A study on the differences in morphology and corrosion resistance performance between two different bis(2-ethylhexyl) phosphate self-assembled thin films prepared on an iron substrate in water and ethanol solvents

Abstract

Bis(2-ethylhexyl) phosphate (BEP) self-assembled thin films were prepared on iron substrates in water and ethanol solutions respectively and the formation mechanisms were investigated by means of attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR), X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). For the iron samples treated with BEP aqueous and alcoholic solutions, the appearance of characteristic functional groups in the ATR-FTIR spectra indicated the existence of BEP molecules on the iron surface. XPS results further demonstrated that, BEP molecules bound to the iron surface in different ways in the case of water and ethanol as solvents, leading to the formation of two types of self-assembled monolayers (SAMs). In the ethanol solution, the BEP SAMs formed on the iron surface by both electrostatic interaction and chemical covalent binding, denoted as BEP_E SAMs; in the aqueous solution, the BEP SAMs formed on the iron surface only through P–O–Fe (substrate) bond between iron substrate and BEP molecules, denoted as BEP_W SAMs. The BEP_E SAMs showed an island-like surface morphology, while BEP_W SAMs had a nodule-like appearance. When used as ultrathin coatings for corrosion protection, the great differences in microstructure, morphology and compactness between the two BEP SAMs had a strong influence on their anti-corrosive properties, which in turn led to an interesting phenomenon that BEP_W SAMs showed much better corrosion protection ability than BEP_E SAMs under otherwise identical conditions.

---

1. Introduction

Self-assembly of organic molecules on a solid surface or at the solid–liquid interface may form highly ordered molecular assemblies, i.e. self-assembled monolayers (SAMs).¹ Generally, such molecules possess a head group that has a strong affinity to the substrate, such as thiol, phosphonate and silane groups. SAMs are created by the anchoring of head groups onto a specific substrate from either vapor or liquid phase, followed by a relatively slow organization of tail groups to enhance the stability and ordering of molecular arrangement.² Because SAMs are able to act as an ultrathin isolation film that can effectively separate the solid substrate from the damaging environments, SAMs, as a new film coating technique, have received increasing interest in metal corrosion and protection.^3,4 The strong chemisorption of the head groups on the solid substrates plays a crucial role in the formation of SAMs. In the past decades, most studies have focused on self-assembly of alkanethiols on the substrates of coinage metals (gold, silver and copper),^5,6 silanes on silicon oxide surface,^7,8 and carboxylic acids on metal oxide surface.^9,10 In particular, thiol-based SAMs on gold have become classical self-assembly systems since the solid evidence for formation of Au–S covalent bond has been obtained.^11,12 In the field of corrosion science, researchers mainly focus on the corrosion protection of copper substrate by alkanethiol, Schiff base and other SAMs while have paid less attention to the fabrication of molecular self-assembly systems on the surfaces of iron and steels. The key technical challenges of preparing highly ordered SAMs on the iron substrate are: (i) iron is apt to rust under general conditions, and (ii) there is lack of special bonding interaction between iron substrate and common organic molecules.^13,14 The first disadvantage makes it impossible to follow the existing self-assembling procedures to prepare thiol-based SAMs on the iron substrate, and the other disadvantage leads to instability of adsorbate molecules on the iron substrate.¹⁵ Consequently, thus far, no typical self-assembly system, like classical systems of SAMs of thiols/gold, silanes/silicon oxide, and carboxylic acids/metal oxides, has been created on the iron substrate.

Iron and steel materials are the most widely used engineering materials, so it is necessary to create a typical self-assembly system on the iron substrate for corrosion protection, without question.^16–20 In order to achieve the goal, the chief problem need to be solved is what type of inhibitor can serve as the film-forming substance. According to basic theory of organic corrosion inhibitors, a film-forming substance should contain heteroatoms that act as the active centers for the process of adsorption on the metal surfaces.^21–24 Moreover, the inhibition efficiency follows the sequence O < N < S < P. Because they contain both O and P atoms in the molecules, alkyl phosphate esters should be ideal candidates among various film-forming substances.^25–28 The second problem is why the self-assembling process must be carried out in water phase. From the perspective of corrosion protection, aqueous solutions containing corrosive substances are the most commonly used corrosive media. Meanwhile from the perspective of industrial applications, the use of large quantities of organic solvents is very dangerous for operators and also harmful to the environment. It is of great theoretical value and practical significance to develop an advanced self-assembly method based on spontaneous adsorption of inhibitor (alkyl phosphate esters) molecules on the iron surface in water phase. The third problem is why alkyl phosphate esters are able to adsorb preferentially onto the iron surface in an aqueous solution. Most alkyl phosphate esters can be treated as both organic inhibitors and amphiphilic substances,^29,30 so the spreading, adsorption and self-assembly at solid–water interfaces help them to spontaneously adsorb onto the iron surface. Especially, the specific adsorption of phosphate groups on the oxidized iron surface may enable them to adsorb preferentially on the iron surface by excluding water molecules from the iron surface.

