The influence of organic structure and rare earth metal cation on the corrosion efficiency observed on AS 1020 steel compared with La ( 4 OHCin ) 3

Whilst the corrosion protection of steel in aqueous chloride environments by the rare earth inhibitor lanthanum 4-hydroxycinnamate is well known, the influence of the structural variation of the organic component as well as the nature of the metal centre has not previously been addressed. Herein we show that praseodymium 4-hydroxy cinnamate is comparable to its lanthanum counterpart in aqueous solution. On the other hand, cerium 4-hydroxycinnamate and lanthanum 2-hydroxycinnamate show poor corrosion protection performance while lanthanum 3-hydroxycinnamate provides a level of inhibition between these. These differences are shown to be related to the speciation in solution and are postulated to be linked to steric influences which are likely to affect the bonding environment within the rare earth compound itself, as well as its bonding with the steel substrate.


Introduction
In order to ameliorate the deterioration of metal substrates from the effects of harsh environments, environmentally friendly inhibitors are an ideal alternative to the carcinogenic and toxic chromates [1−4].Such inhibitors are able to reinforce the metal oxide film or alternatively deposit complexes onto the metal substrate [4,5,6].There is an increased demand for more environmentally friendly and cost effective protection against corrosion; such alternatives include inorganic compounds, e.g.nitrates, molybdates, and rare earths, and organic compounds such as carboxylates, phosphonates, phosphates and sulphonates [7,8,9].Over the past decade Forsyth and co-workers have developed a range of rare earth organic compounds that show promise as environmentally friendly corrosion inhibitors for steel substrates, in particular lanthanum 4-hydroxycinnamate [10−16].
Recently, lanthanum 4-hydroxycinnamate in solution has shown remarkable corrosion inhibition properties with mild steel substrates [15].This compound performs better as an inhibitor of steel at higher concentrations and under alkaline conditions.Under acidic pH conditions dissociation of the lanthanum 4-hydroxycinnamate, (La(4OHCin) 3 , occurs where the predominant species are lanthanum chloride [LaCl 4 ] -and 4-hydroxylcinnamic acid as separate entities in solution.This leads to a rather low mixed inhibition effect compared with the strong anodic inhibition observed at higher pH values when the complex [LaL 4 ] -is present [15].The observation that slight cathodic inhibition was more evident under acidic conditions (except for high concentrations, i.e 1000 ppm where the inhibition was via both anodic and cathodic processes), and that this correlated with an excess of the LaCl 4 -species, strongly suggests that the speciation of La(4OHCin) 3 is critical to the anodic nature and high efficiency of the inhibition and confirms the synergy of the organic and REM components previously suggested by Forsyth et al. [17].
It was certainly interesting to discover that this inhibitor might be capable of functioning well over a wide pH range, and that the inhibition mechanism could also vary over that range.
The study of the speciation for La(4OHCin) 3 and its influence on the corrosion inhibition efficiency for steel substrates over a wide range of environmental conditions has provided an in depth perception of the working mechanism of the inhibitor.The aim of this current work was to investigate the influence of the position of substitution of the aromatic ring of the organic component of the inhibitor, and the influence of the rare earth metal coordinated by the cinnamate ion.This included analysis of lanthanum 2-hydroxycinnamate (La(2OHCin) 3 ), lanthanum 3-hydroxycinnamate (La(3OHCin) 3 ), praseodymium 4-hydroxycinnamate (Pr(4OHCin) 3 ) and cerium 4-hydroxycinnamate (Ce(4OHCin) 3 ).These studies could lead to more clever choices of inhibitor combinations, and a better understanding of the influence of chemical structure and bonding which could allow design of improved inhibition properties for this type of inhibitor compound.

Chemicals and Materials
The x-hydroxy cinnamic acid, where x is 2, 3 and 4, and the lanthanum, praseodymium, and cerium chlorides were purchased from Sigma Aldrich.The rare earth 4-hydroxycinnamates were prepared according to already published methodology [18] while the La(2OHCin) 3 and La(3OHCin) 3 were prepared by similar metathesis synthesis between the corresponding sodium salts and lanthanum chloride, refer to Electronic Supplementary Information (ESI) for experimental details, including the powder X-ray diffraction pattern for La(3OHCin) 3 .All test solutions were prepared using reagent grade sodium chloride (Chem Supply), sodium hydroxide (Merck) and hydrochloric acid (Merck) solutions.The use of deionised water for the preparation of all solutions was adopted.The working electrodes used for the potentiodynamic polarisation experiments were machined from a cylindrical mild steel rod, Australian standard AS1020, to a 1.0 cm diameter (0.78 cm 2 ).These electrodes were coated with an epoxy resin and attached to a teflon holder.The steel coupons used for the constant immersion corrosion tests were prepared also from a 2.0 cm diameter mild steel rod, AS1020, to a thickness of 0.3 cm.

