Preparation of anodized aluminium oxide at high temperatures using low purity aluminium (Al6082)

Preparation of anodized aluminium oxide at high temperatures using low purity aluminium (Al6082)

Surface & Coatings Technology 378 (2019) 124970 Contents lists available at ScienceDirect Surface & Coatings Technology journal homepage: www.elsevi...

4MB Sizes 0 Downloads 7 Views

Surface & Coatings Technology 378 (2019) 124970

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage:

Preparation of anodized aluminium oxide at high temperatures using low purity aluminium (Al6082)


A.E. Kozhukhova, S.P. du Preez , D.G. Bessarabov HySA Infrastructure Centre, Faculty of Engineering, North-West University (NWU), Private Bag X6001, Potchefstroom, 2520, South Africa



Keywords: Anodized aluminium oxide Oxalic acid Low purity aluminium Al6082

A rapid two-step anodization process was developed to prepare semi-ordered nanoporous anodized aluminium oxide (AAO) layers using an inexpensive aluminium (Al) alloy, Al6082 (97.53% Al) as the Al source material. Al anodizing was performed at various anodization voltages (30, 45, 60 V) and temperatures (20, 30, 40 °C) using 0.4 M oxalic acid as the electrolyte. The effects of temperature and voltage on the morphological characteristics of the obtained AAO formed by one- and two-step anodization were investigated. The obtained AAO surfaces were characterized by scanning electron microscopy. Morphological characteristics of importance here were the pore diameter, inter-pore distance, porosity and pore density. An AAO layer with a semi-ordered pore arrangement was prepared using a two-step anodization process, which included an AAO etching step before the second anodization. The obtained AAO had a pore diameter of 43.8 ± 6.0 nm, inter-pore distance of 82.6 ± 19 nm, 25% porosity, pore density of 169 pores/μm2 and layer thickness of 53 μm. Energy-dispersive Xray spectroscopy results suggested that non-Al elements present in Al6082 were present on the surface of AAO. From an economic perspective, the AAO preparation process proposed in this study makes the fabrication of AAO more attractive.

1. Introduction It is becoming increasingly attractive to use aluminium (Al) and its alloys as the source material for the fabrication of catalyst support nanostructures. By subjecting Al to anodic anodizing, an oxide film with hexagonally arranged pores forms on its surface. This oxide layer has increased hardness, thermal conductivity, abrasion and corrosion resistance compared with Al's native oxide layer [1]. Anodized aluminium oxide (AAO) is a structure with a large surface area, suitable for the preparation of various catalysts and for support applications [2,3]. For instance, AAO can be used to prepare nanoparticles (Au [4], Pd [5]), nanotubes (TiO2 [6]; Fe, Co, Ni, with an outer diameter of 50–100 nm [7]; Au and multi-segmented AueNi nanotubes [8]) and nanodots (Ni [9], Cu [10], Au and AueAg [11]). In addition, AAO structures have become widely used in various environmental catalysis applications. One such example is the use of noble metals supported on honeycomb-structured AAO catalysts; they are used to reduce toxic NOx gases, and oxidized volatile organic compounds and CO gas present in exhausts gas originating from vehicles and industrial processes [12,13]. Application of the porous Al support for ruthenium-based catalysts for ammonia decomposition microreactors has also been demonstrated [14].

The choice of the electrolyte and the electrochemical parameters (e.g., temperature, anodizing time, electrolyte type/concentration, voltage and current) governs the morphological characteristics of AAO, thereby providing the necessary morphological properties depending on the intended application. The most commonly applied electrolytes for Al anodizing are sulphuric acid [15,16], oxalic acid [17,18] and phosphoric acid [19,20]. Specialized AAO structures, other than hexagonally arranged pores, may be obtained by applying a modulated signal during the anodization process, or by imprinting the surface of Al with an imprint stamp prior to anodization [21,22]. The two-step anodization process proposed by Masuda and Fukuda afforded highly ordered porous AAO when using high purity Al (99.99% Al) with pore diameters of 67–99 nm and a layer thickness of approximately 1.2 μm [23]. This method allowed the fabrication of AAO with highly ordered pores when using 0.3 M oxalic acid as the electrolyte. However, this AAO fabrication method required an anodization temperature and time of 0 °C and 160 h, respectively. In practical terms, maintaining 0 °C for extended periods of time is an energy intensive process. The Al anodization process may be accelerated by increasing the electrolyte concentration and/or temperature. By doing so, the current density is increased, which accelerates Al's anodization process. It is common to perform anodization under high-field (hard)

Corresponding author. E-mail addresses: [email protected] (S.P. du Preez), [email protected] (D.G. Bessarabov). Received 25 June 2019; Received in revised form 16 August 2019; Accepted 8 September 2019 Available online 10 September 2019 0257-8972/ © 2019 Elsevier B.V. All rights reserved.

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

form a thin protective oxide layer on the surface to avoid pitting and non-uniform oxide layer formation, by preventing localized high and heterogeneous electrical current flows [17]. The parameters chosen for the first step of the two-step anodization procedure are widely applied [18,34–37]. Though, a study by Sulka and Stepniowski stated that the pore order of two-step anodized AAO layers could be improved by increasing the time of the first anodization step from 30 to 60 min [34]. Considering a general goal of this study was to prepare AAO layers rapidly, a more conservative approach was applied. Thus, the time of the first anodization step was kept to 30 min and the conditions for the first anodization step of the two-step anodization procedure were kept constant. The distance between the working and counter electrode was kept at 3 cm. An adjustable DC power source (RS PRO, IPS603) with a voltage range of 0–60 V and current range of 0–5 A was used as the power source. On completion of the anodization procedures, samples were thoroughly rinsed with deionized water and stored in an air-tight container until further use. The morphology and/or surface chemical composition of Al6082 and AAO were determined by scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS), respectively. A FEI Quanta 200 scanning electron microscope with an integrated Oxford Instruments INCA 200 energy dispersive X-ray spectroscopy microanalysis system was used. All micrographs were prepared in secondary electron mode. AAO layer thickness was determined by abrading the side of a sample with 1200 grit sanding paper. The abraded surface was then polished using a MD-Dac cloth assisted by 3 μm polishing paste; followed by polishing using MD-Nap cloth and 1 μm polishing paste until a mirror finish was achieved. Polishing cloths and pastes (DiaDuo-2) were obtained from Struers, Denmark. An image processing program ImageJ [38–40] was used to determine pore diameters (Dp), inter-pore distance (Dc), and AAO layer thickness from the obtained SEM micrographs. Parameters such as porosity (P) and pore density (n) were used to characterize AAO structure. P is defined as a ratio of a surface area occupied by pores to the whole surface area. For a hexagonally arranged nanostructure, P can be calculated from Eq. (1) [41]:

