- Email: [email protected]
Electrospun anatase-phase TiO2 nanofibers with different morphological structures and specific surface areas
Journal of Colloid and Interface Science 398 (2013) 103–111
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Electrospun anatase-phase TiO2 nanofibers with different morphological structures and specific surface areas Guangfei He a, Yibing Cai b,c,⇑, Yong Zhao b, Xiaoxu Wang d, Chuilin Lai a, Min Xi a, Zhengtao Zhu a,b,d, Hao Fong a,b,d,⇑ a
Program of Materials Engineering and Science, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA Department of Chemistry and Applied Biological Sciences, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA Key Laboratory of Eco-Textiles, Ministry of Education, Jiangnan University, Wuxi, Jiangsu 214122, China d Program of Nanoscience and Nanoengineering, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA b c
a r t i c l e
i n f o
Article history: Received 29 December 2012 Accepted 4 February 2013 Available online 21 February 2013 Keywords: Titania Electrospinning Nanofiber Morphological structure Specific surface area
a b s t r a c t Electrospun anatase-phase TiO2 nanofibers with desired morphological structure and relatively high specific surface area are expected to outperform other nanostructures (e.g., powder and film) of TiO2 for various applications (particularly dye-sensitized solar cell and photo-catalysis). In this study, systematic investigations were carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures (e.g., solid, hollow/tubular, and porous) and specific surface areas. The TiO2 nanofibers were generally prepared via electrospinning of precursor nanofibers followed by pyrolysis at 500 °C. For making hollow/tubular TiO2 nanofibers, the technique of co-axial electrospinning was utilized; while for making porous TiO2 nanofibers, the etching treatment in NaOH aqueous solution was adopted. The results indicated that the hollow/tubular TiO2 nanofibers (with diameters of 300–500 nm and wall-thickness in the range from tens of nanometers to 200 nm) had the BET specific surface area of 27.3 m2/g, which was approximately twice as that of the solid TiO2 nanofibers (15.2 m2/g) with diameters of 200–300 nm and lengths of at least tens of microns. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers had the BET specific surface area of 106.5 m2/g, whereas porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers had the highest BET specific surface area of 148.6 m2/g. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction The anatase-phase titanium (IV) oxide (TiO2) has attracted extensive research interests in the recent years, because it has a wide range of applications such as environmental remediation, electronics, sensor technology, solar cell, and photo-catalysis [1–5]. The ceramic of TiO2 has several crystalline phases including anatase and rutile, and the band-gap of anatase-phase TiO2 is 3.2 eV, which is equivalent to the energy of light with the wavelength of 388 nm. Hence, the anatase-phase TiO2 can acquire energy either directly from sunlight or indirectly from light sensitizers (e.g., dyes), causing electrons to be excited to conduction band and concomitantly creating positive holes in valence band. This is termed as charge separation in solar cell, and the
⇑ Corresponding authors. Addresses: Key Laboratory of Eco-Textiles, Ministry of Education, Jiangnan University, Wuxi, Jiangsu 214122, China. Fax: +86 510 8591 2009 (Y. Cai), Program of Materials Engineering and Science, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA. Fax: +1 605 394 1232 (H. Fong). E-mail addresses: [email protected] (Y. Cai), [email protected] (H. Fong). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.02.009
separated charges can be turned into electricity. Additionally, the anatase-phase TiO2 can also be applied to the remediation of a variety of organic compounds and heavy metal ions (e.g., Pt4+, Pd2+, and Cr3+) as well as the destruction of micro-organisms such as bacteria, viruses, and molds from aqueous environments; this is fundamentally owing to the oxidizing capability of positive holes in the valence band. It is noteworthy that many applications of anatase-phase TiO2 are strongly influenced by its morphological structure and specific surface area. The materials-processing technique of electrospinning utilizes the electric force to drive the spinning process and to produce the fibers with diameters typically in the range from tens of nanometers to microns (commonly known as-electrospun nanofibers). The applications of electrospun nanofibers include, but not limited to, composites, filtrations/separations, biomedical applications (e.g., wound dressing, tissue engineering, and drug delivery), electronic applications (e.g., sensors, transistors, and detectors), and energy-related applications (e.g., solar cells, fuel cells, batteries, and supercapacitors) [6–8]. Unlike nanowires, nanorods, and nanotubes that are prepared by bottom-up synthetic methods and usually require further purifications, electrospun nanofibers are
104
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
produced via a top-down manufacturing process; they are inexpensive, continuous, and also relatively easy to align, assemble, and process into applications [6–10]. The electrospun TiO2 nanofibers (with diameters being hundreds of nanometers, and consisting of anatase-phase TiO2 crystallites with sizes of 10 nm), particularly the porous ones with relatively high specific surface area, are expected to outperform other nanostructures (e.g., powder and film) of TiO2 for the applications in solar cells and photocatalyses. In the recent decade, there have been numerous research endeavors devoted on preparations, structures, and properties of electrospun TiO2 nanofibers; in general, these nanofibers are made by electrospinning spin dopes containing TiO2 precursors (e.g., titanium (IV) n-butoxide) followed by pyrolyzing the as-electrospun precursor nanofibers at high temperature [11–17]. Through judiciously adjusting/controlling the properties of spin dopes, as well as the processing parameters during electrospinning and the subsequent pyrolysis, the electrospun TiO2 nanofibers with various morphological structures have been prepared. For example, Xia’s group fabricated TiO2 nanofibers with solid, core-sheath, hollow, and porous structures via the modified electrospinning techniques (e.g., the coaxial electrospinning technique) [11,18–20]; Cheng et al. prepared hollow TiO2 nanofibers (i.e., TiO2 nanotubes) via tailoring the composition of spin dope and adjusting the heating rate during pyrolysis [21]. Additionally, electrospun polymer (e.g., polylactide) nanofibers could be surface-coated with amorphous TiO2 via sol–gel method; upon removal of the thermally degradable polymer, the hollow TiO2 fibers/tubes would be acquired [22]; this suggested that different inorganic/ceramic tubular nanostructures with varied diameter and wall-thickness could be prepared by using electrospun polymer nanofibers as templates. Thereafter, Kim et al. made sub-micron tubes of anatase-phase TiO2 by a template-directed method [23]; in specific, electrospun polyvinylpyrrolidone nanofibers were used as the template for surface-coating of TiO2 using the technique of atomic layer deposition. Zhan et al. fabricated long hollow TiO2 fibers/tubes with mesoporous walls via the combination of a sol–gel method and the coaxial electrospinning technique by using a triblock copolymer as the poredirecting/creating agent, and these hollow fibers/tubes exhibited higher photo-catalytic activities toward degradation of Methylene Blue and formaldehyde than other nanostructured TiO2 materials such as commercial TiO2 nanoparticles (P25, Degussa) and mesoporous TiO2 films [24]. Furthermore, the mesoporous TiO2/SiO2 composite nanofibers have also been prepared, and these nanofibers showed selective photo-catalytic activities on the decompositions of Methylene Blue, Active Yellow, and Disperse Red [25]. In another example, Kanjwal et al. fabricated ZnO-doped TiO2 nanofibers via electrospinning, and the introduced ZnO–TiO2 hierarchical nanostructure could eliminate the dye of Methyl Red within 90 min and the dye of Rhodamine B within 105 min [16]; they also prepared electrospun TiO2 nanofibers with silver nanoparticles, and the photo-catalytic performance was very high [26]. It is also noteworthy that Li et al. [27] made porous TiO2 nanofibers by alkali-dissolution of SiO2 from TiO2/SiO2 composite nanofibers to improve the surface-to-volume ratio, and the resulting nanofibers exhibited superior photo-catalytic activity (the best photo-catalytic efficiency obtained in the study was 77% upon irradiation under the simulated sunlight for 1 h). Additionally, Liu et al. [28] reported that the electrospinning process with side-by-side dual spinnerets could be a simple approach for fabrication of bi-component TiO2/SnO2 nanofibers with controlled hetero-junctions, and such a structure would lead to an increase in the charge separation of the photo-generated electrons and holes within the bi-component system, allowing both electrons and holes to participate in the overall photo-catalytic reaction (i.e., the bi-component TiO2/ SnO2 nanofiber mat would be an excellent photo-catalytic system).
In another example, Stengl et al. [29] made the tungsten-doped TiO2 nanofibers by thermally hydrolyzing the aqueous solutions of peroxo complexes containing titanium and tungsten, and these nanofibers exhibited fast reaction rate toward photo-degradation of Orange II dye. It is well-known that the photovoltaic efficiency and photo-catalytic activity are strongly dependent on the specific surface area and morphological structure of anatase-phase TiO2. The electrospun mats consisting of overlaid anatase-phase TiO2 nanofibers possess the following advantageous properties for those applications: (1) high specific surface area (particularly if the nanofibers are porous), (2) controllable pore sizes among the nanofibers (ranging from tens to hundreds of nanometers), and (3) the thickness of a nanofiber mat can be readily tailored/controlled. Unlike nanoscale TiO2 particles/rods which are in loose granular form, electrospun TiO2 nanofibers are well-contained in the mat. The high specific surface areas of electrospun TiO2 nanofiber mats (particularly the ones consisting of porous TiO2 nanofibers) lead to a large number of reaction sites which would enhance the photovoltaic efficiency and photo-catalytic activity; additionally, the nanofiber mat would also result in great accessibility for reactants during the applications. In this study, systematic investigations have been carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures of solid, hollow/tubular, and porous. The objectives were (1) to reveal the correlations between the preparation methods and the morphological structures of electrospun TiO2 nanofibers and (2) to fabricate the TiO2 nanofibers with increased specific surface areas for future photovoltaic and photo-catalytic studies. Specifically, the solid TiO2 nanofibers were made by electrospinning the spin dope consisting of titanium (IV) n-butoxide (TNBT) and polyvinylpyrrolidone (PVP) in ethanol with the TNBT/PVP mass ratio being 10/1 followed by pyrolysis at 500 °C. The hollow/tubular TiO2 nanofibers were prepared via the technique of coaxial electrospinning; the spin dope for making the sheath component (of as-electrospun precursor nanofibers) consisted of TNBT and PVP in ethanol, while the spin dope for making the core component was paraffin oil. To prepare porous TiO2 nanofibers, two approaches were explored: (1) to prepare the composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., aluminum isopropoxide); after the TiO2/ Al2O3 composite nanofibers were made, the Al2O3 component would then be removed via chemical etching with NaOH aqueous solution, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers and (2) to prepare the mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm); the hypothesis was that, during the sol–gel process of TNBT and the subsequent pyrolysis, the existence of ZnO nanoparticles would hinder the volumetric contraction/shrinkage, resulting in the formation of nanoscale pores. Moreover, the ZnO component in TiO2/ZnO composite nanofibers could also be removed via chemical etching treatment, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The morphological and crystalline structures as well as specific surface areas of the solid, hollow/tubular, and two types of porous TiO2 nanofibers were characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), and a Brunauer-Emmett-Teller (BET) surface area analyzer. The acquired results indicated that all of electrospun TiO2 nanofibers had diameters in hundreds of nanometers and consisted of anatase-phase TiO2 crystallites with sizes of 10 nm; as compared to the solid TiO2 nanofibers with BET surface area of 15.2 m2/g, the hollow/tubular TiO2 nanofibers had the BET surface area of 27.3 m2/g, whereas the BET surface areas of porous TiO2 nanofibers were 106.5 m2/g and 148.6 m2/g upon the preparation methods.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
105
then mixed together followed by addition of 2.5 g TNBT. The final mixture was magnetically stirred for 24 h prior to electrospinning.