Electrode and Coupon Preparation
The working electrodes for use in the potentiodynamic polarisation studies and the corrosion test coupons were abraded with silicon carbide paper to a P4000 grit finish from P320 grit.They were washed with tap water followed by DI water, and dried using nitrogen gas.Potentiodynamic polarisation and corrosion tests were carried out immediately after specimen preparation, which aided reproducibility.
The test solutions were prepared by dissolving the appropriate weights of each rare earth x-hydroxy cinnamate in 1 L to make concentrations of 400 and 600 ppm in 0.01 M sodium chloride solution.The pH of the solutions was adjusted using 0.01 M hydrochloric acid solution to achieve pH 2.5.The unadjusted pH of the prepared solution was pH 5.5.The solutions were exposed to a temperature-controlled laboratory atmosphere at approximately 22 °C.

Potentiodynamic Polarisation
Using potentiodynamic polarization experiments it is possible to determine the inhibitive nature of the rare earth cinnamates at different concentrations and solution pH in an open-to-air environment.
All polarisation experiments were carried out open-to-air using 100 mL of inhibitor solution in a standard three-electrode system consisting of a saturated calomel reference electrode, a titanium mesh counter electrode, and the working electrode.The cell assembly was located in a Faraday cage to prevent electrical interference.The exposed area of the working electrode was 0.78 cm 2 .After immersion the open circuit potential (OCP) was monitored for 12 hours, and then the potentiodynamic polarisation experiment was carried out at a scanning rate of 0.1667 mV s -1 .The scans started at 200 mV more negative than E corr and continued through E corr for a range of 1.0 V in the positive direction.Polarisation curves were replicated in triplicate to assess reproducibility.All electrochemical measurements were conducted using a Biologic VMP multichannel potentiostat with EC-Lab v 10.32 software used for analysis.The current density (i corr ) values were calculated using Tafel extrapolation methods.Where anodic and cathodic curves were near linear and symmetrical within +/− 50 mV from E corr , both Tafel slopes were extrapolated until the lines intersected at E corr .Where polarisation occurred for the cathodic reaction under acidic conditions, the slope for the cathodic region was extrapolated until it intersected the E corr value.

Nuclear Magnetic Resonance Spectroscopy (NMR)
A Jeol 270 MHz spectrometer was used to collect 139 La (38.2 MHz) NMR spectra.A double pulse spin echo experiment was used with a 180° pulse duration of 26.4 µs followed by a 90° pulse duration of 13.2 µs.Spectra were collected using 4096 transients, a 0.36 s acquisition time, and a 0.30 s relaxation delay.Each spectrum had the subsequent processing parameters applied using MestReNova 7.0.3:smoothing via Whittaker Smoother method, phase correction, and baseline correction.Peak widths at half height were measured.All NMR spectra are reported as their corresponding chemical shifts δ.A solution of 0.01 M lanthanum chloride in D 2 O was used as an external standard to reference the 139 La chemical shifts.
NMR experiments were run at room temperature in 10 mm NMR tubes for the 400 ppm lanthanum x-hydroxycinnamate solutions at pH 5.5.

Electrospray Mass Spectroscopy (ESMS)
An Agilent Technologies LC/MSD TOF with acetonitrile as the mobile phase with a flow rate of 25 μL min -1 was used to collect mass spectrometry data.All inhibitor solutions were directly injected into the spectrometer, with nitrogen gas used for nebulisation.Data were collected using Agilent MassHunter Workstation Data Acquisition software and analysed using Agilent MassHunter Qualitative Analysis software.

