Corrosion Science 76 (2013) 317–324

Contents lists available at SciVerse ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Experimental and theoretical studies for mild steel corrosion inhibition

in 1 M HCl by two new benzothiazine derivatives
T. Ghailane a, R.A. Balkhmima b, R. Ghailane a,⇑, A. Souizi a, R. Touir b, M. Ebn Touhami b, K. Marakchi c,
N. Komiha c
a
Laboratoire de Synthèse Organique, Organométallique et Théorique, Faculté des Sciences, Université Ibn Tofail, B.P 13314 000 Kénitra, Morocco
b
Laboratoire des Matériaux, Electrochimie et Environnement, Faculté des Sciences, Université Ibn Tofail, B.P 13314 000 Kénitra, Morocco
c
LS3ME-Equipe de Chimie Théorique et Modélisation, Faculté des Sciences, Université Mohamed V-Agdal, Avenue Ibn Batouta, B.P 1014 Rabat, Morocco

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

Article history: The inhibition effect of new heterocyclic compounds, namely 2-aryl-benzothiazin-3-one (P1) and 3-aryl-
Received 24 December 2012 benzothiazin-2-one (P2) on mild steel corrosion in 1 M HCl was investigated using electrochemical mea-
Accepted 28 June 2013 surements. The results indicated that the inhibition efficiency depends on concentration and molecular
Available online 9 July 2013
structure of the investigated compounds. It is also found that the inhibition of P1 is greater than P2.
The molecular structure effect on the corrosion inhibition efficiency was investigated using DFT calcula-
Keywords: tions. The structural and electronic parameters were calculated and discussed. The obtained results show
A. Acid solutions
that the experimental and theoretical studies agree well and confirm that P1 is the better inhibitor.
A. Mild steel
B. EIS
Ó 2013 Elsevier Ltd. All rights reserved.
B. Polarization
C. Acid corrosion
C. Acid inhibition

1. Introduction The aim of this work is to study the inhibition efficiencies of 2-

aryl-benzothiazin-3-one (P1) and 3-aryl-benzothiazin-2-one (P2)
The search and application for new and efficient corrosion synthesized by Souizi et al. [13] on mild steel corrosion in 1 M
inhibitors becomes an obligation to secure metallic materials HCl using potentiodynamic polarization curves tests and electro-
against corrosion. The selection of appropriate inhibitors mainly chemical impedance spectroscopy (EIS) measurements. The tradi-
depends on the formulation as well as rational use in various envi- tional technique such as electrochemical and quantum chemical
ronments. Organic compounds with nitrogen and oxygen func- method has been employed in this study. The quantum chemical
tional groups as well as multiple bonds or aromatic rings are parameters such as highest occupied molecular orbital energy
considered to be one of the effective chemicals corrosion inhibi- (EHOMO), lowest unoccupied molecular orbital energy (ELUMO),
tions of metal [1–3]. Many research works have been developed (LUMO–HOMO) energy gap (DE), dipole moment (l), electronega-
in order to correlate the substituent effect and the inhibition effi- tivity (v), ionization potential (I), electron affinity (A) global hard-
ciency of organic molecules [4–7]. Earlier reports [8] have been ness (g) and fraction of electron transfer (DN) from the inhibitor
shown that inhibition efficiency of organic compounds containing molecules to iron were determined. The electronic parameters
heterocyclic nitrogen increases with the number of aromatic rings such as natural atomic populations and Fukui indices as indicator
and the availability of electronegative atoms in the molecule. of the reactive centers were also obtained by DFT method in order
Quantum chemical calculations have been widely used to study to understand the adsorption properties of benzothiazine deriva-
reactional mechanisms and to interpret the experimental results tives. In order to take account of the solvent effects, the structural
as well as to resolve chemical ambiguities [9]. However, the inhibi- and electronic parameters were also determined in aqueous
tion efficiency was reinforced by the presence of heteroatoms such solution.
as sulfur, nitrogen and oxygen in the ring which facilitates its
adsorption on metallic surface following the sequence O < N < S
2. Experimental details
[10–12].
2.1. Material preparation and inhibitors

⇑ Corresponding author. Tel.: +212 661631386. The electrolysis cell was a borrosilcate glass (PyrexÒ) cylinder
E-mail address: ghailane_r@yahoo.fr (R. Ghailane). closed by cap with five apertures. Three of them were used for

0010-938X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.corsci.2013.06.052
318 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324

the electrode insertions (Fig. 1). The working electrode was mild Table 1
steel which its composition is summarized in Table 1. The investi- Chemical composition of low carbon steel in wt%.

gated area of the surface was 1 cm2. C Si Mn Cr Mo Ni Al Cu Co V W Fe

Prior to immersion test, the electrode was abraded using emery 0.11 0.24 0.47 0.12 0.02 0.1 0.03 0.14 <0.0012 <0.003 0.06 Balance
paper up to 1200 grade, cleaned with ethanol, washed with dis-
tilled water, and dried finally. The aggressive solution of 1 M HCl
was prepared by dilution of analytical grade 37% HCl with distilled
water. The pH of solution was about 1.2 ± 0.1. The structural for- program [14]. The inhibition efficiency, gi, was evaluated with
mulas of the examined inhibitors in this study are shown in Fig. 2. the relationship:

Rct  R0ct
2.2. Electrochemical measurements gi ¼  100 ð3Þ
Rct
The working electrode was pressure fitted into a polytetrafluo- where R0ct and Rct are the charge transfer resistance values in the ab-
roethylene holder (PTFE) exposing only 1 cm2 of surface to the sence and in the presence of inhibitors, respectively.
solution. Platinum and saturated calomel were used as counter
and reference electrode (sce), respectively. All potentials were 2.3. Computational details
measured against this electrode.
The potentiodynamic polarization curves were recorded by For the theoretical study, complete geometry optimizations of
changing the electrode potential automatically from negative val- the benzothiazine derivatives were performed using the Density
ues (900 mV/sce) to positive values (100 mV/sce) versus Ecorr Functional Theory (DFT) with the Beck’s three parameter exchange
using a Potentiostat/Galvanostat type PGZ 100, at a scan rate of functional and the Lee–Yang–Parr non-local correlation functional
1 mV/s after 1 h of immersion time until reaching steady state. (B3LYP) [15–17] with 6-311++G (2d, 2p) basis set of atomic orbi-
The temperature of solutions was held at 25 ± 1 °C by a thermostat tals as implemented in Gaussian 03 program package [18]. This ap-
water bath. All measurements were carried out in the acid solu- proach was shown to yield geometries in agreement with
tions in air atmosphere without bubbling. experiment for a wide variety of systems. Furthermore, DFT was
The evaluation of corrosion kinetics parameters was obtained considered a very useful technique to probe the inhibitor/surface
using a non-linear regression calculation according to Stern–Geary interaction as well as to analyze the experimental data.
equation: Any physical or chemical property measured in a solvent could
be different if it is measured in another solvent or even in vacuum
i ¼ icorr  fexp½2:3  ðE  Ecorr Þ=ba   exp½2:3  ðE
[19–22]. Moreover, it is well known that the phenomenon of elec-
 Ecorr Þ=bc g ð1Þ trochemical corrosion appears in liquid phase and it is expected
that the inhibiting molecules in solution behave differently from
where icorr is the corrosion current density (A cm2), and ba and bc that in vacuum. Thus, it is necessary to include the solvent effect
are respectively the Tafel slope anodic and cathodic reactions (V/ in the computational calculations. In a similar way, it is important
decade). to take the solvent effect which could appear also in the geometric
The inhibition efficiency, gi, was evaluated from icorr values and in the electrical properties. In fact, the solvent effect on the
using the following relationship: molecular structures of the investigated compounds was taking ac-
count in the present work by a model known as polarized contin-
0
icorr  icorr uum model (PCM) [23] implemented in Gaussian 03 program. In
gi ¼ 0
 100 ð2Þ
icorr this model, the solvent was treated as a continuum dielectric med-
ia and the solute is considered as a trapped molecule in a cavity
where the superscript 0 indicates the corrosion current densities surrounded by solvent.
relative to that determined in the absence of inhibitor. The geometry of the investigated compounds was determined
The electrochemical impedance spectroscopy measurements by optimizing all geometrical variables in gas and aqueous phases
were carried out using a transfer function analyzer (Voltalab PGZ without any symmetry constraint, employing analytic energy gra-
100, Radiometer Analytical), over the frequency range from dients. The harmonic frequencies were computed from analytical
100 kHz to 0.01 Hz with 10 points per decade. The applied ampli- derivatives for all species in order to check the minimum-energy
tude of AC signal was 10 mVrms. All experiments were performed at structures. The quantum chemical parameters obtained from this
the open circuit potential. The obtained impedance data were ana- optimized structure have been determined and analyzed in order
lyzed in term of equivalent electrical circuit using Bouckamp’s to explain the interaction between the inhibitor molecules and

Fig. 1. Schematic of test cell.

 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324 319

Cl
Me

S H
N

N O
H S O
P1: 2-aryl-benzothiazin-3-one ( aryl = PhCl) P2 : 3-aryl-benzothiazin-2-one (aryl = PhCH3)
Fig. 2. Chemical structure of the investigated benzothiazine derivatives.

the metal surface. According to DFT-Koopmans’ theorem [24–26], 3. Results and discussion
the HOMO energy is related to the ionization potential (I) whereas
the LUMO energy is linked to the electron affinity (A), as follows: 3.1. Potentiodynamic polarization curves

