A novel nanostructure with hexagonal-prism pores fabricated under vacuum circumstance

Materials Research Bulletin 50 (2014) 209–212

Contents lists available at ScienceDirect

Materials Research Bulletin journal homepage: www.elsevier.com/locate/matresbu

A novel nanostructure with hexagonal-prism pores fabricated under vacuum circumstance Hongxin Tong a, Zhiyuan Cheng a, Xiang Lv a, Ye Song b,*, Dongliang Yu b, Xufei Zhu a,* a b

School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China Key Laboratory of Soft Chemistry and Functional Materials of Education Ministry, Nanjing University of Science and Technology, Nanjing 210094, China

A R T I C L E I N F O

A B S T R A C T

Article history: Received 9 August 2013 Received in revised form 18 October 2013 Accepted 21 October 2013 Available online 28 October 2013

It is well known that electrochemical anodizations of valve metals are performed under the normal pressure (1 atm = 0.1 MPa), and only columnar pores could be usually obtained in porous anodic alumina (PAA). So far, there have been very few reports involving the atmosphere pressure and its effect on the nanostructure of PAA. Here, a novel PAA nanostructure with hexagonal-prism pores has been successfully fabricated under a vacuum system (0.01 MPa), and its forming process is clarified by the anion-contaminated alumina model and oxygen bubble mold. The present results can provide unique insights into the inherent relations between the structural features and anodizing parameters in all anodizing process. ß 2013 Elsevier Ltd. All rights reserved.

Keywords: Microporous materials Nanostructures Oxides Thin films

1. Introduction Self-organized porous nanostructures of valve metal oxides have been extensively investigated due to their various applications [1–3]. The formation mechanisms and fabricating methods of porous anodic alumina (PAA) and porous anodic titania (PAT) have received considerable attention until now [4–8]. In order to fabricate various nanostructures, many techniques have been utilized, such as the well-known two-step anodization [4,9], hard and pulse anodization [4,10] and changing the anodizing parameters or electrolyte composition [4,11,12]. However, as Hebert et al. [8,13] indicated, no model has successfully explained the relationships between film morphology and processing parameters. And the influence of processing parameters (e.g., temperature, time, anodizing voltage and rotation rate) on the film morphology has been studied extensively [11–14]. However, the influence of atmosphere pressure on the film morphology has been ignored, because the pressure has no direct contribution to the film morphology according to the classical theories of pore formation mechanism [4,7,8,15–18]. All above anodizing processes were performed at the normal atmosphere (0.1 MPa). In fact, the pressure of electrolytic bath could influence the PAA morphology [19]. Moreover, the structural features of PAA and anion-contaminated

* Corresponding authors. Tel.: +86 25 84315949; fax: +86 25 84276082. E-mail addresses: [email protected] (Y. Song), [email protected] (X. Zhu). 0025-5408/$ – see front matter ß 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.materresbull.2013.10.038

alumina (ACA) layer around the pore walls have not been successfully elucidated so far [4,8,18,20,21]. We have proposed a new growth model of PAA, which emphasizes the close relationships among pore generation, ACA layer and oxygen bubble mold (OBM) [22,23]. In our latest work, a two-tier nanostructure with narrow mouth and wide abdomen has been successfully fabricated [12], which was clarified by the combination of viscous flow model [13,18] and oxygen bubble mold. Herein, in order to demonstrate the influence of oxygen evolution on the PAA morphology, two anodizing processes of aluminum under the normal pressure and vacuum system were investigated by comparison. As a result, a novel nanostructure with hexagonal-prism pores is first assembled by anodization under a vacuum system. These hexagonal-prism pores help to identify the dominating factors of pore generation and development. Based on the ACA model [20,21] and the OBM model [12,22,23], the forming process of the hexagonal-prism pores is discussed in detail. This new anodizing technology under vacuum system should be particularly useful for assembling special nanostructures. 2. Experimental Aluminum (99.99%) sheet of 0.2 mm thickness was employed. The sheet was dipped into 2 wt% NaOH solution at 70 8C for chemical polishing for 120 s and then electropolished at a constant current density of 120 mA cm 2 for about 90 s in an electrolyte composed of phosphoric acid (80 wt%), chromium trioxide

