Cycle Stability of Dual-Phase Lithium Titanate (LTO)/TiO2 Nanowires as Lithium Battery Anode

This work studied cycle stability of dual-phase Lithium Titanate (LTO)/TiO2 nanowires as a lithium battery anode. Dual-phase LTO/TiO2 nanowires were successfully synthesized by hydrothermal method at various times lithiation of 10, 24, and 48 h at 80 °C. SEM images show that the morphology of dual-phase LTO/TiO2 is nanowires with a size around 100-200 nm in diameter. The XRD analysis result indicates nanowires main components are anatase (TiO2) and spinel Li4Ti5O12. The first discharge specific capacity of LTO/TiO2-10, LTO/TiO2-24, and LTO/TiO2-48 was 181.68, 175.29, and 154.30 mAh/g, respectively. After the rate capacity testing, the LTO/TiO2-10, LTO/TiO2-24, and LTO/TiO2-48 have maintained 161.25, 165.25, and 152.53 mAh/g separately. The retentions for each sample were 86.71, 92.86, and 89.79 %. Based on the results of electrochemical performance, increased LTO content helped increase samples cycle stability. However, the prolonged lithiation time also produced impurities, which reduced the cycle stability.


INTRODUCTION
Nowadays, many electric and electronic devices have been developed, such as mobile phones, computer ignition, and lighting systems. With the growing power and durability of electric vehicles and portable electronic equipment, the production of high energy density batteries and reliable cycling efficiency is urgent [1] [2]. Optimization of the batteries is one of the most significant methods. Among all practical batteries (lead-acid, nickelcadmium, nickel-metal hydride, et al.), lithium ion batteries (LIBs) have effectively dominated the battery industry based on lithium intercalation chemistry and have essentially revolutionized our daily existence. Almost all of our mobile devices are powered by lithium-ion batteries, and people want larger equipment, especially cars, to function better [3] [4].
However, when it comes to the car scale, the traditional lithium-ion batteries' ability cannot fully match the requirements. Currently, depending on the types of lithium-ion batteries, there are one or more of the following problems: small specific capacity, low safety, low cycle life, high price, and low working temperature. Many improvement schemes are applied to the batteries, including modifying the electrode, improving electrolyte, developing the packaging technology, and so on [5] - [8].
In the process of exploring more suitable materials for lithium-ion batteries anodes, Lithium titanate (Li 4 Ti 5 O 12 / LTO) has attracted tremendous attention due to its specific properties [9] [10]. Compared with the commonly used graphite materials, it performs an operation voltage at around 1.55 V vs. Li+/Li, much higher than graphite (0.1 V vs. Li+/Li) [11]. It enables the Li 4 Ti 5 O 12 to avoid the self-discharge when working as an anode and thus enhances its safety and stability [12]. By comparing with other titaniumbased anodes like TiO 2 , the Li 4 Ti 5 O 12 shows higher cycle stability due to its "zero-strain" property, which means there is almost no volume change during the insertion and extraction of lithium-ions of the crystal [13] [14]. On the other hand, as a precursor to producing LTO, theoretically TiO 2 possesses a specific capacity of 335 mAh/g, comparable to that of graphite [15]- [17]. Noerochim et al. demonstrates that dual-phase LTO/TiO 2 has a high potential as anode material for high-rate application of lithium-ion battery [18].
Liao et al. synthesize dual phase LTO/TiO 2 nanowire arrays as integrated anodes for high-rate lithium-ion batteries. The facile hydrothermal and ion exchange route developed to synthesize dual phase LTO/TiO 2 nanowire arrays. The resulting LTO/TiO 2 has a ratio of about 1:2. The introduction of TiO 2 to Li 4 Ti 5 O 12 increases the specific capacity and creates a dual-phase nanostructure with a high grain boundary density that facilitates electrons and Li-ions transport. The resulting dual-phase nanowire electrode has good rate capability compared to pure Lithium titanate and TiO 2 . The results obtained also explain that the dual-phase LTO/TiO 2 produced has good cycle stability.
In this work, the LTO-TiO 2 dual-phase nanowires synthesized using TiO 2 nanowires as precursors through a hydrothermal process at various times lithiation of 10, 24 and 48 h. SEM and XRD tests were carried out to characterize samples. The electrochemical tests were done through cycling and rate test to study the cycle stability.

Methods 2.2.1. Hydrothermal synthesis LTO-TiO 2 nanowires
TiO 2 nanowires synthesized by a hydrothermal method. In this approach, 2 g titanium (IV) butoxide (97 %, Sigma) dissolved in 12 g ethanol. Sodium hydroxide (1.2 g) and 0.5 g glucose (≥ 99.5 %, Sigma) dissolved in 15 g water. These two solutions then mixed to form a white suspension. After stirring and ultrasonication, the suspension transferred to autoclave and heated to 260 ºC for 24 h. Afterwards, 2.5 ml hydrochloric acid (35 %, Sigma) and 200 ml water added to the suspension. The suspension stood 24 h and washed with water and ethanol. After drying the product, the powders collected and calcined at 600 ºC with Ar protection for one h. The product then noted as TiO 2 nanowires.
To synthesize LTO/TiO 2 nanowires, the previous synthesized TiO 2 nanowires used as precursors. In this process, 0.05 g TiO 2 nanowires dispersed in water with the aid of ultrasonication.

