Effect of interlayer spacing on sodium ion insertion in nanostructured titanium hydrogeno phosphates/carbon nanotube composites

Abstract

In sodium ion batteries, the ease of insertion and extraction of sodium ions in the electrode materials is one of the key parameters for the overall performance. In this article, the electrochemical sodium ion insertion in layered titanium hydrogeno phosphates (TiP) has been studied. In this material, the interlayer spacing and the particle morphology can be controlled by the choice of synthesis methods. Both nanostructured TiP and the coarse grained bulk counterpart were synthesized and properties were compared. While the specific capacity of nanostructured TiP materials was found to be not sensitive to the interlayer spacing, the specific capacity of coarse grained bulk TiP materials was significantly increased as the interlayer spacing was increased with the intercalation of water molecules in the layered host structure. These results indicate that interlayer spacing may not be the primary factor for Na-ion diffusion in nanostructured materials, where many interstitials are available for Na-ion diffusion. It is shown that nanostructured TiP materials can deliver excellent rate capability, and long term cycle stability with stable reversible capacity without the need of interlayer spacing expansion. The electrochemical properties of nanostructured materials were further enhanced when prepared as composites with carbon nanotubes that enhance the overall conductivity of the electrode materials.

---

Introduction

Lithium (Li)-ion battery (LIB) technologies have matured, having developed desirable features such as high energy density, long cycle life, and lightweight for applications ranging from portable electronics to electric vehicles.^1–3 The scarcity of lithium resources, however, is driving attention to developing alternative energy storage systems.^4,5 Sodium (Na)-ion batteries (NIBs) are a potential replacement for LIBs largely due to the low cost and natural abundance of Na precursors.^6–8 However, not all of the electrode materials that show Li storage can be directly used for Na-ion storage due to the steric limitation caused by the bulkier Na⁺ (1.02 Å) compared to Li⁺ (0.76 Å).⁹ Noticeably, recent studies demonstrated reversible Na-ion insertion into various materials with high capacity including V₂O₅, and MoS₂.^10,11 In most cases, however, electrode materials suffer from unsatisfactory cycle stability compared to Li/Li⁺. Therefore, realization of novel electrode materials that have excellent reversible Na-ion insertion properties is very important to advance NIB technologies.

Titanium hydrogeno phosphates, of a formula Ti(HPO₄)₂·xH₂O (TiP–xH₂O), are built on a two-dimensional structure consisting of PO₃(OH) tetrahedra and TiO₆ octahedra.^12,13 They have been proposed as possible Li- and Na-ion conductors through both cation exchange reactions and direct cation insertion into the structure.¹³ The interlayer spacing of TiP–xH₂O varies with the amounts of water in the structure.^14,15 Recently, several groups reported the beneficial roles of water intercalation in layered structured electrode materials. The common conclusions from the earlier work is that water intercalation not only expands the interlayer spacing but also screens the strong interaction between elements in host layers and the charge carriers especially those with larger size or higher charge such as Na⁺ and Mg²⁺.^10,16,17 Therefore, water intercalation could improve the kinetics of charge mobility in the electrode as well as charge storage capacity and long term stability.^10,16,17 In this regard, the amenable interlayer properties with water intercalation provide sound rationale to explore applicability of TiP–xH₂O as electrode materials for Na-ion storage. Their Na-ion conduction and storage behaviors, however, have not been demonstrated. To the best of our knowledge, this is the first study on Na-ion storage in materials based on transition metal hydrogeno phosphates.

In this work, we investigated the reversible insertion of Na ions in Ti(HPO₄)₂·xH₂O (x = 0, 1), denoted as TiP–xH₂O. The interlayer spacing of TiP–xH₂O was tuned with the amounts of water intercalation and the degree of crystallinity of samples was controlled by the synthesis routes. The expanded interlayer spacing with water intercalation incorporation enables better Na-ion diffusion in the bulk hydrated TiP–H₂O materials. However, this beneficial effect of larger interlayer spacing was less pronounced when the size of TiP–xH₂O particles was reduced into nanoscale. Both hydrated and dehydrated nanostructured TiP–xH₂O displayed high specific capacity regardless the interlayer spacing. These results imply that ion diffusion in such nanostructured materials is likely independent of interlayer spacing varied with water intercalation. As literatures have shown that CNTs can be used to fabricate composites with improved mechanical properties,^18,19 we chose CNTs as a stable and conductive substrate for the composite electrode. The electrochemical properties of nanostructured TiP–xH₂O materials were further enhanced when prepared as composites with CNTs, exhibiting excellent rate capability, and long cycle stability.

