In situ hybridization of an MXene/TiO2/NiFeCo-layered double hydroxide composite for electrochemical and photoelectrochemical oxygen evolution

Electrochemical and photoelectrochemical (PEC) oxygen evolution reactions (OER) are receiving considerable attention owing to their important roles in the overall water splitting reaction. In this contribution, ternary NiFeCo-layered double hydroxide (LDH) nanoplates were in situ hybridized with Ti3C2Tx (the MXene phase) via a simple solvothermal process during which Ti3C2Tx was partially oxidized to form anatase TiO2 nanoparticles. The obtained Ti3C2Tx/TiO2/NiFeCo-LDH composite (denoted as TTL) showed a superb OER performance as compared with pristine NiFeCo-LDH and comercial IrO2 catalyst, achieving a current density of 10 mA cm−2 at a potential of 1.55 V versus a reversible hydrogen electrode (vs. RHE) in 0.1 M KOH. Importantly, the composite was further deposited on a standard BiVO4 film to construct a TTL/BiVO4 photoanode which showed a significantly enhanced photocurrent density of 2.25 mA cm−2 at 1.23 V vs. RHE under 100 mW cm−2 illumination. The excellent PEC-OER performance can be attributed to the presence of TiO2 nanoparticles which broadened the light adsorption to improve the generation of electron/hole pairs, while the ternary LDH nanoplates were efficient hole scavengers and the metallic Ti3C2Tx nanosheets were effective shuttles for transporting electrons/ions. Our in situ synthetic method provides a facile way to prepare multi-component catalysts for effective water oxidation and solar energy conversion.


Introduction
Over the past years, considerable efforts have been devoted to the exploration of clean and renewable energy sources. The electrochemical and photoelectrochemical (PEC) splitting of water is highly anticipated for the sustainable production of hydrogen and/or oxygen. 1,2 The oxygen evolution reaction (OER), as one of the two half reactions of water splitting, involving complex electron and ion transfer that usually leads to sluggish kinetics and poor energy conversion efficiency. 3,4 Traditional precious-metal-based oxides such as RuO 2 and IrO 2 are among the most active electrocatalysts towards the OER. 5,6 However, their scarcity and accompanying high cost have limited their mass production and wide application.
Recently, MXenesa large family of layered materials have drawn considerable attention, which are commonly produced by the extraction of A from the ternary carbides or nitrides with a formula of M n+1 AX n , where M is an early transition metal, A is an A-group element and X is C and/or N. [31][32][33] As one of the most widely studied MXene, Ti 3 C 2 T x (T x represents the terminal groups such as -(OH) x and -F x ), 34 has demonstrated outstanding performance in various electrochemical applications, 35,36 thanks to its good conductivity and hydrophilicity, as well as high electronegativity. [37][38][39] Importantly, previous studies found that Ti 3 C 2 T x nanosheets/akes could provide Ti source for the surface growth of TiO 2 for solar energy harvesting. [40][41][42] Therefore, combining Ti 3 C 2 T x /TiO 2 and co-catalyst such as LDHs is expected to produce high performance PEC photocatalyst.
In this work, a composite of Ti 3 C 2 T x /TiO 2 /NiFeCo-LDH (denoted as TTL) was prepared by growth of NiFeCo-LDH nanoplates on surfaces of Ti 3 C 2 T x nanosheets solvothermally, during which TiO 2 nanoparticles were simultaneously formed. The hybrid material showed excellent electrocatalytic activity toward OER and achieved a current density of 10 mA cm À2 in 0.1 M KOH at 1.55 V vs. RHE, which is among the best reported OER catalysts. In addition, the composite material was combined with BiVO 4 for PEC-OER and exhibited the much enhanced activity in comparison with the pristine BiVO 4 . The excellent performance for both OER and PEC-OER resulted from a synergistic effect from all its components.
2.2 Preparation of Ti 3 C 2 T x nanosheets 5 mg of Ti 3 AlC 2 powder was mixed with 30 mL of ethanol and ball-milled to form a homogeneous slurry. The slurry was then centrifuged and dried at 60 C for 12 h before being ground into a ne powder. Then, 180 mg of this powder, 600 mg of 1, 4-H 2 NDC, 3 mL of 40% HF, and 60 mL of DI water were mixed in a 150 mL Teon, heated and maintained at 180 C for 6 h. Aer being cooled down to room temperature, the collected precipitate was washed with DMF for several times, and then sonicated in DMF for 16 h to yield isolated Ti 3 C 2 T x nanosheets.