LTO-TiO 2 Nanowires Characterization
LTO-TiO 2 nanowires that have produced are characterized using different instrument techniques. The surface morphology of nanoparticles studied using scanning electron microscopy (SEM, FEI Nova NanoSEM 450). The nanoparticles' crystal structure studied using X-ray diffraction (XRD, PANalytical Xpert Multipurpose X-ray Diffraction System) techniques. The range of 2θ set from 10 to 100, and the material for the target was cupper.

Electrochemical performance study
Li/Li 2 O electrodes chosen as the counter electrode for the tested electrodes. 25µm thick glass fibres used as the separators. 1M lithium hexafluorophosphate in ethylene carbonate and ethyl methyl carbonate (5/5 by volume) selected as the electrolyte. All coin cells assembled in an Ar filled dry glove box.
The electrical properties analyzed by Netware CT-4008-5V10mA-164 batteries testing system, and the charge-discharge cycling performed between 1-3 V (vs Li/Li+) at room temperature using different C rates (1 C=175 mA/g).

SEM and XRD Characterization
The morphology of the nanowires examined with SEM. As shown in Figure 1 (a), after 10 h lithiation process, there were little small particles grown on the wires' surface. The mean diameter of nanowire samples determined from SEM imaging was around 100 nm. With the increasing of lithiation time, the amount of the small particles grew. After a 24 h reaction, the nanowires' structure can still be recognized (Figure 1 b). After 48 h reaction, most wires fully covered by the small lithiated particles (Figure 1 c). This process caused most of the separated nanowires to be connected and increased the nanowires diameter to around 200 nm. This phenomenon changed the morphology of the nanowires to more nanosheets-like or sponge- like structure [19]. XRD used to identify the crystallographic structure. As shown in Figure 2, after 10 h reaction, there a small number of Li 4 Ti 5 O 12 produced. However, a large number of TiO 2 phase remained. All major Bragg peaks well matched with those of anatase TiO 2 (ICDD no. 00-004-0477) [20] and cubic spinel Li 4 Ti 5 O 12 (ICDD no. 00-049-0207) [21] respectively. As the reaction time extends to 24 h, more TiO 2 phase consumed, and more LTO produced. The main components for the nanowires were still anatase TiO 2 and spinel Li 4 Ti 5 O 12 ( Figure  2). After 48 h lithiation process, most TiO 2 reacted to form LTO. Because excessive LiOH was used as a lithium source during the hydrothermal reaction, the prolonged reaction enabled more lithium ions to be involved in the process [22].
Consequently, Li 2 TiO 3 (ICSD no. 01-075-1602) [23] with a higher Li to Ti ratio than Li 4 Ti 5 O 12 was formed ( Figure 2). The observed major peaks at 18.4437° and 43.2608° representatives for spinel Li 4 Ti 5 O 12 , while the peak at 43.6510° contributed by (4 0 0) face of monoclinic Li 2 TiO 3 . Compared with 10 and 24 h samples, the major peak intensity for anatase TiO 2 , which located at around 25.2872°, severely weakened, indicating the low content of TiO 2 in the nanowires. Also, the excessive lithium ions can be inserted into the lattice gaps in both Li 2 TiO 3 and Li 4 Ti 5 O 12 to form LTO with different Li to Ti ratio. As a result, some miscellaneous peaks were observed in the XRD pattern ( Figure 2. Unmatched peaks belongs to LTO with different phases). These XRD results confirm the SEM results where with increasing lithiation time, nanosheets-like or sponge-like structures are formed. The structure formed indicates the Li 4 Ti 5 O 12 produced during the process.

