Mechanism of corrosion protection in chloride solution by an apple-based green inhibitor: experimental and theoretical studies

Preservation of metals in infrastructures and other assets requires cost-effective and sustainable solutions such as green corrosion inhibitors. This study assesses an apple pomace-derived green inhibitor synthesized by an innovative zero-waste method. Electrochemical measurements revealed the high performance of this liquid extract in reducing the corrosion of carbon steel in NaCl brine. The chemical composition of this inhibitor was characterized by liquid chromatography mass spectroscopy (LC-MS) to shed light on the corrosion inhibition mechanism. Based on LC-MS analysis, the results of surface analysis were interpreted. Specifically, the major corrosion inhibitor agent in the apple pomace extract was determined as C26H50NO7P (1-Linoleoylsn-glycero-3-phosphocholine), which can adsorb onto the steel surface to form a barrier layer and serve as a blocker of active anodic sites. Further study showed that the apple extract adsorption follows the Langmuir isotherm, and physical adsorption is dominant (vs. chemical adsorption). Theoretical calculations using quantum chemistry proposed a physisorption mechanism for the protection of steel by C26H50NO7P molecules.


Introduction
Chloride-based deicers have been used extensively in winter road maintenance operations [24,37]. Since these chemicals are water soluble and corrosive, they can pose a great risk on the integrity, performance, and service life of steel structures (and components) [7,43]. They may induce premature failures, which have substantial implications on the reliability, resilience, serviceability, and environmental footprints of the structure [29,69]. One of the cost-effective methods for preventing or reducing the corrosion damage entails the application of corrosion inhibitors [20,59,78]. This method has been widely employed by maintenance agencies to mitigate the corrosion risk of deicers [50]. The efficiency of corrosion inhibitors is defined by their physicochemical properties such as active functional groups, electronic structure, electron density, non-bonding (lone pair) electrons, and π-electrons [64,66].
Traditional corrosion inhibitors contain chemical compounds that can have toxicological effects and pose harmful impacts on the receiving environment, especially surface waters [18,30]. In recent years, green corrosion inhibitors have been introduced as alternatives to traditional inhibitors [13,38,40]. Bio-based materials have been used in the winter maintenance operations of roadways, either as freezing point depressant, or as green corrosion inhibitor [53]. These materials are mostly produced by fermentation and processing of desugared molasses, beet juice, corn, and other agricultural products [2,16].
Recently, the authors have introduced an innovative zero-waste method for producing agro-based corrosion inhibitors as an additive to deicer/anti-icer chemicals. The major advantage of the chemicals prepared by this method over the commercial green chemicals used by the departments of transportation (e.g., beet juice) is that the former ones feature much lower chemical oxygen demand, which makes them more eco-friendly [31]. An example of inhibitors synthesized by the innovative method is a liquid extract made from peony leaf waste, exhibiting a corrosion inhibition efficiency of 65.8% for carbon steel in NaCl brine [51]. Due to relatively low corrosion inhibition efficiency (IE) of the peony leaf extract, the authors further explored a variety of other local agro-based wastes as feedstock (such as grass, dandelion leaf, blueberry pomace, cherry pomace, and apple pomace) to produce highefficiency "green" corrosion inhibitors. After a preliminary study, the apple pomace seemed to be the most promising agricultural waste for producing green inhibitor among the tested wastes. Therefore, we explored this feedstock further. In addition, the amount of apple waste in Washington State is 27,794 tons of dry biomass per year [27]. Therefore, the current research may lay the foundation of converting a huge amount of waste to an ecofriendly product, which can notably benefit the society and the environment.
Apple pomace has been used as a food ingredient, and for producing jam and jelly [25,41]. Some researchers have used commercial apple-derived pectin (C 6 H 10 O 7 ) as a green inhibitor [79]. Pectin can be extracted from apple pomace using different acid extraction methods which produce solvent waste [42,81]. However, no waste is produced during the extraction process of producing the apple pomace extract herein, which makes it a more environmentally friendly corrosion inhibitor than commercial pectin.
This work investigated the effect of an apple pomace liquid extract on the corrosion of C1010 mild steel in 3.5% NaCl brine. A zero-waste process developed by the authors was adopted to produce this "green" corrosion inhibitor. Chemical, electrochemical, and surface analyses were used to shed light on the corrosion behavior. Finally, using quantum chemical calculations we propose a mechanism of corrosion inhibition.

