Inhibition Effect of Glycerol on the Corrosion of Copper in NaCl Solutions at Different pH Values

The inhibitory effect of glycerol on copper corrosion in aerated NaCl (0.5M) solutions at three pH values (4, 7, and 10) was evaluated. Inhibition efficiencywas assessedwith conventional electrochemical techniques: open circuit potential, potentiodynamic polarization, and electrochemical impedance analysis. Glycerol reduced the corrosion rate of copper in NaCl solutions. The best inhibition effect (η ≈ 83%) was produced in alkaline (pH 10) chloride media. This effect can be ascribed to increased viscosity and the presence of copper-glycerol complexes.


Introduction
As worldwide biodiesel production increases so does production of the byproduct glycerol [1,2].For every gallon (3.78 L) of biodiesel produced approximately 0.3 Kg glycerol results.By 2016, the global biodiesel market is expected to be 37 billion gallons [3], consequently resulting in production of 11.1 billion kilograms of glycerol.This looming glycerol glut means that new uses for glycerol need to be found or conversion processes developed to convert it into more valuable chemicals [4].Finding new uses for glycerol would also help to lower the cost of biodiesel, the main factor behind decreasing production rates.Electrochemical studies have shown glycerol to be useful as an additive in electrochemical baths for coating formation [5][6][7].It has also been used in combination with traditional corrosion inhibitors to improve efficiency, but very few studies have focused on glycerol alone as a corrosion inhibitor.Shaker and Abdel-Rahman [8] reported that the corrosion rate of copper decreases at higher glycerol proportions in water-glycerol solutions.The authors show that the reactions were controlled by diffusion.A key aspect of glycerol is its viscosity and its potential to form a metal-glycerol complex.Viscosity in water-glycerol solutions increases as glycerol concentration increases [9]; therefore, an increase in viscosity can also be expected to decrease mass transfer of ions.A decrease in oxygen concentration can affect the cathodic reaction.In alkaline solutions, glycerol is known to form a Me-glycerol complex (Me = Zn, Fe, and Cu) [10,11] which may act as a barrier to inhibit corrosion of metals.The present study aim was to evaluate the inhibitory effect of glycerol (Figure 1) alone on corrosion of copper exposed to NaCl solutions at three pH values.

Experimental Procedure
Deionized water and reagent grade NaCl, HCl, and NaOH were used to prepare NaCl solutions with three different pH values.Glycerol (99.5%) was purchased from Sigma-Aldrich (CAS 56-81-5).Samples were cut from a pure (99.999%) copper rod (Goodfellow, 5.0 mm diameter) and embedded in epoxy resin.A bar cross section was exposed to the NaCl solution.Before each exposure, the sample was abraded with a series of different grade emery papers (up to 1200), rinsed with water and ethanol, and dried with hot air.
All electrochemical measurements were done using a standard three-electrode cell configuration.Electrochemical experiments were run in the following order: open corrosion potential (OCP), electrochemical impedance spectroscopy (EIS), and potentiodynamic scan.All experiments were carried out using a Gamry PCI4 Potentiostat/Galvanostat/ZRA instrument with CMS 100 and 300 software.OCP was measured for 15 min prior to each impedance experiment, which was done at 10 kHz to 20 mHz with a 10 mV peak-topeak amplitude using ac signals at OCP.The potentiodynamic scan was begun immediately after EIS by scanning from −0.25 to 0.25 V versus OCP at a 1 mVs −1 sweep rate.A standard calomel electrode (SCE) was used as a reference in all electrochemical experiments.Impedance data were examined for causality, stability, and linearity with the Kramer-Kronig relationship using the method described by Boukamp [12] and adapted in the Gamry Echem analyst software.
A scanning electron microscope (SEM, Philips XL30 ESEM) was used to examine any inhibitory effect of glycerol on copper corrosion.Samples examined with SEM were abraded with emery paper (up to 2000), polished with a 0.25 m diamond solution, rinsed with water and acetone, and dried with hot air.Copper samples were immersed for 120 hours in 0.5 M NaCl (pH 7) solution with and without 2 M glycerol.