anodization conditions [17]. Furthermore, AAO fabrication using expensive high purity Al (> 99.9%) as the Al source material is disadvantageous for large-scale commercial AAO fabrication due to economic considerations. Significant cost savings can be achieved by using Al of lower purity. Several studies have been conducted using AA1050 Al (99.5% Al) [24–30]. However, as described in Refs [24–29], the required AAO morphological characteristics were not achieved. Furthermore, cracks and defects appeared on the AAO surface. Fernandez-Romero et al. found that 1 h was the optimal AAl1050 anodization time to minimize the number of surface flaws and to avoid the local dissolution of AAO [28]. The reason is that the presence of non-Al elements in low purity Al alloys affects the formation of AAO layers [31,32]. Studies by FratilaApachitei et al. found that in the presence of Si and Fe, the morphology of the AAO layer contained deflected pores [31,32]. Furthermore, the presence of Cu promotes cracking of the AAO layer, which then allows the electrolyte to access the underlying Al surface, facilitating unwanted anodizing [33]. If the successful use of AAO produced from low purity Al for catalyst preparation and support applications is to be realized, it is important that a relatively high degree of pore order be achieved. Therefore, identifying the optimal anodization conditions and procedure for anodizing low purity Al is an urgent task. Michalska-Domanska et al. found no significant difference in the pore size distribution and pore order of AAO prepared from 99.999% and 99.5% pure Al in a solution of sulphuric acid, water and glycol [30]. They obtained relatively ordered AAO. However, this anodization process required an anodization temperature and time of −1 °C and 18 h, respectively. In this study, we present a rapid AAO fabrication method to prepare semi-organized AAO using the Al6082 alloy (97.53% Al) as the Al source material. The Al6082 Al alloy is approximately four orders of magnitude less costly than 99.99% pure Al, making it an economically feasible Al source material for large-scale AAO fabrication. We explored the use of one- and two-step anodization, using 0.4 M oxalic acid as the electrolyte. The effects of the anodization temperature (20, 30, 40 °C) and voltage (30, 45, 60 V) on the morphological features of the AAO layers were investigated, as well as the effects of AAO etching. It was determined that the etching step is crucial in obtaining AAO with a semi-ordered pore distribution.

Dp 2 P = 0.907 ⎛ ⎞ (%) ⎝ Dc ⎠ ⎜


The P value of AAO can be easily controlled by the anodization parameters. Factors governing P include the rate of oxide growth and dissolution in electrolyte. The n is defined as an overall number of pores per μm2 and is expressed by Eq. (2) [41]:

2. Experimental Engineering grade Al (Al6082) was obtained from Aluminium and Metal Traders, South Africa. The chemical composition of the Al was as follows: 97.53% Al, 0.96% Si, 0.7% Mg, 0.54% Mn, 0.19% Fe and ~0.08% other elements (C, Cr, Zn and Ti). The Al had a thickness of 0.2 cm and it was cut into 1 cm × 3 cm samples (7.6 cm2 total surface area). Prior to the anodization procedures, these samples were degreased by ultrasonication in acetone for 5 min and then in water for 5 min. Thereafter, samples were electropolished at room temperature in a 1:4 v/v mixture of HClO4:ethanol for 1 min at a constant voltage of 20 V. Samples were then thoroughly rinsed with deionized water. All anodization procedures were performed using a two-electrode electrochemical cell with a 5 cm × 15 cm stainless steel grid as a cathode (counter electrode). Isothermal conditions were maintained using a refrigerating–heating circulating system (CORIO CD-200F, Julabo, Germany). The electrolytic solution was continuously stirred with a magnetic stirrer. Two anodization procedures were utilized: i) pre-treated samples were anodized for 1 h in a 0.4 M oxalic acid solution at temperatures of 20, 30 and 40 °C and at 30, 45 and 60 V, and ii) pre-treated samples were anodized at 30 °C and 30 V for 30 min, followed by etching, by submerging the anodized samples in a 5% H3PO4/1.8% CrO3 etching solution. Thereafter, a second anodization step was performed as described in (i). Prior to the second anodization step, samples were anodized at 30 °C and 30 V for 10 min in the 0.4 M oxalic acid solution to


2 ∗ 10 6 (pores/μm2) 3 × D2c


As can be expected from Eqs. (1) and (2), a decrease in Dc leads to an increase in the number of pores formed within the AAO layer. 3. Results and discussion 3.1. Current density of low purity Al During Al anodization, the current is typically dynamic and depends on the process occurring on Al's surface, which can be divided into three stages: i) dense surficial alumina layer formation, ii) pore nucleation at alumina layer concavities, and iii) pore growth and fieldassisted localized alumina dissolution [30,42]. An additional stage has been observed by Michalska-Domanska and Norek [30]. This stage, referred to as Stage iv), can be identified by an increase in current density between Stages ii) and iii). The occurrence of Stage iv) can be ascribed to the incorporation of electrolyte anions to the AAO layer. Such incorporation replaces oxygen atoms in the AAO layer and can significantly increase the unit cell size, which in the case of sulphuric acid based anodization resulted in localized AAO layer deformities caused by hillock formation [30]. Fig. 1 presents the current 2

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Fig. 1. Current density (mA/cm2) as a function of time of samples anodized at constant voltages of 30, 45 and 60 V and temperatures of 20, 30 and 40 °C.