2. Experimental 2.1. Materials Titanium (IV) n-butoxide (TNBT), aluminum isopropoxide (Al[OCH(CH3)2]3), zinc oxide (ZnO, with particle sizes smaller than 30 nm), polyvinylpyrrolidone (PVP, Mw = 1,300,000), N,N-dimethylformamide (DMF), isopropanol (IPA), ethanol, paraffin oil, acetic acid (HAc), and sodium hydroxide (NaOH) were purchased from the Sigma–Aldrich Co. (St. Louis, MO) and used without further purification.
2.2.4. Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers The optimal spin dope (determined from the properties of aselectrospun nanofibers, ZnO/TiO2 composite nanofibers, and final TiO2 nanofibers) with the molar ratio of Ti/Zn being 7/3 was prepared by using the following procedure: PVP (2 g), TNBT (1.59 g), and ZnO (0.41 g) were mixed together with DMF (8 g) in a sealed container followed by being magnetically stirred for 24 h.
2.2. Preparation of spin dopes
2.3. Electrospinning
During the preparations of the following spin dopes, a trace amount (1 wt.%) of HAc was added in each spin dope to control the hydrolysis/gelation of TNBT.
2.3.1. Solid and porous TiO2 nanofibers A solution was filled in a 30 ml plastic syringe having a bluntend stainless-steel needle with the inside diameter of 0.3 mm. The electrospinning setup included a high voltage power supply (Model No.: ES30P), purchased from the Gamma High Voltage Research, Inc. (Ormond Beach, FL), and a nanofiber collector of electrically grounded aluminum foil that covered a laboratory-produced roller with diameter of 25 cm; the rotating speed was set at 100 rpm. The collector was placed at 25 cm below the tip of needle. During electrospinning, a positive high voltage of 20 kV was applied to the needle, and the solution feeding rates of 0.6 ml/h for solid TiO2 nanofibers, 2.0 ml/h for porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers, and 1.0 ml/h for porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers were maintained using a syringe pump (Model No.: KDS 200) purchased from the KD Scientific Inc. (Holliston, MA). The electrospinning was carried out in open environment at room temperature of 25 °C. The electrospinning setup with single (i.e., non-coaxial) spinneret for solid and/or porous TiO2 nanofibers is shown in Fig. 1.
2.2.1. Solid TiO2 nanofibers The spin dope for making solid TiO2 nanofibers was prepared by dissolving 0.3 g PVP and 3 g TNBT in 5 ml ethanol. The spin dope was magnetically stirred at room temperature for 24 h prior to electrospinning. 2.2.2. Hollow/tubular TiO2 nanofibers For making hollow/tubular TiO2 nanofibers, the spin dope for making solid TiO2 nanofibers was placed in a syringe connected to the outer capillary of a coaxial spinneret; while paraffin oil was placed in another syringe connected to the inner capillary. 2.2.3. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers The optimal spin dope (determined from the properties of aselectrospun nanofibers, Al2O3/TiO2 composite nanofibers, and final TiO2 nanofibers) with the molar ratio of Ti/Al being 8/2 was prepared by using the following procedure: (1) 1.4 g PVP was dissolved in 11.75 g IPA in a sealed container via magnetically stirring for 0.5 h; (2) 0.375 g Al[OCH(CH3)2]3 was dissolved in 7.125 g IPA in another sealed container via ultrasonication for 15 min; (3) the PVP/IPA and Al[OCH(CH3)2]3/IPA solutions were
2.3.2. Hollow/tubular TiO2 nanofibers The setup with coaxial spinneret for electrospinning of hollow/ tubular TiO2 nanofibers is also shown in Fig. 1. The coaxial electrospinning is similar to the conventional electrospinning except for the use of a spinneret containing two coaxially arranged
Single spinneret
Coaxial spinneret Fig. 1. Schematic illustration of the setup for electrospinning nanofibers.
106
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
capillaries. The coaxial spinneret used in this study was purchased from the Nisco Engineering AG in Switzerland; the inner capillary had the diameter of 0.3 mm, while the outer capillary had the diameter of 0.8 mm. During coaxial electrospinning, two spin dopes were fed into the inner (core) and outer (sheath) capillaries, respectively. A coaxial jet/filament would be generated between the spinneret and the collector, when a high voltage was applied, and the jet/filament would subsequently be stretched by electrostatic force to generate core-sheath nanofibers [6]. In this study, the feeding rates for the inner spin dope (paraffin oil) and outer spin dope (TNBT and PVP in ethanol) were set at 0.6 ml/h and 0.8 ml/h, respectively. A high voltage of 15 kV was applied to the coaxial spinneret during the electrospinning process, and a laboratory-produced roller (with diameter of 25 cm) covered by electrically grounded aluminum foil was operating at 100 rpm as nanofiber collector. The electrospinning was carried out at room temperature of 25 °C in open environment, and the as-electrospun nanofibers (in the form of overlaid nanofiber mat) were kept in ambient condition for 48 h to allow TNBT (and Al[OCH(CH3)2]3) to completely hydrolyze into three dimensional network (i.e., gel) by the moisture in atmosphere. 2.4. Pyrolysis An electrospun nanofiber mat was placed in a Lindberg 54453 Heavy Duty Tube Furnace purchased from the TPS Co. (Watertown, WI) for pyrolysis. The following was the procedure for pyrolysis: (1) increasing the temperature at 10 °C/min from room temperature to 500 °C, (2) maintaining the temperature at 500 °C for 6 h to completely burn/remove organic components in the fibers and to allow TiO2 to crystallize, and (3) naturally cooling off to room temperature. A constant flow of air was maintained through the furnace during the pyrolysis.
water before being dried at 80 °C in vacuum. Fig. 2 schematically illustrates the morphological structures of different electrospun TiO2 nanofibers. 2.6. Characterization A Zeiss Supra 40VP field-emission SEM and a Rigaku Ultima Plus XRD at the South Dakota School of Mines and Technology, as well as a Hitachi HF-3300 high-resolution TEM at the Oak Ridge National Laboratory, were employed to characterize the morphological and structural properties of both as-electrospun nanofibers and the (final) TiO2 nanofibers. Prior to SEM examination, the specimens were sputter-coated with gold to avoid charge accumulations. A rotating X-ray generator (40 kW, 40 mA) with Cu Ka radiation (wavelength k = 1.54 Å) was used during the XRD experiment. The XRD profiles were recorded from 10° to 70° with the scanning speed of 2°/min. For the TEM characterizations, an acceleration voltage of 100 kV was selected for the as-electrospun nanofibers, and an acceleration voltage of 300 kV was selected for the (final) TiO2 nanofibers. The TEM specimens were prepared by dispersing nanofibers onto lacey carbon films supported on 200-mesh copper grids. Additionally, the Brunauer–Emmett–Teller (BET) method was adopted to determine the specific surface area of each type of electrospun TiO2 nanofibers in the relative pressure (P/P0) range of 0.05–0.35. The desorption data of N2 isotherm were analyzed by the Barrett–Joyner–Halenda (BJH) method to acquire the value of specific surface area. The BET specific surface area of each type of electrospun TiO2 nanofibers was measured for three times, and the average value was reported; it is noteworthy that the variation of values for the same sample was small. 3. Results
2.5. Chemical etching treatment with NaOH aqueous solution
3.1. Solid TiO2 nanofibers
For the preparations of porous TiO2 nanofibers, the TiO2/Al2O3 and TiO2/ZnO composite nanofibers were first prepared; thereafter, each type of composite nanofibers was immersed in 2.5 M NaOH aqueous solution for 24 h under magnetic stirring to selectively remove the Al2O3 or ZnO components, respectively. Finally, the samples were centrifuged and rinsed for 3–5 times with distilled
Morphologies and structures of electrospun solid TiO2 nanofibers were investigated by SEM, TEM, and XRD. The SEM image in Fig. 3a shows that the solid TiO2 nanofibers had the cylindrical morphology with diameters of 200–300 nm and lengths of at least tens of microns. Note that the average fiber length would be affected by the properties (e.g., viscosity, and concentration of
(a)
Solid TiO2 nanofiber
Porous TiO2 nanofibers
(b)
Pyrolysis
1.0 µm Representative morphology of as-electrospun nanofibers
(c)
Hollow/tubular TiO2 nanofiber Removal of Al2O3 TiO2/Al2O3 composite nanofiber
(d)
Removal of ZnO
TiO2/ZnO composite nanofiber Fig. 2. Schematic representation showing the preparations of electrospun TiO2 nanofibers with different morphological structures.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
107
(sheath) capillaries of coaxial spinneret (as shown in Fig. 1); hence, a compound Taylor cone [9] and then a coaxial jet/filament would be generated when a high voltage was applied to the spinneret. The jet/filament would subsequently be stretched by electrostatic force to become thinner while following a looping/spiraling trajectory; during such a process, the solvent in electrospinning jet/filament would evaporate very rapidly, resulting in the formation of coresheath nanofibers. In specific, the outer (sheath) spin dope contained an alkoxide precursor of TiO2 (i.e., TNBT), while the inner (core) spin dope was paraffin oil that would be burnt/removed during the subsequent pyrolysis in air at 500 °C. A representative SEM image of the hollow/tubular TiO2 nanofibers is shown in Fig. 4a. It was evident that these hollow/tubular TiO2 nanofibers had diameters of 300–500 nm and wall-thickness in the range from tens of nanometers to 100–200 nm; under the scope of this study, all of the prepared TiO2 nanofibers appeared to possess the hollow/tubular morphology. The structures of these hollow/tubular TiO2 nanofibers were further characterized by
Fig. 3. Representative SEM (a) and TEM (b) images, and XRD pattern (c) of electrospun solid TiO2 nanofibers.
