Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles

international journal of hydrogen energy xxx (xxxx) xxx

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles Thomas Favet a,b, Val
erie Keller a, Thomas Cottineau a,*, My Ali El Khakani b,**
Institut de Chimie et Proc
ed
es pour l’Energie, l'Environnement et la Sant
e, (ICPEES UMR7515), CNRS, Universit
e de Strasbourg, 25 Rue Becquerel, 67087 Strasbourg, France b Institut National de Recherche Scientifique, Centre Energie, Mat
eriaux et T
el
ecommunication (INRS-EMT), 1650, Boulevard Lionel-Boulet Varennes Qu
ebec J3X 1S2 Canada a

highlights

graphical abstract


Pulsed laser deposition was used to decorate TiO2 nanotubes with CoNi nanoparticles.
Photo-electrochemical activity is enhanced (~50%) by CoNi nanoparticles under sunlight.
Chemical composition and loading of CoNi nanoparticles can be easily adjusted.
Optimum catalyst loading was identified.

article info

abstract

Article history:

The pulsed laser deposition (PLD) technique has been used to decorate TiO2 nanotubes

Received 7 June 2019

(NTs) with cobalt-nickel (CoNi) nanoparticles (NPs). The TiO2 NTs were produced before-

Received in revised form

hand through the controlled anodic oxidation of titanium substrates. The effect of the

19 August 2019

nature of the PLD background gas (Vacuum, O2 and He) on the microstructure, composition

Accepted 22 August 2019

and chemical bondings of the CoNi-NPs deposited onto the TiO2-NTs has been investi-

Available online xxx

gated. We found that the PLD CoNi-NPs have a core/shell (oxide/metal) structure when deposited under vacuum, while they are fully oxidized when deposited under O2. On the

Keywords:

other hand, by varying the CoNi-NPs loading of the TiO2-NTs (through the increase of the

TiO2 nanotubes

number of laser ablation pulses (NLP)), we have systematically studied their photocatalytic

Pulsed laser deposition

effect by means of cyclic-voltammetry (CV) measurements under both AM1.5 simulated

Cobalt/nickel nanoparticles

solar light and filtered visible light. We show that depositing CoNi-NPs on the substrate

Photo-electrochemistry

under vacuum and He increases the photo-electrochemical conversion effectiveness (PCE)

Water-splitting

by 600% (at NLP ¼ 10,000) in the visible light domain, while their overall PCE degrades with

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (T. Cottineau), [email protected] (M.A. El Khakani). https://doi.org/10.1016/j.ijhydene.2019.08.179 0360-3199/© 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

2

international journal of hydrogen energy xxx (xxxx) xxx

NLP under solar illumination. In contrast, the fully oxidized CoNi-NPs (deposited under O2) are found to be the most effective catalyst under sunlight with an overall increase of more than 50% of the PCE at the optimum loading around NLP ~1000. Such catalytic enhancement is believed to result from both an enhanced light absorption by CoO (of which bandgap is of ~2.4 eV) and the formation of a heterojunction between NiO/CoO nanoparticles and TiO2 nanotubes. © 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Hydrogen is a promising alternative fuel that could be used as a clean and sustainable energy source. Due to its high specific energy density and its non-carbonaceous reaction products, it continues to attract much interest and research activities. Most of the hydrogen produced nowadays is based on steam methane reforming and water gas-shift reaction. Both processes are energy intensive, costly, and produce carbon dioxide as a reaction product [1]. Therefore, clean and environmentally friendly hydrogen production methods are extensively investigated, such as water electrolysis using renewable energy sources [2], or direct photo-electrochemical (PEC) water-splitting using solar light for off-grid applications [3]. The latter process relies on the use of a photocatalytic material that can absorb sunlight photons and convert them into electron-hole pairs, which must be spatially separated to trigger oxidation and reduction of water into O2 and H2. From an industrial viability perspective, it has been suggested that PEC cell should have a photoconversion efficiency of at least 10%, last for at least 10 years lifetime and produce hydrogen for 2e4$ per kg [4]. At present, solar panels coupled with conventional water electrolysis are closer to these objectives at least in terms of efficiency, but the hydrogen production cost stays too high because of the expensive costs of the materials used for both PV and the electrolyzer. It is believed that, on the mid-to-long term, photo-electrochemical approaches should be able to meet the challenge of reducing the production costs of solar hydrogen, provided that materials combining high efficiency and stability are discovered and could be integrated in a single system. This study aims at developing new photoanodes with enhanced photoconversion efficiency. A widely studied material as a photoanode for photoelectrochemical water splitting is TiO2, due to its low-cost, high photochemical stability, non-toxicity and its suitable energy band positions for the water-splitting reactions [5,6]. However, due to its large band gap (3.2 eV in its anatase phase), TiO2 light absorption is restricted to the UV part of the solar spectrum (l < 400 nm). Recently, modifying TiO2 to extend its activity in the visible has attracted considerable interest, with different strategies such as noble metal deposition [7,8], cation doping [9], anion doping [10,11], co-alloying [12,13], and dye sensitization [14]. Separation of carriers and charge transport properties must also be considered to enhance the overall efficiency of a material for photoelectrochemical water-splitting. Designing TiO2 at the

nanometric scale is an interesting way to improve such properties, specifically with titania nanotubes array that could be relatively easily prepared by anodization [15]. Indeed, this vertically aligned nanostructure allows a unidirectional path to collect photogenerated electrons, and exhibit few grain boundaries, acting as recombination centers. The production of O2 is commonly called the oxygen evolution reaction (OER). This key step of the water splitting is a four-step reaction, resulting in a slow kinetic and a large overpotential. Therefore, the use of a catalyst for OER is necessary to overcome this energy barrier, in order to facilitate this sluggish reaction. RuO2 and IrO2are certainly the most promising candidates as an OER catalyst [16,17]. However, iridium and ruthenium are rare and quite expensive to envision a scale up of this catalyzed OER process. Currently, many efforts are devoted to develop low-cost and effective catalysts. Decoration with transition metals oxydes such as MnO2 [18], NiO [19] or CoO [20] have been investigated as co-catalysts for the OER reaction. Recently, Cheng et al. have investigated photocatalytic performances of TiO2 nanotubes (NTs) decorated with NiO nanoparticles (NPs), prepared by an electrodeposition method. They have shown that the photocatalytic degradation of 4-chlorophenol with their NiO-NPs/TiO2-NTs composite material is twice more efficient than with a pure TiO2NTs. Similarly, Peng et al. have studied the photocatalytic properties of a cobalt co-catalyst deposited on TiO2-NTs arrays for H2 production. Their cobalt NPs were deposited by a photo-reduction method and their material exhibited a hydrogen production similar to that of TiO2-NTs loaded with platinum as a co-catalyst. All these experimental results point out the interest of the use of Co or Ni nanoparticles as co-catalyst for the enhancement of the photocatalytic activity of TiO2-NTs. In the present work, we have used the pulsed laser deposition (PLD) technique to decorate TiO2-NTs with cobalt-nickel (CoNi) nanoparticles. The PLD is known as a highly efficient physical method for in-situ decoration of any material surface by highly-pure metallic or semiconducting NPs [21e23]. In addition of being a one-step direct synthesis of the NPs directly on the NTs' surface, the PLD is very flexible, offers a large process latitude, and lead to the growth of NPs with no impurities and an intimate interface with the underlying substrates, which favours a more effective charge transfer between the NPs and the NTs [24]. Therefore, photocatalytic activity of materials decorated with PLD-deposited NPs should be higher than for NPs grown by more conventional wet-