Anodic polarisation of steel in inhibitor solutions
Typical potentiodynamic polarisation curves for the inhibitors at 400 ppm and 600 ppm at pH 2.5 and 5.5 revealed that they all suppressed the anodic reaction to some degree (Figures 1-4).At these concentrations for pH 2.5, only a slight shift of E corr was observed toward more noble potentials, except for La(2OHcin) 3 which showed a slight cathodic shift in the potential.This is most likely due to the dissociation of the inhibitor into its rare earth chloride and cinnamate ions, a similar trend was seen in previous work for La(4OHCin) 3 at pH 2.5 [15].According to Table 1 and 3, E corr and I corr values calculated from the potentiodynamic polarisation curves for the inhibitors at concentrations of 400 and 600 ppm and pH 2.5 show that La(4OHCin) 3 was by far the best inhibitor under these conditions with I corr being reduced compared to control solution values (Figure 1).Interestingly, the Pr(4OHCin) 3 and La(3OHCin) 3 showed comparable behaviour, to the results for the La(4OHCin) 3 .In contrast however, the E corr and I corr values calculated for the La(2OHCin) 3 and Ce(4OHCin) 3 are very poor, being almost the same as for the control solution values.
Increasing the pH of the inhibitor solution to pH 5.5 gave a much clearer indication of the trend seen from the potentiodynamic polarisation curves found at pH 2.5, suggesting that the Pr(4OHCin) 3 and La(3OHCin) 3 were almost as efficient as the La(4OHCin) 3 (Figure 2 and 4).This is supported by the Tafel extrapolation data collected at both concentrations, as presented in Table 2 and 4

Characterisation of species in inhibitor solutions
Previous work has shown that the use of NMR spectroscopy in combination with electrospray mass spectrometry gives valuable information about the species present in solution in a given environment [15].
NMR is a powerful chemical characterisation tool and has been used to evaluate the change in solution speciation of the La(4OHCin) 3 under various pH and concentration conditions [15].However, NMR spectroscopy has limitations in that it can only analyse NMR active nuclei, therefore in this case only the lanthanum species can be evaluated as the cerium and praseodymium nuclei are not active [19].A spectrum of lanthanum chloride in D 2 O was used as a reference where the signal resonates at δ 4.89 ppm.From Figure 5 it can be seen that there was a clear shift to higher resonances as the position of the hydroxyl group was changed from 4-hydroxy-to 3-hydroxy-and then the 2-hydroxycinnamate respectively.
Previous work has shown that an increase in concentration of the La(4OHCin) 3 at any given pH results in a shift to higher frequencies and increases the broadness of the peak [15].It was concluded from the speciation study of the La(4OHCin) 3 that a shift to higher frequencies and increasing broadness were an indication of the presence of a number of different species in solution, and that the exchange between the species could lead to short relaxation times (T2) and lifetime broadening [20,21].Table 5 presents the observed shifts found for each solution of lanthanum 2, 3, and 4-hydroxy cinnamate at a concentration of 400 ppm in a solution of pH 5.5 in 0.01 M NaCl.These data clearly indicate that, the resonance shifts to higher frequencies and the peak width at half height increased in going from La(4OHCin) 3 to La(2OHCin) 3 .The shift in NMR frequency signifies a change in the lanthanum environment possibly brought about by differences in the coordination.The changing width could indicate an increasing number of species and possible exchange between these.Being a quadrupolar nucleus, the width could also indicate an increasingly asymmetric environment.This Volume 2, Issue 1, 1-15.
speciation behaviour was further investigated using ESMS as discussed below.12.5 500 La(4OHCin) 3 9.8 470 ESMS measures the mass-to-charge ratio of charged particles.It has been a valuable analytical technique to determine the mass of particles and the elemental composition of a sample in our previous work [15].The behaviour of the chemical equilibrium will change when there are changes to the reaction conditions, for example a change in the concentration, and therefore a change in the mass-to-charge ratio of the charged species is expected to be observed.Table 6 presents a summary of the rare earth complex species that are expected in the solutions investigated here.Each species containing rare earth metal and/or chloride ions will appear with a particular isotope pattern, these clusters found in the mass spectra, in ESI Figures S1−S5, have been supported by simulated mass spectra and can be found in ESI Figure S6 to S14.Additionally, all reported peaks are the main peak of the associated cluster.
ESI Figures S1 to S5 show, the electrospray mass spectrum obtained for 600 ppm La(xOH-Cin) 3 and RE(4OHCin) 3 (RE = Ce, Pr) at pH 5.5.This data demonstrates the change in the mass-to-charge ratio of the species present according to the expected species in Table 6, and also shows that not all inhibitors have the same species present under the given conditions.Previous work published for the speciation work on La(4OHCin) 3 suggested the presence of species as listed in Table 6 (given the mass-to-charge ratios) in varying amounts depends on environmental conditions.An equilibrium between these species was postulated as shown in Equation (1).It was found that the concentration and pH condition was linked to the species present in solution (also supported by NMR spectroscopy).The highest inhibition efficiency observed from potentiodynamic polarisation correlated with the speciation tending towards the right of this equilibrium.This suggests that when species such as [LaL 3 Cl] -and [LaL 4 ] -(L = xOHCin) were present, the observed inhibition was greater than when they were absent.
It was found from the spectrum of La(2OHCin) 3 (ESI Figure S4) that the majority of the species present at this particular concentration and solution pH is of lower molecular weight, and consists mainly of species that lie to the left of the equilibrium reaction (Equation 1).In particular and worth noting is the low concentration of the [LaL 4 ] -species at 791 mass units.Some of the low molecular weight species present in the spectrum have been attributed to species containing lanthanum and hydroxides and may include species such as LaL 2 (OH) 3 (H 2 O) 6 Cl] 3-, [LaL 2 (OH) 3 (H 2 O) 6 Cl 2 ] 4-and [LaL 2 (OH) 3 (H 2 O) 6 Cl 3 ] 4-.These species not prevalent in the electrospray mass spectrum of La(4OHCin) 3 .
The electrospray mass spectrum obtained for La(3OHCin) 3 (ESI Figure S3) also shows a number of different species, not unlike the La(2OHCin) 3 spectrum.However, the apparent concentrations of these species were more evenly distributed along with the high mass-to-charge ratios of species such as [LaL 3 Cl] -and [LaL 4 ] -found in the La(4OHCin) 3 spectrum.
As can be seen from the mass spectrum obtained for Pr(4OHCin) 3 (Figure 6), it appears similar to that found for La(4OHCin) 3 from previous studies.In particular, the peak at 793 mass units is representative of [PrL 4 ] -which appears to be the species contributing to the best inhibition seen for mild steel.The spectrum collected for Ce(4OHCin) 3 (ESI Figure S1) appears to be very different from the other compounds; there are fewer species present in solution, which may indicate that the solution equilibria for this inhibitor are quite different.Worth noting at this point is that the solubility of the Ce(4OHCin) 3 is high in comparison with the other compounds.

