Inhibition Effect of 3D Nanostructures on the Corrosion Resistance of 1-Dodecanethiol Self-Assembled Monolayers on Copper in Nacl Solution

A novel and simple method to improve the corrosion resistance of copper by constructing a 3D 1-dodecanethiol self-assembled monolayers (SAMs) in 3.5% NaCl solution is reported in this study. Several drops of 1% H3PO4 solution are thinly and uniformly distributed on copper surface to form a 3D nanostructure constituted by Cu3(PO4)2 nanoflowers. The anticorrosion properties of 1-dodecanethiol SAMs on copper surface and on copper surface treated with H3PO4 solution were evaluated. Results demonstrated that 1-dodecanethiol SAMs on bare copper surface exhibit good protection capacity, whereas a copper surface pretreated with H3PO4 solution can substantially enhance the corrosion resistance of 1-dodecanethiol SAMs.


Introduction
Copper is widely used in microelectronic packaging due to its advantages, such as high electrical and thermal conductivities, low cost, and ease of manufacture [1].
Nevertheless, copper readily undergoes corrosion in practice.
These compact layers are constituted by highly ordered molecules and formed spontaneously by chemisorption on metal surface. Self-assembly has been recognized as a prospective technology for creating functional materials owing to the extended and two-dimensional molecule layers that can provide excellent corrosion resistance and surface superhydrophobicity [6][7][8]. Compared with the traditional corrosion inhibition methods, SAMs exhibit the advantages of high coverage, few defects, and high inhibition efficiency [9,10]. Yamamoto et al. [11] reported that copper with alkanethiol self-assembled layers obtained excellent anticorrosion abilities.
Alkanethiols were chemisorbed on the copper surface by covalent linking between Cu and S atoms, forming densely-packed, hydrophobic monolayers on the surface. Zhang et al. [12] studied the inhibition effect of Schiff base SAMs on copper. The maximum inhibition efficiency reached 93.9% for CO2-saturated simulative oilfield water after a 3 h self-assembly. Zhang et al. [13] revealed that ammonium pyrrolidine dithiocarbamate SAMs was a mixed-type inhibitor for copper in 3% NaCl solution and sulfur atoms acted as the active adsorption sites during the self-assembly.
Reportedly, the protective properties of SAMs are closely related to thickness and chain length. Laibinis et al. [8] firstly reported that thicker SAMs retarded the oxidation of copper more obviously than thinner SAMs, and an increase in SAM thickness by about 6 Å led to a corresponding decrease in the rate of oxidation by around 60%. Furthermore, Itoh et al. [14] demonstrated that further chemical modification of an 11-mercapto-1-undecanol SAMs with alkyltrichlorosilanes improved the protective capability of monolayers against corrosion in aqueous and atmospheric environments. The anodic process of corrosion was inhibited by network structures owing to two-dimensional polymerization with lateral siloxane linkage between molecules absorbed on copper.
Wang et al. [17] reported a novel method for fabricating an effective inhibition film on copper. Phenylthiourea was absorbed in a copper surface, 1-dodecanethiol was used for subsequent modification, and alternating-current voltage was applied on copper covered with the mixed film for further modification.
Structures and compositions of the metal surface markedly affect the protective property of SAMs. Inhomogeneity and instability of the metal substrate typically result in disordered adsorption layers and poor anticorrosion capability. Rohwerder et al. [18] reported that clean metal surface and metal surface covered with stable oxides were more beneficial for self-assembly than metal surface covered with unstable oxides.
To date, few papers have reported on the effect of surface composition and structure on SAMs for the corrosion inhibition of copper. In this paper, the preliminary goal is the fabrication of stable nanostructures on the copper surface.
According to previous research, several methods of constructing nanostructures on the copper surface were established. For example, Wang et al. [19] obtained desirable 3D nanostructures on copper surface by immersing copper in H3PO4 solution. Structure and chemical composition were tunable by simply changing the concentration of H3PO4 solution and immersion time. He et al. [20] reported the Cu3(PO4)2 nanoflowers were formed through the interfacial reaction between copper foil and phosphate-buffered saline and found that the formation of nanoflowers was related to the concentration of dissolved oxygen, chloride ions, and phosphate ions.
Herein, we reported a simple method for improving the corrosion resistance of SAMs on copper via a pretreatment method using H3PO4 solution. 1-Dodecanethiol was selected for the formation of SAMs. The copper surface was characterized using scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), and X-ray diffraction (XRD). The corrosion resistance of 1-dodecanethiol SAMs on H3PO4-treated copper surface was studied by using potentiodynamic polarization curves and electrochemical impedance spectroscopy (EIS).