In view of the above-mentioned reasons, recently we have put forward a preliminary scheme of preparing short-chain alkyl phosphate SAMs on the iron substrate in water phase and proved the technical feasibility.¹⁵ However, several important scientific questions have been left unanswered. First of all, it is essential to determine whether alkyl phosphate SAMs form on the iron surface through the direct reaction of alkyl phosphate with iron substrate. For this, contrast experiments need to be carried out in organic solvents. Secondly, it is unclear that alkyl phosphate molecules bind to the iron surface by electrostatic interaction or chemical covalent binding. Thirdly, we want to further know how the bonding mode between alkyl phosphate molecules and iron substrate affects the stability of SAMs. With these questions, selecting an alkyl phosphate with two alkyl chains and only a –OH group, bis(2-ethylhexyl) phosphate (BEP), as the research subject, we investigated in what manner BEP molecules self-assembled into ordered molecular aggregates in both water and ethanol solutions respectively. It is worth mentioning that BEP molecules could be self-assembled into two different types of SAMs in water and ethanol solvents. Due to great differences in microstructure, morphology and compactness, the as-obtained two SAMs showed quite different protection effects on the corrosion of iron substrates. The present study provides helpful insights into the self-assembling behavior of the same organic molecules in different solvents.

2. Experimental section

2.1 Materials

A 2 mm-diameter iron rod (99.995%) was received from Alfa Aesar. Bis(2-ethylhexyl) phosphate (BEP) (95 wt%) was purchased from Aladdin China Ltd. Before use, the BEP product was purified according to the procedure described previously.¹⁵ Absolute ethanol (99.5 wt%) and H₂SO₄ (98 wt%) were obtained from Sinopharm Chemical Reagent Co., Ltd. All chemicals and solvents were analytical reagent grade. 0.5 M H₂SO₄ solution was prepared by diluting H₂SO₄ (98 wt%) with ultrapure water (electrical resistivity: ∼18 MΩ cm) and used as corrosive solution. Other aqueous solutions were prepared using ultrapure water.

2.2 Preparation of SAMs

Appropriate amounts of BEP was dispersed in absolute ethanol to form 1.0 wt% BEP ethanol solution and in ultrapure water to form supersaturated aqueous solution of BEP (∼0.1 wt%) respectively, under action of strong ultrasonic vibration.

The above-mentioned iron rod was employed to prepare the working electrodes. The iron rod specimen was embedded in epoxy resin mould, leaving its cross-section only exposed to the BEP aqueous (or alcoholic) solutions and corrosive electrolyte solutions. The iron electrode was ground with SiC emery papers of decreasing particle sizes (400, 2000, 3000 and 4000) to a mirror-like surface, and then rinsed with 95% ethanol and ultrapure water, and finally dried under nitrogen stream at room temperature. Iron electrode was immersed in the BEP alcoholic or supersaturated aqueous solutions for a designated period of time, allowing BEP molecules to spontaneously self-assemble onto the iron substrate. When an iron electrode was taken out from the BEP aqueous or alcoholic solution, it was rinsed repeatedly with absolute ethanol or ultrapure water to eliminate the BEP molecules physically adsorbed on the iron surface, followed by drying with a flow of nitrogen.

2.3 Surface analysis and characterization

Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), atomic force microscopy (AFM) and metallurgical microscope were employed to determine characteristic functional groups, compositions, microstructure and morphology of the BEP SAMs formed on the iron surface, respectively. Here the formation time of BEP SAMs was 1 h in any case. The ATR-FTIR spectra of bare iron samples and BEP SAMs modified iron samples were obtained using a Bruker TENSOR 27 FTIR in wave number range between 4500 and 450 cm⁻¹. Meanwhile, FTIR spectrum of pure BEP substance was also measured and used as the reference. The XPS measurements of the untreated and treated iron samples were performed using a Perkin Elmer PHI 5300 system under an Mg Kα X-ray source by energy analyzer in 35.75 eV. And the obtained XPS spectra were analyzed using XPS software (XPS peak 4.1). Here all binding energies were referenced versus the C–C/C–H components of the C 1s peak at a binding energy of 284.6 eV. AFM images were collected using a Nanoscope IIIa AFM with a fixed flexible micro cantilever and automatic modulation control of the distance between probe and the sample surface according to atomic force under a tapping mode. The surface morphologies of corroded samples were observed by using a Motic metallurgical microscope. Before observation, several iron plate samples, including bare, and BEP_E and BEP_E SAMs modified iron samples, were potentiostatically treated at −0.45 V (vs. SCE) for 300 s.

2.4 Electrochemical measurements

A conventional three-electrode test cell was used for electrochemical measurements at room temperature (22 °C). A bare iron electrode or a BEP SAMs modified iron electrode was employed as working electrode, and a large platinum plate as counter electrode and a saturated calomel electrode (SCE) as reference electrode. Polarization curves were recorded with a CHI 604 electrochemical workstation by scanning the polarization potential from negative to positive at a slow sweep speed of 0.2 mV s⁻¹ to make the electrochemical system close to its steady state during the linear potential scan. Electrochemical impedance spectroscopic (EIS) measurements were performed with an ACM electrochemical workstation in frequency range from 100 kHz to 0.05 Hz with eight points per decade under excitation of a sinusoidal wave of ±5 mV amplitude.

3. Results and discussion

3.1 ATR-FTIR characterization

ATR-FTIR spectroscopy was used as an important tool for determination of whether the BEP thin films were formed on the substrates by spontaneous adsorption of BEP molecules from ethanol or water solution. Meanwhile, FTIR spectrum of pure BEP liquid was also measured and used as a reference spectrum to better interpret how the solvent affected the self-assembling behavior of BEP molecules (Fig. 1).