Fig. 3 presents potentiodynamic polarization curves for mild

I ¼ EHOMO ð4Þ
steel in 1 M HCl containing different concentrations of P1 and P2.
It can be remarked that the anodic branches do not exhibit well-
defined Tafel region. It is seen also that the inhibitors addition hin-
A ¼ ELUMO ð5Þ
dered the acid attack on mild steel. Indeed, an increase in their
Then, the electronegativity (v), the chemical potential (l) and concentrations gives a decrease in anodic current densities indicat-
the global hardness (g) were evaluated, based on the finite differ- ing that these inhibitors act on the electrode dissolution reaction
ence approximation, as linear combinations of the calculated I such as anodic type inhibitors. In addition, it is noted that the ano-
and A: dic reaction appears inhibited in both cases, but the cathodic reac-
tion is inhibited only in the P1 case.
However, icorr, Ecorr, ba and bc were evaluated from the experi-
IþA
v ¼ l ¼ ð6Þ mental results using a user defined function of ‘‘Non-linear least
2 squares curve fit’’ (Eq. (1)) of graphic software (Origin, OriginLab).
In all cases, the correlation factor R2 is greater than 0.99 indicating
a reliable result. Fig. 4 shows, as for an example, the results of
IA
g¼ ð7Þ regressions calculations for the cathodic and anodic scan in the
2 presence of different concentration of P1. In this calculation, the
The fraction of transferred electrons (DN), evaluating the elec- potential domain is limited to Ecorr ± 100 mV/sce. It can be seen
tronic flow in a reaction of two systems with different electroneg- in this figure a good agreement between the experimental and fit-
ativities, in particular case; a metallic surface and an inhibitor ting data.
molecules, was calculated according to Pearson theory [27] as Table 2 summarizes the obtained corrosion kinetic parameters.
follows: Ecorr and the kinetic parameters calculated by a non-linear regres-
sion calculation from the anodic polarization scans near the open
vFe  vinh circuit potential are similar to those determined from the cathodic
DN ¼ ð8Þ polarization scans, though icorr is slightly higher. It can be seen that
2ðgFe þ ginh Þ
the icorr values decreased with concentrations of both inhibitors
and the least value was obtained at 103 M, due to the increase
where the indices Fe and inh refer to iron atom and inhibitor mole- in the blocked fraction of the electrode surface by adsorption. This
cule, respectively. decrease is very marked in the case of P1 explaining by the

(a) 10 2 (b)
102 Blank
-3
10 M -5 10 M
-5
10 M
Blank 10 1
101
i (mA cm )

i (mA cm )
-2

-2

10 0 -5
10 M -3
-4
10 M 10 M
0
10
10 -1

10-1
10 -2

10-2 10 -3
-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1
E (V/sce) E (V/sce)

Fig. 3. Potentiodynamic polarization curves of mild steel in 1 M HCl at various concentration of (a) P1 and (b) P2.
320 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324

101 500 60
Blank solution
-5
Blank
10 M of P1 50
-4
10 M of P1

- ZIm ( Ω cm2)
40
400 -3
10 M of P1
-5
0 10 M
10 -4 30
10 M
(a) 20

- ZIm ( Ω cm )
31.62

2
300 10
i (mA cm )

10
-2

1K 1 0.1
-3
10 M 0
10-1 6.3 0 10 20 30 40 50 60
200 ZRe (Ω cm )
2

630.95
63.09 0.398
100
10-2
1.99 158.48
1.99K 12.58
125.58
0.1
0
Squater : Experimental curves
Line : Fitting curves 0 100 200 300 400 500 600 700
10-3 2
-0.7 -0.6 -0.5 -0.4 -0.3
ZRe (Ω cm )
E (V/sce)
400 60
Fig. 4. Comparison of experimental and fitting data using a non-linear fitting with Blank solution
-5 50
Stern–Geary equation (in the presence of different concentration of P1). 10 M of P2 100
-4

- ZIm ( Ω cm2)
10 M of P2 40
125.82
50.11

300 -3
10 M of P2 30
presence of chloride atom which gives more inhibition. This behav-
ior could be attributed to strong interaction via adsorption be-
2
- ZIm ( Ω cm ) (b) 20 31.62

10
10
tween compound and metallic surface [28]. This may be also due
200
to the adsorption of inhibitor at mild steel surface through non- 0
0.1

0 10 20 30 40 50 60
10
bonding electron pairs present on nitrogen, oxygen and sulfur ZRe (Ω cm )
2

atoms as well as p-electrons [29]. Thus, the variation of inhibition 10

efficiency mainly depends on the substitution of ACH3 group by 100 2.51

7.94
ACl atom. However, the high inhibitive performance of P1 suggests 1.25 1.258
a higher bonding ability on mild steel surface, which possess high-
er number of lone pairs on heteroatoms and p-orbitals. The elec- 0.1
0
tron lone pair on the nitrogen will co-ordinate with the metal 0 100 200 300 400 500
atoms of active sites. Indeed, the chloride atom provides a high 2
ZRe (Ω cm )
electron density compared to ACH3 group, caused a strong interac-
tion with metal surface. Fig. 5. Nyquist plots for mild steel in 1 M HCl at various concentrations of (a) P1
and (b) P2.

3.2. Electrochemical impedance spectroscopy of water molecules originally and/or chlorine ions on mild steel
surface [31]. This effect leads to a protective film, inhibiting species
In order to confirm the results extracted from polarization curve or both on mild steel surface, then decreasing on the dissolution
and to acquire more information about corrosion mechanisms, EIS reaction extend [32,33]. In addition, the decrease of Cct with inhib-
measurements were carried out at corrosion potential. The ob- itors concentration is probably due to a decrease in local dielectric
tained results after immersion in 1 M HCl with and without P1 or constant and/or an increase in the thickness of a protective layer at
P2 at different for 1 h are presented in Fig. 5. It can be seen that electrode surface. The inhibition efficiencies, got from electro-
the impedance diagrams exhibit one capacitive loop which its cen- chemical impedance measurements, show the same trend as those
ter lays under the abscissa. This difference has been attributed to obtained from polarization measurements.
frequency dispersion [30] imputed to different physical phenom-
ena such as roughness and non-homogeneities of the surface.
These diagrams have similar form for all tested concentrations, 3.3. Quantum chemical study
indicating that almost no change in the corrosion mechanism.
Table 3 shows that Rct values increase while Cct decrease with 3.3.1. Molecular geometry
inhibitors concentration indicating that these inhibitors act by The molecules were built with the Gauss View 3.0 implemented
adsorption at the metal/solution interface by gradual displacement in Gaussian 03 package [18], their corresponding geometries were

Table 2
Electrochemical parameter values evaluated from potentiodynamic polarization curves for mild steel in 1 M HCl at various inhibitors concentrations.

