Proanthocyanidins with Corrosion Inhibition Activity for AISI 1020 Carbon Steel under Neutral pH Conditions of Coconut (Cocos nucifera L.) Husk Fibers.

Cocos nucifera L. is a palm tree (Arecaceae) with a high economic value. The coconut husk fibers are nonedible, thick, and abrasion-resistant and correspond up to 85% of biomass discarded as solid waste residue. Therefore, the husk fibers are an underexploited byproduct with a high content of extractives of unreported nature. Two varieties of C. nucifera L. husk extracts were investigated to uncover bioactive metabolites and their possible application as a green corrosion inhibitor for carbon steel AISI 1020 under neutral pH conditions. The chemical analysis indicated 3% (w/w) of proanthocyanidins in the husk fibers with a high B-type procyanidin content. The husk fibers' crude extract showed promising results as an eco-friendly corrosion inhibitor for carbon steel AISI 1020 under neutral pH conditions. Although it formed a film on the metal surface in all tested concentrations (0.4, 0.8, 1.2, and 1.6 g L-1), the highest protective efficiency was shown at a concentration of 1.2 g L-1, determined by electrochemical techniques and mass loss. This was the first comprehensive report on coconut husk fibers' chemical composition, which was similar between the two varieties with potential for industrial application.

Introduction

The coconut palm tree (Cocos nucifera L., Arecaceae) is considered a vital tropical crop. It is widely distributed in coastal vegetation across the tropics, and its fruits have a significant economic value for the food industry. From the coconut copra, the dried kernel, is produced coconut oil and other products.[1] The Brazilian production of coconuts was over 2.3 million tons in 2018, making Brazil the fifth largest coconut producer in the world (FAO; http://www.fao.org/statistics/).

The coconut husk fibers are nonedible, thick, and abrasion-resistant and correspond to up to 85% of biomass weight.[2,3] They are the leading solid waste residue from coconut production and are rich in cellulose, hemicellulose, lignin, and have a high extractive content.[4] Finding ways to maximize the exploitation of coconuts waste would reduce its accumulation and environmental impact, adding value to the supply chain and generating profits in a biorefinery process.

Coconut husk fiber extracts exhibited potential biological activities[5−8] due to proanthocyanidins (PAs) in the extracts.[6,7,9] The demand for PAs with a high degree of polymerization (DP) has increased, predominantly from abundant, inexpensive, and underexploited byproduct sources for various industrial applications.[10−12] Some studies have explored PAs as corrosion inhibitor alternatives to control and prevent corrosion processes.[13−15]

Corrosion happens spontaneously and naturally. In this process, a metal returns to its most stable chemical form (as oxide) found in nature, making the opposite way of the steelmaking procedure.[16] Thus, considering industrial productions, the steel tools may have their lifetime reduced, resulting in severe economic impact. Carbon steel AISI 1020 is widely used in industry due to its low cost and physical properties. However, it is very susceptible to corrosive processes and an interesting target on anticorrosion techniques study.[17] A method for control, protection, and prevention of the corrosion process depends on many factors, and it is crucial to search for eco-friendly compounds with high efficiency.

The interest in corrosion inhibitors that are ecologically friendly has become a tendency. As the coconut husk fibers, agro-industrial residues have turned into a source of bioactive compounds with low toxicity and large availability.[18] Corrosion inhibitors are metallic surface protectors. They can act inhibiting anodic reactions (anodic inhibitors), cathodic reactions (cathodic inhibitors), or both (adsorption inhibitors). Although compounds from natural sources with a corrosion inhibition potential are usually absorption inhibitors, all mechanisms of action can be observed for different metal types and alloys.[18−21]

The effect of PAs as an eco-friendly corrosion inhibitor has already been described, but for acid solutions.[14,22−27] However, if these compounds were effective in other pH ranges, it would be possible to apply them in cooling systems. Hence, this study proposes the further chemical investigation of coconut husk fiber PAs for uncovering the structural diversity of bioactive metabolites and its possible application as a green corrosion inhibitor for carbon steel AISI 1020 under neutral pH conditions.

Results and Discussion

The general information on extraction procedures is found in Table . The extraction of CCR and CCO ground husk fibers yielded 11.7 and 9.7% (w/w), respectively, twofold, as reported previously.[7] The following liquid–liquid partition with ethyl acetate and water 1:1 (v/v) separated lower weight PAs and other phenolics (because they are soluble in ethyl acetate) and the polymeric PAs, which are soluble in water. The CCR crude extract yielded 8.5% of ethyl acetate and 60.9% of aqueous fractions, while the CCO crude extract yielded 4.5% of ethyl acetate and 80.0% of aqueous fractions. These preliminary results indicate that at least 71–80% of the crude extract is composed of polymeric PAs.