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

international journal of hydrogen energy xxx (xxxx) xxx

chemistry techniques (which also require the use and engineering of ligands). Among the various PEC cell configurations [4], we have chosen the case where the photoanode is made out of an ntype semi-conductor (i.e., TiO2 nanotubes decorated with CoNi nanoparticles in the present study), separated from a platinum counter-electrode with a membrane. A schematic representation of our PEC cell with the main chemical reaction involved in the water-splitting process are illustrated in Fig. 1. In this work, we have studied the PLD-based nanodecoration of TiO2-NTs by CoNi-NPs under both different background gas atmospheres and at increasing CoNieNPs catalyst loading through the variation of the number of ablation laser pulses (NLP). The photoconversion efficiency of these newly synthesized CoNi-NPs/TiO2-NTs nanohybrids was systematically investigated under both visible light and simulated sunlight. We were thus able to identify the optimal CoNiNPs growth conditions (background atmosphere and loading) leading to a significant enhancement of the photoconversion performance, under sunlight, compared to bare TiO2 nanotubes.

Experimental methods Synthesis of TiO2 nanotubes Synthesis of TiO2-NTs was performed by using electrochemical anodization of titanium foil (50 mm thick; 99.6%; MaTecK). Prior to anodization, titanium foils were degreased by ultrasonication in acetone, ethanol and Deionized (DI) water (18.2 MU cm) for 10 min in each liquid, and then dried under nitrogen gas flow. Electrochemical anodization of titanium foil was performed in an electrolyte of ethylene glycol (EG; 99.8%; Sigma-Aldrich) with 1%v/v DI-water and 0.3%w/w

3

NH4F (98%; Sigma-Aldrich). Adhesion between NTs and titanium was improved using a pre-anodization step, performed during 1h30 under a constant 45 V potential, applied with a Bio-Logic SP300 potentiostat at room temperature. The firstly grown TiO2-NTs were removed from the surface, and this freshly stripped Ti foil surface was reused for a second anodization step. Thus, self-organized TiO2-NTs were grown on the pre-anodized titanium foil at 20  C, under a constant 45 V potential until a charge of 5C.cm2 was reached. Samples were then carefully rinsed with DI-water and dried under nitrogen flow. The resulting amorphous TiO2-NTs were annealed at 500  C during 4 h under 100 cm3 min1 of air flow, at a heating rate of 5  C.min1 in order to crystallize their structure.

Cobalt/nickel nanoparticles deposition Cobalt and nickel (CoNi) NPs were deposited onto the TiO2NTs by using the Pulsed-Laser Deposition (PLD) technique. Briefly, a KrF excimer laser (l ¼ 248 nm; pulse duration ¼ 15 ns; repetition rate ¼ 20 Hz) is focused with a 45 incidence angle on a polycrystalline 200 -diam. CoNi target (commercially available Ni0.5Co0.5 alloy disk). The on-target laser intensity was of ~1.8  108 W.cm2. The PLD deposition was carried out under three different background gas atmospheres, namely 300 mTorr of O2; 300 mTorr of He and vacuum (2.5  105Torr). The TiO2-NTs/Ti samples were mounted on a rotating substrate holder (5 rpm) placed parallel to the CoNi target, at a distance of 7 cm. The KrF laser beam spot was continuously swept over the target while keeping the rotation motion of the samples holder in order to achieve a uniform NPs deposition during the process. Prior to each PLD deposition, the target was ablated during 5 min to clean its surface. Samples were protected with a shutter during this cleaning step. The CoNi-NPs decoration was

Fig. 1 e Sketch of the photo-electrochemical cell used in this study, with TiO2 nanotubes/CoNi nanoparticles as a photoanode and a platinum counter electrode. Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

4

international journal of hydrogen energy xxx (xxxx) xxx

investigated at different numbers of laser ablation pulses (NLP) varying from 500 to 10,000.

Material characterizations X-Ray diffraction (XRD) measurements were performed on a Bruker D8 Advance equipped with a LynxEye detector and using X-Ray at l ¼ 1.5418  A produced by the Cu ka line of a copper anticathode. X-ray Photoelectron spectroscopy (XPS) measurements were performed with a Thermo VG Scientific spectrophotometer (Pass Energy 20 eV). Etching of our samples was performed with Arþ ions (10 mA; 1 KeV; 5.106 mbar). The acquired XPS spectra were analyzed with CasaXPS software. The energy position of the XPS spectra was eventually corrected using the adventitious carbon peak as an internal reference (285.0 eV). The morphology of our samples was examined by using a Gemini-SEM 500 (Zeiss) field emission gun scanning electron microscope (SEM). Their micro/nanostructure was observed by means of a JEOL 2100 transmission electron microscope (TEM) operating at an applied voltage of 200 kV (with a nominal resolution of 0.2 nm). Cyclic Voltammetry was performed in a homemade 3 electrode cell, with a platinum wire as a counter-electrode, and a Mercury Sulfate Electrode (MSE þ0.64 V vs NHE) as a reference electrode. Light source was a Xe arc lamp (200 W; Newport) equipped with a water filter to avoid heating of the cell and an AM1.5G filter that simulate solar spectral distribution. Measurements were performed with this AM1.5 solar irradiation, but also with the addition of a 400 nm high-pass filter to isolate the response of our samples in the visible spectral region. Power density reaching the cell was calibrated with a power-meter at 100 mW cm2 (AM1.5 filter) and 70 mW cm2 (AM1.5 þ 400 nm filter). Voltammograms were recorded with a scan rate of 10 mV s1, in a potential range from 1.5 V to 0.3 V vs MSE (0.86 Ve0.34 V vs. NHE), using a BioLogic SP300 potentiostat. All measurements were performed in a deoxygenated 0.01 M NaOH þ 0.1 M Na2SO4 electrolyte.

