Titanium (IV) oxide composite hollow nanofibres with silver oxide outgrowth by combined sol–gel and electrospinning techniques and their potential applications in energy and environment

Abstract

A two-dimensional nanostructure composed of Ag₂O nanobranches attached to TiO₂ hollow nanofibres (denoted as 2D Ag₂O/TiO₂) was successfully prepared via sol–gel and coaxial electrospinning techniques. The Ag/TiO₂ hollow nanofibres were fabricated and calcined in ambient air at 500 °C. By calcination, the removal of organic materials and the formation of anatase TiO₂ were achieved with a well-retained hollow structure. The embedded Ag nanoparticles functioned as seeds for the Ag₂O outgrowth on the TiO₂ surface using a hydrothermal treatment at different times and temperatures, which caused the change in physical appearance, surface area, and electrical conductivity of 2D Ag₂O/TiO₂. A high quantity of Ag₂O nanobranches on TiO₂ nanofibres were obtained with increasing the temperature from 110 to 115 °C and the reaction time to 60 min. Consequently, the electrochemical active surface area (EASA) value was maximised to 65.25 cm² per cm² with an enhanced electrical conductivity of 91.3 ± 3.9 × 10⁻² S cm⁻¹. Further studies on the depth-profiles of Ag, Ti, and O revealed the presence of Ag₂O attached to the core structure of TiO₂. The photoelectrochemical and photocatalytic tests confirmed the excellent physical and electrochemical properties of 2D Ag₂O/TiO₂ (@115 °C, 60 min) for use in energy and environmental applications.

Graphical Abstract

1 Introduction

Titanium (IV) oxide or titania (TiO₂) is a wide bandgap n-type semiconductor material (energy band gap, E_g ≈ 3.2 eV) that has been extensively studied due to its high charge capacity, transparency, high reflective index, and chemical stability, with minimal toxicity and great availability at low cost. It has been widely used as an active material in several applications such as catalytic devices, gas sensors, batteries, solar cells, ceramic microporous membranes, and optoelectronic devices [1, 2]. In general, TiO₂ has many polymorphs including anatase, rutile, and brookite. Among these three polymorphs, anatase TiO₂ is the most electroactive structure and can be applied to optoelectronic and electronic devices. However, TiO₂ has short lifetime as the photogenerated electrons and holes recombine quickly leading to a reduction in its photocatalytic performance. Due to its broad bandgap, TiO₂ is only effective under ultraviolet (UV) radiation that occupies about 5% of the solar spectrum compared to visible light (45%). To overcome the constraints of TiO₂, the TiO₂ absorption spectrum needs to shift towards the visible region to fully harness the energy from sunlight as a cheap and renewable energy source. Different approaches have been developed to boost its photocatalytic activity and to extend the range of applications including incorporation of transition metal, non-metal or noble metal doping, surface and morphology modification, and heterojunction fabrication. Enhancing TiO₂ photocatalyst performance by doping or coupling with noble metals nanoparticles (Au, Ag, Pt, Pd) seems promising in solving the recombination of the electron–hole pair.

A variety of synthesis methods for anatase TiO₂-based materials have been developed. The main methods such as laser pyrolysis, microwave, reverse micelle, hydrothermal, chemical, or physical vapour deposition and aerosol process rely on severe synthesis conditions (i.e. high temperature and/or high pressure) and complex protocols [2]. By these processes, the formation of TiO₂-based materials is only limited to either nanocrystalline powder or films. An alternative synthesis pathway of anatase TiO₂-based materials is the sol–gel method [3] due to its simple operation, mild synthesis conditions, good chemical homogeneity, and high purity of the final product. Basically, a titanium precursor undergoes hydrolysis and condensation reactions to produce a liquid sol and then a solid gel, respectively. The conventional sol–gel synthesis can be employed with an electrospinning technique, thereby yielding nanofibrous structure of TiO₂-based materials with adjustable surface properties [4, 5]. This dramatically improves the electrochemical and catalytic activity of TiO₂-based materials [6].

Electrospinning is a simple and versatile method for fibre formation. It can be employed to produce continuous fibres with diameters ranging from tens of nanometres to a few micrometres using electrostatic forces [7]. It can be used to produce ultra-thin nanofibres as well as other morphologies for various materials such as polymers, ceramics, and composites. A typical electrospinning apparatus consists of a syringe, a grounded metal collector, and a high voltage power supply. A melt or blended polymer solution is filled into the syringe and is then extruded from the syringe nozzle under the influence of a strong electrostatic field. The force generated by an interaction between the charged polymer fluid and an external applied electric field induces the formation of a charged jet extruded from the tip of a Taylor cone. Then, it is deposited on the grounded collector. The resulting non-woven mat of ultrathin fibres provides high surface area with relatively small pore sizes. A typical spinnable precursor solution for fabricating TiO₂-based nanofibres comprises of four main components: Ti precursor, polymer, additives, and solvent [8]. The starting precursors have a pivotal impact on the morphology and structure of the resultant TiO₂ nanofibres. The important metal alkoxide or metal salt precursors of TiO₂ are titanium (IV) isopropoxide (Ti(Iso)), tetrabutyl titanate, and titanium tetrachloride. A carrier polymer is usually added to the solution to facilitate the fibre formation during electrospinning by adjusting the rheological properties of the spinning solution. The most commonly used polymers that are compatible with the Ti precursors are poly(vinyl pyrolidone), poly(vinyl acetate) (PVAc), poly(ethylene oxide), poly acrylonitrile, poly(methyl methacrylate), and polystyrene. The polymer of choice can be dissolved in a moderately volatile solvent such as hexafluoroisopropanol, trifluoroethanol, dichloromethane, chloroform, and N, N-dimethylformamide (DMF). An additive such as acetic acid or acetyl acetone is sometimes added to the mixture for different purposes: adjusting the solution properties, stabilising the precursors, and/or electrospinning jet. Moreover, metal nanoparticles can be mixed with the precursor solution to provide seed-assisted growth in metal–organic framework (MOF) nanostructures for use in energy and environment applications. Among various chemicals, the mixture of Ti(Iso) and PVAc dissolved in DMF is typically employed by several research works [9, 10] related to MOFs structured into nanofibers as a result of its excellent spinnability and stretchability at low cost.

Electrospun TiO₂-based nanofibres have been developed and show fascinating new properties and features for a wide range of potential applications. For lithium-ion applications, it has been reported that electrospun (pristine) TiO₂ nanofibres could be produced with a lower specific capacity than TiO₂ nanoparticles. On the other hand, the electrochemical cycling stability of TiO₂-based material was greatly improved by the incorporation of nanofibrous structure. It was reported that the capacity fading of TiO₂ nanofibres decreased from ∼ 63 (for TiO₂ nanoparticles) to ∼ 23% [11]. Metallic nanoparticles (e.g. Au, Ag, or Zn) were then incorporated into electrospun TiO₂ composite nanofibres to promote lithium-ion diffusibility and charge transfer processes. Hence, the specific capacity of metal nanoparticle-embedded TiO₂ nanofibre composites was increased by 20% [12]. Doping TiO₂ with Ag nanoparticles is one of the successful approaches to retard the recombination of photogenerated electron–hole pairs and to extend the TiO2 photoresponse to the visible light region [13]. The heterogeneous Ag/TiO₂ photocatalysts could combine multiple functions in one nanostructure: the high photocatalytic efficiency of the TiO₂ nanocrystals and the ability of Ag nanoparticles to shift the photocatalytic activity to the visible light. A variety of research studies also considered their excellent photocatalytic properties for environmental remediation such as photodegradation of organic pollutants like industrial dye pollutants [14, 15]. Zhao et al. [16] successfully prepared an Ag/TiO₂/graphene composite through a combination of sol–gel and solvothermal methods with improved superficial area and photocatalytic activity to decompose more than 79% of methylene blue (MB) compounds when compared to the TiO₂/graphene with lower removal efficiency of 53%. The Ag-doped TiO₂ displayed an enhanced photocatalytic activity for aqueous medium photocatalytic degradation of phthalic acid (100% removal efficiency) [17] and pentachlorophenol (80% removal efficiency) [18] in the presence of UV radiation for 90 min. However, most studies on Ag/TiO₂ nanocomposites rely on zero-dimensional (0D), e.g. nanosized particles, and one-dimensional (1D) nanostructures, e.g. nanotubes.

