A Facile Pre-Lithiated Strategy towards High-Performance Li2Se-LiTiO2 Composite Cathode for Li-Se Batteries.

Conventional lithium-ion batteries with a limited energy density are unable to assume the responsibility of energy-structure innovation. Lithium-selenium (Li-Se) batteries are considered to be the next generation energy storage devices since Se cathodes have high volumetric energy density. However, the shuttle effect and volume expansion of Se cathodes severely restrict the commercialization of Li-Se batteries. Herein, a facile solid-phase synthesis method is successfully developed to fabricate novel pre-lithiated Li2Se-LiTiO2 composite cathode materials. Impressively, the rationally designed Li2Se-LiTiO2 composites demonstrate significantly enhanced electrochemical performance. On the one hand, the overpotential of Li2Se-LiTiO2 cathode extremely decreases from 2.93 V to 2.15 V. On the other hand, the specific discharge capacity of Li2Se-LiTiO2 cathode is two times higher than that of Li2Se. Such enhancement is mainly accounted to the emergence of oxygen vacancies during the conversion of Ti4+ into Ti3+, as well as the strong chemisorption of LiTiO2 particles for polyselenides. This facile pre-lithiated strategy underscores the potential importance of embedding Li into Se for boosting electrochemical performance of Se cathode, which is highly expected for high-performance Li-Se batteries to cover a wide range of practical applications.

1. Introduction

In recent years, the conventional lithium-ion batteries cannot meet the current development demand due to the limited energy density [1,2,3,4,5,6,7,8,9]. In this respect, sulfur cathode offers a high theoretical specific capacity of 1675 mA h g−1 when paired with lithium metal anode [10,11,12]. Ironically, lithium-sulfur batteries are mainly limited by non-conductive feature of sulfur cathode [13,14,15]. Selenium (Se), as an agnate element of sulfur, has similar charge-discharge reaction and volumetric capacity to sulfur, and the electronic conductivity of Se (1 × 10−3 S m−1) is much higher than sulfur (5 × 10−28 S m−1) [16,17]. Hence, Li-Se batteries are considered as one of the new generations of promising electrochemical energy storage devices [18,19,20,21]. However, the development of Li-Se batteries still faces many problems, such as notorious shuttle effect and volume expansion in Li+ insert/extract processes, which cause the loss of active material and low Coulombic efficiency. Meanwhile, the intermediate products (polyselenides) dissolve in ether-based electrolyte and further migrate to the anode side, which will not only passive the surface of lithium metal anode, but also severely reduce the cyclic performance [22,23,24,25,26,27].

Up to now, many strategies are adopted to circumvent the abovementioned concerns to facilitate the electrochemical performance of Li-Se batteries. The introduction of TiO2 as secondary phase for Li-S/Se batteries is demonstrated to be an effective strategy on account of its strong chemisorption of lithium polysulfides/lithium polyselenides [14,28,29,30,31]. However, the synthesis of TiO2/Se composite is cumbersome, and the volume expansion is still unresolvable. It is worth mentioning that the pre-lithiation of Se is a valid strategy to mitigate the volume expansion [32,33,34]. Unfortunately, the synthesis of fully lithiated Li2Se often involves complex processes and expensive chemical reagents, which remains huge challenges in great urgency [33,34,35]. Therefore, the key to break this predicament is the concise and efficient synthesis of fully pre-lithiated Se-based composites.

In this work, as illustrated in Figure 1, we propose a novel Li2Se-LiTiO2 composite cathode, which is obtained via a two-step solid-phase method. During the synthesis process, LiH synchronously reacts with Se and TiO2 to in-situ form Li2Se-LiTiO2 composite and by-product H2 gas, which needs no subsequent separation. As the fully pre-lithiated cathode, Li2Se-LiTiO2 composite has many unique features. On the one hand, Li2Se could easily ameliorate the volume expansion, thereby leading to the superior structure stability and enhanced cycling stability. On the other hand, the introduced LiTiO2 not only could accelerate the reaction kinetics and electrochemical activities of Li2Se, reducing the reaction overpotential, but also offer the abundant chemical adsorption sites for tapping polyselenides, suppressing the notorious shuttle effect. Therefore, this rationally designed Li2Se-LiTiO2 composite is a promising Se-based cathode for high-performance Li-Se batteries.

Figure 1

Schematic illustration of the preparation process of Li2Se-LiTiO2 composites.

