4.1. Coating structure analysis
The coating network structures after film formation have been studied through FTIR and shown in Fig. 1. The presence of multiple peaks in the range of 1000-1130cm− 1 in the silane films (Fig. 1(a)-(d)) indicates, the formation of Si-O-Si bonds due to the curing of all the silane coatings[51, 52]. The peaks in the range of 900-1000cm− 1 indicate the formation of Si-O-Metal bond formation[52] which is a covalent interaction of the siloxane network to the metal substrate as shown in Fig. 1 (a),(b),(c) and (d).
Two peaks in the range of 2830-2950cm− 1 for the silane systems-1G,1M, and 1A (Fig. 1(a),(b),(c)) for both the sol and the film systems represent the -CH2 symmetric and asymmetric vibrations characteristic of the alkyl chains present in all these silanes except for TEOS[53].
The presence of a peak at 1255 cm− 1 in the film of 1G system (Fig. 1(a)) corresponds to an absorption peak of the Si-CH2 bond in GPTMS[54] whilst the shoulder peak at 915cm− 1 can be indicative of the C-O bond of the oxirane ring characteristic of GPTMS[55]. In the case of the 1M system (Fig. 1(b)), a weak peak at 2569 cm− 1 is observed in the sol but is reportedly missing in the film. This weak peak is representative of the thiol group (-SH) of MPTMS[56, 57]. Meanwhile for the 1A system (Fig. 1(c)), similar kinds of weak peaks are observed in the soil in the range of 1550–1600 cm− 1 which is found to be absent in the film. weak amino (-NH2) peak (1550–1600 cm− 1) of APTES[58, 59]. The absence of weak peaks in both the films (Fig. 1 (b), (c)) could indicate that the bonding to the metal substrate could be through these groups along with the silanol linkage. Moreover, the presence of the ‘S’ and ‘N’ atoms in their respective chains would enhance the chances of intramolecular hydrogen bonding and hence allow for a dense, compact siloxane network. In the case of TEOS-coated steel (Fig. 1d), the presence of only the Si-O-Si peak at 1074cm− 1 and Si-O-M peak at 940cm− 1 and the absence of any other characteristic peak indicates the formation of a siloxane network over the film.
4.2 Surface Morphology of Coating
The evaluation of the microstructure of four organo-silane coatings was conducted through a scanning electron microscope (SEM) using CB340-EDS, JEOL makes. Figure 3 represents the surface microstructure and cross-sectional appearance of four different coatings coated on galvanized IF steel substrate. Evidently, the fracture of the coating film was found for 1T whereas there is no such sign noticed for other coatings (Fig. 3(a-d)). Following the potentiodynamic polarisation study, this is evident for the disintegration of the protection layer for 1T and a path for corroding species to initiate corrosion. In addition to this, the average coating thickness found for 1G,1A,1M, and 1T was 13.01, 16.46, 21, and 24.01 ~ µm respectively.
4.3. Electrochemical results
4.3.1. Potentiodynamic polarisation studies
We have evaluated the electrochemical behavior of different coatings in a 3.5% NaCl solution by using the potentiodynamic polarization test as per ASTM G59-97(2014). The electrochemical behavior in terms of corrosion current density(icorr), corrosion potential (Ecorr), and anodic (βa) and cathodic (βc) Tafel slopes of the coated substrates was calculated from the potentiodynamic polarisation curves(Fig. 4) through Tafel extrapolation method[60, 61]. We have also used Gamry Echem Analyst software to evaluate the different parameters as shown in Table 2. The corrosion inhibition efficiency (η%)[62] which is defined in Eq. (8) has been calculated for all the samples to compare the corrosion resistance properties of the different silane-coated film to that of the uncoated steel.
$${\eta }\text{%}=\frac{{i}_{corr}^{o}-{i}_{corr}}{{i}_{corr}^{o}} X 100$$
8
where \({i}_{corr}^{o}\) and \({i}_{corr}\) represents the corrosion current density of the bare steel and silane-coated film, respectively.
