Adsorption and anti-corrosion characteristics of vanillin Schiff bases on mild steel in 1 M HCl: Experimental and theoretical study

Article abstract of DOI:10.1039/c9ra07982c

Herein, two Schiff base derivatives of vanillin and divanillin with 2-picolylamine, namely, 2-methoxy-4-((pyridin-2-ylmethylimino)methyl)phenol (compound A) and 3,3′-dimethoxy-5,5′-bis-((pyridin-2-ylmethylimino)methyl)-[1,1′-biphenyl]-2,2′-diol (compound B), respectively, were synthesized. Additionally, their adsorption characteristics and corrosion inhibition behavior were compared for mild steel in 1 M HCl using electrochemical impedance spectroscopy, potentiodynamic polarization and weight loss methods. Compound B was found to impart a better anti-corrosive effect (around 95% inhibition efficiency at 313 K) than compound A. The inhibitors act as effective mixed-type inhibitors and exhibit Langmuir-type adsorption behaviour. The kinetic-thermodynamic parameters together with the data obtained from density functional theory (DFT) and molecular dynamics (MD) simulations illustrate the mechanism of corrosion and mode of adsorption of both inhibitors on the metal surface. The better corrosion mitigation propensity of the dimeric form of the inhibitor (compound B) over the monomeric form (compound A) was tested experimentally and explained according to the theoretical data.

Full text of DOI:10.1039/c9ra07982c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Adsorption and anti-corrosion characteristics of
vanillin Schiff bases on mild steel in 1 M HCl:
experimental and theoretical study†
Cite this: RSC Adv., 2020, 10, 9258
bd
bc
Sanjoy Satpati,a Sourav Kr. Saha,
Aditya Suhasaria,a Priyabrata Banerjee
*
a
*
and Dipankar Sukul
Herein, two Schiff base derivatives of vanillin and divanillin with 2-picolylamine, namely, 2-methoxy-4-
((pyridin-2-ylmethylimino)methyl)phenol (compound A) and 3,30-dimethoxy-5,50-bis-((pyridin-2-
ylmethylimino)methyl)-[1,10-biphenyl]-2,20-diol (compound B), respectively, were synthesized.
Additionally, their adsorption characteristics and corrosion inhibition behavior were compared for mild
steel in 1 M HCl using electrochemical impedance spectroscopy, potentiodynamic polarization and
weight loss methods. Compound B was found to impart a better anti-corrosive effect (around 95%
inhibition efficiency at 313 K) than compound A. The inhibitors act as effective mixed-type inhibitors and
exhibit Langmuir-type adsorption behaviour. The kinetic–thermodynamic parameters together with the
data obtained from density functional theory (DFT) and molecular dynamics (MD) simulations illustrate
the mechanism of corrosion and mode of adsorption of both inhibitors on the metal surface. The better
corrosion mitigation propensity of the dimeric form of the inhibitor (compound B) over the monomeric
form (compound A) was tested experimentally and explained according to the theoretical data.
Received 1st October 2019
Accepted 11th February 2020
DOI: 10.1039/c9ra07982c
rsc.li/rsc-advances
corrosion inhibitors. Schiff bases are versatile compounds,
which are synthesized via the condensation of primary amines
Introduction
and carbonyl compounds, and used widely in pharmaceuticals,
agrochemicals and materials science.23–25 In this work, we
aimed to investigate the corrosion inhibition properties of two
newly synthesized Schiff base derivatives of vanillin and diva-
nillin with 2-picolylamine for mild steel in 1 M HCl.
The application of suitable corrosion inhibitors for the control
of corrosion in metals and alloys is very important.1–4 However,
due to ecological concerns, the use of inorganic inhibitors is
gradually being restricted. This has resulted in a surge of
studies involving organic corrosion inhibitors. Organic
compounds containing N, S, and O atoms generally show good
inhibition efficiency for mild steel in acidic media.5–12 In addi-
tion to various heterocycles,13 amines14 and imines,15 different
other classes of organics, such as amino acids,16,17 vitamins,18
polysaccharides,19 surfactants,20 polypeptides,21 lipids,22 poly-
phenols1 and others, have been reported to act as efficient
Vanillin, a biomass-derived phenolic aldehyde, is widely
used as a avoring agent in foods, beverages and pharmaceu-
ticals owing to its anti-microbial and anti-oxidant properties.26
It was rst extracted from vanilla beans, which are primarily
obtained from the orchid Vanilla planifolia. The synthetic
production of vanillin from the abundant lignin via metal-
catalyzed air oxidation converts it into a potential renewable
feedstock chemical.27,28 Herein, we provide further value to
vanillin and explore its potential for applications in a new arena
of green corrosion inhibitors, which are essentially of bio-origin
and less toxic to the environment. To date, vanillin has been
tested for its anti-corrosive propensity for aluminum in acid
solutions.29 However, since vanillin itself failed in this effort for
ferrous metal, it was derivatized into a Schiff base, i.e., 2-
methoxy-4-((pyridin-2-ylmethylimino)methyl)phenol
aDepartment of Chemistry, National Institute of Technology, Durgapur, West Bengal,
713209, India. E-mail: dipankar.sukul@ch.nitdgp.ac.in; Tel: +91 9434788066
bSurface Engineering
& Tribology Group, CSIR-Central Mechanical Engineering
Research Institute, Durgapur 713209, India
cAcademy of Scientic and Innovative Research, CSIR-CMERI Campus, Durgapur
713209, India. E-mail: pr_banerjee@cmeri.res.in; Tel: +91-343-6452220
dFaculty of Engineering, Hokkaido University, Kita-13, Nishi-8, Kita-ku, Sapporo,
Hokkaido 060-8628, Japan
(compound A). Further, we synthesized divanillin, which was
subsequently converted to another Schiff base, 3,30-dimethoxy-
5,50-bis(((pyridin-2-ylmethyl)imino)methyl)-[1,10-biphenyl]-2,20-
diol (compound B). The molecular formulae of these two Schiff
bases are shown in Fig. 1. Compound B is essentially the
dimeric form of compound A. One of our main intentions of
† Electronic supplementary information (ESI) available: NMR, FTIR and mass
spectra of the synthesized inhibitors, potentiodynamic polarization plots and
impedance spectra in the presence of the inhibitors at different temperatures,
variation in the free energy of adsorption with temperature, Arrhenius plots,
electronic distribution in the frontier molecular orbitals for the protonated
forms of the inhibitors, closest distance between the inhibitor atoms, metal
surface data and various other relevant data. See DOI: 10.1039/c9ra07982c
9258 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
divanillin soluble in the medium. This was followed by the
dropwise addition of 1.082 g of 2-(aminomethyl)pyridine (10
mmol). The mixture was then allowed to reux for 18 h. A
greenish solid precipitate appeared aer cooling the reaction
mixture. The product was collected by ltration and washed
with methanol and then diethyl ether several times and nally
dried (yield 70%).
Compound A, divanillin and compound B were character-
ized via 1H NMR (Fig. S1–S3 in the ESI,†), FTIR (Fig. S4 in ESI†),
and ESI mass spectroscopy for compounds A and B (Fig. S5 and
S6 in ESI†) and EI mass spectroscopy for divanillin (Fig. S7 in
ESI†).
Fig. 1 Chemical structures of the vanillin-based Schiff bases.
1H NMR of compound A
this work is to compare the corrosion inhibition properties of
the dimeric and monomeric forms of the vanillin Schiff bases
and explain the phenomena from a theoretical viewpoint.
Electrochemical impedance spectroscopy, potentiodynamic
polarisation and the weight loss method were used to study the
anticorrosive behavior of these Schiff bases for mild steel in an
HCl medium. The corresponding corrosion inhibition efficiency
and thermodynamic and activation parameters involved in the
process in the presence and absence of inhibitors were calcu-
lated. Also, a detailed theoretical study using DFT and molec-
ular dynamics (MD) simulations was performed to determine
the quantum chemical parameters and to establish the experi-
mentally obtained results.
d ppm (DMSO-d6, 400 MHz): d ppm (DMSO-d6, 400 MHz):
3.734 ppm (s, 3H, H atoms of –OCH3 group), 4.737 ppm (s, 2H,
H atoms of methylene group attached to pyridine ring), 6.777 to
6.797 ppm (d, 1H, H atom of benzene ring nearest to –OH
group), 6.935 to 6.964 ppm (d, 1H, d-H atom of pyridine ring),
7.132 to 7.159 ppm (t, 1H, b-H of pyridine ring), 7.206 to 7.225
(d, 1H, 1H, H atom of benzene ring opposite to –OCH3 group),
7.333 ppm (s, 1H, H atom of benzene ring nearest to –OCH3
group), 7.707–7.731 ppm (t, 1H, g-H of pyridine ring), 8.313 to
8.322 ppm (d, 1H, a-H of pyridine ring), 8.460 ppm (s, 1H, H
atom attached to C atom of imine bond).
1H NMR of divanillin
d ppm (DMSO-d6, 400 MHz): 3.867 ppm (s, 6H, H atoms of
–OCH3 group), 7.367 ppm (s, 4H, H-atoms of aromatic rings),
9.775 ppm (s, 2H, H-atoms of aldehyde group).31,32
Experimental
Synthesis of compound A
0.761 g vanillin (5 mmol) was dissolved in 40 mL of methanol
followed by the dropwise addition of 0.541 g of 2-(aminomethyl)
pyridine (5 mmol). The mixture was then allowed to reux for
15 h. The pale yellow colour product obtained aer evaporation
of the solvent was puried by solvent extraction using chloro-
form and water. The resulting product aer the evaporation of
chloroform was washed with diethyl ether to remove any
unreacted vanillin. Finally, the product was dried and used for
further study (yield 70%).
