Study of the inhibiting effect of a quaternary ammonium surfactants mixture synthesized from petroleum fraction (reformate) against the carbon steel corrosion in HCl 1 M
Houria Hamitouche•Abdellah Khelifa• Amel Kouache•Saaˆd Moulay
Received: 7 December 2012 / Accepted: 1 March 2013 ÓSpringer Science+Business Media Dordrecht 2013
Abstract Quaternary ammonium cationic surfactants were synthesized from reformate, a liquid mixture of hydrocarbons (aromatics, naphthenes and paraffins), via chloromethylation/quaternization sequences. The petroleum surfactants thus obtained were evaluated as corrosion inhibitors for carbon steel in 1 M HCl, by gravimetry, potentiodynamic polarization and electrochemical impedance spec- troscopy. The corrosion inhibiting efficiency was assessed as functions of surfactant concentration. The results showed that the inhibiting efficiency increased with surfactant concentration; its optimal value of 70 % was for a surfactant concen- tration of 320 mg/L at 25°C. Potentiodynamic polarization measurements showed that the mixture acts as a mixed type inhibitor. The corrosion inhibiting mechanism is thought to proceed via an adsorption of the surfactant molecules on the steel surface, generating a film and hindering the active sites. Our experimental adsorption data were found to obey the Langmuir adsorption isotherm. SEM images of the treated specimens, revealing the likely formation of a protective film, dem- onstrated the inhibiting capacity of the petroleum quaternary ammonium surfactants against the carbon steel corrosion.
Keywords Corrosion inhibition Carbon steel Petroleum quaternary ammoniumCationic surfactants mixture
H. Hamitouche (&)A. KhelifaA. Kouache
Laboratoire de Ge´nie chimique, De´partement de chimie industrielle, Faculte´ de Technologie, Universite´ Saaˆd Dahlab de Blida, BP 270, Route de Soumaaˆ, 09000 Blida, Algeria e-mail: hamitouchehouria@yahoo.fr
S. Moulay
Laboratoire de Chimie-Physique Mole´culaire et Macromole´culaire, De´partement de chimie industrielle, Faculte´ de Technologie, Universite´ Saaˆd Dahlab de Blida, BP 270, Route de Soumaaˆ, 09000 Blida, Algeria
DOI 10.1007/s11164-013-1133-0
Introduction
Because of the huge economic losses as a result of acid corrosion of steel employed in numerous industrial applications, preventative strategies to minimize this deleterious facet are unequivocally necessary. Of the various means for inhibiting the corrosion, the use of organic surfactants remains the most practically efficient [1]. A special interest has been devoted to the study of the corrosion inhibition at diverse metal/electrolyte interfaces [2–7] by using organic surface-active agents because of their many inherent advantages. Indeed, the surfactants are attractive mostly for their hydrophobic–hydrophilic structure, relatively high molecular weights, easy production, moderate cost, and low toxicity. Surfactant molecules physically prevent the corrosion of metals by adsorption onto their surfaces; the corrosion speed is thereby significantly reduced [8–11]. Their activity as corrosion inhibitors generally increased with their concentrations and was optimal near critical micellar concentrations (CMC) [12–14]. It is customary to combine the surfactant with an additive for enhancing the inhibiting efficiency, when both the surfactant and the additive are not efficient individually [15].
Quaternary ammonium-based cationic surfactants are well known as excellent inhibitors for iron dissolution and for the steel corrosion in acidic media [5,16–18], in various industrial processes [17]. Different quaternary ammonium salts have been studied, such as cetylpyridinium chloride (CPC) [19], 3-methylbenzo [d]thiazol-3-ium bromide [20], decyldimethylisopropylammoniumhydroxide [21], cetyl trimethyl ammonium bromide [16], 1,2-ethane bis(dimethyl alkyl (CnH2n?1) ammonium bromide) (n=10, 12 and 16) [22], and diquaternary ammonium surfactant [23]. It has been proposed that quaternary ammonium salts act as inhibitors by adsorption on the metal surface, and the adsorption takes place through electrostatic attraction between positively charged-N?ion and the induced negative charges on the metal surface. Zhang et al. [24] reported that the negative charge on a solid surface could be increased by inorganic anions, and this will help cationic organic molecules to be easily adsorbed on the surface of the solid. However, the inhibition mechanism of the quaternary ammonium inhibitors for iron and steel corrosion in acidic media has still not been completely understood [25]. The main advantage of quaternary ammonium salts, besides the include N-atom as the adsorption center, is their solubility in water [26] and their stability in acid [27]. They can be used alone or in conjunction with inorganic anions ions such as halide ions [28] or organic compounds such as 2-mercaptobenzoxazole [29]. There are reports [28,30–33] that put forward the synergetic effect of combined surfactants of different natures in the corrosion inhibition of copper, brass, carbon steel, and stainless steel in different corrosive media.