It is observed from Fig. 4 that all the coated steel has similar anodic and cathodic reaction kinetics except 1T. 1T coating has shown a more -ve potential shift from bare substrates. All the coated steel shows higher cathodic Tafel slopes than the anodic slopes which can be attributed to the anodic controlled corrosion process. Bare steel shows an even greater cathodic Tafel slope than coated steel. It indicates that the surface gets oxidized as soon as it is exposed to electrolyte solutions. Therefore, all the samples show increased anodic dissolution behavior with little increase in the anodic polarization potential. The increase in the cathodic Tafel slope for the coated steel indicates a reduction in the kinetics of the cathodic reaction;
2HO + O + 4e → 4(OH) (9)
It is observed (Table 2) that corrosion potential (Ecorr) for 1A and 1M coated samples have been shifted to the positive potential with respect to bare steel but, for 1G and 1T, it shifted towards negative. The combined effect of anodic dissolution and suppression of cathodic reactions can explain the decrease in the corrosion current density [63, 64]. The decrease in the corrosion rate as a result of the change in the current density might also be due to the corrosion inhibition offered by the three-dimensional networked silane film[45] which is acting as a barrier protective layer against oxygen and chloride ions.
Table 2
Electrochemical parameters of uncoated and coated steel obtained through Tafel extrapolation
Samples | Ecorr (V/SCE) | icorr (µA/cm2) | βa (mV/dec) | -βc (mV/dec) | η (%) |
Bare GI | -0.892 | 6.01 | 57.3 | 1015 | |
1G | -0.886 | 0.313 | 70 | 264 | 94.79 |
1A | -0.922 | 0.339 | 75 | 141 | 94.36 |
1M | -0.905 | 0.25 | 62.4 | 222 | 95.84 |
1T | -0.980 | 2.19 | 53.04 | 632 | 63.56 |
Samples | Ecorr (V/SCE) | icorr (µA/cm2) | βa (mV/dec) | -βc (mV/dec) | η (%) |
Bare GI | -0.892 | 6.01 | 57.3 | 1015 | |
1G | -0.886 | 0.313 | 70 | 264 | 94.79 |
1A | -0.922 | 0.339 | 75 | 141 | 94.36 |
1M | -0.905 | 0.25 | 62.4 | 222 | 95.84 |
1T | -0.980 | 2.19 | 53.04 | 632 | 63.56 |
The trend in the corrosion inhibition efficiency values for the different organo-silanes formulation proves that the functional organo-silanes like 1G (GPTMS), 1M (MPTMS), and 1A (APTES) provide better corrosion protection to the galvanized IF steel than non-functional organo-silanes like 1T (TEOS). This phenomenon can be explained through the presence of electron-donating groups like -SH, -NH, and epoxy in the silane chemistry which allows for better adhesion through coordination bonding to the vacant d-orbital of metallic zinc coating on the steel[65, 66]. Moreover, these organo-silanes allow dense film formation by the networked siloxane structure which can be seen in microstructure images but in the case of 1T the coating film was found to be cracked (Fig. 3). Furthermore, in the case of functional silane coatings, this crosslinked dense film is developed in the process of intramolecular hydrogen bonding within the molecules due to the presence of hetero-atoms of different electronegativity values. Stronger hydrogen bonding is evident in epoxy-functional silane due to the presence of oxygen (O) linkage in the epoxy-functional silane (1G) compared to a weaker one in amino-functional silane (1A). Thiol-functional silane (1M) coating exhibited the best corrosion resistance property because of the combined effect of a strong hydrogen-bonded silane network coupled with its inherent corrosion inhibitive property which is further explained employing quantum chemical studies in subsequent sections. The mechanisms of interaction have been elucidated through DFT calculations and explained in subsequent sections.
