1.4 The Li–Mn–Ni–O Face of the Pyramid
11
LiNi x Mn 2−x O 4 with 0 ≤ x ≤ 0.5 [6, 34], while samples along line I
are spinel
at lower temperatures [7, 8] . The other lines are layered structures: line II is the
lithium-rich layered line Li[Li (1−2x)/3 Ni x Mn (2−x)/3 ]O 2 ; 0 ≤ x ≤ 0.5 [2], III is
Li[Ni 1−x Mn x ]O 2 ; 0 ≤ x ≤ 0.5 [35], and IV is Li x Ni 2−x O 2 ; 0 ≤ x ≤ 1 [36]. Another composition line in the triangle that has previously been studied in detail is the
Li x Ni 2−x O 2 line from x = 0 to 1 (from the Ni corner to LiNiO 2 ). For x < 0.62, samples are cubic rocksalt where the metal atoms occupy a face-centered cubic lattice
and every metal site is randomly occupied. Though these materials are not interesting
as positive electrodes they warrant study because they often appear as contaminants
in the synthesis of either spinel materials or layered–spinel composites as discussed
below. Above x = 0.62, the structures are hexagonal with the lithium and nickel
preferentially ordering on alternating layers as illustrated in Fig. 1.2a [36]. The transition at x = 0.62 can be identified by fitting the XRD patterns as hexagonal and
plotting the c/a lattice parameter ratio versus x. Extrapolating to the point where the
ratio reaches
√
24, the expected value for a cubic structure [36] gives the position of
the phase transition. This method will be used in Chap. 5.
1.4.1 Li–Mn–Ni–O Spinel Solid Solutions
The spinel solid solution (line I in Fig. 1.8) has been thoroughly studied. The nickelrich endpoint, LiNi 0.5 Mn 1.5 O 4 , is single-phase when prepared in oxygen but shows a
contaminant rocksalt phase when heated in air [34]. More recently, Ma et al. attributed
the contaminant to Ni 6 MnO 8 [37] while Cabana obtained a cubic lattice parameter of
4.15 Å leaving the stoichiometry of the contaminant phase undetermined [38]. This
confusion about the actual composition of this contaminant will be resolved here by
mapping out the entire spinel–rocksalt coexistence region carefully IV.
It is also important to be mindful that LiNi 0.5 Mn 1.5 O 4 spinel can sustain oxygen
vacancies at high temperature and that oxygen returns into the sample during slow
cooling [39, 40]. Compositions of LiNi 0.5 Mn 1.5 O 4−δ with δ = 0.1 were obtained
at temperatures at or above 750
◦ C [41]. However, it is hard to distinguish this
from phase separation with the formation of the rocksalt phase since both occur
simultaneously in air and both result in an increase in the lattice parameter as well as
a mass decrease due to oxygen loss. The observations in Chap. 5 comparing quenched
and slower cooled samples over wide composition ranges helps distinguish these two
forms of oxygen loss V.
1.4.2 Li–Mn–Ni–O Layered Solid Solutions
As previously mentioned, the lithium-rich layered oxide structures have alternating
lithium and transition metal hexagonal layers, with some excess lithium on the TM
layers. Figure 1.9 and Ref. [2] show that when synthesized in air these structures
11
LiNi x Mn 2−x O 4 with 0 ≤ x ≤ 0.5 [6, 34], while samples along line I
are spinel
at lower temperatures [7, 8] . The other lines are layered structures: line II is the
lithium-rich layered line Li[Li (1−2x)/3 Ni x Mn (2−x)/3 ]O 2 ; 0 ≤ x ≤ 0.5 [2], III is
Li[Ni 1−x Mn x ]O 2 ; 0 ≤ x ≤ 0.5 [35], and IV is Li x Ni 2−x O 2 ; 0 ≤ x ≤ 1 [36]. Another composition line in the triangle that has previously been studied in detail is the
Li x Ni 2−x O 2 line from x = 0 to 1 (from the Ni corner to LiNiO 2 ). For x < 0.62, samples are cubic rocksalt where the metal atoms occupy a face-centered cubic lattice
and every metal site is randomly occupied. Though these materials are not interesting
as positive electrodes they warrant study because they often appear as contaminants
in the synthesis of either spinel materials or layered–spinel composites as discussed
below. Above x = 0.62, the structures are hexagonal with the lithium and nickel
preferentially ordering on alternating layers as illustrated in Fig. 1.2a [36]. The transition at x = 0.62 can be identified by fitting the XRD patterns as hexagonal and
plotting the c/a lattice parameter ratio versus x. Extrapolating to the point where the
ratio reaches
√
24, the expected value for a cubic structure [36] gives the position of
the phase transition. This method will be used in Chap. 5.
1.4.1 Li–Mn–Ni–O Spinel Solid Solutions
The spinel solid solution (line I in Fig. 1.8) has been thoroughly studied. The nickelrich endpoint, LiNi 0.5 Mn 1.5 O 4 , is single-phase when prepared in oxygen but shows a
contaminant rocksalt phase when heated in air [34]. More recently, Ma et al. attributed
the contaminant to Ni 6 MnO 8 [37] while Cabana obtained a cubic lattice parameter of
4.15 Å leaving the stoichiometry of the contaminant phase undetermined [38]. This
confusion about the actual composition of this contaminant will be resolved here by
mapping out the entire spinel–rocksalt coexistence region carefully IV.
It is also important to be mindful that LiNi 0.5 Mn 1.5 O 4 spinel can sustain oxygen
vacancies at high temperature and that oxygen returns into the sample during slow
cooling [39, 40]. Compositions of LiNi 0.5 Mn 1.5 O 4−δ with δ = 0.1 were obtained
at temperatures at or above 750
◦ C [41]. However, it is hard to distinguish this
from phase separation with the formation of the rocksalt phase since both occur
simultaneously in air and both result in an increase in the lattice parameter as well as
a mass decrease due to oxygen loss. The observations in Chap. 5 comparing quenched
and slower cooled samples over wide composition ranges helps distinguish these two
forms of oxygen loss V.
1.4.2 Li–Mn–Ni–O Layered Solid Solutions
As previously mentioned, the lithium-rich layered oxide structures have alternating
lithium and transition metal hexagonal layers, with some excess lithium on the TM
layers. Figure 1.9 and Ref. [2] show that when synthesized in air these structures