Several innovative electrospinning techniques have been applied to expand the function and properties of electrospun fibres. For instance, forced air-assisted electrospinning, air-gap electrospinning, and mixed electrospinning can be employed for this purpose. In addition to these techniques, coaxial electrospinning is a promising method that uses two concentrically aligned nozzles, enabling two polymer solutions to be co-electrospun with no mixing (i.e. shell and core materials should be immiscible). Selective removal of the cores from such coaxial fibres results in hollow fibres which are beneficial for surface-related applications such as electrochemical devices [19]. Several previous works reported over twofold increase in surface area of hollow nanofibres as compared to that of solid nanofibres [15, 20, 21]. In addition to the hollow structure, materials with dimensions outside of the nanoscale range (or known as two-dimensional (2D) nanostructures) tremendously increase active surface area with positive effects on their electrical and electrochemical properties. This occurs since the contact area between an electrode and an electrolyte is greatly improved in comparison with microscale materials. However, not much research has been conducted in 2D nanostructures due to the difficulty in synthesis and fabrication. At present, the most common 2D nanomaterials are made of boron nitride, graphene, and molybdenum disulphide which mainly rely on nanofabrication of natural moulding followed by the separation or extraction of the desired nanostructures [22, 23]. For example, a large quantity of solvent is usually needed in the fabrication of graphene from natural graphite, and the chemical vapour deposition method is expensive and relatively inefficient; hence, it is difficult to scale up for mass production. Therefore, coaxial electrospinning is a promising tool for preparing 2D nanomaterials with universal applicability, high production efficiency, and easy modification of product performance. However, the fabrication of electrospun 2D hollow nanofibres with metallic nanobranches has not been paid much attention. A few publications focused on nanosheets, nanoplates, and nanobelts [24,25,26,27], but not many on nanobranches [28, 29]. There is still room for improvement on the 2D structural fabrication and performance. To the best of our knowledge, no study has been reported on the Ag₂O/TiO₂ system with an incorporation of hollow and hierarchical nanobranch structures to enhance its electrochemical properties and photocatalytic activities. Easy fabrication of 2D nanomaterials is also a concern and one of the key challenges before they can be widely used. Consequently, surface modification and hydrothermal synthesis should be used to produce Ag₂O nanobranches attached to the surface of hollow TiO₂ nanofibres (denoted as 2D Ag₂O/TiO₂).

In the current study, 1D TiO₂ hollow nanofibres with Ag nanoparticles incorporated into their structure were initially fabricated by a combination of coaxial electrospinning and a sol–gel method. The effect of spinning conditions on morphological appearance of the resulting titania fibres was investigated. After that, a calcination process was carried out in air at 500 °C for 1 h to remove organic phases from the titania fibres and to cause a phase transformation. The silver nanoparticles embedded in the hollow 1D TiO₂ nanofibres increased their electrical conductivity with an additional function as nuclei for silver crystal outgrowth via a facile one-step hydrothermal synthesis with no requirements for surfactants or templates. Next, the hydrothermal reactions were performed in an aqueous medium at different reaction temperatures and times to crystallise Ag₂O directly onto the TiO₂ surfaces. Consequently, 2D Ag₂O/TiO₂ was successfully fabricated for the first time. The effects of hydrothermal treatment conditions on their morphological structures and properties were examined using an attenuated total reflectance Fourier transform-infrared spectroscopy (ATR/FT-IR), field-emission scanning electron microscopy with energy dispersive X-ray spectroscopy (FE-SEM/EDS), transmission electron microscopy (TEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), nitrogen adsorption–desorption, cyclic voltammetry, and a four-point probe electrical measurement. Finally, the photoelectrochemical and photocatalytic properties of the as-prepared 2D Ag₂O/TiO₂ samples were studied by electrochemical impedance spectroscopy (EIS), ultraviolet–visible (UV–Vis) spectroscopy, and photoluminescence (PL) analysis to verify their prospective applications in solar cells, semiconductor lasers, surface acoustic wave devices, and secondary batteries.

2 Materials and methods

2.1 Materials

Ti(Iso) (98.0%) and bis(hexamethylene)triamine (BHT, > 95.0%) were purchased from TCI Co., Ltd., Japan. N,N,N’,N’-tetramethyl-p-phenylenediamine (TMPD, 99%), PVAc (average MW ~500,000 by GPC), tetrabutylammonium perchlorate (TBAP, ≥ 95.0%), silver powder (particle size < 100 nm, 99.5% trace metals basis), and MB (92.6%) were purchased from Sigma Aldrich. Acetic acid (99.7%) and DMF (98.0%) were obtained from RCI Labscan Group, Thailand. Mineral oil (100%) was purchased from Ajax Finechem Pty Ltd. Silver nitrate (AgNO₃, 100%) was obtained from VWR International. All the chemicals were of analytical reagent grade and were used with no further purification. Distilled water was used in most experiments except for electrochemical and photoelectrochemical tests at which deionised water was employed to avoid ions in the water interfering with the signals.

2.2 Preparation of pre-calcined Ag-Ti(Iso)/PVAc fibres

Initially, 0.7 g of PVAc granules was dissolved in 4.6 ml of DMF solvent at room temperature under vigorous stirring for 2 h. Next, 5.2 ml of Ti(Iso) was mixed with the previously prepared PVAc solution for use as a spinnable polymer additive. After that, a few drops of acetic acid were added to the Ti(Iso)/PVAc mixture until the solution became transparent with a slight yellowish colour. This was followed by a vigorous stirring for 3 h to ensure a complete sol–gel reaction. Subsequently, 0.1 g of silver powder was suspended in the resulting Ti(Iso)/PVAc solution using a magnetic stirrer for 10 min to provide a metal colloidal Ti(Iso)/PVAc spinning suspension. Then, the spinning suspension was quickly loaded into the outer part of coaxial dual-capillary spinneret; mineral oil was used as the inner solution to produce core/shell nanofibres. Coaxial electrospinning was done with a blunt-ended 16-gauge and a blunt-ended 24-gauge stainless steel needles serving as a co-axial nozzle for a shell and core configuration. The co-axial nozzle was connected to a power supply (Gamma High-Voltage Research, Model D-ES30PN/M692). A range of electric field potentials between 15 and 17 kV were applied across the nozzle and the grounded aluminium foil collector, with a working distance of 15–17 cm to find the optimal electrospinning conditions. Spinning was uninterrupted for 6 h to provide a non-woven fibrous sheet made of electrospun Ag-Ti(Iso)/PVAc nanofibres.