1H NMR of compound B
d ppm (DMSO-d6, 400 MHz): 3.717 ppm (s, 6H, H atoms of
–OCH3 group), 4.764 ppm (s, 4H, H atoms of two methylene
group attached to two pyridine ring), 6.921 to 6.950 ppm (d, 2H,
d-H atom of pyridine ring), centered at 7.225 ppm (4H, H atom
of benzene ring of divanillin unit), 7.409 to 7.454 ppm (t, 2H, b-
H of pyridine ring), 7.796 to 7.845 (t, 2H, g-H of pyridine ring),
8.349 to 8.340 ppm (d, 2H, a-H of pyridine ring), 8.492 ppm (s,
2H, H atom attached to C atom of imine bond).
Synthesis of divanillin
Divanillin, the symmetrical dimer of vanillin, was synthesized
from vanillin using FeCl3.30 2.973 g (11 mmol) FeCl3$6H2O was
dissolved in 50 mL water in a round-bottom ask equipped with
a magnetic stirrer, oil bath and condenser. To this solution,
1.521 g of vanillin was added and suspended under constant
FTIR spectrum of divanillin
Divanillin consists of a benzene rings together with aldehyde
groups, phenolic –OH groups and ether (–OCH3) groups. The
broad absorption peak at around 3261 cmꢁ1 indicates the
presence of –OH groups (O–H stretching). The weak peak
around 3050 cmꢁ1 is due to the presence of aromatic C–H
stretching. Another weak band at around 2969 cmꢁ1 and
2942 cmꢁ1 is due to the C–H stretching of the –OCH3 group. The
strong and sharp peak at 1675 cmꢁ1 is due to the C–O stretching
vibration of the aldehyde group attached to aromatic rings.32,33
The sharp and strong peaks in the region of 1587 cmꢁ1 to
1509 cmꢁ1 are attributed to the presence of aromatic C]C. The
ꢀ
stirring. The mixture was heated to 50 C under constant stir-
ring for 6 h. The mixture was then cooled in an ice bath and the
resultant precipitate was ltered and washed with water and
methanol, and then nally dried to obtain an ash coloured
product (yield 55%).
Synthesis of compound B
1.511 g divanillin (5 mmol) was added to a mixture of 70 mL of peaks in the region of 1456 cmꢁ1 to 1400 cmꢁ1 are due to the in-
methanol and 10 mL DMF and stirred under heating to make plane C–H bending vibrations. The appearance of the peak at
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9259

View Article Online
RSC Advances
Paper
1353 cmꢁ1 is due to the C–C stretching. The peak at 1281 cmꢁ1 electrode (SCE) as the reference electrode (RE). Before the
corresponds to the C–O bending and the peaks at 1259 and measurements, the working electrode in the cell was kept in
1192 cmꢁ1 the C–O stretching vibrations.33 The other peaks are contact with 350 mL test solution for 35 min to achieve the steady
due to the different modes of the C–H, C–C, O–H, C–O bending state condition.
vibrations.33
Potentiodynamic polarization measurements were per-
formed at a scan rate of 0.5 mV sꢁ1 with in the potential range of
ꢃ250 mV from the respective rest potential values. Inhibition
efficiency, hP (%), is dened as:
FTIR spectra of compounds A and B
In the vanillin moiety of compound A and compound B, the
aldehyde group is converted into an imine group and a new
pyridine ring is introduced. Thus, the peak for the aldehyde is
absent and new peaks appear at 1639 cmꢁ1 for compound A and
1636 cmꢁ1 for compound B due to the C]N (imine bond)
stretching. The broad peak centered at ꢂ3415 cmꢁ1 indicates
the presence of –OH groups (O–H stretching). The appearance
of peaks in the region of 1595 cmꢁ1 to 1435 cmꢁ1 is attributed to
the presence of both benzene and pyridine aromatic rings. The
other peaks are due to the different modes of the C–H, C–C, O–
H, and C–O stretching and bending vibrations.
icorr ꢁ i
hPð%Þ ¼
corrðinhÞ ꢄ 100
(1)
icorr
where, icorr and icorr(inh) are the corrosion current density of the
uninhibited and inhibited specimens, respectively, which were
estimated using the Tafel extrapolation method.
EIS measurement was performed within the frequency range
of 0.1 MHz to 10 MHz with an AC amplitude of ꢃ10 mV (rms) at
the rest potential. The inhibition efficiency of the inhibitor is
dened as follows:
Rp ꢁ R0p
Rp
hZð%Þ ¼
ꢄ 100
(2)
ESI-mass spectra of compounds A and B
where Rp and R0p represent the impedance value in the presence
and absence of the inhibitor, respectively. Rp accounts for
a combined effect of different resistances operating at the
metal-electrolyte interface, such as the charge transfer resis-
tance and resistance due to adsorbed corrosion products as well
as inhibitors.
The appearance of a prominent molecular ion peak at m/z 243.2
[L + H+] conrms the formation of compound A. Due to the very
low solubility of compound B in common organic solvent, the
mass spectrum of compound B is not very clear and contains
noise. However, the presence of a molecular ion peak at m/z at
483.1264 (ꢂ50%) conrms the formation of compound B. The
presence of peaks at m/z 451.127 (L-CH3OH + H), 405.1625 (L-py
unit + H), 328.1636 (L-2py unit + H), 314.1520 (L-py unit ꢁ py-
CH2 unit + H), 511.1565 (2L + 3H2O + 2K), and 600.2269 (L +
DMSO + K) also conrms the formation of compound B.
Weight loss measurements
The effect of temperature on the corrosion inhibition efficiency,
rate of corrosion of the inhibited and uninhibited specimens,
different thermodynamics and activation parameters related
with the corrosion inhibition process were evaluated using the
following method. Mild steel coupons (wt% composition: 0.19 C,
0.21 Si, 0.21 Mn, 0.01 P, 0.01 S and the remainder iron) with the
dimensions of 2.5 cm ꢄ 2.5 cm ꢄ 0.1 cm were treated properly,
as previously described (in specimen and solution section). Aer
taking the initial weight, the coupons were immersed in 350 mL
of aqueous HCl (1 M) in the absence and presence of inhibitors.
Aer 6 h of exposure, the coupons were taken out, and washed
under running water using a bristled brush. Coupons were then
washed with distilled water and acetone and dried in a vacuum
desiccator before taking their nal weight. During the experi-
ment with prolonged exposure, the volume of the solution was
maintained by adding the requisite amount of distilled water
when required. The percentage inhibition efficiency, hW (%), was
calculated using following the relation:
EI-mass spectrum of divanillin
The appearance of a very strong and sharp peak at 302 in the EI
mass spectrum conrms the formation of divanillin.
Specimen and solution
Cylindrical-shaped sample specimens were cut from commercially
available mild steel rods with a diameter of 1 inch having the
following composition: 0.22 C, 0.31 Si, 0.60 Mn, 0.04 P, 0.06 S and
the remainder iron (wt% composition). A shiny metal surface was
prepared by rubbing the cut sample pieces with different grade
silicon carbide paper (from 60 to 1200), and thereaer washing
thoroughly with soap, tap water, distilled water and nally with
acetone, and drying in a vacuum desiccator. All the reagents used
were AR grade. Vanillin was obtained from Merck India, 2-pico-
lylamine from Sigma India and HCl (1 M) solution was prepared
using 37% HCl from Merck India.
W0 ꢁ W
hWð%Þ ¼
ꢄ 100
(3)
W0
Electrochemical measurements
where W0 and W are the weight loss of the metal samples in acid
All electrochemical experiments, i.e. electrochemical impedance medium in the presence or absence of the inhibitor.
spectroscopy (EIS) and Tafel extrapolation method were per-
Since organic inhibitors are mostly adsorption-type inhibitors,
formed using a conventional 3-electrode system (model: Gill AC, a direct correlation between the degree of surface coverage (q) with
ACM Instruments, UK). A mild steel sample with an exposed inhibition efficiency over the whole concentration range is
surface area of 1 cm2 functioned as the working electrode, a Pt generally assumed. Following this assumption, q was calculated
mesh electrode as the auxiliary electrode and saturated calomel using the hW (%) values according to the following equation:
9260 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
with a higher value until their chemical potentials become the
same. To evaluate the value of DN, the electronegativity of Fe
(cFe) was replaced by the work function of the Fe (1 1 0) surface
(most stable densely packed surface among the Fe surfaces), fFe
¼ 4.82 eV (ref. 38 and 39) and the global hardness of iron,
hFe, was taken as zero considering I ¼ A for the metallic bulk.
q ¼ hW (%)/100
(4)
Surface morphology
The morphological study was accomplished by keeping the
mild steel samples in 1 M HCl medium in the presence and
absence of inhibitors for 6 h, and thereaer the sample spec-
imen was gently washed with acetone and distilled water and
dried. The surface morphology was observed using scanning
electron microscopy (Hitachi, S-3000N).