Materials
The corrosion coupons were prepared from C1010 steel with a composition presented in Table 1. The exposed surface area of each coupon was 1 cm 2 . The coupons were polished by 60-to 1500-grit sandpaper sequentially, degreased by ethanol, and then washed with deionized water. In the next step, they were exposed to 3.5% NaCl brine over a 7-day period (168 h). The reason for choosing 3.5% NaCl brine for this research is that it is known that NaCl concentration of around 3% is the most corrosive concentration of NaCl brine for iron in aerated solutions at room temperature [11]. Therefore, the authors used this concentration to simulate a worst-case scenario. In the field, the concentration of 23% NaCl is the common asapplied concentration, for anti-icing or deicing applications. However, due to snow precipitation and ice melting conditions, the stormwater runoff tends to have much lower concentrations of NaCl.
Different concentrations of the apple pomace extract (1%, 2%, and 3% v/v (volume of extract/total volume)) were employed in this research. These values corresponded to dry mass per volume of apple pomace extract (APE) at 2.61, 5.22, and 7.82 g/L, respectively. For drying the APE, it was oven dried at 103°C for 24 h. The code names of the samples are shown in Table 2. While the range of 1-3% v/v is high relative to the range of ppm used by many researchers, it is much lower than 20% v/v that is common for using agrobased byproducts such as beet juice in winter maintenance operations. In addition, some researchers have reported the range of up to 4% v/v of alcohol-based inhibitors for protection of reinforcing steel [36].
The apple pomace extract was prepared using a chemical/biological process that has been described elsewhere in detail [51]. Briefly, in this process the apple pomace, which contained seed and peel, underwent two degradation steps: first, chemical degradation in an alkaline solution containing urea at around − 13°C, then biological degradation by bacteria at room temperature. The resulting solution was heated at 85°C until its volume halved. The concentrated solution was used as the corrosion inhibitor without further treatment or purification. All the chemicals used in this research were of analytical-grade purity.
It should be mentioned that the process used in this research is more complex than common processes of dissolving the agricultural feedstock in a solvent such as a mixture of water and alcohol, or acid [17,76].

Mass spectrometry
The liquid chromatography-mass spectrometry (LC-MS) method was employed for in-depth chemical analysis of the APE. The analysis was conducted in the positive ion mode using Synapt G2-S mass spectrometer. Calculation of the normalized weight percent (wt.%) of each compound was conducted based on the normalized percent surface area of each peak with the aid of the ImageJ software.

Electrochemical measurement
A PARSTAT MC multichannel Potentiostat performed the corrosion measurements using a standard three-electrode system comprising the working electrode (corrosion coupon), a reference electrode (Ag/ AgCl, KCl saturated), and a counter electrode (platinum mesh). The electrochemical impedance spectroscopy (EIS) measurements were carried out using the same system in the frequency range of 100 kHz -10 mHz by superimposing AC signal of ±10 mV around the open-circuit potential (OCP). Tafel polarization test was conducted at a potential sweep rate of 0.167 mV/s in the range of − 250 mV to + 250 mV with respect to OCP. The VersaStudio software was used for running the electrochemical corrosion tests and calculating the Tafel polarization parameters, and ZSimp-Win 3.60 software was employed for interpretation of the EIS data. The tests were at least duplicated to obtain an acceptable reproducibility. All of the electrochemical tests were performed at room temperature.

Surface analysis
The surface layer was examined via a JEOL JXA-8500F electron probe micro-analyzer (EPMA). The EPMA uses an electron probe to generate irradiation on the specimen surface, and then the characteristics of the produced X-ray are analyzed by wavelength-dispersive X-ray spectroscopy (WDS) techniques to qualify the element distributions on the surface. For each area, five elements (e.g. Fe, N, Cl, O, and P) were analyzed and their spatial distribution mapped. In order to study the surface layer more in-depth, Xray photoelectron spectroscopy (XPS) analysis was used. The analysis was carried out using an AXIS-165 multielectron spectrometer from Kratos Analytical Inc. The data acquisition and data analysis were performed using the XPSPEAK 4.1 software.

