Finding a suitable alternative of fossil fuel resources as an energy source is a main concern in the 21st century as the termination of fossil fuels resources, increasing environmental pollution and energy shortages [1-3]. In this context, hydrogen is considered to be one of the proper alternative candidates for future energy because of its high energy density and copious sources [4,5]. Presently, the most hydrogen is prepared by using natural gas steam reforming which consumes energy and creates large greenhouse gas emissions, mainly, carbon dioxide [6,7]. Therefore, expanding efforts have been recently done to search for appropriate hydrogen storage materials. In recent years, sodium borohydride (NaBH4) has been taken into account as an attractive candidate for solid hydrogen storage materials owing to its high hydrogen content (10.8 wt% capacity), nontoxicity, non-flammable properties, pure hydrogen production and good stability in alkaline solutions [8,9]. Also, the byproduct corresponding to the sodium borohydride hydrolysis , i.e. sodium metaborate (NaBO2), is eco-friendly and can be recycled as the raw material for regeneration to NaBH4 . Sodium borohydride can be classified as both a complex metal hydride and a chemical hydride as it can generate hydrogen in two manners: thermolysis, where the stored hydrogen is set free by heating, and hydrolysis, where the stored hydrogen is released through reaction with water [11,12]. Since the thermolysis process requires high temperature, the sodium borohydride hydrolysis has been preferentially used to generate hydrogen . Notwithstanding H2 can be generated from sodium borohydride, the low hydrogen productivity of NaBH4 at room temperature is regarded as a key problem to its commercialization, resulting from the low solubility of NaBH4 and NaBO2 byproduct in highly basic solution . Hence to overcome this problem, suitable catalysts are needed to accelerate the NaBH4 hydrolysis for H2 generation in a more efficient way as follows :
To date, noble metal-based catalysts (Au, Pt, Ru and Rh) are observed to be the most effective catalysts for hydrolyzing NaBH4 [15-17]. However, the high cost of limited noble metal resources makes practical application undefendable . Some low cost non-noble metals, such as Cu, Ni, Fe and Co, have been examined widely towards NaBH4 hydrolysis because of their excellent physical and chemical properties [19-21]. However, the main drawback of the use of such non-noble metals is their agglomeration because of high surface energy involving in the exothermic reduction procedure, leads to a significant loss of catalytic activity . Furthermore, the strong adsorption of resulted NaBO2 byproducts on the catalytic surface leads to a dramatic loss of activity only after a few operative cycles . A facile strategy to solve these problems is to find an appropriate supporting metal catalyst [16,23]. A suitable supporting material not only buttresses the active materials but also provides a large specific surface area along with the chance of better dispersion of the active phases . Amongst diverse materials, TiO2 is extensively applied as an alternative supporting material because of its unique chemical and physical features, exceptional stability, low price, nontoxicity and abundance [25,26]. Also, it is worth noting that there are two kinds of acid sites on the surfaces of TiO2: Brønsted-acid sites (surface-bound hydroxyl group i.e. Ti-OH) and Lewis-acid sites (exposed Tin+ cation) [27,28]. Since NaBH4 hydrolysis can be catalyze by an acid [29,30], so TiO2 can employ as an ideal supported metal catalyst for this reaction. Based on researchers’ studies, two key parameters are found to significantly affect catalytic performance during NaBH4 hydrolysis. These parameters are the surface area and adsorption capacity [21,31]. In comparison with different TiO2 nanostructures, TiO2 nanotubes encompass many benefits such as single-step synthesis, providing great specific surface area, and facilitate the transport of reagents to reactive sites [17, 32]. To date, there are various methods for fabrication of TiO2 nanotubes with diverse arrangements. Generally, disordered filaments of TiO2 nanotubes are synthesized by hydrothermal method while extremely ordered TiO2 nanotube arrays are synthesized through the anodization process . In addition, oriented free-standing TiO2 nanotube arrays can achieve by separating the aligned anodic TiO2 nanotubes from the metallic Ti substrate . Based on our earlier study, catalytic activities of TiO2 nanotubes are affected not only by their structure but also strongly by their morphology and arrangement . According to our knowledge, however, no systematic study has been reported until now, that would undertake efforts to find the influence of different TiO2 nanotubes arrangement as a supporting catalyst on NaBH4 hydrolysis. In addition, it still lacks a comprehensive study which compares the catalytic behavior of disordered filaments of TiO2 nanotubes with that of free-standing TiO2 nanotube arrays especially in generating hydrogen from NaBH4.