TNBT) of spin dope and the processing conditions during pyrolysis; in general, the fiber length would be increased if the amount of structural defects in fibers could be reduced. The TEM image in Fig. 3b indicates that electrospun solid TiO2 nanofibers were polycrystalline with crystallites (grains) having sizes of 10 nm. These grains were self-generated during the pyrolysis of precursor nanofibers due to homogeneous nucleation and space confinement of the nanofibers [10]. Fig. 3c shows the XRD pattern of electrospun solid TiO2 nanofibers, and the diffraction peaks can be indexed to the (1 0 1), (0 0 4), (2 0 0), (2 1 1), and (2 0 4) crystallographic planes of (tetragonal) anatase-phase TiO2 according to JCPDS No. 211272. The measured value of BET specific surface area for electrospun solid TiO2 nanofibers was 15.2 m2/g. 3.2. Hollow/tubular TiO2 nanofibers For preparation of hollow/tubular TiO2 nanofibers, two spin dopes were fed respectively through the inner (core) and outer
Fig. 4. Representative SEM (a) and TEM (b) images, and XRD pattern (c) of electrospun hollow/tubular TiO2 nanofibers.
108
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
TEM. Fig. 4b is a high-resolution TEM image showing that the nanotubes were also polycrystalline with self-generated crystallites (grains) having sizes of 10 nm, and the TiO2 in the nanotubes also possessed the crystalline phase of anatase. The anatase phase of hollow/tubular TiO2 nanofibers was further confirmed by the XRD results, as shown in Fig. 4c. This is expected since the determining factor for crystalline structure of TiO2 is the pyrolysis temperature [10], which is the same for both solid and hollow/tubular TiO2 nanofibers. The BET specific surface area of hollow/tubular TiO2 nanofibers was measured at 27.3 m2/g, almost twice as that of solid TiO2 nanofibers; this was simply because both the outside surface and the inside surface of nanotubes would adsorp N2 during the measurement. 3.3. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers To make the precursor of electrospun Al2O3/TiO2 composite nanofibers, the spin dope containing two alkoxides of TNBT and Al[OCH(CH3)2]3 was prepared as described in the Experimental section. Since the molar ratio of Ti/Al in the composite nanofibers was 8/2, it was our speculation that the Al2O3 component would be distributed randomly in the matrix of TiO2 (as separated and/ or co-continuous nanoscale domains as schematically shown in Fig. 2); upon the subsequent removal of Al2O3 via the etching treatment in NaOH aqueous solution, the (final) porous TiO2 nanofibers would then be obtained. The morphologies and structures of (final) porous TiO2 nanofibers and their precursor of electrospun Al2O3/TiO2 composite nanofibers were characterized by SEM, TEM, and XRD, and the
results are shown in Fig. 5. Intriguingly, the electrospun Al2O3/ TiO2 composite nanofibers possessed the ribbon-shaped (instead of cylindrical) morphology; the ribbons were quite uniform with width being 1.0 lm and thickness being 200 nm. The formation mechanism of such a ribbon-shaped morphology is still under investigation. After the treatment in NaOH aqueous solution to selectively remove the Al2O3 component, the (final) porous TiO2 nanofibers were obtained, and their representative morphology is shown in Fig. 5b. Due to the reasons that (1) the magnetic stirring was applied for 24 h during the NaOH treatment, and (2) the removal of Al2O3 component from the composite nanofibers would result in the formation of pores (which would have the weakening effect to fibers); the (final) porous TiO2 nanofibers became short fibers with lengths of several microns. A representative TEM image and the selected area electron diffraction pattern of porous TiO2 nanofibers are shown in Fig. 5c. Similar to the solid and hollow/ tubular TiO2 nanofibers, the porous TiO2 nanofibers were also polycrystalline; the crystalline structure was analyzed by XRD, and the corresponding XRD pattern is shown in Fig. 5d. These results indicated that this type of porous TiO2 nanofibers possessed the anatase phase with the crystallinity of 90%. The BET specific surface area of this type of porous TiO2 nanofibers was measured at 106.5 m2/g, considerably higher than those of solid TiO2 nanofibers (15.2 m2/g) and hollow/tubular TiO2 nanofibers (27.3 m2/g). 3.4. Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers The hypothesis of this approach was that, during the sol–gel process of TNBT in as-electrospun nanofibers and the subsequent
Fig. 5. SEM images of (a) Al2O3/TiO2 composite nanofibers after pyrolysis at 500 °C for 6 h and (b) porous TiO2 nanofibers after chemical etching treatment with NaOH aqueous solution; (c) TEM image (inset: electron diffraction pattern) and (d) XRD pattern of porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
pyrolysis, the existence of ZnO nanoparticles would hinder the volumetric contraction/shrinkage, resulting in the formation of nanoscale pores; furthermore, the ZnO component in ZnO/TiO2 composite nanofibers could also be removed via chemical etching treatment with NaOH aqueous solution, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The ZnO nanoparticles used in this study had the particle sizes smaller than 30 nm, and they could be uniformly dispersed in DMF to form a colloidal mixture. The molar ratio of Ti/Zn in electrospun ZnO/TiO2 composite nanofibers was set at 7/3; this was due to the reasons that (1) if the amount of ZnO was too low, the increase in specific surface area would be too small, while (2) if the amount of ZnO was too high, the fiber morphology would be completely lost upon removal of the ZnO component. The SEM image in Fig. 6a shows the representative morphological structure of electrospun ZnO/TiO2 composite nanofibers; some nanofibers had large diameters of 200–300 nm, while others had small diameters of tens of nanometers. All of nanofibers had relatively rough surfaces with ZnO nanoparticles dispersed both inside and on the nanofibers. The rough surface was probably because the ZnO nanoparticles might be enriched on the nanofiber surface due to the volumetric shrinkage of fibers during the pyrolysis process. The XRD pattern (acquired from the ZnO/TiO2 composite nanofibers) in Fig. 6b shows the characteristic diffraction peaks of both anatase-phase TiO2 and wurtzite-structured ZnO. After the ZnO/TiO2 composite nanofibers were treated with NaOH aqueous solution and then washed by distilled water to remove the ZnO component, the (final) porous TiO2 nanofibers were obtained, and their representative morphology is shown in the SEM and TEM images of Fig. 6c and d, respectively. Similar to the porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers, the
109
removal of ZnO component from the ZnO/TiO2 composite nanofibers would also weaken the fibers, and the (final) porous TiO2 nanofibers were also short fibers with lengths being microns or shorter. The XRD pattern (not shown) acquired from this type of porous TiO2 nanofibers did not contain the clearly identifiable peaks of anatase-phase TiO2. This might be due to the reasons that (1) the anatase-phase TiO2 crystallites were too small to generate identifiable peaks in XRD, and/or (2) there were many nanoand/or meso-sized pores throughout in the fibers. The BET specific surface area of TiO2/ZnO composite nanofibers was measure as 67.5 m2/g, which was higher than that of solid and/or hollow/ tubular TiO2 nanofibers; whereas the BET specific surface area of porous TiO2 nanofibers after removal of ZnO nanoparticles was 148.6 m2/g.
4. Discussion Since the photovoltaic efficiency and photo-catalytic performance would be strongly affected by the specific surface area and morphological structure of anatase-phase TiO2, the objectives of this study included (1) to reveal the correlations between the preparation methods and the morphological structures of electrospun anatase-phase TiO2 nanofibers and (2) to fabricate the TiO2 nanofibers with increased specific surface areas for future photovoltaic and photo-catalytic studies. The density of anatase-phase TiO2 is 4.23 g/cm3. Hence, the total surface area for solid TiO2 nanofibers with the diameter of 200 nm can be calculated as merely 2 m2/g, while the calculated total surface area for solid TiO2 nanorods with the diameter of 20 nm is 20 m2/g. This is the reason that many types of TiO2 nanorods/nanofibers do not
Fig. 6. SEM image (a) and XRD pattern (b) of ZnO/TiO2 composite nanofibers after pyrolysis at 500 °C for 6 h and (c) SEM and (d) TEM images of porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers.