4.3.2. Electrochemical impedance spectroscopy studies
EIS investigation was carried out to study the barrier properties and phenomena of electrochemical reaction at the coating film and metal interface in 3.5% NaCl aq. solution[67, 68]. The findings from this study were supporting the observations drawn from the potentiodynamic polarization and quantum chemical studies[14, 62, 69, 70]. The Nyquist plots shown in Fig. 5(a) indicate the presence of two capacitive loops which can be further confirmed through two-time constants (at a frequency range of 10− 1 – 1 Hz and another at 102- 103 Hz) which is observed in the corresponding Bode phase angle plot (Fig. 5(c)). The high-frequency spectra correspond to the barrier properties of the silane coating whereas the low-frequency spectra are the contribution of the corrosion process happening at the coating/metal interface. The Bode impedance spectra for the bare steel and the silane coatings can be further delineated into a predominantly resistive behavior at lower frequencies (10− 2 to 10− 1 Hz), capacitive behavior at intermediate frequencies (10− 1 to 103 Hz) followed by another zone of resistive behavior at higher frequencies (103 to 105 Hz). The resistive behavior at lower frequencies can be most prominently seen in the bare steel compared to the coated samples. We observed from the Nyquist plot that there is a remarkable increase in the radius of capacitive loops for the silane coatings on the galvanized IF steel sample with respect to the bare steel indicating higher protective efficiency of the coated film. It was shown that the different silane coatings film agrees well with the trend observed in the polarisation studies. The modulus of the impedance (|Z|) for the different samples (Fig. 5(b)) also corroborates the trend found in the polarisation curves. We could quantitatively evaluate the values of the electrochemical parameters from the impedance spectra to better explain the mechanisms of electrochemical reactions of the different coating films. The electrochemical values were obtained by fitting the impedance spectra using an equivalent electric circuit to emulate the phenomena occurring at the different interfaces (electrolyte/coating & coating/steel) as shown in Fig. 5(d).
The equivalent circuit consists of the following parameters: (i) electrolytic solution resistance RS, (ii) a silane coating layer having resistance Rc in parallel with constant-phase element CPEc and (iii) at low frequency, a constant-phase element (CPEdl) and charge transfer resistance (Rct) corresponding to the corrosion reaction[71–73]. Rc represents the resistance offered by the coating to the penetration of electrolyte and corrosive ions whereas the CPEc represents the dielectric impedance offered by the coating layer. The constant phase elements have been chosen to represent the physical inhomogeneities of the coatings and the impedance value of it is defined as[74]:
$${Z}_{CPE}= \frac{1}{{\left(j\omega \right)}^{n} C}$$
10
where, j=√-1, ω is the angular frequency and calculated as ω = 2πf rad s− 1, C is the capacitance and n - the exponent which lies in the range of 0 ≤ n ≤ 1. These constant phase elements were subsequently converted into their respective capacitive analogues using the following formulae[75]:
$$C= \frac{{(CPE \times R)}^{1/\beta }}{R}$$
11
Where, C is the capacitance value, CPE is the constant phase element value and R is the value of the resistance either in parallel or in series to the constant phase element and β is the index for CPE (m and n). Cc and Cdl are the capacitive values shown in Table 3 for their corresponding constant phase elements. From the fitted parameters, the total resistance offered by the silane coating is calculated as the summation of the charge-transfer (Rct) and the coating resistance (Rc).
Corrosion resistance efficiency (η%) was defined with respect to the impedance data to compare the corrosion resistance properties of the different silane-coated samples to that of the bare steel[62].
$${\eta }\text{\%}=\frac{{R}_{t}^{o}-{R}_{t}}{{R}_{t}^{o}} X 100 \left(12\right)$$
where \({R}_{t}^{o}\) and \({R}_{t}\) represents the resistance to corrosion offered by bare steel and silane-coated steel, respectively. The collated impedance parameters are shown in Table 3.