Quantum chemical calculation
All the calculations were performed using molecular mechanics/MM+ level, and the semi-empirical calculations with PM3 method [74] at 25°C by Hyperchem 7.52 software.

Results and discussion
Chemical composition of the liquid extract Figure 1 and Table 3 show the results of LC-MS analysis for the APE. Based on the mass spectrum data, the APE molecules mainly include C 26 H 50 NO 7 P and C 31 H 43 N 5 O (and other organic compounds), which consist of nitrogen, oxygen, and phosphorus elements. The organic mixtures containing N, P, and O elements have polar functions, so will adsorb onto the metallic surface as a barrier organic layer [19]. The adsorption of the polar atoms on the surface of metal may occur through coordination between the lone pair or π-electrons cloud and the metallic surface [23,71].

Electrochemical characterization OCP measurements
The OCP of steel coupons in the presence and the absence of the APE was measured after different periods of immersion in NaCl brine, as shown in Fig. 2.
The OCP can be a function of corrosion potential and electrical resistivity of the protective layer formed on the sample [83]. It can be seen that the OCPs of the samples exposed to the solutions containing the apple pomace extract shifted to more noble values, which may be a sign of the more thermodynamically steady system [33]. This conclusion has been confirmed by considering the steady trend of the OCP for sample APE3 after the third day of the experiment (72 h). This effect is not the same for all concentrations; for instance, APE2 experiences a sharp decrease in OCP values after 72 h; even more than APE1. This behavior can be due to the probabilistic nature of corrosion which causes fluctuations of OCP [28].
Tafel polarization results Figure 3 presents the Tafel polarization plots of the samples after 7-day immersion in NaCl brines having different concentrations of APE. It is seen that addition of APE shifted the corrosion potential to a more noble potential and decreased the anodic current density. These can be ascribed to the adsorption of organic molecules on the anodic active sites of the steel surface [1,14]. Table 4 shows the electrochemical parameters derived from the Tafel polarization plots. By increasing the concentration of the APE, the corrosion current density (i corr ) was notably decreased and the lowest i corr was obtained at 3% (v/v) apple pomace extract. The maximum inhibition efficiency (IE%) of APE calculated using i corr value was 83%, which is significantly more than that of pectin (78.7%) for X60 steel in 0.5 M HCl solution [79] and that of Sarang Semut (Myrmecodia Pendans) extract (65.21%) for low carbon steel in 3.5% NaCl solution [62]. By adding the extract at 3% (v/v), the corrosion potential (E corr ) of the steel increased by more than 85 mV, which shows that APE is an anodic inhibitor and the corrosion inhibition is attributable to change in the activation energy [14,21]. Based on Table 4, the values of anodic and cathodic Tafel slopes were 104.1 and 384.4 mV/dec, respectively for the APE0 sample. Both anodic and cathodic Tafel constants decreased with increases in the APE concentration. Figure 4 depicts the impedance Nyquist curves of steel samples in NaCl brines with various concentrations of APE. The capacitive arc radius increased by increasing the concentration of the apple pomace extract, which corresponds to improved corrosion resistance [15,49]. The capacitive arcs for samples APE0 and APE1 have similar forms despite their different sizes, which shows that the inhibition mechanism is independent of theextract concentration for these samples. However, the feature of the capacitive arc with a Warburg impedance emerged in samples APE2 and APE3. The observed arcs in all samples are non-ideal semi-circles which is likely due to the frequency dispersion caused by corrosion-induced non-homogeneity of the steel surface [8,12]. In addition, adsorption of the barrier organic layer on the surface of metal can induce surface heterogeneity that can affect the shape of capacitive arc [9,58].

