Layered Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 Nanoarrays for Na/Li-Ion Intercalation Systems: Effect of Ion Size

Vertically oriented Na 2 Ti 2 O 4 (OH) 2 (NTO) and K 2 Ti 2 O 4 (OH) 2 (KTO) nanoarrays on Ti foils are synthesized via one-step hydrother- mal method. These titanates-based layered materials with different cations are utilized to assemble binder-free Li/Na- ion batteries (LIBs/NIBs). We experimentally ﬁnd the ion size difference of K, Na and Li in electrode materials or/and electrolyte can dramatically inﬂuence the performances of LIBs/NIBs. NTO with larger volume of unit cell shows higher capacity in both NIBs (213 mA h g − 1 vs. 165 mA h g − 1 ) and LIBs (509 mA h g − 1 vs. 288 mA h g − 1 ). That is, the ion size in host materials can signiﬁcantly inﬂuence the intercalation/extraction of Na and Li ions. On the other hand, Na ions with larger size and atomic mass in the electrolyte show slower chargetransfer(81.5-144.9 μ svs.0.23 μ s)andiondiffusion(7.52 × 10 − 10 cm 2 s − 1 vs.1.33 × 10 − 9 cm 2 s − 1 ,Na 2 Ti 2 O 4 (OH) 2 -based cells) than that of Li ions. This work clearly demonstrates the inﬂuence of ion size, which can clarify and provide fundamental information for fabricating high energy density and long-life batteries. In addition, it demonstrates that both Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 can be used as anode materials with around two ions storage, as for LIBs (4.6 for Na 2 Ti 2 O 4 (OH) 2 and 2.9 for K 2 Ti 2 O 4 (OH) 2 ), and NIBs (1.9 for Na 2 Ti 2 O 4 (OH) 2 and 1.7 for K 2 Ti 2 O 4 (OH) 2 ). Moreover, it is demonstrated that Li/Na ions are stored between the adjacent layered sheets in Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 .

Energy conversion and storage is a key issue in daily life and industry field. The urgent need of renewable and clean energy has attracted great attention in low-cost, safety and rechargeable battery with adequate voltage, capacity and rate capacity. 1 Nowadays, Li ion battery (LIB) has conquered the portable electronic market, offering the largest energy density and output voltage of all rechargeable batteries in use. 2 However, the amount of storage on earth and higher cost to obtain Li has become a critical barrier to popularization and sustainable application in battery energy storage. 1,3 Due to the wide availability and low-cost of sodium, Na ion battery (NIB) is a promising alternative to LIB in the future. But sodium has a lower reducing voltage (-2.71 V vs. SHE, compared to -3.04 V) and the gravimetric capacity is lower (1165 mA h g −1 compared to 3829 mA h g −1 ) than lithium. 4 The atomic mass of Na is more than three times larger and the ionic radius of Na is 0.3 Å larger than Li. Furthermore, graphite cannot be used as an anodic insertion host for Na ions. 5,6 Hence, Na-based anodes are urgently needed for the development of NIBs.
To date, very few viable NIB anode materials have been reported to be viable. 7 Among these materials, titanate based nanomaterials have the relatively low redox potential of Ti 3+/4+ , such as TiO 2 , Na 2 Ti 6 O 13 , Na 2 Ti 3 O 7 and Na 0.66 [Li 0.22 Ti 0.78 ]O 2 . [8][9][10][11][12][13][14][15][16][17] Layered NaTiO 2 electrode was found to have electrochemical activity with ∼0.5 Na transfer in the voltage window 0.6-1.6 V. 18,19 Palacin et al. found two Na ions could be inserted into Na 2 Ti 3 O 7 (0.67 Na per Ti, 200 mA h g −1 ). 13 Na 2 Ti 6 O 13 nanorods are reported to have ultrafast (30 C) rate capability and impressive long cycle life. 10 Recently, ultrathin Na 2 Ti 2 O 4 (OH) 2 nanosheets were synthesized by hydrothermal approach and used as anodes for NIBs/LIBs. 20 Amazingly, this anode shows Na ion storage performance is much better than that in LIBs. The authors attributed this to slower ion transfer in LIBs than in NIBs, which differs from the general knowledge on the size/atomic mass effects of Na and Li. Therefore, it is very important to investigate/clarify how the ion size in electrode materials or/and electrolyte affects the performance of NIBs/LIBs. Na 2 Ti 2 O 4 (OH) 2 (NTO) is a layered sodium titanate with the bodycentered orthorhombic crystal structure, which can be synthesized through one-step hydrothermal method. 21,22 NTO is made up of TiO 6 z E-mail: hong1979@swu.edu.cn octahedral that share edges to form two dimensional sheets (Fig.  S1a). 21,23 The lattice constants of NTO are: a = 19.26 Å, b = 3.78 Å and c = 3.00 Å. 23 The spacing between the adjacent sheets is 9.63 Å, which is larger than that of layered Na 4 Ti 9 O 20 · xH 2 O (8.6 Å) and Li 1.81 H 0.19 Ti 2 O 5 · 2.2H 2 O (8.33 Å). 24,25 And this spacing can be adjusted by changing the cationic types or/and the relative amounts of Na + and OH − . For example, K 2 Ti 2 O 4 (OH) 2 (KTO) can be synthesized with the same crystal structure but with different lattice constants. Thus, NTO and KTO are good candidates for studying the effects of ion size.