Experimental

Synthesis of Ti(HPO₄)₂·xH₂O and Ti(HPO₄)₂·xH₂O/CNT composites

Bulk Ti(HPO₄)₂·xH₂O (denoted as bulk-TiP–H₂O) was synthesized via hydrothermal processes. Ti metal powder (75 mg) was dissolved into 5 mL of phosphoric acid (85%). Then, the mixture solution underwent a hydrothermal reaction at 160 °C for 12 hours. The product was naturally cooled down to room temperature and then collected by vacuum filtration. Nanostructured Ti(HPO₄)₂·xH₂O (denoted as nano-TiP–xH₂O) was synthesized via a solvothermal reaction. Initially, 20 μL of titanium(IV) butoxide (97%) was added to 5 mL of ethanol and the mixture was stirred for 1 hour at room temperature. Then, 0.2 mL of phosphoric acid (85%) was added drop wise and the reaction mixture was transferred to a Teflon-line autoclave vessel with a capacity of 10 mL. The reaction mixture underwent a solvothermal reaction at 120 °C for 20 hours in a vertical furnace. The product was naturally cooled down to room temperature and then collected by vacuum filtration. For nano-TiP–xH₂O and carbon nanotube composite, 5 mL of CNT–ethanol dispersion (0.1 mg mL⁻¹) was used instead of pure ethanol in solvothermal synthesis. The dehydrated phase for Ti(HPO₄)₂ (bulk-TiP and nano-TiP) were obtained by heating the samples at 280 °C in nitrogen atmosphere for 3 hours.

Characterization

X-ray diffraction (XRD) patterns were obtained using a Panalytical X'pert PRO MRD HR X-ray diffraction system with Cu-Kα radiation (λ = 1.5405 Å). Fourier transformed infrared (FTIR) transmittance spectra were recorded using a Thermo Scientific Nicolet 6700 FTIR spectrometer. The thermal analysis of samples was done by using a thermogravimetry (TA Instruments TGA Q5000) system in air atmosphere in the temperature range from 38 °C to 800 °C with a heating rate of 10 °C min⁻¹. Scanning electron microscopy (SEM) was performed using a FEI XL30 SEM-FEG. Transmission electron microscopy (TEM) was carried out by FEI Tecnai G² Twin. The specific surface area of bulk-TiP–xH₂O and nano-TiP–xH₂O was determined by nitrogen adsorption at 77 K.

Electrochemical measurements

The electrochemical properties of TiP–xH₂O and TiP–xH₂O–CNT materials were evaluated using CR 2032-type coin cells. A small thin square shaped sodium metal sheet with a thickness of about 1 mm was used as a counter electrode. The working electrodes were fabricated by following. TiP–xH₂O (or TiP–xH₂O–CNT) was mixed with polyvinylidene difluoride (PVDF) and Super P carbon with a mass ratio of 70 : 20 : 10 followed by casting the mixture paste on aluminum foil. The areal density of electrode was controlled to be between 2.7–3.5 mg cm⁻². The cell assembly was done in a pure argon filled-glove box. 1 M NaClO₄ in ethylene carbonate (EC) and propylene carbonate (PC) (1/1 = v/v) was used as an electrolyte. A glass microfiber filter (grade GF/F; Whatman, U.S.) was used as a separator. Galvanostatic charge–discharge measurements were performed at room temperature using MacPile Battery Analyzer. Cells were tested at various current rates (0.1, 0.2, 0.5, 1.0, 2.0, 5.0C) in the voltage range of 1.5–2.8 V. Cyclic voltammetry (CV) at 0.1 mV s⁻¹ and electrochemical impedance spectroscopy (EIS) measurements were conducted after 50 cycles of charge–discharge in a frequency range of 100 kHz to 0.01 Hz with an AC voltage amplitude of 10 mV (Bio-logic SP300 instrument). To perform ex situ XRD, Swagelok-type cells were used. Swagelok-type cells in charged or discharged states were disassembled in the glove box. Electrodes were rinsed with PC to remove the residual electrolyte salts.

