Control over the Hierarchical Structure of Titanate Nanotube

Mar 31, 2011 - Agglomerates. Dmitry V. Bavykin,* Alexander N. Kulak, and Frank C. Walsh. Materials Engineering and Energy Technology Research Groups, ...
0 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/Langmuir

Control over the Hierarchical Structure of Titanate Nanotube Agglomerates Dmitry V. Bavykin,* Alexander N. Kulak, and Frank C. Walsh Materials Engineering and Energy Technology Research Groups, School of Engineering Sciences, University of Southampton, Southampton SO17 1BJ, United Kingdom

bS Supporting Information ABSTRACT: An alkaline hydrothermal treatment of several types of ordered macroporous TiO2 structures, namely, microtubes, sea urchin shapes, and anodic nanotube arrays, has been investigated under stationary conditions. The effect of the size and geometry of these structures on the morphology of the forming hierarchical agglomerates of titanate nanotubes has been systematically studied. It has been revealed that, at sizes larger than the critical value (ca. 1 μm), the whole geometry of the initial ordered TiO2 structure is maintained under reaction conditions leading to formation of hierarchical structures, in which bulk TiO2 is replaced with titanate nanotube agglomerates. This principle provides a convenient route for the preparation of multiscale micro- and nanostructures of TiO2 based materials. The analysis of critical size suggests that, under reaction conditions, due to the limited transport of dissolved Ti(IV) species, the growth of nanotubes occurs locally.

’ INTRODUCTION The tubular-shaped, solid hydrated form of TiO2 materials with characteristic diameter of a few nanometers, frequently referred to as titanate nanotubes1 (TiNT), has become a subject of many investigations including studies of their physicochemical2 and crystallographic3 properties, their mechanism of formation,4 and their applications.5 Since the discovery of TiNT by Kasuga et al.6 in 1997, many efforts have been made to control the morphological and geometrical properties of formed nanostructured titanates by adjusting the synthesis conditions, which include temperature, mass/liquid ratio, duration of the experiment, composition of the alkaline solution, and presence of additives.1 It has been revealed that increase of temperature in the range 110150 °C results in an increase in the average diameter of the nanotubes,7 whereas a further rise in temperature causes the formation of solid nanofibers8 instead of nanotubes. Effective ways to adjust the average length of the nanotubes include ultrasonic treatment of initial raw TiO29,10 or the improvement to some extent of mass transport conditions during alkaline hydrothermal treatment,11,12 which can affect the dynamics of nanotube growth in the axial direction due to the availability of dissolved titanium(IV) species. Although the effects of the synthetic conditions on the geometry of a single nanotube have been comprehensively studied,1 the principles governing the agglomeration of nanotubes and methods of regulation of the shape and geometry of such secondary structures have not been thoroughly investigated yet. Wider application of nanotubular titanates requires not only better control of morphology, surface properties, and crystal structure of single nanotubes, but also development of efficient and facile methods to manipulate their agglomerates including control of the order or r 2011 American Chemical Society

alignment of the nanotube bundle, packaging of nanotubes into a secondary structure, and development of hierarchical structures. Such hierarchical structures with a well-developed system of nanometer and submicrometer-sized pores applied as an electrode could potentially improve, for example, power characteristics of lithium batteries13,14 or current collection efficiency in dye sensitized solar cells,15,16 by improvement of the transport characteristics of ions in liquid phase and charge carriers in semiconductors. At present, a degree of control over the shape of the nanotubular agglomerates can be achieved by using hydrogen peroxide,17 adjusting the initial particle size distribution for the raw TiO2,9 or selecting suitable mass to volume ratio between initial TiO2 and NaOH.7 However, no comprehensive systematic studies regarding the formation and control of the geometry of the titanate nanotubes agglomerates obtained under alkaline hydrothermal condition have been reported yet. In this work, the correlation between morphology of initial TiO2 and shape of obtained titanate nanotubes agglomerates has been studied in a wide range of characteristic sizes of the structures; from tens of nanometers to tents of micrometers. Three types of ordered structures of TiO2 have been tested, namely, (a) macroporous tubes of amorphous TiO2 obtained by solgel hydrolysis of Ti(O-iPr)4 in the presence of NH3,18 (b) microstructured TiO2 synthesized in the pores of sea urchin templates,19 and (c) TiO2 nanotubes array prepared by anodization of titanium in fluoride-containing electrolyte.20,21 Received data allow identification of the smallest characteristic size of the Received: February 9, 2011 Revised: March 21, 2011 Published: March 31, 2011 5644

dx.doi.org/10.1021/la200527p | Langmuir 2011, 27, 5644–5649

Langmuir

ARTICLE

ordered structures above which the morphology of the initial material is maintained during the alkaline hydrothermal reaction.