210

H. Tong et al. / Materials Research Bulletin 50 (2014) 209–212

(12 wt%) and water (8 wt%) at 70 8C. The electropolished sheets were anodized in an ethylene glycol solution containing 6 wt% phosphoric acid at constant current (15 mA cm 2) and the room temperature for 3600 s under two different pressures (0.1 MPa and 0.01 MPa). A digital thermometer (through a thermocouple) was used for measuring the temperature of the electrolyte. No stirring and cooling devices were adopted and the temperature rise of the electrolyte was less than 23 8C. The nanostructure was characterized by field-emission SEM (Hitachi S4800). To get the crosssection images, the anodized samples with Al substrate were bent into V-shape before observation. 3. Results and discussion Fig. 1 shows the micrographs of the nanostructure with hexagonal-prism pores in PAA obtained under vacuum condition (0.01 MPa). There are many hexagonal and polygonal pores in Fig. 1a. Only a few columnar pores, which have small porediameter and thick pore-wall, can be seen in Fig. 1a. This nanostructure was obtained by anodization under a vacuum system (0.01 MPa). However, different nanostructure with columnar pores in PAA anodized under the normal pressure (0.1 MPa) was obtained as shown in Fig. 2a and b. There are significant differences between Figs. 1a and 2a. Obviously, besides hexagonalprism and columnar pores, the pore-diameter and pore-wall are very different. The diameters are 270 nm and 100 nm in Figs. 1b and 2b, respectively. According to the classical growth models of PAA (e.g., filed-assisted dissolution model, volume expansion model and self-organized model) [4,15–18], it is obviously hard to explain the above differences. Because other anodizing conditions (e.g., electrolyte, temperature, anodizing current or voltage and time) are all the same except for the system pressure, these significant differences may be attributed to the different pressures in two anodizing systems, which can be clarified by the ACA model and oxygen evolution in the following text. In fact, when pure aluminum was anodized in 6 wt% H3PO4 ethylene glycol solution under the normal pressure (0.1 MPa), the evolution of oxygen bubbles on the anode and H2 bubbles on the cathode could be observed easily by the naked eyes, and Fig. 2c shows the difference of the bubbling gas between the anode and the cathode. Fig. 3 shows the difference of the bubbling gas on the anode under different pressures (0.1 MPa and 0.01 MPa) and at different moments. Fig. 3a and b show the different distributions of oxygen evolution around the anode at different moments under

Fig. 1. SEM images of the nanostructure with hexagonal-prism pores in PAA anodized for 3600 s under 0.01 MPa.

0.1 MPa. For the same anodizing process under a vacuum system (0.01 MPa), oxygen evolution became very intense and much more oxygen bubbles could be seen in Fig. 3c and d. Fig. 2d shows the voltage-time curves of the anodizing processes under 0.1 MPa and 0.01 MPa. Under the vacuum circumstance, intense gas bubbles will alter the electrical conductivity in the pores. The variation of bubble quantity can result in the change of anodizing voltage. Therefore, the voltage under 0.01 MPa is higher than that under 0.1 MPa. We consider that the hexagonal-prism pores in Fig. 1 result from both intense oxygen evolution and the existence of ACA layer around the columnar pore walls. In fact, the porous layer of PAA is composed of two structurally different regions, anion-free pure alumina and ACA layer [4,20,21], as shown in Fig. 4a. The ACA layer was first described in detail by Thompson and Wood [20]. Diggle et al. [24] considered that anions incorporation into the barrier oxide was closely associated with the degree of pore formation. However, this important viewpoint has been ignored for many years, because the existence of ACA layer is contrary to

Fig. 2. SEM images of the surface morphology (a) and the cross section (b) of the nanostructure with columnar pores in PAA anodized for 3600 s under 0.1 MPa. (c) Photograph showing the gas bubbles around the anode and the cathode under 0.1 MPa. (d) Voltage–time curves of the anodizing processes under 0.01 MPa and 0.1 MPa.