Local reactivity analysis (Fukui indices)
The local reactivity of the molecules was analysed through
evaluation of the Fukui indices using the Dmol3 module in
Material Studio™ version 6.1 by Accelrys Inc, San Diego, CA. All
calculations were performed by applying the B3LYP exchange–
correlation function and the double numeric with polarization
(DNP) basis set. Here, the Fukui functions were obtained
through the nite difference approximation using Hirscheld
population analysis (HPA).40–42 By applying the concept HSAB
principle43 in a local manner, it can be interpreted that the
regions of a molecule where the Fukui function is large are
chemically soer than the regions where the Fukui function is
small. The Fukui function at a point, r, in the space around the
molecule is dened as the rst order derivative of electron
density at that point with respect to the number of electrons, N,
present in the molecule at a constant external potential, v.40–42
Computational details for quantum chemical calculation
The quantum chemical parameters for the inhibitor molecules
were calculated via the density functional theory (DFT) using
the ORCA program (version 2.7.0). Geometrical optimisation of
the inhibitors was performed at the B3LYP34 functional level
using the triple-z quality basis set TZV (P) with one set of
polarization functions on the N, O and S atoms. For atoms such
carbon and hydrogen, we used slightly smaller polarized split-
valence SV (P) basis sets, which are double-z quality in the
valence region and have a polarizing set of d functions on atoms
other than hydrogen.1,34 SCF calculations were converged tightly
(1 ꢄ 10ꢁ8 Eh in energy, 1 ꢄ 10ꢁ7 Eh in density and 1 ꢄ 10ꢁ7 in
maximum element of the DIIS error vector). All theoretical
calculations were performed in the aqueous phase because
electrochemical corrosion occurs in the aqueous phase
considering the solvent as a continuum of uniform dielectric
constant (3), where the solute is placed as a uniform series of
interlocking atomic spheres.35 Various intrinsic molecular
parameters such as electron affinity (A), ionization potential (I),
electronegativity (c, indicates the ability of a group of atoms to
attract electrons towards itself), global hardness (h, measure of
the resistance of an atom towards charge transfer) and global
soness (s, susceptibility of inhibitor molecules towards charge
transfer) were calculated using the following equations36,37
ꢀ
ꢁ
vrðrÞ
vN
f ðrÞ ¼
(11)
v
Since the electron density is discontinuous with respect to
the number of electrons, le hand and right hand side deriva-
tives were introduced as:
ꢀ
ꢀ
ꢁþ
vrðrÞ
vN
f þðrÞ ¼
(12)
v
ꢁꢁ
vrðrÞ
vN
f ꢁðrÞ ¼
(13)
v
where f+ can be used to probe the reactivity when electrons are
added to the system (attack of a nucleophile) and fꢁ to probe the
reactivity when electrons are extracted from the system (attack
of an electrophile). Taking the nite different approximation,
a condensed form of these functions is proposed as:44
c ¼ (I + A)/2
h ¼ (I ꢁ A)/2
(5)
(6)
(7)
(8)
(9)
s ¼ 1/h ¼ 2/(I ꢁ A)
I ¼ ꢁEHOMO
f+k z qk(N + 1) ꢁ qk(N) (for nucleophilic attack)
fꢁk z qk(N) ꢁ qk(N ꢁ 1) (for electrophilic attack)
(14)
(15)
A ¼ ꢁELUMO
where qk(N + 1), qk(N) and qk(N ꢁ 1) are the gross atomic pop-
ulation (i.e. gross atomic charge) of the atom k in the N + 1, N
and N ꢁ 1 electron systems, respectively.
The fraction of electrons transferred from the inhibitor
molecule to the metallic atom (DN) was calculated using the
following relation:36
ðcFe ꢁ cinh
Þ
Molecular dynamics simulation
DN ¼
(10)
2ðhFe þ hinh
Þ
To calculate the interaction energies between the Fe (1 1 0)
surface and our inhibitor molecules, we used the molecular
The concept of DN is based on the assumption that between dynamics (MD) simulation technique. Accordingly, we used the
two interacting systems of different electronegativities (here, Material Studio™ soware 6.1 (from Accelrys Inc.). In this
the metallic surface and an inhibitor molecule), the electron simulation process, the interaction between the studied mole-
will ow from the molecule of lower electronegativity to that cules and iron surface was carried out in a simulation box of
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9261

View Article Online
RSC Advances
Paper
ꢀ
ꢀ
ꢀ
(40.11 A ꢄ 40.11 A ꢄ 78.00 A) with periodic boundary condi-
tions to avoid any arbitrary boundary effect. Here, we used ten
layers of iron atoms to provide sufficient depth to overcome the
issue related to the cut-off radius. In this investigation, a three-
layered simulation box was created. The rst layer contained an
Fe slab and the second layer was the solution slab, which con-
tained H2O (150), together with H3O+ and Clꢁ ions (15 each) as
well as the inhibitor molecule, and the remaining part of the
box was the vacuum layer. The presence of H3O+ and Clꢁ ions
makes the MD simulation closer to the real system. Aer
construction of the simulation box, MD simulation was carried
out using the COMPASS (Condensed Phase-optimized Molec-
ular Potentials for Atomistic Simulation Studies) ab initio force
eld. In general, the parameterization procedure was divided
into two phases: (i) ab initio parameterization and (ii) empirical
optimization. The MD simulation was performed at 298.0 K
using a canonical ensemble (NVT) with a time step of 1.0 fs and
for a simulation time of 100 ps. All the bulk atoms in the Fe (1 1
0) surface were kept frozen and all the concerned molecules
were allowed to interact with the metal surface freely during the
entire simulation process. The interaction energy (Einteraction
)
between the inhibitor molecule and the Fe (1 1 0) surface was
calculated using the following equation:1,9,10
+
ꢁ
Einteraction ¼ Etotal ꢁ (Esurface+H O+H O +Cl + Einhibitor
)
(16)
2
3
where the total energy of the simulation system is dened as
Etotal, the energy of the iron surface together with H2O, H3O+,
Clꢁ is classied as Esurface+
and that of the free
H2O+H3O++Clꢁ
inhibitor molecule as Einhibitor
.
Results and discussion
Fig. 2 Potentiodynamic polarization plots for mild steel in 1 M HCl in
the presence of compound A (top) and compound B (bottom) at 303 K.
Polarization measurements
The potentiodynamic polarization measurement was per-
formed with a range of concentrations for both inhibitor
molecules and at two different temperatures of 303 and 323 K.
The polarization plots for mild steel in 1 M HCl at 303 K in the
presence of compounds A and B are shown in Fig. 2. Various
corrosion parameters such as the corrosion potential (Ecorr),
corrosion current density (icorr) and anodic and cathodic Tafel
slopes (ba and bc, respectively) were derived following the Tafel
extrapolation method (representative examples shown in Fig. S8
in the ESI†), and tabulated in Table 1. The polarization plots
and corrosion parameters correspond to the typical mixed-type
corrosion inhibitor behavior for both compounds.45 Both the
cathodic and anodic currents diminished systematically with an
increase in the concentration of the inhibitors, without much
alteration in the cathodic and anodic slopes. The corrosion
potentials also remained within a narrow range of ꢃ10 mV.
Mixed-type inhibition by any inhibitor indicates the retar-
dation of both the cathodic hydrogen evolution reaction and
anodic metal oxidation reaction at the cathodic and anodic
reaction sites, respectively. When the inhibitor is capable of
accepting electrons from the metal surface, a deciency in
electrons results in the inhibition of the rate of reduction of
hydronium ions producing hydrogen gas. Similarly, a reduced
rate of metal oxidation reaction is indicative of the fact that the
inhibitors are prone to donate electrons at the anodic sites of
the metal surface, thereby increasing the electron density at the
metal-electrolyte interface. Thus, it can be concluded that both
inhibitors are capable of donating electrons at the anodic
reaction sites and accepting electrons at the cathodic reaction
sites.
Due to the extremely low solubility of divanillin in 1 M HCl,
its electrochemical study could not be carried out. The other two
reactants, i.e., vanillin and 2-(aminomethyl)pyridine did not
exhibit corrosion protection to any signicant extent (inhibition
efficiency being 60–70%, Fig. S9 and Table S1 in ESI†). Thus, the
enhancement of corrosion protection of mild steel in aqueous
HCl was evident when vanillin and 2-(aminomethyl)pyridine
reacted and produced the Schiff base, compound A. Accord-
ingly, the resultant imine group present in compound A must
play a role in providing this signicantly greater effect. The
other possible molecular parameters will be considered
following the quantum molecular result.
9262 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 1 Potentiodynamic polarisation data for mild steel in 1 M HCl in the presence and absence of compounds A and B at 303 K
Conc. (mM)
Uninhibited
ꢁEcorr (mV/SCE)
ba (mV decꢁ1
)
ꢁbc (mV decꢁ1
)
icorr (mA cmꢁ2
)
hP, %
521
121.4
158.7
3254
—
With compound A
0.001
0.010
0.10
0.25
0.50
1.0
514
518
518
513
529
517
532
103.8
104.2
103.4
106.3
102.9
101.1
100.6
129.5
130.7
129.5
131.2
132.6
128.2
126.3
2257
1681
1140
709
675
479
22.3
48.3
64.9
78.2
79.3
85.3
92.0
2.0
260
With compound B
0.001
0.010
0.10
0.25
0.50
1.0
518
529
520
523
520
520
102.9
102.3
97.1
100.3
99.2
121.6
122.5
119.9
119.8
120.1
121.4
1435
545
296
280
249
55.9
83.2
90.9
91.4
92.3
94.0
99.5
194.8
Further, at the elevated temperature of 323 K, the rate of
corrosion was higher in the presence of both compounds A and
B, which is common for any activation energy-controlled reac-
tion. However, the inhibition efficiencies at this higher
temperature for both compounds showed almost the same
value as that at 303 K (Fig. S10 and Table S2 in ESI†). Thus, these
compounds can be applied as effective mixed-type corrosion
inhibitors, at least up to 323 K.
