Journal Information
Vol. 8. Issue 3.
Pages 2481-3388 (May - June 2019)
Share
Share
Download PDF
More article options
Visits
166
Vol. 8. Issue 3.
Pages 2481-3388 (May - June 2019)
Original Article
DOI: 10.1016/j.jmrt.2018.05.030
Open Access
Corrosion protection of aluminum by smart coatings containing layered double hydroxide (LDH) nanocontainers
Visits
166
Iman Imanieh
Corresponding author
iman.imanieh@gmail.com

Corresponding author.
, Abdollah Afshar
Department of Materials Science and Engineering, Sharif University of Technology, P.O. Box 11365-9466, Tehran, Islamic Republic of Iran
This item has received
166
Visits

Under a Creative Commons license
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (24)
Show moreShow less
Tables (4)
Table 1. Design of experiments to optimize the properties of layered double hydroxide.
Table 2. Parameters extracted from the five fitting of the kinetic models. The experiments are selected from Table 1.
Table 3. Weibull parameters which are extracted from the fitted curves for some experiments. The experiments are selected from Table 1.
Table 4. EIS results extracted from the equivalent circuit for the considered coatings after various days of immersion in 3.5% NaCl solution.
Show moreShow less
Abstract

In this study, the LDH containing a kind of green inhibitor (sodium molybdate) was synthesized and utilized in a poly-vinyl alcohol (PVA) base coating applied on the aluminum (5054 series) surface. After evaluating the LDH performance, the inhibitor release from the LDH crystals was examined by changing significant parameters like MII/MIII (the molar ratio of divalent ions into trivalent ions), pH, aging time and temperature. In order to find the kinetic release of intercalated inhibitor a prediction model was introduced in 3.5% NaCl solution. The model expressed that the pH as the processing parameter has special effect on the kinetic releasing of the inhibitor. It can enhance the amount of released inhibitor up to 6 times larger due to its effect on the LDH morphology and structure. Also MII/MIII has serious effect on the amount of intercalated inhibitor and the release rate. In fact, the ratio alteration can affect the amount of final released inhibitor by a factor of 2 and its interaction with the pH is noticeable. The whole mechanism of releasing was recognized as Fickian diffusion and the released data were well fitted with the Weibull model. Corrosion measurements show the mechanism activation of the synthesized smart coating and improvement of the corrosion properties (the polarization resistance has been enhanced more than 3 times) in compare with the regular coatings. More LDH in the coating network result in better resistance (39,250Ωcm−1) to corrosion for the produced coating up to approximately 5.7wt% of the Cl ions.

Keywords:
Aluminum
Smart coating
Layered double hydroxide
Sodium molybdate
Kinetic release
Full Text
1Introduction

Recently, smart coatings have been used in various applications due to their functional abilities [1,2]. Various mechanisms are introduced to activate the mitigation actions against corrosion in times of necessity. In recent years, smart coatings have been supported by nanotechnology and nanomaterials [3–5].

Layered double hydroxides (LDH) are biocompatible inorganic materials which can release the intercalated anions in appropriate time. In other words various smart systems are introduced based on using this structure [6–8]. Using LDH in smart coating is very popular and practical. Many inhibitors are intercalated into the LDH and their behavior has been evaluated [9–12]. One of the most important points in such a structure is the ion exchange ability and the kinetic of the reaction. The ion exchange with high speed lead to loss of intercalated inhibitors in a short time and the ion exchange with low speed lead to poor corrosion prevention. In other words an optimum speed is required to obtain the best protection with standard long life of the coating. Several studies have been reported describing the intercalation and controlled release of the desired material [13–16]. In this regard, synthesis parameters and other variables were found as effective.

In this paper, it is tried to synthesize LDH in order to release an environmentally friendly inhibitor (sodium molybdate) with an appropriate rate just in times of necessity. This reservoir was utilized in a polymeric base coating (PVA) in order to provide a functional or smart coating for protecting aluminum alloys. It is obvious that many parameters can control the function ability of the coating and the kinetic of the release reactions during the coating life. These parameters were modeled by mathematical reactions and the final corrosion properties of the smart coating were evaluated accurately.

2Materials and methods2.1LDH synthesis

There are various synthesizing methods for the LDH which the co-precipitation method is the most popular. By using this methodology, 0.05M and 0.025M magnesium and aluminum nitrate (1:2) resolved in 100ml of distilled water. The water was previously boiled in order to remove carbonate from the solutions. This solution is then slowly added to 200ml of a solution containing sodium molybdate as inhibitor (the pH was controlled by sodium hydroxide and kept constant in 9.5). This process was performed at room temperature within 90min and the solution stirred vigorously. Ar pump into the solution was continued in all the time. In the next step the prepared solution was aged in 65°C for 24h and then centrifuged for 10min. For complete removal of water, the freeze drier was employed.

2.2Significant parameters and kinetic studied

The molar ratio of the divalent and trivalent ions (MII/MIII), pH of the solutions, aging time and temperature were the selected important parameters. These parameters were evaluated by design of experiments (DOE) method based on central composite design (CCD). The MII/MIII, pH, aging time and temperature were changed in the range of 1–5, 8–12, 2–48h, 300–423K (27–150°C) respectively. The CCD designed the experiments in order to identify the effect of these parameters on the kinetic of the inhibitor releasing from the LDH in a certain time. The designed experiments are sorted in Table 1 (Design Expert 7 software was used). In order to gain more insight into the kinetics of the release, five commonly used models were fitted to the release data. Models used are the (i) Avrami-Erofe’ev, (ii) the Elovich model, (iii) first-order rate model, (iv) the modified Freundlich Eq. and a (v) parabolic diffusion model.

Table 1.

Design of experiments to optimize the properties of layered double hydroxide.