’ EXPERIMENTAL PROCEDURES Sodium hydroxide (NaOH), potassium hydroxide (KOH), hydrochloric acid (HCl), sulfuric acid (H2SO4), and hydrogen peroxide (H2O2), titanium tetra iso-propoxy ester (Ti-(O-iPr)4), aqueous solution of NH3 (25 wt %), ethanol (EtOH), glycerol (Gly), and ammonium fluoride (NH4F) pure grade were all obtained from Aldrich and used without further purification. Titanium dioxide (P25, TiO2) was obtained from Degussa. Titanium foil was purchased from Alfa Aesar. The synthesis of macroporous tubes TiO2 was based on the method reported elsewhere.14 10 cm3 of Ti-(O-iPr)4 was slowly added to 100 cm3 of 5% wt aqueous solution of NH3 without stirring at room temperature. After 30 min, the white precipitates were separated by filtration, rinsed with water, and dried at 120 °C overnight. The method of preparation of sea urchin shaped microstructured TiO2 was adopted from the synthesis of macroporous inorganic solids using a natural sea urchin template.15 An ∼1 cm2 sea urchin skeletal plate was flooded with Ti-(O-iPr)4 at reduced pressure in the desiccator for 30 min at room temperature followed by transfer to the chamber saturated with water vapor (100% humidity), where it was left at room temperate for 12 h. The procedure was repeated 4 times for complete saturation of the pores of the sea urchin with amorphous TiO2. The removal of the calcium carbonate template was achieved by treatment of the obtained material with excess of 10% wt HCl followed by rinsing with water and drying at 120 °C overnight. The array of TiO2 nanotubes was prepared by anodic oxidation of titanium foil in waterglycerol electrolyte containing fluoride ions.22 5 cm2 of titanium foil was degreased by sonication with acetone and rinsed with distilled water. Anodic oxidation of titanium foil was undertaken in glycerolwater (90:10 vol) electrolyte containing 0.27 mol dm3 NH4F at room temperature. The electrochemical setup consisted of a two-electrode cell with graphite counter electrode and power supply TTi EX752M. The film of TiO2 nanotube array was prepared by anodization at 40 V for 5 h. The average current density was 2 mA cm2. After anodization, the film was thoroughly rinsed with distilled water. Transformation of all the above ordered structures of TiO2 into hierarchical structures of titanate nanotubes was undertaken under alkaline hydrothermal conditions using the binary NaOH/KOH aqueous mixture as a solvent.23 Two grams of ordered TiO2 was mixed with of 24 cm3 of 10 mol dm3 NaOH and 1 cm3 of 10 mol dm3 KOH aqueous solutions, then placed in a PFA (perfluoroalkoxy polymer) round-bottom flask (Bohlender GmbH) equipped with a thermometer and water jacketed condenser. The mixture was refluxed for 4 days at 110 °C without stirring. After reaction, the white powdery titanates were separated and thoroughly washed with distilled water until the washing solution reached pH 7. In order to convert titanate nanotubes into their protonated form, the powder was washed with an excess of 0.1 mol dm3 HCl for over 30 min until a stable pH value of 2, followed by water washing to pH 5. The sample was dried overnight in air at 120 °C. The morphology and structure of the obtained hierarchical titanate nanotubes were characterized using a JEOL 6500 FEG-SEM scanning electron microscope.