Results and discussion

The powder X-ray diffraction patterns of as-prepared bulk and nanostructured TiP–xH₂O samples are shown in Fig. 1A. The observed patterns matched that of titanium hydrogeno phosphates in the literature.²⁰ The peak at 11.6° was assigned as (002) reflection of bulk-TiP–H₂O. This (002) peak shifted to 12.5° after heating bulk-TiP–H₂O at 280 °C to obtain dehydrated bulk-TiP. The corresponding interlayer distance decreased from 7.7 Å to 6.9 Å. This decrease is due to the loss of the water molecules during heating. Peaks with reduced intensities were observed in nanostructured TiP–xH₂O samples, suggesting the small grain size and the low degree of crystallinity of these samples. The calculated interlayer distances were slightly larger than those of bulk samples. This slightly larger interlayer spacing of 7.9 Å, compared with that of bulk sample (7.7 Å), is likely due to the randomly stacked layers in nanostructured TiP–xH₂O. With this random stacking of layers, the interlayer spacing of nano-TiP–xH₂O could only slightly decrease as intercalation water molecules are removed. The peaks were broader in nano-TiP–H₂O–CNT and nano-TiP–CNT (Fig. S1A†).

Fig. 1 XRD patterns of (A) bulk-TiP–H₂O, bulk-TiP, nano-TiP–H₂O, and nano-TiP, (B) FT-IR spectra of nano-TiP–H₂O–CNT, and nano-TiP–CNT, and (C) TGA curves of bulk-TiP–H₂O, nano-TiP–H₂O, and nano-TiP–H₂O–CNT.

FT-IR spectra also confirmed the structure according the interlayer spacing (Fig. 1B and S1†). The hydrated samples, bulk-TiP–H₂O, nano-TiP–H₂O, and nano-TiP–H₂O–CNT shows characteristic bands of water intercalation. The narrow bands at 3556 cm⁻¹ and 3480 cm⁻¹ are assigned to vibrational mode of water molecule.²¹ The narrow peak at 1613 cm⁻¹ corresponds to the bending mode of water molecule.^21,22 A broad band around 3100 cm⁻¹ was observed in dehydrated samples, bulk-TiP, nano-TiP, and nano-TiP–CNT, instead of the characteristic bands of water intercalation. This band corresponds to the P–OH stretching mode of the hydrogen bond.²³

Fig. 1C shows TGA curves of bulk-TiP–H₂O, nano-TiP–H₂O, and nano-TiP–H₂O–CNT. The weight loss around 280 °C is due to the dehydration process with the water intercalation removal. The weight loss above 400 °C corresponds to the dehydroxylation, accompanied with conversion into the pyrophosphate, TiP₂O₇. The total weight loss of ∼15% indicated the removal of two moles of water per mole of TiP–xH₂O. The additional weight loss of ∼5% after 600 °C observed in nano-TiP–H₂O–CNT suggested that the TiP–CNT composite samples contain ∼5 wt% of CNTs. Notably, both dehydration and dehydroxylation processes occurred at a lower temperature in nanostructured samples. For example, the dehydration started at ∼200 °C and ∼100 °C for bulk-TiP–H₂O and nano-TiP–H₂O, respectively. This result indicate that nanostructured TiP–xH₂O is more defective and more disordered. This is consistent with our previous observation, where the weight loss attributed to the conversion reaction started earlier in amorphous Ni(OH)₂ than crystalline Ni(OH)₂.²⁴