Results and discussion Fig. 2 shows the XRD patterns of TiO2-NTs before and after their PLD-decoration with CoNi-NPs (at an NLP ¼ 10,000) under the three different background atmospheres (300mTorr of O2, 300 mTorr of O2 and vacuum). The XRD patterns of Fig. 2 displays various diffraction peaks confirming the effective crystallization of the TiO2-NTs in the anatase phase, with a small amount of rutile. Indeed, the main anatase (110) crystalline peak, at 2q ¼ 25.32 (JCPDS 03-065-5714), is clearly seen on every XRD pattern. As the annealing temperature reached 500  C, the anataseto-rutile transition phase began to occur, and is observed on each sample, with the appearance of the most intense (100) diffraction line of rutile at 2q ¼ 27.40 (JCPDS 01-089-06975). The mixture of a majority anatase and a minority rutile phases has been reported to be beneficial for the TiO2 based photocatalysis [25]. Moreover, several diffraction peaks related to the underlying metallic titanium (Ti-a) substrate are also present in the XRD pattern. It is to be noted that no

Fig. 2 e XRD patterns of TiO2 nanotubes before and after their PLD-decoration with 10,000 laser pulses of CoNi under O2, He and Vacuum atmospheres.

diffraction peaks related to either metallic or oxide forms of nickel or cobalt were clearly observed. This is not surprising given the tiny amount of laser deposited CoNi nanoparticles, in comparison to the ~3.8 mm-thick TiO2-NTs layer. (NLP of 10,000 leads to a ~510 nm-thick film of CoNi-NPs when deposited onto atomically smooth and flat Si substrate, but these CoNi are actually dispersed all over the high surface area of the TiO2-NTs array.). Another hypothesis is the too low crystallinity of CoNi NPs, as it has already been observed for Co3O4 coatings prepared by PLD in similar experimental conditions [26]. To gain more insights on the surface composition of our TiO2-NTs/CoNi-NPs nanohybrid materials and their chemical bonding states, we have performed systematic X-ray photoelectron spectroscopy (XPS) measurements on the samples. Fig. 3a shows the Ti 2p XPS spectra of the TiO2-NTs sample decorated with CoNi-NPs (NLP ¼ 10,000, under 300 mTorr of O2). The main line is the doublet Ti 2p3/2 and Ti 2p1/2 with the respective binding energies of 458.9 eV and 464.6eV, which are characteristics of Ti4þ in TiO2 [27]. The O 1s XPS region of the samples prepared under various atmospheres (at NLP ¼ 10,000) is presented in Fig. 3b. For each sample, three contributions are observed. Binding energy positions of those peaks for each sample are summarized in Table 1. The energy position of the O1s-A peak (red component in Fig. 3b) is in accordance with oxygen involved in TieO bonds [28]. The O1s-B peak (orange component in Fig. 3b) is coherent with oxygen engaged in NieO and CoeO bond [29].

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

5

international journal of hydrogen energy xxx (xxxx) xxx

Table 1 e Binding energies of the different elements presents in the TiO2-NTs/CoNi-NPs samples prepared under 300 mTorr of O2, 300 mTorr of He and vacuum (all decorated with NLP ¼ 10 000). Peak TiO2eCoNi (under O2)

Ti

TiO2eCoNi (under vacuum)

TiO2eCoNi (under He)

Position (eV)

Position (eV)

Position (eV)

458.9

458.4

459.0

464.6

464.1

464.9

528.3

528.0

529.2

TieO

529.3 530.7

529.7 531.0

530.7 532.4

NieO/CoeO OH/C]O

781.1

780.6

779.5

Co2þ

786.6

785.6

785.4

796.8

796.7

795.5

802.3

803.5

802.1

855.9

855.9

854.8

862.5

861.9

861.1

873.4

873.3

872.4

880.1

879.8

879.3

Chemical state assignation

Ti4þ

2p3/ 2

Ti 2p1/ 2

O 1s A O 1s B O 1s C Co 2p3/ 2

Co 2p3/ 2sat Co 2p1/

Co2þ

2

Co 2p1/ 2

Fig. 3 e XPS spectra of (a) Ti 2p region, (b) O 1s region, (c) Co 2p region and (d) Ni 2p region for the TiO2-NTs samples decorated with CoNi-NPs under vacuum, 300 mTorr of O2 and 300mTorr of He at an NLP of 10,000 pulses.

sat Ni 2p3/

Ni2þ

2

Ni 2p3/ 2

Finally, the O1s-C broad peak (blue component) is attributed to surface OeH bond and C]O from organic pollution [30]. The XPS core-level spectra of Co 2p region presented on Fig. 3c display a doublet Co 2p3/2 and Co 2p1/2 with respective binding energies of 780.4 ± 0.8 eV and 795.6 ± 0.7 eV [31,32]. The binding energies of this doublet and the energy split (DE ¼ 15.2 eV) are consistent with Co2þ in CoO [33]. The Co 2p doublet is also associated with two shake-up satellite peaks, found at 785.9 ± 0.6 eV and 802.6 ± 0.8 eV, which are also characteristic of Co2þ [34,35]. On the other hand, Fig. 3d shows the Ni 2p XPS core-level spectra which also present a main line doublet (Ni 2p3/2 and Ni 2p1/2 with respective binding energies of 855.5 ± 0.6 eV and 873.0 ± 0.6 eV), which is consistent with Ni2þ in NiO [36,37]. This doublet is also associated with two shake-up satellite peaks, found at 861.8 ± 0.7 eV and 879.7 ± 0.4 eV respectively, characteristic of Ni2þ [31,38,39]. XPS results indicate that nickel and cobalt species are always oxidized (NiO and CoO) regardless of the used background atmosphere during PLD deposition. Such an oxidation is expected for the CoNi-NPS deposited under O2 atmosphere, but in the case of vacuum and He atmospheres, the oxidation could result from the air-exposure of the TiO2-NTs/CoNi-NPs samples films, since XPS is mainly probing the very shallow surface layer of the samples.

sat Ni 2p1/

Ni2þ

2

Ni 2p1/ 2

sat

To obtain more information on the in-depth oxidation of the CoNi-NPs, we proceeded with in-situ Arþ ion etching of the surface of the samples prepared under vacuum and under 300 mTorr of O2. The XPS spectra were recorded before etching and after every 5 min of ion etching. Fig. 4 presents the XPS core-level spectra of Co 2p and Ni 2p of the TiO2-NTs/CoNiNPs samples (decorated with NLP ¼ 5000 under vacuum). It is clearly seen in Fig. 4a that just after 5 min of ion etching, a new (Co 2p3/2, Co 2p1/2) doublet appears in addition to the previously observed Co 2p3/2 and Co 2p1/2 doublet associated with Co oxide bonding. The new doublet is found at the respective binding energies of 779.0 eV and 794.2 eV, which are associated with metallic cobalt Co0 [31,40]. This metallic cobalt doublet becomes more prominent as the etching time is increased to 10 and 15 min. The same

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

6

international journal of hydrogen energy xxx (xxxx) xxx

Fig. 4 e XPS spectra of (a) Co 2p region before and after etching and (b) Ni 2p region before and after etching for the TiO2-NTs/CoNi-NPs samples decorated at NLP ¼ 5000, under vacuum).