H. Tong et al. / Materials Research Bulletin 50 (2014) 209–212

211

Fig. 3. Photographs showing the gas bubbles around the anode at different moments under 0.1 MPa (a and b) and 0.01 MPa (c and d).

expectations of filed-assisted dissolution model [18]. Until 2006, Garcia-Vergara et al. [18] proposed that the incorporated anions may have key influences on the barrier oxide flow, thereby affecting cell and pore dimensions. The oxide flow model is compatible with anions incorporation [18]. Therefore, the migration of electrolyte species and ACA layer requires much more extensive studies [4,18,24]. However, due to the difficulties inherent to direct measurements, indirect methods were employed in some cases to identify the existence of ACA layer [17,21]. Based on the oxide flow and ACA models [18,20], we have proposed the oxygen bubble mold of PAA and the growth model of ionic current and electronic current [12,22,23]. Curioni et al. [25] recently indicated that ionic migration was responsible for oxide growth, and electronic conduction mainly resulted in oxygen evolution. At high electric fields, the electronic current is low for aluminum oxide (insulator) and comparatively high for titanium oxide (semiconductor) [25]. Inward-migrating oxygen ions can lose their electrons and be converted to O2, producing gas bubbles that are occluded within the oxide and become an obstacle to ionic migration [25,26]. Oxygen evolution has been observed directly during the anodizing processes of titanium and Al alloys [25,26]. Although oxygen bubbles generated in the nanopores cannot be seen by the naked eyes in most cases, many researchers also reported oxygen evolution in the anodizing process of pure aluminum [27,28].

Fig. 4. (a) Schematic diagram of the distributions of ACA layer and viscous flow of barrier oxide around the OBM at the pore bottom. (b) Schematic diagram shows the forming process of the hexagonal-prism pores.

In summary, it is evident that pore formation should be closely associated with ACA layer, electronic current and oxygen evolution. However, there are very few reports involving the combination of ACA and oxygen evolution. It is difficult to identify the existence of gas bubbles in the pore bottom practically, because the formation mechanism is hardly derived by direct in situ experimental methods [17]. Herein, as an indirect method, anodization under a vacuum system was first employed to identify the existence of gas bubbles. The system pressure, as a predominant infuencing factor on gas evolution, should be taken into account in the anodizing process. Fig. 4a shows the distributions of ACA layer and viscous flow of barrier oxide around the oxygen bubble at the pore bottom. The driving forces of ACA layer, barrier oxide and oxygen evolution are incorporation current ( jc), ionic current ( jion) and electronic current ( je), respectively [23,29,30]. Theoretical derivation of ionic current and electronic current, and comparison between fitting curves and measured curves in the anodizing processes of titanium and aluminum have been studied in detail in our recent work [30]. The ionic current jion mainly comes from the inward migrating O2 ions and outward migrating Al3+, leading to the barrier oxide growth, the white regions (Al2O3) as shown in Fig. 4a. The electrolyte species also migrate inward and contribute to the jc. The jc is considered to be a constant fraction g of jion (i.e., jc = gjion) [23,29,30], because the electrolyte species migrate slower than O2 ions. Once these contaminant species (e.g., PO43 ) are incorporated into the barrier oxide and result in the thin ACA layer, they behave as impurity centers and initiate the generation of je and continuous gas evolution [23]. And many viewpoints (e.g., amorphous to crystalline transition, impurities or second phase particles) about the generation of electronic current and oxygen bubbles have been proposed in other references [25–28]. More recently, Kruse et al. [31] demonstrated that oxygen bubbles formed at the anode had a strong effect on the porous layer formation, whereas grain structure, defect density and metal purity were not found to impact the pore forming process. In fact, the viewpoints about oxygen bubble and electronic current have been demonstrated and cited by other research groups [2,31–35]. According to Fig. 4a, oxygen bubbles must move upwards along the pore wall, thus the micro-gas-flows can be formed in the nanopore, and the ACA layer at the pore bottom is divided into some fragments by the rupture. The ACA fragments are carried by the micro-gas-flows from the pore bottom, which become easy to dissolve into the electrolyte. This is the main driving force for renovating the ACA layer at the pore bottom.