Making surface-active agents from petroleum fractions to produce mixtures of surfactants is indubitably of paramount importance as no prior separation of individual components is required. Such mixture can be used originally and advantageously as corrosion inhibitors against the corrosion of steel in acidic media.
We wish to report here the synthesis of petroleum quaternary ammonium surfactants through chloromethylation/quaternization of the reformate fraction, RE, and the results of their inhibiting effects on the corrosion of carbon steel in acidic
condition. The investigation was performed by means of weight loss method, potentiodynamic polarization and electrochemical impedance spectroscopy (EIS).
Experimental
Materials and equipment
Reformate (RE), a petroleum fraction obtained by catalytic reforming of Naphtha, was supplied from Algiers’ refinery. It is a liquid mixture of hydrocarbons composed of aromatics (%A), naphthenics (%N), and paraffinics (%P). Its composition of these hydrocarbons was determined using the rda method of [r,d, and a stand for refractive index (nD20), density (d420), and aniline point (a.p.)] by means of Eqs.1,2and3[34,35], and the results are shown in Table1.
%A¼1039:4n20
D 470:4d20
4 0:315 a:p:1094:3 ð1Þ
%N¼ 1573:3n20
D þ840:15d20
4 0:4619 a:p:þ166:2 ð2Þ
%P¼100ð%Aþ%NÞ ð3Þ The chemicals involved in the syntheses were purchased from Fluka, Panreac Quimica or Cheminova, and were used without prior purification. The reactants and the products of the syntheses were analyzed by FT-IR (Shimadzu type FTIR 8900) and UV–visible (Shimadzu 1700). For FT-IR analysis, the solid sample was mixed with KBr to make a pellet, and drops of the liquid sample were spread between two plates of potassium bromide. Ethanol and water were the solvents for UV–visible measurements. Melting points were measured with Boe¨tius apparatus.
The working electrode was made of carbon steel (CS). Its chemical composition was as follows: 0.176 % C, 1.5 % Mn, 0.32 % Si, 0.043 % S, 0.09 % P, 0.042 % Cu, 0.018 % Nb, 0.012 % Cr,\0.02 % Ni, and Fe.
For every experimental trial, a polishing pretreatment of the electrode surface was performed with silicon carbide grit of increasing grain size from 280 to 1,200 mesh, followed by ethanolic degreasing, and finally rinsing with plenty of distilled water.
The concentrations of inhibitors were screened from 80 to 480 mg/L.
Synthesis of reformate quaternary ammoniums
Into a 250-mL three-necked and round-bottomed flask equipped with a condenser, a thermometer, and a magnetic stirrer, 50 mL of reformate, 4 g of paraformaldehyde, 1 mL of glacial acetic acid, 15 mL of HCl (37 %), and 5 g of ZnCl2were charged.
Then, the mixture has been bubbled with hydrogen chloride, while stirring, for
Table 1 Physico-chemical properties and hydrocarbons composition of reformate d4
20 nD
20 a.p. (°C) A (%) N (%) P (%)
0.7539 1.4430 29.20 42.03 11.36 46.61
about 2 h at 50°C. A volume of 30 mL of CRE was isolated and a drop of it was used for analysis.
Then, a mixture of 10 mL of chloromethylated petroleum fraction and 20 mL of triethylamine in 10 mL of DMF was heated at 55°C for 1 h. Afterwards, a vacuum distillation was applied to the mixture, and the remaining product was recrystallized from acetone to afford a white powder. The solid product was dried at 50°C for 3 days and weighed, finally, 6 g.
Corrosion measurement techniques Weight loss measurements
A rectangular piece of carbon steel (291.590.3 cm3) was immersed into 60 mL of corrosive solution (1 M HCl) in a beaker. The beaker was placed in a thermostated water bath. The weight loss of the steel specimen was measured at different inhibitor concentrations.