phenomenon is observed on the Ni 2p XPS spectra (Fig. 4b). As the etching time is increased from 5 to 15 min, the Ni 2p spectrum features a new doublet with binding energies of 853.9 eV and 871.1 eV respectively, characteristic of metallic nickel Ni0 [32,41,42]. On the other hand, the XPS spectra of Ti 2p and O 1s regions before and after etching steps (cf. Fig. S1 in Supplementary Information) also indicate a reduction of Ti4þ into Ti3þ and Ti2þ after bombardment with Arþ ions, due to the removal of oxygen atoms from the lattice. XPS results have shown a reduction of titanium to lower oxidation states, and the presence of cobalt and nickel under metallic species after the etching process. The same phenomenon was observed for the samples prepared under 300 mTorr of O2 (Shown on Fig. S2 in Supplementary Information). This could be due to a preferential removal of oxygen atoms during the Arþ ion etching process or to the presence of CoNi-NPs with an oxide/metal shell/core structure formed during the PLD synthesis of the NPs. To better understand the extent of the oxide layer and its dependence of the deposition background atmosphere, more detailed analysis of the XPS spectra were undertaken. Fig. 5a shows the evolution of the metallic cobalt state (Co0) with respect to the overall amount of cobalt (Co0þCo2þ), as derived from XPS analysis. Before etching, no metallic cobalt is observed. After 5 min of ion etching, there is a clear increase of this ratio for the sample with CoNi-NPs deposited under vacuum, whereas there is no change after 5 min of etching for the sample deposited under oxygen. A significant shift from Co2þ to Co0

Fig. 5 e Variation of (a) the (Co0:Cototal) ratio with etching time, (b)The (Ni0:Nitotal) ratio with etching time and (c) (Nitotal:Cototal) ratio with etching time. All ratios are determined based on XPS quantification, for the TiO2-NTs samples decorate with CoNi-NPs at NLP ¼ 10,000(under O2) and at NLP¼5000 (under vacuum).

only occurs after 10 min of etching for the sample prepared under O2. Nevertheless, for the same etching time, the relative proportion of metallic cobalt Co0 always stays higher for the CoNi-NPs deposited under vacuum than for those deposited under O2. Similarly, Fig. 5b compares the (Ni0:Nitotal) ratio for samples prepared under vacuum and O2 atmospheres. As observed for cobalt before etching, no metallic nickel is observed for both samples, but after 5 min of ion etching, this ratio increases and then seems to reach a plateau that is somewhat lower in the case of CoNi deposited under O2, indicating a lower relative content of the metallic state. Both Co and Ni results would suggest that the CoNi NPs prepared in vacuum could exhibit a core-shell structure, with an oxide layer on its surface, and a metallic core. Such a structure is in accordance with the work of Shen et al. [43], where they have reported a core/shell structure of Zn/ZnO nanoparticles synthesized by laser ablation method in an inert atmosphere. This metallic chemical state has also been observed on Co nanoparticles embedded on boron thin film, prepared by PLD under vacuum condition, by Patel et.al [44]. Furthermore, metallic Ni0 and Co0 components appearing for the samples prepared under oxygen atmosphere could be related to the etching process, which removes preferentially oxygen atoms, and triggers reduction of those species. Therefore, we believe that the core/shell structure occurs when the NPs are grown under vacuum, but is likely not present when the CoNi-NPs are synthesized under O2 atmosphere in which a bulk oxidization of NPs is expected. Finally, by comparing the total Ni to Co atoms ratio (Nitotal:Cototal), it is seen that while there is

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

international journal of hydrogen energy xxx (xxxx) xxx

slight difference in the outer surface (before etching), the (Nitotal:Cototal) ratio stabilizes around 1 after 5 min of etching time, confirming the stoichiometry of the CoNi in the target used for their synthesis (PLD is known for its congruent stoichiometry transfer even for complex multi-elements materials [45]). In sum, XPS results confirm a successful decoration of TiO2-NTs with CoNi NPs, which could not be observed by XRD analysis. Moreover, it can be inferred from the XPS results that the CoNi-NPs prepared under vacuum exhibit a core/shell structure while the NPs prepared under O2 are presumably fully composed of metal oxide due to the more oxidative atmosphere during the PLD process. To study the morphology of the TiO2-NTs and CoNi-NPs, Scanning Electron Microscopy (SEM) has been performed on our samples. Electrochemical anodization of titanium foil resulted in growth of self-ordered TiO2-NTs, as can be seen on Fig. 6a. The average length of NTs has been estimated to 3.7 ± 0.2 mm, their outer diameter is of 100 ± 5 nm, and their inner diameter is of 50 ± 5 nm. It has been reported in the literature that 4 mm long TiO2 nanotubes is in the range of optimal size to have sufficient light absorption and low recombination of electron-hole pairs [46]. Fig. 6b, c are topview SEM images of the TiO2-NTs before and after their decoration with CoNi-NPs, respectively. Before the deposition (Fig. 6b), the TiO2-NTs are seen to organize into a compact hexagonal pattern improved by the pre-anodization step done at the beginning of the synthesis. After the PLD decoration process, the CoNi-NPs are seen to be on the top of the walls of the NTs where they tend to agglomerate (see Fig. 6c). Even if the SEM images do not provide the necessary resolution to determine accurately the size distribution of the CoNi-NPs, we can infer

7

that the CoNi-NPs generally have a size below 20 nm. As catalysis strongly depends on the surface area, the small size of nanoparticles would have beneficial effect toward OER, due to a high surface to volume ratio. To examine more accurately the nanostructure of the TiO2-NTs/CoNi-NPs samples, Transmission Electron Microscopy (TEM) observations have been performed. Fig. 6d show an whole chunk constituted of aligned and unbroken TiO2-NTs detached from a sample decorated with CoNi-NPs (at NLP ¼ 20 000, under 300 mTorr of O2). From these randomly positioned TiO2-NTs blocks and the presence of several overlapping layers of NTs on the TEM grid, it is difficult to distinguish the top of the NTs (where there is a higher probability to find CoNi-NPs) from their bottom, even if electrochemical anodization synthesis is known to produce slightly thinner walls on the top of NTs than at their bottom [47]. To better locate the CoNi-NPs, local Energy Dispersive Xray spectroscopy (EDX) quantifications have been performed all along the TiO2-NTs blocks. Fig. 7 shows the variation of the atomic percentages of CoNi, Co, and Ni normalized with the overall amount of cation (Co, Ni, Ti) estimated after EDX analysis, with respect to their position in the nanotubes. At 0 mm, corresponding to the surface of the NTs (1st red circle on Fig. 6d), cobalt and nickel represent half of the cations, with 23.3%at of Co and 21.4%at of Ni, confirming a [Co]/ [Ni] atomic ration of ~1, in agreement with the abovediscussed XPS results after ion etching. The second EDX analysis (red circle 2 on Fig. 6d) was performed 280 nm from the edge of nanotubes, where the ratio of CoNi with respect to the overall amount of cations drastically drops to 2.2%at. This point out that top of the nanotubes are on the down part of

Fig. 6 e Typical SEM images of (a) cross-sectional view of TiO2-NTs tilted at an observation angle of 45 ; top-view of the TiO2NTs before (b) and after (c) their CoNi-NPs decoration (with NLP ¼ 2000, under O2 atmosphere); TEM image of (d) an agglomerate of unbroken TiO2-NTs with CoNi deposited under O2, at NLP ¼ 20,000 (the red circles with their associated numbers indicate the areas where the EDX analysis was performed), (e) a TEM image and (f) a high resolution TEM image of CoNi nanoparticle seen on top of TiO2-NTS. The inset represents the Selected Area Electron Diffraction (SAED) pattern associated with figure f. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article). Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