Conc. (M) Ecorr (mV/sce) icorr (mA cm2) bc (mV/decade) ba (mV/decade) gi (%)
Blank solution 00 634 0.6768 168.1 83.4
P1 105 476 0.1540 178.4 120.6 77
104 398 0.0598 239.6 86.8 91
103 430 0.0291 673.4 79.1 96
P2 105 405 0.1594 214.4 47.8 76
104 399 0.0642 236.7 57.7 90
103 399 0.0515 314.2 53.6 92
 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324 321

Table 3 In addition, it is noted that the solvent effect is slightly weak on

Parameters data evaluated from impedance diagrams for mild steel in 1 M HCl at the dihedrals and bonds angles except for C24AC12AC15AC16
various inhibitors concentrations.
dihedral angle in P1 which is increased by about 6°. However,
Conc. (M) Rct (X cm2) Cct (lF cm2) gi (%) the greatest increase in bonds lengths (about 0.01 Å) was observed
Blank solution 00 38 146.9 – for C21AS in P2. This is probably due to the polarization of CAS
P1 105 254 80 85 bond by the attraction of the carbonyl group. The bond length va-
104 325 69.44 90 lue of C@O increases by 0.009 Å and 0.007 Å in P1 and P2, respec-
103 596 56.8 94 tively. Indeed, the bond length value of C26ACl for P1 is lengthiest
P2 105 287 104.6 88 by about 0.005 Å in aqueous phase while the C24AN bond length
104 324 77.94 89 decreases by about 0.009 Å.
103 404 64.41 92

3.3.2. Global molecular reactivity

fully optimized at B3LYP/6-311++G (2d, 2p) level of theory and the Frontier orbital theory is useful in predicting adsorption centers
vibrational calculations probe that their equilibrium structures of the inhibitor molecules responsible of the interaction metallic
correspond to the minima energy for each benzothiazine deriva- surface/molecule [34–36]. The terms involving the frontier molec-
tives (absence of imaginary frequencies). Moreover, the geometries ular orbitals (FMO) could provide dominative contribution, be-
of the investigated molecules were re-optimized in aqueous phase cause of the inverse dependence of stabilization energy on
at the same level of theory using PCM model for a better approach orbital energy difference (DE = ELUMO EHOMO).
of the experimental parameters. The final geometries and the com- The HOMO energy (EHOMO) is often associated to the electron
putational results parameters of P1 and P2 are given in Fig. 6 and donating ability of the molecule, therefore, inhibitors with high
Table 4, respectively. It is observed from the computational results values of EHOMO have a tendency to donate electrons to appropriate
that the substitution of Cl atom by CH3 group and the exchange the acceptor with low empty molecular orbital energy. Conversely, the
NH group with S atom in thiazine ring in P1 molecule involve sig- LUMO energy (ELUMO) indicates the electron accepting ability of the
nificant modifications on the structural parameters such as bond molecule, the lowest its value the higher the capability of accepting
distances, bond angles and dihedral angles of the studied electrons. The gap energy between the frontier orbitals (DE) is an-
inhibitors. other important factor in describing the molecular activity, so
In addition, the comparison between the optimized geometrical when the gap energy decreased, the inhibitor efficiency is im-
structures of P1 and P2 compounds in vacuum showed that the proved [37]. Khaled has investigated the adsorption of thiourea
bond length C@O is longer in P1 than in P2 by about 0.015 Å and and their derivatives on iron surface in nitric acid medium and
the C24AN14 bond in P1 is shorter than the single C11AN25 in he has found that the gap energy decreases with increasing of inhi-
P2 by about 0.093 Å. These differences can be explained by the bition efficiency [38].
delocalization in O25@C24AN14 region in P1 compound. The The calculated quantum chemical parameters related to the
CAC bond where the first carbon is linked to oxygen atom and inhibition efficiency of the studied molecules, such as the FMO
the second to S in P1 and N in P2, show a length difference of about energies (EHOMO, ELUMO), the gap energy (DE), the dipole moment
0.01 Å. (l), the electronegativity (v), the ionization potential (I), the elec-
Large values of the bond lengths were observed for CAS and tron affinity (A), the global hardness (g), and the fraction of elec-
CACl in P1. The C21AS26 bond length in P2 is seen to be shorter tron transfer (DN) from the inhibitor molecules to iron, are
than C12AS13 in P1 compound. For this, it can be conclude that collected in Table 5. It is seen that the highest EHOMO is obtained
the adsorption on metallic surface is clearly easy with CAS and for P2, this result is not sufficient to conclude that P2 is more effi-
CACl for P1 and only with CAS for P2. cient than P1. Therefore, the high values of EHOMO are likely to indi-
The compounds under investigation are not planar, this result is cate a tendency of the molecule to donate electrons to appropriate
confirmed by the values of the C4AS13AC12 and S13AC12AC15 acceptor molecules with low empty molecular orbitals. Indeed, the
bond angles for the P1 compound and C3AS26AC21 and excellent corrosion inhibitors are usually the organic compounds
N25AC11AC21 bond angles for P2 compound (Table 4). The values which not only give electrons to the unoccupied orbital of metal,
of dihedral angle C24AC12AC15AC16 in P1 and but also to accept free electrons from it [34].
C21AC11AC12AC13 in P2 exhibit a distortion by about 18°. Thus, the ELUMO and gap energy values obtained for P1 show that
However, a distortion by about 10° within the thiazine ring is it has the highest inhibition efficiency. This result is in good agree-
also observed in P1 compared with C11AN25AC4AC3 dihedrals ment with the experimental efficiency gi (%) (Table 5).
angles values in P2. This result explains that the adsorption on The dipole moment (l) provides information on the polarity of
metallic surface was easier for P1 than P2. the whole molecule. High dipole moment values are reported to
facilitate adsorption (and therefore inhibition) by influencing the