Preparation of NiFeCo-LDH nanoplates
NiFeCo-LDH nanoplates were synthesized using a solvothermal method in a reverse microemulsion system based on a previous report. 42 Typically, 1.65 mL of n-butanol and 2.65 mL of N,Ndimethyltetradecylamine were mixed under sonication to form a reverse emulsion solution, into which, 450 mL of NiCl 2 aqueous solution (0.2 mol L À1 ), 150 mL of FeCl 3 aqueous solution (0.2 mol L À1 ), and 75 mL of CoCl 2 aqueous solution (0.2 mol L À1 ) were added. The resulting mixture was then heated and maintained at 120 C for 12 h in a 5 mL Teon. Aer being cooled down to room temperature, the product was collected by centrifugation and washed with ethanol for at least 5 times.

Preparation of Ti 3 C 2 T x /TiO 2 /NiFeCo-LDH composite
The preparation method of Ti 3 C 2 T x /TiO 2 /NiFeCo-LDH was similar to that of NiFeCo-LDH nanoplates, except that 4 mg of Ti 3 C 2 T x nanosheets were mixed with the aforementioned growth solution for NiFeCo-LDH and then vibrated at 850 rpm in an oscillator for 2 h before being transferred to a 5 mL Teon and heated at 120 C for 12 h.

Characterizations
X-ray diffraction (XRD, SmartLab Rigaku) was performed with Cu Ka radiation (l ¼ 1.54Å) as the X-ray source. Scanning electron microscopy (SEM, Hitachi S-4800) was used for the morphological analysis. To gain the microstructure and composition information, transmission electron microscopy (TEM, Hitachi HT7700), high resolution transmission electron microscopy (HRTEM, JEOL 2100F), and energy-dispersive X-ray (EDX) spectroscopy analyses were performed. The elemental analysis and oxidation state study of the samples were carried out by X-ray photoelectron spectroscopy (XPS, PHI 5000 Ver-saProbe), and the binding energies were corrected for specimen charging effects using the C 1s level at 284.6 eV as the reference.

Electrochemical measurements
Typically, 2.5 mg of the active material, i.e. TTL, NiFeCo-LDH, Ti 3 C 2 T x or IrO 2 , was mixed with 400 mL deionized water, 100 mL ethanol and 10 mL Naon, followed by sonication to form a uniform dispersion. Then, 3 mL of such dispersion was dropcasted onto the surface of a pre-polished glassy carbon (GC) electrode (with a diameter of 3 mm) and then dried naturally at room temperature overnight. Cyclic voltammetry (CV) and linear sweep voltammetry (LSV) measurements were conducted in a three-electrode system on an electrochemical station (Autolab 302N). A Pt foil was used as the counter electrode, a 3 M Ag/AgCl electrode as the reference electrode, and the catalyst-modied GC rotating disk electrode (RDE) as the working electrode. All measured potentials versus Ag/AgCl were converted to the reversible hydrogen electrode (RHE) based on the Nernst equation below: 45 where E (vs. RHE) is the potential referred to RHE, E (vs. Ag/ AgCl) is the applied potential against 3 M Ag/AgCl reference electrode, and E Ag/AgCl is the standard potential of Ag/AgCl reference electrode. All the measurements were conducted in O 2 -saturated 0.1 M KOH solution (pH ¼ 13). The CV and LSV curves were measured at scan rate of 100 mV s À1 and 2 mV s À1 , respectively, at a rotating speed of 1600 rpm. Electrochemical impedance spectroscopy (EIS) measurements were conducted at a xed potential of 1.57 V (vs. RHE) by applying an AC voltage with the amplitude of 5 mV over a frequency range of 100 kHz to 0.01 Hz. The electrical double-layer capacitance (C dl ) of the catalyst was measured from the double-layer charging curves using CV in a small potential range of 1.411-1.464 V (vs. RHE) without apparent faradaic processes occurring (Fig. S7 †). The plot of the current density difference at 1.438 V (vs. RHE) against the scan rate (20, 40, 60, 80, 100 mV s À1 ) was linearly tted, and its slope was the C dl of the tested catalyst (Fig. 4c).