Electrochemical impedance measurements
Fig. 3 depicts the Nyquist plots of mild steel in 1 M HCl with
different concentrations of compounds A and B together with
the uninhibited specimen at 303 K. The corresponding
impedance and phase angle plots (Bode plots) are shown in ESI
(Fig. S11 and S12†). These plots clearly reveal that an increase in
inhibitor concentration (both compounds A and B) leads to
a concomitant enhancement in the diameter of the capacitance
loops. Consequently, this indicates an increase in polarization
resistance (Rp). Rp is a collective parameter comprising of
various resistive factors operating at the metal-electrolyte
interface. These factors include charge transfer resistance (Rct)
for both metal ion oxidation and hydrogen ion reduction,
resistance of the adsorbed organic inhibitor layer and resis-
tance of the adsorbed corrosion products. All these parameters
are interrelated and cannot be factorized individually.46,47
Overall, with a gradual increase in inhibitor concentration, the
degree of surface coverage and thickness of the adsorbed
inhibitor layer increase, replacing the pre-adsorbed water
molecules and ions present in the electrolyte solution, and thus
increasing the Rp value.46–48 The non-perfect semi-circle nature
of the Nyquist plots can be explained in terms of surface
heterogeneity due to the formation of an adsorbed layer of the
inhibitors together with the corrosion product.48 The Nyquist
plots in the absence and presence of inhibitors reveal the exis-
tence of one capacitive loop, and the absence of an inductive
loop. The corresponding Bode phase angle plots contain one
Fig. 3 Nyquist plots for mild steel in 1 M HCl in the presence of
compound A (top) and compound B (bottom) at 303 K.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9263

View Article Online
RSC Advances
Paper
compound B. The concept of electrochemical double-layer
capacitance was rst introduced by Helmholtz, and then elab-
orated by Stern and Geary. Different parameters such as the
electrolyte dielectric constant (3), surface area accessible to ions
(A) and distance between the ion and the charged electrode
surface (d) determine the capacitance, Cdl, according to the
following equation:
Fig. 4 Equivalent circuit model used to fit the impedance spectra.
maximum without any hump, while the Bode impedance plots
show only one negative uctuation. All these phenomena
correspond to the involvement of only one time constant.
Accordingly, the observed Nyquist plots were tted with an
equivalent circuit model, as shown in Fig. 4, where Rp stands for
the polarization resistance and Rs stands for the solution
resistance. The constant phase element (CPE) was used in the
equivalent circuit instead of the double-layer capacity (Cdl) to
accommodate the observed depressed semi-circle nature in the
Nyquist plots.49 Accordingly, the chi-square value in the order of
10ꢁ4 conrms the validity of the model and goodness of the t.
The tted parameters together with the inhibition efficiency is
presented in Table 2.
Cdl ¼ 303A/d
(19)
where, 30 is the vacuum permittivity.52,53 An increase in inhibitor
concentration leads to a higher degree of surface coverage, and
thereby reduces the effective value of A. Further, organic
inhibitor molecules are adsorbed on the metal surface, replac-
ing pre-adsorbed ions and water molecules. This reduces the
dielectric constant of the solution adjacent to the metal elec-
trode. Since the pre-adsorbed ions forming the double layer
diffuse away from the metal surface, the effective thickness of
the double layer increases.54,55 All these factors act cooperatively
in reducing the electrochemical double layer capacitance per
unit surface area in the presence of an organic corrosion
inhibitor. The various corrosion rate-reducing effects were more
dominant for compound B, making it the better inhibitor. This
is also evident from the higher Rp and lower Cdl values for
compound B compared to that at the same concentration for
compound A.
The EIS data for vanillin and 2-picolyl amine is presented in
the ESI (Fig. S13 and Table S3 in ESI†), which again conrms
that compounds A and B have much better corrosion resistivity
for mild steel than their precursors. With an increase in
temperature, the Rp value of the both uninhibited and inhibited
systems decreased. However, this did not affect their inhibition
efficiencies. Thus, despite the lower degree of protection
offered, these vanillin derivatives stand out as good corrosion
inhibitors with high inhibition efficiencies at elevated temper-
ature (Fig. S14 and Table S4 in ESI†).
The relation between CPE and impedance is given as
ZCPE ¼ Qꢁ1(iu)ꢁn
(17)
where Q is a proportionality coefficient, u is the angular
frequency, n is a measure of surface irregularity. CPE is con-
verted into the classical lumped elements capacitor (C), resis-
tance (R), and inductance (L) when n ¼ 1, 0, and ꢁ1,
respectively.
The relation between polarization resistance (Rp) and double
layer capacitance (Cdl), is given by the relation.50,51
1ꢁn 1/n
Cdl ¼ (QRp
)
(18)
Table 2 demonstrates that the capacitance (Cdl) values
decreased gradually with an increase in inhibitor concentra-
tion. The rate of the decrease in capacitance was greater for
Table 2 EIS data for mild steel in 1 M HCl in the presence and absence of compounds A and B at 303 K
Conc. (mM)
Uninhibited
Rp (U cm2)
Q (mUꢁ1 sn cmꢁ2
)
n
Cdl (mF cmꢁ2
)
hz, %
c2 ꢄ 104
3.8
1010
0.81
273.9
4.21
With compound A
0.001
0.010
0.10
0.25
0.50
1.0
4.8
7.8
738
553
353
276
229
190
147
0.82
0.84
0.85
0.85
0.85
0.85
0.86
213.8
196.0
131.7
104.5
90.45
76.1
20.8
51.3
64.2
74.3
83.2
87.2
94.8
1.62
2.82
1.65
0.7
2.97
3.36
1.22
10.6
14.8
22.6
29.6
72.4
2.0
70.2
With compound B
0.001
0.010
0.10
0.25
0.50
1.0
7.1
25.6
53.5
63.0
71.2
80.9
400
207
169
146
127
125
0.82
0.84
0.85
0.85
0.86
0.86
110.4
76.3
73.6
63.8
59.0
59.2
46.5
85.2
92.9
93.9
94.7
95.3
0.6
1.37
3.43
2.62
3.05
2.17
9264 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 3 Corrosion rate of mild steel in 1 M HCl in the presence and absence of compounds A and B at different temperatures (exposure time 6 h)
Corrosion rate (CR, mg
cmꢁ2 hꢁ1
)
hW
%
q
Temp. (K)
293
Conc. (mM)
Comp. A
Comp. B
Comp. A
Comp. B
Comp. A
Comp. B
0
2.011
1.535
0.980
0.679
0.469
0.385
0.294
0.197
3.207
2.428
1.496
1.036
0.727
0.537
0.386
0.225
5.437
4.035
2.491
1.714
1.189
0.819
0.604
0.324
9.012
7.202
4.511
3.170
2.385
1.846
1.478
1.071
2.011
1.073
0.457
0.236
0.205
0.167
0.135
—
3.207
1.656
0.535
0.264
0.221
0.187
0.140
—
5.437
2.715
0.785
0.341
0.235
0.193
0.153
—
9.012
4.519
1.471
0.736
0.603
0.513
0.355
—
0.001
0.01
0.1
0.25
0.5
1
23.7
51.3
66.2
76.7
80.8
85.4
90.2
46.3
77.3
88.3
89.8
91.7
93.3
—
0.237
0.513
0.662
0.767
0.808
0.854
0.902
0.463
0.773
0.883
0.898
0.917
0.933
—
2
0
303
313
323
0.001
0.01
0.1
0.25
0.5
1
24.3
53.4
67.7
77.3
83.3
88.0
93.0
48.6
83.2
91.8
93.1
94.2
95.6
—
0.243
0.534
0.677
0.773
0.833
0.880
0.930
0.486
0.832
0.918
0.931
0.942
0.956
—
2
0
0.001
0.01
0.1
0.25
0.5
1
25.8
54.2
68.5
78.1
84.9
88.9
94.1
50.0
85.4
93.7
95.6
96.4
97.2
—
0.258
0.542
0.685
0.781
0.849
0.889
0.941
0.5
0.854
0.937
0.956
0.964
0.972
—
2
0
0.001
0.01
0.1
0.25
0.5
1
20.1
49.9
64.8
73.5
79.5
83.6
88.1
49.9
83.7
91.8
93.31
94.31
96.0
—
0.201
0.499
0.648
0.735
0.795
0.836
0.881
0.499
0.837
0.918
0.933
0.943
0.960
—
2
a high degree of effectiveness even in an elevated temperature
range and high exposure time. This indicates the strong
adsorptive interaction between the inhibitor molecules and
the metal surface, leading to a high value of fraction of surface
covered (q, Table 3).
Weight loss measurement
The results obtained from the electrochemical methods were
veried using the weight loss method. We tested our inhibi-
tors at four different temperatures within the range of 293 K to
323 K with a variable inhibitor concentration aer 6 h of
exposure of the test specimens in HCl solution. The impor-
tance of the weight loss method to determine the rate of
corrosion comes from the fact it encompasses different types
of corrosion, including overall corrosion, galvanic corrosion
and all possible forms of localized corrosion. The results
shown in Table 3 support the conclusion derived from the
electrochemical methods. Compound B exhibited better
corrosion resistivity than compound A. The maximum
inhibitory effect was observed at 313 K, which decreased.
