1 Introduction

Production of hydrogen/oxygen fuels through electrochemical water splitting is considered one of the most efficient green technologies, although large-scale synthesis of cost-effective electrocatalysts used in this process still remains a huge challenge [1,2,3,4,5]. Platinum (Pt)/Pt-based alloys and iridium/ruthenium oxides (IrO2/RuO2) are regarded as the most efficient electrocatalysts for electrochemical hydrogen evolution reaction (HER) and oxygen evolution reaction (OER), respectively [6,7,8,9,10]. However, their high cost and compromised stability as well as the low earth abundance of these metals impede their widespread application [11,12,13,14,15]. Therefore, it is highly desirable to fabricate alternative noble-metal-free and durable electrocatalysts for both OER and HER systems. Although transition metal oxides and hydroxides (NiO, CoO, Ni(OH)2, etc.) [16] are being widely investigated, they mostly have intrinsically low electrical conductivity and their composites with carbon additives should be prepared to improve the electrical conductivity. Metal sulfides, such as nickel sulfide (Ni3S2), are more attractive candidates for electrochemical water splitting, owing to their intrinsic high conductivity, rich catalytic activity, and superior electrochemical stability when applied in HER/OER [17, 18]. Currently, a wide range of nanostructured Ni3S2 (such as Fe-doped Ni3S2 [19] and nanorods [20]) and composites (e.g., Ni3S2 nanosheets/Ni [21], Ni3S2 nanotube/Ni [18]) has been prepared by different methods. They demonstrate improved performance in HER or OER owing to increased exposure of the active sites and boosted ion/electron transfer. Despite these efforts, the overall water-splitting activity of the same Ni3S2-based catalysts for both HER and OER has been rarely reported. In addition, the aforementioned Ni3S2 electrocatalysts are usually synthesized via chemical vapor deposition (CVD) and hydrothermal methods. However, these methods require high-temperature treatment or the use of polluted thiourea or trithiocyanuric acid. Moreover, the high-temperature treatment may cause the coverage or loss of the active sites of Ni3S2 [22,23,24,25]. In this context, a facile and green low-temperature synthesis method for Ni3S2 electrocatalysts is highly desirable.

Low-temperature (< 100 °C) sulfurization using a Na2S solution is a green way for the large-scale synthesis of nanostructured metal sulfides owing to easy processing, high efficiency, and cost-effectiveness. Moreover, this method is particularly suitable for the direct synthesis of metal sulfides arrays with tailored nanostructures. Meanwhile, it has been demonstrated that a high-index-faceted Ni3S2 nanosheet could have superior HER activity owing to possible synergistic catalytic effects arising from the nanosheet array and the exposed {\(\bar{2}10\)} high-index facets [26]. Inspired by these encouraging results, we set out to employ a low-temperature synthesis route to produce Ni3S2 nanoarrays with preferentially exposed {\(\bar{2}10\)} high-index facets as a binder-free electrocatalyst. In addition, in order to further increase the areal load of the active material, we aimed to grow the Ni3S2 arrays as branches on a conductive scaffold to form a core–branch array structure, instead of directly depositing them on carbon cloth.

Herein, we report a facile low-temperature (< 100 °C) sulfurization strategy to synthesize large-scale TiO2@Ni3S2 core/branch arrays as a binder-free electrode for a water-splitting electrolyzer in an alkaline solution. An induced growth process for growing Ni3S2 nanobranch on a TiO2 core obtained by atomic layer deposition (ALD) is proposed, which leads to the in situ growth of {\(\bar{2}10\)} high-index facets of Ni3S2. The as-prepared TiO2@Ni3S2 core/branch arrays possess large active areas, uniform porous structures, and rich active sites of the exposed {\(\bar{2}10\)} high-index facet. These features lead to substantial enhancements in HER and OER activities compared to those of most of the reported Ni3S2-based catalysts. Low overpotentials and fast kinetics as well as superior long-term durability of TiO2@Ni3S2 core/branch arrays are demonstrated. A low-water-splitting voltage of 1.58 V at 10 mA cm−2 is obtained upon using the TiO2@Ni3S2 array electrode as both a cathode and an anode. Our new electrode design strategy paves a green way for the construction of large-scale nickel sulfides with high electrocatalytic efficiency for electrochemical energy storage and conversion applications.