  MII/MIII  pH  Aging temp. K (˚C)  Aging time (h) 
2.00  8.00  318 (45.00)  15.00 
4.00  8.00  318 (45.00)  15.00 
2.00  11.00  318 (45.00)  15.00 
4.00  11.00  318 (45.00)  15.00 
2.00  8.00  353 (80.00)  15.00 
4.00  8.00  353 (80.00)  15.00 
2.00  11.00  353 (80.00)  15.00 
4.00  11.00  353 (80.00)  15.00 
2.00  8.00  318 (45.00)  40.00 
10  4.00  8.00  318 (45.00)  40.00 
11  2.00  11.00  318 (45.00)  40.00 
12  4.00  11.00  318 (45.00)  40.00 
13  2.00  8.00  353 (80.00)  40.00 
14  4.00  8.00  353 (80.00)  40.00 
15  2.00  11.00  353 (80.00)  40.00 
16  4.00  11.00  353 (80.00)  40.00 
17  1.00  9.50  335 (62.00)  27.50 
18  5.00  9.50  335 (62.00)  27.50 
19  3.00  6.50  335 (62.00)  27.50 
20  3.00  12.50  335 (62.00)  27.50 
21  3.00  9.50  300 (27.00)  27.50 
22  3.00  9.50  370 (97.00)  27.50 
23  3.00  9.50  335 (62.00)  2.50 
24  3.00  9.50  335 (62.00)  52.50 
25  3.00  9.50  335 (62.00)  27.50 
26  3.00  9.50  335 (62.00)  27.50 
27  3.00  9.50  335 (62.00)  27.50 
28  3.00  9.50  335 (62.00)  27.50 

In each model, C0 is the amount of guest molybdate in the Inh–LDH at t=0, Ct is the amount of guest molybdate in the Inh–LDH at time t, and kd is the rate of release. a, b, and n are constants and α is the extent of reaction.

2.3Characterization

The crystallinity of the synthesized LDH was determined by X-ray diffraction (XRD) using a STOE transmission diffractometer (STOE & Cie GmbH, Darmstadt, Germany) using Cu Kα radiation (40kV, 30mA). The powder was scanned from 5° to 90° 2θ at a scanning rate of 3°min−1. Infrared spectra in the range of framework and OH vibrations (400–4000cm−1) were recorded on KBr pellets using a Nicolet Magna 550 spectrometer (Thermo Scientific, Dreieich, Germany) in order to identify the LDH bonding. The morphology and particle size were characterized by scanning electron microscope (SEM) (SM-300, TOPCON).

2.4Applying the smart coating

The polyvinyl alcohol (PVA) was selected as composite matrix in the smart coating. Polyvinyl Alcohol (PVA, sometimes referred to as PVOH) is a water soluble polymer used widely in adhesives, paints, sealants, coatings, textiles, plastics and etc. Therefore, the synthesized LDH with different concentration were dispersed in the prepared 10% PVA solution and coated on the aluminum substrate. The dispersed LDH in PVA solution was coated on degreased and deoxidized 5054 Al substrates by a dip coater. The coating was allowed to dry and was then cured for 2h at 150°C to remove residual water and form a three-dimensional polymer network (2μm). During curing the PVA polymerizes becomes insoluble in water.

2.5Corrosion studies

Corrosion testing of coated surfaces was carried out using electrochemical impedance spectroscopy (EIS)-based methods. Tests were conducted in a three electrode cell (platinum as counter and SCE as reference electrode) filled with 3.5% NaCl solution that was open to air, but quiescent (room temperature). Impedance spectra were collected in various times interval (1–3–6–10–14 and 20 days) using a 10mV sine wave voltage perturbation, which was varied in frequency from 105 to 10−2Hz. The experiments were carried out using a Princeton Applied Research 273A potentiostat, and a Solartron Model 1255 frequency response analyzer. The experiment and data logging were coordinated by Scribner Associate's Zplot™ impedance software. For evaluating the final results, appropriate equivalent circuits were found and used to interpret the extracted data. By interpreting the changes of each element in equivalent circuit, the coating mechanism and other phenomena were recognized.

Corrosion behavior of the smart synthesized coating was also studied in 3.5% NaCl solution by using polarization curve. The Anodic and cathodic polarization curve were potentiodynamically measured by a potentiostat/galvanostat (AUTOLAB model 302N) in an appropriate scan rate (0.1mVs−1) and potential range (−0.85 to −0.60V).

3Results and discussion3.1Characterization

After synthesizing the LDH, different test must be carried out on the powder to confirm the correctness of the process. Therefore, XRD analysis of the powder was taken and the results are presented in Fig. 1. Showing sharp or broad peaks at the low angles and weak peaks at the high angles is one of the layered structures characteristics [17] which can be seen in X-ray result of the synthesized powder. As it can be seen the XRD spectrum of the synthesized powder shows the presence of Mg-Al LDH (Mg6Al2CO3(OH)16·4H2O) and a kind of molybdenum oxide. This evidence confirms the correctness of the LDH synthesizing although more analysis is needed to approve this claim. Some extra peaks are observable in the synthesized LDH which are related to other kind of molybdenum oxide.

Fig. 1.

XRD spectrum of the synthesized LDH, MII/MIII 2, pH 8, aging temperature 45°C and aging time 15h in compare with standard patterns of MoO4 and Mg-Al LDH.

(0.13MB).

In order to confirm the X-ray results and also identify the special bonding in the synthesized compound, FTIR analysis was performed and shown in Fig. 2. Strong peaks at 428, 1385, and 1637cm−1 confirm the presence of NO3− groups and water molecules which are present between the layers. A broad peak at 3430 is associated with the stretching of the OH groups of the water molecules or hydroxides. It has been reported that the peak at 1637 can be related to the presence of water molecules which are located between the layers. In addition, the broad peak at 677–613 is linked to Al-O bonding of the hydroxide layer [9]. Therefore, detected bondings confirm the correct synthesis of the LDH. Similar results are observed in the literature [18,19].