Fig. 1 FTIR spectrum of pure BEP substance and the ATR-FTIR spectra of the iron plate samples untreated and treated with BEP-containing ethanol and water solutions respectively.

The analysis for the FTIR spectrum of pure BEP shows that, the absorption peaks at 2961.8 cm⁻¹, 1464.52 cm⁻¹ and 1380.84 cm⁻¹ originate respectively from C–H stretching, scissor bending and out-of-plane bending vibrations of 2-ethylhexyl group.³¹ The absorption peaks at 1021.16 cm⁻¹ and 886.13 cm⁻¹ are the characteristic peaks related to the P–O–C stretching vibrations while the peak at 1226.24 cm⁻¹ is due to the P=O stretching mode.^19,31 Besides, the peaks at 1678.52 cm⁻¹, 2170.02 cm⁻¹ and 2330.35 cm⁻¹ are associated with the P–OH stretching mode,³¹ and the absorption peak at 3450.00 cm⁻¹ corresponds to the O–H stretching mode.³¹

ATR-FTIR spectra for the iron samples treated with BEP alcoholic and aqueous solutions both display the characteristic peaks of 2-ethylhexyl group and ones corresponding to the P–O–C stretching mode. By contrast, ATR-FTIR spectrum of the untreated iron sample does not show any characteristic absorption peak. This indicates the existence of BEP molecules on the iron substrates, no matter whether the self-assembling process was carried out in ethanol or water solvent. On the other hand, by comparing the IR spectrum of pure BEP substance with those of BEP SAMs modified iron samples, we are sure that there exist big differences in infrared absorption properties between free BEP molecules and adsorbed BEP molecules. At first, the characteristic peak at 1226.24 cm⁻¹, which corresponds to the P=O stretching mode of free BEP molecules, disappeared after the BEP molecules self-assembled onto the iron substrates from ethanol or water solution. This change implies that, for adsorbed BEP molecules, the P=O double bonds were opened and new chemical bonds were formed, which is confirmed by the following XPS results. Secondly, infrared absorption behavior of the BEP_W SAMs is quite different from that of the BEP_E SAMs. For the former SAMs, disappearance of the characteristic absorption peak at 3450.00 cm⁻¹ related to the O–H stretching mode of free BEP molecules indicates that there are no free O–H groups in the BEP_W SAMs. For the latter SAMs, the above-mentioned absorption peak did not disappear, but the peak position moved to 3240.68 cm⁻¹. This phenomenon indicates that there are intermolecular hydrogen bonds between the adsorbed phosphate molecule and iron substrate in the BEP_E SAMs. Moreover, the appearance of a new, broad absorption peak at 2530.33 cm⁻¹ associated with O=P–OH⋯O offers beneficial evidence for the formation of hydrogen bonds.³¹

These obviously differences reveals that the BEP molecules self-assembled onto the iron surfaces in different ways depending on the solvent to be used. To be sure, the O–H, P–OH and P=O groups play important roles in the self-assembly process of BEP molecules on the iron substrate. As for how BEP molecules bind to the iron surface in ethanol and water solvents, a detailed study was performed by means of XPS technique.

3.2 XPS analysis

The compositions of two kinds of BEP thin films, BEP_E and BEP_W, were analyzed by core level XPS. Fig. 2 shows the high resolution XPS spectra of the major elements (P, O and Fe) on the surfaces of iron samples modified with BEP_E and BEP_W SAMs, respectively, in which raw data were marked with dotted black lines, and the complete spectrum for each element acquired by curve fitting was drawn with a red solid line.

Fig. 2 High resolution P 2p, O 1s, Fe 2p3 XPS spectra of major elements on the iron surface modified with BEP_E (a) and BEP_W (b) films.

The P 2p core level spectra for BEP_E and BEP_W SAMs are both composed of two peaks, as seen in Fig. 2. For BEP_E SAMs, the peak located at 132.2 eV is corresponded to P–O–C bond in phosphate groups of BEP_E thin films, and the peak with binding energy at 134.8 eV is associated with P–O–Fe bond and contained newly formed phosphate species rather than the previous phosphate.^32–34 Besides, the peak area for P–O–Fe is approximately half of that for P–O–C. In other word, the ratio of chemical bonds P–O–Fe and P–O–C coincides with the condition of P–O–Fe : P–O–C = 1 : 2. In contrast, for BEP_W SAMs, the P–O–Fe and P–O–C peaks are located at 134.6 eV and 131.9 eV, respectively. The peak positions slightly shifted towards lower binding energy compared to those of BEP_E SAMs. In particular, the peak areas of the two peaks are almost equal, which means that the proportion of P–O–Fe and P–O–C meets P–O–Fe : P–O–C = 1 : 1.

The O 1s core level spectra consist of three peaks whether BEP SAMs was formed in ethanol or water. For BEP_E SAMs, the lowest peak is located at binding energy of 529.6 eV which is assigned to oxygen atoms bonded to ferric oxides such as Fe₂O₃ and Fe₃O₄,^19,33–37 and the highest peak is located at binding energy of 533.0 eV which is corresponding to the existence of oxygen in the chemisorbed H₂O.^19,33–37 The middle peak centered at 531.6 eV is the minimal component and mainly originates from the OH⁻ of hydroxide such as FeO(OH) and the O–P bond in phosphate group.^19,33–37 However, for BEP_W SAMs, the middle peak located at binding energy of 531.5 eV consists mainly of the O–P bond in the phosphate group such as PO₄³⁻,^19,33–37 which is corresponding to the satellite peak of iron phosphate located at binding energy of 711.4 eV. The locations of the remaining two peaks are almost the same with peak positions for the BEP_E SAMs, but the differences of the peak areas between the two peaks are more significant.