2. Experimental Section

2.1. Preparation of Li2Se-LiTiO2 Composites

Selenium powder (Se, Aladdin Holdings Group Co. Ltd., Shanghai, China, purity 99.99%), anatase phase titanium dioxide (TiO2, Aladdin Holdings Group Co. Ltd., Shanghai, China, purity 99.8%) and lithium hydride (LiH, Aladdin Holdings Group Co. Ltd., Shanghai, China, purity 97%) were used without further purification. Li2Se-LiTiO2 samples were synthesized via two-step solid-phase reaction. In detail, Se, TiO2 and LiH powders were transferred into stainless-steel milling jars with a molar ratio of 4:1:9. The milling jars were milled on a planetary ball mill (QM-1SP2, Nanjing university instrument factory, Nanjing, China) at 500 rpm for 20 h. After that, the precursor was heated in a custom tube reactor at 500 °C for 3 h under vacuum to obtain Li2Se-LiTiO2 composites.

2.2. Materials Characterization

X-ray diffraction (XRD) patterns were carried out by an X-ray powder diffractometer (X’Pert Pro, Cu Kα radiation, Rigaku Corporation, Tokyo, Japan). All the samples were sealed with Kapton film to avoid air exposure. The morphology and microstructure were observed by scanning electron microscopy (SEM, Nova Nano 450, FEI, Hillsboro, OR, USA) with an energy dispersive spectroscopy (EDS, Oxford X-Max 80, Oxford Instruments, Oxford, UK) detector and transmission electron microscopy (TEM, Tecnai G2 F30, FEI, Hillsboro, OR, USA). X-ray photoelectron spectroscopy (XPS, ESCALAB 250X, Themo, Shanghai, China) measurements were performed on an Axis Ultra DLD system (Kratos) with a monochromatic Al Kα (1486.6 eV) X-ray source. Ultraviolet-visible (UV-vis) adsorption spectra were recorded on SHIMADSU UV-2550 spectrophotometer (SHIMADSU, Chengdu, China) to assess the polyselenide adsorption ability.

2.3. Electrochemical Measurements

Electrochemical properties of Li2Se-LiTiO2 electrodes were conducted on CR2025-type coin cells by using lithium metal as anode. The uniform slurry was composed of 60 wt.% Li2Se-LiTiO2 composite as active material, 30 wt.% conductive carbon (Super P, SP) and 10 wt.% ethyl cellulose, which were mixed by toluene as dispersant. The electrolyte was 1.0 mol L−1 lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) in a co-solvent of 1,3-dioxolane (DOL)/1,2-dimethoxyethane (DME) (1:1, v/v) with 2 wt.% lithium nitrate (LiNO3). The electrolyte dosage in each cell was 20 μL mg−1 of electrolyte-to-active material ratio. All the cells were assembled in an Ar-filled glove box (H2O < 0.1 ppm, O2 < 0.1 ppm). The galvanostatic charge-discharge tests were performed on a battery testing system (Shenzhen Neware Technology Co. Ltd., Shenzhen, China). The cells were firstly charged to 3.8 V at 50 mA g−1, and subsequently the voltage window was adjusted to 1.7–2.6 V. For cyclic voltammogram (CV) analysis, the first forward scan started from open circuit voltage to 3.8 V, and a backward scan ended to 1.7 V. The successive scans were in the voltage range of 1.7–2.6 V at a scan rate of 0.1 mV s−1 on CHI650B electrochemical workstation (Chenhua, Shanghai, China). Electrochemical impedance spectra (EIS) were recorded in the frequency range from 0.1 Hz to 1.0 MHz on CHI650B electrochemical workstation.

