- Email: [email protected]
Tailoring the multi-functionalities of one-dimensional ceria nanostructures via oxygen vacancy modulation
Journal of Colloid and Interface Science 504 (2017) 305–314
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis
Regular Article
Tailoring the multi-functionalities of one-dimensional ceria nanostructures via oxygen vacancy modulation Haiwei Du a, Tao Wan a, Bo Qu a, Jason Scott b, Xi Lin a, Adnan Younis a, Dewei Chu a,⇑ a b
School of Materials Science and Engineering, University of New South Wales, Sydney, NSW 2052, Australia Particles and Catalysis Research Group, School of Chemical Engineering, University of New South Wales, Sydney, NSW 2052, Australia
g r a p h i c a l a b s t r a c t
a r t i c l e
i n f o
Article history: Received 6 April 2017 Revised 16 May 2017 Accepted 17 May 2017 Available online 21 May 2017 Keywords: CeO2 nanostructure Oxygen vacancy Catalytic properties Resistance switching
a b s t r a c t Lattice defects, for example oxygen vacancies in cerium oxide (CeO2), usually play a vital role in determining physical and chemical properties, including catalytic performance and resistance switching behaviour. Here, tin (Sn) was introduced as a dopant in one dimensional CeO2 nanostructures to investigate oxygen vacancy modulation and achieve improved catalytic properties and a tunable electrical performance. Our findings revealed that the Sn-doped CeO2 nanorods maintained their morphology while the aspect ratio decreased gradually with increasing Sn content. The variation in oxygen vacancy concentration with Sn doping was confirmed by Raman and X-ray photoelectron spectroscopies and enhanced thermal catalytic and photo-catalytic performances were attained for the Sn-doped CeO2 nanorods. The variation in oxygen vacancy concentration with Sn doping was also found to influence its electrical properties. Hysteresis loops expressing resistance switching behaviour were observed in Sn-doped CeO2d nanorods. The results detailed in this study can help to rationally design nanostructures with the potential to provide desirable multi-functionalities. Ó 2017 Elsevier Inc. All rights reserved.
1. Introduction With the significant progress of nanotechnology, the past decades have witnessed the great potential of metal oxide semicon⇑ Corresponding author. E-mail address: [email protected] (D. Chu). http://dx.doi.org/10.1016/j.jcis.2017.05.057 0021-9797/Ó 2017 Elsevier Inc. All rights reserved.
ductor nanomaterials for catalytic applications. Generally, the unique characteristics of nanocrystals emerged from their size effect and morphology dependence. It is believed that, by reducing the size of nanocrystals, high surface to volume ratios can be achieved which are more effective for catalysis than their bulk counterparts. Additionally, the catalytic properties of nanocrystals also possess a strong shape/morphology-dependency [1,2]. The
306
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
designed morphology can ensure a selective exposure of actively reactive facets with a specific surface atomic arrangement and coordination to generate more active sites [3]. Among various nanostructures, one-dimensional (1-D) nanocrystals usually exhibit superior performances for catalytic applications as they both share certain common characteristics (such as quantum size effect and large specific surface area) with 0-D nanoparticles and 2-D nanosheets and can achieve the 1-D carrier transport efficiently [4]. As a consequence of these advantages, many metal oxides with rod-like morphologies have shown outstanding and exceptional catalytic activities with greater stability [3]. Another superiority of metal oxide nanomaterials is their capacity for defect engineering [5], which is frequently applied to tailor and further modify their material properties. In particular, as a representative lattice defect in metal oxides, oxygen vacancies unquestionably play a critical role in modulating either thermocatalytic (e.g. carbon monoxide (CO) oxidation) [6] or photocatalytic activities [7]. They can act as active sites to promote CO oxidation by facilitating oxygen migration or provide enhanced photocatalytic efficiencies whereby the oxygen vacancies within the structure can serve as electron traps to inhibit electron-hole pair recombination [8]. Cerium oxide (CeO2) nanocrystals have been widely studied as catalysts [9] or catalyst supports [10] (for example, metal particles) as they are recognized to possess extraordinary redox properties and a strong oxygen storage capacity (OSC) in nature. The intrinsic redox behaviour originates from the charge transfer between two oxidation states (Ce3+/Ce4+), consequently producing an oxygendeficient form, CeO2d, which is rich in oxygen vacancies and can more easily chemisorb and activate molecular oxygen compared with the stoichiometric CeO2 [11]. Generally, the concentration of oxygen vacancies in CeO2 nanomaterials can be modulated by three approaches: morphology design, doping or thermal annealing in a reducing atmosphere. The former two approaches have gained more attention from the scientific community in the field of nanocrystals synthesis. Firstly, with a cubic fluorite structure, the morphology evolution of CeO2 nanocrystals is usually associated with selectively exposed facets. As the formation energy of oxygen vacancies follows the sequence: (1 0 0) < (1 1 0) < (1 1 1) [12], structures with a {1 0 0}/{1 1 0}-dominated surface are catalytically more reactive for CO oxidation [13]. That is, nanorods preferentially exposing {1 0 0} and {1 1 0} facets with a higher OSC exhibit a better catalytic performance than nanocubes with only {1 0 0} facets or nano-octahedrons with only {1 1 1} facets, respectively [14]. Secondly, cation substitution with trivalent acceptor dopants possessing a similar ionic radius is often utilized to generate more oxygen vacancies in the lattice structure [15]. This strategy, based on forming oxygen vacancies induced by an extrinsic dopant, is designed to maintain charge neutrality. Additionally, doping is beneficial for band gap tuning, further facilitating photocatalytic efficiency. Among the many dopants, tin (Sn) has been shown to theoretically lower the formation energy of oxygen vacancies by electronic modification and structural distortion [16], and also experimentally to enhance the CO catalytic properties [17]. Despite the perceived benefits, and to the best of our knowledge, there currently are limited reports available on Sn-doped CeO2 1-D nanostructures which examine their catalytic properties. In this work, 1-D Sn-doped CeO2d nanorods are synthesized by a hydrothermal method and the effects of Sn doping on morphology, crystal structure and especially the oxygen vacancy concentration are studied. As the amount of Sn increases, Ce1xSnxO2d is found to maintain its rod-like morphology although the length to diameter (L/D) ratio decreases. Additionally, doping Sn into the ceria lattice enhanced the oxygen vacancy concentration, which was confirmed by Raman and XPS spectra. The Sn-doped CeO2d nanorods exhibited better catalytic CO oxidation and pho-
tocatalytic performances as compared to neat ceria. Furthermore, as the resistance switching (RS) behaviour in metal oxides is typically accompanied by the migration of oxygen vacancies [18] the RS characteristics of the Ce1xSnxO2d nanorods were investigated, to further reveal the role of oxygen vacancies on performance. 2. Experimental procedure 2.1. Materials Cerium chloride heptahydrate (CeCl37H2O, Mw: 372.58), tin(II) acetate (SnC4H6O4, Mw: 236.80) 1,2-propanediol (q: 1.036 g/mL), acetic acid (CH3COOH, Mw: 60.05) and ammonia hydroxide solution (q: 0.9 g/mL). All chemicals were purchased from SigmaAldrich and used without further purification. 2.2. Synthesis of Ce1xSnxO2d nanorods (x = 0, 0.05 and 0.10) To obtain CeO2d nanorods, 0.3 M CeCl37H2O was dissolved in 5 mL deionized (DI) water after which 100 lL acetic acid and 15 mL 1,2-propanediol were added into the solution in that order. For the synthesis of Ce1xSnxO2d nanorods, CeCl37H2O and SnC4H6O4 were weighed stoichiometrically without using acetic acid. After stirring for 5 min, 10 mL of ammonia hydroxide solution was added dropwise to the mixed solution under magnetic stirring, whereby the colour of the solution turned to light orange. After stirring for 30 min, the solution was transferred into a Teflon flask which was sealed tightly in a stainless-steel autoclave. Hydrothermal treatment was conducted at 160 °C for 12 h. After cooling, the white precipitate was collected and washed with DI water and then centrifuged at 5000 rpm for 2 min for five cycles, and then dried in an oven at 70 °C for 24 h. 2.3. Fabrication of FTO/Ce1xSnxO2d/Au devices The as-prepared Ce1-xSnxO2-d nanorods were dispersed into 5 mL of ethanol with assistance of ultrasound for 5 min. Afterwards, the suspension was drop-coated (a drop of 10 lL) once onto FTO glass, which acted as the bottom electrode, to provide a selfassembled film. The film was dried in an oven at 75 °C for 2 min. Finally, a patterned Au electrode with a size of 250 lm in diameter was sputtered through a shadow mask onto the film surface to act as the top electrode. 2.4. Materials characterization Structural analysis of the as-synthesized Ce1xSnxO2d nanorods was performed using an X-ray diffractometer with Cu Ka radiation (k = 0.1541 nm). The microstructures were visually observed by transmission electron microscopy, TEM (FEI Tecnai G2 and Philips CM200). Raman spectra were collected on a Renishaw inVia Raman Microscope with a 514 nm laser. The band gap was calculated by measuring the absorption spectra using a PerkinElmer UV–Visible Spectrometer. The chemical bonding states were investigated by X-ray photoelectron spectroscopy, XPS (ESCALAB250Xi spectrometer). The specific surface area (SSA) was determined by the Bru nauer–Emmett–Teller (BET) method using a Micromeritics ASAP 2000 Gas Sorption Analyser. The current-voltage (I-V) characteristics of the FTO/Ce1xSnxO2d/Au devices were tested by a Keysight B2902A source-meter. 2.5. H2-temperature-programmed reduction (H2-TPR) measurements the samples were analysed on a Micromeritics Autochem II. Approximately 100 mg of sample was added to a quartz U-cell
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
and placed in the instrument. The sample was initially flushed with 40 mL/min air at room temperature for 20 min after which it was heated to 300 °C at 15 °C/min where it was held for 30 min. The sample was then cooled to 50 °C and flushed with argon to remove the air. A 5 vol% H2 in Ar gas mix, flowing at 40 mL/min, was introduced to the sample where it was kept at 50 °C for 20 min. The sample was then heated to 1000 °C at 15 °C /min. It was held at this temperature for a further 5 min and then cooled to room temperature. H2 consumed during reduction was measured by a thermal conductivity detector (TCD) located within the instrument. 2.6. Catalytic evaluation Carbon monoxide oxidation activities were measured using a fixed bed flow micro-reactor at atmospheric pressure. In a typical experiment, the system was first purged with high purity nitrogen gas for 15 min and then a gas mixture comprising CO/O2/N2 (0.8:20:79.2) with a flow rate of 260 mL/min was introduced into the reactor, which contained 100 mg of the catalyst. Real-time concentrations of CO (at a resolution of 10 ppm) in the effluent gas were analysed by an online infrared gas analyzer (Thermo fisher Nicolet Is10). Photocatalytic activities of the prepared materials were assessed by decolourisation of an organic dye, methyl orange (MO). A 200 ml quartz round bottom flask containing a mixture of 35 mg of catalyst and 100 mL of a 15.5 mg/L dye solution. First, the reaction mixture was stirred in the dark for 30 min to achieve equilibration of the adsorption/desorption of the organic dye by
307
the catalyst. The mixture was then irradiated with the light source for 20 min, after which 5 mL aliquots were collected, followed by centrifugation to remove the catalyst particles. The supernatant was used to determine the concentration of residual dye by UVvis spectroscopy based on the absorbance at 464 nm for MO. This procedure was triplicated for each sample. 3. Results and discussion The morphologies and microstructures of the as-synthesized Ce1xSnxO2d nanorods were observed by TEM. As shown in Fig. 1(a), undoped CeO2d nanorods with a length around 50– 200 nm and uniform diameters in the range of 5–10 nm were fabricated. Fig. 1(e, f) and (i, j) show the TEM images for the Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d nanorods. Upon the introduction of Sn, the nanorod morphology changes slightly while the aspect ratio changes significantly as the length decreases and the diameter increases. The mean grain length and diameter were quantified by selecting more than 50 particles in an individual image. The mean grain length decreases from 99.7 nm to 36.1 nm and 33.6 nm, respectively, on increasing the Sn dopant content (Fig. 1c, g and k), while the of nanorod diameter increases from to 6.9 nm to approximately 11.3 nm for Ce0.95Sn0.05O2d and 16.19 nm for Ce0.90Sn0.10O2d (Fig. 1d, h and l), providing a decreased L/D ratio. A similar trend has also been observed for Ce1xYxO2 and Ce1xZrxO2 nanorods [19,20]. The high resolution TEM (HRTEM) images are shown in Fig. 2. It is reported that numerous {1 1 1}-type nanofacets exist on the (1 1 0) planes of ceria
Fig. 1. TEM images of Ce1xSnxO2d nanorods: CeO2d (a, b), Ce0.95Sn0.05O2d (e, f) and Ce0.90Sn0.10O2d (i, j). The corresponding grain length and diameter distribution of CeO2d (c, d), Ce0.95Sn0.05O2d (g, h) and Ce0.90Sn0.10O2d (k, l).
