Size-controllable Ni5TiO7 nanowires as promising catalysts for CO oxidation

Ni5TiO7 nanowires with controllable sizes are synthesized using PEO method combined with impregnation and annealing at 1050oC in air, with adjustment of different concentrations of impregnating solution to control the dimension of nanowires. The resulting nanowires are characterized in details using X-ray diffraction, scanning electron microscopy, transmission electron microscopy and energy dispersive X-ray analysis. In addition, the CO catalytic oxidation performance of the Ni5TiO7 nanowires is investigated using a fixed-bed quartz tubular reactor and an on-line gas chromatography system, indicating that the activity of this catalytic system for CO oxidation is a strong dependency upon the nanocrystal size.When the size of the Ni5TiO7 nanowires is induced from 4 μm to 50 nm, the corresponding maximum conversion temperature is lowered by ~100 oC.

W, Co) and (CeO x , ZrO 2 )/TiO 2 /Ti coatings 15 , with its obvious advantages of an easy operation, low-cost and versatility in dimension controlling and composition tailoring 16 .
Ni 5 TiO 7 , a promising catalyst for tar conversion, was first synthesized with flux method by Fumio Shimura and Tsutomu Kawamura in 1976 17,18 . It possesses an orthorhombic symmetry with a unit cell lattice constants a = 9.20 ± 0.02 , Å b = 2.99 ± 0.01 Å and c = 12.17 ± 0.04 Å 17 . The structural formula of Ni 5 TiO 7 can be simply regarded as the combination of NiO and TiO 2 with a stoichiometric ratio of NiO:TiO 2 = 5:1 (or Ni:Ti:O = 5:1:7). However, the formation of Ni 5 TiO 7 crystal cannot be achieved by simply heating the NiO and TiO 2 powders at high temperature. Until very recently, Jiang and his co-workers in Siegen University (Germany) reported the fabrication of nanostructured Ni 5 TiO 7 catalyst using PEO method and demonstrated its outstanding performance in biomass gasification process 18 . However, no further work regarding to the size controlling of Ni 5 TiO 7 nanocatalyst, which is crucial for understanding its nucleation, crystallization and further catalytic performance improvement, was preceded. In addition, exploring its multi-functions in environment-related issues and long-term stability is still essential for considering its future practical applications in environment processing.
In this context, the size controlling of Ni 5 TiO 7 catalysts fabricated by PEO method is systemically investigated through tailoring the concentration of Ni ion in impregnation solutions. The morphology and size evolution of Ni 5 TiO 7 nanostructures as a dependence of Ni ions are also discussed. As a key part, the CO catalytic oxidation behaviour of the as-prepared Ni 5 TiO 7 nanowires used as catalysts is also evaluated for exploring their promising applications in CO conversion. In addition, the size influence of Ni 5 TiO 7 catalysts on their CO catalytic efficiency and the stability at high temperature are also studied. It is demonstrated that the nanoscaled Ni 5 TiO 7 catalysts with huge surface area and excellent crystallinity exhibit broad application potential and incomparable advantage in CO oxidation and biomass gasification, especially in car exhaust processing or the degradation of waste gas pollutions, for the sake of their high catalytic performance, low cost and easy preparation process.