Table 1

General Results Obtained for Both Varieties of C. nucifera L.

		liquid–liquid partitioning weight (mg)
sample	extraction yield (%)	EtOAc	H2O	mDP	overall cis/trans molar ratio	PAs content (%)
CCR	11.73	102.43	730.98	4.51 ± 0.17 [3.74]b	66.64 ± 0.02 [2.60]b	2.95 ± 0.72a
CCO	9.92	44.69	802.87	4.58 ± 0.13 [2.86]b	49.73 ± 0.01 [2.31]b	2.99 ± 0.41a

Mean ± standard deviation.

Mean ± standard deviation [relative standard deviation].

PAs are oligomers or polymers of flavan-3-ol units. They can be divided into several classes based on hydroxylation patterns in A- and B-rings of the constitutive units, the stereochemistry of the chiral C3 carbon on the C ring, the inteflavan linkage between them, and the DP.[28] PAs are frequently linked via B-type bonds (a C4 → C8 or C4 → C6 linkages). An additional C2 → O7 or C2 → O5 linkage produces doubly bonded A-type PAs. As such, PAs often occur as a complex mixture. The lack of appropriate standards and efficient analytical methods poses several challenges for the appropriate PA’s analytical determination. Therefore, for the coconut husk fibers, a combination of analytical techniques was used to provide a comprehensive characterization of PAs.

The PA quantification by the n-butanol/hydrochloric acid test is also displayed in Table . PAs appeared in almost 3% (w/w) of both varieties of coconut husk fibers. An aqueous CCO extract has been analyzed previously by vanillin–HCl assay, showing around 10% of detected PAs in the crude extract.[6] Nevertheless, the first report with the PA content concerning the biomass weight is exhibited here.

Phloroglucinolysis was used to assess chemical composition (nature of extension and terminal units) and the chain extent (the mDP) of the coconut PAs. The aqueous fractions were used to avoid interference of monomers and other phenolics that might be found in the crude extract. Regarding the composition (Table and Figure S1), the main constituent of terminal subunits was a mixture of catechin and epicatechin, and CCR had a higher concentration of epicatechin (almost 67%) than a CCO aqueous fraction (nearly 50%). For the extension subunits, (−)-epicatechin was the only one detected, showing a significant specificity. The calculated mDP for CCR and CCO aqueous fractions were very similar at 4.5. To our knowledge, this is the first time that this methodology was used to describe coconut husk fiber PAs. This result might explain their health benefits cited in previous studies[5−9,29,30] since PAs with a small DP (DP ≤ 4) are the ones absorbable by the gut barrier and mostly responsible for biological activities.[31]

Major compounds from the ethyl acetate fractions were characterized by direct infusion ESI-MS/MS in the negative ionization mode. The PA identification was based on molecular ion masses and the MS2 fragmentation pattern of the most abundant ones (Table S1). PA dimers have three specific fragmentation patterns: Retro–Diels–Alder fragmentation, providing ions at m/z 425.1 and m/z 407.1 (H2O elimination); heterocyclic ring fissions are observed at m/z 451.1; and quinone methide cleavage, producing fragment ions at m/z 287.1 and m/z 289.1. The most abundant ion for both samples was m/z 577.2, which corresponded to a B-type procyanidin dimer, with all the characteristic ions in MS2 fragmentation.[32]

The CCR ethyl acetate fraction displayed a molecular ion at m/z 289.2 and MS2 characteristic fragments of a flavan-3-ol monomer. This finding follows previous reports, which show catechin and epicatechin in CCR and CCO aqueous crude extracts.[5,7,9] It also displayed an m/z 421.3 (289 + 132 Da) characteristic of a flavan-3-ol-pentoside monomer; an m/z 561.2 (289 + 272) for a heterogeneous propelargonidin/procyanidin dimer; an m/z 575.2, characteristic of A-type procyanidin dimer; and an m/z 545.2, indicative of a B-type propelargonidin dimer. Other polyphenolics were also identified, such as monocaffeoylquinic acid (chlorogenic acid), with m/z 353.16, and monocaffeoylshikimic acid, with m/z 335.16 and an MS2 fragment at m/z 179 (caffeoyl ion). Monocaffeoylshikimic acid may be considered as a chemosystematic marker for the Arecaceae family.[33] Also, caffeoylshikimic acid isomers, chlorogenic acid, and dicaffeoylquinic acid were detected in a mesocarp methanolic extract of C. nucifera L., collected in India.[35] The CCO ethyl acetate fraction has also displayed ions at m/z 577.2 (B-type) and at m/z 575.2 (A-type), related to procyanidin dimers. Also, in the CCO fraction, it is possible to observe an ion m/z 561.2 for the heterogeneous propelargonidin/procyanidin dimer, an ion at m/z 289.15 for a flavan-3-ol monomer, and the monocaffeoylshikimic acid molecular ion at m/z 335.18.