Therefore in the present work, the catalytic activities of different TiO2 nanotubes arrangements were investigated systematically for the first time. As the main novelty of this work, the catalytic activity of disordered filaments and ordered free-standing TiO2 nanotubes in hydrogen generation from NaBH4 hydrolysis were examined and compared. Moreover, as an original contribution of this work, the kinetic study of the catalytic hydrolysis of NaBH4 over free-standing TiO2 nanotubes was perused. In addition, in the current study, tuning the arrangement of TiO2 nanotubes by chemical-assisted two-step anodization process was employed as a simple and facile route to enhance the hydrogen generation rate from catalytic hydrolysis of NaBH4.
MATERIALS AND METHODS
Preparation disordered filaments of TiO2 nanotubes
Disordered filaments of TiO2 nanotubes (DF-NTs) were prepared through an alkaline hydrothermal treatment. First, 2 g of commercial TiO2 powder (P25, Evonik, Germany) was poured in 100 ml of concentrated caustic NaOH solution (10 M) while stirring for 30 min. subsequently, the mixture was ultrasonicated for 30 min at ambient temperature to become homogenizer. Following the homogenized suspension was transferred to a Teflon-lined autoclave. The obtained solution was then heated at 150 °C under static condition for 48 h. The resulting white precipitates were rinsed with 0.1 M HCl solution and distillated water repeatedly until the solution pH entered near neutral pH. The final output was centrifuged, dried at 80 °C, and annealed for 2 h at 450 °C.
Fabrication free-standing TiO2 nanotube arrays
Free-standing TiO2 nanotube arrays (FS-NTs) were synthesized through chemical-assisted two-step anodization procedure in a two-electrode configuration with the Ti sheet as both anode and cathode. First, Ti sheet was degreased and cleaned by immersion in acetone, ethanol and deionized water for 10 min during each step. After drying, to obtain a completely smooth and sleek surface, the chemical polishing process was taken place in a solution containing nitric acid 65% and hydrofluoric acid 40% with a volume ratio of 3:1 for 2 min. Ethylene glycol with 0.5 wt.% ammonium fluoride and 2 vol% deionized water was used as the electrolyte. The first-step anodization started with the growth of TiO2 nanotubes on Ti sheet with a constant voltage of 50 V for 1.5 hours at room temperature. The as-prepared TiO2 nanotubes were then removed by acid treating process via diluted hydrofluoric acid to expose the substrate. Subsequently, the etched Ti substrate was anodized for the second time in the same condition as the first step for 12 h. After completing the second anodization, the voltage was gradually decreased manually from adjusted potential to 10 V, over several seconds. Subsequently, the fabricated TiO2 nanotubes layer was separated using ultrasonic treatment in methanol solution to achieve free-standing TiO2 nanotubes. Afterward, to obtain a crystalline structure, the annealing of free-standing TiO2 nanotubes was done for 2 h at 450 °C.
The fabricated catalysts were analyzed using an X-ray diffractometer (XRD, Philips X’Pert Pro X-ray diffractometer, Germany) to obtain crystallographic information. The XRD patterns were recorded by means of Cu Kα radiation (the wavelength is 1.54 Å) in the 2q range of 10-80. The surface morphologies, structural characteristics and elements present in the as-prepared catalysts were examined using a field-emission scanning electron microscope (FESEM, TESCAN MIRA3-XMU, Czech Republic) with an energy dispersive spectrometry (EDS) detector. To determine the functional groups available in the catalysts, the Fourier transform infrared analysis (FTIR; Bruker alpha-2000, USA) was performed over a wavenumber range of 500-4000 cm-1. The BET analysis of the catalysts was studied by means of a micrometer ASAP 2000 (USA). The BET (Brunauer-Emmet-Teller) method measures the specific surface area and the pore volume. Furthermore, the average pore diameter was estimated from adsorption isotherms by Barrette Joynere Halenda (BJH) analysis. The principle corresponding this technique is on the basis of multiple points of nitrogen gas isothermal adsorptions-desorption.