Fig. 6. Optimized molecular structures of P1 and P2.

322 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324

Table 4
Geometrical parameters of the studied inhibitors calculated at B3LYP/6-311++G (2d, 2p) in gas (G) and aqueous (A) phases.

Inhibitor Phase Bond lengh (Å) Bond angle (°) Dihedral angle (°)
P1 G C4AS 1.772 C4ASAC12 97.9 C12ASAC4AC3 34.0
A 1.773 98.0 33.7
G C12AS 1.854 SAC12AC15 109.2 C24AC12ASAC4 55.2
A 1.852 109.2 54.2
G C3AN 1.403 SAC12AC24 109.6 C3ANAC24AO 177.8
A 1.405 109.9 179.2
G C24AN 1.375 C3ANAC24 128.6 C3ANAC24AC12 4.9
A 1.364 128.4 3.6
G C12AC24 1.531 NAC3AC4 120.8 C24AC12AC15AC16 52.5
A 1.530 121.0 58.3
G C24AO 1.214 NAC24AC12 115.5 C15AC12ASAC4 178.4
A 1.223 116.1 179.5
G C26ACl 1.757 OAC24AN 121.0 C15AC12AC24AN 169.2
A 1.762 121.1 167.7
P2 G C3AS 1.780 C3ASAC21 104.0 C11ANAC4AC3 43.8
A 1.781 104.1 44.4
G C21AS 1.799 SAC3AC4 120.9 C4ANAC11AC21 67.0
A 1.789 120.6 66.6
G C4AN 1.396 SAC21AC11 116.4 NAC11AC21AO 142.5
A 1.396 116.8 143.4
G C11AN 1.468 C4ANAC11 118.2 C21AC11AC12AC13 70.3
A 1.469 117.9 72.0
G C11AC21 1.541 NAC4AC3 119.5 SAC21AC11AC12 162.8
A 1.538 119.7 162.0
G C21AO 1.199 NAC11AC21 110.0 C12AC11AC21AO 19.1
A 1.206 110.1 19.8
G C23AC28 1.508 OAC21AS 119.1 C12AC11ANAC4 168.6
A 1.508 118.7 167.0

transport process through the adsorbed layer [39]. Several authors (Table 5), however, a slight modification was usually obtained for
have stated that the inhibition efficiency increases with dipole mo- these parameters. Accordingly, it is acceptable to do calculations
ments values [40,41]. On the other hand, literature survey reveals in the gas phase because it reduces the time and the calculations
that many irregularities appear in the correlation of dipole mo- cost and no significant differences are found.
ments with inhibition efficiency [36,42]. The dipole moments of The optimized geometries of the studied compounds in the neu-
P1 and P2 are 6.1983 D and 5.4420 D, respectively, which are high- tral form including their HOMO and LUMO distributions density
er than that of H2O (l = 1.88 D). The high dipole moment value of are shown in Fig. 7. It can be seen that P1 and P2 have a similar
these compounds probably indicates strong dipole–dipole interac- HOMO distribution, which are all mainly located on the benzothia-
tions between them and metallic surface [43]. Accordingly, the zin-3-one in P1 moiety and benzothiazin-2-one in P2 moiety.
benzothiazine molecules adsorption in aqueous solution can be re- These results indicate that the interaction between these mole-
garded as a quasi-substitution process of the water molecules by cules and the steel surface concerns their moieties. In addition,
the inhibitors molecules at metal surface (H2Oads). the LUMO electronic density in P1 was distributed on its entire
The fraction of electrons transferred (DN) from inhibitor to mild while in P2, it was restricted to its benzothiazin-2-one moiety.
steel surface was also calculated using a theoretical vFe and gFe val-
ues for mild steel of 7 eV mol1 and 0 eV mol1, respectively [44].
3.3.3. Local reactivity of the two benzothiazine derivatives
The DN values are correlated to the inhibition efficiency resulting
The local reactivity of the studied inhibitors is investigated
from electron donation. According to Lukovits et al. study [45], if
using the condensed Fukui indices (FI) [46] which are an indication
DN < 3.6, the inhibition efficiency increases with increasing elec-
of the reactive centers of molecules (nucleophilic and electrophilic
tron-donating ability at the metal surface. The obtained values of
centers). Therefore, the molecule regions where the Fukui function
DN reported in Table 5; show that the P2 have the highest value
is large are chemically softer than the regions where the Fukui
of DN (0.693) in aqueous phase. Therefore, the highest inhibition
function is small. The condensed Fukui functions [47] are found
efficiency obtained experimentally for P1 can be explained by the
by taking the finite difference of approximated atomic charges,
tendency of molecule to receive the electron by S atom in unoccu-
computed by natural population analysis (NPA) which was carried
pied orbital (3d). This ability to receive the electron from the
out in terms of localized electron-pair ‘bonding’ units [46].
metallic surface increases the inhibition efficiency.
Natural bond orbital (NBO) analysis [46] was performed to eval-
The calculated quantum chemical parameters in the presence of
uate the electron-density distributions. Thus, the electron density
solvent (water) and gas phase do not exhibit important differences
plays an important role in calculating the chemical reactivity

