Page 163 - Read Online

Liu et al. Energy Mater 2023;3:300011 https://dx.doi.org/10.20517/energymater.2022.68 Page 5 of 10

amount of Ti and Ru was 3.1% and 43.3% (atom weight percentage) [Figure 2G and H]. The specific surface
2
area of the Ru/TiO /CNTs is 968.3 m /g.
2
The XRD pattern of CNTs showed typical broad diffraction peaks of the (002) and (100) planes of carbon,
while TiO /CNTs delivered several diffraction peaks attributed to (101), (200), (211), (204), and (440)
2
planes, which were assigned to rutile TiO (JCPDS (21-1272) [Figure 3A] [29,30] . Notably, no peak for Ru was
2
found in the results of the Ru/TiO /CNTs due to its low content. X-ray photoelectron spectroscopy (XPS) of
2
the Ru/TiO /CNTs confirmed the co-existence of Ru, Ti, O, and C elements [Figure 3B], and the binding
2
3/2
energy of Ti 2p½ and Ru3p overlapped at approximately 465 eV. Four individual component peaks can be
identified as C-C (284.6 eV), C-O (285.3 eV), C=O (287.5eV), and O-C=O (288.5 eV), respectively
[Figure 3C] . O 1s orbital BE spectra of the Ru/TiO /CNTs sample[Figure 3D] identified two peaks at
[25]
2
530.8, and 532.5 eV, indicating the O 1s binding energy of lattice oxygen (O1) and oxygen vacancy (O2),
respectively [14,31] . The peaks at 459.0 eV and 465.08 eV in the fine spectrum of Ti proved the successful
formation of TiO [Figure 3E]. The peaks for Ru 3p1/2 at 464.8 eV were assigned to the photoemission from
2
(Ru ) and the peak at 465.8 eV was attributed to RuO (Ru ) [Figure 3F], which could be due to partial
4+
0
2
oxidation of Ru [32-34] .
Electrochemical performance analysis
The initial discharge-charge curves of cells with the three samples in the potential range of 2.0-4.5V (vs.
Li/Li ) at a current density of 0.1 mA g were measured [Figure 4A]. Cells with Ru/CNTs and Ru/TiO /
-1
+
2
CNTs cathodes showed obviously smaller charge overpotential and larger capacities than those of CNTs,
because of the high catalytic efficiency of Ru catalyst. Meanwhile, the capacity of Ru/TiO /CNTs was almost
2
1000 mAh g higher than that of Ru/CNTs. Both cathodes showed similar initial discharge-charge behavior
-1
at a fixed discharge depth. It is worth noting that the discharge overpotential of the Ru/TiO /CNTs was a bit
2
larger than that without TiO [Figure 4B]. We speculate that the low-conductive TiO showed a weak effect
2
2
on the Li O nucleation during the first cycle. However, the discharge terminal potential recovered in the
2
2
subsequent cycling process. Compared with the cell of Ru/CNTs [Figure 4C], the discharge and charge plots
in fixed capacity kept quite stable [Figure 4D]. A rather clear comparison could be observed from the
evolution process of the discharge terminal potential with a fixed capacity of 500 mAh g . A cycle life of
-1
almost 110 was achieved when using the Ru/TiO /CNTs cathode [Figure 4E]. The prolonged lifespan could
2
be approximately benefited from the thin TiO layer that effectively inhibited carbon corrosion and
2
electrolyte loss.
Discharge and charge mechanism analysis
To further explore how the as-prepared Ru/TiO /CNTs cathode worked on the mechanism of the Li-O
2
2
battery, the morphological evolution of the cathode was characterized by SEM at different cycling states.
The fresh electrode showed a porous structure [Figure 5A]. Snow-like material appeared on the whole
surface of the electrode after the first discharge [Figure 5B], but disappeared after recharge [Figure 5C]. The
reversible formation and disappearance of the discharge product indicated that the Ru/TiO /CNTs cathode
2
could probably catalyze the Li O formation and decomposition. More importantly, unlike traditional
2
2
toroidal-like or film-like discharge products, which either needed large charge overpotential to decompose
or totally passivate the electrode surface, the as-prepared Ru/TiO /CNTs cathode still featured a porous state
2
after discharge and the Li O showed crystalline-like structure with premium size. Thus, the ORR and
2
2
especially OER kinetics based on such structured-Li O should be remarkably boosted and the above-
2
2
mentioned electrochemical results powerfully verified this. The electrode surface became not as sharp as the
fresh one after 30 cycles, but still porous [Figure 5D]. Agglomeration of porous particles appeared on the
cathode when the cycle life reached 60 [Figure 5E], which was probably due to the gradual accumulation of
the side product and the uncompleted decomposition of the discharge product. After another further 40