Electrochemical measurements
All electrochemical experiments were realized in a single-compartment three- electrode and double-walled Pyrex-made cell. The working electrode was made of carbon steel and was coated with methyl methacrylate-based resin; its exposed surface area to the solution was 0.44 cm2. A platinum grid electrode (494 cm2) was used as an auxiliary electrode, and saturated calomel electrode SCE (?0.24 V/ESH) as reference electrode. The three electrodes were connected to the system EG&G composed of a potensiostat/galvanostat (Princeton Applied Research 273A) and a transfer function analyzer (model 5210). The whole system was connected to a personal computer, and the measurements were computed from two programs:
SoftCorr III for polarization curve plotting, and PowerSuite for impedance diagrams.
Potentiodynamic polarization
After immersing for 1 h, the potentiodynamic anodic and cathodic polarization curves were recorded in the potential range of-250 to?250 mV with respect to the corrosion potential at a sweep rate of 0.5 mV/s.
Electrochemical impedance spectroscopy
The impedance diagrams were plotted in potentiostatic mode at the corrosion potential in the frequency interval of 100 kHz to 350 MHz, with a sinusoidal potential perturbation of 5 mV and ten data points per decade were recorded.
Inhibition efficiency
The inhibition efficiency of the petroleum surfactant (IE %) was calculated using Eq. (4) [36]:
IE%¼½1X=X0100 ð4Þ whereX0andXare either the corrosion rate (from weight loss measurements), or the corrosion current density (from polarization curves) or the charge transfer resistance (EIS), in the absence and presence of the inhibitor, respectively.
SEM visualization
The morphology of the surface of carbon steel specimen was visualized after its immersion in the corrosive solution, in the absence and presence of the inhibitor, for 24 h using scanning electron microscopy (JEOL 5300).
Results and discussion
Synthesis of the mixture of reformate-based petroleum quaternary ammonium surfactants
Reformate with a composition in hydrocarbons shown in Table1was subjected to conventional chloromethylation as traced in a representative Eq. (5) [37]; the aromatic fraction underwent transformation under the reaction conditions to afford chloromethylated crude product (CRE). The in situ quaternization was effected with triethylamine in N,N-dimethylformamide to yield white recrystallizable crystals, melting at 216–232°C, and readily soluble in water; their solubility in water was determined to be 21.81 g/L at 25°C. The actual product (reformate quaternary ammonium salt, QRE) is a mixture with different alkyl chains and different aryl groups.
ð5Þ
Figure1illustrates the FT-IR spectra of RE, CRE and QRE. The new bands at 808.02 and 1,263.14 cm-1in the spectrum of CRE (Fig.1b) are characteristic bands of the chloromethyl group (–CH2Cl) and proved its attachment to the aryl group [38, 39]. The success of the quaternization is confirmed by the appearance three bands at 1,033.87, 1,068.59, and 1,168.90 (Fig.1c), which are located within 900–1,300 cm-1 [40]. The band at about 260 nm in the UV–visible spectra of RE, CRE and QRE shown in Fig.2clearly indicates the presence of aromatics.
Inhibition efficiency determination Weight loss measurements
Table2 gathers the results of the inhibition efficiency (IE %) estimated by the weight loss measurements under conditions cited above. As can be seen, IE % varied with QRE concentration. It increased with increasing QRE concentration and reached a maximum value of 70 % at the concentration of 320 mg/L. Beyond this concentration, IE % remained almost unchanged. The corrosion rate, however, decreased with QRE concentration up to 0.710 mg/cm2h for the concentration of 320 mg/L, and was nearly unvaried afterwards. The observed corrosion inhibition was probably a result of the adsorption of QRE molecules onto the carbon steel surface, and their plugging capacity of the dissolution sites of the metal by forming a protective film [41]. At concentrationzs higher than 320 mg/L, the efficiency seems to approach a constant value. This means that the critical micellar
Fig. 1 FT-IR spectra:aRE (liquid form),bCRE (liquid form),cQRE (solid form)
concentration has been attained. The CMC value of QRE was determined in corrosive solution at 25°C using the conductivity method, being around 350 mg/L.
Electrochemical measurements
Open circuit potential (OCP) The variations of the corrosion potential (Ecorr) as functions of time and QRE concentration, in the absence and presence of QRE, are presented in Fig.3. It can be noticed that the potential in the presence of inhibitor was higher that that in the absence, and rose with time. No systematic dependence of the potential on the QRE concentration was observed; yet, a stable value of -507 mV/SCE was attained after 30 min of immersion. Ecorr in the absence of inhibitor was-507 mV/SCE after this time. Thus, it appears that the QRE did not appreciably affect the corrosion potential.