density–time transient of the first anodization step of samples anodized at temperatures of 20, 30 and 40 °C and voltages of 30, 45 and 60 V. The insert included in Fig. 1 presents the current density-time transients during the first 45 s of anodization. Relevant stages occurring during anodization was included. The anodization time was kept at 1 h. Fig. 1 shows that all of the samples followed current evolution, i.e. Stage i) – iii). Here, the 40 °C–45 V, 40 °C–60 V and 30 °C–60 V samples included Stage iv). The importance hereof is contextualized in Section 3.2. Bruera et al. made relatively similar current density-time transient observations using Al1050 as the Al source and using similar anodization parameters [43]. Fig. 1 further shows that samples anodized at high temperature and voltage exhibited higher current densities after several minutes of anodization. For example, the 40 °C–45 V, 40 °C–60 V and 30 °C–60 V samples had current densities of between 46.6 and 104.6 mA/cm2 after 5 min of anodization. The current densities of these samples showed a near exponential decrease as a function of time. The anodization current is related to the mobility of ionic O2−, OH– and Al3+ species through the oxide layer at the bottom of the pores [17,44,45]. According to Lee et al., the current density is determined by the mass transport of oxygen-containing anodic species from the electrolyte to the oxide/metal interface [17]. Thus, the current density is expected to decrease as the diffusion path increases (i.e., the oxide layer thickness increases). This was observed for the 40 °C–45 V, 40 °C–60 V and 30 °C–60 V samples. Samples anodized under milder conditions (all samples anodized at 30 V, 20 °C–45 V, 30 °C–45 V and 20 °C–60 V) had current densities of ≤29.3 mA/cm2 after 5 min of anodization. These samples did not exhibit any significant decrease in current density after 1 h of anodization if compared to samples anodized at 40 °C–45 V, 40 °C–60 V and 30 °C–60 V, which suggested that the diffusion path remained relatively short. According to reports by Lee et al., Sulka and Stepniowski and Bruera et al., the AAO layer thickness of samples anodized under similar conditions were in the range 4–144 μm, approximately [17,34,43]. The AAO layer thickness of samples prepared here were presented in Fig. 2. The layer thickness ranged between 6 and 67 μm and increased linearly with an increase in anodization voltage for each of the temperatures. The thicknesses reported here coincided with results presented by Refs. [17, 34, 43] using pure Al (≥99.5% Al) as the Al source.

Fig. 2. AAO layer thickness of samples anodized at 20, 30 and 40 °C and 30, 45, 60 V for 1 h.

3.2. Morphology of one-step anodized samples Low purity Al will not react in altogether the same way as pure Al during anodization due to the inclusion of various non-Al elements (hereafter referred to as ‘impurities’). According to Fratila-Apachitei et al., the presence of impurities in Al alloys can adversely affect the AAO morphology by preventing the formation of uniform pores [32,46]. As a result of this interfering effect, branching and deflection of pores can take place, and cracks and flaws can be formed. To elucidate the effects of impurities on the formation kinetics of AAO prepared from Al6082, the surface morphology of samples prepared according to the anodization procedure (i) (Section 2) was characterized by SEM. Fig. 3 shows representative micrographs of such samples anodized at temperatures of 20 to 40 °C and voltages of 30 to 60 V. Fig. 3 shows that the overall pore distribution is disordered, which is common for one-step anodized Al. Chung et al. prepared AAO layers with a similar pore distribution than the layers presented in Fig. 3 using high purity Al as the Al source [47]. In most cases, porous AAO layers were obtained, except (to a certain extent) for the surfaces of the 40 °C–45 V (Fig. 3b) and 40 °C–60 V (Fig. 3c) samples. Considering that 3

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Fig. 3. SEM micrographs of Al6082 anodized at 40 °C–30 V (a), 40 °C–45 V (b), 40 °C–60 V (c), 30 °C–30 V (d), 30 °C–45 V (e), 30 °C–60 V (f), 20 °C–30 V (g), 20 °C–45 V (h), and 20 °C–60 V (i). Anodization time was kept at 1 h. All images were recorded at 100,000 times magnification, and the scale bar was 1 μm for all images.

deformed the AAO layer significantly [52]. In another study by the authors, a 0.5 wt% Mn containing Al alloy was used for formation wellordered porous oxide layer; it was clear that Mn did not prevent the formation of ordered pores [53]. When considering Fig. 3, there was no evidence that the presence of impurities severely affected the AAO layers that were obtained and that these layers were comparable to typical one-step anodized AAO layers prepared from pure Al, as reported elsewhere [17,47,54–61]. The strand-like material observed on the surfaces of the 40 °C–45 V (Fig. 3b) and 40 °C–60 V (Fig. 3c) samples was alumina nanowires that formed during the anodization process [62,63]. These nanowires originated from the corners of the hexagonal cells because the walls of the AAO pores were dissolved as a result of extensive anodization. At higher current densities, the AAO dissolution process is preferred over AAO formation. Thus, the presence of nanowires is an indication of over-anodization. No such nanowires were observed on the surface of the 40 °C–30 V (Fig. 3a) sample, suggesting that AAO formation was favoured under the anodization conditions considered. Increasing the anodization potential to 45 V resulted in significant nanowire formation (Fig. 3b). These nanowires covered the majority of the surface of the AAO layer, and the underlying porous structure was only revealed in some areas. The nanowires observed for 40 °C–60 V (Fig. 3c) represents nanowires that have undergone extensive dissolution. By dissolving the surficial nanowires, the underlying porous structure was revealed and it is evident that these pores were relatively well ordered. Furthermore,

Si, Mn and Mg were the largest impurities of Al6082, it was expected that the presence thereof would adversely affect the AAO layer morphology. Fratila et al. found that the Si content of Al-10 wt% Si afforded the formation of SiO2 particles, disorganizing the AAO layer [31]. Ganley et al. prepared semi-ordered AAO layers from the Al alloy Al1100 (0.65 wt% Si) [48]. In this case, SiO2 particle formation and the detrimental effect thereof were not observed for any of the surfaces presented in Fig. 3; likely due to the low Si content of Al6082. Chung et al. stated that anodized Mg-rich Al alloys (such as AA5052, 96.6 wt% Al-2.6. wt% Mg) contained surficial sub-holes caused by branching, and voids in the vertical pore structure. The authors ascribed the latter occurrences to heat and hydrogen formation to the exothermic hydrolysis of Mg. The hydrogen bubbles prevented heat transfer from the immediate area to its surroundings; causing heat accumulation that disproportionately accelerated AAO layer formation [49]. Liu et al. found that Mg could be incorporated into the AAO layer and were subsequently dissolved at the AAO layer surface:electrolyte interface [50]. Habazaki et al. stated that AAO films can become detached from Zn and Mg binary alloys [51]. As with Si, the detrimental effect of Mg was not observed in Fig. 3 and it is also likely due to the low Mg content of Al6082. The effect of the third largest alloying element, Mn, has also been investigated. Voon et al. prepared AAO layers with ordered pore arrangements using Al-x wt% Mn (x = 0.5 to 2) alloys. The authors observed voids and cracks within the AAO layer in alloys containing > 0.5 wt% Mn caused by secondary phase formations, which 4

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Table 1 Morphological parameters (Dp, Dc, P and n) of AAO obtained during one-step anodization of Al6082. Temperature (°C)




a b


Dp (nm) Dc (nm) P (%) n (pores/μm2) Dp (nm) Dc (nm) P (%) n (pores/μm2) Dp (nm) Dc (nm) P (%) n (pores/μm2)

Voltage (V) 30



8.7 ± 3.0 53.0 ± 27.0 2 410 14.2 ± 4.5 62.8 ± 25.6 5 293 19.3 ± 7.1 92.2 ± 34.6 4 136

9.2 ± 3.0 53.4 ± 27.5 3 404 18.1 ± 6.1 69.5 ± 29.7 6 239 88.6 ± 8.5a 116.5 ± 20.7a 52b 85

11.6 ± 3.5 53.4 ± 18.9 4 405 24.6 ± 8.5 79.2 ± 42.1 9 184 133.4 ± 15.3a 146.7 ± 17.1a 75b 54

Manually determined using ImageJ. Determined from area with limited number of nanowires.