Fig. 2.

FTIR result of the synthesized LDH containing molybdate anions as green inhibitor.

(0.18MB).
3.2Kinetic studies

Now, kinetic studies on well synthesized LDH can be performed. To have a better sense from the kinetics of the release, five frequently used models were fitted on the release data (Eqs. (i)–(v)) shown in Fig. 3.

Fig. 3.

Fitting of different models into the release of molybdate from LDH in 0.05M NaCl solution (Experiment 1 from Table 1). Fits for (i) the Avrami-Erofe’ev model, (ii) the Elovich model, (iii) the first-order model, (iv) the Freundlich model, and (v) the parabolic diffusion model.

(0.36MB).

Fig. 3 indicates that all models describe the kinetic of the molybdate release from the LDH (Experiment 1 in Table 1) but the parabolic diffusion and also Avrami-Erofe‘ev models are the best fitted models. The extracted parameters from the five introduced models are shown in Table 2. In these models, according to the R2 values and visual inspection of “linear” plots, it appears that the parabolic diffusion and Avrami-Erofe‘ev models can be the most appropriate for describing molybdate release from the synthesized LDH. It was found that the Avrami-Erofe‘ev model with a little modification can better explain the phenomena (Fig. 4). In fact the ion exchange of the chloride ions with the intercalated molybdate ions is very similar to drug release in drug delivery applications. More details about the introduced model (Weibull model) for this application are provided at the literature. The Weibull model is defined as:

where kw and nw values describe the time scale of the process and indicator of the mechanism, respectively. This model overcomes the limitation of other conventional models, and it has been widely employed in the kinetic studies.

Table 2.

Parameters extracted from the five fitting of the kinetic models. The experiments are selected from Table 1.

Experiments  Fitted models  Parameters
    Kd  n  a  b 
Experiment 1Avrami-Erofe’ev  188.3  −0.2127  –  – 
Elovich  –  –  −0.0066  0.0315 
First-order  0.0005  –  –  – 
Freundlich  0.28  –  −0.1963  – 
Parabolic diffusion  12.175  –  0.0892  – 
Experiment 6Avrami-Erofe’ev  21.94  −0.2079  –  – 
Elovich  –  –  −0.0398  0.0664 
First-order  0.0009  –  –  – 
Freundlich  0.42  –  −0.1838  – 
Parabolic diffusion  0.1137  –  −0.0097  – 
Experiment 7Avrami-Erofe’ev  0.508  −0.5509  –  – 
Elovich  –  –  −0.0155  0.0736 
First-order  0.0028  –  –  – 
Freundlich  1.42  –  −0.4723  – 
Parabolic diffusion  0.1866  –  −0.0172  – 
Experiment 21Avrami-Erofe’ev  0.903  −0.332  –  – 
Elovich  –  –  −0.0122  0.059 
First-order  0.0018  –  –  – 
Freundlich  0.75  –  −0.2862  – 
Parabolic diffusion  0.1463  –  −0.0126  – 
Fig. 4.

Linear fitting of Weibull model to the release of molybdate from LDH in 0.05M NaCl solution (experiments 3, 9, 12, 15, 18 in Table 3), (a) Experiment 3: MII/MIII=2, pH=11, aging temperature=45°C, aging time=15h, (b) Experiment 9: MII/MIII=2, pH=8, aging temperature=45°C, aging time=40h, (c) Experiment 12: MII/MIII=4, pH=11, aging temperature=45°C, aging time=40h, (d) Experiment 15: MII/MIII=2, pH=11, aging temperature=80°C, aging time=40h, and (e) Experiment 18: MII/MIII=5, pH=9.5, aging temperature=62°C, aging time=27h.

(0.32MB).

Previously, Monte Carlo simulation techniques were utilized in both euclidian and fractal spaces for studying a drug release system. It was found that the Weibull model properly expresses the drug release profile in both cases when the Fickian diffusion is the drug release mechanism. In case of release from euclidian membrane [20] and the two-dimensional percolation fractal [21], the values of nw were in the range of 0.69–0.75 and 0.35–0.39, respectively.

As it is observable the nw for all experiments (Table 3) are in the range of Fickian diffusion mechanism and the only difference is in their amount. As it mentioned the lower value of nw indicates the diffusion process with slower rate in the disordered medium. The shape uniformity and other effective parameters will be discussed at the following.

Table 3.

Weibull parameters which are extracted from the fitted curves for some experiments. The experiments are selected from Table 1.

Experiments  Fitted model  Parameters
    Kw  nw 
Experiment 1  Weibull  3.047  0.2127 
Experiment 3  Weibull  2.203  0.2064 
Experiment 6  Weibull  1.900  0.2079 
Experiment 7  Weibull  0.444  0.5509 
Experiment 9  Weibull  2.498  0.1457 
Experiment 12  Weibull  2.751  0.1378 
Experiment 15  Weibull  0.989  0.3471 
Experiment 18  Weibull  2.586  0.1285 
Experiment 21  Weibull  1.034  0.332 

Fig. 5 illustrates the diffusion mechanisms based on concentration alteration versus time. While a drug or any other molecules can diffuse and move through the membrane without any pore size limitation (when that molecule is much smaller than the pore size) the concentration profile of the molecule basically follows Fick's law. It can be seen that the released data of molybdate from synthesized LDH follows this trend. As a matter of fact, the concentration profile of a released molecule generally depends on the ratio of the channel size/molecule size when the interaction of the molecule and the channel surface is not sensible [1]. Similarly in the current case (LDH with the intercalated inhibitor), the release mechanism is recognized as Fickian diffusion. Therefore it is well to consider the size of the container and the size of intercalated molecules. Microscopic considerations show that (Fig. 6) the layer thickness of LDH crystals (the size of exit path of the intercalated inhibitors) is about 40nm while the molybdenum molecules have smaller sizes. Therefore as mentioned the dominant mechanism of releasing can be Fickian diffusion. The extracted data (Table 3) confirm this hypothesis and show that changing the synthesizing parameters (in the studied range) cannot affect the whole release mechanism but it can change some details of the dominant mechanism.