3. Results and Discussion

In order to reveal the reaction mechanism of Li2Se-LiTiO2 composite by reacting LiH with Se and TiO2, Figure 2a,b illustrate the time-pressure curves for the ball-milling and heating processes, respectively. During the initial ball-milling process, the pressure increases dramatically to 3.39 bar within 1 h, suggesting the reaction generates a large amount of gas. Subsequently, the pressure slowly raises to 3.63 bar from 1 h to 10 h. After 10 h, the gas is no longer produced, indicating the ball-milling reaction is completed. Meanwhile, during the heating process, a conspicuous variation in pressure can be detected, in which the pressure is rapidly increased in the temperature ranging from 250 °C to 400 °C. This result implies that LiH reacts with Se and TiO2 in this temperature range. In order to prove the phase transformation after ball-milling and heating processes, the corresponding XRD patterns of Li2Se-LiTiO2, Li2Se and LiTiO2 are depicted in Figure 2c. The diffraction peaks of Li2Se-LiTiO2 composite shows a superposition of Li2Se (PDF#23-0072) and LiTiO2 (PDF#16-0223). It is worth mentioning that the relative intensity of the diffraction peaks of LiTiO2 in Li2Se-LiTiO2 composite becomes stronger with increasing proportion of LiTiO2 (Figure S1). According to our previous work [34], LiH could react with Se to generate Li2Se (2LiH + Se = Li2Se + H2 ↑). Analogously, LiTiO2 could be obtained by reacting LiH with TiO2 (2LiH + 2TiO2 = 2LiTiO2 + H2 ↑), which is verified by XRD results in Figure 2c and Figure S1. The distinct characteristic peaks of LiTiO2 match well with the cubic phase (Fm3m space group) with lattice parameters of a = b = c = 4.140 Å, which Li and Ti ions are octahedrally coordinated by O [36,37,38]. The comparison of our work with other recent studies on Li-Se batteries is shown in Table S1. On the on hand, the introduction of TiO2 alone could improve the cycling performance, but could not relieve the volume expansion. On the other hand, the traditional pre-lithiation process is excessively complex, and a large number of toxic solvents are used in the preparation process, which seriously pollutes the environment. The synthetic route we adopted is simple, green and environmentally friendly, which has great application prospects.

Figure 2

(a) Time-pressure curve during ball milling process. The insert table is the pressure and the number of moles in the gas at certain sampling points. (b) Time-temperature and time-pressure curves during heating process. The insert table is the pressure and the number of moles in the gas at certain sampling points. (c) XRD patterns of Li2Se-LiTiO2, Li2Se and LiTiO2.

The microstructure, morphology and elemental distribution of Li2Se-LiTiO2 are further investigated by SEM, TEM and EDS. Compared to pristine Li2Se (Figure S2a), Li2Se-LiTiO2 composite has more regular morphology with small particle size ranging from 60 to 100 nm (Figure 3a,b), which will be favorable to significantly enhance the electrochemical reactivity and structural stability [34]. Additionally, as illustrated in Figure 3d,e, Se signal is uniformly distributed in Li2Se-LiTiO2 composite, whereas Ti and O signals are tightly conjunct with each other. This result vividly demonstrates Li2Se has a good distribution in LiTiO2. Moreover, the surface chemical state of Li2Se-LiTiO2 composite is elucidated by XPS test. As shown in Figure 3c,f, the peaks of Se-3d and Ti-2p both slightly shift towards the high binding energy since some chemotactic bonds are formed between Li2Se and LiTiO2, which will weaken the shielding effect of the electron atmosphere. It is worth noting that compared to commercial anatase TiO2 (Figure S2e), LiTiO2 (Figure 3g) has slightly larger particle size in the range of 60–80 nm with rougher surface and particle agglomeration. This is attributed to the strong collision of TiO2 particles during the synthesis process. Additionally, the lattice spacings of TiO2 (Figure S2f,g) are 0.148 nm and 0.149 nm, respectively, corresponding to the (204) and (213) interplanar distances of anatase TiO2 phase (PDF#21-1272). In contrast, the lattice distances are 0.239 nm and 0.206 nm as shown in Figure 3i, respectively, indexing the (111) and (200) crystal planes of LiTiO2 (PDF#16-0223), thereby further providing the evidence of the existence of LiTiO2 in Li2Se-LiTiO2 composite.

Figure 3

(a,b) SEM images of Li2Se-LiTiO2. (c,f) High-resolution XPS spectra of Se 3d and Ti 2p of Li2Se-LiTiO2. (d) TEM image of Li2Se-LiTiO2. (e) EDS mapping of Li2Se-LiTiO2. (g) SEM image of LiTiO2. (h) TEM image of LiTiO2. (i) HRTEM images of LiTiO2.

Figure 4a–c presents the CV profiles of Li2Se-LiTiO2, Li2Se and LiTiO2 cathodes in the voltage range of 1.7–3.8 V at a scan rate of 0.1 mV s−1. At the first Li+ extraction step, Li2Se-LiTiO2 cathode has two adjacent oxidation peaks at 2.27 V and 2.31 V. According to previous works [17,21,39], the oxidation peaks of LiTiO2 and Li2Se are generally located at 2.18 V (Figure 4c) and 2.31 V (Figure 4b), which are assigned to the conversion of Ti3+ to Ti4+ in LiTiO2 and the formation of hexagonal element Se0 in Li2Se, respectively. Meanwhile, two-stage processes are observed during the first lithiation step, corresponding to the reduction from Se to polyselenides (2.05 V) and then to Li2Se/Li2Se2 (1.85 V) [19,34]. Noteworthy, the position of reduction peaks after the first cycle slightly moves to the higher potential, implying the overpotential is decreased. This is mainly due to the lower de-lithiated potential of LiTiO2, which triggers the Li+ extraction of Li2Se.