308
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
nanorods [21]. As shown in Fig. 2, the inter-planar spacing corresponding to the (1 1 1) plane decreases gradually with the increasing Sn dopant concentration since the ionic radius of Sn2+ (0.93 Å) is slightly smaller than that of Ce4+ (0.97 Å) [22]. This phenomenon also indicates successful doping of Sn into the lattice. Moreover, the nitrogen adsorption - desorption isotherms and pore size distributions of the Ce1xSnxO2d nanorods are shown in Fig. S1. The SSA decreases gradually from 156.6 to 106.0 m2/g (Table 1) as the Sn amount increases. First, it is well-known that a larger surface area can be achieved in particles with smaller pore diameter. Second, it is reported that the BET surface area decreases with the gradual contracted pore volume [23]. Thus, the decreased SSA of the Ce1xSnxO2d nanorods is probably attributed to the gradual increase of pore diameter (Table 1) and the decreased pore volume (as shown in (Fig. S1) d–f). Fig. 3(a) shows the XRD patterns of the Ce1xSnxO2d nanorods where the diffraction peaks can be indexed to a cubic fluorite structure (JCPDS No. 34-0394). The calculated lattice parameters are shown in Table 1. For the undoped CeO2d nanorods, the lattice parameter (a = 5.4296 Å) is larger than that of bulk CeO2 (a = 5.411 Å), which is attributed to the lattice expansion caused by a partial presence of Ce3+. As the ionic radius of Sn2+ (0.93 Å) is slightly smaller than that of Ce4+ (0.97 Å) [22], the lattice parameter is expected to be smaller following Sn2+ ion introduction. However, no obvious peak shift is observed in the enlarged XRD patterns and the lattice parameter is found to slightly increase upon doping 5 mol% and 10 mol% Sn. As mentioned above, bivalent dopants can lead to oxygen vacancy formation to achieve the charge neutrality, and this process is usually accompanied by a lattice expansion within the CeO2 nanoparticles. Moreover, grain size could also play a role in varying the lattice parameter of the CeO2 nanocrystals as lattice relaxation takes place with decreasing particle size [24,25]. According to the TEM results, Sn doping signifi-
cantly decreases grain size. Thus, the small variation in lattice parameter is attributed to the interaction between lattice shrinkage associated with atomic substitution and lattice expansion due to the formation of oxygen vacancies and the size effect. For the CeO2 nanoparticles, the atomic substitution and size effect both influence the lattice parameter. It has been reported that the lattice parameter increases gradually with the decreasing particle size and significantly when particle size decreases below 10 nm [25,26]. Thus, Sn doping is suggested to play the dominant role in increasing the lattice parameter in this work. Fig. 3(b and c) depict the Raman spectra of the Ce1xSnxO2d nanorods. A sharp peak ascribed to the pure CeO2d nanorods is located at 460 cm1, which corresponds to the Raman-active F2g mode and represents a symmetrical stretching of the Ce–O vibration in the fluorite structure. On doping Sn into the lattice, the F2g mode peak broadens and decreases in intensity, which could be attributed to smaller grain size and increased compressive strain induced in the lattice structure [27]. Additionally, the F2g peak gradually shifts towards a lower wavenumber (Fig. 3c), suggesting that Sn has diffused into the CeO2 lattice which has also been observed in other systems comprising CeO2 doped with Mn2+ [28] or Fe3+ [29]. Generally, as the physical properties of the dopant ions differ from those of the Ce ions in the matrix, embedding dopants with lower valence states into the lattice usually results in higher local structure distortion in the sublattice [30] and a more defective structure. The broad peak (green stripe) at around 600 cm1 is assigned to the intrinsic oxygen vacancies generated by the presence of Ce3+ [26]. Moreover, another peak (red stripe) at around 570 cm1 appears and the intensity increases with the increasing Sn amount. This peak could be ascribed to extrinsic oxygen vacancies that are caused by the replacement of Ce4+ with divalent Sn2+ [31]. It is well known that the relative peak intensity or area ratio between the peak at around 570–600 cm1 and the F2g mode peak
Fig. 2. HRTEM images of Ce1xSnxO2d nanorods: CeO2d (a), Ce0.95Sn0.05O2d (b) and Ce0.90Sn0.10O2d (c).
Table 1 Summary of the results from XRD, Raman, XPS and BET analyses. Sample
XRD
Raman a
CeO2d Ce0.95Sn0.05O2d Ce0.90Sn0.10O2d a b
XPS 3+
BET
a (Å)
I580/ IF2g
Area ratio of peak580/F2g
Ce / Cetotb
Area ratio of Osur/Olatt
Sn 3d5/2 position (eV)
Specific surface area (m2/g)
Adsorption average pore diameter (nm)
Desorption average pore diameter (nm)
5.4296 5.4298 5.4316
0.2157 0.4288 0.4900
0.2332 0.3286 0.3885
0.3191 0.3281 0.3456
0.5224 0.6195 0.7888
– 486.59 486.64
156.6342 113.2152 105.9526
5.9380 6.3266 6.3391
7.2029 7.2116 7.8431
Calculated by Jade 6.0 software. Ce3+/Cetot = Ce3+/(Ce3++Ce4+).
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
309
Fig. 3. XRD patterns (a), Raman spectra (b) and enlarged region at 450 cm1 (c) of Ce1xSnxO2d nanorods.
is related to the concentration of oxygen vacancies in CeO2. As is shown in Table 1, both the intensity ratio and area ratio increase with increasing Sn content, suggesting that the diffusion of Sn2+ into the CeO2 generates more oxygen vacancies in the lattice structure.
The chemical bonding and oxidation states of the elements were examined using XPS analysis, as shown in Fig. 4. From the survey spectrum, no Sn peaks are detected in CeO2d while two peaks appear at 480–500 eV after introducing Sn into the nanorods (Table 1). From the Ce 3d core-level spectra (Fig. 4b), the multiple
Fig. 4. XPS survey spectrum (a), Ce 3d (b), O 1s (c) and Sn 3d (d) spectra of Ce1xSnxO2d nanorods.
310
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
peaks are split to 10 sub peaks by curve-fitting, with u and v referring to Ce 3d3/2 and Ce 3d5/2 components, respectively. The peaks labelled v00, v000, u, u00 and uv000 are characteristic of Ce4+ while v0, u0, v0 and u0 are assigned to Ce3+. It is known that the oxygen vacancy concentration is reflected by the relative content of Ce3+, which can be determined according to the equation:
½Ce3þ ¼ ¼
Ce3þ 3þ
Ce
þ Ce4þ
Aðv0 þ
v0
Aðv0 þ v0 þ u0 þ u0 Þ þ u0 þ u0 Þ þ Aðv þ v00 þ v 000 þ u þ u00 þ u000 Þ
where A is the integrated area. The calculated percentages of Ce3+ for the various catalysts are listed in Table 1. It can be seen that the relative content of Ce3+ increases gradually with increasing Sn
content. Fig. 4c shows the O 1s spectra of the Ce1xSnxO2d nanorods. On de-convolution, the primary peaks can be split to three sub peaks: bulk O2 in the lattice (529–530 eV, OI), surface oxygen assigned to defect oxide or the weakly bonded oxygen species (530.7–531.5 eV, OII) and absorbed oxygen on the surface derived from hydroxyl species, water or carbonates (532–533 eV, OIII). The ratio of OII/OI is typically used to evaluate the proportion of oxygen species as a higher OII/OI indicates more active vacant oxygen species [32]. The calculated peak area ratio of OII/OI is observed to increase with Sn doping. The results are in accordance with the Ce 3d spectra, suggesting that the Sn-doped CeO2d nanorods have more O-vacancy defects than the neat nanorods. Additionally, the valence state of Sn can be determined via XPS, as shown in Fig. 4d. The Sn 3d5/2 positions for Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d are located at 486.59 and 486.64 eV, respectively, giving an
Fig. 5. CO conversion versus temperature (a) over Ce1xSnxO2d nanorods and H 2 TPR profiles (b) of Ce1xSnxO2d (x = 0 and x = 0.1).
Fig. 6. Photocatalytic decolourisation of methyl orange dye (a) under UV irradiation at room temperature. Pseudo first order plots of ln(C/C0) versus reaction time for the degradation of methyl orange dye (b). Cyclic runs for the photocatalytic decolourisation of methyl orange dye in the presence of Ce0.90Sn0.10O2d nanorods under UV irradiation at room temperature (c). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 7. Proposed mechanisms for the CO oxidation (a) and photocatalytic decolourisation (b) by Sn-doped CeO2d nanorods possessing oxygen vacancies. Assuming the oxygen vacancy is formed at the nearest site. O* is the active surface oxygen species.