Electrochemical impedance spectroscopy (EIS) measurements Figure4shows the Nyquist impedance diagrams at the corrosion potential after 1 h-immersion. The
Fig. 2 UV spectra:aRE, in ethanol,bCRE, in ethanol,cQRE, in water
Table 2 Corrosion rate and inhibition efficiency obtained by weight loss measurements for different concentrations of QRE in 1 M HCl solution at 25°C
[QRE] (mg/L) Vcorr(mg/cm2h) IE (%)
Blank 2.44 –
80 1.29 47.13
160 1.10 54.92
240 0.90 63.11
320 0.710 70.90
400 0.759 68.89
480 0.735 69.88
curves were half-loop-shaped. The equivalent electrical circuit that may account for such behavior is represented in Fig.5 [42]. From the impedance diagrams, the electrochemical parameters, such as IE %, charge transfer resistance Rt, and capacitance of the double layer (CPE), can be calculated and are compiled in Table3. The inhibiting effect of QRE toward the corrosion of carbon steel in 1 M HCl is quantified by Rt increase and CPE decrease with increasing QRE concentration. As reported [43, 44], these findings are merely related to the QRE adsorption on the steel surface.
Potentiodynamic polarization measurements Figure6illustrates the cathodic and anodic Tafel polarization curves of the carbon steel in 1 M HCl at 25°C, in the absence and presence of QRE. The different electrochemical parameters, that is corrosion current density (icorr), cathodic and anodic slopes (bc and ba), and inhibition efficiency (IE %), that can be deduced from these curves, are gathered in Table4. The decline of icorr in the presence of QRE, and with increasing QRE concentration up to a certain value, clearly demonstrates the inhibiting capacity of the latter. However, the cathodic and anodic Tafel slopes remained almost unaffected upon addition of this inhibitor. The adsorption phenomenon that occurred during the corrosion course would explain these results, and the mechanism of the corrosion reactions could be about the same [23].
For a matter of comparison, the variations of inhibition efficiency (IE %) as a function of QRE concentration from the different measurements are plotted in Fig.7. As evidenced, all measurements were in a good agreement with slight deviations.
-540 -530 -520 -510 -500
0 600 1200 1800 2400 3000 3600
OCP potential (mV/SCE)
Time (s)
Blanc 80 mg/L 160 mg/L 240 mg/L 320 mg/L
Fig. 3 Open circuit potential-time curves for carbon steel electrode immersed in 1 M HCl solution in the absence and presence of various QRE concentrations at 25°C
0 50 100 150 200 250
0 50 100 150 200 250
-Zim (Ω.cm2)
Zre (Ω.cm2) Blanc
80 mg/L 160 mg/L 240 mg/L 320 mg/L
Fig. 4 Nyquist plots for carbon steel in 1 M HCl solution in the absence and presence of various QRE concentrations at 25°C
Fig. 5 Equivalent circuit used to simulate the EIS diagram
Table 3 Electrochemical impedance parameters and the corresponding inhibition efficiencies for carbon steel in 1 M HCl solution in the absence and presence of different QRE concentrations at 25°C
[QRE] (mg/L) CPE (lF/cm2) Rt(Xcm2) IE (%)
Blank 85.79 73.85 –
80 57.03 151.9 51.38
160 46.82 171.1 56.84
240 43.08 194.4 62.01
320 42.91 210.7 64.95
400 43.28 202.4 63.51
480 42.96 209.2 64.7
Adsorption isotherm
As explained above, the inhibition power of QRE is related mainly to its adsorption capacity onto the carbon steel surface. The commensurate adsorption mode and its model isotherm would provide an insight into the interactions types between adsorption inhibitor molecules and the steel surface. The value of the metal surface coverage extent (h), calculated ash=IE %/100 and computed from the weight loss measurements, was used for establishing the corresponding adsorption isotherm.