Results and Discussion
X-ray diffraction (XRD) is first used to examine the crystalline structures and phase purities of as-synthesized samples obtained at PEO stage and subsequent annealing. Figure 1 shows the XRD patterns of PEO coating and its subsequent dipping into 0.5 M Ni(NO 3 ) 2 solution and annealing at 1050 o C . It is found that mixed anatase and rutile TiO 2 phases have been directly formed in PEO coating after the arc-discharging process in electrolyte. The peak from Ti substrate is not observed because of the excessive thickness of PEO coating (~10 μ m). In addition, a dome peak in the range of 20-40 o is also observed, which indicates the formation of some amorphous phase. After dipping the PEO coating in Ni(NO 3 ) 2 solution for 1 hour and heated at 1050 o C, Ni 5 TiO 7 phase with good crystallinity appears in the surface of the PEO coating and shows predominant diffraction intensity (Fig. 1), in good agreement with the standard structure data of Ni 5 TiO 7 crystal (No. JCPDS: 31-0927). Meanwhile, the preformed anatase TiO 2 phase has completely disappeared from the PEO coating and only rutile TiO 2 phase is maintained due to the metastable feature of anatase TiO 2 at a critical phase transition temperature in the range of 400-1050 o C. 23 The formation of Ni 5 TiO 7 phase implies that Ni ions incorporated into porous PEO surface/matrix have reacted with Ti/TiO 2 at high temperature. Figure 2 shows the morphologies and compositions of the PEO coating and the as-synthesized Ni 5 TiO 7 nanostructures annealed at 1050 o C. It's found that the PEO coating with a thickness of 10 μ m has a rough morphology with concaved pores uniformly distributing on the surface (Fig. 2a,b). The pores are formed during micro-arc discharging under high voltage and high surface temperature, which produces reduplicate melt and solidification reactions on Ti surface. Composition analysis performed on the PEO coating shows that the layer is mainly made of Ti and O elements (TiO 2 ) with Ni, P and W etc as minorities (Fig. 2c), indicating the formation of porous TiO 2 film. These Ni, P and W elements are directly from the electrolytes we used (see experimental section). The detection of Ni etc elements in the PEO coating suggests that the amorphous phase corresponding to the broad dome in XRD pattern ( Fig. 1) may be assigned to some Ni-containing compound. To obtain Ni 5 TiO 7 nanocrystals, the porous PEO coating dipped into Ni(NO 3 ) 2 solution is annealed at 1050 o C for 1 hour, and it is found that needle-like nanowires with diameters ranging from 100 nm to 1 μ m are densely covered on the PEO coating surface (Fig. 2d). Cross-section SEM observation presented in Fig. 2e clearly shows that these nanowires are directly grown from the top layer of the porous coating and are tightly combined with the coating. In contrast to the predominant Ti and O constituents in initial PEO coating, the chemical compositions detected from the local nanowires are identified as Ni, Ti and O with a stoichiometric ratio close to Ni 5 TiO 7 (Fig. 2f), further demonstrating the formation of Ni 5 TiO 7 phase after high-temperature annealing. Figure 3a shows a representative TEM image of Ni 5 TiO 7 nanowires transferred from the porous TiO 2 coating. The smoothness of nanowire surface and the uniformity of nanowire diameter in the range of 200-500 nm can be clearly confirmed. No particles in nano-scaled size can be found at the nanowire surface and tip-end. Careful EDS analyses performed on different positions of the nanowire conclude that the nanowires are indeed made of Ni, Ti and O with an atomic ratio of Ni:Ti:O = 5:1:7 ( Fig. 3b and Fig. S1), matching well with the composition value of standard Ni 5 TiO 7 crystal. The Cu peak with negligible signal is directly from the Cu TEM grid for supporting the samples. Absence of any other impurity peak related to P and W elements that are introduced during PEO process demonstrates again the high phase and chemical purity of Ni 5 TiO 7 nanowires. The high-temperature annealing of PEO coating at 1050 °C in ambient environment only leads to the nucleation and crystallization of Ni 5 TiO 7 phase and does not introduce any possible contamination. The superior crystallinity of Ni 5 TiO 7 nanowires, proved by the succinct selective area electron diffraction (SAED) pattern (Fig. 3c) and HRTEM data (Fig. 3d), verify the single crystalline nature of Ni 5 TiO 7 nanowires. The separated and periodic diffraction spots and well-aligned atom ordering confirm that Ni 5 TiO 7 nanowires prepared by PEO method are free of structural defects like microtwins and stacking faults, indicating the obvious advance of this technique in the crystallinity controlling in comparison with other methods for metal oxide nanostructure preparation 19 . The lattice distance between two adjacent lattice planes parallel and perpendicular to the growth direction are measured to be 0.44 nm and 0.29 nm, respectively, matching well with the d-spacing value of (200) and (010) planes of orthorhombic Ni 5 TiO 7 crystal. The growth orientation of Ni 5 TiO 7 nanowires, along [010] direction, can also be confirmed as a result of a strong growth competition of the (010) surface against other low-indexed planes of the side facets. In addition, an outer layer with a thickness of few nanometers can be found on the surface of the Ni 5 TiO 7 nanowire.