8

international journal of hydrogen energy xxx (xxxx) xxx

Fig. 7 e Atomic percentages of CoNi, Co and Ni normalized with the overall amount of cation (Co þ Ni þ Ti), with respect to their location in the nanotube. Fig. 6d (red circle n 1), where the CoNi concentration is much higher. This ratio thereafter stabilizes when measured at 1 mm deep in the nanotube, reaching 0.5%at. It is worthy to notice that the amount of nickel is very few compared with cobalt when they are measured deeper in the nanotubes. This is in contradiction with the stoichiometric ratio measured previously. A possible explanation for this phenomenon is that the overall amount of cobalt is identical to nickel, but nanoparticles can be enriched with either of these elements. Thorough EDX measurements at different locations on the TiO2-NTs blocks have confirmed the presence of CoNi-NPs all along the NTs with a more important density on the top-side of the hexagonal titania array. It is also clear that CoNi-NPs enters into the inner porosity of the TiO2-NTs. Fig. 6e shows a high resolution TEM image of the top-side of TiO2-NTs where agglomerated CoNi nanoparticles are clearly seen. The thickness of the CoNi NPs layer is in the range of 50e100 nm. To have a better insight of the morphology of these nanoparticles, a magnified TEM image of the co-catalysts is shown on Fig. 6f. These nanoparticles are seen to present a high crystallinity with clearly apparent lattice fringes, and display a diameter of <5 nm. This small observed diameter could also explains why no diffraction peak associated with cobalt or nickel on the XRD pattern of CoNi decorated samples. EDX measurement performed on this area at the top of the tube show that the overall amount of cat ions are 10.7%at of Ti, 47.8%at of Co and 41.5%at of Ni. The above discussed XPS binding energies indicate that cobalt and

Table 2 e Interplanar distances of CoNi nanoparticles measured at different area on Fig. 6f. Area Interplanar distance measured ( A) Attribution 1 2 3 4

2.07 2.02 2.00 1.66

CoO (200) - 2.13  A NiO (200) - 2.09  A CoO (220) - 1.51  A NiO (220) - 1.48  A

nickel are in oxidation state 2þ. Taking this into account, measurements of the interplanar distances performed on several areas (red circles on Fig. 6e) sum up in Table 2 are consistent with NiO [48,49] and CoO [50]. In addition, the SAED pattern showed on the inset of Fig. 6f reveal diffraction rings of CoNi nanoparticles, in good agreement with CoO and NiO lattice (Measured radius of each ring is shown on Table S1 in Supplementary Informations). The presence of these diffraction rings is ascribed to the different orientations of nanoparticles analyzed in SAED. One can also notice diffraction spots on the SAED pattern, corresponding to the TiO2 nanotubes. TEM analysis has pointed out the small size of CoNi of nanoparticles. In addition, SAED measurements confirms crystallization of NPs into CoO and NiO, in good agreements with the Ni2þ/Co2þ chemical state revealed by XPS analyses. To study the photo-electrochemical performance of our TiO2-NTs based samples, cyclic voltammetry (CV) measurements have been performed before and after their PLD decoration with CoNi-NPs. Fig. 8a shows a typical voltammogram of a TiO2-NTs sample before and after its PLD-decoration with CoNi-NPs. The first measurement is always performed in the dark, to estimate current density without illumination. A large triangle is observed at most negative potentials, which arises from the insertion of protons in the TiO2 lattice. Under irradiation, when the applied potential increases, band bending of TiO2 occurs and enhances charge separation and therefore the photo-current density increases. Electron collection will eventually reach a limit imposed by the photon flux, which results in a current density plateau, observed at most positive potentials. The difference between current density of the plateau under illumination and in the dark represents the net produced photocurrent Jph (Fig. 8a). Several calculations can be used to estimate the efficiency of a PEC cell. The most common one is the applied bias photoconversion efficiency (ABPE), which represent the percentage conversion of light energy into chemical energy minus the energy applied by the bias external voltage, according to the following formula: hABPE ¼

ðE0  DEÞ  Jph  100 P

(1)

With DE ¼ |Eapp| - EOCV|, E0 is the standard potential of water splitting (1.23 V), Eapp is the applied potential (V), EOCV is the open circuit voltage (V) under irradiation, Jph is the photocurrent density (mA.cm2), and P is the power density of the light source (mW.cm2). Applying this formula to the cyclic voltammetry data allows to determine the operating point of our PEC cell i.e. the applied bias for which the efficiency is the best. The average photoconversion efficiency for raw TiO2 nanotubes is ~0.15% under solar light according to this calculation. Nevertheless, here, in order to compare the performances of each electrode before and after deposition, we have chosen to present our results with ε (described below), which expresses the photocurrent density enhancement, instead of giving absolute values of photoconversion efficiency (ABPE). This choice was made, because when NPs are deposited at the surface the EOCV value can change and electrochemical reaction can occurs and result in a small current that sums up to the photocurrent and introduces an error in the efficiency

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

international journal of hydrogen energy xxx (xxxx) xxx

9

Fig. 8 e (a) Typical cyclic voltammogram recorded in (0.01 M NaOHþ0.1 M Na2SO4) under both dark and AM1.5 illumination for the TiO2-NTs samples before and after their decoration by CoNi-NPs (300 mtorr of O2; NLP ¼ 500); The variation of (b) ε and (c) εvis of the TiO2-NTs samples decorated with different CoNi-NPs catalyst loading (i.e.; NLP) pulsed-laser-deposited under three different background gas atmospheres: (b) vacuum, (c) 300 mTorr of He and (d) 300 mTorr of O2.

determination. For these reasons, the photocurrent values at high applied potential, where parasitic effects have no influence, were used in the analysis. In order to highlight the contribution of the CoNi-NPs catalyst to the photocurrent production, the following calculations, have been made from CV results: ε¼

aðTiO2 þCoNiÞ  aðTiO2 Þ  100 aðTiO2 Þ

εvis ¼

a¼

avisðTiO2 þCoNiÞ  avisðTiO2 Þ avisðTiO2 Þ

Jph P

avis ¼

Jphvis Pvis

 100

(2)

(3)

(4)

(5)

where ε; and εvis are the photo-electrochemical enhancement under sun simulated and visible filtered lights, respectively; Jph and Jphvis are the photocurrent densities (expressed in mA.cm2) measured at 0.34 V vs. NHE during cyclic voltammetry under AM1.5 illumination (UV þ Vis) and under visible light illumination (AM1.5 plus 400 nm high-pass filter). The average photocurrent density for TiO2 NTs prior to their decoration by CoNi-NPs is of 0.17 ± 0.01 mA.cm2 under solar light, and 0.009 ± 0.002 mA.cm2 under visible light. P and Pvis are the light power density (mW.cm2) reaching the sample under solar and visible light illumination. Equations (2) and (3) allow to quantify the photo-electrochemical enhancement brought by the presence of CoNi-NPs catalyst on the TiO2-NTs samples illuminated with AM1.5 solar simulator and filtered visible light, respectively. For a rigorous comparison, Jph was normalized with light power density. Fig. 8b, c shows the respective variations of ε and εvis with the CoNi-NPs loading (i.e.; NLP) of the TiO2-NTs samples, under the three PLD background atmospheres. Fig. 8b clearly shows, for the TiO2-NTs samples decorated with CoNi-NPs under vacuum atmosphere (blue curve in