The Fe 2p core level spectra were all fitted into three peaks. For BEP_E SAMs, as shown in Fig. 2, the first peak with the smallest area at 706.9 eV is connected with the uncovered sites on the iron surface,^19,35,37,38 the second peak at binding energy of 711.2 eV is ascribed to the ferric compounds-magnetite (Fe₂O₃), iron hydroxides FeOOH and some iron phosphates;^34–38 the third peak at 714.1 eV can be viewed as the satellite of the ferric iron.^35,37,38 For BEP_W SAMs, the satellite peak of iron phosphate located at binding energy of 711.4 eV appeared.³⁵ Similarly, the rest peaks at 710.2 and 714.0 eV respectively represent ferriferous oxide and Fe(III).^35,37,38

According to the above-mentioned analysis and comparison, it is clear that both iron hydroxides and iron phosphates really coexist when self-assembly of BEP was implemented in the ethanol solution, while the main products were iron phosphates for the SAMs obtained from the BEP-containing aqueous solution. In another word, the BEP_E SAMs was formed on the iron substrate in BEP-containing ethanol solution through electrostatic interaction (specific adsorption in the form of undissociated molecules) and formation of chemical bonding, while the self-assembly of BEP in water solution depended strongly on the generation of chemical bonding such as P–O–Fe and PO₄³⁻. Based on the above XPS analysis, the formation of BEP SAMs in ethanol and water can be illustrated with Fig. 3.

Fig. 3 Schematic diagram for self-assembly of BEP molecules from ethanol (a) and water (b) solvents.

In ethanol solution, an individual BEP molecule firstly adsorbed on iron surface through electrostatic interaction and hydrogen bonds between oxygen atoms from P=O and P–OH groups in the molecule and hydroxyl groups on the iron substrate. Subsequently, the BEP molecule anchored at the iron surface via a dehydration reaction (P–O–Fe : P–O–C = 1 : 2). The formation mechanism of BEP_E SAMs is supported the infrared absorption properties of BEP_E SAMs and the related XPS results. But in aqueous solution, due to the release of H⁺ ion from the P–OH group and the electron delocalization along O–P–O segment, negatively charged O atoms were able to directly bind to iron surface or form hydrogen bonds with hydroxyl groups on the iron substrate. Moreover, a BEP molecule can firmly bind to the iron surface through P–O–Fe (substrate) bonds after the dehydration (P–O–Fe : P–O–C = 1 : 1). Because of lack of free –OH groups, the corresponding ATR-FTIR spectrum did not show the characteristic peak corresponding to the O–H stretching mode.

3.3 AFM observations

AFM was used to capture topographical maps of the iron surfaces before and after modification with different BEP SAMs in a tapping mode.³⁹ Generally, when the tip of AFM works in a tapping mode, AFM piezoelectric microcantilever will vibrate with large amplitude. Thus, when the needle tip gently taps on the sample and intermittently touches the sample surface, topographical data from a sample surface can be obtained by means of detecting and controlling the signal which derives from the force between tip and sample surface.

Fig. 4 shows the AFM images for the iron surfaces unmodified and modified with two kinds of BEP SAMs respectively. It is observed that great changes in surface morphology took place after the iron surface was covered by the two BEP SAMs. Firstly, for the unmodified iron sample, its surface seemed relatively flat, with a root-mean-square (RMS) roughness value of 5.10 nm, even though the surface was covered by a homogenous iron oxide film caused by the spontaneous oxidation of iron surface in common atmosphere environment.⁴⁰ However, for the BEP SAM modified iron surfaces, the surface roughness was observed to increase at different levels depending on the used solvent. The roughness value was 15.0 nm and 20.4 nm for the iron surfaces covered with the BEP_E and BEP_W SAMs respectively. Secondly, although BEP molecules can be self-assembled onto the iron substrate in ethanol or water, the as-obtained BEP SAMs still show big differences in morphology and microstructure, as AFM images in Fig. 4b and c indicated. The self-assembly of BEP molecules in ethanol solution produced island-like molecular aggregates, which covered most of the iron surface and only left small areas unoccupied, as demonstrated by high resolution XPS spectra of Fe 2p3 at 706.9 eV. However, self-assembling of BEP molecules on the iron surface in aqueous solution led to the formation of nodules aggregates (BEP_W SAMs), which seemed coarser than the former BEP_E SAMs.⁴¹ According to Schwartz and co-workers' viewpoint, the monolayer thin films usually form by nucleation, growth, and coalescence of submonolayer islands.^42,43 Thus, in ethanol solution, individual BEP molecules firstly adsorbed onto the iron substrate via electrostatic interaction, and subsequently, the adsorbed BEP molecules were gradually organized into large ordered domains, companied by the formation of P–O–Fe bonds, depending on the van der Waals force between 2-ethylhexyl groups. The self-assembling process should be similar to the formation of octadecyl phosphonic acid (OPA) monolayers on mica substrates reported by Schwartz.^42,43 But in water solution, BEP tended to behave like an organic acid and directly react with the iron substrate. In this way, the formation of both P–O–Fe bonds and iron phosphate on the iron surface played a crucial role in the self-assembling of BEP molecules. For the above reasons, it is not difficult to understand why there exist such large differences in morphology and microstructure between BEP_E and BEP_W SAMs. And conversely, such big differences in morphology between BEP_E and BEP_W SAMs also support the different formation mechanisms of BEP SAMs in ethanol and water shown in Fig. 3.