Figure8 depicts the variations of Cinh/h as a function of Cinh (the inhibiting concentration) in the range of the concentrations studied. A linear dependence was
Table 4 Potentiodynamic polarization parameters and the corresponding inhibition efficiencies for carbon steel in 1 M HCl solution in the absence and presence of different QRE concentrations, at 25°C [QRE] (mg/L) -Ecorr(mV/ECS) icorr(lA/cm2) -bc(mV/dec) ba(mV/dec) IE (%)
Blank -472.5 160.4 113.6 84.0 –
80 -486.3 82.89 104.0 83.34 48.32
160 -485.5 68.91 98.82 83.65 57.04
240 -499.0 56.39 109.3 85.74 64.84
320 -488.9 45.26 106.4 80.52 71.78
400 -488.1 46.87 100.0 76.23 70.78
480 -494.6 46.97 99.16 83.63 70.72
-6 -5 -4 -3 -2 -1
-800 -700 -600 -500 -400 -300 -200
log i (µA/cm2)
E (mV/SCE)
Blanc 80 mg/L 160 mg/L 240 mg/L 320 mg/L
Fig. 6 Potentiodynamic polarization curves of carbon steel in 1 M HCl solution in the absence and presence of various QRE concentrations at 25°C
found, with a correlation coefficient of 0.99 and a slope of 1.1, suggesting that the QRE adsorption obeyed the Langmuir adsorption isotherm governed by Eq.6[45]:
Cinh=h¼Cinhþ1=Kads ð6Þ
whereKadsis the adsorption–desorption equilibrium constant.
0 10 20 30 40 50 60 70 80
0 50 100 150 200 250 300 350
Inhibition efficiency IE%
C (mg/L)
Weight loss EIS Tafel
Fig. 7 Comparison between the inhibiting effectiveness determined by weight loss, EIS, and potentiodynamic techniques
0.1 0.3 0.5
0.05 0.15 0.25 0.35
C (mg/L)
Fig. 8 Langmuir adsorption plot of carbon steel in 1 M HCl solution containing various QRE concentrations at 25°C
Scanning electron microscopy (SEM)
SEM micrographs of carbon steel samples are shown in Fig.9. The image in Fig.9b indicates a strong corrosion attack after immersing the carbon steel sample into 1 M
Fig. 9 SEM micrographs of carbon steel samples:aafter polishing,bafter 24 h of immersion in 1 M HCl in the absence of inhibitor,cafter 24 h of immersion in 1 M HCl in the presence of 320 mg/L of QRE
HCl for 24 h and in the absence of QRE. Indeed, the high level of attack is revealed by the formation of deep cavities. Such a corrosion attack was prevented by the same solution and in the presence of 320 mg/L of QRE after the same immersion time (Fig.9c).
Conclusion
A number of concluding remarks can be drawn from this present study, which are:
1. It is possible to produce a cationic surfactant mixture of quaternary ammonium type from local petroleum product, reformate, which can be used, without prior separation, as an inhibitor for acidic steel corrosion.
2. The synthesized reformate-based petroleum quaternary ammonium salts are endowed with a great inhibiting efficiency of carbon steel corrosion in 1 M HCl solution.
3. All techniques employed in ascertaining the corrosion inhibition afforded inhibition efficiency of about the same order. The results were in good agreement with each other.
4. Potentiodynamic polarization measurements showed that the mixture acts as a mixed type inhibitor.
5. The inhibition mechanism is the adsorption of inhibitor molecules onto the metal surface. Experimental data obey the Langmuir adsorption isotherm.
References
1. M.A. Malik, M.A. Hashim, F. Nabi, S.A. AL-Thabaiti, Z. Khan, Int. J. Electrochem. Sci. 6, 1927–1948 (2011)