2.3 Calcination process

The pre-calcined Ag-Ti(Iso)/PVAc fibrous material was initially dried in a convection oven at 80 °C for 48 h to remove the remaining solvent and to complete the hydrolysis reaction. Subsequently, the sample was calcined in air at 500 °C for 1 h with a heating rate of 5 °C min⁻¹. After calcination, the resulting electrospun material consisted of hollow 1D Ag/TiO₂ composite fibres.

2.4 Hydrothermal treatment

In the hydrothermal treatment, the calcined hollow 1D Ag/TiO₂ was treated with BHT and silver nitrate to allow outgrowth of Ag₂O. Briefly, 1.076 g of BHT and 0.849 g of silver nitrate were dissolved in 50 g of distilled water in two separate bottles. Then, both solutions were mixed, followed by adding 10 mg of the calcined hollow Ag/TiO₂ composite fibres. The suspension was continuously stirred for 4 h and finally placed in an autoclave to perform the hydrothermal reaction. The hydrothermal process was studied using two different controlled conditions: (i) a reaction for 60 min with a varied temperature of 110–120 °C and (ii) a reaction temperature of 115 °C with a varied reaction time of 30–60 min. Next, the resulting mixture was allowed to naturally cool to room temperature. The 2D TiO₂-based fibrous material was filtered off, subsequently rinsed several times with distilled water and then dried in a convection oven at 60 °C for 12 h before further characterisation.

2.5 Characterisation

Surface morphology of the pre-calcined hollow Ag-Ti(Iso)/PVAc composite fibres was examined using FE-SEM/EDS (JEOL JSM 5200) to establish the optimal conditions (i.e. applied voltage and working distance) in the electrospinning process. The functional groups of the pre-calcined and as-calcined samples were assessed using ATR/FT-IR (Thermo Nicolet Nexus 670). TEM (JEOL JEM 2010) was used to investigate the hollow structure with high-resolution images and to provide selected area electron diffraction (SAED) patterns of 1D and 2D TiO₂-based nanofibres.

Several analytical techniques were employed to study the effects of hydrothermal conditions on physical, electrochemical, and electrical properties of the as-prepared hollow 2D Ag₂O/TiO₂ composite fibres. Their crystal structures and crystallinity were investigated and then compared with the pre-calcined and the as-calcined composite nanofibres using XRD (PANalytical X’pert PRO) with Cu Kα (λ = 1.54056 Å) radiation over a range of 2θ angles from 20 to 80º. The surface area of 0D, 1D, and 2D electrospun samples was characterised using two approaches including a Brunauer–Emmett–Teller (BET) method and an electrochemical method. BET measurements of specific surface area (S_BET) and pore size distribution were performed using nitrogen adsorption and desorption isotherms (Autosorb-1C) at 77 K. The samples were degassed under inert nitrogen at 200 °C for 12 h prior to BET measurements. The BET measurements of each electrospun TiO₂-based sample were conducted three times, and the average S_BET value was reported. It was notable that the variation of S_BET values was very small in each sample. Electrochemical active surface area (EASA) together with electrochemical activity was evaluated in a mixture of 1 mM TMPD and 0.1 M TBAP in acetonitrile using cyclic voltammetry at various scan rates from 25 to 400 mV s⁻¹. Electrochemical studies were carried out using a three-electrode system consisting of a platinum (Pt) rod, Ag/AgCl, and one of the as-prepared TiO₂-based samples as counter, reference, and working electrodes, respectively.

The XPS results were obtained using an AXIS Ultra DLD instrument of Kratos with monochromatic Al K radiation (15 mA, 15 kV). It operated in the lens-defined mode swept for 3 times, at the pass energy of 20 eV, the maximum X-ray source power of 150 W, and the energy step size of 0.1 eV. All the binding energies were calibrated relative to the C 1 s peak at 284.6 eV. The fibre samples were hot-mounted using an epoxy resin at a curing temperature of 100 °C. A vacuum impregnation was needed to ensure that the hollow and nanobranch structures connected to the surface were fully filled with the resin. To perform in-depth profile analysis, all the mounted specimens were bombarded with argon ions operating at 3 keV for 10 min by a Hitachi IM-4000 ion milling machine. The current density used in the argon ion beam was ~150 μA cm⁻² rastered over a 2 × 2 mm area. The sputter rate was found to be ~2.5 nm min⁻¹ as calibrated by a reference silicon oxide thin film on silicon substrate. After each etching cycle (10 min per cycle), the XPS spectra were collected for all elements assigned in the survey spectra including titanium (Ti 2p), oxygen (O 1 s), and silver (Ag 3d). The surface etching was performed until either the etch depth of 250 μm or the core structure was reached.

The electric conductivity of each sample was measured at room temperature by a Keithley electrometer (Model 2400, USA) in a collinear four-point probe technique. The fibre samples were pressed into 1-mm thick pellets with a diameter of ~ 1.1 cm. The pellets were placed on a stage of a Pro-4 probing station (LUCAS LABS) with an installation of an SP4 4-point probe head. For each type of sample, six replicates were performed with three different readings (a sampling interval of 1 s). The average thickness of each sample was used to calculate the bulk/volume conductivity (σ) determined at five different positions using a digital thickness gauge (Mitutoyo).

A photoelectrochemical test was done using a combination of EIS and UV illumination. The EIS measurements were performed in the three-electrode system as mentioned above. A 0.5 M Na₂SO₄ aqueous solution was used as the electrolyte. The photocatalytic activity was then measured under UV light irradiation generated by an 8-W low-pressure mercury lamp (253.7 nm) with an intensity of 11 mW cm⁻², while a sinusoidal AC excitation voltage with a 10 mV amplitude was applied to the electrochemical system at frequencies scanned from 10⁴ to 0.01 Hz. Measurements under short circuit conditions (i.e. the potential between working and counter electrodes was zero) provided an easily obtainable and reproducible impedance for comparison for all samples.

The PL analysis was performed at room temperature by a JASCO spectrofluorometer (FP-8500) with an excitation wavelength of 325 nm. The photocatalytic measurement was also carried out based on photocatalytic degradation reactions of MB. 40 mg of fibre samples was dispersed in 100 ml of 4%w MB solution in a double jacket cylindrical reactor with water circulation to maintain the ambient temperature for the reaction. The mixture was kept in a dark place for 15 min to reach an adsorption/desorption equilibrium. Next, the suspension was exposed to the UV light irradiation. Aliquots of the suspension were collected and filtered through a syringe filter (pore size ~0.45 μm) at an interval of 15 min. The MB decolourisation was measured immediately by a Spectroquant^® Pharo 300 spectrophotometer. Apart from the removal efficiency, a preliminary evaluation of the photocatalytic stability of 2D Ag₂O/TiO₂ (@115 °C, 60 min) was undertaken in 15 illumination cycles. The duration of UV light irradiation was controlled at 90 min for each cycle. The surface morphology of the recycled 2D Ag₂O/TiO₂ (@115 °C, 60 min) was examined by SEM after 15 cycles of photocatalytic stability test.

3 Results and discussion

3.1 Spinning conditions of pre-calcined hollow Ag-Ti(Iso)/PVAc fibres

An inorganic hollow fibrous structure was fabricated from a mixture of sol–gel precursors and a polymer binder in an organic solvent using a combination of sol–gel and electrospinning processes. The viscous polymer binder played a significant role in increasing the spinnability of inorganic nanofibres by increasing the solution viscosity. This significantly affects morphology of the fibrous material. The effect of the applied potential across the syringe needle and the collector plate as well as the working distance between them is investigated using SEM and TEM as demonstrated in Figs. 1 and 2. It was found that pre-calcined hollow Ag-Ti(Iso)/PVAc composite fibres were successfully produced in this work via a coaxial electrospinning technique. The appropriate spinning conditions included an operating voltage of 15 kV and a working distance of 17 cm. A uniform and continuous hollow fibre structure with no bead formation was fabricated with an average diameter of 870 ± 10 nm. The morphology of pre-calcined hollow Ag-Ti(Iso)/PVAc composite fibres was not considerably affected by the presence of silver powder in the spinning solution despite the appearance of relatively rough surfaces. It was also found that fibrous Ag-Ti(Iso)/PVAc non-woven mats with an average thickness of 70 ± 5 µm were obtained after 6 h of electrospinning.