EIS results
The Bode curves in Fig. 5 reveal that by increasing the APE concentration, |Z| value at the frequency impedance of 10 mHz (Log f = − 2) increased. Since this value corresponds to the corrosion protection provided by the inhibitor [65], it shows that by increasing the inhibitor concentration, the corrosion protection was improved.
The equivalent electrical circuits used for studying the impedance spectra are shown in Fig. 6. The circuits consist of the following elements: R s (solution resistance), R f (corrosion film resistance), Q f (constant phase element (CPE) of film), R ct (charge transfer resistance of metal/ solution interface), and Q dl (CPE of metal/electrolyte interface), and W (Warburg impedance).  Some of the values of the fitted parameters based on the equivalent electrical circuit model are summarized in Table 5. In the equivalent electrical circuit, Q dl is the CPE of the electrical double layer at the metal/electrolyte interface. The CPE represents an non-ideal capacitor in the EIS results, and its impedance is stated as follows [73].
Where Y 0 is the CPE magnitude, n is the phase shift due to surface roughness (− 1 ≤ n ≤ 1), and ω is the angular frequency. The values of C dl , represented in Table 5, were measured as follows [44].
The C dl values decreased with increases in the concentration of APE, which can be ascribed to the reduction of double layer dielectric constant [10]. This means a thicker electrical double layer, which is caused by more displacement of corrosive species by APE molecules at the steel/electrolyte interface [57,75]. By increasing the concentration of APE, the values of R ct increased which could translate to higher inhibition efficiencies [26]. In addition, R f increased in the presence of higher concentrations of APE, which may pertain to the decrease in the number of ionically conducting paths. This implies a slowed-down migration rate of aggressive species (such  as Cl − ions) through the adsorbed protective layer towards the surface of steel. The Warburg impedance was observed in APE2 and APE3 samples, which shows a resistance due to the diffusion process over the porosity of the corrosion product film [32,35]. The fitted curves for Nyquist and Bode curves of samples APE0 and APE3 are shown in Fig. 7. As can be seen, there is a good match between the modeled and experimental results. For modeling the electrochemical behavior of samples APE0 and APE3, the equivalent circuits presented in Fig. 6 (a) and 6 (b) were used, respectively. The maximum IE of APE calculated using EIS data was 98.8%, which was much more than that of pectin (77.1%) for X60 steel in 0.5 M HCl solution [79] and that of Myrmecodia Pendans extract (79.7%) for the corrosion of mild steel API 5 L Grade B in 3.5% NaCl [63].
The Nyquist curves of the APE3 sample at various immersion periods are shown in Fig. 8. The protection responses were modified over time, controlled by the charge transfer mechanism for all time periods. The biggest capacitance arc was observed at day 7, which showed the stability of the protective layer over time. The Bode curves depicted in Fig. 9 demonstrate the increasing of the |Z| 10mHz with APE concentration, which shows improvement in anti-corrosion performance of the inhibitor and its stability over time. After five days of immersion, the Bode curve showed the fingerprint of two time constants, which can be explained by the changing of the protection mechanism from geometry blocking to energy effect [14]. This is in good agreement with Tafel results that linked the APE corrosion inhibition to changing of the activation energy.

Adsorption isotherm
The interaction of APE molecules with the steel surface can be determined by using an adsorption isotherm. In this regard, the surface coverage degree (θ) was assumed to be equivalent to inhibition efficiency (IE) measured by the EIS technique. The IE versus APE concentration was fitted to Langmuir, Freundlich, Flory-Huggins, and Temkin adsorption isotherms. The Langmuir isotherm exhibited the best fit to the data, with the correlation coefficient of 0.997 (Fig. 10). The Langmuir isotherm follows the following equation [54].
Where C is the concentration of APE in g/L (dry mass/volume of extract solution), K is the adsorption coefficient (L/g). Classical adsorption isotherms (such as Langmuir) use the model of non-penetrable interface in which a solvent can be substituted by the molecules of an adsorbate [45].  The change in the free energy of adsorption (ΔG 0 ads ) was measured as follows [54,68].
Where R designates the gas constant, T shows the absolute temperature, K is the adsorption coefficient (L/g), and C solvent is the water concentration (1000 g/L). The Gibbs free energy was estimated to be − 16.33 kJ mol − 1 . The negative value of ΔG 0 ads suggested the spontaneous occurrence of adsorption process of APE species onto the steel surface. The absolute value of Gibbs free energy was smaller than 20 kJ mol − 1 , which suggests that the physical adsorption is dominant, rather than chemical adsorption [72]. In this condition, the electronic structure of the adsorbate is disturbed because of the  electrostatic interaction between APE molecules and the steel surface [70].

Surface analyses SEM micrographs
The SEM images of as-washed coupons are provided in Fig. 11. A rough surface was observed on the uninhibited sample (APE0) and smoother surfaces on the samples exposed to APE containing solutions. Other researchers for the uninhibited and inhibited samples in Cl − containing media reported similar results [3]. The amount of metallic surface covered by the organic layer increased with increases in the concentration of APE. This is in good agreement with the results obtained by electrochemical measurements and adsorption isotherm.