The morphology of bulk and nanostructured TiP–xH₂O samples were characterized using SEM and TEM (Fig. 2). As shown in Fig. 2A, bulk-TiP–H₂O samples had large plates with the size around 5 μm obtained after hydrothermal synthesis. The morphology of nanostructured nano-TiP–H₂O prepared using the ethanol mediated solvothermal synthesis was found to significantly differ from that of bulk counterparts. The nano-TiP–H₂O particles consist of flakes with the size of 200–300 nm as shown in Fig. 2C. This small flake-like morphology enabled the feasibility to form uniform composites with CNTs. These TiP–xH₂O nanoflakes appear to grow along the CNTs as observed in both SEM and TEM images (Fig. 2E and G). The dehydrated TiP–xH₂O samples maintained the same morphology of their hydrated precursors (Fig. 2B, D, F and H).

Fig. 2 SEM images of (A) bulk-TiP–H₂O, (B) bulk-TiP, (C) nano-TiP–H₂O, (D) nano-TiP, (E) nano-TiP–H₂O–CNT, and (F) nano-TiP–CNT, and TEM images of (G) nano-TiP–H₂O–CNT, and (H) nano-TiP–CNT.

The electrochemical properties of different TiP–xH₂O materials were evaluated using coin-type cells and they showed significant differences in the properties. Fig. 3 shows the comparisons of charge–discharge profiles of TiP–xH₂O materials in Na-ion half cells at 0.1C, respectively. The plateau extends over 1.8–2.4 V indicated Na ions insertion operating the Ti³⁺/Ti⁴⁺ redox pair for both bulk and nanostructured TiP–xH₂O materials. As shown in Fig. S2A,† a pair of broad redox peaks was observed in TiP–xH₂O materials, corresponding to the Ti³⁺/Ti⁴⁺ redox pair. It is consistent with the phase transition observed in charge–discharge profiles. For bulk TiP materials, bulk-TiP–H₂O shows the smaller potential gap between anodic and cathodic peaks, indicating that the better electrochemical reversibility of bulk-TiP–H₂O compared to bulk-TiP materials. This result support the idea that the larger interlayer spacing by water intercalation can improve Na-ion mobility in the electrode materials. Na-ion storage capacity according to the interlayer spacing, however, largely depends on the microstructure of TiP–xH₂O materials. In the case of bulk TiP–xH₂O materials, the hydrated phase, bulk-TiP–H₂O exhibited higher specific capacity of 56 mA h g⁻¹ than the dehydrated phase, bulk-TiP (31 mA h g⁻¹). The active material density is slightly higher in dehydrated TiP than hydrated phase, TiP–H₂O due to the absence of water intercalation in the crystal structure.²⁵ Thus, considering this active material density difference, the specific capacity increase by ∼55% with the water intercalation incorporation is significant. This capacity increase can be explained by the expanded interlayer spacing by the water intercalation. The larger interlayer spacing can lead to the faster Na-ion diffusion and better active material utilization during the sodium insertion.^26,27

Fig. 3 Comparisons of galvanostatic charge–discharge curves of bulk-TiP–H₂O, bulk-TiP, nano-TiP–H₂O, and nano-TiP.