Matrix-assisted laser desorption ionization-time-of-flight (MALDI-TOF) analysis of the aqueous fractions proved to be a rich source of structurally essential data. Table displays the peak assignments from the mass spectra (Figures S2 and S3), in which they correspond to sodium ion adducts (+23 Da) in the positive ionization mode. Both varieties exhibited a repetitive pattern corresponding to the (epi)catechin moiety (288 Da). Therefore, the principal [M + Na]+ series corresponded to structures composed exclusively of (epi)catechin units associated with B-type procyanidin. CCR series displayed ion peaks from the trimer up to the hexamer (m/z 889.2, 1177.3, 1465.3, and 1753.4), while CCO displayed the trimer up to the pentamer (m/z 889.2, 1177.2, and 1465.3).

Table 2

CCR and CCO Aqueous Fractions MALDI-TOF [M + Na]+ Mass Spectrum

		CCR	CCO	monomeric units			linkage
DP	calculated m/z (Da)	observed m/z (Da)	observed m/z (Da)	(epi) afzelechin	(epi) catechin	(epi) gallocatechin	A	B
3	855.8	855.2		2	1	0	1	1
	857.8	857.2		2	1	0	0	2
	869.7	869.2	869.1	1	2	0	2	0
	871.8	871.2	871.2	1	2	0	1	1
	873.8	873.2	873.2	1	2	0	0	2
	885.7	885.2	885.1	0	3	0	2	0
	887.8	887.2	887.1	0	3	0	1	1
	889.8	889.2	889.2	0	3	0	0	2
	901.7	901.2	901.1	0	2	1	2	0
	903.8	903.2	903.1	0	2	1	1	1
	905.8	905.2	905.2	0	2	1	0	2
	919.8	919.2	919.1	0	1	2	1	1
	921.8	921.2	921.2	0	1	2	0	2
	935.7	935.2	935.2	0	0	3	1	1
	937.8	937.2	937.2	0	0	3	0	2
4	1142.0		1141.2	2	2	0	2	1
	1144.0	1143.3	1143.2	2	2	0	1	2
	1146.0	1145.3	1145.2	2	2	0	0	3
	1158.0	1157.3	1157.2	1	3	0	2	1
	1160.0	1159.3	1159.2	1	3	0	1	2
	1162.0	1161.3	1161.3	1	3	0	0	3
	1174.0	1173.2	1173.2	0	4	0	2	1
	1176.0	1175.3	1175.2	0	4	0	1	2
	1178.0	1177.3	1177.2	0	4	0	0	3
	1190.0	1189.2	1189.2	0	3	1	2	1
	1192.0	1191.2	1191.2	0	3	1	1	2
	1194.0	1193.3	1193.2	0	3	1	0	3
	1206.0	1205.3		0	2	2	2	1
	1208.0	1207.2	1207.2	0	2	2	1	2
	1210.0	1209.3	1209.2	0	2	2	0	3
	1224.0		1223.2	0	1	3	1	2
	1226.0		1225.2	0	1	3	0	3
5	1432.3	1431.3	1431.4	2	3	0	1	3
	1434.3	1433.3	1433.4	2	3	0	0	4
	1446.2	1445.3	1445.6	1	4	0	2	2
	1448.3	1447.3	1447.3	1	4	0	1	3
	1450.3	1449.3	1449.3	1	4	0	0	4
	1462.2	1461.3	1461.4	0	5	0	2	2
	1464.3	1463.3	1463.3	0	5	0	1	3
	1466.3	1465.3	1465.3	0	5	0	0	4
	1478.2	1477.3	1477.4	0	4	1	2	2
	1480.3	1479.3	1479.3	0	4	1	1	3
	1482.3	1481.3	1481.3	0	4	1	0	4
	1494.2	1493.3		0	3	2	2	2
	1496.2	1495.3	1495.3	0	3	2	1	3
	1498.3	1497.3	1497.3	0	3	2	0	4
6	1720.5	1719.5		2	4	0	1	4
	1722.5	1721.4		2	4	0	0	5
	1734.5	1733.4		1	5	0	2	3
	1736.5	1735.4		1	5	0	1	4
	1738.5	1737.5		1	5	0	0	5
	1750.5	1749.4		0	6	0	2	3
	1752.5	1751.4		0	6	0	1	4
	1754.5	1753.4		0	6	0	0	5
	1766.5	1765.7		0	5	1	2	3
	1768.5	1767.4		0	5	1	1	4
	1770.5	1769.4		0	5	1	0	5
	1782.5	1781.4		0	4	2	2	3
	1784.5	1783.4		0	4	2	1	4
	1786.5	1785.4		0	4	2	0	5