Fig. 5.

The cumulative drug concentration with time. The Fickian (solid line), constraint (dashed line), and single file (dotted line) diffusion models (left side), the released data of molybdate from synthesized LDH [21].

(0.12MB).
Fig. 6.

SEM image of the synthesized LDH and molybdenum oxide molecules (schematic view) intercalated into the layered structure.

(0.29MB).

Considering the nw for different experiments show that the diffusion was occurred with different rates. According to the literature [22], n values less than 0.75 show the release with Fickian diffusion either in Euclidian (0.69<n<0.75) or fractal space, n<0.69. It can be seen all data show that the release are occurring in fractal spaces with Fickian discussion mechanism. Fig. 6 and other presented SEM images confirm that the inhibitor matrixes are fractal and the morphology and the kinetic data are in good agreement. There are some cases in which their nw are more than others. In example, the nw for experiments 7, 15 and 21 are 0.55, 0.35 and 0.33, respectively. It means that the ion exchange processes are occurring more quickly in these synthesized LDH than other ones. Considering the synthesizing condition of these experiments can validate the observed kinetic results. For having a better sense from synthesizing parameters effect on the kinetic behavior, a model was introduced and considered with more precision at the following.

Eq. (1) predicts the released inhibitor after 10min in 3.5% NaCl solution. It can be seen that there is a cubic relation between the synthesis parameters and the released inhibitor in a certain time.

which variables A, B, C, D are the ratio of divalent to trivalent ions (MII/MIII), pH solution, aging temperature K (°C) and time (h), respectively. By focusing on the factors coefficients, the importance of each factor can be determined. One way to check the accuracy of the prediction equation is using normal probability plots (Fig. 7). As can be seen, there is no serious deviation from the standard line, show that the presented model has good accuracy. To have a better sense from the equation accuracy, other criteria can also be used. For example, R-square, which is a number between zero and one reports the accuracy of the prediction equation. The standard expressed that the equation with the R-square more than 0.75 is acceptable and well accurate enough and can be used as a good reference [23]. The equation presented in this study has the R-square of 0.97. This shows the extreme accuracy of the founded model. It will be more sensible if take a look at Fig. 8 which shows the actual values versus predicted values calculated by the software. It can be observed that all data are approximately complied with the y=x. These are all evidence of acceptable accuracy and precision of the introduced equation.

Fig. 7.

Normal probability plot of residuals for the introduced model to predict the released inhibitor from the LDH.

(0.18MB).
Fig. 8.

Predicted versus actual data obtained from introduced model for prediction of released inhibitor from the LDH.

(0.15MB).

Now based on final application of the synthesized LDH, it is possible to select appropriate values for each parameter. Sometimes there is need to minimize the kinetic release of the intercalated inhibitor or sometime there is need to maximize the releasing speed or even sometimes there is a target for example 15ppm releasing in 10min. All these goals are achievable by using the introduced model (the calculation can be easily performed by the software). Fig. 9 illustrates the appropriate values for the two effective factors (pH and MII/MIII) in order to reach the mentioned goals (demonstrated as desirability). The model represents the parameters condition for releasing the molybdate in just 10min. But if there is need other times, the founded releasing mechanism in previous discussions can be very helpful. In other words, by the introduced kinetic models and a simulation software, a very good real sense of ion exchange or releasing reaction of molybdate from the LDH will be obtained. Due to similarity of many releasing or ion exchange phenomena, the results can be developed to many other reactions. Now, the effect of each parameter on the released medium is evaluated individually by the surface plots at the following.

Fig. 9.

Desirability of the determined targets in introduced model for prediction of released inhibitor from the LDH (a) minimum releasing; (b) maximum releasing; (c) 15ppm releasing, after 10min.

(1.1MB).

Fig. 10 shows the effect of two variables (MII/MIII and pH) on the amount of released inhibitor into the 3.5% NaCl solution after 10min. It can be seen that these two variables have significant effect on the behavior of each other. In other words, they have serious interaction with each other. As can be seen in the low pH close to neutral, increasing the ratio of MII/MIII up to 3 cause reduction of released inhibitor. But using the ratio of MII/MIII more than 3 result in more released inhibitor in the solution. In fact, the highest released inhibitor is obtained in the low and high values of MII/MIII. This behavior completely changes in pHs more than 10. It means that the maximum released inhibitor achieves in MII/MIII ratio of 3. It is worth to mention that more inhibitor is released in high pHs. This can be related to the better formation of the LDH in high acidities because as was explained earlier these structures are metal hydroxide which naturally forms better in alkaline pHs. Definitely, this must be noticeable by XRD analysis. Fig. 11 demonstrates the XRD spectrums of two synthesized LDH which are synthesized in different pH. The synthesizing pH for experiment 1 was about 8 and the experiment 3 was synthesized at pH 11. In fact XRD pattern of LDH with high pH shows more sharp and intense peaks which might indicate more crystallinity of the LDH crystals or better formation of them. As mentioned, since the LDH is a hydroxide compound, it has greater chance for the formation at high levels of pH and the XRD spectrums confirm this claim. Another point which is detectable in XRD spectrums is the peak shape alteration at around 35°. In fact it is noticeable that there is splitting of the peak in this region. This implies a difference in the structure by changing the pH. These claims can be clarified by more analysis and considerations.

