Surface & Coatings Technology 309 (2017) 609–620

Contents lists available at ScienceDirect

Surface & Coatings Technology

journal homepage: www.elsevier.com/locate/surfcoat

Self-healing ability of nanoclay-based hybrid sol-gel coatings on

magnesium alloy AZ91D
Swapnil H. Adsul a,c, T. Siva b, S. Sathiyanarayanan b, Shirish H. Sonawane c, R. Subasri a,⁎
a
Centre for Sol-Gel Coatings, International Advanced Research Centre for Powder Metallurgy and New Materials (ARCI), Balapur, Hyderabad, 500005 Telangana State, India
b
Corrosion and Materials Protection Division, CSIR-Central Electrochemical Research Institute, Karaikudi, 630006 Tamil Nadu, India
c
Department of Chemical Engineering, National Institute of Technology, Warangal, 506004 Telangana State, India

a r t i c l e i n f o a b s t r a c t

Article history: The main objective of the study was to investigate the effect of loading cationic corrosion inhibitors into
Received 9 August 2016 halloysite nanotubes on the corrosion resistance of self-healing coatings derived from these materials.
Revised 28 November 2016 For this purpose, corrosion inhibitors Ce3+/Zr4 + were encapsulated into halloysite clay nanotubes. The en-
Accepted in revised form 4 December 2016
capsulated nanotubes were dispersed into a hybrid silica sol-gel matrix and deposited on magnesium alloy
Available online 7 December 2016
AZ91D substrates by dip coating method. The coatings were cured at 130 °C for 1 h in air. The bare, matrix
Keywords:
sol coated and self-healing sol coated substrates were evaluated for their self-healing and corrosion protec-
Self-healing tion abilities using weight loss experiments, potentiodynamic polarization and electrochemical impedance
Sol-gel coatings spectroscopy measurements after immersion in 3.5% NaCl solution for varying time durations between 1 h
Halloysite nanotubes to 24 h. These results were further corroborated with those obtained from scanning vibrating electrode
Mg alloy AZ91D technique. The self-healing property and the promising nature of cationic corrosion inhibitor loaded
Scanning vibrating electrode technique (SVET) halloysite nanotube based corrosion protection coatings on AZ91D substrates could be confirmed from
Corrosion resistance the present investigations.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction resistance of Mg alloys, e.g. (a) by producing more homogeneous

microstructure by rapid solidification; (b) by decreasing impurities
Magnesium alloys, owing to their high strength-to-weight ratio level; (c) by increasing Al content or (d) by deposition of protective
are being extensively used in automobile and aerospace industries coatings [5]. Surface treatment or coatings is the most commonly
as a substitute material to steels and aluminum alloys [1]. The need used strategy to protect the metal structures. Different methods
to reduce the weight of automobile and aircraft components is a re- have been investigated for surface protection such as conversion
sult of the requirement with regard to bringing down the emissions coatings, vapor deposition methods, anodizing, plasma treatment,
[2,3]. However, the poor resistance of Mg alloys to corrosion, wear laser/electron beam treatment, hot-diffusion alloying, sol-gel coat-
and creep in addition to high chemical reactivity, has prevented ing and organic coating, etc. [1].
their use to a large extent in any industry. One of the most impor- The most promising strategy to overcome the problem of corro-
tant reasons for the poor corrosion resistance of Mg alloys is inter- sion in magnesium alloys is to use hybrid sol-gel coatings. Sol-gel
nal galvanic corrosion caused by secondary phases of alloying coatings exhibit a high potential to replace environmentally hazard-
elements such as Mg 17 Al 12 , AlMn and Al 8 Mn 5 . Another reason ous chromate conversion coatings by providing a barrier against cor-
that contributes to the poor corrosion resistance of Mg alloys is rosive media. The major advantage of hybrid sol-gel coatings is that
the presence of thin film of oxides/hydroxides on their surface, crack-free thick coatings with controlled composition and morphol-
which is less protective than the film formed on aluminum and its ogy can be obtained at low curing temperatures. Further, sol-gel
alloys [4]. Therefore, improving the corrosion resistance of Mg al- technology also provides advantages such as low cost, low life-
loys is very essential when they are used for industrial applications. cycle environmental impact and ease of application on industrial
Several attempts have been made to improve the corrosion scale [6]. Sol-gel coatings can be applied to metal substrates by sim-
ple techniques such dip coating, spray coating and spin coating.
However, even if a small defect in the coating is present or developed
during service, the corrosive medium is easily able to reach the metal
⁎ Corresponding author. surface, which enhances the corrosion rate. Therefore, development
E-mail address: [email protected] (R. Subasri). of coatings with an ability to provide prolonged barrier protection