EDS results
The main elements of the surface layer formed on the aswashed APE0 -APE3 steel samples are shown in Fig. 12.
The main peaks for all samples corresponded to iron and oxygen, which suggests the formation of iron oxides on the surface. This is in good agreement with the results obtained by EIS measurement that suggested the formation of a corrosion product layer on the surface of steel. In the case of APE3, a relatively large peak related to phosphorus can be seen around 2 keV, which is attributed to the adsorption of inhibitor molecules onto the steel surface.

EPMA results
For more in-depth analysis of the corrosion surface layer, EPMA analysis was employed which enables simultaneous detection of different elements and mapping of their distribution on a defined surface area. The EPMA elemental maps for the surface layer formed on the APE0 and APE3 steel samples are shown in Fig. 13. On the steel surface exposed to uninhibited NaCl brine (APE0), there is little presence of N or P and the corrosion product layer consisted of an iron oxide/hydroxide layer containing Cl. The interface between the corroded surface (left) and the intact steel substrate (right) is clearly illustrated by the elemental maps of Fe, O and Cl. In contrast, the surface layer in NaCl brine with high inhibitor concentration (APE at 3%) clearly consisted of a phosphorous compound. These results are in very good agreement with the results obtained by EDS. Therefore, from these results and the LC-MS data it can be concluded that C 26 H 50 NO 7 P is the major corrosion inhibitor agent in the APE. The charge density of phosphates is more negative than that of amines [34]; as such, the bond of former with metallic surface is more stable. Figure 14 depicts the results of XPS analysis for WPE0 and WPE3 samples. In the plot associated with Fe2p in WPE0, there is a peak at 708.37 eV which is related to Fe 3 O 4 [46]. The XPS analysis of O1s unveils two peaks at 530.65 eV and 528.8 eV, which are associated with Fe 3 O 4 and Fe 3 O 4 complexes, respectively [4,5]. In the presence of APE (sample APE3), two peaks appeared in the Fe2p plot. The major peak seen at 709.66 eV comes from Fe 2 O 3 [60]. The second peak at 708.63 is due to Fe 3 O 4 [46]. Therefore, in the presence of APE, the major part of corrosion product layer is converted to Fe 2 O 3 from Fe 3 O 4 . It is known that Fe 2 O 3 is more corrosionresistant than Fe 3 O 4 [39]. As such, one of the reasons that APE3 has a less corrosion rate than APE0 is the catalytic effect of APE in converting Fe 3 O 4 to Fe 2 O 3 in corrosion surface layer. It is known that phosphates can catalyze the oxygen evolution at near neutral pH [77]. In addition, they can stabilize the corrosion product layer [47]. Therefore, C 26 H 50 NO 7 P played the main role in converting Fe 3 O 4 to Fe 2 O 3 . Observation of the N1s plot shows a peak at 400.75 eV that indicates adsorption of an amine-containing compound on the steel surface [82], which may be C 31 H 43 N 5 O. In addition, the plot of P2p shows a peak at 134.72 eV that reveals an aminephosphate based constituent [61], which may have resulted from the adsorption of C 26 H 50 NO 7 P compound.