The catalytic behaviors of the as-synthesized catalysts were investigated through hydrolyzing sodium borohydride solution. Fig. 1 indicates the typical experimental setup of producing hydrogen. The detailed procedure is as following: 10 mL NaBH4 solution (containing 5 wt% NaBH4 and 1.5 wt% NaOH) was poured into a sealed flask and placed in the thermostatic water-bath maintained at 30 °C. A water bath was utilized to keep the system under isothermal condition (30 °C). Then 10 mg of as-prepared catalysts were poured in a sealed reaction flask. The reaction proceeded at a stirring rate of 800 rpm and the concentration of hydrogen generated with time was measured immediately. The volume of hydrogen generated was determined by water displacement method, where the hydrogen volume was equal to the displaced water whose weight was recorded by a balance. The rate of hydrogen generation and the factors affecting the the catalyst activity were examined. At the end of the reaction, the catalyst was retained through filtration, washed thoroughly and dried at 60 °C for 10 h. Then the reusability experiment was repeated for 5 cycles.
RESULTS AND DISCUSSION
Fig. 2 represents the FESEM photographs of commercial TiO2 powder, DF-NTs and FS-NTs, respectively. As presented in Fig. 2(a), TiO2 nanopowder has a non-uniform granular shape. After hydrothermal treatment, the DF-NTs are formed with tubular morphology and a layered shape. The DF-NTs have a diameter of 5-10 nm and length of several hundred nanometers, as presented in Fig. 2(b) and 2(c). Fig. 2(d) and 2(e) shows a top-view and barrier layer images corresponding to FS-NTs, respectively. It can be observed that the FS-NTs are well open at the top end and closed at the bottom. Based on the top view image in Fig. 2(d), the FS-NTs have an average internal diameter of about 75 nm and wall thickness of 15 nm, respectively. The cross-sectional FESEM image of the FS-NTs, as presented in Fig 2(f), shows the formation of tubular arrays with ordered and aligned structures through the two-step anodization procedure. Based on Fig. 2(f), FS-NTs have an average length about 8 mm. The elemental relative compositions of DF-NTs and FS-NTs acquired from EDS analysis are illustrated in Fig. 2 (g) and 2(h). Fig. 2(g) and 2(h) clearly show that the peaks corresponding to Ti and O are the prominent peaks in both spectra. Based on Fig. 2(g), it can be found that DF-NTs mainly consisted of Ti and O atoms which the weight contribution of titanium and oxygen was 51.27% and 41.52%, respectively. Both the elements together contributed 92.79% of the total weight. The very small remaining percentage in DF-NTs catalyst is ascribed to the existence of residual Na+ ions that not be replaced by the H+ ions provided from the acid treatment process. Based on Fig. 2(h), FS-NTs mainly consisted of Ti (42.92 wt%) and O (41.57 wt%) with some impurity including C and F. The existence of residual contaminants like carbon and fluorine is ascribed to the precursors, i.e. ethylene glycol and ammonium fluoride, respectively.
The XRD patterns corresponding to the annealed commercial TiO2 powder, DF-NTs and FS-NTs are shown in Fig. 3, respectively. All XRD patterns confirm the crystalline nature of the as-prepared catalysts and also indicate the existence of anatase phase (JCPDS-21-1272), as expected for TiO2 nanotubes annealed at the 450 °C. Regarding Fig. 3, the diffraction peaks which were observed at 25.3º, 37.9°, 48.1°, 53.8°, 55.2°, 62.2°, 68.8°,70.02° and 75. 09° realize nanocrystalline of the anatase phase that can be indexed to the (101), (004), (200), (105), (211), (213), (116), (220) and (215) planes, respectively . The average crystallite size of the fabricated samples was calculated based on the line broadening of the XRD peak observed around 25.3°, i.e. the (101) plane, using the Scherer’s equation [37-40]:
where D is the average crystallite size, k is a constant (=0.89, here), λ is the wavelength of the X-ray radiation source (=1.5406 Å), q is the Bragg’s angle, and β is full width of the diffraction peak at half its maximum intensity in radian. Furthermore, the lattice parameters were estimated using the following equations [41,42]:
where h, k and l are the crystal plane indices, d(hkl) is the distance between the (hkl) crystal planes, θ is the diffraction angle of the (hkl) crystal plane, and a, b, and c are lattice parameters (anatase crystal: a = b ≠c). Also, volume of the crystal cells (tetragonal) was calculated through Equation (5).