Table 3
Electrochemical Parameters for bare and silane-coated galvanized steels obtained from EIS spectra
Samples | Rs Ω.cm2 | Rc kΩ.cm2 | Cc µF.cm-2 | m | Rct kΩ.cm2 | Cdl µF.cm-2 | n | Rt kΩ.cm2 | η% |
Bare | 20.8 | 0.69 | 10.4 | 0.88 | 0.625 | 3.41 | 0.57 | 6.94 | - |
1G | 19.7 | 7.58 | 0.26 | 0.34 | 64.2 | 0.29 | 0.89 | 71.7 | 90.3 |
1A | 18.1 | 7.03 | 1.11 | 0.76 | 19.4 | 0.71 | 0.61 | 26.4 | 73.7 |
1M | 20.0 | 16.01 | 0.153 | 0.89 | 76.9 | 0.152 | 0.37 | 92.9 | 92.5 |
1T | 19.4 | 7.00 | 9.83 | 0.79 | 5.34 | 2.23 | 0.54 | 12.3 | 43.8 |
The value of Rt is an indication of the hindrance to the ingress of the corrosive ions to the metal surface. The trend of the Rt values for the functional organosilane coatings is in total agreement with the potentiodynamic studies. A lower value of Cc (higher capacitive impedance) would mean a dense film possessing low permittivity and hence, improved corrosion resistance property. The protective trends of these coating films are in the following orders 1M (0.15 µF.cm2) > 1G (0.26 µF.cm2) > 1A (1.11 µF.cm2) > 1T (9.83 µF.cm2). This indicates that the thiol-functionalized organosilane is providing good interfacial adhesion as compared to epoxy and amine-functionalized silanes. This can be explained by the logic that the atoms of higher electron negative value tend to form strong chemical bonding with the hydrogen atoms. However, oxygen-bearing epoxy functional groups are supposed to provide strong bonding as compared to thiol-bearing functional groups. Meanwhile, quantum chemical studies (explained in the next section) have further shown that the thiol is found to be the best adhesion promotor, in terms of electronic structures of organo-silane compounds and global reactivity parameters.
4.4. Quantum Chemical Studies
Optimized molecular structures with total energies and distributions of HOMO (Highest Occupied Molecular Orbital) and LUMO (Lowest Unoccupied Molecular Orbital) of the investigated molecules in the gas phase are shown in Fig. 6. Mulliken charges of atoms in four Silane molecules are summarized in Table 4, while quantum chemical parameters are summarized in Table 5.
According to frontier molecular orbital theory, EHOMO (energy of HOMO) represents the ability of a molecule to donate electrons to the empty molecular orbital of metal to facilitate the adsorption process. In contrast, ELUMO (energy of LUMO) represents the ability of a molecule to accept electrons which are related to electron affinity. Moreover, frontier orbitals can be used to predict the adsorption centers of coating molecules. In Fig. 6, the HOMO and LUMO of four different Silane molecules have different distribution centers that indicate that the functional group attached to the alkyl chain influences the overall reactivity of the four molecules. HOMOs of 1G, 1A, and 1M molecules are located on the end functional groups such as epoxy, amine, and thiol, respectively, indicating that adsorption on the metal surface occurs through end functional groups as these functional groups in organo-silane molecules can provide electron to form a chemical bond with the metal of substrate and it was experimentally proved by the infrared spectra[76]. For 1T, the HOMO is distributed in the Si-O bonds and a part of HOMO is in the alkyl chain. As sown in Fig. 6, the LUMO distribution of 1G and 1A are similar and is mainly distributed in the Si-O bond while the LUMO of 1T is distributed in the Si-O bond as well as in the alkyl chain. The charge density of the LUMO in 1M is in the Si-O bond and in the thiol group, indicating that thiol is a good electron acceptor too to form a back-donation interaction with metal that ultimately facilitates the adsorption of 1M molecule on the metal surface.
Generally, an atom with more negative charge in a molecule can provide electron to metal to form a coordinate bond and thus facilitates the adsorption process[77]. Such atoms with negative charges are named as active adsorption sites of a molecule. Mulliken atomic charges for important atoms in four organo-silane molecules are summarized in Table 4.