Interestingly, nanostructured TiP–xH₂O materials showed considerably different Na-ion storage characteristics from those of bulk counterpart. Both hydrated phase, nano-TiP–H₂O and dehydrated phase, nano-TiP displayed higher but similar specific capacity regardless the presence of water intercalation. The specific capacity values of nano-TiP–H₂O and nano-TiP were 66 mA h g⁻¹ and 71 mA h g⁻¹, respectively. These results implied that interlayer spacing expansion is less effective for ion diffusion in nanostructured materials. This result is surprising but can be explained based on the fact that a lot of interstitials in nanostructured materials are available for Na ions to diffuse through and thus can participate in redox process primarily in both hydrated and dehydrated samples. Thus, interlayer spacing has negligible impacts on the overall performance. However, nanostructured TiP material showed better rate capability compared to bulk-TiP–xH₂O materials with interlayer expansion (Fig. S2B and S3†). In case of bulk TiP–xH₂O materials, it lost almost all capacity at 0.5C, indicating limited ion diffusion. The particle size reduction to nanoscale greatly shortens Na-ion diffusion distances, substantially enhancing the rate-dependent capacity retention.^28,29 It is possible to expect nanostructured materials have higher surface area that can participate in surface redox reactions. To confirm this possibility, BET surface area of bulk-TiP–H₂O and nano-TiP–H₂O was measure and N₂ adsorption–desorption isotherms of these samples were shown in Fig. S4.† However, the measured BET surface area of 338 cm³ g⁻¹ for bulk-TiP–H₂O was more than 2 times higher than 132 cm³ g⁻¹ for nano-TiP–H₂O. The higher surface area of bulk-TiP–H₂O can be due to the well-developed two dimensional plate-like morphology as seen in the SEM image (Fig. 2A). Despite the higher surface area, bulk-TiP–H₂O materials showed lower capacity than nano-TiP–H₂O materials. These results, thus, support the major charge storage mechanism in layered TiP·xH₂O materials is Na-ion diffusion and intercalation into the host lattice, with minor contributions from surface redox reactions.

Since nano-TiP–xH₂O materials showed higher Na-ion storage capacity, this nano-TiP–xH₂O material and CNTs composite was prepared for the improved conductivity, and their electrochemical properties were further studied. CV curves of nano-TiP–H₂O–CNT and nano-TiP–CNT composites are displayed in Fig. 4A. The pairs of current peak positioned at 2.0 and 2.3 V for nano-TiP–H₂O–CNT and 1.9 and 2.3 V for nano-TiP–CNT correspond to the redox pair of Ti³⁺/Ti⁴⁺. Fig. 4B and S3† display the representative charge–discharge profiles of nano-TiP–H₂O–CNT and nano-TiP–CNT cells at various current rates of 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0C. Both nano-TiP–H₂O–CNT and nano-TiP–CNT composites exhibited excellent rate capability at various current rates (Fig. 4C and S3†). A large capacity drop after the first discharge was observed in all cells and the possible reasons for such a behavior will be discussed later in detail. The high capacity values of 92 mA h g⁻¹ and 84 mA h g⁻¹ were obtained from nano-TiP–H₂O–CNT and nano-TiP–CNT, respectively, in the second discharge cycle at 0.1C. These capacity values indicated about 0.88 and 0.75 Na-ion insertion per unit formula based on the calculation using the equation n = (3.6MC)/F, where n is the number of moles of Na-ion inserted (mol), F is Faraday constant (C mol⁻¹), C is the specific capacity (mA h g⁻¹), and M is the molecular weight of TiP–xH₂O (g mol⁻¹).³⁰ Note that all the capacity values for nano-TiP–xH₂O–CNT materials are calculated based on the total mass including both nano-TiP–xH₂O and CNT. The specific capacity values of nano-TiP–H₂O–CNT and nano-TiP–CNT composites were 57 mA h g⁻¹ and 56 mA h g⁻¹, respectively, at a high current rate of 5.0C. This excellent rate performance can be attributed to the synergistic contributions from shortened Na-ion diffusion length due to the nanostructure and improved electrical conductivity provided by CNTs.

Fig. 4 Comparisons of (A) CV curves of nano-TiP–H₂O–CNT and nano-TiP–CNT at 0.1 mV s⁻¹, (B) charge–discharge curves of nano-TiP–H₂O, nano-TiP, nano-TiP–H₂O–CNT, and nano-TiP–CNT, (C) specific capacity of nano-TiP–H₂O–CNT and nano-TiP–CNT at various current rates of 0.1, 0.2, 0.5, 1.0, 2.0, and 5.0C, (D) long cycle stability test results of nano-TiP–H₂O, nano-TiP, nano-TiP–H₂O–CNT, and nano-TiP–CNT at 2.0C.