http://dx.doi.org/10.1016/j.surfcoat.2016.12.018
0257-8972/© 2016 Elsevier B.V. All rights reserved.
610 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

and active corrosion protection (self-healing ability) at the corroded

site is the need of the hour. In this context, generally two approaches
have been employed; either introducing the corrosion inhibiting ma-
terial directly into the coating matrix [7,8] or encapsulating it in
micro/nanocontainers, such as carbon nanotubes, nanoporous
metal oxide particles, polymeric capsules to facilitate their con-
trolled release [9–11]. However, since a direct introduction of the in-
hibitor in the coating causes its degradation and deactivation of
inhibitor, this method is generally not used for achieving prolonged
corrosion protection. As an alternative, encapsulating inhibiting ma-
terial into nanocontainers for controlled release is considered to be
an effective strategy for achieving corrosion protection with good
barrier properties and self-healing ability. However, for large scale
application of nanocontainer based self-healing coatings on the sub-
strate, there are limitations with respect to costs. Hence, cost-
effective alternatives for these nanocontainers could be the use of
naturally available clays [12]. Naturally occurring organic clay min-
erals such as halloysite and montmorillonite have advantages such
as being non-toxic, cheap, easily available and eco-friendly. Past in-
vestigations have revealed that, 1–5 wt% addition of clays into coat- Fig. 2. (a) TEM image and (b) EDS analysis of as-received HNT; (c) TEM image of etched
ing formulations enhance the anticorrosive, barrier, thermal and HNT.
mechanical properties [13–15]. Halloysite nanoclays are double-
layered aluminosilicate with concentric tubular structure, chemical-
ly similar to kaolin. They have concentric positively and negatively 2.2. Sol synthesis and coating deposition
charged inner and outer layers of Al2O3 and SiO2, respectively. More-
over, they have the capacity to entrap wide range of active materials Halloysite nanotubes (Sigma Aldrich, USA) having a lumen diam-
such as drug particles, catalysts, inhibitors, marine biocides, etc. eter of ~ 10 nm were etched in 1 M H2SO4 (SD Fine Chemicals, India,
within the lumen and void spaces in the multilayered aluminosili- 98%) for 3 days at 60 °C, followed by washing with deionized water
cate structure [12]. and drying, in order to get enlarged lumen diameter [17]. This was
Though there have been reports on the use of halloysite nano- carried out in order to incorporate more amount of corrosion inhib-
tubes (HNT) as nanocontainers for corrosion inhibitors and appli- itors such as cerium nitrate hexahydrate (Loba Chemie, India,
cation on Al alloys such as AA 2024 [12,16–19], there are no 99.9%) and zirconium n-propoxide (Gelest Inc., USA, 70% in
reports so far, to the best of our knowledge on the use of HNT for propanol) which were taken in a molar ratio of 1:32. The mixture
encapsulating corrosion inhibitors for application on Mg alloys. was then dried in vacuum desiccator. The organic-inorganic hybrid
Hence, the objective of this work was to study the self-healing matrix (hereinafter labeled as MAT) sol was prepared by hydrolysis
corrosion protection ability of coatings that comprise cationic cor- of 3-Glycidoxypropyltrimethoxysilane (GPTMS, Gelest Inc., USA,
rosion inhibitor loaded HNTs dispersed into a hybrid sol-gel 98%) with tetraethoxysilane in molar ratio of 3:5:1 with 0.1 N HCl
matrix. as catalyst. The polymeric microcapsules which were used as cap-
ping agents to stopper the ends of the HNTs were synthesized
using the method reported by White et al. [20,21] and processed
2. Materials and methods using the method reported in [22]. The self-healing (SH) sol com-
prises 2 wt% inhibitor loaded halloysite nanotubes stoppered with
2.1. Substrate polymeric microcapsules dispersed in hybrid matrix sol. The poly-
meric microcapsules acted as a capping agent for the halloysite
The substrate used was magnesium alloy AZ91D with chemical nanotubes. Coatings were generated using dip coating technique
composition (in wt%) by wet analysis: Al-9.14; Zn-0.86; Mn-0.30; using a withdrawal speed of 1 mm/s followed by curing at 130 °C
Cu-0.09; Si-0.08; Fe-0.01; Ni-0.01 and rest Mg. The AZ91D sub- for 1 h in a hot air oven. The coating thickness was measured using
strates were polished with 1000 grit emery sheet followed by re- a non-destructive coating thickness gauge PosiTector® 6000 sup-
peated degreasing with acetone for 30 min and finally drying in plied by DeFelsko Corporation, USA. Adhesion testing was carried
air. out using a cross-hatch cutter according to ASTM D 3359.