Fig. 10.

The simultaneous effects of MII/MIII and pH on the released inhibitors from LDH in solution NaCl 3.5% after 10min.

(0.61MB).
Fig. 11.

XRD spectrum of the synthesized LDH (Experiment 1: MII/MIII 2, pH 8, aging temperature 45°C and aging time 15h and Experiment 3: MII/MIII 2, pH 11, aging temperature 45°C and aging time 15h).

(0.38MB).

Fig. 12(a) illustrates the SEM result of the LDH with the following synthesis condition. The sample is aged at temperature of 318K (45°C) for 15h. The ratio of MII/MIII is 2 and the solution pH was 8. As seen the containers are similar to flakes with nanometric scale in one dimension (nano dimension is more visible in other cases). The same morphologies were observed in others researches [24,25]. Fig. 12(b) shows the morphology changes by changing the MII/MIII to 4. Although flake morphology is perfectly preserved but it looks more thin flakes are obtained. The nanometric scale of the flakes is more visible in this figure. Increasing the MII/MIII ratio means to reduce the concentration of trivalent ion. In other words, less divalent ions have been replaced by trivalent ions and accordingly the network will not be able to absorb molybdate anions with previous capacity. Because this event cause electrostatic interaction decrement between the positive layers and the negative interlayers. Fig. 13 illustrates the XRD spectrums of these two synthesized LDH. In a comparative study, it can be noticed that three peaks which are located at 10–20 intervals, have suffered a sharp drop (marked in the figure by circle). These peaks are related to the intercalated molybdenum oxide. Thus, the mentioned hypothesis can be confirmed.

Fig. 12.

SEM image of the synthesized LDH, (a) MII/MIII: 2, pH solution: 8, aging temperature: 45°C for 15h; (b) MII/MIII: 4, pH solution: 8, aging temperature: 45°C for 15h; (c) MII/MIII: 2, pH solution: 11, aging temperature: 45°C for 15h.

(0.46MB).
Fig. 13.

XRD spectrum of the synthesized LDH (Experiment 1: MII/MIII 2, pH 8, aging temperature 45°C and aging time 15h and Experiment 2: MII/MIII 4, pH 8, aging temperature 45°C and aging time 15h).

(0.1MB).

But the point is complete change of the behavior when the pH varies from 8 to 11. Fig. 12(c) illustrates the morphology of the LDH synthesized in pH 11. It can be seen that, the flakes are accumulated in different regions and create special forms. It seems that there are spheres with the flake surface morphology. In other words, internal arrangement of atoms or ions does not affect seriously but particle size and morphology is affected by pH alteration. Panda et al. [26] previously reported the same observation. Panda et al. reported that use of sodium hydroxide significantly lowers the growth rate of the LDH nucleus. They claim that it occurs due to the formation of capping layer of Na+ all around the LDH nucleus (Fig. 14). In this study also the pH adjustment was performed using sodium hydroxide. In fact, based on Panda et al. founding high pHs with more Na+ concentration, encourage the growth pattern to a specific shapes as demonstrated in Fig. 15. The observed morphology and the XRD spectrum show that the LDH crystals are better form and also more molybdate were successfully intercalated between the layers in high levels of pH. Therefore the reported trend in Fig. 10 seems acceptable.

Fig. 14.

Schematic representation of in situ Na+ passivation layer around the LDHs nuclei [26].

(0.1MB).
Fig. 15.

SEM image of the synthesized LDH, MII/MIII: 2, pH solution: 11, aging temperature: 45°C for 15h.

(0.52MB).

Fig. 16 indicates the aging temperature effect beside the pH effect. The pH effect was previously discussed and here there is the same behavior. But in case of aging effect, it can be seen that aging temperature increment result in less released inhibitor in the solution. It can be related to various reasons. One reason can be related to temperature sensitivity of the intercalated anions. In other words releasing some inhibitors during the aging process is not farfetched. XRD analysis can be useful to confirm this claim. Fig. 17 compares the XRD spectrums of two synthesized LDH with different aging temperature and time. It can be seen that the peaks intensity are declined and also some crystallographic planes had lost their peak in the XRD spectrum by increasing the aging time and temperature (marked in figure by arrows). In other words, crystal or grain growth (shown in Fig. 18) or any other major change in the LDH structure due to the long heat treatment may lead to release of the intercalated molybdate during the aging process. The grains growth is more noticeable when the aging time is long. Fig. 19 indicates the LDH crystals morphology after 40h aging treatment.

Fig. 16.

The simultaneous effects of pH and aging temperature on the released inhibitors from LDH in solution NaCl 3.5% after 10min.

(0.68MB).
Fig. 17.

XRD spectrum of the synthesized LDH (blue spectrum: MII/MIII 2, pH 8, aging temperature 45°C and aging time 15h and red spectrum: MII/MIII 2, pH 11, aging temperature 60°C and aging time 40h), arrows show the disappeared peaks after long aging. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

(0.09MB).
Fig. 18.

SEM image of the synthesized LDH, MII/MIII: 4, pH solution: 8, aging temperature: 45°C for 15h. (a) Magnification 2k×; (b) magnification 50k×.

(0.39MB).
Fig. 19.

SEM image of the synthesized LDH, MII/MIII: 2, pH solution: 8, aging temperature: 45°C for 40h. (a) Magnification 3k×; (b) Magnification 20k×.

(0.3MB).
3.3Corrosion studies

Electrochemical impedance spectroscopy was carried out in various days of immersion on the synthesized smart coatings. The HLDH samples contain 5wt % of LDH and LLDH samples contain 1wt% LDH. The equivalent circuits used in EIS observations are illustrated in Fig. 20(a).

Fig. 20.