Fig. 4 AFM images for the iron surfaces unmodified (a) and modified with BEP_E (b) and BEP_W (c) SAMs.

3.4 Metallographic observations

Fig. 5 shows the surface topography of untreated iron sample and the BEP SAMs coated iron ones after being polarized 300 s at an anodic potential of −0.45 V in 0.5 M H₂SO₄ solution. For the unmodified iron, the iron sample suffered from obvious corrosion as indicated by the rough surface in Fig. 5a. In comparison, when the iron surface was covered by ultrathin BEP SAMs, the modified iron surface seemed still smooth, flat and shiny after being treated at the applied anodic potential. And moreover, the protective effect of BEP_W SAMs was better than that of BEP_E SAMs. This confirms that under acidic condition, the BEP SAMs are able to provide an effective barrier to block corrosive media from contacting the iron substrate. Due to the strong chemical bonding between BEP_W SAMs and iron substrate, the BEP_W SAMs exhibit the stronger anti-corrosion performance than the BEP_E SAMs.

Fig. 5 Metallographic graphs of the iron plate samples unmodified (a) and modified with BEP_E (b) and BEP_W (c) SAMs.

3.5 Polarization curves

The method of steady-state polarization curve was employed to evaluate corrosion protection performance of the BEP SAMs under different conditions. Fig. 6 shows the polarization curves for the unmodified iron and iron electrodes modified by the BEP_E and BEP_W SAMs in 0.5 M H₂SO₄ solutions. The film-forming time used to prepare the two BEP SAMs changed from 10 to 60 min. It is seen from Fig. 6 that, BEP_E SAMs had comparatively small impact on the corrosion potentials, especially in the case of shorter film-forming time, but suppressed both cathodic and anodic currents to some extent. Compared with the polarization curves of unmodified iron electrode, the presence of the BEP_E SAMs lowered the cathodic current more effectively than it did to anodic current when the self-assembly time (SA time) was within 30 min. However, if the SA time increased to 60 min, the corrosion potential shifted towards more positive values, and the anodic dissolution of iron was also inhibited significantly.

Fig. 6 Polarization curves for the unmodified, BEP_E SAMs modified (a) and BEP_W SAMs modified iron electrodes (b) in 0.5 M H₂SO₄ solutions.

It is of interest to note that the BEP SAMs prepared in water, i.e. BEP_W SAMs, presented the better corrosion protection performance than the above BEP_E SAMs. One of the main differences in anti-corrosion property between BEP_E SAMs and BEP_W SAMs is that the BEP_W SAMs suppressed both anodic and cathodic currents to a greater degree whether the SA time was short or long. Of course, the longer SA time was favorable for the enhancement of anti-corrosion performance of the BEP_W SAMs.

The electrochemical parameters of each corrosion system, including corrosion potential (E_corr), corrosion current density (i_corr), cathodic Tafel slope (b_c), and anodic Tafel slope (b_a), may be determined from the corresponding curve by using the Tafel extrapolation method. Thus, the electrochemical parameters for the BEP_E and BEP_W SAMs modified electrodes in each case, along with those of the unmodified iron electrode, were collected in Table 1. Protection efficiencies (PE, expressed as η) of the two BEP SAMs to the iron substrates under different conditions were calculated according to the following eqn (1):

where i_corr and i′_corr represent the corrosion current densities of unmodified iron and BEP SAMs modified iron electrode, respectively. To be convenient for comparison, the PE values of the BEP_E and BEP_W SAMs in each case were also listed in Table 1. High PE value indicates good anti-corrosion performance. It is clear that the BEP_W SAMs had the stronger corrosion protective performance than the BEP_E SAMs under otherwise identical conditions.

Table 1 Corrosion electrochemical parameters for the unmodified iron electrode, the BEP_E SAMs and BEP_W SAMs modified iron electrodes in 0.5 M H₂SO₄ solution

	SA time (min)	Ecorr (mV)	bc (mV dec^−1)	ba (mV dec^−1)	icorr (A cm^−2)	η (%)
Unmodified	0	−544	129.8	55.11	1.991 × 10^−4	—
BEPE SAMs	10	−535	139.6	53.17	1.278 × 10^−4	35.78
	20	−529	136.2	52.37	1.248 × 10^−4	37.31
	30	−533	138.7	53.83	1.199 × 10^−4	39.78
	60	−498	165.3	61.75	9.771 × 10^−5	50.93
BEPW SAMs	10	−516	133.8	56.38	1.314 × 10^−4	33.99
	20	−540	122.9	52.22	5.583 × 10^−5	71.96
	30	−498	122.6	56.32	4.125 × 10^−5	79.28
	60	−496	117.5	42.35	2.369 × 10^−5	88.10