Fig. 1

FE-SEM images (magnification = 5000X ; scale bar = 10.0 µm) of pre-calcined hollow Ag-Ti(Iso)/PVAc composite fibres under different applied voltages of 15–20 kV and working distances of 15–17 cm

Fig. 2

TEM images (magnification = 4000X ; scale bar = 500 nm) of pre-calcined hollow Ag-Ti(Iso)/PVAc composite fibres under different applied voltages of A 15, B 17, and C 20 kV at a working distance of 17 cm

3.2 Properties of as-calcined hollow titania composite fibres

After calcination, the as-prepared hollow Ag-Ti(Iso)/PVAc nanofibres were converted to hollow Ag/TiO₂ composite fibres (denoted as 1D Ag/TiO₂). The morphology, cross section, and hollow structure of the as-calcined 1D Ag/TiO₂ fibres are examined using FE-SEM and TEM as illustrated in Fig. 3. The hollow Ag/TiO₂ composite fibres turned to a grey colour during calcination. In the cross-sectional FE-SEM and TEM images, hollow structured composite fibres were observed. This was attributed to elimination of the mineral oil core. It was also found that the removal of organic contents including the PVAc matrix and mineral oil from the Ag-Ti(Iso)/PVAc composite nanofibres resulted in a decreased average fibre diameter from 870 ± 10 to 705 ± 10 nm with core diameters ranging from 343 ± 7 and 477 ± 12 nm.

Fig. 3

The morphology, cross-sectional and hollow structure of the as-calcined hollow Ag/TiO₂ fibres are revealed by A low magnification SEM, B cross-sectional SEM images and high-resolution TEM (inset)

In addition to FE-SEM and TEM analysis, IR spectroscopic analysis was employed to verify the removal of organic contents after calcination. Figure 4 depicts the FTIR spectra of PVAc observed at ~3400, 2920, 1740, 1430–1340, and 1025 cm⁻¹. These peaks are assigned to the O–H stretching vibration of the hydroxy group, CH₂ asymmetric stretching vibration, C = O carbonyl stretch, C–H bending/deformation vibration of CH₂, and C–O stretching of acetyl groups, respectively. The FTIR spectra of the pre-calcined composite fibres reveal that the adsorption bands in the range of 1000–1750 cm⁻¹ (corresponding to the bending and stretching vibration of PVAc) nearly vanish in the calcined samples. All calcined fibres contained a characteristic band located at a wavenumber of approximately 470 cm⁻¹, which is associated with the O-Ti–O bonding of the anatase phase. The completeness of calcination (removal of the organic part) is confirmed by thermogravimetric analysis (TGA) as shown in Supplement Information S1.

Fig. 4

FTIR spectra of a pre-calcined solid Ti(Iso) fibres, b as-calcined solid TiO₂ fibres, c pre-calcined hollow Ti(Iso) fibres, d as-calcined hollow TiO₂ fibres, e pre-calcined hollow Ag/Ti(Iso) fibres, and f as-calcined hollow Ag/TiO₂ fibres

Elemental analysis of the hollow Ag/TiO₂ composite fibres is performed using EDS as shown in Fig. 5a. According to the EDS results, the presence of elemental oxygen in the composite fibres implies that the titanium transition metal was successfully oxidised to TiO₂ during calcination. It has been reported that the silver seed metal could be converted to various multivalent Ag–O forms such as Ag₂O, AgO, Ag₃O₄, and Ag₂O₃ [12, 13]. These Ag oxides have different crystalline structures that can be identified using XRD as detailed in Section. 3.3.1. The crystal structures and crystal defects of the calcined hollow fibres were evaluated from their SAED patterns (Fig. 5b, c) using TEM. Several sets of diffraction spots observed in all calcined samples indicated that the calcined hollow TiO₂ fibres and hollow Ag/TiO₂ composite fibres had the polycrystalline nature of anatase TiO₂ according to JCPDS Card No. 21–1272. The brightness and intensity of the polymorphic ring in the composite hollow fibres were weaker than those of calcined hollow (pure) TiO₂ fibres. This indicates that an introduction of Ag seed metal to TiO₂ determined the growth orientation of TiO₂, causing a change from highly crystalline TiO₂ fibres to partially crystalline and amorphous Ag/TiO₂ composite fibres.

Fig. 5

Elemental analysis of a hollow Ag/TiO₂ composite fibres using EDS and SAED patterns of b the as-calcined hollow TiO₂ fibres compared with c the as-calcined hollow Ag/TiO₂ composite fibres

3.3 Fabrication of hollow titanium (IV) oxide composite fibres with Ag–O outgrowth structures

The formation of Ag–O outgrowth structures is known to be affected by variations in reaction time and temperature in hydrothermal synthesis. In this work, the effect of hydrothermal conditions was studied.

3.3.1 Phase structure of hollow composite fibres

The XRD patterns of hydrothermally treated hollow Ag–O/TiO₂ composite fibres are demonstrated in Fig. 6 which are also compared with the pre-calcined Ag-Ti(Iso)/PVAc fibres and the as-calcined hollow Ag/TiO₂ fibres. The hydrothermal processes were carried out under two different controlled conditions. Firstly, the reaction time was 60 min with a variation of temperature such as 110, 115, and 120 °C. Secondly, a constant reaction temperature of 115 °C was employed with a variation of reaction time of 30, 45, and 60 min. The corresponding crystal plane orientations are given above the peaks in this figure. The pre-calcined Ag-Ti(Iso)/PVAc nanofibres (Fig. 6a) revealed a prominent peak at approximately 19° 2θ due to the superior crystalline properties of high molecular weight PVAc. This result confirmed that the PVAc nanofibres loaded with metal seed particles and Ti(Iso) could retain their crystalline structure even under the application of strong electrostatic fields during electrospinning. As observed in Fig. 6b, the calcined hollow Ag/TiO₂ composite fibres contain strong anatase peaks (JCPDS Card No. 21–1272) of TiO₂ at 2θ degree values of 25.211º, 37.771º, 38.427º, 47.947º, 53.874º, 54.993º, 62.105º, 62.666º, 68.752º, 70.191º, 75.208º, and 76.039º, corresponding to the crystal planes of (101), (004), (112), (200), (105), (211), (213), (204), (116), (220), (215), and (301), respectively. No evidence of the rutile TiO₂ structure was found in the XRD patterns. This ensures that the calcined hollow Ag/TiO₂ composite fibres were composed of only the anatase phase, which is the most electroactive in optoelectronic and electronic devices. Figure 6b shows the Ag₂O peaks at 2θ degree values of 37.991º, 44.140º, 64.372º, and 77.301º corresponding to the crystal planes of (111), (200), (220), and (311), respectively. The weak Ag₂O peaks indicated that Ag₂O was in the form of well-dispersed nanoparticles after calcination.