Fig. 3. FESEM images of polymeric microcapsule (a) at 5000× magnification and (b) at 50
Fig. 1. FESEM image of as received HNTs. kX magnification.
 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620 611

200 G2, FEI, Netherlands). The length of the nanotubes was also
confirmed using Scanning Electron Microscope (SEM) Hitachi,
model-S3400 N. BET surface area and pore volume measurements
for the as-received, etched and the loaded HNTs were carried out
using a Micromeritics ASAP 2020 surface area and porosimetry
analyzer.

2.4. Weight loss measurements

Weight loss measurements were carried out as per ASTM G31 [23,
24]. AZ91D substrates of dimensions 2.5 cm × 2 cm were used for
weight loss experiment. Bare and coated substrates of AZ91D were im-
mersed in 200 ml 1 wt% NaCl solution for 120 h. Substrates were
washed with deionized water and ethanol followed by immersion in
chromic acid (200 g/l) for 2 min to remove the corrosion products
from the surface. The substrates were again washed with deionized
water and ethanol. The weights of substrates before and after the expo-
sure were measured and the corrosion rates were obtained from the
Fig. 4. Pore volume vs. pore Diameter for as-received, loaded and etched HNTs. weight loss values. The corrosion rates were checked for consistency
by measuring weight losses of three samples each of bare and coated
substrates.
2.3. Transmission electron microscopy, scanning electron microscopy and
BET pore volume analysis 2.5. Salt immersion tests

The lumen diameter and length of the nanotubes was Salt immersion tests were carried out on scribed substrates of di-
ascertained using a transmission electron microscope (TEM, Tecnai mensions 2.5 cm × 2 cm. MAT and SH sol coated substrates were

Fig. 5. Photographs of (a) MAT sol and (b) SH sol coated AZ91 before salt immersion; (c) MAT sol and (d) SH sol coated AZ91 after salt immersion.
612 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

Table 1
Elemental analysis [in wt%] of MAT and SH sol coated substrates before and after 3.5 wt% salt immersion for 120 h.

Elements MAT SH

Before salt immersion After salt immersion Before salt immersion After salt immersion

Scratch Surface Scratch Surface Scratch Surface Scratch Surface

Mg 78.2 17.5 61.0 56.6 88.8 38.3 25.9 65.2

Al 10.5 1.3 7.1 6.4 6.7 1.9 0.8 2.6
Si 5.7 18.1 3.3 5.2 0.1 9.2 0.8 1.4
O 5.6 23.0 24.7 16.3 0.7 20.9 43.2 17.2
C 0.0 40.1 3.9 15.5 1.5 27.0 28.0 10.7
Ce – – – – 0.5 0.6 0.7 0.8
Zr – – – – 0.3 1.1 0.3 0.3

exposed to 3.5 wt% NaCl solution for 120 h. Substrates were scribed so morphology and elemental mapping using Scanning Electron Micro-
as to remove the coating on the scribed area and expose the metal sur- scope (SEM) Hitachi, model-S3400 N. The elemental compositions of
face to the corrosive medium. The scribed substrates prior to and after coatings and substrates were analyzed for the changes by using SEM-
immersion in NaCl solution were then analyzed for their surface EDS.

Fig. 6. Elemental mapping using EDS for SH sol on scratched area (a) before salt immersion; (b) after salt immersion and on surface (c) before salt immersion; (d) after salt immersion.
 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620 613

Fig. 6 (continued).