Figure 4

(a–c) CV profiles of Li2Se-LiTiO2, Li2Se and LiTiO2 cathodes. (d–g) SEM images of Li2Se-LiTiO2 and Li2Se cathodes at fully charged/discharged state. The insets are the digital photos of separators. (h) UV-vis spectra of Li2Se6 solution with LiTiO2 before/after adsorption test.

To further verify the role of LiTiO2 in the Li+ insert/extract processes of Li2Se, the color change of separators is examined by digital photos and SEM images. As shown in Figure 4d, in the fully charged state, the separator of Li2Se-LiTiO2 turns light reddish-brown, as well as only a few of particles are observed at the surface of separator. In contrast, the separator of Li2Se turns conspicuous brick-red, and many particles appear on the surface of separator as illustrated in Figure 4f. Meanwhile, during the fully discharged state, it shows the similar phenomenon, which the separator of Li2Se-LiTiO2 sample is much cleaner than that of Li2Se sample (Figure 4e,g). Apparently, LiTiO2 facilitates the conversion of Li2Se to Se, and it inhibits the dissolution of polyselenides.

Generally, TiO2 has good chemisorption on polysulfides/polyselenides for Li-S and Li-Se batteries [14,28,29,30,31]. In this work, LiTiO2 is a fully lithiated state of TiO2, which has very different chemical state to TiO2. Therefore, it is necessary to inspect the polyselenide adsorption capability of LiTiO2. As depicted in Figure 4h, we design the simulated polyselenide adsorption experiment, which LiTiO2 is added to the simulative Li2Se6 solution. After 24 h, the color of Li2Se6 solution changes from dark brown to light brown with adding LiTiO2. Moreover, on the basis of UV-vis result, the characteristic peak intensity is dramatically decreased, reflecting the reduced concentration of Li2Se6 in simulated solution. This result clearly indicates that LiTiO2 has superior polyselenides trapping ability, which is favorable to achieve stable electrochemical performance of Li-Se batteries.

The electrochemical performance of Li2Se-LiTiO2 composite is evaluated by coin cells. Figure 5a–c exhibit the galvanostatic charge-discharge profiles of Li2Se-LiTiO2, Li2Se and LiTiO2 cathodes. A noticeable overpotential (2.93 V) can be observed in Li2Se cathode during the first charge profile, which is related to the obstruction of Li+ extraction from crystalline Li2Se and the formation of a new interface [40]. Impressively, the overpotential of Li2Se-LiTiO2 cathode is dramatically diminished to only 2.15 V. It is probably attributed to the presence of oxygen vacancies, where Ti4+ is converted to Ti3+ in the synthesis process [41,42,43]. In the charging process, these oxygen vacancies stabilize the free ions, acting as Lewis acid sites that can interact strongly with polyselenides and release more Li+ [44,45]. Meanwhile, the overpotential of Li2Se-TiO2 cathode is still present (Figure S3a), which further supports our assumptions. In the subsequent charging process, a long plateau at ~2.18 V is observed, which is related to the reversible conversion of Li2Se2/Li2Se to Li2Se (8 ≥ n ≥ 4) and Se [46,47]. Meanwhile, during the discharge process, the Li2Se, Li2Se-LiTiO2 and Li2Se-TiO2 cathodes have two voltage plateaus located at ~2.12 V and ~2.05 V, respectively, which are ascribed to the multistep phase transitions of Se to soluble long-chain Li2Se (8 ≥ n ≥ 4) and further to insoluble short-chain Li2Se2/Li2Se [48,49]. It is worth noting that when the first charging voltage is 2.6 V, Li2Se is nonactivated and the Li+ is not completely detached from Li2Se, exhibiting terrible cycling performance (Figure S4).

Figure 5

(a–c) The initial charge-discharge profiles of Li2Se-LiTiO2, Li2Se and LiTiO2 electrodes at a current density of 50 mA g−1. (d,e) Cycling stability at a current density of 50 mA g−1 and multi-rate cycling performance of Li2Se-LiTiO2, Li2Se and LiTiO2 electrodes.