The rotating ring-disk electrode (RRDE) measurements were carried out in a three-electrode cell (CHI 700e, Shanghai, China) using a Pt foil as the counter electrode, a 3 M Ag/AgCl electrode as the reference electrode, and a catalyst-modied RRDE (Garmy RDE710, Beijing, China) as the working electrode with a rotating speed of 1600 rpm in O 2 -saturated 0.1 M KOH solution. The RRDE includes a glassy carbon disk with a diameter of 5 mm and an area of 0.1963 cm 2 , and a Pt ring with an area of 0.1859 cm 2 .

Photoelectrochemical measurements
All PEC-OER tests were conducted on an electrochemical workstation (CHI 660E, CH Instruments, Shanghai) in a threeelectrode cell using a 3 M Ag/AgCl standard electrode as the reference and a Pt foil (1.5 cm Â 1.5 cm) as the counter electrode. Typically, 2.5 mg active material such as LDH, Ti 3 C 2 T x / TiO 2 or TTL composite was mixed with 800 mL deionized water, 200 mL ethanol and 20 mL Naon, followed by sonication to form a uniform dispersion. The concentration of Ti, Ni, Fe and Co elements in LDH, Ti 3 C 2 T x /TiO 2 and TTL dispersions were measured by inductively coupled plasma optical emission spectrometer (ICP-OES), respectively (Table S1 †). For fabrication of a typical photoanode, the loading of TTL onto the 1 cm Â 1 cm BiVO 4 /FTO substrate was 175 mg by drop-wise casting 70 mL of 2.5 mg mL À1 dispersion, which contained $142 mg LDH and $33 mg Ti 3 C 2 T x /TiO 2 . In our control experiments, the loading of LDH on BiVO 4 /FTO was then ensured to be $142 mg (i.e. 57 mL of 2.5 mg mL À1 dispersion) and that of Ti 3 C 2 T x /TiO 2 was $33 mg (i.e. 13 mL of 2.5 mg mL À1 dispersion). Aer the dispersion was drop-casted onto BiVO 4 /FTO, the photoanode was dried naturally at room temperature overnight.
Fluorine-doped SnO 2 (FTO) glass substrates with surface deposited BiVO 4 (BiVO 4 /FTO, purchased from TOEI, Hangzhou, China) and TTL/BiVO 4 /FTO (both with a lm area of 1 cm Â 1 cm) acted as the photoanode. All measurements were performed at room temperature in 0.5 M potassium phosphate (pH ¼ 7) solution. A 300 W xenon lamp (PLX-SEX300, PerfectLight, Beijing) was used to irradiate the photoanode from the back at an intensity of 100 mW cm À2 determined by a power meter (PL-MW2000, PerfectLight, Beijing). The photo energy conversion efficiency (h) of the photoanodes was calculated by the LSV curves using the following equation: 46,47 where 1.23 V is the equilibrium potential for OER, E (vs. RHE) is the applied potential vs. the RHE, J is the photocurrent density at the measured potential, and P light (100 mW cm À2 ) is the power density of illumination. EIS measurements of the photoanodes under illumination were performed on the electrochemical workstation with a 5 mV amplitude perturbation between 100 kHz and 0.01 Hz at open circuit potential.
The incident photon-to-current efficiency (IPCE) was calculated using the equation below: 46 where J light and J dark are the measured photocurrent and dark current density (mA cm À2 ) obtained at 1.23 V vs. RHE, respectively. P is the measured irradiance at a specic wavelength (mW cm À2 ), and l is the wavelength of the incident light (nm).