Fig. 6

XRD patterns of pre-calcined Ag-Ti(Iso)/PVAc fibres and calcined hollow Ag/TiO₂ composite fibres compared with hydrothermally treated hollow 2D Ag–O/TiO₂ composite fibres using two different controlled conditions: at a fixed reaction time of 60 min with a varied temperature of 110–120 °C and at a fixed reaction temperature of 115 °C with a varied time of 30–60 min

In accordance with the XRD patterns of the hydrothermally treated hollow 2D Ag–O/TiO₂ composite fibres in Fig. 6c–g, all peaks indexed as TiO₂ crystals are weakened. Alternatively, the characteristic diffraction peaks of silver (3C silver with a cubic crystal structure, JCPDS Card No. 41–1104 as detailed in Supplement Information S2) emerged. The crystallite sizes of Ag₂O obtained after hydrothermal treatment at various temperatures and reaction times are calculated from the Scherrer’s equation with regard to the (111) plane of Ag₂O as listed in Table 1. The intensity of the Ag₂O peaks was found to increase with hydrothermal temperature from 110 to 120 °C. The crystallite size of Ag₂O hence increased significantly with the reaction temperature. This was contrary to the effect of reaction time at which the Ag₂O crystallite size was relatively unchanged by the hydrothermal reaction after 60 min.

Table 1 Crystallite sizes of Ag₂O outgrowths calculated from the Scherrer’s equation with regard to the (111) plane of Ag₂O and the change in physical appearance, surface area, EASA, and electrical conductivity of various TiO₂-based nanofibres together with the reported values from relevant literature

The formation of Ag₂O outgrowth during the hydrothermal process could be explained using the following mechanisms. The as-calcined hollow Ag/TiO₂ fibres were treated with BHT in an aqueous medium to generate hydroxide ions ${\mathrm{OH}}^{-}$ as shown in Eq. (1).

$$\begin{aligned}&{\mathrm{NH}}_2{\left({\mathrm{CH}}_2\right)}_6\mathrm{NH}{\left({\mathrm{CH}}_2\right)}_6{\mathrm{NH}}_2+{2\mathrm H}_2\mathrm O\\&\quad\leftrightarrow\mathrm{NH}_3^+{\left({\mathrm{CH}}_2\right)}_6\mathrm{NH}{\left({\mathrm{CH}}_2\right)}_6\mathrm{NH}_3^++{2\mathrm{OH}}^-\end{aligned}$$

(1)

The hydroxide ions then reacted with silver ions ${\mathrm{Ag}}^{+}$ to form hydroxyl complexes ([AgOH]) followed by a transformation into solid-phase Ag₂O as shown in Eq. (2).

$$2\left[\mathrm{AgOH}\right]\rightarrow{\mathrm{Ag}}_2\mathrm O+{\mathrm H}_2\mathrm O$$

(2)

The overall reaction can be written as:

$$\begin{aligned}&{2\mathrm{AgNO}}_3+{\mathrm{NH}}_2{\left({\mathrm{CH}}_2\right)}_6\mathrm{NH}{\left({\mathrm{CH}}_2\right)}_6{\mathrm{NH}}_2+{\mathrm H}_2\mathrm O\rightarrow{\mathrm{Ag}}_2\mathrm O\\&\quad+{\mathrm{NO}}_3{\mathrm{NH}}_3{\left({\mathrm{CH}}_2\right)}_6\mathrm{NH}{\left({\mathrm{CH}}_2\right)}_6{\mathrm{NH}}_3{\mathrm{NO}}_3\end{aligned}$$

(3)

3.3.2 Nanostructures of hydrothermally treated hollow Ag2O/TiO2 composite fibres

The effect of hydrothermal temperature and reaction time on the Ag₂O outgrowth structure Is evaluated based on surface and cross-sectional FE-SEM images and their corresponding SAED patterns as shown in Fig. 7. Generally, the FE-SEM images of hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres depict abundant Ag₂O crystals attached to hollow TiO₂ (anatase) nanofibres. The hollow 1D nanostructure of TiO₂ nanofibres was not adversely affected by the hydrothermal processes. The variation of hydrothermal temperatures had a significant effect on particle morphology and particle size distribution of Ag₂O outgrowth, along with the size and weight of the resulting nanofibres. At 110 °C, cubic-like Ag₂O crystals were uniformly coated on the outer surface of hollow TiO₂ nanofibres. Rapid growth of Ag₂O occurred at higher temperatures of 115 and 120 °C, resulting in several particle morphologies of the Ag₂O crystals with particle sizes of ~320 nm for the cubes, ~450 nm for the truncated cubes, ~ 540 nm for the rhombicuboctahedra, and ~ 850 nm for the edge-truncated octahedra. According to the particle size distribution of Ag₂O nanobranches (see the insets of Fig. 7b, e, h, k and m), the distributions became broader and moved towards larger particle sizes when the hydrothermal temperature was increased from 110 to 120 °C. Thus, an increased outer diameter and weight of hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres were obtained with elevated temperature. The longer reaction time had similar effects on particle size growth but was much less dominant in comparison to hydrothermal temperature.

Fig. 7

Cross-sectional and surface FE-SEM images and the corresponding SAED patterns of hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres using various reaction temperatures of a–c 110, d–f 115, and g–i 120 °C with a controlled reaction time of 60 min. The insets are histograms of particle size distribution for the Ag₂O outgrowth. At a reaction temperature of 115 °C, changes in Ag₂O/TiO₂ nanostructure with reaction times of j–k 30 and l–m 45 min are demonstrated

The SAED patterns of the hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres using various reaction temperatures are illustrated in Fig. 7c, f, and i. All the patterns contained several bright spots situated around the diffuse rings, representing polycrystalline structures. The transformation of SAED patterns into broader line profiles with elevated ageing temperature, from 110 to 120 °C, implies a higher degree of disorder in the polycrystalline structure with the presence of many domains and multiple orientations. Still, the crystallinity of Ag₂O nanobranches was maximised at 120 °C due to a decrease in crystallisation activation energy, considering the intensity of the (111) and (200) planes of Ag₂O in the XRD results.

3.4 Physical and electrochemical properties of hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres

Anatase-phase TiO₂ has attracted much research interest for energy and environmental applications since its band gap of approximately 3.2 eV is similar to the energy of light at a wavelength of 388 nm. In solar cells, direct exposure of anatase-phase TiO₂ to sunlight can excite an electron in its bound state to its free state so that the electron can participate in conduction. The lack of an electron in the conductive band creates positive holes in the valence band which are highly active in environmental remediation activities. The photovoltaic efficiency, photocatalytic activity, and electrochemical activity of anatase-phase TiO₂ in energy and environmental applications are strongly dependent on its morphology and surface area. Hence, surface area of the resulting TiO₂ nanostructures was examined using two methods, i.e. BET and electrochemical methods. Typical BET diagrams and the corresponding pore size distribution curves are demonstrated in Fig. 8. The adsorption isotherms of all TiO₂ samples were type IV in the IUPAC classification scheme with H2 hysteresis loops in the range of 0.3–1 P/P_o, generated by capillary condensation of the adsorbate within mesopores of the samples. The shift of hysteresis loops to higher relative pressures when approaching 1 P/P_o was correlated with an increased average pore size (D_p) as well as the broader range of D_p. The pore size distribution of Ag-TiO₂ composite nanofibres (curves a–f in Fig. 8) was wider than that of hollow 1D TiO₂ nanofibres (curve g in Fig. 8) and solid 0D TiO₂ nanofibres (curve h in Fig. 8). As listed in Table 1, the variation in the average D_p from 3.34 to 5.89 nm arises from metal incorporation and especially from hydrothermal treatment. According to FE-SEM results, several particle morphologies and sizes of the Ag₂O nanobranches were produced at higher temperatures of 115 and 120 °C. These combinations enlarged the size of mesopores formed in the hollow 2D Ag₂O/TiO₂ composite fibres.