Materials and solutions
Working electrodes were prepared from a copper sheet of purity 99.9%. For electrochemical studies, copper specimens were embedded in epoxy resin, with a surface area of 1 cm 2 exposed to the electrolyte. The surface of the samples were initially ground with 400-grit emery paper and continued with 800-and 1200-grit emery papers successively. Then the samples were washed with distilled water, degreased with ethanol and acetone, and finally dried with a flow of nitrogen gas.
1-Dodecanethiol (C12H25SH) from Aladdin with ≥ 98% purity was dissolved in absolute ethanol (AR grade) to a concentration of 80 g/L. An aqueous solution of 3.5% NaCl solution was prepared by dissolving NaCl (AR grade) in double-distilled water.
H3PO4 (AR grade, ≥ 85%) concentrated solution was diluted to a concentration of 1%, and the pH value of the solution was adjusted to 2.5 by using NaOH solution.

Fabricating 3D nanostructures on copper surface
Several drops of 1% H3PO4 solution were uniformly spread on the copper surface to form an ultrathin liquid membrane. Fresh H3PO4 solution was added to maintain the liquid membrane during the process. After 1 h, the sample was immersed in distilled water for 2 min to terminate the reaction. The temperature was controlled at 25 o C during preparation.

Formation of 1-dodecanethiol SAMs
The samples (copper and copper treated with H3PO4 solution) were dried in a vacuum dryer for 12 h to thoroughly remove water on the surface. The SAMs were formed by immersing the samples in 1-dodecanethiol ethanol solution for 3 h. Then the samples were washed with flowing distilled water and dried under nitrogen gas flow.

Characterization
Surface morphologies were obtained via scanning electron microscopy (SEM, Phillips Quanta 200) coupled with energy dispersive X-ray spectroscopy (EDS). XPS spectra were measured using a commercial VG Multilab 2000 system. An Al Kα radiation source (1486.6 eV) was equipped for the spectrum measurements, and the electron energy resolution was 0.45 eV. Spectral decomposition was performed using background subtraction and a least-squares fitting program. XRD measurement was conducted on an Empyrean X-ray diffractometer from PANalytical (a Cu Kα irradiation source, λ=1.5418 Å). The FTIR spectrum of H3PO4-treated copper surface with 1-dodecanethiol SAMs was recorded using a Bruker VERTEX 70 Fourier transform infrared spectrophotometer within a range of 4000 cm −1 to 400 cm −1 . The infrared spectra of pure 1-dodecanethiol and copper surface treated with H3PO4 solution were measured for comparison.
Contact angle measurements were carried out for the wettability evaluations of copper surface. A small water droplet was placed on the surface by using the microsyringe and the volume was controlled at 2 µL. The contact angle was determined by averaging the values obtained from three positions of each sample.

Electrochemical measurements
Electrochemical measurements were conducted using an IM6e electrochemical workstation in a conventional three-electrode cell with 3.5% NaCl solution.
Potentiodynamic polarization curves were obtained at a sweep rate of 0.5 mV/s in the potential range of −250 mV to +800 mV versus the open circuit potential. EIS measurements were conducted at the open circuit potential with a 10-mV amplitude perturbation at frequencies from 100 kHz to 10 mHz, with 10 points per decade.
Impedance data were fitted to the appropriate equivalent circuits by using Zview software. All potential values in this paper refer to the saturated calomel electrode (SCE).  by the strong Cu 2p3/2 peak corresponding to Cu(I) or Cu(0) [22,23]. The XPS spectra of O 1s core level that deconvoluted into two components were located at 531.0 and 532.0 eV, which correspond to copper oxides and phosphate, respectively [24,25].