2. R.F.V. Villamil, P. Corio, S.M.L. Agostinho, J.C. Rubim, J. Electroanal. Chem.472, 112–119 (1999) 3. S.S. Abd El Rehim, H.H. Hassan, M.A. Amin, Mater. Chem. Phys.78, 337–348 (2002)
4. E.E. Foad El-Sherbini, S.M. Abdel Wahaab, M. Deyab, Mater. Chem. Phys.89, 183–191 (2005) 5. M.M. Osman, R.A. El-Ghazawy, A.M. Al-Sabagh, Mater. Chem. Phys.80, 55–62 (2003) 6. M. Knag, J. Sjo¨blom, G. Øye, E. Gulbrandsen, Colloids Surf A250, 269–278 (2004) 7. X. Li, G. Mu, Appl. Surf. Sci.252, 1254–1265 (2005)
8. L.-G. Qiu, A.-J. Xie, Y.-H. Shen, Mater. Chem. Phys.87, 237–240 (2004)
9. M.A. Migahed, M. Abd-El-Raouf, A.M. Al-Sabagh, H.M. Abd-El-Bary, Electrochim. Acta 50, 4683–4689 (2005)
10. L.-G. Qiu, A.-J. Xie, Y.-H. Shen, Mater. Chem. Phys.91, 269–273 (2005) 11. G. Mu, X. Li, J. Colloid Interf. Sci.289, 184–192 (2005)
12. R. Fuchs-Godec, Colloids Surf. A280, 130–139 (2006)
13. E.F. Moura, A.O.W. Neto, T.N.C. Dantas, H.S. Ju´nior, A. Gurgel, Colloids Surf. A340, 199–207 (2009)
14. Q. Zhang, Z. Gao, F. Xu, X. Zou, Colloids Surf. A380, 191–200 (2011)
15. S. Rajendran, S.M. Reenkala, N. Anthony, R. Ramaraj, Corros. Sci.44, 2243–2252 (2002) 16. Y. Soror, M.A. El-Ziady, Mater. Chem. Phys.77, 697–703 (2002)
17. L.-G. Qiu, Y. Wu, Y.-M. Wang, X. Jiang, Corros. Sci.50, 576–582 (2008) 18. A. Popova, M. Christov, A. Vasilev, Corros. Sci.53, 1770–1777 (2011) 19. A. Khamis, M.M. Saleh, M.I. Awad, Corros. Sci.66, 343–349 (2013)
20. A. Popova, M. Christov, A. Vasilev, A. Zwetanova, Corros. Sci.53, 679–686 (2011)
21. A.M. Badawi, M.A. Hegazy, A.A. El-Sawy, H.M. Ahmed, W.M. Kamel, Mater. Chem. Phys.124, 458–465 (2010)
22. L.G. Qiu, A.J. Xie, Y.H. Shen, Corros. Sci.47, 273–278 (2005)
23. N.A. Negm, A.M. Al Sabagh, M.A. Migahed, H.M. Abdel Bary, H.M. El Din, Corros. Sci. 52, 2122–2132 (2010)
24. Z.H. Zhang, Z. Sen, C.S. Shao, Chin. J. Inorg. Chem.15, 451 (1999)
25. L. Niu, H. Zhang, F. Wei, S. Wu, X. Cao, P. Liu, Appl. Surf. Sci.252, 1634–1642 (2005) 26. A. El-Shafei, M.N.H. Moussa, A.A. El-Far, Mater. Chem. Phys.70, 175–180 (2001) 27. M.A. Hegazy, M. Abdallah, H. Ahmed, Corros. Sci.52, 2897–2904 (2010) 28. M.A. Deyab, Corros. Sci.49, 2315–2328 (2007)
29. H. Tavakoli, T. Shahrabi, M.G. Hosseini, Mater. Chem. Phys.109, 281–286 (2008) 30. R. Fuchs-Godec, Electrochim. Acta54, 2171–2179 (2009)
31. L.B. Tang, X.M. Li, G.N. Mu, G.H. Liu, L. Li, H.C. Liu, Y.S. Si, J. Mater. Sci.41, 3063–3069 (2006) 32. G.N. Mu, X.M. Li, G.H. Liu, Corros. Sci.47, 1932–1952 (2005)
33. X.M. Li, L.B. Tang, Mater. Chem. Phys.90, 286–297 (2005) 34. L. Robert, Rev. I.F.P.7, 55–59 (1952)
35. L. Robert, ibid8, 586–590 (1953)
36. A. Khadraoui, A. Khelifa, Res. Chem. Intermed. (2012). doi:10.1007/s11164-012-0910-5 37. S. Moulay, Des. Monomers Polym.14, 179–220 (2011)
38. S. Moulay, N. Mehdi, Chin. J. Polym. Sci.24, 1–9 (2006)
39. V. Neagu, E.C. Avram, G. Lisa, React. Funct. Polym.70, 88–97 (2010) 40. B. Gao, X. Zhang, Y. Sheng, Mater. Chem. Phys.108, 375–381 (2008) 41. X. Li, S. Deng, H. Fu, Corros. Sci.52, 2786–2792 (2010)
42. R. Solmaz, E. Altunbas, G. Kardas, Mater. Chem. Phys.125, 796–801 (2011)
43. A. Khadraoui, A. Khelifa, H. Hamitouche, R. Mehdaoui, Res. Chem. Intermed (2013). doi:
10.1007/s11164-012-1014-y
44. H. Mirghasem, F.L. Stijn, M. Mertens, R. Arshadi, Corros. Sci.45, 1473–1489 (2003) 45. M.J. Bahrami, S.M.A. Hosseini, P. Pilvar, Corros. Sci.52, 2793–2803 (2010)
