*3.4. Surface Analysis*

The surface morphology of the samples exposed for 24 h at different CO2 partial pressures in the absence and presence of 1.0 g L−<sup>1</sup> of GA are presented in Figures 7–9. For instance, it can be seen that the surface morphology of the samples exposed to the blank and inhibited solution differs significantly. For the blank solution, at PCO2 = 1 bar, the microstructure of the sample is clearly visible (Figure 7a). The metal surface appears corroded resulting from the selective dissolution of the ferritic phase over the cementite contained in the perlitic phase. By contrast, Figure 7c,d show that after the addition of the inhibitor the metal surface becomes much smoother. It is clear from the image that the metal surface was partially covered by a protective layer, although some areas of the surface still show signs of corrosion attacks.

**Figure 7.** SEM images of the N80 carbon steel surface morphology after 24 h of immersion in the uninhibited ((**a**) **a** lower and **b** higher magnification) and inhibited ((**b**) **c** lower and **d** higher magnification) solution at 25 ◦C and PCO2 = 1 bar CO2.

**Figure 8.** SEM images of the N80 carbon steel surface morphology after 24 h of immersion in the uninhibited ((**a**) **a** lower and **b** higher magnification) and inhibited ((**b**) **c** lower and **d** higher magnification) solution at 25 ◦C and PCO2 = 20 bar CO2.

**Figure 9.** SEM images of the N80 carbon steel surface morphology after 24 h of immersion in the uninhibited ((**a**) **a** lower and **b** higher magnification) and inhibited ((**b**) **c** lower and **d** higher magnification) solution at 25 ◦C and PCO2 = 40 bar CO2.

The severity of the corrosion attack increases with an increase in CO2 partial pressure in the blank solution, as shown in Figure 8a,b and Figure 9a,b respectively carried out at PCO2 = 20 bar and PCO2 = 40 bar. However, it can be seen that in the presence of GA, an increase in CO2 partial pressure led to a gradual increase in the surface coverage on the metal surface, as a result of an increase of GA molecules adsorbed onto the metal surface (Figure 8c,d and Figure 9c,d). At higher CO2 partial pressure (e.g., PCO2 = 40 bar, Figure 9) the protective action of the inhibitor is even more evident. The images show that for the uninhibited solution, the surface of the metal appears severely corroded, while the one obtained in the presence of the inhibitor shows the formation of a uniform protective layer over its entire metal surface. The results indicate that in the presence of GA and with a gradual increase in CO2 partial pressure, the protective layer gradually becomes more compact and thicker [4]. As discussed in Section 3.1, the solubility of CO2 increases with its partial pressure, and as a result of this, the concentration of H<sup>+</sup> ions into the solution also increases, hence the number of the inhibitor molecules that can be protonated and adsorbed onto the metal surface also increases according to the Equation (12), leading to a substantial reduction of the corrosion rate of the metal.

The morphology of the metal surface was also analyzed with the help of an energy-dispersive spectroscopy with the result listed in Table 4. In the absence of GA, the metal surface was characterized by a corrosion product layer mainly consisting of carbon, iron, and a small amount of oxygen elements, indicating that this corrosion layer is mainly composed of Fe3C. These results are in agreement with that previously observed in the literature [5,7,39,40]. Other researchers reported that at a temperature below 40 ◦C, the corrosion product layer is generally composed of Fe3C, and only little traces of FeCO3 were observed on the metal surface [4,7,39,40], as also confirmed by the GIXRD measurements shown in Figure 10. The presence of Fe3C on the metal surface is due to the anodic dissolution of the ferrite phase over the cementite in the perlitic phase, which leads to an accumulation of the cementite on the metal surface.


**Table 4.** Weight percentage of the elements calculated from EDS analyses.

**Figure 10.** XRD spectra of corrosion product film formed on the metal surface after been exposed for 24 h without and with the presence of 1.0 g L−<sup>1</sup> of GA at different CO2 partial pressures at 25 ◦C.

It is worth mentioning that in the presence of GA the content of carbon and oxygen was found to be higher than those observed for the blank solution. It should be noted that carbon and oxygen are also the main constituents of the tested inhibitor and therefore, their higher concentration on the protective layer formed in the presence of the inhibitor can be attributed to its adsorption onto the metal surface, as also reported by other studies [4,7,20]. Moreover, it can be seen from the table that the percentage of Fe decreased in the presence of GA, likely due to the overlying effect of the inhibitor layer.

The GIXRD analysis for the samples corroded in an inhibited and uninhibited solution at PCO2 = 40 bar and at 25 ◦C (Figure 10) shows the presence of cementite on the metal surface, although in the presence of GA the intensity of these peaks is much weaker. This result can be explained as follows: Fe3C accumulates on the metal surface after the dissolution of the ferritic phase. However, in the presence of the inhibitor, it only accumulates in small amounts on the bare metal surface at the early stage of the experiment, since the dissolution of the ferritic phase is quickly suppressed by the absorption of the inhibitor on the surface of the metal.