110
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
Table 1 The descriptions of different electrospun anatase-phase TiO2 nanofibers. Sample
Preparation method
BET specific surface area
Solid TiO2 nanofibers
Solid TiO2 nanofibers were made via electrospinning the spin dope consisting of TNBT and PVP in ethanol with the TNBT/PVP mass ratio of 10/1 followed by pyrolyzing the as-electrospun nanofibers in air at 500 °C. Hollow/tubular TiO2 nanofibers were prepared via the technique of coaxial electrospinning; the spin dope for making the sheath component (of as-electrospun precursor nanofibers) consisted of TNBT and PVP in ethanol, while the spin dope for making the core component was paraffin oil. The as-electrospun precursor nanofibers were pyrolyzed in air at 500 °C. The composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., aluminum isopropoxide) was electrospun into nanofibers, and then the TiO2/Al2O3 composite nanofibers were made by pyrolysis; thereafter, the Al2O3 component would be removed via etching treatment, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers. The mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm) was electrospun into nanofibers, and then the ZnO/TiO2 composite nanofibers were made by pyrolysis; the ZnO component in the resulting TiO2/ZnO composite nanofibers could also be selectively removed, resulting in nanoscale pores in the (final) TiO2 nanofibers.
15.2 (m2/g)
Hollow/tubular TiO2 nanofibers
Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers
Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers
outperform TiO2 nanoparticles in the applications such as dye-sensitized solar cell and photo-catalysis, since TiO2 nanoparticles (such as those purchased from Solaronix in Switzerland) have the average size of 9 nm and the BET specific surface area of 160 m2/g. Note that the BET specific surface area is usually several times larger than the total surface area, because the multilayer (instead of mono-layer) adsorption occurs during the measurement. It was our speculation that electrospun porous TiO2 nanofibers (with the BET specific surface area of at least 100 m2/ g) in the form of overlaid fiber mat would outperform TiO2 nanoparticles, particularly for making the photo-anode in dye-sensitized solar cell, because the nanofibers would provide continuous pathways for the photo-generated electrons without substantial scarification of dye loading amount. To further increase the specific surface area of electrospun anatase-phase TiO2 nanofibers, several approaches were investigated during this study. The first approach was to vary the concentration of carrying polymer (i.e., PVP) and/or to introduce thermally degradable polymers such as polymethylmethacrylate (PMMA) into the spin dope. Such an approach was revealed not to be very effective, and the resulting TiO2 nanofibers generally had the BET specific surface area smaller than 50 m2/g. This was because the decomposition of polymers (e.g., PVP and PMMA) occurred prior to the homogeneous nucleation and growth of anatase-phase TiO2 crystallites during the pyrolysis [10]. In other words, small pores created from the decomposition of polymers would unlikely be retained in the final TiO2 nanofibers, because the TiO2 was structurally amorphous and relatively easy to move molecularly at the stage. The second approach was to prepare the hollow/tubular TiO2 nanofibers. Nevertheless, the BET specific surface area of hollow/tubular TiO2 nanofibers was still considerably lower than 100 m2/g, despite that the value was almost twice as that of solid TiO2 nanofibers. We therefore concluded that small pores had to be generated throughout the nanofibers after pyrolysis to substantially increase the specific surface of anatase-phase TiO2 nanofibers. As described in the section of Results, two approaches were investigated. The first approach was to prepare the composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., Al[OCH(CH3)2]3); after the TiO2/Al2O3 composite nanofibers were made, the Al2O3 component would then be removed via chemical etching treatment with NaOH aqueous solution, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers. The second approach was to prepare the mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm); the hypothesis was that, during the sol–gel process of TNBT and the subsequent pyrolysis, the existence of ZnO nanoparticles would hinder the
27.3 (m2/g)
106.5 (m2/g)
148.6 (m2/g)
volumetric contraction/shrinkage, resulting in the formation of nanoscale pores. Furthermore, the ZnO component in TiO2/ZnO composite nanofibers could also be removed by using the NaOH aqueous solution, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The acquired results indicated that the second approach led to porous TiO2 nanofibers with BET specific surface area of 148.6 m2/g, while the first approach led to porous TiO2 nanofibers with BET specific surface area of 106.5 m2/g. The descriptions of different electrospun TiO2 nanofibers are summarized in Table 1. It is noteworthy that the heating procedure in pyrolysis could have the significant impact on structure (particularly the formation and amount of defects), while it would not appreciably impact on the BET specific surface area. For example, with slower heating rate, the amount of structural defects in both solid and hollow/tubular nanofibers might be less, since the average length of the fibers/tubes was longer, whereas the variation of BET specific surface area was very small. We also attempted to prepare the hollow/tubular TiO2 nanofibers with pores in the walls by using the composite spin dope containing both TNBT and the precursor of a removable component (e.g., Al2O3 or ZnO); nonetheless, such an attempt was not very successful to further increase the BET specific surface area significantly. It also needs to clarify that, if the specific surface area of TiO2 nanofibers is too large (i.e., 300 m2/g or higher), the nano- and/or meso-scaled pores in the nanofibers might lead to too many surface trapping states; if such TiO2 nanofibers are used for making the photo-anode in dye-sensitized solar cell, the photo-generated electrons might not move along the nanofibers for a long distance before the recombination would occur.
5. Summary In summary, the systematic investigations were carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures (e.g., solid, hollow/ tubular, and porous) and specific surface areas. The results indicated that electrospun TiO2 nanofibers with hollow/tubular structure had the BET specific surface area of 27.3 m2/g, which was approximately twice as that of solid TiO2 nanofibers (15.2 m2/ g). The porous TiO2 nanofibers made from the precursor of Al2O3/ TiO2 composite nanofibers had the BET specific surface area of 106.5 m2/g, while the porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers had the highest BET specific surface area of 148.6 m2/g. It is envisioned that electrospun anatase-phase TiO2 nanofibers with desired morphological structure and relatively high BET specific surface area would outperform
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
other nanostructures (e.g., powder and film) of TiO2 for various applications (particularly dye-sensitized solar cell and photocatalysis). Acknowledgments This research was supported by the National Science Foundation (Grant No.: EPS-0903804), the National Aeronautics and Space Administration (Cooperative Agreement No.: NNX10AN34A), and the State of South Dakota. Y. Cai would acknowledge the National Natural Science Foundation of China (Grant No.: 51006046), the Natural Science Foundation of Jiangsu Province (Grant No.: BK2010140), and the National High-Tech R&D Program of China (Grant No.: 2012AA030313). References [1] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735. [2] U. Bach, D. Lupo, P. Comte, J.E. Moser, F. Weissorte, J. Salbeck, H. Spreitzer, M. Gratzel, Nature 395 (1998) 583. [3] D. Vu, Z. Li, H. Zhang, W. Wang, Z. Wang, X. Xu, B. Dong, C. Wang, J. Colloid Interface Sci. 367 (2012) 429. [4] A. Wold, Chem. Mater. 5 (1993) 280. [5] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69. [6] A. Greiner, J.H. Wendorff, Angew. Chem. Int. Ed. 46 (2007) 5670.
111
[7] S. Ramakrishna, K. Fujihara, W.E. Teo, T. Yong, Z.W. Ma, R. Ramaseshan, Mater. Today 9 (2006) 40. [8] V. Thavasi, G. Singh, S. Ramakrishna, Energy Environ. Sci. 1 (2008) 205. [9] D.H. Reneker, I. Chun, Nanotechnology 7 (1996) 216. [10] R. Chandrasekar, L. Zhang, J.Y. Howe, N.E. Hedin, Y. Zhang, H. Fong, J. Mater. Sci. 44 (2009) 1198. [11] D. Li, Y.N. Xia, Nano Lett. 3 (2003) 555. [12] C. Tekmen, A. Suslu, U. Cocen, Mater. Lett. 62 (2008) 4470. [13] B. Ding, C.K. Kim, H.Y. Kim, M.K. Seo, S.J. Park, Fibers Polym. 5 (2004) 105. [14] S.H. Lee, C. Tekmen, W.M. Sigmund, Mater. Sci. Eng. 398 (2005) 77. [15] K. Nakane, K. Yasuda, T. Ogihara, N. Ogata, S. Yamaguchi, J. Appl. Polym. Sci. 104 (2007) 1232. [16] M.A. Kanjwal, N.A. Barakat, F.A. Sheikh, S.J. Park, H.Y. Kim, Macromol. Res. 18 (2010) 233. [17] A.S. Nair, S.Y. Yang, P.N. Zhu, S. Ramakrishna, Chem. Commun. 46 (2010) 7421. [18] D. Li, Y.N. Xia, Nano Lett. 4 (2004) 933. [19] J.T. McCann, D. Li, Y.N. Xia, J. Mater. Chem. 15 (2005) 735. [20] D. Li, J.T. McCann, Y.N. Xia, Small 1 (2005) 83. [21] Y.L. Cheng, W.Z. Huang, Y.F. Zhang, L. Zhu, Y.J. Liu, X.Z. Fan, X.Q. Cao, CrystEngComm 12 (2010) 2256. [22] R.A. Caruso, J.H. Schattka, A. Greiner, Adv. Mater. 13 (2001) 1577. [23] G.M. Kim, S.M. Lee, G.H. Michler, H. Roggendorf, U. Gosele, M. Knez, Chem. Mater. 20 (2008) 3085. [24] S.H. Zhan, D.R. Chen, X.L. Jiao, C.H. Tao, J. Phys. Chem. B 110 (2006) 11199. [25] S.H. Zhan, D.R. Chen, X.L. Jiao, Y. Song, Chem. Commun. 20 (2007) 2043. [26] M.A. Kanjwal, N.A. Barakat, F.A. Sheikh, M.S. Khil, H.Y. Kim, Int. J. Appl. Ceram. Technol. 7 (2010) 54. [27] Q. Li, D. Sun, H. Kim, Mater. Res. Bull. 46 (2011) 2094. [28] Z.Y. Liu, D.D. Sun, P. Guo, J.O. Leckie, Nano Lett. 7 (2007) 1081. [29] V. Stengl, J. Velicka, M. Marikova, T.M. Grygar, ACS Appl. Mater. Interfaces 3 (2011) 4014.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Electrospun anatase-phase TiO2 nanofibers with different morphological structures and specific surface areas Guangfei He a, Yibing Cai b,c,⇑, Yong Zhao b, Xiaoxu Wang d, Chuilin Lai a, Min Xi a, Zhengtao Zhu a,b,d, Hao Fong a,b,d,⇑ a
Program of Materials Engineering and Science, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA Department of Chemistry and Applied Biological Sciences, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA Key Laboratory of Eco-Textiles, Ministry of Education, Jiangnan University, Wuxi, Jiangsu 214122, China d Program of Nanoscience and Nanoengineering, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA b c
a r t i c l e
i n f o
Article history: Received 29 December 2012 Accepted 4 February 2013 Available online 21 February 2013 Keywords: Titania Electrospinning Nanofiber Morphological structure Specific surface area