2 Experimental

2.1 Material Synthesis

In the first step, Ni2(OH)2CO3 nanosheet arrays were obtained by a one-step hydrothermal method using commercial nickel foam as the substrate. For this, Ni(NO3)2 (0.9 g), NH4F (0.23 g), and urea (0.9 g) were dissolved in 75 mL of deionized (DI) water to form a reaction solution. Then, the solution was transferred to a Teflon-lined steel autoclave, and the autoclave was placed in an oven at 120 °C for 8 h. After natural cooling, the sample was rinsed thoroughly with DI water.

In order to synthesize TiO2@Ni2(OH)2CO3 nanoflake arrays, the prepared Ni2(OH)2CO3 nanosheet arrays were placed in an ALD reactor (ALD PICOSUN P-300F) along with TiCl4 and H2O as the Ti and O source, respectively. Then, TiO2 was deposited at 120 °C for 140 cycles. Argon was used as the carrier gas. The final step was the sulfurization process. Typically, the obtained TiO2@Ni2(OH)2CO3 nanoflake array samples were transferred to a 0.1 M Na2S solution and heated at 90 °C for 9 h. After natural cooling and rinsing with DI water, the TiO2@Ni3S2 core/branch arrays were obtained. For comparison, Ni3S2 nanoflake arrays were also synthesized by the direct sulfurization of the Ni2(OH)2CO3 nanosheet arrays on nickel foam (without the ALD TiO2 step) using the same sulfurization conditions mentioned above.

2.2 Material Characterization

Morphologies and microstructures of all samples were investigated using a field-emission scanning electron microscope (FESEM, Hitachi SU8010) and a transmission electron microscope (TEM, JEOL 2100F). The crystal structure was characterized by X-ray diffraction (XRD) with Cu Kα radiation (RigakuD/Max-2550). Raman spectra were collected using a Renishaw-inVia Raman microscope with 514-nm laser excitation. X-Ray photoelectron spectroscopy was performed on an ESCALAB_250Xi X-Ray photoelectron spectrometer with an Al Kα source. Specific surface area distributions were obtained using a porosity instrument (BET, JW-BK112).

2.3 Electrochemical Measurements

HER and OER experiments were conducted using an electrochemical workstation (CH Instrument 660D) with a standard three-electrode setup at room temperature; the as-prepared samples, carbon rod (D = 8 mm), and saturated calomel electrode were used as the working electrode, counter electrode, and reference electrode, respectively. A 1 M KOH solution was used as the electrolyte for both HER and OER tests. All potentials in this work are referred to the reversible hydrogen electrode. All measurements were first carried out following a cyclic voltammetry (CV) test at 100 mV s−1 to stabilize the current. Linear sweep voltammetry (LSV) tests were performed at a scan rate of 5 mV s−1. The Tafel plots of the samples were obtained from the LSV curves acquired with a scan rate of 1 mV s−1. Electrochemical impedance spectroscopy (EIS) was performed at the polarization voltage being indexed to the current density of 10 mA cm−2, in the frequency range of 100 kHz to 50 mHz with an AC amplitude of 10 mV. The stability test was carried out after 10,000 CV cycles. These results were obtained by iR compensation. The overall water splitting was performed in a two-electrode catalyzer, where two pieces of TiO2@Ni3S3 samples with a geometric area of 1 cm2 were used as the electrodes for HER and OER.

3 Results and Discussion

3.1 Physicochemical Properties of TiO2@Ni3S2 Core/Branch Arrays

The core/branch structure of the TiO2@Ni3S2 arrays is schematically illustrated in Fig. 1a. Ni2(OH)2CO3 nanoflake arrays were synthesized on commercial nickel foam via a standard hydrothermal process (see details in Sect. 2.1). A TiO2 layer with 10 nm thickness was deposited on the surface of the Ni2(OH)2CO3 nanoflakes using a simple ALD method. The obtained TiO2@Ni2(OH)2CO3 arrays were converted to TiO2@Ni3S2 core/branch arrays by immersing them into a Na2S solution and heated. We applied this unique-structured material as electrocatalyst and propose the following advantages in enhancing the HER and OER:

Fig. 1
figure 1

a Schematic illustration of the core/branch structure of the TiO2@Ni3S2 arrays. b Optical photograph of the sample. c–g SEM images of the TiO2@Ni3S2 core/branch arrays

Full size image
  1. 1.

    Branched Ni3S2 nanoflakes possess a high surface area and higher porosity than those of the pure Ni3S2 nanoflakes grown directly on Ni foam. Further, the open structure of the interconnected nanoflakes will facilitate ion diffusion and H2/O2 detachment during the HER/OER processes. This is particularly beneficial for large-current electrocatalysis.