Fig. 4. SEM micrographs of Al6082 anodized at 40 °C–60 V, for 15 min (a), 30 min (b, 50,000× magnification) and 45 min (c, 50,000× magnification), and of Al6082 anodized at 40 °C–45 V for 15 min (d), 30 min (e) and 45 min (f). All images (excluding b and c) were recorded at 100000 times magnification, and the scale bar was 1 μm for all images.

samples anodized at 40 °C showed an increase in Dc from 92.2 to 146.7 nm. Is it further evident from Table 1 that the P remained ≤9% for all samples, except for the 40 °C–45 V and 40 °C–60 V samples, which had values of 52 and 75%, respectively. Furthermore, the samples anodized at 20 °C had a consistent n value of approximately 405 pores/μm2. With an increase in anodization potential, the samples anodized at 30 and 40 °C showed steady decreases in n of 293 to 184 pores/μm2 and 136 to 54 pores/μm2, respectively. The decrease in the determined n value was expected because n is inversely proportional to Dc, as indicated in Eq. (2). The Dp, Dc and n values presented in Table 1 were in agreement with the results presented by Bruera et al., using 99.5% Al and 0.3–0.9 M oxalic acid [43]. The following deductions regarding the effects of anodization on the morphological parameters were made from Fig. 3 and Table 1: i) the morphological characteristics of samples anodized at 20 °C were not appreciably affected by an increase in anodization voltage, ii) steady changes in the morphological parameters of samples anodized at 30 °C

the unusual current densities (Stage iv) observed for Fig. 1 did not result in the formation of hillocks. Thus, if electrolyte anions were incorporated in the AAO layer, the presence thereof had an insignificant effect on the layer morphology. The average values for Dp, Dc, P and n of the samples presented in Fig. 3 were determined. Results are tabulated in Table 1. Table 1 shows that the prepared one-step AAO layers had the following values: Dp 8.7 ± 3–133.4 ± 15.3 nm and Dc 53.0 ± 27.0–146.7 ± 17.1 nm. The values reported in Table 1 were similar to results published by Bruera et al. [43]. The samples anodized at 40 °C–45 V and 40 °C–60 V yielded AAO layers with Dp values of 88.6 and 146.7 nm, respectively. It is probable that these layers may have been subjected to a certain degree of pore widening during the late stages of anodization [64,65]. All other anodization parameters that we investigated yielded pores with diameters of < 24.6 nm. No significant increase in Dc was observed for samples anodized at 20 °C; values remained at approximately 53 nm. The Dc of samples anodized at 30 °C increased from 62.8 to 79.2 nm as the voltage was increased, whereas 5

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Fig. 5. SEM micrographs of the 30 °C–60 V sample etched for 1 min (a), 5 min (c) and 10 min (e), and the corresponding surfaces after the second anodization step (b, d, f).

SEM micrographs of these resulting samples. Fig. 4 shows that the pore arrangement of the 40 °C–45 V and 40 °C–60 V samples was not improved by reducing the anodization time to 15, 30 and 45 min. The 40 °C–45 V samples retained their porous structure; no alumina nanowires or large Dp values were observed (as seen in Fig. 4d, e and f). The surface morphology of the 40 °C–60 V sample was dynamic and significant changes occurred as the anodization time was increased from 15 to 45 min. Here, it was evident that a high rate of AAO dissolution prevailed over the AAO growth rate as the anodization time was increased. The pore wall thickness of the 40 °C–60 V–30 min (Fig. 4b) sample was non-uniform and the onset of nanowires formation is evident. The inception of nanowire formation is promoted by non-uniformly sized, randomly arranged pores [66,67]. Xiao et al. found that nanowires can be observed at areas where the

took place when the voltage was increased, and iii) exponential changes in morphological parameters took place in the samples anodized at 40 °C when the voltage was increased, resulting in the formation and partial dissolution of nanowires—suggesting that the anodization time could be decreased to < 1 h. 3.3. Effect of anodization time on one-step anodized samples As mentioned in Section 3.2, the presence of alumina nanowires on the surfaces of 40 °C–45 V and 40 °C–60 V samples suggested over-anodization. Furthermore, the underlying pores appeared to be semi-ordered. Therefore, further samples were prepared under similar anodization temperatures and voltages as used thus far, but now shorter anodization times were used: 15, 30 and 45 min. Fig. 4 presents the 6

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Table 2 Morphological parameters (Dp, Dc, P and n) of AAO obtained during the two-step anodization of Al6082. Etching time (min)

Dp (nm)

Dc (nm)

P (%)

n (pores/μm2)

5 10

43.8 ± 6.0 (24.6) 26.0 ± 10.0 (24.6)

82.6 ± 19.0 (79.2) 86.7 ± 22.1 (79.2)

25 (9) 8 (9)

169 (184) 153 (184)

technological applications (e.g., catalyst support and filtration). Nevertheless, the observed improvement in the pore arrangement is the main reason why the two-step anodization is typically performed. A smooth surface was obtained by increasing the etching time to 10 min—the concave structures (hereafter referred to as ‘concaves’) observed in Fig. 5c were no longer evident. Re-anodization of the 10 min etched samples yielded an AAO with disordered pore arrangement similar to samples presented in Fig. 3 (excluding the 40 °C–45 V and 40 °C–60 V samples). Thus, the benefit of using two-step anodization and the AAO etching step is lost if the sample is over-etched. The morphological parameters of the two-step anodized samples presented in Fig. 5d and f are summarized in Table 2. The values presented in parentheses are the values of the one-step anodized 30 °C–60 V sample (as presented in Table 1) and included here in Table 2 as reference values. Table 2 shows that the Dp of the 5 min etched two-step anodized sample was calculated to be 43.8 nm, which is a significant increase from 24.6 nm for the one-step anodized counterpart. The 5 min etched two-step anodized sample had a Dc of 82.6 nm, whereas its one-step counterpart had a Dc value of 79.2 nm, suggesting that the distance between pores for the one- and two-step anodized samples remained more or less consistent. A significant increase in P was also observed for the 5 min etched sample, which coincided with the increase in Dp. The fact that Dc remained relatively unchanged, while P increased, suggested that the AAO thickness between pores decreased, which was evident from the increase in Dp. A slight decrease in n was also observed. This may be ascribed to the ordered nature of the AAO observed in Fig. 5d. The decrease in n was caused by the absence of numerous random irregular-sized pores present in the disordered AAO of the onestep anodized counterpart. The AAO layer thickness of the two-step anodized sample coincided with its one-step counterpart. The 10 min etched sample had morphological characteristics similar to the one-step anodized 30 °C–60 V AAO sample and was not considered further. The AAO presented in Fig. 5d suggests that Al6082 may be used as the Al source material to prepare AAO with relatively ordered pore distribution by utilizing the anodization procedure applied here.