Fig. 8b), that ε decreases with NLP (NLP is the parameter that controls simultaneously the catalyst loading and the average size of the NPs [23,51]). Same behavior is observed for the TiO2NTs samples decorated with CoNi-NPs deposited under 300mTorr of He (green curve in Fig. 8b). This could be due to the core/shell structure of the NPs, suggested by XPS analyses, where the metallic core could reflect or backscatter incident photons, leading to a decrease of the photo-electrochemical activity. The variation of ε with NLP for the samples decorated with CoNi-NPs deposited under 300 mTorr of O2 is shown in red in Fig. 8b. At lower NLP values (2500), a significant increase of the photo-electrochemical activity is observed under solar illumination (UV þ Vis). Indeed, ε was found to reach up to 54% and 49% at NLP ¼ 500 and NLP ¼ 2000, respectively. For higher NLP values (5000), ε is seen around 0%, indicating that any beneficial effect that originates from the presence of CoNi-NPs catalyst is counteracted by a lesser absorption of the light by the TiO2-NTs, of which surface is likely over coated by an excess of NPs. Indeed, as the number of laser pulse increase, both catalyst loading and NPs size increase. At low NLP, the smaller size of deposited NPs allows a higher surface area with more active sites, more favorable for the OER catalysis. Therefore, the better use of the photogenerated charges can lower the recombination rate, resulting in the increased observed ε at NLP ¼ 500 and NLP ¼ 2000. In contrast, at higher NLP, catalysts loading and NPs size increase. This smaller surface to volume ratio is less efficient toward the OER catalysis. In addition, bigger nanoparticles agglomerates are more likely to diffuse or backscatter incident light and obstruct its absorption by TiO2 nanotubes, leading to a decrease of ε. These results clearly point up that there is an optimal catalyst loading (NLP~2000), where a balance is reached between the amount of NPs required to catalyze the photoconversion process while not hindering light absorption by the underlying TiO2-NTs. The enhancement of photoelectrochemical activity with NLP of TiO2 NTs decorated with CoNi-NPs was also investigated under visible light illumination (εvis), as shown on Fig. 8c. εvis of samples prepared under vacuum (blue curve in Fig. 8c) and 300 mtorr of He (green curve

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

10

international journal of hydrogen energy xxx (xxxx) xxx

in Fig. 8c) is found to increase almost linearly with number of laser pulses. The enhancement of photoconversion under visible illumination brought by CoNi NPs reaches almost 600% (300 mtorr He) and 700% (vacuum) at NLP ¼ 10,000. Nonetheless, this catalytic improvement under visible light doesn't lead to an increase of the overall activity. In fact, the metallic core of NPs deposited under vacuum or helium atmosphere can absorb visible light due to surface plasmon effect [52], and transfer excited electrons to the surface of TiO2-NTs, thus participating to the increase of the observed photocurrent under visible light. J.Gao et al. have studied photoconductivity of Co-doped amorphous carbon/silicon heterostructures, prepared by pulsed laser ablation. Optical absorption spectra of their material shows a peak centered around ~710 nm, signature of surface plasmon resonance in their samples [53]. Another explanation could be that CoO absorbs photons in the visible, due to its narrow band gap (e.g., ~2.4 eV) and actively participate to photo-electrochemical reaction [54]. NiO semiconductor has a larger bandgap (e.g., ~3.5 eV) [19], and couldn't be responsible of this enhanced activity in the visible. Under visible light, the photo-electrochemical activity enhancement εvis of CoNi-NPs deposited under O2 (red curve in Fig. 8c) is found to be less exacerbated than in the case of the CoNi-NPs deposited under either vacuum or He. Since the CoNi-NPs deposited under O2 are likely fully oxidized, one would expect a beneficial contribution from the CoO for the harvesting of visible photons that would not be otherwise photoconverted by TiO2. Another phenomenon that can be involved in the significant increase of ε under solar illumination could be the formation of a heterojunction between NiO/ CoO nanoparticles and TiO2 nanotubes [19,55,56]. The photoelectrochemical water-splitting reaction relies on 3 steps: (i) absorption of photons leading to electron-hole pair formation, (ii) photogenerated carriers separation and (iii) interface oxydo-reduction reactions. Valence and conduction band of n-TiO2 have energy of 7.2 eV and 4.0 eV respectively, whereas NiO valence and conduction bands have a potential of 1.8 eV and 5.2 eV [57]. Such heterojunctions could form when nanoparticles are fully oxidized, and create an internal electric field due to the space charged layer formed after the equilibrium of Fermi level of both semi-conductor [58,59]. This heterojunction would allow a better charge separation, and therefore a lower recombination rate, leading to an increase of the overall photo-electrochemical activity. On the other hand, the enhancement under solar illumination could also be due to a catalytic effect of CoO and NiO NPs, which could lower the energy barrier of the OER and increase the kinetic of the reaction leading to an increase of the activity of these samples under solar illumination.

a core-shell structure where the metallic core is covered by an oxide (CoO and NiO) outer shell, when the PLD is carried out under vaccum or He. In contrast, the CoNi NPs are fully oxidized when deposited under O2 atmosphere. The EDX analyses have shown that the CoNi-NPs are not only present at the top of the TiO2-NTs but they also infiltrate the porous structure of the nanotubes. By performing cyclic voltammetry measurements under solar light (AM1.5 filter) and visible light (AM1.5 filterþ400 nm filter), we were able to point out an increase of more than 50% of the overall photoconversion efficiency of the TiO2-NTs when decorated with PLD-deposited CoNi-NPs under O2 background gas. This photoelectrochemical enhancement reaches its maximum for NLP  2000 pulses. This enhancement under solar illumination could arise from a catalytic effect of CoO/NiO NPs and/or from the formation of heterojunctions between CoO/NiO nanoparticles and TiO2 nanotubes. In sum, our results clearly show that the nature of the background gas used during the PLD process have a strong impact on the structure and chemical bonding of the CoNi-NPs. This, in turn, influences the catalytic properties of the CoNi-NPs and therefore the overall photoelectrochemical activity of the CoNi/TiO2 nanohybrid structures.

Acknowledgement The authors would like to acknowledge the financial support from the Natural Sciences and Engineering Research Council (NSERC) of Canada through the Discovery grants
bec program, and the FRQNT (Le Fonds de Recherche du Que Nature et Technologies) through its strategic Network
bec”. TF is grateful for the joint fellowship “Plasma-Que
gion Alsace, France” and INRS (Qc, Canfrom both the “Re ada). This project was partially funded through the ANR BAGETE (ANR-16-CE05-0001-01). Thierry Dintzer, the SEMCRO Platform, Driss Ihiawakrim, Vasiliki Papaefthymiou, €l Lavoie-Leblanc are gratefully Florian Gelb and Joe acknowledged for their help for the measurements and data treatments of SEM, TEM, XPS, CV and technical assistance on the PLD system, respectively.

Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijhydene.2019.08.179.

references

Conclusion The PLD technique was successfully used to decorate TiO2NTs with different loadings of CoNi-NPs synthesized under three different gas backgrounds (vacuum, He and O2). The structural and compositional results, derived from XPS and EDX analyses, not only confirmed that the CoNi-NPs have a stoichiometric composition (with 50:50 atomic ratio), but more interestingly revealed that the NPs essentially consist of

[1] Simpson A, Lutz A. Exergy analysis of hydrogen production via steam methane reforming. Int J Hydrogen Energy Dec. 2007;32(18):4811e20. [2] Chi J, Yu H. Water electrolysis based on renewable energy for hydrogen production. Chin J Catal Mar. 2018;39(3):390e4. [3] Ge M, et al. A review of TiO2 nanostructured catalysts for sustainable H2 generation. Int J Hydrog Energy Jan. 2017;42:8418e49.

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

international journal of hydrogen energy xxx (xxxx) xxx

[4] Ahmed M, Dincer I. A review on photoelectrochemical hydrogen production systems: challenges and future directions. Int J Hydrogen Energy Jan. 2019;44(5):2474e507. [5] Tsui L, Zangari G. Titania nanotubes by electrochemical anodization for solar energy conversion. J Electrochem Soc 2014;161(7):D3066e77. [6] Joy J, Mathew J, George SC. “Nanomaterials for photoelectrochemical water splitting e review. Int J Hydrogen Energy Mar. 2018;43(10):4804e17. [7] Ye M, Gong J, Lai Y, Lin C, Lin Z. High-efficiency photoelectrocatalytic hydrogen generation enabled by palladium quantum dots-sensitized TiO 2 nanotube Arrays. J Am Chem Soc Sep. 2012;134(38):15720e3. [8] Zhang L, Pan N, Lin S. Influence of Pt deposition on watersplitting hydrogen generation by highly-ordered TiO2 nanotube arrays. Int J Hydrogen Energy Aug. 2014;39(25):13474e80. [9] Ozkan S, Mazare A, Schmuki P. Extracting the limiting factors in photocurrent measurements on TiO2 nanotubes and enhancing the photoelectrochemical properties by Nb doping. Electrochim Acta Sep. 2015;176:819e26. [10] Vitiello RP, Macak JM, Ghicov A, Tsuchiya H, Dick LFP, Schmuki P. N-Doping of anodic TiO2 nanotubes using heat treatment in ammonia. Electrochem Commun Apr. 2006;8(4):544e8. [11] Liu S-H, Tang W-T, Lin W-X. Self-assembled ionic liquid synthesis of nitrogen-doped mesoporous TiO2 for visiblelight-responsive hydrogen production. Int J Hydrogen Energy Sep. 2017;42(38):24006e13. [12] Favet T, Ihiawakrim D, Keller V, Cottineau T. Anions and cations distribution in M 5þ/N 3- co-alloyed TiO 2 nanotubular structures for photo-electrochemical water splitting. Mater Sci Semicond Process Jan. 2018;73:22e9. [13] Delegan N, Pandiyan R, Komtchou S, Dirany A, Drogui P, El Khakani MA. In-situ co-doping of sputter-deposited TiO2:WN films for the development of photoanodes intended for visible-light electro-photocatalytic degradation of emerging pollutants. J Appl Phys May 2018;123(20):205101. [14] Myahkostupov M, Zamkov M, Castellano FN. Dye-sensitized photovoltaic properties of hydrothermally prepared TiO2 nanotubes. Energy Environ Sci 2011;4(3):998. [15] Macak JM, Schmuki P. Anodic growth of self-organized anodic TiO2 nanotubes in viscous electrolytes. Electrochim Acta Nov. 2006;52(3):1258e64. [16] Liu Y, Wang C, Lei Y, Liu F, Tian B, Wang J. Investigation of high-performance IrO2 electrocatalysts prepared by Adams method,. Int J Hydrogen Energy Oct. 2018;43(42):19460e7. [17] Audichon T, et al. “Electroactivity of RuO2eIrO2 mixed nanocatalysts toward the oxygen evolution reaction in a water electrolyzer supplied by a solar profile. Int J Hydrogen Energy Oct. 2014;39(30):16785e96. [18] Li X, et al. Controlled synthesis of MnO2@TiO2 hybrid nanotube arrays with enhanced oxygen evolution reaction performance. Int J Hydrogen Energy Aug. 2018;43(31):14369e78. [19] Deng X, et al. Fabrication of p-NiO/n-TiO 2 nano-tube arrays photoelectrode and its enhanced photocatalytic performance for degradation of 4-chlorphenol. Separ Purif Technol Oct. 2017;186:1e9. [20] Wang X, Zhang S, Peng B, Wang H, Yu H, Peng F. Enhancing the photocatalytic efficiency of TiO2 nanotube arrays for H2 production by using non-noble metal cobalt as co-catalyst. Mater Lett Feb. 2016;165:37e40. [21] Ka I, Gonfa B, Le Borgne V, Ma D, El Khakani MA. Pulsed laser ablation based synthesis of PbS-quantum dot-decorated onedimensional nanostructures and their direct integration into highly efficient nanohybrid heterojunction-based solar cells. Adv Funct Mater Jul. 2014;24(26):4042e50.