212

H. Tong et al. / Materials Research Bulletin 50 (2014) 209–212

4. Conclusion A novel nanostructure with hexagonal-prism pores and an improved anodizing technology under the vacuum system were presented for the first time. This special nanostructure can be clarified by the combination of the ACA layer and the OBM models. The hexagonal-prism pores result from the washout of ACA layer due to the continuous flush of the intense gas-flow and liquid-flow under vacuum system. Attention must be paid to both the pressure and the ACA layer, which may open new possibilities for assembling the porous nanostructures and for understanding the growing kinetics. Furthermore, to confirm the composition of the bubbles will help to clarify the formation mechanism of many porous anodic oxides. Acknowledgments This work was supported financially by the National Natural Science Foundation of China (Grant Nos. 61171043, 51077072) and the Priority Academic Program Development of Jiangsu Higher Education Institutions. Fig. 5. Two EDS images showing the different elements in the cross section of PAA with (a) hexagonal-prism pores and (b) columnar pores.

Another function of the micro-gas-flow is to renovate the electrolyte in the nanopore. The rising micro-gas-flow will carry a little electrolyte to rise. The upwards moving electrolyte then forms an up micro-liquid-flow in the nanopore. When the upward micro-gas-flow or micro-liquid-flow leaves the nanopore, the local vacuum is formed in the nanopore, and then the outer electrolyte will form another micro-liquid-flow which enters the nanopore. The markers of the up or down micro-liquid-flows are shown in Fig. 4a. The up and down micro-liquid-flows will renovate continuously the electrolyte. The renovation of the solution and electrolyte species ensures to renew the ACA layer at the bottom, thus, the driving forces of jion, jc and je can be maintained continually. Accordingly, the forming process of the hexagonal-prism pores can be schematically shown in Fig. 4b. (I) The columnar cells and pores are firstly originated by the oxide viscous flow around the OBM. A void is inevitably formed among three neighboring columnar cells. (II) The columnar cells are transformed into the hexagonal-prism cells with columnar pores, which can be explained by the repulsive forces due to the volume expansion, plastic flow and compressive stress [7,13,16,18]. Notably, the gaps or voids among neighboring cells are closely associated with the oxide expansion coefficients. So the voids are hardly observed in PAA and easily observed in TiO2 nanotubes. (III) The micro-gasflow and micro-liquid-flow are further enhanced by the vacuum anodization. The ACA layer around the columnar pore walls is inevitably flushed and continuously washed out by the gas-flow and liquid-flow. (IV) As a result, the ACA layer is swept into the bulk solution and the hexagonal-prism pores are formed after a long time under the vacuum anodization. However, the SEM and other experimental methods are not capable of giving direct evidence of the micro-gas-flow [17]. In order to provide qualitative information, two EDS images for the two samples shown in Figs. 1c and 2b are supplied in Fig. 5. The spectrums in Fig. 5a and b show the different compositions of the hexagonal-prism pores and the columnar pores, respectively. The phosphorus element in the hexagonal-prism pores is much less than that in the columnar pores. This result may indicate that the hexagonal-prism pores are made up of more anion-free alumina and less ACA layer.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