wall thickness is much larger than the average value [66]. It has also been found that higher current densities associated with higher temperature have an appreciable effect on nanowire growth [63]. As a consequence, a high density of AAO nanowires was formed in the case of the 40 °C–60 V–45 min sample (Fig. 4c). By further increasing the anodizing time to 1 h, the nanowires were partially dissolved, as is evident from Fig. 3c (Section 3.2). When comparing 40 °C–45 V–45 min (Fig. 4f) and 40 °C–45 V–60 min (Fig. 3b), it is evident that a significant difference in surface morphology exists. It is further evident from Fig. 1 that the 40 °C–60 V and 40 °C–45 V samples had the largest current densities and followed hard anodization regime [17,68]. Hard anodization is associated with rapid AAO layer formation and development and is typically applied to reduce the preparation time of AAO layers. This layer formation and development was evident when considering the following 40 °C–60 V AAO layer surfaces during 1 h of anodization: i) AAO layer formation (Fig. 4a – after 15 min of anodization), ii) nanowire formation initiation (Fig. 4b – after 30 min of anodization), iii) progressed nanowire formation (Fig. 4c – after 45 min of anodization), and iv) nanowire dissolution (Fig. 3c – after 1 h of anodization). Relatively similar observations may be made for 40 °C–45 V; i) AAO layer formation (Fig. 4d to f – after 45 min of anodization), and ii) nanowire formation (Fig. 3b – after 1 h of anodization). The rate of AAO layer development was not as significant for 40 °C–45 V when compared to 40 °C–60 V due to the lower anodization potential. Nevertheless, studies by Ganley et al. and Voon et al. stated that the current efficiency (defined as the ratio of current used for oxide formation relative to the total amount of current passed through the sample) decreased appreciably with an increase in temperatures due to accelerated AAO dissolution [48,52,53]. 3.4. Effect of two-step anodization Results presented in Section 3.2 suggested that the 30 °C–60 V sample had the largest pore diameter of the AAO layers (without the presence of nanowires) prepared thus far. In an attempt to obtain an ordered pore arrangement, a sample was anodized according to the anodization procedure (ii) (Section 2). The second step was performed at a temperature and voltage of 30 °C and 60 V, respectively. During the etching step, it was noticed that the anodized sample exhibited continuous surface bubble formation. It indicated that the Al6082 alloy was soluble in the etching solution used. This therefore suggested that the etching tim-e would probably have a critical effect on the AAO etching step. Thus, etching times of 1, 5 and 10 min were considered. The etched surfaces were investigated by SEM before and after being reanodized (Fig. 5). It is evident from Fig. 5 that the etching time had an appreciable effect on the morphology of the oxide layer formed during the second anodization procedure. Results presented in Fig. 5a suggested that 1 min of etching did not remove the initial oxide layer, and re-anodization resulted in an AAO with a distorted surface (Fig. 5b). After increasing the etching time to 5 min, the AAO layer was completely removed and a surface with concave morphology was obtained (Fig. 5c). The pitting observed in Fig. 5c formed during the anodization procedure. Re-anodization of the 5 min etched sample yielded an AAO with semi-ordered pore arrangement (Fig. 5d). Although the pore arrangement shown in Fig. 5d is far from an ideal hexagonal packing, defined by a central pore surrounded by six neighbouring pores, the observed pore arrangement is not a limiting factor to promote certain

3.5. Surface chemical composition To determine whether impurities present in Al6082 migrated along the AAO layer during anodization procedures, the surface chemical composition of as-received Al6082 and the 30 °C–60 V sample was determined by EDS mapping. The mappings were converted to black and white images, with black representing the background and white the detected elemental points. Thereafter, the total area occupied by white pixels was determined (using ImageJ) and represented the amount of an element relevant to the background. Al and the three major impurities were considered: Si, Mg and Mn. Fig. 6 presents these EDS mappings. Results showed that the Al, Si, Mg and Mn contents of the 30 °C–60 V sample were 4.36, 1.19, 0.54 and 0.58% lower that of asreceived Al6082, respectively. This suggested that the surface of the AAO layer had a similar chemical composition to that of the as-received Al, indicating that Sn, Mg and Mn cations became incorporated into the AAO matrix. The decrease in Al was expected due to the incorporation of O atoms. Habazaki et al. stated that impurities present in various Al alloys can be incorporated into the AAO layer at the alloy/AAO layer interface. Incorporation first requires the formation of a thin impurityenriched layer on the alloy surface, immediately below the AAO layer. 7

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Fig. 6. EDS mappings of the Al (a), Si (c), Mg (e) and Mn (g) contents of as-received Al6082 and the mappings of the Al (b), Si (d), Mg (f) and Mn (h) contents of the 30 °C–60 V sample.


Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

Fig. 7. A cross-section micrograph of the AAO layer/alloy interface of the 5 min etched, two-step anodized sample anodized at 30 °C–60 V. The inset shows the chemical composition (wt%) of the areas considered for EDS analysis (solid black line).

temperature and voltage. It was found that Dp, Dc, P and AAO layer thickness increased with increasing temperature and voltage, whereas n decreased with increasing temperature and voltage. iii. Nanowire formation was evident in samples anodized at 40 °C–45 V and 40 °C–60 V for a period of 1 h. Reducing the anodization time to 30 and 45 min did not prevent nanowires from forming in the 40 °C–60 V sample. In the case of the 40 °C–45 V sample, no nanowires were observed when the anodizing time was reduced; however, the obtained AAO did not have pores with notable pore diameters. iv. A sample prepared by the two-step anodization process had a significantly improved pore arrangement. This process included an etching step, which removed the AAO formed during the first anodization procedure, revealing concaves on the Al surface. The AAO obtained during the second anodization step formed from these concaves. In order to preserve the latter mentioned concaves (taking into account that Al6082 is soluble in the etching solution), an optimal etching time of 5 min was determined suitable. v. A two-step anodized sample with semi-ordered pore arrangement had the following morphological characteristics: D p, 43.8 ± 6.0 nm; Dc, 82.6 ± 19.0 nm; P, 25% and n, 169 pores/μm2. The AAO layer thickness was approximately 53 μm.