’ RESULTS AND DISCUSSION Figure 1 a shows an SEM image of amorphous TiO2 obtained by hydrolysis of Ti-(O-iPr)4 in the presence of NH3.14 The material is characterized by long, straight cylindrical channels (pores) originating from peculiarities of flow pattern of mixing reagents during hydrolysis of titanium alkoxide and arranged in an ordered structure. The characteristic diameter of the channel

Figure 1. SEM images of (a) microtubular amorphous TiO2 obtained via solgel hydrolysis of Ti(i-OPr)4 in the presence of NH3 and (b) hierarchical titanate nanotubes obtained from TiO2 microtubes via alkaline hydrothermal route. The insets correspond to magnification of the indicated areas.

varies in the range 13 μm, and the typical distance between channels is 24 μm. Higher magnification (see inset in Figure 1a) of the solid wall of the tube shows that it is composed of spheroidal particles of amorphous TiO2 of approximately 100 nm diameter, which consist of compactly packed, smaller ca. 35 nm particles (see Supporting Information Figure S1 a). A hydrothermal treatment of this material in aqueous mixture of KOH and NaOH without stirring results in transformation of spheroidal amorphous TiO2 to titanate nanotubes (see Figure 1b inset) with typical external diameter of 1015 nm and length 5645

dx.doi.org/10.1021/la200527p |Langmuir 2011, 27, 5644–5649

Langmuir

ARTICLE

Figure 3. SEM images of hierarchical CaTiO3 obtained via solgel hydrolysis of Ti(i-OPr)4 in the pores of sea urchin templates followed by alkaline hydrothermal reaction. The removal of a CaCO3 template between precipitation of TiO2 and alkaline treatment was omitted.

Figure 2. SEM images of (a) microstructured amorphous TiO2 obtained via solgel hydrolysis of Ti(i-OPr)4 inside sea urchin templates and (b) hierarchical titanate nanotubes obtained from TiO2 microstructures via alkaline hydrothermal route. The insets correspond to magnification of the indicated areas.

exceeding several hundred nanometers. Each nanotube has a characteristic multilayered wall structure and the internal diameter varies in the range 35 nm (see Supporting Information Figure 1S b). Figure 1b shows that these nanotubes are randomly agglomerated into the microtubular structure of original ordered TiO2. It is possible that an additional factor which stabilizes the agglomerate is the steric effect resulting from the elongated shape of the nanotubes. Similar behavior of retaining the original microstructure is observed during the transformation of amorphous structured

TiO2 obtained using a porous sea urchin shell as a template (see Figure 2). The echinoid material of sea urchin shells with characteristic interconnected channels of 1015 μm diameter and volume fraction 0.5 can be replicated using the solgel technique24 utilizing the hydrolysis of Ti-(O-iPr)4 and resulting in the formation of macroporous structures of amorphous TiO2 with template geometry similar to that of the original (see Figure 2a). The characteristic size of the interconnected channels in the sea urchin TiO2 is 1520 μm, whereas the size of TiO2 solid blocks is ca. 710 μm due to probable shrinkage during aging of the gel. The solid TiO2 also consists of amorphous globules 100 nm in diameter. The alkaline hydrothermal treatment of such sea urchin TiO2 does not change the macroporous structure and micrometer-scale geometry of the material; however, the solid TiO2 is replaced by a bundle of randomly oriented titanate nanotubes (see Figure 2b). High magnification (inset in Figure 2b) shows that titanate nanotubes are characterized by a morphology similar to the previous case. In both cases (hierarchical microtubes and sea urchin template structures), the obtained samples form pellet-shaped solid monoliths of several millimeters in size. Mechanically, these pellets are relatively fragile due to the brittle nature of the ceramics; however, it is possible to handle them. Most of the physicochemical properties of hierarchical TiNT structures are similar to that of randomly oriented TiNT.1,4 The annealing of TiNT results in their dehydration in the range of temperatures 100350 °C accompanied by transformation of titanates to monoclinic TiO2(B) followed by losing tubular morphology and formation of anatase nanorods at 400 °C. The XRD pattern of hierarchical nanotubes (see Supporting Information Figure S2) is also similar to that of randomly oriented nanotubes (usually attributed to one of the polytitanic acids3) indicating that nanotubes are randomly oriented in the 5646

dx.doi.org/10.1021/la200527p |Langmuir 2011, 27, 5644–5649

Langmuir

Figure 4. SEM images of (a) amorphous TiO2 nanotube array obtained by anodization of Ti in fluoride containing waterglycerol electrolyte and (b) film of random titanate nanotubes obtained from TiO2 nanotubes array via alkaline hydrothermal route. The insets correspond to magnification of the indicated areas.