  2. 2.

    The ALD-TiO2 skeleton not only serves a mechanical support for the Ni3S2 branch, but also induces the nucleation for the directional growth of Ni3S2. Without the ALD-TiO2 skeleton, no Ni3S2-branch can be formed. The TiO2 and Ni3S2 act synergistically to provide better mechanical stability and enhanced specific surface area and larger porosity [27, 28].

  3. 3.

    One important feature of this unique branched Ni3S2 nanoflakes is the exposure of their highly active {\(\bar{2}10\)} high-index facets, which can further improve the HER/OER activities leading to a lower overpotential and Tafel slope.

The morphological evolution of the samples at different stages of the synthesis is revealed by the SEM images (see Fig. S1). The hydrothermally synthesized Ni2(OH)2CO3 nanoflakes with thicknesses between 40 and 60 nm are found aligned vertically on the nickel foam surface, forming an architecture with a porous network (Fig. S1a, b). After the ALD of TiO2, the twisted nanoflakes of Ni2(OH)2CO3 smoothened to form TiO2@Ni2(OH)2CO3 core/shell arrays. Further, the thickness of the TiO2@Ni2(OH)2CO3 core/shell arrays increased to 50–70 nm. However, the 3D porous structure is still preserved, which is not surprising since the ALD generally results in a uniform and conformal deposition of a smooth thin film of amorphous TiO2 (Fig. S1c, d). However, after the final sulfurization in Na2S solution at 90 °C, the morphology changed radically; the previous core/shell structure of TiO2@Ni2(OH)2CO3 transformed into a new type of branched structure of TiO2@Ni3S2. It is observed that the TiO2@Ni3S2 sample is black and the display area is ~ 45 cm2. This process can be easily adapted for large-scale production (Fig. 1b). Meanwhile, the internal TiO2 core is homogeneously coated by the cross-linked Ni3S2 nanoflake shell with 10–15 nm thickness (Fig. 1c, f). Furthermore, the porous morphology remained well preserved in the TiO2@Ni3S2 core/branch arrays. These unique porous structural features provide a number of tunnels to boost electron/ion transfer. As shown in Fig. 1g, the TiO2@Ni3S2 core/branch arrays grew quasi-vertically with respect to the substrate with a height of ~ 1 μm.

The branched microstructure of the TiO2@Ni3S2 arrays was also explored by TEM observation. The Ni2(OH)2CO3 nanoflake presents a thin and smooth appearance (Fig. S2a). The measured interplanar d-spacing of Ni2(OH)2CO3 is about 0.26 nm, which corresponds well with that of the (−201) plane of Ni2(OH)2CO3 (JCPDS 35-0501) (Fig. S2b) [29]. After the ALD of TiO2, the Ni2(OH)2CO3 is completely coated with a thin layer of TiO2 with ~ 10 nm thickness (Fig. S2c, d), forming a TiO2@Ni2(OH)2CO3 core/shell structure. Additionally, the thin TiO2 is amorphous in nature and the interplanar d-spacing of 0.26 nm is still noticed for Ni2(OH)2CO3. As for the TiO2@Ni3S2 sample, the pristine TiO2@Ni2(OH)2CO3 core/shell structure changed to core/branch array, in which the TiO2 core is homogeneously covered by cross-linked Ni3S2 nanoflakes (Fig. 2a). A clear interplanar d-spacing of ~ 0.71 nm is observed, which may be due to the expansion of the c axis of Ni3S2. In addition, the selected area electron diffraction (SAED) pattern shows polycrystalline diffraction rings of the TiO2@Ni3S2 sample (Fig. 2b), which is in good agreement with the (001), (101), (110), and (021) planes of Ni3S2.

Fig. 2
figure 2

a–d TEM-HRTEM images and SAED pattern of TiO2@Ni3S2 core/branch arrays. e EDS elemental maps of O, S, Ni, and Ti in TiO2@Ni3S2 core/branch arrays

Full size image

HRTEM investigation was performed along the [100] zone axis of Ni3S2, and interestingly, the interplanar d-spacing of 0.24 and 0.23 nm matched well with the (003) and (021) planes of the hexagonal Ni3S2 phase (JCPDS 44-1418). Further, the angle between the (003) and (021) facets is approximately 70°, which corresponds to the theoretical value of 70.8° (Fig. 2c, d). This implies that the exposed facets of the Ni3S2 nanoflakes are {\(\bar{2}10\)} high-index facets. According to a previous report by Fang et al. [26], this facet shows superior catalytic performance. Energy-dispersive X-ray spectroscopy (EDS) maps (Fig. 2e) confirm the presence and uniform distribution of O, S, Ni, and Ti in the TiO2@Ni3S2 arrays.