a b s t r a c t Electrospun anatase-phase TiO2 nanofibers with desired morphological structure and relatively high specific surface area are expected to outperform other nanostructures (e.g., powder and film) of TiO2 for various applications (particularly dye-sensitized solar cell and photo-catalysis). In this study, systematic investigations were carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures (e.g., solid, hollow/tubular, and porous) and specific surface areas. The TiO2 nanofibers were generally prepared via electrospinning of precursor nanofibers followed by pyrolysis at 500 °C. For making hollow/tubular TiO2 nanofibers, the technique of co-axial electrospinning was utilized; while for making porous TiO2 nanofibers, the etching treatment in NaOH aqueous solution was adopted. The results indicated that the hollow/tubular TiO2 nanofibers (with diameters of 300–500 nm and wall-thickness in the range from tens of nanometers to 200 nm) had the BET specific surface area of 27.3 m2/g, which was approximately twice as that of the solid TiO2 nanofibers (15.2 m2/g) with diameters of 200–300 nm and lengths of at least tens of microns. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers had the BET specific surface area of 106.5 m2/g, whereas porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers had the highest BET specific surface area of 148.6 m2/g. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction The anatase-phase titanium (IV) oxide (TiO2) has attracted extensive research interests in the recent years, because it has a wide range of applications such as environmental remediation, electronics, sensor technology, solar cell, and photo-catalysis [1–5]. The ceramic of TiO2 has several crystalline phases including anatase and rutile, and the band-gap of anatase-phase TiO2 is 3.2 eV, which is equivalent to the energy of light with the wavelength of 388 nm. Hence, the anatase-phase TiO2 can acquire energy either directly from sunlight or indirectly from light sensitizers (e.g., dyes), causing electrons to be excited to conduction band and concomitantly creating positive holes in valence band. This is termed as charge separation in solar cell, and the
⇑ Corresponding authors. Addresses: Key Laboratory of Eco-Textiles, Ministry of Education, Jiangnan University, Wuxi, Jiangsu 214122, China. Fax: +86 510 8591 2009 (Y. Cai), Program of Materials Engineering and Science, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA. Fax: +1 605 394 1232 (H. Fong). E-mail addresses: [email protected] (Y. Cai), [email protected] (H. Fong). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.02.009
separated charges can be turned into electricity. Additionally, the anatase-phase TiO2 can also be applied to the remediation of a variety of organic compounds and heavy metal ions (e.g., Pt4+, Pd2+, and Cr3+) as well as the destruction of micro-organisms such as bacteria, viruses, and molds from aqueous environments; this is fundamentally owing to the oxidizing capability of positive holes in the valence band. It is noteworthy that many applications of anatase-phase TiO2 are strongly influenced by its morphological structure and specific surface area. The materials-processing technique of electrospinning utilizes the electric force to drive the spinning process and to produce the fibers with diameters typically in the range from tens of nanometers to microns (commonly known as-electrospun nanofibers). The applications of electrospun nanofibers include, but not limited to, composites, filtrations/separations, biomedical applications (e.g., wound dressing, tissue engineering, and drug delivery), electronic applications (e.g., sensors, transistors, and detectors), and energy-related applications (e.g., solar cells, fuel cells, batteries, and supercapacitors) [6–8]. Unlike nanowires, nanorods, and nanotubes that are prepared by bottom-up synthetic methods and usually require further purifications, electrospun nanofibers are
104
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
produced via a top-down manufacturing process; they are inexpensive, continuous, and also relatively easy to align, assemble, and process into applications [6–10]. The electrospun TiO2 nanofibers (with diameters being hundreds of nanometers, and consisting of anatase-phase TiO2 crystallites with sizes of 10 nm), particularly the porous ones with relatively high specific surface area, are expected to outperform other nanostructures (e.g., powder and film) of TiO2 for the applications in solar cells and photocatalyses. In the recent decade, there have been numerous research endeavors devoted on preparations, structures, and properties of electrospun TiO2 nanofibers; in general, these nanofibers are made by electrospinning spin dopes containing TiO2 precursors (e.g., titanium (IV) n-butoxide) followed by pyrolyzing the as-electrospun precursor nanofibers at high temperature [11–17]. Through judiciously adjusting/controlling the properties of spin dopes, as well as the processing parameters during electrospinning and the subsequent pyrolysis, the electrospun TiO2 nanofibers with various morphological structures have been prepared. For example, Xia’s group fabricated TiO2 nanofibers with solid, core-sheath, hollow, and porous structures via the modified electrospinning techniques (e.g., the coaxial electrospinning technique) [11,18–20]; Cheng et al. prepared hollow TiO2 nanofibers (i.e., TiO2 nanotubes) via tailoring the composition of spin dope and adjusting the heating rate during pyrolysis [21]. Additionally, electrospun polymer (e.g., polylactide) nanofibers could be surface-coated with amorphous TiO2 via sol–gel method; upon removal of the thermally degradable polymer, the hollow TiO2 fibers/tubes would be acquired [22]; this suggested that different inorganic/ceramic tubular nanostructures with varied diameter and wall-thickness could be prepared by using electrospun polymer nanofibers as templates. Thereafter, Kim et al. made sub-micron tubes of anatase-phase TiO2 by a template-directed method [23]; in specific, electrospun polyvinylpyrrolidone nanofibers were used as the template for surface-coating of TiO2 using the technique of atomic layer deposition. Zhan et al. fabricated long hollow TiO2 fibers/tubes with mesoporous walls via the combination of a sol–gel method and the coaxial electrospinning technique by using a triblock copolymer as the poredirecting/creating agent, and these hollow fibers/tubes exhibited higher photo-catalytic activities toward degradation of Methylene Blue and formaldehyde than other nanostructured TiO2 materials such as commercial TiO2 nanoparticles (P25, Degussa) and mesoporous TiO2 films [24]. Furthermore, the mesoporous TiO2/SiO2 composite nanofibers have also been prepared, and these nanofibers showed selective photo-catalytic activities on the decompositions of Methylene Blue, Active Yellow, and Disperse Red [25]. In another example, Kanjwal et al. fabricated ZnO-doped TiO2 nanofibers via electrospinning, and the introduced ZnO–TiO2 hierarchical nanostructure could eliminate the dye of Methyl Red within 90 min and the dye of Rhodamine B within 105 min [16]; they also prepared electrospun TiO2 nanofibers with silver nanoparticles, and the photo-catalytic performance was very high [26]. It is also noteworthy that Li et al. [27] made porous TiO2 nanofibers by alkali-dissolution of SiO2 from TiO2/SiO2 composite nanofibers to improve the surface-to-volume ratio, and the resulting nanofibers exhibited superior photo-catalytic activity (the best photo-catalytic efficiency obtained in the study was 77% upon irradiation under the simulated sunlight for 1 h). Additionally, Liu et al. [28] reported that the electrospinning process with side-by-side dual spinnerets could be a simple approach for fabrication of bi-component TiO2/SnO2 nanofibers with controlled hetero-junctions, and such a structure would lead to an increase in the charge separation of the photo-generated electrons and holes within the bi-component system, allowing both electrons and holes to participate in the overall photo-catalytic reaction (i.e., the bi-component TiO2/ SnO2 nanofiber mat would be an excellent photo-catalytic system).
In another example, Stengl et al. [29] made the tungsten-doped TiO2 nanofibers by thermally hydrolyzing the aqueous solutions of peroxo complexes containing titanium and tungsten, and these nanofibers exhibited fast reaction rate toward photo-degradation of Orange II dye. It is well-known that the photovoltaic efficiency and photo-catalytic activity are strongly dependent on the specific surface area and morphological structure of anatase-phase TiO2. The electrospun mats consisting of overlaid anatase-phase TiO2 nanofibers possess the following advantageous properties for those applications: (1) high specific surface area (particularly if the nanofibers are porous), (2) controllable pore sizes among the nanofibers (ranging from tens to hundreds of nanometers), and (3) the thickness of a nanofiber mat can be readily tailored/controlled. Unlike nanoscale TiO2 particles/rods which are in loose granular form, electrospun TiO2 nanofibers are well-contained in the mat. The high specific surface areas of electrospun TiO2 nanofiber mats (particularly the ones consisting of porous TiO2 nanofibers) lead to a large number of reaction sites which would enhance the photovoltaic efficiency and photo-catalytic activity; additionally, the nanofiber mat would also result in great accessibility for reactants during the applications. In this study, systematic investigations have been carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures of solid, hollow/tubular, and porous. The objectives were (1) to reveal the correlations between the preparation methods and the morphological structures of electrospun TiO2 nanofibers and (2) to fabricate the TiO2 nanofibers with increased specific surface areas for future photovoltaic and photo-catalytic studies. Specifically, the solid TiO2 nanofibers were made by electrospinning the spin dope consisting of titanium (IV) n-butoxide (TNBT) and polyvinylpyrrolidone (PVP) in ethanol with the TNBT/PVP mass ratio being 10/1 followed by pyrolysis at 500 °C. The hollow/tubular TiO2 nanofibers were prepared via the technique of coaxial electrospinning; the spin dope for making the sheath component (of as-electrospun precursor nanofibers) consisted of TNBT and PVP in ethanol, while the spin dope for making the core component was paraffin oil. To prepare porous TiO2 nanofibers, two approaches were explored: (1) to prepare the composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., aluminum isopropoxide); after the TiO2/ Al2O3 composite nanofibers were made, the Al2O3 component would then be removed via chemical etching with NaOH aqueous solution, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers and (2) to prepare the mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm); the hypothesis was that, during the sol–gel process of TNBT and the subsequent pyrolysis, the existence of ZnO nanoparticles would hinder the volumetric contraction/shrinkage, resulting in the formation of nanoscale pores. Moreover, the ZnO component in TiO2/ZnO composite nanofibers could also be removed via chemical etching treatment, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The morphological and crystalline structures as well as specific surface areas of the solid, hollow/tubular, and two types of porous TiO2 nanofibers were characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), and a Brunauer-Emmett-Teller (BET) surface area analyzer. The acquired results indicated that all of electrospun TiO2 nanofibers had diameters in hundreds of nanometers and consisted of anatase-phase TiO2 crystallites with sizes of 10 nm; as compared to the solid TiO2 nanofibers with BET surface area of 15.2 m2/g, the hollow/tubular TiO2 nanofibers had the BET surface area of 27.3 m2/g, whereas the BET surface areas of porous TiO2 nanofibers were 106.5 m2/g and 148.6 m2/g upon the preparation methods.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
105
then mixed together followed by addition of 2.5 g TNBT. The final mixture was magnetically stirred for 24 h prior to electrospinning.