Nyquist curves of aluminum samples after 10 days of immersion in a solution of 3.5% NaCl. Samples are 5054 uncoated aluminum, coated with PVA, coated with smart coating containing a low concentration of LDH (LLDH), coated with smart coating containing a high concentration of LDH (HLDH). (a) Equivalent circuit; (b) Nyquist plot; (c) Bode plot.

(0.4MB).

Fig. 20(b) shows the obtained Nyquist curves from 4 samples after 10 days of immersion. Other plots are not presented and just their data are shown in Table 4. In most cases it is observable that the solution resistance is found as about 35Ω. The most important element in all cases which can be correlated with the corrosion resistance is Rct which called charge transfer or polarization resistance. It can be seen that by applying the PVA coating the polarization resistance is enhanced from 10kΩ to 20kΩ. Using LDH nanocontainers containing molybdate inhibitors also significantly improved the polarization resistance.

Table 4.

EIS results extracted from the equivalent circuit for the considered coatings after various days of immersion in 3.5% NaCl solution.

Sample  Elements  Days of immersion
    3 days  6 days  10 days  14 days  20 days 
Uncoated aluminumRΩ or Rsolution  41.23  34.93  37.47  29.42  30.78 
Zf or Rct  10,412  11,840  12,689  10,901  10,713 
P1  2.01E−05  2.43E−05  2.58E−05  2.30E−05  2.67E−05 
n1  0.84  0.85  0.78  0.81  0.74 
W  –  852.19  575.24  731.39  390.02 
Al coated with PVARΩ or Rsolution  150.95  158.77  150.58  148.94  148.91 
Zf or Rct  20,510  29,004  13,749  9861.2  9550.4 
RL  3679  6096.70  4889.3  5686.1  5877.5 
P1  7.46E−06  9.31E−06  1.06E−05  1.11E−05  1.12E−05 
n1  0.85  0.83  0.82  0.82  0.82 
P2  2.76E−05  4.92E−05  5.10E−05  5.44E−05  5.53E−05 
n2  0.51  0.53  0.47  0.54  0.56 
Al coated with PVA/LLDHRΩ or Rsolution  147.29  192.08  144.17  137.5  122.32 
Zf or Rct  47,451  33,766  20,126  19,543  19,406 
RL  246.04  268  214.89  228.83  233.3 
P1  3.08E−05  3.13E−05  4.24E−05  4.16E−05  4.79E−05 
n1  0.56  0.60287  0.58  0.60  0.51 
C  2.41E−06  2.61E−06  2.96E−06  3.01E−06  5.03E−06 
Al coated with PVA/HLDHRΩ or Rsolution  136.33  73.45  54.304  59.03  56.529 
Zf or Rct  38,655  45,214  35,567  25,587  33,441 
RL  229.33  71.08  82.43  70.74  74.421 
P1  6.47E−05  6.90E−05  7.19E−05  7.18E−05  7.6E−05 
n1  0.56  0.65  0.65  0.69  0.65 
C  3.09E−06  1.21E−08  7.50E−09  9.53E−09  9.10E−09 

Results show that, passing the time leads to change the corrosion properties of the all samples. In fact, the corrosion resistance has improved by applying the PVA coating from 10kΩ to 20kΩ. But after some days of immersion, the polymer structure is destroyed by diffusion of the chloride ions and the polarization resistance is declined. In other words, after some days of immersion, the PVA coating cannot protect the substrates and the corrosion properties become similar to the uncoated condition. Only in loaded coatings by LDH, the conditions are different. The LLDH sample with low concentration of inhibitor cannot stand against the chloride ions attacks, but the sample with high percentage of LDH has shown better behavior. It seems that, unlike early days, HLDH sample has more favorable result than other ones.

The capacitance or admittance of the capacitor or constant phase element in the equivalent circuit for the HLDH samples was monitored and a declining trend was found for them. This can be related to replacing the inhibitor molecules with the water molecules [27]. Another reason of decrease in capacitance can be related to inhibitor adsorbing on the metal surface. In fact, since the capacitance is inversely proportional to the thickness of the double layer, a decreasing in the capacitance values could be attributed to the adsorption of the inhibitor onto the metal surface. By entering the inhibitor molecules into the coating network, the active sites will be covered by them and forming a smooth surface. Consequently the capacitance will be decreased due to changes in the surface area. In general, the surface area of the rough and porous electrode is larger than smooth surface. Therefore, it seems that by passing the time, molybdate molecules will be adsorbed on the metal surface and the protective film will be grown until it covers all the possible corroding sites and cause decrement in the capacitance values. As can be seen, the capacitance values are significantly decreased by passing the times for the HLDH samples and this is a good evidence of inhibitor releasing during the immersion days because it doesn’t see in the pure PVA or even LLDH samples.

Fig. 21 shows the plots of the capacitance versus time with low and high concentration of LDH in the coating network. Generally in corrosion tests without inhibitor the diameter of the high frequency semicircle is treated as the charge transfer resistance, and in corrosion tests with inhibitor the high frequency capacitive loop is related to the barrier and properties of the inhibitor layer. From Fig. 21 it can be seen that the capacitance of the HLDH sample shows a sharp decrease in their values as time elapsed. However, in case of LLDH sample, the capacitance values are higher, and their values are approximately constant. It is clear that the presence of the inhibitor with enough concentration has a marked effect on the values of the capacitance. Capacitance values in LLDH sample have almost the same values and all shown the same trend. It is known that a decrease in the capacitance can happen if the inhibitor molecules (low dielectric constant) replace the water molecules (high dielectric constant) on the coating network. Therefore it can be said that this behavior is the sign of coating mechanism activation during the immersion days.

Fig. 21.

The capacitance changes of the capacitor element used in HLDH sample equivalent circuit during 20 days of immersion.