Regarding the effect of exposure time on the inhibition effi-
ciency, compound B demonstrated more than 90% inhibition
efficiency with a concentration of 1 mM, even aer the mild
steel samples were exposed to 1 M HCl for 72 h (Table S5 in
the ESI†). Thus, according to all these observations, it can be
concluded that the synthesized inhibitors are useful in
retarding the rate of corrosion of mild steel in 1 M HCl with
Adsorption isotherm
The Langmuir adsorption isotherm model was found to be best
suited to describe the adsorption behaviour of the vanillin
derivatives on the surface of mild steel in 1 M HCl (Fig. 5). The
Langmuir adsorption isotherm, which relies on the monolayer
formation hypothesis by energetically equivalent adsorption
sites, is represented by the following equation:
C/q ¼ 1/Kads + C
(20)
where C is the concentration of inhibitors used and q the frac-
tion of surface covered at the particular inhibitor concentration
(as tabulated in Table 3). The free energy change in the
adsorption (DG0ads) is related to the adsorption equilibrium
constant (Kads) according to the following equation:
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9265

View Article Online
RSC Advances
Paper
Table 5 DH0ads and DSa0ds data for compounds A and B
DHa0ds (kJ molꢁ1 DS0ads (J Kꢁ1 molꢁ1
)
)
Temp. (K)
Comp. A
ꢁ0.656
Comp. B
Comp. A
Comp. B
293–313
313–323
24.7
ꢁ34.6
116
218
29
intercept of the plot of C/q vs. C gives the Kads and DG0ads values
according to eqn (21). These values are tabulated in Table 4. The
other adsorption parameters such as the enthalpy of adsorption
(DH0ads) and free energy of adsorption (DSa0ds) were evaluated
according to eqn (22) and tabulated in Table 5.
DG0ads ¼ DH0ads ꢁ TDSa0ds
(22)
To construct the adsorption isotherms, the initial concen-
tration of inhibitor was used rather than its equilibrium value.
Hence, more importance should be given towards the relative
variation in the adsorption constants and other adsorption
parameters between the two inhibitors rather than their
absolute values. The equilibrium adsorption constant was
maximum at 313 K for both inhibitors and decreased there-
aer. This indicates that the extent of adsorption is greater
when induced by some degree of thermal energy. Further, the
Kads values for compound B are higher than that for compound
A by many times and they are pronounced mostly at 313 K. For
compound A, DG0ads became gradually more negative with
temperature, and the variation is linear in nature (Fig. S15 in
ESI†). The corresponding heat and entropy of adsorption data
is shown in Table 5, which reveals the exothermic nature of
adsorption with a higher degree of randomness during
adsorption. In contrast, for compound B, the dependency of
DG0ads on temperature is not linear over the whole temperature
range. A break in the slope was observed between the lower
and higher temperature ranges (Fig. S15 in ESI†). Accordingly,
two sets of data for DHa0ds and DSa0ds were obtained in the lower
and higher temperature ranges (Table 5). Inspection of Tables
4 and 5 reveals some important facts regarding the adsorption
of the vanillin derivatives on the surface of mild steel in 1 M
HCl. The DG0ads values for compound B are more negative than
that of compound A, suggesting the stronger adsorption of
compound B. Further, the DG0ads values for compound B
Fig. 5 Langmuir adsorption isotherms involving compounds A (top)
and B (bottom).
DG0ads ¼ ꢁRT ln(55.56Kads
)
(21)
where R is the universal gas constant, T is the experimental
temperature and the factor 55.56 is the molarity of water
(introduced as adsorption occurs in aqueous solution).56 For
a broad range of degree of surface coverage, the nearly unity
slope with a correlation coefficient (r2) value in the order 0.999
for all the plots of C/q against C justies the suitability of the
Langmuir model in the present case. The reciprocal of the
Table 4 Adsorption parameters in the presence of compounds A and B at different temperatures
Slope
Kads ꢄ 10ꢁ3 (molꢁ1
)
DG0ads (kJ molꢁ1
Comp. A
)
Comp.
A
Temp. (K)
Comp. B
Comp. A
Comp. B
Comp. B
293
303
313
323
1.104
1.070
1.058
1.129
1.070
1.045
1.028
1.049
28.65
28.09
28.98
27.62
186.92
253.16
355.87
235.85
34.79
35.92
37.19
38.25
39.35
41.46
43.72
44.01
9266 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
decreased gradually with an increase in temperature, but rate
of decrement was lower at a higher range of temperature
(Fig. S15 in ESI†). This observed trend is characteristic of
chemisorption, which is activated adsorption, and thus
requires a minimum amount of energy to proceed. The varia-
tion in the adsorption constant with temperature also
supports this view. Thus, up to a certain temperature range,
adsorption remains essentially kinetically controlled. At
sufficiently high temperature, adsorption becomes thermo-
dynamically controlled and decreases with an increase in
temperature.21,57–60 DSa0ds was positive for both inhibitors over
the whole temperature range. This seems contrary to any
adsorption process, which should result in a decrease in
randomness. Thus, the results suggest that the removal of pre-
adsorbed water molecules and other ions present in the elec-
trolytic solution from the metal surface occurs simultaneously
with the adsorption of the inhibitor molecules.58 The desorp-
tion of solvent molecules needs to be considered while
explaining the variation in DH0ads. At a lower temperature
range, it was positive for compound B, suggesting endo-
thermic adsorption, whereas it became negative at a higher
temperature. This observation can be interpreted similarly as
Activation parameters
To determine the activation parameters related to the corrosion
process, we plotted the corrosion rate data against temperature
using the following equations:21,59
E*
log CR ¼ log l ꢁ
(23)
(24)
2:303RT
ꢀ
ꢁ
ꢀ
ꢁ
RT
DS*
ꢁDH*
CR ¼
exp
exp
NAh
R
RT
where l is the pre-exponential factor (Arrhenius frequency
factor), E* is the activation energy related to the corrosion
process, R is the universal gas constant, h is Plank's constant, NA
is Avogadro's number, and T is the temperature. DS* and DH*
are the entropy of the activation and enthalpy of activation,
respectively. The plot of log(CR) vs. 1/T (Fig. 6) provides the
value of l and E* from the intercept and slope, respectively. On
the other hand, DS* and DH* are calculated from the plot of
log(CR/T) vs. 1/T (Fig. S16 in ESI†). From the slope of the plot,
DH* can be calculated, and the intercept gives the value of DS*.
The experimentally obtained data is presented in the Table 6.
The activation parameters calculated from the corrosion rate
data using the weight loss method give valuable information
regarding the probable mechanism of adsorption of the inhib-
itor on the metal surface. It was observed that the activation
energy in the presence of compound A was slightly higher than
that for DS0ads
.
The literature suggests that the desorption energy (which is
positive) of water molecules from the iron oxide surface varies
from 50 to 100 kJ molꢁ1, depending on the temperature and
limit of surface coverage.61 For compound A, the energy
released for adsorption exceeds the energy required for the
desorption of pre-adsorbed water molecules for the whole
temperature range, making it an exothermic process.
However, for the other inhibitor, at a lower temperature range,
the opposite effect is more dominant, making the whole
process endothermic. At an elevated temperature, the effect of
adsorption becomes more dominant and exothermic adsorp-
tion behaviour is observed. According to the literature,
chemisorption is accompanied with a heat of adsorption in the
order of ꢁ100 kJ molꢁ1
. Since the heat of desorption of pre-
62
adsorbed water molecules also contributed during the evalu-
ation of the heat of adsorption of the inhibitor molecules, the
overall DH0ads value calculated was much lower than that for
the chemisorption process. However, the trend in the variation
of the different adsorption parameters, as discussed above,
conforms with the conclusion that the adsorption of the
vanillin derivatives on the surface of mild steel in 1 M HCl is
essentially chemisorption, with compound B interacting with
the metal surface more strongly than compound A. Overall, the
adsorption of compound A is both enthalpically and entropi-
cally favourable. For compound B, adsorption at a lower
temperature is entropy driven, whereas at elevated tempera-
ture it is mostly enthalpy driven.
The change in the nature of adsorption for compound B
going from lower to higher temperature (endothermic to
exothermic behaviour) was observed for other systems also.21,59
Nevertheless, it still requires further experimentation at even
higher temperatures to establish the phenomena
unequivocally.
Fig. 6 Arrhenius plots for mild steel in 1 M HCl with compounds A (top)
and B (bottom).