Many other low-intensity peaks were attributable to heterogeneous series. A shift of 16 Da in ion peak series was associated with the replacement of the (epi)catechin unit for an (epi)afzelechin unit (with a 16 Da reduction) or an (epi)gallocatechin unit (with a 16 Da increase). A-type bonded PAs were identified from B-type peaks with a variation of −2 Da.[34,35]

CCR exhibited two [M + Na]+ ion series from the trimer up to the hexamer, both with B-type linkages: the first one with the substitution of (epi)catechin units for one and two (epi)afzelechin units (m/z 873.2, 857.2, 1161.3, 1145.3, 1449.3, 1433.3, 1737.5, and 1721.4) and another with one and two (epi)gallocatechin units (m/z 905.2, 921.2, 1193.3, 1209.3, 1481.3, 1497.3, 1769.4, and 1785.4). Furthermore, one and two A-type linkages were also identified for all the series.

CCO displayed similar [M + Na]+ ion peaks, from the trimer up to the pentamer, with the replacement of the (epi)catechin moiety for one and two (epi)afzelechin units (Figure S3). The same variation was observed for one and two (epi)gallocatechin units. A-type linkage mass shift was also identified for one and two bonds.

The phloroglucinolysis results only showed B-type procyanidins, indicating that they are in higher concentration in the extract. However, the MALDI-TOF analysis found the existence of low-intensity heterogeneous PAs with B- and A-type linkages for both C. nucifera L. varieties, apart from the B-type (epi)catechin series. The diverse types of oligomers, especially heterogeneous PA with one and two A-type linkages, are uncommon.[36]

Our research group’s previous results have determined the ecotoxicity of coconut PAs in assays using seed germination and root elongation as parameters, for which no adverse effect was observed.[37] Associated with the chemical characterization, this result was essential to determine if the extract would be toxic to the environment and whether it could be considered eco-friendly. Once the chemical profile between both varieties is very similar and considering the industrial applications of using coconut husk fibers as a source of PAs, CCO was chosen for the corrosion inhibition assays because of its accessibility.

According to the value of open circuit potential (OCP), it is possible to know if there is an active corrosive process on the metal surface.[38] The screening test to evaluate the CCO crude extract as an eco-friendly corrosion inhibitor affected OCP over 24 h. The control (without the inhibitor) started with −400 mVECS, which indicated an active corrosion process, and reduced its potential to −720 mVECS within 24 h. The different CCO crude extract concentrations maintained the OCP near −700 mVECS during all 24 h. According to the iron pourbaix diagram, values above −600 mVECS in neutral pH have none or just a few corrosive reactions. The OCP screening test to evaluate the CCO crude extract suggests that it could be a green corrosion inhibitor in neutral pH.

The OCP reduction in the absence of the inhibitor (control) is caused by corrosion product formation, promoting a temporary and unstable protector film on a metal surface. The crude extract promoted a protective effect on the metal surface since the beginning of the immersion time, and the OCP valued remained in the metal immunity region.[19,38] The results obtained with OCP monitoring suggest an activity as a corrosion inhibitor. A previous study found that amounts higher than 10 ppm of rice bran oil reduced the OCP for the immunity region for at least 24 h.[39] However, the mechanism of action determination and inhibition efficiency need other techniques such as potentiodynamic curves, linear polarization resistance (LPR) (with Tafel data acquisition), and electrochemical impedance spectroscopy (EIS).[16,18,19]

The potentiodynamic curves (Figure ) and the Tafel data (Table ) are necessary measurements of CCO crude extract electrochemical behavior on the AISI 1020 carbon steel surface under neutral pH solution conditions. Figure shows that the PAs reduced the corrosion potential and corrosion current density, even though the main effect was on the cathodic branch.