(0.06MB).

Fig. 22 shows the potentiodynamic curves of the Al alloys with various coating system immersed in 3.5% NaCl solution. The results confirmed the previous results of EIS measurements. As it can be seen, applying the PVA coating whether with and without LDH crystals shift the potentials into the less negative potentials. In addition, the corrosion current densities have been decreased especially in case of LDH loaded samples. It means that the corrosion phenomenon is thermodynamically and kinetically controlled by applying the PVA coating (with and without LDH). This can be related to the passivation role of the PVA coating and also inhibition role of the intercalated inhibitors. From thermodynamic view, PVA coating without LDH crystals shows fewer tendencies to corrode (less negative potential) due to better polymeric structure (as mentioned, LDH intercalating into the PVA will destroy the polymer network). However, from kinetic view, LDH crystals significantly decrease the corrosion rate due to their effect on surface condition.

Fig. 22.

Polarization curves of uncoated and coated Al alloy (coated with PVA, PVA contain low concentration of LDH and high concentration of LDH) in 3.5% NaCl solution.

(0.08MB).

The noticeable point, in coatings contain LDH crystals, is the current fluctuation in anodic branch of the polarization curve. In similar cases it was reported that these instabilities can be associated with initiation and repassivation of the sites for pits prior to the propagation of stable pits. The instabilities or current fluctuations by increasing the potential observable in Fig. 21, indicates a special case (corrosion initiation followed by repassivation) which can be studied or analyzed based on other founded data. As mentioned previously the corrosion prevention mechanism of the provided smart coatings is based on inhibitor releasing at the necessary times. Therefore observing the current fluctuation by increasing the potential is not farfetched because the surface condition of the coated aluminum by provided smart coating is similar to the surface exposed to pitting corrosion. Maybe Fig. 23[28] can schematically show the mentioned condition.

Fig. 23.

Schematic illustration of anion exchange in hydrotalcite compounds [28].

(0.32MB).

The hosts which are open channels, discussed previously, can accommodate anions and solvent molecules. In the initial state, LDH crystals contain inhibiting anions in their open channels. As is depicted schematically in Fig. 23, on contact with an aggressive electrolyte containing chloride ions, an exchange reaction will be occurred in which the inhibitor anion will be released, and chloride ions will be trapped into the LDH structure [29]:

HTInh+NaCl(aq)HTCl+NaInh(aq)

Thus, the released and adsorbed inhibitors and also the trapped anions can be the main reasons of the current fluctuation due to their effect on the substrate passivation or the generated interruption in the corrosion process.

The polarization measurements were performed in various days of immersion in order to detect the changes in corrosion properties by passing the time. Fig. 24 illustrates the polarization curves for the HLDH sample during 20 days of immersion in 3.5% NaCl solution. As it can be seen all curves indicate the current fluctuation in their anodic branch although after 20 days this is not highlighted yet. In other words, the previous interpretations are confirmed by these curves and the activation of the coating mechanism can be sensed apparently.

Fig. 24.

Polarization curves of coated Al alloy (PVA contain high concentration of LDH (5wt%) in 3.5% NaCl solution during different immersion times).

(0.09MB).
4Conclusion

Based on observations and released profiles a prediction model was introduced in order to estimate the amount of released inhibitor from the LDH after a certain time in NaCl 3.5% solution. The model expresses that the pH of the synthesis environment has special effect on the kinetic of the inhibitor releasing. Also MII/MIII has serious effect and its interaction with the pH is noticeable. Although the aging parameters have tangible effect on the structure and morphology of the LDH, but the pH and MII/MIII are critical factors in controlling the desired morphology and structure. The introduced model has the ability of access any target by establishing a certain value for each factor. The whole mechanism of releasing molybdate from the Mg-Al LDH was recognized as Fickian diffusion and the released data were well fitted with the Weibull model. In order to evaluate the activation of smart species in the coating network EIS was used and interpreted. It was observed that the capacitance values as an equivalent circuit element are significantly decreased by passing times. This fact shows that the release action is performing during the immersion days. More LDH in the coating network result in better resistance to corrosion for the produced coating up to approximately 5.7wt% of the Cl ions. Then adding more LDH to the coating, result in resistance decrement due to the defect formation in the polymer network. Apparently this value depends on the chloride ions concentration and changing the invasive ions concentration will change the optimum value of the required LDH amount.

Conflicts of interest

The authors declare no conflicts of interest.