For the purpose to inspect the long-term cycling stability and multi-rate ability of Li2Se-LiTiO2 cathode, Li2Se and LiTiO2 cathodes are employed as counterparts. As shown in Figure 5d, Li2Se-LiTiO2 cathode delivers a high initial discharge capacity of 398 mA h g−1, which are higher than that of Li2Se cathode (292 mA h g−1). Noteworthy, the capacity fading of Li2Se is 19% after 1 cycle, whereas Li2Se-LiTiO2 cathode is only 9%. The reversible specific capacity of Li2Se-LiTiO2 cathode still remains at 134 mA h g−1 after 100 cycles. In sharp contrast, the specific discharge capacity of Li2Se cathode rapidly decays to 65 mA h g−1 after 100 cycles. Additionally, Li2Se-LiTiO2 cathode exhibits the better multi-rate capability compared to Li2Se cathode, which delivers reversible capacities of 330, 206, 176, 154, 129 and 120 mA h g−1 with upward current densities of 50, 100, 200, 400, 800 and 1000 mA g−1, respectively. Notably, when the current density is returned to 50 mA g−1, a reversible capacity of 160 mA h g−1 is recovered. It should be mentioned that the initial discharge capacity of LiTiO2 is 109 mA h g−1, and it delivers 57 mA h g−1 after 100 cycles. However, the LiTiO2 content in composite is 20%, thereby contributing a small capacity in Li2Se-LiTiO2 cathode. The conductive carbon in the cathode provides few capacities (Figure S5). When the molar ratio of Li2Se to LiTiO2 is 7 to 3, the discharge capacity of Li2Se-LiTiO2 cathode drops rapidly to only 152 mA h g−1 after the 1st cycle, which is related to the high content of LiTiO2 with low specific capacity. Moreover, the cycling stability and multi-rate performance comparison of Li2Se-LiTiO2 composites with various LiTiO2 contents are illustrated in Figure S3c. Apparently, the sample with 20% LiTiO2 demonstrates the best cycling stability and rate capability compared to other counterparts.

Nyquist plots further are performed to confirm the reaction kinetics and electrochemical activities of Li2Se-LiTiO2 cathode before/after cycling. As depicted in Figure S3b, Li2Se-LiTiO2 and Li2Se cathodes both have a semicircle in high frequency along with a sloping straight line in low frequency. Generally, the formation of soluble polyselenides during the charging/discharging process causing an irreversible loss of active material and the retention of polyselenides in the electrolyte, thus inhibiting the transfer of charge. After 100 cycles, Li2Se-LiTiO2 cathode has the smaller semicircle, corresponding to the lower charge transfer resistance than Li2Se cathode. This is attributed to the introduction of LiTiO2 with strong chemisorption on polyselenides, which significantly reduces the internal resistance and further facilitates the mass transfer process.

To further demonstrate the sustained effect of LiTiO2 in stabilizing cycling performance of Li2Se, the separators after cycling are examined by digital photos, SEM and EDS mapping (Figure 6a–d). In the fully-charged state, the diffraction peaks of Li2Se disappear with the conversion of Li2Se to Se, and the newly appeared peaks may be related to the formation of SEI layer and the production of amorphous Se (Figure S6). A large number of reddish-brown particles are observed on the separator of Li2Se cathode. However, only a few of particles are detected on the separator of Li2Se-LiTiO2 cathode. These particles are proved to be polyselenides by EDS mapping characterization. After fully discharging, Se combines with Li+ to regenerate Li2Se and diffraction peaks of Li2Se are observed again (Figure S6). Many particles still remain on the surface of the separator in Li2Se based cell, whereas the surface of the separator of Li2Se-LiTiO2 based cell is spotlessly clean that virtually no particles are observed. This result clearly suggests a long-term role of LiTiO2 to suppress the polyselenide migration and strictly confine polyselenides in the cathode side, matching well with the results in Figure 4d–g.

Figure 6

Digital photos, SEM images and EDS mapping of the separators assembled in Li2Se-LiTiO2 (a,c) and Li2Se (b,d) based cells at fully charged/discharged states after 10 cycles.

4. Conclusions

In summary, an innovative Li2Se-LiTiO2 composite cathode material is successfully developed by a two-step solid-phase method for advanced Li-Se batteries. LiTiO2 with strong chemical adsorption of polyselenides, emerges oxygen vacancies during the conversion of Ti4+ into Ti3+. These oxygen vacancies release of Li+ from polyselenide and improve the utilization of Se. Li-Se battery with this novel cathode exhibits remarkable electrochemical performance in terms of the reduced overpotential from 2.93 V to 2.15 V and high specific discharge capacity that cathode is two times higher than Li2Se cathode. This work provides fantastic inspiration for rationally designing fully pre-lithiated Se-based cathodes in advanced Li-Se batteries.