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
estimated Sn valence state of +2.5 [33] with only mild oxidation occurring during the hydrothermal process. The increased oxygen vacancies can be ascribed to the replacement of Ce with Sn. The replacement of Ce with Sn is favourable for the formation of oxygen vacancy because of keeping charge neutrality and the decreased O-vacancy formation energy [16]. As the ionic radius of Sn2+ (0.93 Å) is very close to Ce4+ (0.97 Å) [22], Sn2+ is expected to substitute Ce4+ in the lattice and an oxygen vacancy ðVO Þ forms so as to maintain the charge balance. The oxygen vacancy formation reaction can be expressed in terms of Kröger-Vink notation, as shown below: CeO2
SnO ! Sn00Ce þ VO þ OO Additionally more Ce3+ can be created by a strong interaction between Ce4+/Ce3+ and Sn4+/Sn2+ via the redox equilibrium of 2Ce4+ + Sn2+ M 2Ce3+ + Sn4+ [34]. The increased concentration of Ce3+ usually results in an oxygen vacancy increase. In addition, it is reported that the longer M-O bonds in oxygen deficient CeO2 usually result in weaker bonded oxygen, and these oxygen are activated with a higher oxygen storage capacity [35,36]. When compared to Ce-O bonds, the average length of Sn-O bonds is longer [37]. Thus, the Sn-doped CeO2d nanorods potentially contain weak bonds which can contribute to the extraction of lattice oxygen, leaving the vacant sites. For CeO2 nanomaterials, both surface oxygen vacancies and bulk oxygen vacancies contribute to the catalytic performance. As the oxygen vacancy concentration decreases from the surface to the bulk [38], the oxygen defects induced by Sn dopants predominantly occur at the particle surface. To evaluate the catalytic performance, the Ce1xSnxO2d nanorods were employed to catalytically oxidise CO or photodecolourise methyl orange (MO) dye. As shown in Fig. 5a, the CO oxidation activity follows the order CeO2d < Ce0.95Sn0.05O2d < Ce0.90Sn0.10O2d. For CO oxidation catalysis, nanocrystals with a higher SSA usually exhibit a better conversion efficiency as the higher SSA provides more active sites for CO molecule adsorption. However, the SSA of the Ce1xSnxO2d nanorods decreases upon doping with Sn. Thus the contribution of increased oxygen vacancies [39] (as discussed in the Raman spectra and XPS analysis) appears to be a more dominant feature for improving CO catalytic properties. To support the CO catalytic activity, H2-TPR analyses (Fig. 5b) were
311
conducted to identify the contribution of the Sn dopant. The CeO2d nanorods exhibit two broad reduction peaks at around 450 °C and 800 °C, corresponding to surface/subsurface and bulk reduction, respectively. Upon doping with 10 mol% of Sn, three TPR peaks are observed which are similar to findings reported elsewhere [34]. The reduction peak of pure SnO2 is usually located at 680 °C [40] which derives from the Sn doping. This reduction peak indicates that more surface oxygen vacancies are generated at a lower temperature by an induced presence of Ce3+. The increased Ce3+ presence is attributed to the interaction between Ce4+/Ce3+ and Sn4+/Sn2+ via the redox equilibrium of 2Ce4+ + Sn2+ M 2Ce3+ + Sn4+. The two peaks above 500 °C for the Sn-doped ceria may be assigned to the reduction of Sn4+ formed in the CeO2 lattice and the bulk reduction of CeO2 [34]. The proposed mechanism for CO oxidation is shown in Fig. 7a. Oxygen vacancies in the defective Sn-doped nanorods can act as oxygen adsorption sites which can generate more active oxygen species [41]. These surface oxygen species actively facilitate the reaction of CO at the surface, thus improving the CO conversion efficiency. Fig. 6a shows the photocatalytic decolourisation of MO by the CeO2d nanorods with different dopant concentrations as a function of irradiation time. Without Sn doping, the photodecolourisation is low with only 27.2% of the MO being decolourised during 240 min of irradiation. However, on doping with Sn, the photocatalytic performance is improved. At a Sn dopant concentration of 10 mol%, the photocatalytic activity is significantly enhanced, with 70.5% of the MO being decolorized. Fig. 6b shows the plots of ln (C/C0) versus reaction time, according to the equation:
lnðC=C0 Þ ¼ kt where C0 is the initial concentration of the dye, C is the concentration of the dye at a certain time (t) and k is the apparent rate constant (min1). Generally, the k for dye decolourisation by a photocatalyst is used to assess the photocatalytic efficiency. As can be seen in Fig. 6b, the photocatalytic decolourisation rate can be fitted to follow pseudo-first-order kinetics. The highest rate constant (4.41 103 min1) of Ce0.90Sn0.10O2d is 2.4-fold and 3.3-fold higher than that of Ce0.95Sn0.05O2d and undoped CeO2d, respectively. It is known that the SSA of CeO2 nanocrystal-based phocatalysts generally plays a vital role in tuning the photocatalytic properties as the higher SSA of nanoparticles provides more active sites for the strong capture of dye molecules [42]. However, in this
Fig. 8. Schematic of FTO/Ce1xSnxO2d/Au device (a) and current-voltage (I-V) characteristics of FTO/Ce1xSnxO2d/Au device: CeO2d (b), Ce0.95Sn0.05O2d (c) and Ce0.90Sn0.10O2d (d). I-V characteristics of FTO/Ce1xSnxO2d/Au devices under different voltage ranges: 5 V 5 V (e), 10 V 10 V (f), 15 V 15 V (g) and 20 V 20 V (h). The sweeping direction is shown by the arrows.
312
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
work the SSA decreases upon introducing Sn dopant. In a similar vein to the CO catalytic results, the enhanced photocatalytic activity can be attributed to the increase in oxygen vacancies. The band gap, as estimated by the Tauc plot according to UV–Vis spectra, is shown in Fig. S2. CeO2d, Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d possess direct band gap values of 3.37, 3.33 and 3.26 eV, respectively. The shift in band gap can be ascribed to the increased oxygen vacancy concentration which can provide a red shift in the band gap [43]. The band edge position of the conduction band (CB) and valence band (VB) can be determined by the following equations: [44].
1 EVB ¼ v Ee þ Eg 2 ECB ¼ EVB Eg where v is the absolute electronegativity of the semiconductor (v = 5.57 eV for CeO2), Ee is the energy of free electrons on the hydrogen scale (4.5 eV) and Eg is the band gap [45]. The calculated VB and CB are 2.70 and 0.56 eV for Ce0.90Sn0.10O2d. Fig. 7b shows the proposed photocatalytic mechanism of the Sn-doped CeO2 nanorods. During UV irradiation, the photoelectrons in the valence band are excited to the CB and the same amount of holes in the VB are simultaneously generated (i.e. electron-hole pair formation). The photoinduced holes react with OH to form hydroxyl radicals (OH) while photogenerated electrons are trapped by oxygen vacancies as the surface traps [46], and further captured by oxygen molecules adsorbed at the oxygen-deficient sites to generate superoxide anions ( O 2 ). The MO is then oxidized by the resulting hydroxyl radicals and superoxide anions at the surface. Thus, the greater number of oxygen vacancies possessed by the Sn-doped CeO2 nanorods could facilitate the photocatalytic efficiency [47] and provide the enhanced photocatalytic performance. To further prove that the doping-generated oxygen vacancies predominantly contribute to the photocatalytic performance, the influence of UV irradation and a higher temperature on MO decolourisation are shown in Fig. S3. It can be seen that under UV irradation and at 120 °C Ce0.90Sn0.10O2d displays an optimum photocatalytic performance compared to under visible-light illumination or at room temperature. This phenomenon is attributed to the more active oxygen vacancies as oxygen vacancies can usually be activated by UV irradation [48] and their mobility can also be enhanced by a higher temperature [10,49]. Additionally, when compared with lattice vibration, the transportation mobility of lattice oxygen ions is more beneficial for separating photoinduced electrons and holes [50]. Consequently, oxygen vacancies are
believed to be predominantly responsible for the improved photocatalytic activity. In addition, to evaluate the photocatalytic stability, Ce0.90Sn0.10O2d nanorods were selected to deolourise MO over four consecutive cycles (Fig. 6c). It is apparent that the photocatalytic activity remains consistent across the four cyclic runs, indicating that the Sn-doped CeO2d possess excellent stability. It has been widely accepted that, in the instance of ceria based materials, oxygen vacancies can strongly influence their electrical properties as the distributed oxygen vacancies inside CeO2 films are usually crucial for modulating resistance [51] as well as can promote a resistive switching phenomenon especially when oxygen vacancies can form a filament as a conductive path [52]. The critical role of oxygen vacancies in defining the electrical properties of our fabricated un-doped and doped CeO2 nanostructures was evaluated by conducting current–voltage (I-V) measurements at various positive and negative potentials. To study the I-V characteristics, devices based on Ce1xSnxO2d (x = 0, 0.05 and 0.1) nanorods were fabricated using a drop-coating approach. Fig. 8a shows a schematic of a FTO/Ce1xSnxO2d/Au device with the asymmetric IV curves exhibited by each device arising from the asymmetric electrode configuration. As shown in Fig. 8b, the film comprising undoped CeO2d rods shows non-linear I-V characteristics, which can be considered as a typical non-linear resistor. In contrast, hysteresis loops are observed for the Sn-doped CeO2d devices (Fig. 8e–h). The behavioral change in I-V characteristics implys that charge carriers other than electrons are probably involved in driving conduction. Similar hysteresis loops have been seen in other materials [53–55] including thin/thick films and single crystals. Unlike previous studies where the resistance switching was a consequence of UV illumination [55], a similar phenomenon is present here, which instead arisies from the increased oxygen vacancy concentration due to the doping effect. In addition, the calculated resistance decreases by an order of magnitude, from 210 MX for the undoped CeO2d to 30–50 MX for the Ce0.95Sn0.05O2d and 10–20 MX for the Ce0.90Sn0.10O2d films. In the case of the neat material, the anionic defect concentration is not sufficiently high to provide a considerable RS effect even with the different voltage ranges. Upon introducing Sn, more oxygen vacancies are available to promote a considerable RS phenomenon as the loops become more obvious. By applying a positive potential, the vacancies begin to migrate towards the counter electrode and reduce the local resistance in the film (as shown in Fig. S4). The gradual increase in the current level of the devices also provides evidence on the increased amount of anionic defects (oxygen vacancies) [56]. It is noted that UV irradation has a significant influence on the oxygen vacancies within CeO2 as more oxygen vacancies can be
Fig. 9. Current-voltage characteristics of FTO/Ce1xSnxO2d/Au devices before and after UV irradation: (a) CeO2d and (b) Ce0.90Sn0.10O2d.
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
generated by UV irradation which are readily detected in UV Raman spectra [31,57]. Thus, I-V characteristics before and after UV irradation are shown in Fig. 9 to further demonstrate the contribution of oxygen vacancies in the Ce1xSnxO2d nanorods. From Fig. 9a, it can be seen clearly that the undoped device performs very differently following exposure to UV irradiation, as an apparent loop with a higher current is observed when compared with the pristine device. In the case of the 10 mol% Sn-doped CeO2d device (Fig. 9b), the current is also observed to increase two-fold compared to when operated without UV exposure. Raman spectra was also conducted for both conditions (Fig. S5) and the relative peak intensity or peak area ratio of peak580/F2g becomes larger after UV irradation, indicating more oxygen vacancies are generated. The findings demonstrate that oxygen vacancy concentrations to increase within the ceria nanostructure upon introducing the Sn dopant whereby the doped ceria nanostructures exhibit enhanced catalytic activities for dye degradation and CO oxidation as well as provide improved RS performances.
[8] [9]
[10]
[11]
[12] [13]
[14] [15]
4. Conclusions
[16]
In this work, one-dimensional Ce1xSnxO2d nanostructures were synthesized by a facile hydrothermal method with the interplay between Sn dopant and oxygen vacancies, and studied by means of catalytic and current-voltage (I-V) measurements for potential multifunctional applications. It was found that the introduction of Sn into the lattice only slightly altered the nanorod morphology although it invoked a lower aspect ratio and decreased specific surface area. Importantly, Sn doping promoted the amount of oxygen vacancies as evidenced by Raman and XPS spectra analyses. The increased active oxygen vacancies in the Sn-doped CeO2d nanorods facilitated both carbon monoxide catalytic oxidation and photocatalytic oxidation activities. In addition, obvious loops with resistance switching behaviour were observed in I-V curves of the Sn-doped CeO2d nanorods, further illustrating the higher oxygen vacancy concentration. The results indicate that modulating oxygen vacancies by compositional design can be utilized to achieve desirable multi-functionalities, especially for catalytic and memory applications.
[17]
Acknowledgement This work is funded by the Australian Research Council Project (grant no. FT140100032). The authors thank Ms. Katie Levick, Dr. Anne Rich and Dr. Bill Gong for assistance with TEM, Raman spectra and XPS measurements. H. Du thanks the China Scholarship Council (CSC) for financial support (No. 201406410060).