In our experiment, the ALD-TiO2 skeleton serves as a nucleation core for the directional growth of Ni3S2 nanosheets. Without the ALD-TiO2 skin, no Ni3S2 branch can be formed. Comparatively, only common Ni3S2 nanoflake arrays are formed in the absence of the TiO2 layer support (Fig. S3a, b). Interestingly, exposed {\(\bar{2}10\)} facets are also found in common Ni3S2 nanosheet arrays, indicating that the low-temperature sulfurization method is favorable for the growth of the high-index {\(\bar{2}10\)} facet, which is also confirmed by TEM and XRD (Fig. S3c, d). During the sulfurization process, the Ni ions would diffuse outward and combine with sulfur-containing groups (S2−, HS−, etc.) along the outer surface of ALD-TiO2 to form Ni3S2 nanocrystal nuclei. This outward diffusion process might be related to the Ostwald ripening effect, in which the energy of the interior is higher than that on outer surface. Ni3S2 species are spontaneously attached to the ALD-TiO2 surface, which induces the growth of active nucleation centers and decreases the interfacial energy barrier for the self-assembly of the Ni3S2 nanoflake branches.

To further demonstrate the benefits of the core/branch arrays, the specific surface area was determined by BET (Fig. S4). The common Ni3S2 arrays and TiO2@Ni3S2 branch nanosheet arrays show specific surface areas of 1.594 and 4.623 m2 g−1, respectively, implying that branching leads to significantly increased surface area. Furthermore, the branched nanoflakes are beneficial in that they provide increased exposed active area/sites, leading to increased utilization of the active Ni3S2 catalyst.

In order to identify the phase and composition of the final product, XRD, Raman spectroscopy, and XPS were carried out. Figures S5 and 3a show the typical XRD patterns of Ni2(OH)2CO3, TiO2@Ni2(OH)2CO3, and TiO2@Ni3S2. Except for the diffraction peaks of Ni foam, other diffraction peaks in the XRD pattern of Ni2(OH)2CO3 correspond well with those of the monoclinic Ni2(OH)2CO3 phase (JCPDS 35-0501), suggesting the high crystallinity of Ni2(OH)2CO3. For the TiO2/Ni2(OH)2CO3 arrays, the diffraction peaks of Ni2(OH)2CO3 can be still detected, but the strength of them decreases. No peaks of the TiO2 are noticed, indicating the amorphous nature of the TiO2 skeleton (Fig. S5). After the low-temperature sulfurization using the Na2S solution, the diffraction peaks of Ni2(OH)2CO3 disappear, and other diffraction peaks that can be indexed well with the Ni3S2 phase (JCPDS 44-1418) are observed apart from the peaks of metallic Ni foam substrate, demonstrating that the as-obtained TiO2@Ni3S2 sample is of high purity [30]. It is worth noting that the strong diffraction peaks of (021) and (003) plane can be observed clearly (Fig. 3a). Meanwhile, the Raman analysis further confirms the formation of the TiO2@Ni3S2 phase. The TiO2@Ni3S2 arrays show five characteristic peaks at ~ 203, 223, 305, and 348 cm−1, which match well with those of the Ni3S2 phase. The characteristic peak at ~ 150 cm−1 could be indexed well with that of amorphous TiO2 (Fig. 3b) [1], further manifesting the successful preparation of TiO2@Ni3S2 core/branch arrays. This conclusion is also supported by XPS results. Figure 3c shows the high-resolution Ni 2p spectra of the TiO2@Ni3S2 sample. Two main core levels (Ni 2p3/2 and Ni 2p1/2) that are characteristic of the Ni state in Ni3S2 are located at 873.28 and 855.78 eV, respectively [31]. As for the S 2p spectra, two characteristic peaks are detected at 163.28 eV (S 2p1/2) and 161.28 eV (S 2p3/2) corresponding to S2− (Fig. 3d) [32]. Moreover, the presence of TiO2 in the TiO2@Ni3S2 core/branch arrays is also confirmed by Ti 2p and O 1s spectra (Fig. 3e, f). Two core-level peaks are located at 529.8 and 531.1 eV in the O 1s spectra, which could be indexed well with Ti–O and O–H bonds, respectively [33, 34]. Ti 2p1/2 (463.8 5 eV) and Ti 2p3/2 (458.0 eV), the characteristic peaks of TiO2 are observed in the Ti 2p spectra. The presence of O–H bond may be due to surface oxidation of Ni3S2 [35,36,37]. All these results mutually confirm the successful fabrication of TiO2@Ni3S2 core/branch arrays via our powerful low-temperature sulfurization strategy.