2.6. Scanning vibrating electrode technique measurements to 3.5 wt% NaCl solution after creation of artificial defects with
sizes ranging from 0.1 to 0.3 mm2. Current density data was obtain-
Scanning Vibrating Electrode Technique (SVET) with Biologic ed after different durations of exposure. Anodic and cathodic re-
SCV 470 control unit was used to analyze the self-healing ability gions represented by positive and negative current densities
of bare and coated AZ91D substrates. SVET measurements were respectively were depicted in the form of 3D current density
carried on samples with dimensions 10 mm × 10 mm and a scan maps for each scan area.
area of 4 mm × 4 mm. Bare and coated substrates were exposed
2.7. Electrochemical tests

Electrochemical Analyzer, CH Instruments Model 600E Series

was used for the potentiodynamic polarization and electrochemi-
Table 2
cal impedance measurements. The system had a three electrode
Results of weight loss measurements of bare and coated AZ91D after exposure to 1 wt%
NaCl for 120 h. configuration including saturated calomel electrode (SCE) as refer-
ence electrode, platinum electrode as counter electrode and the
Substrate Wt. loss (g) Corrosion rate (mmpy)
sample as working electrode with an exposure area of 1 cm 2. The
Bare 0.0433 ± 0.0056 4.14 ± 0.08 electrochemical impedance spectroscopy (EIS) was carried out
MATRIX 0.0403 ± 0.0044 3.19 ± 0.41 over a frequency range of 1 MHz to 0.01 Hz using an AC signal of
SH 0.0245 ± 0.0038 2.02 ± 0.30
10 mV amplitude. The potentiodynamic polarization and open
614 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

Fig. 7. Current density maps for the bare AZ91after exposure to 3.5 wt% NaCl: (a) Initial, (b) after 1 h, (c) after 12 h and (d) after 24 h.

circuit potential measurements were carried out on AZ91D after confirmed that there is an increase in the pore volume of the etched
immersion in 3.5 wt% NaCl solution for varying time durations. HNTs, which has decreased substantially (by nearly 80%) when
The potentiodynamic polarization measurements were carried compared to that of the as-received halloysite nanotubes, after
out using 0.8 mV/s as scan rate. loading with inhibitors and capping their ends. The decrease in
pore volume unequivocally confirmed that the loading of the inhib-
3. Results and discussion itors has indeed taken place inside the lumen of the HNTs.

3.1. Scanning electron microscopy, transmission electron microscopy, and 3.2. Salt immersion tests
BET pore volume analysis:
Scribed substrates after exposure to 3.5 wt% NaCl solution for
The FESEM analysis of as-received HNTs is shown in Fig. 1. As- 120 h were washed with deionized water to remove the accumulat-
received halloysite nanotubes when observed through TEM shows ed salt particles from the surface. The substrates were dried in air
a tubular structure with empty lumen which is ~ 10 nm diameter and observed for their surface morphology. Thicknesses of the ma-
and confirmed that the lengths were varying between 0.1 and trix sol coatings were measured to be in the range of 3–3.5 μm and
0.5 μm, as shown in Fig. 2(a). The elemental analysis carried out those for the self-healing sol coatings were found to be in the range
on the HNT as shown in Fig. 2(b) confirmed the composition of of 4–5 μm. Adhesion of coatings was determined to be of rank 5B
the clay to be aluminosilicate. The TEM analysis of the etched (0% removal after removal of tape as observed under an optical
tubes with increased lumen diameter (~ 30–35 nm) is shown in microscope).
Fig. 2(c). The changes in lumen diameter after etching and In case of MAT coated AZ91D, elemental analysis shown that a
after loading with corrosion inhibitors were corroborated negligible amount of oxides was present on the scribed area before
through BET pore volume and surface area analysis. SEM images salt immersion, which increased in large amount after salt immer-
of polymeric microcapsules that were used as stoppers of the HNT sion of 120 h in scratched area. In addition, decreased amount of car-
are shown in Fig. 3. The diameter of the microcapsules varied bon, oxygen and silica on the surface depicts loss of coating
from 0.1–1.5 μm. properties. Fig. 5(a) and (c) show the images of the scribed area
Pore volume as a function of pore diameter for as-received, and the surface of the MAT coated substrate, before and after salt im-
etched and inhibitor loaded HNT after etching as shown in Fig. 4 mersion respectively.
 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620 615

Fig. 8. Current density maps for the matrix sol coated AZ91 exposure time to NaCl: (a) Initial, (b) after 1 h, (c) after 12 h and (d) after 24 h.