The long term cycle stability of nano-TiP–H₂O, nano-TiP, nano-TiP–H₂O–CNT, and nano-TiP–CNT was evaluated at 2.0C. As discussed earlier, the water intercalation does not significantly affect the specific capacity of these nanostructured samples. In the case of nano-TiP–xH₂O cells without CNT, the capacity faded rapidly in 200 cycles (Fig. 4D and S5†). On the other hand, nano-TiP–xH₂O and CNT composite cells showed outstanding cycle stability. The charge storage capacity became stabilized after a capacity drop in the first discharge, exhibiting a stable reversible capacity of approximately 50 mA h g⁻¹. This reversible capacity was almost retained over 1000 cycles (Fig. 4D and S5†). The observed excellent long cycle stability is thought to be due to CNT networks in the sample. The CNT network can provide mechanical stability to maintain the integrity of the electrodes against the repeated volume expansion and shrinkage during charge and discharge. Our solvothermal synthesis condition produced significantly reduced size of TiP–xH₂O nanoflakes, which enables direct deposition on CNTs as well as more efficient active materials utilization during charge and discharge cycles.

The conductivity of different TiP–xH₂O Na-ion half cells were measured by performing electrochemical impedance spectroscopy (EIS). Fig. 5 shows EIS results performed on stabilized cells after 20 cycles of charge and discharge. The Nyquist plot can be interpreted based on the equivalent circuit (inset of Fig. 5). R₁ corresponds to the combination of ionic resistance of electrolyte, intrinsic resistance of the substrate, and the contact resistance at the interface between active material and the substrate.³¹ R₁ appeared as the intercept at the real part in the high frequency region. The resistor, R₂ paralleled with the constant phase elements, Q₂, represents the charge transfer resistance corresponding to the diameter of semicircle.³² The ion diffusion in the host structure is described with the Warburg element (Z_w).³³ As presented, the simulated data (dotted line) from the equivalent circuit fit well the impedance data (symbol) for all TiP–xH₂O samples. The resistance values obtained from fitting and Na-ion diffusion coefficient (D_Na⁺) estimated using Fig. S6† are summarized in Table 1. As is shown, resistance originated from the intrinsic properties (R₁) such as electrolyte and substrate is similar for all the electrodes. However, the charge transfer resistances (R₂) of nanostructured materials was almost half those of bulk crystalline materials. This significantly reduced resistance is likely due to the shortened electron transport distance in nanostructured materials. More importantly, Na-ion storage properties of bulk-TiP–xH₂O that depend on the interlayer spacing was also confirmed by EIS results. The hydrated bulk-TiP–H₂O exhibited much smaller charge transfer resistance (1212 Ω) than dehydrated bulk-TiP (1622 Ω). In addition, the D_Na⁺ calculated for bulk-TiP–H₂O (7.5 × 10⁻¹² cm² s⁻¹) is about 100 times higher than that of bulk-TiP (8.7 × 10⁻¹⁴ cm² s⁻¹). These results imply that Na-ion diffusion kinetics in such hydrated form can be enhanced due to the expanded interlayer spacing with the water intercalation. Meanwhile, the water intercalation can screen the interaction between elements in host layer and Na ions.¹⁷ However, the calculated D_Na⁺ values for nano-TiP and nano-TiP–H₂O materials are not affected by the water intercalation, supporting the idea that nano-TiP–xH₂O could utilize interstitials for Na-ion storage rather than rely on insertion facilitated by the enlarged interlayer spacing. Moreover, nano-TiP–xH₂O–CNT composites showed further decreased charge transfer resistance. This further reduced resistance is attributed to the fact that CNT network provides conducting pathways for electrons and thus enhances the charge transfer processes.

Fig. 5 Nyquist plots and equivalent circuit of bulk-TiP, bulk-TiP–H₂O, nano-TiP, nano-TiP–H₂O, nano-TiP–CNT, and nano-TiP–H₂O–CNT.