After sufficient enrichment of an impurity, the impurity may be oxidized. The required enrichment varies from impurity to impurity and the degree of enrichment may be correlated with the Gibbs free energy per equivalent for formation of the relevant element's oxide [51]. The bond energies of MgO and MnO are less than that of Al2O3 [69]; whereas the bond energy of SiO2 is higher than that of Al2O3 [70]. Thus, the rate of Mg and Mn transfer from Al6082 to the AAO layer should be kinetically fast when compared to Si; this was evident considering that Mg and Mn contents present on the surface of the AAO layer was approximately double that of Si. To evaluate the mobility of impurities present in Al6082 during anodization, the 5 min etched, two step-anodized 30 °C–60 V was cross sectioned and the AAO layer/alloy interface was analysed by EDS (Fig. 7). It is clear from Fig. 7 that only Al and O were detected during EDS analysis and that no significant concentrations of impurities were evident; impurities were likely present below the detection limit of the instrument (< 0.5 wt%). Thus, significant accumulation of impurities in the area immediately above and below the AAO layer/alloy interface was insignificant. The white particles observed in the alloy section of Fig. 7 were identified as residual polishing paste particles. When considering that impurities migrated along the AAO layer and were present on the AAO layer surface (Fig. 6) and the absence of major morphological defects in Fig. 5d, it appeared that the impurities present in Al6082 did not appreciably affect the AAO layer prepared from Al6082 as the Al source material.

Acknowledgements The Department of Science and Innovation (DSI), South Africa and HySA Infrastructure Center of Competence at the North-West University (NWU), South Africa, are acknowledged for financial support through KP5 program.

4. Conclusions In this study, an AAO layer with semi-ordered pore arrangement was prepared on the surface of a low cost, low purity Al alloy (Al6082) via a rapid two-step anodizing process using 0.4 M oxalic acid as the electrolyte. Morphological characteristics such as pore diameter (Dp), inter-pore distance (Dc), porosity (P) and pore density (n) were determined. The following conclusions can be drawn from this study.

References [1] W. Lee, The anodization of aluminum for nanotechnology applications, JOM 62 (6) (2010) 57–63, [2] G. Wieβmeier, D. Hönicke, Microfabricated components for heterogeneously catalysed reactions, J. Micromech. Microeng. 6 (1996) 285–289, 1088/0960-1317/6/2/012. [3] G. Patermarakis, N. Nicolopoulos, Catalysis over porous anodic alumina film catalysts with different pore surface concentrations, J. Catal. 187 (2) (1999) 311–320, [4] M. Lahav, T. Sehayek, A. Vaskevich, I. Rubinstein, Nanoparticle nanotubes, Angew. Chem. Int. Ed. 42 (2003) 5576–5579,

i. Disordered porous AAO layers with Dp ranging between 8.7 and 133.4 nm and Dc ranging between 53.0 and 146.7 nm could be prepared by a one-step anodizing process at temperature of 20–40 °C and voltages of 30–60 V. ii. Dp, Dc, P, n, and AAO layer thickness greatly varied as a function of 9

Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

(2013) 424–432, [31] L.E. Fratila-Apachitei, F.D. Tichelaar, G.E. Thompson, H. Terryn, P. Skeldon, J. Duszczyk, L. Katgerman, A transmission electron microscopy study of hard anodic oxide layers on AlSi(Cu) alloys, Electrochim. Acta 49 (2004) 3169–3177, https:// [32] L.E. Fratila-Apachitei, H. Terryn, P. Skeldon, G.E. Thompson, J. Duszczyk, L. Katgerman, Influence of substrate microstructure on the growth of anodic oxide layers, Electrochim. Acta 49 (2004) 1127–1140, electacta.2003.10.024. [33] K. Shimizu, K. Kobayashi, G.E. Thompson, P. Skeldon, G.C. Wood, The influence of ‘θ’ precipitates on the anodizing behaviour of binary Al-Cu alloys, Corros. Sci. 39 (2) (1997) 281–284, [34] G.D. Sulka, W.J. Stepniowski, Structural features of self-organized nanopore arrays formed by anodization of aluminum in oxalic acid at relatively high temperatures, Electrochim. Acta 54 (2009) 3683–3691, 01.046. [35] W.J. Stepniowski, D. Zasada, Z. Bojar, First step of anodization influences the final nanopore arrangement in anodized alumina, Surf. Coat. Technol. 206 (2011) 1416–1422, [36] R. Zhang, K. Jiang, Y. Zhu, H. Qi, G. Ding, Ultrasond-assisted anodization of aluminium in oxalic acid, Appl. Surf. Sci. 258 (2011) 586–589, 1016/j.apsusc.2011.08.041. [37] H. Takahashi, M. Nagayama, The determination of the porosity of anodic oxide films on aluminium by the pore-filling method, Coros. Sci. 18 (1978) 911–925, [38] C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH image to ImageJ: 25 years of image analysis, Nat. Methods 9 (2012) 671–675, 2089. [39] L. Vojkuvka, L.F. Marsal, J. Ferré-Borrull, P. Formentin, J. Pallarés, Self-ordered porous alumina membranes with large lattice constant fabricated by hard anodization, Superlattice. Microst. 44 (2008) 577–582, 2007.10.005. [40] G.D. Sulka, K.G. Parkola, Anodising potential influence on well-ordered nanostructures formed by anodisation of aluminium in sulphuric acid, Thin Solid Films 515 (2006) 338–345, [41] G.D. Sulka, Highly ordered anodic porous alumina formation by self-organised anodizing, in: A. Eftekhari (Ed.), Nanostructured Materials in Electrochemistry, Ch. 1 Wiley-VCH, Weinheim, 2008. [42] J.S. Suh, J.S. Lee, Highly ordered two-dimensional carbon nanotube arrays, Appl. Phys. Lett. 75 (1999) 2047–2049, [43] F.A. Bruera, G.R. Kramer, M.L. Vera, A.E. Ares, Synthesis and morphological characterization of nanoporous aluminium oxide films by using a single anodization step, Coatings 9 (2019) 115, [44] O. Jessensky, F. Müller, U. Gösele, Self-organized formation of hexagonal pore arrays in anodic alumina, Appl. Phys. Lett. 72 (10) (1998) 1173–1175, https://doi. org/10.1063/1.121004. [45] V.P. Parkhutik, V.I. Shershulsky, Theoretical modelling of porous oxide growth on aluminium, J. Phys. D. Appl. Phys. (D25) (1992) 1258–1263, 1088/0022-3727/25/8/017. [46] L.E. Fratila-Apachitei, J. Duszczyk, L. Katgerman, Voltage transients and morphology of AlSi(Cu) anodic oxide layers formed in H2SO4 at low temperature, Surf. Coat. Technol. 157 (2002) 80–94, 00144-5. [47] C.K. Chung, M.W. Liao, H.C. Chang, C.T. Lee, Effects of temperature and voltage mode on nanoporous anodic aluminum oxide films by one-step anodization, Thin Solid Films 520 (2011) 1554–1558, [48] J.C. Ganley, K.L. Riechmann, E.G. Seebauer, R.I. Masel, Porous anodic alumina optimized as a catalyst support for microreactors, J. Catal. 227 (2004) 26–32, [49] C.K. Chung, C.H. Tsai, C.R. Hsu, E.H. Kuo, Y. Chen, I.C. Chung, Impurity and temperature enhanced growth behaviour of anodic aluminium oxide from AA5052 Al-Mg alloy using hybrid pulse anodization at room temperature, Corros. Sci. 125 (2017) 40–47, [50] Y. Liu, P. Skeldon, G.E. Thompson, H. Habazaki, K. Shimizu, Anodic film growth on an Al–21at.%Mg alloy, Corros. Sci. 44 (2002) 1133–1142, 1016/S0010-938X(01)00115-9. [51] H. Habazaki, K. Shimizu, P. Skeldon, G.E. Thompson, G.C. Wood, X. Zhou, Effects of alloying elements in anodizing of aluminium, Trans. Int. Met. Finish. 75 (1997) 18–23, [52] C.H. Voon, M.N. Derman, U. Hashim, Effect of manganese content on the fabrication of porous anodic alumina, J. Nanomater. 2012 (2012) 1–9, 1155/2012/752926. [53] C.H. Voon, M.N. Derman, U. Hashim, K.R. Ahmad, K.L. Foo, Effect of temperature of oxalic acid on the fabrication of porous anodic alumina from Al-Mn alloys, J. Nanomater. (2013) (2013) 40–48, [54] C.H. Voon, M.N. Derman, U. Hashim, K.R. Ahmad, L.N. Ho, A simple one-step anodising method for the synthesis of ordered porous anodic alumina, J. Exp. Nanosci. 9 (2014) 106–112, [55] C.K. Chung, B.Y. Chu, C.H. Tsai, C.R. Hsu, Photoluminescence enhancement of nanoporous alumina using one-step anodization of high- and low-purity aluminum at room temperature, Mater. Lett. 190 (2017) 157–160, matlet.2017.01.015. [56] M. Almasi Kashi, A. Ramazani, H. Abbasian, A. Khayyatian, Capacitive humidity sensors based on large diameter porous alumina prepared by high current anodization, Sens. Actuators, A 174 (2012) 69–74, 11.033.