Fig. 3
figure 3

a XRD patterns, b Raman spectrum, c core-level Ni 2p XPS spectra, d core-level S 2p XPS spectra, e O 1s XPS spectra, and f Ti 2p XPS spectra of TiO2@Ni3S2 core/branch arrays

Full size image

3.2 Electrocatalytic Properties of TiO2@Ni3S2 Core/Branch Arrays

The electrocatalytic activity of the three samples (Ni2(OH)2CO3, Ni3S2, and TiO2@Ni3S2 electrodes) was studied using a three-electrode system in a 1 M KOH solution [38,39,40,41]. As presented in Fig. 4a, significantly, the TiO2@Ni3S2 electrode displays the best HER activity with the smallest overpotential (112 mV at the current density of 10 mA cm−2), better than that of the Ni2(OH)2CO3 nanoflake arrays (154 mV) and Ni3S2 (149 mV) nanoflake arrays at the current density of 10 mA cm−2. Meanwhile, the TiO2@Ni3S2 core/branch arrays also show a large current density with the lowest overpotential (177 mV at the current density of 100 mA cm−2), superior to those of the Ni2(OH)2CO3 (259 mV) and Ni3S2 (213 mV) counterparts. Additionally, the enhanced HER performance is further confirmed by the Tafel slopes (Fig. 4b) derived from the previous LSV curves. Obviously, the Ni2(OH)2CO3 and Ni3S2 electrodes show large Tafel slopes (105 and 77 mV/decade), while the TiO2@Ni3S2 electrode exhibits a substantially lower Tafel slope of 69 mV per decade. It is well accepted that a lower Tafel slope implies a faster HER rate. As a result, the TiO2@Ni3S2 electrode leads to the fastest HER process.

Fig. 4
figure 4

Evaluation of the HER performance and comparison of Ni2(OH)2CO3, Ni3S2, and TiO2@Ni3S2 electrodes: a LSV curves, b Tafel plots, c current density as a function of scan rate, and d Nyquist plots of the Ni2(OH)2CO3, Ni3S2, and TiO2@Ni3S2 electrodes. e Electrochemical stability of the TiO2@Ni3S2 electrode at a current density of 100 mA cm−2

Full size image

Furthermore, the HER performance of our designed high-index faceted Ni3S2 nanoflake arrays is also excellent. It is well known that HER involves three principal steps including Tafel (30 mV per decade) reactions, Heyrovsky (40 mV per decade), and Volmer (120 mV per decade) mechanisms [42]. Hence, it can be inferred that the HER with Ni3S2 and TiO2@Ni3S2 electrodes in the alkaline water splitting is based on the Volmer mechanism. Simultaneously, the long-cycle durability of electrocatalysts plays an important role in practical application. The electrochemical stability test of the TiO2@Ni3S2 arrays was carried out continuously at the scan rate of 50 mV s−1 for 10,000 CV cycles. The LSV curves of the TiO2@Ni3S2 electrode before and after 10,000 CV cycles of electrolysis nearly overlap with each other, suggesting the excellent stability of TiO2@Ni3S2 electrode (Fig. S6a).