11

[22] Ka I, Le Borgne V, Ma D, El Khakani MA. Pulsed laser ablation based direct synthesis of single-wall carbon nanotube/PbS quantum dot nanohybrids exhibiting strong, spectrally wide and fast photoresponse. Adv Mater Dec. 2012;24(47):6289e94. [23] Anouar M, Jbilat R, Le Borgne V, Ma D, El Khakani MA. Ag nanoparticle-decorated single wall carbon nanotube films for photovoltaic applications. Mater Renew Sustain Energy Feb. 2016;5:1. UNSP 1. [24] Imbrogno A, Pandiyan R, Barberio M, Macario A, Bonanno A, El Khakani MA. Pulsed-laser-ablation based nanodecoration of multi-wall-carbon nanotubes by Co-Ni nanoparticles for dye-sensitized solar cell counter electrode applications. Mater Renew Sustain Energy May 2017;6(2):11. [25] Cottineau T, Cachet H, Keller V, Sutter EMM. Influence of the anatase/rutile ratio on the charge transport properties of TiO2-NTs arrays studied by dual wavelength optoelectrochemical impedance spectroscopy. Phys Chem Chem Phys Nov. 2017;19(46):31469e78. [26] Warang T, Patel N, Fernandes R, Bazzanella N, Miotello A. Co3O4 nanoparticles assembled coatings synthesized by different techniques for photo-degradation of methylene blue dye. Appl Catal B Environ Mar. 2013;132(133):204e11.  S, Babonneau F. XPS studies of SiO2 TiO2 [27] Ingo GM, Dire powders prepared by sol gel process. Appl Surf Sci 1993;70:230e4. [28] Raja KS, Gandhi T, Misra M. Effect of water content of ethylene glycol as electrolyte for synthesis of ordered titania nanotubes. Electrochem Commun May 2007;9(5):1069e76. [29] Jayakumar A, Antony RP, Zhao J, Lee J-M. MOF-derived nickel and cobalt metal nanoparticles in a N-doped coral shaped carbon matrix of coconut leaf sheath origin for high performance supercapacitors and OER catalysis. Electrochim Acta Mar. 2018;265:336e47. € zinger H. An X Ray photoelectron [30] Lange F, Schmelz H, Kno spectroscopy study of oxides of arsenic supported on TiO2. J Electron Spectrosc Relat Phenom 1991;57:307e15. [31] Wang Y, et al. Bimetallic Ni-M (M ¼ Co, Cu and Zn) supported on attapulgite as catalysts for hydrogen production from glycerol steam reforming. Appl Catal Gen Jan. 2018;550:214e27. [32] Espinos JP, Fernandez A, Gonzalez-Elipe AR. Oxidation and diffusion processes in nickel-titanium oxide systems. Surf Sci 1993;295(3):402e10. [33] Liu L, Zhao C, Zhao H, Zhang Q, Li Y. ZnO-CoO nanoparticles encapsulated in 3D porous carbon microspheres for highperformance lithium-ion battery anodes. Electrochim Acta Jul. 2014;135:224e31. [34] Liu T, et al. CoO nanoparticles embedded in threedimensional nitrogen/sulfur co-doped carbon nanofiber networks as a bifunctional catalyst for oxygen reduction/ evolution reactions. Carbon Sep. 2016;106:84e92. [35] Zhang M, Wang Y, Jia M. Three-Dimensional reduced graphene oxides hydrogel anchored with ultrafine CoO nanoparticles as anode for lithium ion batteries. Electrochim Acta May 2014;129:425e32. [36] Wang H, et al. Ni nanoparticles encapsulated in the channel of titanate nanotubes: efficient noble-metal-free catalysts for selective hydrogen generation from hydrous hydrazine. Chem Eng J Jan. 2018;332:637e46. [37] Liu M, et al. Encapsulation of NiO nanoparticles in mesoporous carbon nanospheres for advanced energy storage. Chem Eng J Jan. 2017;308:240e7. [38] Chen W-T, Chan A, Sun-Waterhouse D, Moriga T, Idriss H, Waterhouse GIN. “Ni/TiO2: a promising low-cost photocatalytic system for solar H2 production from ethanolewater mixtures. J Catal Jun. 2015;326:43e53. [39] Khan MA, Yang O-B. Photocatalytic water splitting for hydrogen production under visible light on Ir and Co ionized titania nanotube. Catal Today Aug. 2009;146(1e2):177e82.

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179

12

international journal of hydrogen energy xxx (xxxx) xxx

[40] Bazylewski P, et al. The characterization of Co-nanoparticles supported on graphene. RSC Adv 2015;5(92):75600e6. [41] Barrios CE, Albiter E, Gracia y Jimenez JM, Tiznado H, RomoHerrera J, Zanella R. “Photocatalytic hydrogen production over titania modified by gold e metal (palladium, nickel and cobalt) catalysts. Int J Hydrogen Energy Dec. 2016;41(48):23287e300. [42] Uhlenbrock S, Scharfschwerdt C, Neumann M, Illing G, Freund H-J. The influence of defects on the Ni 2p and O 1s XPS of NiO. J Phys Condens Matter 1992;4(40):7973. [43] Song L, Wang Y, Ma J, Zhang Q, Shen Z. Core/shell structured Zn/ZnO nanoparticles synthesized by gaseous laser ablation with enhanced photocatalysis efficiency. Appl Surf Sci Jun. 2018;442:101e5. [44] Patel N, Miotello A, Bello V. Pulsed Laser Deposition of Conanoparticles embedded on B-thin film: a very efficient catalyst produced in a single-step process. Appl Catal B Environ Mar. 2011;103(1):31e8. [45] Pandiyan R, Oulad Elhmaidi Z, Sekkat Z, Abd-lefdil M, El Khakani MA. Reconstructing the energy band electronic structure of pulsed laser deposited CZTS thin films intended for solar cell absorber applications. Appl Surf Sci Feb. 2017;396:1562e70. [46] Marien CBD, Cottineau T, Robert D, Drogui P. TiO2 Nanotube arrays: influence of tube length on the photocatalytic degradation of Paraquat. Appl Catal B Environ Oct. 2016;194:1e6. [47] Roy P, Berger S, Schmuki P. TiO2 nanotubes: synthesis and applications. Angew Chem Int Ed Mar. 2011;50(13):2904e39. [48] Peng S, Zeng X, Li Y. Titanate nanotube modified with different nickel precursors for enhanced Eosin Y-sensitized photocatalytic hydrogen evolution. Int J Hydrogen Energy May 2015;40(18):6038e49.

[49] Soundharrajan V, et al. Metal organic frameworkcombustion: a one-pot strategy to NiO nanoparticles with excellent anode properties for lithium ion batteries. J Energy Chem Jan. 2018;27(1):300e5. [50] Zhan X, et al. Efficient CoO nanowire array photocatalysts for H 2 generation. Appl Phys Lett Oct. 2014;105(15):153903. [51] Kukreja LM, Verma S, Pathrose DA, Rao BT. Pulsed laser deposition of plasmonic-metal nanostructures. J Phys Appl Phys Jan. 2014;47(3):34015. [52] Liu Y, Wang Z, Fan W, Geng Z, Feng L. Enhancement of the photocatalytic performance of Ni-loaded TiO2 photocatalyst under sunlight. Ceram Int Apr. 2014;40(3):3887e93. [53] Jiang YC, Wang JF, Gao J. Giant photoconductivity induced by plasmonic Co nanoparticles in Co-doped amorphous carbon/ silicon heterostructures. Carbon Jun. 2014;72:106e13. [54] Zhao X, Cai Z, Wang T, O'Reilly SE, Liu W, Zhao D. A new type of cobalt-deposited titanate nanotubes for enhanced photocatalytic degradation of phenanthrene. Appl Catal B Environ Jun. 2016;187:134e43. [55] Yu X, et al. “NiOeTiO2 pen heterostructured nanocables bridged by zero-bandgap rGO for highly efficient photocatalytic water splitting. Nano Energy Sep. 2015;16:207e17. [56] Zhang G, Huang H, Li W, Yu F, Wu H, Zhou L. Enhanced photocatalytic activity of CoO/TiO2 nanotube composite. Electrochim Acta Oct. 2012;81:117e22. [57] Ye M, et al. Recent advances in interfacial engineering of perovskite solar cells. J Phys Appl Phys Sep. 2017;50(37):373002. [58] Wang H, et al. Semiconductor heterojunction photocatalysts: design, construction, and photocatalytic performances. Chem Soc Rev Jul. 2014;43(15):5234e44. [59] Zhang L, Jaroniec M. Toward designing semiconductorsemiconductor heterojunctions for photocatalytic applications. Appl Surf Sci Feb. 2018;430:2e17.

Please cite this article as: Favet T et al., Enhanced visible-light-photoconversion efficiency of TiO2 nanotubes decorated by pulsed laser deposited CoNi nanoparticles, International Journal of Hydrogen Energy, https://doi.org/10.1016/j.ijhydene.2019.08.179