The calculated crystal sizes, lattice parameters and cell volumes of the fabricated catalysts are presented in Table 1. Regarding Table 1, the values of the lattice parameters are in line with the anatase structure of TiO2 (JCPDS card 21-1272). Also, the crystallite size and the cell volume of all samples are similar which their values are in the range of 20 nm and 134 Å3, respectively. These results could infer that there is no impurity in the TiO2 crystalline lattices.
The FTIR was used to determine the functional groups that present in synthesized catalysts as presented in Fig. 4. The peaks found at about 3420 cm-1 and 2962 cm-1 are assigned to the Ti-OH bond . For all the catalysts, the spectra indicate a roughly strong band at about 1635 cm-1 to 1622 cm-1 which are because of the OH bending vibration corresponding to chemisorbed and/or physisorbed water molecule on the catalysts surface . The broad absorption peaks from 800 cm-1 to 700 cm-1 and 490 cm-1 to 433 cm-1 are ascribed to the asymmetric stretching, symmetric stretching, and the bending modes of Ti-O-Ti bond, respectively . For all catalysts, a small band observed near 3780 cm-1 is related to O-H stretching corresponding to the hydroxyl group to Ti atoms . Also, FTIR spectra are represented several types of vibrations at 1100 cm-1 to 1360 that they are assigned to the Ti-O stretching and Ti-OH bond, respectively . The low-intensity broad band at around the 840 cm-1 for DF-NTs sample is indicative of the Na-O bending of the Na-O-Ti bond of sodium titanates . These results confirmed that P25, DF-NTs and FS-NTs are mainly consisted of Ti and O species which is well in agree with EDS results.
Specific surface areas, pore volume and average pore diameter of the catalysts were determined using the BET process analysis under N2 adsorptive gas, as shown in Fig. 5. Table 2 presents the BET surface area and the corresponding pore volume for catalysts. Based on Fig. 5 (a) and 5 (b), the DF-NTs catalyst presented type IV isotherms with H3 type hysteresis loop at the relative pressure (P/P0) range of 0.4-1, based on the IUPAC classification , with the mean pore diameter of 9.54 nm. This result reflects the slit-like mesoporous geometry of DF-NTs catalyst. The slit-like pore geometry formed by DF-NTs is due to their aggregation and filament arrangement, which is expected for mesoporous materials . According to Fig. 5 (c) and 5 (d) The as-synthesized FS-NTs catalyst showed a type IV isotherm with H1 hysteresis loop in the relative pressure (P/P0) range of 0.4–1 and the average pore diameter was 11.75 nm, demonstrating the characteristics corresponding to mesoporous compounds. The H1 hysteresis loop of FS-NTs catalyst suggests a narrow range of uniform mesopores in the cylindrical pores form, which is in excellent accordance with the oriented tubular structures of FS-NTs. According to Table 2, both FS-NTs and DF-NTs have the highest surface area and pore volume than that commercial TiO2 powder. Furthermore, the BET surface area (112.77 m2/g) and the related pore volume (0.34 cm3/g) of FS-NTs is considerably largest comparing to the surface area (86.54 m2/g) and pore volume (0.25 cm3/g) of DF-NTs. It is worth notable that in catalytic applications, the catalyst surface area is essential to provide enough active places for the reagents to react. In the current study, since the FS-NTs catalysts have the highest surface area comparing to that two other catalysts, so we expect the FS-NTs catalysts show the higher catalytic behavior in generating hydrogen from the NaBH4 solution.