In the synthesis of Ni 5 TiO 7 nanowires, it is found that Ni ions in the impregnation solution play a key role in organizing the initial nucleation locations and are essential for the nucleation (or growth) density  Figure 4 shows the typical morphology of Ni 5 TiO 7 nanowires evolved from as-prepared PEO coating and pre-treated samples annealed at 1050 o C for 1h. Notably, a significant difference in the morphology and size of Ni 5 TiO 7 nanowires can be distinguished from these four samples. When the PEO coating is annealed without dipping into Ni(NO 3 ) 2 solution, only a small fraction of nanowires and nanoparticles with strip shapes and huge size can be found on its surface, as shown in Fig. 4a. XRD and EDS results (Supporting Information Figure S2) clearly demonstrate that the nanowires are Ni 5 TiO 7 , suggesting that Ni ions in the electrolyte has been absorbed/stored in the PEO coating during arc-discharging process and further reacted with Ti/TiO 2 to lead to the formation of Ni 5 TiO 7 phase. Figure 4b-d show the SEM images of Ni 5 TiO 7 nanowires corresponding to Ni(NO 3 ) 2 solution concentrations of 0.1M, 0.5 M and 2M, respectively and annealed at 1050 o C for 1 hour. It can be found that the dipping process has a significant effect on the morphology and size evolutions of the Ni 5 TiO 7 nanowires. Apparently, impregnating the PEO coating with 0.1M Ni(NO 3 ) 2 solution results in the drastic size reduction of Ni 5 TiO 7 nanowires from 2-5 μ m for as-prepared one to an average diameter of 50 nm (Fig. 4b). Inset of Fig. 4b gives a clear morphology and dimension size of Ni 5 TiO 7 nanowires protruding from the PEO coating surface. When the Ni concentration is increased to 0.5 M, both the diameter and length of Ni 5 TiO 7 nanowires have evolved to ~300 nm and ~10 μ m correspondingly, as shown in Fig. 4c. Further increase of the Ni concentration to 2M directly induces the fast growth of Ni 5 TiO 7 nanowires to an average diameter of 4 μ m (Fig. 4d). All these SEM observations clearly demonstrate that the dipping process is of great significance in tailoring the morphology and general size of Ni 5 TiO 7 nanowires. Consequently, the morphology and size of Ni 5 TiO 7 nanowires can be correspondingly modified by controlling the concentrations of Ni(NO 3 ) 2 solution. It is reasonable to understand the morphology and size evolution of Ni 5 TiO 7 nanowires as a dependence of Ni(NO 3 ) 2 solution concentrations. In the case of lower content of Ni(NO 3 ) 2 solution, a thin layer of Ni(NO 3 ) 2 or Ni will cover on the PEO coating surface and aggregate into small islands under high-temperature annealing. These tiny Ni-containing islands with higher surface energy are preferential nucleation sites for the initial nucleation of Ni 5 TiO 7 phase and serve as the seeds for controlling their subsequent crystallization. It should be strengthened that the initial size of these Ni-containing nanoparticles is of crucial importance in dominating the size of Ni 5 TiO 7 nanowires. With the concentration increase of the Ni(NO 3 ) 2 solution, more Ni(NO 3 ) 2 or Ni will be deposited on the PEO coating surface and will give rise to the corresponding thickness enhancement of Ni-containing film. As a result, the nanoislands evolved from the thin film will correspondingly possess a large size for guiding the nucleation and the growth of Ni 5 TiO 7 nanowires. The same phenomenon and process can also be expected in the case of Based on detailed SEM observations and TEM characterizations, a tentative reaction mechanism describing the formations of the PEO oxide coating and the Ni 5 TiO 7 nanowires is proposed 20,21 . At the first stage of PEO process, the titanium will be charged to Ti ions and its surrounding aqueous solution (mainly H 2 O) will be decomposed into H + and OHunder high voltage and temperature, as described below 22 : Following this step, the Ti 4+ will react with OHto