Hypothesised mechanism of action of inhibitors of steel
We have previously discussed the La(4OHCin) 3 inhibition behaviour in the context of the interaction on the metal surface [13,14,15], where a bimetallic film involving the lanthanum complex and the underlying iron from the steel surface forms, and in turn reduces the rate of further metal dissolution at these sites.However, the corrosion is not completely inhibited and some level of corrosion still occurs, leading to changes in the local pH.This change in pH, in particular at the cathodic sites, leads to the hydrolysis of the La(4OHCin) 3 species forming a dynamic mixture of lanthanum hydroxide, iron oxide and bimetallic species.
This mechanism was supported in previous work using Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) vibrational spectroscopy [16,17].This technique proved to be a valuable tool in developing an understanding of the nature of the species on the inhibited surface [22], allowing the examination of the inhibiting complexes attached to the surface without the need to remove the film.
In the current study, the reduction in the observed I corr from La(2OHCin) 3 , Ce(4OHCin) 3 , La(3OHCin) 3 , Pr(4OHCin) 3 to La(4OHCin) 3 may be related to an increase in the availability of hydroxyl groups as donors in bonding to neighbouring molecules on the steel surface, which therefore influences the bimetallic complex formed.
Based on the schematic protective film shown in Figure 7, the orientations of the hydroxyl group are important for further interactions beyond the initial layer of bimetallic complex.As described in previous reports, the terminal hydroxyl group from the 4-hydroxycinnamate ligand (Fig 7a) is capable of intermolecular interactions (through hydrogen bonding) to nearby molecules of the inhibitor.These effectively increase the thickness of the protective film and result in the precipitation of the rare earth metal hydroxide and iron oxides/hydroxides onto the surface due to local changes in the pH.With this ongoing process, it would be reasonable to suggest that the protective film could ultimately lead to a more uniform surface coverage [23].
For the La(3OHCin) 3 , the mechanism of film formation could be very similar in that the hydroxyl group is positioned in a geometry that is accessible to promote further interactions either with the La ion or the underlying substrate, as shown in Figure 7b.Whilst for the La(2OHCin) 3 inhibitor, the ortho-substitution causes the hydroxyl groups to be shielded by the aromatic ring (Fig 7c ), sterically hindering further interactions, and therefore eliminating the possibility of forming a multilayer surface film that might be possible for either the La(3OHCin) 3 and the La(4OHCin) 3 inhibitors.
In the case of Pr(4OHCin) 3 and Ce(4OHCin) 3 , the hydroxy group is in the para-position, as in the La(4OHCin) 3 , and therefore capable of further intermolecular interactions.This would explain the excellent inhibition shown by the Pr(4OHCin) 3 .In contrast, Ce(4OHCin) 3 does not show good inhibiting properties consistent with the differences in speciation observed in the electrospray mass spectroscopy.Interestingly, we observed that the solubility of the cerium compounds is generally significantly higher than the praseodymium and lanthanum analogues (for example Ce(4OHCin) 3 at 1000 ppm readily dissolves in the sodium chloride solution whereas Pr(4OHCin) 3 and La(4OHCin) 3 compounds are difficult to dissolve beyond 600 ppm at neutral pH).This is contrary to the solubility data in water for La(4OHCin) 3 and Ce(4OHCin) 3 which were found to be 961 ppm and 673 ppm, respectively, where the lanthanum compound was marginally more soluble than the cerium derivative [24].The difference in solubility is often related to the size of the rare earth metal cation, however given that La 3+ > Ce 3+ > Pr 3+ , this cannot explain our observations here.Furthermore, the X-ray powder diffraction patterns show that the compounds are all iso-structural having similar d-spacings and intensities and that the isomorphous RE(4OHCin) 3 have very similar ν(OH) absorptions [18].Additional information on the solid state compositions and structures of the rare earth n-hydroxycinnamates can be found in the ESI.The other differences amongst these rare earth metal cations is the existence of two redox states for Ce(III/IV) compared with Pr(III) and La(III).It has been previously shown that Ce(III) can readily oxidise from the Ce 3+ to Ce 4+ in the presence of oxygen and water with some ligands [25].The presence of this redox couple may result in the formation of different complex species that would not otherwise be seen in the lanthanum and/or praseodymium systems.This possibility is supported by the difference seen in the mass spectrum of the cerium compound, ESI Figure S1, compared to lanthanum 4-hydroxycinnamate (ESI Fig S5).However, further work is required to fully understand the complexation in this system, as we are unable to assign the spectrum obtained in this case.Thus the mechanism of inhibition for the Ce(4OHCin) 3 complex differs from that of La(4OHCin) 3 and Pr(4OHCin) 3 due to both the solubility and the oxidation states (Ce 3+ to Ce 4+ ).In our previous work, we have shown that the nature of the metal -ligand complexes in solution are critical for the corrosion inhibition efficiency; in particular the presence of species such as [LnL 4 ] -and [LnL 3 Cl] -.The absence of these in the case of Ce(4OHCin) 3 may contribute to the poorer inhibition.Thus the model proposed in Figure 6 may not apply to Ce(4OHCin) 3 .