2. Experimental 2.1. Materials Titanium (IV) n-butoxide (TNBT), aluminum isopropoxide (Al[OCH(CH3)2]3), zinc oxide (ZnO, with particle sizes smaller than 30 nm), polyvinylpyrrolidone (PVP, Mw = 1,300,000), N,N-dimethylformamide (DMF), isopropanol (IPA), ethanol, paraffin oil, acetic acid (HAc), and sodium hydroxide (NaOH) were purchased from the Sigma–Aldrich Co. (St. Louis, MO) and used without further purification.
2.2.4. Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers The optimal spin dope (determined from the properties of aselectrospun nanofibers, ZnO/TiO2 composite nanofibers, and final TiO2 nanofibers) with the molar ratio of Ti/Zn being 7/3 was prepared by using the following procedure: PVP (2 g), TNBT (1.59 g), and ZnO (0.41 g) were mixed together with DMF (8 g) in a sealed container followed by being magnetically stirred for 24 h.
2.2. Preparation of spin dopes
2.3. Electrospinning
During the preparations of the following spin dopes, a trace amount (1 wt.%) of HAc was added in each spin dope to control the hydrolysis/gelation of TNBT.
2.3.1. Solid and porous TiO2 nanofibers A solution was filled in a 30 ml plastic syringe having a bluntend stainless-steel needle with the inside diameter of 0.3 mm. The electrospinning setup included a high voltage power supply (Model No.: ES30P), purchased from the Gamma High Voltage Research, Inc. (Ormond Beach, FL), and a nanofiber collector of electrically grounded aluminum foil that covered a laboratory-produced roller with diameter of 25 cm; the rotating speed was set at 100 rpm. The collector was placed at 25 cm below the tip of needle. During electrospinning, a positive high voltage of 20 kV was applied to the needle, and the solution feeding rates of 0.6 ml/h for solid TiO2 nanofibers, 2.0 ml/h for porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers, and 1.0 ml/h for porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers were maintained using a syringe pump (Model No.: KDS 200) purchased from the KD Scientific Inc. (Holliston, MA). The electrospinning was carried out in open environment at room temperature of 25 °C. The electrospinning setup with single (i.e., non-coaxial) spinneret for solid and/or porous TiO2 nanofibers is shown in Fig. 1.
2.2.1. Solid TiO2 nanofibers The spin dope for making solid TiO2 nanofibers was prepared by dissolving 0.3 g PVP and 3 g TNBT in 5 ml ethanol. The spin dope was magnetically stirred at room temperature for 24 h prior to electrospinning. 2.2.2. Hollow/tubular TiO2 nanofibers For making hollow/tubular TiO2 nanofibers, the spin dope for making solid TiO2 nanofibers was placed in a syringe connected to the outer capillary of a coaxial spinneret; while paraffin oil was placed in another syringe connected to the inner capillary. 2.2.3. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers The optimal spin dope (determined from the properties of aselectrospun nanofibers, Al2O3/TiO2 composite nanofibers, and final TiO2 nanofibers) with the molar ratio of Ti/Al being 8/2 was prepared by using the following procedure: (1) 1.4 g PVP was dissolved in 11.75 g IPA in a sealed container via magnetically stirring for 0.5 h; (2) 0.375 g Al[OCH(CH3)2]3 was dissolved in 7.125 g IPA in another sealed container via ultrasonication for 15 min; (3) the PVP/IPA and Al[OCH(CH3)2]3/IPA solutions were
2.3.2. Hollow/tubular TiO2 nanofibers The setup with coaxial spinneret for electrospinning of hollow/ tubular TiO2 nanofibers is also shown in Fig. 1. The coaxial electrospinning is similar to the conventional electrospinning except for the use of a spinneret containing two coaxially arranged
Single spinneret
Coaxial spinneret Fig. 1. Schematic illustration of the setup for electrospinning nanofibers.
106
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
capillaries. The coaxial spinneret used in this study was purchased from the Nisco Engineering AG in Switzerland; the inner capillary had the diameter of 0.3 mm, while the outer capillary had the diameter of 0.8 mm. During coaxial electrospinning, two spin dopes were fed into the inner (core) and outer (sheath) capillaries, respectively. A coaxial jet/filament would be generated between the spinneret and the collector, when a high voltage was applied, and the jet/filament would subsequently be stretched by electrostatic force to generate core-sheath nanofibers [6]. In this study, the feeding rates for the inner spin dope (paraffin oil) and outer spin dope (TNBT and PVP in ethanol) were set at 0.6 ml/h and 0.8 ml/h, respectively. A high voltage of 15 kV was applied to the coaxial spinneret during the electrospinning process, and a laboratory-produced roller (with diameter of 25 cm) covered by electrically grounded aluminum foil was operating at 100 rpm as nanofiber collector. The electrospinning was carried out at room temperature of 25 °C in open environment, and the as-electrospun nanofibers (in the form of overlaid nanofiber mat) were kept in ambient condition for 48 h to allow TNBT (and Al[OCH(CH3)2]3) to completely hydrolyze into three dimensional network (i.e., gel) by the moisture in atmosphere. 2.4. Pyrolysis An electrospun nanofiber mat was placed in a Lindberg 54453 Heavy Duty Tube Furnace purchased from the TPS Co. (Watertown, WI) for pyrolysis. The following was the procedure for pyrolysis: (1) increasing the temperature at 10 °C/min from room temperature to 500 °C, (2) maintaining the temperature at 500 °C for 6 h to completely burn/remove organic components in the fibers and to allow TiO2 to crystallize, and (3) naturally cooling off to room temperature. A constant flow of air was maintained through the furnace during the pyrolysis.
water before being dried at 80 °C in vacuum. Fig. 2 schematically illustrates the morphological structures of different electrospun TiO2 nanofibers. 2.6. Characterization A Zeiss Supra 40VP field-emission SEM and a Rigaku Ultima Plus XRD at the South Dakota School of Mines and Technology, as well as a Hitachi HF-3300 high-resolution TEM at the Oak Ridge National Laboratory, were employed to characterize the morphological and structural properties of both as-electrospun nanofibers and the (final) TiO2 nanofibers. Prior to SEM examination, the specimens were sputter-coated with gold to avoid charge accumulations. A rotating X-ray generator (40 kW, 40 mA) with Cu Ka radiation (wavelength k = 1.54 Å) was used during the XRD experiment. The XRD profiles were recorded from 10° to 70° with the scanning speed of 2°/min. For the TEM characterizations, an acceleration voltage of 100 kV was selected for the as-electrospun nanofibers, and an acceleration voltage of 300 kV was selected for the (final) TiO2 nanofibers. The TEM specimens were prepared by dispersing nanofibers onto lacey carbon films supported on 200-mesh copper grids. Additionally, the Brunauer–Emmett–Teller (BET) method was adopted to determine the specific surface area of each type of electrospun TiO2 nanofibers in the relative pressure (P/P0) range of 0.05–0.35. The desorption data of N2 isotherm were analyzed by the Barrett–Joyner–Halenda (BJH) method to acquire the value of specific surface area. The BET specific surface area of each type of electrospun TiO2 nanofibers was measured for three times, and the average value was reported; it is noteworthy that the variation of values for the same sample was small. 3. Results
2.5. Chemical etching treatment with NaOH aqueous solution
3.1. Solid TiO2 nanofibers
For the preparations of porous TiO2 nanofibers, the TiO2/Al2O3 and TiO2/ZnO composite nanofibers were first prepared; thereafter, each type of composite nanofibers was immersed in 2.5 M NaOH aqueous solution for 24 h under magnetic stirring to selectively remove the Al2O3 or ZnO components, respectively. Finally, the samples were centrifuged and rinsed for 3–5 times with distilled
Morphologies and structures of electrospun solid TiO2 nanofibers were investigated by SEM, TEM, and XRD. The SEM image in Fig. 3a shows that the solid TiO2 nanofibers had the cylindrical morphology with diameters of 200–300 nm and lengths of at least tens of microns. Note that the average fiber length would be affected by the properties (e.g., viscosity, and concentration of
(a)
Solid TiO2 nanofiber
Porous TiO2 nanofibers
(b)
Pyrolysis
1.0 µm Representative morphology of as-electrospun nanofibers
(c)
Hollow/tubular TiO2 nanofiber Removal of Al2O3 TiO2/Al2O3 composite nanofiber
(d)
Removal of ZnO
TiO2/ZnO composite nanofiber Fig. 2. Schematic representation showing the preparations of electrospun TiO2 nanofibers with different morphological structures.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
107
(sheath) capillaries of coaxial spinneret (as shown in Fig. 1); hence, a compound Taylor cone [9] and then a coaxial jet/filament would be generated when a high voltage was applied to the spinneret. The jet/filament would subsequently be stretched by electrostatic force to become thinner while following a looping/spiraling trajectory; during such a process, the solvent in electrospinning jet/filament would evaporate very rapidly, resulting in the formation of coresheath nanofibers. In specific, the outer (sheath) spin dope contained an alkoxide precursor of TiO2 (i.e., TNBT), while the inner (core) spin dope was paraffin oil that would be burnt/removed during the subsequent pyrolysis in air at 500 °C. A representative SEM image of the hollow/tubular TiO2 nanofibers is shown in Fig. 4a. It was evident that these hollow/tubular TiO2 nanofibers had diameters of 300–500 nm and wall-thickness in the range from tens of nanometers to 100–200 nm; under the scope of this study, all of the prepared TiO2 nanofibers appeared to possess the hollow/tubular morphology. The structures of these hollow/tubular TiO2 nanofibers were further characterized by
Fig. 3. Representative SEM (a) and TEM (b) images, and XRD pattern (c) of electrospun solid TiO2 nanofibers.
