Corrosion Behavior of Electrodeposited
Nanocrystalline Ni in Aqueous Environments
S. Wang*, J.K. Lewis*, P.R. Roberge* and U. Erb**
*-Department of Chemistry and Chemical Engineering, Royal
Military College of Canada, Kingston, Canada, K7K 5L0.
**-Departmentof Materials and Metallurgical Engineering Queen's University,
Kingston, Canada, K7L 3N6
email: wang_s@rmc.ca
The unique properties observed in recently developed nanocrystalline materials can be useful for applications such as advanced corrosion and wear resistant coatings, soft magnetic materials for magnetic recording, and elecrocatalysts for hydrogen evolution and oxidation reactions. In this paper, the results on the corrosion behavior of electrodeposited nanocrystalline nickel in aqueous environment will be presented. These studies showed that the high corrosion resistance of conventional polycrystalline Ni is conserved in its electrodeposited bulk nanocrystalline counterpart. The passive current density of nanocrystalline Ni was found to be higher in acidic, alkaline and neutral solutions than polycrystalline nickel. This was attributed to the high surface fraction of grain boundaries and triple junctions in nanocrystalline nickel. There was no difference in corrosion current density at high anodic overpotential but conventional polycrystalline Ni was found to be more susceptible to localized corrosion.
A crystallographic model will be considered to determine the intercrystalline defect density at the corroding surface. Based on this model a reduction in grain size from 1 µm to 5 nm corresponds to a change of intercrystalline volume fraction from 0.3% to 50%. The high degree of intercrystallinity provides an increased number of preferential attack sites and a higher average current density, but with reduced current densities at individual attack sites. The resulting decrease in microscopic heterogeneity is shown to inhibit localized corrosion of nanocrystalline materials.
Recently developed nanocrystalline materials showed several improved properties compared with their conventional polycrystalline counterparts. Such enhanced properties can be useful for applications such as corrosion and wear resistant coatings, soft magnetic materials for magnetic recording, and elecrocatalysts for hydrogen evolution and oxidation reactions. For example, Table 1 summarizes the main findings of studies on nanocrystalline Ni produced by electrodeposition.
Nanocrystalline materials may be produced in a number of ways, including gas condensation, spray conversion, ball milling and electrochemical deposition. However, three basic synthesis routes have been applied to produce nanostructured materials for corrosion studies. Most of the early work (1975-90) dealt with materials produced by crystallization of amorphous precursor materials. The second group of materials which has been extensively studied since 1990 are nanostructures produced by electrodeposition. Finally, materials produced by sputter deposition have been studied since about 1992. In this paper, results of polarization studies of electrodeposited nanocrystalline Ni in acidic, alkaline and neutral solutions will be presented. Using a crystallographic model, an attempt to correlated the intercrystalline component fraction of nanocrystalline materials to their localized corrosion behaviors will be discussed.
| 5 times harder. | ||
| 170 times increase. | ||
| cut friction in half. | ||
| reduce or stop localized corrosion. | ||
| 3 to 10 times stronger. | ||
| lower coercivity, resistivity 3 times increased, Ms reduced by 5%. | ||
| higher hydrogen diffusion. | ||
| improved electrocatalytic activities for hydrogen evolution and hydrogen oxidation reactions. |
2-1. Acidic solution
In previous studies, potentiodynamic and potentiostatic polarizations in deaerated 2N H2SO4 were conducted on bulk (2cm square coupons, 0.2mm thick) nanocrystalline pure Ni at grain sizes of 32, 50 and 500nm and compared with polycrystalline pure Ni (grain size of 100mm) [6,7]. Fig. 1 shows the potentiodynamic anodic polarization curves of these specimens in the above mentioned solution.
The nanocrystalline specimens exhibited the same active-passive-transpassive behaviour typical of conventional Ni. However, differences are evident in the passive current density and the open circuit potential. The nanocrystalline specimens show a higher current density in the passive region resulting in higher corrosion rates. These higher current densities were attributed to the higher grain boundary and triple junction content in the nanocrystalline specimens. Based on theoretical considerations, it has been shown that the intercrystalline (grain boundary and triple junction) content of a material increases from a value of 0.3% at a grain size of 1µm to more than 50% at grain sizes less than 5nm [17]. Therefore, this increased current density in the passive region can be considered the result of the substantial increase in the intercrystalline region which provides sites for electrochemical activity. However, this difference in current density diminishes at higher potentials (1100mV SCE) at which the overall dissolution rate overwhelms the structure-controlled dissolution rate observed at lower potentials.