Conclusion
The present work has focussed on understanding the influence of the rare earth metal cations and the hydroxycinnamate structure on the corrosion inhibition properties for the rare earth cinnamate compounds.Consistent with previous work, the La(4OHCin) 3 performed best with Pr(4OHCin) 3 and La(3OHCin) 3 showing near comparable inhibition efficiencies using potentiodynamic polarisation measurements.On the other hand, the Ce(4OHCin) 3 showed poorer inhibition properties correlating with the markedly different species observed in the mass spectrum.We hypothesise that this is related in some way to increased solubility and the two accessible oxidation states for cerium.Furthermore, the availability of the phenol group of the rare earth hydroxycinnamates for intermolecular interactions appears to be imperative for good corrosion inhibition as demonstrated by the poorer performance of La(2OHCin) 3 with the phenol group shielded by the phenyl-ring. .

Figure 1 .
Figure 1.Potentiodynamic polarisation curves for rare earth cinnamates in 0.01 M NaCl solution at 400 ppm and pH 2.5.Polarisation curves are compared to a control experiment consisting of 0.01M NaCl solution at pH 2.5.

Figure 2 .
Figure 2. Potentiodynamic polarisation curves for rare earth cinnamates in a 0.01M NaCl solution at 400 ppm and pH 5.5.Polarisation curves are compared to a control experiment consisting of 0.01M NaCl solution at pH 5.5.

Figure 3 .
Figure 3. Potentiodynamic polarisation curves for rare earth cinnamates in a 0.01M NaCl solution at 600 ppm and pH 2.5.Polarisation curves are compared to a control experiment consisting of 0.01M NaCl solution at pH 2.5.

Figure 4 .
Figure 4. Potentiodynamic polarisation curves for rare earth cinnamates in a 0.01M NaCl solution at 600 ppm and pH 5.5.Polarisation curves are compared to a control experiment consisting of 0.01M NaCl solution at pH 5.5.

Figure 7 .
Figure 7.A schematic hypothesised mechanism of action for the protection of a steel substrate using the inhibitor (a) RE(4OHCin) 3 where RE = La, Ce and Pr, (b) La(3OHCin) 3 and (c) La(2OHCin) 3 .