Size Dependent Fast Li Ion Storage Based on Size Regulated TiO2(B) Nanosheet Electrodes with Vertical, Horizontal and Random Alignment

TiO2(B) has a high theoretical capacity of 335mAh g −1 for Li intercalation and thus has been considered as a candidate for lithium-ion capacitor and Li-ion battery negative electrodes. For high rate lithium storage, i.e. high power density, it is important to shorten the Li diffusion path by using nanostructured TiO2(B). In this work, TiO2(B) nanosheet with different equivalent diameter of 300 nm and 30 nm were prepared. In addition, the orientation of the TiO2(B) nanosheets was manipulated by altering the deposition method and drying process. Smaller size TiO2(B) nanosheets had better Li + intercalation ability compared to larger sized TiO2(B) nanosheets. The effect of alignment of the TiO2(B) nanosheets was evident for small-sized TiO2(B) nanosheets; vertical or random alignment of small-sized TiO2(B) afforded higher capacity compared to horizontally oriented nanosheets. © The Author(s) 2020. Published by ECSJ. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any medium provided the original work is properly cited. [DOI: 10.5796/electrochemistry.20-00055]. Uploading "PDF file created by publishers" to institutional repositories or public websites is not permitted by the copyright license agreement.


Introduction
The fast and safe electrochemical Li storage properties associated with the high intercalation/de-intercalation potential of ³1.5 V vs. Li/Li + makes titanium-based oxides attractive as negative electrode materials for lithium-ion batteries as well as lithium-ion capacitors. 1,2 The suppression of thick SEI (solid electrolyte interface) layer and small volume expansion allows faster charging and discharging than graphite negative electrodes. Of the many polymorphs of titania, TiO 2 (B) has attracted interest owing to the high theoretical Li + storage capacity of 335 mAh g ¹1 , which is comparable to graphite and two times higher than commercially used Li 4 Ti 5 O 12 (theoretical capacity of 175 mAh g ¹1 ). Furthermore, the open-framework and 1D tunnel crystal structure facilitates fast Li intercalation/de-intercalation with small volume change.
Synthesis of high-quality TiO 2 (B) is not simple since TiO 2 (B) is a metastable phase. TiO 2 (B) is typically synthesized by interlayer dehydration condensation via thermal treatment of layered titanic acid (e.g. H 2 Ti 3 O 7 , H 2 Ti 4 O 9 , H 2 Ti 5 O 11 , etc.) or hydrothermal treatment of TiO 2 in alkaline conditions. Thermal treatment of layered titanic acid leads to TiO 2 (B) characterized by whisker-or platelet-like microcrystals with an open channel parallel to the b axis. This characteristic morphology is replicated from the parent layered titanic acid which is prepared by classical high temperature ceramic processing techniques. The b axis of TiO 2 (B) is the direction of the Li + diffusion, 3 thus a long diffusion path coupled with a limited amount of insertion sites translates to slow kinetics. The thickness of TiO 2 (B) along the c axis direction can be controlled by de-laminating and re-stacking of exfoliated TiO 2 nanosheets derived from layered titanic acid as the precursor. 4 However, the Li + diffusion path is still governed by the particle size of the parent ceramic since the length along the b axis is not affected much by the exfoliation process. In addition, TiO 2 (B) electrodes fabricated from layered titanic acid normally show high orientation due to the characteristic platelet-like morphology where the b axis lies horizontally to the current collector, 4 thus the Li + diffusion path is much longer than the actual thickness of the electrode.
Hydrothermal synthesis of nanostructured TiO 2 (B) such as nanowires, nanotubes, and nanoparticles have been reported to show high capacity and improved rate performance compared to micro-sized bulk TiO 2 (B). [5][6][7] The size and shape of such TiO 2 (B) nanomaterials seems to have a strong effect on the lithiation behavior. Nonetheless, it is difficult to control the size of nano-sized TiO 2 (B) through hydrothermal synthesis. In addition, due to the poor electronic conductivity of TiO 2 (B), conducting additives are usually added or coated onto the surface in fairly large amounts for high power studies, which further complicates the fundamental electrochemistry of the material.
In this study, we attempt to elucidate the governing factors of Li + intercalation in TiO 2 (B) using a series of electrodes with wellcharacterized lateral size and orientation without the use of polymeric binders or conductive additives. Since the Li + diffusion in TiO 2 (B) is along the b axis, size-regulated TiO 2 (B) nanosheets with long and short b axis should act as a model material to understand the kinetics of Li + intercalation into TiO 2 (B). The Li + diffusion length along the b axis is controlled by size regulation of TiO 2 (B) nanosheets, which is varied by using TiO 2 nanosheet precursors with different lateral size. Furthermore, the orientation (horizontal, vertical, random alignment) of TiO 2 (B) nanosheets was manipulated by utilizing advanced nanosheet film processing techniques. Based on the controlled morphologies, the kinetics of Li + storage in TiO 2 (B) is discussed.