Table 4
Mullikena) charge (in e-) of atoms/groups in Silane molecules 1G-1T.
Molecules | Si | O1 | O2 | O3 | End atoms |
1G | + 1.3 | -0.8 | -0.8 | -0.8 | -0.3 (O) |
1A | + 1.3 | -0.8 | -0.8 | -0.8 | -0.8 (N) |
1M | + 1.3 | -0.8 | -0.8 | -0.8 | -0.1 (S) |
1T | + 1.4 | -0.6 | -0.6 | -0.6 | - |
a)DFT (B3PW91)/6-311G* calculations |
There are no differences in negative charges of O atoms directly connected with the Si center in 1G, 1A, and 1M, indicating that the reactivity of alkyl chains connected with the Si center is similar in 1G-1M although different functional groups are attached at the end. Differences in charges come at the end atoms connected with the alkyl chain for different functional groups in 1G-1M. Si center and O atoms attached to it become slightly more positive in 1T compared to those in the other three silane molecules. The atomic charges of two O end atoms in 1G are − 0.3 and those of N are − 0.8 in 1A and S is -0.1 in 1M, indicating that these end atoms with negative charges will facilitate the adsorption process in molecules 1G, 1A, and 1M.
4.4.1. Global Reactivity Parameters
A high value of EHOMO and a low value of ELUMO of a molecule, therefore, indicates a high inhibition performance of that molecule against metal corrosion[77]. Further, the gap between the HOMO and LUMO energy levels of a molecule (∆E = ELUMO-EHOMO) determines the chemical stability of the metal complex which measures the interaction between the adsorbed molecule and substrate surface. A molecule with a low ∆E value always provides better inhibition performance because the excitation energy to remove an electron from the last occupied orbital will be always low. A molecule with a low ∆E value is also more polarizable, which is usually associated with a high chemical reactivity and low kinetic stability. Such molecule is termed as soft molecule[78, 79]. In Table 5, the values of EHOMO in both gas and aqueous phases follow the order of 1M ≈ 1A > 1G > 1T while the values of ELUMO follow the order of 1M < 1G < 1A < < 1T. The trend of ∆E values in Table 5 is 1M < 1A < 1G < < 1T in both of gas and solution phases. Based on the above results and above explanations, we can write the corrosion inhibition efficiency (in both gas and aqueous phases) of four organo-silane molecules following the order of 1M > 1A > 1G > 1T. Note that corrosion inhibition efficiencies of all four molecules 1G-1T decrease in the aqueous phase compared to that in the gas phase due to higher ELUMO, lower EHOMO, and higher ∆E values in the aqueous phase than those in the gas phase.
Table 5
Quantum Chemical Parametersa) (in gas/water) for organosilane molecules.
Molecules | ELUMO (eV) | EHOMO (eV) | ∆E (eV) | ƞ (eV) | χ (eV) | ∆N | µ (D) |
1G | 0.5/0.8 | -7.2/-7.3 | 7.7/8.0 | 3.8/4.0 | 3.3/3.2 | 0.6/0.6 | 3.2/3.8 |
1A | 0.6/0.8 | -6.4/-6.6 | 7.0/7.4 | 3.5/3.7 | 2.9/2.9 | 0.8/0.7 | 2.3/3.0 |
1M | 0.4/0.4 | -6.4/-6.7 | 6.9/7.1 | 3.4/3.6 | 3.0/3.2 | 0.7/0.7 | 3.4/4.2 |
1T | 1.2/1.1 | -7.7/-7.9 | 8.9/9.0 | 4.4/4.5 | 3.2/3.4 | 0.6/0.5 | 0.02/0.02 |
a)DFT(B3PW91)/6-311G* calculations |
Global hardness ƞ (see Eq. 2) and global softness σ (see Eq. 6) measure the molecular stability and reactivity. A hard molecule has a large ∆E and a soft molecule has a small ∆E. Soft molecules are more reactive than hard ones because they could easily offer electrons to an acceptor. Here, the coating molecules are Lewis bases, while the metal is a Lewis acid. Bulk metals are soft acids and thus according to HSAB (Hard Soft Acid Base) theory[80], soft base molecules are most effective for corrosion inhibition of metals. Therefore, molecules with the lowest value of global hardness (highest value of global softness) will show the highest corrosion inhibition efficiency[81]. It is clear from Table 5 that the order of ƞ values in both gas and aqueous phases is 1M ≈ 1A < 1G < 1T while the order of σ values in both gas and aqueous phases is 1M ≈ 1A > 1G > 1T. All these results suggest that the sequence of corrosion inhibition efficiency for studied organosilane molecules is 1M ≈ 1A > 1G > 1T, indicating organosilanes with mercapto and amino functional groups are the best for corrosion protection of metal. Note that the inhibitive efficiency of molecules is lower in the aqueous phase than that in gas phase as all molecules have higher ƞ and lower σ values in the aqueous phase than those in gas phase.