SEM surface morphologies
The peak at 133.6 eV in P 2p spectrum is attributed to P 2p1/2 of Cu3(PO4)2 [23]. The binding energy of S peak for 1-dodecanethiol absorbed on copper is 162.2 eV, indicating that 1-dodecanethiol chemisorbed on the surface by thiolate formation, whereas the peak at 163.3 eV can be attributed to the 1-dodecanethiol molecule [26].   Table 1.
Compared with the chemical composition of bare copper surface with 1-dodecanethiol SAMs, the contents of elements C and S evidently increased for the H3PO4-treated copper surface with 1-dodecanethiol SAMs. The results indicated that the copper surface absorbed additional 1-dodecanethiol molecules when pretreated using H3PO4 solution.  respectively [29]. The bands at 1464 and 722 cm −1 correspond to the bending vibration of S-CH2 and stretching vibration of S-C, respectively. Although slight shifts of these wave numbers occurred due to the disordered conformation of alkyl chains were observed, the result was in good agreement with those of C5-C21 alkanethiol SAMs absorbed on gold [30,31]. The bands in Fig. 4c    Cu → Cu + + e -(1) The typically anodic polarization curves for bare copper in neutral NaCl solution can be illuminated by dividing the curves into three parts according to the potential region [33]. Section I: From the Tafel region to the maximum current density. Section II: a region of the current density decreased to the minimum. Section III: a region of increasing in current density to the limited value. The increase in current density at Section I was due to the oxidation of copper to the cuprous ion. With the increase in potential, the decrease in current density to the minimum resulted from the formation of CuCl film. Then CuCl was transformed to soluble CuCl2 at higher potentials resulting in a limiting current density. Above the limiting-current region, the increase in current density might be caused by the formation of Cu(II) species. This phenomenon was also observed in the anodic behaviors of both electrodes modified with 1-dodecanethiol SAMs. Moreover, the cathodic reaction can be related to the reduction of hydrogen ion or dissolved oxygen. Given that the equilibrium potential for hydrogen evolution in a solution at pH 7 is −662 mV versus SCE, the occurrence of hydrogen ion reduction is impossible. Correspondingly, the reduction potential of oxygen is 568 mV versus SCE, indicating that the reduction of oxygen is thermodynamically possible [34]. A small hump appeared on the cathodic curves for bare copper electrode and copper electrode with 1-dodecanethiol SAMs at a potential around −300 mV versus SCE, which may be caused by the reduction of CuCl and/or Cu2O [35][36][37].
As seen from this image, the corrosion rate of bare copper electrode with 1-dodecanethiol SAMs was effectively decreased. By contrast, the corrosion rate of H3PO4-treated electrode with 1-dodecanethiol SAMs unexpectedly decreased by one order of magnitude. In addition, the corrosion potential of the H3PO4-treated electrode with 1-dodecanethiol SAMs notably shifted to a positive region. The corresponding electrochemical results, such as corrosion potential (Ecorr), Tafel slope (ba, bc), and corrosion current density (icorr), are summarized in Table 2. Inhibition efficiency (IE%) were calculated using the following equation: where and ′ are the corrosion current densities of bare copper electrode and electrodes with 1-dodecanethiol SAMs in 3.5% NaCl solution, respectively. The inhibition efficiency increased from 73.1% to 97.2% when the electrode was processed with H3PO4 solution, which indicated that H3PO4 treatment of the copper surface prior to self-assembly is favorable for corrosion protection.      Rs is the solution resistance, Rf is the resistance of 1-dodecanethiol SAMs on the copper electrode surface, Rct represents the charge-transfer resistance, and CPE is the constant phase element. Given the lack of pure capacitance in the non-ideal electrochemical behavior of the surface, a CPE was used to substitute the pure capacitance. In general, a CPE is applicable to the case of surface inhomogeneity (e.g.electrode roughness and adsorbate) [40]. The impedance of CPE is defined as follows: where Y is the magnitude of CPE, ω is the angular frequency, j is an imaginary number, and n is the exponential term represents the degree of the surface inhomogeneity. The corresponding values of these impedance parameters are listed in Table 3. As shown in Table 3, Rct of bare copper increases with immersion time. This trend was relative to the formation of limited protective Cu2O and CuO films [34].