Fig. 8

i Nitrogen adsorption–desorption isotherms and ii the corresponding pore size distribution curves of hydrothermally treated hollow 2D Ag₂O/TiO₂ composite fibres using various reaction temperatures and times compared with the as-calcined hollow 1D Ag/TiO₂ composite nanofibres, hollow 1D TiO₂ nanofibres and solid 0D TiO₂ nanofibres

The S_BET values (in Table 1) were derived from the adsorption isotherms. The results suggested that the morphological structure was of great significance in enhancing S_BET from 30.81 m² g⁻¹ in the solid 0D TiO₂ fibres to 39.07 m² g⁻¹ in the hollow 1D TiO₂ fibres. The small decrease in S_BET from 39.07 m² g⁻¹ in the hollow 1D TiO₂ fibres to 36.72 m² g⁻¹ in the hollow 1D Ag/TiO₂ fibres was attributed to partial pore blockage with the addition of Ag nanoparticles, which were then converted to Ag₂O after calcination. The 2D nanostructure was comprised of Ag₂O nanobranches attached to hollow TiO₂ nanofibres having a higher surface area of up to 44.96 m² g⁻¹ in comparison with the 0D and 1D nanostructures. In consideration of the influence of hydrothermal conditions, the S_BET values of hollow 2D Ag₂O/TiO₂ composite fibres were strongly affected by reaction temperature since these values decreased significantly as reaction temperature was increased from 110 to 120 °C. This was supported by the particle size distribution of Ag₂O nanobranches, which became broader and moved towards larger particle sizes with elevated temperatures. Lengthening reaction time from 30 to 60 min had a minimal effect on the morphological structures and S_BET values of hollow 2D Ag₂O/TiO₂ composite fibres, which were almost identical in shape and size.

An electrochemical technique is a useful tool to determine the electrochemical activity of the resulting TiO₂-based samples in terms of EASA. As detailed in the experimental section, electrochemical characterisation was undertaken using cyclic voltammetry of the well-studied one-electron oxidation of TMPD [30]. The inset in Fig. 9a illustrates a quasi-reversible one-electron redox couple of TMPD/TMPD⁺. However, Faradaic reactions can be either chemically reversible or irreversible, depending on the electrode material and the reactants involved. For the Ag₂O containing samples including 2D Ag₂O/TiO₂ and (pure) Ag₂O particles in Fig. 9a, the peak ratios between cathodic and anodic peaks are near unity indicating that the redox reaction of TMPD/TMPD⁺ couple was stable at the electrode interface under the test conditions. The deterioration of cathodic peak response could be observed in those 1D and 0D TiO₂-based samples at which their electrode surface was mostly composed of TiO₂. The smaller and broader cathodic peaks were possibly caused by the existence of follow-up unfavourable chemical reactions and surface adsorption after the first electron transfer, leading to fewer species available for the cathodic scans. This means that the novel branched nanostructure of hollow 2D Ag₂O/TiO₂ composite fibres was beneficial to improve charge transfer, yielding a completely quasi-reversible redox reaction for TMPD. However, the forward reaction was more favourable.

Fig. 9

a–c Representative cyclic voltammograms measured at a set scan rate of 200 mV s⁻¹ and d a plot of peak current and ν^1/2 extracted from cyclic voltammograms at various scan rates according to the Randles–Sevcik equation of solid 0D TiO₂, hollow 1D TiO₂, hollow 1D Ag/TiO₂, hollow 2D Ag₂O/TiO₂ nanofibres using various reaction temperatures of 110–120 °C or various reaction times of 30–60 min in a solution of 1 mM TMPD and 0.1 M TBAP

The cyclic voltammograms (see Fig. 9a–c) depict a variation in the peak current density of TiO₂-based samples with different nanostructures, reaction temperatures, and reaction times. In Fig. 9a, the peak potential for one-electron TMPD oxidation shifts from ̴ 530 mV (for both 0D and 1D pristine TiO₂ nanofibres) to the lower values of ̴ 420–450 mV in the Ag₂O-containing samples due to the change in the nature of electrode surface from TiO₂ to Ag₂O. Not only that, the increase in peak current also confirmed that the introduction of highly conductive Ag₂O nanoparticles and nanobranches enabled rapid electron transfer in the hollow 1D and 2D Ag₂O/TiO₂ composite fibres. This was attributed to the higher electrical conductivity of pristine Ag₂O (~6.9 × 10⁻¹ S cm⁻¹ [31]) compared to Ti₂O (i.e. 5.5 × 10⁻⁷ [32] or 58.0 ± 3.1 × 10⁻⁸ S cm⁻¹ based on the conductivity test). Figure 9b and c show similar profiles of cyclic voltammograms obtained from all 2D Ag₂O/TiO₂ samples, indicating that there was no significant deterioration of the electrochemical response resulting from hydrothermal reaction conditions. When increasing the hydrothermal temperature from 110 to 120 °C, the peak current was found to increase significantly. It is contrary that the lower S_BET values were obtained due to the aggregation of Ag₂O nanobranches on the TiO₂ surface which enlarged the crystallite size of Ag₂O (see also Table 1). It is known that the magnitude of current generated depends strongly on electrochemically active surface area, enabling higher reaction sites for reactants or a larger storage capacity for electrons per unit volume of electrodes. As a result, a higher current density can be generated. It was believed that the balance between the degree of surface coverage by highly conductive Ag₂O (or Ag₂O content) and total electrode surface area was achieved by using the optimal hydrothermal conditions at 115 °C and 60 min. At higher temperature of 120 °C, the TiO₂-based fibres were mostly covered by the excessive amount of Ag₂O outgrowth. The electron transfer in this case was mainly driven by the highly conductive Ag₂O outgrowth but limited by the reduced specific surface area and large fibre diameters. Hence, the current density was slightly decreased.

The EASA value (given in Table 1) was determined from cyclic voltammograms of each TiO₂-based sample recorded at different scan rates according to the Randles–Sevcik equation:

$${I}_{p}=\left({26.87\times 10}^{4}\right){n}^{3/2}{v}^{1/2}{D}^{1/2}Ac$$

(4)

where i_p is the peak current of the cyclic voltammograms, n is the number of electrons appearing in the half-cell reaction of TMPD (i.e. n = 1), c is the analyte concentration (1 × 10^–6 mol cm⁻³), A is the EASA (cm²), and D is the analyte diffusion coefficient (5 × 10⁻⁵ cm² s⁻¹ [30]). The value i_p/v^1/2 is the slope obtained from a plot of the peak current values and square roots of the scan rates, as shown in Fig. 9d. The linear relationship observed between i_p and v^1/2 indicates that the electrochemical system followed the Randles–Sevcik relation.