---

The reason why there exist such large difference in anti-corrosion performance between the two BEP SAMs should be attributed to the influences of solvents on ordering, compactness, and defect numbers of the BEP SAMs, especially on the bonding strength of adsorbed BEP molecules to the iron substrate. According to XPS analysis and AFM observation, the BEP_E SAMs formed through both electrostatic interaction and chemical covalent binding, and the resulting SAMs had small uncovered areas. The as-fabricated BEP_E SAMs were able to behave as a corrosion barrier layer when the BEP_E SAMs modified iron electrode was immersed in H₂SO₄ solution. As can be seen in Fig. 6, the BEP_E SAMs effectively inhibited the hydrogen evolution reaction (HER) on the iron surface by the blocking effect. However, due to the weak bonding strength, the adsorbed BEP molecules could be readily desorbed from the iron substrate once an anodic potential was applied to the iron electrode, even though the potential was relatively low. So it is not difficult to understand why the BEP_E SAMs showed the low inhibition effect on the anodic dissolution of iron. In fact, even if the SA time was relatively long (e.g. 60 min), the PE of the BEP_E SAMs was still very low, only 50.93%. In strong contrast to the BEP_E SAMs, the PE of BEP_W SAMs could reach 88.10%, almost twice as much as that of BEP_E SAMs, under otherwise identical conditions. The main reason is that the BEP_W SAMs formed on the iron surface through strong P–O–Fe (substrate) bond between BEP molecules and surface iron atoms. Different from the BEP_E SAMs, the BEP_W SAMs can attach more firmly to the iron substrate and also possess good compactness (see Fig. 3b and c). As a result, the BEP_W SAMs show the strong suppressing effect on both cathodic and anodic reactions compared to BEP_E SAMs.

3.6 EIS investigation

EIS measurements were carried out to have a better understanding of the corrosion behavior of the iron electrodes modified with the two BEP SAMs under acidic condition. Fig. 7 shows Nyquist impedance spectra for a unmodified iron electrode and two BEP SAMs modified iron electrodes at the open-circuit potential (E_ocp) in 0.5 M H₂SO₄ solution. This impedance spectrum for unmodified iron consists of a high frequency capacitive and a low frequency inductive loop. The corrosion of iron at E_ocp involves the hydrogen evolution reaction (HER) at cathodic sites and the iron dissolution at anodic sites. The coupling between the two conjugated reactions leads to the corrosion of iron under open-circuit condition. The capacitive loop is related to the relaxation time constant of double-layer capacitance (C_dl) with the overall charger-transfer resistance (R_ct) during the self-corrosion process of iron in H₂SO₄. The inductive loop is associated with the relaxation time constant of an adsorbed intermediate species involving in the corrosion process of iron. The diameter of the capacitive loop reflects the corrosion rate. Generally, the larger the diameter is corresponding to the slower the rate of corrosion reaction.

Fig. 7 A series of Nyquist impedance spectra for the BEP_E SAMs (a) and BEP_W SAMs (b) modified iron electrodes at the respective open-circuit potentials in 0.5 M H₂SO₄ solutions. Self-assembling (SA) time of two BEP SAMs was indicated in two figures respectively. Nyquist impedance spectrum of the unmodified iron electrode was placed in two figures for comparison.

For the BEP_E SAMs modified iron electrodes, the shape of the corresponding impedance spectra remained unchanged although the capacitive loop increased in size at different levels depending on the SA time. Due to the presence of BEP_E SAMs on the iron substrate, corrosion rate of the iron substrate was reduced significantly, which caused the increase of the diameter of the capacitive loop, implying the increase of R_ct. A BEP molecule possesses a head group and two tails. The BEP_E SAMs are created by the adsorption of head groups onto the iron substrate from ethanol phase, followed by a slow reorganization of tail groups and formation of P–O–Fe bond on iron substrate. Obviously, the longer the SA time means the better the ordering and compactness of BEP_E SAMs. As a corrosion barrier layer, a long SA time may help the BEP_E SAMs to enhance the anti-corrosion performance. It is evident that the diameter of the capacitive loop increased with the SA time.

The existence of inductive loops indicates that corrosion reaction still occurred at the BEP_E SAMs coated iron surface. This suggests that BEP_E SAMs did not cover the entire iron surface, or there were a certain amount of defects in BEP_E SAMs. The inference was supported well by the AFM image shown in Fig. 4b.

The same is true of EIS results for the BEP_W SAMs modified iron electrodes in H₂SO₄ solution. It should be noted that, the Nyquist spectra for the BEP_W SAMs modified electrodes show the larger capacitive loop compared to those of the BEP_E SAMs modified ones. Because the BEP_W SAMs are superior to BEP_E SAMs in ordering and compactness, it is not surprising that BEP_W SAMs exhibit much stronger corrosion protection performance. Even so, the corrosion of iron substrate was not completely inhibited.

As mentioned above, the inductive behavior arises from the relaxation time constant of adsorbed species generated in the corrosion reaction of iron. Because this time constant is much larger than that one caused by the charge-transfer process, the inductive loop usually appears in the lowest frequency region. Here what we concerned more is how BEP SAMs affect the charge-transfer process at iron substrate/solution interface, so we may neglect the inductive behavior and focus on the impedance behavior of iron electrode in mid-to-high-frequency region. In this sense, the impedance behavior of unmodified iron can be interpreted by means of equivalent circuit shown in Fig. 8a. In this circuit, R_s represents the solution resistance, R_ct charge-transfer resistance, and C_dl double-layer capacitance. Here a constant phase element (CPE) is substituted for the capacitors in order to fit more exactly the depressed semicircle. Admittance and impedance of a CPE are, respectively, defined as

Y_CPE = Y₀(jω)ⁿ (2)

and

Z_CPE = (1/Y₀)(jω)⁻ⁿ (3)

where Y₀ the modulus, ω the angular frequency and n the phase.