Figure 1

Potentiodynamic curves of carbon steel AISI 1020 immersed in a corrosive solution with different crude extract concentrations for 24 h after immersion.

Table 3

Tafel Polarization Parameters of Carbon Steel AISI 1020 in the Presence of Different CCO Crude Extract Concentrations

test	ba (mV/dec)	bc (mV/dec)	Ecorr (V)	Jcorr (A/cm2)	corrosion rate (mm/year)	polarization resistance (Ω)
Control	147.63	536.79	–0.73264	7.60 × 10–5	0.88306	1163.7
0.4 g L–1	107.04	579.12	–0.77564	2.63 × 10–5	0.30571	1657
0.8 g L–1	55.91	183.92	–0.77719	1.13 × 10–5	0.13074	2043.3
1.2 g L–1	35.19	103.61	–0.77121	7.50 × 10–6	0.08711	1521.7
1.6 g L–1	35.47	137.34	–0.74523	5.41 × 10–6	0.06286	2262.8

The addition of different CCO crude extract concentrations (0.4, 0.8, 1.2, and 1.6 gL–1) reduced the corrosion potential, and the difference between corrosion potential of the control (without the inhibitor) and tests (with the inhibitor) indicates the inhibitor action mechanism. Therefore, since the corrosion potential difference was less than 85 mV, the crude extract could be classified as an adsorption inhibitor.[19] PAs isolated from different natural materials have already been described as a green adsorption corrosion inhibitor.[14,24,26,40,41]

After 24 h of immersion, the Tafel parameters were calculated by fitting an LPR curve with NOVA software and are registered in Table . The data show that the cathodic slope (βc) is higher than the anodic slope (βa) for all the conditions tested. This result suggests that the operational conditions promote more corrosion activity on the cathodic branch. Comparing the βc and βa in the presence of 0.8, 1.2, and 1.6 g L–1 CCO crude extracts with the control (without the inhibitor), an impressive reduction of corrosion reactions is observed.

The corrosion inhibitor efficiency is related to the ideal range of molecule concentration in the solution. The lack or excess of corrosion inhibitor can cause noneffective protective film formation by the absence of sufficient molecules or competition over metal surface interaction, respectively.[19] The presence of 0.4 g L–1 CCO crude extract showed results that suggest the occurrence of a more intensive corrosion process. This effect is common in corrosion inhibitor studies, probably due to the lack of a sufficient amount of inhibitor, which promotes an electric potential difference and stimulates the corrosion process.[20,21,39] Moreover, Table also shows a reduction in corrosion current density (Jcorr) and the corrosion rate in the presence of an inhibitor. Unlike Jcorr and the corrosion rate, the polarization resistance increases with the crude extract addition. This behavior of Jcorr reduction and increasing polarization resistance suggests that the inhibitor forms a protector film on the metal surface.[15,19]

The formation of a black film on all working electrode surfaces was observed with the crude extract addition, which is a previously reported phenomenon. This protector film was described in the presence of phenolic compounds extracted from different plant materials tested as an eco-friendly corrosion inhibitor due to molecule deposition and metal surface interaction.[23,24,27]

Two methods were used to calculate the inhibition efficiency: one measuring the weight loss caused by metal dissociation during corrosion reactions (eq ) and the other estimated by Tafel data (eq ). The results for corrosion inhibitor efficiency are illustrated in Figure . The crude extract showed the activity as a corrosion inhibitor with more than 50% of corrosion process inhibition for all concentrations tested (0.4, 0.8, 1.2, and 1.6 g L–1). However, the best concentrations were 0.8, 1.2, and 1.6 g L–1, with similar results in both methods.

Figure 2

Corrosion inhibitor efficiency of the CCO crude extract determined by weight loss (gray bars) and Jcorr (black bars) for carbon steel AISI 1020 in a neutral pH corrosive solution.