Herein, layered NTO and KTO nano-array films directly on Ti foils via one-step hydrothermal method were synthesized. Both materials have the same crystal structure and almost the same specific surface area/pore size distribution, but the lattice constants have some differences due to the difference of Na and K cations in crystal. This provides convenience to study the effects of ion size in both electrode material and electrolyte. This work clearly demonstrates the influence of ion size in NIBs and LIBs, which can provide fundamental information for fabricating high energy density and long-life batteries.

Experimental
Material preparation.-NTO was prepared by basic hydrothermal growth process of Ti foil in NaOH solution. 26,27 Typically, a piece of Ti foil (2.5×3.0 cm 2 ) was ultrasonically cleaned in water, ethanol and 4% HCl for 15 min each, and then placed against the wall of a 50 mL Teflon-lined stainless steel autoclave filled with 30 mL 1 M NaOH. After that, the sealed autoclave was kept inside in an electric oven at 220 • C for 24 h. After the hydrothermal reaction, Ti foil covered with NTO nanoarrays was washed with deionized water. Finally, the obtained NTO films were dried at 60 • C for 30 min. KTO was synthesized as same as above, except for using 1 M KOH solution.
Structural characterization.-Scanning electron microscope (SEM) equipment with energy dispersive spectrometer (EDS) and transmission electron microscope (TEM) experiments were performed on a JSM-6510LV and a JEM-2010 electron microscope, respectively. X-ray diffractometer (XRD) measurements were carried out on a XRD-7000 X-ray diffractometer using Cu K α radiation at λ = 1.54 Å. Nitrogen adsorption/desorption isotherms for surface area and pore analysis were measured at 77 K using a NOVA 1200e instrument. Thermal stability was examined with thermogravimetric analysis (TGA), Q50 in N 2 . To obtain the lattice constants, XRD and selected area electron diffraction (SAED) patterns are used to calculate lattice constants.
Electrochemical measurement.-Ti foils covered NTO on one side were cut into disc with tableting machine. Electrochemical tests were performed with 2035 coin cells with Na or Li metal as counter electrodes. The electrolyte was 1 M NaClO 4 dissolved in a mixture of ethylene carbonate (EC) and ethylmethyl carbonate (EMC) (1:1 by v.), and the separator was a microporous membrane (Celgard 2135). The cells were assembled in an argon-filled glove box. The galvanostatic charge-discharge tests were conducted at a voltage interval of 0.01-2.5 V with a Land System. Cyclic voltammetry (CV) measurements were carried out at a scan rate of 0.05 mV s −1 using a CHI 660D electrochemical workstation. Electrochemical impedance spectroscopy (EIS) measurements were also conducted on CHI 660D in the frequency range of 0.01-10 5 Hz at open circuit voltage. The measurement conditions of KTO were identical to NTO. For LIBs, the electrolyte is 1 M LiPF 6 in a mixture of EC, diethyl carbonate (DEC) and dimethyl carbonate (DMC) (1:1:1 by v).