agglomerates without the appearance of additional reflections in XRD at small angles caused by ordering of TiNT. The acid-assisted removal of the sea urchin template (mainly CaCO3) between stages of solgel precipitation of TiO2 and alkaline hydrothermal treatment is essential for successful preparation of hierarchical TiNT. Otherwise, the presence of CaCO3 during transformation of TiO2 to TiNT results in formation of spinel-type CaTiO3 nanocrystals instead, which are agglomerated into the macroporous ordered structure with the morphology of the original sea urchin template (see Figure 3). The nanocrystals are characterized by a hexagonal pellet shape, of 100 nm thickness and

ARTICLE

1 μm size. Most of the nanocrystals are assembled into the stack (see inset in Figure3). When the characteristic size of the ordered TiO2 structures is too small, the whole architecture of the solid material can collapse during transformation of titania to TiNT. Figure 4 a shows a TiO2 nanotube array obtained by electrochemical oxidation of titanium surface in waterglycerol electrolyte containing fluoride ions.18 The nanotubes are arranged in a distorted hexagonal packing. The diameter of the nanotube channel varies between 80 and 150 nm and the typical thickness of the wall is 35 nm at the top and ca. 20 nm at the bottom of the coating. The thickness of the film, i.e., the length of the nanotubes is approximately 12 μm. The alkaline hydrothermal treatment of such TiO2 nanotubes results in complete disappearance of the ordered structure of the TiO2 nanotube array and the formation of titanate nanotubes randomly oriented in spheroidal agglomerates (see Figure 4b). These agglomerates are randomly arranged into a porous film of TiNT; the size of each agglomerate is approximately 1 μm. The mechanism of transformation of TiO2 into TiNT under alkaline hydrothermal conditions consists of several stages including dissolution of initial TiO2, accompanied by crystallization of dissolved Ti(IV) into multilayer titanate nanosheets, which can fold into tubular structures.4,7,25 Crystallization of nanostructured titanate probably occurs at close proximity to initial TiO2 resulting in preservation of the original ordered structure. However, due to the diffusion of dissolved Ti(IV) in alkaline solution, such ordered geometry can be distorted. If the characteristic size (l) of initial ordered TiO2 is too small, then such random diffusion of Ti(IV) in solution can cause the total disappearance of the initial morphology and formation of randomly oriented TiNT without certain structural hierarchy (see Figure 5). A similar correlation between the shape of original TiO2 and the shape of TiNT agglomerates has been also observed for spheroidal particles of 500 nm26 or microspheres (46 μm).27 In contrast, the porous film of TiO2 obtained in the presence of poly(ethylene glycol) with characteristic pore size of 100 nm loses its structure after alkaline hydrothermal treatment, resulting in formation of interconnected bundles of TiNT.28 These data are in agreement with our observations allowing the estimation of the critical characteristic size of initial structure as 1 μm. Such hierarchical structures are only observed when an alkaline hydrothermal reaction is undertaken without stirring, which otherwise not only can cause formation of randomly oriented nanotubes,23 but also can result in formation of nanofibers.29 Since the temperature variation between the center and the edges of the reactor is less than 5 °C, the appearance of natural convective flow within the slurry can be excluded, leaving only a diffusive mechanism of transport of dissolved Ti(IV). The fact that 1 μm structures preserve their geometry under transformation of TiO2 to TiNT without stirring probably means that the length of the diffusion of dissolved Ti(IV) during such a transformation does not exceed 1 μm. Using the Einstein formula l2 = Dτ and the typical value for the diffusion coefficient in aqueous solution D = 105 cm2 s1, it is possible to estimate the time (τ) required for diffusion in such length as 103 s. The overall time required for complete reconstruction of TiO2 to TiNT in a motionless slurry is relatively long, approximately several hours. This could mean that the rate of crystallization of dissolved Ti(IV) into nanostructured titanate is high enough to limit the lifetime of Ti(IV) in solution to values smaller than 103 s. The combination of several factors including absence of stirring and rapid crystallization of aqueous Ti(IV) can probably result in 5647

dx.doi.org/10.1021/la200527p |Langmuir 2011, 27, 5644–5649

Langmuir

ARTICLE

Figure 5. Scheme of transformation of structured (with characteristic size of the structure l) TiO2 into a hierarchical structure at l > 1 μm or random at l < 1 μm bundle of titanate nanotubes.

