O system lose a significant amount
of lithium during synthesis. Typically, researchers counter this by adding a small
amount of excess lithium and assume it is lost during synthesis (e.g. Refs. [51, 52]).
However, the phase diagrams here show that most lithium-rich structures in the Li–
Mn–Ni–O system can accommodate more lithium without phase separating. It can
no longer be assumed that all excess lithium is lost during synthesis. Elemental analysis is therefore required to determine the composition of samples after heating. The
contour plots generated in Chap. 7 can also be used to obtain approximate compositions. These contours show that as one adds excess lithium the c axis decreases
rapidly, while the a axis is reduced more gradually. As such, it is vital to use both
parameters in identifying final materials. Recognizing the actual compositions of the
sample also helps explain differences in electrochemistry found in the literature. For
example, Lin et al. [53] made a LiNi 0.5 Mn 0.5 O 2 sample without excess lithium that
showed a lower capacity than materials made by Lu et al. [2]. The samples made
by Lu used LiOH to cause coprecipitation of the mixed transition metal hydroxide
precursor such that a small amount of excess lithium was inevitable. As such, the Lu
sample was just within the layered region while the Lin sample was slightly outside.
In the previous chapter, it was shown that this results in the single-phase material
having a significantly higher capacity, consistent with the published results from Lu
and Lin. A more extreme example of making LiNi 0.5 Mn 0.5 O 2 with excess lithium
was published by Ohzuku et al. [52]. The contour plots were used to demonstrate that
this sample lost little of the excess lithium during synthesis and the electrochemical
performance was therefore attributed to being a lithium-rich layered material.
Two articles show very strange XRD patterns for LiNi 0.5 Mn 0.5 O 2 .
IX The first was published by Jo et al. [54] and showed extreme peak broadening
when regular cooled in oxygen. This was attributed to crystallites on the order of
14 nm. Here, it was shown that samples phase separate into layered-layered composites under such conditions. Therefore, the peak broadening seen by Jo was due to
phase separation and not small crystallites since 14 nm is extremely small for having
been heated for 3 h at 800
◦ C.
X The second strange result was obtained by Hinuma et al. [55] who found
that a single-phase LiNi 0.5 Mn 0.5 O 2 underwent a complex phase separation upon
annealing at 600
◦ C and a return to single-phase upon reheating to 1,000
◦ C. The
new phases appearing upon annealing were not identified in the original paper. The
transformations can now be understood. The layered boundary is high in the Gibbs