Table 5
Quantum chemical parameters of P1 and P2 calculated at B3LYP/6-311++G (2d, 2p) in gas (G) and aqueous (A) phases.

Phase EHOMO (eV) ELUMO (eV) DE (eV) l (D) A (eV) I (eV) v (eV) g (eV) DN gi (%)
P1 G 6.299 1.435 4.864 4.4557 1.435 6.299 3.867 2.432 0.644 96
A 6.266 1.425 4.841 6.1983 1.425 6.266 3.846 2.421 0.651
P2 G 6.056 1.155 4.901 3.8752 1.155 6.056 3.606 2.45 0.693 92
A 6.119 1.379 4.740 5.4420 1.379 6.119 3.749 2.37 0.690
 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324 323

Fig. 7. Optimized molecular structures (left), HOMO (center) and LUMO (right) distribution for P1 and P2.

parameters. For a neutral system of N electrons, independent sin- Pk ðN þ 1Þ  Pk ðN  1Þ

fk0 ¼ ðFor radical attackÞ ð11Þ
gle-point calculations were made at the same level of theory for 2
corresponding N + 1 and N  1 electron systems. In order to char- The calculated Fukui indices for all the charged species (N + 1
acterize the electronic population and the capacity of charge trans- and N  1) as well as the neutrals (N) of the two benzothiazine
fer on each nuclear center, natural population analysis (NPA) [48] derivatives (P1 and P2) are presented in Table 6. For simplicity,
were done. only the charges and Fukui functions on the nitrogen (N), oxygen
fkþ ¼ Pk ðN þ 1Þ  P k ðNÞ ðFor nucleophilic attackÞ ð9Þ (O), sulfur (S) and some carbon atoms are presented. The local
reactivity is analyzed by means of the condensed Fukui functions
fk ¼ Pk ðNÞ  Pk ðN  1Þ ðFor electrophilic attackÞ ð10Þ which allow distinguishing each part of the molecule on the basis
of its distinct chemical behavior due to the substituted functional

Table 6
Pertinent natural population and Fukui functions of P2 and P1 calculated at B3LYP/6-311++G (2d, 2p) in gas (G) and aqueous (A) phases.

Inhibitor Atom Phase P (N) P (N + 1) P (N  1) fkþ fk fk0

P1 C3 G 5.862 6.002 5.848 0.140 0.014 0.077

A 5.866 6.014 5.820 0.147 0.046 0.097
C4 G 6.213 6.166 6.045 0.048 0.168 0.060
A 6.215 6.177 6.022 0.038 0.193 0.078
C12 G 6.403 6.242 6.209 0.161 0.195 0.017
A 6.406 6.255 6.210 0.151 0.196 0.022
S13 G 15.698 15.972 15.575 0.275 0.123 0.199
A 15.711 16.016 15.486 0.305 0.225 0.265
N14 G 7.607 7.292 7.231 0.315 0.376 0.031
A 7.594 7.286 7.220 0.308 0.374 0.033
C24 G 5.305 5.722 5.666 0.417 0.361 0.028
A 5.295 5.766 5.651 0.472 0.356 0.058
O25 G 8.600 8.372 8.229 0.228 0.371 0.072
A 8.655 8.420 8.254 0.235 0.402 0.083
Cl28 G 16.993 17.027 16.921 0.034 0.072 0.053
A 17.007 17.020 16.995 0.012 0.012 0.013
P2 C3 G 6.224 6.122 6.004 0.102 0.220 0.059
A 6.229 6.132 5.988 0.097 0.241 0.072
C4 G 5.850 5.933 5.841 0.083 0.009 0.046
A 5.843 5.954 5.831 0.111 0.012 0.062
C11 G 6.110 6.057 6.063 0.052 0.047 0.003
A 6.113 6.049 6.066 0.063 0.046 0.009
C21 G 5.574 5.923 5.799 0.349 0.225 0.062
A 5.567 6.192 5.784 0.625 0.216 0.204
O22 G 8.540 8.359 8.211 0.182 0.330 0.074
A 8.590 8.527 8.254 0.063 0.336 0.137
N25 G 7.627 7.326 7.114 0.301 0.513 0.106
A 7.634 7.336 7.051 0.298 0.583 0.142
S26 G 15.704 15.873 15.666 0.170 0.038 0.104
A 15.700 15.912 15.713 0.212 0.013 0.099
C28 G 6.588 6.304 6.298 0.284 0.290 0.003
A 6.591 6.295 6.296 0.296 0.295 0.001
324 T. Ghailane et al. / Corrosion Science 76 (2013) 317–324