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9267

View Article Online
RSC Advances
Paper
Table 6 Activation parameters in the presence of compounds A and B
l ꢄ 10ꢁ4 (mg cmꢁ2 hꢁ1
)
E* (kJ molꢁ1
Comp. A
)
DH* (kJ molꢁ1
Comp. A
)
DS* (J Kꢁ1 molꢁ1
)
Conc. (mM)
Comp. A
Comp. B
Comp. B
Comp. B
Comp. A
Comp. B
Uninhibited
0.001
0.01
0.10
0.25
0.50
1.00
2.0
1212.83
1318.56
697.11
542.13
788.32
284.90
366.61
348.34
1212.83
323.07
7.23
38.09
38.98
38.55
38.84
40.68
38.80
40.17
41.24
38.09
36.40
29.46
27.83
25.00
25.93
22.76
—
36.97
37.85
37.38
37.67
39.54
37.45
38.81
39.66
36.97
35.21
27.79
25.97
22.87
23.82
20.60
—
ꢁ113.1
ꢁ112.4
ꢁ117.9
ꢁ119.9
ꢁ116.8
ꢁ125.9
ꢁ123.8
ꢁ125.0
ꢁ113.1
ꢁ124.4
ꢁ157.5
ꢁ169.4
ꢁ181.4
ꢁ179.8
ꢁ192.6
—
1.86
0.49
0.59
0.13
—
that of the uninhibited sample, while the presence of considered to be the rate determining step.65 Following the
compound B resulted in a decrease in the activation energy cathodic Tafel slope tabulated in Table 1, we may presume that
compared to that of the uninhibited specimen to a signicant in the present case, the reaction shown in eqn (25) is the rate-
extent. The Arrhenius equation suggests that the variation in determining step. In fact, all these intermediate steps lead to
the reaction rate with temperature will be more with a higher a decrease in the randomness in their corresponding transition
activation energy. This explains the observation regarding the states, which is reected in the negative value for DS*. Again,
relative variation in the reaction rate with temperature in the the more negative value of DS* in the presence of the inhibitor
absence and presence of the inhibitors. For the uninhibited compared to that of the uninhibited specimen can be explained
sample and in the presence of compound A, the rate of corro- in terms of the blocking of the cathodic and anodic reaction
sion increased to a greater extent with an increase in tempera- sites by the inhibitor molecules. For occurrence of any type of
ture than that with compound B. Furthermore, it is argued that surface reaction, reactants require closer approach or interac-
when the values of E* and DH* of the inhibited system is tion with a metal surface possessing an inhibitor layer, resulting
comparable or lower than that of the uninhibited system, the in more order during the intermediate stages of the reaction.
adsorption may be predicted to be chemisorption, whereas This also results in a decrease in the pre-exponential frequency
a higher value of E* and DH* of the inhibited system compared factor.1
to that of the uninhibited system leads to physisorption.21,59,63
The Arrhenius plots show a deviation from linearity in the
Inspection of the table leads to the conclusion that chemi- case of higher inhibitor concentrations. This non-Arrhenius
sorption is preferred over physisorption. The rate of the behavior is observed for complex processes when the mecha-
decrease in corrosion with an increase in inhibitor concentra- nism of the reaction, and hence the activation parameters,
tion can be explained by the decreased value of the pre- depends on temperature.66 In case of the inhibitor-adsorbed
exponential frequency factor. Again, according to Table 6, DS* metal surface, the extent of adsorption depends on the
is negative for the uninhibited system, and it is even more temperature. This will inuence the mechanism of the corro-
negative in the presence of the inhibitors. This means that the sion process to a certain extent and will be the guiding factor for
transition state in the corrosion process, which is the rate- the non-linearity observed in Fig. 6.
determining step, is more ordered than the reactants, and the
degree of order of the transition state is higher when an
inhibitor layer is present on the metal surface. This can be
explained considering the detailed mechanism of the hydrogen
evolution reaction (eqn (25)–(27)) together with the adsorption
of the inhibitor on the metal surface.64,65
Comparison of the inhibitory action of the inhibitors
Since compound B is essentially the dimeric form of compound
A, one should expect a similar corrosion inhibitory effect of
compound B at half of the concentration of compound A.
However, the analysis of the corrosion current density (icorr
)
M + H3O+ + eꢁ / MHads + H2O
2MHads 4 2M + H2
(25)
(26)
(27)
values reveals that the icorr at 0.5 mM concentration of
compound A is 675 mA cmꢁ2, while that at 0.25 mM concen-
tration of compound B is 280 mA cmꢁ2 (Table 1). This is much
lower than the half of the other value. Similarly, the Rp value in
the presence of 0.5 mM compound A is 22.6 ohm cm2, whereas
for 0.25 mM compound B, it is 63 ohm cm2 (Table 2). Thus, the
inhibitory effect for compound B is more than twice that of
compound A, and this trend is valid throughout the lower
concentration range. A similar conclusion can be derived from
the weight loss measurements (Table 3). When we compared
the icorr and Rp values for 2 mM compound A and 1 mM
compound B, the inhibitory effect was still higher for
MHads + H3O+ + eꢁ 4 M + H2 + H2O
Any one of these intermediate steps may be the rate-
determining step, depending on the nature of the catalytic
surface and other reaction conditions. In general, when the
cathodic Tafel slope is close to ꢁ120 mV decꢁ1 for Fe in acidic
solution, the rst reaction, which is a surface-catalyzed charged
transfer reaction and designated as the Volmer reaction, is
9268 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Fig. 7 SEM images of the surface of the mild steel sample after
immersing it for 6 h in 1 M HCl at room temperature.
compound B, but it not twice with respect to the other. This
suggests a leveling effect at a higher concentration towards the
corrosion inhibitory property, which originates from the fact
that at higher concentration, the metal surface is almost fully
covered by the inhibitors, and therefore they exert very
comparable inhibitory action. The better corrosion inhibitory
propensity of compound B over compound A can be explained
from the result obtained from the DFT calculation.
Fig. 8 Optimized geometry and electron distribution in the HOMO
and LUMO of compounds A (left) and B (right) as obtained from the
DFT study.
SEM morphology
By comparing the SEM images of the uninhibited and inhibited
(0.5 mM of compounds A and B) steel samples aer immersion
in 1 M HCl medium for 6 h, the anti-corrosive activity of the
studied Schiff base derivatives was determined. In the presence
of compound B, the metal surface was corroded less, reecting
the better corrosion mitigation propensity of compound B over
that of compound A for mild steel in aqueous HCl (Fig. 7).
higher the energy of the HOMO and the lower the energy of the
LUMO favour the forward and back electron donation, respec-
tively. Table 7 provides details of the quantum chemical
parameters of the inhibitors studied. Fig. 8, S17 and S18 (in the
ESI†) show the energy-optimized spatial congurations of the
inhibitors and the electronic distribution in the HOMO and
LUMOs of the inhibitors. It was observed that when vanillin is
conjugated with picolylamine through imine bond formation
(i.e. compound A), both EHOMO and ELUMO increase from that of
vanillin and picolylamine. Thus, the observed higher inhibition
efficiency of compound A compared to vanillin and picolyl-
amine may be attributed to the more facile forward electron
transfer for compound A than the retro-electron transfer. This is
reected by the fraction of electrons transferred from the
DFT study results
DFT study can be used to explain the nature of bonding between
the organic inhibitor molecules and metal. Electron transfer
from the inhibitor to the vacant 3d orbitals of Fe occurs through
the HOMO of inhibitor, whereas its LUMO is responsible for
retro-electron transfer from the lled 4s metal orbital.15,67 The
Table 7 Calculated molecular parameters from the DFT study
Inhibitors
EHOMO (eV)
ELUMO (eV)
DE (eV)
m (D)
I (eV)
A (eV)
X (eV)
h (eV)
s (eVꢁ1
)
DN
Neutral form
Compound A
Compound B
Vanillin
ꢁ6.1852
ꢁ6.0556
ꢁ6.5700
ꢁ6.8094
ꢁ1.3545
ꢁ1.4355
ꢁ1.8976
ꢁ0.9280
4.8307
4.6201
4.6724
5.8814
3.37
4.43
5.55
4.50
6.1852
6.0556
6.5700
6.8094
1.3545
1.4355
1.8976
0.9280
3.7698
3.7456
4.2338
3.8687
2.4154
2.3101
2.3362
2.9407
0.414
0.433
0.428
0.340
0.217
0.232
0.125
0.161
Picolyl amine
Protonated form
Compound A
Compound B
ꢁ6.8999
ꢁ7.2086
ꢁ1.9273
ꢁ2.4362
4.9726
4.7724
23.37
29.91
6.8999
7.2086
1.9273
2.4362
4.4136
4.8224
2.4863
2.3862
0.402
0.419
0.082
ꢁ0.0005
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9269

View Article Online
RSC Advances
Paper
inhibitor to the metal (DN). According to the electronic distri- atoms attached to the benzene rings (i.e., O atoms of the –OH
bution in the HOMO of vanillin, the electron cloud is distrib- and OCH3 substituents attached to the benzene group), C atoms
uted over the whole molecular surface, including the phenyl of the benzene rings and N atom of the imine group have higher
ring and the oxygen atoms present in the substituent groups f+k and fkꢁ values for both compounds. Thus, the vanillin moiety
attached to it. In compound A, it is extended up to the imine present in these compounds together with the imine bond are
group present. This elaborates the specic role of the imine mostly involved in electrophilic and nucleophilic reactions.
bond present in compound A towards its corrosion inhibition This tallies totally with the conclusion derived from the elec-
propensity. Thus, it can be concluded that the better forward tronic distribution in the HOMO and LUMO levels obtained
electron donation from compound A to the metal and the from the DFT calculation.
involvement of the imine group are the main reasons for the
greater inhibition efficacy of compound A compared to its
precursor molecules.