Figure 6. Schematic drawing of the asymmetrical environment due to the gradient of Ti(IV) concentration stimulating a difference in crystallization rate for different layers and resulting in the curvature of peeled multilayered nanosheets.

an uneven distribution of Ti(IV) concentration causing a higher concentration near the surface of an amorphous TiO2 precursor. Such an inhomogeneity supports a peeling mechanism of nanosheet growth under alkaline hydrothermal conditions in which the multilayered nanosheets of sodium titanate are formed at the surface of initial TiO2 followed by their peeling off the surface.30 During sequential growth, the layers of titanate nanosheets located in close proximity to TiO2 are exposed to the higher concentration of Ti(IV), resulting in their higher growth rate due to accelerated crystallization. This can result in an imbalance in layer widths, which drives the curvature of multilayered nanosheets into nanotubes.7 Figure 6 schematically shows the mechanism of peeling of multilayered titanate nanosheets, followed by uneven crystallization on different layers resulting in their curvature into nanotubes. This mechanism of localized growth is also consistent with observations that stirring of the reactants promotes the formation of nanofibers instead of nanotubes24 due to a reduced gradient of Ti(IV) concentration, which causes homogeneous crystallization and growth of titanate nanosheet layers. Additional support suggesting the localized growth of nanotubes in the absence of stirring is that no nanotubes have been observed on the wall of the flask.

’ CONCLUSIONS Alkaline hydrothermal transformation of ordered TiO2 including microtubes and sea urchin-shaped and anodic arrays of

nanotubes into hierarchical structures of titanate nanotubes has been studied for different sizes of initial TiO2 at stationary conditions without stimulated stirring of the reacting slurry. Results have shown that when the characteristic size of the ordered structure is more than 1 μm, reaction leads to the formation of TiNT bundles agglomerated into the hierarchical structure with a similar shape to the original ordered TiO2. When the characteristic size is not as small, the obtained TiNT form random agglomerates. The analysis of the diffusion rate suggests an uneven distribution of dissolved Ti(IV) during reaction, resulting in an asymmetrical environment for multilayered titanate nanosheets which causes their curvature into nanotubes. The identification of the smallest characteristic size of the ordered TiO2 offers a new synthetic route for the preparation of hierarchical TiNT, which can be useful in a wider range of applications including lithium batteries, dye sensitized solar cells, biomedical coatings, and so forth. Further systematic studies of the effect of diffusive and convective mass transport during alkaline hydrothermal reaction on the formation of hierarchical structures could extend the range of possible shapes.

’ ASSOCIATED CONTENT

bS

Supporting Information. Additional figures as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*To whom correspondence should be addressed. Tel: þ 44 2380598358, Fax: þ 44 2380598754, e-mail: D.Bavykin@soton. ac.uk.

’ ACKNOWLEDGMENT The authors gratefully acknowledge financial support from the EPSRC, UK (grant EP/F044445/1: “A hydrothermal route to metal oxide nanotubes: synthesis and energy conversion applications”). Authors acknowledge Prof. F. C. Meldrum and Dr. Y. Y. Kim from School of Chemistry, University of Leeds, for providing samples of sea urchin skeletal plates and useful discussion. 5648