[5] A. Johansson, J. Lu, J.-O. Carlsson, M. Boman, Deposition of palladium nanoparticles on the pore walls of anodic alumina using sequential electroless deposition, J. Appl. Phys. 96 (2004) 5189–5194, [6] M.S. Sander, M.J. Cote, W. Gu, B.M. Kile, C.P. Tripp, Template-assisted fabrication of dense, aligned arrays of titania nanotubes with well-controlled dimensions on substrates, Adv. Mater. 16 (2004) 2052–2057, 200400446. [7] H. Cao, L. Wang, Y. Qiu, Q. Wu, G. Wang, L. Zhang, X. Liu, Generation and growth mechanism of metal (Fe, Co, Ni) nanotube arrays, ChemPhysChem 7 (2006) 1500–1504, [8] W. Lee, R. Scholz, K. Nielsch, U. Gösele, A template-based electrochemical method for the synthesis of multisegmented metallic nanotubes, Angew. Chem. Int. Ed. 44 (2005) 6050–6054, [9] H. Chik, J.M. Xu, Nanometric superlattices: non-lithographic fabrication, materials, and prospects, Mater. Sci. Eng. R 43 (2004) 103–138, mser.2003.12.001. [10] T. Shimizu, M. Nagayanagi, T. Ishida, O. Sakata, T. Oku, H. Sakaue, T. Takahagi, S. Shingubara, Epitaxial growth of Cu nanodot arrays using an AAO template on a Si substrate, Electrochem. Solid State Lett. 9 (2006) J13–J16, 1149/1.2176891. [11] H. Masuda, K. Yasui, K. Nishio, Fabrication of ordered arrays of multiple nanodots using anodic porous alumina as an evaporation mask, Adv. Mater. 12 (2000) 1031–1033,<1031::AIDADMA1031>3.0.CO;2-R. [12] L. Wang, T.P. Tran, D.V. Vo, M. Sakurai, H. Kameyama, Design of novel Pt-structured catalyst on anodic aluminum support for VOC’s catalytic combustion, Appl. Catal., A 350 (2008) 150–156, [13] O. Sanz, L.C. Almeida, J.M. Zamaro, M.A. Ulla, E.E. Miro, M. Montes, Washcoating of Pt-ZSM5 onto aluminium foams, Appl. Catal., B 78 (2008) 166–175, https://doi. org/10.1016/j.apcatb.2007.09.024. [14] Z. Ni, E.G. Seebauer, R.I. Masel, Effects of microreactor geometry on performance: differences between posted reactors and channel reactors, Ind. Eng. Chem. Res. 44 (2005) 4267–4271, [15] K. Schwirn, W. Lee, R. Hillebrand, M. Steinhart, K. Nielsch, U. Gösele, Self-ordered anodic aluminum oxide formed by H2SO4 hard anodization, ACS Nano 2 (2008) 302–310, [16] H. Masuda, F. Hasegwa, S. Ono, Self-ordering of cell arrangement of anodic porous alumina formed in sulphuric acid solution, J. Electrochem. Soc. 144 (1997) L127–L130, [17] W. Lee, R. Ji, U. Gösele, K. Nielsch, Fast fabrication of long-range ordered porous alumina membranes by hard anodization, Nat. Mater. 5 (2006) 741–747, https:// [18] W.J. Stepniowski, Z. Bojar, Synthesis of AAO at relatively high temperatures. Study of the influence of anodizing conditions on the alumina structural features, Surf. Coat. Technol. 206 (2011) 265–272, 020. [19] C. Sun, J. Luo, L. Wu, J. Zhang, Self-ordered anodic alumina with continuously tunable pore intervals from 410 to 530 nm, ACS Appl. Mater. Interfaces 2 (5) (2010) 1299–1302, [20] H. Masuda, K. Yada, A. Osaka, Self-ordering of cell configuration of anodic porous alumina with large-size pores in phosphoric acid solution, Jpn. J. Appl. Phys. 37 (1998) L1340–L1342, [21] H. Masuda, H. Asoh, M. Watanabe, K. Nishio, M. Nakao, T. Tamamura, Square and triangular nanohole array architectures in anodic alumina, Adv. Mater. 13 (2001) 189–192,<189::AIDADMA189>3.0.CO;2-Z. [22] W. Lee, J.-C. Kim, U. Gösele, Spontaneous current oscillations during hard anodization of aluminum under potentiostatic conditions, Adv. Funct. Mater. 20 (2010) 21–27, [23] H. Masuda, K. Fukuda, Ordered metal nanohole arrays made by a two-step replication of honeycomb structures of anodic alumina, Science 268 (5216) (1995) 1466–1468, [24] L. Zaraska, G.D. Sulka, J. Szeremeta, M. Jaskuła, Porous anodic alumina formed by anodization of aluminum alloy (AA1050) and high purity aluminum, Electrochim. Acta 55 (2010) 4377–4386, [25] J.M. Montero-Moreno, M. Sarret, C. Muller, Influence of the aluminum surface on the final results of a two-step anodizing, Surf. Coat. Technol. 201 (14) (2007) 6352–6357, [26] T. Aerts, T. Dimogerontakis, I. De Graeve, J. Fransaer, H. Terryn, Influence of the anodizing temperature on the porosity and the mechanical properties of the porous anodic oxide film, Surf. Coat. Technol. 201 (2007) 7310–7317, 1016/j.surfcoat.2007.01.044. [27] C.-U. Yu, C.-C. Hu, A. Bai, Y.-F. Yang, Pore-size dependence of AAO films on surface roughness of Al-1050 sheets controlled by electropolishing coupled with fractional factorial design, Surf. Coat. Technol. 201 (2007) 7259–7265, 1016/j.surfcoat.2007.01.033. [28] L. Fernandez-Romero, J.M. Montero-Moreno, E. Pellicer, F. Peiro, A. Cornet, J.R. Morante, M. Sarret, C. Muller, Assessment of the thermal stability of anodic alumina membranes at high temperatures, Mater. Chem. Phys. 111 (2008) 542–547, [29] T.-S. Shih, P.-S. Wei, Y.-S. Huang, Optical properties of anodic aluminum oxide films on Al1050 alloys, Surf. Coat. Technol. 202 (2008) 3298–3305, https://doi. org/10.1016/j.surfcoat.2007.12.002. [30] M. Michalska-Domanska, M. Norek, W.J. Stepniowski, B. Budner, Fabrication of high quality anodic aluminum oxide (AAO) on low purity aluminum–a comparative study with the AAO produced on high purity aluminum, Electrochim. Acta 105