[18] [19]
[20]
[21]
[22] [23]
[24] [25] [26]
[27]
[28]
[29]
Appendix A. Supplementary material [30]
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2017.05.057. [31]
References [1] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chemistry and properties of nanocrystals of different shapes, Chem. Rev. 105 (2005) 1025–1102. [2] J. Pal, T. Pal, Faceted metal and metal oxide nanoparticles: design, fabrication and catalysis, Nanoscale 7 (2015) 14159–14190. [3] Y. Li, W. Shen, Morphology-dependent nanocatalysts: Rod-shaped oxides, Chem. Soc. Rev. 43 (2014) 1543–1574. [4] X. Wang, Z. Li, J. Shi, Y. Yu, One-dimensional titanium dioxide nanomaterials: nanowires, nanorods, and nanobelts, Chem. Rev. 114 (2014) 9346–9384. [5] E.W. McFarland, H. Metiu, Catalysis by doped oxides, Chem. Rev. 113 (2013) 4391–4427. [6] F. Zaera, Nanostructured materials for applications in heterogeneous catalysis, Chem. Soc. Rev. 42 (2013) 2746–2762. [7] J. Liqiang, Q. Yichun, W. Baiqi, L. Shudan, J. Baojiang, Y. Libin, F. Wei, F. Honggang, S. Jiazhong, Review of photoluminescence performance of nano-
[32]
[33]
[34]
[35]
[36]
313
sized semiconductor materials and its relationships with photocatalytic activity, Sol. Energ. Mat. Sol. Cells 90 (2006) 1773–1787. C.T. Campbell, C.H. Peden, Oxygen vacancies and catalysis on ceria surfaces, Science 309 (2005) 713–714. H. Du, Y. Wang, H. Arandiyan, A. Younis, J. Scott, B. Qu, T. Wan, X. Lin, J. Chen, D. Chu, Design and synthesis of CeO2 nanowire/MnO2 nanosheet heterogeneous structure for enhanced catalytic properties, Mater. Today Commun. 11 (2017) 103–111. T. Montini, M. Melchionna, M. Monai, P. Fornasiero, Fundamentals and catalytic applications of CeO2-based materials, Chem. Rev. 116 (2016) 5987– 6041. Y. Gao, R. Li, S. Chen, L. Luo, T. Cao, W. Huang, Morphology-dependent interplay of reduction behaviors, oxygen vacancies and hydroxyl reactivity of CeO2 nanocrystals, Phys. Chem. Chem. Phys. 17 (2015) 31862–31871. M. Nolan, S.C. Parker, G.W. Watson, The electronic structure of oxygen vacancy defects at the low index surfaces of ceria, Surf. Sci. 595 (2005) 223–232. H.-X. Mai, L.-D. Sun, Y.-W. Zhang, R. Si, W. Feng, H.-P. Zhang, H.-C. Liu, C.-H. Yan, Shape-selective synthesis and oxygen storage behavior of ceria nanopolyhedra, nanorods, and nanocubes, J. Phys. Chem. B 109 (2005) 24380–24385. Z. Wu, M. Li, S.H. Overbury, On the structure dependence of CO oxidation over CeO2 nanocrystals with well-defined surface planes, J. Catal. 285 (2012) 61–73. D.A. Andersson, S.I. Simak, N.V. Skorodumova, I.A. Abrikosov, B. Johansson, Optimization of ionic conductivity in doped ceria, Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 3518–3521. Y. Tang, H. Zhang, L. Cui, C. Ouyang, S. Shi, W. Tang, H. Li, J.-S. Lee, L. Chen, Firstprinciples investigation on redox properties of M-doped CeO2 (M= Mn, Pr, Sn, Zr), Phys. Rev. B 82 (2010) 125104. J. Ayastuy, A. Iglesias-González, M. Gutiérrez-Ortiz, Synthesis and characterization of low amount tin-doped ceria (CeXSn1-XO2-d) for catalytic CO oxidation, Chem. Eng. J. 244 (2014) 372–381. A. Sawa, Resistive switching in transition metal oxides, Mater. Today 11 (2008) 28–36. A.D. Liyanage, S.D. Perera, K. Tan, Y. Chabal, K.J. Balkus Jr, Synthesis, characterization, and photocatalytic activity of Y-doped CeO2 nanorods, ACS Catal. 4 (2014) 577–584. X. Liu, J. Ding, X. Lin, R. Gao, Z. Li, W.-L. Dai, Zr-doped CeO2 nanorods as versatile catalyst in the epoxidation of styrene with tert-butyl hydroperoxide as the oxidant, Appl. Catal. A 503 (2015) 117–123. C. Yang, X. Yu, S. Heißler, A. Nefedov, S. Colussi, J. Llorca, A. Trovarelli, Y. Wang, C. Wöll, Surface Faceting and Reconstruction of Ceria Nanoparticles, Angew. Chem. Int. Ed. 56 (2017) 375–379. C.P. Poole Jr, Encyclopedic dictionary of condensed matter physics, Academic Press, 2004. Y.H. Tan, J.A. Davis, K. Fujikawa, N.V. Ganesh, A.V. Demchenko, K.J. Stine, Surface area and pore size characteristics of nanoporous gold subjected to thermal, mechanical, or surface modification studied using gas adsorption isotherms, cyclic voltammetry, thermogravimetric analysis, and scanning electron microscopy, J. Mater. Chem. 22 (2012) 6733–6745. S. Tsunekawa, R. Sahara, Y. Kawazoe, K. Ishikawa, Lattice relaxation of monosize CeO2-x nanocrystalline particles, Appl. Surf. Sci. 152 (1999) 53–56. S. Tsunekawa, K. Ishikawa, Z.-Q. Li, Y. Kawazoe, A. Kasuya, Origin of anomalous lattice expansion in oxide nanoparticles, Phys. Rev. Lett. 85 (2000) 3440. S. Deshpande, S. Patil, S.V. Kuchibhatla, S. Seal, Size dependency variation in lattice parameter and valency states in nanocrystalline cerium oxide, Appl. Phys. Lett. 87 (2005). 133113 133113. J.E. Spanier, R.D. Robinson, F. Zhang, S.-W. Chan, I.P. Herman, Size-dependent properties of CeO2-y nanoparticles as studied by Raman scattering, Phys. Rev. B 64 (2001) 245407. X. Zhang, J. Wei, H. Yang, X. Liu, W. Liu, C. Zhang, Y. Yang, One-pot synthesis of Mn-doped CeO2 nanospheres for CO oxidation, Eur. J. Inorg. Chem. 2013 (2013) 4443–4449. Z. Zhang, D. Han, S. Wei, Y. Zhang, Determination of active site densities and mechanisms for soot combustion with O2 on Fe-doped CeO2 mixed oxides, J. Catal. 276 (2010) 16–23. P. Sudarsanam, B. Hillary, D.K. Deepa, M.H. Amin, B. Mallesham, B.M. Reddy, S. K. Bhargava, Highly efficient cerium dioxide nanocube-based catalysts for low temperature diesel soot oxidation: the cooperative effect of cerium-and cobalt-oxides, Cataly. Sci. Technol. 5 (2015) 3496–3500. M. Guo, J. Lu, Y. Wu, Y. Wang, M. Luo, UV and visible Raman studies of oxygen vacancies in rare-earth-doped ceria, Langmuir 27 (2011) 3872–3877. Y. Zhang, Z. Qin, G. Wang, H. Zhu, M. Dong, S. Li, Z. Wu, Z. Li, Z. Wu, J. Zhang, Catalytic performance of MnOx-NiO composite oxide in lean methane combustion at low temperature, Appl. Cataly. B 129 (2013) 172–181. C. Liang, Y. Shimizu, T. Sasaki, N. Koshizaki, Synthesis of ultrafine SnO2-x nanocrystals by pulsed laser-induced reactive quenching in liquid medium, J. Phys. Chem. B 107 (2003) 9220–9225. C. Li, G. Zeng, Y. Zhou, X. Zhang, The selective catalytic reduction of NO with NH3 over a novel Ce-Sn-Ti mixed oxides catalyst: promotional effect of SnO2, Appl. Surf. Sci. 342 (2015) 174–182. A. Gupta, M. Hegde, K. Priolkar, U. Waghmare, P. Sarode, S. Emura, Structural investigation of activated lattice oxygen in Ce1-xSnxO2 and Ce1-x-ySnxPdyO2-d by EXAFS and DFT calculation, Chem. Mater. 21 (2009) 5836–5847. G. Dutta, U.V. Waghmare, T. Baidya, M. Hegde, K. Priolkar, P. Sarode, Origin of enhanced reducibility/oxygen storage capacity of Ce1-xTixO2 compared to CeO2 or TiO2, Chem. Mater. 18 (2006) 3249–3256.
314
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
[37] A. Gupta, A. Kumar, M. Hegde, U. Waghmare, Structure of Ce1-xSnxO2 and its relation to oxygen storage property from first-principles analysis, J. Chem. Phys. 132 (2010) 194702. [38] S. Turner, S. Lazar, B. Freitag, R. Egoavil, J. Verbeeck, S. Put, Y. Strauven, G. Van Tendeloo, High resolution mapping of surface reduction in ceria nanoparticles, Nanoscale 3 (2011) 3385–3390. [39] Z.-Y. Pu, X.-S. Liu, A.-P. Jia, Y.-L. Xie, J.-Q. Lu, M.-F. Luo, Enhanced activity for CO oxidation over Pr-and Cu-doped CeO2 catalysts: effect of oxygen vacancies, J. Phys. Chem. C 112 (2008) 15045–15051. [40] X. Yao, C. Tang, Z. Ji, Y. Dai, Y. Cao, F. Gao, L. Dong, Y. Chen, Investigation of the physicochemical properties and catalytic activities of Ce0.67M0.33O2 (M= Zr4+, Ti4+, Sn4+) solid solutions for NO removal by CO, Cataly, Sci. Technol. 3 (2013) 688–698. [41] X. Liu, K. Zhou, L. Wang, B. Wang, Y. Li, Oxygen vacancy clusters promoting reducibility and activity of ceria nanorods, J. Am. Chem. Soc. 131 (2009) 3140– 3141. [42] J. Tian, Y. Sang, Z. Zhao, W. Zhou, D. Wang, X. Kang, H. Liu, J. Wang, S. Chen, H. Cai, Enhanced photocatalytic performances of CeO2/TiO2 nanobelt heterostructures, Small 9 (2013) 3864–3872. [43] B. Tatar, E. Sam, K. Kutlu, M. Ürgen, Synthesis and optical properties of CeO2 nanocrystalline films grown by pulsed electron beam deposition, J. Mater. Sci. 43 (2008) 5102. [44] Y. Xu, M.A. Schoonen, The absolute energy positions of conduction and valence bands of selected semiconducting minerals, Am. Mineral. 85 (2000) 543–556. [45] Z.-M. Yang, G.-F. Huang, W.-Q. Huang, J.-M. Wei, X.-G. Yan, Y.-Y. Liu, C. Jiao, Z. Wan, A. Pan, Novel Ag3PO4/CeO2 composite with high efficiency and stability for photocatalytic applications, J. Mater. Chem. A 2 (2014) 1750– 1756. [46] A.L. Linsebigler, G. Lu, J.T. Yates Jr, Photocatalysis on TiO2 surfaces: principles, mechanisms, and selected results, Chem. Rev. 95 (1995) 735–758.
[47] W. Lei, T. Zhang, L. Gu, P. Liu, J.A. Rodriguez, G. Liu, M. Liu, Surface-structure sensitivity of CeO2 nanocrystals in photocatalysis and enhancing the reactivity with nanogold, ACS Catal. 5 (2015) 4385–4393. [48] Z. Wu, M. Li, J. Howe, H.M. Meyer III, S.H. Overbury, Probing defect sites on CeO2 nanocrystals with well-defined surface planes by Raman spectroscopy and O2 adsorption, Langmuir 26 (2010) 16595–16606. [49] A. Trovarelli, Catalytic properties of ceria and CeO2-containing materials, Cataly. Rev. 38 (1996) 439–520. [50] Y. Li, Q. Sun, M. Kong, W. Shi, J. Huang, J. Tang, X. Zhao, Coupling oxygen ion conduction to photocatalysis in mesoporous nanorod-like ceria significantly improves photocatalytic efficiency, J. Phys. Chem. C 115 (2011) 14050–14057. [51] P. Gao, Z. Wang, W. Fu, Z. Liao, K. Liu, W. Wang, X. Bai, E. Wang, In situ TEM studies of oxygen vacancy migration for electrically induced resistance change effect in cerium oxides, Micron 41 (2010) 301–305. [52] X. Sun, B. Sun, L. Liu, N. Xu, X. Liu, R. Han, J. Kang, G. Xiong, T. Ma, Resistive switching in CeOx films for nonvolatile memory application, IEEE Electron Device Lett. 30 (2009) 334–336. [53] F. Nardi, S. Balatti, S. Larentis, D.C. Gilmer, D. Ielmini, Complementary switching in oxide-based bipolar resistive-switching random memory, IEEE Trans. Electron Dev. 60 (2013) 70–77. [54] T. Lee, K. Sung, T. Kim, J.-H. Ko, J. Jung, Observation of three crystalline layers in hydrothermally grown BiFeO3 thick films, J. Appl. Phys. 116 (2014) 194101. [55] J. Li, C. Ma, K. Jin, C. Ge, L. Gu, X. He, W. Zhou, Q. Zhang, H. Lu, G. Yang, Temperature-dependent resistance switching in SrTiO3, Appl. Phys. Lett. 108 (2016) 242901. [56] X. Tang, X. Zhu, J. Dai, J. Yang, L. Chen, Y. Sun, Evolution of the resistive switching in chemical solution deposited-derived BiFeO3 thin films with dwell time and annealing temperature, J. Appl. Phys. 113 (2013) 043706. [57] Y. Kang, Q. Leng, D. Guo, D. Yang, Y. Pu, C. Hu, Room-temperature magnetism of ceria nanocubes by inductively transferring electrons to Ce atoms from nearby oxygen vacancy, Nano-Micro Lett. 8 (2016) 13–19.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis
Regular Article
Tailoring the multi-functionalities of one-dimensional ceria nanostructures via oxygen vacancy modulation Haiwei Du a, Tao Wan a, Bo Qu a, Jason Scott b, Xi Lin a, Adnan Younis a, Dewei Chu a,⇑ a b
School of Materials Science and Engineering, University of New South Wales, Sydney, NSW 2052, Australia Particles and Catalysis Research Group, School of Chemical Engineering, University of New South Wales, Sydney, NSW 2052, Australia
g r a p h i c a l a b s t r a c t
a r t i c l e
i n f o
Article history: Received 6 April 2017 Revised 16 May 2017 Accepted 17 May 2017 Available online 21 May 2017 Keywords: CeO2 nanostructure Oxygen vacancy Catalytic properties Resistance switching
a b s t r a c t Lattice defects, for example oxygen vacancies in cerium oxide (CeO2), usually play a vital role in determining physical and chemical properties, including catalytic performance and resistance switching behaviour. Here, tin (Sn) was introduced as a dopant in one dimensional CeO2 nanostructures to investigate oxygen vacancy modulation and achieve improved catalytic properties and a tunable electrical performance. Our findings revealed that the Sn-doped CeO2 nanorods maintained their morphology while the aspect ratio decreased gradually with increasing Sn content. The variation in oxygen vacancy concentration with Sn doping was confirmed by Raman and X-ray photoelectron spectroscopies and enhanced thermal catalytic and photo-catalytic performances were attained for the Sn-doped CeO2 nanorods. The variation in oxygen vacancy concentration with Sn doping was also found to influence its electrical properties. Hysteresis loops expressing resistance switching behaviour were observed in Sn-doped CeO2d nanorods. The results detailed in this study can help to rationally design nanostructures with the potential to provide desirable multi-functionalities. Ó 2017 Elsevier Inc. All rights reserved.