Another notable difference in the potentiodynamic response of nanocrystalline and polycrystalline specimens is the open circuit potential. The positive shift of the open circuit potential for the nanocrystalline specimens is thought to be the result of the catalysis of the hydrogen evolution reaction. Nanocrystalline metals in general have been shown to have unique catalytic characteristics which are the result of the high defect content in form of interfaces and triple junctions.
Fig. 2 shows scanning electron micrographs of Ni with (a) 32nm and (b) 100µm grain size held potentiostatically at 1200mV (SCE) in 2N H2SO4 for 2,000 sec.
Both specimens exhibit extensive corrosion but the nanocrystalline Ni is more uniformly corroded while the specimen with 100µm grain size shows extensive localized attack along the grain boundaries and triple junctions. X-ray photoelectron spectroscopy of specimens polarized in the passive region showed that the passive film formed on the nanostructured material is more defective than that formed on the polycrystalline surface while the thickness of the passive layer was the same on both specimens [18]. This higher defective film on the nanocrystalline specimen allows for a more uniform breakdown of the passive film which in turn leads to a more uniform corrosion. In contrast, as has been previously shown [19], in coarse grained Ni, the breakdown of the passive film occurs first at the grain boundaries and triple junctions rather than the crystal surface, leading to preferential attack at these defects.
2-2. Alkaline solution
Nanocrystalline Ni with grain sizes of 16 and 32nm were recently tested in alkaline solution of deaerated 30 wt% KOH (pH=14.8) at 24°C and compared with conventional polycrystalline Ni (grain size 50µm) [8]. Fig. 3 shows the potentiodynamic polarization behaviour of nanocrystalline and polycrystalline samples. The scans (scan rate = 5mV/sec) were conducted to investigate the corrosion behavior of Ni samples covered with a natural oxide film.
As for acidic media, the nanocrystalline materials exhibited a higher passive current density than their polycrystalline counterpart. However, they all converge to the same value at a potential of about 200 mV (SCE). The higher passive current density of nanocrystalline Ni was explained as follows: in the low overpotential region, corrosion attack is initiated more easily at defects sites, i.e. grain boundaries and triple junctions in the nanocrystalline metals. In this case the number of defects sites controls the dissolution mechanism. With an increase in anodic polarization, the corrosion becomes less structure dependent as observed for the case of acidic solutions.
Scanning electron microscopy observation of corroded specimens
(see Fig. 4) showed uniform attack on nanostructured Ni and localized
attack on polycrystalline samples similar to what was observed
on samples corroded in H2SO4.
3-1. Experimental
Three different working electrodes were used. The first was industrially produced plate nickel (Inco 270) which was thermally treated at 600°C for 24 hours. The grain size for this polycrystalline nickel sample was determined by optical microscopy of the etched surface to be of the order of 500 µm. The other two working electrodes used were obtained from electrodeposited bulk Ni [20] with grain sizes of 17 and 30 nm, respectively, as determinedby the X-ray line broadening method.
The electrode surfaces were polished with progressively finer grades of abrasive, starting with emery paper and finishing with alumina powder with a grit of 0.05 µm. The surfaces were rinsed and thoroughly degreased with acetone before introduction into the electrolyte solution.
The electrolyte solution contained 3% by mass sodium chloride in distilled water. This solution had a measured pH of ~6. A single compartment, cylindrical acrylic cell with an volume of 1000 ml was used. The cell temperature was not controlled, but the ambient room temperature was reasonably constant at 22°C.
All current-potential experiments were performed using an EG&G Princeton Applied Research model 173 potentiostat with a model 276 interface which was linked to an IBM compatible personal computer. The electrochemical data acquisition and analysis were conducted and automated with the software package 352/252, Corrosion Analysis from Princeton Applied Research (see figure 5).