References
[1]
G. Jeon, S.Y. Yang, J.K. Kim.
Functional nanoporous membranes for drug delivery.
J Mater Chem, 22 (2012), pp. 14814-14834
[2]
M.F. Montemor.
Functional and smart coatings for corrosion protection: a review of recent advances.
Surf Coat Technol, 258 (2014), pp. 17-37
[3]
M.L. Zheludkevich, J. Tedim, M.G.S. Ferreira.
Smart coatings for active corrosion protection based on multi-functional micro and nanocontainers.
Electrochim Acta, 82 (2012), pp. 314-323
[4]
D. Raps, T. Hack, M. Kolb, M.L. Zheludkevich, O. Nuyken, J. Baghdachi, et al.
American Chemical Society, (2010), pp. 165
[5]
N.Y. Abu-Thabit, A.S. Hamdy.
Stimuli-responsive polyelectrolyte multilayers for fabrication of self-healing coatings – a review.
[6]
M.P. Nikiforov, M.V. Chernysheva, A.A. Eliseev, A.V. Lukashin, Y.D. Tretyakov, Y.V. Maksimov, et al.
Synthesis of iron oxide nanocomposites using layered double hydroxides.
Mater Sci Eng B, 109 (2004), pp. 226-231
[7]
J. Arratia-Quijada, C. Sánchez Jiménez, A. Gurinov, A. Pérez Centeno, I. Ceja Andrade, G.G. Carbajal Arízaga.
Dysprosium-containing layered double hydroxides nanoparticles intercalated with biologically active species as an approach for theranostic systems.
Mater Sci Eng B, 203 (2016), pp. 7-12
[8]
M. Herrero, P. Benito, F.M. Labajos, V. Rives.
Stabilization of Co2+ in layered double hydroxides (LDHs) by microwave-assisted ageing.
J Solid State Chem, 180 (2007), pp. 873-884
[9]
T.T.X. Hang, T.A. Truc, N.T. Duong, P.G. Vu, T. Hoang.
Preparation and characterization of nanocontainers of corrosion inhibitor based on layered double hydroxides.
Appl Clay Sci, 67 (2012), pp. 18-25
[10]
M.F. Montemor, D.V. Snihirova, M.G. Taryba, S.V. Lamaka, I.A. Kartsonakis, A.C. Balaskas, et al.
Evaluation of self-healing ability in protective coatings modified with combinations of layered double hydroxides and cerium molybdate nanocontainers filled with corrosion inhibitors.
Electrochim Acta, 60 (2012), pp. 31-40
[11]
M.L. Zheludkevich, S.K. Poznyak, L.M. Rodrigues, D. Raps, T. Hack, L.F. Dick, et al.
Active protection coatings with layered double hydroxide nanocontainers of corrosion inhibitor.
Corros Sci, 52 (2010), pp. 602-611
[12]
M.L. Zheludkevich, J. Tedim, C.S.R. Freire, S.C.M. Fernandes, S. Kallip, A. Lisenkov, et al.
Self-healing protective coatings with “green” chitosan based pre-layer reservoir of corrosion inhibitor.
J Mater Chem, 21 (2011), pp. 4805-4812
[13]
X. Gao, L. Lei, D. O’Hare, J. Xie, P. Gao, T. Chang.
Intercalation and controlled release properties of vitamin C intercalated layered double hydroxide.
J Solid State Chem, 203 (2013), pp. 174-180
[14]
F. Khodam, Z. Rezvani, A.R. Amani-Ghadim.
Enhanced adsorption of Acid Red 14 by co-assembled LDH/MWCNTs nanohybrid: optimization, kinetic and isotherm.
J Ind Eng Chem, 21 (2015), pp. 1286-1294
[15]
R. Rojas, A.F. Jimenez-Kairuz, R.H. Manzo, C.E. Giacomelli.
Release kinetics from LDH-drug hybrids: effect of layers stacking and drug solubility and polarity.
Colloids Surf A Physicochem Eng Aspects, 463 (2014), pp. 37-43
[16]
O.K. Abiola, A.O. James.
The effects of Aloe vera extract on corrosion and kinetics of corrosion process of zinc in HCl solution.
Corros Sci, 52 (2010), pp. 661-664
[17]
W.T. Reichle.
Synthesis of anionic clay minerals (mixed metal hydroxides, hydrotalcite).
Solid State Ion, 22 (1986), pp. 135-141
[18]
A.N. Ay, B. Zümreoglu-Karan, A. Temel, L. Mafra.
Layered double hydroxides with interlayer borate anions: a critical evaluation of synthesis methodology and pH-independent orientations in nano-galleries.
Appl Clay Sci, 51 (2011), pp. 308-316
[19]
J. Tedim, M.L. Zheludkevich, A.N. Salak, A. Lisenkov, M.G.S. Ferreira.
Nanostructured LDH-container layer with active protection functionality.
J Mater Chem, 21 (2011), pp. 15464-15470
[20]
K. Kosmidis, P. Argyrakis, P. Macheras.
A reappraisal of drug release laws using Monte Carlo simulations: the prevalence of the Weibull function.
Pharm Res, 20 (2003), pp. 988-995
[21]
K. Kosmidis, P. Argyrakis, P. Macheras.
Fractal kinetics in drug release from finite fractal matrices.
J Chem Phys, 119 (2003), pp. 6373-6377
[22]
V. Papadopoulou, K. Kosmidis, M. Vlachou, P. Macheras.
On the use of the Weibull function for the discernment of drug release mechanisms.
Int J Pharm, 309 (2006), pp. 44-50
[23]
R.G. Brereton.
Applied chemometrics for scientists.
John Wiley & Sons, Ltd., (2007),
[24]
Y. Kuang, L. Zhao, S. Zhang, F. Zhang, M. Dong, S. Xu.
Morphologies preparations and applications of layered double hydroxide micro-/nanostructures.
Materials, 3 (2010), pp. 5220
[25]
J. Tedim, S.K. Poznyak, A. Kuznetsova, D. Raps, T. Hack, M.L. Zheludkevich, et al.
Enhancement of active corrosion protection via combination of inhibitor-loaded nanocontainers.
ACS Appl Mater Interfaces, 2 (2010), pp. 1528-1535
[26]
H.S. Panda, R. Srivastava, D. Bahadur.
Synthesis and in situ mechanism of nuclei growth of layered double hydroxides.
Bull Mater Sci, 34 (2011), pp. 1599-1604
[27]
A.Y. Musa, R.T.T. Jalgham, A.B. Mohamad.
Molecular dynamic and quantum chemical calculations for phthalazine derivatives as corrosion inhibitors of mild steel in 1M HCl.
Corros Sci, 56 (2012), pp. 176-183
[28]
R.G. Buchheit, H. Guan, S. Mahajanam, F. Wong.
Active corrosion protection and corrosion sensing in chromate-free organic coatings.
Prog Org Coat, 47 (2003), pp. 174-182
[29]
S. Miyata.
Anion-exchange properties of hydrotalcite-like compounds.
Clays Clay Miner, 31 (1983), pp. 305-311
Copyright © 2019. The Authors
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Tools
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.