Results and Discussion
NTO and KTO have a layered structure (Fig. S1a) and belong to an orthorhombic system. 28,29 The body-centered orthorhombic NTO (or KTO) consists of edge-sharing TiO 6 octahedral that form corrugated layers, between which hydrated Na (or K) cations and protons (H + ) are located. 30 The Na and K ions can be easily replaced with other elements by ion exchange reaction. 22 The thermal stability of NTO was firstly characterized by thermogravimetric analysis under a nitrogen atmosphere. At 500 • C, only ∼6% mass loss is recorded due to dehydroxylation (Fig. S1b), which is consistent with similar layered Ni 3 Si 2 O 5 (OH) 4 anode material. 31 This shows the prepared nanoarrays have good thermal stability. Fig. 1 shows SEM images of as-prepared titanate-based films and charge/discharged electrode materials after 500 cycles. Figs. 1a and 1c depict the variation in the surface topography on Ti foils caused by the alkali solution changing. NTO film is made of long and uniform nanowires, while short/thin nanowires and some nanobelts exist in KTO film. Due to the participation of Na and K ions in the formation of titanate-based nanostructures, the morphological differences must result from the induced influence by the size difference between Na and K ions on Na and K ions ability to intercalate. This phenomenon is consistent with the previously reported result. 32 As shown in Fig.  1b, the average diameter and length of NTO nanowires are ∼79 nm and ∼5.6 μm, respectively. The well aligned nanowire array structure can also be observed clearly. After cycling for 500 times, the morphology of NTO nanowire arrays (Figs. 1e and 1g) undergo significant changes. The end of nanowires bunch together and form the microscale texture but the structure of single nanowire is intact. In Figs. 1c and 1d, the average diameter and length of KTO nanowire/nanobelt observed from SEM are ∼75 nm and ∼3.6 μm, respectively. But the nanoarrays structure of KTO is not aligned regularly like NTO. While the morphology of NTO changes dramatically, there is only a visible change in KTO after cycling. As shown in Figs. 1f and 1h, the surface becomes rougher, wires become agglomerated and less distinguishable.
The morphological and structural characters of as-prepared NTO and KTO further characterized using TEM.  (220), respectively. The other peaks marked with square come from the Ti substrate. 33 The lattice constants of NTO calculated from XRD are: a = 1.741 nm and b = 0.377 nm. For KOH-treated sample, the XRD pattern complies with the formula of K 2 Ti 2 O 4 (OH) 2 , of which the phase is identical to Na 2 Ti 2 O 4 (OH) 2 . This result is the same as what C. Kim et al. reported. 33 The observed diffraction peaks of K 2 Ti 2 O 4 (OH) 2 at 2θ = 10.63 • , 24.1 • , 28.4 • and 47.7 • , correspond to the lattice planes (200), (110), (310) and (020), respectively. The lattice constants of KTO calculated from XRD are: a = 1.663 nm and b = 0.381 nm. The difference of lattice constants between NTO and KTO can be attributed to the different radius and bond energy between Na and K ions. This provides the opportunity to investigate the effects of ion size. The asterisk marked peaks in Figs. 3a and 3b resemble a phase with crystal structure, and is quite similar to that of mineral kassite (CaTi 2 O 4 (OH) 2 ), [34][35][36] so it means the K may not be adequate. C. Kim et al. also observed the similar XRD peaks. 33 After the Li or Na insertion, the asterisk peaks disappear, which means the inadequate K is supplied with Li or Na.   After the first full discharge, the (200) and (020) peaks have visible movement. The magnified XRD patterns of (200) and (020) peaks are shown in supplementary Fig. S2. Based on this and TEM measurements, the lattice constants of as-prepared and intercalated NTO/KTO (Table I) can be obtained (see Table I note). As shown in Table I, lattice constant a shows a larger shift than that of lattice constant b, e.g. 0.034 nm vs. 0.001 nm for NTO in LIBs. This can directly and powerfully demonstrate that the Li/Na ions are stored between the adjacent layered sheets in NTO and KTO. Element contents of various electrode materials are also illustrated in Table I, confirming the ion transfer of Li, Na and K during the first full discharge. The volume of NTO unit cell is 0.217 nm 3 , which is 8.5% larger than that of KTO unit cell. Li insertion makes less change on the values of lattice constants a and b than Na insertion. And the lattice constant a is more likely to be affected, indicating the Na/Li ions are mainly stored between the adjacent sheets in layered NTO/KTO.