In order to further understand the origin of the superior HER activity of the TiO2@Ni3S2 core/branch nanoflake arrays, the effective electrochemical active surface area (ECSA) of the three samples was calculated by determining the double-layer capacitance (Cdl) based on the CV results at different scan rates (Fig. S6b–d). The obtained current density is plotted as a function of scan rates in Fig. 4c. The ECSA value is considered to be linearly proportional to the Cdl value, equaling half the slope value. Notably, the TiO2@Ni3S2 electrode possesses a high capacitance, up to 42 mF cm−2, which is much higher than those of Ni2(OH)2CO3 (24 mF cm−2) and Ni3S2 (29 mF cm−2) electrodes. The above results indicate that the TiO2@Ni3S2 electrode has more exposed active sites. EIS tests were performed to further probe the electrochemical behavior during the HER process. Figure 4d exhibits the Nyquist plots of all electrodes. The semicircle represents the charge transfer resistance (Rct) of the hydrogen evolution reaction. Remarkably, the TiO2@Ni3S2 electrode shows the smallest Rct value among the three electrodes, which suggests that it facilitates the fastest dynamics during HER. Moreover, the solution resistance (Rs) values of Ni2(OH)2CO3, Ni3S2, and TiO2@Ni3S2 electrodes are 1.49, 1.46, and 1.45 Ω, respectively. These findings further verify that TiO2@Ni3S2 still possesses high electronic conductivity and charge transfer ability during the entire hydrogen evolution reaction. In addition, the TiO2@Ni3S2 electrode also shows long-term durability with no decay after 10 h at a large current density of 100 mA cm−2 (Fig. 4e).

As shown in Fig. 5a, the electrolysis cell of the two-electrode system consists of the TiO2@Ni3S2 electrocatalyst as both anode and cathode in 1 M KOH solution (denoted as TiO2@Ni3S2 || TiO2@Ni3S2). Apart from the outstanding HER activity, the as-prepared TiO2@Ni3S2 electrode also delivers excellent OER catalytic performance in the alkaline solution.

Fig. 5
figure 5

a Schematic illustration of the overall water-splitting process using the bifunctional electrocatalyst. b LSV curves at 5 mV s−1 for the OER performances using the Ni2(OH)2CO3, Ni3S2, and TiO2@Ni3S2 electrodes. c LSV curves of the overall water-splitting performance of the TiO2@Ni3S2||TiO2@Ni3S2 electrolyzer. d Comparison of the overall water-splitting performance of our TiO2@Ni3S2||TiO2@Ni3S2 electrolyzer with those of other electrocatalysts in the literature, and e electrochemical stability at 10 mA cm−2

Full size image

As shown in Fig. 5b, the TiO2@Ni3S2 electrode exhibits a remarkably low OER overpotential of 220 mV at 10 mA cm−2 and 300 mV at 100 mA cm−2, superior to those of the Ni2(OH)2CO3 (330 mV, 390 mV) and Ni3S2 (280 mV, 360 mV) electrodes. Owing to its excellent catalytic activities in both OER and HER, the TiO2@Ni3S2 electrode could be utilized as an attractive bifunctional electrocatalyst for water splitting in an alkaline medium. Impressively, a noticeably low cell voltage of 1.58 V is gained at the current density of 10 mA cm−2 (Fig. 5c), better than those of the other reported bifunctional electrocatalysts (Fig. 5d) [1, 9, 26, 43,44,45,46,47]. In order to investigate the change in the chemical composition of TiO2@Ni3S2, high-resolution Ni 2p spectra were acquired after HER and OER tests (Fig. S7). After HER tests, the Ni 2p spectrum remained almost the same as before with a slight redshift owing to the cathodic H2 reduction. However, after the OER test, the peak at 853.08 eV disappeared and the intensity of the satellite peak (2p3/2) decreased because of the formation of hydrated nickel oxide. Furthermore, the TiO2@Ni3S2 || TiO2@Ni3S2 catalyzer cell shows long-term durability with no decay after 10 h (Fig. 5c, e). All the above results demonstrate that the TiO2@Ni3S2 core/branch arrays possess superior electrochemical activity in both HER and OER, suggesting that the designed TiO2@Ni3S2 core/branch arrays would be promising electrocatalysts for practical application in alkaline water splitting.

4 Conclusion

We developed a facile and high-efficiency low-temperature sulfurization method for the large-scale synthesis of novel binder-free TiO2@Ni3S2 core/branch arrays. Impressively, the as-obtained Ni3S2 nanoflake branches exposed the highly active {\(\bar{2}10\)} high-index facet. Strong support and induced directional growth of Ni3S2 nanoflakes are realized with the aid of the ALD-TiO2 skeleton. Owing to large surface area of the core/branch arrays, large porosity, and binder-free adhesion as well as richer active sites of the exposed {\(\bar{2}10\)} high-indexed facets of Ni3S2 nanoflakes, the designed TiO2@Ni3S2 core/branch arrays deliver low overpotentials and Tafel slopes for both OER and HER as well as cycling stability in an alkaline medium superior to those of the other Ni3S2 counterparts. Our work offers a facile low-temperature strategy to construct advanced metal sulfide catalysts for electrochemical energy conversion and storage.