In this work, the number of electrons transferred (∆N) between a metal substrate and four organosilane molecules was calculated using Eq. (5), and the results are summarized in Table 5. According to Sanderson’s electronegativity equalization principle[82], the charge transfer process between metal and coating molecule will continue until their electronegativity (χ) values are equal with each other and ∆N is considered as a derived descriptor from the electronegativity/hardness equalization principle. It was pointed out that the positive value of ∆N indicates that the coating molecule act as an electron donor and the corrosion rate decreases with an increasing value of ∆N because metal-molecule bond strength increases with the increase of ∆N value[83]. Based on the results in Table 5, the order of ∆N values is 1A ≈ 1M > 1G ≈ 1T regardless of phase, indicating organosilanes with thiol and amino functional groups provide the best metal-silane bond strengths.
Dipole moment (µ) is another descriptor of the corrosion protection efficiency of the coating molecule. It is extensively used to define the polarity of a molecule. Efficiency against corrosion increases with the increase of dipole moment[84, 85] as the increasing value of dipole moment facilitates the adsorption of the molecule on the inorganic substrate surface through physical forces, accompanied by the physical desorption of water molecule. As shown in Table 5, molecules 1G, 1M and 1A have µ values higher than that of the water molecule (1.88D). Therefore, all these molecules with functional groups are expected to be strongly adsorbed on the substrate surface through physical forces which are in good agreement with FTIR results described in the experimental section. The order of µ values is 1M > 1G > 1A regardless of phase, indicating organosilane molecule 1M containing thiol functional group provides the strongest adsorption through physical forces onto the galvanized steel substrate compared to that for other silanes.
Overall, based on the quantum chemical calculations of global descriptors considered for each molecule as described above, the order of efficiencies of coating molecules against corrosion of galvanized steel is 1M ≈ 1A > 1G > > 1T. These results clearly establish the fact that functional organosilanes such as MPTMS, APTES, and GPTMS offer much better corrosion protection of galvanized IF steel than normal silanes as TEOS. All these computational results are in good agreement with experimental results. The trends of the curve using capacitive impedance from the experiments and the ∆N and µ values from the quantum chemical calculation are mutually satisfying the observation as shown in Fig. 7. Therefore, quantum chemical calculations can be used in silane-based sol-gel coating formulation to select the best possible coating recipe for corrosion protection of steel. By doing so, there will be needed limited experiments in a short time to establish a good coating.
Figure 7 represents the correlation between the charge transfer values and the dipole moment offered by the different silane-coated systems with that of its coating capacitance. The higher value for both the charge transfer and dipole moment would indicate better protection of galvanized steel against corrosion which is also evident from the lower coating capacitance value from the Fig. 7 (a). The coating capacitance for the functional silane coatings (1G, 1M, 1A) is almost 10 orders of magnitude lower than that of the non-functional silane coating (1T). 1M coating offers the highest impedance, which reflects the higher corrosion resistance observed in the potentiodynamic studies as well.