The metal surface protection is not only affected by inhibitor concentration. The chemical structure is also an essential factor, and this is independent of the metal type.[13,21−23,26] The chemical structure is related to the stabilization of molecule–metal surface bonding, facilitating the formation of a protective film. Thys, polymeric PAs found in the coconut husk fibers tend to be more effective as corrosion inhibitors than the monomeric ones.[13,21,22]

Usually, the evaluation of green corrosion inhibitors is studied in acid solutions since acid pH facilitates the interaction between natural molecules and the metal surface.[42] However, this study used neutral pH conditions, and even so, the CCO crude extract achieved equal or higher inhibition efficiency compared to studies with acid solutions and a shorter immersion time.[22−24,26,40,41,43]

The effect of PAs on the metal surface was analyzed through EIS displayed as Nyquist (Figure ) and Bode (Figure ) plots. The impedance consists of the relation between the alternating potential and the alternating current in the response of known frequency applications. Moreover, although the impedance is inversely proportional to the current, higher impedance modules promote lesser corrosion susceptibility.[19,44,45]

Figure 3

Nyquist plot for carbon steel AISI 1020 in a neutral pH corrosive solution for different CCO crude extract concentrations (A). The control (B) represents the absence of any inhibitor.

Figure 4

Bode plot for carbon steel AISI 1020 in a neutral pH corrosive solution for different CCO crude extract concentrations. The control represents the absence of any inhibitor. (A) Relation between log frequency (Hz) and phase angle (°); (B) relation between frequency (Hz) and log impedance modules (Ω·cm–2).

Figure shows Nyquist plots for all conditions in the presence of the CCO crude extract (0.4, 0.8, 1.2, and 1.6 g L–1). The semicircles around them suggest the formation of a protective layer, blocking or reducing the effect of corroding molecules of the electrolyte on the metal surface. The absence of a semicircle in the inhibitor indicates an active corrosion process. The semicircle sizes show the capacitive behavior of the PA protector film, related to the corrosion efficiency inhibition.

A previous study showed the effect of 1 M diphenol solution as a corrosion inhibitor for AISI 1020 in acid solution and at different temperatures. The published data described a low efficiency for temperatures below 353.15 K and an impedance module of 150 Ω.[21] Here, we do not evaluate the effect of temperature, but the CCO crude extract (containing polymeric PAs) exhibited an inhibition efficiency of over 90% under neutral pH conditions and room temperature, with impedance modules of 3000 Ω. These results emphasize that larger molecules can promote the formation of a protective film on the metal surface.

The Bode plots showed the metal susceptibility to corrosion and the resistance of electric circuits formed on the metal surface. As a response to the applied known frequencies, these two parameters are demonstrated through the relation between the impedance and phase angle modules. Figure shows the Bode plots in the absence of the inhibitor (control) and different CCO crude extract concentrations (0.4, 0.8, 1.2, and 1.6 g L–1). In the absence of an inhibitor, the curve without a defined phase angle suggests an active corrosion process. In the presence of the CCO crude extract, there is a peak of the phase angles at low frequency, corroborating the protective film formation on the metallic surface hypothesis. Thus, as observed in the Nyquist plot, the presence of the CCO crude extract inhibits the corrosion process.

The high impedance results show 1.2 g L–1 CCO crude extract as the most protective concentration with a higher impedance module, followed by 0.8, 0.4, and 1.6 g L–1 contents. The corrosion inhibition efficiency (based on Tafel data and weight loss) shows similar results for 0.8, 1.2, and 1.6 g L–1 (Figure ). These data evidence that it is impossible to establish a relation between inhibitor content and a protective layer formation on the metal surface. The PAs should have an ideal concentration for carbon steel AISI 1020 protection around 1.2 g L–1. Lower concentrations (0.4 and 0.8 g L–1) probably do not have a sufficient number of molecules to form an efficient protective layer, and at higher concentrations (1.6 g L–1), the molecules probably dispute the metal surface for adsorption, disturbing the protective film formation.

Experimental Procedures

Standards and Chemicals

Solvents used were of HPLC grade, except for the extraction process. Methanol (MeOH), ethyl acetate (EtOAc), hydrochloric acid (HCl), formic acid, acetic acid, n-butanol, and trifluoroacetic acid (TFA) were obtained from a national supplier, Tedia (Rio de Janeiro, Brazil). Deionized water was obtained from a Milli-Q water purification system (Millipore Corporation, UK). Ascorbic acid, phloroglucinol, (+)-catechin, (−)-epicatechin, ammonium ferric sulfate dodecahydrate (NH4Fe(SO4)2·12H2O), 2,5-dihydroxybenzoic acid (DHB), sodium chloride (NaCl), sodium bicarbonate (NaHCO3), sodium sulfate anhydrous (Na2SO4), and calcium carbonate (CaCO3) were purchased from Sigma-Aldrich (USA).