The hollow structure of 1D TiO₂ and 1D Ag/TiO₂ nanofibres led to a twofold increase in the EASA values compared to the solid 0D TiO₂ fibres. It was found that Ag nanoparticles embedded in the hollow 1D Ag/TiO₂ nanofibres caused a five orders-of-magnitude improved electrical conductivity (σ) over the pristine hollow 1D TiO₂ fibres. However, their electrochemical activity was relatively similar which suggested that the Ag nanoparticles were only present under the surface of TiO₂ fibres. Consequently, the Ag nanoparticles acted as a conductive filler that facilitated electron transfer within the TiO₂ nanofibre matrix but did not participate directly in the charge transfer processes. It was apparent that the presence of Ag₂O nanobranches in the 2D Ag₂O/TiO₂ samples provided a large number of high-conductive pathways and shorter diffusion routes for the reactant to undergo the redox reaction at the electrode surface. Therefore, their EASA values were approximately two times higher than the EASA value of 1D Ag/TiO₂. A large quantity of Ag₂O nanobranches on TiO₂ nanofibres was attained with elevated hydrothermal temperatures from 110 to 115 °C and prolonged hydrothermal treatment from 30 to 60 min. This optimised the EASA value to 65.25 cm² per cm² of geometric surface area of TiO₂-based samples with an enhanced electrical conductivity of 91.3 ± 3.9 × 10⁻² S cm⁻¹. The extensive Ag₂O outgrowth synthesised at a hydrothermal temperature of 115 °C showed the greatest electrical conductivity of 26.1 ± 1.5 × 10⁻¹ S cm⁻¹ at which the electron transfer was mainly accommodated by the highly conductive Ag₂O. In contrast, its EASA significantly decreased which was associated with the large particle sizes of Ag₂O nanobranches along with the corresponding lowest S_BET.

In summary, hollow 2D Ag₂O/TiO₂ composite fibres should be one of the promising novel nanomaterials for energy applications when compared their physical and electrochemical properties with most relevant research works on electrospun Ag-Ti composite nanofibres. There is scarce information on Ag₂O/TiO₂ composite hollow fibres. For example, the study on Ag₂O/TiO₂ nanotubes [33] focused only on the structural fabrication and their photocatalytic properties. Electrospun solid Ag₂O/TiO₂ fibres proposed by Liu et al. [34] had relatively low S_BET of 13.20 m² g⁻¹ and large average Dp of 14.30 nm as the Ag₂O was present on TiO₂ solid nanofibres in the form of metal oxide nanoparticles. The study also reported that the aggregation of Ag₂O nanoparticles at high Ag₂O loading exhibited the lowest S_BET of 2.2 m² g⁻¹. Despite being incorporated with highly conductive Ag, the TiO₂-based nanofibres containing Ag dopant as a trace had low electrical conductivity of approximately 4.51 × 10⁻³ S cm⁻¹ [35]; the lower value of 1.99 × 10⁻⁵ S cm⁻¹ [36] was attained with 1D mesoporous Ag nanoparticle-embedded TiO₂ nanofibres. The higher S_BET and electrical conductivity observed in the present work on 2D Ag₂O/TiO₂ (Table 1) confirmed the benefits of combining Ag₂O nanobranches with hollow TiO₂-based structure.

Figure 10a shows the XPS survey scanned on the surface of hollow 2D Ag₂O/TiO₂ samples. All the survey scans revealed the similar atomic peaks of titanium (Ti 3p, Ti 3 s, Ti 2p, Ti 2 s), carbon (C 1 s), silver (Ag 3d), and oxygen (O 1 s) on the sample surface. Except for C impurity from the instruments and the environment, all the elements were part of the hollow 2D Ag₂O/TiO₂. The binding energies of 464.6 and 458.8 eV verified the bonding of Ti to O as can be seen in the high-resolution XPS spectrum of Ti 2p (Fig. 10b). When compared to those of pure TiO₂ (i.e. 459.33 and 465.28 eV, respectively [37]), the slight decrease in the binding energies for Ti 2p_3/2 and Ti 2p_1/2 implied that one of the oxygen atoms in the TiO₂ lattice was partially substituted by Ag and thus the excited electrons in the TiO₂ were possibly transferred to Ag [38]. The Ag 3d spectrum in Fig. 10c is comprised of two peaks at 367.5 and 373.5 eV corresponding to the core level of Ag 3d_3/2 and Ag 3d_5/2, respectively, in the hollow 2D Ag₂O/TiO₂ samples. These peaks confirmed the existence of the monovalent Ag⁺ state in Ag₂O compound at which the deconvoluted peak of Ag 3d_5/2 in Ag₂O compound is typically located at 367.8 eV [39]. This agreed well with the XRD results and those reported in literature for a typical TiO₂ [37] and Ag₂O [39]. Moreover, there was no metallic Ag as the binding energy of Ag⁰ at ~ 368.2 eV [39] could not be observed.

Fig. 10

XPS spectra on the surface of hollow 2D Ag₂O/TiO₂ samples: a survey, b Ti 2p, c Ag 3d together with the XPS depth profiles for d Ti 2p, e Ag 3d, and f O 1 s

Further studies on XPS depth profiles of Ag, Ti, and O atomic compositions were performed to investigate the distribution of Ag in the hollow structure by etching the sample surface using low energy Ar ions for 10 min repeatedly. The variation in chemical composition with XPS depth profiles of Ag, Ti, and O obtained from the XPS survey scans for the five 2D Ag₂O/TiO₂ samples is shown in Fig. 10d–f. The highest level of the O element was present on the sample surface associated with the oxygen richness in both Ag₂O nanobranches and TiO₂ core structure. The quantities of Ag and O elements were found to decrease from the sample surface to the etch depth of ~100–150 μm and then significantly increase to some extent. This observation implied the presence of Ag₂O attached to the core structure of TiO₂. Accordingly, the Ti concentration increased with the etch depth and then slightly dropped when approaching the core surface.

3.5 Evaluation of photoelectrochemical and photocatalytic properties

A potential application of TiO₂-based samples for photocatalysis was assessed through EIS measurements with and without UV illumination (photoelectrochemical test). The EIS results are presented in complex impedance plane plots (also known as Nyquist plots) as illustrated in Fig. 11a and b. A single semicircle was obtained in all samples as expected for polycrystalline semiconductor materials with relatively low electronic conductivity. The impedance spectra are fitted well with the proposed equivalent circuit model as depicted in the inset of Fig. 11a. The model is comprised of two elements connected in series, a solution resistance (R_s) and a typical resistor–capacitor circuit (R_ct-CPE). The charge transfer resistance R_ct and a constant phase element CPE are connected in parallel which refer to the electronic conductivity of the electrode material (i.e. TiO₂-based samples) and the behaviour of a double layer that mimics an imperfect capacitor, respectively. CPEs are commonly used instead of pure capacitors to model the non-ideal capacitance in the double layer located at the solid/electrolyte interface. The inhomogeneities in the surface of metal oxide nanofibres resulted in the observed non-ideal behaviour due to the surface roughness and irregularities in surface termination, porosity, and complexity in the double layer structure. The solution resistance is often read from the high frequency intercept with the real impedance axis (near the plot origin). It was found that all samples had similar solution resistance of approximately 1.5 kOhm cm² because an identical electrolyte system was employed. The arc diameter with respect to the real part (X-axis) gave an estimate of the charge transfer resistance at the interface between a particular TiO₂-based sample and a Pt counter electrode. In general, the charge transfer was enhanced with the presence of Ag or Ag₂O in the hollow TiO₂ nanofibres. The great improvement was obtained with the hollow 2D Ag₂O/TiO₂ composite fibres having high EASA. The diagrams in Fig. 11b show much smaller semicircles upon UV illumination when compared to those of the dark conditions (Fig. 11a). A two-fold decrease in the charge transfer resistance for all hollow 2D Ag₂O/TiO₂ composite fibres was also consistent with their excellent electrochemical activities in the presence of Ag₂O nanobranches. The electrode activity was promoted even further with the use of a higher hydrothermal temperature. This can imply that the electrode morphology of hollow 2D Ag₂O/TiO₂ composite fibres along with its phase purity and EASA is promising, thus demonstrating its potential utilisation in dye-synthetised solar cells, adsorbents, and photocatalysts.