TNBT) of spin dope and the processing conditions during pyrolysis; in general, the fiber length would be increased if the amount of structural defects in fibers could be reduced. The TEM image in Fig. 3b indicates that electrospun solid TiO2 nanofibers were polycrystalline with crystallites (grains) having sizes of 10 nm. These grains were self-generated during the pyrolysis of precursor nanofibers due to homogeneous nucleation and space confinement of the nanofibers [10]. Fig. 3c shows the XRD pattern of electrospun solid TiO2 nanofibers, and the diffraction peaks can be indexed to the (1 0 1), (0 0 4), (2 0 0), (2 1 1), and (2 0 4) crystallographic planes of (tetragonal) anatase-phase TiO2 according to JCPDS No. 211272. The measured value of BET specific surface area for electrospun solid TiO2 nanofibers was 15.2 m2/g. 3.2. Hollow/tubular TiO2 nanofibers For preparation of hollow/tubular TiO2 nanofibers, two spin dopes were fed respectively through the inner (core) and outer
Fig. 4. Representative SEM (a) and TEM (b) images, and XRD pattern (c) of electrospun hollow/tubular TiO2 nanofibers.
108
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
TEM. Fig. 4b is a high-resolution TEM image showing that the nanotubes were also polycrystalline with self-generated crystallites (grains) having sizes of 10 nm, and the TiO2 in the nanotubes also possessed the crystalline phase of anatase. The anatase phase of hollow/tubular TiO2 nanofibers was further confirmed by the XRD results, as shown in Fig. 4c. This is expected since the determining factor for crystalline structure of TiO2 is the pyrolysis temperature [10], which is the same for both solid and hollow/tubular TiO2 nanofibers. The BET specific surface area of hollow/tubular TiO2 nanofibers was measured at 27.3 m2/g, almost twice as that of solid TiO2 nanofibers; this was simply because both the outside surface and the inside surface of nanotubes would adsorp N2 during the measurement. 3.3. Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers To make the precursor of electrospun Al2O3/TiO2 composite nanofibers, the spin dope containing two alkoxides of TNBT and Al[OCH(CH3)2]3 was prepared as described in the Experimental section. Since the molar ratio of Ti/Al in the composite nanofibers was 8/2, it was our speculation that the Al2O3 component would be distributed randomly in the matrix of TiO2 (as separated and/ or co-continuous nanoscale domains as schematically shown in Fig. 2); upon the subsequent removal of Al2O3 via the etching treatment in NaOH aqueous solution, the (final) porous TiO2 nanofibers would then be obtained. The morphologies and structures of (final) porous TiO2 nanofibers and their precursor of electrospun Al2O3/TiO2 composite nanofibers were characterized by SEM, TEM, and XRD, and the
results are shown in Fig. 5. Intriguingly, the electrospun Al2O3/ TiO2 composite nanofibers possessed the ribbon-shaped (instead of cylindrical) morphology; the ribbons were quite uniform with width being 1.0 lm and thickness being 200 nm. The formation mechanism of such a ribbon-shaped morphology is still under investigation. After the treatment in NaOH aqueous solution to selectively remove the Al2O3 component, the (final) porous TiO2 nanofibers were obtained, and their representative morphology is shown in Fig. 5b. Due to the reasons that (1) the magnetic stirring was applied for 24 h during the NaOH treatment, and (2) the removal of Al2O3 component from the composite nanofibers would result in the formation of pores (which would have the weakening effect to fibers); the (final) porous TiO2 nanofibers became short fibers with lengths of several microns. A representative TEM image and the selected area electron diffraction pattern of porous TiO2 nanofibers are shown in Fig. 5c. Similar to the solid and hollow/ tubular TiO2 nanofibers, the porous TiO2 nanofibers were also polycrystalline; the crystalline structure was analyzed by XRD, and the corresponding XRD pattern is shown in Fig. 5d. These results indicated that this type of porous TiO2 nanofibers possessed the anatase phase with the crystallinity of 90%. The BET specific surface area of this type of porous TiO2 nanofibers was measured at 106.5 m2/g, considerably higher than those of solid TiO2 nanofibers (15.2 m2/g) and hollow/tubular TiO2 nanofibers (27.3 m2/g). 3.4. Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers The hypothesis of this approach was that, during the sol–gel process of TNBT in as-electrospun nanofibers and the subsequent
Fig. 5. SEM images of (a) Al2O3/TiO2 composite nanofibers after pyrolysis at 500 °C for 6 h and (b) porous TiO2 nanofibers after chemical etching treatment with NaOH aqueous solution; (c) TEM image (inset: electron diffraction pattern) and (d) XRD pattern of porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers.
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
pyrolysis, the existence of ZnO nanoparticles would hinder the volumetric contraction/shrinkage, resulting in the formation of nanoscale pores; furthermore, the ZnO component in ZnO/TiO2 composite nanofibers could also be removed via chemical etching treatment with NaOH aqueous solution, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The ZnO nanoparticles used in this study had the particle sizes smaller than 30 nm, and they could be uniformly dispersed in DMF to form a colloidal mixture. The molar ratio of Ti/Zn in electrospun ZnO/TiO2 composite nanofibers was set at 7/3; this was due to the reasons that (1) if the amount of ZnO was too low, the increase in specific surface area would be too small, while (2) if the amount of ZnO was too high, the fiber morphology would be completely lost upon removal of the ZnO component. The SEM image in Fig. 6a shows the representative morphological structure of electrospun ZnO/TiO2 composite nanofibers; some nanofibers had large diameters of 200–300 nm, while others had small diameters of tens of nanometers. All of nanofibers had relatively rough surfaces with ZnO nanoparticles dispersed both inside and on the nanofibers. The rough surface was probably because the ZnO nanoparticles might be enriched on the nanofiber surface due to the volumetric shrinkage of fibers during the pyrolysis process. The XRD pattern (acquired from the ZnO/TiO2 composite nanofibers) in Fig. 6b shows the characteristic diffraction peaks of both anatase-phase TiO2 and wurtzite-structured ZnO. After the ZnO/TiO2 composite nanofibers were treated with NaOH aqueous solution and then washed by distilled water to remove the ZnO component, the (final) porous TiO2 nanofibers were obtained, and their representative morphology is shown in the SEM and TEM images of Fig. 6c and d, respectively. Similar to the porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers, the
109
removal of ZnO component from the ZnO/TiO2 composite nanofibers would also weaken the fibers, and the (final) porous TiO2 nanofibers were also short fibers with lengths being microns or shorter. The XRD pattern (not shown) acquired from this type of porous TiO2 nanofibers did not contain the clearly identifiable peaks of anatase-phase TiO2. This might be due to the reasons that (1) the anatase-phase TiO2 crystallites were too small to generate identifiable peaks in XRD, and/or (2) there were many nanoand/or meso-sized pores throughout in the fibers. The BET specific surface area of TiO2/ZnO composite nanofibers was measure as 67.5 m2/g, which was higher than that of solid and/or hollow/ tubular TiO2 nanofibers; whereas the BET specific surface area of porous TiO2 nanofibers after removal of ZnO nanoparticles was 148.6 m2/g.
4. Discussion Since the photovoltaic efficiency and photo-catalytic performance would be strongly affected by the specific surface area and morphological structure of anatase-phase TiO2, the objectives of this study included (1) to reveal the correlations between the preparation methods and the morphological structures of electrospun anatase-phase TiO2 nanofibers and (2) to fabricate the TiO2 nanofibers with increased specific surface areas for future photovoltaic and photo-catalytic studies. The density of anatase-phase TiO2 is 4.23 g/cm3. Hence, the total surface area for solid TiO2 nanofibers with the diameter of 200 nm can be calculated as merely 2 m2/g, while the calculated total surface area for solid TiO2 nanorods with the diameter of 20 nm is 20 m2/g. This is the reason that many types of TiO2 nanorods/nanofibers do not
Fig. 6. SEM image (a) and XRD pattern (b) of ZnO/TiO2 composite nanofibers after pyrolysis at 500 °C for 6 h and (c) SEM and (d) TEM images of porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers.
110
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
Table 1 The descriptions of different electrospun anatase-phase TiO2 nanofibers. Sample
Preparation method
BET specific surface area
Solid TiO2 nanofibers
Solid TiO2 nanofibers were made via electrospinning the spin dope consisting of TNBT and PVP in ethanol with the TNBT/PVP mass ratio of 10/1 followed by pyrolyzing the as-electrospun nanofibers in air at 500 °C. Hollow/tubular TiO2 nanofibers were prepared via the technique of coaxial electrospinning; the spin dope for making the sheath component (of as-electrospun precursor nanofibers) consisted of TNBT and PVP in ethanol, while the spin dope for making the core component was paraffin oil. The as-electrospun precursor nanofibers were pyrolyzed in air at 500 °C. The composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., aluminum isopropoxide) was electrospun into nanofibers, and then the TiO2/Al2O3 composite nanofibers were made by pyrolysis; thereafter, the Al2O3 component would be removed via etching treatment, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers. The mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm) was electrospun into nanofibers, and then the ZnO/TiO2 composite nanofibers were made by pyrolysis; the ZnO component in the resulting TiO2/ZnO composite nanofibers could also be selectively removed, resulting in nanoscale pores in the (final) TiO2 nanofibers.