lead to the formation of anatase and rutile TiO 2 phases under a huge temperature gradient generated by micro-arc discharge and aqueous cooling. The formation reaction of TiO 2 can be described as follows 23 : At the third stage of PEO process, Ni(CH 3 COO) 2 in the electrolyte can be ionized to Ni 2+ and (CH 3 COO)by the energetic arc discharge. The Ni 2+ ions will react with other ions such as Ti 4+ and OHunder high temperature and high voltage, which leads to the formation of Ni-containing phase (like amorphous Ni 5 TiO 7 ) in the PEO coating, similar to the formation of CaTiO 3 produced by the PEO method 24,25 . It should be noted that the assertion of the formation of amorphous Ni-containing phase can be deduced from the dome peak in the XRD pattern (Fig. 1), which results from the repeatedly heating and cooling processes under high surface temperature produced by the micro-arc discharge in the electrolytic solution 26 . It has been reported that the momentary temperature in the local discharge zones can be up to 4000~8000 K in a short time 26 . When the PEO coating is subject to a high-temperature annealing at 1050 o C, the Ni-contained amorphous phase will crystallize and the other excessive Ni ions involved in the porous coating layer will diffuse into the top layer to provide Ni precursor and participate in the reaction with Ti/TiO 2 and oxygen. Finally, Ni 5 TiO 7 nanostructures with high density can be achieved.
Previous study has demonstrated the outstanding performance of Ni 5 TiO 7 catalysts in tar conversion with naphthalene as the target sample 18 . For considering its potential application in car exhaust processing, the CO oxidation over the as-prepared Ni 5 TiO 7 nanowires is also evaluated. In order to characterize the "real" catalytic performance of the catalysts, Ni 5 TiO 7 nanowires attached on TiO 2 coatings are peeled off from the Ti substrate. On the other hand, the Ni 5 TiO 7 nanowires in the areas with uniform morphology and high growth density are used for the CO evaluation test in order to depress the influence of porous TiO 2 coating to the greatest extent. Figure 5a summarizes the CO conversion rate as a dependence of reaction temperature for all Ni 5 TiO 7 /TiO 2 composite catalysts prepared under different concentrations of Ni(NO 3 ) 2 solutions. It is noticed that the Ni 5 TiO 7 nanowires corresponding to a 0.1M Ni(NO 3 ) 2 solution exhibit the best CO conversion capability in comparison with the other samples. Considering the morphology or growth density dependence on the concentration of Ni(NO 3 ) 2 solution, it is found that the Ni 5 TiO 7 nanowires corresponding to 0.1M Ni(NO 3 ) 2 solution have the smallest nanowire size and largest growth density than the other samples, indicating that the CO conversion efficiency of Ni 5 TiO 7 nanowires depends strongly on their size and growth density (Fig. 4). From Fig. 5a, it can be seen that the start temperature for CO conversion for Ni 5 TiO 7 nanowires in the case of 0.1M Ni(NO 3 ) 2 solution is around 230 o C. This critical conversion temperature is still higher than that of metal or metal oxide like NiO/Au 27 , Pt/CeO 2 2 , but it should be kept in mind that the catalytic reaction of Ni 5 TiO 7 nanowires is performed without any assistance of highly efficient noble metal catalysts. When the reaction temperature is increased to 440 o C, the CO oxidation capability is enhanced sharply and the catalytic activity approaches to the maximum. Almost all the CO has been completely converted into CO 2 . For Ni 5 TiO 7 nanowires corresponding to a Ni(NO 3 ) 2 concentration of 0.5 M, they exhibit the second best catalytic performance in all the four samples, which shows a starting reaction temperature at 300 o C and an end conversion temperature at 550 o C. However, the Ni 5 TiO 7 nanowires corresponding to 2M Ni(NO 3 ) 2 solution show quite similar/comparable catalytic performance to the