Preparation and characterization of Ti 3 C 2 T x /TiO 2 / NiFeCo-LDH composite
In a typical process, Ti 3 C 2 T x nanosheets were rstly prepared by selectively removing Al layers in bulk Ti 3 AlC 2 crystals with HF to obtain the layered structures as shown in the scanning electron microscopy (SEM) image in Fig. 1a, followed by sonication to produce exfoliated Ti 3 C 2 T x nanosheets (Fig. 1b and c), and a typical nanosheet showed a thickness of $6 nm (Fig. S1 †). The Ti 3 C 2 T x nanosheets and Ti 3 AlC 2 were further characterized by Xray diffraction (XRD, Fig. 1d). The diffraction peak at 39 , which corresponds to the (104) planes of Ti 3 AlC 2 , 48 is not observed in the pattern of Ti 3 C 2 T x , suggesting the complete removal of Al layers. In addition, the (002) peak of Ti 3 C 2 T x compared to that of Ti 3 AlC 2 shied from 9.3 to 8.6 and became broadened, indicating an enlarged interlayer spacing (from 0.94 nm to 1 nm) and possibly a reduced thickness. The chemical composition and oxidization states of the as-exfoliated Ti 3 C 2 T x nanosheets were studied by X-ray photoelectron spectroscopy (XPS) (Fig. S2 †), and surface-terminated groups like -(OH) x and -F x were identied, 49,50 which rendered the Ti 3 C 2 T x nanosheets hydrophilic and negatively charge.
The as-prepared Ti 3 C 2 T x nanosheets were then used as synthetic templates for the growth of NiFeCo-LDH via a solvothermal reaction as illustrated in Scheme 1a. The synthetic process began with mixing Ni 2+ , Fe 3+ and Co 2+ precursors as well as Ti 3 C 2 T x in a reverse microemulsion system, 44,51 in which N,N-dimethyltetradecylamine and n-butanol were mixed with a volume ratio of 8 : 5. The surface-terminated groups (i.e. -(OH) x and -F x ) of Ti 3 C 2 T x nanosheets might absorb Ni 2+ , Fe 3+ , and Co 2+ ions via the electrostatic interaction. Aer the mixture was heated at 120 C in an autoclave for 12 h, NiFeCo-LDH nanoplates together with TiO 2 nanoparticles were formed on the surfaces of Ti 3 C 2 T x nanosheets (Fig. 2a-c). Importantly, different from the NiFeCo-LDH nanoplates synthesized directly in solution and deposited randomly on a surface (e.g. on a copper grid) as shown in Fig. S3, † the NiFeCo-LDH nanoplates tend to stand up on the surfaces of Ti 3 C 2 T x with edges largely exposed. Differently, TiO 2 nanoparticles show spindle-like morphology (Fig. 2c). Our control experiment result indicated that Ti 3 C 2 T x nanosheets which underwent a similar solvothermal treatment also showed surface deposited TiO 2 nanospindles (Fig. S4 †). It has been reported previously that surface defects might form on Ti 3 C 2 T x nanosheets as a result of HF treatment, which provided preferential nucleation sites and Ti source for the growth of TiO 2 in presence of air at elevated temperatures. 41,52,53 Moreover, EDX mapping of a typical hybrid material in Fig. 2d shows the uniform distribution of Ti, Ni, Co, Fe and O elements, further indicating the successful preparation of the composite. The EDX spectrum in Fig. S5a † shows that the atomic ratio of Ni : Fe : Co in LDH was about 3 : 1 : 0.5, and Cl element was also detected. In addition, from the FT-IR spectrum of LDH (Fig. S5b †), a strong vibration of CO 3 2À at $1377 cm À1 was observed. 56,57 These imply that both Cl À and CO 3 2À ions may act as the charge-balancing anions in the LDH nanoplates. The crystal structure of the composite was analyzed by XRD and compared with pristine Ti 3 C 2 T x and LDH (Fig. 2e). The observed diffraction peak at 11.1 can be assigned to the (003) planes of NiFeCo-LDH (JCPDS no. 51-0463) based on previous reports as well as our control experiment (Fig. S6 †). 54,55 The peak present at 25.2 can be assigned to the (101) planes of TiO 2 with the anatase phase (JCPDS no. 21-1272). 