The N 2 adsorption/desorption isotherms of NTO and KTO are shown in Fig. 3c. The surface area values are 92.8 and 96.5 m 2 g −1 by Brunauer-Emmett-Teller (BET) method, referring to NTO and KTO, respectively. Due to the small difference on surface area, we assume that the surface area cannot significantly affect the device performance. Distinct hysteresis loop can be observed in the range of 0.4−1.0 P/P 0 , which suggests the existence of a mesoporous structure for nanostructured NTO and KTO films. The pore size distribution of the samples calculated from desorption isotherm using Barret−Joyner−Halenda (BJH) method is shown in Fig. 3d. The most probable pore diameter is about 3.80 nm, which is the same as in both NTO and KTO. The samples have a number of mesopores and macropores, facilitating the penetration of electrolyte and the transportation of Li + or Na + ions. Moreover, both one-dimensional nanoarrays are all directly grown on Ti foils, which serve as current collectors. It is known that one-dimensional nanostructure facilitates the fast electron transport and collection. 37 The one-dimensional nanoarrays with high BET surface area and porous structure provide a large interface for the penetration of electrolyte, hence expecting to achieve high specific capacity in both LIBs and NIBs.
After the first full discharge, the electrode materials were also collected and characterized by TEM. Figs. 4a and 4b show the first full-intercalated NTO with Na. The surface of NTO nanowire has some nanoparticles and ∼7.9 nm external layer which is distinct from the inside of nanowire. This should be attributed to the fact that the Na intercalation stimulates the volume expansion and the Na 2 O nanoparticles (SEI film) forms on the external layer. The similar case can be observed on the first full-intercalated KTO with Na (Figs. 4c and 4d). However, in LIBs, it has not been observed. This maybe a result of the smaller size of Li ions, but it also needs further investigation to reveal the inherent mechanism. After Na insertion, the variation value of the lattice distance of (020) plane is 0.151 nm and 0.155 nm for NTO and KTO, respectively. As for Li insertion process, the variation value of the interplanar spacing of (020) is 0.011 nm and 0.031 nm for NTO and KTO, respectively. This clearly and powerfully demonstrates Na with larger size will induce large volume expansion. And NTO with large volume of unit cell is favorable to Li/Na storage and has small volume expansion.
To examine the electrochemical properties, CV was firstly conducted. For Li intercalation system, there are three visible redox peaks of NTO-based LIBs marked in Fig. 5a. But the three redox peaks of KTO-based LIBs are barely visible. During CV measurements, both  cathodic curves show great decay between first and second cycles (e.g. from −0.5 A g −1 to −0.25 A g −1 ), suggesting the formation of the SEI films. On first and second cycles, the redox peaks overlap so they could not be well defined. Some peaks such as peak II or/and peak III only existed during the first cycle, indirectly suggesting them are from the side reaction. NTO-based LIBs show higher intensity of oxidation peak I, indicating more Li ions can be efficiently extracted from NTO. NTO shows more anodic redox peak potential (1.48 V) than that of KTO (1.1 V). The difference is consistent with the fact that the reducing voltage of potassium is lower than that of sodium. The redox reactions of MTO (M = Na or K) in Li intercalation system can be illustrated as, In Fig. 5b, for NIBs, two broader redox peaks can be observed. Among the first three cycles, all the intensity of redox peaks decrease gradually. But the magnitude of this decrease is not large as LIBs and the side reaction is not significant. Moreover, NTO-based NIBs show higher intensity of oxidation peak I and peak II, suggesting higher electrochemical activity and more extraction of Na ions. Compared with LIBs, NIBs deliver a lower redox peak current density, indicating poorer charge transfer and ion diffusion kinetics behavior in NIBs. In NIBs, the redox reaction must happen between Ti 4+ and Ti 3+ . The reaction equation may be written as, M 2 Ti 2 O 4 (OH) 2 + xNa + + xe − = Na x M 2 Ti 2 O 4 (OH) 2 [2] Figs. 5c and 5d show the first three charge/discharge curves of NTO-and KTO-based NIBs/LIBs. Assuming that there is one electron redox reaction between Ti 4+ and Ti 3+ (one Na ion storage), the theoretical capacity of NTO and KTO are 111.8 mA h g −1 and 98.6 mA h g −1 , respectively. In Figs. 5c and 5d, the first discharge capacity is 1379 mA h g −1 (NTO) and 701 mA h g −1 (KTO) for LIBs, and 472 mA h g −1 (NTO) and 503 mA h g −1 (KTO) for NIBs at 10 mA g −1 .
These much larger capacities of the first cycle are mainly attributed to the formation of SEI films. For the reversible discharge curves (the third cycle), the number of Na insertion is 1.9 (NTO) and 1.7 (KTO). And the number of Li insertion is 4.6 (NTO) and 2.9 (KTO). Thus, Li ions can be stored easily in NTO and more Li ions can be inserted into the crystal lattice than Na ions.