Investigation catalysts activity and hydrolysis mechanism
The catalytic behavior of three different nanostructures, namely commercial TiO2 powder, DF-NTs and FS-NTs, were investigated through measurement of the volume of hydrogen generated through hydrolyzing an alkaline NaBH4 solution. The results of these experiments are presented in Fig. 6. After adding the catalysts, the hydrogen was released immediately without any induction period. The rates of hydrogen generation (ml.min-1) were calculated from the fitting lines slopes that were achieved from the Volume-time curves. Fig. 6, showed that TiO2 in three different nanostructures itself has catalytic activity for producing hydrogen. Previous studies indicated that there are two kinds of acid sites on the surface of TiO2 nanotubes, namely Brønsted-acid sites (surface-bound hydroxyl group i.e. Ti-OH) and Lewis-acid sites (exposed Tin+ cation) [27,28]. Since NaBH4 hydrolysis can catalyze by acid sites [29,30], so TiO2 nanotubes can be added to reduce the activation energy and accelerate the generation of hydrogen. The hydrolysis reactions of sodium borohydride in existence of catalysts proceed via a seven-step reaction mechanism as presented in reactions (2)–(9) [9,49]:
In the aforementioned reactions, the M symbol illustrates the TiO2 catalysts. Based on the reactions above, we proposed a schematic illustration to clarify the mechanism of the hydrogen evolution from the catalytic hydrolysis of NaBH4, which is demonstrated in Fig. 7. The first step of the reaction is reversible chemisorption of borohydride ions (BH4-) and H2O molecules to the active sites of TiO2 atoms, respectively (reaction (2) and (3)). In the second step, a proton on the borohydride ion then exchanges with a hydroxide ion to form M-H and comparatively stable intermediate BH3(OH)- based on reaction (4). Simultaneously, hydridic H- of BH4- reacts with protic hydrogen of H2O leads to the generation H2 as represented by reaction (5). The generated intermediate ion (BH3(OH)-) is expected to have similar reactivity as borohydride ion and the cycle of charge transfer goes through similar steps as BH3(OH)-→BH2(OH)2- → BH(OH)3- → B(OH)4, releasing H2 at each step(reaction (6)-(9)).
By comparing two different TiO2 nanotubes with commercial TiO2 powder, it can be observed both DF-NTs and FS-NTs catalysts improved the hydrogen generation rates from 3.5 ml.min-1 (for TiO2 powder) to 11 ml.min-1 (for DF-NTs) and 18.4 (for FS-NTs) ml.min-1, respectively. Based on Fig. 6, FS-NTs catalyst showed better catalytic activity than the DF-NTs catalyst which it could cause more than 1.5-fold improvement on the hydrogen generation volume and rate from 80 to 140 ml and from 11 to 18.4 ml.min-1, respectively. This remarkable enhancement in catalytic activity of FS-NTs than DF-NTs catalyst can result from two factors. Firstly, based on Table 2, FS-NTs have the highest surface area than that DF-NTs catalyst. It is evident that the catalyst surface area and its catalytic performance are only related together indirectly. A catalyst with the larger surface area is assumed to have more exposed catalytically active places comparing to the same catalyst with the smaller surface area. Hence FS-NTs catalyst showed better catalytic performance than two other catalysts in hydrogen generation as a result of the improvement in the surface area. Secondly, more orderliness and oriented structure of FS-NTs than DF-NTs catalyst lead to easily transfer of reagents and more adsorbed on the nanotubes surfaces. More adsorption of reagents on reactive sites causes more accelerations of catalytic hydrolysis rate and consequently generates more hydrogen over FS-NTs catalysts.
Kinetic study of the catalytic hydrolysis of NaBH4
To investigate the kinetic properties of the catalytic hydrolysis of NaBH4 with FS-NTs catalysts, the reaction rates were studied as a function of catalyst concentration and amount of NaBH4. The influence of NaBH4 level on the rate of hydrogen generation was examined with various initial levels of NaBH4 ranging from 1 wt% to 10 wt% while the constant level of NaOH (1.5 wt%) was used. Fig. 8(a) represents the diagram corresponding to the volume of hydrogen generated versus time over the hydrolysis of NaBH4 using diverse initial NaBH4 concentrations at 30 °C. The volume of generated hydrogen increased by rising up the NaBH4 concentration from 1 wt.% to 5 wt.%, and next it started to reduce regarding the rise of initial NaBH4 concentration to 10 wt.%. This phenomenon can be explained by the way that when the NaBH4 level is lower than the value where the maximum rate of hydrogen generation is obtained, more BH4- and H2O can contact the active place on catalyst surface to generate hydrogen at higher NaBH4 concentration. However, it should be noticed that NaBO2 is produced simultaneously with hydrogen, and higher initial NaBH4concentration can lead to more NaBO2 accumulated on the surface of catalyst and in the solution because of the low solubility of NaBO2 in alkaline solution. Consequently, the catalytic active places on the catalyst will be blocked, the solution viscosity will increase and the mass transfer will be impeded [23, 50]. Fig. 8(b) shows the diagrams corresponding to the rate of hydrogen generation versus NaBH4 concentration, both in logarithmic scales. The slope corresponding to the resultant straight line is observed to be 0.016, which is close to zero, implying that the NaBH4catalytic hydrolysis over FS-NTs is a zero-order reaction regarding the concentration of NaBH4.