Results and Discussion
The OCP ( oc ) recorded as a function of time for copper in NaCl (0.5 M) with and without different glycerol concentrations at three pH values shows that overall the presence of glycerol at pH 4 shifted the OCP towards negative potentials (Figures 2(a)-2(c)).The drop in OCP during the first 200 seconds is frequently ascribed to dissolution of a native oxide previously formed on the copper surface as a result of contact with the atmosphere.A shift to the negative region is also an indication of an active surface.The general behavior of  oc at pH 7 consisted of a shift to noble potentials and was similar with or without glycerol (Figure 2(b)).At pH 7, increasing glycerol concentration had no effect on OCP.At pH 10, glycerol concentration had a clear effect on  oc , causing it to move gradually toward the positive potential region and indicating formation of corrosion products or adsorption of species onto the copper surface.
The polarization curves for copper in NaCl (0.5 M) solutions with and without glycerol in different concentrations at three pH values showed how the Cu corrosion potential ( corr ) changed slightly from negative to positive as a function of pH (Figures 3(a)-3(c)).This trend coincided with the shift observed for OCP at all pH values.The cathodic branch of the polarization curves exhibited behavior typical of a reduction reaction with mass transfer limitations.Reduction of dissolved oxygen with formation of hydroxide [13], or reduction of water molecules, is commonly responsible for the cathodic reaction [14].Another cathodic reaction in acid is the reduction of oxygen to form water molecules.Reduction of hydrogen is an unlikely cathodic reaction because hydrogen is more active than copper in the electromotive series.In the present results, neither pH nor glycerol concentration affected the shape of the cathodic branch (Figures 3(a)-3(c)), indicating that pH and glycerol did not influence cathodic reaction type.Glycerol concentration did reduce cathodic current very slightly, a reaction more noticeable at pH 10.This reduction may be explained by reduction of mass transfer due to increased viscosity as glycerol concentration increased.
No significant change was observed in the polarization curve anodic current at pH 4, indicating that the anodic reaction was unaffected by addition of glycerol.Glycerol's effect became apparent at pH 7 and even more obvious at pH 10.In all cases, a linear portion of the polarization curve was not easily determined.Extrapolation of both sides of the polarization curves to calculate corrosion parameters is preferable, although occasionally just one side is used, mainly when a Tafel linear region is too small or is difficult to observe  [15].Therefore, in the present study the cathodic branch was chosen to calculate corrosion parameters using the Tafel extrapolation method [16,17].Of note is that at pH 10 the anodic branch exhibited a major reduction in anodic current as glycerol concentration increased.The corrosion parameters calculated from extrapolation of the cathodic branch showed that, at all pH values, corrosion current decreased as glycerol concentration increased (Table 1).The percentage inhibition efficiency was calculated using corrosion current and the following equation [18,19]: where  0 corr is the corrosion current measured in a NaCl solution in absence of glycerol and  corr is the corrosion current measured in a NaCl solution containing glycerol as inhibitor.
The best inhibition efficiency was attained at pH 10, possibly due to formation of a layer of copper-glycerol complexes near the copper surface, effectively reducing the reaction area and producing an inhibitory effect [11].The two main copper-glycerol complexes are CuGl(OH) 3 −2 and CuGl 2 (OH) 2 −2 , where Gl − is a glycerol anion [20].In the example shown in Figures 4(a)-4(c), the calculated Kramer-Kronig data fit well in both the Bode and Nyquist diagrams.Residual error was within 0.5%, and goodness-offit was 24.86 × 10 −6 .However, residual error was generally NaCl ( <0.5% for both acid and neutral NaCl solutions.Residual error in alkaline (pH 10) solutions (with and without glycerol) was <1.5%.This increase in residual error was probably caused by formation of copper-glycerol complexes on the copper surface.An increase in error reported in the literature was ascribed to passive film formation [21]; in this study, goodness-of-fit in all solutions was on the order of 10 −6 .The Nyquist impedance plots for copper in NaCl (0.5 M) solutions with and without glycerol at three pH values show a depressed semicircle the diameter of which can be obtained by extrapolating toward the low frequency value limit ( → 0), which increases with addition of glycerol (Figures 5(a)-5(c)).Semicircle diameter is correlated to the polarization resistance (  ) value, which is considered inversely proportional to corrosion rate.Then, independent of the equivalent circuit used to model the EIS data, the glycerol's inhibitory performance can be observed by comparing   from the diameter of the impedance spectra.Various equivalent circuits have been reported for modeling impedance data of copper in NaCl solutions [20][21][22].The equivalent circuit shown in Figure 6 was used in analysis of the Nyquist plots because it has been used to study coating for corrosion protection, pitting corrosion [22], or copper corrosion in the presence of inhibitors [23][24][25][26].Fitting the equivalent circuit model produced estimated parameters (Table 2).  represents the solution resistance, while CPE 1 and CPE 2 are constant phase elements that absorb inhomogeneities of the double layer capacitance and  improve model fit.The characteristic parameters of the constant phase elements are  and .Estimated   from the semicircle diameter corresponded to the impedance of an anodic reaction occurring in two stages with magnitudes represented as  1 +  2 [27].Although an exact difference between both resistances has been neither fully understood nor fully described [28], a number of studies use these parameters for modeling corrosion behavior of copper in the presence of an inhibitor.Inhibition efficiency was calculated using polarization resistance ( 1 +  2 ) and the following equation [19,29]: where  0  is the polarization resistance measured in a NaCl solution in the absence of glycerol and   is the polarization resistance measured in a NaCl solution containing glycerol as inhibitor.
The fit was generated using the equivalent circuit in Figure 6.Typical results of fitting analysis are shown in Bode (Figure 7(a)) and Nyquist (Figures 7(b)-7(c)) plots (symbols represent measured data and solid lines represent fitted curves).In general, there was a good fit between the calculated and experimental impedance data.
The most significant parameters are total polarization resistance (  =  1 +  2 ), the constant phase element, which represents the double layer capacitance (CPE 1 ), and inhibition efficiency.Other parameters also exhibit glycerol's inhibitory effect but are less notable.Polarization resistance increased and CPE 1,2 values decreased in response to pH levels and glycerol concentration (Table 2).The   and CPE behaviors agree with the potentiodynamic polarization measurement results, and the values of both parameters suggest the presence of a film thickening process that could act as a barrier to corrosion and thus be responsible for the inhibitory effect.Presence of a film was more evident at pH 10; indeed, the best agreement between the inhibition efficiency values determined by the Tafel extrapolation of the cathodic current and the impedance data was observed at pH 10.These findings support the assumption of formation of a copperglycerol complex on the copper surface.
Comparison of the inhibition efficiency obtained by potentiodynamic polarization and the impedance data shows that both results follow the same trend.The very slight difference in values was probably caused by differences in measurement time [30].The order of the electrochemical measurements was open circuit potential, impedance, potentiodynamic polarization, meaning the layer of adsorbed molecules during each measurement would be slightly different.However, impedance was most effective at demonstrating glycerol's inhibitory effect on copper corrosion.
Micrographs (SEM) of copper samples exposed to 0.5 M NaCl at pH 7 with and without added glycerol showed no visible indications of ion chloride attack on the copper surface in the presence of glycerol (Figure 8(a)).In contrast, the copper exposed to NaCl in the absence of glycerol exhibited evidence of localized corrosion, such as micro-and macropitting across the entire copper surface (Figure 8  These observations agree with the electrochemical results and confirm that glycerol effectively inhibits copper corrosion.