The images of the scribed area and the surface for the SH coated 3.3. Weight loss experiments
AZ91D before and after salt immersion are shown in Fig. 5(b) and
(d) respectively. In this case, even though oxides have formed in Weight loss measurements can provide the reliable information
the scribed region after salt immersion, an increase in the amount regarding the corrosion resistance of a given material (metal/
of Ce was observed in the scribed area, as indicating that a passive alloy). In this case, corrosion resistance in the form of corrosion
layer consisting of cerium oxide could have formed that could rate has been calculated for bare, matrix and SH sol coated AZ91D
offer a self-healing effect and provide corrosion protection. The in 1 wt% NaCl for 120 h. The corrosion rate was calculated using
change in elemental compositions of substrates after salt immer- equation as given below [23]:
sion is presented in Table 1. The EDS mapping images of SH sol
coated substrates before and after salt immersion are shown in
8:76  104  w
Fig. 6. Corrosion rate ¼
tAρ
It should be noted that the amount of carbon and oxygen have
also increased in the scribed area after salt immersion. The reason
for this could be that the polymerization of the epoxy groups is Table 2 shows average weight losses and corrosion rates for bare
catalyzed by cations (Ce 3 + /Zr 4 + ) [25] released from the HNTs to and coated substrates. It can be discerned that the corrosion rate
yield polyethylene oxide (PEO) which can in turn provide a good and hence, weight loss is the least for the self-healing sol coated
barrier against the corrosive medium thereby preventing further substrate, followed by that for the hybrid matrix sol and highest
corrosion. This effect in addition to formation of passive oxide for the bare substrate. A nearly 50% reduction in the corrosion
layer of cerium/zirconium can enhance the self-healing effect, rate could be observed for the self-healing sol, whereas only a
which improves and prolongs the corrosion resistance even 11% reduction in the corrosion rate could be observed for the ma-
when there is damage to the coating. Further corroborations to trix sol coated substrate when compared to the bare substrate.
this expectation can be obtained from weight loss and electro- This clearly indicated that the coated substrates exhibited lesser
chemical analysis results of which are presented in the subsequent corrosion rates since the coatings initially act as a barrier to the cor-
sections. rosive medium.
616 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

Fig. 9. Current density maps for the SH Sol coated AZ91 exposure time to NaCl: (a) Initial, (b) after 1 h, (c) after 12 h and (d) after 24 h.

3.4. SVET measurements are shown in Figs. 7–9. Corrosion rate can be seen to increase at
the defect area, which can be inferred from the high anodic current
The SVET results of bare and coated magnesium alloy AZ91 flow indicated by red color, due to small anode large cathode
after immersion in 3.5% NaCl solution for varying time durations configuration. This indicates destruction of the passive oxide

Fig. 10. Equivalent electric circuits used for analyzing EIS data obtained for (a) bare and (b) coated substrates.
 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620 617

Table 3 3.5. Electrochemical measurements

Fitting parameters of bare and coated AZ91D substrates obtained from the EIS studies.