The rate performance of NTO-and KTO-based LIBs/NIBs is summarized in Figs. 5e and 5f. The cells were cycled at six current densities of 10, 20, 50, 100, 200 and 500 mA g −1 in a voltage window of 0.01-2.5 V. As expected, in both LIBs and NIBs, the discharge capacity decreases as the current density increase. Upon decreasing the current density from 500 mA g −1 to 100 mA g −1 , in LIBs, 223 mA h g −1 (NTO) and 75 mA h g −1 (KTO) are obtained, which are about 95.1% and 94.8% capacity retentions, respectively. In NIBs, when the current changed from 500 mA g −1 to 100 mA g −1 , 110 mA h g −1 (NTO) and 68 mA h g −1 (KTO) are obtained, which are about 99.3% and 100% capacity retentions, respectively. These results indicate that both LIBs and NIBs have an acceptable rate performance. Comparing NTO with KTO, NTO-based cells possess higher capacities, which is in agreement with the CV curves in Figs. 5a and 5b. At 100 mA g −1 , the ratio of NTO capacity to KTO capacity is 2.98 in LIBs, while the ratio is 1.63 in NIBs. Compared with NIBs, LIBs show superior capacity at all current densities. In addition, the Coulombic efficiencies during the rate performance tests are shown in Fig. S3. LIBs have higher efficiencies, even KTO-based LIBs exhibit more than 100% efficiency. This may be due to more quantity of electricity contributed by some K ions in KTO crystal extraction. The long-term cycling stability is shown in Fig. S4. To quantitatively reveal the stability, the ratio of the 500 th cycle capacity to the 5 th cycle capacity is calculated. For LIBs, 32.48% and 49.28% are obtained for KTO and NTO, respectively. While 38.46% and 22.48% are obtained in NIBs for KTO and NTO, respectively. Moreover, NTO shows around 100% Coulombic efficiency, while KTO shows less than 100% and more than 100% efficiencies in NIBs and LIBs, respectively. More than 100% efficiency achieved by KTO-based LIBs is agreement with the results in rate performance measurements, suggesting this is repeatable and reliable.
Since the results from the CV curves hint at a diffusion limited response, further investigations were carried out using electrochemical impedance spectroscopy (EIS). EIS is a non-destructive technique to study the electrode kinetics of electrode materials. 38 Figs. 6a-6c presents the Nyquist plots of NTO-and KTO-based LIBs/NIBs at open circuit voltage. To distinguish various phenomena taking place in an electrode at different time scales, an equivalent circuit for these cells was used to fit the Nyquist plots, as shown in Fig. 6d. In this circuit, R e represents the equivalent series resistance. R (sf+ct) depicts the SEI film (sf) and charge transfer resistance (ct). R b is the bulk (b) resistance. The constant phase elements CPE (sf+dl) is due to the SEI film and double layer (dl) capacitance, whereas CPE b refers to bulk capacitance. In addition, W s and C int represent the Warburg impedance and intercalation capacitance. Detailed discussion of this equivalent circuit can be found in the references. [38][39][40][41] The extracted impedance parameters are listed in Table II. Compared with LIBs, NIBs exhibit larger R e , R (sf+ct) and W s , indicating a larger resistance of electrolyte in NIBs, a slower charge transfer in NIBs, and a slower ion diffusion in NIBs. These results demonstrate, in electrolyte, the Li/Na ions can significantly influence the electrode processes due to the radius and atomic mass differences. For fresh cells, the values of R b in both cells deliver similar results, demonstrating the NTO-and KTOcovered Ti foils possess almost the same conductivity. But KTO-based cells exhibit higher resistance of R (sf+ct) , demonstrating the lower  electrochemical activity and kinetical processes. This is in agreement with the results from CV tests in Figs. 5a and 5b. Compared with asprepared cells, the cycled cells show R e and R (sf+ct) increase (except KTO-based LIBs), but W s decreases (except NTO-based NIBs). The lower W s should be attributed to the fact that there is no good contact between electrode materials and electrolyte. The increased R e and R (sf+ct) are as a result that electrode materials have partially fallen off from Ti foil, as shown in Fig. S5. The decreased values of CPE b -P after cycling test provide additional evidences, e.g. 0.9 vs. 0.48 for as-prepared and cycled NTO NIBs, respectively. The case CPE b -P = 1 describes an ideal capacitor while the case CPE b -P = 0 describes a pure resistor. The reason for the decreased R (sf+ct) of KTO-based LIBs can be attributed to the significant size difference between Li and K ions. The inserted Li ions will replace the K ions due to ion exchange and leave a large space for further Li ions insertion. This will decrease the steric hindrance and lead to the decrease of charge transfer resistance. Due to the similar size of Na vs. Li and K vs. Na, this case does not happen.