Synthesis of large and small size TiO 2 nanosheets
Ultrapure water (Milli-Q, >18 M³ cm) was used for all synthesis and characterization. TiO 2 nanosheets were derived from layered K 2 Ti 4 O 9 ·H 2 O (see Supporting Information S1 for details of synthesis). Bi-layered TiO 2 nanosheets were obtained by centrifugal separation. 8 TiO 2 nanosheets with reduced lateral size were obtained by ultrasonicating the TiO 2 nanosheet colloid with a homogenizer (225 W power at 20 kHz, 5 s pulse and 10 s interval of off-power for 30 min).
2.2 Fabrication of TiO 2 (B) nanosheets with different orientation TiO 2 nanosheet electrodes with horizontal, random, and vertical alignment were prepared by different deposition processes and subsequently transformed into horizontal, random and vertically aligned TiO 2 (B) nanosheet electrodes by heat treatment at 350°C under N 2 for 2 hours. The deposition conditions were adjusted so that the amount of the active material per geometric surface was roughly constant at 0.350 mg cm ¹2 .
Horizontally-aligned TiO 2 nanosheet electrodes were fabricated by electrophoretic deposition based on a process similar to a previous study. 4 A pair of Pt plates were placed in the TiO 2 nanosheet colloid (0.5 g L ¹1 ) with a distance of 1 cm. A voltage of 4 V was applied for 30 minutes, and then vacuum-dried. TiO 2 nanosheets deposit on the positive electrode due to the negative charge (negative Zeta potential) of the TiO 2 nanosheets. Verticallyaligned TiO 2 nanosheet electrodes were fabricated by a process developed for fabrication of vertically-aligned graphene oxide electrodes. 9 Following electrophoretic deposition (EPD) of TiO 2 nanosheets, the as-deposited film was freeze dried by transferring the EPD cell quickly to a liquid N 2 bath and then vacuum dried with a freeze dryer (DRW240DA, ADVANTEC Co.). Randomly-oriented electrodes were fabricated by flocculation of TiO 2 nanosheet directly on the current collector. The TiO 2 nanosheet colloid was dropped on a Pt substrate and then 10 mol L ¹1 (M) HCl was dropped to induce flocculation of the nanosheet on the Pt substrate.
Information on methods and apparatus for morphological characterization of the materials are given in Supporting Information S2.

Electrochemical measurements
The electrochemical lithiation/de-lithiation behavior of the TiO 2 (B) nanosheet films were studied with a 3 electrode flat cell (60 mm diameter, 3E-CELL SUS, Eager Corp.) in 1 M LiPF 6 in ethylene carbonate/diethyl carbonate (EC/DEC = 1/1 in volume). Li foil was used as the counter and reference electrodes. Galvanostatic charge/discharge testing was conducted with a voltage window of 1.0-3.0 V vs. Li/Li + at 0.2-10 C (1 C = 335 mA (g-TiO 2 ) ¹1 ) using a 580 Battery Test System (Scribner Associates Inc.). The lateral size of individual bi-layered [Ti 4 O 9 ] 2¹ nanosheets were analyzed from AFM images (Fig. 1) using the equivalent diameter (D e ) as an index. D e was calculated according to the following equation,

Results and Discussion
where A is the area and P is the perimeter of individual nanosheets.
The D e of bi-layered [Ti 4 O 9 ] 2¹ nanosheets without ultranosication had a wide distribution from 100-500 nm with an average D e of 300 nm. The TiO 2 nanosheets obtained with ultrasonication had a Electrochemistry, 88(4), 305-309 (2020) much narrower size distribution of D e = 10-60 nm with an average D e of 30 nm (Fig. S4). Bi-layered [Ti 4 O 9 ] 2¹ nanosheets with different lateral size were converted to TiO 2 (B) nanosheets by thermal treatment. The samples will be denoted as TiO 2 (B)(D e ) nanosheets, where D e represents the equivalent diameter of the nanosheets. Cross-sectional SEM images of horizontally, randomly, and vertically aligned TiO 2 (B)(D e ) nanosheets are shown in Fig. 2. During the drying process, bilayered [Ti 4 O 9 ] 2¹ nanosheets re-stack. In the case of vacuum-dried EPD film, bi-layered [Ti 4 O 9 ] 2¹ nanosheets collapse to form a relatively smooth film with a thickness of ³10 µm. Based on the theoretical density of 3.64 g cm ¹3 for TiO 2 (B) and thickness of individual [Ti 4 O 9 ] 2¹ nanosheets, the film should be composed of 1 © 10 7 TiO 6 octahedra in terms of thickness. The horizontally aligned TiO 2 (B)(D e ) films appear fairly dense in the SEM images, but is actually quite porous; the electrode density estimated from the cross-sectional SEM analysis and mass loading was ³0.35 g cm ¹3 , or 90% porosity based on the theoretical density for TiO 2 (B). By conducting freeze-drying instead of vacuum-drying after EPD, the bi-layered [Ti 4 O 9 ] 2¹ nanosheets align vertically and thus can be converted to vertically-aligned TiO 2 (B)(D e ) without collapse of the macroscopic orientation. The mechanism of vertical alignment of nanosheets via EPD-freeze drying has been discussed earlier. 9 The experimental conditions in this study were controlled so that the mass loading of all electrodes are constant at 0.350 mg cm ¹2 . The thickness of TiO 2 (B)(D e ) nanosheet electrodes with horizontal, random, and vertical alignment are 10, 20, to 100 µm, respectively. Note that as long as the film fabrication process is the same, the difference in the nanosheet size D e of the precursor does not affect the film thickness and tap density. Thus, by decreasing the D e of TiO 2 (B) nanosheets from 300 nm to 30 nm, the b axis, which is the direction of Li + diffusion, should be decreased to 1/10 and the relative number of Li + insertion sites should increase by 10 times.