Fig. 11

Photoelectrochemical and photocatalytic properties of electrospun TiO₂-based samples including EIS Nyquist plots obtained from photoelectrochemical tests with a an absence or b a presence of UV illumination, c room temperature PL spectra, d photocatalytic degradation efficiency between C_t/C₀ and the irradiation time, e 15 cycles of photocatalytic stability test under UV irradiation for 90 min, and f the representative SEM image of the recycled 2D Ag₂O/TiO₂ photocatalyst after 15th cycle

The PL emission measurement was carried out in order to investigate the electron–hole recombination of the TiO₂-based samples as photocatalysts. Figure 11c shows similar patterns of PL spectra in hollow 1D TiO₂ and 1D Ag/TiO₂ which are composed of two main peaks including 358–360 nm and 467–468 nm. The peaks observed at 358–360 nm were raised from the recombination or sub-band gaps induced by oxygen vacancies. The small emission peaks located at 467–468 nm were obtained in all samples associated with the surface oxygen vacancies when the Ti³⁺ charge was transferred to the oxygen vacancies in TiO₆⁸⁻ octahedra [40]. In general, the recombination rate of electrons and holes is found to be lower as a result of an enhanced separation efficiency; this leads to weak PL intensity. It can be seen from Fig. 11c that the sequence of the PL intensities is 2D Ag₂O/TiO₂@115 °C, 60 min < 2D Ag₂O/TiO₂@115 °C, 45 min < 2D Ag₂O/TiO₂@115 °C, 30 min < 2D Ag₂O/TiO₂@110 °C, 60 min < 2D Ag₂O/TiO₂@120 °C, 60 min < 1D Ag/TiO₂ < 1D TiO₂. Accordingly, the Ag₂O outgrowth on the TiO₂ surface exhibited a lower electron–hole recombination rate and higher photocatalytic activity when compared to the pristine TiO₂. The PL intensity decreased gradually with increasing the Ag₂O content to serve as trapping centres for electrons, corresponding to the elevated hydrothermal temperature from 110 to 115 °C and the longer reaction time from 30 to 60 min. However, the excessive quantity of Ag₂O nanobranches in 2D Ag₂O/TiO₂@120 °C, 60 min could decrease the photocatalytic activity of TiO₂ (as seen from its highest PL intensity) because the trapping sites would also act as recombination sites for electrons and holes.

Since the PL results were an indirect tool to study the photocatalytic performance of the 2D Ag₂O/TiO₂, the photocatalytic activity was further proved by the photocatalytic degradation of MB (see Fig. 11d) as a model organic pollutant. It was found that the hollow 2D Ag₂O/TiO₂ composite fibres exhibited the superior activity than the pristine hollow 1D TiO₂ fibres and the photolysis under UV illumination. The as-prepared 2D Ag₂O/TiO₂ composite fibres (4 mg) could decompose 97.57% MB (100 ml, 4%w) in 90 min which had greater photocatalytic property when compared with other previous works. For example, Ag/TiO₂ nanocomposites composed of Ag nanoparticles deposited on anatase TiO₂ clusters could only degrade about 58% MB (10 mg of catalyst in 10 ml solution containing 30 µM or ~0.001%w MB) in 90 min [41] or 94.21% MB (39.47 mg of catalyst in 100 ml solution containing 0.001%w MB) in 80 min [42]. Electrospun TiO₂ nanofibres with Ag nanoparticles developed by Raffi et al. [43] had the decolourisation efficiency of up to 90.5% in 30 min with the 50-ml test solution containing 25 mg of catalyst and 0.1%w MB. It is well-known that MB can be photocatalytically oxidised in the presence of TiO₂ under illumination. The photogenerated electrons from the conduction band of TiO₂ can reduce the pre-adsorbed O₂ at the electrode surface to O₂¯. Subsequently, the photoinduced valence-band holes at the TiO₂ surface have an energy suitable to drive H₂O oxidation, yielding OH radicals that are the active species in photocatalytic MB degradation. Based on this study, the rate of MB degradation was greatly increased with the combination of Ag₂O nanobranches and TiO₂ hollow structure in the hollow 2D Ag₂O/TiO₂ composite fibres. The increase in the reaction rate for MB oxidation was additionally caused by an interfacial electron transfer from excited MB species to Ag₂O. Thus, it is believed that the rate and degree of photocatalytic activity in the Ag₂O/TiO₂ composite fibres are strongly dependent on the TiO₂ content and the active surface area of Ag₂O available for electron transfer. This finding agreed with the excellent physical and electrochemical properties of hollow 2D Ag₂O/TiO₂ composite fibres observed earlier in this work.

Besides their efficiency, the preliminary evaluation of the photocatalytic stability of 2D Ag₂O/TiO₂ (@115 °C, 60 min) was conducted to guarantee its functionality. Figure 11e shows the removal efficiency (%) of MB solution using an UV illumination of 90 min in each cycle. Over 15 cycles of the photocatalytic stability test, no significant loss in the photocatalytic activity of 2D Ag₂O/TiO₂ (@115 °C, 60 min) is observed. The SEM image in Fig. 11f reveals that the recycled photocatalyst could maintain their morphology and nanobranch structure after use. Additionally, there were some closed ends of hollow nanofibres occurring from the catalyst utilisation but no adverse effect on its efficiency in photocatalytic degradation of MB. However, further studies and characterisation on the long-term photocatalytic stability (> 100 cycles) are still needed to verify the life span of this novel nanostructure material as an efficient and stable photocatalyst, adsorbent, or electrocatalyst.

4 Conclusions

Hollow 2D Ag₂O/TiO₂ composite fibres were successfully fabricated using a coaxial electrospinning technique employing a Ag-Ti(Iso)/PVAc colloidal solution. The as-prepared Ag-Ti(Iso)/PVAc mats were subsequently calcined at 500 °C for 1 h to remove PVAc polymer and mineral oil located at the core. FTIR and EDS elemental analysis were employed to verify the removal of the organic contents and TiO₂ formation after calcination. The calcined hollow 1D Ag/TiO₂ composite fibres retained their uniformity with an average outer diameter of ~0.7 µm and a wall thickness of ~ 0.15 µm. The silver nanoparticles embedded in the hollow 1D Ag/TiO₂ nanofibres acted as seeds to provide nucleation sites for facilitating crystal growth on TiO₂ surface during hydrothermal treatment. Ag₂O nanobranches presented cubic-like outgrowths attached to the hollow TiO₂ (anatase) nanofibres according to the natural crystal structure of Ag₂O as assessed by XRD, XPS, FE-SEM, and TEM. The increased fibre size and weight was affected by the reaction temperature to a greater degree than reaction time. The photoelectrochemical and photocatalytic tests confirmed that the hollow 2D Ag₂O/TiO₂ composite fibres were applicable for energy and environmental applications. This was due to the excellent physical and electrochemical properties of 2D Ag₂O/TiO₂ observed using cross-sectional and surface FE-SEM images, SAED patterns, BET, electrochemical methods, and a collinear four-point probe technique. A high quantity of Ag₂O nanobranches on TiO₂ nanofibres was obtained with increasing the hydrothermal temperature from 110 to 115 °C and the reaction time from 30 to 60 min. This was found to maximise the EASA value to 65.25 cm² per cm² of geometric surface area of TiO₂-based samples with an enhanced electrical conductivity of 91.3 ± 3.9 × 10⁻² S cm⁻¹.