Electrochemical Characterization
Half-cell batteries tests carried out with metal lithium as both the counter and the reference electrode to study the electrochemical performance of all three kinds of dual-phase nanowires. Fig. 3 has shown the charge-discharge curves of the first three cycles of all three kinds of samples. During the charge and discharge process, the C rate was 0.5 C (1 C=175 mA/g), so the current in this process a constant value for all three samples. All samples except LTO/TiO 2 -48 display two distinguishable gradient changes during their charge and discharge processes (Figure 3 a, b). The turning points at around 2.0 V and 1.78 V referred to the insertion and extraction of lithium ions in the TiO 2 crystal. The turning points, which around 1.6 V and 1.5 V, on the other hand, were contributed by Li 4 Ti 5 O 12 . In the batteries tests, capacity calculated by multiplication of time, current and voltage, since the current a fixed value, the area under the chargedischarge curve had represented the electrode's capacity. Naturally, a higher and longer plateau produces more capacity than a low and short plane [24].
In the case of LTO/TiO 2 -48, there no TiO 2 charge and discharge planes observed in its curve due to low TiO 2 content (Figure 3 c). Li 4 Ti 5 O 12 provided almost all capacity. As for the LTO/TiO 2 -10, whose TiO 2 content was the highest, the curve shifted most (Figure 3 a). Both the charge and discharge planes of TiO 2 shortened while no significant variation happened to the Li 4 Ti 5 O 12 plane. The reason lies in the difference in electrochemical stability between these two contents. The Li 4 Ti 5 O 12 is known as zero strain materials in lithium-ion batteries, which means there is almost no volume change during the insertion and extraction of lithium ions [25]. Its crystal structure can be preserved to the greatest extent in the cyclic test. On the other hand, compared with Li 4 Ti 5 O 12 , TiO 2 can store more lithium ions per formula unit. This property makes it inevitable that the crystal volume will change Figure 7. Rate performance at various current rate from 0.2 C to 20 C and then back to 0.5 C and 0.2 C of LTO/TiO 2--10, LTO/TiO 2--24, and LTO/TiO 2--48 significantly in cycling. As a result, the crystal structure is more likely to be damaged, and the capacity it provided is also reduced at a higher rate [26]. Figure 4 show the voltage-specific capacity curves at 0.1 C. As the cycle number increased, each sample had a decrease in both charge and discharged specific capacity. Considering the LTO/ TiO 2 -10 and the LTO/TiO 2 -24 (Figure 4 a, b), in the first 50th cycle, the TiO 2 charge and discharge planes have shrunk to about half of their initial length. The LTO/TiO 2 -48 had almost no TiO 2 content, so the TiO 2 charge plane was a slope and cannot be distinguished even at the first cycle (Fig.  4 c) discharge plane of TiO 2 in the curve of LTO/ TiO 2 -48. In Li 4 Ti 5 O 12 , the lengths of charge and discharge plateaus also reduced in all three samples after the first cycles. Because not all lithium ions inserted in the lattice could be extracted during the discharge process, the length of the plane decreased with cycling [27].
Although the capacity of TiO 2 decayed relatively quicker compared with Li 4 Ti 5 O 12 , it still provided the valuable capacity to the electrode. As shown in Fig. 5, the first discharge specific capacity of LTO/ TiO 2 -10 was 181.68 mAh/g, while the LTO/TiO 2 -24 had 175.29 mAh/g as initial. On the other hand, the LTO/TiO 2 -48, which short in TiO 2 content, have only 154.30 mAh/g for its first cycle. After 100 cycle testing, the capacity for LTO/TiO 2 -10, LTO/TiO 2 -24, and LTO/TiO 2 -48 was 135.16 mAh/ g, 142.65 mAh/g, and 121.83 mAh/g. The retentions of the first cycle were 74.7 %, 81.87 %, and 79.32 %, respectively.
The rated capacity is also an essential consideration in lithium-ion batteries. To examine the effect of the different composite electrodes on rate capability, these three Ti-based nanowires studied by charging/discharging at different current rates which were 0.2 C, 0.5 C, 1 C, 2 C, 5 C, 10 C, 20 C, 0.5 C, 0.2 C (1 C=175 mA/g). Figure 6 shows LTO/TiO 2 -10 has the highest initial capacity, which is 184.75 mAh/g. The LTO/TiO 2 -24 slightly lower than the LTO/TiO 2 -10, and it was 177.35 mAh/g. The LTO/TiO 2 -48 have the lowest initial capacity. However, when it at high current rates like 10 C or 20 C, the specific capacity of LTO/TiO 2 -10 and LTO/TiO 2 -24 seriously hindered, because when it at the high current rate, the spinel TiO 2 stepped to an unstable electrochemical situation and cannot be fully involved in the electrode reaction.
Furthermore, its crystal structure was more likely to be damaged. Because the Li 4  These impurities are less electrochemically stable than the pure spinel Li 4 Ti 5 O 12 . As a result, the rate capacity performance of LTO/TiO 2 -48 was worse than it should be.

CONCLUSION
The LTO/TiO 2 dual-phase nanowires synthesized by hydrothermal reaction. LTO/TiO 2 nanowires had different morphologies due to various lithiation time. The nanowires initial thickness was around 100 nm, and the 48 h lithiated sample had a thickness of around 200 nm, and the particles almost entirely covered the wires. The xrd results showed that the longer the lithiation time, the LTO levels increased. The first discharge specific capacity of LTO/TiO 2 -10, LTO/TiO 2 -24, and LTO/TiO 2 -48 was 181.68, 175.29 and 154.30 mAh/g, respectively. After the rate capacity testing, the LTO/TiO 2 -10, LTO/TiO 2 -24, and LTO/TiO 2 -48 have maintained 161. 25