sample without any dipping. The significant difference in catalytic performance undoubtedly comes from the discrepancy of Ni 5 TiO 7 nanowires in size and growth density, which provides different active sites for CO catalytic reaction. From above results, it can be concluded that the catalytic properties of Ni 5 TiO 7 nanowires have close relation to their morphology and surface. SEM observation shown in Fig. 4 has demonstrated that increasing the concentrations of Ni(NO 3 ) 2 solutions can lead to their size increase correspondingly, directly resulting in the weakening of the CO conversion rate and performance. Like most metal oxide catalysts, the CO catalytic oxidation of Ni 5 TiO 7 nanowires is closely related to nanowire size, the exposed surface area, mean diameter and the contact structure between the catalyst and the support metal oxides 28 . It is believed that the desorption of CO and O 2 on the surface of Ni 5 TiO 7 nanowires will react each other on the effect of Ni-O bonds and results in the catalytic oxidation 29 . In addition, the evaluation of the long-term stability of Ni 5 TiO 7 nanowires is also carried out. Figure 5b shows the continuous CO conversion test curve of Ni 5 TiO 7 nanowires, which is annealed at 1050 o C and corresponding to a 0.1M Ni(NO 3 ) 2 solution. The long-term stability test is carried out at 440 o C since it shows the best activity in CO oxidation (Fig. 5a). Notably, the Ni 5 TiO 7 nanowires exhibit extremely stable catalytic performance and no degradation of CO conversion efficiency is observed in the 20-hour test, further demonstrating the excellent high-temperature adaptation ability and thus it can be compared with some metal oxide catalysts like Co 3 O 4 and NiCo 2 O 4 30,31 .
To comprehensively evaluate the application potential of Ni 5 TiO 7 nanowire catalysts, the structural stability of Ni 5 TiO 7 nanowires after several cycle high-temperature catalysis tests is investigated. The Ni 5 TiO 7 nanowires continuously working at 440 o C for 20 hours are used as the samples for structure and crystallinity evaluations using TEM, as shown in Fig. 6. It is found that the morphology and appearance of the nanowires are not destroyed after the catalytic reaction (Fig. 6a). The nanowires still maintain their initial surface smoothness as the one shown in Fig. 3a. Further TEM observation on high-magnifications reveals that the nanowires still show superior crystallinity and the ordering of atoms is not deteriorated (Fig. 6d). Especially, the outer surface of the Ni 5 TiO 7 nanowires still keeps the same feature as the one before catalytic reaction. Absence of any nanoparticles on the nanowire surface implies that the nanowires are not decomposed or involved into any phase transition reactions. Additionally, we do not observe any clues that verify the formation of polycrystalline or amorphous phases in the Fast Fourier Transition (FFT) (Fig. 6c). Composition analysis using high-resolution EDS confirms the same constitutions as the as-synthesized nanowires before catalytic reaction and impurity peaks from any possible phases are not found (Fig. 6b), further demonstrating the chemical stability of our nanowires involved in the CO oxidation. Apart from these merits, it should be strengthened that the Ni 5 TiO 7 nanowires are in-situ fabricated on porous TiO 2 surface with improved substrate adherence. This peculiar advantage will in turn promote their direct fabrications on honeycomb-like metal substrate using PEO method and accelerate their practical utilization. With its predominant advantages in low-cost and easy synthesis process, as well as the superior performance in CO conversion and biomass gasification, the Ni 5 TiO 7 nanowires will certainly open up more space as an efficient and chemically/structurally stable catalyst for environment-friendly applications ranging from car exhaust processing to chemical pollutant cleaning.