Figure 11a,b show the surface morphology for specimens corroded in the blank and inhibited solution carried out at 60 ◦C and PCO2 = 40 bar, after immersion the samples for 24 h in the tested solution, without and with the presence of GA, respectively. The corrosion product layer appears to be different for the inhibited solution compared to one observed in the presence of GA. Figure 11a shows the presence of a porous corrosion product layer formed onto the metal surface corroded in a free-inhibitor solution, pores which create paths for the solution to penetrate it and thereby leading to the dissolution of the underlying metal. On the other hand, the surface of the metal corroded in the presence of GA (Figure 11b) shows the formation of a more compact layer, which forms a better protective barrier and thereby greatly reducing the corrosion rate of the metal.

**Figure 11.** SEM images of the N80 carbon steel morphology after 24 h of immersion in the tested solution at PCO2 = 40 bar and at 60 ◦C, without (**a**) and with (**b**) the presence of GA.

EDS analysis reports high content of carbon, oxygen, and iron elements in both layers (C:11.11%, O:6.02% and C:13.69%, O:11.0%, in the blank and inhibited solution, respectively). The GIXRD measurements presented in Figure 12 show the characteristic XRD diffraction patterns associated with FeCO3. By contrast, the intensity of the iron carbonate peaks observed in the presence of GA is almost negligible. These results suggest that the layer observed for the uninhibited solution is mainly composed of Fe3C and FeCO3, while in the presence of GA is mainly composed of Fe3C with little traces of FeCO3 [4]. Similar behavior was also reported by Ding et al. [41] related to the study of the effect of an imidazoline-type inhibitor against CO2 corrosion of mild steel. The authors suggested that the formation of the corrosion inhibitor layer was able to suppress the formation of the iron carbonate. The precipitation of FeCO3 depends on the concentration of the Fe2<sup>+</sup> and CO2<sup>−</sup> <sup>3</sup> ions, pH, and temperature. When the concentrations of Fe2<sup>+</sup> and CO2<sup>−</sup> <sup>3</sup> ions exceed the solubility limit, FeCO3 will precipitate on the surface [9,40,42]. At higher temperatures, its solubility decreases, and therefore the likelihood of its precipitation will be also higher. In a free-inhibitor solution, the dissolution of the ferrite phase may lead to an increase in the concentration of Fe2<sup>+</sup> ions in the bulk solution and thereby favoring the precipitation of FeCO3 onto the surface of the metal. Conversely, in the presence of the inhibitor, the protective layer formed onto the surface of the metal slows down the corrosion processes, and thereby reducing the concentrations of Fe2<sup>+</sup> ions available for the formation of FeCO3.

**Figure 12.** XRD spectra of corrosion product film formed on the metal surface after been exposed for 24 h without and with the presence of 1.0 g L−<sup>1</sup> of GA at PCO2 = 40 bar and at 60 ◦C.

Figure 13a–d show the SEM analysis of the metal surface after 168 h of immersion in the absence and presence of 1.0 g L−<sup>1</sup> GA at PCO2 = 40 bar, respectively. It is apparent from the figures that a thick porous layer covers both surface samples; although it seems that in the presence of the GA, this layer appears denser, thus providing a higher level of protection. To analyze the condition of the metal surface, these porous layers were removed with the help of Clark's solution. It can be seen that both surfaces show clear signs of corrosion attacks (Figure 13b; however, it is also clear from the figures that in the presence of the inhibitor (Figure 13d) the surface of the metal appears to be less damaged and smoother, with the ground scratches still visible on the surface. This result was also confirmed by the atomic force microscopy experiments performed by Azzaoui et al. [28] concerning the use of GA as a corrosion inhibitor in a 1 M HCl solution. The authors reported that in the uninhibited solution the surface of the metal was found to be more corroded with an average roughness of 1.3 μm, while in the presence of GA the average roughness was reduced to 500 nm. The authors justified this behavior due to the formation of a more compact protective layer on the metal surface that strongly reduced the diffusion of the aggressive substances to the metal, and thereby reducing the corrosion rate of the metal.

**Figure 13.** SEM images of the N80 carbon steel morphology after 168 h of immersion in the tested solution in the presence of 1.0 g L−<sup>1</sup> of GA at PCO2 = 40 bar. Without (**a**,**b**) and with the inhibitor (**c**,**d**) at 25 ◦C.

The SEM-EDS and GIXRD result confirm that GA provides adequate protection to the metal surface from sweet corrosion even at high CO2 partial pressures and after long immersion times. The results are in agreement with the findings obtained with the weight loss measurements, confirming the high inhibition efficiency value observed after a long immersion time.

X-ray photoelectron spectroscopy analysis was employed as a means to confirm the adsorption of the tested inhibitor on the carbon steel surface. The analysis was carried out on the native inhibitor and the steel surface after 24 h of immersion in the tested solution in the presence of 1.0 g L−<sup>1</sup> of GA at PCO2 = 40 bar and at 25 ◦C. The XPS results presented in Figure 14a showed evidence of the presence of O, C, N, and Fe on the carbon steel surface, where the O and C contents displayed the highest amount, while the signal of N was detected with small intensity. The high-resolution peaks core levels were analyzed through a deconvolution fitting of the complex spectra. The binding energies and the corresponding quantification (%) of each peak component are presented in Table S4.