Table 1 Electrode resistance and Na-ion diffusion coefficient (D_Na⁺) values for bulk-TiP, bulk-TiP–H₂O, nano-TiP, nano-TiP–H₂O, nano-TiP–CNT, and nano-TiP–H₂O–CNT obtained EIS measurements

	Bulk		Nanostructured
	TiP	TiP–H2O	TiP	TiP–H2O	TiP–CNT	TiP–H2O–CNT
R1 (Ω)	26.11	11.58	12.1	26.9	39.08	39.12
R2 (Ω)	1622	1212	660.3	658	522.4	474.6
DNa^+ (cm^2 s^−1)	8.7 × 10^−14	7.5 × 10^−12	1.0 × 10^−12	1.4 × 10^−12	1.6 × 10^−11	3.7 × 10^−11

---

To understand the origin of the large capacity drop after the first discharge cycle observed in all Na-ion half cells, ex situ XRD was performed on the electrode materials and resulting patterns are shown in Fig. 6. For this measurement, bulk sample, bulk-TiP–H₂O was selected to clearly see the peak position change. As is shown, (002) obviously shifted from 11.6° to 10.6° after the cell was fully discharged, indicating the intercalation of Na-ion into TiP host. In the fully charged state, however, XRD patterns showed two peaks at 11.6° and 10.6°, respectively. This result suggested that the inserted Na ions are not completely extracted from TiP host when charging. Thus, active site is only partially available for the next discharge and charge cycles, resulting in large decrease in capacity. Importantly, however, the recovered (002) peak at 11.6° after the first cycle indicated that the water intercalation remains between layers, indicating the layered structure with the water intercalation can be maintained in the organic solvent-based electrolyte.

Fig. 6 Ex situ XRD performed on as-prepared, discharged, and charged bulk-TiP–H₂O.

Conclusions

In conclusion, as electrode materials, Ti(HPO₄)₂·xH₂O (TiP–xH₂O) with different microstructure can be synthesized through both hydrothermal and solvothermal reactions. The specific capacity of bulk-TiP–xH₂O materials significantly increases as the interlayer separation is expanded by the water intercalation. This increase in capacity with the water intercalation becomes less significant, as the particle size is reduced to nanometer scale. These results implied that interlayer space is not the primary route for Na-ion diffusion in nanostructured materials, where many interstitials are available for Na-ion diffusion. The electrochemical characteristics of nano-TiP–xH₂O can be further improved using carbon nanotubes as dual functional additives. The strong network formed by CNTs ensures the highly reversible sodium storage capacity and electrical conductivity provided by CNTs enhances the rate capability of nano-TiP–xH₂O materials. This work demonstrates that factors determining electrode microstructures such as size reduction and interlayer spacing can lead to unique electrochemical properties when combined and balanced, and thus provides insights into nanostructured materials for future energy storage for Na-ion and other future batteries.

Acknowledgements

This work is supported by a research grant from Army Research Office (ARO) under contract W911NF-04-D-0001. However, any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the ARO. Thermogravimetric analysis was done with the help of Cameron Bloomquist of University of North Caroline at Chapel Hill. The authors also acknowledge the support from Duke SMIF (Shared Materials Instrumentation Facilities).