DFT study result with protonated form of inhibitors
The local reactivity analysis revealed that the O atom of the –OH
group attached to the benzene ring (O-3 in compound A and O-32
and 34 in compound B, Table S6 in the ESI†) possess the maximum
fꢁk value, making them the most prone towards protonation, i.e.
attack from an electrophile, H+. In highly acidic medium, it is ex-
pected that any organic molecule having heteroatoms with free lone
pairs of electrons should remain in the protonated form. Accord-
ingly, we performed the DFT calculation with monoprotonated
compound A and diprotonated compound B. It was observed that
protonation does not alter the electronic distribution to any
appreciable extent in the HOMO and LUMO levels for compound A
(Fig. S18 in the ESI†), which are mostly distributed over the vanillin
moiety and imine group. Thus, the interaction pattern for the
neutral and protonated compound A should more or less be the
same with the metal surface in acidic medium. For compound B,
there is denite change in the possible interaction mode (Fig. S18 in
the ESI†). The electronic distribution in the HOMO is localized
mostly on the imine and pyridine groups. The LUMO is distributed
all over the divanillin moiety, similar to that for the neutral state.
Here, according to the molecular parameters, it was observed that
protonation results in a decrease in both EHOMO and ELUMO. This
suggests that both compounds in their protonated state are more
reactive for acceptance of electrons from the metal surface rather
than forward electron transfer. As a result, value of the fraction of
electrons transferred, DN, decreases and even becomes negative for
compound B. This is contrary to the fact that experimentally,
compound B provides much better corrosion inhibition potentiality
than compound A. This supports the conclusion that despite the
existence of the protonated form of the inhibitor molecules in
acidic aqueous medium, during interaction with the metal surface,
the neutral form of the molecules is mostly involved. This conclu-
sion is also supported from the experimental observation that the
formation constants of the Schiff base–metal ion complexes are
Comparing compounds A and B, it is evident that the HOMO
energy is higher for compound B, whereas the LUMO energy is
lower for the dimeric form (compound B) than the monomeric
form (compound A). Thus, two-way electron transfer (inhibitor
to the metal and vice versa) is more facilitated for compound B
than compound A. This resulted in a higher soness, s, and
higher fraction of electrons transferred, DN, for compound B
compared to compound A. The higher soness of compound B
makes it more susceptible towards charge transfer. Further-
more, the dipole moment of compound B is greater than that of
compound A. Thus, the dimeric form is more prone to elec-
trostatic interaction. All these favourable molecular parameters
make compound
B the better corrosion inhibitor than
compound A. Analysis of the electronic distribution in the
HOMO and LUMO for both compounds revealed some inter-
esting information regarding the possible mode of interaction
between the inhibitors and the metal. In compound B, two
vanillin moieties lie in a planar orientation. The HOMO and
LUMO in compound A are distributed mostly over the vanillin
moiety and the imine group. In contrast, for compound B, these
encompass both the vanillin moieties and the two imine groups
attached with vanillin moieties. This implies that if compound
A is capable of interacting with either the cathodic or anodic
reaction site one at a time, compound B can interact with both
reaction sites simultaneously. This will lead to the cooperative
mode of interaction for the two vanillin moieties present in
compound B. When one moiety donates electrons from its
HOMO to the anodic reaction site (thereby reducing the rate of
the metal oxidation reaction by enhancing the electron density
in the anodic site), its ability to take up electrons in its LUMO
from the cathodic reaction site of the metal increases, thereby
reducing the rate of the hydrogen evolution reaction at the
cathodic site to a greater extent. The reverse phenomenon also
has equal possibility. Thus, both compounds essentially act as
mixed-type inhibitors, but compound B, by virtue of the coop-
erative interaction between its two vanillin moieties, exhibits
better prociency towards the mitigation of corrosion.19,39,68
greater than that of the corresponding protonated Schiff bases.69
similar conclusion was derived for other heterocyclic bases.9,10,34
A
Possible mode of interaction between the inhibitor molecules
and metal surface
From all the above observations, we proposed a plausible model
for the mode of adsorption of the studied Schiff bases on the
Local reactivity analysis
To establish the local reactivity of the atoms present in surface of mild steel in acidic medium. It was experimentally
compounds A and B against nucleophile and electrophile veried that the surface of mild steel bears an excess positive
attack, positive and negative Fukui indices (i.e., f+k and fkꢁ) values charge in 1 M HCl at the equilibrium potential.11 Chloride ions
for the individual atoms were calculated and the data is pre- and water molecules form an adsorbed layer on the positively
sented in Table S6 (in ESI†). This data clearly shows that the O charged metal surface.11 The protonated Schiff bases are
9270 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 8 Interaction energies of compound A and compound B on the
Fe (1 1 0) surface
attracted to the metal surface by the surface adsorbed chloride
ions. When these inhibitor molecules come sufficiently close to
the positively charged metal surface to form a chemisorbed
layer, the Schiff bases are deprotonated and interact directly
with the metal surface, replacing the pre-adsorbed chloride and
water molecules. The inhibitor molecules through the vanillin
moiety and the imine group donate electrons to the anodic sites
on the metal, whereas these same groups attract electrons from
the cathodic sites (Fig. 9). For compound A, electron donation
and acceptance involve two different inhibitor molecules,
whereas for compound B (the dimeric form of compound A),
a single molecule is capable of bi-directional electron transfer.
Electron donation and acceptance in a single molecule rein-
force each other, and is the basis for the observed superior
corrosion protection ability by compound B.
System
Einteraction (kJ molꢁ1
)
Fe + compound A
Fe + compound B
ꢁ625.19
ꢁ1199.05
the molecular dynamics simulation study is shown in Fig. 10,
and the corresponding interaction energy data is tabulated in
Table 8. The interaction energy for compound B was found to
be almost twice that of compound A. This veries the experi-
mental observation that compound B has stronger suscepti-
bility to interact with the Fe surface in 1 M aqueous HCl.
Furthermore, we estimated the closeness of approach of
various atoms present in both inhibitors with the metal
surface in the equilibrated state (Table S7 in the ESI†). It was
revealed that the O atoms present in the substituted –OH and
–OCH3 groups and N atoms of the imine and pyridine groups
MD simulation result
The equilibrium conguration of compounds A and B on the
Fe (1 1 0) surface in aqueous HCl medium as obtained from
ꢀ
form a distance with the Fe surface in the range of 2.8 to 3.8 A.
This is close enough to conclude that the inhibitor molecules
form a chemisorbed layer on the metal surface.1,15,35 A similar
conclusion was derived based on the calculated kinetic–ther-
modynamic parameters.
Conclusion
Vanillin is an important edible avour and available from bio-
resources. Vanillin and its dimer, divanillin, were derivatized
with picolyl amine into the corresponding Schiff base
compounds.
The adsorption behaviour and anti-corrosion potential of
these compounds towards mild steel in 1 M aqueous HCl were
compared. The main conclusions of our work are summarized
below.
Fig. 9 Pictorial representation of the adsorption of compound B on
the surface of mild steel in 1 M HCl medium.
ꢅ Compounds A and B (Schiff bases of vanillin and divanillin,
respectively) bestow good protection to mild steel from corro-
sion in 1 M HCl medium from room temperature to as high as
323 K at the mM concentration level.
ꢅ These Schiff base derivatives act as mixed-type corrosion
inhibitors, as evident from the potentiodynamic polarisation
study.
ꢅ The extent of the adsorption and corrosion inhibition
propensity of compound B is more than twice that of compound
A at a lower concentration level.
ꢅ These molecules are adsorbed on the metal surface by
chemical mode of adsorption through the vanillin moiety and
imine group.
ꢅ The entire adsorption process is governed by both enthalpy
and entropy changes.
ꢅ Due to its higher HOMO and lower LUMO energy,
compound B acts as the better corrosion inhibitor. The other
intrinsic molecular parameters such as global soness and
dipole moment also impart a favourable impact during the
interaction of compound B with the surface of mild steel in
aqueous HCl.
Fig. 10 Equilibrium adsorption configurations of compound A (a and
b) and compound B (c and d) on the Fe (1 1 0) surface obtained by
molecular dynamics simulation. Top: top view and Bottom: side view
(water molecules are not shown for clarity).
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9271

View Article Online
RSC Advances
Paper
¨
ꢅ The Langmuir adsorption model can be suitably applied to 16 E. Ozdemir and G. Gece, Key Eng. Mater., 2019, 800, 108–112.
calculate the different thermodynamic parameters related to 17 G. L. Mendonça, S. N. Costa, V. N. Freire, P. N. Casciano,
the adsorption process.
A. N. Correia and P. de Lima-Neto, Corros. Sci., 2017, 115,
41–56.
ꢅ The interaction energy obtained from the molecular
dynamics simulation is nearly two times higher for compound B 18 Y.-J. Yang, Y.-K. Li, L. Wang, H. Liu, D.-M. Lu and L. Peng,
than compound A, revealing that the simulation method is an
Int. J. Electrochem. Sci., 2019, 14, 3375–3392.
important theoretical tool to compare the anti-corrosive 19 P. Roy, P. Karfa, U. Adhikari and D. Sukul, Corros. Sci., 2014,
behaviour of structurally comparable corrosion inhibitors.
88, 246–253.
20 P. Roy and D. Sukul, RSC Adv., 2015, 5, 1359–1365.
21 P. Roy, T. Maji, S. Dey and D. Sukul, RSC Adv., 2015, 5,
61170–61178.
Conflicts of interest
There are no conicts to declare.
22 D. Sukul, A. Pal, S. Mukhopadhyay, S. K. Saha and
P. Banerjee, J. Mol. Liq., 2018, 249, 930–940.
23 W. Al Zoubi, A. A. S. Al-Hamdani and M. Kaseem, Appl.
Organomet. Chem., 2016, 30, 810–817.