15.2 (m2/g)
Hollow/tubular TiO2 nanofibers
Porous TiO2 nanofibers made from the precursor of Al2O3/TiO2 composite nanofibers
Porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers
outperform TiO2 nanoparticles in the applications such as dye-sensitized solar cell and photo-catalysis, since TiO2 nanoparticles (such as those purchased from Solaronix in Switzerland) have the average size of 9 nm and the BET specific surface area of 160 m2/g. Note that the BET specific surface area is usually several times larger than the total surface area, because the multilayer (instead of mono-layer) adsorption occurs during the measurement. It was our speculation that electrospun porous TiO2 nanofibers (with the BET specific surface area of at least 100 m2/ g) in the form of overlaid fiber mat would outperform TiO2 nanoparticles, particularly for making the photo-anode in dye-sensitized solar cell, because the nanofibers would provide continuous pathways for the photo-generated electrons without substantial scarification of dye loading amount. To further increase the specific surface area of electrospun anatase-phase TiO2 nanofibers, several approaches were investigated during this study. The first approach was to vary the concentration of carrying polymer (i.e., PVP) and/or to introduce thermally degradable polymers such as polymethylmethacrylate (PMMA) into the spin dope. Such an approach was revealed not to be very effective, and the resulting TiO2 nanofibers generally had the BET specific surface area smaller than 50 m2/g. This was because the decomposition of polymers (e.g., PVP and PMMA) occurred prior to the homogeneous nucleation and growth of anatase-phase TiO2 crystallites during the pyrolysis [10]. In other words, small pores created from the decomposition of polymers would unlikely be retained in the final TiO2 nanofibers, because the TiO2 was structurally amorphous and relatively easy to move molecularly at the stage. The second approach was to prepare the hollow/tubular TiO2 nanofibers. Nevertheless, the BET specific surface area of hollow/tubular TiO2 nanofibers was still considerably lower than 100 m2/g, despite that the value was almost twice as that of solid TiO2 nanofibers. We therefore concluded that small pores had to be generated throughout the nanofibers after pyrolysis to substantially increase the specific surface of anatase-phase TiO2 nanofibers. As described in the section of Results, two approaches were investigated. The first approach was to prepare the composite spin dope containing both TNBT and the precursor of Al2O3 (i.e., Al[OCH(CH3)2]3); after the TiO2/Al2O3 composite nanofibers were made, the Al2O3 component would then be removed via chemical etching treatment with NaOH aqueous solution, resulting in the formation of nanoscale pores throughout the (final) TiO2 nanofibers. The second approach was to prepare the mixture spin dope containing both TNBT and ZnO nanoparticles (with particle sizes smaller than 30 nm); the hypothesis was that, during the sol–gel process of TNBT and the subsequent pyrolysis, the existence of ZnO nanoparticles would hinder the
27.3 (m2/g)
106.5 (m2/g)
148.6 (m2/g)
volumetric contraction/shrinkage, resulting in the formation of nanoscale pores. Furthermore, the ZnO component in TiO2/ZnO composite nanofibers could also be removed by using the NaOH aqueous solution, resulting in more nanoscale pores in the (final) TiO2 nanofibers. The acquired results indicated that the second approach led to porous TiO2 nanofibers with BET specific surface area of 148.6 m2/g, while the first approach led to porous TiO2 nanofibers with BET specific surface area of 106.5 m2/g. The descriptions of different electrospun TiO2 nanofibers are summarized in Table 1. It is noteworthy that the heating procedure in pyrolysis could have the significant impact on structure (particularly the formation and amount of defects), while it would not appreciably impact on the BET specific surface area. For example, with slower heating rate, the amount of structural defects in both solid and hollow/tubular nanofibers might be less, since the average length of the fibers/tubes was longer, whereas the variation of BET specific surface area was very small. We also attempted to prepare the hollow/tubular TiO2 nanofibers with pores in the walls by using the composite spin dope containing both TNBT and the precursor of a removable component (e.g., Al2O3 or ZnO); nonetheless, such an attempt was not very successful to further increase the BET specific surface area significantly. It also needs to clarify that, if the specific surface area of TiO2 nanofibers is too large (i.e., 300 m2/g or higher), the nano- and/or meso-scaled pores in the nanofibers might lead to too many surface trapping states; if such TiO2 nanofibers are used for making the photo-anode in dye-sensitized solar cell, the photo-generated electrons might not move along the nanofibers for a long distance before the recombination would occur.
5. Summary In summary, the systematic investigations were carried out to prepare and characterize electrospun anatase-phase TiO2 nanofibers with different morphological structures (e.g., solid, hollow/ tubular, and porous) and specific surface areas. The results indicated that electrospun TiO2 nanofibers with hollow/tubular structure had the BET specific surface area of 27.3 m2/g, which was approximately twice as that of solid TiO2 nanofibers (15.2 m2/ g). The porous TiO2 nanofibers made from the precursor of Al2O3/ TiO2 composite nanofibers had the BET specific surface area of 106.5 m2/g, while the porous TiO2 nanofibers made from the precursor of ZnO/TiO2 composite nanofibers had the highest BET specific surface area of 148.6 m2/g. It is envisioned that electrospun anatase-phase TiO2 nanofibers with desired morphological structure and relatively high BET specific surface area would outperform
G. He et al. / Journal of Colloid and Interface Science 398 (2013) 103–111
other nanostructures (e.g., powder and film) of TiO2 for various applications (particularly dye-sensitized solar cell and photocatalysis). Acknowledgments This research was supported by the National Science Foundation (Grant No.: EPS-0903804), the National Aeronautics and Space Administration (Cooperative Agreement No.: NNX10AN34A), and the State of South Dakota. Y. Cai would acknowledge the National Natural Science Foundation of China (Grant No.: 51006046), the Natural Science Foundation of Jiangsu Province (Grant No.: BK2010140), and the National High-Tech R&D Program of China (Grant No.: 2012AA030313). References [1] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735. [2] U. Bach, D. Lupo, P. Comte, J.E. Moser, F. Weissorte, J. Salbeck, H. Spreitzer, M. Gratzel, Nature 395 (1998) 583. [3] D. Vu, Z. Li, H. Zhang, W. Wang, Z. Wang, X. Xu, B. Dong, C. Wang, J. Colloid Interface Sci. 367 (2012) 429. [4] A. Wold, Chem. Mater. 5 (1993) 280. [5] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69. [6] A. Greiner, J.H. Wendorff, Angew. Chem. Int. Ed. 46 (2007) 5670.
111
[7] S. Ramakrishna, K. Fujihara, W.E. Teo, T. Yong, Z.W. Ma, R. Ramaseshan, Mater. Today 9 (2006) 40. [8] V. Thavasi, G. Singh, S. Ramakrishna, Energy Environ. Sci. 1 (2008) 205. [9] D.H. Reneker, I. Chun, Nanotechnology 7 (1996) 216. [10] R. Chandrasekar, L. Zhang, J.Y. Howe, N.E. Hedin, Y. Zhang, H. Fong, J. Mater. Sci. 44 (2009) 1198. [11] D. Li, Y.N. Xia, Nano Lett. 3 (2003) 555. [12] C. Tekmen, A. Suslu, U. Cocen, Mater. Lett. 62 (2008) 4470. [13] B. Ding, C.K. Kim, H.Y. Kim, M.K. Seo, S.J. Park, Fibers Polym. 5 (2004) 105. [14] S.H. Lee, C. Tekmen, W.M. Sigmund, Mater. Sci. Eng. 398 (2005) 77. [15] K. Nakane, K. Yasuda, T. Ogihara, N. Ogata, S. Yamaguchi, J. Appl. Polym. Sci. 104 (2007) 1232. [16] M.A. Kanjwal, N.A. Barakat, F.A. Sheikh, S.J. Park, H.Y. Kim, Macromol. Res. 18 (2010) 233. [17] A.S. Nair, S.Y. Yang, P.N. Zhu, S. Ramakrishna, Chem. Commun. 46 (2010) 7421. [18] D. Li, Y.N. Xia, Nano Lett. 4 (2004) 933. [19] J.T. McCann, D. Li, Y.N. Xia, J. Mater. Chem. 15 (2005) 735. [20] D. Li, J.T. McCann, Y.N. Xia, Small 1 (2005) 83. [21] Y.L. Cheng, W.Z. Huang, Y.F. Zhang, L. Zhu, Y.J. Liu, X.Z. Fan, X.Q. Cao, CrystEngComm 12 (2010) 2256. [22] R.A. Caruso, J.H. Schattka, A. Greiner, Adv. Mater. 13 (2001) 1577. [23] G.M. Kim, S.M. Lee, G.H. Michler, H. Roggendorf, U. Gosele, M. Knez, Chem. Mater. 20 (2008) 3085. [24] S.H. Zhan, D.R. Chen, X.L. Jiao, C.H. Tao, J. Phys. Chem. B 110 (2006) 11199. [25] S.H. Zhan, D.R. Chen, X.L. Jiao, Y. Song, Chem. Commun. 20 (2007) 2043. [26] M.A. Kanjwal, N.A. Barakat, F.A. Sheikh, M.S. Khil, H.Y. Kim, Int. J. Appl. Ceram. Technol. 7 (2010) 54. [27] Q. Li, D. Sun, H. Kim, Mater. Res. Bull. 46 (2011) 2094. [28] Z.Y. Liu, D.D. Sun, P. Guo, J.O. Leckie, Nano Lett. 7 (2007) 1081. [29] V. Stengl, J. Velicka, M. Marikova, T.M. Grygar, ACS Appl. Mater. Interfaces 3 (2011) 4014.