58,59 It is interesting to note that the peak for the (002) planes of Ti 3 C 2 T x shis from 8.6 to 6.3 aer hybridization, indicating an increased interlayer spacing (from 1 nm to 1.4 nm). This can be attributed to the intercalation of metal ions and surface oxidation of Ti 3 C 2 T x during the solvothermal process. Selected area electron diffraction (SAED) and high resolution transmission electron microscopy (HRTEM) characterization were further applied to study the microstructure of the composite. As shown in Fig. 2f, the SAED pattern on a typical hybrid material lying atly on a copper grid clearly shows the (100) and (110) spots for Ti 3 C 2 T x with the six-fold symmetry. Besides, discontinued rings for (200) TiO 2 , (220) TiO 2 , (006) LDH and (009) LDH planes are also observed. Fig. 2g shows a typical square lattice pattern of anatase TiO 2 along the [001] zone axis, with a lattice spacing of 3.5Å for (101) planes. On the same image, a folded Ti 3 C 2 T x could also be observed which shows enlarged interlayer spacing of $1.4 nm, consistent with the XRD peak at $6.3 (Fig. 2e). A   Fig. 2h. From this image, the thickness of the LDH nanoplates can be estimated to be $3 nm. The composition and oxidization states of the hybrid material were further analyzed with XPS and shown in Fig. 3. The Ti 2p spectrum (Fig. 3a) shows a doublet at binding energies of 458.3 eV and 464 eV, which can be assigned to TiO 2 . The other peak at 455.2 eV is attributable to C-Ti-T x (T is O, OH or F). [60][61][62] It is important to note that compared with the Ti 2p spectrum of Ti 3 C 2 T x nanosheets before hybridization (Fig. S1a †), the amount of oxide greatly increased aer the solvothermal process. The Ni 2p spectrum shows two sets of doublets for Ni 2+ (855.4 eV and 872.8 eV) and Ni 3+ (857 eV and 874.5 eV), along with two satellite peaks (denoted as Sat.) at 861.5 eV and 879.4 eV, respectively (Fig. 3b). [63][64][65] For the Fe 2p spectrum, the binding energies of 712.4 eV and 724.6 eV are corresponding to Fe 2p 3/2 and Fe 2p 1/2 bands of Fe 3+ , respectively (Fig. 3c). 66 Similarly, the deconvolution of the Co 2p spectrum suggests the presence of Co 2+ (782.5 eV and 797.8 eV) and Co 3+ (780.6 eV and 796.4 eV) species, along with two satellite peaks at 786.1 eV and 804 eV, respectively (Fig. 3d). 63,65 As a result, Co and Ni exist as multiple valence state and Fe provides +3 species in the hybrid material which are consistent with the pure LDH (Fig. S7 †). Based on the XPS and ICP-OES results, the concentrations of the various components in TTL were calculated as shown in Tables S1-S3. †

Electrochemical measurements
We rst examined the electrocatalytic activity of the TTL composite toward OER in O 2 -saturated 0.1 M KOH solution. For comparison, the performance of NiFeCo-LDH, Ti 3 C 2 T x , and commercial IrO 2 towards OER was also tested on the basis of equal loading of active catalysts (i.e. 0.21 mg cm À2 ). Fig. 4a shows the linear-sweep voltammetry (LSV) measurements of the various catalysts deposited on glassy carbon (GC) electrode. The TTL modied electrode showed an oxidation peak at 1.4-1.5 V vs. RHE, which can be assigned to the Ni 2+/3+ to Ni 3+/4+ as well as the Co 2+/3+ to Co 3+/4+ redox processes, 67,68 consistent with their corresponding cyclic voltammetry (CV) curves (Fig. S8 †). Above this potential, current density rises sharply with O 2 evolution. 67,[69][70][71] In addition, the TTL hybrid catalyst achieved a current density of 10 mA cm À2 at a potential of 1.55 V vs. RHE, which outperforms the state-of-the-art IrO 2 catalyst (1.67 V vs. RHE) and is comparable among the currently best OER catalysts under the same measurement conditions (Table S4 †). Furthermore, the Tafel slope for the TTL composite catalyst (98.4 mV dec À1 ) was lower than that of NiFeCo-LDH (114.8 mV dec À1 ), IrO 2 (120.9 mV dec À1 ), and Ti 3 C 2 T x (230.8 mV dec À1 ) (Fig. 4b), suggesting a more favored OER kinetics of the TTL hybrid catalysts.