Surface morphology of the corroded samples was studied using photographs
taken on a JEOL JSM-T100 scanning electron microscope.
3-2. Results and discussion
Figure 6 presents the potentiodynamic polarization behaviour of the nano and polycrystalline Ni electrodes. The scans ran from -0.8V to 0.8V with respect to SCE, at a scan rate of 5 mV/s. The Ecorr of the electrodes was in all three cases very close to their respective, experimentally determined Eocv. At a potential of around 0.6 V (SCE), the passivation is suddenly deactivated. According to the Pourbaix diagram for the system, this may be due to the oxidation of the nickel oxide layer to a higher valence.
As for acidic and basic solutions, the current density in the passive range is higher for nanocrystalline Ni than for polycrystalline material. However, as for the other solutions the corrosion was found to be much more uniform and not localized as observed in polycrystalline material. Scanning electron microscopy (see Figure 7) showed that the corroded polycrystalline samples displayed a few large, deep pits, while the corroded nanocrystalline samples showed a multitude of smaller pits. The surface morphology was examined following anodic polarization from -800 mV (SCE) to 800 mV (SCE) at a rate of 0.5 mV/s.
This result agrees with the contention that corrosion is concentrated at triple junctions and grain boundaries. The polycrystalline sample, having very large grains, has proportionally very few atoms located in triple junctions or grain boundaries. Corrosion, when localized at these sparse locations tends to result in large, deep pits. As the grain size decreases, the number of triple junctions and grain boundaries increases. Hence, the number of pits increases while their size tends to decrease. The same amount of nickel dissolves into solution, but the corrosion is more evenly distributed.
The high localized corrosion resistance of nanocrystalline materials can be further explained by the examination of the microstructure and grain boundary characteristics in the electrodeposited nanocrystalline materials. By considering a regular fourteen faced tatrakaidecahedron as the grain shape, Palumbo et al [17] calculated intercrystal volume fraction (Vic), grain boundary volume fraction (Vgb) and triple junction volume fraction (Vtj) as a function as grain size (see figure 8). Considering a grain boundary thickness of 1 nm, it is shown that a reduction of grain size from 1 µm to 5 nm corresponds to an increase in Vic from ~0.3% to ~50%.
Grain boundaries and other defects in materials have been identified to likely provide preferential attack sites when the materials expose to a corrosive environment. In this case, the nature of the high intercrystalline surface fraction or defect density in nanocrystalline materials provides a more beneficial cathode/anode surface ratio against localized corrosion, although the overall corrosion current or passive current of nanocrystalline materials may be higher than for their polycrystalline counterparts. A geometric calculation showed that the value of matrix to grain boundary defect surface ratio is less than 10 for nanocrystaline materials, while with a grain size of ~50µm, the value reaches thousands in polycrystalline materials [21]. If the matrix to grain boundary defect surface ratio is directly related to the cathode/anode surface ratio, the reduced penetration current density at each individual defect site in nanocrystalline materials will result in localized corrosion to vanish. In addition, the nanocrystalline materials fabricated by electrodeposition have been reported to have relatively narrow grain size distribution and, thus, evenly distributed defect sites on the corroding surface [22, 23]. These characteristics observed in nanocrystalline materials give a better localized corrosion resistance resulted from the evenly distributed corrosion current. Results of the polarization studies presented previously are in general agreement with this analysis. It was found that nanocrystalline metals corroded more generally on all exposed surface, while polycrystalline metals suffered severe localized corrosion.
Although the correlation between the surface microstructure and the corrosion behavior is not directly obvious when a passive film is present on a corroding surface, the strong effects of the surface defects, e.g. grain boundary defects, on the film properties still exist. Using the X-ray photoelectron spectroscopy technique, Rofagha et al [18] examined the passive films of pure polycrystalline and nanocrystalline nickel, after anodic polarization in the passive region in a sulfuric acid solution. No difference in the film thickness was found, but the passive film on the nanocrystalline nickel had a higher defect density and porosity compared with the passive film on the polycrystalline Ni. These were explained as the results of high contents of grain boundaries and triple junctions under the passive film covered surface of nanocrystalline Ni. These uniformly distributed defects contribute to the higher passive current density observed in nanocrystalline samples.