Sl. Exposure Substrate Rcoat Rct Ccoat/CPE Cedl/CPE Fitting In EIS, the substrates were exposed to 3.5 wt% NaCl and
no. time Ω·cm2 Ω·cm2 F/cm2 F/cm2 error impedance was measured and analyzed by using suitable equiva-
01 Initial Bare – 1572 – 4.5e−6 0.03 lent electric circuits for bare and coated substrates as shown in
Matrix 28.4 3107 8.1e−9 5.6e−6 0.02 Fig. 10(a) and (b) respectively. Several equivalent electric circuits
SH 38.5 3716 5.1e−9 9.5e−6 0.03 were analyzed and tested to fit the experimental impedance data.
02 1h Bare – 4025 – 4.8e−6 0.04
The circuit which gave the least error after fitting was selected. In
Matrix 39.6 5668 7.0e−8 6.9e−6 0.01
SH 71.8 9666 1.7e−7 7.1e−6 0.02 order to fit the impedance data, a constant phase element is used
03 12 h Bare – 6246 – 4.7e−6 0.05 rather than pure capacitance corresponding for the coating and
Matrix 56.1 1.1e4 2.9e−7 5.4e−6 0.01 electrical double layer. During initial time of exposure, the charge
SH 54.6 1.2e4 3.2e−8 5.9e−6 0.02 transfer resistance of SH sol coated substrate is the highest when
04 24 h Bare – 1.4e4 – 3.6e−6 0.17
Matrix 59.9 2.4e4 1.0e−5 6.6e−7 0.13
compared to that of bare and matrix sol coated substrate
SH 244.1 9.7e4 6.6e−7 1.9e−6 0.01 (Table 3). As shown in Fig. 11, SH sol coated substrates shows
higher impedance in low frequency region indicating lowest corro-
sion rate on the surface. The SH sol coated substrate shows highest
corrosion resistance even after 24 h, as shown in Fig. 12.
The results of the potentiodynamic polarization analysis for bare
layer due to which, the substrate surface is exposed to the corro- and coated AZ91D after immersion in 3.5 wt% NaCl for 1 h, 12 h and
sive medium. 24 h are shown in Fig. 13. The corrosion currents and potentials for
Fig. 8 shows current density maps for the matrix sol coated sub- different exposure times are shown in Table 4. Fig. 13(a) shows in-
strates after exposure to NaCl solution. At defect area, anodic current crease in corrosion potential of the bare substrate after initial expo-
has decreased after initial increment due to partial recovery of barrier sure, which tends to become more positive after 1 h, possibly due
property after 24 h. to the formation of a passive layer of oxides of magnesium on the
Fig. 9 indicates current density maps for Ce3 +-Zr 4 + loaded substrate. Corrosion current tends to decrease after 24 h, because
nanoclay based self-healing sol coated substrates after immersion of the destruction of this passive layer of oxides. Fig. 13(b) depicts
in the corrosive medium. High initial anodic current at the defect the polarization data for the matrix sol coated AZ91D exposed to
area was seen to decrease after 12 h, due to self-healing activity 3.5 wt% NaCl. It can be observed that after 24 h, the corrosion poten-
and after 24 h, current densities have reduced substantially, indicat- tial has become more positive indicating ennoblement of potential.
ing very low corrosion rates. The SVET data unequivocally confirms Fig. 13(c) shows the polarization plots for corrosion inhibitor loaded
corrosion inhibition due to the release of inhibitors from the lumen nanoclay based SH coatings on AZ91D. Initially, the corrosion current
of the clay nanotubes. These coatings have barrier property due to is high because of the poor barrier property of coatings due to the
the matrix sol and self-healing ability from the inhibitor-loaded dispersion of nanoclay particles. However, after 1 h, due to self-
nanoclays, which becomes effective when barrier property gets healing properties of inhibitor loaded nanoclay based sol, both corro-
destroyed due to change in pH at the defect area. The results of the sion potential and current drops down drastically. This confirms that
SVET measurements clearly indicate that the self-healing sol coated the nanoclay-based self-healing coatings are able to render corrosion
AZ91 substrate is able to autonomically heal at the defect area due protection even after longer duration of exposure to corrosive
to the self-healing effect that is triggered by change in the local pH. medium.
This leads to very low current densities even after prolonged expo- The potentiodynamic polarization and electrochemical imped-
sure to the corrosive medium, which is not the case with the uncoat- ance studies clearly confirm that the nanoclay-based self-healing
ed substrate. In case of the bare substrates, current density around coatings are able to provide improved corrosion resistance as com-
the defect area has increased on prolonged exposure to the corrosive pared to the coatings derived from the matrix sol only after exposure
medium. to corrosive medium. Even when there is a defect in the coating, the

Fig. 11. Nyquist plot for bare and coated substrates after immersion in 3.5% NaCl solution Fig. 12. Comparison of corrosion resistance of bare and coated AZ91D obtained from EIS
for 24 h. data.
618 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

Fig. 13. Potentiodynamic polarization data for (a) bare, (b) matrix sol coated and (c) SH sol coated AZ91D after immersion in 3.5 wt% NaCl solution as a function of time.