**Figure 14.** *Cont.*

**Figure 14.** XPS spectra of the native gum Arabic: (**a**,**c**,**e**,**g**). XPS spectra of the film formed on the N80 carbon steel after 24 h exposure in CO2 at PCO2 = 40 bar in the presence of 1.0 g L−<sup>1</sup> of GA at 25 ◦C: (**b**,**d**,**f**,**h**).

The deconvoluted Fe2p3/<sup>2</sup> peaks (Figure 14b) at 710.5 and 713.0 eV could be associated with the α-Fe2O3 or/and γ- Fe2O3 [13]. The presence of these species is likely due to the partial decomposition of iron carbonate. The literature reported that FeCO3 begins decomposing at temperatures below 100 ◦C according to the following reaction [5,42]:

$$\text{FeCO}\_3 \rightarrow \text{FeO} + \text{CO}\_2 \tag{15}$$

In the presence of CO2 or water vapor, FeO transforms into Fe3O4 [5,42].

$$\text{3FeO} + \text{CO}\_2 \rightarrow \text{Fe}\_3\text{O}\_4 + \text{CO} \tag{16}$$

$$\text{C}3\text{FeO} + \text{H}\_2\text{O} \to \text{Fe}\_3\text{O}\_4 + \text{H}\_2\tag{17}$$

However, in the presence of oxygen, FeO and Fe3O4 transform into Fe2O3 [5,42].

$$\text{2FeO} + \text{O}\_2 \rightarrow 2\text{Fe}\_2\text{O}\_3\tag{18}$$

$$\text{In the air}: \ 4\text{Fe}\_3\text{O}\_4 + \text{O}\_2 \to 6\text{Fe}\_2\text{O}\_3\tag{19}$$

The C1s spectra of the native gum arabic and the adsorbed one (Figure 14c,d, respectively) show three main peaks. The C1s peak with binding energy at 284.8 eV could be attributed to the C–C/C–H bonds [26,28]. The C1s peak at 286.2 eV could be attributed to the C–OH/C=O bonds related to the different groups of GA [26,43]. This peak may also be assigned to the carbon atom bonded to nitrogen in C–N bond [13,44] and could be related to the glycoprotein and/or to the arabinogalactan-protein fractions of the inhibitor (Figure 15b,c, respectively). The last C1s peak with a binding energy of 287.7 eV could be associated with the presence of carbonyl type groups O–C=O/N–C=O that result from the protonation of the GA molecule in the acid environment [28].

It is worth mentioning that no peaks assigned to Fe3C were found with the XPS analysis in contrast to the results reported from the GIXRD analysis, where the characteristic peaks assigned to this compound can be seen in the presence of GA (Figure 10). Fe3C cannot be detected since the average depth of analysis for an XPS measurement is approximately 5 nm however, the cementite formed on the metal surface at the early stage of the experiment is covered by a thicker layer of inhibitor (Figure 9c,d).

The deconvoluted O1s spectra of the native and adsorbed inhibitor are displayed in Figure 14e,f, respectively. The peaks at 531.2 and 532.7 eV could be attributed to the single bonded oxygen in C–O and the double bonded oxygen C=O and/or to the single bonded oxygen in O–C–O respectively [4,13,26,28]. The latter peak may correspond to the carbonyl type groups and/or to the glycosidic C(1)-O-C(4)/C(1)-O-C(6) linkages of the GA molecules (Figure 15a), as well as, in

the case of the sample exposed to the tested solution, to FeCO3 formed on the metals surface, respectively [4,26,28]. Moreover, some authors reported that the peak at 231.2 eV could also be attributed to the oxygen of the hydroxyl groups (–OH) [5,43], likely due to the hydroxyl groups of the tested polysaccharide. The O1s spectrum of the adsorbed inhibitor (Figure 14f) displays an extra peak at 529.7 eV corresponding to O2<sup>−</sup> related to the oxygen atoms bonded with Fe3<sup>+</sup> in the Fe2O3 oxide [4,43,44]. The O1s results are in good agreement with the findings of the Fe2p spectrum.

**Figure 15.** Structure of gum arabic: (**a**) arabinogalactan; (**b**) glycoprotein; (**c**) arabinogalactan-protein.

The presence of N1s peak in the survey for the adsorbed GA on the carbon steel surface (Figure 14a) provides evidence that gum arabic was effectively adsorbed on the tested substrate surface since the N80 carbon steel substrate does not contain nitrogen in its chemical composition. The N1s spectra of the native and adsorbed inhibitor are presented in Figure 14g,h. Both images show the presence of a peak at 400 and 399.8 eV attributed to the nitrogen atom bonded with the carbon atom, for the native and adsorbed inhibitor. However, as it can be seen that the high-resolution N1s spectrum of the tested substrate sample after the addition of GA depicts an extra peak at 397.6 eV. This extra peak can be ascribed to the coordinated nitrogen atom of the amino group with the metal surface (N–Fe bond) [44]. Other authors also suggested that this peak could be attributed to the bond between the nitrogen of the amino groups and the oxide layer on the metal surface (FeOx) [45].