Botanical Samples

C. nucifera L. var. typica A (CCR) was collected in Aracajú, Brazil, and authenticated by Dr. Benedito Calheiros Dias, from Centro de Pesquisas do Cacau, Bahia, Brazil. There, a voucher specimen was deposited under the code CPC 2190. C. nucifera L. var. typica (CCO), commonly known as “Coco-da-Bahia”, was collected in the Campo Experimental de Itaporanga, Embrapa Tabuleiros Costeiros, Sergipe, Brazil. The plant materials were authenticated by the agronomist Humberto Rollemberg Fontes, and a voucher specimen was deposited in the Herbarium of Universidade Federal de Sergipe (ASE 13.631).

Extraction and Purification Procedures

The grounded husk fibers were extracted exhaustively, according to the published protocol.[16] CCR (15.0 g) and 328.38 g of CCO husk fibers were extracted with 1 L of water at 343.15 K, cooled to room temperature, filtered, and then lyophilized, yielding 1.76 and 32.6 g of crude extracts, respectively. Part of the extract (1 g) was dissolved in 500 mL of water and then partitioned with the same volume of ethyl acetate three times. The organic phase was evaporated in a rotatory evaporator, and the aqueous fraction was lyophilized.

Coconut Husk Fiber PA Characterization

PA Quantification with n-Butanol/Hydrochloric Acid

PA content was determined using a published protocol[46] with modifications. First, ground husk fibers (0.2 g in triplicate) were submitted to ultrasound-assisted extraction with acetone/water 7:3 (10 mL) for 10 min. A 5 μL aliquot was added to a sealed test tube containing 3 mL of n-butanol/hydrochloric acid (95:5) and 100 μL of a 2% solution of NH4Fe(SO4)2·12H2O in 2 N hydrochloric acid and then heated for 60 min (368.15 K). After cooling, each replicate absorbance was measured at 550 nm (UV-1601 Shimadzu spectrophotometer), and the PA concentration was calculated in the samples by linear regression.

Direct Infusion ESI-MS/MS Analysis of Ethyl Acetate Fractions

The ethyl acetate fractions were analyzed by direct infusion on an Amazon SL ESI-Iontrap spectrometer (Bruker). The samples (1 mg) were diluted 1:100 with methanol and analyzed in the negative mode. Infusion flow rate: 3 μL/min, nebulizer pressure: 10 psi, N2: 5 L/min, and source temperature: 473.15 K.

MALDI-TOF-MS Analysis of Aqueous Fractions

A MALDI-TOF Autoflex Speed Bruker spectrometer was used for MALDI-TOF analysis using an already published methodology[36] with modifications. Samples (2 mg) were dissolved with 0.1% aqueous TFA and diluted to 1:10 with DHB matrix solution (0.1% TFA). Aqueous sodium chloride (1 mg/mL, 1 μL) was added to the solution. The samples were applied to the MALDI plate (duplicate) and analyzed after drying. The analyses were made in the positive mode, using a peptide calibration standard (Bruker). Data were processed using mMass 5.5.0 software (Strohalm M.).

Phloroglucinolysis of Aqueous Fractions and Sequential HPLC Analysis

Firstly, the analysis followed an already published methodology.[47] The phloroglucinol solution was prepared using phloroglucinol (50 g L−1), ascorbic acid (10 g L−1), and 0.1 N HCl in methanol up to 100 mL. A NaHCO3 solution (40 mM) was prepared with Milli-Q water up to 100 mL. The aqueous fraction (5 mg, in triplicate) was weighted in a capped glass test tube, and then, 1 mL of the phloroglucinol solution was added and heated for 20 min, in a water bath at 323.15 K. An aliquot of 200 μL was transferred to a 2 mL vial, and 1 mL of the NaHCO3 solution was added. HPLC-DAD analysis occurred in an Agilent 1200 series, and a ReproSil-Pur RP-18 column (250 × 4.6 mm, 5 μm, Dr. Maisch GmbH, Germany) with a guard column on the same material was used, flow rate: 1 mL min−1, λ = 280 nm. Mobile phase (A) was aqueous 1% acetic acid and (B) was methanol. Gradient mode: 5% B for 10 min, 5–20% B in 20 min, 20–40% B in 25 min, 90% B for 10 min, and 5% B for 5 min. The mean DP (mDP) was calculated as the sum of all subunits (flavan-3-ol monomer and phloroglucinol adducts, in moles) divided by the sum of all flavan-3-ol monomers (in moles).