References

1. J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
2. C. Liu, F. Li, L. P. Ma and H. M. Cheng, Adv. Mater., 2010, 22, E28–E62.
3. C. Chen, Y. Wen, X. Hu, X. Ji, M. Yan, L. Mai, P. Hu, B. Shan and Y. Huang, Nat. Commun., 2015, 6, 6929.
4. J. Wang, C. Luo, T. Gao, A. Langrock, A. C. Mignerey and C. Wang, Small, 2015, 11, 473–481.
5. X. Xiong, W. Luo, X. Hu, C. Chen, L. Qie, D. Hou and Y. Huang, Sci. Rep., 2015, 5, 9254.
6. T. S. Sahu and S. Mitra, Sci. Rep., 2015, 5, 12571.
7. H.-G. Wang, S. Yuan, D.-L. Ma, X.-B. Zhang and J.-M. Yan, Energy Environ. Sci., 2015, 8, 1660–1681.
8. M. D. Slater, D. Kim, E. Lee and C. S. Johnson, Adv. Funct. Mater., 2013, 23, 947–958.
9. H. Xiong, M. D. Slater, M. Balasubramanian, C. S. Johnson and T. Rajh, J. Phys. Chem. Lett., 2011, 2, 2560–2565.
10. Q. Wei, J. Liu, W. Feng, J. Sheng, X. Tian, L. He, Q. An and L. Mai, J. Mater. Chem. A, 2015, 3, 8070–8075.
11. C. Zhu, X. Mu, P. A. van Aken, Y. Yu and J. Maier, Angew. Chem., Int. Ed. Engl., 2014, 53, 2152–2156.
12. S. Bruque, M. A. G. Aranda, E. R. Losilla, P. Olivera-Pastor and P. Maireles-Torres, Inorg. Chem., 1995, 34, 893–899.
13. V. Pralong, V. Caignaert and B. Raveau, J. Mater. Chem., 2011, 21, 12188.
14. P. Olivia, J. Leonardi and J. F. Laurent, J. Power Sources, 1982, 8, 229–255.
15. R. Llavona, M. SuBrez, J. R. Garza and J. Rodriguez, Inorg. Chem., 1989, 28, 2863–2868.
16. K. W. Nam, S. Kim, S. Lee, M. Salama, I. Shterenberg, Y. Gofer, J. S. Kim, E. Yang, C. S. Park, J. S. Kim, S. S. Lee, W. S. Chang, S. G. Doo, Y. N. Jo, Y. Jung, D. Aurbach and J. W. Choi, Nano Lett., 2015, 15, 4071–4079.
17. K. W. Nam, S. Kim, E. Yang, Y. Jung, E. Levi, D. Aurbach and J. W. Choi, Chem. Mater., 2015, 27, 3721–3725.
18. Z. An, L. He, M. Toda, G. Yamamoto, T. Hashida and T. Ono, J. Nanotechnol., 2014, 26, 195601.
19. P. Zhou, X. Yang, L. He, Z. Hao, W. Luo, B. Xiong, X. Xu, C. Niu, M. Yan and L. Mai, Appl. Phys. Lett., 2015, 106, 111908.
20. M. S. Kishore, V. Pralong, V. Caignaert, U. V. Varadaraju and B. Raveau, J. Power Sources, 2007, 169, 355–360.
21. S. Shakshooki, A. Rhuma and F. Siala, Int. J. Biosci., Biochem. Bioinf., 2014, 326, 137–141.
22. A. Espina, E. Jaimez, S. A. Khainakov, C. Trobajo, J. Gracia and J. Rodriguez, J. Inclusion Phenom. Mol. Recognit. Chem., 1998, 30, 333–352.
23. M. V. Maslova and L. G. Gerasimova, Russ. J. Appl. Chem., 2011, 84, 1–8.
24. G. Lee, C. V. Varanasi and J. Liu, Nanoscale, 2015, 7, 3181–3188.
25. P. P. Prosini, M. Lisi, S. Scaccia, M. Carewska, F. Cardellini and M. Pasquali, J. Electrochem. Soc., 2002, 149, A297.
26. X. Zhou, L. J. Wan and Y. G. Guo, Nanoscale, 2012, 4, 5868–5871.
27. G. Du, Z. Guo, S. Wang, R. Zeng, Z. Chen and H. Liu, Chem. Commun., 2010, 46, 1106–1108.
28. K. Chang and W. Chen, Chem. Commun., 2011, 47, 4252–4254.
29. S. K. Das, R. Mallavajula, N. Jayaprakash and L. A. Archer, J. Mater. Chem., 2012, 22, 12988.
30. X. Tian, X. Xu, L. He, Q. Wei, M. Yan, L. Xu, Y. Zhao, C. Yang and L. Mai, J. Power Sources, 2014, 255, 235–241.
31. M.-S. Wu, Y.-A. Huang and C.-H. Yang, J. Electrochem. Soc., 2008, 155, A798.
32. G. Lee, Y. Cheng, C. V. Varanasi and J. Liu, J. Phys. Chem. C, 2014, 118, 2281–2286.
33. Y. Zhu, Y. Xu, Y. Liu, C. Luo and C. Wang, Nanoscale, 2013, 5, 780–787.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra08242d

---