1. Introduction With the significant progress of nanotechnology, the past decades have witnessed the great potential of metal oxide semicon⇑ Corresponding author. E-mail address: [email protected] (D. Chu). http://dx.doi.org/10.1016/j.jcis.2017.05.057 0021-9797/Ó 2017 Elsevier Inc. All rights reserved.
ductor nanomaterials for catalytic applications. Generally, the unique characteristics of nanocrystals emerged from their size effect and morphology dependence. It is believed that, by reducing the size of nanocrystals, high surface to volume ratios can be achieved which are more effective for catalysis than their bulk counterparts. Additionally, the catalytic properties of nanocrystals also possess a strong shape/morphology-dependency [1,2]. The
306
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
designed morphology can ensure a selective exposure of actively reactive facets with a specific surface atomic arrangement and coordination to generate more active sites [3]. Among various nanostructures, one-dimensional (1-D) nanocrystals usually exhibit superior performances for catalytic applications as they both share certain common characteristics (such as quantum size effect and large specific surface area) with 0-D nanoparticles and 2-D nanosheets and can achieve the 1-D carrier transport efficiently [4]. As a consequence of these advantages, many metal oxides with rod-like morphologies have shown outstanding and exceptional catalytic activities with greater stability [3]. Another superiority of metal oxide nanomaterials is their capacity for defect engineering [5], which is frequently applied to tailor and further modify their material properties. In particular, as a representative lattice defect in metal oxides, oxygen vacancies unquestionably play a critical role in modulating either thermocatalytic (e.g. carbon monoxide (CO) oxidation) [6] or photocatalytic activities [7]. They can act as active sites to promote CO oxidation by facilitating oxygen migration or provide enhanced photocatalytic efficiencies whereby the oxygen vacancies within the structure can serve as electron traps to inhibit electron-hole pair recombination [8]. Cerium oxide (CeO2) nanocrystals have been widely studied as catalysts [9] or catalyst supports [10] (for example, metal particles) as they are recognized to possess extraordinary redox properties and a strong oxygen storage capacity (OSC) in nature. The intrinsic redox behaviour originates from the charge transfer between two oxidation states (Ce3+/Ce4+), consequently producing an oxygendeficient form, CeO2d, which is rich in oxygen vacancies and can more easily chemisorb and activate molecular oxygen compared with the stoichiometric CeO2 [11]. Generally, the concentration of oxygen vacancies in CeO2 nanomaterials can be modulated by three approaches: morphology design, doping or thermal annealing in a reducing atmosphere. The former two approaches have gained more attention from the scientific community in the field of nanocrystals synthesis. Firstly, with a cubic fluorite structure, the morphology evolution of CeO2 nanocrystals is usually associated with selectively exposed facets. As the formation energy of oxygen vacancies follows the sequence: (1 0 0) < (1 1 0) < (1 1 1) [12], structures with a {1 0 0}/{1 1 0}-dominated surface are catalytically more reactive for CO oxidation [13]. That is, nanorods preferentially exposing {1 0 0} and {1 1 0} facets with a higher OSC exhibit a better catalytic performance than nanocubes with only {1 0 0} facets or nano-octahedrons with only {1 1 1} facets, respectively [14]. Secondly, cation substitution with trivalent acceptor dopants possessing a similar ionic radius is often utilized to generate more oxygen vacancies in the lattice structure [15]. This strategy, based on forming oxygen vacancies induced by an extrinsic dopant, is designed to maintain charge neutrality. Additionally, doping is beneficial for band gap tuning, further facilitating photocatalytic efficiency. Among the many dopants, tin (Sn) has been shown to theoretically lower the formation energy of oxygen vacancies by electronic modification and structural distortion [16], and also experimentally to enhance the CO catalytic properties [17]. Despite the perceived benefits, and to the best of our knowledge, there currently are limited reports available on Sn-doped CeO2 1-D nanostructures which examine their catalytic properties. In this work, 1-D Sn-doped CeO2d nanorods are synthesized by a hydrothermal method and the effects of Sn doping on morphology, crystal structure and especially the oxygen vacancy concentration are studied. As the amount of Sn increases, Ce1xSnxO2d is found to maintain its rod-like morphology although the length to diameter (L/D) ratio decreases. Additionally, doping Sn into the ceria lattice enhanced the oxygen vacancy concentration, which was confirmed by Raman and XPS spectra. The Sn-doped CeO2d nanorods exhibited better catalytic CO oxidation and pho-
tocatalytic performances as compared to neat ceria. Furthermore, as the resistance switching (RS) behaviour in metal oxides is typically accompanied by the migration of oxygen vacancies [18] the RS characteristics of the Ce1xSnxO2d nanorods were investigated, to further reveal the role of oxygen vacancies on performance. 2. Experimental procedure 2.1. Materials Cerium chloride heptahydrate (CeCl37H2O, Mw: 372.58), tin(II) acetate (SnC4H6O4, Mw: 236.80) 1,2-propanediol (q: 1.036 g/mL), acetic acid (CH3COOH, Mw: 60.05) and ammonia hydroxide solution (q: 0.9 g/mL). All chemicals were purchased from SigmaAldrich and used without further purification. 2.2. Synthesis of Ce1xSnxO2d nanorods (x = 0, 0.05 and 0.10) To obtain CeO2d nanorods, 0.3 M CeCl37H2O was dissolved in 5 mL deionized (DI) water after which 100 lL acetic acid and 15 mL 1,2-propanediol were added into the solution in that order. For the synthesis of Ce1xSnxO2d nanorods, CeCl37H2O and SnC4H6O4 were weighed stoichiometrically without using acetic acid. After stirring for 5 min, 10 mL of ammonia hydroxide solution was added dropwise to the mixed solution under magnetic stirring, whereby the colour of the solution turned to light orange. After stirring for 30 min, the solution was transferred into a Teflon flask which was sealed tightly in a stainless-steel autoclave. Hydrothermal treatment was conducted at 160 °C for 12 h. After cooling, the white precipitate was collected and washed with DI water and then centrifuged at 5000 rpm for 2 min for five cycles, and then dried in an oven at 70 °C for 24 h. 2.3. Fabrication of FTO/Ce1xSnxO2d/Au devices The as-prepared Ce1-xSnxO2-d nanorods were dispersed into 5 mL of ethanol with assistance of ultrasound for 5 min. Afterwards, the suspension was drop-coated (a drop of 10 lL) once onto FTO glass, which acted as the bottom electrode, to provide a selfassembled film. The film was dried in an oven at 75 °C for 2 min. Finally, a patterned Au electrode with a size of 250 lm in diameter was sputtered through a shadow mask onto the film surface to act as the top electrode. 2.4. Materials characterization Structural analysis of the as-synthesized Ce1xSnxO2d nanorods was performed using an X-ray diffractometer with Cu Ka radiation (k = 0.1541 nm). The microstructures were visually observed by transmission electron microscopy, TEM (FEI Tecnai G2 and Philips CM200). Raman spectra were collected on a Renishaw inVia Raman Microscope with a 514 nm laser. The band gap was calculated by measuring the absorption spectra using a PerkinElmer UV–Visible Spectrometer. The chemical bonding states were investigated by X-ray photoelectron spectroscopy, XPS (ESCALAB250Xi spectrometer). The specific surface area (SSA) was determined by the Bru nauer–Emmett–Teller (BET) method using a Micromeritics ASAP 2000 Gas Sorption Analyser. The current-voltage (I-V) characteristics of the FTO/Ce1xSnxO2d/Au devices were tested by a Keysight B2902A source-meter. 2.5. H2-temperature-programmed reduction (H2-TPR) measurements the samples were analysed on a Micromeritics Autochem II. Approximately 100 mg of sample was added to a quartz U-cell
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
and placed in the instrument. The sample was initially flushed with 40 mL/min air at room temperature for 20 min after which it was heated to 300 °C at 15 °C/min where it was held for 30 min. The sample was then cooled to 50 °C and flushed with argon to remove the air. A 5 vol% H2 in Ar gas mix, flowing at 40 mL/min, was introduced to the sample where it was kept at 50 °C for 20 min. The sample was then heated to 1000 °C at 15 °C /min. It was held at this temperature for a further 5 min and then cooled to room temperature. H2 consumed during reduction was measured by a thermal conductivity detector (TCD) located within the instrument. 2.6. Catalytic evaluation Carbon monoxide oxidation activities were measured using a fixed bed flow micro-reactor at atmospheric pressure. In a typical experiment, the system was first purged with high purity nitrogen gas for 15 min and then a gas mixture comprising CO/O2/N2 (0.8:20:79.2) with a flow rate of 260 mL/min was introduced into the reactor, which contained 100 mg of the catalyst. Real-time concentrations of CO (at a resolution of 10 ppm) in the effluent gas were analysed by an online infrared gas analyzer (Thermo fisher Nicolet Is10). Photocatalytic activities of the prepared materials were assessed by decolourisation of an organic dye, methyl orange (MO). A 200 ml quartz round bottom flask containing a mixture of 35 mg of catalyst and 100 mL of a 15.5 mg/L dye solution. First, the reaction mixture was stirred in the dark for 30 min to achieve equilibration of the adsorption/desorption of the organic dye by
307
the catalyst. The mixture was then irradiated with the light source for 20 min, after which 5 mL aliquots were collected, followed by centrifugation to remove the catalyst particles. The supernatant was used to determine the concentration of residual dye by UVvis spectroscopy based on the absorbance at 464 nm for MO. This procedure was triplicated for each sample. 3. Results and discussion The morphologies and microstructures of the as-synthesized Ce1xSnxO2d nanorods were observed by TEM. As shown in Fig. 1(a), undoped CeO2d nanorods with a length around 50– 200 nm and uniform diameters in the range of 5–10 nm were fabricated. Fig. 1(e, f) and (i, j) show the TEM images for the Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d nanorods. Upon the introduction of Sn, the nanorod morphology changes slightly while the aspect ratio changes significantly as the length decreases and the diameter increases. The mean grain length and diameter were quantified by selecting more than 50 particles in an individual image. The mean grain length decreases from 99.7 nm to 36.1 nm and 33.6 nm, respectively, on increasing the Sn dopant content (Fig. 1c, g and k), while the of nanorod diameter increases from to 6.9 nm to approximately 11.3 nm for Ce0.95Sn0.05O2d and 16.19 nm for Ce0.90Sn0.10O2d (Fig. 1d, h and l), providing a decreased L/D ratio. A similar trend has also been observed for Ce1xYxO2 and Ce1xZrxO2 nanorods [19,20]. The high resolution TEM (HRTEM) images are shown in Fig. 2. It is reported that numerous {1 1 1}-type nanofacets exist on the (1 1 0) planes of ceria
Fig. 1. TEM images of Ce1xSnxO2d nanorods: CeO2d (a, b), Ce0.95Sn0.05O2d (e, f) and Ce0.90Sn0.10O2d (i, j). The corresponding grain length and diameter distribution of CeO2d (c, d), Ce0.95Sn0.05O2d (g, h) and Ce0.90Sn0.10O2d (k, l).