Corrosion Inhibitor Evaluation

Corrosion Testing Solution

For experimental procedures, the lyophilized extract of C. nucifera L. var. typica (CCO) was resuspended in distilled water and sterilized by membrane filtration (0.22 mm pore). A solution composed of 500 ppm of chloride (829 mg/L NaCl), 150 ppm of sulfate (222 mg/L Na2SO4), and 150 ppm of calcium carbonate (CaCO3) was used as a corrosive environment, simulating water from a cooling system. The pH solution was adjusted and maintained in a neutral range (7.0 ± 0.2) during the tests. The corrosion inhibition activity was analyzed by comparing results obtained in the absence and presence of various amounts of crude extract added to the corrosion test solution. The amounts of inhibitor added were 0.4, 0.8, 1.2, and 1.6 g L–1.

Weight Loss Method

Weight loss is a simple method to measure the corrosion rate for any metal under any environmental condition. Before the test, the coupons of carbon steel AISI 1020 were mechanically polished using SiC papers of different grades (80–600 grit), washed with distilled water and acetone, dried, and weighted. Previously, different solutions with pH adjustment to 7.0 (±0.2) were prepared, one without the inhibitor (control) and four with the crude extract (0.4, 0.8, 1.2, and 1.6 g L–1).

The carbon steel AISI 1020 coupons were immersed in the test solutions (each test was performed in triplicate) at 298.15 ± 2 K for 24 h. After this time, the coupons were removed from the solution, immersed in HCl concentrated for 15 s, washed with distilled water and ethanol, dried with cold air, and weighted. The initial and final weights were used to calculate the inhibition efficiency using eq , where IE%WL is perceptual of inhibition efficiency calculated with weight loss data and W0 and W′ are the mass values without and with the coconut crude extract, respectively.[15,48]

Electrochemical Analysis

The tests were carried out in 100 mL glass cells, maintained at controlled temperature (298.15 ± 2 K) and slow shaking for 24 h. The electrochemical performance of the coconut husk fiber crude extract as a green corrosion inhibitor was evaluated in a conventional three-electrode cell connected to an Autolab PGSTAT302N potentiostat/galvanostat. The corrosion test solution was used as the electrolyte, and an AISI 1020 carbon steel (with 1 cm2 of the exposed area), a platinum electrode, and a KCl-saturated calomel electrode (SCE) were selected as working, reference, and counter electrodes, respectively. Before all tests, the working electrodes were mechanically polished using SiC papers of different grades (80–600 grit).

The corrosion behavior of the working electrodes for each coconut crude extract concentration was monitored by OCP, potentiodynamic polarization curve, LPR, and EIS.

OCP data were used for screening the coconut crude extract as a corrosion inhibitor. Then, the working electrode was immersed in solutions without and with the CCO crude extract (0.4, 0.8, 1.2, and 1.6 g L–1) in neutral pH (7.0 ± 0.2), and the OCP was measured for 24 h.

The potentiodynamic polarization curves were carried out by scanning the working electrode from −800 to −300 mVSCE at a scan rate of 0.33 mV s–1. Moreover, the LPR, used for Tafel data acquisition, was performed from −20 to 20 mVSCE around the corrosion potential at a scan rate of 0.33 mV/s. The corrosion current density obtained by LPR was used for estimating the inhibition efficiency with eq , where IE%T is perceptual of inhibition efficiency calculated with Tafel data and J0 and J′ are the corrosion current density values without and with the coconut crude extract, respectively.

The impedance of alternating current was obtained at the OCP in a frequency range of 0.01–10000 Hz with 10 mV of amplitude. The data obtained were illustrated by the Nyquist plot.[10,28,29]

Conclusions

Husk fibers of two C. nucifera L. varieties exhibited 3% (w/w) of PAs.

C. nucifera L. PAs exhibited an mDP of 4.5, majorly composed of (−)-epicatechin.

Mass spectrometry also identified A- and B-type linkages, heterogeneous PAs, and other phenolics for both varieties.

PAs showed promising results as an eco-friendly corrosion inhibitor for carbon steel AISI 1020 under neutral pH conditions.

PAs exhibited a protective film on the carbon steel AISI 1020 surface, which indicates that they are adsorption inhibitors.

The inhibition efficiency was more than 90%, suggesting the husk fiber PAs as an eco-friendly corrosion inhibitor.