K.P. Lee, D. Mattia, J. Membr. Sci. 435 (2013) 52–61. C.J. Ingham, J. ter. Maat, W.M. de Vos, Biotechnol. Adv. 30 (2012) 1089–1099. S. Yeu, J.D. Lunn, H.M. Rangel, D.F. Shantz, J. Membr. Sci. 327 (2009) 108–117. G.D. Sulka, in: A. Eftekhari (Ed.), Nanostructured Materials in Electrochemistry, Wiley-VCH, Wienheim, 2008, pp. 1–116. D.M. Kanani, W.H. Fissell, S. Roy, A. Dubnisheva, A. Fleischman, A.L. Zydney, J. Membr. Sci. 349 (2010) 405–410. Y. Jo, I. Jung, I. Lee, J. Choi, Y. Tak, Electrochem. Commun. 12 (2010) 616–619. P. Roy, S. Berger, P. Schmuki, Angew. Chem. Int. Ed. 50 (2011) 2904–2939. K.R. Hebert, S.P. Albu, I. Paramasivam, P. Schmuki, Nat. Mater. 11 (2012) 162–166. Y. Li, Z.Y. Ling, S.S. Chen, J.C. Wang, Nanotechnology 19 (2008) 225604. Y. Liu, D. Wang, L. Cao, S. Chen, Electrochem. Commun. 23 (2012) 68–71. X. Hu, Z.Y. Ling, K. Wang, Y. Li, Electrochem. Commun. 13 (2011) 1082–1085. X.F. Zhu, Y. Song, D.L. Yu, C.S. Zhang, W. Yao, Electrochem. Commun. 29 (2013) 71–74. J.E. Houser, K.R. Hebert, Nat. Mater. 8 (2009) 415–420. S.P. Albu, P. Schmuki, Electrochim. Acta 91 (2013) 90–95. V.P. Parkhutik, V.I. Shershulsky, J. Phys. D: Appl. Phys. 25 (1992) 1258–1263. O. Jessensky, F. Mu¨ller, U. Go¨sele, Appl. Phys. Lett. 72 (1998) 1173–1175. G. Patermarakis, K. Moussoutzanis, Electrochim. Acta 54 (2009) 2434–2443. S.J. Garcia-Vergara, P. Skeldon, G.E. Thompson, H. Habazaki, Electrochim. Acta 52 (2006) 681–687. X.F. Zhu, L. Liu, Y. Song, H. Jia, H. Yu, X. Xiao, X. Yang, Mater. Lett. 62 (2008) 4038– 4040. G.E. Thompson, G.C. Wood, Nature 290 (1981) 230–232. L. Gonzalez-Rovira, M. Lopez-Haro, A.B.K. Hungria, El. Amrani, J.M. SanchezAmaya, J.J. Calvino, F.J. Botana, Corros. Sci. 52 (2010) 3763–3773. X.F. Zhu, L. Liu, Y. Song, H. Jia, H. Yu, X. Xiao, X. Yang, Monatsh Chem. 139 (2008) 999–1003. X.F. Zhu, Y. Song, L. Liu, C. Wang, J. Zheng, H. Jia, X. Wang, Nanotechnology 20 (2009) 475303. J.W. Diggle, T.C. Downie, C.W. Goulding, Chem. Rev. 69 (1969) 365–405. A. Mazzarolo, M. Curioni, A. Vicenzo, P. Skeldon, G.E. Thompson, Electrochim. Acta 75 (2012) 288–295. D. Regonini, A. Satka, A. Jaroenworaluck, D.W.E. Allsopp, C.R. Bowen, R. Stevens, Electrochim. Acta 74 (2012) 244–253. K. Schwirn, W. Lee, R. Hillebrand, M. Steinhart, K. Nielsch, U. Go¨sele, ACS Nano 2 (2008) 302–310. D.D. Li, L. Zhao, C.H. Jiang, J.G. Lu, Nano Lett. 10 (2010) 2766–2771. J.M. Albella, I. Montero, J.M. Martinez-Duart, Electrochim. Acta 32 (1987) 255– 258. R.Q. Yang, L.F. Jiang, X.F. Zhu, Y. Song, D.L. Yu, A. Han, RSC Adv. 2 (2012) 12474– 12481. A.D. Pauric, S.A. Baig, A.N. Pantaleo, Y. Wang, P. Kruse, J. Electrochem. Soc. 160 (2013) C12–C18. G. Patermarakis, J. Diakonikolaou, J. Solid State Electrochem. 16 (2012) 2921– 2939. W. Lee, K. Schwirn, M. Steinhart, E. Pippel, R. Scholz, U. Go¨sele, Nat. Nanotechnol. 3 (2008) 234–239. C.K. Chung, R.X. Zhou, T.Y. Liu, W.T. Chang, Nanotechnology 20 (2009) 055301. R. Zhang, K.M. Jiang, Y. Zhu, H.Y. Qi, G.Q. Ding, Appl. Surf. Sci. 258 (2011) 586–589.