dx.doi.org/10.1021/la200527p |Langmuir 2011, 27, 5644–5649

Langmuir

ARTICLE

’ REFERENCES (1) Bavykin, D. V.; Walsh, F. C. Titanate and Titania nanotubes: Synthesis, Properties and Applications; Royal Society of Chemistry, 2010. (2) Chen, X.; Mao, S. S. Chem. Rev. 2007, 107, 2891–2959. (3) Chen, Q.; Peng, L.-M. Int. J. Nanotechnol. 2007, 4, 44–65. (4) Bavykin, D. V.; Friedrich, J. M.; Walsh, F. C. Adv. Mater. 2006, 18, 2807–2824. (5) Bavykin, D. V.; Walsh, F. C. Eur. J. Inorg. Chem. 2009, 8, 977–997. (6) Kasuga, T.; Hiramatsu, M.; Hoson, A.; Sekino, T.; Niihara, K. Langmuir 1998, 14, 3160–3163. (7) Bavykin, D. V.; Parmon, V. N.; Lapkin, A. A.; Walsh, F. C. J. Mater. Chem. 2004, 14, 3370–3377. (8) Yuan, Z. Y.; Su, B. L. Colloids Surf., A 2004, 241, 173–183. (9) Viriyaempikul, N.; Sano, N.; Charinpanitkul, T.; Kikuchi, T.; Tanthapanichakoon, W. Nanotechnology 2008, 19, 035601–6. (10) Bavykin, D. V.; Walsh, F. C. J. Phys. Chem. C 2007, 111, 14644–14651. (11) Kukovecz, A.; Hodos, M.; Horvath, E.; Radnoczi, G.; Konya, Z.; Kiricsi, I. J. Phys. Chem. B 2005, 109, 17781–17783. (12) Torrente-Murciano, L.; Lapkin, A. A.; Chadwick, D. J. Mater. Chem. 2010, 20, 6484–6489. (13) Long, J. W.; Dunn, B.; Rolison, D. R.; White, H. S. Chem. Rev. 2004, 104, 4463–4492. (14) Armstrong, G.; Armstrong, A. R.; Canales, J.; Bruce, P. G. Chem. Commun. 2005, 19, 2454–2456. (15) Wang, Z. S.; Kawauchi, H.; Kashima, T.; Arakawa, H. Coord. Chem. Rev. 2004, 248, 1381–1389. (16) Uchida, S.; Chiba, R.; Tomiha, M.; Masaki, N.; Shirai, M. Electrochemistry 2002, 70, 418–420. (17) Mao, Y.; Kanungo, M.; Hemraj-Benny, T.; Wong, S. S. J. Phys. Chem. B 2006, 110, 702–710. (18) Collins, A. M.; Carriazo, D.; Davis, S. A.; Mann, S. Chem. Commun. 2004, 568–569. (19) Yue, W.; Park, R. J.; Kulak, A. N.; Meldrum, F. C. J. Cryst. Growth. 2006, 294, 66–77. (20) Grimes, C. A. J. Mater. Chem. 2007, 17, 1451–1457. (21) Macak, J. M.; Tsuchiya, H.; Ghicov, A.; Yasuda, K.; Hahn, R.; Bauer, S.; Schmuki, P. Curr. Opin. Solid State Mater. Sci. 2007, 11, 3–18. (22) Macak, J. M.; Hildebrand, H.; Marten-Jahns, U.; Schmuki, P. J. Electroanal. Chem. 2008, 621, 254–266. (23) Bavykin, D. V.; Cressey, B. A.; Light, M. E.; Walsh, F. C. Nanotechnology 2008, 19, 275604–5. (24) Meldrum, F. C.; Colfen, H. Chem. Rev. 2008, 108, 4332–4432. (25) Zhang, S.; Peng, L. M.; Chen, Q.; Du, G. H.; Dawson, G.; Zhou, W. Z. Phys. Rev. Lett. 2003, 91, 256103–256107. (26) Tang, Y.; Yang, L.; Chen, J.; Qiu, Z. Langmuir 2010, 26, 10111–10114. (27) Zheng, Z.; Huang, B.; Qin, X.; Zhang, X.; Dai, Y. Chem.—Eur. J. 2010, 16, 11266–11270. (28) Zhang, H.; Liu, P.; Wang, H.; Yu, H.; Zhang, S.; Zhu, H.; Peng, F.; Zhao, H. Langmuir 2010, 26, 1574–1578. (29) Bavykin, D. V.; Kulak, A. N.; Walsh, F. C. Cryst. Growth Des. 2010, 10, 4421–4427. (30) Yang, J. J.; Jin, Z. S.; Wang, X. D.; Li, W.; Zhang, J. W.; Zhang, S. L.; Guo, X. Y.; Zhang, Z. J. Dalton Trans. 2003, 3898–3901.

5649

dx.doi.org/10.1021/la200527p |Langmuir 2011, 27, 5644–5649