Quantum chemical calculations
Theoretical studies have prospective applications in designing and developing of many organic inhibitors in the field of corrosion inhibition chemistry. The capability of the inhibitor for corrosion prevention bases on its spatial distribution of molecule and its electronic arrangement over frontier molecular orbitals. Molecular geometry systems of compounds C 26 H 50 NO 7 P and C 31 H 43 N 5 O as the main compounds of the APE were determined. The stable geometry of the compounds was obtained by PM3 method, which was carried out in Hyperchem software (Fig. 15). The calculated chemical parameters for C 26 H 50 NO 7 P and C 31 H 43 N 5 O molecules are given in Table 6. It is seen that the dipole moments (μ) for C 26  form electrostatic interactions with the metal surface, that result in a strong adsorption on the surface of metal [55]. In this research, the value of μ for C 26 [80]. In addition, the μ value of C 26 H 50 NO 7 P (17.0 D) is more than that of pectin, 11.725 D [79], which justifies the better anti-corrosion performance of APE than pectin. The energy gap (ΔE = E LUMO -E HOMO ) between the E LUMO and E HOMO levels of inhibitor molecules is another important electronic parameter which is given in Table 6. The values of energy gap for C 26   surface and better inhibition efficiency than VB1 with the ΔE of 6.555 eV [6]. Figure 16 displays the electrostatic potential maps for C 26 H 50 NO 7 P and C 31 H 43 N 5 O molecules. More dark red (negative) regions associated with nucleophilic reactivity in the potential map of C 26 H 50 NO 7 P and C 31 H 43 N 5 O indicate that these molecules are readily sharing free electron pairs with the metal surface (electrophilic agent) through electrostatic interactions. Also, the molecules with blue (positive) region have electrophilic reactivity [22]. Figure 17 displays the frontier orbitals including the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) for C 26  It was observed that nitrogen and oxygen atoms in the studied molecules have high electron density. These sites are nominated to be the active sites to proper electrophiles and can be attached within the environment [48]. Based on this, a physisorption mechanism was proposed for adsorption of C 26 H 50 NO 7 P and C 31 H 43 N 5 O molecules on the metallic surface, as depicted in Fig. 18.

Mechanism of corrosion inhibition
The inhibition mechanism in the presence of APE was adsorption of an organic protective film, and chemical conversion of corrosion product layer to a more corrosion-resistant film. The adsorption type was Langmuir isotherm model in which a monolayer of organic molecules form a non-penetrable interface. In the Langmuir model, each active site on the metal surface can be related equally to an adsorbate molecule, and there is no interaction among the adsorbed molecules [67]. In addition, the saturation coverage occurs when all of the surface active cites are occupied by the adsorbate molecules [67]. According to the experimental results, the saturation behavior in the inhibitive characteristics occurred in 3% v/v of APE in this research, since the corrosion efficiency at this concentration of APE in close to 100%, thus it cannot increase considerably.
1-Linoleoyl-sn-glycero-3-phosphocholine (C 26 H 50 NO 7 P) has a major role in the inhibition properties of APE. The corrosion inhibitive behavior of C 26 H 50 NO 7 P can be related to the functional group of its molecular structure [52]. It was also responsible for converting Fe 3 O 4 in corrosion product layer to a more corrosion-resistant iron oxide (Fe 2 O 3 ). APE molecules may attach to the surface of steel though their high electron density sites [70]. The APE adsorption entails the interaction of non-bonding electrons on nitrogen and oxygen with the steel surface.
More APE was associated with higher R ct and R f values which shows the formation of a more protective surface layer. On the other hand, more APE corresponded to lower values of C dl , due to displacement of corrosive agents by organic molecules via an adsorption process at the metal/ media interface. The inhibition efficiency was improved by increasing of APE concentration, which indicates an active corrosion protection. The corrosion protection mechanism changed over time from geometry blocking to energy effect. APE molecules blocked the anodic active sites and shifted the corrosion potential to more passive values. The amount of this shift was more than 85 mV, and this suggests thatAPE is an anodic inhibitor and inhibition property is due to the change of activation energy.

Conclusions
This research assessed the corrosion behavior of C1010 steel in 3.5% NaCl brine in the presence of an apple pomace liquid extract. The corrosion mechanism in the presence of this extract was elucidated using electrochemical measurements, mass spectroscopy, and surface analyses and the following key findings have been obtained.
Apple pomace extract acted as a "green" corrosion inhibitor. The inhibition efficiency increased with increasing the concentration of the extract and over time.
An excellent inhibition efficiency of 98.8% was achieved at 3% of this green inhibitor after 7-day of immersion in NaCl media. Apple pomace extract molecules adsorbed onto the surface of steel following the Langmuir isotherm and physical adsorption was dominant (vs. chemical adsorption). The apple pomace extract blocked anodic active sites on the surface of steel, and converted Fe 3 O 4 to Fe 2 O 3 , which is a more corrosion-resistant iron oxide. The main corrosion inhibitor agent in the apple pomace extract was C 26 H 50 NO 7 P molecule. The protection mechanism improved over time from geometry blocking to energy effect. The quantum chemical calculations confirm the experimental results and suggest the physisorption mechanism of this green corrosion inhibitor.