The much enhanced performance of TTL catalyst compared to their individual components can be attributed to the synergistic coupling effect. The effective surface areas for the various electrocatalysts for OER were rst analyzed based on the double-layer capacitance (C dl ), which was measured by CV with different scan rates over 1.411-1.464 V vs. RHE, that is, a potential range with no apparent faradaic processes (Fig. S9 †). Fig. 4c shows the current density plotted against scan rate at a potential of 1.438 V vs. RHE, from which C dl can be calculated. As expected, the as exfoliated Ti 3 C 2 T x nanosheets showed a low C dl of 3.4 Â 10 À3 mF cm À2 , whereas NiFeCo-LDH nanoplates showed a slightly larger C dl of 2.1 mF cm À2 , and compared to the LDH, the TTL composite exhibited a more than 4 times increased C dl of 8.5 mF cm À2 , suggesting that the Ti 3 C 2 T x nanosheets not only supported the NiFeCo-LDH nanoplates but also facilitated the easy access of aqueous electrolyte to the active surfaces of LDH nanoplates. 28 Furthermore, the metallike Ti 3 C 2 T x nanosheets could enable fast charge/ion transport within the hybrid catalyst lm, as indicated by the electrochemical impedance spectroscopy (EIS) measurement results (Fig. 4d). In addition, rotating ring-disk electrode (RRDE)  measurement was carried out on the TTL-based electrode. A much smaller current was measured on the ring electrode compared to that on the disk electrode (Fig. S10 †), suggesting a four-electron pathway for water oxidation (4OH À / O 2 + 2H 2 O + 4e À ) with negligible formation of peroxide intermediate. 50,72 The Nyquist plots of the various electrodes were tted by the RC circuit model as shown in the inset of Fig. 4d, which includes a solution resistance (R s ), a charge transfer resistance (R ct ) and a constant phase component (CPE). The obtained R ct values are listed in Table S5 †. As consistent with the LSV curves and Tafel slopes, the R ct of TTL is smaller than NiFeCo-LDH and IrO 2 , demonstrating a faster charge transfer during OER.

Photoelectrochemical measurements
Considering the fact that TiO 2 is a typical semiconductor for light harvesting 46,73 and transition metal LDHs have been known as promising co-catalysts or hole scavengers for PEC-OER, 46,74 the simultaneous formation of TiO 2 and NiFeCo-LDH on Ti 3 C 2 T x is expected to generate an efficient hybrid catalyst for PEC-OER. Therefore, we used BiVO 4 as a model PEC-OER catalyst to study the performance of TTL composite by casting it on pristine BiVO 4 /FTO as a photoanode. The PEC-OER tests were performed in 0.5 M potassium phosphate buffer (pH ¼ 7). As shown in the photocurrent-potential curves in Fig. 5a, under 100 mW cm À2 illumination, the TTL/BiVO 4 photoanode showed a much higher photocurrent density compared with the pristine BiVO 4 over a potential window of 0.2 to 1.4 V (vs. RHE), and achieved a current density of 2.25 mA cm À2 at 1.23 V (vs. RHE), which is about 5 times higher than that of the pristine BiVO 4 (0.39 mA cm À2 ). This also outperforms some previously reported LDH-based hybrids like N-decient C 3 N 4 /N-doped graphene/NiFe-LDH hybrid and reduced titania@LDH hybrid. 60,75 Additionally, the onset potential (potential at which photocurrent exceeds to 0.02 mA cm À2 ) 76 shied from 0.51 V for the pristine BiVO 4 to 0.25 V vs. RHE, indicating the enhanced PEC-OER performance in TTL/BiVO 4 photoanode. To further study the photoresponse of TTL/BiVO 4 in comparison with pristine BiVO 4 , chronoamperometry measurements were carried out at 1.23 V (vs. RHE) under on-off illumination cycles (Fig. 5b). The current density of TTL/BiVO 4 is at least 5 times higher than that of the pristine BiVO 4 , in addition to a prompt and steady response over consequent cycles. However, a spike was observed at the beginning of each illumination cycle, and quickly reduced to a steady state plateau, which can be attributed to a sudden generation of charge carriers and partial recombination. 21,77 This can be due to the fact the TTL was dropcasted onto pristine BiVO 4 , which might result in poorer interface compared to those directly grown on BiVO 4 as reported previously. 