Electrochemical properties of TiO 2 (B) nanosheets with
different size and orientation Galvanostatic charge/discharge curves and the corresponding dQ/dE plots calculated from the potetioamperograms are shown in Fig. 3. The dQ/dE plots show typical characteristics of TiO 2 (B) with a broad redox peak at E 1/2 ³ 1.55 V vs. Li/Li + . A small shoulder peak is observed at E 1/2 ³ 2.0 V vs. Li/Li + , which may be ascribed to minor anatase TiO 2 impurities. The discharge capacity is plotted as a function of the C rate for TiO 2 (B)(300 nm) and TiO 2 (B)(30 nm) in Fig. 4. For the case of TiO 2 (B)(300 nm), the orientation of TiO 2 (B) does not strongly affect the Li + intercalation kinetics and the amount of Li + intercalated at 0.2 C is 1/5 of the theoretical capacity (Li 0.2 TiO 2 ). The effect of orientation becomes clear by downsizing the lateral dimension of TiO 2 (B). When electrodes with the same orientation but different nanosheet size are compared, the reduction in Li + diffusion path results in a higher capacity at the same C rate. The capacity at 0.2 C for TiO 2 (B)(30 nm) is increased to Li x TiO 2 (x = 0.28-0.42). The lack of full charging of TiO 2 (B)(D e ) indicates that the penetration depth of Li + into the TiO 2 (B) structure is very shallow, even for TiO 2 (B)(30 nm). This result suggests that the poor electronic conductivity does not allow electrons to conduct through the entire film, and only 30 to 40 percent of the film is active for electrochemical lithiation. The lack of full charging may also be due to the presence of anatase TiO 2 as impurity.
Another important trend that can be pointed out for the TiO 2 (B)(30 nm) electrodes is that vertical and random alignment is superior to horizontal alignment. Vertical alignment shows the highest capacity at low C rate (<1 C), but at C rates above 1 C, the randomly aligned electrodes outperforms vertical alignment. Since the Li + diffusion length of these three electrodes should be the same, the results again implies that the electronic conductivity of the electrodes is the dominating factor at high C; 100 µm thick vertical electrode performs well at low C rate but is a poor material at high C rate.
Electrochemistry, 88(4), 305-309 (2020) nanosheets and small-sized TiO 2 (B) nanosheets with an equivalent diameter of 300 nm and 30 nm were prepared by downsizing the dimension of the precursor, bi-layered [Ti 4 O 9 ] 2¹ nanosheets, by ultrasonication of the nanosheet colloid. TiO 2 (B) nanosheet electrodes with three different orientation (horizontal, random, and vertical) was prepared by vacuum-drying or freeze-drying of electrophoretically deposited electrodes or via flocculation. For large-sized TiO 2 (B)(300 nm) nanosheet electrodes, the orientation effect was not clear, suggesting that the kinetics is dominated by the Li + conductivity through the TiO 2 (B) tunnel structure. By down-  . The discharge capacity plotted as a function of C rate for horizontally, randomly, and vertically aligned TiO 2 (B) nanosheets with equivalent diameter of (a) 300 nm and (b) 30 nm.
Electrochemistry, 88(4), 305-309 (2020) sizing TiO 2 (B) to 30 nm, the power capability was improved. A two times increase in rate performance was obtained for horizontallyaligned electrodes by downsizing, while for vertically-and randomly-aligned electrodes, higher capacity was obtained at the same C rate for TiO 2 (B)(30 nm) nanosheets compared with TiO 2 (B)(300 nm) nanosheets. The orientation of small-sized TiO 2 (B) had a large influence on Li + storage kinetics. Vertically-aligned TiO 2 (B) showed much faster Li + intercalation/de-intercalation behavior compared to horizontally aligned TiO 2 (B) with the same mass loading.

Supporting Information
The Supporting Information is available on the website at DOI: https://doi.org/10.5796/electrochemistry.20-00055.