Fig. 8 Equivalent circuits used to model impedance behavior of iron electrodes unmodified and modified by two BEP SAMs.

For the BEP_E SAMs-coated iron electrodes, their impedance behavior may be modeled by using equivalent circuit B shown in Fig. 8b. Because there are a certain amount of defects in BEP_E SAMs, the corrosion reaction is effectively suppressed in the regions covered by the SAMs but still takes place in uncovered regions. In circuit B, R_s, CPE_dl and R_ct have the same physical meaning as what they represent in the circuit shown Fig. 8a, while CPE_(BEP-E) stands for the capacitance of BEP_E SAMs. In fact, circuit B is essentially equivalent to circuit B′, in which CPE′_dl element contains the contributions from double-layer capacitance and the capacitance of BEP_E SAMs.

Although BEP_W SAMs are more compact than BEP_E SAMs, coverage of BEP_E SAMs on the iron substrate strongly depends on the formation time of SAMs (i.e. SA time). When SA time was relatively short (<30 min), some regions remained uncovered on the iron substrate. In this case, the impedance behavior of the BEP_W SAMs-coated iron electrodes can still be modeled using equivalent circuit B or B′. However, with the increase of SA time (≥30 min), the entire surface of iron substrate was almost covered with BEP_W SAMs, and the corrosion could only occur at small amount of defective sites in the BEP_W SAMs. In theory, equivalent circuit C or C′ is more suited to modeling the impedance behavior of the BEP_W SAMs-coated iron electrodes at this moment. In circuit C (or C′), CPE_(BEP-W) represents the capacitance of BEP_W SAMs, R_(BEP-W) represents the resistance of the SAMs, which reflects the corrosion resistance ability of the SAMs, and other elements have the previous physical meaning.

The impedance spectra in Fig. 7 were analyzed by the above-mentioned three circuits respectively, and the values of these elements were obtained and collected in Table 2.

Table 2 Electrochemical impedance spectra parameters for the unmodified iron electrode, BEP_E SAMs and BEP_W SAMs modified iron electrodes in 0.5 M H₂SO₄ solution obtained by analyzing the impedance spectra shown in Fig. 7

	SA time (min)	Rs (Ω cm^2)	RBEP-W (Ω cm^2)	CPEBEP-W		Rct (Ω cm^2)	CPEdl		η (%)
				Y0 (Ω^−1 cm^−2 s^n)	n		Y0 (Ω^−1 cm^−2 s^n)	n
Unmodified	0	0.25	—	—	—	148.2	2.127 × 10^−4	0.8564
BEPE SAMs	10	0.21	—	—	—	249.2	2.036 × 10^−4	0.7945	40.53
	20	0.24	—	—	—	275.9	1.972 × 10^−4	0.8128	46.28
	30	0.18	—	—	—	290.9	1.886 × 10^−4	0.8133	49.06
	60	0.16	—	—	—	374.3	1.744 × 10^−4	0.7961	60.40
BEPW SAMs	10	0.30	—	—	—	262.9	1.046 × 10^−4	0.8870	43.63
	20	0.28	—	—	—	538.2	1.614 × 10^−5	0.8712	72.46
	30	0.35	257.1	6.826 × 10^−6	0.8525	893.6	1.361 × 10^−5	0.8561	87.12
	60	0.23	250.6	6.811 × 10^−6	0.9752	959.3	9.954 × 10^−6	0.9375	87.76

---

On the basis of R_ct values in different situations, the protection efficiency (PE) of two BEP SAMs can be determined according to the following formula:

where R_ct and R′_ct are the values of charge-transfer resistance for the unmodified iron electrodes and BEP SAMs modified iron electrodes respectively. The PE values of BEP_E and BEP_W SAMs in all case were also listed in Table 2.

By comparing PE values of two BEP SAMs in Table 2 under identical conditions, it is concluded that BEP_W SAMs have the stronger corrosion protection performance than BEP_E SAMs, which is in good agreement with the conclusion based on the results of polarization curves.

4. Conclusion

BEP molecules can be self-assembled onto the iron substrates in water or ethanol. However, formation mechanisms of BEP SAMs in water and in ethanol are different from each other. In the former case, BEP SAMs formed spontaneously on the iron surface only through the P–O–Fe bond between iron substrate and BEP molecules, but in the latter case, BEP molecules adsorbed onto the iron surface via both electrostatic interaction and chemical covalent binding, forming another kind of BEP SAMs. Seeing that BEP_W SAMs are more compact and bind more firmly to the iron substrate than BEP_E SAMs, the former SAMs show stronger corrosion protection performance than the latter SAMs, as demonstrated by electrochemical results.

Acknowledgements

This project is supported by the National Natural Science Foundation of China (21373129).