308
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
nanorods [21]. As shown in Fig. 2, the inter-planar spacing corresponding to the (1 1 1) plane decreases gradually with the increasing Sn dopant concentration since the ionic radius of Sn2+ (0.93 Å) is slightly smaller than that of Ce4+ (0.97 Å) [22]. This phenomenon also indicates successful doping of Sn into the lattice. Moreover, the nitrogen adsorption - desorption isotherms and pore size distributions of the Ce1xSnxO2d nanorods are shown in Fig. S1. The SSA decreases gradually from 156.6 to 106.0 m2/g (Table 1) as the Sn amount increases. First, it is well-known that a larger surface area can be achieved in particles with smaller pore diameter. Second, it is reported that the BET surface area decreases with the gradual contracted pore volume [23]. Thus, the decreased SSA of the Ce1xSnxO2d nanorods is probably attributed to the gradual increase of pore diameter (Table 1) and the decreased pore volume (as shown in (Fig. S1) d–f). Fig. 3(a) shows the XRD patterns of the Ce1xSnxO2d nanorods where the diffraction peaks can be indexed to a cubic fluorite structure (JCPDS No. 34-0394). The calculated lattice parameters are shown in Table 1. For the undoped CeO2d nanorods, the lattice parameter (a = 5.4296 Å) is larger than that of bulk CeO2 (a = 5.411 Å), which is attributed to the lattice expansion caused by a partial presence of Ce3+. As the ionic radius of Sn2+ (0.93 Å) is slightly smaller than that of Ce4+ (0.97 Å) [22], the lattice parameter is expected to be smaller following Sn2+ ion introduction. However, no obvious peak shift is observed in the enlarged XRD patterns and the lattice parameter is found to slightly increase upon doping 5 mol% and 10 mol% Sn. As mentioned above, bivalent dopants can lead to oxygen vacancy formation to achieve the charge neutrality, and this process is usually accompanied by a lattice expansion within the CeO2 nanoparticles. Moreover, grain size could also play a role in varying the lattice parameter of the CeO2 nanocrystals as lattice relaxation takes place with decreasing particle size [24,25]. According to the TEM results, Sn doping signifi-
cantly decreases grain size. Thus, the small variation in lattice parameter is attributed to the interaction between lattice shrinkage associated with atomic substitution and lattice expansion due to the formation of oxygen vacancies and the size effect. For the CeO2 nanoparticles, the atomic substitution and size effect both influence the lattice parameter. It has been reported that the lattice parameter increases gradually with the decreasing particle size and significantly when particle size decreases below 10 nm [25,26]. Thus, Sn doping is suggested to play the dominant role in increasing the lattice parameter in this work. Fig. 3(b and c) depict the Raman spectra of the Ce1xSnxO2d nanorods. A sharp peak ascribed to the pure CeO2d nanorods is located at 460 cm1, which corresponds to the Raman-active F2g mode and represents a symmetrical stretching of the Ce–O vibration in the fluorite structure. On doping Sn into the lattice, the F2g mode peak broadens and decreases in intensity, which could be attributed to smaller grain size and increased compressive strain induced in the lattice structure [27]. Additionally, the F2g peak gradually shifts towards a lower wavenumber (Fig. 3c), suggesting that Sn has diffused into the CeO2 lattice which has also been observed in other systems comprising CeO2 doped with Mn2+ [28] or Fe3+ [29]. Generally, as the physical properties of the dopant ions differ from those of the Ce ions in the matrix, embedding dopants with lower valence states into the lattice usually results in higher local structure distortion in the sublattice [30] and a more defective structure. The broad peak (green stripe) at around 600 cm1 is assigned to the intrinsic oxygen vacancies generated by the presence of Ce3+ [26]. Moreover, another peak (red stripe) at around 570 cm1 appears and the intensity increases with the increasing Sn amount. This peak could be ascribed to extrinsic oxygen vacancies that are caused by the replacement of Ce4+ with divalent Sn2+ [31]. It is well known that the relative peak intensity or area ratio between the peak at around 570–600 cm1 and the F2g mode peak
Fig. 2. HRTEM images of Ce1xSnxO2d nanorods: CeO2d (a), Ce0.95Sn0.05O2d (b) and Ce0.90Sn0.10O2d (c).
Table 1 Summary of the results from XRD, Raman, XPS and BET analyses. Sample
XRD
Raman a
CeO2d Ce0.95Sn0.05O2d Ce0.90Sn0.10O2d a b
XPS 3+
BET
a (Å)
I580/ IF2g
Area ratio of peak580/F2g
Ce / Cetotb
Area ratio of Osur/Olatt
Sn 3d5/2 position (eV)
Specific surface area (m2/g)
Adsorption average pore diameter (nm)
Desorption average pore diameter (nm)
5.4296 5.4298 5.4316
0.2157 0.4288 0.4900
0.2332 0.3286 0.3885
0.3191 0.3281 0.3456
0.5224 0.6195 0.7888
– 486.59 486.64
156.6342 113.2152 105.9526
5.9380 6.3266 6.3391
7.2029 7.2116 7.8431
Calculated by Jade 6.0 software. Ce3+/Cetot = Ce3+/(Ce3++Ce4+).
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
309
Fig. 3. XRD patterns (a), Raman spectra (b) and enlarged region at 450 cm1 (c) of Ce1xSnxO2d nanorods.
is related to the concentration of oxygen vacancies in CeO2. As is shown in Table 1, both the intensity ratio and area ratio increase with increasing Sn content, suggesting that the diffusion of Sn2+ into the CeO2 generates more oxygen vacancies in the lattice structure.
The chemical bonding and oxidation states of the elements were examined using XPS analysis, as shown in Fig. 4. From the survey spectrum, no Sn peaks are detected in CeO2d while two peaks appear at 480–500 eV after introducing Sn into the nanorods (Table 1). From the Ce 3d core-level spectra (Fig. 4b), the multiple
Fig. 4. XPS survey spectrum (a), Ce 3d (b), O 1s (c) and Sn 3d (d) spectra of Ce1xSnxO2d nanorods.
310
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
peaks are split to 10 sub peaks by curve-fitting, with u and v referring to Ce 3d3/2 and Ce 3d5/2 components, respectively. The peaks labelled v00, v000, u, u00 and uv000 are characteristic of Ce4+ while v0, u0, v0 and u0 are assigned to Ce3+. It is known that the oxygen vacancy concentration is reflected by the relative content of Ce3+, which can be determined according to the equation:
½Ce3þ ¼ ¼
Ce3þ 3þ
Ce
þ Ce4þ
Aðv0 þ
v0
Aðv0 þ v0 þ u0 þ u0 Þ þ u0 þ u0 Þ þ Aðv þ v00 þ v 000 þ u þ u00 þ u000 Þ
where A is the integrated area. The calculated percentages of Ce3+ for the various catalysts are listed in Table 1. It can be seen that the relative content of Ce3+ increases gradually with increasing Sn
content. Fig. 4c shows the O 1s spectra of the Ce1xSnxO2d nanorods. On de-convolution, the primary peaks can be split to three sub peaks: bulk O2 in the lattice (529–530 eV, OI), surface oxygen assigned to defect oxide or the weakly bonded oxygen species (530.7–531.5 eV, OII) and absorbed oxygen on the surface derived from hydroxyl species, water or carbonates (532–533 eV, OIII). The ratio of OII/OI is typically used to evaluate the proportion of oxygen species as a higher OII/OI indicates more active vacant oxygen species [32]. The calculated peak area ratio of OII/OI is observed to increase with Sn doping. The results are in accordance with the Ce 3d spectra, suggesting that the Sn-doped CeO2d nanorods have more O-vacancy defects than the neat nanorods. Additionally, the valence state of Sn can be determined via XPS, as shown in Fig. 4d. The Sn 3d5/2 positions for Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d are located at 486.59 and 486.64 eV, respectively, giving an
Fig. 5. CO conversion versus temperature (a) over Ce1xSnxO2d nanorods and H 2 TPR profiles (b) of Ce1xSnxO2d (x = 0 and x = 0.1).
Fig. 6. Photocatalytic decolourisation of methyl orange dye (a) under UV irradiation at room temperature. Pseudo first order plots of ln(C/C0) versus reaction time for the degradation of methyl orange dye (b). Cyclic runs for the photocatalytic decolourisation of methyl orange dye in the presence of Ce0.90Sn0.10O2d nanorods under UV irradiation at room temperature (c). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 7. Proposed mechanisms for the CO oxidation (a) and photocatalytic decolourisation (b) by Sn-doped CeO2d nanorods possessing oxygen vacancies. Assuming the oxygen vacancy is formed at the nearest site. O* is the active surface oxygen species.
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
estimated Sn valence state of +2.5 [33] with only mild oxidation occurring during the hydrothermal process. The increased oxygen vacancies can be ascribed to the replacement of Ce with Sn. The replacement of Ce with Sn is favourable for the formation of oxygen vacancy because of keeping charge neutrality and the decreased O-vacancy formation energy [16]. As the ionic radius of Sn2+ (0.93 Å) is very close to Ce4+ (0.97 Å) [22], Sn2+ is expected to substitute Ce4+ in the lattice and an oxygen vacancy ðVO Þ forms so as to maintain the charge balance. The oxygen vacancy formation reaction can be expressed in terms of Kröger-Vink notation, as shown below: CeO2
SnO ! Sn00Ce þ VO þ OO Additionally more Ce3+ can be created by a strong interaction between Ce4+/Ce3+ and Sn4+/Sn2+ via the redox equilibrium of 2Ce4+ + Sn2+ M 2Ce3+ + Sn4+ [34]. The increased concentration of Ce3+ usually results in an oxygen vacancy increase. In addition, it is reported that the longer M-O bonds in oxygen deficient CeO2 usually result in weaker bonded oxygen, and these oxygen are activated with a higher oxygen storage capacity [35,36]. When compared to Ce-O bonds, the average length of Sn-O bonds is longer [37]. Thus, the Sn-doped CeO2d nanorods potentially contain weak bonds which can contribute to the extraction of lattice oxygen, leaving the vacant sites. For CeO2 nanomaterials, both surface oxygen vacancies and bulk oxygen vacancies contribute to the catalytic performance. As the oxygen vacancy concentration decreases from the surface to the bulk [38], the oxygen defects induced by Sn dopants predominantly occur at the particle surface. To evaluate the catalytic performance, the Ce1xSnxO2d nanorods were employed to catalytically oxidise CO or photodecolourise methyl orange (MO) dye. As shown in Fig. 5a, the CO oxidation activity follows the order CeO2d < Ce0.95Sn0.05O2d < Ce0.90Sn0.10O2d. For CO oxidation catalysis, nanocrystals with a higher SSA usually exhibit a better conversion efficiency as the higher SSA provides more active sites for CO molecule adsorption. However, the SSA of the Ce1xSnxO2d nanorods decreases upon doping with Sn. Thus the contribution of increased oxygen vacancies [39] (as discussed in the Raman spectra and XPS analysis) appears to be a more dominant feature for improving CO catalytic properties. To support the CO catalytic activity, H2-TPR analyses (Fig. 5b) were
311
conducted to identify the contribution of the Sn dopant. The CeO2d nanorods exhibit two broad reduction peaks at around 450 °C and 800 °C, corresponding to surface/subsurface and bulk reduction, respectively. Upon doping with 10 mol% of Sn, three TPR peaks are observed which are similar to findings reported elsewhere [34]. The reduction peak of pure SnO2 is usually located at 680 °C [40] which derives from the Sn doping. This reduction peak indicates that more surface oxygen vacancies are generated at a lower temperature by an induced presence of Ce3+. The increased Ce3+ presence is attributed to the interaction between Ce4+/Ce3+ and Sn4+/Sn2+ via the redox equilibrium of 2Ce4+ + Sn2+ M 2Ce3+ + Sn4+. The two peaks above 500 °C for the Sn-doped ceria may be assigned to the reduction of Sn4+ formed in the CeO2 lattice and the bulk reduction of CeO2 [34]. The proposed mechanism for CO oxidation is shown in Fig. 7a. Oxygen vacancies in the defective Sn-doped nanorods can act as oxygen adsorption sites which can generate more active oxygen species [41]. These surface oxygen species actively facilitate the reaction of CO at the surface, thus improving the CO conversion efficiency. Fig. 6a shows the photocatalytic decolourisation of MO by the CeO2d nanorods with different dopant concentrations as a function of irradiation time. Without Sn doping, the photodecolourisation is low with only 27.2% of the MO being decolourised during 240 min of irradiation. However, on doping with Sn, the photocatalytic performance is improved. At a Sn dopant concentration of 10 mol%, the photocatalytic activity is significantly enhanced, with 70.5% of the MO being decolorized. Fig. 6b shows the plots of ln (C/C0) versus reaction time, according to the equation:
lnðC=C0 Þ ¼ kt where C0 is the initial concentration of the dye, C is the concentration of the dye at a certain time (t) and k is the apparent rate constant (min1). Generally, the k for dye decolourisation by a photocatalyst is used to assess the photocatalytic efficiency. As can be seen in Fig. 6b, the photocatalytic decolourisation rate can be fitted to follow pseudo-first-order kinetics. The highest rate constant (4.41 103 min1) of Ce0.90Sn0.10O2d is 2.4-fold and 3.3-fold higher than that of Ce0.95Sn0.05O2d and undoped CeO2d, respectively. It is known that the SSA of CeO2 nanocrystal-based phocatalysts generally plays a vital role in tuning the photocatalytic properties as the higher SSA of nanoparticles provides more active sites for the strong capture of dye molecules [42]. However, in this
Fig. 8. Schematic of FTO/Ce1xSnxO2d/Au device (a) and current-voltage (I-V) characteristics of FTO/Ce1xSnxO2d/Au device: CeO2d (b), Ce0.95Sn0.05O2d (c) and Ce0.90Sn0.10O2d (d). I-V characteristics of FTO/Ce1xSnxO2d/Au devices under different voltage ranges: 5 V 5 V (e), 10 V 10 V (f), 15 V 15 V (g) and 20 V 20 V (h). The sweeping direction is shown by the arrows.