Surface & Coatings Technology 378 (2019) 124970

A.E. Kozhukhova, et al.

[64] J.P. O’Sullivan, G.C. Wood, The morphology and mechanism of formation of porous anodic films on aluminium, Proc. R. Soc., A 317 (1970) 511–543, 10.1098/rspa.1970.0129. [65] G.D. Sulka, K.G. Parkola, Temperature influence on well-ordered nanopore structures grown by anodization of aluminium in sulphuric acid, Electrochim. Acta 52 (2007) 1880–1888, [66] Z.L. Xiao, C.Y. Han, U. Welp, H.H. Wang, W.K. Kwok, G.A. Willing, J.M. Hiller, R.E. Cook, D.J. Miller, G.W. Crabtree, Fabrication of alumina nanotubes and nanowires by etching porous alumina membranes, Nano Lett. 2 (11) (2002) 1293–1297, [67] Y.F. Mei, G.G. Siu, R.K.Y. Fu, P. Chen, X.L. Wu, T.F. Hung, P.K. Chu, Y. Yang, Formation mechanism of alumina nanotubes and nanowires from highly ordered porous anodic alumina template, J. Appl. Phys. 97 (2005) 034305, , https://doi. org/10.1063/1.1846137. [68] W. Lee, K. Nielsch, U. Gosele, Self-ordering behaviour of nanoporous anodic aluminium oxide (AAO) in malonic acid anodization, Nanotechnology 18 (2007) 475713, , [69] A.C. Crossland, G.E. Thompson, C.J.E. Smith, H. Habazaki, K. Shimizu, P. Skeldon, Formation of manganese-rich layers during anodizing of Al-Mn alloys, Corros. Sci. 41 (1999) 2053–2069, [70] K. Shimizu, K. Kobayashi, Anodic oxide growth on valve metals-future prospects, J. Surf. Finish. Soc. 46 (1995) 402–409,

[57] Alaa M. Abd-Elnaiem, A.M. Mebed, Waleed Ahmed El-Said, M.A. Abdel-Rahim, Porous and mesh alumina formed by anodization of high purity aluminum films at low anodizing voltage, Thin Solid Films 570 (2014) 49–56, 1016/j.tsf.2014.08.046. [58] R. Zhang, K. Jiang, Y. Zhu, H. Qi, G. Ding, Ultrasound-assisted anodization of aluminum in oxalic acid, Appl. Surf. Sci. 258 (2011) 586–589, 1016/j.apsusc.2011.08.041. [59] X. Li, Y. He, T. Zhang, L. Que, Aluminum oxide nanostructure-based substrates for fluorescence enhancement, Opt. Express 20 (2012) 21272–21277, 10.1364/OE.20.021272. [60] B. Zhang, X. Zhao, Y. Li, B. Hou, Fabrication of durable anticorrosion superhydrophobic surfaces on aluminium substrates via a facile one-step electrodeposition approach, RSC Adv. 6 (2016) 35455–35465, C6RA05484F. [61] A.O. Araoyinbo, A.F.M. Noor, S. Sreekantan, A. Aziz, Voltage effect on electrochemical anodization of aluminum at ambient temperature, Int. J. Mech. Mat. Eng. 5 (2010) 53–58. [62] P. Erdogan, B. Yuksel, Y. Birol, Effect of chemical etching on the morphology of anodic aluminum oxides in the two-step anodization process, Appl. Surf. Sci. 258 (2012) 4544–4550, [63] N.-Q. Zhao, X.-X. Jiang, C.-S. Shi, J.-J. Li, Z.-G. Zhao, X.-W. Du, Effects of anodizing conditions on anodic alumina structure, J. Mater. Sci. 42 (2007) 3878–3882,