The influence of catalyst concentration on generating hydrogen for the catalytic hydrolysis of NaBH4 using FS-NTs was further investigated by performing a set of control experiments using diverse levels of FS-NTs catalysts in a range of 50-200 mg while the other reaction conditions were constant. Fig. 8(c) indicates the plot corresponding to the hydrogen generation volume versus time for hydrolyzing NaBH4 using diverse initial catalyst amount at 30 °C. As shown in Fig. 8(c), increasing the catalyst amount could promote the rate of reaction apparently. Based on the Fig. 7(c), as the amount of the catalyst was enhanced from 50 to200 mg, the hydrogen generation volume increased from 56 mL to 216 mL after 15 min of reaction. Fig. 8(d) illustrate the diagram corresponding to the rate of hydrogen generation versus FS-NTs amount, both in logarithmic scales. As shown in Fig. 8(d), the catalytic rate exhibits a linear growth with catalyst amount. Achieving a fitthed straight line with a slope very close to 1 (1.05) demonstrated that hydrolyzing NaBH4 using the catalyst of FS-NTs is a first-order reaction respecting to the concentration of the catalyst.
Hence as a result of the kinetic study, the rate law for hydrolyzing of NaBH4 over using the catalyst of FS-NTs under our reaction conditions can be represented as follows:
Where r is the rate of reaction and k represents the rate constant. This equation is in good agreement with the results reported by Shen et al. for an Ag-activated TiO2 catalyst , by Cheng et al. for a Co-B-TiO2 catalyst , and by Rakap et al. for a Co-Ni-P/Pd-TiO2 catalyst .
The durability and recyclability of a catalyst is another important parameter which determines its potential in practical hydrogen generation systems. The durability and reusability of FS-NTs for generating hydrogen from NaBH4 were evaluated in aqueous solution up to the fifth run at 30 °C. Fig. 9 represents the volumes of generated hydrogen and the rates of hydrogen generation. The catalytic performance of FS-NTs in hydrogen generation volumes remained almost unchanged, preserving 87% of its initial catalytic performance after five cycles. As shown in Fig. 9, no significant decay in the rate of hydrogen generation was observed even after fifth runs in the reusability test, which demonstrates the superior catalytic durability of FS-NTs to accelerate the NaBH4hydrolysis.
In summary, the generating hydrogen from the catalytic hydrolysis of NaBH4 as a function of TiO2 nanostructures was examined. For this purpose, DF-NTs and FS-NTs were prepared via the hydrothermal and chemical-assisted two-step anodization processes, respectively. The FESEM results confirmed that hydrothermal method leads to preparation of disordered filaments TiO2nanotubes while chemical-assisted two-step anodization method causes the production of ordered free-standing TiO2 nanotubes. Furthermore, the BET results showed that the FS-NTs catalyst has the highest specific surface area (112.77 m2/g) which it was 1.30 and 2.25 times greater than that DF-NTs (86.54 m2/g) and spherical TiO2 (50 m2/g) samples, respectively. As a result of enhancement in the specific surface area, the FS-NTs catalysts showed preferable activity in the hydrogen generation comparing with the two other catalysts. In this regard, FS-NTs could generate 140 ml H2 with the rate 18.4 ml.min-1 that it was 1.75 and 7 times more than the volume corresponding to generated H2 by DF-NTs and spherical TiO2, respectively. Moreover, the kinetic studies indicated that the NaBH4 catalytic hydrolysis by FS-NTs was a first-order reaction regarding the catalyst weight, whereas it was a zero-order reaction as a function of NaBH4 concentration. Furthermore, the FS-NTs catalyst revealed good stability and was reused for five runs without the significant loss of catalytic performance. To sum up, this research exhibited that using free-standing TiO2 nanotubes as a potent support catalyst can dramatically increase the rate of generating hydrogen from NaBH4 catalytic hydrolysis application.
CONFLICT OF INTEREST
The authors declare that there are no conflicts of interest regarding the publication of this manuscript.