312
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
work the SSA decreases upon introducing Sn dopant. In a similar vein to the CO catalytic results, the enhanced photocatalytic activity can be attributed to the increase in oxygen vacancies. The band gap, as estimated by the Tauc plot according to UV–Vis spectra, is shown in Fig. S2. CeO2d, Ce0.95Sn0.05O2d and Ce0.90Sn0.10O2d possess direct band gap values of 3.37, 3.33 and 3.26 eV, respectively. The shift in band gap can be ascribed to the increased oxygen vacancy concentration which can provide a red shift in the band gap [43]. The band edge position of the conduction band (CB) and valence band (VB) can be determined by the following equations: [44].
1 EVB ¼ v Ee þ Eg 2 ECB ¼ EVB Eg where v is the absolute electronegativity of the semiconductor (v = 5.57 eV for CeO2), Ee is the energy of free electrons on the hydrogen scale (4.5 eV) and Eg is the band gap [45]. The calculated VB and CB are 2.70 and 0.56 eV for Ce0.90Sn0.10O2d. Fig. 7b shows the proposed photocatalytic mechanism of the Sn-doped CeO2 nanorods. During UV irradiation, the photoelectrons in the valence band are excited to the CB and the same amount of holes in the VB are simultaneously generated (i.e. electron-hole pair formation). The photoinduced holes react with OH to form hydroxyl radicals (OH) while photogenerated electrons are trapped by oxygen vacancies as the surface traps [46], and further captured by oxygen molecules adsorbed at the oxygen-deficient sites to generate superoxide anions ( O 2 ). The MO is then oxidized by the resulting hydroxyl radicals and superoxide anions at the surface. Thus, the greater number of oxygen vacancies possessed by the Sn-doped CeO2 nanorods could facilitate the photocatalytic efficiency [47] and provide the enhanced photocatalytic performance. To further prove that the doping-generated oxygen vacancies predominantly contribute to the photocatalytic performance, the influence of UV irradation and a higher temperature on MO decolourisation are shown in Fig. S3. It can be seen that under UV irradation and at 120 °C Ce0.90Sn0.10O2d displays an optimum photocatalytic performance compared to under visible-light illumination or at room temperature. This phenomenon is attributed to the more active oxygen vacancies as oxygen vacancies can usually be activated by UV irradation [48] and their mobility can also be enhanced by a higher temperature [10,49]. Additionally, when compared with lattice vibration, the transportation mobility of lattice oxygen ions is more beneficial for separating photoinduced electrons and holes [50]. Consequently, oxygen vacancies are
believed to be predominantly responsible for the improved photocatalytic activity. In addition, to evaluate the photocatalytic stability, Ce0.90Sn0.10O2d nanorods were selected to deolourise MO over four consecutive cycles (Fig. 6c). It is apparent that the photocatalytic activity remains consistent across the four cyclic runs, indicating that the Sn-doped CeO2d possess excellent stability. It has been widely accepted that, in the instance of ceria based materials, oxygen vacancies can strongly influence their electrical properties as the distributed oxygen vacancies inside CeO2 films are usually crucial for modulating resistance [51] as well as can promote a resistive switching phenomenon especially when oxygen vacancies can form a filament as a conductive path [52]. The critical role of oxygen vacancies in defining the electrical properties of our fabricated un-doped and doped CeO2 nanostructures was evaluated by conducting current–voltage (I-V) measurements at various positive and negative potentials. To study the I-V characteristics, devices based on Ce1xSnxO2d (x = 0, 0.05 and 0.1) nanorods were fabricated using a drop-coating approach. Fig. 8a shows a schematic of a FTO/Ce1xSnxO2d/Au device with the asymmetric IV curves exhibited by each device arising from the asymmetric electrode configuration. As shown in Fig. 8b, the film comprising undoped CeO2d rods shows non-linear I-V characteristics, which can be considered as a typical non-linear resistor. In contrast, hysteresis loops are observed for the Sn-doped CeO2d devices (Fig. 8e–h). The behavioral change in I-V characteristics implys that charge carriers other than electrons are probably involved in driving conduction. Similar hysteresis loops have been seen in other materials [53–55] including thin/thick films and single crystals. Unlike previous studies where the resistance switching was a consequence of UV illumination [55], a similar phenomenon is present here, which instead arisies from the increased oxygen vacancy concentration due to the doping effect. In addition, the calculated resistance decreases by an order of magnitude, from 210 MX for the undoped CeO2d to 30–50 MX for the Ce0.95Sn0.05O2d and 10–20 MX for the Ce0.90Sn0.10O2d films. In the case of the neat material, the anionic defect concentration is not sufficiently high to provide a considerable RS effect even with the different voltage ranges. Upon introducing Sn, more oxygen vacancies are available to promote a considerable RS phenomenon as the loops become more obvious. By applying a positive potential, the vacancies begin to migrate towards the counter electrode and reduce the local resistance in the film (as shown in Fig. S4). The gradual increase in the current level of the devices also provides evidence on the increased amount of anionic defects (oxygen vacancies) [56]. It is noted that UV irradation has a significant influence on the oxygen vacancies within CeO2 as more oxygen vacancies can be
Fig. 9. Current-voltage characteristics of FTO/Ce1xSnxO2d/Au devices before and after UV irradation: (a) CeO2d and (b) Ce0.90Sn0.10O2d.
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
generated by UV irradation which are readily detected in UV Raman spectra [31,57]. Thus, I-V characteristics before and after UV irradation are shown in Fig. 9 to further demonstrate the contribution of oxygen vacancies in the Ce1xSnxO2d nanorods. From Fig. 9a, it can be seen clearly that the undoped device performs very differently following exposure to UV irradiation, as an apparent loop with a higher current is observed when compared with the pristine device. In the case of the 10 mol% Sn-doped CeO2d device (Fig. 9b), the current is also observed to increase two-fold compared to when operated without UV exposure. Raman spectra was also conducted for both conditions (Fig. S5) and the relative peak intensity or peak area ratio of peak580/F2g becomes larger after UV irradation, indicating more oxygen vacancies are generated. The findings demonstrate that oxygen vacancy concentrations to increase within the ceria nanostructure upon introducing the Sn dopant whereby the doped ceria nanostructures exhibit enhanced catalytic activities for dye degradation and CO oxidation as well as provide improved RS performances.
[8] [9]
[10]
[11]
[12] [13]
[14] [15]
4. Conclusions
[16]
In this work, one-dimensional Ce1xSnxO2d nanostructures were synthesized by a facile hydrothermal method with the interplay between Sn dopant and oxygen vacancies, and studied by means of catalytic and current-voltage (I-V) measurements for potential multifunctional applications. It was found that the introduction of Sn into the lattice only slightly altered the nanorod morphology although it invoked a lower aspect ratio and decreased specific surface area. Importantly, Sn doping promoted the amount of oxygen vacancies as evidenced by Raman and XPS spectra analyses. The increased active oxygen vacancies in the Sn-doped CeO2d nanorods facilitated both carbon monoxide catalytic oxidation and photocatalytic oxidation activities. In addition, obvious loops with resistance switching behaviour were observed in I-V curves of the Sn-doped CeO2d nanorods, further illustrating the higher oxygen vacancy concentration. The results indicate that modulating oxygen vacancies by compositional design can be utilized to achieve desirable multi-functionalities, especially for catalytic and memory applications.
[17]
Acknowledgement This work is funded by the Australian Research Council Project (grant no. FT140100032). The authors thank Ms. Katie Levick, Dr. Anne Rich and Dr. Bill Gong for assistance with TEM, Raman spectra and XPS measurements. H. Du thanks the China Scholarship Council (CSC) for financial support (No. 201406410060).