33,43,78 To explore the charge transport properties of the PEC catalysts, EIS analysis was conducted at open circuit potential. From the obtained Nyquist plots in Fig. 5c, it can be seen that the semicircle for TTL/BiVO 4 photoanode under illumination is smaller than that of the pristine BiVO 4 photoanode, pointing to a much improved charge transfer between the anode and electrolyte, which is in line with the trend observed in Fig. 5a. Furthermore, the applied bias photon-to-current efficiency (ABPE) of pristine BiVO 4 and TTL/ BiVO 4 photoanodes was plotted as a function of applied potential vs. RHE (Fig. 5d), which was derived from the LSV curves in Fig. 5a. It is obvious that the photoconversion efficiency of TTL/BiVO 4 is much higher than that of the pristine BiVO 4 photoanode over the bias window of 0.2-1.23 V vs. RHE. The maximum efficiency of TTL/BiVO 4 is 0.5% (at 0.84 V vs. RHE) which is over 20 times higher than that of pristine BiVO 4 photoanode (0.02% at 1.0 V vs. RHE). We further measured the incident photon-to-current efficiency (IPCE) of the various photoanodes at a constant applied potential of 1.23 V vs. RHE under monochromatic irradiation from 350 nm to 600 nm (Fig. S11 †). All the hybrid photoanodes demonstrated much higher IPCEs than the pristine BiVO 4 photoanode. The TTL/ BiVO 4 achieved the highest performance with a maximum efficiency of 44.6% at 380 nm. It is worth noting that the efficiency for both the Ti 3 C 2 T x /TiO 2 /BiVO 4 and TTL/BiVO 4 photoanodes peaked at $380 nm, which agreed with the fact that TiO 2 with a band-gap of $3.2 eV acted as a UV responsive photocatalyst.
The possible charge-transfer pathways in the TTL/BiVO 4 photoanode during PEC-OER process was proposed as shown in Fig. 6. The much improved PEC performance of TTL/BiVO 4 photoanode can be attributed to the following reasons. First, the anatase TiO 2 nanoparticles with a wide bandgap of $3.2 eV absorbed light mostly in the UV region (Fig. S12 †), 25,43,79 which supplement the absorption of BiVO 4 with a relatively narrow bandgap ($2.4 eV). [78][79][80] This was also evidenced by the IPCE results (Fig. S11 †). Besides, LDHs have also shown weak semiconducting properties for light harvesting. 21,23-25,79 Second, the conductive and hydrophilic Ti 3 C 2 T x nanosheets might act as effective shuttles for electron/ion transport, which was also reected in the EIS analysis results (Fig. 5c). Most importantly, considering the fact that the valence band levels of most CoFe or NiFe-based LDHs (about À5 to À6 eV) 21,46,81 are higher than that of TiO 2 (about À7.2 eV) 44 and BiVO 4 (about À7.1 eV) 82,83 (all the band level are related to the vacuum level), holes generated in BiVO 4 or TiO 2 upon light irradiation could be effectively scavenged by the LDH nanoplates for the oxidation reactions of Ni 2+/3+ to Ni 3+/4+ as well as the Co 2+/3+ to Co 3+/4+ to take place.

Conclusions
In summary, exfoliated Ti 3 C 2 T x nanosheets were used as synthetic templates for the growth of NiFeCo-LDH nanoplates, and also provided Ti source for the formation of anatase TiO 2 nanoparticles. The resulting composite showed excellent performance in OER. The NiFeCo-LDH nanoplates were mostly standing up on Ti 3 C 2 T x with their $3 nm thick edges largely exposed, leading to a high active surface area for redox reactions. In addition, the Ti 3 C 2 T x nanosheets were highly conductive and hydrophilic, allowing for the easy access of electrolyte and transport of electrons/ions. When the composite was further combined with the standard BiVO 4 lm, excellent performance in PEC-OER was also achieved. In addition to the high catalytic activity of NiFeCo-LDH nanoplates and conductivity of Ti 3 C 2 T x , TiO 2 nanoparticles which were uniformly distributed on Ti 3 C 2 T x provided additional light-harvesting ability. We believe that high performance electrochemical and photoelectrochemical OER catalysts could be achieved by rational design and combination of dissimilar functional 2D materials.

Conflicts of interest
There are no conicts to declare.