The rate of charge-transfer reaction can be calculated from the maximum peak frequency from Bode plots (Fig. S6). A faster charge transfer must happen due to a higher peak frequency. For fresh cells, NIBs show slower charge transfer rate in the range of 81.5-144.9 μs, while LIBs have 0.23 μs reaction rate. This clearly demonstrates the insertion/extraction of Na ions is not kinetics favorable in comparison with Li ions. After cycling, the charge transfer rate of NTObased cells shows more significant decrease than that of KTO-based cells.
To further reveal the effects of ion size, the relationship between Z' and ω −0.5 at low frequencies is plotted in Fig. 7. The Warburg coefficient σ ω can be obtained from the slope of these plots. 42 The diffusion coefficient (D) can be obtained from Equation 3, D = 0.5(RT /AF 2 σ ω C) 2 * n −0.5 [3] where R is the gas constant (8.314 J mol −1 K −1 ), T is the temperature (298.5 K), A is the area of the electrode surface, F is the Faraday's constant (96,500 C mol −1 ) and C is the molar concentration of Li/Na ions. For LIBs, the D is 1.33 × 10 −9 cm 2 s −1 of as-prepared NTO cell, 1.32 × 10 −10 cm 2 s −1 of cycled NTO cell, 7.52 × 10 −10 cm 2 s −1 of as-prepared KTO cell, and 1.34 × 10 −10 cm 2 s −1 of as-prepared KTO cell. For NIBs, the D is 7.33 × 10 −10 cm 2 s −1 of as-prepared NTO cell, 1.27 × 10 −9 cm 2 s −1 of cycled NTO cell, 1.36 × 10 −10 cm 2 s −1 of as-prepared KTO cell, and 1.80 × 10 −10 cm 2 s −1 of as-prepared KTO cell. NTO-based cells show larger D than that of KTO-based cells, directly demonstrating the large unit cell of NTO is beneficial to the Li/Na ions diffusion. Comparing LIBs to NIBs, LIBs cells have larger D than that of NIBs cells, indicating the smaller and low-weight Li ions has superior diffusion ability. As expected, the as-prepared LIBs have larger D than cycled. However, the cycled NIBs show larger D than as-prepared. This might be due to the relative larger Na ions induced more volumetric expansion and opened diffusion pathway.

Conclusions
In summary, we synthesized layered Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 nanoarray films directly on Ti foil substrates via onestep hydrothermal method. Li/Na ions are stored between the adjacent layered sheets in Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 , which is demonstrated by XRD characterizations. Based on these two layered materials, the effects of ion size were investigated in LIBs and NIBs cells. Na 2 Ti 2 O 4 (OH) 2 with larger volume of unit cell shows 1.76 times and 1.29 times higher capacity in both LIBs and NIBs than that of K 2 Ti 2 O 4 (OH) 2 based cells, respectively. That is, the ion size in host materials can significantly influence the intercalation/extraction process of Na and Li ions. On the other hand, Li ions with smaller size and atomic mass in the electrolyte show 1.81 times and 5.53 times faster ion diffusion than that of Na ions in LIBs and NIBs, respectively. The charge transfer of LIBs is 0.23 μs, which is significant faster than that of NIBs (81.5-144.9 μs). In addition, both Na 2 Ti 2 O 4 (OH) 2 and K 2 Ti 2 O 4 (OH) 2 can be used as anode materials with around two ions storage for LIBs (4.6 for Na 2 Ti 2 O 4 (OH) 2 and 2.9 for K 2 Ti 2 O 4 (OH) 2 ) and NIBs (1.9 for Na 2 Ti 2 O 4 (OH) 2 and 1.7 for K 2 Ti 2 O 4 (OH) 2 ), but the long-term stability should be further improved. This work has clarified how the performance of NIBs/LIBs is affected by the ion size in electrode materials or/and electrolyte.