Conclusions
Glycerol inhibits corrosion of copper in aerated NaCl (0.5 M) solutions.Inhibition efficiency increases with increasing glycerol concentration and is especially notable at high pH values.Glycerol does not affect the anodic or cathodic reactions, suggesting that increased viscosity reducing mass transport apparently explains the reduction in corrosion rate from acid to neutral solutions.The reduced anodic corrosion rate at pH 10 is probably due to two causes: an increase in solution viscosity and presence of a film on the copper surface, most likely composed of copper-glycerol complexes.SEM images confirm the effectiveness of glycerol as an inhibitor of copper corrosion in NaCl solutions.

Figure 3 :
Figure 3: Polarization curves for Cu in NaCl (0.5 M) solutions with and without glycerol at three pH values: (a) pH 4, (b) pH 7, and (c) pH 10.

Figure 4 :
Figure 4: Typical Kramer-Kronig analysis results for Cu in NaCl (0.5 M, pH 10) solution containing glycerol (2 M).(a) Nyquist diagram, (b) Bode diagram, and (c) residual error (Δ/) for the real and imaginary components of the impedance data.

Figure 8 :
Figure 8: SEM micrographs of copper surface after 120 hours of immersion at room temperature (a) in aerated 0.5 M NaCl (pH 10) + 2 M glycerol and (b) in aerated 0.5 M NaCl (pH 10).

Table 1 :
Electrochemical parameters determined by Tafel extrapolation of the cathodic branch of the potentiodynamic polarization curves for copper in aerated NaCl (0.5 M) with and without glycerol at three pH values.

Table 2 :
Electrochemical impedance parameters obtained by fitting the Nyquist plots for Cu in aerated NaCl (0.5 M) with and without glycerol at three pH values.