corrosion inhibitor is able to be released from the nanotubes which 4. Conclusion

offer corrosion protection by two methods. One is by formation of a
protective passive oxide layer and the other by formation of PEO 1) Cation corrosion inhibitors namely Ce3 +/Zr4 + loaded halloysite
due to polymerization of epoxy functional group from the matrix nanotubes dispersed in a hybrid silica sol-gel matrix was able to pro-
sol. Both these phenomena are initiated by corrosion inhibitor cat- vide prolonged corrosion protection when exposed to corrosive me-
ions, which render good barrier protection. The above-mentioned dium such as 3.5% NaCl.
autonomic-healing mechanism could be affirmed by the elemental 2) The autonomic healing properties of the coatings could be con-
mapping after exposing a scribed coated substrate to the corrosive firmed through weight loss, SVET and electrochemical
medium as well as by SVET measurements and electrochemical anal- measurements.
ysis. A schematic representation of self-healing mechanism of coat- 3) The hybrid sol-gel matrix could provide very good barrier protec-
ing showing crack initiation and self-healing of crack after tion, which could also be verified by the electrochemical
formation of passive layer along with formation of a layer of PEO in measurement.
presence of catalysts is depicted in Fig. 14. 4) The self-healing material studied in the present investigation
can be used as an additive to a sol-gel matrix as well as in con-
junction with paints for improving the barrier properties of
paints when they are applied over the primer layer.
Table 4 5) The nanoclay-based self-healing material was found to have promis-
Corrosion current and potential values of bare, matrix and SH sol coated AZ91D substrates. ing applications and can be a potential replacement for the hazard-
ous chromate-conversion coatings.
Sl. no. Exposure time Substrate Ecorr, V (vs. SCE) Icorr, A/cm2

01 Initial Bare −1.581 1.1e−4

Matrix −1.505 1.5e−5
Acknowledgements
SH −1.532 3.4e−5
02 1h Bare −1.495 2.8e−4 The authors would like to thank Director, ARCI, Hyderabad for the
Matrix −1.511 1.3e−5 constant encouragement and support throughout the entire duration of
SH −1.476 7.5e−7
work. The authors would like to thank G.V.R. Reddy and K.R.C. Soma
03 12 h Bare −1.520 2.4e−4
Matrix −1.460 5.0e−7 Raju for the SEM-EDS analysis and the adhesion testing respectively.
SH −1.508 1.6e−6 Swapnil H. Adsul and R. Subasri would like to acknowledge the financial
04 24 h Bare −1.562 6.0e−5 support from SERB, DST for the funding provided through grant number
Matrix −0.595 1.4e−6 SB/S3/ME/007/2014. T. Siva and S. Sathiyanarayanan thank CSIR - New
SH −1.507 0.9e−6
Delhi for financial support through 12 FYP project Intelcoat (CSC0114).
 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620 619

Fig. 14. Schematic representation of self-healing mechanism of coating: (a) as-generated coating; (b) after crack initiation and (c) self-healing layer formation.