[18] [19]
[20]
[21]
[22] [23]
[24] [25] [26]
[27]
[28]
[29]
Appendix A. Supplementary material [30]
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2017.05.057. [31]
References [1] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chemistry and properties of nanocrystals of different shapes, Chem. Rev. 105 (2005) 1025–1102. [2] J. Pal, T. Pal, Faceted metal and metal oxide nanoparticles: design, fabrication and catalysis, Nanoscale 7 (2015) 14159–14190. [3] Y. Li, W. Shen, Morphology-dependent nanocatalysts: Rod-shaped oxides, Chem. Soc. Rev. 43 (2014) 1543–1574. [4] X. Wang, Z. Li, J. Shi, Y. Yu, One-dimensional titanium dioxide nanomaterials: nanowires, nanorods, and nanobelts, Chem. Rev. 114 (2014) 9346–9384. [5] E.W. McFarland, H. Metiu, Catalysis by doped oxides, Chem. Rev. 113 (2013) 4391–4427. [6] F. Zaera, Nanostructured materials for applications in heterogeneous catalysis, Chem. Soc. Rev. 42 (2013) 2746–2762. [7] J. Liqiang, Q. Yichun, W. Baiqi, L. Shudan, J. Baojiang, Y. Libin, F. Wei, F. Honggang, S. Jiazhong, Review of photoluminescence performance of nano-
[32]
[33]
[34]
[35]
[36]
313
sized semiconductor materials and its relationships with photocatalytic activity, Sol. Energ. Mat. Sol. Cells 90 (2006) 1773–1787. C.T. Campbell, C.H. Peden, Oxygen vacancies and catalysis on ceria surfaces, Science 309 (2005) 713–714. H. Du, Y. Wang, H. Arandiyan, A. Younis, J. Scott, B. Qu, T. Wan, X. Lin, J. Chen, D. Chu, Design and synthesis of CeO2 nanowire/MnO2 nanosheet heterogeneous structure for enhanced catalytic properties, Mater. Today Commun. 11 (2017) 103–111. T. Montini, M. Melchionna, M. Monai, P. Fornasiero, Fundamentals and catalytic applications of CeO2-based materials, Chem. Rev. 116 (2016) 5987– 6041. Y. Gao, R. Li, S. Chen, L. Luo, T. Cao, W. Huang, Morphology-dependent interplay of reduction behaviors, oxygen vacancies and hydroxyl reactivity of CeO2 nanocrystals, Phys. Chem. Chem. Phys. 17 (2015) 31862–31871. M. Nolan, S.C. Parker, G.W. Watson, The electronic structure of oxygen vacancy defects at the low index surfaces of ceria, Surf. Sci. 595 (2005) 223–232. H.-X. Mai, L.-D. Sun, Y.-W. Zhang, R. Si, W. Feng, H.-P. Zhang, H.-C. Liu, C.-H. Yan, Shape-selective synthesis and oxygen storage behavior of ceria nanopolyhedra, nanorods, and nanocubes, J. Phys. Chem. B 109 (2005) 24380–24385. Z. Wu, M. Li, S.H. Overbury, On the structure dependence of CO oxidation over CeO2 nanocrystals with well-defined surface planes, J. Catal. 285 (2012) 61–73. D.A. Andersson, S.I. Simak, N.V. Skorodumova, I.A. Abrikosov, B. Johansson, Optimization of ionic conductivity in doped ceria, Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 3518–3521. Y. Tang, H. Zhang, L. Cui, C. Ouyang, S. Shi, W. Tang, H. Li, J.-S. Lee, L. Chen, Firstprinciples investigation on redox properties of M-doped CeO2 (M= Mn, Pr, Sn, Zr), Phys. Rev. B 82 (2010) 125104. J. Ayastuy, A. Iglesias-González, M. Gutiérrez-Ortiz, Synthesis and characterization of low amount tin-doped ceria (CeXSn1-XO2-d) for catalytic CO oxidation, Chem. Eng. J. 244 (2014) 372–381. A. Sawa, Resistive switching in transition metal oxides, Mater. Today 11 (2008) 28–36. A.D. Liyanage, S.D. Perera, K. Tan, Y. Chabal, K.J. Balkus Jr, Synthesis, characterization, and photocatalytic activity of Y-doped CeO2 nanorods, ACS Catal. 4 (2014) 577–584. X. Liu, J. Ding, X. Lin, R. Gao, Z. Li, W.-L. Dai, Zr-doped CeO2 nanorods as versatile catalyst in the epoxidation of styrene with tert-butyl hydroperoxide as the oxidant, Appl. Catal. A 503 (2015) 117–123. C. Yang, X. Yu, S. Heißler, A. Nefedov, S. Colussi, J. Llorca, A. Trovarelli, Y. Wang, C. Wöll, Surface Faceting and Reconstruction of Ceria Nanoparticles, Angew. Chem. Int. Ed. 56 (2017) 375–379. C.P. Poole Jr, Encyclopedic dictionary of condensed matter physics, Academic Press, 2004. Y.H. Tan, J.A. Davis, K. Fujikawa, N.V. Ganesh, A.V. Demchenko, K.J. Stine, Surface area and pore size characteristics of nanoporous gold subjected to thermal, mechanical, or surface modification studied using gas adsorption isotherms, cyclic voltammetry, thermogravimetric analysis, and scanning electron microscopy, J. Mater. Chem. 22 (2012) 6733–6745. S. Tsunekawa, R. Sahara, Y. Kawazoe, K. Ishikawa, Lattice relaxation of monosize CeO2-x nanocrystalline particles, Appl. Surf. Sci. 152 (1999) 53–56. S. Tsunekawa, K. Ishikawa, Z.-Q. Li, Y. Kawazoe, A. Kasuya, Origin of anomalous lattice expansion in oxide nanoparticles, Phys. Rev. Lett. 85 (2000) 3440. S. Deshpande, S. Patil, S.V. Kuchibhatla, S. Seal, Size dependency variation in lattice parameter and valency states in nanocrystalline cerium oxide, Appl. Phys. Lett. 87 (2005). 133113 133113. J.E. Spanier, R.D. Robinson, F. Zhang, S.-W. Chan, I.P. Herman, Size-dependent properties of CeO2-y nanoparticles as studied by Raman scattering, Phys. Rev. B 64 (2001) 245407. X. Zhang, J. Wei, H. Yang, X. Liu, W. Liu, C. Zhang, Y. Yang, One-pot synthesis of Mn-doped CeO2 nanospheres for CO oxidation, Eur. J. Inorg. Chem. 2013 (2013) 4443–4449. Z. Zhang, D. Han, S. Wei, Y. Zhang, Determination of active site densities and mechanisms for soot combustion with O2 on Fe-doped CeO2 mixed oxides, J. Catal. 276 (2010) 16–23. P. Sudarsanam, B. Hillary, D.K. Deepa, M.H. Amin, B. Mallesham, B.M. Reddy, S. K. Bhargava, Highly efficient cerium dioxide nanocube-based catalysts for low temperature diesel soot oxidation: the cooperative effect of cerium-and cobalt-oxides, Cataly. Sci. Technol. 5 (2015) 3496–3500. M. Guo, J. Lu, Y. Wu, Y. Wang, M. Luo, UV and visible Raman studies of oxygen vacancies in rare-earth-doped ceria, Langmuir 27 (2011) 3872–3877. Y. Zhang, Z. Qin, G. Wang, H. Zhu, M. Dong, S. Li, Z. Wu, Z. Li, Z. Wu, J. Zhang, Catalytic performance of MnOx-NiO composite oxide in lean methane combustion at low temperature, Appl. Cataly. B 129 (2013) 172–181. C. Liang, Y. Shimizu, T. Sasaki, N. Koshizaki, Synthesis of ultrafine SnO2-x nanocrystals by pulsed laser-induced reactive quenching in liquid medium, J. Phys. Chem. B 107 (2003) 9220–9225. C. Li, G. Zeng, Y. Zhou, X. Zhang, The selective catalytic reduction of NO with NH3 over a novel Ce-Sn-Ti mixed oxides catalyst: promotional effect of SnO2, Appl. Surf. Sci. 342 (2015) 174–182. A. Gupta, M. Hegde, K. Priolkar, U. Waghmare, P. Sarode, S. Emura, Structural investigation of activated lattice oxygen in Ce1-xSnxO2 and Ce1-x-ySnxPdyO2-d by EXAFS and DFT calculation, Chem. Mater. 21 (2009) 5836–5847. G. Dutta, U.V. Waghmare, T. Baidya, M. Hegde, K. Priolkar, P. Sarode, Origin of enhanced reducibility/oxygen storage capacity of Ce1-xTixO2 compared to CeO2 or TiO2, Chem. Mater. 18 (2006) 3249–3256.
314
H. Du et al. / Journal of Colloid and Interface Science 504 (2017) 305–314
[37] A. Gupta, A. Kumar, M. Hegde, U. Waghmare, Structure of Ce1-xSnxO2 and its relation to oxygen storage property from first-principles analysis, J. Chem. Phys. 132 (2010) 194702. [38] S. Turner, S. Lazar, B. Freitag, R. Egoavil, J. Verbeeck, S. Put, Y. Strauven, G. Van Tendeloo, High resolution mapping of surface reduction in ceria nanoparticles, Nanoscale 3 (2011) 3385–3390. [39] Z.-Y. Pu, X.-S. Liu, A.-P. Jia, Y.-L. Xie, J.-Q. Lu, M.-F. Luo, Enhanced activity for CO oxidation over Pr-and Cu-doped CeO2 catalysts: effect of oxygen vacancies, J. Phys. Chem. C 112 (2008) 15045–15051. [40] X. Yao, C. Tang, Z. Ji, Y. Dai, Y. Cao, F. Gao, L. Dong, Y. Chen, Investigation of the physicochemical properties and catalytic activities of Ce0.67M0.33O2 (M= Zr4+, Ti4+, Sn4+) solid solutions for NO removal by CO, Cataly, Sci. Technol. 3 (2013) 688–698. [41] X. Liu, K. Zhou, L. Wang, B. Wang, Y. Li, Oxygen vacancy clusters promoting reducibility and activity of ceria nanorods, J. Am. Chem. Soc. 131 (2009) 3140– 3141. [42] J. Tian, Y. Sang, Z. Zhao, W. Zhou, D. Wang, X. Kang, H. Liu, J. Wang, S. Chen, H. Cai, Enhanced photocatalytic performances of CeO2/TiO2 nanobelt heterostructures, Small 9 (2013) 3864–3872. [43] B. Tatar, E. Sam, K. Kutlu, M. Ürgen, Synthesis and optical properties of CeO2 nanocrystalline films grown by pulsed electron beam deposition, J. Mater. Sci. 43 (2008) 5102. [44] Y. Xu, M.A. Schoonen, The absolute energy positions of conduction and valence bands of selected semiconducting minerals, Am. Mineral. 85 (2000) 543–556. [45] Z.-M. Yang, G.-F. Huang, W.-Q. Huang, J.-M. Wei, X.-G. Yan, Y.-Y. Liu, C. Jiao, Z. Wan, A. Pan, Novel Ag3PO4/CeO2 composite with high efficiency and stability for photocatalytic applications, J. Mater. Chem. A 2 (2014) 1750– 1756. [46] A.L. Linsebigler, G. Lu, J.T. Yates Jr, Photocatalysis on TiO2 surfaces: principles, mechanisms, and selected results, Chem. Rev. 95 (1995) 735–758.
[47] W. Lei, T. Zhang, L. Gu, P. Liu, J.A. Rodriguez, G. Liu, M. Liu, Surface-structure sensitivity of CeO2 nanocrystals in photocatalysis and enhancing the reactivity with nanogold, ACS Catal. 5 (2015) 4385–4393. [48] Z. Wu, M. Li, J. Howe, H.M. Meyer III, S.H. Overbury, Probing defect sites on CeO2 nanocrystals with well-defined surface planes by Raman spectroscopy and O2 adsorption, Langmuir 26 (2010) 16595–16606. [49] A. Trovarelli, Catalytic properties of ceria and CeO2-containing materials, Cataly. Rev. 38 (1996) 439–520. [50] Y. Li, Q. Sun, M. Kong, W. Shi, J. Huang, J. Tang, X. Zhao, Coupling oxygen ion conduction to photocatalysis in mesoporous nanorod-like ceria significantly improves photocatalytic efficiency, J. Phys. Chem. C 115 (2011) 14050–14057. [51] P. Gao, Z. Wang, W. Fu, Z. Liao, K. Liu, W. Wang, X. Bai, E. Wang, In situ TEM studies of oxygen vacancy migration for electrically induced resistance change effect in cerium oxides, Micron 41 (2010) 301–305. [52] X. Sun, B. Sun, L. Liu, N. Xu, X. Liu, R. Han, J. Kang, G. Xiong, T. Ma, Resistive switching in CeOx films for nonvolatile memory application, IEEE Electron Device Lett. 30 (2009) 334–336. [53] F. Nardi, S. Balatti, S. Larentis, D.C. Gilmer, D. Ielmini, Complementary switching in oxide-based bipolar resistive-switching random memory, IEEE Trans. Electron Dev. 60 (2013) 70–77. [54] T. Lee, K. Sung, T. Kim, J.-H. Ko, J. Jung, Observation of three crystalline layers in hydrothermally grown BiFeO3 thick films, J. Appl. Phys. 116 (2014) 194101. [55] J. Li, C. Ma, K. Jin, C. Ge, L. Gu, X. He, W. Zhou, Q. Zhang, H. Lu, G. Yang, Temperature-dependent resistance switching in SrTiO3, Appl. Phys. Lett. 108 (2016) 242901. [56] X. Tang, X. Zhu, J. Dai, J. Yang, L. Chen, Y. Sun, Evolution of the resistive switching in chemical solution deposited-derived BiFeO3 thin films with dwell time and annealing temperature, J. Appl. Phys. 113 (2013) 043706. [57] Y. Kang, Q. Leng, D. Guo, D. Yang, Y. Pu, C. Hu, Room-temperature magnetism of ceria nanocubes by inductively transferring electrons to Ce atoms from nearby oxygen vacancy, Nano-Micro Lett. 8 (2016) 13–19.