References

1. A. Ulman, Chem. Rev., 1996, 96, 1533–1554.
2. W. Guo, S. Chen, Y. Feng and C. Yang, J. Phys. Chem. C, 2007, 111, 3109–3115.
3. Y. Ein-Eli, M. Auinat and D. Starosvetsky, J. Power Sources, 2003, 114, 330–337.
4. T. Shimura and K. Aramaki, Corros. Sci., 2008, 50, 2407–2414.
5. J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo and G. M. Whitesides, Chem. Rev., 2005, 105, 1103–1170.
6. H. Ma, C. Yang, S. Chen, Y. Jiao, S. Huang, D. Li and J. Luo, Electrochim. Acta, 2003, 48, 4277–4289.
7. S. Onclin, B. J. Ravoo and D. N. Reinhoudt, Angew. Chem., Int. Ed., 2005, 44, 6282–6304.
8. N. Tillman, A. Ulman and T. L. Penner, Langmuir, 1989, 5, 101–111.
9. P. E. Laibinis, J. J. Hickman, M. S. Wrighton and G. M. Whitesides, Science, 1989, 245, 845.
10. R. J. Jouet, A. D. Warren, D. M. Rosenberg, V. J. Bellitto, K. Park and M. R. Zachariah, Chem. Mater., 2005, 17, 2987–2996.
11. O. Neuman and R. Naaman, J. Phys. Chem. B, 2006, 110, 5163–5165.
12. D. Fischer, A. Curioni and W. Andreoni, Langmuir, 2003, 19, 3567–3571.
13. S. Ramachandran, B.-L. Tsai, M. Blanco, H. Chen, Y. Tang and W. A. Goddard, Langmuir, 1996, 12, 6419–6428.
14. K. Aramaki, Corrosion, 1999, 55, 1020–1030.
15. C. Zhao, X. Gao, H. Lu, R. Yan, C. Wang and H. Ma, RSC Adv., 2015, 5, 54420–54432.
16. H. Ju, Z.-P. Kai and Y. Li, Corros. Sci., 2008, 50, 865–871.
17. P. Bommersbach, C. Alemany-Dumont, J.-P. Millet and B. Normand, Electrochim. Acta, 2005, 51, 1076–1084.
18. I. Ahamad and M. Quraishi, Corros. Sci., 2009, 51, 2006–2013.
19. X. Gao, C. Zhao, H. Lu, F. Gao and H. Ma, Electrochim. Acta, 2014, 150, 188–196.
20. M. Quraishi and J. Rawat, Mater. Chem. Phys., 2003, 77, 43–47.
21. S. Ramesh, S. Rajeswari and S. Maruthamuthu, Appl. Surf. Sci., 2004, 229, 214–225.
22. H. Ashassi-Sorkhabi, M. Majidi and K. Seyyedi, Appl. Surf. Sci., 2004, 225, 176–185.
23. K. Khaled and N. Hackerman, Electrochim. Acta, 2003, 48, 2715–2723.
24. A. Samui and S. M. Phadnis, Prog. Org. Coat., 2005, 54, 263–267.
25. S. Refaey, F. Taha and A. A. El-Malak, Appl. Surf. Sci., 2004, 236, 175–185.
26. M. Textor, L. Ruiz, R. Hofer, A. Rossi, K. Feldman, G. Hähner and N. D. Spencer, Langmuir, 2000, 16, 3257–3271.
27. M. Dubey, T. Weidner, L. J. Gamble and D. G. Castner, Langmuir, 2010, 26, 14747–14754.
28. P. B. Paramonov, S. A. Paniagua, P. J. Hotchkiss, S. C. Jones, N. R. Armstrong, S. R. Marder and J.-L. Brédas, Chem. Mater., 2008, 20, 5131–5133.
29. B. T. Outlaw, B. A. O. Alink, J. A. Kelley and C. S. Claywell, US Pat., 4, 511,480[P], 1985.
30. P. A. Sørensen, S. Kiil, K. Dam-Johansen and C. Weinell, J. Coat. Technol. Res., 2009, 6, 135–176.
31. D. L. Perry, Handbook of inorganic compounds, CRC Press, 2011.
32. S. Sun and G. J. Leggett, Nano Lett., 2007, 7, 3753–3758.
33. S. Feliu and V. Barranco, JCT Res., 2004, 1, 93–102.
34. X. Gao, S. Liu, H. Lu, F. Gao and H. Ma, Ind. Eng. Chem. Res., 2015, 54, 1941–1952.
35. J. A. Rotole and P. M. Sherwood, Chem. Mater., 2001, 13, 3933–3942.
36. K. Babic-Samardzija, C. Lupu, N. Hackerman, A. R. Barron and A. Luttge, Langmuir, 2005, 21, 12187–12196.
37. H. Zarrok, A. Zarrouk, B. Hammouti, R. Salghi, C. Jama and F. Bentiss, Corros. Sci., 2012, 64, 243–252.
38. A. Galtayries, R. Warocquier-Clerout, M. D. Nagel and P. Marcus, Surf. Interface Anal., 2006, 38, 186–190.
39. R. Quiñones and E. S. Gawalt, Langmuir, 2007, 23, 10123–10130.
40. B. A. Rao, M. Y. Iqbal and B. Sreedhar, Corros. Sci., 2009, 51, 1441–1452.
41. R. Bhure, A. Mahapatro, C. Bonner and T. M. Abdel-Fattah, Mater. Sci. Eng., C, 2013, 33, 2050–2058.
42. J. T. Woodward, A. Ulman and D. K. Schwartz, Langmuir, 1996, 12, 3626–3629.
43. J. Woodward and D. Schwartz, J. Am. Chem. Soc., 1996, 118, 10944.

---