Acknowledgements
PB thanks Department of Science and Technology (DST), Govt. 24 A. W. Jeevadason, K. K. Murugavel and M. A. Neelakantan,
of India for research grant no. GAP-225612. DS, PB and AS thank Renewable Sustainable Energy Rev., 2014, 36, 220–227.
Department of Science & Technology and Biotechnology, 25 P. Shetty, Chem. Eng. Commun., 2019, DOI: 10.1080/
Government of West Bengal, India for supporting research
00986445.2019.1630387.
projects no. 217(Sanc.)/ST/P/S&T/15G-10/2017 and 78(Sanc.)/ST/ 26 A. K. Sinha, U. K. Sharma and N. A. Sharma, Int. J. Food Sci.
P/S&T/6G-1/2018. Infrastructural Development of the Depart- Nutr., 2008, 59, 299–326.
ment of Chemistry, NIT Durgapur through DST-FIST grant of 27 J. D. P. Araujo, C. A. Grande and A. E. Rodrigues, Chem. Eng.
´
DST, Govt of India (no. SR/FST/CSI-267/2015) is also gratefully
acknowledged.
Res. Des., 2010, 88, 1024–1032.
28 M. Fache, B. Boutevin and S. Caillol, ACS Sustainable Chem.
Eng., 2016, 4, 35–46.
29 A. Y. El-Etre, Corros. Sci., 2001, 43, 1031–1039.
30 H. Yamamoto, T. Hoshino and T. Uchiyama, Biosci.,
Biotechnol., Biochem., 1999, 63, 390–394.
31 A. S. Amarasekara, B. Wiredua and A. Razzaq, Green Chem.,
2012, 14, 2395–2397.
32 A. S. Amarasekara and A. Razzaq, ISRN Polym. Sci., 2012,
DOI: 10.5402/2012/532171.
33 V. Balachandan and K. Parimala, Spectrochim. Acta, Part A,
2012, 95, 354–368.
References
1 D. Sukul, A. Pal, S. K. Saha, S. Satpati, U. Adhikari and
P. Banerjee, Phys. Chem. Chem. Phys., 2018, 20, 6562–6574.
2 Z. Tang, Curr. Opin. Solid State Mater. Sci., 2019, DOI:
10.1016/j.cossms.2019.06.003.
3 S. Gangopadhyay and P. A. Mahanwar, J. Coat. Technol. Res.,
2018, 15, 789–807.
´
4 T. J. Harvey, F. C. Walsh and A. H. Nahle, J. Mol. Liq., 2018,
266, 160–175.
34 A. Dutta, S. K. Saha, U. Adhikari, P. Banerjee and D. Sukul,
Corros. Sci., 2017, 123, 256–266.
35 S. K Saha and P. Banerjee, Mater. Chem. Front., 2018, 2, 1674–
1691.
5 S. K. Saha, A. Dutta, P. Ghosh and D. Sukul, Phys. Chem.
Chem. Phys., 2015, 17, 5679–5690.
6 B. Chugh, A. K. Singh, S. Thakur, B. Pani, A. K. Pandey,
H. Lgaz, I.-M. Chung and E. E. Ebenso, J. Phys. Chem. C, 36 I. B. Obot, D. Macdonald and Z. M. Gasem, Corros. Sci., 2015,
2019, 123, 22897–22917. 99, 1–30.
7 W. Zhang, Y. Wang, H.-J. Li, Y. Liu, R. Tao, S. Guan, Y. Li and 37 A. Kokalj, Chem. Phys., 2012, 393, 1–12.
Y.-C. Wu, J. Phys. Chem. C, 2019, 123, 14480–14489.
8 C. Verma, L. O. Olasunkanmi, E. E. Ebenso, M. A. Quraishi
and I. B. Obot, J. Phys. Chem. C, 2016, 120, 11598–11611.
9 A. Dutta, S. K. Saha, P. Banerjee and D. Sukul, Corros. Sci.,
2015, 98, 541–550.
10 A. Dutta, S. K. Saha, P. Banerjee, A. K. Patra and D. Sukul,
RSC Adv., 2016, 6, 74833–74844.
11 R. Yıldız, Ionics, 2019, 25, 859–870.
38 G. Fatti, P. Restuccia, C. Calandra and M. C. Righi, J. Phys.
Chem. C, 2018, 122, 28105–28112.
39 P. Roy, S. K. Saha, P. Banerjee, S. Dey and D. Sukul, Res.
Chem. Intermed., 2017, 43, 4423–4444.
40 P. Geerlings, F. De. Pro and W. Langenaeker, Chem. Rev.,
2003, 103, 1793–1873.
41 S. Saha, R. K. Roy and P. W. Ayers, Int. J. Quantum Chem.,
2009, 109, 1790–1806.
12 W. Gong, X. Yin, Y. Liu, Y. Chen and W. Yang, Prog. Org. 42 F. D. Pro, C. V. Alsenoy, A. Peeters, W. Langenaeker and
Coat., 2019, 126, 150–161.
P. Geerlings, J. Comput. Chem., 2002, 23, 1198–1209.
13 A. Singh, K. R. Ansari, M. A. Quraishi, S. Kaya and 43 R. G. Parr and R. G. Pearson, J. Am. Chem. Soc., 1983, 105,
P. Banerjee, New J. Chem., 2019, 43, 6303–6313. 7512–7516.
14 J. Baux, N. Causse, J. Esvan, S. Delaunay, J. Tireau, M. Roy, 44 W. Yang and W. J. Mortier, J. Am. Chem. Soc., 1986, 108,
´
´ `
D. You and N. Pebere, Electrochim. Acta, 2018, 283, 699–707.
5708–5711.
15 M. Murmu, S. K. Saha, N. C. Murmu and P. Banerjee, Corros.
Sci., 2019, 146, 134–151.
9272 | RSC Adv., 2020, 10, 9258–9273
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
45 M. Chevalier, M. Lebrini, F. Robert, S. Sutour, F. Tomi, 57 G. Ehrlich, in Chemistry and Physics of Solid Surfaces VII, ed.
C. Jama, F. Bentiss and C. Roos, Int. J. Electrochem. Sci.,
2019, 14, 1208–1223.
R. Vanselow and R. F. Howe, Springer, Berlin, Heidelberg,
1988, ch. 1, pp. 1–64.
46 U. Rammelt and G. Reinhard, Corros. Sci., 1987, 27, 373–382. 58 W. Durnie, R. De. Marco, A. Jefferson and B. Kinsella, J.
47 T. Pajkossy, Solid State Ionics, 2005, 176, 1997–2003. Electrochem. Soc., 1999, 146, 1751–1756.
48 S. Martinez and M. Metikos-Hukovic, J. Appl. Electrochem., 59 A. Pal, S. Dey and D. Sukul, Res. Chem. Intermed., 2016, 42,
2003, 33, 1137–1142. 4531–4549.
49 S. S. Abdel-Rehim, K. F. Khaled and N. S. Abd-Elsha, 60 A. Zarrouk, B. Hammouti, H. Zarrok, S. S. Al-Deyab and
Electrochim. Acta, 2006, 51, 3269–3277. M. Messali, Int. J. Electrochem. Sci., 2011, 6, 6261–6274.
50 M. Lebrini, F. Robert and C. Roos, Int. J. Electrochem. Sci., 61 E. Zaki, F. Mirabella, F. Ivars-Barcelo, J. Seifert, S. Carey,
ˇ
´
2010, 5, 1698–1712.
S. Shaikhutdinov, H.-J. Freund, X. Li, J. Paier and J. Sauer,
51 H. Lgaz, I. M. Chung, R. Salghi, I. H. Ali, A. Chaouiki,
Phys. Chem. Chem. Phys., 2018, 20, 15764–15774.
Y. El. Aour and M. I. Khan, Appl. Surf. Sci., 2019, 463, 62 D. J. Lavrich, S. M. Wetterer, S. L. Bernasek and G. Scoles, J.
647–658. Phys. Chem. B, 1998, 102, 3456–3465.
52 A. J. Bard and L. R. Faulkner, Electrochemical Methods: 63 A. Popova, E. Sokolova, S. Raicheva and M. Christov, Corros.
Fundamentals and applications, John Wiley & Sons, Inc,
USA, 2nd edn, 2001, pp. 544–572.
53 P. Simon and A. Burke, Interface, 2008, 17, 38–43.
Sci., 2003, 45, 33–58.
64 R. Idouhli, Y. Koumya, M. Khadiri, A. Aityoub, A. Abouelda
and A. Benyaich, Int. J. Ind. Chem., 2019, 10, 133–143.
54 E. McCafferty and N. Hackerman, J. Electrochem. Soc., 1972, 65 S. Szabo and I. Bakos, Corros. Rev., 2004, 22, 183–192.
119, 146–154. 66 W. Wang and C. J. Roberts, AAPS J., 2013, 15, 840–851.
55 E. E. Oguzie, S. G. Wang, Y. Li and F. H. Wang, J. Phys. Chem. 67 G. Gece, Corros. Sci., 2008, 50, 2981–2992.
C, 2009, 113, 8420–8429.
68 D.-Q. Zhang, B. Xie, L.-X. Gao, Q.-R. Cai, H. Goun, J. Kang
56 Y. Liu, J. Chem. Eng. Data, 2009, 54, 1981–1985.
and Y. Lee, Thin Solid Films, 2011, 520, 356–361.
69 K. Rengaraj, B. Sivasankar, M. Anbu and M. Palanichamy, J.
Chem. Sci., 1991, 103, 707–713.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 9258–9273 | 9273