Corrosion behaviors of copper treated with H3PO4 solution
Figs. 9 and 10 show the potentiodynamic polarization curves and EIS of copper electrode with or without H3PO4 solution treatment, respectively. Electrochemical measurements were performed after the system was stabilized in NaCl solution for 30 min. The corresponding potentiodynamic polarization parameters are listed in Table 4.  Electrochemical impedance parameters were obtained using the equivalent circuit mode in Fig. 8a and are summarized in Table 5. The results showed that the bare electrode showed superior corrosion resistance than H3PO4-treated electrode.
This finding suggests that the 3D nanoflowers on the surface did not improve the corrosion resistance of copper without 1-dodecanethiol SAM modification.

Discussion
The pretreatment of copper surface with H3PO4 solution is necessary to obtain excellent protective properties. In our previous unsuccessful trials, the H3PO4 solution was placed on the copper surface without spreading, and the nanoflowers were found to grow on the margin of the liquid membrane on the copper surface, whereas nothing was found in the center except certain trails of corrosion. According to these results and those of previous reports [20,34], the formation of the Cu3(PO4)2 nanoflowers is closely linked to the concentration of dissolved oxygen and PO4 3− ions. The formation of Cu3(PO4)2 nanoflowers is illustrated in Fig. 11, and the corresponding equations are as follows: Cu−2e − → Cu 2+ (anodic reaction) The released Cu 2+ ions react with PO4 3− ions immediately, generating a solubilized layer of phosphate complex intermediate. The free phosphoric acid then corrodes the intermediate, resulting in selective crystallization into nanoparticles, which function as nuclei in the subsequent crystallization [19]. As the reactions progress, additional nanosheets are generated and aggregated into a flower-like sphere. Eventually, the surface is covered by the interconnected flower-like nanosheets. During the fabrication of nanoflowers, a paper-thin liquid membrane must be spread uniformly on the surface to ensure sufficient oxygen. Meanwhile, fresh H3PO4 solution is needed to maintain the liquid membrane and supplement adequate PO4 3− ions. In the EIS measurements, for the electrodes modified with 1-dodecanethiol SAMs, the values of impedance gradually decreased with time, indicating that 1-dodecanethiol SAMs started to deteriorate with the attack of dissolved oxygen or chloride ions. According to a previous report [41], 1-dodecanethiol SAMs are sensitive to air exposure, which causes an increase in the density of defects and a decrease in film thickness. The mechanism for deterioration of 1-dodecanethiol SAMs can be ascribed to the oxidation of thiolates to less-adherent sulfonates, leading to the roughening of the copper surface. The rougher surface may distort the structure of the hydrocarbon lattice and increase the permeability of the SAMs [8,42]. Compared with the bare copper surface, copper surface treated with H3PO4 solution exhibited a large 3D network structure that adsorbed and accommodated additional 1-dodecanethiol molecules. Correspondingly, the 2D 1-dodecanethiol SAMs on the bare copper was turned into 3D SAMs. The interconnected nanosheets modified with 1-dodecanethiol monolayers constructed a 3D hydrophobic barrier to separate the substrate from the aggressive dissolved oxygen and chloride ions. Meanwhile, owing to the thick protective layer and the release of excess 1-dodecanethiol molecules in the cavities, the film exhibited increased durability in the NaCl solution. The schematic is shown in Fig. 12.

Conclusion
The corrosion resistance of 1-dodecanethiol SAMs on bare copper surface and copper surface treated with H3PO4 solution was studied in 3.5% NaCl solution.
1-Dodecanethiol SAMs on both electrodes provide effective protection at the initial time but gradually deteriorate in the subsequent 48 h. Notably, the corrosion rate values of H3PO4-treated copper with 1-dodecanethiol SAMs were one order of magnitude smaller than those of bare copper with 1-dodecanethiol SAMs. These results demonstrate that the 3D network constituted of Cu3(PO4)2 nanoflower considerably improves the corrosion resistance of copper in NaCl solution when modified with 1-dodecanethiol SAMs.

Author Contributions
Z.C conceived the work; S.H performed the experiments and wrote this paper; and X.G made valuable comments on this manuscript.