References [12] D.G. Shchukin, S.V. Lamaka, K.A. Yasakau, M.L. Zheludkevich, M.G.S. Ferreira, H.
Möhwald, Active anticorrosion coatings with halloysite nanocontainers, J. Phys.
[1] R.G. Hu, S. Zhang, J.F. Bu, C.J. Lin, G.L. Song, Recent progress in corrosion protection of Chem. C112 (2008) 958–964.
magnesium alloys by organic coatings, Prog. Org. Coat. 73 (2012) 129–141. [13] E. Darmiani, G.R. Rashed, D. Zaarei, I. Danaee, Synergistic effects of montmorillonite/
[2] B.L. Mordike, T. Ebert, Magnesium-properties, applications and potential, Mater. Sci. cerium nitrate additives on corrosion performance of epoxy-clay nanocomposite
Eng. A302 (2001) 37–45. coatings, Polym.-Plast. Technol. Eng. 52 (2013) 980–990.
[3] G.L. Song, Corrosion electrochemistry of magnesium and its alloys, in: G.L. Song [14] E. Darmiani, I. Danaee, G.R. Rashed, D. Zaarei, Formulation and study of corrosion
(Ed.), Corrosion of Magnesium Alloys, Woodhead Publishing Lim, UK 2011, pp. 3–4. prevention behavior of epoxy cerium nitrate–montmorillonite nanocomposite coat-
[4] J.E. Gray, B. Luan, Protective coatings on magnesium and its alloys- a critical review, ed carbon steel, J. Coat. Technol. Res. 10 (4) (2013) 493–502.
J. Alloys Compd. 336 (2002) 88–113. [15] M. Fedel, E. Callone, S. Dire, F. Deflorian, M.G. Oliver, M. Poelman, Effect of Na-
[5] G.L. Song, A. Atrens, Understanding magnesium corrosion-a framework for im- montmorillonite sonication on the protective properties of hybrid silica coatings,
proved alloy performance, Adv. Eng. Mater. 5 (2003) 837–858. Electrochim. Acta 124 (2014) 90–99.
[6] X. Zhong, Q. Li, J. Hu, S. Zhang, B. Chen, S. Xu, F. Luo, A novel approach to heal the sol- [16] Y.M. Lvov, D.G. Shchukin, H. Möhwald, R.R. Price, Halloysite clay nanotubes for con-
gel coating system on magnesium alloy for corrosion protection, Electrochim. Acta trolled release of protective agents, ACS Nano 2 (2008) 814–820.
55 (2010) 2424–2429. [17] E. Abdullayev, Y. Lvov, Clay nanotubes for controlled release of protective agents - a
[7] S.V. Lamaka, M.F. Montemor, A.F. Galio, M.L. Zheludkevich, C. Trindade, L.F. Dick, review, J. Nanosci. Nanotechnol. 11 (2011) 10007–10026.
M.G.S. Ferreira, Novel hybrid sol-gel coatings for corrosion protection of AZ31B [18] E. Abdullayev, Y. Lvov, Clay nanotubes for corrosion inhibitor encapsulation: release
magnesium alloy, Electrochim. Acta 53 (2008) 4773–4783. control with end stoppers, J. Mater. Chem. 20 (2010) 6681–6687.
[8] X. Zhong, Q. Li, J. Hu, Y. Lu, Characterization and corrosion studies of ceria thin [19] D. Fix, D.V. Andreeva, Y.M. Lvov, D.G. Shchukin, H. Möhwald, Application of inhibi-
film based on fluorinated AZ91D magnesium alloy, Corros. Sci. 50 (2008) tor-loaded halloysite nanotubes in active anti-corrosive coatings, Adv. Funct.
2304–2309. Mater. 19 (2009) 1720–1727.
[9] H. Wei, Y. Wang, J. Guo, N.Z. Shen, D. Jiang, et al., Advanced micro/nanocapsules [20] E.N. Brown, S.R. White, N.R. Sottos, Microcapsule induced toughening in a self-
for self-healing smart anticorrosion coatings, J. Mater. Chem. A 03 (2015) healing polymer composite, J. Mater. Sci. 39 (2004) 1703–1710.
469–480. [21] S.R. White, N.R. Sottos, P.H. Guebelle, J.S. Moore, M.R. Kessler, S.R. Sriram, E.N.
[10] M.F. Montemor, Functional and smart coatings for corrosion protection- a review of Brown, S. Viswanathan, Autonomic healing of polymer composites, Nature 409
recent advances, Surf. Coat. Technol. 258 (2014) 17–37. (2001) 794–797.
[11] T. Nesterova, K.D. Johansen, S. Kiil, Synthesis of durable microcapsules for self- [22] C. Suryanarayana, K.C. Rao, D. Kumar, Preparation and characterization of microcap-
healing anticorrosive coatings: a comparison of selected methods, Prog. Org. Coat. sules containing linseed oil and its use in self-healing coatings, Prog. Org. Coat. 63
70 (2011) 342–352. (2008) 72–78.
620 S.H. Adsul et al. / Surface & Coatings Technology 309 (2017) 609–620

[23] ASTMG31 - 72, Standard Guide for Laboratory Immersion Corrosion Testing of [25] P. Kiruthika, R. Subasri, A. Jyothirmayi, K. Sarvani, N.Y. Hebalkar, Effect of plasma
Metals, 03-02, 2004. surface treatment on mechanical and corrosion protection properties of UV-curable
[24] M.-C. Zhao, P. Schmutz, S. Brunner, M. Liu, G.-l. Song, A. Atrens, An exploratory study Sol-gel based GPTS-ZrO2 coatings on mild steel, Surf. Coat. Technol. 204 (2010)
of the corrosion of Mg alloys during interrupted salt spray testing, Corros. Sci. 51 1270–1276.
(2009) 1277–1292.