groups and the position of the ANH group in thiazine ring. Thus, 5. The reactive site for electrophilic and nucleophilic attacks,
the site for nucleophilic attack will be the place where the fkþ value determined from the Fukui indices, show that the adsorption
is a maximum. In turn, the site for electrophilic attack was con- is more important for the P1 compound.
trolled by the fk value. Therefore, the favored sites for electrophilic
attack and the most reactive of P1 are C4, C12 atoms of the thiazine
ring and N14, S13 and O25 heteroatoms. For P2, the most reactive References
sites are C3, C28, N25 and O22.
Thus, for the nucleophilic attack the most reactive site of P1 are [1] D.K. Yadav, M.A. Quraishi, B. Maiti, Corros. Sci. 50 (2012) 254–266.
[2] Aysße Ongun Yüce, Gülfeza Kardasß, Corros. Sci. 58 (2012) 86–94.
C3, C24 and S13 atoms while for P2 the sites which can accepted [3] H. Zarrok, A. Zarrouk, R. Salghi, M. Ebn Touhami, H. Oudda, B. Hammouti, R.
the electron are C21 and S26. Touir, F. Bentiss, S.S. Al-Deyab, Int. J. Electrochem. Sci. 8 (2013) 6014–6032.
Based on the discussion above, it can be concluded that P1 and P2 [4] M.J. Bahrami, S.M.A. Hosseini, P. Pilvar, Corros. Sci. 52 (2010) 2793–2803.
[5] F. Zhang, Y. Tang, Z. Cao, W. Jing, Z. Wu, Y. Chen, Corros. Sci. 61 (2012) 1–9.
have many active centers for adsorption on mild steel surface. Thus, [6] A. Döner, R. Solmaz, M. Özcan, G. Kardasß, Corros. Sci. 53 (2011) 2902–2913.
the areas containing N, O and S atoms are the most favored sites for [7] E. Jamalizadeh, A.H. Jafari, S.M.A. Hosseini, J. Mol. Struct. (Theochem) 870
bonding to mild steel surface through donating electrons. However, (2008) 23–30.
[8] Z. El Adnani, M. Mcharfi, M. Sfaira, M. Benzakour, A.T. Benjelloun, M. Ebn
the S atom can give and receive the electron to and from metal, Touhami, Corros. Sci. 68 (2013) 223–230.
respectively. The last process reinforces the inhibitor molecule [9] D. Wang, S. Li, Y. Ying, M. Wang, H. Xiao, Z. Chen, Corros. Sci. 41 (1999) 1911–
adsorption on metallic surface and the both process are more accen- 1919.
[10] B. Donnelly, T.C. Downie, R. Grzeskowiak, H.R. Hamburg, D. Short, Corros. Sci.
tuated in P1 than in P2. This result agrees with the experimental
18 (1977) 109–116.
data. [11] I.B. Obot, N.O. Obi-Egbedi, S.A. Umoren, Corros. Sci. 51 (2009) 276–282.
Taking into account the solvent effect, the Fukui indices for the [12] A.B. Tadros, B.A. Abd El-Nabey, J. Electroanal. Chem. 246 (1988) 433–439.
[13] T. Ghailane, M. Saadouni, S. Boukhris, N. Habbadi, A. Hassikou, A. Kerbal, B.
nucleophilic attack increases for P1 on C24 directly bonded to the
Garrigues, A. Souizi, Heterocycles 83 (2011) 357–363.
oxygen and sulfur atoms. The Fukui function values fkþ of C24 site is [14] A. Bouckamp, Users Manual Equivalent Circuit, Ver. 4.51, 1993.
about 0.417 and 0.472 in gas and aqueous phases, respectively and [15] A.D. Becke, J. Chem. Phys. 96 (1992) 9489–9495.
of the sulfur atom is about 0.275 and 0.305 in gas and aqueous [16] A.D. Becke, J. Chem. Phys. 98 (1993) 1372–1377.
[17] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
phases, respectively. For P2 compound, the effect solvent on the [18] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
Fukui function fkþ is more pronounced on C21 bonded to sulfur J.A. Montgomery, Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
atom S. Their values on C21 and on S are 0.349, 0.625, 0.170 and J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
0.212 in gas and aqueous phases, respectively. The electron dona- M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox,
tion from metallic surface to inhibitor molecule is favored on S and H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann,
C atoms bonded to oxygen atom in aqueous phase for both P1 and O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K.
Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S.
P2. The Fukui function values fkþ for electrophilic attack of P1 are Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
stronger in solution than in gas phase for C4, O and S atoms. In Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J.
