38
3 Optimization of the Synthesis of Combinatorial Samples
3.2 Thermo-Gravimetric Analyzer (TGA) Results
for Lithium Loss During Synthesis
The bulk sample was characterized using a JD-2000 diffractometer and Rietveld
refinement to determine that it was single-phase Li x Ni 2−x O 2 with x = 0.965.
Figure 3.2 (c) shows the X-ray diffraction (XRD) scan of this sample after being
heated to 900
◦ C in argon in the TGA. The scan contains two phases: Li x Ni 2−x O 2
with x < 0.62, and Li 2 O (JCPDS #77-2144). Lithium oxide is therefore one product
of the thermal decomposition of lithium nickel oxide and balancing the equation with
oxygen yields:
Li y Ni 2−y O 2 → aLi x Ni 2−x O 2 + b[Li 2 O +
1
2
O 2 ]
(3.1)
where a = (2 − y)/(2 − x), b = (y − x)/(2 − x) and y > x. This reaction
was first proposed by Antolini while studying the thermal decomposition of lithium
nickel oxide, Li y Ni 2−y O 2 , with an initial lithium content of y < 0.6 using thermogravimetric analysis at temperatures of 900
◦ C and higher [70].
In order to confirm Eq. 3.1 and to verify that both lithium oxide and oxygen are
lost, the X-ray scattering from certain samples was measured after the TGA runs
used to produce Fig. 3.3. The sample that was heated in air initially had a mass of
11.66 mg and 0.70 mg was lost during the TGA measurements. Equation 3.1 and
the assumption that both O 2 and Li 2 O are lost were used to determine that the final
state corresponded to x = 0.828. XRD with the JD-2000 diffractometer analyzed
with Rietveld refinement yielded x = 0.839. Similarly, the sample heated in oxygen
had a lithium content of x = 0.887 according to XRD in good agreement with 0.870
obtained from Eq. 3.1.
Furthermore, Sata found that lithium oxide then reacts with oxygen according to:
2Li 2 O (s) + O 2(g) → 2Li 2 O 2(g) [31]. Thus, the means for lithium loss during synthesis
of the combinatorial samples may be due to the production of lithium oxide which
is then converted to lithium peroxide vapor. The fact that solid Li 2 O was only seen
in the XRD scans after heating in argon is consistent with the loss of lithium arising
from the reaction of lithium oxide with oxygen identified by Sata [31]. Since lithium
oxide does not decompose or evaporate at these temperatures and the TGA data are
consistent with losing both oxygen and lithium oxide, the decomposition of lithium
nickel oxide followed by the formation of lithium peroxide vapor was the likeliest
source of lithium loss in the bulk samples.
Figure 3.3 shows the mass loss of samples heated in argon, air and oxygen. Clearly,
the reaction for lithium loss was slowed by oxygen. Furthermore, the sample heated
in argon showed continued mass loss during cooling after the hold at 800
◦ C while
the samples heated in oxygen and air showed no such loss and perhaps even a small
mass increase during the initial stages of cooling. This effect suggests that oxygen
re-entered the sample during cooling in order to counter a deficiency which emerged
during heating and only samples in argon continued to lose mass.
3 Optimization of the Synthesis of Combinatorial Samples
3.2 Thermo-Gravimetric Analyzer (TGA) Results
for Lithium Loss During Synthesis
The bulk sample was characterized using a JD-2000 diffractometer and Rietveld
refinement to determine that it was single-phase Li x Ni 2−x O 2 with x = 0.965.
Figure 3.2 (c) shows the X-ray diffraction (XRD) scan of this sample after being
heated to 900
◦ C in argon in the TGA. The scan contains two phases: Li x Ni 2−x O 2
with x < 0.62, and Li 2 O (JCPDS #77-2144). Lithium oxide is therefore one product
of the thermal decomposition of lithium nickel oxide and balancing the equation with
oxygen yields:
Li y Ni 2−y O 2 → aLi x Ni 2−x O 2 + b[Li 2 O +
1
2
O 2 ]
(3.1)
where a = (2 − y)/(2 − x), b = (y − x)/(2 − x) and y > x. This reaction
was first proposed by Antolini while studying the thermal decomposition of lithium
nickel oxide, Li y Ni 2−y O 2 , with an initial lithium content of y < 0.6 using thermogravimetric analysis at temperatures of 900
◦ C and higher [70].
In order to confirm Eq. 3.1 and to verify that both lithium oxide and oxygen are
lost, the X-ray scattering from certain samples was measured after the TGA runs
used to produce Fig. 3.3. The sample that was heated in air initially had a mass of
11.66 mg and 0.70 mg was lost during the TGA measurements. Equation 3.1 and
the assumption that both O 2 and Li 2 O are lost were used to determine that the final
state corresponded to x = 0.828. XRD with the JD-2000 diffractometer analyzed
with Rietveld refinement yielded x = 0.839. Similarly, the sample heated in oxygen
had a lithium content of x = 0.887 according to XRD in good agreement with 0.870
obtained from Eq. 3.1.
Furthermore, Sata found that lithium oxide then reacts with oxygen according to:
2Li 2 O (s) + O 2(g) → 2Li 2 O 2(g) [31]. Thus, the means for lithium loss during synthesis
of the combinatorial samples may be due to the production of lithium oxide which
is then converted to lithium peroxide vapor. The fact that solid Li 2 O was only seen
in the XRD scans after heating in argon is consistent with the loss of lithium arising
from the reaction of lithium oxide with oxygen identified by Sata [31]. Since lithium
oxide does not decompose or evaporate at these temperatures and the TGA data are
consistent with losing both oxygen and lithium oxide, the decomposition of lithium
nickel oxide followed by the formation of lithium peroxide vapor was the likeliest
source of lithium loss in the bulk samples.
Figure 3.3 shows the mass loss of samples heated in argon, air and oxygen. Clearly,
the reaction for lithium loss was slowed by oxygen. Furthermore, the sample heated
in argon showed continued mass loss during cooling after the hold at 800
◦ C while
the samples heated in oxygen and air showed no such loss and perhaps even a small
mass increase during the initial stages of cooling. This effect suggests that oxygen
re-entered the sample during cooling in order to counter a deficiency which emerged
during heating and only samples in argon continued to lose mass.