Conclusion
In summary, Ni 5 TiO 7 nanowires with single crystal feature have been synthesized using conventional PEO method with subsequent impregnation and annealing processes. The size of Ni 5 TiO 7 nanowires can be selectively tailored through simply controlling the Ni(NO 3 ) 2 concentration in dipping solution and an average minimized diameter as small as 50 nm can be obtained. CO oxidation tests demonstrates that the catalytic performance of Ni 5 TiO 7 nanowires is strongly dependent on their size and growth density and the Ni 5 TiO 7 nanowires with an average diameter of 50 nm show the best CO conversion efficiency. Long-term catalytic evaluation with a total reaction time of 20 hours and the precise HRTEM analysis verify the superior stability in catalytic performance and microstructure, which allows the current Ni 5 TiO 7 nanowires as promising catalysts for diverse applications ranging from biomass gasification to car exhaust process.

Experimental Section
Samples preparation. Commercially available pure grade I titanium sheets (10mm × 10mm × 1mm, Shanxi Baotai group) and graphite were connected to the positive and negative poles of power supply as anode and cathode, respectively. Titanium sheets are also used as the substrates to form porous TiO 2 coatings and the Ti precursor. Before the PEO process, the Ti substrates are ground using #1000 sand paper to remove the contaminated surface (TiO 2 or other impurity) and are cleaned with acetone, ethyl alcohol and distilled water. To promote the PEO process, mixed solutions comprising of sodium phosphate (Na 3  , WO 4 2and B 4 O 7 2ions can promote a uniform micro-arc generation, which is essential for producing homogeneous PEO coating 32,33 . A detailed experimental setup for the PEO reaction can be found in Figure S3. The PEO process was carried out in a fixed applied current density mode of 0.1 A cm −2 . Other parameters including PEO time and pulse frequency were set at 10 min and 1000 Hz, respectively, for all samples after several rounds optimization. The electrolyte temperature was controlled under 50 o C for a full PEO reaction through circulating water cooling system during the PEO process. After the completion of the porous PEO coating, it was dipped into Ni(NO 3 ) 2 solutions with different concentrations for 1 hour for improving the Ni ion concentrations in the PEO coating. Following this step, the PEO coating were transferred to a resistance furnace for thermal annealing to promote the nucleation and crystallization of Ni 5 TiO 7 catalysts. After maintaining the reaction at 1050 o C for 1 hour, high density of Ni 5 TiO 7 nanostructures featured with a wire-like morphology on the surface of PEO coating were obtained.
Structure and composition characterization. Phase identification of the PEO coating and Ni 5 TiO 7 nanostructures was examined by an X-ray diffractions (XRD, Rigaku D/max 2400) using Cuk α (λ kα = 0.154056 nm) as the x-ray source over a 2θ angel of 20-60 o . The morphology of the coatings and Ni 5 TiO 7 nanostructures are characterized by a field-emission scanning electron microscopy (FE-SEM, FEI Inspect F50) equipped with a Quanta 600 Energy Dispersed X-ray spectrometer (EDS) system. The microstructure, crystallinity and elemental compositions of Ni 5 TiO 7 nanowires scratched from the coating surfaces were characterized by a transmission electron microscopy (TEM, Tecnai, F20) under an accelerated voltage of 200 kV.
Catalytic tests. All the catalytic activity evaluations of Ni 5 TiO 7 /TiO 2 composite catalysts for CO oxidation were carried out by using a fixed-bed quartz tubular reactor. The catalysts peeled off from the Ti substrate were placed in the active zone of the quartz tubular reactor. Mixed gases (1.0% CO, 10% O 2 , balanced with helium) are transported into the reactor at a flowing rate of 10 ml/min. The reactants and products were analyzed by using an on-line gas chromatography system (Agilent 7890A) with a molecular sieve column. CO and CO 2 gases were detected with a thermal conductivity detector. All the catalytic tests were carried out at the temperatures ranging from 25 o C to 650 o C.