solution, this function is larger for C3 and N sites and decreases Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M.
for S atom. From this analysis it is obvious that, in the aqueous
Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A.
solution, the S atom is more likely to be a site for electrophilic at- Pople, Gaussian 03, Revision B.01, Gaussian, Inc., Pittsburgh, PA, 2003.
tack in P1 than in P2 and this leads to easy adsorption by donating [19] M.A. Shaker, H.H. Abdel-Rahman, Am. J. Appl. Sci. 4 (2007) 554–564.
[20] M.A. Shaker, M.A. Khalifa, Alex. J. Pharm. Sci. 9 (1995) 159–161.
electron from P1 to metallic surface.
[21] M.S. Masoud, A.E. Ali, M.A. Shaker, M. Abdul Ghani, Spectrochim. Acta A 60
(2004) 3155–3159.
4. Conclusion [22] M.S. Masoud, A.E. Ali, M.A. Shaker, M. Abdul Ghani, Spectrochim. Acta A 61
(2005) 3102–3107.
[23] E. Cancès, B. Mennucci, J. Tomasi, J. Chem. Phys. 107 (1997) 3032–3042.
The inhibition efficiency of mild steel corrosion in 1 M HCl by [24] W.J. Hehre, L. Radom, P.v.R. Schleyer, A.J. Pople, Ab Initio Molecular Orbital
benzothiazine derivatives has been investigated using electro- Theory, Wiley-Interscience, New York, 1986.
[25] J.F. Janak, Phys. Rev. B 18 (1978) 7165–7168.
chemical measurement and quantum chemical calculations at [26] R. Stowasser, R. Hoffmann, J. Am. Chem. Soc. 121 (1999) 3414–3420.
(DFT/B3LYP/6-311++G (2d, 2p)) level of theory. The following con- [27] R.G. Pearson, Inorg. Chem. 27 (1988) 734–740.
clusions were drawn from this study: [28] R.S. Chaudhary, S. Sharma, Indian J. Chem. Technol. 6 (1999) 202–206.
[29] F. Bentiss, M. Traisnel, M. Lagrenée, Appl. Surf. Sci. 161 (2000) 196–202.
[30] F. Mansfeld, Corrosion 36 (1981) 301–307.
1. Reasonably good agreement was observed between potentiody- [31] R. Touir, M. Cenoui, M. El Bakri, M. Ebn Touhami, Corros. Sci. 50 (2008) 1530–
namic polarization and electrochemical impedance spectros- 1537.
copy techniques. [32] K.S. Khairou, A. El-Sayed, J. Appl. Polym. Sci. 88 (2003) 866–871.
[33] C.H. Hsu, F. Mansfeld, Corrosion 57 (2001) 747–748.
2. The geometrical parameters show that the P2 molecule is more [34] P. Zhao, Q. Liang, Y. Li, Appl. Surf. Sci. 252 (2005) 1596–1607.
distorted than P1, thus the last one is the most efficient for the [35] J. Fang, J. Li, J. Mol. Struct. (Theochem) 593 (2002) 179–185.
corrosion inhibition. This result is in good agreement with [36] G. Bereket, E. Hür, C. Öğretir, J. Mol. Struct. (Theochem) 578 (2002) 79–88.
[37] A.Y. Musa, A.B. Mohamad, A.A.H. Kadhum, M.S. Takriff, W. Ahmoda, J. Indus.
experiment results. Eng. Chem. 18 (2012) 551–555.
3. A good correlation was found between the quantum chemical [38] F.K. Khaled, Appl. Surf. Sci. 256 (2010) 6753–6763.
parameters (ELUMO, gap energy (DE) and dipole moment (l)) [39] A. Popova, M. Christov, T. Deligeorigiev, Corrosion 59 (2003) 756–764.
[40] M. Sahin, G. Gece, F. Karci, S. Bilgic, J. Appl. Electrochem. 38 (2008) 809–815.
and inhibition efficiency. However, the fraction of electron [41] M.A. Quraishi, R. Sardar, J. Appl. Electrochem. 33 (2003) 1163–1168.
transfer (DN) and ELUMO decrease for P1 which is the most effi- [42] K.F. Khaled, N.K. Babic-Samardzija, N. Hackerman, Electrochim. Acta 50 (2005)
cient experimentally. This can be explained by the fact that P1 2515–2520.
[43] K. Ramji, D.R. Cairns, S. Rajeswari, Appl. Surf. Sci. 254 (2008) 4483–4493.
compound can give and receive electrons from the metal and [44] I.B. Obot, N.O. Obi-Egbedi, Corros. Sci. 52 (2010) 657–660.
these processes reinforce the adsorption of P1 molecules on [45] I. Lukovits, E. Kalman, F. Zucchi, Corrosion 57 (2001) 3–9.
metallic surface. [46] R.G. Parr, W. Yang, J. Am. Chem. Soc. 106 (1984) 4049–4050.
[47] W. Yang, W.J. Mortier, J. Am. Chem. Soc. 108 (1986) 5708–5711.
4. The density distributions of the frontier molecular orbitals
[48] A.E. Reed, L.A. Curtiss, F. Weinhold, Chem. Rev. 88 (1988) 899–926.
(HOMO and LUMO) show that P1 and P2 adsorb through the
actives centers nitrogen, oxygen and p electrons of the benzo-
thiazine ring.